diff --git a/.gitattributes b/.gitattributes
index 9e61fa067cc885befa2a18a893ef8fae961d89f4..b4eec751814281bcf34c3171f4196f745470dc74 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -4721,3 +4721,12 @@ Qwen2-7B-Instruct_int4_flare-en-fpb_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_
 Qwen2-7B-Instruct_int4_flare-en-fpb_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-1.0-num-3100-sd-10000/checkpoint-762/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 Qwen2-7B-Instruct_int4_flare-en-fpb_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-1.0-num-3100-sd-10000/checkpoint-952/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 Qwen2-7B-Instruct_int4_flare-en-fpb_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-1.0-num-3100-sd-10000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/tokenizer.json filter=lfs diff=lfs merge=lfs -text
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/README.md b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..835e31ab8469ee39ddc8b2b6b2143a8c66dad510
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/README.md
@@ -0,0 +1,202 @@
+---
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.13.1
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/adapter_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..406c5a08dc4a2a33b52c62a482f98c217c417215
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/adapter_config.json
@@ -0,0 +1,29 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/adapter_model.safetensors b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..894306c21591f0f9fc1ad6edef9d7f95864e6e88
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6e6524f0f0d768c7962609d66799cf291b435521c1246c8b6c63ea99e47dce1
+size 109069176
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/README.md b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..835e31ab8469ee39ddc8b2b6b2143a8c66dad510
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/README.md
@@ -0,0 +1,202 @@
+---
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.13.1
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/adapter_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..406c5a08dc4a2a33b52c62a482f98c217c417215
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/adapter_config.json
@@ -0,0 +1,29 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/adapter_model.safetensors b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..7f7fb8974bbe9e3ded294ba9f6549ebe451d4e28
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f0dffcebfe38a83d54dd012bc62f82ee7d2b7b9157304b819748919869de6f4
+size 109069176
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/optimizer.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..422b84716e6e5590c9161a0d19a972ea120a6407
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6508de495fd367d7e871fa8e4fc75836e030f326d435064d21860c3a88186250
+size 55532666
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/rng_state.pth b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..873378f7f032804548ed4fb5cb640548d2d63d40
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3200755ef5e97483bb817f7cf48836d884fb26788ef30ca0156da782ef1dc412
+size 14244
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/scheduler.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5e3d28beeab0e8c386cfeb3f8e978d2b211dbdf3
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23ff4515f17a8f43b8ac627a744199be2a7a561be94af3e0dbcf8fc48ffce02f
+size 1064
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/special_tokens_map.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/special_tokens_map.json
@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/tokenizer.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..86a33946b0c77216d2cce91bb28c8fada4a5e80b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/tokenizer_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..061e40d9db3253624f86e8e364c15ef546527c9d
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/tokenizer_config.json
@@ -0,0 +1,2063 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/trainer_state.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..bc5c9e545f9a414ef2d71332375f729465530003
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/trainer_state.json
@@ -0,0 +1,8708 @@
+{
+  "best_metric": 1.0871200561523438,
+  "best_model_checkpoint": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185",
+  "epoch": 2.0,
+  "eval_steps": 10,
+  "global_step": 12371,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016166841807452913,
+      "grad_norm": 0.9894065856933594,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032333683614905826,
+      "grad_norm": 1.7810699939727783,
+      "learning_rate": 0.0002,
+      "loss": 1.1528,
+      "step": 20
+    },
+    {
+      "epoch": 0.004850052542235874,
+      "grad_norm": 0.5969577431678772,
+      "learning_rate": 0.0002,
+      "loss": 0.9767,
+      "step": 30
+    },
+    {
+      "epoch": 0.006466736722981165,
+      "grad_norm": 0.6354120969772339,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 40
+    },
+    {
+      "epoch": 0.008083420903726457,
+      "grad_norm": 0.5604607462882996,
+      "learning_rate": 0.0002,
+      "loss": 0.8643,
+      "step": 50
+    },
+    {
+      "epoch": 0.009700105084471748,
+      "grad_norm": 0.4676193594932556,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 60
+    },
+    {
+      "epoch": 0.01131678926521704,
+      "grad_norm": 0.6099211573600769,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 70
+    },
+    {
+      "epoch": 0.01293347344596233,
+      "grad_norm": 0.48639994859695435,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 80
+    },
+    {
+      "epoch": 0.014550157626707623,
+      "grad_norm": 0.4904264509677887,
+      "learning_rate": 0.0002,
+      "loss": 0.8704,
+      "step": 90
+    },
+    {
+      "epoch": 0.016166841807452915,
+      "grad_norm": 2.8334362506866455,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 100
+    },
+    {
+      "epoch": 0.017783525988198205,
+      "grad_norm": 0.43221670389175415,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 110
+    },
+    {
+      "epoch": 0.019400210168943496,
+      "grad_norm": 0.42244166135787964,
+      "learning_rate": 0.0002,
+      "loss": 0.8412,
+      "step": 120
+    },
+    {
+      "epoch": 0.02101689434968879,
+      "grad_norm": 0.45363298058509827,
+      "learning_rate": 0.0002,
+      "loss": 0.8467,
+      "step": 130
+    },
+    {
+      "epoch": 0.02263357853043408,
+      "grad_norm": 0.44816508889198303,
+      "learning_rate": 0.0002,
+      "loss": 0.8641,
+      "step": 140
+    },
+    {
+      "epoch": 0.02425026271117937,
+      "grad_norm": 0.43308213353157043,
+      "learning_rate": 0.0002,
+      "loss": 0.8496,
+      "step": 150
+    },
+    {
+      "epoch": 0.02586694689192466,
+      "grad_norm": 0.4084763526916504,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 160
+    },
+    {
+      "epoch": 0.027483631072669955,
+      "grad_norm": 0.5363703966140747,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 170
+    },
+    {
+      "epoch": 0.029100315253415245,
+      "grad_norm": 0.4619699716567993,
+      "learning_rate": 0.0002,
+      "loss": 0.8558,
+      "step": 180
+    },
+    {
+      "epoch": 0.030716999434160536,
+      "grad_norm": 0.49069908261299133,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 190
+    },
+    {
+      "epoch": 0.03233368361490583,
+      "grad_norm": 0.4645835757255554,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 200
+    },
+    {
+      "epoch": 0.03395036779565112,
+      "grad_norm": 1.2411243915557861,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 210
+    },
+    {
+      "epoch": 0.03556705197639641,
+      "grad_norm": 0.5211851596832275,
+      "learning_rate": 0.0002,
+      "loss": 0.8245,
+      "step": 220
+    },
+    {
+      "epoch": 0.037183736157141704,
+      "grad_norm": 0.5253691673278809,
+      "learning_rate": 0.0002,
+      "loss": 0.8194,
+      "step": 230
+    },
+    {
+      "epoch": 0.03880042033788699,
+      "grad_norm": 0.4567478895187378,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 240
+    },
+    {
+      "epoch": 0.040417104518632285,
+      "grad_norm": 0.5472128391265869,
+      "learning_rate": 0.0002,
+      "loss": 0.838,
+      "step": 250
+    },
+    {
+      "epoch": 0.04203378869937758,
+      "grad_norm": 0.42978546023368835,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 260
+    },
+    {
+      "epoch": 0.043650472880122866,
+      "grad_norm": 0.601734459400177,
+      "learning_rate": 0.0002,
+      "loss": 0.8334,
+      "step": 270
+    },
+    {
+      "epoch": 0.04526715706086816,
+      "grad_norm": 0.4286513328552246,
+      "learning_rate": 0.0002,
+      "loss": 0.815,
+      "step": 280
+    },
+    {
+      "epoch": 0.046883841241613454,
+      "grad_norm": 0.5230861902236938,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 290
+    },
+    {
+      "epoch": 0.04850052542235874,
+      "grad_norm": 0.6504611968994141,
+      "learning_rate": 0.0002,
+      "loss": 0.8636,
+      "step": 300
+    },
+    {
+      "epoch": 0.050117209603104035,
+      "grad_norm": 0.43485215306282043,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 310
+    },
+    {
+      "epoch": 0.05173389378384932,
+      "grad_norm": 0.4717007875442505,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 320
+    },
+    {
+      "epoch": 0.053350577964594616,
+      "grad_norm": 0.4059787690639496,
+      "learning_rate": 0.0002,
+      "loss": 0.8469,
+      "step": 330
+    },
+    {
+      "epoch": 0.05496726214533991,
+      "grad_norm": 0.4366913437843323,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 340
+    },
+    {
+      "epoch": 0.0565839463260852,
+      "grad_norm": 0.4233848452568054,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 350
+    },
+    {
+      "epoch": 0.05820063050683049,
+      "grad_norm": 0.4209108352661133,
+      "learning_rate": 0.0002,
+      "loss": 0.8456,
+      "step": 360
+    },
+    {
+      "epoch": 0.059817314687575784,
+      "grad_norm": 0.41637396812438965,
+      "learning_rate": 0.0002,
+      "loss": 0.816,
+      "step": 370
+    },
+    {
+      "epoch": 0.06143399886832107,
+      "grad_norm": 0.46235376596450806,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 380
+    },
+    {
+      "epoch": 0.06305068304906636,
+      "grad_norm": 0.4013484716415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 390
+    },
+    {
+      "epoch": 0.06466736722981166,
+      "grad_norm": 0.47443896532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 400
+    },
+    {
+      "epoch": 0.06628405141055695,
+      "grad_norm": 0.3942156434059143,
+      "learning_rate": 0.0002,
+      "loss": 0.8666,
+      "step": 410
+    },
+    {
+      "epoch": 0.06790073559130223,
+      "grad_norm": 0.4965320825576782,
+      "learning_rate": 0.0002,
+      "loss": 0.8402,
+      "step": 420
+    },
+    {
+      "epoch": 0.06951741977204753,
+      "grad_norm": 0.4304835796356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8317,
+      "step": 430
+    },
+    {
+      "epoch": 0.07113410395279282,
+      "grad_norm": 0.511726975440979,
+      "learning_rate": 0.0002,
+      "loss": 0.8528,
+      "step": 440
+    },
+    {
+      "epoch": 0.07275078813353811,
+      "grad_norm": 0.4040689170360565,
+      "learning_rate": 0.0002,
+      "loss": 0.8675,
+      "step": 450
+    },
+    {
+      "epoch": 0.07436747231428341,
+      "grad_norm": 0.5402171015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 460
+    },
+    {
+      "epoch": 0.0759841564950287,
+      "grad_norm": 0.4174517095088959,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 470
+    },
+    {
+      "epoch": 0.07760084067577398,
+      "grad_norm": 0.4306182265281677,
+      "learning_rate": 0.0002,
+      "loss": 0.7605,
+      "step": 480
+    },
+    {
+      "epoch": 0.07921752485651928,
+      "grad_norm": 0.535210132598877,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 490
+    },
+    {
+      "epoch": 0.08083420903726457,
+      "grad_norm": 0.5339109897613525,
+      "learning_rate": 0.0002,
+      "loss": 0.7825,
+      "step": 500
+    },
+    {
+      "epoch": 0.08245089321800986,
+      "grad_norm": 0.45754891633987427,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 510
+    },
+    {
+      "epoch": 0.08406757739875516,
+      "grad_norm": 0.43820783495903015,
+      "learning_rate": 0.0002,
+      "loss": 0.8144,
+      "step": 520
+    },
+    {
+      "epoch": 0.08568426157950045,
+      "grad_norm": 0.4434749186038971,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 530
+    },
+    {
+      "epoch": 0.08730094576024573,
+      "grad_norm": 0.43111467361450195,
+      "learning_rate": 0.0002,
+      "loss": 0.7857,
+      "step": 540
+    },
+    {
+      "epoch": 0.08891762994099103,
+      "grad_norm": 0.4378940165042877,
+      "learning_rate": 0.0002,
+      "loss": 0.8418,
+      "step": 550
+    },
+    {
+      "epoch": 0.09053431412173632,
+      "grad_norm": 0.4772215187549591,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 560
+    },
+    {
+      "epoch": 0.09215099830248161,
+      "grad_norm": 0.6837629079818726,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 570
+    },
+    {
+      "epoch": 0.09376768248322691,
+      "grad_norm": 0.42241212725639343,
+      "learning_rate": 0.0002,
+      "loss": 0.8607,
+      "step": 580
+    },
+    {
+      "epoch": 0.0953843666639722,
+      "grad_norm": 0.5165936350822449,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 590
+    },
+    {
+      "epoch": 0.09700105084471748,
+      "grad_norm": 0.48737478256225586,
+      "learning_rate": 0.0002,
+      "loss": 0.8664,
+      "step": 600
+    },
+    {
+      "epoch": 0.09861773502546278,
+      "grad_norm": 0.47419852018356323,
+      "learning_rate": 0.0002,
+      "loss": 0.8806,
+      "step": 610
+    },
+    {
+      "epoch": 0.10023441920620807,
+      "grad_norm": 0.4975486099720001,
+      "learning_rate": 0.0002,
+      "loss": 0.8254,
+      "step": 620
+    },
+    {
+      "epoch": 0.10185110338695336,
+      "grad_norm": 0.49123844504356384,
+      "learning_rate": 0.0002,
+      "loss": 0.8548,
+      "step": 630
+    },
+    {
+      "epoch": 0.10346778756769864,
+      "grad_norm": 0.6288952827453613,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 640
+    },
+    {
+      "epoch": 0.10508447174844394,
+      "grad_norm": 0.4277345836162567,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 650
+    },
+    {
+      "epoch": 0.10670115592918923,
+      "grad_norm": 0.4021061956882477,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 660
+    },
+    {
+      "epoch": 0.10831784010993452,
+      "grad_norm": 0.3492237329483032,
+      "learning_rate": 0.0002,
+      "loss": 0.87,
+      "step": 670
+    },
+    {
+      "epoch": 0.10993452429067982,
+      "grad_norm": 0.4341012239456177,
+      "learning_rate": 0.0002,
+      "loss": 0.8698,
+      "step": 680
+    },
+    {
+      "epoch": 0.1115512084714251,
+      "grad_norm": 0.7296304106712341,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 690
+    },
+    {
+      "epoch": 0.1131678926521704,
+      "grad_norm": 0.397494912147522,
+      "learning_rate": 0.0002,
+      "loss": 0.8433,
+      "step": 700
+    },
+    {
+      "epoch": 0.1147845768329157,
+      "grad_norm": 0.396431028842926,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 710
+    },
+    {
+      "epoch": 0.11640126101366098,
+      "grad_norm": 0.48842838406562805,
+      "learning_rate": 0.0002,
+      "loss": 0.8379,
+      "step": 720
+    },
+    {
+      "epoch": 0.11801794519440627,
+      "grad_norm": 0.46322616934776306,
+      "learning_rate": 0.0002,
+      "loss": 0.8238,
+      "step": 730
+    },
+    {
+      "epoch": 0.11963462937515157,
+      "grad_norm": 0.47990912199020386,
+      "learning_rate": 0.0002,
+      "loss": 0.8041,
+      "step": 740
+    },
+    {
+      "epoch": 0.12125131355589686,
+      "grad_norm": 0.4997142255306244,
+      "learning_rate": 0.0002,
+      "loss": 0.82,
+      "step": 750
+    },
+    {
+      "epoch": 0.12286799773664214,
+      "grad_norm": 0.4040526747703552,
+      "learning_rate": 0.0002,
+      "loss": 0.7702,
+      "step": 760
+    },
+    {
+      "epoch": 0.12448468191738744,
+      "grad_norm": 0.453095942735672,
+      "learning_rate": 0.0002,
+      "loss": 0.863,
+      "step": 770
+    },
+    {
+      "epoch": 0.12610136609813272,
+      "grad_norm": 0.4636971950531006,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 780
+    },
+    {
+      "epoch": 0.12771805027887803,
+      "grad_norm": 0.4279276132583618,
+      "learning_rate": 0.0002,
+      "loss": 0.8112,
+      "step": 790
+    },
+    {
+      "epoch": 0.12933473445962332,
+      "grad_norm": 0.46212655305862427,
+      "learning_rate": 0.0002,
+      "loss": 0.8711,
+      "step": 800
+    },
+    {
+      "epoch": 0.1309514186403686,
+      "grad_norm": 0.43127650022506714,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 810
+    },
+    {
+      "epoch": 0.1325681028211139,
+      "grad_norm": 0.4201301336288452,
+      "learning_rate": 0.0002,
+      "loss": 0.8476,
+      "step": 820
+    },
+    {
+      "epoch": 0.13418478700185918,
+      "grad_norm": 0.42583167552948,
+      "learning_rate": 0.0002,
+      "loss": 0.8078,
+      "step": 830
+    },
+    {
+      "epoch": 0.13580147118260447,
+      "grad_norm": 0.4535622000694275,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 840
+    },
+    {
+      "epoch": 0.13741815536334978,
+      "grad_norm": 0.4116036891937256,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 850
+    },
+    {
+      "epoch": 0.13903483954409507,
+      "grad_norm": 0.45997580885887146,
+      "learning_rate": 0.0002,
+      "loss": 0.8466,
+      "step": 860
+    },
+    {
+      "epoch": 0.14065152372484035,
+      "grad_norm": 0.4487837255001068,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 870
+    },
+    {
+      "epoch": 0.14226820790558564,
+      "grad_norm": 0.43650057911872864,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 880
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 0.5335358381271362,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 890
+    },
+    {
+      "epoch": 0.14550157626707622,
+      "grad_norm": 0.5989000201225281,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 900
+    },
+    {
+      "epoch": 0.14711826044782153,
+      "grad_norm": 0.517179012298584,
+      "learning_rate": 0.0002,
+      "loss": 0.8385,
+      "step": 910
+    },
+    {
+      "epoch": 0.14873494462856682,
+      "grad_norm": 0.44435232877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.8255,
+      "step": 920
+    },
+    {
+      "epoch": 0.1503516288093121,
+      "grad_norm": 0.42635923624038696,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 930
+    },
+    {
+      "epoch": 0.1519683129900574,
+      "grad_norm": 0.49603334069252014,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 940
+    },
+    {
+      "epoch": 0.15358499717080268,
+      "grad_norm": 0.40639808773994446,
+      "learning_rate": 0.0002,
+      "loss": 0.8377,
+      "step": 950
+    },
+    {
+      "epoch": 0.15520168135154797,
+      "grad_norm": 0.4850759208202362,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 960
+    },
+    {
+      "epoch": 0.15681836553229328,
+      "grad_norm": 0.4427442252635956,
+      "learning_rate": 0.0002,
+      "loss": 0.846,
+      "step": 970
+    },
+    {
+      "epoch": 0.15843504971303857,
+      "grad_norm": 0.3760930001735687,
+      "learning_rate": 0.0002,
+      "loss": 0.8705,
+      "step": 980
+    },
+    {
+      "epoch": 0.16005173389378385,
+      "grad_norm": 0.4794144332408905,
+      "learning_rate": 0.0002,
+      "loss": 0.8644,
+      "step": 990
+    },
+    {
+      "epoch": 0.16166841807452914,
+      "grad_norm": 0.45828768610954285,
+      "learning_rate": 0.0002,
+      "loss": 0.8002,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16328510225527443,
+      "grad_norm": 0.6313053369522095,
+      "learning_rate": 0.0002,
+      "loss": 0.7658,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16490178643601971,
+      "grad_norm": 0.45041006803512573,
+      "learning_rate": 0.0002,
+      "loss": 0.8047,
+      "step": 1020
+    },
+    {
+      "epoch": 0.166518470616765,
+      "grad_norm": 0.441403865814209,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16813515479751032,
+      "grad_norm": 0.8171296119689941,
+      "learning_rate": 0.0002,
+      "loss": 0.8475,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1697518389782556,
+      "grad_norm": 0.7137420773506165,
+      "learning_rate": 0.0002,
+      "loss": 0.845,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1713685231590009,
+      "grad_norm": 0.5236809849739075,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17298520733974618,
+      "grad_norm": 0.5021864175796509,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17460189152049146,
+      "grad_norm": 0.47347521781921387,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17621857570123675,
+      "grad_norm": 0.4631653428077698,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17783525988198207,
+      "grad_norm": 0.49169182777404785,
+      "learning_rate": 0.0002,
+      "loss": 0.8166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17945194406272735,
+      "grad_norm": 0.5019739270210266,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18106862824347264,
+      "grad_norm": 0.5100422501564026,
+      "learning_rate": 0.0002,
+      "loss": 0.8247,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18268531242421793,
+      "grad_norm": 0.3888324499130249,
+      "learning_rate": 0.0002,
+      "loss": 0.8142,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18430199660496321,
+      "grad_norm": 0.39765217900276184,
+      "learning_rate": 0.0002,
+      "loss": 0.8533,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1859186807857085,
+      "grad_norm": 0.47190186381340027,
+      "learning_rate": 0.0002,
+      "loss": 0.8541,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18753536496645382,
+      "grad_norm": 0.4464188814163208,
+      "learning_rate": 0.0002,
+      "loss": 0.8301,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1891520491471991,
+      "grad_norm": 0.5153930187225342,
+      "learning_rate": 0.0002,
+      "loss": 0.8341,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1907687333279444,
+      "grad_norm": 0.4779708683490753,
+      "learning_rate": 0.0002,
+      "loss": 0.8033,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19238541750868968,
+      "grad_norm": 0.4834315776824951,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19400210168943496,
+      "grad_norm": 0.402357816696167,
+      "learning_rate": 0.0002,
+      "loss": 0.7721,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19561878587018025,
+      "grad_norm": 0.45899084210395813,
+      "learning_rate": 0.0002,
+      "loss": 0.7941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19723547005092557,
+      "grad_norm": 0.5106529593467712,
+      "learning_rate": 0.0002,
+      "loss": 0.8353,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19885215423167085,
+      "grad_norm": 0.45261722803115845,
+      "learning_rate": 0.0002,
+      "loss": 0.7816,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20046883841241614,
+      "grad_norm": 0.4647127091884613,
+      "learning_rate": 0.0002,
+      "loss": 0.8068,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20208552259316143,
+      "grad_norm": 0.4849368929862976,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2037022067739067,
+      "grad_norm": 0.4518061578273773,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 1260
+    },
+    {
+      "epoch": 0.205318890954652,
+      "grad_norm": 0.49535325169563293,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2069355751353973,
+      "grad_norm": 0.4835205376148224,
+      "learning_rate": 0.0002,
+      "loss": 0.8348,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2085522593161426,
+      "grad_norm": 0.45308539271354675,
+      "learning_rate": 0.0002,
+      "loss": 0.8428,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2101689434968879,
+      "grad_norm": 0.5369905233383179,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21178562767763318,
+      "grad_norm": 0.5031622052192688,
+      "learning_rate": 0.0002,
+      "loss": 0.8676,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21340231185837846,
+      "grad_norm": 0.48010334372520447,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21501899603912375,
+      "grad_norm": 0.4905701279640198,
+      "learning_rate": 0.0002,
+      "loss": 0.806,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21663568021986904,
+      "grad_norm": 0.43531742691993713,
+      "learning_rate": 0.0002,
+      "loss": 0.7885,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21825236440061435,
+      "grad_norm": 0.44330692291259766,
+      "learning_rate": 0.0002,
+      "loss": 0.8191,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21986904858135964,
+      "grad_norm": 0.5384416580200195,
+      "learning_rate": 0.0002,
+      "loss": 0.8205,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22148573276210493,
+      "grad_norm": 0.4181833863258362,
+      "learning_rate": 0.0002,
+      "loss": 0.7726,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2231024169428502,
+      "grad_norm": 0.523833692073822,
+      "learning_rate": 0.0002,
+      "loss": 0.8311,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.5528736710548401,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2263357853043408,
+      "grad_norm": 0.43515023589134216,
+      "learning_rate": 0.0002,
+      "loss": 0.8079,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2279524694850861,
+      "grad_norm": 0.48809877038002014,
+      "learning_rate": 0.0002,
+      "loss": 0.8403,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2295691536658314,
+      "grad_norm": 0.43591251969337463,
+      "learning_rate": 0.0002,
+      "loss": 0.8165,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23118583784657668,
+      "grad_norm": 0.44625312089920044,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23280252202732196,
+      "grad_norm": 0.4390665292739868,
+      "learning_rate": 0.0002,
+      "loss": 0.8134,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23441920620806725,
+      "grad_norm": 0.48496049642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.8465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23603589038881254,
+      "grad_norm": 0.45919957756996155,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23765257456955785,
+      "grad_norm": 0.5471845865249634,
+      "learning_rate": 0.0002,
+      "loss": 0.8659,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23926925875030314,
+      "grad_norm": 0.47269317507743835,
+      "learning_rate": 0.0002,
+      "loss": 0.8164,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24088594293104842,
+      "grad_norm": 0.4930245578289032,
+      "learning_rate": 0.0002,
+      "loss": 0.854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2425026271117937,
+      "grad_norm": 0.5605630278587341,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 1500
+    },
+    {
+      "epoch": 0.244119311292539,
+      "grad_norm": 0.4435870945453644,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24573599547328429,
+      "grad_norm": 0.4941999912261963,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24735267965402957,
+      "grad_norm": 0.5100624561309814,
+      "learning_rate": 0.0002,
+      "loss": 0.8427,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2489693638347749,
+      "grad_norm": 0.4638267457485199,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25058604801552015,
+      "grad_norm": 0.5071570873260498,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25220273219626543,
+      "grad_norm": 0.4291319251060486,
+      "learning_rate": 0.0002,
+      "loss": 0.7724,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2538194163770108,
+      "grad_norm": 0.5388049483299255,
+      "learning_rate": 0.0002,
+      "loss": 0.7984,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25543610055775606,
+      "grad_norm": 0.5083683729171753,
+      "learning_rate": 0.0002,
+      "loss": 0.8176,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25705278473850135,
+      "grad_norm": 0.4824463725090027,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25866946891924664,
+      "grad_norm": 0.41177722811698914,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2602861530999919,
+      "grad_norm": 0.5656219124794006,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2619028372807372,
+      "grad_norm": 0.41063204407691956,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2635195214614825,
+      "grad_norm": 0.4897061288356781,
+      "learning_rate": 0.0002,
+      "loss": 0.7998,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2651362056422278,
+      "grad_norm": 0.4454376697540283,
+      "learning_rate": 0.0002,
+      "loss": 0.8198,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26675288982297307,
+      "grad_norm": 0.4355238378047943,
+      "learning_rate": 0.0002,
+      "loss": 0.8684,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26836957400371836,
+      "grad_norm": 0.458310067653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26998625818446365,
+      "grad_norm": 0.4752083718776703,
+      "learning_rate": 0.0002,
+      "loss": 0.7935,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27160294236520893,
+      "grad_norm": 0.4666106402873993,
+      "learning_rate": 0.0002,
+      "loss": 0.8267,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2732196265459543,
+      "grad_norm": 0.4213818609714508,
+      "learning_rate": 0.0002,
+      "loss": 0.8252,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27483631072669956,
+      "grad_norm": 0.5768913626670837,
+      "learning_rate": 0.0002,
+      "loss": 0.8559,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27645299490744485,
+      "grad_norm": 0.4209914803504944,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27806967908819014,
+      "grad_norm": 0.501909613609314,
+      "learning_rate": 0.0002,
+      "loss": 0.8167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2796863632689354,
+      "grad_norm": 0.5266261100769043,
+      "learning_rate": 0.0002,
+      "loss": 0.7832,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2813030474496807,
+      "grad_norm": 0.43806859850883484,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 1740
+    },
+    {
+      "epoch": 0.282919731630426,
+      "grad_norm": 0.46048814058303833,
+      "learning_rate": 0.0002,
+      "loss": 0.8157,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2845364158111713,
+      "grad_norm": 0.44972819089889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8596,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28615309999191657,
+      "grad_norm": 0.5114831328392029,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.47931742668151855,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28938646835340714,
+      "grad_norm": 0.5092599987983704,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29100315253415243,
+      "grad_norm": 0.37581443786621094,
+      "learning_rate": 0.0002,
+      "loss": 0.8506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2926198367148977,
+      "grad_norm": 0.47097381949424744,
+      "learning_rate": 0.0002,
+      "loss": 0.7932,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29423652089564306,
+      "grad_norm": 0.48300236463546753,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29585320507638835,
+      "grad_norm": 0.5600419640541077,
+      "learning_rate": 0.0002,
+      "loss": 0.8391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29746988925713364,
+      "grad_norm": 0.48555272817611694,
+      "learning_rate": 0.0002,
+      "loss": 0.8507,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2990865734378789,
+      "grad_norm": 0.3752668499946594,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3007032576186242,
+      "grad_norm": 0.5328747034072876,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3023199417993695,
+      "grad_norm": 0.48716455698013306,
+      "learning_rate": 0.0002,
+      "loss": 0.8426,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3039366259801148,
+      "grad_norm": 0.5011493563652039,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30555331016086007,
+      "grad_norm": 0.46461427211761475,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30716999434160536,
+      "grad_norm": 0.36630210280418396,
+      "learning_rate": 0.0002,
+      "loss": 0.8478,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30878667852235064,
+      "grad_norm": 0.4217296242713928,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31040336270309593,
+      "grad_norm": 0.4394875466823578,
+      "learning_rate": 0.0002,
+      "loss": 0.8128,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3120200468838412,
+      "grad_norm": 0.6587965488433838,
+      "learning_rate": 0.0002,
+      "loss": 0.8471,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31363673106458656,
+      "grad_norm": 0.5469298958778381,
+      "learning_rate": 0.0002,
+      "loss": 0.8565,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31525341524533185,
+      "grad_norm": 0.4371595084667206,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31687009942607713,
+      "grad_norm": 0.4809541404247284,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3184867836068224,
+      "grad_norm": 0.6061086654663086,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3201034677875677,
+      "grad_norm": 0.5342657566070557,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 1980
+    },
+    {
+      "epoch": 0.321720151968313,
+      "grad_norm": 0.5057743787765503,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3233368361490583,
+      "grad_norm": 0.528626024723053,
+      "learning_rate": 0.0002,
+      "loss": 0.7774,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32495352032980357,
+      "grad_norm": 0.46742770075798035,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32657020451054886,
+      "grad_norm": 0.515101432800293,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32818688869129414,
+      "grad_norm": 0.41941216588020325,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32980357287203943,
+      "grad_norm": 0.49902522563934326,
+      "learning_rate": 0.0002,
+      "loss": 0.7637,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3314202570527847,
+      "grad_norm": 0.4120897650718689,
+      "learning_rate": 0.0002,
+      "loss": 0.7822,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33303694123353,
+      "grad_norm": 0.45352041721343994,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33465362541427535,
+      "grad_norm": 0.523199737071991,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33627030959502063,
+      "grad_norm": 0.4390358626842499,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3378869937757659,
+      "grad_norm": 0.6752901077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3395036779565112,
+      "grad_norm": 0.547821044921875,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3411203621372565,
+      "grad_norm": 0.5161308646202087,
+      "learning_rate": 0.0002,
+      "loss": 0.8561,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3427370463180018,
+      "grad_norm": 0.4565401077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34435373049874707,
+      "grad_norm": 0.4666115939617157,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34597041467949236,
+      "grad_norm": 0.4090428352355957,
+      "learning_rate": 0.0002,
+      "loss": 0.8189,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34758709886023764,
+      "grad_norm": 0.510845422744751,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34920378304098293,
+      "grad_norm": 0.42861923575401306,
+      "learning_rate": 0.0002,
+      "loss": 0.8398,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3508204672217282,
+      "grad_norm": 0.4476332664489746,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3524371514024735,
+      "grad_norm": 0.6065791249275208,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35405383558321885,
+      "grad_norm": 0.42335066199302673,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35567051976396413,
+      "grad_norm": 0.5094629526138306,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3572872039447094,
+      "grad_norm": 0.5476373434066772,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3589038881254547,
+      "grad_norm": 0.3911719024181366,
+      "learning_rate": 0.0002,
+      "loss": 0.814,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3605205723062,
+      "grad_norm": 0.6599636077880859,
+      "learning_rate": 0.0002,
+      "loss": 0.8599,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3621372564869453,
+      "grad_norm": 0.40381914377212524,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36375394066769057,
+      "grad_norm": 0.4433908462524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36537062484843585,
+      "grad_norm": 0.578326940536499,
+      "learning_rate": 0.0002,
+      "loss": 0.8503,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36698730902918114,
+      "grad_norm": 0.5734784007072449,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36860399320992643,
+      "grad_norm": 0.45555487275123596,
+      "learning_rate": 0.0002,
+      "loss": 0.8193,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3702206773906717,
+      "grad_norm": 0.5666276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.7929,
+      "step": 2290
+    },
+    {
+      "epoch": 0.371837361571417,
+      "grad_norm": 0.5461117625236511,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3734540457521623,
+      "grad_norm": 0.6318911910057068,
+      "learning_rate": 0.0002,
+      "loss": 0.8204,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37507072993290763,
+      "grad_norm": 0.493263304233551,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3766874141136529,
+      "grad_norm": 0.5888760089874268,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3783040982943982,
+      "grad_norm": 0.48671841621398926,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3799207824751435,
+      "grad_norm": 0.4385145306587219,
+      "learning_rate": 0.0002,
+      "loss": 0.8367,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3815374666558888,
+      "grad_norm": 0.5523318648338318,
+      "learning_rate": 0.0002,
+      "loss": 0.812,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38315415083663407,
+      "grad_norm": 0.7308220267295837,
+      "learning_rate": 0.0002,
+      "loss": 0.8351,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38477083501737935,
+      "grad_norm": 0.554214358329773,
+      "learning_rate": 0.0002,
+      "loss": 0.859,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38638751919812464,
+      "grad_norm": 0.5425800085067749,
+      "learning_rate": 0.0002,
+      "loss": 0.8146,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3880042033788699,
+      "grad_norm": 0.48811158537864685,
+      "learning_rate": 0.0002,
+      "loss": 0.8282,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3896208875596152,
+      "grad_norm": 0.49212366342544556,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3912375717403605,
+      "grad_norm": 0.5222218632698059,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3928542559211058,
+      "grad_norm": 0.4699819087982178,
+      "learning_rate": 0.0002,
+      "loss": 0.8182,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39447094010185113,
+      "grad_norm": 0.46153587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7919,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3960876242825964,
+      "grad_norm": 0.4150611162185669,
+      "learning_rate": 0.0002,
+      "loss": 0.8111,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3977043084633417,
+      "grad_norm": 0.5799614787101746,
+      "learning_rate": 0.0002,
+      "loss": 0.8589,
+      "step": 2460
+    },
+    {
+      "epoch": 0.399320992644087,
+      "grad_norm": 0.56536865234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4009376768248323,
+      "grad_norm": 0.5451247096061707,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40255436100557757,
+      "grad_norm": 0.5914521217346191,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40417104518632285,
+      "grad_norm": 0.4428117275238037,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40578772936706814,
+      "grad_norm": 0.48580947518348694,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4074044135478134,
+      "grad_norm": 0.436734676361084,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4090210977285587,
+      "grad_norm": 0.5752223134040833,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 2530
+    },
+    {
+      "epoch": 0.410637781909304,
+      "grad_norm": 0.4271308183670044,
+      "learning_rate": 0.0002,
+      "loss": 0.8181,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4122544660900493,
+      "grad_norm": 0.46294718980789185,
+      "learning_rate": 0.0002,
+      "loss": 0.8058,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4138711502707946,
+      "grad_norm": 0.49407583475112915,
+      "learning_rate": 0.0002,
+      "loss": 0.8473,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4154878344515399,
+      "grad_norm": 0.4729035496711731,
+      "learning_rate": 0.0002,
+      "loss": 0.7881,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4171045186322852,
+      "grad_norm": 0.4129747152328491,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4187212028130305,
+      "grad_norm": 0.5684236288070679,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4203378869937758,
+      "grad_norm": 0.4862157106399536,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42195457117452106,
+      "grad_norm": 0.46567976474761963,
+      "learning_rate": 0.0002,
+      "loss": 0.7582,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42357125535526635,
+      "grad_norm": 0.5710650682449341,
+      "learning_rate": 0.0002,
+      "loss": 0.7755,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42518793953601164,
+      "grad_norm": 0.5660041570663452,
+      "learning_rate": 0.0002,
+      "loss": 0.8573,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4268046237167569,
+      "grad_norm": 0.47944375872612,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4284213078975022,
+      "grad_norm": 0.537223756313324,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4300379920782475,
+      "grad_norm": 0.41669997572898865,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 0.44727686047554016,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4332713604397381,
+      "grad_norm": 0.5600888729095459,
+      "learning_rate": 0.0002,
+      "loss": 0.8241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4348880446204834,
+      "grad_norm": 0.39820605516433716,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4365047288012287,
+      "grad_norm": 0.5637655854225159,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 2700
+    },
+    {
+      "epoch": 0.438121412981974,
+      "grad_norm": 0.6363666653633118,
+      "learning_rate": 0.0002,
+      "loss": 0.855,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4397380971627193,
+      "grad_norm": 0.5656129121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.8468,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44135478134346456,
+      "grad_norm": 0.5600156188011169,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44297146552420985,
+      "grad_norm": 0.5506579875946045,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44458814970495514,
+      "grad_norm": 0.49878305196762085,
+      "learning_rate": 0.0002,
+      "loss": 0.7725,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4462048338857004,
+      "grad_norm": 0.4569213092327118,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4478215180664457,
+      "grad_norm": 0.6056680083274841,
+      "learning_rate": 0.0002,
+      "loss": 0.8028,
+      "step": 2770
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.44474557042121887,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4510548864279363,
+      "grad_norm": 0.46055394411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.801,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4526715706086816,
+      "grad_norm": 0.4904133379459381,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45428825478942686,
+      "grad_norm": 0.5647031664848328,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4559049389701722,
+      "grad_norm": 0.5759473443031311,
+      "learning_rate": 0.0002,
+      "loss": 0.8622,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4575216231509175,
+      "grad_norm": 0.5161895751953125,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4591383073316628,
+      "grad_norm": 0.4248254597187042,
+      "learning_rate": 0.0002,
+      "loss": 0.8045,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46075499151240806,
+      "grad_norm": 0.45395001769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.7838,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46237167569315335,
+      "grad_norm": 0.5358697772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.8208,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46398835987389864,
+      "grad_norm": 0.5379165410995483,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4656050440546439,
+      "grad_norm": 0.4601989686489105,
+      "learning_rate": 0.0002,
+      "loss": 0.7403,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4672217282353892,
+      "grad_norm": 0.671115517616272,
+      "learning_rate": 0.0002,
+      "loss": 0.8523,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4688384124161345,
+      "grad_norm": 0.4425133168697357,
+      "learning_rate": 0.0002,
+      "loss": 0.8262,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4704550965968798,
+      "grad_norm": 0.5446155071258545,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47207178077762507,
+      "grad_norm": 0.603306233882904,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47368846495837036,
+      "grad_norm": 0.5377997159957886,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4753051491391157,
+      "grad_norm": 0.4931027591228485,
+      "learning_rate": 0.0002,
+      "loss": 0.8075,
+      "step": 2940
+    },
+    {
+      "epoch": 0.476921833319861,
+      "grad_norm": 0.4711960256099701,
+      "learning_rate": 0.0002,
+      "loss": 0.8004,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4785385175006063,
+      "grad_norm": 0.5020492672920227,
+      "learning_rate": 0.0002,
+      "loss": 0.8121,
+      "step": 2960
+    },
+    {
+      "epoch": 0.48015520168135156,
+      "grad_norm": 0.5428946614265442,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48177188586209685,
+      "grad_norm": 0.5294089317321777,
+      "learning_rate": 0.0002,
+      "loss": 0.7849,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48338857004284214,
+      "grad_norm": 0.648289144039154,
+      "learning_rate": 0.0002,
+      "loss": 0.8553,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4850052542235874,
+      "grad_norm": 0.47916680574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.7874,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4866219384043327,
+      "grad_norm": 0.43849772214889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8087,
+      "step": 3010
+    },
+    {
+      "epoch": 0.488238622585078,
+      "grad_norm": 0.47007861733436584,
+      "learning_rate": 0.0002,
+      "loss": 0.7662,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4898553067658233,
+      "grad_norm": 0.6314331293106079,
+      "learning_rate": 0.0002,
+      "loss": 0.757,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49147199094656857,
+      "grad_norm": 0.49211493134498596,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49308867512731386,
+      "grad_norm": 0.4537973403930664,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49470535930805914,
+      "grad_norm": 0.47326919436454773,
+      "learning_rate": 0.0002,
+      "loss": 0.8095,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4963220434888045,
+      "grad_norm": 0.525874137878418,
+      "learning_rate": 0.0002,
+      "loss": 0.8447,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4979387276695498,
+      "grad_norm": 0.6361091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49955541185029506,
+      "grad_norm": 0.5850642919540405,
+      "learning_rate": 0.0002,
+      "loss": 0.821,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5011720960310403,
+      "grad_norm": 0.47299543023109436,
+      "learning_rate": 0.0002,
+      "loss": 0.8279,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5027887802117856,
+      "grad_norm": 0.473099946975708,
+      "learning_rate": 0.0002,
+      "loss": 0.8681,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5044054643925309,
+      "grad_norm": 0.48186397552490234,
+      "learning_rate": 0.0002,
+      "loss": 0.8223,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5060221485732762,
+      "grad_norm": 0.5015401840209961,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5076388327540216,
+      "grad_norm": 0.5617750287055969,
+      "learning_rate": 0.0002,
+      "loss": 0.7692,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5092555169347668,
+      "grad_norm": 0.5169327259063721,
+      "learning_rate": 0.0002,
+      "loss": 0.8708,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5108722011155121,
+      "grad_norm": 0.545657753944397,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5124888852962574,
+      "grad_norm": 0.512864351272583,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5141055694770027,
+      "grad_norm": 0.4113546311855316,
+      "learning_rate": 0.0002,
+      "loss": 0.7794,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5157222536577479,
+      "grad_norm": 0.44532445073127747,
+      "learning_rate": 0.0002,
+      "loss": 0.8206,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5173389378384933,
+      "grad_norm": 0.5623497366905212,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5189556220192385,
+      "grad_norm": 0.5084741115570068,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5205723061999838,
+      "grad_norm": 0.5305403470993042,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5221889903807291,
+      "grad_norm": 0.4708254337310791,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5238056745614744,
+      "grad_norm": 0.43827131390571594,
+      "learning_rate": 0.0002,
+      "loss": 0.7639,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5254223587422197,
+      "grad_norm": 0.5630002617835999,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 3250
+    },
+    {
+      "epoch": 0.527039042922965,
+      "grad_norm": 0.5010961890220642,
+      "learning_rate": 0.0002,
+      "loss": 0.7522,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5286557271037103,
+      "grad_norm": 0.6303122043609619,
+      "learning_rate": 0.0002,
+      "loss": 0.8374,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5302724112844556,
+      "grad_norm": 0.5107331275939941,
+      "learning_rate": 0.0002,
+      "loss": 0.7727,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5318890954652009,
+      "grad_norm": 0.5700443387031555,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5335057796459461,
+      "grad_norm": 0.46296367049217224,
+      "learning_rate": 0.0002,
+      "loss": 0.7776,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5351224638266915,
+      "grad_norm": 0.531568706035614,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5367391480074367,
+      "grad_norm": 0.4686741530895233,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5383558321881821,
+      "grad_norm": 0.5404331088066101,
+      "learning_rate": 0.0002,
+      "loss": 0.8104,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5399725163689273,
+      "grad_norm": 0.6368790864944458,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5415892005496726,
+      "grad_norm": 0.42300888895988464,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5432058847304179,
+      "grad_norm": 0.5362542867660522,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5448225689111632,
+      "grad_norm": 0.497128963470459,
+      "learning_rate": 0.0002,
+      "loss": 0.858,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5464392530919085,
+      "grad_norm": 0.5006386041641235,
+      "learning_rate": 0.0002,
+      "loss": 0.8519,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5480559372726538,
+      "grad_norm": 0.44136837124824524,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5496726214533991,
+      "grad_norm": 0.5897833108901978,
+      "learning_rate": 0.0002,
+      "loss": 0.773,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5512893056341444,
+      "grad_norm": 0.641075611114502,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5529059898148897,
+      "grad_norm": 0.7251322269439697,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5545226739956349,
+      "grad_norm": 0.47411349415779114,
+      "learning_rate": 0.0002,
+      "loss": 0.7626,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5561393581763803,
+      "grad_norm": 0.4994310438632965,
+      "learning_rate": 0.0002,
+      "loss": 0.8196,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5577560423571255,
+      "grad_norm": 0.5814438462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5593727265378708,
+      "grad_norm": 0.6278898119926453,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5609894107186161,
+      "grad_norm": 0.46208274364471436,
+      "learning_rate": 0.0002,
+      "loss": 0.813,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5626060948993614,
+      "grad_norm": 0.5718930959701538,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5642227790801067,
+      "grad_norm": 0.48178744316101074,
+      "learning_rate": 0.0002,
+      "loss": 0.8152,
+      "step": 3490
+    },
+    {
+      "epoch": 0.565839463260852,
+      "grad_norm": 0.47336965799331665,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5674561474415973,
+      "grad_norm": 0.43442684412002563,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5690728316223426,
+      "grad_norm": 0.6463358998298645,
+      "learning_rate": 0.0002,
+      "loss": 0.7564,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5706895158030879,
+      "grad_norm": 0.5286486744880676,
+      "learning_rate": 0.0002,
+      "loss": 0.836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5723061999838331,
+      "grad_norm": 0.5405499935150146,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5739228841645785,
+      "grad_norm": 0.6654391884803772,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.5081980228424072,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5771562525260691,
+      "grad_norm": 0.48978179693222046,
+      "learning_rate": 0.0002,
+      "loss": 0.7753,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5787729367068143,
+      "grad_norm": 0.5840612053871155,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5803896208875596,
+      "grad_norm": 0.5235261917114258,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5820063050683049,
+      "grad_norm": 0.5672075748443604,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5836229892490502,
+      "grad_norm": 0.5613429546356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5852396734297954,
+      "grad_norm": 0.4032273590564728,
+      "learning_rate": 0.0002,
+      "loss": 0.8274,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5868563576105408,
+      "grad_norm": 0.49559324979782104,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5884730417912861,
+      "grad_norm": 0.6895697712898254,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5900897259720314,
+      "grad_norm": 0.4750136435031891,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5917064101527767,
+      "grad_norm": 0.5176819562911987,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5933230943335219,
+      "grad_norm": 0.5817760229110718,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5949397785142673,
+      "grad_norm": 0.6064626574516296,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5965564626950125,
+      "grad_norm": 0.6728700995445251,
+      "learning_rate": 0.0002,
+      "loss": 0.8422,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5981731468757578,
+      "grad_norm": 0.609305202960968,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5997898310565031,
+      "grad_norm": 0.4615488350391388,
+      "learning_rate": 0.0002,
+      "loss": 0.8048,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6014065152372484,
+      "grad_norm": 2.0531179904937744,
+      "learning_rate": 0.0002,
+      "loss": 0.8214,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6030231994179936,
+      "grad_norm": 0.5091132521629333,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 3730
+    },
+    {
+      "epoch": 0.604639883598739,
+      "grad_norm": 0.5951124429702759,
+      "learning_rate": 0.0002,
+      "loss": 0.7833,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6062565677794842,
+      "grad_norm": 0.5870208144187927,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6078732519602296,
+      "grad_norm": 0.6254619359970093,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6094899361409749,
+      "grad_norm": 0.5577626824378967,
+      "learning_rate": 0.0002,
+      "loss": 0.7868,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6111066203217201,
+      "grad_norm": 0.5004405379295349,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6127233045024655,
+      "grad_norm": 0.5527383685112,
+      "learning_rate": 0.0002,
+      "loss": 0.8092,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6143399886832107,
+      "grad_norm": 0.49116113781929016,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6159566728639561,
+      "grad_norm": 0.5299299359321594,
+      "learning_rate": 0.0002,
+      "loss": 0.8352,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6175733570447013,
+      "grad_norm": 0.464897483587265,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6191900412254466,
+      "grad_norm": 0.6505740880966187,
+      "learning_rate": 0.0002,
+      "loss": 0.7923,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6208067254061919,
+      "grad_norm": 0.5512559413909912,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6224234095869372,
+      "grad_norm": 0.49427518248558044,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6240400937676824,
+      "grad_norm": 0.3839147090911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6256567779484278,
+      "grad_norm": 0.5760218501091003,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6272734621291731,
+      "grad_norm": 0.7226507067680359,
+      "learning_rate": 0.0002,
+      "loss": 0.7836,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6288901463099184,
+      "grad_norm": 0.676781415939331,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6305068304906637,
+      "grad_norm": 0.4284018278121948,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6321235146714089,
+      "grad_norm": 0.5060628056526184,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6337401988521543,
+      "grad_norm": 0.5524522066116333,
+      "learning_rate": 0.0002,
+      "loss": 0.8089,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6353568830328995,
+      "grad_norm": 0.6099881529808044,
+      "learning_rate": 0.0002,
+      "loss": 0.8276,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6369735672136448,
+      "grad_norm": 0.43155938386917114,
+      "learning_rate": 0.0002,
+      "loss": 0.809,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6385902513943901,
+      "grad_norm": 0.6427084803581238,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6402069355751354,
+      "grad_norm": 0.541220486164093,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6418236197558806,
+      "grad_norm": 0.5414294600486755,
+      "learning_rate": 0.0002,
+      "loss": 0.8539,
+      "step": 3970
+    },
+    {
+      "epoch": 0.643440303936626,
+      "grad_norm": 0.46344003081321716,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6450569881173712,
+      "grad_norm": 0.45209285616874695,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6466736722981166,
+      "grad_norm": 0.5417284369468689,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6482903564788619,
+      "grad_norm": 0.7995685935020447,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6499070406596071,
+      "grad_norm": 0.6384002566337585,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6515237248403525,
+      "grad_norm": 0.4472815692424774,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6531404090210977,
+      "grad_norm": 0.6834294199943542,
+      "learning_rate": 0.0002,
+      "loss": 0.8071,
+      "step": 4040
+    },
+    {
+      "epoch": 0.654757093201843,
+      "grad_norm": 0.4612339735031128,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6563737773825883,
+      "grad_norm": 0.9266576170921326,
+      "learning_rate": 0.0002,
+      "loss": 0.8141,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6579904615633336,
+      "grad_norm": 0.4470861852169037,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6596071457440789,
+      "grad_norm": 0.45544925332069397,
+      "learning_rate": 0.0002,
+      "loss": 0.8293,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6612238299248242,
+      "grad_norm": 0.6144481301307678,
+      "learning_rate": 0.0002,
+      "loss": 0.8455,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6628405141055694,
+      "grad_norm": 0.5936288237571716,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6644571982863148,
+      "grad_norm": 0.4822963774204254,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 4110
+    },
+    {
+      "epoch": 0.66607388246706,
+      "grad_norm": 0.48432496190071106,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6676905666478054,
+      "grad_norm": 0.4901607930660248,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6693072508285507,
+      "grad_norm": 0.5018393397331238,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6709239350092959,
+      "grad_norm": 0.6946378946304321,
+      "learning_rate": 0.0002,
+      "loss": 0.8065,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6725406191900413,
+      "grad_norm": 0.5997390747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.6738849878311157,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6757739875515318,
+      "grad_norm": 0.6110581159591675,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6773906717322771,
+      "grad_norm": 0.5703322291374207,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6790073559130224,
+      "grad_norm": 0.4686066210269928,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6806240400937676,
+      "grad_norm": 0.6394643783569336,
+      "learning_rate": 0.0002,
+      "loss": 0.8441,
+      "step": 4210
+    },
+    {
+      "epoch": 0.682240724274513,
+      "grad_norm": 0.5454841256141663,
+      "learning_rate": 0.0002,
+      "loss": 0.8011,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6838574084552582,
+      "grad_norm": 0.4859732985496521,
+      "learning_rate": 0.0002,
+      "loss": 0.8307,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6854740926360036,
+      "grad_norm": 0.5544065833091736,
+      "learning_rate": 0.0002,
+      "loss": 0.8161,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6870907768167488,
+      "grad_norm": 0.4902505576610565,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6887074609974941,
+      "grad_norm": 0.4768051505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7977,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6903241451782395,
+      "grad_norm": 0.49982190132141113,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6919408293589847,
+      "grad_norm": 0.6351838111877441,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 4280
+    },
+    {
+      "epoch": 0.69355751353973,
+      "grad_norm": 0.5647561550140381,
+      "learning_rate": 0.0002,
+      "loss": 0.7664,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6951741977204753,
+      "grad_norm": 0.5340486764907837,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6967908819012206,
+      "grad_norm": 0.5649092793464661,
+      "learning_rate": 0.0002,
+      "loss": 0.8526,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6984075660819659,
+      "grad_norm": 0.6183916926383972,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7000242502627112,
+      "grad_norm": 0.6154509782791138,
+      "learning_rate": 0.0002,
+      "loss": 0.792,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7016409344434564,
+      "grad_norm": 0.5156264305114746,
+      "learning_rate": 0.0002,
+      "loss": 0.8397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7032576186242018,
+      "grad_norm": 0.562171459197998,
+      "learning_rate": 0.0002,
+      "loss": 0.8512,
+      "step": 4350
+    },
+    {
+      "epoch": 0.704874302804947,
+      "grad_norm": 0.4949502646923065,
+      "learning_rate": 0.0002,
+      "loss": 0.7882,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7064909869856923,
+      "grad_norm": 0.5171684622764587,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7081076711664377,
+      "grad_norm": 0.6198443174362183,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7097243553471829,
+      "grad_norm": 0.5802276134490967,
+      "learning_rate": 0.0002,
+      "loss": 0.7606,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7113410395279283,
+      "grad_norm": 0.41096967458724976,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7129577237086735,
+      "grad_norm": 0.4397392272949219,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7145744078894188,
+      "grad_norm": 0.45228442549705505,
+      "learning_rate": 0.0002,
+      "loss": 0.7651,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7161910920701641,
+      "grad_norm": 0.4839673936367035,
+      "learning_rate": 0.0002,
+      "loss": 0.7938,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7178077762509094,
+      "grad_norm": 0.6140755414962769,
+      "learning_rate": 0.0002,
+      "loss": 0.8362,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 0.6841378808021545,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7210411446124,
+      "grad_norm": 0.6664239168167114,
+      "learning_rate": 0.0002,
+      "loss": 0.8177,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7226578287931452,
+      "grad_norm": 0.47552719712257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7983,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7242745129738906,
+      "grad_norm": 0.6649776101112366,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7258911971546358,
+      "grad_norm": 0.5159541964530945,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7275078813353811,
+      "grad_norm": 0.6693112850189209,
+      "learning_rate": 0.0002,
+      "loss": 0.7786,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7291245655161265,
+      "grad_norm": 0.48870977759361267,
+      "learning_rate": 0.0002,
+      "loss": 0.8655,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7307412496968717,
+      "grad_norm": 0.4857887923717499,
+      "learning_rate": 0.0002,
+      "loss": 0.7337,
+      "step": 4520
+    },
+    {
+      "epoch": 0.732357933877617,
+      "grad_norm": 0.5515662431716919,
+      "learning_rate": 0.0002,
+      "loss": 0.8026,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7339746180583623,
+      "grad_norm": 0.6292222738265991,
+      "learning_rate": 0.0002,
+      "loss": 0.8031,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7355913022391076,
+      "grad_norm": 0.48265689611434937,
+      "learning_rate": 0.0002,
+      "loss": 0.7749,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7372079864198529,
+      "grad_norm": 0.8044266104698181,
+      "learning_rate": 0.0002,
+      "loss": 0.8499,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7388246706005982,
+      "grad_norm": 0.6111769676208496,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7404413547813434,
+      "grad_norm": 0.5229553580284119,
+      "learning_rate": 0.0002,
+      "loss": 0.7291,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7420580389620888,
+      "grad_norm": 0.6054152250289917,
+      "learning_rate": 0.0002,
+      "loss": 0.8038,
+      "step": 4590
+    },
+    {
+      "epoch": 0.743674723142834,
+      "grad_norm": 0.5574966669082642,
+      "learning_rate": 0.0002,
+      "loss": 0.8169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7452914073235793,
+      "grad_norm": 0.5395817160606384,
+      "learning_rate": 0.0002,
+      "loss": 0.8439,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7469080915043246,
+      "grad_norm": 0.7116472721099854,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7485247756850699,
+      "grad_norm": 0.5618700981140137,
+      "learning_rate": 0.0002,
+      "loss": 0.7743,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7501414598658153,
+      "grad_norm": 0.5802770853042603,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7517581440465605,
+      "grad_norm": 0.5690428018569946,
+      "learning_rate": 0.0002,
+      "loss": 0.7924,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7533748282273058,
+      "grad_norm": 0.4813360273838043,
+      "learning_rate": 0.0002,
+      "loss": 0.8017,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7549915124080511,
+      "grad_norm": 0.5434042811393738,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7566081965887964,
+      "grad_norm": 0.5502099990844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7582248807695416,
+      "grad_norm": 0.6020621061325073,
+      "learning_rate": 0.0002,
+      "loss": 0.8598,
+      "step": 4690
+    },
+    {
+      "epoch": 0.759841564950287,
+      "grad_norm": 0.4922301471233368,
+      "learning_rate": 0.0002,
+      "loss": 0.7937,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7614582491310322,
+      "grad_norm": 0.6492828726768494,
+      "learning_rate": 0.0002,
+      "loss": 0.788,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7630749333117776,
+      "grad_norm": 0.4865580201148987,
+      "learning_rate": 0.0002,
+      "loss": 0.8313,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7646916174925228,
+      "grad_norm": 0.5971422791481018,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7663083016732681,
+      "grad_norm": 0.6832674145698547,
+      "learning_rate": 0.0002,
+      "loss": 0.8298,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7679249858540134,
+      "grad_norm": 0.500908613204956,
+      "learning_rate": 0.0002,
+      "loss": 0.8156,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7695416700347587,
+      "grad_norm": 0.6112465858459473,
+      "learning_rate": 0.0002,
+      "loss": 0.8383,
+      "step": 4760
+    },
+    {
+      "epoch": 0.771158354215504,
+      "grad_norm": 0.5753506422042847,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7727750383962493,
+      "grad_norm": 0.6529405117034912,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7743917225769946,
+      "grad_norm": 0.5916843414306641,
+      "learning_rate": 0.0002,
+      "loss": 0.8171,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7760084067577399,
+      "grad_norm": 0.4821224510669708,
+      "learning_rate": 0.0002,
+      "loss": 0.83,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7776250909384852,
+      "grad_norm": 0.5532580018043518,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7792417751192304,
+      "grad_norm": 0.4604877233505249,
+      "learning_rate": 0.0002,
+      "loss": 0.7363,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7808584592999758,
+      "grad_norm": 0.5009613037109375,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 4830
+    },
+    {
+      "epoch": 0.782475143480721,
+      "grad_norm": 0.6448560357093811,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7840918276614663,
+      "grad_norm": 0.44327953457832336,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7857085118422116,
+      "grad_norm": 0.5355411171913147,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7873251960229569,
+      "grad_norm": 0.5635677576065063,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7889418802037023,
+      "grad_norm": 0.5417491793632507,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7905585643844475,
+      "grad_norm": 0.4567430913448334,
+      "learning_rate": 0.0002,
+      "loss": 0.7819,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7921752485651928,
+      "grad_norm": 0.44651296734809875,
+      "learning_rate": 0.0002,
+      "loss": 0.8454,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7937919327459381,
+      "grad_norm": 0.5741217136383057,
+      "learning_rate": 0.0002,
+      "loss": 0.7959,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7954086169266834,
+      "grad_norm": 0.6605045199394226,
+      "learning_rate": 0.0002,
+      "loss": 0.8093,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7970253011074286,
+      "grad_norm": 0.5126531720161438,
+      "learning_rate": 0.0002,
+      "loss": 0.77,
+      "step": 4930
+    },
+    {
+      "epoch": 0.798641985288174,
+      "grad_norm": 0.513648271560669,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8002586694689192,
+      "grad_norm": 0.5350404381752014,
+      "learning_rate": 0.0002,
+      "loss": 0.8314,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8018753536496646,
+      "grad_norm": 0.5731674432754517,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8034920378304098,
+      "grad_norm": 0.5974258184432983,
+      "learning_rate": 0.0002,
+      "loss": 0.8572,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8051087220111551,
+      "grad_norm": 0.8774799704551697,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8067254061919004,
+      "grad_norm": 0.5994430184364319,
+      "learning_rate": 0.0002,
+      "loss": 0.7899,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8083420903726457,
+      "grad_norm": 0.4894903004169464,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5000
+    },
+    {
+      "epoch": 0.809958774553391,
+      "grad_norm": 0.5218459367752075,
+      "learning_rate": 0.0002,
+      "loss": 0.78,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8115754587341363,
+      "grad_norm": 0.5232468843460083,
+      "learning_rate": 0.0002,
+      "loss": 0.817,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8131921429148816,
+      "grad_norm": 0.44358372688293457,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8148088270956269,
+      "grad_norm": 0.6202037334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8164255112763722,
+      "grad_norm": 0.7721474170684814,
+      "learning_rate": 0.0002,
+      "loss": 0.7351,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8180421954571174,
+      "grad_norm": 0.5568501353263855,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8196588796378628,
+      "grad_norm": 0.49148809909820557,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 5070
+    },
+    {
+      "epoch": 0.821275563818608,
+      "grad_norm": 0.4956012964248657,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8228922479993533,
+      "grad_norm": 0.6078833937644958,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8245089321800986,
+      "grad_norm": 0.46906954050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.828,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8261256163608439,
+      "grad_norm": 0.50812166929245,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8277423005415891,
+      "grad_norm": 0.5319661498069763,
+      "learning_rate": 0.0002,
+      "loss": 0.8243,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8293589847223345,
+      "grad_norm": 0.4949689209461212,
+      "learning_rate": 0.0002,
+      "loss": 0.7798,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8309756689030798,
+      "grad_norm": 0.5151591300964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8325923530838251,
+      "grad_norm": 0.5530214309692383,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8342090372645704,
+      "grad_norm": 0.6297410130500793,
+      "learning_rate": 0.0002,
+      "loss": 0.8251,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8358257214453156,
+      "grad_norm": 0.5466840267181396,
+      "learning_rate": 0.0002,
+      "loss": 0.8067,
+      "step": 5170
+    },
+    {
+      "epoch": 0.837442405626061,
+      "grad_norm": 0.652913510799408,
+      "learning_rate": 0.0002,
+      "loss": 0.7875,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8390590898068062,
+      "grad_norm": 0.5811293125152588,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8406757739875516,
+      "grad_norm": 0.5109550952911377,
+      "learning_rate": 0.0002,
+      "loss": 0.7412,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8422924581682968,
+      "grad_norm": 0.4551706612110138,
+      "learning_rate": 0.0002,
+      "loss": 0.8077,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8439091423490421,
+      "grad_norm": 0.5813754200935364,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8455258265297874,
+      "grad_norm": 0.5856947898864746,
+      "learning_rate": 0.0002,
+      "loss": 0.802,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8471425107105327,
+      "grad_norm": 0.5482739210128784,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8487591948912779,
+      "grad_norm": 0.49023720622062683,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8503758790720233,
+      "grad_norm": 0.49472475051879883,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8519925632527686,
+      "grad_norm": 0.5490226745605469,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8536092474335139,
+      "grad_norm": 0.5340665578842163,
+      "learning_rate": 0.0002,
+      "loss": 0.8333,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8552259316142592,
+      "grad_norm": 0.5962483882904053,
+      "learning_rate": 0.0002,
+      "loss": 0.8277,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8568426157950044,
+      "grad_norm": 0.586358368396759,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8584592999757498,
+      "grad_norm": 0.49120277166366577,
+      "learning_rate": 0.0002,
+      "loss": 0.7831,
+      "step": 5310
+    },
+    {
+      "epoch": 0.860075984156495,
+      "grad_norm": 0.5887332558631897,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8616926683372403,
+      "grad_norm": 0.42496153712272644,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.5489874482154846,
+      "learning_rate": 0.0002,
+      "loss": 0.7905,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8649260366987309,
+      "grad_norm": 0.5850813984870911,
+      "learning_rate": 0.0002,
+      "loss": 0.7958,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8665427208794761,
+      "grad_norm": 0.517487108707428,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8681594050602215,
+      "grad_norm": 0.5339142680168152,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8697760892409668,
+      "grad_norm": 0.6236387491226196,
+      "learning_rate": 0.0002,
+      "loss": 0.818,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8713927734217121,
+      "grad_norm": 0.5752192735671997,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8730094576024574,
+      "grad_norm": 0.6724614500999451,
+      "learning_rate": 0.0002,
+      "loss": 0.8542,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8746261417832026,
+      "grad_norm": 0.5280613303184509,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 5410
+    },
+    {
+      "epoch": 0.876242825963948,
+      "grad_norm": 0.44033288955688477,
+      "learning_rate": 0.0002,
+      "loss": 0.8231,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8778595101446932,
+      "grad_norm": 0.5199708342552185,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8794761943254386,
+      "grad_norm": 0.46778348088264465,
+      "learning_rate": 0.0002,
+      "loss": 0.7852,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8810928785061838,
+      "grad_norm": 0.4657754898071289,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8827095626869291,
+      "grad_norm": 0.5472902655601501,
+      "learning_rate": 0.0002,
+      "loss": 0.7799,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8843262468676744,
+      "grad_norm": 0.4876766800880432,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8859429310484197,
+      "grad_norm": 0.5057248473167419,
+      "learning_rate": 0.0002,
+      "loss": 0.7906,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8875596152291649,
+      "grad_norm": 0.4637320637702942,
+      "learning_rate": 0.0002,
+      "loss": 0.8124,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8891762994099103,
+      "grad_norm": 0.471955806016922,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8907929835906556,
+      "grad_norm": 0.5209813714027405,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8924096677714008,
+      "grad_norm": 0.6213834285736084,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8940263519521462,
+      "grad_norm": 0.5215408205986023,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8956430361328914,
+      "grad_norm": 0.580478310585022,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8972597203136368,
+      "grad_norm": 0.49102169275283813,
+      "learning_rate": 0.0002,
+      "loss": 0.8371,
+      "step": 5550
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.6043479442596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7806,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9004930886751273,
+      "grad_norm": 0.5636463165283203,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9021097728558726,
+      "grad_norm": 0.5620124340057373,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9037264570366179,
+      "grad_norm": 0.5206354856491089,
+      "learning_rate": 0.0002,
+      "loss": 0.8083,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9053431412173631,
+      "grad_norm": 0.5798229575157166,
+      "learning_rate": 0.0002,
+      "loss": 0.8557,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9069598253981085,
+      "grad_norm": 0.6428212523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.8097,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9085765095788537,
+      "grad_norm": 0.48064687848091125,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9101931937595991,
+      "grad_norm": 0.6347860097885132,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9118098779403444,
+      "grad_norm": 0.5353913307189941,
+      "learning_rate": 0.0002,
+      "loss": 0.851,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9134265621210896,
+      "grad_norm": 0.5323944091796875,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5650
+    },
+    {
+      "epoch": 0.915043246301835,
+      "grad_norm": 0.5261843204498291,
+      "learning_rate": 0.0002,
+      "loss": 0.8393,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9166599304825802,
+      "grad_norm": 0.5451326966285706,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9182766146633256,
+      "grad_norm": 0.5183324217796326,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9198932988440708,
+      "grad_norm": 0.47229018807411194,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9215099830248161,
+      "grad_norm": 0.49180513620376587,
+      "learning_rate": 0.0002,
+      "loss": 0.7757,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9231266672055614,
+      "grad_norm": 0.5419785380363464,
+      "learning_rate": 0.0002,
+      "loss": 0.8735,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9247433513863067,
+      "grad_norm": 0.5408698916435242,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9263600355670519,
+      "grad_norm": 0.5286232829093933,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9279767197477973,
+      "grad_norm": 0.7539758086204529,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9295934039285425,
+      "grad_norm": 0.5166944861412048,
+      "learning_rate": 0.0002,
+      "loss": 0.8118,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9312100881092878,
+      "grad_norm": 0.6601425409317017,
+      "learning_rate": 0.0002,
+      "loss": 0.783,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9328267722900332,
+      "grad_norm": 0.5029960870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9344434564707784,
+      "grad_norm": 0.4926645755767822,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9360601406515238,
+      "grad_norm": 0.5739615559577942,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5790
+    },
+    {
+      "epoch": 0.937676824832269,
+      "grad_norm": 0.5058279037475586,
+      "learning_rate": 0.0002,
+      "loss": 0.8037,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9392935090130143,
+      "grad_norm": 0.5260962247848511,
+      "learning_rate": 0.0002,
+      "loss": 0.8537,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9409101931937596,
+      "grad_norm": 0.5768588185310364,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9425268773745049,
+      "grad_norm": 0.5170126557350159,
+      "learning_rate": 0.0002,
+      "loss": 0.8215,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9441435615552501,
+      "grad_norm": 0.5745864510536194,
+      "learning_rate": 0.0002,
+      "loss": 0.7422,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9457602457359955,
+      "grad_norm": 0.5551357865333557,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9473769299167407,
+      "grad_norm": 0.5776078701019287,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9489936140974861,
+      "grad_norm": 0.5340062379837036,
+      "learning_rate": 0.0002,
+      "loss": 0.8527,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9506102982782314,
+      "grad_norm": 0.6447290182113647,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9522269824589766,
+      "grad_norm": 0.5123815536499023,
+      "learning_rate": 0.0002,
+      "loss": 0.7945,
+      "step": 5890
+    },
+    {
+      "epoch": 0.953843666639722,
+      "grad_norm": 0.48547613620758057,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9554603508204672,
+      "grad_norm": 0.5791414976119995,
+      "learning_rate": 0.0002,
+      "loss": 0.7896,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9570770350012126,
+      "grad_norm": 0.6195011734962463,
+      "learning_rate": 0.0002,
+      "loss": 0.8408,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9586937191819578,
+      "grad_norm": 0.6323803067207336,
+      "learning_rate": 0.0002,
+      "loss": 0.7805,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9603104033627031,
+      "grad_norm": 0.45552879571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9619270875434484,
+      "grad_norm": 0.5796473622322083,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9635437717241937,
+      "grad_norm": 0.647261381149292,
+      "learning_rate": 0.0002,
+      "loss": 0.7672,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9651604559049389,
+      "grad_norm": 0.5487682819366455,
+      "learning_rate": 0.0002,
+      "loss": 0.8086,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9667771400856843,
+      "grad_norm": 0.5743663907051086,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9683938242664295,
+      "grad_norm": 0.5470591187477112,
+      "learning_rate": 0.0002,
+      "loss": 0.8153,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9700105084471748,
+      "grad_norm": 0.5901660323143005,
+      "learning_rate": 0.0002,
+      "loss": 0.8119,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9716271926279202,
+      "grad_norm": 0.6544759273529053,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9732438768086654,
+      "grad_norm": 0.6288470029830933,
+      "learning_rate": 0.0002,
+      "loss": 0.7536,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9748605609894108,
+      "grad_norm": 0.673153817653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 6030
+    },
+    {
+      "epoch": 0.976477245170156,
+      "grad_norm": 0.42854753136634827,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9780939293509013,
+      "grad_norm": 0.5227066278457642,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9797106135316466,
+      "grad_norm": 0.5372416973114014,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9813272977123919,
+      "grad_norm": 0.6026402115821838,
+      "learning_rate": 0.0002,
+      "loss": 0.7591,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9829439818931371,
+      "grad_norm": 0.49547791481018066,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9845606660738825,
+      "grad_norm": 0.4641951322555542,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9861773502546277,
+      "grad_norm": 0.5818535089492798,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9877940344353731,
+      "grad_norm": 0.63955157995224,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9894107186161183,
+      "grad_norm": 0.5649438500404358,
+      "learning_rate": 0.0002,
+      "loss": 0.7547,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9910274027968636,
+      "grad_norm": 0.5290433168411255,
+      "learning_rate": 0.0002,
+      "loss": 0.7861,
+      "step": 6130
+    },
+    {
+      "epoch": 0.992644086977609,
+      "grad_norm": 0.6399374008178711,
+      "learning_rate": 0.0002,
+      "loss": 0.8109,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9942607711583542,
+      "grad_norm": 0.6736576557159424,
+      "learning_rate": 0.0002,
+      "loss": 0.8373,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9958774553390995,
+      "grad_norm": 0.515420138835907,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9974941395198448,
+      "grad_norm": 0.562677800655365,
+      "learning_rate": 0.0002,
+      "loss": 0.8032,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9991108237005901,
+      "grad_norm": 0.7113858461380005,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9999191657909627,
+      "eval_loss": 1.0871200561523438,
+      "eval_runtime": 122.2071,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0007275078813354,
+      "grad_norm": 0.7111801505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0023441920620806,
+      "grad_norm": 0.5402125716209412,
+      "learning_rate": 0.0002,
+      "loss": 0.6865,
+      "step": 6200
+    },
+    {
+      "epoch": 1.003960876242826,
+      "grad_norm": 0.6098830103874207,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0055775604235713,
+      "grad_norm": 0.5829983353614807,
+      "learning_rate": 0.0002,
+      "loss": 0.7631,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0071942446043165,
+      "grad_norm": 0.5614621043205261,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0088109287850617,
+      "grad_norm": 0.5954238772392273,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0104276129658072,
+      "grad_norm": 0.6480574607849121,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0120442971465524,
+      "grad_norm": 0.6051128506660461,
+      "learning_rate": 0.0002,
+      "loss": 0.7514,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0136609813272976,
+      "grad_norm": 0.6318870782852173,
+      "learning_rate": 0.0002,
+      "loss": 0.7237,
+      "step": 6270
+    },
+    {
+      "epoch": 1.015277665508043,
+      "grad_norm": 0.5048980116844177,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0168943496887883,
+      "grad_norm": 0.6346936225891113,
+      "learning_rate": 0.0002,
+      "loss": 0.7391,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0185110338695336,
+      "grad_norm": 0.5711665749549866,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0201277180502788,
+      "grad_norm": 0.5175361037254333,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0217444022310243,
+      "grad_norm": 0.5360831618309021,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0233610864117695,
+      "grad_norm": 0.614675760269165,
+      "learning_rate": 0.0002,
+      "loss": 0.7112,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0249777705925147,
+      "grad_norm": 0.5626118183135986,
+      "learning_rate": 0.0002,
+      "loss": 0.7748,
+      "step": 6340
+    },
+    {
+      "epoch": 1.02659445477326,
+      "grad_norm": 0.574897289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.7375,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0282111389540054,
+      "grad_norm": 0.7185447812080383,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0298278231347506,
+      "grad_norm": 0.6705799698829651,
+      "learning_rate": 0.0002,
+      "loss": 0.703,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0314445073154959,
+      "grad_norm": 0.6740428805351257,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0330611914962413,
+      "grad_norm": 0.663902759552002,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0346778756769865,
+      "grad_norm": 0.5029543042182922,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0362945598577318,
+      "grad_norm": 0.7813863158226013,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 6410
+    },
+    {
+      "epoch": 1.037911244038477,
+      "grad_norm": 0.5396282076835632,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0395279282192225,
+      "grad_norm": 0.5253293514251709,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0411446123999677,
+      "grad_norm": 0.7236770987510681,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 6440
+    },
+    {
+      "epoch": 1.042761296580713,
+      "grad_norm": 0.5670917630195618,
+      "learning_rate": 0.0002,
+      "loss": 0.7259,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0443779807614582,
+      "grad_norm": 0.6031978726387024,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0459946649422036,
+      "grad_norm": 0.5309213399887085,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0476113491229488,
+      "grad_norm": 0.7114651799201965,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 6480
+    },
+    {
+      "epoch": 1.049228033303694,
+      "grad_norm": 0.5591610670089722,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0508447174844395,
+      "grad_norm": 0.5185961127281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0524614016651848,
+      "grad_norm": 0.6510552167892456,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6510
+    },
+    {
+      "epoch": 1.05407808584593,
+      "grad_norm": 0.6557928919792175,
+      "learning_rate": 0.0002,
+      "loss": 0.7057,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0556947700266752,
+      "grad_norm": 0.6973192691802979,
+      "learning_rate": 0.0002,
+      "loss": 0.8056,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0573114542074207,
+      "grad_norm": 0.6226583123207092,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 6540
+    },
+    {
+      "epoch": 1.058928138388166,
+      "grad_norm": 0.5633195638656616,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0605448225689111,
+      "grad_norm": 0.7466658353805542,
+      "learning_rate": 0.0002,
+      "loss": 0.7082,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0621615067496564,
+      "grad_norm": 0.6462772488594055,
+      "learning_rate": 0.0002,
+      "loss": 0.7059,
+      "step": 6570
+    },
+    {
+      "epoch": 1.0637781909304018,
+      "grad_norm": 0.5266856551170349,
+      "learning_rate": 0.0002,
+      "loss": 0.7046,
+      "step": 6580
+    },
+    {
+      "epoch": 1.065394875111147,
+      "grad_norm": 0.534392774105072,
+      "learning_rate": 0.0002,
+      "loss": 0.7157,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0670115592918923,
+      "grad_norm": 0.7514177560806274,
+      "learning_rate": 0.0002,
+      "loss": 0.7115,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0686282434726375,
+      "grad_norm": 0.7593035697937012,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 6610
+    },
+    {
+      "epoch": 1.070244927653383,
+      "grad_norm": 0.5277858972549438,
+      "learning_rate": 0.0002,
+      "loss": 0.6836,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0718616118341282,
+      "grad_norm": 0.5573670268058777,
+      "learning_rate": 0.0002,
+      "loss": 0.7405,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0734782960148734,
+      "grad_norm": 0.6802396774291992,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 6640
+    },
+    {
+      "epoch": 1.0750949801956189,
+      "grad_norm": 0.7367215752601624,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0767116643763641,
+      "grad_norm": 0.5961891412734985,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0783283485571094,
+      "grad_norm": 0.5736313462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0799450327378546,
+      "grad_norm": 0.619219183921814,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0815617169186,
+      "grad_norm": 0.6214390993118286,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0831784010993453,
+      "grad_norm": 0.564536988735199,
+      "learning_rate": 0.0002,
+      "loss": 0.7554,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0847950852800905,
+      "grad_norm": 0.5838140249252319,
+      "learning_rate": 0.0002,
+      "loss": 0.7362,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0864117694608357,
+      "grad_norm": 0.7000553607940674,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0880284536415812,
+      "grad_norm": 0.7078263759613037,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 6730
+    },
+    {
+      "epoch": 1.0896451378223264,
+      "grad_norm": 0.8353848457336426,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6740
+    },
+    {
+      "epoch": 1.0912618220030716,
+      "grad_norm": 0.5615518689155579,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 6750
+    },
+    {
+      "epoch": 1.0928785061838169,
+      "grad_norm": 0.5475581288337708,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 6760
+    },
+    {
+      "epoch": 1.0944951903645623,
+      "grad_norm": 0.5835978388786316,
+      "learning_rate": 0.0002,
+      "loss": 0.7652,
+      "step": 6770
+    },
+    {
+      "epoch": 1.0961118745453076,
+      "grad_norm": 0.5516105890274048,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 6780
+    },
+    {
+      "epoch": 1.0977285587260528,
+      "grad_norm": 0.5875251889228821,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 6790
+    },
+    {
+      "epoch": 1.0993452429067982,
+      "grad_norm": 0.7376947999000549,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1009619270875435,
+      "grad_norm": 0.5656165480613708,
+      "learning_rate": 0.0002,
+      "loss": 0.7512,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1025786112682887,
+      "grad_norm": 0.6365954279899597,
+      "learning_rate": 0.0002,
+      "loss": 0.7409,
+      "step": 6820
+    },
+    {
+      "epoch": 1.104195295449034,
+      "grad_norm": 0.5033080577850342,
+      "learning_rate": 0.0002,
+      "loss": 0.7392,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1058119796297794,
+      "grad_norm": 0.617396891117096,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1074286638105246,
+      "grad_norm": 0.6395374536514282,
+      "learning_rate": 0.0002,
+      "loss": 0.7006,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1090453479912699,
+      "grad_norm": 0.6775295734405518,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1106620321720153,
+      "grad_norm": 0.6655223965644836,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1122787163527605,
+      "grad_norm": 0.676655113697052,
+      "learning_rate": 0.0002,
+      "loss": 0.7553,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1138954005335058,
+      "grad_norm": 0.6062718629837036,
+      "learning_rate": 0.0002,
+      "loss": 0.7342,
+      "step": 6890
+    },
+    {
+      "epoch": 1.115512084714251,
+      "grad_norm": 0.590943455696106,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1171287688949965,
+      "grad_norm": 0.6315317153930664,
+      "learning_rate": 0.0002,
+      "loss": 0.6705,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1187454530757417,
+      "grad_norm": 0.47979024052619934,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 6920
+    },
+    {
+      "epoch": 1.120362137256487,
+      "grad_norm": 0.647298276424408,
+      "learning_rate": 0.0002,
+      "loss": 0.7002,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1219788214372322,
+      "grad_norm": 0.7336484789848328,
+      "learning_rate": 0.0002,
+      "loss": 0.7502,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1235955056179776,
+      "grad_norm": 0.5071424245834351,
+      "learning_rate": 0.0002,
+      "loss": 0.693,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1252121897987228,
+      "grad_norm": 0.6527144312858582,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 6960
+    },
+    {
+      "epoch": 1.126828873979468,
+      "grad_norm": 0.6935935020446777,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1284455581602133,
+      "grad_norm": 0.8026931881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1300622423409588,
+      "grad_norm": 0.5210393667221069,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 6990
+    },
+    {
+      "epoch": 1.131678926521704,
+      "grad_norm": 0.60475093126297,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1332956107024492,
+      "grad_norm": 0.6417073607444763,
+      "learning_rate": 0.0002,
+      "loss": 0.7495,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1349122948831947,
+      "grad_norm": 0.6732175946235657,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 7020
+    },
+    {
+      "epoch": 1.13652897906394,
+      "grad_norm": 0.6719491481781006,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1381456632446851,
+      "grad_norm": 0.5708295106887817,
+      "learning_rate": 0.0002,
+      "loss": 0.7694,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1397623474254304,
+      "grad_norm": 0.7141719460487366,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1413790316061758,
+      "grad_norm": 0.6187017560005188,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 7060
+    },
+    {
+      "epoch": 1.142995715786921,
+      "grad_norm": 0.50581294298172,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1446123999676663,
+      "grad_norm": 0.5620143413543701,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1462290841484115,
+      "grad_norm": 0.6231929659843445,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 7090
+    },
+    {
+      "epoch": 1.147845768329157,
+      "grad_norm": 0.5775774121284485,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1494624525099022,
+      "grad_norm": 0.6492809653282166,
+      "learning_rate": 0.0002,
+      "loss": 0.7728,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1510791366906474,
+      "grad_norm": 0.6434972286224365,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 7120
+    },
+    {
+      "epoch": 1.1526958208713927,
+      "grad_norm": 0.6191812753677368,
+      "learning_rate": 0.0002,
+      "loss": 0.7374,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1543125050521381,
+      "grad_norm": 0.6690331697463989,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 7140
+    },
+    {
+      "epoch": 1.1559291892328833,
+      "grad_norm": 0.5977938175201416,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1575458734136286,
+      "grad_norm": 0.6195854544639587,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7160
+    },
+    {
+      "epoch": 1.159162557594374,
+      "grad_norm": 0.5752048492431641,
+      "learning_rate": 0.0002,
+      "loss": 0.7249,
+      "step": 7170
+    },
+    {
+      "epoch": 1.1607792417751193,
+      "grad_norm": 0.589081883430481,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1623959259558645,
+      "grad_norm": 0.756996750831604,
+      "learning_rate": 0.0002,
+      "loss": 0.704,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1640126101366097,
+      "grad_norm": 0.7614967226982117,
+      "learning_rate": 0.0002,
+      "loss": 0.7404,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1656292943173552,
+      "grad_norm": 0.6120437979698181,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1672459784981004,
+      "grad_norm": 0.6210004687309265,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1688626626788456,
+      "grad_norm": 0.6044116020202637,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7230
+    },
+    {
+      "epoch": 1.170479346859591,
+      "grad_norm": 0.5418457388877869,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1720960310403363,
+      "grad_norm": 0.6413537263870239,
+      "learning_rate": 0.0002,
+      "loss": 0.6938,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1737127152210816,
+      "grad_norm": 0.5777867436408997,
+      "learning_rate": 0.0002,
+      "loss": 0.6978,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1753293994018268,
+      "grad_norm": 0.7092402577400208,
+      "learning_rate": 0.0002,
+      "loss": 0.7503,
+      "step": 7270
+    },
+    {
+      "epoch": 1.176946083582572,
+      "grad_norm": 0.6351709365844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1785627677633175,
+      "grad_norm": 0.6172189712524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7527,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1801794519440627,
+      "grad_norm": 0.6801714897155762,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7300
+    },
+    {
+      "epoch": 1.181796136124808,
+      "grad_norm": 0.6044712066650391,
+      "learning_rate": 0.0002,
+      "loss": 0.6941,
+      "step": 7310
+    },
+    {
+      "epoch": 1.1834128203055534,
+      "grad_norm": 0.7413212060928345,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 7320
+    },
+    {
+      "epoch": 1.1850295044862986,
+      "grad_norm": 0.5303856134414673,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 7330
+    },
+    {
+      "epoch": 1.1866461886670439,
+      "grad_norm": 0.5647098422050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6915,
+      "step": 7340
+    },
+    {
+      "epoch": 1.188262872847789,
+      "grad_norm": 0.7374135255813599,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 7350
+    },
+    {
+      "epoch": 1.1898795570285345,
+      "grad_norm": 0.5710089206695557,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1914962412092798,
+      "grad_norm": 0.6073619723320007,
+      "learning_rate": 0.0002,
+      "loss": 0.8289,
+      "step": 7370
+    },
+    {
+      "epoch": 1.193112925390025,
+      "grad_norm": 0.5899916887283325,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1947296095707705,
+      "grad_norm": 0.7762434482574463,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1963462937515157,
+      "grad_norm": 0.679949939250946,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7400
+    },
+    {
+      "epoch": 1.197962977932261,
+      "grad_norm": 0.6106849312782288,
+      "learning_rate": 0.0002,
+      "loss": 0.7599,
+      "step": 7410
+    },
+    {
+      "epoch": 1.1995796621130062,
+      "grad_norm": 0.682461678981781,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2011963462937516,
+      "grad_norm": 0.6087017059326172,
+      "learning_rate": 0.0002,
+      "loss": 0.7741,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2028130304744968,
+      "grad_norm": 0.63739013671875,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 7440
+    },
+    {
+      "epoch": 1.204429714655242,
+      "grad_norm": 0.6154777407646179,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2060463988359873,
+      "grad_norm": 0.7491534948348999,
+      "learning_rate": 0.0002,
+      "loss": 0.7565,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2076630830167328,
+      "grad_norm": 0.6664797067642212,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 7470
+    },
+    {
+      "epoch": 1.209279767197478,
+      "grad_norm": 0.6660266518592834,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2108964513782232,
+      "grad_norm": 0.6972551345825195,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2125131355589684,
+      "grad_norm": 0.6157945990562439,
+      "learning_rate": 0.0002,
+      "loss": 0.7023,
+      "step": 7500
+    },
+    {
+      "epoch": 1.214129819739714,
+      "grad_norm": 0.5199310183525085,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2157465039204591,
+      "grad_norm": 0.577610433101654,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2173631881012044,
+      "grad_norm": 0.53652423620224,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2189798722819498,
+      "grad_norm": 0.6479050517082214,
+      "learning_rate": 0.0002,
+      "loss": 0.7393,
+      "step": 7540
+    },
+    {
+      "epoch": 1.220596556462695,
+      "grad_norm": 0.618748128414154,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2222132406434403,
+      "grad_norm": 0.6311424374580383,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2238299248241855,
+      "grad_norm": 0.6595825552940369,
+      "learning_rate": 0.0002,
+      "loss": 0.7272,
+      "step": 7570
+    },
+    {
+      "epoch": 1.225446609004931,
+      "grad_norm": 0.5198960900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2270632931856762,
+      "grad_norm": 0.578650712966919,
+      "learning_rate": 0.0002,
+      "loss": 0.674,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2286799773664214,
+      "grad_norm": 0.6080220937728882,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2302966615471669,
+      "grad_norm": 0.7050248384475708,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2319133457279121,
+      "grad_norm": 0.6652196049690247,
+      "learning_rate": 0.0002,
+      "loss": 0.7032,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2335300299086573,
+      "grad_norm": 0.7322776317596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7085,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2351467140894026,
+      "grad_norm": 0.4998728036880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2367633982701478,
+      "grad_norm": 0.6428788900375366,
+      "learning_rate": 0.0002,
+      "loss": 0.7214,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2383800824508933,
+      "grad_norm": 0.585242509841919,
+      "learning_rate": 0.0002,
+      "loss": 0.7699,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2399967666316385,
+      "grad_norm": 0.5211917757987976,
+      "learning_rate": 0.0002,
+      "loss": 0.7621,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2416134508123837,
+      "grad_norm": 0.6490384340286255,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2432301349931292,
+      "grad_norm": 0.6249763369560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2448468191738744,
+      "grad_norm": 0.71870356798172,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2464635033546196,
+      "grad_norm": 0.6761967539787292,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2480801875353649,
+      "grad_norm": 0.6500617265701294,
+      "learning_rate": 0.0002,
+      "loss": 0.7501,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2496968717161103,
+      "grad_norm": 0.8069869875907898,
+      "learning_rate": 0.0002,
+      "loss": 0.7903,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2513135558968556,
+      "grad_norm": 0.6044608950614929,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2529302400776008,
+      "grad_norm": 0.6573283076286316,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2545469242583462,
+      "grad_norm": 0.625430166721344,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2561636084390915,
+      "grad_norm": 0.5442022681236267,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7770
+    },
+    {
+      "epoch": 1.2577802926198367,
+      "grad_norm": 0.6818386912345886,
+      "learning_rate": 0.0002,
+      "loss": 0.7172,
+      "step": 7780
+    },
+    {
+      "epoch": 1.259396976800582,
+      "grad_norm": 0.6381874084472656,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2610136609813272,
+      "grad_norm": 0.6269212961196899,
+      "learning_rate": 0.0002,
+      "loss": 0.6834,
+      "step": 7800
+    },
+    {
+      "epoch": 1.2626303451620726,
+      "grad_norm": 0.600121259689331,
+      "learning_rate": 0.0002,
+      "loss": 0.7821,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2642470293428179,
+      "grad_norm": 0.6337703466415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2658637135235633,
+      "grad_norm": 0.7234963774681091,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 7830
+    },
+    {
+      "epoch": 1.2674803977043085,
+      "grad_norm": 0.800184965133667,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2690970818850538,
+      "grad_norm": 0.7539464831352234,
+      "learning_rate": 0.0002,
+      "loss": 0.7426,
+      "step": 7850
+    },
+    {
+      "epoch": 1.270713766065799,
+      "grad_norm": 0.5493760704994202,
+      "learning_rate": 0.0002,
+      "loss": 0.7496,
+      "step": 7860
+    },
+    {
+      "epoch": 1.2723304502465442,
+      "grad_norm": 0.7477145791053772,
+      "learning_rate": 0.0002,
+      "loss": 0.7537,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2739471344272897,
+      "grad_norm": 0.6366362571716309,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 7880
+    },
+    {
+      "epoch": 1.275563818608035,
+      "grad_norm": 0.7419533729553223,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2771805027887801,
+      "grad_norm": 0.6141223311424255,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 7900
+    },
+    {
+      "epoch": 1.2787971869695256,
+      "grad_norm": 0.7522598505020142,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 7910
+    },
+    {
+      "epoch": 1.2804138711502708,
+      "grad_norm": 0.6935804486274719,
+      "learning_rate": 0.0002,
+      "loss": 0.7097,
+      "step": 7920
+    },
+    {
+      "epoch": 1.282030555331016,
+      "grad_norm": 0.7239290475845337,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 7930
+    },
+    {
+      "epoch": 1.2836472395117613,
+      "grad_norm": 0.8800187110900879,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 7940
+    },
+    {
+      "epoch": 1.2852639236925067,
+      "grad_norm": 0.540458083152771,
+      "learning_rate": 0.0002,
+      "loss": 0.6991,
+      "step": 7950
+    },
+    {
+      "epoch": 1.286880607873252,
+      "grad_norm": 0.6492934226989746,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 7960
+    },
+    {
+      "epoch": 1.2884972920539972,
+      "grad_norm": 0.6543959379196167,
+      "learning_rate": 0.0002,
+      "loss": 0.7742,
+      "step": 7970
+    },
+    {
+      "epoch": 1.2901139762347427,
+      "grad_norm": 0.5804705619812012,
+      "learning_rate": 0.0002,
+      "loss": 0.7316,
+      "step": 7980
+    },
+    {
+      "epoch": 1.291730660415488,
+      "grad_norm": 0.7074727416038513,
+      "learning_rate": 0.0002,
+      "loss": 0.796,
+      "step": 7990
+    },
+    {
+      "epoch": 1.2933473445962331,
+      "grad_norm": 0.5347974300384521,
+      "learning_rate": 0.0002,
+      "loss": 0.7034,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2949640287769784,
+      "grad_norm": 0.6457298398017883,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 8010
+    },
+    {
+      "epoch": 1.2965807129577236,
+      "grad_norm": 0.6407219171524048,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8020
+    },
+    {
+      "epoch": 1.298197397138469,
+      "grad_norm": 0.828439474105835,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2998140813192143,
+      "grad_norm": 0.4840380549430847,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3014307654999595,
+      "grad_norm": 0.5921024680137634,
+      "learning_rate": 0.0002,
+      "loss": 0.7283,
+      "step": 8050
+    },
+    {
+      "epoch": 1.303047449680705,
+      "grad_norm": 0.6170315146446228,
+      "learning_rate": 0.0002,
+      "loss": 0.7477,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3046641338614502,
+      "grad_norm": 0.5374847054481506,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3062808180421954,
+      "grad_norm": 0.545758068561554,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3078975022229407,
+      "grad_norm": 0.55641770362854,
+      "learning_rate": 0.0002,
+      "loss": 0.7463,
+      "step": 8090
+    },
+    {
+      "epoch": 1.309514186403686,
+      "grad_norm": 0.6724897027015686,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3111308705844313,
+      "grad_norm": 0.6923972368240356,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3127475547651766,
+      "grad_norm": 0.5136841535568237,
+      "learning_rate": 0.0002,
+      "loss": 0.7149,
+      "step": 8120
+    },
+    {
+      "epoch": 1.314364238945922,
+      "grad_norm": 0.6766283512115479,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3159809231266673,
+      "grad_norm": 0.6283926367759705,
+      "learning_rate": 0.0002,
+      "loss": 0.7489,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3175976073074125,
+      "grad_norm": 0.644216001033783,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3192142914881577,
+      "grad_norm": 0.7827503085136414,
+      "learning_rate": 0.0002,
+      "loss": 0.7125,
+      "step": 8160
+    },
+    {
+      "epoch": 1.320830975668903,
+      "grad_norm": 0.6651390790939331,
+      "learning_rate": 0.0002,
+      "loss": 0.7271,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3224476598496484,
+      "grad_norm": 0.5547412633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7778,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3240643440303936,
+      "grad_norm": 0.6765179634094238,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 8190
+    },
+    {
+      "epoch": 1.325681028211139,
+      "grad_norm": 0.6822077035903931,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3272977123918843,
+      "grad_norm": 0.5941002368927002,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3289143965726296,
+      "grad_norm": 0.4850037097930908,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3305310807533748,
+      "grad_norm": 0.6162990927696228,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 8230
+    },
+    {
+      "epoch": 1.33214776493412,
+      "grad_norm": 0.6665613651275635,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3337644491148655,
+      "grad_norm": 0.618192732334137,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3353811332956107,
+      "grad_norm": 0.710418701171875,
+      "learning_rate": 0.0002,
+      "loss": 0.7532,
+      "step": 8260
+    },
+    {
+      "epoch": 1.336997817476356,
+      "grad_norm": 0.5109876990318298,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3386145016571014,
+      "grad_norm": 0.6791711449623108,
+      "learning_rate": 0.0002,
+      "loss": 0.7303,
+      "step": 8280
+    },
+    {
+      "epoch": 1.3402311858378466,
+      "grad_norm": 0.6836432814598083,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3418478700185918,
+      "grad_norm": 0.5579386353492737,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8300
+    },
+    {
+      "epoch": 1.343464554199337,
+      "grad_norm": 0.6713546514511108,
+      "learning_rate": 0.0002,
+      "loss": 0.7377,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3450812383800825,
+      "grad_norm": 0.5353720188140869,
+      "learning_rate": 0.0002,
+      "loss": 0.7756,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3466979225608278,
+      "grad_norm": 0.5813682675361633,
+      "learning_rate": 0.0002,
+      "loss": 0.718,
+      "step": 8330
+    },
+    {
+      "epoch": 1.348314606741573,
+      "grad_norm": 0.8158791661262512,
+      "learning_rate": 0.0002,
+      "loss": 0.7294,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3499312909223184,
+      "grad_norm": 0.6193785071372986,
+      "learning_rate": 0.0002,
+      "loss": 0.6992,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3515479751030637,
+      "grad_norm": 0.6353939771652222,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 8360
+    },
+    {
+      "epoch": 1.353164659283809,
+      "grad_norm": 0.6925048232078552,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 8370
+    },
+    {
+      "epoch": 1.3547813434645541,
+      "grad_norm": 0.988264799118042,
+      "learning_rate": 0.0002,
+      "loss": 0.736,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3563980276452994,
+      "grad_norm": 0.6476002931594849,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3580147118260448,
+      "grad_norm": 0.7120398879051208,
+      "learning_rate": 0.0002,
+      "loss": 0.776,
+      "step": 8400
+    },
+    {
+      "epoch": 1.35963139600679,
+      "grad_norm": 0.9048416614532471,
+      "learning_rate": 0.0002,
+      "loss": 0.7368,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3612480801875353,
+      "grad_norm": 0.7000672817230225,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3628647643682807,
+      "grad_norm": 0.6015632152557373,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 8430
+    },
+    {
+      "epoch": 1.364481448549026,
+      "grad_norm": 0.612516462802887,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3660981327297712,
+      "grad_norm": 0.5969301462173462,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 8450
+    },
+    {
+      "epoch": 1.3677148169105164,
+      "grad_norm": 0.6730654239654541,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 8460
+    },
+    {
+      "epoch": 1.369331501091262,
+      "grad_norm": 0.6386392116546631,
+      "learning_rate": 0.0002,
+      "loss": 0.7465,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3709481852720071,
+      "grad_norm": 0.739544153213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3725648694527524,
+      "grad_norm": 0.6462782621383667,
+      "learning_rate": 0.0002,
+      "loss": 0.7892,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3741815536334978,
+      "grad_norm": 0.7346843481063843,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 8500
+    },
+    {
+      "epoch": 1.375798237814243,
+      "grad_norm": 0.6884821057319641,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8510
+    },
+    {
+      "epoch": 1.3774149219949883,
+      "grad_norm": 0.6999333500862122,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3790316061757335,
+      "grad_norm": 0.5378713011741638,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 8530
+    },
+    {
+      "epoch": 1.3806482903564787,
+      "grad_norm": 0.5417906641960144,
+      "learning_rate": 0.0002,
+      "loss": 0.6797,
+      "step": 8540
+    },
+    {
+      "epoch": 1.3822649745372242,
+      "grad_norm": 0.6602526307106018,
+      "learning_rate": 0.0002,
+      "loss": 0.7499,
+      "step": 8550
+    },
+    {
+      "epoch": 1.3838816587179694,
+      "grad_norm": 0.7073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.7356,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3854983428987149,
+      "grad_norm": 0.5841707587242126,
+      "learning_rate": 0.0002,
+      "loss": 0.75,
+      "step": 8570
+    },
+    {
+      "epoch": 1.38711502707946,
+      "grad_norm": 0.7031095027923584,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 8580
+    },
+    {
+      "epoch": 1.3887317112602053,
+      "grad_norm": 0.5198570489883423,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 8590
+    },
+    {
+      "epoch": 1.3903483954409506,
+      "grad_norm": 0.7261320352554321,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 8600
+    },
+    {
+      "epoch": 1.3919650796216958,
+      "grad_norm": 0.5616350173950195,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3935817638024413,
+      "grad_norm": 0.5185914635658264,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 8620
+    },
+    {
+      "epoch": 1.3951984479831865,
+      "grad_norm": 0.5814694762229919,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 8630
+    },
+    {
+      "epoch": 1.3968151321639317,
+      "grad_norm": 0.6977371573448181,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 8640
+    },
+    {
+      "epoch": 1.3984318163446772,
+      "grad_norm": 0.6855689883232117,
+      "learning_rate": 0.0002,
+      "loss": 0.7574,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4000485005254224,
+      "grad_norm": 0.5414357781410217,
+      "learning_rate": 0.0002,
+      "loss": 0.7802,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4016651847061676,
+      "grad_norm": 0.6970012784004211,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4032818688869129,
+      "grad_norm": 0.526079535484314,
+      "learning_rate": 0.0002,
+      "loss": 0.7421,
+      "step": 8680
+    },
+    {
+      "epoch": 1.404898553067658,
+      "grad_norm": 0.758712887763977,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4065152372484035,
+      "grad_norm": 0.7118762731552124,
+      "learning_rate": 0.0002,
+      "loss": 0.7612,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4081319214291488,
+      "grad_norm": 0.5696909427642822,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4097486056098942,
+      "grad_norm": 0.7995436787605286,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4113652897906395,
+      "grad_norm": 0.7237521409988403,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4129819739713847,
+      "grad_norm": 0.744628369808197,
+      "learning_rate": 0.0002,
+      "loss": 0.7661,
+      "step": 8740
+    },
+    {
+      "epoch": 1.41459865815213,
+      "grad_norm": 0.6082926988601685,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4162153423328752,
+      "grad_norm": 0.5185243487358093,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4178320265136206,
+      "grad_norm": 0.5183082222938538,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4194487106943658,
+      "grad_norm": 0.7326041460037231,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 8780
+    },
+    {
+      "epoch": 1.421065394875111,
+      "grad_norm": 0.7174660563468933,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4226820790558565,
+      "grad_norm": 0.8080165982246399,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4242987632366018,
+      "grad_norm": 0.5061507821083069,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 8810
+    },
+    {
+      "epoch": 1.425915447417347,
+      "grad_norm": 0.801602840423584,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4275321315980922,
+      "grad_norm": 0.6150273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4291488157788377,
+      "grad_norm": 0.8786525726318359,
+      "learning_rate": 0.0002,
+      "loss": 0.7452,
+      "step": 8840
+    },
+    {
+      "epoch": 1.430765499959583,
+      "grad_norm": 0.6371538639068604,
+      "learning_rate": 0.0002,
+      "loss": 0.7257,
+      "step": 8850
+    },
+    {
+      "epoch": 1.4323821841403281,
+      "grad_norm": 0.6409295797348022,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4339988683210736,
+      "grad_norm": 0.6452359557151794,
+      "learning_rate": 0.0002,
+      "loss": 0.7891,
+      "step": 8870
+    },
+    {
+      "epoch": 1.4356155525018188,
+      "grad_norm": 0.5842334628105164,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 8880
+    },
+    {
+      "epoch": 1.437232236682564,
+      "grad_norm": 0.696761965751648,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4388489208633093,
+      "grad_norm": 0.6384600400924683,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4404656050440545,
+      "grad_norm": 0.5981136560440063,
+      "learning_rate": 0.0002,
+      "loss": 0.7049,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4420822892248,
+      "grad_norm": 0.6355637907981873,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4436989734055452,
+      "grad_norm": 0.6374830603599548,
+      "learning_rate": 0.0002,
+      "loss": 0.7653,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4453156575862904,
+      "grad_norm": 0.559013307094574,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 8940
+    },
+    {
+      "epoch": 1.446932341767036,
+      "grad_norm": 0.7289170026779175,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4485490259477811,
+      "grad_norm": 0.8649206757545471,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4501657101285264,
+      "grad_norm": 0.7664689421653748,
+      "learning_rate": 0.0002,
+      "loss": 0.7745,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4517823943092716,
+      "grad_norm": 0.7109952569007874,
+      "learning_rate": 0.0002,
+      "loss": 0.7431,
+      "step": 8980
+    },
+    {
+      "epoch": 1.453399078490017,
+      "grad_norm": 0.6312844753265381,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4550157626707623,
+      "grad_norm": 0.6616617441177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7467,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4566324468515075,
+      "grad_norm": 0.7384068965911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7518,
+      "step": 9010
+    },
+    {
+      "epoch": 1.458249131032253,
+      "grad_norm": 0.6549670100212097,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9020
+    },
+    {
+      "epoch": 1.4598658152129982,
+      "grad_norm": 0.6254119277000427,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4614824993937434,
+      "grad_norm": 0.6806328892707825,
+      "learning_rate": 0.0002,
+      "loss": 0.7645,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4630991835744886,
+      "grad_norm": 0.6803115010261536,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4647158677552339,
+      "grad_norm": 0.48529282212257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7264,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4663325519359793,
+      "grad_norm": 0.5995030999183655,
+      "learning_rate": 0.0002,
+      "loss": 0.7542,
+      "step": 9070
+    },
+    {
+      "epoch": 1.4679492361167246,
+      "grad_norm": 0.6005427837371826,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9080
+    },
+    {
+      "epoch": 1.46956592029747,
+      "grad_norm": 0.718564510345459,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4711826044782153,
+      "grad_norm": 0.7003577351570129,
+      "learning_rate": 0.0002,
+      "loss": 0.7089,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4727992886589605,
+      "grad_norm": 0.5888323783874512,
+      "learning_rate": 0.0002,
+      "loss": 0.8069,
+      "step": 9110
+    },
+    {
+      "epoch": 1.4744159728397057,
+      "grad_norm": 0.6417609453201294,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 9120
+    },
+    {
+      "epoch": 1.476032657020451,
+      "grad_norm": 0.572294294834137,
+      "learning_rate": 0.0002,
+      "loss": 0.7441,
+      "step": 9130
+    },
+    {
+      "epoch": 1.4776493412011964,
+      "grad_norm": 0.8200714588165283,
+      "learning_rate": 0.0002,
+      "loss": 0.8053,
+      "step": 9140
+    },
+    {
+      "epoch": 1.4792660253819416,
+      "grad_norm": 0.6343288421630859,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4808827095626869,
+      "grad_norm": 0.7017961144447327,
+      "learning_rate": 0.0002,
+      "loss": 0.7641,
+      "step": 9160
+    },
+    {
+      "epoch": 1.4824993937434323,
+      "grad_norm": 0.6202912926673889,
+      "learning_rate": 0.0002,
+      "loss": 0.7619,
+      "step": 9170
+    },
+    {
+      "epoch": 1.4841160779241775,
+      "grad_norm": 0.6677869558334351,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 9180
+    },
+    {
+      "epoch": 1.4857327621049228,
+      "grad_norm": 0.6052267551422119,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 9190
+    },
+    {
+      "epoch": 1.487349446285668,
+      "grad_norm": 0.6638872027397156,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 9200
+    },
+    {
+      "epoch": 1.4889661304664135,
+      "grad_norm": 0.6245523691177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 9210
+    },
+    {
+      "epoch": 1.4905828146471587,
+      "grad_norm": 0.5761767625808716,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 9220
+    },
+    {
+      "epoch": 1.492199498827904,
+      "grad_norm": 0.8175981640815735,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 9230
+    },
+    {
+      "epoch": 1.4938161830086494,
+      "grad_norm": 0.9144009947776794,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 9240
+    },
+    {
+      "epoch": 1.4954328671893946,
+      "grad_norm": 0.5742552876472473,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 9250
+    },
+    {
+      "epoch": 1.4970495513701398,
+      "grad_norm": 0.534534215927124,
+      "learning_rate": 0.0002,
+      "loss": 0.7418,
+      "step": 9260
+    },
+    {
+      "epoch": 1.498666235550885,
+      "grad_norm": 0.7836225032806396,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5002829197316303,
+      "grad_norm": 0.5292993187904358,
+      "learning_rate": 0.0002,
+      "loss": 0.7453,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5018996039123758,
+      "grad_norm": 0.8044071793556213,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9290
+    },
+    {
+      "epoch": 1.503516288093121,
+      "grad_norm": 0.6185805201530457,
+      "learning_rate": 0.0002,
+      "loss": 0.7229,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5051329722738664,
+      "grad_norm": 0.6093607544898987,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5067496564546117,
+      "grad_norm": 0.5891730189323425,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 9320
+    },
+    {
+      "epoch": 1.508366340635357,
+      "grad_norm": 0.6331129670143127,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5099830248161021,
+      "grad_norm": 0.7690958380699158,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5115997089968474,
+      "grad_norm": 0.6548877358436584,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5132163931775926,
+      "grad_norm": 0.6545143127441406,
+      "learning_rate": 0.0002,
+      "loss": 0.7408,
+      "step": 9360
+    },
+    {
+      "epoch": 1.514833077358338,
+      "grad_norm": 0.553247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.7432,
+      "step": 9370
+    },
+    {
+      "epoch": 1.5164497615390833,
+      "grad_norm": 0.8145074844360352,
+      "learning_rate": 0.0002,
+      "loss": 0.7265,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5180664457198287,
+      "grad_norm": 0.7636994123458862,
+      "learning_rate": 0.0002,
+      "loss": 0.7379,
+      "step": 9390
+    },
+    {
+      "epoch": 1.519683129900574,
+      "grad_norm": 0.6838982701301575,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5212998140813192,
+      "grad_norm": 0.8599441647529602,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5229164982620644,
+      "grad_norm": 0.7020329833030701,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5245331824428097,
+      "grad_norm": 0.6964772343635559,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5261498666235551,
+      "grad_norm": 0.6916600465774536,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9440
+    },
+    {
+      "epoch": 1.5277665508043003,
+      "grad_norm": 0.7282621264457703,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5293832349850458,
+      "grad_norm": 0.5363983511924744,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 9460
+    },
+    {
+      "epoch": 1.530999919165791,
+      "grad_norm": 0.6184861063957214,
+      "learning_rate": 0.0002,
+      "loss": 0.7154,
+      "step": 9470
+    },
+    {
+      "epoch": 1.5326166033465363,
+      "grad_norm": 0.5991285443305969,
+      "learning_rate": 0.0002,
+      "loss": 0.7837,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5342332875272815,
+      "grad_norm": 0.8176587820053101,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5358499717080267,
+      "grad_norm": 0.6473721861839294,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 9500
+    },
+    {
+      "epoch": 1.5374666558887722,
+      "grad_norm": 0.7319952845573425,
+      "learning_rate": 0.0002,
+      "loss": 0.7632,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5390833400695174,
+      "grad_norm": 0.702900230884552,
+      "learning_rate": 0.0002,
+      "loss": 0.7706,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5407000242502629,
+      "grad_norm": 0.7971600294113159,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 9530
+    },
+    {
+      "epoch": 1.542316708431008,
+      "grad_norm": 0.6527525186538696,
+      "learning_rate": 0.0002,
+      "loss": 0.7352,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5439333926117533,
+      "grad_norm": 0.5791676044464111,
+      "learning_rate": 0.0002,
+      "loss": 0.7425,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5455500767924986,
+      "grad_norm": 0.5619390606880188,
+      "learning_rate": 0.0002,
+      "loss": 0.7585,
+      "step": 9560
+    },
+    {
+      "epoch": 1.5471667609732438,
+      "grad_norm": 0.5701689124107361,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9570
+    },
+    {
+      "epoch": 1.548783445153989,
+      "grad_norm": 0.47549352049827576,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5504001293347345,
+      "grad_norm": 0.8730611205101013,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5520168135154797,
+      "grad_norm": 0.6842091083526611,
+      "learning_rate": 0.0002,
+      "loss": 0.798,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5536334976962252,
+      "grad_norm": 0.6675129532814026,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5552501818769704,
+      "grad_norm": 0.8173956274986267,
+      "learning_rate": 0.0002,
+      "loss": 0.7954,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5568668660577156,
+      "grad_norm": 0.724947452545166,
+      "learning_rate": 0.0002,
+      "loss": 0.7535,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5584835502384609,
+      "grad_norm": 0.6154758930206299,
+      "learning_rate": 0.0002,
+      "loss": 0.7738,
+      "step": 9640
+    },
+    {
+      "epoch": 1.560100234419206,
+      "grad_norm": 0.6072008013725281,
+      "learning_rate": 0.0002,
+      "loss": 0.7568,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5617169185999515,
+      "grad_norm": 0.659010648727417,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 9660
+    },
+    {
+      "epoch": 1.5633336027806968,
+      "grad_norm": 0.65857994556427,
+      "learning_rate": 0.0002,
+      "loss": 0.673,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5649502869614422,
+      "grad_norm": 0.5914267301559448,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5665669711421875,
+      "grad_norm": 0.6248020529747009,
+      "learning_rate": 0.0002,
+      "loss": 0.7414,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5681836553229327,
+      "grad_norm": 0.7147795557975769,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 9700
+    },
+    {
+      "epoch": 1.569800339503678,
+      "grad_norm": 0.7076232433319092,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 9710
+    },
+    {
+      "epoch": 1.5714170236844232,
+      "grad_norm": 0.6217400431632996,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5730337078651684,
+      "grad_norm": 0.6709911227226257,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 9730
+    },
+    {
+      "epoch": 1.5746503920459138,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 9740
+    },
+    {
+      "epoch": 1.576267076226659,
+      "grad_norm": 0.6241145730018616,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 9750
+    },
+    {
+      "epoch": 1.5778837604074045,
+      "grad_norm": 0.4960934817790985,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 9760
+    },
+    {
+      "epoch": 1.5795004445881498,
+      "grad_norm": 0.6593309640884399,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 9770
+    },
+    {
+      "epoch": 1.581117128768895,
+      "grad_norm": 0.5814042091369629,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 9780
+    },
+    {
+      "epoch": 1.5827338129496402,
+      "grad_norm": 0.5936070680618286,
+      "learning_rate": 0.0002,
+      "loss": 0.7109,
+      "step": 9790
+    },
+    {
+      "epoch": 1.5843504971303854,
+      "grad_norm": 0.6454403400421143,
+      "learning_rate": 0.0002,
+      "loss": 0.7769,
+      "step": 9800
+    },
+    {
+      "epoch": 1.585967181311131,
+      "grad_norm": 0.7612107992172241,
+      "learning_rate": 0.0002,
+      "loss": 0.7677,
+      "step": 9810
+    },
+    {
+      "epoch": 1.5875838654918761,
+      "grad_norm": 0.6494482755661011,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5892005496726216,
+      "grad_norm": 0.7825694680213928,
+      "learning_rate": 0.0002,
+      "loss": 0.7569,
+      "step": 9830
+    },
+    {
+      "epoch": 1.5908172338533668,
+      "grad_norm": 0.6757757663726807,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 9840
+    },
+    {
+      "epoch": 1.592433918034112,
+      "grad_norm": 0.7105609178543091,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 9850
+    },
+    {
+      "epoch": 1.5940506022148573,
+      "grad_norm": 0.7596991062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 9860
+    },
+    {
+      "epoch": 1.5956672863956025,
+      "grad_norm": 0.5681525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7108,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5972839705763477,
+      "grad_norm": 0.6090980768203735,
+      "learning_rate": 0.0002,
+      "loss": 0.7811,
+      "step": 9880
+    },
+    {
+      "epoch": 1.5989006547570932,
+      "grad_norm": 0.6271613240242004,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6005173389378387,
+      "grad_norm": 0.7656369805335999,
+      "learning_rate": 0.0002,
+      "loss": 0.7419,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6021340231185839,
+      "grad_norm": 0.7504446506500244,
+      "learning_rate": 0.0002,
+      "loss": 0.7336,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6037507072993291,
+      "grad_norm": 0.659656286239624,
+      "learning_rate": 0.0002,
+      "loss": 0.7479,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6053673914800743,
+      "grad_norm": 0.6006826162338257,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6069840756608196,
+      "grad_norm": 0.7872757911682129,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6086007598415648,
+      "grad_norm": 0.5545852780342102,
+      "learning_rate": 0.0002,
+      "loss": 0.768,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6102174440223103,
+      "grad_norm": 0.7429468631744385,
+      "learning_rate": 0.0002,
+      "loss": 0.8064,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6118341282030555,
+      "grad_norm": 0.6873556971549988,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 9970
+    },
+    {
+      "epoch": 1.613450812383801,
+      "grad_norm": 0.5874287486076355,
+      "learning_rate": 0.0002,
+      "loss": 0.7324,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6150674965645462,
+      "grad_norm": 0.6039386987686157,
+      "learning_rate": 0.0002,
+      "loss": 0.7141,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6166841807452914,
+      "grad_norm": 0.6233575940132141,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6183008649260366,
+      "grad_norm": 0.7676448225975037,
+      "learning_rate": 0.0002,
+      "loss": 0.7602,
+      "step": 10010
+    },
+    {
+      "epoch": 1.6199175491067819,
+      "grad_norm": 0.6565698385238647,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6215342332875273,
+      "grad_norm": 0.6787590384483337,
+      "learning_rate": 0.0002,
+      "loss": 0.7104,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6231509174682726,
+      "grad_norm": 0.6137678027153015,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 10040
+    },
+    {
+      "epoch": 1.624767601649018,
+      "grad_norm": 0.5236800312995911,
+      "learning_rate": 0.0002,
+      "loss": 0.7646,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6263842858297632,
+      "grad_norm": 0.7626367807388306,
+      "learning_rate": 0.0002,
+      "loss": 0.7437,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6280009700105085,
+      "grad_norm": 0.5657260417938232,
+      "learning_rate": 0.0002,
+      "loss": 0.7273,
+      "step": 10070
+    },
+    {
+      "epoch": 1.6296176541912537,
+      "grad_norm": 0.4913991391658783,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 10080
+    },
+    {
+      "epoch": 1.631234338371999,
+      "grad_norm": 0.7715556621551514,
+      "learning_rate": 0.0002,
+      "loss": 0.7596,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6328510225527442,
+      "grad_norm": 0.6509000062942505,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6344677067334896,
+      "grad_norm": 0.6215850114822388,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6360843909142349,
+      "grad_norm": 0.6956844329833984,
+      "learning_rate": 0.0002,
+      "loss": 0.7705,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6377010750949803,
+      "grad_norm": 0.6111597418785095,
+      "learning_rate": 0.0002,
+      "loss": 0.7129,
+      "step": 10130
+    },
+    {
+      "epoch": 1.6393177592757255,
+      "grad_norm": 0.6518288850784302,
+      "learning_rate": 0.0002,
+      "loss": 0.6955,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6409344434564708,
+      "grad_norm": 0.6914522051811218,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10150
+    },
+    {
+      "epoch": 1.642551127637216,
+      "grad_norm": 0.63785719871521,
+      "learning_rate": 0.0002,
+      "loss": 0.7295,
+      "step": 10160
+    },
+    {
+      "epoch": 1.6441678118179612,
+      "grad_norm": 0.6379287838935852,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6457844959987067,
+      "grad_norm": 0.6793403029441833,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 10180
+    },
+    {
+      "epoch": 1.647401180179452,
+      "grad_norm": 0.6099132895469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6490178643601974,
+      "grad_norm": 0.5869854092597961,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6506345485409426,
+      "grad_norm": 0.7716999053955078,
+      "learning_rate": 0.0002,
+      "loss": 0.8308,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6522512327216878,
+      "grad_norm": 0.6854110360145569,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 10220
+    },
+    {
+      "epoch": 1.653867916902433,
+      "grad_norm": 0.6957170367240906,
+      "learning_rate": 0.0002,
+      "loss": 0.782,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6554846010831783,
+      "grad_norm": 0.6932903528213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6571012852639235,
+      "grad_norm": 0.7713165283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10250
+    },
+    {
+      "epoch": 1.658717969444669,
+      "grad_norm": 0.7455793619155884,
+      "learning_rate": 0.0002,
+      "loss": 0.7099,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6603346536254144,
+      "grad_norm": 0.5464168190956116,
+      "learning_rate": 0.0002,
+      "loss": 0.7524,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6619513378061597,
+      "grad_norm": 0.6782926321029663,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 10280
+    },
+    {
+      "epoch": 1.663568021986905,
+      "grad_norm": 0.7962649464607239,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6651847061676501,
+      "grad_norm": 0.6814526319503784,
+      "learning_rate": 0.0002,
+      "loss": 0.7142,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6668013903483954,
+      "grad_norm": 0.656895101070404,
+      "learning_rate": 0.0002,
+      "loss": 0.7285,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6684180745291406,
+      "grad_norm": 0.6085672378540039,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 10320
+    },
+    {
+      "epoch": 1.670034758709886,
+      "grad_norm": 0.585508406162262,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 10330
+    },
+    {
+      "epoch": 1.6716514428906313,
+      "grad_norm": 0.6930184364318848,
+      "learning_rate": 0.0002,
+      "loss": 0.7604,
+      "step": 10340
+    },
+    {
+      "epoch": 1.6732681270713767,
+      "grad_norm": 0.575663149356842,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10350
+    },
+    {
+      "epoch": 1.674884811252122,
+      "grad_norm": 0.582502543926239,
+      "learning_rate": 0.0002,
+      "loss": 0.7198,
+      "step": 10360
+    },
+    {
+      "epoch": 1.6765014954328672,
+      "grad_norm": 0.5668916702270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 10370
+    },
+    {
+      "epoch": 1.6781181796136124,
+      "grad_norm": 0.6070065498352051,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10380
+    },
+    {
+      "epoch": 1.6797348637943577,
+      "grad_norm": 0.6141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.7939,
+      "step": 10390
+    },
+    {
+      "epoch": 1.6813515479751031,
+      "grad_norm": 0.8359124064445496,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 10400
+    },
+    {
+      "epoch": 1.6829682321558483,
+      "grad_norm": 0.5378185510635376,
+      "learning_rate": 0.0002,
+      "loss": 0.7488,
+      "step": 10410
+    },
+    {
+      "epoch": 1.6845849163365938,
+      "grad_norm": 0.6959536075592041,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 10420
+    },
+    {
+      "epoch": 1.686201600517339,
+      "grad_norm": 0.6514357328414917,
+      "learning_rate": 0.0002,
+      "loss": 0.7872,
+      "step": 10430
+    },
+    {
+      "epoch": 1.6878182846980843,
+      "grad_norm": 0.7706646919250488,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10440
+    },
+    {
+      "epoch": 1.6894349688788295,
+      "grad_norm": 0.6183337569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.7673,
+      "step": 10450
+    },
+    {
+      "epoch": 1.6910516530595747,
+      "grad_norm": 0.6123278141021729,
+      "learning_rate": 0.0002,
+      "loss": 0.7566,
+      "step": 10460
+    },
+    {
+      "epoch": 1.69266833724032,
+      "grad_norm": 0.6894851326942444,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10470
+    },
+    {
+      "epoch": 1.6942850214210654,
+      "grad_norm": 0.7497312426567078,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 10480
+    },
+    {
+      "epoch": 1.6959017056018106,
+      "grad_norm": 0.5968214273452759,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10490
+    },
+    {
+      "epoch": 1.697518389782556,
+      "grad_norm": 0.6747927069664001,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 10500
+    },
+    {
+      "epoch": 1.6991350739633013,
+      "grad_norm": 0.5708310008049011,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7007517581440466,
+      "grad_norm": 0.606526792049408,
+      "learning_rate": 0.0002,
+      "loss": 0.7385,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7023684423247918,
+      "grad_norm": 0.662011981010437,
+      "learning_rate": 0.0002,
+      "loss": 0.7204,
+      "step": 10530
+    },
+    {
+      "epoch": 1.703985126505537,
+      "grad_norm": 0.7583045363426208,
+      "learning_rate": 0.0002,
+      "loss": 0.7999,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7056018106862825,
+      "grad_norm": 0.721632182598114,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7072184948670277,
+      "grad_norm": 0.6107715368270874,
+      "learning_rate": 0.0002,
+      "loss": 0.7407,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7088351790477732,
+      "grad_norm": 0.6652471423149109,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7104518632285184,
+      "grad_norm": 0.6308087110519409,
+      "learning_rate": 0.0002,
+      "loss": 0.7767,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7120685474092636,
+      "grad_norm": 0.5464386940002441,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7136852315900089,
+      "grad_norm": 0.6558911204338074,
+      "learning_rate": 0.0002,
+      "loss": 0.7063,
+      "step": 10600
+    },
+    {
+      "epoch": 1.715301915770754,
+      "grad_norm": 0.5665024518966675,
+      "learning_rate": 0.0002,
+      "loss": 0.7126,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7169185999514993,
+      "grad_norm": 0.7888094186782837,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7185352841322448,
+      "grad_norm": 0.7084909081459045,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 10630
+    },
+    {
+      "epoch": 1.7201519683129902,
+      "grad_norm": 0.7982324361801147,
+      "learning_rate": 0.0002,
+      "loss": 0.7557,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7217686524937355,
+      "grad_norm": 0.6418732404708862,
+      "learning_rate": 0.0002,
+      "loss": 0.7345,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7233853366744807,
+      "grad_norm": 0.7636681795120239,
+      "learning_rate": 0.0002,
+      "loss": 0.7734,
+      "step": 10660
+    },
+    {
+      "epoch": 1.725002020855226,
+      "grad_norm": 0.5646875500679016,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7266187050359711,
+      "grad_norm": 0.5231260657310486,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7282353892167164,
+      "grad_norm": 0.7635011672973633,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7298520733974618,
+      "grad_norm": 0.7518259286880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 10700
+    },
+    {
+      "epoch": 1.731468757578207,
+      "grad_norm": 0.7295602560043335,
+      "learning_rate": 0.0002,
+      "loss": 0.751,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7330854417589525,
+      "grad_norm": 0.6984632015228271,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7347021259396977,
+      "grad_norm": 0.6198219060897827,
+      "learning_rate": 0.0002,
+      "loss": 0.7921,
+      "step": 10730
+    },
+    {
+      "epoch": 1.736318810120443,
+      "grad_norm": 0.6957576274871826,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7379354943011882,
+      "grad_norm": 0.6430263519287109,
+      "learning_rate": 0.0002,
+      "loss": 0.7917,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7395521784819334,
+      "grad_norm": 0.6134995222091675,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 10760
+    },
+    {
+      "epoch": 1.741168862662679,
+      "grad_norm": 0.7209452986717224,
+      "learning_rate": 0.0002,
+      "loss": 0.7584,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7427855468434241,
+      "grad_norm": 0.6735447645187378,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7444022310241696,
+      "grad_norm": 0.5605693459510803,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 10790
+    },
+    {
+      "epoch": 1.7460189152049148,
+      "grad_norm": 0.6882363557815552,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 10800
+    },
+    {
+      "epoch": 1.74763559938566,
+      "grad_norm": 0.6386259198188782,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7492522835664053,
+      "grad_norm": 0.6529015302658081,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7508689677471505,
+      "grad_norm": 0.5664082765579224,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7524856519278957,
+      "grad_norm": 0.7532684206962585,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7541023361086412,
+      "grad_norm": 0.77171391248703,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7557190202893864,
+      "grad_norm": 0.7255431413650513,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7573357044701319,
+      "grad_norm": 0.763083279132843,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10870
+    },
+    {
+      "epoch": 1.758952388650877,
+      "grad_norm": 0.6042402982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7605690728316223,
+      "grad_norm": 0.7642518281936646,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7621857570123676,
+      "grad_norm": 0.6347904801368713,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7638024411931128,
+      "grad_norm": 0.5371627807617188,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7654191253738583,
+      "grad_norm": 0.6840225458145142,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7670358095546035,
+      "grad_norm": 0.5288469195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10930
+    },
+    {
+      "epoch": 1.768652493735349,
+      "grad_norm": 0.69020676612854,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 10940
+    },
+    {
+      "epoch": 1.7702691779160942,
+      "grad_norm": 0.5943242311477661,
+      "learning_rate": 0.0002,
+      "loss": 0.7468,
+      "step": 10950
+    },
+    {
+      "epoch": 1.7718858620968394,
+      "grad_norm": 0.5616418123245239,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 10960
+    },
+    {
+      "epoch": 1.7735025462775846,
+      "grad_norm": 0.7209470868110657,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 10970
+    },
+    {
+      "epoch": 1.7751192304583299,
+      "grad_norm": 0.6657957434654236,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 10980
+    },
+    {
+      "epoch": 1.776735914639075,
+      "grad_norm": 0.6469064950942993,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 10990
+    },
+    {
+      "epoch": 1.7783525988198206,
+      "grad_norm": 0.6615678071975708,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 11000
+    },
+    {
+      "epoch": 1.779969283000566,
+      "grad_norm": 0.6722439527511597,
+      "learning_rate": 0.0002,
+      "loss": 0.747,
+      "step": 11010
+    },
+    {
+      "epoch": 1.7815859671813112,
+      "grad_norm": 0.634136974811554,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 11020
+    },
+    {
+      "epoch": 1.7832026513620565,
+      "grad_norm": 0.6024377346038818,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 11030
+    },
+    {
+      "epoch": 1.7848193355428017,
+      "grad_norm": 0.6909403800964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 11040
+    },
+    {
+      "epoch": 1.786436019723547,
+      "grad_norm": 0.7148767709732056,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 11050
+    },
+    {
+      "epoch": 1.7880527039042922,
+      "grad_norm": 0.7442979216575623,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 11060
+    },
+    {
+      "epoch": 1.7896693880850376,
+      "grad_norm": 0.6830431818962097,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 11070
+    },
+    {
+      "epoch": 1.7912860722657828,
+      "grad_norm": 0.9172667264938354,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 11080
+    },
+    {
+      "epoch": 1.7929027564465283,
+      "grad_norm": 0.6799490451812744,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 11090
+    },
+    {
+      "epoch": 1.7945194406272735,
+      "grad_norm": 0.7617024779319763,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 11100
+    },
+    {
+      "epoch": 1.7961361248080188,
+      "grad_norm": 0.7701810002326965,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 11110
+    },
+    {
+      "epoch": 1.797752808988764,
+      "grad_norm": 0.7454385757446289,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11120
+    },
+    {
+      "epoch": 1.7993694931695092,
+      "grad_norm": 0.6121436953544617,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8009861773502547,
+      "grad_norm": 0.6237571835517883,
+      "learning_rate": 0.0002,
+      "loss": 0.7305,
+      "step": 11140
+    },
+    {
+      "epoch": 1.802602861531,
+      "grad_norm": 0.6818515658378601,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 11150
+    },
+    {
+      "epoch": 1.8042195457117454,
+      "grad_norm": 0.7768308520317078,
+      "learning_rate": 0.0002,
+      "loss": 0.6876,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8058362298924906,
+      "grad_norm": 0.6875537633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 11170
+    },
+    {
+      "epoch": 1.8074529140732358,
+      "grad_norm": 0.7950584888458252,
+      "learning_rate": 0.0002,
+      "loss": 0.761,
+      "step": 11180
+    },
+    {
+      "epoch": 1.809069598253981,
+      "grad_norm": 0.8210248351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8106862824347263,
+      "grad_norm": 0.6674110889434814,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8123029666154715,
+      "grad_norm": 0.6261674761772156,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 11210
+    },
+    {
+      "epoch": 1.813919650796217,
+      "grad_norm": 0.6484741568565369,
+      "learning_rate": 0.0002,
+      "loss": 0.7122,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8155363349769622,
+      "grad_norm": 0.6231244206428528,
+      "learning_rate": 0.0002,
+      "loss": 0.7718,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8171530191577077,
+      "grad_norm": 0.7243146896362305,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11240
+    },
+    {
+      "epoch": 1.818769703338453,
+      "grad_norm": 0.6776193380355835,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8203863875191981,
+      "grad_norm": 0.5973618030548096,
+      "learning_rate": 0.0002,
+      "loss": 0.7317,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8220030716999434,
+      "grad_norm": 0.6451361179351807,
+      "learning_rate": 0.0002,
+      "loss": 0.7961,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8236197558806886,
+      "grad_norm": 0.5963068008422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 11280
+    },
+    {
+      "epoch": 1.825236440061434,
+      "grad_norm": 0.536902129650116,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8268531242421793,
+      "grad_norm": 0.6993787288665771,
+      "learning_rate": 0.0002,
+      "loss": 0.708,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8284698084229247,
+      "grad_norm": 0.6135255098342896,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11310
+    },
+    {
+      "epoch": 1.83008649260367,
+      "grad_norm": 0.6057423949241638,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8317031767844152,
+      "grad_norm": 0.6598812341690063,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8333198609651604,
+      "grad_norm": 0.6075948476791382,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 11340
+    },
+    {
+      "epoch": 1.8349365451459057,
+      "grad_norm": 0.7065447568893433,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8365532293266509,
+      "grad_norm": 0.680526614189148,
+      "learning_rate": 0.0002,
+      "loss": 0.7365,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8381699135073963,
+      "grad_norm": 0.6356695294380188,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8397865976881416,
+      "grad_norm": 0.6399052143096924,
+      "learning_rate": 0.0002,
+      "loss": 0.721,
+      "step": 11380
+    },
+    {
+      "epoch": 1.841403281868887,
+      "grad_norm": 0.6125704050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 11390
+    },
+    {
+      "epoch": 1.8430199660496323,
+      "grad_norm": 0.7124643325805664,
+      "learning_rate": 0.0002,
+      "loss": 0.755,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8446366502303775,
+      "grad_norm": 0.6099604964256287,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8462533344111227,
+      "grad_norm": 0.7338208556175232,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 11420
+    },
+    {
+      "epoch": 1.847870018591868,
+      "grad_norm": 0.7534668445587158,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8494867027726134,
+      "grad_norm": 0.6135470271110535,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8511033869533586,
+      "grad_norm": 0.6229309439659119,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 11450
+    },
+    {
+      "epoch": 1.852720071134104,
+      "grad_norm": 0.706423282623291,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 11460
+    },
+    {
+      "epoch": 1.8543367553148493,
+      "grad_norm": 0.5460049510002136,
+      "learning_rate": 0.0002,
+      "loss": 0.7411,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8559534394955945,
+      "grad_norm": 0.6616711020469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7416,
+      "step": 11480
+    },
+    {
+      "epoch": 1.8575701236763398,
+      "grad_norm": 0.6372783184051514,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 11490
+    },
+    {
+      "epoch": 1.859186807857085,
+      "grad_norm": 0.7162668108940125,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8608034920378305,
+      "grad_norm": 0.6605209708213806,
+      "learning_rate": 0.0002,
+      "loss": 0.7747,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8624201762185757,
+      "grad_norm": 0.6933956742286682,
+      "learning_rate": 0.0002,
+      "loss": 0.7258,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8640368603993211,
+      "grad_norm": 0.6582090854644775,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8656535445800664,
+      "grad_norm": 0.6416500806808472,
+      "learning_rate": 0.0002,
+      "loss": 0.7313,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8672702287608116,
+      "grad_norm": 0.5434312224388123,
+      "learning_rate": 0.0002,
+      "loss": 0.7372,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8688869129415568,
+      "grad_norm": 0.6827567219734192,
+      "learning_rate": 0.0002,
+      "loss": 0.7635,
+      "step": 11560
+    },
+    {
+      "epoch": 1.870503597122302,
+      "grad_norm": 0.7354370951652527,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 11570
+    },
+    {
+      "epoch": 1.8721202813030473,
+      "grad_norm": 0.590372622013092,
+      "learning_rate": 0.0002,
+      "loss": 0.7526,
+      "step": 11580
+    },
+    {
+      "epoch": 1.8737369654837928,
+      "grad_norm": 0.853183925151825,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 11590
+    },
+    {
+      "epoch": 1.875353649664538,
+      "grad_norm": 0.822678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 11600
+    },
+    {
+      "epoch": 1.8769703338452834,
+      "grad_norm": 0.6591550707817078,
+      "learning_rate": 0.0002,
+      "loss": 0.7427,
+      "step": 11610
+    },
+    {
+      "epoch": 1.8785870180260287,
+      "grad_norm": 0.7475301623344421,
+      "learning_rate": 0.0002,
+      "loss": 0.7054,
+      "step": 11620
+    },
+    {
+      "epoch": 1.880203702206774,
+      "grad_norm": 0.6390765309333801,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 11630
+    },
+    {
+      "epoch": 1.8818203863875191,
+      "grad_norm": 0.6589758992195129,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 11640
+    },
+    {
+      "epoch": 1.8834370705682644,
+      "grad_norm": 0.6765508651733398,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 11650
+    },
+    {
+      "epoch": 1.8850537547490098,
+      "grad_norm": 0.6527857780456543,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 11660
+    },
+    {
+      "epoch": 1.886670438929755,
+      "grad_norm": 0.6642923951148987,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 11670
+    },
+    {
+      "epoch": 1.8882871231105005,
+      "grad_norm": 0.6945584416389465,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 11680
+    },
+    {
+      "epoch": 1.8899038072912457,
+      "grad_norm": 0.694018542766571,
+      "learning_rate": 0.0002,
+      "loss": 0.7711,
+      "step": 11690
+    },
+    {
+      "epoch": 1.891520491471991,
+      "grad_norm": 0.7237417101860046,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 11700
+    },
+    {
+      "epoch": 1.8931371756527362,
+      "grad_norm": 0.7401309609413147,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 11710
+    },
+    {
+      "epoch": 1.8947538598334814,
+      "grad_norm": 0.6537784337997437,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 11720
+    },
+    {
+      "epoch": 1.8963705440142267,
+      "grad_norm": 0.7398539185523987,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 11730
+    },
+    {
+      "epoch": 1.8979872281949721,
+      "grad_norm": 0.6696075797080994,
+      "learning_rate": 0.0002,
+      "loss": 0.7561,
+      "step": 11740
+    },
+    {
+      "epoch": 1.8996039123757174,
+      "grad_norm": 0.6014142036437988,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9012205965564628,
+      "grad_norm": 0.7023524641990662,
+      "learning_rate": 0.0002,
+      "loss": 0.7714,
+      "step": 11760
+    },
+    {
+      "epoch": 1.902837280737208,
+      "grad_norm": 0.739973783493042,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9044539649179533,
+      "grad_norm": 0.5576770901679993,
+      "learning_rate": 0.0002,
+      "loss": 0.7848,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9060706490986985,
+      "grad_norm": 0.6907393932342529,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 11790
+    },
+    {
+      "epoch": 1.9076873332794437,
+      "grad_norm": 0.6934581995010376,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9093040174601892,
+      "grad_norm": 0.591774582862854,
+      "learning_rate": 0.0002,
+      "loss": 0.7199,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9109207016409344,
+      "grad_norm": 0.6249791383743286,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11820
+    },
+    {
+      "epoch": 1.9125373858216799,
+      "grad_norm": 0.6755744218826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 11830
+    },
+    {
+      "epoch": 1.914154070002425,
+      "grad_norm": 0.7286285161972046,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9157707541831703,
+      "grad_norm": 0.7867850065231323,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9173874383639156,
+      "grad_norm": 0.6283972859382629,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11860
+    },
+    {
+      "epoch": 1.9190041225446608,
+      "grad_norm": 0.605823814868927,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 11870
+    },
+    {
+      "epoch": 1.920620806725406,
+      "grad_norm": 0.5927976965904236,
+      "learning_rate": 0.0002,
+      "loss": 0.6598,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9222374909061515,
+      "grad_norm": 0.5974002480506897,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 11890
+    },
+    {
+      "epoch": 1.923854175086897,
+      "grad_norm": 0.7091866135597229,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9254708592676422,
+      "grad_norm": 0.72496497631073,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9270875434483874,
+      "grad_norm": 0.6131896376609802,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9287042276291326,
+      "grad_norm": 0.6556436419487,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9303209118098779,
+      "grad_norm": 0.622932493686676,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 11940
+    },
+    {
+      "epoch": 1.931937595990623,
+      "grad_norm": 0.6618631482124329,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9335542801713685,
+      "grad_norm": 0.630966305732727,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 11960
+    },
+    {
+      "epoch": 1.9351709643521138,
+      "grad_norm": 0.6336734890937805,
+      "learning_rate": 0.0002,
+      "loss": 0.6854,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9367876485328592,
+      "grad_norm": 0.655403196811676,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9384043327136045,
+      "grad_norm": 0.5640574097633362,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 11990
+    },
+    {
+      "epoch": 1.9400210168943497,
+      "grad_norm": 0.6322951316833496,
+      "learning_rate": 0.0002,
+      "loss": 0.7289,
+      "step": 12000
+    },
+    {
+      "epoch": 1.941637701075095,
+      "grad_norm": 0.615703821182251,
+      "learning_rate": 0.0002,
+      "loss": 0.7627,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9432543852558402,
+      "grad_norm": 0.6487536430358887,
+      "learning_rate": 0.0002,
+      "loss": 0.786,
+      "step": 12020
+    },
+    {
+      "epoch": 1.9448710694365856,
+      "grad_norm": 0.9209630489349365,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 12030
+    },
+    {
+      "epoch": 1.9464877536173308,
+      "grad_norm": 0.67485511302948,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9481044377980763,
+      "grad_norm": 0.6831230521202087,
+      "learning_rate": 0.0002,
+      "loss": 0.7551,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9497211219788215,
+      "grad_norm": 0.6578302383422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7546,
+      "step": 12060
+    },
+    {
+      "epoch": 1.9513378061595668,
+      "grad_norm": 0.9975938200950623,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12070
+    },
+    {
+      "epoch": 1.952954490340312,
+      "grad_norm": 0.6637365221977234,
+      "learning_rate": 0.0002,
+      "loss": 0.7952,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9545711745210572,
+      "grad_norm": 0.605707049369812,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9561878587018025,
+      "grad_norm": 0.6584440469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7768,
+      "step": 12100
+    },
+    {
+      "epoch": 1.957804542882548,
+      "grad_norm": 0.6070835590362549,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9594212270632931,
+      "grad_norm": 0.7862601280212402,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9610379112440386,
+      "grad_norm": 0.8175255060195923,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 12130
+    },
+    {
+      "epoch": 1.9626545954247838,
+      "grad_norm": 0.5648472905158997,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 12140
+    },
+    {
+      "epoch": 1.964271279605529,
+      "grad_norm": 0.6591973304748535,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9658879637862743,
+      "grad_norm": 0.5960676074028015,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9675046479670195,
+      "grad_norm": 0.7272544503211975,
+      "learning_rate": 0.0002,
+      "loss": 0.7254,
+      "step": 12170
+    },
+    {
+      "epoch": 1.969121332147765,
+      "grad_norm": 0.7176699042320251,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9707380163285102,
+      "grad_norm": 0.6927123665809631,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9723547005092557,
+      "grad_norm": 0.5536034107208252,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 12200
+    },
+    {
+      "epoch": 1.9739713846900009,
+      "grad_norm": 0.8348390460014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 12210
+    },
+    {
+      "epoch": 1.9755880688707461,
+      "grad_norm": 0.6591181755065918,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 12220
+    },
+    {
+      "epoch": 1.9772047530514913,
+      "grad_norm": 1.0624109506607056,
+      "learning_rate": 0.0002,
+      "loss": 0.763,
+      "step": 12230
+    },
+    {
+      "epoch": 1.9788214372322366,
+      "grad_norm": 0.9265586137771606,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 12240
+    },
+    {
+      "epoch": 1.9804381214129818,
+      "grad_norm": 0.5998196005821228,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 12250
+    },
+    {
+      "epoch": 1.9820548055937273,
+      "grad_norm": 0.6960851550102234,
+      "learning_rate": 0.0002,
+      "loss": 0.7225,
+      "step": 12260
+    },
+    {
+      "epoch": 1.9836714897744727,
+      "grad_norm": 0.7674502730369568,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 12270
+    },
+    {
+      "epoch": 1.985288173955218,
+      "grad_norm": 0.6407275795936584,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 12280
+    },
+    {
+      "epoch": 1.9869048581359632,
+      "grad_norm": 0.6673079133033752,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 12290
+    },
+    {
+      "epoch": 1.9885215423167084,
+      "grad_norm": 0.6989844441413879,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 12300
+    },
+    {
+      "epoch": 1.9901382264974536,
+      "grad_norm": 0.7564442157745361,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 12310
+    },
+    {
+      "epoch": 1.9917549106781989,
+      "grad_norm": 0.6385478973388672,
+      "learning_rate": 0.0002,
+      "loss": 0.7719,
+      "step": 12320
+    },
+    {
+      "epoch": 1.9933715948589443,
+      "grad_norm": 0.7193717956542969,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 12330
+    },
+    {
+      "epoch": 1.9949882790396896,
+      "grad_norm": 0.7987112402915955,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 12340
+    },
+    {
+      "epoch": 1.996604963220435,
+      "grad_norm": 0.7260826826095581,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 12350
+    },
+    {
+      "epoch": 1.9982216474011802,
+      "grad_norm": 0.7968255281448364,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 12360
+    },
+    {
+      "epoch": 1.9998383315819255,
+      "grad_norm": 0.6893062591552734,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1044032573699951,
+      "eval_runtime": 122.1508,
+      "eval_samples_per_second": 6.001,
+      "eval_steps_per_second": 0.753,
+      "step": 12371
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 49480,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.7250218020909875e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/training_args.bin b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..155b12fa9acbc6e71dba75c92bfa79e152397ebf
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-12371/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28694d5564a2b5c7d6881d4ba2af103356aa22489d2c22768ebbe47283c0f4a1
+size 5560
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/README.md b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..835e31ab8469ee39ddc8b2b6b2143a8c66dad510
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/README.md
@@ -0,0 +1,202 @@
+---
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.13.1
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/adapter_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..406c5a08dc4a2a33b52c62a482f98c217c417215
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/adapter_config.json
@@ -0,0 +1,29 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/adapter_model.safetensors b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..3ed21d20216a5ce4568a175156a5ddbb9cb33669
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d9f75e2ee0bafc7ec9c12b204485f764703d363e45fbf4d172349d03b0c4653
+size 109069176
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/optimizer.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b993ee39e96ec08b19bf65bc33ec96629b5cc736
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cad21d963ec0206ac34fc4ec45037ceb484173249d20e8c8a3ef5d7f9f971ef8
+size 55532666
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/rng_state.pth b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..45b03f0a02348d0c09b9c35820ade27d2825d1be
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a94f50547bb5416c60eade4d015f2d1426e23d331062664b231faf5aeeb61aee
+size 14244
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/scheduler.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..98cb302180982cc3bd57a359a66df9af4df06ef1
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e862db497a76925f2f33bbe167cf8fc7ad8a5bd750c2ca881ec45670205b20d
+size 1064
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/special_tokens_map.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/special_tokens_map.json
@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/tokenizer.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..86a33946b0c77216d2cce91bb28c8fada4a5e80b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/tokenizer_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..061e40d9db3253624f86e8e364c15ef546527c9d
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/tokenizer_config.json
@@ -0,0 +1,2063 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/trainer_state.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..cad81d8e99afadcc095799183f259473cbc9a987
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/trainer_state.json
@@ -0,0 +1,13042 @@
+{
+  "best_metric": 1.0871200561523438,
+  "best_model_checkpoint": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185",
+  "epoch": 2.9999191657909625,
+  "eval_steps": 10,
+  "global_step": 18556,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016166841807452913,
+      "grad_norm": 0.9894065856933594,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032333683614905826,
+      "grad_norm": 1.7810699939727783,
+      "learning_rate": 0.0002,
+      "loss": 1.1528,
+      "step": 20
+    },
+    {
+      "epoch": 0.004850052542235874,
+      "grad_norm": 0.5969577431678772,
+      "learning_rate": 0.0002,
+      "loss": 0.9767,
+      "step": 30
+    },
+    {
+      "epoch": 0.006466736722981165,
+      "grad_norm": 0.6354120969772339,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 40
+    },
+    {
+      "epoch": 0.008083420903726457,
+      "grad_norm": 0.5604607462882996,
+      "learning_rate": 0.0002,
+      "loss": 0.8643,
+      "step": 50
+    },
+    {
+      "epoch": 0.009700105084471748,
+      "grad_norm": 0.4676193594932556,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 60
+    },
+    {
+      "epoch": 0.01131678926521704,
+      "grad_norm": 0.6099211573600769,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 70
+    },
+    {
+      "epoch": 0.01293347344596233,
+      "grad_norm": 0.48639994859695435,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 80
+    },
+    {
+      "epoch": 0.014550157626707623,
+      "grad_norm": 0.4904264509677887,
+      "learning_rate": 0.0002,
+      "loss": 0.8704,
+      "step": 90
+    },
+    {
+      "epoch": 0.016166841807452915,
+      "grad_norm": 2.8334362506866455,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 100
+    },
+    {
+      "epoch": 0.017783525988198205,
+      "grad_norm": 0.43221670389175415,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 110
+    },
+    {
+      "epoch": 0.019400210168943496,
+      "grad_norm": 0.42244166135787964,
+      "learning_rate": 0.0002,
+      "loss": 0.8412,
+      "step": 120
+    },
+    {
+      "epoch": 0.02101689434968879,
+      "grad_norm": 0.45363298058509827,
+      "learning_rate": 0.0002,
+      "loss": 0.8467,
+      "step": 130
+    },
+    {
+      "epoch": 0.02263357853043408,
+      "grad_norm": 0.44816508889198303,
+      "learning_rate": 0.0002,
+      "loss": 0.8641,
+      "step": 140
+    },
+    {
+      "epoch": 0.02425026271117937,
+      "grad_norm": 0.43308213353157043,
+      "learning_rate": 0.0002,
+      "loss": 0.8496,
+      "step": 150
+    },
+    {
+      "epoch": 0.02586694689192466,
+      "grad_norm": 0.4084763526916504,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 160
+    },
+    {
+      "epoch": 0.027483631072669955,
+      "grad_norm": 0.5363703966140747,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 170
+    },
+    {
+      "epoch": 0.029100315253415245,
+      "grad_norm": 0.4619699716567993,
+      "learning_rate": 0.0002,
+      "loss": 0.8558,
+      "step": 180
+    },
+    {
+      "epoch": 0.030716999434160536,
+      "grad_norm": 0.49069908261299133,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 190
+    },
+    {
+      "epoch": 0.03233368361490583,
+      "grad_norm": 0.4645835757255554,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 200
+    },
+    {
+      "epoch": 0.03395036779565112,
+      "grad_norm": 1.2411243915557861,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 210
+    },
+    {
+      "epoch": 0.03556705197639641,
+      "grad_norm": 0.5211851596832275,
+      "learning_rate": 0.0002,
+      "loss": 0.8245,
+      "step": 220
+    },
+    {
+      "epoch": 0.037183736157141704,
+      "grad_norm": 0.5253691673278809,
+      "learning_rate": 0.0002,
+      "loss": 0.8194,
+      "step": 230
+    },
+    {
+      "epoch": 0.03880042033788699,
+      "grad_norm": 0.4567478895187378,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 240
+    },
+    {
+      "epoch": 0.040417104518632285,
+      "grad_norm": 0.5472128391265869,
+      "learning_rate": 0.0002,
+      "loss": 0.838,
+      "step": 250
+    },
+    {
+      "epoch": 0.04203378869937758,
+      "grad_norm": 0.42978546023368835,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 260
+    },
+    {
+      "epoch": 0.043650472880122866,
+      "grad_norm": 0.601734459400177,
+      "learning_rate": 0.0002,
+      "loss": 0.8334,
+      "step": 270
+    },
+    {
+      "epoch": 0.04526715706086816,
+      "grad_norm": 0.4286513328552246,
+      "learning_rate": 0.0002,
+      "loss": 0.815,
+      "step": 280
+    },
+    {
+      "epoch": 0.046883841241613454,
+      "grad_norm": 0.5230861902236938,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 290
+    },
+    {
+      "epoch": 0.04850052542235874,
+      "grad_norm": 0.6504611968994141,
+      "learning_rate": 0.0002,
+      "loss": 0.8636,
+      "step": 300
+    },
+    {
+      "epoch": 0.050117209603104035,
+      "grad_norm": 0.43485215306282043,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 310
+    },
+    {
+      "epoch": 0.05173389378384932,
+      "grad_norm": 0.4717007875442505,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 320
+    },
+    {
+      "epoch": 0.053350577964594616,
+      "grad_norm": 0.4059787690639496,
+      "learning_rate": 0.0002,
+      "loss": 0.8469,
+      "step": 330
+    },
+    {
+      "epoch": 0.05496726214533991,
+      "grad_norm": 0.4366913437843323,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 340
+    },
+    {
+      "epoch": 0.0565839463260852,
+      "grad_norm": 0.4233848452568054,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 350
+    },
+    {
+      "epoch": 0.05820063050683049,
+      "grad_norm": 0.4209108352661133,
+      "learning_rate": 0.0002,
+      "loss": 0.8456,
+      "step": 360
+    },
+    {
+      "epoch": 0.059817314687575784,
+      "grad_norm": 0.41637396812438965,
+      "learning_rate": 0.0002,
+      "loss": 0.816,
+      "step": 370
+    },
+    {
+      "epoch": 0.06143399886832107,
+      "grad_norm": 0.46235376596450806,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 380
+    },
+    {
+      "epoch": 0.06305068304906636,
+      "grad_norm": 0.4013484716415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 390
+    },
+    {
+      "epoch": 0.06466736722981166,
+      "grad_norm": 0.47443896532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 400
+    },
+    {
+      "epoch": 0.06628405141055695,
+      "grad_norm": 0.3942156434059143,
+      "learning_rate": 0.0002,
+      "loss": 0.8666,
+      "step": 410
+    },
+    {
+      "epoch": 0.06790073559130223,
+      "grad_norm": 0.4965320825576782,
+      "learning_rate": 0.0002,
+      "loss": 0.8402,
+      "step": 420
+    },
+    {
+      "epoch": 0.06951741977204753,
+      "grad_norm": 0.4304835796356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8317,
+      "step": 430
+    },
+    {
+      "epoch": 0.07113410395279282,
+      "grad_norm": 0.511726975440979,
+      "learning_rate": 0.0002,
+      "loss": 0.8528,
+      "step": 440
+    },
+    {
+      "epoch": 0.07275078813353811,
+      "grad_norm": 0.4040689170360565,
+      "learning_rate": 0.0002,
+      "loss": 0.8675,
+      "step": 450
+    },
+    {
+      "epoch": 0.07436747231428341,
+      "grad_norm": 0.5402171015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 460
+    },
+    {
+      "epoch": 0.0759841564950287,
+      "grad_norm": 0.4174517095088959,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 470
+    },
+    {
+      "epoch": 0.07760084067577398,
+      "grad_norm": 0.4306182265281677,
+      "learning_rate": 0.0002,
+      "loss": 0.7605,
+      "step": 480
+    },
+    {
+      "epoch": 0.07921752485651928,
+      "grad_norm": 0.535210132598877,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 490
+    },
+    {
+      "epoch": 0.08083420903726457,
+      "grad_norm": 0.5339109897613525,
+      "learning_rate": 0.0002,
+      "loss": 0.7825,
+      "step": 500
+    },
+    {
+      "epoch": 0.08245089321800986,
+      "grad_norm": 0.45754891633987427,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 510
+    },
+    {
+      "epoch": 0.08406757739875516,
+      "grad_norm": 0.43820783495903015,
+      "learning_rate": 0.0002,
+      "loss": 0.8144,
+      "step": 520
+    },
+    {
+      "epoch": 0.08568426157950045,
+      "grad_norm": 0.4434749186038971,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 530
+    },
+    {
+      "epoch": 0.08730094576024573,
+      "grad_norm": 0.43111467361450195,
+      "learning_rate": 0.0002,
+      "loss": 0.7857,
+      "step": 540
+    },
+    {
+      "epoch": 0.08891762994099103,
+      "grad_norm": 0.4378940165042877,
+      "learning_rate": 0.0002,
+      "loss": 0.8418,
+      "step": 550
+    },
+    {
+      "epoch": 0.09053431412173632,
+      "grad_norm": 0.4772215187549591,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 560
+    },
+    {
+      "epoch": 0.09215099830248161,
+      "grad_norm": 0.6837629079818726,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 570
+    },
+    {
+      "epoch": 0.09376768248322691,
+      "grad_norm": 0.42241212725639343,
+      "learning_rate": 0.0002,
+      "loss": 0.8607,
+      "step": 580
+    },
+    {
+      "epoch": 0.0953843666639722,
+      "grad_norm": 0.5165936350822449,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 590
+    },
+    {
+      "epoch": 0.09700105084471748,
+      "grad_norm": 0.48737478256225586,
+      "learning_rate": 0.0002,
+      "loss": 0.8664,
+      "step": 600
+    },
+    {
+      "epoch": 0.09861773502546278,
+      "grad_norm": 0.47419852018356323,
+      "learning_rate": 0.0002,
+      "loss": 0.8806,
+      "step": 610
+    },
+    {
+      "epoch": 0.10023441920620807,
+      "grad_norm": 0.4975486099720001,
+      "learning_rate": 0.0002,
+      "loss": 0.8254,
+      "step": 620
+    },
+    {
+      "epoch": 0.10185110338695336,
+      "grad_norm": 0.49123844504356384,
+      "learning_rate": 0.0002,
+      "loss": 0.8548,
+      "step": 630
+    },
+    {
+      "epoch": 0.10346778756769864,
+      "grad_norm": 0.6288952827453613,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 640
+    },
+    {
+      "epoch": 0.10508447174844394,
+      "grad_norm": 0.4277345836162567,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 650
+    },
+    {
+      "epoch": 0.10670115592918923,
+      "grad_norm": 0.4021061956882477,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 660
+    },
+    {
+      "epoch": 0.10831784010993452,
+      "grad_norm": 0.3492237329483032,
+      "learning_rate": 0.0002,
+      "loss": 0.87,
+      "step": 670
+    },
+    {
+      "epoch": 0.10993452429067982,
+      "grad_norm": 0.4341012239456177,
+      "learning_rate": 0.0002,
+      "loss": 0.8698,
+      "step": 680
+    },
+    {
+      "epoch": 0.1115512084714251,
+      "grad_norm": 0.7296304106712341,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 690
+    },
+    {
+      "epoch": 0.1131678926521704,
+      "grad_norm": 0.397494912147522,
+      "learning_rate": 0.0002,
+      "loss": 0.8433,
+      "step": 700
+    },
+    {
+      "epoch": 0.1147845768329157,
+      "grad_norm": 0.396431028842926,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 710
+    },
+    {
+      "epoch": 0.11640126101366098,
+      "grad_norm": 0.48842838406562805,
+      "learning_rate": 0.0002,
+      "loss": 0.8379,
+      "step": 720
+    },
+    {
+      "epoch": 0.11801794519440627,
+      "grad_norm": 0.46322616934776306,
+      "learning_rate": 0.0002,
+      "loss": 0.8238,
+      "step": 730
+    },
+    {
+      "epoch": 0.11963462937515157,
+      "grad_norm": 0.47990912199020386,
+      "learning_rate": 0.0002,
+      "loss": 0.8041,
+      "step": 740
+    },
+    {
+      "epoch": 0.12125131355589686,
+      "grad_norm": 0.4997142255306244,
+      "learning_rate": 0.0002,
+      "loss": 0.82,
+      "step": 750
+    },
+    {
+      "epoch": 0.12286799773664214,
+      "grad_norm": 0.4040526747703552,
+      "learning_rate": 0.0002,
+      "loss": 0.7702,
+      "step": 760
+    },
+    {
+      "epoch": 0.12448468191738744,
+      "grad_norm": 0.453095942735672,
+      "learning_rate": 0.0002,
+      "loss": 0.863,
+      "step": 770
+    },
+    {
+      "epoch": 0.12610136609813272,
+      "grad_norm": 0.4636971950531006,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 780
+    },
+    {
+      "epoch": 0.12771805027887803,
+      "grad_norm": 0.4279276132583618,
+      "learning_rate": 0.0002,
+      "loss": 0.8112,
+      "step": 790
+    },
+    {
+      "epoch": 0.12933473445962332,
+      "grad_norm": 0.46212655305862427,
+      "learning_rate": 0.0002,
+      "loss": 0.8711,
+      "step": 800
+    },
+    {
+      "epoch": 0.1309514186403686,
+      "grad_norm": 0.43127650022506714,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 810
+    },
+    {
+      "epoch": 0.1325681028211139,
+      "grad_norm": 0.4201301336288452,
+      "learning_rate": 0.0002,
+      "loss": 0.8476,
+      "step": 820
+    },
+    {
+      "epoch": 0.13418478700185918,
+      "grad_norm": 0.42583167552948,
+      "learning_rate": 0.0002,
+      "loss": 0.8078,
+      "step": 830
+    },
+    {
+      "epoch": 0.13580147118260447,
+      "grad_norm": 0.4535622000694275,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 840
+    },
+    {
+      "epoch": 0.13741815536334978,
+      "grad_norm": 0.4116036891937256,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 850
+    },
+    {
+      "epoch": 0.13903483954409507,
+      "grad_norm": 0.45997580885887146,
+      "learning_rate": 0.0002,
+      "loss": 0.8466,
+      "step": 860
+    },
+    {
+      "epoch": 0.14065152372484035,
+      "grad_norm": 0.4487837255001068,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 870
+    },
+    {
+      "epoch": 0.14226820790558564,
+      "grad_norm": 0.43650057911872864,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 880
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 0.5335358381271362,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 890
+    },
+    {
+      "epoch": 0.14550157626707622,
+      "grad_norm": 0.5989000201225281,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 900
+    },
+    {
+      "epoch": 0.14711826044782153,
+      "grad_norm": 0.517179012298584,
+      "learning_rate": 0.0002,
+      "loss": 0.8385,
+      "step": 910
+    },
+    {
+      "epoch": 0.14873494462856682,
+      "grad_norm": 0.44435232877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.8255,
+      "step": 920
+    },
+    {
+      "epoch": 0.1503516288093121,
+      "grad_norm": 0.42635923624038696,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 930
+    },
+    {
+      "epoch": 0.1519683129900574,
+      "grad_norm": 0.49603334069252014,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 940
+    },
+    {
+      "epoch": 0.15358499717080268,
+      "grad_norm": 0.40639808773994446,
+      "learning_rate": 0.0002,
+      "loss": 0.8377,
+      "step": 950
+    },
+    {
+      "epoch": 0.15520168135154797,
+      "grad_norm": 0.4850759208202362,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 960
+    },
+    {
+      "epoch": 0.15681836553229328,
+      "grad_norm": 0.4427442252635956,
+      "learning_rate": 0.0002,
+      "loss": 0.846,
+      "step": 970
+    },
+    {
+      "epoch": 0.15843504971303857,
+      "grad_norm": 0.3760930001735687,
+      "learning_rate": 0.0002,
+      "loss": 0.8705,
+      "step": 980
+    },
+    {
+      "epoch": 0.16005173389378385,
+      "grad_norm": 0.4794144332408905,
+      "learning_rate": 0.0002,
+      "loss": 0.8644,
+      "step": 990
+    },
+    {
+      "epoch": 0.16166841807452914,
+      "grad_norm": 0.45828768610954285,
+      "learning_rate": 0.0002,
+      "loss": 0.8002,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16328510225527443,
+      "grad_norm": 0.6313053369522095,
+      "learning_rate": 0.0002,
+      "loss": 0.7658,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16490178643601971,
+      "grad_norm": 0.45041006803512573,
+      "learning_rate": 0.0002,
+      "loss": 0.8047,
+      "step": 1020
+    },
+    {
+      "epoch": 0.166518470616765,
+      "grad_norm": 0.441403865814209,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16813515479751032,
+      "grad_norm": 0.8171296119689941,
+      "learning_rate": 0.0002,
+      "loss": 0.8475,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1697518389782556,
+      "grad_norm": 0.7137420773506165,
+      "learning_rate": 0.0002,
+      "loss": 0.845,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1713685231590009,
+      "grad_norm": 0.5236809849739075,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17298520733974618,
+      "grad_norm": 0.5021864175796509,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17460189152049146,
+      "grad_norm": 0.47347521781921387,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17621857570123675,
+      "grad_norm": 0.4631653428077698,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17783525988198207,
+      "grad_norm": 0.49169182777404785,
+      "learning_rate": 0.0002,
+      "loss": 0.8166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17945194406272735,
+      "grad_norm": 0.5019739270210266,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18106862824347264,
+      "grad_norm": 0.5100422501564026,
+      "learning_rate": 0.0002,
+      "loss": 0.8247,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18268531242421793,
+      "grad_norm": 0.3888324499130249,
+      "learning_rate": 0.0002,
+      "loss": 0.8142,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18430199660496321,
+      "grad_norm": 0.39765217900276184,
+      "learning_rate": 0.0002,
+      "loss": 0.8533,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1859186807857085,
+      "grad_norm": 0.47190186381340027,
+      "learning_rate": 0.0002,
+      "loss": 0.8541,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18753536496645382,
+      "grad_norm": 0.4464188814163208,
+      "learning_rate": 0.0002,
+      "loss": 0.8301,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1891520491471991,
+      "grad_norm": 0.5153930187225342,
+      "learning_rate": 0.0002,
+      "loss": 0.8341,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1907687333279444,
+      "grad_norm": 0.4779708683490753,
+      "learning_rate": 0.0002,
+      "loss": 0.8033,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19238541750868968,
+      "grad_norm": 0.4834315776824951,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19400210168943496,
+      "grad_norm": 0.402357816696167,
+      "learning_rate": 0.0002,
+      "loss": 0.7721,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19561878587018025,
+      "grad_norm": 0.45899084210395813,
+      "learning_rate": 0.0002,
+      "loss": 0.7941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19723547005092557,
+      "grad_norm": 0.5106529593467712,
+      "learning_rate": 0.0002,
+      "loss": 0.8353,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19885215423167085,
+      "grad_norm": 0.45261722803115845,
+      "learning_rate": 0.0002,
+      "loss": 0.7816,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20046883841241614,
+      "grad_norm": 0.4647127091884613,
+      "learning_rate": 0.0002,
+      "loss": 0.8068,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20208552259316143,
+      "grad_norm": 0.4849368929862976,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2037022067739067,
+      "grad_norm": 0.4518061578273773,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 1260
+    },
+    {
+      "epoch": 0.205318890954652,
+      "grad_norm": 0.49535325169563293,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2069355751353973,
+      "grad_norm": 0.4835205376148224,
+      "learning_rate": 0.0002,
+      "loss": 0.8348,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2085522593161426,
+      "grad_norm": 0.45308539271354675,
+      "learning_rate": 0.0002,
+      "loss": 0.8428,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2101689434968879,
+      "grad_norm": 0.5369905233383179,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21178562767763318,
+      "grad_norm": 0.5031622052192688,
+      "learning_rate": 0.0002,
+      "loss": 0.8676,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21340231185837846,
+      "grad_norm": 0.48010334372520447,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21501899603912375,
+      "grad_norm": 0.4905701279640198,
+      "learning_rate": 0.0002,
+      "loss": 0.806,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21663568021986904,
+      "grad_norm": 0.43531742691993713,
+      "learning_rate": 0.0002,
+      "loss": 0.7885,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21825236440061435,
+      "grad_norm": 0.44330692291259766,
+      "learning_rate": 0.0002,
+      "loss": 0.8191,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21986904858135964,
+      "grad_norm": 0.5384416580200195,
+      "learning_rate": 0.0002,
+      "loss": 0.8205,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22148573276210493,
+      "grad_norm": 0.4181833863258362,
+      "learning_rate": 0.0002,
+      "loss": 0.7726,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2231024169428502,
+      "grad_norm": 0.523833692073822,
+      "learning_rate": 0.0002,
+      "loss": 0.8311,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.5528736710548401,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2263357853043408,
+      "grad_norm": 0.43515023589134216,
+      "learning_rate": 0.0002,
+      "loss": 0.8079,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2279524694850861,
+      "grad_norm": 0.48809877038002014,
+      "learning_rate": 0.0002,
+      "loss": 0.8403,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2295691536658314,
+      "grad_norm": 0.43591251969337463,
+      "learning_rate": 0.0002,
+      "loss": 0.8165,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23118583784657668,
+      "grad_norm": 0.44625312089920044,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23280252202732196,
+      "grad_norm": 0.4390665292739868,
+      "learning_rate": 0.0002,
+      "loss": 0.8134,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23441920620806725,
+      "grad_norm": 0.48496049642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.8465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23603589038881254,
+      "grad_norm": 0.45919957756996155,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23765257456955785,
+      "grad_norm": 0.5471845865249634,
+      "learning_rate": 0.0002,
+      "loss": 0.8659,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23926925875030314,
+      "grad_norm": 0.47269317507743835,
+      "learning_rate": 0.0002,
+      "loss": 0.8164,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24088594293104842,
+      "grad_norm": 0.4930245578289032,
+      "learning_rate": 0.0002,
+      "loss": 0.854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2425026271117937,
+      "grad_norm": 0.5605630278587341,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 1500
+    },
+    {
+      "epoch": 0.244119311292539,
+      "grad_norm": 0.4435870945453644,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24573599547328429,
+      "grad_norm": 0.4941999912261963,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24735267965402957,
+      "grad_norm": 0.5100624561309814,
+      "learning_rate": 0.0002,
+      "loss": 0.8427,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2489693638347749,
+      "grad_norm": 0.4638267457485199,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25058604801552015,
+      "grad_norm": 0.5071570873260498,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25220273219626543,
+      "grad_norm": 0.4291319251060486,
+      "learning_rate": 0.0002,
+      "loss": 0.7724,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2538194163770108,
+      "grad_norm": 0.5388049483299255,
+      "learning_rate": 0.0002,
+      "loss": 0.7984,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25543610055775606,
+      "grad_norm": 0.5083683729171753,
+      "learning_rate": 0.0002,
+      "loss": 0.8176,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25705278473850135,
+      "grad_norm": 0.4824463725090027,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25866946891924664,
+      "grad_norm": 0.41177722811698914,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2602861530999919,
+      "grad_norm": 0.5656219124794006,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2619028372807372,
+      "grad_norm": 0.41063204407691956,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2635195214614825,
+      "grad_norm": 0.4897061288356781,
+      "learning_rate": 0.0002,
+      "loss": 0.7998,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2651362056422278,
+      "grad_norm": 0.4454376697540283,
+      "learning_rate": 0.0002,
+      "loss": 0.8198,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26675288982297307,
+      "grad_norm": 0.4355238378047943,
+      "learning_rate": 0.0002,
+      "loss": 0.8684,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26836957400371836,
+      "grad_norm": 0.458310067653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26998625818446365,
+      "grad_norm": 0.4752083718776703,
+      "learning_rate": 0.0002,
+      "loss": 0.7935,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27160294236520893,
+      "grad_norm": 0.4666106402873993,
+      "learning_rate": 0.0002,
+      "loss": 0.8267,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2732196265459543,
+      "grad_norm": 0.4213818609714508,
+      "learning_rate": 0.0002,
+      "loss": 0.8252,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27483631072669956,
+      "grad_norm": 0.5768913626670837,
+      "learning_rate": 0.0002,
+      "loss": 0.8559,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27645299490744485,
+      "grad_norm": 0.4209914803504944,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27806967908819014,
+      "grad_norm": 0.501909613609314,
+      "learning_rate": 0.0002,
+      "loss": 0.8167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2796863632689354,
+      "grad_norm": 0.5266261100769043,
+      "learning_rate": 0.0002,
+      "loss": 0.7832,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2813030474496807,
+      "grad_norm": 0.43806859850883484,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 1740
+    },
+    {
+      "epoch": 0.282919731630426,
+      "grad_norm": 0.46048814058303833,
+      "learning_rate": 0.0002,
+      "loss": 0.8157,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2845364158111713,
+      "grad_norm": 0.44972819089889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8596,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28615309999191657,
+      "grad_norm": 0.5114831328392029,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.47931742668151855,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28938646835340714,
+      "grad_norm": 0.5092599987983704,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29100315253415243,
+      "grad_norm": 0.37581443786621094,
+      "learning_rate": 0.0002,
+      "loss": 0.8506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2926198367148977,
+      "grad_norm": 0.47097381949424744,
+      "learning_rate": 0.0002,
+      "loss": 0.7932,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29423652089564306,
+      "grad_norm": 0.48300236463546753,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29585320507638835,
+      "grad_norm": 0.5600419640541077,
+      "learning_rate": 0.0002,
+      "loss": 0.8391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29746988925713364,
+      "grad_norm": 0.48555272817611694,
+      "learning_rate": 0.0002,
+      "loss": 0.8507,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2990865734378789,
+      "grad_norm": 0.3752668499946594,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3007032576186242,
+      "grad_norm": 0.5328747034072876,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3023199417993695,
+      "grad_norm": 0.48716455698013306,
+      "learning_rate": 0.0002,
+      "loss": 0.8426,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3039366259801148,
+      "grad_norm": 0.5011493563652039,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30555331016086007,
+      "grad_norm": 0.46461427211761475,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30716999434160536,
+      "grad_norm": 0.36630210280418396,
+      "learning_rate": 0.0002,
+      "loss": 0.8478,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30878667852235064,
+      "grad_norm": 0.4217296242713928,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31040336270309593,
+      "grad_norm": 0.4394875466823578,
+      "learning_rate": 0.0002,
+      "loss": 0.8128,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3120200468838412,
+      "grad_norm": 0.6587965488433838,
+      "learning_rate": 0.0002,
+      "loss": 0.8471,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31363673106458656,
+      "grad_norm": 0.5469298958778381,
+      "learning_rate": 0.0002,
+      "loss": 0.8565,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31525341524533185,
+      "grad_norm": 0.4371595084667206,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31687009942607713,
+      "grad_norm": 0.4809541404247284,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3184867836068224,
+      "grad_norm": 0.6061086654663086,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3201034677875677,
+      "grad_norm": 0.5342657566070557,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 1980
+    },
+    {
+      "epoch": 0.321720151968313,
+      "grad_norm": 0.5057743787765503,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3233368361490583,
+      "grad_norm": 0.528626024723053,
+      "learning_rate": 0.0002,
+      "loss": 0.7774,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32495352032980357,
+      "grad_norm": 0.46742770075798035,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32657020451054886,
+      "grad_norm": 0.515101432800293,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32818688869129414,
+      "grad_norm": 0.41941216588020325,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32980357287203943,
+      "grad_norm": 0.49902522563934326,
+      "learning_rate": 0.0002,
+      "loss": 0.7637,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3314202570527847,
+      "grad_norm": 0.4120897650718689,
+      "learning_rate": 0.0002,
+      "loss": 0.7822,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33303694123353,
+      "grad_norm": 0.45352041721343994,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33465362541427535,
+      "grad_norm": 0.523199737071991,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33627030959502063,
+      "grad_norm": 0.4390358626842499,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3378869937757659,
+      "grad_norm": 0.6752901077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3395036779565112,
+      "grad_norm": 0.547821044921875,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3411203621372565,
+      "grad_norm": 0.5161308646202087,
+      "learning_rate": 0.0002,
+      "loss": 0.8561,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3427370463180018,
+      "grad_norm": 0.4565401077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34435373049874707,
+      "grad_norm": 0.4666115939617157,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34597041467949236,
+      "grad_norm": 0.4090428352355957,
+      "learning_rate": 0.0002,
+      "loss": 0.8189,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34758709886023764,
+      "grad_norm": 0.510845422744751,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34920378304098293,
+      "grad_norm": 0.42861923575401306,
+      "learning_rate": 0.0002,
+      "loss": 0.8398,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3508204672217282,
+      "grad_norm": 0.4476332664489746,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3524371514024735,
+      "grad_norm": 0.6065791249275208,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35405383558321885,
+      "grad_norm": 0.42335066199302673,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35567051976396413,
+      "grad_norm": 0.5094629526138306,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3572872039447094,
+      "grad_norm": 0.5476373434066772,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3589038881254547,
+      "grad_norm": 0.3911719024181366,
+      "learning_rate": 0.0002,
+      "loss": 0.814,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3605205723062,
+      "grad_norm": 0.6599636077880859,
+      "learning_rate": 0.0002,
+      "loss": 0.8599,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3621372564869453,
+      "grad_norm": 0.40381914377212524,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36375394066769057,
+      "grad_norm": 0.4433908462524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36537062484843585,
+      "grad_norm": 0.578326940536499,
+      "learning_rate": 0.0002,
+      "loss": 0.8503,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36698730902918114,
+      "grad_norm": 0.5734784007072449,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36860399320992643,
+      "grad_norm": 0.45555487275123596,
+      "learning_rate": 0.0002,
+      "loss": 0.8193,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3702206773906717,
+      "grad_norm": 0.5666276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.7929,
+      "step": 2290
+    },
+    {
+      "epoch": 0.371837361571417,
+      "grad_norm": 0.5461117625236511,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3734540457521623,
+      "grad_norm": 0.6318911910057068,
+      "learning_rate": 0.0002,
+      "loss": 0.8204,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37507072993290763,
+      "grad_norm": 0.493263304233551,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3766874141136529,
+      "grad_norm": 0.5888760089874268,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3783040982943982,
+      "grad_norm": 0.48671841621398926,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3799207824751435,
+      "grad_norm": 0.4385145306587219,
+      "learning_rate": 0.0002,
+      "loss": 0.8367,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3815374666558888,
+      "grad_norm": 0.5523318648338318,
+      "learning_rate": 0.0002,
+      "loss": 0.812,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38315415083663407,
+      "grad_norm": 0.7308220267295837,
+      "learning_rate": 0.0002,
+      "loss": 0.8351,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38477083501737935,
+      "grad_norm": 0.554214358329773,
+      "learning_rate": 0.0002,
+      "loss": 0.859,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38638751919812464,
+      "grad_norm": 0.5425800085067749,
+      "learning_rate": 0.0002,
+      "loss": 0.8146,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3880042033788699,
+      "grad_norm": 0.48811158537864685,
+      "learning_rate": 0.0002,
+      "loss": 0.8282,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3896208875596152,
+      "grad_norm": 0.49212366342544556,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3912375717403605,
+      "grad_norm": 0.5222218632698059,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3928542559211058,
+      "grad_norm": 0.4699819087982178,
+      "learning_rate": 0.0002,
+      "loss": 0.8182,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39447094010185113,
+      "grad_norm": 0.46153587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7919,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3960876242825964,
+      "grad_norm": 0.4150611162185669,
+      "learning_rate": 0.0002,
+      "loss": 0.8111,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3977043084633417,
+      "grad_norm": 0.5799614787101746,
+      "learning_rate": 0.0002,
+      "loss": 0.8589,
+      "step": 2460
+    },
+    {
+      "epoch": 0.399320992644087,
+      "grad_norm": 0.56536865234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4009376768248323,
+      "grad_norm": 0.5451247096061707,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40255436100557757,
+      "grad_norm": 0.5914521217346191,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40417104518632285,
+      "grad_norm": 0.4428117275238037,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40578772936706814,
+      "grad_norm": 0.48580947518348694,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4074044135478134,
+      "grad_norm": 0.436734676361084,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4090210977285587,
+      "grad_norm": 0.5752223134040833,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 2530
+    },
+    {
+      "epoch": 0.410637781909304,
+      "grad_norm": 0.4271308183670044,
+      "learning_rate": 0.0002,
+      "loss": 0.8181,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4122544660900493,
+      "grad_norm": 0.46294718980789185,
+      "learning_rate": 0.0002,
+      "loss": 0.8058,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4138711502707946,
+      "grad_norm": 0.49407583475112915,
+      "learning_rate": 0.0002,
+      "loss": 0.8473,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4154878344515399,
+      "grad_norm": 0.4729035496711731,
+      "learning_rate": 0.0002,
+      "loss": 0.7881,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4171045186322852,
+      "grad_norm": 0.4129747152328491,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4187212028130305,
+      "grad_norm": 0.5684236288070679,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4203378869937758,
+      "grad_norm": 0.4862157106399536,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42195457117452106,
+      "grad_norm": 0.46567976474761963,
+      "learning_rate": 0.0002,
+      "loss": 0.7582,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42357125535526635,
+      "grad_norm": 0.5710650682449341,
+      "learning_rate": 0.0002,
+      "loss": 0.7755,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42518793953601164,
+      "grad_norm": 0.5660041570663452,
+      "learning_rate": 0.0002,
+      "loss": 0.8573,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4268046237167569,
+      "grad_norm": 0.47944375872612,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4284213078975022,
+      "grad_norm": 0.537223756313324,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4300379920782475,
+      "grad_norm": 0.41669997572898865,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 0.44727686047554016,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4332713604397381,
+      "grad_norm": 0.5600888729095459,
+      "learning_rate": 0.0002,
+      "loss": 0.8241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4348880446204834,
+      "grad_norm": 0.39820605516433716,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4365047288012287,
+      "grad_norm": 0.5637655854225159,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 2700
+    },
+    {
+      "epoch": 0.438121412981974,
+      "grad_norm": 0.6363666653633118,
+      "learning_rate": 0.0002,
+      "loss": 0.855,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4397380971627193,
+      "grad_norm": 0.5656129121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.8468,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44135478134346456,
+      "grad_norm": 0.5600156188011169,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44297146552420985,
+      "grad_norm": 0.5506579875946045,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44458814970495514,
+      "grad_norm": 0.49878305196762085,
+      "learning_rate": 0.0002,
+      "loss": 0.7725,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4462048338857004,
+      "grad_norm": 0.4569213092327118,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4478215180664457,
+      "grad_norm": 0.6056680083274841,
+      "learning_rate": 0.0002,
+      "loss": 0.8028,
+      "step": 2770
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.44474557042121887,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4510548864279363,
+      "grad_norm": 0.46055394411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.801,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4526715706086816,
+      "grad_norm": 0.4904133379459381,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45428825478942686,
+      "grad_norm": 0.5647031664848328,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4559049389701722,
+      "grad_norm": 0.5759473443031311,
+      "learning_rate": 0.0002,
+      "loss": 0.8622,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4575216231509175,
+      "grad_norm": 0.5161895751953125,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4591383073316628,
+      "grad_norm": 0.4248254597187042,
+      "learning_rate": 0.0002,
+      "loss": 0.8045,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46075499151240806,
+      "grad_norm": 0.45395001769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.7838,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46237167569315335,
+      "grad_norm": 0.5358697772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.8208,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46398835987389864,
+      "grad_norm": 0.5379165410995483,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4656050440546439,
+      "grad_norm": 0.4601989686489105,
+      "learning_rate": 0.0002,
+      "loss": 0.7403,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4672217282353892,
+      "grad_norm": 0.671115517616272,
+      "learning_rate": 0.0002,
+      "loss": 0.8523,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4688384124161345,
+      "grad_norm": 0.4425133168697357,
+      "learning_rate": 0.0002,
+      "loss": 0.8262,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4704550965968798,
+      "grad_norm": 0.5446155071258545,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47207178077762507,
+      "grad_norm": 0.603306233882904,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47368846495837036,
+      "grad_norm": 0.5377997159957886,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4753051491391157,
+      "grad_norm": 0.4931027591228485,
+      "learning_rate": 0.0002,
+      "loss": 0.8075,
+      "step": 2940
+    },
+    {
+      "epoch": 0.476921833319861,
+      "grad_norm": 0.4711960256099701,
+      "learning_rate": 0.0002,
+      "loss": 0.8004,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4785385175006063,
+      "grad_norm": 0.5020492672920227,
+      "learning_rate": 0.0002,
+      "loss": 0.8121,
+      "step": 2960
+    },
+    {
+      "epoch": 0.48015520168135156,
+      "grad_norm": 0.5428946614265442,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48177188586209685,
+      "grad_norm": 0.5294089317321777,
+      "learning_rate": 0.0002,
+      "loss": 0.7849,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48338857004284214,
+      "grad_norm": 0.648289144039154,
+      "learning_rate": 0.0002,
+      "loss": 0.8553,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4850052542235874,
+      "grad_norm": 0.47916680574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.7874,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4866219384043327,
+      "grad_norm": 0.43849772214889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8087,
+      "step": 3010
+    },
+    {
+      "epoch": 0.488238622585078,
+      "grad_norm": 0.47007861733436584,
+      "learning_rate": 0.0002,
+      "loss": 0.7662,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4898553067658233,
+      "grad_norm": 0.6314331293106079,
+      "learning_rate": 0.0002,
+      "loss": 0.757,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49147199094656857,
+      "grad_norm": 0.49211493134498596,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49308867512731386,
+      "grad_norm": 0.4537973403930664,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49470535930805914,
+      "grad_norm": 0.47326919436454773,
+      "learning_rate": 0.0002,
+      "loss": 0.8095,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4963220434888045,
+      "grad_norm": 0.525874137878418,
+      "learning_rate": 0.0002,
+      "loss": 0.8447,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4979387276695498,
+      "grad_norm": 0.6361091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49955541185029506,
+      "grad_norm": 0.5850642919540405,
+      "learning_rate": 0.0002,
+      "loss": 0.821,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5011720960310403,
+      "grad_norm": 0.47299543023109436,
+      "learning_rate": 0.0002,
+      "loss": 0.8279,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5027887802117856,
+      "grad_norm": 0.473099946975708,
+      "learning_rate": 0.0002,
+      "loss": 0.8681,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5044054643925309,
+      "grad_norm": 0.48186397552490234,
+      "learning_rate": 0.0002,
+      "loss": 0.8223,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5060221485732762,
+      "grad_norm": 0.5015401840209961,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5076388327540216,
+      "grad_norm": 0.5617750287055969,
+      "learning_rate": 0.0002,
+      "loss": 0.7692,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5092555169347668,
+      "grad_norm": 0.5169327259063721,
+      "learning_rate": 0.0002,
+      "loss": 0.8708,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5108722011155121,
+      "grad_norm": 0.545657753944397,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5124888852962574,
+      "grad_norm": 0.512864351272583,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5141055694770027,
+      "grad_norm": 0.4113546311855316,
+      "learning_rate": 0.0002,
+      "loss": 0.7794,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5157222536577479,
+      "grad_norm": 0.44532445073127747,
+      "learning_rate": 0.0002,
+      "loss": 0.8206,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5173389378384933,
+      "grad_norm": 0.5623497366905212,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5189556220192385,
+      "grad_norm": 0.5084741115570068,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5205723061999838,
+      "grad_norm": 0.5305403470993042,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5221889903807291,
+      "grad_norm": 0.4708254337310791,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5238056745614744,
+      "grad_norm": 0.43827131390571594,
+      "learning_rate": 0.0002,
+      "loss": 0.7639,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5254223587422197,
+      "grad_norm": 0.5630002617835999,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 3250
+    },
+    {
+      "epoch": 0.527039042922965,
+      "grad_norm": 0.5010961890220642,
+      "learning_rate": 0.0002,
+      "loss": 0.7522,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5286557271037103,
+      "grad_norm": 0.6303122043609619,
+      "learning_rate": 0.0002,
+      "loss": 0.8374,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5302724112844556,
+      "grad_norm": 0.5107331275939941,
+      "learning_rate": 0.0002,
+      "loss": 0.7727,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5318890954652009,
+      "grad_norm": 0.5700443387031555,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5335057796459461,
+      "grad_norm": 0.46296367049217224,
+      "learning_rate": 0.0002,
+      "loss": 0.7776,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5351224638266915,
+      "grad_norm": 0.531568706035614,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5367391480074367,
+      "grad_norm": 0.4686741530895233,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5383558321881821,
+      "grad_norm": 0.5404331088066101,
+      "learning_rate": 0.0002,
+      "loss": 0.8104,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5399725163689273,
+      "grad_norm": 0.6368790864944458,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5415892005496726,
+      "grad_norm": 0.42300888895988464,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5432058847304179,
+      "grad_norm": 0.5362542867660522,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5448225689111632,
+      "grad_norm": 0.497128963470459,
+      "learning_rate": 0.0002,
+      "loss": 0.858,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5464392530919085,
+      "grad_norm": 0.5006386041641235,
+      "learning_rate": 0.0002,
+      "loss": 0.8519,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5480559372726538,
+      "grad_norm": 0.44136837124824524,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5496726214533991,
+      "grad_norm": 0.5897833108901978,
+      "learning_rate": 0.0002,
+      "loss": 0.773,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5512893056341444,
+      "grad_norm": 0.641075611114502,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5529059898148897,
+      "grad_norm": 0.7251322269439697,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5545226739956349,
+      "grad_norm": 0.47411349415779114,
+      "learning_rate": 0.0002,
+      "loss": 0.7626,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5561393581763803,
+      "grad_norm": 0.4994310438632965,
+      "learning_rate": 0.0002,
+      "loss": 0.8196,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5577560423571255,
+      "grad_norm": 0.5814438462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5593727265378708,
+      "grad_norm": 0.6278898119926453,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5609894107186161,
+      "grad_norm": 0.46208274364471436,
+      "learning_rate": 0.0002,
+      "loss": 0.813,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5626060948993614,
+      "grad_norm": 0.5718930959701538,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5642227790801067,
+      "grad_norm": 0.48178744316101074,
+      "learning_rate": 0.0002,
+      "loss": 0.8152,
+      "step": 3490
+    },
+    {
+      "epoch": 0.565839463260852,
+      "grad_norm": 0.47336965799331665,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5674561474415973,
+      "grad_norm": 0.43442684412002563,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5690728316223426,
+      "grad_norm": 0.6463358998298645,
+      "learning_rate": 0.0002,
+      "loss": 0.7564,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5706895158030879,
+      "grad_norm": 0.5286486744880676,
+      "learning_rate": 0.0002,
+      "loss": 0.836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5723061999838331,
+      "grad_norm": 0.5405499935150146,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5739228841645785,
+      "grad_norm": 0.6654391884803772,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.5081980228424072,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5771562525260691,
+      "grad_norm": 0.48978179693222046,
+      "learning_rate": 0.0002,
+      "loss": 0.7753,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5787729367068143,
+      "grad_norm": 0.5840612053871155,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5803896208875596,
+      "grad_norm": 0.5235261917114258,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5820063050683049,
+      "grad_norm": 0.5672075748443604,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5836229892490502,
+      "grad_norm": 0.5613429546356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5852396734297954,
+      "grad_norm": 0.4032273590564728,
+      "learning_rate": 0.0002,
+      "loss": 0.8274,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5868563576105408,
+      "grad_norm": 0.49559324979782104,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5884730417912861,
+      "grad_norm": 0.6895697712898254,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5900897259720314,
+      "grad_norm": 0.4750136435031891,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5917064101527767,
+      "grad_norm": 0.5176819562911987,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5933230943335219,
+      "grad_norm": 0.5817760229110718,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5949397785142673,
+      "grad_norm": 0.6064626574516296,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5965564626950125,
+      "grad_norm": 0.6728700995445251,
+      "learning_rate": 0.0002,
+      "loss": 0.8422,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5981731468757578,
+      "grad_norm": 0.609305202960968,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5997898310565031,
+      "grad_norm": 0.4615488350391388,
+      "learning_rate": 0.0002,
+      "loss": 0.8048,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6014065152372484,
+      "grad_norm": 2.0531179904937744,
+      "learning_rate": 0.0002,
+      "loss": 0.8214,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6030231994179936,
+      "grad_norm": 0.5091132521629333,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 3730
+    },
+    {
+      "epoch": 0.604639883598739,
+      "grad_norm": 0.5951124429702759,
+      "learning_rate": 0.0002,
+      "loss": 0.7833,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6062565677794842,
+      "grad_norm": 0.5870208144187927,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6078732519602296,
+      "grad_norm": 0.6254619359970093,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6094899361409749,
+      "grad_norm": 0.5577626824378967,
+      "learning_rate": 0.0002,
+      "loss": 0.7868,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6111066203217201,
+      "grad_norm": 0.5004405379295349,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6127233045024655,
+      "grad_norm": 0.5527383685112,
+      "learning_rate": 0.0002,
+      "loss": 0.8092,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6143399886832107,
+      "grad_norm": 0.49116113781929016,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6159566728639561,
+      "grad_norm": 0.5299299359321594,
+      "learning_rate": 0.0002,
+      "loss": 0.8352,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6175733570447013,
+      "grad_norm": 0.464897483587265,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6191900412254466,
+      "grad_norm": 0.6505740880966187,
+      "learning_rate": 0.0002,
+      "loss": 0.7923,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6208067254061919,
+      "grad_norm": 0.5512559413909912,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6224234095869372,
+      "grad_norm": 0.49427518248558044,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6240400937676824,
+      "grad_norm": 0.3839147090911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6256567779484278,
+      "grad_norm": 0.5760218501091003,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6272734621291731,
+      "grad_norm": 0.7226507067680359,
+      "learning_rate": 0.0002,
+      "loss": 0.7836,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6288901463099184,
+      "grad_norm": 0.676781415939331,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6305068304906637,
+      "grad_norm": 0.4284018278121948,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6321235146714089,
+      "grad_norm": 0.5060628056526184,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6337401988521543,
+      "grad_norm": 0.5524522066116333,
+      "learning_rate": 0.0002,
+      "loss": 0.8089,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6353568830328995,
+      "grad_norm": 0.6099881529808044,
+      "learning_rate": 0.0002,
+      "loss": 0.8276,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6369735672136448,
+      "grad_norm": 0.43155938386917114,
+      "learning_rate": 0.0002,
+      "loss": 0.809,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6385902513943901,
+      "grad_norm": 0.6427084803581238,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6402069355751354,
+      "grad_norm": 0.541220486164093,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6418236197558806,
+      "grad_norm": 0.5414294600486755,
+      "learning_rate": 0.0002,
+      "loss": 0.8539,
+      "step": 3970
+    },
+    {
+      "epoch": 0.643440303936626,
+      "grad_norm": 0.46344003081321716,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6450569881173712,
+      "grad_norm": 0.45209285616874695,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6466736722981166,
+      "grad_norm": 0.5417284369468689,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6482903564788619,
+      "grad_norm": 0.7995685935020447,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6499070406596071,
+      "grad_norm": 0.6384002566337585,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6515237248403525,
+      "grad_norm": 0.4472815692424774,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6531404090210977,
+      "grad_norm": 0.6834294199943542,
+      "learning_rate": 0.0002,
+      "loss": 0.8071,
+      "step": 4040
+    },
+    {
+      "epoch": 0.654757093201843,
+      "grad_norm": 0.4612339735031128,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6563737773825883,
+      "grad_norm": 0.9266576170921326,
+      "learning_rate": 0.0002,
+      "loss": 0.8141,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6579904615633336,
+      "grad_norm": 0.4470861852169037,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6596071457440789,
+      "grad_norm": 0.45544925332069397,
+      "learning_rate": 0.0002,
+      "loss": 0.8293,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6612238299248242,
+      "grad_norm": 0.6144481301307678,
+      "learning_rate": 0.0002,
+      "loss": 0.8455,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6628405141055694,
+      "grad_norm": 0.5936288237571716,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6644571982863148,
+      "grad_norm": 0.4822963774204254,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 4110
+    },
+    {
+      "epoch": 0.66607388246706,
+      "grad_norm": 0.48432496190071106,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6676905666478054,
+      "grad_norm": 0.4901607930660248,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6693072508285507,
+      "grad_norm": 0.5018393397331238,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6709239350092959,
+      "grad_norm": 0.6946378946304321,
+      "learning_rate": 0.0002,
+      "loss": 0.8065,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6725406191900413,
+      "grad_norm": 0.5997390747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.6738849878311157,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6757739875515318,
+      "grad_norm": 0.6110581159591675,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6773906717322771,
+      "grad_norm": 0.5703322291374207,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6790073559130224,
+      "grad_norm": 0.4686066210269928,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6806240400937676,
+      "grad_norm": 0.6394643783569336,
+      "learning_rate": 0.0002,
+      "loss": 0.8441,
+      "step": 4210
+    },
+    {
+      "epoch": 0.682240724274513,
+      "grad_norm": 0.5454841256141663,
+      "learning_rate": 0.0002,
+      "loss": 0.8011,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6838574084552582,
+      "grad_norm": 0.4859732985496521,
+      "learning_rate": 0.0002,
+      "loss": 0.8307,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6854740926360036,
+      "grad_norm": 0.5544065833091736,
+      "learning_rate": 0.0002,
+      "loss": 0.8161,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6870907768167488,
+      "grad_norm": 0.4902505576610565,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6887074609974941,
+      "grad_norm": 0.4768051505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7977,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6903241451782395,
+      "grad_norm": 0.49982190132141113,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6919408293589847,
+      "grad_norm": 0.6351838111877441,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 4280
+    },
+    {
+      "epoch": 0.69355751353973,
+      "grad_norm": 0.5647561550140381,
+      "learning_rate": 0.0002,
+      "loss": 0.7664,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6951741977204753,
+      "grad_norm": 0.5340486764907837,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6967908819012206,
+      "grad_norm": 0.5649092793464661,
+      "learning_rate": 0.0002,
+      "loss": 0.8526,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6984075660819659,
+      "grad_norm": 0.6183916926383972,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7000242502627112,
+      "grad_norm": 0.6154509782791138,
+      "learning_rate": 0.0002,
+      "loss": 0.792,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7016409344434564,
+      "grad_norm": 0.5156264305114746,
+      "learning_rate": 0.0002,
+      "loss": 0.8397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7032576186242018,
+      "grad_norm": 0.562171459197998,
+      "learning_rate": 0.0002,
+      "loss": 0.8512,
+      "step": 4350
+    },
+    {
+      "epoch": 0.704874302804947,
+      "grad_norm": 0.4949502646923065,
+      "learning_rate": 0.0002,
+      "loss": 0.7882,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7064909869856923,
+      "grad_norm": 0.5171684622764587,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7081076711664377,
+      "grad_norm": 0.6198443174362183,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7097243553471829,
+      "grad_norm": 0.5802276134490967,
+      "learning_rate": 0.0002,
+      "loss": 0.7606,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7113410395279283,
+      "grad_norm": 0.41096967458724976,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7129577237086735,
+      "grad_norm": 0.4397392272949219,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7145744078894188,
+      "grad_norm": 0.45228442549705505,
+      "learning_rate": 0.0002,
+      "loss": 0.7651,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7161910920701641,
+      "grad_norm": 0.4839673936367035,
+      "learning_rate": 0.0002,
+      "loss": 0.7938,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7178077762509094,
+      "grad_norm": 0.6140755414962769,
+      "learning_rate": 0.0002,
+      "loss": 0.8362,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 0.6841378808021545,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7210411446124,
+      "grad_norm": 0.6664239168167114,
+      "learning_rate": 0.0002,
+      "loss": 0.8177,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7226578287931452,
+      "grad_norm": 0.47552719712257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7983,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7242745129738906,
+      "grad_norm": 0.6649776101112366,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7258911971546358,
+      "grad_norm": 0.5159541964530945,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7275078813353811,
+      "grad_norm": 0.6693112850189209,
+      "learning_rate": 0.0002,
+      "loss": 0.7786,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7291245655161265,
+      "grad_norm": 0.48870977759361267,
+      "learning_rate": 0.0002,
+      "loss": 0.8655,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7307412496968717,
+      "grad_norm": 0.4857887923717499,
+      "learning_rate": 0.0002,
+      "loss": 0.7337,
+      "step": 4520
+    },
+    {
+      "epoch": 0.732357933877617,
+      "grad_norm": 0.5515662431716919,
+      "learning_rate": 0.0002,
+      "loss": 0.8026,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7339746180583623,
+      "grad_norm": 0.6292222738265991,
+      "learning_rate": 0.0002,
+      "loss": 0.8031,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7355913022391076,
+      "grad_norm": 0.48265689611434937,
+      "learning_rate": 0.0002,
+      "loss": 0.7749,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7372079864198529,
+      "grad_norm": 0.8044266104698181,
+      "learning_rate": 0.0002,
+      "loss": 0.8499,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7388246706005982,
+      "grad_norm": 0.6111769676208496,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7404413547813434,
+      "grad_norm": 0.5229553580284119,
+      "learning_rate": 0.0002,
+      "loss": 0.7291,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7420580389620888,
+      "grad_norm": 0.6054152250289917,
+      "learning_rate": 0.0002,
+      "loss": 0.8038,
+      "step": 4590
+    },
+    {
+      "epoch": 0.743674723142834,
+      "grad_norm": 0.5574966669082642,
+      "learning_rate": 0.0002,
+      "loss": 0.8169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7452914073235793,
+      "grad_norm": 0.5395817160606384,
+      "learning_rate": 0.0002,
+      "loss": 0.8439,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7469080915043246,
+      "grad_norm": 0.7116472721099854,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7485247756850699,
+      "grad_norm": 0.5618700981140137,
+      "learning_rate": 0.0002,
+      "loss": 0.7743,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7501414598658153,
+      "grad_norm": 0.5802770853042603,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7517581440465605,
+      "grad_norm": 0.5690428018569946,
+      "learning_rate": 0.0002,
+      "loss": 0.7924,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7533748282273058,
+      "grad_norm": 0.4813360273838043,
+      "learning_rate": 0.0002,
+      "loss": 0.8017,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7549915124080511,
+      "grad_norm": 0.5434042811393738,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7566081965887964,
+      "grad_norm": 0.5502099990844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7582248807695416,
+      "grad_norm": 0.6020621061325073,
+      "learning_rate": 0.0002,
+      "loss": 0.8598,
+      "step": 4690
+    },
+    {
+      "epoch": 0.759841564950287,
+      "grad_norm": 0.4922301471233368,
+      "learning_rate": 0.0002,
+      "loss": 0.7937,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7614582491310322,
+      "grad_norm": 0.6492828726768494,
+      "learning_rate": 0.0002,
+      "loss": 0.788,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7630749333117776,
+      "grad_norm": 0.4865580201148987,
+      "learning_rate": 0.0002,
+      "loss": 0.8313,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7646916174925228,
+      "grad_norm": 0.5971422791481018,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7663083016732681,
+      "grad_norm": 0.6832674145698547,
+      "learning_rate": 0.0002,
+      "loss": 0.8298,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7679249858540134,
+      "grad_norm": 0.500908613204956,
+      "learning_rate": 0.0002,
+      "loss": 0.8156,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7695416700347587,
+      "grad_norm": 0.6112465858459473,
+      "learning_rate": 0.0002,
+      "loss": 0.8383,
+      "step": 4760
+    },
+    {
+      "epoch": 0.771158354215504,
+      "grad_norm": 0.5753506422042847,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7727750383962493,
+      "grad_norm": 0.6529405117034912,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7743917225769946,
+      "grad_norm": 0.5916843414306641,
+      "learning_rate": 0.0002,
+      "loss": 0.8171,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7760084067577399,
+      "grad_norm": 0.4821224510669708,
+      "learning_rate": 0.0002,
+      "loss": 0.83,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7776250909384852,
+      "grad_norm": 0.5532580018043518,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7792417751192304,
+      "grad_norm": 0.4604877233505249,
+      "learning_rate": 0.0002,
+      "loss": 0.7363,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7808584592999758,
+      "grad_norm": 0.5009613037109375,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 4830
+    },
+    {
+      "epoch": 0.782475143480721,
+      "grad_norm": 0.6448560357093811,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7840918276614663,
+      "grad_norm": 0.44327953457832336,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7857085118422116,
+      "grad_norm": 0.5355411171913147,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7873251960229569,
+      "grad_norm": 0.5635677576065063,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7889418802037023,
+      "grad_norm": 0.5417491793632507,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7905585643844475,
+      "grad_norm": 0.4567430913448334,
+      "learning_rate": 0.0002,
+      "loss": 0.7819,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7921752485651928,
+      "grad_norm": 0.44651296734809875,
+      "learning_rate": 0.0002,
+      "loss": 0.8454,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7937919327459381,
+      "grad_norm": 0.5741217136383057,
+      "learning_rate": 0.0002,
+      "loss": 0.7959,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7954086169266834,
+      "grad_norm": 0.6605045199394226,
+      "learning_rate": 0.0002,
+      "loss": 0.8093,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7970253011074286,
+      "grad_norm": 0.5126531720161438,
+      "learning_rate": 0.0002,
+      "loss": 0.77,
+      "step": 4930
+    },
+    {
+      "epoch": 0.798641985288174,
+      "grad_norm": 0.513648271560669,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8002586694689192,
+      "grad_norm": 0.5350404381752014,
+      "learning_rate": 0.0002,
+      "loss": 0.8314,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8018753536496646,
+      "grad_norm": 0.5731674432754517,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8034920378304098,
+      "grad_norm": 0.5974258184432983,
+      "learning_rate": 0.0002,
+      "loss": 0.8572,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8051087220111551,
+      "grad_norm": 0.8774799704551697,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8067254061919004,
+      "grad_norm": 0.5994430184364319,
+      "learning_rate": 0.0002,
+      "loss": 0.7899,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8083420903726457,
+      "grad_norm": 0.4894903004169464,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5000
+    },
+    {
+      "epoch": 0.809958774553391,
+      "grad_norm": 0.5218459367752075,
+      "learning_rate": 0.0002,
+      "loss": 0.78,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8115754587341363,
+      "grad_norm": 0.5232468843460083,
+      "learning_rate": 0.0002,
+      "loss": 0.817,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8131921429148816,
+      "grad_norm": 0.44358372688293457,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8148088270956269,
+      "grad_norm": 0.6202037334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8164255112763722,
+      "grad_norm": 0.7721474170684814,
+      "learning_rate": 0.0002,
+      "loss": 0.7351,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8180421954571174,
+      "grad_norm": 0.5568501353263855,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8196588796378628,
+      "grad_norm": 0.49148809909820557,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 5070
+    },
+    {
+      "epoch": 0.821275563818608,
+      "grad_norm": 0.4956012964248657,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8228922479993533,
+      "grad_norm": 0.6078833937644958,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8245089321800986,
+      "grad_norm": 0.46906954050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.828,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8261256163608439,
+      "grad_norm": 0.50812166929245,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8277423005415891,
+      "grad_norm": 0.5319661498069763,
+      "learning_rate": 0.0002,
+      "loss": 0.8243,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8293589847223345,
+      "grad_norm": 0.4949689209461212,
+      "learning_rate": 0.0002,
+      "loss": 0.7798,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8309756689030798,
+      "grad_norm": 0.5151591300964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8325923530838251,
+      "grad_norm": 0.5530214309692383,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8342090372645704,
+      "grad_norm": 0.6297410130500793,
+      "learning_rate": 0.0002,
+      "loss": 0.8251,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8358257214453156,
+      "grad_norm": 0.5466840267181396,
+      "learning_rate": 0.0002,
+      "loss": 0.8067,
+      "step": 5170
+    },
+    {
+      "epoch": 0.837442405626061,
+      "grad_norm": 0.652913510799408,
+      "learning_rate": 0.0002,
+      "loss": 0.7875,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8390590898068062,
+      "grad_norm": 0.5811293125152588,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8406757739875516,
+      "grad_norm": 0.5109550952911377,
+      "learning_rate": 0.0002,
+      "loss": 0.7412,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8422924581682968,
+      "grad_norm": 0.4551706612110138,
+      "learning_rate": 0.0002,
+      "loss": 0.8077,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8439091423490421,
+      "grad_norm": 0.5813754200935364,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8455258265297874,
+      "grad_norm": 0.5856947898864746,
+      "learning_rate": 0.0002,
+      "loss": 0.802,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8471425107105327,
+      "grad_norm": 0.5482739210128784,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8487591948912779,
+      "grad_norm": 0.49023720622062683,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8503758790720233,
+      "grad_norm": 0.49472475051879883,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8519925632527686,
+      "grad_norm": 0.5490226745605469,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8536092474335139,
+      "grad_norm": 0.5340665578842163,
+      "learning_rate": 0.0002,
+      "loss": 0.8333,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8552259316142592,
+      "grad_norm": 0.5962483882904053,
+      "learning_rate": 0.0002,
+      "loss": 0.8277,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8568426157950044,
+      "grad_norm": 0.586358368396759,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8584592999757498,
+      "grad_norm": 0.49120277166366577,
+      "learning_rate": 0.0002,
+      "loss": 0.7831,
+      "step": 5310
+    },
+    {
+      "epoch": 0.860075984156495,
+      "grad_norm": 0.5887332558631897,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8616926683372403,
+      "grad_norm": 0.42496153712272644,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.5489874482154846,
+      "learning_rate": 0.0002,
+      "loss": 0.7905,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8649260366987309,
+      "grad_norm": 0.5850813984870911,
+      "learning_rate": 0.0002,
+      "loss": 0.7958,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8665427208794761,
+      "grad_norm": 0.517487108707428,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8681594050602215,
+      "grad_norm": 0.5339142680168152,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8697760892409668,
+      "grad_norm": 0.6236387491226196,
+      "learning_rate": 0.0002,
+      "loss": 0.818,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8713927734217121,
+      "grad_norm": 0.5752192735671997,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8730094576024574,
+      "grad_norm": 0.6724614500999451,
+      "learning_rate": 0.0002,
+      "loss": 0.8542,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8746261417832026,
+      "grad_norm": 0.5280613303184509,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 5410
+    },
+    {
+      "epoch": 0.876242825963948,
+      "grad_norm": 0.44033288955688477,
+      "learning_rate": 0.0002,
+      "loss": 0.8231,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8778595101446932,
+      "grad_norm": 0.5199708342552185,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8794761943254386,
+      "grad_norm": 0.46778348088264465,
+      "learning_rate": 0.0002,
+      "loss": 0.7852,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8810928785061838,
+      "grad_norm": 0.4657754898071289,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8827095626869291,
+      "grad_norm": 0.5472902655601501,
+      "learning_rate": 0.0002,
+      "loss": 0.7799,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8843262468676744,
+      "grad_norm": 0.4876766800880432,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8859429310484197,
+      "grad_norm": 0.5057248473167419,
+      "learning_rate": 0.0002,
+      "loss": 0.7906,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8875596152291649,
+      "grad_norm": 0.4637320637702942,
+      "learning_rate": 0.0002,
+      "loss": 0.8124,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8891762994099103,
+      "grad_norm": 0.471955806016922,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8907929835906556,
+      "grad_norm": 0.5209813714027405,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8924096677714008,
+      "grad_norm": 0.6213834285736084,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8940263519521462,
+      "grad_norm": 0.5215408205986023,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8956430361328914,
+      "grad_norm": 0.580478310585022,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8972597203136368,
+      "grad_norm": 0.49102169275283813,
+      "learning_rate": 0.0002,
+      "loss": 0.8371,
+      "step": 5550
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.6043479442596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7806,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9004930886751273,
+      "grad_norm": 0.5636463165283203,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9021097728558726,
+      "grad_norm": 0.5620124340057373,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9037264570366179,
+      "grad_norm": 0.5206354856491089,
+      "learning_rate": 0.0002,
+      "loss": 0.8083,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9053431412173631,
+      "grad_norm": 0.5798229575157166,
+      "learning_rate": 0.0002,
+      "loss": 0.8557,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9069598253981085,
+      "grad_norm": 0.6428212523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.8097,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9085765095788537,
+      "grad_norm": 0.48064687848091125,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9101931937595991,
+      "grad_norm": 0.6347860097885132,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9118098779403444,
+      "grad_norm": 0.5353913307189941,
+      "learning_rate": 0.0002,
+      "loss": 0.851,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9134265621210896,
+      "grad_norm": 0.5323944091796875,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5650
+    },
+    {
+      "epoch": 0.915043246301835,
+      "grad_norm": 0.5261843204498291,
+      "learning_rate": 0.0002,
+      "loss": 0.8393,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9166599304825802,
+      "grad_norm": 0.5451326966285706,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9182766146633256,
+      "grad_norm": 0.5183324217796326,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9198932988440708,
+      "grad_norm": 0.47229018807411194,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9215099830248161,
+      "grad_norm": 0.49180513620376587,
+      "learning_rate": 0.0002,
+      "loss": 0.7757,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9231266672055614,
+      "grad_norm": 0.5419785380363464,
+      "learning_rate": 0.0002,
+      "loss": 0.8735,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9247433513863067,
+      "grad_norm": 0.5408698916435242,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9263600355670519,
+      "grad_norm": 0.5286232829093933,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9279767197477973,
+      "grad_norm": 0.7539758086204529,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9295934039285425,
+      "grad_norm": 0.5166944861412048,
+      "learning_rate": 0.0002,
+      "loss": 0.8118,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9312100881092878,
+      "grad_norm": 0.6601425409317017,
+      "learning_rate": 0.0002,
+      "loss": 0.783,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9328267722900332,
+      "grad_norm": 0.5029960870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9344434564707784,
+      "grad_norm": 0.4926645755767822,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9360601406515238,
+      "grad_norm": 0.5739615559577942,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5790
+    },
+    {
+      "epoch": 0.937676824832269,
+      "grad_norm": 0.5058279037475586,
+      "learning_rate": 0.0002,
+      "loss": 0.8037,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9392935090130143,
+      "grad_norm": 0.5260962247848511,
+      "learning_rate": 0.0002,
+      "loss": 0.8537,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9409101931937596,
+      "grad_norm": 0.5768588185310364,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9425268773745049,
+      "grad_norm": 0.5170126557350159,
+      "learning_rate": 0.0002,
+      "loss": 0.8215,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9441435615552501,
+      "grad_norm": 0.5745864510536194,
+      "learning_rate": 0.0002,
+      "loss": 0.7422,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9457602457359955,
+      "grad_norm": 0.5551357865333557,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9473769299167407,
+      "grad_norm": 0.5776078701019287,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9489936140974861,
+      "grad_norm": 0.5340062379837036,
+      "learning_rate": 0.0002,
+      "loss": 0.8527,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9506102982782314,
+      "grad_norm": 0.6447290182113647,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9522269824589766,
+      "grad_norm": 0.5123815536499023,
+      "learning_rate": 0.0002,
+      "loss": 0.7945,
+      "step": 5890
+    },
+    {
+      "epoch": 0.953843666639722,
+      "grad_norm": 0.48547613620758057,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9554603508204672,
+      "grad_norm": 0.5791414976119995,
+      "learning_rate": 0.0002,
+      "loss": 0.7896,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9570770350012126,
+      "grad_norm": 0.6195011734962463,
+      "learning_rate": 0.0002,
+      "loss": 0.8408,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9586937191819578,
+      "grad_norm": 0.6323803067207336,
+      "learning_rate": 0.0002,
+      "loss": 0.7805,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9603104033627031,
+      "grad_norm": 0.45552879571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9619270875434484,
+      "grad_norm": 0.5796473622322083,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9635437717241937,
+      "grad_norm": 0.647261381149292,
+      "learning_rate": 0.0002,
+      "loss": 0.7672,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9651604559049389,
+      "grad_norm": 0.5487682819366455,
+      "learning_rate": 0.0002,
+      "loss": 0.8086,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9667771400856843,
+      "grad_norm": 0.5743663907051086,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9683938242664295,
+      "grad_norm": 0.5470591187477112,
+      "learning_rate": 0.0002,
+      "loss": 0.8153,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9700105084471748,
+      "grad_norm": 0.5901660323143005,
+      "learning_rate": 0.0002,
+      "loss": 0.8119,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9716271926279202,
+      "grad_norm": 0.6544759273529053,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9732438768086654,
+      "grad_norm": 0.6288470029830933,
+      "learning_rate": 0.0002,
+      "loss": 0.7536,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9748605609894108,
+      "grad_norm": 0.673153817653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 6030
+    },
+    {
+      "epoch": 0.976477245170156,
+      "grad_norm": 0.42854753136634827,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9780939293509013,
+      "grad_norm": 0.5227066278457642,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9797106135316466,
+      "grad_norm": 0.5372416973114014,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9813272977123919,
+      "grad_norm": 0.6026402115821838,
+      "learning_rate": 0.0002,
+      "loss": 0.7591,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9829439818931371,
+      "grad_norm": 0.49547791481018066,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9845606660738825,
+      "grad_norm": 0.4641951322555542,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9861773502546277,
+      "grad_norm": 0.5818535089492798,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9877940344353731,
+      "grad_norm": 0.63955157995224,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9894107186161183,
+      "grad_norm": 0.5649438500404358,
+      "learning_rate": 0.0002,
+      "loss": 0.7547,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9910274027968636,
+      "grad_norm": 0.5290433168411255,
+      "learning_rate": 0.0002,
+      "loss": 0.7861,
+      "step": 6130
+    },
+    {
+      "epoch": 0.992644086977609,
+      "grad_norm": 0.6399374008178711,
+      "learning_rate": 0.0002,
+      "loss": 0.8109,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9942607711583542,
+      "grad_norm": 0.6736576557159424,
+      "learning_rate": 0.0002,
+      "loss": 0.8373,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9958774553390995,
+      "grad_norm": 0.515420138835907,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9974941395198448,
+      "grad_norm": 0.562677800655365,
+      "learning_rate": 0.0002,
+      "loss": 0.8032,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9991108237005901,
+      "grad_norm": 0.7113858461380005,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9999191657909627,
+      "eval_loss": 1.0871200561523438,
+      "eval_runtime": 122.2071,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0007275078813354,
+      "grad_norm": 0.7111801505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0023441920620806,
+      "grad_norm": 0.5402125716209412,
+      "learning_rate": 0.0002,
+      "loss": 0.6865,
+      "step": 6200
+    },
+    {
+      "epoch": 1.003960876242826,
+      "grad_norm": 0.6098830103874207,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0055775604235713,
+      "grad_norm": 0.5829983353614807,
+      "learning_rate": 0.0002,
+      "loss": 0.7631,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0071942446043165,
+      "grad_norm": 0.5614621043205261,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0088109287850617,
+      "grad_norm": 0.5954238772392273,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0104276129658072,
+      "grad_norm": 0.6480574607849121,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0120442971465524,
+      "grad_norm": 0.6051128506660461,
+      "learning_rate": 0.0002,
+      "loss": 0.7514,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0136609813272976,
+      "grad_norm": 0.6318870782852173,
+      "learning_rate": 0.0002,
+      "loss": 0.7237,
+      "step": 6270
+    },
+    {
+      "epoch": 1.015277665508043,
+      "grad_norm": 0.5048980116844177,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0168943496887883,
+      "grad_norm": 0.6346936225891113,
+      "learning_rate": 0.0002,
+      "loss": 0.7391,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0185110338695336,
+      "grad_norm": 0.5711665749549866,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0201277180502788,
+      "grad_norm": 0.5175361037254333,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0217444022310243,
+      "grad_norm": 0.5360831618309021,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0233610864117695,
+      "grad_norm": 0.614675760269165,
+      "learning_rate": 0.0002,
+      "loss": 0.7112,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0249777705925147,
+      "grad_norm": 0.5626118183135986,
+      "learning_rate": 0.0002,
+      "loss": 0.7748,
+      "step": 6340
+    },
+    {
+      "epoch": 1.02659445477326,
+      "grad_norm": 0.574897289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.7375,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0282111389540054,
+      "grad_norm": 0.7185447812080383,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0298278231347506,
+      "grad_norm": 0.6705799698829651,
+      "learning_rate": 0.0002,
+      "loss": 0.703,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0314445073154959,
+      "grad_norm": 0.6740428805351257,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0330611914962413,
+      "grad_norm": 0.663902759552002,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0346778756769865,
+      "grad_norm": 0.5029543042182922,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0362945598577318,
+      "grad_norm": 0.7813863158226013,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 6410
+    },
+    {
+      "epoch": 1.037911244038477,
+      "grad_norm": 0.5396282076835632,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0395279282192225,
+      "grad_norm": 0.5253293514251709,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0411446123999677,
+      "grad_norm": 0.7236770987510681,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 6440
+    },
+    {
+      "epoch": 1.042761296580713,
+      "grad_norm": 0.5670917630195618,
+      "learning_rate": 0.0002,
+      "loss": 0.7259,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0443779807614582,
+      "grad_norm": 0.6031978726387024,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0459946649422036,
+      "grad_norm": 0.5309213399887085,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0476113491229488,
+      "grad_norm": 0.7114651799201965,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 6480
+    },
+    {
+      "epoch": 1.049228033303694,
+      "grad_norm": 0.5591610670089722,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0508447174844395,
+      "grad_norm": 0.5185961127281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0524614016651848,
+      "grad_norm": 0.6510552167892456,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6510
+    },
+    {
+      "epoch": 1.05407808584593,
+      "grad_norm": 0.6557928919792175,
+      "learning_rate": 0.0002,
+      "loss": 0.7057,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0556947700266752,
+      "grad_norm": 0.6973192691802979,
+      "learning_rate": 0.0002,
+      "loss": 0.8056,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0573114542074207,
+      "grad_norm": 0.6226583123207092,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 6540
+    },
+    {
+      "epoch": 1.058928138388166,
+      "grad_norm": 0.5633195638656616,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0605448225689111,
+      "grad_norm": 0.7466658353805542,
+      "learning_rate": 0.0002,
+      "loss": 0.7082,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0621615067496564,
+      "grad_norm": 0.6462772488594055,
+      "learning_rate": 0.0002,
+      "loss": 0.7059,
+      "step": 6570
+    },
+    {
+      "epoch": 1.0637781909304018,
+      "grad_norm": 0.5266856551170349,
+      "learning_rate": 0.0002,
+      "loss": 0.7046,
+      "step": 6580
+    },
+    {
+      "epoch": 1.065394875111147,
+      "grad_norm": 0.534392774105072,
+      "learning_rate": 0.0002,
+      "loss": 0.7157,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0670115592918923,
+      "grad_norm": 0.7514177560806274,
+      "learning_rate": 0.0002,
+      "loss": 0.7115,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0686282434726375,
+      "grad_norm": 0.7593035697937012,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 6610
+    },
+    {
+      "epoch": 1.070244927653383,
+      "grad_norm": 0.5277858972549438,
+      "learning_rate": 0.0002,
+      "loss": 0.6836,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0718616118341282,
+      "grad_norm": 0.5573670268058777,
+      "learning_rate": 0.0002,
+      "loss": 0.7405,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0734782960148734,
+      "grad_norm": 0.6802396774291992,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 6640
+    },
+    {
+      "epoch": 1.0750949801956189,
+      "grad_norm": 0.7367215752601624,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0767116643763641,
+      "grad_norm": 0.5961891412734985,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0783283485571094,
+      "grad_norm": 0.5736313462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0799450327378546,
+      "grad_norm": 0.619219183921814,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0815617169186,
+      "grad_norm": 0.6214390993118286,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0831784010993453,
+      "grad_norm": 0.564536988735199,
+      "learning_rate": 0.0002,
+      "loss": 0.7554,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0847950852800905,
+      "grad_norm": 0.5838140249252319,
+      "learning_rate": 0.0002,
+      "loss": 0.7362,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0864117694608357,
+      "grad_norm": 0.7000553607940674,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0880284536415812,
+      "grad_norm": 0.7078263759613037,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 6730
+    },
+    {
+      "epoch": 1.0896451378223264,
+      "grad_norm": 0.8353848457336426,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6740
+    },
+    {
+      "epoch": 1.0912618220030716,
+      "grad_norm": 0.5615518689155579,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 6750
+    },
+    {
+      "epoch": 1.0928785061838169,
+      "grad_norm": 0.5475581288337708,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 6760
+    },
+    {
+      "epoch": 1.0944951903645623,
+      "grad_norm": 0.5835978388786316,
+      "learning_rate": 0.0002,
+      "loss": 0.7652,
+      "step": 6770
+    },
+    {
+      "epoch": 1.0961118745453076,
+      "grad_norm": 0.5516105890274048,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 6780
+    },
+    {
+      "epoch": 1.0977285587260528,
+      "grad_norm": 0.5875251889228821,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 6790
+    },
+    {
+      "epoch": 1.0993452429067982,
+      "grad_norm": 0.7376947999000549,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1009619270875435,
+      "grad_norm": 0.5656165480613708,
+      "learning_rate": 0.0002,
+      "loss": 0.7512,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1025786112682887,
+      "grad_norm": 0.6365954279899597,
+      "learning_rate": 0.0002,
+      "loss": 0.7409,
+      "step": 6820
+    },
+    {
+      "epoch": 1.104195295449034,
+      "grad_norm": 0.5033080577850342,
+      "learning_rate": 0.0002,
+      "loss": 0.7392,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1058119796297794,
+      "grad_norm": 0.617396891117096,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1074286638105246,
+      "grad_norm": 0.6395374536514282,
+      "learning_rate": 0.0002,
+      "loss": 0.7006,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1090453479912699,
+      "grad_norm": 0.6775295734405518,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1106620321720153,
+      "grad_norm": 0.6655223965644836,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1122787163527605,
+      "grad_norm": 0.676655113697052,
+      "learning_rate": 0.0002,
+      "loss": 0.7553,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1138954005335058,
+      "grad_norm": 0.6062718629837036,
+      "learning_rate": 0.0002,
+      "loss": 0.7342,
+      "step": 6890
+    },
+    {
+      "epoch": 1.115512084714251,
+      "grad_norm": 0.590943455696106,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1171287688949965,
+      "grad_norm": 0.6315317153930664,
+      "learning_rate": 0.0002,
+      "loss": 0.6705,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1187454530757417,
+      "grad_norm": 0.47979024052619934,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 6920
+    },
+    {
+      "epoch": 1.120362137256487,
+      "grad_norm": 0.647298276424408,
+      "learning_rate": 0.0002,
+      "loss": 0.7002,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1219788214372322,
+      "grad_norm": 0.7336484789848328,
+      "learning_rate": 0.0002,
+      "loss": 0.7502,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1235955056179776,
+      "grad_norm": 0.5071424245834351,
+      "learning_rate": 0.0002,
+      "loss": 0.693,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1252121897987228,
+      "grad_norm": 0.6527144312858582,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 6960
+    },
+    {
+      "epoch": 1.126828873979468,
+      "grad_norm": 0.6935935020446777,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1284455581602133,
+      "grad_norm": 0.8026931881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1300622423409588,
+      "grad_norm": 0.5210393667221069,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 6990
+    },
+    {
+      "epoch": 1.131678926521704,
+      "grad_norm": 0.60475093126297,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1332956107024492,
+      "grad_norm": 0.6417073607444763,
+      "learning_rate": 0.0002,
+      "loss": 0.7495,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1349122948831947,
+      "grad_norm": 0.6732175946235657,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 7020
+    },
+    {
+      "epoch": 1.13652897906394,
+      "grad_norm": 0.6719491481781006,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1381456632446851,
+      "grad_norm": 0.5708295106887817,
+      "learning_rate": 0.0002,
+      "loss": 0.7694,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1397623474254304,
+      "grad_norm": 0.7141719460487366,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1413790316061758,
+      "grad_norm": 0.6187017560005188,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 7060
+    },
+    {
+      "epoch": 1.142995715786921,
+      "grad_norm": 0.50581294298172,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1446123999676663,
+      "grad_norm": 0.5620143413543701,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1462290841484115,
+      "grad_norm": 0.6231929659843445,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 7090
+    },
+    {
+      "epoch": 1.147845768329157,
+      "grad_norm": 0.5775774121284485,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1494624525099022,
+      "grad_norm": 0.6492809653282166,
+      "learning_rate": 0.0002,
+      "loss": 0.7728,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1510791366906474,
+      "grad_norm": 0.6434972286224365,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 7120
+    },
+    {
+      "epoch": 1.1526958208713927,
+      "grad_norm": 0.6191812753677368,
+      "learning_rate": 0.0002,
+      "loss": 0.7374,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1543125050521381,
+      "grad_norm": 0.6690331697463989,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 7140
+    },
+    {
+      "epoch": 1.1559291892328833,
+      "grad_norm": 0.5977938175201416,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1575458734136286,
+      "grad_norm": 0.6195854544639587,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7160
+    },
+    {
+      "epoch": 1.159162557594374,
+      "grad_norm": 0.5752048492431641,
+      "learning_rate": 0.0002,
+      "loss": 0.7249,
+      "step": 7170
+    },
+    {
+      "epoch": 1.1607792417751193,
+      "grad_norm": 0.589081883430481,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1623959259558645,
+      "grad_norm": 0.756996750831604,
+      "learning_rate": 0.0002,
+      "loss": 0.704,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1640126101366097,
+      "grad_norm": 0.7614967226982117,
+      "learning_rate": 0.0002,
+      "loss": 0.7404,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1656292943173552,
+      "grad_norm": 0.6120437979698181,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1672459784981004,
+      "grad_norm": 0.6210004687309265,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1688626626788456,
+      "grad_norm": 0.6044116020202637,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7230
+    },
+    {
+      "epoch": 1.170479346859591,
+      "grad_norm": 0.5418457388877869,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1720960310403363,
+      "grad_norm": 0.6413537263870239,
+      "learning_rate": 0.0002,
+      "loss": 0.6938,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1737127152210816,
+      "grad_norm": 0.5777867436408997,
+      "learning_rate": 0.0002,
+      "loss": 0.6978,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1753293994018268,
+      "grad_norm": 0.7092402577400208,
+      "learning_rate": 0.0002,
+      "loss": 0.7503,
+      "step": 7270
+    },
+    {
+      "epoch": 1.176946083582572,
+      "grad_norm": 0.6351709365844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1785627677633175,
+      "grad_norm": 0.6172189712524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7527,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1801794519440627,
+      "grad_norm": 0.6801714897155762,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7300
+    },
+    {
+      "epoch": 1.181796136124808,
+      "grad_norm": 0.6044712066650391,
+      "learning_rate": 0.0002,
+      "loss": 0.6941,
+      "step": 7310
+    },
+    {
+      "epoch": 1.1834128203055534,
+      "grad_norm": 0.7413212060928345,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 7320
+    },
+    {
+      "epoch": 1.1850295044862986,
+      "grad_norm": 0.5303856134414673,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 7330
+    },
+    {
+      "epoch": 1.1866461886670439,
+      "grad_norm": 0.5647098422050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6915,
+      "step": 7340
+    },
+    {
+      "epoch": 1.188262872847789,
+      "grad_norm": 0.7374135255813599,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 7350
+    },
+    {
+      "epoch": 1.1898795570285345,
+      "grad_norm": 0.5710089206695557,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1914962412092798,
+      "grad_norm": 0.6073619723320007,
+      "learning_rate": 0.0002,
+      "loss": 0.8289,
+      "step": 7370
+    },
+    {
+      "epoch": 1.193112925390025,
+      "grad_norm": 0.5899916887283325,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1947296095707705,
+      "grad_norm": 0.7762434482574463,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1963462937515157,
+      "grad_norm": 0.679949939250946,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7400
+    },
+    {
+      "epoch": 1.197962977932261,
+      "grad_norm": 0.6106849312782288,
+      "learning_rate": 0.0002,
+      "loss": 0.7599,
+      "step": 7410
+    },
+    {
+      "epoch": 1.1995796621130062,
+      "grad_norm": 0.682461678981781,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2011963462937516,
+      "grad_norm": 0.6087017059326172,
+      "learning_rate": 0.0002,
+      "loss": 0.7741,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2028130304744968,
+      "grad_norm": 0.63739013671875,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 7440
+    },
+    {
+      "epoch": 1.204429714655242,
+      "grad_norm": 0.6154777407646179,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2060463988359873,
+      "grad_norm": 0.7491534948348999,
+      "learning_rate": 0.0002,
+      "loss": 0.7565,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2076630830167328,
+      "grad_norm": 0.6664797067642212,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 7470
+    },
+    {
+      "epoch": 1.209279767197478,
+      "grad_norm": 0.6660266518592834,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2108964513782232,
+      "grad_norm": 0.6972551345825195,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2125131355589684,
+      "grad_norm": 0.6157945990562439,
+      "learning_rate": 0.0002,
+      "loss": 0.7023,
+      "step": 7500
+    },
+    {
+      "epoch": 1.214129819739714,
+      "grad_norm": 0.5199310183525085,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2157465039204591,
+      "grad_norm": 0.577610433101654,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2173631881012044,
+      "grad_norm": 0.53652423620224,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2189798722819498,
+      "grad_norm": 0.6479050517082214,
+      "learning_rate": 0.0002,
+      "loss": 0.7393,
+      "step": 7540
+    },
+    {
+      "epoch": 1.220596556462695,
+      "grad_norm": 0.618748128414154,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2222132406434403,
+      "grad_norm": 0.6311424374580383,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2238299248241855,
+      "grad_norm": 0.6595825552940369,
+      "learning_rate": 0.0002,
+      "loss": 0.7272,
+      "step": 7570
+    },
+    {
+      "epoch": 1.225446609004931,
+      "grad_norm": 0.5198960900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2270632931856762,
+      "grad_norm": 0.578650712966919,
+      "learning_rate": 0.0002,
+      "loss": 0.674,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2286799773664214,
+      "grad_norm": 0.6080220937728882,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2302966615471669,
+      "grad_norm": 0.7050248384475708,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2319133457279121,
+      "grad_norm": 0.6652196049690247,
+      "learning_rate": 0.0002,
+      "loss": 0.7032,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2335300299086573,
+      "grad_norm": 0.7322776317596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7085,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2351467140894026,
+      "grad_norm": 0.4998728036880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2367633982701478,
+      "grad_norm": 0.6428788900375366,
+      "learning_rate": 0.0002,
+      "loss": 0.7214,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2383800824508933,
+      "grad_norm": 0.585242509841919,
+      "learning_rate": 0.0002,
+      "loss": 0.7699,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2399967666316385,
+      "grad_norm": 0.5211917757987976,
+      "learning_rate": 0.0002,
+      "loss": 0.7621,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2416134508123837,
+      "grad_norm": 0.6490384340286255,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2432301349931292,
+      "grad_norm": 0.6249763369560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2448468191738744,
+      "grad_norm": 0.71870356798172,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2464635033546196,
+      "grad_norm": 0.6761967539787292,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2480801875353649,
+      "grad_norm": 0.6500617265701294,
+      "learning_rate": 0.0002,
+      "loss": 0.7501,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2496968717161103,
+      "grad_norm": 0.8069869875907898,
+      "learning_rate": 0.0002,
+      "loss": 0.7903,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2513135558968556,
+      "grad_norm": 0.6044608950614929,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2529302400776008,
+      "grad_norm": 0.6573283076286316,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2545469242583462,
+      "grad_norm": 0.625430166721344,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2561636084390915,
+      "grad_norm": 0.5442022681236267,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7770
+    },
+    {
+      "epoch": 1.2577802926198367,
+      "grad_norm": 0.6818386912345886,
+      "learning_rate": 0.0002,
+      "loss": 0.7172,
+      "step": 7780
+    },
+    {
+      "epoch": 1.259396976800582,
+      "grad_norm": 0.6381874084472656,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2610136609813272,
+      "grad_norm": 0.6269212961196899,
+      "learning_rate": 0.0002,
+      "loss": 0.6834,
+      "step": 7800
+    },
+    {
+      "epoch": 1.2626303451620726,
+      "grad_norm": 0.600121259689331,
+      "learning_rate": 0.0002,
+      "loss": 0.7821,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2642470293428179,
+      "grad_norm": 0.6337703466415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2658637135235633,
+      "grad_norm": 0.7234963774681091,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 7830
+    },
+    {
+      "epoch": 1.2674803977043085,
+      "grad_norm": 0.800184965133667,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2690970818850538,
+      "grad_norm": 0.7539464831352234,
+      "learning_rate": 0.0002,
+      "loss": 0.7426,
+      "step": 7850
+    },
+    {
+      "epoch": 1.270713766065799,
+      "grad_norm": 0.5493760704994202,
+      "learning_rate": 0.0002,
+      "loss": 0.7496,
+      "step": 7860
+    },
+    {
+      "epoch": 1.2723304502465442,
+      "grad_norm": 0.7477145791053772,
+      "learning_rate": 0.0002,
+      "loss": 0.7537,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2739471344272897,
+      "grad_norm": 0.6366362571716309,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 7880
+    },
+    {
+      "epoch": 1.275563818608035,
+      "grad_norm": 0.7419533729553223,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2771805027887801,
+      "grad_norm": 0.6141223311424255,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 7900
+    },
+    {
+      "epoch": 1.2787971869695256,
+      "grad_norm": 0.7522598505020142,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 7910
+    },
+    {
+      "epoch": 1.2804138711502708,
+      "grad_norm": 0.6935804486274719,
+      "learning_rate": 0.0002,
+      "loss": 0.7097,
+      "step": 7920
+    },
+    {
+      "epoch": 1.282030555331016,
+      "grad_norm": 0.7239290475845337,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 7930
+    },
+    {
+      "epoch": 1.2836472395117613,
+      "grad_norm": 0.8800187110900879,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 7940
+    },
+    {
+      "epoch": 1.2852639236925067,
+      "grad_norm": 0.540458083152771,
+      "learning_rate": 0.0002,
+      "loss": 0.6991,
+      "step": 7950
+    },
+    {
+      "epoch": 1.286880607873252,
+      "grad_norm": 0.6492934226989746,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 7960
+    },
+    {
+      "epoch": 1.2884972920539972,
+      "grad_norm": 0.6543959379196167,
+      "learning_rate": 0.0002,
+      "loss": 0.7742,
+      "step": 7970
+    },
+    {
+      "epoch": 1.2901139762347427,
+      "grad_norm": 0.5804705619812012,
+      "learning_rate": 0.0002,
+      "loss": 0.7316,
+      "step": 7980
+    },
+    {
+      "epoch": 1.291730660415488,
+      "grad_norm": 0.7074727416038513,
+      "learning_rate": 0.0002,
+      "loss": 0.796,
+      "step": 7990
+    },
+    {
+      "epoch": 1.2933473445962331,
+      "grad_norm": 0.5347974300384521,
+      "learning_rate": 0.0002,
+      "loss": 0.7034,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2949640287769784,
+      "grad_norm": 0.6457298398017883,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 8010
+    },
+    {
+      "epoch": 1.2965807129577236,
+      "grad_norm": 0.6407219171524048,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8020
+    },
+    {
+      "epoch": 1.298197397138469,
+      "grad_norm": 0.828439474105835,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2998140813192143,
+      "grad_norm": 0.4840380549430847,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3014307654999595,
+      "grad_norm": 0.5921024680137634,
+      "learning_rate": 0.0002,
+      "loss": 0.7283,
+      "step": 8050
+    },
+    {
+      "epoch": 1.303047449680705,
+      "grad_norm": 0.6170315146446228,
+      "learning_rate": 0.0002,
+      "loss": 0.7477,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3046641338614502,
+      "grad_norm": 0.5374847054481506,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3062808180421954,
+      "grad_norm": 0.545758068561554,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3078975022229407,
+      "grad_norm": 0.55641770362854,
+      "learning_rate": 0.0002,
+      "loss": 0.7463,
+      "step": 8090
+    },
+    {
+      "epoch": 1.309514186403686,
+      "grad_norm": 0.6724897027015686,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3111308705844313,
+      "grad_norm": 0.6923972368240356,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3127475547651766,
+      "grad_norm": 0.5136841535568237,
+      "learning_rate": 0.0002,
+      "loss": 0.7149,
+      "step": 8120
+    },
+    {
+      "epoch": 1.314364238945922,
+      "grad_norm": 0.6766283512115479,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3159809231266673,
+      "grad_norm": 0.6283926367759705,
+      "learning_rate": 0.0002,
+      "loss": 0.7489,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3175976073074125,
+      "grad_norm": 0.644216001033783,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3192142914881577,
+      "grad_norm": 0.7827503085136414,
+      "learning_rate": 0.0002,
+      "loss": 0.7125,
+      "step": 8160
+    },
+    {
+      "epoch": 1.320830975668903,
+      "grad_norm": 0.6651390790939331,
+      "learning_rate": 0.0002,
+      "loss": 0.7271,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3224476598496484,
+      "grad_norm": 0.5547412633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7778,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3240643440303936,
+      "grad_norm": 0.6765179634094238,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 8190
+    },
+    {
+      "epoch": 1.325681028211139,
+      "grad_norm": 0.6822077035903931,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3272977123918843,
+      "grad_norm": 0.5941002368927002,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3289143965726296,
+      "grad_norm": 0.4850037097930908,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3305310807533748,
+      "grad_norm": 0.6162990927696228,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 8230
+    },
+    {
+      "epoch": 1.33214776493412,
+      "grad_norm": 0.6665613651275635,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3337644491148655,
+      "grad_norm": 0.618192732334137,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3353811332956107,
+      "grad_norm": 0.710418701171875,
+      "learning_rate": 0.0002,
+      "loss": 0.7532,
+      "step": 8260
+    },
+    {
+      "epoch": 1.336997817476356,
+      "grad_norm": 0.5109876990318298,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3386145016571014,
+      "grad_norm": 0.6791711449623108,
+      "learning_rate": 0.0002,
+      "loss": 0.7303,
+      "step": 8280
+    },
+    {
+      "epoch": 1.3402311858378466,
+      "grad_norm": 0.6836432814598083,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3418478700185918,
+      "grad_norm": 0.5579386353492737,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8300
+    },
+    {
+      "epoch": 1.343464554199337,
+      "grad_norm": 0.6713546514511108,
+      "learning_rate": 0.0002,
+      "loss": 0.7377,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3450812383800825,
+      "grad_norm": 0.5353720188140869,
+      "learning_rate": 0.0002,
+      "loss": 0.7756,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3466979225608278,
+      "grad_norm": 0.5813682675361633,
+      "learning_rate": 0.0002,
+      "loss": 0.718,
+      "step": 8330
+    },
+    {
+      "epoch": 1.348314606741573,
+      "grad_norm": 0.8158791661262512,
+      "learning_rate": 0.0002,
+      "loss": 0.7294,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3499312909223184,
+      "grad_norm": 0.6193785071372986,
+      "learning_rate": 0.0002,
+      "loss": 0.6992,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3515479751030637,
+      "grad_norm": 0.6353939771652222,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 8360
+    },
+    {
+      "epoch": 1.353164659283809,
+      "grad_norm": 0.6925048232078552,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 8370
+    },
+    {
+      "epoch": 1.3547813434645541,
+      "grad_norm": 0.988264799118042,
+      "learning_rate": 0.0002,
+      "loss": 0.736,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3563980276452994,
+      "grad_norm": 0.6476002931594849,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3580147118260448,
+      "grad_norm": 0.7120398879051208,
+      "learning_rate": 0.0002,
+      "loss": 0.776,
+      "step": 8400
+    },
+    {
+      "epoch": 1.35963139600679,
+      "grad_norm": 0.9048416614532471,
+      "learning_rate": 0.0002,
+      "loss": 0.7368,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3612480801875353,
+      "grad_norm": 0.7000672817230225,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3628647643682807,
+      "grad_norm": 0.6015632152557373,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 8430
+    },
+    {
+      "epoch": 1.364481448549026,
+      "grad_norm": 0.612516462802887,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3660981327297712,
+      "grad_norm": 0.5969301462173462,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 8450
+    },
+    {
+      "epoch": 1.3677148169105164,
+      "grad_norm": 0.6730654239654541,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 8460
+    },
+    {
+      "epoch": 1.369331501091262,
+      "grad_norm": 0.6386392116546631,
+      "learning_rate": 0.0002,
+      "loss": 0.7465,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3709481852720071,
+      "grad_norm": 0.739544153213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3725648694527524,
+      "grad_norm": 0.6462782621383667,
+      "learning_rate": 0.0002,
+      "loss": 0.7892,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3741815536334978,
+      "grad_norm": 0.7346843481063843,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 8500
+    },
+    {
+      "epoch": 1.375798237814243,
+      "grad_norm": 0.6884821057319641,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8510
+    },
+    {
+      "epoch": 1.3774149219949883,
+      "grad_norm": 0.6999333500862122,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3790316061757335,
+      "grad_norm": 0.5378713011741638,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 8530
+    },
+    {
+      "epoch": 1.3806482903564787,
+      "grad_norm": 0.5417906641960144,
+      "learning_rate": 0.0002,
+      "loss": 0.6797,
+      "step": 8540
+    },
+    {
+      "epoch": 1.3822649745372242,
+      "grad_norm": 0.6602526307106018,
+      "learning_rate": 0.0002,
+      "loss": 0.7499,
+      "step": 8550
+    },
+    {
+      "epoch": 1.3838816587179694,
+      "grad_norm": 0.7073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.7356,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3854983428987149,
+      "grad_norm": 0.5841707587242126,
+      "learning_rate": 0.0002,
+      "loss": 0.75,
+      "step": 8570
+    },
+    {
+      "epoch": 1.38711502707946,
+      "grad_norm": 0.7031095027923584,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 8580
+    },
+    {
+      "epoch": 1.3887317112602053,
+      "grad_norm": 0.5198570489883423,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 8590
+    },
+    {
+      "epoch": 1.3903483954409506,
+      "grad_norm": 0.7261320352554321,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 8600
+    },
+    {
+      "epoch": 1.3919650796216958,
+      "grad_norm": 0.5616350173950195,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3935817638024413,
+      "grad_norm": 0.5185914635658264,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 8620
+    },
+    {
+      "epoch": 1.3951984479831865,
+      "grad_norm": 0.5814694762229919,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 8630
+    },
+    {
+      "epoch": 1.3968151321639317,
+      "grad_norm": 0.6977371573448181,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 8640
+    },
+    {
+      "epoch": 1.3984318163446772,
+      "grad_norm": 0.6855689883232117,
+      "learning_rate": 0.0002,
+      "loss": 0.7574,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4000485005254224,
+      "grad_norm": 0.5414357781410217,
+      "learning_rate": 0.0002,
+      "loss": 0.7802,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4016651847061676,
+      "grad_norm": 0.6970012784004211,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4032818688869129,
+      "grad_norm": 0.526079535484314,
+      "learning_rate": 0.0002,
+      "loss": 0.7421,
+      "step": 8680
+    },
+    {
+      "epoch": 1.404898553067658,
+      "grad_norm": 0.758712887763977,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4065152372484035,
+      "grad_norm": 0.7118762731552124,
+      "learning_rate": 0.0002,
+      "loss": 0.7612,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4081319214291488,
+      "grad_norm": 0.5696909427642822,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4097486056098942,
+      "grad_norm": 0.7995436787605286,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4113652897906395,
+      "grad_norm": 0.7237521409988403,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4129819739713847,
+      "grad_norm": 0.744628369808197,
+      "learning_rate": 0.0002,
+      "loss": 0.7661,
+      "step": 8740
+    },
+    {
+      "epoch": 1.41459865815213,
+      "grad_norm": 0.6082926988601685,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4162153423328752,
+      "grad_norm": 0.5185243487358093,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4178320265136206,
+      "grad_norm": 0.5183082222938538,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4194487106943658,
+      "grad_norm": 0.7326041460037231,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 8780
+    },
+    {
+      "epoch": 1.421065394875111,
+      "grad_norm": 0.7174660563468933,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4226820790558565,
+      "grad_norm": 0.8080165982246399,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4242987632366018,
+      "grad_norm": 0.5061507821083069,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 8810
+    },
+    {
+      "epoch": 1.425915447417347,
+      "grad_norm": 0.801602840423584,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4275321315980922,
+      "grad_norm": 0.6150273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4291488157788377,
+      "grad_norm": 0.8786525726318359,
+      "learning_rate": 0.0002,
+      "loss": 0.7452,
+      "step": 8840
+    },
+    {
+      "epoch": 1.430765499959583,
+      "grad_norm": 0.6371538639068604,
+      "learning_rate": 0.0002,
+      "loss": 0.7257,
+      "step": 8850
+    },
+    {
+      "epoch": 1.4323821841403281,
+      "grad_norm": 0.6409295797348022,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4339988683210736,
+      "grad_norm": 0.6452359557151794,
+      "learning_rate": 0.0002,
+      "loss": 0.7891,
+      "step": 8870
+    },
+    {
+      "epoch": 1.4356155525018188,
+      "grad_norm": 0.5842334628105164,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 8880
+    },
+    {
+      "epoch": 1.437232236682564,
+      "grad_norm": 0.696761965751648,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4388489208633093,
+      "grad_norm": 0.6384600400924683,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4404656050440545,
+      "grad_norm": 0.5981136560440063,
+      "learning_rate": 0.0002,
+      "loss": 0.7049,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4420822892248,
+      "grad_norm": 0.6355637907981873,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4436989734055452,
+      "grad_norm": 0.6374830603599548,
+      "learning_rate": 0.0002,
+      "loss": 0.7653,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4453156575862904,
+      "grad_norm": 0.559013307094574,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 8940
+    },
+    {
+      "epoch": 1.446932341767036,
+      "grad_norm": 0.7289170026779175,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4485490259477811,
+      "grad_norm": 0.8649206757545471,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4501657101285264,
+      "grad_norm": 0.7664689421653748,
+      "learning_rate": 0.0002,
+      "loss": 0.7745,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4517823943092716,
+      "grad_norm": 0.7109952569007874,
+      "learning_rate": 0.0002,
+      "loss": 0.7431,
+      "step": 8980
+    },
+    {
+      "epoch": 1.453399078490017,
+      "grad_norm": 0.6312844753265381,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4550157626707623,
+      "grad_norm": 0.6616617441177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7467,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4566324468515075,
+      "grad_norm": 0.7384068965911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7518,
+      "step": 9010
+    },
+    {
+      "epoch": 1.458249131032253,
+      "grad_norm": 0.6549670100212097,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9020
+    },
+    {
+      "epoch": 1.4598658152129982,
+      "grad_norm": 0.6254119277000427,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4614824993937434,
+      "grad_norm": 0.6806328892707825,
+      "learning_rate": 0.0002,
+      "loss": 0.7645,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4630991835744886,
+      "grad_norm": 0.6803115010261536,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4647158677552339,
+      "grad_norm": 0.48529282212257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7264,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4663325519359793,
+      "grad_norm": 0.5995030999183655,
+      "learning_rate": 0.0002,
+      "loss": 0.7542,
+      "step": 9070
+    },
+    {
+      "epoch": 1.4679492361167246,
+      "grad_norm": 0.6005427837371826,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9080
+    },
+    {
+      "epoch": 1.46956592029747,
+      "grad_norm": 0.718564510345459,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4711826044782153,
+      "grad_norm": 0.7003577351570129,
+      "learning_rate": 0.0002,
+      "loss": 0.7089,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4727992886589605,
+      "grad_norm": 0.5888323783874512,
+      "learning_rate": 0.0002,
+      "loss": 0.8069,
+      "step": 9110
+    },
+    {
+      "epoch": 1.4744159728397057,
+      "grad_norm": 0.6417609453201294,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 9120
+    },
+    {
+      "epoch": 1.476032657020451,
+      "grad_norm": 0.572294294834137,
+      "learning_rate": 0.0002,
+      "loss": 0.7441,
+      "step": 9130
+    },
+    {
+      "epoch": 1.4776493412011964,
+      "grad_norm": 0.8200714588165283,
+      "learning_rate": 0.0002,
+      "loss": 0.8053,
+      "step": 9140
+    },
+    {
+      "epoch": 1.4792660253819416,
+      "grad_norm": 0.6343288421630859,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4808827095626869,
+      "grad_norm": 0.7017961144447327,
+      "learning_rate": 0.0002,
+      "loss": 0.7641,
+      "step": 9160
+    },
+    {
+      "epoch": 1.4824993937434323,
+      "grad_norm": 0.6202912926673889,
+      "learning_rate": 0.0002,
+      "loss": 0.7619,
+      "step": 9170
+    },
+    {
+      "epoch": 1.4841160779241775,
+      "grad_norm": 0.6677869558334351,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 9180
+    },
+    {
+      "epoch": 1.4857327621049228,
+      "grad_norm": 0.6052267551422119,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 9190
+    },
+    {
+      "epoch": 1.487349446285668,
+      "grad_norm": 0.6638872027397156,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 9200
+    },
+    {
+      "epoch": 1.4889661304664135,
+      "grad_norm": 0.6245523691177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 9210
+    },
+    {
+      "epoch": 1.4905828146471587,
+      "grad_norm": 0.5761767625808716,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 9220
+    },
+    {
+      "epoch": 1.492199498827904,
+      "grad_norm": 0.8175981640815735,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 9230
+    },
+    {
+      "epoch": 1.4938161830086494,
+      "grad_norm": 0.9144009947776794,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 9240
+    },
+    {
+      "epoch": 1.4954328671893946,
+      "grad_norm": 0.5742552876472473,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 9250
+    },
+    {
+      "epoch": 1.4970495513701398,
+      "grad_norm": 0.534534215927124,
+      "learning_rate": 0.0002,
+      "loss": 0.7418,
+      "step": 9260
+    },
+    {
+      "epoch": 1.498666235550885,
+      "grad_norm": 0.7836225032806396,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5002829197316303,
+      "grad_norm": 0.5292993187904358,
+      "learning_rate": 0.0002,
+      "loss": 0.7453,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5018996039123758,
+      "grad_norm": 0.8044071793556213,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9290
+    },
+    {
+      "epoch": 1.503516288093121,
+      "grad_norm": 0.6185805201530457,
+      "learning_rate": 0.0002,
+      "loss": 0.7229,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5051329722738664,
+      "grad_norm": 0.6093607544898987,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5067496564546117,
+      "grad_norm": 0.5891730189323425,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 9320
+    },
+    {
+      "epoch": 1.508366340635357,
+      "grad_norm": 0.6331129670143127,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5099830248161021,
+      "grad_norm": 0.7690958380699158,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5115997089968474,
+      "grad_norm": 0.6548877358436584,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5132163931775926,
+      "grad_norm": 0.6545143127441406,
+      "learning_rate": 0.0002,
+      "loss": 0.7408,
+      "step": 9360
+    },
+    {
+      "epoch": 1.514833077358338,
+      "grad_norm": 0.553247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.7432,
+      "step": 9370
+    },
+    {
+      "epoch": 1.5164497615390833,
+      "grad_norm": 0.8145074844360352,
+      "learning_rate": 0.0002,
+      "loss": 0.7265,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5180664457198287,
+      "grad_norm": 0.7636994123458862,
+      "learning_rate": 0.0002,
+      "loss": 0.7379,
+      "step": 9390
+    },
+    {
+      "epoch": 1.519683129900574,
+      "grad_norm": 0.6838982701301575,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5212998140813192,
+      "grad_norm": 0.8599441647529602,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5229164982620644,
+      "grad_norm": 0.7020329833030701,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5245331824428097,
+      "grad_norm": 0.6964772343635559,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5261498666235551,
+      "grad_norm": 0.6916600465774536,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9440
+    },
+    {
+      "epoch": 1.5277665508043003,
+      "grad_norm": 0.7282621264457703,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5293832349850458,
+      "grad_norm": 0.5363983511924744,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 9460
+    },
+    {
+      "epoch": 1.530999919165791,
+      "grad_norm": 0.6184861063957214,
+      "learning_rate": 0.0002,
+      "loss": 0.7154,
+      "step": 9470
+    },
+    {
+      "epoch": 1.5326166033465363,
+      "grad_norm": 0.5991285443305969,
+      "learning_rate": 0.0002,
+      "loss": 0.7837,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5342332875272815,
+      "grad_norm": 0.8176587820053101,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5358499717080267,
+      "grad_norm": 0.6473721861839294,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 9500
+    },
+    {
+      "epoch": 1.5374666558887722,
+      "grad_norm": 0.7319952845573425,
+      "learning_rate": 0.0002,
+      "loss": 0.7632,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5390833400695174,
+      "grad_norm": 0.702900230884552,
+      "learning_rate": 0.0002,
+      "loss": 0.7706,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5407000242502629,
+      "grad_norm": 0.7971600294113159,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 9530
+    },
+    {
+      "epoch": 1.542316708431008,
+      "grad_norm": 0.6527525186538696,
+      "learning_rate": 0.0002,
+      "loss": 0.7352,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5439333926117533,
+      "grad_norm": 0.5791676044464111,
+      "learning_rate": 0.0002,
+      "loss": 0.7425,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5455500767924986,
+      "grad_norm": 0.5619390606880188,
+      "learning_rate": 0.0002,
+      "loss": 0.7585,
+      "step": 9560
+    },
+    {
+      "epoch": 1.5471667609732438,
+      "grad_norm": 0.5701689124107361,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9570
+    },
+    {
+      "epoch": 1.548783445153989,
+      "grad_norm": 0.47549352049827576,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5504001293347345,
+      "grad_norm": 0.8730611205101013,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5520168135154797,
+      "grad_norm": 0.6842091083526611,
+      "learning_rate": 0.0002,
+      "loss": 0.798,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5536334976962252,
+      "grad_norm": 0.6675129532814026,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5552501818769704,
+      "grad_norm": 0.8173956274986267,
+      "learning_rate": 0.0002,
+      "loss": 0.7954,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5568668660577156,
+      "grad_norm": 0.724947452545166,
+      "learning_rate": 0.0002,
+      "loss": 0.7535,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5584835502384609,
+      "grad_norm": 0.6154758930206299,
+      "learning_rate": 0.0002,
+      "loss": 0.7738,
+      "step": 9640
+    },
+    {
+      "epoch": 1.560100234419206,
+      "grad_norm": 0.6072008013725281,
+      "learning_rate": 0.0002,
+      "loss": 0.7568,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5617169185999515,
+      "grad_norm": 0.659010648727417,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 9660
+    },
+    {
+      "epoch": 1.5633336027806968,
+      "grad_norm": 0.65857994556427,
+      "learning_rate": 0.0002,
+      "loss": 0.673,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5649502869614422,
+      "grad_norm": 0.5914267301559448,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5665669711421875,
+      "grad_norm": 0.6248020529747009,
+      "learning_rate": 0.0002,
+      "loss": 0.7414,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5681836553229327,
+      "grad_norm": 0.7147795557975769,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 9700
+    },
+    {
+      "epoch": 1.569800339503678,
+      "grad_norm": 0.7076232433319092,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 9710
+    },
+    {
+      "epoch": 1.5714170236844232,
+      "grad_norm": 0.6217400431632996,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5730337078651684,
+      "grad_norm": 0.6709911227226257,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 9730
+    },
+    {
+      "epoch": 1.5746503920459138,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 9740
+    },
+    {
+      "epoch": 1.576267076226659,
+      "grad_norm": 0.6241145730018616,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 9750
+    },
+    {
+      "epoch": 1.5778837604074045,
+      "grad_norm": 0.4960934817790985,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 9760
+    },
+    {
+      "epoch": 1.5795004445881498,
+      "grad_norm": 0.6593309640884399,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 9770
+    },
+    {
+      "epoch": 1.581117128768895,
+      "grad_norm": 0.5814042091369629,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 9780
+    },
+    {
+      "epoch": 1.5827338129496402,
+      "grad_norm": 0.5936070680618286,
+      "learning_rate": 0.0002,
+      "loss": 0.7109,
+      "step": 9790
+    },
+    {
+      "epoch": 1.5843504971303854,
+      "grad_norm": 0.6454403400421143,
+      "learning_rate": 0.0002,
+      "loss": 0.7769,
+      "step": 9800
+    },
+    {
+      "epoch": 1.585967181311131,
+      "grad_norm": 0.7612107992172241,
+      "learning_rate": 0.0002,
+      "loss": 0.7677,
+      "step": 9810
+    },
+    {
+      "epoch": 1.5875838654918761,
+      "grad_norm": 0.6494482755661011,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5892005496726216,
+      "grad_norm": 0.7825694680213928,
+      "learning_rate": 0.0002,
+      "loss": 0.7569,
+      "step": 9830
+    },
+    {
+      "epoch": 1.5908172338533668,
+      "grad_norm": 0.6757757663726807,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 9840
+    },
+    {
+      "epoch": 1.592433918034112,
+      "grad_norm": 0.7105609178543091,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 9850
+    },
+    {
+      "epoch": 1.5940506022148573,
+      "grad_norm": 0.7596991062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 9860
+    },
+    {
+      "epoch": 1.5956672863956025,
+      "grad_norm": 0.5681525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7108,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5972839705763477,
+      "grad_norm": 0.6090980768203735,
+      "learning_rate": 0.0002,
+      "loss": 0.7811,
+      "step": 9880
+    },
+    {
+      "epoch": 1.5989006547570932,
+      "grad_norm": 0.6271613240242004,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6005173389378387,
+      "grad_norm": 0.7656369805335999,
+      "learning_rate": 0.0002,
+      "loss": 0.7419,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6021340231185839,
+      "grad_norm": 0.7504446506500244,
+      "learning_rate": 0.0002,
+      "loss": 0.7336,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6037507072993291,
+      "grad_norm": 0.659656286239624,
+      "learning_rate": 0.0002,
+      "loss": 0.7479,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6053673914800743,
+      "grad_norm": 0.6006826162338257,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6069840756608196,
+      "grad_norm": 0.7872757911682129,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6086007598415648,
+      "grad_norm": 0.5545852780342102,
+      "learning_rate": 0.0002,
+      "loss": 0.768,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6102174440223103,
+      "grad_norm": 0.7429468631744385,
+      "learning_rate": 0.0002,
+      "loss": 0.8064,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6118341282030555,
+      "grad_norm": 0.6873556971549988,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 9970
+    },
+    {
+      "epoch": 1.613450812383801,
+      "grad_norm": 0.5874287486076355,
+      "learning_rate": 0.0002,
+      "loss": 0.7324,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6150674965645462,
+      "grad_norm": 0.6039386987686157,
+      "learning_rate": 0.0002,
+      "loss": 0.7141,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6166841807452914,
+      "grad_norm": 0.6233575940132141,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6183008649260366,
+      "grad_norm": 0.7676448225975037,
+      "learning_rate": 0.0002,
+      "loss": 0.7602,
+      "step": 10010
+    },
+    {
+      "epoch": 1.6199175491067819,
+      "grad_norm": 0.6565698385238647,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6215342332875273,
+      "grad_norm": 0.6787590384483337,
+      "learning_rate": 0.0002,
+      "loss": 0.7104,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6231509174682726,
+      "grad_norm": 0.6137678027153015,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 10040
+    },
+    {
+      "epoch": 1.624767601649018,
+      "grad_norm": 0.5236800312995911,
+      "learning_rate": 0.0002,
+      "loss": 0.7646,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6263842858297632,
+      "grad_norm": 0.7626367807388306,
+      "learning_rate": 0.0002,
+      "loss": 0.7437,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6280009700105085,
+      "grad_norm": 0.5657260417938232,
+      "learning_rate": 0.0002,
+      "loss": 0.7273,
+      "step": 10070
+    },
+    {
+      "epoch": 1.6296176541912537,
+      "grad_norm": 0.4913991391658783,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 10080
+    },
+    {
+      "epoch": 1.631234338371999,
+      "grad_norm": 0.7715556621551514,
+      "learning_rate": 0.0002,
+      "loss": 0.7596,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6328510225527442,
+      "grad_norm": 0.6509000062942505,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6344677067334896,
+      "grad_norm": 0.6215850114822388,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6360843909142349,
+      "grad_norm": 0.6956844329833984,
+      "learning_rate": 0.0002,
+      "loss": 0.7705,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6377010750949803,
+      "grad_norm": 0.6111597418785095,
+      "learning_rate": 0.0002,
+      "loss": 0.7129,
+      "step": 10130
+    },
+    {
+      "epoch": 1.6393177592757255,
+      "grad_norm": 0.6518288850784302,
+      "learning_rate": 0.0002,
+      "loss": 0.6955,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6409344434564708,
+      "grad_norm": 0.6914522051811218,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10150
+    },
+    {
+      "epoch": 1.642551127637216,
+      "grad_norm": 0.63785719871521,
+      "learning_rate": 0.0002,
+      "loss": 0.7295,
+      "step": 10160
+    },
+    {
+      "epoch": 1.6441678118179612,
+      "grad_norm": 0.6379287838935852,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6457844959987067,
+      "grad_norm": 0.6793403029441833,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 10180
+    },
+    {
+      "epoch": 1.647401180179452,
+      "grad_norm": 0.6099132895469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6490178643601974,
+      "grad_norm": 0.5869854092597961,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6506345485409426,
+      "grad_norm": 0.7716999053955078,
+      "learning_rate": 0.0002,
+      "loss": 0.8308,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6522512327216878,
+      "grad_norm": 0.6854110360145569,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 10220
+    },
+    {
+      "epoch": 1.653867916902433,
+      "grad_norm": 0.6957170367240906,
+      "learning_rate": 0.0002,
+      "loss": 0.782,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6554846010831783,
+      "grad_norm": 0.6932903528213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6571012852639235,
+      "grad_norm": 0.7713165283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10250
+    },
+    {
+      "epoch": 1.658717969444669,
+      "grad_norm": 0.7455793619155884,
+      "learning_rate": 0.0002,
+      "loss": 0.7099,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6603346536254144,
+      "grad_norm": 0.5464168190956116,
+      "learning_rate": 0.0002,
+      "loss": 0.7524,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6619513378061597,
+      "grad_norm": 0.6782926321029663,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 10280
+    },
+    {
+      "epoch": 1.663568021986905,
+      "grad_norm": 0.7962649464607239,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6651847061676501,
+      "grad_norm": 0.6814526319503784,
+      "learning_rate": 0.0002,
+      "loss": 0.7142,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6668013903483954,
+      "grad_norm": 0.656895101070404,
+      "learning_rate": 0.0002,
+      "loss": 0.7285,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6684180745291406,
+      "grad_norm": 0.6085672378540039,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 10320
+    },
+    {
+      "epoch": 1.670034758709886,
+      "grad_norm": 0.585508406162262,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 10330
+    },
+    {
+      "epoch": 1.6716514428906313,
+      "grad_norm": 0.6930184364318848,
+      "learning_rate": 0.0002,
+      "loss": 0.7604,
+      "step": 10340
+    },
+    {
+      "epoch": 1.6732681270713767,
+      "grad_norm": 0.575663149356842,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10350
+    },
+    {
+      "epoch": 1.674884811252122,
+      "grad_norm": 0.582502543926239,
+      "learning_rate": 0.0002,
+      "loss": 0.7198,
+      "step": 10360
+    },
+    {
+      "epoch": 1.6765014954328672,
+      "grad_norm": 0.5668916702270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 10370
+    },
+    {
+      "epoch": 1.6781181796136124,
+      "grad_norm": 0.6070065498352051,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10380
+    },
+    {
+      "epoch": 1.6797348637943577,
+      "grad_norm": 0.6141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.7939,
+      "step": 10390
+    },
+    {
+      "epoch": 1.6813515479751031,
+      "grad_norm": 0.8359124064445496,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 10400
+    },
+    {
+      "epoch": 1.6829682321558483,
+      "grad_norm": 0.5378185510635376,
+      "learning_rate": 0.0002,
+      "loss": 0.7488,
+      "step": 10410
+    },
+    {
+      "epoch": 1.6845849163365938,
+      "grad_norm": 0.6959536075592041,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 10420
+    },
+    {
+      "epoch": 1.686201600517339,
+      "grad_norm": 0.6514357328414917,
+      "learning_rate": 0.0002,
+      "loss": 0.7872,
+      "step": 10430
+    },
+    {
+      "epoch": 1.6878182846980843,
+      "grad_norm": 0.7706646919250488,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10440
+    },
+    {
+      "epoch": 1.6894349688788295,
+      "grad_norm": 0.6183337569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.7673,
+      "step": 10450
+    },
+    {
+      "epoch": 1.6910516530595747,
+      "grad_norm": 0.6123278141021729,
+      "learning_rate": 0.0002,
+      "loss": 0.7566,
+      "step": 10460
+    },
+    {
+      "epoch": 1.69266833724032,
+      "grad_norm": 0.6894851326942444,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10470
+    },
+    {
+      "epoch": 1.6942850214210654,
+      "grad_norm": 0.7497312426567078,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 10480
+    },
+    {
+      "epoch": 1.6959017056018106,
+      "grad_norm": 0.5968214273452759,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10490
+    },
+    {
+      "epoch": 1.697518389782556,
+      "grad_norm": 0.6747927069664001,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 10500
+    },
+    {
+      "epoch": 1.6991350739633013,
+      "grad_norm": 0.5708310008049011,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7007517581440466,
+      "grad_norm": 0.606526792049408,
+      "learning_rate": 0.0002,
+      "loss": 0.7385,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7023684423247918,
+      "grad_norm": 0.662011981010437,
+      "learning_rate": 0.0002,
+      "loss": 0.7204,
+      "step": 10530
+    },
+    {
+      "epoch": 1.703985126505537,
+      "grad_norm": 0.7583045363426208,
+      "learning_rate": 0.0002,
+      "loss": 0.7999,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7056018106862825,
+      "grad_norm": 0.721632182598114,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7072184948670277,
+      "grad_norm": 0.6107715368270874,
+      "learning_rate": 0.0002,
+      "loss": 0.7407,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7088351790477732,
+      "grad_norm": 0.6652471423149109,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7104518632285184,
+      "grad_norm": 0.6308087110519409,
+      "learning_rate": 0.0002,
+      "loss": 0.7767,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7120685474092636,
+      "grad_norm": 0.5464386940002441,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7136852315900089,
+      "grad_norm": 0.6558911204338074,
+      "learning_rate": 0.0002,
+      "loss": 0.7063,
+      "step": 10600
+    },
+    {
+      "epoch": 1.715301915770754,
+      "grad_norm": 0.5665024518966675,
+      "learning_rate": 0.0002,
+      "loss": 0.7126,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7169185999514993,
+      "grad_norm": 0.7888094186782837,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7185352841322448,
+      "grad_norm": 0.7084909081459045,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 10630
+    },
+    {
+      "epoch": 1.7201519683129902,
+      "grad_norm": 0.7982324361801147,
+      "learning_rate": 0.0002,
+      "loss": 0.7557,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7217686524937355,
+      "grad_norm": 0.6418732404708862,
+      "learning_rate": 0.0002,
+      "loss": 0.7345,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7233853366744807,
+      "grad_norm": 0.7636681795120239,
+      "learning_rate": 0.0002,
+      "loss": 0.7734,
+      "step": 10660
+    },
+    {
+      "epoch": 1.725002020855226,
+      "grad_norm": 0.5646875500679016,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7266187050359711,
+      "grad_norm": 0.5231260657310486,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7282353892167164,
+      "grad_norm": 0.7635011672973633,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7298520733974618,
+      "grad_norm": 0.7518259286880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 10700
+    },
+    {
+      "epoch": 1.731468757578207,
+      "grad_norm": 0.7295602560043335,
+      "learning_rate": 0.0002,
+      "loss": 0.751,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7330854417589525,
+      "grad_norm": 0.6984632015228271,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7347021259396977,
+      "grad_norm": 0.6198219060897827,
+      "learning_rate": 0.0002,
+      "loss": 0.7921,
+      "step": 10730
+    },
+    {
+      "epoch": 1.736318810120443,
+      "grad_norm": 0.6957576274871826,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7379354943011882,
+      "grad_norm": 0.6430263519287109,
+      "learning_rate": 0.0002,
+      "loss": 0.7917,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7395521784819334,
+      "grad_norm": 0.6134995222091675,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 10760
+    },
+    {
+      "epoch": 1.741168862662679,
+      "grad_norm": 0.7209452986717224,
+      "learning_rate": 0.0002,
+      "loss": 0.7584,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7427855468434241,
+      "grad_norm": 0.6735447645187378,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7444022310241696,
+      "grad_norm": 0.5605693459510803,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 10790
+    },
+    {
+      "epoch": 1.7460189152049148,
+      "grad_norm": 0.6882363557815552,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 10800
+    },
+    {
+      "epoch": 1.74763559938566,
+      "grad_norm": 0.6386259198188782,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7492522835664053,
+      "grad_norm": 0.6529015302658081,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7508689677471505,
+      "grad_norm": 0.5664082765579224,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7524856519278957,
+      "grad_norm": 0.7532684206962585,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7541023361086412,
+      "grad_norm": 0.77171391248703,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7557190202893864,
+      "grad_norm": 0.7255431413650513,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7573357044701319,
+      "grad_norm": 0.763083279132843,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10870
+    },
+    {
+      "epoch": 1.758952388650877,
+      "grad_norm": 0.6042402982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7605690728316223,
+      "grad_norm": 0.7642518281936646,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7621857570123676,
+      "grad_norm": 0.6347904801368713,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7638024411931128,
+      "grad_norm": 0.5371627807617188,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7654191253738583,
+      "grad_norm": 0.6840225458145142,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7670358095546035,
+      "grad_norm": 0.5288469195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10930
+    },
+    {
+      "epoch": 1.768652493735349,
+      "grad_norm": 0.69020676612854,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 10940
+    },
+    {
+      "epoch": 1.7702691779160942,
+      "grad_norm": 0.5943242311477661,
+      "learning_rate": 0.0002,
+      "loss": 0.7468,
+      "step": 10950
+    },
+    {
+      "epoch": 1.7718858620968394,
+      "grad_norm": 0.5616418123245239,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 10960
+    },
+    {
+      "epoch": 1.7735025462775846,
+      "grad_norm": 0.7209470868110657,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 10970
+    },
+    {
+      "epoch": 1.7751192304583299,
+      "grad_norm": 0.6657957434654236,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 10980
+    },
+    {
+      "epoch": 1.776735914639075,
+      "grad_norm": 0.6469064950942993,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 10990
+    },
+    {
+      "epoch": 1.7783525988198206,
+      "grad_norm": 0.6615678071975708,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 11000
+    },
+    {
+      "epoch": 1.779969283000566,
+      "grad_norm": 0.6722439527511597,
+      "learning_rate": 0.0002,
+      "loss": 0.747,
+      "step": 11010
+    },
+    {
+      "epoch": 1.7815859671813112,
+      "grad_norm": 0.634136974811554,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 11020
+    },
+    {
+      "epoch": 1.7832026513620565,
+      "grad_norm": 0.6024377346038818,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 11030
+    },
+    {
+      "epoch": 1.7848193355428017,
+      "grad_norm": 0.6909403800964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 11040
+    },
+    {
+      "epoch": 1.786436019723547,
+      "grad_norm": 0.7148767709732056,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 11050
+    },
+    {
+      "epoch": 1.7880527039042922,
+      "grad_norm": 0.7442979216575623,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 11060
+    },
+    {
+      "epoch": 1.7896693880850376,
+      "grad_norm": 0.6830431818962097,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 11070
+    },
+    {
+      "epoch": 1.7912860722657828,
+      "grad_norm": 0.9172667264938354,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 11080
+    },
+    {
+      "epoch": 1.7929027564465283,
+      "grad_norm": 0.6799490451812744,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 11090
+    },
+    {
+      "epoch": 1.7945194406272735,
+      "grad_norm": 0.7617024779319763,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 11100
+    },
+    {
+      "epoch": 1.7961361248080188,
+      "grad_norm": 0.7701810002326965,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 11110
+    },
+    {
+      "epoch": 1.797752808988764,
+      "grad_norm": 0.7454385757446289,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11120
+    },
+    {
+      "epoch": 1.7993694931695092,
+      "grad_norm": 0.6121436953544617,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8009861773502547,
+      "grad_norm": 0.6237571835517883,
+      "learning_rate": 0.0002,
+      "loss": 0.7305,
+      "step": 11140
+    },
+    {
+      "epoch": 1.802602861531,
+      "grad_norm": 0.6818515658378601,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 11150
+    },
+    {
+      "epoch": 1.8042195457117454,
+      "grad_norm": 0.7768308520317078,
+      "learning_rate": 0.0002,
+      "loss": 0.6876,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8058362298924906,
+      "grad_norm": 0.6875537633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 11170
+    },
+    {
+      "epoch": 1.8074529140732358,
+      "grad_norm": 0.7950584888458252,
+      "learning_rate": 0.0002,
+      "loss": 0.761,
+      "step": 11180
+    },
+    {
+      "epoch": 1.809069598253981,
+      "grad_norm": 0.8210248351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8106862824347263,
+      "grad_norm": 0.6674110889434814,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8123029666154715,
+      "grad_norm": 0.6261674761772156,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 11210
+    },
+    {
+      "epoch": 1.813919650796217,
+      "grad_norm": 0.6484741568565369,
+      "learning_rate": 0.0002,
+      "loss": 0.7122,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8155363349769622,
+      "grad_norm": 0.6231244206428528,
+      "learning_rate": 0.0002,
+      "loss": 0.7718,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8171530191577077,
+      "grad_norm": 0.7243146896362305,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11240
+    },
+    {
+      "epoch": 1.818769703338453,
+      "grad_norm": 0.6776193380355835,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8203863875191981,
+      "grad_norm": 0.5973618030548096,
+      "learning_rate": 0.0002,
+      "loss": 0.7317,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8220030716999434,
+      "grad_norm": 0.6451361179351807,
+      "learning_rate": 0.0002,
+      "loss": 0.7961,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8236197558806886,
+      "grad_norm": 0.5963068008422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 11280
+    },
+    {
+      "epoch": 1.825236440061434,
+      "grad_norm": 0.536902129650116,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8268531242421793,
+      "grad_norm": 0.6993787288665771,
+      "learning_rate": 0.0002,
+      "loss": 0.708,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8284698084229247,
+      "grad_norm": 0.6135255098342896,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11310
+    },
+    {
+      "epoch": 1.83008649260367,
+      "grad_norm": 0.6057423949241638,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8317031767844152,
+      "grad_norm": 0.6598812341690063,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8333198609651604,
+      "grad_norm": 0.6075948476791382,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 11340
+    },
+    {
+      "epoch": 1.8349365451459057,
+      "grad_norm": 0.7065447568893433,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8365532293266509,
+      "grad_norm": 0.680526614189148,
+      "learning_rate": 0.0002,
+      "loss": 0.7365,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8381699135073963,
+      "grad_norm": 0.6356695294380188,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8397865976881416,
+      "grad_norm": 0.6399052143096924,
+      "learning_rate": 0.0002,
+      "loss": 0.721,
+      "step": 11380
+    },
+    {
+      "epoch": 1.841403281868887,
+      "grad_norm": 0.6125704050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 11390
+    },
+    {
+      "epoch": 1.8430199660496323,
+      "grad_norm": 0.7124643325805664,
+      "learning_rate": 0.0002,
+      "loss": 0.755,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8446366502303775,
+      "grad_norm": 0.6099604964256287,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8462533344111227,
+      "grad_norm": 0.7338208556175232,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 11420
+    },
+    {
+      "epoch": 1.847870018591868,
+      "grad_norm": 0.7534668445587158,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8494867027726134,
+      "grad_norm": 0.6135470271110535,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8511033869533586,
+      "grad_norm": 0.6229309439659119,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 11450
+    },
+    {
+      "epoch": 1.852720071134104,
+      "grad_norm": 0.706423282623291,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 11460
+    },
+    {
+      "epoch": 1.8543367553148493,
+      "grad_norm": 0.5460049510002136,
+      "learning_rate": 0.0002,
+      "loss": 0.7411,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8559534394955945,
+      "grad_norm": 0.6616711020469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7416,
+      "step": 11480
+    },
+    {
+      "epoch": 1.8575701236763398,
+      "grad_norm": 0.6372783184051514,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 11490
+    },
+    {
+      "epoch": 1.859186807857085,
+      "grad_norm": 0.7162668108940125,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8608034920378305,
+      "grad_norm": 0.6605209708213806,
+      "learning_rate": 0.0002,
+      "loss": 0.7747,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8624201762185757,
+      "grad_norm": 0.6933956742286682,
+      "learning_rate": 0.0002,
+      "loss": 0.7258,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8640368603993211,
+      "grad_norm": 0.6582090854644775,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8656535445800664,
+      "grad_norm": 0.6416500806808472,
+      "learning_rate": 0.0002,
+      "loss": 0.7313,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8672702287608116,
+      "grad_norm": 0.5434312224388123,
+      "learning_rate": 0.0002,
+      "loss": 0.7372,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8688869129415568,
+      "grad_norm": 0.6827567219734192,
+      "learning_rate": 0.0002,
+      "loss": 0.7635,
+      "step": 11560
+    },
+    {
+      "epoch": 1.870503597122302,
+      "grad_norm": 0.7354370951652527,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 11570
+    },
+    {
+      "epoch": 1.8721202813030473,
+      "grad_norm": 0.590372622013092,
+      "learning_rate": 0.0002,
+      "loss": 0.7526,
+      "step": 11580
+    },
+    {
+      "epoch": 1.8737369654837928,
+      "grad_norm": 0.853183925151825,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 11590
+    },
+    {
+      "epoch": 1.875353649664538,
+      "grad_norm": 0.822678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 11600
+    },
+    {
+      "epoch": 1.8769703338452834,
+      "grad_norm": 0.6591550707817078,
+      "learning_rate": 0.0002,
+      "loss": 0.7427,
+      "step": 11610
+    },
+    {
+      "epoch": 1.8785870180260287,
+      "grad_norm": 0.7475301623344421,
+      "learning_rate": 0.0002,
+      "loss": 0.7054,
+      "step": 11620
+    },
+    {
+      "epoch": 1.880203702206774,
+      "grad_norm": 0.6390765309333801,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 11630
+    },
+    {
+      "epoch": 1.8818203863875191,
+      "grad_norm": 0.6589758992195129,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 11640
+    },
+    {
+      "epoch": 1.8834370705682644,
+      "grad_norm": 0.6765508651733398,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 11650
+    },
+    {
+      "epoch": 1.8850537547490098,
+      "grad_norm": 0.6527857780456543,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 11660
+    },
+    {
+      "epoch": 1.886670438929755,
+      "grad_norm": 0.6642923951148987,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 11670
+    },
+    {
+      "epoch": 1.8882871231105005,
+      "grad_norm": 0.6945584416389465,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 11680
+    },
+    {
+      "epoch": 1.8899038072912457,
+      "grad_norm": 0.694018542766571,
+      "learning_rate": 0.0002,
+      "loss": 0.7711,
+      "step": 11690
+    },
+    {
+      "epoch": 1.891520491471991,
+      "grad_norm": 0.7237417101860046,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 11700
+    },
+    {
+      "epoch": 1.8931371756527362,
+      "grad_norm": 0.7401309609413147,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 11710
+    },
+    {
+      "epoch": 1.8947538598334814,
+      "grad_norm": 0.6537784337997437,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 11720
+    },
+    {
+      "epoch": 1.8963705440142267,
+      "grad_norm": 0.7398539185523987,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 11730
+    },
+    {
+      "epoch": 1.8979872281949721,
+      "grad_norm": 0.6696075797080994,
+      "learning_rate": 0.0002,
+      "loss": 0.7561,
+      "step": 11740
+    },
+    {
+      "epoch": 1.8996039123757174,
+      "grad_norm": 0.6014142036437988,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9012205965564628,
+      "grad_norm": 0.7023524641990662,
+      "learning_rate": 0.0002,
+      "loss": 0.7714,
+      "step": 11760
+    },
+    {
+      "epoch": 1.902837280737208,
+      "grad_norm": 0.739973783493042,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9044539649179533,
+      "grad_norm": 0.5576770901679993,
+      "learning_rate": 0.0002,
+      "loss": 0.7848,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9060706490986985,
+      "grad_norm": 0.6907393932342529,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 11790
+    },
+    {
+      "epoch": 1.9076873332794437,
+      "grad_norm": 0.6934581995010376,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9093040174601892,
+      "grad_norm": 0.591774582862854,
+      "learning_rate": 0.0002,
+      "loss": 0.7199,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9109207016409344,
+      "grad_norm": 0.6249791383743286,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11820
+    },
+    {
+      "epoch": 1.9125373858216799,
+      "grad_norm": 0.6755744218826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 11830
+    },
+    {
+      "epoch": 1.914154070002425,
+      "grad_norm": 0.7286285161972046,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9157707541831703,
+      "grad_norm": 0.7867850065231323,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9173874383639156,
+      "grad_norm": 0.6283972859382629,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11860
+    },
+    {
+      "epoch": 1.9190041225446608,
+      "grad_norm": 0.605823814868927,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 11870
+    },
+    {
+      "epoch": 1.920620806725406,
+      "grad_norm": 0.5927976965904236,
+      "learning_rate": 0.0002,
+      "loss": 0.6598,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9222374909061515,
+      "grad_norm": 0.5974002480506897,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 11890
+    },
+    {
+      "epoch": 1.923854175086897,
+      "grad_norm": 0.7091866135597229,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9254708592676422,
+      "grad_norm": 0.72496497631073,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9270875434483874,
+      "grad_norm": 0.6131896376609802,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9287042276291326,
+      "grad_norm": 0.6556436419487,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9303209118098779,
+      "grad_norm": 0.622932493686676,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 11940
+    },
+    {
+      "epoch": 1.931937595990623,
+      "grad_norm": 0.6618631482124329,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9335542801713685,
+      "grad_norm": 0.630966305732727,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 11960
+    },
+    {
+      "epoch": 1.9351709643521138,
+      "grad_norm": 0.6336734890937805,
+      "learning_rate": 0.0002,
+      "loss": 0.6854,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9367876485328592,
+      "grad_norm": 0.655403196811676,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9384043327136045,
+      "grad_norm": 0.5640574097633362,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 11990
+    },
+    {
+      "epoch": 1.9400210168943497,
+      "grad_norm": 0.6322951316833496,
+      "learning_rate": 0.0002,
+      "loss": 0.7289,
+      "step": 12000
+    },
+    {
+      "epoch": 1.941637701075095,
+      "grad_norm": 0.615703821182251,
+      "learning_rate": 0.0002,
+      "loss": 0.7627,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9432543852558402,
+      "grad_norm": 0.6487536430358887,
+      "learning_rate": 0.0002,
+      "loss": 0.786,
+      "step": 12020
+    },
+    {
+      "epoch": 1.9448710694365856,
+      "grad_norm": 0.9209630489349365,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 12030
+    },
+    {
+      "epoch": 1.9464877536173308,
+      "grad_norm": 0.67485511302948,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9481044377980763,
+      "grad_norm": 0.6831230521202087,
+      "learning_rate": 0.0002,
+      "loss": 0.7551,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9497211219788215,
+      "grad_norm": 0.6578302383422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7546,
+      "step": 12060
+    },
+    {
+      "epoch": 1.9513378061595668,
+      "grad_norm": 0.9975938200950623,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12070
+    },
+    {
+      "epoch": 1.952954490340312,
+      "grad_norm": 0.6637365221977234,
+      "learning_rate": 0.0002,
+      "loss": 0.7952,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9545711745210572,
+      "grad_norm": 0.605707049369812,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9561878587018025,
+      "grad_norm": 0.6584440469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7768,
+      "step": 12100
+    },
+    {
+      "epoch": 1.957804542882548,
+      "grad_norm": 0.6070835590362549,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9594212270632931,
+      "grad_norm": 0.7862601280212402,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9610379112440386,
+      "grad_norm": 0.8175255060195923,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 12130
+    },
+    {
+      "epoch": 1.9626545954247838,
+      "grad_norm": 0.5648472905158997,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 12140
+    },
+    {
+      "epoch": 1.964271279605529,
+      "grad_norm": 0.6591973304748535,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9658879637862743,
+      "grad_norm": 0.5960676074028015,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9675046479670195,
+      "grad_norm": 0.7272544503211975,
+      "learning_rate": 0.0002,
+      "loss": 0.7254,
+      "step": 12170
+    },
+    {
+      "epoch": 1.969121332147765,
+      "grad_norm": 0.7176699042320251,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9707380163285102,
+      "grad_norm": 0.6927123665809631,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9723547005092557,
+      "grad_norm": 0.5536034107208252,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 12200
+    },
+    {
+      "epoch": 1.9739713846900009,
+      "grad_norm": 0.8348390460014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 12210
+    },
+    {
+      "epoch": 1.9755880688707461,
+      "grad_norm": 0.6591181755065918,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 12220
+    },
+    {
+      "epoch": 1.9772047530514913,
+      "grad_norm": 1.0624109506607056,
+      "learning_rate": 0.0002,
+      "loss": 0.763,
+      "step": 12230
+    },
+    {
+      "epoch": 1.9788214372322366,
+      "grad_norm": 0.9265586137771606,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 12240
+    },
+    {
+      "epoch": 1.9804381214129818,
+      "grad_norm": 0.5998196005821228,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 12250
+    },
+    {
+      "epoch": 1.9820548055937273,
+      "grad_norm": 0.6960851550102234,
+      "learning_rate": 0.0002,
+      "loss": 0.7225,
+      "step": 12260
+    },
+    {
+      "epoch": 1.9836714897744727,
+      "grad_norm": 0.7674502730369568,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 12270
+    },
+    {
+      "epoch": 1.985288173955218,
+      "grad_norm": 0.6407275795936584,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 12280
+    },
+    {
+      "epoch": 1.9869048581359632,
+      "grad_norm": 0.6673079133033752,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 12290
+    },
+    {
+      "epoch": 1.9885215423167084,
+      "grad_norm": 0.6989844441413879,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 12300
+    },
+    {
+      "epoch": 1.9901382264974536,
+      "grad_norm": 0.7564442157745361,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 12310
+    },
+    {
+      "epoch": 1.9917549106781989,
+      "grad_norm": 0.6385478973388672,
+      "learning_rate": 0.0002,
+      "loss": 0.7719,
+      "step": 12320
+    },
+    {
+      "epoch": 1.9933715948589443,
+      "grad_norm": 0.7193717956542969,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 12330
+    },
+    {
+      "epoch": 1.9949882790396896,
+      "grad_norm": 0.7987112402915955,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 12340
+    },
+    {
+      "epoch": 1.996604963220435,
+      "grad_norm": 0.7260826826095581,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 12350
+    },
+    {
+      "epoch": 1.9982216474011802,
+      "grad_norm": 0.7968255281448364,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 12360
+    },
+    {
+      "epoch": 1.9998383315819255,
+      "grad_norm": 0.6893062591552734,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1044032573699951,
+      "eval_runtime": 122.1508,
+      "eval_samples_per_second": 6.001,
+      "eval_steps_per_second": 0.753,
+      "step": 12371
+    },
+    {
+      "epoch": 2.0014550157626707,
+      "grad_norm": 0.7775409817695618,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 12380
+    },
+    {
+      "epoch": 2.003071699943416,
+      "grad_norm": 0.76218581199646,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 12390
+    },
+    {
+      "epoch": 2.004688384124161,
+      "grad_norm": 0.5677764415740967,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 12400
+    },
+    {
+      "epoch": 2.006305068304907,
+      "grad_norm": 0.808442234992981,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 12410
+    },
+    {
+      "epoch": 2.007921752485652,
+      "grad_norm": 0.7144765257835388,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 12420
+    },
+    {
+      "epoch": 2.0095384366663973,
+      "grad_norm": 0.6914031505584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 12430
+    },
+    {
+      "epoch": 2.0111551208471425,
+      "grad_norm": 0.7581454515457153,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 12440
+    },
+    {
+      "epoch": 2.0127718050278878,
+      "grad_norm": 0.8388504981994629,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 12450
+    },
+    {
+      "epoch": 2.014388489208633,
+      "grad_norm": 0.6716406941413879,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 12460
+    },
+    {
+      "epoch": 2.0160051733893782,
+      "grad_norm": 0.898902416229248,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 12470
+    },
+    {
+      "epoch": 2.0176218575701235,
+      "grad_norm": 0.6432679891586304,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 12480
+    },
+    {
+      "epoch": 2.019238541750869,
+      "grad_norm": 0.8021109104156494,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12490
+    },
+    {
+      "epoch": 2.0208552259316144,
+      "grad_norm": 0.7039216756820679,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 12500
+    },
+    {
+      "epoch": 2.0224719101123596,
+      "grad_norm": 0.646531879901886,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12510
+    },
+    {
+      "epoch": 2.024088594293105,
+      "grad_norm": 0.783704400062561,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 12520
+    },
+    {
+      "epoch": 2.02570527847385,
+      "grad_norm": 0.8805046677589417,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12530
+    },
+    {
+      "epoch": 2.0273219626545953,
+      "grad_norm": 0.7289270758628845,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0289386468353405,
+      "grad_norm": 0.71653151512146,
+      "learning_rate": 0.0002,
+      "loss": 0.6663,
+      "step": 12550
+    },
+    {
+      "epoch": 2.030555331016086,
+      "grad_norm": 0.73281329870224,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 12560
+    },
+    {
+      "epoch": 2.0321720151968314,
+      "grad_norm": 0.6657090187072754,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 12570
+    },
+    {
+      "epoch": 2.0337886993775767,
+      "grad_norm": 0.8241133093833923,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 12580
+    },
+    {
+      "epoch": 2.035405383558322,
+      "grad_norm": 0.5834135413169861,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 12590
+    },
+    {
+      "epoch": 2.037022067739067,
+      "grad_norm": 0.84502112865448,
+      "learning_rate": 0.0002,
+      "loss": 0.6188,
+      "step": 12600
+    },
+    {
+      "epoch": 2.0386387519198124,
+      "grad_norm": 0.8952481746673584,
+      "learning_rate": 0.0002,
+      "loss": 0.6349,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0402554361005576,
+      "grad_norm": 0.7801461815834045,
+      "learning_rate": 0.0002,
+      "loss": 0.6923,
+      "step": 12620
+    },
+    {
+      "epoch": 2.041872120281303,
+      "grad_norm": 0.6788367033004761,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 12630
+    },
+    {
+      "epoch": 2.0434888044620485,
+      "grad_norm": 0.7241756319999695,
+      "learning_rate": 0.0002,
+      "loss": 0.6162,
+      "step": 12640
+    },
+    {
+      "epoch": 2.0451054886427937,
+      "grad_norm": 0.6933388113975525,
+      "learning_rate": 0.0002,
+      "loss": 0.655,
+      "step": 12650
+    },
+    {
+      "epoch": 2.046722172823539,
+      "grad_norm": 0.8029746413230896,
+      "learning_rate": 0.0002,
+      "loss": 0.6431,
+      "step": 12660
+    },
+    {
+      "epoch": 2.048338857004284,
+      "grad_norm": 0.946399986743927,
+      "learning_rate": 0.0002,
+      "loss": 0.7164,
+      "step": 12670
+    },
+    {
+      "epoch": 2.0499555411850294,
+      "grad_norm": 0.7072678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.638,
+      "step": 12680
+    },
+    {
+      "epoch": 2.0515722253657747,
+      "grad_norm": 0.6810618042945862,
+      "learning_rate": 0.0002,
+      "loss": 0.6487,
+      "step": 12690
+    },
+    {
+      "epoch": 2.05318890954652,
+      "grad_norm": 0.7661160230636597,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 12700
+    },
+    {
+      "epoch": 2.0548055937272656,
+      "grad_norm": 0.6350653767585754,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 12710
+    },
+    {
+      "epoch": 2.056422277908011,
+      "grad_norm": 0.861890971660614,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 12720
+    },
+    {
+      "epoch": 2.058038962088756,
+      "grad_norm": 0.6489875912666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 12730
+    },
+    {
+      "epoch": 2.0596556462695013,
+      "grad_norm": 0.8268506526947021,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 12740
+    },
+    {
+      "epoch": 2.0612723304502465,
+      "grad_norm": 0.607679545879364,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 12750
+    },
+    {
+      "epoch": 2.0628890146309917,
+      "grad_norm": 0.6754153370857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6649,
+      "step": 12760
+    },
+    {
+      "epoch": 2.064505698811737,
+      "grad_norm": 0.7263124585151672,
+      "learning_rate": 0.0002,
+      "loss": 0.6549,
+      "step": 12770
+    },
+    {
+      "epoch": 2.0661223829924826,
+      "grad_norm": 0.6986154317855835,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 12780
+    },
+    {
+      "epoch": 2.067739067173228,
+      "grad_norm": 0.7768576741218567,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 12790
+    },
+    {
+      "epoch": 2.069355751353973,
+      "grad_norm": 0.7546762824058533,
+      "learning_rate": 0.0002,
+      "loss": 0.677,
+      "step": 12800
+    },
+    {
+      "epoch": 2.0709724355347183,
+      "grad_norm": 0.7588880062103271,
+      "learning_rate": 0.0002,
+      "loss": 0.6485,
+      "step": 12810
+    },
+    {
+      "epoch": 2.0725891197154636,
+      "grad_norm": 0.7457242608070374,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12820
+    },
+    {
+      "epoch": 2.074205803896209,
+      "grad_norm": 0.6983516812324524,
+      "learning_rate": 0.0002,
+      "loss": 0.6489,
+      "step": 12830
+    },
+    {
+      "epoch": 2.075822488076954,
+      "grad_norm": 0.7950928807258606,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 12840
+    },
+    {
+      "epoch": 2.0774391722576993,
+      "grad_norm": 0.9248087406158447,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 12850
+    },
+    {
+      "epoch": 2.079055856438445,
+      "grad_norm": 0.7229493260383606,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 12860
+    },
+    {
+      "epoch": 2.08067254061919,
+      "grad_norm": 0.5710847973823547,
+      "learning_rate": 0.0002,
+      "loss": 0.6702,
+      "step": 12870
+    },
+    {
+      "epoch": 2.0822892247999354,
+      "grad_norm": 0.9580423831939697,
+      "learning_rate": 0.0002,
+      "loss": 0.6974,
+      "step": 12880
+    },
+    {
+      "epoch": 2.0839059089806806,
+      "grad_norm": 0.7399665713310242,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 12890
+    },
+    {
+      "epoch": 2.085522593161426,
+      "grad_norm": 0.7981410622596741,
+      "learning_rate": 0.0002,
+      "loss": 0.6993,
+      "step": 12900
+    },
+    {
+      "epoch": 2.087139277342171,
+      "grad_norm": 0.870759904384613,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 12910
+    },
+    {
+      "epoch": 2.0887559615229163,
+      "grad_norm": 0.7001481652259827,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 12920
+    },
+    {
+      "epoch": 2.090372645703662,
+      "grad_norm": 0.6745418310165405,
+      "learning_rate": 0.0002,
+      "loss": 0.6383,
+      "step": 12930
+    },
+    {
+      "epoch": 2.0919893298844072,
+      "grad_norm": 0.7739067673683167,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 12940
+    },
+    {
+      "epoch": 2.0936060140651525,
+      "grad_norm": 0.6742934584617615,
+      "learning_rate": 0.0002,
+      "loss": 0.6856,
+      "step": 12950
+    },
+    {
+      "epoch": 2.0952226982458977,
+      "grad_norm": 0.7270349860191345,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 12960
+    },
+    {
+      "epoch": 2.096839382426643,
+      "grad_norm": 0.7150624394416809,
+      "learning_rate": 0.0002,
+      "loss": 0.6783,
+      "step": 12970
+    },
+    {
+      "epoch": 2.098456066607388,
+      "grad_norm": 0.7734767198562622,
+      "learning_rate": 0.0002,
+      "loss": 0.6093,
+      "step": 12980
+    },
+    {
+      "epoch": 2.1000727507881334,
+      "grad_norm": 0.7618662118911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 12990
+    },
+    {
+      "epoch": 2.101689434968879,
+      "grad_norm": 0.6557944416999817,
+      "learning_rate": 0.0002,
+      "loss": 0.6707,
+      "step": 13000
+    },
+    {
+      "epoch": 2.1033061191496243,
+      "grad_norm": 0.8786448240280151,
+      "learning_rate": 0.0002,
+      "loss": 0.7268,
+      "step": 13010
+    },
+    {
+      "epoch": 2.1049228033303695,
+      "grad_norm": 0.6878724098205566,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 13020
+    },
+    {
+      "epoch": 2.1065394875111147,
+      "grad_norm": 0.822318971157074,
+      "learning_rate": 0.0002,
+      "loss": 0.6824,
+      "step": 13030
+    },
+    {
+      "epoch": 2.10815617169186,
+      "grad_norm": 0.831468939781189,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 13040
+    },
+    {
+      "epoch": 2.109772855872605,
+      "grad_norm": 0.7699505686759949,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 13050
+    },
+    {
+      "epoch": 2.1113895400533504,
+      "grad_norm": 0.7559016346931458,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1130062242340957,
+      "grad_norm": 0.6942209601402283,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 13070
+    },
+    {
+      "epoch": 2.1146229084148414,
+      "grad_norm": 0.6098947525024414,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 13080
+    },
+    {
+      "epoch": 2.1162395925955866,
+      "grad_norm": 0.6499016284942627,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 13090
+    },
+    {
+      "epoch": 2.117856276776332,
+      "grad_norm": 0.7719953060150146,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 13100
+    },
+    {
+      "epoch": 2.119472960957077,
+      "grad_norm": 0.6708134412765503,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1210896451378223,
+      "grad_norm": 0.8119585514068604,
+      "learning_rate": 0.0002,
+      "loss": 0.6056,
+      "step": 13120
+    },
+    {
+      "epoch": 2.1227063293185675,
+      "grad_norm": 0.6947157979011536,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 13130
+    },
+    {
+      "epoch": 2.1243230134993127,
+      "grad_norm": 0.8831837773323059,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 13140
+    },
+    {
+      "epoch": 2.1259396976800584,
+      "grad_norm": 0.7266910672187805,
+      "learning_rate": 0.0002,
+      "loss": 0.6997,
+      "step": 13150
+    },
+    {
+      "epoch": 2.1275563818608036,
+      "grad_norm": 0.8864351511001587,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 13160
+    },
+    {
+      "epoch": 2.129173066041549,
+      "grad_norm": 0.8104248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.6762,
+      "step": 13170
+    },
+    {
+      "epoch": 2.130789750222294,
+      "grad_norm": 0.6077079772949219,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 13180
+    },
+    {
+      "epoch": 2.1324064344030393,
+      "grad_norm": 0.6874213814735413,
+      "learning_rate": 0.0002,
+      "loss": 0.6572,
+      "step": 13190
+    },
+    {
+      "epoch": 2.1340231185837846,
+      "grad_norm": 0.7134367823600769,
+      "learning_rate": 0.0002,
+      "loss": 0.642,
+      "step": 13200
+    },
+    {
+      "epoch": 2.13563980276453,
+      "grad_norm": 0.6101235151290894,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 13210
+    },
+    {
+      "epoch": 2.137256486945275,
+      "grad_norm": 0.6042411923408508,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13220
+    },
+    {
+      "epoch": 2.1388731711260207,
+      "grad_norm": 0.914601743221283,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 13230
+    },
+    {
+      "epoch": 2.140489855306766,
+      "grad_norm": 0.7104284167289734,
+      "learning_rate": 0.0002,
+      "loss": 0.6513,
+      "step": 13240
+    },
+    {
+      "epoch": 2.142106539487511,
+      "grad_norm": 0.664395272731781,
+      "learning_rate": 0.0002,
+      "loss": 0.6607,
+      "step": 13250
+    },
+    {
+      "epoch": 2.1437232236682564,
+      "grad_norm": 0.6991241574287415,
+      "learning_rate": 0.0002,
+      "loss": 0.7211,
+      "step": 13260
+    },
+    {
+      "epoch": 2.1453399078490016,
+      "grad_norm": 0.5469560623168945,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 13270
+    },
+    {
+      "epoch": 2.146956592029747,
+      "grad_norm": 0.8454998135566711,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13280
+    },
+    {
+      "epoch": 2.148573276210492,
+      "grad_norm": 0.7088868618011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 13290
+    },
+    {
+      "epoch": 2.1501899603912378,
+      "grad_norm": 0.7002687454223633,
+      "learning_rate": 0.0002,
+      "loss": 0.6835,
+      "step": 13300
+    },
+    {
+      "epoch": 2.151806644571983,
+      "grad_norm": 0.7785214781761169,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 13310
+    },
+    {
+      "epoch": 2.1534233287527282,
+      "grad_norm": 0.8049132227897644,
+      "learning_rate": 0.0002,
+      "loss": 0.67,
+      "step": 13320
+    },
+    {
+      "epoch": 2.1550400129334735,
+      "grad_norm": 0.8062595129013062,
+      "learning_rate": 0.0002,
+      "loss": 0.6495,
+      "step": 13330
+    },
+    {
+      "epoch": 2.1566566971142187,
+      "grad_norm": 0.6208319067955017,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 13340
+    },
+    {
+      "epoch": 2.158273381294964,
+      "grad_norm": 0.7519655823707581,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 13350
+    },
+    {
+      "epoch": 2.159890065475709,
+      "grad_norm": 0.7645747065544128,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 13360
+    },
+    {
+      "epoch": 2.1615067496564544,
+      "grad_norm": 0.6847302913665771,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 13370
+    },
+    {
+      "epoch": 2.1631234338372,
+      "grad_norm": 0.8630441427230835,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 13380
+    },
+    {
+      "epoch": 2.1647401180179453,
+      "grad_norm": 0.7947702407836914,
+      "learning_rate": 0.0002,
+      "loss": 0.6742,
+      "step": 13390
+    },
+    {
+      "epoch": 2.1663568021986905,
+      "grad_norm": 0.6836977005004883,
+      "learning_rate": 0.0002,
+      "loss": 0.7206,
+      "step": 13400
+    },
+    {
+      "epoch": 2.1679734863794358,
+      "grad_norm": 0.7340566515922546,
+      "learning_rate": 0.0002,
+      "loss": 0.6304,
+      "step": 13410
+    },
+    {
+      "epoch": 2.169590170560181,
+      "grad_norm": 0.7075738906860352,
+      "learning_rate": 0.0002,
+      "loss": 0.6528,
+      "step": 13420
+    },
+    {
+      "epoch": 2.1712068547409262,
+      "grad_norm": 0.7080879807472229,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 13430
+    },
+    {
+      "epoch": 2.1728235389216715,
+      "grad_norm": 0.6218613386154175,
+      "learning_rate": 0.0002,
+      "loss": 0.6615,
+      "step": 13440
+    },
+    {
+      "epoch": 2.174440223102417,
+      "grad_norm": 0.8211479187011719,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 13450
+    },
+    {
+      "epoch": 2.1760569072831624,
+      "grad_norm": 0.864466667175293,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 13460
+    },
+    {
+      "epoch": 2.1776735914639076,
+      "grad_norm": 0.7943857908248901,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 13470
+    },
+    {
+      "epoch": 2.179290275644653,
+      "grad_norm": 0.78728187084198,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 13480
+    },
+    {
+      "epoch": 2.180906959825398,
+      "grad_norm": 0.697527289390564,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 13490
+    },
+    {
+      "epoch": 2.1825236440061433,
+      "grad_norm": 0.8205804228782654,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 13500
+    },
+    {
+      "epoch": 2.1841403281868885,
+      "grad_norm": 0.8709042072296143,
+      "learning_rate": 0.0002,
+      "loss": 0.7227,
+      "step": 13510
+    },
+    {
+      "epoch": 2.1857570123676338,
+      "grad_norm": 0.6228537559509277,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 13520
+    },
+    {
+      "epoch": 2.1873736965483794,
+      "grad_norm": 0.9566980004310608,
+      "learning_rate": 0.0002,
+      "loss": 0.7025,
+      "step": 13530
+    },
+    {
+      "epoch": 2.1889903807291247,
+      "grad_norm": 0.7128894329071045,
+      "learning_rate": 0.0002,
+      "loss": 0.6755,
+      "step": 13540
+    },
+    {
+      "epoch": 2.19060706490987,
+      "grad_norm": 0.6888654232025146,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 13550
+    },
+    {
+      "epoch": 2.192223749090615,
+      "grad_norm": 0.6444337368011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6961,
+      "step": 13560
+    },
+    {
+      "epoch": 2.1938404332713604,
+      "grad_norm": 0.8008806705474854,
+      "learning_rate": 0.0002,
+      "loss": 0.656,
+      "step": 13570
+    },
+    {
+      "epoch": 2.1954571174521056,
+      "grad_norm": 0.8482748866081238,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 13580
+    },
+    {
+      "epoch": 2.197073801632851,
+      "grad_norm": 0.8584157228469849,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 13590
+    },
+    {
+      "epoch": 2.1986904858135965,
+      "grad_norm": 0.7513734698295593,
+      "learning_rate": 0.0002,
+      "loss": 0.7014,
+      "step": 13600
+    },
+    {
+      "epoch": 2.2003071699943417,
+      "grad_norm": 0.7864262461662292,
+      "learning_rate": 0.0002,
+      "loss": 0.6632,
+      "step": 13610
+    },
+    {
+      "epoch": 2.201923854175087,
+      "grad_norm": 0.8493645191192627,
+      "learning_rate": 0.0002,
+      "loss": 0.6879,
+      "step": 13620
+    },
+    {
+      "epoch": 2.203540538355832,
+      "grad_norm": 0.6902140974998474,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 13630
+    },
+    {
+      "epoch": 2.2051572225365774,
+      "grad_norm": 0.8711254596710205,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 13640
+    },
+    {
+      "epoch": 2.2067739067173227,
+      "grad_norm": 0.7832191586494446,
+      "learning_rate": 0.0002,
+      "loss": 0.6359,
+      "step": 13650
+    },
+    {
+      "epoch": 2.208390590898068,
+      "grad_norm": 0.5668176412582397,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2100072750788136,
+      "grad_norm": 0.8648375272750854,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13670
+    },
+    {
+      "epoch": 2.211623959259559,
+      "grad_norm": 0.7643089890480042,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13680
+    },
+    {
+      "epoch": 2.213240643440304,
+      "grad_norm": 0.6293777823448181,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13690
+    },
+    {
+      "epoch": 2.2148573276210493,
+      "grad_norm": 0.6459372639656067,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 13700
+    },
+    {
+      "epoch": 2.2164740118017945,
+      "grad_norm": 0.7060744166374207,
+      "learning_rate": 0.0002,
+      "loss": 0.6526,
+      "step": 13710
+    },
+    {
+      "epoch": 2.2180906959825397,
+      "grad_norm": 0.674109160900116,
+      "learning_rate": 0.0002,
+      "loss": 0.7101,
+      "step": 13720
+    },
+    {
+      "epoch": 2.219707380163285,
+      "grad_norm": 0.830392062664032,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13730
+    },
+    {
+      "epoch": 2.2213240643440306,
+      "grad_norm": 0.6474477052688599,
+      "learning_rate": 0.0002,
+      "loss": 0.6733,
+      "step": 13740
+    },
+    {
+      "epoch": 2.222940748524776,
+      "grad_norm": 0.7037909626960754,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13750
+    },
+    {
+      "epoch": 2.224557432705521,
+      "grad_norm": 0.6554131507873535,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 13760
+    },
+    {
+      "epoch": 2.2261741168862663,
+      "grad_norm": 0.7822230458259583,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 13770
+    },
+    {
+      "epoch": 2.2277908010670116,
+      "grad_norm": 0.9082167744636536,
+      "learning_rate": 0.0002,
+      "loss": 0.6505,
+      "step": 13780
+    },
+    {
+      "epoch": 2.229407485247757,
+      "grad_norm": 0.7918276190757751,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 13790
+    },
+    {
+      "epoch": 2.231024169428502,
+      "grad_norm": 0.7354569435119629,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 13800
+    },
+    {
+      "epoch": 2.2326408536092472,
+      "grad_norm": 0.8265249133110046,
+      "learning_rate": 0.0002,
+      "loss": 0.6503,
+      "step": 13810
+    },
+    {
+      "epoch": 2.234257537789993,
+      "grad_norm": 0.6653847098350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6871,
+      "step": 13820
+    },
+    {
+      "epoch": 2.235874221970738,
+      "grad_norm": 0.7157923579216003,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13830
+    },
+    {
+      "epoch": 2.2374909061514834,
+      "grad_norm": 0.7110323309898376,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 13840
+    },
+    {
+      "epoch": 2.2391075903322286,
+      "grad_norm": 0.7155357599258423,
+      "learning_rate": 0.0002,
+      "loss": 0.6913,
+      "step": 13850
+    },
+    {
+      "epoch": 2.240724274512974,
+      "grad_norm": 1.0177817344665527,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 13860
+    },
+    {
+      "epoch": 2.242340958693719,
+      "grad_norm": 0.7601948380470276,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13870
+    },
+    {
+      "epoch": 2.2439576428744643,
+      "grad_norm": 0.7628820538520813,
+      "learning_rate": 0.0002,
+      "loss": 0.6679,
+      "step": 13880
+    },
+    {
+      "epoch": 2.24557432705521,
+      "grad_norm": 0.7089297771453857,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 13890
+    },
+    {
+      "epoch": 2.247191011235955,
+      "grad_norm": 0.695178210735321,
+      "learning_rate": 0.0002,
+      "loss": 0.7236,
+      "step": 13900
+    },
+    {
+      "epoch": 2.2488076954167004,
+      "grad_norm": 0.7631948590278625,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 13910
+    },
+    {
+      "epoch": 2.2504243795974457,
+      "grad_norm": 0.8203101754188538,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 13920
+    },
+    {
+      "epoch": 2.252041063778191,
+      "grad_norm": 0.8099079728126526,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13930
+    },
+    {
+      "epoch": 2.253657747958936,
+      "grad_norm": 0.6498546004295349,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 13940
+    },
+    {
+      "epoch": 2.2552744321396814,
+      "grad_norm": 0.7797415256500244,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 13950
+    },
+    {
+      "epoch": 2.2568911163204266,
+      "grad_norm": 0.8254124522209167,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 13960
+    },
+    {
+      "epoch": 2.2585078005011723,
+      "grad_norm": 0.6327953338623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 13970
+    },
+    {
+      "epoch": 2.2601244846819175,
+      "grad_norm": 0.734194278717041,
+      "learning_rate": 0.0002,
+      "loss": 0.668,
+      "step": 13980
+    },
+    {
+      "epoch": 2.2617411688626627,
+      "grad_norm": 0.9014202952384949,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 13990
+    },
+    {
+      "epoch": 2.263357853043408,
+      "grad_norm": 0.7643631100654602,
+      "learning_rate": 0.0002,
+      "loss": 0.692,
+      "step": 14000
+    },
+    {
+      "epoch": 2.264974537224153,
+      "grad_norm": 0.8882834911346436,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 14010
+    },
+    {
+      "epoch": 2.2665912214048984,
+      "grad_norm": 0.7975873351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14020
+    },
+    {
+      "epoch": 2.2682079055856437,
+      "grad_norm": 0.7765783071517944,
+      "learning_rate": 0.0002,
+      "loss": 0.7193,
+      "step": 14030
+    },
+    {
+      "epoch": 2.2698245897663893,
+      "grad_norm": 0.8846288323402405,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 14040
+    },
+    {
+      "epoch": 2.2714412739471346,
+      "grad_norm": 0.9006744027137756,
+      "learning_rate": 0.0002,
+      "loss": 0.6494,
+      "step": 14050
+    },
+    {
+      "epoch": 2.27305795812788,
+      "grad_norm": 0.7420173287391663,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 14060
+    },
+    {
+      "epoch": 2.274674642308625,
+      "grad_norm": 0.7956424951553345,
+      "learning_rate": 0.0002,
+      "loss": 0.7068,
+      "step": 14070
+    },
+    {
+      "epoch": 2.2762913264893703,
+      "grad_norm": 0.7783209085464478,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 14080
+    },
+    {
+      "epoch": 2.2779080106701155,
+      "grad_norm": 0.7597188949584961,
+      "learning_rate": 0.0002,
+      "loss": 0.7202,
+      "step": 14090
+    },
+    {
+      "epoch": 2.2795246948508607,
+      "grad_norm": 0.6718921661376953,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14100
+    },
+    {
+      "epoch": 2.281141379031606,
+      "grad_norm": 0.7528082132339478,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 14110
+    },
+    {
+      "epoch": 2.2827580632123516,
+      "grad_norm": 0.8379864692687988,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 14120
+    },
+    {
+      "epoch": 2.284374747393097,
+      "grad_norm": 0.748613715171814,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 14130
+    },
+    {
+      "epoch": 2.285991431573842,
+      "grad_norm": 0.7435423135757446,
+      "learning_rate": 0.0002,
+      "loss": 0.6641,
+      "step": 14140
+    },
+    {
+      "epoch": 2.2876081157545873,
+      "grad_norm": 0.7580803632736206,
+      "learning_rate": 0.0002,
+      "loss": 0.6849,
+      "step": 14150
+    },
+    {
+      "epoch": 2.2892247999353326,
+      "grad_norm": 0.6278321146965027,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 14160
+    },
+    {
+      "epoch": 2.290841484116078,
+      "grad_norm": 0.7663896083831787,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 14170
+    },
+    {
+      "epoch": 2.292458168296823,
+      "grad_norm": 0.9716812372207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 14180
+    },
+    {
+      "epoch": 2.2940748524775687,
+      "grad_norm": 0.8993458151817322,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14190
+    },
+    {
+      "epoch": 2.295691536658314,
+      "grad_norm": 0.6156117916107178,
+      "learning_rate": 0.0002,
+      "loss": 0.6172,
+      "step": 14200
+    },
+    {
+      "epoch": 2.297308220839059,
+      "grad_norm": 0.8911278247833252,
+      "learning_rate": 0.0002,
+      "loss": 0.6318,
+      "step": 14210
+    },
+    {
+      "epoch": 2.2989249050198044,
+      "grad_norm": 0.6422147154808044,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 14220
+    },
+    {
+      "epoch": 2.3005415892005496,
+      "grad_norm": 0.6866879463195801,
+      "learning_rate": 0.0002,
+      "loss": 0.6795,
+      "step": 14230
+    },
+    {
+      "epoch": 2.302158273381295,
+      "grad_norm": 0.9297130107879639,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 14240
+    },
+    {
+      "epoch": 2.30377495756204,
+      "grad_norm": 0.7501356601715088,
+      "learning_rate": 0.0002,
+      "loss": 0.6823,
+      "step": 14250
+    },
+    {
+      "epoch": 2.3053916417427853,
+      "grad_norm": 0.8363515138626099,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 14260
+    },
+    {
+      "epoch": 2.307008325923531,
+      "grad_norm": 0.9083868265151978,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3086250101042762,
+      "grad_norm": 0.7791516780853271,
+      "learning_rate": 0.0002,
+      "loss": 0.6862,
+      "step": 14280
+    },
+    {
+      "epoch": 2.3102416942850215,
+      "grad_norm": 0.8766953349113464,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14290
+    },
+    {
+      "epoch": 2.3118583784657667,
+      "grad_norm": 0.7916635274887085,
+      "learning_rate": 0.0002,
+      "loss": 0.6698,
+      "step": 14300
+    },
+    {
+      "epoch": 2.313475062646512,
+      "grad_norm": 0.627525269985199,
+      "learning_rate": 0.0002,
+      "loss": 0.6927,
+      "step": 14310
+    },
+    {
+      "epoch": 2.315091746827257,
+      "grad_norm": 0.8856783509254456,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 14320
+    },
+    {
+      "epoch": 2.316708431008003,
+      "grad_norm": 0.6758689284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 14330
+    },
+    {
+      "epoch": 2.318325115188748,
+      "grad_norm": 0.6428321003913879,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 14340
+    },
+    {
+      "epoch": 2.3199417993694933,
+      "grad_norm": 0.9032121300697327,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 14350
+    },
+    {
+      "epoch": 2.3215584835502385,
+      "grad_norm": 0.8035986423492432,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14360
+    },
+    {
+      "epoch": 2.3231751677309838,
+      "grad_norm": 0.7974579334259033,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14370
+    },
+    {
+      "epoch": 2.324791851911729,
+      "grad_norm": 0.8356034755706787,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 14380
+    },
+    {
+      "epoch": 2.326408536092474,
+      "grad_norm": 0.998760998249054,
+      "learning_rate": 0.0002,
+      "loss": 0.6623,
+      "step": 14390
+    },
+    {
+      "epoch": 2.3280252202732195,
+      "grad_norm": 0.6518142223358154,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 14400
+    },
+    {
+      "epoch": 2.3296419044539647,
+      "grad_norm": 0.7443506717681885,
+      "learning_rate": 0.0002,
+      "loss": 0.7146,
+      "step": 14410
+    },
+    {
+      "epoch": 2.3312585886347104,
+      "grad_norm": 0.8436172604560852,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14420
+    },
+    {
+      "epoch": 2.3328752728154556,
+      "grad_norm": 0.7411080598831177,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 14430
+    },
+    {
+      "epoch": 2.334491956996201,
+      "grad_norm": 0.8839048743247986,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 14440
+    },
+    {
+      "epoch": 2.336108641176946,
+      "grad_norm": 0.8360885977745056,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 14450
+    },
+    {
+      "epoch": 2.3377253253576913,
+      "grad_norm": 0.7608986496925354,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 14460
+    },
+    {
+      "epoch": 2.3393420095384365,
+      "grad_norm": 0.8179867267608643,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14470
+    },
+    {
+      "epoch": 2.340958693719182,
+      "grad_norm": 0.5989999771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14480
+    },
+    {
+      "epoch": 2.3425753778999274,
+      "grad_norm": 0.9450054168701172,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 14490
+    },
+    {
+      "epoch": 2.3441920620806727,
+      "grad_norm": 0.7885149717330933,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 14500
+    },
+    {
+      "epoch": 2.345808746261418,
+      "grad_norm": 0.8152616620063782,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 14510
+    },
+    {
+      "epoch": 2.347425430442163,
+      "grad_norm": 0.7193838953971863,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 14520
+    },
+    {
+      "epoch": 2.3490421146229084,
+      "grad_norm": 0.6701092720031738,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 14530
+    },
+    {
+      "epoch": 2.3506587988036536,
+      "grad_norm": 0.7529364228248596,
+      "learning_rate": 0.0002,
+      "loss": 0.6559,
+      "step": 14540
+    },
+    {
+      "epoch": 2.352275482984399,
+      "grad_norm": 0.6599733829498291,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 14550
+    },
+    {
+      "epoch": 2.353892167165144,
+      "grad_norm": 0.9502474069595337,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14560
+    },
+    {
+      "epoch": 2.3555088513458897,
+      "grad_norm": 0.7619650959968567,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 14570
+    },
+    {
+      "epoch": 2.357125535526635,
+      "grad_norm": 0.9854652285575867,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14580
+    },
+    {
+      "epoch": 2.35874221970738,
+      "grad_norm": 0.727439284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 14590
+    },
+    {
+      "epoch": 2.3603589038881254,
+      "grad_norm": 0.6994746327400208,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 14600
+    },
+    {
+      "epoch": 2.3619755880688706,
+      "grad_norm": 0.7117531299591064,
+      "learning_rate": 0.0002,
+      "loss": 0.6587,
+      "step": 14610
+    },
+    {
+      "epoch": 2.363592272249616,
+      "grad_norm": 0.6403067708015442,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 14620
+    },
+    {
+      "epoch": 2.3652089564303616,
+      "grad_norm": 0.8377841711044312,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14630
+    },
+    {
+      "epoch": 2.366825640611107,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14640
+    },
+    {
+      "epoch": 2.368442324791852,
+      "grad_norm": 0.8418586254119873,
+      "learning_rate": 0.0002,
+      "loss": 0.6552,
+      "step": 14650
+    },
+    {
+      "epoch": 2.3700590089725972,
+      "grad_norm": 0.6178573369979858,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14660
+    },
+    {
+      "epoch": 2.3716756931533425,
+      "grad_norm": 0.6368302702903748,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 14670
+    },
+    {
+      "epoch": 2.3732923773340877,
+      "grad_norm": 0.9122977256774902,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14680
+    },
+    {
+      "epoch": 2.374909061514833,
+      "grad_norm": 0.7086195349693298,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 14690
+    },
+    {
+      "epoch": 2.376525745695578,
+      "grad_norm": 0.7500800490379333,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 14700
+    },
+    {
+      "epoch": 2.378142429876324,
+      "grad_norm": 0.6634900569915771,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 14710
+    },
+    {
+      "epoch": 2.379759114057069,
+      "grad_norm": 0.839898407459259,
+      "learning_rate": 0.0002,
+      "loss": 0.6614,
+      "step": 14720
+    },
+    {
+      "epoch": 2.3813757982378143,
+      "grad_norm": 0.7578426003456116,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14730
+    },
+    {
+      "epoch": 2.3829924824185595,
+      "grad_norm": 1.0213173627853394,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3846091665993048,
+      "grad_norm": 0.7855949401855469,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 14750
+    },
+    {
+      "epoch": 2.38622585078005,
+      "grad_norm": 0.7224128842353821,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 14760
+    },
+    {
+      "epoch": 2.3878425349607952,
+      "grad_norm": 0.8040381669998169,
+      "learning_rate": 0.0002,
+      "loss": 0.7017,
+      "step": 14770
+    },
+    {
+      "epoch": 2.389459219141541,
+      "grad_norm": 0.7705281376838684,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 14780
+    },
+    {
+      "epoch": 2.391075903322286,
+      "grad_norm": 0.667966902256012,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 14790
+    },
+    {
+      "epoch": 2.3926925875030314,
+      "grad_norm": 0.6611011028289795,
+      "learning_rate": 0.0002,
+      "loss": 0.7061,
+      "step": 14800
+    },
+    {
+      "epoch": 2.3943092716837766,
+      "grad_norm": 0.6862651705741882,
+      "learning_rate": 0.0002,
+      "loss": 0.6527,
+      "step": 14810
+    },
+    {
+      "epoch": 2.395925955864522,
+      "grad_norm": 0.8086010217666626,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 14820
+    },
+    {
+      "epoch": 2.397542640045267,
+      "grad_norm": 0.7189689874649048,
+      "learning_rate": 0.0002,
+      "loss": 0.7189,
+      "step": 14830
+    },
+    {
+      "epoch": 2.3991593242260123,
+      "grad_norm": 0.6280009150505066,
+      "learning_rate": 0.0002,
+      "loss": 0.6709,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4007760084067575,
+      "grad_norm": 0.7826612591743469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14850
+    },
+    {
+      "epoch": 2.402392692587503,
+      "grad_norm": 0.7681610584259033,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4040093767682484,
+      "grad_norm": 0.720966100692749,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 14870
+    },
+    {
+      "epoch": 2.4056260609489937,
+      "grad_norm": 0.8202250599861145,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 14880
+    },
+    {
+      "epoch": 2.407242745129739,
+      "grad_norm": 0.786212682723999,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 14890
+    },
+    {
+      "epoch": 2.408859429310484,
+      "grad_norm": 0.6647164821624756,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 14900
+    },
+    {
+      "epoch": 2.4104761134912294,
+      "grad_norm": 0.7566399574279785,
+      "learning_rate": 0.0002,
+      "loss": 0.6706,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4120927976719746,
+      "grad_norm": 0.748814582824707,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 14920
+    },
+    {
+      "epoch": 2.4137094818527203,
+      "grad_norm": 0.7624038457870483,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14930
+    },
+    {
+      "epoch": 2.4153261660334655,
+      "grad_norm": 0.8267335295677185,
+      "learning_rate": 0.0002,
+      "loss": 0.6483,
+      "step": 14940
+    },
+    {
+      "epoch": 2.4169428502142107,
+      "grad_norm": 0.8785360455513,
+      "learning_rate": 0.0002,
+      "loss": 0.6612,
+      "step": 14950
+    },
+    {
+      "epoch": 2.418559534394956,
+      "grad_norm": 0.679887592792511,
+      "learning_rate": 0.0002,
+      "loss": 0.6718,
+      "step": 14960
+    },
+    {
+      "epoch": 2.420176218575701,
+      "grad_norm": 0.7218474745750427,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4217929027564464,
+      "grad_norm": 0.6342799663543701,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14980
+    },
+    {
+      "epoch": 2.4234095869371917,
+      "grad_norm": 0.7098712921142578,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 14990
+    },
+    {
+      "epoch": 2.425026271117937,
+      "grad_norm": 0.7497431635856628,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4266429552986826,
+      "grad_norm": 0.934836208820343,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15010
+    },
+    {
+      "epoch": 2.428259639479428,
+      "grad_norm": 0.8430966734886169,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 15020
+    },
+    {
+      "epoch": 2.429876323660173,
+      "grad_norm": 0.7032104730606079,
+      "learning_rate": 0.0002,
+      "loss": 0.6985,
+      "step": 15030
+    },
+    {
+      "epoch": 2.4314930078409183,
+      "grad_norm": 0.7746111750602722,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 15040
+    },
+    {
+      "epoch": 2.4331096920216635,
+      "grad_norm": 0.7661406397819519,
+      "learning_rate": 0.0002,
+      "loss": 0.7177,
+      "step": 15050
+    },
+    {
+      "epoch": 2.4347263762024087,
+      "grad_norm": 0.6941645741462708,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 15060
+    },
+    {
+      "epoch": 2.436343060383154,
+      "grad_norm": 0.7487249374389648,
+      "learning_rate": 0.0002,
+      "loss": 0.6421,
+      "step": 15070
+    },
+    {
+      "epoch": 2.4379597445638996,
+      "grad_norm": 0.7639912962913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 15080
+    },
+    {
+      "epoch": 2.439576428744645,
+      "grad_norm": 0.7708953619003296,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 15090
+    },
+    {
+      "epoch": 2.44119311292539,
+      "grad_norm": 0.9135832190513611,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15100
+    },
+    {
+      "epoch": 2.4428097971061353,
+      "grad_norm": 0.8283005356788635,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 15110
+    },
+    {
+      "epoch": 2.4444264812868806,
+      "grad_norm": 0.925299346446991,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 15120
+    },
+    {
+      "epoch": 2.446043165467626,
+      "grad_norm": 0.7013528943061829,
+      "learning_rate": 0.0002,
+      "loss": 0.687,
+      "step": 15130
+    },
+    {
+      "epoch": 2.447659849648371,
+      "grad_norm": 0.622303307056427,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 15140
+    },
+    {
+      "epoch": 2.4492765338291163,
+      "grad_norm": 0.876569390296936,
+      "learning_rate": 0.0002,
+      "loss": 0.6676,
+      "step": 15150
+    },
+    {
+      "epoch": 2.450893218009862,
+      "grad_norm": 0.6836351752281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 15160
+    },
+    {
+      "epoch": 2.452509902190607,
+      "grad_norm": 0.7886684536933899,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 15170
+    },
+    {
+      "epoch": 2.4541265863713524,
+      "grad_norm": 0.6647440791130066,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 15180
+    },
+    {
+      "epoch": 2.4557432705520976,
+      "grad_norm": 0.7477722764015198,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 15190
+    },
+    {
+      "epoch": 2.457359954732843,
+      "grad_norm": 0.8192033767700195,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 15200
+    },
+    {
+      "epoch": 2.458976638913588,
+      "grad_norm": 0.847537100315094,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 15210
+    },
+    {
+      "epoch": 2.4605933230943338,
+      "grad_norm": 0.9027776122093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 15220
+    },
+    {
+      "epoch": 2.462210007275079,
+      "grad_norm": 0.7217772006988525,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 15230
+    },
+    {
+      "epoch": 2.4638266914558242,
+      "grad_norm": 0.7994546294212341,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 15240
+    },
+    {
+      "epoch": 2.4654433756365695,
+      "grad_norm": 0.939916729927063,
+      "learning_rate": 0.0002,
+      "loss": 0.6828,
+      "step": 15250
+    },
+    {
+      "epoch": 2.4670600598173147,
+      "grad_norm": 1.0009053945541382,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15260
+    },
+    {
+      "epoch": 2.46867674399806,
+      "grad_norm": 0.625555694103241,
+      "learning_rate": 0.0002,
+      "loss": 0.643,
+      "step": 15270
+    },
+    {
+      "epoch": 2.470293428178805,
+      "grad_norm": 0.7924878597259521,
+      "learning_rate": 0.0002,
+      "loss": 0.688,
+      "step": 15280
+    },
+    {
+      "epoch": 2.4719101123595504,
+      "grad_norm": 0.8536689877510071,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 15290
+    },
+    {
+      "epoch": 2.4735267965402956,
+      "grad_norm": 0.8572589755058289,
+      "learning_rate": 0.0002,
+      "loss": 0.6924,
+      "step": 15300
+    },
+    {
+      "epoch": 2.4751434807210413,
+      "grad_norm": 0.773279070854187,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 15310
+    },
+    {
+      "epoch": 2.4767601649017865,
+      "grad_norm": 0.7708749771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 15320
+    },
+    {
+      "epoch": 2.4783768490825318,
+      "grad_norm": 0.770905077457428,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15330
+    },
+    {
+      "epoch": 2.479993533263277,
+      "grad_norm": 0.8238571882247925,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 15340
+    },
+    {
+      "epoch": 2.481610217444022,
+      "grad_norm": 0.7670477032661438,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15350
+    },
+    {
+      "epoch": 2.4832269016247674,
+      "grad_norm": 0.905036985874176,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 15360
+    },
+    {
+      "epoch": 2.484843585805513,
+      "grad_norm": 0.6672089695930481,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 15370
+    },
+    {
+      "epoch": 2.4864602699862584,
+      "grad_norm": 0.625095784664154,
+      "learning_rate": 0.0002,
+      "loss": 0.6722,
+      "step": 15380
+    },
+    {
+      "epoch": 2.4880769541670036,
+      "grad_norm": 0.679772675037384,
+      "learning_rate": 0.0002,
+      "loss": 0.6396,
+      "step": 15390
+    },
+    {
+      "epoch": 2.489693638347749,
+      "grad_norm": 0.711492121219635,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 15400
+    },
+    {
+      "epoch": 2.491310322528494,
+      "grad_norm": 0.876189112663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6966,
+      "step": 15410
+    },
+    {
+      "epoch": 2.4929270067092393,
+      "grad_norm": 0.7236915230751038,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 15420
+    },
+    {
+      "epoch": 2.4945436908899845,
+      "grad_norm": 0.6629832983016968,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 15430
+    },
+    {
+      "epoch": 2.4961603750707297,
+      "grad_norm": 0.9756859540939331,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 15440
+    },
+    {
+      "epoch": 2.4977770592514754,
+      "grad_norm": 0.6896940469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 15450
+    },
+    {
+      "epoch": 2.4993937434322206,
+      "grad_norm": 0.7105149626731873,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 15460
+    },
+    {
+      "epoch": 2.501010427612966,
+      "grad_norm": 0.8374546766281128,
+      "learning_rate": 0.0002,
+      "loss": 0.6859,
+      "step": 15470
+    },
+    {
+      "epoch": 2.502627111793711,
+      "grad_norm": 0.7320070266723633,
+      "learning_rate": 0.0002,
+      "loss": 0.6512,
+      "step": 15480
+    },
+    {
+      "epoch": 2.5042437959744563,
+      "grad_norm": 0.8306367993354797,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 15490
+    },
+    {
+      "epoch": 2.5058604801552016,
+      "grad_norm": 0.7472721338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.7253,
+      "step": 15500
+    },
+    {
+      "epoch": 2.507477164335947,
+      "grad_norm": 0.6147692203521729,
+      "learning_rate": 0.0002,
+      "loss": 0.6699,
+      "step": 15510
+    },
+    {
+      "epoch": 2.5090938485166925,
+      "grad_norm": 0.7788505554199219,
+      "learning_rate": 0.0002,
+      "loss": 0.7158,
+      "step": 15520
+    },
+    {
+      "epoch": 2.5107105326974377,
+      "grad_norm": 0.8807527422904968,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 15530
+    },
+    {
+      "epoch": 2.512327216878183,
+      "grad_norm": 0.7521643042564392,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 15540
+    },
+    {
+      "epoch": 2.513943901058928,
+      "grad_norm": 0.6900225281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15550
+    },
+    {
+      "epoch": 2.5155605852396734,
+      "grad_norm": 0.6601938605308533,
+      "learning_rate": 0.0002,
+      "loss": 0.6769,
+      "step": 15560
+    },
+    {
+      "epoch": 2.5171772694204186,
+      "grad_norm": 0.8179984092712402,
+      "learning_rate": 0.0002,
+      "loss": 0.6648,
+      "step": 15570
+    },
+    {
+      "epoch": 2.518793953601164,
+      "grad_norm": 0.792556881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.7028,
+      "step": 15580
+    },
+    {
+      "epoch": 2.520410637781909,
+      "grad_norm": 0.7081938982009888,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 15590
+    },
+    {
+      "epoch": 2.5220273219626543,
+      "grad_norm": 0.8733121156692505,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 15600
+    },
+    {
+      "epoch": 2.5236440061434,
+      "grad_norm": 0.7980992794036865,
+      "learning_rate": 0.0002,
+      "loss": 0.6969,
+      "step": 15610
+    },
+    {
+      "epoch": 2.5252606903241452,
+      "grad_norm": 0.883664071559906,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 15620
+    },
+    {
+      "epoch": 2.5268773745048905,
+      "grad_norm": 0.6963341236114502,
+      "learning_rate": 0.0002,
+      "loss": 0.7022,
+      "step": 15630
+    },
+    {
+      "epoch": 2.5284940586856357,
+      "grad_norm": 0.6433573365211487,
+      "learning_rate": 0.0002,
+      "loss": 0.7334,
+      "step": 15640
+    },
+    {
+      "epoch": 2.530110742866381,
+      "grad_norm": 0.8538183569908142,
+      "learning_rate": 0.0002,
+      "loss": 0.6889,
+      "step": 15650
+    },
+    {
+      "epoch": 2.5317274270471266,
+      "grad_norm": 0.9748201370239258,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 15660
+    },
+    {
+      "epoch": 2.533344111227872,
+      "grad_norm": 0.7670575380325317,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 15670
+    },
+    {
+      "epoch": 2.534960795408617,
+      "grad_norm": 0.8738890290260315,
+      "learning_rate": 0.0002,
+      "loss": 0.6435,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5365774795893623,
+      "grad_norm": 0.8391636610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.6802,
+      "step": 15690
+    },
+    {
+      "epoch": 2.5381941637701075,
+      "grad_norm": 0.7239366769790649,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 15700
+    },
+    {
+      "epoch": 2.5398108479508528,
+      "grad_norm": 0.8498379588127136,
+      "learning_rate": 0.0002,
+      "loss": 0.7011,
+      "step": 15710
+    },
+    {
+      "epoch": 2.541427532131598,
+      "grad_norm": 0.8029484152793884,
+      "learning_rate": 0.0002,
+      "loss": 0.6998,
+      "step": 15720
+    },
+    {
+      "epoch": 2.5430442163123432,
+      "grad_norm": 1.0639333724975586,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 15730
+    },
+    {
+      "epoch": 2.5446609004930885,
+      "grad_norm": 0.6401297450065613,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 15740
+    },
+    {
+      "epoch": 2.5462775846738337,
+      "grad_norm": 0.7123814821243286,
+      "learning_rate": 0.0002,
+      "loss": 0.7196,
+      "step": 15750
+    },
+    {
+      "epoch": 2.5478942688545794,
+      "grad_norm": 0.7874974608421326,
+      "learning_rate": 0.0002,
+      "loss": 0.654,
+      "step": 15760
+    },
+    {
+      "epoch": 2.5495109530353246,
+      "grad_norm": 0.8046808838844299,
+      "learning_rate": 0.0002,
+      "loss": 0.6721,
+      "step": 15770
+    },
+    {
+      "epoch": 2.55112763721607,
+      "grad_norm": 0.7888661623001099,
+      "learning_rate": 0.0002,
+      "loss": 0.6665,
+      "step": 15780
+    },
+    {
+      "epoch": 2.552744321396815,
+      "grad_norm": 0.8445866107940674,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15790
+    },
+    {
+      "epoch": 2.5543610055775603,
+      "grad_norm": 0.7475846409797668,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 15800
+    },
+    {
+      "epoch": 2.555977689758306,
+      "grad_norm": 0.7455102801322937,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 15810
+    },
+    {
+      "epoch": 2.557594373939051,
+      "grad_norm": 0.8226983547210693,
+      "learning_rate": 0.0002,
+      "loss": 0.6932,
+      "step": 15820
+    },
+    {
+      "epoch": 2.5592110581197964,
+      "grad_norm": 0.8920368552207947,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 15830
+    },
+    {
+      "epoch": 2.5608277423005417,
+      "grad_norm": 0.8413904905319214,
+      "learning_rate": 0.0002,
+      "loss": 0.6297,
+      "step": 15840
+    },
+    {
+      "epoch": 2.562444426481287,
+      "grad_norm": 0.8483649492263794,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 15850
+    },
+    {
+      "epoch": 2.564061110662032,
+      "grad_norm": 0.5923284292221069,
+      "learning_rate": 0.0002,
+      "loss": 0.6957,
+      "step": 15860
+    },
+    {
+      "epoch": 2.5656777948427774,
+      "grad_norm": 0.8518726229667664,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 15870
+    },
+    {
+      "epoch": 2.5672944790235226,
+      "grad_norm": 0.731235146522522,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 15880
+    },
+    {
+      "epoch": 2.568911163204268,
+      "grad_norm": 0.7517194151878357,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 15890
+    },
+    {
+      "epoch": 2.5705278473850135,
+      "grad_norm": 0.8378692269325256,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 15900
+    },
+    {
+      "epoch": 2.5721445315657587,
+      "grad_norm": 0.843701958656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 15910
+    },
+    {
+      "epoch": 2.573761215746504,
+      "grad_norm": 0.7254629731178284,
+      "learning_rate": 0.0002,
+      "loss": 0.6882,
+      "step": 15920
+    },
+    {
+      "epoch": 2.575377899927249,
+      "grad_norm": 0.8863335847854614,
+      "learning_rate": 0.0002,
+      "loss": 0.6872,
+      "step": 15930
+    },
+    {
+      "epoch": 2.5769945841079944,
+      "grad_norm": 0.7675097584724426,
+      "learning_rate": 0.0002,
+      "loss": 0.6813,
+      "step": 15940
+    },
+    {
+      "epoch": 2.5786112682887397,
+      "grad_norm": 0.82063889503479,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 15950
+    },
+    {
+      "epoch": 2.5802279524694853,
+      "grad_norm": 0.7729717493057251,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 15960
+    },
+    {
+      "epoch": 2.5818446366502306,
+      "grad_norm": 0.8301846981048584,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 15970
+    },
+    {
+      "epoch": 2.583461320830976,
+      "grad_norm": 0.7906861305236816,
+      "learning_rate": 0.0002,
+      "loss": 0.6897,
+      "step": 15980
+    },
+    {
+      "epoch": 2.585078005011721,
+      "grad_norm": 0.6749057173728943,
+      "learning_rate": 0.0002,
+      "loss": 0.7175,
+      "step": 15990
+    },
+    {
+      "epoch": 2.5866946891924663,
+      "grad_norm": 0.9386842846870422,
+      "learning_rate": 0.0002,
+      "loss": 0.7212,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5883113733732115,
+      "grad_norm": 0.7868891358375549,
+      "learning_rate": 0.0002,
+      "loss": 0.6934,
+      "step": 16010
+    },
+    {
+      "epoch": 2.5899280575539567,
+      "grad_norm": 0.8674671053886414,
+      "learning_rate": 0.0002,
+      "loss": 0.7036,
+      "step": 16020
+    },
+    {
+      "epoch": 2.591544741734702,
+      "grad_norm": 0.7043559551239014,
+      "learning_rate": 0.0002,
+      "loss": 0.7217,
+      "step": 16030
+    },
+    {
+      "epoch": 2.593161425915447,
+      "grad_norm": 0.5846083760261536,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 16040
+    },
+    {
+      "epoch": 2.594778110096193,
+      "grad_norm": 0.7323982119560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 16050
+    },
+    {
+      "epoch": 2.596394794276938,
+      "grad_norm": 0.9069556593894958,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 16060
+    },
+    {
+      "epoch": 2.5980114784576833,
+      "grad_norm": 0.7522736191749573,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 16070
+    },
+    {
+      "epoch": 2.5996281626384286,
+      "grad_norm": 0.8149648308753967,
+      "learning_rate": 0.0002,
+      "loss": 0.6477,
+      "step": 16080
+    },
+    {
+      "epoch": 2.601244846819174,
+      "grad_norm": 0.6214233040809631,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 16090
+    },
+    {
+      "epoch": 2.602861530999919,
+      "grad_norm": 0.6803743839263916,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 16100
+    },
+    {
+      "epoch": 2.6044782151806647,
+      "grad_norm": 0.7223997116088867,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 16110
+    },
+    {
+      "epoch": 2.60609489936141,
+      "grad_norm": 0.7324174642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.6867,
+      "step": 16120
+    },
+    {
+      "epoch": 2.607711583542155,
+      "grad_norm": 0.9594739675521851,
+      "learning_rate": 0.0002,
+      "loss": 0.7159,
+      "step": 16130
+    },
+    {
+      "epoch": 2.6093282677229004,
+      "grad_norm": 0.9485327005386353,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 16140
+    },
+    {
+      "epoch": 2.6109449519036456,
+      "grad_norm": 0.8449000120162964,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 16150
+    },
+    {
+      "epoch": 2.612561636084391,
+      "grad_norm": 0.8520140051841736,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 16160
+    },
+    {
+      "epoch": 2.614178320265136,
+      "grad_norm": 0.7456524968147278,
+      "learning_rate": 0.0002,
+      "loss": 0.6759,
+      "step": 16170
+    },
+    {
+      "epoch": 2.6157950044458813,
+      "grad_norm": 0.9912857413291931,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 16180
+    },
+    {
+      "epoch": 2.6174116886266265,
+      "grad_norm": 0.9001946449279785,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 16190
+    },
+    {
+      "epoch": 2.619028372807372,
+      "grad_norm": 0.6568667888641357,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16200
+    },
+    {
+      "epoch": 2.6206450569881174,
+      "grad_norm": 1.0248128175735474,
+      "learning_rate": 0.0002,
+      "loss": 0.7013,
+      "step": 16210
+    },
+    {
+      "epoch": 2.6222617411688627,
+      "grad_norm": 0.6509039998054504,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 16220
+    },
+    {
+      "epoch": 2.623878425349608,
+      "grad_norm": 0.7626351118087769,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 16230
+    },
+    {
+      "epoch": 2.625495109530353,
+      "grad_norm": 0.6938552260398865,
+      "learning_rate": 0.0002,
+      "loss": 0.6556,
+      "step": 16240
+    },
+    {
+      "epoch": 2.6271117937110984,
+      "grad_norm": 0.6434680819511414,
+      "learning_rate": 0.0002,
+      "loss": 0.65,
+      "step": 16250
+    },
+    {
+      "epoch": 2.628728477891844,
+      "grad_norm": 0.7111515998840332,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 16260
+    },
+    {
+      "epoch": 2.6303451620725893,
+      "grad_norm": 0.7712395787239075,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 16270
+    },
+    {
+      "epoch": 2.6319618462533345,
+      "grad_norm": 0.792209267616272,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 16280
+    },
+    {
+      "epoch": 2.6335785304340797,
+      "grad_norm": 0.6801066398620605,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 16290
+    },
+    {
+      "epoch": 2.635195214614825,
+      "grad_norm": 0.7802573442459106,
+      "learning_rate": 0.0002,
+      "loss": 0.73,
+      "step": 16300
+    },
+    {
+      "epoch": 2.63681189879557,
+      "grad_norm": 0.7742244601249695,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 16310
+    },
+    {
+      "epoch": 2.6384285829763154,
+      "grad_norm": 0.664184033870697,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 16320
+    },
+    {
+      "epoch": 2.6400452671570607,
+      "grad_norm": 0.9242228865623474,
+      "learning_rate": 0.0002,
+      "loss": 0.6442,
+      "step": 16330
+    },
+    {
+      "epoch": 2.641661951337806,
+      "grad_norm": 0.9661325216293335,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 16340
+    },
+    {
+      "epoch": 2.6432786355185516,
+      "grad_norm": 0.837526798248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 16350
+    },
+    {
+      "epoch": 2.644895319699297,
+      "grad_norm": 1.1834373474121094,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 16360
+    },
+    {
+      "epoch": 2.646512003880042,
+      "grad_norm": 0.7467831373214722,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 16370
+    },
+    {
+      "epoch": 2.6481286880607873,
+      "grad_norm": 0.8627146482467651,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 16380
+    },
+    {
+      "epoch": 2.6497453722415325,
+      "grad_norm": 0.790447473526001,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 16390
+    },
+    {
+      "epoch": 2.651362056422278,
+      "grad_norm": 0.8447365164756775,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 16400
+    },
+    {
+      "epoch": 2.6529787406030234,
+      "grad_norm": 0.7831417918205261,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 16410
+    },
+    {
+      "epoch": 2.6545954247837686,
+      "grad_norm": 0.6837952136993408,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 16420
+    },
+    {
+      "epoch": 2.656212108964514,
+      "grad_norm": 0.7031801342964172,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 16430
+    },
+    {
+      "epoch": 2.657828793145259,
+      "grad_norm": 0.8963770866394043,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 16440
+    },
+    {
+      "epoch": 2.6594454773260043,
+      "grad_norm": 0.6852328181266785,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 16450
+    },
+    {
+      "epoch": 2.6610621615067496,
+      "grad_norm": 0.8069294095039368,
+      "learning_rate": 0.0002,
+      "loss": 0.6278,
+      "step": 16460
+    },
+    {
+      "epoch": 2.662678845687495,
+      "grad_norm": 0.7503686547279358,
+      "learning_rate": 0.0002,
+      "loss": 0.6939,
+      "step": 16470
+    },
+    {
+      "epoch": 2.66429552986824,
+      "grad_norm": 0.6430956125259399,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16480
+    },
+    {
+      "epoch": 2.6659122140489853,
+      "grad_norm": 0.7894312739372253,
+      "learning_rate": 0.0002,
+      "loss": 0.6863,
+      "step": 16490
+    },
+    {
+      "epoch": 2.667528898229731,
+      "grad_norm": 0.7277431488037109,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 16500
+    },
+    {
+      "epoch": 2.669145582410476,
+      "grad_norm": 0.6816153526306152,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 16510
+    },
+    {
+      "epoch": 2.6707622665912214,
+      "grad_norm": 0.8145235776901245,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 16520
+    },
+    {
+      "epoch": 2.6723789507719666,
+      "grad_norm": 0.8645890355110168,
+      "learning_rate": 0.0002,
+      "loss": 0.709,
+      "step": 16530
+    },
+    {
+      "epoch": 2.673995634952712,
+      "grad_norm": 0.704393208026886,
+      "learning_rate": 0.0002,
+      "loss": 0.6946,
+      "step": 16540
+    },
+    {
+      "epoch": 2.6756123191334575,
+      "grad_norm": 1.0120846033096313,
+      "learning_rate": 0.0002,
+      "loss": 0.6378,
+      "step": 16550
+    },
+    {
+      "epoch": 2.6772290033142028,
+      "grad_norm": 0.6919328570365906,
+      "learning_rate": 0.0002,
+      "loss": 0.7241,
+      "step": 16560
+    },
+    {
+      "epoch": 2.678845687494948,
+      "grad_norm": 0.6924574971199036,
+      "learning_rate": 0.0002,
+      "loss": 0.7098,
+      "step": 16570
+    },
+    {
+      "epoch": 2.6804623716756932,
+      "grad_norm": 0.9679301381111145,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 16580
+    },
+    {
+      "epoch": 2.6820790558564385,
+      "grad_norm": 0.6810211539268494,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 16590
+    },
+    {
+      "epoch": 2.6836957400371837,
+      "grad_norm": 0.9730555415153503,
+      "learning_rate": 0.0002,
+      "loss": 0.6688,
+      "step": 16600
+    },
+    {
+      "epoch": 2.685312424217929,
+      "grad_norm": 0.7852821350097656,
+      "learning_rate": 0.0002,
+      "loss": 0.7344,
+      "step": 16610
+    },
+    {
+      "epoch": 2.686929108398674,
+      "grad_norm": 0.6059057116508484,
+      "learning_rate": 0.0002,
+      "loss": 0.6401,
+      "step": 16620
+    },
+    {
+      "epoch": 2.6885457925794194,
+      "grad_norm": 0.9395958781242371,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 16630
+    },
+    {
+      "epoch": 2.690162476760165,
+      "grad_norm": 0.7473729848861694,
+      "learning_rate": 0.0002,
+      "loss": 0.7174,
+      "step": 16640
+    },
+    {
+      "epoch": 2.6917791609409103,
+      "grad_norm": 0.765934407711029,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 16650
+    },
+    {
+      "epoch": 2.6933958451216555,
+      "grad_norm": 0.8496677279472351,
+      "learning_rate": 0.0002,
+      "loss": 0.707,
+      "step": 16660
+    },
+    {
+      "epoch": 2.6950125293024008,
+      "grad_norm": 0.7641879916191101,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 16670
+    },
+    {
+      "epoch": 2.696629213483146,
+      "grad_norm": 0.8471952676773071,
+      "learning_rate": 0.0002,
+      "loss": 0.6566,
+      "step": 16680
+    },
+    {
+      "epoch": 2.6982458976638912,
+      "grad_norm": 0.6946060657501221,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 16690
+    },
+    {
+      "epoch": 2.699862581844637,
+      "grad_norm": 0.7361312508583069,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 16700
+    },
+    {
+      "epoch": 2.701479266025382,
+      "grad_norm": 0.6605038046836853,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 16710
+    },
+    {
+      "epoch": 2.7030959502061274,
+      "grad_norm": 0.7164411544799805,
+      "learning_rate": 0.0002,
+      "loss": 0.6885,
+      "step": 16720
+    },
+    {
+      "epoch": 2.7047126343868726,
+      "grad_norm": 0.6496201157569885,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 16730
+    },
+    {
+      "epoch": 2.706329318567618,
+      "grad_norm": 0.7826663851737976,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 16740
+    },
+    {
+      "epoch": 2.707946002748363,
+      "grad_norm": 0.7639131546020508,
+      "learning_rate": 0.0002,
+      "loss": 0.6773,
+      "step": 16750
+    },
+    {
+      "epoch": 2.7095626869291083,
+      "grad_norm": 0.7976210713386536,
+      "learning_rate": 0.0002,
+      "loss": 0.69,
+      "step": 16760
+    },
+    {
+      "epoch": 2.7111793711098535,
+      "grad_norm": 0.6836577653884888,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 16770
+    },
+    {
+      "epoch": 2.7127960552905988,
+      "grad_norm": 0.8025202751159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 16780
+    },
+    {
+      "epoch": 2.7144127394713444,
+      "grad_norm": 0.7636463642120361,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7160294236520897,
+      "grad_norm": 0.7481677532196045,
+      "learning_rate": 0.0002,
+      "loss": 0.6227,
+      "step": 16800
+    },
+    {
+      "epoch": 2.717646107832835,
+      "grad_norm": 0.7566834688186646,
+      "learning_rate": 0.0002,
+      "loss": 0.6925,
+      "step": 16810
+    },
+    {
+      "epoch": 2.71926279201358,
+      "grad_norm": 0.7931267619132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6531,
+      "step": 16820
+    },
+    {
+      "epoch": 2.7208794761943254,
+      "grad_norm": 0.8811662197113037,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 16830
+    },
+    {
+      "epoch": 2.7224961603750706,
+      "grad_norm": 0.8561240434646606,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 16840
+    },
+    {
+      "epoch": 2.7241128445558163,
+      "grad_norm": 0.7121599316596985,
+      "learning_rate": 0.0002,
+      "loss": 0.7135,
+      "step": 16850
+    },
+    {
+      "epoch": 2.7257295287365615,
+      "grad_norm": 0.8066257238388062,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16860
+    },
+    {
+      "epoch": 2.7273462129173067,
+      "grad_norm": 0.7699271440505981,
+      "learning_rate": 0.0002,
+      "loss": 0.6839,
+      "step": 16870
+    },
+    {
+      "epoch": 2.728962897098052,
+      "grad_norm": 1.1828432083129883,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 16880
+    },
+    {
+      "epoch": 2.730579581278797,
+      "grad_norm": 0.9989302754402161,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 16890
+    },
+    {
+      "epoch": 2.7321962654595424,
+      "grad_norm": 0.8100560307502747,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 16900
+    },
+    {
+      "epoch": 2.7338129496402876,
+      "grad_norm": 0.8615233898162842,
+      "learning_rate": 0.0002,
+      "loss": 0.6851,
+      "step": 16910
+    },
+    {
+      "epoch": 2.735429633821033,
+      "grad_norm": 0.8633756041526794,
+      "learning_rate": 0.0002,
+      "loss": 0.6322,
+      "step": 16920
+    },
+    {
+      "epoch": 2.737046318001778,
+      "grad_norm": 0.7769348621368408,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 16930
+    },
+    {
+      "epoch": 2.738663002182524,
+      "grad_norm": 0.6943058371543884,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 16940
+    },
+    {
+      "epoch": 2.740279686363269,
+      "grad_norm": 0.8510736227035522,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 16950
+    },
+    {
+      "epoch": 2.7418963705440142,
+      "grad_norm": 0.7732602953910828,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 16960
+    },
+    {
+      "epoch": 2.7435130547247595,
+      "grad_norm": 0.5981788635253906,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 16970
+    },
+    {
+      "epoch": 2.7451297389055047,
+      "grad_norm": 0.7604416012763977,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16980
+    },
+    {
+      "epoch": 2.74674642308625,
+      "grad_norm": 0.7377738356590271,
+      "learning_rate": 0.0002,
+      "loss": 0.6981,
+      "step": 16990
+    },
+    {
+      "epoch": 2.7483631072669956,
+      "grad_norm": 0.9400289058685303,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 17000
+    },
+    {
+      "epoch": 2.749979791447741,
+      "grad_norm": 0.6340599656105042,
+      "learning_rate": 0.0002,
+      "loss": 0.6952,
+      "step": 17010
+    },
+    {
+      "epoch": 2.751596475628486,
+      "grad_norm": 0.7297601103782654,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 17020
+    },
+    {
+      "epoch": 2.7532131598092313,
+      "grad_norm": 0.9479979872703552,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 17030
+    },
+    {
+      "epoch": 2.7548298439899765,
+      "grad_norm": 0.8461511135101318,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 17040
+    },
+    {
+      "epoch": 2.7564465281707218,
+      "grad_norm": 0.7477551698684692,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17050
+    },
+    {
+      "epoch": 2.758063212351467,
+      "grad_norm": 1.019270420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 17060
+    },
+    {
+      "epoch": 2.7596798965322122,
+      "grad_norm": 0.7730235457420349,
+      "learning_rate": 0.0002,
+      "loss": 0.7119,
+      "step": 17070
+    },
+    {
+      "epoch": 2.7612965807129575,
+      "grad_norm": 0.8216866254806519,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 17080
+    },
+    {
+      "epoch": 2.762913264893703,
+      "grad_norm": 0.7235931754112244,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17090
+    },
+    {
+      "epoch": 2.7645299490744484,
+      "grad_norm": 0.7352296710014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7031,
+      "step": 17100
+    },
+    {
+      "epoch": 2.7661466332551936,
+      "grad_norm": 0.8129373788833618,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 17110
+    },
+    {
+      "epoch": 2.767763317435939,
+      "grad_norm": 0.7387019991874695,
+      "learning_rate": 0.0002,
+      "loss": 0.6703,
+      "step": 17120
+    },
+    {
+      "epoch": 2.769380001616684,
+      "grad_norm": 0.9149190187454224,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 17130
+    },
+    {
+      "epoch": 2.7709966857974297,
+      "grad_norm": 0.7352971434593201,
+      "learning_rate": 0.0002,
+      "loss": 0.6038,
+      "step": 17140
+    },
+    {
+      "epoch": 2.772613369978175,
+      "grad_norm": 0.7903780341148376,
+      "learning_rate": 0.0002,
+      "loss": 0.6728,
+      "step": 17150
+    },
+    {
+      "epoch": 2.77423005415892,
+      "grad_norm": 0.8255927562713623,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17160
+    },
+    {
+      "epoch": 2.7758467383396654,
+      "grad_norm": 0.7235927581787109,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 17170
+    },
+    {
+      "epoch": 2.7774634225204107,
+      "grad_norm": 0.8281434774398804,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 17180
+    },
+    {
+      "epoch": 2.779080106701156,
+      "grad_norm": 0.7586921453475952,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 17190
+    },
+    {
+      "epoch": 2.780696790881901,
+      "grad_norm": 0.7161715030670166,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 17200
+    },
+    {
+      "epoch": 2.7823134750626464,
+      "grad_norm": 0.762868344783783,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 17210
+    },
+    {
+      "epoch": 2.7839301592433916,
+      "grad_norm": 0.9285483360290527,
+      "learning_rate": 0.0002,
+      "loss": 0.705,
+      "step": 17220
+    },
+    {
+      "epoch": 2.785546843424137,
+      "grad_norm": 0.6900462508201599,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 17230
+    },
+    {
+      "epoch": 2.7871635276048825,
+      "grad_norm": 0.780384361743927,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17240
+    },
+    {
+      "epoch": 2.7887802117856277,
+      "grad_norm": 0.7580406665802002,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17250
+    },
+    {
+      "epoch": 2.790396895966373,
+      "grad_norm": 0.8145199418067932,
+      "learning_rate": 0.0002,
+      "loss": 0.6833,
+      "step": 17260
+    },
+    {
+      "epoch": 2.792013580147118,
+      "grad_norm": 0.9159596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 17270
+    },
+    {
+      "epoch": 2.7936302643278634,
+      "grad_norm": 0.9590014219284058,
+      "learning_rate": 0.0002,
+      "loss": 0.6008,
+      "step": 17280
+    },
+    {
+      "epoch": 2.795246948508609,
+      "grad_norm": 0.7603529691696167,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 17290
+    },
+    {
+      "epoch": 2.7968636326893543,
+      "grad_norm": 0.8039976358413696,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 17300
+    },
+    {
+      "epoch": 2.7984803168700996,
+      "grad_norm": 0.8364847302436829,
+      "learning_rate": 0.0002,
+      "loss": 0.7037,
+      "step": 17310
+    },
+    {
+      "epoch": 2.800097001050845,
+      "grad_norm": 0.8763046860694885,
+      "learning_rate": 0.0002,
+      "loss": 0.6749,
+      "step": 17320
+    },
+    {
+      "epoch": 2.80171368523159,
+      "grad_norm": 0.8409647941589355,
+      "learning_rate": 0.0002,
+      "loss": 0.6844,
+      "step": 17330
+    },
+    {
+      "epoch": 2.8033303694123353,
+      "grad_norm": 0.7649006247520447,
+      "learning_rate": 0.0002,
+      "loss": 0.6936,
+      "step": 17340
+    },
+    {
+      "epoch": 2.8049470535930805,
+      "grad_norm": 0.7970262169837952,
+      "learning_rate": 0.0002,
+      "loss": 0.7051,
+      "step": 17350
+    },
+    {
+      "epoch": 2.8065637377738257,
+      "grad_norm": 0.9088607430458069,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 17360
+    },
+    {
+      "epoch": 2.808180421954571,
+      "grad_norm": 0.6454846858978271,
+      "learning_rate": 0.0002,
+      "loss": 0.675,
+      "step": 17370
+    },
+    {
+      "epoch": 2.809797106135316,
+      "grad_norm": 0.7744787931442261,
+      "learning_rate": 0.0002,
+      "loss": 0.7069,
+      "step": 17380
+    },
+    {
+      "epoch": 2.811413790316062,
+      "grad_norm": 0.6678640842437744,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 17390
+    },
+    {
+      "epoch": 2.813030474496807,
+      "grad_norm": 0.772676944732666,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 17400
+    },
+    {
+      "epoch": 2.8146471586775523,
+      "grad_norm": 0.7088175415992737,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 17410
+    },
+    {
+      "epoch": 2.8162638428582976,
+      "grad_norm": 0.8280573487281799,
+      "learning_rate": 0.0002,
+      "loss": 0.7086,
+      "step": 17420
+    },
+    {
+      "epoch": 2.817880527039043,
+      "grad_norm": 0.6665388345718384,
+      "learning_rate": 0.0002,
+      "loss": 0.6732,
+      "step": 17430
+    },
+    {
+      "epoch": 2.8194972112197885,
+      "grad_norm": 0.6427883505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 17440
+    },
+    {
+      "epoch": 2.8211138954005337,
+      "grad_norm": 0.9697760343551636,
+      "learning_rate": 0.0002,
+      "loss": 0.6972,
+      "step": 17450
+    },
+    {
+      "epoch": 2.822730579581279,
+      "grad_norm": 0.7573966383934021,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 17460
+    },
+    {
+      "epoch": 2.824347263762024,
+      "grad_norm": 0.878688633441925,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 17470
+    },
+    {
+      "epoch": 2.8259639479427694,
+      "grad_norm": 0.7752242684364319,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 17480
+    },
+    {
+      "epoch": 2.8275806321235146,
+      "grad_norm": 0.6135398745536804,
+      "learning_rate": 0.0002,
+      "loss": 0.6638,
+      "step": 17490
+    },
+    {
+      "epoch": 2.82919731630426,
+      "grad_norm": 0.6924924850463867,
+      "learning_rate": 0.0002,
+      "loss": 0.6829,
+      "step": 17500
+    },
+    {
+      "epoch": 2.830814000485005,
+      "grad_norm": 0.7471627593040466,
+      "learning_rate": 0.0002,
+      "loss": 0.6731,
+      "step": 17510
+    },
+    {
+      "epoch": 2.8324306846657503,
+      "grad_norm": 0.7145499587059021,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 17520
+    },
+    {
+      "epoch": 2.834047368846496,
+      "grad_norm": 0.7415414452552795,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17530
+    },
+    {
+      "epoch": 2.8356640530272412,
+      "grad_norm": 0.7328441739082336,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17540
+    },
+    {
+      "epoch": 2.8372807372079865,
+      "grad_norm": 0.8267839550971985,
+      "learning_rate": 0.0002,
+      "loss": 0.6866,
+      "step": 17550
+    },
+    {
+      "epoch": 2.8388974213887317,
+      "grad_norm": 0.8877885341644287,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17560
+    },
+    {
+      "epoch": 2.840514105569477,
+      "grad_norm": 0.857138454914093,
+      "learning_rate": 0.0002,
+      "loss": 0.7136,
+      "step": 17570
+    },
+    {
+      "epoch": 2.842130789750222,
+      "grad_norm": 0.8470779657363892,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 17580
+    },
+    {
+      "epoch": 2.843747473930968,
+      "grad_norm": 0.8553254008293152,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 17590
+    },
+    {
+      "epoch": 2.845364158111713,
+      "grad_norm": 0.8033196926116943,
+      "learning_rate": 0.0002,
+      "loss": 0.7297,
+      "step": 17600
+    },
+    {
+      "epoch": 2.8469808422924583,
+      "grad_norm": 0.7949087023735046,
+      "learning_rate": 0.0002,
+      "loss": 0.7062,
+      "step": 17610
+    },
+    {
+      "epoch": 2.8485975264732035,
+      "grad_norm": 0.9241406321525574,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 17620
+    },
+    {
+      "epoch": 2.8502142106539488,
+      "grad_norm": 0.7721285223960876,
+      "learning_rate": 0.0002,
+      "loss": 0.6601,
+      "step": 17630
+    },
+    {
+      "epoch": 2.851830894834694,
+      "grad_norm": 1.0246692895889282,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 17640
+    },
+    {
+      "epoch": 2.853447579015439,
+      "grad_norm": 0.9244589805603027,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 17650
+    },
+    {
+      "epoch": 2.8550642631961844,
+      "grad_norm": 0.7243508696556091,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 17660
+    },
+    {
+      "epoch": 2.8566809473769297,
+      "grad_norm": 0.8943371176719666,
+      "learning_rate": 0.0002,
+      "loss": 0.6471,
+      "step": 17670
+    },
+    {
+      "epoch": 2.8582976315576754,
+      "grad_norm": 0.6531758904457092,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 17680
+    },
+    {
+      "epoch": 2.8599143157384206,
+      "grad_norm": 0.8367000818252563,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 17690
+    },
+    {
+      "epoch": 2.861530999919166,
+      "grad_norm": 0.7868556380271912,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 17700
+    },
+    {
+      "epoch": 2.863147684099911,
+      "grad_norm": 0.7213859558105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 17710
+    },
+    {
+      "epoch": 2.8647643682806563,
+      "grad_norm": 0.7383931279182434,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 17720
+    },
+    {
+      "epoch": 2.8663810524614015,
+      "grad_norm": 0.7566812634468079,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 17730
+    },
+    {
+      "epoch": 2.867997736642147,
+      "grad_norm": 0.6930373311042786,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 17740
+    },
+    {
+      "epoch": 2.8696144208228924,
+      "grad_norm": 0.7911090850830078,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17750
+    },
+    {
+      "epoch": 2.8712311050036377,
+      "grad_norm": 0.8484548926353455,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 17760
+    },
+    {
+      "epoch": 2.872847789184383,
+      "grad_norm": 0.7647597193717957,
+      "learning_rate": 0.0002,
+      "loss": 0.6931,
+      "step": 17770
+    },
+    {
+      "epoch": 2.874464473365128,
+      "grad_norm": 0.8791151642799377,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 17780
+    },
+    {
+      "epoch": 2.8760811575458733,
+      "grad_norm": 0.7253178358078003,
+      "learning_rate": 0.0002,
+      "loss": 0.7078,
+      "step": 17790
+    },
+    {
+      "epoch": 2.8776978417266186,
+      "grad_norm": 0.7956077456474304,
+      "learning_rate": 0.0002,
+      "loss": 0.6474,
+      "step": 17800
+    },
+    {
+      "epoch": 2.879314525907364,
+      "grad_norm": 0.8657688498497009,
+      "learning_rate": 0.0002,
+      "loss": 0.6687,
+      "step": 17810
+    },
+    {
+      "epoch": 2.880931210088109,
+      "grad_norm": 0.7059141993522644,
+      "learning_rate": 0.0002,
+      "loss": 0.7171,
+      "step": 17820
+    },
+    {
+      "epoch": 2.8825478942688547,
+      "grad_norm": 0.8886896967887878,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 17830
+    },
+    {
+      "epoch": 2.8841645784496,
+      "grad_norm": 0.821032702922821,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 17840
+    },
+    {
+      "epoch": 2.885781262630345,
+      "grad_norm": 0.7183963656425476,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 17850
+    },
+    {
+      "epoch": 2.8873979468110904,
+      "grad_norm": 0.6222899556159973,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 17860
+    },
+    {
+      "epoch": 2.8890146309918356,
+      "grad_norm": 0.8187434077262878,
+      "learning_rate": 0.0002,
+      "loss": 0.6626,
+      "step": 17870
+    },
+    {
+      "epoch": 2.890631315172581,
+      "grad_norm": 0.9838479161262512,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 17880
+    },
+    {
+      "epoch": 2.8922479993533265,
+      "grad_norm": 0.7567742466926575,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 17890
+    },
+    {
+      "epoch": 2.893864683534072,
+      "grad_norm": 0.6875903606414795,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17900
+    },
+    {
+      "epoch": 2.895481367714817,
+      "grad_norm": 0.8043789267539978,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 17910
+    },
+    {
+      "epoch": 2.8970980518955622,
+      "grad_norm": 0.8062626719474792,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 17920
+    },
+    {
+      "epoch": 2.8987147360763075,
+      "grad_norm": 1.0251191854476929,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 17930
+    },
+    {
+      "epoch": 2.9003314202570527,
+      "grad_norm": 0.882253110408783,
+      "learning_rate": 0.0002,
+      "loss": 0.6605,
+      "step": 17940
+    },
+    {
+      "epoch": 2.901948104437798,
+      "grad_norm": 0.8683299422264099,
+      "learning_rate": 0.0002,
+      "loss": 0.6719,
+      "step": 17950
+    },
+    {
+      "epoch": 2.903564788618543,
+      "grad_norm": 0.7167282104492188,
+      "learning_rate": 0.0002,
+      "loss": 0.6896,
+      "step": 17960
+    },
+    {
+      "epoch": 2.9051814727992884,
+      "grad_norm": 0.7093694806098938,
+      "learning_rate": 0.0002,
+      "loss": 0.663,
+      "step": 17970
+    },
+    {
+      "epoch": 2.906798156980034,
+      "grad_norm": 0.8549879193305969,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 17980
+    },
+    {
+      "epoch": 2.9084148411607793,
+      "grad_norm": 0.6989606618881226,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 17990
+    },
+    {
+      "epoch": 2.9100315253415245,
+      "grad_norm": 0.9482976794242859,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 18000
+    },
+    {
+      "epoch": 2.9116482095222698,
+      "grad_norm": 0.7182440161705017,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 18010
+    },
+    {
+      "epoch": 2.913264893703015,
+      "grad_norm": 0.7732226252555847,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 18020
+    },
+    {
+      "epoch": 2.9148815778837607,
+      "grad_norm": 0.7936875224113464,
+      "learning_rate": 0.0002,
+      "loss": 0.7123,
+      "step": 18030
+    },
+    {
+      "epoch": 2.916498262064506,
+      "grad_norm": 0.8825615644454956,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 18040
+    },
+    {
+      "epoch": 2.918114946245251,
+      "grad_norm": 0.6778587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 18050
+    },
+    {
+      "epoch": 2.9197316304259964,
+      "grad_norm": 0.7529265880584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 18060
+    },
+    {
+      "epoch": 2.9213483146067416,
+      "grad_norm": 0.7111883163452148,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 18070
+    },
+    {
+      "epoch": 2.922964998787487,
+      "grad_norm": 0.7214767932891846,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 18080
+    },
+    {
+      "epoch": 2.924581682968232,
+      "grad_norm": 0.800417423248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 18090
+    },
+    {
+      "epoch": 2.9261983671489773,
+      "grad_norm": 1.248575210571289,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 18100
+    },
+    {
+      "epoch": 2.9278150513297225,
+      "grad_norm": 0.757788360118866,
+      "learning_rate": 0.0002,
+      "loss": 0.7004,
+      "step": 18110
+    },
+    {
+      "epoch": 2.9294317355104678,
+      "grad_norm": 1.0583995580673218,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 18120
+    },
+    {
+      "epoch": 2.9310484196912134,
+      "grad_norm": 0.8228777647018433,
+      "learning_rate": 0.0002,
+      "loss": 0.6365,
+      "step": 18130
+    },
+    {
+      "epoch": 2.9326651038719587,
+      "grad_norm": 0.8374035358428955,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 18140
+    },
+    {
+      "epoch": 2.934281788052704,
+      "grad_norm": 0.7976473569869995,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 18150
+    },
+    {
+      "epoch": 2.935898472233449,
+      "grad_norm": 0.8009907603263855,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 18160
+    },
+    {
+      "epoch": 2.9375151564141944,
+      "grad_norm": 0.835213303565979,
+      "learning_rate": 0.0002,
+      "loss": 0.7485,
+      "step": 18170
+    },
+    {
+      "epoch": 2.93913184059494,
+      "grad_norm": 0.7982219457626343,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 18180
+    },
+    {
+      "epoch": 2.9407485247756853,
+      "grad_norm": 0.7070978879928589,
+      "learning_rate": 0.0002,
+      "loss": 0.6348,
+      "step": 18190
+    },
+    {
+      "epoch": 2.9423652089564305,
+      "grad_norm": 0.8619440197944641,
+      "learning_rate": 0.0002,
+      "loss": 0.6608,
+      "step": 18200
+    },
+    {
+      "epoch": 2.9439818931371757,
+      "grad_norm": 0.6693987250328064,
+      "learning_rate": 0.0002,
+      "loss": 0.666,
+      "step": 18210
+    },
+    {
+      "epoch": 2.945598577317921,
+      "grad_norm": 0.6747021079063416,
+      "learning_rate": 0.0002,
+      "loss": 0.728,
+      "step": 18220
+    },
+    {
+      "epoch": 2.947215261498666,
+      "grad_norm": 0.860387921333313,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 18230
+    },
+    {
+      "epoch": 2.9488319456794114,
+      "grad_norm": 0.799976646900177,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 18240
+    },
+    {
+      "epoch": 2.9504486298601567,
+      "grad_norm": 0.7864769101142883,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 18250
+    },
+    {
+      "epoch": 2.952065314040902,
+      "grad_norm": 0.6713884472846985,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 18260
+    },
+    {
+      "epoch": 2.9536819982216476,
+      "grad_norm": 0.9031508564949036,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 18270
+    },
+    {
+      "epoch": 2.955298682402393,
+      "grad_norm": 0.7205073237419128,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 18280
+    },
+    {
+      "epoch": 2.956915366583138,
+      "grad_norm": 0.7746205925941467,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 18290
+    },
+    {
+      "epoch": 2.9585320507638833,
+      "grad_norm": 0.6533427834510803,
+      "learning_rate": 0.0002,
+      "loss": 0.6218,
+      "step": 18300
+    },
+    {
+      "epoch": 2.9601487349446285,
+      "grad_norm": 0.9083208441734314,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 18310
+    },
+    {
+      "epoch": 2.9617654191253737,
+      "grad_norm": 0.7446991801261902,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 18320
+    },
+    {
+      "epoch": 2.9633821033061194,
+      "grad_norm": 0.6514461636543274,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 18330
+    },
+    {
+      "epoch": 2.9649987874868646,
+      "grad_norm": 0.8580465912818909,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18340
+    },
+    {
+      "epoch": 2.96661547166761,
+      "grad_norm": 0.7074266076087952,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 18350
+    },
+    {
+      "epoch": 2.968232155848355,
+      "grad_norm": 0.899892270565033,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 18360
+    },
+    {
+      "epoch": 2.9698488400291003,
+      "grad_norm": 0.8217641711235046,
+      "learning_rate": 0.0002,
+      "loss": 0.7094,
+      "step": 18370
+    },
+    {
+      "epoch": 2.9714655242098456,
+      "grad_norm": 0.8611799478530884,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 18380
+    },
+    {
+      "epoch": 2.973082208390591,
+      "grad_norm": 0.6909302473068237,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18390
+    },
+    {
+      "epoch": 2.974698892571336,
+      "grad_norm": 0.6554358005523682,
+      "learning_rate": 0.0002,
+      "loss": 0.7247,
+      "step": 18400
+    },
+    {
+      "epoch": 2.9763155767520812,
+      "grad_norm": 0.7803071737289429,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 18410
+    },
+    {
+      "epoch": 2.977932260932827,
+      "grad_norm": 0.7838954925537109,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 18420
+    },
+    {
+      "epoch": 2.979548945113572,
+      "grad_norm": 0.7098495364189148,
+      "learning_rate": 0.0002,
+      "loss": 0.6522,
+      "step": 18430
+    },
+    {
+      "epoch": 2.9811656292943174,
+      "grad_norm": 0.8981785774230957,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 18440
+    },
+    {
+      "epoch": 2.9827823134750626,
+      "grad_norm": 0.7197171449661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6689,
+      "step": 18450
+    },
+    {
+      "epoch": 2.984398997655808,
+      "grad_norm": 0.793185293674469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 18460
+    },
+    {
+      "epoch": 2.986015681836553,
+      "grad_norm": 0.8531473875045776,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 18470
+    },
+    {
+      "epoch": 2.9876323660172988,
+      "grad_norm": 0.6627361178398132,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 18480
+    },
+    {
+      "epoch": 2.989249050198044,
+      "grad_norm": 0.5708155035972595,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 18490
+    },
+    {
+      "epoch": 2.990865734378789,
+      "grad_norm": 0.8227280378341675,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 18500
+    },
+    {
+      "epoch": 2.9924824185595345,
+      "grad_norm": 0.7102749943733215,
+      "learning_rate": 0.0002,
+      "loss": 0.6701,
+      "step": 18510
+    },
+    {
+      "epoch": 2.9940991027402797,
+      "grad_norm": 0.839485228061676,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 18520
+    },
+    {
+      "epoch": 2.995715786921025,
+      "grad_norm": 0.9038704037666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 18530
+    },
+    {
+      "epoch": 2.99733247110177,
+      "grad_norm": 0.8737510442733765,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 18540
+    },
+    {
+      "epoch": 2.9989491552825154,
+      "grad_norm": 0.7323142886161804,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 18550
+    },
+    {
+      "epoch": 2.9999191657909625,
+      "eval_loss": 1.1262480020523071,
+      "eval_runtime": 122.0868,
+      "eval_samples_per_second": 6.004,
+      "eval_steps_per_second": 0.754,
+      "step": 18556
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 49480,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.587532703136481e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/training_args.bin b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..155b12fa9acbc6e71dba75c92bfa79e152397ebf
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-18556/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28694d5564a2b5c7d6881d4ba2af103356aa22489d2c22768ebbe47283c0f4a1
+size 5560
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/README.md b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..835e31ab8469ee39ddc8b2b6b2143a8c66dad510
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/README.md
@@ -0,0 +1,202 @@
+---
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.13.1
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/adapter_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..406c5a08dc4a2a33b52c62a482f98c217c417215
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/adapter_config.json
@@ -0,0 +1,29 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/adapter_model.safetensors b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..b4728a8d41cdf93d8b529cdfd1f99a5be2c754d8
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:932a3fd701339ac0d6d24098de64692c58e2f5819c1c254b5170b90f9786ef3f
+size 109069176
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/optimizer.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f141976366a4602a7f1e902746a2e9c591c03cdc
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4abee7ce2c656dbc963450bea7ad045f80818ca885875b7270228650d2d85e87
+size 55532666
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/rng_state.pth b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..8aa782afef5fb506c1085024c63d0b3f35500787
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f176776c36fdc2d6595bb77fc99e5132f347416e5d12b98e23c7bd4498df3581
+size 14244
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/scheduler.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1bcab7a111927143da964bbcdfcdd3a7fb414748
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b949cbb2a6d4962c67bf28344950c416a55c88bdaab30dddd07052abcc682b3a
+size 1064
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/special_tokens_map.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/special_tokens_map.json
@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/tokenizer.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..86a33946b0c77216d2cce91bb28c8fada4a5e80b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/tokenizer_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..061e40d9db3253624f86e8e364c15ef546527c9d
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/tokenizer_config.json
@@ -0,0 +1,2063 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/trainer_state.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..683a20286dac73f4a80befa8618856aa28edcb05
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/trainer_state.json
@@ -0,0 +1,17383 @@
+{
+  "best_metric": 1.0871200561523438,
+  "best_model_checkpoint": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185",
+  "epoch": 4.0,
+  "eval_steps": 10,
+  "global_step": 24742,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016166841807452913,
+      "grad_norm": 0.9894065856933594,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032333683614905826,
+      "grad_norm": 1.7810699939727783,
+      "learning_rate": 0.0002,
+      "loss": 1.1528,
+      "step": 20
+    },
+    {
+      "epoch": 0.004850052542235874,
+      "grad_norm": 0.5969577431678772,
+      "learning_rate": 0.0002,
+      "loss": 0.9767,
+      "step": 30
+    },
+    {
+      "epoch": 0.006466736722981165,
+      "grad_norm": 0.6354120969772339,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 40
+    },
+    {
+      "epoch": 0.008083420903726457,
+      "grad_norm": 0.5604607462882996,
+      "learning_rate": 0.0002,
+      "loss": 0.8643,
+      "step": 50
+    },
+    {
+      "epoch": 0.009700105084471748,
+      "grad_norm": 0.4676193594932556,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 60
+    },
+    {
+      "epoch": 0.01131678926521704,
+      "grad_norm": 0.6099211573600769,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 70
+    },
+    {
+      "epoch": 0.01293347344596233,
+      "grad_norm": 0.48639994859695435,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 80
+    },
+    {
+      "epoch": 0.014550157626707623,
+      "grad_norm": 0.4904264509677887,
+      "learning_rate": 0.0002,
+      "loss": 0.8704,
+      "step": 90
+    },
+    {
+      "epoch": 0.016166841807452915,
+      "grad_norm": 2.8334362506866455,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 100
+    },
+    {
+      "epoch": 0.017783525988198205,
+      "grad_norm": 0.43221670389175415,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 110
+    },
+    {
+      "epoch": 0.019400210168943496,
+      "grad_norm": 0.42244166135787964,
+      "learning_rate": 0.0002,
+      "loss": 0.8412,
+      "step": 120
+    },
+    {
+      "epoch": 0.02101689434968879,
+      "grad_norm": 0.45363298058509827,
+      "learning_rate": 0.0002,
+      "loss": 0.8467,
+      "step": 130
+    },
+    {
+      "epoch": 0.02263357853043408,
+      "grad_norm": 0.44816508889198303,
+      "learning_rate": 0.0002,
+      "loss": 0.8641,
+      "step": 140
+    },
+    {
+      "epoch": 0.02425026271117937,
+      "grad_norm": 0.43308213353157043,
+      "learning_rate": 0.0002,
+      "loss": 0.8496,
+      "step": 150
+    },
+    {
+      "epoch": 0.02586694689192466,
+      "grad_norm": 0.4084763526916504,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 160
+    },
+    {
+      "epoch": 0.027483631072669955,
+      "grad_norm": 0.5363703966140747,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 170
+    },
+    {
+      "epoch": 0.029100315253415245,
+      "grad_norm": 0.4619699716567993,
+      "learning_rate": 0.0002,
+      "loss": 0.8558,
+      "step": 180
+    },
+    {
+      "epoch": 0.030716999434160536,
+      "grad_norm": 0.49069908261299133,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 190
+    },
+    {
+      "epoch": 0.03233368361490583,
+      "grad_norm": 0.4645835757255554,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 200
+    },
+    {
+      "epoch": 0.03395036779565112,
+      "grad_norm": 1.2411243915557861,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 210
+    },
+    {
+      "epoch": 0.03556705197639641,
+      "grad_norm": 0.5211851596832275,
+      "learning_rate": 0.0002,
+      "loss": 0.8245,
+      "step": 220
+    },
+    {
+      "epoch": 0.037183736157141704,
+      "grad_norm": 0.5253691673278809,
+      "learning_rate": 0.0002,
+      "loss": 0.8194,
+      "step": 230
+    },
+    {
+      "epoch": 0.03880042033788699,
+      "grad_norm": 0.4567478895187378,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 240
+    },
+    {
+      "epoch": 0.040417104518632285,
+      "grad_norm": 0.5472128391265869,
+      "learning_rate": 0.0002,
+      "loss": 0.838,
+      "step": 250
+    },
+    {
+      "epoch": 0.04203378869937758,
+      "grad_norm": 0.42978546023368835,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 260
+    },
+    {
+      "epoch": 0.043650472880122866,
+      "grad_norm": 0.601734459400177,
+      "learning_rate": 0.0002,
+      "loss": 0.8334,
+      "step": 270
+    },
+    {
+      "epoch": 0.04526715706086816,
+      "grad_norm": 0.4286513328552246,
+      "learning_rate": 0.0002,
+      "loss": 0.815,
+      "step": 280
+    },
+    {
+      "epoch": 0.046883841241613454,
+      "grad_norm": 0.5230861902236938,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 290
+    },
+    {
+      "epoch": 0.04850052542235874,
+      "grad_norm": 0.6504611968994141,
+      "learning_rate": 0.0002,
+      "loss": 0.8636,
+      "step": 300
+    },
+    {
+      "epoch": 0.050117209603104035,
+      "grad_norm": 0.43485215306282043,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 310
+    },
+    {
+      "epoch": 0.05173389378384932,
+      "grad_norm": 0.4717007875442505,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 320
+    },
+    {
+      "epoch": 0.053350577964594616,
+      "grad_norm": 0.4059787690639496,
+      "learning_rate": 0.0002,
+      "loss": 0.8469,
+      "step": 330
+    },
+    {
+      "epoch": 0.05496726214533991,
+      "grad_norm": 0.4366913437843323,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 340
+    },
+    {
+      "epoch": 0.0565839463260852,
+      "grad_norm": 0.4233848452568054,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 350
+    },
+    {
+      "epoch": 0.05820063050683049,
+      "grad_norm": 0.4209108352661133,
+      "learning_rate": 0.0002,
+      "loss": 0.8456,
+      "step": 360
+    },
+    {
+      "epoch": 0.059817314687575784,
+      "grad_norm": 0.41637396812438965,
+      "learning_rate": 0.0002,
+      "loss": 0.816,
+      "step": 370
+    },
+    {
+      "epoch": 0.06143399886832107,
+      "grad_norm": 0.46235376596450806,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 380
+    },
+    {
+      "epoch": 0.06305068304906636,
+      "grad_norm": 0.4013484716415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 390
+    },
+    {
+      "epoch": 0.06466736722981166,
+      "grad_norm": 0.47443896532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 400
+    },
+    {
+      "epoch": 0.06628405141055695,
+      "grad_norm": 0.3942156434059143,
+      "learning_rate": 0.0002,
+      "loss": 0.8666,
+      "step": 410
+    },
+    {
+      "epoch": 0.06790073559130223,
+      "grad_norm": 0.4965320825576782,
+      "learning_rate": 0.0002,
+      "loss": 0.8402,
+      "step": 420
+    },
+    {
+      "epoch": 0.06951741977204753,
+      "grad_norm": 0.4304835796356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8317,
+      "step": 430
+    },
+    {
+      "epoch": 0.07113410395279282,
+      "grad_norm": 0.511726975440979,
+      "learning_rate": 0.0002,
+      "loss": 0.8528,
+      "step": 440
+    },
+    {
+      "epoch": 0.07275078813353811,
+      "grad_norm": 0.4040689170360565,
+      "learning_rate": 0.0002,
+      "loss": 0.8675,
+      "step": 450
+    },
+    {
+      "epoch": 0.07436747231428341,
+      "grad_norm": 0.5402171015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 460
+    },
+    {
+      "epoch": 0.0759841564950287,
+      "grad_norm": 0.4174517095088959,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 470
+    },
+    {
+      "epoch": 0.07760084067577398,
+      "grad_norm": 0.4306182265281677,
+      "learning_rate": 0.0002,
+      "loss": 0.7605,
+      "step": 480
+    },
+    {
+      "epoch": 0.07921752485651928,
+      "grad_norm": 0.535210132598877,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 490
+    },
+    {
+      "epoch": 0.08083420903726457,
+      "grad_norm": 0.5339109897613525,
+      "learning_rate": 0.0002,
+      "loss": 0.7825,
+      "step": 500
+    },
+    {
+      "epoch": 0.08245089321800986,
+      "grad_norm": 0.45754891633987427,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 510
+    },
+    {
+      "epoch": 0.08406757739875516,
+      "grad_norm": 0.43820783495903015,
+      "learning_rate": 0.0002,
+      "loss": 0.8144,
+      "step": 520
+    },
+    {
+      "epoch": 0.08568426157950045,
+      "grad_norm": 0.4434749186038971,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 530
+    },
+    {
+      "epoch": 0.08730094576024573,
+      "grad_norm": 0.43111467361450195,
+      "learning_rate": 0.0002,
+      "loss": 0.7857,
+      "step": 540
+    },
+    {
+      "epoch": 0.08891762994099103,
+      "grad_norm": 0.4378940165042877,
+      "learning_rate": 0.0002,
+      "loss": 0.8418,
+      "step": 550
+    },
+    {
+      "epoch": 0.09053431412173632,
+      "grad_norm": 0.4772215187549591,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 560
+    },
+    {
+      "epoch": 0.09215099830248161,
+      "grad_norm": 0.6837629079818726,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 570
+    },
+    {
+      "epoch": 0.09376768248322691,
+      "grad_norm": 0.42241212725639343,
+      "learning_rate": 0.0002,
+      "loss": 0.8607,
+      "step": 580
+    },
+    {
+      "epoch": 0.0953843666639722,
+      "grad_norm": 0.5165936350822449,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 590
+    },
+    {
+      "epoch": 0.09700105084471748,
+      "grad_norm": 0.48737478256225586,
+      "learning_rate": 0.0002,
+      "loss": 0.8664,
+      "step": 600
+    },
+    {
+      "epoch": 0.09861773502546278,
+      "grad_norm": 0.47419852018356323,
+      "learning_rate": 0.0002,
+      "loss": 0.8806,
+      "step": 610
+    },
+    {
+      "epoch": 0.10023441920620807,
+      "grad_norm": 0.4975486099720001,
+      "learning_rate": 0.0002,
+      "loss": 0.8254,
+      "step": 620
+    },
+    {
+      "epoch": 0.10185110338695336,
+      "grad_norm": 0.49123844504356384,
+      "learning_rate": 0.0002,
+      "loss": 0.8548,
+      "step": 630
+    },
+    {
+      "epoch": 0.10346778756769864,
+      "grad_norm": 0.6288952827453613,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 640
+    },
+    {
+      "epoch": 0.10508447174844394,
+      "grad_norm": 0.4277345836162567,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 650
+    },
+    {
+      "epoch": 0.10670115592918923,
+      "grad_norm": 0.4021061956882477,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 660
+    },
+    {
+      "epoch": 0.10831784010993452,
+      "grad_norm": 0.3492237329483032,
+      "learning_rate": 0.0002,
+      "loss": 0.87,
+      "step": 670
+    },
+    {
+      "epoch": 0.10993452429067982,
+      "grad_norm": 0.4341012239456177,
+      "learning_rate": 0.0002,
+      "loss": 0.8698,
+      "step": 680
+    },
+    {
+      "epoch": 0.1115512084714251,
+      "grad_norm": 0.7296304106712341,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 690
+    },
+    {
+      "epoch": 0.1131678926521704,
+      "grad_norm": 0.397494912147522,
+      "learning_rate": 0.0002,
+      "loss": 0.8433,
+      "step": 700
+    },
+    {
+      "epoch": 0.1147845768329157,
+      "grad_norm": 0.396431028842926,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 710
+    },
+    {
+      "epoch": 0.11640126101366098,
+      "grad_norm": 0.48842838406562805,
+      "learning_rate": 0.0002,
+      "loss": 0.8379,
+      "step": 720
+    },
+    {
+      "epoch": 0.11801794519440627,
+      "grad_norm": 0.46322616934776306,
+      "learning_rate": 0.0002,
+      "loss": 0.8238,
+      "step": 730
+    },
+    {
+      "epoch": 0.11963462937515157,
+      "grad_norm": 0.47990912199020386,
+      "learning_rate": 0.0002,
+      "loss": 0.8041,
+      "step": 740
+    },
+    {
+      "epoch": 0.12125131355589686,
+      "grad_norm": 0.4997142255306244,
+      "learning_rate": 0.0002,
+      "loss": 0.82,
+      "step": 750
+    },
+    {
+      "epoch": 0.12286799773664214,
+      "grad_norm": 0.4040526747703552,
+      "learning_rate": 0.0002,
+      "loss": 0.7702,
+      "step": 760
+    },
+    {
+      "epoch": 0.12448468191738744,
+      "grad_norm": 0.453095942735672,
+      "learning_rate": 0.0002,
+      "loss": 0.863,
+      "step": 770
+    },
+    {
+      "epoch": 0.12610136609813272,
+      "grad_norm": 0.4636971950531006,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 780
+    },
+    {
+      "epoch": 0.12771805027887803,
+      "grad_norm": 0.4279276132583618,
+      "learning_rate": 0.0002,
+      "loss": 0.8112,
+      "step": 790
+    },
+    {
+      "epoch": 0.12933473445962332,
+      "grad_norm": 0.46212655305862427,
+      "learning_rate": 0.0002,
+      "loss": 0.8711,
+      "step": 800
+    },
+    {
+      "epoch": 0.1309514186403686,
+      "grad_norm": 0.43127650022506714,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 810
+    },
+    {
+      "epoch": 0.1325681028211139,
+      "grad_norm": 0.4201301336288452,
+      "learning_rate": 0.0002,
+      "loss": 0.8476,
+      "step": 820
+    },
+    {
+      "epoch": 0.13418478700185918,
+      "grad_norm": 0.42583167552948,
+      "learning_rate": 0.0002,
+      "loss": 0.8078,
+      "step": 830
+    },
+    {
+      "epoch": 0.13580147118260447,
+      "grad_norm": 0.4535622000694275,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 840
+    },
+    {
+      "epoch": 0.13741815536334978,
+      "grad_norm": 0.4116036891937256,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 850
+    },
+    {
+      "epoch": 0.13903483954409507,
+      "grad_norm": 0.45997580885887146,
+      "learning_rate": 0.0002,
+      "loss": 0.8466,
+      "step": 860
+    },
+    {
+      "epoch": 0.14065152372484035,
+      "grad_norm": 0.4487837255001068,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 870
+    },
+    {
+      "epoch": 0.14226820790558564,
+      "grad_norm": 0.43650057911872864,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 880
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 0.5335358381271362,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 890
+    },
+    {
+      "epoch": 0.14550157626707622,
+      "grad_norm": 0.5989000201225281,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 900
+    },
+    {
+      "epoch": 0.14711826044782153,
+      "grad_norm": 0.517179012298584,
+      "learning_rate": 0.0002,
+      "loss": 0.8385,
+      "step": 910
+    },
+    {
+      "epoch": 0.14873494462856682,
+      "grad_norm": 0.44435232877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.8255,
+      "step": 920
+    },
+    {
+      "epoch": 0.1503516288093121,
+      "grad_norm": 0.42635923624038696,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 930
+    },
+    {
+      "epoch": 0.1519683129900574,
+      "grad_norm": 0.49603334069252014,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 940
+    },
+    {
+      "epoch": 0.15358499717080268,
+      "grad_norm": 0.40639808773994446,
+      "learning_rate": 0.0002,
+      "loss": 0.8377,
+      "step": 950
+    },
+    {
+      "epoch": 0.15520168135154797,
+      "grad_norm": 0.4850759208202362,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 960
+    },
+    {
+      "epoch": 0.15681836553229328,
+      "grad_norm": 0.4427442252635956,
+      "learning_rate": 0.0002,
+      "loss": 0.846,
+      "step": 970
+    },
+    {
+      "epoch": 0.15843504971303857,
+      "grad_norm": 0.3760930001735687,
+      "learning_rate": 0.0002,
+      "loss": 0.8705,
+      "step": 980
+    },
+    {
+      "epoch": 0.16005173389378385,
+      "grad_norm": 0.4794144332408905,
+      "learning_rate": 0.0002,
+      "loss": 0.8644,
+      "step": 990
+    },
+    {
+      "epoch": 0.16166841807452914,
+      "grad_norm": 0.45828768610954285,
+      "learning_rate": 0.0002,
+      "loss": 0.8002,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16328510225527443,
+      "grad_norm": 0.6313053369522095,
+      "learning_rate": 0.0002,
+      "loss": 0.7658,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16490178643601971,
+      "grad_norm": 0.45041006803512573,
+      "learning_rate": 0.0002,
+      "loss": 0.8047,
+      "step": 1020
+    },
+    {
+      "epoch": 0.166518470616765,
+      "grad_norm": 0.441403865814209,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16813515479751032,
+      "grad_norm": 0.8171296119689941,
+      "learning_rate": 0.0002,
+      "loss": 0.8475,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1697518389782556,
+      "grad_norm": 0.7137420773506165,
+      "learning_rate": 0.0002,
+      "loss": 0.845,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1713685231590009,
+      "grad_norm": 0.5236809849739075,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17298520733974618,
+      "grad_norm": 0.5021864175796509,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17460189152049146,
+      "grad_norm": 0.47347521781921387,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17621857570123675,
+      "grad_norm": 0.4631653428077698,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17783525988198207,
+      "grad_norm": 0.49169182777404785,
+      "learning_rate": 0.0002,
+      "loss": 0.8166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17945194406272735,
+      "grad_norm": 0.5019739270210266,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18106862824347264,
+      "grad_norm": 0.5100422501564026,
+      "learning_rate": 0.0002,
+      "loss": 0.8247,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18268531242421793,
+      "grad_norm": 0.3888324499130249,
+      "learning_rate": 0.0002,
+      "loss": 0.8142,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18430199660496321,
+      "grad_norm": 0.39765217900276184,
+      "learning_rate": 0.0002,
+      "loss": 0.8533,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1859186807857085,
+      "grad_norm": 0.47190186381340027,
+      "learning_rate": 0.0002,
+      "loss": 0.8541,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18753536496645382,
+      "grad_norm": 0.4464188814163208,
+      "learning_rate": 0.0002,
+      "loss": 0.8301,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1891520491471991,
+      "grad_norm": 0.5153930187225342,
+      "learning_rate": 0.0002,
+      "loss": 0.8341,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1907687333279444,
+      "grad_norm": 0.4779708683490753,
+      "learning_rate": 0.0002,
+      "loss": 0.8033,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19238541750868968,
+      "grad_norm": 0.4834315776824951,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19400210168943496,
+      "grad_norm": 0.402357816696167,
+      "learning_rate": 0.0002,
+      "loss": 0.7721,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19561878587018025,
+      "grad_norm": 0.45899084210395813,
+      "learning_rate": 0.0002,
+      "loss": 0.7941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19723547005092557,
+      "grad_norm": 0.5106529593467712,
+      "learning_rate": 0.0002,
+      "loss": 0.8353,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19885215423167085,
+      "grad_norm": 0.45261722803115845,
+      "learning_rate": 0.0002,
+      "loss": 0.7816,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20046883841241614,
+      "grad_norm": 0.4647127091884613,
+      "learning_rate": 0.0002,
+      "loss": 0.8068,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20208552259316143,
+      "grad_norm": 0.4849368929862976,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2037022067739067,
+      "grad_norm": 0.4518061578273773,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 1260
+    },
+    {
+      "epoch": 0.205318890954652,
+      "grad_norm": 0.49535325169563293,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2069355751353973,
+      "grad_norm": 0.4835205376148224,
+      "learning_rate": 0.0002,
+      "loss": 0.8348,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2085522593161426,
+      "grad_norm": 0.45308539271354675,
+      "learning_rate": 0.0002,
+      "loss": 0.8428,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2101689434968879,
+      "grad_norm": 0.5369905233383179,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21178562767763318,
+      "grad_norm": 0.5031622052192688,
+      "learning_rate": 0.0002,
+      "loss": 0.8676,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21340231185837846,
+      "grad_norm": 0.48010334372520447,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21501899603912375,
+      "grad_norm": 0.4905701279640198,
+      "learning_rate": 0.0002,
+      "loss": 0.806,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21663568021986904,
+      "grad_norm": 0.43531742691993713,
+      "learning_rate": 0.0002,
+      "loss": 0.7885,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21825236440061435,
+      "grad_norm": 0.44330692291259766,
+      "learning_rate": 0.0002,
+      "loss": 0.8191,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21986904858135964,
+      "grad_norm": 0.5384416580200195,
+      "learning_rate": 0.0002,
+      "loss": 0.8205,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22148573276210493,
+      "grad_norm": 0.4181833863258362,
+      "learning_rate": 0.0002,
+      "loss": 0.7726,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2231024169428502,
+      "grad_norm": 0.523833692073822,
+      "learning_rate": 0.0002,
+      "loss": 0.8311,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.5528736710548401,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2263357853043408,
+      "grad_norm": 0.43515023589134216,
+      "learning_rate": 0.0002,
+      "loss": 0.8079,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2279524694850861,
+      "grad_norm": 0.48809877038002014,
+      "learning_rate": 0.0002,
+      "loss": 0.8403,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2295691536658314,
+      "grad_norm": 0.43591251969337463,
+      "learning_rate": 0.0002,
+      "loss": 0.8165,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23118583784657668,
+      "grad_norm": 0.44625312089920044,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23280252202732196,
+      "grad_norm": 0.4390665292739868,
+      "learning_rate": 0.0002,
+      "loss": 0.8134,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23441920620806725,
+      "grad_norm": 0.48496049642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.8465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23603589038881254,
+      "grad_norm": 0.45919957756996155,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23765257456955785,
+      "grad_norm": 0.5471845865249634,
+      "learning_rate": 0.0002,
+      "loss": 0.8659,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23926925875030314,
+      "grad_norm": 0.47269317507743835,
+      "learning_rate": 0.0002,
+      "loss": 0.8164,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24088594293104842,
+      "grad_norm": 0.4930245578289032,
+      "learning_rate": 0.0002,
+      "loss": 0.854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2425026271117937,
+      "grad_norm": 0.5605630278587341,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 1500
+    },
+    {
+      "epoch": 0.244119311292539,
+      "grad_norm": 0.4435870945453644,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24573599547328429,
+      "grad_norm": 0.4941999912261963,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24735267965402957,
+      "grad_norm": 0.5100624561309814,
+      "learning_rate": 0.0002,
+      "loss": 0.8427,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2489693638347749,
+      "grad_norm": 0.4638267457485199,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25058604801552015,
+      "grad_norm": 0.5071570873260498,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25220273219626543,
+      "grad_norm": 0.4291319251060486,
+      "learning_rate": 0.0002,
+      "loss": 0.7724,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2538194163770108,
+      "grad_norm": 0.5388049483299255,
+      "learning_rate": 0.0002,
+      "loss": 0.7984,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25543610055775606,
+      "grad_norm": 0.5083683729171753,
+      "learning_rate": 0.0002,
+      "loss": 0.8176,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25705278473850135,
+      "grad_norm": 0.4824463725090027,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25866946891924664,
+      "grad_norm": 0.41177722811698914,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2602861530999919,
+      "grad_norm": 0.5656219124794006,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2619028372807372,
+      "grad_norm": 0.41063204407691956,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2635195214614825,
+      "grad_norm": 0.4897061288356781,
+      "learning_rate": 0.0002,
+      "loss": 0.7998,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2651362056422278,
+      "grad_norm": 0.4454376697540283,
+      "learning_rate": 0.0002,
+      "loss": 0.8198,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26675288982297307,
+      "grad_norm": 0.4355238378047943,
+      "learning_rate": 0.0002,
+      "loss": 0.8684,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26836957400371836,
+      "grad_norm": 0.458310067653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26998625818446365,
+      "grad_norm": 0.4752083718776703,
+      "learning_rate": 0.0002,
+      "loss": 0.7935,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27160294236520893,
+      "grad_norm": 0.4666106402873993,
+      "learning_rate": 0.0002,
+      "loss": 0.8267,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2732196265459543,
+      "grad_norm": 0.4213818609714508,
+      "learning_rate": 0.0002,
+      "loss": 0.8252,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27483631072669956,
+      "grad_norm": 0.5768913626670837,
+      "learning_rate": 0.0002,
+      "loss": 0.8559,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27645299490744485,
+      "grad_norm": 0.4209914803504944,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27806967908819014,
+      "grad_norm": 0.501909613609314,
+      "learning_rate": 0.0002,
+      "loss": 0.8167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2796863632689354,
+      "grad_norm": 0.5266261100769043,
+      "learning_rate": 0.0002,
+      "loss": 0.7832,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2813030474496807,
+      "grad_norm": 0.43806859850883484,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 1740
+    },
+    {
+      "epoch": 0.282919731630426,
+      "grad_norm": 0.46048814058303833,
+      "learning_rate": 0.0002,
+      "loss": 0.8157,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2845364158111713,
+      "grad_norm": 0.44972819089889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8596,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28615309999191657,
+      "grad_norm": 0.5114831328392029,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.47931742668151855,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28938646835340714,
+      "grad_norm": 0.5092599987983704,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29100315253415243,
+      "grad_norm": 0.37581443786621094,
+      "learning_rate": 0.0002,
+      "loss": 0.8506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2926198367148977,
+      "grad_norm": 0.47097381949424744,
+      "learning_rate": 0.0002,
+      "loss": 0.7932,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29423652089564306,
+      "grad_norm": 0.48300236463546753,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29585320507638835,
+      "grad_norm": 0.5600419640541077,
+      "learning_rate": 0.0002,
+      "loss": 0.8391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29746988925713364,
+      "grad_norm": 0.48555272817611694,
+      "learning_rate": 0.0002,
+      "loss": 0.8507,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2990865734378789,
+      "grad_norm": 0.3752668499946594,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3007032576186242,
+      "grad_norm": 0.5328747034072876,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3023199417993695,
+      "grad_norm": 0.48716455698013306,
+      "learning_rate": 0.0002,
+      "loss": 0.8426,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3039366259801148,
+      "grad_norm": 0.5011493563652039,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30555331016086007,
+      "grad_norm": 0.46461427211761475,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30716999434160536,
+      "grad_norm": 0.36630210280418396,
+      "learning_rate": 0.0002,
+      "loss": 0.8478,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30878667852235064,
+      "grad_norm": 0.4217296242713928,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31040336270309593,
+      "grad_norm": 0.4394875466823578,
+      "learning_rate": 0.0002,
+      "loss": 0.8128,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3120200468838412,
+      "grad_norm": 0.6587965488433838,
+      "learning_rate": 0.0002,
+      "loss": 0.8471,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31363673106458656,
+      "grad_norm": 0.5469298958778381,
+      "learning_rate": 0.0002,
+      "loss": 0.8565,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31525341524533185,
+      "grad_norm": 0.4371595084667206,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31687009942607713,
+      "grad_norm": 0.4809541404247284,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3184867836068224,
+      "grad_norm": 0.6061086654663086,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3201034677875677,
+      "grad_norm": 0.5342657566070557,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 1980
+    },
+    {
+      "epoch": 0.321720151968313,
+      "grad_norm": 0.5057743787765503,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3233368361490583,
+      "grad_norm": 0.528626024723053,
+      "learning_rate": 0.0002,
+      "loss": 0.7774,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32495352032980357,
+      "grad_norm": 0.46742770075798035,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32657020451054886,
+      "grad_norm": 0.515101432800293,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32818688869129414,
+      "grad_norm": 0.41941216588020325,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32980357287203943,
+      "grad_norm": 0.49902522563934326,
+      "learning_rate": 0.0002,
+      "loss": 0.7637,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3314202570527847,
+      "grad_norm": 0.4120897650718689,
+      "learning_rate": 0.0002,
+      "loss": 0.7822,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33303694123353,
+      "grad_norm": 0.45352041721343994,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33465362541427535,
+      "grad_norm": 0.523199737071991,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33627030959502063,
+      "grad_norm": 0.4390358626842499,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3378869937757659,
+      "grad_norm": 0.6752901077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3395036779565112,
+      "grad_norm": 0.547821044921875,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3411203621372565,
+      "grad_norm": 0.5161308646202087,
+      "learning_rate": 0.0002,
+      "loss": 0.8561,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3427370463180018,
+      "grad_norm": 0.4565401077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34435373049874707,
+      "grad_norm": 0.4666115939617157,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34597041467949236,
+      "grad_norm": 0.4090428352355957,
+      "learning_rate": 0.0002,
+      "loss": 0.8189,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34758709886023764,
+      "grad_norm": 0.510845422744751,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34920378304098293,
+      "grad_norm": 0.42861923575401306,
+      "learning_rate": 0.0002,
+      "loss": 0.8398,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3508204672217282,
+      "grad_norm": 0.4476332664489746,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3524371514024735,
+      "grad_norm": 0.6065791249275208,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35405383558321885,
+      "grad_norm": 0.42335066199302673,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35567051976396413,
+      "grad_norm": 0.5094629526138306,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3572872039447094,
+      "grad_norm": 0.5476373434066772,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3589038881254547,
+      "grad_norm": 0.3911719024181366,
+      "learning_rate": 0.0002,
+      "loss": 0.814,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3605205723062,
+      "grad_norm": 0.6599636077880859,
+      "learning_rate": 0.0002,
+      "loss": 0.8599,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3621372564869453,
+      "grad_norm": 0.40381914377212524,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36375394066769057,
+      "grad_norm": 0.4433908462524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36537062484843585,
+      "grad_norm": 0.578326940536499,
+      "learning_rate": 0.0002,
+      "loss": 0.8503,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36698730902918114,
+      "grad_norm": 0.5734784007072449,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36860399320992643,
+      "grad_norm": 0.45555487275123596,
+      "learning_rate": 0.0002,
+      "loss": 0.8193,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3702206773906717,
+      "grad_norm": 0.5666276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.7929,
+      "step": 2290
+    },
+    {
+      "epoch": 0.371837361571417,
+      "grad_norm": 0.5461117625236511,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3734540457521623,
+      "grad_norm": 0.6318911910057068,
+      "learning_rate": 0.0002,
+      "loss": 0.8204,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37507072993290763,
+      "grad_norm": 0.493263304233551,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3766874141136529,
+      "grad_norm": 0.5888760089874268,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3783040982943982,
+      "grad_norm": 0.48671841621398926,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3799207824751435,
+      "grad_norm": 0.4385145306587219,
+      "learning_rate": 0.0002,
+      "loss": 0.8367,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3815374666558888,
+      "grad_norm": 0.5523318648338318,
+      "learning_rate": 0.0002,
+      "loss": 0.812,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38315415083663407,
+      "grad_norm": 0.7308220267295837,
+      "learning_rate": 0.0002,
+      "loss": 0.8351,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38477083501737935,
+      "grad_norm": 0.554214358329773,
+      "learning_rate": 0.0002,
+      "loss": 0.859,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38638751919812464,
+      "grad_norm": 0.5425800085067749,
+      "learning_rate": 0.0002,
+      "loss": 0.8146,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3880042033788699,
+      "grad_norm": 0.48811158537864685,
+      "learning_rate": 0.0002,
+      "loss": 0.8282,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3896208875596152,
+      "grad_norm": 0.49212366342544556,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3912375717403605,
+      "grad_norm": 0.5222218632698059,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3928542559211058,
+      "grad_norm": 0.4699819087982178,
+      "learning_rate": 0.0002,
+      "loss": 0.8182,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39447094010185113,
+      "grad_norm": 0.46153587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7919,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3960876242825964,
+      "grad_norm": 0.4150611162185669,
+      "learning_rate": 0.0002,
+      "loss": 0.8111,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3977043084633417,
+      "grad_norm": 0.5799614787101746,
+      "learning_rate": 0.0002,
+      "loss": 0.8589,
+      "step": 2460
+    },
+    {
+      "epoch": 0.399320992644087,
+      "grad_norm": 0.56536865234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4009376768248323,
+      "grad_norm": 0.5451247096061707,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40255436100557757,
+      "grad_norm": 0.5914521217346191,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40417104518632285,
+      "grad_norm": 0.4428117275238037,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40578772936706814,
+      "grad_norm": 0.48580947518348694,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4074044135478134,
+      "grad_norm": 0.436734676361084,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4090210977285587,
+      "grad_norm": 0.5752223134040833,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 2530
+    },
+    {
+      "epoch": 0.410637781909304,
+      "grad_norm": 0.4271308183670044,
+      "learning_rate": 0.0002,
+      "loss": 0.8181,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4122544660900493,
+      "grad_norm": 0.46294718980789185,
+      "learning_rate": 0.0002,
+      "loss": 0.8058,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4138711502707946,
+      "grad_norm": 0.49407583475112915,
+      "learning_rate": 0.0002,
+      "loss": 0.8473,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4154878344515399,
+      "grad_norm": 0.4729035496711731,
+      "learning_rate": 0.0002,
+      "loss": 0.7881,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4171045186322852,
+      "grad_norm": 0.4129747152328491,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4187212028130305,
+      "grad_norm": 0.5684236288070679,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4203378869937758,
+      "grad_norm": 0.4862157106399536,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42195457117452106,
+      "grad_norm": 0.46567976474761963,
+      "learning_rate": 0.0002,
+      "loss": 0.7582,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42357125535526635,
+      "grad_norm": 0.5710650682449341,
+      "learning_rate": 0.0002,
+      "loss": 0.7755,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42518793953601164,
+      "grad_norm": 0.5660041570663452,
+      "learning_rate": 0.0002,
+      "loss": 0.8573,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4268046237167569,
+      "grad_norm": 0.47944375872612,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4284213078975022,
+      "grad_norm": 0.537223756313324,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4300379920782475,
+      "grad_norm": 0.41669997572898865,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 0.44727686047554016,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4332713604397381,
+      "grad_norm": 0.5600888729095459,
+      "learning_rate": 0.0002,
+      "loss": 0.8241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4348880446204834,
+      "grad_norm": 0.39820605516433716,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4365047288012287,
+      "grad_norm": 0.5637655854225159,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 2700
+    },
+    {
+      "epoch": 0.438121412981974,
+      "grad_norm": 0.6363666653633118,
+      "learning_rate": 0.0002,
+      "loss": 0.855,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4397380971627193,
+      "grad_norm": 0.5656129121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.8468,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44135478134346456,
+      "grad_norm": 0.5600156188011169,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44297146552420985,
+      "grad_norm": 0.5506579875946045,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44458814970495514,
+      "grad_norm": 0.49878305196762085,
+      "learning_rate": 0.0002,
+      "loss": 0.7725,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4462048338857004,
+      "grad_norm": 0.4569213092327118,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4478215180664457,
+      "grad_norm": 0.6056680083274841,
+      "learning_rate": 0.0002,
+      "loss": 0.8028,
+      "step": 2770
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.44474557042121887,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4510548864279363,
+      "grad_norm": 0.46055394411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.801,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4526715706086816,
+      "grad_norm": 0.4904133379459381,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45428825478942686,
+      "grad_norm": 0.5647031664848328,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4559049389701722,
+      "grad_norm": 0.5759473443031311,
+      "learning_rate": 0.0002,
+      "loss": 0.8622,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4575216231509175,
+      "grad_norm": 0.5161895751953125,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4591383073316628,
+      "grad_norm": 0.4248254597187042,
+      "learning_rate": 0.0002,
+      "loss": 0.8045,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46075499151240806,
+      "grad_norm": 0.45395001769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.7838,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46237167569315335,
+      "grad_norm": 0.5358697772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.8208,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46398835987389864,
+      "grad_norm": 0.5379165410995483,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4656050440546439,
+      "grad_norm": 0.4601989686489105,
+      "learning_rate": 0.0002,
+      "loss": 0.7403,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4672217282353892,
+      "grad_norm": 0.671115517616272,
+      "learning_rate": 0.0002,
+      "loss": 0.8523,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4688384124161345,
+      "grad_norm": 0.4425133168697357,
+      "learning_rate": 0.0002,
+      "loss": 0.8262,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4704550965968798,
+      "grad_norm": 0.5446155071258545,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47207178077762507,
+      "grad_norm": 0.603306233882904,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47368846495837036,
+      "grad_norm": 0.5377997159957886,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4753051491391157,
+      "grad_norm": 0.4931027591228485,
+      "learning_rate": 0.0002,
+      "loss": 0.8075,
+      "step": 2940
+    },
+    {
+      "epoch": 0.476921833319861,
+      "grad_norm": 0.4711960256099701,
+      "learning_rate": 0.0002,
+      "loss": 0.8004,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4785385175006063,
+      "grad_norm": 0.5020492672920227,
+      "learning_rate": 0.0002,
+      "loss": 0.8121,
+      "step": 2960
+    },
+    {
+      "epoch": 0.48015520168135156,
+      "grad_norm": 0.5428946614265442,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48177188586209685,
+      "grad_norm": 0.5294089317321777,
+      "learning_rate": 0.0002,
+      "loss": 0.7849,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48338857004284214,
+      "grad_norm": 0.648289144039154,
+      "learning_rate": 0.0002,
+      "loss": 0.8553,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4850052542235874,
+      "grad_norm": 0.47916680574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.7874,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4866219384043327,
+      "grad_norm": 0.43849772214889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8087,
+      "step": 3010
+    },
+    {
+      "epoch": 0.488238622585078,
+      "grad_norm": 0.47007861733436584,
+      "learning_rate": 0.0002,
+      "loss": 0.7662,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4898553067658233,
+      "grad_norm": 0.6314331293106079,
+      "learning_rate": 0.0002,
+      "loss": 0.757,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49147199094656857,
+      "grad_norm": 0.49211493134498596,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49308867512731386,
+      "grad_norm": 0.4537973403930664,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49470535930805914,
+      "grad_norm": 0.47326919436454773,
+      "learning_rate": 0.0002,
+      "loss": 0.8095,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4963220434888045,
+      "grad_norm": 0.525874137878418,
+      "learning_rate": 0.0002,
+      "loss": 0.8447,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4979387276695498,
+      "grad_norm": 0.6361091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49955541185029506,
+      "grad_norm": 0.5850642919540405,
+      "learning_rate": 0.0002,
+      "loss": 0.821,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5011720960310403,
+      "grad_norm": 0.47299543023109436,
+      "learning_rate": 0.0002,
+      "loss": 0.8279,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5027887802117856,
+      "grad_norm": 0.473099946975708,
+      "learning_rate": 0.0002,
+      "loss": 0.8681,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5044054643925309,
+      "grad_norm": 0.48186397552490234,
+      "learning_rate": 0.0002,
+      "loss": 0.8223,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5060221485732762,
+      "grad_norm": 0.5015401840209961,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5076388327540216,
+      "grad_norm": 0.5617750287055969,
+      "learning_rate": 0.0002,
+      "loss": 0.7692,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5092555169347668,
+      "grad_norm": 0.5169327259063721,
+      "learning_rate": 0.0002,
+      "loss": 0.8708,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5108722011155121,
+      "grad_norm": 0.545657753944397,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5124888852962574,
+      "grad_norm": 0.512864351272583,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5141055694770027,
+      "grad_norm": 0.4113546311855316,
+      "learning_rate": 0.0002,
+      "loss": 0.7794,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5157222536577479,
+      "grad_norm": 0.44532445073127747,
+      "learning_rate": 0.0002,
+      "loss": 0.8206,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5173389378384933,
+      "grad_norm": 0.5623497366905212,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5189556220192385,
+      "grad_norm": 0.5084741115570068,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5205723061999838,
+      "grad_norm": 0.5305403470993042,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5221889903807291,
+      "grad_norm": 0.4708254337310791,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5238056745614744,
+      "grad_norm": 0.43827131390571594,
+      "learning_rate": 0.0002,
+      "loss": 0.7639,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5254223587422197,
+      "grad_norm": 0.5630002617835999,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 3250
+    },
+    {
+      "epoch": 0.527039042922965,
+      "grad_norm": 0.5010961890220642,
+      "learning_rate": 0.0002,
+      "loss": 0.7522,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5286557271037103,
+      "grad_norm": 0.6303122043609619,
+      "learning_rate": 0.0002,
+      "loss": 0.8374,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5302724112844556,
+      "grad_norm": 0.5107331275939941,
+      "learning_rate": 0.0002,
+      "loss": 0.7727,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5318890954652009,
+      "grad_norm": 0.5700443387031555,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5335057796459461,
+      "grad_norm": 0.46296367049217224,
+      "learning_rate": 0.0002,
+      "loss": 0.7776,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5351224638266915,
+      "grad_norm": 0.531568706035614,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5367391480074367,
+      "grad_norm": 0.4686741530895233,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5383558321881821,
+      "grad_norm": 0.5404331088066101,
+      "learning_rate": 0.0002,
+      "loss": 0.8104,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5399725163689273,
+      "grad_norm": 0.6368790864944458,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5415892005496726,
+      "grad_norm": 0.42300888895988464,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5432058847304179,
+      "grad_norm": 0.5362542867660522,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5448225689111632,
+      "grad_norm": 0.497128963470459,
+      "learning_rate": 0.0002,
+      "loss": 0.858,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5464392530919085,
+      "grad_norm": 0.5006386041641235,
+      "learning_rate": 0.0002,
+      "loss": 0.8519,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5480559372726538,
+      "grad_norm": 0.44136837124824524,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5496726214533991,
+      "grad_norm": 0.5897833108901978,
+      "learning_rate": 0.0002,
+      "loss": 0.773,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5512893056341444,
+      "grad_norm": 0.641075611114502,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5529059898148897,
+      "grad_norm": 0.7251322269439697,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5545226739956349,
+      "grad_norm": 0.47411349415779114,
+      "learning_rate": 0.0002,
+      "loss": 0.7626,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5561393581763803,
+      "grad_norm": 0.4994310438632965,
+      "learning_rate": 0.0002,
+      "loss": 0.8196,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5577560423571255,
+      "grad_norm": 0.5814438462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5593727265378708,
+      "grad_norm": 0.6278898119926453,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5609894107186161,
+      "grad_norm": 0.46208274364471436,
+      "learning_rate": 0.0002,
+      "loss": 0.813,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5626060948993614,
+      "grad_norm": 0.5718930959701538,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5642227790801067,
+      "grad_norm": 0.48178744316101074,
+      "learning_rate": 0.0002,
+      "loss": 0.8152,
+      "step": 3490
+    },
+    {
+      "epoch": 0.565839463260852,
+      "grad_norm": 0.47336965799331665,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5674561474415973,
+      "grad_norm": 0.43442684412002563,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5690728316223426,
+      "grad_norm": 0.6463358998298645,
+      "learning_rate": 0.0002,
+      "loss": 0.7564,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5706895158030879,
+      "grad_norm": 0.5286486744880676,
+      "learning_rate": 0.0002,
+      "loss": 0.836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5723061999838331,
+      "grad_norm": 0.5405499935150146,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5739228841645785,
+      "grad_norm": 0.6654391884803772,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.5081980228424072,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5771562525260691,
+      "grad_norm": 0.48978179693222046,
+      "learning_rate": 0.0002,
+      "loss": 0.7753,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5787729367068143,
+      "grad_norm": 0.5840612053871155,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5803896208875596,
+      "grad_norm": 0.5235261917114258,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5820063050683049,
+      "grad_norm": 0.5672075748443604,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5836229892490502,
+      "grad_norm": 0.5613429546356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5852396734297954,
+      "grad_norm": 0.4032273590564728,
+      "learning_rate": 0.0002,
+      "loss": 0.8274,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5868563576105408,
+      "grad_norm": 0.49559324979782104,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5884730417912861,
+      "grad_norm": 0.6895697712898254,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5900897259720314,
+      "grad_norm": 0.4750136435031891,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5917064101527767,
+      "grad_norm": 0.5176819562911987,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5933230943335219,
+      "grad_norm": 0.5817760229110718,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5949397785142673,
+      "grad_norm": 0.6064626574516296,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5965564626950125,
+      "grad_norm": 0.6728700995445251,
+      "learning_rate": 0.0002,
+      "loss": 0.8422,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5981731468757578,
+      "grad_norm": 0.609305202960968,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5997898310565031,
+      "grad_norm": 0.4615488350391388,
+      "learning_rate": 0.0002,
+      "loss": 0.8048,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6014065152372484,
+      "grad_norm": 2.0531179904937744,
+      "learning_rate": 0.0002,
+      "loss": 0.8214,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6030231994179936,
+      "grad_norm": 0.5091132521629333,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 3730
+    },
+    {
+      "epoch": 0.604639883598739,
+      "grad_norm": 0.5951124429702759,
+      "learning_rate": 0.0002,
+      "loss": 0.7833,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6062565677794842,
+      "grad_norm": 0.5870208144187927,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6078732519602296,
+      "grad_norm": 0.6254619359970093,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6094899361409749,
+      "grad_norm": 0.5577626824378967,
+      "learning_rate": 0.0002,
+      "loss": 0.7868,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6111066203217201,
+      "grad_norm": 0.5004405379295349,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6127233045024655,
+      "grad_norm": 0.5527383685112,
+      "learning_rate": 0.0002,
+      "loss": 0.8092,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6143399886832107,
+      "grad_norm": 0.49116113781929016,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6159566728639561,
+      "grad_norm": 0.5299299359321594,
+      "learning_rate": 0.0002,
+      "loss": 0.8352,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6175733570447013,
+      "grad_norm": 0.464897483587265,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6191900412254466,
+      "grad_norm": 0.6505740880966187,
+      "learning_rate": 0.0002,
+      "loss": 0.7923,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6208067254061919,
+      "grad_norm": 0.5512559413909912,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6224234095869372,
+      "grad_norm": 0.49427518248558044,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6240400937676824,
+      "grad_norm": 0.3839147090911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6256567779484278,
+      "grad_norm": 0.5760218501091003,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6272734621291731,
+      "grad_norm": 0.7226507067680359,
+      "learning_rate": 0.0002,
+      "loss": 0.7836,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6288901463099184,
+      "grad_norm": 0.676781415939331,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6305068304906637,
+      "grad_norm": 0.4284018278121948,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6321235146714089,
+      "grad_norm": 0.5060628056526184,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6337401988521543,
+      "grad_norm": 0.5524522066116333,
+      "learning_rate": 0.0002,
+      "loss": 0.8089,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6353568830328995,
+      "grad_norm": 0.6099881529808044,
+      "learning_rate": 0.0002,
+      "loss": 0.8276,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6369735672136448,
+      "grad_norm": 0.43155938386917114,
+      "learning_rate": 0.0002,
+      "loss": 0.809,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6385902513943901,
+      "grad_norm": 0.6427084803581238,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6402069355751354,
+      "grad_norm": 0.541220486164093,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6418236197558806,
+      "grad_norm": 0.5414294600486755,
+      "learning_rate": 0.0002,
+      "loss": 0.8539,
+      "step": 3970
+    },
+    {
+      "epoch": 0.643440303936626,
+      "grad_norm": 0.46344003081321716,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6450569881173712,
+      "grad_norm": 0.45209285616874695,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6466736722981166,
+      "grad_norm": 0.5417284369468689,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6482903564788619,
+      "grad_norm": 0.7995685935020447,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6499070406596071,
+      "grad_norm": 0.6384002566337585,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6515237248403525,
+      "grad_norm": 0.4472815692424774,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6531404090210977,
+      "grad_norm": 0.6834294199943542,
+      "learning_rate": 0.0002,
+      "loss": 0.8071,
+      "step": 4040
+    },
+    {
+      "epoch": 0.654757093201843,
+      "grad_norm": 0.4612339735031128,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6563737773825883,
+      "grad_norm": 0.9266576170921326,
+      "learning_rate": 0.0002,
+      "loss": 0.8141,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6579904615633336,
+      "grad_norm": 0.4470861852169037,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6596071457440789,
+      "grad_norm": 0.45544925332069397,
+      "learning_rate": 0.0002,
+      "loss": 0.8293,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6612238299248242,
+      "grad_norm": 0.6144481301307678,
+      "learning_rate": 0.0002,
+      "loss": 0.8455,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6628405141055694,
+      "grad_norm": 0.5936288237571716,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6644571982863148,
+      "grad_norm": 0.4822963774204254,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 4110
+    },
+    {
+      "epoch": 0.66607388246706,
+      "grad_norm": 0.48432496190071106,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6676905666478054,
+      "grad_norm": 0.4901607930660248,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6693072508285507,
+      "grad_norm": 0.5018393397331238,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6709239350092959,
+      "grad_norm": 0.6946378946304321,
+      "learning_rate": 0.0002,
+      "loss": 0.8065,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6725406191900413,
+      "grad_norm": 0.5997390747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.6738849878311157,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6757739875515318,
+      "grad_norm": 0.6110581159591675,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6773906717322771,
+      "grad_norm": 0.5703322291374207,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6790073559130224,
+      "grad_norm": 0.4686066210269928,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6806240400937676,
+      "grad_norm": 0.6394643783569336,
+      "learning_rate": 0.0002,
+      "loss": 0.8441,
+      "step": 4210
+    },
+    {
+      "epoch": 0.682240724274513,
+      "grad_norm": 0.5454841256141663,
+      "learning_rate": 0.0002,
+      "loss": 0.8011,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6838574084552582,
+      "grad_norm": 0.4859732985496521,
+      "learning_rate": 0.0002,
+      "loss": 0.8307,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6854740926360036,
+      "grad_norm": 0.5544065833091736,
+      "learning_rate": 0.0002,
+      "loss": 0.8161,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6870907768167488,
+      "grad_norm": 0.4902505576610565,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6887074609974941,
+      "grad_norm": 0.4768051505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7977,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6903241451782395,
+      "grad_norm": 0.49982190132141113,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6919408293589847,
+      "grad_norm": 0.6351838111877441,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 4280
+    },
+    {
+      "epoch": 0.69355751353973,
+      "grad_norm": 0.5647561550140381,
+      "learning_rate": 0.0002,
+      "loss": 0.7664,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6951741977204753,
+      "grad_norm": 0.5340486764907837,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6967908819012206,
+      "grad_norm": 0.5649092793464661,
+      "learning_rate": 0.0002,
+      "loss": 0.8526,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6984075660819659,
+      "grad_norm": 0.6183916926383972,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7000242502627112,
+      "grad_norm": 0.6154509782791138,
+      "learning_rate": 0.0002,
+      "loss": 0.792,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7016409344434564,
+      "grad_norm": 0.5156264305114746,
+      "learning_rate": 0.0002,
+      "loss": 0.8397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7032576186242018,
+      "grad_norm": 0.562171459197998,
+      "learning_rate": 0.0002,
+      "loss": 0.8512,
+      "step": 4350
+    },
+    {
+      "epoch": 0.704874302804947,
+      "grad_norm": 0.4949502646923065,
+      "learning_rate": 0.0002,
+      "loss": 0.7882,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7064909869856923,
+      "grad_norm": 0.5171684622764587,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7081076711664377,
+      "grad_norm": 0.6198443174362183,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7097243553471829,
+      "grad_norm": 0.5802276134490967,
+      "learning_rate": 0.0002,
+      "loss": 0.7606,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7113410395279283,
+      "grad_norm": 0.41096967458724976,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7129577237086735,
+      "grad_norm": 0.4397392272949219,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7145744078894188,
+      "grad_norm": 0.45228442549705505,
+      "learning_rate": 0.0002,
+      "loss": 0.7651,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7161910920701641,
+      "grad_norm": 0.4839673936367035,
+      "learning_rate": 0.0002,
+      "loss": 0.7938,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7178077762509094,
+      "grad_norm": 0.6140755414962769,
+      "learning_rate": 0.0002,
+      "loss": 0.8362,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 0.6841378808021545,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7210411446124,
+      "grad_norm": 0.6664239168167114,
+      "learning_rate": 0.0002,
+      "loss": 0.8177,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7226578287931452,
+      "grad_norm": 0.47552719712257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7983,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7242745129738906,
+      "grad_norm": 0.6649776101112366,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7258911971546358,
+      "grad_norm": 0.5159541964530945,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7275078813353811,
+      "grad_norm": 0.6693112850189209,
+      "learning_rate": 0.0002,
+      "loss": 0.7786,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7291245655161265,
+      "grad_norm": 0.48870977759361267,
+      "learning_rate": 0.0002,
+      "loss": 0.8655,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7307412496968717,
+      "grad_norm": 0.4857887923717499,
+      "learning_rate": 0.0002,
+      "loss": 0.7337,
+      "step": 4520
+    },
+    {
+      "epoch": 0.732357933877617,
+      "grad_norm": 0.5515662431716919,
+      "learning_rate": 0.0002,
+      "loss": 0.8026,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7339746180583623,
+      "grad_norm": 0.6292222738265991,
+      "learning_rate": 0.0002,
+      "loss": 0.8031,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7355913022391076,
+      "grad_norm": 0.48265689611434937,
+      "learning_rate": 0.0002,
+      "loss": 0.7749,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7372079864198529,
+      "grad_norm": 0.8044266104698181,
+      "learning_rate": 0.0002,
+      "loss": 0.8499,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7388246706005982,
+      "grad_norm": 0.6111769676208496,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7404413547813434,
+      "grad_norm": 0.5229553580284119,
+      "learning_rate": 0.0002,
+      "loss": 0.7291,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7420580389620888,
+      "grad_norm": 0.6054152250289917,
+      "learning_rate": 0.0002,
+      "loss": 0.8038,
+      "step": 4590
+    },
+    {
+      "epoch": 0.743674723142834,
+      "grad_norm": 0.5574966669082642,
+      "learning_rate": 0.0002,
+      "loss": 0.8169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7452914073235793,
+      "grad_norm": 0.5395817160606384,
+      "learning_rate": 0.0002,
+      "loss": 0.8439,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7469080915043246,
+      "grad_norm": 0.7116472721099854,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7485247756850699,
+      "grad_norm": 0.5618700981140137,
+      "learning_rate": 0.0002,
+      "loss": 0.7743,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7501414598658153,
+      "grad_norm": 0.5802770853042603,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7517581440465605,
+      "grad_norm": 0.5690428018569946,
+      "learning_rate": 0.0002,
+      "loss": 0.7924,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7533748282273058,
+      "grad_norm": 0.4813360273838043,
+      "learning_rate": 0.0002,
+      "loss": 0.8017,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7549915124080511,
+      "grad_norm": 0.5434042811393738,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7566081965887964,
+      "grad_norm": 0.5502099990844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7582248807695416,
+      "grad_norm": 0.6020621061325073,
+      "learning_rate": 0.0002,
+      "loss": 0.8598,
+      "step": 4690
+    },
+    {
+      "epoch": 0.759841564950287,
+      "grad_norm": 0.4922301471233368,
+      "learning_rate": 0.0002,
+      "loss": 0.7937,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7614582491310322,
+      "grad_norm": 0.6492828726768494,
+      "learning_rate": 0.0002,
+      "loss": 0.788,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7630749333117776,
+      "grad_norm": 0.4865580201148987,
+      "learning_rate": 0.0002,
+      "loss": 0.8313,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7646916174925228,
+      "grad_norm": 0.5971422791481018,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7663083016732681,
+      "grad_norm": 0.6832674145698547,
+      "learning_rate": 0.0002,
+      "loss": 0.8298,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7679249858540134,
+      "grad_norm": 0.500908613204956,
+      "learning_rate": 0.0002,
+      "loss": 0.8156,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7695416700347587,
+      "grad_norm": 0.6112465858459473,
+      "learning_rate": 0.0002,
+      "loss": 0.8383,
+      "step": 4760
+    },
+    {
+      "epoch": 0.771158354215504,
+      "grad_norm": 0.5753506422042847,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7727750383962493,
+      "grad_norm": 0.6529405117034912,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7743917225769946,
+      "grad_norm": 0.5916843414306641,
+      "learning_rate": 0.0002,
+      "loss": 0.8171,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7760084067577399,
+      "grad_norm": 0.4821224510669708,
+      "learning_rate": 0.0002,
+      "loss": 0.83,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7776250909384852,
+      "grad_norm": 0.5532580018043518,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7792417751192304,
+      "grad_norm": 0.4604877233505249,
+      "learning_rate": 0.0002,
+      "loss": 0.7363,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7808584592999758,
+      "grad_norm": 0.5009613037109375,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 4830
+    },
+    {
+      "epoch": 0.782475143480721,
+      "grad_norm": 0.6448560357093811,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7840918276614663,
+      "grad_norm": 0.44327953457832336,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7857085118422116,
+      "grad_norm": 0.5355411171913147,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7873251960229569,
+      "grad_norm": 0.5635677576065063,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7889418802037023,
+      "grad_norm": 0.5417491793632507,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7905585643844475,
+      "grad_norm": 0.4567430913448334,
+      "learning_rate": 0.0002,
+      "loss": 0.7819,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7921752485651928,
+      "grad_norm": 0.44651296734809875,
+      "learning_rate": 0.0002,
+      "loss": 0.8454,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7937919327459381,
+      "grad_norm": 0.5741217136383057,
+      "learning_rate": 0.0002,
+      "loss": 0.7959,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7954086169266834,
+      "grad_norm": 0.6605045199394226,
+      "learning_rate": 0.0002,
+      "loss": 0.8093,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7970253011074286,
+      "grad_norm": 0.5126531720161438,
+      "learning_rate": 0.0002,
+      "loss": 0.77,
+      "step": 4930
+    },
+    {
+      "epoch": 0.798641985288174,
+      "grad_norm": 0.513648271560669,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8002586694689192,
+      "grad_norm": 0.5350404381752014,
+      "learning_rate": 0.0002,
+      "loss": 0.8314,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8018753536496646,
+      "grad_norm": 0.5731674432754517,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8034920378304098,
+      "grad_norm": 0.5974258184432983,
+      "learning_rate": 0.0002,
+      "loss": 0.8572,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8051087220111551,
+      "grad_norm": 0.8774799704551697,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8067254061919004,
+      "grad_norm": 0.5994430184364319,
+      "learning_rate": 0.0002,
+      "loss": 0.7899,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8083420903726457,
+      "grad_norm": 0.4894903004169464,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5000
+    },
+    {
+      "epoch": 0.809958774553391,
+      "grad_norm": 0.5218459367752075,
+      "learning_rate": 0.0002,
+      "loss": 0.78,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8115754587341363,
+      "grad_norm": 0.5232468843460083,
+      "learning_rate": 0.0002,
+      "loss": 0.817,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8131921429148816,
+      "grad_norm": 0.44358372688293457,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8148088270956269,
+      "grad_norm": 0.6202037334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8164255112763722,
+      "grad_norm": 0.7721474170684814,
+      "learning_rate": 0.0002,
+      "loss": 0.7351,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8180421954571174,
+      "grad_norm": 0.5568501353263855,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8196588796378628,
+      "grad_norm": 0.49148809909820557,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 5070
+    },
+    {
+      "epoch": 0.821275563818608,
+      "grad_norm": 0.4956012964248657,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8228922479993533,
+      "grad_norm": 0.6078833937644958,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8245089321800986,
+      "grad_norm": 0.46906954050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.828,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8261256163608439,
+      "grad_norm": 0.50812166929245,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8277423005415891,
+      "grad_norm": 0.5319661498069763,
+      "learning_rate": 0.0002,
+      "loss": 0.8243,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8293589847223345,
+      "grad_norm": 0.4949689209461212,
+      "learning_rate": 0.0002,
+      "loss": 0.7798,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8309756689030798,
+      "grad_norm": 0.5151591300964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8325923530838251,
+      "grad_norm": 0.5530214309692383,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8342090372645704,
+      "grad_norm": 0.6297410130500793,
+      "learning_rate": 0.0002,
+      "loss": 0.8251,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8358257214453156,
+      "grad_norm": 0.5466840267181396,
+      "learning_rate": 0.0002,
+      "loss": 0.8067,
+      "step": 5170
+    },
+    {
+      "epoch": 0.837442405626061,
+      "grad_norm": 0.652913510799408,
+      "learning_rate": 0.0002,
+      "loss": 0.7875,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8390590898068062,
+      "grad_norm": 0.5811293125152588,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8406757739875516,
+      "grad_norm": 0.5109550952911377,
+      "learning_rate": 0.0002,
+      "loss": 0.7412,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8422924581682968,
+      "grad_norm": 0.4551706612110138,
+      "learning_rate": 0.0002,
+      "loss": 0.8077,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8439091423490421,
+      "grad_norm": 0.5813754200935364,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8455258265297874,
+      "grad_norm": 0.5856947898864746,
+      "learning_rate": 0.0002,
+      "loss": 0.802,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8471425107105327,
+      "grad_norm": 0.5482739210128784,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8487591948912779,
+      "grad_norm": 0.49023720622062683,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8503758790720233,
+      "grad_norm": 0.49472475051879883,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8519925632527686,
+      "grad_norm": 0.5490226745605469,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8536092474335139,
+      "grad_norm": 0.5340665578842163,
+      "learning_rate": 0.0002,
+      "loss": 0.8333,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8552259316142592,
+      "grad_norm": 0.5962483882904053,
+      "learning_rate": 0.0002,
+      "loss": 0.8277,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8568426157950044,
+      "grad_norm": 0.586358368396759,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8584592999757498,
+      "grad_norm": 0.49120277166366577,
+      "learning_rate": 0.0002,
+      "loss": 0.7831,
+      "step": 5310
+    },
+    {
+      "epoch": 0.860075984156495,
+      "grad_norm": 0.5887332558631897,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8616926683372403,
+      "grad_norm": 0.42496153712272644,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.5489874482154846,
+      "learning_rate": 0.0002,
+      "loss": 0.7905,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8649260366987309,
+      "grad_norm": 0.5850813984870911,
+      "learning_rate": 0.0002,
+      "loss": 0.7958,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8665427208794761,
+      "grad_norm": 0.517487108707428,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8681594050602215,
+      "grad_norm": 0.5339142680168152,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8697760892409668,
+      "grad_norm": 0.6236387491226196,
+      "learning_rate": 0.0002,
+      "loss": 0.818,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8713927734217121,
+      "grad_norm": 0.5752192735671997,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8730094576024574,
+      "grad_norm": 0.6724614500999451,
+      "learning_rate": 0.0002,
+      "loss": 0.8542,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8746261417832026,
+      "grad_norm": 0.5280613303184509,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 5410
+    },
+    {
+      "epoch": 0.876242825963948,
+      "grad_norm": 0.44033288955688477,
+      "learning_rate": 0.0002,
+      "loss": 0.8231,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8778595101446932,
+      "grad_norm": 0.5199708342552185,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8794761943254386,
+      "grad_norm": 0.46778348088264465,
+      "learning_rate": 0.0002,
+      "loss": 0.7852,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8810928785061838,
+      "grad_norm": 0.4657754898071289,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8827095626869291,
+      "grad_norm": 0.5472902655601501,
+      "learning_rate": 0.0002,
+      "loss": 0.7799,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8843262468676744,
+      "grad_norm": 0.4876766800880432,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8859429310484197,
+      "grad_norm": 0.5057248473167419,
+      "learning_rate": 0.0002,
+      "loss": 0.7906,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8875596152291649,
+      "grad_norm": 0.4637320637702942,
+      "learning_rate": 0.0002,
+      "loss": 0.8124,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8891762994099103,
+      "grad_norm": 0.471955806016922,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8907929835906556,
+      "grad_norm": 0.5209813714027405,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8924096677714008,
+      "grad_norm": 0.6213834285736084,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8940263519521462,
+      "grad_norm": 0.5215408205986023,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8956430361328914,
+      "grad_norm": 0.580478310585022,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8972597203136368,
+      "grad_norm": 0.49102169275283813,
+      "learning_rate": 0.0002,
+      "loss": 0.8371,
+      "step": 5550
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.6043479442596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7806,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9004930886751273,
+      "grad_norm": 0.5636463165283203,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9021097728558726,
+      "grad_norm": 0.5620124340057373,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9037264570366179,
+      "grad_norm": 0.5206354856491089,
+      "learning_rate": 0.0002,
+      "loss": 0.8083,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9053431412173631,
+      "grad_norm": 0.5798229575157166,
+      "learning_rate": 0.0002,
+      "loss": 0.8557,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9069598253981085,
+      "grad_norm": 0.6428212523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.8097,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9085765095788537,
+      "grad_norm": 0.48064687848091125,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9101931937595991,
+      "grad_norm": 0.6347860097885132,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9118098779403444,
+      "grad_norm": 0.5353913307189941,
+      "learning_rate": 0.0002,
+      "loss": 0.851,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9134265621210896,
+      "grad_norm": 0.5323944091796875,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5650
+    },
+    {
+      "epoch": 0.915043246301835,
+      "grad_norm": 0.5261843204498291,
+      "learning_rate": 0.0002,
+      "loss": 0.8393,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9166599304825802,
+      "grad_norm": 0.5451326966285706,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9182766146633256,
+      "grad_norm": 0.5183324217796326,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9198932988440708,
+      "grad_norm": 0.47229018807411194,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9215099830248161,
+      "grad_norm": 0.49180513620376587,
+      "learning_rate": 0.0002,
+      "loss": 0.7757,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9231266672055614,
+      "grad_norm": 0.5419785380363464,
+      "learning_rate": 0.0002,
+      "loss": 0.8735,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9247433513863067,
+      "grad_norm": 0.5408698916435242,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9263600355670519,
+      "grad_norm": 0.5286232829093933,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9279767197477973,
+      "grad_norm": 0.7539758086204529,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9295934039285425,
+      "grad_norm": 0.5166944861412048,
+      "learning_rate": 0.0002,
+      "loss": 0.8118,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9312100881092878,
+      "grad_norm": 0.6601425409317017,
+      "learning_rate": 0.0002,
+      "loss": 0.783,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9328267722900332,
+      "grad_norm": 0.5029960870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9344434564707784,
+      "grad_norm": 0.4926645755767822,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9360601406515238,
+      "grad_norm": 0.5739615559577942,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5790
+    },
+    {
+      "epoch": 0.937676824832269,
+      "grad_norm": 0.5058279037475586,
+      "learning_rate": 0.0002,
+      "loss": 0.8037,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9392935090130143,
+      "grad_norm": 0.5260962247848511,
+      "learning_rate": 0.0002,
+      "loss": 0.8537,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9409101931937596,
+      "grad_norm": 0.5768588185310364,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9425268773745049,
+      "grad_norm": 0.5170126557350159,
+      "learning_rate": 0.0002,
+      "loss": 0.8215,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9441435615552501,
+      "grad_norm": 0.5745864510536194,
+      "learning_rate": 0.0002,
+      "loss": 0.7422,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9457602457359955,
+      "grad_norm": 0.5551357865333557,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9473769299167407,
+      "grad_norm": 0.5776078701019287,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9489936140974861,
+      "grad_norm": 0.5340062379837036,
+      "learning_rate": 0.0002,
+      "loss": 0.8527,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9506102982782314,
+      "grad_norm": 0.6447290182113647,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9522269824589766,
+      "grad_norm": 0.5123815536499023,
+      "learning_rate": 0.0002,
+      "loss": 0.7945,
+      "step": 5890
+    },
+    {
+      "epoch": 0.953843666639722,
+      "grad_norm": 0.48547613620758057,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9554603508204672,
+      "grad_norm": 0.5791414976119995,
+      "learning_rate": 0.0002,
+      "loss": 0.7896,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9570770350012126,
+      "grad_norm": 0.6195011734962463,
+      "learning_rate": 0.0002,
+      "loss": 0.8408,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9586937191819578,
+      "grad_norm": 0.6323803067207336,
+      "learning_rate": 0.0002,
+      "loss": 0.7805,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9603104033627031,
+      "grad_norm": 0.45552879571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9619270875434484,
+      "grad_norm": 0.5796473622322083,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9635437717241937,
+      "grad_norm": 0.647261381149292,
+      "learning_rate": 0.0002,
+      "loss": 0.7672,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9651604559049389,
+      "grad_norm": 0.5487682819366455,
+      "learning_rate": 0.0002,
+      "loss": 0.8086,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9667771400856843,
+      "grad_norm": 0.5743663907051086,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9683938242664295,
+      "grad_norm": 0.5470591187477112,
+      "learning_rate": 0.0002,
+      "loss": 0.8153,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9700105084471748,
+      "grad_norm": 0.5901660323143005,
+      "learning_rate": 0.0002,
+      "loss": 0.8119,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9716271926279202,
+      "grad_norm": 0.6544759273529053,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9732438768086654,
+      "grad_norm": 0.6288470029830933,
+      "learning_rate": 0.0002,
+      "loss": 0.7536,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9748605609894108,
+      "grad_norm": 0.673153817653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 6030
+    },
+    {
+      "epoch": 0.976477245170156,
+      "grad_norm": 0.42854753136634827,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9780939293509013,
+      "grad_norm": 0.5227066278457642,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9797106135316466,
+      "grad_norm": 0.5372416973114014,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9813272977123919,
+      "grad_norm": 0.6026402115821838,
+      "learning_rate": 0.0002,
+      "loss": 0.7591,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9829439818931371,
+      "grad_norm": 0.49547791481018066,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9845606660738825,
+      "grad_norm": 0.4641951322555542,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9861773502546277,
+      "grad_norm": 0.5818535089492798,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9877940344353731,
+      "grad_norm": 0.63955157995224,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9894107186161183,
+      "grad_norm": 0.5649438500404358,
+      "learning_rate": 0.0002,
+      "loss": 0.7547,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9910274027968636,
+      "grad_norm": 0.5290433168411255,
+      "learning_rate": 0.0002,
+      "loss": 0.7861,
+      "step": 6130
+    },
+    {
+      "epoch": 0.992644086977609,
+      "grad_norm": 0.6399374008178711,
+      "learning_rate": 0.0002,
+      "loss": 0.8109,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9942607711583542,
+      "grad_norm": 0.6736576557159424,
+      "learning_rate": 0.0002,
+      "loss": 0.8373,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9958774553390995,
+      "grad_norm": 0.515420138835907,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9974941395198448,
+      "grad_norm": 0.562677800655365,
+      "learning_rate": 0.0002,
+      "loss": 0.8032,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9991108237005901,
+      "grad_norm": 0.7113858461380005,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9999191657909627,
+      "eval_loss": 1.0871200561523438,
+      "eval_runtime": 122.2071,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0007275078813354,
+      "grad_norm": 0.7111801505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0023441920620806,
+      "grad_norm": 0.5402125716209412,
+      "learning_rate": 0.0002,
+      "loss": 0.6865,
+      "step": 6200
+    },
+    {
+      "epoch": 1.003960876242826,
+      "grad_norm": 0.6098830103874207,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0055775604235713,
+      "grad_norm": 0.5829983353614807,
+      "learning_rate": 0.0002,
+      "loss": 0.7631,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0071942446043165,
+      "grad_norm": 0.5614621043205261,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0088109287850617,
+      "grad_norm": 0.5954238772392273,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0104276129658072,
+      "grad_norm": 0.6480574607849121,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0120442971465524,
+      "grad_norm": 0.6051128506660461,
+      "learning_rate": 0.0002,
+      "loss": 0.7514,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0136609813272976,
+      "grad_norm": 0.6318870782852173,
+      "learning_rate": 0.0002,
+      "loss": 0.7237,
+      "step": 6270
+    },
+    {
+      "epoch": 1.015277665508043,
+      "grad_norm": 0.5048980116844177,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0168943496887883,
+      "grad_norm": 0.6346936225891113,
+      "learning_rate": 0.0002,
+      "loss": 0.7391,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0185110338695336,
+      "grad_norm": 0.5711665749549866,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0201277180502788,
+      "grad_norm": 0.5175361037254333,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0217444022310243,
+      "grad_norm": 0.5360831618309021,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0233610864117695,
+      "grad_norm": 0.614675760269165,
+      "learning_rate": 0.0002,
+      "loss": 0.7112,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0249777705925147,
+      "grad_norm": 0.5626118183135986,
+      "learning_rate": 0.0002,
+      "loss": 0.7748,
+      "step": 6340
+    },
+    {
+      "epoch": 1.02659445477326,
+      "grad_norm": 0.574897289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.7375,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0282111389540054,
+      "grad_norm": 0.7185447812080383,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0298278231347506,
+      "grad_norm": 0.6705799698829651,
+      "learning_rate": 0.0002,
+      "loss": 0.703,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0314445073154959,
+      "grad_norm": 0.6740428805351257,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0330611914962413,
+      "grad_norm": 0.663902759552002,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0346778756769865,
+      "grad_norm": 0.5029543042182922,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0362945598577318,
+      "grad_norm": 0.7813863158226013,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 6410
+    },
+    {
+      "epoch": 1.037911244038477,
+      "grad_norm": 0.5396282076835632,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0395279282192225,
+      "grad_norm": 0.5253293514251709,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0411446123999677,
+      "grad_norm": 0.7236770987510681,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 6440
+    },
+    {
+      "epoch": 1.042761296580713,
+      "grad_norm": 0.5670917630195618,
+      "learning_rate": 0.0002,
+      "loss": 0.7259,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0443779807614582,
+      "grad_norm": 0.6031978726387024,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0459946649422036,
+      "grad_norm": 0.5309213399887085,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0476113491229488,
+      "grad_norm": 0.7114651799201965,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 6480
+    },
+    {
+      "epoch": 1.049228033303694,
+      "grad_norm": 0.5591610670089722,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0508447174844395,
+      "grad_norm": 0.5185961127281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0524614016651848,
+      "grad_norm": 0.6510552167892456,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6510
+    },
+    {
+      "epoch": 1.05407808584593,
+      "grad_norm": 0.6557928919792175,
+      "learning_rate": 0.0002,
+      "loss": 0.7057,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0556947700266752,
+      "grad_norm": 0.6973192691802979,
+      "learning_rate": 0.0002,
+      "loss": 0.8056,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0573114542074207,
+      "grad_norm": 0.6226583123207092,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 6540
+    },
+    {
+      "epoch": 1.058928138388166,
+      "grad_norm": 0.5633195638656616,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0605448225689111,
+      "grad_norm": 0.7466658353805542,
+      "learning_rate": 0.0002,
+      "loss": 0.7082,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0621615067496564,
+      "grad_norm": 0.6462772488594055,
+      "learning_rate": 0.0002,
+      "loss": 0.7059,
+      "step": 6570
+    },
+    {
+      "epoch": 1.0637781909304018,
+      "grad_norm": 0.5266856551170349,
+      "learning_rate": 0.0002,
+      "loss": 0.7046,
+      "step": 6580
+    },
+    {
+      "epoch": 1.065394875111147,
+      "grad_norm": 0.534392774105072,
+      "learning_rate": 0.0002,
+      "loss": 0.7157,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0670115592918923,
+      "grad_norm": 0.7514177560806274,
+      "learning_rate": 0.0002,
+      "loss": 0.7115,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0686282434726375,
+      "grad_norm": 0.7593035697937012,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 6610
+    },
+    {
+      "epoch": 1.070244927653383,
+      "grad_norm": 0.5277858972549438,
+      "learning_rate": 0.0002,
+      "loss": 0.6836,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0718616118341282,
+      "grad_norm": 0.5573670268058777,
+      "learning_rate": 0.0002,
+      "loss": 0.7405,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0734782960148734,
+      "grad_norm": 0.6802396774291992,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 6640
+    },
+    {
+      "epoch": 1.0750949801956189,
+      "grad_norm": 0.7367215752601624,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0767116643763641,
+      "grad_norm": 0.5961891412734985,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0783283485571094,
+      "grad_norm": 0.5736313462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0799450327378546,
+      "grad_norm": 0.619219183921814,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0815617169186,
+      "grad_norm": 0.6214390993118286,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0831784010993453,
+      "grad_norm": 0.564536988735199,
+      "learning_rate": 0.0002,
+      "loss": 0.7554,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0847950852800905,
+      "grad_norm": 0.5838140249252319,
+      "learning_rate": 0.0002,
+      "loss": 0.7362,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0864117694608357,
+      "grad_norm": 0.7000553607940674,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0880284536415812,
+      "grad_norm": 0.7078263759613037,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 6730
+    },
+    {
+      "epoch": 1.0896451378223264,
+      "grad_norm": 0.8353848457336426,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6740
+    },
+    {
+      "epoch": 1.0912618220030716,
+      "grad_norm": 0.5615518689155579,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 6750
+    },
+    {
+      "epoch": 1.0928785061838169,
+      "grad_norm": 0.5475581288337708,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 6760
+    },
+    {
+      "epoch": 1.0944951903645623,
+      "grad_norm": 0.5835978388786316,
+      "learning_rate": 0.0002,
+      "loss": 0.7652,
+      "step": 6770
+    },
+    {
+      "epoch": 1.0961118745453076,
+      "grad_norm": 0.5516105890274048,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 6780
+    },
+    {
+      "epoch": 1.0977285587260528,
+      "grad_norm": 0.5875251889228821,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 6790
+    },
+    {
+      "epoch": 1.0993452429067982,
+      "grad_norm": 0.7376947999000549,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1009619270875435,
+      "grad_norm": 0.5656165480613708,
+      "learning_rate": 0.0002,
+      "loss": 0.7512,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1025786112682887,
+      "grad_norm": 0.6365954279899597,
+      "learning_rate": 0.0002,
+      "loss": 0.7409,
+      "step": 6820
+    },
+    {
+      "epoch": 1.104195295449034,
+      "grad_norm": 0.5033080577850342,
+      "learning_rate": 0.0002,
+      "loss": 0.7392,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1058119796297794,
+      "grad_norm": 0.617396891117096,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1074286638105246,
+      "grad_norm": 0.6395374536514282,
+      "learning_rate": 0.0002,
+      "loss": 0.7006,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1090453479912699,
+      "grad_norm": 0.6775295734405518,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1106620321720153,
+      "grad_norm": 0.6655223965644836,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1122787163527605,
+      "grad_norm": 0.676655113697052,
+      "learning_rate": 0.0002,
+      "loss": 0.7553,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1138954005335058,
+      "grad_norm": 0.6062718629837036,
+      "learning_rate": 0.0002,
+      "loss": 0.7342,
+      "step": 6890
+    },
+    {
+      "epoch": 1.115512084714251,
+      "grad_norm": 0.590943455696106,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1171287688949965,
+      "grad_norm": 0.6315317153930664,
+      "learning_rate": 0.0002,
+      "loss": 0.6705,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1187454530757417,
+      "grad_norm": 0.47979024052619934,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 6920
+    },
+    {
+      "epoch": 1.120362137256487,
+      "grad_norm": 0.647298276424408,
+      "learning_rate": 0.0002,
+      "loss": 0.7002,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1219788214372322,
+      "grad_norm": 0.7336484789848328,
+      "learning_rate": 0.0002,
+      "loss": 0.7502,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1235955056179776,
+      "grad_norm": 0.5071424245834351,
+      "learning_rate": 0.0002,
+      "loss": 0.693,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1252121897987228,
+      "grad_norm": 0.6527144312858582,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 6960
+    },
+    {
+      "epoch": 1.126828873979468,
+      "grad_norm": 0.6935935020446777,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1284455581602133,
+      "grad_norm": 0.8026931881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1300622423409588,
+      "grad_norm": 0.5210393667221069,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 6990
+    },
+    {
+      "epoch": 1.131678926521704,
+      "grad_norm": 0.60475093126297,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1332956107024492,
+      "grad_norm": 0.6417073607444763,
+      "learning_rate": 0.0002,
+      "loss": 0.7495,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1349122948831947,
+      "grad_norm": 0.6732175946235657,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 7020
+    },
+    {
+      "epoch": 1.13652897906394,
+      "grad_norm": 0.6719491481781006,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1381456632446851,
+      "grad_norm": 0.5708295106887817,
+      "learning_rate": 0.0002,
+      "loss": 0.7694,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1397623474254304,
+      "grad_norm": 0.7141719460487366,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1413790316061758,
+      "grad_norm": 0.6187017560005188,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 7060
+    },
+    {
+      "epoch": 1.142995715786921,
+      "grad_norm": 0.50581294298172,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1446123999676663,
+      "grad_norm": 0.5620143413543701,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1462290841484115,
+      "grad_norm": 0.6231929659843445,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 7090
+    },
+    {
+      "epoch": 1.147845768329157,
+      "grad_norm": 0.5775774121284485,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1494624525099022,
+      "grad_norm": 0.6492809653282166,
+      "learning_rate": 0.0002,
+      "loss": 0.7728,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1510791366906474,
+      "grad_norm": 0.6434972286224365,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 7120
+    },
+    {
+      "epoch": 1.1526958208713927,
+      "grad_norm": 0.6191812753677368,
+      "learning_rate": 0.0002,
+      "loss": 0.7374,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1543125050521381,
+      "grad_norm": 0.6690331697463989,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 7140
+    },
+    {
+      "epoch": 1.1559291892328833,
+      "grad_norm": 0.5977938175201416,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1575458734136286,
+      "grad_norm": 0.6195854544639587,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7160
+    },
+    {
+      "epoch": 1.159162557594374,
+      "grad_norm": 0.5752048492431641,
+      "learning_rate": 0.0002,
+      "loss": 0.7249,
+      "step": 7170
+    },
+    {
+      "epoch": 1.1607792417751193,
+      "grad_norm": 0.589081883430481,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1623959259558645,
+      "grad_norm": 0.756996750831604,
+      "learning_rate": 0.0002,
+      "loss": 0.704,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1640126101366097,
+      "grad_norm": 0.7614967226982117,
+      "learning_rate": 0.0002,
+      "loss": 0.7404,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1656292943173552,
+      "grad_norm": 0.6120437979698181,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1672459784981004,
+      "grad_norm": 0.6210004687309265,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1688626626788456,
+      "grad_norm": 0.6044116020202637,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7230
+    },
+    {
+      "epoch": 1.170479346859591,
+      "grad_norm": 0.5418457388877869,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1720960310403363,
+      "grad_norm": 0.6413537263870239,
+      "learning_rate": 0.0002,
+      "loss": 0.6938,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1737127152210816,
+      "grad_norm": 0.5777867436408997,
+      "learning_rate": 0.0002,
+      "loss": 0.6978,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1753293994018268,
+      "grad_norm": 0.7092402577400208,
+      "learning_rate": 0.0002,
+      "loss": 0.7503,
+      "step": 7270
+    },
+    {
+      "epoch": 1.176946083582572,
+      "grad_norm": 0.6351709365844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1785627677633175,
+      "grad_norm": 0.6172189712524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7527,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1801794519440627,
+      "grad_norm": 0.6801714897155762,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7300
+    },
+    {
+      "epoch": 1.181796136124808,
+      "grad_norm": 0.6044712066650391,
+      "learning_rate": 0.0002,
+      "loss": 0.6941,
+      "step": 7310
+    },
+    {
+      "epoch": 1.1834128203055534,
+      "grad_norm": 0.7413212060928345,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 7320
+    },
+    {
+      "epoch": 1.1850295044862986,
+      "grad_norm": 0.5303856134414673,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 7330
+    },
+    {
+      "epoch": 1.1866461886670439,
+      "grad_norm": 0.5647098422050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6915,
+      "step": 7340
+    },
+    {
+      "epoch": 1.188262872847789,
+      "grad_norm": 0.7374135255813599,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 7350
+    },
+    {
+      "epoch": 1.1898795570285345,
+      "grad_norm": 0.5710089206695557,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1914962412092798,
+      "grad_norm": 0.6073619723320007,
+      "learning_rate": 0.0002,
+      "loss": 0.8289,
+      "step": 7370
+    },
+    {
+      "epoch": 1.193112925390025,
+      "grad_norm": 0.5899916887283325,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1947296095707705,
+      "grad_norm": 0.7762434482574463,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1963462937515157,
+      "grad_norm": 0.679949939250946,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7400
+    },
+    {
+      "epoch": 1.197962977932261,
+      "grad_norm": 0.6106849312782288,
+      "learning_rate": 0.0002,
+      "loss": 0.7599,
+      "step": 7410
+    },
+    {
+      "epoch": 1.1995796621130062,
+      "grad_norm": 0.682461678981781,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2011963462937516,
+      "grad_norm": 0.6087017059326172,
+      "learning_rate": 0.0002,
+      "loss": 0.7741,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2028130304744968,
+      "grad_norm": 0.63739013671875,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 7440
+    },
+    {
+      "epoch": 1.204429714655242,
+      "grad_norm": 0.6154777407646179,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2060463988359873,
+      "grad_norm": 0.7491534948348999,
+      "learning_rate": 0.0002,
+      "loss": 0.7565,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2076630830167328,
+      "grad_norm": 0.6664797067642212,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 7470
+    },
+    {
+      "epoch": 1.209279767197478,
+      "grad_norm": 0.6660266518592834,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2108964513782232,
+      "grad_norm": 0.6972551345825195,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2125131355589684,
+      "grad_norm": 0.6157945990562439,
+      "learning_rate": 0.0002,
+      "loss": 0.7023,
+      "step": 7500
+    },
+    {
+      "epoch": 1.214129819739714,
+      "grad_norm": 0.5199310183525085,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2157465039204591,
+      "grad_norm": 0.577610433101654,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2173631881012044,
+      "grad_norm": 0.53652423620224,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2189798722819498,
+      "grad_norm": 0.6479050517082214,
+      "learning_rate": 0.0002,
+      "loss": 0.7393,
+      "step": 7540
+    },
+    {
+      "epoch": 1.220596556462695,
+      "grad_norm": 0.618748128414154,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2222132406434403,
+      "grad_norm": 0.6311424374580383,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2238299248241855,
+      "grad_norm": 0.6595825552940369,
+      "learning_rate": 0.0002,
+      "loss": 0.7272,
+      "step": 7570
+    },
+    {
+      "epoch": 1.225446609004931,
+      "grad_norm": 0.5198960900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2270632931856762,
+      "grad_norm": 0.578650712966919,
+      "learning_rate": 0.0002,
+      "loss": 0.674,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2286799773664214,
+      "grad_norm": 0.6080220937728882,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2302966615471669,
+      "grad_norm": 0.7050248384475708,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2319133457279121,
+      "grad_norm": 0.6652196049690247,
+      "learning_rate": 0.0002,
+      "loss": 0.7032,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2335300299086573,
+      "grad_norm": 0.7322776317596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7085,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2351467140894026,
+      "grad_norm": 0.4998728036880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2367633982701478,
+      "grad_norm": 0.6428788900375366,
+      "learning_rate": 0.0002,
+      "loss": 0.7214,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2383800824508933,
+      "grad_norm": 0.585242509841919,
+      "learning_rate": 0.0002,
+      "loss": 0.7699,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2399967666316385,
+      "grad_norm": 0.5211917757987976,
+      "learning_rate": 0.0002,
+      "loss": 0.7621,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2416134508123837,
+      "grad_norm": 0.6490384340286255,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2432301349931292,
+      "grad_norm": 0.6249763369560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2448468191738744,
+      "grad_norm": 0.71870356798172,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2464635033546196,
+      "grad_norm": 0.6761967539787292,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2480801875353649,
+      "grad_norm": 0.6500617265701294,
+      "learning_rate": 0.0002,
+      "loss": 0.7501,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2496968717161103,
+      "grad_norm": 0.8069869875907898,
+      "learning_rate": 0.0002,
+      "loss": 0.7903,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2513135558968556,
+      "grad_norm": 0.6044608950614929,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2529302400776008,
+      "grad_norm": 0.6573283076286316,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2545469242583462,
+      "grad_norm": 0.625430166721344,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2561636084390915,
+      "grad_norm": 0.5442022681236267,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7770
+    },
+    {
+      "epoch": 1.2577802926198367,
+      "grad_norm": 0.6818386912345886,
+      "learning_rate": 0.0002,
+      "loss": 0.7172,
+      "step": 7780
+    },
+    {
+      "epoch": 1.259396976800582,
+      "grad_norm": 0.6381874084472656,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2610136609813272,
+      "grad_norm": 0.6269212961196899,
+      "learning_rate": 0.0002,
+      "loss": 0.6834,
+      "step": 7800
+    },
+    {
+      "epoch": 1.2626303451620726,
+      "grad_norm": 0.600121259689331,
+      "learning_rate": 0.0002,
+      "loss": 0.7821,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2642470293428179,
+      "grad_norm": 0.6337703466415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2658637135235633,
+      "grad_norm": 0.7234963774681091,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 7830
+    },
+    {
+      "epoch": 1.2674803977043085,
+      "grad_norm": 0.800184965133667,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2690970818850538,
+      "grad_norm": 0.7539464831352234,
+      "learning_rate": 0.0002,
+      "loss": 0.7426,
+      "step": 7850
+    },
+    {
+      "epoch": 1.270713766065799,
+      "grad_norm": 0.5493760704994202,
+      "learning_rate": 0.0002,
+      "loss": 0.7496,
+      "step": 7860
+    },
+    {
+      "epoch": 1.2723304502465442,
+      "grad_norm": 0.7477145791053772,
+      "learning_rate": 0.0002,
+      "loss": 0.7537,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2739471344272897,
+      "grad_norm": 0.6366362571716309,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 7880
+    },
+    {
+      "epoch": 1.275563818608035,
+      "grad_norm": 0.7419533729553223,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2771805027887801,
+      "grad_norm": 0.6141223311424255,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 7900
+    },
+    {
+      "epoch": 1.2787971869695256,
+      "grad_norm": 0.7522598505020142,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 7910
+    },
+    {
+      "epoch": 1.2804138711502708,
+      "grad_norm": 0.6935804486274719,
+      "learning_rate": 0.0002,
+      "loss": 0.7097,
+      "step": 7920
+    },
+    {
+      "epoch": 1.282030555331016,
+      "grad_norm": 0.7239290475845337,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 7930
+    },
+    {
+      "epoch": 1.2836472395117613,
+      "grad_norm": 0.8800187110900879,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 7940
+    },
+    {
+      "epoch": 1.2852639236925067,
+      "grad_norm": 0.540458083152771,
+      "learning_rate": 0.0002,
+      "loss": 0.6991,
+      "step": 7950
+    },
+    {
+      "epoch": 1.286880607873252,
+      "grad_norm": 0.6492934226989746,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 7960
+    },
+    {
+      "epoch": 1.2884972920539972,
+      "grad_norm": 0.6543959379196167,
+      "learning_rate": 0.0002,
+      "loss": 0.7742,
+      "step": 7970
+    },
+    {
+      "epoch": 1.2901139762347427,
+      "grad_norm": 0.5804705619812012,
+      "learning_rate": 0.0002,
+      "loss": 0.7316,
+      "step": 7980
+    },
+    {
+      "epoch": 1.291730660415488,
+      "grad_norm": 0.7074727416038513,
+      "learning_rate": 0.0002,
+      "loss": 0.796,
+      "step": 7990
+    },
+    {
+      "epoch": 1.2933473445962331,
+      "grad_norm": 0.5347974300384521,
+      "learning_rate": 0.0002,
+      "loss": 0.7034,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2949640287769784,
+      "grad_norm": 0.6457298398017883,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 8010
+    },
+    {
+      "epoch": 1.2965807129577236,
+      "grad_norm": 0.6407219171524048,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8020
+    },
+    {
+      "epoch": 1.298197397138469,
+      "grad_norm": 0.828439474105835,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2998140813192143,
+      "grad_norm": 0.4840380549430847,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3014307654999595,
+      "grad_norm": 0.5921024680137634,
+      "learning_rate": 0.0002,
+      "loss": 0.7283,
+      "step": 8050
+    },
+    {
+      "epoch": 1.303047449680705,
+      "grad_norm": 0.6170315146446228,
+      "learning_rate": 0.0002,
+      "loss": 0.7477,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3046641338614502,
+      "grad_norm": 0.5374847054481506,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3062808180421954,
+      "grad_norm": 0.545758068561554,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3078975022229407,
+      "grad_norm": 0.55641770362854,
+      "learning_rate": 0.0002,
+      "loss": 0.7463,
+      "step": 8090
+    },
+    {
+      "epoch": 1.309514186403686,
+      "grad_norm": 0.6724897027015686,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3111308705844313,
+      "grad_norm": 0.6923972368240356,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3127475547651766,
+      "grad_norm": 0.5136841535568237,
+      "learning_rate": 0.0002,
+      "loss": 0.7149,
+      "step": 8120
+    },
+    {
+      "epoch": 1.314364238945922,
+      "grad_norm": 0.6766283512115479,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3159809231266673,
+      "grad_norm": 0.6283926367759705,
+      "learning_rate": 0.0002,
+      "loss": 0.7489,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3175976073074125,
+      "grad_norm": 0.644216001033783,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3192142914881577,
+      "grad_norm": 0.7827503085136414,
+      "learning_rate": 0.0002,
+      "loss": 0.7125,
+      "step": 8160
+    },
+    {
+      "epoch": 1.320830975668903,
+      "grad_norm": 0.6651390790939331,
+      "learning_rate": 0.0002,
+      "loss": 0.7271,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3224476598496484,
+      "grad_norm": 0.5547412633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7778,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3240643440303936,
+      "grad_norm": 0.6765179634094238,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 8190
+    },
+    {
+      "epoch": 1.325681028211139,
+      "grad_norm": 0.6822077035903931,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3272977123918843,
+      "grad_norm": 0.5941002368927002,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3289143965726296,
+      "grad_norm": 0.4850037097930908,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3305310807533748,
+      "grad_norm": 0.6162990927696228,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 8230
+    },
+    {
+      "epoch": 1.33214776493412,
+      "grad_norm": 0.6665613651275635,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3337644491148655,
+      "grad_norm": 0.618192732334137,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3353811332956107,
+      "grad_norm": 0.710418701171875,
+      "learning_rate": 0.0002,
+      "loss": 0.7532,
+      "step": 8260
+    },
+    {
+      "epoch": 1.336997817476356,
+      "grad_norm": 0.5109876990318298,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3386145016571014,
+      "grad_norm": 0.6791711449623108,
+      "learning_rate": 0.0002,
+      "loss": 0.7303,
+      "step": 8280
+    },
+    {
+      "epoch": 1.3402311858378466,
+      "grad_norm": 0.6836432814598083,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3418478700185918,
+      "grad_norm": 0.5579386353492737,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8300
+    },
+    {
+      "epoch": 1.343464554199337,
+      "grad_norm": 0.6713546514511108,
+      "learning_rate": 0.0002,
+      "loss": 0.7377,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3450812383800825,
+      "grad_norm": 0.5353720188140869,
+      "learning_rate": 0.0002,
+      "loss": 0.7756,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3466979225608278,
+      "grad_norm": 0.5813682675361633,
+      "learning_rate": 0.0002,
+      "loss": 0.718,
+      "step": 8330
+    },
+    {
+      "epoch": 1.348314606741573,
+      "grad_norm": 0.8158791661262512,
+      "learning_rate": 0.0002,
+      "loss": 0.7294,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3499312909223184,
+      "grad_norm": 0.6193785071372986,
+      "learning_rate": 0.0002,
+      "loss": 0.6992,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3515479751030637,
+      "grad_norm": 0.6353939771652222,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 8360
+    },
+    {
+      "epoch": 1.353164659283809,
+      "grad_norm": 0.6925048232078552,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 8370
+    },
+    {
+      "epoch": 1.3547813434645541,
+      "grad_norm": 0.988264799118042,
+      "learning_rate": 0.0002,
+      "loss": 0.736,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3563980276452994,
+      "grad_norm": 0.6476002931594849,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3580147118260448,
+      "grad_norm": 0.7120398879051208,
+      "learning_rate": 0.0002,
+      "loss": 0.776,
+      "step": 8400
+    },
+    {
+      "epoch": 1.35963139600679,
+      "grad_norm": 0.9048416614532471,
+      "learning_rate": 0.0002,
+      "loss": 0.7368,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3612480801875353,
+      "grad_norm": 0.7000672817230225,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3628647643682807,
+      "grad_norm": 0.6015632152557373,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 8430
+    },
+    {
+      "epoch": 1.364481448549026,
+      "grad_norm": 0.612516462802887,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3660981327297712,
+      "grad_norm": 0.5969301462173462,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 8450
+    },
+    {
+      "epoch": 1.3677148169105164,
+      "grad_norm": 0.6730654239654541,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 8460
+    },
+    {
+      "epoch": 1.369331501091262,
+      "grad_norm": 0.6386392116546631,
+      "learning_rate": 0.0002,
+      "loss": 0.7465,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3709481852720071,
+      "grad_norm": 0.739544153213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3725648694527524,
+      "grad_norm": 0.6462782621383667,
+      "learning_rate": 0.0002,
+      "loss": 0.7892,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3741815536334978,
+      "grad_norm": 0.7346843481063843,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 8500
+    },
+    {
+      "epoch": 1.375798237814243,
+      "grad_norm": 0.6884821057319641,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8510
+    },
+    {
+      "epoch": 1.3774149219949883,
+      "grad_norm": 0.6999333500862122,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3790316061757335,
+      "grad_norm": 0.5378713011741638,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 8530
+    },
+    {
+      "epoch": 1.3806482903564787,
+      "grad_norm": 0.5417906641960144,
+      "learning_rate": 0.0002,
+      "loss": 0.6797,
+      "step": 8540
+    },
+    {
+      "epoch": 1.3822649745372242,
+      "grad_norm": 0.6602526307106018,
+      "learning_rate": 0.0002,
+      "loss": 0.7499,
+      "step": 8550
+    },
+    {
+      "epoch": 1.3838816587179694,
+      "grad_norm": 0.7073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.7356,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3854983428987149,
+      "grad_norm": 0.5841707587242126,
+      "learning_rate": 0.0002,
+      "loss": 0.75,
+      "step": 8570
+    },
+    {
+      "epoch": 1.38711502707946,
+      "grad_norm": 0.7031095027923584,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 8580
+    },
+    {
+      "epoch": 1.3887317112602053,
+      "grad_norm": 0.5198570489883423,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 8590
+    },
+    {
+      "epoch": 1.3903483954409506,
+      "grad_norm": 0.7261320352554321,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 8600
+    },
+    {
+      "epoch": 1.3919650796216958,
+      "grad_norm": 0.5616350173950195,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3935817638024413,
+      "grad_norm": 0.5185914635658264,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 8620
+    },
+    {
+      "epoch": 1.3951984479831865,
+      "grad_norm": 0.5814694762229919,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 8630
+    },
+    {
+      "epoch": 1.3968151321639317,
+      "grad_norm": 0.6977371573448181,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 8640
+    },
+    {
+      "epoch": 1.3984318163446772,
+      "grad_norm": 0.6855689883232117,
+      "learning_rate": 0.0002,
+      "loss": 0.7574,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4000485005254224,
+      "grad_norm": 0.5414357781410217,
+      "learning_rate": 0.0002,
+      "loss": 0.7802,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4016651847061676,
+      "grad_norm": 0.6970012784004211,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4032818688869129,
+      "grad_norm": 0.526079535484314,
+      "learning_rate": 0.0002,
+      "loss": 0.7421,
+      "step": 8680
+    },
+    {
+      "epoch": 1.404898553067658,
+      "grad_norm": 0.758712887763977,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4065152372484035,
+      "grad_norm": 0.7118762731552124,
+      "learning_rate": 0.0002,
+      "loss": 0.7612,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4081319214291488,
+      "grad_norm": 0.5696909427642822,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4097486056098942,
+      "grad_norm": 0.7995436787605286,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4113652897906395,
+      "grad_norm": 0.7237521409988403,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4129819739713847,
+      "grad_norm": 0.744628369808197,
+      "learning_rate": 0.0002,
+      "loss": 0.7661,
+      "step": 8740
+    },
+    {
+      "epoch": 1.41459865815213,
+      "grad_norm": 0.6082926988601685,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4162153423328752,
+      "grad_norm": 0.5185243487358093,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4178320265136206,
+      "grad_norm": 0.5183082222938538,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4194487106943658,
+      "grad_norm": 0.7326041460037231,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 8780
+    },
+    {
+      "epoch": 1.421065394875111,
+      "grad_norm": 0.7174660563468933,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4226820790558565,
+      "grad_norm": 0.8080165982246399,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4242987632366018,
+      "grad_norm": 0.5061507821083069,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 8810
+    },
+    {
+      "epoch": 1.425915447417347,
+      "grad_norm": 0.801602840423584,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4275321315980922,
+      "grad_norm": 0.6150273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4291488157788377,
+      "grad_norm": 0.8786525726318359,
+      "learning_rate": 0.0002,
+      "loss": 0.7452,
+      "step": 8840
+    },
+    {
+      "epoch": 1.430765499959583,
+      "grad_norm": 0.6371538639068604,
+      "learning_rate": 0.0002,
+      "loss": 0.7257,
+      "step": 8850
+    },
+    {
+      "epoch": 1.4323821841403281,
+      "grad_norm": 0.6409295797348022,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4339988683210736,
+      "grad_norm": 0.6452359557151794,
+      "learning_rate": 0.0002,
+      "loss": 0.7891,
+      "step": 8870
+    },
+    {
+      "epoch": 1.4356155525018188,
+      "grad_norm": 0.5842334628105164,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 8880
+    },
+    {
+      "epoch": 1.437232236682564,
+      "grad_norm": 0.696761965751648,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4388489208633093,
+      "grad_norm": 0.6384600400924683,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4404656050440545,
+      "grad_norm": 0.5981136560440063,
+      "learning_rate": 0.0002,
+      "loss": 0.7049,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4420822892248,
+      "grad_norm": 0.6355637907981873,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4436989734055452,
+      "grad_norm": 0.6374830603599548,
+      "learning_rate": 0.0002,
+      "loss": 0.7653,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4453156575862904,
+      "grad_norm": 0.559013307094574,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 8940
+    },
+    {
+      "epoch": 1.446932341767036,
+      "grad_norm": 0.7289170026779175,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4485490259477811,
+      "grad_norm": 0.8649206757545471,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4501657101285264,
+      "grad_norm": 0.7664689421653748,
+      "learning_rate": 0.0002,
+      "loss": 0.7745,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4517823943092716,
+      "grad_norm": 0.7109952569007874,
+      "learning_rate": 0.0002,
+      "loss": 0.7431,
+      "step": 8980
+    },
+    {
+      "epoch": 1.453399078490017,
+      "grad_norm": 0.6312844753265381,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4550157626707623,
+      "grad_norm": 0.6616617441177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7467,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4566324468515075,
+      "grad_norm": 0.7384068965911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7518,
+      "step": 9010
+    },
+    {
+      "epoch": 1.458249131032253,
+      "grad_norm": 0.6549670100212097,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9020
+    },
+    {
+      "epoch": 1.4598658152129982,
+      "grad_norm": 0.6254119277000427,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4614824993937434,
+      "grad_norm": 0.6806328892707825,
+      "learning_rate": 0.0002,
+      "loss": 0.7645,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4630991835744886,
+      "grad_norm": 0.6803115010261536,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4647158677552339,
+      "grad_norm": 0.48529282212257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7264,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4663325519359793,
+      "grad_norm": 0.5995030999183655,
+      "learning_rate": 0.0002,
+      "loss": 0.7542,
+      "step": 9070
+    },
+    {
+      "epoch": 1.4679492361167246,
+      "grad_norm": 0.6005427837371826,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9080
+    },
+    {
+      "epoch": 1.46956592029747,
+      "grad_norm": 0.718564510345459,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4711826044782153,
+      "grad_norm": 0.7003577351570129,
+      "learning_rate": 0.0002,
+      "loss": 0.7089,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4727992886589605,
+      "grad_norm": 0.5888323783874512,
+      "learning_rate": 0.0002,
+      "loss": 0.8069,
+      "step": 9110
+    },
+    {
+      "epoch": 1.4744159728397057,
+      "grad_norm": 0.6417609453201294,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 9120
+    },
+    {
+      "epoch": 1.476032657020451,
+      "grad_norm": 0.572294294834137,
+      "learning_rate": 0.0002,
+      "loss": 0.7441,
+      "step": 9130
+    },
+    {
+      "epoch": 1.4776493412011964,
+      "grad_norm": 0.8200714588165283,
+      "learning_rate": 0.0002,
+      "loss": 0.8053,
+      "step": 9140
+    },
+    {
+      "epoch": 1.4792660253819416,
+      "grad_norm": 0.6343288421630859,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4808827095626869,
+      "grad_norm": 0.7017961144447327,
+      "learning_rate": 0.0002,
+      "loss": 0.7641,
+      "step": 9160
+    },
+    {
+      "epoch": 1.4824993937434323,
+      "grad_norm": 0.6202912926673889,
+      "learning_rate": 0.0002,
+      "loss": 0.7619,
+      "step": 9170
+    },
+    {
+      "epoch": 1.4841160779241775,
+      "grad_norm": 0.6677869558334351,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 9180
+    },
+    {
+      "epoch": 1.4857327621049228,
+      "grad_norm": 0.6052267551422119,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 9190
+    },
+    {
+      "epoch": 1.487349446285668,
+      "grad_norm": 0.6638872027397156,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 9200
+    },
+    {
+      "epoch": 1.4889661304664135,
+      "grad_norm": 0.6245523691177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 9210
+    },
+    {
+      "epoch": 1.4905828146471587,
+      "grad_norm": 0.5761767625808716,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 9220
+    },
+    {
+      "epoch": 1.492199498827904,
+      "grad_norm": 0.8175981640815735,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 9230
+    },
+    {
+      "epoch": 1.4938161830086494,
+      "grad_norm": 0.9144009947776794,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 9240
+    },
+    {
+      "epoch": 1.4954328671893946,
+      "grad_norm": 0.5742552876472473,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 9250
+    },
+    {
+      "epoch": 1.4970495513701398,
+      "grad_norm": 0.534534215927124,
+      "learning_rate": 0.0002,
+      "loss": 0.7418,
+      "step": 9260
+    },
+    {
+      "epoch": 1.498666235550885,
+      "grad_norm": 0.7836225032806396,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5002829197316303,
+      "grad_norm": 0.5292993187904358,
+      "learning_rate": 0.0002,
+      "loss": 0.7453,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5018996039123758,
+      "grad_norm": 0.8044071793556213,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9290
+    },
+    {
+      "epoch": 1.503516288093121,
+      "grad_norm": 0.6185805201530457,
+      "learning_rate": 0.0002,
+      "loss": 0.7229,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5051329722738664,
+      "grad_norm": 0.6093607544898987,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5067496564546117,
+      "grad_norm": 0.5891730189323425,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 9320
+    },
+    {
+      "epoch": 1.508366340635357,
+      "grad_norm": 0.6331129670143127,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5099830248161021,
+      "grad_norm": 0.7690958380699158,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5115997089968474,
+      "grad_norm": 0.6548877358436584,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5132163931775926,
+      "grad_norm": 0.6545143127441406,
+      "learning_rate": 0.0002,
+      "loss": 0.7408,
+      "step": 9360
+    },
+    {
+      "epoch": 1.514833077358338,
+      "grad_norm": 0.553247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.7432,
+      "step": 9370
+    },
+    {
+      "epoch": 1.5164497615390833,
+      "grad_norm": 0.8145074844360352,
+      "learning_rate": 0.0002,
+      "loss": 0.7265,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5180664457198287,
+      "grad_norm": 0.7636994123458862,
+      "learning_rate": 0.0002,
+      "loss": 0.7379,
+      "step": 9390
+    },
+    {
+      "epoch": 1.519683129900574,
+      "grad_norm": 0.6838982701301575,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5212998140813192,
+      "grad_norm": 0.8599441647529602,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5229164982620644,
+      "grad_norm": 0.7020329833030701,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5245331824428097,
+      "grad_norm": 0.6964772343635559,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5261498666235551,
+      "grad_norm": 0.6916600465774536,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9440
+    },
+    {
+      "epoch": 1.5277665508043003,
+      "grad_norm": 0.7282621264457703,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5293832349850458,
+      "grad_norm": 0.5363983511924744,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 9460
+    },
+    {
+      "epoch": 1.530999919165791,
+      "grad_norm": 0.6184861063957214,
+      "learning_rate": 0.0002,
+      "loss": 0.7154,
+      "step": 9470
+    },
+    {
+      "epoch": 1.5326166033465363,
+      "grad_norm": 0.5991285443305969,
+      "learning_rate": 0.0002,
+      "loss": 0.7837,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5342332875272815,
+      "grad_norm": 0.8176587820053101,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5358499717080267,
+      "grad_norm": 0.6473721861839294,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 9500
+    },
+    {
+      "epoch": 1.5374666558887722,
+      "grad_norm": 0.7319952845573425,
+      "learning_rate": 0.0002,
+      "loss": 0.7632,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5390833400695174,
+      "grad_norm": 0.702900230884552,
+      "learning_rate": 0.0002,
+      "loss": 0.7706,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5407000242502629,
+      "grad_norm": 0.7971600294113159,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 9530
+    },
+    {
+      "epoch": 1.542316708431008,
+      "grad_norm": 0.6527525186538696,
+      "learning_rate": 0.0002,
+      "loss": 0.7352,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5439333926117533,
+      "grad_norm": 0.5791676044464111,
+      "learning_rate": 0.0002,
+      "loss": 0.7425,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5455500767924986,
+      "grad_norm": 0.5619390606880188,
+      "learning_rate": 0.0002,
+      "loss": 0.7585,
+      "step": 9560
+    },
+    {
+      "epoch": 1.5471667609732438,
+      "grad_norm": 0.5701689124107361,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9570
+    },
+    {
+      "epoch": 1.548783445153989,
+      "grad_norm": 0.47549352049827576,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5504001293347345,
+      "grad_norm": 0.8730611205101013,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5520168135154797,
+      "grad_norm": 0.6842091083526611,
+      "learning_rate": 0.0002,
+      "loss": 0.798,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5536334976962252,
+      "grad_norm": 0.6675129532814026,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5552501818769704,
+      "grad_norm": 0.8173956274986267,
+      "learning_rate": 0.0002,
+      "loss": 0.7954,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5568668660577156,
+      "grad_norm": 0.724947452545166,
+      "learning_rate": 0.0002,
+      "loss": 0.7535,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5584835502384609,
+      "grad_norm": 0.6154758930206299,
+      "learning_rate": 0.0002,
+      "loss": 0.7738,
+      "step": 9640
+    },
+    {
+      "epoch": 1.560100234419206,
+      "grad_norm": 0.6072008013725281,
+      "learning_rate": 0.0002,
+      "loss": 0.7568,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5617169185999515,
+      "grad_norm": 0.659010648727417,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 9660
+    },
+    {
+      "epoch": 1.5633336027806968,
+      "grad_norm": 0.65857994556427,
+      "learning_rate": 0.0002,
+      "loss": 0.673,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5649502869614422,
+      "grad_norm": 0.5914267301559448,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5665669711421875,
+      "grad_norm": 0.6248020529747009,
+      "learning_rate": 0.0002,
+      "loss": 0.7414,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5681836553229327,
+      "grad_norm": 0.7147795557975769,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 9700
+    },
+    {
+      "epoch": 1.569800339503678,
+      "grad_norm": 0.7076232433319092,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 9710
+    },
+    {
+      "epoch": 1.5714170236844232,
+      "grad_norm": 0.6217400431632996,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5730337078651684,
+      "grad_norm": 0.6709911227226257,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 9730
+    },
+    {
+      "epoch": 1.5746503920459138,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 9740
+    },
+    {
+      "epoch": 1.576267076226659,
+      "grad_norm": 0.6241145730018616,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 9750
+    },
+    {
+      "epoch": 1.5778837604074045,
+      "grad_norm": 0.4960934817790985,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 9760
+    },
+    {
+      "epoch": 1.5795004445881498,
+      "grad_norm": 0.6593309640884399,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 9770
+    },
+    {
+      "epoch": 1.581117128768895,
+      "grad_norm": 0.5814042091369629,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 9780
+    },
+    {
+      "epoch": 1.5827338129496402,
+      "grad_norm": 0.5936070680618286,
+      "learning_rate": 0.0002,
+      "loss": 0.7109,
+      "step": 9790
+    },
+    {
+      "epoch": 1.5843504971303854,
+      "grad_norm": 0.6454403400421143,
+      "learning_rate": 0.0002,
+      "loss": 0.7769,
+      "step": 9800
+    },
+    {
+      "epoch": 1.585967181311131,
+      "grad_norm": 0.7612107992172241,
+      "learning_rate": 0.0002,
+      "loss": 0.7677,
+      "step": 9810
+    },
+    {
+      "epoch": 1.5875838654918761,
+      "grad_norm": 0.6494482755661011,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5892005496726216,
+      "grad_norm": 0.7825694680213928,
+      "learning_rate": 0.0002,
+      "loss": 0.7569,
+      "step": 9830
+    },
+    {
+      "epoch": 1.5908172338533668,
+      "grad_norm": 0.6757757663726807,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 9840
+    },
+    {
+      "epoch": 1.592433918034112,
+      "grad_norm": 0.7105609178543091,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 9850
+    },
+    {
+      "epoch": 1.5940506022148573,
+      "grad_norm": 0.7596991062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 9860
+    },
+    {
+      "epoch": 1.5956672863956025,
+      "grad_norm": 0.5681525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7108,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5972839705763477,
+      "grad_norm": 0.6090980768203735,
+      "learning_rate": 0.0002,
+      "loss": 0.7811,
+      "step": 9880
+    },
+    {
+      "epoch": 1.5989006547570932,
+      "grad_norm": 0.6271613240242004,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6005173389378387,
+      "grad_norm": 0.7656369805335999,
+      "learning_rate": 0.0002,
+      "loss": 0.7419,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6021340231185839,
+      "grad_norm": 0.7504446506500244,
+      "learning_rate": 0.0002,
+      "loss": 0.7336,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6037507072993291,
+      "grad_norm": 0.659656286239624,
+      "learning_rate": 0.0002,
+      "loss": 0.7479,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6053673914800743,
+      "grad_norm": 0.6006826162338257,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6069840756608196,
+      "grad_norm": 0.7872757911682129,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6086007598415648,
+      "grad_norm": 0.5545852780342102,
+      "learning_rate": 0.0002,
+      "loss": 0.768,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6102174440223103,
+      "grad_norm": 0.7429468631744385,
+      "learning_rate": 0.0002,
+      "loss": 0.8064,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6118341282030555,
+      "grad_norm": 0.6873556971549988,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 9970
+    },
+    {
+      "epoch": 1.613450812383801,
+      "grad_norm": 0.5874287486076355,
+      "learning_rate": 0.0002,
+      "loss": 0.7324,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6150674965645462,
+      "grad_norm": 0.6039386987686157,
+      "learning_rate": 0.0002,
+      "loss": 0.7141,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6166841807452914,
+      "grad_norm": 0.6233575940132141,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6183008649260366,
+      "grad_norm": 0.7676448225975037,
+      "learning_rate": 0.0002,
+      "loss": 0.7602,
+      "step": 10010
+    },
+    {
+      "epoch": 1.6199175491067819,
+      "grad_norm": 0.6565698385238647,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6215342332875273,
+      "grad_norm": 0.6787590384483337,
+      "learning_rate": 0.0002,
+      "loss": 0.7104,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6231509174682726,
+      "grad_norm": 0.6137678027153015,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 10040
+    },
+    {
+      "epoch": 1.624767601649018,
+      "grad_norm": 0.5236800312995911,
+      "learning_rate": 0.0002,
+      "loss": 0.7646,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6263842858297632,
+      "grad_norm": 0.7626367807388306,
+      "learning_rate": 0.0002,
+      "loss": 0.7437,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6280009700105085,
+      "grad_norm": 0.5657260417938232,
+      "learning_rate": 0.0002,
+      "loss": 0.7273,
+      "step": 10070
+    },
+    {
+      "epoch": 1.6296176541912537,
+      "grad_norm": 0.4913991391658783,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 10080
+    },
+    {
+      "epoch": 1.631234338371999,
+      "grad_norm": 0.7715556621551514,
+      "learning_rate": 0.0002,
+      "loss": 0.7596,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6328510225527442,
+      "grad_norm": 0.6509000062942505,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6344677067334896,
+      "grad_norm": 0.6215850114822388,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6360843909142349,
+      "grad_norm": 0.6956844329833984,
+      "learning_rate": 0.0002,
+      "loss": 0.7705,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6377010750949803,
+      "grad_norm": 0.6111597418785095,
+      "learning_rate": 0.0002,
+      "loss": 0.7129,
+      "step": 10130
+    },
+    {
+      "epoch": 1.6393177592757255,
+      "grad_norm": 0.6518288850784302,
+      "learning_rate": 0.0002,
+      "loss": 0.6955,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6409344434564708,
+      "grad_norm": 0.6914522051811218,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10150
+    },
+    {
+      "epoch": 1.642551127637216,
+      "grad_norm": 0.63785719871521,
+      "learning_rate": 0.0002,
+      "loss": 0.7295,
+      "step": 10160
+    },
+    {
+      "epoch": 1.6441678118179612,
+      "grad_norm": 0.6379287838935852,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6457844959987067,
+      "grad_norm": 0.6793403029441833,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 10180
+    },
+    {
+      "epoch": 1.647401180179452,
+      "grad_norm": 0.6099132895469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6490178643601974,
+      "grad_norm": 0.5869854092597961,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6506345485409426,
+      "grad_norm": 0.7716999053955078,
+      "learning_rate": 0.0002,
+      "loss": 0.8308,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6522512327216878,
+      "grad_norm": 0.6854110360145569,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 10220
+    },
+    {
+      "epoch": 1.653867916902433,
+      "grad_norm": 0.6957170367240906,
+      "learning_rate": 0.0002,
+      "loss": 0.782,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6554846010831783,
+      "grad_norm": 0.6932903528213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6571012852639235,
+      "grad_norm": 0.7713165283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10250
+    },
+    {
+      "epoch": 1.658717969444669,
+      "grad_norm": 0.7455793619155884,
+      "learning_rate": 0.0002,
+      "loss": 0.7099,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6603346536254144,
+      "grad_norm": 0.5464168190956116,
+      "learning_rate": 0.0002,
+      "loss": 0.7524,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6619513378061597,
+      "grad_norm": 0.6782926321029663,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 10280
+    },
+    {
+      "epoch": 1.663568021986905,
+      "grad_norm": 0.7962649464607239,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6651847061676501,
+      "grad_norm": 0.6814526319503784,
+      "learning_rate": 0.0002,
+      "loss": 0.7142,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6668013903483954,
+      "grad_norm": 0.656895101070404,
+      "learning_rate": 0.0002,
+      "loss": 0.7285,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6684180745291406,
+      "grad_norm": 0.6085672378540039,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 10320
+    },
+    {
+      "epoch": 1.670034758709886,
+      "grad_norm": 0.585508406162262,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 10330
+    },
+    {
+      "epoch": 1.6716514428906313,
+      "grad_norm": 0.6930184364318848,
+      "learning_rate": 0.0002,
+      "loss": 0.7604,
+      "step": 10340
+    },
+    {
+      "epoch": 1.6732681270713767,
+      "grad_norm": 0.575663149356842,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10350
+    },
+    {
+      "epoch": 1.674884811252122,
+      "grad_norm": 0.582502543926239,
+      "learning_rate": 0.0002,
+      "loss": 0.7198,
+      "step": 10360
+    },
+    {
+      "epoch": 1.6765014954328672,
+      "grad_norm": 0.5668916702270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 10370
+    },
+    {
+      "epoch": 1.6781181796136124,
+      "grad_norm": 0.6070065498352051,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10380
+    },
+    {
+      "epoch": 1.6797348637943577,
+      "grad_norm": 0.6141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.7939,
+      "step": 10390
+    },
+    {
+      "epoch": 1.6813515479751031,
+      "grad_norm": 0.8359124064445496,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 10400
+    },
+    {
+      "epoch": 1.6829682321558483,
+      "grad_norm": 0.5378185510635376,
+      "learning_rate": 0.0002,
+      "loss": 0.7488,
+      "step": 10410
+    },
+    {
+      "epoch": 1.6845849163365938,
+      "grad_norm": 0.6959536075592041,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 10420
+    },
+    {
+      "epoch": 1.686201600517339,
+      "grad_norm": 0.6514357328414917,
+      "learning_rate": 0.0002,
+      "loss": 0.7872,
+      "step": 10430
+    },
+    {
+      "epoch": 1.6878182846980843,
+      "grad_norm": 0.7706646919250488,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10440
+    },
+    {
+      "epoch": 1.6894349688788295,
+      "grad_norm": 0.6183337569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.7673,
+      "step": 10450
+    },
+    {
+      "epoch": 1.6910516530595747,
+      "grad_norm": 0.6123278141021729,
+      "learning_rate": 0.0002,
+      "loss": 0.7566,
+      "step": 10460
+    },
+    {
+      "epoch": 1.69266833724032,
+      "grad_norm": 0.6894851326942444,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10470
+    },
+    {
+      "epoch": 1.6942850214210654,
+      "grad_norm": 0.7497312426567078,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 10480
+    },
+    {
+      "epoch": 1.6959017056018106,
+      "grad_norm": 0.5968214273452759,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10490
+    },
+    {
+      "epoch": 1.697518389782556,
+      "grad_norm": 0.6747927069664001,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 10500
+    },
+    {
+      "epoch": 1.6991350739633013,
+      "grad_norm": 0.5708310008049011,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7007517581440466,
+      "grad_norm": 0.606526792049408,
+      "learning_rate": 0.0002,
+      "loss": 0.7385,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7023684423247918,
+      "grad_norm": 0.662011981010437,
+      "learning_rate": 0.0002,
+      "loss": 0.7204,
+      "step": 10530
+    },
+    {
+      "epoch": 1.703985126505537,
+      "grad_norm": 0.7583045363426208,
+      "learning_rate": 0.0002,
+      "loss": 0.7999,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7056018106862825,
+      "grad_norm": 0.721632182598114,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7072184948670277,
+      "grad_norm": 0.6107715368270874,
+      "learning_rate": 0.0002,
+      "loss": 0.7407,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7088351790477732,
+      "grad_norm": 0.6652471423149109,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7104518632285184,
+      "grad_norm": 0.6308087110519409,
+      "learning_rate": 0.0002,
+      "loss": 0.7767,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7120685474092636,
+      "grad_norm": 0.5464386940002441,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7136852315900089,
+      "grad_norm": 0.6558911204338074,
+      "learning_rate": 0.0002,
+      "loss": 0.7063,
+      "step": 10600
+    },
+    {
+      "epoch": 1.715301915770754,
+      "grad_norm": 0.5665024518966675,
+      "learning_rate": 0.0002,
+      "loss": 0.7126,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7169185999514993,
+      "grad_norm": 0.7888094186782837,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7185352841322448,
+      "grad_norm": 0.7084909081459045,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 10630
+    },
+    {
+      "epoch": 1.7201519683129902,
+      "grad_norm": 0.7982324361801147,
+      "learning_rate": 0.0002,
+      "loss": 0.7557,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7217686524937355,
+      "grad_norm": 0.6418732404708862,
+      "learning_rate": 0.0002,
+      "loss": 0.7345,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7233853366744807,
+      "grad_norm": 0.7636681795120239,
+      "learning_rate": 0.0002,
+      "loss": 0.7734,
+      "step": 10660
+    },
+    {
+      "epoch": 1.725002020855226,
+      "grad_norm": 0.5646875500679016,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7266187050359711,
+      "grad_norm": 0.5231260657310486,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7282353892167164,
+      "grad_norm": 0.7635011672973633,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7298520733974618,
+      "grad_norm": 0.7518259286880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 10700
+    },
+    {
+      "epoch": 1.731468757578207,
+      "grad_norm": 0.7295602560043335,
+      "learning_rate": 0.0002,
+      "loss": 0.751,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7330854417589525,
+      "grad_norm": 0.6984632015228271,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7347021259396977,
+      "grad_norm": 0.6198219060897827,
+      "learning_rate": 0.0002,
+      "loss": 0.7921,
+      "step": 10730
+    },
+    {
+      "epoch": 1.736318810120443,
+      "grad_norm": 0.6957576274871826,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7379354943011882,
+      "grad_norm": 0.6430263519287109,
+      "learning_rate": 0.0002,
+      "loss": 0.7917,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7395521784819334,
+      "grad_norm": 0.6134995222091675,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 10760
+    },
+    {
+      "epoch": 1.741168862662679,
+      "grad_norm": 0.7209452986717224,
+      "learning_rate": 0.0002,
+      "loss": 0.7584,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7427855468434241,
+      "grad_norm": 0.6735447645187378,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7444022310241696,
+      "grad_norm": 0.5605693459510803,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 10790
+    },
+    {
+      "epoch": 1.7460189152049148,
+      "grad_norm": 0.6882363557815552,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 10800
+    },
+    {
+      "epoch": 1.74763559938566,
+      "grad_norm": 0.6386259198188782,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7492522835664053,
+      "grad_norm": 0.6529015302658081,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7508689677471505,
+      "grad_norm": 0.5664082765579224,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7524856519278957,
+      "grad_norm": 0.7532684206962585,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7541023361086412,
+      "grad_norm": 0.77171391248703,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7557190202893864,
+      "grad_norm": 0.7255431413650513,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7573357044701319,
+      "grad_norm": 0.763083279132843,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10870
+    },
+    {
+      "epoch": 1.758952388650877,
+      "grad_norm": 0.6042402982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7605690728316223,
+      "grad_norm": 0.7642518281936646,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7621857570123676,
+      "grad_norm": 0.6347904801368713,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7638024411931128,
+      "grad_norm": 0.5371627807617188,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7654191253738583,
+      "grad_norm": 0.6840225458145142,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7670358095546035,
+      "grad_norm": 0.5288469195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10930
+    },
+    {
+      "epoch": 1.768652493735349,
+      "grad_norm": 0.69020676612854,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 10940
+    },
+    {
+      "epoch": 1.7702691779160942,
+      "grad_norm": 0.5943242311477661,
+      "learning_rate": 0.0002,
+      "loss": 0.7468,
+      "step": 10950
+    },
+    {
+      "epoch": 1.7718858620968394,
+      "grad_norm": 0.5616418123245239,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 10960
+    },
+    {
+      "epoch": 1.7735025462775846,
+      "grad_norm": 0.7209470868110657,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 10970
+    },
+    {
+      "epoch": 1.7751192304583299,
+      "grad_norm": 0.6657957434654236,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 10980
+    },
+    {
+      "epoch": 1.776735914639075,
+      "grad_norm": 0.6469064950942993,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 10990
+    },
+    {
+      "epoch": 1.7783525988198206,
+      "grad_norm": 0.6615678071975708,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 11000
+    },
+    {
+      "epoch": 1.779969283000566,
+      "grad_norm": 0.6722439527511597,
+      "learning_rate": 0.0002,
+      "loss": 0.747,
+      "step": 11010
+    },
+    {
+      "epoch": 1.7815859671813112,
+      "grad_norm": 0.634136974811554,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 11020
+    },
+    {
+      "epoch": 1.7832026513620565,
+      "grad_norm": 0.6024377346038818,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 11030
+    },
+    {
+      "epoch": 1.7848193355428017,
+      "grad_norm": 0.6909403800964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 11040
+    },
+    {
+      "epoch": 1.786436019723547,
+      "grad_norm": 0.7148767709732056,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 11050
+    },
+    {
+      "epoch": 1.7880527039042922,
+      "grad_norm": 0.7442979216575623,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 11060
+    },
+    {
+      "epoch": 1.7896693880850376,
+      "grad_norm": 0.6830431818962097,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 11070
+    },
+    {
+      "epoch": 1.7912860722657828,
+      "grad_norm": 0.9172667264938354,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 11080
+    },
+    {
+      "epoch": 1.7929027564465283,
+      "grad_norm": 0.6799490451812744,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 11090
+    },
+    {
+      "epoch": 1.7945194406272735,
+      "grad_norm": 0.7617024779319763,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 11100
+    },
+    {
+      "epoch": 1.7961361248080188,
+      "grad_norm": 0.7701810002326965,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 11110
+    },
+    {
+      "epoch": 1.797752808988764,
+      "grad_norm": 0.7454385757446289,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11120
+    },
+    {
+      "epoch": 1.7993694931695092,
+      "grad_norm": 0.6121436953544617,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8009861773502547,
+      "grad_norm": 0.6237571835517883,
+      "learning_rate": 0.0002,
+      "loss": 0.7305,
+      "step": 11140
+    },
+    {
+      "epoch": 1.802602861531,
+      "grad_norm": 0.6818515658378601,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 11150
+    },
+    {
+      "epoch": 1.8042195457117454,
+      "grad_norm": 0.7768308520317078,
+      "learning_rate": 0.0002,
+      "loss": 0.6876,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8058362298924906,
+      "grad_norm": 0.6875537633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 11170
+    },
+    {
+      "epoch": 1.8074529140732358,
+      "grad_norm": 0.7950584888458252,
+      "learning_rate": 0.0002,
+      "loss": 0.761,
+      "step": 11180
+    },
+    {
+      "epoch": 1.809069598253981,
+      "grad_norm": 0.8210248351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8106862824347263,
+      "grad_norm": 0.6674110889434814,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8123029666154715,
+      "grad_norm": 0.6261674761772156,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 11210
+    },
+    {
+      "epoch": 1.813919650796217,
+      "grad_norm": 0.6484741568565369,
+      "learning_rate": 0.0002,
+      "loss": 0.7122,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8155363349769622,
+      "grad_norm": 0.6231244206428528,
+      "learning_rate": 0.0002,
+      "loss": 0.7718,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8171530191577077,
+      "grad_norm": 0.7243146896362305,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11240
+    },
+    {
+      "epoch": 1.818769703338453,
+      "grad_norm": 0.6776193380355835,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8203863875191981,
+      "grad_norm": 0.5973618030548096,
+      "learning_rate": 0.0002,
+      "loss": 0.7317,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8220030716999434,
+      "grad_norm": 0.6451361179351807,
+      "learning_rate": 0.0002,
+      "loss": 0.7961,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8236197558806886,
+      "grad_norm": 0.5963068008422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 11280
+    },
+    {
+      "epoch": 1.825236440061434,
+      "grad_norm": 0.536902129650116,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8268531242421793,
+      "grad_norm": 0.6993787288665771,
+      "learning_rate": 0.0002,
+      "loss": 0.708,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8284698084229247,
+      "grad_norm": 0.6135255098342896,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11310
+    },
+    {
+      "epoch": 1.83008649260367,
+      "grad_norm": 0.6057423949241638,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8317031767844152,
+      "grad_norm": 0.6598812341690063,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8333198609651604,
+      "grad_norm": 0.6075948476791382,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 11340
+    },
+    {
+      "epoch": 1.8349365451459057,
+      "grad_norm": 0.7065447568893433,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8365532293266509,
+      "grad_norm": 0.680526614189148,
+      "learning_rate": 0.0002,
+      "loss": 0.7365,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8381699135073963,
+      "grad_norm": 0.6356695294380188,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8397865976881416,
+      "grad_norm": 0.6399052143096924,
+      "learning_rate": 0.0002,
+      "loss": 0.721,
+      "step": 11380
+    },
+    {
+      "epoch": 1.841403281868887,
+      "grad_norm": 0.6125704050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 11390
+    },
+    {
+      "epoch": 1.8430199660496323,
+      "grad_norm": 0.7124643325805664,
+      "learning_rate": 0.0002,
+      "loss": 0.755,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8446366502303775,
+      "grad_norm": 0.6099604964256287,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8462533344111227,
+      "grad_norm": 0.7338208556175232,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 11420
+    },
+    {
+      "epoch": 1.847870018591868,
+      "grad_norm": 0.7534668445587158,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8494867027726134,
+      "grad_norm": 0.6135470271110535,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8511033869533586,
+      "grad_norm": 0.6229309439659119,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 11450
+    },
+    {
+      "epoch": 1.852720071134104,
+      "grad_norm": 0.706423282623291,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 11460
+    },
+    {
+      "epoch": 1.8543367553148493,
+      "grad_norm": 0.5460049510002136,
+      "learning_rate": 0.0002,
+      "loss": 0.7411,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8559534394955945,
+      "grad_norm": 0.6616711020469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7416,
+      "step": 11480
+    },
+    {
+      "epoch": 1.8575701236763398,
+      "grad_norm": 0.6372783184051514,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 11490
+    },
+    {
+      "epoch": 1.859186807857085,
+      "grad_norm": 0.7162668108940125,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8608034920378305,
+      "grad_norm": 0.6605209708213806,
+      "learning_rate": 0.0002,
+      "loss": 0.7747,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8624201762185757,
+      "grad_norm": 0.6933956742286682,
+      "learning_rate": 0.0002,
+      "loss": 0.7258,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8640368603993211,
+      "grad_norm": 0.6582090854644775,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8656535445800664,
+      "grad_norm": 0.6416500806808472,
+      "learning_rate": 0.0002,
+      "loss": 0.7313,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8672702287608116,
+      "grad_norm": 0.5434312224388123,
+      "learning_rate": 0.0002,
+      "loss": 0.7372,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8688869129415568,
+      "grad_norm": 0.6827567219734192,
+      "learning_rate": 0.0002,
+      "loss": 0.7635,
+      "step": 11560
+    },
+    {
+      "epoch": 1.870503597122302,
+      "grad_norm": 0.7354370951652527,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 11570
+    },
+    {
+      "epoch": 1.8721202813030473,
+      "grad_norm": 0.590372622013092,
+      "learning_rate": 0.0002,
+      "loss": 0.7526,
+      "step": 11580
+    },
+    {
+      "epoch": 1.8737369654837928,
+      "grad_norm": 0.853183925151825,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 11590
+    },
+    {
+      "epoch": 1.875353649664538,
+      "grad_norm": 0.822678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 11600
+    },
+    {
+      "epoch": 1.8769703338452834,
+      "grad_norm": 0.6591550707817078,
+      "learning_rate": 0.0002,
+      "loss": 0.7427,
+      "step": 11610
+    },
+    {
+      "epoch": 1.8785870180260287,
+      "grad_norm": 0.7475301623344421,
+      "learning_rate": 0.0002,
+      "loss": 0.7054,
+      "step": 11620
+    },
+    {
+      "epoch": 1.880203702206774,
+      "grad_norm": 0.6390765309333801,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 11630
+    },
+    {
+      "epoch": 1.8818203863875191,
+      "grad_norm": 0.6589758992195129,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 11640
+    },
+    {
+      "epoch": 1.8834370705682644,
+      "grad_norm": 0.6765508651733398,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 11650
+    },
+    {
+      "epoch": 1.8850537547490098,
+      "grad_norm": 0.6527857780456543,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 11660
+    },
+    {
+      "epoch": 1.886670438929755,
+      "grad_norm": 0.6642923951148987,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 11670
+    },
+    {
+      "epoch": 1.8882871231105005,
+      "grad_norm": 0.6945584416389465,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 11680
+    },
+    {
+      "epoch": 1.8899038072912457,
+      "grad_norm": 0.694018542766571,
+      "learning_rate": 0.0002,
+      "loss": 0.7711,
+      "step": 11690
+    },
+    {
+      "epoch": 1.891520491471991,
+      "grad_norm": 0.7237417101860046,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 11700
+    },
+    {
+      "epoch": 1.8931371756527362,
+      "grad_norm": 0.7401309609413147,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 11710
+    },
+    {
+      "epoch": 1.8947538598334814,
+      "grad_norm": 0.6537784337997437,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 11720
+    },
+    {
+      "epoch": 1.8963705440142267,
+      "grad_norm": 0.7398539185523987,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 11730
+    },
+    {
+      "epoch": 1.8979872281949721,
+      "grad_norm": 0.6696075797080994,
+      "learning_rate": 0.0002,
+      "loss": 0.7561,
+      "step": 11740
+    },
+    {
+      "epoch": 1.8996039123757174,
+      "grad_norm": 0.6014142036437988,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9012205965564628,
+      "grad_norm": 0.7023524641990662,
+      "learning_rate": 0.0002,
+      "loss": 0.7714,
+      "step": 11760
+    },
+    {
+      "epoch": 1.902837280737208,
+      "grad_norm": 0.739973783493042,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9044539649179533,
+      "grad_norm": 0.5576770901679993,
+      "learning_rate": 0.0002,
+      "loss": 0.7848,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9060706490986985,
+      "grad_norm": 0.6907393932342529,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 11790
+    },
+    {
+      "epoch": 1.9076873332794437,
+      "grad_norm": 0.6934581995010376,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9093040174601892,
+      "grad_norm": 0.591774582862854,
+      "learning_rate": 0.0002,
+      "loss": 0.7199,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9109207016409344,
+      "grad_norm": 0.6249791383743286,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11820
+    },
+    {
+      "epoch": 1.9125373858216799,
+      "grad_norm": 0.6755744218826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 11830
+    },
+    {
+      "epoch": 1.914154070002425,
+      "grad_norm": 0.7286285161972046,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9157707541831703,
+      "grad_norm": 0.7867850065231323,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9173874383639156,
+      "grad_norm": 0.6283972859382629,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11860
+    },
+    {
+      "epoch": 1.9190041225446608,
+      "grad_norm": 0.605823814868927,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 11870
+    },
+    {
+      "epoch": 1.920620806725406,
+      "grad_norm": 0.5927976965904236,
+      "learning_rate": 0.0002,
+      "loss": 0.6598,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9222374909061515,
+      "grad_norm": 0.5974002480506897,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 11890
+    },
+    {
+      "epoch": 1.923854175086897,
+      "grad_norm": 0.7091866135597229,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9254708592676422,
+      "grad_norm": 0.72496497631073,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9270875434483874,
+      "grad_norm": 0.6131896376609802,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9287042276291326,
+      "grad_norm": 0.6556436419487,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9303209118098779,
+      "grad_norm": 0.622932493686676,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 11940
+    },
+    {
+      "epoch": 1.931937595990623,
+      "grad_norm": 0.6618631482124329,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9335542801713685,
+      "grad_norm": 0.630966305732727,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 11960
+    },
+    {
+      "epoch": 1.9351709643521138,
+      "grad_norm": 0.6336734890937805,
+      "learning_rate": 0.0002,
+      "loss": 0.6854,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9367876485328592,
+      "grad_norm": 0.655403196811676,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9384043327136045,
+      "grad_norm": 0.5640574097633362,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 11990
+    },
+    {
+      "epoch": 1.9400210168943497,
+      "grad_norm": 0.6322951316833496,
+      "learning_rate": 0.0002,
+      "loss": 0.7289,
+      "step": 12000
+    },
+    {
+      "epoch": 1.941637701075095,
+      "grad_norm": 0.615703821182251,
+      "learning_rate": 0.0002,
+      "loss": 0.7627,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9432543852558402,
+      "grad_norm": 0.6487536430358887,
+      "learning_rate": 0.0002,
+      "loss": 0.786,
+      "step": 12020
+    },
+    {
+      "epoch": 1.9448710694365856,
+      "grad_norm": 0.9209630489349365,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 12030
+    },
+    {
+      "epoch": 1.9464877536173308,
+      "grad_norm": 0.67485511302948,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9481044377980763,
+      "grad_norm": 0.6831230521202087,
+      "learning_rate": 0.0002,
+      "loss": 0.7551,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9497211219788215,
+      "grad_norm": 0.6578302383422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7546,
+      "step": 12060
+    },
+    {
+      "epoch": 1.9513378061595668,
+      "grad_norm": 0.9975938200950623,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12070
+    },
+    {
+      "epoch": 1.952954490340312,
+      "grad_norm": 0.6637365221977234,
+      "learning_rate": 0.0002,
+      "loss": 0.7952,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9545711745210572,
+      "grad_norm": 0.605707049369812,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9561878587018025,
+      "grad_norm": 0.6584440469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7768,
+      "step": 12100
+    },
+    {
+      "epoch": 1.957804542882548,
+      "grad_norm": 0.6070835590362549,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9594212270632931,
+      "grad_norm": 0.7862601280212402,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9610379112440386,
+      "grad_norm": 0.8175255060195923,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 12130
+    },
+    {
+      "epoch": 1.9626545954247838,
+      "grad_norm": 0.5648472905158997,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 12140
+    },
+    {
+      "epoch": 1.964271279605529,
+      "grad_norm": 0.6591973304748535,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9658879637862743,
+      "grad_norm": 0.5960676074028015,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9675046479670195,
+      "grad_norm": 0.7272544503211975,
+      "learning_rate": 0.0002,
+      "loss": 0.7254,
+      "step": 12170
+    },
+    {
+      "epoch": 1.969121332147765,
+      "grad_norm": 0.7176699042320251,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9707380163285102,
+      "grad_norm": 0.6927123665809631,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9723547005092557,
+      "grad_norm": 0.5536034107208252,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 12200
+    },
+    {
+      "epoch": 1.9739713846900009,
+      "grad_norm": 0.8348390460014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 12210
+    },
+    {
+      "epoch": 1.9755880688707461,
+      "grad_norm": 0.6591181755065918,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 12220
+    },
+    {
+      "epoch": 1.9772047530514913,
+      "grad_norm": 1.0624109506607056,
+      "learning_rate": 0.0002,
+      "loss": 0.763,
+      "step": 12230
+    },
+    {
+      "epoch": 1.9788214372322366,
+      "grad_norm": 0.9265586137771606,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 12240
+    },
+    {
+      "epoch": 1.9804381214129818,
+      "grad_norm": 0.5998196005821228,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 12250
+    },
+    {
+      "epoch": 1.9820548055937273,
+      "grad_norm": 0.6960851550102234,
+      "learning_rate": 0.0002,
+      "loss": 0.7225,
+      "step": 12260
+    },
+    {
+      "epoch": 1.9836714897744727,
+      "grad_norm": 0.7674502730369568,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 12270
+    },
+    {
+      "epoch": 1.985288173955218,
+      "grad_norm": 0.6407275795936584,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 12280
+    },
+    {
+      "epoch": 1.9869048581359632,
+      "grad_norm": 0.6673079133033752,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 12290
+    },
+    {
+      "epoch": 1.9885215423167084,
+      "grad_norm": 0.6989844441413879,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 12300
+    },
+    {
+      "epoch": 1.9901382264974536,
+      "grad_norm": 0.7564442157745361,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 12310
+    },
+    {
+      "epoch": 1.9917549106781989,
+      "grad_norm": 0.6385478973388672,
+      "learning_rate": 0.0002,
+      "loss": 0.7719,
+      "step": 12320
+    },
+    {
+      "epoch": 1.9933715948589443,
+      "grad_norm": 0.7193717956542969,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 12330
+    },
+    {
+      "epoch": 1.9949882790396896,
+      "grad_norm": 0.7987112402915955,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 12340
+    },
+    {
+      "epoch": 1.996604963220435,
+      "grad_norm": 0.7260826826095581,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 12350
+    },
+    {
+      "epoch": 1.9982216474011802,
+      "grad_norm": 0.7968255281448364,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 12360
+    },
+    {
+      "epoch": 1.9998383315819255,
+      "grad_norm": 0.6893062591552734,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1044032573699951,
+      "eval_runtime": 122.1508,
+      "eval_samples_per_second": 6.001,
+      "eval_steps_per_second": 0.753,
+      "step": 12371
+    },
+    {
+      "epoch": 2.0014550157626707,
+      "grad_norm": 0.7775409817695618,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 12380
+    },
+    {
+      "epoch": 2.003071699943416,
+      "grad_norm": 0.76218581199646,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 12390
+    },
+    {
+      "epoch": 2.004688384124161,
+      "grad_norm": 0.5677764415740967,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 12400
+    },
+    {
+      "epoch": 2.006305068304907,
+      "grad_norm": 0.808442234992981,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 12410
+    },
+    {
+      "epoch": 2.007921752485652,
+      "grad_norm": 0.7144765257835388,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 12420
+    },
+    {
+      "epoch": 2.0095384366663973,
+      "grad_norm": 0.6914031505584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 12430
+    },
+    {
+      "epoch": 2.0111551208471425,
+      "grad_norm": 0.7581454515457153,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 12440
+    },
+    {
+      "epoch": 2.0127718050278878,
+      "grad_norm": 0.8388504981994629,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 12450
+    },
+    {
+      "epoch": 2.014388489208633,
+      "grad_norm": 0.6716406941413879,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 12460
+    },
+    {
+      "epoch": 2.0160051733893782,
+      "grad_norm": 0.898902416229248,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 12470
+    },
+    {
+      "epoch": 2.0176218575701235,
+      "grad_norm": 0.6432679891586304,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 12480
+    },
+    {
+      "epoch": 2.019238541750869,
+      "grad_norm": 0.8021109104156494,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12490
+    },
+    {
+      "epoch": 2.0208552259316144,
+      "grad_norm": 0.7039216756820679,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 12500
+    },
+    {
+      "epoch": 2.0224719101123596,
+      "grad_norm": 0.646531879901886,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12510
+    },
+    {
+      "epoch": 2.024088594293105,
+      "grad_norm": 0.783704400062561,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 12520
+    },
+    {
+      "epoch": 2.02570527847385,
+      "grad_norm": 0.8805046677589417,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12530
+    },
+    {
+      "epoch": 2.0273219626545953,
+      "grad_norm": 0.7289270758628845,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0289386468353405,
+      "grad_norm": 0.71653151512146,
+      "learning_rate": 0.0002,
+      "loss": 0.6663,
+      "step": 12550
+    },
+    {
+      "epoch": 2.030555331016086,
+      "grad_norm": 0.73281329870224,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 12560
+    },
+    {
+      "epoch": 2.0321720151968314,
+      "grad_norm": 0.6657090187072754,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 12570
+    },
+    {
+      "epoch": 2.0337886993775767,
+      "grad_norm": 0.8241133093833923,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 12580
+    },
+    {
+      "epoch": 2.035405383558322,
+      "grad_norm": 0.5834135413169861,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 12590
+    },
+    {
+      "epoch": 2.037022067739067,
+      "grad_norm": 0.84502112865448,
+      "learning_rate": 0.0002,
+      "loss": 0.6188,
+      "step": 12600
+    },
+    {
+      "epoch": 2.0386387519198124,
+      "grad_norm": 0.8952481746673584,
+      "learning_rate": 0.0002,
+      "loss": 0.6349,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0402554361005576,
+      "grad_norm": 0.7801461815834045,
+      "learning_rate": 0.0002,
+      "loss": 0.6923,
+      "step": 12620
+    },
+    {
+      "epoch": 2.041872120281303,
+      "grad_norm": 0.6788367033004761,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 12630
+    },
+    {
+      "epoch": 2.0434888044620485,
+      "grad_norm": 0.7241756319999695,
+      "learning_rate": 0.0002,
+      "loss": 0.6162,
+      "step": 12640
+    },
+    {
+      "epoch": 2.0451054886427937,
+      "grad_norm": 0.6933388113975525,
+      "learning_rate": 0.0002,
+      "loss": 0.655,
+      "step": 12650
+    },
+    {
+      "epoch": 2.046722172823539,
+      "grad_norm": 0.8029746413230896,
+      "learning_rate": 0.0002,
+      "loss": 0.6431,
+      "step": 12660
+    },
+    {
+      "epoch": 2.048338857004284,
+      "grad_norm": 0.946399986743927,
+      "learning_rate": 0.0002,
+      "loss": 0.7164,
+      "step": 12670
+    },
+    {
+      "epoch": 2.0499555411850294,
+      "grad_norm": 0.7072678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.638,
+      "step": 12680
+    },
+    {
+      "epoch": 2.0515722253657747,
+      "grad_norm": 0.6810618042945862,
+      "learning_rate": 0.0002,
+      "loss": 0.6487,
+      "step": 12690
+    },
+    {
+      "epoch": 2.05318890954652,
+      "grad_norm": 0.7661160230636597,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 12700
+    },
+    {
+      "epoch": 2.0548055937272656,
+      "grad_norm": 0.6350653767585754,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 12710
+    },
+    {
+      "epoch": 2.056422277908011,
+      "grad_norm": 0.861890971660614,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 12720
+    },
+    {
+      "epoch": 2.058038962088756,
+      "grad_norm": 0.6489875912666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 12730
+    },
+    {
+      "epoch": 2.0596556462695013,
+      "grad_norm": 0.8268506526947021,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 12740
+    },
+    {
+      "epoch": 2.0612723304502465,
+      "grad_norm": 0.607679545879364,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 12750
+    },
+    {
+      "epoch": 2.0628890146309917,
+      "grad_norm": 0.6754153370857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6649,
+      "step": 12760
+    },
+    {
+      "epoch": 2.064505698811737,
+      "grad_norm": 0.7263124585151672,
+      "learning_rate": 0.0002,
+      "loss": 0.6549,
+      "step": 12770
+    },
+    {
+      "epoch": 2.0661223829924826,
+      "grad_norm": 0.6986154317855835,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 12780
+    },
+    {
+      "epoch": 2.067739067173228,
+      "grad_norm": 0.7768576741218567,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 12790
+    },
+    {
+      "epoch": 2.069355751353973,
+      "grad_norm": 0.7546762824058533,
+      "learning_rate": 0.0002,
+      "loss": 0.677,
+      "step": 12800
+    },
+    {
+      "epoch": 2.0709724355347183,
+      "grad_norm": 0.7588880062103271,
+      "learning_rate": 0.0002,
+      "loss": 0.6485,
+      "step": 12810
+    },
+    {
+      "epoch": 2.0725891197154636,
+      "grad_norm": 0.7457242608070374,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12820
+    },
+    {
+      "epoch": 2.074205803896209,
+      "grad_norm": 0.6983516812324524,
+      "learning_rate": 0.0002,
+      "loss": 0.6489,
+      "step": 12830
+    },
+    {
+      "epoch": 2.075822488076954,
+      "grad_norm": 0.7950928807258606,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 12840
+    },
+    {
+      "epoch": 2.0774391722576993,
+      "grad_norm": 0.9248087406158447,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 12850
+    },
+    {
+      "epoch": 2.079055856438445,
+      "grad_norm": 0.7229493260383606,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 12860
+    },
+    {
+      "epoch": 2.08067254061919,
+      "grad_norm": 0.5710847973823547,
+      "learning_rate": 0.0002,
+      "loss": 0.6702,
+      "step": 12870
+    },
+    {
+      "epoch": 2.0822892247999354,
+      "grad_norm": 0.9580423831939697,
+      "learning_rate": 0.0002,
+      "loss": 0.6974,
+      "step": 12880
+    },
+    {
+      "epoch": 2.0839059089806806,
+      "grad_norm": 0.7399665713310242,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 12890
+    },
+    {
+      "epoch": 2.085522593161426,
+      "grad_norm": 0.7981410622596741,
+      "learning_rate": 0.0002,
+      "loss": 0.6993,
+      "step": 12900
+    },
+    {
+      "epoch": 2.087139277342171,
+      "grad_norm": 0.870759904384613,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 12910
+    },
+    {
+      "epoch": 2.0887559615229163,
+      "grad_norm": 0.7001481652259827,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 12920
+    },
+    {
+      "epoch": 2.090372645703662,
+      "grad_norm": 0.6745418310165405,
+      "learning_rate": 0.0002,
+      "loss": 0.6383,
+      "step": 12930
+    },
+    {
+      "epoch": 2.0919893298844072,
+      "grad_norm": 0.7739067673683167,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 12940
+    },
+    {
+      "epoch": 2.0936060140651525,
+      "grad_norm": 0.6742934584617615,
+      "learning_rate": 0.0002,
+      "loss": 0.6856,
+      "step": 12950
+    },
+    {
+      "epoch": 2.0952226982458977,
+      "grad_norm": 0.7270349860191345,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 12960
+    },
+    {
+      "epoch": 2.096839382426643,
+      "grad_norm": 0.7150624394416809,
+      "learning_rate": 0.0002,
+      "loss": 0.6783,
+      "step": 12970
+    },
+    {
+      "epoch": 2.098456066607388,
+      "grad_norm": 0.7734767198562622,
+      "learning_rate": 0.0002,
+      "loss": 0.6093,
+      "step": 12980
+    },
+    {
+      "epoch": 2.1000727507881334,
+      "grad_norm": 0.7618662118911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 12990
+    },
+    {
+      "epoch": 2.101689434968879,
+      "grad_norm": 0.6557944416999817,
+      "learning_rate": 0.0002,
+      "loss": 0.6707,
+      "step": 13000
+    },
+    {
+      "epoch": 2.1033061191496243,
+      "grad_norm": 0.8786448240280151,
+      "learning_rate": 0.0002,
+      "loss": 0.7268,
+      "step": 13010
+    },
+    {
+      "epoch": 2.1049228033303695,
+      "grad_norm": 0.6878724098205566,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 13020
+    },
+    {
+      "epoch": 2.1065394875111147,
+      "grad_norm": 0.822318971157074,
+      "learning_rate": 0.0002,
+      "loss": 0.6824,
+      "step": 13030
+    },
+    {
+      "epoch": 2.10815617169186,
+      "grad_norm": 0.831468939781189,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 13040
+    },
+    {
+      "epoch": 2.109772855872605,
+      "grad_norm": 0.7699505686759949,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 13050
+    },
+    {
+      "epoch": 2.1113895400533504,
+      "grad_norm": 0.7559016346931458,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1130062242340957,
+      "grad_norm": 0.6942209601402283,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 13070
+    },
+    {
+      "epoch": 2.1146229084148414,
+      "grad_norm": 0.6098947525024414,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 13080
+    },
+    {
+      "epoch": 2.1162395925955866,
+      "grad_norm": 0.6499016284942627,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 13090
+    },
+    {
+      "epoch": 2.117856276776332,
+      "grad_norm": 0.7719953060150146,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 13100
+    },
+    {
+      "epoch": 2.119472960957077,
+      "grad_norm": 0.6708134412765503,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1210896451378223,
+      "grad_norm": 0.8119585514068604,
+      "learning_rate": 0.0002,
+      "loss": 0.6056,
+      "step": 13120
+    },
+    {
+      "epoch": 2.1227063293185675,
+      "grad_norm": 0.6947157979011536,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 13130
+    },
+    {
+      "epoch": 2.1243230134993127,
+      "grad_norm": 0.8831837773323059,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 13140
+    },
+    {
+      "epoch": 2.1259396976800584,
+      "grad_norm": 0.7266910672187805,
+      "learning_rate": 0.0002,
+      "loss": 0.6997,
+      "step": 13150
+    },
+    {
+      "epoch": 2.1275563818608036,
+      "grad_norm": 0.8864351511001587,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 13160
+    },
+    {
+      "epoch": 2.129173066041549,
+      "grad_norm": 0.8104248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.6762,
+      "step": 13170
+    },
+    {
+      "epoch": 2.130789750222294,
+      "grad_norm": 0.6077079772949219,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 13180
+    },
+    {
+      "epoch": 2.1324064344030393,
+      "grad_norm": 0.6874213814735413,
+      "learning_rate": 0.0002,
+      "loss": 0.6572,
+      "step": 13190
+    },
+    {
+      "epoch": 2.1340231185837846,
+      "grad_norm": 0.7134367823600769,
+      "learning_rate": 0.0002,
+      "loss": 0.642,
+      "step": 13200
+    },
+    {
+      "epoch": 2.13563980276453,
+      "grad_norm": 0.6101235151290894,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 13210
+    },
+    {
+      "epoch": 2.137256486945275,
+      "grad_norm": 0.6042411923408508,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13220
+    },
+    {
+      "epoch": 2.1388731711260207,
+      "grad_norm": 0.914601743221283,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 13230
+    },
+    {
+      "epoch": 2.140489855306766,
+      "grad_norm": 0.7104284167289734,
+      "learning_rate": 0.0002,
+      "loss": 0.6513,
+      "step": 13240
+    },
+    {
+      "epoch": 2.142106539487511,
+      "grad_norm": 0.664395272731781,
+      "learning_rate": 0.0002,
+      "loss": 0.6607,
+      "step": 13250
+    },
+    {
+      "epoch": 2.1437232236682564,
+      "grad_norm": 0.6991241574287415,
+      "learning_rate": 0.0002,
+      "loss": 0.7211,
+      "step": 13260
+    },
+    {
+      "epoch": 2.1453399078490016,
+      "grad_norm": 0.5469560623168945,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 13270
+    },
+    {
+      "epoch": 2.146956592029747,
+      "grad_norm": 0.8454998135566711,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13280
+    },
+    {
+      "epoch": 2.148573276210492,
+      "grad_norm": 0.7088868618011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 13290
+    },
+    {
+      "epoch": 2.1501899603912378,
+      "grad_norm": 0.7002687454223633,
+      "learning_rate": 0.0002,
+      "loss": 0.6835,
+      "step": 13300
+    },
+    {
+      "epoch": 2.151806644571983,
+      "grad_norm": 0.7785214781761169,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 13310
+    },
+    {
+      "epoch": 2.1534233287527282,
+      "grad_norm": 0.8049132227897644,
+      "learning_rate": 0.0002,
+      "loss": 0.67,
+      "step": 13320
+    },
+    {
+      "epoch": 2.1550400129334735,
+      "grad_norm": 0.8062595129013062,
+      "learning_rate": 0.0002,
+      "loss": 0.6495,
+      "step": 13330
+    },
+    {
+      "epoch": 2.1566566971142187,
+      "grad_norm": 0.6208319067955017,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 13340
+    },
+    {
+      "epoch": 2.158273381294964,
+      "grad_norm": 0.7519655823707581,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 13350
+    },
+    {
+      "epoch": 2.159890065475709,
+      "grad_norm": 0.7645747065544128,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 13360
+    },
+    {
+      "epoch": 2.1615067496564544,
+      "grad_norm": 0.6847302913665771,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 13370
+    },
+    {
+      "epoch": 2.1631234338372,
+      "grad_norm": 0.8630441427230835,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 13380
+    },
+    {
+      "epoch": 2.1647401180179453,
+      "grad_norm": 0.7947702407836914,
+      "learning_rate": 0.0002,
+      "loss": 0.6742,
+      "step": 13390
+    },
+    {
+      "epoch": 2.1663568021986905,
+      "grad_norm": 0.6836977005004883,
+      "learning_rate": 0.0002,
+      "loss": 0.7206,
+      "step": 13400
+    },
+    {
+      "epoch": 2.1679734863794358,
+      "grad_norm": 0.7340566515922546,
+      "learning_rate": 0.0002,
+      "loss": 0.6304,
+      "step": 13410
+    },
+    {
+      "epoch": 2.169590170560181,
+      "grad_norm": 0.7075738906860352,
+      "learning_rate": 0.0002,
+      "loss": 0.6528,
+      "step": 13420
+    },
+    {
+      "epoch": 2.1712068547409262,
+      "grad_norm": 0.7080879807472229,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 13430
+    },
+    {
+      "epoch": 2.1728235389216715,
+      "grad_norm": 0.6218613386154175,
+      "learning_rate": 0.0002,
+      "loss": 0.6615,
+      "step": 13440
+    },
+    {
+      "epoch": 2.174440223102417,
+      "grad_norm": 0.8211479187011719,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 13450
+    },
+    {
+      "epoch": 2.1760569072831624,
+      "grad_norm": 0.864466667175293,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 13460
+    },
+    {
+      "epoch": 2.1776735914639076,
+      "grad_norm": 0.7943857908248901,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 13470
+    },
+    {
+      "epoch": 2.179290275644653,
+      "grad_norm": 0.78728187084198,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 13480
+    },
+    {
+      "epoch": 2.180906959825398,
+      "grad_norm": 0.697527289390564,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 13490
+    },
+    {
+      "epoch": 2.1825236440061433,
+      "grad_norm": 0.8205804228782654,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 13500
+    },
+    {
+      "epoch": 2.1841403281868885,
+      "grad_norm": 0.8709042072296143,
+      "learning_rate": 0.0002,
+      "loss": 0.7227,
+      "step": 13510
+    },
+    {
+      "epoch": 2.1857570123676338,
+      "grad_norm": 0.6228537559509277,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 13520
+    },
+    {
+      "epoch": 2.1873736965483794,
+      "grad_norm": 0.9566980004310608,
+      "learning_rate": 0.0002,
+      "loss": 0.7025,
+      "step": 13530
+    },
+    {
+      "epoch": 2.1889903807291247,
+      "grad_norm": 0.7128894329071045,
+      "learning_rate": 0.0002,
+      "loss": 0.6755,
+      "step": 13540
+    },
+    {
+      "epoch": 2.19060706490987,
+      "grad_norm": 0.6888654232025146,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 13550
+    },
+    {
+      "epoch": 2.192223749090615,
+      "grad_norm": 0.6444337368011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6961,
+      "step": 13560
+    },
+    {
+      "epoch": 2.1938404332713604,
+      "grad_norm": 0.8008806705474854,
+      "learning_rate": 0.0002,
+      "loss": 0.656,
+      "step": 13570
+    },
+    {
+      "epoch": 2.1954571174521056,
+      "grad_norm": 0.8482748866081238,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 13580
+    },
+    {
+      "epoch": 2.197073801632851,
+      "grad_norm": 0.8584157228469849,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 13590
+    },
+    {
+      "epoch": 2.1986904858135965,
+      "grad_norm": 0.7513734698295593,
+      "learning_rate": 0.0002,
+      "loss": 0.7014,
+      "step": 13600
+    },
+    {
+      "epoch": 2.2003071699943417,
+      "grad_norm": 0.7864262461662292,
+      "learning_rate": 0.0002,
+      "loss": 0.6632,
+      "step": 13610
+    },
+    {
+      "epoch": 2.201923854175087,
+      "grad_norm": 0.8493645191192627,
+      "learning_rate": 0.0002,
+      "loss": 0.6879,
+      "step": 13620
+    },
+    {
+      "epoch": 2.203540538355832,
+      "grad_norm": 0.6902140974998474,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 13630
+    },
+    {
+      "epoch": 2.2051572225365774,
+      "grad_norm": 0.8711254596710205,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 13640
+    },
+    {
+      "epoch": 2.2067739067173227,
+      "grad_norm": 0.7832191586494446,
+      "learning_rate": 0.0002,
+      "loss": 0.6359,
+      "step": 13650
+    },
+    {
+      "epoch": 2.208390590898068,
+      "grad_norm": 0.5668176412582397,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2100072750788136,
+      "grad_norm": 0.8648375272750854,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13670
+    },
+    {
+      "epoch": 2.211623959259559,
+      "grad_norm": 0.7643089890480042,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13680
+    },
+    {
+      "epoch": 2.213240643440304,
+      "grad_norm": 0.6293777823448181,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13690
+    },
+    {
+      "epoch": 2.2148573276210493,
+      "grad_norm": 0.6459372639656067,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 13700
+    },
+    {
+      "epoch": 2.2164740118017945,
+      "grad_norm": 0.7060744166374207,
+      "learning_rate": 0.0002,
+      "loss": 0.6526,
+      "step": 13710
+    },
+    {
+      "epoch": 2.2180906959825397,
+      "grad_norm": 0.674109160900116,
+      "learning_rate": 0.0002,
+      "loss": 0.7101,
+      "step": 13720
+    },
+    {
+      "epoch": 2.219707380163285,
+      "grad_norm": 0.830392062664032,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13730
+    },
+    {
+      "epoch": 2.2213240643440306,
+      "grad_norm": 0.6474477052688599,
+      "learning_rate": 0.0002,
+      "loss": 0.6733,
+      "step": 13740
+    },
+    {
+      "epoch": 2.222940748524776,
+      "grad_norm": 0.7037909626960754,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13750
+    },
+    {
+      "epoch": 2.224557432705521,
+      "grad_norm": 0.6554131507873535,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 13760
+    },
+    {
+      "epoch": 2.2261741168862663,
+      "grad_norm": 0.7822230458259583,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 13770
+    },
+    {
+      "epoch": 2.2277908010670116,
+      "grad_norm": 0.9082167744636536,
+      "learning_rate": 0.0002,
+      "loss": 0.6505,
+      "step": 13780
+    },
+    {
+      "epoch": 2.229407485247757,
+      "grad_norm": 0.7918276190757751,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 13790
+    },
+    {
+      "epoch": 2.231024169428502,
+      "grad_norm": 0.7354569435119629,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 13800
+    },
+    {
+      "epoch": 2.2326408536092472,
+      "grad_norm": 0.8265249133110046,
+      "learning_rate": 0.0002,
+      "loss": 0.6503,
+      "step": 13810
+    },
+    {
+      "epoch": 2.234257537789993,
+      "grad_norm": 0.6653847098350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6871,
+      "step": 13820
+    },
+    {
+      "epoch": 2.235874221970738,
+      "grad_norm": 0.7157923579216003,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13830
+    },
+    {
+      "epoch": 2.2374909061514834,
+      "grad_norm": 0.7110323309898376,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 13840
+    },
+    {
+      "epoch": 2.2391075903322286,
+      "grad_norm": 0.7155357599258423,
+      "learning_rate": 0.0002,
+      "loss": 0.6913,
+      "step": 13850
+    },
+    {
+      "epoch": 2.240724274512974,
+      "grad_norm": 1.0177817344665527,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 13860
+    },
+    {
+      "epoch": 2.242340958693719,
+      "grad_norm": 0.7601948380470276,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13870
+    },
+    {
+      "epoch": 2.2439576428744643,
+      "grad_norm": 0.7628820538520813,
+      "learning_rate": 0.0002,
+      "loss": 0.6679,
+      "step": 13880
+    },
+    {
+      "epoch": 2.24557432705521,
+      "grad_norm": 0.7089297771453857,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 13890
+    },
+    {
+      "epoch": 2.247191011235955,
+      "grad_norm": 0.695178210735321,
+      "learning_rate": 0.0002,
+      "loss": 0.7236,
+      "step": 13900
+    },
+    {
+      "epoch": 2.2488076954167004,
+      "grad_norm": 0.7631948590278625,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 13910
+    },
+    {
+      "epoch": 2.2504243795974457,
+      "grad_norm": 0.8203101754188538,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 13920
+    },
+    {
+      "epoch": 2.252041063778191,
+      "grad_norm": 0.8099079728126526,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13930
+    },
+    {
+      "epoch": 2.253657747958936,
+      "grad_norm": 0.6498546004295349,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 13940
+    },
+    {
+      "epoch": 2.2552744321396814,
+      "grad_norm": 0.7797415256500244,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 13950
+    },
+    {
+      "epoch": 2.2568911163204266,
+      "grad_norm": 0.8254124522209167,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 13960
+    },
+    {
+      "epoch": 2.2585078005011723,
+      "grad_norm": 0.6327953338623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 13970
+    },
+    {
+      "epoch": 2.2601244846819175,
+      "grad_norm": 0.734194278717041,
+      "learning_rate": 0.0002,
+      "loss": 0.668,
+      "step": 13980
+    },
+    {
+      "epoch": 2.2617411688626627,
+      "grad_norm": 0.9014202952384949,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 13990
+    },
+    {
+      "epoch": 2.263357853043408,
+      "grad_norm": 0.7643631100654602,
+      "learning_rate": 0.0002,
+      "loss": 0.692,
+      "step": 14000
+    },
+    {
+      "epoch": 2.264974537224153,
+      "grad_norm": 0.8882834911346436,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 14010
+    },
+    {
+      "epoch": 2.2665912214048984,
+      "grad_norm": 0.7975873351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14020
+    },
+    {
+      "epoch": 2.2682079055856437,
+      "grad_norm": 0.7765783071517944,
+      "learning_rate": 0.0002,
+      "loss": 0.7193,
+      "step": 14030
+    },
+    {
+      "epoch": 2.2698245897663893,
+      "grad_norm": 0.8846288323402405,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 14040
+    },
+    {
+      "epoch": 2.2714412739471346,
+      "grad_norm": 0.9006744027137756,
+      "learning_rate": 0.0002,
+      "loss": 0.6494,
+      "step": 14050
+    },
+    {
+      "epoch": 2.27305795812788,
+      "grad_norm": 0.7420173287391663,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 14060
+    },
+    {
+      "epoch": 2.274674642308625,
+      "grad_norm": 0.7956424951553345,
+      "learning_rate": 0.0002,
+      "loss": 0.7068,
+      "step": 14070
+    },
+    {
+      "epoch": 2.2762913264893703,
+      "grad_norm": 0.7783209085464478,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 14080
+    },
+    {
+      "epoch": 2.2779080106701155,
+      "grad_norm": 0.7597188949584961,
+      "learning_rate": 0.0002,
+      "loss": 0.7202,
+      "step": 14090
+    },
+    {
+      "epoch": 2.2795246948508607,
+      "grad_norm": 0.6718921661376953,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14100
+    },
+    {
+      "epoch": 2.281141379031606,
+      "grad_norm": 0.7528082132339478,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 14110
+    },
+    {
+      "epoch": 2.2827580632123516,
+      "grad_norm": 0.8379864692687988,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 14120
+    },
+    {
+      "epoch": 2.284374747393097,
+      "grad_norm": 0.748613715171814,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 14130
+    },
+    {
+      "epoch": 2.285991431573842,
+      "grad_norm": 0.7435423135757446,
+      "learning_rate": 0.0002,
+      "loss": 0.6641,
+      "step": 14140
+    },
+    {
+      "epoch": 2.2876081157545873,
+      "grad_norm": 0.7580803632736206,
+      "learning_rate": 0.0002,
+      "loss": 0.6849,
+      "step": 14150
+    },
+    {
+      "epoch": 2.2892247999353326,
+      "grad_norm": 0.6278321146965027,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 14160
+    },
+    {
+      "epoch": 2.290841484116078,
+      "grad_norm": 0.7663896083831787,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 14170
+    },
+    {
+      "epoch": 2.292458168296823,
+      "grad_norm": 0.9716812372207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 14180
+    },
+    {
+      "epoch": 2.2940748524775687,
+      "grad_norm": 0.8993458151817322,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14190
+    },
+    {
+      "epoch": 2.295691536658314,
+      "grad_norm": 0.6156117916107178,
+      "learning_rate": 0.0002,
+      "loss": 0.6172,
+      "step": 14200
+    },
+    {
+      "epoch": 2.297308220839059,
+      "grad_norm": 0.8911278247833252,
+      "learning_rate": 0.0002,
+      "loss": 0.6318,
+      "step": 14210
+    },
+    {
+      "epoch": 2.2989249050198044,
+      "grad_norm": 0.6422147154808044,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 14220
+    },
+    {
+      "epoch": 2.3005415892005496,
+      "grad_norm": 0.6866879463195801,
+      "learning_rate": 0.0002,
+      "loss": 0.6795,
+      "step": 14230
+    },
+    {
+      "epoch": 2.302158273381295,
+      "grad_norm": 0.9297130107879639,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 14240
+    },
+    {
+      "epoch": 2.30377495756204,
+      "grad_norm": 0.7501356601715088,
+      "learning_rate": 0.0002,
+      "loss": 0.6823,
+      "step": 14250
+    },
+    {
+      "epoch": 2.3053916417427853,
+      "grad_norm": 0.8363515138626099,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 14260
+    },
+    {
+      "epoch": 2.307008325923531,
+      "grad_norm": 0.9083868265151978,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3086250101042762,
+      "grad_norm": 0.7791516780853271,
+      "learning_rate": 0.0002,
+      "loss": 0.6862,
+      "step": 14280
+    },
+    {
+      "epoch": 2.3102416942850215,
+      "grad_norm": 0.8766953349113464,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14290
+    },
+    {
+      "epoch": 2.3118583784657667,
+      "grad_norm": 0.7916635274887085,
+      "learning_rate": 0.0002,
+      "loss": 0.6698,
+      "step": 14300
+    },
+    {
+      "epoch": 2.313475062646512,
+      "grad_norm": 0.627525269985199,
+      "learning_rate": 0.0002,
+      "loss": 0.6927,
+      "step": 14310
+    },
+    {
+      "epoch": 2.315091746827257,
+      "grad_norm": 0.8856783509254456,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 14320
+    },
+    {
+      "epoch": 2.316708431008003,
+      "grad_norm": 0.6758689284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 14330
+    },
+    {
+      "epoch": 2.318325115188748,
+      "grad_norm": 0.6428321003913879,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 14340
+    },
+    {
+      "epoch": 2.3199417993694933,
+      "grad_norm": 0.9032121300697327,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 14350
+    },
+    {
+      "epoch": 2.3215584835502385,
+      "grad_norm": 0.8035986423492432,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14360
+    },
+    {
+      "epoch": 2.3231751677309838,
+      "grad_norm": 0.7974579334259033,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14370
+    },
+    {
+      "epoch": 2.324791851911729,
+      "grad_norm": 0.8356034755706787,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 14380
+    },
+    {
+      "epoch": 2.326408536092474,
+      "grad_norm": 0.998760998249054,
+      "learning_rate": 0.0002,
+      "loss": 0.6623,
+      "step": 14390
+    },
+    {
+      "epoch": 2.3280252202732195,
+      "grad_norm": 0.6518142223358154,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 14400
+    },
+    {
+      "epoch": 2.3296419044539647,
+      "grad_norm": 0.7443506717681885,
+      "learning_rate": 0.0002,
+      "loss": 0.7146,
+      "step": 14410
+    },
+    {
+      "epoch": 2.3312585886347104,
+      "grad_norm": 0.8436172604560852,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14420
+    },
+    {
+      "epoch": 2.3328752728154556,
+      "grad_norm": 0.7411080598831177,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 14430
+    },
+    {
+      "epoch": 2.334491956996201,
+      "grad_norm": 0.8839048743247986,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 14440
+    },
+    {
+      "epoch": 2.336108641176946,
+      "grad_norm": 0.8360885977745056,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 14450
+    },
+    {
+      "epoch": 2.3377253253576913,
+      "grad_norm": 0.7608986496925354,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 14460
+    },
+    {
+      "epoch": 2.3393420095384365,
+      "grad_norm": 0.8179867267608643,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14470
+    },
+    {
+      "epoch": 2.340958693719182,
+      "grad_norm": 0.5989999771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14480
+    },
+    {
+      "epoch": 2.3425753778999274,
+      "grad_norm": 0.9450054168701172,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 14490
+    },
+    {
+      "epoch": 2.3441920620806727,
+      "grad_norm": 0.7885149717330933,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 14500
+    },
+    {
+      "epoch": 2.345808746261418,
+      "grad_norm": 0.8152616620063782,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 14510
+    },
+    {
+      "epoch": 2.347425430442163,
+      "grad_norm": 0.7193838953971863,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 14520
+    },
+    {
+      "epoch": 2.3490421146229084,
+      "grad_norm": 0.6701092720031738,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 14530
+    },
+    {
+      "epoch": 2.3506587988036536,
+      "grad_norm": 0.7529364228248596,
+      "learning_rate": 0.0002,
+      "loss": 0.6559,
+      "step": 14540
+    },
+    {
+      "epoch": 2.352275482984399,
+      "grad_norm": 0.6599733829498291,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 14550
+    },
+    {
+      "epoch": 2.353892167165144,
+      "grad_norm": 0.9502474069595337,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14560
+    },
+    {
+      "epoch": 2.3555088513458897,
+      "grad_norm": 0.7619650959968567,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 14570
+    },
+    {
+      "epoch": 2.357125535526635,
+      "grad_norm": 0.9854652285575867,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14580
+    },
+    {
+      "epoch": 2.35874221970738,
+      "grad_norm": 0.727439284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 14590
+    },
+    {
+      "epoch": 2.3603589038881254,
+      "grad_norm": 0.6994746327400208,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 14600
+    },
+    {
+      "epoch": 2.3619755880688706,
+      "grad_norm": 0.7117531299591064,
+      "learning_rate": 0.0002,
+      "loss": 0.6587,
+      "step": 14610
+    },
+    {
+      "epoch": 2.363592272249616,
+      "grad_norm": 0.6403067708015442,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 14620
+    },
+    {
+      "epoch": 2.3652089564303616,
+      "grad_norm": 0.8377841711044312,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14630
+    },
+    {
+      "epoch": 2.366825640611107,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14640
+    },
+    {
+      "epoch": 2.368442324791852,
+      "grad_norm": 0.8418586254119873,
+      "learning_rate": 0.0002,
+      "loss": 0.6552,
+      "step": 14650
+    },
+    {
+      "epoch": 2.3700590089725972,
+      "grad_norm": 0.6178573369979858,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14660
+    },
+    {
+      "epoch": 2.3716756931533425,
+      "grad_norm": 0.6368302702903748,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 14670
+    },
+    {
+      "epoch": 2.3732923773340877,
+      "grad_norm": 0.9122977256774902,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14680
+    },
+    {
+      "epoch": 2.374909061514833,
+      "grad_norm": 0.7086195349693298,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 14690
+    },
+    {
+      "epoch": 2.376525745695578,
+      "grad_norm": 0.7500800490379333,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 14700
+    },
+    {
+      "epoch": 2.378142429876324,
+      "grad_norm": 0.6634900569915771,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 14710
+    },
+    {
+      "epoch": 2.379759114057069,
+      "grad_norm": 0.839898407459259,
+      "learning_rate": 0.0002,
+      "loss": 0.6614,
+      "step": 14720
+    },
+    {
+      "epoch": 2.3813757982378143,
+      "grad_norm": 0.7578426003456116,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14730
+    },
+    {
+      "epoch": 2.3829924824185595,
+      "grad_norm": 1.0213173627853394,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3846091665993048,
+      "grad_norm": 0.7855949401855469,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 14750
+    },
+    {
+      "epoch": 2.38622585078005,
+      "grad_norm": 0.7224128842353821,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 14760
+    },
+    {
+      "epoch": 2.3878425349607952,
+      "grad_norm": 0.8040381669998169,
+      "learning_rate": 0.0002,
+      "loss": 0.7017,
+      "step": 14770
+    },
+    {
+      "epoch": 2.389459219141541,
+      "grad_norm": 0.7705281376838684,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 14780
+    },
+    {
+      "epoch": 2.391075903322286,
+      "grad_norm": 0.667966902256012,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 14790
+    },
+    {
+      "epoch": 2.3926925875030314,
+      "grad_norm": 0.6611011028289795,
+      "learning_rate": 0.0002,
+      "loss": 0.7061,
+      "step": 14800
+    },
+    {
+      "epoch": 2.3943092716837766,
+      "grad_norm": 0.6862651705741882,
+      "learning_rate": 0.0002,
+      "loss": 0.6527,
+      "step": 14810
+    },
+    {
+      "epoch": 2.395925955864522,
+      "grad_norm": 0.8086010217666626,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 14820
+    },
+    {
+      "epoch": 2.397542640045267,
+      "grad_norm": 0.7189689874649048,
+      "learning_rate": 0.0002,
+      "loss": 0.7189,
+      "step": 14830
+    },
+    {
+      "epoch": 2.3991593242260123,
+      "grad_norm": 0.6280009150505066,
+      "learning_rate": 0.0002,
+      "loss": 0.6709,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4007760084067575,
+      "grad_norm": 0.7826612591743469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14850
+    },
+    {
+      "epoch": 2.402392692587503,
+      "grad_norm": 0.7681610584259033,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4040093767682484,
+      "grad_norm": 0.720966100692749,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 14870
+    },
+    {
+      "epoch": 2.4056260609489937,
+      "grad_norm": 0.8202250599861145,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 14880
+    },
+    {
+      "epoch": 2.407242745129739,
+      "grad_norm": 0.786212682723999,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 14890
+    },
+    {
+      "epoch": 2.408859429310484,
+      "grad_norm": 0.6647164821624756,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 14900
+    },
+    {
+      "epoch": 2.4104761134912294,
+      "grad_norm": 0.7566399574279785,
+      "learning_rate": 0.0002,
+      "loss": 0.6706,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4120927976719746,
+      "grad_norm": 0.748814582824707,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 14920
+    },
+    {
+      "epoch": 2.4137094818527203,
+      "grad_norm": 0.7624038457870483,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14930
+    },
+    {
+      "epoch": 2.4153261660334655,
+      "grad_norm": 0.8267335295677185,
+      "learning_rate": 0.0002,
+      "loss": 0.6483,
+      "step": 14940
+    },
+    {
+      "epoch": 2.4169428502142107,
+      "grad_norm": 0.8785360455513,
+      "learning_rate": 0.0002,
+      "loss": 0.6612,
+      "step": 14950
+    },
+    {
+      "epoch": 2.418559534394956,
+      "grad_norm": 0.679887592792511,
+      "learning_rate": 0.0002,
+      "loss": 0.6718,
+      "step": 14960
+    },
+    {
+      "epoch": 2.420176218575701,
+      "grad_norm": 0.7218474745750427,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4217929027564464,
+      "grad_norm": 0.6342799663543701,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14980
+    },
+    {
+      "epoch": 2.4234095869371917,
+      "grad_norm": 0.7098712921142578,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 14990
+    },
+    {
+      "epoch": 2.425026271117937,
+      "grad_norm": 0.7497431635856628,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4266429552986826,
+      "grad_norm": 0.934836208820343,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15010
+    },
+    {
+      "epoch": 2.428259639479428,
+      "grad_norm": 0.8430966734886169,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 15020
+    },
+    {
+      "epoch": 2.429876323660173,
+      "grad_norm": 0.7032104730606079,
+      "learning_rate": 0.0002,
+      "loss": 0.6985,
+      "step": 15030
+    },
+    {
+      "epoch": 2.4314930078409183,
+      "grad_norm": 0.7746111750602722,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 15040
+    },
+    {
+      "epoch": 2.4331096920216635,
+      "grad_norm": 0.7661406397819519,
+      "learning_rate": 0.0002,
+      "loss": 0.7177,
+      "step": 15050
+    },
+    {
+      "epoch": 2.4347263762024087,
+      "grad_norm": 0.6941645741462708,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 15060
+    },
+    {
+      "epoch": 2.436343060383154,
+      "grad_norm": 0.7487249374389648,
+      "learning_rate": 0.0002,
+      "loss": 0.6421,
+      "step": 15070
+    },
+    {
+      "epoch": 2.4379597445638996,
+      "grad_norm": 0.7639912962913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 15080
+    },
+    {
+      "epoch": 2.439576428744645,
+      "grad_norm": 0.7708953619003296,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 15090
+    },
+    {
+      "epoch": 2.44119311292539,
+      "grad_norm": 0.9135832190513611,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15100
+    },
+    {
+      "epoch": 2.4428097971061353,
+      "grad_norm": 0.8283005356788635,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 15110
+    },
+    {
+      "epoch": 2.4444264812868806,
+      "grad_norm": 0.925299346446991,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 15120
+    },
+    {
+      "epoch": 2.446043165467626,
+      "grad_norm": 0.7013528943061829,
+      "learning_rate": 0.0002,
+      "loss": 0.687,
+      "step": 15130
+    },
+    {
+      "epoch": 2.447659849648371,
+      "grad_norm": 0.622303307056427,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 15140
+    },
+    {
+      "epoch": 2.4492765338291163,
+      "grad_norm": 0.876569390296936,
+      "learning_rate": 0.0002,
+      "loss": 0.6676,
+      "step": 15150
+    },
+    {
+      "epoch": 2.450893218009862,
+      "grad_norm": 0.6836351752281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 15160
+    },
+    {
+      "epoch": 2.452509902190607,
+      "grad_norm": 0.7886684536933899,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 15170
+    },
+    {
+      "epoch": 2.4541265863713524,
+      "grad_norm": 0.6647440791130066,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 15180
+    },
+    {
+      "epoch": 2.4557432705520976,
+      "grad_norm": 0.7477722764015198,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 15190
+    },
+    {
+      "epoch": 2.457359954732843,
+      "grad_norm": 0.8192033767700195,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 15200
+    },
+    {
+      "epoch": 2.458976638913588,
+      "grad_norm": 0.847537100315094,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 15210
+    },
+    {
+      "epoch": 2.4605933230943338,
+      "grad_norm": 0.9027776122093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 15220
+    },
+    {
+      "epoch": 2.462210007275079,
+      "grad_norm": 0.7217772006988525,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 15230
+    },
+    {
+      "epoch": 2.4638266914558242,
+      "grad_norm": 0.7994546294212341,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 15240
+    },
+    {
+      "epoch": 2.4654433756365695,
+      "grad_norm": 0.939916729927063,
+      "learning_rate": 0.0002,
+      "loss": 0.6828,
+      "step": 15250
+    },
+    {
+      "epoch": 2.4670600598173147,
+      "grad_norm": 1.0009053945541382,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15260
+    },
+    {
+      "epoch": 2.46867674399806,
+      "grad_norm": 0.625555694103241,
+      "learning_rate": 0.0002,
+      "loss": 0.643,
+      "step": 15270
+    },
+    {
+      "epoch": 2.470293428178805,
+      "grad_norm": 0.7924878597259521,
+      "learning_rate": 0.0002,
+      "loss": 0.688,
+      "step": 15280
+    },
+    {
+      "epoch": 2.4719101123595504,
+      "grad_norm": 0.8536689877510071,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 15290
+    },
+    {
+      "epoch": 2.4735267965402956,
+      "grad_norm": 0.8572589755058289,
+      "learning_rate": 0.0002,
+      "loss": 0.6924,
+      "step": 15300
+    },
+    {
+      "epoch": 2.4751434807210413,
+      "grad_norm": 0.773279070854187,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 15310
+    },
+    {
+      "epoch": 2.4767601649017865,
+      "grad_norm": 0.7708749771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 15320
+    },
+    {
+      "epoch": 2.4783768490825318,
+      "grad_norm": 0.770905077457428,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15330
+    },
+    {
+      "epoch": 2.479993533263277,
+      "grad_norm": 0.8238571882247925,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 15340
+    },
+    {
+      "epoch": 2.481610217444022,
+      "grad_norm": 0.7670477032661438,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15350
+    },
+    {
+      "epoch": 2.4832269016247674,
+      "grad_norm": 0.905036985874176,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 15360
+    },
+    {
+      "epoch": 2.484843585805513,
+      "grad_norm": 0.6672089695930481,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 15370
+    },
+    {
+      "epoch": 2.4864602699862584,
+      "grad_norm": 0.625095784664154,
+      "learning_rate": 0.0002,
+      "loss": 0.6722,
+      "step": 15380
+    },
+    {
+      "epoch": 2.4880769541670036,
+      "grad_norm": 0.679772675037384,
+      "learning_rate": 0.0002,
+      "loss": 0.6396,
+      "step": 15390
+    },
+    {
+      "epoch": 2.489693638347749,
+      "grad_norm": 0.711492121219635,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 15400
+    },
+    {
+      "epoch": 2.491310322528494,
+      "grad_norm": 0.876189112663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6966,
+      "step": 15410
+    },
+    {
+      "epoch": 2.4929270067092393,
+      "grad_norm": 0.7236915230751038,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 15420
+    },
+    {
+      "epoch": 2.4945436908899845,
+      "grad_norm": 0.6629832983016968,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 15430
+    },
+    {
+      "epoch": 2.4961603750707297,
+      "grad_norm": 0.9756859540939331,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 15440
+    },
+    {
+      "epoch": 2.4977770592514754,
+      "grad_norm": 0.6896940469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 15450
+    },
+    {
+      "epoch": 2.4993937434322206,
+      "grad_norm": 0.7105149626731873,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 15460
+    },
+    {
+      "epoch": 2.501010427612966,
+      "grad_norm": 0.8374546766281128,
+      "learning_rate": 0.0002,
+      "loss": 0.6859,
+      "step": 15470
+    },
+    {
+      "epoch": 2.502627111793711,
+      "grad_norm": 0.7320070266723633,
+      "learning_rate": 0.0002,
+      "loss": 0.6512,
+      "step": 15480
+    },
+    {
+      "epoch": 2.5042437959744563,
+      "grad_norm": 0.8306367993354797,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 15490
+    },
+    {
+      "epoch": 2.5058604801552016,
+      "grad_norm": 0.7472721338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.7253,
+      "step": 15500
+    },
+    {
+      "epoch": 2.507477164335947,
+      "grad_norm": 0.6147692203521729,
+      "learning_rate": 0.0002,
+      "loss": 0.6699,
+      "step": 15510
+    },
+    {
+      "epoch": 2.5090938485166925,
+      "grad_norm": 0.7788505554199219,
+      "learning_rate": 0.0002,
+      "loss": 0.7158,
+      "step": 15520
+    },
+    {
+      "epoch": 2.5107105326974377,
+      "grad_norm": 0.8807527422904968,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 15530
+    },
+    {
+      "epoch": 2.512327216878183,
+      "grad_norm": 0.7521643042564392,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 15540
+    },
+    {
+      "epoch": 2.513943901058928,
+      "grad_norm": 0.6900225281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15550
+    },
+    {
+      "epoch": 2.5155605852396734,
+      "grad_norm": 0.6601938605308533,
+      "learning_rate": 0.0002,
+      "loss": 0.6769,
+      "step": 15560
+    },
+    {
+      "epoch": 2.5171772694204186,
+      "grad_norm": 0.8179984092712402,
+      "learning_rate": 0.0002,
+      "loss": 0.6648,
+      "step": 15570
+    },
+    {
+      "epoch": 2.518793953601164,
+      "grad_norm": 0.792556881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.7028,
+      "step": 15580
+    },
+    {
+      "epoch": 2.520410637781909,
+      "grad_norm": 0.7081938982009888,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 15590
+    },
+    {
+      "epoch": 2.5220273219626543,
+      "grad_norm": 0.8733121156692505,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 15600
+    },
+    {
+      "epoch": 2.5236440061434,
+      "grad_norm": 0.7980992794036865,
+      "learning_rate": 0.0002,
+      "loss": 0.6969,
+      "step": 15610
+    },
+    {
+      "epoch": 2.5252606903241452,
+      "grad_norm": 0.883664071559906,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 15620
+    },
+    {
+      "epoch": 2.5268773745048905,
+      "grad_norm": 0.6963341236114502,
+      "learning_rate": 0.0002,
+      "loss": 0.7022,
+      "step": 15630
+    },
+    {
+      "epoch": 2.5284940586856357,
+      "grad_norm": 0.6433573365211487,
+      "learning_rate": 0.0002,
+      "loss": 0.7334,
+      "step": 15640
+    },
+    {
+      "epoch": 2.530110742866381,
+      "grad_norm": 0.8538183569908142,
+      "learning_rate": 0.0002,
+      "loss": 0.6889,
+      "step": 15650
+    },
+    {
+      "epoch": 2.5317274270471266,
+      "grad_norm": 0.9748201370239258,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 15660
+    },
+    {
+      "epoch": 2.533344111227872,
+      "grad_norm": 0.7670575380325317,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 15670
+    },
+    {
+      "epoch": 2.534960795408617,
+      "grad_norm": 0.8738890290260315,
+      "learning_rate": 0.0002,
+      "loss": 0.6435,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5365774795893623,
+      "grad_norm": 0.8391636610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.6802,
+      "step": 15690
+    },
+    {
+      "epoch": 2.5381941637701075,
+      "grad_norm": 0.7239366769790649,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 15700
+    },
+    {
+      "epoch": 2.5398108479508528,
+      "grad_norm": 0.8498379588127136,
+      "learning_rate": 0.0002,
+      "loss": 0.7011,
+      "step": 15710
+    },
+    {
+      "epoch": 2.541427532131598,
+      "grad_norm": 0.8029484152793884,
+      "learning_rate": 0.0002,
+      "loss": 0.6998,
+      "step": 15720
+    },
+    {
+      "epoch": 2.5430442163123432,
+      "grad_norm": 1.0639333724975586,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 15730
+    },
+    {
+      "epoch": 2.5446609004930885,
+      "grad_norm": 0.6401297450065613,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 15740
+    },
+    {
+      "epoch": 2.5462775846738337,
+      "grad_norm": 0.7123814821243286,
+      "learning_rate": 0.0002,
+      "loss": 0.7196,
+      "step": 15750
+    },
+    {
+      "epoch": 2.5478942688545794,
+      "grad_norm": 0.7874974608421326,
+      "learning_rate": 0.0002,
+      "loss": 0.654,
+      "step": 15760
+    },
+    {
+      "epoch": 2.5495109530353246,
+      "grad_norm": 0.8046808838844299,
+      "learning_rate": 0.0002,
+      "loss": 0.6721,
+      "step": 15770
+    },
+    {
+      "epoch": 2.55112763721607,
+      "grad_norm": 0.7888661623001099,
+      "learning_rate": 0.0002,
+      "loss": 0.6665,
+      "step": 15780
+    },
+    {
+      "epoch": 2.552744321396815,
+      "grad_norm": 0.8445866107940674,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15790
+    },
+    {
+      "epoch": 2.5543610055775603,
+      "grad_norm": 0.7475846409797668,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 15800
+    },
+    {
+      "epoch": 2.555977689758306,
+      "grad_norm": 0.7455102801322937,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 15810
+    },
+    {
+      "epoch": 2.557594373939051,
+      "grad_norm": 0.8226983547210693,
+      "learning_rate": 0.0002,
+      "loss": 0.6932,
+      "step": 15820
+    },
+    {
+      "epoch": 2.5592110581197964,
+      "grad_norm": 0.8920368552207947,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 15830
+    },
+    {
+      "epoch": 2.5608277423005417,
+      "grad_norm": 0.8413904905319214,
+      "learning_rate": 0.0002,
+      "loss": 0.6297,
+      "step": 15840
+    },
+    {
+      "epoch": 2.562444426481287,
+      "grad_norm": 0.8483649492263794,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 15850
+    },
+    {
+      "epoch": 2.564061110662032,
+      "grad_norm": 0.5923284292221069,
+      "learning_rate": 0.0002,
+      "loss": 0.6957,
+      "step": 15860
+    },
+    {
+      "epoch": 2.5656777948427774,
+      "grad_norm": 0.8518726229667664,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 15870
+    },
+    {
+      "epoch": 2.5672944790235226,
+      "grad_norm": 0.731235146522522,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 15880
+    },
+    {
+      "epoch": 2.568911163204268,
+      "grad_norm": 0.7517194151878357,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 15890
+    },
+    {
+      "epoch": 2.5705278473850135,
+      "grad_norm": 0.8378692269325256,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 15900
+    },
+    {
+      "epoch": 2.5721445315657587,
+      "grad_norm": 0.843701958656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 15910
+    },
+    {
+      "epoch": 2.573761215746504,
+      "grad_norm": 0.7254629731178284,
+      "learning_rate": 0.0002,
+      "loss": 0.6882,
+      "step": 15920
+    },
+    {
+      "epoch": 2.575377899927249,
+      "grad_norm": 0.8863335847854614,
+      "learning_rate": 0.0002,
+      "loss": 0.6872,
+      "step": 15930
+    },
+    {
+      "epoch": 2.5769945841079944,
+      "grad_norm": 0.7675097584724426,
+      "learning_rate": 0.0002,
+      "loss": 0.6813,
+      "step": 15940
+    },
+    {
+      "epoch": 2.5786112682887397,
+      "grad_norm": 0.82063889503479,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 15950
+    },
+    {
+      "epoch": 2.5802279524694853,
+      "grad_norm": 0.7729717493057251,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 15960
+    },
+    {
+      "epoch": 2.5818446366502306,
+      "grad_norm": 0.8301846981048584,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 15970
+    },
+    {
+      "epoch": 2.583461320830976,
+      "grad_norm": 0.7906861305236816,
+      "learning_rate": 0.0002,
+      "loss": 0.6897,
+      "step": 15980
+    },
+    {
+      "epoch": 2.585078005011721,
+      "grad_norm": 0.6749057173728943,
+      "learning_rate": 0.0002,
+      "loss": 0.7175,
+      "step": 15990
+    },
+    {
+      "epoch": 2.5866946891924663,
+      "grad_norm": 0.9386842846870422,
+      "learning_rate": 0.0002,
+      "loss": 0.7212,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5883113733732115,
+      "grad_norm": 0.7868891358375549,
+      "learning_rate": 0.0002,
+      "loss": 0.6934,
+      "step": 16010
+    },
+    {
+      "epoch": 2.5899280575539567,
+      "grad_norm": 0.8674671053886414,
+      "learning_rate": 0.0002,
+      "loss": 0.7036,
+      "step": 16020
+    },
+    {
+      "epoch": 2.591544741734702,
+      "grad_norm": 0.7043559551239014,
+      "learning_rate": 0.0002,
+      "loss": 0.7217,
+      "step": 16030
+    },
+    {
+      "epoch": 2.593161425915447,
+      "grad_norm": 0.5846083760261536,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 16040
+    },
+    {
+      "epoch": 2.594778110096193,
+      "grad_norm": 0.7323982119560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 16050
+    },
+    {
+      "epoch": 2.596394794276938,
+      "grad_norm": 0.9069556593894958,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 16060
+    },
+    {
+      "epoch": 2.5980114784576833,
+      "grad_norm": 0.7522736191749573,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 16070
+    },
+    {
+      "epoch": 2.5996281626384286,
+      "grad_norm": 0.8149648308753967,
+      "learning_rate": 0.0002,
+      "loss": 0.6477,
+      "step": 16080
+    },
+    {
+      "epoch": 2.601244846819174,
+      "grad_norm": 0.6214233040809631,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 16090
+    },
+    {
+      "epoch": 2.602861530999919,
+      "grad_norm": 0.6803743839263916,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 16100
+    },
+    {
+      "epoch": 2.6044782151806647,
+      "grad_norm": 0.7223997116088867,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 16110
+    },
+    {
+      "epoch": 2.60609489936141,
+      "grad_norm": 0.7324174642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.6867,
+      "step": 16120
+    },
+    {
+      "epoch": 2.607711583542155,
+      "grad_norm": 0.9594739675521851,
+      "learning_rate": 0.0002,
+      "loss": 0.7159,
+      "step": 16130
+    },
+    {
+      "epoch": 2.6093282677229004,
+      "grad_norm": 0.9485327005386353,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 16140
+    },
+    {
+      "epoch": 2.6109449519036456,
+      "grad_norm": 0.8449000120162964,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 16150
+    },
+    {
+      "epoch": 2.612561636084391,
+      "grad_norm": 0.8520140051841736,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 16160
+    },
+    {
+      "epoch": 2.614178320265136,
+      "grad_norm": 0.7456524968147278,
+      "learning_rate": 0.0002,
+      "loss": 0.6759,
+      "step": 16170
+    },
+    {
+      "epoch": 2.6157950044458813,
+      "grad_norm": 0.9912857413291931,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 16180
+    },
+    {
+      "epoch": 2.6174116886266265,
+      "grad_norm": 0.9001946449279785,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 16190
+    },
+    {
+      "epoch": 2.619028372807372,
+      "grad_norm": 0.6568667888641357,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16200
+    },
+    {
+      "epoch": 2.6206450569881174,
+      "grad_norm": 1.0248128175735474,
+      "learning_rate": 0.0002,
+      "loss": 0.7013,
+      "step": 16210
+    },
+    {
+      "epoch": 2.6222617411688627,
+      "grad_norm": 0.6509039998054504,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 16220
+    },
+    {
+      "epoch": 2.623878425349608,
+      "grad_norm": 0.7626351118087769,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 16230
+    },
+    {
+      "epoch": 2.625495109530353,
+      "grad_norm": 0.6938552260398865,
+      "learning_rate": 0.0002,
+      "loss": 0.6556,
+      "step": 16240
+    },
+    {
+      "epoch": 2.6271117937110984,
+      "grad_norm": 0.6434680819511414,
+      "learning_rate": 0.0002,
+      "loss": 0.65,
+      "step": 16250
+    },
+    {
+      "epoch": 2.628728477891844,
+      "grad_norm": 0.7111515998840332,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 16260
+    },
+    {
+      "epoch": 2.6303451620725893,
+      "grad_norm": 0.7712395787239075,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 16270
+    },
+    {
+      "epoch": 2.6319618462533345,
+      "grad_norm": 0.792209267616272,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 16280
+    },
+    {
+      "epoch": 2.6335785304340797,
+      "grad_norm": 0.6801066398620605,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 16290
+    },
+    {
+      "epoch": 2.635195214614825,
+      "grad_norm": 0.7802573442459106,
+      "learning_rate": 0.0002,
+      "loss": 0.73,
+      "step": 16300
+    },
+    {
+      "epoch": 2.63681189879557,
+      "grad_norm": 0.7742244601249695,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 16310
+    },
+    {
+      "epoch": 2.6384285829763154,
+      "grad_norm": 0.664184033870697,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 16320
+    },
+    {
+      "epoch": 2.6400452671570607,
+      "grad_norm": 0.9242228865623474,
+      "learning_rate": 0.0002,
+      "loss": 0.6442,
+      "step": 16330
+    },
+    {
+      "epoch": 2.641661951337806,
+      "grad_norm": 0.9661325216293335,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 16340
+    },
+    {
+      "epoch": 2.6432786355185516,
+      "grad_norm": 0.837526798248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 16350
+    },
+    {
+      "epoch": 2.644895319699297,
+      "grad_norm": 1.1834373474121094,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 16360
+    },
+    {
+      "epoch": 2.646512003880042,
+      "grad_norm": 0.7467831373214722,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 16370
+    },
+    {
+      "epoch": 2.6481286880607873,
+      "grad_norm": 0.8627146482467651,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 16380
+    },
+    {
+      "epoch": 2.6497453722415325,
+      "grad_norm": 0.790447473526001,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 16390
+    },
+    {
+      "epoch": 2.651362056422278,
+      "grad_norm": 0.8447365164756775,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 16400
+    },
+    {
+      "epoch": 2.6529787406030234,
+      "grad_norm": 0.7831417918205261,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 16410
+    },
+    {
+      "epoch": 2.6545954247837686,
+      "grad_norm": 0.6837952136993408,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 16420
+    },
+    {
+      "epoch": 2.656212108964514,
+      "grad_norm": 0.7031801342964172,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 16430
+    },
+    {
+      "epoch": 2.657828793145259,
+      "grad_norm": 0.8963770866394043,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 16440
+    },
+    {
+      "epoch": 2.6594454773260043,
+      "grad_norm": 0.6852328181266785,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 16450
+    },
+    {
+      "epoch": 2.6610621615067496,
+      "grad_norm": 0.8069294095039368,
+      "learning_rate": 0.0002,
+      "loss": 0.6278,
+      "step": 16460
+    },
+    {
+      "epoch": 2.662678845687495,
+      "grad_norm": 0.7503686547279358,
+      "learning_rate": 0.0002,
+      "loss": 0.6939,
+      "step": 16470
+    },
+    {
+      "epoch": 2.66429552986824,
+      "grad_norm": 0.6430956125259399,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16480
+    },
+    {
+      "epoch": 2.6659122140489853,
+      "grad_norm": 0.7894312739372253,
+      "learning_rate": 0.0002,
+      "loss": 0.6863,
+      "step": 16490
+    },
+    {
+      "epoch": 2.667528898229731,
+      "grad_norm": 0.7277431488037109,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 16500
+    },
+    {
+      "epoch": 2.669145582410476,
+      "grad_norm": 0.6816153526306152,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 16510
+    },
+    {
+      "epoch": 2.6707622665912214,
+      "grad_norm": 0.8145235776901245,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 16520
+    },
+    {
+      "epoch": 2.6723789507719666,
+      "grad_norm": 0.8645890355110168,
+      "learning_rate": 0.0002,
+      "loss": 0.709,
+      "step": 16530
+    },
+    {
+      "epoch": 2.673995634952712,
+      "grad_norm": 0.704393208026886,
+      "learning_rate": 0.0002,
+      "loss": 0.6946,
+      "step": 16540
+    },
+    {
+      "epoch": 2.6756123191334575,
+      "grad_norm": 1.0120846033096313,
+      "learning_rate": 0.0002,
+      "loss": 0.6378,
+      "step": 16550
+    },
+    {
+      "epoch": 2.6772290033142028,
+      "grad_norm": 0.6919328570365906,
+      "learning_rate": 0.0002,
+      "loss": 0.7241,
+      "step": 16560
+    },
+    {
+      "epoch": 2.678845687494948,
+      "grad_norm": 0.6924574971199036,
+      "learning_rate": 0.0002,
+      "loss": 0.7098,
+      "step": 16570
+    },
+    {
+      "epoch": 2.6804623716756932,
+      "grad_norm": 0.9679301381111145,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 16580
+    },
+    {
+      "epoch": 2.6820790558564385,
+      "grad_norm": 0.6810211539268494,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 16590
+    },
+    {
+      "epoch": 2.6836957400371837,
+      "grad_norm": 0.9730555415153503,
+      "learning_rate": 0.0002,
+      "loss": 0.6688,
+      "step": 16600
+    },
+    {
+      "epoch": 2.685312424217929,
+      "grad_norm": 0.7852821350097656,
+      "learning_rate": 0.0002,
+      "loss": 0.7344,
+      "step": 16610
+    },
+    {
+      "epoch": 2.686929108398674,
+      "grad_norm": 0.6059057116508484,
+      "learning_rate": 0.0002,
+      "loss": 0.6401,
+      "step": 16620
+    },
+    {
+      "epoch": 2.6885457925794194,
+      "grad_norm": 0.9395958781242371,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 16630
+    },
+    {
+      "epoch": 2.690162476760165,
+      "grad_norm": 0.7473729848861694,
+      "learning_rate": 0.0002,
+      "loss": 0.7174,
+      "step": 16640
+    },
+    {
+      "epoch": 2.6917791609409103,
+      "grad_norm": 0.765934407711029,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 16650
+    },
+    {
+      "epoch": 2.6933958451216555,
+      "grad_norm": 0.8496677279472351,
+      "learning_rate": 0.0002,
+      "loss": 0.707,
+      "step": 16660
+    },
+    {
+      "epoch": 2.6950125293024008,
+      "grad_norm": 0.7641879916191101,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 16670
+    },
+    {
+      "epoch": 2.696629213483146,
+      "grad_norm": 0.8471952676773071,
+      "learning_rate": 0.0002,
+      "loss": 0.6566,
+      "step": 16680
+    },
+    {
+      "epoch": 2.6982458976638912,
+      "grad_norm": 0.6946060657501221,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 16690
+    },
+    {
+      "epoch": 2.699862581844637,
+      "grad_norm": 0.7361312508583069,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 16700
+    },
+    {
+      "epoch": 2.701479266025382,
+      "grad_norm": 0.6605038046836853,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 16710
+    },
+    {
+      "epoch": 2.7030959502061274,
+      "grad_norm": 0.7164411544799805,
+      "learning_rate": 0.0002,
+      "loss": 0.6885,
+      "step": 16720
+    },
+    {
+      "epoch": 2.7047126343868726,
+      "grad_norm": 0.6496201157569885,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 16730
+    },
+    {
+      "epoch": 2.706329318567618,
+      "grad_norm": 0.7826663851737976,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 16740
+    },
+    {
+      "epoch": 2.707946002748363,
+      "grad_norm": 0.7639131546020508,
+      "learning_rate": 0.0002,
+      "loss": 0.6773,
+      "step": 16750
+    },
+    {
+      "epoch": 2.7095626869291083,
+      "grad_norm": 0.7976210713386536,
+      "learning_rate": 0.0002,
+      "loss": 0.69,
+      "step": 16760
+    },
+    {
+      "epoch": 2.7111793711098535,
+      "grad_norm": 0.6836577653884888,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 16770
+    },
+    {
+      "epoch": 2.7127960552905988,
+      "grad_norm": 0.8025202751159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 16780
+    },
+    {
+      "epoch": 2.7144127394713444,
+      "grad_norm": 0.7636463642120361,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7160294236520897,
+      "grad_norm": 0.7481677532196045,
+      "learning_rate": 0.0002,
+      "loss": 0.6227,
+      "step": 16800
+    },
+    {
+      "epoch": 2.717646107832835,
+      "grad_norm": 0.7566834688186646,
+      "learning_rate": 0.0002,
+      "loss": 0.6925,
+      "step": 16810
+    },
+    {
+      "epoch": 2.71926279201358,
+      "grad_norm": 0.7931267619132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6531,
+      "step": 16820
+    },
+    {
+      "epoch": 2.7208794761943254,
+      "grad_norm": 0.8811662197113037,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 16830
+    },
+    {
+      "epoch": 2.7224961603750706,
+      "grad_norm": 0.8561240434646606,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 16840
+    },
+    {
+      "epoch": 2.7241128445558163,
+      "grad_norm": 0.7121599316596985,
+      "learning_rate": 0.0002,
+      "loss": 0.7135,
+      "step": 16850
+    },
+    {
+      "epoch": 2.7257295287365615,
+      "grad_norm": 0.8066257238388062,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16860
+    },
+    {
+      "epoch": 2.7273462129173067,
+      "grad_norm": 0.7699271440505981,
+      "learning_rate": 0.0002,
+      "loss": 0.6839,
+      "step": 16870
+    },
+    {
+      "epoch": 2.728962897098052,
+      "grad_norm": 1.1828432083129883,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 16880
+    },
+    {
+      "epoch": 2.730579581278797,
+      "grad_norm": 0.9989302754402161,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 16890
+    },
+    {
+      "epoch": 2.7321962654595424,
+      "grad_norm": 0.8100560307502747,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 16900
+    },
+    {
+      "epoch": 2.7338129496402876,
+      "grad_norm": 0.8615233898162842,
+      "learning_rate": 0.0002,
+      "loss": 0.6851,
+      "step": 16910
+    },
+    {
+      "epoch": 2.735429633821033,
+      "grad_norm": 0.8633756041526794,
+      "learning_rate": 0.0002,
+      "loss": 0.6322,
+      "step": 16920
+    },
+    {
+      "epoch": 2.737046318001778,
+      "grad_norm": 0.7769348621368408,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 16930
+    },
+    {
+      "epoch": 2.738663002182524,
+      "grad_norm": 0.6943058371543884,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 16940
+    },
+    {
+      "epoch": 2.740279686363269,
+      "grad_norm": 0.8510736227035522,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 16950
+    },
+    {
+      "epoch": 2.7418963705440142,
+      "grad_norm": 0.7732602953910828,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 16960
+    },
+    {
+      "epoch": 2.7435130547247595,
+      "grad_norm": 0.5981788635253906,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 16970
+    },
+    {
+      "epoch": 2.7451297389055047,
+      "grad_norm": 0.7604416012763977,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16980
+    },
+    {
+      "epoch": 2.74674642308625,
+      "grad_norm": 0.7377738356590271,
+      "learning_rate": 0.0002,
+      "loss": 0.6981,
+      "step": 16990
+    },
+    {
+      "epoch": 2.7483631072669956,
+      "grad_norm": 0.9400289058685303,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 17000
+    },
+    {
+      "epoch": 2.749979791447741,
+      "grad_norm": 0.6340599656105042,
+      "learning_rate": 0.0002,
+      "loss": 0.6952,
+      "step": 17010
+    },
+    {
+      "epoch": 2.751596475628486,
+      "grad_norm": 0.7297601103782654,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 17020
+    },
+    {
+      "epoch": 2.7532131598092313,
+      "grad_norm": 0.9479979872703552,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 17030
+    },
+    {
+      "epoch": 2.7548298439899765,
+      "grad_norm": 0.8461511135101318,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 17040
+    },
+    {
+      "epoch": 2.7564465281707218,
+      "grad_norm": 0.7477551698684692,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17050
+    },
+    {
+      "epoch": 2.758063212351467,
+      "grad_norm": 1.019270420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 17060
+    },
+    {
+      "epoch": 2.7596798965322122,
+      "grad_norm": 0.7730235457420349,
+      "learning_rate": 0.0002,
+      "loss": 0.7119,
+      "step": 17070
+    },
+    {
+      "epoch": 2.7612965807129575,
+      "grad_norm": 0.8216866254806519,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 17080
+    },
+    {
+      "epoch": 2.762913264893703,
+      "grad_norm": 0.7235931754112244,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17090
+    },
+    {
+      "epoch": 2.7645299490744484,
+      "grad_norm": 0.7352296710014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7031,
+      "step": 17100
+    },
+    {
+      "epoch": 2.7661466332551936,
+      "grad_norm": 0.8129373788833618,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 17110
+    },
+    {
+      "epoch": 2.767763317435939,
+      "grad_norm": 0.7387019991874695,
+      "learning_rate": 0.0002,
+      "loss": 0.6703,
+      "step": 17120
+    },
+    {
+      "epoch": 2.769380001616684,
+      "grad_norm": 0.9149190187454224,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 17130
+    },
+    {
+      "epoch": 2.7709966857974297,
+      "grad_norm": 0.7352971434593201,
+      "learning_rate": 0.0002,
+      "loss": 0.6038,
+      "step": 17140
+    },
+    {
+      "epoch": 2.772613369978175,
+      "grad_norm": 0.7903780341148376,
+      "learning_rate": 0.0002,
+      "loss": 0.6728,
+      "step": 17150
+    },
+    {
+      "epoch": 2.77423005415892,
+      "grad_norm": 0.8255927562713623,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17160
+    },
+    {
+      "epoch": 2.7758467383396654,
+      "grad_norm": 0.7235927581787109,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 17170
+    },
+    {
+      "epoch": 2.7774634225204107,
+      "grad_norm": 0.8281434774398804,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 17180
+    },
+    {
+      "epoch": 2.779080106701156,
+      "grad_norm": 0.7586921453475952,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 17190
+    },
+    {
+      "epoch": 2.780696790881901,
+      "grad_norm": 0.7161715030670166,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 17200
+    },
+    {
+      "epoch": 2.7823134750626464,
+      "grad_norm": 0.762868344783783,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 17210
+    },
+    {
+      "epoch": 2.7839301592433916,
+      "grad_norm": 0.9285483360290527,
+      "learning_rate": 0.0002,
+      "loss": 0.705,
+      "step": 17220
+    },
+    {
+      "epoch": 2.785546843424137,
+      "grad_norm": 0.6900462508201599,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 17230
+    },
+    {
+      "epoch": 2.7871635276048825,
+      "grad_norm": 0.780384361743927,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17240
+    },
+    {
+      "epoch": 2.7887802117856277,
+      "grad_norm": 0.7580406665802002,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17250
+    },
+    {
+      "epoch": 2.790396895966373,
+      "grad_norm": 0.8145199418067932,
+      "learning_rate": 0.0002,
+      "loss": 0.6833,
+      "step": 17260
+    },
+    {
+      "epoch": 2.792013580147118,
+      "grad_norm": 0.9159596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 17270
+    },
+    {
+      "epoch": 2.7936302643278634,
+      "grad_norm": 0.9590014219284058,
+      "learning_rate": 0.0002,
+      "loss": 0.6008,
+      "step": 17280
+    },
+    {
+      "epoch": 2.795246948508609,
+      "grad_norm": 0.7603529691696167,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 17290
+    },
+    {
+      "epoch": 2.7968636326893543,
+      "grad_norm": 0.8039976358413696,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 17300
+    },
+    {
+      "epoch": 2.7984803168700996,
+      "grad_norm": 0.8364847302436829,
+      "learning_rate": 0.0002,
+      "loss": 0.7037,
+      "step": 17310
+    },
+    {
+      "epoch": 2.800097001050845,
+      "grad_norm": 0.8763046860694885,
+      "learning_rate": 0.0002,
+      "loss": 0.6749,
+      "step": 17320
+    },
+    {
+      "epoch": 2.80171368523159,
+      "grad_norm": 0.8409647941589355,
+      "learning_rate": 0.0002,
+      "loss": 0.6844,
+      "step": 17330
+    },
+    {
+      "epoch": 2.8033303694123353,
+      "grad_norm": 0.7649006247520447,
+      "learning_rate": 0.0002,
+      "loss": 0.6936,
+      "step": 17340
+    },
+    {
+      "epoch": 2.8049470535930805,
+      "grad_norm": 0.7970262169837952,
+      "learning_rate": 0.0002,
+      "loss": 0.7051,
+      "step": 17350
+    },
+    {
+      "epoch": 2.8065637377738257,
+      "grad_norm": 0.9088607430458069,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 17360
+    },
+    {
+      "epoch": 2.808180421954571,
+      "grad_norm": 0.6454846858978271,
+      "learning_rate": 0.0002,
+      "loss": 0.675,
+      "step": 17370
+    },
+    {
+      "epoch": 2.809797106135316,
+      "grad_norm": 0.7744787931442261,
+      "learning_rate": 0.0002,
+      "loss": 0.7069,
+      "step": 17380
+    },
+    {
+      "epoch": 2.811413790316062,
+      "grad_norm": 0.6678640842437744,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 17390
+    },
+    {
+      "epoch": 2.813030474496807,
+      "grad_norm": 0.772676944732666,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 17400
+    },
+    {
+      "epoch": 2.8146471586775523,
+      "grad_norm": 0.7088175415992737,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 17410
+    },
+    {
+      "epoch": 2.8162638428582976,
+      "grad_norm": 0.8280573487281799,
+      "learning_rate": 0.0002,
+      "loss": 0.7086,
+      "step": 17420
+    },
+    {
+      "epoch": 2.817880527039043,
+      "grad_norm": 0.6665388345718384,
+      "learning_rate": 0.0002,
+      "loss": 0.6732,
+      "step": 17430
+    },
+    {
+      "epoch": 2.8194972112197885,
+      "grad_norm": 0.6427883505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 17440
+    },
+    {
+      "epoch": 2.8211138954005337,
+      "grad_norm": 0.9697760343551636,
+      "learning_rate": 0.0002,
+      "loss": 0.6972,
+      "step": 17450
+    },
+    {
+      "epoch": 2.822730579581279,
+      "grad_norm": 0.7573966383934021,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 17460
+    },
+    {
+      "epoch": 2.824347263762024,
+      "grad_norm": 0.878688633441925,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 17470
+    },
+    {
+      "epoch": 2.8259639479427694,
+      "grad_norm": 0.7752242684364319,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 17480
+    },
+    {
+      "epoch": 2.8275806321235146,
+      "grad_norm": 0.6135398745536804,
+      "learning_rate": 0.0002,
+      "loss": 0.6638,
+      "step": 17490
+    },
+    {
+      "epoch": 2.82919731630426,
+      "grad_norm": 0.6924924850463867,
+      "learning_rate": 0.0002,
+      "loss": 0.6829,
+      "step": 17500
+    },
+    {
+      "epoch": 2.830814000485005,
+      "grad_norm": 0.7471627593040466,
+      "learning_rate": 0.0002,
+      "loss": 0.6731,
+      "step": 17510
+    },
+    {
+      "epoch": 2.8324306846657503,
+      "grad_norm": 0.7145499587059021,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 17520
+    },
+    {
+      "epoch": 2.834047368846496,
+      "grad_norm": 0.7415414452552795,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17530
+    },
+    {
+      "epoch": 2.8356640530272412,
+      "grad_norm": 0.7328441739082336,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17540
+    },
+    {
+      "epoch": 2.8372807372079865,
+      "grad_norm": 0.8267839550971985,
+      "learning_rate": 0.0002,
+      "loss": 0.6866,
+      "step": 17550
+    },
+    {
+      "epoch": 2.8388974213887317,
+      "grad_norm": 0.8877885341644287,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17560
+    },
+    {
+      "epoch": 2.840514105569477,
+      "grad_norm": 0.857138454914093,
+      "learning_rate": 0.0002,
+      "loss": 0.7136,
+      "step": 17570
+    },
+    {
+      "epoch": 2.842130789750222,
+      "grad_norm": 0.8470779657363892,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 17580
+    },
+    {
+      "epoch": 2.843747473930968,
+      "grad_norm": 0.8553254008293152,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 17590
+    },
+    {
+      "epoch": 2.845364158111713,
+      "grad_norm": 0.8033196926116943,
+      "learning_rate": 0.0002,
+      "loss": 0.7297,
+      "step": 17600
+    },
+    {
+      "epoch": 2.8469808422924583,
+      "grad_norm": 0.7949087023735046,
+      "learning_rate": 0.0002,
+      "loss": 0.7062,
+      "step": 17610
+    },
+    {
+      "epoch": 2.8485975264732035,
+      "grad_norm": 0.9241406321525574,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 17620
+    },
+    {
+      "epoch": 2.8502142106539488,
+      "grad_norm": 0.7721285223960876,
+      "learning_rate": 0.0002,
+      "loss": 0.6601,
+      "step": 17630
+    },
+    {
+      "epoch": 2.851830894834694,
+      "grad_norm": 1.0246692895889282,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 17640
+    },
+    {
+      "epoch": 2.853447579015439,
+      "grad_norm": 0.9244589805603027,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 17650
+    },
+    {
+      "epoch": 2.8550642631961844,
+      "grad_norm": 0.7243508696556091,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 17660
+    },
+    {
+      "epoch": 2.8566809473769297,
+      "grad_norm": 0.8943371176719666,
+      "learning_rate": 0.0002,
+      "loss": 0.6471,
+      "step": 17670
+    },
+    {
+      "epoch": 2.8582976315576754,
+      "grad_norm": 0.6531758904457092,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 17680
+    },
+    {
+      "epoch": 2.8599143157384206,
+      "grad_norm": 0.8367000818252563,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 17690
+    },
+    {
+      "epoch": 2.861530999919166,
+      "grad_norm": 0.7868556380271912,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 17700
+    },
+    {
+      "epoch": 2.863147684099911,
+      "grad_norm": 0.7213859558105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 17710
+    },
+    {
+      "epoch": 2.8647643682806563,
+      "grad_norm": 0.7383931279182434,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 17720
+    },
+    {
+      "epoch": 2.8663810524614015,
+      "grad_norm": 0.7566812634468079,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 17730
+    },
+    {
+      "epoch": 2.867997736642147,
+      "grad_norm": 0.6930373311042786,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 17740
+    },
+    {
+      "epoch": 2.8696144208228924,
+      "grad_norm": 0.7911090850830078,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17750
+    },
+    {
+      "epoch": 2.8712311050036377,
+      "grad_norm": 0.8484548926353455,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 17760
+    },
+    {
+      "epoch": 2.872847789184383,
+      "grad_norm": 0.7647597193717957,
+      "learning_rate": 0.0002,
+      "loss": 0.6931,
+      "step": 17770
+    },
+    {
+      "epoch": 2.874464473365128,
+      "grad_norm": 0.8791151642799377,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 17780
+    },
+    {
+      "epoch": 2.8760811575458733,
+      "grad_norm": 0.7253178358078003,
+      "learning_rate": 0.0002,
+      "loss": 0.7078,
+      "step": 17790
+    },
+    {
+      "epoch": 2.8776978417266186,
+      "grad_norm": 0.7956077456474304,
+      "learning_rate": 0.0002,
+      "loss": 0.6474,
+      "step": 17800
+    },
+    {
+      "epoch": 2.879314525907364,
+      "grad_norm": 0.8657688498497009,
+      "learning_rate": 0.0002,
+      "loss": 0.6687,
+      "step": 17810
+    },
+    {
+      "epoch": 2.880931210088109,
+      "grad_norm": 0.7059141993522644,
+      "learning_rate": 0.0002,
+      "loss": 0.7171,
+      "step": 17820
+    },
+    {
+      "epoch": 2.8825478942688547,
+      "grad_norm": 0.8886896967887878,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 17830
+    },
+    {
+      "epoch": 2.8841645784496,
+      "grad_norm": 0.821032702922821,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 17840
+    },
+    {
+      "epoch": 2.885781262630345,
+      "grad_norm": 0.7183963656425476,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 17850
+    },
+    {
+      "epoch": 2.8873979468110904,
+      "grad_norm": 0.6222899556159973,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 17860
+    },
+    {
+      "epoch": 2.8890146309918356,
+      "grad_norm": 0.8187434077262878,
+      "learning_rate": 0.0002,
+      "loss": 0.6626,
+      "step": 17870
+    },
+    {
+      "epoch": 2.890631315172581,
+      "grad_norm": 0.9838479161262512,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 17880
+    },
+    {
+      "epoch": 2.8922479993533265,
+      "grad_norm": 0.7567742466926575,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 17890
+    },
+    {
+      "epoch": 2.893864683534072,
+      "grad_norm": 0.6875903606414795,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17900
+    },
+    {
+      "epoch": 2.895481367714817,
+      "grad_norm": 0.8043789267539978,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 17910
+    },
+    {
+      "epoch": 2.8970980518955622,
+      "grad_norm": 0.8062626719474792,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 17920
+    },
+    {
+      "epoch": 2.8987147360763075,
+      "grad_norm": 1.0251191854476929,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 17930
+    },
+    {
+      "epoch": 2.9003314202570527,
+      "grad_norm": 0.882253110408783,
+      "learning_rate": 0.0002,
+      "loss": 0.6605,
+      "step": 17940
+    },
+    {
+      "epoch": 2.901948104437798,
+      "grad_norm": 0.8683299422264099,
+      "learning_rate": 0.0002,
+      "loss": 0.6719,
+      "step": 17950
+    },
+    {
+      "epoch": 2.903564788618543,
+      "grad_norm": 0.7167282104492188,
+      "learning_rate": 0.0002,
+      "loss": 0.6896,
+      "step": 17960
+    },
+    {
+      "epoch": 2.9051814727992884,
+      "grad_norm": 0.7093694806098938,
+      "learning_rate": 0.0002,
+      "loss": 0.663,
+      "step": 17970
+    },
+    {
+      "epoch": 2.906798156980034,
+      "grad_norm": 0.8549879193305969,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 17980
+    },
+    {
+      "epoch": 2.9084148411607793,
+      "grad_norm": 0.6989606618881226,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 17990
+    },
+    {
+      "epoch": 2.9100315253415245,
+      "grad_norm": 0.9482976794242859,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 18000
+    },
+    {
+      "epoch": 2.9116482095222698,
+      "grad_norm": 0.7182440161705017,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 18010
+    },
+    {
+      "epoch": 2.913264893703015,
+      "grad_norm": 0.7732226252555847,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 18020
+    },
+    {
+      "epoch": 2.9148815778837607,
+      "grad_norm": 0.7936875224113464,
+      "learning_rate": 0.0002,
+      "loss": 0.7123,
+      "step": 18030
+    },
+    {
+      "epoch": 2.916498262064506,
+      "grad_norm": 0.8825615644454956,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 18040
+    },
+    {
+      "epoch": 2.918114946245251,
+      "grad_norm": 0.6778587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 18050
+    },
+    {
+      "epoch": 2.9197316304259964,
+      "grad_norm": 0.7529265880584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 18060
+    },
+    {
+      "epoch": 2.9213483146067416,
+      "grad_norm": 0.7111883163452148,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 18070
+    },
+    {
+      "epoch": 2.922964998787487,
+      "grad_norm": 0.7214767932891846,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 18080
+    },
+    {
+      "epoch": 2.924581682968232,
+      "grad_norm": 0.800417423248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 18090
+    },
+    {
+      "epoch": 2.9261983671489773,
+      "grad_norm": 1.248575210571289,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 18100
+    },
+    {
+      "epoch": 2.9278150513297225,
+      "grad_norm": 0.757788360118866,
+      "learning_rate": 0.0002,
+      "loss": 0.7004,
+      "step": 18110
+    },
+    {
+      "epoch": 2.9294317355104678,
+      "grad_norm": 1.0583995580673218,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 18120
+    },
+    {
+      "epoch": 2.9310484196912134,
+      "grad_norm": 0.8228777647018433,
+      "learning_rate": 0.0002,
+      "loss": 0.6365,
+      "step": 18130
+    },
+    {
+      "epoch": 2.9326651038719587,
+      "grad_norm": 0.8374035358428955,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 18140
+    },
+    {
+      "epoch": 2.934281788052704,
+      "grad_norm": 0.7976473569869995,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 18150
+    },
+    {
+      "epoch": 2.935898472233449,
+      "grad_norm": 0.8009907603263855,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 18160
+    },
+    {
+      "epoch": 2.9375151564141944,
+      "grad_norm": 0.835213303565979,
+      "learning_rate": 0.0002,
+      "loss": 0.7485,
+      "step": 18170
+    },
+    {
+      "epoch": 2.93913184059494,
+      "grad_norm": 0.7982219457626343,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 18180
+    },
+    {
+      "epoch": 2.9407485247756853,
+      "grad_norm": 0.7070978879928589,
+      "learning_rate": 0.0002,
+      "loss": 0.6348,
+      "step": 18190
+    },
+    {
+      "epoch": 2.9423652089564305,
+      "grad_norm": 0.8619440197944641,
+      "learning_rate": 0.0002,
+      "loss": 0.6608,
+      "step": 18200
+    },
+    {
+      "epoch": 2.9439818931371757,
+      "grad_norm": 0.6693987250328064,
+      "learning_rate": 0.0002,
+      "loss": 0.666,
+      "step": 18210
+    },
+    {
+      "epoch": 2.945598577317921,
+      "grad_norm": 0.6747021079063416,
+      "learning_rate": 0.0002,
+      "loss": 0.728,
+      "step": 18220
+    },
+    {
+      "epoch": 2.947215261498666,
+      "grad_norm": 0.860387921333313,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 18230
+    },
+    {
+      "epoch": 2.9488319456794114,
+      "grad_norm": 0.799976646900177,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 18240
+    },
+    {
+      "epoch": 2.9504486298601567,
+      "grad_norm": 0.7864769101142883,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 18250
+    },
+    {
+      "epoch": 2.952065314040902,
+      "grad_norm": 0.6713884472846985,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 18260
+    },
+    {
+      "epoch": 2.9536819982216476,
+      "grad_norm": 0.9031508564949036,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 18270
+    },
+    {
+      "epoch": 2.955298682402393,
+      "grad_norm": 0.7205073237419128,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 18280
+    },
+    {
+      "epoch": 2.956915366583138,
+      "grad_norm": 0.7746205925941467,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 18290
+    },
+    {
+      "epoch": 2.9585320507638833,
+      "grad_norm": 0.6533427834510803,
+      "learning_rate": 0.0002,
+      "loss": 0.6218,
+      "step": 18300
+    },
+    {
+      "epoch": 2.9601487349446285,
+      "grad_norm": 0.9083208441734314,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 18310
+    },
+    {
+      "epoch": 2.9617654191253737,
+      "grad_norm": 0.7446991801261902,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 18320
+    },
+    {
+      "epoch": 2.9633821033061194,
+      "grad_norm": 0.6514461636543274,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 18330
+    },
+    {
+      "epoch": 2.9649987874868646,
+      "grad_norm": 0.8580465912818909,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18340
+    },
+    {
+      "epoch": 2.96661547166761,
+      "grad_norm": 0.7074266076087952,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 18350
+    },
+    {
+      "epoch": 2.968232155848355,
+      "grad_norm": 0.899892270565033,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 18360
+    },
+    {
+      "epoch": 2.9698488400291003,
+      "grad_norm": 0.8217641711235046,
+      "learning_rate": 0.0002,
+      "loss": 0.7094,
+      "step": 18370
+    },
+    {
+      "epoch": 2.9714655242098456,
+      "grad_norm": 0.8611799478530884,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 18380
+    },
+    {
+      "epoch": 2.973082208390591,
+      "grad_norm": 0.6909302473068237,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18390
+    },
+    {
+      "epoch": 2.974698892571336,
+      "grad_norm": 0.6554358005523682,
+      "learning_rate": 0.0002,
+      "loss": 0.7247,
+      "step": 18400
+    },
+    {
+      "epoch": 2.9763155767520812,
+      "grad_norm": 0.7803071737289429,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 18410
+    },
+    {
+      "epoch": 2.977932260932827,
+      "grad_norm": 0.7838954925537109,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 18420
+    },
+    {
+      "epoch": 2.979548945113572,
+      "grad_norm": 0.7098495364189148,
+      "learning_rate": 0.0002,
+      "loss": 0.6522,
+      "step": 18430
+    },
+    {
+      "epoch": 2.9811656292943174,
+      "grad_norm": 0.8981785774230957,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 18440
+    },
+    {
+      "epoch": 2.9827823134750626,
+      "grad_norm": 0.7197171449661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6689,
+      "step": 18450
+    },
+    {
+      "epoch": 2.984398997655808,
+      "grad_norm": 0.793185293674469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 18460
+    },
+    {
+      "epoch": 2.986015681836553,
+      "grad_norm": 0.8531473875045776,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 18470
+    },
+    {
+      "epoch": 2.9876323660172988,
+      "grad_norm": 0.6627361178398132,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 18480
+    },
+    {
+      "epoch": 2.989249050198044,
+      "grad_norm": 0.5708155035972595,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 18490
+    },
+    {
+      "epoch": 2.990865734378789,
+      "grad_norm": 0.8227280378341675,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 18500
+    },
+    {
+      "epoch": 2.9924824185595345,
+      "grad_norm": 0.7102749943733215,
+      "learning_rate": 0.0002,
+      "loss": 0.6701,
+      "step": 18510
+    },
+    {
+      "epoch": 2.9940991027402797,
+      "grad_norm": 0.839485228061676,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 18520
+    },
+    {
+      "epoch": 2.995715786921025,
+      "grad_norm": 0.9038704037666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 18530
+    },
+    {
+      "epoch": 2.99733247110177,
+      "grad_norm": 0.8737510442733765,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 18540
+    },
+    {
+      "epoch": 2.9989491552825154,
+      "grad_norm": 0.7323142886161804,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 18550
+    },
+    {
+      "epoch": 2.9999191657909625,
+      "eval_loss": 1.1262480020523071,
+      "eval_runtime": 122.0868,
+      "eval_samples_per_second": 6.004,
+      "eval_steps_per_second": 0.754,
+      "step": 18556
+    },
+    {
+      "epoch": 3.000565839463261,
+      "grad_norm": 0.8465463519096375,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 18560
+    },
+    {
+      "epoch": 3.0021825236440063,
+      "grad_norm": 0.9134138822555542,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 18570
+    },
+    {
+      "epoch": 3.0037992078247515,
+      "grad_norm": 0.760715126991272,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 18580
+    },
+    {
+      "epoch": 3.0054158920054967,
+      "grad_norm": 0.9208743572235107,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18590
+    },
+    {
+      "epoch": 3.007032576186242,
+      "grad_norm": 0.9232364892959595,
+      "learning_rate": 0.0002,
+      "loss": 0.5731,
+      "step": 18600
+    },
+    {
+      "epoch": 3.008649260366987,
+      "grad_norm": 1.1881544589996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6299,
+      "step": 18610
+    },
+    {
+      "epoch": 3.0102659445477324,
+      "grad_norm": 0.9372987747192383,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 18620
+    },
+    {
+      "epoch": 3.0118826287284777,
+      "grad_norm": 0.6900241374969482,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 18630
+    },
+    {
+      "epoch": 3.0134993129092233,
+      "grad_norm": 0.8451071381568909,
+      "learning_rate": 0.0002,
+      "loss": 0.5256,
+      "step": 18640
+    },
+    {
+      "epoch": 3.0151159970899686,
+      "grad_norm": 0.7763112187385559,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 18650
+    },
+    {
+      "epoch": 3.016732681270714,
+      "grad_norm": 1.043653964996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 18660
+    },
+    {
+      "epoch": 3.018349365451459,
+      "grad_norm": 1.0170660018920898,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 18670
+    },
+    {
+      "epoch": 3.0199660496322043,
+      "grad_norm": 0.7534180283546448,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 18680
+    },
+    {
+      "epoch": 3.0215827338129495,
+      "grad_norm": 0.7507367730140686,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 18690
+    },
+    {
+      "epoch": 3.0231994179936947,
+      "grad_norm": 0.7861620187759399,
+      "learning_rate": 0.0002,
+      "loss": 0.6201,
+      "step": 18700
+    },
+    {
+      "epoch": 3.0248161021744404,
+      "grad_norm": 1.0580339431762695,
+      "learning_rate": 0.0002,
+      "loss": 0.5802,
+      "step": 18710
+    },
+    {
+      "epoch": 3.0264327863551856,
+      "grad_norm": 0.7542710900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 18720
+    },
+    {
+      "epoch": 3.028049470535931,
+      "grad_norm": 0.8189544677734375,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 18730
+    },
+    {
+      "epoch": 3.029666154716676,
+      "grad_norm": 0.9126611351966858,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 18740
+    },
+    {
+      "epoch": 3.0312828388974213,
+      "grad_norm": 0.8891341686248779,
+      "learning_rate": 0.0002,
+      "loss": 0.6443,
+      "step": 18750
+    },
+    {
+      "epoch": 3.0328995230781666,
+      "grad_norm": 0.8419283032417297,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 18760
+    },
+    {
+      "epoch": 3.034516207258912,
+      "grad_norm": 0.8048048615455627,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18770
+    },
+    {
+      "epoch": 3.0361328914396575,
+      "grad_norm": 0.7820217609405518,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 18780
+    },
+    {
+      "epoch": 3.0377495756204027,
+      "grad_norm": 0.854721188545227,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 18790
+    },
+    {
+      "epoch": 3.039366259801148,
+      "grad_norm": 0.912092924118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5784,
+      "step": 18800
+    },
+    {
+      "epoch": 3.040982943981893,
+      "grad_norm": 0.6596226096153259,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 18810
+    },
+    {
+      "epoch": 3.0425996281626384,
+      "grad_norm": 0.6351348757743835,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 18820
+    },
+    {
+      "epoch": 3.0442163123433836,
+      "grad_norm": 0.778188943862915,
+      "learning_rate": 0.0002,
+      "loss": 0.5953,
+      "step": 18830
+    },
+    {
+      "epoch": 3.045832996524129,
+      "grad_norm": 0.68234783411026,
+      "learning_rate": 0.0002,
+      "loss": 0.602,
+      "step": 18840
+    },
+    {
+      "epoch": 3.047449680704874,
+      "grad_norm": 0.998628556728363,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 18850
+    },
+    {
+      "epoch": 3.0490663648856198,
+      "grad_norm": 0.7393841743469238,
+      "learning_rate": 0.0002,
+      "loss": 0.6231,
+      "step": 18860
+    },
+    {
+      "epoch": 3.050683049066365,
+      "grad_norm": 0.84438556432724,
+      "learning_rate": 0.0002,
+      "loss": 0.568,
+      "step": 18870
+    },
+    {
+      "epoch": 3.0522997332471102,
+      "grad_norm": 0.8857501745223999,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 18880
+    },
+    {
+      "epoch": 3.0539164174278555,
+      "grad_norm": 0.7208474278450012,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 18890
+    },
+    {
+      "epoch": 3.0555331016086007,
+      "grad_norm": 0.7135229110717773,
+      "learning_rate": 0.0002,
+      "loss": 0.5998,
+      "step": 18900
+    },
+    {
+      "epoch": 3.057149785789346,
+      "grad_norm": 0.9130001664161682,
+      "learning_rate": 0.0002,
+      "loss": 0.5575,
+      "step": 18910
+    },
+    {
+      "epoch": 3.058766469970091,
+      "grad_norm": 0.9001716375350952,
+      "learning_rate": 0.0002,
+      "loss": 0.5955,
+      "step": 18920
+    },
+    {
+      "epoch": 3.060383154150837,
+      "grad_norm": 0.8667559623718262,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 18930
+    },
+    {
+      "epoch": 3.061999838331582,
+      "grad_norm": 0.8943959474563599,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18940
+    },
+    {
+      "epoch": 3.0636165225123273,
+      "grad_norm": 0.8298377990722656,
+      "learning_rate": 0.0002,
+      "loss": 0.5978,
+      "step": 18950
+    },
+    {
+      "epoch": 3.0652332066930725,
+      "grad_norm": 0.7935267686843872,
+      "learning_rate": 0.0002,
+      "loss": 0.5782,
+      "step": 18960
+    },
+    {
+      "epoch": 3.0668498908738178,
+      "grad_norm": 1.1506379842758179,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 18970
+    },
+    {
+      "epoch": 3.068466575054563,
+      "grad_norm": 0.7693049907684326,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18980
+    },
+    {
+      "epoch": 3.0700832592353082,
+      "grad_norm": 0.8040135502815247,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 18990
+    },
+    {
+      "epoch": 3.0716999434160535,
+      "grad_norm": 0.828404426574707,
+      "learning_rate": 0.0002,
+      "loss": 0.5541,
+      "step": 19000
+    },
+    {
+      "epoch": 3.073316627596799,
+      "grad_norm": 0.8811164498329163,
+      "learning_rate": 0.0002,
+      "loss": 0.6048,
+      "step": 19010
+    },
+    {
+      "epoch": 3.0749333117775444,
+      "grad_norm": 1.036205768585205,
+      "learning_rate": 0.0002,
+      "loss": 0.5845,
+      "step": 19020
+    },
+    {
+      "epoch": 3.0765499959582896,
+      "grad_norm": 0.8857285976409912,
+      "learning_rate": 0.0002,
+      "loss": 0.5838,
+      "step": 19030
+    },
+    {
+      "epoch": 3.078166680139035,
+      "grad_norm": 0.8392079472541809,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19040
+    },
+    {
+      "epoch": 3.07978336431978,
+      "grad_norm": 1.0287401676177979,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 19050
+    },
+    {
+      "epoch": 3.0814000485005253,
+      "grad_norm": 1.0086315870285034,
+      "learning_rate": 0.0002,
+      "loss": 0.5964,
+      "step": 19060
+    },
+    {
+      "epoch": 3.0830167326812705,
+      "grad_norm": 0.9245324730873108,
+      "learning_rate": 0.0002,
+      "loss": 0.5567,
+      "step": 19070
+    },
+    {
+      "epoch": 3.084633416862016,
+      "grad_norm": 0.8680877089500427,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 19080
+    },
+    {
+      "epoch": 3.0862501010427614,
+      "grad_norm": 0.8814793825149536,
+      "learning_rate": 0.0002,
+      "loss": 0.5611,
+      "step": 19090
+    },
+    {
+      "epoch": 3.0878667852235067,
+      "grad_norm": 0.9234458208084106,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19100
+    },
+    {
+      "epoch": 3.089483469404252,
+      "grad_norm": 1.1291664838790894,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 19110
+    },
+    {
+      "epoch": 3.091100153584997,
+      "grad_norm": 0.9191402792930603,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 19120
+    },
+    {
+      "epoch": 3.0927168377657424,
+      "grad_norm": 0.7103154063224792,
+      "learning_rate": 0.0002,
+      "loss": 0.5856,
+      "step": 19130
+    },
+    {
+      "epoch": 3.0943335219464876,
+      "grad_norm": 0.9368883967399597,
+      "learning_rate": 0.0002,
+      "loss": 0.6479,
+      "step": 19140
+    },
+    {
+      "epoch": 3.095950206127233,
+      "grad_norm": 0.9676656723022461,
+      "learning_rate": 0.0002,
+      "loss": 0.6167,
+      "step": 19150
+    },
+    {
+      "epoch": 3.0975668903079785,
+      "grad_norm": 0.8739792704582214,
+      "learning_rate": 0.0002,
+      "loss": 0.5794,
+      "step": 19160
+    },
+    {
+      "epoch": 3.0991835744887237,
+      "grad_norm": 0.8530174493789673,
+      "learning_rate": 0.0002,
+      "loss": 0.6112,
+      "step": 19170
+    },
+    {
+      "epoch": 3.100800258669469,
+      "grad_norm": 0.794945478439331,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 19180
+    },
+    {
+      "epoch": 3.102416942850214,
+      "grad_norm": 0.9508888125419617,
+      "learning_rate": 0.0002,
+      "loss": 0.5928,
+      "step": 19190
+    },
+    {
+      "epoch": 3.1040336270309594,
+      "grad_norm": 1.0599955320358276,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 19200
+    },
+    {
+      "epoch": 3.1056503112117047,
+      "grad_norm": 1.0673625469207764,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 19210
+    },
+    {
+      "epoch": 3.10726699539245,
+      "grad_norm": 0.7739115953445435,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 19220
+    },
+    {
+      "epoch": 3.1088836795731956,
+      "grad_norm": 0.9884951114654541,
+      "learning_rate": 0.0002,
+      "loss": 0.6046,
+      "step": 19230
+    },
+    {
+      "epoch": 3.110500363753941,
+      "grad_norm": 0.862260103225708,
+      "learning_rate": 0.0002,
+      "loss": 0.5932,
+      "step": 19240
+    },
+    {
+      "epoch": 3.112117047934686,
+      "grad_norm": 0.7690284848213196,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 19250
+    },
+    {
+      "epoch": 3.1137337321154313,
+      "grad_norm": 0.8758958578109741,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 19260
+    },
+    {
+      "epoch": 3.1153504162961765,
+      "grad_norm": 1.0356395244598389,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 19270
+    },
+    {
+      "epoch": 3.1169671004769217,
+      "grad_norm": 0.6950937509536743,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 19280
+    },
+    {
+      "epoch": 3.118583784657667,
+      "grad_norm": 0.760998010635376,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19290
+    },
+    {
+      "epoch": 3.1202004688384126,
+      "grad_norm": 0.9335789084434509,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 19300
+    },
+    {
+      "epoch": 3.121817153019158,
+      "grad_norm": 0.9636204242706299,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 19310
+    },
+    {
+      "epoch": 3.123433837199903,
+      "grad_norm": 1.0820997953414917,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 19320
+    },
+    {
+      "epoch": 3.1250505213806483,
+      "grad_norm": 0.7333487272262573,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 19330
+    },
+    {
+      "epoch": 3.1266672055613935,
+      "grad_norm": 1.0417509078979492,
+      "learning_rate": 0.0002,
+      "loss": 0.6178,
+      "step": 19340
+    },
+    {
+      "epoch": 3.128283889742139,
+      "grad_norm": 0.9267749190330505,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 19350
+    },
+    {
+      "epoch": 3.129900573922884,
+      "grad_norm": 0.777798593044281,
+      "learning_rate": 0.0002,
+      "loss": 0.6063,
+      "step": 19360
+    },
+    {
+      "epoch": 3.1315172581036297,
+      "grad_norm": 0.8425456881523132,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 19370
+    },
+    {
+      "epoch": 3.133133942284375,
+      "grad_norm": 0.9617102146148682,
+      "learning_rate": 0.0002,
+      "loss": 0.6042,
+      "step": 19380
+    },
+    {
+      "epoch": 3.13475062646512,
+      "grad_norm": 1.0052828788757324,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 19390
+    },
+    {
+      "epoch": 3.1363673106458654,
+      "grad_norm": 0.7637009024620056,
+      "learning_rate": 0.0002,
+      "loss": 0.5713,
+      "step": 19400
+    },
+    {
+      "epoch": 3.1379839948266106,
+      "grad_norm": 0.7958088517189026,
+      "learning_rate": 0.0002,
+      "loss": 0.5497,
+      "step": 19410
+    },
+    {
+      "epoch": 3.139600679007356,
+      "grad_norm": 0.9161727428436279,
+      "learning_rate": 0.0002,
+      "loss": 0.6283,
+      "step": 19420
+    },
+    {
+      "epoch": 3.141217363188101,
+      "grad_norm": 0.8402149677276611,
+      "learning_rate": 0.0002,
+      "loss": 0.5638,
+      "step": 19430
+    },
+    {
+      "epoch": 3.1428340473688463,
+      "grad_norm": 1.0056525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.5848,
+      "step": 19440
+    },
+    {
+      "epoch": 3.144450731549592,
+      "grad_norm": 1.0129190683364868,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 19450
+    },
+    {
+      "epoch": 3.146067415730337,
+      "grad_norm": 0.790825366973877,
+      "learning_rate": 0.0002,
+      "loss": 0.5808,
+      "step": 19460
+    },
+    {
+      "epoch": 3.1476840999110824,
+      "grad_norm": 1.441665530204773,
+      "learning_rate": 0.0002,
+      "loss": 0.5607,
+      "step": 19470
+    },
+    {
+      "epoch": 3.1493007840918277,
+      "grad_norm": 0.7846331596374512,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19480
+    },
+    {
+      "epoch": 3.150917468272573,
+      "grad_norm": 0.7915332913398743,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 19490
+    },
+    {
+      "epoch": 3.152534152453318,
+      "grad_norm": 0.933982253074646,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 19500
+    },
+    {
+      "epoch": 3.1541508366340634,
+      "grad_norm": 1.038408637046814,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 19510
+    },
+    {
+      "epoch": 3.155767520814809,
+      "grad_norm": 1.018935203552246,
+      "learning_rate": 0.0002,
+      "loss": 0.6271,
+      "step": 19520
+    },
+    {
+      "epoch": 3.1573842049955543,
+      "grad_norm": 0.9618112444877625,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 19530
+    },
+    {
+      "epoch": 3.1590008891762995,
+      "grad_norm": 0.8900452852249146,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 19540
+    },
+    {
+      "epoch": 3.1606175733570447,
+      "grad_norm": 0.8254160284996033,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 19550
+    },
+    {
+      "epoch": 3.16223425753779,
+      "grad_norm": 1.004376769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19560
+    },
+    {
+      "epoch": 3.163850941718535,
+      "grad_norm": 1.0490446090698242,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 19570
+    },
+    {
+      "epoch": 3.1654676258992804,
+      "grad_norm": 0.7387403845787048,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19580
+    },
+    {
+      "epoch": 3.1670843100800257,
+      "grad_norm": 0.7611538171768188,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 19590
+    },
+    {
+      "epoch": 3.1687009942607713,
+      "grad_norm": 0.8239886164665222,
+      "learning_rate": 0.0002,
+      "loss": 0.5937,
+      "step": 19600
+    },
+    {
+      "epoch": 3.1703176784415166,
+      "grad_norm": 0.9327243566513062,
+      "learning_rate": 0.0002,
+      "loss": 0.6068,
+      "step": 19610
+    },
+    {
+      "epoch": 3.171934362622262,
+      "grad_norm": 0.9662560224533081,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 19620
+    },
+    {
+      "epoch": 3.173551046803007,
+      "grad_norm": 0.9183341860771179,
+      "learning_rate": 0.0002,
+      "loss": 0.5988,
+      "step": 19630
+    },
+    {
+      "epoch": 3.1751677309837523,
+      "grad_norm": 0.875066876411438,
+      "learning_rate": 0.0002,
+      "loss": 0.5909,
+      "step": 19640
+    },
+    {
+      "epoch": 3.1767844151644975,
+      "grad_norm": 0.8567508459091187,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 19650
+    },
+    {
+      "epoch": 3.1784010993452427,
+      "grad_norm": 0.6805780529975891,
+      "learning_rate": 0.0002,
+      "loss": 0.5805,
+      "step": 19660
+    },
+    {
+      "epoch": 3.1800177835259884,
+      "grad_norm": 0.8776944279670715,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 19670
+    },
+    {
+      "epoch": 3.1816344677067336,
+      "grad_norm": 0.9036329984664917,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 19680
+    },
+    {
+      "epoch": 3.183251151887479,
+      "grad_norm": 0.8527372479438782,
+      "learning_rate": 0.0002,
+      "loss": 0.6238,
+      "step": 19690
+    },
+    {
+      "epoch": 3.184867836068224,
+      "grad_norm": 1.1045585870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 19700
+    },
+    {
+      "epoch": 3.1864845202489693,
+      "grad_norm": 0.9213830828666687,
+      "learning_rate": 0.0002,
+      "loss": 0.5491,
+      "step": 19710
+    },
+    {
+      "epoch": 3.1881012044297146,
+      "grad_norm": 0.8865814805030823,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 19720
+    },
+    {
+      "epoch": 3.18971788861046,
+      "grad_norm": 0.7939388751983643,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19730
+    },
+    {
+      "epoch": 3.191334572791205,
+      "grad_norm": 0.6966729760169983,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 19740
+    },
+    {
+      "epoch": 3.1929512569719507,
+      "grad_norm": 0.8023673295974731,
+      "learning_rate": 0.0002,
+      "loss": 0.5839,
+      "step": 19750
+    },
+    {
+      "epoch": 3.194567941152696,
+      "grad_norm": 0.7992037534713745,
+      "learning_rate": 0.0002,
+      "loss": 0.6267,
+      "step": 19760
+    },
+    {
+      "epoch": 3.196184625333441,
+      "grad_norm": 0.7412247657775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 19770
+    },
+    {
+      "epoch": 3.1978013095141864,
+      "grad_norm": 0.9598729014396667,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 19780
+    },
+    {
+      "epoch": 3.1994179936949316,
+      "grad_norm": 0.8331366777420044,
+      "learning_rate": 0.0002,
+      "loss": 0.5685,
+      "step": 19790
+    },
+    {
+      "epoch": 3.201034677875677,
+      "grad_norm": 0.8939169645309448,
+      "learning_rate": 0.0002,
+      "loss": 0.6104,
+      "step": 19800
+    },
+    {
+      "epoch": 3.202651362056422,
+      "grad_norm": 0.9219734072685242,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 19810
+    },
+    {
+      "epoch": 3.2042680462371678,
+      "grad_norm": 0.869490385055542,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19820
+    },
+    {
+      "epoch": 3.205884730417913,
+      "grad_norm": 0.8989706635475159,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 19830
+    },
+    {
+      "epoch": 3.2075014145986582,
+      "grad_norm": 0.8477165102958679,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 19840
+    },
+    {
+      "epoch": 3.2091180987794035,
+      "grad_norm": 0.8720678687095642,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 19850
+    },
+    {
+      "epoch": 3.2107347829601487,
+      "grad_norm": 0.861406683921814,
+      "learning_rate": 0.0002,
+      "loss": 0.5694,
+      "step": 19860
+    },
+    {
+      "epoch": 3.212351467140894,
+      "grad_norm": 0.8228686451911926,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 19870
+    },
+    {
+      "epoch": 3.213968151321639,
+      "grad_norm": 0.7936596870422363,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19880
+    },
+    {
+      "epoch": 3.2155848355023844,
+      "grad_norm": 1.097377896308899,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 19890
+    },
+    {
+      "epoch": 3.21720151968313,
+      "grad_norm": 0.9544782638549805,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 19900
+    },
+    {
+      "epoch": 3.2188182038638753,
+      "grad_norm": 0.8240751624107361,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 19910
+    },
+    {
+      "epoch": 3.2204348880446205,
+      "grad_norm": 0.8332096338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 19920
+    },
+    {
+      "epoch": 3.2220515722253658,
+      "grad_norm": 1.0954567193984985,
+      "learning_rate": 0.0002,
+      "loss": 0.6274,
+      "step": 19930
+    },
+    {
+      "epoch": 3.223668256406111,
+      "grad_norm": 0.7790525555610657,
+      "learning_rate": 0.0002,
+      "loss": 0.652,
+      "step": 19940
+    },
+    {
+      "epoch": 3.225284940586856,
+      "grad_norm": 0.7966814041137695,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19950
+    },
+    {
+      "epoch": 3.2269016247676015,
+      "grad_norm": 0.9751881957054138,
+      "learning_rate": 0.0002,
+      "loss": 0.5911,
+      "step": 19960
+    },
+    {
+      "epoch": 3.228518308948347,
+      "grad_norm": 0.9856047630310059,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 19970
+    },
+    {
+      "epoch": 3.2301349931290924,
+      "grad_norm": 1.3062353134155273,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 19980
+    },
+    {
+      "epoch": 3.2317516773098376,
+      "grad_norm": 0.9510692358016968,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 19990
+    },
+    {
+      "epoch": 3.233368361490583,
+      "grad_norm": 0.8630342483520508,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 20000
+    },
+    {
+      "epoch": 3.234985045671328,
+      "grad_norm": 0.8966519236564636,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20010
+    },
+    {
+      "epoch": 3.2366017298520733,
+      "grad_norm": 0.7093510627746582,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 20020
+    },
+    {
+      "epoch": 3.2382184140328185,
+      "grad_norm": 0.7771096229553223,
+      "learning_rate": 0.0002,
+      "loss": 0.585,
+      "step": 20030
+    },
+    {
+      "epoch": 3.2398350982135637,
+      "grad_norm": 0.841058075428009,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 20040
+    },
+    {
+      "epoch": 3.2414517823943094,
+      "grad_norm": 0.909712553024292,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 20050
+    },
+    {
+      "epoch": 3.2430684665750547,
+      "grad_norm": 0.8321019411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20060
+    },
+    {
+      "epoch": 3.2446851507558,
+      "grad_norm": 0.779901921749115,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 20070
+    },
+    {
+      "epoch": 3.246301834936545,
+      "grad_norm": 0.6249170303344727,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 20080
+    },
+    {
+      "epoch": 3.2479185191172903,
+      "grad_norm": 0.8000940680503845,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 20090
+    },
+    {
+      "epoch": 3.2495352032980356,
+      "grad_norm": 0.7627735137939453,
+      "learning_rate": 0.0002,
+      "loss": 0.6273,
+      "step": 20100
+    },
+    {
+      "epoch": 3.2511518874787813,
+      "grad_norm": 0.8780747056007385,
+      "learning_rate": 0.0002,
+      "loss": 0.6223,
+      "step": 20110
+    },
+    {
+      "epoch": 3.2527685716595265,
+      "grad_norm": 0.772037148475647,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 20120
+    },
+    {
+      "epoch": 3.2543852558402717,
+      "grad_norm": 1.0086580514907837,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 20130
+    },
+    {
+      "epoch": 3.256001940021017,
+      "grad_norm": 0.9360289573669434,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20140
+    },
+    {
+      "epoch": 3.257618624201762,
+      "grad_norm": 1.2099586725234985,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20150
+    },
+    {
+      "epoch": 3.2592353083825074,
+      "grad_norm": 0.8368481397628784,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 20160
+    },
+    {
+      "epoch": 3.2608519925632526,
+      "grad_norm": 0.7391039133071899,
+      "learning_rate": 0.0002,
+      "loss": 0.5626,
+      "step": 20170
+    },
+    {
+      "epoch": 3.262468676743998,
+      "grad_norm": 0.9122273325920105,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 20180
+    },
+    {
+      "epoch": 3.264085360924743,
+      "grad_norm": 0.8502281904220581,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 20190
+    },
+    {
+      "epoch": 3.265702045105489,
+      "grad_norm": 1.0926852226257324,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 20200
+    },
+    {
+      "epoch": 3.267318729286234,
+      "grad_norm": 0.7902828454971313,
+      "learning_rate": 0.0002,
+      "loss": 0.6027,
+      "step": 20210
+    },
+    {
+      "epoch": 3.2689354134669792,
+      "grad_norm": 0.8724729418754578,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20220
+    },
+    {
+      "epoch": 3.2705520976477245,
+      "grad_norm": 0.8469277024269104,
+      "learning_rate": 0.0002,
+      "loss": 0.6242,
+      "step": 20230
+    },
+    {
+      "epoch": 3.2721687818284697,
+      "grad_norm": 0.8865092992782593,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 20240
+    },
+    {
+      "epoch": 3.273785466009215,
+      "grad_norm": 1.0979334115982056,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20250
+    },
+    {
+      "epoch": 3.2754021501899606,
+      "grad_norm": 1.0860793590545654,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 20260
+    },
+    {
+      "epoch": 3.277018834370706,
+      "grad_norm": 0.981745183467865,
+      "learning_rate": 0.0002,
+      "loss": 0.6105,
+      "step": 20270
+    },
+    {
+      "epoch": 3.278635518551451,
+      "grad_norm": 0.9155020713806152,
+      "learning_rate": 0.0002,
+      "loss": 0.627,
+      "step": 20280
+    },
+    {
+      "epoch": 3.2802522027321963,
+      "grad_norm": 0.8436718583106995,
+      "learning_rate": 0.0002,
+      "loss": 0.5899,
+      "step": 20290
+    },
+    {
+      "epoch": 3.2818688869129415,
+      "grad_norm": 1.0329409837722778,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 20300
+    },
+    {
+      "epoch": 3.2834855710936868,
+      "grad_norm": 0.9876394271850586,
+      "learning_rate": 0.0002,
+      "loss": 0.6,
+      "step": 20310
+    },
+    {
+      "epoch": 3.285102255274432,
+      "grad_norm": 0.8052917718887329,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 20320
+    },
+    {
+      "epoch": 3.2867189394551772,
+      "grad_norm": 0.8390680551528931,
+      "learning_rate": 0.0002,
+      "loss": 0.5949,
+      "step": 20330
+    },
+    {
+      "epoch": 3.288335623635923,
+      "grad_norm": 0.9515735507011414,
+      "learning_rate": 0.0002,
+      "loss": 0.6492,
+      "step": 20340
+    },
+    {
+      "epoch": 3.289952307816668,
+      "grad_norm": 0.8028870224952698,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 20350
+    },
+    {
+      "epoch": 3.2915689919974134,
+      "grad_norm": 0.862592339515686,
+      "learning_rate": 0.0002,
+      "loss": 0.634,
+      "step": 20360
+    },
+    {
+      "epoch": 3.2931856761781586,
+      "grad_norm": 0.7451621890068054,
+      "learning_rate": 0.0002,
+      "loss": 0.6345,
+      "step": 20370
+    },
+    {
+      "epoch": 3.294802360358904,
+      "grad_norm": 0.8966776728630066,
+      "learning_rate": 0.0002,
+      "loss": 0.6458,
+      "step": 20380
+    },
+    {
+      "epoch": 3.296419044539649,
+      "grad_norm": 0.9289216995239258,
+      "learning_rate": 0.0002,
+      "loss": 0.5967,
+      "step": 20390
+    },
+    {
+      "epoch": 3.2980357287203943,
+      "grad_norm": 0.9649626612663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6599,
+      "step": 20400
+    },
+    {
+      "epoch": 3.29965241290114,
+      "grad_norm": 1.1953798532485962,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 20410
+    },
+    {
+      "epoch": 3.301269097081885,
+      "grad_norm": 0.8929083943367004,
+      "learning_rate": 0.0002,
+      "loss": 0.5997,
+      "step": 20420
+    },
+    {
+      "epoch": 3.3028857812626304,
+      "grad_norm": 0.8922014236450195,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 20430
+    },
+    {
+      "epoch": 3.3045024654433757,
+      "grad_norm": 0.9754860401153564,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 20440
+    },
+    {
+      "epoch": 3.306119149624121,
+      "grad_norm": 0.8873140215873718,
+      "learning_rate": 0.0002,
+      "loss": 0.5653,
+      "step": 20450
+    },
+    {
+      "epoch": 3.307735833804866,
+      "grad_norm": 0.857271671295166,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20460
+    },
+    {
+      "epoch": 3.3093525179856114,
+      "grad_norm": 0.9022141098976135,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 20470
+    },
+    {
+      "epoch": 3.3109692021663566,
+      "grad_norm": 0.8614798188209534,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 20480
+    },
+    {
+      "epoch": 3.3125858863471023,
+      "grad_norm": 0.8838164210319519,
+      "learning_rate": 0.0002,
+      "loss": 0.6254,
+      "step": 20490
+    },
+    {
+      "epoch": 3.3142025705278475,
+      "grad_norm": 0.8709736466407776,
+      "learning_rate": 0.0002,
+      "loss": 0.5849,
+      "step": 20500
+    },
+    {
+      "epoch": 3.3158192547085927,
+      "grad_norm": 0.9533300995826721,
+      "learning_rate": 0.0002,
+      "loss": 0.6146,
+      "step": 20510
+    },
+    {
+      "epoch": 3.317435938889338,
+      "grad_norm": 0.8259269595146179,
+      "learning_rate": 0.0002,
+      "loss": 0.6029,
+      "step": 20520
+    },
+    {
+      "epoch": 3.319052623070083,
+      "grad_norm": 0.8607608079910278,
+      "learning_rate": 0.0002,
+      "loss": 0.6268,
+      "step": 20530
+    },
+    {
+      "epoch": 3.3206693072508284,
+      "grad_norm": 1.0863020420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.5676,
+      "step": 20540
+    },
+    {
+      "epoch": 3.3222859914315737,
+      "grad_norm": 1.011489987373352,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 20550
+    },
+    {
+      "epoch": 3.3239026756123193,
+      "grad_norm": 0.6952177882194519,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 20560
+    },
+    {
+      "epoch": 3.3255193597930646,
+      "grad_norm": 0.9638974070549011,
+      "learning_rate": 0.0002,
+      "loss": 0.6229,
+      "step": 20570
+    },
+    {
+      "epoch": 3.32713604397381,
+      "grad_norm": 1.0310138463974,
+      "learning_rate": 0.0002,
+      "loss": 0.5882,
+      "step": 20580
+    },
+    {
+      "epoch": 3.328752728154555,
+      "grad_norm": 0.9371318221092224,
+      "learning_rate": 0.0002,
+      "loss": 0.594,
+      "step": 20590
+    },
+    {
+      "epoch": 3.3303694123353003,
+      "grad_norm": 0.8756691813468933,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 20600
+    },
+    {
+      "epoch": 3.3319860965160455,
+      "grad_norm": 1.054175853729248,
+      "learning_rate": 0.0002,
+      "loss": 0.5994,
+      "step": 20610
+    },
+    {
+      "epoch": 3.3336027806967907,
+      "grad_norm": 0.9074128270149231,
+      "learning_rate": 0.0002,
+      "loss": 0.6169,
+      "step": 20620
+    },
+    {
+      "epoch": 3.335219464877536,
+      "grad_norm": 0.906900942325592,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20630
+    },
+    {
+      "epoch": 3.3368361490582816,
+      "grad_norm": 0.8689333200454712,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 20640
+    },
+    {
+      "epoch": 3.338452833239027,
+      "grad_norm": 0.9889747500419617,
+      "learning_rate": 0.0002,
+      "loss": 0.6079,
+      "step": 20650
+    },
+    {
+      "epoch": 3.340069517419772,
+      "grad_norm": 1.0685805082321167,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20660
+    },
+    {
+      "epoch": 3.3416862016005173,
+      "grad_norm": 0.7495010495185852,
+      "learning_rate": 0.0002,
+      "loss": 0.6091,
+      "step": 20670
+    },
+    {
+      "epoch": 3.3433028857812626,
+      "grad_norm": 0.8747848272323608,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 20680
+    },
+    {
+      "epoch": 3.344919569962008,
+      "grad_norm": 0.9762673377990723,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 20690
+    },
+    {
+      "epoch": 3.346536254142753,
+      "grad_norm": 1.0284489393234253,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 20700
+    },
+    {
+      "epoch": 3.3481529383234987,
+      "grad_norm": 0.7293812036514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20710
+    },
+    {
+      "epoch": 3.349769622504244,
+      "grad_norm": 0.8330199122428894,
+      "learning_rate": 0.0002,
+      "loss": 0.609,
+      "step": 20720
+    },
+    {
+      "epoch": 3.351386306684989,
+      "grad_norm": 0.9808499217033386,
+      "learning_rate": 0.0002,
+      "loss": 0.5729,
+      "step": 20730
+    },
+    {
+      "epoch": 3.3530029908657344,
+      "grad_norm": 0.9508825540542603,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 20740
+    },
+    {
+      "epoch": 3.3546196750464796,
+      "grad_norm": 0.790483832359314,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 20750
+    },
+    {
+      "epoch": 3.356236359227225,
+      "grad_norm": 1.022793173789978,
+      "learning_rate": 0.0002,
+      "loss": 0.6327,
+      "step": 20760
+    },
+    {
+      "epoch": 3.35785304340797,
+      "grad_norm": 0.8318950533866882,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 20770
+    },
+    {
+      "epoch": 3.3594697275887153,
+      "grad_norm": 0.7980858087539673,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 20780
+    },
+    {
+      "epoch": 3.361086411769461,
+      "grad_norm": 0.8114802241325378,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 20790
+    },
+    {
+      "epoch": 3.3627030959502062,
+      "grad_norm": 0.8522519469261169,
+      "learning_rate": 0.0002,
+      "loss": 0.6017,
+      "step": 20800
+    },
+    {
+      "epoch": 3.3643197801309515,
+      "grad_norm": 0.9142431616783142,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 20810
+    },
+    {
+      "epoch": 3.3659364643116967,
+      "grad_norm": 0.771170437335968,
+      "learning_rate": 0.0002,
+      "loss": 0.6331,
+      "step": 20820
+    },
+    {
+      "epoch": 3.367553148492442,
+      "grad_norm": 1.0628231763839722,
+      "learning_rate": 0.0002,
+      "loss": 0.5879,
+      "step": 20830
+    },
+    {
+      "epoch": 3.369169832673187,
+      "grad_norm": 0.9384352564811707,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 20840
+    },
+    {
+      "epoch": 3.370786516853933,
+      "grad_norm": 1.1286591291427612,
+      "learning_rate": 0.0002,
+      "loss": 0.6292,
+      "step": 20850
+    },
+    {
+      "epoch": 3.372403201034678,
+      "grad_norm": 1.1349513530731201,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 20860
+    },
+    {
+      "epoch": 3.3740198852154233,
+      "grad_norm": 1.0127464532852173,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 20870
+    },
+    {
+      "epoch": 3.3756365693961685,
+      "grad_norm": 0.9111971855163574,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 20880
+    },
+    {
+      "epoch": 3.3772532535769137,
+      "grad_norm": 0.871356725692749,
+      "learning_rate": 0.0002,
+      "loss": 0.6101,
+      "step": 20890
+    },
+    {
+      "epoch": 3.378869937757659,
+      "grad_norm": 0.7774117588996887,
+      "learning_rate": 0.0002,
+      "loss": 0.5995,
+      "step": 20900
+    },
+    {
+      "epoch": 3.380486621938404,
+      "grad_norm": 1.0089964866638184,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 20910
+    },
+    {
+      "epoch": 3.3821033061191494,
+      "grad_norm": 0.7855867147445679,
+      "learning_rate": 0.0002,
+      "loss": 0.5908,
+      "step": 20920
+    },
+    {
+      "epoch": 3.3837199902998947,
+      "grad_norm": 1.3713710308074951,
+      "learning_rate": 0.0002,
+      "loss": 0.6373,
+      "step": 20930
+    },
+    {
+      "epoch": 3.3853366744806404,
+      "grad_norm": 0.8599116206169128,
+      "learning_rate": 0.0002,
+      "loss": 0.6627,
+      "step": 20940
+    },
+    {
+      "epoch": 3.3869533586613856,
+      "grad_norm": 0.9392673373222351,
+      "learning_rate": 0.0002,
+      "loss": 0.6224,
+      "step": 20950
+    },
+    {
+      "epoch": 3.388570042842131,
+      "grad_norm": 0.8764075040817261,
+      "learning_rate": 0.0002,
+      "loss": 0.5855,
+      "step": 20960
+    },
+    {
+      "epoch": 3.390186727022876,
+      "grad_norm": 0.8240136504173279,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 20970
+    },
+    {
+      "epoch": 3.3918034112036213,
+      "grad_norm": 1.0982369184494019,
+      "learning_rate": 0.0002,
+      "loss": 0.5783,
+      "step": 20980
+    },
+    {
+      "epoch": 3.3934200953843665,
+      "grad_norm": 1.0599013566970825,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 20990
+    },
+    {
+      "epoch": 3.395036779565112,
+      "grad_norm": 0.895438015460968,
+      "learning_rate": 0.0002,
+      "loss": 0.6356,
+      "step": 21000
+    },
+    {
+      "epoch": 3.3966534637458574,
+      "grad_norm": 0.6974841356277466,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 21010
+    },
+    {
+      "epoch": 3.3982701479266026,
+      "grad_norm": 0.9571719765663147,
+      "learning_rate": 0.0002,
+      "loss": 0.5704,
+      "step": 21020
+    },
+    {
+      "epoch": 3.399886832107348,
+      "grad_norm": 0.831912636756897,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 21030
+    },
+    {
+      "epoch": 3.401503516288093,
+      "grad_norm": 0.831936240196228,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 21040
+    },
+    {
+      "epoch": 3.4031202004688383,
+      "grad_norm": 0.7388373613357544,
+      "learning_rate": 0.0002,
+      "loss": 0.5857,
+      "step": 21050
+    },
+    {
+      "epoch": 3.4047368846495836,
+      "grad_norm": 0.938667356967926,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21060
+    },
+    {
+      "epoch": 3.406353568830329,
+      "grad_norm": 0.9202313423156738,
+      "learning_rate": 0.0002,
+      "loss": 0.6121,
+      "step": 21070
+    },
+    {
+      "epoch": 3.4079702530110745,
+      "grad_norm": 0.9888381958007812,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 21080
+    },
+    {
+      "epoch": 3.4095869371918197,
+      "grad_norm": 0.8526970744132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21090
+    },
+    {
+      "epoch": 3.411203621372565,
+      "grad_norm": 0.7939383387565613,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 21100
+    },
+    {
+      "epoch": 3.41282030555331,
+      "grad_norm": 0.9986352920532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 21110
+    },
+    {
+      "epoch": 3.4144369897340554,
+      "grad_norm": 0.8895300030708313,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 21120
+    },
+    {
+      "epoch": 3.4160536739148006,
+      "grad_norm": 0.9559482932090759,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 21130
+    },
+    {
+      "epoch": 3.417670358095546,
+      "grad_norm": 0.8351506590843201,
+      "learning_rate": 0.0002,
+      "loss": 0.6491,
+      "step": 21140
+    },
+    {
+      "epoch": 3.4192870422762915,
+      "grad_norm": 0.8224456906318665,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 21150
+    },
+    {
+      "epoch": 3.4209037264570368,
+      "grad_norm": 1.0110299587249756,
+      "learning_rate": 0.0002,
+      "loss": 0.5871,
+      "step": 21160
+    },
+    {
+      "epoch": 3.422520410637782,
+      "grad_norm": 0.82564777135849,
+      "learning_rate": 0.0002,
+      "loss": 0.6116,
+      "step": 21170
+    },
+    {
+      "epoch": 3.4241370948185272,
+      "grad_norm": 1.004738688468933,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 21180
+    },
+    {
+      "epoch": 3.4257537789992725,
+      "grad_norm": 0.7545676827430725,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 21190
+    },
+    {
+      "epoch": 3.4273704631800177,
+      "grad_norm": 0.8918704390525818,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 21200
+    },
+    {
+      "epoch": 3.428987147360763,
+      "grad_norm": 0.8336876034736633,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 21210
+    },
+    {
+      "epoch": 3.430603831541508,
+      "grad_norm": 0.8928771018981934,
+      "learning_rate": 0.0002,
+      "loss": 0.5824,
+      "step": 21220
+    },
+    {
+      "epoch": 3.432220515722254,
+      "grad_norm": 0.7663705945014954,
+      "learning_rate": 0.0002,
+      "loss": 0.6468,
+      "step": 21230
+    },
+    {
+      "epoch": 3.433837199902999,
+      "grad_norm": 0.8392598628997803,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 21240
+    },
+    {
+      "epoch": 3.4354538840837443,
+      "grad_norm": 0.8819600343704224,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 21250
+    },
+    {
+      "epoch": 3.4370705682644895,
+      "grad_norm": 0.9124642014503479,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 21260
+    },
+    {
+      "epoch": 3.4386872524452348,
+      "grad_norm": 0.8329763412475586,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 21270
+    },
+    {
+      "epoch": 3.44030393662598,
+      "grad_norm": 0.9982839822769165,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 21280
+    },
+    {
+      "epoch": 3.4419206208067252,
+      "grad_norm": 0.9105954766273499,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 21290
+    },
+    {
+      "epoch": 3.443537304987471,
+      "grad_norm": 0.8182359337806702,
+      "learning_rate": 0.0002,
+      "loss": 0.6028,
+      "step": 21300
+    },
+    {
+      "epoch": 3.445153989168216,
+      "grad_norm": 1.0568904876708984,
+      "learning_rate": 0.0002,
+      "loss": 0.5991,
+      "step": 21310
+    },
+    {
+      "epoch": 3.4467706733489614,
+      "grad_norm": 0.968539834022522,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 21320
+    },
+    {
+      "epoch": 3.4483873575297066,
+      "grad_norm": 0.8774511218070984,
+      "learning_rate": 0.0002,
+      "loss": 0.6219,
+      "step": 21330
+    },
+    {
+      "epoch": 3.450004041710452,
+      "grad_norm": 0.7598156332969666,
+      "learning_rate": 0.0002,
+      "loss": 0.6438,
+      "step": 21340
+    },
+    {
+      "epoch": 3.451620725891197,
+      "grad_norm": 1.1012897491455078,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 21350
+    },
+    {
+      "epoch": 3.4532374100719423,
+      "grad_norm": 0.8040637373924255,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 21360
+    },
+    {
+      "epoch": 3.4548540942526875,
+      "grad_norm": 0.8497496247291565,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 21370
+    },
+    {
+      "epoch": 3.456470778433433,
+      "grad_norm": 0.8429915904998779,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 21380
+    },
+    {
+      "epoch": 3.4580874626141784,
+      "grad_norm": 0.8107112646102905,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 21390
+    },
+    {
+      "epoch": 3.4597041467949237,
+      "grad_norm": 1.00872004032135,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 21400
+    },
+    {
+      "epoch": 3.461320830975669,
+      "grad_norm": 0.8266542553901672,
+      "learning_rate": 0.0002,
+      "loss": 0.5712,
+      "step": 21410
+    },
+    {
+      "epoch": 3.462937515156414,
+      "grad_norm": 0.8972568511962891,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 21420
+    },
+    {
+      "epoch": 3.4645541993371594,
+      "grad_norm": 1.0781476497650146,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 21430
+    },
+    {
+      "epoch": 3.4661708835179046,
+      "grad_norm": 0.9571592807769775,
+      "learning_rate": 0.0002,
+      "loss": 0.6303,
+      "step": 21440
+    },
+    {
+      "epoch": 3.4677875676986503,
+      "grad_norm": 0.881547212600708,
+      "learning_rate": 0.0002,
+      "loss": 0.6309,
+      "step": 21450
+    },
+    {
+      "epoch": 3.4694042518793955,
+      "grad_norm": 0.6955338716506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 21460
+    },
+    {
+      "epoch": 3.4710209360601407,
+      "grad_norm": 0.901187539100647,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 21470
+    },
+    {
+      "epoch": 3.472637620240886,
+      "grad_norm": 0.7063511610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.639,
+      "step": 21480
+    },
+    {
+      "epoch": 3.474254304421631,
+      "grad_norm": 0.8462792038917542,
+      "learning_rate": 0.0002,
+      "loss": 0.6154,
+      "step": 21490
+    },
+    {
+      "epoch": 3.4758709886023764,
+      "grad_norm": 1.1861060857772827,
+      "learning_rate": 0.0002,
+      "loss": 0.61,
+      "step": 21500
+    },
+    {
+      "epoch": 3.4774876727831217,
+      "grad_norm": 0.70503169298172,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 21510
+    },
+    {
+      "epoch": 3.479104356963867,
+      "grad_norm": 0.9650066494941711,
+      "learning_rate": 0.0002,
+      "loss": 0.6475,
+      "step": 21520
+    },
+    {
+      "epoch": 3.4807210411446126,
+      "grad_norm": 1.0266852378845215,
+      "learning_rate": 0.0002,
+      "loss": 0.6452,
+      "step": 21530
+    },
+    {
+      "epoch": 3.482337725325358,
+      "grad_norm": 0.956372857093811,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 21540
+    },
+    {
+      "epoch": 3.483954409506103,
+      "grad_norm": 0.8848432898521423,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 21550
+    },
+    {
+      "epoch": 3.4855710936868483,
+      "grad_norm": 1.0805351734161377,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 21560
+    },
+    {
+      "epoch": 3.4871877778675935,
+      "grad_norm": 0.9279725551605225,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 21570
+    },
+    {
+      "epoch": 3.4888044620483387,
+      "grad_norm": 0.9049562215805054,
+      "learning_rate": 0.0002,
+      "loss": 0.6354,
+      "step": 21580
+    },
+    {
+      "epoch": 3.4904211462290844,
+      "grad_norm": 0.9619429111480713,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 21590
+    },
+    {
+      "epoch": 3.4920378304098296,
+      "grad_norm": 0.8508906960487366,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 21600
+    },
+    {
+      "epoch": 3.493654514590575,
+      "grad_norm": 0.8692502379417419,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 21610
+    },
+    {
+      "epoch": 3.49527119877132,
+      "grad_norm": 0.8187332153320312,
+      "learning_rate": 0.0002,
+      "loss": 0.5878,
+      "step": 21620
+    },
+    {
+      "epoch": 3.4968878829520653,
+      "grad_norm": 1.145400047302246,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 21630
+    },
+    {
+      "epoch": 3.4985045671328105,
+      "grad_norm": 0.8281388282775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 21640
+    },
+    {
+      "epoch": 3.500121251313556,
+      "grad_norm": 0.82256019115448,
+      "learning_rate": 0.0002,
+      "loss": 0.6624,
+      "step": 21650
+    },
+    {
+      "epoch": 3.501737935494301,
+      "grad_norm": 0.9315484762191772,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 21660
+    },
+    {
+      "epoch": 3.5033546196750462,
+      "grad_norm": 0.7626111507415771,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 21670
+    },
+    {
+      "epoch": 3.504971303855792,
+      "grad_norm": 0.9275059103965759,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 21680
+    },
+    {
+      "epoch": 3.506587988036537,
+      "grad_norm": 0.7906724810600281,
+      "learning_rate": 0.0002,
+      "loss": 0.64,
+      "step": 21690
+    },
+    {
+      "epoch": 3.5082046722172824,
+      "grad_norm": 0.8289761543273926,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 21700
+    },
+    {
+      "epoch": 3.5098213563980276,
+      "grad_norm": 0.8316431045532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 21710
+    },
+    {
+      "epoch": 3.511438040578773,
+      "grad_norm": 1.0451812744140625,
+      "learning_rate": 0.0002,
+      "loss": 0.619,
+      "step": 21720
+    },
+    {
+      "epoch": 3.513054724759518,
+      "grad_norm": 0.928252637386322,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 21730
+    },
+    {
+      "epoch": 3.5146714089402638,
+      "grad_norm": 0.7985895276069641,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 21740
+    },
+    {
+      "epoch": 3.516288093121009,
+      "grad_norm": 0.6740974187850952,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 21750
+    },
+    {
+      "epoch": 3.517904777301754,
+      "grad_norm": 0.8482223749160767,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 21760
+    },
+    {
+      "epoch": 3.5195214614824994,
+      "grad_norm": 0.889947772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.6277,
+      "step": 21770
+    },
+    {
+      "epoch": 3.5211381456632447,
+      "grad_norm": 0.8304598927497864,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 21780
+    },
+    {
+      "epoch": 3.52275482984399,
+      "grad_norm": 0.8002981543540955,
+      "learning_rate": 0.0002,
+      "loss": 0.6156,
+      "step": 21790
+    },
+    {
+      "epoch": 3.524371514024735,
+      "grad_norm": 0.8115083575248718,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21800
+    },
+    {
+      "epoch": 3.5259881982054804,
+      "grad_norm": 0.9715048670768738,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 21810
+    },
+    {
+      "epoch": 3.5276048823862256,
+      "grad_norm": 1.0910786390304565,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 21820
+    },
+    {
+      "epoch": 3.5292215665669713,
+      "grad_norm": 0.8438942432403564,
+      "learning_rate": 0.0002,
+      "loss": 0.6543,
+      "step": 21830
+    },
+    {
+      "epoch": 3.5308382507477165,
+      "grad_norm": 0.8813382983207703,
+      "learning_rate": 0.0002,
+      "loss": 0.6509,
+      "step": 21840
+    },
+    {
+      "epoch": 3.5324549349284617,
+      "grad_norm": 0.7092908024787903,
+      "learning_rate": 0.0002,
+      "loss": 0.6049,
+      "step": 21850
+    },
+    {
+      "epoch": 3.534071619109207,
+      "grad_norm": 0.8332187533378601,
+      "learning_rate": 0.0002,
+      "loss": 0.5678,
+      "step": 21860
+    },
+    {
+      "epoch": 3.535688303289952,
+      "grad_norm": 0.8958209156990051,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21870
+    },
+    {
+      "epoch": 3.5373049874706974,
+      "grad_norm": 0.824138879776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 21880
+    },
+    {
+      "epoch": 3.538921671651443,
+      "grad_norm": 0.8375158309936523,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 21890
+    },
+    {
+      "epoch": 3.5405383558321883,
+      "grad_norm": 1.0274608135223389,
+      "learning_rate": 0.0002,
+      "loss": 0.6019,
+      "step": 21900
+    },
+    {
+      "epoch": 3.5421550400129336,
+      "grad_norm": 0.7088932394981384,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 21910
+    },
+    {
+      "epoch": 3.543771724193679,
+      "grad_norm": 0.8172445297241211,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 21920
+    },
+    {
+      "epoch": 3.545388408374424,
+      "grad_norm": 0.9904135465621948,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 21930
+    },
+    {
+      "epoch": 3.5470050925551693,
+      "grad_norm": 0.9900432229042053,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 21940
+    },
+    {
+      "epoch": 3.5486217767359145,
+      "grad_norm": 0.8963301181793213,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 21950
+    },
+    {
+      "epoch": 3.5502384609166597,
+      "grad_norm": 0.8551464676856995,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 21960
+    },
+    {
+      "epoch": 3.551855145097405,
+      "grad_norm": 1.0916603803634644,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 21970
+    },
+    {
+      "epoch": 3.5534718292781506,
+      "grad_norm": 0.841598391532898,
+      "learning_rate": 0.0002,
+      "loss": 0.6523,
+      "step": 21980
+    },
+    {
+      "epoch": 3.555088513458896,
+      "grad_norm": 0.8566757440567017,
+      "learning_rate": 0.0002,
+      "loss": 0.617,
+      "step": 21990
+    },
+    {
+      "epoch": 3.556705197639641,
+      "grad_norm": 1.0145052671432495,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 22000
+    },
+    {
+      "epoch": 3.5583218818203863,
+      "grad_norm": 0.9293754696846008,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 22010
+    },
+    {
+      "epoch": 3.5599385660011316,
+      "grad_norm": 0.9568536281585693,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 22020
+    },
+    {
+      "epoch": 3.5615552501818772,
+      "grad_norm": 0.8613139986991882,
+      "learning_rate": 0.0002,
+      "loss": 0.641,
+      "step": 22030
+    },
+    {
+      "epoch": 3.5631719343626225,
+      "grad_norm": 0.8179237246513367,
+      "learning_rate": 0.0002,
+      "loss": 0.6496,
+      "step": 22040
+    },
+    {
+      "epoch": 3.5647886185433677,
+      "grad_norm": 0.9059830904006958,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 22050
+    },
+    {
+      "epoch": 3.566405302724113,
+      "grad_norm": 1.0068252086639404,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 22060
+    },
+    {
+      "epoch": 3.568021986904858,
+      "grad_norm": 0.9682072997093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6239,
+      "step": 22070
+    },
+    {
+      "epoch": 3.5696386710856034,
+      "grad_norm": 0.8514005541801453,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 22080
+    },
+    {
+      "epoch": 3.5712553552663486,
+      "grad_norm": 0.8327770829200745,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 22090
+    },
+    {
+      "epoch": 3.572872039447094,
+      "grad_norm": 1.024976372718811,
+      "learning_rate": 0.0002,
+      "loss": 0.5976,
+      "step": 22100
+    },
+    {
+      "epoch": 3.574488723627839,
+      "grad_norm": 0.7721174955368042,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 22110
+    },
+    {
+      "epoch": 3.5761054078085843,
+      "grad_norm": 1.0351054668426514,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 22120
+    },
+    {
+      "epoch": 3.57772209198933,
+      "grad_norm": 0.9680907130241394,
+      "learning_rate": 0.0002,
+      "loss": 0.6379,
+      "step": 22130
+    },
+    {
+      "epoch": 3.5793387761700752,
+      "grad_norm": 0.8016974925994873,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 22140
+    },
+    {
+      "epoch": 3.5809554603508205,
+      "grad_norm": 1.0109003782272339,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 22150
+    },
+    {
+      "epoch": 3.5825721445315657,
+      "grad_norm": 1.0473392009735107,
+      "learning_rate": 0.0002,
+      "loss": 0.6368,
+      "step": 22160
+    },
+    {
+      "epoch": 3.584188828712311,
+      "grad_norm": 0.8686613440513611,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 22170
+    },
+    {
+      "epoch": 3.5858055128930566,
+      "grad_norm": 0.869149923324585,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 22180
+    },
+    {
+      "epoch": 3.587422197073802,
+      "grad_norm": 0.9769062995910645,
+      "learning_rate": 0.0002,
+      "loss": 0.5895,
+      "step": 22190
+    },
+    {
+      "epoch": 3.589038881254547,
+      "grad_norm": 0.779636561870575,
+      "learning_rate": 0.0002,
+      "loss": 0.5939,
+      "step": 22200
+    },
+    {
+      "epoch": 3.5906555654352923,
+      "grad_norm": 0.9063841104507446,
+      "learning_rate": 0.0002,
+      "loss": 0.5875,
+      "step": 22210
+    },
+    {
+      "epoch": 3.5922722496160375,
+      "grad_norm": 0.9216037392616272,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 22220
+    },
+    {
+      "epoch": 3.5938889337967828,
+      "grad_norm": 1.0217336416244507,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 22230
+    },
+    {
+      "epoch": 3.595505617977528,
+      "grad_norm": 0.8513161540031433,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 22240
+    },
+    {
+      "epoch": 3.597122302158273,
+      "grad_norm": 0.8084813952445984,
+      "learning_rate": 0.0002,
+      "loss": 0.6301,
+      "step": 22250
+    },
+    {
+      "epoch": 3.5987389863390185,
+      "grad_norm": 0.8524802923202515,
+      "learning_rate": 0.0002,
+      "loss": 0.6197,
+      "step": 22260
+    },
+    {
+      "epoch": 3.600355670519764,
+      "grad_norm": 0.9356237649917603,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 22270
+    },
+    {
+      "epoch": 3.6019723547005094,
+      "grad_norm": 1.009600281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.628,
+      "step": 22280
+    },
+    {
+      "epoch": 3.6035890388812546,
+      "grad_norm": 0.9900581240653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 22290
+    },
+    {
+      "epoch": 3.605205723062,
+      "grad_norm": 1.062495231628418,
+      "learning_rate": 0.0002,
+      "loss": 0.5725,
+      "step": 22300
+    },
+    {
+      "epoch": 3.606822407242745,
+      "grad_norm": 0.8832381367683411,
+      "learning_rate": 0.0002,
+      "loss": 0.607,
+      "step": 22310
+    },
+    {
+      "epoch": 3.6084390914234903,
+      "grad_norm": 0.9284297823905945,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 22320
+    },
+    {
+      "epoch": 3.610055775604236,
+      "grad_norm": 1.2381829023361206,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 22330
+    },
+    {
+      "epoch": 3.611672459784981,
+      "grad_norm": 0.929434597492218,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 22340
+    },
+    {
+      "epoch": 3.6132891439657264,
+      "grad_norm": 0.9714490175247192,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 22350
+    },
+    {
+      "epoch": 3.6149058281464717,
+      "grad_norm": 0.808014988899231,
+      "learning_rate": 0.0002,
+      "loss": 0.6861,
+      "step": 22360
+    },
+    {
+      "epoch": 3.616522512327217,
+      "grad_norm": 1.0364398956298828,
+      "learning_rate": 0.0002,
+      "loss": 0.6428,
+      "step": 22370
+    },
+    {
+      "epoch": 3.618139196507962,
+      "grad_norm": 0.7858489751815796,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22380
+    },
+    {
+      "epoch": 3.6197558806887074,
+      "grad_norm": 0.9920870065689087,
+      "learning_rate": 0.0002,
+      "loss": 0.6214,
+      "step": 22390
+    },
+    {
+      "epoch": 3.6213725648694526,
+      "grad_norm": 0.9183220863342285,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 22400
+    },
+    {
+      "epoch": 3.622989249050198,
+      "grad_norm": 0.9826246500015259,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22410
+    },
+    {
+      "epoch": 3.6246059332309435,
+      "grad_norm": 0.8632931113243103,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 22420
+    },
+    {
+      "epoch": 3.6262226174116887,
+      "grad_norm": 0.8468965291976929,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 22430
+    },
+    {
+      "epoch": 3.627839301592434,
+      "grad_norm": 0.8466871976852417,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 22440
+    },
+    {
+      "epoch": 3.629455985773179,
+      "grad_norm": 0.9501169919967651,
+      "learning_rate": 0.0002,
+      "loss": 0.5941,
+      "step": 22450
+    },
+    {
+      "epoch": 3.6310726699539244,
+      "grad_norm": 0.8906720876693726,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 22460
+    },
+    {
+      "epoch": 3.6326893541346696,
+      "grad_norm": 0.7400227189064026,
+      "learning_rate": 0.0002,
+      "loss": 0.6928,
+      "step": 22470
+    },
+    {
+      "epoch": 3.6343060383154153,
+      "grad_norm": 0.9756355881690979,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22480
+    },
+    {
+      "epoch": 3.6359227224961606,
+      "grad_norm": 0.7504993081092834,
+      "learning_rate": 0.0002,
+      "loss": 0.6203,
+      "step": 22490
+    },
+    {
+      "epoch": 3.637539406676906,
+      "grad_norm": 0.9270039200782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 22500
+    },
+    {
+      "epoch": 3.639156090857651,
+      "grad_norm": 0.8841686844825745,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 22510
+    },
+    {
+      "epoch": 3.6407727750383962,
+      "grad_norm": 0.8533213138580322,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 22520
+    },
+    {
+      "epoch": 3.6423894592191415,
+      "grad_norm": 1.0052043199539185,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 22530
+    },
+    {
+      "epoch": 3.6440061433998867,
+      "grad_norm": 1.0323461294174194,
+      "learning_rate": 0.0002,
+      "loss": 0.6363,
+      "step": 22540
+    },
+    {
+      "epoch": 3.645622827580632,
+      "grad_norm": 0.8654312491416931,
+      "learning_rate": 0.0002,
+      "loss": 0.6545,
+      "step": 22550
+    },
+    {
+      "epoch": 3.647239511761377,
+      "grad_norm": 0.6400038003921509,
+      "learning_rate": 0.0002,
+      "loss": 0.6155,
+      "step": 22560
+    },
+    {
+      "epoch": 3.648856195942123,
+      "grad_norm": 0.8061298727989197,
+      "learning_rate": 0.0002,
+      "loss": 0.5829,
+      "step": 22570
+    },
+    {
+      "epoch": 3.650472880122868,
+      "grad_norm": 0.9257854223251343,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 22580
+    },
+    {
+      "epoch": 3.6520895643036133,
+      "grad_norm": 0.8439396619796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6409,
+      "step": 22590
+    },
+    {
+      "epoch": 3.6537062484843585,
+      "grad_norm": 0.7764544486999512,
+      "learning_rate": 0.0002,
+      "loss": 0.5996,
+      "step": 22600
+    },
+    {
+      "epoch": 3.6553229326651038,
+      "grad_norm": 1.125451683998108,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 22610
+    },
+    {
+      "epoch": 3.656939616845849,
+      "grad_norm": 0.7523018717765808,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 22620
+    },
+    {
+      "epoch": 3.6585563010265947,
+      "grad_norm": 1.071026086807251,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 22630
+    },
+    {
+      "epoch": 3.66017298520734,
+      "grad_norm": 0.945791482925415,
+      "learning_rate": 0.0002,
+      "loss": 0.6459,
+      "step": 22640
+    },
+    {
+      "epoch": 3.661789669388085,
+      "grad_norm": 0.8001811504364014,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 22650
+    },
+    {
+      "epoch": 3.6634063535688304,
+      "grad_norm": 0.9700816869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.6385,
+      "step": 22660
+    },
+    {
+      "epoch": 3.6650230377495756,
+      "grad_norm": 0.9053242206573486,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22670
+    },
+    {
+      "epoch": 3.666639721930321,
+      "grad_norm": 0.944362461566925,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 22680
+    },
+    {
+      "epoch": 3.668256406111066,
+      "grad_norm": 1.067489504814148,
+      "learning_rate": 0.0002,
+      "loss": 0.6235,
+      "step": 22690
+    },
+    {
+      "epoch": 3.6698730902918113,
+      "grad_norm": 1.0984995365142822,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 22700
+    },
+    {
+      "epoch": 3.6714897744725565,
+      "grad_norm": 0.9336317777633667,
+      "learning_rate": 0.0002,
+      "loss": 0.6717,
+      "step": 22710
+    },
+    {
+      "epoch": 3.673106458653302,
+      "grad_norm": 0.9261918663978577,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 22720
+    },
+    {
+      "epoch": 3.6747231428340474,
+      "grad_norm": 0.8648008704185486,
+      "learning_rate": 0.0002,
+      "loss": 0.6332,
+      "step": 22730
+    },
+    {
+      "epoch": 3.6763398270147927,
+      "grad_norm": 0.7225083708763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6576,
+      "step": 22740
+    },
+    {
+      "epoch": 3.677956511195538,
+      "grad_norm": 0.9258282780647278,
+      "learning_rate": 0.0002,
+      "loss": 0.6406,
+      "step": 22750
+    },
+    {
+      "epoch": 3.679573195376283,
+      "grad_norm": 0.70876145362854,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 22760
+    },
+    {
+      "epoch": 3.681189879557029,
+      "grad_norm": 0.8780210018157959,
+      "learning_rate": 0.0002,
+      "loss": 0.6821,
+      "step": 22770
+    },
+    {
+      "epoch": 3.682806563737774,
+      "grad_norm": 0.8075440526008606,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22780
+    },
+    {
+      "epoch": 3.6844232479185193,
+      "grad_norm": 0.8503130674362183,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22790
+    },
+    {
+      "epoch": 3.6860399320992645,
+      "grad_norm": 0.8413618206977844,
+      "learning_rate": 0.0002,
+      "loss": 0.6082,
+      "step": 22800
+    },
+    {
+      "epoch": 3.6876566162800097,
+      "grad_norm": 0.8675165176391602,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 22810
+    },
+    {
+      "epoch": 3.689273300460755,
+      "grad_norm": 0.8235884308815002,
+      "learning_rate": 0.0002,
+      "loss": 0.6157,
+      "step": 22820
+    },
+    {
+      "epoch": 3.6908899846415,
+      "grad_norm": 0.9477725625038147,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 22830
+    },
+    {
+      "epoch": 3.6925066688222454,
+      "grad_norm": 0.7883533835411072,
+      "learning_rate": 0.0002,
+      "loss": 0.6481,
+      "step": 22840
+    },
+    {
+      "epoch": 3.6941233530029907,
+      "grad_norm": 1.047913908958435,
+      "learning_rate": 0.0002,
+      "loss": 0.5872,
+      "step": 22850
+    },
+    {
+      "epoch": 3.695740037183736,
+      "grad_norm": 0.9171528816223145,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 22860
+    },
+    {
+      "epoch": 3.6973567213644816,
+      "grad_norm": 0.9338192343711853,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 22870
+    },
+    {
+      "epoch": 3.698973405545227,
+      "grad_norm": 0.8799443244934082,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 22880
+    },
+    {
+      "epoch": 3.700590089725972,
+      "grad_norm": 0.8515434861183167,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 22890
+    },
+    {
+      "epoch": 3.7022067739067173,
+      "grad_norm": 0.7805591821670532,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 22900
+    },
+    {
+      "epoch": 3.7038234580874625,
+      "grad_norm": 0.8470911979675293,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 22910
+    },
+    {
+      "epoch": 3.705440142268208,
+      "grad_norm": 0.9452309012413025,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 22920
+    },
+    {
+      "epoch": 3.7070568264489534,
+      "grad_norm": 0.950243353843689,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 22930
+    },
+    {
+      "epoch": 3.7086735106296986,
+      "grad_norm": 0.7882499098777771,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 22940
+    },
+    {
+      "epoch": 3.710290194810444,
+      "grad_norm": 0.8307787775993347,
+      "learning_rate": 0.0002,
+      "loss": 0.6462,
+      "step": 22950
+    },
+    {
+      "epoch": 3.711906878991189,
+      "grad_norm": 1.0970630645751953,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 22960
+    },
+    {
+      "epoch": 3.7135235631719343,
+      "grad_norm": 0.8269566297531128,
+      "learning_rate": 0.0002,
+      "loss": 0.6281,
+      "step": 22970
+    },
+    {
+      "epoch": 3.7151402473526796,
+      "grad_norm": 0.8306704759597778,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22980
+    },
+    {
+      "epoch": 3.716756931533425,
+      "grad_norm": 0.9710225462913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6418,
+      "step": 22990
+    },
+    {
+      "epoch": 3.71837361571417,
+      "grad_norm": 0.8890530467033386,
+      "learning_rate": 0.0002,
+      "loss": 0.6639,
+      "step": 23000
+    },
+    {
+      "epoch": 3.7199902998949153,
+      "grad_norm": 0.883522629737854,
+      "learning_rate": 0.0002,
+      "loss": 0.6084,
+      "step": 23010
+    },
+    {
+      "epoch": 3.721606984075661,
+      "grad_norm": 0.8662652373313904,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 23020
+    },
+    {
+      "epoch": 3.723223668256406,
+      "grad_norm": 0.7228406667709351,
+      "learning_rate": 0.0002,
+      "loss": 0.6266,
+      "step": 23030
+    },
+    {
+      "epoch": 3.7248403524371514,
+      "grad_norm": 1.060792088508606,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23040
+    },
+    {
+      "epoch": 3.7264570366178966,
+      "grad_norm": 1.0119613409042358,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 23050
+    },
+    {
+      "epoch": 3.728073720798642,
+      "grad_norm": 0.9212996959686279,
+      "learning_rate": 0.0002,
+      "loss": 0.6466,
+      "step": 23060
+    },
+    {
+      "epoch": 3.7296904049793875,
+      "grad_norm": 0.925690233707428,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 23070
+    },
+    {
+      "epoch": 3.7313070891601328,
+      "grad_norm": 0.8323310613632202,
+      "learning_rate": 0.0002,
+      "loss": 0.615,
+      "step": 23080
+    },
+    {
+      "epoch": 3.732923773340878,
+      "grad_norm": 0.8966048955917358,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 23090
+    },
+    {
+      "epoch": 3.7345404575216232,
+      "grad_norm": 0.8995837569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23100
+    },
+    {
+      "epoch": 3.7361571417023685,
+      "grad_norm": 0.8748890161514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23110
+    },
+    {
+      "epoch": 3.7377738258831137,
+      "grad_norm": 0.7985540628433228,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 23120
+    },
+    {
+      "epoch": 3.739390510063859,
+      "grad_norm": 1.0240917205810547,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 23130
+    },
+    {
+      "epoch": 3.741007194244604,
+      "grad_norm": 0.9181789755821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 23140
+    },
+    {
+      "epoch": 3.7426238784253494,
+      "grad_norm": 0.8896583914756775,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 23150
+    },
+    {
+      "epoch": 3.744240562606095,
+      "grad_norm": 0.8635515570640564,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 23160
+    },
+    {
+      "epoch": 3.7458572467868403,
+      "grad_norm": 0.8873575329780579,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 23170
+    },
+    {
+      "epoch": 3.7474739309675855,
+      "grad_norm": 0.9807148575782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23180
+    },
+    {
+      "epoch": 3.7490906151483308,
+      "grad_norm": 0.900477945804596,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 23190
+    },
+    {
+      "epoch": 3.750707299329076,
+      "grad_norm": 0.9379992485046387,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23200
+    },
+    {
+      "epoch": 3.752323983509821,
+      "grad_norm": 0.9649890661239624,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 23210
+    },
+    {
+      "epoch": 3.753940667690567,
+      "grad_norm": 0.824442446231842,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 23220
+    },
+    {
+      "epoch": 3.755557351871312,
+      "grad_norm": 0.8896150588989258,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 23230
+    },
+    {
+      "epoch": 3.7571740360520574,
+      "grad_norm": 0.751249372959137,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 23240
+    },
+    {
+      "epoch": 3.7587907202328026,
+      "grad_norm": 0.9392193555831909,
+      "learning_rate": 0.0002,
+      "loss": 0.629,
+      "step": 23250
+    },
+    {
+      "epoch": 3.760407404413548,
+      "grad_norm": 0.9284586310386658,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 23260
+    },
+    {
+      "epoch": 3.762024088594293,
+      "grad_norm": 0.7738175392150879,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23270
+    },
+    {
+      "epoch": 3.7636407727750383,
+      "grad_norm": 0.9252978563308716,
+      "learning_rate": 0.0002,
+      "loss": 0.6743,
+      "step": 23280
+    },
+    {
+      "epoch": 3.7652574569557835,
+      "grad_norm": 0.9501895904541016,
+      "learning_rate": 0.0002,
+      "loss": 0.5984,
+      "step": 23290
+    },
+    {
+      "epoch": 3.7668741411365287,
+      "grad_norm": 0.9416276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 23300
+    },
+    {
+      "epoch": 3.7684908253172744,
+      "grad_norm": 0.7076631784439087,
+      "learning_rate": 0.0002,
+      "loss": 0.6507,
+      "step": 23310
+    },
+    {
+      "epoch": 3.7701075094980196,
+      "grad_norm": 0.9864492416381836,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 23320
+    },
+    {
+      "epoch": 3.771724193678765,
+      "grad_norm": 0.8450456261634827,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 23330
+    },
+    {
+      "epoch": 3.77334087785951,
+      "grad_norm": 1.0768941640853882,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23340
+    },
+    {
+      "epoch": 3.7749575620402553,
+      "grad_norm": 0.9956819415092468,
+      "learning_rate": 0.0002,
+      "loss": 0.6408,
+      "step": 23350
+    },
+    {
+      "epoch": 3.7765742462210006,
+      "grad_norm": 0.9234658479690552,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 23360
+    },
+    {
+      "epoch": 3.7781909304017463,
+      "grad_norm": 1.0993858575820923,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23370
+    },
+    {
+      "epoch": 3.7798076145824915,
+      "grad_norm": 0.923159658908844,
+      "learning_rate": 0.0002,
+      "loss": 0.6391,
+      "step": 23380
+    },
+    {
+      "epoch": 3.7814242987632367,
+      "grad_norm": 0.9311541318893433,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23390
+    },
+    {
+      "epoch": 3.783040982943982,
+      "grad_norm": 0.919681191444397,
+      "learning_rate": 0.0002,
+      "loss": 0.6535,
+      "step": 23400
+    },
+    {
+      "epoch": 3.784657667124727,
+      "grad_norm": 1.7406195402145386,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 23410
+    },
+    {
+      "epoch": 3.7862743513054724,
+      "grad_norm": 0.7789074182510376,
+      "learning_rate": 0.0002,
+      "loss": 0.657,
+      "step": 23420
+    },
+    {
+      "epoch": 3.7878910354862176,
+      "grad_norm": 0.8302814960479736,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23430
+    },
+    {
+      "epoch": 3.789507719666963,
+      "grad_norm": 0.8089349269866943,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23440
+    },
+    {
+      "epoch": 3.791124403847708,
+      "grad_norm": 0.9006284475326538,
+      "learning_rate": 0.0002,
+      "loss": 0.6682,
+      "step": 23450
+    },
+    {
+      "epoch": 3.7927410880284538,
+      "grad_norm": 0.8426766991615295,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 23460
+    },
+    {
+      "epoch": 3.794357772209199,
+      "grad_norm": 1.2576252222061157,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 23470
+    },
+    {
+      "epoch": 3.7959744563899442,
+      "grad_norm": 1.0307610034942627,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 23480
+    },
+    {
+      "epoch": 3.7975911405706895,
+      "grad_norm": 0.8525972962379456,
+      "learning_rate": 0.0002,
+      "loss": 0.6262,
+      "step": 23490
+    },
+    {
+      "epoch": 3.7992078247514347,
+      "grad_norm": 1.159039855003357,
+      "learning_rate": 0.0002,
+      "loss": 0.6757,
+      "step": 23500
+    },
+    {
+      "epoch": 3.80082450893218,
+      "grad_norm": 1.4193549156188965,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23510
+    },
+    {
+      "epoch": 3.8024411931129256,
+      "grad_norm": 0.8245543837547302,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 23520
+    },
+    {
+      "epoch": 3.804057877293671,
+      "grad_norm": 0.8847230076789856,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23530
+    },
+    {
+      "epoch": 3.805674561474416,
+      "grad_norm": 0.9574624300003052,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 23540
+    },
+    {
+      "epoch": 3.8072912456551613,
+      "grad_norm": 1.048020601272583,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 23550
+    },
+    {
+      "epoch": 3.8089079298359065,
+      "grad_norm": 0.8302255868911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6497,
+      "step": 23560
+    },
+    {
+      "epoch": 3.8105246140166518,
+      "grad_norm": 0.8269215822219849,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 23570
+    },
+    {
+      "epoch": 3.812141298197397,
+      "grad_norm": 0.9375753402709961,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 23580
+    },
+    {
+      "epoch": 3.8137579823781422,
+      "grad_norm": 1.0234097242355347,
+      "learning_rate": 0.0002,
+      "loss": 0.6132,
+      "step": 23590
+    },
+    {
+      "epoch": 3.8153746665588875,
+      "grad_norm": 0.8978445529937744,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23600
+    },
+    {
+      "epoch": 3.816991350739633,
+      "grad_norm": 0.7929515838623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 23610
+    },
+    {
+      "epoch": 3.8186080349203784,
+      "grad_norm": 1.3255881071090698,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23620
+    },
+    {
+      "epoch": 3.8202247191011236,
+      "grad_norm": 0.9188598990440369,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 23630
+    },
+    {
+      "epoch": 3.821841403281869,
+      "grad_norm": 0.8811675906181335,
+      "learning_rate": 0.0002,
+      "loss": 0.6152,
+      "step": 23640
+    },
+    {
+      "epoch": 3.823458087462614,
+      "grad_norm": 0.8061038255691528,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 23650
+    },
+    {
+      "epoch": 3.8250747716433597,
+      "grad_norm": 0.9975376129150391,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 23660
+    },
+    {
+      "epoch": 3.826691455824105,
+      "grad_norm": 0.8036105036735535,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 23670
+    },
+    {
+      "epoch": 3.82830814000485,
+      "grad_norm": 0.7401984333992004,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 23680
+    },
+    {
+      "epoch": 3.8299248241855954,
+      "grad_norm": 0.829753041267395,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 23690
+    },
+    {
+      "epoch": 3.8315415083663407,
+      "grad_norm": 0.8753240704536438,
+      "learning_rate": 0.0002,
+      "loss": 0.6611,
+      "step": 23700
+    },
+    {
+      "epoch": 3.833158192547086,
+      "grad_norm": 0.8157842755317688,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 23710
+    },
+    {
+      "epoch": 3.834774876727831,
+      "grad_norm": 0.6183798909187317,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 23720
+    },
+    {
+      "epoch": 3.8363915609085764,
+      "grad_norm": 0.9548442363739014,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 23730
+    },
+    {
+      "epoch": 3.8380082450893216,
+      "grad_norm": 0.8319669961929321,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23740
+    },
+    {
+      "epoch": 3.839624929270067,
+      "grad_norm": 0.9718693494796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 23750
+    },
+    {
+      "epoch": 3.8412416134508125,
+      "grad_norm": 0.8672235012054443,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 23760
+    },
+    {
+      "epoch": 3.8428582976315577,
+      "grad_norm": 1.1210707426071167,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 23770
+    },
+    {
+      "epoch": 3.844474981812303,
+      "grad_norm": 0.9177767634391785,
+      "learning_rate": 0.0002,
+      "loss": 0.5923,
+      "step": 23780
+    },
+    {
+      "epoch": 3.846091665993048,
+      "grad_norm": 0.8714171648025513,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 23790
+    },
+    {
+      "epoch": 3.8477083501737934,
+      "grad_norm": 1.1853246688842773,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 23800
+    },
+    {
+      "epoch": 3.849325034354539,
+      "grad_norm": 0.8091260194778442,
+      "learning_rate": 0.0002,
+      "loss": 0.6144,
+      "step": 23810
+    },
+    {
+      "epoch": 3.8509417185352843,
+      "grad_norm": 0.9710774421691895,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23820
+    },
+    {
+      "epoch": 3.8525584027160296,
+      "grad_norm": 0.7648707628250122,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23830
+    },
+    {
+      "epoch": 3.854175086896775,
+      "grad_norm": 0.7809253931045532,
+      "learning_rate": 0.0002,
+      "loss": 0.6013,
+      "step": 23840
+    },
+    {
+      "epoch": 3.85579177107752,
+      "grad_norm": 0.8337951898574829,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 23850
+    },
+    {
+      "epoch": 3.8574084552582653,
+      "grad_norm": 0.9271913170814514,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23860
+    },
+    {
+      "epoch": 3.8590251394390105,
+      "grad_norm": 0.985334038734436,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 23870
+    },
+    {
+      "epoch": 3.8606418236197557,
+      "grad_norm": 0.8458583354949951,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 23880
+    },
+    {
+      "epoch": 3.862258507800501,
+      "grad_norm": 1.015348196029663,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 23890
+    },
+    {
+      "epoch": 3.8638751919812466,
+      "grad_norm": 1.0121688842773438,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23900
+    },
+    {
+      "epoch": 3.865491876161992,
+      "grad_norm": 0.8883971571922302,
+      "learning_rate": 0.0002,
+      "loss": 0.5921,
+      "step": 23910
+    },
+    {
+      "epoch": 3.867108560342737,
+      "grad_norm": 1.028086543083191,
+      "learning_rate": 0.0002,
+      "loss": 0.6597,
+      "step": 23920
+    },
+    {
+      "epoch": 3.8687252445234823,
+      "grad_norm": 0.9645734429359436,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 23930
+    },
+    {
+      "epoch": 3.8703419287042276,
+      "grad_norm": 0.8235350251197815,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 23940
+    },
+    {
+      "epoch": 3.871958612884973,
+      "grad_norm": 1.0298916101455688,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23950
+    },
+    {
+      "epoch": 3.8735752970657185,
+      "grad_norm": 1.0063377618789673,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 23960
+    },
+    {
+      "epoch": 3.8751919812464637,
+      "grad_norm": 0.9230626821517944,
+      "learning_rate": 0.0002,
+      "loss": 0.6234,
+      "step": 23970
+    },
+    {
+      "epoch": 3.876808665427209,
+      "grad_norm": 0.9243063926696777,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 23980
+    },
+    {
+      "epoch": 3.878425349607954,
+      "grad_norm": 1.0211291313171387,
+      "learning_rate": 0.0002,
+      "loss": 0.6035,
+      "step": 23990
+    },
+    {
+      "epoch": 3.8800420337886994,
+      "grad_norm": 0.7800535559654236,
+      "learning_rate": 0.0002,
+      "loss": 0.6351,
+      "step": 24000
+    },
+    {
+      "epoch": 3.8816587179694446,
+      "grad_norm": 0.7904248833656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 24010
+    },
+    {
+      "epoch": 3.88327540215019,
+      "grad_norm": 1.1975988149642944,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 24020
+    },
+    {
+      "epoch": 3.884892086330935,
+      "grad_norm": 1.0626593828201294,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 24030
+    },
+    {
+      "epoch": 3.8865087705116803,
+      "grad_norm": 0.9012193083763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 24040
+    },
+    {
+      "epoch": 3.888125454692426,
+      "grad_norm": 1.1159172058105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6786,
+      "step": 24050
+    },
+    {
+      "epoch": 3.889742138873171,
+      "grad_norm": 1.276838779449463,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 24060
+    },
+    {
+      "epoch": 3.8913588230539164,
+      "grad_norm": 0.8467690348625183,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 24070
+    },
+    {
+      "epoch": 3.8929755072346617,
+      "grad_norm": 0.9862841963768005,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 24080
+    },
+    {
+      "epoch": 3.894592191415407,
+      "grad_norm": 0.7134621739387512,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 24090
+    },
+    {
+      "epoch": 3.896208875596152,
+      "grad_norm": 0.8178175091743469,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 24100
+    },
+    {
+      "epoch": 3.897825559776898,
+      "grad_norm": 0.9229172468185425,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 24110
+    },
+    {
+      "epoch": 3.899442243957643,
+      "grad_norm": 1.0878316164016724,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 24120
+    },
+    {
+      "epoch": 3.9010589281383883,
+      "grad_norm": 0.971645712852478,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24130
+    },
+    {
+      "epoch": 3.9026756123191335,
+      "grad_norm": 0.8862188458442688,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 24140
+    },
+    {
+      "epoch": 3.9042922964998787,
+      "grad_norm": 0.9126982688903809,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 24150
+    },
+    {
+      "epoch": 3.905908980680624,
+      "grad_norm": 0.8833470940589905,
+      "learning_rate": 0.0002,
+      "loss": 0.6734,
+      "step": 24160
+    },
+    {
+      "epoch": 3.907525664861369,
+      "grad_norm": 0.8320947885513306,
+      "learning_rate": 0.0002,
+      "loss": 0.5832,
+      "step": 24170
+    },
+    {
+      "epoch": 3.9091423490421144,
+      "grad_norm": 0.9156602025032043,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 24180
+    },
+    {
+      "epoch": 3.9107590332228597,
+      "grad_norm": 1.029181957244873,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 24190
+    },
+    {
+      "epoch": 3.9123757174036053,
+      "grad_norm": 0.9052802324295044,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 24200
+    },
+    {
+      "epoch": 3.9139924015843506,
+      "grad_norm": 0.8847255110740662,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 24210
+    },
+    {
+      "epoch": 3.915609085765096,
+      "grad_norm": 0.9642062187194824,
+      "learning_rate": 0.0002,
+      "loss": 0.6343,
+      "step": 24220
+    },
+    {
+      "epoch": 3.917225769945841,
+      "grad_norm": 0.8629093766212463,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 24230
+    },
+    {
+      "epoch": 3.9188424541265863,
+      "grad_norm": 0.8674976825714111,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 24240
+    },
+    {
+      "epoch": 3.9204591383073315,
+      "grad_norm": 1.104846477508545,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 24250
+    },
+    {
+      "epoch": 3.922075822488077,
+      "grad_norm": 1.0874955654144287,
+      "learning_rate": 0.0002,
+      "loss": 0.6501,
+      "step": 24260
+    },
+    {
+      "epoch": 3.9236925066688224,
+      "grad_norm": 0.8689812421798706,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 24270
+    },
+    {
+      "epoch": 3.9253091908495676,
+      "grad_norm": 0.9724617004394531,
+      "learning_rate": 0.0002,
+      "loss": 0.5893,
+      "step": 24280
+    },
+    {
+      "epoch": 3.926925875030313,
+      "grad_norm": 0.9165538549423218,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24290
+    },
+    {
+      "epoch": 3.928542559211058,
+      "grad_norm": 0.9307710528373718,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 24300
+    },
+    {
+      "epoch": 3.9301592433918033,
+      "grad_norm": 0.8589295148849487,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 24310
+    },
+    {
+      "epoch": 3.9317759275725486,
+      "grad_norm": 0.9151099920272827,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 24320
+    },
+    {
+      "epoch": 3.933392611753294,
+      "grad_norm": 0.9633517265319824,
+      "learning_rate": 0.0002,
+      "loss": 0.7075,
+      "step": 24330
+    },
+    {
+      "epoch": 3.935009295934039,
+      "grad_norm": 0.9521116018295288,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24340
+    },
+    {
+      "epoch": 3.9366259801147847,
+      "grad_norm": 0.8366776704788208,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 24350
+    },
+    {
+      "epoch": 3.93824266429553,
+      "grad_norm": 0.8972663283348083,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 24360
+    },
+    {
+      "epoch": 3.939859348476275,
+      "grad_norm": 0.8102919459342957,
+      "learning_rate": 0.0002,
+      "loss": 0.661,
+      "step": 24370
+    },
+    {
+      "epoch": 3.9414760326570204,
+      "grad_norm": 0.8189975023269653,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 24380
+    },
+    {
+      "epoch": 3.9430927168377656,
+      "grad_norm": 0.9569464921951294,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 24390
+    },
+    {
+      "epoch": 3.9447094010185113,
+      "grad_norm": 0.7459101676940918,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 24400
+    },
+    {
+      "epoch": 3.9463260851992565,
+      "grad_norm": 0.8536974787712097,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 24410
+    },
+    {
+      "epoch": 3.9479427693800018,
+      "grad_norm": 0.8763698935508728,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 24420
+    },
+    {
+      "epoch": 3.949559453560747,
+      "grad_norm": 0.9381106495857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6478,
+      "step": 24430
+    },
+    {
+      "epoch": 3.9511761377414922,
+      "grad_norm": 0.934440016746521,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 24440
+    },
+    {
+      "epoch": 3.9527928219222375,
+      "grad_norm": 0.903918981552124,
+      "learning_rate": 0.0002,
+      "loss": 0.6393,
+      "step": 24450
+    },
+    {
+      "epoch": 3.9544095061029827,
+      "grad_norm": 0.8771953582763672,
+      "learning_rate": 0.0002,
+      "loss": 0.6175,
+      "step": 24460
+    },
+    {
+      "epoch": 3.956026190283728,
+      "grad_norm": 1.0375410318374634,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 24470
+    },
+    {
+      "epoch": 3.957642874464473,
+      "grad_norm": 0.9439185261726379,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 24480
+    },
+    {
+      "epoch": 3.9592595586452184,
+      "grad_norm": 0.935467004776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 24490
+    },
+    {
+      "epoch": 3.960876242825964,
+      "grad_norm": 0.6900772452354431,
+      "learning_rate": 0.0002,
+      "loss": 0.6437,
+      "step": 24500
+    },
+    {
+      "epoch": 3.9624929270067093,
+      "grad_norm": 1.0172916650772095,
+      "learning_rate": 0.0002,
+      "loss": 0.6445,
+      "step": 24510
+    },
+    {
+      "epoch": 3.9641096111874545,
+      "grad_norm": 0.9167046546936035,
+      "learning_rate": 0.0002,
+      "loss": 0.6308,
+      "step": 24520
+    },
+    {
+      "epoch": 3.9657262953681998,
+      "grad_norm": 0.7230527997016907,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 24530
+    },
+    {
+      "epoch": 3.967342979548945,
+      "grad_norm": 0.8980403542518616,
+      "learning_rate": 0.0002,
+      "loss": 0.6564,
+      "step": 24540
+    },
+    {
+      "epoch": 3.9689596637296907,
+      "grad_norm": 0.8555465936660767,
+      "learning_rate": 0.0002,
+      "loss": 0.6099,
+      "step": 24550
+    },
+    {
+      "epoch": 3.970576347910436,
+      "grad_norm": 0.7825445532798767,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 24560
+    },
+    {
+      "epoch": 3.972193032091181,
+      "grad_norm": 0.7273133993148804,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 24570
+    },
+    {
+      "epoch": 3.9738097162719264,
+      "grad_norm": 0.9612047672271729,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 24580
+    },
+    {
+      "epoch": 3.9754264004526716,
+      "grad_norm": 0.9865460991859436,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 24590
+    },
+    {
+      "epoch": 3.977043084633417,
+      "grad_norm": 0.8638762831687927,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 24600
+    },
+    {
+      "epoch": 3.978659768814162,
+      "grad_norm": 1.0096198320388794,
+      "learning_rate": 0.0002,
+      "loss": 0.6097,
+      "step": 24610
+    },
+    {
+      "epoch": 3.9802764529949073,
+      "grad_norm": 0.8475532531738281,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 24620
+    },
+    {
+      "epoch": 3.9818931371756525,
+      "grad_norm": 0.9696195721626282,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 24630
+    },
+    {
+      "epoch": 3.9835098213563978,
+      "grad_norm": 0.7499843239784241,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 24640
+    },
+    {
+      "epoch": 3.9851265055371434,
+      "grad_norm": 0.8865424990653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6054,
+      "step": 24650
+    },
+    {
+      "epoch": 3.9867431897178887,
+      "grad_norm": 0.8089959025382996,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 24660
+    },
+    {
+      "epoch": 3.988359873898634,
+      "grad_norm": 0.6946012377738953,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 24670
+    },
+    {
+      "epoch": 3.989976558079379,
+      "grad_norm": 0.7991759181022644,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 24680
+    },
+    {
+      "epoch": 3.9915932422601244,
+      "grad_norm": 0.8803931474685669,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 24690
+    },
+    {
+      "epoch": 3.99320992644087,
+      "grad_norm": 0.8848299980163574,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 24700
+    },
+    {
+      "epoch": 3.9948266106216153,
+      "grad_norm": 0.7448889017105103,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 24710
+    },
+    {
+      "epoch": 3.9964432948023605,
+      "grad_norm": 0.9361620545387268,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24720
+    },
+    {
+      "epoch": 3.9980599789831057,
+      "grad_norm": 0.9958081245422363,
+      "learning_rate": 0.0002,
+      "loss": 0.5917,
+      "step": 24730
+    },
+    {
+      "epoch": 3.999676663163851,
+      "grad_norm": 1.026004672050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6567,
+      "step": 24740
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.1524168252944946,
+      "eval_runtime": 122.1585,
+      "eval_samples_per_second": 6.0,
+      "eval_steps_per_second": 0.753,
+      "step": 24742
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 49480,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1450043604181975e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/training_args.bin b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..155b12fa9acbc6e71dba75c92bfa79e152397ebf
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-24742/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28694d5564a2b5c7d6881d4ba2af103356aa22489d2c22768ebbe47283c0f4a1
+size 5560
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/README.md b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..835e31ab8469ee39ddc8b2b6b2143a8c66dad510
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/README.md
@@ -0,0 +1,202 @@
+---
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.13.1
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/adapter_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..406c5a08dc4a2a33b52c62a482f98c217c417215
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/adapter_config.json
@@ -0,0 +1,29 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/adapter_model.safetensors b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..ce80f21ed1430250bb2ab7d46153081cf1e22877
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e04d544e4567e29a67e15f9106733f490e4c61532a4c007dbebfeecbd9ff4ec
+size 109069176
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/optimizer.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..3342d83a2f98f5c73c723fb6acad633f3bb5c270
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6c898a033af64e51f08c03ec5d2a5e2b26f6c694581bc3ab8111a68bae869bd
+size 55532666
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/rng_state.pth b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1a80f7f758218b04a02a67ab397d6b0e8797938b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b421df5e4949ec6ebb6abd76b5b9cb0ccf6cbe8d92c3f93f3da670bfcfed14a
+size 14244
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/scheduler.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..46397082e444ae5e9364813106f3681d944fd9b2
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3901a3f254a869f9b316e3e369e0229c1e3af371b3b635bc272919551abe800b
+size 1064
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/special_tokens_map.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/special_tokens_map.json
@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/tokenizer.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..86a33946b0c77216d2cce91bb28c8fada4a5e80b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/tokenizer_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..061e40d9db3253624f86e8e364c15ef546527c9d
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/tokenizer_config.json
@@ -0,0 +1,2063 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/trainer_state.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..a280dfb0b2cea5818094da1a41a0c216d16a5214
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/trainer_state.json
@@ -0,0 +1,21717 @@
+{
+  "best_metric": 1.0871200561523438,
+  "best_model_checkpoint": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185",
+  "epoch": 4.9999191657909625,
+  "eval_steps": 10,
+  "global_step": 30927,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016166841807452913,
+      "grad_norm": 0.9894065856933594,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032333683614905826,
+      "grad_norm": 1.7810699939727783,
+      "learning_rate": 0.0002,
+      "loss": 1.1528,
+      "step": 20
+    },
+    {
+      "epoch": 0.004850052542235874,
+      "grad_norm": 0.5969577431678772,
+      "learning_rate": 0.0002,
+      "loss": 0.9767,
+      "step": 30
+    },
+    {
+      "epoch": 0.006466736722981165,
+      "grad_norm": 0.6354120969772339,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 40
+    },
+    {
+      "epoch": 0.008083420903726457,
+      "grad_norm": 0.5604607462882996,
+      "learning_rate": 0.0002,
+      "loss": 0.8643,
+      "step": 50
+    },
+    {
+      "epoch": 0.009700105084471748,
+      "grad_norm": 0.4676193594932556,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 60
+    },
+    {
+      "epoch": 0.01131678926521704,
+      "grad_norm": 0.6099211573600769,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 70
+    },
+    {
+      "epoch": 0.01293347344596233,
+      "grad_norm": 0.48639994859695435,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 80
+    },
+    {
+      "epoch": 0.014550157626707623,
+      "grad_norm": 0.4904264509677887,
+      "learning_rate": 0.0002,
+      "loss": 0.8704,
+      "step": 90
+    },
+    {
+      "epoch": 0.016166841807452915,
+      "grad_norm": 2.8334362506866455,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 100
+    },
+    {
+      "epoch": 0.017783525988198205,
+      "grad_norm": 0.43221670389175415,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 110
+    },
+    {
+      "epoch": 0.019400210168943496,
+      "grad_norm": 0.42244166135787964,
+      "learning_rate": 0.0002,
+      "loss": 0.8412,
+      "step": 120
+    },
+    {
+      "epoch": 0.02101689434968879,
+      "grad_norm": 0.45363298058509827,
+      "learning_rate": 0.0002,
+      "loss": 0.8467,
+      "step": 130
+    },
+    {
+      "epoch": 0.02263357853043408,
+      "grad_norm": 0.44816508889198303,
+      "learning_rate": 0.0002,
+      "loss": 0.8641,
+      "step": 140
+    },
+    {
+      "epoch": 0.02425026271117937,
+      "grad_norm": 0.43308213353157043,
+      "learning_rate": 0.0002,
+      "loss": 0.8496,
+      "step": 150
+    },
+    {
+      "epoch": 0.02586694689192466,
+      "grad_norm": 0.4084763526916504,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 160
+    },
+    {
+      "epoch": 0.027483631072669955,
+      "grad_norm": 0.5363703966140747,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 170
+    },
+    {
+      "epoch": 0.029100315253415245,
+      "grad_norm": 0.4619699716567993,
+      "learning_rate": 0.0002,
+      "loss": 0.8558,
+      "step": 180
+    },
+    {
+      "epoch": 0.030716999434160536,
+      "grad_norm": 0.49069908261299133,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 190
+    },
+    {
+      "epoch": 0.03233368361490583,
+      "grad_norm": 0.4645835757255554,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 200
+    },
+    {
+      "epoch": 0.03395036779565112,
+      "grad_norm": 1.2411243915557861,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 210
+    },
+    {
+      "epoch": 0.03556705197639641,
+      "grad_norm": 0.5211851596832275,
+      "learning_rate": 0.0002,
+      "loss": 0.8245,
+      "step": 220
+    },
+    {
+      "epoch": 0.037183736157141704,
+      "grad_norm": 0.5253691673278809,
+      "learning_rate": 0.0002,
+      "loss": 0.8194,
+      "step": 230
+    },
+    {
+      "epoch": 0.03880042033788699,
+      "grad_norm": 0.4567478895187378,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 240
+    },
+    {
+      "epoch": 0.040417104518632285,
+      "grad_norm": 0.5472128391265869,
+      "learning_rate": 0.0002,
+      "loss": 0.838,
+      "step": 250
+    },
+    {
+      "epoch": 0.04203378869937758,
+      "grad_norm": 0.42978546023368835,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 260
+    },
+    {
+      "epoch": 0.043650472880122866,
+      "grad_norm": 0.601734459400177,
+      "learning_rate": 0.0002,
+      "loss": 0.8334,
+      "step": 270
+    },
+    {
+      "epoch": 0.04526715706086816,
+      "grad_norm": 0.4286513328552246,
+      "learning_rate": 0.0002,
+      "loss": 0.815,
+      "step": 280
+    },
+    {
+      "epoch": 0.046883841241613454,
+      "grad_norm": 0.5230861902236938,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 290
+    },
+    {
+      "epoch": 0.04850052542235874,
+      "grad_norm": 0.6504611968994141,
+      "learning_rate": 0.0002,
+      "loss": 0.8636,
+      "step": 300
+    },
+    {
+      "epoch": 0.050117209603104035,
+      "grad_norm": 0.43485215306282043,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 310
+    },
+    {
+      "epoch": 0.05173389378384932,
+      "grad_norm": 0.4717007875442505,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 320
+    },
+    {
+      "epoch": 0.053350577964594616,
+      "grad_norm": 0.4059787690639496,
+      "learning_rate": 0.0002,
+      "loss": 0.8469,
+      "step": 330
+    },
+    {
+      "epoch": 0.05496726214533991,
+      "grad_norm": 0.4366913437843323,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 340
+    },
+    {
+      "epoch": 0.0565839463260852,
+      "grad_norm": 0.4233848452568054,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 350
+    },
+    {
+      "epoch": 0.05820063050683049,
+      "grad_norm": 0.4209108352661133,
+      "learning_rate": 0.0002,
+      "loss": 0.8456,
+      "step": 360
+    },
+    {
+      "epoch": 0.059817314687575784,
+      "grad_norm": 0.41637396812438965,
+      "learning_rate": 0.0002,
+      "loss": 0.816,
+      "step": 370
+    },
+    {
+      "epoch": 0.06143399886832107,
+      "grad_norm": 0.46235376596450806,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 380
+    },
+    {
+      "epoch": 0.06305068304906636,
+      "grad_norm": 0.4013484716415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 390
+    },
+    {
+      "epoch": 0.06466736722981166,
+      "grad_norm": 0.47443896532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 400
+    },
+    {
+      "epoch": 0.06628405141055695,
+      "grad_norm": 0.3942156434059143,
+      "learning_rate": 0.0002,
+      "loss": 0.8666,
+      "step": 410
+    },
+    {
+      "epoch": 0.06790073559130223,
+      "grad_norm": 0.4965320825576782,
+      "learning_rate": 0.0002,
+      "loss": 0.8402,
+      "step": 420
+    },
+    {
+      "epoch": 0.06951741977204753,
+      "grad_norm": 0.4304835796356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8317,
+      "step": 430
+    },
+    {
+      "epoch": 0.07113410395279282,
+      "grad_norm": 0.511726975440979,
+      "learning_rate": 0.0002,
+      "loss": 0.8528,
+      "step": 440
+    },
+    {
+      "epoch": 0.07275078813353811,
+      "grad_norm": 0.4040689170360565,
+      "learning_rate": 0.0002,
+      "loss": 0.8675,
+      "step": 450
+    },
+    {
+      "epoch": 0.07436747231428341,
+      "grad_norm": 0.5402171015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 460
+    },
+    {
+      "epoch": 0.0759841564950287,
+      "grad_norm": 0.4174517095088959,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 470
+    },
+    {
+      "epoch": 0.07760084067577398,
+      "grad_norm": 0.4306182265281677,
+      "learning_rate": 0.0002,
+      "loss": 0.7605,
+      "step": 480
+    },
+    {
+      "epoch": 0.07921752485651928,
+      "grad_norm": 0.535210132598877,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 490
+    },
+    {
+      "epoch": 0.08083420903726457,
+      "grad_norm": 0.5339109897613525,
+      "learning_rate": 0.0002,
+      "loss": 0.7825,
+      "step": 500
+    },
+    {
+      "epoch": 0.08245089321800986,
+      "grad_norm": 0.45754891633987427,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 510
+    },
+    {
+      "epoch": 0.08406757739875516,
+      "grad_norm": 0.43820783495903015,
+      "learning_rate": 0.0002,
+      "loss": 0.8144,
+      "step": 520
+    },
+    {
+      "epoch": 0.08568426157950045,
+      "grad_norm": 0.4434749186038971,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 530
+    },
+    {
+      "epoch": 0.08730094576024573,
+      "grad_norm": 0.43111467361450195,
+      "learning_rate": 0.0002,
+      "loss": 0.7857,
+      "step": 540
+    },
+    {
+      "epoch": 0.08891762994099103,
+      "grad_norm": 0.4378940165042877,
+      "learning_rate": 0.0002,
+      "loss": 0.8418,
+      "step": 550
+    },
+    {
+      "epoch": 0.09053431412173632,
+      "grad_norm": 0.4772215187549591,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 560
+    },
+    {
+      "epoch": 0.09215099830248161,
+      "grad_norm": 0.6837629079818726,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 570
+    },
+    {
+      "epoch": 0.09376768248322691,
+      "grad_norm": 0.42241212725639343,
+      "learning_rate": 0.0002,
+      "loss": 0.8607,
+      "step": 580
+    },
+    {
+      "epoch": 0.0953843666639722,
+      "grad_norm": 0.5165936350822449,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 590
+    },
+    {
+      "epoch": 0.09700105084471748,
+      "grad_norm": 0.48737478256225586,
+      "learning_rate": 0.0002,
+      "loss": 0.8664,
+      "step": 600
+    },
+    {
+      "epoch": 0.09861773502546278,
+      "grad_norm": 0.47419852018356323,
+      "learning_rate": 0.0002,
+      "loss": 0.8806,
+      "step": 610
+    },
+    {
+      "epoch": 0.10023441920620807,
+      "grad_norm": 0.4975486099720001,
+      "learning_rate": 0.0002,
+      "loss": 0.8254,
+      "step": 620
+    },
+    {
+      "epoch": 0.10185110338695336,
+      "grad_norm": 0.49123844504356384,
+      "learning_rate": 0.0002,
+      "loss": 0.8548,
+      "step": 630
+    },
+    {
+      "epoch": 0.10346778756769864,
+      "grad_norm": 0.6288952827453613,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 640
+    },
+    {
+      "epoch": 0.10508447174844394,
+      "grad_norm": 0.4277345836162567,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 650
+    },
+    {
+      "epoch": 0.10670115592918923,
+      "grad_norm": 0.4021061956882477,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 660
+    },
+    {
+      "epoch": 0.10831784010993452,
+      "grad_norm": 0.3492237329483032,
+      "learning_rate": 0.0002,
+      "loss": 0.87,
+      "step": 670
+    },
+    {
+      "epoch": 0.10993452429067982,
+      "grad_norm": 0.4341012239456177,
+      "learning_rate": 0.0002,
+      "loss": 0.8698,
+      "step": 680
+    },
+    {
+      "epoch": 0.1115512084714251,
+      "grad_norm": 0.7296304106712341,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 690
+    },
+    {
+      "epoch": 0.1131678926521704,
+      "grad_norm": 0.397494912147522,
+      "learning_rate": 0.0002,
+      "loss": 0.8433,
+      "step": 700
+    },
+    {
+      "epoch": 0.1147845768329157,
+      "grad_norm": 0.396431028842926,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 710
+    },
+    {
+      "epoch": 0.11640126101366098,
+      "grad_norm": 0.48842838406562805,
+      "learning_rate": 0.0002,
+      "loss": 0.8379,
+      "step": 720
+    },
+    {
+      "epoch": 0.11801794519440627,
+      "grad_norm": 0.46322616934776306,
+      "learning_rate": 0.0002,
+      "loss": 0.8238,
+      "step": 730
+    },
+    {
+      "epoch": 0.11963462937515157,
+      "grad_norm": 0.47990912199020386,
+      "learning_rate": 0.0002,
+      "loss": 0.8041,
+      "step": 740
+    },
+    {
+      "epoch": 0.12125131355589686,
+      "grad_norm": 0.4997142255306244,
+      "learning_rate": 0.0002,
+      "loss": 0.82,
+      "step": 750
+    },
+    {
+      "epoch": 0.12286799773664214,
+      "grad_norm": 0.4040526747703552,
+      "learning_rate": 0.0002,
+      "loss": 0.7702,
+      "step": 760
+    },
+    {
+      "epoch": 0.12448468191738744,
+      "grad_norm": 0.453095942735672,
+      "learning_rate": 0.0002,
+      "loss": 0.863,
+      "step": 770
+    },
+    {
+      "epoch": 0.12610136609813272,
+      "grad_norm": 0.4636971950531006,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 780
+    },
+    {
+      "epoch": 0.12771805027887803,
+      "grad_norm": 0.4279276132583618,
+      "learning_rate": 0.0002,
+      "loss": 0.8112,
+      "step": 790
+    },
+    {
+      "epoch": 0.12933473445962332,
+      "grad_norm": 0.46212655305862427,
+      "learning_rate": 0.0002,
+      "loss": 0.8711,
+      "step": 800
+    },
+    {
+      "epoch": 0.1309514186403686,
+      "grad_norm": 0.43127650022506714,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 810
+    },
+    {
+      "epoch": 0.1325681028211139,
+      "grad_norm": 0.4201301336288452,
+      "learning_rate": 0.0002,
+      "loss": 0.8476,
+      "step": 820
+    },
+    {
+      "epoch": 0.13418478700185918,
+      "grad_norm": 0.42583167552948,
+      "learning_rate": 0.0002,
+      "loss": 0.8078,
+      "step": 830
+    },
+    {
+      "epoch": 0.13580147118260447,
+      "grad_norm": 0.4535622000694275,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 840
+    },
+    {
+      "epoch": 0.13741815536334978,
+      "grad_norm": 0.4116036891937256,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 850
+    },
+    {
+      "epoch": 0.13903483954409507,
+      "grad_norm": 0.45997580885887146,
+      "learning_rate": 0.0002,
+      "loss": 0.8466,
+      "step": 860
+    },
+    {
+      "epoch": 0.14065152372484035,
+      "grad_norm": 0.4487837255001068,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 870
+    },
+    {
+      "epoch": 0.14226820790558564,
+      "grad_norm": 0.43650057911872864,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 880
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 0.5335358381271362,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 890
+    },
+    {
+      "epoch": 0.14550157626707622,
+      "grad_norm": 0.5989000201225281,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 900
+    },
+    {
+      "epoch": 0.14711826044782153,
+      "grad_norm": 0.517179012298584,
+      "learning_rate": 0.0002,
+      "loss": 0.8385,
+      "step": 910
+    },
+    {
+      "epoch": 0.14873494462856682,
+      "grad_norm": 0.44435232877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.8255,
+      "step": 920
+    },
+    {
+      "epoch": 0.1503516288093121,
+      "grad_norm": 0.42635923624038696,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 930
+    },
+    {
+      "epoch": 0.1519683129900574,
+      "grad_norm": 0.49603334069252014,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 940
+    },
+    {
+      "epoch": 0.15358499717080268,
+      "grad_norm": 0.40639808773994446,
+      "learning_rate": 0.0002,
+      "loss": 0.8377,
+      "step": 950
+    },
+    {
+      "epoch": 0.15520168135154797,
+      "grad_norm": 0.4850759208202362,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 960
+    },
+    {
+      "epoch": 0.15681836553229328,
+      "grad_norm": 0.4427442252635956,
+      "learning_rate": 0.0002,
+      "loss": 0.846,
+      "step": 970
+    },
+    {
+      "epoch": 0.15843504971303857,
+      "grad_norm": 0.3760930001735687,
+      "learning_rate": 0.0002,
+      "loss": 0.8705,
+      "step": 980
+    },
+    {
+      "epoch": 0.16005173389378385,
+      "grad_norm": 0.4794144332408905,
+      "learning_rate": 0.0002,
+      "loss": 0.8644,
+      "step": 990
+    },
+    {
+      "epoch": 0.16166841807452914,
+      "grad_norm": 0.45828768610954285,
+      "learning_rate": 0.0002,
+      "loss": 0.8002,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16328510225527443,
+      "grad_norm": 0.6313053369522095,
+      "learning_rate": 0.0002,
+      "loss": 0.7658,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16490178643601971,
+      "grad_norm": 0.45041006803512573,
+      "learning_rate": 0.0002,
+      "loss": 0.8047,
+      "step": 1020
+    },
+    {
+      "epoch": 0.166518470616765,
+      "grad_norm": 0.441403865814209,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16813515479751032,
+      "grad_norm": 0.8171296119689941,
+      "learning_rate": 0.0002,
+      "loss": 0.8475,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1697518389782556,
+      "grad_norm": 0.7137420773506165,
+      "learning_rate": 0.0002,
+      "loss": 0.845,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1713685231590009,
+      "grad_norm": 0.5236809849739075,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17298520733974618,
+      "grad_norm": 0.5021864175796509,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17460189152049146,
+      "grad_norm": 0.47347521781921387,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17621857570123675,
+      "grad_norm": 0.4631653428077698,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17783525988198207,
+      "grad_norm": 0.49169182777404785,
+      "learning_rate": 0.0002,
+      "loss": 0.8166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17945194406272735,
+      "grad_norm": 0.5019739270210266,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18106862824347264,
+      "grad_norm": 0.5100422501564026,
+      "learning_rate": 0.0002,
+      "loss": 0.8247,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18268531242421793,
+      "grad_norm": 0.3888324499130249,
+      "learning_rate": 0.0002,
+      "loss": 0.8142,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18430199660496321,
+      "grad_norm": 0.39765217900276184,
+      "learning_rate": 0.0002,
+      "loss": 0.8533,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1859186807857085,
+      "grad_norm": 0.47190186381340027,
+      "learning_rate": 0.0002,
+      "loss": 0.8541,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18753536496645382,
+      "grad_norm": 0.4464188814163208,
+      "learning_rate": 0.0002,
+      "loss": 0.8301,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1891520491471991,
+      "grad_norm": 0.5153930187225342,
+      "learning_rate": 0.0002,
+      "loss": 0.8341,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1907687333279444,
+      "grad_norm": 0.4779708683490753,
+      "learning_rate": 0.0002,
+      "loss": 0.8033,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19238541750868968,
+      "grad_norm": 0.4834315776824951,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19400210168943496,
+      "grad_norm": 0.402357816696167,
+      "learning_rate": 0.0002,
+      "loss": 0.7721,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19561878587018025,
+      "grad_norm": 0.45899084210395813,
+      "learning_rate": 0.0002,
+      "loss": 0.7941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19723547005092557,
+      "grad_norm": 0.5106529593467712,
+      "learning_rate": 0.0002,
+      "loss": 0.8353,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19885215423167085,
+      "grad_norm": 0.45261722803115845,
+      "learning_rate": 0.0002,
+      "loss": 0.7816,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20046883841241614,
+      "grad_norm": 0.4647127091884613,
+      "learning_rate": 0.0002,
+      "loss": 0.8068,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20208552259316143,
+      "grad_norm": 0.4849368929862976,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2037022067739067,
+      "grad_norm": 0.4518061578273773,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 1260
+    },
+    {
+      "epoch": 0.205318890954652,
+      "grad_norm": 0.49535325169563293,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2069355751353973,
+      "grad_norm": 0.4835205376148224,
+      "learning_rate": 0.0002,
+      "loss": 0.8348,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2085522593161426,
+      "grad_norm": 0.45308539271354675,
+      "learning_rate": 0.0002,
+      "loss": 0.8428,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2101689434968879,
+      "grad_norm": 0.5369905233383179,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21178562767763318,
+      "grad_norm": 0.5031622052192688,
+      "learning_rate": 0.0002,
+      "loss": 0.8676,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21340231185837846,
+      "grad_norm": 0.48010334372520447,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21501899603912375,
+      "grad_norm": 0.4905701279640198,
+      "learning_rate": 0.0002,
+      "loss": 0.806,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21663568021986904,
+      "grad_norm": 0.43531742691993713,
+      "learning_rate": 0.0002,
+      "loss": 0.7885,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21825236440061435,
+      "grad_norm": 0.44330692291259766,
+      "learning_rate": 0.0002,
+      "loss": 0.8191,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21986904858135964,
+      "grad_norm": 0.5384416580200195,
+      "learning_rate": 0.0002,
+      "loss": 0.8205,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22148573276210493,
+      "grad_norm": 0.4181833863258362,
+      "learning_rate": 0.0002,
+      "loss": 0.7726,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2231024169428502,
+      "grad_norm": 0.523833692073822,
+      "learning_rate": 0.0002,
+      "loss": 0.8311,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.5528736710548401,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2263357853043408,
+      "grad_norm": 0.43515023589134216,
+      "learning_rate": 0.0002,
+      "loss": 0.8079,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2279524694850861,
+      "grad_norm": 0.48809877038002014,
+      "learning_rate": 0.0002,
+      "loss": 0.8403,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2295691536658314,
+      "grad_norm": 0.43591251969337463,
+      "learning_rate": 0.0002,
+      "loss": 0.8165,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23118583784657668,
+      "grad_norm": 0.44625312089920044,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23280252202732196,
+      "grad_norm": 0.4390665292739868,
+      "learning_rate": 0.0002,
+      "loss": 0.8134,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23441920620806725,
+      "grad_norm": 0.48496049642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.8465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23603589038881254,
+      "grad_norm": 0.45919957756996155,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23765257456955785,
+      "grad_norm": 0.5471845865249634,
+      "learning_rate": 0.0002,
+      "loss": 0.8659,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23926925875030314,
+      "grad_norm": 0.47269317507743835,
+      "learning_rate": 0.0002,
+      "loss": 0.8164,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24088594293104842,
+      "grad_norm": 0.4930245578289032,
+      "learning_rate": 0.0002,
+      "loss": 0.854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2425026271117937,
+      "grad_norm": 0.5605630278587341,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 1500
+    },
+    {
+      "epoch": 0.244119311292539,
+      "grad_norm": 0.4435870945453644,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24573599547328429,
+      "grad_norm": 0.4941999912261963,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24735267965402957,
+      "grad_norm": 0.5100624561309814,
+      "learning_rate": 0.0002,
+      "loss": 0.8427,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2489693638347749,
+      "grad_norm": 0.4638267457485199,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25058604801552015,
+      "grad_norm": 0.5071570873260498,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25220273219626543,
+      "grad_norm": 0.4291319251060486,
+      "learning_rate": 0.0002,
+      "loss": 0.7724,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2538194163770108,
+      "grad_norm": 0.5388049483299255,
+      "learning_rate": 0.0002,
+      "loss": 0.7984,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25543610055775606,
+      "grad_norm": 0.5083683729171753,
+      "learning_rate": 0.0002,
+      "loss": 0.8176,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25705278473850135,
+      "grad_norm": 0.4824463725090027,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25866946891924664,
+      "grad_norm": 0.41177722811698914,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2602861530999919,
+      "grad_norm": 0.5656219124794006,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2619028372807372,
+      "grad_norm": 0.41063204407691956,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2635195214614825,
+      "grad_norm": 0.4897061288356781,
+      "learning_rate": 0.0002,
+      "loss": 0.7998,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2651362056422278,
+      "grad_norm": 0.4454376697540283,
+      "learning_rate": 0.0002,
+      "loss": 0.8198,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26675288982297307,
+      "grad_norm": 0.4355238378047943,
+      "learning_rate": 0.0002,
+      "loss": 0.8684,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26836957400371836,
+      "grad_norm": 0.458310067653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26998625818446365,
+      "grad_norm": 0.4752083718776703,
+      "learning_rate": 0.0002,
+      "loss": 0.7935,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27160294236520893,
+      "grad_norm": 0.4666106402873993,
+      "learning_rate": 0.0002,
+      "loss": 0.8267,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2732196265459543,
+      "grad_norm": 0.4213818609714508,
+      "learning_rate": 0.0002,
+      "loss": 0.8252,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27483631072669956,
+      "grad_norm": 0.5768913626670837,
+      "learning_rate": 0.0002,
+      "loss": 0.8559,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27645299490744485,
+      "grad_norm": 0.4209914803504944,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27806967908819014,
+      "grad_norm": 0.501909613609314,
+      "learning_rate": 0.0002,
+      "loss": 0.8167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2796863632689354,
+      "grad_norm": 0.5266261100769043,
+      "learning_rate": 0.0002,
+      "loss": 0.7832,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2813030474496807,
+      "grad_norm": 0.43806859850883484,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 1740
+    },
+    {
+      "epoch": 0.282919731630426,
+      "grad_norm": 0.46048814058303833,
+      "learning_rate": 0.0002,
+      "loss": 0.8157,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2845364158111713,
+      "grad_norm": 0.44972819089889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8596,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28615309999191657,
+      "grad_norm": 0.5114831328392029,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.47931742668151855,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28938646835340714,
+      "grad_norm": 0.5092599987983704,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29100315253415243,
+      "grad_norm": 0.37581443786621094,
+      "learning_rate": 0.0002,
+      "loss": 0.8506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2926198367148977,
+      "grad_norm": 0.47097381949424744,
+      "learning_rate": 0.0002,
+      "loss": 0.7932,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29423652089564306,
+      "grad_norm": 0.48300236463546753,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29585320507638835,
+      "grad_norm": 0.5600419640541077,
+      "learning_rate": 0.0002,
+      "loss": 0.8391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29746988925713364,
+      "grad_norm": 0.48555272817611694,
+      "learning_rate": 0.0002,
+      "loss": 0.8507,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2990865734378789,
+      "grad_norm": 0.3752668499946594,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3007032576186242,
+      "grad_norm": 0.5328747034072876,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3023199417993695,
+      "grad_norm": 0.48716455698013306,
+      "learning_rate": 0.0002,
+      "loss": 0.8426,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3039366259801148,
+      "grad_norm": 0.5011493563652039,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30555331016086007,
+      "grad_norm": 0.46461427211761475,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30716999434160536,
+      "grad_norm": 0.36630210280418396,
+      "learning_rate": 0.0002,
+      "loss": 0.8478,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30878667852235064,
+      "grad_norm": 0.4217296242713928,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31040336270309593,
+      "grad_norm": 0.4394875466823578,
+      "learning_rate": 0.0002,
+      "loss": 0.8128,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3120200468838412,
+      "grad_norm": 0.6587965488433838,
+      "learning_rate": 0.0002,
+      "loss": 0.8471,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31363673106458656,
+      "grad_norm": 0.5469298958778381,
+      "learning_rate": 0.0002,
+      "loss": 0.8565,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31525341524533185,
+      "grad_norm": 0.4371595084667206,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31687009942607713,
+      "grad_norm": 0.4809541404247284,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3184867836068224,
+      "grad_norm": 0.6061086654663086,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3201034677875677,
+      "grad_norm": 0.5342657566070557,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 1980
+    },
+    {
+      "epoch": 0.321720151968313,
+      "grad_norm": 0.5057743787765503,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3233368361490583,
+      "grad_norm": 0.528626024723053,
+      "learning_rate": 0.0002,
+      "loss": 0.7774,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32495352032980357,
+      "grad_norm": 0.46742770075798035,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32657020451054886,
+      "grad_norm": 0.515101432800293,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32818688869129414,
+      "grad_norm": 0.41941216588020325,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32980357287203943,
+      "grad_norm": 0.49902522563934326,
+      "learning_rate": 0.0002,
+      "loss": 0.7637,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3314202570527847,
+      "grad_norm": 0.4120897650718689,
+      "learning_rate": 0.0002,
+      "loss": 0.7822,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33303694123353,
+      "grad_norm": 0.45352041721343994,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33465362541427535,
+      "grad_norm": 0.523199737071991,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33627030959502063,
+      "grad_norm": 0.4390358626842499,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3378869937757659,
+      "grad_norm": 0.6752901077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3395036779565112,
+      "grad_norm": 0.547821044921875,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3411203621372565,
+      "grad_norm": 0.5161308646202087,
+      "learning_rate": 0.0002,
+      "loss": 0.8561,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3427370463180018,
+      "grad_norm": 0.4565401077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34435373049874707,
+      "grad_norm": 0.4666115939617157,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34597041467949236,
+      "grad_norm": 0.4090428352355957,
+      "learning_rate": 0.0002,
+      "loss": 0.8189,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34758709886023764,
+      "grad_norm": 0.510845422744751,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34920378304098293,
+      "grad_norm": 0.42861923575401306,
+      "learning_rate": 0.0002,
+      "loss": 0.8398,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3508204672217282,
+      "grad_norm": 0.4476332664489746,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3524371514024735,
+      "grad_norm": 0.6065791249275208,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35405383558321885,
+      "grad_norm": 0.42335066199302673,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35567051976396413,
+      "grad_norm": 0.5094629526138306,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3572872039447094,
+      "grad_norm": 0.5476373434066772,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3589038881254547,
+      "grad_norm": 0.3911719024181366,
+      "learning_rate": 0.0002,
+      "loss": 0.814,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3605205723062,
+      "grad_norm": 0.6599636077880859,
+      "learning_rate": 0.0002,
+      "loss": 0.8599,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3621372564869453,
+      "grad_norm": 0.40381914377212524,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36375394066769057,
+      "grad_norm": 0.4433908462524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36537062484843585,
+      "grad_norm": 0.578326940536499,
+      "learning_rate": 0.0002,
+      "loss": 0.8503,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36698730902918114,
+      "grad_norm": 0.5734784007072449,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36860399320992643,
+      "grad_norm": 0.45555487275123596,
+      "learning_rate": 0.0002,
+      "loss": 0.8193,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3702206773906717,
+      "grad_norm": 0.5666276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.7929,
+      "step": 2290
+    },
+    {
+      "epoch": 0.371837361571417,
+      "grad_norm": 0.5461117625236511,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3734540457521623,
+      "grad_norm": 0.6318911910057068,
+      "learning_rate": 0.0002,
+      "loss": 0.8204,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37507072993290763,
+      "grad_norm": 0.493263304233551,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3766874141136529,
+      "grad_norm": 0.5888760089874268,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3783040982943982,
+      "grad_norm": 0.48671841621398926,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3799207824751435,
+      "grad_norm": 0.4385145306587219,
+      "learning_rate": 0.0002,
+      "loss": 0.8367,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3815374666558888,
+      "grad_norm": 0.5523318648338318,
+      "learning_rate": 0.0002,
+      "loss": 0.812,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38315415083663407,
+      "grad_norm": 0.7308220267295837,
+      "learning_rate": 0.0002,
+      "loss": 0.8351,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38477083501737935,
+      "grad_norm": 0.554214358329773,
+      "learning_rate": 0.0002,
+      "loss": 0.859,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38638751919812464,
+      "grad_norm": 0.5425800085067749,
+      "learning_rate": 0.0002,
+      "loss": 0.8146,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3880042033788699,
+      "grad_norm": 0.48811158537864685,
+      "learning_rate": 0.0002,
+      "loss": 0.8282,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3896208875596152,
+      "grad_norm": 0.49212366342544556,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3912375717403605,
+      "grad_norm": 0.5222218632698059,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3928542559211058,
+      "grad_norm": 0.4699819087982178,
+      "learning_rate": 0.0002,
+      "loss": 0.8182,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39447094010185113,
+      "grad_norm": 0.46153587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7919,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3960876242825964,
+      "grad_norm": 0.4150611162185669,
+      "learning_rate": 0.0002,
+      "loss": 0.8111,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3977043084633417,
+      "grad_norm": 0.5799614787101746,
+      "learning_rate": 0.0002,
+      "loss": 0.8589,
+      "step": 2460
+    },
+    {
+      "epoch": 0.399320992644087,
+      "grad_norm": 0.56536865234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4009376768248323,
+      "grad_norm": 0.5451247096061707,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40255436100557757,
+      "grad_norm": 0.5914521217346191,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40417104518632285,
+      "grad_norm": 0.4428117275238037,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40578772936706814,
+      "grad_norm": 0.48580947518348694,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4074044135478134,
+      "grad_norm": 0.436734676361084,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4090210977285587,
+      "grad_norm": 0.5752223134040833,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 2530
+    },
+    {
+      "epoch": 0.410637781909304,
+      "grad_norm": 0.4271308183670044,
+      "learning_rate": 0.0002,
+      "loss": 0.8181,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4122544660900493,
+      "grad_norm": 0.46294718980789185,
+      "learning_rate": 0.0002,
+      "loss": 0.8058,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4138711502707946,
+      "grad_norm": 0.49407583475112915,
+      "learning_rate": 0.0002,
+      "loss": 0.8473,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4154878344515399,
+      "grad_norm": 0.4729035496711731,
+      "learning_rate": 0.0002,
+      "loss": 0.7881,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4171045186322852,
+      "grad_norm": 0.4129747152328491,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4187212028130305,
+      "grad_norm": 0.5684236288070679,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4203378869937758,
+      "grad_norm": 0.4862157106399536,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42195457117452106,
+      "grad_norm": 0.46567976474761963,
+      "learning_rate": 0.0002,
+      "loss": 0.7582,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42357125535526635,
+      "grad_norm": 0.5710650682449341,
+      "learning_rate": 0.0002,
+      "loss": 0.7755,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42518793953601164,
+      "grad_norm": 0.5660041570663452,
+      "learning_rate": 0.0002,
+      "loss": 0.8573,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4268046237167569,
+      "grad_norm": 0.47944375872612,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4284213078975022,
+      "grad_norm": 0.537223756313324,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4300379920782475,
+      "grad_norm": 0.41669997572898865,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 0.44727686047554016,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4332713604397381,
+      "grad_norm": 0.5600888729095459,
+      "learning_rate": 0.0002,
+      "loss": 0.8241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4348880446204834,
+      "grad_norm": 0.39820605516433716,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4365047288012287,
+      "grad_norm": 0.5637655854225159,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 2700
+    },
+    {
+      "epoch": 0.438121412981974,
+      "grad_norm": 0.6363666653633118,
+      "learning_rate": 0.0002,
+      "loss": 0.855,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4397380971627193,
+      "grad_norm": 0.5656129121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.8468,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44135478134346456,
+      "grad_norm": 0.5600156188011169,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44297146552420985,
+      "grad_norm": 0.5506579875946045,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44458814970495514,
+      "grad_norm": 0.49878305196762085,
+      "learning_rate": 0.0002,
+      "loss": 0.7725,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4462048338857004,
+      "grad_norm": 0.4569213092327118,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4478215180664457,
+      "grad_norm": 0.6056680083274841,
+      "learning_rate": 0.0002,
+      "loss": 0.8028,
+      "step": 2770
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.44474557042121887,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4510548864279363,
+      "grad_norm": 0.46055394411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.801,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4526715706086816,
+      "grad_norm": 0.4904133379459381,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45428825478942686,
+      "grad_norm": 0.5647031664848328,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4559049389701722,
+      "grad_norm": 0.5759473443031311,
+      "learning_rate": 0.0002,
+      "loss": 0.8622,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4575216231509175,
+      "grad_norm": 0.5161895751953125,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4591383073316628,
+      "grad_norm": 0.4248254597187042,
+      "learning_rate": 0.0002,
+      "loss": 0.8045,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46075499151240806,
+      "grad_norm": 0.45395001769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.7838,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46237167569315335,
+      "grad_norm": 0.5358697772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.8208,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46398835987389864,
+      "grad_norm": 0.5379165410995483,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4656050440546439,
+      "grad_norm": 0.4601989686489105,
+      "learning_rate": 0.0002,
+      "loss": 0.7403,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4672217282353892,
+      "grad_norm": 0.671115517616272,
+      "learning_rate": 0.0002,
+      "loss": 0.8523,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4688384124161345,
+      "grad_norm": 0.4425133168697357,
+      "learning_rate": 0.0002,
+      "loss": 0.8262,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4704550965968798,
+      "grad_norm": 0.5446155071258545,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47207178077762507,
+      "grad_norm": 0.603306233882904,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47368846495837036,
+      "grad_norm": 0.5377997159957886,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4753051491391157,
+      "grad_norm": 0.4931027591228485,
+      "learning_rate": 0.0002,
+      "loss": 0.8075,
+      "step": 2940
+    },
+    {
+      "epoch": 0.476921833319861,
+      "grad_norm": 0.4711960256099701,
+      "learning_rate": 0.0002,
+      "loss": 0.8004,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4785385175006063,
+      "grad_norm": 0.5020492672920227,
+      "learning_rate": 0.0002,
+      "loss": 0.8121,
+      "step": 2960
+    },
+    {
+      "epoch": 0.48015520168135156,
+      "grad_norm": 0.5428946614265442,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48177188586209685,
+      "grad_norm": 0.5294089317321777,
+      "learning_rate": 0.0002,
+      "loss": 0.7849,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48338857004284214,
+      "grad_norm": 0.648289144039154,
+      "learning_rate": 0.0002,
+      "loss": 0.8553,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4850052542235874,
+      "grad_norm": 0.47916680574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.7874,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4866219384043327,
+      "grad_norm": 0.43849772214889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8087,
+      "step": 3010
+    },
+    {
+      "epoch": 0.488238622585078,
+      "grad_norm": 0.47007861733436584,
+      "learning_rate": 0.0002,
+      "loss": 0.7662,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4898553067658233,
+      "grad_norm": 0.6314331293106079,
+      "learning_rate": 0.0002,
+      "loss": 0.757,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49147199094656857,
+      "grad_norm": 0.49211493134498596,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49308867512731386,
+      "grad_norm": 0.4537973403930664,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49470535930805914,
+      "grad_norm": 0.47326919436454773,
+      "learning_rate": 0.0002,
+      "loss": 0.8095,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4963220434888045,
+      "grad_norm": 0.525874137878418,
+      "learning_rate": 0.0002,
+      "loss": 0.8447,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4979387276695498,
+      "grad_norm": 0.6361091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49955541185029506,
+      "grad_norm": 0.5850642919540405,
+      "learning_rate": 0.0002,
+      "loss": 0.821,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5011720960310403,
+      "grad_norm": 0.47299543023109436,
+      "learning_rate": 0.0002,
+      "loss": 0.8279,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5027887802117856,
+      "grad_norm": 0.473099946975708,
+      "learning_rate": 0.0002,
+      "loss": 0.8681,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5044054643925309,
+      "grad_norm": 0.48186397552490234,
+      "learning_rate": 0.0002,
+      "loss": 0.8223,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5060221485732762,
+      "grad_norm": 0.5015401840209961,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5076388327540216,
+      "grad_norm": 0.5617750287055969,
+      "learning_rate": 0.0002,
+      "loss": 0.7692,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5092555169347668,
+      "grad_norm": 0.5169327259063721,
+      "learning_rate": 0.0002,
+      "loss": 0.8708,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5108722011155121,
+      "grad_norm": 0.545657753944397,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5124888852962574,
+      "grad_norm": 0.512864351272583,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5141055694770027,
+      "grad_norm": 0.4113546311855316,
+      "learning_rate": 0.0002,
+      "loss": 0.7794,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5157222536577479,
+      "grad_norm": 0.44532445073127747,
+      "learning_rate": 0.0002,
+      "loss": 0.8206,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5173389378384933,
+      "grad_norm": 0.5623497366905212,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5189556220192385,
+      "grad_norm": 0.5084741115570068,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5205723061999838,
+      "grad_norm": 0.5305403470993042,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5221889903807291,
+      "grad_norm": 0.4708254337310791,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5238056745614744,
+      "grad_norm": 0.43827131390571594,
+      "learning_rate": 0.0002,
+      "loss": 0.7639,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5254223587422197,
+      "grad_norm": 0.5630002617835999,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 3250
+    },
+    {
+      "epoch": 0.527039042922965,
+      "grad_norm": 0.5010961890220642,
+      "learning_rate": 0.0002,
+      "loss": 0.7522,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5286557271037103,
+      "grad_norm": 0.6303122043609619,
+      "learning_rate": 0.0002,
+      "loss": 0.8374,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5302724112844556,
+      "grad_norm": 0.5107331275939941,
+      "learning_rate": 0.0002,
+      "loss": 0.7727,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5318890954652009,
+      "grad_norm": 0.5700443387031555,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5335057796459461,
+      "grad_norm": 0.46296367049217224,
+      "learning_rate": 0.0002,
+      "loss": 0.7776,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5351224638266915,
+      "grad_norm": 0.531568706035614,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5367391480074367,
+      "grad_norm": 0.4686741530895233,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5383558321881821,
+      "grad_norm": 0.5404331088066101,
+      "learning_rate": 0.0002,
+      "loss": 0.8104,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5399725163689273,
+      "grad_norm": 0.6368790864944458,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5415892005496726,
+      "grad_norm": 0.42300888895988464,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5432058847304179,
+      "grad_norm": 0.5362542867660522,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5448225689111632,
+      "grad_norm": 0.497128963470459,
+      "learning_rate": 0.0002,
+      "loss": 0.858,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5464392530919085,
+      "grad_norm": 0.5006386041641235,
+      "learning_rate": 0.0002,
+      "loss": 0.8519,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5480559372726538,
+      "grad_norm": 0.44136837124824524,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5496726214533991,
+      "grad_norm": 0.5897833108901978,
+      "learning_rate": 0.0002,
+      "loss": 0.773,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5512893056341444,
+      "grad_norm": 0.641075611114502,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5529059898148897,
+      "grad_norm": 0.7251322269439697,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5545226739956349,
+      "grad_norm": 0.47411349415779114,
+      "learning_rate": 0.0002,
+      "loss": 0.7626,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5561393581763803,
+      "grad_norm": 0.4994310438632965,
+      "learning_rate": 0.0002,
+      "loss": 0.8196,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5577560423571255,
+      "grad_norm": 0.5814438462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5593727265378708,
+      "grad_norm": 0.6278898119926453,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5609894107186161,
+      "grad_norm": 0.46208274364471436,
+      "learning_rate": 0.0002,
+      "loss": 0.813,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5626060948993614,
+      "grad_norm": 0.5718930959701538,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5642227790801067,
+      "grad_norm": 0.48178744316101074,
+      "learning_rate": 0.0002,
+      "loss": 0.8152,
+      "step": 3490
+    },
+    {
+      "epoch": 0.565839463260852,
+      "grad_norm": 0.47336965799331665,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5674561474415973,
+      "grad_norm": 0.43442684412002563,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5690728316223426,
+      "grad_norm": 0.6463358998298645,
+      "learning_rate": 0.0002,
+      "loss": 0.7564,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5706895158030879,
+      "grad_norm": 0.5286486744880676,
+      "learning_rate": 0.0002,
+      "loss": 0.836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5723061999838331,
+      "grad_norm": 0.5405499935150146,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5739228841645785,
+      "grad_norm": 0.6654391884803772,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.5081980228424072,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5771562525260691,
+      "grad_norm": 0.48978179693222046,
+      "learning_rate": 0.0002,
+      "loss": 0.7753,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5787729367068143,
+      "grad_norm": 0.5840612053871155,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5803896208875596,
+      "grad_norm": 0.5235261917114258,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5820063050683049,
+      "grad_norm": 0.5672075748443604,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5836229892490502,
+      "grad_norm": 0.5613429546356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5852396734297954,
+      "grad_norm": 0.4032273590564728,
+      "learning_rate": 0.0002,
+      "loss": 0.8274,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5868563576105408,
+      "grad_norm": 0.49559324979782104,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5884730417912861,
+      "grad_norm": 0.6895697712898254,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5900897259720314,
+      "grad_norm": 0.4750136435031891,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5917064101527767,
+      "grad_norm": 0.5176819562911987,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5933230943335219,
+      "grad_norm": 0.5817760229110718,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5949397785142673,
+      "grad_norm": 0.6064626574516296,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5965564626950125,
+      "grad_norm": 0.6728700995445251,
+      "learning_rate": 0.0002,
+      "loss": 0.8422,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5981731468757578,
+      "grad_norm": 0.609305202960968,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5997898310565031,
+      "grad_norm": 0.4615488350391388,
+      "learning_rate": 0.0002,
+      "loss": 0.8048,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6014065152372484,
+      "grad_norm": 2.0531179904937744,
+      "learning_rate": 0.0002,
+      "loss": 0.8214,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6030231994179936,
+      "grad_norm": 0.5091132521629333,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 3730
+    },
+    {
+      "epoch": 0.604639883598739,
+      "grad_norm": 0.5951124429702759,
+      "learning_rate": 0.0002,
+      "loss": 0.7833,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6062565677794842,
+      "grad_norm": 0.5870208144187927,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6078732519602296,
+      "grad_norm": 0.6254619359970093,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6094899361409749,
+      "grad_norm": 0.5577626824378967,
+      "learning_rate": 0.0002,
+      "loss": 0.7868,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6111066203217201,
+      "grad_norm": 0.5004405379295349,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6127233045024655,
+      "grad_norm": 0.5527383685112,
+      "learning_rate": 0.0002,
+      "loss": 0.8092,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6143399886832107,
+      "grad_norm": 0.49116113781929016,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6159566728639561,
+      "grad_norm": 0.5299299359321594,
+      "learning_rate": 0.0002,
+      "loss": 0.8352,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6175733570447013,
+      "grad_norm": 0.464897483587265,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6191900412254466,
+      "grad_norm": 0.6505740880966187,
+      "learning_rate": 0.0002,
+      "loss": 0.7923,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6208067254061919,
+      "grad_norm": 0.5512559413909912,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6224234095869372,
+      "grad_norm": 0.49427518248558044,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6240400937676824,
+      "grad_norm": 0.3839147090911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6256567779484278,
+      "grad_norm": 0.5760218501091003,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6272734621291731,
+      "grad_norm": 0.7226507067680359,
+      "learning_rate": 0.0002,
+      "loss": 0.7836,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6288901463099184,
+      "grad_norm": 0.676781415939331,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6305068304906637,
+      "grad_norm": 0.4284018278121948,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6321235146714089,
+      "grad_norm": 0.5060628056526184,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6337401988521543,
+      "grad_norm": 0.5524522066116333,
+      "learning_rate": 0.0002,
+      "loss": 0.8089,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6353568830328995,
+      "grad_norm": 0.6099881529808044,
+      "learning_rate": 0.0002,
+      "loss": 0.8276,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6369735672136448,
+      "grad_norm": 0.43155938386917114,
+      "learning_rate": 0.0002,
+      "loss": 0.809,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6385902513943901,
+      "grad_norm": 0.6427084803581238,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6402069355751354,
+      "grad_norm": 0.541220486164093,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6418236197558806,
+      "grad_norm": 0.5414294600486755,
+      "learning_rate": 0.0002,
+      "loss": 0.8539,
+      "step": 3970
+    },
+    {
+      "epoch": 0.643440303936626,
+      "grad_norm": 0.46344003081321716,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6450569881173712,
+      "grad_norm": 0.45209285616874695,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6466736722981166,
+      "grad_norm": 0.5417284369468689,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6482903564788619,
+      "grad_norm": 0.7995685935020447,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6499070406596071,
+      "grad_norm": 0.6384002566337585,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6515237248403525,
+      "grad_norm": 0.4472815692424774,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6531404090210977,
+      "grad_norm": 0.6834294199943542,
+      "learning_rate": 0.0002,
+      "loss": 0.8071,
+      "step": 4040
+    },
+    {
+      "epoch": 0.654757093201843,
+      "grad_norm": 0.4612339735031128,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6563737773825883,
+      "grad_norm": 0.9266576170921326,
+      "learning_rate": 0.0002,
+      "loss": 0.8141,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6579904615633336,
+      "grad_norm": 0.4470861852169037,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6596071457440789,
+      "grad_norm": 0.45544925332069397,
+      "learning_rate": 0.0002,
+      "loss": 0.8293,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6612238299248242,
+      "grad_norm": 0.6144481301307678,
+      "learning_rate": 0.0002,
+      "loss": 0.8455,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6628405141055694,
+      "grad_norm": 0.5936288237571716,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6644571982863148,
+      "grad_norm": 0.4822963774204254,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 4110
+    },
+    {
+      "epoch": 0.66607388246706,
+      "grad_norm": 0.48432496190071106,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6676905666478054,
+      "grad_norm": 0.4901607930660248,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6693072508285507,
+      "grad_norm": 0.5018393397331238,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6709239350092959,
+      "grad_norm": 0.6946378946304321,
+      "learning_rate": 0.0002,
+      "loss": 0.8065,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6725406191900413,
+      "grad_norm": 0.5997390747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.6738849878311157,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6757739875515318,
+      "grad_norm": 0.6110581159591675,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6773906717322771,
+      "grad_norm": 0.5703322291374207,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6790073559130224,
+      "grad_norm": 0.4686066210269928,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6806240400937676,
+      "grad_norm": 0.6394643783569336,
+      "learning_rate": 0.0002,
+      "loss": 0.8441,
+      "step": 4210
+    },
+    {
+      "epoch": 0.682240724274513,
+      "grad_norm": 0.5454841256141663,
+      "learning_rate": 0.0002,
+      "loss": 0.8011,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6838574084552582,
+      "grad_norm": 0.4859732985496521,
+      "learning_rate": 0.0002,
+      "loss": 0.8307,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6854740926360036,
+      "grad_norm": 0.5544065833091736,
+      "learning_rate": 0.0002,
+      "loss": 0.8161,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6870907768167488,
+      "grad_norm": 0.4902505576610565,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6887074609974941,
+      "grad_norm": 0.4768051505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7977,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6903241451782395,
+      "grad_norm": 0.49982190132141113,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6919408293589847,
+      "grad_norm": 0.6351838111877441,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 4280
+    },
+    {
+      "epoch": 0.69355751353973,
+      "grad_norm": 0.5647561550140381,
+      "learning_rate": 0.0002,
+      "loss": 0.7664,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6951741977204753,
+      "grad_norm": 0.5340486764907837,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6967908819012206,
+      "grad_norm": 0.5649092793464661,
+      "learning_rate": 0.0002,
+      "loss": 0.8526,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6984075660819659,
+      "grad_norm": 0.6183916926383972,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7000242502627112,
+      "grad_norm": 0.6154509782791138,
+      "learning_rate": 0.0002,
+      "loss": 0.792,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7016409344434564,
+      "grad_norm": 0.5156264305114746,
+      "learning_rate": 0.0002,
+      "loss": 0.8397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7032576186242018,
+      "grad_norm": 0.562171459197998,
+      "learning_rate": 0.0002,
+      "loss": 0.8512,
+      "step": 4350
+    },
+    {
+      "epoch": 0.704874302804947,
+      "grad_norm": 0.4949502646923065,
+      "learning_rate": 0.0002,
+      "loss": 0.7882,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7064909869856923,
+      "grad_norm": 0.5171684622764587,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7081076711664377,
+      "grad_norm": 0.6198443174362183,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7097243553471829,
+      "grad_norm": 0.5802276134490967,
+      "learning_rate": 0.0002,
+      "loss": 0.7606,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7113410395279283,
+      "grad_norm": 0.41096967458724976,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7129577237086735,
+      "grad_norm": 0.4397392272949219,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7145744078894188,
+      "grad_norm": 0.45228442549705505,
+      "learning_rate": 0.0002,
+      "loss": 0.7651,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7161910920701641,
+      "grad_norm": 0.4839673936367035,
+      "learning_rate": 0.0002,
+      "loss": 0.7938,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7178077762509094,
+      "grad_norm": 0.6140755414962769,
+      "learning_rate": 0.0002,
+      "loss": 0.8362,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 0.6841378808021545,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7210411446124,
+      "grad_norm": 0.6664239168167114,
+      "learning_rate": 0.0002,
+      "loss": 0.8177,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7226578287931452,
+      "grad_norm": 0.47552719712257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7983,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7242745129738906,
+      "grad_norm": 0.6649776101112366,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7258911971546358,
+      "grad_norm": 0.5159541964530945,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7275078813353811,
+      "grad_norm": 0.6693112850189209,
+      "learning_rate": 0.0002,
+      "loss": 0.7786,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7291245655161265,
+      "grad_norm": 0.48870977759361267,
+      "learning_rate": 0.0002,
+      "loss": 0.8655,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7307412496968717,
+      "grad_norm": 0.4857887923717499,
+      "learning_rate": 0.0002,
+      "loss": 0.7337,
+      "step": 4520
+    },
+    {
+      "epoch": 0.732357933877617,
+      "grad_norm": 0.5515662431716919,
+      "learning_rate": 0.0002,
+      "loss": 0.8026,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7339746180583623,
+      "grad_norm": 0.6292222738265991,
+      "learning_rate": 0.0002,
+      "loss": 0.8031,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7355913022391076,
+      "grad_norm": 0.48265689611434937,
+      "learning_rate": 0.0002,
+      "loss": 0.7749,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7372079864198529,
+      "grad_norm": 0.8044266104698181,
+      "learning_rate": 0.0002,
+      "loss": 0.8499,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7388246706005982,
+      "grad_norm": 0.6111769676208496,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7404413547813434,
+      "grad_norm": 0.5229553580284119,
+      "learning_rate": 0.0002,
+      "loss": 0.7291,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7420580389620888,
+      "grad_norm": 0.6054152250289917,
+      "learning_rate": 0.0002,
+      "loss": 0.8038,
+      "step": 4590
+    },
+    {
+      "epoch": 0.743674723142834,
+      "grad_norm": 0.5574966669082642,
+      "learning_rate": 0.0002,
+      "loss": 0.8169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7452914073235793,
+      "grad_norm": 0.5395817160606384,
+      "learning_rate": 0.0002,
+      "loss": 0.8439,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7469080915043246,
+      "grad_norm": 0.7116472721099854,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7485247756850699,
+      "grad_norm": 0.5618700981140137,
+      "learning_rate": 0.0002,
+      "loss": 0.7743,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7501414598658153,
+      "grad_norm": 0.5802770853042603,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7517581440465605,
+      "grad_norm": 0.5690428018569946,
+      "learning_rate": 0.0002,
+      "loss": 0.7924,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7533748282273058,
+      "grad_norm": 0.4813360273838043,
+      "learning_rate": 0.0002,
+      "loss": 0.8017,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7549915124080511,
+      "grad_norm": 0.5434042811393738,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7566081965887964,
+      "grad_norm": 0.5502099990844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7582248807695416,
+      "grad_norm": 0.6020621061325073,
+      "learning_rate": 0.0002,
+      "loss": 0.8598,
+      "step": 4690
+    },
+    {
+      "epoch": 0.759841564950287,
+      "grad_norm": 0.4922301471233368,
+      "learning_rate": 0.0002,
+      "loss": 0.7937,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7614582491310322,
+      "grad_norm": 0.6492828726768494,
+      "learning_rate": 0.0002,
+      "loss": 0.788,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7630749333117776,
+      "grad_norm": 0.4865580201148987,
+      "learning_rate": 0.0002,
+      "loss": 0.8313,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7646916174925228,
+      "grad_norm": 0.5971422791481018,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7663083016732681,
+      "grad_norm": 0.6832674145698547,
+      "learning_rate": 0.0002,
+      "loss": 0.8298,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7679249858540134,
+      "grad_norm": 0.500908613204956,
+      "learning_rate": 0.0002,
+      "loss": 0.8156,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7695416700347587,
+      "grad_norm": 0.6112465858459473,
+      "learning_rate": 0.0002,
+      "loss": 0.8383,
+      "step": 4760
+    },
+    {
+      "epoch": 0.771158354215504,
+      "grad_norm": 0.5753506422042847,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7727750383962493,
+      "grad_norm": 0.6529405117034912,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7743917225769946,
+      "grad_norm": 0.5916843414306641,
+      "learning_rate": 0.0002,
+      "loss": 0.8171,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7760084067577399,
+      "grad_norm": 0.4821224510669708,
+      "learning_rate": 0.0002,
+      "loss": 0.83,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7776250909384852,
+      "grad_norm": 0.5532580018043518,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7792417751192304,
+      "grad_norm": 0.4604877233505249,
+      "learning_rate": 0.0002,
+      "loss": 0.7363,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7808584592999758,
+      "grad_norm": 0.5009613037109375,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 4830
+    },
+    {
+      "epoch": 0.782475143480721,
+      "grad_norm": 0.6448560357093811,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7840918276614663,
+      "grad_norm": 0.44327953457832336,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7857085118422116,
+      "grad_norm": 0.5355411171913147,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7873251960229569,
+      "grad_norm": 0.5635677576065063,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7889418802037023,
+      "grad_norm": 0.5417491793632507,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7905585643844475,
+      "grad_norm": 0.4567430913448334,
+      "learning_rate": 0.0002,
+      "loss": 0.7819,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7921752485651928,
+      "grad_norm": 0.44651296734809875,
+      "learning_rate": 0.0002,
+      "loss": 0.8454,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7937919327459381,
+      "grad_norm": 0.5741217136383057,
+      "learning_rate": 0.0002,
+      "loss": 0.7959,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7954086169266834,
+      "grad_norm": 0.6605045199394226,
+      "learning_rate": 0.0002,
+      "loss": 0.8093,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7970253011074286,
+      "grad_norm": 0.5126531720161438,
+      "learning_rate": 0.0002,
+      "loss": 0.77,
+      "step": 4930
+    },
+    {
+      "epoch": 0.798641985288174,
+      "grad_norm": 0.513648271560669,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8002586694689192,
+      "grad_norm": 0.5350404381752014,
+      "learning_rate": 0.0002,
+      "loss": 0.8314,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8018753536496646,
+      "grad_norm": 0.5731674432754517,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8034920378304098,
+      "grad_norm": 0.5974258184432983,
+      "learning_rate": 0.0002,
+      "loss": 0.8572,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8051087220111551,
+      "grad_norm": 0.8774799704551697,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8067254061919004,
+      "grad_norm": 0.5994430184364319,
+      "learning_rate": 0.0002,
+      "loss": 0.7899,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8083420903726457,
+      "grad_norm": 0.4894903004169464,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5000
+    },
+    {
+      "epoch": 0.809958774553391,
+      "grad_norm": 0.5218459367752075,
+      "learning_rate": 0.0002,
+      "loss": 0.78,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8115754587341363,
+      "grad_norm": 0.5232468843460083,
+      "learning_rate": 0.0002,
+      "loss": 0.817,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8131921429148816,
+      "grad_norm": 0.44358372688293457,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8148088270956269,
+      "grad_norm": 0.6202037334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8164255112763722,
+      "grad_norm": 0.7721474170684814,
+      "learning_rate": 0.0002,
+      "loss": 0.7351,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8180421954571174,
+      "grad_norm": 0.5568501353263855,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8196588796378628,
+      "grad_norm": 0.49148809909820557,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 5070
+    },
+    {
+      "epoch": 0.821275563818608,
+      "grad_norm": 0.4956012964248657,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8228922479993533,
+      "grad_norm": 0.6078833937644958,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8245089321800986,
+      "grad_norm": 0.46906954050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.828,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8261256163608439,
+      "grad_norm": 0.50812166929245,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8277423005415891,
+      "grad_norm": 0.5319661498069763,
+      "learning_rate": 0.0002,
+      "loss": 0.8243,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8293589847223345,
+      "grad_norm": 0.4949689209461212,
+      "learning_rate": 0.0002,
+      "loss": 0.7798,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8309756689030798,
+      "grad_norm": 0.5151591300964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8325923530838251,
+      "grad_norm": 0.5530214309692383,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8342090372645704,
+      "grad_norm": 0.6297410130500793,
+      "learning_rate": 0.0002,
+      "loss": 0.8251,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8358257214453156,
+      "grad_norm": 0.5466840267181396,
+      "learning_rate": 0.0002,
+      "loss": 0.8067,
+      "step": 5170
+    },
+    {
+      "epoch": 0.837442405626061,
+      "grad_norm": 0.652913510799408,
+      "learning_rate": 0.0002,
+      "loss": 0.7875,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8390590898068062,
+      "grad_norm": 0.5811293125152588,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8406757739875516,
+      "grad_norm": 0.5109550952911377,
+      "learning_rate": 0.0002,
+      "loss": 0.7412,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8422924581682968,
+      "grad_norm": 0.4551706612110138,
+      "learning_rate": 0.0002,
+      "loss": 0.8077,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8439091423490421,
+      "grad_norm": 0.5813754200935364,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8455258265297874,
+      "grad_norm": 0.5856947898864746,
+      "learning_rate": 0.0002,
+      "loss": 0.802,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8471425107105327,
+      "grad_norm": 0.5482739210128784,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8487591948912779,
+      "grad_norm": 0.49023720622062683,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8503758790720233,
+      "grad_norm": 0.49472475051879883,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8519925632527686,
+      "grad_norm": 0.5490226745605469,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8536092474335139,
+      "grad_norm": 0.5340665578842163,
+      "learning_rate": 0.0002,
+      "loss": 0.8333,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8552259316142592,
+      "grad_norm": 0.5962483882904053,
+      "learning_rate": 0.0002,
+      "loss": 0.8277,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8568426157950044,
+      "grad_norm": 0.586358368396759,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8584592999757498,
+      "grad_norm": 0.49120277166366577,
+      "learning_rate": 0.0002,
+      "loss": 0.7831,
+      "step": 5310
+    },
+    {
+      "epoch": 0.860075984156495,
+      "grad_norm": 0.5887332558631897,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8616926683372403,
+      "grad_norm": 0.42496153712272644,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.5489874482154846,
+      "learning_rate": 0.0002,
+      "loss": 0.7905,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8649260366987309,
+      "grad_norm": 0.5850813984870911,
+      "learning_rate": 0.0002,
+      "loss": 0.7958,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8665427208794761,
+      "grad_norm": 0.517487108707428,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8681594050602215,
+      "grad_norm": 0.5339142680168152,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8697760892409668,
+      "grad_norm": 0.6236387491226196,
+      "learning_rate": 0.0002,
+      "loss": 0.818,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8713927734217121,
+      "grad_norm": 0.5752192735671997,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8730094576024574,
+      "grad_norm": 0.6724614500999451,
+      "learning_rate": 0.0002,
+      "loss": 0.8542,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8746261417832026,
+      "grad_norm": 0.5280613303184509,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 5410
+    },
+    {
+      "epoch": 0.876242825963948,
+      "grad_norm": 0.44033288955688477,
+      "learning_rate": 0.0002,
+      "loss": 0.8231,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8778595101446932,
+      "grad_norm": 0.5199708342552185,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8794761943254386,
+      "grad_norm": 0.46778348088264465,
+      "learning_rate": 0.0002,
+      "loss": 0.7852,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8810928785061838,
+      "grad_norm": 0.4657754898071289,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8827095626869291,
+      "grad_norm": 0.5472902655601501,
+      "learning_rate": 0.0002,
+      "loss": 0.7799,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8843262468676744,
+      "grad_norm": 0.4876766800880432,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8859429310484197,
+      "grad_norm": 0.5057248473167419,
+      "learning_rate": 0.0002,
+      "loss": 0.7906,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8875596152291649,
+      "grad_norm": 0.4637320637702942,
+      "learning_rate": 0.0002,
+      "loss": 0.8124,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8891762994099103,
+      "grad_norm": 0.471955806016922,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8907929835906556,
+      "grad_norm": 0.5209813714027405,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8924096677714008,
+      "grad_norm": 0.6213834285736084,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8940263519521462,
+      "grad_norm": 0.5215408205986023,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8956430361328914,
+      "grad_norm": 0.580478310585022,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8972597203136368,
+      "grad_norm": 0.49102169275283813,
+      "learning_rate": 0.0002,
+      "loss": 0.8371,
+      "step": 5550
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.6043479442596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7806,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9004930886751273,
+      "grad_norm": 0.5636463165283203,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9021097728558726,
+      "grad_norm": 0.5620124340057373,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9037264570366179,
+      "grad_norm": 0.5206354856491089,
+      "learning_rate": 0.0002,
+      "loss": 0.8083,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9053431412173631,
+      "grad_norm": 0.5798229575157166,
+      "learning_rate": 0.0002,
+      "loss": 0.8557,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9069598253981085,
+      "grad_norm": 0.6428212523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.8097,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9085765095788537,
+      "grad_norm": 0.48064687848091125,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9101931937595991,
+      "grad_norm": 0.6347860097885132,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9118098779403444,
+      "grad_norm": 0.5353913307189941,
+      "learning_rate": 0.0002,
+      "loss": 0.851,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9134265621210896,
+      "grad_norm": 0.5323944091796875,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5650
+    },
+    {
+      "epoch": 0.915043246301835,
+      "grad_norm": 0.5261843204498291,
+      "learning_rate": 0.0002,
+      "loss": 0.8393,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9166599304825802,
+      "grad_norm": 0.5451326966285706,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9182766146633256,
+      "grad_norm": 0.5183324217796326,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9198932988440708,
+      "grad_norm": 0.47229018807411194,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9215099830248161,
+      "grad_norm": 0.49180513620376587,
+      "learning_rate": 0.0002,
+      "loss": 0.7757,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9231266672055614,
+      "grad_norm": 0.5419785380363464,
+      "learning_rate": 0.0002,
+      "loss": 0.8735,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9247433513863067,
+      "grad_norm": 0.5408698916435242,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9263600355670519,
+      "grad_norm": 0.5286232829093933,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9279767197477973,
+      "grad_norm": 0.7539758086204529,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9295934039285425,
+      "grad_norm": 0.5166944861412048,
+      "learning_rate": 0.0002,
+      "loss": 0.8118,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9312100881092878,
+      "grad_norm": 0.6601425409317017,
+      "learning_rate": 0.0002,
+      "loss": 0.783,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9328267722900332,
+      "grad_norm": 0.5029960870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9344434564707784,
+      "grad_norm": 0.4926645755767822,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9360601406515238,
+      "grad_norm": 0.5739615559577942,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5790
+    },
+    {
+      "epoch": 0.937676824832269,
+      "grad_norm": 0.5058279037475586,
+      "learning_rate": 0.0002,
+      "loss": 0.8037,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9392935090130143,
+      "grad_norm": 0.5260962247848511,
+      "learning_rate": 0.0002,
+      "loss": 0.8537,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9409101931937596,
+      "grad_norm": 0.5768588185310364,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9425268773745049,
+      "grad_norm": 0.5170126557350159,
+      "learning_rate": 0.0002,
+      "loss": 0.8215,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9441435615552501,
+      "grad_norm": 0.5745864510536194,
+      "learning_rate": 0.0002,
+      "loss": 0.7422,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9457602457359955,
+      "grad_norm": 0.5551357865333557,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9473769299167407,
+      "grad_norm": 0.5776078701019287,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9489936140974861,
+      "grad_norm": 0.5340062379837036,
+      "learning_rate": 0.0002,
+      "loss": 0.8527,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9506102982782314,
+      "grad_norm": 0.6447290182113647,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9522269824589766,
+      "grad_norm": 0.5123815536499023,
+      "learning_rate": 0.0002,
+      "loss": 0.7945,
+      "step": 5890
+    },
+    {
+      "epoch": 0.953843666639722,
+      "grad_norm": 0.48547613620758057,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9554603508204672,
+      "grad_norm": 0.5791414976119995,
+      "learning_rate": 0.0002,
+      "loss": 0.7896,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9570770350012126,
+      "grad_norm": 0.6195011734962463,
+      "learning_rate": 0.0002,
+      "loss": 0.8408,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9586937191819578,
+      "grad_norm": 0.6323803067207336,
+      "learning_rate": 0.0002,
+      "loss": 0.7805,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9603104033627031,
+      "grad_norm": 0.45552879571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9619270875434484,
+      "grad_norm": 0.5796473622322083,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9635437717241937,
+      "grad_norm": 0.647261381149292,
+      "learning_rate": 0.0002,
+      "loss": 0.7672,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9651604559049389,
+      "grad_norm": 0.5487682819366455,
+      "learning_rate": 0.0002,
+      "loss": 0.8086,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9667771400856843,
+      "grad_norm": 0.5743663907051086,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9683938242664295,
+      "grad_norm": 0.5470591187477112,
+      "learning_rate": 0.0002,
+      "loss": 0.8153,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9700105084471748,
+      "grad_norm": 0.5901660323143005,
+      "learning_rate": 0.0002,
+      "loss": 0.8119,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9716271926279202,
+      "grad_norm": 0.6544759273529053,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9732438768086654,
+      "grad_norm": 0.6288470029830933,
+      "learning_rate": 0.0002,
+      "loss": 0.7536,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9748605609894108,
+      "grad_norm": 0.673153817653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 6030
+    },
+    {
+      "epoch": 0.976477245170156,
+      "grad_norm": 0.42854753136634827,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9780939293509013,
+      "grad_norm": 0.5227066278457642,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9797106135316466,
+      "grad_norm": 0.5372416973114014,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9813272977123919,
+      "grad_norm": 0.6026402115821838,
+      "learning_rate": 0.0002,
+      "loss": 0.7591,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9829439818931371,
+      "grad_norm": 0.49547791481018066,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9845606660738825,
+      "grad_norm": 0.4641951322555542,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9861773502546277,
+      "grad_norm": 0.5818535089492798,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9877940344353731,
+      "grad_norm": 0.63955157995224,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9894107186161183,
+      "grad_norm": 0.5649438500404358,
+      "learning_rate": 0.0002,
+      "loss": 0.7547,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9910274027968636,
+      "grad_norm": 0.5290433168411255,
+      "learning_rate": 0.0002,
+      "loss": 0.7861,
+      "step": 6130
+    },
+    {
+      "epoch": 0.992644086977609,
+      "grad_norm": 0.6399374008178711,
+      "learning_rate": 0.0002,
+      "loss": 0.8109,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9942607711583542,
+      "grad_norm": 0.6736576557159424,
+      "learning_rate": 0.0002,
+      "loss": 0.8373,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9958774553390995,
+      "grad_norm": 0.515420138835907,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9974941395198448,
+      "grad_norm": 0.562677800655365,
+      "learning_rate": 0.0002,
+      "loss": 0.8032,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9991108237005901,
+      "grad_norm": 0.7113858461380005,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9999191657909627,
+      "eval_loss": 1.0871200561523438,
+      "eval_runtime": 122.2071,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0007275078813354,
+      "grad_norm": 0.7111801505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0023441920620806,
+      "grad_norm": 0.5402125716209412,
+      "learning_rate": 0.0002,
+      "loss": 0.6865,
+      "step": 6200
+    },
+    {
+      "epoch": 1.003960876242826,
+      "grad_norm": 0.6098830103874207,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0055775604235713,
+      "grad_norm": 0.5829983353614807,
+      "learning_rate": 0.0002,
+      "loss": 0.7631,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0071942446043165,
+      "grad_norm": 0.5614621043205261,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0088109287850617,
+      "grad_norm": 0.5954238772392273,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0104276129658072,
+      "grad_norm": 0.6480574607849121,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0120442971465524,
+      "grad_norm": 0.6051128506660461,
+      "learning_rate": 0.0002,
+      "loss": 0.7514,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0136609813272976,
+      "grad_norm": 0.6318870782852173,
+      "learning_rate": 0.0002,
+      "loss": 0.7237,
+      "step": 6270
+    },
+    {
+      "epoch": 1.015277665508043,
+      "grad_norm": 0.5048980116844177,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0168943496887883,
+      "grad_norm": 0.6346936225891113,
+      "learning_rate": 0.0002,
+      "loss": 0.7391,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0185110338695336,
+      "grad_norm": 0.5711665749549866,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0201277180502788,
+      "grad_norm": 0.5175361037254333,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0217444022310243,
+      "grad_norm": 0.5360831618309021,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0233610864117695,
+      "grad_norm": 0.614675760269165,
+      "learning_rate": 0.0002,
+      "loss": 0.7112,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0249777705925147,
+      "grad_norm": 0.5626118183135986,
+      "learning_rate": 0.0002,
+      "loss": 0.7748,
+      "step": 6340
+    },
+    {
+      "epoch": 1.02659445477326,
+      "grad_norm": 0.574897289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.7375,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0282111389540054,
+      "grad_norm": 0.7185447812080383,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0298278231347506,
+      "grad_norm": 0.6705799698829651,
+      "learning_rate": 0.0002,
+      "loss": 0.703,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0314445073154959,
+      "grad_norm": 0.6740428805351257,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0330611914962413,
+      "grad_norm": 0.663902759552002,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0346778756769865,
+      "grad_norm": 0.5029543042182922,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0362945598577318,
+      "grad_norm": 0.7813863158226013,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 6410
+    },
+    {
+      "epoch": 1.037911244038477,
+      "grad_norm": 0.5396282076835632,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0395279282192225,
+      "grad_norm": 0.5253293514251709,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0411446123999677,
+      "grad_norm": 0.7236770987510681,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 6440
+    },
+    {
+      "epoch": 1.042761296580713,
+      "grad_norm": 0.5670917630195618,
+      "learning_rate": 0.0002,
+      "loss": 0.7259,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0443779807614582,
+      "grad_norm": 0.6031978726387024,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0459946649422036,
+      "grad_norm": 0.5309213399887085,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0476113491229488,
+      "grad_norm": 0.7114651799201965,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 6480
+    },
+    {
+      "epoch": 1.049228033303694,
+      "grad_norm": 0.5591610670089722,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0508447174844395,
+      "grad_norm": 0.5185961127281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0524614016651848,
+      "grad_norm": 0.6510552167892456,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6510
+    },
+    {
+      "epoch": 1.05407808584593,
+      "grad_norm": 0.6557928919792175,
+      "learning_rate": 0.0002,
+      "loss": 0.7057,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0556947700266752,
+      "grad_norm": 0.6973192691802979,
+      "learning_rate": 0.0002,
+      "loss": 0.8056,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0573114542074207,
+      "grad_norm": 0.6226583123207092,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 6540
+    },
+    {
+      "epoch": 1.058928138388166,
+      "grad_norm": 0.5633195638656616,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0605448225689111,
+      "grad_norm": 0.7466658353805542,
+      "learning_rate": 0.0002,
+      "loss": 0.7082,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0621615067496564,
+      "grad_norm": 0.6462772488594055,
+      "learning_rate": 0.0002,
+      "loss": 0.7059,
+      "step": 6570
+    },
+    {
+      "epoch": 1.0637781909304018,
+      "grad_norm": 0.5266856551170349,
+      "learning_rate": 0.0002,
+      "loss": 0.7046,
+      "step": 6580
+    },
+    {
+      "epoch": 1.065394875111147,
+      "grad_norm": 0.534392774105072,
+      "learning_rate": 0.0002,
+      "loss": 0.7157,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0670115592918923,
+      "grad_norm": 0.7514177560806274,
+      "learning_rate": 0.0002,
+      "loss": 0.7115,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0686282434726375,
+      "grad_norm": 0.7593035697937012,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 6610
+    },
+    {
+      "epoch": 1.070244927653383,
+      "grad_norm": 0.5277858972549438,
+      "learning_rate": 0.0002,
+      "loss": 0.6836,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0718616118341282,
+      "grad_norm": 0.5573670268058777,
+      "learning_rate": 0.0002,
+      "loss": 0.7405,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0734782960148734,
+      "grad_norm": 0.6802396774291992,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 6640
+    },
+    {
+      "epoch": 1.0750949801956189,
+      "grad_norm": 0.7367215752601624,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0767116643763641,
+      "grad_norm": 0.5961891412734985,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0783283485571094,
+      "grad_norm": 0.5736313462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0799450327378546,
+      "grad_norm": 0.619219183921814,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0815617169186,
+      "grad_norm": 0.6214390993118286,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0831784010993453,
+      "grad_norm": 0.564536988735199,
+      "learning_rate": 0.0002,
+      "loss": 0.7554,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0847950852800905,
+      "grad_norm": 0.5838140249252319,
+      "learning_rate": 0.0002,
+      "loss": 0.7362,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0864117694608357,
+      "grad_norm": 0.7000553607940674,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0880284536415812,
+      "grad_norm": 0.7078263759613037,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 6730
+    },
+    {
+      "epoch": 1.0896451378223264,
+      "grad_norm": 0.8353848457336426,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6740
+    },
+    {
+      "epoch": 1.0912618220030716,
+      "grad_norm": 0.5615518689155579,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 6750
+    },
+    {
+      "epoch": 1.0928785061838169,
+      "grad_norm": 0.5475581288337708,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 6760
+    },
+    {
+      "epoch": 1.0944951903645623,
+      "grad_norm": 0.5835978388786316,
+      "learning_rate": 0.0002,
+      "loss": 0.7652,
+      "step": 6770
+    },
+    {
+      "epoch": 1.0961118745453076,
+      "grad_norm": 0.5516105890274048,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 6780
+    },
+    {
+      "epoch": 1.0977285587260528,
+      "grad_norm": 0.5875251889228821,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 6790
+    },
+    {
+      "epoch": 1.0993452429067982,
+      "grad_norm": 0.7376947999000549,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1009619270875435,
+      "grad_norm": 0.5656165480613708,
+      "learning_rate": 0.0002,
+      "loss": 0.7512,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1025786112682887,
+      "grad_norm": 0.6365954279899597,
+      "learning_rate": 0.0002,
+      "loss": 0.7409,
+      "step": 6820
+    },
+    {
+      "epoch": 1.104195295449034,
+      "grad_norm": 0.5033080577850342,
+      "learning_rate": 0.0002,
+      "loss": 0.7392,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1058119796297794,
+      "grad_norm": 0.617396891117096,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1074286638105246,
+      "grad_norm": 0.6395374536514282,
+      "learning_rate": 0.0002,
+      "loss": 0.7006,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1090453479912699,
+      "grad_norm": 0.6775295734405518,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1106620321720153,
+      "grad_norm": 0.6655223965644836,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1122787163527605,
+      "grad_norm": 0.676655113697052,
+      "learning_rate": 0.0002,
+      "loss": 0.7553,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1138954005335058,
+      "grad_norm": 0.6062718629837036,
+      "learning_rate": 0.0002,
+      "loss": 0.7342,
+      "step": 6890
+    },
+    {
+      "epoch": 1.115512084714251,
+      "grad_norm": 0.590943455696106,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1171287688949965,
+      "grad_norm": 0.6315317153930664,
+      "learning_rate": 0.0002,
+      "loss": 0.6705,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1187454530757417,
+      "grad_norm": 0.47979024052619934,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 6920
+    },
+    {
+      "epoch": 1.120362137256487,
+      "grad_norm": 0.647298276424408,
+      "learning_rate": 0.0002,
+      "loss": 0.7002,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1219788214372322,
+      "grad_norm": 0.7336484789848328,
+      "learning_rate": 0.0002,
+      "loss": 0.7502,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1235955056179776,
+      "grad_norm": 0.5071424245834351,
+      "learning_rate": 0.0002,
+      "loss": 0.693,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1252121897987228,
+      "grad_norm": 0.6527144312858582,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 6960
+    },
+    {
+      "epoch": 1.126828873979468,
+      "grad_norm": 0.6935935020446777,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1284455581602133,
+      "grad_norm": 0.8026931881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1300622423409588,
+      "grad_norm": 0.5210393667221069,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 6990
+    },
+    {
+      "epoch": 1.131678926521704,
+      "grad_norm": 0.60475093126297,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1332956107024492,
+      "grad_norm": 0.6417073607444763,
+      "learning_rate": 0.0002,
+      "loss": 0.7495,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1349122948831947,
+      "grad_norm": 0.6732175946235657,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 7020
+    },
+    {
+      "epoch": 1.13652897906394,
+      "grad_norm": 0.6719491481781006,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1381456632446851,
+      "grad_norm": 0.5708295106887817,
+      "learning_rate": 0.0002,
+      "loss": 0.7694,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1397623474254304,
+      "grad_norm": 0.7141719460487366,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1413790316061758,
+      "grad_norm": 0.6187017560005188,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 7060
+    },
+    {
+      "epoch": 1.142995715786921,
+      "grad_norm": 0.50581294298172,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1446123999676663,
+      "grad_norm": 0.5620143413543701,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1462290841484115,
+      "grad_norm": 0.6231929659843445,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 7090
+    },
+    {
+      "epoch": 1.147845768329157,
+      "grad_norm": 0.5775774121284485,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1494624525099022,
+      "grad_norm": 0.6492809653282166,
+      "learning_rate": 0.0002,
+      "loss": 0.7728,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1510791366906474,
+      "grad_norm": 0.6434972286224365,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 7120
+    },
+    {
+      "epoch": 1.1526958208713927,
+      "grad_norm": 0.6191812753677368,
+      "learning_rate": 0.0002,
+      "loss": 0.7374,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1543125050521381,
+      "grad_norm": 0.6690331697463989,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 7140
+    },
+    {
+      "epoch": 1.1559291892328833,
+      "grad_norm": 0.5977938175201416,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1575458734136286,
+      "grad_norm": 0.6195854544639587,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7160
+    },
+    {
+      "epoch": 1.159162557594374,
+      "grad_norm": 0.5752048492431641,
+      "learning_rate": 0.0002,
+      "loss": 0.7249,
+      "step": 7170
+    },
+    {
+      "epoch": 1.1607792417751193,
+      "grad_norm": 0.589081883430481,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1623959259558645,
+      "grad_norm": 0.756996750831604,
+      "learning_rate": 0.0002,
+      "loss": 0.704,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1640126101366097,
+      "grad_norm": 0.7614967226982117,
+      "learning_rate": 0.0002,
+      "loss": 0.7404,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1656292943173552,
+      "grad_norm": 0.6120437979698181,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1672459784981004,
+      "grad_norm": 0.6210004687309265,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1688626626788456,
+      "grad_norm": 0.6044116020202637,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7230
+    },
+    {
+      "epoch": 1.170479346859591,
+      "grad_norm": 0.5418457388877869,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1720960310403363,
+      "grad_norm": 0.6413537263870239,
+      "learning_rate": 0.0002,
+      "loss": 0.6938,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1737127152210816,
+      "grad_norm": 0.5777867436408997,
+      "learning_rate": 0.0002,
+      "loss": 0.6978,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1753293994018268,
+      "grad_norm": 0.7092402577400208,
+      "learning_rate": 0.0002,
+      "loss": 0.7503,
+      "step": 7270
+    },
+    {
+      "epoch": 1.176946083582572,
+      "grad_norm": 0.6351709365844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1785627677633175,
+      "grad_norm": 0.6172189712524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7527,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1801794519440627,
+      "grad_norm": 0.6801714897155762,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7300
+    },
+    {
+      "epoch": 1.181796136124808,
+      "grad_norm": 0.6044712066650391,
+      "learning_rate": 0.0002,
+      "loss": 0.6941,
+      "step": 7310
+    },
+    {
+      "epoch": 1.1834128203055534,
+      "grad_norm": 0.7413212060928345,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 7320
+    },
+    {
+      "epoch": 1.1850295044862986,
+      "grad_norm": 0.5303856134414673,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 7330
+    },
+    {
+      "epoch": 1.1866461886670439,
+      "grad_norm": 0.5647098422050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6915,
+      "step": 7340
+    },
+    {
+      "epoch": 1.188262872847789,
+      "grad_norm": 0.7374135255813599,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 7350
+    },
+    {
+      "epoch": 1.1898795570285345,
+      "grad_norm": 0.5710089206695557,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1914962412092798,
+      "grad_norm": 0.6073619723320007,
+      "learning_rate": 0.0002,
+      "loss": 0.8289,
+      "step": 7370
+    },
+    {
+      "epoch": 1.193112925390025,
+      "grad_norm": 0.5899916887283325,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1947296095707705,
+      "grad_norm": 0.7762434482574463,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1963462937515157,
+      "grad_norm": 0.679949939250946,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7400
+    },
+    {
+      "epoch": 1.197962977932261,
+      "grad_norm": 0.6106849312782288,
+      "learning_rate": 0.0002,
+      "loss": 0.7599,
+      "step": 7410
+    },
+    {
+      "epoch": 1.1995796621130062,
+      "grad_norm": 0.682461678981781,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2011963462937516,
+      "grad_norm": 0.6087017059326172,
+      "learning_rate": 0.0002,
+      "loss": 0.7741,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2028130304744968,
+      "grad_norm": 0.63739013671875,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 7440
+    },
+    {
+      "epoch": 1.204429714655242,
+      "grad_norm": 0.6154777407646179,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2060463988359873,
+      "grad_norm": 0.7491534948348999,
+      "learning_rate": 0.0002,
+      "loss": 0.7565,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2076630830167328,
+      "grad_norm": 0.6664797067642212,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 7470
+    },
+    {
+      "epoch": 1.209279767197478,
+      "grad_norm": 0.6660266518592834,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2108964513782232,
+      "grad_norm": 0.6972551345825195,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2125131355589684,
+      "grad_norm": 0.6157945990562439,
+      "learning_rate": 0.0002,
+      "loss": 0.7023,
+      "step": 7500
+    },
+    {
+      "epoch": 1.214129819739714,
+      "grad_norm": 0.5199310183525085,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2157465039204591,
+      "grad_norm": 0.577610433101654,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2173631881012044,
+      "grad_norm": 0.53652423620224,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2189798722819498,
+      "grad_norm": 0.6479050517082214,
+      "learning_rate": 0.0002,
+      "loss": 0.7393,
+      "step": 7540
+    },
+    {
+      "epoch": 1.220596556462695,
+      "grad_norm": 0.618748128414154,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2222132406434403,
+      "grad_norm": 0.6311424374580383,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2238299248241855,
+      "grad_norm": 0.6595825552940369,
+      "learning_rate": 0.0002,
+      "loss": 0.7272,
+      "step": 7570
+    },
+    {
+      "epoch": 1.225446609004931,
+      "grad_norm": 0.5198960900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2270632931856762,
+      "grad_norm": 0.578650712966919,
+      "learning_rate": 0.0002,
+      "loss": 0.674,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2286799773664214,
+      "grad_norm": 0.6080220937728882,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2302966615471669,
+      "grad_norm": 0.7050248384475708,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2319133457279121,
+      "grad_norm": 0.6652196049690247,
+      "learning_rate": 0.0002,
+      "loss": 0.7032,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2335300299086573,
+      "grad_norm": 0.7322776317596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7085,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2351467140894026,
+      "grad_norm": 0.4998728036880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2367633982701478,
+      "grad_norm": 0.6428788900375366,
+      "learning_rate": 0.0002,
+      "loss": 0.7214,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2383800824508933,
+      "grad_norm": 0.585242509841919,
+      "learning_rate": 0.0002,
+      "loss": 0.7699,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2399967666316385,
+      "grad_norm": 0.5211917757987976,
+      "learning_rate": 0.0002,
+      "loss": 0.7621,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2416134508123837,
+      "grad_norm": 0.6490384340286255,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2432301349931292,
+      "grad_norm": 0.6249763369560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2448468191738744,
+      "grad_norm": 0.71870356798172,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2464635033546196,
+      "grad_norm": 0.6761967539787292,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2480801875353649,
+      "grad_norm": 0.6500617265701294,
+      "learning_rate": 0.0002,
+      "loss": 0.7501,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2496968717161103,
+      "grad_norm": 0.8069869875907898,
+      "learning_rate": 0.0002,
+      "loss": 0.7903,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2513135558968556,
+      "grad_norm": 0.6044608950614929,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2529302400776008,
+      "grad_norm": 0.6573283076286316,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2545469242583462,
+      "grad_norm": 0.625430166721344,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2561636084390915,
+      "grad_norm": 0.5442022681236267,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7770
+    },
+    {
+      "epoch": 1.2577802926198367,
+      "grad_norm": 0.6818386912345886,
+      "learning_rate": 0.0002,
+      "loss": 0.7172,
+      "step": 7780
+    },
+    {
+      "epoch": 1.259396976800582,
+      "grad_norm": 0.6381874084472656,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2610136609813272,
+      "grad_norm": 0.6269212961196899,
+      "learning_rate": 0.0002,
+      "loss": 0.6834,
+      "step": 7800
+    },
+    {
+      "epoch": 1.2626303451620726,
+      "grad_norm": 0.600121259689331,
+      "learning_rate": 0.0002,
+      "loss": 0.7821,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2642470293428179,
+      "grad_norm": 0.6337703466415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2658637135235633,
+      "grad_norm": 0.7234963774681091,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 7830
+    },
+    {
+      "epoch": 1.2674803977043085,
+      "grad_norm": 0.800184965133667,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2690970818850538,
+      "grad_norm": 0.7539464831352234,
+      "learning_rate": 0.0002,
+      "loss": 0.7426,
+      "step": 7850
+    },
+    {
+      "epoch": 1.270713766065799,
+      "grad_norm": 0.5493760704994202,
+      "learning_rate": 0.0002,
+      "loss": 0.7496,
+      "step": 7860
+    },
+    {
+      "epoch": 1.2723304502465442,
+      "grad_norm": 0.7477145791053772,
+      "learning_rate": 0.0002,
+      "loss": 0.7537,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2739471344272897,
+      "grad_norm": 0.6366362571716309,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 7880
+    },
+    {
+      "epoch": 1.275563818608035,
+      "grad_norm": 0.7419533729553223,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2771805027887801,
+      "grad_norm": 0.6141223311424255,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 7900
+    },
+    {
+      "epoch": 1.2787971869695256,
+      "grad_norm": 0.7522598505020142,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 7910
+    },
+    {
+      "epoch": 1.2804138711502708,
+      "grad_norm": 0.6935804486274719,
+      "learning_rate": 0.0002,
+      "loss": 0.7097,
+      "step": 7920
+    },
+    {
+      "epoch": 1.282030555331016,
+      "grad_norm": 0.7239290475845337,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 7930
+    },
+    {
+      "epoch": 1.2836472395117613,
+      "grad_norm": 0.8800187110900879,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 7940
+    },
+    {
+      "epoch": 1.2852639236925067,
+      "grad_norm": 0.540458083152771,
+      "learning_rate": 0.0002,
+      "loss": 0.6991,
+      "step": 7950
+    },
+    {
+      "epoch": 1.286880607873252,
+      "grad_norm": 0.6492934226989746,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 7960
+    },
+    {
+      "epoch": 1.2884972920539972,
+      "grad_norm": 0.6543959379196167,
+      "learning_rate": 0.0002,
+      "loss": 0.7742,
+      "step": 7970
+    },
+    {
+      "epoch": 1.2901139762347427,
+      "grad_norm": 0.5804705619812012,
+      "learning_rate": 0.0002,
+      "loss": 0.7316,
+      "step": 7980
+    },
+    {
+      "epoch": 1.291730660415488,
+      "grad_norm": 0.7074727416038513,
+      "learning_rate": 0.0002,
+      "loss": 0.796,
+      "step": 7990
+    },
+    {
+      "epoch": 1.2933473445962331,
+      "grad_norm": 0.5347974300384521,
+      "learning_rate": 0.0002,
+      "loss": 0.7034,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2949640287769784,
+      "grad_norm": 0.6457298398017883,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 8010
+    },
+    {
+      "epoch": 1.2965807129577236,
+      "grad_norm": 0.6407219171524048,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8020
+    },
+    {
+      "epoch": 1.298197397138469,
+      "grad_norm": 0.828439474105835,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2998140813192143,
+      "grad_norm": 0.4840380549430847,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3014307654999595,
+      "grad_norm": 0.5921024680137634,
+      "learning_rate": 0.0002,
+      "loss": 0.7283,
+      "step": 8050
+    },
+    {
+      "epoch": 1.303047449680705,
+      "grad_norm": 0.6170315146446228,
+      "learning_rate": 0.0002,
+      "loss": 0.7477,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3046641338614502,
+      "grad_norm": 0.5374847054481506,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3062808180421954,
+      "grad_norm": 0.545758068561554,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3078975022229407,
+      "grad_norm": 0.55641770362854,
+      "learning_rate": 0.0002,
+      "loss": 0.7463,
+      "step": 8090
+    },
+    {
+      "epoch": 1.309514186403686,
+      "grad_norm": 0.6724897027015686,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3111308705844313,
+      "grad_norm": 0.6923972368240356,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3127475547651766,
+      "grad_norm": 0.5136841535568237,
+      "learning_rate": 0.0002,
+      "loss": 0.7149,
+      "step": 8120
+    },
+    {
+      "epoch": 1.314364238945922,
+      "grad_norm": 0.6766283512115479,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3159809231266673,
+      "grad_norm": 0.6283926367759705,
+      "learning_rate": 0.0002,
+      "loss": 0.7489,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3175976073074125,
+      "grad_norm": 0.644216001033783,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3192142914881577,
+      "grad_norm": 0.7827503085136414,
+      "learning_rate": 0.0002,
+      "loss": 0.7125,
+      "step": 8160
+    },
+    {
+      "epoch": 1.320830975668903,
+      "grad_norm": 0.6651390790939331,
+      "learning_rate": 0.0002,
+      "loss": 0.7271,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3224476598496484,
+      "grad_norm": 0.5547412633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7778,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3240643440303936,
+      "grad_norm": 0.6765179634094238,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 8190
+    },
+    {
+      "epoch": 1.325681028211139,
+      "grad_norm": 0.6822077035903931,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3272977123918843,
+      "grad_norm": 0.5941002368927002,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3289143965726296,
+      "grad_norm": 0.4850037097930908,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3305310807533748,
+      "grad_norm": 0.6162990927696228,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 8230
+    },
+    {
+      "epoch": 1.33214776493412,
+      "grad_norm": 0.6665613651275635,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3337644491148655,
+      "grad_norm": 0.618192732334137,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3353811332956107,
+      "grad_norm": 0.710418701171875,
+      "learning_rate": 0.0002,
+      "loss": 0.7532,
+      "step": 8260
+    },
+    {
+      "epoch": 1.336997817476356,
+      "grad_norm": 0.5109876990318298,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3386145016571014,
+      "grad_norm": 0.6791711449623108,
+      "learning_rate": 0.0002,
+      "loss": 0.7303,
+      "step": 8280
+    },
+    {
+      "epoch": 1.3402311858378466,
+      "grad_norm": 0.6836432814598083,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3418478700185918,
+      "grad_norm": 0.5579386353492737,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8300
+    },
+    {
+      "epoch": 1.343464554199337,
+      "grad_norm": 0.6713546514511108,
+      "learning_rate": 0.0002,
+      "loss": 0.7377,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3450812383800825,
+      "grad_norm": 0.5353720188140869,
+      "learning_rate": 0.0002,
+      "loss": 0.7756,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3466979225608278,
+      "grad_norm": 0.5813682675361633,
+      "learning_rate": 0.0002,
+      "loss": 0.718,
+      "step": 8330
+    },
+    {
+      "epoch": 1.348314606741573,
+      "grad_norm": 0.8158791661262512,
+      "learning_rate": 0.0002,
+      "loss": 0.7294,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3499312909223184,
+      "grad_norm": 0.6193785071372986,
+      "learning_rate": 0.0002,
+      "loss": 0.6992,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3515479751030637,
+      "grad_norm": 0.6353939771652222,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 8360
+    },
+    {
+      "epoch": 1.353164659283809,
+      "grad_norm": 0.6925048232078552,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 8370
+    },
+    {
+      "epoch": 1.3547813434645541,
+      "grad_norm": 0.988264799118042,
+      "learning_rate": 0.0002,
+      "loss": 0.736,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3563980276452994,
+      "grad_norm": 0.6476002931594849,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3580147118260448,
+      "grad_norm": 0.7120398879051208,
+      "learning_rate": 0.0002,
+      "loss": 0.776,
+      "step": 8400
+    },
+    {
+      "epoch": 1.35963139600679,
+      "grad_norm": 0.9048416614532471,
+      "learning_rate": 0.0002,
+      "loss": 0.7368,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3612480801875353,
+      "grad_norm": 0.7000672817230225,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3628647643682807,
+      "grad_norm": 0.6015632152557373,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 8430
+    },
+    {
+      "epoch": 1.364481448549026,
+      "grad_norm": 0.612516462802887,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3660981327297712,
+      "grad_norm": 0.5969301462173462,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 8450
+    },
+    {
+      "epoch": 1.3677148169105164,
+      "grad_norm": 0.6730654239654541,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 8460
+    },
+    {
+      "epoch": 1.369331501091262,
+      "grad_norm": 0.6386392116546631,
+      "learning_rate": 0.0002,
+      "loss": 0.7465,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3709481852720071,
+      "grad_norm": 0.739544153213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3725648694527524,
+      "grad_norm": 0.6462782621383667,
+      "learning_rate": 0.0002,
+      "loss": 0.7892,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3741815536334978,
+      "grad_norm": 0.7346843481063843,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 8500
+    },
+    {
+      "epoch": 1.375798237814243,
+      "grad_norm": 0.6884821057319641,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8510
+    },
+    {
+      "epoch": 1.3774149219949883,
+      "grad_norm": 0.6999333500862122,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3790316061757335,
+      "grad_norm": 0.5378713011741638,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 8530
+    },
+    {
+      "epoch": 1.3806482903564787,
+      "grad_norm": 0.5417906641960144,
+      "learning_rate": 0.0002,
+      "loss": 0.6797,
+      "step": 8540
+    },
+    {
+      "epoch": 1.3822649745372242,
+      "grad_norm": 0.6602526307106018,
+      "learning_rate": 0.0002,
+      "loss": 0.7499,
+      "step": 8550
+    },
+    {
+      "epoch": 1.3838816587179694,
+      "grad_norm": 0.7073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.7356,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3854983428987149,
+      "grad_norm": 0.5841707587242126,
+      "learning_rate": 0.0002,
+      "loss": 0.75,
+      "step": 8570
+    },
+    {
+      "epoch": 1.38711502707946,
+      "grad_norm": 0.7031095027923584,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 8580
+    },
+    {
+      "epoch": 1.3887317112602053,
+      "grad_norm": 0.5198570489883423,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 8590
+    },
+    {
+      "epoch": 1.3903483954409506,
+      "grad_norm": 0.7261320352554321,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 8600
+    },
+    {
+      "epoch": 1.3919650796216958,
+      "grad_norm": 0.5616350173950195,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3935817638024413,
+      "grad_norm": 0.5185914635658264,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 8620
+    },
+    {
+      "epoch": 1.3951984479831865,
+      "grad_norm": 0.5814694762229919,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 8630
+    },
+    {
+      "epoch": 1.3968151321639317,
+      "grad_norm": 0.6977371573448181,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 8640
+    },
+    {
+      "epoch": 1.3984318163446772,
+      "grad_norm": 0.6855689883232117,
+      "learning_rate": 0.0002,
+      "loss": 0.7574,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4000485005254224,
+      "grad_norm": 0.5414357781410217,
+      "learning_rate": 0.0002,
+      "loss": 0.7802,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4016651847061676,
+      "grad_norm": 0.6970012784004211,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4032818688869129,
+      "grad_norm": 0.526079535484314,
+      "learning_rate": 0.0002,
+      "loss": 0.7421,
+      "step": 8680
+    },
+    {
+      "epoch": 1.404898553067658,
+      "grad_norm": 0.758712887763977,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4065152372484035,
+      "grad_norm": 0.7118762731552124,
+      "learning_rate": 0.0002,
+      "loss": 0.7612,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4081319214291488,
+      "grad_norm": 0.5696909427642822,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4097486056098942,
+      "grad_norm": 0.7995436787605286,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4113652897906395,
+      "grad_norm": 0.7237521409988403,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4129819739713847,
+      "grad_norm": 0.744628369808197,
+      "learning_rate": 0.0002,
+      "loss": 0.7661,
+      "step": 8740
+    },
+    {
+      "epoch": 1.41459865815213,
+      "grad_norm": 0.6082926988601685,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4162153423328752,
+      "grad_norm": 0.5185243487358093,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4178320265136206,
+      "grad_norm": 0.5183082222938538,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4194487106943658,
+      "grad_norm": 0.7326041460037231,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 8780
+    },
+    {
+      "epoch": 1.421065394875111,
+      "grad_norm": 0.7174660563468933,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4226820790558565,
+      "grad_norm": 0.8080165982246399,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4242987632366018,
+      "grad_norm": 0.5061507821083069,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 8810
+    },
+    {
+      "epoch": 1.425915447417347,
+      "grad_norm": 0.801602840423584,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4275321315980922,
+      "grad_norm": 0.6150273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4291488157788377,
+      "grad_norm": 0.8786525726318359,
+      "learning_rate": 0.0002,
+      "loss": 0.7452,
+      "step": 8840
+    },
+    {
+      "epoch": 1.430765499959583,
+      "grad_norm": 0.6371538639068604,
+      "learning_rate": 0.0002,
+      "loss": 0.7257,
+      "step": 8850
+    },
+    {
+      "epoch": 1.4323821841403281,
+      "grad_norm": 0.6409295797348022,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4339988683210736,
+      "grad_norm": 0.6452359557151794,
+      "learning_rate": 0.0002,
+      "loss": 0.7891,
+      "step": 8870
+    },
+    {
+      "epoch": 1.4356155525018188,
+      "grad_norm": 0.5842334628105164,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 8880
+    },
+    {
+      "epoch": 1.437232236682564,
+      "grad_norm": 0.696761965751648,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4388489208633093,
+      "grad_norm": 0.6384600400924683,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4404656050440545,
+      "grad_norm": 0.5981136560440063,
+      "learning_rate": 0.0002,
+      "loss": 0.7049,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4420822892248,
+      "grad_norm": 0.6355637907981873,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4436989734055452,
+      "grad_norm": 0.6374830603599548,
+      "learning_rate": 0.0002,
+      "loss": 0.7653,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4453156575862904,
+      "grad_norm": 0.559013307094574,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 8940
+    },
+    {
+      "epoch": 1.446932341767036,
+      "grad_norm": 0.7289170026779175,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4485490259477811,
+      "grad_norm": 0.8649206757545471,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4501657101285264,
+      "grad_norm": 0.7664689421653748,
+      "learning_rate": 0.0002,
+      "loss": 0.7745,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4517823943092716,
+      "grad_norm": 0.7109952569007874,
+      "learning_rate": 0.0002,
+      "loss": 0.7431,
+      "step": 8980
+    },
+    {
+      "epoch": 1.453399078490017,
+      "grad_norm": 0.6312844753265381,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4550157626707623,
+      "grad_norm": 0.6616617441177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7467,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4566324468515075,
+      "grad_norm": 0.7384068965911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7518,
+      "step": 9010
+    },
+    {
+      "epoch": 1.458249131032253,
+      "grad_norm": 0.6549670100212097,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9020
+    },
+    {
+      "epoch": 1.4598658152129982,
+      "grad_norm": 0.6254119277000427,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4614824993937434,
+      "grad_norm": 0.6806328892707825,
+      "learning_rate": 0.0002,
+      "loss": 0.7645,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4630991835744886,
+      "grad_norm": 0.6803115010261536,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4647158677552339,
+      "grad_norm": 0.48529282212257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7264,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4663325519359793,
+      "grad_norm": 0.5995030999183655,
+      "learning_rate": 0.0002,
+      "loss": 0.7542,
+      "step": 9070
+    },
+    {
+      "epoch": 1.4679492361167246,
+      "grad_norm": 0.6005427837371826,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9080
+    },
+    {
+      "epoch": 1.46956592029747,
+      "grad_norm": 0.718564510345459,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4711826044782153,
+      "grad_norm": 0.7003577351570129,
+      "learning_rate": 0.0002,
+      "loss": 0.7089,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4727992886589605,
+      "grad_norm": 0.5888323783874512,
+      "learning_rate": 0.0002,
+      "loss": 0.8069,
+      "step": 9110
+    },
+    {
+      "epoch": 1.4744159728397057,
+      "grad_norm": 0.6417609453201294,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 9120
+    },
+    {
+      "epoch": 1.476032657020451,
+      "grad_norm": 0.572294294834137,
+      "learning_rate": 0.0002,
+      "loss": 0.7441,
+      "step": 9130
+    },
+    {
+      "epoch": 1.4776493412011964,
+      "grad_norm": 0.8200714588165283,
+      "learning_rate": 0.0002,
+      "loss": 0.8053,
+      "step": 9140
+    },
+    {
+      "epoch": 1.4792660253819416,
+      "grad_norm": 0.6343288421630859,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4808827095626869,
+      "grad_norm": 0.7017961144447327,
+      "learning_rate": 0.0002,
+      "loss": 0.7641,
+      "step": 9160
+    },
+    {
+      "epoch": 1.4824993937434323,
+      "grad_norm": 0.6202912926673889,
+      "learning_rate": 0.0002,
+      "loss": 0.7619,
+      "step": 9170
+    },
+    {
+      "epoch": 1.4841160779241775,
+      "grad_norm": 0.6677869558334351,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 9180
+    },
+    {
+      "epoch": 1.4857327621049228,
+      "grad_norm": 0.6052267551422119,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 9190
+    },
+    {
+      "epoch": 1.487349446285668,
+      "grad_norm": 0.6638872027397156,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 9200
+    },
+    {
+      "epoch": 1.4889661304664135,
+      "grad_norm": 0.6245523691177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 9210
+    },
+    {
+      "epoch": 1.4905828146471587,
+      "grad_norm": 0.5761767625808716,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 9220
+    },
+    {
+      "epoch": 1.492199498827904,
+      "grad_norm": 0.8175981640815735,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 9230
+    },
+    {
+      "epoch": 1.4938161830086494,
+      "grad_norm": 0.9144009947776794,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 9240
+    },
+    {
+      "epoch": 1.4954328671893946,
+      "grad_norm": 0.5742552876472473,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 9250
+    },
+    {
+      "epoch": 1.4970495513701398,
+      "grad_norm": 0.534534215927124,
+      "learning_rate": 0.0002,
+      "loss": 0.7418,
+      "step": 9260
+    },
+    {
+      "epoch": 1.498666235550885,
+      "grad_norm": 0.7836225032806396,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5002829197316303,
+      "grad_norm": 0.5292993187904358,
+      "learning_rate": 0.0002,
+      "loss": 0.7453,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5018996039123758,
+      "grad_norm": 0.8044071793556213,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9290
+    },
+    {
+      "epoch": 1.503516288093121,
+      "grad_norm": 0.6185805201530457,
+      "learning_rate": 0.0002,
+      "loss": 0.7229,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5051329722738664,
+      "grad_norm": 0.6093607544898987,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5067496564546117,
+      "grad_norm": 0.5891730189323425,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 9320
+    },
+    {
+      "epoch": 1.508366340635357,
+      "grad_norm": 0.6331129670143127,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5099830248161021,
+      "grad_norm": 0.7690958380699158,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5115997089968474,
+      "grad_norm": 0.6548877358436584,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5132163931775926,
+      "grad_norm": 0.6545143127441406,
+      "learning_rate": 0.0002,
+      "loss": 0.7408,
+      "step": 9360
+    },
+    {
+      "epoch": 1.514833077358338,
+      "grad_norm": 0.553247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.7432,
+      "step": 9370
+    },
+    {
+      "epoch": 1.5164497615390833,
+      "grad_norm": 0.8145074844360352,
+      "learning_rate": 0.0002,
+      "loss": 0.7265,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5180664457198287,
+      "grad_norm": 0.7636994123458862,
+      "learning_rate": 0.0002,
+      "loss": 0.7379,
+      "step": 9390
+    },
+    {
+      "epoch": 1.519683129900574,
+      "grad_norm": 0.6838982701301575,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5212998140813192,
+      "grad_norm": 0.8599441647529602,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5229164982620644,
+      "grad_norm": 0.7020329833030701,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5245331824428097,
+      "grad_norm": 0.6964772343635559,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5261498666235551,
+      "grad_norm": 0.6916600465774536,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9440
+    },
+    {
+      "epoch": 1.5277665508043003,
+      "grad_norm": 0.7282621264457703,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5293832349850458,
+      "grad_norm": 0.5363983511924744,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 9460
+    },
+    {
+      "epoch": 1.530999919165791,
+      "grad_norm": 0.6184861063957214,
+      "learning_rate": 0.0002,
+      "loss": 0.7154,
+      "step": 9470
+    },
+    {
+      "epoch": 1.5326166033465363,
+      "grad_norm": 0.5991285443305969,
+      "learning_rate": 0.0002,
+      "loss": 0.7837,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5342332875272815,
+      "grad_norm": 0.8176587820053101,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5358499717080267,
+      "grad_norm": 0.6473721861839294,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 9500
+    },
+    {
+      "epoch": 1.5374666558887722,
+      "grad_norm": 0.7319952845573425,
+      "learning_rate": 0.0002,
+      "loss": 0.7632,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5390833400695174,
+      "grad_norm": 0.702900230884552,
+      "learning_rate": 0.0002,
+      "loss": 0.7706,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5407000242502629,
+      "grad_norm": 0.7971600294113159,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 9530
+    },
+    {
+      "epoch": 1.542316708431008,
+      "grad_norm": 0.6527525186538696,
+      "learning_rate": 0.0002,
+      "loss": 0.7352,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5439333926117533,
+      "grad_norm": 0.5791676044464111,
+      "learning_rate": 0.0002,
+      "loss": 0.7425,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5455500767924986,
+      "grad_norm": 0.5619390606880188,
+      "learning_rate": 0.0002,
+      "loss": 0.7585,
+      "step": 9560
+    },
+    {
+      "epoch": 1.5471667609732438,
+      "grad_norm": 0.5701689124107361,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9570
+    },
+    {
+      "epoch": 1.548783445153989,
+      "grad_norm": 0.47549352049827576,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5504001293347345,
+      "grad_norm": 0.8730611205101013,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5520168135154797,
+      "grad_norm": 0.6842091083526611,
+      "learning_rate": 0.0002,
+      "loss": 0.798,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5536334976962252,
+      "grad_norm": 0.6675129532814026,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5552501818769704,
+      "grad_norm": 0.8173956274986267,
+      "learning_rate": 0.0002,
+      "loss": 0.7954,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5568668660577156,
+      "grad_norm": 0.724947452545166,
+      "learning_rate": 0.0002,
+      "loss": 0.7535,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5584835502384609,
+      "grad_norm": 0.6154758930206299,
+      "learning_rate": 0.0002,
+      "loss": 0.7738,
+      "step": 9640
+    },
+    {
+      "epoch": 1.560100234419206,
+      "grad_norm": 0.6072008013725281,
+      "learning_rate": 0.0002,
+      "loss": 0.7568,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5617169185999515,
+      "grad_norm": 0.659010648727417,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 9660
+    },
+    {
+      "epoch": 1.5633336027806968,
+      "grad_norm": 0.65857994556427,
+      "learning_rate": 0.0002,
+      "loss": 0.673,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5649502869614422,
+      "grad_norm": 0.5914267301559448,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5665669711421875,
+      "grad_norm": 0.6248020529747009,
+      "learning_rate": 0.0002,
+      "loss": 0.7414,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5681836553229327,
+      "grad_norm": 0.7147795557975769,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 9700
+    },
+    {
+      "epoch": 1.569800339503678,
+      "grad_norm": 0.7076232433319092,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 9710
+    },
+    {
+      "epoch": 1.5714170236844232,
+      "grad_norm": 0.6217400431632996,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5730337078651684,
+      "grad_norm": 0.6709911227226257,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 9730
+    },
+    {
+      "epoch": 1.5746503920459138,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 9740
+    },
+    {
+      "epoch": 1.576267076226659,
+      "grad_norm": 0.6241145730018616,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 9750
+    },
+    {
+      "epoch": 1.5778837604074045,
+      "grad_norm": 0.4960934817790985,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 9760
+    },
+    {
+      "epoch": 1.5795004445881498,
+      "grad_norm": 0.6593309640884399,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 9770
+    },
+    {
+      "epoch": 1.581117128768895,
+      "grad_norm": 0.5814042091369629,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 9780
+    },
+    {
+      "epoch": 1.5827338129496402,
+      "grad_norm": 0.5936070680618286,
+      "learning_rate": 0.0002,
+      "loss": 0.7109,
+      "step": 9790
+    },
+    {
+      "epoch": 1.5843504971303854,
+      "grad_norm": 0.6454403400421143,
+      "learning_rate": 0.0002,
+      "loss": 0.7769,
+      "step": 9800
+    },
+    {
+      "epoch": 1.585967181311131,
+      "grad_norm": 0.7612107992172241,
+      "learning_rate": 0.0002,
+      "loss": 0.7677,
+      "step": 9810
+    },
+    {
+      "epoch": 1.5875838654918761,
+      "grad_norm": 0.6494482755661011,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5892005496726216,
+      "grad_norm": 0.7825694680213928,
+      "learning_rate": 0.0002,
+      "loss": 0.7569,
+      "step": 9830
+    },
+    {
+      "epoch": 1.5908172338533668,
+      "grad_norm": 0.6757757663726807,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 9840
+    },
+    {
+      "epoch": 1.592433918034112,
+      "grad_norm": 0.7105609178543091,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 9850
+    },
+    {
+      "epoch": 1.5940506022148573,
+      "grad_norm": 0.7596991062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 9860
+    },
+    {
+      "epoch": 1.5956672863956025,
+      "grad_norm": 0.5681525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7108,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5972839705763477,
+      "grad_norm": 0.6090980768203735,
+      "learning_rate": 0.0002,
+      "loss": 0.7811,
+      "step": 9880
+    },
+    {
+      "epoch": 1.5989006547570932,
+      "grad_norm": 0.6271613240242004,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6005173389378387,
+      "grad_norm": 0.7656369805335999,
+      "learning_rate": 0.0002,
+      "loss": 0.7419,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6021340231185839,
+      "grad_norm": 0.7504446506500244,
+      "learning_rate": 0.0002,
+      "loss": 0.7336,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6037507072993291,
+      "grad_norm": 0.659656286239624,
+      "learning_rate": 0.0002,
+      "loss": 0.7479,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6053673914800743,
+      "grad_norm": 0.6006826162338257,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6069840756608196,
+      "grad_norm": 0.7872757911682129,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6086007598415648,
+      "grad_norm": 0.5545852780342102,
+      "learning_rate": 0.0002,
+      "loss": 0.768,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6102174440223103,
+      "grad_norm": 0.7429468631744385,
+      "learning_rate": 0.0002,
+      "loss": 0.8064,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6118341282030555,
+      "grad_norm": 0.6873556971549988,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 9970
+    },
+    {
+      "epoch": 1.613450812383801,
+      "grad_norm": 0.5874287486076355,
+      "learning_rate": 0.0002,
+      "loss": 0.7324,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6150674965645462,
+      "grad_norm": 0.6039386987686157,
+      "learning_rate": 0.0002,
+      "loss": 0.7141,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6166841807452914,
+      "grad_norm": 0.6233575940132141,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6183008649260366,
+      "grad_norm": 0.7676448225975037,
+      "learning_rate": 0.0002,
+      "loss": 0.7602,
+      "step": 10010
+    },
+    {
+      "epoch": 1.6199175491067819,
+      "grad_norm": 0.6565698385238647,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6215342332875273,
+      "grad_norm": 0.6787590384483337,
+      "learning_rate": 0.0002,
+      "loss": 0.7104,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6231509174682726,
+      "grad_norm": 0.6137678027153015,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 10040
+    },
+    {
+      "epoch": 1.624767601649018,
+      "grad_norm": 0.5236800312995911,
+      "learning_rate": 0.0002,
+      "loss": 0.7646,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6263842858297632,
+      "grad_norm": 0.7626367807388306,
+      "learning_rate": 0.0002,
+      "loss": 0.7437,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6280009700105085,
+      "grad_norm": 0.5657260417938232,
+      "learning_rate": 0.0002,
+      "loss": 0.7273,
+      "step": 10070
+    },
+    {
+      "epoch": 1.6296176541912537,
+      "grad_norm": 0.4913991391658783,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 10080
+    },
+    {
+      "epoch": 1.631234338371999,
+      "grad_norm": 0.7715556621551514,
+      "learning_rate": 0.0002,
+      "loss": 0.7596,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6328510225527442,
+      "grad_norm": 0.6509000062942505,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6344677067334896,
+      "grad_norm": 0.6215850114822388,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6360843909142349,
+      "grad_norm": 0.6956844329833984,
+      "learning_rate": 0.0002,
+      "loss": 0.7705,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6377010750949803,
+      "grad_norm": 0.6111597418785095,
+      "learning_rate": 0.0002,
+      "loss": 0.7129,
+      "step": 10130
+    },
+    {
+      "epoch": 1.6393177592757255,
+      "grad_norm": 0.6518288850784302,
+      "learning_rate": 0.0002,
+      "loss": 0.6955,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6409344434564708,
+      "grad_norm": 0.6914522051811218,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10150
+    },
+    {
+      "epoch": 1.642551127637216,
+      "grad_norm": 0.63785719871521,
+      "learning_rate": 0.0002,
+      "loss": 0.7295,
+      "step": 10160
+    },
+    {
+      "epoch": 1.6441678118179612,
+      "grad_norm": 0.6379287838935852,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6457844959987067,
+      "grad_norm": 0.6793403029441833,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 10180
+    },
+    {
+      "epoch": 1.647401180179452,
+      "grad_norm": 0.6099132895469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6490178643601974,
+      "grad_norm": 0.5869854092597961,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6506345485409426,
+      "grad_norm": 0.7716999053955078,
+      "learning_rate": 0.0002,
+      "loss": 0.8308,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6522512327216878,
+      "grad_norm": 0.6854110360145569,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 10220
+    },
+    {
+      "epoch": 1.653867916902433,
+      "grad_norm": 0.6957170367240906,
+      "learning_rate": 0.0002,
+      "loss": 0.782,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6554846010831783,
+      "grad_norm": 0.6932903528213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6571012852639235,
+      "grad_norm": 0.7713165283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10250
+    },
+    {
+      "epoch": 1.658717969444669,
+      "grad_norm": 0.7455793619155884,
+      "learning_rate": 0.0002,
+      "loss": 0.7099,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6603346536254144,
+      "grad_norm": 0.5464168190956116,
+      "learning_rate": 0.0002,
+      "loss": 0.7524,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6619513378061597,
+      "grad_norm": 0.6782926321029663,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 10280
+    },
+    {
+      "epoch": 1.663568021986905,
+      "grad_norm": 0.7962649464607239,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6651847061676501,
+      "grad_norm": 0.6814526319503784,
+      "learning_rate": 0.0002,
+      "loss": 0.7142,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6668013903483954,
+      "grad_norm": 0.656895101070404,
+      "learning_rate": 0.0002,
+      "loss": 0.7285,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6684180745291406,
+      "grad_norm": 0.6085672378540039,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 10320
+    },
+    {
+      "epoch": 1.670034758709886,
+      "grad_norm": 0.585508406162262,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 10330
+    },
+    {
+      "epoch": 1.6716514428906313,
+      "grad_norm": 0.6930184364318848,
+      "learning_rate": 0.0002,
+      "loss": 0.7604,
+      "step": 10340
+    },
+    {
+      "epoch": 1.6732681270713767,
+      "grad_norm": 0.575663149356842,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10350
+    },
+    {
+      "epoch": 1.674884811252122,
+      "grad_norm": 0.582502543926239,
+      "learning_rate": 0.0002,
+      "loss": 0.7198,
+      "step": 10360
+    },
+    {
+      "epoch": 1.6765014954328672,
+      "grad_norm": 0.5668916702270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 10370
+    },
+    {
+      "epoch": 1.6781181796136124,
+      "grad_norm": 0.6070065498352051,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10380
+    },
+    {
+      "epoch": 1.6797348637943577,
+      "grad_norm": 0.6141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.7939,
+      "step": 10390
+    },
+    {
+      "epoch": 1.6813515479751031,
+      "grad_norm": 0.8359124064445496,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 10400
+    },
+    {
+      "epoch": 1.6829682321558483,
+      "grad_norm": 0.5378185510635376,
+      "learning_rate": 0.0002,
+      "loss": 0.7488,
+      "step": 10410
+    },
+    {
+      "epoch": 1.6845849163365938,
+      "grad_norm": 0.6959536075592041,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 10420
+    },
+    {
+      "epoch": 1.686201600517339,
+      "grad_norm": 0.6514357328414917,
+      "learning_rate": 0.0002,
+      "loss": 0.7872,
+      "step": 10430
+    },
+    {
+      "epoch": 1.6878182846980843,
+      "grad_norm": 0.7706646919250488,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10440
+    },
+    {
+      "epoch": 1.6894349688788295,
+      "grad_norm": 0.6183337569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.7673,
+      "step": 10450
+    },
+    {
+      "epoch": 1.6910516530595747,
+      "grad_norm": 0.6123278141021729,
+      "learning_rate": 0.0002,
+      "loss": 0.7566,
+      "step": 10460
+    },
+    {
+      "epoch": 1.69266833724032,
+      "grad_norm": 0.6894851326942444,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10470
+    },
+    {
+      "epoch": 1.6942850214210654,
+      "grad_norm": 0.7497312426567078,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 10480
+    },
+    {
+      "epoch": 1.6959017056018106,
+      "grad_norm": 0.5968214273452759,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10490
+    },
+    {
+      "epoch": 1.697518389782556,
+      "grad_norm": 0.6747927069664001,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 10500
+    },
+    {
+      "epoch": 1.6991350739633013,
+      "grad_norm": 0.5708310008049011,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7007517581440466,
+      "grad_norm": 0.606526792049408,
+      "learning_rate": 0.0002,
+      "loss": 0.7385,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7023684423247918,
+      "grad_norm": 0.662011981010437,
+      "learning_rate": 0.0002,
+      "loss": 0.7204,
+      "step": 10530
+    },
+    {
+      "epoch": 1.703985126505537,
+      "grad_norm": 0.7583045363426208,
+      "learning_rate": 0.0002,
+      "loss": 0.7999,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7056018106862825,
+      "grad_norm": 0.721632182598114,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7072184948670277,
+      "grad_norm": 0.6107715368270874,
+      "learning_rate": 0.0002,
+      "loss": 0.7407,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7088351790477732,
+      "grad_norm": 0.6652471423149109,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7104518632285184,
+      "grad_norm": 0.6308087110519409,
+      "learning_rate": 0.0002,
+      "loss": 0.7767,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7120685474092636,
+      "grad_norm": 0.5464386940002441,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7136852315900089,
+      "grad_norm": 0.6558911204338074,
+      "learning_rate": 0.0002,
+      "loss": 0.7063,
+      "step": 10600
+    },
+    {
+      "epoch": 1.715301915770754,
+      "grad_norm": 0.5665024518966675,
+      "learning_rate": 0.0002,
+      "loss": 0.7126,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7169185999514993,
+      "grad_norm": 0.7888094186782837,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7185352841322448,
+      "grad_norm": 0.7084909081459045,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 10630
+    },
+    {
+      "epoch": 1.7201519683129902,
+      "grad_norm": 0.7982324361801147,
+      "learning_rate": 0.0002,
+      "loss": 0.7557,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7217686524937355,
+      "grad_norm": 0.6418732404708862,
+      "learning_rate": 0.0002,
+      "loss": 0.7345,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7233853366744807,
+      "grad_norm": 0.7636681795120239,
+      "learning_rate": 0.0002,
+      "loss": 0.7734,
+      "step": 10660
+    },
+    {
+      "epoch": 1.725002020855226,
+      "grad_norm": 0.5646875500679016,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7266187050359711,
+      "grad_norm": 0.5231260657310486,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7282353892167164,
+      "grad_norm": 0.7635011672973633,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7298520733974618,
+      "grad_norm": 0.7518259286880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 10700
+    },
+    {
+      "epoch": 1.731468757578207,
+      "grad_norm": 0.7295602560043335,
+      "learning_rate": 0.0002,
+      "loss": 0.751,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7330854417589525,
+      "grad_norm": 0.6984632015228271,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7347021259396977,
+      "grad_norm": 0.6198219060897827,
+      "learning_rate": 0.0002,
+      "loss": 0.7921,
+      "step": 10730
+    },
+    {
+      "epoch": 1.736318810120443,
+      "grad_norm": 0.6957576274871826,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7379354943011882,
+      "grad_norm": 0.6430263519287109,
+      "learning_rate": 0.0002,
+      "loss": 0.7917,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7395521784819334,
+      "grad_norm": 0.6134995222091675,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 10760
+    },
+    {
+      "epoch": 1.741168862662679,
+      "grad_norm": 0.7209452986717224,
+      "learning_rate": 0.0002,
+      "loss": 0.7584,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7427855468434241,
+      "grad_norm": 0.6735447645187378,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7444022310241696,
+      "grad_norm": 0.5605693459510803,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 10790
+    },
+    {
+      "epoch": 1.7460189152049148,
+      "grad_norm": 0.6882363557815552,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 10800
+    },
+    {
+      "epoch": 1.74763559938566,
+      "grad_norm": 0.6386259198188782,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7492522835664053,
+      "grad_norm": 0.6529015302658081,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7508689677471505,
+      "grad_norm": 0.5664082765579224,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7524856519278957,
+      "grad_norm": 0.7532684206962585,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7541023361086412,
+      "grad_norm": 0.77171391248703,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7557190202893864,
+      "grad_norm": 0.7255431413650513,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7573357044701319,
+      "grad_norm": 0.763083279132843,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10870
+    },
+    {
+      "epoch": 1.758952388650877,
+      "grad_norm": 0.6042402982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7605690728316223,
+      "grad_norm": 0.7642518281936646,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7621857570123676,
+      "grad_norm": 0.6347904801368713,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7638024411931128,
+      "grad_norm": 0.5371627807617188,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7654191253738583,
+      "grad_norm": 0.6840225458145142,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7670358095546035,
+      "grad_norm": 0.5288469195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10930
+    },
+    {
+      "epoch": 1.768652493735349,
+      "grad_norm": 0.69020676612854,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 10940
+    },
+    {
+      "epoch": 1.7702691779160942,
+      "grad_norm": 0.5943242311477661,
+      "learning_rate": 0.0002,
+      "loss": 0.7468,
+      "step": 10950
+    },
+    {
+      "epoch": 1.7718858620968394,
+      "grad_norm": 0.5616418123245239,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 10960
+    },
+    {
+      "epoch": 1.7735025462775846,
+      "grad_norm": 0.7209470868110657,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 10970
+    },
+    {
+      "epoch": 1.7751192304583299,
+      "grad_norm": 0.6657957434654236,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 10980
+    },
+    {
+      "epoch": 1.776735914639075,
+      "grad_norm": 0.6469064950942993,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 10990
+    },
+    {
+      "epoch": 1.7783525988198206,
+      "grad_norm": 0.6615678071975708,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 11000
+    },
+    {
+      "epoch": 1.779969283000566,
+      "grad_norm": 0.6722439527511597,
+      "learning_rate": 0.0002,
+      "loss": 0.747,
+      "step": 11010
+    },
+    {
+      "epoch": 1.7815859671813112,
+      "grad_norm": 0.634136974811554,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 11020
+    },
+    {
+      "epoch": 1.7832026513620565,
+      "grad_norm": 0.6024377346038818,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 11030
+    },
+    {
+      "epoch": 1.7848193355428017,
+      "grad_norm": 0.6909403800964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 11040
+    },
+    {
+      "epoch": 1.786436019723547,
+      "grad_norm": 0.7148767709732056,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 11050
+    },
+    {
+      "epoch": 1.7880527039042922,
+      "grad_norm": 0.7442979216575623,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 11060
+    },
+    {
+      "epoch": 1.7896693880850376,
+      "grad_norm": 0.6830431818962097,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 11070
+    },
+    {
+      "epoch": 1.7912860722657828,
+      "grad_norm": 0.9172667264938354,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 11080
+    },
+    {
+      "epoch": 1.7929027564465283,
+      "grad_norm": 0.6799490451812744,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 11090
+    },
+    {
+      "epoch": 1.7945194406272735,
+      "grad_norm": 0.7617024779319763,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 11100
+    },
+    {
+      "epoch": 1.7961361248080188,
+      "grad_norm": 0.7701810002326965,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 11110
+    },
+    {
+      "epoch": 1.797752808988764,
+      "grad_norm": 0.7454385757446289,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11120
+    },
+    {
+      "epoch": 1.7993694931695092,
+      "grad_norm": 0.6121436953544617,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8009861773502547,
+      "grad_norm": 0.6237571835517883,
+      "learning_rate": 0.0002,
+      "loss": 0.7305,
+      "step": 11140
+    },
+    {
+      "epoch": 1.802602861531,
+      "grad_norm": 0.6818515658378601,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 11150
+    },
+    {
+      "epoch": 1.8042195457117454,
+      "grad_norm": 0.7768308520317078,
+      "learning_rate": 0.0002,
+      "loss": 0.6876,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8058362298924906,
+      "grad_norm": 0.6875537633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 11170
+    },
+    {
+      "epoch": 1.8074529140732358,
+      "grad_norm": 0.7950584888458252,
+      "learning_rate": 0.0002,
+      "loss": 0.761,
+      "step": 11180
+    },
+    {
+      "epoch": 1.809069598253981,
+      "grad_norm": 0.8210248351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8106862824347263,
+      "grad_norm": 0.6674110889434814,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8123029666154715,
+      "grad_norm": 0.6261674761772156,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 11210
+    },
+    {
+      "epoch": 1.813919650796217,
+      "grad_norm": 0.6484741568565369,
+      "learning_rate": 0.0002,
+      "loss": 0.7122,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8155363349769622,
+      "grad_norm": 0.6231244206428528,
+      "learning_rate": 0.0002,
+      "loss": 0.7718,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8171530191577077,
+      "grad_norm": 0.7243146896362305,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11240
+    },
+    {
+      "epoch": 1.818769703338453,
+      "grad_norm": 0.6776193380355835,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8203863875191981,
+      "grad_norm": 0.5973618030548096,
+      "learning_rate": 0.0002,
+      "loss": 0.7317,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8220030716999434,
+      "grad_norm": 0.6451361179351807,
+      "learning_rate": 0.0002,
+      "loss": 0.7961,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8236197558806886,
+      "grad_norm": 0.5963068008422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 11280
+    },
+    {
+      "epoch": 1.825236440061434,
+      "grad_norm": 0.536902129650116,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8268531242421793,
+      "grad_norm": 0.6993787288665771,
+      "learning_rate": 0.0002,
+      "loss": 0.708,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8284698084229247,
+      "grad_norm": 0.6135255098342896,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11310
+    },
+    {
+      "epoch": 1.83008649260367,
+      "grad_norm": 0.6057423949241638,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8317031767844152,
+      "grad_norm": 0.6598812341690063,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8333198609651604,
+      "grad_norm": 0.6075948476791382,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 11340
+    },
+    {
+      "epoch": 1.8349365451459057,
+      "grad_norm": 0.7065447568893433,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8365532293266509,
+      "grad_norm": 0.680526614189148,
+      "learning_rate": 0.0002,
+      "loss": 0.7365,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8381699135073963,
+      "grad_norm": 0.6356695294380188,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8397865976881416,
+      "grad_norm": 0.6399052143096924,
+      "learning_rate": 0.0002,
+      "loss": 0.721,
+      "step": 11380
+    },
+    {
+      "epoch": 1.841403281868887,
+      "grad_norm": 0.6125704050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 11390
+    },
+    {
+      "epoch": 1.8430199660496323,
+      "grad_norm": 0.7124643325805664,
+      "learning_rate": 0.0002,
+      "loss": 0.755,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8446366502303775,
+      "grad_norm": 0.6099604964256287,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8462533344111227,
+      "grad_norm": 0.7338208556175232,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 11420
+    },
+    {
+      "epoch": 1.847870018591868,
+      "grad_norm": 0.7534668445587158,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8494867027726134,
+      "grad_norm": 0.6135470271110535,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8511033869533586,
+      "grad_norm": 0.6229309439659119,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 11450
+    },
+    {
+      "epoch": 1.852720071134104,
+      "grad_norm": 0.706423282623291,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 11460
+    },
+    {
+      "epoch": 1.8543367553148493,
+      "grad_norm": 0.5460049510002136,
+      "learning_rate": 0.0002,
+      "loss": 0.7411,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8559534394955945,
+      "grad_norm": 0.6616711020469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7416,
+      "step": 11480
+    },
+    {
+      "epoch": 1.8575701236763398,
+      "grad_norm": 0.6372783184051514,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 11490
+    },
+    {
+      "epoch": 1.859186807857085,
+      "grad_norm": 0.7162668108940125,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8608034920378305,
+      "grad_norm": 0.6605209708213806,
+      "learning_rate": 0.0002,
+      "loss": 0.7747,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8624201762185757,
+      "grad_norm": 0.6933956742286682,
+      "learning_rate": 0.0002,
+      "loss": 0.7258,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8640368603993211,
+      "grad_norm": 0.6582090854644775,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8656535445800664,
+      "grad_norm": 0.6416500806808472,
+      "learning_rate": 0.0002,
+      "loss": 0.7313,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8672702287608116,
+      "grad_norm": 0.5434312224388123,
+      "learning_rate": 0.0002,
+      "loss": 0.7372,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8688869129415568,
+      "grad_norm": 0.6827567219734192,
+      "learning_rate": 0.0002,
+      "loss": 0.7635,
+      "step": 11560
+    },
+    {
+      "epoch": 1.870503597122302,
+      "grad_norm": 0.7354370951652527,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 11570
+    },
+    {
+      "epoch": 1.8721202813030473,
+      "grad_norm": 0.590372622013092,
+      "learning_rate": 0.0002,
+      "loss": 0.7526,
+      "step": 11580
+    },
+    {
+      "epoch": 1.8737369654837928,
+      "grad_norm": 0.853183925151825,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 11590
+    },
+    {
+      "epoch": 1.875353649664538,
+      "grad_norm": 0.822678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 11600
+    },
+    {
+      "epoch": 1.8769703338452834,
+      "grad_norm": 0.6591550707817078,
+      "learning_rate": 0.0002,
+      "loss": 0.7427,
+      "step": 11610
+    },
+    {
+      "epoch": 1.8785870180260287,
+      "grad_norm": 0.7475301623344421,
+      "learning_rate": 0.0002,
+      "loss": 0.7054,
+      "step": 11620
+    },
+    {
+      "epoch": 1.880203702206774,
+      "grad_norm": 0.6390765309333801,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 11630
+    },
+    {
+      "epoch": 1.8818203863875191,
+      "grad_norm": 0.6589758992195129,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 11640
+    },
+    {
+      "epoch": 1.8834370705682644,
+      "grad_norm": 0.6765508651733398,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 11650
+    },
+    {
+      "epoch": 1.8850537547490098,
+      "grad_norm": 0.6527857780456543,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 11660
+    },
+    {
+      "epoch": 1.886670438929755,
+      "grad_norm": 0.6642923951148987,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 11670
+    },
+    {
+      "epoch": 1.8882871231105005,
+      "grad_norm": 0.6945584416389465,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 11680
+    },
+    {
+      "epoch": 1.8899038072912457,
+      "grad_norm": 0.694018542766571,
+      "learning_rate": 0.0002,
+      "loss": 0.7711,
+      "step": 11690
+    },
+    {
+      "epoch": 1.891520491471991,
+      "grad_norm": 0.7237417101860046,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 11700
+    },
+    {
+      "epoch": 1.8931371756527362,
+      "grad_norm": 0.7401309609413147,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 11710
+    },
+    {
+      "epoch": 1.8947538598334814,
+      "grad_norm": 0.6537784337997437,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 11720
+    },
+    {
+      "epoch": 1.8963705440142267,
+      "grad_norm": 0.7398539185523987,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 11730
+    },
+    {
+      "epoch": 1.8979872281949721,
+      "grad_norm": 0.6696075797080994,
+      "learning_rate": 0.0002,
+      "loss": 0.7561,
+      "step": 11740
+    },
+    {
+      "epoch": 1.8996039123757174,
+      "grad_norm": 0.6014142036437988,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9012205965564628,
+      "grad_norm": 0.7023524641990662,
+      "learning_rate": 0.0002,
+      "loss": 0.7714,
+      "step": 11760
+    },
+    {
+      "epoch": 1.902837280737208,
+      "grad_norm": 0.739973783493042,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9044539649179533,
+      "grad_norm": 0.5576770901679993,
+      "learning_rate": 0.0002,
+      "loss": 0.7848,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9060706490986985,
+      "grad_norm": 0.6907393932342529,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 11790
+    },
+    {
+      "epoch": 1.9076873332794437,
+      "grad_norm": 0.6934581995010376,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9093040174601892,
+      "grad_norm": 0.591774582862854,
+      "learning_rate": 0.0002,
+      "loss": 0.7199,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9109207016409344,
+      "grad_norm": 0.6249791383743286,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11820
+    },
+    {
+      "epoch": 1.9125373858216799,
+      "grad_norm": 0.6755744218826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 11830
+    },
+    {
+      "epoch": 1.914154070002425,
+      "grad_norm": 0.7286285161972046,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9157707541831703,
+      "grad_norm": 0.7867850065231323,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9173874383639156,
+      "grad_norm": 0.6283972859382629,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11860
+    },
+    {
+      "epoch": 1.9190041225446608,
+      "grad_norm": 0.605823814868927,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 11870
+    },
+    {
+      "epoch": 1.920620806725406,
+      "grad_norm": 0.5927976965904236,
+      "learning_rate": 0.0002,
+      "loss": 0.6598,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9222374909061515,
+      "grad_norm": 0.5974002480506897,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 11890
+    },
+    {
+      "epoch": 1.923854175086897,
+      "grad_norm": 0.7091866135597229,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9254708592676422,
+      "grad_norm": 0.72496497631073,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9270875434483874,
+      "grad_norm": 0.6131896376609802,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9287042276291326,
+      "grad_norm": 0.6556436419487,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9303209118098779,
+      "grad_norm": 0.622932493686676,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 11940
+    },
+    {
+      "epoch": 1.931937595990623,
+      "grad_norm": 0.6618631482124329,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9335542801713685,
+      "grad_norm": 0.630966305732727,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 11960
+    },
+    {
+      "epoch": 1.9351709643521138,
+      "grad_norm": 0.6336734890937805,
+      "learning_rate": 0.0002,
+      "loss": 0.6854,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9367876485328592,
+      "grad_norm": 0.655403196811676,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9384043327136045,
+      "grad_norm": 0.5640574097633362,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 11990
+    },
+    {
+      "epoch": 1.9400210168943497,
+      "grad_norm": 0.6322951316833496,
+      "learning_rate": 0.0002,
+      "loss": 0.7289,
+      "step": 12000
+    },
+    {
+      "epoch": 1.941637701075095,
+      "grad_norm": 0.615703821182251,
+      "learning_rate": 0.0002,
+      "loss": 0.7627,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9432543852558402,
+      "grad_norm": 0.6487536430358887,
+      "learning_rate": 0.0002,
+      "loss": 0.786,
+      "step": 12020
+    },
+    {
+      "epoch": 1.9448710694365856,
+      "grad_norm": 0.9209630489349365,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 12030
+    },
+    {
+      "epoch": 1.9464877536173308,
+      "grad_norm": 0.67485511302948,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9481044377980763,
+      "grad_norm": 0.6831230521202087,
+      "learning_rate": 0.0002,
+      "loss": 0.7551,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9497211219788215,
+      "grad_norm": 0.6578302383422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7546,
+      "step": 12060
+    },
+    {
+      "epoch": 1.9513378061595668,
+      "grad_norm": 0.9975938200950623,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12070
+    },
+    {
+      "epoch": 1.952954490340312,
+      "grad_norm": 0.6637365221977234,
+      "learning_rate": 0.0002,
+      "loss": 0.7952,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9545711745210572,
+      "grad_norm": 0.605707049369812,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9561878587018025,
+      "grad_norm": 0.6584440469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7768,
+      "step": 12100
+    },
+    {
+      "epoch": 1.957804542882548,
+      "grad_norm": 0.6070835590362549,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9594212270632931,
+      "grad_norm": 0.7862601280212402,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9610379112440386,
+      "grad_norm": 0.8175255060195923,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 12130
+    },
+    {
+      "epoch": 1.9626545954247838,
+      "grad_norm": 0.5648472905158997,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 12140
+    },
+    {
+      "epoch": 1.964271279605529,
+      "grad_norm": 0.6591973304748535,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9658879637862743,
+      "grad_norm": 0.5960676074028015,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9675046479670195,
+      "grad_norm": 0.7272544503211975,
+      "learning_rate": 0.0002,
+      "loss": 0.7254,
+      "step": 12170
+    },
+    {
+      "epoch": 1.969121332147765,
+      "grad_norm": 0.7176699042320251,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9707380163285102,
+      "grad_norm": 0.6927123665809631,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9723547005092557,
+      "grad_norm": 0.5536034107208252,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 12200
+    },
+    {
+      "epoch": 1.9739713846900009,
+      "grad_norm": 0.8348390460014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 12210
+    },
+    {
+      "epoch": 1.9755880688707461,
+      "grad_norm": 0.6591181755065918,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 12220
+    },
+    {
+      "epoch": 1.9772047530514913,
+      "grad_norm": 1.0624109506607056,
+      "learning_rate": 0.0002,
+      "loss": 0.763,
+      "step": 12230
+    },
+    {
+      "epoch": 1.9788214372322366,
+      "grad_norm": 0.9265586137771606,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 12240
+    },
+    {
+      "epoch": 1.9804381214129818,
+      "grad_norm": 0.5998196005821228,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 12250
+    },
+    {
+      "epoch": 1.9820548055937273,
+      "grad_norm": 0.6960851550102234,
+      "learning_rate": 0.0002,
+      "loss": 0.7225,
+      "step": 12260
+    },
+    {
+      "epoch": 1.9836714897744727,
+      "grad_norm": 0.7674502730369568,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 12270
+    },
+    {
+      "epoch": 1.985288173955218,
+      "grad_norm": 0.6407275795936584,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 12280
+    },
+    {
+      "epoch": 1.9869048581359632,
+      "grad_norm": 0.6673079133033752,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 12290
+    },
+    {
+      "epoch": 1.9885215423167084,
+      "grad_norm": 0.6989844441413879,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 12300
+    },
+    {
+      "epoch": 1.9901382264974536,
+      "grad_norm": 0.7564442157745361,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 12310
+    },
+    {
+      "epoch": 1.9917549106781989,
+      "grad_norm": 0.6385478973388672,
+      "learning_rate": 0.0002,
+      "loss": 0.7719,
+      "step": 12320
+    },
+    {
+      "epoch": 1.9933715948589443,
+      "grad_norm": 0.7193717956542969,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 12330
+    },
+    {
+      "epoch": 1.9949882790396896,
+      "grad_norm": 0.7987112402915955,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 12340
+    },
+    {
+      "epoch": 1.996604963220435,
+      "grad_norm": 0.7260826826095581,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 12350
+    },
+    {
+      "epoch": 1.9982216474011802,
+      "grad_norm": 0.7968255281448364,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 12360
+    },
+    {
+      "epoch": 1.9998383315819255,
+      "grad_norm": 0.6893062591552734,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1044032573699951,
+      "eval_runtime": 122.1508,
+      "eval_samples_per_second": 6.001,
+      "eval_steps_per_second": 0.753,
+      "step": 12371
+    },
+    {
+      "epoch": 2.0014550157626707,
+      "grad_norm": 0.7775409817695618,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 12380
+    },
+    {
+      "epoch": 2.003071699943416,
+      "grad_norm": 0.76218581199646,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 12390
+    },
+    {
+      "epoch": 2.004688384124161,
+      "grad_norm": 0.5677764415740967,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 12400
+    },
+    {
+      "epoch": 2.006305068304907,
+      "grad_norm": 0.808442234992981,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 12410
+    },
+    {
+      "epoch": 2.007921752485652,
+      "grad_norm": 0.7144765257835388,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 12420
+    },
+    {
+      "epoch": 2.0095384366663973,
+      "grad_norm": 0.6914031505584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 12430
+    },
+    {
+      "epoch": 2.0111551208471425,
+      "grad_norm": 0.7581454515457153,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 12440
+    },
+    {
+      "epoch": 2.0127718050278878,
+      "grad_norm": 0.8388504981994629,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 12450
+    },
+    {
+      "epoch": 2.014388489208633,
+      "grad_norm": 0.6716406941413879,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 12460
+    },
+    {
+      "epoch": 2.0160051733893782,
+      "grad_norm": 0.898902416229248,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 12470
+    },
+    {
+      "epoch": 2.0176218575701235,
+      "grad_norm": 0.6432679891586304,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 12480
+    },
+    {
+      "epoch": 2.019238541750869,
+      "grad_norm": 0.8021109104156494,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12490
+    },
+    {
+      "epoch": 2.0208552259316144,
+      "grad_norm": 0.7039216756820679,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 12500
+    },
+    {
+      "epoch": 2.0224719101123596,
+      "grad_norm": 0.646531879901886,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12510
+    },
+    {
+      "epoch": 2.024088594293105,
+      "grad_norm": 0.783704400062561,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 12520
+    },
+    {
+      "epoch": 2.02570527847385,
+      "grad_norm": 0.8805046677589417,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12530
+    },
+    {
+      "epoch": 2.0273219626545953,
+      "grad_norm": 0.7289270758628845,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0289386468353405,
+      "grad_norm": 0.71653151512146,
+      "learning_rate": 0.0002,
+      "loss": 0.6663,
+      "step": 12550
+    },
+    {
+      "epoch": 2.030555331016086,
+      "grad_norm": 0.73281329870224,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 12560
+    },
+    {
+      "epoch": 2.0321720151968314,
+      "grad_norm": 0.6657090187072754,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 12570
+    },
+    {
+      "epoch": 2.0337886993775767,
+      "grad_norm": 0.8241133093833923,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 12580
+    },
+    {
+      "epoch": 2.035405383558322,
+      "grad_norm": 0.5834135413169861,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 12590
+    },
+    {
+      "epoch": 2.037022067739067,
+      "grad_norm": 0.84502112865448,
+      "learning_rate": 0.0002,
+      "loss": 0.6188,
+      "step": 12600
+    },
+    {
+      "epoch": 2.0386387519198124,
+      "grad_norm": 0.8952481746673584,
+      "learning_rate": 0.0002,
+      "loss": 0.6349,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0402554361005576,
+      "grad_norm": 0.7801461815834045,
+      "learning_rate": 0.0002,
+      "loss": 0.6923,
+      "step": 12620
+    },
+    {
+      "epoch": 2.041872120281303,
+      "grad_norm": 0.6788367033004761,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 12630
+    },
+    {
+      "epoch": 2.0434888044620485,
+      "grad_norm": 0.7241756319999695,
+      "learning_rate": 0.0002,
+      "loss": 0.6162,
+      "step": 12640
+    },
+    {
+      "epoch": 2.0451054886427937,
+      "grad_norm": 0.6933388113975525,
+      "learning_rate": 0.0002,
+      "loss": 0.655,
+      "step": 12650
+    },
+    {
+      "epoch": 2.046722172823539,
+      "grad_norm": 0.8029746413230896,
+      "learning_rate": 0.0002,
+      "loss": 0.6431,
+      "step": 12660
+    },
+    {
+      "epoch": 2.048338857004284,
+      "grad_norm": 0.946399986743927,
+      "learning_rate": 0.0002,
+      "loss": 0.7164,
+      "step": 12670
+    },
+    {
+      "epoch": 2.0499555411850294,
+      "grad_norm": 0.7072678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.638,
+      "step": 12680
+    },
+    {
+      "epoch": 2.0515722253657747,
+      "grad_norm": 0.6810618042945862,
+      "learning_rate": 0.0002,
+      "loss": 0.6487,
+      "step": 12690
+    },
+    {
+      "epoch": 2.05318890954652,
+      "grad_norm": 0.7661160230636597,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 12700
+    },
+    {
+      "epoch": 2.0548055937272656,
+      "grad_norm": 0.6350653767585754,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 12710
+    },
+    {
+      "epoch": 2.056422277908011,
+      "grad_norm": 0.861890971660614,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 12720
+    },
+    {
+      "epoch": 2.058038962088756,
+      "grad_norm": 0.6489875912666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 12730
+    },
+    {
+      "epoch": 2.0596556462695013,
+      "grad_norm": 0.8268506526947021,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 12740
+    },
+    {
+      "epoch": 2.0612723304502465,
+      "grad_norm": 0.607679545879364,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 12750
+    },
+    {
+      "epoch": 2.0628890146309917,
+      "grad_norm": 0.6754153370857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6649,
+      "step": 12760
+    },
+    {
+      "epoch": 2.064505698811737,
+      "grad_norm": 0.7263124585151672,
+      "learning_rate": 0.0002,
+      "loss": 0.6549,
+      "step": 12770
+    },
+    {
+      "epoch": 2.0661223829924826,
+      "grad_norm": 0.6986154317855835,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 12780
+    },
+    {
+      "epoch": 2.067739067173228,
+      "grad_norm": 0.7768576741218567,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 12790
+    },
+    {
+      "epoch": 2.069355751353973,
+      "grad_norm": 0.7546762824058533,
+      "learning_rate": 0.0002,
+      "loss": 0.677,
+      "step": 12800
+    },
+    {
+      "epoch": 2.0709724355347183,
+      "grad_norm": 0.7588880062103271,
+      "learning_rate": 0.0002,
+      "loss": 0.6485,
+      "step": 12810
+    },
+    {
+      "epoch": 2.0725891197154636,
+      "grad_norm": 0.7457242608070374,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12820
+    },
+    {
+      "epoch": 2.074205803896209,
+      "grad_norm": 0.6983516812324524,
+      "learning_rate": 0.0002,
+      "loss": 0.6489,
+      "step": 12830
+    },
+    {
+      "epoch": 2.075822488076954,
+      "grad_norm": 0.7950928807258606,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 12840
+    },
+    {
+      "epoch": 2.0774391722576993,
+      "grad_norm": 0.9248087406158447,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 12850
+    },
+    {
+      "epoch": 2.079055856438445,
+      "grad_norm": 0.7229493260383606,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 12860
+    },
+    {
+      "epoch": 2.08067254061919,
+      "grad_norm": 0.5710847973823547,
+      "learning_rate": 0.0002,
+      "loss": 0.6702,
+      "step": 12870
+    },
+    {
+      "epoch": 2.0822892247999354,
+      "grad_norm": 0.9580423831939697,
+      "learning_rate": 0.0002,
+      "loss": 0.6974,
+      "step": 12880
+    },
+    {
+      "epoch": 2.0839059089806806,
+      "grad_norm": 0.7399665713310242,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 12890
+    },
+    {
+      "epoch": 2.085522593161426,
+      "grad_norm": 0.7981410622596741,
+      "learning_rate": 0.0002,
+      "loss": 0.6993,
+      "step": 12900
+    },
+    {
+      "epoch": 2.087139277342171,
+      "grad_norm": 0.870759904384613,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 12910
+    },
+    {
+      "epoch": 2.0887559615229163,
+      "grad_norm": 0.7001481652259827,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 12920
+    },
+    {
+      "epoch": 2.090372645703662,
+      "grad_norm": 0.6745418310165405,
+      "learning_rate": 0.0002,
+      "loss": 0.6383,
+      "step": 12930
+    },
+    {
+      "epoch": 2.0919893298844072,
+      "grad_norm": 0.7739067673683167,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 12940
+    },
+    {
+      "epoch": 2.0936060140651525,
+      "grad_norm": 0.6742934584617615,
+      "learning_rate": 0.0002,
+      "loss": 0.6856,
+      "step": 12950
+    },
+    {
+      "epoch": 2.0952226982458977,
+      "grad_norm": 0.7270349860191345,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 12960
+    },
+    {
+      "epoch": 2.096839382426643,
+      "grad_norm": 0.7150624394416809,
+      "learning_rate": 0.0002,
+      "loss": 0.6783,
+      "step": 12970
+    },
+    {
+      "epoch": 2.098456066607388,
+      "grad_norm": 0.7734767198562622,
+      "learning_rate": 0.0002,
+      "loss": 0.6093,
+      "step": 12980
+    },
+    {
+      "epoch": 2.1000727507881334,
+      "grad_norm": 0.7618662118911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 12990
+    },
+    {
+      "epoch": 2.101689434968879,
+      "grad_norm": 0.6557944416999817,
+      "learning_rate": 0.0002,
+      "loss": 0.6707,
+      "step": 13000
+    },
+    {
+      "epoch": 2.1033061191496243,
+      "grad_norm": 0.8786448240280151,
+      "learning_rate": 0.0002,
+      "loss": 0.7268,
+      "step": 13010
+    },
+    {
+      "epoch": 2.1049228033303695,
+      "grad_norm": 0.6878724098205566,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 13020
+    },
+    {
+      "epoch": 2.1065394875111147,
+      "grad_norm": 0.822318971157074,
+      "learning_rate": 0.0002,
+      "loss": 0.6824,
+      "step": 13030
+    },
+    {
+      "epoch": 2.10815617169186,
+      "grad_norm": 0.831468939781189,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 13040
+    },
+    {
+      "epoch": 2.109772855872605,
+      "grad_norm": 0.7699505686759949,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 13050
+    },
+    {
+      "epoch": 2.1113895400533504,
+      "grad_norm": 0.7559016346931458,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1130062242340957,
+      "grad_norm": 0.6942209601402283,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 13070
+    },
+    {
+      "epoch": 2.1146229084148414,
+      "grad_norm": 0.6098947525024414,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 13080
+    },
+    {
+      "epoch": 2.1162395925955866,
+      "grad_norm": 0.6499016284942627,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 13090
+    },
+    {
+      "epoch": 2.117856276776332,
+      "grad_norm": 0.7719953060150146,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 13100
+    },
+    {
+      "epoch": 2.119472960957077,
+      "grad_norm": 0.6708134412765503,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1210896451378223,
+      "grad_norm": 0.8119585514068604,
+      "learning_rate": 0.0002,
+      "loss": 0.6056,
+      "step": 13120
+    },
+    {
+      "epoch": 2.1227063293185675,
+      "grad_norm": 0.6947157979011536,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 13130
+    },
+    {
+      "epoch": 2.1243230134993127,
+      "grad_norm": 0.8831837773323059,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 13140
+    },
+    {
+      "epoch": 2.1259396976800584,
+      "grad_norm": 0.7266910672187805,
+      "learning_rate": 0.0002,
+      "loss": 0.6997,
+      "step": 13150
+    },
+    {
+      "epoch": 2.1275563818608036,
+      "grad_norm": 0.8864351511001587,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 13160
+    },
+    {
+      "epoch": 2.129173066041549,
+      "grad_norm": 0.8104248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.6762,
+      "step": 13170
+    },
+    {
+      "epoch": 2.130789750222294,
+      "grad_norm": 0.6077079772949219,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 13180
+    },
+    {
+      "epoch": 2.1324064344030393,
+      "grad_norm": 0.6874213814735413,
+      "learning_rate": 0.0002,
+      "loss": 0.6572,
+      "step": 13190
+    },
+    {
+      "epoch": 2.1340231185837846,
+      "grad_norm": 0.7134367823600769,
+      "learning_rate": 0.0002,
+      "loss": 0.642,
+      "step": 13200
+    },
+    {
+      "epoch": 2.13563980276453,
+      "grad_norm": 0.6101235151290894,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 13210
+    },
+    {
+      "epoch": 2.137256486945275,
+      "grad_norm": 0.6042411923408508,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13220
+    },
+    {
+      "epoch": 2.1388731711260207,
+      "grad_norm": 0.914601743221283,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 13230
+    },
+    {
+      "epoch": 2.140489855306766,
+      "grad_norm": 0.7104284167289734,
+      "learning_rate": 0.0002,
+      "loss": 0.6513,
+      "step": 13240
+    },
+    {
+      "epoch": 2.142106539487511,
+      "grad_norm": 0.664395272731781,
+      "learning_rate": 0.0002,
+      "loss": 0.6607,
+      "step": 13250
+    },
+    {
+      "epoch": 2.1437232236682564,
+      "grad_norm": 0.6991241574287415,
+      "learning_rate": 0.0002,
+      "loss": 0.7211,
+      "step": 13260
+    },
+    {
+      "epoch": 2.1453399078490016,
+      "grad_norm": 0.5469560623168945,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 13270
+    },
+    {
+      "epoch": 2.146956592029747,
+      "grad_norm": 0.8454998135566711,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13280
+    },
+    {
+      "epoch": 2.148573276210492,
+      "grad_norm": 0.7088868618011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 13290
+    },
+    {
+      "epoch": 2.1501899603912378,
+      "grad_norm": 0.7002687454223633,
+      "learning_rate": 0.0002,
+      "loss": 0.6835,
+      "step": 13300
+    },
+    {
+      "epoch": 2.151806644571983,
+      "grad_norm": 0.7785214781761169,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 13310
+    },
+    {
+      "epoch": 2.1534233287527282,
+      "grad_norm": 0.8049132227897644,
+      "learning_rate": 0.0002,
+      "loss": 0.67,
+      "step": 13320
+    },
+    {
+      "epoch": 2.1550400129334735,
+      "grad_norm": 0.8062595129013062,
+      "learning_rate": 0.0002,
+      "loss": 0.6495,
+      "step": 13330
+    },
+    {
+      "epoch": 2.1566566971142187,
+      "grad_norm": 0.6208319067955017,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 13340
+    },
+    {
+      "epoch": 2.158273381294964,
+      "grad_norm": 0.7519655823707581,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 13350
+    },
+    {
+      "epoch": 2.159890065475709,
+      "grad_norm": 0.7645747065544128,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 13360
+    },
+    {
+      "epoch": 2.1615067496564544,
+      "grad_norm": 0.6847302913665771,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 13370
+    },
+    {
+      "epoch": 2.1631234338372,
+      "grad_norm": 0.8630441427230835,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 13380
+    },
+    {
+      "epoch": 2.1647401180179453,
+      "grad_norm": 0.7947702407836914,
+      "learning_rate": 0.0002,
+      "loss": 0.6742,
+      "step": 13390
+    },
+    {
+      "epoch": 2.1663568021986905,
+      "grad_norm": 0.6836977005004883,
+      "learning_rate": 0.0002,
+      "loss": 0.7206,
+      "step": 13400
+    },
+    {
+      "epoch": 2.1679734863794358,
+      "grad_norm": 0.7340566515922546,
+      "learning_rate": 0.0002,
+      "loss": 0.6304,
+      "step": 13410
+    },
+    {
+      "epoch": 2.169590170560181,
+      "grad_norm": 0.7075738906860352,
+      "learning_rate": 0.0002,
+      "loss": 0.6528,
+      "step": 13420
+    },
+    {
+      "epoch": 2.1712068547409262,
+      "grad_norm": 0.7080879807472229,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 13430
+    },
+    {
+      "epoch": 2.1728235389216715,
+      "grad_norm": 0.6218613386154175,
+      "learning_rate": 0.0002,
+      "loss": 0.6615,
+      "step": 13440
+    },
+    {
+      "epoch": 2.174440223102417,
+      "grad_norm": 0.8211479187011719,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 13450
+    },
+    {
+      "epoch": 2.1760569072831624,
+      "grad_norm": 0.864466667175293,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 13460
+    },
+    {
+      "epoch": 2.1776735914639076,
+      "grad_norm": 0.7943857908248901,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 13470
+    },
+    {
+      "epoch": 2.179290275644653,
+      "grad_norm": 0.78728187084198,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 13480
+    },
+    {
+      "epoch": 2.180906959825398,
+      "grad_norm": 0.697527289390564,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 13490
+    },
+    {
+      "epoch": 2.1825236440061433,
+      "grad_norm": 0.8205804228782654,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 13500
+    },
+    {
+      "epoch": 2.1841403281868885,
+      "grad_norm": 0.8709042072296143,
+      "learning_rate": 0.0002,
+      "loss": 0.7227,
+      "step": 13510
+    },
+    {
+      "epoch": 2.1857570123676338,
+      "grad_norm": 0.6228537559509277,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 13520
+    },
+    {
+      "epoch": 2.1873736965483794,
+      "grad_norm": 0.9566980004310608,
+      "learning_rate": 0.0002,
+      "loss": 0.7025,
+      "step": 13530
+    },
+    {
+      "epoch": 2.1889903807291247,
+      "grad_norm": 0.7128894329071045,
+      "learning_rate": 0.0002,
+      "loss": 0.6755,
+      "step": 13540
+    },
+    {
+      "epoch": 2.19060706490987,
+      "grad_norm": 0.6888654232025146,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 13550
+    },
+    {
+      "epoch": 2.192223749090615,
+      "grad_norm": 0.6444337368011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6961,
+      "step": 13560
+    },
+    {
+      "epoch": 2.1938404332713604,
+      "grad_norm": 0.8008806705474854,
+      "learning_rate": 0.0002,
+      "loss": 0.656,
+      "step": 13570
+    },
+    {
+      "epoch": 2.1954571174521056,
+      "grad_norm": 0.8482748866081238,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 13580
+    },
+    {
+      "epoch": 2.197073801632851,
+      "grad_norm": 0.8584157228469849,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 13590
+    },
+    {
+      "epoch": 2.1986904858135965,
+      "grad_norm": 0.7513734698295593,
+      "learning_rate": 0.0002,
+      "loss": 0.7014,
+      "step": 13600
+    },
+    {
+      "epoch": 2.2003071699943417,
+      "grad_norm": 0.7864262461662292,
+      "learning_rate": 0.0002,
+      "loss": 0.6632,
+      "step": 13610
+    },
+    {
+      "epoch": 2.201923854175087,
+      "grad_norm": 0.8493645191192627,
+      "learning_rate": 0.0002,
+      "loss": 0.6879,
+      "step": 13620
+    },
+    {
+      "epoch": 2.203540538355832,
+      "grad_norm": 0.6902140974998474,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 13630
+    },
+    {
+      "epoch": 2.2051572225365774,
+      "grad_norm": 0.8711254596710205,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 13640
+    },
+    {
+      "epoch": 2.2067739067173227,
+      "grad_norm": 0.7832191586494446,
+      "learning_rate": 0.0002,
+      "loss": 0.6359,
+      "step": 13650
+    },
+    {
+      "epoch": 2.208390590898068,
+      "grad_norm": 0.5668176412582397,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2100072750788136,
+      "grad_norm": 0.8648375272750854,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13670
+    },
+    {
+      "epoch": 2.211623959259559,
+      "grad_norm": 0.7643089890480042,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13680
+    },
+    {
+      "epoch": 2.213240643440304,
+      "grad_norm": 0.6293777823448181,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13690
+    },
+    {
+      "epoch": 2.2148573276210493,
+      "grad_norm": 0.6459372639656067,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 13700
+    },
+    {
+      "epoch": 2.2164740118017945,
+      "grad_norm": 0.7060744166374207,
+      "learning_rate": 0.0002,
+      "loss": 0.6526,
+      "step": 13710
+    },
+    {
+      "epoch": 2.2180906959825397,
+      "grad_norm": 0.674109160900116,
+      "learning_rate": 0.0002,
+      "loss": 0.7101,
+      "step": 13720
+    },
+    {
+      "epoch": 2.219707380163285,
+      "grad_norm": 0.830392062664032,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13730
+    },
+    {
+      "epoch": 2.2213240643440306,
+      "grad_norm": 0.6474477052688599,
+      "learning_rate": 0.0002,
+      "loss": 0.6733,
+      "step": 13740
+    },
+    {
+      "epoch": 2.222940748524776,
+      "grad_norm": 0.7037909626960754,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13750
+    },
+    {
+      "epoch": 2.224557432705521,
+      "grad_norm": 0.6554131507873535,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 13760
+    },
+    {
+      "epoch": 2.2261741168862663,
+      "grad_norm": 0.7822230458259583,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 13770
+    },
+    {
+      "epoch": 2.2277908010670116,
+      "grad_norm": 0.9082167744636536,
+      "learning_rate": 0.0002,
+      "loss": 0.6505,
+      "step": 13780
+    },
+    {
+      "epoch": 2.229407485247757,
+      "grad_norm": 0.7918276190757751,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 13790
+    },
+    {
+      "epoch": 2.231024169428502,
+      "grad_norm": 0.7354569435119629,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 13800
+    },
+    {
+      "epoch": 2.2326408536092472,
+      "grad_norm": 0.8265249133110046,
+      "learning_rate": 0.0002,
+      "loss": 0.6503,
+      "step": 13810
+    },
+    {
+      "epoch": 2.234257537789993,
+      "grad_norm": 0.6653847098350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6871,
+      "step": 13820
+    },
+    {
+      "epoch": 2.235874221970738,
+      "grad_norm": 0.7157923579216003,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13830
+    },
+    {
+      "epoch": 2.2374909061514834,
+      "grad_norm": 0.7110323309898376,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 13840
+    },
+    {
+      "epoch": 2.2391075903322286,
+      "grad_norm": 0.7155357599258423,
+      "learning_rate": 0.0002,
+      "loss": 0.6913,
+      "step": 13850
+    },
+    {
+      "epoch": 2.240724274512974,
+      "grad_norm": 1.0177817344665527,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 13860
+    },
+    {
+      "epoch": 2.242340958693719,
+      "grad_norm": 0.7601948380470276,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13870
+    },
+    {
+      "epoch": 2.2439576428744643,
+      "grad_norm": 0.7628820538520813,
+      "learning_rate": 0.0002,
+      "loss": 0.6679,
+      "step": 13880
+    },
+    {
+      "epoch": 2.24557432705521,
+      "grad_norm": 0.7089297771453857,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 13890
+    },
+    {
+      "epoch": 2.247191011235955,
+      "grad_norm": 0.695178210735321,
+      "learning_rate": 0.0002,
+      "loss": 0.7236,
+      "step": 13900
+    },
+    {
+      "epoch": 2.2488076954167004,
+      "grad_norm": 0.7631948590278625,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 13910
+    },
+    {
+      "epoch": 2.2504243795974457,
+      "grad_norm": 0.8203101754188538,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 13920
+    },
+    {
+      "epoch": 2.252041063778191,
+      "grad_norm": 0.8099079728126526,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13930
+    },
+    {
+      "epoch": 2.253657747958936,
+      "grad_norm": 0.6498546004295349,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 13940
+    },
+    {
+      "epoch": 2.2552744321396814,
+      "grad_norm": 0.7797415256500244,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 13950
+    },
+    {
+      "epoch": 2.2568911163204266,
+      "grad_norm": 0.8254124522209167,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 13960
+    },
+    {
+      "epoch": 2.2585078005011723,
+      "grad_norm": 0.6327953338623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 13970
+    },
+    {
+      "epoch": 2.2601244846819175,
+      "grad_norm": 0.734194278717041,
+      "learning_rate": 0.0002,
+      "loss": 0.668,
+      "step": 13980
+    },
+    {
+      "epoch": 2.2617411688626627,
+      "grad_norm": 0.9014202952384949,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 13990
+    },
+    {
+      "epoch": 2.263357853043408,
+      "grad_norm": 0.7643631100654602,
+      "learning_rate": 0.0002,
+      "loss": 0.692,
+      "step": 14000
+    },
+    {
+      "epoch": 2.264974537224153,
+      "grad_norm": 0.8882834911346436,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 14010
+    },
+    {
+      "epoch": 2.2665912214048984,
+      "grad_norm": 0.7975873351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14020
+    },
+    {
+      "epoch": 2.2682079055856437,
+      "grad_norm": 0.7765783071517944,
+      "learning_rate": 0.0002,
+      "loss": 0.7193,
+      "step": 14030
+    },
+    {
+      "epoch": 2.2698245897663893,
+      "grad_norm": 0.8846288323402405,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 14040
+    },
+    {
+      "epoch": 2.2714412739471346,
+      "grad_norm": 0.9006744027137756,
+      "learning_rate": 0.0002,
+      "loss": 0.6494,
+      "step": 14050
+    },
+    {
+      "epoch": 2.27305795812788,
+      "grad_norm": 0.7420173287391663,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 14060
+    },
+    {
+      "epoch": 2.274674642308625,
+      "grad_norm": 0.7956424951553345,
+      "learning_rate": 0.0002,
+      "loss": 0.7068,
+      "step": 14070
+    },
+    {
+      "epoch": 2.2762913264893703,
+      "grad_norm": 0.7783209085464478,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 14080
+    },
+    {
+      "epoch": 2.2779080106701155,
+      "grad_norm": 0.7597188949584961,
+      "learning_rate": 0.0002,
+      "loss": 0.7202,
+      "step": 14090
+    },
+    {
+      "epoch": 2.2795246948508607,
+      "grad_norm": 0.6718921661376953,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14100
+    },
+    {
+      "epoch": 2.281141379031606,
+      "grad_norm": 0.7528082132339478,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 14110
+    },
+    {
+      "epoch": 2.2827580632123516,
+      "grad_norm": 0.8379864692687988,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 14120
+    },
+    {
+      "epoch": 2.284374747393097,
+      "grad_norm": 0.748613715171814,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 14130
+    },
+    {
+      "epoch": 2.285991431573842,
+      "grad_norm": 0.7435423135757446,
+      "learning_rate": 0.0002,
+      "loss": 0.6641,
+      "step": 14140
+    },
+    {
+      "epoch": 2.2876081157545873,
+      "grad_norm": 0.7580803632736206,
+      "learning_rate": 0.0002,
+      "loss": 0.6849,
+      "step": 14150
+    },
+    {
+      "epoch": 2.2892247999353326,
+      "grad_norm": 0.6278321146965027,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 14160
+    },
+    {
+      "epoch": 2.290841484116078,
+      "grad_norm": 0.7663896083831787,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 14170
+    },
+    {
+      "epoch": 2.292458168296823,
+      "grad_norm": 0.9716812372207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 14180
+    },
+    {
+      "epoch": 2.2940748524775687,
+      "grad_norm": 0.8993458151817322,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14190
+    },
+    {
+      "epoch": 2.295691536658314,
+      "grad_norm": 0.6156117916107178,
+      "learning_rate": 0.0002,
+      "loss": 0.6172,
+      "step": 14200
+    },
+    {
+      "epoch": 2.297308220839059,
+      "grad_norm": 0.8911278247833252,
+      "learning_rate": 0.0002,
+      "loss": 0.6318,
+      "step": 14210
+    },
+    {
+      "epoch": 2.2989249050198044,
+      "grad_norm": 0.6422147154808044,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 14220
+    },
+    {
+      "epoch": 2.3005415892005496,
+      "grad_norm": 0.6866879463195801,
+      "learning_rate": 0.0002,
+      "loss": 0.6795,
+      "step": 14230
+    },
+    {
+      "epoch": 2.302158273381295,
+      "grad_norm": 0.9297130107879639,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 14240
+    },
+    {
+      "epoch": 2.30377495756204,
+      "grad_norm": 0.7501356601715088,
+      "learning_rate": 0.0002,
+      "loss": 0.6823,
+      "step": 14250
+    },
+    {
+      "epoch": 2.3053916417427853,
+      "grad_norm": 0.8363515138626099,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 14260
+    },
+    {
+      "epoch": 2.307008325923531,
+      "grad_norm": 0.9083868265151978,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3086250101042762,
+      "grad_norm": 0.7791516780853271,
+      "learning_rate": 0.0002,
+      "loss": 0.6862,
+      "step": 14280
+    },
+    {
+      "epoch": 2.3102416942850215,
+      "grad_norm": 0.8766953349113464,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14290
+    },
+    {
+      "epoch": 2.3118583784657667,
+      "grad_norm": 0.7916635274887085,
+      "learning_rate": 0.0002,
+      "loss": 0.6698,
+      "step": 14300
+    },
+    {
+      "epoch": 2.313475062646512,
+      "grad_norm": 0.627525269985199,
+      "learning_rate": 0.0002,
+      "loss": 0.6927,
+      "step": 14310
+    },
+    {
+      "epoch": 2.315091746827257,
+      "grad_norm": 0.8856783509254456,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 14320
+    },
+    {
+      "epoch": 2.316708431008003,
+      "grad_norm": 0.6758689284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 14330
+    },
+    {
+      "epoch": 2.318325115188748,
+      "grad_norm": 0.6428321003913879,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 14340
+    },
+    {
+      "epoch": 2.3199417993694933,
+      "grad_norm": 0.9032121300697327,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 14350
+    },
+    {
+      "epoch": 2.3215584835502385,
+      "grad_norm": 0.8035986423492432,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14360
+    },
+    {
+      "epoch": 2.3231751677309838,
+      "grad_norm": 0.7974579334259033,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14370
+    },
+    {
+      "epoch": 2.324791851911729,
+      "grad_norm": 0.8356034755706787,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 14380
+    },
+    {
+      "epoch": 2.326408536092474,
+      "grad_norm": 0.998760998249054,
+      "learning_rate": 0.0002,
+      "loss": 0.6623,
+      "step": 14390
+    },
+    {
+      "epoch": 2.3280252202732195,
+      "grad_norm": 0.6518142223358154,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 14400
+    },
+    {
+      "epoch": 2.3296419044539647,
+      "grad_norm": 0.7443506717681885,
+      "learning_rate": 0.0002,
+      "loss": 0.7146,
+      "step": 14410
+    },
+    {
+      "epoch": 2.3312585886347104,
+      "grad_norm": 0.8436172604560852,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14420
+    },
+    {
+      "epoch": 2.3328752728154556,
+      "grad_norm": 0.7411080598831177,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 14430
+    },
+    {
+      "epoch": 2.334491956996201,
+      "grad_norm": 0.8839048743247986,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 14440
+    },
+    {
+      "epoch": 2.336108641176946,
+      "grad_norm": 0.8360885977745056,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 14450
+    },
+    {
+      "epoch": 2.3377253253576913,
+      "grad_norm": 0.7608986496925354,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 14460
+    },
+    {
+      "epoch": 2.3393420095384365,
+      "grad_norm": 0.8179867267608643,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14470
+    },
+    {
+      "epoch": 2.340958693719182,
+      "grad_norm": 0.5989999771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14480
+    },
+    {
+      "epoch": 2.3425753778999274,
+      "grad_norm": 0.9450054168701172,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 14490
+    },
+    {
+      "epoch": 2.3441920620806727,
+      "grad_norm": 0.7885149717330933,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 14500
+    },
+    {
+      "epoch": 2.345808746261418,
+      "grad_norm": 0.8152616620063782,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 14510
+    },
+    {
+      "epoch": 2.347425430442163,
+      "grad_norm": 0.7193838953971863,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 14520
+    },
+    {
+      "epoch": 2.3490421146229084,
+      "grad_norm": 0.6701092720031738,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 14530
+    },
+    {
+      "epoch": 2.3506587988036536,
+      "grad_norm": 0.7529364228248596,
+      "learning_rate": 0.0002,
+      "loss": 0.6559,
+      "step": 14540
+    },
+    {
+      "epoch": 2.352275482984399,
+      "grad_norm": 0.6599733829498291,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 14550
+    },
+    {
+      "epoch": 2.353892167165144,
+      "grad_norm": 0.9502474069595337,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14560
+    },
+    {
+      "epoch": 2.3555088513458897,
+      "grad_norm": 0.7619650959968567,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 14570
+    },
+    {
+      "epoch": 2.357125535526635,
+      "grad_norm": 0.9854652285575867,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14580
+    },
+    {
+      "epoch": 2.35874221970738,
+      "grad_norm": 0.727439284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 14590
+    },
+    {
+      "epoch": 2.3603589038881254,
+      "grad_norm": 0.6994746327400208,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 14600
+    },
+    {
+      "epoch": 2.3619755880688706,
+      "grad_norm": 0.7117531299591064,
+      "learning_rate": 0.0002,
+      "loss": 0.6587,
+      "step": 14610
+    },
+    {
+      "epoch": 2.363592272249616,
+      "grad_norm": 0.6403067708015442,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 14620
+    },
+    {
+      "epoch": 2.3652089564303616,
+      "grad_norm": 0.8377841711044312,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14630
+    },
+    {
+      "epoch": 2.366825640611107,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14640
+    },
+    {
+      "epoch": 2.368442324791852,
+      "grad_norm": 0.8418586254119873,
+      "learning_rate": 0.0002,
+      "loss": 0.6552,
+      "step": 14650
+    },
+    {
+      "epoch": 2.3700590089725972,
+      "grad_norm": 0.6178573369979858,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14660
+    },
+    {
+      "epoch": 2.3716756931533425,
+      "grad_norm": 0.6368302702903748,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 14670
+    },
+    {
+      "epoch": 2.3732923773340877,
+      "grad_norm": 0.9122977256774902,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14680
+    },
+    {
+      "epoch": 2.374909061514833,
+      "grad_norm": 0.7086195349693298,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 14690
+    },
+    {
+      "epoch": 2.376525745695578,
+      "grad_norm": 0.7500800490379333,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 14700
+    },
+    {
+      "epoch": 2.378142429876324,
+      "grad_norm": 0.6634900569915771,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 14710
+    },
+    {
+      "epoch": 2.379759114057069,
+      "grad_norm": 0.839898407459259,
+      "learning_rate": 0.0002,
+      "loss": 0.6614,
+      "step": 14720
+    },
+    {
+      "epoch": 2.3813757982378143,
+      "grad_norm": 0.7578426003456116,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14730
+    },
+    {
+      "epoch": 2.3829924824185595,
+      "grad_norm": 1.0213173627853394,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3846091665993048,
+      "grad_norm": 0.7855949401855469,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 14750
+    },
+    {
+      "epoch": 2.38622585078005,
+      "grad_norm": 0.7224128842353821,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 14760
+    },
+    {
+      "epoch": 2.3878425349607952,
+      "grad_norm": 0.8040381669998169,
+      "learning_rate": 0.0002,
+      "loss": 0.7017,
+      "step": 14770
+    },
+    {
+      "epoch": 2.389459219141541,
+      "grad_norm": 0.7705281376838684,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 14780
+    },
+    {
+      "epoch": 2.391075903322286,
+      "grad_norm": 0.667966902256012,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 14790
+    },
+    {
+      "epoch": 2.3926925875030314,
+      "grad_norm": 0.6611011028289795,
+      "learning_rate": 0.0002,
+      "loss": 0.7061,
+      "step": 14800
+    },
+    {
+      "epoch": 2.3943092716837766,
+      "grad_norm": 0.6862651705741882,
+      "learning_rate": 0.0002,
+      "loss": 0.6527,
+      "step": 14810
+    },
+    {
+      "epoch": 2.395925955864522,
+      "grad_norm": 0.8086010217666626,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 14820
+    },
+    {
+      "epoch": 2.397542640045267,
+      "grad_norm": 0.7189689874649048,
+      "learning_rate": 0.0002,
+      "loss": 0.7189,
+      "step": 14830
+    },
+    {
+      "epoch": 2.3991593242260123,
+      "grad_norm": 0.6280009150505066,
+      "learning_rate": 0.0002,
+      "loss": 0.6709,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4007760084067575,
+      "grad_norm": 0.7826612591743469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14850
+    },
+    {
+      "epoch": 2.402392692587503,
+      "grad_norm": 0.7681610584259033,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4040093767682484,
+      "grad_norm": 0.720966100692749,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 14870
+    },
+    {
+      "epoch": 2.4056260609489937,
+      "grad_norm": 0.8202250599861145,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 14880
+    },
+    {
+      "epoch": 2.407242745129739,
+      "grad_norm": 0.786212682723999,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 14890
+    },
+    {
+      "epoch": 2.408859429310484,
+      "grad_norm": 0.6647164821624756,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 14900
+    },
+    {
+      "epoch": 2.4104761134912294,
+      "grad_norm": 0.7566399574279785,
+      "learning_rate": 0.0002,
+      "loss": 0.6706,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4120927976719746,
+      "grad_norm": 0.748814582824707,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 14920
+    },
+    {
+      "epoch": 2.4137094818527203,
+      "grad_norm": 0.7624038457870483,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14930
+    },
+    {
+      "epoch": 2.4153261660334655,
+      "grad_norm": 0.8267335295677185,
+      "learning_rate": 0.0002,
+      "loss": 0.6483,
+      "step": 14940
+    },
+    {
+      "epoch": 2.4169428502142107,
+      "grad_norm": 0.8785360455513,
+      "learning_rate": 0.0002,
+      "loss": 0.6612,
+      "step": 14950
+    },
+    {
+      "epoch": 2.418559534394956,
+      "grad_norm": 0.679887592792511,
+      "learning_rate": 0.0002,
+      "loss": 0.6718,
+      "step": 14960
+    },
+    {
+      "epoch": 2.420176218575701,
+      "grad_norm": 0.7218474745750427,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4217929027564464,
+      "grad_norm": 0.6342799663543701,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14980
+    },
+    {
+      "epoch": 2.4234095869371917,
+      "grad_norm": 0.7098712921142578,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 14990
+    },
+    {
+      "epoch": 2.425026271117937,
+      "grad_norm": 0.7497431635856628,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4266429552986826,
+      "grad_norm": 0.934836208820343,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15010
+    },
+    {
+      "epoch": 2.428259639479428,
+      "grad_norm": 0.8430966734886169,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 15020
+    },
+    {
+      "epoch": 2.429876323660173,
+      "grad_norm": 0.7032104730606079,
+      "learning_rate": 0.0002,
+      "loss": 0.6985,
+      "step": 15030
+    },
+    {
+      "epoch": 2.4314930078409183,
+      "grad_norm": 0.7746111750602722,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 15040
+    },
+    {
+      "epoch": 2.4331096920216635,
+      "grad_norm": 0.7661406397819519,
+      "learning_rate": 0.0002,
+      "loss": 0.7177,
+      "step": 15050
+    },
+    {
+      "epoch": 2.4347263762024087,
+      "grad_norm": 0.6941645741462708,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 15060
+    },
+    {
+      "epoch": 2.436343060383154,
+      "grad_norm": 0.7487249374389648,
+      "learning_rate": 0.0002,
+      "loss": 0.6421,
+      "step": 15070
+    },
+    {
+      "epoch": 2.4379597445638996,
+      "grad_norm": 0.7639912962913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 15080
+    },
+    {
+      "epoch": 2.439576428744645,
+      "grad_norm": 0.7708953619003296,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 15090
+    },
+    {
+      "epoch": 2.44119311292539,
+      "grad_norm": 0.9135832190513611,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15100
+    },
+    {
+      "epoch": 2.4428097971061353,
+      "grad_norm": 0.8283005356788635,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 15110
+    },
+    {
+      "epoch": 2.4444264812868806,
+      "grad_norm": 0.925299346446991,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 15120
+    },
+    {
+      "epoch": 2.446043165467626,
+      "grad_norm": 0.7013528943061829,
+      "learning_rate": 0.0002,
+      "loss": 0.687,
+      "step": 15130
+    },
+    {
+      "epoch": 2.447659849648371,
+      "grad_norm": 0.622303307056427,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 15140
+    },
+    {
+      "epoch": 2.4492765338291163,
+      "grad_norm": 0.876569390296936,
+      "learning_rate": 0.0002,
+      "loss": 0.6676,
+      "step": 15150
+    },
+    {
+      "epoch": 2.450893218009862,
+      "grad_norm": 0.6836351752281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 15160
+    },
+    {
+      "epoch": 2.452509902190607,
+      "grad_norm": 0.7886684536933899,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 15170
+    },
+    {
+      "epoch": 2.4541265863713524,
+      "grad_norm": 0.6647440791130066,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 15180
+    },
+    {
+      "epoch": 2.4557432705520976,
+      "grad_norm": 0.7477722764015198,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 15190
+    },
+    {
+      "epoch": 2.457359954732843,
+      "grad_norm": 0.8192033767700195,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 15200
+    },
+    {
+      "epoch": 2.458976638913588,
+      "grad_norm": 0.847537100315094,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 15210
+    },
+    {
+      "epoch": 2.4605933230943338,
+      "grad_norm": 0.9027776122093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 15220
+    },
+    {
+      "epoch": 2.462210007275079,
+      "grad_norm": 0.7217772006988525,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 15230
+    },
+    {
+      "epoch": 2.4638266914558242,
+      "grad_norm": 0.7994546294212341,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 15240
+    },
+    {
+      "epoch": 2.4654433756365695,
+      "grad_norm": 0.939916729927063,
+      "learning_rate": 0.0002,
+      "loss": 0.6828,
+      "step": 15250
+    },
+    {
+      "epoch": 2.4670600598173147,
+      "grad_norm": 1.0009053945541382,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15260
+    },
+    {
+      "epoch": 2.46867674399806,
+      "grad_norm": 0.625555694103241,
+      "learning_rate": 0.0002,
+      "loss": 0.643,
+      "step": 15270
+    },
+    {
+      "epoch": 2.470293428178805,
+      "grad_norm": 0.7924878597259521,
+      "learning_rate": 0.0002,
+      "loss": 0.688,
+      "step": 15280
+    },
+    {
+      "epoch": 2.4719101123595504,
+      "grad_norm": 0.8536689877510071,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 15290
+    },
+    {
+      "epoch": 2.4735267965402956,
+      "grad_norm": 0.8572589755058289,
+      "learning_rate": 0.0002,
+      "loss": 0.6924,
+      "step": 15300
+    },
+    {
+      "epoch": 2.4751434807210413,
+      "grad_norm": 0.773279070854187,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 15310
+    },
+    {
+      "epoch": 2.4767601649017865,
+      "grad_norm": 0.7708749771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 15320
+    },
+    {
+      "epoch": 2.4783768490825318,
+      "grad_norm": 0.770905077457428,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15330
+    },
+    {
+      "epoch": 2.479993533263277,
+      "grad_norm": 0.8238571882247925,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 15340
+    },
+    {
+      "epoch": 2.481610217444022,
+      "grad_norm": 0.7670477032661438,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15350
+    },
+    {
+      "epoch": 2.4832269016247674,
+      "grad_norm": 0.905036985874176,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 15360
+    },
+    {
+      "epoch": 2.484843585805513,
+      "grad_norm": 0.6672089695930481,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 15370
+    },
+    {
+      "epoch": 2.4864602699862584,
+      "grad_norm": 0.625095784664154,
+      "learning_rate": 0.0002,
+      "loss": 0.6722,
+      "step": 15380
+    },
+    {
+      "epoch": 2.4880769541670036,
+      "grad_norm": 0.679772675037384,
+      "learning_rate": 0.0002,
+      "loss": 0.6396,
+      "step": 15390
+    },
+    {
+      "epoch": 2.489693638347749,
+      "grad_norm": 0.711492121219635,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 15400
+    },
+    {
+      "epoch": 2.491310322528494,
+      "grad_norm": 0.876189112663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6966,
+      "step": 15410
+    },
+    {
+      "epoch": 2.4929270067092393,
+      "grad_norm": 0.7236915230751038,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 15420
+    },
+    {
+      "epoch": 2.4945436908899845,
+      "grad_norm": 0.6629832983016968,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 15430
+    },
+    {
+      "epoch": 2.4961603750707297,
+      "grad_norm": 0.9756859540939331,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 15440
+    },
+    {
+      "epoch": 2.4977770592514754,
+      "grad_norm": 0.6896940469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 15450
+    },
+    {
+      "epoch": 2.4993937434322206,
+      "grad_norm": 0.7105149626731873,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 15460
+    },
+    {
+      "epoch": 2.501010427612966,
+      "grad_norm": 0.8374546766281128,
+      "learning_rate": 0.0002,
+      "loss": 0.6859,
+      "step": 15470
+    },
+    {
+      "epoch": 2.502627111793711,
+      "grad_norm": 0.7320070266723633,
+      "learning_rate": 0.0002,
+      "loss": 0.6512,
+      "step": 15480
+    },
+    {
+      "epoch": 2.5042437959744563,
+      "grad_norm": 0.8306367993354797,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 15490
+    },
+    {
+      "epoch": 2.5058604801552016,
+      "grad_norm": 0.7472721338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.7253,
+      "step": 15500
+    },
+    {
+      "epoch": 2.507477164335947,
+      "grad_norm": 0.6147692203521729,
+      "learning_rate": 0.0002,
+      "loss": 0.6699,
+      "step": 15510
+    },
+    {
+      "epoch": 2.5090938485166925,
+      "grad_norm": 0.7788505554199219,
+      "learning_rate": 0.0002,
+      "loss": 0.7158,
+      "step": 15520
+    },
+    {
+      "epoch": 2.5107105326974377,
+      "grad_norm": 0.8807527422904968,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 15530
+    },
+    {
+      "epoch": 2.512327216878183,
+      "grad_norm": 0.7521643042564392,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 15540
+    },
+    {
+      "epoch": 2.513943901058928,
+      "grad_norm": 0.6900225281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15550
+    },
+    {
+      "epoch": 2.5155605852396734,
+      "grad_norm": 0.6601938605308533,
+      "learning_rate": 0.0002,
+      "loss": 0.6769,
+      "step": 15560
+    },
+    {
+      "epoch": 2.5171772694204186,
+      "grad_norm": 0.8179984092712402,
+      "learning_rate": 0.0002,
+      "loss": 0.6648,
+      "step": 15570
+    },
+    {
+      "epoch": 2.518793953601164,
+      "grad_norm": 0.792556881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.7028,
+      "step": 15580
+    },
+    {
+      "epoch": 2.520410637781909,
+      "grad_norm": 0.7081938982009888,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 15590
+    },
+    {
+      "epoch": 2.5220273219626543,
+      "grad_norm": 0.8733121156692505,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 15600
+    },
+    {
+      "epoch": 2.5236440061434,
+      "grad_norm": 0.7980992794036865,
+      "learning_rate": 0.0002,
+      "loss": 0.6969,
+      "step": 15610
+    },
+    {
+      "epoch": 2.5252606903241452,
+      "grad_norm": 0.883664071559906,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 15620
+    },
+    {
+      "epoch": 2.5268773745048905,
+      "grad_norm": 0.6963341236114502,
+      "learning_rate": 0.0002,
+      "loss": 0.7022,
+      "step": 15630
+    },
+    {
+      "epoch": 2.5284940586856357,
+      "grad_norm": 0.6433573365211487,
+      "learning_rate": 0.0002,
+      "loss": 0.7334,
+      "step": 15640
+    },
+    {
+      "epoch": 2.530110742866381,
+      "grad_norm": 0.8538183569908142,
+      "learning_rate": 0.0002,
+      "loss": 0.6889,
+      "step": 15650
+    },
+    {
+      "epoch": 2.5317274270471266,
+      "grad_norm": 0.9748201370239258,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 15660
+    },
+    {
+      "epoch": 2.533344111227872,
+      "grad_norm": 0.7670575380325317,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 15670
+    },
+    {
+      "epoch": 2.534960795408617,
+      "grad_norm": 0.8738890290260315,
+      "learning_rate": 0.0002,
+      "loss": 0.6435,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5365774795893623,
+      "grad_norm": 0.8391636610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.6802,
+      "step": 15690
+    },
+    {
+      "epoch": 2.5381941637701075,
+      "grad_norm": 0.7239366769790649,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 15700
+    },
+    {
+      "epoch": 2.5398108479508528,
+      "grad_norm": 0.8498379588127136,
+      "learning_rate": 0.0002,
+      "loss": 0.7011,
+      "step": 15710
+    },
+    {
+      "epoch": 2.541427532131598,
+      "grad_norm": 0.8029484152793884,
+      "learning_rate": 0.0002,
+      "loss": 0.6998,
+      "step": 15720
+    },
+    {
+      "epoch": 2.5430442163123432,
+      "grad_norm": 1.0639333724975586,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 15730
+    },
+    {
+      "epoch": 2.5446609004930885,
+      "grad_norm": 0.6401297450065613,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 15740
+    },
+    {
+      "epoch": 2.5462775846738337,
+      "grad_norm": 0.7123814821243286,
+      "learning_rate": 0.0002,
+      "loss": 0.7196,
+      "step": 15750
+    },
+    {
+      "epoch": 2.5478942688545794,
+      "grad_norm": 0.7874974608421326,
+      "learning_rate": 0.0002,
+      "loss": 0.654,
+      "step": 15760
+    },
+    {
+      "epoch": 2.5495109530353246,
+      "grad_norm": 0.8046808838844299,
+      "learning_rate": 0.0002,
+      "loss": 0.6721,
+      "step": 15770
+    },
+    {
+      "epoch": 2.55112763721607,
+      "grad_norm": 0.7888661623001099,
+      "learning_rate": 0.0002,
+      "loss": 0.6665,
+      "step": 15780
+    },
+    {
+      "epoch": 2.552744321396815,
+      "grad_norm": 0.8445866107940674,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15790
+    },
+    {
+      "epoch": 2.5543610055775603,
+      "grad_norm": 0.7475846409797668,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 15800
+    },
+    {
+      "epoch": 2.555977689758306,
+      "grad_norm": 0.7455102801322937,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 15810
+    },
+    {
+      "epoch": 2.557594373939051,
+      "grad_norm": 0.8226983547210693,
+      "learning_rate": 0.0002,
+      "loss": 0.6932,
+      "step": 15820
+    },
+    {
+      "epoch": 2.5592110581197964,
+      "grad_norm": 0.8920368552207947,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 15830
+    },
+    {
+      "epoch": 2.5608277423005417,
+      "grad_norm": 0.8413904905319214,
+      "learning_rate": 0.0002,
+      "loss": 0.6297,
+      "step": 15840
+    },
+    {
+      "epoch": 2.562444426481287,
+      "grad_norm": 0.8483649492263794,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 15850
+    },
+    {
+      "epoch": 2.564061110662032,
+      "grad_norm": 0.5923284292221069,
+      "learning_rate": 0.0002,
+      "loss": 0.6957,
+      "step": 15860
+    },
+    {
+      "epoch": 2.5656777948427774,
+      "grad_norm": 0.8518726229667664,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 15870
+    },
+    {
+      "epoch": 2.5672944790235226,
+      "grad_norm": 0.731235146522522,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 15880
+    },
+    {
+      "epoch": 2.568911163204268,
+      "grad_norm": 0.7517194151878357,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 15890
+    },
+    {
+      "epoch": 2.5705278473850135,
+      "grad_norm": 0.8378692269325256,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 15900
+    },
+    {
+      "epoch": 2.5721445315657587,
+      "grad_norm": 0.843701958656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 15910
+    },
+    {
+      "epoch": 2.573761215746504,
+      "grad_norm": 0.7254629731178284,
+      "learning_rate": 0.0002,
+      "loss": 0.6882,
+      "step": 15920
+    },
+    {
+      "epoch": 2.575377899927249,
+      "grad_norm": 0.8863335847854614,
+      "learning_rate": 0.0002,
+      "loss": 0.6872,
+      "step": 15930
+    },
+    {
+      "epoch": 2.5769945841079944,
+      "grad_norm": 0.7675097584724426,
+      "learning_rate": 0.0002,
+      "loss": 0.6813,
+      "step": 15940
+    },
+    {
+      "epoch": 2.5786112682887397,
+      "grad_norm": 0.82063889503479,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 15950
+    },
+    {
+      "epoch": 2.5802279524694853,
+      "grad_norm": 0.7729717493057251,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 15960
+    },
+    {
+      "epoch": 2.5818446366502306,
+      "grad_norm": 0.8301846981048584,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 15970
+    },
+    {
+      "epoch": 2.583461320830976,
+      "grad_norm": 0.7906861305236816,
+      "learning_rate": 0.0002,
+      "loss": 0.6897,
+      "step": 15980
+    },
+    {
+      "epoch": 2.585078005011721,
+      "grad_norm": 0.6749057173728943,
+      "learning_rate": 0.0002,
+      "loss": 0.7175,
+      "step": 15990
+    },
+    {
+      "epoch": 2.5866946891924663,
+      "grad_norm": 0.9386842846870422,
+      "learning_rate": 0.0002,
+      "loss": 0.7212,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5883113733732115,
+      "grad_norm": 0.7868891358375549,
+      "learning_rate": 0.0002,
+      "loss": 0.6934,
+      "step": 16010
+    },
+    {
+      "epoch": 2.5899280575539567,
+      "grad_norm": 0.8674671053886414,
+      "learning_rate": 0.0002,
+      "loss": 0.7036,
+      "step": 16020
+    },
+    {
+      "epoch": 2.591544741734702,
+      "grad_norm": 0.7043559551239014,
+      "learning_rate": 0.0002,
+      "loss": 0.7217,
+      "step": 16030
+    },
+    {
+      "epoch": 2.593161425915447,
+      "grad_norm": 0.5846083760261536,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 16040
+    },
+    {
+      "epoch": 2.594778110096193,
+      "grad_norm": 0.7323982119560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 16050
+    },
+    {
+      "epoch": 2.596394794276938,
+      "grad_norm": 0.9069556593894958,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 16060
+    },
+    {
+      "epoch": 2.5980114784576833,
+      "grad_norm": 0.7522736191749573,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 16070
+    },
+    {
+      "epoch": 2.5996281626384286,
+      "grad_norm": 0.8149648308753967,
+      "learning_rate": 0.0002,
+      "loss": 0.6477,
+      "step": 16080
+    },
+    {
+      "epoch": 2.601244846819174,
+      "grad_norm": 0.6214233040809631,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 16090
+    },
+    {
+      "epoch": 2.602861530999919,
+      "grad_norm": 0.6803743839263916,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 16100
+    },
+    {
+      "epoch": 2.6044782151806647,
+      "grad_norm": 0.7223997116088867,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 16110
+    },
+    {
+      "epoch": 2.60609489936141,
+      "grad_norm": 0.7324174642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.6867,
+      "step": 16120
+    },
+    {
+      "epoch": 2.607711583542155,
+      "grad_norm": 0.9594739675521851,
+      "learning_rate": 0.0002,
+      "loss": 0.7159,
+      "step": 16130
+    },
+    {
+      "epoch": 2.6093282677229004,
+      "grad_norm": 0.9485327005386353,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 16140
+    },
+    {
+      "epoch": 2.6109449519036456,
+      "grad_norm": 0.8449000120162964,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 16150
+    },
+    {
+      "epoch": 2.612561636084391,
+      "grad_norm": 0.8520140051841736,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 16160
+    },
+    {
+      "epoch": 2.614178320265136,
+      "grad_norm": 0.7456524968147278,
+      "learning_rate": 0.0002,
+      "loss": 0.6759,
+      "step": 16170
+    },
+    {
+      "epoch": 2.6157950044458813,
+      "grad_norm": 0.9912857413291931,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 16180
+    },
+    {
+      "epoch": 2.6174116886266265,
+      "grad_norm": 0.9001946449279785,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 16190
+    },
+    {
+      "epoch": 2.619028372807372,
+      "grad_norm": 0.6568667888641357,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16200
+    },
+    {
+      "epoch": 2.6206450569881174,
+      "grad_norm": 1.0248128175735474,
+      "learning_rate": 0.0002,
+      "loss": 0.7013,
+      "step": 16210
+    },
+    {
+      "epoch": 2.6222617411688627,
+      "grad_norm": 0.6509039998054504,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 16220
+    },
+    {
+      "epoch": 2.623878425349608,
+      "grad_norm": 0.7626351118087769,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 16230
+    },
+    {
+      "epoch": 2.625495109530353,
+      "grad_norm": 0.6938552260398865,
+      "learning_rate": 0.0002,
+      "loss": 0.6556,
+      "step": 16240
+    },
+    {
+      "epoch": 2.6271117937110984,
+      "grad_norm": 0.6434680819511414,
+      "learning_rate": 0.0002,
+      "loss": 0.65,
+      "step": 16250
+    },
+    {
+      "epoch": 2.628728477891844,
+      "grad_norm": 0.7111515998840332,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 16260
+    },
+    {
+      "epoch": 2.6303451620725893,
+      "grad_norm": 0.7712395787239075,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 16270
+    },
+    {
+      "epoch": 2.6319618462533345,
+      "grad_norm": 0.792209267616272,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 16280
+    },
+    {
+      "epoch": 2.6335785304340797,
+      "grad_norm": 0.6801066398620605,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 16290
+    },
+    {
+      "epoch": 2.635195214614825,
+      "grad_norm": 0.7802573442459106,
+      "learning_rate": 0.0002,
+      "loss": 0.73,
+      "step": 16300
+    },
+    {
+      "epoch": 2.63681189879557,
+      "grad_norm": 0.7742244601249695,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 16310
+    },
+    {
+      "epoch": 2.6384285829763154,
+      "grad_norm": 0.664184033870697,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 16320
+    },
+    {
+      "epoch": 2.6400452671570607,
+      "grad_norm": 0.9242228865623474,
+      "learning_rate": 0.0002,
+      "loss": 0.6442,
+      "step": 16330
+    },
+    {
+      "epoch": 2.641661951337806,
+      "grad_norm": 0.9661325216293335,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 16340
+    },
+    {
+      "epoch": 2.6432786355185516,
+      "grad_norm": 0.837526798248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 16350
+    },
+    {
+      "epoch": 2.644895319699297,
+      "grad_norm": 1.1834373474121094,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 16360
+    },
+    {
+      "epoch": 2.646512003880042,
+      "grad_norm": 0.7467831373214722,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 16370
+    },
+    {
+      "epoch": 2.6481286880607873,
+      "grad_norm": 0.8627146482467651,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 16380
+    },
+    {
+      "epoch": 2.6497453722415325,
+      "grad_norm": 0.790447473526001,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 16390
+    },
+    {
+      "epoch": 2.651362056422278,
+      "grad_norm": 0.8447365164756775,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 16400
+    },
+    {
+      "epoch": 2.6529787406030234,
+      "grad_norm": 0.7831417918205261,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 16410
+    },
+    {
+      "epoch": 2.6545954247837686,
+      "grad_norm": 0.6837952136993408,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 16420
+    },
+    {
+      "epoch": 2.656212108964514,
+      "grad_norm": 0.7031801342964172,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 16430
+    },
+    {
+      "epoch": 2.657828793145259,
+      "grad_norm": 0.8963770866394043,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 16440
+    },
+    {
+      "epoch": 2.6594454773260043,
+      "grad_norm": 0.6852328181266785,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 16450
+    },
+    {
+      "epoch": 2.6610621615067496,
+      "grad_norm": 0.8069294095039368,
+      "learning_rate": 0.0002,
+      "loss": 0.6278,
+      "step": 16460
+    },
+    {
+      "epoch": 2.662678845687495,
+      "grad_norm": 0.7503686547279358,
+      "learning_rate": 0.0002,
+      "loss": 0.6939,
+      "step": 16470
+    },
+    {
+      "epoch": 2.66429552986824,
+      "grad_norm": 0.6430956125259399,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16480
+    },
+    {
+      "epoch": 2.6659122140489853,
+      "grad_norm": 0.7894312739372253,
+      "learning_rate": 0.0002,
+      "loss": 0.6863,
+      "step": 16490
+    },
+    {
+      "epoch": 2.667528898229731,
+      "grad_norm": 0.7277431488037109,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 16500
+    },
+    {
+      "epoch": 2.669145582410476,
+      "grad_norm": 0.6816153526306152,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 16510
+    },
+    {
+      "epoch": 2.6707622665912214,
+      "grad_norm": 0.8145235776901245,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 16520
+    },
+    {
+      "epoch": 2.6723789507719666,
+      "grad_norm": 0.8645890355110168,
+      "learning_rate": 0.0002,
+      "loss": 0.709,
+      "step": 16530
+    },
+    {
+      "epoch": 2.673995634952712,
+      "grad_norm": 0.704393208026886,
+      "learning_rate": 0.0002,
+      "loss": 0.6946,
+      "step": 16540
+    },
+    {
+      "epoch": 2.6756123191334575,
+      "grad_norm": 1.0120846033096313,
+      "learning_rate": 0.0002,
+      "loss": 0.6378,
+      "step": 16550
+    },
+    {
+      "epoch": 2.6772290033142028,
+      "grad_norm": 0.6919328570365906,
+      "learning_rate": 0.0002,
+      "loss": 0.7241,
+      "step": 16560
+    },
+    {
+      "epoch": 2.678845687494948,
+      "grad_norm": 0.6924574971199036,
+      "learning_rate": 0.0002,
+      "loss": 0.7098,
+      "step": 16570
+    },
+    {
+      "epoch": 2.6804623716756932,
+      "grad_norm": 0.9679301381111145,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 16580
+    },
+    {
+      "epoch": 2.6820790558564385,
+      "grad_norm": 0.6810211539268494,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 16590
+    },
+    {
+      "epoch": 2.6836957400371837,
+      "grad_norm": 0.9730555415153503,
+      "learning_rate": 0.0002,
+      "loss": 0.6688,
+      "step": 16600
+    },
+    {
+      "epoch": 2.685312424217929,
+      "grad_norm": 0.7852821350097656,
+      "learning_rate": 0.0002,
+      "loss": 0.7344,
+      "step": 16610
+    },
+    {
+      "epoch": 2.686929108398674,
+      "grad_norm": 0.6059057116508484,
+      "learning_rate": 0.0002,
+      "loss": 0.6401,
+      "step": 16620
+    },
+    {
+      "epoch": 2.6885457925794194,
+      "grad_norm": 0.9395958781242371,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 16630
+    },
+    {
+      "epoch": 2.690162476760165,
+      "grad_norm": 0.7473729848861694,
+      "learning_rate": 0.0002,
+      "loss": 0.7174,
+      "step": 16640
+    },
+    {
+      "epoch": 2.6917791609409103,
+      "grad_norm": 0.765934407711029,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 16650
+    },
+    {
+      "epoch": 2.6933958451216555,
+      "grad_norm": 0.8496677279472351,
+      "learning_rate": 0.0002,
+      "loss": 0.707,
+      "step": 16660
+    },
+    {
+      "epoch": 2.6950125293024008,
+      "grad_norm": 0.7641879916191101,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 16670
+    },
+    {
+      "epoch": 2.696629213483146,
+      "grad_norm": 0.8471952676773071,
+      "learning_rate": 0.0002,
+      "loss": 0.6566,
+      "step": 16680
+    },
+    {
+      "epoch": 2.6982458976638912,
+      "grad_norm": 0.6946060657501221,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 16690
+    },
+    {
+      "epoch": 2.699862581844637,
+      "grad_norm": 0.7361312508583069,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 16700
+    },
+    {
+      "epoch": 2.701479266025382,
+      "grad_norm": 0.6605038046836853,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 16710
+    },
+    {
+      "epoch": 2.7030959502061274,
+      "grad_norm": 0.7164411544799805,
+      "learning_rate": 0.0002,
+      "loss": 0.6885,
+      "step": 16720
+    },
+    {
+      "epoch": 2.7047126343868726,
+      "grad_norm": 0.6496201157569885,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 16730
+    },
+    {
+      "epoch": 2.706329318567618,
+      "grad_norm": 0.7826663851737976,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 16740
+    },
+    {
+      "epoch": 2.707946002748363,
+      "grad_norm": 0.7639131546020508,
+      "learning_rate": 0.0002,
+      "loss": 0.6773,
+      "step": 16750
+    },
+    {
+      "epoch": 2.7095626869291083,
+      "grad_norm": 0.7976210713386536,
+      "learning_rate": 0.0002,
+      "loss": 0.69,
+      "step": 16760
+    },
+    {
+      "epoch": 2.7111793711098535,
+      "grad_norm": 0.6836577653884888,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 16770
+    },
+    {
+      "epoch": 2.7127960552905988,
+      "grad_norm": 0.8025202751159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 16780
+    },
+    {
+      "epoch": 2.7144127394713444,
+      "grad_norm": 0.7636463642120361,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7160294236520897,
+      "grad_norm": 0.7481677532196045,
+      "learning_rate": 0.0002,
+      "loss": 0.6227,
+      "step": 16800
+    },
+    {
+      "epoch": 2.717646107832835,
+      "grad_norm": 0.7566834688186646,
+      "learning_rate": 0.0002,
+      "loss": 0.6925,
+      "step": 16810
+    },
+    {
+      "epoch": 2.71926279201358,
+      "grad_norm": 0.7931267619132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6531,
+      "step": 16820
+    },
+    {
+      "epoch": 2.7208794761943254,
+      "grad_norm": 0.8811662197113037,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 16830
+    },
+    {
+      "epoch": 2.7224961603750706,
+      "grad_norm": 0.8561240434646606,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 16840
+    },
+    {
+      "epoch": 2.7241128445558163,
+      "grad_norm": 0.7121599316596985,
+      "learning_rate": 0.0002,
+      "loss": 0.7135,
+      "step": 16850
+    },
+    {
+      "epoch": 2.7257295287365615,
+      "grad_norm": 0.8066257238388062,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16860
+    },
+    {
+      "epoch": 2.7273462129173067,
+      "grad_norm": 0.7699271440505981,
+      "learning_rate": 0.0002,
+      "loss": 0.6839,
+      "step": 16870
+    },
+    {
+      "epoch": 2.728962897098052,
+      "grad_norm": 1.1828432083129883,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 16880
+    },
+    {
+      "epoch": 2.730579581278797,
+      "grad_norm": 0.9989302754402161,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 16890
+    },
+    {
+      "epoch": 2.7321962654595424,
+      "grad_norm": 0.8100560307502747,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 16900
+    },
+    {
+      "epoch": 2.7338129496402876,
+      "grad_norm": 0.8615233898162842,
+      "learning_rate": 0.0002,
+      "loss": 0.6851,
+      "step": 16910
+    },
+    {
+      "epoch": 2.735429633821033,
+      "grad_norm": 0.8633756041526794,
+      "learning_rate": 0.0002,
+      "loss": 0.6322,
+      "step": 16920
+    },
+    {
+      "epoch": 2.737046318001778,
+      "grad_norm": 0.7769348621368408,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 16930
+    },
+    {
+      "epoch": 2.738663002182524,
+      "grad_norm": 0.6943058371543884,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 16940
+    },
+    {
+      "epoch": 2.740279686363269,
+      "grad_norm": 0.8510736227035522,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 16950
+    },
+    {
+      "epoch": 2.7418963705440142,
+      "grad_norm": 0.7732602953910828,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 16960
+    },
+    {
+      "epoch": 2.7435130547247595,
+      "grad_norm": 0.5981788635253906,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 16970
+    },
+    {
+      "epoch": 2.7451297389055047,
+      "grad_norm": 0.7604416012763977,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16980
+    },
+    {
+      "epoch": 2.74674642308625,
+      "grad_norm": 0.7377738356590271,
+      "learning_rate": 0.0002,
+      "loss": 0.6981,
+      "step": 16990
+    },
+    {
+      "epoch": 2.7483631072669956,
+      "grad_norm": 0.9400289058685303,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 17000
+    },
+    {
+      "epoch": 2.749979791447741,
+      "grad_norm": 0.6340599656105042,
+      "learning_rate": 0.0002,
+      "loss": 0.6952,
+      "step": 17010
+    },
+    {
+      "epoch": 2.751596475628486,
+      "grad_norm": 0.7297601103782654,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 17020
+    },
+    {
+      "epoch": 2.7532131598092313,
+      "grad_norm": 0.9479979872703552,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 17030
+    },
+    {
+      "epoch": 2.7548298439899765,
+      "grad_norm": 0.8461511135101318,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 17040
+    },
+    {
+      "epoch": 2.7564465281707218,
+      "grad_norm": 0.7477551698684692,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17050
+    },
+    {
+      "epoch": 2.758063212351467,
+      "grad_norm": 1.019270420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 17060
+    },
+    {
+      "epoch": 2.7596798965322122,
+      "grad_norm": 0.7730235457420349,
+      "learning_rate": 0.0002,
+      "loss": 0.7119,
+      "step": 17070
+    },
+    {
+      "epoch": 2.7612965807129575,
+      "grad_norm": 0.8216866254806519,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 17080
+    },
+    {
+      "epoch": 2.762913264893703,
+      "grad_norm": 0.7235931754112244,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17090
+    },
+    {
+      "epoch": 2.7645299490744484,
+      "grad_norm": 0.7352296710014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7031,
+      "step": 17100
+    },
+    {
+      "epoch": 2.7661466332551936,
+      "grad_norm": 0.8129373788833618,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 17110
+    },
+    {
+      "epoch": 2.767763317435939,
+      "grad_norm": 0.7387019991874695,
+      "learning_rate": 0.0002,
+      "loss": 0.6703,
+      "step": 17120
+    },
+    {
+      "epoch": 2.769380001616684,
+      "grad_norm": 0.9149190187454224,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 17130
+    },
+    {
+      "epoch": 2.7709966857974297,
+      "grad_norm": 0.7352971434593201,
+      "learning_rate": 0.0002,
+      "loss": 0.6038,
+      "step": 17140
+    },
+    {
+      "epoch": 2.772613369978175,
+      "grad_norm": 0.7903780341148376,
+      "learning_rate": 0.0002,
+      "loss": 0.6728,
+      "step": 17150
+    },
+    {
+      "epoch": 2.77423005415892,
+      "grad_norm": 0.8255927562713623,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17160
+    },
+    {
+      "epoch": 2.7758467383396654,
+      "grad_norm": 0.7235927581787109,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 17170
+    },
+    {
+      "epoch": 2.7774634225204107,
+      "grad_norm": 0.8281434774398804,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 17180
+    },
+    {
+      "epoch": 2.779080106701156,
+      "grad_norm": 0.7586921453475952,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 17190
+    },
+    {
+      "epoch": 2.780696790881901,
+      "grad_norm": 0.7161715030670166,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 17200
+    },
+    {
+      "epoch": 2.7823134750626464,
+      "grad_norm": 0.762868344783783,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 17210
+    },
+    {
+      "epoch": 2.7839301592433916,
+      "grad_norm": 0.9285483360290527,
+      "learning_rate": 0.0002,
+      "loss": 0.705,
+      "step": 17220
+    },
+    {
+      "epoch": 2.785546843424137,
+      "grad_norm": 0.6900462508201599,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 17230
+    },
+    {
+      "epoch": 2.7871635276048825,
+      "grad_norm": 0.780384361743927,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17240
+    },
+    {
+      "epoch": 2.7887802117856277,
+      "grad_norm": 0.7580406665802002,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17250
+    },
+    {
+      "epoch": 2.790396895966373,
+      "grad_norm": 0.8145199418067932,
+      "learning_rate": 0.0002,
+      "loss": 0.6833,
+      "step": 17260
+    },
+    {
+      "epoch": 2.792013580147118,
+      "grad_norm": 0.9159596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 17270
+    },
+    {
+      "epoch": 2.7936302643278634,
+      "grad_norm": 0.9590014219284058,
+      "learning_rate": 0.0002,
+      "loss": 0.6008,
+      "step": 17280
+    },
+    {
+      "epoch": 2.795246948508609,
+      "grad_norm": 0.7603529691696167,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 17290
+    },
+    {
+      "epoch": 2.7968636326893543,
+      "grad_norm": 0.8039976358413696,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 17300
+    },
+    {
+      "epoch": 2.7984803168700996,
+      "grad_norm": 0.8364847302436829,
+      "learning_rate": 0.0002,
+      "loss": 0.7037,
+      "step": 17310
+    },
+    {
+      "epoch": 2.800097001050845,
+      "grad_norm": 0.8763046860694885,
+      "learning_rate": 0.0002,
+      "loss": 0.6749,
+      "step": 17320
+    },
+    {
+      "epoch": 2.80171368523159,
+      "grad_norm": 0.8409647941589355,
+      "learning_rate": 0.0002,
+      "loss": 0.6844,
+      "step": 17330
+    },
+    {
+      "epoch": 2.8033303694123353,
+      "grad_norm": 0.7649006247520447,
+      "learning_rate": 0.0002,
+      "loss": 0.6936,
+      "step": 17340
+    },
+    {
+      "epoch": 2.8049470535930805,
+      "grad_norm": 0.7970262169837952,
+      "learning_rate": 0.0002,
+      "loss": 0.7051,
+      "step": 17350
+    },
+    {
+      "epoch": 2.8065637377738257,
+      "grad_norm": 0.9088607430458069,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 17360
+    },
+    {
+      "epoch": 2.808180421954571,
+      "grad_norm": 0.6454846858978271,
+      "learning_rate": 0.0002,
+      "loss": 0.675,
+      "step": 17370
+    },
+    {
+      "epoch": 2.809797106135316,
+      "grad_norm": 0.7744787931442261,
+      "learning_rate": 0.0002,
+      "loss": 0.7069,
+      "step": 17380
+    },
+    {
+      "epoch": 2.811413790316062,
+      "grad_norm": 0.6678640842437744,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 17390
+    },
+    {
+      "epoch": 2.813030474496807,
+      "grad_norm": 0.772676944732666,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 17400
+    },
+    {
+      "epoch": 2.8146471586775523,
+      "grad_norm": 0.7088175415992737,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 17410
+    },
+    {
+      "epoch": 2.8162638428582976,
+      "grad_norm": 0.8280573487281799,
+      "learning_rate": 0.0002,
+      "loss": 0.7086,
+      "step": 17420
+    },
+    {
+      "epoch": 2.817880527039043,
+      "grad_norm": 0.6665388345718384,
+      "learning_rate": 0.0002,
+      "loss": 0.6732,
+      "step": 17430
+    },
+    {
+      "epoch": 2.8194972112197885,
+      "grad_norm": 0.6427883505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 17440
+    },
+    {
+      "epoch": 2.8211138954005337,
+      "grad_norm": 0.9697760343551636,
+      "learning_rate": 0.0002,
+      "loss": 0.6972,
+      "step": 17450
+    },
+    {
+      "epoch": 2.822730579581279,
+      "grad_norm": 0.7573966383934021,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 17460
+    },
+    {
+      "epoch": 2.824347263762024,
+      "grad_norm": 0.878688633441925,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 17470
+    },
+    {
+      "epoch": 2.8259639479427694,
+      "grad_norm": 0.7752242684364319,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 17480
+    },
+    {
+      "epoch": 2.8275806321235146,
+      "grad_norm": 0.6135398745536804,
+      "learning_rate": 0.0002,
+      "loss": 0.6638,
+      "step": 17490
+    },
+    {
+      "epoch": 2.82919731630426,
+      "grad_norm": 0.6924924850463867,
+      "learning_rate": 0.0002,
+      "loss": 0.6829,
+      "step": 17500
+    },
+    {
+      "epoch": 2.830814000485005,
+      "grad_norm": 0.7471627593040466,
+      "learning_rate": 0.0002,
+      "loss": 0.6731,
+      "step": 17510
+    },
+    {
+      "epoch": 2.8324306846657503,
+      "grad_norm": 0.7145499587059021,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 17520
+    },
+    {
+      "epoch": 2.834047368846496,
+      "grad_norm": 0.7415414452552795,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17530
+    },
+    {
+      "epoch": 2.8356640530272412,
+      "grad_norm": 0.7328441739082336,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17540
+    },
+    {
+      "epoch": 2.8372807372079865,
+      "grad_norm": 0.8267839550971985,
+      "learning_rate": 0.0002,
+      "loss": 0.6866,
+      "step": 17550
+    },
+    {
+      "epoch": 2.8388974213887317,
+      "grad_norm": 0.8877885341644287,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17560
+    },
+    {
+      "epoch": 2.840514105569477,
+      "grad_norm": 0.857138454914093,
+      "learning_rate": 0.0002,
+      "loss": 0.7136,
+      "step": 17570
+    },
+    {
+      "epoch": 2.842130789750222,
+      "grad_norm": 0.8470779657363892,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 17580
+    },
+    {
+      "epoch": 2.843747473930968,
+      "grad_norm": 0.8553254008293152,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 17590
+    },
+    {
+      "epoch": 2.845364158111713,
+      "grad_norm": 0.8033196926116943,
+      "learning_rate": 0.0002,
+      "loss": 0.7297,
+      "step": 17600
+    },
+    {
+      "epoch": 2.8469808422924583,
+      "grad_norm": 0.7949087023735046,
+      "learning_rate": 0.0002,
+      "loss": 0.7062,
+      "step": 17610
+    },
+    {
+      "epoch": 2.8485975264732035,
+      "grad_norm": 0.9241406321525574,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 17620
+    },
+    {
+      "epoch": 2.8502142106539488,
+      "grad_norm": 0.7721285223960876,
+      "learning_rate": 0.0002,
+      "loss": 0.6601,
+      "step": 17630
+    },
+    {
+      "epoch": 2.851830894834694,
+      "grad_norm": 1.0246692895889282,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 17640
+    },
+    {
+      "epoch": 2.853447579015439,
+      "grad_norm": 0.9244589805603027,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 17650
+    },
+    {
+      "epoch": 2.8550642631961844,
+      "grad_norm": 0.7243508696556091,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 17660
+    },
+    {
+      "epoch": 2.8566809473769297,
+      "grad_norm": 0.8943371176719666,
+      "learning_rate": 0.0002,
+      "loss": 0.6471,
+      "step": 17670
+    },
+    {
+      "epoch": 2.8582976315576754,
+      "grad_norm": 0.6531758904457092,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 17680
+    },
+    {
+      "epoch": 2.8599143157384206,
+      "grad_norm": 0.8367000818252563,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 17690
+    },
+    {
+      "epoch": 2.861530999919166,
+      "grad_norm": 0.7868556380271912,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 17700
+    },
+    {
+      "epoch": 2.863147684099911,
+      "grad_norm": 0.7213859558105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 17710
+    },
+    {
+      "epoch": 2.8647643682806563,
+      "grad_norm": 0.7383931279182434,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 17720
+    },
+    {
+      "epoch": 2.8663810524614015,
+      "grad_norm": 0.7566812634468079,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 17730
+    },
+    {
+      "epoch": 2.867997736642147,
+      "grad_norm": 0.6930373311042786,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 17740
+    },
+    {
+      "epoch": 2.8696144208228924,
+      "grad_norm": 0.7911090850830078,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17750
+    },
+    {
+      "epoch": 2.8712311050036377,
+      "grad_norm": 0.8484548926353455,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 17760
+    },
+    {
+      "epoch": 2.872847789184383,
+      "grad_norm": 0.7647597193717957,
+      "learning_rate": 0.0002,
+      "loss": 0.6931,
+      "step": 17770
+    },
+    {
+      "epoch": 2.874464473365128,
+      "grad_norm": 0.8791151642799377,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 17780
+    },
+    {
+      "epoch": 2.8760811575458733,
+      "grad_norm": 0.7253178358078003,
+      "learning_rate": 0.0002,
+      "loss": 0.7078,
+      "step": 17790
+    },
+    {
+      "epoch": 2.8776978417266186,
+      "grad_norm": 0.7956077456474304,
+      "learning_rate": 0.0002,
+      "loss": 0.6474,
+      "step": 17800
+    },
+    {
+      "epoch": 2.879314525907364,
+      "grad_norm": 0.8657688498497009,
+      "learning_rate": 0.0002,
+      "loss": 0.6687,
+      "step": 17810
+    },
+    {
+      "epoch": 2.880931210088109,
+      "grad_norm": 0.7059141993522644,
+      "learning_rate": 0.0002,
+      "loss": 0.7171,
+      "step": 17820
+    },
+    {
+      "epoch": 2.8825478942688547,
+      "grad_norm": 0.8886896967887878,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 17830
+    },
+    {
+      "epoch": 2.8841645784496,
+      "grad_norm": 0.821032702922821,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 17840
+    },
+    {
+      "epoch": 2.885781262630345,
+      "grad_norm": 0.7183963656425476,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 17850
+    },
+    {
+      "epoch": 2.8873979468110904,
+      "grad_norm": 0.6222899556159973,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 17860
+    },
+    {
+      "epoch": 2.8890146309918356,
+      "grad_norm": 0.8187434077262878,
+      "learning_rate": 0.0002,
+      "loss": 0.6626,
+      "step": 17870
+    },
+    {
+      "epoch": 2.890631315172581,
+      "grad_norm": 0.9838479161262512,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 17880
+    },
+    {
+      "epoch": 2.8922479993533265,
+      "grad_norm": 0.7567742466926575,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 17890
+    },
+    {
+      "epoch": 2.893864683534072,
+      "grad_norm": 0.6875903606414795,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17900
+    },
+    {
+      "epoch": 2.895481367714817,
+      "grad_norm": 0.8043789267539978,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 17910
+    },
+    {
+      "epoch": 2.8970980518955622,
+      "grad_norm": 0.8062626719474792,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 17920
+    },
+    {
+      "epoch": 2.8987147360763075,
+      "grad_norm": 1.0251191854476929,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 17930
+    },
+    {
+      "epoch": 2.9003314202570527,
+      "grad_norm": 0.882253110408783,
+      "learning_rate": 0.0002,
+      "loss": 0.6605,
+      "step": 17940
+    },
+    {
+      "epoch": 2.901948104437798,
+      "grad_norm": 0.8683299422264099,
+      "learning_rate": 0.0002,
+      "loss": 0.6719,
+      "step": 17950
+    },
+    {
+      "epoch": 2.903564788618543,
+      "grad_norm": 0.7167282104492188,
+      "learning_rate": 0.0002,
+      "loss": 0.6896,
+      "step": 17960
+    },
+    {
+      "epoch": 2.9051814727992884,
+      "grad_norm": 0.7093694806098938,
+      "learning_rate": 0.0002,
+      "loss": 0.663,
+      "step": 17970
+    },
+    {
+      "epoch": 2.906798156980034,
+      "grad_norm": 0.8549879193305969,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 17980
+    },
+    {
+      "epoch": 2.9084148411607793,
+      "grad_norm": 0.6989606618881226,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 17990
+    },
+    {
+      "epoch": 2.9100315253415245,
+      "grad_norm": 0.9482976794242859,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 18000
+    },
+    {
+      "epoch": 2.9116482095222698,
+      "grad_norm": 0.7182440161705017,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 18010
+    },
+    {
+      "epoch": 2.913264893703015,
+      "grad_norm": 0.7732226252555847,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 18020
+    },
+    {
+      "epoch": 2.9148815778837607,
+      "grad_norm": 0.7936875224113464,
+      "learning_rate": 0.0002,
+      "loss": 0.7123,
+      "step": 18030
+    },
+    {
+      "epoch": 2.916498262064506,
+      "grad_norm": 0.8825615644454956,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 18040
+    },
+    {
+      "epoch": 2.918114946245251,
+      "grad_norm": 0.6778587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 18050
+    },
+    {
+      "epoch": 2.9197316304259964,
+      "grad_norm": 0.7529265880584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 18060
+    },
+    {
+      "epoch": 2.9213483146067416,
+      "grad_norm": 0.7111883163452148,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 18070
+    },
+    {
+      "epoch": 2.922964998787487,
+      "grad_norm": 0.7214767932891846,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 18080
+    },
+    {
+      "epoch": 2.924581682968232,
+      "grad_norm": 0.800417423248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 18090
+    },
+    {
+      "epoch": 2.9261983671489773,
+      "grad_norm": 1.248575210571289,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 18100
+    },
+    {
+      "epoch": 2.9278150513297225,
+      "grad_norm": 0.757788360118866,
+      "learning_rate": 0.0002,
+      "loss": 0.7004,
+      "step": 18110
+    },
+    {
+      "epoch": 2.9294317355104678,
+      "grad_norm": 1.0583995580673218,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 18120
+    },
+    {
+      "epoch": 2.9310484196912134,
+      "grad_norm": 0.8228777647018433,
+      "learning_rate": 0.0002,
+      "loss": 0.6365,
+      "step": 18130
+    },
+    {
+      "epoch": 2.9326651038719587,
+      "grad_norm": 0.8374035358428955,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 18140
+    },
+    {
+      "epoch": 2.934281788052704,
+      "grad_norm": 0.7976473569869995,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 18150
+    },
+    {
+      "epoch": 2.935898472233449,
+      "grad_norm": 0.8009907603263855,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 18160
+    },
+    {
+      "epoch": 2.9375151564141944,
+      "grad_norm": 0.835213303565979,
+      "learning_rate": 0.0002,
+      "loss": 0.7485,
+      "step": 18170
+    },
+    {
+      "epoch": 2.93913184059494,
+      "grad_norm": 0.7982219457626343,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 18180
+    },
+    {
+      "epoch": 2.9407485247756853,
+      "grad_norm": 0.7070978879928589,
+      "learning_rate": 0.0002,
+      "loss": 0.6348,
+      "step": 18190
+    },
+    {
+      "epoch": 2.9423652089564305,
+      "grad_norm": 0.8619440197944641,
+      "learning_rate": 0.0002,
+      "loss": 0.6608,
+      "step": 18200
+    },
+    {
+      "epoch": 2.9439818931371757,
+      "grad_norm": 0.6693987250328064,
+      "learning_rate": 0.0002,
+      "loss": 0.666,
+      "step": 18210
+    },
+    {
+      "epoch": 2.945598577317921,
+      "grad_norm": 0.6747021079063416,
+      "learning_rate": 0.0002,
+      "loss": 0.728,
+      "step": 18220
+    },
+    {
+      "epoch": 2.947215261498666,
+      "grad_norm": 0.860387921333313,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 18230
+    },
+    {
+      "epoch": 2.9488319456794114,
+      "grad_norm": 0.799976646900177,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 18240
+    },
+    {
+      "epoch": 2.9504486298601567,
+      "grad_norm": 0.7864769101142883,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 18250
+    },
+    {
+      "epoch": 2.952065314040902,
+      "grad_norm": 0.6713884472846985,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 18260
+    },
+    {
+      "epoch": 2.9536819982216476,
+      "grad_norm": 0.9031508564949036,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 18270
+    },
+    {
+      "epoch": 2.955298682402393,
+      "grad_norm": 0.7205073237419128,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 18280
+    },
+    {
+      "epoch": 2.956915366583138,
+      "grad_norm": 0.7746205925941467,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 18290
+    },
+    {
+      "epoch": 2.9585320507638833,
+      "grad_norm": 0.6533427834510803,
+      "learning_rate": 0.0002,
+      "loss": 0.6218,
+      "step": 18300
+    },
+    {
+      "epoch": 2.9601487349446285,
+      "grad_norm": 0.9083208441734314,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 18310
+    },
+    {
+      "epoch": 2.9617654191253737,
+      "grad_norm": 0.7446991801261902,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 18320
+    },
+    {
+      "epoch": 2.9633821033061194,
+      "grad_norm": 0.6514461636543274,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 18330
+    },
+    {
+      "epoch": 2.9649987874868646,
+      "grad_norm": 0.8580465912818909,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18340
+    },
+    {
+      "epoch": 2.96661547166761,
+      "grad_norm": 0.7074266076087952,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 18350
+    },
+    {
+      "epoch": 2.968232155848355,
+      "grad_norm": 0.899892270565033,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 18360
+    },
+    {
+      "epoch": 2.9698488400291003,
+      "grad_norm": 0.8217641711235046,
+      "learning_rate": 0.0002,
+      "loss": 0.7094,
+      "step": 18370
+    },
+    {
+      "epoch": 2.9714655242098456,
+      "grad_norm": 0.8611799478530884,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 18380
+    },
+    {
+      "epoch": 2.973082208390591,
+      "grad_norm": 0.6909302473068237,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18390
+    },
+    {
+      "epoch": 2.974698892571336,
+      "grad_norm": 0.6554358005523682,
+      "learning_rate": 0.0002,
+      "loss": 0.7247,
+      "step": 18400
+    },
+    {
+      "epoch": 2.9763155767520812,
+      "grad_norm": 0.7803071737289429,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 18410
+    },
+    {
+      "epoch": 2.977932260932827,
+      "grad_norm": 0.7838954925537109,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 18420
+    },
+    {
+      "epoch": 2.979548945113572,
+      "grad_norm": 0.7098495364189148,
+      "learning_rate": 0.0002,
+      "loss": 0.6522,
+      "step": 18430
+    },
+    {
+      "epoch": 2.9811656292943174,
+      "grad_norm": 0.8981785774230957,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 18440
+    },
+    {
+      "epoch": 2.9827823134750626,
+      "grad_norm": 0.7197171449661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6689,
+      "step": 18450
+    },
+    {
+      "epoch": 2.984398997655808,
+      "grad_norm": 0.793185293674469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 18460
+    },
+    {
+      "epoch": 2.986015681836553,
+      "grad_norm": 0.8531473875045776,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 18470
+    },
+    {
+      "epoch": 2.9876323660172988,
+      "grad_norm": 0.6627361178398132,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 18480
+    },
+    {
+      "epoch": 2.989249050198044,
+      "grad_norm": 0.5708155035972595,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 18490
+    },
+    {
+      "epoch": 2.990865734378789,
+      "grad_norm": 0.8227280378341675,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 18500
+    },
+    {
+      "epoch": 2.9924824185595345,
+      "grad_norm": 0.7102749943733215,
+      "learning_rate": 0.0002,
+      "loss": 0.6701,
+      "step": 18510
+    },
+    {
+      "epoch": 2.9940991027402797,
+      "grad_norm": 0.839485228061676,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 18520
+    },
+    {
+      "epoch": 2.995715786921025,
+      "grad_norm": 0.9038704037666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 18530
+    },
+    {
+      "epoch": 2.99733247110177,
+      "grad_norm": 0.8737510442733765,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 18540
+    },
+    {
+      "epoch": 2.9989491552825154,
+      "grad_norm": 0.7323142886161804,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 18550
+    },
+    {
+      "epoch": 2.9999191657909625,
+      "eval_loss": 1.1262480020523071,
+      "eval_runtime": 122.0868,
+      "eval_samples_per_second": 6.004,
+      "eval_steps_per_second": 0.754,
+      "step": 18556
+    },
+    {
+      "epoch": 3.000565839463261,
+      "grad_norm": 0.8465463519096375,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 18560
+    },
+    {
+      "epoch": 3.0021825236440063,
+      "grad_norm": 0.9134138822555542,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 18570
+    },
+    {
+      "epoch": 3.0037992078247515,
+      "grad_norm": 0.760715126991272,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 18580
+    },
+    {
+      "epoch": 3.0054158920054967,
+      "grad_norm": 0.9208743572235107,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18590
+    },
+    {
+      "epoch": 3.007032576186242,
+      "grad_norm": 0.9232364892959595,
+      "learning_rate": 0.0002,
+      "loss": 0.5731,
+      "step": 18600
+    },
+    {
+      "epoch": 3.008649260366987,
+      "grad_norm": 1.1881544589996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6299,
+      "step": 18610
+    },
+    {
+      "epoch": 3.0102659445477324,
+      "grad_norm": 0.9372987747192383,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 18620
+    },
+    {
+      "epoch": 3.0118826287284777,
+      "grad_norm": 0.6900241374969482,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 18630
+    },
+    {
+      "epoch": 3.0134993129092233,
+      "grad_norm": 0.8451071381568909,
+      "learning_rate": 0.0002,
+      "loss": 0.5256,
+      "step": 18640
+    },
+    {
+      "epoch": 3.0151159970899686,
+      "grad_norm": 0.7763112187385559,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 18650
+    },
+    {
+      "epoch": 3.016732681270714,
+      "grad_norm": 1.043653964996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 18660
+    },
+    {
+      "epoch": 3.018349365451459,
+      "grad_norm": 1.0170660018920898,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 18670
+    },
+    {
+      "epoch": 3.0199660496322043,
+      "grad_norm": 0.7534180283546448,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 18680
+    },
+    {
+      "epoch": 3.0215827338129495,
+      "grad_norm": 0.7507367730140686,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 18690
+    },
+    {
+      "epoch": 3.0231994179936947,
+      "grad_norm": 0.7861620187759399,
+      "learning_rate": 0.0002,
+      "loss": 0.6201,
+      "step": 18700
+    },
+    {
+      "epoch": 3.0248161021744404,
+      "grad_norm": 1.0580339431762695,
+      "learning_rate": 0.0002,
+      "loss": 0.5802,
+      "step": 18710
+    },
+    {
+      "epoch": 3.0264327863551856,
+      "grad_norm": 0.7542710900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 18720
+    },
+    {
+      "epoch": 3.028049470535931,
+      "grad_norm": 0.8189544677734375,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 18730
+    },
+    {
+      "epoch": 3.029666154716676,
+      "grad_norm": 0.9126611351966858,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 18740
+    },
+    {
+      "epoch": 3.0312828388974213,
+      "grad_norm": 0.8891341686248779,
+      "learning_rate": 0.0002,
+      "loss": 0.6443,
+      "step": 18750
+    },
+    {
+      "epoch": 3.0328995230781666,
+      "grad_norm": 0.8419283032417297,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 18760
+    },
+    {
+      "epoch": 3.034516207258912,
+      "grad_norm": 0.8048048615455627,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18770
+    },
+    {
+      "epoch": 3.0361328914396575,
+      "grad_norm": 0.7820217609405518,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 18780
+    },
+    {
+      "epoch": 3.0377495756204027,
+      "grad_norm": 0.854721188545227,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 18790
+    },
+    {
+      "epoch": 3.039366259801148,
+      "grad_norm": 0.912092924118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5784,
+      "step": 18800
+    },
+    {
+      "epoch": 3.040982943981893,
+      "grad_norm": 0.6596226096153259,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 18810
+    },
+    {
+      "epoch": 3.0425996281626384,
+      "grad_norm": 0.6351348757743835,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 18820
+    },
+    {
+      "epoch": 3.0442163123433836,
+      "grad_norm": 0.778188943862915,
+      "learning_rate": 0.0002,
+      "loss": 0.5953,
+      "step": 18830
+    },
+    {
+      "epoch": 3.045832996524129,
+      "grad_norm": 0.68234783411026,
+      "learning_rate": 0.0002,
+      "loss": 0.602,
+      "step": 18840
+    },
+    {
+      "epoch": 3.047449680704874,
+      "grad_norm": 0.998628556728363,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 18850
+    },
+    {
+      "epoch": 3.0490663648856198,
+      "grad_norm": 0.7393841743469238,
+      "learning_rate": 0.0002,
+      "loss": 0.6231,
+      "step": 18860
+    },
+    {
+      "epoch": 3.050683049066365,
+      "grad_norm": 0.84438556432724,
+      "learning_rate": 0.0002,
+      "loss": 0.568,
+      "step": 18870
+    },
+    {
+      "epoch": 3.0522997332471102,
+      "grad_norm": 0.8857501745223999,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 18880
+    },
+    {
+      "epoch": 3.0539164174278555,
+      "grad_norm": 0.7208474278450012,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 18890
+    },
+    {
+      "epoch": 3.0555331016086007,
+      "grad_norm": 0.7135229110717773,
+      "learning_rate": 0.0002,
+      "loss": 0.5998,
+      "step": 18900
+    },
+    {
+      "epoch": 3.057149785789346,
+      "grad_norm": 0.9130001664161682,
+      "learning_rate": 0.0002,
+      "loss": 0.5575,
+      "step": 18910
+    },
+    {
+      "epoch": 3.058766469970091,
+      "grad_norm": 0.9001716375350952,
+      "learning_rate": 0.0002,
+      "loss": 0.5955,
+      "step": 18920
+    },
+    {
+      "epoch": 3.060383154150837,
+      "grad_norm": 0.8667559623718262,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 18930
+    },
+    {
+      "epoch": 3.061999838331582,
+      "grad_norm": 0.8943959474563599,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18940
+    },
+    {
+      "epoch": 3.0636165225123273,
+      "grad_norm": 0.8298377990722656,
+      "learning_rate": 0.0002,
+      "loss": 0.5978,
+      "step": 18950
+    },
+    {
+      "epoch": 3.0652332066930725,
+      "grad_norm": 0.7935267686843872,
+      "learning_rate": 0.0002,
+      "loss": 0.5782,
+      "step": 18960
+    },
+    {
+      "epoch": 3.0668498908738178,
+      "grad_norm": 1.1506379842758179,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 18970
+    },
+    {
+      "epoch": 3.068466575054563,
+      "grad_norm": 0.7693049907684326,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18980
+    },
+    {
+      "epoch": 3.0700832592353082,
+      "grad_norm": 0.8040135502815247,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 18990
+    },
+    {
+      "epoch": 3.0716999434160535,
+      "grad_norm": 0.828404426574707,
+      "learning_rate": 0.0002,
+      "loss": 0.5541,
+      "step": 19000
+    },
+    {
+      "epoch": 3.073316627596799,
+      "grad_norm": 0.8811164498329163,
+      "learning_rate": 0.0002,
+      "loss": 0.6048,
+      "step": 19010
+    },
+    {
+      "epoch": 3.0749333117775444,
+      "grad_norm": 1.036205768585205,
+      "learning_rate": 0.0002,
+      "loss": 0.5845,
+      "step": 19020
+    },
+    {
+      "epoch": 3.0765499959582896,
+      "grad_norm": 0.8857285976409912,
+      "learning_rate": 0.0002,
+      "loss": 0.5838,
+      "step": 19030
+    },
+    {
+      "epoch": 3.078166680139035,
+      "grad_norm": 0.8392079472541809,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19040
+    },
+    {
+      "epoch": 3.07978336431978,
+      "grad_norm": 1.0287401676177979,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 19050
+    },
+    {
+      "epoch": 3.0814000485005253,
+      "grad_norm": 1.0086315870285034,
+      "learning_rate": 0.0002,
+      "loss": 0.5964,
+      "step": 19060
+    },
+    {
+      "epoch": 3.0830167326812705,
+      "grad_norm": 0.9245324730873108,
+      "learning_rate": 0.0002,
+      "loss": 0.5567,
+      "step": 19070
+    },
+    {
+      "epoch": 3.084633416862016,
+      "grad_norm": 0.8680877089500427,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 19080
+    },
+    {
+      "epoch": 3.0862501010427614,
+      "grad_norm": 0.8814793825149536,
+      "learning_rate": 0.0002,
+      "loss": 0.5611,
+      "step": 19090
+    },
+    {
+      "epoch": 3.0878667852235067,
+      "grad_norm": 0.9234458208084106,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19100
+    },
+    {
+      "epoch": 3.089483469404252,
+      "grad_norm": 1.1291664838790894,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 19110
+    },
+    {
+      "epoch": 3.091100153584997,
+      "grad_norm": 0.9191402792930603,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 19120
+    },
+    {
+      "epoch": 3.0927168377657424,
+      "grad_norm": 0.7103154063224792,
+      "learning_rate": 0.0002,
+      "loss": 0.5856,
+      "step": 19130
+    },
+    {
+      "epoch": 3.0943335219464876,
+      "grad_norm": 0.9368883967399597,
+      "learning_rate": 0.0002,
+      "loss": 0.6479,
+      "step": 19140
+    },
+    {
+      "epoch": 3.095950206127233,
+      "grad_norm": 0.9676656723022461,
+      "learning_rate": 0.0002,
+      "loss": 0.6167,
+      "step": 19150
+    },
+    {
+      "epoch": 3.0975668903079785,
+      "grad_norm": 0.8739792704582214,
+      "learning_rate": 0.0002,
+      "loss": 0.5794,
+      "step": 19160
+    },
+    {
+      "epoch": 3.0991835744887237,
+      "grad_norm": 0.8530174493789673,
+      "learning_rate": 0.0002,
+      "loss": 0.6112,
+      "step": 19170
+    },
+    {
+      "epoch": 3.100800258669469,
+      "grad_norm": 0.794945478439331,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 19180
+    },
+    {
+      "epoch": 3.102416942850214,
+      "grad_norm": 0.9508888125419617,
+      "learning_rate": 0.0002,
+      "loss": 0.5928,
+      "step": 19190
+    },
+    {
+      "epoch": 3.1040336270309594,
+      "grad_norm": 1.0599955320358276,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 19200
+    },
+    {
+      "epoch": 3.1056503112117047,
+      "grad_norm": 1.0673625469207764,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 19210
+    },
+    {
+      "epoch": 3.10726699539245,
+      "grad_norm": 0.7739115953445435,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 19220
+    },
+    {
+      "epoch": 3.1088836795731956,
+      "grad_norm": 0.9884951114654541,
+      "learning_rate": 0.0002,
+      "loss": 0.6046,
+      "step": 19230
+    },
+    {
+      "epoch": 3.110500363753941,
+      "grad_norm": 0.862260103225708,
+      "learning_rate": 0.0002,
+      "loss": 0.5932,
+      "step": 19240
+    },
+    {
+      "epoch": 3.112117047934686,
+      "grad_norm": 0.7690284848213196,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 19250
+    },
+    {
+      "epoch": 3.1137337321154313,
+      "grad_norm": 0.8758958578109741,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 19260
+    },
+    {
+      "epoch": 3.1153504162961765,
+      "grad_norm": 1.0356395244598389,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 19270
+    },
+    {
+      "epoch": 3.1169671004769217,
+      "grad_norm": 0.6950937509536743,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 19280
+    },
+    {
+      "epoch": 3.118583784657667,
+      "grad_norm": 0.760998010635376,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19290
+    },
+    {
+      "epoch": 3.1202004688384126,
+      "grad_norm": 0.9335789084434509,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 19300
+    },
+    {
+      "epoch": 3.121817153019158,
+      "grad_norm": 0.9636204242706299,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 19310
+    },
+    {
+      "epoch": 3.123433837199903,
+      "grad_norm": 1.0820997953414917,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 19320
+    },
+    {
+      "epoch": 3.1250505213806483,
+      "grad_norm": 0.7333487272262573,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 19330
+    },
+    {
+      "epoch": 3.1266672055613935,
+      "grad_norm": 1.0417509078979492,
+      "learning_rate": 0.0002,
+      "loss": 0.6178,
+      "step": 19340
+    },
+    {
+      "epoch": 3.128283889742139,
+      "grad_norm": 0.9267749190330505,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 19350
+    },
+    {
+      "epoch": 3.129900573922884,
+      "grad_norm": 0.777798593044281,
+      "learning_rate": 0.0002,
+      "loss": 0.6063,
+      "step": 19360
+    },
+    {
+      "epoch": 3.1315172581036297,
+      "grad_norm": 0.8425456881523132,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 19370
+    },
+    {
+      "epoch": 3.133133942284375,
+      "grad_norm": 0.9617102146148682,
+      "learning_rate": 0.0002,
+      "loss": 0.6042,
+      "step": 19380
+    },
+    {
+      "epoch": 3.13475062646512,
+      "grad_norm": 1.0052828788757324,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 19390
+    },
+    {
+      "epoch": 3.1363673106458654,
+      "grad_norm": 0.7637009024620056,
+      "learning_rate": 0.0002,
+      "loss": 0.5713,
+      "step": 19400
+    },
+    {
+      "epoch": 3.1379839948266106,
+      "grad_norm": 0.7958088517189026,
+      "learning_rate": 0.0002,
+      "loss": 0.5497,
+      "step": 19410
+    },
+    {
+      "epoch": 3.139600679007356,
+      "grad_norm": 0.9161727428436279,
+      "learning_rate": 0.0002,
+      "loss": 0.6283,
+      "step": 19420
+    },
+    {
+      "epoch": 3.141217363188101,
+      "grad_norm": 0.8402149677276611,
+      "learning_rate": 0.0002,
+      "loss": 0.5638,
+      "step": 19430
+    },
+    {
+      "epoch": 3.1428340473688463,
+      "grad_norm": 1.0056525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.5848,
+      "step": 19440
+    },
+    {
+      "epoch": 3.144450731549592,
+      "grad_norm": 1.0129190683364868,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 19450
+    },
+    {
+      "epoch": 3.146067415730337,
+      "grad_norm": 0.790825366973877,
+      "learning_rate": 0.0002,
+      "loss": 0.5808,
+      "step": 19460
+    },
+    {
+      "epoch": 3.1476840999110824,
+      "grad_norm": 1.441665530204773,
+      "learning_rate": 0.0002,
+      "loss": 0.5607,
+      "step": 19470
+    },
+    {
+      "epoch": 3.1493007840918277,
+      "grad_norm": 0.7846331596374512,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19480
+    },
+    {
+      "epoch": 3.150917468272573,
+      "grad_norm": 0.7915332913398743,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 19490
+    },
+    {
+      "epoch": 3.152534152453318,
+      "grad_norm": 0.933982253074646,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 19500
+    },
+    {
+      "epoch": 3.1541508366340634,
+      "grad_norm": 1.038408637046814,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 19510
+    },
+    {
+      "epoch": 3.155767520814809,
+      "grad_norm": 1.018935203552246,
+      "learning_rate": 0.0002,
+      "loss": 0.6271,
+      "step": 19520
+    },
+    {
+      "epoch": 3.1573842049955543,
+      "grad_norm": 0.9618112444877625,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 19530
+    },
+    {
+      "epoch": 3.1590008891762995,
+      "grad_norm": 0.8900452852249146,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 19540
+    },
+    {
+      "epoch": 3.1606175733570447,
+      "grad_norm": 0.8254160284996033,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 19550
+    },
+    {
+      "epoch": 3.16223425753779,
+      "grad_norm": 1.004376769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19560
+    },
+    {
+      "epoch": 3.163850941718535,
+      "grad_norm": 1.0490446090698242,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 19570
+    },
+    {
+      "epoch": 3.1654676258992804,
+      "grad_norm": 0.7387403845787048,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19580
+    },
+    {
+      "epoch": 3.1670843100800257,
+      "grad_norm": 0.7611538171768188,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 19590
+    },
+    {
+      "epoch": 3.1687009942607713,
+      "grad_norm": 0.8239886164665222,
+      "learning_rate": 0.0002,
+      "loss": 0.5937,
+      "step": 19600
+    },
+    {
+      "epoch": 3.1703176784415166,
+      "grad_norm": 0.9327243566513062,
+      "learning_rate": 0.0002,
+      "loss": 0.6068,
+      "step": 19610
+    },
+    {
+      "epoch": 3.171934362622262,
+      "grad_norm": 0.9662560224533081,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 19620
+    },
+    {
+      "epoch": 3.173551046803007,
+      "grad_norm": 0.9183341860771179,
+      "learning_rate": 0.0002,
+      "loss": 0.5988,
+      "step": 19630
+    },
+    {
+      "epoch": 3.1751677309837523,
+      "grad_norm": 0.875066876411438,
+      "learning_rate": 0.0002,
+      "loss": 0.5909,
+      "step": 19640
+    },
+    {
+      "epoch": 3.1767844151644975,
+      "grad_norm": 0.8567508459091187,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 19650
+    },
+    {
+      "epoch": 3.1784010993452427,
+      "grad_norm": 0.6805780529975891,
+      "learning_rate": 0.0002,
+      "loss": 0.5805,
+      "step": 19660
+    },
+    {
+      "epoch": 3.1800177835259884,
+      "grad_norm": 0.8776944279670715,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 19670
+    },
+    {
+      "epoch": 3.1816344677067336,
+      "grad_norm": 0.9036329984664917,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 19680
+    },
+    {
+      "epoch": 3.183251151887479,
+      "grad_norm": 0.8527372479438782,
+      "learning_rate": 0.0002,
+      "loss": 0.6238,
+      "step": 19690
+    },
+    {
+      "epoch": 3.184867836068224,
+      "grad_norm": 1.1045585870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 19700
+    },
+    {
+      "epoch": 3.1864845202489693,
+      "grad_norm": 0.9213830828666687,
+      "learning_rate": 0.0002,
+      "loss": 0.5491,
+      "step": 19710
+    },
+    {
+      "epoch": 3.1881012044297146,
+      "grad_norm": 0.8865814805030823,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 19720
+    },
+    {
+      "epoch": 3.18971788861046,
+      "grad_norm": 0.7939388751983643,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19730
+    },
+    {
+      "epoch": 3.191334572791205,
+      "grad_norm": 0.6966729760169983,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 19740
+    },
+    {
+      "epoch": 3.1929512569719507,
+      "grad_norm": 0.8023673295974731,
+      "learning_rate": 0.0002,
+      "loss": 0.5839,
+      "step": 19750
+    },
+    {
+      "epoch": 3.194567941152696,
+      "grad_norm": 0.7992037534713745,
+      "learning_rate": 0.0002,
+      "loss": 0.6267,
+      "step": 19760
+    },
+    {
+      "epoch": 3.196184625333441,
+      "grad_norm": 0.7412247657775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 19770
+    },
+    {
+      "epoch": 3.1978013095141864,
+      "grad_norm": 0.9598729014396667,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 19780
+    },
+    {
+      "epoch": 3.1994179936949316,
+      "grad_norm": 0.8331366777420044,
+      "learning_rate": 0.0002,
+      "loss": 0.5685,
+      "step": 19790
+    },
+    {
+      "epoch": 3.201034677875677,
+      "grad_norm": 0.8939169645309448,
+      "learning_rate": 0.0002,
+      "loss": 0.6104,
+      "step": 19800
+    },
+    {
+      "epoch": 3.202651362056422,
+      "grad_norm": 0.9219734072685242,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 19810
+    },
+    {
+      "epoch": 3.2042680462371678,
+      "grad_norm": 0.869490385055542,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19820
+    },
+    {
+      "epoch": 3.205884730417913,
+      "grad_norm": 0.8989706635475159,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 19830
+    },
+    {
+      "epoch": 3.2075014145986582,
+      "grad_norm": 0.8477165102958679,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 19840
+    },
+    {
+      "epoch": 3.2091180987794035,
+      "grad_norm": 0.8720678687095642,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 19850
+    },
+    {
+      "epoch": 3.2107347829601487,
+      "grad_norm": 0.861406683921814,
+      "learning_rate": 0.0002,
+      "loss": 0.5694,
+      "step": 19860
+    },
+    {
+      "epoch": 3.212351467140894,
+      "grad_norm": 0.8228686451911926,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 19870
+    },
+    {
+      "epoch": 3.213968151321639,
+      "grad_norm": 0.7936596870422363,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19880
+    },
+    {
+      "epoch": 3.2155848355023844,
+      "grad_norm": 1.097377896308899,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 19890
+    },
+    {
+      "epoch": 3.21720151968313,
+      "grad_norm": 0.9544782638549805,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 19900
+    },
+    {
+      "epoch": 3.2188182038638753,
+      "grad_norm": 0.8240751624107361,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 19910
+    },
+    {
+      "epoch": 3.2204348880446205,
+      "grad_norm": 0.8332096338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 19920
+    },
+    {
+      "epoch": 3.2220515722253658,
+      "grad_norm": 1.0954567193984985,
+      "learning_rate": 0.0002,
+      "loss": 0.6274,
+      "step": 19930
+    },
+    {
+      "epoch": 3.223668256406111,
+      "grad_norm": 0.7790525555610657,
+      "learning_rate": 0.0002,
+      "loss": 0.652,
+      "step": 19940
+    },
+    {
+      "epoch": 3.225284940586856,
+      "grad_norm": 0.7966814041137695,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19950
+    },
+    {
+      "epoch": 3.2269016247676015,
+      "grad_norm": 0.9751881957054138,
+      "learning_rate": 0.0002,
+      "loss": 0.5911,
+      "step": 19960
+    },
+    {
+      "epoch": 3.228518308948347,
+      "grad_norm": 0.9856047630310059,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 19970
+    },
+    {
+      "epoch": 3.2301349931290924,
+      "grad_norm": 1.3062353134155273,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 19980
+    },
+    {
+      "epoch": 3.2317516773098376,
+      "grad_norm": 0.9510692358016968,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 19990
+    },
+    {
+      "epoch": 3.233368361490583,
+      "grad_norm": 0.8630342483520508,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 20000
+    },
+    {
+      "epoch": 3.234985045671328,
+      "grad_norm": 0.8966519236564636,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20010
+    },
+    {
+      "epoch": 3.2366017298520733,
+      "grad_norm": 0.7093510627746582,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 20020
+    },
+    {
+      "epoch": 3.2382184140328185,
+      "grad_norm": 0.7771096229553223,
+      "learning_rate": 0.0002,
+      "loss": 0.585,
+      "step": 20030
+    },
+    {
+      "epoch": 3.2398350982135637,
+      "grad_norm": 0.841058075428009,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 20040
+    },
+    {
+      "epoch": 3.2414517823943094,
+      "grad_norm": 0.909712553024292,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 20050
+    },
+    {
+      "epoch": 3.2430684665750547,
+      "grad_norm": 0.8321019411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20060
+    },
+    {
+      "epoch": 3.2446851507558,
+      "grad_norm": 0.779901921749115,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 20070
+    },
+    {
+      "epoch": 3.246301834936545,
+      "grad_norm": 0.6249170303344727,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 20080
+    },
+    {
+      "epoch": 3.2479185191172903,
+      "grad_norm": 0.8000940680503845,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 20090
+    },
+    {
+      "epoch": 3.2495352032980356,
+      "grad_norm": 0.7627735137939453,
+      "learning_rate": 0.0002,
+      "loss": 0.6273,
+      "step": 20100
+    },
+    {
+      "epoch": 3.2511518874787813,
+      "grad_norm": 0.8780747056007385,
+      "learning_rate": 0.0002,
+      "loss": 0.6223,
+      "step": 20110
+    },
+    {
+      "epoch": 3.2527685716595265,
+      "grad_norm": 0.772037148475647,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 20120
+    },
+    {
+      "epoch": 3.2543852558402717,
+      "grad_norm": 1.0086580514907837,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 20130
+    },
+    {
+      "epoch": 3.256001940021017,
+      "grad_norm": 0.9360289573669434,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20140
+    },
+    {
+      "epoch": 3.257618624201762,
+      "grad_norm": 1.2099586725234985,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20150
+    },
+    {
+      "epoch": 3.2592353083825074,
+      "grad_norm": 0.8368481397628784,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 20160
+    },
+    {
+      "epoch": 3.2608519925632526,
+      "grad_norm": 0.7391039133071899,
+      "learning_rate": 0.0002,
+      "loss": 0.5626,
+      "step": 20170
+    },
+    {
+      "epoch": 3.262468676743998,
+      "grad_norm": 0.9122273325920105,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 20180
+    },
+    {
+      "epoch": 3.264085360924743,
+      "grad_norm": 0.8502281904220581,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 20190
+    },
+    {
+      "epoch": 3.265702045105489,
+      "grad_norm": 1.0926852226257324,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 20200
+    },
+    {
+      "epoch": 3.267318729286234,
+      "grad_norm": 0.7902828454971313,
+      "learning_rate": 0.0002,
+      "loss": 0.6027,
+      "step": 20210
+    },
+    {
+      "epoch": 3.2689354134669792,
+      "grad_norm": 0.8724729418754578,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20220
+    },
+    {
+      "epoch": 3.2705520976477245,
+      "grad_norm": 0.8469277024269104,
+      "learning_rate": 0.0002,
+      "loss": 0.6242,
+      "step": 20230
+    },
+    {
+      "epoch": 3.2721687818284697,
+      "grad_norm": 0.8865092992782593,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 20240
+    },
+    {
+      "epoch": 3.273785466009215,
+      "grad_norm": 1.0979334115982056,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20250
+    },
+    {
+      "epoch": 3.2754021501899606,
+      "grad_norm": 1.0860793590545654,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 20260
+    },
+    {
+      "epoch": 3.277018834370706,
+      "grad_norm": 0.981745183467865,
+      "learning_rate": 0.0002,
+      "loss": 0.6105,
+      "step": 20270
+    },
+    {
+      "epoch": 3.278635518551451,
+      "grad_norm": 0.9155020713806152,
+      "learning_rate": 0.0002,
+      "loss": 0.627,
+      "step": 20280
+    },
+    {
+      "epoch": 3.2802522027321963,
+      "grad_norm": 0.8436718583106995,
+      "learning_rate": 0.0002,
+      "loss": 0.5899,
+      "step": 20290
+    },
+    {
+      "epoch": 3.2818688869129415,
+      "grad_norm": 1.0329409837722778,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 20300
+    },
+    {
+      "epoch": 3.2834855710936868,
+      "grad_norm": 0.9876394271850586,
+      "learning_rate": 0.0002,
+      "loss": 0.6,
+      "step": 20310
+    },
+    {
+      "epoch": 3.285102255274432,
+      "grad_norm": 0.8052917718887329,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 20320
+    },
+    {
+      "epoch": 3.2867189394551772,
+      "grad_norm": 0.8390680551528931,
+      "learning_rate": 0.0002,
+      "loss": 0.5949,
+      "step": 20330
+    },
+    {
+      "epoch": 3.288335623635923,
+      "grad_norm": 0.9515735507011414,
+      "learning_rate": 0.0002,
+      "loss": 0.6492,
+      "step": 20340
+    },
+    {
+      "epoch": 3.289952307816668,
+      "grad_norm": 0.8028870224952698,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 20350
+    },
+    {
+      "epoch": 3.2915689919974134,
+      "grad_norm": 0.862592339515686,
+      "learning_rate": 0.0002,
+      "loss": 0.634,
+      "step": 20360
+    },
+    {
+      "epoch": 3.2931856761781586,
+      "grad_norm": 0.7451621890068054,
+      "learning_rate": 0.0002,
+      "loss": 0.6345,
+      "step": 20370
+    },
+    {
+      "epoch": 3.294802360358904,
+      "grad_norm": 0.8966776728630066,
+      "learning_rate": 0.0002,
+      "loss": 0.6458,
+      "step": 20380
+    },
+    {
+      "epoch": 3.296419044539649,
+      "grad_norm": 0.9289216995239258,
+      "learning_rate": 0.0002,
+      "loss": 0.5967,
+      "step": 20390
+    },
+    {
+      "epoch": 3.2980357287203943,
+      "grad_norm": 0.9649626612663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6599,
+      "step": 20400
+    },
+    {
+      "epoch": 3.29965241290114,
+      "grad_norm": 1.1953798532485962,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 20410
+    },
+    {
+      "epoch": 3.301269097081885,
+      "grad_norm": 0.8929083943367004,
+      "learning_rate": 0.0002,
+      "loss": 0.5997,
+      "step": 20420
+    },
+    {
+      "epoch": 3.3028857812626304,
+      "grad_norm": 0.8922014236450195,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 20430
+    },
+    {
+      "epoch": 3.3045024654433757,
+      "grad_norm": 0.9754860401153564,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 20440
+    },
+    {
+      "epoch": 3.306119149624121,
+      "grad_norm": 0.8873140215873718,
+      "learning_rate": 0.0002,
+      "loss": 0.5653,
+      "step": 20450
+    },
+    {
+      "epoch": 3.307735833804866,
+      "grad_norm": 0.857271671295166,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20460
+    },
+    {
+      "epoch": 3.3093525179856114,
+      "grad_norm": 0.9022141098976135,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 20470
+    },
+    {
+      "epoch": 3.3109692021663566,
+      "grad_norm": 0.8614798188209534,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 20480
+    },
+    {
+      "epoch": 3.3125858863471023,
+      "grad_norm": 0.8838164210319519,
+      "learning_rate": 0.0002,
+      "loss": 0.6254,
+      "step": 20490
+    },
+    {
+      "epoch": 3.3142025705278475,
+      "grad_norm": 0.8709736466407776,
+      "learning_rate": 0.0002,
+      "loss": 0.5849,
+      "step": 20500
+    },
+    {
+      "epoch": 3.3158192547085927,
+      "grad_norm": 0.9533300995826721,
+      "learning_rate": 0.0002,
+      "loss": 0.6146,
+      "step": 20510
+    },
+    {
+      "epoch": 3.317435938889338,
+      "grad_norm": 0.8259269595146179,
+      "learning_rate": 0.0002,
+      "loss": 0.6029,
+      "step": 20520
+    },
+    {
+      "epoch": 3.319052623070083,
+      "grad_norm": 0.8607608079910278,
+      "learning_rate": 0.0002,
+      "loss": 0.6268,
+      "step": 20530
+    },
+    {
+      "epoch": 3.3206693072508284,
+      "grad_norm": 1.0863020420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.5676,
+      "step": 20540
+    },
+    {
+      "epoch": 3.3222859914315737,
+      "grad_norm": 1.011489987373352,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 20550
+    },
+    {
+      "epoch": 3.3239026756123193,
+      "grad_norm": 0.6952177882194519,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 20560
+    },
+    {
+      "epoch": 3.3255193597930646,
+      "grad_norm": 0.9638974070549011,
+      "learning_rate": 0.0002,
+      "loss": 0.6229,
+      "step": 20570
+    },
+    {
+      "epoch": 3.32713604397381,
+      "grad_norm": 1.0310138463974,
+      "learning_rate": 0.0002,
+      "loss": 0.5882,
+      "step": 20580
+    },
+    {
+      "epoch": 3.328752728154555,
+      "grad_norm": 0.9371318221092224,
+      "learning_rate": 0.0002,
+      "loss": 0.594,
+      "step": 20590
+    },
+    {
+      "epoch": 3.3303694123353003,
+      "grad_norm": 0.8756691813468933,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 20600
+    },
+    {
+      "epoch": 3.3319860965160455,
+      "grad_norm": 1.054175853729248,
+      "learning_rate": 0.0002,
+      "loss": 0.5994,
+      "step": 20610
+    },
+    {
+      "epoch": 3.3336027806967907,
+      "grad_norm": 0.9074128270149231,
+      "learning_rate": 0.0002,
+      "loss": 0.6169,
+      "step": 20620
+    },
+    {
+      "epoch": 3.335219464877536,
+      "grad_norm": 0.906900942325592,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20630
+    },
+    {
+      "epoch": 3.3368361490582816,
+      "grad_norm": 0.8689333200454712,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 20640
+    },
+    {
+      "epoch": 3.338452833239027,
+      "grad_norm": 0.9889747500419617,
+      "learning_rate": 0.0002,
+      "loss": 0.6079,
+      "step": 20650
+    },
+    {
+      "epoch": 3.340069517419772,
+      "grad_norm": 1.0685805082321167,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20660
+    },
+    {
+      "epoch": 3.3416862016005173,
+      "grad_norm": 0.7495010495185852,
+      "learning_rate": 0.0002,
+      "loss": 0.6091,
+      "step": 20670
+    },
+    {
+      "epoch": 3.3433028857812626,
+      "grad_norm": 0.8747848272323608,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 20680
+    },
+    {
+      "epoch": 3.344919569962008,
+      "grad_norm": 0.9762673377990723,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 20690
+    },
+    {
+      "epoch": 3.346536254142753,
+      "grad_norm": 1.0284489393234253,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 20700
+    },
+    {
+      "epoch": 3.3481529383234987,
+      "grad_norm": 0.7293812036514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20710
+    },
+    {
+      "epoch": 3.349769622504244,
+      "grad_norm": 0.8330199122428894,
+      "learning_rate": 0.0002,
+      "loss": 0.609,
+      "step": 20720
+    },
+    {
+      "epoch": 3.351386306684989,
+      "grad_norm": 0.9808499217033386,
+      "learning_rate": 0.0002,
+      "loss": 0.5729,
+      "step": 20730
+    },
+    {
+      "epoch": 3.3530029908657344,
+      "grad_norm": 0.9508825540542603,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 20740
+    },
+    {
+      "epoch": 3.3546196750464796,
+      "grad_norm": 0.790483832359314,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 20750
+    },
+    {
+      "epoch": 3.356236359227225,
+      "grad_norm": 1.022793173789978,
+      "learning_rate": 0.0002,
+      "loss": 0.6327,
+      "step": 20760
+    },
+    {
+      "epoch": 3.35785304340797,
+      "grad_norm": 0.8318950533866882,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 20770
+    },
+    {
+      "epoch": 3.3594697275887153,
+      "grad_norm": 0.7980858087539673,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 20780
+    },
+    {
+      "epoch": 3.361086411769461,
+      "grad_norm": 0.8114802241325378,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 20790
+    },
+    {
+      "epoch": 3.3627030959502062,
+      "grad_norm": 0.8522519469261169,
+      "learning_rate": 0.0002,
+      "loss": 0.6017,
+      "step": 20800
+    },
+    {
+      "epoch": 3.3643197801309515,
+      "grad_norm": 0.9142431616783142,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 20810
+    },
+    {
+      "epoch": 3.3659364643116967,
+      "grad_norm": 0.771170437335968,
+      "learning_rate": 0.0002,
+      "loss": 0.6331,
+      "step": 20820
+    },
+    {
+      "epoch": 3.367553148492442,
+      "grad_norm": 1.0628231763839722,
+      "learning_rate": 0.0002,
+      "loss": 0.5879,
+      "step": 20830
+    },
+    {
+      "epoch": 3.369169832673187,
+      "grad_norm": 0.9384352564811707,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 20840
+    },
+    {
+      "epoch": 3.370786516853933,
+      "grad_norm": 1.1286591291427612,
+      "learning_rate": 0.0002,
+      "loss": 0.6292,
+      "step": 20850
+    },
+    {
+      "epoch": 3.372403201034678,
+      "grad_norm": 1.1349513530731201,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 20860
+    },
+    {
+      "epoch": 3.3740198852154233,
+      "grad_norm": 1.0127464532852173,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 20870
+    },
+    {
+      "epoch": 3.3756365693961685,
+      "grad_norm": 0.9111971855163574,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 20880
+    },
+    {
+      "epoch": 3.3772532535769137,
+      "grad_norm": 0.871356725692749,
+      "learning_rate": 0.0002,
+      "loss": 0.6101,
+      "step": 20890
+    },
+    {
+      "epoch": 3.378869937757659,
+      "grad_norm": 0.7774117588996887,
+      "learning_rate": 0.0002,
+      "loss": 0.5995,
+      "step": 20900
+    },
+    {
+      "epoch": 3.380486621938404,
+      "grad_norm": 1.0089964866638184,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 20910
+    },
+    {
+      "epoch": 3.3821033061191494,
+      "grad_norm": 0.7855867147445679,
+      "learning_rate": 0.0002,
+      "loss": 0.5908,
+      "step": 20920
+    },
+    {
+      "epoch": 3.3837199902998947,
+      "grad_norm": 1.3713710308074951,
+      "learning_rate": 0.0002,
+      "loss": 0.6373,
+      "step": 20930
+    },
+    {
+      "epoch": 3.3853366744806404,
+      "grad_norm": 0.8599116206169128,
+      "learning_rate": 0.0002,
+      "loss": 0.6627,
+      "step": 20940
+    },
+    {
+      "epoch": 3.3869533586613856,
+      "grad_norm": 0.9392673373222351,
+      "learning_rate": 0.0002,
+      "loss": 0.6224,
+      "step": 20950
+    },
+    {
+      "epoch": 3.388570042842131,
+      "grad_norm": 0.8764075040817261,
+      "learning_rate": 0.0002,
+      "loss": 0.5855,
+      "step": 20960
+    },
+    {
+      "epoch": 3.390186727022876,
+      "grad_norm": 0.8240136504173279,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 20970
+    },
+    {
+      "epoch": 3.3918034112036213,
+      "grad_norm": 1.0982369184494019,
+      "learning_rate": 0.0002,
+      "loss": 0.5783,
+      "step": 20980
+    },
+    {
+      "epoch": 3.3934200953843665,
+      "grad_norm": 1.0599013566970825,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 20990
+    },
+    {
+      "epoch": 3.395036779565112,
+      "grad_norm": 0.895438015460968,
+      "learning_rate": 0.0002,
+      "loss": 0.6356,
+      "step": 21000
+    },
+    {
+      "epoch": 3.3966534637458574,
+      "grad_norm": 0.6974841356277466,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 21010
+    },
+    {
+      "epoch": 3.3982701479266026,
+      "grad_norm": 0.9571719765663147,
+      "learning_rate": 0.0002,
+      "loss": 0.5704,
+      "step": 21020
+    },
+    {
+      "epoch": 3.399886832107348,
+      "grad_norm": 0.831912636756897,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 21030
+    },
+    {
+      "epoch": 3.401503516288093,
+      "grad_norm": 0.831936240196228,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 21040
+    },
+    {
+      "epoch": 3.4031202004688383,
+      "grad_norm": 0.7388373613357544,
+      "learning_rate": 0.0002,
+      "loss": 0.5857,
+      "step": 21050
+    },
+    {
+      "epoch": 3.4047368846495836,
+      "grad_norm": 0.938667356967926,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21060
+    },
+    {
+      "epoch": 3.406353568830329,
+      "grad_norm": 0.9202313423156738,
+      "learning_rate": 0.0002,
+      "loss": 0.6121,
+      "step": 21070
+    },
+    {
+      "epoch": 3.4079702530110745,
+      "grad_norm": 0.9888381958007812,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 21080
+    },
+    {
+      "epoch": 3.4095869371918197,
+      "grad_norm": 0.8526970744132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21090
+    },
+    {
+      "epoch": 3.411203621372565,
+      "grad_norm": 0.7939383387565613,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 21100
+    },
+    {
+      "epoch": 3.41282030555331,
+      "grad_norm": 0.9986352920532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 21110
+    },
+    {
+      "epoch": 3.4144369897340554,
+      "grad_norm": 0.8895300030708313,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 21120
+    },
+    {
+      "epoch": 3.4160536739148006,
+      "grad_norm": 0.9559482932090759,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 21130
+    },
+    {
+      "epoch": 3.417670358095546,
+      "grad_norm": 0.8351506590843201,
+      "learning_rate": 0.0002,
+      "loss": 0.6491,
+      "step": 21140
+    },
+    {
+      "epoch": 3.4192870422762915,
+      "grad_norm": 0.8224456906318665,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 21150
+    },
+    {
+      "epoch": 3.4209037264570368,
+      "grad_norm": 1.0110299587249756,
+      "learning_rate": 0.0002,
+      "loss": 0.5871,
+      "step": 21160
+    },
+    {
+      "epoch": 3.422520410637782,
+      "grad_norm": 0.82564777135849,
+      "learning_rate": 0.0002,
+      "loss": 0.6116,
+      "step": 21170
+    },
+    {
+      "epoch": 3.4241370948185272,
+      "grad_norm": 1.004738688468933,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 21180
+    },
+    {
+      "epoch": 3.4257537789992725,
+      "grad_norm": 0.7545676827430725,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 21190
+    },
+    {
+      "epoch": 3.4273704631800177,
+      "grad_norm": 0.8918704390525818,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 21200
+    },
+    {
+      "epoch": 3.428987147360763,
+      "grad_norm": 0.8336876034736633,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 21210
+    },
+    {
+      "epoch": 3.430603831541508,
+      "grad_norm": 0.8928771018981934,
+      "learning_rate": 0.0002,
+      "loss": 0.5824,
+      "step": 21220
+    },
+    {
+      "epoch": 3.432220515722254,
+      "grad_norm": 0.7663705945014954,
+      "learning_rate": 0.0002,
+      "loss": 0.6468,
+      "step": 21230
+    },
+    {
+      "epoch": 3.433837199902999,
+      "grad_norm": 0.8392598628997803,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 21240
+    },
+    {
+      "epoch": 3.4354538840837443,
+      "grad_norm": 0.8819600343704224,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 21250
+    },
+    {
+      "epoch": 3.4370705682644895,
+      "grad_norm": 0.9124642014503479,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 21260
+    },
+    {
+      "epoch": 3.4386872524452348,
+      "grad_norm": 0.8329763412475586,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 21270
+    },
+    {
+      "epoch": 3.44030393662598,
+      "grad_norm": 0.9982839822769165,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 21280
+    },
+    {
+      "epoch": 3.4419206208067252,
+      "grad_norm": 0.9105954766273499,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 21290
+    },
+    {
+      "epoch": 3.443537304987471,
+      "grad_norm": 0.8182359337806702,
+      "learning_rate": 0.0002,
+      "loss": 0.6028,
+      "step": 21300
+    },
+    {
+      "epoch": 3.445153989168216,
+      "grad_norm": 1.0568904876708984,
+      "learning_rate": 0.0002,
+      "loss": 0.5991,
+      "step": 21310
+    },
+    {
+      "epoch": 3.4467706733489614,
+      "grad_norm": 0.968539834022522,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 21320
+    },
+    {
+      "epoch": 3.4483873575297066,
+      "grad_norm": 0.8774511218070984,
+      "learning_rate": 0.0002,
+      "loss": 0.6219,
+      "step": 21330
+    },
+    {
+      "epoch": 3.450004041710452,
+      "grad_norm": 0.7598156332969666,
+      "learning_rate": 0.0002,
+      "loss": 0.6438,
+      "step": 21340
+    },
+    {
+      "epoch": 3.451620725891197,
+      "grad_norm": 1.1012897491455078,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 21350
+    },
+    {
+      "epoch": 3.4532374100719423,
+      "grad_norm": 0.8040637373924255,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 21360
+    },
+    {
+      "epoch": 3.4548540942526875,
+      "grad_norm": 0.8497496247291565,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 21370
+    },
+    {
+      "epoch": 3.456470778433433,
+      "grad_norm": 0.8429915904998779,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 21380
+    },
+    {
+      "epoch": 3.4580874626141784,
+      "grad_norm": 0.8107112646102905,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 21390
+    },
+    {
+      "epoch": 3.4597041467949237,
+      "grad_norm": 1.00872004032135,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 21400
+    },
+    {
+      "epoch": 3.461320830975669,
+      "grad_norm": 0.8266542553901672,
+      "learning_rate": 0.0002,
+      "loss": 0.5712,
+      "step": 21410
+    },
+    {
+      "epoch": 3.462937515156414,
+      "grad_norm": 0.8972568511962891,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 21420
+    },
+    {
+      "epoch": 3.4645541993371594,
+      "grad_norm": 1.0781476497650146,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 21430
+    },
+    {
+      "epoch": 3.4661708835179046,
+      "grad_norm": 0.9571592807769775,
+      "learning_rate": 0.0002,
+      "loss": 0.6303,
+      "step": 21440
+    },
+    {
+      "epoch": 3.4677875676986503,
+      "grad_norm": 0.881547212600708,
+      "learning_rate": 0.0002,
+      "loss": 0.6309,
+      "step": 21450
+    },
+    {
+      "epoch": 3.4694042518793955,
+      "grad_norm": 0.6955338716506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 21460
+    },
+    {
+      "epoch": 3.4710209360601407,
+      "grad_norm": 0.901187539100647,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 21470
+    },
+    {
+      "epoch": 3.472637620240886,
+      "grad_norm": 0.7063511610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.639,
+      "step": 21480
+    },
+    {
+      "epoch": 3.474254304421631,
+      "grad_norm": 0.8462792038917542,
+      "learning_rate": 0.0002,
+      "loss": 0.6154,
+      "step": 21490
+    },
+    {
+      "epoch": 3.4758709886023764,
+      "grad_norm": 1.1861060857772827,
+      "learning_rate": 0.0002,
+      "loss": 0.61,
+      "step": 21500
+    },
+    {
+      "epoch": 3.4774876727831217,
+      "grad_norm": 0.70503169298172,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 21510
+    },
+    {
+      "epoch": 3.479104356963867,
+      "grad_norm": 0.9650066494941711,
+      "learning_rate": 0.0002,
+      "loss": 0.6475,
+      "step": 21520
+    },
+    {
+      "epoch": 3.4807210411446126,
+      "grad_norm": 1.0266852378845215,
+      "learning_rate": 0.0002,
+      "loss": 0.6452,
+      "step": 21530
+    },
+    {
+      "epoch": 3.482337725325358,
+      "grad_norm": 0.956372857093811,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 21540
+    },
+    {
+      "epoch": 3.483954409506103,
+      "grad_norm": 0.8848432898521423,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 21550
+    },
+    {
+      "epoch": 3.4855710936868483,
+      "grad_norm": 1.0805351734161377,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 21560
+    },
+    {
+      "epoch": 3.4871877778675935,
+      "grad_norm": 0.9279725551605225,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 21570
+    },
+    {
+      "epoch": 3.4888044620483387,
+      "grad_norm": 0.9049562215805054,
+      "learning_rate": 0.0002,
+      "loss": 0.6354,
+      "step": 21580
+    },
+    {
+      "epoch": 3.4904211462290844,
+      "grad_norm": 0.9619429111480713,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 21590
+    },
+    {
+      "epoch": 3.4920378304098296,
+      "grad_norm": 0.8508906960487366,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 21600
+    },
+    {
+      "epoch": 3.493654514590575,
+      "grad_norm": 0.8692502379417419,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 21610
+    },
+    {
+      "epoch": 3.49527119877132,
+      "grad_norm": 0.8187332153320312,
+      "learning_rate": 0.0002,
+      "loss": 0.5878,
+      "step": 21620
+    },
+    {
+      "epoch": 3.4968878829520653,
+      "grad_norm": 1.145400047302246,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 21630
+    },
+    {
+      "epoch": 3.4985045671328105,
+      "grad_norm": 0.8281388282775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 21640
+    },
+    {
+      "epoch": 3.500121251313556,
+      "grad_norm": 0.82256019115448,
+      "learning_rate": 0.0002,
+      "loss": 0.6624,
+      "step": 21650
+    },
+    {
+      "epoch": 3.501737935494301,
+      "grad_norm": 0.9315484762191772,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 21660
+    },
+    {
+      "epoch": 3.5033546196750462,
+      "grad_norm": 0.7626111507415771,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 21670
+    },
+    {
+      "epoch": 3.504971303855792,
+      "grad_norm": 0.9275059103965759,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 21680
+    },
+    {
+      "epoch": 3.506587988036537,
+      "grad_norm": 0.7906724810600281,
+      "learning_rate": 0.0002,
+      "loss": 0.64,
+      "step": 21690
+    },
+    {
+      "epoch": 3.5082046722172824,
+      "grad_norm": 0.8289761543273926,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 21700
+    },
+    {
+      "epoch": 3.5098213563980276,
+      "grad_norm": 0.8316431045532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 21710
+    },
+    {
+      "epoch": 3.511438040578773,
+      "grad_norm": 1.0451812744140625,
+      "learning_rate": 0.0002,
+      "loss": 0.619,
+      "step": 21720
+    },
+    {
+      "epoch": 3.513054724759518,
+      "grad_norm": 0.928252637386322,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 21730
+    },
+    {
+      "epoch": 3.5146714089402638,
+      "grad_norm": 0.7985895276069641,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 21740
+    },
+    {
+      "epoch": 3.516288093121009,
+      "grad_norm": 0.6740974187850952,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 21750
+    },
+    {
+      "epoch": 3.517904777301754,
+      "grad_norm": 0.8482223749160767,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 21760
+    },
+    {
+      "epoch": 3.5195214614824994,
+      "grad_norm": 0.889947772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.6277,
+      "step": 21770
+    },
+    {
+      "epoch": 3.5211381456632447,
+      "grad_norm": 0.8304598927497864,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 21780
+    },
+    {
+      "epoch": 3.52275482984399,
+      "grad_norm": 0.8002981543540955,
+      "learning_rate": 0.0002,
+      "loss": 0.6156,
+      "step": 21790
+    },
+    {
+      "epoch": 3.524371514024735,
+      "grad_norm": 0.8115083575248718,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21800
+    },
+    {
+      "epoch": 3.5259881982054804,
+      "grad_norm": 0.9715048670768738,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 21810
+    },
+    {
+      "epoch": 3.5276048823862256,
+      "grad_norm": 1.0910786390304565,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 21820
+    },
+    {
+      "epoch": 3.5292215665669713,
+      "grad_norm": 0.8438942432403564,
+      "learning_rate": 0.0002,
+      "loss": 0.6543,
+      "step": 21830
+    },
+    {
+      "epoch": 3.5308382507477165,
+      "grad_norm": 0.8813382983207703,
+      "learning_rate": 0.0002,
+      "loss": 0.6509,
+      "step": 21840
+    },
+    {
+      "epoch": 3.5324549349284617,
+      "grad_norm": 0.7092908024787903,
+      "learning_rate": 0.0002,
+      "loss": 0.6049,
+      "step": 21850
+    },
+    {
+      "epoch": 3.534071619109207,
+      "grad_norm": 0.8332187533378601,
+      "learning_rate": 0.0002,
+      "loss": 0.5678,
+      "step": 21860
+    },
+    {
+      "epoch": 3.535688303289952,
+      "grad_norm": 0.8958209156990051,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21870
+    },
+    {
+      "epoch": 3.5373049874706974,
+      "grad_norm": 0.824138879776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 21880
+    },
+    {
+      "epoch": 3.538921671651443,
+      "grad_norm": 0.8375158309936523,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 21890
+    },
+    {
+      "epoch": 3.5405383558321883,
+      "grad_norm": 1.0274608135223389,
+      "learning_rate": 0.0002,
+      "loss": 0.6019,
+      "step": 21900
+    },
+    {
+      "epoch": 3.5421550400129336,
+      "grad_norm": 0.7088932394981384,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 21910
+    },
+    {
+      "epoch": 3.543771724193679,
+      "grad_norm": 0.8172445297241211,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 21920
+    },
+    {
+      "epoch": 3.545388408374424,
+      "grad_norm": 0.9904135465621948,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 21930
+    },
+    {
+      "epoch": 3.5470050925551693,
+      "grad_norm": 0.9900432229042053,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 21940
+    },
+    {
+      "epoch": 3.5486217767359145,
+      "grad_norm": 0.8963301181793213,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 21950
+    },
+    {
+      "epoch": 3.5502384609166597,
+      "grad_norm": 0.8551464676856995,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 21960
+    },
+    {
+      "epoch": 3.551855145097405,
+      "grad_norm": 1.0916603803634644,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 21970
+    },
+    {
+      "epoch": 3.5534718292781506,
+      "grad_norm": 0.841598391532898,
+      "learning_rate": 0.0002,
+      "loss": 0.6523,
+      "step": 21980
+    },
+    {
+      "epoch": 3.555088513458896,
+      "grad_norm": 0.8566757440567017,
+      "learning_rate": 0.0002,
+      "loss": 0.617,
+      "step": 21990
+    },
+    {
+      "epoch": 3.556705197639641,
+      "grad_norm": 1.0145052671432495,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 22000
+    },
+    {
+      "epoch": 3.5583218818203863,
+      "grad_norm": 0.9293754696846008,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 22010
+    },
+    {
+      "epoch": 3.5599385660011316,
+      "grad_norm": 0.9568536281585693,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 22020
+    },
+    {
+      "epoch": 3.5615552501818772,
+      "grad_norm": 0.8613139986991882,
+      "learning_rate": 0.0002,
+      "loss": 0.641,
+      "step": 22030
+    },
+    {
+      "epoch": 3.5631719343626225,
+      "grad_norm": 0.8179237246513367,
+      "learning_rate": 0.0002,
+      "loss": 0.6496,
+      "step": 22040
+    },
+    {
+      "epoch": 3.5647886185433677,
+      "grad_norm": 0.9059830904006958,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 22050
+    },
+    {
+      "epoch": 3.566405302724113,
+      "grad_norm": 1.0068252086639404,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 22060
+    },
+    {
+      "epoch": 3.568021986904858,
+      "grad_norm": 0.9682072997093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6239,
+      "step": 22070
+    },
+    {
+      "epoch": 3.5696386710856034,
+      "grad_norm": 0.8514005541801453,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 22080
+    },
+    {
+      "epoch": 3.5712553552663486,
+      "grad_norm": 0.8327770829200745,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 22090
+    },
+    {
+      "epoch": 3.572872039447094,
+      "grad_norm": 1.024976372718811,
+      "learning_rate": 0.0002,
+      "loss": 0.5976,
+      "step": 22100
+    },
+    {
+      "epoch": 3.574488723627839,
+      "grad_norm": 0.7721174955368042,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 22110
+    },
+    {
+      "epoch": 3.5761054078085843,
+      "grad_norm": 1.0351054668426514,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 22120
+    },
+    {
+      "epoch": 3.57772209198933,
+      "grad_norm": 0.9680907130241394,
+      "learning_rate": 0.0002,
+      "loss": 0.6379,
+      "step": 22130
+    },
+    {
+      "epoch": 3.5793387761700752,
+      "grad_norm": 0.8016974925994873,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 22140
+    },
+    {
+      "epoch": 3.5809554603508205,
+      "grad_norm": 1.0109003782272339,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 22150
+    },
+    {
+      "epoch": 3.5825721445315657,
+      "grad_norm": 1.0473392009735107,
+      "learning_rate": 0.0002,
+      "loss": 0.6368,
+      "step": 22160
+    },
+    {
+      "epoch": 3.584188828712311,
+      "grad_norm": 0.8686613440513611,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 22170
+    },
+    {
+      "epoch": 3.5858055128930566,
+      "grad_norm": 0.869149923324585,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 22180
+    },
+    {
+      "epoch": 3.587422197073802,
+      "grad_norm": 0.9769062995910645,
+      "learning_rate": 0.0002,
+      "loss": 0.5895,
+      "step": 22190
+    },
+    {
+      "epoch": 3.589038881254547,
+      "grad_norm": 0.779636561870575,
+      "learning_rate": 0.0002,
+      "loss": 0.5939,
+      "step": 22200
+    },
+    {
+      "epoch": 3.5906555654352923,
+      "grad_norm": 0.9063841104507446,
+      "learning_rate": 0.0002,
+      "loss": 0.5875,
+      "step": 22210
+    },
+    {
+      "epoch": 3.5922722496160375,
+      "grad_norm": 0.9216037392616272,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 22220
+    },
+    {
+      "epoch": 3.5938889337967828,
+      "grad_norm": 1.0217336416244507,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 22230
+    },
+    {
+      "epoch": 3.595505617977528,
+      "grad_norm": 0.8513161540031433,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 22240
+    },
+    {
+      "epoch": 3.597122302158273,
+      "grad_norm": 0.8084813952445984,
+      "learning_rate": 0.0002,
+      "loss": 0.6301,
+      "step": 22250
+    },
+    {
+      "epoch": 3.5987389863390185,
+      "grad_norm": 0.8524802923202515,
+      "learning_rate": 0.0002,
+      "loss": 0.6197,
+      "step": 22260
+    },
+    {
+      "epoch": 3.600355670519764,
+      "grad_norm": 0.9356237649917603,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 22270
+    },
+    {
+      "epoch": 3.6019723547005094,
+      "grad_norm": 1.009600281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.628,
+      "step": 22280
+    },
+    {
+      "epoch": 3.6035890388812546,
+      "grad_norm": 0.9900581240653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 22290
+    },
+    {
+      "epoch": 3.605205723062,
+      "grad_norm": 1.062495231628418,
+      "learning_rate": 0.0002,
+      "loss": 0.5725,
+      "step": 22300
+    },
+    {
+      "epoch": 3.606822407242745,
+      "grad_norm": 0.8832381367683411,
+      "learning_rate": 0.0002,
+      "loss": 0.607,
+      "step": 22310
+    },
+    {
+      "epoch": 3.6084390914234903,
+      "grad_norm": 0.9284297823905945,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 22320
+    },
+    {
+      "epoch": 3.610055775604236,
+      "grad_norm": 1.2381829023361206,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 22330
+    },
+    {
+      "epoch": 3.611672459784981,
+      "grad_norm": 0.929434597492218,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 22340
+    },
+    {
+      "epoch": 3.6132891439657264,
+      "grad_norm": 0.9714490175247192,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 22350
+    },
+    {
+      "epoch": 3.6149058281464717,
+      "grad_norm": 0.808014988899231,
+      "learning_rate": 0.0002,
+      "loss": 0.6861,
+      "step": 22360
+    },
+    {
+      "epoch": 3.616522512327217,
+      "grad_norm": 1.0364398956298828,
+      "learning_rate": 0.0002,
+      "loss": 0.6428,
+      "step": 22370
+    },
+    {
+      "epoch": 3.618139196507962,
+      "grad_norm": 0.7858489751815796,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22380
+    },
+    {
+      "epoch": 3.6197558806887074,
+      "grad_norm": 0.9920870065689087,
+      "learning_rate": 0.0002,
+      "loss": 0.6214,
+      "step": 22390
+    },
+    {
+      "epoch": 3.6213725648694526,
+      "grad_norm": 0.9183220863342285,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 22400
+    },
+    {
+      "epoch": 3.622989249050198,
+      "grad_norm": 0.9826246500015259,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22410
+    },
+    {
+      "epoch": 3.6246059332309435,
+      "grad_norm": 0.8632931113243103,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 22420
+    },
+    {
+      "epoch": 3.6262226174116887,
+      "grad_norm": 0.8468965291976929,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 22430
+    },
+    {
+      "epoch": 3.627839301592434,
+      "grad_norm": 0.8466871976852417,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 22440
+    },
+    {
+      "epoch": 3.629455985773179,
+      "grad_norm": 0.9501169919967651,
+      "learning_rate": 0.0002,
+      "loss": 0.5941,
+      "step": 22450
+    },
+    {
+      "epoch": 3.6310726699539244,
+      "grad_norm": 0.8906720876693726,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 22460
+    },
+    {
+      "epoch": 3.6326893541346696,
+      "grad_norm": 0.7400227189064026,
+      "learning_rate": 0.0002,
+      "loss": 0.6928,
+      "step": 22470
+    },
+    {
+      "epoch": 3.6343060383154153,
+      "grad_norm": 0.9756355881690979,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22480
+    },
+    {
+      "epoch": 3.6359227224961606,
+      "grad_norm": 0.7504993081092834,
+      "learning_rate": 0.0002,
+      "loss": 0.6203,
+      "step": 22490
+    },
+    {
+      "epoch": 3.637539406676906,
+      "grad_norm": 0.9270039200782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 22500
+    },
+    {
+      "epoch": 3.639156090857651,
+      "grad_norm": 0.8841686844825745,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 22510
+    },
+    {
+      "epoch": 3.6407727750383962,
+      "grad_norm": 0.8533213138580322,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 22520
+    },
+    {
+      "epoch": 3.6423894592191415,
+      "grad_norm": 1.0052043199539185,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 22530
+    },
+    {
+      "epoch": 3.6440061433998867,
+      "grad_norm": 1.0323461294174194,
+      "learning_rate": 0.0002,
+      "loss": 0.6363,
+      "step": 22540
+    },
+    {
+      "epoch": 3.645622827580632,
+      "grad_norm": 0.8654312491416931,
+      "learning_rate": 0.0002,
+      "loss": 0.6545,
+      "step": 22550
+    },
+    {
+      "epoch": 3.647239511761377,
+      "grad_norm": 0.6400038003921509,
+      "learning_rate": 0.0002,
+      "loss": 0.6155,
+      "step": 22560
+    },
+    {
+      "epoch": 3.648856195942123,
+      "grad_norm": 0.8061298727989197,
+      "learning_rate": 0.0002,
+      "loss": 0.5829,
+      "step": 22570
+    },
+    {
+      "epoch": 3.650472880122868,
+      "grad_norm": 0.9257854223251343,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 22580
+    },
+    {
+      "epoch": 3.6520895643036133,
+      "grad_norm": 0.8439396619796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6409,
+      "step": 22590
+    },
+    {
+      "epoch": 3.6537062484843585,
+      "grad_norm": 0.7764544486999512,
+      "learning_rate": 0.0002,
+      "loss": 0.5996,
+      "step": 22600
+    },
+    {
+      "epoch": 3.6553229326651038,
+      "grad_norm": 1.125451683998108,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 22610
+    },
+    {
+      "epoch": 3.656939616845849,
+      "grad_norm": 0.7523018717765808,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 22620
+    },
+    {
+      "epoch": 3.6585563010265947,
+      "grad_norm": 1.071026086807251,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 22630
+    },
+    {
+      "epoch": 3.66017298520734,
+      "grad_norm": 0.945791482925415,
+      "learning_rate": 0.0002,
+      "loss": 0.6459,
+      "step": 22640
+    },
+    {
+      "epoch": 3.661789669388085,
+      "grad_norm": 0.8001811504364014,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 22650
+    },
+    {
+      "epoch": 3.6634063535688304,
+      "grad_norm": 0.9700816869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.6385,
+      "step": 22660
+    },
+    {
+      "epoch": 3.6650230377495756,
+      "grad_norm": 0.9053242206573486,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22670
+    },
+    {
+      "epoch": 3.666639721930321,
+      "grad_norm": 0.944362461566925,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 22680
+    },
+    {
+      "epoch": 3.668256406111066,
+      "grad_norm": 1.067489504814148,
+      "learning_rate": 0.0002,
+      "loss": 0.6235,
+      "step": 22690
+    },
+    {
+      "epoch": 3.6698730902918113,
+      "grad_norm": 1.0984995365142822,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 22700
+    },
+    {
+      "epoch": 3.6714897744725565,
+      "grad_norm": 0.9336317777633667,
+      "learning_rate": 0.0002,
+      "loss": 0.6717,
+      "step": 22710
+    },
+    {
+      "epoch": 3.673106458653302,
+      "grad_norm": 0.9261918663978577,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 22720
+    },
+    {
+      "epoch": 3.6747231428340474,
+      "grad_norm": 0.8648008704185486,
+      "learning_rate": 0.0002,
+      "loss": 0.6332,
+      "step": 22730
+    },
+    {
+      "epoch": 3.6763398270147927,
+      "grad_norm": 0.7225083708763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6576,
+      "step": 22740
+    },
+    {
+      "epoch": 3.677956511195538,
+      "grad_norm": 0.9258282780647278,
+      "learning_rate": 0.0002,
+      "loss": 0.6406,
+      "step": 22750
+    },
+    {
+      "epoch": 3.679573195376283,
+      "grad_norm": 0.70876145362854,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 22760
+    },
+    {
+      "epoch": 3.681189879557029,
+      "grad_norm": 0.8780210018157959,
+      "learning_rate": 0.0002,
+      "loss": 0.6821,
+      "step": 22770
+    },
+    {
+      "epoch": 3.682806563737774,
+      "grad_norm": 0.8075440526008606,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22780
+    },
+    {
+      "epoch": 3.6844232479185193,
+      "grad_norm": 0.8503130674362183,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22790
+    },
+    {
+      "epoch": 3.6860399320992645,
+      "grad_norm": 0.8413618206977844,
+      "learning_rate": 0.0002,
+      "loss": 0.6082,
+      "step": 22800
+    },
+    {
+      "epoch": 3.6876566162800097,
+      "grad_norm": 0.8675165176391602,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 22810
+    },
+    {
+      "epoch": 3.689273300460755,
+      "grad_norm": 0.8235884308815002,
+      "learning_rate": 0.0002,
+      "loss": 0.6157,
+      "step": 22820
+    },
+    {
+      "epoch": 3.6908899846415,
+      "grad_norm": 0.9477725625038147,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 22830
+    },
+    {
+      "epoch": 3.6925066688222454,
+      "grad_norm": 0.7883533835411072,
+      "learning_rate": 0.0002,
+      "loss": 0.6481,
+      "step": 22840
+    },
+    {
+      "epoch": 3.6941233530029907,
+      "grad_norm": 1.047913908958435,
+      "learning_rate": 0.0002,
+      "loss": 0.5872,
+      "step": 22850
+    },
+    {
+      "epoch": 3.695740037183736,
+      "grad_norm": 0.9171528816223145,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 22860
+    },
+    {
+      "epoch": 3.6973567213644816,
+      "grad_norm": 0.9338192343711853,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 22870
+    },
+    {
+      "epoch": 3.698973405545227,
+      "grad_norm": 0.8799443244934082,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 22880
+    },
+    {
+      "epoch": 3.700590089725972,
+      "grad_norm": 0.8515434861183167,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 22890
+    },
+    {
+      "epoch": 3.7022067739067173,
+      "grad_norm": 0.7805591821670532,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 22900
+    },
+    {
+      "epoch": 3.7038234580874625,
+      "grad_norm": 0.8470911979675293,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 22910
+    },
+    {
+      "epoch": 3.705440142268208,
+      "grad_norm": 0.9452309012413025,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 22920
+    },
+    {
+      "epoch": 3.7070568264489534,
+      "grad_norm": 0.950243353843689,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 22930
+    },
+    {
+      "epoch": 3.7086735106296986,
+      "grad_norm": 0.7882499098777771,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 22940
+    },
+    {
+      "epoch": 3.710290194810444,
+      "grad_norm": 0.8307787775993347,
+      "learning_rate": 0.0002,
+      "loss": 0.6462,
+      "step": 22950
+    },
+    {
+      "epoch": 3.711906878991189,
+      "grad_norm": 1.0970630645751953,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 22960
+    },
+    {
+      "epoch": 3.7135235631719343,
+      "grad_norm": 0.8269566297531128,
+      "learning_rate": 0.0002,
+      "loss": 0.6281,
+      "step": 22970
+    },
+    {
+      "epoch": 3.7151402473526796,
+      "grad_norm": 0.8306704759597778,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22980
+    },
+    {
+      "epoch": 3.716756931533425,
+      "grad_norm": 0.9710225462913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6418,
+      "step": 22990
+    },
+    {
+      "epoch": 3.71837361571417,
+      "grad_norm": 0.8890530467033386,
+      "learning_rate": 0.0002,
+      "loss": 0.6639,
+      "step": 23000
+    },
+    {
+      "epoch": 3.7199902998949153,
+      "grad_norm": 0.883522629737854,
+      "learning_rate": 0.0002,
+      "loss": 0.6084,
+      "step": 23010
+    },
+    {
+      "epoch": 3.721606984075661,
+      "grad_norm": 0.8662652373313904,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 23020
+    },
+    {
+      "epoch": 3.723223668256406,
+      "grad_norm": 0.7228406667709351,
+      "learning_rate": 0.0002,
+      "loss": 0.6266,
+      "step": 23030
+    },
+    {
+      "epoch": 3.7248403524371514,
+      "grad_norm": 1.060792088508606,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23040
+    },
+    {
+      "epoch": 3.7264570366178966,
+      "grad_norm": 1.0119613409042358,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 23050
+    },
+    {
+      "epoch": 3.728073720798642,
+      "grad_norm": 0.9212996959686279,
+      "learning_rate": 0.0002,
+      "loss": 0.6466,
+      "step": 23060
+    },
+    {
+      "epoch": 3.7296904049793875,
+      "grad_norm": 0.925690233707428,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 23070
+    },
+    {
+      "epoch": 3.7313070891601328,
+      "grad_norm": 0.8323310613632202,
+      "learning_rate": 0.0002,
+      "loss": 0.615,
+      "step": 23080
+    },
+    {
+      "epoch": 3.732923773340878,
+      "grad_norm": 0.8966048955917358,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 23090
+    },
+    {
+      "epoch": 3.7345404575216232,
+      "grad_norm": 0.8995837569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23100
+    },
+    {
+      "epoch": 3.7361571417023685,
+      "grad_norm": 0.8748890161514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23110
+    },
+    {
+      "epoch": 3.7377738258831137,
+      "grad_norm": 0.7985540628433228,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 23120
+    },
+    {
+      "epoch": 3.739390510063859,
+      "grad_norm": 1.0240917205810547,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 23130
+    },
+    {
+      "epoch": 3.741007194244604,
+      "grad_norm": 0.9181789755821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 23140
+    },
+    {
+      "epoch": 3.7426238784253494,
+      "grad_norm": 0.8896583914756775,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 23150
+    },
+    {
+      "epoch": 3.744240562606095,
+      "grad_norm": 0.8635515570640564,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 23160
+    },
+    {
+      "epoch": 3.7458572467868403,
+      "grad_norm": 0.8873575329780579,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 23170
+    },
+    {
+      "epoch": 3.7474739309675855,
+      "grad_norm": 0.9807148575782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23180
+    },
+    {
+      "epoch": 3.7490906151483308,
+      "grad_norm": 0.900477945804596,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 23190
+    },
+    {
+      "epoch": 3.750707299329076,
+      "grad_norm": 0.9379992485046387,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23200
+    },
+    {
+      "epoch": 3.752323983509821,
+      "grad_norm": 0.9649890661239624,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 23210
+    },
+    {
+      "epoch": 3.753940667690567,
+      "grad_norm": 0.824442446231842,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 23220
+    },
+    {
+      "epoch": 3.755557351871312,
+      "grad_norm": 0.8896150588989258,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 23230
+    },
+    {
+      "epoch": 3.7571740360520574,
+      "grad_norm": 0.751249372959137,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 23240
+    },
+    {
+      "epoch": 3.7587907202328026,
+      "grad_norm": 0.9392193555831909,
+      "learning_rate": 0.0002,
+      "loss": 0.629,
+      "step": 23250
+    },
+    {
+      "epoch": 3.760407404413548,
+      "grad_norm": 0.9284586310386658,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 23260
+    },
+    {
+      "epoch": 3.762024088594293,
+      "grad_norm": 0.7738175392150879,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23270
+    },
+    {
+      "epoch": 3.7636407727750383,
+      "grad_norm": 0.9252978563308716,
+      "learning_rate": 0.0002,
+      "loss": 0.6743,
+      "step": 23280
+    },
+    {
+      "epoch": 3.7652574569557835,
+      "grad_norm": 0.9501895904541016,
+      "learning_rate": 0.0002,
+      "loss": 0.5984,
+      "step": 23290
+    },
+    {
+      "epoch": 3.7668741411365287,
+      "grad_norm": 0.9416276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 23300
+    },
+    {
+      "epoch": 3.7684908253172744,
+      "grad_norm": 0.7076631784439087,
+      "learning_rate": 0.0002,
+      "loss": 0.6507,
+      "step": 23310
+    },
+    {
+      "epoch": 3.7701075094980196,
+      "grad_norm": 0.9864492416381836,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 23320
+    },
+    {
+      "epoch": 3.771724193678765,
+      "grad_norm": 0.8450456261634827,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 23330
+    },
+    {
+      "epoch": 3.77334087785951,
+      "grad_norm": 1.0768941640853882,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23340
+    },
+    {
+      "epoch": 3.7749575620402553,
+      "grad_norm": 0.9956819415092468,
+      "learning_rate": 0.0002,
+      "loss": 0.6408,
+      "step": 23350
+    },
+    {
+      "epoch": 3.7765742462210006,
+      "grad_norm": 0.9234658479690552,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 23360
+    },
+    {
+      "epoch": 3.7781909304017463,
+      "grad_norm": 1.0993858575820923,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23370
+    },
+    {
+      "epoch": 3.7798076145824915,
+      "grad_norm": 0.923159658908844,
+      "learning_rate": 0.0002,
+      "loss": 0.6391,
+      "step": 23380
+    },
+    {
+      "epoch": 3.7814242987632367,
+      "grad_norm": 0.9311541318893433,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23390
+    },
+    {
+      "epoch": 3.783040982943982,
+      "grad_norm": 0.919681191444397,
+      "learning_rate": 0.0002,
+      "loss": 0.6535,
+      "step": 23400
+    },
+    {
+      "epoch": 3.784657667124727,
+      "grad_norm": 1.7406195402145386,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 23410
+    },
+    {
+      "epoch": 3.7862743513054724,
+      "grad_norm": 0.7789074182510376,
+      "learning_rate": 0.0002,
+      "loss": 0.657,
+      "step": 23420
+    },
+    {
+      "epoch": 3.7878910354862176,
+      "grad_norm": 0.8302814960479736,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23430
+    },
+    {
+      "epoch": 3.789507719666963,
+      "grad_norm": 0.8089349269866943,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23440
+    },
+    {
+      "epoch": 3.791124403847708,
+      "grad_norm": 0.9006284475326538,
+      "learning_rate": 0.0002,
+      "loss": 0.6682,
+      "step": 23450
+    },
+    {
+      "epoch": 3.7927410880284538,
+      "grad_norm": 0.8426766991615295,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 23460
+    },
+    {
+      "epoch": 3.794357772209199,
+      "grad_norm": 1.2576252222061157,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 23470
+    },
+    {
+      "epoch": 3.7959744563899442,
+      "grad_norm": 1.0307610034942627,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 23480
+    },
+    {
+      "epoch": 3.7975911405706895,
+      "grad_norm": 0.8525972962379456,
+      "learning_rate": 0.0002,
+      "loss": 0.6262,
+      "step": 23490
+    },
+    {
+      "epoch": 3.7992078247514347,
+      "grad_norm": 1.159039855003357,
+      "learning_rate": 0.0002,
+      "loss": 0.6757,
+      "step": 23500
+    },
+    {
+      "epoch": 3.80082450893218,
+      "grad_norm": 1.4193549156188965,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23510
+    },
+    {
+      "epoch": 3.8024411931129256,
+      "grad_norm": 0.8245543837547302,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 23520
+    },
+    {
+      "epoch": 3.804057877293671,
+      "grad_norm": 0.8847230076789856,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23530
+    },
+    {
+      "epoch": 3.805674561474416,
+      "grad_norm": 0.9574624300003052,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 23540
+    },
+    {
+      "epoch": 3.8072912456551613,
+      "grad_norm": 1.048020601272583,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 23550
+    },
+    {
+      "epoch": 3.8089079298359065,
+      "grad_norm": 0.8302255868911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6497,
+      "step": 23560
+    },
+    {
+      "epoch": 3.8105246140166518,
+      "grad_norm": 0.8269215822219849,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 23570
+    },
+    {
+      "epoch": 3.812141298197397,
+      "grad_norm": 0.9375753402709961,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 23580
+    },
+    {
+      "epoch": 3.8137579823781422,
+      "grad_norm": 1.0234097242355347,
+      "learning_rate": 0.0002,
+      "loss": 0.6132,
+      "step": 23590
+    },
+    {
+      "epoch": 3.8153746665588875,
+      "grad_norm": 0.8978445529937744,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23600
+    },
+    {
+      "epoch": 3.816991350739633,
+      "grad_norm": 0.7929515838623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 23610
+    },
+    {
+      "epoch": 3.8186080349203784,
+      "grad_norm": 1.3255881071090698,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23620
+    },
+    {
+      "epoch": 3.8202247191011236,
+      "grad_norm": 0.9188598990440369,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 23630
+    },
+    {
+      "epoch": 3.821841403281869,
+      "grad_norm": 0.8811675906181335,
+      "learning_rate": 0.0002,
+      "loss": 0.6152,
+      "step": 23640
+    },
+    {
+      "epoch": 3.823458087462614,
+      "grad_norm": 0.8061038255691528,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 23650
+    },
+    {
+      "epoch": 3.8250747716433597,
+      "grad_norm": 0.9975376129150391,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 23660
+    },
+    {
+      "epoch": 3.826691455824105,
+      "grad_norm": 0.8036105036735535,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 23670
+    },
+    {
+      "epoch": 3.82830814000485,
+      "grad_norm": 0.7401984333992004,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 23680
+    },
+    {
+      "epoch": 3.8299248241855954,
+      "grad_norm": 0.829753041267395,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 23690
+    },
+    {
+      "epoch": 3.8315415083663407,
+      "grad_norm": 0.8753240704536438,
+      "learning_rate": 0.0002,
+      "loss": 0.6611,
+      "step": 23700
+    },
+    {
+      "epoch": 3.833158192547086,
+      "grad_norm": 0.8157842755317688,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 23710
+    },
+    {
+      "epoch": 3.834774876727831,
+      "grad_norm": 0.6183798909187317,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 23720
+    },
+    {
+      "epoch": 3.8363915609085764,
+      "grad_norm": 0.9548442363739014,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 23730
+    },
+    {
+      "epoch": 3.8380082450893216,
+      "grad_norm": 0.8319669961929321,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23740
+    },
+    {
+      "epoch": 3.839624929270067,
+      "grad_norm": 0.9718693494796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 23750
+    },
+    {
+      "epoch": 3.8412416134508125,
+      "grad_norm": 0.8672235012054443,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 23760
+    },
+    {
+      "epoch": 3.8428582976315577,
+      "grad_norm": 1.1210707426071167,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 23770
+    },
+    {
+      "epoch": 3.844474981812303,
+      "grad_norm": 0.9177767634391785,
+      "learning_rate": 0.0002,
+      "loss": 0.5923,
+      "step": 23780
+    },
+    {
+      "epoch": 3.846091665993048,
+      "grad_norm": 0.8714171648025513,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 23790
+    },
+    {
+      "epoch": 3.8477083501737934,
+      "grad_norm": 1.1853246688842773,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 23800
+    },
+    {
+      "epoch": 3.849325034354539,
+      "grad_norm": 0.8091260194778442,
+      "learning_rate": 0.0002,
+      "loss": 0.6144,
+      "step": 23810
+    },
+    {
+      "epoch": 3.8509417185352843,
+      "grad_norm": 0.9710774421691895,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23820
+    },
+    {
+      "epoch": 3.8525584027160296,
+      "grad_norm": 0.7648707628250122,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23830
+    },
+    {
+      "epoch": 3.854175086896775,
+      "grad_norm": 0.7809253931045532,
+      "learning_rate": 0.0002,
+      "loss": 0.6013,
+      "step": 23840
+    },
+    {
+      "epoch": 3.85579177107752,
+      "grad_norm": 0.8337951898574829,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 23850
+    },
+    {
+      "epoch": 3.8574084552582653,
+      "grad_norm": 0.9271913170814514,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23860
+    },
+    {
+      "epoch": 3.8590251394390105,
+      "grad_norm": 0.985334038734436,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 23870
+    },
+    {
+      "epoch": 3.8606418236197557,
+      "grad_norm": 0.8458583354949951,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 23880
+    },
+    {
+      "epoch": 3.862258507800501,
+      "grad_norm": 1.015348196029663,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 23890
+    },
+    {
+      "epoch": 3.8638751919812466,
+      "grad_norm": 1.0121688842773438,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23900
+    },
+    {
+      "epoch": 3.865491876161992,
+      "grad_norm": 0.8883971571922302,
+      "learning_rate": 0.0002,
+      "loss": 0.5921,
+      "step": 23910
+    },
+    {
+      "epoch": 3.867108560342737,
+      "grad_norm": 1.028086543083191,
+      "learning_rate": 0.0002,
+      "loss": 0.6597,
+      "step": 23920
+    },
+    {
+      "epoch": 3.8687252445234823,
+      "grad_norm": 0.9645734429359436,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 23930
+    },
+    {
+      "epoch": 3.8703419287042276,
+      "grad_norm": 0.8235350251197815,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 23940
+    },
+    {
+      "epoch": 3.871958612884973,
+      "grad_norm": 1.0298916101455688,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23950
+    },
+    {
+      "epoch": 3.8735752970657185,
+      "grad_norm": 1.0063377618789673,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 23960
+    },
+    {
+      "epoch": 3.8751919812464637,
+      "grad_norm": 0.9230626821517944,
+      "learning_rate": 0.0002,
+      "loss": 0.6234,
+      "step": 23970
+    },
+    {
+      "epoch": 3.876808665427209,
+      "grad_norm": 0.9243063926696777,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 23980
+    },
+    {
+      "epoch": 3.878425349607954,
+      "grad_norm": 1.0211291313171387,
+      "learning_rate": 0.0002,
+      "loss": 0.6035,
+      "step": 23990
+    },
+    {
+      "epoch": 3.8800420337886994,
+      "grad_norm": 0.7800535559654236,
+      "learning_rate": 0.0002,
+      "loss": 0.6351,
+      "step": 24000
+    },
+    {
+      "epoch": 3.8816587179694446,
+      "grad_norm": 0.7904248833656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 24010
+    },
+    {
+      "epoch": 3.88327540215019,
+      "grad_norm": 1.1975988149642944,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 24020
+    },
+    {
+      "epoch": 3.884892086330935,
+      "grad_norm": 1.0626593828201294,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 24030
+    },
+    {
+      "epoch": 3.8865087705116803,
+      "grad_norm": 0.9012193083763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 24040
+    },
+    {
+      "epoch": 3.888125454692426,
+      "grad_norm": 1.1159172058105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6786,
+      "step": 24050
+    },
+    {
+      "epoch": 3.889742138873171,
+      "grad_norm": 1.276838779449463,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 24060
+    },
+    {
+      "epoch": 3.8913588230539164,
+      "grad_norm": 0.8467690348625183,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 24070
+    },
+    {
+      "epoch": 3.8929755072346617,
+      "grad_norm": 0.9862841963768005,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 24080
+    },
+    {
+      "epoch": 3.894592191415407,
+      "grad_norm": 0.7134621739387512,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 24090
+    },
+    {
+      "epoch": 3.896208875596152,
+      "grad_norm": 0.8178175091743469,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 24100
+    },
+    {
+      "epoch": 3.897825559776898,
+      "grad_norm": 0.9229172468185425,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 24110
+    },
+    {
+      "epoch": 3.899442243957643,
+      "grad_norm": 1.0878316164016724,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 24120
+    },
+    {
+      "epoch": 3.9010589281383883,
+      "grad_norm": 0.971645712852478,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24130
+    },
+    {
+      "epoch": 3.9026756123191335,
+      "grad_norm": 0.8862188458442688,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 24140
+    },
+    {
+      "epoch": 3.9042922964998787,
+      "grad_norm": 0.9126982688903809,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 24150
+    },
+    {
+      "epoch": 3.905908980680624,
+      "grad_norm": 0.8833470940589905,
+      "learning_rate": 0.0002,
+      "loss": 0.6734,
+      "step": 24160
+    },
+    {
+      "epoch": 3.907525664861369,
+      "grad_norm": 0.8320947885513306,
+      "learning_rate": 0.0002,
+      "loss": 0.5832,
+      "step": 24170
+    },
+    {
+      "epoch": 3.9091423490421144,
+      "grad_norm": 0.9156602025032043,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 24180
+    },
+    {
+      "epoch": 3.9107590332228597,
+      "grad_norm": 1.029181957244873,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 24190
+    },
+    {
+      "epoch": 3.9123757174036053,
+      "grad_norm": 0.9052802324295044,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 24200
+    },
+    {
+      "epoch": 3.9139924015843506,
+      "grad_norm": 0.8847255110740662,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 24210
+    },
+    {
+      "epoch": 3.915609085765096,
+      "grad_norm": 0.9642062187194824,
+      "learning_rate": 0.0002,
+      "loss": 0.6343,
+      "step": 24220
+    },
+    {
+      "epoch": 3.917225769945841,
+      "grad_norm": 0.8629093766212463,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 24230
+    },
+    {
+      "epoch": 3.9188424541265863,
+      "grad_norm": 0.8674976825714111,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 24240
+    },
+    {
+      "epoch": 3.9204591383073315,
+      "grad_norm": 1.104846477508545,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 24250
+    },
+    {
+      "epoch": 3.922075822488077,
+      "grad_norm": 1.0874955654144287,
+      "learning_rate": 0.0002,
+      "loss": 0.6501,
+      "step": 24260
+    },
+    {
+      "epoch": 3.9236925066688224,
+      "grad_norm": 0.8689812421798706,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 24270
+    },
+    {
+      "epoch": 3.9253091908495676,
+      "grad_norm": 0.9724617004394531,
+      "learning_rate": 0.0002,
+      "loss": 0.5893,
+      "step": 24280
+    },
+    {
+      "epoch": 3.926925875030313,
+      "grad_norm": 0.9165538549423218,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24290
+    },
+    {
+      "epoch": 3.928542559211058,
+      "grad_norm": 0.9307710528373718,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 24300
+    },
+    {
+      "epoch": 3.9301592433918033,
+      "grad_norm": 0.8589295148849487,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 24310
+    },
+    {
+      "epoch": 3.9317759275725486,
+      "grad_norm": 0.9151099920272827,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 24320
+    },
+    {
+      "epoch": 3.933392611753294,
+      "grad_norm": 0.9633517265319824,
+      "learning_rate": 0.0002,
+      "loss": 0.7075,
+      "step": 24330
+    },
+    {
+      "epoch": 3.935009295934039,
+      "grad_norm": 0.9521116018295288,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24340
+    },
+    {
+      "epoch": 3.9366259801147847,
+      "grad_norm": 0.8366776704788208,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 24350
+    },
+    {
+      "epoch": 3.93824266429553,
+      "grad_norm": 0.8972663283348083,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 24360
+    },
+    {
+      "epoch": 3.939859348476275,
+      "grad_norm": 0.8102919459342957,
+      "learning_rate": 0.0002,
+      "loss": 0.661,
+      "step": 24370
+    },
+    {
+      "epoch": 3.9414760326570204,
+      "grad_norm": 0.8189975023269653,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 24380
+    },
+    {
+      "epoch": 3.9430927168377656,
+      "grad_norm": 0.9569464921951294,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 24390
+    },
+    {
+      "epoch": 3.9447094010185113,
+      "grad_norm": 0.7459101676940918,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 24400
+    },
+    {
+      "epoch": 3.9463260851992565,
+      "grad_norm": 0.8536974787712097,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 24410
+    },
+    {
+      "epoch": 3.9479427693800018,
+      "grad_norm": 0.8763698935508728,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 24420
+    },
+    {
+      "epoch": 3.949559453560747,
+      "grad_norm": 0.9381106495857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6478,
+      "step": 24430
+    },
+    {
+      "epoch": 3.9511761377414922,
+      "grad_norm": 0.934440016746521,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 24440
+    },
+    {
+      "epoch": 3.9527928219222375,
+      "grad_norm": 0.903918981552124,
+      "learning_rate": 0.0002,
+      "loss": 0.6393,
+      "step": 24450
+    },
+    {
+      "epoch": 3.9544095061029827,
+      "grad_norm": 0.8771953582763672,
+      "learning_rate": 0.0002,
+      "loss": 0.6175,
+      "step": 24460
+    },
+    {
+      "epoch": 3.956026190283728,
+      "grad_norm": 1.0375410318374634,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 24470
+    },
+    {
+      "epoch": 3.957642874464473,
+      "grad_norm": 0.9439185261726379,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 24480
+    },
+    {
+      "epoch": 3.9592595586452184,
+      "grad_norm": 0.935467004776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 24490
+    },
+    {
+      "epoch": 3.960876242825964,
+      "grad_norm": 0.6900772452354431,
+      "learning_rate": 0.0002,
+      "loss": 0.6437,
+      "step": 24500
+    },
+    {
+      "epoch": 3.9624929270067093,
+      "grad_norm": 1.0172916650772095,
+      "learning_rate": 0.0002,
+      "loss": 0.6445,
+      "step": 24510
+    },
+    {
+      "epoch": 3.9641096111874545,
+      "grad_norm": 0.9167046546936035,
+      "learning_rate": 0.0002,
+      "loss": 0.6308,
+      "step": 24520
+    },
+    {
+      "epoch": 3.9657262953681998,
+      "grad_norm": 0.7230527997016907,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 24530
+    },
+    {
+      "epoch": 3.967342979548945,
+      "grad_norm": 0.8980403542518616,
+      "learning_rate": 0.0002,
+      "loss": 0.6564,
+      "step": 24540
+    },
+    {
+      "epoch": 3.9689596637296907,
+      "grad_norm": 0.8555465936660767,
+      "learning_rate": 0.0002,
+      "loss": 0.6099,
+      "step": 24550
+    },
+    {
+      "epoch": 3.970576347910436,
+      "grad_norm": 0.7825445532798767,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 24560
+    },
+    {
+      "epoch": 3.972193032091181,
+      "grad_norm": 0.7273133993148804,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 24570
+    },
+    {
+      "epoch": 3.9738097162719264,
+      "grad_norm": 0.9612047672271729,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 24580
+    },
+    {
+      "epoch": 3.9754264004526716,
+      "grad_norm": 0.9865460991859436,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 24590
+    },
+    {
+      "epoch": 3.977043084633417,
+      "grad_norm": 0.8638762831687927,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 24600
+    },
+    {
+      "epoch": 3.978659768814162,
+      "grad_norm": 1.0096198320388794,
+      "learning_rate": 0.0002,
+      "loss": 0.6097,
+      "step": 24610
+    },
+    {
+      "epoch": 3.9802764529949073,
+      "grad_norm": 0.8475532531738281,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 24620
+    },
+    {
+      "epoch": 3.9818931371756525,
+      "grad_norm": 0.9696195721626282,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 24630
+    },
+    {
+      "epoch": 3.9835098213563978,
+      "grad_norm": 0.7499843239784241,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 24640
+    },
+    {
+      "epoch": 3.9851265055371434,
+      "grad_norm": 0.8865424990653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6054,
+      "step": 24650
+    },
+    {
+      "epoch": 3.9867431897178887,
+      "grad_norm": 0.8089959025382996,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 24660
+    },
+    {
+      "epoch": 3.988359873898634,
+      "grad_norm": 0.6946012377738953,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 24670
+    },
+    {
+      "epoch": 3.989976558079379,
+      "grad_norm": 0.7991759181022644,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 24680
+    },
+    {
+      "epoch": 3.9915932422601244,
+      "grad_norm": 0.8803931474685669,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 24690
+    },
+    {
+      "epoch": 3.99320992644087,
+      "grad_norm": 0.8848299980163574,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 24700
+    },
+    {
+      "epoch": 3.9948266106216153,
+      "grad_norm": 0.7448889017105103,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 24710
+    },
+    {
+      "epoch": 3.9964432948023605,
+      "grad_norm": 0.9361620545387268,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24720
+    },
+    {
+      "epoch": 3.9980599789831057,
+      "grad_norm": 0.9958081245422363,
+      "learning_rate": 0.0002,
+      "loss": 0.5917,
+      "step": 24730
+    },
+    {
+      "epoch": 3.999676663163851,
+      "grad_norm": 1.026004672050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6567,
+      "step": 24740
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.1524168252944946,
+      "eval_runtime": 122.1585,
+      "eval_samples_per_second": 6.0,
+      "eval_steps_per_second": 0.753,
+      "step": 24742
+    },
+    {
+      "epoch": 4.001293347344596,
+      "grad_norm": 1.0664808750152588,
+      "learning_rate": 0.0002,
+      "loss": 0.6057,
+      "step": 24750
+    },
+    {
+      "epoch": 4.002910031525341,
+      "grad_norm": 1.0113720893859863,
+      "learning_rate": 0.0002,
+      "loss": 0.5644,
+      "step": 24760
+    },
+    {
+      "epoch": 4.004526715706087,
+      "grad_norm": 0.991486668586731,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 24770
+    },
+    {
+      "epoch": 4.006143399886832,
+      "grad_norm": 0.951754629611969,
+      "learning_rate": 0.0002,
+      "loss": 0.508,
+      "step": 24780
+    },
+    {
+      "epoch": 4.007760084067577,
+      "grad_norm": 1.13059401512146,
+      "learning_rate": 0.0002,
+      "loss": 0.5314,
+      "step": 24790
+    },
+    {
+      "epoch": 4.009376768248322,
+      "grad_norm": 0.9343926310539246,
+      "learning_rate": 0.0002,
+      "loss": 0.5323,
+      "step": 24800
+    },
+    {
+      "epoch": 4.010993452429068,
+      "grad_norm": 1.0680590867996216,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 24810
+    },
+    {
+      "epoch": 4.012610136609814,
+      "grad_norm": 1.0022706985473633,
+      "learning_rate": 0.0002,
+      "loss": 0.513,
+      "step": 24820
+    },
+    {
+      "epoch": 4.014226820790559,
+      "grad_norm": 1.0285297632217407,
+      "learning_rate": 0.0002,
+      "loss": 0.543,
+      "step": 24830
+    },
+    {
+      "epoch": 4.015843504971304,
+      "grad_norm": 0.8347002863883972,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 24840
+    },
+    {
+      "epoch": 4.017460189152049,
+      "grad_norm": 0.9675396680831909,
+      "learning_rate": 0.0002,
+      "loss": 0.5655,
+      "step": 24850
+    },
+    {
+      "epoch": 4.019076873332795,
+      "grad_norm": 0.9238511323928833,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 24860
+    },
+    {
+      "epoch": 4.02069355751354,
+      "grad_norm": 1.1576941013336182,
+      "learning_rate": 0.0002,
+      "loss": 0.5327,
+      "step": 24870
+    },
+    {
+      "epoch": 4.022310241694285,
+      "grad_norm": 0.8583757281303406,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 24880
+    },
+    {
+      "epoch": 4.02392692587503,
+      "grad_norm": 0.9816817045211792,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 24890
+    },
+    {
+      "epoch": 4.0255436100557755,
+      "grad_norm": 0.955073893070221,
+      "learning_rate": 0.0002,
+      "loss": 0.5605,
+      "step": 24900
+    },
+    {
+      "epoch": 4.027160294236521,
+      "grad_norm": 1.1054974794387817,
+      "learning_rate": 0.0002,
+      "loss": 0.4896,
+      "step": 24910
+    },
+    {
+      "epoch": 4.028776978417266,
+      "grad_norm": 1.1240060329437256,
+      "learning_rate": 0.0002,
+      "loss": 0.5246,
+      "step": 24920
+    },
+    {
+      "epoch": 4.030393662598011,
+      "grad_norm": 0.9512825012207031,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 24930
+    },
+    {
+      "epoch": 4.0320103467787565,
+      "grad_norm": 0.85965496301651,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 24940
+    },
+    {
+      "epoch": 4.033627030959502,
+      "grad_norm": 0.9378061294555664,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 24950
+    },
+    {
+      "epoch": 4.035243715140247,
+      "grad_norm": 0.9655424356460571,
+      "learning_rate": 0.0002,
+      "loss": 0.5008,
+      "step": 24960
+    },
+    {
+      "epoch": 4.036860399320993,
+      "grad_norm": 1.1393707990646362,
+      "learning_rate": 0.0002,
+      "loss": 0.5538,
+      "step": 24970
+    },
+    {
+      "epoch": 4.038477083501738,
+      "grad_norm": 1.0220451354980469,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 24980
+    },
+    {
+      "epoch": 4.0400937676824835,
+      "grad_norm": 0.9785808324813843,
+      "learning_rate": 0.0002,
+      "loss": 0.5813,
+      "step": 24990
+    },
+    {
+      "epoch": 4.041710451863229,
+      "grad_norm": 1.0257649421691895,
+      "learning_rate": 0.0002,
+      "loss": 0.5153,
+      "step": 25000
+    },
+    {
+      "epoch": 4.043327136043974,
+      "grad_norm": 0.9737892150878906,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 25010
+    },
+    {
+      "epoch": 4.044943820224719,
+      "grad_norm": 0.7416959404945374,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 25020
+    },
+    {
+      "epoch": 4.046560504405464,
+      "grad_norm": 0.7909596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 25030
+    },
+    {
+      "epoch": 4.04817718858621,
+      "grad_norm": 0.8923130631446838,
+      "learning_rate": 0.0002,
+      "loss": 0.5265,
+      "step": 25040
+    },
+    {
+      "epoch": 4.049793872766955,
+      "grad_norm": 0.9044941663742065,
+      "learning_rate": 0.0002,
+      "loss": 0.5035,
+      "step": 25050
+    },
+    {
+      "epoch": 4.0514105569477,
+      "grad_norm": 0.866352379322052,
+      "learning_rate": 0.0002,
+      "loss": 0.5135,
+      "step": 25060
+    },
+    {
+      "epoch": 4.053027241128445,
+      "grad_norm": 1.544549822807312,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 25070
+    },
+    {
+      "epoch": 4.054643925309191,
+      "grad_norm": 0.8426995277404785,
+      "learning_rate": 0.0002,
+      "loss": 0.5418,
+      "step": 25080
+    },
+    {
+      "epoch": 4.056260609489936,
+      "grad_norm": 0.9797548651695251,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 25090
+    },
+    {
+      "epoch": 4.057877293670681,
+      "grad_norm": 0.8468434810638428,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 25100
+    },
+    {
+      "epoch": 4.059493977851426,
+      "grad_norm": 0.9294559955596924,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 25110
+    },
+    {
+      "epoch": 4.061110662032172,
+      "grad_norm": 0.9686688780784607,
+      "learning_rate": 0.0002,
+      "loss": 0.5295,
+      "step": 25120
+    },
+    {
+      "epoch": 4.062727346212918,
+      "grad_norm": 0.8042728304862976,
+      "learning_rate": 0.0002,
+      "loss": 0.5642,
+      "step": 25130
+    },
+    {
+      "epoch": 4.064344030393663,
+      "grad_norm": 1.165160894393921,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 25140
+    },
+    {
+      "epoch": 4.065960714574408,
+      "grad_norm": 1.2161961793899536,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 25150
+    },
+    {
+      "epoch": 4.067577398755153,
+      "grad_norm": 1.0762810707092285,
+      "learning_rate": 0.0002,
+      "loss": 0.5217,
+      "step": 25160
+    },
+    {
+      "epoch": 4.069194082935899,
+      "grad_norm": 0.7580869793891907,
+      "learning_rate": 0.0002,
+      "loss": 0.5886,
+      "step": 25170
+    },
+    {
+      "epoch": 4.070810767116644,
+      "grad_norm": 0.9630117416381836,
+      "learning_rate": 0.0002,
+      "loss": 0.5401,
+      "step": 25180
+    },
+    {
+      "epoch": 4.072427451297389,
+      "grad_norm": 0.9049716591835022,
+      "learning_rate": 0.0002,
+      "loss": 0.5378,
+      "step": 25190
+    },
+    {
+      "epoch": 4.074044135478134,
+      "grad_norm": 1.1536930799484253,
+      "learning_rate": 0.0002,
+      "loss": 0.5266,
+      "step": 25200
+    },
+    {
+      "epoch": 4.0756608196588795,
+      "grad_norm": 0.901461124420166,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 25210
+    },
+    {
+      "epoch": 4.077277503839625,
+      "grad_norm": 1.3318437337875366,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 25220
+    },
+    {
+      "epoch": 4.07889418802037,
+      "grad_norm": 0.8811455368995667,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 25230
+    },
+    {
+      "epoch": 4.080510872201115,
+      "grad_norm": 1.0564165115356445,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 25240
+    },
+    {
+      "epoch": 4.08212755638186,
+      "grad_norm": 1.1008027791976929,
+      "learning_rate": 0.0002,
+      "loss": 0.5472,
+      "step": 25250
+    },
+    {
+      "epoch": 4.083744240562606,
+      "grad_norm": 1.150097131729126,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 25260
+    },
+    {
+      "epoch": 4.085360924743352,
+      "grad_norm": 0.9339924454689026,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 25270
+    },
+    {
+      "epoch": 4.086977608924097,
+      "grad_norm": 1.0902045965194702,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 25280
+    },
+    {
+      "epoch": 4.088594293104842,
+      "grad_norm": 0.8483911156654358,
+      "learning_rate": 0.0002,
+      "loss": 0.5203,
+      "step": 25290
+    },
+    {
+      "epoch": 4.0902109772855875,
+      "grad_norm": 0.9477024674415588,
+      "learning_rate": 0.0002,
+      "loss": 0.5697,
+      "step": 25300
+    },
+    {
+      "epoch": 4.091827661466333,
+      "grad_norm": 0.9500215649604797,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 25310
+    },
+    {
+      "epoch": 4.093444345647078,
+      "grad_norm": 1.040468454360962,
+      "learning_rate": 0.0002,
+      "loss": 0.5045,
+      "step": 25320
+    },
+    {
+      "epoch": 4.095061029827823,
+      "grad_norm": 0.7457592487335205,
+      "learning_rate": 0.0002,
+      "loss": 0.5488,
+      "step": 25330
+    },
+    {
+      "epoch": 4.096677714008568,
+      "grad_norm": 1.2092097997665405,
+      "learning_rate": 0.0002,
+      "loss": 0.609,
+      "step": 25340
+    },
+    {
+      "epoch": 4.098294398189314,
+      "grad_norm": 0.9652107954025269,
+      "learning_rate": 0.0002,
+      "loss": 0.5174,
+      "step": 25350
+    },
+    {
+      "epoch": 4.099911082370059,
+      "grad_norm": 0.8464955687522888,
+      "learning_rate": 0.0002,
+      "loss": 0.5559,
+      "step": 25360
+    },
+    {
+      "epoch": 4.101527766550804,
+      "grad_norm": 0.875026285648346,
+      "learning_rate": 0.0002,
+      "loss": 0.5635,
+      "step": 25370
+    },
+    {
+      "epoch": 4.103144450731549,
+      "grad_norm": 0.9241740107536316,
+      "learning_rate": 0.0002,
+      "loss": 0.5774,
+      "step": 25380
+    },
+    {
+      "epoch": 4.1047611349122946,
+      "grad_norm": 0.9769546389579773,
+      "learning_rate": 0.0002,
+      "loss": 0.5578,
+      "step": 25390
+    },
+    {
+      "epoch": 4.10637781909304,
+      "grad_norm": 1.1501960754394531,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 25400
+    },
+    {
+      "epoch": 4.107994503273786,
+      "grad_norm": 0.9135243892669678,
+      "learning_rate": 0.0002,
+      "loss": 0.5241,
+      "step": 25410
+    },
+    {
+      "epoch": 4.109611187454531,
+      "grad_norm": 0.9905396103858948,
+      "learning_rate": 0.0002,
+      "loss": 0.5152,
+      "step": 25420
+    },
+    {
+      "epoch": 4.111227871635276,
+      "grad_norm": 0.9845104217529297,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 25430
+    },
+    {
+      "epoch": 4.112844555816022,
+      "grad_norm": 0.8326883912086487,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 25440
+    },
+    {
+      "epoch": 4.114461239996767,
+      "grad_norm": 0.9264556765556335,
+      "learning_rate": 0.0002,
+      "loss": 0.5312,
+      "step": 25450
+    },
+    {
+      "epoch": 4.116077924177512,
+      "grad_norm": 1.043080449104309,
+      "learning_rate": 0.0002,
+      "loss": 0.5968,
+      "step": 25460
+    },
+    {
+      "epoch": 4.117694608358257,
+      "grad_norm": 0.8533386588096619,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 25470
+    },
+    {
+      "epoch": 4.1193112925390025,
+      "grad_norm": 1.0133965015411377,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 25480
+    },
+    {
+      "epoch": 4.120927976719748,
+      "grad_norm": 0.7476310133934021,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 25490
+    },
+    {
+      "epoch": 4.122544660900493,
+      "grad_norm": 1.1247259378433228,
+      "learning_rate": 0.0002,
+      "loss": 0.5189,
+      "step": 25500
+    },
+    {
+      "epoch": 4.124161345081238,
+      "grad_norm": 1.0764678716659546,
+      "learning_rate": 0.0002,
+      "loss": 0.5751,
+      "step": 25510
+    },
+    {
+      "epoch": 4.1257780292619834,
+      "grad_norm": 0.7679798007011414,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 25520
+    },
+    {
+      "epoch": 4.127394713442729,
+      "grad_norm": 0.8877071142196655,
+      "learning_rate": 0.0002,
+      "loss": 0.5233,
+      "step": 25530
+    },
+    {
+      "epoch": 4.129011397623474,
+      "grad_norm": 1.0440239906311035,
+      "learning_rate": 0.0002,
+      "loss": 0.5769,
+      "step": 25540
+    },
+    {
+      "epoch": 4.130628081804219,
+      "grad_norm": 0.984145998954773,
+      "learning_rate": 0.0002,
+      "loss": 0.5723,
+      "step": 25550
+    },
+    {
+      "epoch": 4.132244765984965,
+      "grad_norm": 0.8667055368423462,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 25560
+    },
+    {
+      "epoch": 4.1338614501657105,
+      "grad_norm": 1.1300835609436035,
+      "learning_rate": 0.0002,
+      "loss": 0.5816,
+      "step": 25570
+    },
+    {
+      "epoch": 4.135478134346456,
+      "grad_norm": 0.9314348101615906,
+      "learning_rate": 0.0002,
+      "loss": 0.524,
+      "step": 25580
+    },
+    {
+      "epoch": 4.137094818527201,
+      "grad_norm": 0.7731879949569702,
+      "learning_rate": 0.0002,
+      "loss": 0.5283,
+      "step": 25590
+    },
+    {
+      "epoch": 4.138711502707946,
+      "grad_norm": 1.0080097913742065,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 25600
+    },
+    {
+      "epoch": 4.140328186888691,
+      "grad_norm": 1.2475038766860962,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 25610
+    },
+    {
+      "epoch": 4.141944871069437,
+      "grad_norm": 0.9912930727005005,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 25620
+    },
+    {
+      "epoch": 4.143561555250182,
+      "grad_norm": 0.9088651537895203,
+      "learning_rate": 0.0002,
+      "loss": 0.5624,
+      "step": 25630
+    },
+    {
+      "epoch": 4.145178239430927,
+      "grad_norm": 0.8940697312355042,
+      "learning_rate": 0.0002,
+      "loss": 0.5393,
+      "step": 25640
+    },
+    {
+      "epoch": 4.146794923611672,
+      "grad_norm": 1.0798203945159912,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 25650
+    },
+    {
+      "epoch": 4.148411607792418,
+      "grad_norm": 0.955172061920166,
+      "learning_rate": 0.0002,
+      "loss": 0.5987,
+      "step": 25660
+    },
+    {
+      "epoch": 4.150028291973163,
+      "grad_norm": 0.9692716002464294,
+      "learning_rate": 0.0002,
+      "loss": 0.569,
+      "step": 25670
+    },
+    {
+      "epoch": 4.151644976153908,
+      "grad_norm": 1.0813939571380615,
+      "learning_rate": 0.0002,
+      "loss": 0.5478,
+      "step": 25680
+    },
+    {
+      "epoch": 4.153261660334653,
+      "grad_norm": 1.135675072669983,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 25690
+    },
+    {
+      "epoch": 4.1548783445153985,
+      "grad_norm": 1.0392236709594727,
+      "learning_rate": 0.0002,
+      "loss": 0.5247,
+      "step": 25700
+    },
+    {
+      "epoch": 4.156495028696145,
+      "grad_norm": 0.9473116993904114,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 25710
+    },
+    {
+      "epoch": 4.15811171287689,
+      "grad_norm": 0.712493896484375,
+      "learning_rate": 0.0002,
+      "loss": 0.5339,
+      "step": 25720
+    },
+    {
+      "epoch": 4.159728397057635,
+      "grad_norm": 0.8724465370178223,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 25730
+    },
+    {
+      "epoch": 4.16134508123838,
+      "grad_norm": 0.9870015978813171,
+      "learning_rate": 0.0002,
+      "loss": 0.5325,
+      "step": 25740
+    },
+    {
+      "epoch": 4.1629617654191255,
+      "grad_norm": 1.025273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 25750
+    },
+    {
+      "epoch": 4.164578449599871,
+      "grad_norm": 0.9243090152740479,
+      "learning_rate": 0.0002,
+      "loss": 0.5223,
+      "step": 25760
+    },
+    {
+      "epoch": 4.166195133780616,
+      "grad_norm": 1.1656451225280762,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 25770
+    },
+    {
+      "epoch": 4.167811817961361,
+      "grad_norm": 0.936358630657196,
+      "learning_rate": 0.0002,
+      "loss": 0.5334,
+      "step": 25780
+    },
+    {
+      "epoch": 4.1694285021421065,
+      "grad_norm": 0.8618208169937134,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 25790
+    },
+    {
+      "epoch": 4.171045186322852,
+      "grad_norm": 0.8580600023269653,
+      "learning_rate": 0.0002,
+      "loss": 0.5186,
+      "step": 25800
+    },
+    {
+      "epoch": 4.172661870503597,
+      "grad_norm": 1.0128562450408936,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 25810
+    },
+    {
+      "epoch": 4.174278554684342,
+      "grad_norm": 0.854865312576294,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 25820
+    },
+    {
+      "epoch": 4.175895238865087,
+      "grad_norm": 1.235082745552063,
+      "learning_rate": 0.0002,
+      "loss": 0.5377,
+      "step": 25830
+    },
+    {
+      "epoch": 4.177511923045833,
+      "grad_norm": 0.9796220660209656,
+      "learning_rate": 0.0002,
+      "loss": 0.5614,
+      "step": 25840
+    },
+    {
+      "epoch": 4.179128607226578,
+      "grad_norm": 0.8922094702720642,
+      "learning_rate": 0.0002,
+      "loss": 0.5689,
+      "step": 25850
+    },
+    {
+      "epoch": 4.180745291407324,
+      "grad_norm": 0.9672530293464661,
+      "learning_rate": 0.0002,
+      "loss": 0.5806,
+      "step": 25860
+    },
+    {
+      "epoch": 4.182361975588069,
+      "grad_norm": 0.8662548661231995,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 25870
+    },
+    {
+      "epoch": 4.1839786597688144,
+      "grad_norm": 0.7938798069953918,
+      "learning_rate": 0.0002,
+      "loss": 0.5329,
+      "step": 25880
+    },
+    {
+      "epoch": 4.18559534394956,
+      "grad_norm": 1.0517958402633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5427,
+      "step": 25890
+    },
+    {
+      "epoch": 4.187212028130305,
+      "grad_norm": 0.8939275145530701,
+      "learning_rate": 0.0002,
+      "loss": 0.5147,
+      "step": 25900
+    },
+    {
+      "epoch": 4.18882871231105,
+      "grad_norm": 1.0296672582626343,
+      "learning_rate": 0.0002,
+      "loss": 0.5199,
+      "step": 25910
+    },
+    {
+      "epoch": 4.190445396491795,
+      "grad_norm": 0.8104017972946167,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 25920
+    },
+    {
+      "epoch": 4.192062080672541,
+      "grad_norm": 0.9984509944915771,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 25930
+    },
+    {
+      "epoch": 4.193678764853286,
+      "grad_norm": 0.9844784736633301,
+      "learning_rate": 0.0002,
+      "loss": 0.5356,
+      "step": 25940
+    },
+    {
+      "epoch": 4.195295449034031,
+      "grad_norm": 0.8168622255325317,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 25950
+    },
+    {
+      "epoch": 4.196912133214776,
+      "grad_norm": 1.0878913402557373,
+      "learning_rate": 0.0002,
+      "loss": 0.542,
+      "step": 25960
+    },
+    {
+      "epoch": 4.1985288173955215,
+      "grad_norm": 0.927126407623291,
+      "learning_rate": 0.0002,
+      "loss": 0.5414,
+      "step": 25970
+    },
+    {
+      "epoch": 4.200145501576267,
+      "grad_norm": 0.838586688041687,
+      "learning_rate": 0.0002,
+      "loss": 0.5794,
+      "step": 25980
+    },
+    {
+      "epoch": 4.201762185757012,
+      "grad_norm": 1.2572145462036133,
+      "learning_rate": 0.0002,
+      "loss": 0.5454,
+      "step": 25990
+    },
+    {
+      "epoch": 4.203378869937758,
+      "grad_norm": 1.0476740598678589,
+      "learning_rate": 0.0002,
+      "loss": 0.5048,
+      "step": 26000
+    },
+    {
+      "epoch": 4.204995554118503,
+      "grad_norm": 1.0873368978500366,
+      "learning_rate": 0.0002,
+      "loss": 0.5127,
+      "step": 26010
+    },
+    {
+      "epoch": 4.206612238299249,
+      "grad_norm": 1.2664896249771118,
+      "learning_rate": 0.0002,
+      "loss": 0.5679,
+      "step": 26020
+    },
+    {
+      "epoch": 4.208228922479994,
+      "grad_norm": 1.0312391519546509,
+      "learning_rate": 0.0002,
+      "loss": 0.5814,
+      "step": 26030
+    },
+    {
+      "epoch": 4.209845606660739,
+      "grad_norm": 1.0235042572021484,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 26040
+    },
+    {
+      "epoch": 4.211462290841484,
+      "grad_norm": 0.8882219195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 26050
+    },
+    {
+      "epoch": 4.2130789750222295,
+      "grad_norm": 0.9115961790084839,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 26060
+    },
+    {
+      "epoch": 4.214695659202975,
+      "grad_norm": 1.0218228101730347,
+      "learning_rate": 0.0002,
+      "loss": 0.5455,
+      "step": 26070
+    },
+    {
+      "epoch": 4.21631234338372,
+      "grad_norm": 1.0802232027053833,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 26080
+    },
+    {
+      "epoch": 4.217929027564465,
+      "grad_norm": 1.1488053798675537,
+      "learning_rate": 0.0002,
+      "loss": 0.557,
+      "step": 26090
+    },
+    {
+      "epoch": 4.21954571174521,
+      "grad_norm": 1.0487725734710693,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 26100
+    },
+    {
+      "epoch": 4.221162395925956,
+      "grad_norm": 0.9131165742874146,
+      "learning_rate": 0.0002,
+      "loss": 0.5568,
+      "step": 26110
+    },
+    {
+      "epoch": 4.222779080106701,
+      "grad_norm": 0.9012845158576965,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 26120
+    },
+    {
+      "epoch": 4.224395764287446,
+      "grad_norm": 0.8389840126037598,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 26130
+    },
+    {
+      "epoch": 4.226012448468191,
+      "grad_norm": 0.8924660682678223,
+      "learning_rate": 0.0002,
+      "loss": 0.5268,
+      "step": 26140
+    },
+    {
+      "epoch": 4.2276291326489375,
+      "grad_norm": 0.8556463718414307,
+      "learning_rate": 0.0002,
+      "loss": 0.5715,
+      "step": 26150
+    },
+    {
+      "epoch": 4.229245816829683,
+      "grad_norm": 0.9643129110336304,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 26160
+    },
+    {
+      "epoch": 4.230862501010428,
+      "grad_norm": 0.9865712523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 26170
+    },
+    {
+      "epoch": 4.232479185191173,
+      "grad_norm": 1.152641773223877,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 26180
+    },
+    {
+      "epoch": 4.234095869371918,
+      "grad_norm": 0.9157698154449463,
+      "learning_rate": 0.0002,
+      "loss": 0.5632,
+      "step": 26190
+    },
+    {
+      "epoch": 4.235712553552664,
+      "grad_norm": 0.8418048620223999,
+      "learning_rate": 0.0002,
+      "loss": 0.5717,
+      "step": 26200
+    },
+    {
+      "epoch": 4.237329237733409,
+      "grad_norm": 0.9430168867111206,
+      "learning_rate": 0.0002,
+      "loss": 0.5624,
+      "step": 26210
+    },
+    {
+      "epoch": 4.238945921914154,
+      "grad_norm": 1.012582778930664,
+      "learning_rate": 0.0002,
+      "loss": 0.5574,
+      "step": 26220
+    },
+    {
+      "epoch": 4.240562606094899,
+      "grad_norm": 1.112619400024414,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 26230
+    },
+    {
+      "epoch": 4.2421792902756446,
+      "grad_norm": 0.9243621826171875,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 26240
+    },
+    {
+      "epoch": 4.24379597445639,
+      "grad_norm": 0.6977595686912537,
+      "learning_rate": 0.0002,
+      "loss": 0.569,
+      "step": 26250
+    },
+    {
+      "epoch": 4.245412658637135,
+      "grad_norm": 0.9600721597671509,
+      "learning_rate": 0.0002,
+      "loss": 0.5379,
+      "step": 26260
+    },
+    {
+      "epoch": 4.24702934281788,
+      "grad_norm": 0.882641613483429,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 26270
+    },
+    {
+      "epoch": 4.2486460269986255,
+      "grad_norm": 1.010920763015747,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 26280
+    },
+    {
+      "epoch": 4.250262711179371,
+      "grad_norm": 0.9289400577545166,
+      "learning_rate": 0.0002,
+      "loss": 0.5803,
+      "step": 26290
+    },
+    {
+      "epoch": 4.251879395360117,
+      "grad_norm": 1.137397289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.541,
+      "step": 26300
+    },
+    {
+      "epoch": 4.253496079540862,
+      "grad_norm": 1.0136182308197021,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 26310
+    },
+    {
+      "epoch": 4.255112763721607,
+      "grad_norm": 0.9387356042861938,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 26320
+    },
+    {
+      "epoch": 4.2567294479023525,
+      "grad_norm": 1.1833957433700562,
+      "learning_rate": 0.0002,
+      "loss": 0.5948,
+      "step": 26330
+    },
+    {
+      "epoch": 4.258346132083098,
+      "grad_norm": 0.9415934681892395,
+      "learning_rate": 0.0002,
+      "loss": 0.5905,
+      "step": 26340
+    },
+    {
+      "epoch": 4.259962816263843,
+      "grad_norm": 0.8550165891647339,
+      "learning_rate": 0.0002,
+      "loss": 0.5539,
+      "step": 26350
+    },
+    {
+      "epoch": 4.261579500444588,
+      "grad_norm": 9.924622535705566,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 26360
+    },
+    {
+      "epoch": 4.2631961846253335,
+      "grad_norm": 1.0104902982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.5689,
+      "step": 26370
+    },
+    {
+      "epoch": 4.264812868806079,
+      "grad_norm": 0.890794038772583,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 26380
+    },
+    {
+      "epoch": 4.266429552986824,
+      "grad_norm": 1.0560191869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.563,
+      "step": 26390
+    },
+    {
+      "epoch": 4.268046237167569,
+      "grad_norm": 1.0135581493377686,
+      "learning_rate": 0.0002,
+      "loss": 0.5119,
+      "step": 26400
+    },
+    {
+      "epoch": 4.269662921348314,
+      "grad_norm": 1.1304140090942383,
+      "learning_rate": 0.0002,
+      "loss": 0.5359,
+      "step": 26410
+    },
+    {
+      "epoch": 4.27127960552906,
+      "grad_norm": 0.9899303913116455,
+      "learning_rate": 0.0002,
+      "loss": 0.5615,
+      "step": 26420
+    },
+    {
+      "epoch": 4.272896289709805,
+      "grad_norm": 1.0505329370498657,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 26430
+    },
+    {
+      "epoch": 4.27451297389055,
+      "grad_norm": 0.9389396905899048,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 26440
+    },
+    {
+      "epoch": 4.276129658071296,
+      "grad_norm": 0.875328779220581,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 26450
+    },
+    {
+      "epoch": 4.277746342252041,
+      "grad_norm": 1.0689256191253662,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 26460
+    },
+    {
+      "epoch": 4.279363026432787,
+      "grad_norm": 0.9988957643508911,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 26470
+    },
+    {
+      "epoch": 4.280979710613532,
+      "grad_norm": 0.8721813559532166,
+      "learning_rate": 0.0002,
+      "loss": 0.5478,
+      "step": 26480
+    },
+    {
+      "epoch": 4.282596394794277,
+      "grad_norm": 1.100109577178955,
+      "learning_rate": 0.0002,
+      "loss": 0.5424,
+      "step": 26490
+    },
+    {
+      "epoch": 4.284213078975022,
+      "grad_norm": 1.1607271432876587,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 26500
+    },
+    {
+      "epoch": 4.285829763155768,
+      "grad_norm": 0.879088819026947,
+      "learning_rate": 0.0002,
+      "loss": 0.6287,
+      "step": 26510
+    },
+    {
+      "epoch": 4.287446447336513,
+      "grad_norm": 0.9891700744628906,
+      "learning_rate": 0.0002,
+      "loss": 0.573,
+      "step": 26520
+    },
+    {
+      "epoch": 4.289063131517258,
+      "grad_norm": 1.0831127166748047,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 26530
+    },
+    {
+      "epoch": 4.290679815698003,
+      "grad_norm": 1.4108285903930664,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 26540
+    },
+    {
+      "epoch": 4.2922964998787485,
+      "grad_norm": 1.0630289316177368,
+      "learning_rate": 0.0002,
+      "loss": 0.5888,
+      "step": 26550
+    },
+    {
+      "epoch": 4.293913184059494,
+      "grad_norm": 1.0854572057724,
+      "learning_rate": 0.0002,
+      "loss": 0.5817,
+      "step": 26560
+    },
+    {
+      "epoch": 4.295529868240239,
+      "grad_norm": 0.9561646580696106,
+      "learning_rate": 0.0002,
+      "loss": 0.5586,
+      "step": 26570
+    },
+    {
+      "epoch": 4.297146552420984,
+      "grad_norm": 0.9064981937408447,
+      "learning_rate": 0.0002,
+      "loss": 0.5674,
+      "step": 26580
+    },
+    {
+      "epoch": 4.298763236601729,
+      "grad_norm": 1.0082972049713135,
+      "learning_rate": 0.0002,
+      "loss": 0.5847,
+      "step": 26590
+    },
+    {
+      "epoch": 4.3003799207824756,
+      "grad_norm": 1.1613214015960693,
+      "learning_rate": 0.0002,
+      "loss": 0.5711,
+      "step": 26600
+    },
+    {
+      "epoch": 4.301996604963221,
+      "grad_norm": 0.9847695231437683,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 26610
+    },
+    {
+      "epoch": 4.303613289143966,
+      "grad_norm": 1.0980697870254517,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 26620
+    },
+    {
+      "epoch": 4.305229973324711,
+      "grad_norm": 0.8861175179481506,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 26630
+    },
+    {
+      "epoch": 4.3068466575054565,
+      "grad_norm": 0.8917363286018372,
+      "learning_rate": 0.0002,
+      "loss": 0.5716,
+      "step": 26640
+    },
+    {
+      "epoch": 4.308463341686202,
+      "grad_norm": 1.0458378791809082,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 26650
+    },
+    {
+      "epoch": 4.310080025866947,
+      "grad_norm": 1.4859240055084229,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 26660
+    },
+    {
+      "epoch": 4.311696710047692,
+      "grad_norm": 1.1376359462738037,
+      "learning_rate": 0.0002,
+      "loss": 0.5296,
+      "step": 26670
+    },
+    {
+      "epoch": 4.313313394228437,
+      "grad_norm": 0.991349995136261,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 26680
+    },
+    {
+      "epoch": 4.314930078409183,
+      "grad_norm": 0.9995543956756592,
+      "learning_rate": 0.0002,
+      "loss": 0.5338,
+      "step": 26690
+    },
+    {
+      "epoch": 4.316546762589928,
+      "grad_norm": 1.0515851974487305,
+      "learning_rate": 0.0002,
+      "loss": 0.5542,
+      "step": 26700
+    },
+    {
+      "epoch": 4.318163446770673,
+      "grad_norm": 1.008023977279663,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 26710
+    },
+    {
+      "epoch": 4.319780130951418,
+      "grad_norm": 1.0184582471847534,
+      "learning_rate": 0.0002,
+      "loss": 0.5506,
+      "step": 26720
+    },
+    {
+      "epoch": 4.321396815132164,
+      "grad_norm": 1.161071538925171,
+      "learning_rate": 0.0002,
+      "loss": 0.5828,
+      "step": 26730
+    },
+    {
+      "epoch": 4.323013499312909,
+      "grad_norm": 0.9580779671669006,
+      "learning_rate": 0.0002,
+      "loss": 0.5633,
+      "step": 26740
+    },
+    {
+      "epoch": 4.324630183493655,
+      "grad_norm": 1.0189911127090454,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 26750
+    },
+    {
+      "epoch": 4.3262468676744,
+      "grad_norm": 0.7484358549118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 26760
+    },
+    {
+      "epoch": 4.327863551855145,
+      "grad_norm": 1.0015908479690552,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 26770
+    },
+    {
+      "epoch": 4.329480236035891,
+      "grad_norm": 0.8972945809364319,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 26780
+    },
+    {
+      "epoch": 4.331096920216636,
+      "grad_norm": 1.01099693775177,
+      "learning_rate": 0.0002,
+      "loss": 0.5857,
+      "step": 26790
+    },
+    {
+      "epoch": 4.332713604397381,
+      "grad_norm": 0.846958339214325,
+      "learning_rate": 0.0002,
+      "loss": 0.5591,
+      "step": 26800
+    },
+    {
+      "epoch": 4.334330288578126,
+      "grad_norm": 1.0792603492736816,
+      "learning_rate": 0.0002,
+      "loss": 0.5547,
+      "step": 26810
+    },
+    {
+      "epoch": 4.3359469727588715,
+      "grad_norm": 1.0373345613479614,
+      "learning_rate": 0.0002,
+      "loss": 0.5747,
+      "step": 26820
+    },
+    {
+      "epoch": 4.337563656939617,
+      "grad_norm": 0.9779167771339417,
+      "learning_rate": 0.0002,
+      "loss": 0.558,
+      "step": 26830
+    },
+    {
+      "epoch": 4.339180341120362,
+      "grad_norm": 1.0235520601272583,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 26840
+    },
+    {
+      "epoch": 4.340797025301107,
+      "grad_norm": 1.04195237159729,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 26850
+    },
+    {
+      "epoch": 4.3424137094818525,
+      "grad_norm": 0.9479565620422363,
+      "learning_rate": 0.0002,
+      "loss": 0.5474,
+      "step": 26860
+    },
+    {
+      "epoch": 4.344030393662598,
+      "grad_norm": 0.9526172280311584,
+      "learning_rate": 0.0002,
+      "loss": 0.5646,
+      "step": 26870
+    },
+    {
+      "epoch": 4.345647077843343,
+      "grad_norm": 0.8571456074714661,
+      "learning_rate": 0.0002,
+      "loss": 0.521,
+      "step": 26880
+    },
+    {
+      "epoch": 4.347263762024088,
+      "grad_norm": 0.9475828409194946,
+      "learning_rate": 0.0002,
+      "loss": 0.5846,
+      "step": 26890
+    },
+    {
+      "epoch": 4.348880446204834,
+      "grad_norm": 1.0529576539993286,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 26900
+    },
+    {
+      "epoch": 4.3504971303855795,
+      "grad_norm": 0.9648140072822571,
+      "learning_rate": 0.0002,
+      "loss": 0.56,
+      "step": 26910
+    },
+    {
+      "epoch": 4.352113814566325,
+      "grad_norm": 1.0488841533660889,
+      "learning_rate": 0.0002,
+      "loss": 0.5162,
+      "step": 26920
+    },
+    {
+      "epoch": 4.35373049874707,
+      "grad_norm": 0.8771942257881165,
+      "learning_rate": 0.0002,
+      "loss": 0.5842,
+      "step": 26930
+    },
+    {
+      "epoch": 4.355347182927815,
+      "grad_norm": 0.9411202073097229,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 26940
+    },
+    {
+      "epoch": 4.35696386710856,
+      "grad_norm": 1.0997588634490967,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 26950
+    },
+    {
+      "epoch": 4.358580551289306,
+      "grad_norm": 0.968754768371582,
+      "learning_rate": 0.0002,
+      "loss": 0.5528,
+      "step": 26960
+    },
+    {
+      "epoch": 4.360197235470051,
+      "grad_norm": 0.9990773797035217,
+      "learning_rate": 0.0002,
+      "loss": 0.5881,
+      "step": 26970
+    },
+    {
+      "epoch": 4.361813919650796,
+      "grad_norm": 1.0210620164871216,
+      "learning_rate": 0.0002,
+      "loss": 0.5761,
+      "step": 26980
+    },
+    {
+      "epoch": 4.363430603831541,
+      "grad_norm": 0.855462908744812,
+      "learning_rate": 0.0002,
+      "loss": 0.5768,
+      "step": 26990
+    },
+    {
+      "epoch": 4.365047288012287,
+      "grad_norm": 0.9169660806655884,
+      "learning_rate": 0.0002,
+      "loss": 0.5493,
+      "step": 27000
+    },
+    {
+      "epoch": 4.366663972193032,
+      "grad_norm": 1.089629888534546,
+      "learning_rate": 0.0002,
+      "loss": 0.5697,
+      "step": 27010
+    },
+    {
+      "epoch": 4.368280656373777,
+      "grad_norm": 1.0932867527008057,
+      "learning_rate": 0.0002,
+      "loss": 0.5854,
+      "step": 27020
+    },
+    {
+      "epoch": 4.369897340554522,
+      "grad_norm": 0.9290956854820251,
+      "learning_rate": 0.0002,
+      "loss": 0.5656,
+      "step": 27030
+    },
+    {
+      "epoch": 4.3715140247352675,
+      "grad_norm": 1.2800624370574951,
+      "learning_rate": 0.0002,
+      "loss": 0.5727,
+      "step": 27040
+    },
+    {
+      "epoch": 4.373130708916014,
+      "grad_norm": 0.8993493318557739,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 27050
+    },
+    {
+      "epoch": 4.374747393096759,
+      "grad_norm": 1.1566431522369385,
+      "learning_rate": 0.0002,
+      "loss": 0.6232,
+      "step": 27060
+    },
+    {
+      "epoch": 4.376364077277504,
+      "grad_norm": 0.9479052424430847,
+      "learning_rate": 0.0002,
+      "loss": 0.5902,
+      "step": 27070
+    },
+    {
+      "epoch": 4.377980761458249,
+      "grad_norm": 1.0063648223876953,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 27080
+    },
+    {
+      "epoch": 4.379597445638995,
+      "grad_norm": 0.8342045545578003,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 27090
+    },
+    {
+      "epoch": 4.38121412981974,
+      "grad_norm": 1.1390739679336548,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 27100
+    },
+    {
+      "epoch": 4.382830814000485,
+      "grad_norm": 0.9547637104988098,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 27110
+    },
+    {
+      "epoch": 4.38444749818123,
+      "grad_norm": 1.0503804683685303,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 27120
+    },
+    {
+      "epoch": 4.3860641823619755,
+      "grad_norm": 0.9064017534255981,
+      "learning_rate": 0.0002,
+      "loss": 0.5787,
+      "step": 27130
+    },
+    {
+      "epoch": 4.387680866542721,
+      "grad_norm": 0.9382519125938416,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 27140
+    },
+    {
+      "epoch": 4.389297550723466,
+      "grad_norm": 1.0410341024398804,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 27150
+    },
+    {
+      "epoch": 4.390914234904211,
+      "grad_norm": 0.9218655824661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6034,
+      "step": 27160
+    },
+    {
+      "epoch": 4.392530919084956,
+      "grad_norm": 0.8119737505912781,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 27170
+    },
+    {
+      "epoch": 4.394147603265702,
+      "grad_norm": 0.8584722876548767,
+      "learning_rate": 0.0002,
+      "loss": 0.5612,
+      "step": 27180
+    },
+    {
+      "epoch": 4.395764287446447,
+      "grad_norm": 0.9668293595314026,
+      "learning_rate": 0.0002,
+      "loss": 0.5772,
+      "step": 27190
+    },
+    {
+      "epoch": 4.397380971627193,
+      "grad_norm": 1.022334098815918,
+      "learning_rate": 0.0002,
+      "loss": 0.6009,
+      "step": 27200
+    },
+    {
+      "epoch": 4.398997655807938,
+      "grad_norm": 0.9553216099739075,
+      "learning_rate": 0.0002,
+      "loss": 0.5573,
+      "step": 27210
+    },
+    {
+      "epoch": 4.4006143399886835,
+      "grad_norm": 0.9282339215278625,
+      "learning_rate": 0.0002,
+      "loss": 0.5604,
+      "step": 27220
+    },
+    {
+      "epoch": 4.402231024169429,
+      "grad_norm": 1.0232292413711548,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 27230
+    },
+    {
+      "epoch": 4.403847708350174,
+      "grad_norm": 0.9915700554847717,
+      "learning_rate": 0.0002,
+      "loss": 0.6078,
+      "step": 27240
+    },
+    {
+      "epoch": 4.405464392530919,
+      "grad_norm": 1.0014961957931519,
+      "learning_rate": 0.0002,
+      "loss": 0.5778,
+      "step": 27250
+    },
+    {
+      "epoch": 4.407081076711664,
+      "grad_norm": 1.1172103881835938,
+      "learning_rate": 0.0002,
+      "loss": 0.5824,
+      "step": 27260
+    },
+    {
+      "epoch": 4.40869776089241,
+      "grad_norm": 0.8583093285560608,
+      "learning_rate": 0.0002,
+      "loss": 0.5286,
+      "step": 27270
+    },
+    {
+      "epoch": 4.410314445073155,
+      "grad_norm": 0.7609201669692993,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 27280
+    },
+    {
+      "epoch": 4.4119311292539,
+      "grad_norm": 1.0619351863861084,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 27290
+    },
+    {
+      "epoch": 4.413547813434645,
+      "grad_norm": 1.0177674293518066,
+      "learning_rate": 0.0002,
+      "loss": 0.5579,
+      "step": 27300
+    },
+    {
+      "epoch": 4.4151644976153905,
+      "grad_norm": 0.9921218156814575,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 27310
+    },
+    {
+      "epoch": 4.416781181796136,
+      "grad_norm": 1.126244306564331,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 27320
+    },
+    {
+      "epoch": 4.418397865976881,
+      "grad_norm": 1.0678540468215942,
+      "learning_rate": 0.0002,
+      "loss": 0.5743,
+      "step": 27330
+    },
+    {
+      "epoch": 4.420014550157627,
+      "grad_norm": 0.8705704212188721,
+      "learning_rate": 0.0002,
+      "loss": 0.5665,
+      "step": 27340
+    },
+    {
+      "epoch": 4.421631234338372,
+      "grad_norm": 1.272074818611145,
+      "learning_rate": 0.0002,
+      "loss": 0.5763,
+      "step": 27350
+    },
+    {
+      "epoch": 4.423247918519118,
+      "grad_norm": 0.8740444183349609,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 27360
+    },
+    {
+      "epoch": 4.424864602699863,
+      "grad_norm": 1.0584250688552856,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 27370
+    },
+    {
+      "epoch": 4.426481286880608,
+      "grad_norm": 1.059870719909668,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 27380
+    },
+    {
+      "epoch": 4.428097971061353,
+      "grad_norm": 1.072265863418579,
+      "learning_rate": 0.0002,
+      "loss": 0.5551,
+      "step": 27390
+    },
+    {
+      "epoch": 4.4297146552420985,
+      "grad_norm": 0.871481716632843,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 27400
+    },
+    {
+      "epoch": 4.431331339422844,
+      "grad_norm": 0.9555448293685913,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 27410
+    },
+    {
+      "epoch": 4.432948023603589,
+      "grad_norm": 1.0402292013168335,
+      "learning_rate": 0.0002,
+      "loss": 0.5593,
+      "step": 27420
+    },
+    {
+      "epoch": 4.434564707784334,
+      "grad_norm": 1.12587571144104,
+      "learning_rate": 0.0002,
+      "loss": 0.5532,
+      "step": 27430
+    },
+    {
+      "epoch": 4.436181391965079,
+      "grad_norm": 1.0783193111419678,
+      "learning_rate": 0.0002,
+      "loss": 0.5403,
+      "step": 27440
+    },
+    {
+      "epoch": 4.437798076145825,
+      "grad_norm": 1.024133563041687,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 27450
+    },
+    {
+      "epoch": 4.43941476032657,
+      "grad_norm": 0.9156768918037415,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 27460
+    },
+    {
+      "epoch": 4.441031444507315,
+      "grad_norm": 1.0215224027633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 27470
+    },
+    {
+      "epoch": 4.442648128688061,
+      "grad_norm": 1.082116961479187,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 27480
+    },
+    {
+      "epoch": 4.4442648128688065,
+      "grad_norm": 1.0412873029708862,
+      "learning_rate": 0.0002,
+      "loss": 0.6203,
+      "step": 27490
+    },
+    {
+      "epoch": 4.445881497049552,
+      "grad_norm": 1.0509289503097534,
+      "learning_rate": 0.0002,
+      "loss": 0.5939,
+      "step": 27500
+    },
+    {
+      "epoch": 4.447498181230297,
+      "grad_norm": 0.9291498064994812,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 27510
+    },
+    {
+      "epoch": 4.449114865411042,
+      "grad_norm": 0.970184326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.5408,
+      "step": 27520
+    },
+    {
+      "epoch": 4.450731549591787,
+      "grad_norm": 0.8418883681297302,
+      "learning_rate": 0.0002,
+      "loss": 0.5705,
+      "step": 27530
+    },
+    {
+      "epoch": 4.452348233772533,
+      "grad_norm": 0.8823825120925903,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 27540
+    },
+    {
+      "epoch": 4.453964917953278,
+      "grad_norm": 1.1909019947052002,
+      "learning_rate": 0.0002,
+      "loss": 0.5867,
+      "step": 27550
+    },
+    {
+      "epoch": 4.455581602134023,
+      "grad_norm": 1.0317302942276,
+      "learning_rate": 0.0002,
+      "loss": 0.5685,
+      "step": 27560
+    },
+    {
+      "epoch": 4.457198286314768,
+      "grad_norm": 0.9977751970291138,
+      "learning_rate": 0.0002,
+      "loss": 0.5538,
+      "step": 27570
+    },
+    {
+      "epoch": 4.458814970495514,
+      "grad_norm": 0.8909519910812378,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 27580
+    },
+    {
+      "epoch": 4.460431654676259,
+      "grad_norm": 0.8653029799461365,
+      "learning_rate": 0.0002,
+      "loss": 0.6099,
+      "step": 27590
+    },
+    {
+      "epoch": 4.462048338857004,
+      "grad_norm": 1.0783653259277344,
+      "learning_rate": 0.0002,
+      "loss": 0.5622,
+      "step": 27600
+    },
+    {
+      "epoch": 4.463665023037749,
+      "grad_norm": 1.1235394477844238,
+      "learning_rate": 0.0002,
+      "loss": 0.579,
+      "step": 27610
+    },
+    {
+      "epoch": 4.4652817072184945,
+      "grad_norm": 0.9386643767356873,
+      "learning_rate": 0.0002,
+      "loss": 0.5545,
+      "step": 27620
+    },
+    {
+      "epoch": 4.466898391399241,
+      "grad_norm": 1.0605148077011108,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 27630
+    },
+    {
+      "epoch": 4.468515075579986,
+      "grad_norm": 1.1283893585205078,
+      "learning_rate": 0.0002,
+      "loss": 0.5886,
+      "step": 27640
+    },
+    {
+      "epoch": 4.470131759760731,
+      "grad_norm": 1.0583468675613403,
+      "learning_rate": 0.0002,
+      "loss": 0.5801,
+      "step": 27650
+    },
+    {
+      "epoch": 4.471748443941476,
+      "grad_norm": 0.9563992023468018,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 27660
+    },
+    {
+      "epoch": 4.4733651281222215,
+      "grad_norm": 1.100598931312561,
+      "learning_rate": 0.0002,
+      "loss": 0.5687,
+      "step": 27670
+    },
+    {
+      "epoch": 4.474981812302967,
+      "grad_norm": 0.9386957287788391,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 27680
+    },
+    {
+      "epoch": 4.476598496483712,
+      "grad_norm": 1.2946288585662842,
+      "learning_rate": 0.0002,
+      "loss": 0.6241,
+      "step": 27690
+    },
+    {
+      "epoch": 4.478215180664457,
+      "grad_norm": 1.0325199365615845,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 27700
+    },
+    {
+      "epoch": 4.4798318648452025,
+      "grad_norm": 1.0318928956985474,
+      "learning_rate": 0.0002,
+      "loss": 0.588,
+      "step": 27710
+    },
+    {
+      "epoch": 4.481448549025948,
+      "grad_norm": 0.8721024394035339,
+      "learning_rate": 0.0002,
+      "loss": 0.5656,
+      "step": 27720
+    },
+    {
+      "epoch": 4.483065233206693,
+      "grad_norm": 1.17376708984375,
+      "learning_rate": 0.0002,
+      "loss": 0.5421,
+      "step": 27730
+    },
+    {
+      "epoch": 4.484681917387438,
+      "grad_norm": 1.0926326513290405,
+      "learning_rate": 0.0002,
+      "loss": 0.5657,
+      "step": 27740
+    },
+    {
+      "epoch": 4.486298601568183,
+      "grad_norm": 0.9043852686882019,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 27750
+    },
+    {
+      "epoch": 4.487915285748929,
+      "grad_norm": 1.064600944519043,
+      "learning_rate": 0.0002,
+      "loss": 0.582,
+      "step": 27760
+    },
+    {
+      "epoch": 4.489531969929674,
+      "grad_norm": 0.7833460569381714,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 27770
+    },
+    {
+      "epoch": 4.49114865411042,
+      "grad_norm": 1.1073496341705322,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 27780
+    },
+    {
+      "epoch": 4.492765338291165,
+      "grad_norm": 1.0799397230148315,
+      "learning_rate": 0.0002,
+      "loss": 0.5577,
+      "step": 27790
+    },
+    {
+      "epoch": 4.49438202247191,
+      "grad_norm": 1.1062238216400146,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 27800
+    },
+    {
+      "epoch": 4.495998706652656,
+      "grad_norm": 1.0568242073059082,
+      "learning_rate": 0.0002,
+      "loss": 0.6126,
+      "step": 27810
+    },
+    {
+      "epoch": 4.497615390833401,
+      "grad_norm": 0.8861091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 27820
+    },
+    {
+      "epoch": 4.499232075014146,
+      "grad_norm": 1.2297543287277222,
+      "learning_rate": 0.0002,
+      "loss": 0.5858,
+      "step": 27830
+    },
+    {
+      "epoch": 4.500848759194891,
+      "grad_norm": 0.9600302577018738,
+      "learning_rate": 0.0002,
+      "loss": 0.5859,
+      "step": 27840
+    },
+    {
+      "epoch": 4.502465443375637,
+      "grad_norm": 1.057051181793213,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 27850
+    },
+    {
+      "epoch": 4.504082127556382,
+      "grad_norm": 0.9839690923690796,
+      "learning_rate": 0.0002,
+      "loss": 0.5788,
+      "step": 27860
+    },
+    {
+      "epoch": 4.505698811737127,
+      "grad_norm": 1.1479853391647339,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 27870
+    },
+    {
+      "epoch": 4.507315495917872,
+      "grad_norm": 1.0550768375396729,
+      "learning_rate": 0.0002,
+      "loss": 0.6039,
+      "step": 27880
+    },
+    {
+      "epoch": 4.5089321800986175,
+      "grad_norm": 0.898209273815155,
+      "learning_rate": 0.0002,
+      "loss": 0.563,
+      "step": 27890
+    },
+    {
+      "epoch": 4.510548864279363,
+      "grad_norm": 0.9460315108299255,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 27900
+    },
+    {
+      "epoch": 4.512165548460108,
+      "grad_norm": 0.9499884247779846,
+      "learning_rate": 0.0002,
+      "loss": 0.5702,
+      "step": 27910
+    },
+    {
+      "epoch": 4.513782232640853,
+      "grad_norm": 0.7801318764686584,
+      "learning_rate": 0.0002,
+      "loss": 0.5385,
+      "step": 27920
+    },
+    {
+      "epoch": 4.515398916821599,
+      "grad_norm": 0.9286966323852539,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 27930
+    },
+    {
+      "epoch": 4.517015601002345,
+      "grad_norm": 0.9539980292320251,
+      "learning_rate": 0.0002,
+      "loss": 0.5717,
+      "step": 27940
+    },
+    {
+      "epoch": 4.51863228518309,
+      "grad_norm": 1.1053401231765747,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 27950
+    },
+    {
+      "epoch": 4.520248969363835,
+      "grad_norm": 0.7535534501075745,
+      "learning_rate": 0.0002,
+      "loss": 0.6087,
+      "step": 27960
+    },
+    {
+      "epoch": 4.52186565354458,
+      "grad_norm": 1.076926589012146,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 27970
+    },
+    {
+      "epoch": 4.5234823377253255,
+      "grad_norm": 1.181935429573059,
+      "learning_rate": 0.0002,
+      "loss": 0.6028,
+      "step": 27980
+    },
+    {
+      "epoch": 4.525099021906071,
+      "grad_norm": 0.9293407201766968,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 27990
+    },
+    {
+      "epoch": 4.526715706086816,
+      "grad_norm": 0.8953009247779846,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 28000
+    },
+    {
+      "epoch": 4.528332390267561,
+      "grad_norm": 1.0850225687026978,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 28010
+    },
+    {
+      "epoch": 4.529949074448306,
+      "grad_norm": 0.9125663042068481,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 28020
+    },
+    {
+      "epoch": 4.531565758629052,
+      "grad_norm": 0.8745216727256775,
+      "learning_rate": 0.0002,
+      "loss": 0.5922,
+      "step": 28030
+    },
+    {
+      "epoch": 4.533182442809797,
+      "grad_norm": 1.0783463716506958,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 28040
+    },
+    {
+      "epoch": 4.534799126990542,
+      "grad_norm": 0.7513844966888428,
+      "learning_rate": 0.0002,
+      "loss": 0.5754,
+      "step": 28050
+    },
+    {
+      "epoch": 4.536415811171287,
+      "grad_norm": 1.0135776996612549,
+      "learning_rate": 0.0002,
+      "loss": 0.5608,
+      "step": 28060
+    },
+    {
+      "epoch": 4.538032495352033,
+      "grad_norm": 0.8886825442314148,
+      "learning_rate": 0.0002,
+      "loss": 0.5827,
+      "step": 28070
+    },
+    {
+      "epoch": 4.539649179532779,
+      "grad_norm": 0.8153995275497437,
+      "learning_rate": 0.0002,
+      "loss": 0.5605,
+      "step": 28080
+    },
+    {
+      "epoch": 4.541265863713524,
+      "grad_norm": 0.9853341579437256,
+      "learning_rate": 0.0002,
+      "loss": 0.6377,
+      "step": 28090
+    },
+    {
+      "epoch": 4.542882547894269,
+      "grad_norm": 0.9365800023078918,
+      "learning_rate": 0.0002,
+      "loss": 0.5957,
+      "step": 28100
+    },
+    {
+      "epoch": 4.544499232075014,
+      "grad_norm": 0.9765017628669739,
+      "learning_rate": 0.0002,
+      "loss": 0.5477,
+      "step": 28110
+    },
+    {
+      "epoch": 4.54611591625576,
+      "grad_norm": 0.9811279773712158,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 28120
+    },
+    {
+      "epoch": 4.547732600436505,
+      "grad_norm": 1.0387924909591675,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 28130
+    },
+    {
+      "epoch": 4.54934928461725,
+      "grad_norm": 1.0684878826141357,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 28140
+    },
+    {
+      "epoch": 4.550965968797995,
+      "grad_norm": 1.0000102519989014,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 28150
+    },
+    {
+      "epoch": 4.5525826529787405,
+      "grad_norm": 1.0717930793762207,
+      "learning_rate": 0.0002,
+      "loss": 0.5327,
+      "step": 28160
+    },
+    {
+      "epoch": 4.554199337159486,
+      "grad_norm": 0.990074634552002,
+      "learning_rate": 0.0002,
+      "loss": 0.5594,
+      "step": 28170
+    },
+    {
+      "epoch": 4.555816021340231,
+      "grad_norm": 0.8673754930496216,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 28180
+    },
+    {
+      "epoch": 4.557432705520976,
+      "grad_norm": 0.864247739315033,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 28190
+    },
+    {
+      "epoch": 4.5590493897017215,
+      "grad_norm": 0.8280200958251953,
+      "learning_rate": 0.0002,
+      "loss": 0.5516,
+      "step": 28200
+    },
+    {
+      "epoch": 4.560666073882467,
+      "grad_norm": 1.1312172412872314,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 28210
+    },
+    {
+      "epoch": 4.562282758063212,
+      "grad_norm": 0.9147403240203857,
+      "learning_rate": 0.0002,
+      "loss": 0.5776,
+      "step": 28220
+    },
+    {
+      "epoch": 4.563899442243958,
+      "grad_norm": 1.0321218967437744,
+      "learning_rate": 0.0002,
+      "loss": 0.5591,
+      "step": 28230
+    },
+    {
+      "epoch": 4.565516126424703,
+      "grad_norm": 1.168332815170288,
+      "learning_rate": 0.0002,
+      "loss": 0.5508,
+      "step": 28240
+    },
+    {
+      "epoch": 4.5671328106054485,
+      "grad_norm": 1.0067222118377686,
+      "learning_rate": 0.0002,
+      "loss": 0.5649,
+      "step": 28250
+    },
+    {
+      "epoch": 4.568749494786194,
+      "grad_norm": 1.0283393859863281,
+      "learning_rate": 0.0002,
+      "loss": 0.5853,
+      "step": 28260
+    },
+    {
+      "epoch": 4.570366178966939,
+      "grad_norm": 0.9912363886833191,
+      "learning_rate": 0.0002,
+      "loss": 0.5772,
+      "step": 28270
+    },
+    {
+      "epoch": 4.571982863147684,
+      "grad_norm": 1.108032464981079,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 28280
+    },
+    {
+      "epoch": 4.573599547328429,
+      "grad_norm": 0.8260078430175781,
+      "learning_rate": 0.0002,
+      "loss": 0.5529,
+      "step": 28290
+    },
+    {
+      "epoch": 4.575216231509175,
+      "grad_norm": 0.8946247100830078,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 28300
+    },
+    {
+      "epoch": 4.57683291568992,
+      "grad_norm": 0.8273587822914124,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 28310
+    },
+    {
+      "epoch": 4.578449599870665,
+      "grad_norm": 0.9040093421936035,
+      "learning_rate": 0.0002,
+      "loss": 0.6058,
+      "step": 28320
+    },
+    {
+      "epoch": 4.58006628405141,
+      "grad_norm": 0.8435290455818176,
+      "learning_rate": 0.0002,
+      "loss": 0.5521,
+      "step": 28330
+    },
+    {
+      "epoch": 4.581682968232156,
+      "grad_norm": 1.164088249206543,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 28340
+    },
+    {
+      "epoch": 4.583299652412901,
+      "grad_norm": 0.9861085414886475,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28350
+    },
+    {
+      "epoch": 4.584916336593646,
+      "grad_norm": 0.8892980813980103,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 28360
+    },
+    {
+      "epoch": 4.586533020774391,
+      "grad_norm": 1.240574836730957,
+      "learning_rate": 0.0002,
+      "loss": 0.598,
+      "step": 28370
+    },
+    {
+      "epoch": 4.588149704955137,
+      "grad_norm": 0.8669408559799194,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 28380
+    },
+    {
+      "epoch": 4.589766389135883,
+      "grad_norm": 0.9145985841751099,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28390
+    },
+    {
+      "epoch": 4.591383073316628,
+      "grad_norm": 0.8584614992141724,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 28400
+    },
+    {
+      "epoch": 4.592999757497373,
+      "grad_norm": 1.118829369544983,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 28410
+    },
+    {
+      "epoch": 4.594616441678118,
+      "grad_norm": 1.1411553621292114,
+      "learning_rate": 0.0002,
+      "loss": 0.5641,
+      "step": 28420
+    },
+    {
+      "epoch": 4.596233125858864,
+      "grad_norm": 0.9433278441429138,
+      "learning_rate": 0.0002,
+      "loss": 0.549,
+      "step": 28430
+    },
+    {
+      "epoch": 4.597849810039609,
+      "grad_norm": 0.816830039024353,
+      "learning_rate": 0.0002,
+      "loss": 0.5496,
+      "step": 28440
+    },
+    {
+      "epoch": 4.599466494220354,
+      "grad_norm": 1.2124968767166138,
+      "learning_rate": 0.0002,
+      "loss": 0.5543,
+      "step": 28450
+    },
+    {
+      "epoch": 4.601083178401099,
+      "grad_norm": 0.9658762216567993,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 28460
+    },
+    {
+      "epoch": 4.6026998625818445,
+      "grad_norm": 0.836100161075592,
+      "learning_rate": 0.0002,
+      "loss": 0.5902,
+      "step": 28470
+    },
+    {
+      "epoch": 4.60431654676259,
+      "grad_norm": 0.9989104270935059,
+      "learning_rate": 0.0002,
+      "loss": 0.5749,
+      "step": 28480
+    },
+    {
+      "epoch": 4.605933230943335,
+      "grad_norm": 1.1298956871032715,
+      "learning_rate": 0.0002,
+      "loss": 0.5616,
+      "step": 28490
+    },
+    {
+      "epoch": 4.60754991512408,
+      "grad_norm": 1.1731704473495483,
+      "learning_rate": 0.0002,
+      "loss": 0.5846,
+      "step": 28500
+    },
+    {
+      "epoch": 4.609166599304825,
+      "grad_norm": 0.9624714255332947,
+      "learning_rate": 0.0002,
+      "loss": 0.5816,
+      "step": 28510
+    },
+    {
+      "epoch": 4.610783283485571,
+      "grad_norm": 1.364073634147644,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 28520
+    },
+    {
+      "epoch": 4.612399967666317,
+      "grad_norm": 1.1827356815338135,
+      "learning_rate": 0.0002,
+      "loss": 0.6237,
+      "step": 28530
+    },
+    {
+      "epoch": 4.614016651847062,
+      "grad_norm": 0.6651531457901001,
+      "learning_rate": 0.0002,
+      "loss": 0.5643,
+      "step": 28540
+    },
+    {
+      "epoch": 4.615633336027807,
+      "grad_norm": 1.1640995740890503,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 28550
+    },
+    {
+      "epoch": 4.6172500202085525,
+      "grad_norm": 1.028918743133545,
+      "learning_rate": 0.0002,
+      "loss": 0.5995,
+      "step": 28560
+    },
+    {
+      "epoch": 4.618866704389298,
+      "grad_norm": 0.8252120614051819,
+      "learning_rate": 0.0002,
+      "loss": 0.5607,
+      "step": 28570
+    },
+    {
+      "epoch": 4.620483388570043,
+      "grad_norm": 1.3536735773086548,
+      "learning_rate": 0.0002,
+      "loss": 0.5769,
+      "step": 28580
+    },
+    {
+      "epoch": 4.622100072750788,
+      "grad_norm": 1.2146915197372437,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 28590
+    },
+    {
+      "epoch": 4.623716756931533,
+      "grad_norm": 1.0122549533843994,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 28600
+    },
+    {
+      "epoch": 4.625333441112279,
+      "grad_norm": 0.9977872967720032,
+      "learning_rate": 0.0002,
+      "loss": 0.6072,
+      "step": 28610
+    },
+    {
+      "epoch": 4.626950125293024,
+      "grad_norm": 1.0159751176834106,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 28620
+    },
+    {
+      "epoch": 4.628566809473769,
+      "grad_norm": 1.0028325319290161,
+      "learning_rate": 0.0002,
+      "loss": 0.5935,
+      "step": 28630
+    },
+    {
+      "epoch": 4.630183493654514,
+      "grad_norm": 0.901638388633728,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 28640
+    },
+    {
+      "epoch": 4.6318001778352595,
+      "grad_norm": 0.9450507164001465,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 28650
+    },
+    {
+      "epoch": 4.633416862016006,
+      "grad_norm": 0.9987545013427734,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 28660
+    },
+    {
+      "epoch": 4.63503354619675,
+      "grad_norm": 0.9574332237243652,
+      "learning_rate": 0.0002,
+      "loss": 0.5863,
+      "step": 28670
+    },
+    {
+      "epoch": 4.636650230377496,
+      "grad_norm": 1.2215653657913208,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 28680
+    },
+    {
+      "epoch": 4.638266914558241,
+      "grad_norm": 0.9798858761787415,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 28690
+    },
+    {
+      "epoch": 4.639883598738987,
+      "grad_norm": 1.0648466348648071,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 28700
+    },
+    {
+      "epoch": 4.641500282919732,
+      "grad_norm": 1.0606504678726196,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 28710
+    },
+    {
+      "epoch": 4.643116967100477,
+      "grad_norm": 1.0892442464828491,
+      "learning_rate": 0.0002,
+      "loss": 0.5801,
+      "step": 28720
+    },
+    {
+      "epoch": 4.644733651281222,
+      "grad_norm": 0.914391040802002,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 28730
+    },
+    {
+      "epoch": 4.6463503354619675,
+      "grad_norm": 0.9782370328903198,
+      "learning_rate": 0.0002,
+      "loss": 0.5439,
+      "step": 28740
+    },
+    {
+      "epoch": 4.647967019642713,
+      "grad_norm": 1.0344339609146118,
+      "learning_rate": 0.0002,
+      "loss": 0.6035,
+      "step": 28750
+    },
+    {
+      "epoch": 4.649583703823458,
+      "grad_norm": 1.0513931512832642,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 28760
+    },
+    {
+      "epoch": 4.651200388004203,
+      "grad_norm": 0.9711475968360901,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 28770
+    },
+    {
+      "epoch": 4.652817072184948,
+      "grad_norm": 0.977519690990448,
+      "learning_rate": 0.0002,
+      "loss": 0.5472,
+      "step": 28780
+    },
+    {
+      "epoch": 4.654433756365694,
+      "grad_norm": 0.9150224924087524,
+      "learning_rate": 0.0002,
+      "loss": 0.5826,
+      "step": 28790
+    },
+    {
+      "epoch": 4.656050440546439,
+      "grad_norm": 1.0973542928695679,
+      "learning_rate": 0.0002,
+      "loss": 0.5382,
+      "step": 28800
+    },
+    {
+      "epoch": 4.657667124727185,
+      "grad_norm": 0.944877564907074,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 28810
+    },
+    {
+      "epoch": 4.659283808907929,
+      "grad_norm": 0.9508748650550842,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 28820
+    },
+    {
+      "epoch": 4.6609004930886755,
+      "grad_norm": 0.9681721329689026,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 28830
+    },
+    {
+      "epoch": 4.662517177269421,
+      "grad_norm": 1.0214351415634155,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 28840
+    },
+    {
+      "epoch": 4.664133861450166,
+      "grad_norm": 0.9748611450195312,
+      "learning_rate": 0.0002,
+      "loss": 0.6031,
+      "step": 28850
+    },
+    {
+      "epoch": 4.665750545630911,
+      "grad_norm": 0.8484147191047668,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 28860
+    },
+    {
+      "epoch": 4.667367229811656,
+      "grad_norm": 1.1252986192703247,
+      "learning_rate": 0.0002,
+      "loss": 0.5699,
+      "step": 28870
+    },
+    {
+      "epoch": 4.668983913992402,
+      "grad_norm": 0.8706206679344177,
+      "learning_rate": 0.0002,
+      "loss": 0.5724,
+      "step": 28880
+    },
+    {
+      "epoch": 4.670600598173147,
+      "grad_norm": 1.1432424783706665,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 28890
+    },
+    {
+      "epoch": 4.672217282353892,
+      "grad_norm": 1.017029047012329,
+      "learning_rate": 0.0002,
+      "loss": 0.5675,
+      "step": 28900
+    },
+    {
+      "epoch": 4.673833966534637,
+      "grad_norm": 1.085597038269043,
+      "learning_rate": 0.0002,
+      "loss": 0.5831,
+      "step": 28910
+    },
+    {
+      "epoch": 4.675450650715383,
+      "grad_norm": 0.9275796413421631,
+      "learning_rate": 0.0002,
+      "loss": 0.5678,
+      "step": 28920
+    },
+    {
+      "epoch": 4.677067334896128,
+      "grad_norm": 0.9518964886665344,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28930
+    },
+    {
+      "epoch": 4.678684019076873,
+      "grad_norm": 1.0352122783660889,
+      "learning_rate": 0.0002,
+      "loss": 0.6232,
+      "step": 28940
+    },
+    {
+      "epoch": 4.680300703257618,
+      "grad_norm": 1.090124249458313,
+      "learning_rate": 0.0002,
+      "loss": 0.5786,
+      "step": 28950
+    },
+    {
+      "epoch": 4.681917387438364,
+      "grad_norm": 0.8799563050270081,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 28960
+    },
+    {
+      "epoch": 4.683534071619109,
+      "grad_norm": 1.0929821729660034,
+      "learning_rate": 0.0002,
+      "loss": 0.5787,
+      "step": 28970
+    },
+    {
+      "epoch": 4.685150755799855,
+      "grad_norm": 0.903727650642395,
+      "learning_rate": 0.0002,
+      "loss": 0.6134,
+      "step": 28980
+    },
+    {
+      "epoch": 4.6867674399806,
+      "grad_norm": 0.9752424955368042,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 28990
+    },
+    {
+      "epoch": 4.688384124161345,
+      "grad_norm": 0.9351571202278137,
+      "learning_rate": 0.0002,
+      "loss": 0.5762,
+      "step": 29000
+    },
+    {
+      "epoch": 4.6900008083420905,
+      "grad_norm": 0.923877477645874,
+      "learning_rate": 0.0002,
+      "loss": 0.5811,
+      "step": 29010
+    },
+    {
+      "epoch": 4.691617492522836,
+      "grad_norm": 1.045389175415039,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 29020
+    },
+    {
+      "epoch": 4.693234176703581,
+      "grad_norm": 1.0200831890106201,
+      "learning_rate": 0.0002,
+      "loss": 0.584,
+      "step": 29030
+    },
+    {
+      "epoch": 4.694850860884326,
+      "grad_norm": 1.1499706506729126,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 29040
+    },
+    {
+      "epoch": 4.6964675450650715,
+      "grad_norm": 0.860118567943573,
+      "learning_rate": 0.0002,
+      "loss": 0.5745,
+      "step": 29050
+    },
+    {
+      "epoch": 4.698084229245817,
+      "grad_norm": 0.9774864315986633,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 29060
+    },
+    {
+      "epoch": 4.699700913426562,
+      "grad_norm": 1.0323210954666138,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 29070
+    },
+    {
+      "epoch": 4.701317597607307,
+      "grad_norm": 0.8492481112480164,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 29080
+    },
+    {
+      "epoch": 4.702934281788052,
+      "grad_norm": 1.131951093673706,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 29090
+    },
+    {
+      "epoch": 4.704550965968798,
+      "grad_norm": 0.8763113021850586,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 29100
+    },
+    {
+      "epoch": 4.706167650149544,
+      "grad_norm": 1.045028805732727,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 29110
+    },
+    {
+      "epoch": 4.707784334330288,
+      "grad_norm": 0.9961401224136353,
+      "learning_rate": 0.0002,
+      "loss": 0.5548,
+      "step": 29120
+    },
+    {
+      "epoch": 4.709401018511034,
+      "grad_norm": 0.9282503724098206,
+      "learning_rate": 0.0002,
+      "loss": 0.559,
+      "step": 29130
+    },
+    {
+      "epoch": 4.711017702691779,
+      "grad_norm": 1.1418932676315308,
+      "learning_rate": 0.0002,
+      "loss": 0.5744,
+      "step": 29140
+    },
+    {
+      "epoch": 4.712634386872525,
+      "grad_norm": 0.9950099587440491,
+      "learning_rate": 0.0002,
+      "loss": 0.5394,
+      "step": 29150
+    },
+    {
+      "epoch": 4.71425107105327,
+      "grad_norm": 0.8304893374443054,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 29160
+    },
+    {
+      "epoch": 4.715867755234015,
+      "grad_norm": 1.115626335144043,
+      "learning_rate": 0.0002,
+      "loss": 0.6074,
+      "step": 29170
+    },
+    {
+      "epoch": 4.71748443941476,
+      "grad_norm": 1.079818606376648,
+      "learning_rate": 0.0002,
+      "loss": 0.6265,
+      "step": 29180
+    },
+    {
+      "epoch": 4.719101123595506,
+      "grad_norm": 1.1929082870483398,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 29190
+    },
+    {
+      "epoch": 4.720717807776251,
+      "grad_norm": 0.9621080756187439,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 29200
+    },
+    {
+      "epoch": 4.722334491956996,
+      "grad_norm": 0.8549222350120544,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 29210
+    },
+    {
+      "epoch": 4.723951176137741,
+      "grad_norm": 0.9341941475868225,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 29220
+    },
+    {
+      "epoch": 4.7255678603184865,
+      "grad_norm": 1.075406789779663,
+      "learning_rate": 0.0002,
+      "loss": 0.5436,
+      "step": 29230
+    },
+    {
+      "epoch": 4.727184544499232,
+      "grad_norm": 1.0859880447387695,
+      "learning_rate": 0.0002,
+      "loss": 0.576,
+      "step": 29240
+    },
+    {
+      "epoch": 4.728801228679977,
+      "grad_norm": 0.8475605249404907,
+      "learning_rate": 0.0002,
+      "loss": 0.5525,
+      "step": 29250
+    },
+    {
+      "epoch": 4.730417912860723,
+      "grad_norm": 0.9331845641136169,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 29260
+    },
+    {
+      "epoch": 4.7320345970414674,
+      "grad_norm": 0.9279314279556274,
+      "learning_rate": 0.0002,
+      "loss": 0.5901,
+      "step": 29270
+    },
+    {
+      "epoch": 4.733651281222214,
+      "grad_norm": 0.7803558707237244,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 29280
+    },
+    {
+      "epoch": 4.735267965402959,
+      "grad_norm": 1.0159329175949097,
+      "learning_rate": 0.0002,
+      "loss": 0.5968,
+      "step": 29290
+    },
+    {
+      "epoch": 4.736884649583704,
+      "grad_norm": 0.9448670744895935,
+      "learning_rate": 0.0002,
+      "loss": 0.5333,
+      "step": 29300
+    },
+    {
+      "epoch": 4.738501333764449,
+      "grad_norm": 1.0732197761535645,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 29310
+    },
+    {
+      "epoch": 4.7401180179451945,
+      "grad_norm": 0.901830792427063,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 29320
+    },
+    {
+      "epoch": 4.74173470212594,
+      "grad_norm": 0.9141789674758911,
+      "learning_rate": 0.0002,
+      "loss": 0.6105,
+      "step": 29330
+    },
+    {
+      "epoch": 4.743351386306685,
+      "grad_norm": 0.9733418226242065,
+      "learning_rate": 0.0002,
+      "loss": 0.5481,
+      "step": 29340
+    },
+    {
+      "epoch": 4.74496807048743,
+      "grad_norm": 0.909810483455658,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 29350
+    },
+    {
+      "epoch": 4.746584754668175,
+      "grad_norm": 0.909541666507721,
+      "learning_rate": 0.0002,
+      "loss": 0.5911,
+      "step": 29360
+    },
+    {
+      "epoch": 4.748201438848921,
+      "grad_norm": 0.9383015632629395,
+      "learning_rate": 0.0002,
+      "loss": 0.5579,
+      "step": 29370
+    },
+    {
+      "epoch": 4.749818123029666,
+      "grad_norm": 0.9275668263435364,
+      "learning_rate": 0.0002,
+      "loss": 0.5529,
+      "step": 29380
+    },
+    {
+      "epoch": 4.751434807210411,
+      "grad_norm": 1.1146225929260254,
+      "learning_rate": 0.0002,
+      "loss": 0.5623,
+      "step": 29390
+    },
+    {
+      "epoch": 4.753051491391156,
+      "grad_norm": 1.0062453746795654,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 29400
+    },
+    {
+      "epoch": 4.7546681755719025,
+      "grad_norm": 0.9451895952224731,
+      "learning_rate": 0.0002,
+      "loss": 0.5872,
+      "step": 29410
+    },
+    {
+      "epoch": 4.756284859752648,
+      "grad_norm": 0.870457649230957,
+      "learning_rate": 0.0002,
+      "loss": 0.5767,
+      "step": 29420
+    },
+    {
+      "epoch": 4.757901543933393,
+      "grad_norm": 1.0411282777786255,
+      "learning_rate": 0.0002,
+      "loss": 0.57,
+      "step": 29430
+    },
+    {
+      "epoch": 4.759518228114138,
+      "grad_norm": 1.1648986339569092,
+      "learning_rate": 0.0002,
+      "loss": 0.5688,
+      "step": 29440
+    },
+    {
+      "epoch": 4.761134912294883,
+      "grad_norm": 0.8999572992324829,
+      "learning_rate": 0.0002,
+      "loss": 0.5432,
+      "step": 29450
+    },
+    {
+      "epoch": 4.762751596475629,
+      "grad_norm": 0.9863559007644653,
+      "learning_rate": 0.0002,
+      "loss": 0.5667,
+      "step": 29460
+    },
+    {
+      "epoch": 4.764368280656374,
+      "grad_norm": 0.9676542282104492,
+      "learning_rate": 0.0002,
+      "loss": 0.5779,
+      "step": 29470
+    },
+    {
+      "epoch": 4.765984964837119,
+      "grad_norm": 1.004775047302246,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 29480
+    },
+    {
+      "epoch": 4.767601649017864,
+      "grad_norm": 1.0937515497207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6044,
+      "step": 29490
+    },
+    {
+      "epoch": 4.7692183331986095,
+      "grad_norm": 0.9551598429679871,
+      "learning_rate": 0.0002,
+      "loss": 0.5433,
+      "step": 29500
+    },
+    {
+      "epoch": 4.770835017379355,
+      "grad_norm": 1.0757228136062622,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 29510
+    },
+    {
+      "epoch": 4.7724517015601,
+      "grad_norm": 1.0588841438293457,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 29520
+    },
+    {
+      "epoch": 4.774068385740845,
+      "grad_norm": 1.0744032859802246,
+      "learning_rate": 0.0002,
+      "loss": 0.5814,
+      "step": 29530
+    },
+    {
+      "epoch": 4.7756850699215905,
+      "grad_norm": 1.0066277980804443,
+      "learning_rate": 0.0002,
+      "loss": 0.5681,
+      "step": 29540
+    },
+    {
+      "epoch": 4.777301754102336,
+      "grad_norm": 1.082319736480713,
+      "learning_rate": 0.0002,
+      "loss": 0.545,
+      "step": 29550
+    },
+    {
+      "epoch": 4.778918438283082,
+      "grad_norm": 0.8252472877502441,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 29560
+    },
+    {
+      "epoch": 4.780535122463827,
+      "grad_norm": 0.9855340123176575,
+      "learning_rate": 0.0002,
+      "loss": 0.5666,
+      "step": 29570
+    },
+    {
+      "epoch": 4.782151806644572,
+      "grad_norm": 0.9991421699523926,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 29580
+    },
+    {
+      "epoch": 4.7837684908253175,
+      "grad_norm": 1.316841959953308,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 29590
+    },
+    {
+      "epoch": 4.785385175006063,
+      "grad_norm": 1.1513035297393799,
+      "learning_rate": 0.0002,
+      "loss": 0.6102,
+      "step": 29600
+    },
+    {
+      "epoch": 4.787001859186808,
+      "grad_norm": 0.9767683744430542,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 29610
+    },
+    {
+      "epoch": 4.788618543367553,
+      "grad_norm": 0.9786278605461121,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 29620
+    },
+    {
+      "epoch": 4.7902352275482984,
+      "grad_norm": 0.8004973530769348,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 29630
+    },
+    {
+      "epoch": 4.791851911729044,
+      "grad_norm": 1.0997767448425293,
+      "learning_rate": 0.0002,
+      "loss": 0.5932,
+      "step": 29640
+    },
+    {
+      "epoch": 4.793468595909789,
+      "grad_norm": 0.9752856492996216,
+      "learning_rate": 0.0002,
+      "loss": 0.5655,
+      "step": 29650
+    },
+    {
+      "epoch": 4.795085280090534,
+      "grad_norm": 1.0518392324447632,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 29660
+    },
+    {
+      "epoch": 4.796701964271279,
+      "grad_norm": 1.1050055027008057,
+      "learning_rate": 0.0002,
+      "loss": 0.6042,
+      "step": 29670
+    },
+    {
+      "epoch": 4.798318648452025,
+      "grad_norm": 0.9933857917785645,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 29680
+    },
+    {
+      "epoch": 4.79993533263277,
+      "grad_norm": 1.2804018259048462,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 29690
+    },
+    {
+      "epoch": 4.801552016813515,
+      "grad_norm": 1.0133371353149414,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 29700
+    },
+    {
+      "epoch": 4.803168700994261,
+      "grad_norm": 1.080350637435913,
+      "learning_rate": 0.0002,
+      "loss": 0.5662,
+      "step": 29710
+    },
+    {
+      "epoch": 4.804785385175006,
+      "grad_norm": 0.9986529350280762,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 29720
+    },
+    {
+      "epoch": 4.806402069355752,
+      "grad_norm": 0.975665807723999,
+      "learning_rate": 0.0002,
+      "loss": 0.5894,
+      "step": 29730
+    },
+    {
+      "epoch": 4.808018753536497,
+      "grad_norm": 0.8458138704299927,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 29740
+    },
+    {
+      "epoch": 4.809635437717242,
+      "grad_norm": 0.99330073595047,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 29750
+    },
+    {
+      "epoch": 4.811252121897987,
+      "grad_norm": 0.898274302482605,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 29760
+    },
+    {
+      "epoch": 4.812868806078733,
+      "grad_norm": 1.0504480600357056,
+      "learning_rate": 0.0002,
+      "loss": 0.5842,
+      "step": 29770
+    },
+    {
+      "epoch": 4.814485490259478,
+      "grad_norm": 0.937919020652771,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 29780
+    },
+    {
+      "epoch": 4.816102174440223,
+      "grad_norm": 0.9593307971954346,
+      "learning_rate": 0.0002,
+      "loss": 0.5885,
+      "step": 29790
+    },
+    {
+      "epoch": 4.817718858620968,
+      "grad_norm": 0.9431198835372925,
+      "learning_rate": 0.0002,
+      "loss": 0.578,
+      "step": 29800
+    },
+    {
+      "epoch": 4.8193355428017135,
+      "grad_norm": 1.2729957103729248,
+      "learning_rate": 0.0002,
+      "loss": 0.5739,
+      "step": 29810
+    },
+    {
+      "epoch": 4.820952226982459,
+      "grad_norm": 0.8876838684082031,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 29820
+    },
+    {
+      "epoch": 4.822568911163204,
+      "grad_norm": 1.0185000896453857,
+      "learning_rate": 0.0002,
+      "loss": 0.5583,
+      "step": 29830
+    },
+    {
+      "epoch": 4.824185595343949,
+      "grad_norm": 1.064276099205017,
+      "learning_rate": 0.0002,
+      "loss": 0.5686,
+      "step": 29840
+    },
+    {
+      "epoch": 4.825802279524694,
+      "grad_norm": 0.9774803519248962,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 29850
+    },
+    {
+      "epoch": 4.8274189637054405,
+      "grad_norm": 1.131646990776062,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 29860
+    },
+    {
+      "epoch": 4.829035647886186,
+      "grad_norm": 1.081455945968628,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 29870
+    },
+    {
+      "epoch": 4.830652332066931,
+      "grad_norm": 0.990538477897644,
+      "learning_rate": 0.0002,
+      "loss": 0.5793,
+      "step": 29880
+    },
+    {
+      "epoch": 4.832269016247676,
+      "grad_norm": 0.9750600457191467,
+      "learning_rate": 0.0002,
+      "loss": 0.5833,
+      "step": 29890
+    },
+    {
+      "epoch": 4.8338857004284215,
+      "grad_norm": 1.0600621700286865,
+      "learning_rate": 0.0002,
+      "loss": 0.619,
+      "step": 29900
+    },
+    {
+      "epoch": 4.835502384609167,
+      "grad_norm": 0.9237320423126221,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 29910
+    },
+    {
+      "epoch": 4.837119068789912,
+      "grad_norm": 0.9739177227020264,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 29920
+    },
+    {
+      "epoch": 4.838735752970657,
+      "grad_norm": 1.128677248954773,
+      "learning_rate": 0.0002,
+      "loss": 0.587,
+      "step": 29930
+    },
+    {
+      "epoch": 4.840352437151402,
+      "grad_norm": 1.042604923248291,
+      "learning_rate": 0.0002,
+      "loss": 0.564,
+      "step": 29940
+    },
+    {
+      "epoch": 4.841969121332148,
+      "grad_norm": 0.849758505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.5885,
+      "step": 29950
+    },
+    {
+      "epoch": 4.843585805512893,
+      "grad_norm": 1.2809888124465942,
+      "learning_rate": 0.0002,
+      "loss": 0.5952,
+      "step": 29960
+    },
+    {
+      "epoch": 4.845202489693638,
+      "grad_norm": 1.0177865028381348,
+      "learning_rate": 0.0002,
+      "loss": 0.5703,
+      "step": 29970
+    },
+    {
+      "epoch": 4.846819173874383,
+      "grad_norm": 1.0026639699935913,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 29980
+    },
+    {
+      "epoch": 4.8484358580551286,
+      "grad_norm": 0.9679505228996277,
+      "learning_rate": 0.0002,
+      "loss": 0.5897,
+      "step": 29990
+    },
+    {
+      "epoch": 4.850052542235874,
+      "grad_norm": 0.8939532041549683,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 30000
+    },
+    {
+      "epoch": 4.85166922641662,
+      "grad_norm": 0.9957457780838013,
+      "learning_rate": 0.0002,
+      "loss": 0.5852,
+      "step": 30010
+    },
+    {
+      "epoch": 4.853285910597365,
+      "grad_norm": 1.1646790504455566,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 30020
+    },
+    {
+      "epoch": 4.85490259477811,
+      "grad_norm": 0.8804680705070496,
+      "learning_rate": 0.0002,
+      "loss": 0.5711,
+      "step": 30030
+    },
+    {
+      "epoch": 4.856519278958856,
+      "grad_norm": 1.161970853805542,
+      "learning_rate": 0.0002,
+      "loss": 0.5397,
+      "step": 30040
+    },
+    {
+      "epoch": 4.858135963139601,
+      "grad_norm": 0.9081037640571594,
+      "learning_rate": 0.0002,
+      "loss": 0.5552,
+      "step": 30050
+    },
+    {
+      "epoch": 4.859752647320346,
+      "grad_norm": 0.9402848482131958,
+      "learning_rate": 0.0002,
+      "loss": 0.6024,
+      "step": 30060
+    },
+    {
+      "epoch": 4.861369331501091,
+      "grad_norm": 0.9023865461349487,
+      "learning_rate": 0.0002,
+      "loss": 0.6256,
+      "step": 30070
+    },
+    {
+      "epoch": 4.8629860156818365,
+      "grad_norm": 1.0173414945602417,
+      "learning_rate": 0.0002,
+      "loss": 0.5926,
+      "step": 30080
+    },
+    {
+      "epoch": 4.864602699862582,
+      "grad_norm": 1.084402322769165,
+      "learning_rate": 0.0002,
+      "loss": 0.6274,
+      "step": 30090
+    },
+    {
+      "epoch": 4.866219384043327,
+      "grad_norm": 0.9577937126159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6311,
+      "step": 30100
+    },
+    {
+      "epoch": 4.867836068224072,
+      "grad_norm": 0.9807606935501099,
+      "learning_rate": 0.0002,
+      "loss": 0.5724,
+      "step": 30110
+    },
+    {
+      "epoch": 4.8694527524048175,
+      "grad_norm": 0.978784441947937,
+      "learning_rate": 0.0002,
+      "loss": 0.5786,
+      "step": 30120
+    },
+    {
+      "epoch": 4.871069436585563,
+      "grad_norm": 0.9762914776802063,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 30130
+    },
+    {
+      "epoch": 4.872686120766308,
+      "grad_norm": 0.9404871463775635,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 30140
+    },
+    {
+      "epoch": 4.874302804947053,
+      "grad_norm": 1.0069509744644165,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 30150
+    },
+    {
+      "epoch": 4.875919489127799,
+      "grad_norm": 1.1770923137664795,
+      "learning_rate": 0.0002,
+      "loss": 0.6225,
+      "step": 30160
+    },
+    {
+      "epoch": 4.8775361733085445,
+      "grad_norm": 1.021210789680481,
+      "learning_rate": 0.0002,
+      "loss": 0.5657,
+      "step": 30170
+    },
+    {
+      "epoch": 4.87915285748929,
+      "grad_norm": 0.8512648940086365,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 30180
+    },
+    {
+      "epoch": 4.880769541670035,
+      "grad_norm": 0.9345870018005371,
+      "learning_rate": 0.0002,
+      "loss": 0.5519,
+      "step": 30190
+    },
+    {
+      "epoch": 4.88238622585078,
+      "grad_norm": 1.0224418640136719,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 30200
+    },
+    {
+      "epoch": 4.884002910031525,
+      "grad_norm": 1.0316044092178345,
+      "learning_rate": 0.0002,
+      "loss": 0.5807,
+      "step": 30210
+    },
+    {
+      "epoch": 4.885619594212271,
+      "grad_norm": 1.102437973022461,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 30220
+    },
+    {
+      "epoch": 4.887236278393016,
+      "grad_norm": 1.0220023393630981,
+      "learning_rate": 0.0002,
+      "loss": 0.586,
+      "step": 30230
+    },
+    {
+      "epoch": 4.888852962573761,
+      "grad_norm": 1.0934523344039917,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 30240
+    },
+    {
+      "epoch": 4.890469646754506,
+      "grad_norm": 1.264630913734436,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 30250
+    },
+    {
+      "epoch": 4.892086330935252,
+      "grad_norm": 1.0999879837036133,
+      "learning_rate": 0.0002,
+      "loss": 0.5712,
+      "step": 30260
+    },
+    {
+      "epoch": 4.893703015115997,
+      "grad_norm": 0.9124550223350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 30270
+    },
+    {
+      "epoch": 4.895319699296742,
+      "grad_norm": 0.9853624105453491,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 30280
+    },
+    {
+      "epoch": 4.896936383477488,
+      "grad_norm": 1.0589802265167236,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 30290
+    },
+    {
+      "epoch": 4.8985530676582325,
+      "grad_norm": 0.8487226366996765,
+      "learning_rate": 0.0002,
+      "loss": 0.6129,
+      "step": 30300
+    },
+    {
+      "epoch": 4.900169751838979,
+      "grad_norm": 1.0212191343307495,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 30310
+    },
+    {
+      "epoch": 4.901786436019724,
+      "grad_norm": 1.0187491178512573,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 30320
+    },
+    {
+      "epoch": 4.903403120200469,
+      "grad_norm": 1.0013091564178467,
+      "learning_rate": 0.0002,
+      "loss": 0.5809,
+      "step": 30330
+    },
+    {
+      "epoch": 4.905019804381214,
+      "grad_norm": 1.0017542839050293,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 30340
+    },
+    {
+      "epoch": 4.9066364885619596,
+      "grad_norm": 0.9665151238441467,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 30350
+    },
+    {
+      "epoch": 4.908253172742705,
+      "grad_norm": 0.8774822950363159,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 30360
+    },
+    {
+      "epoch": 4.90986985692345,
+      "grad_norm": 0.9449850916862488,
+      "learning_rate": 0.0002,
+      "loss": 0.5771,
+      "step": 30370
+    },
+    {
+      "epoch": 4.911486541104195,
+      "grad_norm": 0.7368341088294983,
+      "learning_rate": 0.0002,
+      "loss": 0.58,
+      "step": 30380
+    },
+    {
+      "epoch": 4.9131032252849405,
+      "grad_norm": 0.9669167995452881,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 30390
+    },
+    {
+      "epoch": 4.914719909465686,
+      "grad_norm": 1.1227794885635376,
+      "learning_rate": 0.0002,
+      "loss": 0.6202,
+      "step": 30400
+    },
+    {
+      "epoch": 4.916336593646431,
+      "grad_norm": 0.9884361028671265,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 30410
+    },
+    {
+      "epoch": 4.917953277827176,
+      "grad_norm": 0.9949551224708557,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 30420
+    },
+    {
+      "epoch": 4.919569962007921,
+      "grad_norm": 0.9491621851921082,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 30430
+    },
+    {
+      "epoch": 4.9211866461886675,
+      "grad_norm": 0.78848797082901,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 30440
+    },
+    {
+      "epoch": 4.922803330369412,
+      "grad_norm": 1.0693835020065308,
+      "learning_rate": 0.0002,
+      "loss": 0.5561,
+      "step": 30450
+    },
+    {
+      "epoch": 4.924420014550158,
+      "grad_norm": 0.9573729634284973,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 30460
+    },
+    {
+      "epoch": 4.926036698730903,
+      "grad_norm": 0.9975152611732483,
+      "learning_rate": 0.0002,
+      "loss": 0.6084,
+      "step": 30470
+    },
+    {
+      "epoch": 4.9276533829116484,
+      "grad_norm": 0.8695693016052246,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 30480
+    },
+    {
+      "epoch": 4.929270067092394,
+      "grad_norm": 1.145394206047058,
+      "learning_rate": 0.0002,
+      "loss": 0.6144,
+      "step": 30490
+    },
+    {
+      "epoch": 4.930886751273139,
+      "grad_norm": 0.7668989896774292,
+      "learning_rate": 0.0002,
+      "loss": 0.5736,
+      "step": 30500
+    },
+    {
+      "epoch": 4.932503435453884,
+      "grad_norm": 0.9630151391029358,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 30510
+    },
+    {
+      "epoch": 4.934120119634629,
+      "grad_norm": 0.940705418586731,
+      "learning_rate": 0.0002,
+      "loss": 0.6461,
+      "step": 30520
+    },
+    {
+      "epoch": 4.935736803815375,
+      "grad_norm": 1.3243348598480225,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 30530
+    },
+    {
+      "epoch": 4.93735348799612,
+      "grad_norm": 1.004347801208496,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 30540
+    },
+    {
+      "epoch": 4.938970172176865,
+      "grad_norm": 0.8711541295051575,
+      "learning_rate": 0.0002,
+      "loss": 0.583,
+      "step": 30550
+    },
+    {
+      "epoch": 4.94058685635761,
+      "grad_norm": 0.8980631828308105,
+      "learning_rate": 0.0002,
+      "loss": 0.599,
+      "step": 30560
+    },
+    {
+      "epoch": 4.9422035405383555,
+      "grad_norm": 0.8388893604278564,
+      "learning_rate": 0.0002,
+      "loss": 0.6024,
+      "step": 30570
+    },
+    {
+      "epoch": 4.943820224719101,
+      "grad_norm": 1.0991183519363403,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 30580
+    },
+    {
+      "epoch": 4.945436908899847,
+      "grad_norm": 0.9731075763702393,
+      "learning_rate": 0.0002,
+      "loss": 0.5906,
+      "step": 30590
+    },
+    {
+      "epoch": 4.947053593080591,
+      "grad_norm": 1.3904452323913574,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 30600
+    },
+    {
+      "epoch": 4.948670277261337,
+      "grad_norm": 1.2489882707595825,
+      "learning_rate": 0.0002,
+      "loss": 0.5952,
+      "step": 30610
+    },
+    {
+      "epoch": 4.950286961442083,
+      "grad_norm": 1.240072250366211,
+      "learning_rate": 0.0002,
+      "loss": 0.5887,
+      "step": 30620
+    },
+    {
+      "epoch": 4.951903645622828,
+      "grad_norm": 0.9191411733627319,
+      "learning_rate": 0.0002,
+      "loss": 0.5762,
+      "step": 30630
+    },
+    {
+      "epoch": 4.953520329803573,
+      "grad_norm": 0.8888895511627197,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 30640
+    },
+    {
+      "epoch": 4.955137013984318,
+      "grad_norm": 0.9001450538635254,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 30650
+    },
+    {
+      "epoch": 4.9567536981650635,
+      "grad_norm": 1.053971767425537,
+      "learning_rate": 0.0002,
+      "loss": 0.6047,
+      "step": 30660
+    },
+    {
+      "epoch": 4.958370382345809,
+      "grad_norm": 1.2224042415618896,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 30670
+    },
+    {
+      "epoch": 4.959987066526554,
+      "grad_norm": 0.8855111598968506,
+      "learning_rate": 0.0002,
+      "loss": 0.6211,
+      "step": 30680
+    },
+    {
+      "epoch": 4.961603750707299,
+      "grad_norm": 0.9489575624465942,
+      "learning_rate": 0.0002,
+      "loss": 0.5764,
+      "step": 30690
+    },
+    {
+      "epoch": 4.963220434888044,
+      "grad_norm": 0.9635404944419861,
+      "learning_rate": 0.0002,
+      "loss": 0.5371,
+      "step": 30700
+    },
+    {
+      "epoch": 4.96483711906879,
+      "grad_norm": 1.1784121990203857,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 30710
+    },
+    {
+      "epoch": 4.966453803249535,
+      "grad_norm": 1.0059462785720825,
+      "learning_rate": 0.0002,
+      "loss": 0.5803,
+      "step": 30720
+    },
+    {
+      "epoch": 4.96807048743028,
+      "grad_norm": 0.9479738473892212,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 30730
+    },
+    {
+      "epoch": 4.969687171611026,
+      "grad_norm": 1.0624593496322632,
+      "learning_rate": 0.0002,
+      "loss": 0.584,
+      "step": 30740
+    },
+    {
+      "epoch": 4.971303855791771,
+      "grad_norm": 1.1429259777069092,
+      "learning_rate": 0.0002,
+      "loss": 0.6202,
+      "step": 30750
+    },
+    {
+      "epoch": 4.972920539972517,
+      "grad_norm": 0.9102491140365601,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 30760
+    },
+    {
+      "epoch": 4.974537224153262,
+      "grad_norm": 1.1262688636779785,
+      "learning_rate": 0.0002,
+      "loss": 0.6025,
+      "step": 30770
+    },
+    {
+      "epoch": 4.976153908334007,
+      "grad_norm": 1.1415393352508545,
+      "learning_rate": 0.0002,
+      "loss": 0.588,
+      "step": 30780
+    },
+    {
+      "epoch": 4.977770592514752,
+      "grad_norm": 1.083078384399414,
+      "learning_rate": 0.0002,
+      "loss": 0.5832,
+      "step": 30790
+    },
+    {
+      "epoch": 4.979387276695498,
+      "grad_norm": 0.964859127998352,
+      "learning_rate": 0.0002,
+      "loss": 0.6025,
+      "step": 30800
+    },
+    {
+      "epoch": 4.981003960876243,
+      "grad_norm": 0.8704743385314941,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 30810
+    },
+    {
+      "epoch": 4.982620645056988,
+      "grad_norm": 1.0714856386184692,
+      "learning_rate": 0.0002,
+      "loss": 0.5666,
+      "step": 30820
+    },
+    {
+      "epoch": 4.984237329237733,
+      "grad_norm": 0.6818771362304688,
+      "learning_rate": 0.0002,
+      "loss": 0.565,
+      "step": 30830
+    },
+    {
+      "epoch": 4.985854013418479,
+      "grad_norm": 1.0454156398773193,
+      "learning_rate": 0.0002,
+      "loss": 0.5999,
+      "step": 30840
+    },
+    {
+      "epoch": 4.987470697599224,
+      "grad_norm": 0.9410776495933533,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 30850
+    },
+    {
+      "epoch": 4.989087381779969,
+      "grad_norm": 1.0878902673721313,
+      "learning_rate": 0.0002,
+      "loss": 0.5899,
+      "step": 30860
+    },
+    {
+      "epoch": 4.990704065960714,
+      "grad_norm": 0.8916727304458618,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 30870
+    },
+    {
+      "epoch": 4.9923207501414595,
+      "grad_norm": 1.045776128768921,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 30880
+    },
+    {
+      "epoch": 4.993937434322206,
+      "grad_norm": 0.9861903786659241,
+      "learning_rate": 0.0002,
+      "loss": 0.5767,
+      "step": 30890
+    },
+    {
+      "epoch": 4.995554118502951,
+      "grad_norm": 0.9275050759315491,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 30900
+    },
+    {
+      "epoch": 4.997170802683696,
+      "grad_norm": 0.94013911485672,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 30910
+    },
+    {
+      "epoch": 4.998787486864441,
+      "grad_norm": 0.9771268367767334,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 30920
+    },
+    {
+      "epoch": 4.9999191657909625,
+      "eval_loss": 1.1968598365783691,
+      "eval_runtime": 122.2519,
+      "eval_samples_per_second": 5.996,
+      "eval_steps_per_second": 0.753,
+      "step": 30927
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 49480,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.431255450522747e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/training_args.bin b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..155b12fa9acbc6e71dba75c92bfa79e152397ebf
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-30927/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28694d5564a2b5c7d6881d4ba2af103356aa22489d2c22768ebbe47283c0f4a1
+size 5560
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/README.md b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..835e31ab8469ee39ddc8b2b6b2143a8c66dad510
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/README.md
@@ -0,0 +1,202 @@
+---
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.13.1
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/adapter_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..406c5a08dc4a2a33b52c62a482f98c217c417215
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/adapter_config.json
@@ -0,0 +1,29 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/adapter_model.safetensors b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..bf74e91baddd9bf12e2d15e86e17b59b8709c514
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae20f88a8da1f25d309771826b573d955d59bf638bfd7ebe56d06432eb2bcd57
+size 109069176
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/optimizer.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..000e7d4eeba61672fe5b70b5762a146650730c8a
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d770a15916660cc35507ef27edaa08ece197ee3ae29b8d7c8422d0dc1def1a85
+size 55532666
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/rng_state.pth b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..f6f721479c01f456f61ca73426e82ddb0ae7df41
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59aa160eb003909b59eb576bcfcd615685360c81e70469d31aeea715df40d745
+size 14244
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/scheduler.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..10639d46460f861cbf1977ce29feebed3e09bd6f
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a13545656639db3126c168548c9c958ee9173553480b5b708d0d8b5933af97ad
+size 1064
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/special_tokens_map.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/special_tokens_map.json
@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/tokenizer.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..86a33946b0c77216d2cce91bb28c8fada4a5e80b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/tokenizer_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..061e40d9db3253624f86e8e364c15ef546527c9d
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/tokenizer_config.json
@@ -0,0 +1,2063 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/trainer_state.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..da5362fe48a08695048e61a59e11aefcd7700e25
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/trainer_state.json
@@ -0,0 +1,26058 @@
+{
+  "best_metric": 1.0871200561523438,
+  "best_model_checkpoint": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185",
+  "epoch": 6.0,
+  "eval_steps": 10,
+  "global_step": 37113,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016166841807452913,
+      "grad_norm": 0.9894065856933594,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032333683614905826,
+      "grad_norm": 1.7810699939727783,
+      "learning_rate": 0.0002,
+      "loss": 1.1528,
+      "step": 20
+    },
+    {
+      "epoch": 0.004850052542235874,
+      "grad_norm": 0.5969577431678772,
+      "learning_rate": 0.0002,
+      "loss": 0.9767,
+      "step": 30
+    },
+    {
+      "epoch": 0.006466736722981165,
+      "grad_norm": 0.6354120969772339,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 40
+    },
+    {
+      "epoch": 0.008083420903726457,
+      "grad_norm": 0.5604607462882996,
+      "learning_rate": 0.0002,
+      "loss": 0.8643,
+      "step": 50
+    },
+    {
+      "epoch": 0.009700105084471748,
+      "grad_norm": 0.4676193594932556,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 60
+    },
+    {
+      "epoch": 0.01131678926521704,
+      "grad_norm": 0.6099211573600769,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 70
+    },
+    {
+      "epoch": 0.01293347344596233,
+      "grad_norm": 0.48639994859695435,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 80
+    },
+    {
+      "epoch": 0.014550157626707623,
+      "grad_norm": 0.4904264509677887,
+      "learning_rate": 0.0002,
+      "loss": 0.8704,
+      "step": 90
+    },
+    {
+      "epoch": 0.016166841807452915,
+      "grad_norm": 2.8334362506866455,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 100
+    },
+    {
+      "epoch": 0.017783525988198205,
+      "grad_norm": 0.43221670389175415,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 110
+    },
+    {
+      "epoch": 0.019400210168943496,
+      "grad_norm": 0.42244166135787964,
+      "learning_rate": 0.0002,
+      "loss": 0.8412,
+      "step": 120
+    },
+    {
+      "epoch": 0.02101689434968879,
+      "grad_norm": 0.45363298058509827,
+      "learning_rate": 0.0002,
+      "loss": 0.8467,
+      "step": 130
+    },
+    {
+      "epoch": 0.02263357853043408,
+      "grad_norm": 0.44816508889198303,
+      "learning_rate": 0.0002,
+      "loss": 0.8641,
+      "step": 140
+    },
+    {
+      "epoch": 0.02425026271117937,
+      "grad_norm": 0.43308213353157043,
+      "learning_rate": 0.0002,
+      "loss": 0.8496,
+      "step": 150
+    },
+    {
+      "epoch": 0.02586694689192466,
+      "grad_norm": 0.4084763526916504,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 160
+    },
+    {
+      "epoch": 0.027483631072669955,
+      "grad_norm": 0.5363703966140747,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 170
+    },
+    {
+      "epoch": 0.029100315253415245,
+      "grad_norm": 0.4619699716567993,
+      "learning_rate": 0.0002,
+      "loss": 0.8558,
+      "step": 180
+    },
+    {
+      "epoch": 0.030716999434160536,
+      "grad_norm": 0.49069908261299133,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 190
+    },
+    {
+      "epoch": 0.03233368361490583,
+      "grad_norm": 0.4645835757255554,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 200
+    },
+    {
+      "epoch": 0.03395036779565112,
+      "grad_norm": 1.2411243915557861,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 210
+    },
+    {
+      "epoch": 0.03556705197639641,
+      "grad_norm": 0.5211851596832275,
+      "learning_rate": 0.0002,
+      "loss": 0.8245,
+      "step": 220
+    },
+    {
+      "epoch": 0.037183736157141704,
+      "grad_norm": 0.5253691673278809,
+      "learning_rate": 0.0002,
+      "loss": 0.8194,
+      "step": 230
+    },
+    {
+      "epoch": 0.03880042033788699,
+      "grad_norm": 0.4567478895187378,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 240
+    },
+    {
+      "epoch": 0.040417104518632285,
+      "grad_norm": 0.5472128391265869,
+      "learning_rate": 0.0002,
+      "loss": 0.838,
+      "step": 250
+    },
+    {
+      "epoch": 0.04203378869937758,
+      "grad_norm": 0.42978546023368835,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 260
+    },
+    {
+      "epoch": 0.043650472880122866,
+      "grad_norm": 0.601734459400177,
+      "learning_rate": 0.0002,
+      "loss": 0.8334,
+      "step": 270
+    },
+    {
+      "epoch": 0.04526715706086816,
+      "grad_norm": 0.4286513328552246,
+      "learning_rate": 0.0002,
+      "loss": 0.815,
+      "step": 280
+    },
+    {
+      "epoch": 0.046883841241613454,
+      "grad_norm": 0.5230861902236938,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 290
+    },
+    {
+      "epoch": 0.04850052542235874,
+      "grad_norm": 0.6504611968994141,
+      "learning_rate": 0.0002,
+      "loss": 0.8636,
+      "step": 300
+    },
+    {
+      "epoch": 0.050117209603104035,
+      "grad_norm": 0.43485215306282043,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 310
+    },
+    {
+      "epoch": 0.05173389378384932,
+      "grad_norm": 0.4717007875442505,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 320
+    },
+    {
+      "epoch": 0.053350577964594616,
+      "grad_norm": 0.4059787690639496,
+      "learning_rate": 0.0002,
+      "loss": 0.8469,
+      "step": 330
+    },
+    {
+      "epoch": 0.05496726214533991,
+      "grad_norm": 0.4366913437843323,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 340
+    },
+    {
+      "epoch": 0.0565839463260852,
+      "grad_norm": 0.4233848452568054,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 350
+    },
+    {
+      "epoch": 0.05820063050683049,
+      "grad_norm": 0.4209108352661133,
+      "learning_rate": 0.0002,
+      "loss": 0.8456,
+      "step": 360
+    },
+    {
+      "epoch": 0.059817314687575784,
+      "grad_norm": 0.41637396812438965,
+      "learning_rate": 0.0002,
+      "loss": 0.816,
+      "step": 370
+    },
+    {
+      "epoch": 0.06143399886832107,
+      "grad_norm": 0.46235376596450806,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 380
+    },
+    {
+      "epoch": 0.06305068304906636,
+      "grad_norm": 0.4013484716415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 390
+    },
+    {
+      "epoch": 0.06466736722981166,
+      "grad_norm": 0.47443896532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 400
+    },
+    {
+      "epoch": 0.06628405141055695,
+      "grad_norm": 0.3942156434059143,
+      "learning_rate": 0.0002,
+      "loss": 0.8666,
+      "step": 410
+    },
+    {
+      "epoch": 0.06790073559130223,
+      "grad_norm": 0.4965320825576782,
+      "learning_rate": 0.0002,
+      "loss": 0.8402,
+      "step": 420
+    },
+    {
+      "epoch": 0.06951741977204753,
+      "grad_norm": 0.4304835796356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8317,
+      "step": 430
+    },
+    {
+      "epoch": 0.07113410395279282,
+      "grad_norm": 0.511726975440979,
+      "learning_rate": 0.0002,
+      "loss": 0.8528,
+      "step": 440
+    },
+    {
+      "epoch": 0.07275078813353811,
+      "grad_norm": 0.4040689170360565,
+      "learning_rate": 0.0002,
+      "loss": 0.8675,
+      "step": 450
+    },
+    {
+      "epoch": 0.07436747231428341,
+      "grad_norm": 0.5402171015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 460
+    },
+    {
+      "epoch": 0.0759841564950287,
+      "grad_norm": 0.4174517095088959,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 470
+    },
+    {
+      "epoch": 0.07760084067577398,
+      "grad_norm": 0.4306182265281677,
+      "learning_rate": 0.0002,
+      "loss": 0.7605,
+      "step": 480
+    },
+    {
+      "epoch": 0.07921752485651928,
+      "grad_norm": 0.535210132598877,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 490
+    },
+    {
+      "epoch": 0.08083420903726457,
+      "grad_norm": 0.5339109897613525,
+      "learning_rate": 0.0002,
+      "loss": 0.7825,
+      "step": 500
+    },
+    {
+      "epoch": 0.08245089321800986,
+      "grad_norm": 0.45754891633987427,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 510
+    },
+    {
+      "epoch": 0.08406757739875516,
+      "grad_norm": 0.43820783495903015,
+      "learning_rate": 0.0002,
+      "loss": 0.8144,
+      "step": 520
+    },
+    {
+      "epoch": 0.08568426157950045,
+      "grad_norm": 0.4434749186038971,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 530
+    },
+    {
+      "epoch": 0.08730094576024573,
+      "grad_norm": 0.43111467361450195,
+      "learning_rate": 0.0002,
+      "loss": 0.7857,
+      "step": 540
+    },
+    {
+      "epoch": 0.08891762994099103,
+      "grad_norm": 0.4378940165042877,
+      "learning_rate": 0.0002,
+      "loss": 0.8418,
+      "step": 550
+    },
+    {
+      "epoch": 0.09053431412173632,
+      "grad_norm": 0.4772215187549591,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 560
+    },
+    {
+      "epoch": 0.09215099830248161,
+      "grad_norm": 0.6837629079818726,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 570
+    },
+    {
+      "epoch": 0.09376768248322691,
+      "grad_norm": 0.42241212725639343,
+      "learning_rate": 0.0002,
+      "loss": 0.8607,
+      "step": 580
+    },
+    {
+      "epoch": 0.0953843666639722,
+      "grad_norm": 0.5165936350822449,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 590
+    },
+    {
+      "epoch": 0.09700105084471748,
+      "grad_norm": 0.48737478256225586,
+      "learning_rate": 0.0002,
+      "loss": 0.8664,
+      "step": 600
+    },
+    {
+      "epoch": 0.09861773502546278,
+      "grad_norm": 0.47419852018356323,
+      "learning_rate": 0.0002,
+      "loss": 0.8806,
+      "step": 610
+    },
+    {
+      "epoch": 0.10023441920620807,
+      "grad_norm": 0.4975486099720001,
+      "learning_rate": 0.0002,
+      "loss": 0.8254,
+      "step": 620
+    },
+    {
+      "epoch": 0.10185110338695336,
+      "grad_norm": 0.49123844504356384,
+      "learning_rate": 0.0002,
+      "loss": 0.8548,
+      "step": 630
+    },
+    {
+      "epoch": 0.10346778756769864,
+      "grad_norm": 0.6288952827453613,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 640
+    },
+    {
+      "epoch": 0.10508447174844394,
+      "grad_norm": 0.4277345836162567,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 650
+    },
+    {
+      "epoch": 0.10670115592918923,
+      "grad_norm": 0.4021061956882477,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 660
+    },
+    {
+      "epoch": 0.10831784010993452,
+      "grad_norm": 0.3492237329483032,
+      "learning_rate": 0.0002,
+      "loss": 0.87,
+      "step": 670
+    },
+    {
+      "epoch": 0.10993452429067982,
+      "grad_norm": 0.4341012239456177,
+      "learning_rate": 0.0002,
+      "loss": 0.8698,
+      "step": 680
+    },
+    {
+      "epoch": 0.1115512084714251,
+      "grad_norm": 0.7296304106712341,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 690
+    },
+    {
+      "epoch": 0.1131678926521704,
+      "grad_norm": 0.397494912147522,
+      "learning_rate": 0.0002,
+      "loss": 0.8433,
+      "step": 700
+    },
+    {
+      "epoch": 0.1147845768329157,
+      "grad_norm": 0.396431028842926,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 710
+    },
+    {
+      "epoch": 0.11640126101366098,
+      "grad_norm": 0.48842838406562805,
+      "learning_rate": 0.0002,
+      "loss": 0.8379,
+      "step": 720
+    },
+    {
+      "epoch": 0.11801794519440627,
+      "grad_norm": 0.46322616934776306,
+      "learning_rate": 0.0002,
+      "loss": 0.8238,
+      "step": 730
+    },
+    {
+      "epoch": 0.11963462937515157,
+      "grad_norm": 0.47990912199020386,
+      "learning_rate": 0.0002,
+      "loss": 0.8041,
+      "step": 740
+    },
+    {
+      "epoch": 0.12125131355589686,
+      "grad_norm": 0.4997142255306244,
+      "learning_rate": 0.0002,
+      "loss": 0.82,
+      "step": 750
+    },
+    {
+      "epoch": 0.12286799773664214,
+      "grad_norm": 0.4040526747703552,
+      "learning_rate": 0.0002,
+      "loss": 0.7702,
+      "step": 760
+    },
+    {
+      "epoch": 0.12448468191738744,
+      "grad_norm": 0.453095942735672,
+      "learning_rate": 0.0002,
+      "loss": 0.863,
+      "step": 770
+    },
+    {
+      "epoch": 0.12610136609813272,
+      "grad_norm": 0.4636971950531006,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 780
+    },
+    {
+      "epoch": 0.12771805027887803,
+      "grad_norm": 0.4279276132583618,
+      "learning_rate": 0.0002,
+      "loss": 0.8112,
+      "step": 790
+    },
+    {
+      "epoch": 0.12933473445962332,
+      "grad_norm": 0.46212655305862427,
+      "learning_rate": 0.0002,
+      "loss": 0.8711,
+      "step": 800
+    },
+    {
+      "epoch": 0.1309514186403686,
+      "grad_norm": 0.43127650022506714,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 810
+    },
+    {
+      "epoch": 0.1325681028211139,
+      "grad_norm": 0.4201301336288452,
+      "learning_rate": 0.0002,
+      "loss": 0.8476,
+      "step": 820
+    },
+    {
+      "epoch": 0.13418478700185918,
+      "grad_norm": 0.42583167552948,
+      "learning_rate": 0.0002,
+      "loss": 0.8078,
+      "step": 830
+    },
+    {
+      "epoch": 0.13580147118260447,
+      "grad_norm": 0.4535622000694275,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 840
+    },
+    {
+      "epoch": 0.13741815536334978,
+      "grad_norm": 0.4116036891937256,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 850
+    },
+    {
+      "epoch": 0.13903483954409507,
+      "grad_norm": 0.45997580885887146,
+      "learning_rate": 0.0002,
+      "loss": 0.8466,
+      "step": 860
+    },
+    {
+      "epoch": 0.14065152372484035,
+      "grad_norm": 0.4487837255001068,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 870
+    },
+    {
+      "epoch": 0.14226820790558564,
+      "grad_norm": 0.43650057911872864,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 880
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 0.5335358381271362,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 890
+    },
+    {
+      "epoch": 0.14550157626707622,
+      "grad_norm": 0.5989000201225281,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 900
+    },
+    {
+      "epoch": 0.14711826044782153,
+      "grad_norm": 0.517179012298584,
+      "learning_rate": 0.0002,
+      "loss": 0.8385,
+      "step": 910
+    },
+    {
+      "epoch": 0.14873494462856682,
+      "grad_norm": 0.44435232877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.8255,
+      "step": 920
+    },
+    {
+      "epoch": 0.1503516288093121,
+      "grad_norm": 0.42635923624038696,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 930
+    },
+    {
+      "epoch": 0.1519683129900574,
+      "grad_norm": 0.49603334069252014,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 940
+    },
+    {
+      "epoch": 0.15358499717080268,
+      "grad_norm": 0.40639808773994446,
+      "learning_rate": 0.0002,
+      "loss": 0.8377,
+      "step": 950
+    },
+    {
+      "epoch": 0.15520168135154797,
+      "grad_norm": 0.4850759208202362,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 960
+    },
+    {
+      "epoch": 0.15681836553229328,
+      "grad_norm": 0.4427442252635956,
+      "learning_rate": 0.0002,
+      "loss": 0.846,
+      "step": 970
+    },
+    {
+      "epoch": 0.15843504971303857,
+      "grad_norm": 0.3760930001735687,
+      "learning_rate": 0.0002,
+      "loss": 0.8705,
+      "step": 980
+    },
+    {
+      "epoch": 0.16005173389378385,
+      "grad_norm": 0.4794144332408905,
+      "learning_rate": 0.0002,
+      "loss": 0.8644,
+      "step": 990
+    },
+    {
+      "epoch": 0.16166841807452914,
+      "grad_norm": 0.45828768610954285,
+      "learning_rate": 0.0002,
+      "loss": 0.8002,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16328510225527443,
+      "grad_norm": 0.6313053369522095,
+      "learning_rate": 0.0002,
+      "loss": 0.7658,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16490178643601971,
+      "grad_norm": 0.45041006803512573,
+      "learning_rate": 0.0002,
+      "loss": 0.8047,
+      "step": 1020
+    },
+    {
+      "epoch": 0.166518470616765,
+      "grad_norm": 0.441403865814209,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16813515479751032,
+      "grad_norm": 0.8171296119689941,
+      "learning_rate": 0.0002,
+      "loss": 0.8475,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1697518389782556,
+      "grad_norm": 0.7137420773506165,
+      "learning_rate": 0.0002,
+      "loss": 0.845,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1713685231590009,
+      "grad_norm": 0.5236809849739075,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17298520733974618,
+      "grad_norm": 0.5021864175796509,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17460189152049146,
+      "grad_norm": 0.47347521781921387,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17621857570123675,
+      "grad_norm": 0.4631653428077698,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17783525988198207,
+      "grad_norm": 0.49169182777404785,
+      "learning_rate": 0.0002,
+      "loss": 0.8166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17945194406272735,
+      "grad_norm": 0.5019739270210266,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18106862824347264,
+      "grad_norm": 0.5100422501564026,
+      "learning_rate": 0.0002,
+      "loss": 0.8247,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18268531242421793,
+      "grad_norm": 0.3888324499130249,
+      "learning_rate": 0.0002,
+      "loss": 0.8142,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18430199660496321,
+      "grad_norm": 0.39765217900276184,
+      "learning_rate": 0.0002,
+      "loss": 0.8533,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1859186807857085,
+      "grad_norm": 0.47190186381340027,
+      "learning_rate": 0.0002,
+      "loss": 0.8541,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18753536496645382,
+      "grad_norm": 0.4464188814163208,
+      "learning_rate": 0.0002,
+      "loss": 0.8301,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1891520491471991,
+      "grad_norm": 0.5153930187225342,
+      "learning_rate": 0.0002,
+      "loss": 0.8341,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1907687333279444,
+      "grad_norm": 0.4779708683490753,
+      "learning_rate": 0.0002,
+      "loss": 0.8033,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19238541750868968,
+      "grad_norm": 0.4834315776824951,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19400210168943496,
+      "grad_norm": 0.402357816696167,
+      "learning_rate": 0.0002,
+      "loss": 0.7721,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19561878587018025,
+      "grad_norm": 0.45899084210395813,
+      "learning_rate": 0.0002,
+      "loss": 0.7941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19723547005092557,
+      "grad_norm": 0.5106529593467712,
+      "learning_rate": 0.0002,
+      "loss": 0.8353,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19885215423167085,
+      "grad_norm": 0.45261722803115845,
+      "learning_rate": 0.0002,
+      "loss": 0.7816,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20046883841241614,
+      "grad_norm": 0.4647127091884613,
+      "learning_rate": 0.0002,
+      "loss": 0.8068,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20208552259316143,
+      "grad_norm": 0.4849368929862976,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2037022067739067,
+      "grad_norm": 0.4518061578273773,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 1260
+    },
+    {
+      "epoch": 0.205318890954652,
+      "grad_norm": 0.49535325169563293,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2069355751353973,
+      "grad_norm": 0.4835205376148224,
+      "learning_rate": 0.0002,
+      "loss": 0.8348,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2085522593161426,
+      "grad_norm": 0.45308539271354675,
+      "learning_rate": 0.0002,
+      "loss": 0.8428,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2101689434968879,
+      "grad_norm": 0.5369905233383179,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21178562767763318,
+      "grad_norm": 0.5031622052192688,
+      "learning_rate": 0.0002,
+      "loss": 0.8676,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21340231185837846,
+      "grad_norm": 0.48010334372520447,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21501899603912375,
+      "grad_norm": 0.4905701279640198,
+      "learning_rate": 0.0002,
+      "loss": 0.806,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21663568021986904,
+      "grad_norm": 0.43531742691993713,
+      "learning_rate": 0.0002,
+      "loss": 0.7885,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21825236440061435,
+      "grad_norm": 0.44330692291259766,
+      "learning_rate": 0.0002,
+      "loss": 0.8191,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21986904858135964,
+      "grad_norm": 0.5384416580200195,
+      "learning_rate": 0.0002,
+      "loss": 0.8205,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22148573276210493,
+      "grad_norm": 0.4181833863258362,
+      "learning_rate": 0.0002,
+      "loss": 0.7726,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2231024169428502,
+      "grad_norm": 0.523833692073822,
+      "learning_rate": 0.0002,
+      "loss": 0.8311,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.5528736710548401,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2263357853043408,
+      "grad_norm": 0.43515023589134216,
+      "learning_rate": 0.0002,
+      "loss": 0.8079,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2279524694850861,
+      "grad_norm": 0.48809877038002014,
+      "learning_rate": 0.0002,
+      "loss": 0.8403,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2295691536658314,
+      "grad_norm": 0.43591251969337463,
+      "learning_rate": 0.0002,
+      "loss": 0.8165,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23118583784657668,
+      "grad_norm": 0.44625312089920044,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23280252202732196,
+      "grad_norm": 0.4390665292739868,
+      "learning_rate": 0.0002,
+      "loss": 0.8134,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23441920620806725,
+      "grad_norm": 0.48496049642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.8465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23603589038881254,
+      "grad_norm": 0.45919957756996155,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23765257456955785,
+      "grad_norm": 0.5471845865249634,
+      "learning_rate": 0.0002,
+      "loss": 0.8659,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23926925875030314,
+      "grad_norm": 0.47269317507743835,
+      "learning_rate": 0.0002,
+      "loss": 0.8164,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24088594293104842,
+      "grad_norm": 0.4930245578289032,
+      "learning_rate": 0.0002,
+      "loss": 0.854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2425026271117937,
+      "grad_norm": 0.5605630278587341,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 1500
+    },
+    {
+      "epoch": 0.244119311292539,
+      "grad_norm": 0.4435870945453644,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24573599547328429,
+      "grad_norm": 0.4941999912261963,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24735267965402957,
+      "grad_norm": 0.5100624561309814,
+      "learning_rate": 0.0002,
+      "loss": 0.8427,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2489693638347749,
+      "grad_norm": 0.4638267457485199,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25058604801552015,
+      "grad_norm": 0.5071570873260498,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25220273219626543,
+      "grad_norm": 0.4291319251060486,
+      "learning_rate": 0.0002,
+      "loss": 0.7724,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2538194163770108,
+      "grad_norm": 0.5388049483299255,
+      "learning_rate": 0.0002,
+      "loss": 0.7984,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25543610055775606,
+      "grad_norm": 0.5083683729171753,
+      "learning_rate": 0.0002,
+      "loss": 0.8176,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25705278473850135,
+      "grad_norm": 0.4824463725090027,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25866946891924664,
+      "grad_norm": 0.41177722811698914,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2602861530999919,
+      "grad_norm": 0.5656219124794006,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2619028372807372,
+      "grad_norm": 0.41063204407691956,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2635195214614825,
+      "grad_norm": 0.4897061288356781,
+      "learning_rate": 0.0002,
+      "loss": 0.7998,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2651362056422278,
+      "grad_norm": 0.4454376697540283,
+      "learning_rate": 0.0002,
+      "loss": 0.8198,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26675288982297307,
+      "grad_norm": 0.4355238378047943,
+      "learning_rate": 0.0002,
+      "loss": 0.8684,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26836957400371836,
+      "grad_norm": 0.458310067653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26998625818446365,
+      "grad_norm": 0.4752083718776703,
+      "learning_rate": 0.0002,
+      "loss": 0.7935,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27160294236520893,
+      "grad_norm": 0.4666106402873993,
+      "learning_rate": 0.0002,
+      "loss": 0.8267,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2732196265459543,
+      "grad_norm": 0.4213818609714508,
+      "learning_rate": 0.0002,
+      "loss": 0.8252,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27483631072669956,
+      "grad_norm": 0.5768913626670837,
+      "learning_rate": 0.0002,
+      "loss": 0.8559,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27645299490744485,
+      "grad_norm": 0.4209914803504944,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27806967908819014,
+      "grad_norm": 0.501909613609314,
+      "learning_rate": 0.0002,
+      "loss": 0.8167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2796863632689354,
+      "grad_norm": 0.5266261100769043,
+      "learning_rate": 0.0002,
+      "loss": 0.7832,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2813030474496807,
+      "grad_norm": 0.43806859850883484,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 1740
+    },
+    {
+      "epoch": 0.282919731630426,
+      "grad_norm": 0.46048814058303833,
+      "learning_rate": 0.0002,
+      "loss": 0.8157,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2845364158111713,
+      "grad_norm": 0.44972819089889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8596,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28615309999191657,
+      "grad_norm": 0.5114831328392029,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.47931742668151855,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28938646835340714,
+      "grad_norm": 0.5092599987983704,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29100315253415243,
+      "grad_norm": 0.37581443786621094,
+      "learning_rate": 0.0002,
+      "loss": 0.8506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2926198367148977,
+      "grad_norm": 0.47097381949424744,
+      "learning_rate": 0.0002,
+      "loss": 0.7932,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29423652089564306,
+      "grad_norm": 0.48300236463546753,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29585320507638835,
+      "grad_norm": 0.5600419640541077,
+      "learning_rate": 0.0002,
+      "loss": 0.8391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29746988925713364,
+      "grad_norm": 0.48555272817611694,
+      "learning_rate": 0.0002,
+      "loss": 0.8507,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2990865734378789,
+      "grad_norm": 0.3752668499946594,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3007032576186242,
+      "grad_norm": 0.5328747034072876,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3023199417993695,
+      "grad_norm": 0.48716455698013306,
+      "learning_rate": 0.0002,
+      "loss": 0.8426,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3039366259801148,
+      "grad_norm": 0.5011493563652039,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30555331016086007,
+      "grad_norm": 0.46461427211761475,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30716999434160536,
+      "grad_norm": 0.36630210280418396,
+      "learning_rate": 0.0002,
+      "loss": 0.8478,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30878667852235064,
+      "grad_norm": 0.4217296242713928,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31040336270309593,
+      "grad_norm": 0.4394875466823578,
+      "learning_rate": 0.0002,
+      "loss": 0.8128,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3120200468838412,
+      "grad_norm": 0.6587965488433838,
+      "learning_rate": 0.0002,
+      "loss": 0.8471,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31363673106458656,
+      "grad_norm": 0.5469298958778381,
+      "learning_rate": 0.0002,
+      "loss": 0.8565,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31525341524533185,
+      "grad_norm": 0.4371595084667206,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31687009942607713,
+      "grad_norm": 0.4809541404247284,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3184867836068224,
+      "grad_norm": 0.6061086654663086,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3201034677875677,
+      "grad_norm": 0.5342657566070557,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 1980
+    },
+    {
+      "epoch": 0.321720151968313,
+      "grad_norm": 0.5057743787765503,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3233368361490583,
+      "grad_norm": 0.528626024723053,
+      "learning_rate": 0.0002,
+      "loss": 0.7774,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32495352032980357,
+      "grad_norm": 0.46742770075798035,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32657020451054886,
+      "grad_norm": 0.515101432800293,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32818688869129414,
+      "grad_norm": 0.41941216588020325,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32980357287203943,
+      "grad_norm": 0.49902522563934326,
+      "learning_rate": 0.0002,
+      "loss": 0.7637,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3314202570527847,
+      "grad_norm": 0.4120897650718689,
+      "learning_rate": 0.0002,
+      "loss": 0.7822,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33303694123353,
+      "grad_norm": 0.45352041721343994,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33465362541427535,
+      "grad_norm": 0.523199737071991,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33627030959502063,
+      "grad_norm": 0.4390358626842499,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3378869937757659,
+      "grad_norm": 0.6752901077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3395036779565112,
+      "grad_norm": 0.547821044921875,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3411203621372565,
+      "grad_norm": 0.5161308646202087,
+      "learning_rate": 0.0002,
+      "loss": 0.8561,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3427370463180018,
+      "grad_norm": 0.4565401077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34435373049874707,
+      "grad_norm": 0.4666115939617157,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34597041467949236,
+      "grad_norm": 0.4090428352355957,
+      "learning_rate": 0.0002,
+      "loss": 0.8189,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34758709886023764,
+      "grad_norm": 0.510845422744751,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34920378304098293,
+      "grad_norm": 0.42861923575401306,
+      "learning_rate": 0.0002,
+      "loss": 0.8398,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3508204672217282,
+      "grad_norm": 0.4476332664489746,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3524371514024735,
+      "grad_norm": 0.6065791249275208,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35405383558321885,
+      "grad_norm": 0.42335066199302673,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35567051976396413,
+      "grad_norm": 0.5094629526138306,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3572872039447094,
+      "grad_norm": 0.5476373434066772,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3589038881254547,
+      "grad_norm": 0.3911719024181366,
+      "learning_rate": 0.0002,
+      "loss": 0.814,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3605205723062,
+      "grad_norm": 0.6599636077880859,
+      "learning_rate": 0.0002,
+      "loss": 0.8599,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3621372564869453,
+      "grad_norm": 0.40381914377212524,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36375394066769057,
+      "grad_norm": 0.4433908462524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36537062484843585,
+      "grad_norm": 0.578326940536499,
+      "learning_rate": 0.0002,
+      "loss": 0.8503,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36698730902918114,
+      "grad_norm": 0.5734784007072449,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36860399320992643,
+      "grad_norm": 0.45555487275123596,
+      "learning_rate": 0.0002,
+      "loss": 0.8193,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3702206773906717,
+      "grad_norm": 0.5666276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.7929,
+      "step": 2290
+    },
+    {
+      "epoch": 0.371837361571417,
+      "grad_norm": 0.5461117625236511,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3734540457521623,
+      "grad_norm": 0.6318911910057068,
+      "learning_rate": 0.0002,
+      "loss": 0.8204,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37507072993290763,
+      "grad_norm": 0.493263304233551,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3766874141136529,
+      "grad_norm": 0.5888760089874268,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3783040982943982,
+      "grad_norm": 0.48671841621398926,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3799207824751435,
+      "grad_norm": 0.4385145306587219,
+      "learning_rate": 0.0002,
+      "loss": 0.8367,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3815374666558888,
+      "grad_norm": 0.5523318648338318,
+      "learning_rate": 0.0002,
+      "loss": 0.812,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38315415083663407,
+      "grad_norm": 0.7308220267295837,
+      "learning_rate": 0.0002,
+      "loss": 0.8351,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38477083501737935,
+      "grad_norm": 0.554214358329773,
+      "learning_rate": 0.0002,
+      "loss": 0.859,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38638751919812464,
+      "grad_norm": 0.5425800085067749,
+      "learning_rate": 0.0002,
+      "loss": 0.8146,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3880042033788699,
+      "grad_norm": 0.48811158537864685,
+      "learning_rate": 0.0002,
+      "loss": 0.8282,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3896208875596152,
+      "grad_norm": 0.49212366342544556,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3912375717403605,
+      "grad_norm": 0.5222218632698059,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3928542559211058,
+      "grad_norm": 0.4699819087982178,
+      "learning_rate": 0.0002,
+      "loss": 0.8182,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39447094010185113,
+      "grad_norm": 0.46153587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7919,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3960876242825964,
+      "grad_norm": 0.4150611162185669,
+      "learning_rate": 0.0002,
+      "loss": 0.8111,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3977043084633417,
+      "grad_norm": 0.5799614787101746,
+      "learning_rate": 0.0002,
+      "loss": 0.8589,
+      "step": 2460
+    },
+    {
+      "epoch": 0.399320992644087,
+      "grad_norm": 0.56536865234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4009376768248323,
+      "grad_norm": 0.5451247096061707,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40255436100557757,
+      "grad_norm": 0.5914521217346191,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40417104518632285,
+      "grad_norm": 0.4428117275238037,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40578772936706814,
+      "grad_norm": 0.48580947518348694,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4074044135478134,
+      "grad_norm": 0.436734676361084,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4090210977285587,
+      "grad_norm": 0.5752223134040833,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 2530
+    },
+    {
+      "epoch": 0.410637781909304,
+      "grad_norm": 0.4271308183670044,
+      "learning_rate": 0.0002,
+      "loss": 0.8181,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4122544660900493,
+      "grad_norm": 0.46294718980789185,
+      "learning_rate": 0.0002,
+      "loss": 0.8058,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4138711502707946,
+      "grad_norm": 0.49407583475112915,
+      "learning_rate": 0.0002,
+      "loss": 0.8473,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4154878344515399,
+      "grad_norm": 0.4729035496711731,
+      "learning_rate": 0.0002,
+      "loss": 0.7881,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4171045186322852,
+      "grad_norm": 0.4129747152328491,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4187212028130305,
+      "grad_norm": 0.5684236288070679,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4203378869937758,
+      "grad_norm": 0.4862157106399536,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42195457117452106,
+      "grad_norm": 0.46567976474761963,
+      "learning_rate": 0.0002,
+      "loss": 0.7582,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42357125535526635,
+      "grad_norm": 0.5710650682449341,
+      "learning_rate": 0.0002,
+      "loss": 0.7755,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42518793953601164,
+      "grad_norm": 0.5660041570663452,
+      "learning_rate": 0.0002,
+      "loss": 0.8573,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4268046237167569,
+      "grad_norm": 0.47944375872612,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4284213078975022,
+      "grad_norm": 0.537223756313324,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4300379920782475,
+      "grad_norm": 0.41669997572898865,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 0.44727686047554016,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4332713604397381,
+      "grad_norm": 0.5600888729095459,
+      "learning_rate": 0.0002,
+      "loss": 0.8241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4348880446204834,
+      "grad_norm": 0.39820605516433716,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4365047288012287,
+      "grad_norm": 0.5637655854225159,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 2700
+    },
+    {
+      "epoch": 0.438121412981974,
+      "grad_norm": 0.6363666653633118,
+      "learning_rate": 0.0002,
+      "loss": 0.855,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4397380971627193,
+      "grad_norm": 0.5656129121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.8468,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44135478134346456,
+      "grad_norm": 0.5600156188011169,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44297146552420985,
+      "grad_norm": 0.5506579875946045,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44458814970495514,
+      "grad_norm": 0.49878305196762085,
+      "learning_rate": 0.0002,
+      "loss": 0.7725,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4462048338857004,
+      "grad_norm": 0.4569213092327118,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4478215180664457,
+      "grad_norm": 0.6056680083274841,
+      "learning_rate": 0.0002,
+      "loss": 0.8028,
+      "step": 2770
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.44474557042121887,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4510548864279363,
+      "grad_norm": 0.46055394411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.801,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4526715706086816,
+      "grad_norm": 0.4904133379459381,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45428825478942686,
+      "grad_norm": 0.5647031664848328,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4559049389701722,
+      "grad_norm": 0.5759473443031311,
+      "learning_rate": 0.0002,
+      "loss": 0.8622,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4575216231509175,
+      "grad_norm": 0.5161895751953125,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4591383073316628,
+      "grad_norm": 0.4248254597187042,
+      "learning_rate": 0.0002,
+      "loss": 0.8045,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46075499151240806,
+      "grad_norm": 0.45395001769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.7838,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46237167569315335,
+      "grad_norm": 0.5358697772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.8208,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46398835987389864,
+      "grad_norm": 0.5379165410995483,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4656050440546439,
+      "grad_norm": 0.4601989686489105,
+      "learning_rate": 0.0002,
+      "loss": 0.7403,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4672217282353892,
+      "grad_norm": 0.671115517616272,
+      "learning_rate": 0.0002,
+      "loss": 0.8523,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4688384124161345,
+      "grad_norm": 0.4425133168697357,
+      "learning_rate": 0.0002,
+      "loss": 0.8262,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4704550965968798,
+      "grad_norm": 0.5446155071258545,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47207178077762507,
+      "grad_norm": 0.603306233882904,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47368846495837036,
+      "grad_norm": 0.5377997159957886,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4753051491391157,
+      "grad_norm": 0.4931027591228485,
+      "learning_rate": 0.0002,
+      "loss": 0.8075,
+      "step": 2940
+    },
+    {
+      "epoch": 0.476921833319861,
+      "grad_norm": 0.4711960256099701,
+      "learning_rate": 0.0002,
+      "loss": 0.8004,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4785385175006063,
+      "grad_norm": 0.5020492672920227,
+      "learning_rate": 0.0002,
+      "loss": 0.8121,
+      "step": 2960
+    },
+    {
+      "epoch": 0.48015520168135156,
+      "grad_norm": 0.5428946614265442,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48177188586209685,
+      "grad_norm": 0.5294089317321777,
+      "learning_rate": 0.0002,
+      "loss": 0.7849,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48338857004284214,
+      "grad_norm": 0.648289144039154,
+      "learning_rate": 0.0002,
+      "loss": 0.8553,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4850052542235874,
+      "grad_norm": 0.47916680574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.7874,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4866219384043327,
+      "grad_norm": 0.43849772214889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8087,
+      "step": 3010
+    },
+    {
+      "epoch": 0.488238622585078,
+      "grad_norm": 0.47007861733436584,
+      "learning_rate": 0.0002,
+      "loss": 0.7662,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4898553067658233,
+      "grad_norm": 0.6314331293106079,
+      "learning_rate": 0.0002,
+      "loss": 0.757,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49147199094656857,
+      "grad_norm": 0.49211493134498596,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49308867512731386,
+      "grad_norm": 0.4537973403930664,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49470535930805914,
+      "grad_norm": 0.47326919436454773,
+      "learning_rate": 0.0002,
+      "loss": 0.8095,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4963220434888045,
+      "grad_norm": 0.525874137878418,
+      "learning_rate": 0.0002,
+      "loss": 0.8447,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4979387276695498,
+      "grad_norm": 0.6361091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49955541185029506,
+      "grad_norm": 0.5850642919540405,
+      "learning_rate": 0.0002,
+      "loss": 0.821,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5011720960310403,
+      "grad_norm": 0.47299543023109436,
+      "learning_rate": 0.0002,
+      "loss": 0.8279,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5027887802117856,
+      "grad_norm": 0.473099946975708,
+      "learning_rate": 0.0002,
+      "loss": 0.8681,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5044054643925309,
+      "grad_norm": 0.48186397552490234,
+      "learning_rate": 0.0002,
+      "loss": 0.8223,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5060221485732762,
+      "grad_norm": 0.5015401840209961,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5076388327540216,
+      "grad_norm": 0.5617750287055969,
+      "learning_rate": 0.0002,
+      "loss": 0.7692,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5092555169347668,
+      "grad_norm": 0.5169327259063721,
+      "learning_rate": 0.0002,
+      "loss": 0.8708,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5108722011155121,
+      "grad_norm": 0.545657753944397,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5124888852962574,
+      "grad_norm": 0.512864351272583,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5141055694770027,
+      "grad_norm": 0.4113546311855316,
+      "learning_rate": 0.0002,
+      "loss": 0.7794,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5157222536577479,
+      "grad_norm": 0.44532445073127747,
+      "learning_rate": 0.0002,
+      "loss": 0.8206,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5173389378384933,
+      "grad_norm": 0.5623497366905212,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5189556220192385,
+      "grad_norm": 0.5084741115570068,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5205723061999838,
+      "grad_norm": 0.5305403470993042,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5221889903807291,
+      "grad_norm": 0.4708254337310791,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5238056745614744,
+      "grad_norm": 0.43827131390571594,
+      "learning_rate": 0.0002,
+      "loss": 0.7639,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5254223587422197,
+      "grad_norm": 0.5630002617835999,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 3250
+    },
+    {
+      "epoch": 0.527039042922965,
+      "grad_norm": 0.5010961890220642,
+      "learning_rate": 0.0002,
+      "loss": 0.7522,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5286557271037103,
+      "grad_norm": 0.6303122043609619,
+      "learning_rate": 0.0002,
+      "loss": 0.8374,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5302724112844556,
+      "grad_norm": 0.5107331275939941,
+      "learning_rate": 0.0002,
+      "loss": 0.7727,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5318890954652009,
+      "grad_norm": 0.5700443387031555,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5335057796459461,
+      "grad_norm": 0.46296367049217224,
+      "learning_rate": 0.0002,
+      "loss": 0.7776,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5351224638266915,
+      "grad_norm": 0.531568706035614,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5367391480074367,
+      "grad_norm": 0.4686741530895233,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5383558321881821,
+      "grad_norm": 0.5404331088066101,
+      "learning_rate": 0.0002,
+      "loss": 0.8104,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5399725163689273,
+      "grad_norm": 0.6368790864944458,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5415892005496726,
+      "grad_norm": 0.42300888895988464,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5432058847304179,
+      "grad_norm": 0.5362542867660522,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5448225689111632,
+      "grad_norm": 0.497128963470459,
+      "learning_rate": 0.0002,
+      "loss": 0.858,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5464392530919085,
+      "grad_norm": 0.5006386041641235,
+      "learning_rate": 0.0002,
+      "loss": 0.8519,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5480559372726538,
+      "grad_norm": 0.44136837124824524,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5496726214533991,
+      "grad_norm": 0.5897833108901978,
+      "learning_rate": 0.0002,
+      "loss": 0.773,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5512893056341444,
+      "grad_norm": 0.641075611114502,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5529059898148897,
+      "grad_norm": 0.7251322269439697,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5545226739956349,
+      "grad_norm": 0.47411349415779114,
+      "learning_rate": 0.0002,
+      "loss": 0.7626,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5561393581763803,
+      "grad_norm": 0.4994310438632965,
+      "learning_rate": 0.0002,
+      "loss": 0.8196,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5577560423571255,
+      "grad_norm": 0.5814438462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5593727265378708,
+      "grad_norm": 0.6278898119926453,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5609894107186161,
+      "grad_norm": 0.46208274364471436,
+      "learning_rate": 0.0002,
+      "loss": 0.813,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5626060948993614,
+      "grad_norm": 0.5718930959701538,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5642227790801067,
+      "grad_norm": 0.48178744316101074,
+      "learning_rate": 0.0002,
+      "loss": 0.8152,
+      "step": 3490
+    },
+    {
+      "epoch": 0.565839463260852,
+      "grad_norm": 0.47336965799331665,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5674561474415973,
+      "grad_norm": 0.43442684412002563,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5690728316223426,
+      "grad_norm": 0.6463358998298645,
+      "learning_rate": 0.0002,
+      "loss": 0.7564,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5706895158030879,
+      "grad_norm": 0.5286486744880676,
+      "learning_rate": 0.0002,
+      "loss": 0.836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5723061999838331,
+      "grad_norm": 0.5405499935150146,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5739228841645785,
+      "grad_norm": 0.6654391884803772,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.5081980228424072,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5771562525260691,
+      "grad_norm": 0.48978179693222046,
+      "learning_rate": 0.0002,
+      "loss": 0.7753,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5787729367068143,
+      "grad_norm": 0.5840612053871155,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5803896208875596,
+      "grad_norm": 0.5235261917114258,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5820063050683049,
+      "grad_norm": 0.5672075748443604,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5836229892490502,
+      "grad_norm": 0.5613429546356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5852396734297954,
+      "grad_norm": 0.4032273590564728,
+      "learning_rate": 0.0002,
+      "loss": 0.8274,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5868563576105408,
+      "grad_norm": 0.49559324979782104,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5884730417912861,
+      "grad_norm": 0.6895697712898254,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5900897259720314,
+      "grad_norm": 0.4750136435031891,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5917064101527767,
+      "grad_norm": 0.5176819562911987,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5933230943335219,
+      "grad_norm": 0.5817760229110718,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5949397785142673,
+      "grad_norm": 0.6064626574516296,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5965564626950125,
+      "grad_norm": 0.6728700995445251,
+      "learning_rate": 0.0002,
+      "loss": 0.8422,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5981731468757578,
+      "grad_norm": 0.609305202960968,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5997898310565031,
+      "grad_norm": 0.4615488350391388,
+      "learning_rate": 0.0002,
+      "loss": 0.8048,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6014065152372484,
+      "grad_norm": 2.0531179904937744,
+      "learning_rate": 0.0002,
+      "loss": 0.8214,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6030231994179936,
+      "grad_norm": 0.5091132521629333,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 3730
+    },
+    {
+      "epoch": 0.604639883598739,
+      "grad_norm": 0.5951124429702759,
+      "learning_rate": 0.0002,
+      "loss": 0.7833,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6062565677794842,
+      "grad_norm": 0.5870208144187927,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6078732519602296,
+      "grad_norm": 0.6254619359970093,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6094899361409749,
+      "grad_norm": 0.5577626824378967,
+      "learning_rate": 0.0002,
+      "loss": 0.7868,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6111066203217201,
+      "grad_norm": 0.5004405379295349,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6127233045024655,
+      "grad_norm": 0.5527383685112,
+      "learning_rate": 0.0002,
+      "loss": 0.8092,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6143399886832107,
+      "grad_norm": 0.49116113781929016,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6159566728639561,
+      "grad_norm": 0.5299299359321594,
+      "learning_rate": 0.0002,
+      "loss": 0.8352,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6175733570447013,
+      "grad_norm": 0.464897483587265,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6191900412254466,
+      "grad_norm": 0.6505740880966187,
+      "learning_rate": 0.0002,
+      "loss": 0.7923,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6208067254061919,
+      "grad_norm": 0.5512559413909912,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6224234095869372,
+      "grad_norm": 0.49427518248558044,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6240400937676824,
+      "grad_norm": 0.3839147090911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6256567779484278,
+      "grad_norm": 0.5760218501091003,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6272734621291731,
+      "grad_norm": 0.7226507067680359,
+      "learning_rate": 0.0002,
+      "loss": 0.7836,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6288901463099184,
+      "grad_norm": 0.676781415939331,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6305068304906637,
+      "grad_norm": 0.4284018278121948,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6321235146714089,
+      "grad_norm": 0.5060628056526184,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6337401988521543,
+      "grad_norm": 0.5524522066116333,
+      "learning_rate": 0.0002,
+      "loss": 0.8089,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6353568830328995,
+      "grad_norm": 0.6099881529808044,
+      "learning_rate": 0.0002,
+      "loss": 0.8276,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6369735672136448,
+      "grad_norm": 0.43155938386917114,
+      "learning_rate": 0.0002,
+      "loss": 0.809,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6385902513943901,
+      "grad_norm": 0.6427084803581238,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6402069355751354,
+      "grad_norm": 0.541220486164093,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6418236197558806,
+      "grad_norm": 0.5414294600486755,
+      "learning_rate": 0.0002,
+      "loss": 0.8539,
+      "step": 3970
+    },
+    {
+      "epoch": 0.643440303936626,
+      "grad_norm": 0.46344003081321716,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6450569881173712,
+      "grad_norm": 0.45209285616874695,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6466736722981166,
+      "grad_norm": 0.5417284369468689,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6482903564788619,
+      "grad_norm": 0.7995685935020447,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6499070406596071,
+      "grad_norm": 0.6384002566337585,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6515237248403525,
+      "grad_norm": 0.4472815692424774,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6531404090210977,
+      "grad_norm": 0.6834294199943542,
+      "learning_rate": 0.0002,
+      "loss": 0.8071,
+      "step": 4040
+    },
+    {
+      "epoch": 0.654757093201843,
+      "grad_norm": 0.4612339735031128,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6563737773825883,
+      "grad_norm": 0.9266576170921326,
+      "learning_rate": 0.0002,
+      "loss": 0.8141,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6579904615633336,
+      "grad_norm": 0.4470861852169037,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6596071457440789,
+      "grad_norm": 0.45544925332069397,
+      "learning_rate": 0.0002,
+      "loss": 0.8293,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6612238299248242,
+      "grad_norm": 0.6144481301307678,
+      "learning_rate": 0.0002,
+      "loss": 0.8455,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6628405141055694,
+      "grad_norm": 0.5936288237571716,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6644571982863148,
+      "grad_norm": 0.4822963774204254,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 4110
+    },
+    {
+      "epoch": 0.66607388246706,
+      "grad_norm": 0.48432496190071106,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6676905666478054,
+      "grad_norm": 0.4901607930660248,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6693072508285507,
+      "grad_norm": 0.5018393397331238,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6709239350092959,
+      "grad_norm": 0.6946378946304321,
+      "learning_rate": 0.0002,
+      "loss": 0.8065,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6725406191900413,
+      "grad_norm": 0.5997390747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.6738849878311157,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6757739875515318,
+      "grad_norm": 0.6110581159591675,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6773906717322771,
+      "grad_norm": 0.5703322291374207,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6790073559130224,
+      "grad_norm": 0.4686066210269928,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6806240400937676,
+      "grad_norm": 0.6394643783569336,
+      "learning_rate": 0.0002,
+      "loss": 0.8441,
+      "step": 4210
+    },
+    {
+      "epoch": 0.682240724274513,
+      "grad_norm": 0.5454841256141663,
+      "learning_rate": 0.0002,
+      "loss": 0.8011,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6838574084552582,
+      "grad_norm": 0.4859732985496521,
+      "learning_rate": 0.0002,
+      "loss": 0.8307,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6854740926360036,
+      "grad_norm": 0.5544065833091736,
+      "learning_rate": 0.0002,
+      "loss": 0.8161,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6870907768167488,
+      "grad_norm": 0.4902505576610565,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6887074609974941,
+      "grad_norm": 0.4768051505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7977,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6903241451782395,
+      "grad_norm": 0.49982190132141113,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6919408293589847,
+      "grad_norm": 0.6351838111877441,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 4280
+    },
+    {
+      "epoch": 0.69355751353973,
+      "grad_norm": 0.5647561550140381,
+      "learning_rate": 0.0002,
+      "loss": 0.7664,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6951741977204753,
+      "grad_norm": 0.5340486764907837,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6967908819012206,
+      "grad_norm": 0.5649092793464661,
+      "learning_rate": 0.0002,
+      "loss": 0.8526,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6984075660819659,
+      "grad_norm": 0.6183916926383972,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7000242502627112,
+      "grad_norm": 0.6154509782791138,
+      "learning_rate": 0.0002,
+      "loss": 0.792,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7016409344434564,
+      "grad_norm": 0.5156264305114746,
+      "learning_rate": 0.0002,
+      "loss": 0.8397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7032576186242018,
+      "grad_norm": 0.562171459197998,
+      "learning_rate": 0.0002,
+      "loss": 0.8512,
+      "step": 4350
+    },
+    {
+      "epoch": 0.704874302804947,
+      "grad_norm": 0.4949502646923065,
+      "learning_rate": 0.0002,
+      "loss": 0.7882,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7064909869856923,
+      "grad_norm": 0.5171684622764587,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7081076711664377,
+      "grad_norm": 0.6198443174362183,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7097243553471829,
+      "grad_norm": 0.5802276134490967,
+      "learning_rate": 0.0002,
+      "loss": 0.7606,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7113410395279283,
+      "grad_norm": 0.41096967458724976,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7129577237086735,
+      "grad_norm": 0.4397392272949219,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7145744078894188,
+      "grad_norm": 0.45228442549705505,
+      "learning_rate": 0.0002,
+      "loss": 0.7651,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7161910920701641,
+      "grad_norm": 0.4839673936367035,
+      "learning_rate": 0.0002,
+      "loss": 0.7938,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7178077762509094,
+      "grad_norm": 0.6140755414962769,
+      "learning_rate": 0.0002,
+      "loss": 0.8362,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 0.6841378808021545,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7210411446124,
+      "grad_norm": 0.6664239168167114,
+      "learning_rate": 0.0002,
+      "loss": 0.8177,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7226578287931452,
+      "grad_norm": 0.47552719712257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7983,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7242745129738906,
+      "grad_norm": 0.6649776101112366,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7258911971546358,
+      "grad_norm": 0.5159541964530945,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7275078813353811,
+      "grad_norm": 0.6693112850189209,
+      "learning_rate": 0.0002,
+      "loss": 0.7786,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7291245655161265,
+      "grad_norm": 0.48870977759361267,
+      "learning_rate": 0.0002,
+      "loss": 0.8655,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7307412496968717,
+      "grad_norm": 0.4857887923717499,
+      "learning_rate": 0.0002,
+      "loss": 0.7337,
+      "step": 4520
+    },
+    {
+      "epoch": 0.732357933877617,
+      "grad_norm": 0.5515662431716919,
+      "learning_rate": 0.0002,
+      "loss": 0.8026,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7339746180583623,
+      "grad_norm": 0.6292222738265991,
+      "learning_rate": 0.0002,
+      "loss": 0.8031,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7355913022391076,
+      "grad_norm": 0.48265689611434937,
+      "learning_rate": 0.0002,
+      "loss": 0.7749,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7372079864198529,
+      "grad_norm": 0.8044266104698181,
+      "learning_rate": 0.0002,
+      "loss": 0.8499,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7388246706005982,
+      "grad_norm": 0.6111769676208496,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7404413547813434,
+      "grad_norm": 0.5229553580284119,
+      "learning_rate": 0.0002,
+      "loss": 0.7291,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7420580389620888,
+      "grad_norm": 0.6054152250289917,
+      "learning_rate": 0.0002,
+      "loss": 0.8038,
+      "step": 4590
+    },
+    {
+      "epoch": 0.743674723142834,
+      "grad_norm": 0.5574966669082642,
+      "learning_rate": 0.0002,
+      "loss": 0.8169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7452914073235793,
+      "grad_norm": 0.5395817160606384,
+      "learning_rate": 0.0002,
+      "loss": 0.8439,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7469080915043246,
+      "grad_norm": 0.7116472721099854,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7485247756850699,
+      "grad_norm": 0.5618700981140137,
+      "learning_rate": 0.0002,
+      "loss": 0.7743,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7501414598658153,
+      "grad_norm": 0.5802770853042603,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7517581440465605,
+      "grad_norm": 0.5690428018569946,
+      "learning_rate": 0.0002,
+      "loss": 0.7924,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7533748282273058,
+      "grad_norm": 0.4813360273838043,
+      "learning_rate": 0.0002,
+      "loss": 0.8017,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7549915124080511,
+      "grad_norm": 0.5434042811393738,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7566081965887964,
+      "grad_norm": 0.5502099990844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7582248807695416,
+      "grad_norm": 0.6020621061325073,
+      "learning_rate": 0.0002,
+      "loss": 0.8598,
+      "step": 4690
+    },
+    {
+      "epoch": 0.759841564950287,
+      "grad_norm": 0.4922301471233368,
+      "learning_rate": 0.0002,
+      "loss": 0.7937,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7614582491310322,
+      "grad_norm": 0.6492828726768494,
+      "learning_rate": 0.0002,
+      "loss": 0.788,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7630749333117776,
+      "grad_norm": 0.4865580201148987,
+      "learning_rate": 0.0002,
+      "loss": 0.8313,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7646916174925228,
+      "grad_norm": 0.5971422791481018,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7663083016732681,
+      "grad_norm": 0.6832674145698547,
+      "learning_rate": 0.0002,
+      "loss": 0.8298,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7679249858540134,
+      "grad_norm": 0.500908613204956,
+      "learning_rate": 0.0002,
+      "loss": 0.8156,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7695416700347587,
+      "grad_norm": 0.6112465858459473,
+      "learning_rate": 0.0002,
+      "loss": 0.8383,
+      "step": 4760
+    },
+    {
+      "epoch": 0.771158354215504,
+      "grad_norm": 0.5753506422042847,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7727750383962493,
+      "grad_norm": 0.6529405117034912,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7743917225769946,
+      "grad_norm": 0.5916843414306641,
+      "learning_rate": 0.0002,
+      "loss": 0.8171,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7760084067577399,
+      "grad_norm": 0.4821224510669708,
+      "learning_rate": 0.0002,
+      "loss": 0.83,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7776250909384852,
+      "grad_norm": 0.5532580018043518,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7792417751192304,
+      "grad_norm": 0.4604877233505249,
+      "learning_rate": 0.0002,
+      "loss": 0.7363,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7808584592999758,
+      "grad_norm": 0.5009613037109375,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 4830
+    },
+    {
+      "epoch": 0.782475143480721,
+      "grad_norm": 0.6448560357093811,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7840918276614663,
+      "grad_norm": 0.44327953457832336,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7857085118422116,
+      "grad_norm": 0.5355411171913147,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7873251960229569,
+      "grad_norm": 0.5635677576065063,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7889418802037023,
+      "grad_norm": 0.5417491793632507,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7905585643844475,
+      "grad_norm": 0.4567430913448334,
+      "learning_rate": 0.0002,
+      "loss": 0.7819,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7921752485651928,
+      "grad_norm": 0.44651296734809875,
+      "learning_rate": 0.0002,
+      "loss": 0.8454,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7937919327459381,
+      "grad_norm": 0.5741217136383057,
+      "learning_rate": 0.0002,
+      "loss": 0.7959,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7954086169266834,
+      "grad_norm": 0.6605045199394226,
+      "learning_rate": 0.0002,
+      "loss": 0.8093,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7970253011074286,
+      "grad_norm": 0.5126531720161438,
+      "learning_rate": 0.0002,
+      "loss": 0.77,
+      "step": 4930
+    },
+    {
+      "epoch": 0.798641985288174,
+      "grad_norm": 0.513648271560669,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8002586694689192,
+      "grad_norm": 0.5350404381752014,
+      "learning_rate": 0.0002,
+      "loss": 0.8314,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8018753536496646,
+      "grad_norm": 0.5731674432754517,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8034920378304098,
+      "grad_norm": 0.5974258184432983,
+      "learning_rate": 0.0002,
+      "loss": 0.8572,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8051087220111551,
+      "grad_norm": 0.8774799704551697,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8067254061919004,
+      "grad_norm": 0.5994430184364319,
+      "learning_rate": 0.0002,
+      "loss": 0.7899,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8083420903726457,
+      "grad_norm": 0.4894903004169464,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5000
+    },
+    {
+      "epoch": 0.809958774553391,
+      "grad_norm": 0.5218459367752075,
+      "learning_rate": 0.0002,
+      "loss": 0.78,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8115754587341363,
+      "grad_norm": 0.5232468843460083,
+      "learning_rate": 0.0002,
+      "loss": 0.817,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8131921429148816,
+      "grad_norm": 0.44358372688293457,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8148088270956269,
+      "grad_norm": 0.6202037334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8164255112763722,
+      "grad_norm": 0.7721474170684814,
+      "learning_rate": 0.0002,
+      "loss": 0.7351,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8180421954571174,
+      "grad_norm": 0.5568501353263855,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8196588796378628,
+      "grad_norm": 0.49148809909820557,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 5070
+    },
+    {
+      "epoch": 0.821275563818608,
+      "grad_norm": 0.4956012964248657,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8228922479993533,
+      "grad_norm": 0.6078833937644958,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8245089321800986,
+      "grad_norm": 0.46906954050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.828,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8261256163608439,
+      "grad_norm": 0.50812166929245,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8277423005415891,
+      "grad_norm": 0.5319661498069763,
+      "learning_rate": 0.0002,
+      "loss": 0.8243,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8293589847223345,
+      "grad_norm": 0.4949689209461212,
+      "learning_rate": 0.0002,
+      "loss": 0.7798,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8309756689030798,
+      "grad_norm": 0.5151591300964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8325923530838251,
+      "grad_norm": 0.5530214309692383,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8342090372645704,
+      "grad_norm": 0.6297410130500793,
+      "learning_rate": 0.0002,
+      "loss": 0.8251,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8358257214453156,
+      "grad_norm": 0.5466840267181396,
+      "learning_rate": 0.0002,
+      "loss": 0.8067,
+      "step": 5170
+    },
+    {
+      "epoch": 0.837442405626061,
+      "grad_norm": 0.652913510799408,
+      "learning_rate": 0.0002,
+      "loss": 0.7875,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8390590898068062,
+      "grad_norm": 0.5811293125152588,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8406757739875516,
+      "grad_norm": 0.5109550952911377,
+      "learning_rate": 0.0002,
+      "loss": 0.7412,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8422924581682968,
+      "grad_norm": 0.4551706612110138,
+      "learning_rate": 0.0002,
+      "loss": 0.8077,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8439091423490421,
+      "grad_norm": 0.5813754200935364,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8455258265297874,
+      "grad_norm": 0.5856947898864746,
+      "learning_rate": 0.0002,
+      "loss": 0.802,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8471425107105327,
+      "grad_norm": 0.5482739210128784,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8487591948912779,
+      "grad_norm": 0.49023720622062683,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8503758790720233,
+      "grad_norm": 0.49472475051879883,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8519925632527686,
+      "grad_norm": 0.5490226745605469,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8536092474335139,
+      "grad_norm": 0.5340665578842163,
+      "learning_rate": 0.0002,
+      "loss": 0.8333,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8552259316142592,
+      "grad_norm": 0.5962483882904053,
+      "learning_rate": 0.0002,
+      "loss": 0.8277,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8568426157950044,
+      "grad_norm": 0.586358368396759,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8584592999757498,
+      "grad_norm": 0.49120277166366577,
+      "learning_rate": 0.0002,
+      "loss": 0.7831,
+      "step": 5310
+    },
+    {
+      "epoch": 0.860075984156495,
+      "grad_norm": 0.5887332558631897,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8616926683372403,
+      "grad_norm": 0.42496153712272644,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.5489874482154846,
+      "learning_rate": 0.0002,
+      "loss": 0.7905,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8649260366987309,
+      "grad_norm": 0.5850813984870911,
+      "learning_rate": 0.0002,
+      "loss": 0.7958,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8665427208794761,
+      "grad_norm": 0.517487108707428,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8681594050602215,
+      "grad_norm": 0.5339142680168152,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8697760892409668,
+      "grad_norm": 0.6236387491226196,
+      "learning_rate": 0.0002,
+      "loss": 0.818,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8713927734217121,
+      "grad_norm": 0.5752192735671997,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8730094576024574,
+      "grad_norm": 0.6724614500999451,
+      "learning_rate": 0.0002,
+      "loss": 0.8542,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8746261417832026,
+      "grad_norm": 0.5280613303184509,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 5410
+    },
+    {
+      "epoch": 0.876242825963948,
+      "grad_norm": 0.44033288955688477,
+      "learning_rate": 0.0002,
+      "loss": 0.8231,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8778595101446932,
+      "grad_norm": 0.5199708342552185,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8794761943254386,
+      "grad_norm": 0.46778348088264465,
+      "learning_rate": 0.0002,
+      "loss": 0.7852,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8810928785061838,
+      "grad_norm": 0.4657754898071289,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8827095626869291,
+      "grad_norm": 0.5472902655601501,
+      "learning_rate": 0.0002,
+      "loss": 0.7799,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8843262468676744,
+      "grad_norm": 0.4876766800880432,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8859429310484197,
+      "grad_norm": 0.5057248473167419,
+      "learning_rate": 0.0002,
+      "loss": 0.7906,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8875596152291649,
+      "grad_norm": 0.4637320637702942,
+      "learning_rate": 0.0002,
+      "loss": 0.8124,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8891762994099103,
+      "grad_norm": 0.471955806016922,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8907929835906556,
+      "grad_norm": 0.5209813714027405,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8924096677714008,
+      "grad_norm": 0.6213834285736084,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8940263519521462,
+      "grad_norm": 0.5215408205986023,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8956430361328914,
+      "grad_norm": 0.580478310585022,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8972597203136368,
+      "grad_norm": 0.49102169275283813,
+      "learning_rate": 0.0002,
+      "loss": 0.8371,
+      "step": 5550
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.6043479442596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7806,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9004930886751273,
+      "grad_norm": 0.5636463165283203,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9021097728558726,
+      "grad_norm": 0.5620124340057373,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9037264570366179,
+      "grad_norm": 0.5206354856491089,
+      "learning_rate": 0.0002,
+      "loss": 0.8083,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9053431412173631,
+      "grad_norm": 0.5798229575157166,
+      "learning_rate": 0.0002,
+      "loss": 0.8557,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9069598253981085,
+      "grad_norm": 0.6428212523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.8097,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9085765095788537,
+      "grad_norm": 0.48064687848091125,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9101931937595991,
+      "grad_norm": 0.6347860097885132,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9118098779403444,
+      "grad_norm": 0.5353913307189941,
+      "learning_rate": 0.0002,
+      "loss": 0.851,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9134265621210896,
+      "grad_norm": 0.5323944091796875,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5650
+    },
+    {
+      "epoch": 0.915043246301835,
+      "grad_norm": 0.5261843204498291,
+      "learning_rate": 0.0002,
+      "loss": 0.8393,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9166599304825802,
+      "grad_norm": 0.5451326966285706,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9182766146633256,
+      "grad_norm": 0.5183324217796326,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9198932988440708,
+      "grad_norm": 0.47229018807411194,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9215099830248161,
+      "grad_norm": 0.49180513620376587,
+      "learning_rate": 0.0002,
+      "loss": 0.7757,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9231266672055614,
+      "grad_norm": 0.5419785380363464,
+      "learning_rate": 0.0002,
+      "loss": 0.8735,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9247433513863067,
+      "grad_norm": 0.5408698916435242,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9263600355670519,
+      "grad_norm": 0.5286232829093933,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9279767197477973,
+      "grad_norm": 0.7539758086204529,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9295934039285425,
+      "grad_norm": 0.5166944861412048,
+      "learning_rate": 0.0002,
+      "loss": 0.8118,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9312100881092878,
+      "grad_norm": 0.6601425409317017,
+      "learning_rate": 0.0002,
+      "loss": 0.783,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9328267722900332,
+      "grad_norm": 0.5029960870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9344434564707784,
+      "grad_norm": 0.4926645755767822,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9360601406515238,
+      "grad_norm": 0.5739615559577942,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5790
+    },
+    {
+      "epoch": 0.937676824832269,
+      "grad_norm": 0.5058279037475586,
+      "learning_rate": 0.0002,
+      "loss": 0.8037,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9392935090130143,
+      "grad_norm": 0.5260962247848511,
+      "learning_rate": 0.0002,
+      "loss": 0.8537,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9409101931937596,
+      "grad_norm": 0.5768588185310364,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9425268773745049,
+      "grad_norm": 0.5170126557350159,
+      "learning_rate": 0.0002,
+      "loss": 0.8215,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9441435615552501,
+      "grad_norm": 0.5745864510536194,
+      "learning_rate": 0.0002,
+      "loss": 0.7422,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9457602457359955,
+      "grad_norm": 0.5551357865333557,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9473769299167407,
+      "grad_norm": 0.5776078701019287,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9489936140974861,
+      "grad_norm": 0.5340062379837036,
+      "learning_rate": 0.0002,
+      "loss": 0.8527,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9506102982782314,
+      "grad_norm": 0.6447290182113647,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9522269824589766,
+      "grad_norm": 0.5123815536499023,
+      "learning_rate": 0.0002,
+      "loss": 0.7945,
+      "step": 5890
+    },
+    {
+      "epoch": 0.953843666639722,
+      "grad_norm": 0.48547613620758057,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9554603508204672,
+      "grad_norm": 0.5791414976119995,
+      "learning_rate": 0.0002,
+      "loss": 0.7896,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9570770350012126,
+      "grad_norm": 0.6195011734962463,
+      "learning_rate": 0.0002,
+      "loss": 0.8408,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9586937191819578,
+      "grad_norm": 0.6323803067207336,
+      "learning_rate": 0.0002,
+      "loss": 0.7805,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9603104033627031,
+      "grad_norm": 0.45552879571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9619270875434484,
+      "grad_norm": 0.5796473622322083,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9635437717241937,
+      "grad_norm": 0.647261381149292,
+      "learning_rate": 0.0002,
+      "loss": 0.7672,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9651604559049389,
+      "grad_norm": 0.5487682819366455,
+      "learning_rate": 0.0002,
+      "loss": 0.8086,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9667771400856843,
+      "grad_norm": 0.5743663907051086,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9683938242664295,
+      "grad_norm": 0.5470591187477112,
+      "learning_rate": 0.0002,
+      "loss": 0.8153,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9700105084471748,
+      "grad_norm": 0.5901660323143005,
+      "learning_rate": 0.0002,
+      "loss": 0.8119,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9716271926279202,
+      "grad_norm": 0.6544759273529053,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9732438768086654,
+      "grad_norm": 0.6288470029830933,
+      "learning_rate": 0.0002,
+      "loss": 0.7536,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9748605609894108,
+      "grad_norm": 0.673153817653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 6030
+    },
+    {
+      "epoch": 0.976477245170156,
+      "grad_norm": 0.42854753136634827,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9780939293509013,
+      "grad_norm": 0.5227066278457642,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9797106135316466,
+      "grad_norm": 0.5372416973114014,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9813272977123919,
+      "grad_norm": 0.6026402115821838,
+      "learning_rate": 0.0002,
+      "loss": 0.7591,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9829439818931371,
+      "grad_norm": 0.49547791481018066,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9845606660738825,
+      "grad_norm": 0.4641951322555542,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9861773502546277,
+      "grad_norm": 0.5818535089492798,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9877940344353731,
+      "grad_norm": 0.63955157995224,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9894107186161183,
+      "grad_norm": 0.5649438500404358,
+      "learning_rate": 0.0002,
+      "loss": 0.7547,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9910274027968636,
+      "grad_norm": 0.5290433168411255,
+      "learning_rate": 0.0002,
+      "loss": 0.7861,
+      "step": 6130
+    },
+    {
+      "epoch": 0.992644086977609,
+      "grad_norm": 0.6399374008178711,
+      "learning_rate": 0.0002,
+      "loss": 0.8109,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9942607711583542,
+      "grad_norm": 0.6736576557159424,
+      "learning_rate": 0.0002,
+      "loss": 0.8373,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9958774553390995,
+      "grad_norm": 0.515420138835907,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9974941395198448,
+      "grad_norm": 0.562677800655365,
+      "learning_rate": 0.0002,
+      "loss": 0.8032,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9991108237005901,
+      "grad_norm": 0.7113858461380005,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9999191657909627,
+      "eval_loss": 1.0871200561523438,
+      "eval_runtime": 122.2071,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0007275078813354,
+      "grad_norm": 0.7111801505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0023441920620806,
+      "grad_norm": 0.5402125716209412,
+      "learning_rate": 0.0002,
+      "loss": 0.6865,
+      "step": 6200
+    },
+    {
+      "epoch": 1.003960876242826,
+      "grad_norm": 0.6098830103874207,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0055775604235713,
+      "grad_norm": 0.5829983353614807,
+      "learning_rate": 0.0002,
+      "loss": 0.7631,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0071942446043165,
+      "grad_norm": 0.5614621043205261,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0088109287850617,
+      "grad_norm": 0.5954238772392273,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0104276129658072,
+      "grad_norm": 0.6480574607849121,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0120442971465524,
+      "grad_norm": 0.6051128506660461,
+      "learning_rate": 0.0002,
+      "loss": 0.7514,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0136609813272976,
+      "grad_norm": 0.6318870782852173,
+      "learning_rate": 0.0002,
+      "loss": 0.7237,
+      "step": 6270
+    },
+    {
+      "epoch": 1.015277665508043,
+      "grad_norm": 0.5048980116844177,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0168943496887883,
+      "grad_norm": 0.6346936225891113,
+      "learning_rate": 0.0002,
+      "loss": 0.7391,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0185110338695336,
+      "grad_norm": 0.5711665749549866,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0201277180502788,
+      "grad_norm": 0.5175361037254333,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0217444022310243,
+      "grad_norm": 0.5360831618309021,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0233610864117695,
+      "grad_norm": 0.614675760269165,
+      "learning_rate": 0.0002,
+      "loss": 0.7112,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0249777705925147,
+      "grad_norm": 0.5626118183135986,
+      "learning_rate": 0.0002,
+      "loss": 0.7748,
+      "step": 6340
+    },
+    {
+      "epoch": 1.02659445477326,
+      "grad_norm": 0.574897289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.7375,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0282111389540054,
+      "grad_norm": 0.7185447812080383,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0298278231347506,
+      "grad_norm": 0.6705799698829651,
+      "learning_rate": 0.0002,
+      "loss": 0.703,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0314445073154959,
+      "grad_norm": 0.6740428805351257,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0330611914962413,
+      "grad_norm": 0.663902759552002,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0346778756769865,
+      "grad_norm": 0.5029543042182922,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0362945598577318,
+      "grad_norm": 0.7813863158226013,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 6410
+    },
+    {
+      "epoch": 1.037911244038477,
+      "grad_norm": 0.5396282076835632,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0395279282192225,
+      "grad_norm": 0.5253293514251709,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0411446123999677,
+      "grad_norm": 0.7236770987510681,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 6440
+    },
+    {
+      "epoch": 1.042761296580713,
+      "grad_norm": 0.5670917630195618,
+      "learning_rate": 0.0002,
+      "loss": 0.7259,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0443779807614582,
+      "grad_norm": 0.6031978726387024,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0459946649422036,
+      "grad_norm": 0.5309213399887085,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0476113491229488,
+      "grad_norm": 0.7114651799201965,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 6480
+    },
+    {
+      "epoch": 1.049228033303694,
+      "grad_norm": 0.5591610670089722,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0508447174844395,
+      "grad_norm": 0.5185961127281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0524614016651848,
+      "grad_norm": 0.6510552167892456,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6510
+    },
+    {
+      "epoch": 1.05407808584593,
+      "grad_norm": 0.6557928919792175,
+      "learning_rate": 0.0002,
+      "loss": 0.7057,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0556947700266752,
+      "grad_norm": 0.6973192691802979,
+      "learning_rate": 0.0002,
+      "loss": 0.8056,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0573114542074207,
+      "grad_norm": 0.6226583123207092,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 6540
+    },
+    {
+      "epoch": 1.058928138388166,
+      "grad_norm": 0.5633195638656616,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0605448225689111,
+      "grad_norm": 0.7466658353805542,
+      "learning_rate": 0.0002,
+      "loss": 0.7082,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0621615067496564,
+      "grad_norm": 0.6462772488594055,
+      "learning_rate": 0.0002,
+      "loss": 0.7059,
+      "step": 6570
+    },
+    {
+      "epoch": 1.0637781909304018,
+      "grad_norm": 0.5266856551170349,
+      "learning_rate": 0.0002,
+      "loss": 0.7046,
+      "step": 6580
+    },
+    {
+      "epoch": 1.065394875111147,
+      "grad_norm": 0.534392774105072,
+      "learning_rate": 0.0002,
+      "loss": 0.7157,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0670115592918923,
+      "grad_norm": 0.7514177560806274,
+      "learning_rate": 0.0002,
+      "loss": 0.7115,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0686282434726375,
+      "grad_norm": 0.7593035697937012,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 6610
+    },
+    {
+      "epoch": 1.070244927653383,
+      "grad_norm": 0.5277858972549438,
+      "learning_rate": 0.0002,
+      "loss": 0.6836,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0718616118341282,
+      "grad_norm": 0.5573670268058777,
+      "learning_rate": 0.0002,
+      "loss": 0.7405,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0734782960148734,
+      "grad_norm": 0.6802396774291992,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 6640
+    },
+    {
+      "epoch": 1.0750949801956189,
+      "grad_norm": 0.7367215752601624,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0767116643763641,
+      "grad_norm": 0.5961891412734985,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0783283485571094,
+      "grad_norm": 0.5736313462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0799450327378546,
+      "grad_norm": 0.619219183921814,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0815617169186,
+      "grad_norm": 0.6214390993118286,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0831784010993453,
+      "grad_norm": 0.564536988735199,
+      "learning_rate": 0.0002,
+      "loss": 0.7554,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0847950852800905,
+      "grad_norm": 0.5838140249252319,
+      "learning_rate": 0.0002,
+      "loss": 0.7362,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0864117694608357,
+      "grad_norm": 0.7000553607940674,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0880284536415812,
+      "grad_norm": 0.7078263759613037,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 6730
+    },
+    {
+      "epoch": 1.0896451378223264,
+      "grad_norm": 0.8353848457336426,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6740
+    },
+    {
+      "epoch": 1.0912618220030716,
+      "grad_norm": 0.5615518689155579,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 6750
+    },
+    {
+      "epoch": 1.0928785061838169,
+      "grad_norm": 0.5475581288337708,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 6760
+    },
+    {
+      "epoch": 1.0944951903645623,
+      "grad_norm": 0.5835978388786316,
+      "learning_rate": 0.0002,
+      "loss": 0.7652,
+      "step": 6770
+    },
+    {
+      "epoch": 1.0961118745453076,
+      "grad_norm": 0.5516105890274048,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 6780
+    },
+    {
+      "epoch": 1.0977285587260528,
+      "grad_norm": 0.5875251889228821,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 6790
+    },
+    {
+      "epoch": 1.0993452429067982,
+      "grad_norm": 0.7376947999000549,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1009619270875435,
+      "grad_norm": 0.5656165480613708,
+      "learning_rate": 0.0002,
+      "loss": 0.7512,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1025786112682887,
+      "grad_norm": 0.6365954279899597,
+      "learning_rate": 0.0002,
+      "loss": 0.7409,
+      "step": 6820
+    },
+    {
+      "epoch": 1.104195295449034,
+      "grad_norm": 0.5033080577850342,
+      "learning_rate": 0.0002,
+      "loss": 0.7392,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1058119796297794,
+      "grad_norm": 0.617396891117096,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1074286638105246,
+      "grad_norm": 0.6395374536514282,
+      "learning_rate": 0.0002,
+      "loss": 0.7006,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1090453479912699,
+      "grad_norm": 0.6775295734405518,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1106620321720153,
+      "grad_norm": 0.6655223965644836,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1122787163527605,
+      "grad_norm": 0.676655113697052,
+      "learning_rate": 0.0002,
+      "loss": 0.7553,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1138954005335058,
+      "grad_norm": 0.6062718629837036,
+      "learning_rate": 0.0002,
+      "loss": 0.7342,
+      "step": 6890
+    },
+    {
+      "epoch": 1.115512084714251,
+      "grad_norm": 0.590943455696106,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1171287688949965,
+      "grad_norm": 0.6315317153930664,
+      "learning_rate": 0.0002,
+      "loss": 0.6705,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1187454530757417,
+      "grad_norm": 0.47979024052619934,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 6920
+    },
+    {
+      "epoch": 1.120362137256487,
+      "grad_norm": 0.647298276424408,
+      "learning_rate": 0.0002,
+      "loss": 0.7002,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1219788214372322,
+      "grad_norm": 0.7336484789848328,
+      "learning_rate": 0.0002,
+      "loss": 0.7502,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1235955056179776,
+      "grad_norm": 0.5071424245834351,
+      "learning_rate": 0.0002,
+      "loss": 0.693,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1252121897987228,
+      "grad_norm": 0.6527144312858582,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 6960
+    },
+    {
+      "epoch": 1.126828873979468,
+      "grad_norm": 0.6935935020446777,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1284455581602133,
+      "grad_norm": 0.8026931881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1300622423409588,
+      "grad_norm": 0.5210393667221069,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 6990
+    },
+    {
+      "epoch": 1.131678926521704,
+      "grad_norm": 0.60475093126297,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1332956107024492,
+      "grad_norm": 0.6417073607444763,
+      "learning_rate": 0.0002,
+      "loss": 0.7495,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1349122948831947,
+      "grad_norm": 0.6732175946235657,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 7020
+    },
+    {
+      "epoch": 1.13652897906394,
+      "grad_norm": 0.6719491481781006,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1381456632446851,
+      "grad_norm": 0.5708295106887817,
+      "learning_rate": 0.0002,
+      "loss": 0.7694,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1397623474254304,
+      "grad_norm": 0.7141719460487366,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1413790316061758,
+      "grad_norm": 0.6187017560005188,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 7060
+    },
+    {
+      "epoch": 1.142995715786921,
+      "grad_norm": 0.50581294298172,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1446123999676663,
+      "grad_norm": 0.5620143413543701,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1462290841484115,
+      "grad_norm": 0.6231929659843445,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 7090
+    },
+    {
+      "epoch": 1.147845768329157,
+      "grad_norm": 0.5775774121284485,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1494624525099022,
+      "grad_norm": 0.6492809653282166,
+      "learning_rate": 0.0002,
+      "loss": 0.7728,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1510791366906474,
+      "grad_norm": 0.6434972286224365,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 7120
+    },
+    {
+      "epoch": 1.1526958208713927,
+      "grad_norm": 0.6191812753677368,
+      "learning_rate": 0.0002,
+      "loss": 0.7374,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1543125050521381,
+      "grad_norm": 0.6690331697463989,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 7140
+    },
+    {
+      "epoch": 1.1559291892328833,
+      "grad_norm": 0.5977938175201416,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1575458734136286,
+      "grad_norm": 0.6195854544639587,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7160
+    },
+    {
+      "epoch": 1.159162557594374,
+      "grad_norm": 0.5752048492431641,
+      "learning_rate": 0.0002,
+      "loss": 0.7249,
+      "step": 7170
+    },
+    {
+      "epoch": 1.1607792417751193,
+      "grad_norm": 0.589081883430481,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1623959259558645,
+      "grad_norm": 0.756996750831604,
+      "learning_rate": 0.0002,
+      "loss": 0.704,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1640126101366097,
+      "grad_norm": 0.7614967226982117,
+      "learning_rate": 0.0002,
+      "loss": 0.7404,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1656292943173552,
+      "grad_norm": 0.6120437979698181,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1672459784981004,
+      "grad_norm": 0.6210004687309265,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1688626626788456,
+      "grad_norm": 0.6044116020202637,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7230
+    },
+    {
+      "epoch": 1.170479346859591,
+      "grad_norm": 0.5418457388877869,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1720960310403363,
+      "grad_norm": 0.6413537263870239,
+      "learning_rate": 0.0002,
+      "loss": 0.6938,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1737127152210816,
+      "grad_norm": 0.5777867436408997,
+      "learning_rate": 0.0002,
+      "loss": 0.6978,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1753293994018268,
+      "grad_norm": 0.7092402577400208,
+      "learning_rate": 0.0002,
+      "loss": 0.7503,
+      "step": 7270
+    },
+    {
+      "epoch": 1.176946083582572,
+      "grad_norm": 0.6351709365844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1785627677633175,
+      "grad_norm": 0.6172189712524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7527,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1801794519440627,
+      "grad_norm": 0.6801714897155762,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7300
+    },
+    {
+      "epoch": 1.181796136124808,
+      "grad_norm": 0.6044712066650391,
+      "learning_rate": 0.0002,
+      "loss": 0.6941,
+      "step": 7310
+    },
+    {
+      "epoch": 1.1834128203055534,
+      "grad_norm": 0.7413212060928345,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 7320
+    },
+    {
+      "epoch": 1.1850295044862986,
+      "grad_norm": 0.5303856134414673,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 7330
+    },
+    {
+      "epoch": 1.1866461886670439,
+      "grad_norm": 0.5647098422050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6915,
+      "step": 7340
+    },
+    {
+      "epoch": 1.188262872847789,
+      "grad_norm": 0.7374135255813599,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 7350
+    },
+    {
+      "epoch": 1.1898795570285345,
+      "grad_norm": 0.5710089206695557,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1914962412092798,
+      "grad_norm": 0.6073619723320007,
+      "learning_rate": 0.0002,
+      "loss": 0.8289,
+      "step": 7370
+    },
+    {
+      "epoch": 1.193112925390025,
+      "grad_norm": 0.5899916887283325,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1947296095707705,
+      "grad_norm": 0.7762434482574463,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1963462937515157,
+      "grad_norm": 0.679949939250946,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7400
+    },
+    {
+      "epoch": 1.197962977932261,
+      "grad_norm": 0.6106849312782288,
+      "learning_rate": 0.0002,
+      "loss": 0.7599,
+      "step": 7410
+    },
+    {
+      "epoch": 1.1995796621130062,
+      "grad_norm": 0.682461678981781,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2011963462937516,
+      "grad_norm": 0.6087017059326172,
+      "learning_rate": 0.0002,
+      "loss": 0.7741,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2028130304744968,
+      "grad_norm": 0.63739013671875,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 7440
+    },
+    {
+      "epoch": 1.204429714655242,
+      "grad_norm": 0.6154777407646179,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2060463988359873,
+      "grad_norm": 0.7491534948348999,
+      "learning_rate": 0.0002,
+      "loss": 0.7565,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2076630830167328,
+      "grad_norm": 0.6664797067642212,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 7470
+    },
+    {
+      "epoch": 1.209279767197478,
+      "grad_norm": 0.6660266518592834,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2108964513782232,
+      "grad_norm": 0.6972551345825195,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2125131355589684,
+      "grad_norm": 0.6157945990562439,
+      "learning_rate": 0.0002,
+      "loss": 0.7023,
+      "step": 7500
+    },
+    {
+      "epoch": 1.214129819739714,
+      "grad_norm": 0.5199310183525085,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2157465039204591,
+      "grad_norm": 0.577610433101654,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2173631881012044,
+      "grad_norm": 0.53652423620224,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2189798722819498,
+      "grad_norm": 0.6479050517082214,
+      "learning_rate": 0.0002,
+      "loss": 0.7393,
+      "step": 7540
+    },
+    {
+      "epoch": 1.220596556462695,
+      "grad_norm": 0.618748128414154,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2222132406434403,
+      "grad_norm": 0.6311424374580383,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2238299248241855,
+      "grad_norm": 0.6595825552940369,
+      "learning_rate": 0.0002,
+      "loss": 0.7272,
+      "step": 7570
+    },
+    {
+      "epoch": 1.225446609004931,
+      "grad_norm": 0.5198960900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2270632931856762,
+      "grad_norm": 0.578650712966919,
+      "learning_rate": 0.0002,
+      "loss": 0.674,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2286799773664214,
+      "grad_norm": 0.6080220937728882,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2302966615471669,
+      "grad_norm": 0.7050248384475708,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2319133457279121,
+      "grad_norm": 0.6652196049690247,
+      "learning_rate": 0.0002,
+      "loss": 0.7032,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2335300299086573,
+      "grad_norm": 0.7322776317596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7085,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2351467140894026,
+      "grad_norm": 0.4998728036880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2367633982701478,
+      "grad_norm": 0.6428788900375366,
+      "learning_rate": 0.0002,
+      "loss": 0.7214,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2383800824508933,
+      "grad_norm": 0.585242509841919,
+      "learning_rate": 0.0002,
+      "loss": 0.7699,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2399967666316385,
+      "grad_norm": 0.5211917757987976,
+      "learning_rate": 0.0002,
+      "loss": 0.7621,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2416134508123837,
+      "grad_norm": 0.6490384340286255,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2432301349931292,
+      "grad_norm": 0.6249763369560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2448468191738744,
+      "grad_norm": 0.71870356798172,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2464635033546196,
+      "grad_norm": 0.6761967539787292,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2480801875353649,
+      "grad_norm": 0.6500617265701294,
+      "learning_rate": 0.0002,
+      "loss": 0.7501,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2496968717161103,
+      "grad_norm": 0.8069869875907898,
+      "learning_rate": 0.0002,
+      "loss": 0.7903,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2513135558968556,
+      "grad_norm": 0.6044608950614929,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2529302400776008,
+      "grad_norm": 0.6573283076286316,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2545469242583462,
+      "grad_norm": 0.625430166721344,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2561636084390915,
+      "grad_norm": 0.5442022681236267,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7770
+    },
+    {
+      "epoch": 1.2577802926198367,
+      "grad_norm": 0.6818386912345886,
+      "learning_rate": 0.0002,
+      "loss": 0.7172,
+      "step": 7780
+    },
+    {
+      "epoch": 1.259396976800582,
+      "grad_norm": 0.6381874084472656,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2610136609813272,
+      "grad_norm": 0.6269212961196899,
+      "learning_rate": 0.0002,
+      "loss": 0.6834,
+      "step": 7800
+    },
+    {
+      "epoch": 1.2626303451620726,
+      "grad_norm": 0.600121259689331,
+      "learning_rate": 0.0002,
+      "loss": 0.7821,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2642470293428179,
+      "grad_norm": 0.6337703466415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2658637135235633,
+      "grad_norm": 0.7234963774681091,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 7830
+    },
+    {
+      "epoch": 1.2674803977043085,
+      "grad_norm": 0.800184965133667,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2690970818850538,
+      "grad_norm": 0.7539464831352234,
+      "learning_rate": 0.0002,
+      "loss": 0.7426,
+      "step": 7850
+    },
+    {
+      "epoch": 1.270713766065799,
+      "grad_norm": 0.5493760704994202,
+      "learning_rate": 0.0002,
+      "loss": 0.7496,
+      "step": 7860
+    },
+    {
+      "epoch": 1.2723304502465442,
+      "grad_norm": 0.7477145791053772,
+      "learning_rate": 0.0002,
+      "loss": 0.7537,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2739471344272897,
+      "grad_norm": 0.6366362571716309,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 7880
+    },
+    {
+      "epoch": 1.275563818608035,
+      "grad_norm": 0.7419533729553223,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2771805027887801,
+      "grad_norm": 0.6141223311424255,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 7900
+    },
+    {
+      "epoch": 1.2787971869695256,
+      "grad_norm": 0.7522598505020142,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 7910
+    },
+    {
+      "epoch": 1.2804138711502708,
+      "grad_norm": 0.6935804486274719,
+      "learning_rate": 0.0002,
+      "loss": 0.7097,
+      "step": 7920
+    },
+    {
+      "epoch": 1.282030555331016,
+      "grad_norm": 0.7239290475845337,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 7930
+    },
+    {
+      "epoch": 1.2836472395117613,
+      "grad_norm": 0.8800187110900879,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 7940
+    },
+    {
+      "epoch": 1.2852639236925067,
+      "grad_norm": 0.540458083152771,
+      "learning_rate": 0.0002,
+      "loss": 0.6991,
+      "step": 7950
+    },
+    {
+      "epoch": 1.286880607873252,
+      "grad_norm": 0.6492934226989746,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 7960
+    },
+    {
+      "epoch": 1.2884972920539972,
+      "grad_norm": 0.6543959379196167,
+      "learning_rate": 0.0002,
+      "loss": 0.7742,
+      "step": 7970
+    },
+    {
+      "epoch": 1.2901139762347427,
+      "grad_norm": 0.5804705619812012,
+      "learning_rate": 0.0002,
+      "loss": 0.7316,
+      "step": 7980
+    },
+    {
+      "epoch": 1.291730660415488,
+      "grad_norm": 0.7074727416038513,
+      "learning_rate": 0.0002,
+      "loss": 0.796,
+      "step": 7990
+    },
+    {
+      "epoch": 1.2933473445962331,
+      "grad_norm": 0.5347974300384521,
+      "learning_rate": 0.0002,
+      "loss": 0.7034,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2949640287769784,
+      "grad_norm": 0.6457298398017883,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 8010
+    },
+    {
+      "epoch": 1.2965807129577236,
+      "grad_norm": 0.6407219171524048,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8020
+    },
+    {
+      "epoch": 1.298197397138469,
+      "grad_norm": 0.828439474105835,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2998140813192143,
+      "grad_norm": 0.4840380549430847,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3014307654999595,
+      "grad_norm": 0.5921024680137634,
+      "learning_rate": 0.0002,
+      "loss": 0.7283,
+      "step": 8050
+    },
+    {
+      "epoch": 1.303047449680705,
+      "grad_norm": 0.6170315146446228,
+      "learning_rate": 0.0002,
+      "loss": 0.7477,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3046641338614502,
+      "grad_norm": 0.5374847054481506,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3062808180421954,
+      "grad_norm": 0.545758068561554,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3078975022229407,
+      "grad_norm": 0.55641770362854,
+      "learning_rate": 0.0002,
+      "loss": 0.7463,
+      "step": 8090
+    },
+    {
+      "epoch": 1.309514186403686,
+      "grad_norm": 0.6724897027015686,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3111308705844313,
+      "grad_norm": 0.6923972368240356,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3127475547651766,
+      "grad_norm": 0.5136841535568237,
+      "learning_rate": 0.0002,
+      "loss": 0.7149,
+      "step": 8120
+    },
+    {
+      "epoch": 1.314364238945922,
+      "grad_norm": 0.6766283512115479,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3159809231266673,
+      "grad_norm": 0.6283926367759705,
+      "learning_rate": 0.0002,
+      "loss": 0.7489,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3175976073074125,
+      "grad_norm": 0.644216001033783,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3192142914881577,
+      "grad_norm": 0.7827503085136414,
+      "learning_rate": 0.0002,
+      "loss": 0.7125,
+      "step": 8160
+    },
+    {
+      "epoch": 1.320830975668903,
+      "grad_norm": 0.6651390790939331,
+      "learning_rate": 0.0002,
+      "loss": 0.7271,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3224476598496484,
+      "grad_norm": 0.5547412633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7778,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3240643440303936,
+      "grad_norm": 0.6765179634094238,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 8190
+    },
+    {
+      "epoch": 1.325681028211139,
+      "grad_norm": 0.6822077035903931,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3272977123918843,
+      "grad_norm": 0.5941002368927002,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3289143965726296,
+      "grad_norm": 0.4850037097930908,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3305310807533748,
+      "grad_norm": 0.6162990927696228,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 8230
+    },
+    {
+      "epoch": 1.33214776493412,
+      "grad_norm": 0.6665613651275635,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3337644491148655,
+      "grad_norm": 0.618192732334137,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3353811332956107,
+      "grad_norm": 0.710418701171875,
+      "learning_rate": 0.0002,
+      "loss": 0.7532,
+      "step": 8260
+    },
+    {
+      "epoch": 1.336997817476356,
+      "grad_norm": 0.5109876990318298,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3386145016571014,
+      "grad_norm": 0.6791711449623108,
+      "learning_rate": 0.0002,
+      "loss": 0.7303,
+      "step": 8280
+    },
+    {
+      "epoch": 1.3402311858378466,
+      "grad_norm": 0.6836432814598083,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3418478700185918,
+      "grad_norm": 0.5579386353492737,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8300
+    },
+    {
+      "epoch": 1.343464554199337,
+      "grad_norm": 0.6713546514511108,
+      "learning_rate": 0.0002,
+      "loss": 0.7377,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3450812383800825,
+      "grad_norm": 0.5353720188140869,
+      "learning_rate": 0.0002,
+      "loss": 0.7756,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3466979225608278,
+      "grad_norm": 0.5813682675361633,
+      "learning_rate": 0.0002,
+      "loss": 0.718,
+      "step": 8330
+    },
+    {
+      "epoch": 1.348314606741573,
+      "grad_norm": 0.8158791661262512,
+      "learning_rate": 0.0002,
+      "loss": 0.7294,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3499312909223184,
+      "grad_norm": 0.6193785071372986,
+      "learning_rate": 0.0002,
+      "loss": 0.6992,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3515479751030637,
+      "grad_norm": 0.6353939771652222,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 8360
+    },
+    {
+      "epoch": 1.353164659283809,
+      "grad_norm": 0.6925048232078552,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 8370
+    },
+    {
+      "epoch": 1.3547813434645541,
+      "grad_norm": 0.988264799118042,
+      "learning_rate": 0.0002,
+      "loss": 0.736,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3563980276452994,
+      "grad_norm": 0.6476002931594849,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3580147118260448,
+      "grad_norm": 0.7120398879051208,
+      "learning_rate": 0.0002,
+      "loss": 0.776,
+      "step": 8400
+    },
+    {
+      "epoch": 1.35963139600679,
+      "grad_norm": 0.9048416614532471,
+      "learning_rate": 0.0002,
+      "loss": 0.7368,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3612480801875353,
+      "grad_norm": 0.7000672817230225,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3628647643682807,
+      "grad_norm": 0.6015632152557373,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 8430
+    },
+    {
+      "epoch": 1.364481448549026,
+      "grad_norm": 0.612516462802887,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3660981327297712,
+      "grad_norm": 0.5969301462173462,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 8450
+    },
+    {
+      "epoch": 1.3677148169105164,
+      "grad_norm": 0.6730654239654541,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 8460
+    },
+    {
+      "epoch": 1.369331501091262,
+      "grad_norm": 0.6386392116546631,
+      "learning_rate": 0.0002,
+      "loss": 0.7465,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3709481852720071,
+      "grad_norm": 0.739544153213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3725648694527524,
+      "grad_norm": 0.6462782621383667,
+      "learning_rate": 0.0002,
+      "loss": 0.7892,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3741815536334978,
+      "grad_norm": 0.7346843481063843,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 8500
+    },
+    {
+      "epoch": 1.375798237814243,
+      "grad_norm": 0.6884821057319641,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8510
+    },
+    {
+      "epoch": 1.3774149219949883,
+      "grad_norm": 0.6999333500862122,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3790316061757335,
+      "grad_norm": 0.5378713011741638,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 8530
+    },
+    {
+      "epoch": 1.3806482903564787,
+      "grad_norm": 0.5417906641960144,
+      "learning_rate": 0.0002,
+      "loss": 0.6797,
+      "step": 8540
+    },
+    {
+      "epoch": 1.3822649745372242,
+      "grad_norm": 0.6602526307106018,
+      "learning_rate": 0.0002,
+      "loss": 0.7499,
+      "step": 8550
+    },
+    {
+      "epoch": 1.3838816587179694,
+      "grad_norm": 0.7073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.7356,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3854983428987149,
+      "grad_norm": 0.5841707587242126,
+      "learning_rate": 0.0002,
+      "loss": 0.75,
+      "step": 8570
+    },
+    {
+      "epoch": 1.38711502707946,
+      "grad_norm": 0.7031095027923584,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 8580
+    },
+    {
+      "epoch": 1.3887317112602053,
+      "grad_norm": 0.5198570489883423,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 8590
+    },
+    {
+      "epoch": 1.3903483954409506,
+      "grad_norm": 0.7261320352554321,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 8600
+    },
+    {
+      "epoch": 1.3919650796216958,
+      "grad_norm": 0.5616350173950195,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3935817638024413,
+      "grad_norm": 0.5185914635658264,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 8620
+    },
+    {
+      "epoch": 1.3951984479831865,
+      "grad_norm": 0.5814694762229919,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 8630
+    },
+    {
+      "epoch": 1.3968151321639317,
+      "grad_norm": 0.6977371573448181,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 8640
+    },
+    {
+      "epoch": 1.3984318163446772,
+      "grad_norm": 0.6855689883232117,
+      "learning_rate": 0.0002,
+      "loss": 0.7574,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4000485005254224,
+      "grad_norm": 0.5414357781410217,
+      "learning_rate": 0.0002,
+      "loss": 0.7802,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4016651847061676,
+      "grad_norm": 0.6970012784004211,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4032818688869129,
+      "grad_norm": 0.526079535484314,
+      "learning_rate": 0.0002,
+      "loss": 0.7421,
+      "step": 8680
+    },
+    {
+      "epoch": 1.404898553067658,
+      "grad_norm": 0.758712887763977,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4065152372484035,
+      "grad_norm": 0.7118762731552124,
+      "learning_rate": 0.0002,
+      "loss": 0.7612,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4081319214291488,
+      "grad_norm": 0.5696909427642822,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4097486056098942,
+      "grad_norm": 0.7995436787605286,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4113652897906395,
+      "grad_norm": 0.7237521409988403,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4129819739713847,
+      "grad_norm": 0.744628369808197,
+      "learning_rate": 0.0002,
+      "loss": 0.7661,
+      "step": 8740
+    },
+    {
+      "epoch": 1.41459865815213,
+      "grad_norm": 0.6082926988601685,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4162153423328752,
+      "grad_norm": 0.5185243487358093,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4178320265136206,
+      "grad_norm": 0.5183082222938538,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4194487106943658,
+      "grad_norm": 0.7326041460037231,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 8780
+    },
+    {
+      "epoch": 1.421065394875111,
+      "grad_norm": 0.7174660563468933,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4226820790558565,
+      "grad_norm": 0.8080165982246399,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4242987632366018,
+      "grad_norm": 0.5061507821083069,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 8810
+    },
+    {
+      "epoch": 1.425915447417347,
+      "grad_norm": 0.801602840423584,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4275321315980922,
+      "grad_norm": 0.6150273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4291488157788377,
+      "grad_norm": 0.8786525726318359,
+      "learning_rate": 0.0002,
+      "loss": 0.7452,
+      "step": 8840
+    },
+    {
+      "epoch": 1.430765499959583,
+      "grad_norm": 0.6371538639068604,
+      "learning_rate": 0.0002,
+      "loss": 0.7257,
+      "step": 8850
+    },
+    {
+      "epoch": 1.4323821841403281,
+      "grad_norm": 0.6409295797348022,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4339988683210736,
+      "grad_norm": 0.6452359557151794,
+      "learning_rate": 0.0002,
+      "loss": 0.7891,
+      "step": 8870
+    },
+    {
+      "epoch": 1.4356155525018188,
+      "grad_norm": 0.5842334628105164,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 8880
+    },
+    {
+      "epoch": 1.437232236682564,
+      "grad_norm": 0.696761965751648,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4388489208633093,
+      "grad_norm": 0.6384600400924683,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4404656050440545,
+      "grad_norm": 0.5981136560440063,
+      "learning_rate": 0.0002,
+      "loss": 0.7049,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4420822892248,
+      "grad_norm": 0.6355637907981873,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4436989734055452,
+      "grad_norm": 0.6374830603599548,
+      "learning_rate": 0.0002,
+      "loss": 0.7653,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4453156575862904,
+      "grad_norm": 0.559013307094574,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 8940
+    },
+    {
+      "epoch": 1.446932341767036,
+      "grad_norm": 0.7289170026779175,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4485490259477811,
+      "grad_norm": 0.8649206757545471,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4501657101285264,
+      "grad_norm": 0.7664689421653748,
+      "learning_rate": 0.0002,
+      "loss": 0.7745,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4517823943092716,
+      "grad_norm": 0.7109952569007874,
+      "learning_rate": 0.0002,
+      "loss": 0.7431,
+      "step": 8980
+    },
+    {
+      "epoch": 1.453399078490017,
+      "grad_norm": 0.6312844753265381,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4550157626707623,
+      "grad_norm": 0.6616617441177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7467,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4566324468515075,
+      "grad_norm": 0.7384068965911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7518,
+      "step": 9010
+    },
+    {
+      "epoch": 1.458249131032253,
+      "grad_norm": 0.6549670100212097,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9020
+    },
+    {
+      "epoch": 1.4598658152129982,
+      "grad_norm": 0.6254119277000427,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4614824993937434,
+      "grad_norm": 0.6806328892707825,
+      "learning_rate": 0.0002,
+      "loss": 0.7645,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4630991835744886,
+      "grad_norm": 0.6803115010261536,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4647158677552339,
+      "grad_norm": 0.48529282212257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7264,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4663325519359793,
+      "grad_norm": 0.5995030999183655,
+      "learning_rate": 0.0002,
+      "loss": 0.7542,
+      "step": 9070
+    },
+    {
+      "epoch": 1.4679492361167246,
+      "grad_norm": 0.6005427837371826,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9080
+    },
+    {
+      "epoch": 1.46956592029747,
+      "grad_norm": 0.718564510345459,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4711826044782153,
+      "grad_norm": 0.7003577351570129,
+      "learning_rate": 0.0002,
+      "loss": 0.7089,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4727992886589605,
+      "grad_norm": 0.5888323783874512,
+      "learning_rate": 0.0002,
+      "loss": 0.8069,
+      "step": 9110
+    },
+    {
+      "epoch": 1.4744159728397057,
+      "grad_norm": 0.6417609453201294,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 9120
+    },
+    {
+      "epoch": 1.476032657020451,
+      "grad_norm": 0.572294294834137,
+      "learning_rate": 0.0002,
+      "loss": 0.7441,
+      "step": 9130
+    },
+    {
+      "epoch": 1.4776493412011964,
+      "grad_norm": 0.8200714588165283,
+      "learning_rate": 0.0002,
+      "loss": 0.8053,
+      "step": 9140
+    },
+    {
+      "epoch": 1.4792660253819416,
+      "grad_norm": 0.6343288421630859,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4808827095626869,
+      "grad_norm": 0.7017961144447327,
+      "learning_rate": 0.0002,
+      "loss": 0.7641,
+      "step": 9160
+    },
+    {
+      "epoch": 1.4824993937434323,
+      "grad_norm": 0.6202912926673889,
+      "learning_rate": 0.0002,
+      "loss": 0.7619,
+      "step": 9170
+    },
+    {
+      "epoch": 1.4841160779241775,
+      "grad_norm": 0.6677869558334351,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 9180
+    },
+    {
+      "epoch": 1.4857327621049228,
+      "grad_norm": 0.6052267551422119,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 9190
+    },
+    {
+      "epoch": 1.487349446285668,
+      "grad_norm": 0.6638872027397156,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 9200
+    },
+    {
+      "epoch": 1.4889661304664135,
+      "grad_norm": 0.6245523691177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 9210
+    },
+    {
+      "epoch": 1.4905828146471587,
+      "grad_norm": 0.5761767625808716,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 9220
+    },
+    {
+      "epoch": 1.492199498827904,
+      "grad_norm": 0.8175981640815735,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 9230
+    },
+    {
+      "epoch": 1.4938161830086494,
+      "grad_norm": 0.9144009947776794,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 9240
+    },
+    {
+      "epoch": 1.4954328671893946,
+      "grad_norm": 0.5742552876472473,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 9250
+    },
+    {
+      "epoch": 1.4970495513701398,
+      "grad_norm": 0.534534215927124,
+      "learning_rate": 0.0002,
+      "loss": 0.7418,
+      "step": 9260
+    },
+    {
+      "epoch": 1.498666235550885,
+      "grad_norm": 0.7836225032806396,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5002829197316303,
+      "grad_norm": 0.5292993187904358,
+      "learning_rate": 0.0002,
+      "loss": 0.7453,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5018996039123758,
+      "grad_norm": 0.8044071793556213,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9290
+    },
+    {
+      "epoch": 1.503516288093121,
+      "grad_norm": 0.6185805201530457,
+      "learning_rate": 0.0002,
+      "loss": 0.7229,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5051329722738664,
+      "grad_norm": 0.6093607544898987,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5067496564546117,
+      "grad_norm": 0.5891730189323425,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 9320
+    },
+    {
+      "epoch": 1.508366340635357,
+      "grad_norm": 0.6331129670143127,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5099830248161021,
+      "grad_norm": 0.7690958380699158,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5115997089968474,
+      "grad_norm": 0.6548877358436584,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5132163931775926,
+      "grad_norm": 0.6545143127441406,
+      "learning_rate": 0.0002,
+      "loss": 0.7408,
+      "step": 9360
+    },
+    {
+      "epoch": 1.514833077358338,
+      "grad_norm": 0.553247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.7432,
+      "step": 9370
+    },
+    {
+      "epoch": 1.5164497615390833,
+      "grad_norm": 0.8145074844360352,
+      "learning_rate": 0.0002,
+      "loss": 0.7265,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5180664457198287,
+      "grad_norm": 0.7636994123458862,
+      "learning_rate": 0.0002,
+      "loss": 0.7379,
+      "step": 9390
+    },
+    {
+      "epoch": 1.519683129900574,
+      "grad_norm": 0.6838982701301575,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5212998140813192,
+      "grad_norm": 0.8599441647529602,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5229164982620644,
+      "grad_norm": 0.7020329833030701,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5245331824428097,
+      "grad_norm": 0.6964772343635559,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5261498666235551,
+      "grad_norm": 0.6916600465774536,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9440
+    },
+    {
+      "epoch": 1.5277665508043003,
+      "grad_norm": 0.7282621264457703,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5293832349850458,
+      "grad_norm": 0.5363983511924744,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 9460
+    },
+    {
+      "epoch": 1.530999919165791,
+      "grad_norm": 0.6184861063957214,
+      "learning_rate": 0.0002,
+      "loss": 0.7154,
+      "step": 9470
+    },
+    {
+      "epoch": 1.5326166033465363,
+      "grad_norm": 0.5991285443305969,
+      "learning_rate": 0.0002,
+      "loss": 0.7837,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5342332875272815,
+      "grad_norm": 0.8176587820053101,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5358499717080267,
+      "grad_norm": 0.6473721861839294,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 9500
+    },
+    {
+      "epoch": 1.5374666558887722,
+      "grad_norm": 0.7319952845573425,
+      "learning_rate": 0.0002,
+      "loss": 0.7632,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5390833400695174,
+      "grad_norm": 0.702900230884552,
+      "learning_rate": 0.0002,
+      "loss": 0.7706,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5407000242502629,
+      "grad_norm": 0.7971600294113159,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 9530
+    },
+    {
+      "epoch": 1.542316708431008,
+      "grad_norm": 0.6527525186538696,
+      "learning_rate": 0.0002,
+      "loss": 0.7352,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5439333926117533,
+      "grad_norm": 0.5791676044464111,
+      "learning_rate": 0.0002,
+      "loss": 0.7425,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5455500767924986,
+      "grad_norm": 0.5619390606880188,
+      "learning_rate": 0.0002,
+      "loss": 0.7585,
+      "step": 9560
+    },
+    {
+      "epoch": 1.5471667609732438,
+      "grad_norm": 0.5701689124107361,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9570
+    },
+    {
+      "epoch": 1.548783445153989,
+      "grad_norm": 0.47549352049827576,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5504001293347345,
+      "grad_norm": 0.8730611205101013,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5520168135154797,
+      "grad_norm": 0.6842091083526611,
+      "learning_rate": 0.0002,
+      "loss": 0.798,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5536334976962252,
+      "grad_norm": 0.6675129532814026,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5552501818769704,
+      "grad_norm": 0.8173956274986267,
+      "learning_rate": 0.0002,
+      "loss": 0.7954,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5568668660577156,
+      "grad_norm": 0.724947452545166,
+      "learning_rate": 0.0002,
+      "loss": 0.7535,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5584835502384609,
+      "grad_norm": 0.6154758930206299,
+      "learning_rate": 0.0002,
+      "loss": 0.7738,
+      "step": 9640
+    },
+    {
+      "epoch": 1.560100234419206,
+      "grad_norm": 0.6072008013725281,
+      "learning_rate": 0.0002,
+      "loss": 0.7568,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5617169185999515,
+      "grad_norm": 0.659010648727417,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 9660
+    },
+    {
+      "epoch": 1.5633336027806968,
+      "grad_norm": 0.65857994556427,
+      "learning_rate": 0.0002,
+      "loss": 0.673,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5649502869614422,
+      "grad_norm": 0.5914267301559448,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5665669711421875,
+      "grad_norm": 0.6248020529747009,
+      "learning_rate": 0.0002,
+      "loss": 0.7414,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5681836553229327,
+      "grad_norm": 0.7147795557975769,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 9700
+    },
+    {
+      "epoch": 1.569800339503678,
+      "grad_norm": 0.7076232433319092,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 9710
+    },
+    {
+      "epoch": 1.5714170236844232,
+      "grad_norm": 0.6217400431632996,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5730337078651684,
+      "grad_norm": 0.6709911227226257,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 9730
+    },
+    {
+      "epoch": 1.5746503920459138,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 9740
+    },
+    {
+      "epoch": 1.576267076226659,
+      "grad_norm": 0.6241145730018616,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 9750
+    },
+    {
+      "epoch": 1.5778837604074045,
+      "grad_norm": 0.4960934817790985,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 9760
+    },
+    {
+      "epoch": 1.5795004445881498,
+      "grad_norm": 0.6593309640884399,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 9770
+    },
+    {
+      "epoch": 1.581117128768895,
+      "grad_norm": 0.5814042091369629,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 9780
+    },
+    {
+      "epoch": 1.5827338129496402,
+      "grad_norm": 0.5936070680618286,
+      "learning_rate": 0.0002,
+      "loss": 0.7109,
+      "step": 9790
+    },
+    {
+      "epoch": 1.5843504971303854,
+      "grad_norm": 0.6454403400421143,
+      "learning_rate": 0.0002,
+      "loss": 0.7769,
+      "step": 9800
+    },
+    {
+      "epoch": 1.585967181311131,
+      "grad_norm": 0.7612107992172241,
+      "learning_rate": 0.0002,
+      "loss": 0.7677,
+      "step": 9810
+    },
+    {
+      "epoch": 1.5875838654918761,
+      "grad_norm": 0.6494482755661011,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5892005496726216,
+      "grad_norm": 0.7825694680213928,
+      "learning_rate": 0.0002,
+      "loss": 0.7569,
+      "step": 9830
+    },
+    {
+      "epoch": 1.5908172338533668,
+      "grad_norm": 0.6757757663726807,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 9840
+    },
+    {
+      "epoch": 1.592433918034112,
+      "grad_norm": 0.7105609178543091,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 9850
+    },
+    {
+      "epoch": 1.5940506022148573,
+      "grad_norm": 0.7596991062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 9860
+    },
+    {
+      "epoch": 1.5956672863956025,
+      "grad_norm": 0.5681525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7108,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5972839705763477,
+      "grad_norm": 0.6090980768203735,
+      "learning_rate": 0.0002,
+      "loss": 0.7811,
+      "step": 9880
+    },
+    {
+      "epoch": 1.5989006547570932,
+      "grad_norm": 0.6271613240242004,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6005173389378387,
+      "grad_norm": 0.7656369805335999,
+      "learning_rate": 0.0002,
+      "loss": 0.7419,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6021340231185839,
+      "grad_norm": 0.7504446506500244,
+      "learning_rate": 0.0002,
+      "loss": 0.7336,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6037507072993291,
+      "grad_norm": 0.659656286239624,
+      "learning_rate": 0.0002,
+      "loss": 0.7479,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6053673914800743,
+      "grad_norm": 0.6006826162338257,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6069840756608196,
+      "grad_norm": 0.7872757911682129,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6086007598415648,
+      "grad_norm": 0.5545852780342102,
+      "learning_rate": 0.0002,
+      "loss": 0.768,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6102174440223103,
+      "grad_norm": 0.7429468631744385,
+      "learning_rate": 0.0002,
+      "loss": 0.8064,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6118341282030555,
+      "grad_norm": 0.6873556971549988,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 9970
+    },
+    {
+      "epoch": 1.613450812383801,
+      "grad_norm": 0.5874287486076355,
+      "learning_rate": 0.0002,
+      "loss": 0.7324,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6150674965645462,
+      "grad_norm": 0.6039386987686157,
+      "learning_rate": 0.0002,
+      "loss": 0.7141,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6166841807452914,
+      "grad_norm": 0.6233575940132141,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6183008649260366,
+      "grad_norm": 0.7676448225975037,
+      "learning_rate": 0.0002,
+      "loss": 0.7602,
+      "step": 10010
+    },
+    {
+      "epoch": 1.6199175491067819,
+      "grad_norm": 0.6565698385238647,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6215342332875273,
+      "grad_norm": 0.6787590384483337,
+      "learning_rate": 0.0002,
+      "loss": 0.7104,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6231509174682726,
+      "grad_norm": 0.6137678027153015,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 10040
+    },
+    {
+      "epoch": 1.624767601649018,
+      "grad_norm": 0.5236800312995911,
+      "learning_rate": 0.0002,
+      "loss": 0.7646,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6263842858297632,
+      "grad_norm": 0.7626367807388306,
+      "learning_rate": 0.0002,
+      "loss": 0.7437,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6280009700105085,
+      "grad_norm": 0.5657260417938232,
+      "learning_rate": 0.0002,
+      "loss": 0.7273,
+      "step": 10070
+    },
+    {
+      "epoch": 1.6296176541912537,
+      "grad_norm": 0.4913991391658783,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 10080
+    },
+    {
+      "epoch": 1.631234338371999,
+      "grad_norm": 0.7715556621551514,
+      "learning_rate": 0.0002,
+      "loss": 0.7596,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6328510225527442,
+      "grad_norm": 0.6509000062942505,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6344677067334896,
+      "grad_norm": 0.6215850114822388,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6360843909142349,
+      "grad_norm": 0.6956844329833984,
+      "learning_rate": 0.0002,
+      "loss": 0.7705,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6377010750949803,
+      "grad_norm": 0.6111597418785095,
+      "learning_rate": 0.0002,
+      "loss": 0.7129,
+      "step": 10130
+    },
+    {
+      "epoch": 1.6393177592757255,
+      "grad_norm": 0.6518288850784302,
+      "learning_rate": 0.0002,
+      "loss": 0.6955,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6409344434564708,
+      "grad_norm": 0.6914522051811218,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10150
+    },
+    {
+      "epoch": 1.642551127637216,
+      "grad_norm": 0.63785719871521,
+      "learning_rate": 0.0002,
+      "loss": 0.7295,
+      "step": 10160
+    },
+    {
+      "epoch": 1.6441678118179612,
+      "grad_norm": 0.6379287838935852,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6457844959987067,
+      "grad_norm": 0.6793403029441833,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 10180
+    },
+    {
+      "epoch": 1.647401180179452,
+      "grad_norm": 0.6099132895469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6490178643601974,
+      "grad_norm": 0.5869854092597961,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6506345485409426,
+      "grad_norm": 0.7716999053955078,
+      "learning_rate": 0.0002,
+      "loss": 0.8308,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6522512327216878,
+      "grad_norm": 0.6854110360145569,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 10220
+    },
+    {
+      "epoch": 1.653867916902433,
+      "grad_norm": 0.6957170367240906,
+      "learning_rate": 0.0002,
+      "loss": 0.782,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6554846010831783,
+      "grad_norm": 0.6932903528213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6571012852639235,
+      "grad_norm": 0.7713165283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10250
+    },
+    {
+      "epoch": 1.658717969444669,
+      "grad_norm": 0.7455793619155884,
+      "learning_rate": 0.0002,
+      "loss": 0.7099,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6603346536254144,
+      "grad_norm": 0.5464168190956116,
+      "learning_rate": 0.0002,
+      "loss": 0.7524,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6619513378061597,
+      "grad_norm": 0.6782926321029663,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 10280
+    },
+    {
+      "epoch": 1.663568021986905,
+      "grad_norm": 0.7962649464607239,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6651847061676501,
+      "grad_norm": 0.6814526319503784,
+      "learning_rate": 0.0002,
+      "loss": 0.7142,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6668013903483954,
+      "grad_norm": 0.656895101070404,
+      "learning_rate": 0.0002,
+      "loss": 0.7285,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6684180745291406,
+      "grad_norm": 0.6085672378540039,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 10320
+    },
+    {
+      "epoch": 1.670034758709886,
+      "grad_norm": 0.585508406162262,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 10330
+    },
+    {
+      "epoch": 1.6716514428906313,
+      "grad_norm": 0.6930184364318848,
+      "learning_rate": 0.0002,
+      "loss": 0.7604,
+      "step": 10340
+    },
+    {
+      "epoch": 1.6732681270713767,
+      "grad_norm": 0.575663149356842,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10350
+    },
+    {
+      "epoch": 1.674884811252122,
+      "grad_norm": 0.582502543926239,
+      "learning_rate": 0.0002,
+      "loss": 0.7198,
+      "step": 10360
+    },
+    {
+      "epoch": 1.6765014954328672,
+      "grad_norm": 0.5668916702270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 10370
+    },
+    {
+      "epoch": 1.6781181796136124,
+      "grad_norm": 0.6070065498352051,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10380
+    },
+    {
+      "epoch": 1.6797348637943577,
+      "grad_norm": 0.6141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.7939,
+      "step": 10390
+    },
+    {
+      "epoch": 1.6813515479751031,
+      "grad_norm": 0.8359124064445496,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 10400
+    },
+    {
+      "epoch": 1.6829682321558483,
+      "grad_norm": 0.5378185510635376,
+      "learning_rate": 0.0002,
+      "loss": 0.7488,
+      "step": 10410
+    },
+    {
+      "epoch": 1.6845849163365938,
+      "grad_norm": 0.6959536075592041,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 10420
+    },
+    {
+      "epoch": 1.686201600517339,
+      "grad_norm": 0.6514357328414917,
+      "learning_rate": 0.0002,
+      "loss": 0.7872,
+      "step": 10430
+    },
+    {
+      "epoch": 1.6878182846980843,
+      "grad_norm": 0.7706646919250488,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10440
+    },
+    {
+      "epoch": 1.6894349688788295,
+      "grad_norm": 0.6183337569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.7673,
+      "step": 10450
+    },
+    {
+      "epoch": 1.6910516530595747,
+      "grad_norm": 0.6123278141021729,
+      "learning_rate": 0.0002,
+      "loss": 0.7566,
+      "step": 10460
+    },
+    {
+      "epoch": 1.69266833724032,
+      "grad_norm": 0.6894851326942444,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10470
+    },
+    {
+      "epoch": 1.6942850214210654,
+      "grad_norm": 0.7497312426567078,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 10480
+    },
+    {
+      "epoch": 1.6959017056018106,
+      "grad_norm": 0.5968214273452759,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10490
+    },
+    {
+      "epoch": 1.697518389782556,
+      "grad_norm": 0.6747927069664001,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 10500
+    },
+    {
+      "epoch": 1.6991350739633013,
+      "grad_norm": 0.5708310008049011,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7007517581440466,
+      "grad_norm": 0.606526792049408,
+      "learning_rate": 0.0002,
+      "loss": 0.7385,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7023684423247918,
+      "grad_norm": 0.662011981010437,
+      "learning_rate": 0.0002,
+      "loss": 0.7204,
+      "step": 10530
+    },
+    {
+      "epoch": 1.703985126505537,
+      "grad_norm": 0.7583045363426208,
+      "learning_rate": 0.0002,
+      "loss": 0.7999,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7056018106862825,
+      "grad_norm": 0.721632182598114,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7072184948670277,
+      "grad_norm": 0.6107715368270874,
+      "learning_rate": 0.0002,
+      "loss": 0.7407,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7088351790477732,
+      "grad_norm": 0.6652471423149109,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7104518632285184,
+      "grad_norm": 0.6308087110519409,
+      "learning_rate": 0.0002,
+      "loss": 0.7767,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7120685474092636,
+      "grad_norm": 0.5464386940002441,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7136852315900089,
+      "grad_norm": 0.6558911204338074,
+      "learning_rate": 0.0002,
+      "loss": 0.7063,
+      "step": 10600
+    },
+    {
+      "epoch": 1.715301915770754,
+      "grad_norm": 0.5665024518966675,
+      "learning_rate": 0.0002,
+      "loss": 0.7126,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7169185999514993,
+      "grad_norm": 0.7888094186782837,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7185352841322448,
+      "grad_norm": 0.7084909081459045,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 10630
+    },
+    {
+      "epoch": 1.7201519683129902,
+      "grad_norm": 0.7982324361801147,
+      "learning_rate": 0.0002,
+      "loss": 0.7557,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7217686524937355,
+      "grad_norm": 0.6418732404708862,
+      "learning_rate": 0.0002,
+      "loss": 0.7345,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7233853366744807,
+      "grad_norm": 0.7636681795120239,
+      "learning_rate": 0.0002,
+      "loss": 0.7734,
+      "step": 10660
+    },
+    {
+      "epoch": 1.725002020855226,
+      "grad_norm": 0.5646875500679016,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7266187050359711,
+      "grad_norm": 0.5231260657310486,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7282353892167164,
+      "grad_norm": 0.7635011672973633,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7298520733974618,
+      "grad_norm": 0.7518259286880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 10700
+    },
+    {
+      "epoch": 1.731468757578207,
+      "grad_norm": 0.7295602560043335,
+      "learning_rate": 0.0002,
+      "loss": 0.751,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7330854417589525,
+      "grad_norm": 0.6984632015228271,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7347021259396977,
+      "grad_norm": 0.6198219060897827,
+      "learning_rate": 0.0002,
+      "loss": 0.7921,
+      "step": 10730
+    },
+    {
+      "epoch": 1.736318810120443,
+      "grad_norm": 0.6957576274871826,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7379354943011882,
+      "grad_norm": 0.6430263519287109,
+      "learning_rate": 0.0002,
+      "loss": 0.7917,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7395521784819334,
+      "grad_norm": 0.6134995222091675,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 10760
+    },
+    {
+      "epoch": 1.741168862662679,
+      "grad_norm": 0.7209452986717224,
+      "learning_rate": 0.0002,
+      "loss": 0.7584,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7427855468434241,
+      "grad_norm": 0.6735447645187378,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7444022310241696,
+      "grad_norm": 0.5605693459510803,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 10790
+    },
+    {
+      "epoch": 1.7460189152049148,
+      "grad_norm": 0.6882363557815552,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 10800
+    },
+    {
+      "epoch": 1.74763559938566,
+      "grad_norm": 0.6386259198188782,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7492522835664053,
+      "grad_norm": 0.6529015302658081,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7508689677471505,
+      "grad_norm": 0.5664082765579224,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7524856519278957,
+      "grad_norm": 0.7532684206962585,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7541023361086412,
+      "grad_norm": 0.77171391248703,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7557190202893864,
+      "grad_norm": 0.7255431413650513,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7573357044701319,
+      "grad_norm": 0.763083279132843,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10870
+    },
+    {
+      "epoch": 1.758952388650877,
+      "grad_norm": 0.6042402982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7605690728316223,
+      "grad_norm": 0.7642518281936646,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7621857570123676,
+      "grad_norm": 0.6347904801368713,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7638024411931128,
+      "grad_norm": 0.5371627807617188,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7654191253738583,
+      "grad_norm": 0.6840225458145142,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7670358095546035,
+      "grad_norm": 0.5288469195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10930
+    },
+    {
+      "epoch": 1.768652493735349,
+      "grad_norm": 0.69020676612854,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 10940
+    },
+    {
+      "epoch": 1.7702691779160942,
+      "grad_norm": 0.5943242311477661,
+      "learning_rate": 0.0002,
+      "loss": 0.7468,
+      "step": 10950
+    },
+    {
+      "epoch": 1.7718858620968394,
+      "grad_norm": 0.5616418123245239,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 10960
+    },
+    {
+      "epoch": 1.7735025462775846,
+      "grad_norm": 0.7209470868110657,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 10970
+    },
+    {
+      "epoch": 1.7751192304583299,
+      "grad_norm": 0.6657957434654236,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 10980
+    },
+    {
+      "epoch": 1.776735914639075,
+      "grad_norm": 0.6469064950942993,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 10990
+    },
+    {
+      "epoch": 1.7783525988198206,
+      "grad_norm": 0.6615678071975708,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 11000
+    },
+    {
+      "epoch": 1.779969283000566,
+      "grad_norm": 0.6722439527511597,
+      "learning_rate": 0.0002,
+      "loss": 0.747,
+      "step": 11010
+    },
+    {
+      "epoch": 1.7815859671813112,
+      "grad_norm": 0.634136974811554,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 11020
+    },
+    {
+      "epoch": 1.7832026513620565,
+      "grad_norm": 0.6024377346038818,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 11030
+    },
+    {
+      "epoch": 1.7848193355428017,
+      "grad_norm": 0.6909403800964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 11040
+    },
+    {
+      "epoch": 1.786436019723547,
+      "grad_norm": 0.7148767709732056,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 11050
+    },
+    {
+      "epoch": 1.7880527039042922,
+      "grad_norm": 0.7442979216575623,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 11060
+    },
+    {
+      "epoch": 1.7896693880850376,
+      "grad_norm": 0.6830431818962097,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 11070
+    },
+    {
+      "epoch": 1.7912860722657828,
+      "grad_norm": 0.9172667264938354,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 11080
+    },
+    {
+      "epoch": 1.7929027564465283,
+      "grad_norm": 0.6799490451812744,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 11090
+    },
+    {
+      "epoch": 1.7945194406272735,
+      "grad_norm": 0.7617024779319763,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 11100
+    },
+    {
+      "epoch": 1.7961361248080188,
+      "grad_norm": 0.7701810002326965,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 11110
+    },
+    {
+      "epoch": 1.797752808988764,
+      "grad_norm": 0.7454385757446289,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11120
+    },
+    {
+      "epoch": 1.7993694931695092,
+      "grad_norm": 0.6121436953544617,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8009861773502547,
+      "grad_norm": 0.6237571835517883,
+      "learning_rate": 0.0002,
+      "loss": 0.7305,
+      "step": 11140
+    },
+    {
+      "epoch": 1.802602861531,
+      "grad_norm": 0.6818515658378601,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 11150
+    },
+    {
+      "epoch": 1.8042195457117454,
+      "grad_norm": 0.7768308520317078,
+      "learning_rate": 0.0002,
+      "loss": 0.6876,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8058362298924906,
+      "grad_norm": 0.6875537633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 11170
+    },
+    {
+      "epoch": 1.8074529140732358,
+      "grad_norm": 0.7950584888458252,
+      "learning_rate": 0.0002,
+      "loss": 0.761,
+      "step": 11180
+    },
+    {
+      "epoch": 1.809069598253981,
+      "grad_norm": 0.8210248351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8106862824347263,
+      "grad_norm": 0.6674110889434814,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8123029666154715,
+      "grad_norm": 0.6261674761772156,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 11210
+    },
+    {
+      "epoch": 1.813919650796217,
+      "grad_norm": 0.6484741568565369,
+      "learning_rate": 0.0002,
+      "loss": 0.7122,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8155363349769622,
+      "grad_norm": 0.6231244206428528,
+      "learning_rate": 0.0002,
+      "loss": 0.7718,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8171530191577077,
+      "grad_norm": 0.7243146896362305,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11240
+    },
+    {
+      "epoch": 1.818769703338453,
+      "grad_norm": 0.6776193380355835,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8203863875191981,
+      "grad_norm": 0.5973618030548096,
+      "learning_rate": 0.0002,
+      "loss": 0.7317,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8220030716999434,
+      "grad_norm": 0.6451361179351807,
+      "learning_rate": 0.0002,
+      "loss": 0.7961,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8236197558806886,
+      "grad_norm": 0.5963068008422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 11280
+    },
+    {
+      "epoch": 1.825236440061434,
+      "grad_norm": 0.536902129650116,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8268531242421793,
+      "grad_norm": 0.6993787288665771,
+      "learning_rate": 0.0002,
+      "loss": 0.708,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8284698084229247,
+      "grad_norm": 0.6135255098342896,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11310
+    },
+    {
+      "epoch": 1.83008649260367,
+      "grad_norm": 0.6057423949241638,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8317031767844152,
+      "grad_norm": 0.6598812341690063,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8333198609651604,
+      "grad_norm": 0.6075948476791382,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 11340
+    },
+    {
+      "epoch": 1.8349365451459057,
+      "grad_norm": 0.7065447568893433,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8365532293266509,
+      "grad_norm": 0.680526614189148,
+      "learning_rate": 0.0002,
+      "loss": 0.7365,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8381699135073963,
+      "grad_norm": 0.6356695294380188,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8397865976881416,
+      "grad_norm": 0.6399052143096924,
+      "learning_rate": 0.0002,
+      "loss": 0.721,
+      "step": 11380
+    },
+    {
+      "epoch": 1.841403281868887,
+      "grad_norm": 0.6125704050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 11390
+    },
+    {
+      "epoch": 1.8430199660496323,
+      "grad_norm": 0.7124643325805664,
+      "learning_rate": 0.0002,
+      "loss": 0.755,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8446366502303775,
+      "grad_norm": 0.6099604964256287,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8462533344111227,
+      "grad_norm": 0.7338208556175232,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 11420
+    },
+    {
+      "epoch": 1.847870018591868,
+      "grad_norm": 0.7534668445587158,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8494867027726134,
+      "grad_norm": 0.6135470271110535,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8511033869533586,
+      "grad_norm": 0.6229309439659119,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 11450
+    },
+    {
+      "epoch": 1.852720071134104,
+      "grad_norm": 0.706423282623291,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 11460
+    },
+    {
+      "epoch": 1.8543367553148493,
+      "grad_norm": 0.5460049510002136,
+      "learning_rate": 0.0002,
+      "loss": 0.7411,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8559534394955945,
+      "grad_norm": 0.6616711020469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7416,
+      "step": 11480
+    },
+    {
+      "epoch": 1.8575701236763398,
+      "grad_norm": 0.6372783184051514,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 11490
+    },
+    {
+      "epoch": 1.859186807857085,
+      "grad_norm": 0.7162668108940125,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8608034920378305,
+      "grad_norm": 0.6605209708213806,
+      "learning_rate": 0.0002,
+      "loss": 0.7747,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8624201762185757,
+      "grad_norm": 0.6933956742286682,
+      "learning_rate": 0.0002,
+      "loss": 0.7258,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8640368603993211,
+      "grad_norm": 0.6582090854644775,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8656535445800664,
+      "grad_norm": 0.6416500806808472,
+      "learning_rate": 0.0002,
+      "loss": 0.7313,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8672702287608116,
+      "grad_norm": 0.5434312224388123,
+      "learning_rate": 0.0002,
+      "loss": 0.7372,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8688869129415568,
+      "grad_norm": 0.6827567219734192,
+      "learning_rate": 0.0002,
+      "loss": 0.7635,
+      "step": 11560
+    },
+    {
+      "epoch": 1.870503597122302,
+      "grad_norm": 0.7354370951652527,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 11570
+    },
+    {
+      "epoch": 1.8721202813030473,
+      "grad_norm": 0.590372622013092,
+      "learning_rate": 0.0002,
+      "loss": 0.7526,
+      "step": 11580
+    },
+    {
+      "epoch": 1.8737369654837928,
+      "grad_norm": 0.853183925151825,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 11590
+    },
+    {
+      "epoch": 1.875353649664538,
+      "grad_norm": 0.822678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 11600
+    },
+    {
+      "epoch": 1.8769703338452834,
+      "grad_norm": 0.6591550707817078,
+      "learning_rate": 0.0002,
+      "loss": 0.7427,
+      "step": 11610
+    },
+    {
+      "epoch": 1.8785870180260287,
+      "grad_norm": 0.7475301623344421,
+      "learning_rate": 0.0002,
+      "loss": 0.7054,
+      "step": 11620
+    },
+    {
+      "epoch": 1.880203702206774,
+      "grad_norm": 0.6390765309333801,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 11630
+    },
+    {
+      "epoch": 1.8818203863875191,
+      "grad_norm": 0.6589758992195129,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 11640
+    },
+    {
+      "epoch": 1.8834370705682644,
+      "grad_norm": 0.6765508651733398,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 11650
+    },
+    {
+      "epoch": 1.8850537547490098,
+      "grad_norm": 0.6527857780456543,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 11660
+    },
+    {
+      "epoch": 1.886670438929755,
+      "grad_norm": 0.6642923951148987,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 11670
+    },
+    {
+      "epoch": 1.8882871231105005,
+      "grad_norm": 0.6945584416389465,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 11680
+    },
+    {
+      "epoch": 1.8899038072912457,
+      "grad_norm": 0.694018542766571,
+      "learning_rate": 0.0002,
+      "loss": 0.7711,
+      "step": 11690
+    },
+    {
+      "epoch": 1.891520491471991,
+      "grad_norm": 0.7237417101860046,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 11700
+    },
+    {
+      "epoch": 1.8931371756527362,
+      "grad_norm": 0.7401309609413147,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 11710
+    },
+    {
+      "epoch": 1.8947538598334814,
+      "grad_norm": 0.6537784337997437,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 11720
+    },
+    {
+      "epoch": 1.8963705440142267,
+      "grad_norm": 0.7398539185523987,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 11730
+    },
+    {
+      "epoch": 1.8979872281949721,
+      "grad_norm": 0.6696075797080994,
+      "learning_rate": 0.0002,
+      "loss": 0.7561,
+      "step": 11740
+    },
+    {
+      "epoch": 1.8996039123757174,
+      "grad_norm": 0.6014142036437988,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9012205965564628,
+      "grad_norm": 0.7023524641990662,
+      "learning_rate": 0.0002,
+      "loss": 0.7714,
+      "step": 11760
+    },
+    {
+      "epoch": 1.902837280737208,
+      "grad_norm": 0.739973783493042,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9044539649179533,
+      "grad_norm": 0.5576770901679993,
+      "learning_rate": 0.0002,
+      "loss": 0.7848,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9060706490986985,
+      "grad_norm": 0.6907393932342529,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 11790
+    },
+    {
+      "epoch": 1.9076873332794437,
+      "grad_norm": 0.6934581995010376,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9093040174601892,
+      "grad_norm": 0.591774582862854,
+      "learning_rate": 0.0002,
+      "loss": 0.7199,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9109207016409344,
+      "grad_norm": 0.6249791383743286,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11820
+    },
+    {
+      "epoch": 1.9125373858216799,
+      "grad_norm": 0.6755744218826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 11830
+    },
+    {
+      "epoch": 1.914154070002425,
+      "grad_norm": 0.7286285161972046,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9157707541831703,
+      "grad_norm": 0.7867850065231323,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9173874383639156,
+      "grad_norm": 0.6283972859382629,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11860
+    },
+    {
+      "epoch": 1.9190041225446608,
+      "grad_norm": 0.605823814868927,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 11870
+    },
+    {
+      "epoch": 1.920620806725406,
+      "grad_norm": 0.5927976965904236,
+      "learning_rate": 0.0002,
+      "loss": 0.6598,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9222374909061515,
+      "grad_norm": 0.5974002480506897,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 11890
+    },
+    {
+      "epoch": 1.923854175086897,
+      "grad_norm": 0.7091866135597229,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9254708592676422,
+      "grad_norm": 0.72496497631073,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9270875434483874,
+      "grad_norm": 0.6131896376609802,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9287042276291326,
+      "grad_norm": 0.6556436419487,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9303209118098779,
+      "grad_norm": 0.622932493686676,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 11940
+    },
+    {
+      "epoch": 1.931937595990623,
+      "grad_norm": 0.6618631482124329,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9335542801713685,
+      "grad_norm": 0.630966305732727,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 11960
+    },
+    {
+      "epoch": 1.9351709643521138,
+      "grad_norm": 0.6336734890937805,
+      "learning_rate": 0.0002,
+      "loss": 0.6854,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9367876485328592,
+      "grad_norm": 0.655403196811676,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9384043327136045,
+      "grad_norm": 0.5640574097633362,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 11990
+    },
+    {
+      "epoch": 1.9400210168943497,
+      "grad_norm": 0.6322951316833496,
+      "learning_rate": 0.0002,
+      "loss": 0.7289,
+      "step": 12000
+    },
+    {
+      "epoch": 1.941637701075095,
+      "grad_norm": 0.615703821182251,
+      "learning_rate": 0.0002,
+      "loss": 0.7627,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9432543852558402,
+      "grad_norm": 0.6487536430358887,
+      "learning_rate": 0.0002,
+      "loss": 0.786,
+      "step": 12020
+    },
+    {
+      "epoch": 1.9448710694365856,
+      "grad_norm": 0.9209630489349365,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 12030
+    },
+    {
+      "epoch": 1.9464877536173308,
+      "grad_norm": 0.67485511302948,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9481044377980763,
+      "grad_norm": 0.6831230521202087,
+      "learning_rate": 0.0002,
+      "loss": 0.7551,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9497211219788215,
+      "grad_norm": 0.6578302383422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7546,
+      "step": 12060
+    },
+    {
+      "epoch": 1.9513378061595668,
+      "grad_norm": 0.9975938200950623,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12070
+    },
+    {
+      "epoch": 1.952954490340312,
+      "grad_norm": 0.6637365221977234,
+      "learning_rate": 0.0002,
+      "loss": 0.7952,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9545711745210572,
+      "grad_norm": 0.605707049369812,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9561878587018025,
+      "grad_norm": 0.6584440469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7768,
+      "step": 12100
+    },
+    {
+      "epoch": 1.957804542882548,
+      "grad_norm": 0.6070835590362549,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9594212270632931,
+      "grad_norm": 0.7862601280212402,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9610379112440386,
+      "grad_norm": 0.8175255060195923,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 12130
+    },
+    {
+      "epoch": 1.9626545954247838,
+      "grad_norm": 0.5648472905158997,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 12140
+    },
+    {
+      "epoch": 1.964271279605529,
+      "grad_norm": 0.6591973304748535,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9658879637862743,
+      "grad_norm": 0.5960676074028015,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9675046479670195,
+      "grad_norm": 0.7272544503211975,
+      "learning_rate": 0.0002,
+      "loss": 0.7254,
+      "step": 12170
+    },
+    {
+      "epoch": 1.969121332147765,
+      "grad_norm": 0.7176699042320251,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9707380163285102,
+      "grad_norm": 0.6927123665809631,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9723547005092557,
+      "grad_norm": 0.5536034107208252,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 12200
+    },
+    {
+      "epoch": 1.9739713846900009,
+      "grad_norm": 0.8348390460014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 12210
+    },
+    {
+      "epoch": 1.9755880688707461,
+      "grad_norm": 0.6591181755065918,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 12220
+    },
+    {
+      "epoch": 1.9772047530514913,
+      "grad_norm": 1.0624109506607056,
+      "learning_rate": 0.0002,
+      "loss": 0.763,
+      "step": 12230
+    },
+    {
+      "epoch": 1.9788214372322366,
+      "grad_norm": 0.9265586137771606,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 12240
+    },
+    {
+      "epoch": 1.9804381214129818,
+      "grad_norm": 0.5998196005821228,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 12250
+    },
+    {
+      "epoch": 1.9820548055937273,
+      "grad_norm": 0.6960851550102234,
+      "learning_rate": 0.0002,
+      "loss": 0.7225,
+      "step": 12260
+    },
+    {
+      "epoch": 1.9836714897744727,
+      "grad_norm": 0.7674502730369568,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 12270
+    },
+    {
+      "epoch": 1.985288173955218,
+      "grad_norm": 0.6407275795936584,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 12280
+    },
+    {
+      "epoch": 1.9869048581359632,
+      "grad_norm": 0.6673079133033752,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 12290
+    },
+    {
+      "epoch": 1.9885215423167084,
+      "grad_norm": 0.6989844441413879,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 12300
+    },
+    {
+      "epoch": 1.9901382264974536,
+      "grad_norm": 0.7564442157745361,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 12310
+    },
+    {
+      "epoch": 1.9917549106781989,
+      "grad_norm": 0.6385478973388672,
+      "learning_rate": 0.0002,
+      "loss": 0.7719,
+      "step": 12320
+    },
+    {
+      "epoch": 1.9933715948589443,
+      "grad_norm": 0.7193717956542969,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 12330
+    },
+    {
+      "epoch": 1.9949882790396896,
+      "grad_norm": 0.7987112402915955,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 12340
+    },
+    {
+      "epoch": 1.996604963220435,
+      "grad_norm": 0.7260826826095581,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 12350
+    },
+    {
+      "epoch": 1.9982216474011802,
+      "grad_norm": 0.7968255281448364,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 12360
+    },
+    {
+      "epoch": 1.9998383315819255,
+      "grad_norm": 0.6893062591552734,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1044032573699951,
+      "eval_runtime": 122.1508,
+      "eval_samples_per_second": 6.001,
+      "eval_steps_per_second": 0.753,
+      "step": 12371
+    },
+    {
+      "epoch": 2.0014550157626707,
+      "grad_norm": 0.7775409817695618,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 12380
+    },
+    {
+      "epoch": 2.003071699943416,
+      "grad_norm": 0.76218581199646,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 12390
+    },
+    {
+      "epoch": 2.004688384124161,
+      "grad_norm": 0.5677764415740967,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 12400
+    },
+    {
+      "epoch": 2.006305068304907,
+      "grad_norm": 0.808442234992981,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 12410
+    },
+    {
+      "epoch": 2.007921752485652,
+      "grad_norm": 0.7144765257835388,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 12420
+    },
+    {
+      "epoch": 2.0095384366663973,
+      "grad_norm": 0.6914031505584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 12430
+    },
+    {
+      "epoch": 2.0111551208471425,
+      "grad_norm": 0.7581454515457153,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 12440
+    },
+    {
+      "epoch": 2.0127718050278878,
+      "grad_norm": 0.8388504981994629,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 12450
+    },
+    {
+      "epoch": 2.014388489208633,
+      "grad_norm": 0.6716406941413879,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 12460
+    },
+    {
+      "epoch": 2.0160051733893782,
+      "grad_norm": 0.898902416229248,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 12470
+    },
+    {
+      "epoch": 2.0176218575701235,
+      "grad_norm": 0.6432679891586304,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 12480
+    },
+    {
+      "epoch": 2.019238541750869,
+      "grad_norm": 0.8021109104156494,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12490
+    },
+    {
+      "epoch": 2.0208552259316144,
+      "grad_norm": 0.7039216756820679,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 12500
+    },
+    {
+      "epoch": 2.0224719101123596,
+      "grad_norm": 0.646531879901886,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12510
+    },
+    {
+      "epoch": 2.024088594293105,
+      "grad_norm": 0.783704400062561,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 12520
+    },
+    {
+      "epoch": 2.02570527847385,
+      "grad_norm": 0.8805046677589417,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12530
+    },
+    {
+      "epoch": 2.0273219626545953,
+      "grad_norm": 0.7289270758628845,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0289386468353405,
+      "grad_norm": 0.71653151512146,
+      "learning_rate": 0.0002,
+      "loss": 0.6663,
+      "step": 12550
+    },
+    {
+      "epoch": 2.030555331016086,
+      "grad_norm": 0.73281329870224,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 12560
+    },
+    {
+      "epoch": 2.0321720151968314,
+      "grad_norm": 0.6657090187072754,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 12570
+    },
+    {
+      "epoch": 2.0337886993775767,
+      "grad_norm": 0.8241133093833923,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 12580
+    },
+    {
+      "epoch": 2.035405383558322,
+      "grad_norm": 0.5834135413169861,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 12590
+    },
+    {
+      "epoch": 2.037022067739067,
+      "grad_norm": 0.84502112865448,
+      "learning_rate": 0.0002,
+      "loss": 0.6188,
+      "step": 12600
+    },
+    {
+      "epoch": 2.0386387519198124,
+      "grad_norm": 0.8952481746673584,
+      "learning_rate": 0.0002,
+      "loss": 0.6349,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0402554361005576,
+      "grad_norm": 0.7801461815834045,
+      "learning_rate": 0.0002,
+      "loss": 0.6923,
+      "step": 12620
+    },
+    {
+      "epoch": 2.041872120281303,
+      "grad_norm": 0.6788367033004761,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 12630
+    },
+    {
+      "epoch": 2.0434888044620485,
+      "grad_norm": 0.7241756319999695,
+      "learning_rate": 0.0002,
+      "loss": 0.6162,
+      "step": 12640
+    },
+    {
+      "epoch": 2.0451054886427937,
+      "grad_norm": 0.6933388113975525,
+      "learning_rate": 0.0002,
+      "loss": 0.655,
+      "step": 12650
+    },
+    {
+      "epoch": 2.046722172823539,
+      "grad_norm": 0.8029746413230896,
+      "learning_rate": 0.0002,
+      "loss": 0.6431,
+      "step": 12660
+    },
+    {
+      "epoch": 2.048338857004284,
+      "grad_norm": 0.946399986743927,
+      "learning_rate": 0.0002,
+      "loss": 0.7164,
+      "step": 12670
+    },
+    {
+      "epoch": 2.0499555411850294,
+      "grad_norm": 0.7072678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.638,
+      "step": 12680
+    },
+    {
+      "epoch": 2.0515722253657747,
+      "grad_norm": 0.6810618042945862,
+      "learning_rate": 0.0002,
+      "loss": 0.6487,
+      "step": 12690
+    },
+    {
+      "epoch": 2.05318890954652,
+      "grad_norm": 0.7661160230636597,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 12700
+    },
+    {
+      "epoch": 2.0548055937272656,
+      "grad_norm": 0.6350653767585754,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 12710
+    },
+    {
+      "epoch": 2.056422277908011,
+      "grad_norm": 0.861890971660614,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 12720
+    },
+    {
+      "epoch": 2.058038962088756,
+      "grad_norm": 0.6489875912666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 12730
+    },
+    {
+      "epoch": 2.0596556462695013,
+      "grad_norm": 0.8268506526947021,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 12740
+    },
+    {
+      "epoch": 2.0612723304502465,
+      "grad_norm": 0.607679545879364,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 12750
+    },
+    {
+      "epoch": 2.0628890146309917,
+      "grad_norm": 0.6754153370857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6649,
+      "step": 12760
+    },
+    {
+      "epoch": 2.064505698811737,
+      "grad_norm": 0.7263124585151672,
+      "learning_rate": 0.0002,
+      "loss": 0.6549,
+      "step": 12770
+    },
+    {
+      "epoch": 2.0661223829924826,
+      "grad_norm": 0.6986154317855835,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 12780
+    },
+    {
+      "epoch": 2.067739067173228,
+      "grad_norm": 0.7768576741218567,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 12790
+    },
+    {
+      "epoch": 2.069355751353973,
+      "grad_norm": 0.7546762824058533,
+      "learning_rate": 0.0002,
+      "loss": 0.677,
+      "step": 12800
+    },
+    {
+      "epoch": 2.0709724355347183,
+      "grad_norm": 0.7588880062103271,
+      "learning_rate": 0.0002,
+      "loss": 0.6485,
+      "step": 12810
+    },
+    {
+      "epoch": 2.0725891197154636,
+      "grad_norm": 0.7457242608070374,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12820
+    },
+    {
+      "epoch": 2.074205803896209,
+      "grad_norm": 0.6983516812324524,
+      "learning_rate": 0.0002,
+      "loss": 0.6489,
+      "step": 12830
+    },
+    {
+      "epoch": 2.075822488076954,
+      "grad_norm": 0.7950928807258606,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 12840
+    },
+    {
+      "epoch": 2.0774391722576993,
+      "grad_norm": 0.9248087406158447,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 12850
+    },
+    {
+      "epoch": 2.079055856438445,
+      "grad_norm": 0.7229493260383606,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 12860
+    },
+    {
+      "epoch": 2.08067254061919,
+      "grad_norm": 0.5710847973823547,
+      "learning_rate": 0.0002,
+      "loss": 0.6702,
+      "step": 12870
+    },
+    {
+      "epoch": 2.0822892247999354,
+      "grad_norm": 0.9580423831939697,
+      "learning_rate": 0.0002,
+      "loss": 0.6974,
+      "step": 12880
+    },
+    {
+      "epoch": 2.0839059089806806,
+      "grad_norm": 0.7399665713310242,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 12890
+    },
+    {
+      "epoch": 2.085522593161426,
+      "grad_norm": 0.7981410622596741,
+      "learning_rate": 0.0002,
+      "loss": 0.6993,
+      "step": 12900
+    },
+    {
+      "epoch": 2.087139277342171,
+      "grad_norm": 0.870759904384613,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 12910
+    },
+    {
+      "epoch": 2.0887559615229163,
+      "grad_norm": 0.7001481652259827,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 12920
+    },
+    {
+      "epoch": 2.090372645703662,
+      "grad_norm": 0.6745418310165405,
+      "learning_rate": 0.0002,
+      "loss": 0.6383,
+      "step": 12930
+    },
+    {
+      "epoch": 2.0919893298844072,
+      "grad_norm": 0.7739067673683167,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 12940
+    },
+    {
+      "epoch": 2.0936060140651525,
+      "grad_norm": 0.6742934584617615,
+      "learning_rate": 0.0002,
+      "loss": 0.6856,
+      "step": 12950
+    },
+    {
+      "epoch": 2.0952226982458977,
+      "grad_norm": 0.7270349860191345,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 12960
+    },
+    {
+      "epoch": 2.096839382426643,
+      "grad_norm": 0.7150624394416809,
+      "learning_rate": 0.0002,
+      "loss": 0.6783,
+      "step": 12970
+    },
+    {
+      "epoch": 2.098456066607388,
+      "grad_norm": 0.7734767198562622,
+      "learning_rate": 0.0002,
+      "loss": 0.6093,
+      "step": 12980
+    },
+    {
+      "epoch": 2.1000727507881334,
+      "grad_norm": 0.7618662118911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 12990
+    },
+    {
+      "epoch": 2.101689434968879,
+      "grad_norm": 0.6557944416999817,
+      "learning_rate": 0.0002,
+      "loss": 0.6707,
+      "step": 13000
+    },
+    {
+      "epoch": 2.1033061191496243,
+      "grad_norm": 0.8786448240280151,
+      "learning_rate": 0.0002,
+      "loss": 0.7268,
+      "step": 13010
+    },
+    {
+      "epoch": 2.1049228033303695,
+      "grad_norm": 0.6878724098205566,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 13020
+    },
+    {
+      "epoch": 2.1065394875111147,
+      "grad_norm": 0.822318971157074,
+      "learning_rate": 0.0002,
+      "loss": 0.6824,
+      "step": 13030
+    },
+    {
+      "epoch": 2.10815617169186,
+      "grad_norm": 0.831468939781189,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 13040
+    },
+    {
+      "epoch": 2.109772855872605,
+      "grad_norm": 0.7699505686759949,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 13050
+    },
+    {
+      "epoch": 2.1113895400533504,
+      "grad_norm": 0.7559016346931458,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1130062242340957,
+      "grad_norm": 0.6942209601402283,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 13070
+    },
+    {
+      "epoch": 2.1146229084148414,
+      "grad_norm": 0.6098947525024414,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 13080
+    },
+    {
+      "epoch": 2.1162395925955866,
+      "grad_norm": 0.6499016284942627,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 13090
+    },
+    {
+      "epoch": 2.117856276776332,
+      "grad_norm": 0.7719953060150146,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 13100
+    },
+    {
+      "epoch": 2.119472960957077,
+      "grad_norm": 0.6708134412765503,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1210896451378223,
+      "grad_norm": 0.8119585514068604,
+      "learning_rate": 0.0002,
+      "loss": 0.6056,
+      "step": 13120
+    },
+    {
+      "epoch": 2.1227063293185675,
+      "grad_norm": 0.6947157979011536,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 13130
+    },
+    {
+      "epoch": 2.1243230134993127,
+      "grad_norm": 0.8831837773323059,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 13140
+    },
+    {
+      "epoch": 2.1259396976800584,
+      "grad_norm": 0.7266910672187805,
+      "learning_rate": 0.0002,
+      "loss": 0.6997,
+      "step": 13150
+    },
+    {
+      "epoch": 2.1275563818608036,
+      "grad_norm": 0.8864351511001587,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 13160
+    },
+    {
+      "epoch": 2.129173066041549,
+      "grad_norm": 0.8104248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.6762,
+      "step": 13170
+    },
+    {
+      "epoch": 2.130789750222294,
+      "grad_norm": 0.6077079772949219,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 13180
+    },
+    {
+      "epoch": 2.1324064344030393,
+      "grad_norm": 0.6874213814735413,
+      "learning_rate": 0.0002,
+      "loss": 0.6572,
+      "step": 13190
+    },
+    {
+      "epoch": 2.1340231185837846,
+      "grad_norm": 0.7134367823600769,
+      "learning_rate": 0.0002,
+      "loss": 0.642,
+      "step": 13200
+    },
+    {
+      "epoch": 2.13563980276453,
+      "grad_norm": 0.6101235151290894,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 13210
+    },
+    {
+      "epoch": 2.137256486945275,
+      "grad_norm": 0.6042411923408508,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13220
+    },
+    {
+      "epoch": 2.1388731711260207,
+      "grad_norm": 0.914601743221283,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 13230
+    },
+    {
+      "epoch": 2.140489855306766,
+      "grad_norm": 0.7104284167289734,
+      "learning_rate": 0.0002,
+      "loss": 0.6513,
+      "step": 13240
+    },
+    {
+      "epoch": 2.142106539487511,
+      "grad_norm": 0.664395272731781,
+      "learning_rate": 0.0002,
+      "loss": 0.6607,
+      "step": 13250
+    },
+    {
+      "epoch": 2.1437232236682564,
+      "grad_norm": 0.6991241574287415,
+      "learning_rate": 0.0002,
+      "loss": 0.7211,
+      "step": 13260
+    },
+    {
+      "epoch": 2.1453399078490016,
+      "grad_norm": 0.5469560623168945,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 13270
+    },
+    {
+      "epoch": 2.146956592029747,
+      "grad_norm": 0.8454998135566711,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13280
+    },
+    {
+      "epoch": 2.148573276210492,
+      "grad_norm": 0.7088868618011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 13290
+    },
+    {
+      "epoch": 2.1501899603912378,
+      "grad_norm": 0.7002687454223633,
+      "learning_rate": 0.0002,
+      "loss": 0.6835,
+      "step": 13300
+    },
+    {
+      "epoch": 2.151806644571983,
+      "grad_norm": 0.7785214781761169,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 13310
+    },
+    {
+      "epoch": 2.1534233287527282,
+      "grad_norm": 0.8049132227897644,
+      "learning_rate": 0.0002,
+      "loss": 0.67,
+      "step": 13320
+    },
+    {
+      "epoch": 2.1550400129334735,
+      "grad_norm": 0.8062595129013062,
+      "learning_rate": 0.0002,
+      "loss": 0.6495,
+      "step": 13330
+    },
+    {
+      "epoch": 2.1566566971142187,
+      "grad_norm": 0.6208319067955017,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 13340
+    },
+    {
+      "epoch": 2.158273381294964,
+      "grad_norm": 0.7519655823707581,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 13350
+    },
+    {
+      "epoch": 2.159890065475709,
+      "grad_norm": 0.7645747065544128,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 13360
+    },
+    {
+      "epoch": 2.1615067496564544,
+      "grad_norm": 0.6847302913665771,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 13370
+    },
+    {
+      "epoch": 2.1631234338372,
+      "grad_norm": 0.8630441427230835,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 13380
+    },
+    {
+      "epoch": 2.1647401180179453,
+      "grad_norm": 0.7947702407836914,
+      "learning_rate": 0.0002,
+      "loss": 0.6742,
+      "step": 13390
+    },
+    {
+      "epoch": 2.1663568021986905,
+      "grad_norm": 0.6836977005004883,
+      "learning_rate": 0.0002,
+      "loss": 0.7206,
+      "step": 13400
+    },
+    {
+      "epoch": 2.1679734863794358,
+      "grad_norm": 0.7340566515922546,
+      "learning_rate": 0.0002,
+      "loss": 0.6304,
+      "step": 13410
+    },
+    {
+      "epoch": 2.169590170560181,
+      "grad_norm": 0.7075738906860352,
+      "learning_rate": 0.0002,
+      "loss": 0.6528,
+      "step": 13420
+    },
+    {
+      "epoch": 2.1712068547409262,
+      "grad_norm": 0.7080879807472229,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 13430
+    },
+    {
+      "epoch": 2.1728235389216715,
+      "grad_norm": 0.6218613386154175,
+      "learning_rate": 0.0002,
+      "loss": 0.6615,
+      "step": 13440
+    },
+    {
+      "epoch": 2.174440223102417,
+      "grad_norm": 0.8211479187011719,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 13450
+    },
+    {
+      "epoch": 2.1760569072831624,
+      "grad_norm": 0.864466667175293,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 13460
+    },
+    {
+      "epoch": 2.1776735914639076,
+      "grad_norm": 0.7943857908248901,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 13470
+    },
+    {
+      "epoch": 2.179290275644653,
+      "grad_norm": 0.78728187084198,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 13480
+    },
+    {
+      "epoch": 2.180906959825398,
+      "grad_norm": 0.697527289390564,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 13490
+    },
+    {
+      "epoch": 2.1825236440061433,
+      "grad_norm": 0.8205804228782654,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 13500
+    },
+    {
+      "epoch": 2.1841403281868885,
+      "grad_norm": 0.8709042072296143,
+      "learning_rate": 0.0002,
+      "loss": 0.7227,
+      "step": 13510
+    },
+    {
+      "epoch": 2.1857570123676338,
+      "grad_norm": 0.6228537559509277,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 13520
+    },
+    {
+      "epoch": 2.1873736965483794,
+      "grad_norm": 0.9566980004310608,
+      "learning_rate": 0.0002,
+      "loss": 0.7025,
+      "step": 13530
+    },
+    {
+      "epoch": 2.1889903807291247,
+      "grad_norm": 0.7128894329071045,
+      "learning_rate": 0.0002,
+      "loss": 0.6755,
+      "step": 13540
+    },
+    {
+      "epoch": 2.19060706490987,
+      "grad_norm": 0.6888654232025146,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 13550
+    },
+    {
+      "epoch": 2.192223749090615,
+      "grad_norm": 0.6444337368011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6961,
+      "step": 13560
+    },
+    {
+      "epoch": 2.1938404332713604,
+      "grad_norm": 0.8008806705474854,
+      "learning_rate": 0.0002,
+      "loss": 0.656,
+      "step": 13570
+    },
+    {
+      "epoch": 2.1954571174521056,
+      "grad_norm": 0.8482748866081238,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 13580
+    },
+    {
+      "epoch": 2.197073801632851,
+      "grad_norm": 0.8584157228469849,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 13590
+    },
+    {
+      "epoch": 2.1986904858135965,
+      "grad_norm": 0.7513734698295593,
+      "learning_rate": 0.0002,
+      "loss": 0.7014,
+      "step": 13600
+    },
+    {
+      "epoch": 2.2003071699943417,
+      "grad_norm": 0.7864262461662292,
+      "learning_rate": 0.0002,
+      "loss": 0.6632,
+      "step": 13610
+    },
+    {
+      "epoch": 2.201923854175087,
+      "grad_norm": 0.8493645191192627,
+      "learning_rate": 0.0002,
+      "loss": 0.6879,
+      "step": 13620
+    },
+    {
+      "epoch": 2.203540538355832,
+      "grad_norm": 0.6902140974998474,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 13630
+    },
+    {
+      "epoch": 2.2051572225365774,
+      "grad_norm": 0.8711254596710205,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 13640
+    },
+    {
+      "epoch": 2.2067739067173227,
+      "grad_norm": 0.7832191586494446,
+      "learning_rate": 0.0002,
+      "loss": 0.6359,
+      "step": 13650
+    },
+    {
+      "epoch": 2.208390590898068,
+      "grad_norm": 0.5668176412582397,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2100072750788136,
+      "grad_norm": 0.8648375272750854,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13670
+    },
+    {
+      "epoch": 2.211623959259559,
+      "grad_norm": 0.7643089890480042,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13680
+    },
+    {
+      "epoch": 2.213240643440304,
+      "grad_norm": 0.6293777823448181,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13690
+    },
+    {
+      "epoch": 2.2148573276210493,
+      "grad_norm": 0.6459372639656067,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 13700
+    },
+    {
+      "epoch": 2.2164740118017945,
+      "grad_norm": 0.7060744166374207,
+      "learning_rate": 0.0002,
+      "loss": 0.6526,
+      "step": 13710
+    },
+    {
+      "epoch": 2.2180906959825397,
+      "grad_norm": 0.674109160900116,
+      "learning_rate": 0.0002,
+      "loss": 0.7101,
+      "step": 13720
+    },
+    {
+      "epoch": 2.219707380163285,
+      "grad_norm": 0.830392062664032,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13730
+    },
+    {
+      "epoch": 2.2213240643440306,
+      "grad_norm": 0.6474477052688599,
+      "learning_rate": 0.0002,
+      "loss": 0.6733,
+      "step": 13740
+    },
+    {
+      "epoch": 2.222940748524776,
+      "grad_norm": 0.7037909626960754,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13750
+    },
+    {
+      "epoch": 2.224557432705521,
+      "grad_norm": 0.6554131507873535,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 13760
+    },
+    {
+      "epoch": 2.2261741168862663,
+      "grad_norm": 0.7822230458259583,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 13770
+    },
+    {
+      "epoch": 2.2277908010670116,
+      "grad_norm": 0.9082167744636536,
+      "learning_rate": 0.0002,
+      "loss": 0.6505,
+      "step": 13780
+    },
+    {
+      "epoch": 2.229407485247757,
+      "grad_norm": 0.7918276190757751,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 13790
+    },
+    {
+      "epoch": 2.231024169428502,
+      "grad_norm": 0.7354569435119629,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 13800
+    },
+    {
+      "epoch": 2.2326408536092472,
+      "grad_norm": 0.8265249133110046,
+      "learning_rate": 0.0002,
+      "loss": 0.6503,
+      "step": 13810
+    },
+    {
+      "epoch": 2.234257537789993,
+      "grad_norm": 0.6653847098350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6871,
+      "step": 13820
+    },
+    {
+      "epoch": 2.235874221970738,
+      "grad_norm": 0.7157923579216003,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13830
+    },
+    {
+      "epoch": 2.2374909061514834,
+      "grad_norm": 0.7110323309898376,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 13840
+    },
+    {
+      "epoch": 2.2391075903322286,
+      "grad_norm": 0.7155357599258423,
+      "learning_rate": 0.0002,
+      "loss": 0.6913,
+      "step": 13850
+    },
+    {
+      "epoch": 2.240724274512974,
+      "grad_norm": 1.0177817344665527,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 13860
+    },
+    {
+      "epoch": 2.242340958693719,
+      "grad_norm": 0.7601948380470276,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13870
+    },
+    {
+      "epoch": 2.2439576428744643,
+      "grad_norm": 0.7628820538520813,
+      "learning_rate": 0.0002,
+      "loss": 0.6679,
+      "step": 13880
+    },
+    {
+      "epoch": 2.24557432705521,
+      "grad_norm": 0.7089297771453857,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 13890
+    },
+    {
+      "epoch": 2.247191011235955,
+      "grad_norm": 0.695178210735321,
+      "learning_rate": 0.0002,
+      "loss": 0.7236,
+      "step": 13900
+    },
+    {
+      "epoch": 2.2488076954167004,
+      "grad_norm": 0.7631948590278625,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 13910
+    },
+    {
+      "epoch": 2.2504243795974457,
+      "grad_norm": 0.8203101754188538,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 13920
+    },
+    {
+      "epoch": 2.252041063778191,
+      "grad_norm": 0.8099079728126526,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13930
+    },
+    {
+      "epoch": 2.253657747958936,
+      "grad_norm": 0.6498546004295349,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 13940
+    },
+    {
+      "epoch": 2.2552744321396814,
+      "grad_norm": 0.7797415256500244,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 13950
+    },
+    {
+      "epoch": 2.2568911163204266,
+      "grad_norm": 0.8254124522209167,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 13960
+    },
+    {
+      "epoch": 2.2585078005011723,
+      "grad_norm": 0.6327953338623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 13970
+    },
+    {
+      "epoch": 2.2601244846819175,
+      "grad_norm": 0.734194278717041,
+      "learning_rate": 0.0002,
+      "loss": 0.668,
+      "step": 13980
+    },
+    {
+      "epoch": 2.2617411688626627,
+      "grad_norm": 0.9014202952384949,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 13990
+    },
+    {
+      "epoch": 2.263357853043408,
+      "grad_norm": 0.7643631100654602,
+      "learning_rate": 0.0002,
+      "loss": 0.692,
+      "step": 14000
+    },
+    {
+      "epoch": 2.264974537224153,
+      "grad_norm": 0.8882834911346436,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 14010
+    },
+    {
+      "epoch": 2.2665912214048984,
+      "grad_norm": 0.7975873351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14020
+    },
+    {
+      "epoch": 2.2682079055856437,
+      "grad_norm": 0.7765783071517944,
+      "learning_rate": 0.0002,
+      "loss": 0.7193,
+      "step": 14030
+    },
+    {
+      "epoch": 2.2698245897663893,
+      "grad_norm": 0.8846288323402405,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 14040
+    },
+    {
+      "epoch": 2.2714412739471346,
+      "grad_norm": 0.9006744027137756,
+      "learning_rate": 0.0002,
+      "loss": 0.6494,
+      "step": 14050
+    },
+    {
+      "epoch": 2.27305795812788,
+      "grad_norm": 0.7420173287391663,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 14060
+    },
+    {
+      "epoch": 2.274674642308625,
+      "grad_norm": 0.7956424951553345,
+      "learning_rate": 0.0002,
+      "loss": 0.7068,
+      "step": 14070
+    },
+    {
+      "epoch": 2.2762913264893703,
+      "grad_norm": 0.7783209085464478,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 14080
+    },
+    {
+      "epoch": 2.2779080106701155,
+      "grad_norm": 0.7597188949584961,
+      "learning_rate": 0.0002,
+      "loss": 0.7202,
+      "step": 14090
+    },
+    {
+      "epoch": 2.2795246948508607,
+      "grad_norm": 0.6718921661376953,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14100
+    },
+    {
+      "epoch": 2.281141379031606,
+      "grad_norm": 0.7528082132339478,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 14110
+    },
+    {
+      "epoch": 2.2827580632123516,
+      "grad_norm": 0.8379864692687988,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 14120
+    },
+    {
+      "epoch": 2.284374747393097,
+      "grad_norm": 0.748613715171814,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 14130
+    },
+    {
+      "epoch": 2.285991431573842,
+      "grad_norm": 0.7435423135757446,
+      "learning_rate": 0.0002,
+      "loss": 0.6641,
+      "step": 14140
+    },
+    {
+      "epoch": 2.2876081157545873,
+      "grad_norm": 0.7580803632736206,
+      "learning_rate": 0.0002,
+      "loss": 0.6849,
+      "step": 14150
+    },
+    {
+      "epoch": 2.2892247999353326,
+      "grad_norm": 0.6278321146965027,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 14160
+    },
+    {
+      "epoch": 2.290841484116078,
+      "grad_norm": 0.7663896083831787,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 14170
+    },
+    {
+      "epoch": 2.292458168296823,
+      "grad_norm": 0.9716812372207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 14180
+    },
+    {
+      "epoch": 2.2940748524775687,
+      "grad_norm": 0.8993458151817322,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14190
+    },
+    {
+      "epoch": 2.295691536658314,
+      "grad_norm": 0.6156117916107178,
+      "learning_rate": 0.0002,
+      "loss": 0.6172,
+      "step": 14200
+    },
+    {
+      "epoch": 2.297308220839059,
+      "grad_norm": 0.8911278247833252,
+      "learning_rate": 0.0002,
+      "loss": 0.6318,
+      "step": 14210
+    },
+    {
+      "epoch": 2.2989249050198044,
+      "grad_norm": 0.6422147154808044,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 14220
+    },
+    {
+      "epoch": 2.3005415892005496,
+      "grad_norm": 0.6866879463195801,
+      "learning_rate": 0.0002,
+      "loss": 0.6795,
+      "step": 14230
+    },
+    {
+      "epoch": 2.302158273381295,
+      "grad_norm": 0.9297130107879639,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 14240
+    },
+    {
+      "epoch": 2.30377495756204,
+      "grad_norm": 0.7501356601715088,
+      "learning_rate": 0.0002,
+      "loss": 0.6823,
+      "step": 14250
+    },
+    {
+      "epoch": 2.3053916417427853,
+      "grad_norm": 0.8363515138626099,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 14260
+    },
+    {
+      "epoch": 2.307008325923531,
+      "grad_norm": 0.9083868265151978,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3086250101042762,
+      "grad_norm": 0.7791516780853271,
+      "learning_rate": 0.0002,
+      "loss": 0.6862,
+      "step": 14280
+    },
+    {
+      "epoch": 2.3102416942850215,
+      "grad_norm": 0.8766953349113464,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14290
+    },
+    {
+      "epoch": 2.3118583784657667,
+      "grad_norm": 0.7916635274887085,
+      "learning_rate": 0.0002,
+      "loss": 0.6698,
+      "step": 14300
+    },
+    {
+      "epoch": 2.313475062646512,
+      "grad_norm": 0.627525269985199,
+      "learning_rate": 0.0002,
+      "loss": 0.6927,
+      "step": 14310
+    },
+    {
+      "epoch": 2.315091746827257,
+      "grad_norm": 0.8856783509254456,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 14320
+    },
+    {
+      "epoch": 2.316708431008003,
+      "grad_norm": 0.6758689284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 14330
+    },
+    {
+      "epoch": 2.318325115188748,
+      "grad_norm": 0.6428321003913879,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 14340
+    },
+    {
+      "epoch": 2.3199417993694933,
+      "grad_norm": 0.9032121300697327,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 14350
+    },
+    {
+      "epoch": 2.3215584835502385,
+      "grad_norm": 0.8035986423492432,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14360
+    },
+    {
+      "epoch": 2.3231751677309838,
+      "grad_norm": 0.7974579334259033,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14370
+    },
+    {
+      "epoch": 2.324791851911729,
+      "grad_norm": 0.8356034755706787,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 14380
+    },
+    {
+      "epoch": 2.326408536092474,
+      "grad_norm": 0.998760998249054,
+      "learning_rate": 0.0002,
+      "loss": 0.6623,
+      "step": 14390
+    },
+    {
+      "epoch": 2.3280252202732195,
+      "grad_norm": 0.6518142223358154,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 14400
+    },
+    {
+      "epoch": 2.3296419044539647,
+      "grad_norm": 0.7443506717681885,
+      "learning_rate": 0.0002,
+      "loss": 0.7146,
+      "step": 14410
+    },
+    {
+      "epoch": 2.3312585886347104,
+      "grad_norm": 0.8436172604560852,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14420
+    },
+    {
+      "epoch": 2.3328752728154556,
+      "grad_norm": 0.7411080598831177,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 14430
+    },
+    {
+      "epoch": 2.334491956996201,
+      "grad_norm": 0.8839048743247986,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 14440
+    },
+    {
+      "epoch": 2.336108641176946,
+      "grad_norm": 0.8360885977745056,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 14450
+    },
+    {
+      "epoch": 2.3377253253576913,
+      "grad_norm": 0.7608986496925354,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 14460
+    },
+    {
+      "epoch": 2.3393420095384365,
+      "grad_norm": 0.8179867267608643,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14470
+    },
+    {
+      "epoch": 2.340958693719182,
+      "grad_norm": 0.5989999771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14480
+    },
+    {
+      "epoch": 2.3425753778999274,
+      "grad_norm": 0.9450054168701172,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 14490
+    },
+    {
+      "epoch": 2.3441920620806727,
+      "grad_norm": 0.7885149717330933,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 14500
+    },
+    {
+      "epoch": 2.345808746261418,
+      "grad_norm": 0.8152616620063782,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 14510
+    },
+    {
+      "epoch": 2.347425430442163,
+      "grad_norm": 0.7193838953971863,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 14520
+    },
+    {
+      "epoch": 2.3490421146229084,
+      "grad_norm": 0.6701092720031738,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 14530
+    },
+    {
+      "epoch": 2.3506587988036536,
+      "grad_norm": 0.7529364228248596,
+      "learning_rate": 0.0002,
+      "loss": 0.6559,
+      "step": 14540
+    },
+    {
+      "epoch": 2.352275482984399,
+      "grad_norm": 0.6599733829498291,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 14550
+    },
+    {
+      "epoch": 2.353892167165144,
+      "grad_norm": 0.9502474069595337,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14560
+    },
+    {
+      "epoch": 2.3555088513458897,
+      "grad_norm": 0.7619650959968567,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 14570
+    },
+    {
+      "epoch": 2.357125535526635,
+      "grad_norm": 0.9854652285575867,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14580
+    },
+    {
+      "epoch": 2.35874221970738,
+      "grad_norm": 0.727439284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 14590
+    },
+    {
+      "epoch": 2.3603589038881254,
+      "grad_norm": 0.6994746327400208,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 14600
+    },
+    {
+      "epoch": 2.3619755880688706,
+      "grad_norm": 0.7117531299591064,
+      "learning_rate": 0.0002,
+      "loss": 0.6587,
+      "step": 14610
+    },
+    {
+      "epoch": 2.363592272249616,
+      "grad_norm": 0.6403067708015442,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 14620
+    },
+    {
+      "epoch": 2.3652089564303616,
+      "grad_norm": 0.8377841711044312,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14630
+    },
+    {
+      "epoch": 2.366825640611107,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14640
+    },
+    {
+      "epoch": 2.368442324791852,
+      "grad_norm": 0.8418586254119873,
+      "learning_rate": 0.0002,
+      "loss": 0.6552,
+      "step": 14650
+    },
+    {
+      "epoch": 2.3700590089725972,
+      "grad_norm": 0.6178573369979858,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14660
+    },
+    {
+      "epoch": 2.3716756931533425,
+      "grad_norm": 0.6368302702903748,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 14670
+    },
+    {
+      "epoch": 2.3732923773340877,
+      "grad_norm": 0.9122977256774902,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14680
+    },
+    {
+      "epoch": 2.374909061514833,
+      "grad_norm": 0.7086195349693298,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 14690
+    },
+    {
+      "epoch": 2.376525745695578,
+      "grad_norm": 0.7500800490379333,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 14700
+    },
+    {
+      "epoch": 2.378142429876324,
+      "grad_norm": 0.6634900569915771,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 14710
+    },
+    {
+      "epoch": 2.379759114057069,
+      "grad_norm": 0.839898407459259,
+      "learning_rate": 0.0002,
+      "loss": 0.6614,
+      "step": 14720
+    },
+    {
+      "epoch": 2.3813757982378143,
+      "grad_norm": 0.7578426003456116,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14730
+    },
+    {
+      "epoch": 2.3829924824185595,
+      "grad_norm": 1.0213173627853394,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3846091665993048,
+      "grad_norm": 0.7855949401855469,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 14750
+    },
+    {
+      "epoch": 2.38622585078005,
+      "grad_norm": 0.7224128842353821,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 14760
+    },
+    {
+      "epoch": 2.3878425349607952,
+      "grad_norm": 0.8040381669998169,
+      "learning_rate": 0.0002,
+      "loss": 0.7017,
+      "step": 14770
+    },
+    {
+      "epoch": 2.389459219141541,
+      "grad_norm": 0.7705281376838684,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 14780
+    },
+    {
+      "epoch": 2.391075903322286,
+      "grad_norm": 0.667966902256012,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 14790
+    },
+    {
+      "epoch": 2.3926925875030314,
+      "grad_norm": 0.6611011028289795,
+      "learning_rate": 0.0002,
+      "loss": 0.7061,
+      "step": 14800
+    },
+    {
+      "epoch": 2.3943092716837766,
+      "grad_norm": 0.6862651705741882,
+      "learning_rate": 0.0002,
+      "loss": 0.6527,
+      "step": 14810
+    },
+    {
+      "epoch": 2.395925955864522,
+      "grad_norm": 0.8086010217666626,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 14820
+    },
+    {
+      "epoch": 2.397542640045267,
+      "grad_norm": 0.7189689874649048,
+      "learning_rate": 0.0002,
+      "loss": 0.7189,
+      "step": 14830
+    },
+    {
+      "epoch": 2.3991593242260123,
+      "grad_norm": 0.6280009150505066,
+      "learning_rate": 0.0002,
+      "loss": 0.6709,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4007760084067575,
+      "grad_norm": 0.7826612591743469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14850
+    },
+    {
+      "epoch": 2.402392692587503,
+      "grad_norm": 0.7681610584259033,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4040093767682484,
+      "grad_norm": 0.720966100692749,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 14870
+    },
+    {
+      "epoch": 2.4056260609489937,
+      "grad_norm": 0.8202250599861145,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 14880
+    },
+    {
+      "epoch": 2.407242745129739,
+      "grad_norm": 0.786212682723999,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 14890
+    },
+    {
+      "epoch": 2.408859429310484,
+      "grad_norm": 0.6647164821624756,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 14900
+    },
+    {
+      "epoch": 2.4104761134912294,
+      "grad_norm": 0.7566399574279785,
+      "learning_rate": 0.0002,
+      "loss": 0.6706,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4120927976719746,
+      "grad_norm": 0.748814582824707,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 14920
+    },
+    {
+      "epoch": 2.4137094818527203,
+      "grad_norm": 0.7624038457870483,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14930
+    },
+    {
+      "epoch": 2.4153261660334655,
+      "grad_norm": 0.8267335295677185,
+      "learning_rate": 0.0002,
+      "loss": 0.6483,
+      "step": 14940
+    },
+    {
+      "epoch": 2.4169428502142107,
+      "grad_norm": 0.8785360455513,
+      "learning_rate": 0.0002,
+      "loss": 0.6612,
+      "step": 14950
+    },
+    {
+      "epoch": 2.418559534394956,
+      "grad_norm": 0.679887592792511,
+      "learning_rate": 0.0002,
+      "loss": 0.6718,
+      "step": 14960
+    },
+    {
+      "epoch": 2.420176218575701,
+      "grad_norm": 0.7218474745750427,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4217929027564464,
+      "grad_norm": 0.6342799663543701,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14980
+    },
+    {
+      "epoch": 2.4234095869371917,
+      "grad_norm": 0.7098712921142578,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 14990
+    },
+    {
+      "epoch": 2.425026271117937,
+      "grad_norm": 0.7497431635856628,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4266429552986826,
+      "grad_norm": 0.934836208820343,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15010
+    },
+    {
+      "epoch": 2.428259639479428,
+      "grad_norm": 0.8430966734886169,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 15020
+    },
+    {
+      "epoch": 2.429876323660173,
+      "grad_norm": 0.7032104730606079,
+      "learning_rate": 0.0002,
+      "loss": 0.6985,
+      "step": 15030
+    },
+    {
+      "epoch": 2.4314930078409183,
+      "grad_norm": 0.7746111750602722,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 15040
+    },
+    {
+      "epoch": 2.4331096920216635,
+      "grad_norm": 0.7661406397819519,
+      "learning_rate": 0.0002,
+      "loss": 0.7177,
+      "step": 15050
+    },
+    {
+      "epoch": 2.4347263762024087,
+      "grad_norm": 0.6941645741462708,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 15060
+    },
+    {
+      "epoch": 2.436343060383154,
+      "grad_norm": 0.7487249374389648,
+      "learning_rate": 0.0002,
+      "loss": 0.6421,
+      "step": 15070
+    },
+    {
+      "epoch": 2.4379597445638996,
+      "grad_norm": 0.7639912962913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 15080
+    },
+    {
+      "epoch": 2.439576428744645,
+      "grad_norm": 0.7708953619003296,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 15090
+    },
+    {
+      "epoch": 2.44119311292539,
+      "grad_norm": 0.9135832190513611,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15100
+    },
+    {
+      "epoch": 2.4428097971061353,
+      "grad_norm": 0.8283005356788635,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 15110
+    },
+    {
+      "epoch": 2.4444264812868806,
+      "grad_norm": 0.925299346446991,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 15120
+    },
+    {
+      "epoch": 2.446043165467626,
+      "grad_norm": 0.7013528943061829,
+      "learning_rate": 0.0002,
+      "loss": 0.687,
+      "step": 15130
+    },
+    {
+      "epoch": 2.447659849648371,
+      "grad_norm": 0.622303307056427,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 15140
+    },
+    {
+      "epoch": 2.4492765338291163,
+      "grad_norm": 0.876569390296936,
+      "learning_rate": 0.0002,
+      "loss": 0.6676,
+      "step": 15150
+    },
+    {
+      "epoch": 2.450893218009862,
+      "grad_norm": 0.6836351752281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 15160
+    },
+    {
+      "epoch": 2.452509902190607,
+      "grad_norm": 0.7886684536933899,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 15170
+    },
+    {
+      "epoch": 2.4541265863713524,
+      "grad_norm": 0.6647440791130066,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 15180
+    },
+    {
+      "epoch": 2.4557432705520976,
+      "grad_norm": 0.7477722764015198,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 15190
+    },
+    {
+      "epoch": 2.457359954732843,
+      "grad_norm": 0.8192033767700195,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 15200
+    },
+    {
+      "epoch": 2.458976638913588,
+      "grad_norm": 0.847537100315094,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 15210
+    },
+    {
+      "epoch": 2.4605933230943338,
+      "grad_norm": 0.9027776122093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 15220
+    },
+    {
+      "epoch": 2.462210007275079,
+      "grad_norm": 0.7217772006988525,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 15230
+    },
+    {
+      "epoch": 2.4638266914558242,
+      "grad_norm": 0.7994546294212341,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 15240
+    },
+    {
+      "epoch": 2.4654433756365695,
+      "grad_norm": 0.939916729927063,
+      "learning_rate": 0.0002,
+      "loss": 0.6828,
+      "step": 15250
+    },
+    {
+      "epoch": 2.4670600598173147,
+      "grad_norm": 1.0009053945541382,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15260
+    },
+    {
+      "epoch": 2.46867674399806,
+      "grad_norm": 0.625555694103241,
+      "learning_rate": 0.0002,
+      "loss": 0.643,
+      "step": 15270
+    },
+    {
+      "epoch": 2.470293428178805,
+      "grad_norm": 0.7924878597259521,
+      "learning_rate": 0.0002,
+      "loss": 0.688,
+      "step": 15280
+    },
+    {
+      "epoch": 2.4719101123595504,
+      "grad_norm": 0.8536689877510071,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 15290
+    },
+    {
+      "epoch": 2.4735267965402956,
+      "grad_norm": 0.8572589755058289,
+      "learning_rate": 0.0002,
+      "loss": 0.6924,
+      "step": 15300
+    },
+    {
+      "epoch": 2.4751434807210413,
+      "grad_norm": 0.773279070854187,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 15310
+    },
+    {
+      "epoch": 2.4767601649017865,
+      "grad_norm": 0.7708749771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 15320
+    },
+    {
+      "epoch": 2.4783768490825318,
+      "grad_norm": 0.770905077457428,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15330
+    },
+    {
+      "epoch": 2.479993533263277,
+      "grad_norm": 0.8238571882247925,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 15340
+    },
+    {
+      "epoch": 2.481610217444022,
+      "grad_norm": 0.7670477032661438,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15350
+    },
+    {
+      "epoch": 2.4832269016247674,
+      "grad_norm": 0.905036985874176,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 15360
+    },
+    {
+      "epoch": 2.484843585805513,
+      "grad_norm": 0.6672089695930481,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 15370
+    },
+    {
+      "epoch": 2.4864602699862584,
+      "grad_norm": 0.625095784664154,
+      "learning_rate": 0.0002,
+      "loss": 0.6722,
+      "step": 15380
+    },
+    {
+      "epoch": 2.4880769541670036,
+      "grad_norm": 0.679772675037384,
+      "learning_rate": 0.0002,
+      "loss": 0.6396,
+      "step": 15390
+    },
+    {
+      "epoch": 2.489693638347749,
+      "grad_norm": 0.711492121219635,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 15400
+    },
+    {
+      "epoch": 2.491310322528494,
+      "grad_norm": 0.876189112663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6966,
+      "step": 15410
+    },
+    {
+      "epoch": 2.4929270067092393,
+      "grad_norm": 0.7236915230751038,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 15420
+    },
+    {
+      "epoch": 2.4945436908899845,
+      "grad_norm": 0.6629832983016968,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 15430
+    },
+    {
+      "epoch": 2.4961603750707297,
+      "grad_norm": 0.9756859540939331,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 15440
+    },
+    {
+      "epoch": 2.4977770592514754,
+      "grad_norm": 0.6896940469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 15450
+    },
+    {
+      "epoch": 2.4993937434322206,
+      "grad_norm": 0.7105149626731873,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 15460
+    },
+    {
+      "epoch": 2.501010427612966,
+      "grad_norm": 0.8374546766281128,
+      "learning_rate": 0.0002,
+      "loss": 0.6859,
+      "step": 15470
+    },
+    {
+      "epoch": 2.502627111793711,
+      "grad_norm": 0.7320070266723633,
+      "learning_rate": 0.0002,
+      "loss": 0.6512,
+      "step": 15480
+    },
+    {
+      "epoch": 2.5042437959744563,
+      "grad_norm": 0.8306367993354797,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 15490
+    },
+    {
+      "epoch": 2.5058604801552016,
+      "grad_norm": 0.7472721338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.7253,
+      "step": 15500
+    },
+    {
+      "epoch": 2.507477164335947,
+      "grad_norm": 0.6147692203521729,
+      "learning_rate": 0.0002,
+      "loss": 0.6699,
+      "step": 15510
+    },
+    {
+      "epoch": 2.5090938485166925,
+      "grad_norm": 0.7788505554199219,
+      "learning_rate": 0.0002,
+      "loss": 0.7158,
+      "step": 15520
+    },
+    {
+      "epoch": 2.5107105326974377,
+      "grad_norm": 0.8807527422904968,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 15530
+    },
+    {
+      "epoch": 2.512327216878183,
+      "grad_norm": 0.7521643042564392,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 15540
+    },
+    {
+      "epoch": 2.513943901058928,
+      "grad_norm": 0.6900225281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15550
+    },
+    {
+      "epoch": 2.5155605852396734,
+      "grad_norm": 0.6601938605308533,
+      "learning_rate": 0.0002,
+      "loss": 0.6769,
+      "step": 15560
+    },
+    {
+      "epoch": 2.5171772694204186,
+      "grad_norm": 0.8179984092712402,
+      "learning_rate": 0.0002,
+      "loss": 0.6648,
+      "step": 15570
+    },
+    {
+      "epoch": 2.518793953601164,
+      "grad_norm": 0.792556881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.7028,
+      "step": 15580
+    },
+    {
+      "epoch": 2.520410637781909,
+      "grad_norm": 0.7081938982009888,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 15590
+    },
+    {
+      "epoch": 2.5220273219626543,
+      "grad_norm": 0.8733121156692505,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 15600
+    },
+    {
+      "epoch": 2.5236440061434,
+      "grad_norm": 0.7980992794036865,
+      "learning_rate": 0.0002,
+      "loss": 0.6969,
+      "step": 15610
+    },
+    {
+      "epoch": 2.5252606903241452,
+      "grad_norm": 0.883664071559906,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 15620
+    },
+    {
+      "epoch": 2.5268773745048905,
+      "grad_norm": 0.6963341236114502,
+      "learning_rate": 0.0002,
+      "loss": 0.7022,
+      "step": 15630
+    },
+    {
+      "epoch": 2.5284940586856357,
+      "grad_norm": 0.6433573365211487,
+      "learning_rate": 0.0002,
+      "loss": 0.7334,
+      "step": 15640
+    },
+    {
+      "epoch": 2.530110742866381,
+      "grad_norm": 0.8538183569908142,
+      "learning_rate": 0.0002,
+      "loss": 0.6889,
+      "step": 15650
+    },
+    {
+      "epoch": 2.5317274270471266,
+      "grad_norm": 0.9748201370239258,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 15660
+    },
+    {
+      "epoch": 2.533344111227872,
+      "grad_norm": 0.7670575380325317,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 15670
+    },
+    {
+      "epoch": 2.534960795408617,
+      "grad_norm": 0.8738890290260315,
+      "learning_rate": 0.0002,
+      "loss": 0.6435,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5365774795893623,
+      "grad_norm": 0.8391636610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.6802,
+      "step": 15690
+    },
+    {
+      "epoch": 2.5381941637701075,
+      "grad_norm": 0.7239366769790649,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 15700
+    },
+    {
+      "epoch": 2.5398108479508528,
+      "grad_norm": 0.8498379588127136,
+      "learning_rate": 0.0002,
+      "loss": 0.7011,
+      "step": 15710
+    },
+    {
+      "epoch": 2.541427532131598,
+      "grad_norm": 0.8029484152793884,
+      "learning_rate": 0.0002,
+      "loss": 0.6998,
+      "step": 15720
+    },
+    {
+      "epoch": 2.5430442163123432,
+      "grad_norm": 1.0639333724975586,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 15730
+    },
+    {
+      "epoch": 2.5446609004930885,
+      "grad_norm": 0.6401297450065613,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 15740
+    },
+    {
+      "epoch": 2.5462775846738337,
+      "grad_norm": 0.7123814821243286,
+      "learning_rate": 0.0002,
+      "loss": 0.7196,
+      "step": 15750
+    },
+    {
+      "epoch": 2.5478942688545794,
+      "grad_norm": 0.7874974608421326,
+      "learning_rate": 0.0002,
+      "loss": 0.654,
+      "step": 15760
+    },
+    {
+      "epoch": 2.5495109530353246,
+      "grad_norm": 0.8046808838844299,
+      "learning_rate": 0.0002,
+      "loss": 0.6721,
+      "step": 15770
+    },
+    {
+      "epoch": 2.55112763721607,
+      "grad_norm": 0.7888661623001099,
+      "learning_rate": 0.0002,
+      "loss": 0.6665,
+      "step": 15780
+    },
+    {
+      "epoch": 2.552744321396815,
+      "grad_norm": 0.8445866107940674,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15790
+    },
+    {
+      "epoch": 2.5543610055775603,
+      "grad_norm": 0.7475846409797668,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 15800
+    },
+    {
+      "epoch": 2.555977689758306,
+      "grad_norm": 0.7455102801322937,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 15810
+    },
+    {
+      "epoch": 2.557594373939051,
+      "grad_norm": 0.8226983547210693,
+      "learning_rate": 0.0002,
+      "loss": 0.6932,
+      "step": 15820
+    },
+    {
+      "epoch": 2.5592110581197964,
+      "grad_norm": 0.8920368552207947,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 15830
+    },
+    {
+      "epoch": 2.5608277423005417,
+      "grad_norm": 0.8413904905319214,
+      "learning_rate": 0.0002,
+      "loss": 0.6297,
+      "step": 15840
+    },
+    {
+      "epoch": 2.562444426481287,
+      "grad_norm": 0.8483649492263794,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 15850
+    },
+    {
+      "epoch": 2.564061110662032,
+      "grad_norm": 0.5923284292221069,
+      "learning_rate": 0.0002,
+      "loss": 0.6957,
+      "step": 15860
+    },
+    {
+      "epoch": 2.5656777948427774,
+      "grad_norm": 0.8518726229667664,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 15870
+    },
+    {
+      "epoch": 2.5672944790235226,
+      "grad_norm": 0.731235146522522,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 15880
+    },
+    {
+      "epoch": 2.568911163204268,
+      "grad_norm": 0.7517194151878357,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 15890
+    },
+    {
+      "epoch": 2.5705278473850135,
+      "grad_norm": 0.8378692269325256,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 15900
+    },
+    {
+      "epoch": 2.5721445315657587,
+      "grad_norm": 0.843701958656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 15910
+    },
+    {
+      "epoch": 2.573761215746504,
+      "grad_norm": 0.7254629731178284,
+      "learning_rate": 0.0002,
+      "loss": 0.6882,
+      "step": 15920
+    },
+    {
+      "epoch": 2.575377899927249,
+      "grad_norm": 0.8863335847854614,
+      "learning_rate": 0.0002,
+      "loss": 0.6872,
+      "step": 15930
+    },
+    {
+      "epoch": 2.5769945841079944,
+      "grad_norm": 0.7675097584724426,
+      "learning_rate": 0.0002,
+      "loss": 0.6813,
+      "step": 15940
+    },
+    {
+      "epoch": 2.5786112682887397,
+      "grad_norm": 0.82063889503479,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 15950
+    },
+    {
+      "epoch": 2.5802279524694853,
+      "grad_norm": 0.7729717493057251,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 15960
+    },
+    {
+      "epoch": 2.5818446366502306,
+      "grad_norm": 0.8301846981048584,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 15970
+    },
+    {
+      "epoch": 2.583461320830976,
+      "grad_norm": 0.7906861305236816,
+      "learning_rate": 0.0002,
+      "loss": 0.6897,
+      "step": 15980
+    },
+    {
+      "epoch": 2.585078005011721,
+      "grad_norm": 0.6749057173728943,
+      "learning_rate": 0.0002,
+      "loss": 0.7175,
+      "step": 15990
+    },
+    {
+      "epoch": 2.5866946891924663,
+      "grad_norm": 0.9386842846870422,
+      "learning_rate": 0.0002,
+      "loss": 0.7212,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5883113733732115,
+      "grad_norm": 0.7868891358375549,
+      "learning_rate": 0.0002,
+      "loss": 0.6934,
+      "step": 16010
+    },
+    {
+      "epoch": 2.5899280575539567,
+      "grad_norm": 0.8674671053886414,
+      "learning_rate": 0.0002,
+      "loss": 0.7036,
+      "step": 16020
+    },
+    {
+      "epoch": 2.591544741734702,
+      "grad_norm": 0.7043559551239014,
+      "learning_rate": 0.0002,
+      "loss": 0.7217,
+      "step": 16030
+    },
+    {
+      "epoch": 2.593161425915447,
+      "grad_norm": 0.5846083760261536,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 16040
+    },
+    {
+      "epoch": 2.594778110096193,
+      "grad_norm": 0.7323982119560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 16050
+    },
+    {
+      "epoch": 2.596394794276938,
+      "grad_norm": 0.9069556593894958,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 16060
+    },
+    {
+      "epoch": 2.5980114784576833,
+      "grad_norm": 0.7522736191749573,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 16070
+    },
+    {
+      "epoch": 2.5996281626384286,
+      "grad_norm": 0.8149648308753967,
+      "learning_rate": 0.0002,
+      "loss": 0.6477,
+      "step": 16080
+    },
+    {
+      "epoch": 2.601244846819174,
+      "grad_norm": 0.6214233040809631,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 16090
+    },
+    {
+      "epoch": 2.602861530999919,
+      "grad_norm": 0.6803743839263916,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 16100
+    },
+    {
+      "epoch": 2.6044782151806647,
+      "grad_norm": 0.7223997116088867,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 16110
+    },
+    {
+      "epoch": 2.60609489936141,
+      "grad_norm": 0.7324174642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.6867,
+      "step": 16120
+    },
+    {
+      "epoch": 2.607711583542155,
+      "grad_norm": 0.9594739675521851,
+      "learning_rate": 0.0002,
+      "loss": 0.7159,
+      "step": 16130
+    },
+    {
+      "epoch": 2.6093282677229004,
+      "grad_norm": 0.9485327005386353,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 16140
+    },
+    {
+      "epoch": 2.6109449519036456,
+      "grad_norm": 0.8449000120162964,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 16150
+    },
+    {
+      "epoch": 2.612561636084391,
+      "grad_norm": 0.8520140051841736,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 16160
+    },
+    {
+      "epoch": 2.614178320265136,
+      "grad_norm": 0.7456524968147278,
+      "learning_rate": 0.0002,
+      "loss": 0.6759,
+      "step": 16170
+    },
+    {
+      "epoch": 2.6157950044458813,
+      "grad_norm": 0.9912857413291931,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 16180
+    },
+    {
+      "epoch": 2.6174116886266265,
+      "grad_norm": 0.9001946449279785,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 16190
+    },
+    {
+      "epoch": 2.619028372807372,
+      "grad_norm": 0.6568667888641357,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16200
+    },
+    {
+      "epoch": 2.6206450569881174,
+      "grad_norm": 1.0248128175735474,
+      "learning_rate": 0.0002,
+      "loss": 0.7013,
+      "step": 16210
+    },
+    {
+      "epoch": 2.6222617411688627,
+      "grad_norm": 0.6509039998054504,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 16220
+    },
+    {
+      "epoch": 2.623878425349608,
+      "grad_norm": 0.7626351118087769,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 16230
+    },
+    {
+      "epoch": 2.625495109530353,
+      "grad_norm": 0.6938552260398865,
+      "learning_rate": 0.0002,
+      "loss": 0.6556,
+      "step": 16240
+    },
+    {
+      "epoch": 2.6271117937110984,
+      "grad_norm": 0.6434680819511414,
+      "learning_rate": 0.0002,
+      "loss": 0.65,
+      "step": 16250
+    },
+    {
+      "epoch": 2.628728477891844,
+      "grad_norm": 0.7111515998840332,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 16260
+    },
+    {
+      "epoch": 2.6303451620725893,
+      "grad_norm": 0.7712395787239075,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 16270
+    },
+    {
+      "epoch": 2.6319618462533345,
+      "grad_norm": 0.792209267616272,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 16280
+    },
+    {
+      "epoch": 2.6335785304340797,
+      "grad_norm": 0.6801066398620605,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 16290
+    },
+    {
+      "epoch": 2.635195214614825,
+      "grad_norm": 0.7802573442459106,
+      "learning_rate": 0.0002,
+      "loss": 0.73,
+      "step": 16300
+    },
+    {
+      "epoch": 2.63681189879557,
+      "grad_norm": 0.7742244601249695,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 16310
+    },
+    {
+      "epoch": 2.6384285829763154,
+      "grad_norm": 0.664184033870697,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 16320
+    },
+    {
+      "epoch": 2.6400452671570607,
+      "grad_norm": 0.9242228865623474,
+      "learning_rate": 0.0002,
+      "loss": 0.6442,
+      "step": 16330
+    },
+    {
+      "epoch": 2.641661951337806,
+      "grad_norm": 0.9661325216293335,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 16340
+    },
+    {
+      "epoch": 2.6432786355185516,
+      "grad_norm": 0.837526798248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 16350
+    },
+    {
+      "epoch": 2.644895319699297,
+      "grad_norm": 1.1834373474121094,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 16360
+    },
+    {
+      "epoch": 2.646512003880042,
+      "grad_norm": 0.7467831373214722,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 16370
+    },
+    {
+      "epoch": 2.6481286880607873,
+      "grad_norm": 0.8627146482467651,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 16380
+    },
+    {
+      "epoch": 2.6497453722415325,
+      "grad_norm": 0.790447473526001,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 16390
+    },
+    {
+      "epoch": 2.651362056422278,
+      "grad_norm": 0.8447365164756775,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 16400
+    },
+    {
+      "epoch": 2.6529787406030234,
+      "grad_norm": 0.7831417918205261,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 16410
+    },
+    {
+      "epoch": 2.6545954247837686,
+      "grad_norm": 0.6837952136993408,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 16420
+    },
+    {
+      "epoch": 2.656212108964514,
+      "grad_norm": 0.7031801342964172,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 16430
+    },
+    {
+      "epoch": 2.657828793145259,
+      "grad_norm": 0.8963770866394043,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 16440
+    },
+    {
+      "epoch": 2.6594454773260043,
+      "grad_norm": 0.6852328181266785,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 16450
+    },
+    {
+      "epoch": 2.6610621615067496,
+      "grad_norm": 0.8069294095039368,
+      "learning_rate": 0.0002,
+      "loss": 0.6278,
+      "step": 16460
+    },
+    {
+      "epoch": 2.662678845687495,
+      "grad_norm": 0.7503686547279358,
+      "learning_rate": 0.0002,
+      "loss": 0.6939,
+      "step": 16470
+    },
+    {
+      "epoch": 2.66429552986824,
+      "grad_norm": 0.6430956125259399,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16480
+    },
+    {
+      "epoch": 2.6659122140489853,
+      "grad_norm": 0.7894312739372253,
+      "learning_rate": 0.0002,
+      "loss": 0.6863,
+      "step": 16490
+    },
+    {
+      "epoch": 2.667528898229731,
+      "grad_norm": 0.7277431488037109,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 16500
+    },
+    {
+      "epoch": 2.669145582410476,
+      "grad_norm": 0.6816153526306152,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 16510
+    },
+    {
+      "epoch": 2.6707622665912214,
+      "grad_norm": 0.8145235776901245,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 16520
+    },
+    {
+      "epoch": 2.6723789507719666,
+      "grad_norm": 0.8645890355110168,
+      "learning_rate": 0.0002,
+      "loss": 0.709,
+      "step": 16530
+    },
+    {
+      "epoch": 2.673995634952712,
+      "grad_norm": 0.704393208026886,
+      "learning_rate": 0.0002,
+      "loss": 0.6946,
+      "step": 16540
+    },
+    {
+      "epoch": 2.6756123191334575,
+      "grad_norm": 1.0120846033096313,
+      "learning_rate": 0.0002,
+      "loss": 0.6378,
+      "step": 16550
+    },
+    {
+      "epoch": 2.6772290033142028,
+      "grad_norm": 0.6919328570365906,
+      "learning_rate": 0.0002,
+      "loss": 0.7241,
+      "step": 16560
+    },
+    {
+      "epoch": 2.678845687494948,
+      "grad_norm": 0.6924574971199036,
+      "learning_rate": 0.0002,
+      "loss": 0.7098,
+      "step": 16570
+    },
+    {
+      "epoch": 2.6804623716756932,
+      "grad_norm": 0.9679301381111145,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 16580
+    },
+    {
+      "epoch": 2.6820790558564385,
+      "grad_norm": 0.6810211539268494,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 16590
+    },
+    {
+      "epoch": 2.6836957400371837,
+      "grad_norm": 0.9730555415153503,
+      "learning_rate": 0.0002,
+      "loss": 0.6688,
+      "step": 16600
+    },
+    {
+      "epoch": 2.685312424217929,
+      "grad_norm": 0.7852821350097656,
+      "learning_rate": 0.0002,
+      "loss": 0.7344,
+      "step": 16610
+    },
+    {
+      "epoch": 2.686929108398674,
+      "grad_norm": 0.6059057116508484,
+      "learning_rate": 0.0002,
+      "loss": 0.6401,
+      "step": 16620
+    },
+    {
+      "epoch": 2.6885457925794194,
+      "grad_norm": 0.9395958781242371,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 16630
+    },
+    {
+      "epoch": 2.690162476760165,
+      "grad_norm": 0.7473729848861694,
+      "learning_rate": 0.0002,
+      "loss": 0.7174,
+      "step": 16640
+    },
+    {
+      "epoch": 2.6917791609409103,
+      "grad_norm": 0.765934407711029,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 16650
+    },
+    {
+      "epoch": 2.6933958451216555,
+      "grad_norm": 0.8496677279472351,
+      "learning_rate": 0.0002,
+      "loss": 0.707,
+      "step": 16660
+    },
+    {
+      "epoch": 2.6950125293024008,
+      "grad_norm": 0.7641879916191101,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 16670
+    },
+    {
+      "epoch": 2.696629213483146,
+      "grad_norm": 0.8471952676773071,
+      "learning_rate": 0.0002,
+      "loss": 0.6566,
+      "step": 16680
+    },
+    {
+      "epoch": 2.6982458976638912,
+      "grad_norm": 0.6946060657501221,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 16690
+    },
+    {
+      "epoch": 2.699862581844637,
+      "grad_norm": 0.7361312508583069,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 16700
+    },
+    {
+      "epoch": 2.701479266025382,
+      "grad_norm": 0.6605038046836853,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 16710
+    },
+    {
+      "epoch": 2.7030959502061274,
+      "grad_norm": 0.7164411544799805,
+      "learning_rate": 0.0002,
+      "loss": 0.6885,
+      "step": 16720
+    },
+    {
+      "epoch": 2.7047126343868726,
+      "grad_norm": 0.6496201157569885,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 16730
+    },
+    {
+      "epoch": 2.706329318567618,
+      "grad_norm": 0.7826663851737976,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 16740
+    },
+    {
+      "epoch": 2.707946002748363,
+      "grad_norm": 0.7639131546020508,
+      "learning_rate": 0.0002,
+      "loss": 0.6773,
+      "step": 16750
+    },
+    {
+      "epoch": 2.7095626869291083,
+      "grad_norm": 0.7976210713386536,
+      "learning_rate": 0.0002,
+      "loss": 0.69,
+      "step": 16760
+    },
+    {
+      "epoch": 2.7111793711098535,
+      "grad_norm": 0.6836577653884888,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 16770
+    },
+    {
+      "epoch": 2.7127960552905988,
+      "grad_norm": 0.8025202751159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 16780
+    },
+    {
+      "epoch": 2.7144127394713444,
+      "grad_norm": 0.7636463642120361,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7160294236520897,
+      "grad_norm": 0.7481677532196045,
+      "learning_rate": 0.0002,
+      "loss": 0.6227,
+      "step": 16800
+    },
+    {
+      "epoch": 2.717646107832835,
+      "grad_norm": 0.7566834688186646,
+      "learning_rate": 0.0002,
+      "loss": 0.6925,
+      "step": 16810
+    },
+    {
+      "epoch": 2.71926279201358,
+      "grad_norm": 0.7931267619132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6531,
+      "step": 16820
+    },
+    {
+      "epoch": 2.7208794761943254,
+      "grad_norm": 0.8811662197113037,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 16830
+    },
+    {
+      "epoch": 2.7224961603750706,
+      "grad_norm": 0.8561240434646606,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 16840
+    },
+    {
+      "epoch": 2.7241128445558163,
+      "grad_norm": 0.7121599316596985,
+      "learning_rate": 0.0002,
+      "loss": 0.7135,
+      "step": 16850
+    },
+    {
+      "epoch": 2.7257295287365615,
+      "grad_norm": 0.8066257238388062,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16860
+    },
+    {
+      "epoch": 2.7273462129173067,
+      "grad_norm": 0.7699271440505981,
+      "learning_rate": 0.0002,
+      "loss": 0.6839,
+      "step": 16870
+    },
+    {
+      "epoch": 2.728962897098052,
+      "grad_norm": 1.1828432083129883,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 16880
+    },
+    {
+      "epoch": 2.730579581278797,
+      "grad_norm": 0.9989302754402161,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 16890
+    },
+    {
+      "epoch": 2.7321962654595424,
+      "grad_norm": 0.8100560307502747,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 16900
+    },
+    {
+      "epoch": 2.7338129496402876,
+      "grad_norm": 0.8615233898162842,
+      "learning_rate": 0.0002,
+      "loss": 0.6851,
+      "step": 16910
+    },
+    {
+      "epoch": 2.735429633821033,
+      "grad_norm": 0.8633756041526794,
+      "learning_rate": 0.0002,
+      "loss": 0.6322,
+      "step": 16920
+    },
+    {
+      "epoch": 2.737046318001778,
+      "grad_norm": 0.7769348621368408,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 16930
+    },
+    {
+      "epoch": 2.738663002182524,
+      "grad_norm": 0.6943058371543884,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 16940
+    },
+    {
+      "epoch": 2.740279686363269,
+      "grad_norm": 0.8510736227035522,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 16950
+    },
+    {
+      "epoch": 2.7418963705440142,
+      "grad_norm": 0.7732602953910828,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 16960
+    },
+    {
+      "epoch": 2.7435130547247595,
+      "grad_norm": 0.5981788635253906,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 16970
+    },
+    {
+      "epoch": 2.7451297389055047,
+      "grad_norm": 0.7604416012763977,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16980
+    },
+    {
+      "epoch": 2.74674642308625,
+      "grad_norm": 0.7377738356590271,
+      "learning_rate": 0.0002,
+      "loss": 0.6981,
+      "step": 16990
+    },
+    {
+      "epoch": 2.7483631072669956,
+      "grad_norm": 0.9400289058685303,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 17000
+    },
+    {
+      "epoch": 2.749979791447741,
+      "grad_norm": 0.6340599656105042,
+      "learning_rate": 0.0002,
+      "loss": 0.6952,
+      "step": 17010
+    },
+    {
+      "epoch": 2.751596475628486,
+      "grad_norm": 0.7297601103782654,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 17020
+    },
+    {
+      "epoch": 2.7532131598092313,
+      "grad_norm": 0.9479979872703552,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 17030
+    },
+    {
+      "epoch": 2.7548298439899765,
+      "grad_norm": 0.8461511135101318,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 17040
+    },
+    {
+      "epoch": 2.7564465281707218,
+      "grad_norm": 0.7477551698684692,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17050
+    },
+    {
+      "epoch": 2.758063212351467,
+      "grad_norm": 1.019270420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 17060
+    },
+    {
+      "epoch": 2.7596798965322122,
+      "grad_norm": 0.7730235457420349,
+      "learning_rate": 0.0002,
+      "loss": 0.7119,
+      "step": 17070
+    },
+    {
+      "epoch": 2.7612965807129575,
+      "grad_norm": 0.8216866254806519,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 17080
+    },
+    {
+      "epoch": 2.762913264893703,
+      "grad_norm": 0.7235931754112244,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17090
+    },
+    {
+      "epoch": 2.7645299490744484,
+      "grad_norm": 0.7352296710014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7031,
+      "step": 17100
+    },
+    {
+      "epoch": 2.7661466332551936,
+      "grad_norm": 0.8129373788833618,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 17110
+    },
+    {
+      "epoch": 2.767763317435939,
+      "grad_norm": 0.7387019991874695,
+      "learning_rate": 0.0002,
+      "loss": 0.6703,
+      "step": 17120
+    },
+    {
+      "epoch": 2.769380001616684,
+      "grad_norm": 0.9149190187454224,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 17130
+    },
+    {
+      "epoch": 2.7709966857974297,
+      "grad_norm": 0.7352971434593201,
+      "learning_rate": 0.0002,
+      "loss": 0.6038,
+      "step": 17140
+    },
+    {
+      "epoch": 2.772613369978175,
+      "grad_norm": 0.7903780341148376,
+      "learning_rate": 0.0002,
+      "loss": 0.6728,
+      "step": 17150
+    },
+    {
+      "epoch": 2.77423005415892,
+      "grad_norm": 0.8255927562713623,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17160
+    },
+    {
+      "epoch": 2.7758467383396654,
+      "grad_norm": 0.7235927581787109,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 17170
+    },
+    {
+      "epoch": 2.7774634225204107,
+      "grad_norm": 0.8281434774398804,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 17180
+    },
+    {
+      "epoch": 2.779080106701156,
+      "grad_norm": 0.7586921453475952,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 17190
+    },
+    {
+      "epoch": 2.780696790881901,
+      "grad_norm": 0.7161715030670166,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 17200
+    },
+    {
+      "epoch": 2.7823134750626464,
+      "grad_norm": 0.762868344783783,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 17210
+    },
+    {
+      "epoch": 2.7839301592433916,
+      "grad_norm": 0.9285483360290527,
+      "learning_rate": 0.0002,
+      "loss": 0.705,
+      "step": 17220
+    },
+    {
+      "epoch": 2.785546843424137,
+      "grad_norm": 0.6900462508201599,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 17230
+    },
+    {
+      "epoch": 2.7871635276048825,
+      "grad_norm": 0.780384361743927,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17240
+    },
+    {
+      "epoch": 2.7887802117856277,
+      "grad_norm": 0.7580406665802002,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17250
+    },
+    {
+      "epoch": 2.790396895966373,
+      "grad_norm": 0.8145199418067932,
+      "learning_rate": 0.0002,
+      "loss": 0.6833,
+      "step": 17260
+    },
+    {
+      "epoch": 2.792013580147118,
+      "grad_norm": 0.9159596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 17270
+    },
+    {
+      "epoch": 2.7936302643278634,
+      "grad_norm": 0.9590014219284058,
+      "learning_rate": 0.0002,
+      "loss": 0.6008,
+      "step": 17280
+    },
+    {
+      "epoch": 2.795246948508609,
+      "grad_norm": 0.7603529691696167,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 17290
+    },
+    {
+      "epoch": 2.7968636326893543,
+      "grad_norm": 0.8039976358413696,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 17300
+    },
+    {
+      "epoch": 2.7984803168700996,
+      "grad_norm": 0.8364847302436829,
+      "learning_rate": 0.0002,
+      "loss": 0.7037,
+      "step": 17310
+    },
+    {
+      "epoch": 2.800097001050845,
+      "grad_norm": 0.8763046860694885,
+      "learning_rate": 0.0002,
+      "loss": 0.6749,
+      "step": 17320
+    },
+    {
+      "epoch": 2.80171368523159,
+      "grad_norm": 0.8409647941589355,
+      "learning_rate": 0.0002,
+      "loss": 0.6844,
+      "step": 17330
+    },
+    {
+      "epoch": 2.8033303694123353,
+      "grad_norm": 0.7649006247520447,
+      "learning_rate": 0.0002,
+      "loss": 0.6936,
+      "step": 17340
+    },
+    {
+      "epoch": 2.8049470535930805,
+      "grad_norm": 0.7970262169837952,
+      "learning_rate": 0.0002,
+      "loss": 0.7051,
+      "step": 17350
+    },
+    {
+      "epoch": 2.8065637377738257,
+      "grad_norm": 0.9088607430458069,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 17360
+    },
+    {
+      "epoch": 2.808180421954571,
+      "grad_norm": 0.6454846858978271,
+      "learning_rate": 0.0002,
+      "loss": 0.675,
+      "step": 17370
+    },
+    {
+      "epoch": 2.809797106135316,
+      "grad_norm": 0.7744787931442261,
+      "learning_rate": 0.0002,
+      "loss": 0.7069,
+      "step": 17380
+    },
+    {
+      "epoch": 2.811413790316062,
+      "grad_norm": 0.6678640842437744,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 17390
+    },
+    {
+      "epoch": 2.813030474496807,
+      "grad_norm": 0.772676944732666,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 17400
+    },
+    {
+      "epoch": 2.8146471586775523,
+      "grad_norm": 0.7088175415992737,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 17410
+    },
+    {
+      "epoch": 2.8162638428582976,
+      "grad_norm": 0.8280573487281799,
+      "learning_rate": 0.0002,
+      "loss": 0.7086,
+      "step": 17420
+    },
+    {
+      "epoch": 2.817880527039043,
+      "grad_norm": 0.6665388345718384,
+      "learning_rate": 0.0002,
+      "loss": 0.6732,
+      "step": 17430
+    },
+    {
+      "epoch": 2.8194972112197885,
+      "grad_norm": 0.6427883505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 17440
+    },
+    {
+      "epoch": 2.8211138954005337,
+      "grad_norm": 0.9697760343551636,
+      "learning_rate": 0.0002,
+      "loss": 0.6972,
+      "step": 17450
+    },
+    {
+      "epoch": 2.822730579581279,
+      "grad_norm": 0.7573966383934021,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 17460
+    },
+    {
+      "epoch": 2.824347263762024,
+      "grad_norm": 0.878688633441925,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 17470
+    },
+    {
+      "epoch": 2.8259639479427694,
+      "grad_norm": 0.7752242684364319,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 17480
+    },
+    {
+      "epoch": 2.8275806321235146,
+      "grad_norm": 0.6135398745536804,
+      "learning_rate": 0.0002,
+      "loss": 0.6638,
+      "step": 17490
+    },
+    {
+      "epoch": 2.82919731630426,
+      "grad_norm": 0.6924924850463867,
+      "learning_rate": 0.0002,
+      "loss": 0.6829,
+      "step": 17500
+    },
+    {
+      "epoch": 2.830814000485005,
+      "grad_norm": 0.7471627593040466,
+      "learning_rate": 0.0002,
+      "loss": 0.6731,
+      "step": 17510
+    },
+    {
+      "epoch": 2.8324306846657503,
+      "grad_norm": 0.7145499587059021,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 17520
+    },
+    {
+      "epoch": 2.834047368846496,
+      "grad_norm": 0.7415414452552795,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17530
+    },
+    {
+      "epoch": 2.8356640530272412,
+      "grad_norm": 0.7328441739082336,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17540
+    },
+    {
+      "epoch": 2.8372807372079865,
+      "grad_norm": 0.8267839550971985,
+      "learning_rate": 0.0002,
+      "loss": 0.6866,
+      "step": 17550
+    },
+    {
+      "epoch": 2.8388974213887317,
+      "grad_norm": 0.8877885341644287,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17560
+    },
+    {
+      "epoch": 2.840514105569477,
+      "grad_norm": 0.857138454914093,
+      "learning_rate": 0.0002,
+      "loss": 0.7136,
+      "step": 17570
+    },
+    {
+      "epoch": 2.842130789750222,
+      "grad_norm": 0.8470779657363892,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 17580
+    },
+    {
+      "epoch": 2.843747473930968,
+      "grad_norm": 0.8553254008293152,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 17590
+    },
+    {
+      "epoch": 2.845364158111713,
+      "grad_norm": 0.8033196926116943,
+      "learning_rate": 0.0002,
+      "loss": 0.7297,
+      "step": 17600
+    },
+    {
+      "epoch": 2.8469808422924583,
+      "grad_norm": 0.7949087023735046,
+      "learning_rate": 0.0002,
+      "loss": 0.7062,
+      "step": 17610
+    },
+    {
+      "epoch": 2.8485975264732035,
+      "grad_norm": 0.9241406321525574,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 17620
+    },
+    {
+      "epoch": 2.8502142106539488,
+      "grad_norm": 0.7721285223960876,
+      "learning_rate": 0.0002,
+      "loss": 0.6601,
+      "step": 17630
+    },
+    {
+      "epoch": 2.851830894834694,
+      "grad_norm": 1.0246692895889282,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 17640
+    },
+    {
+      "epoch": 2.853447579015439,
+      "grad_norm": 0.9244589805603027,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 17650
+    },
+    {
+      "epoch": 2.8550642631961844,
+      "grad_norm": 0.7243508696556091,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 17660
+    },
+    {
+      "epoch": 2.8566809473769297,
+      "grad_norm": 0.8943371176719666,
+      "learning_rate": 0.0002,
+      "loss": 0.6471,
+      "step": 17670
+    },
+    {
+      "epoch": 2.8582976315576754,
+      "grad_norm": 0.6531758904457092,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 17680
+    },
+    {
+      "epoch": 2.8599143157384206,
+      "grad_norm": 0.8367000818252563,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 17690
+    },
+    {
+      "epoch": 2.861530999919166,
+      "grad_norm": 0.7868556380271912,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 17700
+    },
+    {
+      "epoch": 2.863147684099911,
+      "grad_norm": 0.7213859558105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 17710
+    },
+    {
+      "epoch": 2.8647643682806563,
+      "grad_norm": 0.7383931279182434,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 17720
+    },
+    {
+      "epoch": 2.8663810524614015,
+      "grad_norm": 0.7566812634468079,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 17730
+    },
+    {
+      "epoch": 2.867997736642147,
+      "grad_norm": 0.6930373311042786,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 17740
+    },
+    {
+      "epoch": 2.8696144208228924,
+      "grad_norm": 0.7911090850830078,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17750
+    },
+    {
+      "epoch": 2.8712311050036377,
+      "grad_norm": 0.8484548926353455,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 17760
+    },
+    {
+      "epoch": 2.872847789184383,
+      "grad_norm": 0.7647597193717957,
+      "learning_rate": 0.0002,
+      "loss": 0.6931,
+      "step": 17770
+    },
+    {
+      "epoch": 2.874464473365128,
+      "grad_norm": 0.8791151642799377,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 17780
+    },
+    {
+      "epoch": 2.8760811575458733,
+      "grad_norm": 0.7253178358078003,
+      "learning_rate": 0.0002,
+      "loss": 0.7078,
+      "step": 17790
+    },
+    {
+      "epoch": 2.8776978417266186,
+      "grad_norm": 0.7956077456474304,
+      "learning_rate": 0.0002,
+      "loss": 0.6474,
+      "step": 17800
+    },
+    {
+      "epoch": 2.879314525907364,
+      "grad_norm": 0.8657688498497009,
+      "learning_rate": 0.0002,
+      "loss": 0.6687,
+      "step": 17810
+    },
+    {
+      "epoch": 2.880931210088109,
+      "grad_norm": 0.7059141993522644,
+      "learning_rate": 0.0002,
+      "loss": 0.7171,
+      "step": 17820
+    },
+    {
+      "epoch": 2.8825478942688547,
+      "grad_norm": 0.8886896967887878,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 17830
+    },
+    {
+      "epoch": 2.8841645784496,
+      "grad_norm": 0.821032702922821,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 17840
+    },
+    {
+      "epoch": 2.885781262630345,
+      "grad_norm": 0.7183963656425476,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 17850
+    },
+    {
+      "epoch": 2.8873979468110904,
+      "grad_norm": 0.6222899556159973,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 17860
+    },
+    {
+      "epoch": 2.8890146309918356,
+      "grad_norm": 0.8187434077262878,
+      "learning_rate": 0.0002,
+      "loss": 0.6626,
+      "step": 17870
+    },
+    {
+      "epoch": 2.890631315172581,
+      "grad_norm": 0.9838479161262512,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 17880
+    },
+    {
+      "epoch": 2.8922479993533265,
+      "grad_norm": 0.7567742466926575,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 17890
+    },
+    {
+      "epoch": 2.893864683534072,
+      "grad_norm": 0.6875903606414795,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17900
+    },
+    {
+      "epoch": 2.895481367714817,
+      "grad_norm": 0.8043789267539978,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 17910
+    },
+    {
+      "epoch": 2.8970980518955622,
+      "grad_norm": 0.8062626719474792,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 17920
+    },
+    {
+      "epoch": 2.8987147360763075,
+      "grad_norm": 1.0251191854476929,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 17930
+    },
+    {
+      "epoch": 2.9003314202570527,
+      "grad_norm": 0.882253110408783,
+      "learning_rate": 0.0002,
+      "loss": 0.6605,
+      "step": 17940
+    },
+    {
+      "epoch": 2.901948104437798,
+      "grad_norm": 0.8683299422264099,
+      "learning_rate": 0.0002,
+      "loss": 0.6719,
+      "step": 17950
+    },
+    {
+      "epoch": 2.903564788618543,
+      "grad_norm": 0.7167282104492188,
+      "learning_rate": 0.0002,
+      "loss": 0.6896,
+      "step": 17960
+    },
+    {
+      "epoch": 2.9051814727992884,
+      "grad_norm": 0.7093694806098938,
+      "learning_rate": 0.0002,
+      "loss": 0.663,
+      "step": 17970
+    },
+    {
+      "epoch": 2.906798156980034,
+      "grad_norm": 0.8549879193305969,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 17980
+    },
+    {
+      "epoch": 2.9084148411607793,
+      "grad_norm": 0.6989606618881226,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 17990
+    },
+    {
+      "epoch": 2.9100315253415245,
+      "grad_norm": 0.9482976794242859,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 18000
+    },
+    {
+      "epoch": 2.9116482095222698,
+      "grad_norm": 0.7182440161705017,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 18010
+    },
+    {
+      "epoch": 2.913264893703015,
+      "grad_norm": 0.7732226252555847,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 18020
+    },
+    {
+      "epoch": 2.9148815778837607,
+      "grad_norm": 0.7936875224113464,
+      "learning_rate": 0.0002,
+      "loss": 0.7123,
+      "step": 18030
+    },
+    {
+      "epoch": 2.916498262064506,
+      "grad_norm": 0.8825615644454956,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 18040
+    },
+    {
+      "epoch": 2.918114946245251,
+      "grad_norm": 0.6778587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 18050
+    },
+    {
+      "epoch": 2.9197316304259964,
+      "grad_norm": 0.7529265880584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 18060
+    },
+    {
+      "epoch": 2.9213483146067416,
+      "grad_norm": 0.7111883163452148,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 18070
+    },
+    {
+      "epoch": 2.922964998787487,
+      "grad_norm": 0.7214767932891846,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 18080
+    },
+    {
+      "epoch": 2.924581682968232,
+      "grad_norm": 0.800417423248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 18090
+    },
+    {
+      "epoch": 2.9261983671489773,
+      "grad_norm": 1.248575210571289,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 18100
+    },
+    {
+      "epoch": 2.9278150513297225,
+      "grad_norm": 0.757788360118866,
+      "learning_rate": 0.0002,
+      "loss": 0.7004,
+      "step": 18110
+    },
+    {
+      "epoch": 2.9294317355104678,
+      "grad_norm": 1.0583995580673218,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 18120
+    },
+    {
+      "epoch": 2.9310484196912134,
+      "grad_norm": 0.8228777647018433,
+      "learning_rate": 0.0002,
+      "loss": 0.6365,
+      "step": 18130
+    },
+    {
+      "epoch": 2.9326651038719587,
+      "grad_norm": 0.8374035358428955,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 18140
+    },
+    {
+      "epoch": 2.934281788052704,
+      "grad_norm": 0.7976473569869995,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 18150
+    },
+    {
+      "epoch": 2.935898472233449,
+      "grad_norm": 0.8009907603263855,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 18160
+    },
+    {
+      "epoch": 2.9375151564141944,
+      "grad_norm": 0.835213303565979,
+      "learning_rate": 0.0002,
+      "loss": 0.7485,
+      "step": 18170
+    },
+    {
+      "epoch": 2.93913184059494,
+      "grad_norm": 0.7982219457626343,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 18180
+    },
+    {
+      "epoch": 2.9407485247756853,
+      "grad_norm": 0.7070978879928589,
+      "learning_rate": 0.0002,
+      "loss": 0.6348,
+      "step": 18190
+    },
+    {
+      "epoch": 2.9423652089564305,
+      "grad_norm": 0.8619440197944641,
+      "learning_rate": 0.0002,
+      "loss": 0.6608,
+      "step": 18200
+    },
+    {
+      "epoch": 2.9439818931371757,
+      "grad_norm": 0.6693987250328064,
+      "learning_rate": 0.0002,
+      "loss": 0.666,
+      "step": 18210
+    },
+    {
+      "epoch": 2.945598577317921,
+      "grad_norm": 0.6747021079063416,
+      "learning_rate": 0.0002,
+      "loss": 0.728,
+      "step": 18220
+    },
+    {
+      "epoch": 2.947215261498666,
+      "grad_norm": 0.860387921333313,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 18230
+    },
+    {
+      "epoch": 2.9488319456794114,
+      "grad_norm": 0.799976646900177,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 18240
+    },
+    {
+      "epoch": 2.9504486298601567,
+      "grad_norm": 0.7864769101142883,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 18250
+    },
+    {
+      "epoch": 2.952065314040902,
+      "grad_norm": 0.6713884472846985,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 18260
+    },
+    {
+      "epoch": 2.9536819982216476,
+      "grad_norm": 0.9031508564949036,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 18270
+    },
+    {
+      "epoch": 2.955298682402393,
+      "grad_norm": 0.7205073237419128,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 18280
+    },
+    {
+      "epoch": 2.956915366583138,
+      "grad_norm": 0.7746205925941467,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 18290
+    },
+    {
+      "epoch": 2.9585320507638833,
+      "grad_norm": 0.6533427834510803,
+      "learning_rate": 0.0002,
+      "loss": 0.6218,
+      "step": 18300
+    },
+    {
+      "epoch": 2.9601487349446285,
+      "grad_norm": 0.9083208441734314,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 18310
+    },
+    {
+      "epoch": 2.9617654191253737,
+      "grad_norm": 0.7446991801261902,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 18320
+    },
+    {
+      "epoch": 2.9633821033061194,
+      "grad_norm": 0.6514461636543274,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 18330
+    },
+    {
+      "epoch": 2.9649987874868646,
+      "grad_norm": 0.8580465912818909,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18340
+    },
+    {
+      "epoch": 2.96661547166761,
+      "grad_norm": 0.7074266076087952,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 18350
+    },
+    {
+      "epoch": 2.968232155848355,
+      "grad_norm": 0.899892270565033,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 18360
+    },
+    {
+      "epoch": 2.9698488400291003,
+      "grad_norm": 0.8217641711235046,
+      "learning_rate": 0.0002,
+      "loss": 0.7094,
+      "step": 18370
+    },
+    {
+      "epoch": 2.9714655242098456,
+      "grad_norm": 0.8611799478530884,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 18380
+    },
+    {
+      "epoch": 2.973082208390591,
+      "grad_norm": 0.6909302473068237,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18390
+    },
+    {
+      "epoch": 2.974698892571336,
+      "grad_norm": 0.6554358005523682,
+      "learning_rate": 0.0002,
+      "loss": 0.7247,
+      "step": 18400
+    },
+    {
+      "epoch": 2.9763155767520812,
+      "grad_norm": 0.7803071737289429,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 18410
+    },
+    {
+      "epoch": 2.977932260932827,
+      "grad_norm": 0.7838954925537109,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 18420
+    },
+    {
+      "epoch": 2.979548945113572,
+      "grad_norm": 0.7098495364189148,
+      "learning_rate": 0.0002,
+      "loss": 0.6522,
+      "step": 18430
+    },
+    {
+      "epoch": 2.9811656292943174,
+      "grad_norm": 0.8981785774230957,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 18440
+    },
+    {
+      "epoch": 2.9827823134750626,
+      "grad_norm": 0.7197171449661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6689,
+      "step": 18450
+    },
+    {
+      "epoch": 2.984398997655808,
+      "grad_norm": 0.793185293674469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 18460
+    },
+    {
+      "epoch": 2.986015681836553,
+      "grad_norm": 0.8531473875045776,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 18470
+    },
+    {
+      "epoch": 2.9876323660172988,
+      "grad_norm": 0.6627361178398132,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 18480
+    },
+    {
+      "epoch": 2.989249050198044,
+      "grad_norm": 0.5708155035972595,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 18490
+    },
+    {
+      "epoch": 2.990865734378789,
+      "grad_norm": 0.8227280378341675,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 18500
+    },
+    {
+      "epoch": 2.9924824185595345,
+      "grad_norm": 0.7102749943733215,
+      "learning_rate": 0.0002,
+      "loss": 0.6701,
+      "step": 18510
+    },
+    {
+      "epoch": 2.9940991027402797,
+      "grad_norm": 0.839485228061676,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 18520
+    },
+    {
+      "epoch": 2.995715786921025,
+      "grad_norm": 0.9038704037666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 18530
+    },
+    {
+      "epoch": 2.99733247110177,
+      "grad_norm": 0.8737510442733765,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 18540
+    },
+    {
+      "epoch": 2.9989491552825154,
+      "grad_norm": 0.7323142886161804,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 18550
+    },
+    {
+      "epoch": 2.9999191657909625,
+      "eval_loss": 1.1262480020523071,
+      "eval_runtime": 122.0868,
+      "eval_samples_per_second": 6.004,
+      "eval_steps_per_second": 0.754,
+      "step": 18556
+    },
+    {
+      "epoch": 3.000565839463261,
+      "grad_norm": 0.8465463519096375,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 18560
+    },
+    {
+      "epoch": 3.0021825236440063,
+      "grad_norm": 0.9134138822555542,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 18570
+    },
+    {
+      "epoch": 3.0037992078247515,
+      "grad_norm": 0.760715126991272,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 18580
+    },
+    {
+      "epoch": 3.0054158920054967,
+      "grad_norm": 0.9208743572235107,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18590
+    },
+    {
+      "epoch": 3.007032576186242,
+      "grad_norm": 0.9232364892959595,
+      "learning_rate": 0.0002,
+      "loss": 0.5731,
+      "step": 18600
+    },
+    {
+      "epoch": 3.008649260366987,
+      "grad_norm": 1.1881544589996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6299,
+      "step": 18610
+    },
+    {
+      "epoch": 3.0102659445477324,
+      "grad_norm": 0.9372987747192383,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 18620
+    },
+    {
+      "epoch": 3.0118826287284777,
+      "grad_norm": 0.6900241374969482,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 18630
+    },
+    {
+      "epoch": 3.0134993129092233,
+      "grad_norm": 0.8451071381568909,
+      "learning_rate": 0.0002,
+      "loss": 0.5256,
+      "step": 18640
+    },
+    {
+      "epoch": 3.0151159970899686,
+      "grad_norm": 0.7763112187385559,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 18650
+    },
+    {
+      "epoch": 3.016732681270714,
+      "grad_norm": 1.043653964996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 18660
+    },
+    {
+      "epoch": 3.018349365451459,
+      "grad_norm": 1.0170660018920898,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 18670
+    },
+    {
+      "epoch": 3.0199660496322043,
+      "grad_norm": 0.7534180283546448,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 18680
+    },
+    {
+      "epoch": 3.0215827338129495,
+      "grad_norm": 0.7507367730140686,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 18690
+    },
+    {
+      "epoch": 3.0231994179936947,
+      "grad_norm": 0.7861620187759399,
+      "learning_rate": 0.0002,
+      "loss": 0.6201,
+      "step": 18700
+    },
+    {
+      "epoch": 3.0248161021744404,
+      "grad_norm": 1.0580339431762695,
+      "learning_rate": 0.0002,
+      "loss": 0.5802,
+      "step": 18710
+    },
+    {
+      "epoch": 3.0264327863551856,
+      "grad_norm": 0.7542710900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 18720
+    },
+    {
+      "epoch": 3.028049470535931,
+      "grad_norm": 0.8189544677734375,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 18730
+    },
+    {
+      "epoch": 3.029666154716676,
+      "grad_norm": 0.9126611351966858,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 18740
+    },
+    {
+      "epoch": 3.0312828388974213,
+      "grad_norm": 0.8891341686248779,
+      "learning_rate": 0.0002,
+      "loss": 0.6443,
+      "step": 18750
+    },
+    {
+      "epoch": 3.0328995230781666,
+      "grad_norm": 0.8419283032417297,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 18760
+    },
+    {
+      "epoch": 3.034516207258912,
+      "grad_norm": 0.8048048615455627,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18770
+    },
+    {
+      "epoch": 3.0361328914396575,
+      "grad_norm": 0.7820217609405518,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 18780
+    },
+    {
+      "epoch": 3.0377495756204027,
+      "grad_norm": 0.854721188545227,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 18790
+    },
+    {
+      "epoch": 3.039366259801148,
+      "grad_norm": 0.912092924118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5784,
+      "step": 18800
+    },
+    {
+      "epoch": 3.040982943981893,
+      "grad_norm": 0.6596226096153259,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 18810
+    },
+    {
+      "epoch": 3.0425996281626384,
+      "grad_norm": 0.6351348757743835,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 18820
+    },
+    {
+      "epoch": 3.0442163123433836,
+      "grad_norm": 0.778188943862915,
+      "learning_rate": 0.0002,
+      "loss": 0.5953,
+      "step": 18830
+    },
+    {
+      "epoch": 3.045832996524129,
+      "grad_norm": 0.68234783411026,
+      "learning_rate": 0.0002,
+      "loss": 0.602,
+      "step": 18840
+    },
+    {
+      "epoch": 3.047449680704874,
+      "grad_norm": 0.998628556728363,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 18850
+    },
+    {
+      "epoch": 3.0490663648856198,
+      "grad_norm": 0.7393841743469238,
+      "learning_rate": 0.0002,
+      "loss": 0.6231,
+      "step": 18860
+    },
+    {
+      "epoch": 3.050683049066365,
+      "grad_norm": 0.84438556432724,
+      "learning_rate": 0.0002,
+      "loss": 0.568,
+      "step": 18870
+    },
+    {
+      "epoch": 3.0522997332471102,
+      "grad_norm": 0.8857501745223999,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 18880
+    },
+    {
+      "epoch": 3.0539164174278555,
+      "grad_norm": 0.7208474278450012,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 18890
+    },
+    {
+      "epoch": 3.0555331016086007,
+      "grad_norm": 0.7135229110717773,
+      "learning_rate": 0.0002,
+      "loss": 0.5998,
+      "step": 18900
+    },
+    {
+      "epoch": 3.057149785789346,
+      "grad_norm": 0.9130001664161682,
+      "learning_rate": 0.0002,
+      "loss": 0.5575,
+      "step": 18910
+    },
+    {
+      "epoch": 3.058766469970091,
+      "grad_norm": 0.9001716375350952,
+      "learning_rate": 0.0002,
+      "loss": 0.5955,
+      "step": 18920
+    },
+    {
+      "epoch": 3.060383154150837,
+      "grad_norm": 0.8667559623718262,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 18930
+    },
+    {
+      "epoch": 3.061999838331582,
+      "grad_norm": 0.8943959474563599,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18940
+    },
+    {
+      "epoch": 3.0636165225123273,
+      "grad_norm": 0.8298377990722656,
+      "learning_rate": 0.0002,
+      "loss": 0.5978,
+      "step": 18950
+    },
+    {
+      "epoch": 3.0652332066930725,
+      "grad_norm": 0.7935267686843872,
+      "learning_rate": 0.0002,
+      "loss": 0.5782,
+      "step": 18960
+    },
+    {
+      "epoch": 3.0668498908738178,
+      "grad_norm": 1.1506379842758179,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 18970
+    },
+    {
+      "epoch": 3.068466575054563,
+      "grad_norm": 0.7693049907684326,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18980
+    },
+    {
+      "epoch": 3.0700832592353082,
+      "grad_norm": 0.8040135502815247,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 18990
+    },
+    {
+      "epoch": 3.0716999434160535,
+      "grad_norm": 0.828404426574707,
+      "learning_rate": 0.0002,
+      "loss": 0.5541,
+      "step": 19000
+    },
+    {
+      "epoch": 3.073316627596799,
+      "grad_norm": 0.8811164498329163,
+      "learning_rate": 0.0002,
+      "loss": 0.6048,
+      "step": 19010
+    },
+    {
+      "epoch": 3.0749333117775444,
+      "grad_norm": 1.036205768585205,
+      "learning_rate": 0.0002,
+      "loss": 0.5845,
+      "step": 19020
+    },
+    {
+      "epoch": 3.0765499959582896,
+      "grad_norm": 0.8857285976409912,
+      "learning_rate": 0.0002,
+      "loss": 0.5838,
+      "step": 19030
+    },
+    {
+      "epoch": 3.078166680139035,
+      "grad_norm": 0.8392079472541809,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19040
+    },
+    {
+      "epoch": 3.07978336431978,
+      "grad_norm": 1.0287401676177979,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 19050
+    },
+    {
+      "epoch": 3.0814000485005253,
+      "grad_norm": 1.0086315870285034,
+      "learning_rate": 0.0002,
+      "loss": 0.5964,
+      "step": 19060
+    },
+    {
+      "epoch": 3.0830167326812705,
+      "grad_norm": 0.9245324730873108,
+      "learning_rate": 0.0002,
+      "loss": 0.5567,
+      "step": 19070
+    },
+    {
+      "epoch": 3.084633416862016,
+      "grad_norm": 0.8680877089500427,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 19080
+    },
+    {
+      "epoch": 3.0862501010427614,
+      "grad_norm": 0.8814793825149536,
+      "learning_rate": 0.0002,
+      "loss": 0.5611,
+      "step": 19090
+    },
+    {
+      "epoch": 3.0878667852235067,
+      "grad_norm": 0.9234458208084106,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19100
+    },
+    {
+      "epoch": 3.089483469404252,
+      "grad_norm": 1.1291664838790894,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 19110
+    },
+    {
+      "epoch": 3.091100153584997,
+      "grad_norm": 0.9191402792930603,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 19120
+    },
+    {
+      "epoch": 3.0927168377657424,
+      "grad_norm": 0.7103154063224792,
+      "learning_rate": 0.0002,
+      "loss": 0.5856,
+      "step": 19130
+    },
+    {
+      "epoch": 3.0943335219464876,
+      "grad_norm": 0.9368883967399597,
+      "learning_rate": 0.0002,
+      "loss": 0.6479,
+      "step": 19140
+    },
+    {
+      "epoch": 3.095950206127233,
+      "grad_norm": 0.9676656723022461,
+      "learning_rate": 0.0002,
+      "loss": 0.6167,
+      "step": 19150
+    },
+    {
+      "epoch": 3.0975668903079785,
+      "grad_norm": 0.8739792704582214,
+      "learning_rate": 0.0002,
+      "loss": 0.5794,
+      "step": 19160
+    },
+    {
+      "epoch": 3.0991835744887237,
+      "grad_norm": 0.8530174493789673,
+      "learning_rate": 0.0002,
+      "loss": 0.6112,
+      "step": 19170
+    },
+    {
+      "epoch": 3.100800258669469,
+      "grad_norm": 0.794945478439331,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 19180
+    },
+    {
+      "epoch": 3.102416942850214,
+      "grad_norm": 0.9508888125419617,
+      "learning_rate": 0.0002,
+      "loss": 0.5928,
+      "step": 19190
+    },
+    {
+      "epoch": 3.1040336270309594,
+      "grad_norm": 1.0599955320358276,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 19200
+    },
+    {
+      "epoch": 3.1056503112117047,
+      "grad_norm": 1.0673625469207764,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 19210
+    },
+    {
+      "epoch": 3.10726699539245,
+      "grad_norm": 0.7739115953445435,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 19220
+    },
+    {
+      "epoch": 3.1088836795731956,
+      "grad_norm": 0.9884951114654541,
+      "learning_rate": 0.0002,
+      "loss": 0.6046,
+      "step": 19230
+    },
+    {
+      "epoch": 3.110500363753941,
+      "grad_norm": 0.862260103225708,
+      "learning_rate": 0.0002,
+      "loss": 0.5932,
+      "step": 19240
+    },
+    {
+      "epoch": 3.112117047934686,
+      "grad_norm": 0.7690284848213196,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 19250
+    },
+    {
+      "epoch": 3.1137337321154313,
+      "grad_norm": 0.8758958578109741,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 19260
+    },
+    {
+      "epoch": 3.1153504162961765,
+      "grad_norm": 1.0356395244598389,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 19270
+    },
+    {
+      "epoch": 3.1169671004769217,
+      "grad_norm": 0.6950937509536743,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 19280
+    },
+    {
+      "epoch": 3.118583784657667,
+      "grad_norm": 0.760998010635376,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19290
+    },
+    {
+      "epoch": 3.1202004688384126,
+      "grad_norm": 0.9335789084434509,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 19300
+    },
+    {
+      "epoch": 3.121817153019158,
+      "grad_norm": 0.9636204242706299,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 19310
+    },
+    {
+      "epoch": 3.123433837199903,
+      "grad_norm": 1.0820997953414917,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 19320
+    },
+    {
+      "epoch": 3.1250505213806483,
+      "grad_norm": 0.7333487272262573,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 19330
+    },
+    {
+      "epoch": 3.1266672055613935,
+      "grad_norm": 1.0417509078979492,
+      "learning_rate": 0.0002,
+      "loss": 0.6178,
+      "step": 19340
+    },
+    {
+      "epoch": 3.128283889742139,
+      "grad_norm": 0.9267749190330505,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 19350
+    },
+    {
+      "epoch": 3.129900573922884,
+      "grad_norm": 0.777798593044281,
+      "learning_rate": 0.0002,
+      "loss": 0.6063,
+      "step": 19360
+    },
+    {
+      "epoch": 3.1315172581036297,
+      "grad_norm": 0.8425456881523132,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 19370
+    },
+    {
+      "epoch": 3.133133942284375,
+      "grad_norm": 0.9617102146148682,
+      "learning_rate": 0.0002,
+      "loss": 0.6042,
+      "step": 19380
+    },
+    {
+      "epoch": 3.13475062646512,
+      "grad_norm": 1.0052828788757324,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 19390
+    },
+    {
+      "epoch": 3.1363673106458654,
+      "grad_norm": 0.7637009024620056,
+      "learning_rate": 0.0002,
+      "loss": 0.5713,
+      "step": 19400
+    },
+    {
+      "epoch": 3.1379839948266106,
+      "grad_norm": 0.7958088517189026,
+      "learning_rate": 0.0002,
+      "loss": 0.5497,
+      "step": 19410
+    },
+    {
+      "epoch": 3.139600679007356,
+      "grad_norm": 0.9161727428436279,
+      "learning_rate": 0.0002,
+      "loss": 0.6283,
+      "step": 19420
+    },
+    {
+      "epoch": 3.141217363188101,
+      "grad_norm": 0.8402149677276611,
+      "learning_rate": 0.0002,
+      "loss": 0.5638,
+      "step": 19430
+    },
+    {
+      "epoch": 3.1428340473688463,
+      "grad_norm": 1.0056525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.5848,
+      "step": 19440
+    },
+    {
+      "epoch": 3.144450731549592,
+      "grad_norm": 1.0129190683364868,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 19450
+    },
+    {
+      "epoch": 3.146067415730337,
+      "grad_norm": 0.790825366973877,
+      "learning_rate": 0.0002,
+      "loss": 0.5808,
+      "step": 19460
+    },
+    {
+      "epoch": 3.1476840999110824,
+      "grad_norm": 1.441665530204773,
+      "learning_rate": 0.0002,
+      "loss": 0.5607,
+      "step": 19470
+    },
+    {
+      "epoch": 3.1493007840918277,
+      "grad_norm": 0.7846331596374512,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19480
+    },
+    {
+      "epoch": 3.150917468272573,
+      "grad_norm": 0.7915332913398743,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 19490
+    },
+    {
+      "epoch": 3.152534152453318,
+      "grad_norm": 0.933982253074646,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 19500
+    },
+    {
+      "epoch": 3.1541508366340634,
+      "grad_norm": 1.038408637046814,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 19510
+    },
+    {
+      "epoch": 3.155767520814809,
+      "grad_norm": 1.018935203552246,
+      "learning_rate": 0.0002,
+      "loss": 0.6271,
+      "step": 19520
+    },
+    {
+      "epoch": 3.1573842049955543,
+      "grad_norm": 0.9618112444877625,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 19530
+    },
+    {
+      "epoch": 3.1590008891762995,
+      "grad_norm": 0.8900452852249146,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 19540
+    },
+    {
+      "epoch": 3.1606175733570447,
+      "grad_norm": 0.8254160284996033,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 19550
+    },
+    {
+      "epoch": 3.16223425753779,
+      "grad_norm": 1.004376769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19560
+    },
+    {
+      "epoch": 3.163850941718535,
+      "grad_norm": 1.0490446090698242,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 19570
+    },
+    {
+      "epoch": 3.1654676258992804,
+      "grad_norm": 0.7387403845787048,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19580
+    },
+    {
+      "epoch": 3.1670843100800257,
+      "grad_norm": 0.7611538171768188,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 19590
+    },
+    {
+      "epoch": 3.1687009942607713,
+      "grad_norm": 0.8239886164665222,
+      "learning_rate": 0.0002,
+      "loss": 0.5937,
+      "step": 19600
+    },
+    {
+      "epoch": 3.1703176784415166,
+      "grad_norm": 0.9327243566513062,
+      "learning_rate": 0.0002,
+      "loss": 0.6068,
+      "step": 19610
+    },
+    {
+      "epoch": 3.171934362622262,
+      "grad_norm": 0.9662560224533081,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 19620
+    },
+    {
+      "epoch": 3.173551046803007,
+      "grad_norm": 0.9183341860771179,
+      "learning_rate": 0.0002,
+      "loss": 0.5988,
+      "step": 19630
+    },
+    {
+      "epoch": 3.1751677309837523,
+      "grad_norm": 0.875066876411438,
+      "learning_rate": 0.0002,
+      "loss": 0.5909,
+      "step": 19640
+    },
+    {
+      "epoch": 3.1767844151644975,
+      "grad_norm": 0.8567508459091187,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 19650
+    },
+    {
+      "epoch": 3.1784010993452427,
+      "grad_norm": 0.6805780529975891,
+      "learning_rate": 0.0002,
+      "loss": 0.5805,
+      "step": 19660
+    },
+    {
+      "epoch": 3.1800177835259884,
+      "grad_norm": 0.8776944279670715,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 19670
+    },
+    {
+      "epoch": 3.1816344677067336,
+      "grad_norm": 0.9036329984664917,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 19680
+    },
+    {
+      "epoch": 3.183251151887479,
+      "grad_norm": 0.8527372479438782,
+      "learning_rate": 0.0002,
+      "loss": 0.6238,
+      "step": 19690
+    },
+    {
+      "epoch": 3.184867836068224,
+      "grad_norm": 1.1045585870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 19700
+    },
+    {
+      "epoch": 3.1864845202489693,
+      "grad_norm": 0.9213830828666687,
+      "learning_rate": 0.0002,
+      "loss": 0.5491,
+      "step": 19710
+    },
+    {
+      "epoch": 3.1881012044297146,
+      "grad_norm": 0.8865814805030823,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 19720
+    },
+    {
+      "epoch": 3.18971788861046,
+      "grad_norm": 0.7939388751983643,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19730
+    },
+    {
+      "epoch": 3.191334572791205,
+      "grad_norm": 0.6966729760169983,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 19740
+    },
+    {
+      "epoch": 3.1929512569719507,
+      "grad_norm": 0.8023673295974731,
+      "learning_rate": 0.0002,
+      "loss": 0.5839,
+      "step": 19750
+    },
+    {
+      "epoch": 3.194567941152696,
+      "grad_norm": 0.7992037534713745,
+      "learning_rate": 0.0002,
+      "loss": 0.6267,
+      "step": 19760
+    },
+    {
+      "epoch": 3.196184625333441,
+      "grad_norm": 0.7412247657775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 19770
+    },
+    {
+      "epoch": 3.1978013095141864,
+      "grad_norm": 0.9598729014396667,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 19780
+    },
+    {
+      "epoch": 3.1994179936949316,
+      "grad_norm": 0.8331366777420044,
+      "learning_rate": 0.0002,
+      "loss": 0.5685,
+      "step": 19790
+    },
+    {
+      "epoch": 3.201034677875677,
+      "grad_norm": 0.8939169645309448,
+      "learning_rate": 0.0002,
+      "loss": 0.6104,
+      "step": 19800
+    },
+    {
+      "epoch": 3.202651362056422,
+      "grad_norm": 0.9219734072685242,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 19810
+    },
+    {
+      "epoch": 3.2042680462371678,
+      "grad_norm": 0.869490385055542,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19820
+    },
+    {
+      "epoch": 3.205884730417913,
+      "grad_norm": 0.8989706635475159,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 19830
+    },
+    {
+      "epoch": 3.2075014145986582,
+      "grad_norm": 0.8477165102958679,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 19840
+    },
+    {
+      "epoch": 3.2091180987794035,
+      "grad_norm": 0.8720678687095642,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 19850
+    },
+    {
+      "epoch": 3.2107347829601487,
+      "grad_norm": 0.861406683921814,
+      "learning_rate": 0.0002,
+      "loss": 0.5694,
+      "step": 19860
+    },
+    {
+      "epoch": 3.212351467140894,
+      "grad_norm": 0.8228686451911926,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 19870
+    },
+    {
+      "epoch": 3.213968151321639,
+      "grad_norm": 0.7936596870422363,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19880
+    },
+    {
+      "epoch": 3.2155848355023844,
+      "grad_norm": 1.097377896308899,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 19890
+    },
+    {
+      "epoch": 3.21720151968313,
+      "grad_norm": 0.9544782638549805,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 19900
+    },
+    {
+      "epoch": 3.2188182038638753,
+      "grad_norm": 0.8240751624107361,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 19910
+    },
+    {
+      "epoch": 3.2204348880446205,
+      "grad_norm": 0.8332096338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 19920
+    },
+    {
+      "epoch": 3.2220515722253658,
+      "grad_norm": 1.0954567193984985,
+      "learning_rate": 0.0002,
+      "loss": 0.6274,
+      "step": 19930
+    },
+    {
+      "epoch": 3.223668256406111,
+      "grad_norm": 0.7790525555610657,
+      "learning_rate": 0.0002,
+      "loss": 0.652,
+      "step": 19940
+    },
+    {
+      "epoch": 3.225284940586856,
+      "grad_norm": 0.7966814041137695,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19950
+    },
+    {
+      "epoch": 3.2269016247676015,
+      "grad_norm": 0.9751881957054138,
+      "learning_rate": 0.0002,
+      "loss": 0.5911,
+      "step": 19960
+    },
+    {
+      "epoch": 3.228518308948347,
+      "grad_norm": 0.9856047630310059,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 19970
+    },
+    {
+      "epoch": 3.2301349931290924,
+      "grad_norm": 1.3062353134155273,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 19980
+    },
+    {
+      "epoch": 3.2317516773098376,
+      "grad_norm": 0.9510692358016968,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 19990
+    },
+    {
+      "epoch": 3.233368361490583,
+      "grad_norm": 0.8630342483520508,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 20000
+    },
+    {
+      "epoch": 3.234985045671328,
+      "grad_norm": 0.8966519236564636,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20010
+    },
+    {
+      "epoch": 3.2366017298520733,
+      "grad_norm": 0.7093510627746582,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 20020
+    },
+    {
+      "epoch": 3.2382184140328185,
+      "grad_norm": 0.7771096229553223,
+      "learning_rate": 0.0002,
+      "loss": 0.585,
+      "step": 20030
+    },
+    {
+      "epoch": 3.2398350982135637,
+      "grad_norm": 0.841058075428009,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 20040
+    },
+    {
+      "epoch": 3.2414517823943094,
+      "grad_norm": 0.909712553024292,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 20050
+    },
+    {
+      "epoch": 3.2430684665750547,
+      "grad_norm": 0.8321019411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20060
+    },
+    {
+      "epoch": 3.2446851507558,
+      "grad_norm": 0.779901921749115,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 20070
+    },
+    {
+      "epoch": 3.246301834936545,
+      "grad_norm": 0.6249170303344727,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 20080
+    },
+    {
+      "epoch": 3.2479185191172903,
+      "grad_norm": 0.8000940680503845,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 20090
+    },
+    {
+      "epoch": 3.2495352032980356,
+      "grad_norm": 0.7627735137939453,
+      "learning_rate": 0.0002,
+      "loss": 0.6273,
+      "step": 20100
+    },
+    {
+      "epoch": 3.2511518874787813,
+      "grad_norm": 0.8780747056007385,
+      "learning_rate": 0.0002,
+      "loss": 0.6223,
+      "step": 20110
+    },
+    {
+      "epoch": 3.2527685716595265,
+      "grad_norm": 0.772037148475647,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 20120
+    },
+    {
+      "epoch": 3.2543852558402717,
+      "grad_norm": 1.0086580514907837,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 20130
+    },
+    {
+      "epoch": 3.256001940021017,
+      "grad_norm": 0.9360289573669434,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20140
+    },
+    {
+      "epoch": 3.257618624201762,
+      "grad_norm": 1.2099586725234985,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20150
+    },
+    {
+      "epoch": 3.2592353083825074,
+      "grad_norm": 0.8368481397628784,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 20160
+    },
+    {
+      "epoch": 3.2608519925632526,
+      "grad_norm": 0.7391039133071899,
+      "learning_rate": 0.0002,
+      "loss": 0.5626,
+      "step": 20170
+    },
+    {
+      "epoch": 3.262468676743998,
+      "grad_norm": 0.9122273325920105,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 20180
+    },
+    {
+      "epoch": 3.264085360924743,
+      "grad_norm": 0.8502281904220581,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 20190
+    },
+    {
+      "epoch": 3.265702045105489,
+      "grad_norm": 1.0926852226257324,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 20200
+    },
+    {
+      "epoch": 3.267318729286234,
+      "grad_norm": 0.7902828454971313,
+      "learning_rate": 0.0002,
+      "loss": 0.6027,
+      "step": 20210
+    },
+    {
+      "epoch": 3.2689354134669792,
+      "grad_norm": 0.8724729418754578,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20220
+    },
+    {
+      "epoch": 3.2705520976477245,
+      "grad_norm": 0.8469277024269104,
+      "learning_rate": 0.0002,
+      "loss": 0.6242,
+      "step": 20230
+    },
+    {
+      "epoch": 3.2721687818284697,
+      "grad_norm": 0.8865092992782593,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 20240
+    },
+    {
+      "epoch": 3.273785466009215,
+      "grad_norm": 1.0979334115982056,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20250
+    },
+    {
+      "epoch": 3.2754021501899606,
+      "grad_norm": 1.0860793590545654,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 20260
+    },
+    {
+      "epoch": 3.277018834370706,
+      "grad_norm": 0.981745183467865,
+      "learning_rate": 0.0002,
+      "loss": 0.6105,
+      "step": 20270
+    },
+    {
+      "epoch": 3.278635518551451,
+      "grad_norm": 0.9155020713806152,
+      "learning_rate": 0.0002,
+      "loss": 0.627,
+      "step": 20280
+    },
+    {
+      "epoch": 3.2802522027321963,
+      "grad_norm": 0.8436718583106995,
+      "learning_rate": 0.0002,
+      "loss": 0.5899,
+      "step": 20290
+    },
+    {
+      "epoch": 3.2818688869129415,
+      "grad_norm": 1.0329409837722778,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 20300
+    },
+    {
+      "epoch": 3.2834855710936868,
+      "grad_norm": 0.9876394271850586,
+      "learning_rate": 0.0002,
+      "loss": 0.6,
+      "step": 20310
+    },
+    {
+      "epoch": 3.285102255274432,
+      "grad_norm": 0.8052917718887329,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 20320
+    },
+    {
+      "epoch": 3.2867189394551772,
+      "grad_norm": 0.8390680551528931,
+      "learning_rate": 0.0002,
+      "loss": 0.5949,
+      "step": 20330
+    },
+    {
+      "epoch": 3.288335623635923,
+      "grad_norm": 0.9515735507011414,
+      "learning_rate": 0.0002,
+      "loss": 0.6492,
+      "step": 20340
+    },
+    {
+      "epoch": 3.289952307816668,
+      "grad_norm": 0.8028870224952698,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 20350
+    },
+    {
+      "epoch": 3.2915689919974134,
+      "grad_norm": 0.862592339515686,
+      "learning_rate": 0.0002,
+      "loss": 0.634,
+      "step": 20360
+    },
+    {
+      "epoch": 3.2931856761781586,
+      "grad_norm": 0.7451621890068054,
+      "learning_rate": 0.0002,
+      "loss": 0.6345,
+      "step": 20370
+    },
+    {
+      "epoch": 3.294802360358904,
+      "grad_norm": 0.8966776728630066,
+      "learning_rate": 0.0002,
+      "loss": 0.6458,
+      "step": 20380
+    },
+    {
+      "epoch": 3.296419044539649,
+      "grad_norm": 0.9289216995239258,
+      "learning_rate": 0.0002,
+      "loss": 0.5967,
+      "step": 20390
+    },
+    {
+      "epoch": 3.2980357287203943,
+      "grad_norm": 0.9649626612663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6599,
+      "step": 20400
+    },
+    {
+      "epoch": 3.29965241290114,
+      "grad_norm": 1.1953798532485962,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 20410
+    },
+    {
+      "epoch": 3.301269097081885,
+      "grad_norm": 0.8929083943367004,
+      "learning_rate": 0.0002,
+      "loss": 0.5997,
+      "step": 20420
+    },
+    {
+      "epoch": 3.3028857812626304,
+      "grad_norm": 0.8922014236450195,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 20430
+    },
+    {
+      "epoch": 3.3045024654433757,
+      "grad_norm": 0.9754860401153564,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 20440
+    },
+    {
+      "epoch": 3.306119149624121,
+      "grad_norm": 0.8873140215873718,
+      "learning_rate": 0.0002,
+      "loss": 0.5653,
+      "step": 20450
+    },
+    {
+      "epoch": 3.307735833804866,
+      "grad_norm": 0.857271671295166,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20460
+    },
+    {
+      "epoch": 3.3093525179856114,
+      "grad_norm": 0.9022141098976135,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 20470
+    },
+    {
+      "epoch": 3.3109692021663566,
+      "grad_norm": 0.8614798188209534,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 20480
+    },
+    {
+      "epoch": 3.3125858863471023,
+      "grad_norm": 0.8838164210319519,
+      "learning_rate": 0.0002,
+      "loss": 0.6254,
+      "step": 20490
+    },
+    {
+      "epoch": 3.3142025705278475,
+      "grad_norm": 0.8709736466407776,
+      "learning_rate": 0.0002,
+      "loss": 0.5849,
+      "step": 20500
+    },
+    {
+      "epoch": 3.3158192547085927,
+      "grad_norm": 0.9533300995826721,
+      "learning_rate": 0.0002,
+      "loss": 0.6146,
+      "step": 20510
+    },
+    {
+      "epoch": 3.317435938889338,
+      "grad_norm": 0.8259269595146179,
+      "learning_rate": 0.0002,
+      "loss": 0.6029,
+      "step": 20520
+    },
+    {
+      "epoch": 3.319052623070083,
+      "grad_norm": 0.8607608079910278,
+      "learning_rate": 0.0002,
+      "loss": 0.6268,
+      "step": 20530
+    },
+    {
+      "epoch": 3.3206693072508284,
+      "grad_norm": 1.0863020420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.5676,
+      "step": 20540
+    },
+    {
+      "epoch": 3.3222859914315737,
+      "grad_norm": 1.011489987373352,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 20550
+    },
+    {
+      "epoch": 3.3239026756123193,
+      "grad_norm": 0.6952177882194519,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 20560
+    },
+    {
+      "epoch": 3.3255193597930646,
+      "grad_norm": 0.9638974070549011,
+      "learning_rate": 0.0002,
+      "loss": 0.6229,
+      "step": 20570
+    },
+    {
+      "epoch": 3.32713604397381,
+      "grad_norm": 1.0310138463974,
+      "learning_rate": 0.0002,
+      "loss": 0.5882,
+      "step": 20580
+    },
+    {
+      "epoch": 3.328752728154555,
+      "grad_norm": 0.9371318221092224,
+      "learning_rate": 0.0002,
+      "loss": 0.594,
+      "step": 20590
+    },
+    {
+      "epoch": 3.3303694123353003,
+      "grad_norm": 0.8756691813468933,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 20600
+    },
+    {
+      "epoch": 3.3319860965160455,
+      "grad_norm": 1.054175853729248,
+      "learning_rate": 0.0002,
+      "loss": 0.5994,
+      "step": 20610
+    },
+    {
+      "epoch": 3.3336027806967907,
+      "grad_norm": 0.9074128270149231,
+      "learning_rate": 0.0002,
+      "loss": 0.6169,
+      "step": 20620
+    },
+    {
+      "epoch": 3.335219464877536,
+      "grad_norm": 0.906900942325592,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20630
+    },
+    {
+      "epoch": 3.3368361490582816,
+      "grad_norm": 0.8689333200454712,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 20640
+    },
+    {
+      "epoch": 3.338452833239027,
+      "grad_norm": 0.9889747500419617,
+      "learning_rate": 0.0002,
+      "loss": 0.6079,
+      "step": 20650
+    },
+    {
+      "epoch": 3.340069517419772,
+      "grad_norm": 1.0685805082321167,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20660
+    },
+    {
+      "epoch": 3.3416862016005173,
+      "grad_norm": 0.7495010495185852,
+      "learning_rate": 0.0002,
+      "loss": 0.6091,
+      "step": 20670
+    },
+    {
+      "epoch": 3.3433028857812626,
+      "grad_norm": 0.8747848272323608,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 20680
+    },
+    {
+      "epoch": 3.344919569962008,
+      "grad_norm": 0.9762673377990723,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 20690
+    },
+    {
+      "epoch": 3.346536254142753,
+      "grad_norm": 1.0284489393234253,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 20700
+    },
+    {
+      "epoch": 3.3481529383234987,
+      "grad_norm": 0.7293812036514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20710
+    },
+    {
+      "epoch": 3.349769622504244,
+      "grad_norm": 0.8330199122428894,
+      "learning_rate": 0.0002,
+      "loss": 0.609,
+      "step": 20720
+    },
+    {
+      "epoch": 3.351386306684989,
+      "grad_norm": 0.9808499217033386,
+      "learning_rate": 0.0002,
+      "loss": 0.5729,
+      "step": 20730
+    },
+    {
+      "epoch": 3.3530029908657344,
+      "grad_norm": 0.9508825540542603,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 20740
+    },
+    {
+      "epoch": 3.3546196750464796,
+      "grad_norm": 0.790483832359314,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 20750
+    },
+    {
+      "epoch": 3.356236359227225,
+      "grad_norm": 1.022793173789978,
+      "learning_rate": 0.0002,
+      "loss": 0.6327,
+      "step": 20760
+    },
+    {
+      "epoch": 3.35785304340797,
+      "grad_norm": 0.8318950533866882,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 20770
+    },
+    {
+      "epoch": 3.3594697275887153,
+      "grad_norm": 0.7980858087539673,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 20780
+    },
+    {
+      "epoch": 3.361086411769461,
+      "grad_norm": 0.8114802241325378,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 20790
+    },
+    {
+      "epoch": 3.3627030959502062,
+      "grad_norm": 0.8522519469261169,
+      "learning_rate": 0.0002,
+      "loss": 0.6017,
+      "step": 20800
+    },
+    {
+      "epoch": 3.3643197801309515,
+      "grad_norm": 0.9142431616783142,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 20810
+    },
+    {
+      "epoch": 3.3659364643116967,
+      "grad_norm": 0.771170437335968,
+      "learning_rate": 0.0002,
+      "loss": 0.6331,
+      "step": 20820
+    },
+    {
+      "epoch": 3.367553148492442,
+      "grad_norm": 1.0628231763839722,
+      "learning_rate": 0.0002,
+      "loss": 0.5879,
+      "step": 20830
+    },
+    {
+      "epoch": 3.369169832673187,
+      "grad_norm": 0.9384352564811707,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 20840
+    },
+    {
+      "epoch": 3.370786516853933,
+      "grad_norm": 1.1286591291427612,
+      "learning_rate": 0.0002,
+      "loss": 0.6292,
+      "step": 20850
+    },
+    {
+      "epoch": 3.372403201034678,
+      "grad_norm": 1.1349513530731201,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 20860
+    },
+    {
+      "epoch": 3.3740198852154233,
+      "grad_norm": 1.0127464532852173,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 20870
+    },
+    {
+      "epoch": 3.3756365693961685,
+      "grad_norm": 0.9111971855163574,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 20880
+    },
+    {
+      "epoch": 3.3772532535769137,
+      "grad_norm": 0.871356725692749,
+      "learning_rate": 0.0002,
+      "loss": 0.6101,
+      "step": 20890
+    },
+    {
+      "epoch": 3.378869937757659,
+      "grad_norm": 0.7774117588996887,
+      "learning_rate": 0.0002,
+      "loss": 0.5995,
+      "step": 20900
+    },
+    {
+      "epoch": 3.380486621938404,
+      "grad_norm": 1.0089964866638184,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 20910
+    },
+    {
+      "epoch": 3.3821033061191494,
+      "grad_norm": 0.7855867147445679,
+      "learning_rate": 0.0002,
+      "loss": 0.5908,
+      "step": 20920
+    },
+    {
+      "epoch": 3.3837199902998947,
+      "grad_norm": 1.3713710308074951,
+      "learning_rate": 0.0002,
+      "loss": 0.6373,
+      "step": 20930
+    },
+    {
+      "epoch": 3.3853366744806404,
+      "grad_norm": 0.8599116206169128,
+      "learning_rate": 0.0002,
+      "loss": 0.6627,
+      "step": 20940
+    },
+    {
+      "epoch": 3.3869533586613856,
+      "grad_norm": 0.9392673373222351,
+      "learning_rate": 0.0002,
+      "loss": 0.6224,
+      "step": 20950
+    },
+    {
+      "epoch": 3.388570042842131,
+      "grad_norm": 0.8764075040817261,
+      "learning_rate": 0.0002,
+      "loss": 0.5855,
+      "step": 20960
+    },
+    {
+      "epoch": 3.390186727022876,
+      "grad_norm": 0.8240136504173279,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 20970
+    },
+    {
+      "epoch": 3.3918034112036213,
+      "grad_norm": 1.0982369184494019,
+      "learning_rate": 0.0002,
+      "loss": 0.5783,
+      "step": 20980
+    },
+    {
+      "epoch": 3.3934200953843665,
+      "grad_norm": 1.0599013566970825,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 20990
+    },
+    {
+      "epoch": 3.395036779565112,
+      "grad_norm": 0.895438015460968,
+      "learning_rate": 0.0002,
+      "loss": 0.6356,
+      "step": 21000
+    },
+    {
+      "epoch": 3.3966534637458574,
+      "grad_norm": 0.6974841356277466,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 21010
+    },
+    {
+      "epoch": 3.3982701479266026,
+      "grad_norm": 0.9571719765663147,
+      "learning_rate": 0.0002,
+      "loss": 0.5704,
+      "step": 21020
+    },
+    {
+      "epoch": 3.399886832107348,
+      "grad_norm": 0.831912636756897,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 21030
+    },
+    {
+      "epoch": 3.401503516288093,
+      "grad_norm": 0.831936240196228,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 21040
+    },
+    {
+      "epoch": 3.4031202004688383,
+      "grad_norm": 0.7388373613357544,
+      "learning_rate": 0.0002,
+      "loss": 0.5857,
+      "step": 21050
+    },
+    {
+      "epoch": 3.4047368846495836,
+      "grad_norm": 0.938667356967926,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21060
+    },
+    {
+      "epoch": 3.406353568830329,
+      "grad_norm": 0.9202313423156738,
+      "learning_rate": 0.0002,
+      "loss": 0.6121,
+      "step": 21070
+    },
+    {
+      "epoch": 3.4079702530110745,
+      "grad_norm": 0.9888381958007812,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 21080
+    },
+    {
+      "epoch": 3.4095869371918197,
+      "grad_norm": 0.8526970744132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21090
+    },
+    {
+      "epoch": 3.411203621372565,
+      "grad_norm": 0.7939383387565613,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 21100
+    },
+    {
+      "epoch": 3.41282030555331,
+      "grad_norm": 0.9986352920532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 21110
+    },
+    {
+      "epoch": 3.4144369897340554,
+      "grad_norm": 0.8895300030708313,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 21120
+    },
+    {
+      "epoch": 3.4160536739148006,
+      "grad_norm": 0.9559482932090759,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 21130
+    },
+    {
+      "epoch": 3.417670358095546,
+      "grad_norm": 0.8351506590843201,
+      "learning_rate": 0.0002,
+      "loss": 0.6491,
+      "step": 21140
+    },
+    {
+      "epoch": 3.4192870422762915,
+      "grad_norm": 0.8224456906318665,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 21150
+    },
+    {
+      "epoch": 3.4209037264570368,
+      "grad_norm": 1.0110299587249756,
+      "learning_rate": 0.0002,
+      "loss": 0.5871,
+      "step": 21160
+    },
+    {
+      "epoch": 3.422520410637782,
+      "grad_norm": 0.82564777135849,
+      "learning_rate": 0.0002,
+      "loss": 0.6116,
+      "step": 21170
+    },
+    {
+      "epoch": 3.4241370948185272,
+      "grad_norm": 1.004738688468933,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 21180
+    },
+    {
+      "epoch": 3.4257537789992725,
+      "grad_norm": 0.7545676827430725,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 21190
+    },
+    {
+      "epoch": 3.4273704631800177,
+      "grad_norm": 0.8918704390525818,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 21200
+    },
+    {
+      "epoch": 3.428987147360763,
+      "grad_norm": 0.8336876034736633,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 21210
+    },
+    {
+      "epoch": 3.430603831541508,
+      "grad_norm": 0.8928771018981934,
+      "learning_rate": 0.0002,
+      "loss": 0.5824,
+      "step": 21220
+    },
+    {
+      "epoch": 3.432220515722254,
+      "grad_norm": 0.7663705945014954,
+      "learning_rate": 0.0002,
+      "loss": 0.6468,
+      "step": 21230
+    },
+    {
+      "epoch": 3.433837199902999,
+      "grad_norm": 0.8392598628997803,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 21240
+    },
+    {
+      "epoch": 3.4354538840837443,
+      "grad_norm": 0.8819600343704224,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 21250
+    },
+    {
+      "epoch": 3.4370705682644895,
+      "grad_norm": 0.9124642014503479,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 21260
+    },
+    {
+      "epoch": 3.4386872524452348,
+      "grad_norm": 0.8329763412475586,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 21270
+    },
+    {
+      "epoch": 3.44030393662598,
+      "grad_norm": 0.9982839822769165,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 21280
+    },
+    {
+      "epoch": 3.4419206208067252,
+      "grad_norm": 0.9105954766273499,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 21290
+    },
+    {
+      "epoch": 3.443537304987471,
+      "grad_norm": 0.8182359337806702,
+      "learning_rate": 0.0002,
+      "loss": 0.6028,
+      "step": 21300
+    },
+    {
+      "epoch": 3.445153989168216,
+      "grad_norm": 1.0568904876708984,
+      "learning_rate": 0.0002,
+      "loss": 0.5991,
+      "step": 21310
+    },
+    {
+      "epoch": 3.4467706733489614,
+      "grad_norm": 0.968539834022522,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 21320
+    },
+    {
+      "epoch": 3.4483873575297066,
+      "grad_norm": 0.8774511218070984,
+      "learning_rate": 0.0002,
+      "loss": 0.6219,
+      "step": 21330
+    },
+    {
+      "epoch": 3.450004041710452,
+      "grad_norm": 0.7598156332969666,
+      "learning_rate": 0.0002,
+      "loss": 0.6438,
+      "step": 21340
+    },
+    {
+      "epoch": 3.451620725891197,
+      "grad_norm": 1.1012897491455078,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 21350
+    },
+    {
+      "epoch": 3.4532374100719423,
+      "grad_norm": 0.8040637373924255,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 21360
+    },
+    {
+      "epoch": 3.4548540942526875,
+      "grad_norm": 0.8497496247291565,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 21370
+    },
+    {
+      "epoch": 3.456470778433433,
+      "grad_norm": 0.8429915904998779,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 21380
+    },
+    {
+      "epoch": 3.4580874626141784,
+      "grad_norm": 0.8107112646102905,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 21390
+    },
+    {
+      "epoch": 3.4597041467949237,
+      "grad_norm": 1.00872004032135,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 21400
+    },
+    {
+      "epoch": 3.461320830975669,
+      "grad_norm": 0.8266542553901672,
+      "learning_rate": 0.0002,
+      "loss": 0.5712,
+      "step": 21410
+    },
+    {
+      "epoch": 3.462937515156414,
+      "grad_norm": 0.8972568511962891,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 21420
+    },
+    {
+      "epoch": 3.4645541993371594,
+      "grad_norm": 1.0781476497650146,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 21430
+    },
+    {
+      "epoch": 3.4661708835179046,
+      "grad_norm": 0.9571592807769775,
+      "learning_rate": 0.0002,
+      "loss": 0.6303,
+      "step": 21440
+    },
+    {
+      "epoch": 3.4677875676986503,
+      "grad_norm": 0.881547212600708,
+      "learning_rate": 0.0002,
+      "loss": 0.6309,
+      "step": 21450
+    },
+    {
+      "epoch": 3.4694042518793955,
+      "grad_norm": 0.6955338716506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 21460
+    },
+    {
+      "epoch": 3.4710209360601407,
+      "grad_norm": 0.901187539100647,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 21470
+    },
+    {
+      "epoch": 3.472637620240886,
+      "grad_norm": 0.7063511610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.639,
+      "step": 21480
+    },
+    {
+      "epoch": 3.474254304421631,
+      "grad_norm": 0.8462792038917542,
+      "learning_rate": 0.0002,
+      "loss": 0.6154,
+      "step": 21490
+    },
+    {
+      "epoch": 3.4758709886023764,
+      "grad_norm": 1.1861060857772827,
+      "learning_rate": 0.0002,
+      "loss": 0.61,
+      "step": 21500
+    },
+    {
+      "epoch": 3.4774876727831217,
+      "grad_norm": 0.70503169298172,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 21510
+    },
+    {
+      "epoch": 3.479104356963867,
+      "grad_norm": 0.9650066494941711,
+      "learning_rate": 0.0002,
+      "loss": 0.6475,
+      "step": 21520
+    },
+    {
+      "epoch": 3.4807210411446126,
+      "grad_norm": 1.0266852378845215,
+      "learning_rate": 0.0002,
+      "loss": 0.6452,
+      "step": 21530
+    },
+    {
+      "epoch": 3.482337725325358,
+      "grad_norm": 0.956372857093811,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 21540
+    },
+    {
+      "epoch": 3.483954409506103,
+      "grad_norm": 0.8848432898521423,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 21550
+    },
+    {
+      "epoch": 3.4855710936868483,
+      "grad_norm": 1.0805351734161377,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 21560
+    },
+    {
+      "epoch": 3.4871877778675935,
+      "grad_norm": 0.9279725551605225,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 21570
+    },
+    {
+      "epoch": 3.4888044620483387,
+      "grad_norm": 0.9049562215805054,
+      "learning_rate": 0.0002,
+      "loss": 0.6354,
+      "step": 21580
+    },
+    {
+      "epoch": 3.4904211462290844,
+      "grad_norm": 0.9619429111480713,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 21590
+    },
+    {
+      "epoch": 3.4920378304098296,
+      "grad_norm": 0.8508906960487366,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 21600
+    },
+    {
+      "epoch": 3.493654514590575,
+      "grad_norm": 0.8692502379417419,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 21610
+    },
+    {
+      "epoch": 3.49527119877132,
+      "grad_norm": 0.8187332153320312,
+      "learning_rate": 0.0002,
+      "loss": 0.5878,
+      "step": 21620
+    },
+    {
+      "epoch": 3.4968878829520653,
+      "grad_norm": 1.145400047302246,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 21630
+    },
+    {
+      "epoch": 3.4985045671328105,
+      "grad_norm": 0.8281388282775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 21640
+    },
+    {
+      "epoch": 3.500121251313556,
+      "grad_norm": 0.82256019115448,
+      "learning_rate": 0.0002,
+      "loss": 0.6624,
+      "step": 21650
+    },
+    {
+      "epoch": 3.501737935494301,
+      "grad_norm": 0.9315484762191772,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 21660
+    },
+    {
+      "epoch": 3.5033546196750462,
+      "grad_norm": 0.7626111507415771,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 21670
+    },
+    {
+      "epoch": 3.504971303855792,
+      "grad_norm": 0.9275059103965759,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 21680
+    },
+    {
+      "epoch": 3.506587988036537,
+      "grad_norm": 0.7906724810600281,
+      "learning_rate": 0.0002,
+      "loss": 0.64,
+      "step": 21690
+    },
+    {
+      "epoch": 3.5082046722172824,
+      "grad_norm": 0.8289761543273926,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 21700
+    },
+    {
+      "epoch": 3.5098213563980276,
+      "grad_norm": 0.8316431045532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 21710
+    },
+    {
+      "epoch": 3.511438040578773,
+      "grad_norm": 1.0451812744140625,
+      "learning_rate": 0.0002,
+      "loss": 0.619,
+      "step": 21720
+    },
+    {
+      "epoch": 3.513054724759518,
+      "grad_norm": 0.928252637386322,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 21730
+    },
+    {
+      "epoch": 3.5146714089402638,
+      "grad_norm": 0.7985895276069641,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 21740
+    },
+    {
+      "epoch": 3.516288093121009,
+      "grad_norm": 0.6740974187850952,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 21750
+    },
+    {
+      "epoch": 3.517904777301754,
+      "grad_norm": 0.8482223749160767,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 21760
+    },
+    {
+      "epoch": 3.5195214614824994,
+      "grad_norm": 0.889947772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.6277,
+      "step": 21770
+    },
+    {
+      "epoch": 3.5211381456632447,
+      "grad_norm": 0.8304598927497864,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 21780
+    },
+    {
+      "epoch": 3.52275482984399,
+      "grad_norm": 0.8002981543540955,
+      "learning_rate": 0.0002,
+      "loss": 0.6156,
+      "step": 21790
+    },
+    {
+      "epoch": 3.524371514024735,
+      "grad_norm": 0.8115083575248718,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21800
+    },
+    {
+      "epoch": 3.5259881982054804,
+      "grad_norm": 0.9715048670768738,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 21810
+    },
+    {
+      "epoch": 3.5276048823862256,
+      "grad_norm": 1.0910786390304565,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 21820
+    },
+    {
+      "epoch": 3.5292215665669713,
+      "grad_norm": 0.8438942432403564,
+      "learning_rate": 0.0002,
+      "loss": 0.6543,
+      "step": 21830
+    },
+    {
+      "epoch": 3.5308382507477165,
+      "grad_norm": 0.8813382983207703,
+      "learning_rate": 0.0002,
+      "loss": 0.6509,
+      "step": 21840
+    },
+    {
+      "epoch": 3.5324549349284617,
+      "grad_norm": 0.7092908024787903,
+      "learning_rate": 0.0002,
+      "loss": 0.6049,
+      "step": 21850
+    },
+    {
+      "epoch": 3.534071619109207,
+      "grad_norm": 0.8332187533378601,
+      "learning_rate": 0.0002,
+      "loss": 0.5678,
+      "step": 21860
+    },
+    {
+      "epoch": 3.535688303289952,
+      "grad_norm": 0.8958209156990051,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21870
+    },
+    {
+      "epoch": 3.5373049874706974,
+      "grad_norm": 0.824138879776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 21880
+    },
+    {
+      "epoch": 3.538921671651443,
+      "grad_norm": 0.8375158309936523,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 21890
+    },
+    {
+      "epoch": 3.5405383558321883,
+      "grad_norm": 1.0274608135223389,
+      "learning_rate": 0.0002,
+      "loss": 0.6019,
+      "step": 21900
+    },
+    {
+      "epoch": 3.5421550400129336,
+      "grad_norm": 0.7088932394981384,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 21910
+    },
+    {
+      "epoch": 3.543771724193679,
+      "grad_norm": 0.8172445297241211,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 21920
+    },
+    {
+      "epoch": 3.545388408374424,
+      "grad_norm": 0.9904135465621948,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 21930
+    },
+    {
+      "epoch": 3.5470050925551693,
+      "grad_norm": 0.9900432229042053,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 21940
+    },
+    {
+      "epoch": 3.5486217767359145,
+      "grad_norm": 0.8963301181793213,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 21950
+    },
+    {
+      "epoch": 3.5502384609166597,
+      "grad_norm": 0.8551464676856995,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 21960
+    },
+    {
+      "epoch": 3.551855145097405,
+      "grad_norm": 1.0916603803634644,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 21970
+    },
+    {
+      "epoch": 3.5534718292781506,
+      "grad_norm": 0.841598391532898,
+      "learning_rate": 0.0002,
+      "loss": 0.6523,
+      "step": 21980
+    },
+    {
+      "epoch": 3.555088513458896,
+      "grad_norm": 0.8566757440567017,
+      "learning_rate": 0.0002,
+      "loss": 0.617,
+      "step": 21990
+    },
+    {
+      "epoch": 3.556705197639641,
+      "grad_norm": 1.0145052671432495,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 22000
+    },
+    {
+      "epoch": 3.5583218818203863,
+      "grad_norm": 0.9293754696846008,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 22010
+    },
+    {
+      "epoch": 3.5599385660011316,
+      "grad_norm": 0.9568536281585693,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 22020
+    },
+    {
+      "epoch": 3.5615552501818772,
+      "grad_norm": 0.8613139986991882,
+      "learning_rate": 0.0002,
+      "loss": 0.641,
+      "step": 22030
+    },
+    {
+      "epoch": 3.5631719343626225,
+      "grad_norm": 0.8179237246513367,
+      "learning_rate": 0.0002,
+      "loss": 0.6496,
+      "step": 22040
+    },
+    {
+      "epoch": 3.5647886185433677,
+      "grad_norm": 0.9059830904006958,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 22050
+    },
+    {
+      "epoch": 3.566405302724113,
+      "grad_norm": 1.0068252086639404,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 22060
+    },
+    {
+      "epoch": 3.568021986904858,
+      "grad_norm": 0.9682072997093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6239,
+      "step": 22070
+    },
+    {
+      "epoch": 3.5696386710856034,
+      "grad_norm": 0.8514005541801453,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 22080
+    },
+    {
+      "epoch": 3.5712553552663486,
+      "grad_norm": 0.8327770829200745,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 22090
+    },
+    {
+      "epoch": 3.572872039447094,
+      "grad_norm": 1.024976372718811,
+      "learning_rate": 0.0002,
+      "loss": 0.5976,
+      "step": 22100
+    },
+    {
+      "epoch": 3.574488723627839,
+      "grad_norm": 0.7721174955368042,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 22110
+    },
+    {
+      "epoch": 3.5761054078085843,
+      "grad_norm": 1.0351054668426514,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 22120
+    },
+    {
+      "epoch": 3.57772209198933,
+      "grad_norm": 0.9680907130241394,
+      "learning_rate": 0.0002,
+      "loss": 0.6379,
+      "step": 22130
+    },
+    {
+      "epoch": 3.5793387761700752,
+      "grad_norm": 0.8016974925994873,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 22140
+    },
+    {
+      "epoch": 3.5809554603508205,
+      "grad_norm": 1.0109003782272339,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 22150
+    },
+    {
+      "epoch": 3.5825721445315657,
+      "grad_norm": 1.0473392009735107,
+      "learning_rate": 0.0002,
+      "loss": 0.6368,
+      "step": 22160
+    },
+    {
+      "epoch": 3.584188828712311,
+      "grad_norm": 0.8686613440513611,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 22170
+    },
+    {
+      "epoch": 3.5858055128930566,
+      "grad_norm": 0.869149923324585,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 22180
+    },
+    {
+      "epoch": 3.587422197073802,
+      "grad_norm": 0.9769062995910645,
+      "learning_rate": 0.0002,
+      "loss": 0.5895,
+      "step": 22190
+    },
+    {
+      "epoch": 3.589038881254547,
+      "grad_norm": 0.779636561870575,
+      "learning_rate": 0.0002,
+      "loss": 0.5939,
+      "step": 22200
+    },
+    {
+      "epoch": 3.5906555654352923,
+      "grad_norm": 0.9063841104507446,
+      "learning_rate": 0.0002,
+      "loss": 0.5875,
+      "step": 22210
+    },
+    {
+      "epoch": 3.5922722496160375,
+      "grad_norm": 0.9216037392616272,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 22220
+    },
+    {
+      "epoch": 3.5938889337967828,
+      "grad_norm": 1.0217336416244507,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 22230
+    },
+    {
+      "epoch": 3.595505617977528,
+      "grad_norm": 0.8513161540031433,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 22240
+    },
+    {
+      "epoch": 3.597122302158273,
+      "grad_norm": 0.8084813952445984,
+      "learning_rate": 0.0002,
+      "loss": 0.6301,
+      "step": 22250
+    },
+    {
+      "epoch": 3.5987389863390185,
+      "grad_norm": 0.8524802923202515,
+      "learning_rate": 0.0002,
+      "loss": 0.6197,
+      "step": 22260
+    },
+    {
+      "epoch": 3.600355670519764,
+      "grad_norm": 0.9356237649917603,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 22270
+    },
+    {
+      "epoch": 3.6019723547005094,
+      "grad_norm": 1.009600281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.628,
+      "step": 22280
+    },
+    {
+      "epoch": 3.6035890388812546,
+      "grad_norm": 0.9900581240653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 22290
+    },
+    {
+      "epoch": 3.605205723062,
+      "grad_norm": 1.062495231628418,
+      "learning_rate": 0.0002,
+      "loss": 0.5725,
+      "step": 22300
+    },
+    {
+      "epoch": 3.606822407242745,
+      "grad_norm": 0.8832381367683411,
+      "learning_rate": 0.0002,
+      "loss": 0.607,
+      "step": 22310
+    },
+    {
+      "epoch": 3.6084390914234903,
+      "grad_norm": 0.9284297823905945,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 22320
+    },
+    {
+      "epoch": 3.610055775604236,
+      "grad_norm": 1.2381829023361206,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 22330
+    },
+    {
+      "epoch": 3.611672459784981,
+      "grad_norm": 0.929434597492218,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 22340
+    },
+    {
+      "epoch": 3.6132891439657264,
+      "grad_norm": 0.9714490175247192,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 22350
+    },
+    {
+      "epoch": 3.6149058281464717,
+      "grad_norm": 0.808014988899231,
+      "learning_rate": 0.0002,
+      "loss": 0.6861,
+      "step": 22360
+    },
+    {
+      "epoch": 3.616522512327217,
+      "grad_norm": 1.0364398956298828,
+      "learning_rate": 0.0002,
+      "loss": 0.6428,
+      "step": 22370
+    },
+    {
+      "epoch": 3.618139196507962,
+      "grad_norm": 0.7858489751815796,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22380
+    },
+    {
+      "epoch": 3.6197558806887074,
+      "grad_norm": 0.9920870065689087,
+      "learning_rate": 0.0002,
+      "loss": 0.6214,
+      "step": 22390
+    },
+    {
+      "epoch": 3.6213725648694526,
+      "grad_norm": 0.9183220863342285,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 22400
+    },
+    {
+      "epoch": 3.622989249050198,
+      "grad_norm": 0.9826246500015259,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22410
+    },
+    {
+      "epoch": 3.6246059332309435,
+      "grad_norm": 0.8632931113243103,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 22420
+    },
+    {
+      "epoch": 3.6262226174116887,
+      "grad_norm": 0.8468965291976929,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 22430
+    },
+    {
+      "epoch": 3.627839301592434,
+      "grad_norm": 0.8466871976852417,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 22440
+    },
+    {
+      "epoch": 3.629455985773179,
+      "grad_norm": 0.9501169919967651,
+      "learning_rate": 0.0002,
+      "loss": 0.5941,
+      "step": 22450
+    },
+    {
+      "epoch": 3.6310726699539244,
+      "grad_norm": 0.8906720876693726,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 22460
+    },
+    {
+      "epoch": 3.6326893541346696,
+      "grad_norm": 0.7400227189064026,
+      "learning_rate": 0.0002,
+      "loss": 0.6928,
+      "step": 22470
+    },
+    {
+      "epoch": 3.6343060383154153,
+      "grad_norm": 0.9756355881690979,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22480
+    },
+    {
+      "epoch": 3.6359227224961606,
+      "grad_norm": 0.7504993081092834,
+      "learning_rate": 0.0002,
+      "loss": 0.6203,
+      "step": 22490
+    },
+    {
+      "epoch": 3.637539406676906,
+      "grad_norm": 0.9270039200782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 22500
+    },
+    {
+      "epoch": 3.639156090857651,
+      "grad_norm": 0.8841686844825745,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 22510
+    },
+    {
+      "epoch": 3.6407727750383962,
+      "grad_norm": 0.8533213138580322,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 22520
+    },
+    {
+      "epoch": 3.6423894592191415,
+      "grad_norm": 1.0052043199539185,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 22530
+    },
+    {
+      "epoch": 3.6440061433998867,
+      "grad_norm": 1.0323461294174194,
+      "learning_rate": 0.0002,
+      "loss": 0.6363,
+      "step": 22540
+    },
+    {
+      "epoch": 3.645622827580632,
+      "grad_norm": 0.8654312491416931,
+      "learning_rate": 0.0002,
+      "loss": 0.6545,
+      "step": 22550
+    },
+    {
+      "epoch": 3.647239511761377,
+      "grad_norm": 0.6400038003921509,
+      "learning_rate": 0.0002,
+      "loss": 0.6155,
+      "step": 22560
+    },
+    {
+      "epoch": 3.648856195942123,
+      "grad_norm": 0.8061298727989197,
+      "learning_rate": 0.0002,
+      "loss": 0.5829,
+      "step": 22570
+    },
+    {
+      "epoch": 3.650472880122868,
+      "grad_norm": 0.9257854223251343,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 22580
+    },
+    {
+      "epoch": 3.6520895643036133,
+      "grad_norm": 0.8439396619796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6409,
+      "step": 22590
+    },
+    {
+      "epoch": 3.6537062484843585,
+      "grad_norm": 0.7764544486999512,
+      "learning_rate": 0.0002,
+      "loss": 0.5996,
+      "step": 22600
+    },
+    {
+      "epoch": 3.6553229326651038,
+      "grad_norm": 1.125451683998108,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 22610
+    },
+    {
+      "epoch": 3.656939616845849,
+      "grad_norm": 0.7523018717765808,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 22620
+    },
+    {
+      "epoch": 3.6585563010265947,
+      "grad_norm": 1.071026086807251,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 22630
+    },
+    {
+      "epoch": 3.66017298520734,
+      "grad_norm": 0.945791482925415,
+      "learning_rate": 0.0002,
+      "loss": 0.6459,
+      "step": 22640
+    },
+    {
+      "epoch": 3.661789669388085,
+      "grad_norm": 0.8001811504364014,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 22650
+    },
+    {
+      "epoch": 3.6634063535688304,
+      "grad_norm": 0.9700816869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.6385,
+      "step": 22660
+    },
+    {
+      "epoch": 3.6650230377495756,
+      "grad_norm": 0.9053242206573486,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22670
+    },
+    {
+      "epoch": 3.666639721930321,
+      "grad_norm": 0.944362461566925,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 22680
+    },
+    {
+      "epoch": 3.668256406111066,
+      "grad_norm": 1.067489504814148,
+      "learning_rate": 0.0002,
+      "loss": 0.6235,
+      "step": 22690
+    },
+    {
+      "epoch": 3.6698730902918113,
+      "grad_norm": 1.0984995365142822,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 22700
+    },
+    {
+      "epoch": 3.6714897744725565,
+      "grad_norm": 0.9336317777633667,
+      "learning_rate": 0.0002,
+      "loss": 0.6717,
+      "step": 22710
+    },
+    {
+      "epoch": 3.673106458653302,
+      "grad_norm": 0.9261918663978577,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 22720
+    },
+    {
+      "epoch": 3.6747231428340474,
+      "grad_norm": 0.8648008704185486,
+      "learning_rate": 0.0002,
+      "loss": 0.6332,
+      "step": 22730
+    },
+    {
+      "epoch": 3.6763398270147927,
+      "grad_norm": 0.7225083708763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6576,
+      "step": 22740
+    },
+    {
+      "epoch": 3.677956511195538,
+      "grad_norm": 0.9258282780647278,
+      "learning_rate": 0.0002,
+      "loss": 0.6406,
+      "step": 22750
+    },
+    {
+      "epoch": 3.679573195376283,
+      "grad_norm": 0.70876145362854,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 22760
+    },
+    {
+      "epoch": 3.681189879557029,
+      "grad_norm": 0.8780210018157959,
+      "learning_rate": 0.0002,
+      "loss": 0.6821,
+      "step": 22770
+    },
+    {
+      "epoch": 3.682806563737774,
+      "grad_norm": 0.8075440526008606,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22780
+    },
+    {
+      "epoch": 3.6844232479185193,
+      "grad_norm": 0.8503130674362183,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22790
+    },
+    {
+      "epoch": 3.6860399320992645,
+      "grad_norm": 0.8413618206977844,
+      "learning_rate": 0.0002,
+      "loss": 0.6082,
+      "step": 22800
+    },
+    {
+      "epoch": 3.6876566162800097,
+      "grad_norm": 0.8675165176391602,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 22810
+    },
+    {
+      "epoch": 3.689273300460755,
+      "grad_norm": 0.8235884308815002,
+      "learning_rate": 0.0002,
+      "loss": 0.6157,
+      "step": 22820
+    },
+    {
+      "epoch": 3.6908899846415,
+      "grad_norm": 0.9477725625038147,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 22830
+    },
+    {
+      "epoch": 3.6925066688222454,
+      "grad_norm": 0.7883533835411072,
+      "learning_rate": 0.0002,
+      "loss": 0.6481,
+      "step": 22840
+    },
+    {
+      "epoch": 3.6941233530029907,
+      "grad_norm": 1.047913908958435,
+      "learning_rate": 0.0002,
+      "loss": 0.5872,
+      "step": 22850
+    },
+    {
+      "epoch": 3.695740037183736,
+      "grad_norm": 0.9171528816223145,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 22860
+    },
+    {
+      "epoch": 3.6973567213644816,
+      "grad_norm": 0.9338192343711853,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 22870
+    },
+    {
+      "epoch": 3.698973405545227,
+      "grad_norm": 0.8799443244934082,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 22880
+    },
+    {
+      "epoch": 3.700590089725972,
+      "grad_norm": 0.8515434861183167,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 22890
+    },
+    {
+      "epoch": 3.7022067739067173,
+      "grad_norm": 0.7805591821670532,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 22900
+    },
+    {
+      "epoch": 3.7038234580874625,
+      "grad_norm": 0.8470911979675293,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 22910
+    },
+    {
+      "epoch": 3.705440142268208,
+      "grad_norm": 0.9452309012413025,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 22920
+    },
+    {
+      "epoch": 3.7070568264489534,
+      "grad_norm": 0.950243353843689,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 22930
+    },
+    {
+      "epoch": 3.7086735106296986,
+      "grad_norm": 0.7882499098777771,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 22940
+    },
+    {
+      "epoch": 3.710290194810444,
+      "grad_norm": 0.8307787775993347,
+      "learning_rate": 0.0002,
+      "loss": 0.6462,
+      "step": 22950
+    },
+    {
+      "epoch": 3.711906878991189,
+      "grad_norm": 1.0970630645751953,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 22960
+    },
+    {
+      "epoch": 3.7135235631719343,
+      "grad_norm": 0.8269566297531128,
+      "learning_rate": 0.0002,
+      "loss": 0.6281,
+      "step": 22970
+    },
+    {
+      "epoch": 3.7151402473526796,
+      "grad_norm": 0.8306704759597778,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22980
+    },
+    {
+      "epoch": 3.716756931533425,
+      "grad_norm": 0.9710225462913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6418,
+      "step": 22990
+    },
+    {
+      "epoch": 3.71837361571417,
+      "grad_norm": 0.8890530467033386,
+      "learning_rate": 0.0002,
+      "loss": 0.6639,
+      "step": 23000
+    },
+    {
+      "epoch": 3.7199902998949153,
+      "grad_norm": 0.883522629737854,
+      "learning_rate": 0.0002,
+      "loss": 0.6084,
+      "step": 23010
+    },
+    {
+      "epoch": 3.721606984075661,
+      "grad_norm": 0.8662652373313904,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 23020
+    },
+    {
+      "epoch": 3.723223668256406,
+      "grad_norm": 0.7228406667709351,
+      "learning_rate": 0.0002,
+      "loss": 0.6266,
+      "step": 23030
+    },
+    {
+      "epoch": 3.7248403524371514,
+      "grad_norm": 1.060792088508606,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23040
+    },
+    {
+      "epoch": 3.7264570366178966,
+      "grad_norm": 1.0119613409042358,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 23050
+    },
+    {
+      "epoch": 3.728073720798642,
+      "grad_norm": 0.9212996959686279,
+      "learning_rate": 0.0002,
+      "loss": 0.6466,
+      "step": 23060
+    },
+    {
+      "epoch": 3.7296904049793875,
+      "grad_norm": 0.925690233707428,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 23070
+    },
+    {
+      "epoch": 3.7313070891601328,
+      "grad_norm": 0.8323310613632202,
+      "learning_rate": 0.0002,
+      "loss": 0.615,
+      "step": 23080
+    },
+    {
+      "epoch": 3.732923773340878,
+      "grad_norm": 0.8966048955917358,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 23090
+    },
+    {
+      "epoch": 3.7345404575216232,
+      "grad_norm": 0.8995837569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23100
+    },
+    {
+      "epoch": 3.7361571417023685,
+      "grad_norm": 0.8748890161514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23110
+    },
+    {
+      "epoch": 3.7377738258831137,
+      "grad_norm": 0.7985540628433228,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 23120
+    },
+    {
+      "epoch": 3.739390510063859,
+      "grad_norm": 1.0240917205810547,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 23130
+    },
+    {
+      "epoch": 3.741007194244604,
+      "grad_norm": 0.9181789755821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 23140
+    },
+    {
+      "epoch": 3.7426238784253494,
+      "grad_norm": 0.8896583914756775,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 23150
+    },
+    {
+      "epoch": 3.744240562606095,
+      "grad_norm": 0.8635515570640564,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 23160
+    },
+    {
+      "epoch": 3.7458572467868403,
+      "grad_norm": 0.8873575329780579,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 23170
+    },
+    {
+      "epoch": 3.7474739309675855,
+      "grad_norm": 0.9807148575782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23180
+    },
+    {
+      "epoch": 3.7490906151483308,
+      "grad_norm": 0.900477945804596,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 23190
+    },
+    {
+      "epoch": 3.750707299329076,
+      "grad_norm": 0.9379992485046387,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23200
+    },
+    {
+      "epoch": 3.752323983509821,
+      "grad_norm": 0.9649890661239624,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 23210
+    },
+    {
+      "epoch": 3.753940667690567,
+      "grad_norm": 0.824442446231842,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 23220
+    },
+    {
+      "epoch": 3.755557351871312,
+      "grad_norm": 0.8896150588989258,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 23230
+    },
+    {
+      "epoch": 3.7571740360520574,
+      "grad_norm": 0.751249372959137,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 23240
+    },
+    {
+      "epoch": 3.7587907202328026,
+      "grad_norm": 0.9392193555831909,
+      "learning_rate": 0.0002,
+      "loss": 0.629,
+      "step": 23250
+    },
+    {
+      "epoch": 3.760407404413548,
+      "grad_norm": 0.9284586310386658,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 23260
+    },
+    {
+      "epoch": 3.762024088594293,
+      "grad_norm": 0.7738175392150879,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23270
+    },
+    {
+      "epoch": 3.7636407727750383,
+      "grad_norm": 0.9252978563308716,
+      "learning_rate": 0.0002,
+      "loss": 0.6743,
+      "step": 23280
+    },
+    {
+      "epoch": 3.7652574569557835,
+      "grad_norm": 0.9501895904541016,
+      "learning_rate": 0.0002,
+      "loss": 0.5984,
+      "step": 23290
+    },
+    {
+      "epoch": 3.7668741411365287,
+      "grad_norm": 0.9416276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 23300
+    },
+    {
+      "epoch": 3.7684908253172744,
+      "grad_norm": 0.7076631784439087,
+      "learning_rate": 0.0002,
+      "loss": 0.6507,
+      "step": 23310
+    },
+    {
+      "epoch": 3.7701075094980196,
+      "grad_norm": 0.9864492416381836,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 23320
+    },
+    {
+      "epoch": 3.771724193678765,
+      "grad_norm": 0.8450456261634827,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 23330
+    },
+    {
+      "epoch": 3.77334087785951,
+      "grad_norm": 1.0768941640853882,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23340
+    },
+    {
+      "epoch": 3.7749575620402553,
+      "grad_norm": 0.9956819415092468,
+      "learning_rate": 0.0002,
+      "loss": 0.6408,
+      "step": 23350
+    },
+    {
+      "epoch": 3.7765742462210006,
+      "grad_norm": 0.9234658479690552,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 23360
+    },
+    {
+      "epoch": 3.7781909304017463,
+      "grad_norm": 1.0993858575820923,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23370
+    },
+    {
+      "epoch": 3.7798076145824915,
+      "grad_norm": 0.923159658908844,
+      "learning_rate": 0.0002,
+      "loss": 0.6391,
+      "step": 23380
+    },
+    {
+      "epoch": 3.7814242987632367,
+      "grad_norm": 0.9311541318893433,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23390
+    },
+    {
+      "epoch": 3.783040982943982,
+      "grad_norm": 0.919681191444397,
+      "learning_rate": 0.0002,
+      "loss": 0.6535,
+      "step": 23400
+    },
+    {
+      "epoch": 3.784657667124727,
+      "grad_norm": 1.7406195402145386,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 23410
+    },
+    {
+      "epoch": 3.7862743513054724,
+      "grad_norm": 0.7789074182510376,
+      "learning_rate": 0.0002,
+      "loss": 0.657,
+      "step": 23420
+    },
+    {
+      "epoch": 3.7878910354862176,
+      "grad_norm": 0.8302814960479736,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23430
+    },
+    {
+      "epoch": 3.789507719666963,
+      "grad_norm": 0.8089349269866943,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23440
+    },
+    {
+      "epoch": 3.791124403847708,
+      "grad_norm": 0.9006284475326538,
+      "learning_rate": 0.0002,
+      "loss": 0.6682,
+      "step": 23450
+    },
+    {
+      "epoch": 3.7927410880284538,
+      "grad_norm": 0.8426766991615295,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 23460
+    },
+    {
+      "epoch": 3.794357772209199,
+      "grad_norm": 1.2576252222061157,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 23470
+    },
+    {
+      "epoch": 3.7959744563899442,
+      "grad_norm": 1.0307610034942627,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 23480
+    },
+    {
+      "epoch": 3.7975911405706895,
+      "grad_norm": 0.8525972962379456,
+      "learning_rate": 0.0002,
+      "loss": 0.6262,
+      "step": 23490
+    },
+    {
+      "epoch": 3.7992078247514347,
+      "grad_norm": 1.159039855003357,
+      "learning_rate": 0.0002,
+      "loss": 0.6757,
+      "step": 23500
+    },
+    {
+      "epoch": 3.80082450893218,
+      "grad_norm": 1.4193549156188965,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23510
+    },
+    {
+      "epoch": 3.8024411931129256,
+      "grad_norm": 0.8245543837547302,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 23520
+    },
+    {
+      "epoch": 3.804057877293671,
+      "grad_norm": 0.8847230076789856,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23530
+    },
+    {
+      "epoch": 3.805674561474416,
+      "grad_norm": 0.9574624300003052,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 23540
+    },
+    {
+      "epoch": 3.8072912456551613,
+      "grad_norm": 1.048020601272583,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 23550
+    },
+    {
+      "epoch": 3.8089079298359065,
+      "grad_norm": 0.8302255868911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6497,
+      "step": 23560
+    },
+    {
+      "epoch": 3.8105246140166518,
+      "grad_norm": 0.8269215822219849,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 23570
+    },
+    {
+      "epoch": 3.812141298197397,
+      "grad_norm": 0.9375753402709961,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 23580
+    },
+    {
+      "epoch": 3.8137579823781422,
+      "grad_norm": 1.0234097242355347,
+      "learning_rate": 0.0002,
+      "loss": 0.6132,
+      "step": 23590
+    },
+    {
+      "epoch": 3.8153746665588875,
+      "grad_norm": 0.8978445529937744,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23600
+    },
+    {
+      "epoch": 3.816991350739633,
+      "grad_norm": 0.7929515838623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 23610
+    },
+    {
+      "epoch": 3.8186080349203784,
+      "grad_norm": 1.3255881071090698,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23620
+    },
+    {
+      "epoch": 3.8202247191011236,
+      "grad_norm": 0.9188598990440369,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 23630
+    },
+    {
+      "epoch": 3.821841403281869,
+      "grad_norm": 0.8811675906181335,
+      "learning_rate": 0.0002,
+      "loss": 0.6152,
+      "step": 23640
+    },
+    {
+      "epoch": 3.823458087462614,
+      "grad_norm": 0.8061038255691528,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 23650
+    },
+    {
+      "epoch": 3.8250747716433597,
+      "grad_norm": 0.9975376129150391,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 23660
+    },
+    {
+      "epoch": 3.826691455824105,
+      "grad_norm": 0.8036105036735535,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 23670
+    },
+    {
+      "epoch": 3.82830814000485,
+      "grad_norm": 0.7401984333992004,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 23680
+    },
+    {
+      "epoch": 3.8299248241855954,
+      "grad_norm": 0.829753041267395,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 23690
+    },
+    {
+      "epoch": 3.8315415083663407,
+      "grad_norm": 0.8753240704536438,
+      "learning_rate": 0.0002,
+      "loss": 0.6611,
+      "step": 23700
+    },
+    {
+      "epoch": 3.833158192547086,
+      "grad_norm": 0.8157842755317688,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 23710
+    },
+    {
+      "epoch": 3.834774876727831,
+      "grad_norm": 0.6183798909187317,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 23720
+    },
+    {
+      "epoch": 3.8363915609085764,
+      "grad_norm": 0.9548442363739014,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 23730
+    },
+    {
+      "epoch": 3.8380082450893216,
+      "grad_norm": 0.8319669961929321,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23740
+    },
+    {
+      "epoch": 3.839624929270067,
+      "grad_norm": 0.9718693494796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 23750
+    },
+    {
+      "epoch": 3.8412416134508125,
+      "grad_norm": 0.8672235012054443,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 23760
+    },
+    {
+      "epoch": 3.8428582976315577,
+      "grad_norm": 1.1210707426071167,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 23770
+    },
+    {
+      "epoch": 3.844474981812303,
+      "grad_norm": 0.9177767634391785,
+      "learning_rate": 0.0002,
+      "loss": 0.5923,
+      "step": 23780
+    },
+    {
+      "epoch": 3.846091665993048,
+      "grad_norm": 0.8714171648025513,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 23790
+    },
+    {
+      "epoch": 3.8477083501737934,
+      "grad_norm": 1.1853246688842773,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 23800
+    },
+    {
+      "epoch": 3.849325034354539,
+      "grad_norm": 0.8091260194778442,
+      "learning_rate": 0.0002,
+      "loss": 0.6144,
+      "step": 23810
+    },
+    {
+      "epoch": 3.8509417185352843,
+      "grad_norm": 0.9710774421691895,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23820
+    },
+    {
+      "epoch": 3.8525584027160296,
+      "grad_norm": 0.7648707628250122,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23830
+    },
+    {
+      "epoch": 3.854175086896775,
+      "grad_norm": 0.7809253931045532,
+      "learning_rate": 0.0002,
+      "loss": 0.6013,
+      "step": 23840
+    },
+    {
+      "epoch": 3.85579177107752,
+      "grad_norm": 0.8337951898574829,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 23850
+    },
+    {
+      "epoch": 3.8574084552582653,
+      "grad_norm": 0.9271913170814514,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23860
+    },
+    {
+      "epoch": 3.8590251394390105,
+      "grad_norm": 0.985334038734436,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 23870
+    },
+    {
+      "epoch": 3.8606418236197557,
+      "grad_norm": 0.8458583354949951,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 23880
+    },
+    {
+      "epoch": 3.862258507800501,
+      "grad_norm": 1.015348196029663,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 23890
+    },
+    {
+      "epoch": 3.8638751919812466,
+      "grad_norm": 1.0121688842773438,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23900
+    },
+    {
+      "epoch": 3.865491876161992,
+      "grad_norm": 0.8883971571922302,
+      "learning_rate": 0.0002,
+      "loss": 0.5921,
+      "step": 23910
+    },
+    {
+      "epoch": 3.867108560342737,
+      "grad_norm": 1.028086543083191,
+      "learning_rate": 0.0002,
+      "loss": 0.6597,
+      "step": 23920
+    },
+    {
+      "epoch": 3.8687252445234823,
+      "grad_norm": 0.9645734429359436,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 23930
+    },
+    {
+      "epoch": 3.8703419287042276,
+      "grad_norm": 0.8235350251197815,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 23940
+    },
+    {
+      "epoch": 3.871958612884973,
+      "grad_norm": 1.0298916101455688,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23950
+    },
+    {
+      "epoch": 3.8735752970657185,
+      "grad_norm": 1.0063377618789673,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 23960
+    },
+    {
+      "epoch": 3.8751919812464637,
+      "grad_norm": 0.9230626821517944,
+      "learning_rate": 0.0002,
+      "loss": 0.6234,
+      "step": 23970
+    },
+    {
+      "epoch": 3.876808665427209,
+      "grad_norm": 0.9243063926696777,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 23980
+    },
+    {
+      "epoch": 3.878425349607954,
+      "grad_norm": 1.0211291313171387,
+      "learning_rate": 0.0002,
+      "loss": 0.6035,
+      "step": 23990
+    },
+    {
+      "epoch": 3.8800420337886994,
+      "grad_norm": 0.7800535559654236,
+      "learning_rate": 0.0002,
+      "loss": 0.6351,
+      "step": 24000
+    },
+    {
+      "epoch": 3.8816587179694446,
+      "grad_norm": 0.7904248833656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 24010
+    },
+    {
+      "epoch": 3.88327540215019,
+      "grad_norm": 1.1975988149642944,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 24020
+    },
+    {
+      "epoch": 3.884892086330935,
+      "grad_norm": 1.0626593828201294,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 24030
+    },
+    {
+      "epoch": 3.8865087705116803,
+      "grad_norm": 0.9012193083763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 24040
+    },
+    {
+      "epoch": 3.888125454692426,
+      "grad_norm": 1.1159172058105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6786,
+      "step": 24050
+    },
+    {
+      "epoch": 3.889742138873171,
+      "grad_norm": 1.276838779449463,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 24060
+    },
+    {
+      "epoch": 3.8913588230539164,
+      "grad_norm": 0.8467690348625183,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 24070
+    },
+    {
+      "epoch": 3.8929755072346617,
+      "grad_norm": 0.9862841963768005,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 24080
+    },
+    {
+      "epoch": 3.894592191415407,
+      "grad_norm": 0.7134621739387512,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 24090
+    },
+    {
+      "epoch": 3.896208875596152,
+      "grad_norm": 0.8178175091743469,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 24100
+    },
+    {
+      "epoch": 3.897825559776898,
+      "grad_norm": 0.9229172468185425,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 24110
+    },
+    {
+      "epoch": 3.899442243957643,
+      "grad_norm": 1.0878316164016724,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 24120
+    },
+    {
+      "epoch": 3.9010589281383883,
+      "grad_norm": 0.971645712852478,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24130
+    },
+    {
+      "epoch": 3.9026756123191335,
+      "grad_norm": 0.8862188458442688,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 24140
+    },
+    {
+      "epoch": 3.9042922964998787,
+      "grad_norm": 0.9126982688903809,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 24150
+    },
+    {
+      "epoch": 3.905908980680624,
+      "grad_norm": 0.8833470940589905,
+      "learning_rate": 0.0002,
+      "loss": 0.6734,
+      "step": 24160
+    },
+    {
+      "epoch": 3.907525664861369,
+      "grad_norm": 0.8320947885513306,
+      "learning_rate": 0.0002,
+      "loss": 0.5832,
+      "step": 24170
+    },
+    {
+      "epoch": 3.9091423490421144,
+      "grad_norm": 0.9156602025032043,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 24180
+    },
+    {
+      "epoch": 3.9107590332228597,
+      "grad_norm": 1.029181957244873,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 24190
+    },
+    {
+      "epoch": 3.9123757174036053,
+      "grad_norm": 0.9052802324295044,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 24200
+    },
+    {
+      "epoch": 3.9139924015843506,
+      "grad_norm": 0.8847255110740662,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 24210
+    },
+    {
+      "epoch": 3.915609085765096,
+      "grad_norm": 0.9642062187194824,
+      "learning_rate": 0.0002,
+      "loss": 0.6343,
+      "step": 24220
+    },
+    {
+      "epoch": 3.917225769945841,
+      "grad_norm": 0.8629093766212463,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 24230
+    },
+    {
+      "epoch": 3.9188424541265863,
+      "grad_norm": 0.8674976825714111,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 24240
+    },
+    {
+      "epoch": 3.9204591383073315,
+      "grad_norm": 1.104846477508545,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 24250
+    },
+    {
+      "epoch": 3.922075822488077,
+      "grad_norm": 1.0874955654144287,
+      "learning_rate": 0.0002,
+      "loss": 0.6501,
+      "step": 24260
+    },
+    {
+      "epoch": 3.9236925066688224,
+      "grad_norm": 0.8689812421798706,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 24270
+    },
+    {
+      "epoch": 3.9253091908495676,
+      "grad_norm": 0.9724617004394531,
+      "learning_rate": 0.0002,
+      "loss": 0.5893,
+      "step": 24280
+    },
+    {
+      "epoch": 3.926925875030313,
+      "grad_norm": 0.9165538549423218,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24290
+    },
+    {
+      "epoch": 3.928542559211058,
+      "grad_norm": 0.9307710528373718,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 24300
+    },
+    {
+      "epoch": 3.9301592433918033,
+      "grad_norm": 0.8589295148849487,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 24310
+    },
+    {
+      "epoch": 3.9317759275725486,
+      "grad_norm": 0.9151099920272827,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 24320
+    },
+    {
+      "epoch": 3.933392611753294,
+      "grad_norm": 0.9633517265319824,
+      "learning_rate": 0.0002,
+      "loss": 0.7075,
+      "step": 24330
+    },
+    {
+      "epoch": 3.935009295934039,
+      "grad_norm": 0.9521116018295288,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24340
+    },
+    {
+      "epoch": 3.9366259801147847,
+      "grad_norm": 0.8366776704788208,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 24350
+    },
+    {
+      "epoch": 3.93824266429553,
+      "grad_norm": 0.8972663283348083,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 24360
+    },
+    {
+      "epoch": 3.939859348476275,
+      "grad_norm": 0.8102919459342957,
+      "learning_rate": 0.0002,
+      "loss": 0.661,
+      "step": 24370
+    },
+    {
+      "epoch": 3.9414760326570204,
+      "grad_norm": 0.8189975023269653,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 24380
+    },
+    {
+      "epoch": 3.9430927168377656,
+      "grad_norm": 0.9569464921951294,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 24390
+    },
+    {
+      "epoch": 3.9447094010185113,
+      "grad_norm": 0.7459101676940918,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 24400
+    },
+    {
+      "epoch": 3.9463260851992565,
+      "grad_norm": 0.8536974787712097,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 24410
+    },
+    {
+      "epoch": 3.9479427693800018,
+      "grad_norm": 0.8763698935508728,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 24420
+    },
+    {
+      "epoch": 3.949559453560747,
+      "grad_norm": 0.9381106495857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6478,
+      "step": 24430
+    },
+    {
+      "epoch": 3.9511761377414922,
+      "grad_norm": 0.934440016746521,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 24440
+    },
+    {
+      "epoch": 3.9527928219222375,
+      "grad_norm": 0.903918981552124,
+      "learning_rate": 0.0002,
+      "loss": 0.6393,
+      "step": 24450
+    },
+    {
+      "epoch": 3.9544095061029827,
+      "grad_norm": 0.8771953582763672,
+      "learning_rate": 0.0002,
+      "loss": 0.6175,
+      "step": 24460
+    },
+    {
+      "epoch": 3.956026190283728,
+      "grad_norm": 1.0375410318374634,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 24470
+    },
+    {
+      "epoch": 3.957642874464473,
+      "grad_norm": 0.9439185261726379,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 24480
+    },
+    {
+      "epoch": 3.9592595586452184,
+      "grad_norm": 0.935467004776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 24490
+    },
+    {
+      "epoch": 3.960876242825964,
+      "grad_norm": 0.6900772452354431,
+      "learning_rate": 0.0002,
+      "loss": 0.6437,
+      "step": 24500
+    },
+    {
+      "epoch": 3.9624929270067093,
+      "grad_norm": 1.0172916650772095,
+      "learning_rate": 0.0002,
+      "loss": 0.6445,
+      "step": 24510
+    },
+    {
+      "epoch": 3.9641096111874545,
+      "grad_norm": 0.9167046546936035,
+      "learning_rate": 0.0002,
+      "loss": 0.6308,
+      "step": 24520
+    },
+    {
+      "epoch": 3.9657262953681998,
+      "grad_norm": 0.7230527997016907,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 24530
+    },
+    {
+      "epoch": 3.967342979548945,
+      "grad_norm": 0.8980403542518616,
+      "learning_rate": 0.0002,
+      "loss": 0.6564,
+      "step": 24540
+    },
+    {
+      "epoch": 3.9689596637296907,
+      "grad_norm": 0.8555465936660767,
+      "learning_rate": 0.0002,
+      "loss": 0.6099,
+      "step": 24550
+    },
+    {
+      "epoch": 3.970576347910436,
+      "grad_norm": 0.7825445532798767,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 24560
+    },
+    {
+      "epoch": 3.972193032091181,
+      "grad_norm": 0.7273133993148804,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 24570
+    },
+    {
+      "epoch": 3.9738097162719264,
+      "grad_norm": 0.9612047672271729,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 24580
+    },
+    {
+      "epoch": 3.9754264004526716,
+      "grad_norm": 0.9865460991859436,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 24590
+    },
+    {
+      "epoch": 3.977043084633417,
+      "grad_norm": 0.8638762831687927,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 24600
+    },
+    {
+      "epoch": 3.978659768814162,
+      "grad_norm": 1.0096198320388794,
+      "learning_rate": 0.0002,
+      "loss": 0.6097,
+      "step": 24610
+    },
+    {
+      "epoch": 3.9802764529949073,
+      "grad_norm": 0.8475532531738281,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 24620
+    },
+    {
+      "epoch": 3.9818931371756525,
+      "grad_norm": 0.9696195721626282,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 24630
+    },
+    {
+      "epoch": 3.9835098213563978,
+      "grad_norm": 0.7499843239784241,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 24640
+    },
+    {
+      "epoch": 3.9851265055371434,
+      "grad_norm": 0.8865424990653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6054,
+      "step": 24650
+    },
+    {
+      "epoch": 3.9867431897178887,
+      "grad_norm": 0.8089959025382996,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 24660
+    },
+    {
+      "epoch": 3.988359873898634,
+      "grad_norm": 0.6946012377738953,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 24670
+    },
+    {
+      "epoch": 3.989976558079379,
+      "grad_norm": 0.7991759181022644,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 24680
+    },
+    {
+      "epoch": 3.9915932422601244,
+      "grad_norm": 0.8803931474685669,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 24690
+    },
+    {
+      "epoch": 3.99320992644087,
+      "grad_norm": 0.8848299980163574,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 24700
+    },
+    {
+      "epoch": 3.9948266106216153,
+      "grad_norm": 0.7448889017105103,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 24710
+    },
+    {
+      "epoch": 3.9964432948023605,
+      "grad_norm": 0.9361620545387268,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24720
+    },
+    {
+      "epoch": 3.9980599789831057,
+      "grad_norm": 0.9958081245422363,
+      "learning_rate": 0.0002,
+      "loss": 0.5917,
+      "step": 24730
+    },
+    {
+      "epoch": 3.999676663163851,
+      "grad_norm": 1.026004672050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6567,
+      "step": 24740
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.1524168252944946,
+      "eval_runtime": 122.1585,
+      "eval_samples_per_second": 6.0,
+      "eval_steps_per_second": 0.753,
+      "step": 24742
+    },
+    {
+      "epoch": 4.001293347344596,
+      "grad_norm": 1.0664808750152588,
+      "learning_rate": 0.0002,
+      "loss": 0.6057,
+      "step": 24750
+    },
+    {
+      "epoch": 4.002910031525341,
+      "grad_norm": 1.0113720893859863,
+      "learning_rate": 0.0002,
+      "loss": 0.5644,
+      "step": 24760
+    },
+    {
+      "epoch": 4.004526715706087,
+      "grad_norm": 0.991486668586731,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 24770
+    },
+    {
+      "epoch": 4.006143399886832,
+      "grad_norm": 0.951754629611969,
+      "learning_rate": 0.0002,
+      "loss": 0.508,
+      "step": 24780
+    },
+    {
+      "epoch": 4.007760084067577,
+      "grad_norm": 1.13059401512146,
+      "learning_rate": 0.0002,
+      "loss": 0.5314,
+      "step": 24790
+    },
+    {
+      "epoch": 4.009376768248322,
+      "grad_norm": 0.9343926310539246,
+      "learning_rate": 0.0002,
+      "loss": 0.5323,
+      "step": 24800
+    },
+    {
+      "epoch": 4.010993452429068,
+      "grad_norm": 1.0680590867996216,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 24810
+    },
+    {
+      "epoch": 4.012610136609814,
+      "grad_norm": 1.0022706985473633,
+      "learning_rate": 0.0002,
+      "loss": 0.513,
+      "step": 24820
+    },
+    {
+      "epoch": 4.014226820790559,
+      "grad_norm": 1.0285297632217407,
+      "learning_rate": 0.0002,
+      "loss": 0.543,
+      "step": 24830
+    },
+    {
+      "epoch": 4.015843504971304,
+      "grad_norm": 0.8347002863883972,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 24840
+    },
+    {
+      "epoch": 4.017460189152049,
+      "grad_norm": 0.9675396680831909,
+      "learning_rate": 0.0002,
+      "loss": 0.5655,
+      "step": 24850
+    },
+    {
+      "epoch": 4.019076873332795,
+      "grad_norm": 0.9238511323928833,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 24860
+    },
+    {
+      "epoch": 4.02069355751354,
+      "grad_norm": 1.1576941013336182,
+      "learning_rate": 0.0002,
+      "loss": 0.5327,
+      "step": 24870
+    },
+    {
+      "epoch": 4.022310241694285,
+      "grad_norm": 0.8583757281303406,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 24880
+    },
+    {
+      "epoch": 4.02392692587503,
+      "grad_norm": 0.9816817045211792,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 24890
+    },
+    {
+      "epoch": 4.0255436100557755,
+      "grad_norm": 0.955073893070221,
+      "learning_rate": 0.0002,
+      "loss": 0.5605,
+      "step": 24900
+    },
+    {
+      "epoch": 4.027160294236521,
+      "grad_norm": 1.1054974794387817,
+      "learning_rate": 0.0002,
+      "loss": 0.4896,
+      "step": 24910
+    },
+    {
+      "epoch": 4.028776978417266,
+      "grad_norm": 1.1240060329437256,
+      "learning_rate": 0.0002,
+      "loss": 0.5246,
+      "step": 24920
+    },
+    {
+      "epoch": 4.030393662598011,
+      "grad_norm": 0.9512825012207031,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 24930
+    },
+    {
+      "epoch": 4.0320103467787565,
+      "grad_norm": 0.85965496301651,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 24940
+    },
+    {
+      "epoch": 4.033627030959502,
+      "grad_norm": 0.9378061294555664,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 24950
+    },
+    {
+      "epoch": 4.035243715140247,
+      "grad_norm": 0.9655424356460571,
+      "learning_rate": 0.0002,
+      "loss": 0.5008,
+      "step": 24960
+    },
+    {
+      "epoch": 4.036860399320993,
+      "grad_norm": 1.1393707990646362,
+      "learning_rate": 0.0002,
+      "loss": 0.5538,
+      "step": 24970
+    },
+    {
+      "epoch": 4.038477083501738,
+      "grad_norm": 1.0220451354980469,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 24980
+    },
+    {
+      "epoch": 4.0400937676824835,
+      "grad_norm": 0.9785808324813843,
+      "learning_rate": 0.0002,
+      "loss": 0.5813,
+      "step": 24990
+    },
+    {
+      "epoch": 4.041710451863229,
+      "grad_norm": 1.0257649421691895,
+      "learning_rate": 0.0002,
+      "loss": 0.5153,
+      "step": 25000
+    },
+    {
+      "epoch": 4.043327136043974,
+      "grad_norm": 0.9737892150878906,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 25010
+    },
+    {
+      "epoch": 4.044943820224719,
+      "grad_norm": 0.7416959404945374,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 25020
+    },
+    {
+      "epoch": 4.046560504405464,
+      "grad_norm": 0.7909596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 25030
+    },
+    {
+      "epoch": 4.04817718858621,
+      "grad_norm": 0.8923130631446838,
+      "learning_rate": 0.0002,
+      "loss": 0.5265,
+      "step": 25040
+    },
+    {
+      "epoch": 4.049793872766955,
+      "grad_norm": 0.9044941663742065,
+      "learning_rate": 0.0002,
+      "loss": 0.5035,
+      "step": 25050
+    },
+    {
+      "epoch": 4.0514105569477,
+      "grad_norm": 0.866352379322052,
+      "learning_rate": 0.0002,
+      "loss": 0.5135,
+      "step": 25060
+    },
+    {
+      "epoch": 4.053027241128445,
+      "grad_norm": 1.544549822807312,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 25070
+    },
+    {
+      "epoch": 4.054643925309191,
+      "grad_norm": 0.8426995277404785,
+      "learning_rate": 0.0002,
+      "loss": 0.5418,
+      "step": 25080
+    },
+    {
+      "epoch": 4.056260609489936,
+      "grad_norm": 0.9797548651695251,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 25090
+    },
+    {
+      "epoch": 4.057877293670681,
+      "grad_norm": 0.8468434810638428,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 25100
+    },
+    {
+      "epoch": 4.059493977851426,
+      "grad_norm": 0.9294559955596924,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 25110
+    },
+    {
+      "epoch": 4.061110662032172,
+      "grad_norm": 0.9686688780784607,
+      "learning_rate": 0.0002,
+      "loss": 0.5295,
+      "step": 25120
+    },
+    {
+      "epoch": 4.062727346212918,
+      "grad_norm": 0.8042728304862976,
+      "learning_rate": 0.0002,
+      "loss": 0.5642,
+      "step": 25130
+    },
+    {
+      "epoch": 4.064344030393663,
+      "grad_norm": 1.165160894393921,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 25140
+    },
+    {
+      "epoch": 4.065960714574408,
+      "grad_norm": 1.2161961793899536,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 25150
+    },
+    {
+      "epoch": 4.067577398755153,
+      "grad_norm": 1.0762810707092285,
+      "learning_rate": 0.0002,
+      "loss": 0.5217,
+      "step": 25160
+    },
+    {
+      "epoch": 4.069194082935899,
+      "grad_norm": 0.7580869793891907,
+      "learning_rate": 0.0002,
+      "loss": 0.5886,
+      "step": 25170
+    },
+    {
+      "epoch": 4.070810767116644,
+      "grad_norm": 0.9630117416381836,
+      "learning_rate": 0.0002,
+      "loss": 0.5401,
+      "step": 25180
+    },
+    {
+      "epoch": 4.072427451297389,
+      "grad_norm": 0.9049716591835022,
+      "learning_rate": 0.0002,
+      "loss": 0.5378,
+      "step": 25190
+    },
+    {
+      "epoch": 4.074044135478134,
+      "grad_norm": 1.1536930799484253,
+      "learning_rate": 0.0002,
+      "loss": 0.5266,
+      "step": 25200
+    },
+    {
+      "epoch": 4.0756608196588795,
+      "grad_norm": 0.901461124420166,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 25210
+    },
+    {
+      "epoch": 4.077277503839625,
+      "grad_norm": 1.3318437337875366,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 25220
+    },
+    {
+      "epoch": 4.07889418802037,
+      "grad_norm": 0.8811455368995667,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 25230
+    },
+    {
+      "epoch": 4.080510872201115,
+      "grad_norm": 1.0564165115356445,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 25240
+    },
+    {
+      "epoch": 4.08212755638186,
+      "grad_norm": 1.1008027791976929,
+      "learning_rate": 0.0002,
+      "loss": 0.5472,
+      "step": 25250
+    },
+    {
+      "epoch": 4.083744240562606,
+      "grad_norm": 1.150097131729126,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 25260
+    },
+    {
+      "epoch": 4.085360924743352,
+      "grad_norm": 0.9339924454689026,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 25270
+    },
+    {
+      "epoch": 4.086977608924097,
+      "grad_norm": 1.0902045965194702,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 25280
+    },
+    {
+      "epoch": 4.088594293104842,
+      "grad_norm": 0.8483911156654358,
+      "learning_rate": 0.0002,
+      "loss": 0.5203,
+      "step": 25290
+    },
+    {
+      "epoch": 4.0902109772855875,
+      "grad_norm": 0.9477024674415588,
+      "learning_rate": 0.0002,
+      "loss": 0.5697,
+      "step": 25300
+    },
+    {
+      "epoch": 4.091827661466333,
+      "grad_norm": 0.9500215649604797,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 25310
+    },
+    {
+      "epoch": 4.093444345647078,
+      "grad_norm": 1.040468454360962,
+      "learning_rate": 0.0002,
+      "loss": 0.5045,
+      "step": 25320
+    },
+    {
+      "epoch": 4.095061029827823,
+      "grad_norm": 0.7457592487335205,
+      "learning_rate": 0.0002,
+      "loss": 0.5488,
+      "step": 25330
+    },
+    {
+      "epoch": 4.096677714008568,
+      "grad_norm": 1.2092097997665405,
+      "learning_rate": 0.0002,
+      "loss": 0.609,
+      "step": 25340
+    },
+    {
+      "epoch": 4.098294398189314,
+      "grad_norm": 0.9652107954025269,
+      "learning_rate": 0.0002,
+      "loss": 0.5174,
+      "step": 25350
+    },
+    {
+      "epoch": 4.099911082370059,
+      "grad_norm": 0.8464955687522888,
+      "learning_rate": 0.0002,
+      "loss": 0.5559,
+      "step": 25360
+    },
+    {
+      "epoch": 4.101527766550804,
+      "grad_norm": 0.875026285648346,
+      "learning_rate": 0.0002,
+      "loss": 0.5635,
+      "step": 25370
+    },
+    {
+      "epoch": 4.103144450731549,
+      "grad_norm": 0.9241740107536316,
+      "learning_rate": 0.0002,
+      "loss": 0.5774,
+      "step": 25380
+    },
+    {
+      "epoch": 4.1047611349122946,
+      "grad_norm": 0.9769546389579773,
+      "learning_rate": 0.0002,
+      "loss": 0.5578,
+      "step": 25390
+    },
+    {
+      "epoch": 4.10637781909304,
+      "grad_norm": 1.1501960754394531,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 25400
+    },
+    {
+      "epoch": 4.107994503273786,
+      "grad_norm": 0.9135243892669678,
+      "learning_rate": 0.0002,
+      "loss": 0.5241,
+      "step": 25410
+    },
+    {
+      "epoch": 4.109611187454531,
+      "grad_norm": 0.9905396103858948,
+      "learning_rate": 0.0002,
+      "loss": 0.5152,
+      "step": 25420
+    },
+    {
+      "epoch": 4.111227871635276,
+      "grad_norm": 0.9845104217529297,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 25430
+    },
+    {
+      "epoch": 4.112844555816022,
+      "grad_norm": 0.8326883912086487,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 25440
+    },
+    {
+      "epoch": 4.114461239996767,
+      "grad_norm": 0.9264556765556335,
+      "learning_rate": 0.0002,
+      "loss": 0.5312,
+      "step": 25450
+    },
+    {
+      "epoch": 4.116077924177512,
+      "grad_norm": 1.043080449104309,
+      "learning_rate": 0.0002,
+      "loss": 0.5968,
+      "step": 25460
+    },
+    {
+      "epoch": 4.117694608358257,
+      "grad_norm": 0.8533386588096619,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 25470
+    },
+    {
+      "epoch": 4.1193112925390025,
+      "grad_norm": 1.0133965015411377,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 25480
+    },
+    {
+      "epoch": 4.120927976719748,
+      "grad_norm": 0.7476310133934021,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 25490
+    },
+    {
+      "epoch": 4.122544660900493,
+      "grad_norm": 1.1247259378433228,
+      "learning_rate": 0.0002,
+      "loss": 0.5189,
+      "step": 25500
+    },
+    {
+      "epoch": 4.124161345081238,
+      "grad_norm": 1.0764678716659546,
+      "learning_rate": 0.0002,
+      "loss": 0.5751,
+      "step": 25510
+    },
+    {
+      "epoch": 4.1257780292619834,
+      "grad_norm": 0.7679798007011414,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 25520
+    },
+    {
+      "epoch": 4.127394713442729,
+      "grad_norm": 0.8877071142196655,
+      "learning_rate": 0.0002,
+      "loss": 0.5233,
+      "step": 25530
+    },
+    {
+      "epoch": 4.129011397623474,
+      "grad_norm": 1.0440239906311035,
+      "learning_rate": 0.0002,
+      "loss": 0.5769,
+      "step": 25540
+    },
+    {
+      "epoch": 4.130628081804219,
+      "grad_norm": 0.984145998954773,
+      "learning_rate": 0.0002,
+      "loss": 0.5723,
+      "step": 25550
+    },
+    {
+      "epoch": 4.132244765984965,
+      "grad_norm": 0.8667055368423462,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 25560
+    },
+    {
+      "epoch": 4.1338614501657105,
+      "grad_norm": 1.1300835609436035,
+      "learning_rate": 0.0002,
+      "loss": 0.5816,
+      "step": 25570
+    },
+    {
+      "epoch": 4.135478134346456,
+      "grad_norm": 0.9314348101615906,
+      "learning_rate": 0.0002,
+      "loss": 0.524,
+      "step": 25580
+    },
+    {
+      "epoch": 4.137094818527201,
+      "grad_norm": 0.7731879949569702,
+      "learning_rate": 0.0002,
+      "loss": 0.5283,
+      "step": 25590
+    },
+    {
+      "epoch": 4.138711502707946,
+      "grad_norm": 1.0080097913742065,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 25600
+    },
+    {
+      "epoch": 4.140328186888691,
+      "grad_norm": 1.2475038766860962,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 25610
+    },
+    {
+      "epoch": 4.141944871069437,
+      "grad_norm": 0.9912930727005005,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 25620
+    },
+    {
+      "epoch": 4.143561555250182,
+      "grad_norm": 0.9088651537895203,
+      "learning_rate": 0.0002,
+      "loss": 0.5624,
+      "step": 25630
+    },
+    {
+      "epoch": 4.145178239430927,
+      "grad_norm": 0.8940697312355042,
+      "learning_rate": 0.0002,
+      "loss": 0.5393,
+      "step": 25640
+    },
+    {
+      "epoch": 4.146794923611672,
+      "grad_norm": 1.0798203945159912,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 25650
+    },
+    {
+      "epoch": 4.148411607792418,
+      "grad_norm": 0.955172061920166,
+      "learning_rate": 0.0002,
+      "loss": 0.5987,
+      "step": 25660
+    },
+    {
+      "epoch": 4.150028291973163,
+      "grad_norm": 0.9692716002464294,
+      "learning_rate": 0.0002,
+      "loss": 0.569,
+      "step": 25670
+    },
+    {
+      "epoch": 4.151644976153908,
+      "grad_norm": 1.0813939571380615,
+      "learning_rate": 0.0002,
+      "loss": 0.5478,
+      "step": 25680
+    },
+    {
+      "epoch": 4.153261660334653,
+      "grad_norm": 1.135675072669983,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 25690
+    },
+    {
+      "epoch": 4.1548783445153985,
+      "grad_norm": 1.0392236709594727,
+      "learning_rate": 0.0002,
+      "loss": 0.5247,
+      "step": 25700
+    },
+    {
+      "epoch": 4.156495028696145,
+      "grad_norm": 0.9473116993904114,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 25710
+    },
+    {
+      "epoch": 4.15811171287689,
+      "grad_norm": 0.712493896484375,
+      "learning_rate": 0.0002,
+      "loss": 0.5339,
+      "step": 25720
+    },
+    {
+      "epoch": 4.159728397057635,
+      "grad_norm": 0.8724465370178223,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 25730
+    },
+    {
+      "epoch": 4.16134508123838,
+      "grad_norm": 0.9870015978813171,
+      "learning_rate": 0.0002,
+      "loss": 0.5325,
+      "step": 25740
+    },
+    {
+      "epoch": 4.1629617654191255,
+      "grad_norm": 1.025273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 25750
+    },
+    {
+      "epoch": 4.164578449599871,
+      "grad_norm": 0.9243090152740479,
+      "learning_rate": 0.0002,
+      "loss": 0.5223,
+      "step": 25760
+    },
+    {
+      "epoch": 4.166195133780616,
+      "grad_norm": 1.1656451225280762,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 25770
+    },
+    {
+      "epoch": 4.167811817961361,
+      "grad_norm": 0.936358630657196,
+      "learning_rate": 0.0002,
+      "loss": 0.5334,
+      "step": 25780
+    },
+    {
+      "epoch": 4.1694285021421065,
+      "grad_norm": 0.8618208169937134,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 25790
+    },
+    {
+      "epoch": 4.171045186322852,
+      "grad_norm": 0.8580600023269653,
+      "learning_rate": 0.0002,
+      "loss": 0.5186,
+      "step": 25800
+    },
+    {
+      "epoch": 4.172661870503597,
+      "grad_norm": 1.0128562450408936,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 25810
+    },
+    {
+      "epoch": 4.174278554684342,
+      "grad_norm": 0.854865312576294,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 25820
+    },
+    {
+      "epoch": 4.175895238865087,
+      "grad_norm": 1.235082745552063,
+      "learning_rate": 0.0002,
+      "loss": 0.5377,
+      "step": 25830
+    },
+    {
+      "epoch": 4.177511923045833,
+      "grad_norm": 0.9796220660209656,
+      "learning_rate": 0.0002,
+      "loss": 0.5614,
+      "step": 25840
+    },
+    {
+      "epoch": 4.179128607226578,
+      "grad_norm": 0.8922094702720642,
+      "learning_rate": 0.0002,
+      "loss": 0.5689,
+      "step": 25850
+    },
+    {
+      "epoch": 4.180745291407324,
+      "grad_norm": 0.9672530293464661,
+      "learning_rate": 0.0002,
+      "loss": 0.5806,
+      "step": 25860
+    },
+    {
+      "epoch": 4.182361975588069,
+      "grad_norm": 0.8662548661231995,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 25870
+    },
+    {
+      "epoch": 4.1839786597688144,
+      "grad_norm": 0.7938798069953918,
+      "learning_rate": 0.0002,
+      "loss": 0.5329,
+      "step": 25880
+    },
+    {
+      "epoch": 4.18559534394956,
+      "grad_norm": 1.0517958402633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5427,
+      "step": 25890
+    },
+    {
+      "epoch": 4.187212028130305,
+      "grad_norm": 0.8939275145530701,
+      "learning_rate": 0.0002,
+      "loss": 0.5147,
+      "step": 25900
+    },
+    {
+      "epoch": 4.18882871231105,
+      "grad_norm": 1.0296672582626343,
+      "learning_rate": 0.0002,
+      "loss": 0.5199,
+      "step": 25910
+    },
+    {
+      "epoch": 4.190445396491795,
+      "grad_norm": 0.8104017972946167,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 25920
+    },
+    {
+      "epoch": 4.192062080672541,
+      "grad_norm": 0.9984509944915771,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 25930
+    },
+    {
+      "epoch": 4.193678764853286,
+      "grad_norm": 0.9844784736633301,
+      "learning_rate": 0.0002,
+      "loss": 0.5356,
+      "step": 25940
+    },
+    {
+      "epoch": 4.195295449034031,
+      "grad_norm": 0.8168622255325317,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 25950
+    },
+    {
+      "epoch": 4.196912133214776,
+      "grad_norm": 1.0878913402557373,
+      "learning_rate": 0.0002,
+      "loss": 0.542,
+      "step": 25960
+    },
+    {
+      "epoch": 4.1985288173955215,
+      "grad_norm": 0.927126407623291,
+      "learning_rate": 0.0002,
+      "loss": 0.5414,
+      "step": 25970
+    },
+    {
+      "epoch": 4.200145501576267,
+      "grad_norm": 0.838586688041687,
+      "learning_rate": 0.0002,
+      "loss": 0.5794,
+      "step": 25980
+    },
+    {
+      "epoch": 4.201762185757012,
+      "grad_norm": 1.2572145462036133,
+      "learning_rate": 0.0002,
+      "loss": 0.5454,
+      "step": 25990
+    },
+    {
+      "epoch": 4.203378869937758,
+      "grad_norm": 1.0476740598678589,
+      "learning_rate": 0.0002,
+      "loss": 0.5048,
+      "step": 26000
+    },
+    {
+      "epoch": 4.204995554118503,
+      "grad_norm": 1.0873368978500366,
+      "learning_rate": 0.0002,
+      "loss": 0.5127,
+      "step": 26010
+    },
+    {
+      "epoch": 4.206612238299249,
+      "grad_norm": 1.2664896249771118,
+      "learning_rate": 0.0002,
+      "loss": 0.5679,
+      "step": 26020
+    },
+    {
+      "epoch": 4.208228922479994,
+      "grad_norm": 1.0312391519546509,
+      "learning_rate": 0.0002,
+      "loss": 0.5814,
+      "step": 26030
+    },
+    {
+      "epoch": 4.209845606660739,
+      "grad_norm": 1.0235042572021484,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 26040
+    },
+    {
+      "epoch": 4.211462290841484,
+      "grad_norm": 0.8882219195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 26050
+    },
+    {
+      "epoch": 4.2130789750222295,
+      "grad_norm": 0.9115961790084839,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 26060
+    },
+    {
+      "epoch": 4.214695659202975,
+      "grad_norm": 1.0218228101730347,
+      "learning_rate": 0.0002,
+      "loss": 0.5455,
+      "step": 26070
+    },
+    {
+      "epoch": 4.21631234338372,
+      "grad_norm": 1.0802232027053833,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 26080
+    },
+    {
+      "epoch": 4.217929027564465,
+      "grad_norm": 1.1488053798675537,
+      "learning_rate": 0.0002,
+      "loss": 0.557,
+      "step": 26090
+    },
+    {
+      "epoch": 4.21954571174521,
+      "grad_norm": 1.0487725734710693,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 26100
+    },
+    {
+      "epoch": 4.221162395925956,
+      "grad_norm": 0.9131165742874146,
+      "learning_rate": 0.0002,
+      "loss": 0.5568,
+      "step": 26110
+    },
+    {
+      "epoch": 4.222779080106701,
+      "grad_norm": 0.9012845158576965,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 26120
+    },
+    {
+      "epoch": 4.224395764287446,
+      "grad_norm": 0.8389840126037598,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 26130
+    },
+    {
+      "epoch": 4.226012448468191,
+      "grad_norm": 0.8924660682678223,
+      "learning_rate": 0.0002,
+      "loss": 0.5268,
+      "step": 26140
+    },
+    {
+      "epoch": 4.2276291326489375,
+      "grad_norm": 0.8556463718414307,
+      "learning_rate": 0.0002,
+      "loss": 0.5715,
+      "step": 26150
+    },
+    {
+      "epoch": 4.229245816829683,
+      "grad_norm": 0.9643129110336304,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 26160
+    },
+    {
+      "epoch": 4.230862501010428,
+      "grad_norm": 0.9865712523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 26170
+    },
+    {
+      "epoch": 4.232479185191173,
+      "grad_norm": 1.152641773223877,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 26180
+    },
+    {
+      "epoch": 4.234095869371918,
+      "grad_norm": 0.9157698154449463,
+      "learning_rate": 0.0002,
+      "loss": 0.5632,
+      "step": 26190
+    },
+    {
+      "epoch": 4.235712553552664,
+      "grad_norm": 0.8418048620223999,
+      "learning_rate": 0.0002,
+      "loss": 0.5717,
+      "step": 26200
+    },
+    {
+      "epoch": 4.237329237733409,
+      "grad_norm": 0.9430168867111206,
+      "learning_rate": 0.0002,
+      "loss": 0.5624,
+      "step": 26210
+    },
+    {
+      "epoch": 4.238945921914154,
+      "grad_norm": 1.012582778930664,
+      "learning_rate": 0.0002,
+      "loss": 0.5574,
+      "step": 26220
+    },
+    {
+      "epoch": 4.240562606094899,
+      "grad_norm": 1.112619400024414,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 26230
+    },
+    {
+      "epoch": 4.2421792902756446,
+      "grad_norm": 0.9243621826171875,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 26240
+    },
+    {
+      "epoch": 4.24379597445639,
+      "grad_norm": 0.6977595686912537,
+      "learning_rate": 0.0002,
+      "loss": 0.569,
+      "step": 26250
+    },
+    {
+      "epoch": 4.245412658637135,
+      "grad_norm": 0.9600721597671509,
+      "learning_rate": 0.0002,
+      "loss": 0.5379,
+      "step": 26260
+    },
+    {
+      "epoch": 4.24702934281788,
+      "grad_norm": 0.882641613483429,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 26270
+    },
+    {
+      "epoch": 4.2486460269986255,
+      "grad_norm": 1.010920763015747,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 26280
+    },
+    {
+      "epoch": 4.250262711179371,
+      "grad_norm": 0.9289400577545166,
+      "learning_rate": 0.0002,
+      "loss": 0.5803,
+      "step": 26290
+    },
+    {
+      "epoch": 4.251879395360117,
+      "grad_norm": 1.137397289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.541,
+      "step": 26300
+    },
+    {
+      "epoch": 4.253496079540862,
+      "grad_norm": 1.0136182308197021,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 26310
+    },
+    {
+      "epoch": 4.255112763721607,
+      "grad_norm": 0.9387356042861938,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 26320
+    },
+    {
+      "epoch": 4.2567294479023525,
+      "grad_norm": 1.1833957433700562,
+      "learning_rate": 0.0002,
+      "loss": 0.5948,
+      "step": 26330
+    },
+    {
+      "epoch": 4.258346132083098,
+      "grad_norm": 0.9415934681892395,
+      "learning_rate": 0.0002,
+      "loss": 0.5905,
+      "step": 26340
+    },
+    {
+      "epoch": 4.259962816263843,
+      "grad_norm": 0.8550165891647339,
+      "learning_rate": 0.0002,
+      "loss": 0.5539,
+      "step": 26350
+    },
+    {
+      "epoch": 4.261579500444588,
+      "grad_norm": 9.924622535705566,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 26360
+    },
+    {
+      "epoch": 4.2631961846253335,
+      "grad_norm": 1.0104902982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.5689,
+      "step": 26370
+    },
+    {
+      "epoch": 4.264812868806079,
+      "grad_norm": 0.890794038772583,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 26380
+    },
+    {
+      "epoch": 4.266429552986824,
+      "grad_norm": 1.0560191869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.563,
+      "step": 26390
+    },
+    {
+      "epoch": 4.268046237167569,
+      "grad_norm": 1.0135581493377686,
+      "learning_rate": 0.0002,
+      "loss": 0.5119,
+      "step": 26400
+    },
+    {
+      "epoch": 4.269662921348314,
+      "grad_norm": 1.1304140090942383,
+      "learning_rate": 0.0002,
+      "loss": 0.5359,
+      "step": 26410
+    },
+    {
+      "epoch": 4.27127960552906,
+      "grad_norm": 0.9899303913116455,
+      "learning_rate": 0.0002,
+      "loss": 0.5615,
+      "step": 26420
+    },
+    {
+      "epoch": 4.272896289709805,
+      "grad_norm": 1.0505329370498657,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 26430
+    },
+    {
+      "epoch": 4.27451297389055,
+      "grad_norm": 0.9389396905899048,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 26440
+    },
+    {
+      "epoch": 4.276129658071296,
+      "grad_norm": 0.875328779220581,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 26450
+    },
+    {
+      "epoch": 4.277746342252041,
+      "grad_norm": 1.0689256191253662,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 26460
+    },
+    {
+      "epoch": 4.279363026432787,
+      "grad_norm": 0.9988957643508911,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 26470
+    },
+    {
+      "epoch": 4.280979710613532,
+      "grad_norm": 0.8721813559532166,
+      "learning_rate": 0.0002,
+      "loss": 0.5478,
+      "step": 26480
+    },
+    {
+      "epoch": 4.282596394794277,
+      "grad_norm": 1.100109577178955,
+      "learning_rate": 0.0002,
+      "loss": 0.5424,
+      "step": 26490
+    },
+    {
+      "epoch": 4.284213078975022,
+      "grad_norm": 1.1607271432876587,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 26500
+    },
+    {
+      "epoch": 4.285829763155768,
+      "grad_norm": 0.879088819026947,
+      "learning_rate": 0.0002,
+      "loss": 0.6287,
+      "step": 26510
+    },
+    {
+      "epoch": 4.287446447336513,
+      "grad_norm": 0.9891700744628906,
+      "learning_rate": 0.0002,
+      "loss": 0.573,
+      "step": 26520
+    },
+    {
+      "epoch": 4.289063131517258,
+      "grad_norm": 1.0831127166748047,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 26530
+    },
+    {
+      "epoch": 4.290679815698003,
+      "grad_norm": 1.4108285903930664,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 26540
+    },
+    {
+      "epoch": 4.2922964998787485,
+      "grad_norm": 1.0630289316177368,
+      "learning_rate": 0.0002,
+      "loss": 0.5888,
+      "step": 26550
+    },
+    {
+      "epoch": 4.293913184059494,
+      "grad_norm": 1.0854572057724,
+      "learning_rate": 0.0002,
+      "loss": 0.5817,
+      "step": 26560
+    },
+    {
+      "epoch": 4.295529868240239,
+      "grad_norm": 0.9561646580696106,
+      "learning_rate": 0.0002,
+      "loss": 0.5586,
+      "step": 26570
+    },
+    {
+      "epoch": 4.297146552420984,
+      "grad_norm": 0.9064981937408447,
+      "learning_rate": 0.0002,
+      "loss": 0.5674,
+      "step": 26580
+    },
+    {
+      "epoch": 4.298763236601729,
+      "grad_norm": 1.0082972049713135,
+      "learning_rate": 0.0002,
+      "loss": 0.5847,
+      "step": 26590
+    },
+    {
+      "epoch": 4.3003799207824756,
+      "grad_norm": 1.1613214015960693,
+      "learning_rate": 0.0002,
+      "loss": 0.5711,
+      "step": 26600
+    },
+    {
+      "epoch": 4.301996604963221,
+      "grad_norm": 0.9847695231437683,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 26610
+    },
+    {
+      "epoch": 4.303613289143966,
+      "grad_norm": 1.0980697870254517,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 26620
+    },
+    {
+      "epoch": 4.305229973324711,
+      "grad_norm": 0.8861175179481506,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 26630
+    },
+    {
+      "epoch": 4.3068466575054565,
+      "grad_norm": 0.8917363286018372,
+      "learning_rate": 0.0002,
+      "loss": 0.5716,
+      "step": 26640
+    },
+    {
+      "epoch": 4.308463341686202,
+      "grad_norm": 1.0458378791809082,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 26650
+    },
+    {
+      "epoch": 4.310080025866947,
+      "grad_norm": 1.4859240055084229,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 26660
+    },
+    {
+      "epoch": 4.311696710047692,
+      "grad_norm": 1.1376359462738037,
+      "learning_rate": 0.0002,
+      "loss": 0.5296,
+      "step": 26670
+    },
+    {
+      "epoch": 4.313313394228437,
+      "grad_norm": 0.991349995136261,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 26680
+    },
+    {
+      "epoch": 4.314930078409183,
+      "grad_norm": 0.9995543956756592,
+      "learning_rate": 0.0002,
+      "loss": 0.5338,
+      "step": 26690
+    },
+    {
+      "epoch": 4.316546762589928,
+      "grad_norm": 1.0515851974487305,
+      "learning_rate": 0.0002,
+      "loss": 0.5542,
+      "step": 26700
+    },
+    {
+      "epoch": 4.318163446770673,
+      "grad_norm": 1.008023977279663,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 26710
+    },
+    {
+      "epoch": 4.319780130951418,
+      "grad_norm": 1.0184582471847534,
+      "learning_rate": 0.0002,
+      "loss": 0.5506,
+      "step": 26720
+    },
+    {
+      "epoch": 4.321396815132164,
+      "grad_norm": 1.161071538925171,
+      "learning_rate": 0.0002,
+      "loss": 0.5828,
+      "step": 26730
+    },
+    {
+      "epoch": 4.323013499312909,
+      "grad_norm": 0.9580779671669006,
+      "learning_rate": 0.0002,
+      "loss": 0.5633,
+      "step": 26740
+    },
+    {
+      "epoch": 4.324630183493655,
+      "grad_norm": 1.0189911127090454,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 26750
+    },
+    {
+      "epoch": 4.3262468676744,
+      "grad_norm": 0.7484358549118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 26760
+    },
+    {
+      "epoch": 4.327863551855145,
+      "grad_norm": 1.0015908479690552,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 26770
+    },
+    {
+      "epoch": 4.329480236035891,
+      "grad_norm": 0.8972945809364319,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 26780
+    },
+    {
+      "epoch": 4.331096920216636,
+      "grad_norm": 1.01099693775177,
+      "learning_rate": 0.0002,
+      "loss": 0.5857,
+      "step": 26790
+    },
+    {
+      "epoch": 4.332713604397381,
+      "grad_norm": 0.846958339214325,
+      "learning_rate": 0.0002,
+      "loss": 0.5591,
+      "step": 26800
+    },
+    {
+      "epoch": 4.334330288578126,
+      "grad_norm": 1.0792603492736816,
+      "learning_rate": 0.0002,
+      "loss": 0.5547,
+      "step": 26810
+    },
+    {
+      "epoch": 4.3359469727588715,
+      "grad_norm": 1.0373345613479614,
+      "learning_rate": 0.0002,
+      "loss": 0.5747,
+      "step": 26820
+    },
+    {
+      "epoch": 4.337563656939617,
+      "grad_norm": 0.9779167771339417,
+      "learning_rate": 0.0002,
+      "loss": 0.558,
+      "step": 26830
+    },
+    {
+      "epoch": 4.339180341120362,
+      "grad_norm": 1.0235520601272583,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 26840
+    },
+    {
+      "epoch": 4.340797025301107,
+      "grad_norm": 1.04195237159729,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 26850
+    },
+    {
+      "epoch": 4.3424137094818525,
+      "grad_norm": 0.9479565620422363,
+      "learning_rate": 0.0002,
+      "loss": 0.5474,
+      "step": 26860
+    },
+    {
+      "epoch": 4.344030393662598,
+      "grad_norm": 0.9526172280311584,
+      "learning_rate": 0.0002,
+      "loss": 0.5646,
+      "step": 26870
+    },
+    {
+      "epoch": 4.345647077843343,
+      "grad_norm": 0.8571456074714661,
+      "learning_rate": 0.0002,
+      "loss": 0.521,
+      "step": 26880
+    },
+    {
+      "epoch": 4.347263762024088,
+      "grad_norm": 0.9475828409194946,
+      "learning_rate": 0.0002,
+      "loss": 0.5846,
+      "step": 26890
+    },
+    {
+      "epoch": 4.348880446204834,
+      "grad_norm": 1.0529576539993286,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 26900
+    },
+    {
+      "epoch": 4.3504971303855795,
+      "grad_norm": 0.9648140072822571,
+      "learning_rate": 0.0002,
+      "loss": 0.56,
+      "step": 26910
+    },
+    {
+      "epoch": 4.352113814566325,
+      "grad_norm": 1.0488841533660889,
+      "learning_rate": 0.0002,
+      "loss": 0.5162,
+      "step": 26920
+    },
+    {
+      "epoch": 4.35373049874707,
+      "grad_norm": 0.8771942257881165,
+      "learning_rate": 0.0002,
+      "loss": 0.5842,
+      "step": 26930
+    },
+    {
+      "epoch": 4.355347182927815,
+      "grad_norm": 0.9411202073097229,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 26940
+    },
+    {
+      "epoch": 4.35696386710856,
+      "grad_norm": 1.0997588634490967,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 26950
+    },
+    {
+      "epoch": 4.358580551289306,
+      "grad_norm": 0.968754768371582,
+      "learning_rate": 0.0002,
+      "loss": 0.5528,
+      "step": 26960
+    },
+    {
+      "epoch": 4.360197235470051,
+      "grad_norm": 0.9990773797035217,
+      "learning_rate": 0.0002,
+      "loss": 0.5881,
+      "step": 26970
+    },
+    {
+      "epoch": 4.361813919650796,
+      "grad_norm": 1.0210620164871216,
+      "learning_rate": 0.0002,
+      "loss": 0.5761,
+      "step": 26980
+    },
+    {
+      "epoch": 4.363430603831541,
+      "grad_norm": 0.855462908744812,
+      "learning_rate": 0.0002,
+      "loss": 0.5768,
+      "step": 26990
+    },
+    {
+      "epoch": 4.365047288012287,
+      "grad_norm": 0.9169660806655884,
+      "learning_rate": 0.0002,
+      "loss": 0.5493,
+      "step": 27000
+    },
+    {
+      "epoch": 4.366663972193032,
+      "grad_norm": 1.089629888534546,
+      "learning_rate": 0.0002,
+      "loss": 0.5697,
+      "step": 27010
+    },
+    {
+      "epoch": 4.368280656373777,
+      "grad_norm": 1.0932867527008057,
+      "learning_rate": 0.0002,
+      "loss": 0.5854,
+      "step": 27020
+    },
+    {
+      "epoch": 4.369897340554522,
+      "grad_norm": 0.9290956854820251,
+      "learning_rate": 0.0002,
+      "loss": 0.5656,
+      "step": 27030
+    },
+    {
+      "epoch": 4.3715140247352675,
+      "grad_norm": 1.2800624370574951,
+      "learning_rate": 0.0002,
+      "loss": 0.5727,
+      "step": 27040
+    },
+    {
+      "epoch": 4.373130708916014,
+      "grad_norm": 0.8993493318557739,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 27050
+    },
+    {
+      "epoch": 4.374747393096759,
+      "grad_norm": 1.1566431522369385,
+      "learning_rate": 0.0002,
+      "loss": 0.6232,
+      "step": 27060
+    },
+    {
+      "epoch": 4.376364077277504,
+      "grad_norm": 0.9479052424430847,
+      "learning_rate": 0.0002,
+      "loss": 0.5902,
+      "step": 27070
+    },
+    {
+      "epoch": 4.377980761458249,
+      "grad_norm": 1.0063648223876953,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 27080
+    },
+    {
+      "epoch": 4.379597445638995,
+      "grad_norm": 0.8342045545578003,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 27090
+    },
+    {
+      "epoch": 4.38121412981974,
+      "grad_norm": 1.1390739679336548,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 27100
+    },
+    {
+      "epoch": 4.382830814000485,
+      "grad_norm": 0.9547637104988098,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 27110
+    },
+    {
+      "epoch": 4.38444749818123,
+      "grad_norm": 1.0503804683685303,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 27120
+    },
+    {
+      "epoch": 4.3860641823619755,
+      "grad_norm": 0.9064017534255981,
+      "learning_rate": 0.0002,
+      "loss": 0.5787,
+      "step": 27130
+    },
+    {
+      "epoch": 4.387680866542721,
+      "grad_norm": 0.9382519125938416,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 27140
+    },
+    {
+      "epoch": 4.389297550723466,
+      "grad_norm": 1.0410341024398804,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 27150
+    },
+    {
+      "epoch": 4.390914234904211,
+      "grad_norm": 0.9218655824661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6034,
+      "step": 27160
+    },
+    {
+      "epoch": 4.392530919084956,
+      "grad_norm": 0.8119737505912781,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 27170
+    },
+    {
+      "epoch": 4.394147603265702,
+      "grad_norm": 0.8584722876548767,
+      "learning_rate": 0.0002,
+      "loss": 0.5612,
+      "step": 27180
+    },
+    {
+      "epoch": 4.395764287446447,
+      "grad_norm": 0.9668293595314026,
+      "learning_rate": 0.0002,
+      "loss": 0.5772,
+      "step": 27190
+    },
+    {
+      "epoch": 4.397380971627193,
+      "grad_norm": 1.022334098815918,
+      "learning_rate": 0.0002,
+      "loss": 0.6009,
+      "step": 27200
+    },
+    {
+      "epoch": 4.398997655807938,
+      "grad_norm": 0.9553216099739075,
+      "learning_rate": 0.0002,
+      "loss": 0.5573,
+      "step": 27210
+    },
+    {
+      "epoch": 4.4006143399886835,
+      "grad_norm": 0.9282339215278625,
+      "learning_rate": 0.0002,
+      "loss": 0.5604,
+      "step": 27220
+    },
+    {
+      "epoch": 4.402231024169429,
+      "grad_norm": 1.0232292413711548,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 27230
+    },
+    {
+      "epoch": 4.403847708350174,
+      "grad_norm": 0.9915700554847717,
+      "learning_rate": 0.0002,
+      "loss": 0.6078,
+      "step": 27240
+    },
+    {
+      "epoch": 4.405464392530919,
+      "grad_norm": 1.0014961957931519,
+      "learning_rate": 0.0002,
+      "loss": 0.5778,
+      "step": 27250
+    },
+    {
+      "epoch": 4.407081076711664,
+      "grad_norm": 1.1172103881835938,
+      "learning_rate": 0.0002,
+      "loss": 0.5824,
+      "step": 27260
+    },
+    {
+      "epoch": 4.40869776089241,
+      "grad_norm": 0.8583093285560608,
+      "learning_rate": 0.0002,
+      "loss": 0.5286,
+      "step": 27270
+    },
+    {
+      "epoch": 4.410314445073155,
+      "grad_norm": 0.7609201669692993,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 27280
+    },
+    {
+      "epoch": 4.4119311292539,
+      "grad_norm": 1.0619351863861084,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 27290
+    },
+    {
+      "epoch": 4.413547813434645,
+      "grad_norm": 1.0177674293518066,
+      "learning_rate": 0.0002,
+      "loss": 0.5579,
+      "step": 27300
+    },
+    {
+      "epoch": 4.4151644976153905,
+      "grad_norm": 0.9921218156814575,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 27310
+    },
+    {
+      "epoch": 4.416781181796136,
+      "grad_norm": 1.126244306564331,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 27320
+    },
+    {
+      "epoch": 4.418397865976881,
+      "grad_norm": 1.0678540468215942,
+      "learning_rate": 0.0002,
+      "loss": 0.5743,
+      "step": 27330
+    },
+    {
+      "epoch": 4.420014550157627,
+      "grad_norm": 0.8705704212188721,
+      "learning_rate": 0.0002,
+      "loss": 0.5665,
+      "step": 27340
+    },
+    {
+      "epoch": 4.421631234338372,
+      "grad_norm": 1.272074818611145,
+      "learning_rate": 0.0002,
+      "loss": 0.5763,
+      "step": 27350
+    },
+    {
+      "epoch": 4.423247918519118,
+      "grad_norm": 0.8740444183349609,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 27360
+    },
+    {
+      "epoch": 4.424864602699863,
+      "grad_norm": 1.0584250688552856,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 27370
+    },
+    {
+      "epoch": 4.426481286880608,
+      "grad_norm": 1.059870719909668,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 27380
+    },
+    {
+      "epoch": 4.428097971061353,
+      "grad_norm": 1.072265863418579,
+      "learning_rate": 0.0002,
+      "loss": 0.5551,
+      "step": 27390
+    },
+    {
+      "epoch": 4.4297146552420985,
+      "grad_norm": 0.871481716632843,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 27400
+    },
+    {
+      "epoch": 4.431331339422844,
+      "grad_norm": 0.9555448293685913,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 27410
+    },
+    {
+      "epoch": 4.432948023603589,
+      "grad_norm": 1.0402292013168335,
+      "learning_rate": 0.0002,
+      "loss": 0.5593,
+      "step": 27420
+    },
+    {
+      "epoch": 4.434564707784334,
+      "grad_norm": 1.12587571144104,
+      "learning_rate": 0.0002,
+      "loss": 0.5532,
+      "step": 27430
+    },
+    {
+      "epoch": 4.436181391965079,
+      "grad_norm": 1.0783193111419678,
+      "learning_rate": 0.0002,
+      "loss": 0.5403,
+      "step": 27440
+    },
+    {
+      "epoch": 4.437798076145825,
+      "grad_norm": 1.024133563041687,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 27450
+    },
+    {
+      "epoch": 4.43941476032657,
+      "grad_norm": 0.9156768918037415,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 27460
+    },
+    {
+      "epoch": 4.441031444507315,
+      "grad_norm": 1.0215224027633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 27470
+    },
+    {
+      "epoch": 4.442648128688061,
+      "grad_norm": 1.082116961479187,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 27480
+    },
+    {
+      "epoch": 4.4442648128688065,
+      "grad_norm": 1.0412873029708862,
+      "learning_rate": 0.0002,
+      "loss": 0.6203,
+      "step": 27490
+    },
+    {
+      "epoch": 4.445881497049552,
+      "grad_norm": 1.0509289503097534,
+      "learning_rate": 0.0002,
+      "loss": 0.5939,
+      "step": 27500
+    },
+    {
+      "epoch": 4.447498181230297,
+      "grad_norm": 0.9291498064994812,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 27510
+    },
+    {
+      "epoch": 4.449114865411042,
+      "grad_norm": 0.970184326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.5408,
+      "step": 27520
+    },
+    {
+      "epoch": 4.450731549591787,
+      "grad_norm": 0.8418883681297302,
+      "learning_rate": 0.0002,
+      "loss": 0.5705,
+      "step": 27530
+    },
+    {
+      "epoch": 4.452348233772533,
+      "grad_norm": 0.8823825120925903,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 27540
+    },
+    {
+      "epoch": 4.453964917953278,
+      "grad_norm": 1.1909019947052002,
+      "learning_rate": 0.0002,
+      "loss": 0.5867,
+      "step": 27550
+    },
+    {
+      "epoch": 4.455581602134023,
+      "grad_norm": 1.0317302942276,
+      "learning_rate": 0.0002,
+      "loss": 0.5685,
+      "step": 27560
+    },
+    {
+      "epoch": 4.457198286314768,
+      "grad_norm": 0.9977751970291138,
+      "learning_rate": 0.0002,
+      "loss": 0.5538,
+      "step": 27570
+    },
+    {
+      "epoch": 4.458814970495514,
+      "grad_norm": 0.8909519910812378,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 27580
+    },
+    {
+      "epoch": 4.460431654676259,
+      "grad_norm": 0.8653029799461365,
+      "learning_rate": 0.0002,
+      "loss": 0.6099,
+      "step": 27590
+    },
+    {
+      "epoch": 4.462048338857004,
+      "grad_norm": 1.0783653259277344,
+      "learning_rate": 0.0002,
+      "loss": 0.5622,
+      "step": 27600
+    },
+    {
+      "epoch": 4.463665023037749,
+      "grad_norm": 1.1235394477844238,
+      "learning_rate": 0.0002,
+      "loss": 0.579,
+      "step": 27610
+    },
+    {
+      "epoch": 4.4652817072184945,
+      "grad_norm": 0.9386643767356873,
+      "learning_rate": 0.0002,
+      "loss": 0.5545,
+      "step": 27620
+    },
+    {
+      "epoch": 4.466898391399241,
+      "grad_norm": 1.0605148077011108,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 27630
+    },
+    {
+      "epoch": 4.468515075579986,
+      "grad_norm": 1.1283893585205078,
+      "learning_rate": 0.0002,
+      "loss": 0.5886,
+      "step": 27640
+    },
+    {
+      "epoch": 4.470131759760731,
+      "grad_norm": 1.0583468675613403,
+      "learning_rate": 0.0002,
+      "loss": 0.5801,
+      "step": 27650
+    },
+    {
+      "epoch": 4.471748443941476,
+      "grad_norm": 0.9563992023468018,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 27660
+    },
+    {
+      "epoch": 4.4733651281222215,
+      "grad_norm": 1.100598931312561,
+      "learning_rate": 0.0002,
+      "loss": 0.5687,
+      "step": 27670
+    },
+    {
+      "epoch": 4.474981812302967,
+      "grad_norm": 0.9386957287788391,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 27680
+    },
+    {
+      "epoch": 4.476598496483712,
+      "grad_norm": 1.2946288585662842,
+      "learning_rate": 0.0002,
+      "loss": 0.6241,
+      "step": 27690
+    },
+    {
+      "epoch": 4.478215180664457,
+      "grad_norm": 1.0325199365615845,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 27700
+    },
+    {
+      "epoch": 4.4798318648452025,
+      "grad_norm": 1.0318928956985474,
+      "learning_rate": 0.0002,
+      "loss": 0.588,
+      "step": 27710
+    },
+    {
+      "epoch": 4.481448549025948,
+      "grad_norm": 0.8721024394035339,
+      "learning_rate": 0.0002,
+      "loss": 0.5656,
+      "step": 27720
+    },
+    {
+      "epoch": 4.483065233206693,
+      "grad_norm": 1.17376708984375,
+      "learning_rate": 0.0002,
+      "loss": 0.5421,
+      "step": 27730
+    },
+    {
+      "epoch": 4.484681917387438,
+      "grad_norm": 1.0926326513290405,
+      "learning_rate": 0.0002,
+      "loss": 0.5657,
+      "step": 27740
+    },
+    {
+      "epoch": 4.486298601568183,
+      "grad_norm": 0.9043852686882019,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 27750
+    },
+    {
+      "epoch": 4.487915285748929,
+      "grad_norm": 1.064600944519043,
+      "learning_rate": 0.0002,
+      "loss": 0.582,
+      "step": 27760
+    },
+    {
+      "epoch": 4.489531969929674,
+      "grad_norm": 0.7833460569381714,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 27770
+    },
+    {
+      "epoch": 4.49114865411042,
+      "grad_norm": 1.1073496341705322,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 27780
+    },
+    {
+      "epoch": 4.492765338291165,
+      "grad_norm": 1.0799397230148315,
+      "learning_rate": 0.0002,
+      "loss": 0.5577,
+      "step": 27790
+    },
+    {
+      "epoch": 4.49438202247191,
+      "grad_norm": 1.1062238216400146,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 27800
+    },
+    {
+      "epoch": 4.495998706652656,
+      "grad_norm": 1.0568242073059082,
+      "learning_rate": 0.0002,
+      "loss": 0.6126,
+      "step": 27810
+    },
+    {
+      "epoch": 4.497615390833401,
+      "grad_norm": 0.8861091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 27820
+    },
+    {
+      "epoch": 4.499232075014146,
+      "grad_norm": 1.2297543287277222,
+      "learning_rate": 0.0002,
+      "loss": 0.5858,
+      "step": 27830
+    },
+    {
+      "epoch": 4.500848759194891,
+      "grad_norm": 0.9600302577018738,
+      "learning_rate": 0.0002,
+      "loss": 0.5859,
+      "step": 27840
+    },
+    {
+      "epoch": 4.502465443375637,
+      "grad_norm": 1.057051181793213,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 27850
+    },
+    {
+      "epoch": 4.504082127556382,
+      "grad_norm": 0.9839690923690796,
+      "learning_rate": 0.0002,
+      "loss": 0.5788,
+      "step": 27860
+    },
+    {
+      "epoch": 4.505698811737127,
+      "grad_norm": 1.1479853391647339,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 27870
+    },
+    {
+      "epoch": 4.507315495917872,
+      "grad_norm": 1.0550768375396729,
+      "learning_rate": 0.0002,
+      "loss": 0.6039,
+      "step": 27880
+    },
+    {
+      "epoch": 4.5089321800986175,
+      "grad_norm": 0.898209273815155,
+      "learning_rate": 0.0002,
+      "loss": 0.563,
+      "step": 27890
+    },
+    {
+      "epoch": 4.510548864279363,
+      "grad_norm": 0.9460315108299255,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 27900
+    },
+    {
+      "epoch": 4.512165548460108,
+      "grad_norm": 0.9499884247779846,
+      "learning_rate": 0.0002,
+      "loss": 0.5702,
+      "step": 27910
+    },
+    {
+      "epoch": 4.513782232640853,
+      "grad_norm": 0.7801318764686584,
+      "learning_rate": 0.0002,
+      "loss": 0.5385,
+      "step": 27920
+    },
+    {
+      "epoch": 4.515398916821599,
+      "grad_norm": 0.9286966323852539,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 27930
+    },
+    {
+      "epoch": 4.517015601002345,
+      "grad_norm": 0.9539980292320251,
+      "learning_rate": 0.0002,
+      "loss": 0.5717,
+      "step": 27940
+    },
+    {
+      "epoch": 4.51863228518309,
+      "grad_norm": 1.1053401231765747,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 27950
+    },
+    {
+      "epoch": 4.520248969363835,
+      "grad_norm": 0.7535534501075745,
+      "learning_rate": 0.0002,
+      "loss": 0.6087,
+      "step": 27960
+    },
+    {
+      "epoch": 4.52186565354458,
+      "grad_norm": 1.076926589012146,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 27970
+    },
+    {
+      "epoch": 4.5234823377253255,
+      "grad_norm": 1.181935429573059,
+      "learning_rate": 0.0002,
+      "loss": 0.6028,
+      "step": 27980
+    },
+    {
+      "epoch": 4.525099021906071,
+      "grad_norm": 0.9293407201766968,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 27990
+    },
+    {
+      "epoch": 4.526715706086816,
+      "grad_norm": 0.8953009247779846,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 28000
+    },
+    {
+      "epoch": 4.528332390267561,
+      "grad_norm": 1.0850225687026978,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 28010
+    },
+    {
+      "epoch": 4.529949074448306,
+      "grad_norm": 0.9125663042068481,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 28020
+    },
+    {
+      "epoch": 4.531565758629052,
+      "grad_norm": 0.8745216727256775,
+      "learning_rate": 0.0002,
+      "loss": 0.5922,
+      "step": 28030
+    },
+    {
+      "epoch": 4.533182442809797,
+      "grad_norm": 1.0783463716506958,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 28040
+    },
+    {
+      "epoch": 4.534799126990542,
+      "grad_norm": 0.7513844966888428,
+      "learning_rate": 0.0002,
+      "loss": 0.5754,
+      "step": 28050
+    },
+    {
+      "epoch": 4.536415811171287,
+      "grad_norm": 1.0135776996612549,
+      "learning_rate": 0.0002,
+      "loss": 0.5608,
+      "step": 28060
+    },
+    {
+      "epoch": 4.538032495352033,
+      "grad_norm": 0.8886825442314148,
+      "learning_rate": 0.0002,
+      "loss": 0.5827,
+      "step": 28070
+    },
+    {
+      "epoch": 4.539649179532779,
+      "grad_norm": 0.8153995275497437,
+      "learning_rate": 0.0002,
+      "loss": 0.5605,
+      "step": 28080
+    },
+    {
+      "epoch": 4.541265863713524,
+      "grad_norm": 0.9853341579437256,
+      "learning_rate": 0.0002,
+      "loss": 0.6377,
+      "step": 28090
+    },
+    {
+      "epoch": 4.542882547894269,
+      "grad_norm": 0.9365800023078918,
+      "learning_rate": 0.0002,
+      "loss": 0.5957,
+      "step": 28100
+    },
+    {
+      "epoch": 4.544499232075014,
+      "grad_norm": 0.9765017628669739,
+      "learning_rate": 0.0002,
+      "loss": 0.5477,
+      "step": 28110
+    },
+    {
+      "epoch": 4.54611591625576,
+      "grad_norm": 0.9811279773712158,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 28120
+    },
+    {
+      "epoch": 4.547732600436505,
+      "grad_norm": 1.0387924909591675,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 28130
+    },
+    {
+      "epoch": 4.54934928461725,
+      "grad_norm": 1.0684878826141357,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 28140
+    },
+    {
+      "epoch": 4.550965968797995,
+      "grad_norm": 1.0000102519989014,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 28150
+    },
+    {
+      "epoch": 4.5525826529787405,
+      "grad_norm": 1.0717930793762207,
+      "learning_rate": 0.0002,
+      "loss": 0.5327,
+      "step": 28160
+    },
+    {
+      "epoch": 4.554199337159486,
+      "grad_norm": 0.990074634552002,
+      "learning_rate": 0.0002,
+      "loss": 0.5594,
+      "step": 28170
+    },
+    {
+      "epoch": 4.555816021340231,
+      "grad_norm": 0.8673754930496216,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 28180
+    },
+    {
+      "epoch": 4.557432705520976,
+      "grad_norm": 0.864247739315033,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 28190
+    },
+    {
+      "epoch": 4.5590493897017215,
+      "grad_norm": 0.8280200958251953,
+      "learning_rate": 0.0002,
+      "loss": 0.5516,
+      "step": 28200
+    },
+    {
+      "epoch": 4.560666073882467,
+      "grad_norm": 1.1312172412872314,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 28210
+    },
+    {
+      "epoch": 4.562282758063212,
+      "grad_norm": 0.9147403240203857,
+      "learning_rate": 0.0002,
+      "loss": 0.5776,
+      "step": 28220
+    },
+    {
+      "epoch": 4.563899442243958,
+      "grad_norm": 1.0321218967437744,
+      "learning_rate": 0.0002,
+      "loss": 0.5591,
+      "step": 28230
+    },
+    {
+      "epoch": 4.565516126424703,
+      "grad_norm": 1.168332815170288,
+      "learning_rate": 0.0002,
+      "loss": 0.5508,
+      "step": 28240
+    },
+    {
+      "epoch": 4.5671328106054485,
+      "grad_norm": 1.0067222118377686,
+      "learning_rate": 0.0002,
+      "loss": 0.5649,
+      "step": 28250
+    },
+    {
+      "epoch": 4.568749494786194,
+      "grad_norm": 1.0283393859863281,
+      "learning_rate": 0.0002,
+      "loss": 0.5853,
+      "step": 28260
+    },
+    {
+      "epoch": 4.570366178966939,
+      "grad_norm": 0.9912363886833191,
+      "learning_rate": 0.0002,
+      "loss": 0.5772,
+      "step": 28270
+    },
+    {
+      "epoch": 4.571982863147684,
+      "grad_norm": 1.108032464981079,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 28280
+    },
+    {
+      "epoch": 4.573599547328429,
+      "grad_norm": 0.8260078430175781,
+      "learning_rate": 0.0002,
+      "loss": 0.5529,
+      "step": 28290
+    },
+    {
+      "epoch": 4.575216231509175,
+      "grad_norm": 0.8946247100830078,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 28300
+    },
+    {
+      "epoch": 4.57683291568992,
+      "grad_norm": 0.8273587822914124,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 28310
+    },
+    {
+      "epoch": 4.578449599870665,
+      "grad_norm": 0.9040093421936035,
+      "learning_rate": 0.0002,
+      "loss": 0.6058,
+      "step": 28320
+    },
+    {
+      "epoch": 4.58006628405141,
+      "grad_norm": 0.8435290455818176,
+      "learning_rate": 0.0002,
+      "loss": 0.5521,
+      "step": 28330
+    },
+    {
+      "epoch": 4.581682968232156,
+      "grad_norm": 1.164088249206543,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 28340
+    },
+    {
+      "epoch": 4.583299652412901,
+      "grad_norm": 0.9861085414886475,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28350
+    },
+    {
+      "epoch": 4.584916336593646,
+      "grad_norm": 0.8892980813980103,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 28360
+    },
+    {
+      "epoch": 4.586533020774391,
+      "grad_norm": 1.240574836730957,
+      "learning_rate": 0.0002,
+      "loss": 0.598,
+      "step": 28370
+    },
+    {
+      "epoch": 4.588149704955137,
+      "grad_norm": 0.8669408559799194,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 28380
+    },
+    {
+      "epoch": 4.589766389135883,
+      "grad_norm": 0.9145985841751099,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28390
+    },
+    {
+      "epoch": 4.591383073316628,
+      "grad_norm": 0.8584614992141724,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 28400
+    },
+    {
+      "epoch": 4.592999757497373,
+      "grad_norm": 1.118829369544983,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 28410
+    },
+    {
+      "epoch": 4.594616441678118,
+      "grad_norm": 1.1411553621292114,
+      "learning_rate": 0.0002,
+      "loss": 0.5641,
+      "step": 28420
+    },
+    {
+      "epoch": 4.596233125858864,
+      "grad_norm": 0.9433278441429138,
+      "learning_rate": 0.0002,
+      "loss": 0.549,
+      "step": 28430
+    },
+    {
+      "epoch": 4.597849810039609,
+      "grad_norm": 0.816830039024353,
+      "learning_rate": 0.0002,
+      "loss": 0.5496,
+      "step": 28440
+    },
+    {
+      "epoch": 4.599466494220354,
+      "grad_norm": 1.2124968767166138,
+      "learning_rate": 0.0002,
+      "loss": 0.5543,
+      "step": 28450
+    },
+    {
+      "epoch": 4.601083178401099,
+      "grad_norm": 0.9658762216567993,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 28460
+    },
+    {
+      "epoch": 4.6026998625818445,
+      "grad_norm": 0.836100161075592,
+      "learning_rate": 0.0002,
+      "loss": 0.5902,
+      "step": 28470
+    },
+    {
+      "epoch": 4.60431654676259,
+      "grad_norm": 0.9989104270935059,
+      "learning_rate": 0.0002,
+      "loss": 0.5749,
+      "step": 28480
+    },
+    {
+      "epoch": 4.605933230943335,
+      "grad_norm": 1.1298956871032715,
+      "learning_rate": 0.0002,
+      "loss": 0.5616,
+      "step": 28490
+    },
+    {
+      "epoch": 4.60754991512408,
+      "grad_norm": 1.1731704473495483,
+      "learning_rate": 0.0002,
+      "loss": 0.5846,
+      "step": 28500
+    },
+    {
+      "epoch": 4.609166599304825,
+      "grad_norm": 0.9624714255332947,
+      "learning_rate": 0.0002,
+      "loss": 0.5816,
+      "step": 28510
+    },
+    {
+      "epoch": 4.610783283485571,
+      "grad_norm": 1.364073634147644,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 28520
+    },
+    {
+      "epoch": 4.612399967666317,
+      "grad_norm": 1.1827356815338135,
+      "learning_rate": 0.0002,
+      "loss": 0.6237,
+      "step": 28530
+    },
+    {
+      "epoch": 4.614016651847062,
+      "grad_norm": 0.6651531457901001,
+      "learning_rate": 0.0002,
+      "loss": 0.5643,
+      "step": 28540
+    },
+    {
+      "epoch": 4.615633336027807,
+      "grad_norm": 1.1640995740890503,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 28550
+    },
+    {
+      "epoch": 4.6172500202085525,
+      "grad_norm": 1.028918743133545,
+      "learning_rate": 0.0002,
+      "loss": 0.5995,
+      "step": 28560
+    },
+    {
+      "epoch": 4.618866704389298,
+      "grad_norm": 0.8252120614051819,
+      "learning_rate": 0.0002,
+      "loss": 0.5607,
+      "step": 28570
+    },
+    {
+      "epoch": 4.620483388570043,
+      "grad_norm": 1.3536735773086548,
+      "learning_rate": 0.0002,
+      "loss": 0.5769,
+      "step": 28580
+    },
+    {
+      "epoch": 4.622100072750788,
+      "grad_norm": 1.2146915197372437,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 28590
+    },
+    {
+      "epoch": 4.623716756931533,
+      "grad_norm": 1.0122549533843994,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 28600
+    },
+    {
+      "epoch": 4.625333441112279,
+      "grad_norm": 0.9977872967720032,
+      "learning_rate": 0.0002,
+      "loss": 0.6072,
+      "step": 28610
+    },
+    {
+      "epoch": 4.626950125293024,
+      "grad_norm": 1.0159751176834106,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 28620
+    },
+    {
+      "epoch": 4.628566809473769,
+      "grad_norm": 1.0028325319290161,
+      "learning_rate": 0.0002,
+      "loss": 0.5935,
+      "step": 28630
+    },
+    {
+      "epoch": 4.630183493654514,
+      "grad_norm": 0.901638388633728,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 28640
+    },
+    {
+      "epoch": 4.6318001778352595,
+      "grad_norm": 0.9450507164001465,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 28650
+    },
+    {
+      "epoch": 4.633416862016006,
+      "grad_norm": 0.9987545013427734,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 28660
+    },
+    {
+      "epoch": 4.63503354619675,
+      "grad_norm": 0.9574332237243652,
+      "learning_rate": 0.0002,
+      "loss": 0.5863,
+      "step": 28670
+    },
+    {
+      "epoch": 4.636650230377496,
+      "grad_norm": 1.2215653657913208,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 28680
+    },
+    {
+      "epoch": 4.638266914558241,
+      "grad_norm": 0.9798858761787415,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 28690
+    },
+    {
+      "epoch": 4.639883598738987,
+      "grad_norm": 1.0648466348648071,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 28700
+    },
+    {
+      "epoch": 4.641500282919732,
+      "grad_norm": 1.0606504678726196,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 28710
+    },
+    {
+      "epoch": 4.643116967100477,
+      "grad_norm": 1.0892442464828491,
+      "learning_rate": 0.0002,
+      "loss": 0.5801,
+      "step": 28720
+    },
+    {
+      "epoch": 4.644733651281222,
+      "grad_norm": 0.914391040802002,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 28730
+    },
+    {
+      "epoch": 4.6463503354619675,
+      "grad_norm": 0.9782370328903198,
+      "learning_rate": 0.0002,
+      "loss": 0.5439,
+      "step": 28740
+    },
+    {
+      "epoch": 4.647967019642713,
+      "grad_norm": 1.0344339609146118,
+      "learning_rate": 0.0002,
+      "loss": 0.6035,
+      "step": 28750
+    },
+    {
+      "epoch": 4.649583703823458,
+      "grad_norm": 1.0513931512832642,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 28760
+    },
+    {
+      "epoch": 4.651200388004203,
+      "grad_norm": 0.9711475968360901,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 28770
+    },
+    {
+      "epoch": 4.652817072184948,
+      "grad_norm": 0.977519690990448,
+      "learning_rate": 0.0002,
+      "loss": 0.5472,
+      "step": 28780
+    },
+    {
+      "epoch": 4.654433756365694,
+      "grad_norm": 0.9150224924087524,
+      "learning_rate": 0.0002,
+      "loss": 0.5826,
+      "step": 28790
+    },
+    {
+      "epoch": 4.656050440546439,
+      "grad_norm": 1.0973542928695679,
+      "learning_rate": 0.0002,
+      "loss": 0.5382,
+      "step": 28800
+    },
+    {
+      "epoch": 4.657667124727185,
+      "grad_norm": 0.944877564907074,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 28810
+    },
+    {
+      "epoch": 4.659283808907929,
+      "grad_norm": 0.9508748650550842,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 28820
+    },
+    {
+      "epoch": 4.6609004930886755,
+      "grad_norm": 0.9681721329689026,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 28830
+    },
+    {
+      "epoch": 4.662517177269421,
+      "grad_norm": 1.0214351415634155,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 28840
+    },
+    {
+      "epoch": 4.664133861450166,
+      "grad_norm": 0.9748611450195312,
+      "learning_rate": 0.0002,
+      "loss": 0.6031,
+      "step": 28850
+    },
+    {
+      "epoch": 4.665750545630911,
+      "grad_norm": 0.8484147191047668,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 28860
+    },
+    {
+      "epoch": 4.667367229811656,
+      "grad_norm": 1.1252986192703247,
+      "learning_rate": 0.0002,
+      "loss": 0.5699,
+      "step": 28870
+    },
+    {
+      "epoch": 4.668983913992402,
+      "grad_norm": 0.8706206679344177,
+      "learning_rate": 0.0002,
+      "loss": 0.5724,
+      "step": 28880
+    },
+    {
+      "epoch": 4.670600598173147,
+      "grad_norm": 1.1432424783706665,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 28890
+    },
+    {
+      "epoch": 4.672217282353892,
+      "grad_norm": 1.017029047012329,
+      "learning_rate": 0.0002,
+      "loss": 0.5675,
+      "step": 28900
+    },
+    {
+      "epoch": 4.673833966534637,
+      "grad_norm": 1.085597038269043,
+      "learning_rate": 0.0002,
+      "loss": 0.5831,
+      "step": 28910
+    },
+    {
+      "epoch": 4.675450650715383,
+      "grad_norm": 0.9275796413421631,
+      "learning_rate": 0.0002,
+      "loss": 0.5678,
+      "step": 28920
+    },
+    {
+      "epoch": 4.677067334896128,
+      "grad_norm": 0.9518964886665344,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28930
+    },
+    {
+      "epoch": 4.678684019076873,
+      "grad_norm": 1.0352122783660889,
+      "learning_rate": 0.0002,
+      "loss": 0.6232,
+      "step": 28940
+    },
+    {
+      "epoch": 4.680300703257618,
+      "grad_norm": 1.090124249458313,
+      "learning_rate": 0.0002,
+      "loss": 0.5786,
+      "step": 28950
+    },
+    {
+      "epoch": 4.681917387438364,
+      "grad_norm": 0.8799563050270081,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 28960
+    },
+    {
+      "epoch": 4.683534071619109,
+      "grad_norm": 1.0929821729660034,
+      "learning_rate": 0.0002,
+      "loss": 0.5787,
+      "step": 28970
+    },
+    {
+      "epoch": 4.685150755799855,
+      "grad_norm": 0.903727650642395,
+      "learning_rate": 0.0002,
+      "loss": 0.6134,
+      "step": 28980
+    },
+    {
+      "epoch": 4.6867674399806,
+      "grad_norm": 0.9752424955368042,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 28990
+    },
+    {
+      "epoch": 4.688384124161345,
+      "grad_norm": 0.9351571202278137,
+      "learning_rate": 0.0002,
+      "loss": 0.5762,
+      "step": 29000
+    },
+    {
+      "epoch": 4.6900008083420905,
+      "grad_norm": 0.923877477645874,
+      "learning_rate": 0.0002,
+      "loss": 0.5811,
+      "step": 29010
+    },
+    {
+      "epoch": 4.691617492522836,
+      "grad_norm": 1.045389175415039,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 29020
+    },
+    {
+      "epoch": 4.693234176703581,
+      "grad_norm": 1.0200831890106201,
+      "learning_rate": 0.0002,
+      "loss": 0.584,
+      "step": 29030
+    },
+    {
+      "epoch": 4.694850860884326,
+      "grad_norm": 1.1499706506729126,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 29040
+    },
+    {
+      "epoch": 4.6964675450650715,
+      "grad_norm": 0.860118567943573,
+      "learning_rate": 0.0002,
+      "loss": 0.5745,
+      "step": 29050
+    },
+    {
+      "epoch": 4.698084229245817,
+      "grad_norm": 0.9774864315986633,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 29060
+    },
+    {
+      "epoch": 4.699700913426562,
+      "grad_norm": 1.0323210954666138,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 29070
+    },
+    {
+      "epoch": 4.701317597607307,
+      "grad_norm": 0.8492481112480164,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 29080
+    },
+    {
+      "epoch": 4.702934281788052,
+      "grad_norm": 1.131951093673706,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 29090
+    },
+    {
+      "epoch": 4.704550965968798,
+      "grad_norm": 0.8763113021850586,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 29100
+    },
+    {
+      "epoch": 4.706167650149544,
+      "grad_norm": 1.045028805732727,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 29110
+    },
+    {
+      "epoch": 4.707784334330288,
+      "grad_norm": 0.9961401224136353,
+      "learning_rate": 0.0002,
+      "loss": 0.5548,
+      "step": 29120
+    },
+    {
+      "epoch": 4.709401018511034,
+      "grad_norm": 0.9282503724098206,
+      "learning_rate": 0.0002,
+      "loss": 0.559,
+      "step": 29130
+    },
+    {
+      "epoch": 4.711017702691779,
+      "grad_norm": 1.1418932676315308,
+      "learning_rate": 0.0002,
+      "loss": 0.5744,
+      "step": 29140
+    },
+    {
+      "epoch": 4.712634386872525,
+      "grad_norm": 0.9950099587440491,
+      "learning_rate": 0.0002,
+      "loss": 0.5394,
+      "step": 29150
+    },
+    {
+      "epoch": 4.71425107105327,
+      "grad_norm": 0.8304893374443054,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 29160
+    },
+    {
+      "epoch": 4.715867755234015,
+      "grad_norm": 1.115626335144043,
+      "learning_rate": 0.0002,
+      "loss": 0.6074,
+      "step": 29170
+    },
+    {
+      "epoch": 4.71748443941476,
+      "grad_norm": 1.079818606376648,
+      "learning_rate": 0.0002,
+      "loss": 0.6265,
+      "step": 29180
+    },
+    {
+      "epoch": 4.719101123595506,
+      "grad_norm": 1.1929082870483398,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 29190
+    },
+    {
+      "epoch": 4.720717807776251,
+      "grad_norm": 0.9621080756187439,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 29200
+    },
+    {
+      "epoch": 4.722334491956996,
+      "grad_norm": 0.8549222350120544,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 29210
+    },
+    {
+      "epoch": 4.723951176137741,
+      "grad_norm": 0.9341941475868225,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 29220
+    },
+    {
+      "epoch": 4.7255678603184865,
+      "grad_norm": 1.075406789779663,
+      "learning_rate": 0.0002,
+      "loss": 0.5436,
+      "step": 29230
+    },
+    {
+      "epoch": 4.727184544499232,
+      "grad_norm": 1.0859880447387695,
+      "learning_rate": 0.0002,
+      "loss": 0.576,
+      "step": 29240
+    },
+    {
+      "epoch": 4.728801228679977,
+      "grad_norm": 0.8475605249404907,
+      "learning_rate": 0.0002,
+      "loss": 0.5525,
+      "step": 29250
+    },
+    {
+      "epoch": 4.730417912860723,
+      "grad_norm": 0.9331845641136169,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 29260
+    },
+    {
+      "epoch": 4.7320345970414674,
+      "grad_norm": 0.9279314279556274,
+      "learning_rate": 0.0002,
+      "loss": 0.5901,
+      "step": 29270
+    },
+    {
+      "epoch": 4.733651281222214,
+      "grad_norm": 0.7803558707237244,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 29280
+    },
+    {
+      "epoch": 4.735267965402959,
+      "grad_norm": 1.0159329175949097,
+      "learning_rate": 0.0002,
+      "loss": 0.5968,
+      "step": 29290
+    },
+    {
+      "epoch": 4.736884649583704,
+      "grad_norm": 0.9448670744895935,
+      "learning_rate": 0.0002,
+      "loss": 0.5333,
+      "step": 29300
+    },
+    {
+      "epoch": 4.738501333764449,
+      "grad_norm": 1.0732197761535645,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 29310
+    },
+    {
+      "epoch": 4.7401180179451945,
+      "grad_norm": 0.901830792427063,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 29320
+    },
+    {
+      "epoch": 4.74173470212594,
+      "grad_norm": 0.9141789674758911,
+      "learning_rate": 0.0002,
+      "loss": 0.6105,
+      "step": 29330
+    },
+    {
+      "epoch": 4.743351386306685,
+      "grad_norm": 0.9733418226242065,
+      "learning_rate": 0.0002,
+      "loss": 0.5481,
+      "step": 29340
+    },
+    {
+      "epoch": 4.74496807048743,
+      "grad_norm": 0.909810483455658,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 29350
+    },
+    {
+      "epoch": 4.746584754668175,
+      "grad_norm": 0.909541666507721,
+      "learning_rate": 0.0002,
+      "loss": 0.5911,
+      "step": 29360
+    },
+    {
+      "epoch": 4.748201438848921,
+      "grad_norm": 0.9383015632629395,
+      "learning_rate": 0.0002,
+      "loss": 0.5579,
+      "step": 29370
+    },
+    {
+      "epoch": 4.749818123029666,
+      "grad_norm": 0.9275668263435364,
+      "learning_rate": 0.0002,
+      "loss": 0.5529,
+      "step": 29380
+    },
+    {
+      "epoch": 4.751434807210411,
+      "grad_norm": 1.1146225929260254,
+      "learning_rate": 0.0002,
+      "loss": 0.5623,
+      "step": 29390
+    },
+    {
+      "epoch": 4.753051491391156,
+      "grad_norm": 1.0062453746795654,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 29400
+    },
+    {
+      "epoch": 4.7546681755719025,
+      "grad_norm": 0.9451895952224731,
+      "learning_rate": 0.0002,
+      "loss": 0.5872,
+      "step": 29410
+    },
+    {
+      "epoch": 4.756284859752648,
+      "grad_norm": 0.870457649230957,
+      "learning_rate": 0.0002,
+      "loss": 0.5767,
+      "step": 29420
+    },
+    {
+      "epoch": 4.757901543933393,
+      "grad_norm": 1.0411282777786255,
+      "learning_rate": 0.0002,
+      "loss": 0.57,
+      "step": 29430
+    },
+    {
+      "epoch": 4.759518228114138,
+      "grad_norm": 1.1648986339569092,
+      "learning_rate": 0.0002,
+      "loss": 0.5688,
+      "step": 29440
+    },
+    {
+      "epoch": 4.761134912294883,
+      "grad_norm": 0.8999572992324829,
+      "learning_rate": 0.0002,
+      "loss": 0.5432,
+      "step": 29450
+    },
+    {
+      "epoch": 4.762751596475629,
+      "grad_norm": 0.9863559007644653,
+      "learning_rate": 0.0002,
+      "loss": 0.5667,
+      "step": 29460
+    },
+    {
+      "epoch": 4.764368280656374,
+      "grad_norm": 0.9676542282104492,
+      "learning_rate": 0.0002,
+      "loss": 0.5779,
+      "step": 29470
+    },
+    {
+      "epoch": 4.765984964837119,
+      "grad_norm": 1.004775047302246,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 29480
+    },
+    {
+      "epoch": 4.767601649017864,
+      "grad_norm": 1.0937515497207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6044,
+      "step": 29490
+    },
+    {
+      "epoch": 4.7692183331986095,
+      "grad_norm": 0.9551598429679871,
+      "learning_rate": 0.0002,
+      "loss": 0.5433,
+      "step": 29500
+    },
+    {
+      "epoch": 4.770835017379355,
+      "grad_norm": 1.0757228136062622,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 29510
+    },
+    {
+      "epoch": 4.7724517015601,
+      "grad_norm": 1.0588841438293457,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 29520
+    },
+    {
+      "epoch": 4.774068385740845,
+      "grad_norm": 1.0744032859802246,
+      "learning_rate": 0.0002,
+      "loss": 0.5814,
+      "step": 29530
+    },
+    {
+      "epoch": 4.7756850699215905,
+      "grad_norm": 1.0066277980804443,
+      "learning_rate": 0.0002,
+      "loss": 0.5681,
+      "step": 29540
+    },
+    {
+      "epoch": 4.777301754102336,
+      "grad_norm": 1.082319736480713,
+      "learning_rate": 0.0002,
+      "loss": 0.545,
+      "step": 29550
+    },
+    {
+      "epoch": 4.778918438283082,
+      "grad_norm": 0.8252472877502441,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 29560
+    },
+    {
+      "epoch": 4.780535122463827,
+      "grad_norm": 0.9855340123176575,
+      "learning_rate": 0.0002,
+      "loss": 0.5666,
+      "step": 29570
+    },
+    {
+      "epoch": 4.782151806644572,
+      "grad_norm": 0.9991421699523926,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 29580
+    },
+    {
+      "epoch": 4.7837684908253175,
+      "grad_norm": 1.316841959953308,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 29590
+    },
+    {
+      "epoch": 4.785385175006063,
+      "grad_norm": 1.1513035297393799,
+      "learning_rate": 0.0002,
+      "loss": 0.6102,
+      "step": 29600
+    },
+    {
+      "epoch": 4.787001859186808,
+      "grad_norm": 0.9767683744430542,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 29610
+    },
+    {
+      "epoch": 4.788618543367553,
+      "grad_norm": 0.9786278605461121,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 29620
+    },
+    {
+      "epoch": 4.7902352275482984,
+      "grad_norm": 0.8004973530769348,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 29630
+    },
+    {
+      "epoch": 4.791851911729044,
+      "grad_norm": 1.0997767448425293,
+      "learning_rate": 0.0002,
+      "loss": 0.5932,
+      "step": 29640
+    },
+    {
+      "epoch": 4.793468595909789,
+      "grad_norm": 0.9752856492996216,
+      "learning_rate": 0.0002,
+      "loss": 0.5655,
+      "step": 29650
+    },
+    {
+      "epoch": 4.795085280090534,
+      "grad_norm": 1.0518392324447632,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 29660
+    },
+    {
+      "epoch": 4.796701964271279,
+      "grad_norm": 1.1050055027008057,
+      "learning_rate": 0.0002,
+      "loss": 0.6042,
+      "step": 29670
+    },
+    {
+      "epoch": 4.798318648452025,
+      "grad_norm": 0.9933857917785645,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 29680
+    },
+    {
+      "epoch": 4.79993533263277,
+      "grad_norm": 1.2804018259048462,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 29690
+    },
+    {
+      "epoch": 4.801552016813515,
+      "grad_norm": 1.0133371353149414,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 29700
+    },
+    {
+      "epoch": 4.803168700994261,
+      "grad_norm": 1.080350637435913,
+      "learning_rate": 0.0002,
+      "loss": 0.5662,
+      "step": 29710
+    },
+    {
+      "epoch": 4.804785385175006,
+      "grad_norm": 0.9986529350280762,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 29720
+    },
+    {
+      "epoch": 4.806402069355752,
+      "grad_norm": 0.975665807723999,
+      "learning_rate": 0.0002,
+      "loss": 0.5894,
+      "step": 29730
+    },
+    {
+      "epoch": 4.808018753536497,
+      "grad_norm": 0.8458138704299927,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 29740
+    },
+    {
+      "epoch": 4.809635437717242,
+      "grad_norm": 0.99330073595047,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 29750
+    },
+    {
+      "epoch": 4.811252121897987,
+      "grad_norm": 0.898274302482605,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 29760
+    },
+    {
+      "epoch": 4.812868806078733,
+      "grad_norm": 1.0504480600357056,
+      "learning_rate": 0.0002,
+      "loss": 0.5842,
+      "step": 29770
+    },
+    {
+      "epoch": 4.814485490259478,
+      "grad_norm": 0.937919020652771,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 29780
+    },
+    {
+      "epoch": 4.816102174440223,
+      "grad_norm": 0.9593307971954346,
+      "learning_rate": 0.0002,
+      "loss": 0.5885,
+      "step": 29790
+    },
+    {
+      "epoch": 4.817718858620968,
+      "grad_norm": 0.9431198835372925,
+      "learning_rate": 0.0002,
+      "loss": 0.578,
+      "step": 29800
+    },
+    {
+      "epoch": 4.8193355428017135,
+      "grad_norm": 1.2729957103729248,
+      "learning_rate": 0.0002,
+      "loss": 0.5739,
+      "step": 29810
+    },
+    {
+      "epoch": 4.820952226982459,
+      "grad_norm": 0.8876838684082031,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 29820
+    },
+    {
+      "epoch": 4.822568911163204,
+      "grad_norm": 1.0185000896453857,
+      "learning_rate": 0.0002,
+      "loss": 0.5583,
+      "step": 29830
+    },
+    {
+      "epoch": 4.824185595343949,
+      "grad_norm": 1.064276099205017,
+      "learning_rate": 0.0002,
+      "loss": 0.5686,
+      "step": 29840
+    },
+    {
+      "epoch": 4.825802279524694,
+      "grad_norm": 0.9774803519248962,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 29850
+    },
+    {
+      "epoch": 4.8274189637054405,
+      "grad_norm": 1.131646990776062,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 29860
+    },
+    {
+      "epoch": 4.829035647886186,
+      "grad_norm": 1.081455945968628,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 29870
+    },
+    {
+      "epoch": 4.830652332066931,
+      "grad_norm": 0.990538477897644,
+      "learning_rate": 0.0002,
+      "loss": 0.5793,
+      "step": 29880
+    },
+    {
+      "epoch": 4.832269016247676,
+      "grad_norm": 0.9750600457191467,
+      "learning_rate": 0.0002,
+      "loss": 0.5833,
+      "step": 29890
+    },
+    {
+      "epoch": 4.8338857004284215,
+      "grad_norm": 1.0600621700286865,
+      "learning_rate": 0.0002,
+      "loss": 0.619,
+      "step": 29900
+    },
+    {
+      "epoch": 4.835502384609167,
+      "grad_norm": 0.9237320423126221,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 29910
+    },
+    {
+      "epoch": 4.837119068789912,
+      "grad_norm": 0.9739177227020264,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 29920
+    },
+    {
+      "epoch": 4.838735752970657,
+      "grad_norm": 1.128677248954773,
+      "learning_rate": 0.0002,
+      "loss": 0.587,
+      "step": 29930
+    },
+    {
+      "epoch": 4.840352437151402,
+      "grad_norm": 1.042604923248291,
+      "learning_rate": 0.0002,
+      "loss": 0.564,
+      "step": 29940
+    },
+    {
+      "epoch": 4.841969121332148,
+      "grad_norm": 0.849758505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.5885,
+      "step": 29950
+    },
+    {
+      "epoch": 4.843585805512893,
+      "grad_norm": 1.2809888124465942,
+      "learning_rate": 0.0002,
+      "loss": 0.5952,
+      "step": 29960
+    },
+    {
+      "epoch": 4.845202489693638,
+      "grad_norm": 1.0177865028381348,
+      "learning_rate": 0.0002,
+      "loss": 0.5703,
+      "step": 29970
+    },
+    {
+      "epoch": 4.846819173874383,
+      "grad_norm": 1.0026639699935913,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 29980
+    },
+    {
+      "epoch": 4.8484358580551286,
+      "grad_norm": 0.9679505228996277,
+      "learning_rate": 0.0002,
+      "loss": 0.5897,
+      "step": 29990
+    },
+    {
+      "epoch": 4.850052542235874,
+      "grad_norm": 0.8939532041549683,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 30000
+    },
+    {
+      "epoch": 4.85166922641662,
+      "grad_norm": 0.9957457780838013,
+      "learning_rate": 0.0002,
+      "loss": 0.5852,
+      "step": 30010
+    },
+    {
+      "epoch": 4.853285910597365,
+      "grad_norm": 1.1646790504455566,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 30020
+    },
+    {
+      "epoch": 4.85490259477811,
+      "grad_norm": 0.8804680705070496,
+      "learning_rate": 0.0002,
+      "loss": 0.5711,
+      "step": 30030
+    },
+    {
+      "epoch": 4.856519278958856,
+      "grad_norm": 1.161970853805542,
+      "learning_rate": 0.0002,
+      "loss": 0.5397,
+      "step": 30040
+    },
+    {
+      "epoch": 4.858135963139601,
+      "grad_norm": 0.9081037640571594,
+      "learning_rate": 0.0002,
+      "loss": 0.5552,
+      "step": 30050
+    },
+    {
+      "epoch": 4.859752647320346,
+      "grad_norm": 0.9402848482131958,
+      "learning_rate": 0.0002,
+      "loss": 0.6024,
+      "step": 30060
+    },
+    {
+      "epoch": 4.861369331501091,
+      "grad_norm": 0.9023865461349487,
+      "learning_rate": 0.0002,
+      "loss": 0.6256,
+      "step": 30070
+    },
+    {
+      "epoch": 4.8629860156818365,
+      "grad_norm": 1.0173414945602417,
+      "learning_rate": 0.0002,
+      "loss": 0.5926,
+      "step": 30080
+    },
+    {
+      "epoch": 4.864602699862582,
+      "grad_norm": 1.084402322769165,
+      "learning_rate": 0.0002,
+      "loss": 0.6274,
+      "step": 30090
+    },
+    {
+      "epoch": 4.866219384043327,
+      "grad_norm": 0.9577937126159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6311,
+      "step": 30100
+    },
+    {
+      "epoch": 4.867836068224072,
+      "grad_norm": 0.9807606935501099,
+      "learning_rate": 0.0002,
+      "loss": 0.5724,
+      "step": 30110
+    },
+    {
+      "epoch": 4.8694527524048175,
+      "grad_norm": 0.978784441947937,
+      "learning_rate": 0.0002,
+      "loss": 0.5786,
+      "step": 30120
+    },
+    {
+      "epoch": 4.871069436585563,
+      "grad_norm": 0.9762914776802063,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 30130
+    },
+    {
+      "epoch": 4.872686120766308,
+      "grad_norm": 0.9404871463775635,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 30140
+    },
+    {
+      "epoch": 4.874302804947053,
+      "grad_norm": 1.0069509744644165,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 30150
+    },
+    {
+      "epoch": 4.875919489127799,
+      "grad_norm": 1.1770923137664795,
+      "learning_rate": 0.0002,
+      "loss": 0.6225,
+      "step": 30160
+    },
+    {
+      "epoch": 4.8775361733085445,
+      "grad_norm": 1.021210789680481,
+      "learning_rate": 0.0002,
+      "loss": 0.5657,
+      "step": 30170
+    },
+    {
+      "epoch": 4.87915285748929,
+      "grad_norm": 0.8512648940086365,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 30180
+    },
+    {
+      "epoch": 4.880769541670035,
+      "grad_norm": 0.9345870018005371,
+      "learning_rate": 0.0002,
+      "loss": 0.5519,
+      "step": 30190
+    },
+    {
+      "epoch": 4.88238622585078,
+      "grad_norm": 1.0224418640136719,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 30200
+    },
+    {
+      "epoch": 4.884002910031525,
+      "grad_norm": 1.0316044092178345,
+      "learning_rate": 0.0002,
+      "loss": 0.5807,
+      "step": 30210
+    },
+    {
+      "epoch": 4.885619594212271,
+      "grad_norm": 1.102437973022461,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 30220
+    },
+    {
+      "epoch": 4.887236278393016,
+      "grad_norm": 1.0220023393630981,
+      "learning_rate": 0.0002,
+      "loss": 0.586,
+      "step": 30230
+    },
+    {
+      "epoch": 4.888852962573761,
+      "grad_norm": 1.0934523344039917,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 30240
+    },
+    {
+      "epoch": 4.890469646754506,
+      "grad_norm": 1.264630913734436,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 30250
+    },
+    {
+      "epoch": 4.892086330935252,
+      "grad_norm": 1.0999879837036133,
+      "learning_rate": 0.0002,
+      "loss": 0.5712,
+      "step": 30260
+    },
+    {
+      "epoch": 4.893703015115997,
+      "grad_norm": 0.9124550223350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 30270
+    },
+    {
+      "epoch": 4.895319699296742,
+      "grad_norm": 0.9853624105453491,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 30280
+    },
+    {
+      "epoch": 4.896936383477488,
+      "grad_norm": 1.0589802265167236,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 30290
+    },
+    {
+      "epoch": 4.8985530676582325,
+      "grad_norm": 0.8487226366996765,
+      "learning_rate": 0.0002,
+      "loss": 0.6129,
+      "step": 30300
+    },
+    {
+      "epoch": 4.900169751838979,
+      "grad_norm": 1.0212191343307495,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 30310
+    },
+    {
+      "epoch": 4.901786436019724,
+      "grad_norm": 1.0187491178512573,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 30320
+    },
+    {
+      "epoch": 4.903403120200469,
+      "grad_norm": 1.0013091564178467,
+      "learning_rate": 0.0002,
+      "loss": 0.5809,
+      "step": 30330
+    },
+    {
+      "epoch": 4.905019804381214,
+      "grad_norm": 1.0017542839050293,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 30340
+    },
+    {
+      "epoch": 4.9066364885619596,
+      "grad_norm": 0.9665151238441467,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 30350
+    },
+    {
+      "epoch": 4.908253172742705,
+      "grad_norm": 0.8774822950363159,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 30360
+    },
+    {
+      "epoch": 4.90986985692345,
+      "grad_norm": 0.9449850916862488,
+      "learning_rate": 0.0002,
+      "loss": 0.5771,
+      "step": 30370
+    },
+    {
+      "epoch": 4.911486541104195,
+      "grad_norm": 0.7368341088294983,
+      "learning_rate": 0.0002,
+      "loss": 0.58,
+      "step": 30380
+    },
+    {
+      "epoch": 4.9131032252849405,
+      "grad_norm": 0.9669167995452881,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 30390
+    },
+    {
+      "epoch": 4.914719909465686,
+      "grad_norm": 1.1227794885635376,
+      "learning_rate": 0.0002,
+      "loss": 0.6202,
+      "step": 30400
+    },
+    {
+      "epoch": 4.916336593646431,
+      "grad_norm": 0.9884361028671265,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 30410
+    },
+    {
+      "epoch": 4.917953277827176,
+      "grad_norm": 0.9949551224708557,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 30420
+    },
+    {
+      "epoch": 4.919569962007921,
+      "grad_norm": 0.9491621851921082,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 30430
+    },
+    {
+      "epoch": 4.9211866461886675,
+      "grad_norm": 0.78848797082901,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 30440
+    },
+    {
+      "epoch": 4.922803330369412,
+      "grad_norm": 1.0693835020065308,
+      "learning_rate": 0.0002,
+      "loss": 0.5561,
+      "step": 30450
+    },
+    {
+      "epoch": 4.924420014550158,
+      "grad_norm": 0.9573729634284973,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 30460
+    },
+    {
+      "epoch": 4.926036698730903,
+      "grad_norm": 0.9975152611732483,
+      "learning_rate": 0.0002,
+      "loss": 0.6084,
+      "step": 30470
+    },
+    {
+      "epoch": 4.9276533829116484,
+      "grad_norm": 0.8695693016052246,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 30480
+    },
+    {
+      "epoch": 4.929270067092394,
+      "grad_norm": 1.145394206047058,
+      "learning_rate": 0.0002,
+      "loss": 0.6144,
+      "step": 30490
+    },
+    {
+      "epoch": 4.930886751273139,
+      "grad_norm": 0.7668989896774292,
+      "learning_rate": 0.0002,
+      "loss": 0.5736,
+      "step": 30500
+    },
+    {
+      "epoch": 4.932503435453884,
+      "grad_norm": 0.9630151391029358,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 30510
+    },
+    {
+      "epoch": 4.934120119634629,
+      "grad_norm": 0.940705418586731,
+      "learning_rate": 0.0002,
+      "loss": 0.6461,
+      "step": 30520
+    },
+    {
+      "epoch": 4.935736803815375,
+      "grad_norm": 1.3243348598480225,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 30530
+    },
+    {
+      "epoch": 4.93735348799612,
+      "grad_norm": 1.004347801208496,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 30540
+    },
+    {
+      "epoch": 4.938970172176865,
+      "grad_norm": 0.8711541295051575,
+      "learning_rate": 0.0002,
+      "loss": 0.583,
+      "step": 30550
+    },
+    {
+      "epoch": 4.94058685635761,
+      "grad_norm": 0.8980631828308105,
+      "learning_rate": 0.0002,
+      "loss": 0.599,
+      "step": 30560
+    },
+    {
+      "epoch": 4.9422035405383555,
+      "grad_norm": 0.8388893604278564,
+      "learning_rate": 0.0002,
+      "loss": 0.6024,
+      "step": 30570
+    },
+    {
+      "epoch": 4.943820224719101,
+      "grad_norm": 1.0991183519363403,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 30580
+    },
+    {
+      "epoch": 4.945436908899847,
+      "grad_norm": 0.9731075763702393,
+      "learning_rate": 0.0002,
+      "loss": 0.5906,
+      "step": 30590
+    },
+    {
+      "epoch": 4.947053593080591,
+      "grad_norm": 1.3904452323913574,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 30600
+    },
+    {
+      "epoch": 4.948670277261337,
+      "grad_norm": 1.2489882707595825,
+      "learning_rate": 0.0002,
+      "loss": 0.5952,
+      "step": 30610
+    },
+    {
+      "epoch": 4.950286961442083,
+      "grad_norm": 1.240072250366211,
+      "learning_rate": 0.0002,
+      "loss": 0.5887,
+      "step": 30620
+    },
+    {
+      "epoch": 4.951903645622828,
+      "grad_norm": 0.9191411733627319,
+      "learning_rate": 0.0002,
+      "loss": 0.5762,
+      "step": 30630
+    },
+    {
+      "epoch": 4.953520329803573,
+      "grad_norm": 0.8888895511627197,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 30640
+    },
+    {
+      "epoch": 4.955137013984318,
+      "grad_norm": 0.9001450538635254,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 30650
+    },
+    {
+      "epoch": 4.9567536981650635,
+      "grad_norm": 1.053971767425537,
+      "learning_rate": 0.0002,
+      "loss": 0.6047,
+      "step": 30660
+    },
+    {
+      "epoch": 4.958370382345809,
+      "grad_norm": 1.2224042415618896,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 30670
+    },
+    {
+      "epoch": 4.959987066526554,
+      "grad_norm": 0.8855111598968506,
+      "learning_rate": 0.0002,
+      "loss": 0.6211,
+      "step": 30680
+    },
+    {
+      "epoch": 4.961603750707299,
+      "grad_norm": 0.9489575624465942,
+      "learning_rate": 0.0002,
+      "loss": 0.5764,
+      "step": 30690
+    },
+    {
+      "epoch": 4.963220434888044,
+      "grad_norm": 0.9635404944419861,
+      "learning_rate": 0.0002,
+      "loss": 0.5371,
+      "step": 30700
+    },
+    {
+      "epoch": 4.96483711906879,
+      "grad_norm": 1.1784121990203857,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 30710
+    },
+    {
+      "epoch": 4.966453803249535,
+      "grad_norm": 1.0059462785720825,
+      "learning_rate": 0.0002,
+      "loss": 0.5803,
+      "step": 30720
+    },
+    {
+      "epoch": 4.96807048743028,
+      "grad_norm": 0.9479738473892212,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 30730
+    },
+    {
+      "epoch": 4.969687171611026,
+      "grad_norm": 1.0624593496322632,
+      "learning_rate": 0.0002,
+      "loss": 0.584,
+      "step": 30740
+    },
+    {
+      "epoch": 4.971303855791771,
+      "grad_norm": 1.1429259777069092,
+      "learning_rate": 0.0002,
+      "loss": 0.6202,
+      "step": 30750
+    },
+    {
+      "epoch": 4.972920539972517,
+      "grad_norm": 0.9102491140365601,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 30760
+    },
+    {
+      "epoch": 4.974537224153262,
+      "grad_norm": 1.1262688636779785,
+      "learning_rate": 0.0002,
+      "loss": 0.6025,
+      "step": 30770
+    },
+    {
+      "epoch": 4.976153908334007,
+      "grad_norm": 1.1415393352508545,
+      "learning_rate": 0.0002,
+      "loss": 0.588,
+      "step": 30780
+    },
+    {
+      "epoch": 4.977770592514752,
+      "grad_norm": 1.083078384399414,
+      "learning_rate": 0.0002,
+      "loss": 0.5832,
+      "step": 30790
+    },
+    {
+      "epoch": 4.979387276695498,
+      "grad_norm": 0.964859127998352,
+      "learning_rate": 0.0002,
+      "loss": 0.6025,
+      "step": 30800
+    },
+    {
+      "epoch": 4.981003960876243,
+      "grad_norm": 0.8704743385314941,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 30810
+    },
+    {
+      "epoch": 4.982620645056988,
+      "grad_norm": 1.0714856386184692,
+      "learning_rate": 0.0002,
+      "loss": 0.5666,
+      "step": 30820
+    },
+    {
+      "epoch": 4.984237329237733,
+      "grad_norm": 0.6818771362304688,
+      "learning_rate": 0.0002,
+      "loss": 0.565,
+      "step": 30830
+    },
+    {
+      "epoch": 4.985854013418479,
+      "grad_norm": 1.0454156398773193,
+      "learning_rate": 0.0002,
+      "loss": 0.5999,
+      "step": 30840
+    },
+    {
+      "epoch": 4.987470697599224,
+      "grad_norm": 0.9410776495933533,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 30850
+    },
+    {
+      "epoch": 4.989087381779969,
+      "grad_norm": 1.0878902673721313,
+      "learning_rate": 0.0002,
+      "loss": 0.5899,
+      "step": 30860
+    },
+    {
+      "epoch": 4.990704065960714,
+      "grad_norm": 0.8916727304458618,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 30870
+    },
+    {
+      "epoch": 4.9923207501414595,
+      "grad_norm": 1.045776128768921,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 30880
+    },
+    {
+      "epoch": 4.993937434322206,
+      "grad_norm": 0.9861903786659241,
+      "learning_rate": 0.0002,
+      "loss": 0.5767,
+      "step": 30890
+    },
+    {
+      "epoch": 4.995554118502951,
+      "grad_norm": 0.9275050759315491,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 30900
+    },
+    {
+      "epoch": 4.997170802683696,
+      "grad_norm": 0.94013911485672,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 30910
+    },
+    {
+      "epoch": 4.998787486864441,
+      "grad_norm": 0.9771268367767334,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 30920
+    },
+    {
+      "epoch": 4.9999191657909625,
+      "eval_loss": 1.1968598365783691,
+      "eval_runtime": 122.2519,
+      "eval_samples_per_second": 5.996,
+      "eval_steps_per_second": 0.753,
+      "step": 30927
+    },
+    {
+      "epoch": 5.0004041710451865,
+      "grad_norm": 0.8021580576896667,
+      "learning_rate": 0.0002,
+      "loss": 0.5238,
+      "step": 30930
+    },
+    {
+      "epoch": 5.002020855225932,
+      "grad_norm": 1.0807327032089233,
+      "learning_rate": 0.0002,
+      "loss": 0.4984,
+      "step": 30940
+    },
+    {
+      "epoch": 5.003637539406677,
+      "grad_norm": 1.1638425588607788,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 30950
+    },
+    {
+      "epoch": 5.005254223587422,
+      "grad_norm": 1.1700230836868286,
+      "learning_rate": 0.0002,
+      "loss": 0.4621,
+      "step": 30960
+    },
+    {
+      "epoch": 5.0068709077681675,
+      "grad_norm": 0.9053420424461365,
+      "learning_rate": 0.0002,
+      "loss": 0.4657,
+      "step": 30970
+    },
+    {
+      "epoch": 5.008487591948913,
+      "grad_norm": 0.9226111769676208,
+      "learning_rate": 0.0002,
+      "loss": 0.4865,
+      "step": 30980
+    },
+    {
+      "epoch": 5.010104276129658,
+      "grad_norm": 1.238669514656067,
+      "learning_rate": 0.0002,
+      "loss": 0.5011,
+      "step": 30990
+    },
+    {
+      "epoch": 5.011720960310403,
+      "grad_norm": 1.0668327808380127,
+      "learning_rate": 0.0002,
+      "loss": 0.4754,
+      "step": 31000
+    },
+    {
+      "epoch": 5.013337644491148,
+      "grad_norm": 1.0903944969177246,
+      "learning_rate": 0.0002,
+      "loss": 0.5414,
+      "step": 31010
+    },
+    {
+      "epoch": 5.014954328671894,
+      "grad_norm": 1.0763911008834839,
+      "learning_rate": 0.0002,
+      "loss": 0.5117,
+      "step": 31020
+    },
+    {
+      "epoch": 5.016571012852639,
+      "grad_norm": 1.0108771324157715,
+      "learning_rate": 0.0002,
+      "loss": 0.4908,
+      "step": 31030
+    },
+    {
+      "epoch": 5.018187697033385,
+      "grad_norm": 0.8816103935241699,
+      "learning_rate": 0.0002,
+      "loss": 0.5052,
+      "step": 31040
+    },
+    {
+      "epoch": 5.01980438121413,
+      "grad_norm": 1.11434805393219,
+      "learning_rate": 0.0002,
+      "loss": 0.4985,
+      "step": 31050
+    },
+    {
+      "epoch": 5.021421065394875,
+      "grad_norm": 1.0727789402008057,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 31060
+    },
+    {
+      "epoch": 5.023037749575621,
+      "grad_norm": 1.1480379104614258,
+      "learning_rate": 0.0002,
+      "loss": 0.4938,
+      "step": 31070
+    },
+    {
+      "epoch": 5.024654433756366,
+      "grad_norm": 1.0913071632385254,
+      "learning_rate": 0.0002,
+      "loss": 0.491,
+      "step": 31080
+    },
+    {
+      "epoch": 5.026271117937111,
+      "grad_norm": 0.9891864657402039,
+      "learning_rate": 0.0002,
+      "loss": 0.4896,
+      "step": 31090
+    },
+    {
+      "epoch": 5.027887802117856,
+      "grad_norm": 0.9167473912239075,
+      "learning_rate": 0.0002,
+      "loss": 0.4965,
+      "step": 31100
+    },
+    {
+      "epoch": 5.029504486298602,
+      "grad_norm": 1.2259035110473633,
+      "learning_rate": 0.0002,
+      "loss": 0.5098,
+      "step": 31110
+    },
+    {
+      "epoch": 5.031121170479347,
+      "grad_norm": 1.1812787055969238,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 31120
+    },
+    {
+      "epoch": 5.032737854660092,
+      "grad_norm": 1.0890522003173828,
+      "learning_rate": 0.0002,
+      "loss": 0.4725,
+      "step": 31130
+    },
+    {
+      "epoch": 5.034354538840837,
+      "grad_norm": 1.0521091222763062,
+      "learning_rate": 0.0002,
+      "loss": 0.4768,
+      "step": 31140
+    },
+    {
+      "epoch": 5.0359712230215825,
+      "grad_norm": 1.1274569034576416,
+      "learning_rate": 0.0002,
+      "loss": 0.4718,
+      "step": 31150
+    },
+    {
+      "epoch": 5.037587907202328,
+      "grad_norm": 1.140974998474121,
+      "learning_rate": 0.0002,
+      "loss": 0.4604,
+      "step": 31160
+    },
+    {
+      "epoch": 5.039204591383073,
+      "grad_norm": 1.1215609312057495,
+      "learning_rate": 0.0002,
+      "loss": 0.5077,
+      "step": 31170
+    },
+    {
+      "epoch": 5.040821275563818,
+      "grad_norm": 1.0107218027114868,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 31180
+    },
+    {
+      "epoch": 5.042437959744564,
+      "grad_norm": 1.0198770761489868,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 31190
+    },
+    {
+      "epoch": 5.0440546439253096,
+      "grad_norm": 1.1613430976867676,
+      "learning_rate": 0.0002,
+      "loss": 0.5004,
+      "step": 31200
+    },
+    {
+      "epoch": 5.045671328106055,
+      "grad_norm": 0.8555458188056946,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 31210
+    },
+    {
+      "epoch": 5.0472880122868,
+      "grad_norm": 1.0235545635223389,
+      "learning_rate": 0.0002,
+      "loss": 0.4878,
+      "step": 31220
+    },
+    {
+      "epoch": 5.048904696467545,
+      "grad_norm": 1.0228750705718994,
+      "learning_rate": 0.0002,
+      "loss": 0.499,
+      "step": 31230
+    },
+    {
+      "epoch": 5.0505213806482905,
+      "grad_norm": 0.8216419816017151,
+      "learning_rate": 0.0002,
+      "loss": 0.4544,
+      "step": 31240
+    },
+    {
+      "epoch": 5.052138064829036,
+      "grad_norm": 0.925828218460083,
+      "learning_rate": 0.0002,
+      "loss": 0.4947,
+      "step": 31250
+    },
+    {
+      "epoch": 5.053754749009781,
+      "grad_norm": 0.9229369759559631,
+      "learning_rate": 0.0002,
+      "loss": 0.4835,
+      "step": 31260
+    },
+    {
+      "epoch": 5.055371433190526,
+      "grad_norm": 0.9531727433204651,
+      "learning_rate": 0.0002,
+      "loss": 0.5136,
+      "step": 31270
+    },
+    {
+      "epoch": 5.056988117371271,
+      "grad_norm": 0.7738548517227173,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 31280
+    },
+    {
+      "epoch": 5.058604801552017,
+      "grad_norm": 1.0551451444625854,
+      "learning_rate": 0.0002,
+      "loss": 0.5166,
+      "step": 31290
+    },
+    {
+      "epoch": 5.060221485732762,
+      "grad_norm": 0.9782299399375916,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 31300
+    },
+    {
+      "epoch": 5.061838169913507,
+      "grad_norm": 1.0220632553100586,
+      "learning_rate": 0.0002,
+      "loss": 0.4776,
+      "step": 31310
+    },
+    {
+      "epoch": 5.063454854094252,
+      "grad_norm": 0.9808892607688904,
+      "learning_rate": 0.0002,
+      "loss": 0.5117,
+      "step": 31320
+    },
+    {
+      "epoch": 5.065071538274998,
+      "grad_norm": 1.0662003755569458,
+      "learning_rate": 0.0002,
+      "loss": 0.501,
+      "step": 31330
+    },
+    {
+      "epoch": 5.066688222455744,
+      "grad_norm": 1.0036940574645996,
+      "learning_rate": 0.0002,
+      "loss": 0.4844,
+      "step": 31340
+    },
+    {
+      "epoch": 5.068304906636489,
+      "grad_norm": 1.1931052207946777,
+      "learning_rate": 0.0002,
+      "loss": 0.5299,
+      "step": 31350
+    },
+    {
+      "epoch": 5.069921590817234,
+      "grad_norm": 0.9370693564414978,
+      "learning_rate": 0.0002,
+      "loss": 0.4646,
+      "step": 31360
+    },
+    {
+      "epoch": 5.071538274997979,
+      "grad_norm": 0.9589039087295532,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 31370
+    },
+    {
+      "epoch": 5.073154959178725,
+      "grad_norm": 1.0052711963653564,
+      "learning_rate": 0.0002,
+      "loss": 0.4669,
+      "step": 31380
+    },
+    {
+      "epoch": 5.07477164335947,
+      "grad_norm": 0.9991368651390076,
+      "learning_rate": 0.0002,
+      "loss": 0.5283,
+      "step": 31390
+    },
+    {
+      "epoch": 5.076388327540215,
+      "grad_norm": 0.8539695739746094,
+      "learning_rate": 0.0002,
+      "loss": 0.4579,
+      "step": 31400
+    },
+    {
+      "epoch": 5.07800501172096,
+      "grad_norm": 1.048775553703308,
+      "learning_rate": 0.0002,
+      "loss": 0.4609,
+      "step": 31410
+    },
+    {
+      "epoch": 5.0796216959017055,
+      "grad_norm": 0.9983724355697632,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 31420
+    },
+    {
+      "epoch": 5.081238380082451,
+      "grad_norm": 1.0189813375473022,
+      "learning_rate": 0.0002,
+      "loss": 0.4594,
+      "step": 31430
+    },
+    {
+      "epoch": 5.082855064263196,
+      "grad_norm": 0.9781646728515625,
+      "learning_rate": 0.0002,
+      "loss": 0.5449,
+      "step": 31440
+    },
+    {
+      "epoch": 5.084471748443941,
+      "grad_norm": 0.9424566030502319,
+      "learning_rate": 0.0002,
+      "loss": 0.4698,
+      "step": 31450
+    },
+    {
+      "epoch": 5.0860884326246865,
+      "grad_norm": 1.0036484003067017,
+      "learning_rate": 0.0002,
+      "loss": 0.4768,
+      "step": 31460
+    },
+    {
+      "epoch": 5.087705116805432,
+      "grad_norm": 1.0983147621154785,
+      "learning_rate": 0.0002,
+      "loss": 0.487,
+      "step": 31470
+    },
+    {
+      "epoch": 5.089321800986177,
+      "grad_norm": 1.0856730937957764,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 31480
+    },
+    {
+      "epoch": 5.090938485166923,
+      "grad_norm": 1.2191699743270874,
+      "learning_rate": 0.0002,
+      "loss": 0.485,
+      "step": 31490
+    },
+    {
+      "epoch": 5.092555169347668,
+      "grad_norm": 0.939346194267273,
+      "learning_rate": 0.0002,
+      "loss": 0.4936,
+      "step": 31500
+    },
+    {
+      "epoch": 5.0941718535284135,
+      "grad_norm": 0.9730121493339539,
+      "learning_rate": 0.0002,
+      "loss": 0.5107,
+      "step": 31510
+    },
+    {
+      "epoch": 5.095788537709159,
+      "grad_norm": 0.923686146736145,
+      "learning_rate": 0.0002,
+      "loss": 0.4973,
+      "step": 31520
+    },
+    {
+      "epoch": 5.097405221889904,
+      "grad_norm": 1.1734349727630615,
+      "learning_rate": 0.0002,
+      "loss": 0.4906,
+      "step": 31530
+    },
+    {
+      "epoch": 5.099021906070649,
+      "grad_norm": 1.084509015083313,
+      "learning_rate": 0.0002,
+      "loss": 0.5165,
+      "step": 31540
+    },
+    {
+      "epoch": 5.100638590251394,
+      "grad_norm": 1.0144678354263306,
+      "learning_rate": 0.0002,
+      "loss": 0.5078,
+      "step": 31550
+    },
+    {
+      "epoch": 5.10225527443214,
+      "grad_norm": 0.9958019256591797,
+      "learning_rate": 0.0002,
+      "loss": 0.4719,
+      "step": 31560
+    },
+    {
+      "epoch": 5.103871958612885,
+      "grad_norm": 0.8900736570358276,
+      "learning_rate": 0.0002,
+      "loss": 0.4876,
+      "step": 31570
+    },
+    {
+      "epoch": 5.10548864279363,
+      "grad_norm": 1.0921649932861328,
+      "learning_rate": 0.0002,
+      "loss": 0.463,
+      "step": 31580
+    },
+    {
+      "epoch": 5.107105326974375,
+      "grad_norm": 1.1613792181015015,
+      "learning_rate": 0.0002,
+      "loss": 0.5148,
+      "step": 31590
+    },
+    {
+      "epoch": 5.108722011155121,
+      "grad_norm": 0.9211367964744568,
+      "learning_rate": 0.0002,
+      "loss": 0.5055,
+      "step": 31600
+    },
+    {
+      "epoch": 5.110338695335866,
+      "grad_norm": 1.3315813541412354,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 31610
+    },
+    {
+      "epoch": 5.111955379516611,
+      "grad_norm": 1.3765019178390503,
+      "learning_rate": 0.0002,
+      "loss": 0.5336,
+      "step": 31620
+    },
+    {
+      "epoch": 5.113572063697356,
+      "grad_norm": 1.070198893547058,
+      "learning_rate": 0.0002,
+      "loss": 0.4861,
+      "step": 31630
+    },
+    {
+      "epoch": 5.115188747878102,
+      "grad_norm": 0.947631299495697,
+      "learning_rate": 0.0002,
+      "loss": 0.5046,
+      "step": 31640
+    },
+    {
+      "epoch": 5.116805432058848,
+      "grad_norm": 1.0197371244430542,
+      "learning_rate": 0.0002,
+      "loss": 0.5297,
+      "step": 31650
+    },
+    {
+      "epoch": 5.118422116239593,
+      "grad_norm": 0.8647911548614502,
+      "learning_rate": 0.0002,
+      "loss": 0.5014,
+      "step": 31660
+    },
+    {
+      "epoch": 5.120038800420338,
+      "grad_norm": 0.8944075107574463,
+      "learning_rate": 0.0002,
+      "loss": 0.4705,
+      "step": 31670
+    },
+    {
+      "epoch": 5.121655484601083,
+      "grad_norm": 1.124497652053833,
+      "learning_rate": 0.0002,
+      "loss": 0.5175,
+      "step": 31680
+    },
+    {
+      "epoch": 5.123272168781829,
+      "grad_norm": 0.893131673336029,
+      "learning_rate": 0.0002,
+      "loss": 0.5109,
+      "step": 31690
+    },
+    {
+      "epoch": 5.124888852962574,
+      "grad_norm": 1.0122284889221191,
+      "learning_rate": 0.0002,
+      "loss": 0.4937,
+      "step": 31700
+    },
+    {
+      "epoch": 5.126505537143319,
+      "grad_norm": 0.9493719935417175,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 31710
+    },
+    {
+      "epoch": 5.128122221324064,
+      "grad_norm": 0.9700539112091064,
+      "learning_rate": 0.0002,
+      "loss": 0.5031,
+      "step": 31720
+    },
+    {
+      "epoch": 5.1297389055048095,
+      "grad_norm": 1.111677646636963,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 31730
+    },
+    {
+      "epoch": 5.131355589685555,
+      "grad_norm": 0.8204274773597717,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 31740
+    },
+    {
+      "epoch": 5.1329722738663,
+      "grad_norm": 1.1029267311096191,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 31750
+    },
+    {
+      "epoch": 5.134588958047045,
+      "grad_norm": 1.065575122833252,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 31760
+    },
+    {
+      "epoch": 5.13620564222779,
+      "grad_norm": 0.8208706974983215,
+      "learning_rate": 0.0002,
+      "loss": 0.502,
+      "step": 31770
+    },
+    {
+      "epoch": 5.137822326408536,
+      "grad_norm": 1.0520979166030884,
+      "learning_rate": 0.0002,
+      "loss": 0.5352,
+      "step": 31780
+    },
+    {
+      "epoch": 5.139439010589282,
+      "grad_norm": 0.8585538268089294,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 31790
+    },
+    {
+      "epoch": 5.141055694770027,
+      "grad_norm": 1.1491447687149048,
+      "learning_rate": 0.0002,
+      "loss": 0.5159,
+      "step": 31800
+    },
+    {
+      "epoch": 5.142672378950772,
+      "grad_norm": 0.9441081285476685,
+      "learning_rate": 0.0002,
+      "loss": 0.5157,
+      "step": 31810
+    },
+    {
+      "epoch": 5.1442890631315175,
+      "grad_norm": 1.4146889448165894,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 31820
+    },
+    {
+      "epoch": 5.145905747312263,
+      "grad_norm": 1.0326547622680664,
+      "learning_rate": 0.0002,
+      "loss": 0.5159,
+      "step": 31830
+    },
+    {
+      "epoch": 5.147522431493008,
+      "grad_norm": 0.9879202842712402,
+      "learning_rate": 0.0002,
+      "loss": 0.5348,
+      "step": 31840
+    },
+    {
+      "epoch": 5.149139115673753,
+      "grad_norm": 1.0374281406402588,
+      "learning_rate": 0.0002,
+      "loss": 0.5083,
+      "step": 31850
+    },
+    {
+      "epoch": 5.150755799854498,
+      "grad_norm": 1.181229591369629,
+      "learning_rate": 0.0002,
+      "loss": 0.4827,
+      "step": 31860
+    },
+    {
+      "epoch": 5.152372484035244,
+      "grad_norm": 1.2078537940979004,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 31870
+    },
+    {
+      "epoch": 5.153989168215989,
+      "grad_norm": 0.9599190354347229,
+      "learning_rate": 0.0002,
+      "loss": 0.5329,
+      "step": 31880
+    },
+    {
+      "epoch": 5.155605852396734,
+      "grad_norm": 1.0378568172454834,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 31890
+    },
+    {
+      "epoch": 5.157222536577479,
+      "grad_norm": 0.8746536374092102,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 31900
+    },
+    {
+      "epoch": 5.1588392207582245,
+      "grad_norm": 1.0232136249542236,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 31910
+    },
+    {
+      "epoch": 5.16045590493897,
+      "grad_norm": 0.9827565550804138,
+      "learning_rate": 0.0002,
+      "loss": 0.4844,
+      "step": 31920
+    },
+    {
+      "epoch": 5.162072589119716,
+      "grad_norm": 1.342657208442688,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 31930
+    },
+    {
+      "epoch": 5.163689273300461,
+      "grad_norm": 1.18390691280365,
+      "learning_rate": 0.0002,
+      "loss": 0.513,
+      "step": 31940
+    },
+    {
+      "epoch": 5.165305957481206,
+      "grad_norm": 0.996350109577179,
+      "learning_rate": 0.0002,
+      "loss": 0.5267,
+      "step": 31950
+    },
+    {
+      "epoch": 5.166922641661952,
+      "grad_norm": 0.9710391163825989,
+      "learning_rate": 0.0002,
+      "loss": 0.5063,
+      "step": 31960
+    },
+    {
+      "epoch": 5.168539325842697,
+      "grad_norm": 1.0264002084732056,
+      "learning_rate": 0.0002,
+      "loss": 0.5115,
+      "step": 31970
+    },
+    {
+      "epoch": 5.170156010023442,
+      "grad_norm": 1.0028311014175415,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 31980
+    },
+    {
+      "epoch": 5.171772694204187,
+      "grad_norm": 1.1078234910964966,
+      "learning_rate": 0.0002,
+      "loss": 0.5103,
+      "step": 31990
+    },
+    {
+      "epoch": 5.1733893783849325,
+      "grad_norm": 0.9659610390663147,
+      "learning_rate": 0.0002,
+      "loss": 0.495,
+      "step": 32000
+    },
+    {
+      "epoch": 5.175006062565678,
+      "grad_norm": 0.841986894607544,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 32010
+    },
+    {
+      "epoch": 5.176622746746423,
+      "grad_norm": 1.095332384109497,
+      "learning_rate": 0.0002,
+      "loss": 0.48,
+      "step": 32020
+    },
+    {
+      "epoch": 5.178239430927168,
+      "grad_norm": 1.1242377758026123,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 32030
+    },
+    {
+      "epoch": 5.179856115107913,
+      "grad_norm": 0.9872292280197144,
+      "learning_rate": 0.0002,
+      "loss": 0.5573,
+      "step": 32040
+    },
+    {
+      "epoch": 5.181472799288659,
+      "grad_norm": 0.936161994934082,
+      "learning_rate": 0.0002,
+      "loss": 0.48,
+      "step": 32050
+    },
+    {
+      "epoch": 5.183089483469404,
+      "grad_norm": 1.166100025177002,
+      "learning_rate": 0.0002,
+      "loss": 0.5093,
+      "step": 32060
+    },
+    {
+      "epoch": 5.184706167650149,
+      "grad_norm": 1.0764425992965698,
+      "learning_rate": 0.0002,
+      "loss": 0.5438,
+      "step": 32070
+    },
+    {
+      "epoch": 5.186322851830895,
+      "grad_norm": 1.0480051040649414,
+      "learning_rate": 0.0002,
+      "loss": 0.4843,
+      "step": 32080
+    },
+    {
+      "epoch": 5.1879395360116405,
+      "grad_norm": 1.0874916315078735,
+      "learning_rate": 0.0002,
+      "loss": 0.5386,
+      "step": 32090
+    },
+    {
+      "epoch": 5.189556220192386,
+      "grad_norm": 1.0817396640777588,
+      "learning_rate": 0.0002,
+      "loss": 0.4975,
+      "step": 32100
+    },
+    {
+      "epoch": 5.191172904373131,
+      "grad_norm": 1.054111361503601,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 32110
+    },
+    {
+      "epoch": 5.192789588553876,
+      "grad_norm": 0.9655823707580566,
+      "learning_rate": 0.0002,
+      "loss": 0.5229,
+      "step": 32120
+    },
+    {
+      "epoch": 5.194406272734621,
+      "grad_norm": 1.1384109258651733,
+      "learning_rate": 0.0002,
+      "loss": 0.5105,
+      "step": 32130
+    },
+    {
+      "epoch": 5.196022956915367,
+      "grad_norm": 1.0149348974227905,
+      "learning_rate": 0.0002,
+      "loss": 0.5073,
+      "step": 32140
+    },
+    {
+      "epoch": 5.197639641096112,
+      "grad_norm": 1.1084046363830566,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 32150
+    },
+    {
+      "epoch": 5.199256325276857,
+      "grad_norm": 1.1209309101104736,
+      "learning_rate": 0.0002,
+      "loss": 0.4936,
+      "step": 32160
+    },
+    {
+      "epoch": 5.200873009457602,
+      "grad_norm": 1.133089542388916,
+      "learning_rate": 0.0002,
+      "loss": 0.5101,
+      "step": 32170
+    },
+    {
+      "epoch": 5.202489693638348,
+      "grad_norm": 1.0893020629882812,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 32180
+    },
+    {
+      "epoch": 5.204106377819093,
+      "grad_norm": 0.90018630027771,
+      "learning_rate": 0.0002,
+      "loss": 0.4872,
+      "step": 32190
+    },
+    {
+      "epoch": 5.205723061999838,
+      "grad_norm": 0.977622926235199,
+      "learning_rate": 0.0002,
+      "loss": 0.4999,
+      "step": 32200
+    },
+    {
+      "epoch": 5.207339746180583,
+      "grad_norm": 1.2940177917480469,
+      "learning_rate": 0.0002,
+      "loss": 0.5028,
+      "step": 32210
+    },
+    {
+      "epoch": 5.2089564303613285,
+      "grad_norm": 1.2131710052490234,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 32220
+    },
+    {
+      "epoch": 5.210573114542075,
+      "grad_norm": 1.0234841108322144,
+      "learning_rate": 0.0002,
+      "loss": 0.5189,
+      "step": 32230
+    },
+    {
+      "epoch": 5.21218979872282,
+      "grad_norm": 1.157975435256958,
+      "learning_rate": 0.0002,
+      "loss": 0.5424,
+      "step": 32240
+    },
+    {
+      "epoch": 5.213806482903565,
+      "grad_norm": 1.0381282567977905,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 32250
+    },
+    {
+      "epoch": 5.21542316708431,
+      "grad_norm": 1.0125395059585571,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 32260
+    },
+    {
+      "epoch": 5.2170398512650555,
+      "grad_norm": 1.272691011428833,
+      "learning_rate": 0.0002,
+      "loss": 0.5216,
+      "step": 32270
+    },
+    {
+      "epoch": 5.218656535445801,
+      "grad_norm": 1.0061250925064087,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 32280
+    },
+    {
+      "epoch": 5.220273219626546,
+      "grad_norm": 0.9752234816551208,
+      "learning_rate": 0.0002,
+      "loss": 0.4739,
+      "step": 32290
+    },
+    {
+      "epoch": 5.221889903807291,
+      "grad_norm": 1.1193140745162964,
+      "learning_rate": 0.0002,
+      "loss": 0.5471,
+      "step": 32300
+    },
+    {
+      "epoch": 5.2235065879880365,
+      "grad_norm": 1.0126434564590454,
+      "learning_rate": 0.0002,
+      "loss": 0.4976,
+      "step": 32310
+    },
+    {
+      "epoch": 5.225123272168782,
+      "grad_norm": 1.4338394403457642,
+      "learning_rate": 0.0002,
+      "loss": 0.5257,
+      "step": 32320
+    },
+    {
+      "epoch": 5.226739956349527,
+      "grad_norm": 1.004101276397705,
+      "learning_rate": 0.0002,
+      "loss": 0.5235,
+      "step": 32330
+    },
+    {
+      "epoch": 5.228356640530272,
+      "grad_norm": 0.8744166493415833,
+      "learning_rate": 0.0002,
+      "loss": 0.5091,
+      "step": 32340
+    },
+    {
+      "epoch": 5.229973324711017,
+      "grad_norm": 1.0165376663208008,
+      "learning_rate": 0.0002,
+      "loss": 0.5388,
+      "step": 32350
+    },
+    {
+      "epoch": 5.231590008891763,
+      "grad_norm": 0.8635954260826111,
+      "learning_rate": 0.0002,
+      "loss": 0.5469,
+      "step": 32360
+    },
+    {
+      "epoch": 5.233206693072509,
+      "grad_norm": 1.1392399072647095,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 32370
+    },
+    {
+      "epoch": 5.234823377253254,
+      "grad_norm": 1.0202113389968872,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 32380
+    },
+    {
+      "epoch": 5.236440061433999,
+      "grad_norm": 1.0417983531951904,
+      "learning_rate": 0.0002,
+      "loss": 0.4983,
+      "step": 32390
+    },
+    {
+      "epoch": 5.238056745614744,
+      "grad_norm": 0.8729333877563477,
+      "learning_rate": 0.0002,
+      "loss": 0.507,
+      "step": 32400
+    },
+    {
+      "epoch": 5.23967342979549,
+      "grad_norm": 1.1626229286193848,
+      "learning_rate": 0.0002,
+      "loss": 0.5426,
+      "step": 32410
+    },
+    {
+      "epoch": 5.241290113976235,
+      "grad_norm": 0.9086161851882935,
+      "learning_rate": 0.0002,
+      "loss": 0.5355,
+      "step": 32420
+    },
+    {
+      "epoch": 5.24290679815698,
+      "grad_norm": 1.3999892473220825,
+      "learning_rate": 0.0002,
+      "loss": 0.4927,
+      "step": 32430
+    },
+    {
+      "epoch": 5.244523482337725,
+      "grad_norm": 1.0356311798095703,
+      "learning_rate": 0.0002,
+      "loss": 0.4795,
+      "step": 32440
+    },
+    {
+      "epoch": 5.246140166518471,
+      "grad_norm": 0.9655531644821167,
+      "learning_rate": 0.0002,
+      "loss": 0.5035,
+      "step": 32450
+    },
+    {
+      "epoch": 5.247756850699216,
+      "grad_norm": 1.0411828756332397,
+      "learning_rate": 0.0002,
+      "loss": 0.5166,
+      "step": 32460
+    },
+    {
+      "epoch": 5.249373534879961,
+      "grad_norm": 1.1199816465377808,
+      "learning_rate": 0.0002,
+      "loss": 0.5141,
+      "step": 32470
+    },
+    {
+      "epoch": 5.250990219060706,
+      "grad_norm": 1.260321855545044,
+      "learning_rate": 0.0002,
+      "loss": 0.4864,
+      "step": 32480
+    },
+    {
+      "epoch": 5.2526069032414515,
+      "grad_norm": 1.2950857877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.4893,
+      "step": 32490
+    },
+    {
+      "epoch": 5.254223587422197,
+      "grad_norm": 0.8982820510864258,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 32500
+    },
+    {
+      "epoch": 5.255840271602942,
+      "grad_norm": 0.8512987494468689,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 32510
+    },
+    {
+      "epoch": 5.257456955783688,
+      "grad_norm": 1.067443609237671,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 32520
+    },
+    {
+      "epoch": 5.259073639964433,
+      "grad_norm": 1.0957417488098145,
+      "learning_rate": 0.0002,
+      "loss": 0.4928,
+      "step": 32530
+    },
+    {
+      "epoch": 5.260690324145179,
+      "grad_norm": 1.4161807298660278,
+      "learning_rate": 0.0002,
+      "loss": 0.5169,
+      "step": 32540
+    },
+    {
+      "epoch": 5.262307008325924,
+      "grad_norm": 1.2264093160629272,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 32550
+    },
+    {
+      "epoch": 5.263923692506669,
+      "grad_norm": 1.0015931129455566,
+      "learning_rate": 0.0002,
+      "loss": 0.5221,
+      "step": 32560
+    },
+    {
+      "epoch": 5.265540376687414,
+      "grad_norm": 1.0743094682693481,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 32570
+    },
+    {
+      "epoch": 5.2671570608681595,
+      "grad_norm": 1.1386840343475342,
+      "learning_rate": 0.0002,
+      "loss": 0.5289,
+      "step": 32580
+    },
+    {
+      "epoch": 5.268773745048905,
+      "grad_norm": 1.0093860626220703,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 32590
+    },
+    {
+      "epoch": 5.27039042922965,
+      "grad_norm": 0.9593744874000549,
+      "learning_rate": 0.0002,
+      "loss": 0.5175,
+      "step": 32600
+    },
+    {
+      "epoch": 5.272007113410395,
+      "grad_norm": 1.146021842956543,
+      "learning_rate": 0.0002,
+      "loss": 0.528,
+      "step": 32610
+    },
+    {
+      "epoch": 5.27362379759114,
+      "grad_norm": 0.9579031467437744,
+      "learning_rate": 0.0002,
+      "loss": 0.4983,
+      "step": 32620
+    },
+    {
+      "epoch": 5.275240481771886,
+      "grad_norm": 1.0548793077468872,
+      "learning_rate": 0.0002,
+      "loss": 0.5376,
+      "step": 32630
+    },
+    {
+      "epoch": 5.276857165952631,
+      "grad_norm": 1.0380561351776123,
+      "learning_rate": 0.0002,
+      "loss": 0.5267,
+      "step": 32640
+    },
+    {
+      "epoch": 5.278473850133376,
+      "grad_norm": 1.2119969129562378,
+      "learning_rate": 0.0002,
+      "loss": 0.5182,
+      "step": 32650
+    },
+    {
+      "epoch": 5.280090534314121,
+      "grad_norm": 1.0507797002792358,
+      "learning_rate": 0.0002,
+      "loss": 0.5298,
+      "step": 32660
+    },
+    {
+      "epoch": 5.2817072184948675,
+      "grad_norm": 1.0185176134109497,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 32670
+    },
+    {
+      "epoch": 5.283323902675613,
+      "grad_norm": 1.2358098030090332,
+      "learning_rate": 0.0002,
+      "loss": 0.4904,
+      "step": 32680
+    },
+    {
+      "epoch": 5.284940586856358,
+      "grad_norm": 0.7937114238739014,
+      "learning_rate": 0.0002,
+      "loss": 0.5169,
+      "step": 32690
+    },
+    {
+      "epoch": 5.286557271037103,
+      "grad_norm": 0.9825124740600586,
+      "learning_rate": 0.0002,
+      "loss": 0.495,
+      "step": 32700
+    },
+    {
+      "epoch": 5.288173955217848,
+      "grad_norm": 1.2059301137924194,
+      "learning_rate": 0.0002,
+      "loss": 0.5149,
+      "step": 32710
+    },
+    {
+      "epoch": 5.289790639398594,
+      "grad_norm": 1.0828571319580078,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 32720
+    },
+    {
+      "epoch": 5.291407323579339,
+      "grad_norm": 1.0129735469818115,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 32730
+    },
+    {
+      "epoch": 5.293024007760084,
+      "grad_norm": 1.0591634511947632,
+      "learning_rate": 0.0002,
+      "loss": 0.5216,
+      "step": 32740
+    },
+    {
+      "epoch": 5.294640691940829,
+      "grad_norm": 0.9256815910339355,
+      "learning_rate": 0.0002,
+      "loss": 0.522,
+      "step": 32750
+    },
+    {
+      "epoch": 5.2962573761215745,
+      "grad_norm": 1.0928633213043213,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 32760
+    },
+    {
+      "epoch": 5.29787406030232,
+      "grad_norm": 0.9415594935417175,
+      "learning_rate": 0.0002,
+      "loss": 0.5093,
+      "step": 32770
+    },
+    {
+      "epoch": 5.299490744483065,
+      "grad_norm": 1.141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.5252,
+      "step": 32780
+    },
+    {
+      "epoch": 5.30110742866381,
+      "grad_norm": 1.0646510124206543,
+      "learning_rate": 0.0002,
+      "loss": 0.4837,
+      "step": 32790
+    },
+    {
+      "epoch": 5.3027241128445555,
+      "grad_norm": 1.189661979675293,
+      "learning_rate": 0.0002,
+      "loss": 0.5547,
+      "step": 32800
+    },
+    {
+      "epoch": 5.304340797025301,
+      "grad_norm": 0.9568731188774109,
+      "learning_rate": 0.0002,
+      "loss": 0.5664,
+      "step": 32810
+    },
+    {
+      "epoch": 5.305957481206047,
+      "grad_norm": 1.1556824445724487,
+      "learning_rate": 0.0002,
+      "loss": 0.5344,
+      "step": 32820
+    },
+    {
+      "epoch": 5.307574165386792,
+      "grad_norm": 0.9353463649749756,
+      "learning_rate": 0.0002,
+      "loss": 0.4894,
+      "step": 32830
+    },
+    {
+      "epoch": 5.309190849567537,
+      "grad_norm": 1.1208295822143555,
+      "learning_rate": 0.0002,
+      "loss": 0.5052,
+      "step": 32840
+    },
+    {
+      "epoch": 5.3108075337482825,
+      "grad_norm": 1.0894153118133545,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 32850
+    },
+    {
+      "epoch": 5.312424217929028,
+      "grad_norm": 1.090329647064209,
+      "learning_rate": 0.0002,
+      "loss": 0.5046,
+      "step": 32860
+    },
+    {
+      "epoch": 5.314040902109773,
+      "grad_norm": 1.0781712532043457,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 32870
+    },
+    {
+      "epoch": 5.315657586290518,
+      "grad_norm": 1.1785295009613037,
+      "learning_rate": 0.0002,
+      "loss": 0.57,
+      "step": 32880
+    },
+    {
+      "epoch": 5.317274270471263,
+      "grad_norm": 1.0406851768493652,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 32890
+    },
+    {
+      "epoch": 5.318890954652009,
+      "grad_norm": 1.0982953310012817,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 32900
+    },
+    {
+      "epoch": 5.320507638832754,
+      "grad_norm": 1.2969383001327515,
+      "learning_rate": 0.0002,
+      "loss": 0.4944,
+      "step": 32910
+    },
+    {
+      "epoch": 5.322124323013499,
+      "grad_norm": 0.9687288999557495,
+      "learning_rate": 0.0002,
+      "loss": 0.4786,
+      "step": 32920
+    },
+    {
+      "epoch": 5.323741007194244,
+      "grad_norm": 1.136760950088501,
+      "learning_rate": 0.0002,
+      "loss": 0.5286,
+      "step": 32930
+    },
+    {
+      "epoch": 5.32535769137499,
+      "grad_norm": 1.3045495748519897,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 32940
+    },
+    {
+      "epoch": 5.326974375555735,
+      "grad_norm": 1.221675992012024,
+      "learning_rate": 0.0002,
+      "loss": 0.5413,
+      "step": 32950
+    },
+    {
+      "epoch": 5.32859105973648,
+      "grad_norm": 1.1380633115768433,
+      "learning_rate": 0.0002,
+      "loss": 0.4999,
+      "step": 32960
+    },
+    {
+      "epoch": 5.330207743917226,
+      "grad_norm": 1.1065956354141235,
+      "learning_rate": 0.0002,
+      "loss": 0.5037,
+      "step": 32970
+    },
+    {
+      "epoch": 5.331824428097971,
+      "grad_norm": 1.0187175273895264,
+      "learning_rate": 0.0002,
+      "loss": 0.4913,
+      "step": 32980
+    },
+    {
+      "epoch": 5.333441112278717,
+      "grad_norm": 0.9077118039131165,
+      "learning_rate": 0.0002,
+      "loss": 0.5234,
+      "step": 32990
+    },
+    {
+      "epoch": 5.335057796459462,
+      "grad_norm": 1.0092815160751343,
+      "learning_rate": 0.0002,
+      "loss": 0.5071,
+      "step": 33000
+    },
+    {
+      "epoch": 5.336674480640207,
+      "grad_norm": 1.0168777704238892,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 33010
+    },
+    {
+      "epoch": 5.338291164820952,
+      "grad_norm": 0.996161937713623,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 33020
+    },
+    {
+      "epoch": 5.339907849001698,
+      "grad_norm": 0.794463038444519,
+      "learning_rate": 0.0002,
+      "loss": 0.5024,
+      "step": 33030
+    },
+    {
+      "epoch": 5.341524533182443,
+      "grad_norm": 0.9750674962997437,
+      "learning_rate": 0.0002,
+      "loss": 0.5112,
+      "step": 33040
+    },
+    {
+      "epoch": 5.343141217363188,
+      "grad_norm": 1.2770029306411743,
+      "learning_rate": 0.0002,
+      "loss": 0.528,
+      "step": 33050
+    },
+    {
+      "epoch": 5.344757901543933,
+      "grad_norm": 1.1500186920166016,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 33060
+    },
+    {
+      "epoch": 5.3463745857246785,
+      "grad_norm": 1.0726377964019775,
+      "learning_rate": 0.0002,
+      "loss": 0.4906,
+      "step": 33070
+    },
+    {
+      "epoch": 5.347991269905424,
+      "grad_norm": 0.9314153790473938,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 33080
+    },
+    {
+      "epoch": 5.349607954086169,
+      "grad_norm": 1.344988465309143,
+      "learning_rate": 0.0002,
+      "loss": 0.5434,
+      "step": 33090
+    },
+    {
+      "epoch": 5.351224638266914,
+      "grad_norm": 0.863196611404419,
+      "learning_rate": 0.0002,
+      "loss": 0.4874,
+      "step": 33100
+    },
+    {
+      "epoch": 5.352841322447659,
+      "grad_norm": 1.128100037574768,
+      "learning_rate": 0.0002,
+      "loss": 0.534,
+      "step": 33110
+    },
+    {
+      "epoch": 5.3544580066284055,
+      "grad_norm": 1.1673583984375,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 33120
+    },
+    {
+      "epoch": 5.356074690809151,
+      "grad_norm": 0.9416789412498474,
+      "learning_rate": 0.0002,
+      "loss": 0.4787,
+      "step": 33130
+    },
+    {
+      "epoch": 5.357691374989896,
+      "grad_norm": 1.1855236291885376,
+      "learning_rate": 0.0002,
+      "loss": 0.5155,
+      "step": 33140
+    },
+    {
+      "epoch": 5.359308059170641,
+      "grad_norm": 1.0415170192718506,
+      "learning_rate": 0.0002,
+      "loss": 0.515,
+      "step": 33150
+    },
+    {
+      "epoch": 5.3609247433513865,
+      "grad_norm": 0.9953004121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.545,
+      "step": 33160
+    },
+    {
+      "epoch": 5.362541427532132,
+      "grad_norm": 0.96138596534729,
+      "learning_rate": 0.0002,
+      "loss": 0.5305,
+      "step": 33170
+    },
+    {
+      "epoch": 5.364158111712877,
+      "grad_norm": 1.341979742050171,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 33180
+    },
+    {
+      "epoch": 5.365774795893622,
+      "grad_norm": 1.0136911869049072,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 33190
+    },
+    {
+      "epoch": 5.367391480074367,
+      "grad_norm": 0.8685575127601624,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 33200
+    },
+    {
+      "epoch": 5.369008164255113,
+      "grad_norm": 0.8833574652671814,
+      "learning_rate": 0.0002,
+      "loss": 0.5146,
+      "step": 33210
+    },
+    {
+      "epoch": 5.370624848435858,
+      "grad_norm": 0.9123612642288208,
+      "learning_rate": 0.0002,
+      "loss": 0.4982,
+      "step": 33220
+    },
+    {
+      "epoch": 5.372241532616603,
+      "grad_norm": 1.2720599174499512,
+      "learning_rate": 0.0002,
+      "loss": 0.5047,
+      "step": 33230
+    },
+    {
+      "epoch": 5.373858216797348,
+      "grad_norm": 1.0596648454666138,
+      "learning_rate": 0.0002,
+      "loss": 0.5175,
+      "step": 33240
+    },
+    {
+      "epoch": 5.3754749009780936,
+      "grad_norm": 1.119701623916626,
+      "learning_rate": 0.0002,
+      "loss": 0.5284,
+      "step": 33250
+    },
+    {
+      "epoch": 5.377091585158839,
+      "grad_norm": 1.3000061511993408,
+      "learning_rate": 0.0002,
+      "loss": 0.5217,
+      "step": 33260
+    },
+    {
+      "epoch": 5.378708269339585,
+      "grad_norm": 1.083891749382019,
+      "learning_rate": 0.0002,
+      "loss": 0.5125,
+      "step": 33270
+    },
+    {
+      "epoch": 5.38032495352033,
+      "grad_norm": 0.9402718544006348,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 33280
+    },
+    {
+      "epoch": 5.381941637701075,
+      "grad_norm": 1.3376892805099487,
+      "learning_rate": 0.0002,
+      "loss": 0.5559,
+      "step": 33290
+    },
+    {
+      "epoch": 5.383558321881821,
+      "grad_norm": 1.1600074768066406,
+      "learning_rate": 0.0002,
+      "loss": 0.5193,
+      "step": 33300
+    },
+    {
+      "epoch": 5.385175006062566,
+      "grad_norm": 1.1449427604675293,
+      "learning_rate": 0.0002,
+      "loss": 0.4907,
+      "step": 33310
+    },
+    {
+      "epoch": 5.386791690243311,
+      "grad_norm": 1.3118891716003418,
+      "learning_rate": 0.0002,
+      "loss": 0.5449,
+      "step": 33320
+    },
+    {
+      "epoch": 5.388408374424056,
+      "grad_norm": 0.743449866771698,
+      "learning_rate": 0.0002,
+      "loss": 0.547,
+      "step": 33330
+    },
+    {
+      "epoch": 5.3900250586048015,
+      "grad_norm": 0.9358304142951965,
+      "learning_rate": 0.0002,
+      "loss": 0.5555,
+      "step": 33340
+    },
+    {
+      "epoch": 5.391641742785547,
+      "grad_norm": 1.0447142124176025,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 33350
+    },
+    {
+      "epoch": 5.393258426966292,
+      "grad_norm": 1.1088626384735107,
+      "learning_rate": 0.0002,
+      "loss": 0.5106,
+      "step": 33360
+    },
+    {
+      "epoch": 5.394875111147037,
+      "grad_norm": 1.1267958879470825,
+      "learning_rate": 0.0002,
+      "loss": 0.4929,
+      "step": 33370
+    },
+    {
+      "epoch": 5.3964917953277824,
+      "grad_norm": 0.9709370136260986,
+      "learning_rate": 0.0002,
+      "loss": 0.5165,
+      "step": 33380
+    },
+    {
+      "epoch": 5.398108479508528,
+      "grad_norm": 1.0939103364944458,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 33390
+    },
+    {
+      "epoch": 5.399725163689273,
+      "grad_norm": 0.9559304714202881,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 33400
+    },
+    {
+      "epoch": 5.401341847870018,
+      "grad_norm": 1.199580430984497,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 33410
+    },
+    {
+      "epoch": 5.402958532050764,
+      "grad_norm": 0.9097000360488892,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 33420
+    },
+    {
+      "epoch": 5.4045752162315095,
+      "grad_norm": 1.1940981149673462,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 33430
+    },
+    {
+      "epoch": 5.406191900412255,
+      "grad_norm": 1.0530916452407837,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 33440
+    },
+    {
+      "epoch": 5.407808584593,
+      "grad_norm": 1.0482549667358398,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 33450
+    },
+    {
+      "epoch": 5.409425268773745,
+      "grad_norm": 1.2524714469909668,
+      "learning_rate": 0.0002,
+      "loss": 0.501,
+      "step": 33460
+    },
+    {
+      "epoch": 5.41104195295449,
+      "grad_norm": 1.1091666221618652,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 33470
+    },
+    {
+      "epoch": 5.412658637135236,
+      "grad_norm": 0.9981587529182434,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 33480
+    },
+    {
+      "epoch": 5.414275321315981,
+      "grad_norm": 1.016681432723999,
+      "learning_rate": 0.0002,
+      "loss": 0.4977,
+      "step": 33490
+    },
+    {
+      "epoch": 5.415892005496726,
+      "grad_norm": 1.1456854343414307,
+      "learning_rate": 0.0002,
+      "loss": 0.5388,
+      "step": 33500
+    },
+    {
+      "epoch": 5.417508689677471,
+      "grad_norm": 1.1454259157180786,
+      "learning_rate": 0.0002,
+      "loss": 0.5292,
+      "step": 33510
+    },
+    {
+      "epoch": 5.419125373858217,
+      "grad_norm": 0.9858416318893433,
+      "learning_rate": 0.0002,
+      "loss": 0.5061,
+      "step": 33520
+    },
+    {
+      "epoch": 5.420742058038962,
+      "grad_norm": 0.9764766693115234,
+      "learning_rate": 0.0002,
+      "loss": 0.5139,
+      "step": 33530
+    },
+    {
+      "epoch": 5.422358742219707,
+      "grad_norm": 1.199920892715454,
+      "learning_rate": 0.0002,
+      "loss": 0.5518,
+      "step": 33540
+    },
+    {
+      "epoch": 5.423975426400452,
+      "grad_norm": 1.3107370138168335,
+      "learning_rate": 0.0002,
+      "loss": 0.5182,
+      "step": 33550
+    },
+    {
+      "epoch": 5.4255921105811975,
+      "grad_norm": 0.9637970328330994,
+      "learning_rate": 0.0002,
+      "loss": 0.5149,
+      "step": 33560
+    },
+    {
+      "epoch": 5.427208794761944,
+      "grad_norm": 1.023359775543213,
+      "learning_rate": 0.0002,
+      "loss": 0.526,
+      "step": 33570
+    },
+    {
+      "epoch": 5.428825478942689,
+      "grad_norm": 1.060417652130127,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 33580
+    },
+    {
+      "epoch": 5.430442163123434,
+      "grad_norm": 0.9971120953559875,
+      "learning_rate": 0.0002,
+      "loss": 0.5052,
+      "step": 33590
+    },
+    {
+      "epoch": 5.432058847304179,
+      "grad_norm": 0.9213743209838867,
+      "learning_rate": 0.0002,
+      "loss": 0.5044,
+      "step": 33600
+    },
+    {
+      "epoch": 5.4336755314849245,
+      "grad_norm": 1.1512309312820435,
+      "learning_rate": 0.0002,
+      "loss": 0.5714,
+      "step": 33610
+    },
+    {
+      "epoch": 5.43529221566567,
+      "grad_norm": 1.2198847532272339,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 33620
+    },
+    {
+      "epoch": 5.436908899846415,
+      "grad_norm": 1.0329595804214478,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 33630
+    },
+    {
+      "epoch": 5.43852558402716,
+      "grad_norm": 1.1075750589370728,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 33640
+    },
+    {
+      "epoch": 5.4401422682079055,
+      "grad_norm": 1.006342887878418,
+      "learning_rate": 0.0002,
+      "loss": 0.5295,
+      "step": 33650
+    },
+    {
+      "epoch": 5.441758952388651,
+      "grad_norm": 0.9179885983467102,
+      "learning_rate": 0.0002,
+      "loss": 0.5394,
+      "step": 33660
+    },
+    {
+      "epoch": 5.443375636569396,
+      "grad_norm": 1.2799493074417114,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 33670
+    },
+    {
+      "epoch": 5.444992320750141,
+      "grad_norm": 1.1153863668441772,
+      "learning_rate": 0.0002,
+      "loss": 0.5426,
+      "step": 33680
+    },
+    {
+      "epoch": 5.446609004930886,
+      "grad_norm": 1.0681028366088867,
+      "learning_rate": 0.0002,
+      "loss": 0.5087,
+      "step": 33690
+    },
+    {
+      "epoch": 5.448225689111632,
+      "grad_norm": 0.9788817167282104,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 33700
+    },
+    {
+      "epoch": 5.449842373292377,
+      "grad_norm": 0.8481608629226685,
+      "learning_rate": 0.0002,
+      "loss": 0.5308,
+      "step": 33710
+    },
+    {
+      "epoch": 5.451459057473123,
+      "grad_norm": 1.113756537437439,
+      "learning_rate": 0.0002,
+      "loss": 0.5225,
+      "step": 33720
+    },
+    {
+      "epoch": 5.453075741653868,
+      "grad_norm": 0.8425475358963013,
+      "learning_rate": 0.0002,
+      "loss": 0.5213,
+      "step": 33730
+    },
+    {
+      "epoch": 5.4546924258346134,
+      "grad_norm": 1.0852208137512207,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 33740
+    },
+    {
+      "epoch": 5.456309110015359,
+      "grad_norm": 1.1664748191833496,
+      "learning_rate": 0.0002,
+      "loss": 0.5535,
+      "step": 33750
+    },
+    {
+      "epoch": 5.457925794196104,
+      "grad_norm": 1.217241644859314,
+      "learning_rate": 0.0002,
+      "loss": 0.5419,
+      "step": 33760
+    },
+    {
+      "epoch": 5.459542478376849,
+      "grad_norm": 1.1572928428649902,
+      "learning_rate": 0.0002,
+      "loss": 0.5351,
+      "step": 33770
+    },
+    {
+      "epoch": 5.461159162557594,
+      "grad_norm": 1.0437318086624146,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 33780
+    },
+    {
+      "epoch": 5.46277584673834,
+      "grad_norm": 0.9807571768760681,
+      "learning_rate": 0.0002,
+      "loss": 0.5266,
+      "step": 33790
+    },
+    {
+      "epoch": 5.464392530919085,
+      "grad_norm": 1.1436342000961304,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 33800
+    },
+    {
+      "epoch": 5.46600921509983,
+      "grad_norm": 1.1004794836044312,
+      "learning_rate": 0.0002,
+      "loss": 0.5338,
+      "step": 33810
+    },
+    {
+      "epoch": 5.467625899280575,
+      "grad_norm": 1.2130268812179565,
+      "learning_rate": 0.0002,
+      "loss": 0.4868,
+      "step": 33820
+    },
+    {
+      "epoch": 5.4692425834613205,
+      "grad_norm": 1.3154419660568237,
+      "learning_rate": 0.0002,
+      "loss": 0.516,
+      "step": 33830
+    },
+    {
+      "epoch": 5.470859267642066,
+      "grad_norm": 0.7934383749961853,
+      "learning_rate": 0.0002,
+      "loss": 0.4934,
+      "step": 33840
+    },
+    {
+      "epoch": 5.472475951822812,
+      "grad_norm": 0.7838410139083862,
+      "learning_rate": 0.0002,
+      "loss": 0.5133,
+      "step": 33850
+    },
+    {
+      "epoch": 5.474092636003557,
+      "grad_norm": 1.0415139198303223,
+      "learning_rate": 0.0002,
+      "loss": 0.4926,
+      "step": 33860
+    },
+    {
+      "epoch": 5.475709320184302,
+      "grad_norm": 0.9213164448738098,
+      "learning_rate": 0.0002,
+      "loss": 0.5323,
+      "step": 33870
+    },
+    {
+      "epoch": 5.477326004365048,
+      "grad_norm": 1.0364776849746704,
+      "learning_rate": 0.0002,
+      "loss": 0.5125,
+      "step": 33880
+    },
+    {
+      "epoch": 5.478942688545793,
+      "grad_norm": 0.9994072318077087,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 33890
+    },
+    {
+      "epoch": 5.480559372726538,
+      "grad_norm": 1.196730136871338,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 33900
+    },
+    {
+      "epoch": 5.482176056907283,
+      "grad_norm": 0.9955780506134033,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 33910
+    },
+    {
+      "epoch": 5.4837927410880285,
+      "grad_norm": 1.168188214302063,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 33920
+    },
+    {
+      "epoch": 5.485409425268774,
+      "grad_norm": 1.1816450357437134,
+      "learning_rate": 0.0002,
+      "loss": 0.5548,
+      "step": 33930
+    },
+    {
+      "epoch": 5.487026109449519,
+      "grad_norm": 1.079715609550476,
+      "learning_rate": 0.0002,
+      "loss": 0.5535,
+      "step": 33940
+    },
+    {
+      "epoch": 5.488642793630264,
+      "grad_norm": 1.153850793838501,
+      "learning_rate": 0.0002,
+      "loss": 0.5262,
+      "step": 33950
+    },
+    {
+      "epoch": 5.490259477811009,
+      "grad_norm": 1.0207297801971436,
+      "learning_rate": 0.0002,
+      "loss": 0.5248,
+      "step": 33960
+    },
+    {
+      "epoch": 5.491876161991755,
+      "grad_norm": 1.1290855407714844,
+      "learning_rate": 0.0002,
+      "loss": 0.5142,
+      "step": 33970
+    },
+    {
+      "epoch": 5.4934928461725,
+      "grad_norm": 1.068058967590332,
+      "learning_rate": 0.0002,
+      "loss": 0.5168,
+      "step": 33980
+    },
+    {
+      "epoch": 5.495109530353245,
+      "grad_norm": 0.9789979457855225,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 33990
+    },
+    {
+      "epoch": 5.496726214533991,
+      "grad_norm": 0.9696692824363708,
+      "learning_rate": 0.0002,
+      "loss": 0.5113,
+      "step": 34000
+    },
+    {
+      "epoch": 5.4983428987147365,
+      "grad_norm": 1.0539981126785278,
+      "learning_rate": 0.0002,
+      "loss": 0.5413,
+      "step": 34010
+    },
+    {
+      "epoch": 5.499959582895482,
+      "grad_norm": 1.0249929428100586,
+      "learning_rate": 0.0002,
+      "loss": 0.5783,
+      "step": 34020
+    },
+    {
+      "epoch": 5.501576267076227,
+      "grad_norm": 0.9577504992485046,
+      "learning_rate": 0.0002,
+      "loss": 0.4888,
+      "step": 34030
+    },
+    {
+      "epoch": 5.503192951256972,
+      "grad_norm": 1.0963513851165771,
+      "learning_rate": 0.0002,
+      "loss": 0.5291,
+      "step": 34040
+    },
+    {
+      "epoch": 5.504809635437717,
+      "grad_norm": 0.8339345455169678,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 34050
+    },
+    {
+      "epoch": 5.506426319618463,
+      "grad_norm": 1.0138782262802124,
+      "learning_rate": 0.0002,
+      "loss": 0.5191,
+      "step": 34060
+    },
+    {
+      "epoch": 5.508043003799208,
+      "grad_norm": 1.0180109739303589,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 34070
+    },
+    {
+      "epoch": 5.509659687979953,
+      "grad_norm": 1.2790818214416504,
+      "learning_rate": 0.0002,
+      "loss": 0.5083,
+      "step": 34080
+    },
+    {
+      "epoch": 5.511276372160698,
+      "grad_norm": 1.428247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 34090
+    },
+    {
+      "epoch": 5.5128930563414436,
+      "grad_norm": 1.0926059484481812,
+      "learning_rate": 0.0002,
+      "loss": 0.5291,
+      "step": 34100
+    },
+    {
+      "epoch": 5.514509740522189,
+      "grad_norm": 1.2353343963623047,
+      "learning_rate": 0.0002,
+      "loss": 0.5665,
+      "step": 34110
+    },
+    {
+      "epoch": 5.516126424702934,
+      "grad_norm": 0.935587465763092,
+      "learning_rate": 0.0002,
+      "loss": 0.5331,
+      "step": 34120
+    },
+    {
+      "epoch": 5.517743108883679,
+      "grad_norm": 0.9767586588859558,
+      "learning_rate": 0.0002,
+      "loss": 0.5512,
+      "step": 34130
+    },
+    {
+      "epoch": 5.5193597930644245,
+      "grad_norm": 1.1660610437393188,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 34140
+    },
+    {
+      "epoch": 5.520976477245171,
+      "grad_norm": 0.9828870892524719,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 34150
+    },
+    {
+      "epoch": 5.522593161425916,
+      "grad_norm": 1.0097278356552124,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 34160
+    },
+    {
+      "epoch": 5.524209845606661,
+      "grad_norm": 1.1766167879104614,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 34170
+    },
+    {
+      "epoch": 5.525826529787406,
+      "grad_norm": 0.982292115688324,
+      "learning_rate": 0.0002,
+      "loss": 0.5258,
+      "step": 34180
+    },
+    {
+      "epoch": 5.5274432139681515,
+      "grad_norm": 1.0744609832763672,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 34190
+    },
+    {
+      "epoch": 5.529059898148897,
+      "grad_norm": 1.3831160068511963,
+      "learning_rate": 0.0002,
+      "loss": 0.5469,
+      "step": 34200
+    },
+    {
+      "epoch": 5.530676582329642,
+      "grad_norm": 1.074771761894226,
+      "learning_rate": 0.0002,
+      "loss": 0.5819,
+      "step": 34210
+    },
+    {
+      "epoch": 5.532293266510387,
+      "grad_norm": 1.016652226448059,
+      "learning_rate": 0.0002,
+      "loss": 0.5399,
+      "step": 34220
+    },
+    {
+      "epoch": 5.5339099506911325,
+      "grad_norm": 1.2231552600860596,
+      "learning_rate": 0.0002,
+      "loss": 0.5158,
+      "step": 34230
+    },
+    {
+      "epoch": 5.535526634871878,
+      "grad_norm": 0.8051198720932007,
+      "learning_rate": 0.0002,
+      "loss": 0.5091,
+      "step": 34240
+    },
+    {
+      "epoch": 5.537143319052623,
+      "grad_norm": 1.1779674291610718,
+      "learning_rate": 0.0002,
+      "loss": 0.5583,
+      "step": 34250
+    },
+    {
+      "epoch": 5.538760003233368,
+      "grad_norm": 1.2468291521072388,
+      "learning_rate": 0.0002,
+      "loss": 0.5044,
+      "step": 34260
+    },
+    {
+      "epoch": 5.540376687414113,
+      "grad_norm": 1.14818274974823,
+      "learning_rate": 0.0002,
+      "loss": 0.523,
+      "step": 34270
+    },
+    {
+      "epoch": 5.541993371594859,
+      "grad_norm": 1.2362616062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.5375,
+      "step": 34280
+    },
+    {
+      "epoch": 5.543610055775604,
+      "grad_norm": 1.0206977128982544,
+      "learning_rate": 0.0002,
+      "loss": 0.4996,
+      "step": 34290
+    },
+    {
+      "epoch": 5.54522673995635,
+      "grad_norm": 1.2018457651138306,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 34300
+    },
+    {
+      "epoch": 5.546843424137095,
+      "grad_norm": 1.0349043607711792,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 34310
+    },
+    {
+      "epoch": 5.54846010831784,
+      "grad_norm": 1.2022006511688232,
+      "learning_rate": 0.0002,
+      "loss": 0.5231,
+      "step": 34320
+    },
+    {
+      "epoch": 5.550076792498586,
+      "grad_norm": 1.0810624361038208,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 34330
+    },
+    {
+      "epoch": 5.551693476679331,
+      "grad_norm": 1.3297529220581055,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 34340
+    },
+    {
+      "epoch": 5.553310160860076,
+      "grad_norm": 0.9722549915313721,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 34350
+    },
+    {
+      "epoch": 5.554926845040821,
+      "grad_norm": 0.9903425574302673,
+      "learning_rate": 0.0002,
+      "loss": 0.4823,
+      "step": 34360
+    },
+    {
+      "epoch": 5.556543529221567,
+      "grad_norm": 0.9568067789077759,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 34370
+    },
+    {
+      "epoch": 5.558160213402312,
+      "grad_norm": 1.113870620727539,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 34380
+    },
+    {
+      "epoch": 5.559776897583057,
+      "grad_norm": 1.0557632446289062,
+      "learning_rate": 0.0002,
+      "loss": 0.5278,
+      "step": 34390
+    },
+    {
+      "epoch": 5.561393581763802,
+      "grad_norm": 0.9615673422813416,
+      "learning_rate": 0.0002,
+      "loss": 0.5501,
+      "step": 34400
+    },
+    {
+      "epoch": 5.5630102659445475,
+      "grad_norm": 0.9536027312278748,
+      "learning_rate": 0.0002,
+      "loss": 0.5066,
+      "step": 34410
+    },
+    {
+      "epoch": 5.564626950125293,
+      "grad_norm": 0.8808749318122864,
+      "learning_rate": 0.0002,
+      "loss": 0.4949,
+      "step": 34420
+    },
+    {
+      "epoch": 5.566243634306038,
+      "grad_norm": 1.286132574081421,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 34430
+    },
+    {
+      "epoch": 5.567860318486783,
+      "grad_norm": 1.259644865989685,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 34440
+    },
+    {
+      "epoch": 5.569477002667529,
+      "grad_norm": 0.9920216798782349,
+      "learning_rate": 0.0002,
+      "loss": 0.4922,
+      "step": 34450
+    },
+    {
+      "epoch": 5.5710936868482746,
+      "grad_norm": 1.182926893234253,
+      "learning_rate": 0.0002,
+      "loss": 0.5527,
+      "step": 34460
+    },
+    {
+      "epoch": 5.57271037102902,
+      "grad_norm": 1.1434749364852905,
+      "learning_rate": 0.0002,
+      "loss": 0.5185,
+      "step": 34470
+    },
+    {
+      "epoch": 5.574327055209765,
+      "grad_norm": 1.2420979738235474,
+      "learning_rate": 0.0002,
+      "loss": 0.5256,
+      "step": 34480
+    },
+    {
+      "epoch": 5.57594373939051,
+      "grad_norm": 0.9338384866714478,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 34490
+    },
+    {
+      "epoch": 5.5775604235712555,
+      "grad_norm": 1.0196425914764404,
+      "learning_rate": 0.0002,
+      "loss": 0.5634,
+      "step": 34500
+    },
+    {
+      "epoch": 5.579177107752001,
+      "grad_norm": 0.9586997032165527,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 34510
+    },
+    {
+      "epoch": 5.580793791932746,
+      "grad_norm": 1.2409086227416992,
+      "learning_rate": 0.0002,
+      "loss": 0.5336,
+      "step": 34520
+    },
+    {
+      "epoch": 5.582410476113491,
+      "grad_norm": 1.1483757495880127,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 34530
+    },
+    {
+      "epoch": 5.584027160294236,
+      "grad_norm": 1.1624305248260498,
+      "learning_rate": 0.0002,
+      "loss": 0.5325,
+      "step": 34540
+    },
+    {
+      "epoch": 5.585643844474982,
+      "grad_norm": 1.2635223865509033,
+      "learning_rate": 0.0002,
+      "loss": 0.5342,
+      "step": 34550
+    },
+    {
+      "epoch": 5.587260528655727,
+      "grad_norm": 0.9824051856994629,
+      "learning_rate": 0.0002,
+      "loss": 0.4924,
+      "step": 34560
+    },
+    {
+      "epoch": 5.588877212836472,
+      "grad_norm": 1.0858620405197144,
+      "learning_rate": 0.0002,
+      "loss": 0.5395,
+      "step": 34570
+    },
+    {
+      "epoch": 5.590493897017217,
+      "grad_norm": 1.1452655792236328,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 34580
+    },
+    {
+      "epoch": 5.592110581197963,
+      "grad_norm": 1.110610842704773,
+      "learning_rate": 0.0002,
+      "loss": 0.5746,
+      "step": 34590
+    },
+    {
+      "epoch": 5.593727265378709,
+      "grad_norm": 0.9976194500923157,
+      "learning_rate": 0.0002,
+      "loss": 0.5285,
+      "step": 34600
+    },
+    {
+      "epoch": 5.595343949559454,
+      "grad_norm": 1.0698920488357544,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 34610
+    },
+    {
+      "epoch": 5.596960633740199,
+      "grad_norm": 1.1505171060562134,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 34620
+    },
+    {
+      "epoch": 5.598577317920944,
+      "grad_norm": 1.1014643907546997,
+      "learning_rate": 0.0002,
+      "loss": 0.5471,
+      "step": 34630
+    },
+    {
+      "epoch": 5.60019400210169,
+      "grad_norm": 0.915595293045044,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 34640
+    },
+    {
+      "epoch": 5.601810686282435,
+      "grad_norm": 1.1856765747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 34650
+    },
+    {
+      "epoch": 5.60342737046318,
+      "grad_norm": 1.1357687711715698,
+      "learning_rate": 0.0002,
+      "loss": 0.5502,
+      "step": 34660
+    },
+    {
+      "epoch": 5.605044054643925,
+      "grad_norm": 1.0232492685317993,
+      "learning_rate": 0.0002,
+      "loss": 0.5034,
+      "step": 34670
+    },
+    {
+      "epoch": 5.6066607388246705,
+      "grad_norm": 0.9375017881393433,
+      "learning_rate": 0.0002,
+      "loss": 0.5357,
+      "step": 34680
+    },
+    {
+      "epoch": 5.608277423005416,
+      "grad_norm": 1.0796529054641724,
+      "learning_rate": 0.0002,
+      "loss": 0.5518,
+      "step": 34690
+    },
+    {
+      "epoch": 5.609894107186161,
+      "grad_norm": 1.1383336782455444,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 34700
+    },
+    {
+      "epoch": 5.611510791366906,
+      "grad_norm": 1.0248544216156006,
+      "learning_rate": 0.0002,
+      "loss": 0.5477,
+      "step": 34710
+    },
+    {
+      "epoch": 5.6131274755476515,
+      "grad_norm": 1.0986040830612183,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 34720
+    },
+    {
+      "epoch": 5.614744159728397,
+      "grad_norm": 1.2689568996429443,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 34730
+    },
+    {
+      "epoch": 5.616360843909142,
+      "grad_norm": 1.4044264554977417,
+      "learning_rate": 0.0002,
+      "loss": 0.5136,
+      "step": 34740
+    },
+    {
+      "epoch": 5.617977528089888,
+      "grad_norm": 1.2084474563598633,
+      "learning_rate": 0.0002,
+      "loss": 0.5699,
+      "step": 34750
+    },
+    {
+      "epoch": 5.619594212270633,
+      "grad_norm": 1.061248540878296,
+      "learning_rate": 0.0002,
+      "loss": 0.5377,
+      "step": 34760
+    },
+    {
+      "epoch": 5.6212108964513785,
+      "grad_norm": 1.0220764875411987,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 34770
+    },
+    {
+      "epoch": 5.622827580632124,
+      "grad_norm": 1.0859092473983765,
+      "learning_rate": 0.0002,
+      "loss": 0.54,
+      "step": 34780
+    },
+    {
+      "epoch": 5.624444264812869,
+      "grad_norm": 0.9049732089042664,
+      "learning_rate": 0.0002,
+      "loss": 0.5308,
+      "step": 34790
+    },
+    {
+      "epoch": 5.626060948993614,
+      "grad_norm": 1.2103937864303589,
+      "learning_rate": 0.0002,
+      "loss": 0.5433,
+      "step": 34800
+    },
+    {
+      "epoch": 5.627677633174359,
+      "grad_norm": 0.9854230284690857,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 34810
+    },
+    {
+      "epoch": 5.629294317355105,
+      "grad_norm": 0.9316635131835938,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 34820
+    },
+    {
+      "epoch": 5.63091100153585,
+      "grad_norm": 1.105296015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.5393,
+      "step": 34830
+    },
+    {
+      "epoch": 5.632527685716595,
+      "grad_norm": 0.993383526802063,
+      "learning_rate": 0.0002,
+      "loss": 0.5527,
+      "step": 34840
+    },
+    {
+      "epoch": 5.63414436989734,
+      "grad_norm": 1.1544116735458374,
+      "learning_rate": 0.0002,
+      "loss": 0.5375,
+      "step": 34850
+    },
+    {
+      "epoch": 5.635761054078086,
+      "grad_norm": 1.284475326538086,
+      "learning_rate": 0.0002,
+      "loss": 0.5448,
+      "step": 34860
+    },
+    {
+      "epoch": 5.637377738258831,
+      "grad_norm": 1.121997594833374,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 34870
+    },
+    {
+      "epoch": 5.638994422439576,
+      "grad_norm": 1.213040828704834,
+      "learning_rate": 0.0002,
+      "loss": 0.5335,
+      "step": 34880
+    },
+    {
+      "epoch": 5.640611106620321,
+      "grad_norm": 1.23222017288208,
+      "learning_rate": 0.0002,
+      "loss": 0.5623,
+      "step": 34890
+    },
+    {
+      "epoch": 5.642227790801067,
+      "grad_norm": 0.9793637990951538,
+      "learning_rate": 0.0002,
+      "loss": 0.5622,
+      "step": 34900
+    },
+    {
+      "epoch": 5.643844474981813,
+      "grad_norm": 1.38919997215271,
+      "learning_rate": 0.0002,
+      "loss": 0.5405,
+      "step": 34910
+    },
+    {
+      "epoch": 5.645461159162558,
+      "grad_norm": 0.8390951156616211,
+      "learning_rate": 0.0002,
+      "loss": 0.5007,
+      "step": 34920
+    },
+    {
+      "epoch": 5.647077843343303,
+      "grad_norm": 0.9465909004211426,
+      "learning_rate": 0.0002,
+      "loss": 0.5974,
+      "step": 34930
+    },
+    {
+      "epoch": 5.648694527524048,
+      "grad_norm": 1.066957712173462,
+      "learning_rate": 0.0002,
+      "loss": 0.5264,
+      "step": 34940
+    },
+    {
+      "epoch": 5.650311211704794,
+      "grad_norm": 0.9842154383659363,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 34950
+    },
+    {
+      "epoch": 5.651927895885539,
+      "grad_norm": 1.1766440868377686,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 34960
+    },
+    {
+      "epoch": 5.653544580066284,
+      "grad_norm": 0.9061306118965149,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 34970
+    },
+    {
+      "epoch": 5.655161264247029,
+      "grad_norm": 1.2941309213638306,
+      "learning_rate": 0.0002,
+      "loss": 0.5446,
+      "step": 34980
+    },
+    {
+      "epoch": 5.6567779484277745,
+      "grad_norm": 0.9741247892379761,
+      "learning_rate": 0.0002,
+      "loss": 0.5704,
+      "step": 34990
+    },
+    {
+      "epoch": 5.65839463260852,
+      "grad_norm": 1.0784187316894531,
+      "learning_rate": 0.0002,
+      "loss": 0.5152,
+      "step": 35000
+    },
+    {
+      "epoch": 5.660011316789265,
+      "grad_norm": 0.937889814376831,
+      "learning_rate": 0.0002,
+      "loss": 0.5363,
+      "step": 35010
+    },
+    {
+      "epoch": 5.66162800097001,
+      "grad_norm": 0.9667879939079285,
+      "learning_rate": 0.0002,
+      "loss": 0.5019,
+      "step": 35020
+    },
+    {
+      "epoch": 5.663244685150756,
+      "grad_norm": 1.0554876327514648,
+      "learning_rate": 0.0002,
+      "loss": 0.5209,
+      "step": 35030
+    },
+    {
+      "epoch": 5.664861369331501,
+      "grad_norm": 1.2030539512634277,
+      "learning_rate": 0.0002,
+      "loss": 0.523,
+      "step": 35040
+    },
+    {
+      "epoch": 5.666478053512247,
+      "grad_norm": 1.0849953889846802,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 35050
+    },
+    {
+      "epoch": 5.668094737692992,
+      "grad_norm": 1.1598973274230957,
+      "learning_rate": 0.0002,
+      "loss": 0.5747,
+      "step": 35060
+    },
+    {
+      "epoch": 5.669711421873737,
+      "grad_norm": 1.0233359336853027,
+      "learning_rate": 0.0002,
+      "loss": 0.5488,
+      "step": 35070
+    },
+    {
+      "epoch": 5.6713281060544825,
+      "grad_norm": 1.1124799251556396,
+      "learning_rate": 0.0002,
+      "loss": 0.5409,
+      "step": 35080
+    },
+    {
+      "epoch": 5.672944790235228,
+      "grad_norm": 1.2351475954055786,
+      "learning_rate": 0.0002,
+      "loss": 0.5578,
+      "step": 35090
+    },
+    {
+      "epoch": 5.674561474415973,
+      "grad_norm": 1.0240728855133057,
+      "learning_rate": 0.0002,
+      "loss": 0.5638,
+      "step": 35100
+    },
+    {
+      "epoch": 5.676178158596718,
+      "grad_norm": 1.0223692655563354,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 35110
+    },
+    {
+      "epoch": 5.677794842777463,
+      "grad_norm": 1.4569132328033447,
+      "learning_rate": 0.0002,
+      "loss": 0.524,
+      "step": 35120
+    },
+    {
+      "epoch": 5.679411526958209,
+      "grad_norm": 0.8983587026596069,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 35130
+    },
+    {
+      "epoch": 5.681028211138954,
+      "grad_norm": 1.0775383710861206,
+      "learning_rate": 0.0002,
+      "loss": 0.5439,
+      "step": 35140
+    },
+    {
+      "epoch": 5.682644895319699,
+      "grad_norm": 0.9800270795822144,
+      "learning_rate": 0.0002,
+      "loss": 0.5289,
+      "step": 35150
+    },
+    {
+      "epoch": 5.684261579500444,
+      "grad_norm": 0.9858237504959106,
+      "learning_rate": 0.0002,
+      "loss": 0.533,
+      "step": 35160
+    },
+    {
+      "epoch": 5.6858782636811895,
+      "grad_norm": 1.031087040901184,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 35170
+    },
+    {
+      "epoch": 5.687494947861936,
+      "grad_norm": 1.0294365882873535,
+      "learning_rate": 0.0002,
+      "loss": 0.5528,
+      "step": 35180
+    },
+    {
+      "epoch": 5.68911163204268,
+      "grad_norm": 1.108144760131836,
+      "learning_rate": 0.0002,
+      "loss": 0.5581,
+      "step": 35190
+    },
+    {
+      "epoch": 5.690728316223426,
+      "grad_norm": 1.0813100337982178,
+      "learning_rate": 0.0002,
+      "loss": 0.5373,
+      "step": 35200
+    },
+    {
+      "epoch": 5.692345000404171,
+      "grad_norm": 1.3146867752075195,
+      "learning_rate": 0.0002,
+      "loss": 0.5429,
+      "step": 35210
+    },
+    {
+      "epoch": 5.693961684584917,
+      "grad_norm": 1.16780424118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5297,
+      "step": 35220
+    },
+    {
+      "epoch": 5.695578368765662,
+      "grad_norm": 0.9929125905036926,
+      "learning_rate": 0.0002,
+      "loss": 0.577,
+      "step": 35230
+    },
+    {
+      "epoch": 5.697195052946407,
+      "grad_norm": 0.9049441814422607,
+      "learning_rate": 0.0002,
+      "loss": 0.5441,
+      "step": 35240
+    },
+    {
+      "epoch": 5.698811737127152,
+      "grad_norm": 0.9768866300582886,
+      "learning_rate": 0.0002,
+      "loss": 0.5349,
+      "step": 35250
+    },
+    {
+      "epoch": 5.7004284213078975,
+      "grad_norm": 0.8306029438972473,
+      "learning_rate": 0.0002,
+      "loss": 0.542,
+      "step": 35260
+    },
+    {
+      "epoch": 5.702045105488643,
+      "grad_norm": 0.8417280316352844,
+      "learning_rate": 0.0002,
+      "loss": 0.4771,
+      "step": 35270
+    },
+    {
+      "epoch": 5.703661789669388,
+      "grad_norm": 0.9954485893249512,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 35280
+    },
+    {
+      "epoch": 5.705278473850133,
+      "grad_norm": 1.2417993545532227,
+      "learning_rate": 0.0002,
+      "loss": 0.5469,
+      "step": 35290
+    },
+    {
+      "epoch": 5.706895158030878,
+      "grad_norm": 1.1696544885635376,
+      "learning_rate": 0.0002,
+      "loss": 0.5275,
+      "step": 35300
+    },
+    {
+      "epoch": 5.708511842211624,
+      "grad_norm": 1.2424817085266113,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 35310
+    },
+    {
+      "epoch": 5.710128526392369,
+      "grad_norm": 1.1791106462478638,
+      "learning_rate": 0.0002,
+      "loss": 0.5595,
+      "step": 35320
+    },
+    {
+      "epoch": 5.711745210573115,
+      "grad_norm": 1.202181339263916,
+      "learning_rate": 0.0002,
+      "loss": 0.5076,
+      "step": 35330
+    },
+    {
+      "epoch": 5.713361894753859,
+      "grad_norm": 1.1006861925125122,
+      "learning_rate": 0.0002,
+      "loss": 0.5847,
+      "step": 35340
+    },
+    {
+      "epoch": 5.7149785789346055,
+      "grad_norm": 1.0918344259262085,
+      "learning_rate": 0.0002,
+      "loss": 0.5627,
+      "step": 35350
+    },
+    {
+      "epoch": 5.716595263115351,
+      "grad_norm": 1.0427305698394775,
+      "learning_rate": 0.0002,
+      "loss": 0.5677,
+      "step": 35360
+    },
+    {
+      "epoch": 5.718211947296096,
+      "grad_norm": 1.0818872451782227,
+      "learning_rate": 0.0002,
+      "loss": 0.5288,
+      "step": 35370
+    },
+    {
+      "epoch": 5.719828631476841,
+      "grad_norm": 1.186006784439087,
+      "learning_rate": 0.0002,
+      "loss": 0.5296,
+      "step": 35380
+    },
+    {
+      "epoch": 5.721445315657586,
+      "grad_norm": 1.2073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 35390
+    },
+    {
+      "epoch": 5.723061999838332,
+      "grad_norm": 1.065338134765625,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 35400
+    },
+    {
+      "epoch": 5.724678684019077,
+      "grad_norm": 0.9448973536491394,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 35410
+    },
+    {
+      "epoch": 5.726295368199822,
+      "grad_norm": 1.1487499475479126,
+      "learning_rate": 0.0002,
+      "loss": 0.5276,
+      "step": 35420
+    },
+    {
+      "epoch": 5.727912052380567,
+      "grad_norm": 1.1334216594696045,
+      "learning_rate": 0.0002,
+      "loss": 0.5435,
+      "step": 35430
+    },
+    {
+      "epoch": 5.729528736561313,
+      "grad_norm": 1.1932826042175293,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 35440
+    },
+    {
+      "epoch": 5.731145420742058,
+      "grad_norm": 1.2615786790847778,
+      "learning_rate": 0.0002,
+      "loss": 0.5502,
+      "step": 35450
+    },
+    {
+      "epoch": 5.732762104922803,
+      "grad_norm": 1.2803694009780884,
+      "learning_rate": 0.0002,
+      "loss": 0.5612,
+      "step": 35460
+    },
+    {
+      "epoch": 5.734378789103548,
+      "grad_norm": 0.9271906614303589,
+      "learning_rate": 0.0002,
+      "loss": 0.5458,
+      "step": 35470
+    },
+    {
+      "epoch": 5.735995473284294,
+      "grad_norm": 1.0958917140960693,
+      "learning_rate": 0.0002,
+      "loss": 0.5342,
+      "step": 35480
+    },
+    {
+      "epoch": 5.737612157465039,
+      "grad_norm": 1.1072784662246704,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 35490
+    },
+    {
+      "epoch": 5.739228841645785,
+      "grad_norm": 1.1641002893447876,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 35500
+    },
+    {
+      "epoch": 5.74084552582653,
+      "grad_norm": 1.0246447324752808,
+      "learning_rate": 0.0002,
+      "loss": 0.5252,
+      "step": 35510
+    },
+    {
+      "epoch": 5.742462210007275,
+      "grad_norm": 1.032474398612976,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 35520
+    },
+    {
+      "epoch": 5.7440788941880205,
+      "grad_norm": 1.1600854396820068,
+      "learning_rate": 0.0002,
+      "loss": 0.4965,
+      "step": 35530
+    },
+    {
+      "epoch": 5.745695578368766,
+      "grad_norm": 1.0686054229736328,
+      "learning_rate": 0.0002,
+      "loss": 0.5543,
+      "step": 35540
+    },
+    {
+      "epoch": 5.747312262549511,
+      "grad_norm": 1.2314637899398804,
+      "learning_rate": 0.0002,
+      "loss": 0.5706,
+      "step": 35550
+    },
+    {
+      "epoch": 5.748928946730256,
+      "grad_norm": 0.922134280204773,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 35560
+    },
+    {
+      "epoch": 5.7505456309110015,
+      "grad_norm": 0.933043360710144,
+      "learning_rate": 0.0002,
+      "loss": 0.5495,
+      "step": 35570
+    },
+    {
+      "epoch": 5.752162315091747,
+      "grad_norm": 1.1911931037902832,
+      "learning_rate": 0.0002,
+      "loss": 0.5007,
+      "step": 35580
+    },
+    {
+      "epoch": 5.753778999272492,
+      "grad_norm": 0.8984857797622681,
+      "learning_rate": 0.0002,
+      "loss": 0.5244,
+      "step": 35590
+    },
+    {
+      "epoch": 5.755395683453237,
+      "grad_norm": 0.9495107531547546,
+      "learning_rate": 0.0002,
+      "loss": 0.5493,
+      "step": 35600
+    },
+    {
+      "epoch": 5.757012367633982,
+      "grad_norm": 1.2805472612380981,
+      "learning_rate": 0.0002,
+      "loss": 0.5326,
+      "step": 35610
+    },
+    {
+      "epoch": 5.758629051814728,
+      "grad_norm": 1.1236625909805298,
+      "learning_rate": 0.0002,
+      "loss": 0.5276,
+      "step": 35620
+    },
+    {
+      "epoch": 5.760245735995474,
+      "grad_norm": 1.0552798509597778,
+      "learning_rate": 0.0002,
+      "loss": 0.6102,
+      "step": 35630
+    },
+    {
+      "epoch": 5.761862420176218,
+      "grad_norm": 1.119909644126892,
+      "learning_rate": 0.0002,
+      "loss": 0.5479,
+      "step": 35640
+    },
+    {
+      "epoch": 5.763479104356964,
+      "grad_norm": 0.8786116242408752,
+      "learning_rate": 0.0002,
+      "loss": 0.5282,
+      "step": 35650
+    },
+    {
+      "epoch": 5.765095788537709,
+      "grad_norm": 1.2417117357254028,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 35660
+    },
+    {
+      "epoch": 5.766712472718455,
+      "grad_norm": 1.255200982093811,
+      "learning_rate": 0.0002,
+      "loss": 0.537,
+      "step": 35670
+    },
+    {
+      "epoch": 5.7683291568992,
+      "grad_norm": 1.0611358880996704,
+      "learning_rate": 0.0002,
+      "loss": 0.5308,
+      "step": 35680
+    },
+    {
+      "epoch": 5.769945841079945,
+      "grad_norm": 1.1443911790847778,
+      "learning_rate": 0.0002,
+      "loss": 0.5614,
+      "step": 35690
+    },
+    {
+      "epoch": 5.77156252526069,
+      "grad_norm": 1.1437989473342896,
+      "learning_rate": 0.0002,
+      "loss": 0.5386,
+      "step": 35700
+    },
+    {
+      "epoch": 5.773179209441436,
+      "grad_norm": 1.1375046968460083,
+      "learning_rate": 0.0002,
+      "loss": 0.537,
+      "step": 35710
+    },
+    {
+      "epoch": 5.774795893622181,
+      "grad_norm": 1.0777729749679565,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 35720
+    },
+    {
+      "epoch": 5.776412577802926,
+      "grad_norm": 1.1160215139389038,
+      "learning_rate": 0.0002,
+      "loss": 0.5521,
+      "step": 35730
+    },
+    {
+      "epoch": 5.778029261983671,
+      "grad_norm": 1.1268514394760132,
+      "learning_rate": 0.0002,
+      "loss": 0.5569,
+      "step": 35740
+    },
+    {
+      "epoch": 5.7796459461644165,
+      "grad_norm": 1.2752262353897095,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 35750
+    },
+    {
+      "epoch": 5.781262630345162,
+      "grad_norm": 1.0416184663772583,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 35760
+    },
+    {
+      "epoch": 5.782879314525907,
+      "grad_norm": 1.0622444152832031,
+      "learning_rate": 0.0002,
+      "loss": 0.5438,
+      "step": 35770
+    },
+    {
+      "epoch": 5.784495998706653,
+      "grad_norm": 1.1217877864837646,
+      "learning_rate": 0.0002,
+      "loss": 0.5268,
+      "step": 35780
+    },
+    {
+      "epoch": 5.786112682887398,
+      "grad_norm": 0.9363139867782593,
+      "learning_rate": 0.0002,
+      "loss": 0.5225,
+      "step": 35790
+    },
+    {
+      "epoch": 5.787729367068144,
+      "grad_norm": 0.96628737449646,
+      "learning_rate": 0.0002,
+      "loss": 0.5524,
+      "step": 35800
+    },
+    {
+      "epoch": 5.789346051248889,
+      "grad_norm": 0.9572572112083435,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 35810
+    },
+    {
+      "epoch": 5.790962735429634,
+      "grad_norm": 0.938724935054779,
+      "learning_rate": 0.0002,
+      "loss": 0.5615,
+      "step": 35820
+    },
+    {
+      "epoch": 5.792579419610379,
+      "grad_norm": 1.3314417600631714,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 35830
+    },
+    {
+      "epoch": 5.7941961037911245,
+      "grad_norm": 1.0097602605819702,
+      "learning_rate": 0.0002,
+      "loss": 0.5441,
+      "step": 35840
+    },
+    {
+      "epoch": 5.79581278797187,
+      "grad_norm": 1.1265122890472412,
+      "learning_rate": 0.0002,
+      "loss": 0.591,
+      "step": 35850
+    },
+    {
+      "epoch": 5.797429472152615,
+      "grad_norm": 1.2191909551620483,
+      "learning_rate": 0.0002,
+      "loss": 0.5333,
+      "step": 35860
+    },
+    {
+      "epoch": 5.79904615633336,
+      "grad_norm": 0.9690808057785034,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 35870
+    },
+    {
+      "epoch": 5.800662840514105,
+      "grad_norm": 1.0871665477752686,
+      "learning_rate": 0.0002,
+      "loss": 0.5425,
+      "step": 35880
+    },
+    {
+      "epoch": 5.802279524694851,
+      "grad_norm": 1.1093597412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.5602,
+      "step": 35890
+    },
+    {
+      "epoch": 5.803896208875596,
+      "grad_norm": 1.2434282302856445,
+      "learning_rate": 0.0002,
+      "loss": 0.5475,
+      "step": 35900
+    },
+    {
+      "epoch": 5.805512893056341,
+      "grad_norm": 1.2933623790740967,
+      "learning_rate": 0.0002,
+      "loss": 0.5288,
+      "step": 35910
+    },
+    {
+      "epoch": 5.807129577237086,
+      "grad_norm": 1.0005441904067993,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 35920
+    },
+    {
+      "epoch": 5.8087462614178325,
+      "grad_norm": 1.2373108863830566,
+      "learning_rate": 0.0002,
+      "loss": 0.5318,
+      "step": 35930
+    },
+    {
+      "epoch": 5.810362945598578,
+      "grad_norm": 1.2622692584991455,
+      "learning_rate": 0.0002,
+      "loss": 0.5413,
+      "step": 35940
+    },
+    {
+      "epoch": 5.811979629779323,
+      "grad_norm": 1.0112963914871216,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 35950
+    },
+    {
+      "epoch": 5.813596313960068,
+      "grad_norm": 1.050572395324707,
+      "learning_rate": 0.0002,
+      "loss": 0.5115,
+      "step": 35960
+    },
+    {
+      "epoch": 5.815212998140813,
+      "grad_norm": 0.9774560928344727,
+      "learning_rate": 0.0002,
+      "loss": 0.5288,
+      "step": 35970
+    },
+    {
+      "epoch": 5.816829682321559,
+      "grad_norm": 1.19438898563385,
+      "learning_rate": 0.0002,
+      "loss": 0.585,
+      "step": 35980
+    },
+    {
+      "epoch": 5.818446366502304,
+      "grad_norm": 1.0267130136489868,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 35990
+    },
+    {
+      "epoch": 5.820063050683049,
+      "grad_norm": 0.9813851714134216,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 36000
+    },
+    {
+      "epoch": 5.821679734863794,
+      "grad_norm": 0.9177457094192505,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 36010
+    },
+    {
+      "epoch": 5.8232964190445395,
+      "grad_norm": 1.0020731687545776,
+      "learning_rate": 0.0002,
+      "loss": 0.5453,
+      "step": 36020
+    },
+    {
+      "epoch": 5.824913103225285,
+      "grad_norm": 1.073222041130066,
+      "learning_rate": 0.0002,
+      "loss": 0.5646,
+      "step": 36030
+    },
+    {
+      "epoch": 5.82652978740603,
+      "grad_norm": 1.016337513923645,
+      "learning_rate": 0.0002,
+      "loss": 0.5539,
+      "step": 36040
+    },
+    {
+      "epoch": 5.828146471586775,
+      "grad_norm": 1.267364263534546,
+      "learning_rate": 0.0002,
+      "loss": 0.5592,
+      "step": 36050
+    },
+    {
+      "epoch": 5.8297631557675205,
+      "grad_norm": 1.2730127573013306,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 36060
+    },
+    {
+      "epoch": 5.831379839948266,
+      "grad_norm": 1.108442783355713,
+      "learning_rate": 0.0002,
+      "loss": 0.5247,
+      "step": 36070
+    },
+    {
+      "epoch": 5.832996524129012,
+      "grad_norm": 1.198072075843811,
+      "learning_rate": 0.0002,
+      "loss": 0.5103,
+      "step": 36080
+    },
+    {
+      "epoch": 5.834613208309757,
+      "grad_norm": 1.0458786487579346,
+      "learning_rate": 0.0002,
+      "loss": 0.5479,
+      "step": 36090
+    },
+    {
+      "epoch": 5.836229892490502,
+      "grad_norm": 0.9096664786338806,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 36100
+    },
+    {
+      "epoch": 5.8378465766712475,
+      "grad_norm": 0.9957793951034546,
+      "learning_rate": 0.0002,
+      "loss": 0.5602,
+      "step": 36110
+    },
+    {
+      "epoch": 5.839463260851993,
+      "grad_norm": 1.3693058490753174,
+      "learning_rate": 0.0002,
+      "loss": 0.5799,
+      "step": 36120
+    },
+    {
+      "epoch": 5.841079945032738,
+      "grad_norm": 1.268608808517456,
+      "learning_rate": 0.0002,
+      "loss": 0.5425,
+      "step": 36130
+    },
+    {
+      "epoch": 5.842696629213483,
+      "grad_norm": 0.8516020178794861,
+      "learning_rate": 0.0002,
+      "loss": 0.5653,
+      "step": 36140
+    },
+    {
+      "epoch": 5.844313313394228,
+      "grad_norm": 0.90385502576828,
+      "learning_rate": 0.0002,
+      "loss": 0.5475,
+      "step": 36150
+    },
+    {
+      "epoch": 5.845929997574974,
+      "grad_norm": 1.0910571813583374,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 36160
+    },
+    {
+      "epoch": 5.847546681755719,
+      "grad_norm": 0.9417795538902283,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 36170
+    },
+    {
+      "epoch": 5.849163365936464,
+      "grad_norm": 1.0027360916137695,
+      "learning_rate": 0.0002,
+      "loss": 0.5784,
+      "step": 36180
+    },
+    {
+      "epoch": 5.850780050117209,
+      "grad_norm": 1.1480516195297241,
+      "learning_rate": 0.0002,
+      "loss": 0.5423,
+      "step": 36190
+    },
+    {
+      "epoch": 5.852396734297955,
+      "grad_norm": 1.2431457042694092,
+      "learning_rate": 0.0002,
+      "loss": 0.5517,
+      "step": 36200
+    },
+    {
+      "epoch": 5.8540134184787,
+      "grad_norm": 1.091465950012207,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 36210
+    },
+    {
+      "epoch": 5.855630102659445,
+      "grad_norm": 0.9693930745124817,
+      "learning_rate": 0.0002,
+      "loss": 0.53,
+      "step": 36220
+    },
+    {
+      "epoch": 5.857246786840191,
+      "grad_norm": 0.9937465190887451,
+      "learning_rate": 0.0002,
+      "loss": 0.5453,
+      "step": 36230
+    },
+    {
+      "epoch": 5.858863471020936,
+      "grad_norm": 1.0731011629104614,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 36240
+    },
+    {
+      "epoch": 5.860480155201682,
+      "grad_norm": 1.0869048833847046,
+      "learning_rate": 0.0002,
+      "loss": 0.5687,
+      "step": 36250
+    },
+    {
+      "epoch": 5.862096839382427,
+      "grad_norm": 0.9226390719413757,
+      "learning_rate": 0.0002,
+      "loss": 0.5576,
+      "step": 36260
+    },
+    {
+      "epoch": 5.863713523563172,
+      "grad_norm": 1.1755430698394775,
+      "learning_rate": 0.0002,
+      "loss": 0.531,
+      "step": 36270
+    },
+    {
+      "epoch": 5.865330207743917,
+      "grad_norm": 0.8815974593162537,
+      "learning_rate": 0.0002,
+      "loss": 0.558,
+      "step": 36280
+    },
+    {
+      "epoch": 5.866946891924663,
+      "grad_norm": 1.3648751974105835,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 36290
+    },
+    {
+      "epoch": 5.868563576105408,
+      "grad_norm": 0.8729211091995239,
+      "learning_rate": 0.0002,
+      "loss": 0.536,
+      "step": 36300
+    },
+    {
+      "epoch": 5.870180260286153,
+      "grad_norm": 1.0870907306671143,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 36310
+    },
+    {
+      "epoch": 5.871796944466898,
+      "grad_norm": 1.1164259910583496,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 36320
+    },
+    {
+      "epoch": 5.8734136286476435,
+      "grad_norm": 1.1572535037994385,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 36330
+    },
+    {
+      "epoch": 5.875030312828389,
+      "grad_norm": 1.0456238985061646,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 36340
+    },
+    {
+      "epoch": 5.876646997009134,
+      "grad_norm": 1.1310722827911377,
+      "learning_rate": 0.0002,
+      "loss": 0.5008,
+      "step": 36350
+    },
+    {
+      "epoch": 5.878263681189879,
+      "grad_norm": 1.0004712343215942,
+      "learning_rate": 0.0002,
+      "loss": 0.5352,
+      "step": 36360
+    },
+    {
+      "epoch": 5.879880365370624,
+      "grad_norm": 1.0991777181625366,
+      "learning_rate": 0.0002,
+      "loss": 0.5632,
+      "step": 36370
+    },
+    {
+      "epoch": 5.8814970495513705,
+      "grad_norm": 1.2789239883422852,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 36380
+    },
+    {
+      "epoch": 5.883113733732116,
+      "grad_norm": 0.9524819850921631,
+      "learning_rate": 0.0002,
+      "loss": 0.56,
+      "step": 36390
+    },
+    {
+      "epoch": 5.884730417912861,
+      "grad_norm": 1.1115771532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 36400
+    },
+    {
+      "epoch": 5.886347102093606,
+      "grad_norm": 1.37419855594635,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 36410
+    },
+    {
+      "epoch": 5.8879637862743515,
+      "grad_norm": 1.1449527740478516,
+      "learning_rate": 0.0002,
+      "loss": 0.5675,
+      "step": 36420
+    },
+    {
+      "epoch": 5.889580470455097,
+      "grad_norm": 1.198046326637268,
+      "learning_rate": 0.0002,
+      "loss": 0.5255,
+      "step": 36430
+    },
+    {
+      "epoch": 5.891197154635842,
+      "grad_norm": 1.0180530548095703,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 36440
+    },
+    {
+      "epoch": 5.892813838816587,
+      "grad_norm": 1.0516417026519775,
+      "learning_rate": 0.0002,
+      "loss": 0.5319,
+      "step": 36450
+    },
+    {
+      "epoch": 5.894430522997332,
+      "grad_norm": 1.1658052206039429,
+      "learning_rate": 0.0002,
+      "loss": 0.5782,
+      "step": 36460
+    },
+    {
+      "epoch": 5.896047207178078,
+      "grad_norm": 1.190699577331543,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 36470
+    },
+    {
+      "epoch": 5.897663891358823,
+      "grad_norm": 1.1235495805740356,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 36480
+    },
+    {
+      "epoch": 5.899280575539568,
+      "grad_norm": 1.1926926374435425,
+      "learning_rate": 0.0002,
+      "loss": 0.5284,
+      "step": 36490
+    },
+    {
+      "epoch": 5.900897259720313,
+      "grad_norm": 1.1184662580490112,
+      "learning_rate": 0.0002,
+      "loss": 0.5686,
+      "step": 36500
+    },
+    {
+      "epoch": 5.9025139439010585,
+      "grad_norm": 1.000970721244812,
+      "learning_rate": 0.0002,
+      "loss": 0.5147,
+      "step": 36510
+    },
+    {
+      "epoch": 5.904130628081804,
+      "grad_norm": 1.0373306274414062,
+      "learning_rate": 0.0002,
+      "loss": 0.5351,
+      "step": 36520
+    },
+    {
+      "epoch": 5.90574731226255,
+      "grad_norm": 1.0840669870376587,
+      "learning_rate": 0.0002,
+      "loss": 0.535,
+      "step": 36530
+    },
+    {
+      "epoch": 5.907363996443295,
+      "grad_norm": 0.9908381104469299,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 36540
+    },
+    {
+      "epoch": 5.90898068062404,
+      "grad_norm": 1.0456029176712036,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 36550
+    },
+    {
+      "epoch": 5.910597364804786,
+      "grad_norm": 1.1381454467773438,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 36560
+    },
+    {
+      "epoch": 5.912214048985531,
+      "grad_norm": 0.9440900087356567,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 36570
+    },
+    {
+      "epoch": 5.913830733166276,
+      "grad_norm": 1.1674573421478271,
+      "learning_rate": 0.0002,
+      "loss": 0.5542,
+      "step": 36580
+    },
+    {
+      "epoch": 5.915447417347021,
+      "grad_norm": 1.1226966381072998,
+      "learning_rate": 0.0002,
+      "loss": 0.526,
+      "step": 36590
+    },
+    {
+      "epoch": 5.9170641015277665,
+      "grad_norm": 0.9696915745735168,
+      "learning_rate": 0.0002,
+      "loss": 0.6091,
+      "step": 36600
+    },
+    {
+      "epoch": 5.918680785708512,
+      "grad_norm": 0.9593005180358887,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 36610
+    },
+    {
+      "epoch": 5.920297469889257,
+      "grad_norm": 1.122169852256775,
+      "learning_rate": 0.0002,
+      "loss": 0.5536,
+      "step": 36620
+    },
+    {
+      "epoch": 5.921914154070002,
+      "grad_norm": 0.9923415780067444,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 36630
+    },
+    {
+      "epoch": 5.923530838250747,
+      "grad_norm": 1.063838005065918,
+      "learning_rate": 0.0002,
+      "loss": 0.5893,
+      "step": 36640
+    },
+    {
+      "epoch": 5.925147522431493,
+      "grad_norm": 0.9083505272865295,
+      "learning_rate": 0.0002,
+      "loss": 0.5799,
+      "step": 36650
+    },
+    {
+      "epoch": 5.926764206612239,
+      "grad_norm": 0.9439437985420227,
+      "learning_rate": 0.0002,
+      "loss": 0.5264,
+      "step": 36660
+    },
+    {
+      "epoch": 5.928380890792983,
+      "grad_norm": 0.9778534173965454,
+      "learning_rate": 0.0002,
+      "loss": 0.5891,
+      "step": 36670
+    },
+    {
+      "epoch": 5.929997574973729,
+      "grad_norm": 0.9723961353302002,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 36680
+    },
+    {
+      "epoch": 5.9316142591544745,
+      "grad_norm": 1.162333607673645,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 36690
+    },
+    {
+      "epoch": 5.93323094333522,
+      "grad_norm": 1.2784897089004517,
+      "learning_rate": 0.0002,
+      "loss": 0.5771,
+      "step": 36700
+    },
+    {
+      "epoch": 5.934847627515965,
+      "grad_norm": 1.0924867391586304,
+      "learning_rate": 0.0002,
+      "loss": 0.5343,
+      "step": 36710
+    },
+    {
+      "epoch": 5.93646431169671,
+      "grad_norm": 1.046922206878662,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 36720
+    },
+    {
+      "epoch": 5.938080995877455,
+      "grad_norm": 0.8632535338401794,
+      "learning_rate": 0.0002,
+      "loss": 0.5476,
+      "step": 36730
+    },
+    {
+      "epoch": 5.939697680058201,
+      "grad_norm": 1.358762502670288,
+      "learning_rate": 0.0002,
+      "loss": 0.5456,
+      "step": 36740
+    },
+    {
+      "epoch": 5.941314364238946,
+      "grad_norm": 1.2058624029159546,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 36750
+    },
+    {
+      "epoch": 5.942931048419691,
+      "grad_norm": 1.1396408081054688,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 36760
+    },
+    {
+      "epoch": 5.944547732600436,
+      "grad_norm": 1.1510354280471802,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 36770
+    },
+    {
+      "epoch": 5.946164416781182,
+      "grad_norm": 1.1401607990264893,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 36780
+    },
+    {
+      "epoch": 5.947781100961927,
+      "grad_norm": 1.1871325969696045,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 36790
+    },
+    {
+      "epoch": 5.949397785142672,
+      "grad_norm": 0.9928333163261414,
+      "learning_rate": 0.0002,
+      "loss": 0.4945,
+      "step": 36800
+    },
+    {
+      "epoch": 5.951014469323418,
+      "grad_norm": 1.0549445152282715,
+      "learning_rate": 0.0002,
+      "loss": 0.5303,
+      "step": 36810
+    },
+    {
+      "epoch": 5.9526311535041625,
+      "grad_norm": 0.9791563749313354,
+      "learning_rate": 0.0002,
+      "loss": 0.5532,
+      "step": 36820
+    },
+    {
+      "epoch": 5.954247837684909,
+      "grad_norm": 1.1268441677093506,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 36830
+    },
+    {
+      "epoch": 5.955864521865654,
+      "grad_norm": 1.0533992052078247,
+      "learning_rate": 0.0002,
+      "loss": 0.5585,
+      "step": 36840
+    },
+    {
+      "epoch": 5.957481206046399,
+      "grad_norm": 1.023358941078186,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 36850
+    },
+    {
+      "epoch": 5.959097890227144,
+      "grad_norm": 1.2631961107254028,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 36860
+    },
+    {
+      "epoch": 5.9607145744078895,
+      "grad_norm": 0.9397698640823364,
+      "learning_rate": 0.0002,
+      "loss": 0.5662,
+      "step": 36870
+    },
+    {
+      "epoch": 5.962331258588635,
+      "grad_norm": 1.1678427457809448,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 36880
+    },
+    {
+      "epoch": 5.96394794276938,
+      "grad_norm": 1.1403759717941284,
+      "learning_rate": 0.0002,
+      "loss": 0.5435,
+      "step": 36890
+    },
+    {
+      "epoch": 5.965564626950125,
+      "grad_norm": 1.030572772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.5479,
+      "step": 36900
+    },
+    {
+      "epoch": 5.9671813111308705,
+      "grad_norm": 1.0992497205734253,
+      "learning_rate": 0.0002,
+      "loss": 0.5838,
+      "step": 36910
+    },
+    {
+      "epoch": 5.968797995311616,
+      "grad_norm": 1.075466275215149,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 36920
+    },
+    {
+      "epoch": 5.970414679492361,
+      "grad_norm": 1.0153694152832031,
+      "learning_rate": 0.0002,
+      "loss": 0.5739,
+      "step": 36930
+    },
+    {
+      "epoch": 5.972031363673106,
+      "grad_norm": 0.973193883895874,
+      "learning_rate": 0.0002,
+      "loss": 0.5672,
+      "step": 36940
+    },
+    {
+      "epoch": 5.973648047853851,
+      "grad_norm": 0.8294678926467896,
+      "learning_rate": 0.0002,
+      "loss": 0.5585,
+      "step": 36950
+    },
+    {
+      "epoch": 5.9752647320345975,
+      "grad_norm": 1.0048716068267822,
+      "learning_rate": 0.0002,
+      "loss": 0.5631,
+      "step": 36960
+    },
+    {
+      "epoch": 5.976881416215342,
+      "grad_norm": 0.9714070558547974,
+      "learning_rate": 0.0002,
+      "loss": 0.5471,
+      "step": 36970
+    },
+    {
+      "epoch": 5.978498100396088,
+      "grad_norm": 0.8667682409286499,
+      "learning_rate": 0.0002,
+      "loss": 0.5419,
+      "step": 36980
+    },
+    {
+      "epoch": 5.980114784576833,
+      "grad_norm": 1.0461409091949463,
+      "learning_rate": 0.0002,
+      "loss": 0.5474,
+      "step": 36990
+    },
+    {
+      "epoch": 5.981731468757578,
+      "grad_norm": 0.9229754209518433,
+      "learning_rate": 0.0002,
+      "loss": 0.5454,
+      "step": 37000
+    },
+    {
+      "epoch": 5.983348152938324,
+      "grad_norm": 1.0406876802444458,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 37010
+    },
+    {
+      "epoch": 5.984964837119069,
+      "grad_norm": 0.8993828296661377,
+      "learning_rate": 0.0002,
+      "loss": 0.5569,
+      "step": 37020
+    },
+    {
+      "epoch": 5.986581521299814,
+      "grad_norm": 1.2260479927062988,
+      "learning_rate": 0.0002,
+      "loss": 0.5611,
+      "step": 37030
+    },
+    {
+      "epoch": 5.988198205480559,
+      "grad_norm": 1.0107380151748657,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 37040
+    },
+    {
+      "epoch": 5.989814889661305,
+      "grad_norm": 1.0240139961242676,
+      "learning_rate": 0.0002,
+      "loss": 0.5639,
+      "step": 37050
+    },
+    {
+      "epoch": 5.99143157384205,
+      "grad_norm": 1.0185275077819824,
+      "learning_rate": 0.0002,
+      "loss": 0.5209,
+      "step": 37060
+    },
+    {
+      "epoch": 5.993048258022795,
+      "grad_norm": 1.1361802816390991,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 37070
+    },
+    {
+      "epoch": 5.99466494220354,
+      "grad_norm": 1.0395532846450806,
+      "learning_rate": 0.0002,
+      "loss": 0.5692,
+      "step": 37080
+    },
+    {
+      "epoch": 5.9962816263842855,
+      "grad_norm": 0.9463558197021484,
+      "learning_rate": 0.0002,
+      "loss": 0.594,
+      "step": 37090
+    },
+    {
+      "epoch": 5.997898310565031,
+      "grad_norm": 1.2066948413848877,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 37100
+    },
+    {
+      "epoch": 5.999514994745777,
+      "grad_norm": 0.9749386310577393,
+      "learning_rate": 0.0002,
+      "loss": 0.5356,
+      "step": 37110
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 1.2270219326019287,
+      "eval_runtime": 122.2047,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 37113
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 49480,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.7175065406272963e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/training_args.bin b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..155b12fa9acbc6e71dba75c92bfa79e152397ebf
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-37113/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28694d5564a2b5c7d6881d4ba2af103356aa22489d2c22768ebbe47283c0f4a1
+size 5560
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/README.md b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..835e31ab8469ee39ddc8b2b6b2143a8c66dad510
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/README.md
@@ -0,0 +1,202 @@
+---
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.13.1
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/adapter_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..406c5a08dc4a2a33b52c62a482f98c217c417215
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/adapter_config.json
@@ -0,0 +1,29 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/adapter_model.safetensors b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..28601f2be0b06ae9321ec6060321b6f2ec7a7237
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c1406cb62687223f2383ed0c258f0e930e79fed0698d70c29b14f26bb411939
+size 109069176
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/optimizer.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ce132b138ff281db21e3012de036863389b9c56b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85f6de3550fb5c5e303b3a93ee6be0a88e1684e070e6e05bbc1b53f188580a27
+size 55532666
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/rng_state.pth b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..eb65ab3ffed153e0604aeb2a485badf41287aa40
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a404304c9f5ba871a0af32ddd9cb5af2e5087b5984920bd41be7f2405c4de25
+size 14244
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/scheduler.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..497e899a7927d55a3579d08a7fa7106328985e8c
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75fc1866fc3d98969c78ad715fdff0568f69b0d75b9c4a54d77c7ffc5e910d3b
+size 1064
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/special_tokens_map.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/special_tokens_map.json
@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/tokenizer.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..86a33946b0c77216d2cce91bb28c8fada4a5e80b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/tokenizer_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..061e40d9db3253624f86e8e364c15ef546527c9d
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/tokenizer_config.json
@@ -0,0 +1,2063 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/trainer_state.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..1cca2ea5ee533ef48bb3612e2d7c4deebbfa9104
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/trainer_state.json
@@ -0,0 +1,30392 @@
+{
+  "best_metric": 1.0871200561523438,
+  "best_model_checkpoint": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185",
+  "epoch": 6.9999191657909625,
+  "eval_steps": 10,
+  "global_step": 43298,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016166841807452913,
+      "grad_norm": 0.9894065856933594,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032333683614905826,
+      "grad_norm": 1.7810699939727783,
+      "learning_rate": 0.0002,
+      "loss": 1.1528,
+      "step": 20
+    },
+    {
+      "epoch": 0.004850052542235874,
+      "grad_norm": 0.5969577431678772,
+      "learning_rate": 0.0002,
+      "loss": 0.9767,
+      "step": 30
+    },
+    {
+      "epoch": 0.006466736722981165,
+      "grad_norm": 0.6354120969772339,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 40
+    },
+    {
+      "epoch": 0.008083420903726457,
+      "grad_norm": 0.5604607462882996,
+      "learning_rate": 0.0002,
+      "loss": 0.8643,
+      "step": 50
+    },
+    {
+      "epoch": 0.009700105084471748,
+      "grad_norm": 0.4676193594932556,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 60
+    },
+    {
+      "epoch": 0.01131678926521704,
+      "grad_norm": 0.6099211573600769,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 70
+    },
+    {
+      "epoch": 0.01293347344596233,
+      "grad_norm": 0.48639994859695435,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 80
+    },
+    {
+      "epoch": 0.014550157626707623,
+      "grad_norm": 0.4904264509677887,
+      "learning_rate": 0.0002,
+      "loss": 0.8704,
+      "step": 90
+    },
+    {
+      "epoch": 0.016166841807452915,
+      "grad_norm": 2.8334362506866455,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 100
+    },
+    {
+      "epoch": 0.017783525988198205,
+      "grad_norm": 0.43221670389175415,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 110
+    },
+    {
+      "epoch": 0.019400210168943496,
+      "grad_norm": 0.42244166135787964,
+      "learning_rate": 0.0002,
+      "loss": 0.8412,
+      "step": 120
+    },
+    {
+      "epoch": 0.02101689434968879,
+      "grad_norm": 0.45363298058509827,
+      "learning_rate": 0.0002,
+      "loss": 0.8467,
+      "step": 130
+    },
+    {
+      "epoch": 0.02263357853043408,
+      "grad_norm": 0.44816508889198303,
+      "learning_rate": 0.0002,
+      "loss": 0.8641,
+      "step": 140
+    },
+    {
+      "epoch": 0.02425026271117937,
+      "grad_norm": 0.43308213353157043,
+      "learning_rate": 0.0002,
+      "loss": 0.8496,
+      "step": 150
+    },
+    {
+      "epoch": 0.02586694689192466,
+      "grad_norm": 0.4084763526916504,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 160
+    },
+    {
+      "epoch": 0.027483631072669955,
+      "grad_norm": 0.5363703966140747,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 170
+    },
+    {
+      "epoch": 0.029100315253415245,
+      "grad_norm": 0.4619699716567993,
+      "learning_rate": 0.0002,
+      "loss": 0.8558,
+      "step": 180
+    },
+    {
+      "epoch": 0.030716999434160536,
+      "grad_norm": 0.49069908261299133,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 190
+    },
+    {
+      "epoch": 0.03233368361490583,
+      "grad_norm": 0.4645835757255554,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 200
+    },
+    {
+      "epoch": 0.03395036779565112,
+      "grad_norm": 1.2411243915557861,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 210
+    },
+    {
+      "epoch": 0.03556705197639641,
+      "grad_norm": 0.5211851596832275,
+      "learning_rate": 0.0002,
+      "loss": 0.8245,
+      "step": 220
+    },
+    {
+      "epoch": 0.037183736157141704,
+      "grad_norm": 0.5253691673278809,
+      "learning_rate": 0.0002,
+      "loss": 0.8194,
+      "step": 230
+    },
+    {
+      "epoch": 0.03880042033788699,
+      "grad_norm": 0.4567478895187378,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 240
+    },
+    {
+      "epoch": 0.040417104518632285,
+      "grad_norm": 0.5472128391265869,
+      "learning_rate": 0.0002,
+      "loss": 0.838,
+      "step": 250
+    },
+    {
+      "epoch": 0.04203378869937758,
+      "grad_norm": 0.42978546023368835,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 260
+    },
+    {
+      "epoch": 0.043650472880122866,
+      "grad_norm": 0.601734459400177,
+      "learning_rate": 0.0002,
+      "loss": 0.8334,
+      "step": 270
+    },
+    {
+      "epoch": 0.04526715706086816,
+      "grad_norm": 0.4286513328552246,
+      "learning_rate": 0.0002,
+      "loss": 0.815,
+      "step": 280
+    },
+    {
+      "epoch": 0.046883841241613454,
+      "grad_norm": 0.5230861902236938,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 290
+    },
+    {
+      "epoch": 0.04850052542235874,
+      "grad_norm": 0.6504611968994141,
+      "learning_rate": 0.0002,
+      "loss": 0.8636,
+      "step": 300
+    },
+    {
+      "epoch": 0.050117209603104035,
+      "grad_norm": 0.43485215306282043,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 310
+    },
+    {
+      "epoch": 0.05173389378384932,
+      "grad_norm": 0.4717007875442505,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 320
+    },
+    {
+      "epoch": 0.053350577964594616,
+      "grad_norm": 0.4059787690639496,
+      "learning_rate": 0.0002,
+      "loss": 0.8469,
+      "step": 330
+    },
+    {
+      "epoch": 0.05496726214533991,
+      "grad_norm": 0.4366913437843323,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 340
+    },
+    {
+      "epoch": 0.0565839463260852,
+      "grad_norm": 0.4233848452568054,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 350
+    },
+    {
+      "epoch": 0.05820063050683049,
+      "grad_norm": 0.4209108352661133,
+      "learning_rate": 0.0002,
+      "loss": 0.8456,
+      "step": 360
+    },
+    {
+      "epoch": 0.059817314687575784,
+      "grad_norm": 0.41637396812438965,
+      "learning_rate": 0.0002,
+      "loss": 0.816,
+      "step": 370
+    },
+    {
+      "epoch": 0.06143399886832107,
+      "grad_norm": 0.46235376596450806,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 380
+    },
+    {
+      "epoch": 0.06305068304906636,
+      "grad_norm": 0.4013484716415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 390
+    },
+    {
+      "epoch": 0.06466736722981166,
+      "grad_norm": 0.47443896532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 400
+    },
+    {
+      "epoch": 0.06628405141055695,
+      "grad_norm": 0.3942156434059143,
+      "learning_rate": 0.0002,
+      "loss": 0.8666,
+      "step": 410
+    },
+    {
+      "epoch": 0.06790073559130223,
+      "grad_norm": 0.4965320825576782,
+      "learning_rate": 0.0002,
+      "loss": 0.8402,
+      "step": 420
+    },
+    {
+      "epoch": 0.06951741977204753,
+      "grad_norm": 0.4304835796356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8317,
+      "step": 430
+    },
+    {
+      "epoch": 0.07113410395279282,
+      "grad_norm": 0.511726975440979,
+      "learning_rate": 0.0002,
+      "loss": 0.8528,
+      "step": 440
+    },
+    {
+      "epoch": 0.07275078813353811,
+      "grad_norm": 0.4040689170360565,
+      "learning_rate": 0.0002,
+      "loss": 0.8675,
+      "step": 450
+    },
+    {
+      "epoch": 0.07436747231428341,
+      "grad_norm": 0.5402171015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 460
+    },
+    {
+      "epoch": 0.0759841564950287,
+      "grad_norm": 0.4174517095088959,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 470
+    },
+    {
+      "epoch": 0.07760084067577398,
+      "grad_norm": 0.4306182265281677,
+      "learning_rate": 0.0002,
+      "loss": 0.7605,
+      "step": 480
+    },
+    {
+      "epoch": 0.07921752485651928,
+      "grad_norm": 0.535210132598877,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 490
+    },
+    {
+      "epoch": 0.08083420903726457,
+      "grad_norm": 0.5339109897613525,
+      "learning_rate": 0.0002,
+      "loss": 0.7825,
+      "step": 500
+    },
+    {
+      "epoch": 0.08245089321800986,
+      "grad_norm": 0.45754891633987427,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 510
+    },
+    {
+      "epoch": 0.08406757739875516,
+      "grad_norm": 0.43820783495903015,
+      "learning_rate": 0.0002,
+      "loss": 0.8144,
+      "step": 520
+    },
+    {
+      "epoch": 0.08568426157950045,
+      "grad_norm": 0.4434749186038971,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 530
+    },
+    {
+      "epoch": 0.08730094576024573,
+      "grad_norm": 0.43111467361450195,
+      "learning_rate": 0.0002,
+      "loss": 0.7857,
+      "step": 540
+    },
+    {
+      "epoch": 0.08891762994099103,
+      "grad_norm": 0.4378940165042877,
+      "learning_rate": 0.0002,
+      "loss": 0.8418,
+      "step": 550
+    },
+    {
+      "epoch": 0.09053431412173632,
+      "grad_norm": 0.4772215187549591,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 560
+    },
+    {
+      "epoch": 0.09215099830248161,
+      "grad_norm": 0.6837629079818726,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 570
+    },
+    {
+      "epoch": 0.09376768248322691,
+      "grad_norm": 0.42241212725639343,
+      "learning_rate": 0.0002,
+      "loss": 0.8607,
+      "step": 580
+    },
+    {
+      "epoch": 0.0953843666639722,
+      "grad_norm": 0.5165936350822449,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 590
+    },
+    {
+      "epoch": 0.09700105084471748,
+      "grad_norm": 0.48737478256225586,
+      "learning_rate": 0.0002,
+      "loss": 0.8664,
+      "step": 600
+    },
+    {
+      "epoch": 0.09861773502546278,
+      "grad_norm": 0.47419852018356323,
+      "learning_rate": 0.0002,
+      "loss": 0.8806,
+      "step": 610
+    },
+    {
+      "epoch": 0.10023441920620807,
+      "grad_norm": 0.4975486099720001,
+      "learning_rate": 0.0002,
+      "loss": 0.8254,
+      "step": 620
+    },
+    {
+      "epoch": 0.10185110338695336,
+      "grad_norm": 0.49123844504356384,
+      "learning_rate": 0.0002,
+      "loss": 0.8548,
+      "step": 630
+    },
+    {
+      "epoch": 0.10346778756769864,
+      "grad_norm": 0.6288952827453613,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 640
+    },
+    {
+      "epoch": 0.10508447174844394,
+      "grad_norm": 0.4277345836162567,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 650
+    },
+    {
+      "epoch": 0.10670115592918923,
+      "grad_norm": 0.4021061956882477,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 660
+    },
+    {
+      "epoch": 0.10831784010993452,
+      "grad_norm": 0.3492237329483032,
+      "learning_rate": 0.0002,
+      "loss": 0.87,
+      "step": 670
+    },
+    {
+      "epoch": 0.10993452429067982,
+      "grad_norm": 0.4341012239456177,
+      "learning_rate": 0.0002,
+      "loss": 0.8698,
+      "step": 680
+    },
+    {
+      "epoch": 0.1115512084714251,
+      "grad_norm": 0.7296304106712341,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 690
+    },
+    {
+      "epoch": 0.1131678926521704,
+      "grad_norm": 0.397494912147522,
+      "learning_rate": 0.0002,
+      "loss": 0.8433,
+      "step": 700
+    },
+    {
+      "epoch": 0.1147845768329157,
+      "grad_norm": 0.396431028842926,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 710
+    },
+    {
+      "epoch": 0.11640126101366098,
+      "grad_norm": 0.48842838406562805,
+      "learning_rate": 0.0002,
+      "loss": 0.8379,
+      "step": 720
+    },
+    {
+      "epoch": 0.11801794519440627,
+      "grad_norm": 0.46322616934776306,
+      "learning_rate": 0.0002,
+      "loss": 0.8238,
+      "step": 730
+    },
+    {
+      "epoch": 0.11963462937515157,
+      "grad_norm": 0.47990912199020386,
+      "learning_rate": 0.0002,
+      "loss": 0.8041,
+      "step": 740
+    },
+    {
+      "epoch": 0.12125131355589686,
+      "grad_norm": 0.4997142255306244,
+      "learning_rate": 0.0002,
+      "loss": 0.82,
+      "step": 750
+    },
+    {
+      "epoch": 0.12286799773664214,
+      "grad_norm": 0.4040526747703552,
+      "learning_rate": 0.0002,
+      "loss": 0.7702,
+      "step": 760
+    },
+    {
+      "epoch": 0.12448468191738744,
+      "grad_norm": 0.453095942735672,
+      "learning_rate": 0.0002,
+      "loss": 0.863,
+      "step": 770
+    },
+    {
+      "epoch": 0.12610136609813272,
+      "grad_norm": 0.4636971950531006,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 780
+    },
+    {
+      "epoch": 0.12771805027887803,
+      "grad_norm": 0.4279276132583618,
+      "learning_rate": 0.0002,
+      "loss": 0.8112,
+      "step": 790
+    },
+    {
+      "epoch": 0.12933473445962332,
+      "grad_norm": 0.46212655305862427,
+      "learning_rate": 0.0002,
+      "loss": 0.8711,
+      "step": 800
+    },
+    {
+      "epoch": 0.1309514186403686,
+      "grad_norm": 0.43127650022506714,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 810
+    },
+    {
+      "epoch": 0.1325681028211139,
+      "grad_norm": 0.4201301336288452,
+      "learning_rate": 0.0002,
+      "loss": 0.8476,
+      "step": 820
+    },
+    {
+      "epoch": 0.13418478700185918,
+      "grad_norm": 0.42583167552948,
+      "learning_rate": 0.0002,
+      "loss": 0.8078,
+      "step": 830
+    },
+    {
+      "epoch": 0.13580147118260447,
+      "grad_norm": 0.4535622000694275,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 840
+    },
+    {
+      "epoch": 0.13741815536334978,
+      "grad_norm": 0.4116036891937256,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 850
+    },
+    {
+      "epoch": 0.13903483954409507,
+      "grad_norm": 0.45997580885887146,
+      "learning_rate": 0.0002,
+      "loss": 0.8466,
+      "step": 860
+    },
+    {
+      "epoch": 0.14065152372484035,
+      "grad_norm": 0.4487837255001068,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 870
+    },
+    {
+      "epoch": 0.14226820790558564,
+      "grad_norm": 0.43650057911872864,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 880
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 0.5335358381271362,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 890
+    },
+    {
+      "epoch": 0.14550157626707622,
+      "grad_norm": 0.5989000201225281,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 900
+    },
+    {
+      "epoch": 0.14711826044782153,
+      "grad_norm": 0.517179012298584,
+      "learning_rate": 0.0002,
+      "loss": 0.8385,
+      "step": 910
+    },
+    {
+      "epoch": 0.14873494462856682,
+      "grad_norm": 0.44435232877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.8255,
+      "step": 920
+    },
+    {
+      "epoch": 0.1503516288093121,
+      "grad_norm": 0.42635923624038696,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 930
+    },
+    {
+      "epoch": 0.1519683129900574,
+      "grad_norm": 0.49603334069252014,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 940
+    },
+    {
+      "epoch": 0.15358499717080268,
+      "grad_norm": 0.40639808773994446,
+      "learning_rate": 0.0002,
+      "loss": 0.8377,
+      "step": 950
+    },
+    {
+      "epoch": 0.15520168135154797,
+      "grad_norm": 0.4850759208202362,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 960
+    },
+    {
+      "epoch": 0.15681836553229328,
+      "grad_norm": 0.4427442252635956,
+      "learning_rate": 0.0002,
+      "loss": 0.846,
+      "step": 970
+    },
+    {
+      "epoch": 0.15843504971303857,
+      "grad_norm": 0.3760930001735687,
+      "learning_rate": 0.0002,
+      "loss": 0.8705,
+      "step": 980
+    },
+    {
+      "epoch": 0.16005173389378385,
+      "grad_norm": 0.4794144332408905,
+      "learning_rate": 0.0002,
+      "loss": 0.8644,
+      "step": 990
+    },
+    {
+      "epoch": 0.16166841807452914,
+      "grad_norm": 0.45828768610954285,
+      "learning_rate": 0.0002,
+      "loss": 0.8002,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16328510225527443,
+      "grad_norm": 0.6313053369522095,
+      "learning_rate": 0.0002,
+      "loss": 0.7658,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16490178643601971,
+      "grad_norm": 0.45041006803512573,
+      "learning_rate": 0.0002,
+      "loss": 0.8047,
+      "step": 1020
+    },
+    {
+      "epoch": 0.166518470616765,
+      "grad_norm": 0.441403865814209,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16813515479751032,
+      "grad_norm": 0.8171296119689941,
+      "learning_rate": 0.0002,
+      "loss": 0.8475,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1697518389782556,
+      "grad_norm": 0.7137420773506165,
+      "learning_rate": 0.0002,
+      "loss": 0.845,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1713685231590009,
+      "grad_norm": 0.5236809849739075,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17298520733974618,
+      "grad_norm": 0.5021864175796509,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17460189152049146,
+      "grad_norm": 0.47347521781921387,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17621857570123675,
+      "grad_norm": 0.4631653428077698,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17783525988198207,
+      "grad_norm": 0.49169182777404785,
+      "learning_rate": 0.0002,
+      "loss": 0.8166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17945194406272735,
+      "grad_norm": 0.5019739270210266,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18106862824347264,
+      "grad_norm": 0.5100422501564026,
+      "learning_rate": 0.0002,
+      "loss": 0.8247,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18268531242421793,
+      "grad_norm": 0.3888324499130249,
+      "learning_rate": 0.0002,
+      "loss": 0.8142,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18430199660496321,
+      "grad_norm": 0.39765217900276184,
+      "learning_rate": 0.0002,
+      "loss": 0.8533,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1859186807857085,
+      "grad_norm": 0.47190186381340027,
+      "learning_rate": 0.0002,
+      "loss": 0.8541,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18753536496645382,
+      "grad_norm": 0.4464188814163208,
+      "learning_rate": 0.0002,
+      "loss": 0.8301,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1891520491471991,
+      "grad_norm": 0.5153930187225342,
+      "learning_rate": 0.0002,
+      "loss": 0.8341,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1907687333279444,
+      "grad_norm": 0.4779708683490753,
+      "learning_rate": 0.0002,
+      "loss": 0.8033,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19238541750868968,
+      "grad_norm": 0.4834315776824951,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19400210168943496,
+      "grad_norm": 0.402357816696167,
+      "learning_rate": 0.0002,
+      "loss": 0.7721,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19561878587018025,
+      "grad_norm": 0.45899084210395813,
+      "learning_rate": 0.0002,
+      "loss": 0.7941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19723547005092557,
+      "grad_norm": 0.5106529593467712,
+      "learning_rate": 0.0002,
+      "loss": 0.8353,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19885215423167085,
+      "grad_norm": 0.45261722803115845,
+      "learning_rate": 0.0002,
+      "loss": 0.7816,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20046883841241614,
+      "grad_norm": 0.4647127091884613,
+      "learning_rate": 0.0002,
+      "loss": 0.8068,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20208552259316143,
+      "grad_norm": 0.4849368929862976,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2037022067739067,
+      "grad_norm": 0.4518061578273773,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 1260
+    },
+    {
+      "epoch": 0.205318890954652,
+      "grad_norm": 0.49535325169563293,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2069355751353973,
+      "grad_norm": 0.4835205376148224,
+      "learning_rate": 0.0002,
+      "loss": 0.8348,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2085522593161426,
+      "grad_norm": 0.45308539271354675,
+      "learning_rate": 0.0002,
+      "loss": 0.8428,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2101689434968879,
+      "grad_norm": 0.5369905233383179,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21178562767763318,
+      "grad_norm": 0.5031622052192688,
+      "learning_rate": 0.0002,
+      "loss": 0.8676,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21340231185837846,
+      "grad_norm": 0.48010334372520447,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21501899603912375,
+      "grad_norm": 0.4905701279640198,
+      "learning_rate": 0.0002,
+      "loss": 0.806,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21663568021986904,
+      "grad_norm": 0.43531742691993713,
+      "learning_rate": 0.0002,
+      "loss": 0.7885,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21825236440061435,
+      "grad_norm": 0.44330692291259766,
+      "learning_rate": 0.0002,
+      "loss": 0.8191,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21986904858135964,
+      "grad_norm": 0.5384416580200195,
+      "learning_rate": 0.0002,
+      "loss": 0.8205,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22148573276210493,
+      "grad_norm": 0.4181833863258362,
+      "learning_rate": 0.0002,
+      "loss": 0.7726,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2231024169428502,
+      "grad_norm": 0.523833692073822,
+      "learning_rate": 0.0002,
+      "loss": 0.8311,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.5528736710548401,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2263357853043408,
+      "grad_norm": 0.43515023589134216,
+      "learning_rate": 0.0002,
+      "loss": 0.8079,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2279524694850861,
+      "grad_norm": 0.48809877038002014,
+      "learning_rate": 0.0002,
+      "loss": 0.8403,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2295691536658314,
+      "grad_norm": 0.43591251969337463,
+      "learning_rate": 0.0002,
+      "loss": 0.8165,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23118583784657668,
+      "grad_norm": 0.44625312089920044,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23280252202732196,
+      "grad_norm": 0.4390665292739868,
+      "learning_rate": 0.0002,
+      "loss": 0.8134,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23441920620806725,
+      "grad_norm": 0.48496049642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.8465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23603589038881254,
+      "grad_norm": 0.45919957756996155,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23765257456955785,
+      "grad_norm": 0.5471845865249634,
+      "learning_rate": 0.0002,
+      "loss": 0.8659,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23926925875030314,
+      "grad_norm": 0.47269317507743835,
+      "learning_rate": 0.0002,
+      "loss": 0.8164,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24088594293104842,
+      "grad_norm": 0.4930245578289032,
+      "learning_rate": 0.0002,
+      "loss": 0.854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2425026271117937,
+      "grad_norm": 0.5605630278587341,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 1500
+    },
+    {
+      "epoch": 0.244119311292539,
+      "grad_norm": 0.4435870945453644,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24573599547328429,
+      "grad_norm": 0.4941999912261963,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24735267965402957,
+      "grad_norm": 0.5100624561309814,
+      "learning_rate": 0.0002,
+      "loss": 0.8427,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2489693638347749,
+      "grad_norm": 0.4638267457485199,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25058604801552015,
+      "grad_norm": 0.5071570873260498,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25220273219626543,
+      "grad_norm": 0.4291319251060486,
+      "learning_rate": 0.0002,
+      "loss": 0.7724,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2538194163770108,
+      "grad_norm": 0.5388049483299255,
+      "learning_rate": 0.0002,
+      "loss": 0.7984,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25543610055775606,
+      "grad_norm": 0.5083683729171753,
+      "learning_rate": 0.0002,
+      "loss": 0.8176,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25705278473850135,
+      "grad_norm": 0.4824463725090027,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25866946891924664,
+      "grad_norm": 0.41177722811698914,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2602861530999919,
+      "grad_norm": 0.5656219124794006,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2619028372807372,
+      "grad_norm": 0.41063204407691956,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2635195214614825,
+      "grad_norm": 0.4897061288356781,
+      "learning_rate": 0.0002,
+      "loss": 0.7998,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2651362056422278,
+      "grad_norm": 0.4454376697540283,
+      "learning_rate": 0.0002,
+      "loss": 0.8198,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26675288982297307,
+      "grad_norm": 0.4355238378047943,
+      "learning_rate": 0.0002,
+      "loss": 0.8684,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26836957400371836,
+      "grad_norm": 0.458310067653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26998625818446365,
+      "grad_norm": 0.4752083718776703,
+      "learning_rate": 0.0002,
+      "loss": 0.7935,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27160294236520893,
+      "grad_norm": 0.4666106402873993,
+      "learning_rate": 0.0002,
+      "loss": 0.8267,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2732196265459543,
+      "grad_norm": 0.4213818609714508,
+      "learning_rate": 0.0002,
+      "loss": 0.8252,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27483631072669956,
+      "grad_norm": 0.5768913626670837,
+      "learning_rate": 0.0002,
+      "loss": 0.8559,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27645299490744485,
+      "grad_norm": 0.4209914803504944,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27806967908819014,
+      "grad_norm": 0.501909613609314,
+      "learning_rate": 0.0002,
+      "loss": 0.8167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2796863632689354,
+      "grad_norm": 0.5266261100769043,
+      "learning_rate": 0.0002,
+      "loss": 0.7832,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2813030474496807,
+      "grad_norm": 0.43806859850883484,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 1740
+    },
+    {
+      "epoch": 0.282919731630426,
+      "grad_norm": 0.46048814058303833,
+      "learning_rate": 0.0002,
+      "loss": 0.8157,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2845364158111713,
+      "grad_norm": 0.44972819089889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8596,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28615309999191657,
+      "grad_norm": 0.5114831328392029,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.47931742668151855,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28938646835340714,
+      "grad_norm": 0.5092599987983704,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29100315253415243,
+      "grad_norm": 0.37581443786621094,
+      "learning_rate": 0.0002,
+      "loss": 0.8506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2926198367148977,
+      "grad_norm": 0.47097381949424744,
+      "learning_rate": 0.0002,
+      "loss": 0.7932,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29423652089564306,
+      "grad_norm": 0.48300236463546753,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29585320507638835,
+      "grad_norm": 0.5600419640541077,
+      "learning_rate": 0.0002,
+      "loss": 0.8391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29746988925713364,
+      "grad_norm": 0.48555272817611694,
+      "learning_rate": 0.0002,
+      "loss": 0.8507,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2990865734378789,
+      "grad_norm": 0.3752668499946594,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3007032576186242,
+      "grad_norm": 0.5328747034072876,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3023199417993695,
+      "grad_norm": 0.48716455698013306,
+      "learning_rate": 0.0002,
+      "loss": 0.8426,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3039366259801148,
+      "grad_norm": 0.5011493563652039,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30555331016086007,
+      "grad_norm": 0.46461427211761475,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30716999434160536,
+      "grad_norm": 0.36630210280418396,
+      "learning_rate": 0.0002,
+      "loss": 0.8478,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30878667852235064,
+      "grad_norm": 0.4217296242713928,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31040336270309593,
+      "grad_norm": 0.4394875466823578,
+      "learning_rate": 0.0002,
+      "loss": 0.8128,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3120200468838412,
+      "grad_norm": 0.6587965488433838,
+      "learning_rate": 0.0002,
+      "loss": 0.8471,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31363673106458656,
+      "grad_norm": 0.5469298958778381,
+      "learning_rate": 0.0002,
+      "loss": 0.8565,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31525341524533185,
+      "grad_norm": 0.4371595084667206,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31687009942607713,
+      "grad_norm": 0.4809541404247284,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3184867836068224,
+      "grad_norm": 0.6061086654663086,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3201034677875677,
+      "grad_norm": 0.5342657566070557,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 1980
+    },
+    {
+      "epoch": 0.321720151968313,
+      "grad_norm": 0.5057743787765503,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3233368361490583,
+      "grad_norm": 0.528626024723053,
+      "learning_rate": 0.0002,
+      "loss": 0.7774,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32495352032980357,
+      "grad_norm": 0.46742770075798035,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32657020451054886,
+      "grad_norm": 0.515101432800293,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32818688869129414,
+      "grad_norm": 0.41941216588020325,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32980357287203943,
+      "grad_norm": 0.49902522563934326,
+      "learning_rate": 0.0002,
+      "loss": 0.7637,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3314202570527847,
+      "grad_norm": 0.4120897650718689,
+      "learning_rate": 0.0002,
+      "loss": 0.7822,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33303694123353,
+      "grad_norm": 0.45352041721343994,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33465362541427535,
+      "grad_norm": 0.523199737071991,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33627030959502063,
+      "grad_norm": 0.4390358626842499,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3378869937757659,
+      "grad_norm": 0.6752901077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3395036779565112,
+      "grad_norm": 0.547821044921875,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3411203621372565,
+      "grad_norm": 0.5161308646202087,
+      "learning_rate": 0.0002,
+      "loss": 0.8561,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3427370463180018,
+      "grad_norm": 0.4565401077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34435373049874707,
+      "grad_norm": 0.4666115939617157,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34597041467949236,
+      "grad_norm": 0.4090428352355957,
+      "learning_rate": 0.0002,
+      "loss": 0.8189,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34758709886023764,
+      "grad_norm": 0.510845422744751,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34920378304098293,
+      "grad_norm": 0.42861923575401306,
+      "learning_rate": 0.0002,
+      "loss": 0.8398,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3508204672217282,
+      "grad_norm": 0.4476332664489746,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3524371514024735,
+      "grad_norm": 0.6065791249275208,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35405383558321885,
+      "grad_norm": 0.42335066199302673,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35567051976396413,
+      "grad_norm": 0.5094629526138306,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3572872039447094,
+      "grad_norm": 0.5476373434066772,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3589038881254547,
+      "grad_norm": 0.3911719024181366,
+      "learning_rate": 0.0002,
+      "loss": 0.814,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3605205723062,
+      "grad_norm": 0.6599636077880859,
+      "learning_rate": 0.0002,
+      "loss": 0.8599,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3621372564869453,
+      "grad_norm": 0.40381914377212524,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36375394066769057,
+      "grad_norm": 0.4433908462524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36537062484843585,
+      "grad_norm": 0.578326940536499,
+      "learning_rate": 0.0002,
+      "loss": 0.8503,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36698730902918114,
+      "grad_norm": 0.5734784007072449,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36860399320992643,
+      "grad_norm": 0.45555487275123596,
+      "learning_rate": 0.0002,
+      "loss": 0.8193,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3702206773906717,
+      "grad_norm": 0.5666276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.7929,
+      "step": 2290
+    },
+    {
+      "epoch": 0.371837361571417,
+      "grad_norm": 0.5461117625236511,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3734540457521623,
+      "grad_norm": 0.6318911910057068,
+      "learning_rate": 0.0002,
+      "loss": 0.8204,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37507072993290763,
+      "grad_norm": 0.493263304233551,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3766874141136529,
+      "grad_norm": 0.5888760089874268,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3783040982943982,
+      "grad_norm": 0.48671841621398926,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3799207824751435,
+      "grad_norm": 0.4385145306587219,
+      "learning_rate": 0.0002,
+      "loss": 0.8367,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3815374666558888,
+      "grad_norm": 0.5523318648338318,
+      "learning_rate": 0.0002,
+      "loss": 0.812,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38315415083663407,
+      "grad_norm": 0.7308220267295837,
+      "learning_rate": 0.0002,
+      "loss": 0.8351,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38477083501737935,
+      "grad_norm": 0.554214358329773,
+      "learning_rate": 0.0002,
+      "loss": 0.859,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38638751919812464,
+      "grad_norm": 0.5425800085067749,
+      "learning_rate": 0.0002,
+      "loss": 0.8146,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3880042033788699,
+      "grad_norm": 0.48811158537864685,
+      "learning_rate": 0.0002,
+      "loss": 0.8282,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3896208875596152,
+      "grad_norm": 0.49212366342544556,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3912375717403605,
+      "grad_norm": 0.5222218632698059,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3928542559211058,
+      "grad_norm": 0.4699819087982178,
+      "learning_rate": 0.0002,
+      "loss": 0.8182,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39447094010185113,
+      "grad_norm": 0.46153587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7919,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3960876242825964,
+      "grad_norm": 0.4150611162185669,
+      "learning_rate": 0.0002,
+      "loss": 0.8111,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3977043084633417,
+      "grad_norm": 0.5799614787101746,
+      "learning_rate": 0.0002,
+      "loss": 0.8589,
+      "step": 2460
+    },
+    {
+      "epoch": 0.399320992644087,
+      "grad_norm": 0.56536865234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4009376768248323,
+      "grad_norm": 0.5451247096061707,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40255436100557757,
+      "grad_norm": 0.5914521217346191,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40417104518632285,
+      "grad_norm": 0.4428117275238037,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40578772936706814,
+      "grad_norm": 0.48580947518348694,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4074044135478134,
+      "grad_norm": 0.436734676361084,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4090210977285587,
+      "grad_norm": 0.5752223134040833,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 2530
+    },
+    {
+      "epoch": 0.410637781909304,
+      "grad_norm": 0.4271308183670044,
+      "learning_rate": 0.0002,
+      "loss": 0.8181,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4122544660900493,
+      "grad_norm": 0.46294718980789185,
+      "learning_rate": 0.0002,
+      "loss": 0.8058,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4138711502707946,
+      "grad_norm": 0.49407583475112915,
+      "learning_rate": 0.0002,
+      "loss": 0.8473,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4154878344515399,
+      "grad_norm": 0.4729035496711731,
+      "learning_rate": 0.0002,
+      "loss": 0.7881,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4171045186322852,
+      "grad_norm": 0.4129747152328491,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4187212028130305,
+      "grad_norm": 0.5684236288070679,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4203378869937758,
+      "grad_norm": 0.4862157106399536,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42195457117452106,
+      "grad_norm": 0.46567976474761963,
+      "learning_rate": 0.0002,
+      "loss": 0.7582,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42357125535526635,
+      "grad_norm": 0.5710650682449341,
+      "learning_rate": 0.0002,
+      "loss": 0.7755,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42518793953601164,
+      "grad_norm": 0.5660041570663452,
+      "learning_rate": 0.0002,
+      "loss": 0.8573,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4268046237167569,
+      "grad_norm": 0.47944375872612,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4284213078975022,
+      "grad_norm": 0.537223756313324,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4300379920782475,
+      "grad_norm": 0.41669997572898865,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 0.44727686047554016,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4332713604397381,
+      "grad_norm": 0.5600888729095459,
+      "learning_rate": 0.0002,
+      "loss": 0.8241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4348880446204834,
+      "grad_norm": 0.39820605516433716,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4365047288012287,
+      "grad_norm": 0.5637655854225159,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 2700
+    },
+    {
+      "epoch": 0.438121412981974,
+      "grad_norm": 0.6363666653633118,
+      "learning_rate": 0.0002,
+      "loss": 0.855,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4397380971627193,
+      "grad_norm": 0.5656129121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.8468,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44135478134346456,
+      "grad_norm": 0.5600156188011169,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44297146552420985,
+      "grad_norm": 0.5506579875946045,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44458814970495514,
+      "grad_norm": 0.49878305196762085,
+      "learning_rate": 0.0002,
+      "loss": 0.7725,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4462048338857004,
+      "grad_norm": 0.4569213092327118,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4478215180664457,
+      "grad_norm": 0.6056680083274841,
+      "learning_rate": 0.0002,
+      "loss": 0.8028,
+      "step": 2770
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.44474557042121887,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4510548864279363,
+      "grad_norm": 0.46055394411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.801,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4526715706086816,
+      "grad_norm": 0.4904133379459381,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45428825478942686,
+      "grad_norm": 0.5647031664848328,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4559049389701722,
+      "grad_norm": 0.5759473443031311,
+      "learning_rate": 0.0002,
+      "loss": 0.8622,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4575216231509175,
+      "grad_norm": 0.5161895751953125,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4591383073316628,
+      "grad_norm": 0.4248254597187042,
+      "learning_rate": 0.0002,
+      "loss": 0.8045,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46075499151240806,
+      "grad_norm": 0.45395001769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.7838,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46237167569315335,
+      "grad_norm": 0.5358697772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.8208,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46398835987389864,
+      "grad_norm": 0.5379165410995483,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4656050440546439,
+      "grad_norm": 0.4601989686489105,
+      "learning_rate": 0.0002,
+      "loss": 0.7403,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4672217282353892,
+      "grad_norm": 0.671115517616272,
+      "learning_rate": 0.0002,
+      "loss": 0.8523,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4688384124161345,
+      "grad_norm": 0.4425133168697357,
+      "learning_rate": 0.0002,
+      "loss": 0.8262,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4704550965968798,
+      "grad_norm": 0.5446155071258545,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47207178077762507,
+      "grad_norm": 0.603306233882904,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47368846495837036,
+      "grad_norm": 0.5377997159957886,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4753051491391157,
+      "grad_norm": 0.4931027591228485,
+      "learning_rate": 0.0002,
+      "loss": 0.8075,
+      "step": 2940
+    },
+    {
+      "epoch": 0.476921833319861,
+      "grad_norm": 0.4711960256099701,
+      "learning_rate": 0.0002,
+      "loss": 0.8004,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4785385175006063,
+      "grad_norm": 0.5020492672920227,
+      "learning_rate": 0.0002,
+      "loss": 0.8121,
+      "step": 2960
+    },
+    {
+      "epoch": 0.48015520168135156,
+      "grad_norm": 0.5428946614265442,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48177188586209685,
+      "grad_norm": 0.5294089317321777,
+      "learning_rate": 0.0002,
+      "loss": 0.7849,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48338857004284214,
+      "grad_norm": 0.648289144039154,
+      "learning_rate": 0.0002,
+      "loss": 0.8553,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4850052542235874,
+      "grad_norm": 0.47916680574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.7874,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4866219384043327,
+      "grad_norm": 0.43849772214889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8087,
+      "step": 3010
+    },
+    {
+      "epoch": 0.488238622585078,
+      "grad_norm": 0.47007861733436584,
+      "learning_rate": 0.0002,
+      "loss": 0.7662,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4898553067658233,
+      "grad_norm": 0.6314331293106079,
+      "learning_rate": 0.0002,
+      "loss": 0.757,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49147199094656857,
+      "grad_norm": 0.49211493134498596,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49308867512731386,
+      "grad_norm": 0.4537973403930664,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49470535930805914,
+      "grad_norm": 0.47326919436454773,
+      "learning_rate": 0.0002,
+      "loss": 0.8095,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4963220434888045,
+      "grad_norm": 0.525874137878418,
+      "learning_rate": 0.0002,
+      "loss": 0.8447,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4979387276695498,
+      "grad_norm": 0.6361091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49955541185029506,
+      "grad_norm": 0.5850642919540405,
+      "learning_rate": 0.0002,
+      "loss": 0.821,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5011720960310403,
+      "grad_norm": 0.47299543023109436,
+      "learning_rate": 0.0002,
+      "loss": 0.8279,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5027887802117856,
+      "grad_norm": 0.473099946975708,
+      "learning_rate": 0.0002,
+      "loss": 0.8681,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5044054643925309,
+      "grad_norm": 0.48186397552490234,
+      "learning_rate": 0.0002,
+      "loss": 0.8223,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5060221485732762,
+      "grad_norm": 0.5015401840209961,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5076388327540216,
+      "grad_norm": 0.5617750287055969,
+      "learning_rate": 0.0002,
+      "loss": 0.7692,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5092555169347668,
+      "grad_norm": 0.5169327259063721,
+      "learning_rate": 0.0002,
+      "loss": 0.8708,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5108722011155121,
+      "grad_norm": 0.545657753944397,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5124888852962574,
+      "grad_norm": 0.512864351272583,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5141055694770027,
+      "grad_norm": 0.4113546311855316,
+      "learning_rate": 0.0002,
+      "loss": 0.7794,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5157222536577479,
+      "grad_norm": 0.44532445073127747,
+      "learning_rate": 0.0002,
+      "loss": 0.8206,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5173389378384933,
+      "grad_norm": 0.5623497366905212,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5189556220192385,
+      "grad_norm": 0.5084741115570068,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5205723061999838,
+      "grad_norm": 0.5305403470993042,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5221889903807291,
+      "grad_norm": 0.4708254337310791,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5238056745614744,
+      "grad_norm": 0.43827131390571594,
+      "learning_rate": 0.0002,
+      "loss": 0.7639,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5254223587422197,
+      "grad_norm": 0.5630002617835999,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 3250
+    },
+    {
+      "epoch": 0.527039042922965,
+      "grad_norm": 0.5010961890220642,
+      "learning_rate": 0.0002,
+      "loss": 0.7522,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5286557271037103,
+      "grad_norm": 0.6303122043609619,
+      "learning_rate": 0.0002,
+      "loss": 0.8374,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5302724112844556,
+      "grad_norm": 0.5107331275939941,
+      "learning_rate": 0.0002,
+      "loss": 0.7727,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5318890954652009,
+      "grad_norm": 0.5700443387031555,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5335057796459461,
+      "grad_norm": 0.46296367049217224,
+      "learning_rate": 0.0002,
+      "loss": 0.7776,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5351224638266915,
+      "grad_norm": 0.531568706035614,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5367391480074367,
+      "grad_norm": 0.4686741530895233,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5383558321881821,
+      "grad_norm": 0.5404331088066101,
+      "learning_rate": 0.0002,
+      "loss": 0.8104,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5399725163689273,
+      "grad_norm": 0.6368790864944458,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5415892005496726,
+      "grad_norm": 0.42300888895988464,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5432058847304179,
+      "grad_norm": 0.5362542867660522,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5448225689111632,
+      "grad_norm": 0.497128963470459,
+      "learning_rate": 0.0002,
+      "loss": 0.858,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5464392530919085,
+      "grad_norm": 0.5006386041641235,
+      "learning_rate": 0.0002,
+      "loss": 0.8519,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5480559372726538,
+      "grad_norm": 0.44136837124824524,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5496726214533991,
+      "grad_norm": 0.5897833108901978,
+      "learning_rate": 0.0002,
+      "loss": 0.773,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5512893056341444,
+      "grad_norm": 0.641075611114502,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5529059898148897,
+      "grad_norm": 0.7251322269439697,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5545226739956349,
+      "grad_norm": 0.47411349415779114,
+      "learning_rate": 0.0002,
+      "loss": 0.7626,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5561393581763803,
+      "grad_norm": 0.4994310438632965,
+      "learning_rate": 0.0002,
+      "loss": 0.8196,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5577560423571255,
+      "grad_norm": 0.5814438462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5593727265378708,
+      "grad_norm": 0.6278898119926453,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5609894107186161,
+      "grad_norm": 0.46208274364471436,
+      "learning_rate": 0.0002,
+      "loss": 0.813,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5626060948993614,
+      "grad_norm": 0.5718930959701538,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5642227790801067,
+      "grad_norm": 0.48178744316101074,
+      "learning_rate": 0.0002,
+      "loss": 0.8152,
+      "step": 3490
+    },
+    {
+      "epoch": 0.565839463260852,
+      "grad_norm": 0.47336965799331665,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5674561474415973,
+      "grad_norm": 0.43442684412002563,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5690728316223426,
+      "grad_norm": 0.6463358998298645,
+      "learning_rate": 0.0002,
+      "loss": 0.7564,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5706895158030879,
+      "grad_norm": 0.5286486744880676,
+      "learning_rate": 0.0002,
+      "loss": 0.836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5723061999838331,
+      "grad_norm": 0.5405499935150146,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5739228841645785,
+      "grad_norm": 0.6654391884803772,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.5081980228424072,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5771562525260691,
+      "grad_norm": 0.48978179693222046,
+      "learning_rate": 0.0002,
+      "loss": 0.7753,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5787729367068143,
+      "grad_norm": 0.5840612053871155,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5803896208875596,
+      "grad_norm": 0.5235261917114258,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5820063050683049,
+      "grad_norm": 0.5672075748443604,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5836229892490502,
+      "grad_norm": 0.5613429546356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5852396734297954,
+      "grad_norm": 0.4032273590564728,
+      "learning_rate": 0.0002,
+      "loss": 0.8274,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5868563576105408,
+      "grad_norm": 0.49559324979782104,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5884730417912861,
+      "grad_norm": 0.6895697712898254,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5900897259720314,
+      "grad_norm": 0.4750136435031891,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5917064101527767,
+      "grad_norm": 0.5176819562911987,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5933230943335219,
+      "grad_norm": 0.5817760229110718,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5949397785142673,
+      "grad_norm": 0.6064626574516296,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5965564626950125,
+      "grad_norm": 0.6728700995445251,
+      "learning_rate": 0.0002,
+      "loss": 0.8422,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5981731468757578,
+      "grad_norm": 0.609305202960968,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5997898310565031,
+      "grad_norm": 0.4615488350391388,
+      "learning_rate": 0.0002,
+      "loss": 0.8048,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6014065152372484,
+      "grad_norm": 2.0531179904937744,
+      "learning_rate": 0.0002,
+      "loss": 0.8214,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6030231994179936,
+      "grad_norm": 0.5091132521629333,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 3730
+    },
+    {
+      "epoch": 0.604639883598739,
+      "grad_norm": 0.5951124429702759,
+      "learning_rate": 0.0002,
+      "loss": 0.7833,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6062565677794842,
+      "grad_norm": 0.5870208144187927,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6078732519602296,
+      "grad_norm": 0.6254619359970093,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6094899361409749,
+      "grad_norm": 0.5577626824378967,
+      "learning_rate": 0.0002,
+      "loss": 0.7868,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6111066203217201,
+      "grad_norm": 0.5004405379295349,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6127233045024655,
+      "grad_norm": 0.5527383685112,
+      "learning_rate": 0.0002,
+      "loss": 0.8092,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6143399886832107,
+      "grad_norm": 0.49116113781929016,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6159566728639561,
+      "grad_norm": 0.5299299359321594,
+      "learning_rate": 0.0002,
+      "loss": 0.8352,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6175733570447013,
+      "grad_norm": 0.464897483587265,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6191900412254466,
+      "grad_norm": 0.6505740880966187,
+      "learning_rate": 0.0002,
+      "loss": 0.7923,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6208067254061919,
+      "grad_norm": 0.5512559413909912,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6224234095869372,
+      "grad_norm": 0.49427518248558044,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6240400937676824,
+      "grad_norm": 0.3839147090911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6256567779484278,
+      "grad_norm": 0.5760218501091003,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6272734621291731,
+      "grad_norm": 0.7226507067680359,
+      "learning_rate": 0.0002,
+      "loss": 0.7836,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6288901463099184,
+      "grad_norm": 0.676781415939331,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6305068304906637,
+      "grad_norm": 0.4284018278121948,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6321235146714089,
+      "grad_norm": 0.5060628056526184,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6337401988521543,
+      "grad_norm": 0.5524522066116333,
+      "learning_rate": 0.0002,
+      "loss": 0.8089,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6353568830328995,
+      "grad_norm": 0.6099881529808044,
+      "learning_rate": 0.0002,
+      "loss": 0.8276,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6369735672136448,
+      "grad_norm": 0.43155938386917114,
+      "learning_rate": 0.0002,
+      "loss": 0.809,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6385902513943901,
+      "grad_norm": 0.6427084803581238,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6402069355751354,
+      "grad_norm": 0.541220486164093,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6418236197558806,
+      "grad_norm": 0.5414294600486755,
+      "learning_rate": 0.0002,
+      "loss": 0.8539,
+      "step": 3970
+    },
+    {
+      "epoch": 0.643440303936626,
+      "grad_norm": 0.46344003081321716,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6450569881173712,
+      "grad_norm": 0.45209285616874695,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6466736722981166,
+      "grad_norm": 0.5417284369468689,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6482903564788619,
+      "grad_norm": 0.7995685935020447,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6499070406596071,
+      "grad_norm": 0.6384002566337585,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6515237248403525,
+      "grad_norm": 0.4472815692424774,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6531404090210977,
+      "grad_norm": 0.6834294199943542,
+      "learning_rate": 0.0002,
+      "loss": 0.8071,
+      "step": 4040
+    },
+    {
+      "epoch": 0.654757093201843,
+      "grad_norm": 0.4612339735031128,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6563737773825883,
+      "grad_norm": 0.9266576170921326,
+      "learning_rate": 0.0002,
+      "loss": 0.8141,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6579904615633336,
+      "grad_norm": 0.4470861852169037,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6596071457440789,
+      "grad_norm": 0.45544925332069397,
+      "learning_rate": 0.0002,
+      "loss": 0.8293,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6612238299248242,
+      "grad_norm": 0.6144481301307678,
+      "learning_rate": 0.0002,
+      "loss": 0.8455,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6628405141055694,
+      "grad_norm": 0.5936288237571716,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6644571982863148,
+      "grad_norm": 0.4822963774204254,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 4110
+    },
+    {
+      "epoch": 0.66607388246706,
+      "grad_norm": 0.48432496190071106,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6676905666478054,
+      "grad_norm": 0.4901607930660248,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6693072508285507,
+      "grad_norm": 0.5018393397331238,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6709239350092959,
+      "grad_norm": 0.6946378946304321,
+      "learning_rate": 0.0002,
+      "loss": 0.8065,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6725406191900413,
+      "grad_norm": 0.5997390747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.6738849878311157,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6757739875515318,
+      "grad_norm": 0.6110581159591675,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6773906717322771,
+      "grad_norm": 0.5703322291374207,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6790073559130224,
+      "grad_norm": 0.4686066210269928,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6806240400937676,
+      "grad_norm": 0.6394643783569336,
+      "learning_rate": 0.0002,
+      "loss": 0.8441,
+      "step": 4210
+    },
+    {
+      "epoch": 0.682240724274513,
+      "grad_norm": 0.5454841256141663,
+      "learning_rate": 0.0002,
+      "loss": 0.8011,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6838574084552582,
+      "grad_norm": 0.4859732985496521,
+      "learning_rate": 0.0002,
+      "loss": 0.8307,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6854740926360036,
+      "grad_norm": 0.5544065833091736,
+      "learning_rate": 0.0002,
+      "loss": 0.8161,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6870907768167488,
+      "grad_norm": 0.4902505576610565,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6887074609974941,
+      "grad_norm": 0.4768051505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7977,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6903241451782395,
+      "grad_norm": 0.49982190132141113,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6919408293589847,
+      "grad_norm": 0.6351838111877441,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 4280
+    },
+    {
+      "epoch": 0.69355751353973,
+      "grad_norm": 0.5647561550140381,
+      "learning_rate": 0.0002,
+      "loss": 0.7664,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6951741977204753,
+      "grad_norm": 0.5340486764907837,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6967908819012206,
+      "grad_norm": 0.5649092793464661,
+      "learning_rate": 0.0002,
+      "loss": 0.8526,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6984075660819659,
+      "grad_norm": 0.6183916926383972,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7000242502627112,
+      "grad_norm": 0.6154509782791138,
+      "learning_rate": 0.0002,
+      "loss": 0.792,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7016409344434564,
+      "grad_norm": 0.5156264305114746,
+      "learning_rate": 0.0002,
+      "loss": 0.8397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7032576186242018,
+      "grad_norm": 0.562171459197998,
+      "learning_rate": 0.0002,
+      "loss": 0.8512,
+      "step": 4350
+    },
+    {
+      "epoch": 0.704874302804947,
+      "grad_norm": 0.4949502646923065,
+      "learning_rate": 0.0002,
+      "loss": 0.7882,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7064909869856923,
+      "grad_norm": 0.5171684622764587,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7081076711664377,
+      "grad_norm": 0.6198443174362183,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7097243553471829,
+      "grad_norm": 0.5802276134490967,
+      "learning_rate": 0.0002,
+      "loss": 0.7606,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7113410395279283,
+      "grad_norm": 0.41096967458724976,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7129577237086735,
+      "grad_norm": 0.4397392272949219,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7145744078894188,
+      "grad_norm": 0.45228442549705505,
+      "learning_rate": 0.0002,
+      "loss": 0.7651,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7161910920701641,
+      "grad_norm": 0.4839673936367035,
+      "learning_rate": 0.0002,
+      "loss": 0.7938,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7178077762509094,
+      "grad_norm": 0.6140755414962769,
+      "learning_rate": 0.0002,
+      "loss": 0.8362,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 0.6841378808021545,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7210411446124,
+      "grad_norm": 0.6664239168167114,
+      "learning_rate": 0.0002,
+      "loss": 0.8177,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7226578287931452,
+      "grad_norm": 0.47552719712257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7983,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7242745129738906,
+      "grad_norm": 0.6649776101112366,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7258911971546358,
+      "grad_norm": 0.5159541964530945,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7275078813353811,
+      "grad_norm": 0.6693112850189209,
+      "learning_rate": 0.0002,
+      "loss": 0.7786,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7291245655161265,
+      "grad_norm": 0.48870977759361267,
+      "learning_rate": 0.0002,
+      "loss": 0.8655,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7307412496968717,
+      "grad_norm": 0.4857887923717499,
+      "learning_rate": 0.0002,
+      "loss": 0.7337,
+      "step": 4520
+    },
+    {
+      "epoch": 0.732357933877617,
+      "grad_norm": 0.5515662431716919,
+      "learning_rate": 0.0002,
+      "loss": 0.8026,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7339746180583623,
+      "grad_norm": 0.6292222738265991,
+      "learning_rate": 0.0002,
+      "loss": 0.8031,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7355913022391076,
+      "grad_norm": 0.48265689611434937,
+      "learning_rate": 0.0002,
+      "loss": 0.7749,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7372079864198529,
+      "grad_norm": 0.8044266104698181,
+      "learning_rate": 0.0002,
+      "loss": 0.8499,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7388246706005982,
+      "grad_norm": 0.6111769676208496,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7404413547813434,
+      "grad_norm": 0.5229553580284119,
+      "learning_rate": 0.0002,
+      "loss": 0.7291,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7420580389620888,
+      "grad_norm": 0.6054152250289917,
+      "learning_rate": 0.0002,
+      "loss": 0.8038,
+      "step": 4590
+    },
+    {
+      "epoch": 0.743674723142834,
+      "grad_norm": 0.5574966669082642,
+      "learning_rate": 0.0002,
+      "loss": 0.8169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7452914073235793,
+      "grad_norm": 0.5395817160606384,
+      "learning_rate": 0.0002,
+      "loss": 0.8439,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7469080915043246,
+      "grad_norm": 0.7116472721099854,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7485247756850699,
+      "grad_norm": 0.5618700981140137,
+      "learning_rate": 0.0002,
+      "loss": 0.7743,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7501414598658153,
+      "grad_norm": 0.5802770853042603,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7517581440465605,
+      "grad_norm": 0.5690428018569946,
+      "learning_rate": 0.0002,
+      "loss": 0.7924,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7533748282273058,
+      "grad_norm": 0.4813360273838043,
+      "learning_rate": 0.0002,
+      "loss": 0.8017,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7549915124080511,
+      "grad_norm": 0.5434042811393738,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7566081965887964,
+      "grad_norm": 0.5502099990844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7582248807695416,
+      "grad_norm": 0.6020621061325073,
+      "learning_rate": 0.0002,
+      "loss": 0.8598,
+      "step": 4690
+    },
+    {
+      "epoch": 0.759841564950287,
+      "grad_norm": 0.4922301471233368,
+      "learning_rate": 0.0002,
+      "loss": 0.7937,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7614582491310322,
+      "grad_norm": 0.6492828726768494,
+      "learning_rate": 0.0002,
+      "loss": 0.788,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7630749333117776,
+      "grad_norm": 0.4865580201148987,
+      "learning_rate": 0.0002,
+      "loss": 0.8313,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7646916174925228,
+      "grad_norm": 0.5971422791481018,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7663083016732681,
+      "grad_norm": 0.6832674145698547,
+      "learning_rate": 0.0002,
+      "loss": 0.8298,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7679249858540134,
+      "grad_norm": 0.500908613204956,
+      "learning_rate": 0.0002,
+      "loss": 0.8156,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7695416700347587,
+      "grad_norm": 0.6112465858459473,
+      "learning_rate": 0.0002,
+      "loss": 0.8383,
+      "step": 4760
+    },
+    {
+      "epoch": 0.771158354215504,
+      "grad_norm": 0.5753506422042847,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7727750383962493,
+      "grad_norm": 0.6529405117034912,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7743917225769946,
+      "grad_norm": 0.5916843414306641,
+      "learning_rate": 0.0002,
+      "loss": 0.8171,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7760084067577399,
+      "grad_norm": 0.4821224510669708,
+      "learning_rate": 0.0002,
+      "loss": 0.83,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7776250909384852,
+      "grad_norm": 0.5532580018043518,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7792417751192304,
+      "grad_norm": 0.4604877233505249,
+      "learning_rate": 0.0002,
+      "loss": 0.7363,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7808584592999758,
+      "grad_norm": 0.5009613037109375,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 4830
+    },
+    {
+      "epoch": 0.782475143480721,
+      "grad_norm": 0.6448560357093811,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7840918276614663,
+      "grad_norm": 0.44327953457832336,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7857085118422116,
+      "grad_norm": 0.5355411171913147,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7873251960229569,
+      "grad_norm": 0.5635677576065063,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7889418802037023,
+      "grad_norm": 0.5417491793632507,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7905585643844475,
+      "grad_norm": 0.4567430913448334,
+      "learning_rate": 0.0002,
+      "loss": 0.7819,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7921752485651928,
+      "grad_norm": 0.44651296734809875,
+      "learning_rate": 0.0002,
+      "loss": 0.8454,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7937919327459381,
+      "grad_norm": 0.5741217136383057,
+      "learning_rate": 0.0002,
+      "loss": 0.7959,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7954086169266834,
+      "grad_norm": 0.6605045199394226,
+      "learning_rate": 0.0002,
+      "loss": 0.8093,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7970253011074286,
+      "grad_norm": 0.5126531720161438,
+      "learning_rate": 0.0002,
+      "loss": 0.77,
+      "step": 4930
+    },
+    {
+      "epoch": 0.798641985288174,
+      "grad_norm": 0.513648271560669,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8002586694689192,
+      "grad_norm": 0.5350404381752014,
+      "learning_rate": 0.0002,
+      "loss": 0.8314,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8018753536496646,
+      "grad_norm": 0.5731674432754517,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8034920378304098,
+      "grad_norm": 0.5974258184432983,
+      "learning_rate": 0.0002,
+      "loss": 0.8572,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8051087220111551,
+      "grad_norm": 0.8774799704551697,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8067254061919004,
+      "grad_norm": 0.5994430184364319,
+      "learning_rate": 0.0002,
+      "loss": 0.7899,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8083420903726457,
+      "grad_norm": 0.4894903004169464,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5000
+    },
+    {
+      "epoch": 0.809958774553391,
+      "grad_norm": 0.5218459367752075,
+      "learning_rate": 0.0002,
+      "loss": 0.78,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8115754587341363,
+      "grad_norm": 0.5232468843460083,
+      "learning_rate": 0.0002,
+      "loss": 0.817,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8131921429148816,
+      "grad_norm": 0.44358372688293457,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8148088270956269,
+      "grad_norm": 0.6202037334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8164255112763722,
+      "grad_norm": 0.7721474170684814,
+      "learning_rate": 0.0002,
+      "loss": 0.7351,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8180421954571174,
+      "grad_norm": 0.5568501353263855,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8196588796378628,
+      "grad_norm": 0.49148809909820557,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 5070
+    },
+    {
+      "epoch": 0.821275563818608,
+      "grad_norm": 0.4956012964248657,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8228922479993533,
+      "grad_norm": 0.6078833937644958,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8245089321800986,
+      "grad_norm": 0.46906954050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.828,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8261256163608439,
+      "grad_norm": 0.50812166929245,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8277423005415891,
+      "grad_norm": 0.5319661498069763,
+      "learning_rate": 0.0002,
+      "loss": 0.8243,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8293589847223345,
+      "grad_norm": 0.4949689209461212,
+      "learning_rate": 0.0002,
+      "loss": 0.7798,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8309756689030798,
+      "grad_norm": 0.5151591300964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8325923530838251,
+      "grad_norm": 0.5530214309692383,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8342090372645704,
+      "grad_norm": 0.6297410130500793,
+      "learning_rate": 0.0002,
+      "loss": 0.8251,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8358257214453156,
+      "grad_norm": 0.5466840267181396,
+      "learning_rate": 0.0002,
+      "loss": 0.8067,
+      "step": 5170
+    },
+    {
+      "epoch": 0.837442405626061,
+      "grad_norm": 0.652913510799408,
+      "learning_rate": 0.0002,
+      "loss": 0.7875,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8390590898068062,
+      "grad_norm": 0.5811293125152588,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8406757739875516,
+      "grad_norm": 0.5109550952911377,
+      "learning_rate": 0.0002,
+      "loss": 0.7412,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8422924581682968,
+      "grad_norm": 0.4551706612110138,
+      "learning_rate": 0.0002,
+      "loss": 0.8077,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8439091423490421,
+      "grad_norm": 0.5813754200935364,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8455258265297874,
+      "grad_norm": 0.5856947898864746,
+      "learning_rate": 0.0002,
+      "loss": 0.802,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8471425107105327,
+      "grad_norm": 0.5482739210128784,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8487591948912779,
+      "grad_norm": 0.49023720622062683,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8503758790720233,
+      "grad_norm": 0.49472475051879883,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8519925632527686,
+      "grad_norm": 0.5490226745605469,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8536092474335139,
+      "grad_norm": 0.5340665578842163,
+      "learning_rate": 0.0002,
+      "loss": 0.8333,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8552259316142592,
+      "grad_norm": 0.5962483882904053,
+      "learning_rate": 0.0002,
+      "loss": 0.8277,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8568426157950044,
+      "grad_norm": 0.586358368396759,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8584592999757498,
+      "grad_norm": 0.49120277166366577,
+      "learning_rate": 0.0002,
+      "loss": 0.7831,
+      "step": 5310
+    },
+    {
+      "epoch": 0.860075984156495,
+      "grad_norm": 0.5887332558631897,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8616926683372403,
+      "grad_norm": 0.42496153712272644,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.5489874482154846,
+      "learning_rate": 0.0002,
+      "loss": 0.7905,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8649260366987309,
+      "grad_norm": 0.5850813984870911,
+      "learning_rate": 0.0002,
+      "loss": 0.7958,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8665427208794761,
+      "grad_norm": 0.517487108707428,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8681594050602215,
+      "grad_norm": 0.5339142680168152,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8697760892409668,
+      "grad_norm": 0.6236387491226196,
+      "learning_rate": 0.0002,
+      "loss": 0.818,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8713927734217121,
+      "grad_norm": 0.5752192735671997,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8730094576024574,
+      "grad_norm": 0.6724614500999451,
+      "learning_rate": 0.0002,
+      "loss": 0.8542,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8746261417832026,
+      "grad_norm": 0.5280613303184509,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 5410
+    },
+    {
+      "epoch": 0.876242825963948,
+      "grad_norm": 0.44033288955688477,
+      "learning_rate": 0.0002,
+      "loss": 0.8231,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8778595101446932,
+      "grad_norm": 0.5199708342552185,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8794761943254386,
+      "grad_norm": 0.46778348088264465,
+      "learning_rate": 0.0002,
+      "loss": 0.7852,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8810928785061838,
+      "grad_norm": 0.4657754898071289,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8827095626869291,
+      "grad_norm": 0.5472902655601501,
+      "learning_rate": 0.0002,
+      "loss": 0.7799,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8843262468676744,
+      "grad_norm": 0.4876766800880432,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8859429310484197,
+      "grad_norm": 0.5057248473167419,
+      "learning_rate": 0.0002,
+      "loss": 0.7906,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8875596152291649,
+      "grad_norm": 0.4637320637702942,
+      "learning_rate": 0.0002,
+      "loss": 0.8124,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8891762994099103,
+      "grad_norm": 0.471955806016922,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8907929835906556,
+      "grad_norm": 0.5209813714027405,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8924096677714008,
+      "grad_norm": 0.6213834285736084,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8940263519521462,
+      "grad_norm": 0.5215408205986023,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8956430361328914,
+      "grad_norm": 0.580478310585022,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8972597203136368,
+      "grad_norm": 0.49102169275283813,
+      "learning_rate": 0.0002,
+      "loss": 0.8371,
+      "step": 5550
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.6043479442596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7806,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9004930886751273,
+      "grad_norm": 0.5636463165283203,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9021097728558726,
+      "grad_norm": 0.5620124340057373,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9037264570366179,
+      "grad_norm": 0.5206354856491089,
+      "learning_rate": 0.0002,
+      "loss": 0.8083,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9053431412173631,
+      "grad_norm": 0.5798229575157166,
+      "learning_rate": 0.0002,
+      "loss": 0.8557,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9069598253981085,
+      "grad_norm": 0.6428212523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.8097,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9085765095788537,
+      "grad_norm": 0.48064687848091125,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9101931937595991,
+      "grad_norm": 0.6347860097885132,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9118098779403444,
+      "grad_norm": 0.5353913307189941,
+      "learning_rate": 0.0002,
+      "loss": 0.851,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9134265621210896,
+      "grad_norm": 0.5323944091796875,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5650
+    },
+    {
+      "epoch": 0.915043246301835,
+      "grad_norm": 0.5261843204498291,
+      "learning_rate": 0.0002,
+      "loss": 0.8393,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9166599304825802,
+      "grad_norm": 0.5451326966285706,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9182766146633256,
+      "grad_norm": 0.5183324217796326,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9198932988440708,
+      "grad_norm": 0.47229018807411194,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9215099830248161,
+      "grad_norm": 0.49180513620376587,
+      "learning_rate": 0.0002,
+      "loss": 0.7757,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9231266672055614,
+      "grad_norm": 0.5419785380363464,
+      "learning_rate": 0.0002,
+      "loss": 0.8735,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9247433513863067,
+      "grad_norm": 0.5408698916435242,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9263600355670519,
+      "grad_norm": 0.5286232829093933,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9279767197477973,
+      "grad_norm": 0.7539758086204529,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9295934039285425,
+      "grad_norm": 0.5166944861412048,
+      "learning_rate": 0.0002,
+      "loss": 0.8118,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9312100881092878,
+      "grad_norm": 0.6601425409317017,
+      "learning_rate": 0.0002,
+      "loss": 0.783,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9328267722900332,
+      "grad_norm": 0.5029960870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9344434564707784,
+      "grad_norm": 0.4926645755767822,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9360601406515238,
+      "grad_norm": 0.5739615559577942,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5790
+    },
+    {
+      "epoch": 0.937676824832269,
+      "grad_norm": 0.5058279037475586,
+      "learning_rate": 0.0002,
+      "loss": 0.8037,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9392935090130143,
+      "grad_norm": 0.5260962247848511,
+      "learning_rate": 0.0002,
+      "loss": 0.8537,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9409101931937596,
+      "grad_norm": 0.5768588185310364,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9425268773745049,
+      "grad_norm": 0.5170126557350159,
+      "learning_rate": 0.0002,
+      "loss": 0.8215,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9441435615552501,
+      "grad_norm": 0.5745864510536194,
+      "learning_rate": 0.0002,
+      "loss": 0.7422,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9457602457359955,
+      "grad_norm": 0.5551357865333557,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9473769299167407,
+      "grad_norm": 0.5776078701019287,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9489936140974861,
+      "grad_norm": 0.5340062379837036,
+      "learning_rate": 0.0002,
+      "loss": 0.8527,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9506102982782314,
+      "grad_norm": 0.6447290182113647,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9522269824589766,
+      "grad_norm": 0.5123815536499023,
+      "learning_rate": 0.0002,
+      "loss": 0.7945,
+      "step": 5890
+    },
+    {
+      "epoch": 0.953843666639722,
+      "grad_norm": 0.48547613620758057,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9554603508204672,
+      "grad_norm": 0.5791414976119995,
+      "learning_rate": 0.0002,
+      "loss": 0.7896,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9570770350012126,
+      "grad_norm": 0.6195011734962463,
+      "learning_rate": 0.0002,
+      "loss": 0.8408,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9586937191819578,
+      "grad_norm": 0.6323803067207336,
+      "learning_rate": 0.0002,
+      "loss": 0.7805,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9603104033627031,
+      "grad_norm": 0.45552879571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9619270875434484,
+      "grad_norm": 0.5796473622322083,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9635437717241937,
+      "grad_norm": 0.647261381149292,
+      "learning_rate": 0.0002,
+      "loss": 0.7672,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9651604559049389,
+      "grad_norm": 0.5487682819366455,
+      "learning_rate": 0.0002,
+      "loss": 0.8086,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9667771400856843,
+      "grad_norm": 0.5743663907051086,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9683938242664295,
+      "grad_norm": 0.5470591187477112,
+      "learning_rate": 0.0002,
+      "loss": 0.8153,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9700105084471748,
+      "grad_norm": 0.5901660323143005,
+      "learning_rate": 0.0002,
+      "loss": 0.8119,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9716271926279202,
+      "grad_norm": 0.6544759273529053,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9732438768086654,
+      "grad_norm": 0.6288470029830933,
+      "learning_rate": 0.0002,
+      "loss": 0.7536,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9748605609894108,
+      "grad_norm": 0.673153817653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 6030
+    },
+    {
+      "epoch": 0.976477245170156,
+      "grad_norm": 0.42854753136634827,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9780939293509013,
+      "grad_norm": 0.5227066278457642,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9797106135316466,
+      "grad_norm": 0.5372416973114014,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9813272977123919,
+      "grad_norm": 0.6026402115821838,
+      "learning_rate": 0.0002,
+      "loss": 0.7591,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9829439818931371,
+      "grad_norm": 0.49547791481018066,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9845606660738825,
+      "grad_norm": 0.4641951322555542,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9861773502546277,
+      "grad_norm": 0.5818535089492798,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9877940344353731,
+      "grad_norm": 0.63955157995224,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9894107186161183,
+      "grad_norm": 0.5649438500404358,
+      "learning_rate": 0.0002,
+      "loss": 0.7547,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9910274027968636,
+      "grad_norm": 0.5290433168411255,
+      "learning_rate": 0.0002,
+      "loss": 0.7861,
+      "step": 6130
+    },
+    {
+      "epoch": 0.992644086977609,
+      "grad_norm": 0.6399374008178711,
+      "learning_rate": 0.0002,
+      "loss": 0.8109,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9942607711583542,
+      "grad_norm": 0.6736576557159424,
+      "learning_rate": 0.0002,
+      "loss": 0.8373,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9958774553390995,
+      "grad_norm": 0.515420138835907,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9974941395198448,
+      "grad_norm": 0.562677800655365,
+      "learning_rate": 0.0002,
+      "loss": 0.8032,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9991108237005901,
+      "grad_norm": 0.7113858461380005,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9999191657909627,
+      "eval_loss": 1.0871200561523438,
+      "eval_runtime": 122.2071,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0007275078813354,
+      "grad_norm": 0.7111801505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0023441920620806,
+      "grad_norm": 0.5402125716209412,
+      "learning_rate": 0.0002,
+      "loss": 0.6865,
+      "step": 6200
+    },
+    {
+      "epoch": 1.003960876242826,
+      "grad_norm": 0.6098830103874207,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0055775604235713,
+      "grad_norm": 0.5829983353614807,
+      "learning_rate": 0.0002,
+      "loss": 0.7631,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0071942446043165,
+      "grad_norm": 0.5614621043205261,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0088109287850617,
+      "grad_norm": 0.5954238772392273,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0104276129658072,
+      "grad_norm": 0.6480574607849121,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0120442971465524,
+      "grad_norm": 0.6051128506660461,
+      "learning_rate": 0.0002,
+      "loss": 0.7514,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0136609813272976,
+      "grad_norm": 0.6318870782852173,
+      "learning_rate": 0.0002,
+      "loss": 0.7237,
+      "step": 6270
+    },
+    {
+      "epoch": 1.015277665508043,
+      "grad_norm": 0.5048980116844177,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0168943496887883,
+      "grad_norm": 0.6346936225891113,
+      "learning_rate": 0.0002,
+      "loss": 0.7391,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0185110338695336,
+      "grad_norm": 0.5711665749549866,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0201277180502788,
+      "grad_norm": 0.5175361037254333,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0217444022310243,
+      "grad_norm": 0.5360831618309021,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0233610864117695,
+      "grad_norm": 0.614675760269165,
+      "learning_rate": 0.0002,
+      "loss": 0.7112,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0249777705925147,
+      "grad_norm": 0.5626118183135986,
+      "learning_rate": 0.0002,
+      "loss": 0.7748,
+      "step": 6340
+    },
+    {
+      "epoch": 1.02659445477326,
+      "grad_norm": 0.574897289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.7375,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0282111389540054,
+      "grad_norm": 0.7185447812080383,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0298278231347506,
+      "grad_norm": 0.6705799698829651,
+      "learning_rate": 0.0002,
+      "loss": 0.703,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0314445073154959,
+      "grad_norm": 0.6740428805351257,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0330611914962413,
+      "grad_norm": 0.663902759552002,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0346778756769865,
+      "grad_norm": 0.5029543042182922,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0362945598577318,
+      "grad_norm": 0.7813863158226013,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 6410
+    },
+    {
+      "epoch": 1.037911244038477,
+      "grad_norm": 0.5396282076835632,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0395279282192225,
+      "grad_norm": 0.5253293514251709,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0411446123999677,
+      "grad_norm": 0.7236770987510681,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 6440
+    },
+    {
+      "epoch": 1.042761296580713,
+      "grad_norm": 0.5670917630195618,
+      "learning_rate": 0.0002,
+      "loss": 0.7259,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0443779807614582,
+      "grad_norm": 0.6031978726387024,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0459946649422036,
+      "grad_norm": 0.5309213399887085,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0476113491229488,
+      "grad_norm": 0.7114651799201965,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 6480
+    },
+    {
+      "epoch": 1.049228033303694,
+      "grad_norm": 0.5591610670089722,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0508447174844395,
+      "grad_norm": 0.5185961127281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0524614016651848,
+      "grad_norm": 0.6510552167892456,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6510
+    },
+    {
+      "epoch": 1.05407808584593,
+      "grad_norm": 0.6557928919792175,
+      "learning_rate": 0.0002,
+      "loss": 0.7057,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0556947700266752,
+      "grad_norm": 0.6973192691802979,
+      "learning_rate": 0.0002,
+      "loss": 0.8056,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0573114542074207,
+      "grad_norm": 0.6226583123207092,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 6540
+    },
+    {
+      "epoch": 1.058928138388166,
+      "grad_norm": 0.5633195638656616,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0605448225689111,
+      "grad_norm": 0.7466658353805542,
+      "learning_rate": 0.0002,
+      "loss": 0.7082,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0621615067496564,
+      "grad_norm": 0.6462772488594055,
+      "learning_rate": 0.0002,
+      "loss": 0.7059,
+      "step": 6570
+    },
+    {
+      "epoch": 1.0637781909304018,
+      "grad_norm": 0.5266856551170349,
+      "learning_rate": 0.0002,
+      "loss": 0.7046,
+      "step": 6580
+    },
+    {
+      "epoch": 1.065394875111147,
+      "grad_norm": 0.534392774105072,
+      "learning_rate": 0.0002,
+      "loss": 0.7157,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0670115592918923,
+      "grad_norm": 0.7514177560806274,
+      "learning_rate": 0.0002,
+      "loss": 0.7115,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0686282434726375,
+      "grad_norm": 0.7593035697937012,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 6610
+    },
+    {
+      "epoch": 1.070244927653383,
+      "grad_norm": 0.5277858972549438,
+      "learning_rate": 0.0002,
+      "loss": 0.6836,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0718616118341282,
+      "grad_norm": 0.5573670268058777,
+      "learning_rate": 0.0002,
+      "loss": 0.7405,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0734782960148734,
+      "grad_norm": 0.6802396774291992,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 6640
+    },
+    {
+      "epoch": 1.0750949801956189,
+      "grad_norm": 0.7367215752601624,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0767116643763641,
+      "grad_norm": 0.5961891412734985,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0783283485571094,
+      "grad_norm": 0.5736313462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0799450327378546,
+      "grad_norm": 0.619219183921814,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0815617169186,
+      "grad_norm": 0.6214390993118286,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0831784010993453,
+      "grad_norm": 0.564536988735199,
+      "learning_rate": 0.0002,
+      "loss": 0.7554,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0847950852800905,
+      "grad_norm": 0.5838140249252319,
+      "learning_rate": 0.0002,
+      "loss": 0.7362,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0864117694608357,
+      "grad_norm": 0.7000553607940674,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0880284536415812,
+      "grad_norm": 0.7078263759613037,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 6730
+    },
+    {
+      "epoch": 1.0896451378223264,
+      "grad_norm": 0.8353848457336426,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6740
+    },
+    {
+      "epoch": 1.0912618220030716,
+      "grad_norm": 0.5615518689155579,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 6750
+    },
+    {
+      "epoch": 1.0928785061838169,
+      "grad_norm": 0.5475581288337708,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 6760
+    },
+    {
+      "epoch": 1.0944951903645623,
+      "grad_norm": 0.5835978388786316,
+      "learning_rate": 0.0002,
+      "loss": 0.7652,
+      "step": 6770
+    },
+    {
+      "epoch": 1.0961118745453076,
+      "grad_norm": 0.5516105890274048,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 6780
+    },
+    {
+      "epoch": 1.0977285587260528,
+      "grad_norm": 0.5875251889228821,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 6790
+    },
+    {
+      "epoch": 1.0993452429067982,
+      "grad_norm": 0.7376947999000549,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1009619270875435,
+      "grad_norm": 0.5656165480613708,
+      "learning_rate": 0.0002,
+      "loss": 0.7512,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1025786112682887,
+      "grad_norm": 0.6365954279899597,
+      "learning_rate": 0.0002,
+      "loss": 0.7409,
+      "step": 6820
+    },
+    {
+      "epoch": 1.104195295449034,
+      "grad_norm": 0.5033080577850342,
+      "learning_rate": 0.0002,
+      "loss": 0.7392,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1058119796297794,
+      "grad_norm": 0.617396891117096,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1074286638105246,
+      "grad_norm": 0.6395374536514282,
+      "learning_rate": 0.0002,
+      "loss": 0.7006,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1090453479912699,
+      "grad_norm": 0.6775295734405518,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1106620321720153,
+      "grad_norm": 0.6655223965644836,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1122787163527605,
+      "grad_norm": 0.676655113697052,
+      "learning_rate": 0.0002,
+      "loss": 0.7553,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1138954005335058,
+      "grad_norm": 0.6062718629837036,
+      "learning_rate": 0.0002,
+      "loss": 0.7342,
+      "step": 6890
+    },
+    {
+      "epoch": 1.115512084714251,
+      "grad_norm": 0.590943455696106,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1171287688949965,
+      "grad_norm": 0.6315317153930664,
+      "learning_rate": 0.0002,
+      "loss": 0.6705,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1187454530757417,
+      "grad_norm": 0.47979024052619934,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 6920
+    },
+    {
+      "epoch": 1.120362137256487,
+      "grad_norm": 0.647298276424408,
+      "learning_rate": 0.0002,
+      "loss": 0.7002,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1219788214372322,
+      "grad_norm": 0.7336484789848328,
+      "learning_rate": 0.0002,
+      "loss": 0.7502,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1235955056179776,
+      "grad_norm": 0.5071424245834351,
+      "learning_rate": 0.0002,
+      "loss": 0.693,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1252121897987228,
+      "grad_norm": 0.6527144312858582,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 6960
+    },
+    {
+      "epoch": 1.126828873979468,
+      "grad_norm": 0.6935935020446777,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1284455581602133,
+      "grad_norm": 0.8026931881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1300622423409588,
+      "grad_norm": 0.5210393667221069,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 6990
+    },
+    {
+      "epoch": 1.131678926521704,
+      "grad_norm": 0.60475093126297,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1332956107024492,
+      "grad_norm": 0.6417073607444763,
+      "learning_rate": 0.0002,
+      "loss": 0.7495,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1349122948831947,
+      "grad_norm": 0.6732175946235657,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 7020
+    },
+    {
+      "epoch": 1.13652897906394,
+      "grad_norm": 0.6719491481781006,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1381456632446851,
+      "grad_norm": 0.5708295106887817,
+      "learning_rate": 0.0002,
+      "loss": 0.7694,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1397623474254304,
+      "grad_norm": 0.7141719460487366,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1413790316061758,
+      "grad_norm": 0.6187017560005188,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 7060
+    },
+    {
+      "epoch": 1.142995715786921,
+      "grad_norm": 0.50581294298172,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1446123999676663,
+      "grad_norm": 0.5620143413543701,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1462290841484115,
+      "grad_norm": 0.6231929659843445,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 7090
+    },
+    {
+      "epoch": 1.147845768329157,
+      "grad_norm": 0.5775774121284485,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1494624525099022,
+      "grad_norm": 0.6492809653282166,
+      "learning_rate": 0.0002,
+      "loss": 0.7728,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1510791366906474,
+      "grad_norm": 0.6434972286224365,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 7120
+    },
+    {
+      "epoch": 1.1526958208713927,
+      "grad_norm": 0.6191812753677368,
+      "learning_rate": 0.0002,
+      "loss": 0.7374,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1543125050521381,
+      "grad_norm": 0.6690331697463989,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 7140
+    },
+    {
+      "epoch": 1.1559291892328833,
+      "grad_norm": 0.5977938175201416,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1575458734136286,
+      "grad_norm": 0.6195854544639587,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7160
+    },
+    {
+      "epoch": 1.159162557594374,
+      "grad_norm": 0.5752048492431641,
+      "learning_rate": 0.0002,
+      "loss": 0.7249,
+      "step": 7170
+    },
+    {
+      "epoch": 1.1607792417751193,
+      "grad_norm": 0.589081883430481,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1623959259558645,
+      "grad_norm": 0.756996750831604,
+      "learning_rate": 0.0002,
+      "loss": 0.704,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1640126101366097,
+      "grad_norm": 0.7614967226982117,
+      "learning_rate": 0.0002,
+      "loss": 0.7404,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1656292943173552,
+      "grad_norm": 0.6120437979698181,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1672459784981004,
+      "grad_norm": 0.6210004687309265,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1688626626788456,
+      "grad_norm": 0.6044116020202637,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7230
+    },
+    {
+      "epoch": 1.170479346859591,
+      "grad_norm": 0.5418457388877869,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1720960310403363,
+      "grad_norm": 0.6413537263870239,
+      "learning_rate": 0.0002,
+      "loss": 0.6938,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1737127152210816,
+      "grad_norm": 0.5777867436408997,
+      "learning_rate": 0.0002,
+      "loss": 0.6978,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1753293994018268,
+      "grad_norm": 0.7092402577400208,
+      "learning_rate": 0.0002,
+      "loss": 0.7503,
+      "step": 7270
+    },
+    {
+      "epoch": 1.176946083582572,
+      "grad_norm": 0.6351709365844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1785627677633175,
+      "grad_norm": 0.6172189712524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7527,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1801794519440627,
+      "grad_norm": 0.6801714897155762,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7300
+    },
+    {
+      "epoch": 1.181796136124808,
+      "grad_norm": 0.6044712066650391,
+      "learning_rate": 0.0002,
+      "loss": 0.6941,
+      "step": 7310
+    },
+    {
+      "epoch": 1.1834128203055534,
+      "grad_norm": 0.7413212060928345,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 7320
+    },
+    {
+      "epoch": 1.1850295044862986,
+      "grad_norm": 0.5303856134414673,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 7330
+    },
+    {
+      "epoch": 1.1866461886670439,
+      "grad_norm": 0.5647098422050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6915,
+      "step": 7340
+    },
+    {
+      "epoch": 1.188262872847789,
+      "grad_norm": 0.7374135255813599,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 7350
+    },
+    {
+      "epoch": 1.1898795570285345,
+      "grad_norm": 0.5710089206695557,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1914962412092798,
+      "grad_norm": 0.6073619723320007,
+      "learning_rate": 0.0002,
+      "loss": 0.8289,
+      "step": 7370
+    },
+    {
+      "epoch": 1.193112925390025,
+      "grad_norm": 0.5899916887283325,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1947296095707705,
+      "grad_norm": 0.7762434482574463,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1963462937515157,
+      "grad_norm": 0.679949939250946,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7400
+    },
+    {
+      "epoch": 1.197962977932261,
+      "grad_norm": 0.6106849312782288,
+      "learning_rate": 0.0002,
+      "loss": 0.7599,
+      "step": 7410
+    },
+    {
+      "epoch": 1.1995796621130062,
+      "grad_norm": 0.682461678981781,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2011963462937516,
+      "grad_norm": 0.6087017059326172,
+      "learning_rate": 0.0002,
+      "loss": 0.7741,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2028130304744968,
+      "grad_norm": 0.63739013671875,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 7440
+    },
+    {
+      "epoch": 1.204429714655242,
+      "grad_norm": 0.6154777407646179,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2060463988359873,
+      "grad_norm": 0.7491534948348999,
+      "learning_rate": 0.0002,
+      "loss": 0.7565,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2076630830167328,
+      "grad_norm": 0.6664797067642212,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 7470
+    },
+    {
+      "epoch": 1.209279767197478,
+      "grad_norm": 0.6660266518592834,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2108964513782232,
+      "grad_norm": 0.6972551345825195,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2125131355589684,
+      "grad_norm": 0.6157945990562439,
+      "learning_rate": 0.0002,
+      "loss": 0.7023,
+      "step": 7500
+    },
+    {
+      "epoch": 1.214129819739714,
+      "grad_norm": 0.5199310183525085,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2157465039204591,
+      "grad_norm": 0.577610433101654,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2173631881012044,
+      "grad_norm": 0.53652423620224,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2189798722819498,
+      "grad_norm": 0.6479050517082214,
+      "learning_rate": 0.0002,
+      "loss": 0.7393,
+      "step": 7540
+    },
+    {
+      "epoch": 1.220596556462695,
+      "grad_norm": 0.618748128414154,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2222132406434403,
+      "grad_norm": 0.6311424374580383,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2238299248241855,
+      "grad_norm": 0.6595825552940369,
+      "learning_rate": 0.0002,
+      "loss": 0.7272,
+      "step": 7570
+    },
+    {
+      "epoch": 1.225446609004931,
+      "grad_norm": 0.5198960900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2270632931856762,
+      "grad_norm": 0.578650712966919,
+      "learning_rate": 0.0002,
+      "loss": 0.674,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2286799773664214,
+      "grad_norm": 0.6080220937728882,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2302966615471669,
+      "grad_norm": 0.7050248384475708,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2319133457279121,
+      "grad_norm": 0.6652196049690247,
+      "learning_rate": 0.0002,
+      "loss": 0.7032,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2335300299086573,
+      "grad_norm": 0.7322776317596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7085,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2351467140894026,
+      "grad_norm": 0.4998728036880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2367633982701478,
+      "grad_norm": 0.6428788900375366,
+      "learning_rate": 0.0002,
+      "loss": 0.7214,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2383800824508933,
+      "grad_norm": 0.585242509841919,
+      "learning_rate": 0.0002,
+      "loss": 0.7699,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2399967666316385,
+      "grad_norm": 0.5211917757987976,
+      "learning_rate": 0.0002,
+      "loss": 0.7621,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2416134508123837,
+      "grad_norm": 0.6490384340286255,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2432301349931292,
+      "grad_norm": 0.6249763369560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2448468191738744,
+      "grad_norm": 0.71870356798172,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2464635033546196,
+      "grad_norm": 0.6761967539787292,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2480801875353649,
+      "grad_norm": 0.6500617265701294,
+      "learning_rate": 0.0002,
+      "loss": 0.7501,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2496968717161103,
+      "grad_norm": 0.8069869875907898,
+      "learning_rate": 0.0002,
+      "loss": 0.7903,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2513135558968556,
+      "grad_norm": 0.6044608950614929,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2529302400776008,
+      "grad_norm": 0.6573283076286316,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2545469242583462,
+      "grad_norm": 0.625430166721344,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2561636084390915,
+      "grad_norm": 0.5442022681236267,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7770
+    },
+    {
+      "epoch": 1.2577802926198367,
+      "grad_norm": 0.6818386912345886,
+      "learning_rate": 0.0002,
+      "loss": 0.7172,
+      "step": 7780
+    },
+    {
+      "epoch": 1.259396976800582,
+      "grad_norm": 0.6381874084472656,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2610136609813272,
+      "grad_norm": 0.6269212961196899,
+      "learning_rate": 0.0002,
+      "loss": 0.6834,
+      "step": 7800
+    },
+    {
+      "epoch": 1.2626303451620726,
+      "grad_norm": 0.600121259689331,
+      "learning_rate": 0.0002,
+      "loss": 0.7821,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2642470293428179,
+      "grad_norm": 0.6337703466415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2658637135235633,
+      "grad_norm": 0.7234963774681091,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 7830
+    },
+    {
+      "epoch": 1.2674803977043085,
+      "grad_norm": 0.800184965133667,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2690970818850538,
+      "grad_norm": 0.7539464831352234,
+      "learning_rate": 0.0002,
+      "loss": 0.7426,
+      "step": 7850
+    },
+    {
+      "epoch": 1.270713766065799,
+      "grad_norm": 0.5493760704994202,
+      "learning_rate": 0.0002,
+      "loss": 0.7496,
+      "step": 7860
+    },
+    {
+      "epoch": 1.2723304502465442,
+      "grad_norm": 0.7477145791053772,
+      "learning_rate": 0.0002,
+      "loss": 0.7537,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2739471344272897,
+      "grad_norm": 0.6366362571716309,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 7880
+    },
+    {
+      "epoch": 1.275563818608035,
+      "grad_norm": 0.7419533729553223,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2771805027887801,
+      "grad_norm": 0.6141223311424255,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 7900
+    },
+    {
+      "epoch": 1.2787971869695256,
+      "grad_norm": 0.7522598505020142,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 7910
+    },
+    {
+      "epoch": 1.2804138711502708,
+      "grad_norm": 0.6935804486274719,
+      "learning_rate": 0.0002,
+      "loss": 0.7097,
+      "step": 7920
+    },
+    {
+      "epoch": 1.282030555331016,
+      "grad_norm": 0.7239290475845337,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 7930
+    },
+    {
+      "epoch": 1.2836472395117613,
+      "grad_norm": 0.8800187110900879,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 7940
+    },
+    {
+      "epoch": 1.2852639236925067,
+      "grad_norm": 0.540458083152771,
+      "learning_rate": 0.0002,
+      "loss": 0.6991,
+      "step": 7950
+    },
+    {
+      "epoch": 1.286880607873252,
+      "grad_norm": 0.6492934226989746,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 7960
+    },
+    {
+      "epoch": 1.2884972920539972,
+      "grad_norm": 0.6543959379196167,
+      "learning_rate": 0.0002,
+      "loss": 0.7742,
+      "step": 7970
+    },
+    {
+      "epoch": 1.2901139762347427,
+      "grad_norm": 0.5804705619812012,
+      "learning_rate": 0.0002,
+      "loss": 0.7316,
+      "step": 7980
+    },
+    {
+      "epoch": 1.291730660415488,
+      "grad_norm": 0.7074727416038513,
+      "learning_rate": 0.0002,
+      "loss": 0.796,
+      "step": 7990
+    },
+    {
+      "epoch": 1.2933473445962331,
+      "grad_norm": 0.5347974300384521,
+      "learning_rate": 0.0002,
+      "loss": 0.7034,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2949640287769784,
+      "grad_norm": 0.6457298398017883,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 8010
+    },
+    {
+      "epoch": 1.2965807129577236,
+      "grad_norm": 0.6407219171524048,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8020
+    },
+    {
+      "epoch": 1.298197397138469,
+      "grad_norm": 0.828439474105835,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2998140813192143,
+      "grad_norm": 0.4840380549430847,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3014307654999595,
+      "grad_norm": 0.5921024680137634,
+      "learning_rate": 0.0002,
+      "loss": 0.7283,
+      "step": 8050
+    },
+    {
+      "epoch": 1.303047449680705,
+      "grad_norm": 0.6170315146446228,
+      "learning_rate": 0.0002,
+      "loss": 0.7477,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3046641338614502,
+      "grad_norm": 0.5374847054481506,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3062808180421954,
+      "grad_norm": 0.545758068561554,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3078975022229407,
+      "grad_norm": 0.55641770362854,
+      "learning_rate": 0.0002,
+      "loss": 0.7463,
+      "step": 8090
+    },
+    {
+      "epoch": 1.309514186403686,
+      "grad_norm": 0.6724897027015686,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3111308705844313,
+      "grad_norm": 0.6923972368240356,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3127475547651766,
+      "grad_norm": 0.5136841535568237,
+      "learning_rate": 0.0002,
+      "loss": 0.7149,
+      "step": 8120
+    },
+    {
+      "epoch": 1.314364238945922,
+      "grad_norm": 0.6766283512115479,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3159809231266673,
+      "grad_norm": 0.6283926367759705,
+      "learning_rate": 0.0002,
+      "loss": 0.7489,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3175976073074125,
+      "grad_norm": 0.644216001033783,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3192142914881577,
+      "grad_norm": 0.7827503085136414,
+      "learning_rate": 0.0002,
+      "loss": 0.7125,
+      "step": 8160
+    },
+    {
+      "epoch": 1.320830975668903,
+      "grad_norm": 0.6651390790939331,
+      "learning_rate": 0.0002,
+      "loss": 0.7271,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3224476598496484,
+      "grad_norm": 0.5547412633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7778,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3240643440303936,
+      "grad_norm": 0.6765179634094238,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 8190
+    },
+    {
+      "epoch": 1.325681028211139,
+      "grad_norm": 0.6822077035903931,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3272977123918843,
+      "grad_norm": 0.5941002368927002,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3289143965726296,
+      "grad_norm": 0.4850037097930908,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3305310807533748,
+      "grad_norm": 0.6162990927696228,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 8230
+    },
+    {
+      "epoch": 1.33214776493412,
+      "grad_norm": 0.6665613651275635,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3337644491148655,
+      "grad_norm": 0.618192732334137,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3353811332956107,
+      "grad_norm": 0.710418701171875,
+      "learning_rate": 0.0002,
+      "loss": 0.7532,
+      "step": 8260
+    },
+    {
+      "epoch": 1.336997817476356,
+      "grad_norm": 0.5109876990318298,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3386145016571014,
+      "grad_norm": 0.6791711449623108,
+      "learning_rate": 0.0002,
+      "loss": 0.7303,
+      "step": 8280
+    },
+    {
+      "epoch": 1.3402311858378466,
+      "grad_norm": 0.6836432814598083,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3418478700185918,
+      "grad_norm": 0.5579386353492737,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8300
+    },
+    {
+      "epoch": 1.343464554199337,
+      "grad_norm": 0.6713546514511108,
+      "learning_rate": 0.0002,
+      "loss": 0.7377,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3450812383800825,
+      "grad_norm": 0.5353720188140869,
+      "learning_rate": 0.0002,
+      "loss": 0.7756,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3466979225608278,
+      "grad_norm": 0.5813682675361633,
+      "learning_rate": 0.0002,
+      "loss": 0.718,
+      "step": 8330
+    },
+    {
+      "epoch": 1.348314606741573,
+      "grad_norm": 0.8158791661262512,
+      "learning_rate": 0.0002,
+      "loss": 0.7294,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3499312909223184,
+      "grad_norm": 0.6193785071372986,
+      "learning_rate": 0.0002,
+      "loss": 0.6992,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3515479751030637,
+      "grad_norm": 0.6353939771652222,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 8360
+    },
+    {
+      "epoch": 1.353164659283809,
+      "grad_norm": 0.6925048232078552,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 8370
+    },
+    {
+      "epoch": 1.3547813434645541,
+      "grad_norm": 0.988264799118042,
+      "learning_rate": 0.0002,
+      "loss": 0.736,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3563980276452994,
+      "grad_norm": 0.6476002931594849,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3580147118260448,
+      "grad_norm": 0.7120398879051208,
+      "learning_rate": 0.0002,
+      "loss": 0.776,
+      "step": 8400
+    },
+    {
+      "epoch": 1.35963139600679,
+      "grad_norm": 0.9048416614532471,
+      "learning_rate": 0.0002,
+      "loss": 0.7368,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3612480801875353,
+      "grad_norm": 0.7000672817230225,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3628647643682807,
+      "grad_norm": 0.6015632152557373,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 8430
+    },
+    {
+      "epoch": 1.364481448549026,
+      "grad_norm": 0.612516462802887,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3660981327297712,
+      "grad_norm": 0.5969301462173462,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 8450
+    },
+    {
+      "epoch": 1.3677148169105164,
+      "grad_norm": 0.6730654239654541,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 8460
+    },
+    {
+      "epoch": 1.369331501091262,
+      "grad_norm": 0.6386392116546631,
+      "learning_rate": 0.0002,
+      "loss": 0.7465,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3709481852720071,
+      "grad_norm": 0.739544153213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3725648694527524,
+      "grad_norm": 0.6462782621383667,
+      "learning_rate": 0.0002,
+      "loss": 0.7892,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3741815536334978,
+      "grad_norm": 0.7346843481063843,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 8500
+    },
+    {
+      "epoch": 1.375798237814243,
+      "grad_norm": 0.6884821057319641,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8510
+    },
+    {
+      "epoch": 1.3774149219949883,
+      "grad_norm": 0.6999333500862122,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3790316061757335,
+      "grad_norm": 0.5378713011741638,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 8530
+    },
+    {
+      "epoch": 1.3806482903564787,
+      "grad_norm": 0.5417906641960144,
+      "learning_rate": 0.0002,
+      "loss": 0.6797,
+      "step": 8540
+    },
+    {
+      "epoch": 1.3822649745372242,
+      "grad_norm": 0.6602526307106018,
+      "learning_rate": 0.0002,
+      "loss": 0.7499,
+      "step": 8550
+    },
+    {
+      "epoch": 1.3838816587179694,
+      "grad_norm": 0.7073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.7356,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3854983428987149,
+      "grad_norm": 0.5841707587242126,
+      "learning_rate": 0.0002,
+      "loss": 0.75,
+      "step": 8570
+    },
+    {
+      "epoch": 1.38711502707946,
+      "grad_norm": 0.7031095027923584,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 8580
+    },
+    {
+      "epoch": 1.3887317112602053,
+      "grad_norm": 0.5198570489883423,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 8590
+    },
+    {
+      "epoch": 1.3903483954409506,
+      "grad_norm": 0.7261320352554321,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 8600
+    },
+    {
+      "epoch": 1.3919650796216958,
+      "grad_norm": 0.5616350173950195,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3935817638024413,
+      "grad_norm": 0.5185914635658264,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 8620
+    },
+    {
+      "epoch": 1.3951984479831865,
+      "grad_norm": 0.5814694762229919,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 8630
+    },
+    {
+      "epoch": 1.3968151321639317,
+      "grad_norm": 0.6977371573448181,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 8640
+    },
+    {
+      "epoch": 1.3984318163446772,
+      "grad_norm": 0.6855689883232117,
+      "learning_rate": 0.0002,
+      "loss": 0.7574,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4000485005254224,
+      "grad_norm": 0.5414357781410217,
+      "learning_rate": 0.0002,
+      "loss": 0.7802,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4016651847061676,
+      "grad_norm": 0.6970012784004211,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4032818688869129,
+      "grad_norm": 0.526079535484314,
+      "learning_rate": 0.0002,
+      "loss": 0.7421,
+      "step": 8680
+    },
+    {
+      "epoch": 1.404898553067658,
+      "grad_norm": 0.758712887763977,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4065152372484035,
+      "grad_norm": 0.7118762731552124,
+      "learning_rate": 0.0002,
+      "loss": 0.7612,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4081319214291488,
+      "grad_norm": 0.5696909427642822,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4097486056098942,
+      "grad_norm": 0.7995436787605286,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4113652897906395,
+      "grad_norm": 0.7237521409988403,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4129819739713847,
+      "grad_norm": 0.744628369808197,
+      "learning_rate": 0.0002,
+      "loss": 0.7661,
+      "step": 8740
+    },
+    {
+      "epoch": 1.41459865815213,
+      "grad_norm": 0.6082926988601685,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4162153423328752,
+      "grad_norm": 0.5185243487358093,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4178320265136206,
+      "grad_norm": 0.5183082222938538,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4194487106943658,
+      "grad_norm": 0.7326041460037231,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 8780
+    },
+    {
+      "epoch": 1.421065394875111,
+      "grad_norm": 0.7174660563468933,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4226820790558565,
+      "grad_norm": 0.8080165982246399,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4242987632366018,
+      "grad_norm": 0.5061507821083069,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 8810
+    },
+    {
+      "epoch": 1.425915447417347,
+      "grad_norm": 0.801602840423584,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4275321315980922,
+      "grad_norm": 0.6150273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4291488157788377,
+      "grad_norm": 0.8786525726318359,
+      "learning_rate": 0.0002,
+      "loss": 0.7452,
+      "step": 8840
+    },
+    {
+      "epoch": 1.430765499959583,
+      "grad_norm": 0.6371538639068604,
+      "learning_rate": 0.0002,
+      "loss": 0.7257,
+      "step": 8850
+    },
+    {
+      "epoch": 1.4323821841403281,
+      "grad_norm": 0.6409295797348022,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4339988683210736,
+      "grad_norm": 0.6452359557151794,
+      "learning_rate": 0.0002,
+      "loss": 0.7891,
+      "step": 8870
+    },
+    {
+      "epoch": 1.4356155525018188,
+      "grad_norm": 0.5842334628105164,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 8880
+    },
+    {
+      "epoch": 1.437232236682564,
+      "grad_norm": 0.696761965751648,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4388489208633093,
+      "grad_norm": 0.6384600400924683,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4404656050440545,
+      "grad_norm": 0.5981136560440063,
+      "learning_rate": 0.0002,
+      "loss": 0.7049,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4420822892248,
+      "grad_norm": 0.6355637907981873,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4436989734055452,
+      "grad_norm": 0.6374830603599548,
+      "learning_rate": 0.0002,
+      "loss": 0.7653,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4453156575862904,
+      "grad_norm": 0.559013307094574,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 8940
+    },
+    {
+      "epoch": 1.446932341767036,
+      "grad_norm": 0.7289170026779175,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4485490259477811,
+      "grad_norm": 0.8649206757545471,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4501657101285264,
+      "grad_norm": 0.7664689421653748,
+      "learning_rate": 0.0002,
+      "loss": 0.7745,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4517823943092716,
+      "grad_norm": 0.7109952569007874,
+      "learning_rate": 0.0002,
+      "loss": 0.7431,
+      "step": 8980
+    },
+    {
+      "epoch": 1.453399078490017,
+      "grad_norm": 0.6312844753265381,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4550157626707623,
+      "grad_norm": 0.6616617441177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7467,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4566324468515075,
+      "grad_norm": 0.7384068965911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7518,
+      "step": 9010
+    },
+    {
+      "epoch": 1.458249131032253,
+      "grad_norm": 0.6549670100212097,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9020
+    },
+    {
+      "epoch": 1.4598658152129982,
+      "grad_norm": 0.6254119277000427,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4614824993937434,
+      "grad_norm": 0.6806328892707825,
+      "learning_rate": 0.0002,
+      "loss": 0.7645,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4630991835744886,
+      "grad_norm": 0.6803115010261536,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4647158677552339,
+      "grad_norm": 0.48529282212257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7264,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4663325519359793,
+      "grad_norm": 0.5995030999183655,
+      "learning_rate": 0.0002,
+      "loss": 0.7542,
+      "step": 9070
+    },
+    {
+      "epoch": 1.4679492361167246,
+      "grad_norm": 0.6005427837371826,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9080
+    },
+    {
+      "epoch": 1.46956592029747,
+      "grad_norm": 0.718564510345459,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4711826044782153,
+      "grad_norm": 0.7003577351570129,
+      "learning_rate": 0.0002,
+      "loss": 0.7089,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4727992886589605,
+      "grad_norm": 0.5888323783874512,
+      "learning_rate": 0.0002,
+      "loss": 0.8069,
+      "step": 9110
+    },
+    {
+      "epoch": 1.4744159728397057,
+      "grad_norm": 0.6417609453201294,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 9120
+    },
+    {
+      "epoch": 1.476032657020451,
+      "grad_norm": 0.572294294834137,
+      "learning_rate": 0.0002,
+      "loss": 0.7441,
+      "step": 9130
+    },
+    {
+      "epoch": 1.4776493412011964,
+      "grad_norm": 0.8200714588165283,
+      "learning_rate": 0.0002,
+      "loss": 0.8053,
+      "step": 9140
+    },
+    {
+      "epoch": 1.4792660253819416,
+      "grad_norm": 0.6343288421630859,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4808827095626869,
+      "grad_norm": 0.7017961144447327,
+      "learning_rate": 0.0002,
+      "loss": 0.7641,
+      "step": 9160
+    },
+    {
+      "epoch": 1.4824993937434323,
+      "grad_norm": 0.6202912926673889,
+      "learning_rate": 0.0002,
+      "loss": 0.7619,
+      "step": 9170
+    },
+    {
+      "epoch": 1.4841160779241775,
+      "grad_norm": 0.6677869558334351,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 9180
+    },
+    {
+      "epoch": 1.4857327621049228,
+      "grad_norm": 0.6052267551422119,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 9190
+    },
+    {
+      "epoch": 1.487349446285668,
+      "grad_norm": 0.6638872027397156,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 9200
+    },
+    {
+      "epoch": 1.4889661304664135,
+      "grad_norm": 0.6245523691177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 9210
+    },
+    {
+      "epoch": 1.4905828146471587,
+      "grad_norm": 0.5761767625808716,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 9220
+    },
+    {
+      "epoch": 1.492199498827904,
+      "grad_norm": 0.8175981640815735,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 9230
+    },
+    {
+      "epoch": 1.4938161830086494,
+      "grad_norm": 0.9144009947776794,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 9240
+    },
+    {
+      "epoch": 1.4954328671893946,
+      "grad_norm": 0.5742552876472473,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 9250
+    },
+    {
+      "epoch": 1.4970495513701398,
+      "grad_norm": 0.534534215927124,
+      "learning_rate": 0.0002,
+      "loss": 0.7418,
+      "step": 9260
+    },
+    {
+      "epoch": 1.498666235550885,
+      "grad_norm": 0.7836225032806396,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5002829197316303,
+      "grad_norm": 0.5292993187904358,
+      "learning_rate": 0.0002,
+      "loss": 0.7453,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5018996039123758,
+      "grad_norm": 0.8044071793556213,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9290
+    },
+    {
+      "epoch": 1.503516288093121,
+      "grad_norm": 0.6185805201530457,
+      "learning_rate": 0.0002,
+      "loss": 0.7229,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5051329722738664,
+      "grad_norm": 0.6093607544898987,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5067496564546117,
+      "grad_norm": 0.5891730189323425,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 9320
+    },
+    {
+      "epoch": 1.508366340635357,
+      "grad_norm": 0.6331129670143127,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5099830248161021,
+      "grad_norm": 0.7690958380699158,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5115997089968474,
+      "grad_norm": 0.6548877358436584,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5132163931775926,
+      "grad_norm": 0.6545143127441406,
+      "learning_rate": 0.0002,
+      "loss": 0.7408,
+      "step": 9360
+    },
+    {
+      "epoch": 1.514833077358338,
+      "grad_norm": 0.553247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.7432,
+      "step": 9370
+    },
+    {
+      "epoch": 1.5164497615390833,
+      "grad_norm": 0.8145074844360352,
+      "learning_rate": 0.0002,
+      "loss": 0.7265,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5180664457198287,
+      "grad_norm": 0.7636994123458862,
+      "learning_rate": 0.0002,
+      "loss": 0.7379,
+      "step": 9390
+    },
+    {
+      "epoch": 1.519683129900574,
+      "grad_norm": 0.6838982701301575,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5212998140813192,
+      "grad_norm": 0.8599441647529602,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5229164982620644,
+      "grad_norm": 0.7020329833030701,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5245331824428097,
+      "grad_norm": 0.6964772343635559,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5261498666235551,
+      "grad_norm": 0.6916600465774536,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9440
+    },
+    {
+      "epoch": 1.5277665508043003,
+      "grad_norm": 0.7282621264457703,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5293832349850458,
+      "grad_norm": 0.5363983511924744,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 9460
+    },
+    {
+      "epoch": 1.530999919165791,
+      "grad_norm": 0.6184861063957214,
+      "learning_rate": 0.0002,
+      "loss": 0.7154,
+      "step": 9470
+    },
+    {
+      "epoch": 1.5326166033465363,
+      "grad_norm": 0.5991285443305969,
+      "learning_rate": 0.0002,
+      "loss": 0.7837,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5342332875272815,
+      "grad_norm": 0.8176587820053101,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5358499717080267,
+      "grad_norm": 0.6473721861839294,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 9500
+    },
+    {
+      "epoch": 1.5374666558887722,
+      "grad_norm": 0.7319952845573425,
+      "learning_rate": 0.0002,
+      "loss": 0.7632,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5390833400695174,
+      "grad_norm": 0.702900230884552,
+      "learning_rate": 0.0002,
+      "loss": 0.7706,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5407000242502629,
+      "grad_norm": 0.7971600294113159,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 9530
+    },
+    {
+      "epoch": 1.542316708431008,
+      "grad_norm": 0.6527525186538696,
+      "learning_rate": 0.0002,
+      "loss": 0.7352,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5439333926117533,
+      "grad_norm": 0.5791676044464111,
+      "learning_rate": 0.0002,
+      "loss": 0.7425,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5455500767924986,
+      "grad_norm": 0.5619390606880188,
+      "learning_rate": 0.0002,
+      "loss": 0.7585,
+      "step": 9560
+    },
+    {
+      "epoch": 1.5471667609732438,
+      "grad_norm": 0.5701689124107361,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9570
+    },
+    {
+      "epoch": 1.548783445153989,
+      "grad_norm": 0.47549352049827576,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5504001293347345,
+      "grad_norm": 0.8730611205101013,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5520168135154797,
+      "grad_norm": 0.6842091083526611,
+      "learning_rate": 0.0002,
+      "loss": 0.798,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5536334976962252,
+      "grad_norm": 0.6675129532814026,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5552501818769704,
+      "grad_norm": 0.8173956274986267,
+      "learning_rate": 0.0002,
+      "loss": 0.7954,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5568668660577156,
+      "grad_norm": 0.724947452545166,
+      "learning_rate": 0.0002,
+      "loss": 0.7535,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5584835502384609,
+      "grad_norm": 0.6154758930206299,
+      "learning_rate": 0.0002,
+      "loss": 0.7738,
+      "step": 9640
+    },
+    {
+      "epoch": 1.560100234419206,
+      "grad_norm": 0.6072008013725281,
+      "learning_rate": 0.0002,
+      "loss": 0.7568,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5617169185999515,
+      "grad_norm": 0.659010648727417,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 9660
+    },
+    {
+      "epoch": 1.5633336027806968,
+      "grad_norm": 0.65857994556427,
+      "learning_rate": 0.0002,
+      "loss": 0.673,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5649502869614422,
+      "grad_norm": 0.5914267301559448,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5665669711421875,
+      "grad_norm": 0.6248020529747009,
+      "learning_rate": 0.0002,
+      "loss": 0.7414,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5681836553229327,
+      "grad_norm": 0.7147795557975769,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 9700
+    },
+    {
+      "epoch": 1.569800339503678,
+      "grad_norm": 0.7076232433319092,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 9710
+    },
+    {
+      "epoch": 1.5714170236844232,
+      "grad_norm": 0.6217400431632996,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5730337078651684,
+      "grad_norm": 0.6709911227226257,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 9730
+    },
+    {
+      "epoch": 1.5746503920459138,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 9740
+    },
+    {
+      "epoch": 1.576267076226659,
+      "grad_norm": 0.6241145730018616,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 9750
+    },
+    {
+      "epoch": 1.5778837604074045,
+      "grad_norm": 0.4960934817790985,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 9760
+    },
+    {
+      "epoch": 1.5795004445881498,
+      "grad_norm": 0.6593309640884399,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 9770
+    },
+    {
+      "epoch": 1.581117128768895,
+      "grad_norm": 0.5814042091369629,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 9780
+    },
+    {
+      "epoch": 1.5827338129496402,
+      "grad_norm": 0.5936070680618286,
+      "learning_rate": 0.0002,
+      "loss": 0.7109,
+      "step": 9790
+    },
+    {
+      "epoch": 1.5843504971303854,
+      "grad_norm": 0.6454403400421143,
+      "learning_rate": 0.0002,
+      "loss": 0.7769,
+      "step": 9800
+    },
+    {
+      "epoch": 1.585967181311131,
+      "grad_norm": 0.7612107992172241,
+      "learning_rate": 0.0002,
+      "loss": 0.7677,
+      "step": 9810
+    },
+    {
+      "epoch": 1.5875838654918761,
+      "grad_norm": 0.6494482755661011,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5892005496726216,
+      "grad_norm": 0.7825694680213928,
+      "learning_rate": 0.0002,
+      "loss": 0.7569,
+      "step": 9830
+    },
+    {
+      "epoch": 1.5908172338533668,
+      "grad_norm": 0.6757757663726807,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 9840
+    },
+    {
+      "epoch": 1.592433918034112,
+      "grad_norm": 0.7105609178543091,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 9850
+    },
+    {
+      "epoch": 1.5940506022148573,
+      "grad_norm": 0.7596991062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 9860
+    },
+    {
+      "epoch": 1.5956672863956025,
+      "grad_norm": 0.5681525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7108,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5972839705763477,
+      "grad_norm": 0.6090980768203735,
+      "learning_rate": 0.0002,
+      "loss": 0.7811,
+      "step": 9880
+    },
+    {
+      "epoch": 1.5989006547570932,
+      "grad_norm": 0.6271613240242004,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6005173389378387,
+      "grad_norm": 0.7656369805335999,
+      "learning_rate": 0.0002,
+      "loss": 0.7419,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6021340231185839,
+      "grad_norm": 0.7504446506500244,
+      "learning_rate": 0.0002,
+      "loss": 0.7336,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6037507072993291,
+      "grad_norm": 0.659656286239624,
+      "learning_rate": 0.0002,
+      "loss": 0.7479,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6053673914800743,
+      "grad_norm": 0.6006826162338257,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6069840756608196,
+      "grad_norm": 0.7872757911682129,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6086007598415648,
+      "grad_norm": 0.5545852780342102,
+      "learning_rate": 0.0002,
+      "loss": 0.768,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6102174440223103,
+      "grad_norm": 0.7429468631744385,
+      "learning_rate": 0.0002,
+      "loss": 0.8064,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6118341282030555,
+      "grad_norm": 0.6873556971549988,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 9970
+    },
+    {
+      "epoch": 1.613450812383801,
+      "grad_norm": 0.5874287486076355,
+      "learning_rate": 0.0002,
+      "loss": 0.7324,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6150674965645462,
+      "grad_norm": 0.6039386987686157,
+      "learning_rate": 0.0002,
+      "loss": 0.7141,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6166841807452914,
+      "grad_norm": 0.6233575940132141,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6183008649260366,
+      "grad_norm": 0.7676448225975037,
+      "learning_rate": 0.0002,
+      "loss": 0.7602,
+      "step": 10010
+    },
+    {
+      "epoch": 1.6199175491067819,
+      "grad_norm": 0.6565698385238647,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6215342332875273,
+      "grad_norm": 0.6787590384483337,
+      "learning_rate": 0.0002,
+      "loss": 0.7104,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6231509174682726,
+      "grad_norm": 0.6137678027153015,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 10040
+    },
+    {
+      "epoch": 1.624767601649018,
+      "grad_norm": 0.5236800312995911,
+      "learning_rate": 0.0002,
+      "loss": 0.7646,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6263842858297632,
+      "grad_norm": 0.7626367807388306,
+      "learning_rate": 0.0002,
+      "loss": 0.7437,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6280009700105085,
+      "grad_norm": 0.5657260417938232,
+      "learning_rate": 0.0002,
+      "loss": 0.7273,
+      "step": 10070
+    },
+    {
+      "epoch": 1.6296176541912537,
+      "grad_norm": 0.4913991391658783,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 10080
+    },
+    {
+      "epoch": 1.631234338371999,
+      "grad_norm": 0.7715556621551514,
+      "learning_rate": 0.0002,
+      "loss": 0.7596,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6328510225527442,
+      "grad_norm": 0.6509000062942505,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6344677067334896,
+      "grad_norm": 0.6215850114822388,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6360843909142349,
+      "grad_norm": 0.6956844329833984,
+      "learning_rate": 0.0002,
+      "loss": 0.7705,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6377010750949803,
+      "grad_norm": 0.6111597418785095,
+      "learning_rate": 0.0002,
+      "loss": 0.7129,
+      "step": 10130
+    },
+    {
+      "epoch": 1.6393177592757255,
+      "grad_norm": 0.6518288850784302,
+      "learning_rate": 0.0002,
+      "loss": 0.6955,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6409344434564708,
+      "grad_norm": 0.6914522051811218,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10150
+    },
+    {
+      "epoch": 1.642551127637216,
+      "grad_norm": 0.63785719871521,
+      "learning_rate": 0.0002,
+      "loss": 0.7295,
+      "step": 10160
+    },
+    {
+      "epoch": 1.6441678118179612,
+      "grad_norm": 0.6379287838935852,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6457844959987067,
+      "grad_norm": 0.6793403029441833,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 10180
+    },
+    {
+      "epoch": 1.647401180179452,
+      "grad_norm": 0.6099132895469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6490178643601974,
+      "grad_norm": 0.5869854092597961,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6506345485409426,
+      "grad_norm": 0.7716999053955078,
+      "learning_rate": 0.0002,
+      "loss": 0.8308,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6522512327216878,
+      "grad_norm": 0.6854110360145569,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 10220
+    },
+    {
+      "epoch": 1.653867916902433,
+      "grad_norm": 0.6957170367240906,
+      "learning_rate": 0.0002,
+      "loss": 0.782,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6554846010831783,
+      "grad_norm": 0.6932903528213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6571012852639235,
+      "grad_norm": 0.7713165283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10250
+    },
+    {
+      "epoch": 1.658717969444669,
+      "grad_norm": 0.7455793619155884,
+      "learning_rate": 0.0002,
+      "loss": 0.7099,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6603346536254144,
+      "grad_norm": 0.5464168190956116,
+      "learning_rate": 0.0002,
+      "loss": 0.7524,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6619513378061597,
+      "grad_norm": 0.6782926321029663,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 10280
+    },
+    {
+      "epoch": 1.663568021986905,
+      "grad_norm": 0.7962649464607239,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6651847061676501,
+      "grad_norm": 0.6814526319503784,
+      "learning_rate": 0.0002,
+      "loss": 0.7142,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6668013903483954,
+      "grad_norm": 0.656895101070404,
+      "learning_rate": 0.0002,
+      "loss": 0.7285,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6684180745291406,
+      "grad_norm": 0.6085672378540039,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 10320
+    },
+    {
+      "epoch": 1.670034758709886,
+      "grad_norm": 0.585508406162262,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 10330
+    },
+    {
+      "epoch": 1.6716514428906313,
+      "grad_norm": 0.6930184364318848,
+      "learning_rate": 0.0002,
+      "loss": 0.7604,
+      "step": 10340
+    },
+    {
+      "epoch": 1.6732681270713767,
+      "grad_norm": 0.575663149356842,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10350
+    },
+    {
+      "epoch": 1.674884811252122,
+      "grad_norm": 0.582502543926239,
+      "learning_rate": 0.0002,
+      "loss": 0.7198,
+      "step": 10360
+    },
+    {
+      "epoch": 1.6765014954328672,
+      "grad_norm": 0.5668916702270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 10370
+    },
+    {
+      "epoch": 1.6781181796136124,
+      "grad_norm": 0.6070065498352051,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10380
+    },
+    {
+      "epoch": 1.6797348637943577,
+      "grad_norm": 0.6141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.7939,
+      "step": 10390
+    },
+    {
+      "epoch": 1.6813515479751031,
+      "grad_norm": 0.8359124064445496,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 10400
+    },
+    {
+      "epoch": 1.6829682321558483,
+      "grad_norm": 0.5378185510635376,
+      "learning_rate": 0.0002,
+      "loss": 0.7488,
+      "step": 10410
+    },
+    {
+      "epoch": 1.6845849163365938,
+      "grad_norm": 0.6959536075592041,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 10420
+    },
+    {
+      "epoch": 1.686201600517339,
+      "grad_norm": 0.6514357328414917,
+      "learning_rate": 0.0002,
+      "loss": 0.7872,
+      "step": 10430
+    },
+    {
+      "epoch": 1.6878182846980843,
+      "grad_norm": 0.7706646919250488,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10440
+    },
+    {
+      "epoch": 1.6894349688788295,
+      "grad_norm": 0.6183337569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.7673,
+      "step": 10450
+    },
+    {
+      "epoch": 1.6910516530595747,
+      "grad_norm": 0.6123278141021729,
+      "learning_rate": 0.0002,
+      "loss": 0.7566,
+      "step": 10460
+    },
+    {
+      "epoch": 1.69266833724032,
+      "grad_norm": 0.6894851326942444,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10470
+    },
+    {
+      "epoch": 1.6942850214210654,
+      "grad_norm": 0.7497312426567078,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 10480
+    },
+    {
+      "epoch": 1.6959017056018106,
+      "grad_norm": 0.5968214273452759,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10490
+    },
+    {
+      "epoch": 1.697518389782556,
+      "grad_norm": 0.6747927069664001,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 10500
+    },
+    {
+      "epoch": 1.6991350739633013,
+      "grad_norm": 0.5708310008049011,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7007517581440466,
+      "grad_norm": 0.606526792049408,
+      "learning_rate": 0.0002,
+      "loss": 0.7385,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7023684423247918,
+      "grad_norm": 0.662011981010437,
+      "learning_rate": 0.0002,
+      "loss": 0.7204,
+      "step": 10530
+    },
+    {
+      "epoch": 1.703985126505537,
+      "grad_norm": 0.7583045363426208,
+      "learning_rate": 0.0002,
+      "loss": 0.7999,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7056018106862825,
+      "grad_norm": 0.721632182598114,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7072184948670277,
+      "grad_norm": 0.6107715368270874,
+      "learning_rate": 0.0002,
+      "loss": 0.7407,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7088351790477732,
+      "grad_norm": 0.6652471423149109,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7104518632285184,
+      "grad_norm": 0.6308087110519409,
+      "learning_rate": 0.0002,
+      "loss": 0.7767,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7120685474092636,
+      "grad_norm": 0.5464386940002441,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7136852315900089,
+      "grad_norm": 0.6558911204338074,
+      "learning_rate": 0.0002,
+      "loss": 0.7063,
+      "step": 10600
+    },
+    {
+      "epoch": 1.715301915770754,
+      "grad_norm": 0.5665024518966675,
+      "learning_rate": 0.0002,
+      "loss": 0.7126,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7169185999514993,
+      "grad_norm": 0.7888094186782837,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7185352841322448,
+      "grad_norm": 0.7084909081459045,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 10630
+    },
+    {
+      "epoch": 1.7201519683129902,
+      "grad_norm": 0.7982324361801147,
+      "learning_rate": 0.0002,
+      "loss": 0.7557,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7217686524937355,
+      "grad_norm": 0.6418732404708862,
+      "learning_rate": 0.0002,
+      "loss": 0.7345,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7233853366744807,
+      "grad_norm": 0.7636681795120239,
+      "learning_rate": 0.0002,
+      "loss": 0.7734,
+      "step": 10660
+    },
+    {
+      "epoch": 1.725002020855226,
+      "grad_norm": 0.5646875500679016,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7266187050359711,
+      "grad_norm": 0.5231260657310486,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7282353892167164,
+      "grad_norm": 0.7635011672973633,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7298520733974618,
+      "grad_norm": 0.7518259286880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 10700
+    },
+    {
+      "epoch": 1.731468757578207,
+      "grad_norm": 0.7295602560043335,
+      "learning_rate": 0.0002,
+      "loss": 0.751,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7330854417589525,
+      "grad_norm": 0.6984632015228271,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7347021259396977,
+      "grad_norm": 0.6198219060897827,
+      "learning_rate": 0.0002,
+      "loss": 0.7921,
+      "step": 10730
+    },
+    {
+      "epoch": 1.736318810120443,
+      "grad_norm": 0.6957576274871826,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7379354943011882,
+      "grad_norm": 0.6430263519287109,
+      "learning_rate": 0.0002,
+      "loss": 0.7917,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7395521784819334,
+      "grad_norm": 0.6134995222091675,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 10760
+    },
+    {
+      "epoch": 1.741168862662679,
+      "grad_norm": 0.7209452986717224,
+      "learning_rate": 0.0002,
+      "loss": 0.7584,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7427855468434241,
+      "grad_norm": 0.6735447645187378,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7444022310241696,
+      "grad_norm": 0.5605693459510803,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 10790
+    },
+    {
+      "epoch": 1.7460189152049148,
+      "grad_norm": 0.6882363557815552,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 10800
+    },
+    {
+      "epoch": 1.74763559938566,
+      "grad_norm": 0.6386259198188782,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7492522835664053,
+      "grad_norm": 0.6529015302658081,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7508689677471505,
+      "grad_norm": 0.5664082765579224,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7524856519278957,
+      "grad_norm": 0.7532684206962585,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7541023361086412,
+      "grad_norm": 0.77171391248703,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7557190202893864,
+      "grad_norm": 0.7255431413650513,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7573357044701319,
+      "grad_norm": 0.763083279132843,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10870
+    },
+    {
+      "epoch": 1.758952388650877,
+      "grad_norm": 0.6042402982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7605690728316223,
+      "grad_norm": 0.7642518281936646,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7621857570123676,
+      "grad_norm": 0.6347904801368713,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7638024411931128,
+      "grad_norm": 0.5371627807617188,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7654191253738583,
+      "grad_norm": 0.6840225458145142,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7670358095546035,
+      "grad_norm": 0.5288469195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10930
+    },
+    {
+      "epoch": 1.768652493735349,
+      "grad_norm": 0.69020676612854,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 10940
+    },
+    {
+      "epoch": 1.7702691779160942,
+      "grad_norm": 0.5943242311477661,
+      "learning_rate": 0.0002,
+      "loss": 0.7468,
+      "step": 10950
+    },
+    {
+      "epoch": 1.7718858620968394,
+      "grad_norm": 0.5616418123245239,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 10960
+    },
+    {
+      "epoch": 1.7735025462775846,
+      "grad_norm": 0.7209470868110657,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 10970
+    },
+    {
+      "epoch": 1.7751192304583299,
+      "grad_norm": 0.6657957434654236,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 10980
+    },
+    {
+      "epoch": 1.776735914639075,
+      "grad_norm": 0.6469064950942993,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 10990
+    },
+    {
+      "epoch": 1.7783525988198206,
+      "grad_norm": 0.6615678071975708,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 11000
+    },
+    {
+      "epoch": 1.779969283000566,
+      "grad_norm": 0.6722439527511597,
+      "learning_rate": 0.0002,
+      "loss": 0.747,
+      "step": 11010
+    },
+    {
+      "epoch": 1.7815859671813112,
+      "grad_norm": 0.634136974811554,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 11020
+    },
+    {
+      "epoch": 1.7832026513620565,
+      "grad_norm": 0.6024377346038818,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 11030
+    },
+    {
+      "epoch": 1.7848193355428017,
+      "grad_norm": 0.6909403800964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 11040
+    },
+    {
+      "epoch": 1.786436019723547,
+      "grad_norm": 0.7148767709732056,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 11050
+    },
+    {
+      "epoch": 1.7880527039042922,
+      "grad_norm": 0.7442979216575623,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 11060
+    },
+    {
+      "epoch": 1.7896693880850376,
+      "grad_norm": 0.6830431818962097,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 11070
+    },
+    {
+      "epoch": 1.7912860722657828,
+      "grad_norm": 0.9172667264938354,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 11080
+    },
+    {
+      "epoch": 1.7929027564465283,
+      "grad_norm": 0.6799490451812744,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 11090
+    },
+    {
+      "epoch": 1.7945194406272735,
+      "grad_norm": 0.7617024779319763,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 11100
+    },
+    {
+      "epoch": 1.7961361248080188,
+      "grad_norm": 0.7701810002326965,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 11110
+    },
+    {
+      "epoch": 1.797752808988764,
+      "grad_norm": 0.7454385757446289,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11120
+    },
+    {
+      "epoch": 1.7993694931695092,
+      "grad_norm": 0.6121436953544617,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8009861773502547,
+      "grad_norm": 0.6237571835517883,
+      "learning_rate": 0.0002,
+      "loss": 0.7305,
+      "step": 11140
+    },
+    {
+      "epoch": 1.802602861531,
+      "grad_norm": 0.6818515658378601,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 11150
+    },
+    {
+      "epoch": 1.8042195457117454,
+      "grad_norm": 0.7768308520317078,
+      "learning_rate": 0.0002,
+      "loss": 0.6876,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8058362298924906,
+      "grad_norm": 0.6875537633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 11170
+    },
+    {
+      "epoch": 1.8074529140732358,
+      "grad_norm": 0.7950584888458252,
+      "learning_rate": 0.0002,
+      "loss": 0.761,
+      "step": 11180
+    },
+    {
+      "epoch": 1.809069598253981,
+      "grad_norm": 0.8210248351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8106862824347263,
+      "grad_norm": 0.6674110889434814,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8123029666154715,
+      "grad_norm": 0.6261674761772156,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 11210
+    },
+    {
+      "epoch": 1.813919650796217,
+      "grad_norm": 0.6484741568565369,
+      "learning_rate": 0.0002,
+      "loss": 0.7122,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8155363349769622,
+      "grad_norm": 0.6231244206428528,
+      "learning_rate": 0.0002,
+      "loss": 0.7718,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8171530191577077,
+      "grad_norm": 0.7243146896362305,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11240
+    },
+    {
+      "epoch": 1.818769703338453,
+      "grad_norm": 0.6776193380355835,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8203863875191981,
+      "grad_norm": 0.5973618030548096,
+      "learning_rate": 0.0002,
+      "loss": 0.7317,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8220030716999434,
+      "grad_norm": 0.6451361179351807,
+      "learning_rate": 0.0002,
+      "loss": 0.7961,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8236197558806886,
+      "grad_norm": 0.5963068008422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 11280
+    },
+    {
+      "epoch": 1.825236440061434,
+      "grad_norm": 0.536902129650116,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8268531242421793,
+      "grad_norm": 0.6993787288665771,
+      "learning_rate": 0.0002,
+      "loss": 0.708,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8284698084229247,
+      "grad_norm": 0.6135255098342896,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11310
+    },
+    {
+      "epoch": 1.83008649260367,
+      "grad_norm": 0.6057423949241638,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8317031767844152,
+      "grad_norm": 0.6598812341690063,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8333198609651604,
+      "grad_norm": 0.6075948476791382,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 11340
+    },
+    {
+      "epoch": 1.8349365451459057,
+      "grad_norm": 0.7065447568893433,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8365532293266509,
+      "grad_norm": 0.680526614189148,
+      "learning_rate": 0.0002,
+      "loss": 0.7365,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8381699135073963,
+      "grad_norm": 0.6356695294380188,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8397865976881416,
+      "grad_norm": 0.6399052143096924,
+      "learning_rate": 0.0002,
+      "loss": 0.721,
+      "step": 11380
+    },
+    {
+      "epoch": 1.841403281868887,
+      "grad_norm": 0.6125704050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 11390
+    },
+    {
+      "epoch": 1.8430199660496323,
+      "grad_norm": 0.7124643325805664,
+      "learning_rate": 0.0002,
+      "loss": 0.755,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8446366502303775,
+      "grad_norm": 0.6099604964256287,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8462533344111227,
+      "grad_norm": 0.7338208556175232,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 11420
+    },
+    {
+      "epoch": 1.847870018591868,
+      "grad_norm": 0.7534668445587158,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8494867027726134,
+      "grad_norm": 0.6135470271110535,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8511033869533586,
+      "grad_norm": 0.6229309439659119,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 11450
+    },
+    {
+      "epoch": 1.852720071134104,
+      "grad_norm": 0.706423282623291,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 11460
+    },
+    {
+      "epoch": 1.8543367553148493,
+      "grad_norm": 0.5460049510002136,
+      "learning_rate": 0.0002,
+      "loss": 0.7411,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8559534394955945,
+      "grad_norm": 0.6616711020469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7416,
+      "step": 11480
+    },
+    {
+      "epoch": 1.8575701236763398,
+      "grad_norm": 0.6372783184051514,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 11490
+    },
+    {
+      "epoch": 1.859186807857085,
+      "grad_norm": 0.7162668108940125,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8608034920378305,
+      "grad_norm": 0.6605209708213806,
+      "learning_rate": 0.0002,
+      "loss": 0.7747,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8624201762185757,
+      "grad_norm": 0.6933956742286682,
+      "learning_rate": 0.0002,
+      "loss": 0.7258,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8640368603993211,
+      "grad_norm": 0.6582090854644775,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8656535445800664,
+      "grad_norm": 0.6416500806808472,
+      "learning_rate": 0.0002,
+      "loss": 0.7313,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8672702287608116,
+      "grad_norm": 0.5434312224388123,
+      "learning_rate": 0.0002,
+      "loss": 0.7372,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8688869129415568,
+      "grad_norm": 0.6827567219734192,
+      "learning_rate": 0.0002,
+      "loss": 0.7635,
+      "step": 11560
+    },
+    {
+      "epoch": 1.870503597122302,
+      "grad_norm": 0.7354370951652527,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 11570
+    },
+    {
+      "epoch": 1.8721202813030473,
+      "grad_norm": 0.590372622013092,
+      "learning_rate": 0.0002,
+      "loss": 0.7526,
+      "step": 11580
+    },
+    {
+      "epoch": 1.8737369654837928,
+      "grad_norm": 0.853183925151825,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 11590
+    },
+    {
+      "epoch": 1.875353649664538,
+      "grad_norm": 0.822678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 11600
+    },
+    {
+      "epoch": 1.8769703338452834,
+      "grad_norm": 0.6591550707817078,
+      "learning_rate": 0.0002,
+      "loss": 0.7427,
+      "step": 11610
+    },
+    {
+      "epoch": 1.8785870180260287,
+      "grad_norm": 0.7475301623344421,
+      "learning_rate": 0.0002,
+      "loss": 0.7054,
+      "step": 11620
+    },
+    {
+      "epoch": 1.880203702206774,
+      "grad_norm": 0.6390765309333801,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 11630
+    },
+    {
+      "epoch": 1.8818203863875191,
+      "grad_norm": 0.6589758992195129,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 11640
+    },
+    {
+      "epoch": 1.8834370705682644,
+      "grad_norm": 0.6765508651733398,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 11650
+    },
+    {
+      "epoch": 1.8850537547490098,
+      "grad_norm": 0.6527857780456543,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 11660
+    },
+    {
+      "epoch": 1.886670438929755,
+      "grad_norm": 0.6642923951148987,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 11670
+    },
+    {
+      "epoch": 1.8882871231105005,
+      "grad_norm": 0.6945584416389465,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 11680
+    },
+    {
+      "epoch": 1.8899038072912457,
+      "grad_norm": 0.694018542766571,
+      "learning_rate": 0.0002,
+      "loss": 0.7711,
+      "step": 11690
+    },
+    {
+      "epoch": 1.891520491471991,
+      "grad_norm": 0.7237417101860046,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 11700
+    },
+    {
+      "epoch": 1.8931371756527362,
+      "grad_norm": 0.7401309609413147,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 11710
+    },
+    {
+      "epoch": 1.8947538598334814,
+      "grad_norm": 0.6537784337997437,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 11720
+    },
+    {
+      "epoch": 1.8963705440142267,
+      "grad_norm": 0.7398539185523987,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 11730
+    },
+    {
+      "epoch": 1.8979872281949721,
+      "grad_norm": 0.6696075797080994,
+      "learning_rate": 0.0002,
+      "loss": 0.7561,
+      "step": 11740
+    },
+    {
+      "epoch": 1.8996039123757174,
+      "grad_norm": 0.6014142036437988,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9012205965564628,
+      "grad_norm": 0.7023524641990662,
+      "learning_rate": 0.0002,
+      "loss": 0.7714,
+      "step": 11760
+    },
+    {
+      "epoch": 1.902837280737208,
+      "grad_norm": 0.739973783493042,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9044539649179533,
+      "grad_norm": 0.5576770901679993,
+      "learning_rate": 0.0002,
+      "loss": 0.7848,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9060706490986985,
+      "grad_norm": 0.6907393932342529,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 11790
+    },
+    {
+      "epoch": 1.9076873332794437,
+      "grad_norm": 0.6934581995010376,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9093040174601892,
+      "grad_norm": 0.591774582862854,
+      "learning_rate": 0.0002,
+      "loss": 0.7199,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9109207016409344,
+      "grad_norm": 0.6249791383743286,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11820
+    },
+    {
+      "epoch": 1.9125373858216799,
+      "grad_norm": 0.6755744218826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 11830
+    },
+    {
+      "epoch": 1.914154070002425,
+      "grad_norm": 0.7286285161972046,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9157707541831703,
+      "grad_norm": 0.7867850065231323,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9173874383639156,
+      "grad_norm": 0.6283972859382629,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11860
+    },
+    {
+      "epoch": 1.9190041225446608,
+      "grad_norm": 0.605823814868927,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 11870
+    },
+    {
+      "epoch": 1.920620806725406,
+      "grad_norm": 0.5927976965904236,
+      "learning_rate": 0.0002,
+      "loss": 0.6598,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9222374909061515,
+      "grad_norm": 0.5974002480506897,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 11890
+    },
+    {
+      "epoch": 1.923854175086897,
+      "grad_norm": 0.7091866135597229,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9254708592676422,
+      "grad_norm": 0.72496497631073,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9270875434483874,
+      "grad_norm": 0.6131896376609802,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9287042276291326,
+      "grad_norm": 0.6556436419487,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9303209118098779,
+      "grad_norm": 0.622932493686676,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 11940
+    },
+    {
+      "epoch": 1.931937595990623,
+      "grad_norm": 0.6618631482124329,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9335542801713685,
+      "grad_norm": 0.630966305732727,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 11960
+    },
+    {
+      "epoch": 1.9351709643521138,
+      "grad_norm": 0.6336734890937805,
+      "learning_rate": 0.0002,
+      "loss": 0.6854,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9367876485328592,
+      "grad_norm": 0.655403196811676,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9384043327136045,
+      "grad_norm": 0.5640574097633362,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 11990
+    },
+    {
+      "epoch": 1.9400210168943497,
+      "grad_norm": 0.6322951316833496,
+      "learning_rate": 0.0002,
+      "loss": 0.7289,
+      "step": 12000
+    },
+    {
+      "epoch": 1.941637701075095,
+      "grad_norm": 0.615703821182251,
+      "learning_rate": 0.0002,
+      "loss": 0.7627,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9432543852558402,
+      "grad_norm": 0.6487536430358887,
+      "learning_rate": 0.0002,
+      "loss": 0.786,
+      "step": 12020
+    },
+    {
+      "epoch": 1.9448710694365856,
+      "grad_norm": 0.9209630489349365,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 12030
+    },
+    {
+      "epoch": 1.9464877536173308,
+      "grad_norm": 0.67485511302948,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9481044377980763,
+      "grad_norm": 0.6831230521202087,
+      "learning_rate": 0.0002,
+      "loss": 0.7551,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9497211219788215,
+      "grad_norm": 0.6578302383422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7546,
+      "step": 12060
+    },
+    {
+      "epoch": 1.9513378061595668,
+      "grad_norm": 0.9975938200950623,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12070
+    },
+    {
+      "epoch": 1.952954490340312,
+      "grad_norm": 0.6637365221977234,
+      "learning_rate": 0.0002,
+      "loss": 0.7952,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9545711745210572,
+      "grad_norm": 0.605707049369812,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9561878587018025,
+      "grad_norm": 0.6584440469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7768,
+      "step": 12100
+    },
+    {
+      "epoch": 1.957804542882548,
+      "grad_norm": 0.6070835590362549,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9594212270632931,
+      "grad_norm": 0.7862601280212402,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9610379112440386,
+      "grad_norm": 0.8175255060195923,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 12130
+    },
+    {
+      "epoch": 1.9626545954247838,
+      "grad_norm": 0.5648472905158997,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 12140
+    },
+    {
+      "epoch": 1.964271279605529,
+      "grad_norm": 0.6591973304748535,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9658879637862743,
+      "grad_norm": 0.5960676074028015,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9675046479670195,
+      "grad_norm": 0.7272544503211975,
+      "learning_rate": 0.0002,
+      "loss": 0.7254,
+      "step": 12170
+    },
+    {
+      "epoch": 1.969121332147765,
+      "grad_norm": 0.7176699042320251,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9707380163285102,
+      "grad_norm": 0.6927123665809631,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9723547005092557,
+      "grad_norm": 0.5536034107208252,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 12200
+    },
+    {
+      "epoch": 1.9739713846900009,
+      "grad_norm": 0.8348390460014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 12210
+    },
+    {
+      "epoch": 1.9755880688707461,
+      "grad_norm": 0.6591181755065918,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 12220
+    },
+    {
+      "epoch": 1.9772047530514913,
+      "grad_norm": 1.0624109506607056,
+      "learning_rate": 0.0002,
+      "loss": 0.763,
+      "step": 12230
+    },
+    {
+      "epoch": 1.9788214372322366,
+      "grad_norm": 0.9265586137771606,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 12240
+    },
+    {
+      "epoch": 1.9804381214129818,
+      "grad_norm": 0.5998196005821228,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 12250
+    },
+    {
+      "epoch": 1.9820548055937273,
+      "grad_norm": 0.6960851550102234,
+      "learning_rate": 0.0002,
+      "loss": 0.7225,
+      "step": 12260
+    },
+    {
+      "epoch": 1.9836714897744727,
+      "grad_norm": 0.7674502730369568,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 12270
+    },
+    {
+      "epoch": 1.985288173955218,
+      "grad_norm": 0.6407275795936584,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 12280
+    },
+    {
+      "epoch": 1.9869048581359632,
+      "grad_norm": 0.6673079133033752,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 12290
+    },
+    {
+      "epoch": 1.9885215423167084,
+      "grad_norm": 0.6989844441413879,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 12300
+    },
+    {
+      "epoch": 1.9901382264974536,
+      "grad_norm": 0.7564442157745361,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 12310
+    },
+    {
+      "epoch": 1.9917549106781989,
+      "grad_norm": 0.6385478973388672,
+      "learning_rate": 0.0002,
+      "loss": 0.7719,
+      "step": 12320
+    },
+    {
+      "epoch": 1.9933715948589443,
+      "grad_norm": 0.7193717956542969,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 12330
+    },
+    {
+      "epoch": 1.9949882790396896,
+      "grad_norm": 0.7987112402915955,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 12340
+    },
+    {
+      "epoch": 1.996604963220435,
+      "grad_norm": 0.7260826826095581,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 12350
+    },
+    {
+      "epoch": 1.9982216474011802,
+      "grad_norm": 0.7968255281448364,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 12360
+    },
+    {
+      "epoch": 1.9998383315819255,
+      "grad_norm": 0.6893062591552734,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1044032573699951,
+      "eval_runtime": 122.1508,
+      "eval_samples_per_second": 6.001,
+      "eval_steps_per_second": 0.753,
+      "step": 12371
+    },
+    {
+      "epoch": 2.0014550157626707,
+      "grad_norm": 0.7775409817695618,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 12380
+    },
+    {
+      "epoch": 2.003071699943416,
+      "grad_norm": 0.76218581199646,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 12390
+    },
+    {
+      "epoch": 2.004688384124161,
+      "grad_norm": 0.5677764415740967,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 12400
+    },
+    {
+      "epoch": 2.006305068304907,
+      "grad_norm": 0.808442234992981,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 12410
+    },
+    {
+      "epoch": 2.007921752485652,
+      "grad_norm": 0.7144765257835388,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 12420
+    },
+    {
+      "epoch": 2.0095384366663973,
+      "grad_norm": 0.6914031505584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 12430
+    },
+    {
+      "epoch": 2.0111551208471425,
+      "grad_norm": 0.7581454515457153,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 12440
+    },
+    {
+      "epoch": 2.0127718050278878,
+      "grad_norm": 0.8388504981994629,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 12450
+    },
+    {
+      "epoch": 2.014388489208633,
+      "grad_norm": 0.6716406941413879,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 12460
+    },
+    {
+      "epoch": 2.0160051733893782,
+      "grad_norm": 0.898902416229248,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 12470
+    },
+    {
+      "epoch": 2.0176218575701235,
+      "grad_norm": 0.6432679891586304,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 12480
+    },
+    {
+      "epoch": 2.019238541750869,
+      "grad_norm": 0.8021109104156494,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12490
+    },
+    {
+      "epoch": 2.0208552259316144,
+      "grad_norm": 0.7039216756820679,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 12500
+    },
+    {
+      "epoch": 2.0224719101123596,
+      "grad_norm": 0.646531879901886,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12510
+    },
+    {
+      "epoch": 2.024088594293105,
+      "grad_norm": 0.783704400062561,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 12520
+    },
+    {
+      "epoch": 2.02570527847385,
+      "grad_norm": 0.8805046677589417,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12530
+    },
+    {
+      "epoch": 2.0273219626545953,
+      "grad_norm": 0.7289270758628845,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0289386468353405,
+      "grad_norm": 0.71653151512146,
+      "learning_rate": 0.0002,
+      "loss": 0.6663,
+      "step": 12550
+    },
+    {
+      "epoch": 2.030555331016086,
+      "grad_norm": 0.73281329870224,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 12560
+    },
+    {
+      "epoch": 2.0321720151968314,
+      "grad_norm": 0.6657090187072754,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 12570
+    },
+    {
+      "epoch": 2.0337886993775767,
+      "grad_norm": 0.8241133093833923,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 12580
+    },
+    {
+      "epoch": 2.035405383558322,
+      "grad_norm": 0.5834135413169861,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 12590
+    },
+    {
+      "epoch": 2.037022067739067,
+      "grad_norm": 0.84502112865448,
+      "learning_rate": 0.0002,
+      "loss": 0.6188,
+      "step": 12600
+    },
+    {
+      "epoch": 2.0386387519198124,
+      "grad_norm": 0.8952481746673584,
+      "learning_rate": 0.0002,
+      "loss": 0.6349,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0402554361005576,
+      "grad_norm": 0.7801461815834045,
+      "learning_rate": 0.0002,
+      "loss": 0.6923,
+      "step": 12620
+    },
+    {
+      "epoch": 2.041872120281303,
+      "grad_norm": 0.6788367033004761,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 12630
+    },
+    {
+      "epoch": 2.0434888044620485,
+      "grad_norm": 0.7241756319999695,
+      "learning_rate": 0.0002,
+      "loss": 0.6162,
+      "step": 12640
+    },
+    {
+      "epoch": 2.0451054886427937,
+      "grad_norm": 0.6933388113975525,
+      "learning_rate": 0.0002,
+      "loss": 0.655,
+      "step": 12650
+    },
+    {
+      "epoch": 2.046722172823539,
+      "grad_norm": 0.8029746413230896,
+      "learning_rate": 0.0002,
+      "loss": 0.6431,
+      "step": 12660
+    },
+    {
+      "epoch": 2.048338857004284,
+      "grad_norm": 0.946399986743927,
+      "learning_rate": 0.0002,
+      "loss": 0.7164,
+      "step": 12670
+    },
+    {
+      "epoch": 2.0499555411850294,
+      "grad_norm": 0.7072678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.638,
+      "step": 12680
+    },
+    {
+      "epoch": 2.0515722253657747,
+      "grad_norm": 0.6810618042945862,
+      "learning_rate": 0.0002,
+      "loss": 0.6487,
+      "step": 12690
+    },
+    {
+      "epoch": 2.05318890954652,
+      "grad_norm": 0.7661160230636597,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 12700
+    },
+    {
+      "epoch": 2.0548055937272656,
+      "grad_norm": 0.6350653767585754,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 12710
+    },
+    {
+      "epoch": 2.056422277908011,
+      "grad_norm": 0.861890971660614,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 12720
+    },
+    {
+      "epoch": 2.058038962088756,
+      "grad_norm": 0.6489875912666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 12730
+    },
+    {
+      "epoch": 2.0596556462695013,
+      "grad_norm": 0.8268506526947021,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 12740
+    },
+    {
+      "epoch": 2.0612723304502465,
+      "grad_norm": 0.607679545879364,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 12750
+    },
+    {
+      "epoch": 2.0628890146309917,
+      "grad_norm": 0.6754153370857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6649,
+      "step": 12760
+    },
+    {
+      "epoch": 2.064505698811737,
+      "grad_norm": 0.7263124585151672,
+      "learning_rate": 0.0002,
+      "loss": 0.6549,
+      "step": 12770
+    },
+    {
+      "epoch": 2.0661223829924826,
+      "grad_norm": 0.6986154317855835,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 12780
+    },
+    {
+      "epoch": 2.067739067173228,
+      "grad_norm": 0.7768576741218567,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 12790
+    },
+    {
+      "epoch": 2.069355751353973,
+      "grad_norm": 0.7546762824058533,
+      "learning_rate": 0.0002,
+      "loss": 0.677,
+      "step": 12800
+    },
+    {
+      "epoch": 2.0709724355347183,
+      "grad_norm": 0.7588880062103271,
+      "learning_rate": 0.0002,
+      "loss": 0.6485,
+      "step": 12810
+    },
+    {
+      "epoch": 2.0725891197154636,
+      "grad_norm": 0.7457242608070374,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12820
+    },
+    {
+      "epoch": 2.074205803896209,
+      "grad_norm": 0.6983516812324524,
+      "learning_rate": 0.0002,
+      "loss": 0.6489,
+      "step": 12830
+    },
+    {
+      "epoch": 2.075822488076954,
+      "grad_norm": 0.7950928807258606,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 12840
+    },
+    {
+      "epoch": 2.0774391722576993,
+      "grad_norm": 0.9248087406158447,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 12850
+    },
+    {
+      "epoch": 2.079055856438445,
+      "grad_norm": 0.7229493260383606,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 12860
+    },
+    {
+      "epoch": 2.08067254061919,
+      "grad_norm": 0.5710847973823547,
+      "learning_rate": 0.0002,
+      "loss": 0.6702,
+      "step": 12870
+    },
+    {
+      "epoch": 2.0822892247999354,
+      "grad_norm": 0.9580423831939697,
+      "learning_rate": 0.0002,
+      "loss": 0.6974,
+      "step": 12880
+    },
+    {
+      "epoch": 2.0839059089806806,
+      "grad_norm": 0.7399665713310242,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 12890
+    },
+    {
+      "epoch": 2.085522593161426,
+      "grad_norm": 0.7981410622596741,
+      "learning_rate": 0.0002,
+      "loss": 0.6993,
+      "step": 12900
+    },
+    {
+      "epoch": 2.087139277342171,
+      "grad_norm": 0.870759904384613,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 12910
+    },
+    {
+      "epoch": 2.0887559615229163,
+      "grad_norm": 0.7001481652259827,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 12920
+    },
+    {
+      "epoch": 2.090372645703662,
+      "grad_norm": 0.6745418310165405,
+      "learning_rate": 0.0002,
+      "loss": 0.6383,
+      "step": 12930
+    },
+    {
+      "epoch": 2.0919893298844072,
+      "grad_norm": 0.7739067673683167,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 12940
+    },
+    {
+      "epoch": 2.0936060140651525,
+      "grad_norm": 0.6742934584617615,
+      "learning_rate": 0.0002,
+      "loss": 0.6856,
+      "step": 12950
+    },
+    {
+      "epoch": 2.0952226982458977,
+      "grad_norm": 0.7270349860191345,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 12960
+    },
+    {
+      "epoch": 2.096839382426643,
+      "grad_norm": 0.7150624394416809,
+      "learning_rate": 0.0002,
+      "loss": 0.6783,
+      "step": 12970
+    },
+    {
+      "epoch": 2.098456066607388,
+      "grad_norm": 0.7734767198562622,
+      "learning_rate": 0.0002,
+      "loss": 0.6093,
+      "step": 12980
+    },
+    {
+      "epoch": 2.1000727507881334,
+      "grad_norm": 0.7618662118911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 12990
+    },
+    {
+      "epoch": 2.101689434968879,
+      "grad_norm": 0.6557944416999817,
+      "learning_rate": 0.0002,
+      "loss": 0.6707,
+      "step": 13000
+    },
+    {
+      "epoch": 2.1033061191496243,
+      "grad_norm": 0.8786448240280151,
+      "learning_rate": 0.0002,
+      "loss": 0.7268,
+      "step": 13010
+    },
+    {
+      "epoch": 2.1049228033303695,
+      "grad_norm": 0.6878724098205566,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 13020
+    },
+    {
+      "epoch": 2.1065394875111147,
+      "grad_norm": 0.822318971157074,
+      "learning_rate": 0.0002,
+      "loss": 0.6824,
+      "step": 13030
+    },
+    {
+      "epoch": 2.10815617169186,
+      "grad_norm": 0.831468939781189,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 13040
+    },
+    {
+      "epoch": 2.109772855872605,
+      "grad_norm": 0.7699505686759949,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 13050
+    },
+    {
+      "epoch": 2.1113895400533504,
+      "grad_norm": 0.7559016346931458,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1130062242340957,
+      "grad_norm": 0.6942209601402283,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 13070
+    },
+    {
+      "epoch": 2.1146229084148414,
+      "grad_norm": 0.6098947525024414,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 13080
+    },
+    {
+      "epoch": 2.1162395925955866,
+      "grad_norm": 0.6499016284942627,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 13090
+    },
+    {
+      "epoch": 2.117856276776332,
+      "grad_norm": 0.7719953060150146,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 13100
+    },
+    {
+      "epoch": 2.119472960957077,
+      "grad_norm": 0.6708134412765503,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1210896451378223,
+      "grad_norm": 0.8119585514068604,
+      "learning_rate": 0.0002,
+      "loss": 0.6056,
+      "step": 13120
+    },
+    {
+      "epoch": 2.1227063293185675,
+      "grad_norm": 0.6947157979011536,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 13130
+    },
+    {
+      "epoch": 2.1243230134993127,
+      "grad_norm": 0.8831837773323059,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 13140
+    },
+    {
+      "epoch": 2.1259396976800584,
+      "grad_norm": 0.7266910672187805,
+      "learning_rate": 0.0002,
+      "loss": 0.6997,
+      "step": 13150
+    },
+    {
+      "epoch": 2.1275563818608036,
+      "grad_norm": 0.8864351511001587,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 13160
+    },
+    {
+      "epoch": 2.129173066041549,
+      "grad_norm": 0.8104248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.6762,
+      "step": 13170
+    },
+    {
+      "epoch": 2.130789750222294,
+      "grad_norm": 0.6077079772949219,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 13180
+    },
+    {
+      "epoch": 2.1324064344030393,
+      "grad_norm": 0.6874213814735413,
+      "learning_rate": 0.0002,
+      "loss": 0.6572,
+      "step": 13190
+    },
+    {
+      "epoch": 2.1340231185837846,
+      "grad_norm": 0.7134367823600769,
+      "learning_rate": 0.0002,
+      "loss": 0.642,
+      "step": 13200
+    },
+    {
+      "epoch": 2.13563980276453,
+      "grad_norm": 0.6101235151290894,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 13210
+    },
+    {
+      "epoch": 2.137256486945275,
+      "grad_norm": 0.6042411923408508,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13220
+    },
+    {
+      "epoch": 2.1388731711260207,
+      "grad_norm": 0.914601743221283,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 13230
+    },
+    {
+      "epoch": 2.140489855306766,
+      "grad_norm": 0.7104284167289734,
+      "learning_rate": 0.0002,
+      "loss": 0.6513,
+      "step": 13240
+    },
+    {
+      "epoch": 2.142106539487511,
+      "grad_norm": 0.664395272731781,
+      "learning_rate": 0.0002,
+      "loss": 0.6607,
+      "step": 13250
+    },
+    {
+      "epoch": 2.1437232236682564,
+      "grad_norm": 0.6991241574287415,
+      "learning_rate": 0.0002,
+      "loss": 0.7211,
+      "step": 13260
+    },
+    {
+      "epoch": 2.1453399078490016,
+      "grad_norm": 0.5469560623168945,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 13270
+    },
+    {
+      "epoch": 2.146956592029747,
+      "grad_norm": 0.8454998135566711,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13280
+    },
+    {
+      "epoch": 2.148573276210492,
+      "grad_norm": 0.7088868618011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 13290
+    },
+    {
+      "epoch": 2.1501899603912378,
+      "grad_norm": 0.7002687454223633,
+      "learning_rate": 0.0002,
+      "loss": 0.6835,
+      "step": 13300
+    },
+    {
+      "epoch": 2.151806644571983,
+      "grad_norm": 0.7785214781761169,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 13310
+    },
+    {
+      "epoch": 2.1534233287527282,
+      "grad_norm": 0.8049132227897644,
+      "learning_rate": 0.0002,
+      "loss": 0.67,
+      "step": 13320
+    },
+    {
+      "epoch": 2.1550400129334735,
+      "grad_norm": 0.8062595129013062,
+      "learning_rate": 0.0002,
+      "loss": 0.6495,
+      "step": 13330
+    },
+    {
+      "epoch": 2.1566566971142187,
+      "grad_norm": 0.6208319067955017,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 13340
+    },
+    {
+      "epoch": 2.158273381294964,
+      "grad_norm": 0.7519655823707581,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 13350
+    },
+    {
+      "epoch": 2.159890065475709,
+      "grad_norm": 0.7645747065544128,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 13360
+    },
+    {
+      "epoch": 2.1615067496564544,
+      "grad_norm": 0.6847302913665771,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 13370
+    },
+    {
+      "epoch": 2.1631234338372,
+      "grad_norm": 0.8630441427230835,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 13380
+    },
+    {
+      "epoch": 2.1647401180179453,
+      "grad_norm": 0.7947702407836914,
+      "learning_rate": 0.0002,
+      "loss": 0.6742,
+      "step": 13390
+    },
+    {
+      "epoch": 2.1663568021986905,
+      "grad_norm": 0.6836977005004883,
+      "learning_rate": 0.0002,
+      "loss": 0.7206,
+      "step": 13400
+    },
+    {
+      "epoch": 2.1679734863794358,
+      "grad_norm": 0.7340566515922546,
+      "learning_rate": 0.0002,
+      "loss": 0.6304,
+      "step": 13410
+    },
+    {
+      "epoch": 2.169590170560181,
+      "grad_norm": 0.7075738906860352,
+      "learning_rate": 0.0002,
+      "loss": 0.6528,
+      "step": 13420
+    },
+    {
+      "epoch": 2.1712068547409262,
+      "grad_norm": 0.7080879807472229,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 13430
+    },
+    {
+      "epoch": 2.1728235389216715,
+      "grad_norm": 0.6218613386154175,
+      "learning_rate": 0.0002,
+      "loss": 0.6615,
+      "step": 13440
+    },
+    {
+      "epoch": 2.174440223102417,
+      "grad_norm": 0.8211479187011719,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 13450
+    },
+    {
+      "epoch": 2.1760569072831624,
+      "grad_norm": 0.864466667175293,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 13460
+    },
+    {
+      "epoch": 2.1776735914639076,
+      "grad_norm": 0.7943857908248901,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 13470
+    },
+    {
+      "epoch": 2.179290275644653,
+      "grad_norm": 0.78728187084198,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 13480
+    },
+    {
+      "epoch": 2.180906959825398,
+      "grad_norm": 0.697527289390564,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 13490
+    },
+    {
+      "epoch": 2.1825236440061433,
+      "grad_norm": 0.8205804228782654,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 13500
+    },
+    {
+      "epoch": 2.1841403281868885,
+      "grad_norm": 0.8709042072296143,
+      "learning_rate": 0.0002,
+      "loss": 0.7227,
+      "step": 13510
+    },
+    {
+      "epoch": 2.1857570123676338,
+      "grad_norm": 0.6228537559509277,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 13520
+    },
+    {
+      "epoch": 2.1873736965483794,
+      "grad_norm": 0.9566980004310608,
+      "learning_rate": 0.0002,
+      "loss": 0.7025,
+      "step": 13530
+    },
+    {
+      "epoch": 2.1889903807291247,
+      "grad_norm": 0.7128894329071045,
+      "learning_rate": 0.0002,
+      "loss": 0.6755,
+      "step": 13540
+    },
+    {
+      "epoch": 2.19060706490987,
+      "grad_norm": 0.6888654232025146,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 13550
+    },
+    {
+      "epoch": 2.192223749090615,
+      "grad_norm": 0.6444337368011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6961,
+      "step": 13560
+    },
+    {
+      "epoch": 2.1938404332713604,
+      "grad_norm": 0.8008806705474854,
+      "learning_rate": 0.0002,
+      "loss": 0.656,
+      "step": 13570
+    },
+    {
+      "epoch": 2.1954571174521056,
+      "grad_norm": 0.8482748866081238,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 13580
+    },
+    {
+      "epoch": 2.197073801632851,
+      "grad_norm": 0.8584157228469849,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 13590
+    },
+    {
+      "epoch": 2.1986904858135965,
+      "grad_norm": 0.7513734698295593,
+      "learning_rate": 0.0002,
+      "loss": 0.7014,
+      "step": 13600
+    },
+    {
+      "epoch": 2.2003071699943417,
+      "grad_norm": 0.7864262461662292,
+      "learning_rate": 0.0002,
+      "loss": 0.6632,
+      "step": 13610
+    },
+    {
+      "epoch": 2.201923854175087,
+      "grad_norm": 0.8493645191192627,
+      "learning_rate": 0.0002,
+      "loss": 0.6879,
+      "step": 13620
+    },
+    {
+      "epoch": 2.203540538355832,
+      "grad_norm": 0.6902140974998474,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 13630
+    },
+    {
+      "epoch": 2.2051572225365774,
+      "grad_norm": 0.8711254596710205,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 13640
+    },
+    {
+      "epoch": 2.2067739067173227,
+      "grad_norm": 0.7832191586494446,
+      "learning_rate": 0.0002,
+      "loss": 0.6359,
+      "step": 13650
+    },
+    {
+      "epoch": 2.208390590898068,
+      "grad_norm": 0.5668176412582397,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2100072750788136,
+      "grad_norm": 0.8648375272750854,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13670
+    },
+    {
+      "epoch": 2.211623959259559,
+      "grad_norm": 0.7643089890480042,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13680
+    },
+    {
+      "epoch": 2.213240643440304,
+      "grad_norm": 0.6293777823448181,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13690
+    },
+    {
+      "epoch": 2.2148573276210493,
+      "grad_norm": 0.6459372639656067,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 13700
+    },
+    {
+      "epoch": 2.2164740118017945,
+      "grad_norm": 0.7060744166374207,
+      "learning_rate": 0.0002,
+      "loss": 0.6526,
+      "step": 13710
+    },
+    {
+      "epoch": 2.2180906959825397,
+      "grad_norm": 0.674109160900116,
+      "learning_rate": 0.0002,
+      "loss": 0.7101,
+      "step": 13720
+    },
+    {
+      "epoch": 2.219707380163285,
+      "grad_norm": 0.830392062664032,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13730
+    },
+    {
+      "epoch": 2.2213240643440306,
+      "grad_norm": 0.6474477052688599,
+      "learning_rate": 0.0002,
+      "loss": 0.6733,
+      "step": 13740
+    },
+    {
+      "epoch": 2.222940748524776,
+      "grad_norm": 0.7037909626960754,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13750
+    },
+    {
+      "epoch": 2.224557432705521,
+      "grad_norm": 0.6554131507873535,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 13760
+    },
+    {
+      "epoch": 2.2261741168862663,
+      "grad_norm": 0.7822230458259583,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 13770
+    },
+    {
+      "epoch": 2.2277908010670116,
+      "grad_norm": 0.9082167744636536,
+      "learning_rate": 0.0002,
+      "loss": 0.6505,
+      "step": 13780
+    },
+    {
+      "epoch": 2.229407485247757,
+      "grad_norm": 0.7918276190757751,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 13790
+    },
+    {
+      "epoch": 2.231024169428502,
+      "grad_norm": 0.7354569435119629,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 13800
+    },
+    {
+      "epoch": 2.2326408536092472,
+      "grad_norm": 0.8265249133110046,
+      "learning_rate": 0.0002,
+      "loss": 0.6503,
+      "step": 13810
+    },
+    {
+      "epoch": 2.234257537789993,
+      "grad_norm": 0.6653847098350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6871,
+      "step": 13820
+    },
+    {
+      "epoch": 2.235874221970738,
+      "grad_norm": 0.7157923579216003,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13830
+    },
+    {
+      "epoch": 2.2374909061514834,
+      "grad_norm": 0.7110323309898376,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 13840
+    },
+    {
+      "epoch": 2.2391075903322286,
+      "grad_norm": 0.7155357599258423,
+      "learning_rate": 0.0002,
+      "loss": 0.6913,
+      "step": 13850
+    },
+    {
+      "epoch": 2.240724274512974,
+      "grad_norm": 1.0177817344665527,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 13860
+    },
+    {
+      "epoch": 2.242340958693719,
+      "grad_norm": 0.7601948380470276,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13870
+    },
+    {
+      "epoch": 2.2439576428744643,
+      "grad_norm": 0.7628820538520813,
+      "learning_rate": 0.0002,
+      "loss": 0.6679,
+      "step": 13880
+    },
+    {
+      "epoch": 2.24557432705521,
+      "grad_norm": 0.7089297771453857,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 13890
+    },
+    {
+      "epoch": 2.247191011235955,
+      "grad_norm": 0.695178210735321,
+      "learning_rate": 0.0002,
+      "loss": 0.7236,
+      "step": 13900
+    },
+    {
+      "epoch": 2.2488076954167004,
+      "grad_norm": 0.7631948590278625,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 13910
+    },
+    {
+      "epoch": 2.2504243795974457,
+      "grad_norm": 0.8203101754188538,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 13920
+    },
+    {
+      "epoch": 2.252041063778191,
+      "grad_norm": 0.8099079728126526,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13930
+    },
+    {
+      "epoch": 2.253657747958936,
+      "grad_norm": 0.6498546004295349,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 13940
+    },
+    {
+      "epoch": 2.2552744321396814,
+      "grad_norm": 0.7797415256500244,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 13950
+    },
+    {
+      "epoch": 2.2568911163204266,
+      "grad_norm": 0.8254124522209167,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 13960
+    },
+    {
+      "epoch": 2.2585078005011723,
+      "grad_norm": 0.6327953338623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 13970
+    },
+    {
+      "epoch": 2.2601244846819175,
+      "grad_norm": 0.734194278717041,
+      "learning_rate": 0.0002,
+      "loss": 0.668,
+      "step": 13980
+    },
+    {
+      "epoch": 2.2617411688626627,
+      "grad_norm": 0.9014202952384949,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 13990
+    },
+    {
+      "epoch": 2.263357853043408,
+      "grad_norm": 0.7643631100654602,
+      "learning_rate": 0.0002,
+      "loss": 0.692,
+      "step": 14000
+    },
+    {
+      "epoch": 2.264974537224153,
+      "grad_norm": 0.8882834911346436,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 14010
+    },
+    {
+      "epoch": 2.2665912214048984,
+      "grad_norm": 0.7975873351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14020
+    },
+    {
+      "epoch": 2.2682079055856437,
+      "grad_norm": 0.7765783071517944,
+      "learning_rate": 0.0002,
+      "loss": 0.7193,
+      "step": 14030
+    },
+    {
+      "epoch": 2.2698245897663893,
+      "grad_norm": 0.8846288323402405,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 14040
+    },
+    {
+      "epoch": 2.2714412739471346,
+      "grad_norm": 0.9006744027137756,
+      "learning_rate": 0.0002,
+      "loss": 0.6494,
+      "step": 14050
+    },
+    {
+      "epoch": 2.27305795812788,
+      "grad_norm": 0.7420173287391663,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 14060
+    },
+    {
+      "epoch": 2.274674642308625,
+      "grad_norm": 0.7956424951553345,
+      "learning_rate": 0.0002,
+      "loss": 0.7068,
+      "step": 14070
+    },
+    {
+      "epoch": 2.2762913264893703,
+      "grad_norm": 0.7783209085464478,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 14080
+    },
+    {
+      "epoch": 2.2779080106701155,
+      "grad_norm": 0.7597188949584961,
+      "learning_rate": 0.0002,
+      "loss": 0.7202,
+      "step": 14090
+    },
+    {
+      "epoch": 2.2795246948508607,
+      "grad_norm": 0.6718921661376953,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14100
+    },
+    {
+      "epoch": 2.281141379031606,
+      "grad_norm": 0.7528082132339478,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 14110
+    },
+    {
+      "epoch": 2.2827580632123516,
+      "grad_norm": 0.8379864692687988,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 14120
+    },
+    {
+      "epoch": 2.284374747393097,
+      "grad_norm": 0.748613715171814,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 14130
+    },
+    {
+      "epoch": 2.285991431573842,
+      "grad_norm": 0.7435423135757446,
+      "learning_rate": 0.0002,
+      "loss": 0.6641,
+      "step": 14140
+    },
+    {
+      "epoch": 2.2876081157545873,
+      "grad_norm": 0.7580803632736206,
+      "learning_rate": 0.0002,
+      "loss": 0.6849,
+      "step": 14150
+    },
+    {
+      "epoch": 2.2892247999353326,
+      "grad_norm": 0.6278321146965027,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 14160
+    },
+    {
+      "epoch": 2.290841484116078,
+      "grad_norm": 0.7663896083831787,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 14170
+    },
+    {
+      "epoch": 2.292458168296823,
+      "grad_norm": 0.9716812372207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 14180
+    },
+    {
+      "epoch": 2.2940748524775687,
+      "grad_norm": 0.8993458151817322,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14190
+    },
+    {
+      "epoch": 2.295691536658314,
+      "grad_norm": 0.6156117916107178,
+      "learning_rate": 0.0002,
+      "loss": 0.6172,
+      "step": 14200
+    },
+    {
+      "epoch": 2.297308220839059,
+      "grad_norm": 0.8911278247833252,
+      "learning_rate": 0.0002,
+      "loss": 0.6318,
+      "step": 14210
+    },
+    {
+      "epoch": 2.2989249050198044,
+      "grad_norm": 0.6422147154808044,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 14220
+    },
+    {
+      "epoch": 2.3005415892005496,
+      "grad_norm": 0.6866879463195801,
+      "learning_rate": 0.0002,
+      "loss": 0.6795,
+      "step": 14230
+    },
+    {
+      "epoch": 2.302158273381295,
+      "grad_norm": 0.9297130107879639,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 14240
+    },
+    {
+      "epoch": 2.30377495756204,
+      "grad_norm": 0.7501356601715088,
+      "learning_rate": 0.0002,
+      "loss": 0.6823,
+      "step": 14250
+    },
+    {
+      "epoch": 2.3053916417427853,
+      "grad_norm": 0.8363515138626099,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 14260
+    },
+    {
+      "epoch": 2.307008325923531,
+      "grad_norm": 0.9083868265151978,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3086250101042762,
+      "grad_norm": 0.7791516780853271,
+      "learning_rate": 0.0002,
+      "loss": 0.6862,
+      "step": 14280
+    },
+    {
+      "epoch": 2.3102416942850215,
+      "grad_norm": 0.8766953349113464,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14290
+    },
+    {
+      "epoch": 2.3118583784657667,
+      "grad_norm": 0.7916635274887085,
+      "learning_rate": 0.0002,
+      "loss": 0.6698,
+      "step": 14300
+    },
+    {
+      "epoch": 2.313475062646512,
+      "grad_norm": 0.627525269985199,
+      "learning_rate": 0.0002,
+      "loss": 0.6927,
+      "step": 14310
+    },
+    {
+      "epoch": 2.315091746827257,
+      "grad_norm": 0.8856783509254456,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 14320
+    },
+    {
+      "epoch": 2.316708431008003,
+      "grad_norm": 0.6758689284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 14330
+    },
+    {
+      "epoch": 2.318325115188748,
+      "grad_norm": 0.6428321003913879,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 14340
+    },
+    {
+      "epoch": 2.3199417993694933,
+      "grad_norm": 0.9032121300697327,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 14350
+    },
+    {
+      "epoch": 2.3215584835502385,
+      "grad_norm": 0.8035986423492432,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14360
+    },
+    {
+      "epoch": 2.3231751677309838,
+      "grad_norm": 0.7974579334259033,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14370
+    },
+    {
+      "epoch": 2.324791851911729,
+      "grad_norm": 0.8356034755706787,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 14380
+    },
+    {
+      "epoch": 2.326408536092474,
+      "grad_norm": 0.998760998249054,
+      "learning_rate": 0.0002,
+      "loss": 0.6623,
+      "step": 14390
+    },
+    {
+      "epoch": 2.3280252202732195,
+      "grad_norm": 0.6518142223358154,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 14400
+    },
+    {
+      "epoch": 2.3296419044539647,
+      "grad_norm": 0.7443506717681885,
+      "learning_rate": 0.0002,
+      "loss": 0.7146,
+      "step": 14410
+    },
+    {
+      "epoch": 2.3312585886347104,
+      "grad_norm": 0.8436172604560852,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14420
+    },
+    {
+      "epoch": 2.3328752728154556,
+      "grad_norm": 0.7411080598831177,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 14430
+    },
+    {
+      "epoch": 2.334491956996201,
+      "grad_norm": 0.8839048743247986,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 14440
+    },
+    {
+      "epoch": 2.336108641176946,
+      "grad_norm": 0.8360885977745056,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 14450
+    },
+    {
+      "epoch": 2.3377253253576913,
+      "grad_norm": 0.7608986496925354,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 14460
+    },
+    {
+      "epoch": 2.3393420095384365,
+      "grad_norm": 0.8179867267608643,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14470
+    },
+    {
+      "epoch": 2.340958693719182,
+      "grad_norm": 0.5989999771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14480
+    },
+    {
+      "epoch": 2.3425753778999274,
+      "grad_norm": 0.9450054168701172,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 14490
+    },
+    {
+      "epoch": 2.3441920620806727,
+      "grad_norm": 0.7885149717330933,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 14500
+    },
+    {
+      "epoch": 2.345808746261418,
+      "grad_norm": 0.8152616620063782,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 14510
+    },
+    {
+      "epoch": 2.347425430442163,
+      "grad_norm": 0.7193838953971863,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 14520
+    },
+    {
+      "epoch": 2.3490421146229084,
+      "grad_norm": 0.6701092720031738,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 14530
+    },
+    {
+      "epoch": 2.3506587988036536,
+      "grad_norm": 0.7529364228248596,
+      "learning_rate": 0.0002,
+      "loss": 0.6559,
+      "step": 14540
+    },
+    {
+      "epoch": 2.352275482984399,
+      "grad_norm": 0.6599733829498291,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 14550
+    },
+    {
+      "epoch": 2.353892167165144,
+      "grad_norm": 0.9502474069595337,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14560
+    },
+    {
+      "epoch": 2.3555088513458897,
+      "grad_norm": 0.7619650959968567,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 14570
+    },
+    {
+      "epoch": 2.357125535526635,
+      "grad_norm": 0.9854652285575867,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14580
+    },
+    {
+      "epoch": 2.35874221970738,
+      "grad_norm": 0.727439284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 14590
+    },
+    {
+      "epoch": 2.3603589038881254,
+      "grad_norm": 0.6994746327400208,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 14600
+    },
+    {
+      "epoch": 2.3619755880688706,
+      "grad_norm": 0.7117531299591064,
+      "learning_rate": 0.0002,
+      "loss": 0.6587,
+      "step": 14610
+    },
+    {
+      "epoch": 2.363592272249616,
+      "grad_norm": 0.6403067708015442,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 14620
+    },
+    {
+      "epoch": 2.3652089564303616,
+      "grad_norm": 0.8377841711044312,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14630
+    },
+    {
+      "epoch": 2.366825640611107,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14640
+    },
+    {
+      "epoch": 2.368442324791852,
+      "grad_norm": 0.8418586254119873,
+      "learning_rate": 0.0002,
+      "loss": 0.6552,
+      "step": 14650
+    },
+    {
+      "epoch": 2.3700590089725972,
+      "grad_norm": 0.6178573369979858,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14660
+    },
+    {
+      "epoch": 2.3716756931533425,
+      "grad_norm": 0.6368302702903748,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 14670
+    },
+    {
+      "epoch": 2.3732923773340877,
+      "grad_norm": 0.9122977256774902,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14680
+    },
+    {
+      "epoch": 2.374909061514833,
+      "grad_norm": 0.7086195349693298,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 14690
+    },
+    {
+      "epoch": 2.376525745695578,
+      "grad_norm": 0.7500800490379333,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 14700
+    },
+    {
+      "epoch": 2.378142429876324,
+      "grad_norm": 0.6634900569915771,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 14710
+    },
+    {
+      "epoch": 2.379759114057069,
+      "grad_norm": 0.839898407459259,
+      "learning_rate": 0.0002,
+      "loss": 0.6614,
+      "step": 14720
+    },
+    {
+      "epoch": 2.3813757982378143,
+      "grad_norm": 0.7578426003456116,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14730
+    },
+    {
+      "epoch": 2.3829924824185595,
+      "grad_norm": 1.0213173627853394,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3846091665993048,
+      "grad_norm": 0.7855949401855469,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 14750
+    },
+    {
+      "epoch": 2.38622585078005,
+      "grad_norm": 0.7224128842353821,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 14760
+    },
+    {
+      "epoch": 2.3878425349607952,
+      "grad_norm": 0.8040381669998169,
+      "learning_rate": 0.0002,
+      "loss": 0.7017,
+      "step": 14770
+    },
+    {
+      "epoch": 2.389459219141541,
+      "grad_norm": 0.7705281376838684,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 14780
+    },
+    {
+      "epoch": 2.391075903322286,
+      "grad_norm": 0.667966902256012,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 14790
+    },
+    {
+      "epoch": 2.3926925875030314,
+      "grad_norm": 0.6611011028289795,
+      "learning_rate": 0.0002,
+      "loss": 0.7061,
+      "step": 14800
+    },
+    {
+      "epoch": 2.3943092716837766,
+      "grad_norm": 0.6862651705741882,
+      "learning_rate": 0.0002,
+      "loss": 0.6527,
+      "step": 14810
+    },
+    {
+      "epoch": 2.395925955864522,
+      "grad_norm": 0.8086010217666626,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 14820
+    },
+    {
+      "epoch": 2.397542640045267,
+      "grad_norm": 0.7189689874649048,
+      "learning_rate": 0.0002,
+      "loss": 0.7189,
+      "step": 14830
+    },
+    {
+      "epoch": 2.3991593242260123,
+      "grad_norm": 0.6280009150505066,
+      "learning_rate": 0.0002,
+      "loss": 0.6709,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4007760084067575,
+      "grad_norm": 0.7826612591743469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14850
+    },
+    {
+      "epoch": 2.402392692587503,
+      "grad_norm": 0.7681610584259033,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4040093767682484,
+      "grad_norm": 0.720966100692749,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 14870
+    },
+    {
+      "epoch": 2.4056260609489937,
+      "grad_norm": 0.8202250599861145,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 14880
+    },
+    {
+      "epoch": 2.407242745129739,
+      "grad_norm": 0.786212682723999,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 14890
+    },
+    {
+      "epoch": 2.408859429310484,
+      "grad_norm": 0.6647164821624756,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 14900
+    },
+    {
+      "epoch": 2.4104761134912294,
+      "grad_norm": 0.7566399574279785,
+      "learning_rate": 0.0002,
+      "loss": 0.6706,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4120927976719746,
+      "grad_norm": 0.748814582824707,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 14920
+    },
+    {
+      "epoch": 2.4137094818527203,
+      "grad_norm": 0.7624038457870483,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14930
+    },
+    {
+      "epoch": 2.4153261660334655,
+      "grad_norm": 0.8267335295677185,
+      "learning_rate": 0.0002,
+      "loss": 0.6483,
+      "step": 14940
+    },
+    {
+      "epoch": 2.4169428502142107,
+      "grad_norm": 0.8785360455513,
+      "learning_rate": 0.0002,
+      "loss": 0.6612,
+      "step": 14950
+    },
+    {
+      "epoch": 2.418559534394956,
+      "grad_norm": 0.679887592792511,
+      "learning_rate": 0.0002,
+      "loss": 0.6718,
+      "step": 14960
+    },
+    {
+      "epoch": 2.420176218575701,
+      "grad_norm": 0.7218474745750427,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4217929027564464,
+      "grad_norm": 0.6342799663543701,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14980
+    },
+    {
+      "epoch": 2.4234095869371917,
+      "grad_norm": 0.7098712921142578,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 14990
+    },
+    {
+      "epoch": 2.425026271117937,
+      "grad_norm": 0.7497431635856628,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4266429552986826,
+      "grad_norm": 0.934836208820343,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15010
+    },
+    {
+      "epoch": 2.428259639479428,
+      "grad_norm": 0.8430966734886169,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 15020
+    },
+    {
+      "epoch": 2.429876323660173,
+      "grad_norm": 0.7032104730606079,
+      "learning_rate": 0.0002,
+      "loss": 0.6985,
+      "step": 15030
+    },
+    {
+      "epoch": 2.4314930078409183,
+      "grad_norm": 0.7746111750602722,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 15040
+    },
+    {
+      "epoch": 2.4331096920216635,
+      "grad_norm": 0.7661406397819519,
+      "learning_rate": 0.0002,
+      "loss": 0.7177,
+      "step": 15050
+    },
+    {
+      "epoch": 2.4347263762024087,
+      "grad_norm": 0.6941645741462708,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 15060
+    },
+    {
+      "epoch": 2.436343060383154,
+      "grad_norm": 0.7487249374389648,
+      "learning_rate": 0.0002,
+      "loss": 0.6421,
+      "step": 15070
+    },
+    {
+      "epoch": 2.4379597445638996,
+      "grad_norm": 0.7639912962913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 15080
+    },
+    {
+      "epoch": 2.439576428744645,
+      "grad_norm": 0.7708953619003296,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 15090
+    },
+    {
+      "epoch": 2.44119311292539,
+      "grad_norm": 0.9135832190513611,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15100
+    },
+    {
+      "epoch": 2.4428097971061353,
+      "grad_norm": 0.8283005356788635,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 15110
+    },
+    {
+      "epoch": 2.4444264812868806,
+      "grad_norm": 0.925299346446991,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 15120
+    },
+    {
+      "epoch": 2.446043165467626,
+      "grad_norm": 0.7013528943061829,
+      "learning_rate": 0.0002,
+      "loss": 0.687,
+      "step": 15130
+    },
+    {
+      "epoch": 2.447659849648371,
+      "grad_norm": 0.622303307056427,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 15140
+    },
+    {
+      "epoch": 2.4492765338291163,
+      "grad_norm": 0.876569390296936,
+      "learning_rate": 0.0002,
+      "loss": 0.6676,
+      "step": 15150
+    },
+    {
+      "epoch": 2.450893218009862,
+      "grad_norm": 0.6836351752281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 15160
+    },
+    {
+      "epoch": 2.452509902190607,
+      "grad_norm": 0.7886684536933899,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 15170
+    },
+    {
+      "epoch": 2.4541265863713524,
+      "grad_norm": 0.6647440791130066,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 15180
+    },
+    {
+      "epoch": 2.4557432705520976,
+      "grad_norm": 0.7477722764015198,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 15190
+    },
+    {
+      "epoch": 2.457359954732843,
+      "grad_norm": 0.8192033767700195,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 15200
+    },
+    {
+      "epoch": 2.458976638913588,
+      "grad_norm": 0.847537100315094,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 15210
+    },
+    {
+      "epoch": 2.4605933230943338,
+      "grad_norm": 0.9027776122093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 15220
+    },
+    {
+      "epoch": 2.462210007275079,
+      "grad_norm": 0.7217772006988525,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 15230
+    },
+    {
+      "epoch": 2.4638266914558242,
+      "grad_norm": 0.7994546294212341,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 15240
+    },
+    {
+      "epoch": 2.4654433756365695,
+      "grad_norm": 0.939916729927063,
+      "learning_rate": 0.0002,
+      "loss": 0.6828,
+      "step": 15250
+    },
+    {
+      "epoch": 2.4670600598173147,
+      "grad_norm": 1.0009053945541382,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15260
+    },
+    {
+      "epoch": 2.46867674399806,
+      "grad_norm": 0.625555694103241,
+      "learning_rate": 0.0002,
+      "loss": 0.643,
+      "step": 15270
+    },
+    {
+      "epoch": 2.470293428178805,
+      "grad_norm": 0.7924878597259521,
+      "learning_rate": 0.0002,
+      "loss": 0.688,
+      "step": 15280
+    },
+    {
+      "epoch": 2.4719101123595504,
+      "grad_norm": 0.8536689877510071,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 15290
+    },
+    {
+      "epoch": 2.4735267965402956,
+      "grad_norm": 0.8572589755058289,
+      "learning_rate": 0.0002,
+      "loss": 0.6924,
+      "step": 15300
+    },
+    {
+      "epoch": 2.4751434807210413,
+      "grad_norm": 0.773279070854187,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 15310
+    },
+    {
+      "epoch": 2.4767601649017865,
+      "grad_norm": 0.7708749771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 15320
+    },
+    {
+      "epoch": 2.4783768490825318,
+      "grad_norm": 0.770905077457428,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15330
+    },
+    {
+      "epoch": 2.479993533263277,
+      "grad_norm": 0.8238571882247925,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 15340
+    },
+    {
+      "epoch": 2.481610217444022,
+      "grad_norm": 0.7670477032661438,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15350
+    },
+    {
+      "epoch": 2.4832269016247674,
+      "grad_norm": 0.905036985874176,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 15360
+    },
+    {
+      "epoch": 2.484843585805513,
+      "grad_norm": 0.6672089695930481,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 15370
+    },
+    {
+      "epoch": 2.4864602699862584,
+      "grad_norm": 0.625095784664154,
+      "learning_rate": 0.0002,
+      "loss": 0.6722,
+      "step": 15380
+    },
+    {
+      "epoch": 2.4880769541670036,
+      "grad_norm": 0.679772675037384,
+      "learning_rate": 0.0002,
+      "loss": 0.6396,
+      "step": 15390
+    },
+    {
+      "epoch": 2.489693638347749,
+      "grad_norm": 0.711492121219635,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 15400
+    },
+    {
+      "epoch": 2.491310322528494,
+      "grad_norm": 0.876189112663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6966,
+      "step": 15410
+    },
+    {
+      "epoch": 2.4929270067092393,
+      "grad_norm": 0.7236915230751038,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 15420
+    },
+    {
+      "epoch": 2.4945436908899845,
+      "grad_norm": 0.6629832983016968,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 15430
+    },
+    {
+      "epoch": 2.4961603750707297,
+      "grad_norm": 0.9756859540939331,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 15440
+    },
+    {
+      "epoch": 2.4977770592514754,
+      "grad_norm": 0.6896940469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 15450
+    },
+    {
+      "epoch": 2.4993937434322206,
+      "grad_norm": 0.7105149626731873,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 15460
+    },
+    {
+      "epoch": 2.501010427612966,
+      "grad_norm": 0.8374546766281128,
+      "learning_rate": 0.0002,
+      "loss": 0.6859,
+      "step": 15470
+    },
+    {
+      "epoch": 2.502627111793711,
+      "grad_norm": 0.7320070266723633,
+      "learning_rate": 0.0002,
+      "loss": 0.6512,
+      "step": 15480
+    },
+    {
+      "epoch": 2.5042437959744563,
+      "grad_norm": 0.8306367993354797,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 15490
+    },
+    {
+      "epoch": 2.5058604801552016,
+      "grad_norm": 0.7472721338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.7253,
+      "step": 15500
+    },
+    {
+      "epoch": 2.507477164335947,
+      "grad_norm": 0.6147692203521729,
+      "learning_rate": 0.0002,
+      "loss": 0.6699,
+      "step": 15510
+    },
+    {
+      "epoch": 2.5090938485166925,
+      "grad_norm": 0.7788505554199219,
+      "learning_rate": 0.0002,
+      "loss": 0.7158,
+      "step": 15520
+    },
+    {
+      "epoch": 2.5107105326974377,
+      "grad_norm": 0.8807527422904968,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 15530
+    },
+    {
+      "epoch": 2.512327216878183,
+      "grad_norm": 0.7521643042564392,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 15540
+    },
+    {
+      "epoch": 2.513943901058928,
+      "grad_norm": 0.6900225281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15550
+    },
+    {
+      "epoch": 2.5155605852396734,
+      "grad_norm": 0.6601938605308533,
+      "learning_rate": 0.0002,
+      "loss": 0.6769,
+      "step": 15560
+    },
+    {
+      "epoch": 2.5171772694204186,
+      "grad_norm": 0.8179984092712402,
+      "learning_rate": 0.0002,
+      "loss": 0.6648,
+      "step": 15570
+    },
+    {
+      "epoch": 2.518793953601164,
+      "grad_norm": 0.792556881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.7028,
+      "step": 15580
+    },
+    {
+      "epoch": 2.520410637781909,
+      "grad_norm": 0.7081938982009888,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 15590
+    },
+    {
+      "epoch": 2.5220273219626543,
+      "grad_norm": 0.8733121156692505,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 15600
+    },
+    {
+      "epoch": 2.5236440061434,
+      "grad_norm": 0.7980992794036865,
+      "learning_rate": 0.0002,
+      "loss": 0.6969,
+      "step": 15610
+    },
+    {
+      "epoch": 2.5252606903241452,
+      "grad_norm": 0.883664071559906,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 15620
+    },
+    {
+      "epoch": 2.5268773745048905,
+      "grad_norm": 0.6963341236114502,
+      "learning_rate": 0.0002,
+      "loss": 0.7022,
+      "step": 15630
+    },
+    {
+      "epoch": 2.5284940586856357,
+      "grad_norm": 0.6433573365211487,
+      "learning_rate": 0.0002,
+      "loss": 0.7334,
+      "step": 15640
+    },
+    {
+      "epoch": 2.530110742866381,
+      "grad_norm": 0.8538183569908142,
+      "learning_rate": 0.0002,
+      "loss": 0.6889,
+      "step": 15650
+    },
+    {
+      "epoch": 2.5317274270471266,
+      "grad_norm": 0.9748201370239258,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 15660
+    },
+    {
+      "epoch": 2.533344111227872,
+      "grad_norm": 0.7670575380325317,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 15670
+    },
+    {
+      "epoch": 2.534960795408617,
+      "grad_norm": 0.8738890290260315,
+      "learning_rate": 0.0002,
+      "loss": 0.6435,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5365774795893623,
+      "grad_norm": 0.8391636610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.6802,
+      "step": 15690
+    },
+    {
+      "epoch": 2.5381941637701075,
+      "grad_norm": 0.7239366769790649,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 15700
+    },
+    {
+      "epoch": 2.5398108479508528,
+      "grad_norm": 0.8498379588127136,
+      "learning_rate": 0.0002,
+      "loss": 0.7011,
+      "step": 15710
+    },
+    {
+      "epoch": 2.541427532131598,
+      "grad_norm": 0.8029484152793884,
+      "learning_rate": 0.0002,
+      "loss": 0.6998,
+      "step": 15720
+    },
+    {
+      "epoch": 2.5430442163123432,
+      "grad_norm": 1.0639333724975586,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 15730
+    },
+    {
+      "epoch": 2.5446609004930885,
+      "grad_norm": 0.6401297450065613,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 15740
+    },
+    {
+      "epoch": 2.5462775846738337,
+      "grad_norm": 0.7123814821243286,
+      "learning_rate": 0.0002,
+      "loss": 0.7196,
+      "step": 15750
+    },
+    {
+      "epoch": 2.5478942688545794,
+      "grad_norm": 0.7874974608421326,
+      "learning_rate": 0.0002,
+      "loss": 0.654,
+      "step": 15760
+    },
+    {
+      "epoch": 2.5495109530353246,
+      "grad_norm": 0.8046808838844299,
+      "learning_rate": 0.0002,
+      "loss": 0.6721,
+      "step": 15770
+    },
+    {
+      "epoch": 2.55112763721607,
+      "grad_norm": 0.7888661623001099,
+      "learning_rate": 0.0002,
+      "loss": 0.6665,
+      "step": 15780
+    },
+    {
+      "epoch": 2.552744321396815,
+      "grad_norm": 0.8445866107940674,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15790
+    },
+    {
+      "epoch": 2.5543610055775603,
+      "grad_norm": 0.7475846409797668,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 15800
+    },
+    {
+      "epoch": 2.555977689758306,
+      "grad_norm": 0.7455102801322937,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 15810
+    },
+    {
+      "epoch": 2.557594373939051,
+      "grad_norm": 0.8226983547210693,
+      "learning_rate": 0.0002,
+      "loss": 0.6932,
+      "step": 15820
+    },
+    {
+      "epoch": 2.5592110581197964,
+      "grad_norm": 0.8920368552207947,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 15830
+    },
+    {
+      "epoch": 2.5608277423005417,
+      "grad_norm": 0.8413904905319214,
+      "learning_rate": 0.0002,
+      "loss": 0.6297,
+      "step": 15840
+    },
+    {
+      "epoch": 2.562444426481287,
+      "grad_norm": 0.8483649492263794,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 15850
+    },
+    {
+      "epoch": 2.564061110662032,
+      "grad_norm": 0.5923284292221069,
+      "learning_rate": 0.0002,
+      "loss": 0.6957,
+      "step": 15860
+    },
+    {
+      "epoch": 2.5656777948427774,
+      "grad_norm": 0.8518726229667664,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 15870
+    },
+    {
+      "epoch": 2.5672944790235226,
+      "grad_norm": 0.731235146522522,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 15880
+    },
+    {
+      "epoch": 2.568911163204268,
+      "grad_norm": 0.7517194151878357,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 15890
+    },
+    {
+      "epoch": 2.5705278473850135,
+      "grad_norm": 0.8378692269325256,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 15900
+    },
+    {
+      "epoch": 2.5721445315657587,
+      "grad_norm": 0.843701958656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 15910
+    },
+    {
+      "epoch": 2.573761215746504,
+      "grad_norm": 0.7254629731178284,
+      "learning_rate": 0.0002,
+      "loss": 0.6882,
+      "step": 15920
+    },
+    {
+      "epoch": 2.575377899927249,
+      "grad_norm": 0.8863335847854614,
+      "learning_rate": 0.0002,
+      "loss": 0.6872,
+      "step": 15930
+    },
+    {
+      "epoch": 2.5769945841079944,
+      "grad_norm": 0.7675097584724426,
+      "learning_rate": 0.0002,
+      "loss": 0.6813,
+      "step": 15940
+    },
+    {
+      "epoch": 2.5786112682887397,
+      "grad_norm": 0.82063889503479,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 15950
+    },
+    {
+      "epoch": 2.5802279524694853,
+      "grad_norm": 0.7729717493057251,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 15960
+    },
+    {
+      "epoch": 2.5818446366502306,
+      "grad_norm": 0.8301846981048584,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 15970
+    },
+    {
+      "epoch": 2.583461320830976,
+      "grad_norm": 0.7906861305236816,
+      "learning_rate": 0.0002,
+      "loss": 0.6897,
+      "step": 15980
+    },
+    {
+      "epoch": 2.585078005011721,
+      "grad_norm": 0.6749057173728943,
+      "learning_rate": 0.0002,
+      "loss": 0.7175,
+      "step": 15990
+    },
+    {
+      "epoch": 2.5866946891924663,
+      "grad_norm": 0.9386842846870422,
+      "learning_rate": 0.0002,
+      "loss": 0.7212,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5883113733732115,
+      "grad_norm": 0.7868891358375549,
+      "learning_rate": 0.0002,
+      "loss": 0.6934,
+      "step": 16010
+    },
+    {
+      "epoch": 2.5899280575539567,
+      "grad_norm": 0.8674671053886414,
+      "learning_rate": 0.0002,
+      "loss": 0.7036,
+      "step": 16020
+    },
+    {
+      "epoch": 2.591544741734702,
+      "grad_norm": 0.7043559551239014,
+      "learning_rate": 0.0002,
+      "loss": 0.7217,
+      "step": 16030
+    },
+    {
+      "epoch": 2.593161425915447,
+      "grad_norm": 0.5846083760261536,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 16040
+    },
+    {
+      "epoch": 2.594778110096193,
+      "grad_norm": 0.7323982119560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 16050
+    },
+    {
+      "epoch": 2.596394794276938,
+      "grad_norm": 0.9069556593894958,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 16060
+    },
+    {
+      "epoch": 2.5980114784576833,
+      "grad_norm": 0.7522736191749573,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 16070
+    },
+    {
+      "epoch": 2.5996281626384286,
+      "grad_norm": 0.8149648308753967,
+      "learning_rate": 0.0002,
+      "loss": 0.6477,
+      "step": 16080
+    },
+    {
+      "epoch": 2.601244846819174,
+      "grad_norm": 0.6214233040809631,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 16090
+    },
+    {
+      "epoch": 2.602861530999919,
+      "grad_norm": 0.6803743839263916,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 16100
+    },
+    {
+      "epoch": 2.6044782151806647,
+      "grad_norm": 0.7223997116088867,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 16110
+    },
+    {
+      "epoch": 2.60609489936141,
+      "grad_norm": 0.7324174642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.6867,
+      "step": 16120
+    },
+    {
+      "epoch": 2.607711583542155,
+      "grad_norm": 0.9594739675521851,
+      "learning_rate": 0.0002,
+      "loss": 0.7159,
+      "step": 16130
+    },
+    {
+      "epoch": 2.6093282677229004,
+      "grad_norm": 0.9485327005386353,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 16140
+    },
+    {
+      "epoch": 2.6109449519036456,
+      "grad_norm": 0.8449000120162964,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 16150
+    },
+    {
+      "epoch": 2.612561636084391,
+      "grad_norm": 0.8520140051841736,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 16160
+    },
+    {
+      "epoch": 2.614178320265136,
+      "grad_norm": 0.7456524968147278,
+      "learning_rate": 0.0002,
+      "loss": 0.6759,
+      "step": 16170
+    },
+    {
+      "epoch": 2.6157950044458813,
+      "grad_norm": 0.9912857413291931,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 16180
+    },
+    {
+      "epoch": 2.6174116886266265,
+      "grad_norm": 0.9001946449279785,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 16190
+    },
+    {
+      "epoch": 2.619028372807372,
+      "grad_norm": 0.6568667888641357,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16200
+    },
+    {
+      "epoch": 2.6206450569881174,
+      "grad_norm": 1.0248128175735474,
+      "learning_rate": 0.0002,
+      "loss": 0.7013,
+      "step": 16210
+    },
+    {
+      "epoch": 2.6222617411688627,
+      "grad_norm": 0.6509039998054504,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 16220
+    },
+    {
+      "epoch": 2.623878425349608,
+      "grad_norm": 0.7626351118087769,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 16230
+    },
+    {
+      "epoch": 2.625495109530353,
+      "grad_norm": 0.6938552260398865,
+      "learning_rate": 0.0002,
+      "loss": 0.6556,
+      "step": 16240
+    },
+    {
+      "epoch": 2.6271117937110984,
+      "grad_norm": 0.6434680819511414,
+      "learning_rate": 0.0002,
+      "loss": 0.65,
+      "step": 16250
+    },
+    {
+      "epoch": 2.628728477891844,
+      "grad_norm": 0.7111515998840332,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 16260
+    },
+    {
+      "epoch": 2.6303451620725893,
+      "grad_norm": 0.7712395787239075,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 16270
+    },
+    {
+      "epoch": 2.6319618462533345,
+      "grad_norm": 0.792209267616272,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 16280
+    },
+    {
+      "epoch": 2.6335785304340797,
+      "grad_norm": 0.6801066398620605,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 16290
+    },
+    {
+      "epoch": 2.635195214614825,
+      "grad_norm": 0.7802573442459106,
+      "learning_rate": 0.0002,
+      "loss": 0.73,
+      "step": 16300
+    },
+    {
+      "epoch": 2.63681189879557,
+      "grad_norm": 0.7742244601249695,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 16310
+    },
+    {
+      "epoch": 2.6384285829763154,
+      "grad_norm": 0.664184033870697,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 16320
+    },
+    {
+      "epoch": 2.6400452671570607,
+      "grad_norm": 0.9242228865623474,
+      "learning_rate": 0.0002,
+      "loss": 0.6442,
+      "step": 16330
+    },
+    {
+      "epoch": 2.641661951337806,
+      "grad_norm": 0.9661325216293335,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 16340
+    },
+    {
+      "epoch": 2.6432786355185516,
+      "grad_norm": 0.837526798248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 16350
+    },
+    {
+      "epoch": 2.644895319699297,
+      "grad_norm": 1.1834373474121094,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 16360
+    },
+    {
+      "epoch": 2.646512003880042,
+      "grad_norm": 0.7467831373214722,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 16370
+    },
+    {
+      "epoch": 2.6481286880607873,
+      "grad_norm": 0.8627146482467651,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 16380
+    },
+    {
+      "epoch": 2.6497453722415325,
+      "grad_norm": 0.790447473526001,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 16390
+    },
+    {
+      "epoch": 2.651362056422278,
+      "grad_norm": 0.8447365164756775,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 16400
+    },
+    {
+      "epoch": 2.6529787406030234,
+      "grad_norm": 0.7831417918205261,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 16410
+    },
+    {
+      "epoch": 2.6545954247837686,
+      "grad_norm": 0.6837952136993408,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 16420
+    },
+    {
+      "epoch": 2.656212108964514,
+      "grad_norm": 0.7031801342964172,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 16430
+    },
+    {
+      "epoch": 2.657828793145259,
+      "grad_norm": 0.8963770866394043,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 16440
+    },
+    {
+      "epoch": 2.6594454773260043,
+      "grad_norm": 0.6852328181266785,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 16450
+    },
+    {
+      "epoch": 2.6610621615067496,
+      "grad_norm": 0.8069294095039368,
+      "learning_rate": 0.0002,
+      "loss": 0.6278,
+      "step": 16460
+    },
+    {
+      "epoch": 2.662678845687495,
+      "grad_norm": 0.7503686547279358,
+      "learning_rate": 0.0002,
+      "loss": 0.6939,
+      "step": 16470
+    },
+    {
+      "epoch": 2.66429552986824,
+      "grad_norm": 0.6430956125259399,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16480
+    },
+    {
+      "epoch": 2.6659122140489853,
+      "grad_norm": 0.7894312739372253,
+      "learning_rate": 0.0002,
+      "loss": 0.6863,
+      "step": 16490
+    },
+    {
+      "epoch": 2.667528898229731,
+      "grad_norm": 0.7277431488037109,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 16500
+    },
+    {
+      "epoch": 2.669145582410476,
+      "grad_norm": 0.6816153526306152,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 16510
+    },
+    {
+      "epoch": 2.6707622665912214,
+      "grad_norm": 0.8145235776901245,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 16520
+    },
+    {
+      "epoch": 2.6723789507719666,
+      "grad_norm": 0.8645890355110168,
+      "learning_rate": 0.0002,
+      "loss": 0.709,
+      "step": 16530
+    },
+    {
+      "epoch": 2.673995634952712,
+      "grad_norm": 0.704393208026886,
+      "learning_rate": 0.0002,
+      "loss": 0.6946,
+      "step": 16540
+    },
+    {
+      "epoch": 2.6756123191334575,
+      "grad_norm": 1.0120846033096313,
+      "learning_rate": 0.0002,
+      "loss": 0.6378,
+      "step": 16550
+    },
+    {
+      "epoch": 2.6772290033142028,
+      "grad_norm": 0.6919328570365906,
+      "learning_rate": 0.0002,
+      "loss": 0.7241,
+      "step": 16560
+    },
+    {
+      "epoch": 2.678845687494948,
+      "grad_norm": 0.6924574971199036,
+      "learning_rate": 0.0002,
+      "loss": 0.7098,
+      "step": 16570
+    },
+    {
+      "epoch": 2.6804623716756932,
+      "grad_norm": 0.9679301381111145,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 16580
+    },
+    {
+      "epoch": 2.6820790558564385,
+      "grad_norm": 0.6810211539268494,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 16590
+    },
+    {
+      "epoch": 2.6836957400371837,
+      "grad_norm": 0.9730555415153503,
+      "learning_rate": 0.0002,
+      "loss": 0.6688,
+      "step": 16600
+    },
+    {
+      "epoch": 2.685312424217929,
+      "grad_norm": 0.7852821350097656,
+      "learning_rate": 0.0002,
+      "loss": 0.7344,
+      "step": 16610
+    },
+    {
+      "epoch": 2.686929108398674,
+      "grad_norm": 0.6059057116508484,
+      "learning_rate": 0.0002,
+      "loss": 0.6401,
+      "step": 16620
+    },
+    {
+      "epoch": 2.6885457925794194,
+      "grad_norm": 0.9395958781242371,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 16630
+    },
+    {
+      "epoch": 2.690162476760165,
+      "grad_norm": 0.7473729848861694,
+      "learning_rate": 0.0002,
+      "loss": 0.7174,
+      "step": 16640
+    },
+    {
+      "epoch": 2.6917791609409103,
+      "grad_norm": 0.765934407711029,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 16650
+    },
+    {
+      "epoch": 2.6933958451216555,
+      "grad_norm": 0.8496677279472351,
+      "learning_rate": 0.0002,
+      "loss": 0.707,
+      "step": 16660
+    },
+    {
+      "epoch": 2.6950125293024008,
+      "grad_norm": 0.7641879916191101,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 16670
+    },
+    {
+      "epoch": 2.696629213483146,
+      "grad_norm": 0.8471952676773071,
+      "learning_rate": 0.0002,
+      "loss": 0.6566,
+      "step": 16680
+    },
+    {
+      "epoch": 2.6982458976638912,
+      "grad_norm": 0.6946060657501221,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 16690
+    },
+    {
+      "epoch": 2.699862581844637,
+      "grad_norm": 0.7361312508583069,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 16700
+    },
+    {
+      "epoch": 2.701479266025382,
+      "grad_norm": 0.6605038046836853,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 16710
+    },
+    {
+      "epoch": 2.7030959502061274,
+      "grad_norm": 0.7164411544799805,
+      "learning_rate": 0.0002,
+      "loss": 0.6885,
+      "step": 16720
+    },
+    {
+      "epoch": 2.7047126343868726,
+      "grad_norm": 0.6496201157569885,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 16730
+    },
+    {
+      "epoch": 2.706329318567618,
+      "grad_norm": 0.7826663851737976,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 16740
+    },
+    {
+      "epoch": 2.707946002748363,
+      "grad_norm": 0.7639131546020508,
+      "learning_rate": 0.0002,
+      "loss": 0.6773,
+      "step": 16750
+    },
+    {
+      "epoch": 2.7095626869291083,
+      "grad_norm": 0.7976210713386536,
+      "learning_rate": 0.0002,
+      "loss": 0.69,
+      "step": 16760
+    },
+    {
+      "epoch": 2.7111793711098535,
+      "grad_norm": 0.6836577653884888,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 16770
+    },
+    {
+      "epoch": 2.7127960552905988,
+      "grad_norm": 0.8025202751159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 16780
+    },
+    {
+      "epoch": 2.7144127394713444,
+      "grad_norm": 0.7636463642120361,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7160294236520897,
+      "grad_norm": 0.7481677532196045,
+      "learning_rate": 0.0002,
+      "loss": 0.6227,
+      "step": 16800
+    },
+    {
+      "epoch": 2.717646107832835,
+      "grad_norm": 0.7566834688186646,
+      "learning_rate": 0.0002,
+      "loss": 0.6925,
+      "step": 16810
+    },
+    {
+      "epoch": 2.71926279201358,
+      "grad_norm": 0.7931267619132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6531,
+      "step": 16820
+    },
+    {
+      "epoch": 2.7208794761943254,
+      "grad_norm": 0.8811662197113037,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 16830
+    },
+    {
+      "epoch": 2.7224961603750706,
+      "grad_norm": 0.8561240434646606,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 16840
+    },
+    {
+      "epoch": 2.7241128445558163,
+      "grad_norm": 0.7121599316596985,
+      "learning_rate": 0.0002,
+      "loss": 0.7135,
+      "step": 16850
+    },
+    {
+      "epoch": 2.7257295287365615,
+      "grad_norm": 0.8066257238388062,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16860
+    },
+    {
+      "epoch": 2.7273462129173067,
+      "grad_norm": 0.7699271440505981,
+      "learning_rate": 0.0002,
+      "loss": 0.6839,
+      "step": 16870
+    },
+    {
+      "epoch": 2.728962897098052,
+      "grad_norm": 1.1828432083129883,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 16880
+    },
+    {
+      "epoch": 2.730579581278797,
+      "grad_norm": 0.9989302754402161,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 16890
+    },
+    {
+      "epoch": 2.7321962654595424,
+      "grad_norm": 0.8100560307502747,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 16900
+    },
+    {
+      "epoch": 2.7338129496402876,
+      "grad_norm": 0.8615233898162842,
+      "learning_rate": 0.0002,
+      "loss": 0.6851,
+      "step": 16910
+    },
+    {
+      "epoch": 2.735429633821033,
+      "grad_norm": 0.8633756041526794,
+      "learning_rate": 0.0002,
+      "loss": 0.6322,
+      "step": 16920
+    },
+    {
+      "epoch": 2.737046318001778,
+      "grad_norm": 0.7769348621368408,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 16930
+    },
+    {
+      "epoch": 2.738663002182524,
+      "grad_norm": 0.6943058371543884,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 16940
+    },
+    {
+      "epoch": 2.740279686363269,
+      "grad_norm": 0.8510736227035522,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 16950
+    },
+    {
+      "epoch": 2.7418963705440142,
+      "grad_norm": 0.7732602953910828,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 16960
+    },
+    {
+      "epoch": 2.7435130547247595,
+      "grad_norm": 0.5981788635253906,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 16970
+    },
+    {
+      "epoch": 2.7451297389055047,
+      "grad_norm": 0.7604416012763977,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16980
+    },
+    {
+      "epoch": 2.74674642308625,
+      "grad_norm": 0.7377738356590271,
+      "learning_rate": 0.0002,
+      "loss": 0.6981,
+      "step": 16990
+    },
+    {
+      "epoch": 2.7483631072669956,
+      "grad_norm": 0.9400289058685303,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 17000
+    },
+    {
+      "epoch": 2.749979791447741,
+      "grad_norm": 0.6340599656105042,
+      "learning_rate": 0.0002,
+      "loss": 0.6952,
+      "step": 17010
+    },
+    {
+      "epoch": 2.751596475628486,
+      "grad_norm": 0.7297601103782654,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 17020
+    },
+    {
+      "epoch": 2.7532131598092313,
+      "grad_norm": 0.9479979872703552,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 17030
+    },
+    {
+      "epoch": 2.7548298439899765,
+      "grad_norm": 0.8461511135101318,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 17040
+    },
+    {
+      "epoch": 2.7564465281707218,
+      "grad_norm": 0.7477551698684692,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17050
+    },
+    {
+      "epoch": 2.758063212351467,
+      "grad_norm": 1.019270420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 17060
+    },
+    {
+      "epoch": 2.7596798965322122,
+      "grad_norm": 0.7730235457420349,
+      "learning_rate": 0.0002,
+      "loss": 0.7119,
+      "step": 17070
+    },
+    {
+      "epoch": 2.7612965807129575,
+      "grad_norm": 0.8216866254806519,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 17080
+    },
+    {
+      "epoch": 2.762913264893703,
+      "grad_norm": 0.7235931754112244,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17090
+    },
+    {
+      "epoch": 2.7645299490744484,
+      "grad_norm": 0.7352296710014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7031,
+      "step": 17100
+    },
+    {
+      "epoch": 2.7661466332551936,
+      "grad_norm": 0.8129373788833618,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 17110
+    },
+    {
+      "epoch": 2.767763317435939,
+      "grad_norm": 0.7387019991874695,
+      "learning_rate": 0.0002,
+      "loss": 0.6703,
+      "step": 17120
+    },
+    {
+      "epoch": 2.769380001616684,
+      "grad_norm": 0.9149190187454224,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 17130
+    },
+    {
+      "epoch": 2.7709966857974297,
+      "grad_norm": 0.7352971434593201,
+      "learning_rate": 0.0002,
+      "loss": 0.6038,
+      "step": 17140
+    },
+    {
+      "epoch": 2.772613369978175,
+      "grad_norm": 0.7903780341148376,
+      "learning_rate": 0.0002,
+      "loss": 0.6728,
+      "step": 17150
+    },
+    {
+      "epoch": 2.77423005415892,
+      "grad_norm": 0.8255927562713623,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17160
+    },
+    {
+      "epoch": 2.7758467383396654,
+      "grad_norm": 0.7235927581787109,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 17170
+    },
+    {
+      "epoch": 2.7774634225204107,
+      "grad_norm": 0.8281434774398804,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 17180
+    },
+    {
+      "epoch": 2.779080106701156,
+      "grad_norm": 0.7586921453475952,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 17190
+    },
+    {
+      "epoch": 2.780696790881901,
+      "grad_norm": 0.7161715030670166,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 17200
+    },
+    {
+      "epoch": 2.7823134750626464,
+      "grad_norm": 0.762868344783783,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 17210
+    },
+    {
+      "epoch": 2.7839301592433916,
+      "grad_norm": 0.9285483360290527,
+      "learning_rate": 0.0002,
+      "loss": 0.705,
+      "step": 17220
+    },
+    {
+      "epoch": 2.785546843424137,
+      "grad_norm": 0.6900462508201599,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 17230
+    },
+    {
+      "epoch": 2.7871635276048825,
+      "grad_norm": 0.780384361743927,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17240
+    },
+    {
+      "epoch": 2.7887802117856277,
+      "grad_norm": 0.7580406665802002,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17250
+    },
+    {
+      "epoch": 2.790396895966373,
+      "grad_norm": 0.8145199418067932,
+      "learning_rate": 0.0002,
+      "loss": 0.6833,
+      "step": 17260
+    },
+    {
+      "epoch": 2.792013580147118,
+      "grad_norm": 0.9159596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 17270
+    },
+    {
+      "epoch": 2.7936302643278634,
+      "grad_norm": 0.9590014219284058,
+      "learning_rate": 0.0002,
+      "loss": 0.6008,
+      "step": 17280
+    },
+    {
+      "epoch": 2.795246948508609,
+      "grad_norm": 0.7603529691696167,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 17290
+    },
+    {
+      "epoch": 2.7968636326893543,
+      "grad_norm": 0.8039976358413696,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 17300
+    },
+    {
+      "epoch": 2.7984803168700996,
+      "grad_norm": 0.8364847302436829,
+      "learning_rate": 0.0002,
+      "loss": 0.7037,
+      "step": 17310
+    },
+    {
+      "epoch": 2.800097001050845,
+      "grad_norm": 0.8763046860694885,
+      "learning_rate": 0.0002,
+      "loss": 0.6749,
+      "step": 17320
+    },
+    {
+      "epoch": 2.80171368523159,
+      "grad_norm": 0.8409647941589355,
+      "learning_rate": 0.0002,
+      "loss": 0.6844,
+      "step": 17330
+    },
+    {
+      "epoch": 2.8033303694123353,
+      "grad_norm": 0.7649006247520447,
+      "learning_rate": 0.0002,
+      "loss": 0.6936,
+      "step": 17340
+    },
+    {
+      "epoch": 2.8049470535930805,
+      "grad_norm": 0.7970262169837952,
+      "learning_rate": 0.0002,
+      "loss": 0.7051,
+      "step": 17350
+    },
+    {
+      "epoch": 2.8065637377738257,
+      "grad_norm": 0.9088607430458069,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 17360
+    },
+    {
+      "epoch": 2.808180421954571,
+      "grad_norm": 0.6454846858978271,
+      "learning_rate": 0.0002,
+      "loss": 0.675,
+      "step": 17370
+    },
+    {
+      "epoch": 2.809797106135316,
+      "grad_norm": 0.7744787931442261,
+      "learning_rate": 0.0002,
+      "loss": 0.7069,
+      "step": 17380
+    },
+    {
+      "epoch": 2.811413790316062,
+      "grad_norm": 0.6678640842437744,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 17390
+    },
+    {
+      "epoch": 2.813030474496807,
+      "grad_norm": 0.772676944732666,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 17400
+    },
+    {
+      "epoch": 2.8146471586775523,
+      "grad_norm": 0.7088175415992737,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 17410
+    },
+    {
+      "epoch": 2.8162638428582976,
+      "grad_norm": 0.8280573487281799,
+      "learning_rate": 0.0002,
+      "loss": 0.7086,
+      "step": 17420
+    },
+    {
+      "epoch": 2.817880527039043,
+      "grad_norm": 0.6665388345718384,
+      "learning_rate": 0.0002,
+      "loss": 0.6732,
+      "step": 17430
+    },
+    {
+      "epoch": 2.8194972112197885,
+      "grad_norm": 0.6427883505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 17440
+    },
+    {
+      "epoch": 2.8211138954005337,
+      "grad_norm": 0.9697760343551636,
+      "learning_rate": 0.0002,
+      "loss": 0.6972,
+      "step": 17450
+    },
+    {
+      "epoch": 2.822730579581279,
+      "grad_norm": 0.7573966383934021,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 17460
+    },
+    {
+      "epoch": 2.824347263762024,
+      "grad_norm": 0.878688633441925,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 17470
+    },
+    {
+      "epoch": 2.8259639479427694,
+      "grad_norm": 0.7752242684364319,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 17480
+    },
+    {
+      "epoch": 2.8275806321235146,
+      "grad_norm": 0.6135398745536804,
+      "learning_rate": 0.0002,
+      "loss": 0.6638,
+      "step": 17490
+    },
+    {
+      "epoch": 2.82919731630426,
+      "grad_norm": 0.6924924850463867,
+      "learning_rate": 0.0002,
+      "loss": 0.6829,
+      "step": 17500
+    },
+    {
+      "epoch": 2.830814000485005,
+      "grad_norm": 0.7471627593040466,
+      "learning_rate": 0.0002,
+      "loss": 0.6731,
+      "step": 17510
+    },
+    {
+      "epoch": 2.8324306846657503,
+      "grad_norm": 0.7145499587059021,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 17520
+    },
+    {
+      "epoch": 2.834047368846496,
+      "grad_norm": 0.7415414452552795,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17530
+    },
+    {
+      "epoch": 2.8356640530272412,
+      "grad_norm": 0.7328441739082336,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17540
+    },
+    {
+      "epoch": 2.8372807372079865,
+      "grad_norm": 0.8267839550971985,
+      "learning_rate": 0.0002,
+      "loss": 0.6866,
+      "step": 17550
+    },
+    {
+      "epoch": 2.8388974213887317,
+      "grad_norm": 0.8877885341644287,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17560
+    },
+    {
+      "epoch": 2.840514105569477,
+      "grad_norm": 0.857138454914093,
+      "learning_rate": 0.0002,
+      "loss": 0.7136,
+      "step": 17570
+    },
+    {
+      "epoch": 2.842130789750222,
+      "grad_norm": 0.8470779657363892,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 17580
+    },
+    {
+      "epoch": 2.843747473930968,
+      "grad_norm": 0.8553254008293152,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 17590
+    },
+    {
+      "epoch": 2.845364158111713,
+      "grad_norm": 0.8033196926116943,
+      "learning_rate": 0.0002,
+      "loss": 0.7297,
+      "step": 17600
+    },
+    {
+      "epoch": 2.8469808422924583,
+      "grad_norm": 0.7949087023735046,
+      "learning_rate": 0.0002,
+      "loss": 0.7062,
+      "step": 17610
+    },
+    {
+      "epoch": 2.8485975264732035,
+      "grad_norm": 0.9241406321525574,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 17620
+    },
+    {
+      "epoch": 2.8502142106539488,
+      "grad_norm": 0.7721285223960876,
+      "learning_rate": 0.0002,
+      "loss": 0.6601,
+      "step": 17630
+    },
+    {
+      "epoch": 2.851830894834694,
+      "grad_norm": 1.0246692895889282,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 17640
+    },
+    {
+      "epoch": 2.853447579015439,
+      "grad_norm": 0.9244589805603027,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 17650
+    },
+    {
+      "epoch": 2.8550642631961844,
+      "grad_norm": 0.7243508696556091,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 17660
+    },
+    {
+      "epoch": 2.8566809473769297,
+      "grad_norm": 0.8943371176719666,
+      "learning_rate": 0.0002,
+      "loss": 0.6471,
+      "step": 17670
+    },
+    {
+      "epoch": 2.8582976315576754,
+      "grad_norm": 0.6531758904457092,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 17680
+    },
+    {
+      "epoch": 2.8599143157384206,
+      "grad_norm": 0.8367000818252563,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 17690
+    },
+    {
+      "epoch": 2.861530999919166,
+      "grad_norm": 0.7868556380271912,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 17700
+    },
+    {
+      "epoch": 2.863147684099911,
+      "grad_norm": 0.7213859558105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 17710
+    },
+    {
+      "epoch": 2.8647643682806563,
+      "grad_norm": 0.7383931279182434,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 17720
+    },
+    {
+      "epoch": 2.8663810524614015,
+      "grad_norm": 0.7566812634468079,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 17730
+    },
+    {
+      "epoch": 2.867997736642147,
+      "grad_norm": 0.6930373311042786,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 17740
+    },
+    {
+      "epoch": 2.8696144208228924,
+      "grad_norm": 0.7911090850830078,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17750
+    },
+    {
+      "epoch": 2.8712311050036377,
+      "grad_norm": 0.8484548926353455,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 17760
+    },
+    {
+      "epoch": 2.872847789184383,
+      "grad_norm": 0.7647597193717957,
+      "learning_rate": 0.0002,
+      "loss": 0.6931,
+      "step": 17770
+    },
+    {
+      "epoch": 2.874464473365128,
+      "grad_norm": 0.8791151642799377,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 17780
+    },
+    {
+      "epoch": 2.8760811575458733,
+      "grad_norm": 0.7253178358078003,
+      "learning_rate": 0.0002,
+      "loss": 0.7078,
+      "step": 17790
+    },
+    {
+      "epoch": 2.8776978417266186,
+      "grad_norm": 0.7956077456474304,
+      "learning_rate": 0.0002,
+      "loss": 0.6474,
+      "step": 17800
+    },
+    {
+      "epoch": 2.879314525907364,
+      "grad_norm": 0.8657688498497009,
+      "learning_rate": 0.0002,
+      "loss": 0.6687,
+      "step": 17810
+    },
+    {
+      "epoch": 2.880931210088109,
+      "grad_norm": 0.7059141993522644,
+      "learning_rate": 0.0002,
+      "loss": 0.7171,
+      "step": 17820
+    },
+    {
+      "epoch": 2.8825478942688547,
+      "grad_norm": 0.8886896967887878,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 17830
+    },
+    {
+      "epoch": 2.8841645784496,
+      "grad_norm": 0.821032702922821,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 17840
+    },
+    {
+      "epoch": 2.885781262630345,
+      "grad_norm": 0.7183963656425476,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 17850
+    },
+    {
+      "epoch": 2.8873979468110904,
+      "grad_norm": 0.6222899556159973,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 17860
+    },
+    {
+      "epoch": 2.8890146309918356,
+      "grad_norm": 0.8187434077262878,
+      "learning_rate": 0.0002,
+      "loss": 0.6626,
+      "step": 17870
+    },
+    {
+      "epoch": 2.890631315172581,
+      "grad_norm": 0.9838479161262512,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 17880
+    },
+    {
+      "epoch": 2.8922479993533265,
+      "grad_norm": 0.7567742466926575,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 17890
+    },
+    {
+      "epoch": 2.893864683534072,
+      "grad_norm": 0.6875903606414795,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17900
+    },
+    {
+      "epoch": 2.895481367714817,
+      "grad_norm": 0.8043789267539978,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 17910
+    },
+    {
+      "epoch": 2.8970980518955622,
+      "grad_norm": 0.8062626719474792,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 17920
+    },
+    {
+      "epoch": 2.8987147360763075,
+      "grad_norm": 1.0251191854476929,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 17930
+    },
+    {
+      "epoch": 2.9003314202570527,
+      "grad_norm": 0.882253110408783,
+      "learning_rate": 0.0002,
+      "loss": 0.6605,
+      "step": 17940
+    },
+    {
+      "epoch": 2.901948104437798,
+      "grad_norm": 0.8683299422264099,
+      "learning_rate": 0.0002,
+      "loss": 0.6719,
+      "step": 17950
+    },
+    {
+      "epoch": 2.903564788618543,
+      "grad_norm": 0.7167282104492188,
+      "learning_rate": 0.0002,
+      "loss": 0.6896,
+      "step": 17960
+    },
+    {
+      "epoch": 2.9051814727992884,
+      "grad_norm": 0.7093694806098938,
+      "learning_rate": 0.0002,
+      "loss": 0.663,
+      "step": 17970
+    },
+    {
+      "epoch": 2.906798156980034,
+      "grad_norm": 0.8549879193305969,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 17980
+    },
+    {
+      "epoch": 2.9084148411607793,
+      "grad_norm": 0.6989606618881226,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 17990
+    },
+    {
+      "epoch": 2.9100315253415245,
+      "grad_norm": 0.9482976794242859,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 18000
+    },
+    {
+      "epoch": 2.9116482095222698,
+      "grad_norm": 0.7182440161705017,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 18010
+    },
+    {
+      "epoch": 2.913264893703015,
+      "grad_norm": 0.7732226252555847,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 18020
+    },
+    {
+      "epoch": 2.9148815778837607,
+      "grad_norm": 0.7936875224113464,
+      "learning_rate": 0.0002,
+      "loss": 0.7123,
+      "step": 18030
+    },
+    {
+      "epoch": 2.916498262064506,
+      "grad_norm": 0.8825615644454956,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 18040
+    },
+    {
+      "epoch": 2.918114946245251,
+      "grad_norm": 0.6778587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 18050
+    },
+    {
+      "epoch": 2.9197316304259964,
+      "grad_norm": 0.7529265880584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 18060
+    },
+    {
+      "epoch": 2.9213483146067416,
+      "grad_norm": 0.7111883163452148,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 18070
+    },
+    {
+      "epoch": 2.922964998787487,
+      "grad_norm": 0.7214767932891846,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 18080
+    },
+    {
+      "epoch": 2.924581682968232,
+      "grad_norm": 0.800417423248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 18090
+    },
+    {
+      "epoch": 2.9261983671489773,
+      "grad_norm": 1.248575210571289,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 18100
+    },
+    {
+      "epoch": 2.9278150513297225,
+      "grad_norm": 0.757788360118866,
+      "learning_rate": 0.0002,
+      "loss": 0.7004,
+      "step": 18110
+    },
+    {
+      "epoch": 2.9294317355104678,
+      "grad_norm": 1.0583995580673218,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 18120
+    },
+    {
+      "epoch": 2.9310484196912134,
+      "grad_norm": 0.8228777647018433,
+      "learning_rate": 0.0002,
+      "loss": 0.6365,
+      "step": 18130
+    },
+    {
+      "epoch": 2.9326651038719587,
+      "grad_norm": 0.8374035358428955,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 18140
+    },
+    {
+      "epoch": 2.934281788052704,
+      "grad_norm": 0.7976473569869995,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 18150
+    },
+    {
+      "epoch": 2.935898472233449,
+      "grad_norm": 0.8009907603263855,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 18160
+    },
+    {
+      "epoch": 2.9375151564141944,
+      "grad_norm": 0.835213303565979,
+      "learning_rate": 0.0002,
+      "loss": 0.7485,
+      "step": 18170
+    },
+    {
+      "epoch": 2.93913184059494,
+      "grad_norm": 0.7982219457626343,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 18180
+    },
+    {
+      "epoch": 2.9407485247756853,
+      "grad_norm": 0.7070978879928589,
+      "learning_rate": 0.0002,
+      "loss": 0.6348,
+      "step": 18190
+    },
+    {
+      "epoch": 2.9423652089564305,
+      "grad_norm": 0.8619440197944641,
+      "learning_rate": 0.0002,
+      "loss": 0.6608,
+      "step": 18200
+    },
+    {
+      "epoch": 2.9439818931371757,
+      "grad_norm": 0.6693987250328064,
+      "learning_rate": 0.0002,
+      "loss": 0.666,
+      "step": 18210
+    },
+    {
+      "epoch": 2.945598577317921,
+      "grad_norm": 0.6747021079063416,
+      "learning_rate": 0.0002,
+      "loss": 0.728,
+      "step": 18220
+    },
+    {
+      "epoch": 2.947215261498666,
+      "grad_norm": 0.860387921333313,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 18230
+    },
+    {
+      "epoch": 2.9488319456794114,
+      "grad_norm": 0.799976646900177,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 18240
+    },
+    {
+      "epoch": 2.9504486298601567,
+      "grad_norm": 0.7864769101142883,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 18250
+    },
+    {
+      "epoch": 2.952065314040902,
+      "grad_norm": 0.6713884472846985,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 18260
+    },
+    {
+      "epoch": 2.9536819982216476,
+      "grad_norm": 0.9031508564949036,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 18270
+    },
+    {
+      "epoch": 2.955298682402393,
+      "grad_norm": 0.7205073237419128,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 18280
+    },
+    {
+      "epoch": 2.956915366583138,
+      "grad_norm": 0.7746205925941467,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 18290
+    },
+    {
+      "epoch": 2.9585320507638833,
+      "grad_norm": 0.6533427834510803,
+      "learning_rate": 0.0002,
+      "loss": 0.6218,
+      "step": 18300
+    },
+    {
+      "epoch": 2.9601487349446285,
+      "grad_norm": 0.9083208441734314,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 18310
+    },
+    {
+      "epoch": 2.9617654191253737,
+      "grad_norm": 0.7446991801261902,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 18320
+    },
+    {
+      "epoch": 2.9633821033061194,
+      "grad_norm": 0.6514461636543274,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 18330
+    },
+    {
+      "epoch": 2.9649987874868646,
+      "grad_norm": 0.8580465912818909,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18340
+    },
+    {
+      "epoch": 2.96661547166761,
+      "grad_norm": 0.7074266076087952,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 18350
+    },
+    {
+      "epoch": 2.968232155848355,
+      "grad_norm": 0.899892270565033,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 18360
+    },
+    {
+      "epoch": 2.9698488400291003,
+      "grad_norm": 0.8217641711235046,
+      "learning_rate": 0.0002,
+      "loss": 0.7094,
+      "step": 18370
+    },
+    {
+      "epoch": 2.9714655242098456,
+      "grad_norm": 0.8611799478530884,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 18380
+    },
+    {
+      "epoch": 2.973082208390591,
+      "grad_norm": 0.6909302473068237,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18390
+    },
+    {
+      "epoch": 2.974698892571336,
+      "grad_norm": 0.6554358005523682,
+      "learning_rate": 0.0002,
+      "loss": 0.7247,
+      "step": 18400
+    },
+    {
+      "epoch": 2.9763155767520812,
+      "grad_norm": 0.7803071737289429,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 18410
+    },
+    {
+      "epoch": 2.977932260932827,
+      "grad_norm": 0.7838954925537109,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 18420
+    },
+    {
+      "epoch": 2.979548945113572,
+      "grad_norm": 0.7098495364189148,
+      "learning_rate": 0.0002,
+      "loss": 0.6522,
+      "step": 18430
+    },
+    {
+      "epoch": 2.9811656292943174,
+      "grad_norm": 0.8981785774230957,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 18440
+    },
+    {
+      "epoch": 2.9827823134750626,
+      "grad_norm": 0.7197171449661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6689,
+      "step": 18450
+    },
+    {
+      "epoch": 2.984398997655808,
+      "grad_norm": 0.793185293674469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 18460
+    },
+    {
+      "epoch": 2.986015681836553,
+      "grad_norm": 0.8531473875045776,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 18470
+    },
+    {
+      "epoch": 2.9876323660172988,
+      "grad_norm": 0.6627361178398132,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 18480
+    },
+    {
+      "epoch": 2.989249050198044,
+      "grad_norm": 0.5708155035972595,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 18490
+    },
+    {
+      "epoch": 2.990865734378789,
+      "grad_norm": 0.8227280378341675,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 18500
+    },
+    {
+      "epoch": 2.9924824185595345,
+      "grad_norm": 0.7102749943733215,
+      "learning_rate": 0.0002,
+      "loss": 0.6701,
+      "step": 18510
+    },
+    {
+      "epoch": 2.9940991027402797,
+      "grad_norm": 0.839485228061676,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 18520
+    },
+    {
+      "epoch": 2.995715786921025,
+      "grad_norm": 0.9038704037666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 18530
+    },
+    {
+      "epoch": 2.99733247110177,
+      "grad_norm": 0.8737510442733765,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 18540
+    },
+    {
+      "epoch": 2.9989491552825154,
+      "grad_norm": 0.7323142886161804,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 18550
+    },
+    {
+      "epoch": 2.9999191657909625,
+      "eval_loss": 1.1262480020523071,
+      "eval_runtime": 122.0868,
+      "eval_samples_per_second": 6.004,
+      "eval_steps_per_second": 0.754,
+      "step": 18556
+    },
+    {
+      "epoch": 3.000565839463261,
+      "grad_norm": 0.8465463519096375,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 18560
+    },
+    {
+      "epoch": 3.0021825236440063,
+      "grad_norm": 0.9134138822555542,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 18570
+    },
+    {
+      "epoch": 3.0037992078247515,
+      "grad_norm": 0.760715126991272,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 18580
+    },
+    {
+      "epoch": 3.0054158920054967,
+      "grad_norm": 0.9208743572235107,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18590
+    },
+    {
+      "epoch": 3.007032576186242,
+      "grad_norm": 0.9232364892959595,
+      "learning_rate": 0.0002,
+      "loss": 0.5731,
+      "step": 18600
+    },
+    {
+      "epoch": 3.008649260366987,
+      "grad_norm": 1.1881544589996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6299,
+      "step": 18610
+    },
+    {
+      "epoch": 3.0102659445477324,
+      "grad_norm": 0.9372987747192383,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 18620
+    },
+    {
+      "epoch": 3.0118826287284777,
+      "grad_norm": 0.6900241374969482,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 18630
+    },
+    {
+      "epoch": 3.0134993129092233,
+      "grad_norm": 0.8451071381568909,
+      "learning_rate": 0.0002,
+      "loss": 0.5256,
+      "step": 18640
+    },
+    {
+      "epoch": 3.0151159970899686,
+      "grad_norm": 0.7763112187385559,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 18650
+    },
+    {
+      "epoch": 3.016732681270714,
+      "grad_norm": 1.043653964996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 18660
+    },
+    {
+      "epoch": 3.018349365451459,
+      "grad_norm": 1.0170660018920898,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 18670
+    },
+    {
+      "epoch": 3.0199660496322043,
+      "grad_norm": 0.7534180283546448,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 18680
+    },
+    {
+      "epoch": 3.0215827338129495,
+      "grad_norm": 0.7507367730140686,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 18690
+    },
+    {
+      "epoch": 3.0231994179936947,
+      "grad_norm": 0.7861620187759399,
+      "learning_rate": 0.0002,
+      "loss": 0.6201,
+      "step": 18700
+    },
+    {
+      "epoch": 3.0248161021744404,
+      "grad_norm": 1.0580339431762695,
+      "learning_rate": 0.0002,
+      "loss": 0.5802,
+      "step": 18710
+    },
+    {
+      "epoch": 3.0264327863551856,
+      "grad_norm": 0.7542710900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 18720
+    },
+    {
+      "epoch": 3.028049470535931,
+      "grad_norm": 0.8189544677734375,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 18730
+    },
+    {
+      "epoch": 3.029666154716676,
+      "grad_norm": 0.9126611351966858,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 18740
+    },
+    {
+      "epoch": 3.0312828388974213,
+      "grad_norm": 0.8891341686248779,
+      "learning_rate": 0.0002,
+      "loss": 0.6443,
+      "step": 18750
+    },
+    {
+      "epoch": 3.0328995230781666,
+      "grad_norm": 0.8419283032417297,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 18760
+    },
+    {
+      "epoch": 3.034516207258912,
+      "grad_norm": 0.8048048615455627,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18770
+    },
+    {
+      "epoch": 3.0361328914396575,
+      "grad_norm": 0.7820217609405518,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 18780
+    },
+    {
+      "epoch": 3.0377495756204027,
+      "grad_norm": 0.854721188545227,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 18790
+    },
+    {
+      "epoch": 3.039366259801148,
+      "grad_norm": 0.912092924118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5784,
+      "step": 18800
+    },
+    {
+      "epoch": 3.040982943981893,
+      "grad_norm": 0.6596226096153259,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 18810
+    },
+    {
+      "epoch": 3.0425996281626384,
+      "grad_norm": 0.6351348757743835,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 18820
+    },
+    {
+      "epoch": 3.0442163123433836,
+      "grad_norm": 0.778188943862915,
+      "learning_rate": 0.0002,
+      "loss": 0.5953,
+      "step": 18830
+    },
+    {
+      "epoch": 3.045832996524129,
+      "grad_norm": 0.68234783411026,
+      "learning_rate": 0.0002,
+      "loss": 0.602,
+      "step": 18840
+    },
+    {
+      "epoch": 3.047449680704874,
+      "grad_norm": 0.998628556728363,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 18850
+    },
+    {
+      "epoch": 3.0490663648856198,
+      "grad_norm": 0.7393841743469238,
+      "learning_rate": 0.0002,
+      "loss": 0.6231,
+      "step": 18860
+    },
+    {
+      "epoch": 3.050683049066365,
+      "grad_norm": 0.84438556432724,
+      "learning_rate": 0.0002,
+      "loss": 0.568,
+      "step": 18870
+    },
+    {
+      "epoch": 3.0522997332471102,
+      "grad_norm": 0.8857501745223999,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 18880
+    },
+    {
+      "epoch": 3.0539164174278555,
+      "grad_norm": 0.7208474278450012,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 18890
+    },
+    {
+      "epoch": 3.0555331016086007,
+      "grad_norm": 0.7135229110717773,
+      "learning_rate": 0.0002,
+      "loss": 0.5998,
+      "step": 18900
+    },
+    {
+      "epoch": 3.057149785789346,
+      "grad_norm": 0.9130001664161682,
+      "learning_rate": 0.0002,
+      "loss": 0.5575,
+      "step": 18910
+    },
+    {
+      "epoch": 3.058766469970091,
+      "grad_norm": 0.9001716375350952,
+      "learning_rate": 0.0002,
+      "loss": 0.5955,
+      "step": 18920
+    },
+    {
+      "epoch": 3.060383154150837,
+      "grad_norm": 0.8667559623718262,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 18930
+    },
+    {
+      "epoch": 3.061999838331582,
+      "grad_norm": 0.8943959474563599,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18940
+    },
+    {
+      "epoch": 3.0636165225123273,
+      "grad_norm": 0.8298377990722656,
+      "learning_rate": 0.0002,
+      "loss": 0.5978,
+      "step": 18950
+    },
+    {
+      "epoch": 3.0652332066930725,
+      "grad_norm": 0.7935267686843872,
+      "learning_rate": 0.0002,
+      "loss": 0.5782,
+      "step": 18960
+    },
+    {
+      "epoch": 3.0668498908738178,
+      "grad_norm": 1.1506379842758179,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 18970
+    },
+    {
+      "epoch": 3.068466575054563,
+      "grad_norm": 0.7693049907684326,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18980
+    },
+    {
+      "epoch": 3.0700832592353082,
+      "grad_norm": 0.8040135502815247,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 18990
+    },
+    {
+      "epoch": 3.0716999434160535,
+      "grad_norm": 0.828404426574707,
+      "learning_rate": 0.0002,
+      "loss": 0.5541,
+      "step": 19000
+    },
+    {
+      "epoch": 3.073316627596799,
+      "grad_norm": 0.8811164498329163,
+      "learning_rate": 0.0002,
+      "loss": 0.6048,
+      "step": 19010
+    },
+    {
+      "epoch": 3.0749333117775444,
+      "grad_norm": 1.036205768585205,
+      "learning_rate": 0.0002,
+      "loss": 0.5845,
+      "step": 19020
+    },
+    {
+      "epoch": 3.0765499959582896,
+      "grad_norm": 0.8857285976409912,
+      "learning_rate": 0.0002,
+      "loss": 0.5838,
+      "step": 19030
+    },
+    {
+      "epoch": 3.078166680139035,
+      "grad_norm": 0.8392079472541809,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19040
+    },
+    {
+      "epoch": 3.07978336431978,
+      "grad_norm": 1.0287401676177979,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 19050
+    },
+    {
+      "epoch": 3.0814000485005253,
+      "grad_norm": 1.0086315870285034,
+      "learning_rate": 0.0002,
+      "loss": 0.5964,
+      "step": 19060
+    },
+    {
+      "epoch": 3.0830167326812705,
+      "grad_norm": 0.9245324730873108,
+      "learning_rate": 0.0002,
+      "loss": 0.5567,
+      "step": 19070
+    },
+    {
+      "epoch": 3.084633416862016,
+      "grad_norm": 0.8680877089500427,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 19080
+    },
+    {
+      "epoch": 3.0862501010427614,
+      "grad_norm": 0.8814793825149536,
+      "learning_rate": 0.0002,
+      "loss": 0.5611,
+      "step": 19090
+    },
+    {
+      "epoch": 3.0878667852235067,
+      "grad_norm": 0.9234458208084106,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19100
+    },
+    {
+      "epoch": 3.089483469404252,
+      "grad_norm": 1.1291664838790894,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 19110
+    },
+    {
+      "epoch": 3.091100153584997,
+      "grad_norm": 0.9191402792930603,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 19120
+    },
+    {
+      "epoch": 3.0927168377657424,
+      "grad_norm": 0.7103154063224792,
+      "learning_rate": 0.0002,
+      "loss": 0.5856,
+      "step": 19130
+    },
+    {
+      "epoch": 3.0943335219464876,
+      "grad_norm": 0.9368883967399597,
+      "learning_rate": 0.0002,
+      "loss": 0.6479,
+      "step": 19140
+    },
+    {
+      "epoch": 3.095950206127233,
+      "grad_norm": 0.9676656723022461,
+      "learning_rate": 0.0002,
+      "loss": 0.6167,
+      "step": 19150
+    },
+    {
+      "epoch": 3.0975668903079785,
+      "grad_norm": 0.8739792704582214,
+      "learning_rate": 0.0002,
+      "loss": 0.5794,
+      "step": 19160
+    },
+    {
+      "epoch": 3.0991835744887237,
+      "grad_norm": 0.8530174493789673,
+      "learning_rate": 0.0002,
+      "loss": 0.6112,
+      "step": 19170
+    },
+    {
+      "epoch": 3.100800258669469,
+      "grad_norm": 0.794945478439331,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 19180
+    },
+    {
+      "epoch": 3.102416942850214,
+      "grad_norm": 0.9508888125419617,
+      "learning_rate": 0.0002,
+      "loss": 0.5928,
+      "step": 19190
+    },
+    {
+      "epoch": 3.1040336270309594,
+      "grad_norm": 1.0599955320358276,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 19200
+    },
+    {
+      "epoch": 3.1056503112117047,
+      "grad_norm": 1.0673625469207764,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 19210
+    },
+    {
+      "epoch": 3.10726699539245,
+      "grad_norm": 0.7739115953445435,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 19220
+    },
+    {
+      "epoch": 3.1088836795731956,
+      "grad_norm": 0.9884951114654541,
+      "learning_rate": 0.0002,
+      "loss": 0.6046,
+      "step": 19230
+    },
+    {
+      "epoch": 3.110500363753941,
+      "grad_norm": 0.862260103225708,
+      "learning_rate": 0.0002,
+      "loss": 0.5932,
+      "step": 19240
+    },
+    {
+      "epoch": 3.112117047934686,
+      "grad_norm": 0.7690284848213196,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 19250
+    },
+    {
+      "epoch": 3.1137337321154313,
+      "grad_norm": 0.8758958578109741,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 19260
+    },
+    {
+      "epoch": 3.1153504162961765,
+      "grad_norm": 1.0356395244598389,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 19270
+    },
+    {
+      "epoch": 3.1169671004769217,
+      "grad_norm": 0.6950937509536743,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 19280
+    },
+    {
+      "epoch": 3.118583784657667,
+      "grad_norm": 0.760998010635376,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19290
+    },
+    {
+      "epoch": 3.1202004688384126,
+      "grad_norm": 0.9335789084434509,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 19300
+    },
+    {
+      "epoch": 3.121817153019158,
+      "grad_norm": 0.9636204242706299,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 19310
+    },
+    {
+      "epoch": 3.123433837199903,
+      "grad_norm": 1.0820997953414917,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 19320
+    },
+    {
+      "epoch": 3.1250505213806483,
+      "grad_norm": 0.7333487272262573,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 19330
+    },
+    {
+      "epoch": 3.1266672055613935,
+      "grad_norm": 1.0417509078979492,
+      "learning_rate": 0.0002,
+      "loss": 0.6178,
+      "step": 19340
+    },
+    {
+      "epoch": 3.128283889742139,
+      "grad_norm": 0.9267749190330505,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 19350
+    },
+    {
+      "epoch": 3.129900573922884,
+      "grad_norm": 0.777798593044281,
+      "learning_rate": 0.0002,
+      "loss": 0.6063,
+      "step": 19360
+    },
+    {
+      "epoch": 3.1315172581036297,
+      "grad_norm": 0.8425456881523132,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 19370
+    },
+    {
+      "epoch": 3.133133942284375,
+      "grad_norm": 0.9617102146148682,
+      "learning_rate": 0.0002,
+      "loss": 0.6042,
+      "step": 19380
+    },
+    {
+      "epoch": 3.13475062646512,
+      "grad_norm": 1.0052828788757324,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 19390
+    },
+    {
+      "epoch": 3.1363673106458654,
+      "grad_norm": 0.7637009024620056,
+      "learning_rate": 0.0002,
+      "loss": 0.5713,
+      "step": 19400
+    },
+    {
+      "epoch": 3.1379839948266106,
+      "grad_norm": 0.7958088517189026,
+      "learning_rate": 0.0002,
+      "loss": 0.5497,
+      "step": 19410
+    },
+    {
+      "epoch": 3.139600679007356,
+      "grad_norm": 0.9161727428436279,
+      "learning_rate": 0.0002,
+      "loss": 0.6283,
+      "step": 19420
+    },
+    {
+      "epoch": 3.141217363188101,
+      "grad_norm": 0.8402149677276611,
+      "learning_rate": 0.0002,
+      "loss": 0.5638,
+      "step": 19430
+    },
+    {
+      "epoch": 3.1428340473688463,
+      "grad_norm": 1.0056525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.5848,
+      "step": 19440
+    },
+    {
+      "epoch": 3.144450731549592,
+      "grad_norm": 1.0129190683364868,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 19450
+    },
+    {
+      "epoch": 3.146067415730337,
+      "grad_norm": 0.790825366973877,
+      "learning_rate": 0.0002,
+      "loss": 0.5808,
+      "step": 19460
+    },
+    {
+      "epoch": 3.1476840999110824,
+      "grad_norm": 1.441665530204773,
+      "learning_rate": 0.0002,
+      "loss": 0.5607,
+      "step": 19470
+    },
+    {
+      "epoch": 3.1493007840918277,
+      "grad_norm": 0.7846331596374512,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19480
+    },
+    {
+      "epoch": 3.150917468272573,
+      "grad_norm": 0.7915332913398743,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 19490
+    },
+    {
+      "epoch": 3.152534152453318,
+      "grad_norm": 0.933982253074646,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 19500
+    },
+    {
+      "epoch": 3.1541508366340634,
+      "grad_norm": 1.038408637046814,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 19510
+    },
+    {
+      "epoch": 3.155767520814809,
+      "grad_norm": 1.018935203552246,
+      "learning_rate": 0.0002,
+      "loss": 0.6271,
+      "step": 19520
+    },
+    {
+      "epoch": 3.1573842049955543,
+      "grad_norm": 0.9618112444877625,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 19530
+    },
+    {
+      "epoch": 3.1590008891762995,
+      "grad_norm": 0.8900452852249146,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 19540
+    },
+    {
+      "epoch": 3.1606175733570447,
+      "grad_norm": 0.8254160284996033,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 19550
+    },
+    {
+      "epoch": 3.16223425753779,
+      "grad_norm": 1.004376769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19560
+    },
+    {
+      "epoch": 3.163850941718535,
+      "grad_norm": 1.0490446090698242,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 19570
+    },
+    {
+      "epoch": 3.1654676258992804,
+      "grad_norm": 0.7387403845787048,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19580
+    },
+    {
+      "epoch": 3.1670843100800257,
+      "grad_norm": 0.7611538171768188,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 19590
+    },
+    {
+      "epoch": 3.1687009942607713,
+      "grad_norm": 0.8239886164665222,
+      "learning_rate": 0.0002,
+      "loss": 0.5937,
+      "step": 19600
+    },
+    {
+      "epoch": 3.1703176784415166,
+      "grad_norm": 0.9327243566513062,
+      "learning_rate": 0.0002,
+      "loss": 0.6068,
+      "step": 19610
+    },
+    {
+      "epoch": 3.171934362622262,
+      "grad_norm": 0.9662560224533081,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 19620
+    },
+    {
+      "epoch": 3.173551046803007,
+      "grad_norm": 0.9183341860771179,
+      "learning_rate": 0.0002,
+      "loss": 0.5988,
+      "step": 19630
+    },
+    {
+      "epoch": 3.1751677309837523,
+      "grad_norm": 0.875066876411438,
+      "learning_rate": 0.0002,
+      "loss": 0.5909,
+      "step": 19640
+    },
+    {
+      "epoch": 3.1767844151644975,
+      "grad_norm": 0.8567508459091187,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 19650
+    },
+    {
+      "epoch": 3.1784010993452427,
+      "grad_norm": 0.6805780529975891,
+      "learning_rate": 0.0002,
+      "loss": 0.5805,
+      "step": 19660
+    },
+    {
+      "epoch": 3.1800177835259884,
+      "grad_norm": 0.8776944279670715,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 19670
+    },
+    {
+      "epoch": 3.1816344677067336,
+      "grad_norm": 0.9036329984664917,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 19680
+    },
+    {
+      "epoch": 3.183251151887479,
+      "grad_norm": 0.8527372479438782,
+      "learning_rate": 0.0002,
+      "loss": 0.6238,
+      "step": 19690
+    },
+    {
+      "epoch": 3.184867836068224,
+      "grad_norm": 1.1045585870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 19700
+    },
+    {
+      "epoch": 3.1864845202489693,
+      "grad_norm": 0.9213830828666687,
+      "learning_rate": 0.0002,
+      "loss": 0.5491,
+      "step": 19710
+    },
+    {
+      "epoch": 3.1881012044297146,
+      "grad_norm": 0.8865814805030823,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 19720
+    },
+    {
+      "epoch": 3.18971788861046,
+      "grad_norm": 0.7939388751983643,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19730
+    },
+    {
+      "epoch": 3.191334572791205,
+      "grad_norm": 0.6966729760169983,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 19740
+    },
+    {
+      "epoch": 3.1929512569719507,
+      "grad_norm": 0.8023673295974731,
+      "learning_rate": 0.0002,
+      "loss": 0.5839,
+      "step": 19750
+    },
+    {
+      "epoch": 3.194567941152696,
+      "grad_norm": 0.7992037534713745,
+      "learning_rate": 0.0002,
+      "loss": 0.6267,
+      "step": 19760
+    },
+    {
+      "epoch": 3.196184625333441,
+      "grad_norm": 0.7412247657775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 19770
+    },
+    {
+      "epoch": 3.1978013095141864,
+      "grad_norm": 0.9598729014396667,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 19780
+    },
+    {
+      "epoch": 3.1994179936949316,
+      "grad_norm": 0.8331366777420044,
+      "learning_rate": 0.0002,
+      "loss": 0.5685,
+      "step": 19790
+    },
+    {
+      "epoch": 3.201034677875677,
+      "grad_norm": 0.8939169645309448,
+      "learning_rate": 0.0002,
+      "loss": 0.6104,
+      "step": 19800
+    },
+    {
+      "epoch": 3.202651362056422,
+      "grad_norm": 0.9219734072685242,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 19810
+    },
+    {
+      "epoch": 3.2042680462371678,
+      "grad_norm": 0.869490385055542,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19820
+    },
+    {
+      "epoch": 3.205884730417913,
+      "grad_norm": 0.8989706635475159,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 19830
+    },
+    {
+      "epoch": 3.2075014145986582,
+      "grad_norm": 0.8477165102958679,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 19840
+    },
+    {
+      "epoch": 3.2091180987794035,
+      "grad_norm": 0.8720678687095642,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 19850
+    },
+    {
+      "epoch": 3.2107347829601487,
+      "grad_norm": 0.861406683921814,
+      "learning_rate": 0.0002,
+      "loss": 0.5694,
+      "step": 19860
+    },
+    {
+      "epoch": 3.212351467140894,
+      "grad_norm": 0.8228686451911926,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 19870
+    },
+    {
+      "epoch": 3.213968151321639,
+      "grad_norm": 0.7936596870422363,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19880
+    },
+    {
+      "epoch": 3.2155848355023844,
+      "grad_norm": 1.097377896308899,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 19890
+    },
+    {
+      "epoch": 3.21720151968313,
+      "grad_norm": 0.9544782638549805,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 19900
+    },
+    {
+      "epoch": 3.2188182038638753,
+      "grad_norm": 0.8240751624107361,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 19910
+    },
+    {
+      "epoch": 3.2204348880446205,
+      "grad_norm": 0.8332096338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 19920
+    },
+    {
+      "epoch": 3.2220515722253658,
+      "grad_norm": 1.0954567193984985,
+      "learning_rate": 0.0002,
+      "loss": 0.6274,
+      "step": 19930
+    },
+    {
+      "epoch": 3.223668256406111,
+      "grad_norm": 0.7790525555610657,
+      "learning_rate": 0.0002,
+      "loss": 0.652,
+      "step": 19940
+    },
+    {
+      "epoch": 3.225284940586856,
+      "grad_norm": 0.7966814041137695,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19950
+    },
+    {
+      "epoch": 3.2269016247676015,
+      "grad_norm": 0.9751881957054138,
+      "learning_rate": 0.0002,
+      "loss": 0.5911,
+      "step": 19960
+    },
+    {
+      "epoch": 3.228518308948347,
+      "grad_norm": 0.9856047630310059,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 19970
+    },
+    {
+      "epoch": 3.2301349931290924,
+      "grad_norm": 1.3062353134155273,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 19980
+    },
+    {
+      "epoch": 3.2317516773098376,
+      "grad_norm": 0.9510692358016968,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 19990
+    },
+    {
+      "epoch": 3.233368361490583,
+      "grad_norm": 0.8630342483520508,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 20000
+    },
+    {
+      "epoch": 3.234985045671328,
+      "grad_norm": 0.8966519236564636,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20010
+    },
+    {
+      "epoch": 3.2366017298520733,
+      "grad_norm": 0.7093510627746582,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 20020
+    },
+    {
+      "epoch": 3.2382184140328185,
+      "grad_norm": 0.7771096229553223,
+      "learning_rate": 0.0002,
+      "loss": 0.585,
+      "step": 20030
+    },
+    {
+      "epoch": 3.2398350982135637,
+      "grad_norm": 0.841058075428009,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 20040
+    },
+    {
+      "epoch": 3.2414517823943094,
+      "grad_norm": 0.909712553024292,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 20050
+    },
+    {
+      "epoch": 3.2430684665750547,
+      "grad_norm": 0.8321019411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20060
+    },
+    {
+      "epoch": 3.2446851507558,
+      "grad_norm": 0.779901921749115,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 20070
+    },
+    {
+      "epoch": 3.246301834936545,
+      "grad_norm": 0.6249170303344727,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 20080
+    },
+    {
+      "epoch": 3.2479185191172903,
+      "grad_norm": 0.8000940680503845,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 20090
+    },
+    {
+      "epoch": 3.2495352032980356,
+      "grad_norm": 0.7627735137939453,
+      "learning_rate": 0.0002,
+      "loss": 0.6273,
+      "step": 20100
+    },
+    {
+      "epoch": 3.2511518874787813,
+      "grad_norm": 0.8780747056007385,
+      "learning_rate": 0.0002,
+      "loss": 0.6223,
+      "step": 20110
+    },
+    {
+      "epoch": 3.2527685716595265,
+      "grad_norm": 0.772037148475647,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 20120
+    },
+    {
+      "epoch": 3.2543852558402717,
+      "grad_norm": 1.0086580514907837,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 20130
+    },
+    {
+      "epoch": 3.256001940021017,
+      "grad_norm": 0.9360289573669434,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20140
+    },
+    {
+      "epoch": 3.257618624201762,
+      "grad_norm": 1.2099586725234985,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20150
+    },
+    {
+      "epoch": 3.2592353083825074,
+      "grad_norm": 0.8368481397628784,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 20160
+    },
+    {
+      "epoch": 3.2608519925632526,
+      "grad_norm": 0.7391039133071899,
+      "learning_rate": 0.0002,
+      "loss": 0.5626,
+      "step": 20170
+    },
+    {
+      "epoch": 3.262468676743998,
+      "grad_norm": 0.9122273325920105,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 20180
+    },
+    {
+      "epoch": 3.264085360924743,
+      "grad_norm": 0.8502281904220581,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 20190
+    },
+    {
+      "epoch": 3.265702045105489,
+      "grad_norm": 1.0926852226257324,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 20200
+    },
+    {
+      "epoch": 3.267318729286234,
+      "grad_norm": 0.7902828454971313,
+      "learning_rate": 0.0002,
+      "loss": 0.6027,
+      "step": 20210
+    },
+    {
+      "epoch": 3.2689354134669792,
+      "grad_norm": 0.8724729418754578,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20220
+    },
+    {
+      "epoch": 3.2705520976477245,
+      "grad_norm": 0.8469277024269104,
+      "learning_rate": 0.0002,
+      "loss": 0.6242,
+      "step": 20230
+    },
+    {
+      "epoch": 3.2721687818284697,
+      "grad_norm": 0.8865092992782593,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 20240
+    },
+    {
+      "epoch": 3.273785466009215,
+      "grad_norm": 1.0979334115982056,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20250
+    },
+    {
+      "epoch": 3.2754021501899606,
+      "grad_norm": 1.0860793590545654,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 20260
+    },
+    {
+      "epoch": 3.277018834370706,
+      "grad_norm": 0.981745183467865,
+      "learning_rate": 0.0002,
+      "loss": 0.6105,
+      "step": 20270
+    },
+    {
+      "epoch": 3.278635518551451,
+      "grad_norm": 0.9155020713806152,
+      "learning_rate": 0.0002,
+      "loss": 0.627,
+      "step": 20280
+    },
+    {
+      "epoch": 3.2802522027321963,
+      "grad_norm": 0.8436718583106995,
+      "learning_rate": 0.0002,
+      "loss": 0.5899,
+      "step": 20290
+    },
+    {
+      "epoch": 3.2818688869129415,
+      "grad_norm": 1.0329409837722778,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 20300
+    },
+    {
+      "epoch": 3.2834855710936868,
+      "grad_norm": 0.9876394271850586,
+      "learning_rate": 0.0002,
+      "loss": 0.6,
+      "step": 20310
+    },
+    {
+      "epoch": 3.285102255274432,
+      "grad_norm": 0.8052917718887329,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 20320
+    },
+    {
+      "epoch": 3.2867189394551772,
+      "grad_norm": 0.8390680551528931,
+      "learning_rate": 0.0002,
+      "loss": 0.5949,
+      "step": 20330
+    },
+    {
+      "epoch": 3.288335623635923,
+      "grad_norm": 0.9515735507011414,
+      "learning_rate": 0.0002,
+      "loss": 0.6492,
+      "step": 20340
+    },
+    {
+      "epoch": 3.289952307816668,
+      "grad_norm": 0.8028870224952698,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 20350
+    },
+    {
+      "epoch": 3.2915689919974134,
+      "grad_norm": 0.862592339515686,
+      "learning_rate": 0.0002,
+      "loss": 0.634,
+      "step": 20360
+    },
+    {
+      "epoch": 3.2931856761781586,
+      "grad_norm": 0.7451621890068054,
+      "learning_rate": 0.0002,
+      "loss": 0.6345,
+      "step": 20370
+    },
+    {
+      "epoch": 3.294802360358904,
+      "grad_norm": 0.8966776728630066,
+      "learning_rate": 0.0002,
+      "loss": 0.6458,
+      "step": 20380
+    },
+    {
+      "epoch": 3.296419044539649,
+      "grad_norm": 0.9289216995239258,
+      "learning_rate": 0.0002,
+      "loss": 0.5967,
+      "step": 20390
+    },
+    {
+      "epoch": 3.2980357287203943,
+      "grad_norm": 0.9649626612663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6599,
+      "step": 20400
+    },
+    {
+      "epoch": 3.29965241290114,
+      "grad_norm": 1.1953798532485962,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 20410
+    },
+    {
+      "epoch": 3.301269097081885,
+      "grad_norm": 0.8929083943367004,
+      "learning_rate": 0.0002,
+      "loss": 0.5997,
+      "step": 20420
+    },
+    {
+      "epoch": 3.3028857812626304,
+      "grad_norm": 0.8922014236450195,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 20430
+    },
+    {
+      "epoch": 3.3045024654433757,
+      "grad_norm": 0.9754860401153564,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 20440
+    },
+    {
+      "epoch": 3.306119149624121,
+      "grad_norm": 0.8873140215873718,
+      "learning_rate": 0.0002,
+      "loss": 0.5653,
+      "step": 20450
+    },
+    {
+      "epoch": 3.307735833804866,
+      "grad_norm": 0.857271671295166,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20460
+    },
+    {
+      "epoch": 3.3093525179856114,
+      "grad_norm": 0.9022141098976135,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 20470
+    },
+    {
+      "epoch": 3.3109692021663566,
+      "grad_norm": 0.8614798188209534,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 20480
+    },
+    {
+      "epoch": 3.3125858863471023,
+      "grad_norm": 0.8838164210319519,
+      "learning_rate": 0.0002,
+      "loss": 0.6254,
+      "step": 20490
+    },
+    {
+      "epoch": 3.3142025705278475,
+      "grad_norm": 0.8709736466407776,
+      "learning_rate": 0.0002,
+      "loss": 0.5849,
+      "step": 20500
+    },
+    {
+      "epoch": 3.3158192547085927,
+      "grad_norm": 0.9533300995826721,
+      "learning_rate": 0.0002,
+      "loss": 0.6146,
+      "step": 20510
+    },
+    {
+      "epoch": 3.317435938889338,
+      "grad_norm": 0.8259269595146179,
+      "learning_rate": 0.0002,
+      "loss": 0.6029,
+      "step": 20520
+    },
+    {
+      "epoch": 3.319052623070083,
+      "grad_norm": 0.8607608079910278,
+      "learning_rate": 0.0002,
+      "loss": 0.6268,
+      "step": 20530
+    },
+    {
+      "epoch": 3.3206693072508284,
+      "grad_norm": 1.0863020420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.5676,
+      "step": 20540
+    },
+    {
+      "epoch": 3.3222859914315737,
+      "grad_norm": 1.011489987373352,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 20550
+    },
+    {
+      "epoch": 3.3239026756123193,
+      "grad_norm": 0.6952177882194519,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 20560
+    },
+    {
+      "epoch": 3.3255193597930646,
+      "grad_norm": 0.9638974070549011,
+      "learning_rate": 0.0002,
+      "loss": 0.6229,
+      "step": 20570
+    },
+    {
+      "epoch": 3.32713604397381,
+      "grad_norm": 1.0310138463974,
+      "learning_rate": 0.0002,
+      "loss": 0.5882,
+      "step": 20580
+    },
+    {
+      "epoch": 3.328752728154555,
+      "grad_norm": 0.9371318221092224,
+      "learning_rate": 0.0002,
+      "loss": 0.594,
+      "step": 20590
+    },
+    {
+      "epoch": 3.3303694123353003,
+      "grad_norm": 0.8756691813468933,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 20600
+    },
+    {
+      "epoch": 3.3319860965160455,
+      "grad_norm": 1.054175853729248,
+      "learning_rate": 0.0002,
+      "loss": 0.5994,
+      "step": 20610
+    },
+    {
+      "epoch": 3.3336027806967907,
+      "grad_norm": 0.9074128270149231,
+      "learning_rate": 0.0002,
+      "loss": 0.6169,
+      "step": 20620
+    },
+    {
+      "epoch": 3.335219464877536,
+      "grad_norm": 0.906900942325592,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20630
+    },
+    {
+      "epoch": 3.3368361490582816,
+      "grad_norm": 0.8689333200454712,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 20640
+    },
+    {
+      "epoch": 3.338452833239027,
+      "grad_norm": 0.9889747500419617,
+      "learning_rate": 0.0002,
+      "loss": 0.6079,
+      "step": 20650
+    },
+    {
+      "epoch": 3.340069517419772,
+      "grad_norm": 1.0685805082321167,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20660
+    },
+    {
+      "epoch": 3.3416862016005173,
+      "grad_norm": 0.7495010495185852,
+      "learning_rate": 0.0002,
+      "loss": 0.6091,
+      "step": 20670
+    },
+    {
+      "epoch": 3.3433028857812626,
+      "grad_norm": 0.8747848272323608,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 20680
+    },
+    {
+      "epoch": 3.344919569962008,
+      "grad_norm": 0.9762673377990723,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 20690
+    },
+    {
+      "epoch": 3.346536254142753,
+      "grad_norm": 1.0284489393234253,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 20700
+    },
+    {
+      "epoch": 3.3481529383234987,
+      "grad_norm": 0.7293812036514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20710
+    },
+    {
+      "epoch": 3.349769622504244,
+      "grad_norm": 0.8330199122428894,
+      "learning_rate": 0.0002,
+      "loss": 0.609,
+      "step": 20720
+    },
+    {
+      "epoch": 3.351386306684989,
+      "grad_norm": 0.9808499217033386,
+      "learning_rate": 0.0002,
+      "loss": 0.5729,
+      "step": 20730
+    },
+    {
+      "epoch": 3.3530029908657344,
+      "grad_norm": 0.9508825540542603,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 20740
+    },
+    {
+      "epoch": 3.3546196750464796,
+      "grad_norm": 0.790483832359314,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 20750
+    },
+    {
+      "epoch": 3.356236359227225,
+      "grad_norm": 1.022793173789978,
+      "learning_rate": 0.0002,
+      "loss": 0.6327,
+      "step": 20760
+    },
+    {
+      "epoch": 3.35785304340797,
+      "grad_norm": 0.8318950533866882,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 20770
+    },
+    {
+      "epoch": 3.3594697275887153,
+      "grad_norm": 0.7980858087539673,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 20780
+    },
+    {
+      "epoch": 3.361086411769461,
+      "grad_norm": 0.8114802241325378,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 20790
+    },
+    {
+      "epoch": 3.3627030959502062,
+      "grad_norm": 0.8522519469261169,
+      "learning_rate": 0.0002,
+      "loss": 0.6017,
+      "step": 20800
+    },
+    {
+      "epoch": 3.3643197801309515,
+      "grad_norm": 0.9142431616783142,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 20810
+    },
+    {
+      "epoch": 3.3659364643116967,
+      "grad_norm": 0.771170437335968,
+      "learning_rate": 0.0002,
+      "loss": 0.6331,
+      "step": 20820
+    },
+    {
+      "epoch": 3.367553148492442,
+      "grad_norm": 1.0628231763839722,
+      "learning_rate": 0.0002,
+      "loss": 0.5879,
+      "step": 20830
+    },
+    {
+      "epoch": 3.369169832673187,
+      "grad_norm": 0.9384352564811707,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 20840
+    },
+    {
+      "epoch": 3.370786516853933,
+      "grad_norm": 1.1286591291427612,
+      "learning_rate": 0.0002,
+      "loss": 0.6292,
+      "step": 20850
+    },
+    {
+      "epoch": 3.372403201034678,
+      "grad_norm": 1.1349513530731201,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 20860
+    },
+    {
+      "epoch": 3.3740198852154233,
+      "grad_norm": 1.0127464532852173,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 20870
+    },
+    {
+      "epoch": 3.3756365693961685,
+      "grad_norm": 0.9111971855163574,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 20880
+    },
+    {
+      "epoch": 3.3772532535769137,
+      "grad_norm": 0.871356725692749,
+      "learning_rate": 0.0002,
+      "loss": 0.6101,
+      "step": 20890
+    },
+    {
+      "epoch": 3.378869937757659,
+      "grad_norm": 0.7774117588996887,
+      "learning_rate": 0.0002,
+      "loss": 0.5995,
+      "step": 20900
+    },
+    {
+      "epoch": 3.380486621938404,
+      "grad_norm": 1.0089964866638184,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 20910
+    },
+    {
+      "epoch": 3.3821033061191494,
+      "grad_norm": 0.7855867147445679,
+      "learning_rate": 0.0002,
+      "loss": 0.5908,
+      "step": 20920
+    },
+    {
+      "epoch": 3.3837199902998947,
+      "grad_norm": 1.3713710308074951,
+      "learning_rate": 0.0002,
+      "loss": 0.6373,
+      "step": 20930
+    },
+    {
+      "epoch": 3.3853366744806404,
+      "grad_norm": 0.8599116206169128,
+      "learning_rate": 0.0002,
+      "loss": 0.6627,
+      "step": 20940
+    },
+    {
+      "epoch": 3.3869533586613856,
+      "grad_norm": 0.9392673373222351,
+      "learning_rate": 0.0002,
+      "loss": 0.6224,
+      "step": 20950
+    },
+    {
+      "epoch": 3.388570042842131,
+      "grad_norm": 0.8764075040817261,
+      "learning_rate": 0.0002,
+      "loss": 0.5855,
+      "step": 20960
+    },
+    {
+      "epoch": 3.390186727022876,
+      "grad_norm": 0.8240136504173279,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 20970
+    },
+    {
+      "epoch": 3.3918034112036213,
+      "grad_norm": 1.0982369184494019,
+      "learning_rate": 0.0002,
+      "loss": 0.5783,
+      "step": 20980
+    },
+    {
+      "epoch": 3.3934200953843665,
+      "grad_norm": 1.0599013566970825,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 20990
+    },
+    {
+      "epoch": 3.395036779565112,
+      "grad_norm": 0.895438015460968,
+      "learning_rate": 0.0002,
+      "loss": 0.6356,
+      "step": 21000
+    },
+    {
+      "epoch": 3.3966534637458574,
+      "grad_norm": 0.6974841356277466,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 21010
+    },
+    {
+      "epoch": 3.3982701479266026,
+      "grad_norm": 0.9571719765663147,
+      "learning_rate": 0.0002,
+      "loss": 0.5704,
+      "step": 21020
+    },
+    {
+      "epoch": 3.399886832107348,
+      "grad_norm": 0.831912636756897,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 21030
+    },
+    {
+      "epoch": 3.401503516288093,
+      "grad_norm": 0.831936240196228,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 21040
+    },
+    {
+      "epoch": 3.4031202004688383,
+      "grad_norm": 0.7388373613357544,
+      "learning_rate": 0.0002,
+      "loss": 0.5857,
+      "step": 21050
+    },
+    {
+      "epoch": 3.4047368846495836,
+      "grad_norm": 0.938667356967926,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21060
+    },
+    {
+      "epoch": 3.406353568830329,
+      "grad_norm": 0.9202313423156738,
+      "learning_rate": 0.0002,
+      "loss": 0.6121,
+      "step": 21070
+    },
+    {
+      "epoch": 3.4079702530110745,
+      "grad_norm": 0.9888381958007812,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 21080
+    },
+    {
+      "epoch": 3.4095869371918197,
+      "grad_norm": 0.8526970744132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21090
+    },
+    {
+      "epoch": 3.411203621372565,
+      "grad_norm": 0.7939383387565613,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 21100
+    },
+    {
+      "epoch": 3.41282030555331,
+      "grad_norm": 0.9986352920532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 21110
+    },
+    {
+      "epoch": 3.4144369897340554,
+      "grad_norm": 0.8895300030708313,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 21120
+    },
+    {
+      "epoch": 3.4160536739148006,
+      "grad_norm": 0.9559482932090759,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 21130
+    },
+    {
+      "epoch": 3.417670358095546,
+      "grad_norm": 0.8351506590843201,
+      "learning_rate": 0.0002,
+      "loss": 0.6491,
+      "step": 21140
+    },
+    {
+      "epoch": 3.4192870422762915,
+      "grad_norm": 0.8224456906318665,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 21150
+    },
+    {
+      "epoch": 3.4209037264570368,
+      "grad_norm": 1.0110299587249756,
+      "learning_rate": 0.0002,
+      "loss": 0.5871,
+      "step": 21160
+    },
+    {
+      "epoch": 3.422520410637782,
+      "grad_norm": 0.82564777135849,
+      "learning_rate": 0.0002,
+      "loss": 0.6116,
+      "step": 21170
+    },
+    {
+      "epoch": 3.4241370948185272,
+      "grad_norm": 1.004738688468933,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 21180
+    },
+    {
+      "epoch": 3.4257537789992725,
+      "grad_norm": 0.7545676827430725,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 21190
+    },
+    {
+      "epoch": 3.4273704631800177,
+      "grad_norm": 0.8918704390525818,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 21200
+    },
+    {
+      "epoch": 3.428987147360763,
+      "grad_norm": 0.8336876034736633,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 21210
+    },
+    {
+      "epoch": 3.430603831541508,
+      "grad_norm": 0.8928771018981934,
+      "learning_rate": 0.0002,
+      "loss": 0.5824,
+      "step": 21220
+    },
+    {
+      "epoch": 3.432220515722254,
+      "grad_norm": 0.7663705945014954,
+      "learning_rate": 0.0002,
+      "loss": 0.6468,
+      "step": 21230
+    },
+    {
+      "epoch": 3.433837199902999,
+      "grad_norm": 0.8392598628997803,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 21240
+    },
+    {
+      "epoch": 3.4354538840837443,
+      "grad_norm": 0.8819600343704224,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 21250
+    },
+    {
+      "epoch": 3.4370705682644895,
+      "grad_norm": 0.9124642014503479,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 21260
+    },
+    {
+      "epoch": 3.4386872524452348,
+      "grad_norm": 0.8329763412475586,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 21270
+    },
+    {
+      "epoch": 3.44030393662598,
+      "grad_norm": 0.9982839822769165,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 21280
+    },
+    {
+      "epoch": 3.4419206208067252,
+      "grad_norm": 0.9105954766273499,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 21290
+    },
+    {
+      "epoch": 3.443537304987471,
+      "grad_norm": 0.8182359337806702,
+      "learning_rate": 0.0002,
+      "loss": 0.6028,
+      "step": 21300
+    },
+    {
+      "epoch": 3.445153989168216,
+      "grad_norm": 1.0568904876708984,
+      "learning_rate": 0.0002,
+      "loss": 0.5991,
+      "step": 21310
+    },
+    {
+      "epoch": 3.4467706733489614,
+      "grad_norm": 0.968539834022522,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 21320
+    },
+    {
+      "epoch": 3.4483873575297066,
+      "grad_norm": 0.8774511218070984,
+      "learning_rate": 0.0002,
+      "loss": 0.6219,
+      "step": 21330
+    },
+    {
+      "epoch": 3.450004041710452,
+      "grad_norm": 0.7598156332969666,
+      "learning_rate": 0.0002,
+      "loss": 0.6438,
+      "step": 21340
+    },
+    {
+      "epoch": 3.451620725891197,
+      "grad_norm": 1.1012897491455078,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 21350
+    },
+    {
+      "epoch": 3.4532374100719423,
+      "grad_norm": 0.8040637373924255,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 21360
+    },
+    {
+      "epoch": 3.4548540942526875,
+      "grad_norm": 0.8497496247291565,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 21370
+    },
+    {
+      "epoch": 3.456470778433433,
+      "grad_norm": 0.8429915904998779,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 21380
+    },
+    {
+      "epoch": 3.4580874626141784,
+      "grad_norm": 0.8107112646102905,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 21390
+    },
+    {
+      "epoch": 3.4597041467949237,
+      "grad_norm": 1.00872004032135,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 21400
+    },
+    {
+      "epoch": 3.461320830975669,
+      "grad_norm": 0.8266542553901672,
+      "learning_rate": 0.0002,
+      "loss": 0.5712,
+      "step": 21410
+    },
+    {
+      "epoch": 3.462937515156414,
+      "grad_norm": 0.8972568511962891,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 21420
+    },
+    {
+      "epoch": 3.4645541993371594,
+      "grad_norm": 1.0781476497650146,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 21430
+    },
+    {
+      "epoch": 3.4661708835179046,
+      "grad_norm": 0.9571592807769775,
+      "learning_rate": 0.0002,
+      "loss": 0.6303,
+      "step": 21440
+    },
+    {
+      "epoch": 3.4677875676986503,
+      "grad_norm": 0.881547212600708,
+      "learning_rate": 0.0002,
+      "loss": 0.6309,
+      "step": 21450
+    },
+    {
+      "epoch": 3.4694042518793955,
+      "grad_norm": 0.6955338716506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 21460
+    },
+    {
+      "epoch": 3.4710209360601407,
+      "grad_norm": 0.901187539100647,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 21470
+    },
+    {
+      "epoch": 3.472637620240886,
+      "grad_norm": 0.7063511610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.639,
+      "step": 21480
+    },
+    {
+      "epoch": 3.474254304421631,
+      "grad_norm": 0.8462792038917542,
+      "learning_rate": 0.0002,
+      "loss": 0.6154,
+      "step": 21490
+    },
+    {
+      "epoch": 3.4758709886023764,
+      "grad_norm": 1.1861060857772827,
+      "learning_rate": 0.0002,
+      "loss": 0.61,
+      "step": 21500
+    },
+    {
+      "epoch": 3.4774876727831217,
+      "grad_norm": 0.70503169298172,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 21510
+    },
+    {
+      "epoch": 3.479104356963867,
+      "grad_norm": 0.9650066494941711,
+      "learning_rate": 0.0002,
+      "loss": 0.6475,
+      "step": 21520
+    },
+    {
+      "epoch": 3.4807210411446126,
+      "grad_norm": 1.0266852378845215,
+      "learning_rate": 0.0002,
+      "loss": 0.6452,
+      "step": 21530
+    },
+    {
+      "epoch": 3.482337725325358,
+      "grad_norm": 0.956372857093811,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 21540
+    },
+    {
+      "epoch": 3.483954409506103,
+      "grad_norm": 0.8848432898521423,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 21550
+    },
+    {
+      "epoch": 3.4855710936868483,
+      "grad_norm": 1.0805351734161377,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 21560
+    },
+    {
+      "epoch": 3.4871877778675935,
+      "grad_norm": 0.9279725551605225,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 21570
+    },
+    {
+      "epoch": 3.4888044620483387,
+      "grad_norm": 0.9049562215805054,
+      "learning_rate": 0.0002,
+      "loss": 0.6354,
+      "step": 21580
+    },
+    {
+      "epoch": 3.4904211462290844,
+      "grad_norm": 0.9619429111480713,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 21590
+    },
+    {
+      "epoch": 3.4920378304098296,
+      "grad_norm": 0.8508906960487366,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 21600
+    },
+    {
+      "epoch": 3.493654514590575,
+      "grad_norm": 0.8692502379417419,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 21610
+    },
+    {
+      "epoch": 3.49527119877132,
+      "grad_norm": 0.8187332153320312,
+      "learning_rate": 0.0002,
+      "loss": 0.5878,
+      "step": 21620
+    },
+    {
+      "epoch": 3.4968878829520653,
+      "grad_norm": 1.145400047302246,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 21630
+    },
+    {
+      "epoch": 3.4985045671328105,
+      "grad_norm": 0.8281388282775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 21640
+    },
+    {
+      "epoch": 3.500121251313556,
+      "grad_norm": 0.82256019115448,
+      "learning_rate": 0.0002,
+      "loss": 0.6624,
+      "step": 21650
+    },
+    {
+      "epoch": 3.501737935494301,
+      "grad_norm": 0.9315484762191772,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 21660
+    },
+    {
+      "epoch": 3.5033546196750462,
+      "grad_norm": 0.7626111507415771,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 21670
+    },
+    {
+      "epoch": 3.504971303855792,
+      "grad_norm": 0.9275059103965759,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 21680
+    },
+    {
+      "epoch": 3.506587988036537,
+      "grad_norm": 0.7906724810600281,
+      "learning_rate": 0.0002,
+      "loss": 0.64,
+      "step": 21690
+    },
+    {
+      "epoch": 3.5082046722172824,
+      "grad_norm": 0.8289761543273926,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 21700
+    },
+    {
+      "epoch": 3.5098213563980276,
+      "grad_norm": 0.8316431045532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 21710
+    },
+    {
+      "epoch": 3.511438040578773,
+      "grad_norm": 1.0451812744140625,
+      "learning_rate": 0.0002,
+      "loss": 0.619,
+      "step": 21720
+    },
+    {
+      "epoch": 3.513054724759518,
+      "grad_norm": 0.928252637386322,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 21730
+    },
+    {
+      "epoch": 3.5146714089402638,
+      "grad_norm": 0.7985895276069641,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 21740
+    },
+    {
+      "epoch": 3.516288093121009,
+      "grad_norm": 0.6740974187850952,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 21750
+    },
+    {
+      "epoch": 3.517904777301754,
+      "grad_norm": 0.8482223749160767,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 21760
+    },
+    {
+      "epoch": 3.5195214614824994,
+      "grad_norm": 0.889947772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.6277,
+      "step": 21770
+    },
+    {
+      "epoch": 3.5211381456632447,
+      "grad_norm": 0.8304598927497864,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 21780
+    },
+    {
+      "epoch": 3.52275482984399,
+      "grad_norm": 0.8002981543540955,
+      "learning_rate": 0.0002,
+      "loss": 0.6156,
+      "step": 21790
+    },
+    {
+      "epoch": 3.524371514024735,
+      "grad_norm": 0.8115083575248718,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21800
+    },
+    {
+      "epoch": 3.5259881982054804,
+      "grad_norm": 0.9715048670768738,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 21810
+    },
+    {
+      "epoch": 3.5276048823862256,
+      "grad_norm": 1.0910786390304565,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 21820
+    },
+    {
+      "epoch": 3.5292215665669713,
+      "grad_norm": 0.8438942432403564,
+      "learning_rate": 0.0002,
+      "loss": 0.6543,
+      "step": 21830
+    },
+    {
+      "epoch": 3.5308382507477165,
+      "grad_norm": 0.8813382983207703,
+      "learning_rate": 0.0002,
+      "loss": 0.6509,
+      "step": 21840
+    },
+    {
+      "epoch": 3.5324549349284617,
+      "grad_norm": 0.7092908024787903,
+      "learning_rate": 0.0002,
+      "loss": 0.6049,
+      "step": 21850
+    },
+    {
+      "epoch": 3.534071619109207,
+      "grad_norm": 0.8332187533378601,
+      "learning_rate": 0.0002,
+      "loss": 0.5678,
+      "step": 21860
+    },
+    {
+      "epoch": 3.535688303289952,
+      "grad_norm": 0.8958209156990051,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21870
+    },
+    {
+      "epoch": 3.5373049874706974,
+      "grad_norm": 0.824138879776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 21880
+    },
+    {
+      "epoch": 3.538921671651443,
+      "grad_norm": 0.8375158309936523,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 21890
+    },
+    {
+      "epoch": 3.5405383558321883,
+      "grad_norm": 1.0274608135223389,
+      "learning_rate": 0.0002,
+      "loss": 0.6019,
+      "step": 21900
+    },
+    {
+      "epoch": 3.5421550400129336,
+      "grad_norm": 0.7088932394981384,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 21910
+    },
+    {
+      "epoch": 3.543771724193679,
+      "grad_norm": 0.8172445297241211,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 21920
+    },
+    {
+      "epoch": 3.545388408374424,
+      "grad_norm": 0.9904135465621948,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 21930
+    },
+    {
+      "epoch": 3.5470050925551693,
+      "grad_norm": 0.9900432229042053,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 21940
+    },
+    {
+      "epoch": 3.5486217767359145,
+      "grad_norm": 0.8963301181793213,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 21950
+    },
+    {
+      "epoch": 3.5502384609166597,
+      "grad_norm": 0.8551464676856995,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 21960
+    },
+    {
+      "epoch": 3.551855145097405,
+      "grad_norm": 1.0916603803634644,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 21970
+    },
+    {
+      "epoch": 3.5534718292781506,
+      "grad_norm": 0.841598391532898,
+      "learning_rate": 0.0002,
+      "loss": 0.6523,
+      "step": 21980
+    },
+    {
+      "epoch": 3.555088513458896,
+      "grad_norm": 0.8566757440567017,
+      "learning_rate": 0.0002,
+      "loss": 0.617,
+      "step": 21990
+    },
+    {
+      "epoch": 3.556705197639641,
+      "grad_norm": 1.0145052671432495,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 22000
+    },
+    {
+      "epoch": 3.5583218818203863,
+      "grad_norm": 0.9293754696846008,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 22010
+    },
+    {
+      "epoch": 3.5599385660011316,
+      "grad_norm": 0.9568536281585693,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 22020
+    },
+    {
+      "epoch": 3.5615552501818772,
+      "grad_norm": 0.8613139986991882,
+      "learning_rate": 0.0002,
+      "loss": 0.641,
+      "step": 22030
+    },
+    {
+      "epoch": 3.5631719343626225,
+      "grad_norm": 0.8179237246513367,
+      "learning_rate": 0.0002,
+      "loss": 0.6496,
+      "step": 22040
+    },
+    {
+      "epoch": 3.5647886185433677,
+      "grad_norm": 0.9059830904006958,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 22050
+    },
+    {
+      "epoch": 3.566405302724113,
+      "grad_norm": 1.0068252086639404,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 22060
+    },
+    {
+      "epoch": 3.568021986904858,
+      "grad_norm": 0.9682072997093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6239,
+      "step": 22070
+    },
+    {
+      "epoch": 3.5696386710856034,
+      "grad_norm": 0.8514005541801453,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 22080
+    },
+    {
+      "epoch": 3.5712553552663486,
+      "grad_norm": 0.8327770829200745,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 22090
+    },
+    {
+      "epoch": 3.572872039447094,
+      "grad_norm": 1.024976372718811,
+      "learning_rate": 0.0002,
+      "loss": 0.5976,
+      "step": 22100
+    },
+    {
+      "epoch": 3.574488723627839,
+      "grad_norm": 0.7721174955368042,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 22110
+    },
+    {
+      "epoch": 3.5761054078085843,
+      "grad_norm": 1.0351054668426514,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 22120
+    },
+    {
+      "epoch": 3.57772209198933,
+      "grad_norm": 0.9680907130241394,
+      "learning_rate": 0.0002,
+      "loss": 0.6379,
+      "step": 22130
+    },
+    {
+      "epoch": 3.5793387761700752,
+      "grad_norm": 0.8016974925994873,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 22140
+    },
+    {
+      "epoch": 3.5809554603508205,
+      "grad_norm": 1.0109003782272339,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 22150
+    },
+    {
+      "epoch": 3.5825721445315657,
+      "grad_norm": 1.0473392009735107,
+      "learning_rate": 0.0002,
+      "loss": 0.6368,
+      "step": 22160
+    },
+    {
+      "epoch": 3.584188828712311,
+      "grad_norm": 0.8686613440513611,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 22170
+    },
+    {
+      "epoch": 3.5858055128930566,
+      "grad_norm": 0.869149923324585,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 22180
+    },
+    {
+      "epoch": 3.587422197073802,
+      "grad_norm": 0.9769062995910645,
+      "learning_rate": 0.0002,
+      "loss": 0.5895,
+      "step": 22190
+    },
+    {
+      "epoch": 3.589038881254547,
+      "grad_norm": 0.779636561870575,
+      "learning_rate": 0.0002,
+      "loss": 0.5939,
+      "step": 22200
+    },
+    {
+      "epoch": 3.5906555654352923,
+      "grad_norm": 0.9063841104507446,
+      "learning_rate": 0.0002,
+      "loss": 0.5875,
+      "step": 22210
+    },
+    {
+      "epoch": 3.5922722496160375,
+      "grad_norm": 0.9216037392616272,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 22220
+    },
+    {
+      "epoch": 3.5938889337967828,
+      "grad_norm": 1.0217336416244507,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 22230
+    },
+    {
+      "epoch": 3.595505617977528,
+      "grad_norm": 0.8513161540031433,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 22240
+    },
+    {
+      "epoch": 3.597122302158273,
+      "grad_norm": 0.8084813952445984,
+      "learning_rate": 0.0002,
+      "loss": 0.6301,
+      "step": 22250
+    },
+    {
+      "epoch": 3.5987389863390185,
+      "grad_norm": 0.8524802923202515,
+      "learning_rate": 0.0002,
+      "loss": 0.6197,
+      "step": 22260
+    },
+    {
+      "epoch": 3.600355670519764,
+      "grad_norm": 0.9356237649917603,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 22270
+    },
+    {
+      "epoch": 3.6019723547005094,
+      "grad_norm": 1.009600281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.628,
+      "step": 22280
+    },
+    {
+      "epoch": 3.6035890388812546,
+      "grad_norm": 0.9900581240653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 22290
+    },
+    {
+      "epoch": 3.605205723062,
+      "grad_norm": 1.062495231628418,
+      "learning_rate": 0.0002,
+      "loss": 0.5725,
+      "step": 22300
+    },
+    {
+      "epoch": 3.606822407242745,
+      "grad_norm": 0.8832381367683411,
+      "learning_rate": 0.0002,
+      "loss": 0.607,
+      "step": 22310
+    },
+    {
+      "epoch": 3.6084390914234903,
+      "grad_norm": 0.9284297823905945,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 22320
+    },
+    {
+      "epoch": 3.610055775604236,
+      "grad_norm": 1.2381829023361206,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 22330
+    },
+    {
+      "epoch": 3.611672459784981,
+      "grad_norm": 0.929434597492218,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 22340
+    },
+    {
+      "epoch": 3.6132891439657264,
+      "grad_norm": 0.9714490175247192,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 22350
+    },
+    {
+      "epoch": 3.6149058281464717,
+      "grad_norm": 0.808014988899231,
+      "learning_rate": 0.0002,
+      "loss": 0.6861,
+      "step": 22360
+    },
+    {
+      "epoch": 3.616522512327217,
+      "grad_norm": 1.0364398956298828,
+      "learning_rate": 0.0002,
+      "loss": 0.6428,
+      "step": 22370
+    },
+    {
+      "epoch": 3.618139196507962,
+      "grad_norm": 0.7858489751815796,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22380
+    },
+    {
+      "epoch": 3.6197558806887074,
+      "grad_norm": 0.9920870065689087,
+      "learning_rate": 0.0002,
+      "loss": 0.6214,
+      "step": 22390
+    },
+    {
+      "epoch": 3.6213725648694526,
+      "grad_norm": 0.9183220863342285,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 22400
+    },
+    {
+      "epoch": 3.622989249050198,
+      "grad_norm": 0.9826246500015259,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22410
+    },
+    {
+      "epoch": 3.6246059332309435,
+      "grad_norm": 0.8632931113243103,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 22420
+    },
+    {
+      "epoch": 3.6262226174116887,
+      "grad_norm": 0.8468965291976929,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 22430
+    },
+    {
+      "epoch": 3.627839301592434,
+      "grad_norm": 0.8466871976852417,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 22440
+    },
+    {
+      "epoch": 3.629455985773179,
+      "grad_norm": 0.9501169919967651,
+      "learning_rate": 0.0002,
+      "loss": 0.5941,
+      "step": 22450
+    },
+    {
+      "epoch": 3.6310726699539244,
+      "grad_norm": 0.8906720876693726,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 22460
+    },
+    {
+      "epoch": 3.6326893541346696,
+      "grad_norm": 0.7400227189064026,
+      "learning_rate": 0.0002,
+      "loss": 0.6928,
+      "step": 22470
+    },
+    {
+      "epoch": 3.6343060383154153,
+      "grad_norm": 0.9756355881690979,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22480
+    },
+    {
+      "epoch": 3.6359227224961606,
+      "grad_norm": 0.7504993081092834,
+      "learning_rate": 0.0002,
+      "loss": 0.6203,
+      "step": 22490
+    },
+    {
+      "epoch": 3.637539406676906,
+      "grad_norm": 0.9270039200782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 22500
+    },
+    {
+      "epoch": 3.639156090857651,
+      "grad_norm": 0.8841686844825745,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 22510
+    },
+    {
+      "epoch": 3.6407727750383962,
+      "grad_norm": 0.8533213138580322,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 22520
+    },
+    {
+      "epoch": 3.6423894592191415,
+      "grad_norm": 1.0052043199539185,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 22530
+    },
+    {
+      "epoch": 3.6440061433998867,
+      "grad_norm": 1.0323461294174194,
+      "learning_rate": 0.0002,
+      "loss": 0.6363,
+      "step": 22540
+    },
+    {
+      "epoch": 3.645622827580632,
+      "grad_norm": 0.8654312491416931,
+      "learning_rate": 0.0002,
+      "loss": 0.6545,
+      "step": 22550
+    },
+    {
+      "epoch": 3.647239511761377,
+      "grad_norm": 0.6400038003921509,
+      "learning_rate": 0.0002,
+      "loss": 0.6155,
+      "step": 22560
+    },
+    {
+      "epoch": 3.648856195942123,
+      "grad_norm": 0.8061298727989197,
+      "learning_rate": 0.0002,
+      "loss": 0.5829,
+      "step": 22570
+    },
+    {
+      "epoch": 3.650472880122868,
+      "grad_norm": 0.9257854223251343,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 22580
+    },
+    {
+      "epoch": 3.6520895643036133,
+      "grad_norm": 0.8439396619796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6409,
+      "step": 22590
+    },
+    {
+      "epoch": 3.6537062484843585,
+      "grad_norm": 0.7764544486999512,
+      "learning_rate": 0.0002,
+      "loss": 0.5996,
+      "step": 22600
+    },
+    {
+      "epoch": 3.6553229326651038,
+      "grad_norm": 1.125451683998108,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 22610
+    },
+    {
+      "epoch": 3.656939616845849,
+      "grad_norm": 0.7523018717765808,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 22620
+    },
+    {
+      "epoch": 3.6585563010265947,
+      "grad_norm": 1.071026086807251,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 22630
+    },
+    {
+      "epoch": 3.66017298520734,
+      "grad_norm": 0.945791482925415,
+      "learning_rate": 0.0002,
+      "loss": 0.6459,
+      "step": 22640
+    },
+    {
+      "epoch": 3.661789669388085,
+      "grad_norm": 0.8001811504364014,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 22650
+    },
+    {
+      "epoch": 3.6634063535688304,
+      "grad_norm": 0.9700816869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.6385,
+      "step": 22660
+    },
+    {
+      "epoch": 3.6650230377495756,
+      "grad_norm": 0.9053242206573486,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22670
+    },
+    {
+      "epoch": 3.666639721930321,
+      "grad_norm": 0.944362461566925,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 22680
+    },
+    {
+      "epoch": 3.668256406111066,
+      "grad_norm": 1.067489504814148,
+      "learning_rate": 0.0002,
+      "loss": 0.6235,
+      "step": 22690
+    },
+    {
+      "epoch": 3.6698730902918113,
+      "grad_norm": 1.0984995365142822,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 22700
+    },
+    {
+      "epoch": 3.6714897744725565,
+      "grad_norm": 0.9336317777633667,
+      "learning_rate": 0.0002,
+      "loss": 0.6717,
+      "step": 22710
+    },
+    {
+      "epoch": 3.673106458653302,
+      "grad_norm": 0.9261918663978577,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 22720
+    },
+    {
+      "epoch": 3.6747231428340474,
+      "grad_norm": 0.8648008704185486,
+      "learning_rate": 0.0002,
+      "loss": 0.6332,
+      "step": 22730
+    },
+    {
+      "epoch": 3.6763398270147927,
+      "grad_norm": 0.7225083708763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6576,
+      "step": 22740
+    },
+    {
+      "epoch": 3.677956511195538,
+      "grad_norm": 0.9258282780647278,
+      "learning_rate": 0.0002,
+      "loss": 0.6406,
+      "step": 22750
+    },
+    {
+      "epoch": 3.679573195376283,
+      "grad_norm": 0.70876145362854,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 22760
+    },
+    {
+      "epoch": 3.681189879557029,
+      "grad_norm": 0.8780210018157959,
+      "learning_rate": 0.0002,
+      "loss": 0.6821,
+      "step": 22770
+    },
+    {
+      "epoch": 3.682806563737774,
+      "grad_norm": 0.8075440526008606,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22780
+    },
+    {
+      "epoch": 3.6844232479185193,
+      "grad_norm": 0.8503130674362183,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22790
+    },
+    {
+      "epoch": 3.6860399320992645,
+      "grad_norm": 0.8413618206977844,
+      "learning_rate": 0.0002,
+      "loss": 0.6082,
+      "step": 22800
+    },
+    {
+      "epoch": 3.6876566162800097,
+      "grad_norm": 0.8675165176391602,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 22810
+    },
+    {
+      "epoch": 3.689273300460755,
+      "grad_norm": 0.8235884308815002,
+      "learning_rate": 0.0002,
+      "loss": 0.6157,
+      "step": 22820
+    },
+    {
+      "epoch": 3.6908899846415,
+      "grad_norm": 0.9477725625038147,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 22830
+    },
+    {
+      "epoch": 3.6925066688222454,
+      "grad_norm": 0.7883533835411072,
+      "learning_rate": 0.0002,
+      "loss": 0.6481,
+      "step": 22840
+    },
+    {
+      "epoch": 3.6941233530029907,
+      "grad_norm": 1.047913908958435,
+      "learning_rate": 0.0002,
+      "loss": 0.5872,
+      "step": 22850
+    },
+    {
+      "epoch": 3.695740037183736,
+      "grad_norm": 0.9171528816223145,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 22860
+    },
+    {
+      "epoch": 3.6973567213644816,
+      "grad_norm": 0.9338192343711853,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 22870
+    },
+    {
+      "epoch": 3.698973405545227,
+      "grad_norm": 0.8799443244934082,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 22880
+    },
+    {
+      "epoch": 3.700590089725972,
+      "grad_norm": 0.8515434861183167,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 22890
+    },
+    {
+      "epoch": 3.7022067739067173,
+      "grad_norm": 0.7805591821670532,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 22900
+    },
+    {
+      "epoch": 3.7038234580874625,
+      "grad_norm": 0.8470911979675293,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 22910
+    },
+    {
+      "epoch": 3.705440142268208,
+      "grad_norm": 0.9452309012413025,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 22920
+    },
+    {
+      "epoch": 3.7070568264489534,
+      "grad_norm": 0.950243353843689,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 22930
+    },
+    {
+      "epoch": 3.7086735106296986,
+      "grad_norm": 0.7882499098777771,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 22940
+    },
+    {
+      "epoch": 3.710290194810444,
+      "grad_norm": 0.8307787775993347,
+      "learning_rate": 0.0002,
+      "loss": 0.6462,
+      "step": 22950
+    },
+    {
+      "epoch": 3.711906878991189,
+      "grad_norm": 1.0970630645751953,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 22960
+    },
+    {
+      "epoch": 3.7135235631719343,
+      "grad_norm": 0.8269566297531128,
+      "learning_rate": 0.0002,
+      "loss": 0.6281,
+      "step": 22970
+    },
+    {
+      "epoch": 3.7151402473526796,
+      "grad_norm": 0.8306704759597778,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22980
+    },
+    {
+      "epoch": 3.716756931533425,
+      "grad_norm": 0.9710225462913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6418,
+      "step": 22990
+    },
+    {
+      "epoch": 3.71837361571417,
+      "grad_norm": 0.8890530467033386,
+      "learning_rate": 0.0002,
+      "loss": 0.6639,
+      "step": 23000
+    },
+    {
+      "epoch": 3.7199902998949153,
+      "grad_norm": 0.883522629737854,
+      "learning_rate": 0.0002,
+      "loss": 0.6084,
+      "step": 23010
+    },
+    {
+      "epoch": 3.721606984075661,
+      "grad_norm": 0.8662652373313904,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 23020
+    },
+    {
+      "epoch": 3.723223668256406,
+      "grad_norm": 0.7228406667709351,
+      "learning_rate": 0.0002,
+      "loss": 0.6266,
+      "step": 23030
+    },
+    {
+      "epoch": 3.7248403524371514,
+      "grad_norm": 1.060792088508606,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23040
+    },
+    {
+      "epoch": 3.7264570366178966,
+      "grad_norm": 1.0119613409042358,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 23050
+    },
+    {
+      "epoch": 3.728073720798642,
+      "grad_norm": 0.9212996959686279,
+      "learning_rate": 0.0002,
+      "loss": 0.6466,
+      "step": 23060
+    },
+    {
+      "epoch": 3.7296904049793875,
+      "grad_norm": 0.925690233707428,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 23070
+    },
+    {
+      "epoch": 3.7313070891601328,
+      "grad_norm": 0.8323310613632202,
+      "learning_rate": 0.0002,
+      "loss": 0.615,
+      "step": 23080
+    },
+    {
+      "epoch": 3.732923773340878,
+      "grad_norm": 0.8966048955917358,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 23090
+    },
+    {
+      "epoch": 3.7345404575216232,
+      "grad_norm": 0.8995837569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23100
+    },
+    {
+      "epoch": 3.7361571417023685,
+      "grad_norm": 0.8748890161514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23110
+    },
+    {
+      "epoch": 3.7377738258831137,
+      "grad_norm": 0.7985540628433228,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 23120
+    },
+    {
+      "epoch": 3.739390510063859,
+      "grad_norm": 1.0240917205810547,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 23130
+    },
+    {
+      "epoch": 3.741007194244604,
+      "grad_norm": 0.9181789755821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 23140
+    },
+    {
+      "epoch": 3.7426238784253494,
+      "grad_norm": 0.8896583914756775,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 23150
+    },
+    {
+      "epoch": 3.744240562606095,
+      "grad_norm": 0.8635515570640564,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 23160
+    },
+    {
+      "epoch": 3.7458572467868403,
+      "grad_norm": 0.8873575329780579,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 23170
+    },
+    {
+      "epoch": 3.7474739309675855,
+      "grad_norm": 0.9807148575782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23180
+    },
+    {
+      "epoch": 3.7490906151483308,
+      "grad_norm": 0.900477945804596,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 23190
+    },
+    {
+      "epoch": 3.750707299329076,
+      "grad_norm": 0.9379992485046387,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23200
+    },
+    {
+      "epoch": 3.752323983509821,
+      "grad_norm": 0.9649890661239624,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 23210
+    },
+    {
+      "epoch": 3.753940667690567,
+      "grad_norm": 0.824442446231842,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 23220
+    },
+    {
+      "epoch": 3.755557351871312,
+      "grad_norm": 0.8896150588989258,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 23230
+    },
+    {
+      "epoch": 3.7571740360520574,
+      "grad_norm": 0.751249372959137,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 23240
+    },
+    {
+      "epoch": 3.7587907202328026,
+      "grad_norm": 0.9392193555831909,
+      "learning_rate": 0.0002,
+      "loss": 0.629,
+      "step": 23250
+    },
+    {
+      "epoch": 3.760407404413548,
+      "grad_norm": 0.9284586310386658,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 23260
+    },
+    {
+      "epoch": 3.762024088594293,
+      "grad_norm": 0.7738175392150879,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23270
+    },
+    {
+      "epoch": 3.7636407727750383,
+      "grad_norm": 0.9252978563308716,
+      "learning_rate": 0.0002,
+      "loss": 0.6743,
+      "step": 23280
+    },
+    {
+      "epoch": 3.7652574569557835,
+      "grad_norm": 0.9501895904541016,
+      "learning_rate": 0.0002,
+      "loss": 0.5984,
+      "step": 23290
+    },
+    {
+      "epoch": 3.7668741411365287,
+      "grad_norm": 0.9416276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 23300
+    },
+    {
+      "epoch": 3.7684908253172744,
+      "grad_norm": 0.7076631784439087,
+      "learning_rate": 0.0002,
+      "loss": 0.6507,
+      "step": 23310
+    },
+    {
+      "epoch": 3.7701075094980196,
+      "grad_norm": 0.9864492416381836,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 23320
+    },
+    {
+      "epoch": 3.771724193678765,
+      "grad_norm": 0.8450456261634827,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 23330
+    },
+    {
+      "epoch": 3.77334087785951,
+      "grad_norm": 1.0768941640853882,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23340
+    },
+    {
+      "epoch": 3.7749575620402553,
+      "grad_norm": 0.9956819415092468,
+      "learning_rate": 0.0002,
+      "loss": 0.6408,
+      "step": 23350
+    },
+    {
+      "epoch": 3.7765742462210006,
+      "grad_norm": 0.9234658479690552,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 23360
+    },
+    {
+      "epoch": 3.7781909304017463,
+      "grad_norm": 1.0993858575820923,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23370
+    },
+    {
+      "epoch": 3.7798076145824915,
+      "grad_norm": 0.923159658908844,
+      "learning_rate": 0.0002,
+      "loss": 0.6391,
+      "step": 23380
+    },
+    {
+      "epoch": 3.7814242987632367,
+      "grad_norm": 0.9311541318893433,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23390
+    },
+    {
+      "epoch": 3.783040982943982,
+      "grad_norm": 0.919681191444397,
+      "learning_rate": 0.0002,
+      "loss": 0.6535,
+      "step": 23400
+    },
+    {
+      "epoch": 3.784657667124727,
+      "grad_norm": 1.7406195402145386,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 23410
+    },
+    {
+      "epoch": 3.7862743513054724,
+      "grad_norm": 0.7789074182510376,
+      "learning_rate": 0.0002,
+      "loss": 0.657,
+      "step": 23420
+    },
+    {
+      "epoch": 3.7878910354862176,
+      "grad_norm": 0.8302814960479736,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23430
+    },
+    {
+      "epoch": 3.789507719666963,
+      "grad_norm": 0.8089349269866943,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23440
+    },
+    {
+      "epoch": 3.791124403847708,
+      "grad_norm": 0.9006284475326538,
+      "learning_rate": 0.0002,
+      "loss": 0.6682,
+      "step": 23450
+    },
+    {
+      "epoch": 3.7927410880284538,
+      "grad_norm": 0.8426766991615295,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 23460
+    },
+    {
+      "epoch": 3.794357772209199,
+      "grad_norm": 1.2576252222061157,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 23470
+    },
+    {
+      "epoch": 3.7959744563899442,
+      "grad_norm": 1.0307610034942627,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 23480
+    },
+    {
+      "epoch": 3.7975911405706895,
+      "grad_norm": 0.8525972962379456,
+      "learning_rate": 0.0002,
+      "loss": 0.6262,
+      "step": 23490
+    },
+    {
+      "epoch": 3.7992078247514347,
+      "grad_norm": 1.159039855003357,
+      "learning_rate": 0.0002,
+      "loss": 0.6757,
+      "step": 23500
+    },
+    {
+      "epoch": 3.80082450893218,
+      "grad_norm": 1.4193549156188965,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23510
+    },
+    {
+      "epoch": 3.8024411931129256,
+      "grad_norm": 0.8245543837547302,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 23520
+    },
+    {
+      "epoch": 3.804057877293671,
+      "grad_norm": 0.8847230076789856,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23530
+    },
+    {
+      "epoch": 3.805674561474416,
+      "grad_norm": 0.9574624300003052,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 23540
+    },
+    {
+      "epoch": 3.8072912456551613,
+      "grad_norm": 1.048020601272583,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 23550
+    },
+    {
+      "epoch": 3.8089079298359065,
+      "grad_norm": 0.8302255868911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6497,
+      "step": 23560
+    },
+    {
+      "epoch": 3.8105246140166518,
+      "grad_norm": 0.8269215822219849,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 23570
+    },
+    {
+      "epoch": 3.812141298197397,
+      "grad_norm": 0.9375753402709961,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 23580
+    },
+    {
+      "epoch": 3.8137579823781422,
+      "grad_norm": 1.0234097242355347,
+      "learning_rate": 0.0002,
+      "loss": 0.6132,
+      "step": 23590
+    },
+    {
+      "epoch": 3.8153746665588875,
+      "grad_norm": 0.8978445529937744,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23600
+    },
+    {
+      "epoch": 3.816991350739633,
+      "grad_norm": 0.7929515838623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 23610
+    },
+    {
+      "epoch": 3.8186080349203784,
+      "grad_norm": 1.3255881071090698,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23620
+    },
+    {
+      "epoch": 3.8202247191011236,
+      "grad_norm": 0.9188598990440369,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 23630
+    },
+    {
+      "epoch": 3.821841403281869,
+      "grad_norm": 0.8811675906181335,
+      "learning_rate": 0.0002,
+      "loss": 0.6152,
+      "step": 23640
+    },
+    {
+      "epoch": 3.823458087462614,
+      "grad_norm": 0.8061038255691528,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 23650
+    },
+    {
+      "epoch": 3.8250747716433597,
+      "grad_norm": 0.9975376129150391,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 23660
+    },
+    {
+      "epoch": 3.826691455824105,
+      "grad_norm": 0.8036105036735535,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 23670
+    },
+    {
+      "epoch": 3.82830814000485,
+      "grad_norm": 0.7401984333992004,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 23680
+    },
+    {
+      "epoch": 3.8299248241855954,
+      "grad_norm": 0.829753041267395,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 23690
+    },
+    {
+      "epoch": 3.8315415083663407,
+      "grad_norm": 0.8753240704536438,
+      "learning_rate": 0.0002,
+      "loss": 0.6611,
+      "step": 23700
+    },
+    {
+      "epoch": 3.833158192547086,
+      "grad_norm": 0.8157842755317688,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 23710
+    },
+    {
+      "epoch": 3.834774876727831,
+      "grad_norm": 0.6183798909187317,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 23720
+    },
+    {
+      "epoch": 3.8363915609085764,
+      "grad_norm": 0.9548442363739014,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 23730
+    },
+    {
+      "epoch": 3.8380082450893216,
+      "grad_norm": 0.8319669961929321,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23740
+    },
+    {
+      "epoch": 3.839624929270067,
+      "grad_norm": 0.9718693494796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 23750
+    },
+    {
+      "epoch": 3.8412416134508125,
+      "grad_norm": 0.8672235012054443,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 23760
+    },
+    {
+      "epoch": 3.8428582976315577,
+      "grad_norm": 1.1210707426071167,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 23770
+    },
+    {
+      "epoch": 3.844474981812303,
+      "grad_norm": 0.9177767634391785,
+      "learning_rate": 0.0002,
+      "loss": 0.5923,
+      "step": 23780
+    },
+    {
+      "epoch": 3.846091665993048,
+      "grad_norm": 0.8714171648025513,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 23790
+    },
+    {
+      "epoch": 3.8477083501737934,
+      "grad_norm": 1.1853246688842773,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 23800
+    },
+    {
+      "epoch": 3.849325034354539,
+      "grad_norm": 0.8091260194778442,
+      "learning_rate": 0.0002,
+      "loss": 0.6144,
+      "step": 23810
+    },
+    {
+      "epoch": 3.8509417185352843,
+      "grad_norm": 0.9710774421691895,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23820
+    },
+    {
+      "epoch": 3.8525584027160296,
+      "grad_norm": 0.7648707628250122,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23830
+    },
+    {
+      "epoch": 3.854175086896775,
+      "grad_norm": 0.7809253931045532,
+      "learning_rate": 0.0002,
+      "loss": 0.6013,
+      "step": 23840
+    },
+    {
+      "epoch": 3.85579177107752,
+      "grad_norm": 0.8337951898574829,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 23850
+    },
+    {
+      "epoch": 3.8574084552582653,
+      "grad_norm": 0.9271913170814514,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23860
+    },
+    {
+      "epoch": 3.8590251394390105,
+      "grad_norm": 0.985334038734436,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 23870
+    },
+    {
+      "epoch": 3.8606418236197557,
+      "grad_norm": 0.8458583354949951,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 23880
+    },
+    {
+      "epoch": 3.862258507800501,
+      "grad_norm": 1.015348196029663,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 23890
+    },
+    {
+      "epoch": 3.8638751919812466,
+      "grad_norm": 1.0121688842773438,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23900
+    },
+    {
+      "epoch": 3.865491876161992,
+      "grad_norm": 0.8883971571922302,
+      "learning_rate": 0.0002,
+      "loss": 0.5921,
+      "step": 23910
+    },
+    {
+      "epoch": 3.867108560342737,
+      "grad_norm": 1.028086543083191,
+      "learning_rate": 0.0002,
+      "loss": 0.6597,
+      "step": 23920
+    },
+    {
+      "epoch": 3.8687252445234823,
+      "grad_norm": 0.9645734429359436,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 23930
+    },
+    {
+      "epoch": 3.8703419287042276,
+      "grad_norm": 0.8235350251197815,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 23940
+    },
+    {
+      "epoch": 3.871958612884973,
+      "grad_norm": 1.0298916101455688,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23950
+    },
+    {
+      "epoch": 3.8735752970657185,
+      "grad_norm": 1.0063377618789673,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 23960
+    },
+    {
+      "epoch": 3.8751919812464637,
+      "grad_norm": 0.9230626821517944,
+      "learning_rate": 0.0002,
+      "loss": 0.6234,
+      "step": 23970
+    },
+    {
+      "epoch": 3.876808665427209,
+      "grad_norm": 0.9243063926696777,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 23980
+    },
+    {
+      "epoch": 3.878425349607954,
+      "grad_norm": 1.0211291313171387,
+      "learning_rate": 0.0002,
+      "loss": 0.6035,
+      "step": 23990
+    },
+    {
+      "epoch": 3.8800420337886994,
+      "grad_norm": 0.7800535559654236,
+      "learning_rate": 0.0002,
+      "loss": 0.6351,
+      "step": 24000
+    },
+    {
+      "epoch": 3.8816587179694446,
+      "grad_norm": 0.7904248833656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 24010
+    },
+    {
+      "epoch": 3.88327540215019,
+      "grad_norm": 1.1975988149642944,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 24020
+    },
+    {
+      "epoch": 3.884892086330935,
+      "grad_norm": 1.0626593828201294,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 24030
+    },
+    {
+      "epoch": 3.8865087705116803,
+      "grad_norm": 0.9012193083763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 24040
+    },
+    {
+      "epoch": 3.888125454692426,
+      "grad_norm": 1.1159172058105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6786,
+      "step": 24050
+    },
+    {
+      "epoch": 3.889742138873171,
+      "grad_norm": 1.276838779449463,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 24060
+    },
+    {
+      "epoch": 3.8913588230539164,
+      "grad_norm": 0.8467690348625183,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 24070
+    },
+    {
+      "epoch": 3.8929755072346617,
+      "grad_norm": 0.9862841963768005,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 24080
+    },
+    {
+      "epoch": 3.894592191415407,
+      "grad_norm": 0.7134621739387512,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 24090
+    },
+    {
+      "epoch": 3.896208875596152,
+      "grad_norm": 0.8178175091743469,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 24100
+    },
+    {
+      "epoch": 3.897825559776898,
+      "grad_norm": 0.9229172468185425,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 24110
+    },
+    {
+      "epoch": 3.899442243957643,
+      "grad_norm": 1.0878316164016724,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 24120
+    },
+    {
+      "epoch": 3.9010589281383883,
+      "grad_norm": 0.971645712852478,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24130
+    },
+    {
+      "epoch": 3.9026756123191335,
+      "grad_norm": 0.8862188458442688,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 24140
+    },
+    {
+      "epoch": 3.9042922964998787,
+      "grad_norm": 0.9126982688903809,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 24150
+    },
+    {
+      "epoch": 3.905908980680624,
+      "grad_norm": 0.8833470940589905,
+      "learning_rate": 0.0002,
+      "loss": 0.6734,
+      "step": 24160
+    },
+    {
+      "epoch": 3.907525664861369,
+      "grad_norm": 0.8320947885513306,
+      "learning_rate": 0.0002,
+      "loss": 0.5832,
+      "step": 24170
+    },
+    {
+      "epoch": 3.9091423490421144,
+      "grad_norm": 0.9156602025032043,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 24180
+    },
+    {
+      "epoch": 3.9107590332228597,
+      "grad_norm": 1.029181957244873,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 24190
+    },
+    {
+      "epoch": 3.9123757174036053,
+      "grad_norm": 0.9052802324295044,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 24200
+    },
+    {
+      "epoch": 3.9139924015843506,
+      "grad_norm": 0.8847255110740662,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 24210
+    },
+    {
+      "epoch": 3.915609085765096,
+      "grad_norm": 0.9642062187194824,
+      "learning_rate": 0.0002,
+      "loss": 0.6343,
+      "step": 24220
+    },
+    {
+      "epoch": 3.917225769945841,
+      "grad_norm": 0.8629093766212463,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 24230
+    },
+    {
+      "epoch": 3.9188424541265863,
+      "grad_norm": 0.8674976825714111,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 24240
+    },
+    {
+      "epoch": 3.9204591383073315,
+      "grad_norm": 1.104846477508545,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 24250
+    },
+    {
+      "epoch": 3.922075822488077,
+      "grad_norm": 1.0874955654144287,
+      "learning_rate": 0.0002,
+      "loss": 0.6501,
+      "step": 24260
+    },
+    {
+      "epoch": 3.9236925066688224,
+      "grad_norm": 0.8689812421798706,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 24270
+    },
+    {
+      "epoch": 3.9253091908495676,
+      "grad_norm": 0.9724617004394531,
+      "learning_rate": 0.0002,
+      "loss": 0.5893,
+      "step": 24280
+    },
+    {
+      "epoch": 3.926925875030313,
+      "grad_norm": 0.9165538549423218,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24290
+    },
+    {
+      "epoch": 3.928542559211058,
+      "grad_norm": 0.9307710528373718,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 24300
+    },
+    {
+      "epoch": 3.9301592433918033,
+      "grad_norm": 0.8589295148849487,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 24310
+    },
+    {
+      "epoch": 3.9317759275725486,
+      "grad_norm": 0.9151099920272827,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 24320
+    },
+    {
+      "epoch": 3.933392611753294,
+      "grad_norm": 0.9633517265319824,
+      "learning_rate": 0.0002,
+      "loss": 0.7075,
+      "step": 24330
+    },
+    {
+      "epoch": 3.935009295934039,
+      "grad_norm": 0.9521116018295288,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24340
+    },
+    {
+      "epoch": 3.9366259801147847,
+      "grad_norm": 0.8366776704788208,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 24350
+    },
+    {
+      "epoch": 3.93824266429553,
+      "grad_norm": 0.8972663283348083,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 24360
+    },
+    {
+      "epoch": 3.939859348476275,
+      "grad_norm": 0.8102919459342957,
+      "learning_rate": 0.0002,
+      "loss": 0.661,
+      "step": 24370
+    },
+    {
+      "epoch": 3.9414760326570204,
+      "grad_norm": 0.8189975023269653,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 24380
+    },
+    {
+      "epoch": 3.9430927168377656,
+      "grad_norm": 0.9569464921951294,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 24390
+    },
+    {
+      "epoch": 3.9447094010185113,
+      "grad_norm": 0.7459101676940918,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 24400
+    },
+    {
+      "epoch": 3.9463260851992565,
+      "grad_norm": 0.8536974787712097,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 24410
+    },
+    {
+      "epoch": 3.9479427693800018,
+      "grad_norm": 0.8763698935508728,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 24420
+    },
+    {
+      "epoch": 3.949559453560747,
+      "grad_norm": 0.9381106495857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6478,
+      "step": 24430
+    },
+    {
+      "epoch": 3.9511761377414922,
+      "grad_norm": 0.934440016746521,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 24440
+    },
+    {
+      "epoch": 3.9527928219222375,
+      "grad_norm": 0.903918981552124,
+      "learning_rate": 0.0002,
+      "loss": 0.6393,
+      "step": 24450
+    },
+    {
+      "epoch": 3.9544095061029827,
+      "grad_norm": 0.8771953582763672,
+      "learning_rate": 0.0002,
+      "loss": 0.6175,
+      "step": 24460
+    },
+    {
+      "epoch": 3.956026190283728,
+      "grad_norm": 1.0375410318374634,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 24470
+    },
+    {
+      "epoch": 3.957642874464473,
+      "grad_norm": 0.9439185261726379,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 24480
+    },
+    {
+      "epoch": 3.9592595586452184,
+      "grad_norm": 0.935467004776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 24490
+    },
+    {
+      "epoch": 3.960876242825964,
+      "grad_norm": 0.6900772452354431,
+      "learning_rate": 0.0002,
+      "loss": 0.6437,
+      "step": 24500
+    },
+    {
+      "epoch": 3.9624929270067093,
+      "grad_norm": 1.0172916650772095,
+      "learning_rate": 0.0002,
+      "loss": 0.6445,
+      "step": 24510
+    },
+    {
+      "epoch": 3.9641096111874545,
+      "grad_norm": 0.9167046546936035,
+      "learning_rate": 0.0002,
+      "loss": 0.6308,
+      "step": 24520
+    },
+    {
+      "epoch": 3.9657262953681998,
+      "grad_norm": 0.7230527997016907,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 24530
+    },
+    {
+      "epoch": 3.967342979548945,
+      "grad_norm": 0.8980403542518616,
+      "learning_rate": 0.0002,
+      "loss": 0.6564,
+      "step": 24540
+    },
+    {
+      "epoch": 3.9689596637296907,
+      "grad_norm": 0.8555465936660767,
+      "learning_rate": 0.0002,
+      "loss": 0.6099,
+      "step": 24550
+    },
+    {
+      "epoch": 3.970576347910436,
+      "grad_norm": 0.7825445532798767,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 24560
+    },
+    {
+      "epoch": 3.972193032091181,
+      "grad_norm": 0.7273133993148804,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 24570
+    },
+    {
+      "epoch": 3.9738097162719264,
+      "grad_norm": 0.9612047672271729,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 24580
+    },
+    {
+      "epoch": 3.9754264004526716,
+      "grad_norm": 0.9865460991859436,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 24590
+    },
+    {
+      "epoch": 3.977043084633417,
+      "grad_norm": 0.8638762831687927,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 24600
+    },
+    {
+      "epoch": 3.978659768814162,
+      "grad_norm": 1.0096198320388794,
+      "learning_rate": 0.0002,
+      "loss": 0.6097,
+      "step": 24610
+    },
+    {
+      "epoch": 3.9802764529949073,
+      "grad_norm": 0.8475532531738281,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 24620
+    },
+    {
+      "epoch": 3.9818931371756525,
+      "grad_norm": 0.9696195721626282,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 24630
+    },
+    {
+      "epoch": 3.9835098213563978,
+      "grad_norm": 0.7499843239784241,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 24640
+    },
+    {
+      "epoch": 3.9851265055371434,
+      "grad_norm": 0.8865424990653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6054,
+      "step": 24650
+    },
+    {
+      "epoch": 3.9867431897178887,
+      "grad_norm": 0.8089959025382996,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 24660
+    },
+    {
+      "epoch": 3.988359873898634,
+      "grad_norm": 0.6946012377738953,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 24670
+    },
+    {
+      "epoch": 3.989976558079379,
+      "grad_norm": 0.7991759181022644,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 24680
+    },
+    {
+      "epoch": 3.9915932422601244,
+      "grad_norm": 0.8803931474685669,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 24690
+    },
+    {
+      "epoch": 3.99320992644087,
+      "grad_norm": 0.8848299980163574,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 24700
+    },
+    {
+      "epoch": 3.9948266106216153,
+      "grad_norm": 0.7448889017105103,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 24710
+    },
+    {
+      "epoch": 3.9964432948023605,
+      "grad_norm": 0.9361620545387268,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24720
+    },
+    {
+      "epoch": 3.9980599789831057,
+      "grad_norm": 0.9958081245422363,
+      "learning_rate": 0.0002,
+      "loss": 0.5917,
+      "step": 24730
+    },
+    {
+      "epoch": 3.999676663163851,
+      "grad_norm": 1.026004672050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6567,
+      "step": 24740
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.1524168252944946,
+      "eval_runtime": 122.1585,
+      "eval_samples_per_second": 6.0,
+      "eval_steps_per_second": 0.753,
+      "step": 24742
+    },
+    {
+      "epoch": 4.001293347344596,
+      "grad_norm": 1.0664808750152588,
+      "learning_rate": 0.0002,
+      "loss": 0.6057,
+      "step": 24750
+    },
+    {
+      "epoch": 4.002910031525341,
+      "grad_norm": 1.0113720893859863,
+      "learning_rate": 0.0002,
+      "loss": 0.5644,
+      "step": 24760
+    },
+    {
+      "epoch": 4.004526715706087,
+      "grad_norm": 0.991486668586731,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 24770
+    },
+    {
+      "epoch": 4.006143399886832,
+      "grad_norm": 0.951754629611969,
+      "learning_rate": 0.0002,
+      "loss": 0.508,
+      "step": 24780
+    },
+    {
+      "epoch": 4.007760084067577,
+      "grad_norm": 1.13059401512146,
+      "learning_rate": 0.0002,
+      "loss": 0.5314,
+      "step": 24790
+    },
+    {
+      "epoch": 4.009376768248322,
+      "grad_norm": 0.9343926310539246,
+      "learning_rate": 0.0002,
+      "loss": 0.5323,
+      "step": 24800
+    },
+    {
+      "epoch": 4.010993452429068,
+      "grad_norm": 1.0680590867996216,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 24810
+    },
+    {
+      "epoch": 4.012610136609814,
+      "grad_norm": 1.0022706985473633,
+      "learning_rate": 0.0002,
+      "loss": 0.513,
+      "step": 24820
+    },
+    {
+      "epoch": 4.014226820790559,
+      "grad_norm": 1.0285297632217407,
+      "learning_rate": 0.0002,
+      "loss": 0.543,
+      "step": 24830
+    },
+    {
+      "epoch": 4.015843504971304,
+      "grad_norm": 0.8347002863883972,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 24840
+    },
+    {
+      "epoch": 4.017460189152049,
+      "grad_norm": 0.9675396680831909,
+      "learning_rate": 0.0002,
+      "loss": 0.5655,
+      "step": 24850
+    },
+    {
+      "epoch": 4.019076873332795,
+      "grad_norm": 0.9238511323928833,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 24860
+    },
+    {
+      "epoch": 4.02069355751354,
+      "grad_norm": 1.1576941013336182,
+      "learning_rate": 0.0002,
+      "loss": 0.5327,
+      "step": 24870
+    },
+    {
+      "epoch": 4.022310241694285,
+      "grad_norm": 0.8583757281303406,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 24880
+    },
+    {
+      "epoch": 4.02392692587503,
+      "grad_norm": 0.9816817045211792,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 24890
+    },
+    {
+      "epoch": 4.0255436100557755,
+      "grad_norm": 0.955073893070221,
+      "learning_rate": 0.0002,
+      "loss": 0.5605,
+      "step": 24900
+    },
+    {
+      "epoch": 4.027160294236521,
+      "grad_norm": 1.1054974794387817,
+      "learning_rate": 0.0002,
+      "loss": 0.4896,
+      "step": 24910
+    },
+    {
+      "epoch": 4.028776978417266,
+      "grad_norm": 1.1240060329437256,
+      "learning_rate": 0.0002,
+      "loss": 0.5246,
+      "step": 24920
+    },
+    {
+      "epoch": 4.030393662598011,
+      "grad_norm": 0.9512825012207031,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 24930
+    },
+    {
+      "epoch": 4.0320103467787565,
+      "grad_norm": 0.85965496301651,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 24940
+    },
+    {
+      "epoch": 4.033627030959502,
+      "grad_norm": 0.9378061294555664,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 24950
+    },
+    {
+      "epoch": 4.035243715140247,
+      "grad_norm": 0.9655424356460571,
+      "learning_rate": 0.0002,
+      "loss": 0.5008,
+      "step": 24960
+    },
+    {
+      "epoch": 4.036860399320993,
+      "grad_norm": 1.1393707990646362,
+      "learning_rate": 0.0002,
+      "loss": 0.5538,
+      "step": 24970
+    },
+    {
+      "epoch": 4.038477083501738,
+      "grad_norm": 1.0220451354980469,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 24980
+    },
+    {
+      "epoch": 4.0400937676824835,
+      "grad_norm": 0.9785808324813843,
+      "learning_rate": 0.0002,
+      "loss": 0.5813,
+      "step": 24990
+    },
+    {
+      "epoch": 4.041710451863229,
+      "grad_norm": 1.0257649421691895,
+      "learning_rate": 0.0002,
+      "loss": 0.5153,
+      "step": 25000
+    },
+    {
+      "epoch": 4.043327136043974,
+      "grad_norm": 0.9737892150878906,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 25010
+    },
+    {
+      "epoch": 4.044943820224719,
+      "grad_norm": 0.7416959404945374,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 25020
+    },
+    {
+      "epoch": 4.046560504405464,
+      "grad_norm": 0.7909596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 25030
+    },
+    {
+      "epoch": 4.04817718858621,
+      "grad_norm": 0.8923130631446838,
+      "learning_rate": 0.0002,
+      "loss": 0.5265,
+      "step": 25040
+    },
+    {
+      "epoch": 4.049793872766955,
+      "grad_norm": 0.9044941663742065,
+      "learning_rate": 0.0002,
+      "loss": 0.5035,
+      "step": 25050
+    },
+    {
+      "epoch": 4.0514105569477,
+      "grad_norm": 0.866352379322052,
+      "learning_rate": 0.0002,
+      "loss": 0.5135,
+      "step": 25060
+    },
+    {
+      "epoch": 4.053027241128445,
+      "grad_norm": 1.544549822807312,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 25070
+    },
+    {
+      "epoch": 4.054643925309191,
+      "grad_norm": 0.8426995277404785,
+      "learning_rate": 0.0002,
+      "loss": 0.5418,
+      "step": 25080
+    },
+    {
+      "epoch": 4.056260609489936,
+      "grad_norm": 0.9797548651695251,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 25090
+    },
+    {
+      "epoch": 4.057877293670681,
+      "grad_norm": 0.8468434810638428,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 25100
+    },
+    {
+      "epoch": 4.059493977851426,
+      "grad_norm": 0.9294559955596924,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 25110
+    },
+    {
+      "epoch": 4.061110662032172,
+      "grad_norm": 0.9686688780784607,
+      "learning_rate": 0.0002,
+      "loss": 0.5295,
+      "step": 25120
+    },
+    {
+      "epoch": 4.062727346212918,
+      "grad_norm": 0.8042728304862976,
+      "learning_rate": 0.0002,
+      "loss": 0.5642,
+      "step": 25130
+    },
+    {
+      "epoch": 4.064344030393663,
+      "grad_norm": 1.165160894393921,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 25140
+    },
+    {
+      "epoch": 4.065960714574408,
+      "grad_norm": 1.2161961793899536,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 25150
+    },
+    {
+      "epoch": 4.067577398755153,
+      "grad_norm": 1.0762810707092285,
+      "learning_rate": 0.0002,
+      "loss": 0.5217,
+      "step": 25160
+    },
+    {
+      "epoch": 4.069194082935899,
+      "grad_norm": 0.7580869793891907,
+      "learning_rate": 0.0002,
+      "loss": 0.5886,
+      "step": 25170
+    },
+    {
+      "epoch": 4.070810767116644,
+      "grad_norm": 0.9630117416381836,
+      "learning_rate": 0.0002,
+      "loss": 0.5401,
+      "step": 25180
+    },
+    {
+      "epoch": 4.072427451297389,
+      "grad_norm": 0.9049716591835022,
+      "learning_rate": 0.0002,
+      "loss": 0.5378,
+      "step": 25190
+    },
+    {
+      "epoch": 4.074044135478134,
+      "grad_norm": 1.1536930799484253,
+      "learning_rate": 0.0002,
+      "loss": 0.5266,
+      "step": 25200
+    },
+    {
+      "epoch": 4.0756608196588795,
+      "grad_norm": 0.901461124420166,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 25210
+    },
+    {
+      "epoch": 4.077277503839625,
+      "grad_norm": 1.3318437337875366,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 25220
+    },
+    {
+      "epoch": 4.07889418802037,
+      "grad_norm": 0.8811455368995667,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 25230
+    },
+    {
+      "epoch": 4.080510872201115,
+      "grad_norm": 1.0564165115356445,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 25240
+    },
+    {
+      "epoch": 4.08212755638186,
+      "grad_norm": 1.1008027791976929,
+      "learning_rate": 0.0002,
+      "loss": 0.5472,
+      "step": 25250
+    },
+    {
+      "epoch": 4.083744240562606,
+      "grad_norm": 1.150097131729126,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 25260
+    },
+    {
+      "epoch": 4.085360924743352,
+      "grad_norm": 0.9339924454689026,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 25270
+    },
+    {
+      "epoch": 4.086977608924097,
+      "grad_norm": 1.0902045965194702,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 25280
+    },
+    {
+      "epoch": 4.088594293104842,
+      "grad_norm": 0.8483911156654358,
+      "learning_rate": 0.0002,
+      "loss": 0.5203,
+      "step": 25290
+    },
+    {
+      "epoch": 4.0902109772855875,
+      "grad_norm": 0.9477024674415588,
+      "learning_rate": 0.0002,
+      "loss": 0.5697,
+      "step": 25300
+    },
+    {
+      "epoch": 4.091827661466333,
+      "grad_norm": 0.9500215649604797,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 25310
+    },
+    {
+      "epoch": 4.093444345647078,
+      "grad_norm": 1.040468454360962,
+      "learning_rate": 0.0002,
+      "loss": 0.5045,
+      "step": 25320
+    },
+    {
+      "epoch": 4.095061029827823,
+      "grad_norm": 0.7457592487335205,
+      "learning_rate": 0.0002,
+      "loss": 0.5488,
+      "step": 25330
+    },
+    {
+      "epoch": 4.096677714008568,
+      "grad_norm": 1.2092097997665405,
+      "learning_rate": 0.0002,
+      "loss": 0.609,
+      "step": 25340
+    },
+    {
+      "epoch": 4.098294398189314,
+      "grad_norm": 0.9652107954025269,
+      "learning_rate": 0.0002,
+      "loss": 0.5174,
+      "step": 25350
+    },
+    {
+      "epoch": 4.099911082370059,
+      "grad_norm": 0.8464955687522888,
+      "learning_rate": 0.0002,
+      "loss": 0.5559,
+      "step": 25360
+    },
+    {
+      "epoch": 4.101527766550804,
+      "grad_norm": 0.875026285648346,
+      "learning_rate": 0.0002,
+      "loss": 0.5635,
+      "step": 25370
+    },
+    {
+      "epoch": 4.103144450731549,
+      "grad_norm": 0.9241740107536316,
+      "learning_rate": 0.0002,
+      "loss": 0.5774,
+      "step": 25380
+    },
+    {
+      "epoch": 4.1047611349122946,
+      "grad_norm": 0.9769546389579773,
+      "learning_rate": 0.0002,
+      "loss": 0.5578,
+      "step": 25390
+    },
+    {
+      "epoch": 4.10637781909304,
+      "grad_norm": 1.1501960754394531,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 25400
+    },
+    {
+      "epoch": 4.107994503273786,
+      "grad_norm": 0.9135243892669678,
+      "learning_rate": 0.0002,
+      "loss": 0.5241,
+      "step": 25410
+    },
+    {
+      "epoch": 4.109611187454531,
+      "grad_norm": 0.9905396103858948,
+      "learning_rate": 0.0002,
+      "loss": 0.5152,
+      "step": 25420
+    },
+    {
+      "epoch": 4.111227871635276,
+      "grad_norm": 0.9845104217529297,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 25430
+    },
+    {
+      "epoch": 4.112844555816022,
+      "grad_norm": 0.8326883912086487,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 25440
+    },
+    {
+      "epoch": 4.114461239996767,
+      "grad_norm": 0.9264556765556335,
+      "learning_rate": 0.0002,
+      "loss": 0.5312,
+      "step": 25450
+    },
+    {
+      "epoch": 4.116077924177512,
+      "grad_norm": 1.043080449104309,
+      "learning_rate": 0.0002,
+      "loss": 0.5968,
+      "step": 25460
+    },
+    {
+      "epoch": 4.117694608358257,
+      "grad_norm": 0.8533386588096619,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 25470
+    },
+    {
+      "epoch": 4.1193112925390025,
+      "grad_norm": 1.0133965015411377,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 25480
+    },
+    {
+      "epoch": 4.120927976719748,
+      "grad_norm": 0.7476310133934021,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 25490
+    },
+    {
+      "epoch": 4.122544660900493,
+      "grad_norm": 1.1247259378433228,
+      "learning_rate": 0.0002,
+      "loss": 0.5189,
+      "step": 25500
+    },
+    {
+      "epoch": 4.124161345081238,
+      "grad_norm": 1.0764678716659546,
+      "learning_rate": 0.0002,
+      "loss": 0.5751,
+      "step": 25510
+    },
+    {
+      "epoch": 4.1257780292619834,
+      "grad_norm": 0.7679798007011414,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 25520
+    },
+    {
+      "epoch": 4.127394713442729,
+      "grad_norm": 0.8877071142196655,
+      "learning_rate": 0.0002,
+      "loss": 0.5233,
+      "step": 25530
+    },
+    {
+      "epoch": 4.129011397623474,
+      "grad_norm": 1.0440239906311035,
+      "learning_rate": 0.0002,
+      "loss": 0.5769,
+      "step": 25540
+    },
+    {
+      "epoch": 4.130628081804219,
+      "grad_norm": 0.984145998954773,
+      "learning_rate": 0.0002,
+      "loss": 0.5723,
+      "step": 25550
+    },
+    {
+      "epoch": 4.132244765984965,
+      "grad_norm": 0.8667055368423462,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 25560
+    },
+    {
+      "epoch": 4.1338614501657105,
+      "grad_norm": 1.1300835609436035,
+      "learning_rate": 0.0002,
+      "loss": 0.5816,
+      "step": 25570
+    },
+    {
+      "epoch": 4.135478134346456,
+      "grad_norm": 0.9314348101615906,
+      "learning_rate": 0.0002,
+      "loss": 0.524,
+      "step": 25580
+    },
+    {
+      "epoch": 4.137094818527201,
+      "grad_norm": 0.7731879949569702,
+      "learning_rate": 0.0002,
+      "loss": 0.5283,
+      "step": 25590
+    },
+    {
+      "epoch": 4.138711502707946,
+      "grad_norm": 1.0080097913742065,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 25600
+    },
+    {
+      "epoch": 4.140328186888691,
+      "grad_norm": 1.2475038766860962,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 25610
+    },
+    {
+      "epoch": 4.141944871069437,
+      "grad_norm": 0.9912930727005005,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 25620
+    },
+    {
+      "epoch": 4.143561555250182,
+      "grad_norm": 0.9088651537895203,
+      "learning_rate": 0.0002,
+      "loss": 0.5624,
+      "step": 25630
+    },
+    {
+      "epoch": 4.145178239430927,
+      "grad_norm": 0.8940697312355042,
+      "learning_rate": 0.0002,
+      "loss": 0.5393,
+      "step": 25640
+    },
+    {
+      "epoch": 4.146794923611672,
+      "grad_norm": 1.0798203945159912,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 25650
+    },
+    {
+      "epoch": 4.148411607792418,
+      "grad_norm": 0.955172061920166,
+      "learning_rate": 0.0002,
+      "loss": 0.5987,
+      "step": 25660
+    },
+    {
+      "epoch": 4.150028291973163,
+      "grad_norm": 0.9692716002464294,
+      "learning_rate": 0.0002,
+      "loss": 0.569,
+      "step": 25670
+    },
+    {
+      "epoch": 4.151644976153908,
+      "grad_norm": 1.0813939571380615,
+      "learning_rate": 0.0002,
+      "loss": 0.5478,
+      "step": 25680
+    },
+    {
+      "epoch": 4.153261660334653,
+      "grad_norm": 1.135675072669983,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 25690
+    },
+    {
+      "epoch": 4.1548783445153985,
+      "grad_norm": 1.0392236709594727,
+      "learning_rate": 0.0002,
+      "loss": 0.5247,
+      "step": 25700
+    },
+    {
+      "epoch": 4.156495028696145,
+      "grad_norm": 0.9473116993904114,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 25710
+    },
+    {
+      "epoch": 4.15811171287689,
+      "grad_norm": 0.712493896484375,
+      "learning_rate": 0.0002,
+      "loss": 0.5339,
+      "step": 25720
+    },
+    {
+      "epoch": 4.159728397057635,
+      "grad_norm": 0.8724465370178223,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 25730
+    },
+    {
+      "epoch": 4.16134508123838,
+      "grad_norm": 0.9870015978813171,
+      "learning_rate": 0.0002,
+      "loss": 0.5325,
+      "step": 25740
+    },
+    {
+      "epoch": 4.1629617654191255,
+      "grad_norm": 1.025273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 25750
+    },
+    {
+      "epoch": 4.164578449599871,
+      "grad_norm": 0.9243090152740479,
+      "learning_rate": 0.0002,
+      "loss": 0.5223,
+      "step": 25760
+    },
+    {
+      "epoch": 4.166195133780616,
+      "grad_norm": 1.1656451225280762,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 25770
+    },
+    {
+      "epoch": 4.167811817961361,
+      "grad_norm": 0.936358630657196,
+      "learning_rate": 0.0002,
+      "loss": 0.5334,
+      "step": 25780
+    },
+    {
+      "epoch": 4.1694285021421065,
+      "grad_norm": 0.8618208169937134,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 25790
+    },
+    {
+      "epoch": 4.171045186322852,
+      "grad_norm": 0.8580600023269653,
+      "learning_rate": 0.0002,
+      "loss": 0.5186,
+      "step": 25800
+    },
+    {
+      "epoch": 4.172661870503597,
+      "grad_norm": 1.0128562450408936,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 25810
+    },
+    {
+      "epoch": 4.174278554684342,
+      "grad_norm": 0.854865312576294,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 25820
+    },
+    {
+      "epoch": 4.175895238865087,
+      "grad_norm": 1.235082745552063,
+      "learning_rate": 0.0002,
+      "loss": 0.5377,
+      "step": 25830
+    },
+    {
+      "epoch": 4.177511923045833,
+      "grad_norm": 0.9796220660209656,
+      "learning_rate": 0.0002,
+      "loss": 0.5614,
+      "step": 25840
+    },
+    {
+      "epoch": 4.179128607226578,
+      "grad_norm": 0.8922094702720642,
+      "learning_rate": 0.0002,
+      "loss": 0.5689,
+      "step": 25850
+    },
+    {
+      "epoch": 4.180745291407324,
+      "grad_norm": 0.9672530293464661,
+      "learning_rate": 0.0002,
+      "loss": 0.5806,
+      "step": 25860
+    },
+    {
+      "epoch": 4.182361975588069,
+      "grad_norm": 0.8662548661231995,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 25870
+    },
+    {
+      "epoch": 4.1839786597688144,
+      "grad_norm": 0.7938798069953918,
+      "learning_rate": 0.0002,
+      "loss": 0.5329,
+      "step": 25880
+    },
+    {
+      "epoch": 4.18559534394956,
+      "grad_norm": 1.0517958402633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5427,
+      "step": 25890
+    },
+    {
+      "epoch": 4.187212028130305,
+      "grad_norm": 0.8939275145530701,
+      "learning_rate": 0.0002,
+      "loss": 0.5147,
+      "step": 25900
+    },
+    {
+      "epoch": 4.18882871231105,
+      "grad_norm": 1.0296672582626343,
+      "learning_rate": 0.0002,
+      "loss": 0.5199,
+      "step": 25910
+    },
+    {
+      "epoch": 4.190445396491795,
+      "grad_norm": 0.8104017972946167,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 25920
+    },
+    {
+      "epoch": 4.192062080672541,
+      "grad_norm": 0.9984509944915771,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 25930
+    },
+    {
+      "epoch": 4.193678764853286,
+      "grad_norm": 0.9844784736633301,
+      "learning_rate": 0.0002,
+      "loss": 0.5356,
+      "step": 25940
+    },
+    {
+      "epoch": 4.195295449034031,
+      "grad_norm": 0.8168622255325317,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 25950
+    },
+    {
+      "epoch": 4.196912133214776,
+      "grad_norm": 1.0878913402557373,
+      "learning_rate": 0.0002,
+      "loss": 0.542,
+      "step": 25960
+    },
+    {
+      "epoch": 4.1985288173955215,
+      "grad_norm": 0.927126407623291,
+      "learning_rate": 0.0002,
+      "loss": 0.5414,
+      "step": 25970
+    },
+    {
+      "epoch": 4.200145501576267,
+      "grad_norm": 0.838586688041687,
+      "learning_rate": 0.0002,
+      "loss": 0.5794,
+      "step": 25980
+    },
+    {
+      "epoch": 4.201762185757012,
+      "grad_norm": 1.2572145462036133,
+      "learning_rate": 0.0002,
+      "loss": 0.5454,
+      "step": 25990
+    },
+    {
+      "epoch": 4.203378869937758,
+      "grad_norm": 1.0476740598678589,
+      "learning_rate": 0.0002,
+      "loss": 0.5048,
+      "step": 26000
+    },
+    {
+      "epoch": 4.204995554118503,
+      "grad_norm": 1.0873368978500366,
+      "learning_rate": 0.0002,
+      "loss": 0.5127,
+      "step": 26010
+    },
+    {
+      "epoch": 4.206612238299249,
+      "grad_norm": 1.2664896249771118,
+      "learning_rate": 0.0002,
+      "loss": 0.5679,
+      "step": 26020
+    },
+    {
+      "epoch": 4.208228922479994,
+      "grad_norm": 1.0312391519546509,
+      "learning_rate": 0.0002,
+      "loss": 0.5814,
+      "step": 26030
+    },
+    {
+      "epoch": 4.209845606660739,
+      "grad_norm": 1.0235042572021484,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 26040
+    },
+    {
+      "epoch": 4.211462290841484,
+      "grad_norm": 0.8882219195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 26050
+    },
+    {
+      "epoch": 4.2130789750222295,
+      "grad_norm": 0.9115961790084839,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 26060
+    },
+    {
+      "epoch": 4.214695659202975,
+      "grad_norm": 1.0218228101730347,
+      "learning_rate": 0.0002,
+      "loss": 0.5455,
+      "step": 26070
+    },
+    {
+      "epoch": 4.21631234338372,
+      "grad_norm": 1.0802232027053833,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 26080
+    },
+    {
+      "epoch": 4.217929027564465,
+      "grad_norm": 1.1488053798675537,
+      "learning_rate": 0.0002,
+      "loss": 0.557,
+      "step": 26090
+    },
+    {
+      "epoch": 4.21954571174521,
+      "grad_norm": 1.0487725734710693,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 26100
+    },
+    {
+      "epoch": 4.221162395925956,
+      "grad_norm": 0.9131165742874146,
+      "learning_rate": 0.0002,
+      "loss": 0.5568,
+      "step": 26110
+    },
+    {
+      "epoch": 4.222779080106701,
+      "grad_norm": 0.9012845158576965,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 26120
+    },
+    {
+      "epoch": 4.224395764287446,
+      "grad_norm": 0.8389840126037598,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 26130
+    },
+    {
+      "epoch": 4.226012448468191,
+      "grad_norm": 0.8924660682678223,
+      "learning_rate": 0.0002,
+      "loss": 0.5268,
+      "step": 26140
+    },
+    {
+      "epoch": 4.2276291326489375,
+      "grad_norm": 0.8556463718414307,
+      "learning_rate": 0.0002,
+      "loss": 0.5715,
+      "step": 26150
+    },
+    {
+      "epoch": 4.229245816829683,
+      "grad_norm": 0.9643129110336304,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 26160
+    },
+    {
+      "epoch": 4.230862501010428,
+      "grad_norm": 0.9865712523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 26170
+    },
+    {
+      "epoch": 4.232479185191173,
+      "grad_norm": 1.152641773223877,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 26180
+    },
+    {
+      "epoch": 4.234095869371918,
+      "grad_norm": 0.9157698154449463,
+      "learning_rate": 0.0002,
+      "loss": 0.5632,
+      "step": 26190
+    },
+    {
+      "epoch": 4.235712553552664,
+      "grad_norm": 0.8418048620223999,
+      "learning_rate": 0.0002,
+      "loss": 0.5717,
+      "step": 26200
+    },
+    {
+      "epoch": 4.237329237733409,
+      "grad_norm": 0.9430168867111206,
+      "learning_rate": 0.0002,
+      "loss": 0.5624,
+      "step": 26210
+    },
+    {
+      "epoch": 4.238945921914154,
+      "grad_norm": 1.012582778930664,
+      "learning_rate": 0.0002,
+      "loss": 0.5574,
+      "step": 26220
+    },
+    {
+      "epoch": 4.240562606094899,
+      "grad_norm": 1.112619400024414,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 26230
+    },
+    {
+      "epoch": 4.2421792902756446,
+      "grad_norm": 0.9243621826171875,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 26240
+    },
+    {
+      "epoch": 4.24379597445639,
+      "grad_norm": 0.6977595686912537,
+      "learning_rate": 0.0002,
+      "loss": 0.569,
+      "step": 26250
+    },
+    {
+      "epoch": 4.245412658637135,
+      "grad_norm": 0.9600721597671509,
+      "learning_rate": 0.0002,
+      "loss": 0.5379,
+      "step": 26260
+    },
+    {
+      "epoch": 4.24702934281788,
+      "grad_norm": 0.882641613483429,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 26270
+    },
+    {
+      "epoch": 4.2486460269986255,
+      "grad_norm": 1.010920763015747,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 26280
+    },
+    {
+      "epoch": 4.250262711179371,
+      "grad_norm": 0.9289400577545166,
+      "learning_rate": 0.0002,
+      "loss": 0.5803,
+      "step": 26290
+    },
+    {
+      "epoch": 4.251879395360117,
+      "grad_norm": 1.137397289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.541,
+      "step": 26300
+    },
+    {
+      "epoch": 4.253496079540862,
+      "grad_norm": 1.0136182308197021,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 26310
+    },
+    {
+      "epoch": 4.255112763721607,
+      "grad_norm": 0.9387356042861938,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 26320
+    },
+    {
+      "epoch": 4.2567294479023525,
+      "grad_norm": 1.1833957433700562,
+      "learning_rate": 0.0002,
+      "loss": 0.5948,
+      "step": 26330
+    },
+    {
+      "epoch": 4.258346132083098,
+      "grad_norm": 0.9415934681892395,
+      "learning_rate": 0.0002,
+      "loss": 0.5905,
+      "step": 26340
+    },
+    {
+      "epoch": 4.259962816263843,
+      "grad_norm": 0.8550165891647339,
+      "learning_rate": 0.0002,
+      "loss": 0.5539,
+      "step": 26350
+    },
+    {
+      "epoch": 4.261579500444588,
+      "grad_norm": 9.924622535705566,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 26360
+    },
+    {
+      "epoch": 4.2631961846253335,
+      "grad_norm": 1.0104902982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.5689,
+      "step": 26370
+    },
+    {
+      "epoch": 4.264812868806079,
+      "grad_norm": 0.890794038772583,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 26380
+    },
+    {
+      "epoch": 4.266429552986824,
+      "grad_norm": 1.0560191869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.563,
+      "step": 26390
+    },
+    {
+      "epoch": 4.268046237167569,
+      "grad_norm": 1.0135581493377686,
+      "learning_rate": 0.0002,
+      "loss": 0.5119,
+      "step": 26400
+    },
+    {
+      "epoch": 4.269662921348314,
+      "grad_norm": 1.1304140090942383,
+      "learning_rate": 0.0002,
+      "loss": 0.5359,
+      "step": 26410
+    },
+    {
+      "epoch": 4.27127960552906,
+      "grad_norm": 0.9899303913116455,
+      "learning_rate": 0.0002,
+      "loss": 0.5615,
+      "step": 26420
+    },
+    {
+      "epoch": 4.272896289709805,
+      "grad_norm": 1.0505329370498657,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 26430
+    },
+    {
+      "epoch": 4.27451297389055,
+      "grad_norm": 0.9389396905899048,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 26440
+    },
+    {
+      "epoch": 4.276129658071296,
+      "grad_norm": 0.875328779220581,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 26450
+    },
+    {
+      "epoch": 4.277746342252041,
+      "grad_norm": 1.0689256191253662,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 26460
+    },
+    {
+      "epoch": 4.279363026432787,
+      "grad_norm": 0.9988957643508911,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 26470
+    },
+    {
+      "epoch": 4.280979710613532,
+      "grad_norm": 0.8721813559532166,
+      "learning_rate": 0.0002,
+      "loss": 0.5478,
+      "step": 26480
+    },
+    {
+      "epoch": 4.282596394794277,
+      "grad_norm": 1.100109577178955,
+      "learning_rate": 0.0002,
+      "loss": 0.5424,
+      "step": 26490
+    },
+    {
+      "epoch": 4.284213078975022,
+      "grad_norm": 1.1607271432876587,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 26500
+    },
+    {
+      "epoch": 4.285829763155768,
+      "grad_norm": 0.879088819026947,
+      "learning_rate": 0.0002,
+      "loss": 0.6287,
+      "step": 26510
+    },
+    {
+      "epoch": 4.287446447336513,
+      "grad_norm": 0.9891700744628906,
+      "learning_rate": 0.0002,
+      "loss": 0.573,
+      "step": 26520
+    },
+    {
+      "epoch": 4.289063131517258,
+      "grad_norm": 1.0831127166748047,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 26530
+    },
+    {
+      "epoch": 4.290679815698003,
+      "grad_norm": 1.4108285903930664,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 26540
+    },
+    {
+      "epoch": 4.2922964998787485,
+      "grad_norm": 1.0630289316177368,
+      "learning_rate": 0.0002,
+      "loss": 0.5888,
+      "step": 26550
+    },
+    {
+      "epoch": 4.293913184059494,
+      "grad_norm": 1.0854572057724,
+      "learning_rate": 0.0002,
+      "loss": 0.5817,
+      "step": 26560
+    },
+    {
+      "epoch": 4.295529868240239,
+      "grad_norm": 0.9561646580696106,
+      "learning_rate": 0.0002,
+      "loss": 0.5586,
+      "step": 26570
+    },
+    {
+      "epoch": 4.297146552420984,
+      "grad_norm": 0.9064981937408447,
+      "learning_rate": 0.0002,
+      "loss": 0.5674,
+      "step": 26580
+    },
+    {
+      "epoch": 4.298763236601729,
+      "grad_norm": 1.0082972049713135,
+      "learning_rate": 0.0002,
+      "loss": 0.5847,
+      "step": 26590
+    },
+    {
+      "epoch": 4.3003799207824756,
+      "grad_norm": 1.1613214015960693,
+      "learning_rate": 0.0002,
+      "loss": 0.5711,
+      "step": 26600
+    },
+    {
+      "epoch": 4.301996604963221,
+      "grad_norm": 0.9847695231437683,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 26610
+    },
+    {
+      "epoch": 4.303613289143966,
+      "grad_norm": 1.0980697870254517,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 26620
+    },
+    {
+      "epoch": 4.305229973324711,
+      "grad_norm": 0.8861175179481506,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 26630
+    },
+    {
+      "epoch": 4.3068466575054565,
+      "grad_norm": 0.8917363286018372,
+      "learning_rate": 0.0002,
+      "loss": 0.5716,
+      "step": 26640
+    },
+    {
+      "epoch": 4.308463341686202,
+      "grad_norm": 1.0458378791809082,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 26650
+    },
+    {
+      "epoch": 4.310080025866947,
+      "grad_norm": 1.4859240055084229,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 26660
+    },
+    {
+      "epoch": 4.311696710047692,
+      "grad_norm": 1.1376359462738037,
+      "learning_rate": 0.0002,
+      "loss": 0.5296,
+      "step": 26670
+    },
+    {
+      "epoch": 4.313313394228437,
+      "grad_norm": 0.991349995136261,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 26680
+    },
+    {
+      "epoch": 4.314930078409183,
+      "grad_norm": 0.9995543956756592,
+      "learning_rate": 0.0002,
+      "loss": 0.5338,
+      "step": 26690
+    },
+    {
+      "epoch": 4.316546762589928,
+      "grad_norm": 1.0515851974487305,
+      "learning_rate": 0.0002,
+      "loss": 0.5542,
+      "step": 26700
+    },
+    {
+      "epoch": 4.318163446770673,
+      "grad_norm": 1.008023977279663,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 26710
+    },
+    {
+      "epoch": 4.319780130951418,
+      "grad_norm": 1.0184582471847534,
+      "learning_rate": 0.0002,
+      "loss": 0.5506,
+      "step": 26720
+    },
+    {
+      "epoch": 4.321396815132164,
+      "grad_norm": 1.161071538925171,
+      "learning_rate": 0.0002,
+      "loss": 0.5828,
+      "step": 26730
+    },
+    {
+      "epoch": 4.323013499312909,
+      "grad_norm": 0.9580779671669006,
+      "learning_rate": 0.0002,
+      "loss": 0.5633,
+      "step": 26740
+    },
+    {
+      "epoch": 4.324630183493655,
+      "grad_norm": 1.0189911127090454,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 26750
+    },
+    {
+      "epoch": 4.3262468676744,
+      "grad_norm": 0.7484358549118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 26760
+    },
+    {
+      "epoch": 4.327863551855145,
+      "grad_norm": 1.0015908479690552,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 26770
+    },
+    {
+      "epoch": 4.329480236035891,
+      "grad_norm": 0.8972945809364319,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 26780
+    },
+    {
+      "epoch": 4.331096920216636,
+      "grad_norm": 1.01099693775177,
+      "learning_rate": 0.0002,
+      "loss": 0.5857,
+      "step": 26790
+    },
+    {
+      "epoch": 4.332713604397381,
+      "grad_norm": 0.846958339214325,
+      "learning_rate": 0.0002,
+      "loss": 0.5591,
+      "step": 26800
+    },
+    {
+      "epoch": 4.334330288578126,
+      "grad_norm": 1.0792603492736816,
+      "learning_rate": 0.0002,
+      "loss": 0.5547,
+      "step": 26810
+    },
+    {
+      "epoch": 4.3359469727588715,
+      "grad_norm": 1.0373345613479614,
+      "learning_rate": 0.0002,
+      "loss": 0.5747,
+      "step": 26820
+    },
+    {
+      "epoch": 4.337563656939617,
+      "grad_norm": 0.9779167771339417,
+      "learning_rate": 0.0002,
+      "loss": 0.558,
+      "step": 26830
+    },
+    {
+      "epoch": 4.339180341120362,
+      "grad_norm": 1.0235520601272583,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 26840
+    },
+    {
+      "epoch": 4.340797025301107,
+      "grad_norm": 1.04195237159729,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 26850
+    },
+    {
+      "epoch": 4.3424137094818525,
+      "grad_norm": 0.9479565620422363,
+      "learning_rate": 0.0002,
+      "loss": 0.5474,
+      "step": 26860
+    },
+    {
+      "epoch": 4.344030393662598,
+      "grad_norm": 0.9526172280311584,
+      "learning_rate": 0.0002,
+      "loss": 0.5646,
+      "step": 26870
+    },
+    {
+      "epoch": 4.345647077843343,
+      "grad_norm": 0.8571456074714661,
+      "learning_rate": 0.0002,
+      "loss": 0.521,
+      "step": 26880
+    },
+    {
+      "epoch": 4.347263762024088,
+      "grad_norm": 0.9475828409194946,
+      "learning_rate": 0.0002,
+      "loss": 0.5846,
+      "step": 26890
+    },
+    {
+      "epoch": 4.348880446204834,
+      "grad_norm": 1.0529576539993286,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 26900
+    },
+    {
+      "epoch": 4.3504971303855795,
+      "grad_norm": 0.9648140072822571,
+      "learning_rate": 0.0002,
+      "loss": 0.56,
+      "step": 26910
+    },
+    {
+      "epoch": 4.352113814566325,
+      "grad_norm": 1.0488841533660889,
+      "learning_rate": 0.0002,
+      "loss": 0.5162,
+      "step": 26920
+    },
+    {
+      "epoch": 4.35373049874707,
+      "grad_norm": 0.8771942257881165,
+      "learning_rate": 0.0002,
+      "loss": 0.5842,
+      "step": 26930
+    },
+    {
+      "epoch": 4.355347182927815,
+      "grad_norm": 0.9411202073097229,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 26940
+    },
+    {
+      "epoch": 4.35696386710856,
+      "grad_norm": 1.0997588634490967,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 26950
+    },
+    {
+      "epoch": 4.358580551289306,
+      "grad_norm": 0.968754768371582,
+      "learning_rate": 0.0002,
+      "loss": 0.5528,
+      "step": 26960
+    },
+    {
+      "epoch": 4.360197235470051,
+      "grad_norm": 0.9990773797035217,
+      "learning_rate": 0.0002,
+      "loss": 0.5881,
+      "step": 26970
+    },
+    {
+      "epoch": 4.361813919650796,
+      "grad_norm": 1.0210620164871216,
+      "learning_rate": 0.0002,
+      "loss": 0.5761,
+      "step": 26980
+    },
+    {
+      "epoch": 4.363430603831541,
+      "grad_norm": 0.855462908744812,
+      "learning_rate": 0.0002,
+      "loss": 0.5768,
+      "step": 26990
+    },
+    {
+      "epoch": 4.365047288012287,
+      "grad_norm": 0.9169660806655884,
+      "learning_rate": 0.0002,
+      "loss": 0.5493,
+      "step": 27000
+    },
+    {
+      "epoch": 4.366663972193032,
+      "grad_norm": 1.089629888534546,
+      "learning_rate": 0.0002,
+      "loss": 0.5697,
+      "step": 27010
+    },
+    {
+      "epoch": 4.368280656373777,
+      "grad_norm": 1.0932867527008057,
+      "learning_rate": 0.0002,
+      "loss": 0.5854,
+      "step": 27020
+    },
+    {
+      "epoch": 4.369897340554522,
+      "grad_norm": 0.9290956854820251,
+      "learning_rate": 0.0002,
+      "loss": 0.5656,
+      "step": 27030
+    },
+    {
+      "epoch": 4.3715140247352675,
+      "grad_norm": 1.2800624370574951,
+      "learning_rate": 0.0002,
+      "loss": 0.5727,
+      "step": 27040
+    },
+    {
+      "epoch": 4.373130708916014,
+      "grad_norm": 0.8993493318557739,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 27050
+    },
+    {
+      "epoch": 4.374747393096759,
+      "grad_norm": 1.1566431522369385,
+      "learning_rate": 0.0002,
+      "loss": 0.6232,
+      "step": 27060
+    },
+    {
+      "epoch": 4.376364077277504,
+      "grad_norm": 0.9479052424430847,
+      "learning_rate": 0.0002,
+      "loss": 0.5902,
+      "step": 27070
+    },
+    {
+      "epoch": 4.377980761458249,
+      "grad_norm": 1.0063648223876953,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 27080
+    },
+    {
+      "epoch": 4.379597445638995,
+      "grad_norm": 0.8342045545578003,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 27090
+    },
+    {
+      "epoch": 4.38121412981974,
+      "grad_norm": 1.1390739679336548,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 27100
+    },
+    {
+      "epoch": 4.382830814000485,
+      "grad_norm": 0.9547637104988098,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 27110
+    },
+    {
+      "epoch": 4.38444749818123,
+      "grad_norm": 1.0503804683685303,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 27120
+    },
+    {
+      "epoch": 4.3860641823619755,
+      "grad_norm": 0.9064017534255981,
+      "learning_rate": 0.0002,
+      "loss": 0.5787,
+      "step": 27130
+    },
+    {
+      "epoch": 4.387680866542721,
+      "grad_norm": 0.9382519125938416,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 27140
+    },
+    {
+      "epoch": 4.389297550723466,
+      "grad_norm": 1.0410341024398804,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 27150
+    },
+    {
+      "epoch": 4.390914234904211,
+      "grad_norm": 0.9218655824661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6034,
+      "step": 27160
+    },
+    {
+      "epoch": 4.392530919084956,
+      "grad_norm": 0.8119737505912781,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 27170
+    },
+    {
+      "epoch": 4.394147603265702,
+      "grad_norm": 0.8584722876548767,
+      "learning_rate": 0.0002,
+      "loss": 0.5612,
+      "step": 27180
+    },
+    {
+      "epoch": 4.395764287446447,
+      "grad_norm": 0.9668293595314026,
+      "learning_rate": 0.0002,
+      "loss": 0.5772,
+      "step": 27190
+    },
+    {
+      "epoch": 4.397380971627193,
+      "grad_norm": 1.022334098815918,
+      "learning_rate": 0.0002,
+      "loss": 0.6009,
+      "step": 27200
+    },
+    {
+      "epoch": 4.398997655807938,
+      "grad_norm": 0.9553216099739075,
+      "learning_rate": 0.0002,
+      "loss": 0.5573,
+      "step": 27210
+    },
+    {
+      "epoch": 4.4006143399886835,
+      "grad_norm": 0.9282339215278625,
+      "learning_rate": 0.0002,
+      "loss": 0.5604,
+      "step": 27220
+    },
+    {
+      "epoch": 4.402231024169429,
+      "grad_norm": 1.0232292413711548,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 27230
+    },
+    {
+      "epoch": 4.403847708350174,
+      "grad_norm": 0.9915700554847717,
+      "learning_rate": 0.0002,
+      "loss": 0.6078,
+      "step": 27240
+    },
+    {
+      "epoch": 4.405464392530919,
+      "grad_norm": 1.0014961957931519,
+      "learning_rate": 0.0002,
+      "loss": 0.5778,
+      "step": 27250
+    },
+    {
+      "epoch": 4.407081076711664,
+      "grad_norm": 1.1172103881835938,
+      "learning_rate": 0.0002,
+      "loss": 0.5824,
+      "step": 27260
+    },
+    {
+      "epoch": 4.40869776089241,
+      "grad_norm": 0.8583093285560608,
+      "learning_rate": 0.0002,
+      "loss": 0.5286,
+      "step": 27270
+    },
+    {
+      "epoch": 4.410314445073155,
+      "grad_norm": 0.7609201669692993,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 27280
+    },
+    {
+      "epoch": 4.4119311292539,
+      "grad_norm": 1.0619351863861084,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 27290
+    },
+    {
+      "epoch": 4.413547813434645,
+      "grad_norm": 1.0177674293518066,
+      "learning_rate": 0.0002,
+      "loss": 0.5579,
+      "step": 27300
+    },
+    {
+      "epoch": 4.4151644976153905,
+      "grad_norm": 0.9921218156814575,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 27310
+    },
+    {
+      "epoch": 4.416781181796136,
+      "grad_norm": 1.126244306564331,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 27320
+    },
+    {
+      "epoch": 4.418397865976881,
+      "grad_norm": 1.0678540468215942,
+      "learning_rate": 0.0002,
+      "loss": 0.5743,
+      "step": 27330
+    },
+    {
+      "epoch": 4.420014550157627,
+      "grad_norm": 0.8705704212188721,
+      "learning_rate": 0.0002,
+      "loss": 0.5665,
+      "step": 27340
+    },
+    {
+      "epoch": 4.421631234338372,
+      "grad_norm": 1.272074818611145,
+      "learning_rate": 0.0002,
+      "loss": 0.5763,
+      "step": 27350
+    },
+    {
+      "epoch": 4.423247918519118,
+      "grad_norm": 0.8740444183349609,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 27360
+    },
+    {
+      "epoch": 4.424864602699863,
+      "grad_norm": 1.0584250688552856,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 27370
+    },
+    {
+      "epoch": 4.426481286880608,
+      "grad_norm": 1.059870719909668,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 27380
+    },
+    {
+      "epoch": 4.428097971061353,
+      "grad_norm": 1.072265863418579,
+      "learning_rate": 0.0002,
+      "loss": 0.5551,
+      "step": 27390
+    },
+    {
+      "epoch": 4.4297146552420985,
+      "grad_norm": 0.871481716632843,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 27400
+    },
+    {
+      "epoch": 4.431331339422844,
+      "grad_norm": 0.9555448293685913,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 27410
+    },
+    {
+      "epoch": 4.432948023603589,
+      "grad_norm": 1.0402292013168335,
+      "learning_rate": 0.0002,
+      "loss": 0.5593,
+      "step": 27420
+    },
+    {
+      "epoch": 4.434564707784334,
+      "grad_norm": 1.12587571144104,
+      "learning_rate": 0.0002,
+      "loss": 0.5532,
+      "step": 27430
+    },
+    {
+      "epoch": 4.436181391965079,
+      "grad_norm": 1.0783193111419678,
+      "learning_rate": 0.0002,
+      "loss": 0.5403,
+      "step": 27440
+    },
+    {
+      "epoch": 4.437798076145825,
+      "grad_norm": 1.024133563041687,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 27450
+    },
+    {
+      "epoch": 4.43941476032657,
+      "grad_norm": 0.9156768918037415,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 27460
+    },
+    {
+      "epoch": 4.441031444507315,
+      "grad_norm": 1.0215224027633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 27470
+    },
+    {
+      "epoch": 4.442648128688061,
+      "grad_norm": 1.082116961479187,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 27480
+    },
+    {
+      "epoch": 4.4442648128688065,
+      "grad_norm": 1.0412873029708862,
+      "learning_rate": 0.0002,
+      "loss": 0.6203,
+      "step": 27490
+    },
+    {
+      "epoch": 4.445881497049552,
+      "grad_norm": 1.0509289503097534,
+      "learning_rate": 0.0002,
+      "loss": 0.5939,
+      "step": 27500
+    },
+    {
+      "epoch": 4.447498181230297,
+      "grad_norm": 0.9291498064994812,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 27510
+    },
+    {
+      "epoch": 4.449114865411042,
+      "grad_norm": 0.970184326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.5408,
+      "step": 27520
+    },
+    {
+      "epoch": 4.450731549591787,
+      "grad_norm": 0.8418883681297302,
+      "learning_rate": 0.0002,
+      "loss": 0.5705,
+      "step": 27530
+    },
+    {
+      "epoch": 4.452348233772533,
+      "grad_norm": 0.8823825120925903,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 27540
+    },
+    {
+      "epoch": 4.453964917953278,
+      "grad_norm": 1.1909019947052002,
+      "learning_rate": 0.0002,
+      "loss": 0.5867,
+      "step": 27550
+    },
+    {
+      "epoch": 4.455581602134023,
+      "grad_norm": 1.0317302942276,
+      "learning_rate": 0.0002,
+      "loss": 0.5685,
+      "step": 27560
+    },
+    {
+      "epoch": 4.457198286314768,
+      "grad_norm": 0.9977751970291138,
+      "learning_rate": 0.0002,
+      "loss": 0.5538,
+      "step": 27570
+    },
+    {
+      "epoch": 4.458814970495514,
+      "grad_norm": 0.8909519910812378,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 27580
+    },
+    {
+      "epoch": 4.460431654676259,
+      "grad_norm": 0.8653029799461365,
+      "learning_rate": 0.0002,
+      "loss": 0.6099,
+      "step": 27590
+    },
+    {
+      "epoch": 4.462048338857004,
+      "grad_norm": 1.0783653259277344,
+      "learning_rate": 0.0002,
+      "loss": 0.5622,
+      "step": 27600
+    },
+    {
+      "epoch": 4.463665023037749,
+      "grad_norm": 1.1235394477844238,
+      "learning_rate": 0.0002,
+      "loss": 0.579,
+      "step": 27610
+    },
+    {
+      "epoch": 4.4652817072184945,
+      "grad_norm": 0.9386643767356873,
+      "learning_rate": 0.0002,
+      "loss": 0.5545,
+      "step": 27620
+    },
+    {
+      "epoch": 4.466898391399241,
+      "grad_norm": 1.0605148077011108,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 27630
+    },
+    {
+      "epoch": 4.468515075579986,
+      "grad_norm": 1.1283893585205078,
+      "learning_rate": 0.0002,
+      "loss": 0.5886,
+      "step": 27640
+    },
+    {
+      "epoch": 4.470131759760731,
+      "grad_norm": 1.0583468675613403,
+      "learning_rate": 0.0002,
+      "loss": 0.5801,
+      "step": 27650
+    },
+    {
+      "epoch": 4.471748443941476,
+      "grad_norm": 0.9563992023468018,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 27660
+    },
+    {
+      "epoch": 4.4733651281222215,
+      "grad_norm": 1.100598931312561,
+      "learning_rate": 0.0002,
+      "loss": 0.5687,
+      "step": 27670
+    },
+    {
+      "epoch": 4.474981812302967,
+      "grad_norm": 0.9386957287788391,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 27680
+    },
+    {
+      "epoch": 4.476598496483712,
+      "grad_norm": 1.2946288585662842,
+      "learning_rate": 0.0002,
+      "loss": 0.6241,
+      "step": 27690
+    },
+    {
+      "epoch": 4.478215180664457,
+      "grad_norm": 1.0325199365615845,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 27700
+    },
+    {
+      "epoch": 4.4798318648452025,
+      "grad_norm": 1.0318928956985474,
+      "learning_rate": 0.0002,
+      "loss": 0.588,
+      "step": 27710
+    },
+    {
+      "epoch": 4.481448549025948,
+      "grad_norm": 0.8721024394035339,
+      "learning_rate": 0.0002,
+      "loss": 0.5656,
+      "step": 27720
+    },
+    {
+      "epoch": 4.483065233206693,
+      "grad_norm": 1.17376708984375,
+      "learning_rate": 0.0002,
+      "loss": 0.5421,
+      "step": 27730
+    },
+    {
+      "epoch": 4.484681917387438,
+      "grad_norm": 1.0926326513290405,
+      "learning_rate": 0.0002,
+      "loss": 0.5657,
+      "step": 27740
+    },
+    {
+      "epoch": 4.486298601568183,
+      "grad_norm": 0.9043852686882019,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 27750
+    },
+    {
+      "epoch": 4.487915285748929,
+      "grad_norm": 1.064600944519043,
+      "learning_rate": 0.0002,
+      "loss": 0.582,
+      "step": 27760
+    },
+    {
+      "epoch": 4.489531969929674,
+      "grad_norm": 0.7833460569381714,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 27770
+    },
+    {
+      "epoch": 4.49114865411042,
+      "grad_norm": 1.1073496341705322,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 27780
+    },
+    {
+      "epoch": 4.492765338291165,
+      "grad_norm": 1.0799397230148315,
+      "learning_rate": 0.0002,
+      "loss": 0.5577,
+      "step": 27790
+    },
+    {
+      "epoch": 4.49438202247191,
+      "grad_norm": 1.1062238216400146,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 27800
+    },
+    {
+      "epoch": 4.495998706652656,
+      "grad_norm": 1.0568242073059082,
+      "learning_rate": 0.0002,
+      "loss": 0.6126,
+      "step": 27810
+    },
+    {
+      "epoch": 4.497615390833401,
+      "grad_norm": 0.8861091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 27820
+    },
+    {
+      "epoch": 4.499232075014146,
+      "grad_norm": 1.2297543287277222,
+      "learning_rate": 0.0002,
+      "loss": 0.5858,
+      "step": 27830
+    },
+    {
+      "epoch": 4.500848759194891,
+      "grad_norm": 0.9600302577018738,
+      "learning_rate": 0.0002,
+      "loss": 0.5859,
+      "step": 27840
+    },
+    {
+      "epoch": 4.502465443375637,
+      "grad_norm": 1.057051181793213,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 27850
+    },
+    {
+      "epoch": 4.504082127556382,
+      "grad_norm": 0.9839690923690796,
+      "learning_rate": 0.0002,
+      "loss": 0.5788,
+      "step": 27860
+    },
+    {
+      "epoch": 4.505698811737127,
+      "grad_norm": 1.1479853391647339,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 27870
+    },
+    {
+      "epoch": 4.507315495917872,
+      "grad_norm": 1.0550768375396729,
+      "learning_rate": 0.0002,
+      "loss": 0.6039,
+      "step": 27880
+    },
+    {
+      "epoch": 4.5089321800986175,
+      "grad_norm": 0.898209273815155,
+      "learning_rate": 0.0002,
+      "loss": 0.563,
+      "step": 27890
+    },
+    {
+      "epoch": 4.510548864279363,
+      "grad_norm": 0.9460315108299255,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 27900
+    },
+    {
+      "epoch": 4.512165548460108,
+      "grad_norm": 0.9499884247779846,
+      "learning_rate": 0.0002,
+      "loss": 0.5702,
+      "step": 27910
+    },
+    {
+      "epoch": 4.513782232640853,
+      "grad_norm": 0.7801318764686584,
+      "learning_rate": 0.0002,
+      "loss": 0.5385,
+      "step": 27920
+    },
+    {
+      "epoch": 4.515398916821599,
+      "grad_norm": 0.9286966323852539,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 27930
+    },
+    {
+      "epoch": 4.517015601002345,
+      "grad_norm": 0.9539980292320251,
+      "learning_rate": 0.0002,
+      "loss": 0.5717,
+      "step": 27940
+    },
+    {
+      "epoch": 4.51863228518309,
+      "grad_norm": 1.1053401231765747,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 27950
+    },
+    {
+      "epoch": 4.520248969363835,
+      "grad_norm": 0.7535534501075745,
+      "learning_rate": 0.0002,
+      "loss": 0.6087,
+      "step": 27960
+    },
+    {
+      "epoch": 4.52186565354458,
+      "grad_norm": 1.076926589012146,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 27970
+    },
+    {
+      "epoch": 4.5234823377253255,
+      "grad_norm": 1.181935429573059,
+      "learning_rate": 0.0002,
+      "loss": 0.6028,
+      "step": 27980
+    },
+    {
+      "epoch": 4.525099021906071,
+      "grad_norm": 0.9293407201766968,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 27990
+    },
+    {
+      "epoch": 4.526715706086816,
+      "grad_norm": 0.8953009247779846,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 28000
+    },
+    {
+      "epoch": 4.528332390267561,
+      "grad_norm": 1.0850225687026978,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 28010
+    },
+    {
+      "epoch": 4.529949074448306,
+      "grad_norm": 0.9125663042068481,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 28020
+    },
+    {
+      "epoch": 4.531565758629052,
+      "grad_norm": 0.8745216727256775,
+      "learning_rate": 0.0002,
+      "loss": 0.5922,
+      "step": 28030
+    },
+    {
+      "epoch": 4.533182442809797,
+      "grad_norm": 1.0783463716506958,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 28040
+    },
+    {
+      "epoch": 4.534799126990542,
+      "grad_norm": 0.7513844966888428,
+      "learning_rate": 0.0002,
+      "loss": 0.5754,
+      "step": 28050
+    },
+    {
+      "epoch": 4.536415811171287,
+      "grad_norm": 1.0135776996612549,
+      "learning_rate": 0.0002,
+      "loss": 0.5608,
+      "step": 28060
+    },
+    {
+      "epoch": 4.538032495352033,
+      "grad_norm": 0.8886825442314148,
+      "learning_rate": 0.0002,
+      "loss": 0.5827,
+      "step": 28070
+    },
+    {
+      "epoch": 4.539649179532779,
+      "grad_norm": 0.8153995275497437,
+      "learning_rate": 0.0002,
+      "loss": 0.5605,
+      "step": 28080
+    },
+    {
+      "epoch": 4.541265863713524,
+      "grad_norm": 0.9853341579437256,
+      "learning_rate": 0.0002,
+      "loss": 0.6377,
+      "step": 28090
+    },
+    {
+      "epoch": 4.542882547894269,
+      "grad_norm": 0.9365800023078918,
+      "learning_rate": 0.0002,
+      "loss": 0.5957,
+      "step": 28100
+    },
+    {
+      "epoch": 4.544499232075014,
+      "grad_norm": 0.9765017628669739,
+      "learning_rate": 0.0002,
+      "loss": 0.5477,
+      "step": 28110
+    },
+    {
+      "epoch": 4.54611591625576,
+      "grad_norm": 0.9811279773712158,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 28120
+    },
+    {
+      "epoch": 4.547732600436505,
+      "grad_norm": 1.0387924909591675,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 28130
+    },
+    {
+      "epoch": 4.54934928461725,
+      "grad_norm": 1.0684878826141357,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 28140
+    },
+    {
+      "epoch": 4.550965968797995,
+      "grad_norm": 1.0000102519989014,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 28150
+    },
+    {
+      "epoch": 4.5525826529787405,
+      "grad_norm": 1.0717930793762207,
+      "learning_rate": 0.0002,
+      "loss": 0.5327,
+      "step": 28160
+    },
+    {
+      "epoch": 4.554199337159486,
+      "grad_norm": 0.990074634552002,
+      "learning_rate": 0.0002,
+      "loss": 0.5594,
+      "step": 28170
+    },
+    {
+      "epoch": 4.555816021340231,
+      "grad_norm": 0.8673754930496216,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 28180
+    },
+    {
+      "epoch": 4.557432705520976,
+      "grad_norm": 0.864247739315033,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 28190
+    },
+    {
+      "epoch": 4.5590493897017215,
+      "grad_norm": 0.8280200958251953,
+      "learning_rate": 0.0002,
+      "loss": 0.5516,
+      "step": 28200
+    },
+    {
+      "epoch": 4.560666073882467,
+      "grad_norm": 1.1312172412872314,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 28210
+    },
+    {
+      "epoch": 4.562282758063212,
+      "grad_norm": 0.9147403240203857,
+      "learning_rate": 0.0002,
+      "loss": 0.5776,
+      "step": 28220
+    },
+    {
+      "epoch": 4.563899442243958,
+      "grad_norm": 1.0321218967437744,
+      "learning_rate": 0.0002,
+      "loss": 0.5591,
+      "step": 28230
+    },
+    {
+      "epoch": 4.565516126424703,
+      "grad_norm": 1.168332815170288,
+      "learning_rate": 0.0002,
+      "loss": 0.5508,
+      "step": 28240
+    },
+    {
+      "epoch": 4.5671328106054485,
+      "grad_norm": 1.0067222118377686,
+      "learning_rate": 0.0002,
+      "loss": 0.5649,
+      "step": 28250
+    },
+    {
+      "epoch": 4.568749494786194,
+      "grad_norm": 1.0283393859863281,
+      "learning_rate": 0.0002,
+      "loss": 0.5853,
+      "step": 28260
+    },
+    {
+      "epoch": 4.570366178966939,
+      "grad_norm": 0.9912363886833191,
+      "learning_rate": 0.0002,
+      "loss": 0.5772,
+      "step": 28270
+    },
+    {
+      "epoch": 4.571982863147684,
+      "grad_norm": 1.108032464981079,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 28280
+    },
+    {
+      "epoch": 4.573599547328429,
+      "grad_norm": 0.8260078430175781,
+      "learning_rate": 0.0002,
+      "loss": 0.5529,
+      "step": 28290
+    },
+    {
+      "epoch": 4.575216231509175,
+      "grad_norm": 0.8946247100830078,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 28300
+    },
+    {
+      "epoch": 4.57683291568992,
+      "grad_norm": 0.8273587822914124,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 28310
+    },
+    {
+      "epoch": 4.578449599870665,
+      "grad_norm": 0.9040093421936035,
+      "learning_rate": 0.0002,
+      "loss": 0.6058,
+      "step": 28320
+    },
+    {
+      "epoch": 4.58006628405141,
+      "grad_norm": 0.8435290455818176,
+      "learning_rate": 0.0002,
+      "loss": 0.5521,
+      "step": 28330
+    },
+    {
+      "epoch": 4.581682968232156,
+      "grad_norm": 1.164088249206543,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 28340
+    },
+    {
+      "epoch": 4.583299652412901,
+      "grad_norm": 0.9861085414886475,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28350
+    },
+    {
+      "epoch": 4.584916336593646,
+      "grad_norm": 0.8892980813980103,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 28360
+    },
+    {
+      "epoch": 4.586533020774391,
+      "grad_norm": 1.240574836730957,
+      "learning_rate": 0.0002,
+      "loss": 0.598,
+      "step": 28370
+    },
+    {
+      "epoch": 4.588149704955137,
+      "grad_norm": 0.8669408559799194,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 28380
+    },
+    {
+      "epoch": 4.589766389135883,
+      "grad_norm": 0.9145985841751099,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28390
+    },
+    {
+      "epoch": 4.591383073316628,
+      "grad_norm": 0.8584614992141724,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 28400
+    },
+    {
+      "epoch": 4.592999757497373,
+      "grad_norm": 1.118829369544983,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 28410
+    },
+    {
+      "epoch": 4.594616441678118,
+      "grad_norm": 1.1411553621292114,
+      "learning_rate": 0.0002,
+      "loss": 0.5641,
+      "step": 28420
+    },
+    {
+      "epoch": 4.596233125858864,
+      "grad_norm": 0.9433278441429138,
+      "learning_rate": 0.0002,
+      "loss": 0.549,
+      "step": 28430
+    },
+    {
+      "epoch": 4.597849810039609,
+      "grad_norm": 0.816830039024353,
+      "learning_rate": 0.0002,
+      "loss": 0.5496,
+      "step": 28440
+    },
+    {
+      "epoch": 4.599466494220354,
+      "grad_norm": 1.2124968767166138,
+      "learning_rate": 0.0002,
+      "loss": 0.5543,
+      "step": 28450
+    },
+    {
+      "epoch": 4.601083178401099,
+      "grad_norm": 0.9658762216567993,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 28460
+    },
+    {
+      "epoch": 4.6026998625818445,
+      "grad_norm": 0.836100161075592,
+      "learning_rate": 0.0002,
+      "loss": 0.5902,
+      "step": 28470
+    },
+    {
+      "epoch": 4.60431654676259,
+      "grad_norm": 0.9989104270935059,
+      "learning_rate": 0.0002,
+      "loss": 0.5749,
+      "step": 28480
+    },
+    {
+      "epoch": 4.605933230943335,
+      "grad_norm": 1.1298956871032715,
+      "learning_rate": 0.0002,
+      "loss": 0.5616,
+      "step": 28490
+    },
+    {
+      "epoch": 4.60754991512408,
+      "grad_norm": 1.1731704473495483,
+      "learning_rate": 0.0002,
+      "loss": 0.5846,
+      "step": 28500
+    },
+    {
+      "epoch": 4.609166599304825,
+      "grad_norm": 0.9624714255332947,
+      "learning_rate": 0.0002,
+      "loss": 0.5816,
+      "step": 28510
+    },
+    {
+      "epoch": 4.610783283485571,
+      "grad_norm": 1.364073634147644,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 28520
+    },
+    {
+      "epoch": 4.612399967666317,
+      "grad_norm": 1.1827356815338135,
+      "learning_rate": 0.0002,
+      "loss": 0.6237,
+      "step": 28530
+    },
+    {
+      "epoch": 4.614016651847062,
+      "grad_norm": 0.6651531457901001,
+      "learning_rate": 0.0002,
+      "loss": 0.5643,
+      "step": 28540
+    },
+    {
+      "epoch": 4.615633336027807,
+      "grad_norm": 1.1640995740890503,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 28550
+    },
+    {
+      "epoch": 4.6172500202085525,
+      "grad_norm": 1.028918743133545,
+      "learning_rate": 0.0002,
+      "loss": 0.5995,
+      "step": 28560
+    },
+    {
+      "epoch": 4.618866704389298,
+      "grad_norm": 0.8252120614051819,
+      "learning_rate": 0.0002,
+      "loss": 0.5607,
+      "step": 28570
+    },
+    {
+      "epoch": 4.620483388570043,
+      "grad_norm": 1.3536735773086548,
+      "learning_rate": 0.0002,
+      "loss": 0.5769,
+      "step": 28580
+    },
+    {
+      "epoch": 4.622100072750788,
+      "grad_norm": 1.2146915197372437,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 28590
+    },
+    {
+      "epoch": 4.623716756931533,
+      "grad_norm": 1.0122549533843994,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 28600
+    },
+    {
+      "epoch": 4.625333441112279,
+      "grad_norm": 0.9977872967720032,
+      "learning_rate": 0.0002,
+      "loss": 0.6072,
+      "step": 28610
+    },
+    {
+      "epoch": 4.626950125293024,
+      "grad_norm": 1.0159751176834106,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 28620
+    },
+    {
+      "epoch": 4.628566809473769,
+      "grad_norm": 1.0028325319290161,
+      "learning_rate": 0.0002,
+      "loss": 0.5935,
+      "step": 28630
+    },
+    {
+      "epoch": 4.630183493654514,
+      "grad_norm": 0.901638388633728,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 28640
+    },
+    {
+      "epoch": 4.6318001778352595,
+      "grad_norm": 0.9450507164001465,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 28650
+    },
+    {
+      "epoch": 4.633416862016006,
+      "grad_norm": 0.9987545013427734,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 28660
+    },
+    {
+      "epoch": 4.63503354619675,
+      "grad_norm": 0.9574332237243652,
+      "learning_rate": 0.0002,
+      "loss": 0.5863,
+      "step": 28670
+    },
+    {
+      "epoch": 4.636650230377496,
+      "grad_norm": 1.2215653657913208,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 28680
+    },
+    {
+      "epoch": 4.638266914558241,
+      "grad_norm": 0.9798858761787415,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 28690
+    },
+    {
+      "epoch": 4.639883598738987,
+      "grad_norm": 1.0648466348648071,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 28700
+    },
+    {
+      "epoch": 4.641500282919732,
+      "grad_norm": 1.0606504678726196,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 28710
+    },
+    {
+      "epoch": 4.643116967100477,
+      "grad_norm": 1.0892442464828491,
+      "learning_rate": 0.0002,
+      "loss": 0.5801,
+      "step": 28720
+    },
+    {
+      "epoch": 4.644733651281222,
+      "grad_norm": 0.914391040802002,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 28730
+    },
+    {
+      "epoch": 4.6463503354619675,
+      "grad_norm": 0.9782370328903198,
+      "learning_rate": 0.0002,
+      "loss": 0.5439,
+      "step": 28740
+    },
+    {
+      "epoch": 4.647967019642713,
+      "grad_norm": 1.0344339609146118,
+      "learning_rate": 0.0002,
+      "loss": 0.6035,
+      "step": 28750
+    },
+    {
+      "epoch": 4.649583703823458,
+      "grad_norm": 1.0513931512832642,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 28760
+    },
+    {
+      "epoch": 4.651200388004203,
+      "grad_norm": 0.9711475968360901,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 28770
+    },
+    {
+      "epoch": 4.652817072184948,
+      "grad_norm": 0.977519690990448,
+      "learning_rate": 0.0002,
+      "loss": 0.5472,
+      "step": 28780
+    },
+    {
+      "epoch": 4.654433756365694,
+      "grad_norm": 0.9150224924087524,
+      "learning_rate": 0.0002,
+      "loss": 0.5826,
+      "step": 28790
+    },
+    {
+      "epoch": 4.656050440546439,
+      "grad_norm": 1.0973542928695679,
+      "learning_rate": 0.0002,
+      "loss": 0.5382,
+      "step": 28800
+    },
+    {
+      "epoch": 4.657667124727185,
+      "grad_norm": 0.944877564907074,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 28810
+    },
+    {
+      "epoch": 4.659283808907929,
+      "grad_norm": 0.9508748650550842,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 28820
+    },
+    {
+      "epoch": 4.6609004930886755,
+      "grad_norm": 0.9681721329689026,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 28830
+    },
+    {
+      "epoch": 4.662517177269421,
+      "grad_norm": 1.0214351415634155,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 28840
+    },
+    {
+      "epoch": 4.664133861450166,
+      "grad_norm": 0.9748611450195312,
+      "learning_rate": 0.0002,
+      "loss": 0.6031,
+      "step": 28850
+    },
+    {
+      "epoch": 4.665750545630911,
+      "grad_norm": 0.8484147191047668,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 28860
+    },
+    {
+      "epoch": 4.667367229811656,
+      "grad_norm": 1.1252986192703247,
+      "learning_rate": 0.0002,
+      "loss": 0.5699,
+      "step": 28870
+    },
+    {
+      "epoch": 4.668983913992402,
+      "grad_norm": 0.8706206679344177,
+      "learning_rate": 0.0002,
+      "loss": 0.5724,
+      "step": 28880
+    },
+    {
+      "epoch": 4.670600598173147,
+      "grad_norm": 1.1432424783706665,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 28890
+    },
+    {
+      "epoch": 4.672217282353892,
+      "grad_norm": 1.017029047012329,
+      "learning_rate": 0.0002,
+      "loss": 0.5675,
+      "step": 28900
+    },
+    {
+      "epoch": 4.673833966534637,
+      "grad_norm": 1.085597038269043,
+      "learning_rate": 0.0002,
+      "loss": 0.5831,
+      "step": 28910
+    },
+    {
+      "epoch": 4.675450650715383,
+      "grad_norm": 0.9275796413421631,
+      "learning_rate": 0.0002,
+      "loss": 0.5678,
+      "step": 28920
+    },
+    {
+      "epoch": 4.677067334896128,
+      "grad_norm": 0.9518964886665344,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28930
+    },
+    {
+      "epoch": 4.678684019076873,
+      "grad_norm": 1.0352122783660889,
+      "learning_rate": 0.0002,
+      "loss": 0.6232,
+      "step": 28940
+    },
+    {
+      "epoch": 4.680300703257618,
+      "grad_norm": 1.090124249458313,
+      "learning_rate": 0.0002,
+      "loss": 0.5786,
+      "step": 28950
+    },
+    {
+      "epoch": 4.681917387438364,
+      "grad_norm": 0.8799563050270081,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 28960
+    },
+    {
+      "epoch": 4.683534071619109,
+      "grad_norm": 1.0929821729660034,
+      "learning_rate": 0.0002,
+      "loss": 0.5787,
+      "step": 28970
+    },
+    {
+      "epoch": 4.685150755799855,
+      "grad_norm": 0.903727650642395,
+      "learning_rate": 0.0002,
+      "loss": 0.6134,
+      "step": 28980
+    },
+    {
+      "epoch": 4.6867674399806,
+      "grad_norm": 0.9752424955368042,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 28990
+    },
+    {
+      "epoch": 4.688384124161345,
+      "grad_norm": 0.9351571202278137,
+      "learning_rate": 0.0002,
+      "loss": 0.5762,
+      "step": 29000
+    },
+    {
+      "epoch": 4.6900008083420905,
+      "grad_norm": 0.923877477645874,
+      "learning_rate": 0.0002,
+      "loss": 0.5811,
+      "step": 29010
+    },
+    {
+      "epoch": 4.691617492522836,
+      "grad_norm": 1.045389175415039,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 29020
+    },
+    {
+      "epoch": 4.693234176703581,
+      "grad_norm": 1.0200831890106201,
+      "learning_rate": 0.0002,
+      "loss": 0.584,
+      "step": 29030
+    },
+    {
+      "epoch": 4.694850860884326,
+      "grad_norm": 1.1499706506729126,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 29040
+    },
+    {
+      "epoch": 4.6964675450650715,
+      "grad_norm": 0.860118567943573,
+      "learning_rate": 0.0002,
+      "loss": 0.5745,
+      "step": 29050
+    },
+    {
+      "epoch": 4.698084229245817,
+      "grad_norm": 0.9774864315986633,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 29060
+    },
+    {
+      "epoch": 4.699700913426562,
+      "grad_norm": 1.0323210954666138,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 29070
+    },
+    {
+      "epoch": 4.701317597607307,
+      "grad_norm": 0.8492481112480164,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 29080
+    },
+    {
+      "epoch": 4.702934281788052,
+      "grad_norm": 1.131951093673706,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 29090
+    },
+    {
+      "epoch": 4.704550965968798,
+      "grad_norm": 0.8763113021850586,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 29100
+    },
+    {
+      "epoch": 4.706167650149544,
+      "grad_norm": 1.045028805732727,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 29110
+    },
+    {
+      "epoch": 4.707784334330288,
+      "grad_norm": 0.9961401224136353,
+      "learning_rate": 0.0002,
+      "loss": 0.5548,
+      "step": 29120
+    },
+    {
+      "epoch": 4.709401018511034,
+      "grad_norm": 0.9282503724098206,
+      "learning_rate": 0.0002,
+      "loss": 0.559,
+      "step": 29130
+    },
+    {
+      "epoch": 4.711017702691779,
+      "grad_norm": 1.1418932676315308,
+      "learning_rate": 0.0002,
+      "loss": 0.5744,
+      "step": 29140
+    },
+    {
+      "epoch": 4.712634386872525,
+      "grad_norm": 0.9950099587440491,
+      "learning_rate": 0.0002,
+      "loss": 0.5394,
+      "step": 29150
+    },
+    {
+      "epoch": 4.71425107105327,
+      "grad_norm": 0.8304893374443054,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 29160
+    },
+    {
+      "epoch": 4.715867755234015,
+      "grad_norm": 1.115626335144043,
+      "learning_rate": 0.0002,
+      "loss": 0.6074,
+      "step": 29170
+    },
+    {
+      "epoch": 4.71748443941476,
+      "grad_norm": 1.079818606376648,
+      "learning_rate": 0.0002,
+      "loss": 0.6265,
+      "step": 29180
+    },
+    {
+      "epoch": 4.719101123595506,
+      "grad_norm": 1.1929082870483398,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 29190
+    },
+    {
+      "epoch": 4.720717807776251,
+      "grad_norm": 0.9621080756187439,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 29200
+    },
+    {
+      "epoch": 4.722334491956996,
+      "grad_norm": 0.8549222350120544,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 29210
+    },
+    {
+      "epoch": 4.723951176137741,
+      "grad_norm": 0.9341941475868225,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 29220
+    },
+    {
+      "epoch": 4.7255678603184865,
+      "grad_norm": 1.075406789779663,
+      "learning_rate": 0.0002,
+      "loss": 0.5436,
+      "step": 29230
+    },
+    {
+      "epoch": 4.727184544499232,
+      "grad_norm": 1.0859880447387695,
+      "learning_rate": 0.0002,
+      "loss": 0.576,
+      "step": 29240
+    },
+    {
+      "epoch": 4.728801228679977,
+      "grad_norm": 0.8475605249404907,
+      "learning_rate": 0.0002,
+      "loss": 0.5525,
+      "step": 29250
+    },
+    {
+      "epoch": 4.730417912860723,
+      "grad_norm": 0.9331845641136169,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 29260
+    },
+    {
+      "epoch": 4.7320345970414674,
+      "grad_norm": 0.9279314279556274,
+      "learning_rate": 0.0002,
+      "loss": 0.5901,
+      "step": 29270
+    },
+    {
+      "epoch": 4.733651281222214,
+      "grad_norm": 0.7803558707237244,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 29280
+    },
+    {
+      "epoch": 4.735267965402959,
+      "grad_norm": 1.0159329175949097,
+      "learning_rate": 0.0002,
+      "loss": 0.5968,
+      "step": 29290
+    },
+    {
+      "epoch": 4.736884649583704,
+      "grad_norm": 0.9448670744895935,
+      "learning_rate": 0.0002,
+      "loss": 0.5333,
+      "step": 29300
+    },
+    {
+      "epoch": 4.738501333764449,
+      "grad_norm": 1.0732197761535645,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 29310
+    },
+    {
+      "epoch": 4.7401180179451945,
+      "grad_norm": 0.901830792427063,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 29320
+    },
+    {
+      "epoch": 4.74173470212594,
+      "grad_norm": 0.9141789674758911,
+      "learning_rate": 0.0002,
+      "loss": 0.6105,
+      "step": 29330
+    },
+    {
+      "epoch": 4.743351386306685,
+      "grad_norm": 0.9733418226242065,
+      "learning_rate": 0.0002,
+      "loss": 0.5481,
+      "step": 29340
+    },
+    {
+      "epoch": 4.74496807048743,
+      "grad_norm": 0.909810483455658,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 29350
+    },
+    {
+      "epoch": 4.746584754668175,
+      "grad_norm": 0.909541666507721,
+      "learning_rate": 0.0002,
+      "loss": 0.5911,
+      "step": 29360
+    },
+    {
+      "epoch": 4.748201438848921,
+      "grad_norm": 0.9383015632629395,
+      "learning_rate": 0.0002,
+      "loss": 0.5579,
+      "step": 29370
+    },
+    {
+      "epoch": 4.749818123029666,
+      "grad_norm": 0.9275668263435364,
+      "learning_rate": 0.0002,
+      "loss": 0.5529,
+      "step": 29380
+    },
+    {
+      "epoch": 4.751434807210411,
+      "grad_norm": 1.1146225929260254,
+      "learning_rate": 0.0002,
+      "loss": 0.5623,
+      "step": 29390
+    },
+    {
+      "epoch": 4.753051491391156,
+      "grad_norm": 1.0062453746795654,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 29400
+    },
+    {
+      "epoch": 4.7546681755719025,
+      "grad_norm": 0.9451895952224731,
+      "learning_rate": 0.0002,
+      "loss": 0.5872,
+      "step": 29410
+    },
+    {
+      "epoch": 4.756284859752648,
+      "grad_norm": 0.870457649230957,
+      "learning_rate": 0.0002,
+      "loss": 0.5767,
+      "step": 29420
+    },
+    {
+      "epoch": 4.757901543933393,
+      "grad_norm": 1.0411282777786255,
+      "learning_rate": 0.0002,
+      "loss": 0.57,
+      "step": 29430
+    },
+    {
+      "epoch": 4.759518228114138,
+      "grad_norm": 1.1648986339569092,
+      "learning_rate": 0.0002,
+      "loss": 0.5688,
+      "step": 29440
+    },
+    {
+      "epoch": 4.761134912294883,
+      "grad_norm": 0.8999572992324829,
+      "learning_rate": 0.0002,
+      "loss": 0.5432,
+      "step": 29450
+    },
+    {
+      "epoch": 4.762751596475629,
+      "grad_norm": 0.9863559007644653,
+      "learning_rate": 0.0002,
+      "loss": 0.5667,
+      "step": 29460
+    },
+    {
+      "epoch": 4.764368280656374,
+      "grad_norm": 0.9676542282104492,
+      "learning_rate": 0.0002,
+      "loss": 0.5779,
+      "step": 29470
+    },
+    {
+      "epoch": 4.765984964837119,
+      "grad_norm": 1.004775047302246,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 29480
+    },
+    {
+      "epoch": 4.767601649017864,
+      "grad_norm": 1.0937515497207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6044,
+      "step": 29490
+    },
+    {
+      "epoch": 4.7692183331986095,
+      "grad_norm": 0.9551598429679871,
+      "learning_rate": 0.0002,
+      "loss": 0.5433,
+      "step": 29500
+    },
+    {
+      "epoch": 4.770835017379355,
+      "grad_norm": 1.0757228136062622,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 29510
+    },
+    {
+      "epoch": 4.7724517015601,
+      "grad_norm": 1.0588841438293457,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 29520
+    },
+    {
+      "epoch": 4.774068385740845,
+      "grad_norm": 1.0744032859802246,
+      "learning_rate": 0.0002,
+      "loss": 0.5814,
+      "step": 29530
+    },
+    {
+      "epoch": 4.7756850699215905,
+      "grad_norm": 1.0066277980804443,
+      "learning_rate": 0.0002,
+      "loss": 0.5681,
+      "step": 29540
+    },
+    {
+      "epoch": 4.777301754102336,
+      "grad_norm": 1.082319736480713,
+      "learning_rate": 0.0002,
+      "loss": 0.545,
+      "step": 29550
+    },
+    {
+      "epoch": 4.778918438283082,
+      "grad_norm": 0.8252472877502441,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 29560
+    },
+    {
+      "epoch": 4.780535122463827,
+      "grad_norm": 0.9855340123176575,
+      "learning_rate": 0.0002,
+      "loss": 0.5666,
+      "step": 29570
+    },
+    {
+      "epoch": 4.782151806644572,
+      "grad_norm": 0.9991421699523926,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 29580
+    },
+    {
+      "epoch": 4.7837684908253175,
+      "grad_norm": 1.316841959953308,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 29590
+    },
+    {
+      "epoch": 4.785385175006063,
+      "grad_norm": 1.1513035297393799,
+      "learning_rate": 0.0002,
+      "loss": 0.6102,
+      "step": 29600
+    },
+    {
+      "epoch": 4.787001859186808,
+      "grad_norm": 0.9767683744430542,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 29610
+    },
+    {
+      "epoch": 4.788618543367553,
+      "grad_norm": 0.9786278605461121,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 29620
+    },
+    {
+      "epoch": 4.7902352275482984,
+      "grad_norm": 0.8004973530769348,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 29630
+    },
+    {
+      "epoch": 4.791851911729044,
+      "grad_norm": 1.0997767448425293,
+      "learning_rate": 0.0002,
+      "loss": 0.5932,
+      "step": 29640
+    },
+    {
+      "epoch": 4.793468595909789,
+      "grad_norm": 0.9752856492996216,
+      "learning_rate": 0.0002,
+      "loss": 0.5655,
+      "step": 29650
+    },
+    {
+      "epoch": 4.795085280090534,
+      "grad_norm": 1.0518392324447632,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 29660
+    },
+    {
+      "epoch": 4.796701964271279,
+      "grad_norm": 1.1050055027008057,
+      "learning_rate": 0.0002,
+      "loss": 0.6042,
+      "step": 29670
+    },
+    {
+      "epoch": 4.798318648452025,
+      "grad_norm": 0.9933857917785645,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 29680
+    },
+    {
+      "epoch": 4.79993533263277,
+      "grad_norm": 1.2804018259048462,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 29690
+    },
+    {
+      "epoch": 4.801552016813515,
+      "grad_norm": 1.0133371353149414,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 29700
+    },
+    {
+      "epoch": 4.803168700994261,
+      "grad_norm": 1.080350637435913,
+      "learning_rate": 0.0002,
+      "loss": 0.5662,
+      "step": 29710
+    },
+    {
+      "epoch": 4.804785385175006,
+      "grad_norm": 0.9986529350280762,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 29720
+    },
+    {
+      "epoch": 4.806402069355752,
+      "grad_norm": 0.975665807723999,
+      "learning_rate": 0.0002,
+      "loss": 0.5894,
+      "step": 29730
+    },
+    {
+      "epoch": 4.808018753536497,
+      "grad_norm": 0.8458138704299927,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 29740
+    },
+    {
+      "epoch": 4.809635437717242,
+      "grad_norm": 0.99330073595047,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 29750
+    },
+    {
+      "epoch": 4.811252121897987,
+      "grad_norm": 0.898274302482605,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 29760
+    },
+    {
+      "epoch": 4.812868806078733,
+      "grad_norm": 1.0504480600357056,
+      "learning_rate": 0.0002,
+      "loss": 0.5842,
+      "step": 29770
+    },
+    {
+      "epoch": 4.814485490259478,
+      "grad_norm": 0.937919020652771,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 29780
+    },
+    {
+      "epoch": 4.816102174440223,
+      "grad_norm": 0.9593307971954346,
+      "learning_rate": 0.0002,
+      "loss": 0.5885,
+      "step": 29790
+    },
+    {
+      "epoch": 4.817718858620968,
+      "grad_norm": 0.9431198835372925,
+      "learning_rate": 0.0002,
+      "loss": 0.578,
+      "step": 29800
+    },
+    {
+      "epoch": 4.8193355428017135,
+      "grad_norm": 1.2729957103729248,
+      "learning_rate": 0.0002,
+      "loss": 0.5739,
+      "step": 29810
+    },
+    {
+      "epoch": 4.820952226982459,
+      "grad_norm": 0.8876838684082031,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 29820
+    },
+    {
+      "epoch": 4.822568911163204,
+      "grad_norm": 1.0185000896453857,
+      "learning_rate": 0.0002,
+      "loss": 0.5583,
+      "step": 29830
+    },
+    {
+      "epoch": 4.824185595343949,
+      "grad_norm": 1.064276099205017,
+      "learning_rate": 0.0002,
+      "loss": 0.5686,
+      "step": 29840
+    },
+    {
+      "epoch": 4.825802279524694,
+      "grad_norm": 0.9774803519248962,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 29850
+    },
+    {
+      "epoch": 4.8274189637054405,
+      "grad_norm": 1.131646990776062,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 29860
+    },
+    {
+      "epoch": 4.829035647886186,
+      "grad_norm": 1.081455945968628,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 29870
+    },
+    {
+      "epoch": 4.830652332066931,
+      "grad_norm": 0.990538477897644,
+      "learning_rate": 0.0002,
+      "loss": 0.5793,
+      "step": 29880
+    },
+    {
+      "epoch": 4.832269016247676,
+      "grad_norm": 0.9750600457191467,
+      "learning_rate": 0.0002,
+      "loss": 0.5833,
+      "step": 29890
+    },
+    {
+      "epoch": 4.8338857004284215,
+      "grad_norm": 1.0600621700286865,
+      "learning_rate": 0.0002,
+      "loss": 0.619,
+      "step": 29900
+    },
+    {
+      "epoch": 4.835502384609167,
+      "grad_norm": 0.9237320423126221,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 29910
+    },
+    {
+      "epoch": 4.837119068789912,
+      "grad_norm": 0.9739177227020264,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 29920
+    },
+    {
+      "epoch": 4.838735752970657,
+      "grad_norm": 1.128677248954773,
+      "learning_rate": 0.0002,
+      "loss": 0.587,
+      "step": 29930
+    },
+    {
+      "epoch": 4.840352437151402,
+      "grad_norm": 1.042604923248291,
+      "learning_rate": 0.0002,
+      "loss": 0.564,
+      "step": 29940
+    },
+    {
+      "epoch": 4.841969121332148,
+      "grad_norm": 0.849758505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.5885,
+      "step": 29950
+    },
+    {
+      "epoch": 4.843585805512893,
+      "grad_norm": 1.2809888124465942,
+      "learning_rate": 0.0002,
+      "loss": 0.5952,
+      "step": 29960
+    },
+    {
+      "epoch": 4.845202489693638,
+      "grad_norm": 1.0177865028381348,
+      "learning_rate": 0.0002,
+      "loss": 0.5703,
+      "step": 29970
+    },
+    {
+      "epoch": 4.846819173874383,
+      "grad_norm": 1.0026639699935913,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 29980
+    },
+    {
+      "epoch": 4.8484358580551286,
+      "grad_norm": 0.9679505228996277,
+      "learning_rate": 0.0002,
+      "loss": 0.5897,
+      "step": 29990
+    },
+    {
+      "epoch": 4.850052542235874,
+      "grad_norm": 0.8939532041549683,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 30000
+    },
+    {
+      "epoch": 4.85166922641662,
+      "grad_norm": 0.9957457780838013,
+      "learning_rate": 0.0002,
+      "loss": 0.5852,
+      "step": 30010
+    },
+    {
+      "epoch": 4.853285910597365,
+      "grad_norm": 1.1646790504455566,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 30020
+    },
+    {
+      "epoch": 4.85490259477811,
+      "grad_norm": 0.8804680705070496,
+      "learning_rate": 0.0002,
+      "loss": 0.5711,
+      "step": 30030
+    },
+    {
+      "epoch": 4.856519278958856,
+      "grad_norm": 1.161970853805542,
+      "learning_rate": 0.0002,
+      "loss": 0.5397,
+      "step": 30040
+    },
+    {
+      "epoch": 4.858135963139601,
+      "grad_norm": 0.9081037640571594,
+      "learning_rate": 0.0002,
+      "loss": 0.5552,
+      "step": 30050
+    },
+    {
+      "epoch": 4.859752647320346,
+      "grad_norm": 0.9402848482131958,
+      "learning_rate": 0.0002,
+      "loss": 0.6024,
+      "step": 30060
+    },
+    {
+      "epoch": 4.861369331501091,
+      "grad_norm": 0.9023865461349487,
+      "learning_rate": 0.0002,
+      "loss": 0.6256,
+      "step": 30070
+    },
+    {
+      "epoch": 4.8629860156818365,
+      "grad_norm": 1.0173414945602417,
+      "learning_rate": 0.0002,
+      "loss": 0.5926,
+      "step": 30080
+    },
+    {
+      "epoch": 4.864602699862582,
+      "grad_norm": 1.084402322769165,
+      "learning_rate": 0.0002,
+      "loss": 0.6274,
+      "step": 30090
+    },
+    {
+      "epoch": 4.866219384043327,
+      "grad_norm": 0.9577937126159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6311,
+      "step": 30100
+    },
+    {
+      "epoch": 4.867836068224072,
+      "grad_norm": 0.9807606935501099,
+      "learning_rate": 0.0002,
+      "loss": 0.5724,
+      "step": 30110
+    },
+    {
+      "epoch": 4.8694527524048175,
+      "grad_norm": 0.978784441947937,
+      "learning_rate": 0.0002,
+      "loss": 0.5786,
+      "step": 30120
+    },
+    {
+      "epoch": 4.871069436585563,
+      "grad_norm": 0.9762914776802063,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 30130
+    },
+    {
+      "epoch": 4.872686120766308,
+      "grad_norm": 0.9404871463775635,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 30140
+    },
+    {
+      "epoch": 4.874302804947053,
+      "grad_norm": 1.0069509744644165,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 30150
+    },
+    {
+      "epoch": 4.875919489127799,
+      "grad_norm": 1.1770923137664795,
+      "learning_rate": 0.0002,
+      "loss": 0.6225,
+      "step": 30160
+    },
+    {
+      "epoch": 4.8775361733085445,
+      "grad_norm": 1.021210789680481,
+      "learning_rate": 0.0002,
+      "loss": 0.5657,
+      "step": 30170
+    },
+    {
+      "epoch": 4.87915285748929,
+      "grad_norm": 0.8512648940086365,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 30180
+    },
+    {
+      "epoch": 4.880769541670035,
+      "grad_norm": 0.9345870018005371,
+      "learning_rate": 0.0002,
+      "loss": 0.5519,
+      "step": 30190
+    },
+    {
+      "epoch": 4.88238622585078,
+      "grad_norm": 1.0224418640136719,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 30200
+    },
+    {
+      "epoch": 4.884002910031525,
+      "grad_norm": 1.0316044092178345,
+      "learning_rate": 0.0002,
+      "loss": 0.5807,
+      "step": 30210
+    },
+    {
+      "epoch": 4.885619594212271,
+      "grad_norm": 1.102437973022461,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 30220
+    },
+    {
+      "epoch": 4.887236278393016,
+      "grad_norm": 1.0220023393630981,
+      "learning_rate": 0.0002,
+      "loss": 0.586,
+      "step": 30230
+    },
+    {
+      "epoch": 4.888852962573761,
+      "grad_norm": 1.0934523344039917,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 30240
+    },
+    {
+      "epoch": 4.890469646754506,
+      "grad_norm": 1.264630913734436,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 30250
+    },
+    {
+      "epoch": 4.892086330935252,
+      "grad_norm": 1.0999879837036133,
+      "learning_rate": 0.0002,
+      "loss": 0.5712,
+      "step": 30260
+    },
+    {
+      "epoch": 4.893703015115997,
+      "grad_norm": 0.9124550223350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 30270
+    },
+    {
+      "epoch": 4.895319699296742,
+      "grad_norm": 0.9853624105453491,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 30280
+    },
+    {
+      "epoch": 4.896936383477488,
+      "grad_norm": 1.0589802265167236,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 30290
+    },
+    {
+      "epoch": 4.8985530676582325,
+      "grad_norm": 0.8487226366996765,
+      "learning_rate": 0.0002,
+      "loss": 0.6129,
+      "step": 30300
+    },
+    {
+      "epoch": 4.900169751838979,
+      "grad_norm": 1.0212191343307495,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 30310
+    },
+    {
+      "epoch": 4.901786436019724,
+      "grad_norm": 1.0187491178512573,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 30320
+    },
+    {
+      "epoch": 4.903403120200469,
+      "grad_norm": 1.0013091564178467,
+      "learning_rate": 0.0002,
+      "loss": 0.5809,
+      "step": 30330
+    },
+    {
+      "epoch": 4.905019804381214,
+      "grad_norm": 1.0017542839050293,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 30340
+    },
+    {
+      "epoch": 4.9066364885619596,
+      "grad_norm": 0.9665151238441467,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 30350
+    },
+    {
+      "epoch": 4.908253172742705,
+      "grad_norm": 0.8774822950363159,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 30360
+    },
+    {
+      "epoch": 4.90986985692345,
+      "grad_norm": 0.9449850916862488,
+      "learning_rate": 0.0002,
+      "loss": 0.5771,
+      "step": 30370
+    },
+    {
+      "epoch": 4.911486541104195,
+      "grad_norm": 0.7368341088294983,
+      "learning_rate": 0.0002,
+      "loss": 0.58,
+      "step": 30380
+    },
+    {
+      "epoch": 4.9131032252849405,
+      "grad_norm": 0.9669167995452881,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 30390
+    },
+    {
+      "epoch": 4.914719909465686,
+      "grad_norm": 1.1227794885635376,
+      "learning_rate": 0.0002,
+      "loss": 0.6202,
+      "step": 30400
+    },
+    {
+      "epoch": 4.916336593646431,
+      "grad_norm": 0.9884361028671265,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 30410
+    },
+    {
+      "epoch": 4.917953277827176,
+      "grad_norm": 0.9949551224708557,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 30420
+    },
+    {
+      "epoch": 4.919569962007921,
+      "grad_norm": 0.9491621851921082,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 30430
+    },
+    {
+      "epoch": 4.9211866461886675,
+      "grad_norm": 0.78848797082901,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 30440
+    },
+    {
+      "epoch": 4.922803330369412,
+      "grad_norm": 1.0693835020065308,
+      "learning_rate": 0.0002,
+      "loss": 0.5561,
+      "step": 30450
+    },
+    {
+      "epoch": 4.924420014550158,
+      "grad_norm": 0.9573729634284973,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 30460
+    },
+    {
+      "epoch": 4.926036698730903,
+      "grad_norm": 0.9975152611732483,
+      "learning_rate": 0.0002,
+      "loss": 0.6084,
+      "step": 30470
+    },
+    {
+      "epoch": 4.9276533829116484,
+      "grad_norm": 0.8695693016052246,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 30480
+    },
+    {
+      "epoch": 4.929270067092394,
+      "grad_norm": 1.145394206047058,
+      "learning_rate": 0.0002,
+      "loss": 0.6144,
+      "step": 30490
+    },
+    {
+      "epoch": 4.930886751273139,
+      "grad_norm": 0.7668989896774292,
+      "learning_rate": 0.0002,
+      "loss": 0.5736,
+      "step": 30500
+    },
+    {
+      "epoch": 4.932503435453884,
+      "grad_norm": 0.9630151391029358,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 30510
+    },
+    {
+      "epoch": 4.934120119634629,
+      "grad_norm": 0.940705418586731,
+      "learning_rate": 0.0002,
+      "loss": 0.6461,
+      "step": 30520
+    },
+    {
+      "epoch": 4.935736803815375,
+      "grad_norm": 1.3243348598480225,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 30530
+    },
+    {
+      "epoch": 4.93735348799612,
+      "grad_norm": 1.004347801208496,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 30540
+    },
+    {
+      "epoch": 4.938970172176865,
+      "grad_norm": 0.8711541295051575,
+      "learning_rate": 0.0002,
+      "loss": 0.583,
+      "step": 30550
+    },
+    {
+      "epoch": 4.94058685635761,
+      "grad_norm": 0.8980631828308105,
+      "learning_rate": 0.0002,
+      "loss": 0.599,
+      "step": 30560
+    },
+    {
+      "epoch": 4.9422035405383555,
+      "grad_norm": 0.8388893604278564,
+      "learning_rate": 0.0002,
+      "loss": 0.6024,
+      "step": 30570
+    },
+    {
+      "epoch": 4.943820224719101,
+      "grad_norm": 1.0991183519363403,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 30580
+    },
+    {
+      "epoch": 4.945436908899847,
+      "grad_norm": 0.9731075763702393,
+      "learning_rate": 0.0002,
+      "loss": 0.5906,
+      "step": 30590
+    },
+    {
+      "epoch": 4.947053593080591,
+      "grad_norm": 1.3904452323913574,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 30600
+    },
+    {
+      "epoch": 4.948670277261337,
+      "grad_norm": 1.2489882707595825,
+      "learning_rate": 0.0002,
+      "loss": 0.5952,
+      "step": 30610
+    },
+    {
+      "epoch": 4.950286961442083,
+      "grad_norm": 1.240072250366211,
+      "learning_rate": 0.0002,
+      "loss": 0.5887,
+      "step": 30620
+    },
+    {
+      "epoch": 4.951903645622828,
+      "grad_norm": 0.9191411733627319,
+      "learning_rate": 0.0002,
+      "loss": 0.5762,
+      "step": 30630
+    },
+    {
+      "epoch": 4.953520329803573,
+      "grad_norm": 0.8888895511627197,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 30640
+    },
+    {
+      "epoch": 4.955137013984318,
+      "grad_norm": 0.9001450538635254,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 30650
+    },
+    {
+      "epoch": 4.9567536981650635,
+      "grad_norm": 1.053971767425537,
+      "learning_rate": 0.0002,
+      "loss": 0.6047,
+      "step": 30660
+    },
+    {
+      "epoch": 4.958370382345809,
+      "grad_norm": 1.2224042415618896,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 30670
+    },
+    {
+      "epoch": 4.959987066526554,
+      "grad_norm": 0.8855111598968506,
+      "learning_rate": 0.0002,
+      "loss": 0.6211,
+      "step": 30680
+    },
+    {
+      "epoch": 4.961603750707299,
+      "grad_norm": 0.9489575624465942,
+      "learning_rate": 0.0002,
+      "loss": 0.5764,
+      "step": 30690
+    },
+    {
+      "epoch": 4.963220434888044,
+      "grad_norm": 0.9635404944419861,
+      "learning_rate": 0.0002,
+      "loss": 0.5371,
+      "step": 30700
+    },
+    {
+      "epoch": 4.96483711906879,
+      "grad_norm": 1.1784121990203857,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 30710
+    },
+    {
+      "epoch": 4.966453803249535,
+      "grad_norm": 1.0059462785720825,
+      "learning_rate": 0.0002,
+      "loss": 0.5803,
+      "step": 30720
+    },
+    {
+      "epoch": 4.96807048743028,
+      "grad_norm": 0.9479738473892212,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 30730
+    },
+    {
+      "epoch": 4.969687171611026,
+      "grad_norm": 1.0624593496322632,
+      "learning_rate": 0.0002,
+      "loss": 0.584,
+      "step": 30740
+    },
+    {
+      "epoch": 4.971303855791771,
+      "grad_norm": 1.1429259777069092,
+      "learning_rate": 0.0002,
+      "loss": 0.6202,
+      "step": 30750
+    },
+    {
+      "epoch": 4.972920539972517,
+      "grad_norm": 0.9102491140365601,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 30760
+    },
+    {
+      "epoch": 4.974537224153262,
+      "grad_norm": 1.1262688636779785,
+      "learning_rate": 0.0002,
+      "loss": 0.6025,
+      "step": 30770
+    },
+    {
+      "epoch": 4.976153908334007,
+      "grad_norm": 1.1415393352508545,
+      "learning_rate": 0.0002,
+      "loss": 0.588,
+      "step": 30780
+    },
+    {
+      "epoch": 4.977770592514752,
+      "grad_norm": 1.083078384399414,
+      "learning_rate": 0.0002,
+      "loss": 0.5832,
+      "step": 30790
+    },
+    {
+      "epoch": 4.979387276695498,
+      "grad_norm": 0.964859127998352,
+      "learning_rate": 0.0002,
+      "loss": 0.6025,
+      "step": 30800
+    },
+    {
+      "epoch": 4.981003960876243,
+      "grad_norm": 0.8704743385314941,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 30810
+    },
+    {
+      "epoch": 4.982620645056988,
+      "grad_norm": 1.0714856386184692,
+      "learning_rate": 0.0002,
+      "loss": 0.5666,
+      "step": 30820
+    },
+    {
+      "epoch": 4.984237329237733,
+      "grad_norm": 0.6818771362304688,
+      "learning_rate": 0.0002,
+      "loss": 0.565,
+      "step": 30830
+    },
+    {
+      "epoch": 4.985854013418479,
+      "grad_norm": 1.0454156398773193,
+      "learning_rate": 0.0002,
+      "loss": 0.5999,
+      "step": 30840
+    },
+    {
+      "epoch": 4.987470697599224,
+      "grad_norm": 0.9410776495933533,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 30850
+    },
+    {
+      "epoch": 4.989087381779969,
+      "grad_norm": 1.0878902673721313,
+      "learning_rate": 0.0002,
+      "loss": 0.5899,
+      "step": 30860
+    },
+    {
+      "epoch": 4.990704065960714,
+      "grad_norm": 0.8916727304458618,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 30870
+    },
+    {
+      "epoch": 4.9923207501414595,
+      "grad_norm": 1.045776128768921,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 30880
+    },
+    {
+      "epoch": 4.993937434322206,
+      "grad_norm": 0.9861903786659241,
+      "learning_rate": 0.0002,
+      "loss": 0.5767,
+      "step": 30890
+    },
+    {
+      "epoch": 4.995554118502951,
+      "grad_norm": 0.9275050759315491,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 30900
+    },
+    {
+      "epoch": 4.997170802683696,
+      "grad_norm": 0.94013911485672,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 30910
+    },
+    {
+      "epoch": 4.998787486864441,
+      "grad_norm": 0.9771268367767334,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 30920
+    },
+    {
+      "epoch": 4.9999191657909625,
+      "eval_loss": 1.1968598365783691,
+      "eval_runtime": 122.2519,
+      "eval_samples_per_second": 5.996,
+      "eval_steps_per_second": 0.753,
+      "step": 30927
+    },
+    {
+      "epoch": 5.0004041710451865,
+      "grad_norm": 0.8021580576896667,
+      "learning_rate": 0.0002,
+      "loss": 0.5238,
+      "step": 30930
+    },
+    {
+      "epoch": 5.002020855225932,
+      "grad_norm": 1.0807327032089233,
+      "learning_rate": 0.0002,
+      "loss": 0.4984,
+      "step": 30940
+    },
+    {
+      "epoch": 5.003637539406677,
+      "grad_norm": 1.1638425588607788,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 30950
+    },
+    {
+      "epoch": 5.005254223587422,
+      "grad_norm": 1.1700230836868286,
+      "learning_rate": 0.0002,
+      "loss": 0.4621,
+      "step": 30960
+    },
+    {
+      "epoch": 5.0068709077681675,
+      "grad_norm": 0.9053420424461365,
+      "learning_rate": 0.0002,
+      "loss": 0.4657,
+      "step": 30970
+    },
+    {
+      "epoch": 5.008487591948913,
+      "grad_norm": 0.9226111769676208,
+      "learning_rate": 0.0002,
+      "loss": 0.4865,
+      "step": 30980
+    },
+    {
+      "epoch": 5.010104276129658,
+      "grad_norm": 1.238669514656067,
+      "learning_rate": 0.0002,
+      "loss": 0.5011,
+      "step": 30990
+    },
+    {
+      "epoch": 5.011720960310403,
+      "grad_norm": 1.0668327808380127,
+      "learning_rate": 0.0002,
+      "loss": 0.4754,
+      "step": 31000
+    },
+    {
+      "epoch": 5.013337644491148,
+      "grad_norm": 1.0903944969177246,
+      "learning_rate": 0.0002,
+      "loss": 0.5414,
+      "step": 31010
+    },
+    {
+      "epoch": 5.014954328671894,
+      "grad_norm": 1.0763911008834839,
+      "learning_rate": 0.0002,
+      "loss": 0.5117,
+      "step": 31020
+    },
+    {
+      "epoch": 5.016571012852639,
+      "grad_norm": 1.0108771324157715,
+      "learning_rate": 0.0002,
+      "loss": 0.4908,
+      "step": 31030
+    },
+    {
+      "epoch": 5.018187697033385,
+      "grad_norm": 0.8816103935241699,
+      "learning_rate": 0.0002,
+      "loss": 0.5052,
+      "step": 31040
+    },
+    {
+      "epoch": 5.01980438121413,
+      "grad_norm": 1.11434805393219,
+      "learning_rate": 0.0002,
+      "loss": 0.4985,
+      "step": 31050
+    },
+    {
+      "epoch": 5.021421065394875,
+      "grad_norm": 1.0727789402008057,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 31060
+    },
+    {
+      "epoch": 5.023037749575621,
+      "grad_norm": 1.1480379104614258,
+      "learning_rate": 0.0002,
+      "loss": 0.4938,
+      "step": 31070
+    },
+    {
+      "epoch": 5.024654433756366,
+      "grad_norm": 1.0913071632385254,
+      "learning_rate": 0.0002,
+      "loss": 0.491,
+      "step": 31080
+    },
+    {
+      "epoch": 5.026271117937111,
+      "grad_norm": 0.9891864657402039,
+      "learning_rate": 0.0002,
+      "loss": 0.4896,
+      "step": 31090
+    },
+    {
+      "epoch": 5.027887802117856,
+      "grad_norm": 0.9167473912239075,
+      "learning_rate": 0.0002,
+      "loss": 0.4965,
+      "step": 31100
+    },
+    {
+      "epoch": 5.029504486298602,
+      "grad_norm": 1.2259035110473633,
+      "learning_rate": 0.0002,
+      "loss": 0.5098,
+      "step": 31110
+    },
+    {
+      "epoch": 5.031121170479347,
+      "grad_norm": 1.1812787055969238,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 31120
+    },
+    {
+      "epoch": 5.032737854660092,
+      "grad_norm": 1.0890522003173828,
+      "learning_rate": 0.0002,
+      "loss": 0.4725,
+      "step": 31130
+    },
+    {
+      "epoch": 5.034354538840837,
+      "grad_norm": 1.0521091222763062,
+      "learning_rate": 0.0002,
+      "loss": 0.4768,
+      "step": 31140
+    },
+    {
+      "epoch": 5.0359712230215825,
+      "grad_norm": 1.1274569034576416,
+      "learning_rate": 0.0002,
+      "loss": 0.4718,
+      "step": 31150
+    },
+    {
+      "epoch": 5.037587907202328,
+      "grad_norm": 1.140974998474121,
+      "learning_rate": 0.0002,
+      "loss": 0.4604,
+      "step": 31160
+    },
+    {
+      "epoch": 5.039204591383073,
+      "grad_norm": 1.1215609312057495,
+      "learning_rate": 0.0002,
+      "loss": 0.5077,
+      "step": 31170
+    },
+    {
+      "epoch": 5.040821275563818,
+      "grad_norm": 1.0107218027114868,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 31180
+    },
+    {
+      "epoch": 5.042437959744564,
+      "grad_norm": 1.0198770761489868,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 31190
+    },
+    {
+      "epoch": 5.0440546439253096,
+      "grad_norm": 1.1613430976867676,
+      "learning_rate": 0.0002,
+      "loss": 0.5004,
+      "step": 31200
+    },
+    {
+      "epoch": 5.045671328106055,
+      "grad_norm": 0.8555458188056946,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 31210
+    },
+    {
+      "epoch": 5.0472880122868,
+      "grad_norm": 1.0235545635223389,
+      "learning_rate": 0.0002,
+      "loss": 0.4878,
+      "step": 31220
+    },
+    {
+      "epoch": 5.048904696467545,
+      "grad_norm": 1.0228750705718994,
+      "learning_rate": 0.0002,
+      "loss": 0.499,
+      "step": 31230
+    },
+    {
+      "epoch": 5.0505213806482905,
+      "grad_norm": 0.8216419816017151,
+      "learning_rate": 0.0002,
+      "loss": 0.4544,
+      "step": 31240
+    },
+    {
+      "epoch": 5.052138064829036,
+      "grad_norm": 0.925828218460083,
+      "learning_rate": 0.0002,
+      "loss": 0.4947,
+      "step": 31250
+    },
+    {
+      "epoch": 5.053754749009781,
+      "grad_norm": 0.9229369759559631,
+      "learning_rate": 0.0002,
+      "loss": 0.4835,
+      "step": 31260
+    },
+    {
+      "epoch": 5.055371433190526,
+      "grad_norm": 0.9531727433204651,
+      "learning_rate": 0.0002,
+      "loss": 0.5136,
+      "step": 31270
+    },
+    {
+      "epoch": 5.056988117371271,
+      "grad_norm": 0.7738548517227173,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 31280
+    },
+    {
+      "epoch": 5.058604801552017,
+      "grad_norm": 1.0551451444625854,
+      "learning_rate": 0.0002,
+      "loss": 0.5166,
+      "step": 31290
+    },
+    {
+      "epoch": 5.060221485732762,
+      "grad_norm": 0.9782299399375916,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 31300
+    },
+    {
+      "epoch": 5.061838169913507,
+      "grad_norm": 1.0220632553100586,
+      "learning_rate": 0.0002,
+      "loss": 0.4776,
+      "step": 31310
+    },
+    {
+      "epoch": 5.063454854094252,
+      "grad_norm": 0.9808892607688904,
+      "learning_rate": 0.0002,
+      "loss": 0.5117,
+      "step": 31320
+    },
+    {
+      "epoch": 5.065071538274998,
+      "grad_norm": 1.0662003755569458,
+      "learning_rate": 0.0002,
+      "loss": 0.501,
+      "step": 31330
+    },
+    {
+      "epoch": 5.066688222455744,
+      "grad_norm": 1.0036940574645996,
+      "learning_rate": 0.0002,
+      "loss": 0.4844,
+      "step": 31340
+    },
+    {
+      "epoch": 5.068304906636489,
+      "grad_norm": 1.1931052207946777,
+      "learning_rate": 0.0002,
+      "loss": 0.5299,
+      "step": 31350
+    },
+    {
+      "epoch": 5.069921590817234,
+      "grad_norm": 0.9370693564414978,
+      "learning_rate": 0.0002,
+      "loss": 0.4646,
+      "step": 31360
+    },
+    {
+      "epoch": 5.071538274997979,
+      "grad_norm": 0.9589039087295532,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 31370
+    },
+    {
+      "epoch": 5.073154959178725,
+      "grad_norm": 1.0052711963653564,
+      "learning_rate": 0.0002,
+      "loss": 0.4669,
+      "step": 31380
+    },
+    {
+      "epoch": 5.07477164335947,
+      "grad_norm": 0.9991368651390076,
+      "learning_rate": 0.0002,
+      "loss": 0.5283,
+      "step": 31390
+    },
+    {
+      "epoch": 5.076388327540215,
+      "grad_norm": 0.8539695739746094,
+      "learning_rate": 0.0002,
+      "loss": 0.4579,
+      "step": 31400
+    },
+    {
+      "epoch": 5.07800501172096,
+      "grad_norm": 1.048775553703308,
+      "learning_rate": 0.0002,
+      "loss": 0.4609,
+      "step": 31410
+    },
+    {
+      "epoch": 5.0796216959017055,
+      "grad_norm": 0.9983724355697632,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 31420
+    },
+    {
+      "epoch": 5.081238380082451,
+      "grad_norm": 1.0189813375473022,
+      "learning_rate": 0.0002,
+      "loss": 0.4594,
+      "step": 31430
+    },
+    {
+      "epoch": 5.082855064263196,
+      "grad_norm": 0.9781646728515625,
+      "learning_rate": 0.0002,
+      "loss": 0.5449,
+      "step": 31440
+    },
+    {
+      "epoch": 5.084471748443941,
+      "grad_norm": 0.9424566030502319,
+      "learning_rate": 0.0002,
+      "loss": 0.4698,
+      "step": 31450
+    },
+    {
+      "epoch": 5.0860884326246865,
+      "grad_norm": 1.0036484003067017,
+      "learning_rate": 0.0002,
+      "loss": 0.4768,
+      "step": 31460
+    },
+    {
+      "epoch": 5.087705116805432,
+      "grad_norm": 1.0983147621154785,
+      "learning_rate": 0.0002,
+      "loss": 0.487,
+      "step": 31470
+    },
+    {
+      "epoch": 5.089321800986177,
+      "grad_norm": 1.0856730937957764,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 31480
+    },
+    {
+      "epoch": 5.090938485166923,
+      "grad_norm": 1.2191699743270874,
+      "learning_rate": 0.0002,
+      "loss": 0.485,
+      "step": 31490
+    },
+    {
+      "epoch": 5.092555169347668,
+      "grad_norm": 0.939346194267273,
+      "learning_rate": 0.0002,
+      "loss": 0.4936,
+      "step": 31500
+    },
+    {
+      "epoch": 5.0941718535284135,
+      "grad_norm": 0.9730121493339539,
+      "learning_rate": 0.0002,
+      "loss": 0.5107,
+      "step": 31510
+    },
+    {
+      "epoch": 5.095788537709159,
+      "grad_norm": 0.923686146736145,
+      "learning_rate": 0.0002,
+      "loss": 0.4973,
+      "step": 31520
+    },
+    {
+      "epoch": 5.097405221889904,
+      "grad_norm": 1.1734349727630615,
+      "learning_rate": 0.0002,
+      "loss": 0.4906,
+      "step": 31530
+    },
+    {
+      "epoch": 5.099021906070649,
+      "grad_norm": 1.084509015083313,
+      "learning_rate": 0.0002,
+      "loss": 0.5165,
+      "step": 31540
+    },
+    {
+      "epoch": 5.100638590251394,
+      "grad_norm": 1.0144678354263306,
+      "learning_rate": 0.0002,
+      "loss": 0.5078,
+      "step": 31550
+    },
+    {
+      "epoch": 5.10225527443214,
+      "grad_norm": 0.9958019256591797,
+      "learning_rate": 0.0002,
+      "loss": 0.4719,
+      "step": 31560
+    },
+    {
+      "epoch": 5.103871958612885,
+      "grad_norm": 0.8900736570358276,
+      "learning_rate": 0.0002,
+      "loss": 0.4876,
+      "step": 31570
+    },
+    {
+      "epoch": 5.10548864279363,
+      "grad_norm": 1.0921649932861328,
+      "learning_rate": 0.0002,
+      "loss": 0.463,
+      "step": 31580
+    },
+    {
+      "epoch": 5.107105326974375,
+      "grad_norm": 1.1613792181015015,
+      "learning_rate": 0.0002,
+      "loss": 0.5148,
+      "step": 31590
+    },
+    {
+      "epoch": 5.108722011155121,
+      "grad_norm": 0.9211367964744568,
+      "learning_rate": 0.0002,
+      "loss": 0.5055,
+      "step": 31600
+    },
+    {
+      "epoch": 5.110338695335866,
+      "grad_norm": 1.3315813541412354,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 31610
+    },
+    {
+      "epoch": 5.111955379516611,
+      "grad_norm": 1.3765019178390503,
+      "learning_rate": 0.0002,
+      "loss": 0.5336,
+      "step": 31620
+    },
+    {
+      "epoch": 5.113572063697356,
+      "grad_norm": 1.070198893547058,
+      "learning_rate": 0.0002,
+      "loss": 0.4861,
+      "step": 31630
+    },
+    {
+      "epoch": 5.115188747878102,
+      "grad_norm": 0.947631299495697,
+      "learning_rate": 0.0002,
+      "loss": 0.5046,
+      "step": 31640
+    },
+    {
+      "epoch": 5.116805432058848,
+      "grad_norm": 1.0197371244430542,
+      "learning_rate": 0.0002,
+      "loss": 0.5297,
+      "step": 31650
+    },
+    {
+      "epoch": 5.118422116239593,
+      "grad_norm": 0.8647911548614502,
+      "learning_rate": 0.0002,
+      "loss": 0.5014,
+      "step": 31660
+    },
+    {
+      "epoch": 5.120038800420338,
+      "grad_norm": 0.8944075107574463,
+      "learning_rate": 0.0002,
+      "loss": 0.4705,
+      "step": 31670
+    },
+    {
+      "epoch": 5.121655484601083,
+      "grad_norm": 1.124497652053833,
+      "learning_rate": 0.0002,
+      "loss": 0.5175,
+      "step": 31680
+    },
+    {
+      "epoch": 5.123272168781829,
+      "grad_norm": 0.893131673336029,
+      "learning_rate": 0.0002,
+      "loss": 0.5109,
+      "step": 31690
+    },
+    {
+      "epoch": 5.124888852962574,
+      "grad_norm": 1.0122284889221191,
+      "learning_rate": 0.0002,
+      "loss": 0.4937,
+      "step": 31700
+    },
+    {
+      "epoch": 5.126505537143319,
+      "grad_norm": 0.9493719935417175,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 31710
+    },
+    {
+      "epoch": 5.128122221324064,
+      "grad_norm": 0.9700539112091064,
+      "learning_rate": 0.0002,
+      "loss": 0.5031,
+      "step": 31720
+    },
+    {
+      "epoch": 5.1297389055048095,
+      "grad_norm": 1.111677646636963,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 31730
+    },
+    {
+      "epoch": 5.131355589685555,
+      "grad_norm": 0.8204274773597717,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 31740
+    },
+    {
+      "epoch": 5.1329722738663,
+      "grad_norm": 1.1029267311096191,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 31750
+    },
+    {
+      "epoch": 5.134588958047045,
+      "grad_norm": 1.065575122833252,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 31760
+    },
+    {
+      "epoch": 5.13620564222779,
+      "grad_norm": 0.8208706974983215,
+      "learning_rate": 0.0002,
+      "loss": 0.502,
+      "step": 31770
+    },
+    {
+      "epoch": 5.137822326408536,
+      "grad_norm": 1.0520979166030884,
+      "learning_rate": 0.0002,
+      "loss": 0.5352,
+      "step": 31780
+    },
+    {
+      "epoch": 5.139439010589282,
+      "grad_norm": 0.8585538268089294,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 31790
+    },
+    {
+      "epoch": 5.141055694770027,
+      "grad_norm": 1.1491447687149048,
+      "learning_rate": 0.0002,
+      "loss": 0.5159,
+      "step": 31800
+    },
+    {
+      "epoch": 5.142672378950772,
+      "grad_norm": 0.9441081285476685,
+      "learning_rate": 0.0002,
+      "loss": 0.5157,
+      "step": 31810
+    },
+    {
+      "epoch": 5.1442890631315175,
+      "grad_norm": 1.4146889448165894,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 31820
+    },
+    {
+      "epoch": 5.145905747312263,
+      "grad_norm": 1.0326547622680664,
+      "learning_rate": 0.0002,
+      "loss": 0.5159,
+      "step": 31830
+    },
+    {
+      "epoch": 5.147522431493008,
+      "grad_norm": 0.9879202842712402,
+      "learning_rate": 0.0002,
+      "loss": 0.5348,
+      "step": 31840
+    },
+    {
+      "epoch": 5.149139115673753,
+      "grad_norm": 1.0374281406402588,
+      "learning_rate": 0.0002,
+      "loss": 0.5083,
+      "step": 31850
+    },
+    {
+      "epoch": 5.150755799854498,
+      "grad_norm": 1.181229591369629,
+      "learning_rate": 0.0002,
+      "loss": 0.4827,
+      "step": 31860
+    },
+    {
+      "epoch": 5.152372484035244,
+      "grad_norm": 1.2078537940979004,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 31870
+    },
+    {
+      "epoch": 5.153989168215989,
+      "grad_norm": 0.9599190354347229,
+      "learning_rate": 0.0002,
+      "loss": 0.5329,
+      "step": 31880
+    },
+    {
+      "epoch": 5.155605852396734,
+      "grad_norm": 1.0378568172454834,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 31890
+    },
+    {
+      "epoch": 5.157222536577479,
+      "grad_norm": 0.8746536374092102,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 31900
+    },
+    {
+      "epoch": 5.1588392207582245,
+      "grad_norm": 1.0232136249542236,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 31910
+    },
+    {
+      "epoch": 5.16045590493897,
+      "grad_norm": 0.9827565550804138,
+      "learning_rate": 0.0002,
+      "loss": 0.4844,
+      "step": 31920
+    },
+    {
+      "epoch": 5.162072589119716,
+      "grad_norm": 1.342657208442688,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 31930
+    },
+    {
+      "epoch": 5.163689273300461,
+      "grad_norm": 1.18390691280365,
+      "learning_rate": 0.0002,
+      "loss": 0.513,
+      "step": 31940
+    },
+    {
+      "epoch": 5.165305957481206,
+      "grad_norm": 0.996350109577179,
+      "learning_rate": 0.0002,
+      "loss": 0.5267,
+      "step": 31950
+    },
+    {
+      "epoch": 5.166922641661952,
+      "grad_norm": 0.9710391163825989,
+      "learning_rate": 0.0002,
+      "loss": 0.5063,
+      "step": 31960
+    },
+    {
+      "epoch": 5.168539325842697,
+      "grad_norm": 1.0264002084732056,
+      "learning_rate": 0.0002,
+      "loss": 0.5115,
+      "step": 31970
+    },
+    {
+      "epoch": 5.170156010023442,
+      "grad_norm": 1.0028311014175415,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 31980
+    },
+    {
+      "epoch": 5.171772694204187,
+      "grad_norm": 1.1078234910964966,
+      "learning_rate": 0.0002,
+      "loss": 0.5103,
+      "step": 31990
+    },
+    {
+      "epoch": 5.1733893783849325,
+      "grad_norm": 0.9659610390663147,
+      "learning_rate": 0.0002,
+      "loss": 0.495,
+      "step": 32000
+    },
+    {
+      "epoch": 5.175006062565678,
+      "grad_norm": 0.841986894607544,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 32010
+    },
+    {
+      "epoch": 5.176622746746423,
+      "grad_norm": 1.095332384109497,
+      "learning_rate": 0.0002,
+      "loss": 0.48,
+      "step": 32020
+    },
+    {
+      "epoch": 5.178239430927168,
+      "grad_norm": 1.1242377758026123,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 32030
+    },
+    {
+      "epoch": 5.179856115107913,
+      "grad_norm": 0.9872292280197144,
+      "learning_rate": 0.0002,
+      "loss": 0.5573,
+      "step": 32040
+    },
+    {
+      "epoch": 5.181472799288659,
+      "grad_norm": 0.936161994934082,
+      "learning_rate": 0.0002,
+      "loss": 0.48,
+      "step": 32050
+    },
+    {
+      "epoch": 5.183089483469404,
+      "grad_norm": 1.166100025177002,
+      "learning_rate": 0.0002,
+      "loss": 0.5093,
+      "step": 32060
+    },
+    {
+      "epoch": 5.184706167650149,
+      "grad_norm": 1.0764425992965698,
+      "learning_rate": 0.0002,
+      "loss": 0.5438,
+      "step": 32070
+    },
+    {
+      "epoch": 5.186322851830895,
+      "grad_norm": 1.0480051040649414,
+      "learning_rate": 0.0002,
+      "loss": 0.4843,
+      "step": 32080
+    },
+    {
+      "epoch": 5.1879395360116405,
+      "grad_norm": 1.0874916315078735,
+      "learning_rate": 0.0002,
+      "loss": 0.5386,
+      "step": 32090
+    },
+    {
+      "epoch": 5.189556220192386,
+      "grad_norm": 1.0817396640777588,
+      "learning_rate": 0.0002,
+      "loss": 0.4975,
+      "step": 32100
+    },
+    {
+      "epoch": 5.191172904373131,
+      "grad_norm": 1.054111361503601,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 32110
+    },
+    {
+      "epoch": 5.192789588553876,
+      "grad_norm": 0.9655823707580566,
+      "learning_rate": 0.0002,
+      "loss": 0.5229,
+      "step": 32120
+    },
+    {
+      "epoch": 5.194406272734621,
+      "grad_norm": 1.1384109258651733,
+      "learning_rate": 0.0002,
+      "loss": 0.5105,
+      "step": 32130
+    },
+    {
+      "epoch": 5.196022956915367,
+      "grad_norm": 1.0149348974227905,
+      "learning_rate": 0.0002,
+      "loss": 0.5073,
+      "step": 32140
+    },
+    {
+      "epoch": 5.197639641096112,
+      "grad_norm": 1.1084046363830566,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 32150
+    },
+    {
+      "epoch": 5.199256325276857,
+      "grad_norm": 1.1209309101104736,
+      "learning_rate": 0.0002,
+      "loss": 0.4936,
+      "step": 32160
+    },
+    {
+      "epoch": 5.200873009457602,
+      "grad_norm": 1.133089542388916,
+      "learning_rate": 0.0002,
+      "loss": 0.5101,
+      "step": 32170
+    },
+    {
+      "epoch": 5.202489693638348,
+      "grad_norm": 1.0893020629882812,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 32180
+    },
+    {
+      "epoch": 5.204106377819093,
+      "grad_norm": 0.90018630027771,
+      "learning_rate": 0.0002,
+      "loss": 0.4872,
+      "step": 32190
+    },
+    {
+      "epoch": 5.205723061999838,
+      "grad_norm": 0.977622926235199,
+      "learning_rate": 0.0002,
+      "loss": 0.4999,
+      "step": 32200
+    },
+    {
+      "epoch": 5.207339746180583,
+      "grad_norm": 1.2940177917480469,
+      "learning_rate": 0.0002,
+      "loss": 0.5028,
+      "step": 32210
+    },
+    {
+      "epoch": 5.2089564303613285,
+      "grad_norm": 1.2131710052490234,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 32220
+    },
+    {
+      "epoch": 5.210573114542075,
+      "grad_norm": 1.0234841108322144,
+      "learning_rate": 0.0002,
+      "loss": 0.5189,
+      "step": 32230
+    },
+    {
+      "epoch": 5.21218979872282,
+      "grad_norm": 1.157975435256958,
+      "learning_rate": 0.0002,
+      "loss": 0.5424,
+      "step": 32240
+    },
+    {
+      "epoch": 5.213806482903565,
+      "grad_norm": 1.0381282567977905,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 32250
+    },
+    {
+      "epoch": 5.21542316708431,
+      "grad_norm": 1.0125395059585571,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 32260
+    },
+    {
+      "epoch": 5.2170398512650555,
+      "grad_norm": 1.272691011428833,
+      "learning_rate": 0.0002,
+      "loss": 0.5216,
+      "step": 32270
+    },
+    {
+      "epoch": 5.218656535445801,
+      "grad_norm": 1.0061250925064087,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 32280
+    },
+    {
+      "epoch": 5.220273219626546,
+      "grad_norm": 0.9752234816551208,
+      "learning_rate": 0.0002,
+      "loss": 0.4739,
+      "step": 32290
+    },
+    {
+      "epoch": 5.221889903807291,
+      "grad_norm": 1.1193140745162964,
+      "learning_rate": 0.0002,
+      "loss": 0.5471,
+      "step": 32300
+    },
+    {
+      "epoch": 5.2235065879880365,
+      "grad_norm": 1.0126434564590454,
+      "learning_rate": 0.0002,
+      "loss": 0.4976,
+      "step": 32310
+    },
+    {
+      "epoch": 5.225123272168782,
+      "grad_norm": 1.4338394403457642,
+      "learning_rate": 0.0002,
+      "loss": 0.5257,
+      "step": 32320
+    },
+    {
+      "epoch": 5.226739956349527,
+      "grad_norm": 1.004101276397705,
+      "learning_rate": 0.0002,
+      "loss": 0.5235,
+      "step": 32330
+    },
+    {
+      "epoch": 5.228356640530272,
+      "grad_norm": 0.8744166493415833,
+      "learning_rate": 0.0002,
+      "loss": 0.5091,
+      "step": 32340
+    },
+    {
+      "epoch": 5.229973324711017,
+      "grad_norm": 1.0165376663208008,
+      "learning_rate": 0.0002,
+      "loss": 0.5388,
+      "step": 32350
+    },
+    {
+      "epoch": 5.231590008891763,
+      "grad_norm": 0.8635954260826111,
+      "learning_rate": 0.0002,
+      "loss": 0.5469,
+      "step": 32360
+    },
+    {
+      "epoch": 5.233206693072509,
+      "grad_norm": 1.1392399072647095,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 32370
+    },
+    {
+      "epoch": 5.234823377253254,
+      "grad_norm": 1.0202113389968872,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 32380
+    },
+    {
+      "epoch": 5.236440061433999,
+      "grad_norm": 1.0417983531951904,
+      "learning_rate": 0.0002,
+      "loss": 0.4983,
+      "step": 32390
+    },
+    {
+      "epoch": 5.238056745614744,
+      "grad_norm": 0.8729333877563477,
+      "learning_rate": 0.0002,
+      "loss": 0.507,
+      "step": 32400
+    },
+    {
+      "epoch": 5.23967342979549,
+      "grad_norm": 1.1626229286193848,
+      "learning_rate": 0.0002,
+      "loss": 0.5426,
+      "step": 32410
+    },
+    {
+      "epoch": 5.241290113976235,
+      "grad_norm": 0.9086161851882935,
+      "learning_rate": 0.0002,
+      "loss": 0.5355,
+      "step": 32420
+    },
+    {
+      "epoch": 5.24290679815698,
+      "grad_norm": 1.3999892473220825,
+      "learning_rate": 0.0002,
+      "loss": 0.4927,
+      "step": 32430
+    },
+    {
+      "epoch": 5.244523482337725,
+      "grad_norm": 1.0356311798095703,
+      "learning_rate": 0.0002,
+      "loss": 0.4795,
+      "step": 32440
+    },
+    {
+      "epoch": 5.246140166518471,
+      "grad_norm": 0.9655531644821167,
+      "learning_rate": 0.0002,
+      "loss": 0.5035,
+      "step": 32450
+    },
+    {
+      "epoch": 5.247756850699216,
+      "grad_norm": 1.0411828756332397,
+      "learning_rate": 0.0002,
+      "loss": 0.5166,
+      "step": 32460
+    },
+    {
+      "epoch": 5.249373534879961,
+      "grad_norm": 1.1199816465377808,
+      "learning_rate": 0.0002,
+      "loss": 0.5141,
+      "step": 32470
+    },
+    {
+      "epoch": 5.250990219060706,
+      "grad_norm": 1.260321855545044,
+      "learning_rate": 0.0002,
+      "loss": 0.4864,
+      "step": 32480
+    },
+    {
+      "epoch": 5.2526069032414515,
+      "grad_norm": 1.2950857877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.4893,
+      "step": 32490
+    },
+    {
+      "epoch": 5.254223587422197,
+      "grad_norm": 0.8982820510864258,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 32500
+    },
+    {
+      "epoch": 5.255840271602942,
+      "grad_norm": 0.8512987494468689,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 32510
+    },
+    {
+      "epoch": 5.257456955783688,
+      "grad_norm": 1.067443609237671,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 32520
+    },
+    {
+      "epoch": 5.259073639964433,
+      "grad_norm": 1.0957417488098145,
+      "learning_rate": 0.0002,
+      "loss": 0.4928,
+      "step": 32530
+    },
+    {
+      "epoch": 5.260690324145179,
+      "grad_norm": 1.4161807298660278,
+      "learning_rate": 0.0002,
+      "loss": 0.5169,
+      "step": 32540
+    },
+    {
+      "epoch": 5.262307008325924,
+      "grad_norm": 1.2264093160629272,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 32550
+    },
+    {
+      "epoch": 5.263923692506669,
+      "grad_norm": 1.0015931129455566,
+      "learning_rate": 0.0002,
+      "loss": 0.5221,
+      "step": 32560
+    },
+    {
+      "epoch": 5.265540376687414,
+      "grad_norm": 1.0743094682693481,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 32570
+    },
+    {
+      "epoch": 5.2671570608681595,
+      "grad_norm": 1.1386840343475342,
+      "learning_rate": 0.0002,
+      "loss": 0.5289,
+      "step": 32580
+    },
+    {
+      "epoch": 5.268773745048905,
+      "grad_norm": 1.0093860626220703,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 32590
+    },
+    {
+      "epoch": 5.27039042922965,
+      "grad_norm": 0.9593744874000549,
+      "learning_rate": 0.0002,
+      "loss": 0.5175,
+      "step": 32600
+    },
+    {
+      "epoch": 5.272007113410395,
+      "grad_norm": 1.146021842956543,
+      "learning_rate": 0.0002,
+      "loss": 0.528,
+      "step": 32610
+    },
+    {
+      "epoch": 5.27362379759114,
+      "grad_norm": 0.9579031467437744,
+      "learning_rate": 0.0002,
+      "loss": 0.4983,
+      "step": 32620
+    },
+    {
+      "epoch": 5.275240481771886,
+      "grad_norm": 1.0548793077468872,
+      "learning_rate": 0.0002,
+      "loss": 0.5376,
+      "step": 32630
+    },
+    {
+      "epoch": 5.276857165952631,
+      "grad_norm": 1.0380561351776123,
+      "learning_rate": 0.0002,
+      "loss": 0.5267,
+      "step": 32640
+    },
+    {
+      "epoch": 5.278473850133376,
+      "grad_norm": 1.2119969129562378,
+      "learning_rate": 0.0002,
+      "loss": 0.5182,
+      "step": 32650
+    },
+    {
+      "epoch": 5.280090534314121,
+      "grad_norm": 1.0507797002792358,
+      "learning_rate": 0.0002,
+      "loss": 0.5298,
+      "step": 32660
+    },
+    {
+      "epoch": 5.2817072184948675,
+      "grad_norm": 1.0185176134109497,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 32670
+    },
+    {
+      "epoch": 5.283323902675613,
+      "grad_norm": 1.2358098030090332,
+      "learning_rate": 0.0002,
+      "loss": 0.4904,
+      "step": 32680
+    },
+    {
+      "epoch": 5.284940586856358,
+      "grad_norm": 0.7937114238739014,
+      "learning_rate": 0.0002,
+      "loss": 0.5169,
+      "step": 32690
+    },
+    {
+      "epoch": 5.286557271037103,
+      "grad_norm": 0.9825124740600586,
+      "learning_rate": 0.0002,
+      "loss": 0.495,
+      "step": 32700
+    },
+    {
+      "epoch": 5.288173955217848,
+      "grad_norm": 1.2059301137924194,
+      "learning_rate": 0.0002,
+      "loss": 0.5149,
+      "step": 32710
+    },
+    {
+      "epoch": 5.289790639398594,
+      "grad_norm": 1.0828571319580078,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 32720
+    },
+    {
+      "epoch": 5.291407323579339,
+      "grad_norm": 1.0129735469818115,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 32730
+    },
+    {
+      "epoch": 5.293024007760084,
+      "grad_norm": 1.0591634511947632,
+      "learning_rate": 0.0002,
+      "loss": 0.5216,
+      "step": 32740
+    },
+    {
+      "epoch": 5.294640691940829,
+      "grad_norm": 0.9256815910339355,
+      "learning_rate": 0.0002,
+      "loss": 0.522,
+      "step": 32750
+    },
+    {
+      "epoch": 5.2962573761215745,
+      "grad_norm": 1.0928633213043213,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 32760
+    },
+    {
+      "epoch": 5.29787406030232,
+      "grad_norm": 0.9415594935417175,
+      "learning_rate": 0.0002,
+      "loss": 0.5093,
+      "step": 32770
+    },
+    {
+      "epoch": 5.299490744483065,
+      "grad_norm": 1.141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.5252,
+      "step": 32780
+    },
+    {
+      "epoch": 5.30110742866381,
+      "grad_norm": 1.0646510124206543,
+      "learning_rate": 0.0002,
+      "loss": 0.4837,
+      "step": 32790
+    },
+    {
+      "epoch": 5.3027241128445555,
+      "grad_norm": 1.189661979675293,
+      "learning_rate": 0.0002,
+      "loss": 0.5547,
+      "step": 32800
+    },
+    {
+      "epoch": 5.304340797025301,
+      "grad_norm": 0.9568731188774109,
+      "learning_rate": 0.0002,
+      "loss": 0.5664,
+      "step": 32810
+    },
+    {
+      "epoch": 5.305957481206047,
+      "grad_norm": 1.1556824445724487,
+      "learning_rate": 0.0002,
+      "loss": 0.5344,
+      "step": 32820
+    },
+    {
+      "epoch": 5.307574165386792,
+      "grad_norm": 0.9353463649749756,
+      "learning_rate": 0.0002,
+      "loss": 0.4894,
+      "step": 32830
+    },
+    {
+      "epoch": 5.309190849567537,
+      "grad_norm": 1.1208295822143555,
+      "learning_rate": 0.0002,
+      "loss": 0.5052,
+      "step": 32840
+    },
+    {
+      "epoch": 5.3108075337482825,
+      "grad_norm": 1.0894153118133545,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 32850
+    },
+    {
+      "epoch": 5.312424217929028,
+      "grad_norm": 1.090329647064209,
+      "learning_rate": 0.0002,
+      "loss": 0.5046,
+      "step": 32860
+    },
+    {
+      "epoch": 5.314040902109773,
+      "grad_norm": 1.0781712532043457,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 32870
+    },
+    {
+      "epoch": 5.315657586290518,
+      "grad_norm": 1.1785295009613037,
+      "learning_rate": 0.0002,
+      "loss": 0.57,
+      "step": 32880
+    },
+    {
+      "epoch": 5.317274270471263,
+      "grad_norm": 1.0406851768493652,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 32890
+    },
+    {
+      "epoch": 5.318890954652009,
+      "grad_norm": 1.0982953310012817,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 32900
+    },
+    {
+      "epoch": 5.320507638832754,
+      "grad_norm": 1.2969383001327515,
+      "learning_rate": 0.0002,
+      "loss": 0.4944,
+      "step": 32910
+    },
+    {
+      "epoch": 5.322124323013499,
+      "grad_norm": 0.9687288999557495,
+      "learning_rate": 0.0002,
+      "loss": 0.4786,
+      "step": 32920
+    },
+    {
+      "epoch": 5.323741007194244,
+      "grad_norm": 1.136760950088501,
+      "learning_rate": 0.0002,
+      "loss": 0.5286,
+      "step": 32930
+    },
+    {
+      "epoch": 5.32535769137499,
+      "grad_norm": 1.3045495748519897,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 32940
+    },
+    {
+      "epoch": 5.326974375555735,
+      "grad_norm": 1.221675992012024,
+      "learning_rate": 0.0002,
+      "loss": 0.5413,
+      "step": 32950
+    },
+    {
+      "epoch": 5.32859105973648,
+      "grad_norm": 1.1380633115768433,
+      "learning_rate": 0.0002,
+      "loss": 0.4999,
+      "step": 32960
+    },
+    {
+      "epoch": 5.330207743917226,
+      "grad_norm": 1.1065956354141235,
+      "learning_rate": 0.0002,
+      "loss": 0.5037,
+      "step": 32970
+    },
+    {
+      "epoch": 5.331824428097971,
+      "grad_norm": 1.0187175273895264,
+      "learning_rate": 0.0002,
+      "loss": 0.4913,
+      "step": 32980
+    },
+    {
+      "epoch": 5.333441112278717,
+      "grad_norm": 0.9077118039131165,
+      "learning_rate": 0.0002,
+      "loss": 0.5234,
+      "step": 32990
+    },
+    {
+      "epoch": 5.335057796459462,
+      "grad_norm": 1.0092815160751343,
+      "learning_rate": 0.0002,
+      "loss": 0.5071,
+      "step": 33000
+    },
+    {
+      "epoch": 5.336674480640207,
+      "grad_norm": 1.0168777704238892,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 33010
+    },
+    {
+      "epoch": 5.338291164820952,
+      "grad_norm": 0.996161937713623,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 33020
+    },
+    {
+      "epoch": 5.339907849001698,
+      "grad_norm": 0.794463038444519,
+      "learning_rate": 0.0002,
+      "loss": 0.5024,
+      "step": 33030
+    },
+    {
+      "epoch": 5.341524533182443,
+      "grad_norm": 0.9750674962997437,
+      "learning_rate": 0.0002,
+      "loss": 0.5112,
+      "step": 33040
+    },
+    {
+      "epoch": 5.343141217363188,
+      "grad_norm": 1.2770029306411743,
+      "learning_rate": 0.0002,
+      "loss": 0.528,
+      "step": 33050
+    },
+    {
+      "epoch": 5.344757901543933,
+      "grad_norm": 1.1500186920166016,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 33060
+    },
+    {
+      "epoch": 5.3463745857246785,
+      "grad_norm": 1.0726377964019775,
+      "learning_rate": 0.0002,
+      "loss": 0.4906,
+      "step": 33070
+    },
+    {
+      "epoch": 5.347991269905424,
+      "grad_norm": 0.9314153790473938,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 33080
+    },
+    {
+      "epoch": 5.349607954086169,
+      "grad_norm": 1.344988465309143,
+      "learning_rate": 0.0002,
+      "loss": 0.5434,
+      "step": 33090
+    },
+    {
+      "epoch": 5.351224638266914,
+      "grad_norm": 0.863196611404419,
+      "learning_rate": 0.0002,
+      "loss": 0.4874,
+      "step": 33100
+    },
+    {
+      "epoch": 5.352841322447659,
+      "grad_norm": 1.128100037574768,
+      "learning_rate": 0.0002,
+      "loss": 0.534,
+      "step": 33110
+    },
+    {
+      "epoch": 5.3544580066284055,
+      "grad_norm": 1.1673583984375,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 33120
+    },
+    {
+      "epoch": 5.356074690809151,
+      "grad_norm": 0.9416789412498474,
+      "learning_rate": 0.0002,
+      "loss": 0.4787,
+      "step": 33130
+    },
+    {
+      "epoch": 5.357691374989896,
+      "grad_norm": 1.1855236291885376,
+      "learning_rate": 0.0002,
+      "loss": 0.5155,
+      "step": 33140
+    },
+    {
+      "epoch": 5.359308059170641,
+      "grad_norm": 1.0415170192718506,
+      "learning_rate": 0.0002,
+      "loss": 0.515,
+      "step": 33150
+    },
+    {
+      "epoch": 5.3609247433513865,
+      "grad_norm": 0.9953004121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.545,
+      "step": 33160
+    },
+    {
+      "epoch": 5.362541427532132,
+      "grad_norm": 0.96138596534729,
+      "learning_rate": 0.0002,
+      "loss": 0.5305,
+      "step": 33170
+    },
+    {
+      "epoch": 5.364158111712877,
+      "grad_norm": 1.341979742050171,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 33180
+    },
+    {
+      "epoch": 5.365774795893622,
+      "grad_norm": 1.0136911869049072,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 33190
+    },
+    {
+      "epoch": 5.367391480074367,
+      "grad_norm": 0.8685575127601624,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 33200
+    },
+    {
+      "epoch": 5.369008164255113,
+      "grad_norm": 0.8833574652671814,
+      "learning_rate": 0.0002,
+      "loss": 0.5146,
+      "step": 33210
+    },
+    {
+      "epoch": 5.370624848435858,
+      "grad_norm": 0.9123612642288208,
+      "learning_rate": 0.0002,
+      "loss": 0.4982,
+      "step": 33220
+    },
+    {
+      "epoch": 5.372241532616603,
+      "grad_norm": 1.2720599174499512,
+      "learning_rate": 0.0002,
+      "loss": 0.5047,
+      "step": 33230
+    },
+    {
+      "epoch": 5.373858216797348,
+      "grad_norm": 1.0596648454666138,
+      "learning_rate": 0.0002,
+      "loss": 0.5175,
+      "step": 33240
+    },
+    {
+      "epoch": 5.3754749009780936,
+      "grad_norm": 1.119701623916626,
+      "learning_rate": 0.0002,
+      "loss": 0.5284,
+      "step": 33250
+    },
+    {
+      "epoch": 5.377091585158839,
+      "grad_norm": 1.3000061511993408,
+      "learning_rate": 0.0002,
+      "loss": 0.5217,
+      "step": 33260
+    },
+    {
+      "epoch": 5.378708269339585,
+      "grad_norm": 1.083891749382019,
+      "learning_rate": 0.0002,
+      "loss": 0.5125,
+      "step": 33270
+    },
+    {
+      "epoch": 5.38032495352033,
+      "grad_norm": 0.9402718544006348,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 33280
+    },
+    {
+      "epoch": 5.381941637701075,
+      "grad_norm": 1.3376892805099487,
+      "learning_rate": 0.0002,
+      "loss": 0.5559,
+      "step": 33290
+    },
+    {
+      "epoch": 5.383558321881821,
+      "grad_norm": 1.1600074768066406,
+      "learning_rate": 0.0002,
+      "loss": 0.5193,
+      "step": 33300
+    },
+    {
+      "epoch": 5.385175006062566,
+      "grad_norm": 1.1449427604675293,
+      "learning_rate": 0.0002,
+      "loss": 0.4907,
+      "step": 33310
+    },
+    {
+      "epoch": 5.386791690243311,
+      "grad_norm": 1.3118891716003418,
+      "learning_rate": 0.0002,
+      "loss": 0.5449,
+      "step": 33320
+    },
+    {
+      "epoch": 5.388408374424056,
+      "grad_norm": 0.743449866771698,
+      "learning_rate": 0.0002,
+      "loss": 0.547,
+      "step": 33330
+    },
+    {
+      "epoch": 5.3900250586048015,
+      "grad_norm": 0.9358304142951965,
+      "learning_rate": 0.0002,
+      "loss": 0.5555,
+      "step": 33340
+    },
+    {
+      "epoch": 5.391641742785547,
+      "grad_norm": 1.0447142124176025,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 33350
+    },
+    {
+      "epoch": 5.393258426966292,
+      "grad_norm": 1.1088626384735107,
+      "learning_rate": 0.0002,
+      "loss": 0.5106,
+      "step": 33360
+    },
+    {
+      "epoch": 5.394875111147037,
+      "grad_norm": 1.1267958879470825,
+      "learning_rate": 0.0002,
+      "loss": 0.4929,
+      "step": 33370
+    },
+    {
+      "epoch": 5.3964917953277824,
+      "grad_norm": 0.9709370136260986,
+      "learning_rate": 0.0002,
+      "loss": 0.5165,
+      "step": 33380
+    },
+    {
+      "epoch": 5.398108479508528,
+      "grad_norm": 1.0939103364944458,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 33390
+    },
+    {
+      "epoch": 5.399725163689273,
+      "grad_norm": 0.9559304714202881,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 33400
+    },
+    {
+      "epoch": 5.401341847870018,
+      "grad_norm": 1.199580430984497,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 33410
+    },
+    {
+      "epoch": 5.402958532050764,
+      "grad_norm": 0.9097000360488892,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 33420
+    },
+    {
+      "epoch": 5.4045752162315095,
+      "grad_norm": 1.1940981149673462,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 33430
+    },
+    {
+      "epoch": 5.406191900412255,
+      "grad_norm": 1.0530916452407837,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 33440
+    },
+    {
+      "epoch": 5.407808584593,
+      "grad_norm": 1.0482549667358398,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 33450
+    },
+    {
+      "epoch": 5.409425268773745,
+      "grad_norm": 1.2524714469909668,
+      "learning_rate": 0.0002,
+      "loss": 0.501,
+      "step": 33460
+    },
+    {
+      "epoch": 5.41104195295449,
+      "grad_norm": 1.1091666221618652,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 33470
+    },
+    {
+      "epoch": 5.412658637135236,
+      "grad_norm": 0.9981587529182434,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 33480
+    },
+    {
+      "epoch": 5.414275321315981,
+      "grad_norm": 1.016681432723999,
+      "learning_rate": 0.0002,
+      "loss": 0.4977,
+      "step": 33490
+    },
+    {
+      "epoch": 5.415892005496726,
+      "grad_norm": 1.1456854343414307,
+      "learning_rate": 0.0002,
+      "loss": 0.5388,
+      "step": 33500
+    },
+    {
+      "epoch": 5.417508689677471,
+      "grad_norm": 1.1454259157180786,
+      "learning_rate": 0.0002,
+      "loss": 0.5292,
+      "step": 33510
+    },
+    {
+      "epoch": 5.419125373858217,
+      "grad_norm": 0.9858416318893433,
+      "learning_rate": 0.0002,
+      "loss": 0.5061,
+      "step": 33520
+    },
+    {
+      "epoch": 5.420742058038962,
+      "grad_norm": 0.9764766693115234,
+      "learning_rate": 0.0002,
+      "loss": 0.5139,
+      "step": 33530
+    },
+    {
+      "epoch": 5.422358742219707,
+      "grad_norm": 1.199920892715454,
+      "learning_rate": 0.0002,
+      "loss": 0.5518,
+      "step": 33540
+    },
+    {
+      "epoch": 5.423975426400452,
+      "grad_norm": 1.3107370138168335,
+      "learning_rate": 0.0002,
+      "loss": 0.5182,
+      "step": 33550
+    },
+    {
+      "epoch": 5.4255921105811975,
+      "grad_norm": 0.9637970328330994,
+      "learning_rate": 0.0002,
+      "loss": 0.5149,
+      "step": 33560
+    },
+    {
+      "epoch": 5.427208794761944,
+      "grad_norm": 1.023359775543213,
+      "learning_rate": 0.0002,
+      "loss": 0.526,
+      "step": 33570
+    },
+    {
+      "epoch": 5.428825478942689,
+      "grad_norm": 1.060417652130127,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 33580
+    },
+    {
+      "epoch": 5.430442163123434,
+      "grad_norm": 0.9971120953559875,
+      "learning_rate": 0.0002,
+      "loss": 0.5052,
+      "step": 33590
+    },
+    {
+      "epoch": 5.432058847304179,
+      "grad_norm": 0.9213743209838867,
+      "learning_rate": 0.0002,
+      "loss": 0.5044,
+      "step": 33600
+    },
+    {
+      "epoch": 5.4336755314849245,
+      "grad_norm": 1.1512309312820435,
+      "learning_rate": 0.0002,
+      "loss": 0.5714,
+      "step": 33610
+    },
+    {
+      "epoch": 5.43529221566567,
+      "grad_norm": 1.2198847532272339,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 33620
+    },
+    {
+      "epoch": 5.436908899846415,
+      "grad_norm": 1.0329595804214478,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 33630
+    },
+    {
+      "epoch": 5.43852558402716,
+      "grad_norm": 1.1075750589370728,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 33640
+    },
+    {
+      "epoch": 5.4401422682079055,
+      "grad_norm": 1.006342887878418,
+      "learning_rate": 0.0002,
+      "loss": 0.5295,
+      "step": 33650
+    },
+    {
+      "epoch": 5.441758952388651,
+      "grad_norm": 0.9179885983467102,
+      "learning_rate": 0.0002,
+      "loss": 0.5394,
+      "step": 33660
+    },
+    {
+      "epoch": 5.443375636569396,
+      "grad_norm": 1.2799493074417114,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 33670
+    },
+    {
+      "epoch": 5.444992320750141,
+      "grad_norm": 1.1153863668441772,
+      "learning_rate": 0.0002,
+      "loss": 0.5426,
+      "step": 33680
+    },
+    {
+      "epoch": 5.446609004930886,
+      "grad_norm": 1.0681028366088867,
+      "learning_rate": 0.0002,
+      "loss": 0.5087,
+      "step": 33690
+    },
+    {
+      "epoch": 5.448225689111632,
+      "grad_norm": 0.9788817167282104,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 33700
+    },
+    {
+      "epoch": 5.449842373292377,
+      "grad_norm": 0.8481608629226685,
+      "learning_rate": 0.0002,
+      "loss": 0.5308,
+      "step": 33710
+    },
+    {
+      "epoch": 5.451459057473123,
+      "grad_norm": 1.113756537437439,
+      "learning_rate": 0.0002,
+      "loss": 0.5225,
+      "step": 33720
+    },
+    {
+      "epoch": 5.453075741653868,
+      "grad_norm": 0.8425475358963013,
+      "learning_rate": 0.0002,
+      "loss": 0.5213,
+      "step": 33730
+    },
+    {
+      "epoch": 5.4546924258346134,
+      "grad_norm": 1.0852208137512207,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 33740
+    },
+    {
+      "epoch": 5.456309110015359,
+      "grad_norm": 1.1664748191833496,
+      "learning_rate": 0.0002,
+      "loss": 0.5535,
+      "step": 33750
+    },
+    {
+      "epoch": 5.457925794196104,
+      "grad_norm": 1.217241644859314,
+      "learning_rate": 0.0002,
+      "loss": 0.5419,
+      "step": 33760
+    },
+    {
+      "epoch": 5.459542478376849,
+      "grad_norm": 1.1572928428649902,
+      "learning_rate": 0.0002,
+      "loss": 0.5351,
+      "step": 33770
+    },
+    {
+      "epoch": 5.461159162557594,
+      "grad_norm": 1.0437318086624146,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 33780
+    },
+    {
+      "epoch": 5.46277584673834,
+      "grad_norm": 0.9807571768760681,
+      "learning_rate": 0.0002,
+      "loss": 0.5266,
+      "step": 33790
+    },
+    {
+      "epoch": 5.464392530919085,
+      "grad_norm": 1.1436342000961304,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 33800
+    },
+    {
+      "epoch": 5.46600921509983,
+      "grad_norm": 1.1004794836044312,
+      "learning_rate": 0.0002,
+      "loss": 0.5338,
+      "step": 33810
+    },
+    {
+      "epoch": 5.467625899280575,
+      "grad_norm": 1.2130268812179565,
+      "learning_rate": 0.0002,
+      "loss": 0.4868,
+      "step": 33820
+    },
+    {
+      "epoch": 5.4692425834613205,
+      "grad_norm": 1.3154419660568237,
+      "learning_rate": 0.0002,
+      "loss": 0.516,
+      "step": 33830
+    },
+    {
+      "epoch": 5.470859267642066,
+      "grad_norm": 0.7934383749961853,
+      "learning_rate": 0.0002,
+      "loss": 0.4934,
+      "step": 33840
+    },
+    {
+      "epoch": 5.472475951822812,
+      "grad_norm": 0.7838410139083862,
+      "learning_rate": 0.0002,
+      "loss": 0.5133,
+      "step": 33850
+    },
+    {
+      "epoch": 5.474092636003557,
+      "grad_norm": 1.0415139198303223,
+      "learning_rate": 0.0002,
+      "loss": 0.4926,
+      "step": 33860
+    },
+    {
+      "epoch": 5.475709320184302,
+      "grad_norm": 0.9213164448738098,
+      "learning_rate": 0.0002,
+      "loss": 0.5323,
+      "step": 33870
+    },
+    {
+      "epoch": 5.477326004365048,
+      "grad_norm": 1.0364776849746704,
+      "learning_rate": 0.0002,
+      "loss": 0.5125,
+      "step": 33880
+    },
+    {
+      "epoch": 5.478942688545793,
+      "grad_norm": 0.9994072318077087,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 33890
+    },
+    {
+      "epoch": 5.480559372726538,
+      "grad_norm": 1.196730136871338,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 33900
+    },
+    {
+      "epoch": 5.482176056907283,
+      "grad_norm": 0.9955780506134033,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 33910
+    },
+    {
+      "epoch": 5.4837927410880285,
+      "grad_norm": 1.168188214302063,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 33920
+    },
+    {
+      "epoch": 5.485409425268774,
+      "grad_norm": 1.1816450357437134,
+      "learning_rate": 0.0002,
+      "loss": 0.5548,
+      "step": 33930
+    },
+    {
+      "epoch": 5.487026109449519,
+      "grad_norm": 1.079715609550476,
+      "learning_rate": 0.0002,
+      "loss": 0.5535,
+      "step": 33940
+    },
+    {
+      "epoch": 5.488642793630264,
+      "grad_norm": 1.153850793838501,
+      "learning_rate": 0.0002,
+      "loss": 0.5262,
+      "step": 33950
+    },
+    {
+      "epoch": 5.490259477811009,
+      "grad_norm": 1.0207297801971436,
+      "learning_rate": 0.0002,
+      "loss": 0.5248,
+      "step": 33960
+    },
+    {
+      "epoch": 5.491876161991755,
+      "grad_norm": 1.1290855407714844,
+      "learning_rate": 0.0002,
+      "loss": 0.5142,
+      "step": 33970
+    },
+    {
+      "epoch": 5.4934928461725,
+      "grad_norm": 1.068058967590332,
+      "learning_rate": 0.0002,
+      "loss": 0.5168,
+      "step": 33980
+    },
+    {
+      "epoch": 5.495109530353245,
+      "grad_norm": 0.9789979457855225,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 33990
+    },
+    {
+      "epoch": 5.496726214533991,
+      "grad_norm": 0.9696692824363708,
+      "learning_rate": 0.0002,
+      "loss": 0.5113,
+      "step": 34000
+    },
+    {
+      "epoch": 5.4983428987147365,
+      "grad_norm": 1.0539981126785278,
+      "learning_rate": 0.0002,
+      "loss": 0.5413,
+      "step": 34010
+    },
+    {
+      "epoch": 5.499959582895482,
+      "grad_norm": 1.0249929428100586,
+      "learning_rate": 0.0002,
+      "loss": 0.5783,
+      "step": 34020
+    },
+    {
+      "epoch": 5.501576267076227,
+      "grad_norm": 0.9577504992485046,
+      "learning_rate": 0.0002,
+      "loss": 0.4888,
+      "step": 34030
+    },
+    {
+      "epoch": 5.503192951256972,
+      "grad_norm": 1.0963513851165771,
+      "learning_rate": 0.0002,
+      "loss": 0.5291,
+      "step": 34040
+    },
+    {
+      "epoch": 5.504809635437717,
+      "grad_norm": 0.8339345455169678,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 34050
+    },
+    {
+      "epoch": 5.506426319618463,
+      "grad_norm": 1.0138782262802124,
+      "learning_rate": 0.0002,
+      "loss": 0.5191,
+      "step": 34060
+    },
+    {
+      "epoch": 5.508043003799208,
+      "grad_norm": 1.0180109739303589,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 34070
+    },
+    {
+      "epoch": 5.509659687979953,
+      "grad_norm": 1.2790818214416504,
+      "learning_rate": 0.0002,
+      "loss": 0.5083,
+      "step": 34080
+    },
+    {
+      "epoch": 5.511276372160698,
+      "grad_norm": 1.428247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 34090
+    },
+    {
+      "epoch": 5.5128930563414436,
+      "grad_norm": 1.0926059484481812,
+      "learning_rate": 0.0002,
+      "loss": 0.5291,
+      "step": 34100
+    },
+    {
+      "epoch": 5.514509740522189,
+      "grad_norm": 1.2353343963623047,
+      "learning_rate": 0.0002,
+      "loss": 0.5665,
+      "step": 34110
+    },
+    {
+      "epoch": 5.516126424702934,
+      "grad_norm": 0.935587465763092,
+      "learning_rate": 0.0002,
+      "loss": 0.5331,
+      "step": 34120
+    },
+    {
+      "epoch": 5.517743108883679,
+      "grad_norm": 0.9767586588859558,
+      "learning_rate": 0.0002,
+      "loss": 0.5512,
+      "step": 34130
+    },
+    {
+      "epoch": 5.5193597930644245,
+      "grad_norm": 1.1660610437393188,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 34140
+    },
+    {
+      "epoch": 5.520976477245171,
+      "grad_norm": 0.9828870892524719,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 34150
+    },
+    {
+      "epoch": 5.522593161425916,
+      "grad_norm": 1.0097278356552124,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 34160
+    },
+    {
+      "epoch": 5.524209845606661,
+      "grad_norm": 1.1766167879104614,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 34170
+    },
+    {
+      "epoch": 5.525826529787406,
+      "grad_norm": 0.982292115688324,
+      "learning_rate": 0.0002,
+      "loss": 0.5258,
+      "step": 34180
+    },
+    {
+      "epoch": 5.5274432139681515,
+      "grad_norm": 1.0744609832763672,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 34190
+    },
+    {
+      "epoch": 5.529059898148897,
+      "grad_norm": 1.3831160068511963,
+      "learning_rate": 0.0002,
+      "loss": 0.5469,
+      "step": 34200
+    },
+    {
+      "epoch": 5.530676582329642,
+      "grad_norm": 1.074771761894226,
+      "learning_rate": 0.0002,
+      "loss": 0.5819,
+      "step": 34210
+    },
+    {
+      "epoch": 5.532293266510387,
+      "grad_norm": 1.016652226448059,
+      "learning_rate": 0.0002,
+      "loss": 0.5399,
+      "step": 34220
+    },
+    {
+      "epoch": 5.5339099506911325,
+      "grad_norm": 1.2231552600860596,
+      "learning_rate": 0.0002,
+      "loss": 0.5158,
+      "step": 34230
+    },
+    {
+      "epoch": 5.535526634871878,
+      "grad_norm": 0.8051198720932007,
+      "learning_rate": 0.0002,
+      "loss": 0.5091,
+      "step": 34240
+    },
+    {
+      "epoch": 5.537143319052623,
+      "grad_norm": 1.1779674291610718,
+      "learning_rate": 0.0002,
+      "loss": 0.5583,
+      "step": 34250
+    },
+    {
+      "epoch": 5.538760003233368,
+      "grad_norm": 1.2468291521072388,
+      "learning_rate": 0.0002,
+      "loss": 0.5044,
+      "step": 34260
+    },
+    {
+      "epoch": 5.540376687414113,
+      "grad_norm": 1.14818274974823,
+      "learning_rate": 0.0002,
+      "loss": 0.523,
+      "step": 34270
+    },
+    {
+      "epoch": 5.541993371594859,
+      "grad_norm": 1.2362616062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.5375,
+      "step": 34280
+    },
+    {
+      "epoch": 5.543610055775604,
+      "grad_norm": 1.0206977128982544,
+      "learning_rate": 0.0002,
+      "loss": 0.4996,
+      "step": 34290
+    },
+    {
+      "epoch": 5.54522673995635,
+      "grad_norm": 1.2018457651138306,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 34300
+    },
+    {
+      "epoch": 5.546843424137095,
+      "grad_norm": 1.0349043607711792,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 34310
+    },
+    {
+      "epoch": 5.54846010831784,
+      "grad_norm": 1.2022006511688232,
+      "learning_rate": 0.0002,
+      "loss": 0.5231,
+      "step": 34320
+    },
+    {
+      "epoch": 5.550076792498586,
+      "grad_norm": 1.0810624361038208,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 34330
+    },
+    {
+      "epoch": 5.551693476679331,
+      "grad_norm": 1.3297529220581055,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 34340
+    },
+    {
+      "epoch": 5.553310160860076,
+      "grad_norm": 0.9722549915313721,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 34350
+    },
+    {
+      "epoch": 5.554926845040821,
+      "grad_norm": 0.9903425574302673,
+      "learning_rate": 0.0002,
+      "loss": 0.4823,
+      "step": 34360
+    },
+    {
+      "epoch": 5.556543529221567,
+      "grad_norm": 0.9568067789077759,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 34370
+    },
+    {
+      "epoch": 5.558160213402312,
+      "grad_norm": 1.113870620727539,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 34380
+    },
+    {
+      "epoch": 5.559776897583057,
+      "grad_norm": 1.0557632446289062,
+      "learning_rate": 0.0002,
+      "loss": 0.5278,
+      "step": 34390
+    },
+    {
+      "epoch": 5.561393581763802,
+      "grad_norm": 0.9615673422813416,
+      "learning_rate": 0.0002,
+      "loss": 0.5501,
+      "step": 34400
+    },
+    {
+      "epoch": 5.5630102659445475,
+      "grad_norm": 0.9536027312278748,
+      "learning_rate": 0.0002,
+      "loss": 0.5066,
+      "step": 34410
+    },
+    {
+      "epoch": 5.564626950125293,
+      "grad_norm": 0.8808749318122864,
+      "learning_rate": 0.0002,
+      "loss": 0.4949,
+      "step": 34420
+    },
+    {
+      "epoch": 5.566243634306038,
+      "grad_norm": 1.286132574081421,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 34430
+    },
+    {
+      "epoch": 5.567860318486783,
+      "grad_norm": 1.259644865989685,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 34440
+    },
+    {
+      "epoch": 5.569477002667529,
+      "grad_norm": 0.9920216798782349,
+      "learning_rate": 0.0002,
+      "loss": 0.4922,
+      "step": 34450
+    },
+    {
+      "epoch": 5.5710936868482746,
+      "grad_norm": 1.182926893234253,
+      "learning_rate": 0.0002,
+      "loss": 0.5527,
+      "step": 34460
+    },
+    {
+      "epoch": 5.57271037102902,
+      "grad_norm": 1.1434749364852905,
+      "learning_rate": 0.0002,
+      "loss": 0.5185,
+      "step": 34470
+    },
+    {
+      "epoch": 5.574327055209765,
+      "grad_norm": 1.2420979738235474,
+      "learning_rate": 0.0002,
+      "loss": 0.5256,
+      "step": 34480
+    },
+    {
+      "epoch": 5.57594373939051,
+      "grad_norm": 0.9338384866714478,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 34490
+    },
+    {
+      "epoch": 5.5775604235712555,
+      "grad_norm": 1.0196425914764404,
+      "learning_rate": 0.0002,
+      "loss": 0.5634,
+      "step": 34500
+    },
+    {
+      "epoch": 5.579177107752001,
+      "grad_norm": 0.9586997032165527,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 34510
+    },
+    {
+      "epoch": 5.580793791932746,
+      "grad_norm": 1.2409086227416992,
+      "learning_rate": 0.0002,
+      "loss": 0.5336,
+      "step": 34520
+    },
+    {
+      "epoch": 5.582410476113491,
+      "grad_norm": 1.1483757495880127,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 34530
+    },
+    {
+      "epoch": 5.584027160294236,
+      "grad_norm": 1.1624305248260498,
+      "learning_rate": 0.0002,
+      "loss": 0.5325,
+      "step": 34540
+    },
+    {
+      "epoch": 5.585643844474982,
+      "grad_norm": 1.2635223865509033,
+      "learning_rate": 0.0002,
+      "loss": 0.5342,
+      "step": 34550
+    },
+    {
+      "epoch": 5.587260528655727,
+      "grad_norm": 0.9824051856994629,
+      "learning_rate": 0.0002,
+      "loss": 0.4924,
+      "step": 34560
+    },
+    {
+      "epoch": 5.588877212836472,
+      "grad_norm": 1.0858620405197144,
+      "learning_rate": 0.0002,
+      "loss": 0.5395,
+      "step": 34570
+    },
+    {
+      "epoch": 5.590493897017217,
+      "grad_norm": 1.1452655792236328,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 34580
+    },
+    {
+      "epoch": 5.592110581197963,
+      "grad_norm": 1.110610842704773,
+      "learning_rate": 0.0002,
+      "loss": 0.5746,
+      "step": 34590
+    },
+    {
+      "epoch": 5.593727265378709,
+      "grad_norm": 0.9976194500923157,
+      "learning_rate": 0.0002,
+      "loss": 0.5285,
+      "step": 34600
+    },
+    {
+      "epoch": 5.595343949559454,
+      "grad_norm": 1.0698920488357544,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 34610
+    },
+    {
+      "epoch": 5.596960633740199,
+      "grad_norm": 1.1505171060562134,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 34620
+    },
+    {
+      "epoch": 5.598577317920944,
+      "grad_norm": 1.1014643907546997,
+      "learning_rate": 0.0002,
+      "loss": 0.5471,
+      "step": 34630
+    },
+    {
+      "epoch": 5.60019400210169,
+      "grad_norm": 0.915595293045044,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 34640
+    },
+    {
+      "epoch": 5.601810686282435,
+      "grad_norm": 1.1856765747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 34650
+    },
+    {
+      "epoch": 5.60342737046318,
+      "grad_norm": 1.1357687711715698,
+      "learning_rate": 0.0002,
+      "loss": 0.5502,
+      "step": 34660
+    },
+    {
+      "epoch": 5.605044054643925,
+      "grad_norm": 1.0232492685317993,
+      "learning_rate": 0.0002,
+      "loss": 0.5034,
+      "step": 34670
+    },
+    {
+      "epoch": 5.6066607388246705,
+      "grad_norm": 0.9375017881393433,
+      "learning_rate": 0.0002,
+      "loss": 0.5357,
+      "step": 34680
+    },
+    {
+      "epoch": 5.608277423005416,
+      "grad_norm": 1.0796529054641724,
+      "learning_rate": 0.0002,
+      "loss": 0.5518,
+      "step": 34690
+    },
+    {
+      "epoch": 5.609894107186161,
+      "grad_norm": 1.1383336782455444,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 34700
+    },
+    {
+      "epoch": 5.611510791366906,
+      "grad_norm": 1.0248544216156006,
+      "learning_rate": 0.0002,
+      "loss": 0.5477,
+      "step": 34710
+    },
+    {
+      "epoch": 5.6131274755476515,
+      "grad_norm": 1.0986040830612183,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 34720
+    },
+    {
+      "epoch": 5.614744159728397,
+      "grad_norm": 1.2689568996429443,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 34730
+    },
+    {
+      "epoch": 5.616360843909142,
+      "grad_norm": 1.4044264554977417,
+      "learning_rate": 0.0002,
+      "loss": 0.5136,
+      "step": 34740
+    },
+    {
+      "epoch": 5.617977528089888,
+      "grad_norm": 1.2084474563598633,
+      "learning_rate": 0.0002,
+      "loss": 0.5699,
+      "step": 34750
+    },
+    {
+      "epoch": 5.619594212270633,
+      "grad_norm": 1.061248540878296,
+      "learning_rate": 0.0002,
+      "loss": 0.5377,
+      "step": 34760
+    },
+    {
+      "epoch": 5.6212108964513785,
+      "grad_norm": 1.0220764875411987,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 34770
+    },
+    {
+      "epoch": 5.622827580632124,
+      "grad_norm": 1.0859092473983765,
+      "learning_rate": 0.0002,
+      "loss": 0.54,
+      "step": 34780
+    },
+    {
+      "epoch": 5.624444264812869,
+      "grad_norm": 0.9049732089042664,
+      "learning_rate": 0.0002,
+      "loss": 0.5308,
+      "step": 34790
+    },
+    {
+      "epoch": 5.626060948993614,
+      "grad_norm": 1.2103937864303589,
+      "learning_rate": 0.0002,
+      "loss": 0.5433,
+      "step": 34800
+    },
+    {
+      "epoch": 5.627677633174359,
+      "grad_norm": 0.9854230284690857,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 34810
+    },
+    {
+      "epoch": 5.629294317355105,
+      "grad_norm": 0.9316635131835938,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 34820
+    },
+    {
+      "epoch": 5.63091100153585,
+      "grad_norm": 1.105296015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.5393,
+      "step": 34830
+    },
+    {
+      "epoch": 5.632527685716595,
+      "grad_norm": 0.993383526802063,
+      "learning_rate": 0.0002,
+      "loss": 0.5527,
+      "step": 34840
+    },
+    {
+      "epoch": 5.63414436989734,
+      "grad_norm": 1.1544116735458374,
+      "learning_rate": 0.0002,
+      "loss": 0.5375,
+      "step": 34850
+    },
+    {
+      "epoch": 5.635761054078086,
+      "grad_norm": 1.284475326538086,
+      "learning_rate": 0.0002,
+      "loss": 0.5448,
+      "step": 34860
+    },
+    {
+      "epoch": 5.637377738258831,
+      "grad_norm": 1.121997594833374,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 34870
+    },
+    {
+      "epoch": 5.638994422439576,
+      "grad_norm": 1.213040828704834,
+      "learning_rate": 0.0002,
+      "loss": 0.5335,
+      "step": 34880
+    },
+    {
+      "epoch": 5.640611106620321,
+      "grad_norm": 1.23222017288208,
+      "learning_rate": 0.0002,
+      "loss": 0.5623,
+      "step": 34890
+    },
+    {
+      "epoch": 5.642227790801067,
+      "grad_norm": 0.9793637990951538,
+      "learning_rate": 0.0002,
+      "loss": 0.5622,
+      "step": 34900
+    },
+    {
+      "epoch": 5.643844474981813,
+      "grad_norm": 1.38919997215271,
+      "learning_rate": 0.0002,
+      "loss": 0.5405,
+      "step": 34910
+    },
+    {
+      "epoch": 5.645461159162558,
+      "grad_norm": 0.8390951156616211,
+      "learning_rate": 0.0002,
+      "loss": 0.5007,
+      "step": 34920
+    },
+    {
+      "epoch": 5.647077843343303,
+      "grad_norm": 0.9465909004211426,
+      "learning_rate": 0.0002,
+      "loss": 0.5974,
+      "step": 34930
+    },
+    {
+      "epoch": 5.648694527524048,
+      "grad_norm": 1.066957712173462,
+      "learning_rate": 0.0002,
+      "loss": 0.5264,
+      "step": 34940
+    },
+    {
+      "epoch": 5.650311211704794,
+      "grad_norm": 0.9842154383659363,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 34950
+    },
+    {
+      "epoch": 5.651927895885539,
+      "grad_norm": 1.1766440868377686,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 34960
+    },
+    {
+      "epoch": 5.653544580066284,
+      "grad_norm": 0.9061306118965149,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 34970
+    },
+    {
+      "epoch": 5.655161264247029,
+      "grad_norm": 1.2941309213638306,
+      "learning_rate": 0.0002,
+      "loss": 0.5446,
+      "step": 34980
+    },
+    {
+      "epoch": 5.6567779484277745,
+      "grad_norm": 0.9741247892379761,
+      "learning_rate": 0.0002,
+      "loss": 0.5704,
+      "step": 34990
+    },
+    {
+      "epoch": 5.65839463260852,
+      "grad_norm": 1.0784187316894531,
+      "learning_rate": 0.0002,
+      "loss": 0.5152,
+      "step": 35000
+    },
+    {
+      "epoch": 5.660011316789265,
+      "grad_norm": 0.937889814376831,
+      "learning_rate": 0.0002,
+      "loss": 0.5363,
+      "step": 35010
+    },
+    {
+      "epoch": 5.66162800097001,
+      "grad_norm": 0.9667879939079285,
+      "learning_rate": 0.0002,
+      "loss": 0.5019,
+      "step": 35020
+    },
+    {
+      "epoch": 5.663244685150756,
+      "grad_norm": 1.0554876327514648,
+      "learning_rate": 0.0002,
+      "loss": 0.5209,
+      "step": 35030
+    },
+    {
+      "epoch": 5.664861369331501,
+      "grad_norm": 1.2030539512634277,
+      "learning_rate": 0.0002,
+      "loss": 0.523,
+      "step": 35040
+    },
+    {
+      "epoch": 5.666478053512247,
+      "grad_norm": 1.0849953889846802,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 35050
+    },
+    {
+      "epoch": 5.668094737692992,
+      "grad_norm": 1.1598973274230957,
+      "learning_rate": 0.0002,
+      "loss": 0.5747,
+      "step": 35060
+    },
+    {
+      "epoch": 5.669711421873737,
+      "grad_norm": 1.0233359336853027,
+      "learning_rate": 0.0002,
+      "loss": 0.5488,
+      "step": 35070
+    },
+    {
+      "epoch": 5.6713281060544825,
+      "grad_norm": 1.1124799251556396,
+      "learning_rate": 0.0002,
+      "loss": 0.5409,
+      "step": 35080
+    },
+    {
+      "epoch": 5.672944790235228,
+      "grad_norm": 1.2351475954055786,
+      "learning_rate": 0.0002,
+      "loss": 0.5578,
+      "step": 35090
+    },
+    {
+      "epoch": 5.674561474415973,
+      "grad_norm": 1.0240728855133057,
+      "learning_rate": 0.0002,
+      "loss": 0.5638,
+      "step": 35100
+    },
+    {
+      "epoch": 5.676178158596718,
+      "grad_norm": 1.0223692655563354,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 35110
+    },
+    {
+      "epoch": 5.677794842777463,
+      "grad_norm": 1.4569132328033447,
+      "learning_rate": 0.0002,
+      "loss": 0.524,
+      "step": 35120
+    },
+    {
+      "epoch": 5.679411526958209,
+      "grad_norm": 0.8983587026596069,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 35130
+    },
+    {
+      "epoch": 5.681028211138954,
+      "grad_norm": 1.0775383710861206,
+      "learning_rate": 0.0002,
+      "loss": 0.5439,
+      "step": 35140
+    },
+    {
+      "epoch": 5.682644895319699,
+      "grad_norm": 0.9800270795822144,
+      "learning_rate": 0.0002,
+      "loss": 0.5289,
+      "step": 35150
+    },
+    {
+      "epoch": 5.684261579500444,
+      "grad_norm": 0.9858237504959106,
+      "learning_rate": 0.0002,
+      "loss": 0.533,
+      "step": 35160
+    },
+    {
+      "epoch": 5.6858782636811895,
+      "grad_norm": 1.031087040901184,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 35170
+    },
+    {
+      "epoch": 5.687494947861936,
+      "grad_norm": 1.0294365882873535,
+      "learning_rate": 0.0002,
+      "loss": 0.5528,
+      "step": 35180
+    },
+    {
+      "epoch": 5.68911163204268,
+      "grad_norm": 1.108144760131836,
+      "learning_rate": 0.0002,
+      "loss": 0.5581,
+      "step": 35190
+    },
+    {
+      "epoch": 5.690728316223426,
+      "grad_norm": 1.0813100337982178,
+      "learning_rate": 0.0002,
+      "loss": 0.5373,
+      "step": 35200
+    },
+    {
+      "epoch": 5.692345000404171,
+      "grad_norm": 1.3146867752075195,
+      "learning_rate": 0.0002,
+      "loss": 0.5429,
+      "step": 35210
+    },
+    {
+      "epoch": 5.693961684584917,
+      "grad_norm": 1.16780424118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5297,
+      "step": 35220
+    },
+    {
+      "epoch": 5.695578368765662,
+      "grad_norm": 0.9929125905036926,
+      "learning_rate": 0.0002,
+      "loss": 0.577,
+      "step": 35230
+    },
+    {
+      "epoch": 5.697195052946407,
+      "grad_norm": 0.9049441814422607,
+      "learning_rate": 0.0002,
+      "loss": 0.5441,
+      "step": 35240
+    },
+    {
+      "epoch": 5.698811737127152,
+      "grad_norm": 0.9768866300582886,
+      "learning_rate": 0.0002,
+      "loss": 0.5349,
+      "step": 35250
+    },
+    {
+      "epoch": 5.7004284213078975,
+      "grad_norm": 0.8306029438972473,
+      "learning_rate": 0.0002,
+      "loss": 0.542,
+      "step": 35260
+    },
+    {
+      "epoch": 5.702045105488643,
+      "grad_norm": 0.8417280316352844,
+      "learning_rate": 0.0002,
+      "loss": 0.4771,
+      "step": 35270
+    },
+    {
+      "epoch": 5.703661789669388,
+      "grad_norm": 0.9954485893249512,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 35280
+    },
+    {
+      "epoch": 5.705278473850133,
+      "grad_norm": 1.2417993545532227,
+      "learning_rate": 0.0002,
+      "loss": 0.5469,
+      "step": 35290
+    },
+    {
+      "epoch": 5.706895158030878,
+      "grad_norm": 1.1696544885635376,
+      "learning_rate": 0.0002,
+      "loss": 0.5275,
+      "step": 35300
+    },
+    {
+      "epoch": 5.708511842211624,
+      "grad_norm": 1.2424817085266113,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 35310
+    },
+    {
+      "epoch": 5.710128526392369,
+      "grad_norm": 1.1791106462478638,
+      "learning_rate": 0.0002,
+      "loss": 0.5595,
+      "step": 35320
+    },
+    {
+      "epoch": 5.711745210573115,
+      "grad_norm": 1.202181339263916,
+      "learning_rate": 0.0002,
+      "loss": 0.5076,
+      "step": 35330
+    },
+    {
+      "epoch": 5.713361894753859,
+      "grad_norm": 1.1006861925125122,
+      "learning_rate": 0.0002,
+      "loss": 0.5847,
+      "step": 35340
+    },
+    {
+      "epoch": 5.7149785789346055,
+      "grad_norm": 1.0918344259262085,
+      "learning_rate": 0.0002,
+      "loss": 0.5627,
+      "step": 35350
+    },
+    {
+      "epoch": 5.716595263115351,
+      "grad_norm": 1.0427305698394775,
+      "learning_rate": 0.0002,
+      "loss": 0.5677,
+      "step": 35360
+    },
+    {
+      "epoch": 5.718211947296096,
+      "grad_norm": 1.0818872451782227,
+      "learning_rate": 0.0002,
+      "loss": 0.5288,
+      "step": 35370
+    },
+    {
+      "epoch": 5.719828631476841,
+      "grad_norm": 1.186006784439087,
+      "learning_rate": 0.0002,
+      "loss": 0.5296,
+      "step": 35380
+    },
+    {
+      "epoch": 5.721445315657586,
+      "grad_norm": 1.2073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 35390
+    },
+    {
+      "epoch": 5.723061999838332,
+      "grad_norm": 1.065338134765625,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 35400
+    },
+    {
+      "epoch": 5.724678684019077,
+      "grad_norm": 0.9448973536491394,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 35410
+    },
+    {
+      "epoch": 5.726295368199822,
+      "grad_norm": 1.1487499475479126,
+      "learning_rate": 0.0002,
+      "loss": 0.5276,
+      "step": 35420
+    },
+    {
+      "epoch": 5.727912052380567,
+      "grad_norm": 1.1334216594696045,
+      "learning_rate": 0.0002,
+      "loss": 0.5435,
+      "step": 35430
+    },
+    {
+      "epoch": 5.729528736561313,
+      "grad_norm": 1.1932826042175293,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 35440
+    },
+    {
+      "epoch": 5.731145420742058,
+      "grad_norm": 1.2615786790847778,
+      "learning_rate": 0.0002,
+      "loss": 0.5502,
+      "step": 35450
+    },
+    {
+      "epoch": 5.732762104922803,
+      "grad_norm": 1.2803694009780884,
+      "learning_rate": 0.0002,
+      "loss": 0.5612,
+      "step": 35460
+    },
+    {
+      "epoch": 5.734378789103548,
+      "grad_norm": 0.9271906614303589,
+      "learning_rate": 0.0002,
+      "loss": 0.5458,
+      "step": 35470
+    },
+    {
+      "epoch": 5.735995473284294,
+      "grad_norm": 1.0958917140960693,
+      "learning_rate": 0.0002,
+      "loss": 0.5342,
+      "step": 35480
+    },
+    {
+      "epoch": 5.737612157465039,
+      "grad_norm": 1.1072784662246704,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 35490
+    },
+    {
+      "epoch": 5.739228841645785,
+      "grad_norm": 1.1641002893447876,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 35500
+    },
+    {
+      "epoch": 5.74084552582653,
+      "grad_norm": 1.0246447324752808,
+      "learning_rate": 0.0002,
+      "loss": 0.5252,
+      "step": 35510
+    },
+    {
+      "epoch": 5.742462210007275,
+      "grad_norm": 1.032474398612976,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 35520
+    },
+    {
+      "epoch": 5.7440788941880205,
+      "grad_norm": 1.1600854396820068,
+      "learning_rate": 0.0002,
+      "loss": 0.4965,
+      "step": 35530
+    },
+    {
+      "epoch": 5.745695578368766,
+      "grad_norm": 1.0686054229736328,
+      "learning_rate": 0.0002,
+      "loss": 0.5543,
+      "step": 35540
+    },
+    {
+      "epoch": 5.747312262549511,
+      "grad_norm": 1.2314637899398804,
+      "learning_rate": 0.0002,
+      "loss": 0.5706,
+      "step": 35550
+    },
+    {
+      "epoch": 5.748928946730256,
+      "grad_norm": 0.922134280204773,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 35560
+    },
+    {
+      "epoch": 5.7505456309110015,
+      "grad_norm": 0.933043360710144,
+      "learning_rate": 0.0002,
+      "loss": 0.5495,
+      "step": 35570
+    },
+    {
+      "epoch": 5.752162315091747,
+      "grad_norm": 1.1911931037902832,
+      "learning_rate": 0.0002,
+      "loss": 0.5007,
+      "step": 35580
+    },
+    {
+      "epoch": 5.753778999272492,
+      "grad_norm": 0.8984857797622681,
+      "learning_rate": 0.0002,
+      "loss": 0.5244,
+      "step": 35590
+    },
+    {
+      "epoch": 5.755395683453237,
+      "grad_norm": 0.9495107531547546,
+      "learning_rate": 0.0002,
+      "loss": 0.5493,
+      "step": 35600
+    },
+    {
+      "epoch": 5.757012367633982,
+      "grad_norm": 1.2805472612380981,
+      "learning_rate": 0.0002,
+      "loss": 0.5326,
+      "step": 35610
+    },
+    {
+      "epoch": 5.758629051814728,
+      "grad_norm": 1.1236625909805298,
+      "learning_rate": 0.0002,
+      "loss": 0.5276,
+      "step": 35620
+    },
+    {
+      "epoch": 5.760245735995474,
+      "grad_norm": 1.0552798509597778,
+      "learning_rate": 0.0002,
+      "loss": 0.6102,
+      "step": 35630
+    },
+    {
+      "epoch": 5.761862420176218,
+      "grad_norm": 1.119909644126892,
+      "learning_rate": 0.0002,
+      "loss": 0.5479,
+      "step": 35640
+    },
+    {
+      "epoch": 5.763479104356964,
+      "grad_norm": 0.8786116242408752,
+      "learning_rate": 0.0002,
+      "loss": 0.5282,
+      "step": 35650
+    },
+    {
+      "epoch": 5.765095788537709,
+      "grad_norm": 1.2417117357254028,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 35660
+    },
+    {
+      "epoch": 5.766712472718455,
+      "grad_norm": 1.255200982093811,
+      "learning_rate": 0.0002,
+      "loss": 0.537,
+      "step": 35670
+    },
+    {
+      "epoch": 5.7683291568992,
+      "grad_norm": 1.0611358880996704,
+      "learning_rate": 0.0002,
+      "loss": 0.5308,
+      "step": 35680
+    },
+    {
+      "epoch": 5.769945841079945,
+      "grad_norm": 1.1443911790847778,
+      "learning_rate": 0.0002,
+      "loss": 0.5614,
+      "step": 35690
+    },
+    {
+      "epoch": 5.77156252526069,
+      "grad_norm": 1.1437989473342896,
+      "learning_rate": 0.0002,
+      "loss": 0.5386,
+      "step": 35700
+    },
+    {
+      "epoch": 5.773179209441436,
+      "grad_norm": 1.1375046968460083,
+      "learning_rate": 0.0002,
+      "loss": 0.537,
+      "step": 35710
+    },
+    {
+      "epoch": 5.774795893622181,
+      "grad_norm": 1.0777729749679565,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 35720
+    },
+    {
+      "epoch": 5.776412577802926,
+      "grad_norm": 1.1160215139389038,
+      "learning_rate": 0.0002,
+      "loss": 0.5521,
+      "step": 35730
+    },
+    {
+      "epoch": 5.778029261983671,
+      "grad_norm": 1.1268514394760132,
+      "learning_rate": 0.0002,
+      "loss": 0.5569,
+      "step": 35740
+    },
+    {
+      "epoch": 5.7796459461644165,
+      "grad_norm": 1.2752262353897095,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 35750
+    },
+    {
+      "epoch": 5.781262630345162,
+      "grad_norm": 1.0416184663772583,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 35760
+    },
+    {
+      "epoch": 5.782879314525907,
+      "grad_norm": 1.0622444152832031,
+      "learning_rate": 0.0002,
+      "loss": 0.5438,
+      "step": 35770
+    },
+    {
+      "epoch": 5.784495998706653,
+      "grad_norm": 1.1217877864837646,
+      "learning_rate": 0.0002,
+      "loss": 0.5268,
+      "step": 35780
+    },
+    {
+      "epoch": 5.786112682887398,
+      "grad_norm": 0.9363139867782593,
+      "learning_rate": 0.0002,
+      "loss": 0.5225,
+      "step": 35790
+    },
+    {
+      "epoch": 5.787729367068144,
+      "grad_norm": 0.96628737449646,
+      "learning_rate": 0.0002,
+      "loss": 0.5524,
+      "step": 35800
+    },
+    {
+      "epoch": 5.789346051248889,
+      "grad_norm": 0.9572572112083435,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 35810
+    },
+    {
+      "epoch": 5.790962735429634,
+      "grad_norm": 0.938724935054779,
+      "learning_rate": 0.0002,
+      "loss": 0.5615,
+      "step": 35820
+    },
+    {
+      "epoch": 5.792579419610379,
+      "grad_norm": 1.3314417600631714,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 35830
+    },
+    {
+      "epoch": 5.7941961037911245,
+      "grad_norm": 1.0097602605819702,
+      "learning_rate": 0.0002,
+      "loss": 0.5441,
+      "step": 35840
+    },
+    {
+      "epoch": 5.79581278797187,
+      "grad_norm": 1.1265122890472412,
+      "learning_rate": 0.0002,
+      "loss": 0.591,
+      "step": 35850
+    },
+    {
+      "epoch": 5.797429472152615,
+      "grad_norm": 1.2191909551620483,
+      "learning_rate": 0.0002,
+      "loss": 0.5333,
+      "step": 35860
+    },
+    {
+      "epoch": 5.79904615633336,
+      "grad_norm": 0.9690808057785034,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 35870
+    },
+    {
+      "epoch": 5.800662840514105,
+      "grad_norm": 1.0871665477752686,
+      "learning_rate": 0.0002,
+      "loss": 0.5425,
+      "step": 35880
+    },
+    {
+      "epoch": 5.802279524694851,
+      "grad_norm": 1.1093597412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.5602,
+      "step": 35890
+    },
+    {
+      "epoch": 5.803896208875596,
+      "grad_norm": 1.2434282302856445,
+      "learning_rate": 0.0002,
+      "loss": 0.5475,
+      "step": 35900
+    },
+    {
+      "epoch": 5.805512893056341,
+      "grad_norm": 1.2933623790740967,
+      "learning_rate": 0.0002,
+      "loss": 0.5288,
+      "step": 35910
+    },
+    {
+      "epoch": 5.807129577237086,
+      "grad_norm": 1.0005441904067993,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 35920
+    },
+    {
+      "epoch": 5.8087462614178325,
+      "grad_norm": 1.2373108863830566,
+      "learning_rate": 0.0002,
+      "loss": 0.5318,
+      "step": 35930
+    },
+    {
+      "epoch": 5.810362945598578,
+      "grad_norm": 1.2622692584991455,
+      "learning_rate": 0.0002,
+      "loss": 0.5413,
+      "step": 35940
+    },
+    {
+      "epoch": 5.811979629779323,
+      "grad_norm": 1.0112963914871216,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 35950
+    },
+    {
+      "epoch": 5.813596313960068,
+      "grad_norm": 1.050572395324707,
+      "learning_rate": 0.0002,
+      "loss": 0.5115,
+      "step": 35960
+    },
+    {
+      "epoch": 5.815212998140813,
+      "grad_norm": 0.9774560928344727,
+      "learning_rate": 0.0002,
+      "loss": 0.5288,
+      "step": 35970
+    },
+    {
+      "epoch": 5.816829682321559,
+      "grad_norm": 1.19438898563385,
+      "learning_rate": 0.0002,
+      "loss": 0.585,
+      "step": 35980
+    },
+    {
+      "epoch": 5.818446366502304,
+      "grad_norm": 1.0267130136489868,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 35990
+    },
+    {
+      "epoch": 5.820063050683049,
+      "grad_norm": 0.9813851714134216,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 36000
+    },
+    {
+      "epoch": 5.821679734863794,
+      "grad_norm": 0.9177457094192505,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 36010
+    },
+    {
+      "epoch": 5.8232964190445395,
+      "grad_norm": 1.0020731687545776,
+      "learning_rate": 0.0002,
+      "loss": 0.5453,
+      "step": 36020
+    },
+    {
+      "epoch": 5.824913103225285,
+      "grad_norm": 1.073222041130066,
+      "learning_rate": 0.0002,
+      "loss": 0.5646,
+      "step": 36030
+    },
+    {
+      "epoch": 5.82652978740603,
+      "grad_norm": 1.016337513923645,
+      "learning_rate": 0.0002,
+      "loss": 0.5539,
+      "step": 36040
+    },
+    {
+      "epoch": 5.828146471586775,
+      "grad_norm": 1.267364263534546,
+      "learning_rate": 0.0002,
+      "loss": 0.5592,
+      "step": 36050
+    },
+    {
+      "epoch": 5.8297631557675205,
+      "grad_norm": 1.2730127573013306,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 36060
+    },
+    {
+      "epoch": 5.831379839948266,
+      "grad_norm": 1.108442783355713,
+      "learning_rate": 0.0002,
+      "loss": 0.5247,
+      "step": 36070
+    },
+    {
+      "epoch": 5.832996524129012,
+      "grad_norm": 1.198072075843811,
+      "learning_rate": 0.0002,
+      "loss": 0.5103,
+      "step": 36080
+    },
+    {
+      "epoch": 5.834613208309757,
+      "grad_norm": 1.0458786487579346,
+      "learning_rate": 0.0002,
+      "loss": 0.5479,
+      "step": 36090
+    },
+    {
+      "epoch": 5.836229892490502,
+      "grad_norm": 0.9096664786338806,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 36100
+    },
+    {
+      "epoch": 5.8378465766712475,
+      "grad_norm": 0.9957793951034546,
+      "learning_rate": 0.0002,
+      "loss": 0.5602,
+      "step": 36110
+    },
+    {
+      "epoch": 5.839463260851993,
+      "grad_norm": 1.3693058490753174,
+      "learning_rate": 0.0002,
+      "loss": 0.5799,
+      "step": 36120
+    },
+    {
+      "epoch": 5.841079945032738,
+      "grad_norm": 1.268608808517456,
+      "learning_rate": 0.0002,
+      "loss": 0.5425,
+      "step": 36130
+    },
+    {
+      "epoch": 5.842696629213483,
+      "grad_norm": 0.8516020178794861,
+      "learning_rate": 0.0002,
+      "loss": 0.5653,
+      "step": 36140
+    },
+    {
+      "epoch": 5.844313313394228,
+      "grad_norm": 0.90385502576828,
+      "learning_rate": 0.0002,
+      "loss": 0.5475,
+      "step": 36150
+    },
+    {
+      "epoch": 5.845929997574974,
+      "grad_norm": 1.0910571813583374,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 36160
+    },
+    {
+      "epoch": 5.847546681755719,
+      "grad_norm": 0.9417795538902283,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 36170
+    },
+    {
+      "epoch": 5.849163365936464,
+      "grad_norm": 1.0027360916137695,
+      "learning_rate": 0.0002,
+      "loss": 0.5784,
+      "step": 36180
+    },
+    {
+      "epoch": 5.850780050117209,
+      "grad_norm": 1.1480516195297241,
+      "learning_rate": 0.0002,
+      "loss": 0.5423,
+      "step": 36190
+    },
+    {
+      "epoch": 5.852396734297955,
+      "grad_norm": 1.2431457042694092,
+      "learning_rate": 0.0002,
+      "loss": 0.5517,
+      "step": 36200
+    },
+    {
+      "epoch": 5.8540134184787,
+      "grad_norm": 1.091465950012207,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 36210
+    },
+    {
+      "epoch": 5.855630102659445,
+      "grad_norm": 0.9693930745124817,
+      "learning_rate": 0.0002,
+      "loss": 0.53,
+      "step": 36220
+    },
+    {
+      "epoch": 5.857246786840191,
+      "grad_norm": 0.9937465190887451,
+      "learning_rate": 0.0002,
+      "loss": 0.5453,
+      "step": 36230
+    },
+    {
+      "epoch": 5.858863471020936,
+      "grad_norm": 1.0731011629104614,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 36240
+    },
+    {
+      "epoch": 5.860480155201682,
+      "grad_norm": 1.0869048833847046,
+      "learning_rate": 0.0002,
+      "loss": 0.5687,
+      "step": 36250
+    },
+    {
+      "epoch": 5.862096839382427,
+      "grad_norm": 0.9226390719413757,
+      "learning_rate": 0.0002,
+      "loss": 0.5576,
+      "step": 36260
+    },
+    {
+      "epoch": 5.863713523563172,
+      "grad_norm": 1.1755430698394775,
+      "learning_rate": 0.0002,
+      "loss": 0.531,
+      "step": 36270
+    },
+    {
+      "epoch": 5.865330207743917,
+      "grad_norm": 0.8815974593162537,
+      "learning_rate": 0.0002,
+      "loss": 0.558,
+      "step": 36280
+    },
+    {
+      "epoch": 5.866946891924663,
+      "grad_norm": 1.3648751974105835,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 36290
+    },
+    {
+      "epoch": 5.868563576105408,
+      "grad_norm": 0.8729211091995239,
+      "learning_rate": 0.0002,
+      "loss": 0.536,
+      "step": 36300
+    },
+    {
+      "epoch": 5.870180260286153,
+      "grad_norm": 1.0870907306671143,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 36310
+    },
+    {
+      "epoch": 5.871796944466898,
+      "grad_norm": 1.1164259910583496,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 36320
+    },
+    {
+      "epoch": 5.8734136286476435,
+      "grad_norm": 1.1572535037994385,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 36330
+    },
+    {
+      "epoch": 5.875030312828389,
+      "grad_norm": 1.0456238985061646,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 36340
+    },
+    {
+      "epoch": 5.876646997009134,
+      "grad_norm": 1.1310722827911377,
+      "learning_rate": 0.0002,
+      "loss": 0.5008,
+      "step": 36350
+    },
+    {
+      "epoch": 5.878263681189879,
+      "grad_norm": 1.0004712343215942,
+      "learning_rate": 0.0002,
+      "loss": 0.5352,
+      "step": 36360
+    },
+    {
+      "epoch": 5.879880365370624,
+      "grad_norm": 1.0991777181625366,
+      "learning_rate": 0.0002,
+      "loss": 0.5632,
+      "step": 36370
+    },
+    {
+      "epoch": 5.8814970495513705,
+      "grad_norm": 1.2789239883422852,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 36380
+    },
+    {
+      "epoch": 5.883113733732116,
+      "grad_norm": 0.9524819850921631,
+      "learning_rate": 0.0002,
+      "loss": 0.56,
+      "step": 36390
+    },
+    {
+      "epoch": 5.884730417912861,
+      "grad_norm": 1.1115771532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 36400
+    },
+    {
+      "epoch": 5.886347102093606,
+      "grad_norm": 1.37419855594635,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 36410
+    },
+    {
+      "epoch": 5.8879637862743515,
+      "grad_norm": 1.1449527740478516,
+      "learning_rate": 0.0002,
+      "loss": 0.5675,
+      "step": 36420
+    },
+    {
+      "epoch": 5.889580470455097,
+      "grad_norm": 1.198046326637268,
+      "learning_rate": 0.0002,
+      "loss": 0.5255,
+      "step": 36430
+    },
+    {
+      "epoch": 5.891197154635842,
+      "grad_norm": 1.0180530548095703,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 36440
+    },
+    {
+      "epoch": 5.892813838816587,
+      "grad_norm": 1.0516417026519775,
+      "learning_rate": 0.0002,
+      "loss": 0.5319,
+      "step": 36450
+    },
+    {
+      "epoch": 5.894430522997332,
+      "grad_norm": 1.1658052206039429,
+      "learning_rate": 0.0002,
+      "loss": 0.5782,
+      "step": 36460
+    },
+    {
+      "epoch": 5.896047207178078,
+      "grad_norm": 1.190699577331543,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 36470
+    },
+    {
+      "epoch": 5.897663891358823,
+      "grad_norm": 1.1235495805740356,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 36480
+    },
+    {
+      "epoch": 5.899280575539568,
+      "grad_norm": 1.1926926374435425,
+      "learning_rate": 0.0002,
+      "loss": 0.5284,
+      "step": 36490
+    },
+    {
+      "epoch": 5.900897259720313,
+      "grad_norm": 1.1184662580490112,
+      "learning_rate": 0.0002,
+      "loss": 0.5686,
+      "step": 36500
+    },
+    {
+      "epoch": 5.9025139439010585,
+      "grad_norm": 1.000970721244812,
+      "learning_rate": 0.0002,
+      "loss": 0.5147,
+      "step": 36510
+    },
+    {
+      "epoch": 5.904130628081804,
+      "grad_norm": 1.0373306274414062,
+      "learning_rate": 0.0002,
+      "loss": 0.5351,
+      "step": 36520
+    },
+    {
+      "epoch": 5.90574731226255,
+      "grad_norm": 1.0840669870376587,
+      "learning_rate": 0.0002,
+      "loss": 0.535,
+      "step": 36530
+    },
+    {
+      "epoch": 5.907363996443295,
+      "grad_norm": 0.9908381104469299,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 36540
+    },
+    {
+      "epoch": 5.90898068062404,
+      "grad_norm": 1.0456029176712036,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 36550
+    },
+    {
+      "epoch": 5.910597364804786,
+      "grad_norm": 1.1381454467773438,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 36560
+    },
+    {
+      "epoch": 5.912214048985531,
+      "grad_norm": 0.9440900087356567,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 36570
+    },
+    {
+      "epoch": 5.913830733166276,
+      "grad_norm": 1.1674573421478271,
+      "learning_rate": 0.0002,
+      "loss": 0.5542,
+      "step": 36580
+    },
+    {
+      "epoch": 5.915447417347021,
+      "grad_norm": 1.1226966381072998,
+      "learning_rate": 0.0002,
+      "loss": 0.526,
+      "step": 36590
+    },
+    {
+      "epoch": 5.9170641015277665,
+      "grad_norm": 0.9696915745735168,
+      "learning_rate": 0.0002,
+      "loss": 0.6091,
+      "step": 36600
+    },
+    {
+      "epoch": 5.918680785708512,
+      "grad_norm": 0.9593005180358887,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 36610
+    },
+    {
+      "epoch": 5.920297469889257,
+      "grad_norm": 1.122169852256775,
+      "learning_rate": 0.0002,
+      "loss": 0.5536,
+      "step": 36620
+    },
+    {
+      "epoch": 5.921914154070002,
+      "grad_norm": 0.9923415780067444,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 36630
+    },
+    {
+      "epoch": 5.923530838250747,
+      "grad_norm": 1.063838005065918,
+      "learning_rate": 0.0002,
+      "loss": 0.5893,
+      "step": 36640
+    },
+    {
+      "epoch": 5.925147522431493,
+      "grad_norm": 0.9083505272865295,
+      "learning_rate": 0.0002,
+      "loss": 0.5799,
+      "step": 36650
+    },
+    {
+      "epoch": 5.926764206612239,
+      "grad_norm": 0.9439437985420227,
+      "learning_rate": 0.0002,
+      "loss": 0.5264,
+      "step": 36660
+    },
+    {
+      "epoch": 5.928380890792983,
+      "grad_norm": 0.9778534173965454,
+      "learning_rate": 0.0002,
+      "loss": 0.5891,
+      "step": 36670
+    },
+    {
+      "epoch": 5.929997574973729,
+      "grad_norm": 0.9723961353302002,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 36680
+    },
+    {
+      "epoch": 5.9316142591544745,
+      "grad_norm": 1.162333607673645,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 36690
+    },
+    {
+      "epoch": 5.93323094333522,
+      "grad_norm": 1.2784897089004517,
+      "learning_rate": 0.0002,
+      "loss": 0.5771,
+      "step": 36700
+    },
+    {
+      "epoch": 5.934847627515965,
+      "grad_norm": 1.0924867391586304,
+      "learning_rate": 0.0002,
+      "loss": 0.5343,
+      "step": 36710
+    },
+    {
+      "epoch": 5.93646431169671,
+      "grad_norm": 1.046922206878662,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 36720
+    },
+    {
+      "epoch": 5.938080995877455,
+      "grad_norm": 0.8632535338401794,
+      "learning_rate": 0.0002,
+      "loss": 0.5476,
+      "step": 36730
+    },
+    {
+      "epoch": 5.939697680058201,
+      "grad_norm": 1.358762502670288,
+      "learning_rate": 0.0002,
+      "loss": 0.5456,
+      "step": 36740
+    },
+    {
+      "epoch": 5.941314364238946,
+      "grad_norm": 1.2058624029159546,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 36750
+    },
+    {
+      "epoch": 5.942931048419691,
+      "grad_norm": 1.1396408081054688,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 36760
+    },
+    {
+      "epoch": 5.944547732600436,
+      "grad_norm": 1.1510354280471802,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 36770
+    },
+    {
+      "epoch": 5.946164416781182,
+      "grad_norm": 1.1401607990264893,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 36780
+    },
+    {
+      "epoch": 5.947781100961927,
+      "grad_norm": 1.1871325969696045,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 36790
+    },
+    {
+      "epoch": 5.949397785142672,
+      "grad_norm": 0.9928333163261414,
+      "learning_rate": 0.0002,
+      "loss": 0.4945,
+      "step": 36800
+    },
+    {
+      "epoch": 5.951014469323418,
+      "grad_norm": 1.0549445152282715,
+      "learning_rate": 0.0002,
+      "loss": 0.5303,
+      "step": 36810
+    },
+    {
+      "epoch": 5.9526311535041625,
+      "grad_norm": 0.9791563749313354,
+      "learning_rate": 0.0002,
+      "loss": 0.5532,
+      "step": 36820
+    },
+    {
+      "epoch": 5.954247837684909,
+      "grad_norm": 1.1268441677093506,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 36830
+    },
+    {
+      "epoch": 5.955864521865654,
+      "grad_norm": 1.0533992052078247,
+      "learning_rate": 0.0002,
+      "loss": 0.5585,
+      "step": 36840
+    },
+    {
+      "epoch": 5.957481206046399,
+      "grad_norm": 1.023358941078186,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 36850
+    },
+    {
+      "epoch": 5.959097890227144,
+      "grad_norm": 1.2631961107254028,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 36860
+    },
+    {
+      "epoch": 5.9607145744078895,
+      "grad_norm": 0.9397698640823364,
+      "learning_rate": 0.0002,
+      "loss": 0.5662,
+      "step": 36870
+    },
+    {
+      "epoch": 5.962331258588635,
+      "grad_norm": 1.1678427457809448,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 36880
+    },
+    {
+      "epoch": 5.96394794276938,
+      "grad_norm": 1.1403759717941284,
+      "learning_rate": 0.0002,
+      "loss": 0.5435,
+      "step": 36890
+    },
+    {
+      "epoch": 5.965564626950125,
+      "grad_norm": 1.030572772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.5479,
+      "step": 36900
+    },
+    {
+      "epoch": 5.9671813111308705,
+      "grad_norm": 1.0992497205734253,
+      "learning_rate": 0.0002,
+      "loss": 0.5838,
+      "step": 36910
+    },
+    {
+      "epoch": 5.968797995311616,
+      "grad_norm": 1.075466275215149,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 36920
+    },
+    {
+      "epoch": 5.970414679492361,
+      "grad_norm": 1.0153694152832031,
+      "learning_rate": 0.0002,
+      "loss": 0.5739,
+      "step": 36930
+    },
+    {
+      "epoch": 5.972031363673106,
+      "grad_norm": 0.973193883895874,
+      "learning_rate": 0.0002,
+      "loss": 0.5672,
+      "step": 36940
+    },
+    {
+      "epoch": 5.973648047853851,
+      "grad_norm": 0.8294678926467896,
+      "learning_rate": 0.0002,
+      "loss": 0.5585,
+      "step": 36950
+    },
+    {
+      "epoch": 5.9752647320345975,
+      "grad_norm": 1.0048716068267822,
+      "learning_rate": 0.0002,
+      "loss": 0.5631,
+      "step": 36960
+    },
+    {
+      "epoch": 5.976881416215342,
+      "grad_norm": 0.9714070558547974,
+      "learning_rate": 0.0002,
+      "loss": 0.5471,
+      "step": 36970
+    },
+    {
+      "epoch": 5.978498100396088,
+      "grad_norm": 0.8667682409286499,
+      "learning_rate": 0.0002,
+      "loss": 0.5419,
+      "step": 36980
+    },
+    {
+      "epoch": 5.980114784576833,
+      "grad_norm": 1.0461409091949463,
+      "learning_rate": 0.0002,
+      "loss": 0.5474,
+      "step": 36990
+    },
+    {
+      "epoch": 5.981731468757578,
+      "grad_norm": 0.9229754209518433,
+      "learning_rate": 0.0002,
+      "loss": 0.5454,
+      "step": 37000
+    },
+    {
+      "epoch": 5.983348152938324,
+      "grad_norm": 1.0406876802444458,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 37010
+    },
+    {
+      "epoch": 5.984964837119069,
+      "grad_norm": 0.8993828296661377,
+      "learning_rate": 0.0002,
+      "loss": 0.5569,
+      "step": 37020
+    },
+    {
+      "epoch": 5.986581521299814,
+      "grad_norm": 1.2260479927062988,
+      "learning_rate": 0.0002,
+      "loss": 0.5611,
+      "step": 37030
+    },
+    {
+      "epoch": 5.988198205480559,
+      "grad_norm": 1.0107380151748657,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 37040
+    },
+    {
+      "epoch": 5.989814889661305,
+      "grad_norm": 1.0240139961242676,
+      "learning_rate": 0.0002,
+      "loss": 0.5639,
+      "step": 37050
+    },
+    {
+      "epoch": 5.99143157384205,
+      "grad_norm": 1.0185275077819824,
+      "learning_rate": 0.0002,
+      "loss": 0.5209,
+      "step": 37060
+    },
+    {
+      "epoch": 5.993048258022795,
+      "grad_norm": 1.1361802816390991,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 37070
+    },
+    {
+      "epoch": 5.99466494220354,
+      "grad_norm": 1.0395532846450806,
+      "learning_rate": 0.0002,
+      "loss": 0.5692,
+      "step": 37080
+    },
+    {
+      "epoch": 5.9962816263842855,
+      "grad_norm": 0.9463558197021484,
+      "learning_rate": 0.0002,
+      "loss": 0.594,
+      "step": 37090
+    },
+    {
+      "epoch": 5.997898310565031,
+      "grad_norm": 1.2066948413848877,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 37100
+    },
+    {
+      "epoch": 5.999514994745777,
+      "grad_norm": 0.9749386310577393,
+      "learning_rate": 0.0002,
+      "loss": 0.5356,
+      "step": 37110
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 1.2270219326019287,
+      "eval_runtime": 122.2047,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 37113
+    },
+    {
+      "epoch": 6.001131678926522,
+      "grad_norm": 0.9641092419624329,
+      "learning_rate": 0.0002,
+      "loss": 0.4855,
+      "step": 37120
+    },
+    {
+      "epoch": 6.002748363107267,
+      "grad_norm": 1.103379249572754,
+      "learning_rate": 0.0002,
+      "loss": 0.4112,
+      "step": 37130
+    },
+    {
+      "epoch": 6.004365047288013,
+      "grad_norm": 0.8381665349006653,
+      "learning_rate": 0.0002,
+      "loss": 0.4577,
+      "step": 37140
+    },
+    {
+      "epoch": 6.005981731468758,
+      "grad_norm": 1.245323896408081,
+      "learning_rate": 0.0002,
+      "loss": 0.4794,
+      "step": 37150
+    },
+    {
+      "epoch": 6.007598415649503,
+      "grad_norm": 1.3140289783477783,
+      "learning_rate": 0.0002,
+      "loss": 0.4503,
+      "step": 37160
+    },
+    {
+      "epoch": 6.009215099830248,
+      "grad_norm": 0.8479695916175842,
+      "learning_rate": 0.0002,
+      "loss": 0.4456,
+      "step": 37170
+    },
+    {
+      "epoch": 6.0108317840109935,
+      "grad_norm": 0.8841437101364136,
+      "learning_rate": 0.0002,
+      "loss": 0.4573,
+      "step": 37180
+    },
+    {
+      "epoch": 6.012448468191739,
+      "grad_norm": 0.8900154829025269,
+      "learning_rate": 0.0002,
+      "loss": 0.4565,
+      "step": 37190
+    },
+    {
+      "epoch": 6.014065152372484,
+      "grad_norm": 1.2753345966339111,
+      "learning_rate": 0.0002,
+      "loss": 0.457,
+      "step": 37200
+    },
+    {
+      "epoch": 6.015681836553229,
+      "grad_norm": 1.4625498056411743,
+      "learning_rate": 0.0002,
+      "loss": 0.4365,
+      "step": 37210
+    },
+    {
+      "epoch": 6.017298520733974,
+      "grad_norm": 0.7455034852027893,
+      "learning_rate": 0.0002,
+      "loss": 0.4252,
+      "step": 37220
+    },
+    {
+      "epoch": 6.01891520491472,
+      "grad_norm": 1.1658862829208374,
+      "learning_rate": 0.0002,
+      "loss": 0.4433,
+      "step": 37230
+    },
+    {
+      "epoch": 6.020531889095465,
+      "grad_norm": 0.9785751104354858,
+      "learning_rate": 0.0002,
+      "loss": 0.4499,
+      "step": 37240
+    },
+    {
+      "epoch": 6.02214857327621,
+      "grad_norm": 1.3193122148513794,
+      "learning_rate": 0.0002,
+      "loss": 0.4956,
+      "step": 37250
+    },
+    {
+      "epoch": 6.023765257456955,
+      "grad_norm": 1.038273572921753,
+      "learning_rate": 0.0002,
+      "loss": 0.4727,
+      "step": 37260
+    },
+    {
+      "epoch": 6.0253819416377015,
+      "grad_norm": 1.0550594329833984,
+      "learning_rate": 0.0002,
+      "loss": 0.4395,
+      "step": 37270
+    },
+    {
+      "epoch": 6.026998625818447,
+      "grad_norm": 0.9745930433273315,
+      "learning_rate": 0.0002,
+      "loss": 0.4767,
+      "step": 37280
+    },
+    {
+      "epoch": 6.028615309999192,
+      "grad_norm": 0.9273530840873718,
+      "learning_rate": 0.0002,
+      "loss": 0.4233,
+      "step": 37290
+    },
+    {
+      "epoch": 6.030231994179937,
+      "grad_norm": 1.3844057321548462,
+      "learning_rate": 0.0002,
+      "loss": 0.4195,
+      "step": 37300
+    },
+    {
+      "epoch": 6.031848678360682,
+      "grad_norm": 1.2058762311935425,
+      "learning_rate": 0.0002,
+      "loss": 0.4768,
+      "step": 37310
+    },
+    {
+      "epoch": 6.033465362541428,
+      "grad_norm": 1.242663025856018,
+      "learning_rate": 0.0002,
+      "loss": 0.4499,
+      "step": 37320
+    },
+    {
+      "epoch": 6.035082046722173,
+      "grad_norm": 1.3504270315170288,
+      "learning_rate": 0.0002,
+      "loss": 0.4597,
+      "step": 37330
+    },
+    {
+      "epoch": 6.036698730902918,
+      "grad_norm": 0.8734912276268005,
+      "learning_rate": 0.0002,
+      "loss": 0.4402,
+      "step": 37340
+    },
+    {
+      "epoch": 6.038315415083663,
+      "grad_norm": 1.0182311534881592,
+      "learning_rate": 0.0002,
+      "loss": 0.477,
+      "step": 37350
+    },
+    {
+      "epoch": 6.0399320992644085,
+      "grad_norm": 0.9898499846458435,
+      "learning_rate": 0.0002,
+      "loss": 0.4261,
+      "step": 37360
+    },
+    {
+      "epoch": 6.041548783445154,
+      "grad_norm": 1.0637860298156738,
+      "learning_rate": 0.0002,
+      "loss": 0.4459,
+      "step": 37370
+    },
+    {
+      "epoch": 6.043165467625899,
+      "grad_norm": 1.0099523067474365,
+      "learning_rate": 0.0002,
+      "loss": 0.4958,
+      "step": 37380
+    },
+    {
+      "epoch": 6.044782151806644,
+      "grad_norm": 1.1080750226974487,
+      "learning_rate": 0.0002,
+      "loss": 0.4459,
+      "step": 37390
+    },
+    {
+      "epoch": 6.0463988359873895,
+      "grad_norm": 1.2551289796829224,
+      "learning_rate": 0.0002,
+      "loss": 0.4473,
+      "step": 37400
+    },
+    {
+      "epoch": 6.048015520168136,
+      "grad_norm": 0.8959632515907288,
+      "learning_rate": 0.0002,
+      "loss": 0.468,
+      "step": 37410
+    },
+    {
+      "epoch": 6.049632204348881,
+      "grad_norm": 1.1748892068862915,
+      "learning_rate": 0.0002,
+      "loss": 0.4255,
+      "step": 37420
+    },
+    {
+      "epoch": 6.051248888529626,
+      "grad_norm": 1.3122745752334595,
+      "learning_rate": 0.0002,
+      "loss": 0.4458,
+      "step": 37430
+    },
+    {
+      "epoch": 6.052865572710371,
+      "grad_norm": 1.0227985382080078,
+      "learning_rate": 0.0002,
+      "loss": 0.4676,
+      "step": 37440
+    },
+    {
+      "epoch": 6.0544822568911165,
+      "grad_norm": 1.0380030870437622,
+      "learning_rate": 0.0002,
+      "loss": 0.4503,
+      "step": 37450
+    },
+    {
+      "epoch": 6.056098941071862,
+      "grad_norm": 0.8919622898101807,
+      "learning_rate": 0.0002,
+      "loss": 0.4686,
+      "step": 37460
+    },
+    {
+      "epoch": 6.057715625252607,
+      "grad_norm": 1.4554150104522705,
+      "learning_rate": 0.0002,
+      "loss": 0.4406,
+      "step": 37470
+    },
+    {
+      "epoch": 6.059332309433352,
+      "grad_norm": 1.2853292226791382,
+      "learning_rate": 0.0002,
+      "loss": 0.4688,
+      "step": 37480
+    },
+    {
+      "epoch": 6.0609489936140974,
+      "grad_norm": 1.2951840162277222,
+      "learning_rate": 0.0002,
+      "loss": 0.4489,
+      "step": 37490
+    },
+    {
+      "epoch": 6.062565677794843,
+      "grad_norm": 1.1750973463058472,
+      "learning_rate": 0.0002,
+      "loss": 0.4819,
+      "step": 37500
+    },
+    {
+      "epoch": 6.064182361975588,
+      "grad_norm": 0.9328424334526062,
+      "learning_rate": 0.0002,
+      "loss": 0.4574,
+      "step": 37510
+    },
+    {
+      "epoch": 6.065799046156333,
+      "grad_norm": 1.0353537797927856,
+      "learning_rate": 0.0002,
+      "loss": 0.4597,
+      "step": 37520
+    },
+    {
+      "epoch": 6.067415730337078,
+      "grad_norm": 1.1594274044036865,
+      "learning_rate": 0.0002,
+      "loss": 0.4407,
+      "step": 37530
+    },
+    {
+      "epoch": 6.069032414517824,
+      "grad_norm": 0.9034168124198914,
+      "learning_rate": 0.0002,
+      "loss": 0.4642,
+      "step": 37540
+    },
+    {
+      "epoch": 6.070649098698569,
+      "grad_norm": 1.068617820739746,
+      "learning_rate": 0.0002,
+      "loss": 0.4625,
+      "step": 37550
+    },
+    {
+      "epoch": 6.072265782879315,
+      "grad_norm": 1.0931321382522583,
+      "learning_rate": 0.0002,
+      "loss": 0.4378,
+      "step": 37560
+    },
+    {
+      "epoch": 6.07388246706006,
+      "grad_norm": 1.2542688846588135,
+      "learning_rate": 0.0002,
+      "loss": 0.4527,
+      "step": 37570
+    },
+    {
+      "epoch": 6.075499151240805,
+      "grad_norm": 1.273384928703308,
+      "learning_rate": 0.0002,
+      "loss": 0.4725,
+      "step": 37580
+    },
+    {
+      "epoch": 6.077115835421551,
+      "grad_norm": 1.4771400690078735,
+      "learning_rate": 0.0002,
+      "loss": 0.4928,
+      "step": 37590
+    },
+    {
+      "epoch": 6.078732519602296,
+      "grad_norm": 1.3751444816589355,
+      "learning_rate": 0.0002,
+      "loss": 0.461,
+      "step": 37600
+    },
+    {
+      "epoch": 6.080349203783041,
+      "grad_norm": 1.4532550573349,
+      "learning_rate": 0.0002,
+      "loss": 0.4602,
+      "step": 37610
+    },
+    {
+      "epoch": 6.081965887963786,
+      "grad_norm": 1.3175991773605347,
+      "learning_rate": 0.0002,
+      "loss": 0.4428,
+      "step": 37620
+    },
+    {
+      "epoch": 6.083582572144532,
+      "grad_norm": 1.0624970197677612,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 37630
+    },
+    {
+      "epoch": 6.085199256325277,
+      "grad_norm": 1.099715232849121,
+      "learning_rate": 0.0002,
+      "loss": 0.413,
+      "step": 37640
+    },
+    {
+      "epoch": 6.086815940506022,
+      "grad_norm": 1.0380114316940308,
+      "learning_rate": 0.0002,
+      "loss": 0.4528,
+      "step": 37650
+    },
+    {
+      "epoch": 6.088432624686767,
+      "grad_norm": 1.1136109828948975,
+      "learning_rate": 0.0002,
+      "loss": 0.4373,
+      "step": 37660
+    },
+    {
+      "epoch": 6.0900493088675125,
+      "grad_norm": 0.996498703956604,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 37670
+    },
+    {
+      "epoch": 6.091665993048258,
+      "grad_norm": 1.0552574396133423,
+      "learning_rate": 0.0002,
+      "loss": 0.4713,
+      "step": 37680
+    },
+    {
+      "epoch": 6.093282677229003,
+      "grad_norm": 1.4108527898788452,
+      "learning_rate": 0.0002,
+      "loss": 0.4414,
+      "step": 37690
+    },
+    {
+      "epoch": 6.094899361409748,
+      "grad_norm": 1.1323093175888062,
+      "learning_rate": 0.0002,
+      "loss": 0.4851,
+      "step": 37700
+    },
+    {
+      "epoch": 6.096516045590494,
+      "grad_norm": 0.9364377856254578,
+      "learning_rate": 0.0002,
+      "loss": 0.4455,
+      "step": 37710
+    },
+    {
+      "epoch": 6.0981327297712395,
+      "grad_norm": 1.1300561428070068,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 37720
+    },
+    {
+      "epoch": 6.099749413951985,
+      "grad_norm": 1.0616047382354736,
+      "learning_rate": 0.0002,
+      "loss": 0.4539,
+      "step": 37730
+    },
+    {
+      "epoch": 6.10136609813273,
+      "grad_norm": 1.1205905675888062,
+      "learning_rate": 0.0002,
+      "loss": 0.4516,
+      "step": 37740
+    },
+    {
+      "epoch": 6.102982782313475,
+      "grad_norm": 0.9592534303665161,
+      "learning_rate": 0.0002,
+      "loss": 0.4688,
+      "step": 37750
+    },
+    {
+      "epoch": 6.1045994664942205,
+      "grad_norm": 0.9797531962394714,
+      "learning_rate": 0.0002,
+      "loss": 0.4494,
+      "step": 37760
+    },
+    {
+      "epoch": 6.106216150674966,
+      "grad_norm": 1.093404769897461,
+      "learning_rate": 0.0002,
+      "loss": 0.4237,
+      "step": 37770
+    },
+    {
+      "epoch": 6.107832834855711,
+      "grad_norm": 1.2172642946243286,
+      "learning_rate": 0.0002,
+      "loss": 0.4691,
+      "step": 37780
+    },
+    {
+      "epoch": 6.109449519036456,
+      "grad_norm": 1.0467255115509033,
+      "learning_rate": 0.0002,
+      "loss": 0.4398,
+      "step": 37790
+    },
+    {
+      "epoch": 6.111066203217201,
+      "grad_norm": 1.159318208694458,
+      "learning_rate": 0.0002,
+      "loss": 0.4676,
+      "step": 37800
+    },
+    {
+      "epoch": 6.112682887397947,
+      "grad_norm": 1.0615603923797607,
+      "learning_rate": 0.0002,
+      "loss": 0.4539,
+      "step": 37810
+    },
+    {
+      "epoch": 6.114299571578692,
+      "grad_norm": 1.0542045831680298,
+      "learning_rate": 0.0002,
+      "loss": 0.4957,
+      "step": 37820
+    },
+    {
+      "epoch": 6.115916255759437,
+      "grad_norm": 0.8962697982788086,
+      "learning_rate": 0.0002,
+      "loss": 0.4512,
+      "step": 37830
+    },
+    {
+      "epoch": 6.117532939940182,
+      "grad_norm": 1.106352686882019,
+      "learning_rate": 0.0002,
+      "loss": 0.4519,
+      "step": 37840
+    },
+    {
+      "epoch": 6.1191496241209276,
+      "grad_norm": 1.1660276651382446,
+      "learning_rate": 0.0002,
+      "loss": 0.4421,
+      "step": 37850
+    },
+    {
+      "epoch": 6.120766308301674,
+      "grad_norm": 1.3524385690689087,
+      "learning_rate": 0.0002,
+      "loss": 0.4701,
+      "step": 37860
+    },
+    {
+      "epoch": 6.122382992482419,
+      "grad_norm": 1.1056050062179565,
+      "learning_rate": 0.0002,
+      "loss": 0.4684,
+      "step": 37870
+    },
+    {
+      "epoch": 6.123999676663164,
+      "grad_norm": 1.0772725343704224,
+      "learning_rate": 0.0002,
+      "loss": 0.4518,
+      "step": 37880
+    },
+    {
+      "epoch": 6.125616360843909,
+      "grad_norm": 1.1011115312576294,
+      "learning_rate": 0.0002,
+      "loss": 0.4356,
+      "step": 37890
+    },
+    {
+      "epoch": 6.127233045024655,
+      "grad_norm": 0.8952536582946777,
+      "learning_rate": 0.0002,
+      "loss": 0.4909,
+      "step": 37900
+    },
+    {
+      "epoch": 6.1288497292054,
+      "grad_norm": 1.244398593902588,
+      "learning_rate": 0.0002,
+      "loss": 0.4299,
+      "step": 37910
+    },
+    {
+      "epoch": 6.130466413386145,
+      "grad_norm": 0.9658283591270447,
+      "learning_rate": 0.0002,
+      "loss": 0.4764,
+      "step": 37920
+    },
+    {
+      "epoch": 6.13208309756689,
+      "grad_norm": 1.0649068355560303,
+      "learning_rate": 0.0002,
+      "loss": 0.4378,
+      "step": 37930
+    },
+    {
+      "epoch": 6.1336997817476355,
+      "grad_norm": 0.94698166847229,
+      "learning_rate": 0.0002,
+      "loss": 0.4638,
+      "step": 37940
+    },
+    {
+      "epoch": 6.135316465928381,
+      "grad_norm": 1.1450897455215454,
+      "learning_rate": 0.0002,
+      "loss": 0.488,
+      "step": 37950
+    },
+    {
+      "epoch": 6.136933150109126,
+      "grad_norm": 1.032482624053955,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 37960
+    },
+    {
+      "epoch": 6.138549834289871,
+      "grad_norm": 1.0993428230285645,
+      "learning_rate": 0.0002,
+      "loss": 0.4179,
+      "step": 37970
+    },
+    {
+      "epoch": 6.1401665184706165,
+      "grad_norm": 1.2907029390335083,
+      "learning_rate": 0.0002,
+      "loss": 0.4781,
+      "step": 37980
+    },
+    {
+      "epoch": 6.141783202651362,
+      "grad_norm": 1.1007903814315796,
+      "learning_rate": 0.0002,
+      "loss": 0.4671,
+      "step": 37990
+    },
+    {
+      "epoch": 6.143399886832107,
+      "grad_norm": 0.9286124110221863,
+      "learning_rate": 0.0002,
+      "loss": 0.4213,
+      "step": 38000
+    },
+    {
+      "epoch": 6.145016571012853,
+      "grad_norm": 1.1426366567611694,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 38010
+    },
+    {
+      "epoch": 6.146633255193598,
+      "grad_norm": 1.2608287334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 38020
+    },
+    {
+      "epoch": 6.1482499393743435,
+      "grad_norm": 1.1346837282180786,
+      "learning_rate": 0.0002,
+      "loss": 0.454,
+      "step": 38030
+    },
+    {
+      "epoch": 6.149866623555089,
+      "grad_norm": 1.144080400466919,
+      "learning_rate": 0.0002,
+      "loss": 0.4469,
+      "step": 38040
+    },
+    {
+      "epoch": 6.151483307735834,
+      "grad_norm": 1.3456705808639526,
+      "learning_rate": 0.0002,
+      "loss": 0.4515,
+      "step": 38050
+    },
+    {
+      "epoch": 6.153099991916579,
+      "grad_norm": 1.0517960786819458,
+      "learning_rate": 0.0002,
+      "loss": 0.4775,
+      "step": 38060
+    },
+    {
+      "epoch": 6.154716676097324,
+      "grad_norm": 1.1887445449829102,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 38070
+    },
+    {
+      "epoch": 6.15633336027807,
+      "grad_norm": 1.0449163913726807,
+      "learning_rate": 0.0002,
+      "loss": 0.4516,
+      "step": 38080
+    },
+    {
+      "epoch": 6.157950044458815,
+      "grad_norm": 1.3218743801116943,
+      "learning_rate": 0.0002,
+      "loss": 0.4808,
+      "step": 38090
+    },
+    {
+      "epoch": 6.15956672863956,
+      "grad_norm": 1.003208875656128,
+      "learning_rate": 0.0002,
+      "loss": 0.4632,
+      "step": 38100
+    },
+    {
+      "epoch": 6.161183412820305,
+      "grad_norm": 1.008623719215393,
+      "learning_rate": 0.0002,
+      "loss": 0.4978,
+      "step": 38110
+    },
+    {
+      "epoch": 6.162800097001051,
+      "grad_norm": 1.2122787237167358,
+      "learning_rate": 0.0002,
+      "loss": 0.4608,
+      "step": 38120
+    },
+    {
+      "epoch": 6.164416781181796,
+      "grad_norm": 1.253403902053833,
+      "learning_rate": 0.0002,
+      "loss": 0.4666,
+      "step": 38130
+    },
+    {
+      "epoch": 6.166033465362541,
+      "grad_norm": 1.2289724349975586,
+      "learning_rate": 0.0002,
+      "loss": 0.4778,
+      "step": 38140
+    },
+    {
+      "epoch": 6.167650149543286,
+      "grad_norm": 1.330694556236267,
+      "learning_rate": 0.0002,
+      "loss": 0.4774,
+      "step": 38150
+    },
+    {
+      "epoch": 6.169266833724032,
+      "grad_norm": 1.0946741104125977,
+      "learning_rate": 0.0002,
+      "loss": 0.4699,
+      "step": 38160
+    },
+    {
+      "epoch": 6.170883517904778,
+      "grad_norm": 1.0719934701919556,
+      "learning_rate": 0.0002,
+      "loss": 0.4816,
+      "step": 38170
+    },
+    {
+      "epoch": 6.172500202085523,
+      "grad_norm": 1.1142133474349976,
+      "learning_rate": 0.0002,
+      "loss": 0.4678,
+      "step": 38180
+    },
+    {
+      "epoch": 6.174116886266268,
+      "grad_norm": 1.1221938133239746,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 38190
+    },
+    {
+      "epoch": 6.175733570447013,
+      "grad_norm": 1.1391617059707642,
+      "learning_rate": 0.0002,
+      "loss": 0.4462,
+      "step": 38200
+    },
+    {
+      "epoch": 6.1773502546277586,
+      "grad_norm": 1.2263455390930176,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 38210
+    },
+    {
+      "epoch": 6.178966938808504,
+      "grad_norm": 1.0930434465408325,
+      "learning_rate": 0.0002,
+      "loss": 0.4633,
+      "step": 38220
+    },
+    {
+      "epoch": 6.180583622989249,
+      "grad_norm": 1.3489030599594116,
+      "learning_rate": 0.0002,
+      "loss": 0.4406,
+      "step": 38230
+    },
+    {
+      "epoch": 6.182200307169994,
+      "grad_norm": 1.1383486986160278,
+      "learning_rate": 0.0002,
+      "loss": 0.4994,
+      "step": 38240
+    },
+    {
+      "epoch": 6.1838169913507395,
+      "grad_norm": 1.2408897876739502,
+      "learning_rate": 0.0002,
+      "loss": 0.4851,
+      "step": 38250
+    },
+    {
+      "epoch": 6.185433675531485,
+      "grad_norm": 1.1436222791671753,
+      "learning_rate": 0.0002,
+      "loss": 0.4848,
+      "step": 38260
+    },
+    {
+      "epoch": 6.18705035971223,
+      "grad_norm": 1.370117425918579,
+      "learning_rate": 0.0002,
+      "loss": 0.4594,
+      "step": 38270
+    },
+    {
+      "epoch": 6.188667043892975,
+      "grad_norm": 0.8862423300743103,
+      "learning_rate": 0.0002,
+      "loss": 0.5023,
+      "step": 38280
+    },
+    {
+      "epoch": 6.19028372807372,
+      "grad_norm": 0.9603779315948486,
+      "learning_rate": 0.0002,
+      "loss": 0.4559,
+      "step": 38290
+    },
+    {
+      "epoch": 6.191900412254466,
+      "grad_norm": 1.389291524887085,
+      "learning_rate": 0.0002,
+      "loss": 0.4835,
+      "step": 38300
+    },
+    {
+      "epoch": 6.193517096435212,
+      "grad_norm": 1.0767031908035278,
+      "learning_rate": 0.0002,
+      "loss": 0.4435,
+      "step": 38310
+    },
+    {
+      "epoch": 6.195133780615957,
+      "grad_norm": 1.1800403594970703,
+      "learning_rate": 0.0002,
+      "loss": 0.4683,
+      "step": 38320
+    },
+    {
+      "epoch": 6.196750464796702,
+      "grad_norm": 0.997891366481781,
+      "learning_rate": 0.0002,
+      "loss": 0.4608,
+      "step": 38330
+    },
+    {
+      "epoch": 6.1983671489774474,
+      "grad_norm": 1.1201492547988892,
+      "learning_rate": 0.0002,
+      "loss": 0.4575,
+      "step": 38340
+    },
+    {
+      "epoch": 6.199983833158193,
+      "grad_norm": 0.9769026637077332,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 38350
+    },
+    {
+      "epoch": 6.201600517338938,
+      "grad_norm": 0.9447069764137268,
+      "learning_rate": 0.0002,
+      "loss": 0.4563,
+      "step": 38360
+    },
+    {
+      "epoch": 6.203217201519683,
+      "grad_norm": 1.0959235429763794,
+      "learning_rate": 0.0002,
+      "loss": 0.516,
+      "step": 38370
+    },
+    {
+      "epoch": 6.204833885700428,
+      "grad_norm": 1.2495406866073608,
+      "learning_rate": 0.0002,
+      "loss": 0.4688,
+      "step": 38380
+    },
+    {
+      "epoch": 6.206450569881174,
+      "grad_norm": 0.8589218258857727,
+      "learning_rate": 0.0002,
+      "loss": 0.4445,
+      "step": 38390
+    },
+    {
+      "epoch": 6.208067254061919,
+      "grad_norm": 0.959155797958374,
+      "learning_rate": 0.0002,
+      "loss": 0.4808,
+      "step": 38400
+    },
+    {
+      "epoch": 6.209683938242664,
+      "grad_norm": 1.0105533599853516,
+      "learning_rate": 0.0002,
+      "loss": 0.4622,
+      "step": 38410
+    },
+    {
+      "epoch": 6.211300622423409,
+      "grad_norm": 0.9824615120887756,
+      "learning_rate": 0.0002,
+      "loss": 0.4887,
+      "step": 38420
+    },
+    {
+      "epoch": 6.2129173066041545,
+      "grad_norm": 0.8616500496864319,
+      "learning_rate": 0.0002,
+      "loss": 0.4656,
+      "step": 38430
+    },
+    {
+      "epoch": 6.2145339907849,
+      "grad_norm": 1.2917758226394653,
+      "learning_rate": 0.0002,
+      "loss": 0.449,
+      "step": 38440
+    },
+    {
+      "epoch": 6.216150674965646,
+      "grad_norm": 1.0564531087875366,
+      "learning_rate": 0.0002,
+      "loss": 0.4201,
+      "step": 38450
+    },
+    {
+      "epoch": 6.217767359146391,
+      "grad_norm": 1.152331829071045,
+      "learning_rate": 0.0002,
+      "loss": 0.4849,
+      "step": 38460
+    },
+    {
+      "epoch": 6.219384043327136,
+      "grad_norm": 0.9152206778526306,
+      "learning_rate": 0.0002,
+      "loss": 0.4887,
+      "step": 38470
+    },
+    {
+      "epoch": 6.221000727507882,
+      "grad_norm": 0.9931167960166931,
+      "learning_rate": 0.0002,
+      "loss": 0.4686,
+      "step": 38480
+    },
+    {
+      "epoch": 6.222617411688627,
+      "grad_norm": 1.3248072862625122,
+      "learning_rate": 0.0002,
+      "loss": 0.4765,
+      "step": 38490
+    },
+    {
+      "epoch": 6.224234095869372,
+      "grad_norm": 1.3916507959365845,
+      "learning_rate": 0.0002,
+      "loss": 0.4636,
+      "step": 38500
+    },
+    {
+      "epoch": 6.225850780050117,
+      "grad_norm": 1.1775140762329102,
+      "learning_rate": 0.0002,
+      "loss": 0.506,
+      "step": 38510
+    },
+    {
+      "epoch": 6.2274674642308625,
+      "grad_norm": 1.1581059694290161,
+      "learning_rate": 0.0002,
+      "loss": 0.47,
+      "step": 38520
+    },
+    {
+      "epoch": 6.229084148411608,
+      "grad_norm": 1.359320878982544,
+      "learning_rate": 0.0002,
+      "loss": 0.4679,
+      "step": 38530
+    },
+    {
+      "epoch": 6.230700832592353,
+      "grad_norm": 1.185041904449463,
+      "learning_rate": 0.0002,
+      "loss": 0.4697,
+      "step": 38540
+    },
+    {
+      "epoch": 6.232317516773098,
+      "grad_norm": 1.1861097812652588,
+      "learning_rate": 0.0002,
+      "loss": 0.4815,
+      "step": 38550
+    },
+    {
+      "epoch": 6.233934200953843,
+      "grad_norm": 1.126990556716919,
+      "learning_rate": 0.0002,
+      "loss": 0.4925,
+      "step": 38560
+    },
+    {
+      "epoch": 6.235550885134589,
+      "grad_norm": 0.9744541049003601,
+      "learning_rate": 0.0002,
+      "loss": 0.4414,
+      "step": 38570
+    },
+    {
+      "epoch": 6.237167569315334,
+      "grad_norm": 1.1260887384414673,
+      "learning_rate": 0.0002,
+      "loss": 0.4577,
+      "step": 38580
+    },
+    {
+      "epoch": 6.238784253496079,
+      "grad_norm": 1.1290327310562134,
+      "learning_rate": 0.0002,
+      "loss": 0.4852,
+      "step": 38590
+    },
+    {
+      "epoch": 6.240400937676825,
+      "grad_norm": 1.0952879190444946,
+      "learning_rate": 0.0002,
+      "loss": 0.4805,
+      "step": 38600
+    },
+    {
+      "epoch": 6.2420176218575705,
+      "grad_norm": 1.1037684679031372,
+      "learning_rate": 0.0002,
+      "loss": 0.4436,
+      "step": 38610
+    },
+    {
+      "epoch": 6.243634306038316,
+      "grad_norm": 1.1356085538864136,
+      "learning_rate": 0.0002,
+      "loss": 0.466,
+      "step": 38620
+    },
+    {
+      "epoch": 6.245250990219061,
+      "grad_norm": 1.0677106380462646,
+      "learning_rate": 0.0002,
+      "loss": 0.5129,
+      "step": 38630
+    },
+    {
+      "epoch": 6.246867674399806,
+      "grad_norm": 1.1573411226272583,
+      "learning_rate": 0.0002,
+      "loss": 0.4907,
+      "step": 38640
+    },
+    {
+      "epoch": 6.248484358580551,
+      "grad_norm": 1.2707505226135254,
+      "learning_rate": 0.0002,
+      "loss": 0.5098,
+      "step": 38650
+    },
+    {
+      "epoch": 6.250101042761297,
+      "grad_norm": 1.0480109453201294,
+      "learning_rate": 0.0002,
+      "loss": 0.4926,
+      "step": 38660
+    },
+    {
+      "epoch": 6.251717726942042,
+      "grad_norm": 1.3668724298477173,
+      "learning_rate": 0.0002,
+      "loss": 0.4654,
+      "step": 38670
+    },
+    {
+      "epoch": 6.253334411122787,
+      "grad_norm": 1.217289686203003,
+      "learning_rate": 0.0002,
+      "loss": 0.5128,
+      "step": 38680
+    },
+    {
+      "epoch": 6.254951095303532,
+      "grad_norm": 1.2950236797332764,
+      "learning_rate": 0.0002,
+      "loss": 0.4621,
+      "step": 38690
+    },
+    {
+      "epoch": 6.256567779484278,
+      "grad_norm": 1.4506934881210327,
+      "learning_rate": 0.0002,
+      "loss": 0.5076,
+      "step": 38700
+    },
+    {
+      "epoch": 6.258184463665023,
+      "grad_norm": 1.1248667240142822,
+      "learning_rate": 0.0002,
+      "loss": 0.4803,
+      "step": 38710
+    },
+    {
+      "epoch": 6.259801147845768,
+      "grad_norm": 1.3384023904800415,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 38720
+    },
+    {
+      "epoch": 6.261417832026513,
+      "grad_norm": 1.128074288368225,
+      "learning_rate": 0.0002,
+      "loss": 0.473,
+      "step": 38730
+    },
+    {
+      "epoch": 6.263034516207259,
+      "grad_norm": 1.1169012784957886,
+      "learning_rate": 0.0002,
+      "loss": 0.4638,
+      "step": 38740
+    },
+    {
+      "epoch": 6.264651200388005,
+      "grad_norm": 1.195198893547058,
+      "learning_rate": 0.0002,
+      "loss": 0.4747,
+      "step": 38750
+    },
+    {
+      "epoch": 6.26626788456875,
+      "grad_norm": 1.2471518516540527,
+      "learning_rate": 0.0002,
+      "loss": 0.4906,
+      "step": 38760
+    },
+    {
+      "epoch": 6.267884568749495,
+      "grad_norm": 1.2646394968032837,
+      "learning_rate": 0.0002,
+      "loss": 0.4507,
+      "step": 38770
+    },
+    {
+      "epoch": 6.26950125293024,
+      "grad_norm": 1.0286450386047363,
+      "learning_rate": 0.0002,
+      "loss": 0.4934,
+      "step": 38780
+    },
+    {
+      "epoch": 6.2711179371109855,
+      "grad_norm": 1.2440695762634277,
+      "learning_rate": 0.0002,
+      "loss": 0.4787,
+      "step": 38790
+    },
+    {
+      "epoch": 6.272734621291731,
+      "grad_norm": 0.8941256403923035,
+      "learning_rate": 0.0002,
+      "loss": 0.4806,
+      "step": 38800
+    },
+    {
+      "epoch": 6.274351305472476,
+      "grad_norm": 1.0693447589874268,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 38810
+    },
+    {
+      "epoch": 6.275967989653221,
+      "grad_norm": 1.0936840772628784,
+      "learning_rate": 0.0002,
+      "loss": 0.4408,
+      "step": 38820
+    },
+    {
+      "epoch": 6.2775846738339665,
+      "grad_norm": 1.0961874723434448,
+      "learning_rate": 0.0002,
+      "loss": 0.4729,
+      "step": 38830
+    },
+    {
+      "epoch": 6.279201358014712,
+      "grad_norm": 1.1465433835983276,
+      "learning_rate": 0.0002,
+      "loss": 0.4504,
+      "step": 38840
+    },
+    {
+      "epoch": 6.280818042195457,
+      "grad_norm": 1.2987004518508911,
+      "learning_rate": 0.0002,
+      "loss": 0.4771,
+      "step": 38850
+    },
+    {
+      "epoch": 6.282434726376202,
+      "grad_norm": 1.1310304403305054,
+      "learning_rate": 0.0002,
+      "loss": 0.4945,
+      "step": 38860
+    },
+    {
+      "epoch": 6.284051410556947,
+      "grad_norm": 1.306538462638855,
+      "learning_rate": 0.0002,
+      "loss": 0.5346,
+      "step": 38870
+    },
+    {
+      "epoch": 6.285668094737693,
+      "grad_norm": 1.2405401468276978,
+      "learning_rate": 0.0002,
+      "loss": 0.4873,
+      "step": 38880
+    },
+    {
+      "epoch": 6.287284778918439,
+      "grad_norm": 1.0934767723083496,
+      "learning_rate": 0.0002,
+      "loss": 0.4929,
+      "step": 38890
+    },
+    {
+      "epoch": 6.288901463099184,
+      "grad_norm": 1.3370496034622192,
+      "learning_rate": 0.0002,
+      "loss": 0.4853,
+      "step": 38900
+    },
+    {
+      "epoch": 6.290518147279929,
+      "grad_norm": 1.0319404602050781,
+      "learning_rate": 0.0002,
+      "loss": 0.4892,
+      "step": 38910
+    },
+    {
+      "epoch": 6.292134831460674,
+      "grad_norm": 0.9734271168708801,
+      "learning_rate": 0.0002,
+      "loss": 0.4685,
+      "step": 38920
+    },
+    {
+      "epoch": 6.29375151564142,
+      "grad_norm": 1.0940454006195068,
+      "learning_rate": 0.0002,
+      "loss": 0.5085,
+      "step": 38930
+    },
+    {
+      "epoch": 6.295368199822165,
+      "grad_norm": 1.036500334739685,
+      "learning_rate": 0.0002,
+      "loss": 0.4985,
+      "step": 38940
+    },
+    {
+      "epoch": 6.29698488400291,
+      "grad_norm": 1.020308256149292,
+      "learning_rate": 0.0002,
+      "loss": 0.4878,
+      "step": 38950
+    },
+    {
+      "epoch": 6.298601568183655,
+      "grad_norm": 1.1416399478912354,
+      "learning_rate": 0.0002,
+      "loss": 0.4668,
+      "step": 38960
+    },
+    {
+      "epoch": 6.300218252364401,
+      "grad_norm": 1.2497479915618896,
+      "learning_rate": 0.0002,
+      "loss": 0.4727,
+      "step": 38970
+    },
+    {
+      "epoch": 6.301834936545146,
+      "grad_norm": 1.1692523956298828,
+      "learning_rate": 0.0002,
+      "loss": 0.4721,
+      "step": 38980
+    },
+    {
+      "epoch": 6.303451620725891,
+      "grad_norm": 1.0693109035491943,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 38990
+    },
+    {
+      "epoch": 6.305068304906636,
+      "grad_norm": 0.8883291482925415,
+      "learning_rate": 0.0002,
+      "loss": 0.4875,
+      "step": 39000
+    },
+    {
+      "epoch": 6.3066849890873815,
+      "grad_norm": 1.1445088386535645,
+      "learning_rate": 0.0002,
+      "loss": 0.5371,
+      "step": 39010
+    },
+    {
+      "epoch": 6.308301673268127,
+      "grad_norm": 1.226792335510254,
+      "learning_rate": 0.0002,
+      "loss": 0.5089,
+      "step": 39020
+    },
+    {
+      "epoch": 6.309918357448872,
+      "grad_norm": 1.0498932600021362,
+      "learning_rate": 0.0002,
+      "loss": 0.474,
+      "step": 39030
+    },
+    {
+      "epoch": 6.311535041629618,
+      "grad_norm": 1.0834535360336304,
+      "learning_rate": 0.0002,
+      "loss": 0.4964,
+      "step": 39040
+    },
+    {
+      "epoch": 6.313151725810363,
+      "grad_norm": 1.144666075706482,
+      "learning_rate": 0.0002,
+      "loss": 0.4733,
+      "step": 39050
+    },
+    {
+      "epoch": 6.3147684099911086,
+      "grad_norm": 1.1468489170074463,
+      "learning_rate": 0.0002,
+      "loss": 0.4784,
+      "step": 39060
+    },
+    {
+      "epoch": 6.316385094171854,
+      "grad_norm": 1.290949821472168,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 39070
+    },
+    {
+      "epoch": 6.318001778352599,
+      "grad_norm": 1.087868094444275,
+      "learning_rate": 0.0002,
+      "loss": 0.5002,
+      "step": 39080
+    },
+    {
+      "epoch": 6.319618462533344,
+      "grad_norm": 1.0156296491622925,
+      "learning_rate": 0.0002,
+      "loss": 0.4944,
+      "step": 39090
+    },
+    {
+      "epoch": 6.3212351467140895,
+      "grad_norm": 1.0805060863494873,
+      "learning_rate": 0.0002,
+      "loss": 0.5019,
+      "step": 39100
+    },
+    {
+      "epoch": 6.322851830894835,
+      "grad_norm": 0.9030579924583435,
+      "learning_rate": 0.0002,
+      "loss": 0.4598,
+      "step": 39110
+    },
+    {
+      "epoch": 6.32446851507558,
+      "grad_norm": 1.1488285064697266,
+      "learning_rate": 0.0002,
+      "loss": 0.4635,
+      "step": 39120
+    },
+    {
+      "epoch": 6.326085199256325,
+      "grad_norm": 1.2050796747207642,
+      "learning_rate": 0.0002,
+      "loss": 0.5368,
+      "step": 39130
+    },
+    {
+      "epoch": 6.32770188343707,
+      "grad_norm": 1.093451738357544,
+      "learning_rate": 0.0002,
+      "loss": 0.4854,
+      "step": 39140
+    },
+    {
+      "epoch": 6.329318567617816,
+      "grad_norm": 1.2046772241592407,
+      "learning_rate": 0.0002,
+      "loss": 0.5055,
+      "step": 39150
+    },
+    {
+      "epoch": 6.330935251798561,
+      "grad_norm": 1.045777678489685,
+      "learning_rate": 0.0002,
+      "loss": 0.4703,
+      "step": 39160
+    },
+    {
+      "epoch": 6.332551935979306,
+      "grad_norm": 1.2008492946624756,
+      "learning_rate": 0.0002,
+      "loss": 0.513,
+      "step": 39170
+    },
+    {
+      "epoch": 6.334168620160051,
+      "grad_norm": 1.0613869428634644,
+      "learning_rate": 0.0002,
+      "loss": 0.4909,
+      "step": 39180
+    },
+    {
+      "epoch": 6.3357853043407975,
+      "grad_norm": 1.058440089225769,
+      "learning_rate": 0.0002,
+      "loss": 0.4708,
+      "step": 39190
+    },
+    {
+      "epoch": 6.337401988521543,
+      "grad_norm": 1.195658802986145,
+      "learning_rate": 0.0002,
+      "loss": 0.4719,
+      "step": 39200
+    },
+    {
+      "epoch": 6.339018672702288,
+      "grad_norm": 1.1595174074172974,
+      "learning_rate": 0.0002,
+      "loss": 0.4901,
+      "step": 39210
+    },
+    {
+      "epoch": 6.340635356883033,
+      "grad_norm": 1.0674750804901123,
+      "learning_rate": 0.0002,
+      "loss": 0.4587,
+      "step": 39220
+    },
+    {
+      "epoch": 6.342252041063778,
+      "grad_norm": 1.3306758403778076,
+      "learning_rate": 0.0002,
+      "loss": 0.4801,
+      "step": 39230
+    },
+    {
+      "epoch": 6.343868725244524,
+      "grad_norm": 1.3582593202590942,
+      "learning_rate": 0.0002,
+      "loss": 0.4839,
+      "step": 39240
+    },
+    {
+      "epoch": 6.345485409425269,
+      "grad_norm": 1.2351572513580322,
+      "learning_rate": 0.0002,
+      "loss": 0.4964,
+      "step": 39250
+    },
+    {
+      "epoch": 6.347102093606014,
+      "grad_norm": 1.3623450994491577,
+      "learning_rate": 0.0002,
+      "loss": 0.4806,
+      "step": 39260
+    },
+    {
+      "epoch": 6.348718777786759,
+      "grad_norm": 1.201270580291748,
+      "learning_rate": 0.0002,
+      "loss": 0.466,
+      "step": 39270
+    },
+    {
+      "epoch": 6.3503354619675045,
+      "grad_norm": 0.9300584197044373,
+      "learning_rate": 0.0002,
+      "loss": 0.4899,
+      "step": 39280
+    },
+    {
+      "epoch": 6.35195214614825,
+      "grad_norm": 0.944525957107544,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 39290
+    },
+    {
+      "epoch": 6.353568830328995,
+      "grad_norm": 1.4263732433319092,
+      "learning_rate": 0.0002,
+      "loss": 0.4954,
+      "step": 39300
+    },
+    {
+      "epoch": 6.35518551450974,
+      "grad_norm": 1.392592191696167,
+      "learning_rate": 0.0002,
+      "loss": 0.4982,
+      "step": 39310
+    },
+    {
+      "epoch": 6.3568021986904855,
+      "grad_norm": 1.0753393173217773,
+      "learning_rate": 0.0002,
+      "loss": 0.4868,
+      "step": 39320
+    },
+    {
+      "epoch": 6.358418882871231,
+      "grad_norm": 1.0088151693344116,
+      "learning_rate": 0.0002,
+      "loss": 0.4896,
+      "step": 39330
+    },
+    {
+      "epoch": 6.360035567051977,
+      "grad_norm": 1.1784582138061523,
+      "learning_rate": 0.0002,
+      "loss": 0.4684,
+      "step": 39340
+    },
+    {
+      "epoch": 6.361652251232722,
+      "grad_norm": 1.020526647567749,
+      "learning_rate": 0.0002,
+      "loss": 0.4732,
+      "step": 39350
+    },
+    {
+      "epoch": 6.363268935413467,
+      "grad_norm": 1.1400747299194336,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 39360
+    },
+    {
+      "epoch": 6.3648856195942125,
+      "grad_norm": 0.9960665702819824,
+      "learning_rate": 0.0002,
+      "loss": 0.4976,
+      "step": 39370
+    },
+    {
+      "epoch": 6.366502303774958,
+      "grad_norm": 1.1547569036483765,
+      "learning_rate": 0.0002,
+      "loss": 0.483,
+      "step": 39380
+    },
+    {
+      "epoch": 6.368118987955703,
+      "grad_norm": 1.2180676460266113,
+      "learning_rate": 0.0002,
+      "loss": 0.4861,
+      "step": 39390
+    },
+    {
+      "epoch": 6.369735672136448,
+      "grad_norm": 1.1391799449920654,
+      "learning_rate": 0.0002,
+      "loss": 0.4805,
+      "step": 39400
+    },
+    {
+      "epoch": 6.371352356317193,
+      "grad_norm": 1.2893574237823486,
+      "learning_rate": 0.0002,
+      "loss": 0.5004,
+      "step": 39410
+    },
+    {
+      "epoch": 6.372969040497939,
+      "grad_norm": 1.192878246307373,
+      "learning_rate": 0.0002,
+      "loss": 0.4807,
+      "step": 39420
+    },
+    {
+      "epoch": 6.374585724678684,
+      "grad_norm": 0.9771704077720642,
+      "learning_rate": 0.0002,
+      "loss": 0.4637,
+      "step": 39430
+    },
+    {
+      "epoch": 6.376202408859429,
+      "grad_norm": 1.285387635231018,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 39440
+    },
+    {
+      "epoch": 6.377819093040174,
+      "grad_norm": 1.019957184791565,
+      "learning_rate": 0.0002,
+      "loss": 0.4593,
+      "step": 39450
+    },
+    {
+      "epoch": 6.37943577722092,
+      "grad_norm": 1.2002915143966675,
+      "learning_rate": 0.0002,
+      "loss": 0.473,
+      "step": 39460
+    },
+    {
+      "epoch": 6.381052461401665,
+      "grad_norm": 1.3285092115402222,
+      "learning_rate": 0.0002,
+      "loss": 0.5025,
+      "step": 39470
+    },
+    {
+      "epoch": 6.38266914558241,
+      "grad_norm": 1.097846269607544,
+      "learning_rate": 0.0002,
+      "loss": 0.4626,
+      "step": 39480
+    },
+    {
+      "epoch": 6.384285829763156,
+      "grad_norm": 0.9537988305091858,
+      "learning_rate": 0.0002,
+      "loss": 0.5109,
+      "step": 39490
+    },
+    {
+      "epoch": 6.385902513943901,
+      "grad_norm": 1.0350042581558228,
+      "learning_rate": 0.0002,
+      "loss": 0.4492,
+      "step": 39500
+    },
+    {
+      "epoch": 6.387519198124647,
+      "grad_norm": 0.9559133052825928,
+      "learning_rate": 0.0002,
+      "loss": 0.4824,
+      "step": 39510
+    },
+    {
+      "epoch": 6.389135882305392,
+      "grad_norm": 0.9615123271942139,
+      "learning_rate": 0.0002,
+      "loss": 0.5189,
+      "step": 39520
+    },
+    {
+      "epoch": 6.390752566486137,
+      "grad_norm": 1.0604504346847534,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 39530
+    },
+    {
+      "epoch": 6.392369250666882,
+      "grad_norm": 1.2460750341415405,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 39540
+    },
+    {
+      "epoch": 6.393985934847628,
+      "grad_norm": 1.1496477127075195,
+      "learning_rate": 0.0002,
+      "loss": 0.4929,
+      "step": 39550
+    },
+    {
+      "epoch": 6.395602619028373,
+      "grad_norm": 1.048043966293335,
+      "learning_rate": 0.0002,
+      "loss": 0.4872,
+      "step": 39560
+    },
+    {
+      "epoch": 6.397219303209118,
+      "grad_norm": 1.333539366722107,
+      "learning_rate": 0.0002,
+      "loss": 0.5231,
+      "step": 39570
+    },
+    {
+      "epoch": 6.398835987389863,
+      "grad_norm": 1.0605626106262207,
+      "learning_rate": 0.0002,
+      "loss": 0.4877,
+      "step": 39580
+    },
+    {
+      "epoch": 6.4004526715706085,
+      "grad_norm": 1.163220763206482,
+      "learning_rate": 0.0002,
+      "loss": 0.4643,
+      "step": 39590
+    },
+    {
+      "epoch": 6.402069355751354,
+      "grad_norm": 1.1878494024276733,
+      "learning_rate": 0.0002,
+      "loss": 0.4824,
+      "step": 39600
+    },
+    {
+      "epoch": 6.403686039932099,
+      "grad_norm": 1.4630796909332275,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 39610
+    },
+    {
+      "epoch": 6.405302724112844,
+      "grad_norm": 1.073255181312561,
+      "learning_rate": 0.0002,
+      "loss": 0.4985,
+      "step": 39620
+    },
+    {
+      "epoch": 6.406919408293589,
+      "grad_norm": 1.0538873672485352,
+      "learning_rate": 0.0002,
+      "loss": 0.5108,
+      "step": 39630
+    },
+    {
+      "epoch": 6.4085360924743355,
+      "grad_norm": 1.015525221824646,
+      "learning_rate": 0.0002,
+      "loss": 0.4801,
+      "step": 39640
+    },
+    {
+      "epoch": 6.410152776655081,
+      "grad_norm": 1.1454379558563232,
+      "learning_rate": 0.0002,
+      "loss": 0.4781,
+      "step": 39650
+    },
+    {
+      "epoch": 6.411769460835826,
+      "grad_norm": 1.2801800966262817,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 39660
+    },
+    {
+      "epoch": 6.413386145016571,
+      "grad_norm": 1.077579140663147,
+      "learning_rate": 0.0002,
+      "loss": 0.4804,
+      "step": 39670
+    },
+    {
+      "epoch": 6.4150028291973165,
+      "grad_norm": 1.376662015914917,
+      "learning_rate": 0.0002,
+      "loss": 0.51,
+      "step": 39680
+    },
+    {
+      "epoch": 6.416619513378062,
+      "grad_norm": 1.2064344882965088,
+      "learning_rate": 0.0002,
+      "loss": 0.4956,
+      "step": 39690
+    },
+    {
+      "epoch": 6.418236197558807,
+      "grad_norm": 1.0689115524291992,
+      "learning_rate": 0.0002,
+      "loss": 0.4762,
+      "step": 39700
+    },
+    {
+      "epoch": 6.419852881739552,
+      "grad_norm": 0.9997019171714783,
+      "learning_rate": 0.0002,
+      "loss": 0.4762,
+      "step": 39710
+    },
+    {
+      "epoch": 6.421469565920297,
+      "grad_norm": 1.2368080615997314,
+      "learning_rate": 0.0002,
+      "loss": 0.49,
+      "step": 39720
+    },
+    {
+      "epoch": 6.423086250101043,
+      "grad_norm": 1.2085820436477661,
+      "learning_rate": 0.0002,
+      "loss": 0.4774,
+      "step": 39730
+    },
+    {
+      "epoch": 6.424702934281788,
+      "grad_norm": 1.057246208190918,
+      "learning_rate": 0.0002,
+      "loss": 0.4671,
+      "step": 39740
+    },
+    {
+      "epoch": 6.426319618462533,
+      "grad_norm": 1.1311043500900269,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 39750
+    },
+    {
+      "epoch": 6.427936302643278,
+      "grad_norm": 1.2352231740951538,
+      "learning_rate": 0.0002,
+      "loss": 0.5171,
+      "step": 39760
+    },
+    {
+      "epoch": 6.4295529868240235,
+      "grad_norm": 0.953233540058136,
+      "learning_rate": 0.0002,
+      "loss": 0.466,
+      "step": 39770
+    },
+    {
+      "epoch": 6.431169671004769,
+      "grad_norm": 1.0632505416870117,
+      "learning_rate": 0.0002,
+      "loss": 0.4834,
+      "step": 39780
+    },
+    {
+      "epoch": 6.432786355185515,
+      "grad_norm": 1.0916751623153687,
+      "learning_rate": 0.0002,
+      "loss": 0.5053,
+      "step": 39790
+    },
+    {
+      "epoch": 6.43440303936626,
+      "grad_norm": 0.9732703566551208,
+      "learning_rate": 0.0002,
+      "loss": 0.4788,
+      "step": 39800
+    },
+    {
+      "epoch": 6.436019723547005,
+      "grad_norm": 1.1673705577850342,
+      "learning_rate": 0.0002,
+      "loss": 0.4982,
+      "step": 39810
+    },
+    {
+      "epoch": 6.437636407727751,
+      "grad_norm": 1.1049559116363525,
+      "learning_rate": 0.0002,
+      "loss": 0.4484,
+      "step": 39820
+    },
+    {
+      "epoch": 6.439253091908496,
+      "grad_norm": 1.345277190208435,
+      "learning_rate": 0.0002,
+      "loss": 0.4784,
+      "step": 39830
+    },
+    {
+      "epoch": 6.440869776089241,
+      "grad_norm": 1.1118950843811035,
+      "learning_rate": 0.0002,
+      "loss": 0.4716,
+      "step": 39840
+    },
+    {
+      "epoch": 6.442486460269986,
+      "grad_norm": 1.4872850179672241,
+      "learning_rate": 0.0002,
+      "loss": 0.5133,
+      "step": 39850
+    },
+    {
+      "epoch": 6.4441031444507315,
+      "grad_norm": 1.0763497352600098,
+      "learning_rate": 0.0002,
+      "loss": 0.4532,
+      "step": 39860
+    },
+    {
+      "epoch": 6.445719828631477,
+      "grad_norm": 0.9245555400848389,
+      "learning_rate": 0.0002,
+      "loss": 0.4572,
+      "step": 39870
+    },
+    {
+      "epoch": 6.447336512812222,
+      "grad_norm": 1.4154807329177856,
+      "learning_rate": 0.0002,
+      "loss": 0.4917,
+      "step": 39880
+    },
+    {
+      "epoch": 6.448953196992967,
+      "grad_norm": 1.0885124206542969,
+      "learning_rate": 0.0002,
+      "loss": 0.4852,
+      "step": 39890
+    },
+    {
+      "epoch": 6.450569881173712,
+      "grad_norm": 1.3989344835281372,
+      "learning_rate": 0.0002,
+      "loss": 0.5399,
+      "step": 39900
+    },
+    {
+      "epoch": 6.452186565354458,
+      "grad_norm": 0.9763124585151672,
+      "learning_rate": 0.0002,
+      "loss": 0.509,
+      "step": 39910
+    },
+    {
+      "epoch": 6.453803249535203,
+      "grad_norm": 1.135272741317749,
+      "learning_rate": 0.0002,
+      "loss": 0.5134,
+      "step": 39920
+    },
+    {
+      "epoch": 6.455419933715948,
+      "grad_norm": 1.1140081882476807,
+      "learning_rate": 0.0002,
+      "loss": 0.4941,
+      "step": 39930
+    },
+    {
+      "epoch": 6.457036617896694,
+      "grad_norm": 1.0992448329925537,
+      "learning_rate": 0.0002,
+      "loss": 0.5137,
+      "step": 39940
+    },
+    {
+      "epoch": 6.4586533020774395,
+      "grad_norm": 1.1658501625061035,
+      "learning_rate": 0.0002,
+      "loss": 0.4914,
+      "step": 39950
+    },
+    {
+      "epoch": 6.460269986258185,
+      "grad_norm": 1.1122797727584839,
+      "learning_rate": 0.0002,
+      "loss": 0.5036,
+      "step": 39960
+    },
+    {
+      "epoch": 6.46188667043893,
+      "grad_norm": 0.9664968252182007,
+      "learning_rate": 0.0002,
+      "loss": 0.5159,
+      "step": 39970
+    },
+    {
+      "epoch": 6.463503354619675,
+      "grad_norm": 1.2513965368270874,
+      "learning_rate": 0.0002,
+      "loss": 0.4989,
+      "step": 39980
+    },
+    {
+      "epoch": 6.46512003880042,
+      "grad_norm": 1.1198630332946777,
+      "learning_rate": 0.0002,
+      "loss": 0.4694,
+      "step": 39990
+    },
+    {
+      "epoch": 6.466736722981166,
+      "grad_norm": 0.8783249855041504,
+      "learning_rate": 0.0002,
+      "loss": 0.5023,
+      "step": 40000
+    },
+    {
+      "epoch": 6.468353407161911,
+      "grad_norm": 1.1313109397888184,
+      "learning_rate": 0.0002,
+      "loss": 0.4648,
+      "step": 40010
+    },
+    {
+      "epoch": 6.469970091342656,
+      "grad_norm": 1.0854487419128418,
+      "learning_rate": 0.0002,
+      "loss": 0.4965,
+      "step": 40020
+    },
+    {
+      "epoch": 6.471586775523401,
+      "grad_norm": 1.1738566160202026,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 40030
+    },
+    {
+      "epoch": 6.473203459704147,
+      "grad_norm": 0.9720084071159363,
+      "learning_rate": 0.0002,
+      "loss": 0.4947,
+      "step": 40040
+    },
+    {
+      "epoch": 6.474820143884892,
+      "grad_norm": 1.105618953704834,
+      "learning_rate": 0.0002,
+      "loss": 0.5218,
+      "step": 40050
+    },
+    {
+      "epoch": 6.476436828065637,
+      "grad_norm": 1.2007657289505005,
+      "learning_rate": 0.0002,
+      "loss": 0.4943,
+      "step": 40060
+    },
+    {
+      "epoch": 6.478053512246382,
+      "grad_norm": 1.088402509689331,
+      "learning_rate": 0.0002,
+      "loss": 0.4882,
+      "step": 40070
+    },
+    {
+      "epoch": 6.4796701964271275,
+      "grad_norm": 1.0775291919708252,
+      "learning_rate": 0.0002,
+      "loss": 0.504,
+      "step": 40080
+    },
+    {
+      "epoch": 6.481286880607874,
+      "grad_norm": 1.1018189191818237,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 40090
+    },
+    {
+      "epoch": 6.482903564788619,
+      "grad_norm": 1.1676557064056396,
+      "learning_rate": 0.0002,
+      "loss": 0.488,
+      "step": 40100
+    },
+    {
+      "epoch": 6.484520248969364,
+      "grad_norm": 0.9619805812835693,
+      "learning_rate": 0.0002,
+      "loss": 0.4818,
+      "step": 40110
+    },
+    {
+      "epoch": 6.486136933150109,
+      "grad_norm": 1.2408208847045898,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 40120
+    },
+    {
+      "epoch": 6.4877536173308545,
+      "grad_norm": 1.3488136529922485,
+      "learning_rate": 0.0002,
+      "loss": 0.4668,
+      "step": 40130
+    },
+    {
+      "epoch": 6.4893703015116,
+      "grad_norm": 0.9864488244056702,
+      "learning_rate": 0.0002,
+      "loss": 0.4774,
+      "step": 40140
+    },
+    {
+      "epoch": 6.490986985692345,
+      "grad_norm": 0.9437947273254395,
+      "learning_rate": 0.0002,
+      "loss": 0.4651,
+      "step": 40150
+    },
+    {
+      "epoch": 6.49260366987309,
+      "grad_norm": 1.2005455493927002,
+      "learning_rate": 0.0002,
+      "loss": 0.542,
+      "step": 40160
+    },
+    {
+      "epoch": 6.4942203540538355,
+      "grad_norm": 1.0796732902526855,
+      "learning_rate": 0.0002,
+      "loss": 0.4704,
+      "step": 40170
+    },
+    {
+      "epoch": 6.495837038234581,
+      "grad_norm": 1.1347825527191162,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 40180
+    },
+    {
+      "epoch": 6.497453722415326,
+      "grad_norm": 1.2311455011367798,
+      "learning_rate": 0.0002,
+      "loss": 0.5215,
+      "step": 40190
+    },
+    {
+      "epoch": 6.499070406596071,
+      "grad_norm": 1.068609356880188,
+      "learning_rate": 0.0002,
+      "loss": 0.5043,
+      "step": 40200
+    },
+    {
+      "epoch": 6.500687090776816,
+      "grad_norm": 1.196425437927246,
+      "learning_rate": 0.0002,
+      "loss": 0.4868,
+      "step": 40210
+    },
+    {
+      "epoch": 6.5023037749575625,
+      "grad_norm": 1.183927297592163,
+      "learning_rate": 0.0002,
+      "loss": 0.4881,
+      "step": 40220
+    },
+    {
+      "epoch": 6.503920459138307,
+      "grad_norm": 0.9099724292755127,
+      "learning_rate": 0.0002,
+      "loss": 0.4958,
+      "step": 40230
+    },
+    {
+      "epoch": 6.505537143319053,
+      "grad_norm": 0.9261038899421692,
+      "learning_rate": 0.0002,
+      "loss": 0.4816,
+      "step": 40240
+    },
+    {
+      "epoch": 6.507153827499798,
+      "grad_norm": 1.185491681098938,
+      "learning_rate": 0.0002,
+      "loss": 0.5151,
+      "step": 40250
+    },
+    {
+      "epoch": 6.508770511680543,
+      "grad_norm": 1.1866052150726318,
+      "learning_rate": 0.0002,
+      "loss": 0.4853,
+      "step": 40260
+    },
+    {
+      "epoch": 6.510387195861289,
+      "grad_norm": 1.1600912809371948,
+      "learning_rate": 0.0002,
+      "loss": 0.491,
+      "step": 40270
+    },
+    {
+      "epoch": 6.512003880042034,
+      "grad_norm": 0.9609426259994507,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 40280
+    },
+    {
+      "epoch": 6.513620564222779,
+      "grad_norm": 1.078864336013794,
+      "learning_rate": 0.0002,
+      "loss": 0.4794,
+      "step": 40290
+    },
+    {
+      "epoch": 6.515237248403524,
+      "grad_norm": 1.042761206626892,
+      "learning_rate": 0.0002,
+      "loss": 0.46,
+      "step": 40300
+    },
+    {
+      "epoch": 6.51685393258427,
+      "grad_norm": 0.9742481112480164,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 40310
+    },
+    {
+      "epoch": 6.518470616765015,
+      "grad_norm": 1.2544835805892944,
+      "learning_rate": 0.0002,
+      "loss": 0.5234,
+      "step": 40320
+    },
+    {
+      "epoch": 6.52008730094576,
+      "grad_norm": 1.3019760847091675,
+      "learning_rate": 0.0002,
+      "loss": 0.4815,
+      "step": 40330
+    },
+    {
+      "epoch": 6.521703985126505,
+      "grad_norm": 1.3196964263916016,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 40340
+    },
+    {
+      "epoch": 6.5233206693072505,
+      "grad_norm": 1.2795668840408325,
+      "learning_rate": 0.0002,
+      "loss": 0.4979,
+      "step": 40350
+    },
+    {
+      "epoch": 6.524937353487996,
+      "grad_norm": 1.1618940830230713,
+      "learning_rate": 0.0002,
+      "loss": 0.5075,
+      "step": 40360
+    },
+    {
+      "epoch": 6.526554037668742,
+      "grad_norm": 1.330543041229248,
+      "learning_rate": 0.0002,
+      "loss": 0.5081,
+      "step": 40370
+    },
+    {
+      "epoch": 6.528170721849486,
+      "grad_norm": 1.1946901082992554,
+      "learning_rate": 0.0002,
+      "loss": 0.5055,
+      "step": 40380
+    },
+    {
+      "epoch": 6.529787406030232,
+      "grad_norm": 1.1708201169967651,
+      "learning_rate": 0.0002,
+      "loss": 0.4518,
+      "step": 40390
+    },
+    {
+      "epoch": 6.531404090210978,
+      "grad_norm": 0.894036591053009,
+      "learning_rate": 0.0002,
+      "loss": 0.4556,
+      "step": 40400
+    },
+    {
+      "epoch": 6.533020774391723,
+      "grad_norm": 1.1199041604995728,
+      "learning_rate": 0.0002,
+      "loss": 0.4919,
+      "step": 40410
+    },
+    {
+      "epoch": 6.534637458572468,
+      "grad_norm": 1.180317759513855,
+      "learning_rate": 0.0002,
+      "loss": 0.471,
+      "step": 40420
+    },
+    {
+      "epoch": 6.536254142753213,
+      "grad_norm": 1.37367582321167,
+      "learning_rate": 0.0002,
+      "loss": 0.4914,
+      "step": 40430
+    },
+    {
+      "epoch": 6.5378708269339585,
+      "grad_norm": 1.134791612625122,
+      "learning_rate": 0.0002,
+      "loss": 0.4561,
+      "step": 40440
+    },
+    {
+      "epoch": 6.539487511114704,
+      "grad_norm": 1.1160204410552979,
+      "learning_rate": 0.0002,
+      "loss": 0.5337,
+      "step": 40450
+    },
+    {
+      "epoch": 6.541104195295449,
+      "grad_norm": 1.268347978591919,
+      "learning_rate": 0.0002,
+      "loss": 0.5299,
+      "step": 40460
+    },
+    {
+      "epoch": 6.542720879476194,
+      "grad_norm": 1.1424330472946167,
+      "learning_rate": 0.0002,
+      "loss": 0.5167,
+      "step": 40470
+    },
+    {
+      "epoch": 6.544337563656939,
+      "grad_norm": 1.3098465204238892,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 40480
+    },
+    {
+      "epoch": 6.545954247837685,
+      "grad_norm": 1.3439544439315796,
+      "learning_rate": 0.0002,
+      "loss": 0.4865,
+      "step": 40490
+    },
+    {
+      "epoch": 6.54757093201843,
+      "grad_norm": 1.2708452939987183,
+      "learning_rate": 0.0002,
+      "loss": 0.5183,
+      "step": 40500
+    },
+    {
+      "epoch": 6.549187616199175,
+      "grad_norm": 1.483680248260498,
+      "learning_rate": 0.0002,
+      "loss": 0.5099,
+      "step": 40510
+    },
+    {
+      "epoch": 6.550804300379921,
+      "grad_norm": 1.1697806119918823,
+      "learning_rate": 0.0002,
+      "loss": 0.4811,
+      "step": 40520
+    },
+    {
+      "epoch": 6.5524209845606665,
+      "grad_norm": 1.1665642261505127,
+      "learning_rate": 0.0002,
+      "loss": 0.4814,
+      "step": 40530
+    },
+    {
+      "epoch": 6.554037668741412,
+      "grad_norm": 1.1243325471878052,
+      "learning_rate": 0.0002,
+      "loss": 0.4985,
+      "step": 40540
+    },
+    {
+      "epoch": 6.555654352922157,
+      "grad_norm": 1.0277988910675049,
+      "learning_rate": 0.0002,
+      "loss": 0.4936,
+      "step": 40550
+    },
+    {
+      "epoch": 6.557271037102902,
+      "grad_norm": 1.1466810703277588,
+      "learning_rate": 0.0002,
+      "loss": 0.487,
+      "step": 40560
+    },
+    {
+      "epoch": 6.558887721283647,
+      "grad_norm": 1.1415363550186157,
+      "learning_rate": 0.0002,
+      "loss": 0.4851,
+      "step": 40570
+    },
+    {
+      "epoch": 6.560504405464393,
+      "grad_norm": 1.1923491954803467,
+      "learning_rate": 0.0002,
+      "loss": 0.4631,
+      "step": 40580
+    },
+    {
+      "epoch": 6.562121089645138,
+      "grad_norm": 0.9264549612998962,
+      "learning_rate": 0.0002,
+      "loss": 0.5071,
+      "step": 40590
+    },
+    {
+      "epoch": 6.563737773825883,
+      "grad_norm": 0.8810341954231262,
+      "learning_rate": 0.0002,
+      "loss": 0.466,
+      "step": 40600
+    },
+    {
+      "epoch": 6.565354458006628,
+      "grad_norm": 2.3296701908111572,
+      "learning_rate": 0.0002,
+      "loss": 0.5085,
+      "step": 40610
+    },
+    {
+      "epoch": 6.5669711421873735,
+      "grad_norm": 1.0865163803100586,
+      "learning_rate": 0.0002,
+      "loss": 0.5196,
+      "step": 40620
+    },
+    {
+      "epoch": 6.568587826368119,
+      "grad_norm": 0.9844607710838318,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 40630
+    },
+    {
+      "epoch": 6.570204510548864,
+      "grad_norm": 1.1686855554580688,
+      "learning_rate": 0.0002,
+      "loss": 0.5437,
+      "step": 40640
+    },
+    {
+      "epoch": 6.571821194729609,
+      "grad_norm": 1.016829252243042,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 40650
+    },
+    {
+      "epoch": 6.5734378789103545,
+      "grad_norm": 1.2789337635040283,
+      "learning_rate": 0.0002,
+      "loss": 0.5243,
+      "step": 40660
+    },
+    {
+      "epoch": 6.575054563091101,
+      "grad_norm": 1.0819072723388672,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 40670
+    },
+    {
+      "epoch": 6.576671247271846,
+      "grad_norm": 1.1478345394134521,
+      "learning_rate": 0.0002,
+      "loss": 0.5024,
+      "step": 40680
+    },
+    {
+      "epoch": 6.578287931452591,
+      "grad_norm": 0.7972208857536316,
+      "learning_rate": 0.0002,
+      "loss": 0.5282,
+      "step": 40690
+    },
+    {
+      "epoch": 6.579904615633336,
+      "grad_norm": 1.1481789350509644,
+      "learning_rate": 0.0002,
+      "loss": 0.4877,
+      "step": 40700
+    },
+    {
+      "epoch": 6.5815212998140815,
+      "grad_norm": 1.0921871662139893,
+      "learning_rate": 0.0002,
+      "loss": 0.5143,
+      "step": 40710
+    },
+    {
+      "epoch": 6.583137983994827,
+      "grad_norm": 1.0230315923690796,
+      "learning_rate": 0.0002,
+      "loss": 0.5441,
+      "step": 40720
+    },
+    {
+      "epoch": 6.584754668175572,
+      "grad_norm": 1.151049017906189,
+      "learning_rate": 0.0002,
+      "loss": 0.4734,
+      "step": 40730
+    },
+    {
+      "epoch": 6.586371352356317,
+      "grad_norm": 1.4016883373260498,
+      "learning_rate": 0.0002,
+      "loss": 0.4782,
+      "step": 40740
+    },
+    {
+      "epoch": 6.587988036537062,
+      "grad_norm": 1.2211825847625732,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 40750
+    },
+    {
+      "epoch": 6.589604720717808,
+      "grad_norm": 1.2803404331207275,
+      "learning_rate": 0.0002,
+      "loss": 0.4815,
+      "step": 40760
+    },
+    {
+      "epoch": 6.591221404898553,
+      "grad_norm": 1.1119942665100098,
+      "learning_rate": 0.0002,
+      "loss": 0.5329,
+      "step": 40770
+    },
+    {
+      "epoch": 6.592838089079298,
+      "grad_norm": 1.464650273323059,
+      "learning_rate": 0.0002,
+      "loss": 0.5135,
+      "step": 40780
+    },
+    {
+      "epoch": 6.594454773260043,
+      "grad_norm": 1.1751397848129272,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 40790
+    },
+    {
+      "epoch": 6.596071457440789,
+      "grad_norm": 1.0866316556930542,
+      "learning_rate": 0.0002,
+      "loss": 0.4772,
+      "step": 40800
+    },
+    {
+      "epoch": 6.597688141621534,
+      "grad_norm": 1.1733694076538086,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 40810
+    },
+    {
+      "epoch": 6.59930482580228,
+      "grad_norm": 1.184708833694458,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 40820
+    },
+    {
+      "epoch": 6.600921509983025,
+      "grad_norm": 1.406081199645996,
+      "learning_rate": 0.0002,
+      "loss": 0.4885,
+      "step": 40830
+    },
+    {
+      "epoch": 6.60253819416377,
+      "grad_norm": 0.9658212661743164,
+      "learning_rate": 0.0002,
+      "loss": 0.499,
+      "step": 40840
+    },
+    {
+      "epoch": 6.604154878344516,
+      "grad_norm": 1.1457678079605103,
+      "learning_rate": 0.0002,
+      "loss": 0.5113,
+      "step": 40850
+    },
+    {
+      "epoch": 6.605771562525261,
+      "grad_norm": 1.0487784147262573,
+      "learning_rate": 0.0002,
+      "loss": 0.4916,
+      "step": 40860
+    },
+    {
+      "epoch": 6.607388246706006,
+      "grad_norm": 0.9357177019119263,
+      "learning_rate": 0.0002,
+      "loss": 0.4682,
+      "step": 40870
+    },
+    {
+      "epoch": 6.609004930886751,
+      "grad_norm": 1.1479727029800415,
+      "learning_rate": 0.0002,
+      "loss": 0.4751,
+      "step": 40880
+    },
+    {
+      "epoch": 6.610621615067497,
+      "grad_norm": 1.3729329109191895,
+      "learning_rate": 0.0002,
+      "loss": 0.5493,
+      "step": 40890
+    },
+    {
+      "epoch": 6.612238299248242,
+      "grad_norm": 1.0085599422454834,
+      "learning_rate": 0.0002,
+      "loss": 0.4886,
+      "step": 40900
+    },
+    {
+      "epoch": 6.613854983428987,
+      "grad_norm": 1.2750911712646484,
+      "learning_rate": 0.0002,
+      "loss": 0.516,
+      "step": 40910
+    },
+    {
+      "epoch": 6.615471667609732,
+      "grad_norm": 1.1929547786712646,
+      "learning_rate": 0.0002,
+      "loss": 0.5342,
+      "step": 40920
+    },
+    {
+      "epoch": 6.6170883517904775,
+      "grad_norm": 1.0821375846862793,
+      "learning_rate": 0.0002,
+      "loss": 0.4919,
+      "step": 40930
+    },
+    {
+      "epoch": 6.618705035971223,
+      "grad_norm": 1.197347640991211,
+      "learning_rate": 0.0002,
+      "loss": 0.5057,
+      "step": 40940
+    },
+    {
+      "epoch": 6.620321720151968,
+      "grad_norm": 1.2074699401855469,
+      "learning_rate": 0.0002,
+      "loss": 0.492,
+      "step": 40950
+    },
+    {
+      "epoch": 6.621938404332713,
+      "grad_norm": 1.312009572982788,
+      "learning_rate": 0.0002,
+      "loss": 0.5089,
+      "step": 40960
+    },
+    {
+      "epoch": 6.623555088513459,
+      "grad_norm": 1.4381471872329712,
+      "learning_rate": 0.0002,
+      "loss": 0.5476,
+      "step": 40970
+    },
+    {
+      "epoch": 6.6251717726942045,
+      "grad_norm": 1.1574671268463135,
+      "learning_rate": 0.0002,
+      "loss": 0.4904,
+      "step": 40980
+    },
+    {
+      "epoch": 6.62678845687495,
+      "grad_norm": 0.885661780834198,
+      "learning_rate": 0.0002,
+      "loss": 0.531,
+      "step": 40990
+    },
+    {
+      "epoch": 6.628405141055695,
+      "grad_norm": 1.024571180343628,
+      "learning_rate": 0.0002,
+      "loss": 0.5145,
+      "step": 41000
+    },
+    {
+      "epoch": 6.63002182523644,
+      "grad_norm": 1.103437900543213,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 41010
+    },
+    {
+      "epoch": 6.6316385094171855,
+      "grad_norm": 1.122450828552246,
+      "learning_rate": 0.0002,
+      "loss": 0.4671,
+      "step": 41020
+    },
+    {
+      "epoch": 6.633255193597931,
+      "grad_norm": 1.2256295680999756,
+      "learning_rate": 0.0002,
+      "loss": 0.5134,
+      "step": 41030
+    },
+    {
+      "epoch": 6.634871877778676,
+      "grad_norm": 1.364594578742981,
+      "learning_rate": 0.0002,
+      "loss": 0.4908,
+      "step": 41040
+    },
+    {
+      "epoch": 6.636488561959421,
+      "grad_norm": 0.9550056457519531,
+      "learning_rate": 0.0002,
+      "loss": 0.4964,
+      "step": 41050
+    },
+    {
+      "epoch": 6.638105246140166,
+      "grad_norm": 1.3174707889556885,
+      "learning_rate": 0.0002,
+      "loss": 0.5028,
+      "step": 41060
+    },
+    {
+      "epoch": 6.639721930320912,
+      "grad_norm": 1.0835540294647217,
+      "learning_rate": 0.0002,
+      "loss": 0.4717,
+      "step": 41070
+    },
+    {
+      "epoch": 6.641338614501657,
+      "grad_norm": 1.1432770490646362,
+      "learning_rate": 0.0002,
+      "loss": 0.497,
+      "step": 41080
+    },
+    {
+      "epoch": 6.642955298682402,
+      "grad_norm": 1.2398556470870972,
+      "learning_rate": 0.0002,
+      "loss": 0.4903,
+      "step": 41090
+    },
+    {
+      "epoch": 6.644571982863147,
+      "grad_norm": 1.1147747039794922,
+      "learning_rate": 0.0002,
+      "loss": 0.4991,
+      "step": 41100
+    },
+    {
+      "epoch": 6.6461886670438926,
+      "grad_norm": 1.0730493068695068,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 41110
+    },
+    {
+      "epoch": 6.647805351224639,
+      "grad_norm": 1.3218451738357544,
+      "learning_rate": 0.0002,
+      "loss": 0.486,
+      "step": 41120
+    },
+    {
+      "epoch": 6.649422035405384,
+      "grad_norm": 1.3027331829071045,
+      "learning_rate": 0.0002,
+      "loss": 0.5276,
+      "step": 41130
+    },
+    {
+      "epoch": 6.651038719586129,
+      "grad_norm": 1.0280735492706299,
+      "learning_rate": 0.0002,
+      "loss": 0.5263,
+      "step": 41140
+    },
+    {
+      "epoch": 6.652655403766874,
+      "grad_norm": 1.109916090965271,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 41150
+    },
+    {
+      "epoch": 6.65427208794762,
+      "grad_norm": 1.078734040260315,
+      "learning_rate": 0.0002,
+      "loss": 0.5001,
+      "step": 41160
+    },
+    {
+      "epoch": 6.655888772128365,
+      "grad_norm": 1.1595654487609863,
+      "learning_rate": 0.0002,
+      "loss": 0.484,
+      "step": 41170
+    },
+    {
+      "epoch": 6.65750545630911,
+      "grad_norm": 1.1701031923294067,
+      "learning_rate": 0.0002,
+      "loss": 0.5101,
+      "step": 41180
+    },
+    {
+      "epoch": 6.659122140489855,
+      "grad_norm": 1.0424643754959106,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 41190
+    },
+    {
+      "epoch": 6.6607388246706005,
+      "grad_norm": 1.22880220413208,
+      "learning_rate": 0.0002,
+      "loss": 0.4863,
+      "step": 41200
+    },
+    {
+      "epoch": 6.662355508851346,
+      "grad_norm": 1.1907655000686646,
+      "learning_rate": 0.0002,
+      "loss": 0.4987,
+      "step": 41210
+    },
+    {
+      "epoch": 6.663972193032091,
+      "grad_norm": 1.0765007734298706,
+      "learning_rate": 0.0002,
+      "loss": 0.5343,
+      "step": 41220
+    },
+    {
+      "epoch": 6.665588877212836,
+      "grad_norm": 0.9994917511940002,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 41230
+    },
+    {
+      "epoch": 6.6672055613935814,
+      "grad_norm": 0.968578040599823,
+      "learning_rate": 0.0002,
+      "loss": 0.507,
+      "step": 41240
+    },
+    {
+      "epoch": 6.668822245574327,
+      "grad_norm": 1.0576032400131226,
+      "learning_rate": 0.0002,
+      "loss": 0.5068,
+      "step": 41250
+    },
+    {
+      "epoch": 6.670438929755072,
+      "grad_norm": 1.2183765172958374,
+      "learning_rate": 0.0002,
+      "loss": 0.486,
+      "step": 41260
+    },
+    {
+      "epoch": 6.672055613935818,
+      "grad_norm": 1.2548623085021973,
+      "learning_rate": 0.0002,
+      "loss": 0.4764,
+      "step": 41270
+    },
+    {
+      "epoch": 6.673672298116563,
+      "grad_norm": 1.0848388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.5014,
+      "step": 41280
+    },
+    {
+      "epoch": 6.6752889822973085,
+      "grad_norm": 1.21421217918396,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 41290
+    },
+    {
+      "epoch": 6.676905666478054,
+      "grad_norm": 1.1453598737716675,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 41300
+    },
+    {
+      "epoch": 6.678522350658799,
+      "grad_norm": 1.2682722806930542,
+      "learning_rate": 0.0002,
+      "loss": 0.5033,
+      "step": 41310
+    },
+    {
+      "epoch": 6.680139034839544,
+      "grad_norm": 1.1659725904464722,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 41320
+    },
+    {
+      "epoch": 6.681755719020289,
+      "grad_norm": 1.36194908618927,
+      "learning_rate": 0.0002,
+      "loss": 0.5505,
+      "step": 41330
+    },
+    {
+      "epoch": 6.683372403201035,
+      "grad_norm": 1.1712592840194702,
+      "learning_rate": 0.0002,
+      "loss": 0.5127,
+      "step": 41340
+    },
+    {
+      "epoch": 6.68498908738178,
+      "grad_norm": 1.4168336391448975,
+      "learning_rate": 0.0002,
+      "loss": 0.5082,
+      "step": 41350
+    },
+    {
+      "epoch": 6.686605771562525,
+      "grad_norm": 1.0395328998565674,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 41360
+    },
+    {
+      "epoch": 6.68822245574327,
+      "grad_norm": 1.2511054277420044,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 41370
+    },
+    {
+      "epoch": 6.689839139924016,
+      "grad_norm": 1.0438542366027832,
+      "learning_rate": 0.0002,
+      "loss": 0.5027,
+      "step": 41380
+    },
+    {
+      "epoch": 6.691455824104761,
+      "grad_norm": 1.08684241771698,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 41390
+    },
+    {
+      "epoch": 6.693072508285506,
+      "grad_norm": 1.250788927078247,
+      "learning_rate": 0.0002,
+      "loss": 0.5224,
+      "step": 41400
+    },
+    {
+      "epoch": 6.694689192466251,
+      "grad_norm": 1.313890814781189,
+      "learning_rate": 0.0002,
+      "loss": 0.4921,
+      "step": 41410
+    },
+    {
+      "epoch": 6.696305876646997,
+      "grad_norm": 1.3218982219696045,
+      "learning_rate": 0.0002,
+      "loss": 0.5028,
+      "step": 41420
+    },
+    {
+      "epoch": 6.697922560827743,
+      "grad_norm": 1.0366582870483398,
+      "learning_rate": 0.0002,
+      "loss": 0.4851,
+      "step": 41430
+    },
+    {
+      "epoch": 6.699539245008488,
+      "grad_norm": 1.066121220588684,
+      "learning_rate": 0.0002,
+      "loss": 0.5103,
+      "step": 41440
+    },
+    {
+      "epoch": 6.701155929189233,
+      "grad_norm": 1.0239925384521484,
+      "learning_rate": 0.0002,
+      "loss": 0.4966,
+      "step": 41450
+    },
+    {
+      "epoch": 6.702772613369978,
+      "grad_norm": 0.9402176141738892,
+      "learning_rate": 0.0002,
+      "loss": 0.4767,
+      "step": 41460
+    },
+    {
+      "epoch": 6.7043892975507235,
+      "grad_norm": 1.391718864440918,
+      "learning_rate": 0.0002,
+      "loss": 0.5381,
+      "step": 41470
+    },
+    {
+      "epoch": 6.706005981731469,
+      "grad_norm": 1.215600609779358,
+      "learning_rate": 0.0002,
+      "loss": 0.512,
+      "step": 41480
+    },
+    {
+      "epoch": 6.707622665912214,
+      "grad_norm": 1.063722848892212,
+      "learning_rate": 0.0002,
+      "loss": 0.5219,
+      "step": 41490
+    },
+    {
+      "epoch": 6.709239350092959,
+      "grad_norm": 1.132149577140808,
+      "learning_rate": 0.0002,
+      "loss": 0.492,
+      "step": 41500
+    },
+    {
+      "epoch": 6.7108560342737045,
+      "grad_norm": 1.0302950143814087,
+      "learning_rate": 0.0002,
+      "loss": 0.4812,
+      "step": 41510
+    },
+    {
+      "epoch": 6.71247271845445,
+      "grad_norm": 1.5342752933502197,
+      "learning_rate": 0.0002,
+      "loss": 0.5141,
+      "step": 41520
+    },
+    {
+      "epoch": 6.714089402635195,
+      "grad_norm": 1.177137017250061,
+      "learning_rate": 0.0002,
+      "loss": 0.5123,
+      "step": 41530
+    },
+    {
+      "epoch": 6.71570608681594,
+      "grad_norm": 1.2335538864135742,
+      "learning_rate": 0.0002,
+      "loss": 0.5082,
+      "step": 41540
+    },
+    {
+      "epoch": 6.717322770996686,
+      "grad_norm": 1.140604853630066,
+      "learning_rate": 0.0002,
+      "loss": 0.4864,
+      "step": 41550
+    },
+    {
+      "epoch": 6.718939455177431,
+      "grad_norm": 1.3567465543746948,
+      "learning_rate": 0.0002,
+      "loss": 0.4888,
+      "step": 41560
+    },
+    {
+      "epoch": 6.720556139358177,
+      "grad_norm": 1.0693929195404053,
+      "learning_rate": 0.0002,
+      "loss": 0.5183,
+      "step": 41570
+    },
+    {
+      "epoch": 6.722172823538922,
+      "grad_norm": 1.1592605113983154,
+      "learning_rate": 0.0002,
+      "loss": 0.5131,
+      "step": 41580
+    },
+    {
+      "epoch": 6.723789507719667,
+      "grad_norm": 0.989006519317627,
+      "learning_rate": 0.0002,
+      "loss": 0.5476,
+      "step": 41590
+    },
+    {
+      "epoch": 6.7254061919004124,
+      "grad_norm": 1.04103422164917,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 41600
+    },
+    {
+      "epoch": 6.727022876081158,
+      "grad_norm": 1.1129004955291748,
+      "learning_rate": 0.0002,
+      "loss": 0.4823,
+      "step": 41610
+    },
+    {
+      "epoch": 6.728639560261903,
+      "grad_norm": 1.1473113298416138,
+      "learning_rate": 0.0002,
+      "loss": 0.5032,
+      "step": 41620
+    },
+    {
+      "epoch": 6.730256244442648,
+      "grad_norm": 1.348036527633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 41630
+    },
+    {
+      "epoch": 6.731872928623393,
+      "grad_norm": 1.259942650794983,
+      "learning_rate": 0.0002,
+      "loss": 0.4983,
+      "step": 41640
+    },
+    {
+      "epoch": 6.733489612804139,
+      "grad_norm": 1.0591514110565186,
+      "learning_rate": 0.0002,
+      "loss": 0.5182,
+      "step": 41650
+    },
+    {
+      "epoch": 6.735106296984884,
+      "grad_norm": 0.9737129211425781,
+      "learning_rate": 0.0002,
+      "loss": 0.4886,
+      "step": 41660
+    },
+    {
+      "epoch": 6.736722981165629,
+      "grad_norm": 1.2520451545715332,
+      "learning_rate": 0.0002,
+      "loss": 0.5051,
+      "step": 41670
+    },
+    {
+      "epoch": 6.738339665346374,
+      "grad_norm": 1.0555530786514282,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 41680
+    },
+    {
+      "epoch": 6.7399563495271195,
+      "grad_norm": 1.0025697946548462,
+      "learning_rate": 0.0002,
+      "loss": 0.4954,
+      "step": 41690
+    },
+    {
+      "epoch": 6.741573033707866,
+      "grad_norm": 1.1114100217819214,
+      "learning_rate": 0.0002,
+      "loss": 0.5485,
+      "step": 41700
+    },
+    {
+      "epoch": 6.74318971788861,
+      "grad_norm": 1.1537504196166992,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 41710
+    },
+    {
+      "epoch": 6.744806402069356,
+      "grad_norm": 1.037880539894104,
+      "learning_rate": 0.0002,
+      "loss": 0.5025,
+      "step": 41720
+    },
+    {
+      "epoch": 6.746423086250101,
+      "grad_norm": 1.0691965818405151,
+      "learning_rate": 0.0002,
+      "loss": 0.482,
+      "step": 41730
+    },
+    {
+      "epoch": 6.748039770430847,
+      "grad_norm": 1.376325011253357,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 41740
+    },
+    {
+      "epoch": 6.749656454611592,
+      "grad_norm": 1.4667129516601562,
+      "learning_rate": 0.0002,
+      "loss": 0.5484,
+      "step": 41750
+    },
+    {
+      "epoch": 6.751273138792337,
+      "grad_norm": 1.1517162322998047,
+      "learning_rate": 0.0002,
+      "loss": 0.5139,
+      "step": 41760
+    },
+    {
+      "epoch": 6.752889822973082,
+      "grad_norm": 1.1454511880874634,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 41770
+    },
+    {
+      "epoch": 6.7545065071538275,
+      "grad_norm": 1.6323128938674927,
+      "learning_rate": 0.0002,
+      "loss": 0.4664,
+      "step": 41780
+    },
+    {
+      "epoch": 6.756123191334573,
+      "grad_norm": 1.0951642990112305,
+      "learning_rate": 0.0002,
+      "loss": 0.5153,
+      "step": 41790
+    },
+    {
+      "epoch": 6.757739875515318,
+      "grad_norm": 1.0766983032226562,
+      "learning_rate": 0.0002,
+      "loss": 0.4998,
+      "step": 41800
+    },
+    {
+      "epoch": 6.759356559696063,
+      "grad_norm": 1.3472381830215454,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 41810
+    },
+    {
+      "epoch": 6.760973243876808,
+      "grad_norm": 1.0248444080352783,
+      "learning_rate": 0.0002,
+      "loss": 0.5172,
+      "step": 41820
+    },
+    {
+      "epoch": 6.762589928057554,
+      "grad_norm": 1.1276055574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 41830
+    },
+    {
+      "epoch": 6.764206612238299,
+      "grad_norm": 1.5398495197296143,
+      "learning_rate": 0.0002,
+      "loss": 0.5044,
+      "step": 41840
+    },
+    {
+      "epoch": 6.765823296419045,
+      "grad_norm": 1.1886497735977173,
+      "learning_rate": 0.0002,
+      "loss": 0.5097,
+      "step": 41850
+    },
+    {
+      "epoch": 6.767439980599789,
+      "grad_norm": 1.027198076248169,
+      "learning_rate": 0.0002,
+      "loss": 0.499,
+      "step": 41860
+    },
+    {
+      "epoch": 6.7690566647805355,
+      "grad_norm": 1.4644980430603027,
+      "learning_rate": 0.0002,
+      "loss": 0.5444,
+      "step": 41870
+    },
+    {
+      "epoch": 6.770673348961281,
+      "grad_norm": 0.9633586406707764,
+      "learning_rate": 0.0002,
+      "loss": 0.5009,
+      "step": 41880
+    },
+    {
+      "epoch": 6.772290033142026,
+      "grad_norm": 1.0895354747772217,
+      "learning_rate": 0.0002,
+      "loss": 0.484,
+      "step": 41890
+    },
+    {
+      "epoch": 6.773906717322771,
+      "grad_norm": 1.1887167692184448,
+      "learning_rate": 0.0002,
+      "loss": 0.5172,
+      "step": 41900
+    },
+    {
+      "epoch": 6.775523401503516,
+      "grad_norm": 1.3699820041656494,
+      "learning_rate": 0.0002,
+      "loss": 0.5399,
+      "step": 41910
+    },
+    {
+      "epoch": 6.777140085684262,
+      "grad_norm": 1.0266352891921997,
+      "learning_rate": 0.0002,
+      "loss": 0.5504,
+      "step": 41920
+    },
+    {
+      "epoch": 6.778756769865007,
+      "grad_norm": 1.0919075012207031,
+      "learning_rate": 0.0002,
+      "loss": 0.5105,
+      "step": 41930
+    },
+    {
+      "epoch": 6.780373454045752,
+      "grad_norm": 0.9839563369750977,
+      "learning_rate": 0.0002,
+      "loss": 0.4842,
+      "step": 41940
+    },
+    {
+      "epoch": 6.781990138226497,
+      "grad_norm": 1.2605451345443726,
+      "learning_rate": 0.0002,
+      "loss": 0.5081,
+      "step": 41950
+    },
+    {
+      "epoch": 6.7836068224072426,
+      "grad_norm": 0.9268672466278076,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 41960
+    },
+    {
+      "epoch": 6.785223506587988,
+      "grad_norm": 1.2002313137054443,
+      "learning_rate": 0.0002,
+      "loss": 0.4916,
+      "step": 41970
+    },
+    {
+      "epoch": 6.786840190768733,
+      "grad_norm": 1.2018438577651978,
+      "learning_rate": 0.0002,
+      "loss": 0.5467,
+      "step": 41980
+    },
+    {
+      "epoch": 6.788456874949478,
+      "grad_norm": 1.17646062374115,
+      "learning_rate": 0.0002,
+      "loss": 0.5491,
+      "step": 41990
+    },
+    {
+      "epoch": 6.790073559130224,
+      "grad_norm": 1.1080009937286377,
+      "learning_rate": 0.0002,
+      "loss": 0.5354,
+      "step": 42000
+    },
+    {
+      "epoch": 6.791690243310969,
+      "grad_norm": 1.1606498956680298,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 42010
+    },
+    {
+      "epoch": 6.793306927491715,
+      "grad_norm": 1.2484819889068604,
+      "learning_rate": 0.0002,
+      "loss": 0.4931,
+      "step": 42020
+    },
+    {
+      "epoch": 6.79492361167246,
+      "grad_norm": 1.1363215446472168,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 42030
+    },
+    {
+      "epoch": 6.796540295853205,
+      "grad_norm": 1.4469727277755737,
+      "learning_rate": 0.0002,
+      "loss": 0.5343,
+      "step": 42040
+    },
+    {
+      "epoch": 6.7981569800339505,
+      "grad_norm": 1.0617138147354126,
+      "learning_rate": 0.0002,
+      "loss": 0.5146,
+      "step": 42050
+    },
+    {
+      "epoch": 6.799773664214696,
+      "grad_norm": 1.1459330320358276,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 42060
+    },
+    {
+      "epoch": 6.801390348395441,
+      "grad_norm": 1.2095019817352295,
+      "learning_rate": 0.0002,
+      "loss": 0.5116,
+      "step": 42070
+    },
+    {
+      "epoch": 6.803007032576186,
+      "grad_norm": 1.3200831413269043,
+      "learning_rate": 0.0002,
+      "loss": 0.545,
+      "step": 42080
+    },
+    {
+      "epoch": 6.8046237167569315,
+      "grad_norm": 1.1633318662643433,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 42090
+    },
+    {
+      "epoch": 6.806240400937677,
+      "grad_norm": 0.8986614942550659,
+      "learning_rate": 0.0002,
+      "loss": 0.4938,
+      "step": 42100
+    },
+    {
+      "epoch": 6.807857085118422,
+      "grad_norm": 1.3705275058746338,
+      "learning_rate": 0.0002,
+      "loss": 0.559,
+      "step": 42110
+    },
+    {
+      "epoch": 6.809473769299167,
+      "grad_norm": 1.2418090105056763,
+      "learning_rate": 0.0002,
+      "loss": 0.5022,
+      "step": 42120
+    },
+    {
+      "epoch": 6.811090453479912,
+      "grad_norm": 1.0818954706192017,
+      "learning_rate": 0.0002,
+      "loss": 0.5014,
+      "step": 42130
+    },
+    {
+      "epoch": 6.812707137660658,
+      "grad_norm": 0.9293872117996216,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 42140
+    },
+    {
+      "epoch": 6.814323821841404,
+      "grad_norm": 0.9791894555091858,
+      "learning_rate": 0.0002,
+      "loss": 0.5009,
+      "step": 42150
+    },
+    {
+      "epoch": 6.815940506022149,
+      "grad_norm": 1.1956568956375122,
+      "learning_rate": 0.0002,
+      "loss": 0.5142,
+      "step": 42160
+    },
+    {
+      "epoch": 6.817557190202894,
+      "grad_norm": 0.9643568992614746,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 42170
+    },
+    {
+      "epoch": 6.819173874383639,
+      "grad_norm": 1.2499792575836182,
+      "learning_rate": 0.0002,
+      "loss": 0.5121,
+      "step": 42180
+    },
+    {
+      "epoch": 6.820790558564385,
+      "grad_norm": 1.1779413223266602,
+      "learning_rate": 0.0002,
+      "loss": 0.4942,
+      "step": 42190
+    },
+    {
+      "epoch": 6.82240724274513,
+      "grad_norm": 1.0570595264434814,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 42200
+    },
+    {
+      "epoch": 6.824023926925875,
+      "grad_norm": 1.1393938064575195,
+      "learning_rate": 0.0002,
+      "loss": 0.4997,
+      "step": 42210
+    },
+    {
+      "epoch": 6.82564061110662,
+      "grad_norm": 1.152463436126709,
+      "learning_rate": 0.0002,
+      "loss": 0.4842,
+      "step": 42220
+    },
+    {
+      "epoch": 6.827257295287366,
+      "grad_norm": 1.3353025913238525,
+      "learning_rate": 0.0002,
+      "loss": 0.5234,
+      "step": 42230
+    },
+    {
+      "epoch": 6.828873979468111,
+      "grad_norm": 1.1719051599502563,
+      "learning_rate": 0.0002,
+      "loss": 0.539,
+      "step": 42240
+    },
+    {
+      "epoch": 6.830490663648856,
+      "grad_norm": 1.262141227722168,
+      "learning_rate": 0.0002,
+      "loss": 0.5139,
+      "step": 42250
+    },
+    {
+      "epoch": 6.832107347829601,
+      "grad_norm": 1.240899920463562,
+      "learning_rate": 0.0002,
+      "loss": 0.5021,
+      "step": 42260
+    },
+    {
+      "epoch": 6.8337240320103465,
+      "grad_norm": 1.0505269765853882,
+      "learning_rate": 0.0002,
+      "loss": 0.4961,
+      "step": 42270
+    },
+    {
+      "epoch": 6.835340716191092,
+      "grad_norm": 1.1556071043014526,
+      "learning_rate": 0.0002,
+      "loss": 0.4932,
+      "step": 42280
+    },
+    {
+      "epoch": 6.836957400371837,
+      "grad_norm": 1.1427719593048096,
+      "learning_rate": 0.0002,
+      "loss": 0.5461,
+      "step": 42290
+    },
+    {
+      "epoch": 6.838574084552583,
+      "grad_norm": 1.1540080308914185,
+      "learning_rate": 0.0002,
+      "loss": 0.5199,
+      "step": 42300
+    },
+    {
+      "epoch": 6.840190768733328,
+      "grad_norm": 1.0521200895309448,
+      "learning_rate": 0.0002,
+      "loss": 0.5269,
+      "step": 42310
+    },
+    {
+      "epoch": 6.8418074529140736,
+      "grad_norm": 1.0205531120300293,
+      "learning_rate": 0.0002,
+      "loss": 0.541,
+      "step": 42320
+    },
+    {
+      "epoch": 6.843424137094819,
+      "grad_norm": 1.0010193586349487,
+      "learning_rate": 0.0002,
+      "loss": 0.5225,
+      "step": 42330
+    },
+    {
+      "epoch": 6.845040821275564,
+      "grad_norm": 1.2138770818710327,
+      "learning_rate": 0.0002,
+      "loss": 0.5101,
+      "step": 42340
+    },
+    {
+      "epoch": 6.846657505456309,
+      "grad_norm": 1.3028651475906372,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 42350
+    },
+    {
+      "epoch": 6.8482741896370545,
+      "grad_norm": 1.0326353311538696,
+      "learning_rate": 0.0002,
+      "loss": 0.4894,
+      "step": 42360
+    },
+    {
+      "epoch": 6.8498908738178,
+      "grad_norm": 1.036085605621338,
+      "learning_rate": 0.0002,
+      "loss": 0.5285,
+      "step": 42370
+    },
+    {
+      "epoch": 6.851507557998545,
+      "grad_norm": 1.0575472116470337,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 42380
+    },
+    {
+      "epoch": 6.85312424217929,
+      "grad_norm": 1.1749629974365234,
+      "learning_rate": 0.0002,
+      "loss": 0.4997,
+      "step": 42390
+    },
+    {
+      "epoch": 6.854740926360035,
+      "grad_norm": 1.1747760772705078,
+      "learning_rate": 0.0002,
+      "loss": 0.4961,
+      "step": 42400
+    },
+    {
+      "epoch": 6.856357610540781,
+      "grad_norm": 1.1877071857452393,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 42410
+    },
+    {
+      "epoch": 6.857974294721526,
+      "grad_norm": 1.1209983825683594,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 42420
+    },
+    {
+      "epoch": 6.859590978902271,
+      "grad_norm": 1.2918205261230469,
+      "learning_rate": 0.0002,
+      "loss": 0.4939,
+      "step": 42430
+    },
+    {
+      "epoch": 6.861207663083016,
+      "grad_norm": 1.2443464994430542,
+      "learning_rate": 0.0002,
+      "loss": 0.5012,
+      "step": 42440
+    },
+    {
+      "epoch": 6.8628243472637624,
+      "grad_norm": 0.9336795210838318,
+      "learning_rate": 0.0002,
+      "loss": 0.5226,
+      "step": 42450
+    },
+    {
+      "epoch": 6.864441031444508,
+      "grad_norm": 1.2183542251586914,
+      "learning_rate": 0.0002,
+      "loss": 0.5108,
+      "step": 42460
+    },
+    {
+      "epoch": 6.866057715625253,
+      "grad_norm": 1.0071234703063965,
+      "learning_rate": 0.0002,
+      "loss": 0.5245,
+      "step": 42470
+    },
+    {
+      "epoch": 6.867674399805998,
+      "grad_norm": 1.2914012670516968,
+      "learning_rate": 0.0002,
+      "loss": 0.4753,
+      "step": 42480
+    },
+    {
+      "epoch": 6.869291083986743,
+      "grad_norm": 1.1050426959991455,
+      "learning_rate": 0.0002,
+      "loss": 0.4865,
+      "step": 42490
+    },
+    {
+      "epoch": 6.870907768167489,
+      "grad_norm": 1.1163811683654785,
+      "learning_rate": 0.0002,
+      "loss": 0.5243,
+      "step": 42500
+    },
+    {
+      "epoch": 6.872524452348234,
+      "grad_norm": 1.1575818061828613,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 42510
+    },
+    {
+      "epoch": 6.874141136528979,
+      "grad_norm": 1.11167311668396,
+      "learning_rate": 0.0002,
+      "loss": 0.5353,
+      "step": 42520
+    },
+    {
+      "epoch": 6.875757820709724,
+      "grad_norm": 1.0379102230072021,
+      "learning_rate": 0.0002,
+      "loss": 0.5141,
+      "step": 42530
+    },
+    {
+      "epoch": 6.8773745048904695,
+      "grad_norm": 1.2617160081863403,
+      "learning_rate": 0.0002,
+      "loss": 0.5355,
+      "step": 42540
+    },
+    {
+      "epoch": 6.878991189071215,
+      "grad_norm": 1.1749719381332397,
+      "learning_rate": 0.0002,
+      "loss": 0.4785,
+      "step": 42550
+    },
+    {
+      "epoch": 6.88060787325196,
+      "grad_norm": 1.2284821271896362,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 42560
+    },
+    {
+      "epoch": 6.882224557432705,
+      "grad_norm": 1.1917030811309814,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 42570
+    },
+    {
+      "epoch": 6.8838412416134505,
+      "grad_norm": 1.1943914890289307,
+      "learning_rate": 0.0002,
+      "loss": 0.5176,
+      "step": 42580
+    },
+    {
+      "epoch": 6.885457925794196,
+      "grad_norm": 1.2641394138336182,
+      "learning_rate": 0.0002,
+      "loss": 0.5072,
+      "step": 42590
+    },
+    {
+      "epoch": 6.887074609974942,
+      "grad_norm": 1.1280436515808105,
+      "learning_rate": 0.0002,
+      "loss": 0.5004,
+      "step": 42600
+    },
+    {
+      "epoch": 6.888691294155687,
+      "grad_norm": 0.9865449070930481,
+      "learning_rate": 0.0002,
+      "loss": 0.5328,
+      "step": 42610
+    },
+    {
+      "epoch": 6.890307978336432,
+      "grad_norm": 0.994987428188324,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 42620
+    },
+    {
+      "epoch": 6.8919246625171775,
+      "grad_norm": 0.9900388717651367,
+      "learning_rate": 0.0002,
+      "loss": 0.4805,
+      "step": 42630
+    },
+    {
+      "epoch": 6.893541346697923,
+      "grad_norm": 1.2992421388626099,
+      "learning_rate": 0.0002,
+      "loss": 0.5467,
+      "step": 42640
+    },
+    {
+      "epoch": 6.895158030878668,
+      "grad_norm": 1.0152487754821777,
+      "learning_rate": 0.0002,
+      "loss": 0.5017,
+      "step": 42650
+    },
+    {
+      "epoch": 6.896774715059413,
+      "grad_norm": 1.199453353881836,
+      "learning_rate": 0.0002,
+      "loss": 0.5043,
+      "step": 42660
+    },
+    {
+      "epoch": 6.898391399240158,
+      "grad_norm": 1.100630521774292,
+      "learning_rate": 0.0002,
+      "loss": 0.5106,
+      "step": 42670
+    },
+    {
+      "epoch": 6.900008083420904,
+      "grad_norm": 1.0489764213562012,
+      "learning_rate": 0.0002,
+      "loss": 0.503,
+      "step": 42680
+    },
+    {
+      "epoch": 6.901624767601649,
+      "grad_norm": 1.101407527923584,
+      "learning_rate": 0.0002,
+      "loss": 0.4634,
+      "step": 42690
+    },
+    {
+      "epoch": 6.903241451782394,
+      "grad_norm": 1.3130593299865723,
+      "learning_rate": 0.0002,
+      "loss": 0.5361,
+      "step": 42700
+    },
+    {
+      "epoch": 6.904858135963139,
+      "grad_norm": 0.9906072616577148,
+      "learning_rate": 0.0002,
+      "loss": 0.5119,
+      "step": 42710
+    },
+    {
+      "epoch": 6.906474820143885,
+      "grad_norm": 1.094502329826355,
+      "learning_rate": 0.0002,
+      "loss": 0.5146,
+      "step": 42720
+    },
+    {
+      "epoch": 6.90809150432463,
+      "grad_norm": 1.1025426387786865,
+      "learning_rate": 0.0002,
+      "loss": 0.5165,
+      "step": 42730
+    },
+    {
+      "epoch": 6.909708188505375,
+      "grad_norm": 1.0644042491912842,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 42740
+    },
+    {
+      "epoch": 6.911324872686121,
+      "grad_norm": 1.0709129571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.5024,
+      "step": 42750
+    },
+    {
+      "epoch": 6.912941556866866,
+      "grad_norm": 1.2445871829986572,
+      "learning_rate": 0.0002,
+      "loss": 0.5093,
+      "step": 42760
+    },
+    {
+      "epoch": 6.914558241047612,
+      "grad_norm": 1.020058035850525,
+      "learning_rate": 0.0002,
+      "loss": 0.5305,
+      "step": 42770
+    },
+    {
+      "epoch": 6.916174925228357,
+      "grad_norm": 0.9795091152191162,
+      "learning_rate": 0.0002,
+      "loss": 0.5382,
+      "step": 42780
+    },
+    {
+      "epoch": 6.917791609409102,
+      "grad_norm": 0.9369977116584778,
+      "learning_rate": 0.0002,
+      "loss": 0.5429,
+      "step": 42790
+    },
+    {
+      "epoch": 6.919408293589847,
+      "grad_norm": 1.0741904973983765,
+      "learning_rate": 0.0002,
+      "loss": 0.5444,
+      "step": 42800
+    },
+    {
+      "epoch": 6.921024977770593,
+      "grad_norm": 1.0702799558639526,
+      "learning_rate": 0.0002,
+      "loss": 0.5402,
+      "step": 42810
+    },
+    {
+      "epoch": 6.922641661951338,
+      "grad_norm": 1.0383983850479126,
+      "learning_rate": 0.0002,
+      "loss": 0.5291,
+      "step": 42820
+    },
+    {
+      "epoch": 6.924258346132083,
+      "grad_norm": 1.0761083364486694,
+      "learning_rate": 0.0002,
+      "loss": 0.5106,
+      "step": 42830
+    },
+    {
+      "epoch": 6.925875030312828,
+      "grad_norm": 1.2332350015640259,
+      "learning_rate": 0.0002,
+      "loss": 0.5726,
+      "step": 42840
+    },
+    {
+      "epoch": 6.9274917144935735,
+      "grad_norm": 1.3184348344802856,
+      "learning_rate": 0.0002,
+      "loss": 0.4996,
+      "step": 42850
+    },
+    {
+      "epoch": 6.929108398674319,
+      "grad_norm": 1.0586378574371338,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 42860
+    },
+    {
+      "epoch": 6.930725082855064,
+      "grad_norm": 1.2294201850891113,
+      "learning_rate": 0.0002,
+      "loss": 0.511,
+      "step": 42870
+    },
+    {
+      "epoch": 6.932341767035809,
+      "grad_norm": 1.3097991943359375,
+      "learning_rate": 0.0002,
+      "loss": 0.54,
+      "step": 42880
+    },
+    {
+      "epoch": 6.933958451216554,
+      "grad_norm": 0.9006873965263367,
+      "learning_rate": 0.0002,
+      "loss": 0.5228,
+      "step": 42890
+    },
+    {
+      "epoch": 6.9355751353973005,
+      "grad_norm": 1.265931248664856,
+      "learning_rate": 0.0002,
+      "loss": 0.4617,
+      "step": 42900
+    },
+    {
+      "epoch": 6.937191819578046,
+      "grad_norm": 1.1013522148132324,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 42910
+    },
+    {
+      "epoch": 6.938808503758791,
+      "grad_norm": 0.9910131692886353,
+      "learning_rate": 0.0002,
+      "loss": 0.5334,
+      "step": 42920
+    },
+    {
+      "epoch": 6.940425187939536,
+      "grad_norm": 1.102683424949646,
+      "learning_rate": 0.0002,
+      "loss": 0.5211,
+      "step": 42930
+    },
+    {
+      "epoch": 6.9420418721202815,
+      "grad_norm": 1.232961893081665,
+      "learning_rate": 0.0002,
+      "loss": 0.5588,
+      "step": 42940
+    },
+    {
+      "epoch": 6.943658556301027,
+      "grad_norm": 1.1714650392532349,
+      "learning_rate": 0.0002,
+      "loss": 0.5357,
+      "step": 42950
+    },
+    {
+      "epoch": 6.945275240481772,
+      "grad_norm": 1.1684318780899048,
+      "learning_rate": 0.0002,
+      "loss": 0.5232,
+      "step": 42960
+    },
+    {
+      "epoch": 6.946891924662517,
+      "grad_norm": 1.2074716091156006,
+      "learning_rate": 0.0002,
+      "loss": 0.5035,
+      "step": 42970
+    },
+    {
+      "epoch": 6.948508608843262,
+      "grad_norm": 1.2061275243759155,
+      "learning_rate": 0.0002,
+      "loss": 0.5111,
+      "step": 42980
+    },
+    {
+      "epoch": 6.950125293024008,
+      "grad_norm": 1.1216989755630493,
+      "learning_rate": 0.0002,
+      "loss": 0.5066,
+      "step": 42990
+    },
+    {
+      "epoch": 6.951741977204753,
+      "grad_norm": 1.304117202758789,
+      "learning_rate": 0.0002,
+      "loss": 0.4948,
+      "step": 43000
+    },
+    {
+      "epoch": 6.953358661385498,
+      "grad_norm": 1.2377972602844238,
+      "learning_rate": 0.0002,
+      "loss": 0.5684,
+      "step": 43010
+    },
+    {
+      "epoch": 6.954975345566243,
+      "grad_norm": 1.2332178354263306,
+      "learning_rate": 0.0002,
+      "loss": 0.4792,
+      "step": 43020
+    },
+    {
+      "epoch": 6.956592029746989,
+      "grad_norm": 1.1919599771499634,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 43030
+    },
+    {
+      "epoch": 6.958208713927734,
+      "grad_norm": 1.272700548171997,
+      "learning_rate": 0.0002,
+      "loss": 0.5352,
+      "step": 43040
+    },
+    {
+      "epoch": 6.95982539810848,
+      "grad_norm": 1.4377546310424805,
+      "learning_rate": 0.0002,
+      "loss": 0.5328,
+      "step": 43050
+    },
+    {
+      "epoch": 6.961442082289225,
+      "grad_norm": 1.2070353031158447,
+      "learning_rate": 0.0002,
+      "loss": 0.4894,
+      "step": 43060
+    },
+    {
+      "epoch": 6.96305876646997,
+      "grad_norm": 1.090205430984497,
+      "learning_rate": 0.0002,
+      "loss": 0.525,
+      "step": 43070
+    },
+    {
+      "epoch": 6.964675450650716,
+      "grad_norm": 1.1832911968231201,
+      "learning_rate": 0.0002,
+      "loss": 0.5255,
+      "step": 43080
+    },
+    {
+      "epoch": 6.966292134831461,
+      "grad_norm": 1.2921082973480225,
+      "learning_rate": 0.0002,
+      "loss": 0.5497,
+      "step": 43090
+    },
+    {
+      "epoch": 6.967908819012206,
+      "grad_norm": 1.4303096532821655,
+      "learning_rate": 0.0002,
+      "loss": 0.5527,
+      "step": 43100
+    },
+    {
+      "epoch": 6.969525503192951,
+      "grad_norm": 1.0788004398345947,
+      "learning_rate": 0.0002,
+      "loss": 0.4807,
+      "step": 43110
+    },
+    {
+      "epoch": 6.9711421873736965,
+      "grad_norm": 1.2192047834396362,
+      "learning_rate": 0.0002,
+      "loss": 0.5006,
+      "step": 43120
+    },
+    {
+      "epoch": 6.972758871554442,
+      "grad_norm": 1.0735143423080444,
+      "learning_rate": 0.0002,
+      "loss": 0.4714,
+      "step": 43130
+    },
+    {
+      "epoch": 6.974375555735187,
+      "grad_norm": 1.0317153930664062,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 43140
+    },
+    {
+      "epoch": 6.975992239915932,
+      "grad_norm": 1.0926798582077026,
+      "learning_rate": 0.0002,
+      "loss": 0.5154,
+      "step": 43150
+    },
+    {
+      "epoch": 6.977608924096677,
+      "grad_norm": 1.1660500764846802,
+      "learning_rate": 0.0002,
+      "loss": 0.4976,
+      "step": 43160
+    },
+    {
+      "epoch": 6.979225608277423,
+      "grad_norm": 1.3945232629776,
+      "learning_rate": 0.0002,
+      "loss": 0.5456,
+      "step": 43170
+    },
+    {
+      "epoch": 6.980842292458169,
+      "grad_norm": 1.2684587240219116,
+      "learning_rate": 0.0002,
+      "loss": 0.4979,
+      "step": 43180
+    },
+    {
+      "epoch": 6.982458976638913,
+      "grad_norm": 1.1574004888534546,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 43190
+    },
+    {
+      "epoch": 6.984075660819659,
+      "grad_norm": 1.2534198760986328,
+      "learning_rate": 0.0002,
+      "loss": 0.5629,
+      "step": 43200
+    },
+    {
+      "epoch": 6.9856923450004045,
+      "grad_norm": 1.135245442390442,
+      "learning_rate": 0.0002,
+      "loss": 0.5191,
+      "step": 43210
+    },
+    {
+      "epoch": 6.98730902918115,
+      "grad_norm": 1.3824104070663452,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 43220
+    },
+    {
+      "epoch": 6.988925713361895,
+      "grad_norm": 1.2128452062606812,
+      "learning_rate": 0.0002,
+      "loss": 0.5294,
+      "step": 43230
+    },
+    {
+      "epoch": 6.99054239754264,
+      "grad_norm": 1.0795245170593262,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 43240
+    },
+    {
+      "epoch": 6.992159081723385,
+      "grad_norm": 1.337353229522705,
+      "learning_rate": 0.0002,
+      "loss": 0.4889,
+      "step": 43250
+    },
+    {
+      "epoch": 6.993775765904131,
+      "grad_norm": 1.1731765270233154,
+      "learning_rate": 0.0002,
+      "loss": 0.4749,
+      "step": 43260
+    },
+    {
+      "epoch": 6.995392450084876,
+      "grad_norm": 1.0203192234039307,
+      "learning_rate": 0.0002,
+      "loss": 0.4897,
+      "step": 43270
+    },
+    {
+      "epoch": 6.997009134265621,
+      "grad_norm": 0.9261201620101929,
+      "learning_rate": 0.0002,
+      "loss": 0.5324,
+      "step": 43280
+    },
+    {
+      "epoch": 6.998625818446366,
+      "grad_norm": 1.107865810394287,
+      "learning_rate": 0.0002,
+      "loss": 0.5227,
+      "step": 43290
+    },
+    {
+      "epoch": 6.9999191657909625,
+      "eval_loss": 1.2679380178451538,
+      "eval_runtime": 122.202,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 43298
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 49480,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.0037576307318456e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/training_args.bin b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..155b12fa9acbc6e71dba75c92bfa79e152397ebf
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-43298/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28694d5564a2b5c7d6881d4ba2af103356aa22489d2c22768ebbe47283c0f4a1
+size 5560
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/README.md b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..835e31ab8469ee39ddc8b2b6b2143a8c66dad510
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/README.md
@@ -0,0 +1,202 @@
+---
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.13.1
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/adapter_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..406c5a08dc4a2a33b52c62a482f98c217c417215
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/adapter_config.json
@@ -0,0 +1,29 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/adapter_model.safetensors b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..f828a74fd080647f5b76b649265f019c1dee6822
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ae514d1496760bcd747d5f35daaea1ebe7589d9dd859a867aab3f6c7ee63ffb
+size 109069176
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/optimizer.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a3afc3bb8dbb51f750dab1dac300b22f2bc060ea
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ffad8d9c012795bb87f2c48fc7e292b984d249e84ceb16f5b4353e87aaf6e92
+size 55532666
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/rng_state.pth b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e6f2c4ca9aa9b93bb0507887bfdf97dc98716951
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ffb5fee38f909e20f0db9c8d5d103ccd6be5344dadb2c10329d2136c9aa207c
+size 14244
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/scheduler.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9f94c302b2027faa9c2dc68a9cd9b3395df05d67
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67233313ac4b876ce7a58fbb5923f3e93ad0293bc11cc82e5d9a6d5c6521e23e
+size 1064
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/special_tokens_map.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/special_tokens_map.json
@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/tokenizer.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..86a33946b0c77216d2cce91bb28c8fada4a5e80b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/tokenizer_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..061e40d9db3253624f86e8e364c15ef546527c9d
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/tokenizer_config.json
@@ -0,0 +1,2063 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/trainer_state.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..148526696eab1bc2c6d827bf826dde47f5f93dd3
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/trainer_state.json
@@ -0,0 +1,34733 @@
+{
+  "best_metric": 1.0871200561523438,
+  "best_model_checkpoint": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185",
+  "epoch": 7.999353326327702,
+  "eval_steps": 10,
+  "global_step": 49480,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016166841807452913,
+      "grad_norm": 0.9894065856933594,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032333683614905826,
+      "grad_norm": 1.7810699939727783,
+      "learning_rate": 0.0002,
+      "loss": 1.1528,
+      "step": 20
+    },
+    {
+      "epoch": 0.004850052542235874,
+      "grad_norm": 0.5969577431678772,
+      "learning_rate": 0.0002,
+      "loss": 0.9767,
+      "step": 30
+    },
+    {
+      "epoch": 0.006466736722981165,
+      "grad_norm": 0.6354120969772339,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 40
+    },
+    {
+      "epoch": 0.008083420903726457,
+      "grad_norm": 0.5604607462882996,
+      "learning_rate": 0.0002,
+      "loss": 0.8643,
+      "step": 50
+    },
+    {
+      "epoch": 0.009700105084471748,
+      "grad_norm": 0.4676193594932556,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 60
+    },
+    {
+      "epoch": 0.01131678926521704,
+      "grad_norm": 0.6099211573600769,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 70
+    },
+    {
+      "epoch": 0.01293347344596233,
+      "grad_norm": 0.48639994859695435,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 80
+    },
+    {
+      "epoch": 0.014550157626707623,
+      "grad_norm": 0.4904264509677887,
+      "learning_rate": 0.0002,
+      "loss": 0.8704,
+      "step": 90
+    },
+    {
+      "epoch": 0.016166841807452915,
+      "grad_norm": 2.8334362506866455,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 100
+    },
+    {
+      "epoch": 0.017783525988198205,
+      "grad_norm": 0.43221670389175415,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 110
+    },
+    {
+      "epoch": 0.019400210168943496,
+      "grad_norm": 0.42244166135787964,
+      "learning_rate": 0.0002,
+      "loss": 0.8412,
+      "step": 120
+    },
+    {
+      "epoch": 0.02101689434968879,
+      "grad_norm": 0.45363298058509827,
+      "learning_rate": 0.0002,
+      "loss": 0.8467,
+      "step": 130
+    },
+    {
+      "epoch": 0.02263357853043408,
+      "grad_norm": 0.44816508889198303,
+      "learning_rate": 0.0002,
+      "loss": 0.8641,
+      "step": 140
+    },
+    {
+      "epoch": 0.02425026271117937,
+      "grad_norm": 0.43308213353157043,
+      "learning_rate": 0.0002,
+      "loss": 0.8496,
+      "step": 150
+    },
+    {
+      "epoch": 0.02586694689192466,
+      "grad_norm": 0.4084763526916504,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 160
+    },
+    {
+      "epoch": 0.027483631072669955,
+      "grad_norm": 0.5363703966140747,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 170
+    },
+    {
+      "epoch": 0.029100315253415245,
+      "grad_norm": 0.4619699716567993,
+      "learning_rate": 0.0002,
+      "loss": 0.8558,
+      "step": 180
+    },
+    {
+      "epoch": 0.030716999434160536,
+      "grad_norm": 0.49069908261299133,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 190
+    },
+    {
+      "epoch": 0.03233368361490583,
+      "grad_norm": 0.4645835757255554,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 200
+    },
+    {
+      "epoch": 0.03395036779565112,
+      "grad_norm": 1.2411243915557861,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 210
+    },
+    {
+      "epoch": 0.03556705197639641,
+      "grad_norm": 0.5211851596832275,
+      "learning_rate": 0.0002,
+      "loss": 0.8245,
+      "step": 220
+    },
+    {
+      "epoch": 0.037183736157141704,
+      "grad_norm": 0.5253691673278809,
+      "learning_rate": 0.0002,
+      "loss": 0.8194,
+      "step": 230
+    },
+    {
+      "epoch": 0.03880042033788699,
+      "grad_norm": 0.4567478895187378,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 240
+    },
+    {
+      "epoch": 0.040417104518632285,
+      "grad_norm": 0.5472128391265869,
+      "learning_rate": 0.0002,
+      "loss": 0.838,
+      "step": 250
+    },
+    {
+      "epoch": 0.04203378869937758,
+      "grad_norm": 0.42978546023368835,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 260
+    },
+    {
+      "epoch": 0.043650472880122866,
+      "grad_norm": 0.601734459400177,
+      "learning_rate": 0.0002,
+      "loss": 0.8334,
+      "step": 270
+    },
+    {
+      "epoch": 0.04526715706086816,
+      "grad_norm": 0.4286513328552246,
+      "learning_rate": 0.0002,
+      "loss": 0.815,
+      "step": 280
+    },
+    {
+      "epoch": 0.046883841241613454,
+      "grad_norm": 0.5230861902236938,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 290
+    },
+    {
+      "epoch": 0.04850052542235874,
+      "grad_norm": 0.6504611968994141,
+      "learning_rate": 0.0002,
+      "loss": 0.8636,
+      "step": 300
+    },
+    {
+      "epoch": 0.050117209603104035,
+      "grad_norm": 0.43485215306282043,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 310
+    },
+    {
+      "epoch": 0.05173389378384932,
+      "grad_norm": 0.4717007875442505,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 320
+    },
+    {
+      "epoch": 0.053350577964594616,
+      "grad_norm": 0.4059787690639496,
+      "learning_rate": 0.0002,
+      "loss": 0.8469,
+      "step": 330
+    },
+    {
+      "epoch": 0.05496726214533991,
+      "grad_norm": 0.4366913437843323,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 340
+    },
+    {
+      "epoch": 0.0565839463260852,
+      "grad_norm": 0.4233848452568054,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 350
+    },
+    {
+      "epoch": 0.05820063050683049,
+      "grad_norm": 0.4209108352661133,
+      "learning_rate": 0.0002,
+      "loss": 0.8456,
+      "step": 360
+    },
+    {
+      "epoch": 0.059817314687575784,
+      "grad_norm": 0.41637396812438965,
+      "learning_rate": 0.0002,
+      "loss": 0.816,
+      "step": 370
+    },
+    {
+      "epoch": 0.06143399886832107,
+      "grad_norm": 0.46235376596450806,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 380
+    },
+    {
+      "epoch": 0.06305068304906636,
+      "grad_norm": 0.4013484716415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 390
+    },
+    {
+      "epoch": 0.06466736722981166,
+      "grad_norm": 0.47443896532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 400
+    },
+    {
+      "epoch": 0.06628405141055695,
+      "grad_norm": 0.3942156434059143,
+      "learning_rate": 0.0002,
+      "loss": 0.8666,
+      "step": 410
+    },
+    {
+      "epoch": 0.06790073559130223,
+      "grad_norm": 0.4965320825576782,
+      "learning_rate": 0.0002,
+      "loss": 0.8402,
+      "step": 420
+    },
+    {
+      "epoch": 0.06951741977204753,
+      "grad_norm": 0.4304835796356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8317,
+      "step": 430
+    },
+    {
+      "epoch": 0.07113410395279282,
+      "grad_norm": 0.511726975440979,
+      "learning_rate": 0.0002,
+      "loss": 0.8528,
+      "step": 440
+    },
+    {
+      "epoch": 0.07275078813353811,
+      "grad_norm": 0.4040689170360565,
+      "learning_rate": 0.0002,
+      "loss": 0.8675,
+      "step": 450
+    },
+    {
+      "epoch": 0.07436747231428341,
+      "grad_norm": 0.5402171015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 460
+    },
+    {
+      "epoch": 0.0759841564950287,
+      "grad_norm": 0.4174517095088959,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 470
+    },
+    {
+      "epoch": 0.07760084067577398,
+      "grad_norm": 0.4306182265281677,
+      "learning_rate": 0.0002,
+      "loss": 0.7605,
+      "step": 480
+    },
+    {
+      "epoch": 0.07921752485651928,
+      "grad_norm": 0.535210132598877,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 490
+    },
+    {
+      "epoch": 0.08083420903726457,
+      "grad_norm": 0.5339109897613525,
+      "learning_rate": 0.0002,
+      "loss": 0.7825,
+      "step": 500
+    },
+    {
+      "epoch": 0.08245089321800986,
+      "grad_norm": 0.45754891633987427,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 510
+    },
+    {
+      "epoch": 0.08406757739875516,
+      "grad_norm": 0.43820783495903015,
+      "learning_rate": 0.0002,
+      "loss": 0.8144,
+      "step": 520
+    },
+    {
+      "epoch": 0.08568426157950045,
+      "grad_norm": 0.4434749186038971,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 530
+    },
+    {
+      "epoch": 0.08730094576024573,
+      "grad_norm": 0.43111467361450195,
+      "learning_rate": 0.0002,
+      "loss": 0.7857,
+      "step": 540
+    },
+    {
+      "epoch": 0.08891762994099103,
+      "grad_norm": 0.4378940165042877,
+      "learning_rate": 0.0002,
+      "loss": 0.8418,
+      "step": 550
+    },
+    {
+      "epoch": 0.09053431412173632,
+      "grad_norm": 0.4772215187549591,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 560
+    },
+    {
+      "epoch": 0.09215099830248161,
+      "grad_norm": 0.6837629079818726,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 570
+    },
+    {
+      "epoch": 0.09376768248322691,
+      "grad_norm": 0.42241212725639343,
+      "learning_rate": 0.0002,
+      "loss": 0.8607,
+      "step": 580
+    },
+    {
+      "epoch": 0.0953843666639722,
+      "grad_norm": 0.5165936350822449,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 590
+    },
+    {
+      "epoch": 0.09700105084471748,
+      "grad_norm": 0.48737478256225586,
+      "learning_rate": 0.0002,
+      "loss": 0.8664,
+      "step": 600
+    },
+    {
+      "epoch": 0.09861773502546278,
+      "grad_norm": 0.47419852018356323,
+      "learning_rate": 0.0002,
+      "loss": 0.8806,
+      "step": 610
+    },
+    {
+      "epoch": 0.10023441920620807,
+      "grad_norm": 0.4975486099720001,
+      "learning_rate": 0.0002,
+      "loss": 0.8254,
+      "step": 620
+    },
+    {
+      "epoch": 0.10185110338695336,
+      "grad_norm": 0.49123844504356384,
+      "learning_rate": 0.0002,
+      "loss": 0.8548,
+      "step": 630
+    },
+    {
+      "epoch": 0.10346778756769864,
+      "grad_norm": 0.6288952827453613,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 640
+    },
+    {
+      "epoch": 0.10508447174844394,
+      "grad_norm": 0.4277345836162567,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 650
+    },
+    {
+      "epoch": 0.10670115592918923,
+      "grad_norm": 0.4021061956882477,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 660
+    },
+    {
+      "epoch": 0.10831784010993452,
+      "grad_norm": 0.3492237329483032,
+      "learning_rate": 0.0002,
+      "loss": 0.87,
+      "step": 670
+    },
+    {
+      "epoch": 0.10993452429067982,
+      "grad_norm": 0.4341012239456177,
+      "learning_rate": 0.0002,
+      "loss": 0.8698,
+      "step": 680
+    },
+    {
+      "epoch": 0.1115512084714251,
+      "grad_norm": 0.7296304106712341,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 690
+    },
+    {
+      "epoch": 0.1131678926521704,
+      "grad_norm": 0.397494912147522,
+      "learning_rate": 0.0002,
+      "loss": 0.8433,
+      "step": 700
+    },
+    {
+      "epoch": 0.1147845768329157,
+      "grad_norm": 0.396431028842926,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 710
+    },
+    {
+      "epoch": 0.11640126101366098,
+      "grad_norm": 0.48842838406562805,
+      "learning_rate": 0.0002,
+      "loss": 0.8379,
+      "step": 720
+    },
+    {
+      "epoch": 0.11801794519440627,
+      "grad_norm": 0.46322616934776306,
+      "learning_rate": 0.0002,
+      "loss": 0.8238,
+      "step": 730
+    },
+    {
+      "epoch": 0.11963462937515157,
+      "grad_norm": 0.47990912199020386,
+      "learning_rate": 0.0002,
+      "loss": 0.8041,
+      "step": 740
+    },
+    {
+      "epoch": 0.12125131355589686,
+      "grad_norm": 0.4997142255306244,
+      "learning_rate": 0.0002,
+      "loss": 0.82,
+      "step": 750
+    },
+    {
+      "epoch": 0.12286799773664214,
+      "grad_norm": 0.4040526747703552,
+      "learning_rate": 0.0002,
+      "loss": 0.7702,
+      "step": 760
+    },
+    {
+      "epoch": 0.12448468191738744,
+      "grad_norm": 0.453095942735672,
+      "learning_rate": 0.0002,
+      "loss": 0.863,
+      "step": 770
+    },
+    {
+      "epoch": 0.12610136609813272,
+      "grad_norm": 0.4636971950531006,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 780
+    },
+    {
+      "epoch": 0.12771805027887803,
+      "grad_norm": 0.4279276132583618,
+      "learning_rate": 0.0002,
+      "loss": 0.8112,
+      "step": 790
+    },
+    {
+      "epoch": 0.12933473445962332,
+      "grad_norm": 0.46212655305862427,
+      "learning_rate": 0.0002,
+      "loss": 0.8711,
+      "step": 800
+    },
+    {
+      "epoch": 0.1309514186403686,
+      "grad_norm": 0.43127650022506714,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 810
+    },
+    {
+      "epoch": 0.1325681028211139,
+      "grad_norm": 0.4201301336288452,
+      "learning_rate": 0.0002,
+      "loss": 0.8476,
+      "step": 820
+    },
+    {
+      "epoch": 0.13418478700185918,
+      "grad_norm": 0.42583167552948,
+      "learning_rate": 0.0002,
+      "loss": 0.8078,
+      "step": 830
+    },
+    {
+      "epoch": 0.13580147118260447,
+      "grad_norm": 0.4535622000694275,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 840
+    },
+    {
+      "epoch": 0.13741815536334978,
+      "grad_norm": 0.4116036891937256,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 850
+    },
+    {
+      "epoch": 0.13903483954409507,
+      "grad_norm": 0.45997580885887146,
+      "learning_rate": 0.0002,
+      "loss": 0.8466,
+      "step": 860
+    },
+    {
+      "epoch": 0.14065152372484035,
+      "grad_norm": 0.4487837255001068,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 870
+    },
+    {
+      "epoch": 0.14226820790558564,
+      "grad_norm": 0.43650057911872864,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 880
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 0.5335358381271362,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 890
+    },
+    {
+      "epoch": 0.14550157626707622,
+      "grad_norm": 0.5989000201225281,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 900
+    },
+    {
+      "epoch": 0.14711826044782153,
+      "grad_norm": 0.517179012298584,
+      "learning_rate": 0.0002,
+      "loss": 0.8385,
+      "step": 910
+    },
+    {
+      "epoch": 0.14873494462856682,
+      "grad_norm": 0.44435232877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.8255,
+      "step": 920
+    },
+    {
+      "epoch": 0.1503516288093121,
+      "grad_norm": 0.42635923624038696,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 930
+    },
+    {
+      "epoch": 0.1519683129900574,
+      "grad_norm": 0.49603334069252014,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 940
+    },
+    {
+      "epoch": 0.15358499717080268,
+      "grad_norm": 0.40639808773994446,
+      "learning_rate": 0.0002,
+      "loss": 0.8377,
+      "step": 950
+    },
+    {
+      "epoch": 0.15520168135154797,
+      "grad_norm": 0.4850759208202362,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 960
+    },
+    {
+      "epoch": 0.15681836553229328,
+      "grad_norm": 0.4427442252635956,
+      "learning_rate": 0.0002,
+      "loss": 0.846,
+      "step": 970
+    },
+    {
+      "epoch": 0.15843504971303857,
+      "grad_norm": 0.3760930001735687,
+      "learning_rate": 0.0002,
+      "loss": 0.8705,
+      "step": 980
+    },
+    {
+      "epoch": 0.16005173389378385,
+      "grad_norm": 0.4794144332408905,
+      "learning_rate": 0.0002,
+      "loss": 0.8644,
+      "step": 990
+    },
+    {
+      "epoch": 0.16166841807452914,
+      "grad_norm": 0.45828768610954285,
+      "learning_rate": 0.0002,
+      "loss": 0.8002,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16328510225527443,
+      "grad_norm": 0.6313053369522095,
+      "learning_rate": 0.0002,
+      "loss": 0.7658,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16490178643601971,
+      "grad_norm": 0.45041006803512573,
+      "learning_rate": 0.0002,
+      "loss": 0.8047,
+      "step": 1020
+    },
+    {
+      "epoch": 0.166518470616765,
+      "grad_norm": 0.441403865814209,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16813515479751032,
+      "grad_norm": 0.8171296119689941,
+      "learning_rate": 0.0002,
+      "loss": 0.8475,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1697518389782556,
+      "grad_norm": 0.7137420773506165,
+      "learning_rate": 0.0002,
+      "loss": 0.845,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1713685231590009,
+      "grad_norm": 0.5236809849739075,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17298520733974618,
+      "grad_norm": 0.5021864175796509,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17460189152049146,
+      "grad_norm": 0.47347521781921387,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17621857570123675,
+      "grad_norm": 0.4631653428077698,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17783525988198207,
+      "grad_norm": 0.49169182777404785,
+      "learning_rate": 0.0002,
+      "loss": 0.8166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17945194406272735,
+      "grad_norm": 0.5019739270210266,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18106862824347264,
+      "grad_norm": 0.5100422501564026,
+      "learning_rate": 0.0002,
+      "loss": 0.8247,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18268531242421793,
+      "grad_norm": 0.3888324499130249,
+      "learning_rate": 0.0002,
+      "loss": 0.8142,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18430199660496321,
+      "grad_norm": 0.39765217900276184,
+      "learning_rate": 0.0002,
+      "loss": 0.8533,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1859186807857085,
+      "grad_norm": 0.47190186381340027,
+      "learning_rate": 0.0002,
+      "loss": 0.8541,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18753536496645382,
+      "grad_norm": 0.4464188814163208,
+      "learning_rate": 0.0002,
+      "loss": 0.8301,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1891520491471991,
+      "grad_norm": 0.5153930187225342,
+      "learning_rate": 0.0002,
+      "loss": 0.8341,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1907687333279444,
+      "grad_norm": 0.4779708683490753,
+      "learning_rate": 0.0002,
+      "loss": 0.8033,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19238541750868968,
+      "grad_norm": 0.4834315776824951,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19400210168943496,
+      "grad_norm": 0.402357816696167,
+      "learning_rate": 0.0002,
+      "loss": 0.7721,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19561878587018025,
+      "grad_norm": 0.45899084210395813,
+      "learning_rate": 0.0002,
+      "loss": 0.7941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19723547005092557,
+      "grad_norm": 0.5106529593467712,
+      "learning_rate": 0.0002,
+      "loss": 0.8353,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19885215423167085,
+      "grad_norm": 0.45261722803115845,
+      "learning_rate": 0.0002,
+      "loss": 0.7816,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20046883841241614,
+      "grad_norm": 0.4647127091884613,
+      "learning_rate": 0.0002,
+      "loss": 0.8068,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20208552259316143,
+      "grad_norm": 0.4849368929862976,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2037022067739067,
+      "grad_norm": 0.4518061578273773,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 1260
+    },
+    {
+      "epoch": 0.205318890954652,
+      "grad_norm": 0.49535325169563293,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2069355751353973,
+      "grad_norm": 0.4835205376148224,
+      "learning_rate": 0.0002,
+      "loss": 0.8348,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2085522593161426,
+      "grad_norm": 0.45308539271354675,
+      "learning_rate": 0.0002,
+      "loss": 0.8428,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2101689434968879,
+      "grad_norm": 0.5369905233383179,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21178562767763318,
+      "grad_norm": 0.5031622052192688,
+      "learning_rate": 0.0002,
+      "loss": 0.8676,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21340231185837846,
+      "grad_norm": 0.48010334372520447,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21501899603912375,
+      "grad_norm": 0.4905701279640198,
+      "learning_rate": 0.0002,
+      "loss": 0.806,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21663568021986904,
+      "grad_norm": 0.43531742691993713,
+      "learning_rate": 0.0002,
+      "loss": 0.7885,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21825236440061435,
+      "grad_norm": 0.44330692291259766,
+      "learning_rate": 0.0002,
+      "loss": 0.8191,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21986904858135964,
+      "grad_norm": 0.5384416580200195,
+      "learning_rate": 0.0002,
+      "loss": 0.8205,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22148573276210493,
+      "grad_norm": 0.4181833863258362,
+      "learning_rate": 0.0002,
+      "loss": 0.7726,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2231024169428502,
+      "grad_norm": 0.523833692073822,
+      "learning_rate": 0.0002,
+      "loss": 0.8311,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.5528736710548401,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2263357853043408,
+      "grad_norm": 0.43515023589134216,
+      "learning_rate": 0.0002,
+      "loss": 0.8079,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2279524694850861,
+      "grad_norm": 0.48809877038002014,
+      "learning_rate": 0.0002,
+      "loss": 0.8403,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2295691536658314,
+      "grad_norm": 0.43591251969337463,
+      "learning_rate": 0.0002,
+      "loss": 0.8165,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23118583784657668,
+      "grad_norm": 0.44625312089920044,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23280252202732196,
+      "grad_norm": 0.4390665292739868,
+      "learning_rate": 0.0002,
+      "loss": 0.8134,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23441920620806725,
+      "grad_norm": 0.48496049642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.8465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23603589038881254,
+      "grad_norm": 0.45919957756996155,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23765257456955785,
+      "grad_norm": 0.5471845865249634,
+      "learning_rate": 0.0002,
+      "loss": 0.8659,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23926925875030314,
+      "grad_norm": 0.47269317507743835,
+      "learning_rate": 0.0002,
+      "loss": 0.8164,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24088594293104842,
+      "grad_norm": 0.4930245578289032,
+      "learning_rate": 0.0002,
+      "loss": 0.854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2425026271117937,
+      "grad_norm": 0.5605630278587341,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 1500
+    },
+    {
+      "epoch": 0.244119311292539,
+      "grad_norm": 0.4435870945453644,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24573599547328429,
+      "grad_norm": 0.4941999912261963,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24735267965402957,
+      "grad_norm": 0.5100624561309814,
+      "learning_rate": 0.0002,
+      "loss": 0.8427,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2489693638347749,
+      "grad_norm": 0.4638267457485199,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25058604801552015,
+      "grad_norm": 0.5071570873260498,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25220273219626543,
+      "grad_norm": 0.4291319251060486,
+      "learning_rate": 0.0002,
+      "loss": 0.7724,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2538194163770108,
+      "grad_norm": 0.5388049483299255,
+      "learning_rate": 0.0002,
+      "loss": 0.7984,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25543610055775606,
+      "grad_norm": 0.5083683729171753,
+      "learning_rate": 0.0002,
+      "loss": 0.8176,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25705278473850135,
+      "grad_norm": 0.4824463725090027,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25866946891924664,
+      "grad_norm": 0.41177722811698914,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2602861530999919,
+      "grad_norm": 0.5656219124794006,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2619028372807372,
+      "grad_norm": 0.41063204407691956,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2635195214614825,
+      "grad_norm": 0.4897061288356781,
+      "learning_rate": 0.0002,
+      "loss": 0.7998,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2651362056422278,
+      "grad_norm": 0.4454376697540283,
+      "learning_rate": 0.0002,
+      "loss": 0.8198,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26675288982297307,
+      "grad_norm": 0.4355238378047943,
+      "learning_rate": 0.0002,
+      "loss": 0.8684,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26836957400371836,
+      "grad_norm": 0.458310067653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26998625818446365,
+      "grad_norm": 0.4752083718776703,
+      "learning_rate": 0.0002,
+      "loss": 0.7935,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27160294236520893,
+      "grad_norm": 0.4666106402873993,
+      "learning_rate": 0.0002,
+      "loss": 0.8267,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2732196265459543,
+      "grad_norm": 0.4213818609714508,
+      "learning_rate": 0.0002,
+      "loss": 0.8252,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27483631072669956,
+      "grad_norm": 0.5768913626670837,
+      "learning_rate": 0.0002,
+      "loss": 0.8559,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27645299490744485,
+      "grad_norm": 0.4209914803504944,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27806967908819014,
+      "grad_norm": 0.501909613609314,
+      "learning_rate": 0.0002,
+      "loss": 0.8167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2796863632689354,
+      "grad_norm": 0.5266261100769043,
+      "learning_rate": 0.0002,
+      "loss": 0.7832,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2813030474496807,
+      "grad_norm": 0.43806859850883484,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 1740
+    },
+    {
+      "epoch": 0.282919731630426,
+      "grad_norm": 0.46048814058303833,
+      "learning_rate": 0.0002,
+      "loss": 0.8157,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2845364158111713,
+      "grad_norm": 0.44972819089889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8596,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28615309999191657,
+      "grad_norm": 0.5114831328392029,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.47931742668151855,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28938646835340714,
+      "grad_norm": 0.5092599987983704,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29100315253415243,
+      "grad_norm": 0.37581443786621094,
+      "learning_rate": 0.0002,
+      "loss": 0.8506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2926198367148977,
+      "grad_norm": 0.47097381949424744,
+      "learning_rate": 0.0002,
+      "loss": 0.7932,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29423652089564306,
+      "grad_norm": 0.48300236463546753,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29585320507638835,
+      "grad_norm": 0.5600419640541077,
+      "learning_rate": 0.0002,
+      "loss": 0.8391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29746988925713364,
+      "grad_norm": 0.48555272817611694,
+      "learning_rate": 0.0002,
+      "loss": 0.8507,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2990865734378789,
+      "grad_norm": 0.3752668499946594,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3007032576186242,
+      "grad_norm": 0.5328747034072876,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3023199417993695,
+      "grad_norm": 0.48716455698013306,
+      "learning_rate": 0.0002,
+      "loss": 0.8426,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3039366259801148,
+      "grad_norm": 0.5011493563652039,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30555331016086007,
+      "grad_norm": 0.46461427211761475,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30716999434160536,
+      "grad_norm": 0.36630210280418396,
+      "learning_rate": 0.0002,
+      "loss": 0.8478,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30878667852235064,
+      "grad_norm": 0.4217296242713928,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31040336270309593,
+      "grad_norm": 0.4394875466823578,
+      "learning_rate": 0.0002,
+      "loss": 0.8128,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3120200468838412,
+      "grad_norm": 0.6587965488433838,
+      "learning_rate": 0.0002,
+      "loss": 0.8471,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31363673106458656,
+      "grad_norm": 0.5469298958778381,
+      "learning_rate": 0.0002,
+      "loss": 0.8565,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31525341524533185,
+      "grad_norm": 0.4371595084667206,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31687009942607713,
+      "grad_norm": 0.4809541404247284,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3184867836068224,
+      "grad_norm": 0.6061086654663086,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3201034677875677,
+      "grad_norm": 0.5342657566070557,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 1980
+    },
+    {
+      "epoch": 0.321720151968313,
+      "grad_norm": 0.5057743787765503,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3233368361490583,
+      "grad_norm": 0.528626024723053,
+      "learning_rate": 0.0002,
+      "loss": 0.7774,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32495352032980357,
+      "grad_norm": 0.46742770075798035,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32657020451054886,
+      "grad_norm": 0.515101432800293,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32818688869129414,
+      "grad_norm": 0.41941216588020325,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32980357287203943,
+      "grad_norm": 0.49902522563934326,
+      "learning_rate": 0.0002,
+      "loss": 0.7637,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3314202570527847,
+      "grad_norm": 0.4120897650718689,
+      "learning_rate": 0.0002,
+      "loss": 0.7822,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33303694123353,
+      "grad_norm": 0.45352041721343994,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33465362541427535,
+      "grad_norm": 0.523199737071991,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33627030959502063,
+      "grad_norm": 0.4390358626842499,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3378869937757659,
+      "grad_norm": 0.6752901077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3395036779565112,
+      "grad_norm": 0.547821044921875,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3411203621372565,
+      "grad_norm": 0.5161308646202087,
+      "learning_rate": 0.0002,
+      "loss": 0.8561,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3427370463180018,
+      "grad_norm": 0.4565401077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34435373049874707,
+      "grad_norm": 0.4666115939617157,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34597041467949236,
+      "grad_norm": 0.4090428352355957,
+      "learning_rate": 0.0002,
+      "loss": 0.8189,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34758709886023764,
+      "grad_norm": 0.510845422744751,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34920378304098293,
+      "grad_norm": 0.42861923575401306,
+      "learning_rate": 0.0002,
+      "loss": 0.8398,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3508204672217282,
+      "grad_norm": 0.4476332664489746,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3524371514024735,
+      "grad_norm": 0.6065791249275208,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35405383558321885,
+      "grad_norm": 0.42335066199302673,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35567051976396413,
+      "grad_norm": 0.5094629526138306,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3572872039447094,
+      "grad_norm": 0.5476373434066772,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3589038881254547,
+      "grad_norm": 0.3911719024181366,
+      "learning_rate": 0.0002,
+      "loss": 0.814,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3605205723062,
+      "grad_norm": 0.6599636077880859,
+      "learning_rate": 0.0002,
+      "loss": 0.8599,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3621372564869453,
+      "grad_norm": 0.40381914377212524,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36375394066769057,
+      "grad_norm": 0.4433908462524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36537062484843585,
+      "grad_norm": 0.578326940536499,
+      "learning_rate": 0.0002,
+      "loss": 0.8503,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36698730902918114,
+      "grad_norm": 0.5734784007072449,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36860399320992643,
+      "grad_norm": 0.45555487275123596,
+      "learning_rate": 0.0002,
+      "loss": 0.8193,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3702206773906717,
+      "grad_norm": 0.5666276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.7929,
+      "step": 2290
+    },
+    {
+      "epoch": 0.371837361571417,
+      "grad_norm": 0.5461117625236511,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3734540457521623,
+      "grad_norm": 0.6318911910057068,
+      "learning_rate": 0.0002,
+      "loss": 0.8204,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37507072993290763,
+      "grad_norm": 0.493263304233551,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3766874141136529,
+      "grad_norm": 0.5888760089874268,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3783040982943982,
+      "grad_norm": 0.48671841621398926,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3799207824751435,
+      "grad_norm": 0.4385145306587219,
+      "learning_rate": 0.0002,
+      "loss": 0.8367,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3815374666558888,
+      "grad_norm": 0.5523318648338318,
+      "learning_rate": 0.0002,
+      "loss": 0.812,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38315415083663407,
+      "grad_norm": 0.7308220267295837,
+      "learning_rate": 0.0002,
+      "loss": 0.8351,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38477083501737935,
+      "grad_norm": 0.554214358329773,
+      "learning_rate": 0.0002,
+      "loss": 0.859,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38638751919812464,
+      "grad_norm": 0.5425800085067749,
+      "learning_rate": 0.0002,
+      "loss": 0.8146,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3880042033788699,
+      "grad_norm": 0.48811158537864685,
+      "learning_rate": 0.0002,
+      "loss": 0.8282,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3896208875596152,
+      "grad_norm": 0.49212366342544556,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3912375717403605,
+      "grad_norm": 0.5222218632698059,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3928542559211058,
+      "grad_norm": 0.4699819087982178,
+      "learning_rate": 0.0002,
+      "loss": 0.8182,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39447094010185113,
+      "grad_norm": 0.46153587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7919,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3960876242825964,
+      "grad_norm": 0.4150611162185669,
+      "learning_rate": 0.0002,
+      "loss": 0.8111,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3977043084633417,
+      "grad_norm": 0.5799614787101746,
+      "learning_rate": 0.0002,
+      "loss": 0.8589,
+      "step": 2460
+    },
+    {
+      "epoch": 0.399320992644087,
+      "grad_norm": 0.56536865234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4009376768248323,
+      "grad_norm": 0.5451247096061707,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40255436100557757,
+      "grad_norm": 0.5914521217346191,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40417104518632285,
+      "grad_norm": 0.4428117275238037,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40578772936706814,
+      "grad_norm": 0.48580947518348694,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4074044135478134,
+      "grad_norm": 0.436734676361084,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4090210977285587,
+      "grad_norm": 0.5752223134040833,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 2530
+    },
+    {
+      "epoch": 0.410637781909304,
+      "grad_norm": 0.4271308183670044,
+      "learning_rate": 0.0002,
+      "loss": 0.8181,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4122544660900493,
+      "grad_norm": 0.46294718980789185,
+      "learning_rate": 0.0002,
+      "loss": 0.8058,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4138711502707946,
+      "grad_norm": 0.49407583475112915,
+      "learning_rate": 0.0002,
+      "loss": 0.8473,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4154878344515399,
+      "grad_norm": 0.4729035496711731,
+      "learning_rate": 0.0002,
+      "loss": 0.7881,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4171045186322852,
+      "grad_norm": 0.4129747152328491,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4187212028130305,
+      "grad_norm": 0.5684236288070679,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4203378869937758,
+      "grad_norm": 0.4862157106399536,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42195457117452106,
+      "grad_norm": 0.46567976474761963,
+      "learning_rate": 0.0002,
+      "loss": 0.7582,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42357125535526635,
+      "grad_norm": 0.5710650682449341,
+      "learning_rate": 0.0002,
+      "loss": 0.7755,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42518793953601164,
+      "grad_norm": 0.5660041570663452,
+      "learning_rate": 0.0002,
+      "loss": 0.8573,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4268046237167569,
+      "grad_norm": 0.47944375872612,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4284213078975022,
+      "grad_norm": 0.537223756313324,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4300379920782475,
+      "grad_norm": 0.41669997572898865,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 0.44727686047554016,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4332713604397381,
+      "grad_norm": 0.5600888729095459,
+      "learning_rate": 0.0002,
+      "loss": 0.8241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4348880446204834,
+      "grad_norm": 0.39820605516433716,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4365047288012287,
+      "grad_norm": 0.5637655854225159,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 2700
+    },
+    {
+      "epoch": 0.438121412981974,
+      "grad_norm": 0.6363666653633118,
+      "learning_rate": 0.0002,
+      "loss": 0.855,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4397380971627193,
+      "grad_norm": 0.5656129121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.8468,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44135478134346456,
+      "grad_norm": 0.5600156188011169,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44297146552420985,
+      "grad_norm": 0.5506579875946045,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44458814970495514,
+      "grad_norm": 0.49878305196762085,
+      "learning_rate": 0.0002,
+      "loss": 0.7725,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4462048338857004,
+      "grad_norm": 0.4569213092327118,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4478215180664457,
+      "grad_norm": 0.6056680083274841,
+      "learning_rate": 0.0002,
+      "loss": 0.8028,
+      "step": 2770
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.44474557042121887,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4510548864279363,
+      "grad_norm": 0.46055394411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.801,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4526715706086816,
+      "grad_norm": 0.4904133379459381,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45428825478942686,
+      "grad_norm": 0.5647031664848328,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4559049389701722,
+      "grad_norm": 0.5759473443031311,
+      "learning_rate": 0.0002,
+      "loss": 0.8622,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4575216231509175,
+      "grad_norm": 0.5161895751953125,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4591383073316628,
+      "grad_norm": 0.4248254597187042,
+      "learning_rate": 0.0002,
+      "loss": 0.8045,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46075499151240806,
+      "grad_norm": 0.45395001769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.7838,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46237167569315335,
+      "grad_norm": 0.5358697772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.8208,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46398835987389864,
+      "grad_norm": 0.5379165410995483,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4656050440546439,
+      "grad_norm": 0.4601989686489105,
+      "learning_rate": 0.0002,
+      "loss": 0.7403,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4672217282353892,
+      "grad_norm": 0.671115517616272,
+      "learning_rate": 0.0002,
+      "loss": 0.8523,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4688384124161345,
+      "grad_norm": 0.4425133168697357,
+      "learning_rate": 0.0002,
+      "loss": 0.8262,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4704550965968798,
+      "grad_norm": 0.5446155071258545,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47207178077762507,
+      "grad_norm": 0.603306233882904,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47368846495837036,
+      "grad_norm": 0.5377997159957886,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4753051491391157,
+      "grad_norm": 0.4931027591228485,
+      "learning_rate": 0.0002,
+      "loss": 0.8075,
+      "step": 2940
+    },
+    {
+      "epoch": 0.476921833319861,
+      "grad_norm": 0.4711960256099701,
+      "learning_rate": 0.0002,
+      "loss": 0.8004,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4785385175006063,
+      "grad_norm": 0.5020492672920227,
+      "learning_rate": 0.0002,
+      "loss": 0.8121,
+      "step": 2960
+    },
+    {
+      "epoch": 0.48015520168135156,
+      "grad_norm": 0.5428946614265442,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48177188586209685,
+      "grad_norm": 0.5294089317321777,
+      "learning_rate": 0.0002,
+      "loss": 0.7849,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48338857004284214,
+      "grad_norm": 0.648289144039154,
+      "learning_rate": 0.0002,
+      "loss": 0.8553,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4850052542235874,
+      "grad_norm": 0.47916680574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.7874,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4866219384043327,
+      "grad_norm": 0.43849772214889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8087,
+      "step": 3010
+    },
+    {
+      "epoch": 0.488238622585078,
+      "grad_norm": 0.47007861733436584,
+      "learning_rate": 0.0002,
+      "loss": 0.7662,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4898553067658233,
+      "grad_norm": 0.6314331293106079,
+      "learning_rate": 0.0002,
+      "loss": 0.757,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49147199094656857,
+      "grad_norm": 0.49211493134498596,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49308867512731386,
+      "grad_norm": 0.4537973403930664,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49470535930805914,
+      "grad_norm": 0.47326919436454773,
+      "learning_rate": 0.0002,
+      "loss": 0.8095,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4963220434888045,
+      "grad_norm": 0.525874137878418,
+      "learning_rate": 0.0002,
+      "loss": 0.8447,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4979387276695498,
+      "grad_norm": 0.6361091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49955541185029506,
+      "grad_norm": 0.5850642919540405,
+      "learning_rate": 0.0002,
+      "loss": 0.821,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5011720960310403,
+      "grad_norm": 0.47299543023109436,
+      "learning_rate": 0.0002,
+      "loss": 0.8279,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5027887802117856,
+      "grad_norm": 0.473099946975708,
+      "learning_rate": 0.0002,
+      "loss": 0.8681,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5044054643925309,
+      "grad_norm": 0.48186397552490234,
+      "learning_rate": 0.0002,
+      "loss": 0.8223,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5060221485732762,
+      "grad_norm": 0.5015401840209961,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5076388327540216,
+      "grad_norm": 0.5617750287055969,
+      "learning_rate": 0.0002,
+      "loss": 0.7692,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5092555169347668,
+      "grad_norm": 0.5169327259063721,
+      "learning_rate": 0.0002,
+      "loss": 0.8708,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5108722011155121,
+      "grad_norm": 0.545657753944397,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5124888852962574,
+      "grad_norm": 0.512864351272583,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5141055694770027,
+      "grad_norm": 0.4113546311855316,
+      "learning_rate": 0.0002,
+      "loss": 0.7794,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5157222536577479,
+      "grad_norm": 0.44532445073127747,
+      "learning_rate": 0.0002,
+      "loss": 0.8206,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5173389378384933,
+      "grad_norm": 0.5623497366905212,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5189556220192385,
+      "grad_norm": 0.5084741115570068,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5205723061999838,
+      "grad_norm": 0.5305403470993042,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5221889903807291,
+      "grad_norm": 0.4708254337310791,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5238056745614744,
+      "grad_norm": 0.43827131390571594,
+      "learning_rate": 0.0002,
+      "loss": 0.7639,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5254223587422197,
+      "grad_norm": 0.5630002617835999,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 3250
+    },
+    {
+      "epoch": 0.527039042922965,
+      "grad_norm": 0.5010961890220642,
+      "learning_rate": 0.0002,
+      "loss": 0.7522,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5286557271037103,
+      "grad_norm": 0.6303122043609619,
+      "learning_rate": 0.0002,
+      "loss": 0.8374,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5302724112844556,
+      "grad_norm": 0.5107331275939941,
+      "learning_rate": 0.0002,
+      "loss": 0.7727,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5318890954652009,
+      "grad_norm": 0.5700443387031555,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5335057796459461,
+      "grad_norm": 0.46296367049217224,
+      "learning_rate": 0.0002,
+      "loss": 0.7776,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5351224638266915,
+      "grad_norm": 0.531568706035614,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5367391480074367,
+      "grad_norm": 0.4686741530895233,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5383558321881821,
+      "grad_norm": 0.5404331088066101,
+      "learning_rate": 0.0002,
+      "loss": 0.8104,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5399725163689273,
+      "grad_norm": 0.6368790864944458,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5415892005496726,
+      "grad_norm": 0.42300888895988464,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5432058847304179,
+      "grad_norm": 0.5362542867660522,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5448225689111632,
+      "grad_norm": 0.497128963470459,
+      "learning_rate": 0.0002,
+      "loss": 0.858,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5464392530919085,
+      "grad_norm": 0.5006386041641235,
+      "learning_rate": 0.0002,
+      "loss": 0.8519,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5480559372726538,
+      "grad_norm": 0.44136837124824524,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5496726214533991,
+      "grad_norm": 0.5897833108901978,
+      "learning_rate": 0.0002,
+      "loss": 0.773,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5512893056341444,
+      "grad_norm": 0.641075611114502,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5529059898148897,
+      "grad_norm": 0.7251322269439697,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5545226739956349,
+      "grad_norm": 0.47411349415779114,
+      "learning_rate": 0.0002,
+      "loss": 0.7626,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5561393581763803,
+      "grad_norm": 0.4994310438632965,
+      "learning_rate": 0.0002,
+      "loss": 0.8196,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5577560423571255,
+      "grad_norm": 0.5814438462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5593727265378708,
+      "grad_norm": 0.6278898119926453,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5609894107186161,
+      "grad_norm": 0.46208274364471436,
+      "learning_rate": 0.0002,
+      "loss": 0.813,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5626060948993614,
+      "grad_norm": 0.5718930959701538,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5642227790801067,
+      "grad_norm": 0.48178744316101074,
+      "learning_rate": 0.0002,
+      "loss": 0.8152,
+      "step": 3490
+    },
+    {
+      "epoch": 0.565839463260852,
+      "grad_norm": 0.47336965799331665,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5674561474415973,
+      "grad_norm": 0.43442684412002563,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5690728316223426,
+      "grad_norm": 0.6463358998298645,
+      "learning_rate": 0.0002,
+      "loss": 0.7564,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5706895158030879,
+      "grad_norm": 0.5286486744880676,
+      "learning_rate": 0.0002,
+      "loss": 0.836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5723061999838331,
+      "grad_norm": 0.5405499935150146,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5739228841645785,
+      "grad_norm": 0.6654391884803772,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.5081980228424072,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5771562525260691,
+      "grad_norm": 0.48978179693222046,
+      "learning_rate": 0.0002,
+      "loss": 0.7753,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5787729367068143,
+      "grad_norm": 0.5840612053871155,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5803896208875596,
+      "grad_norm": 0.5235261917114258,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5820063050683049,
+      "grad_norm": 0.5672075748443604,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5836229892490502,
+      "grad_norm": 0.5613429546356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5852396734297954,
+      "grad_norm": 0.4032273590564728,
+      "learning_rate": 0.0002,
+      "loss": 0.8274,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5868563576105408,
+      "grad_norm": 0.49559324979782104,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5884730417912861,
+      "grad_norm": 0.6895697712898254,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5900897259720314,
+      "grad_norm": 0.4750136435031891,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5917064101527767,
+      "grad_norm": 0.5176819562911987,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5933230943335219,
+      "grad_norm": 0.5817760229110718,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5949397785142673,
+      "grad_norm": 0.6064626574516296,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5965564626950125,
+      "grad_norm": 0.6728700995445251,
+      "learning_rate": 0.0002,
+      "loss": 0.8422,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5981731468757578,
+      "grad_norm": 0.609305202960968,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5997898310565031,
+      "grad_norm": 0.4615488350391388,
+      "learning_rate": 0.0002,
+      "loss": 0.8048,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6014065152372484,
+      "grad_norm": 2.0531179904937744,
+      "learning_rate": 0.0002,
+      "loss": 0.8214,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6030231994179936,
+      "grad_norm": 0.5091132521629333,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 3730
+    },
+    {
+      "epoch": 0.604639883598739,
+      "grad_norm": 0.5951124429702759,
+      "learning_rate": 0.0002,
+      "loss": 0.7833,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6062565677794842,
+      "grad_norm": 0.5870208144187927,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6078732519602296,
+      "grad_norm": 0.6254619359970093,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6094899361409749,
+      "grad_norm": 0.5577626824378967,
+      "learning_rate": 0.0002,
+      "loss": 0.7868,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6111066203217201,
+      "grad_norm": 0.5004405379295349,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6127233045024655,
+      "grad_norm": 0.5527383685112,
+      "learning_rate": 0.0002,
+      "loss": 0.8092,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6143399886832107,
+      "grad_norm": 0.49116113781929016,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6159566728639561,
+      "grad_norm": 0.5299299359321594,
+      "learning_rate": 0.0002,
+      "loss": 0.8352,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6175733570447013,
+      "grad_norm": 0.464897483587265,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6191900412254466,
+      "grad_norm": 0.6505740880966187,
+      "learning_rate": 0.0002,
+      "loss": 0.7923,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6208067254061919,
+      "grad_norm": 0.5512559413909912,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6224234095869372,
+      "grad_norm": 0.49427518248558044,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6240400937676824,
+      "grad_norm": 0.3839147090911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6256567779484278,
+      "grad_norm": 0.5760218501091003,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6272734621291731,
+      "grad_norm": 0.7226507067680359,
+      "learning_rate": 0.0002,
+      "loss": 0.7836,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6288901463099184,
+      "grad_norm": 0.676781415939331,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6305068304906637,
+      "grad_norm": 0.4284018278121948,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6321235146714089,
+      "grad_norm": 0.5060628056526184,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6337401988521543,
+      "grad_norm": 0.5524522066116333,
+      "learning_rate": 0.0002,
+      "loss": 0.8089,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6353568830328995,
+      "grad_norm": 0.6099881529808044,
+      "learning_rate": 0.0002,
+      "loss": 0.8276,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6369735672136448,
+      "grad_norm": 0.43155938386917114,
+      "learning_rate": 0.0002,
+      "loss": 0.809,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6385902513943901,
+      "grad_norm": 0.6427084803581238,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6402069355751354,
+      "grad_norm": 0.541220486164093,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6418236197558806,
+      "grad_norm": 0.5414294600486755,
+      "learning_rate": 0.0002,
+      "loss": 0.8539,
+      "step": 3970
+    },
+    {
+      "epoch": 0.643440303936626,
+      "grad_norm": 0.46344003081321716,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6450569881173712,
+      "grad_norm": 0.45209285616874695,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6466736722981166,
+      "grad_norm": 0.5417284369468689,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6482903564788619,
+      "grad_norm": 0.7995685935020447,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6499070406596071,
+      "grad_norm": 0.6384002566337585,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6515237248403525,
+      "grad_norm": 0.4472815692424774,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6531404090210977,
+      "grad_norm": 0.6834294199943542,
+      "learning_rate": 0.0002,
+      "loss": 0.8071,
+      "step": 4040
+    },
+    {
+      "epoch": 0.654757093201843,
+      "grad_norm": 0.4612339735031128,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6563737773825883,
+      "grad_norm": 0.9266576170921326,
+      "learning_rate": 0.0002,
+      "loss": 0.8141,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6579904615633336,
+      "grad_norm": 0.4470861852169037,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6596071457440789,
+      "grad_norm": 0.45544925332069397,
+      "learning_rate": 0.0002,
+      "loss": 0.8293,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6612238299248242,
+      "grad_norm": 0.6144481301307678,
+      "learning_rate": 0.0002,
+      "loss": 0.8455,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6628405141055694,
+      "grad_norm": 0.5936288237571716,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6644571982863148,
+      "grad_norm": 0.4822963774204254,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 4110
+    },
+    {
+      "epoch": 0.66607388246706,
+      "grad_norm": 0.48432496190071106,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6676905666478054,
+      "grad_norm": 0.4901607930660248,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6693072508285507,
+      "grad_norm": 0.5018393397331238,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6709239350092959,
+      "grad_norm": 0.6946378946304321,
+      "learning_rate": 0.0002,
+      "loss": 0.8065,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6725406191900413,
+      "grad_norm": 0.5997390747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.6738849878311157,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6757739875515318,
+      "grad_norm": 0.6110581159591675,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6773906717322771,
+      "grad_norm": 0.5703322291374207,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6790073559130224,
+      "grad_norm": 0.4686066210269928,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6806240400937676,
+      "grad_norm": 0.6394643783569336,
+      "learning_rate": 0.0002,
+      "loss": 0.8441,
+      "step": 4210
+    },
+    {
+      "epoch": 0.682240724274513,
+      "grad_norm": 0.5454841256141663,
+      "learning_rate": 0.0002,
+      "loss": 0.8011,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6838574084552582,
+      "grad_norm": 0.4859732985496521,
+      "learning_rate": 0.0002,
+      "loss": 0.8307,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6854740926360036,
+      "grad_norm": 0.5544065833091736,
+      "learning_rate": 0.0002,
+      "loss": 0.8161,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6870907768167488,
+      "grad_norm": 0.4902505576610565,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6887074609974941,
+      "grad_norm": 0.4768051505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7977,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6903241451782395,
+      "grad_norm": 0.49982190132141113,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6919408293589847,
+      "grad_norm": 0.6351838111877441,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 4280
+    },
+    {
+      "epoch": 0.69355751353973,
+      "grad_norm": 0.5647561550140381,
+      "learning_rate": 0.0002,
+      "loss": 0.7664,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6951741977204753,
+      "grad_norm": 0.5340486764907837,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6967908819012206,
+      "grad_norm": 0.5649092793464661,
+      "learning_rate": 0.0002,
+      "loss": 0.8526,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6984075660819659,
+      "grad_norm": 0.6183916926383972,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7000242502627112,
+      "grad_norm": 0.6154509782791138,
+      "learning_rate": 0.0002,
+      "loss": 0.792,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7016409344434564,
+      "grad_norm": 0.5156264305114746,
+      "learning_rate": 0.0002,
+      "loss": 0.8397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7032576186242018,
+      "grad_norm": 0.562171459197998,
+      "learning_rate": 0.0002,
+      "loss": 0.8512,
+      "step": 4350
+    },
+    {
+      "epoch": 0.704874302804947,
+      "grad_norm": 0.4949502646923065,
+      "learning_rate": 0.0002,
+      "loss": 0.7882,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7064909869856923,
+      "grad_norm": 0.5171684622764587,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7081076711664377,
+      "grad_norm": 0.6198443174362183,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7097243553471829,
+      "grad_norm": 0.5802276134490967,
+      "learning_rate": 0.0002,
+      "loss": 0.7606,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7113410395279283,
+      "grad_norm": 0.41096967458724976,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7129577237086735,
+      "grad_norm": 0.4397392272949219,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7145744078894188,
+      "grad_norm": 0.45228442549705505,
+      "learning_rate": 0.0002,
+      "loss": 0.7651,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7161910920701641,
+      "grad_norm": 0.4839673936367035,
+      "learning_rate": 0.0002,
+      "loss": 0.7938,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7178077762509094,
+      "grad_norm": 0.6140755414962769,
+      "learning_rate": 0.0002,
+      "loss": 0.8362,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 0.6841378808021545,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7210411446124,
+      "grad_norm": 0.6664239168167114,
+      "learning_rate": 0.0002,
+      "loss": 0.8177,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7226578287931452,
+      "grad_norm": 0.47552719712257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7983,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7242745129738906,
+      "grad_norm": 0.6649776101112366,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7258911971546358,
+      "grad_norm": 0.5159541964530945,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7275078813353811,
+      "grad_norm": 0.6693112850189209,
+      "learning_rate": 0.0002,
+      "loss": 0.7786,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7291245655161265,
+      "grad_norm": 0.48870977759361267,
+      "learning_rate": 0.0002,
+      "loss": 0.8655,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7307412496968717,
+      "grad_norm": 0.4857887923717499,
+      "learning_rate": 0.0002,
+      "loss": 0.7337,
+      "step": 4520
+    },
+    {
+      "epoch": 0.732357933877617,
+      "grad_norm": 0.5515662431716919,
+      "learning_rate": 0.0002,
+      "loss": 0.8026,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7339746180583623,
+      "grad_norm": 0.6292222738265991,
+      "learning_rate": 0.0002,
+      "loss": 0.8031,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7355913022391076,
+      "grad_norm": 0.48265689611434937,
+      "learning_rate": 0.0002,
+      "loss": 0.7749,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7372079864198529,
+      "grad_norm": 0.8044266104698181,
+      "learning_rate": 0.0002,
+      "loss": 0.8499,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7388246706005982,
+      "grad_norm": 0.6111769676208496,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7404413547813434,
+      "grad_norm": 0.5229553580284119,
+      "learning_rate": 0.0002,
+      "loss": 0.7291,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7420580389620888,
+      "grad_norm": 0.6054152250289917,
+      "learning_rate": 0.0002,
+      "loss": 0.8038,
+      "step": 4590
+    },
+    {
+      "epoch": 0.743674723142834,
+      "grad_norm": 0.5574966669082642,
+      "learning_rate": 0.0002,
+      "loss": 0.8169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7452914073235793,
+      "grad_norm": 0.5395817160606384,
+      "learning_rate": 0.0002,
+      "loss": 0.8439,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7469080915043246,
+      "grad_norm": 0.7116472721099854,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7485247756850699,
+      "grad_norm": 0.5618700981140137,
+      "learning_rate": 0.0002,
+      "loss": 0.7743,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7501414598658153,
+      "grad_norm": 0.5802770853042603,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7517581440465605,
+      "grad_norm": 0.5690428018569946,
+      "learning_rate": 0.0002,
+      "loss": 0.7924,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7533748282273058,
+      "grad_norm": 0.4813360273838043,
+      "learning_rate": 0.0002,
+      "loss": 0.8017,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7549915124080511,
+      "grad_norm": 0.5434042811393738,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7566081965887964,
+      "grad_norm": 0.5502099990844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7582248807695416,
+      "grad_norm": 0.6020621061325073,
+      "learning_rate": 0.0002,
+      "loss": 0.8598,
+      "step": 4690
+    },
+    {
+      "epoch": 0.759841564950287,
+      "grad_norm": 0.4922301471233368,
+      "learning_rate": 0.0002,
+      "loss": 0.7937,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7614582491310322,
+      "grad_norm": 0.6492828726768494,
+      "learning_rate": 0.0002,
+      "loss": 0.788,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7630749333117776,
+      "grad_norm": 0.4865580201148987,
+      "learning_rate": 0.0002,
+      "loss": 0.8313,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7646916174925228,
+      "grad_norm": 0.5971422791481018,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7663083016732681,
+      "grad_norm": 0.6832674145698547,
+      "learning_rate": 0.0002,
+      "loss": 0.8298,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7679249858540134,
+      "grad_norm": 0.500908613204956,
+      "learning_rate": 0.0002,
+      "loss": 0.8156,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7695416700347587,
+      "grad_norm": 0.6112465858459473,
+      "learning_rate": 0.0002,
+      "loss": 0.8383,
+      "step": 4760
+    },
+    {
+      "epoch": 0.771158354215504,
+      "grad_norm": 0.5753506422042847,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7727750383962493,
+      "grad_norm": 0.6529405117034912,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7743917225769946,
+      "grad_norm": 0.5916843414306641,
+      "learning_rate": 0.0002,
+      "loss": 0.8171,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7760084067577399,
+      "grad_norm": 0.4821224510669708,
+      "learning_rate": 0.0002,
+      "loss": 0.83,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7776250909384852,
+      "grad_norm": 0.5532580018043518,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7792417751192304,
+      "grad_norm": 0.4604877233505249,
+      "learning_rate": 0.0002,
+      "loss": 0.7363,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7808584592999758,
+      "grad_norm": 0.5009613037109375,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 4830
+    },
+    {
+      "epoch": 0.782475143480721,
+      "grad_norm": 0.6448560357093811,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7840918276614663,
+      "grad_norm": 0.44327953457832336,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7857085118422116,
+      "grad_norm": 0.5355411171913147,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7873251960229569,
+      "grad_norm": 0.5635677576065063,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7889418802037023,
+      "grad_norm": 0.5417491793632507,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7905585643844475,
+      "grad_norm": 0.4567430913448334,
+      "learning_rate": 0.0002,
+      "loss": 0.7819,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7921752485651928,
+      "grad_norm": 0.44651296734809875,
+      "learning_rate": 0.0002,
+      "loss": 0.8454,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7937919327459381,
+      "grad_norm": 0.5741217136383057,
+      "learning_rate": 0.0002,
+      "loss": 0.7959,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7954086169266834,
+      "grad_norm": 0.6605045199394226,
+      "learning_rate": 0.0002,
+      "loss": 0.8093,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7970253011074286,
+      "grad_norm": 0.5126531720161438,
+      "learning_rate": 0.0002,
+      "loss": 0.77,
+      "step": 4930
+    },
+    {
+      "epoch": 0.798641985288174,
+      "grad_norm": 0.513648271560669,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8002586694689192,
+      "grad_norm": 0.5350404381752014,
+      "learning_rate": 0.0002,
+      "loss": 0.8314,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8018753536496646,
+      "grad_norm": 0.5731674432754517,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8034920378304098,
+      "grad_norm": 0.5974258184432983,
+      "learning_rate": 0.0002,
+      "loss": 0.8572,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8051087220111551,
+      "grad_norm": 0.8774799704551697,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8067254061919004,
+      "grad_norm": 0.5994430184364319,
+      "learning_rate": 0.0002,
+      "loss": 0.7899,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8083420903726457,
+      "grad_norm": 0.4894903004169464,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5000
+    },
+    {
+      "epoch": 0.809958774553391,
+      "grad_norm": 0.5218459367752075,
+      "learning_rate": 0.0002,
+      "loss": 0.78,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8115754587341363,
+      "grad_norm": 0.5232468843460083,
+      "learning_rate": 0.0002,
+      "loss": 0.817,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8131921429148816,
+      "grad_norm": 0.44358372688293457,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8148088270956269,
+      "grad_norm": 0.6202037334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8164255112763722,
+      "grad_norm": 0.7721474170684814,
+      "learning_rate": 0.0002,
+      "loss": 0.7351,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8180421954571174,
+      "grad_norm": 0.5568501353263855,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8196588796378628,
+      "grad_norm": 0.49148809909820557,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 5070
+    },
+    {
+      "epoch": 0.821275563818608,
+      "grad_norm": 0.4956012964248657,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8228922479993533,
+      "grad_norm": 0.6078833937644958,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8245089321800986,
+      "grad_norm": 0.46906954050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.828,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8261256163608439,
+      "grad_norm": 0.50812166929245,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8277423005415891,
+      "grad_norm": 0.5319661498069763,
+      "learning_rate": 0.0002,
+      "loss": 0.8243,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8293589847223345,
+      "grad_norm": 0.4949689209461212,
+      "learning_rate": 0.0002,
+      "loss": 0.7798,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8309756689030798,
+      "grad_norm": 0.5151591300964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8325923530838251,
+      "grad_norm": 0.5530214309692383,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8342090372645704,
+      "grad_norm": 0.6297410130500793,
+      "learning_rate": 0.0002,
+      "loss": 0.8251,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8358257214453156,
+      "grad_norm": 0.5466840267181396,
+      "learning_rate": 0.0002,
+      "loss": 0.8067,
+      "step": 5170
+    },
+    {
+      "epoch": 0.837442405626061,
+      "grad_norm": 0.652913510799408,
+      "learning_rate": 0.0002,
+      "loss": 0.7875,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8390590898068062,
+      "grad_norm": 0.5811293125152588,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8406757739875516,
+      "grad_norm": 0.5109550952911377,
+      "learning_rate": 0.0002,
+      "loss": 0.7412,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8422924581682968,
+      "grad_norm": 0.4551706612110138,
+      "learning_rate": 0.0002,
+      "loss": 0.8077,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8439091423490421,
+      "grad_norm": 0.5813754200935364,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8455258265297874,
+      "grad_norm": 0.5856947898864746,
+      "learning_rate": 0.0002,
+      "loss": 0.802,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8471425107105327,
+      "grad_norm": 0.5482739210128784,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8487591948912779,
+      "grad_norm": 0.49023720622062683,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8503758790720233,
+      "grad_norm": 0.49472475051879883,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8519925632527686,
+      "grad_norm": 0.5490226745605469,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8536092474335139,
+      "grad_norm": 0.5340665578842163,
+      "learning_rate": 0.0002,
+      "loss": 0.8333,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8552259316142592,
+      "grad_norm": 0.5962483882904053,
+      "learning_rate": 0.0002,
+      "loss": 0.8277,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8568426157950044,
+      "grad_norm": 0.586358368396759,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8584592999757498,
+      "grad_norm": 0.49120277166366577,
+      "learning_rate": 0.0002,
+      "loss": 0.7831,
+      "step": 5310
+    },
+    {
+      "epoch": 0.860075984156495,
+      "grad_norm": 0.5887332558631897,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8616926683372403,
+      "grad_norm": 0.42496153712272644,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.5489874482154846,
+      "learning_rate": 0.0002,
+      "loss": 0.7905,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8649260366987309,
+      "grad_norm": 0.5850813984870911,
+      "learning_rate": 0.0002,
+      "loss": 0.7958,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8665427208794761,
+      "grad_norm": 0.517487108707428,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8681594050602215,
+      "grad_norm": 0.5339142680168152,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8697760892409668,
+      "grad_norm": 0.6236387491226196,
+      "learning_rate": 0.0002,
+      "loss": 0.818,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8713927734217121,
+      "grad_norm": 0.5752192735671997,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8730094576024574,
+      "grad_norm": 0.6724614500999451,
+      "learning_rate": 0.0002,
+      "loss": 0.8542,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8746261417832026,
+      "grad_norm": 0.5280613303184509,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 5410
+    },
+    {
+      "epoch": 0.876242825963948,
+      "grad_norm": 0.44033288955688477,
+      "learning_rate": 0.0002,
+      "loss": 0.8231,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8778595101446932,
+      "grad_norm": 0.5199708342552185,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8794761943254386,
+      "grad_norm": 0.46778348088264465,
+      "learning_rate": 0.0002,
+      "loss": 0.7852,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8810928785061838,
+      "grad_norm": 0.4657754898071289,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8827095626869291,
+      "grad_norm": 0.5472902655601501,
+      "learning_rate": 0.0002,
+      "loss": 0.7799,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8843262468676744,
+      "grad_norm": 0.4876766800880432,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8859429310484197,
+      "grad_norm": 0.5057248473167419,
+      "learning_rate": 0.0002,
+      "loss": 0.7906,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8875596152291649,
+      "grad_norm": 0.4637320637702942,
+      "learning_rate": 0.0002,
+      "loss": 0.8124,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8891762994099103,
+      "grad_norm": 0.471955806016922,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8907929835906556,
+      "grad_norm": 0.5209813714027405,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8924096677714008,
+      "grad_norm": 0.6213834285736084,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8940263519521462,
+      "grad_norm": 0.5215408205986023,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8956430361328914,
+      "grad_norm": 0.580478310585022,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8972597203136368,
+      "grad_norm": 0.49102169275283813,
+      "learning_rate": 0.0002,
+      "loss": 0.8371,
+      "step": 5550
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.6043479442596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7806,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9004930886751273,
+      "grad_norm": 0.5636463165283203,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9021097728558726,
+      "grad_norm": 0.5620124340057373,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9037264570366179,
+      "grad_norm": 0.5206354856491089,
+      "learning_rate": 0.0002,
+      "loss": 0.8083,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9053431412173631,
+      "grad_norm": 0.5798229575157166,
+      "learning_rate": 0.0002,
+      "loss": 0.8557,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9069598253981085,
+      "grad_norm": 0.6428212523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.8097,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9085765095788537,
+      "grad_norm": 0.48064687848091125,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9101931937595991,
+      "grad_norm": 0.6347860097885132,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9118098779403444,
+      "grad_norm": 0.5353913307189941,
+      "learning_rate": 0.0002,
+      "loss": 0.851,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9134265621210896,
+      "grad_norm": 0.5323944091796875,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5650
+    },
+    {
+      "epoch": 0.915043246301835,
+      "grad_norm": 0.5261843204498291,
+      "learning_rate": 0.0002,
+      "loss": 0.8393,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9166599304825802,
+      "grad_norm": 0.5451326966285706,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9182766146633256,
+      "grad_norm": 0.5183324217796326,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9198932988440708,
+      "grad_norm": 0.47229018807411194,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9215099830248161,
+      "grad_norm": 0.49180513620376587,
+      "learning_rate": 0.0002,
+      "loss": 0.7757,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9231266672055614,
+      "grad_norm": 0.5419785380363464,
+      "learning_rate": 0.0002,
+      "loss": 0.8735,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9247433513863067,
+      "grad_norm": 0.5408698916435242,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9263600355670519,
+      "grad_norm": 0.5286232829093933,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9279767197477973,
+      "grad_norm": 0.7539758086204529,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9295934039285425,
+      "grad_norm": 0.5166944861412048,
+      "learning_rate": 0.0002,
+      "loss": 0.8118,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9312100881092878,
+      "grad_norm": 0.6601425409317017,
+      "learning_rate": 0.0002,
+      "loss": 0.783,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9328267722900332,
+      "grad_norm": 0.5029960870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9344434564707784,
+      "grad_norm": 0.4926645755767822,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9360601406515238,
+      "grad_norm": 0.5739615559577942,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5790
+    },
+    {
+      "epoch": 0.937676824832269,
+      "grad_norm": 0.5058279037475586,
+      "learning_rate": 0.0002,
+      "loss": 0.8037,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9392935090130143,
+      "grad_norm": 0.5260962247848511,
+      "learning_rate": 0.0002,
+      "loss": 0.8537,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9409101931937596,
+      "grad_norm": 0.5768588185310364,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9425268773745049,
+      "grad_norm": 0.5170126557350159,
+      "learning_rate": 0.0002,
+      "loss": 0.8215,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9441435615552501,
+      "grad_norm": 0.5745864510536194,
+      "learning_rate": 0.0002,
+      "loss": 0.7422,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9457602457359955,
+      "grad_norm": 0.5551357865333557,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9473769299167407,
+      "grad_norm": 0.5776078701019287,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9489936140974861,
+      "grad_norm": 0.5340062379837036,
+      "learning_rate": 0.0002,
+      "loss": 0.8527,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9506102982782314,
+      "grad_norm": 0.6447290182113647,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9522269824589766,
+      "grad_norm": 0.5123815536499023,
+      "learning_rate": 0.0002,
+      "loss": 0.7945,
+      "step": 5890
+    },
+    {
+      "epoch": 0.953843666639722,
+      "grad_norm": 0.48547613620758057,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9554603508204672,
+      "grad_norm": 0.5791414976119995,
+      "learning_rate": 0.0002,
+      "loss": 0.7896,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9570770350012126,
+      "grad_norm": 0.6195011734962463,
+      "learning_rate": 0.0002,
+      "loss": 0.8408,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9586937191819578,
+      "grad_norm": 0.6323803067207336,
+      "learning_rate": 0.0002,
+      "loss": 0.7805,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9603104033627031,
+      "grad_norm": 0.45552879571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9619270875434484,
+      "grad_norm": 0.5796473622322083,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9635437717241937,
+      "grad_norm": 0.647261381149292,
+      "learning_rate": 0.0002,
+      "loss": 0.7672,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9651604559049389,
+      "grad_norm": 0.5487682819366455,
+      "learning_rate": 0.0002,
+      "loss": 0.8086,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9667771400856843,
+      "grad_norm": 0.5743663907051086,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9683938242664295,
+      "grad_norm": 0.5470591187477112,
+      "learning_rate": 0.0002,
+      "loss": 0.8153,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9700105084471748,
+      "grad_norm": 0.5901660323143005,
+      "learning_rate": 0.0002,
+      "loss": 0.8119,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9716271926279202,
+      "grad_norm": 0.6544759273529053,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9732438768086654,
+      "grad_norm": 0.6288470029830933,
+      "learning_rate": 0.0002,
+      "loss": 0.7536,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9748605609894108,
+      "grad_norm": 0.673153817653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 6030
+    },
+    {
+      "epoch": 0.976477245170156,
+      "grad_norm": 0.42854753136634827,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9780939293509013,
+      "grad_norm": 0.5227066278457642,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9797106135316466,
+      "grad_norm": 0.5372416973114014,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9813272977123919,
+      "grad_norm": 0.6026402115821838,
+      "learning_rate": 0.0002,
+      "loss": 0.7591,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9829439818931371,
+      "grad_norm": 0.49547791481018066,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9845606660738825,
+      "grad_norm": 0.4641951322555542,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9861773502546277,
+      "grad_norm": 0.5818535089492798,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9877940344353731,
+      "grad_norm": 0.63955157995224,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9894107186161183,
+      "grad_norm": 0.5649438500404358,
+      "learning_rate": 0.0002,
+      "loss": 0.7547,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9910274027968636,
+      "grad_norm": 0.5290433168411255,
+      "learning_rate": 0.0002,
+      "loss": 0.7861,
+      "step": 6130
+    },
+    {
+      "epoch": 0.992644086977609,
+      "grad_norm": 0.6399374008178711,
+      "learning_rate": 0.0002,
+      "loss": 0.8109,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9942607711583542,
+      "grad_norm": 0.6736576557159424,
+      "learning_rate": 0.0002,
+      "loss": 0.8373,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9958774553390995,
+      "grad_norm": 0.515420138835907,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9974941395198448,
+      "grad_norm": 0.562677800655365,
+      "learning_rate": 0.0002,
+      "loss": 0.8032,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9991108237005901,
+      "grad_norm": 0.7113858461380005,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9999191657909627,
+      "eval_loss": 1.0871200561523438,
+      "eval_runtime": 122.2071,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0007275078813354,
+      "grad_norm": 0.7111801505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0023441920620806,
+      "grad_norm": 0.5402125716209412,
+      "learning_rate": 0.0002,
+      "loss": 0.6865,
+      "step": 6200
+    },
+    {
+      "epoch": 1.003960876242826,
+      "grad_norm": 0.6098830103874207,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0055775604235713,
+      "grad_norm": 0.5829983353614807,
+      "learning_rate": 0.0002,
+      "loss": 0.7631,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0071942446043165,
+      "grad_norm": 0.5614621043205261,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0088109287850617,
+      "grad_norm": 0.5954238772392273,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0104276129658072,
+      "grad_norm": 0.6480574607849121,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0120442971465524,
+      "grad_norm": 0.6051128506660461,
+      "learning_rate": 0.0002,
+      "loss": 0.7514,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0136609813272976,
+      "grad_norm": 0.6318870782852173,
+      "learning_rate": 0.0002,
+      "loss": 0.7237,
+      "step": 6270
+    },
+    {
+      "epoch": 1.015277665508043,
+      "grad_norm": 0.5048980116844177,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0168943496887883,
+      "grad_norm": 0.6346936225891113,
+      "learning_rate": 0.0002,
+      "loss": 0.7391,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0185110338695336,
+      "grad_norm": 0.5711665749549866,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0201277180502788,
+      "grad_norm": 0.5175361037254333,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0217444022310243,
+      "grad_norm": 0.5360831618309021,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0233610864117695,
+      "grad_norm": 0.614675760269165,
+      "learning_rate": 0.0002,
+      "loss": 0.7112,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0249777705925147,
+      "grad_norm": 0.5626118183135986,
+      "learning_rate": 0.0002,
+      "loss": 0.7748,
+      "step": 6340
+    },
+    {
+      "epoch": 1.02659445477326,
+      "grad_norm": 0.574897289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.7375,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0282111389540054,
+      "grad_norm": 0.7185447812080383,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0298278231347506,
+      "grad_norm": 0.6705799698829651,
+      "learning_rate": 0.0002,
+      "loss": 0.703,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0314445073154959,
+      "grad_norm": 0.6740428805351257,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0330611914962413,
+      "grad_norm": 0.663902759552002,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0346778756769865,
+      "grad_norm": 0.5029543042182922,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0362945598577318,
+      "grad_norm": 0.7813863158226013,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 6410
+    },
+    {
+      "epoch": 1.037911244038477,
+      "grad_norm": 0.5396282076835632,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0395279282192225,
+      "grad_norm": 0.5253293514251709,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0411446123999677,
+      "grad_norm": 0.7236770987510681,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 6440
+    },
+    {
+      "epoch": 1.042761296580713,
+      "grad_norm": 0.5670917630195618,
+      "learning_rate": 0.0002,
+      "loss": 0.7259,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0443779807614582,
+      "grad_norm": 0.6031978726387024,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0459946649422036,
+      "grad_norm": 0.5309213399887085,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0476113491229488,
+      "grad_norm": 0.7114651799201965,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 6480
+    },
+    {
+      "epoch": 1.049228033303694,
+      "grad_norm": 0.5591610670089722,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0508447174844395,
+      "grad_norm": 0.5185961127281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0524614016651848,
+      "grad_norm": 0.6510552167892456,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6510
+    },
+    {
+      "epoch": 1.05407808584593,
+      "grad_norm": 0.6557928919792175,
+      "learning_rate": 0.0002,
+      "loss": 0.7057,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0556947700266752,
+      "grad_norm": 0.6973192691802979,
+      "learning_rate": 0.0002,
+      "loss": 0.8056,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0573114542074207,
+      "grad_norm": 0.6226583123207092,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 6540
+    },
+    {
+      "epoch": 1.058928138388166,
+      "grad_norm": 0.5633195638656616,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0605448225689111,
+      "grad_norm": 0.7466658353805542,
+      "learning_rate": 0.0002,
+      "loss": 0.7082,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0621615067496564,
+      "grad_norm": 0.6462772488594055,
+      "learning_rate": 0.0002,
+      "loss": 0.7059,
+      "step": 6570
+    },
+    {
+      "epoch": 1.0637781909304018,
+      "grad_norm": 0.5266856551170349,
+      "learning_rate": 0.0002,
+      "loss": 0.7046,
+      "step": 6580
+    },
+    {
+      "epoch": 1.065394875111147,
+      "grad_norm": 0.534392774105072,
+      "learning_rate": 0.0002,
+      "loss": 0.7157,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0670115592918923,
+      "grad_norm": 0.7514177560806274,
+      "learning_rate": 0.0002,
+      "loss": 0.7115,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0686282434726375,
+      "grad_norm": 0.7593035697937012,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 6610
+    },
+    {
+      "epoch": 1.070244927653383,
+      "grad_norm": 0.5277858972549438,
+      "learning_rate": 0.0002,
+      "loss": 0.6836,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0718616118341282,
+      "grad_norm": 0.5573670268058777,
+      "learning_rate": 0.0002,
+      "loss": 0.7405,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0734782960148734,
+      "grad_norm": 0.6802396774291992,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 6640
+    },
+    {
+      "epoch": 1.0750949801956189,
+      "grad_norm": 0.7367215752601624,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0767116643763641,
+      "grad_norm": 0.5961891412734985,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0783283485571094,
+      "grad_norm": 0.5736313462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0799450327378546,
+      "grad_norm": 0.619219183921814,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0815617169186,
+      "grad_norm": 0.6214390993118286,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0831784010993453,
+      "grad_norm": 0.564536988735199,
+      "learning_rate": 0.0002,
+      "loss": 0.7554,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0847950852800905,
+      "grad_norm": 0.5838140249252319,
+      "learning_rate": 0.0002,
+      "loss": 0.7362,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0864117694608357,
+      "grad_norm": 0.7000553607940674,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0880284536415812,
+      "grad_norm": 0.7078263759613037,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 6730
+    },
+    {
+      "epoch": 1.0896451378223264,
+      "grad_norm": 0.8353848457336426,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 6740
+    },
+    {
+      "epoch": 1.0912618220030716,
+      "grad_norm": 0.5615518689155579,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 6750
+    },
+    {
+      "epoch": 1.0928785061838169,
+      "grad_norm": 0.5475581288337708,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 6760
+    },
+    {
+      "epoch": 1.0944951903645623,
+      "grad_norm": 0.5835978388786316,
+      "learning_rate": 0.0002,
+      "loss": 0.7652,
+      "step": 6770
+    },
+    {
+      "epoch": 1.0961118745453076,
+      "grad_norm": 0.5516105890274048,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 6780
+    },
+    {
+      "epoch": 1.0977285587260528,
+      "grad_norm": 0.5875251889228821,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 6790
+    },
+    {
+      "epoch": 1.0993452429067982,
+      "grad_norm": 0.7376947999000549,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1009619270875435,
+      "grad_norm": 0.5656165480613708,
+      "learning_rate": 0.0002,
+      "loss": 0.7512,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1025786112682887,
+      "grad_norm": 0.6365954279899597,
+      "learning_rate": 0.0002,
+      "loss": 0.7409,
+      "step": 6820
+    },
+    {
+      "epoch": 1.104195295449034,
+      "grad_norm": 0.5033080577850342,
+      "learning_rate": 0.0002,
+      "loss": 0.7392,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1058119796297794,
+      "grad_norm": 0.617396891117096,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1074286638105246,
+      "grad_norm": 0.6395374536514282,
+      "learning_rate": 0.0002,
+      "loss": 0.7006,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1090453479912699,
+      "grad_norm": 0.6775295734405518,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1106620321720153,
+      "grad_norm": 0.6655223965644836,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1122787163527605,
+      "grad_norm": 0.676655113697052,
+      "learning_rate": 0.0002,
+      "loss": 0.7553,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1138954005335058,
+      "grad_norm": 0.6062718629837036,
+      "learning_rate": 0.0002,
+      "loss": 0.7342,
+      "step": 6890
+    },
+    {
+      "epoch": 1.115512084714251,
+      "grad_norm": 0.590943455696106,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1171287688949965,
+      "grad_norm": 0.6315317153930664,
+      "learning_rate": 0.0002,
+      "loss": 0.6705,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1187454530757417,
+      "grad_norm": 0.47979024052619934,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 6920
+    },
+    {
+      "epoch": 1.120362137256487,
+      "grad_norm": 0.647298276424408,
+      "learning_rate": 0.0002,
+      "loss": 0.7002,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1219788214372322,
+      "grad_norm": 0.7336484789848328,
+      "learning_rate": 0.0002,
+      "loss": 0.7502,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1235955056179776,
+      "grad_norm": 0.5071424245834351,
+      "learning_rate": 0.0002,
+      "loss": 0.693,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1252121897987228,
+      "grad_norm": 0.6527144312858582,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 6960
+    },
+    {
+      "epoch": 1.126828873979468,
+      "grad_norm": 0.6935935020446777,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1284455581602133,
+      "grad_norm": 0.8026931881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1300622423409588,
+      "grad_norm": 0.5210393667221069,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 6990
+    },
+    {
+      "epoch": 1.131678926521704,
+      "grad_norm": 0.60475093126297,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1332956107024492,
+      "grad_norm": 0.6417073607444763,
+      "learning_rate": 0.0002,
+      "loss": 0.7495,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1349122948831947,
+      "grad_norm": 0.6732175946235657,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 7020
+    },
+    {
+      "epoch": 1.13652897906394,
+      "grad_norm": 0.6719491481781006,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1381456632446851,
+      "grad_norm": 0.5708295106887817,
+      "learning_rate": 0.0002,
+      "loss": 0.7694,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1397623474254304,
+      "grad_norm": 0.7141719460487366,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1413790316061758,
+      "grad_norm": 0.6187017560005188,
+      "learning_rate": 0.0002,
+      "loss": 0.764,
+      "step": 7060
+    },
+    {
+      "epoch": 1.142995715786921,
+      "grad_norm": 0.50581294298172,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 7070
+    },
+    {
+      "epoch": 1.1446123999676663,
+      "grad_norm": 0.5620143413543701,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1462290841484115,
+      "grad_norm": 0.6231929659843445,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 7090
+    },
+    {
+      "epoch": 1.147845768329157,
+      "grad_norm": 0.5775774121284485,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1494624525099022,
+      "grad_norm": 0.6492809653282166,
+      "learning_rate": 0.0002,
+      "loss": 0.7728,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1510791366906474,
+      "grad_norm": 0.6434972286224365,
+      "learning_rate": 0.0002,
+      "loss": 0.7545,
+      "step": 7120
+    },
+    {
+      "epoch": 1.1526958208713927,
+      "grad_norm": 0.6191812753677368,
+      "learning_rate": 0.0002,
+      "loss": 0.7374,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1543125050521381,
+      "grad_norm": 0.6690331697463989,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 7140
+    },
+    {
+      "epoch": 1.1559291892328833,
+      "grad_norm": 0.5977938175201416,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1575458734136286,
+      "grad_norm": 0.6195854544639587,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7160
+    },
+    {
+      "epoch": 1.159162557594374,
+      "grad_norm": 0.5752048492431641,
+      "learning_rate": 0.0002,
+      "loss": 0.7249,
+      "step": 7170
+    },
+    {
+      "epoch": 1.1607792417751193,
+      "grad_norm": 0.589081883430481,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1623959259558645,
+      "grad_norm": 0.756996750831604,
+      "learning_rate": 0.0002,
+      "loss": 0.704,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1640126101366097,
+      "grad_norm": 0.7614967226982117,
+      "learning_rate": 0.0002,
+      "loss": 0.7404,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1656292943173552,
+      "grad_norm": 0.6120437979698181,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1672459784981004,
+      "grad_norm": 0.6210004687309265,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1688626626788456,
+      "grad_norm": 0.6044116020202637,
+      "learning_rate": 0.0002,
+      "loss": 0.7251,
+      "step": 7230
+    },
+    {
+      "epoch": 1.170479346859591,
+      "grad_norm": 0.5418457388877869,
+      "learning_rate": 0.0002,
+      "loss": 0.7361,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1720960310403363,
+      "grad_norm": 0.6413537263870239,
+      "learning_rate": 0.0002,
+      "loss": 0.6938,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1737127152210816,
+      "grad_norm": 0.5777867436408997,
+      "learning_rate": 0.0002,
+      "loss": 0.6978,
+      "step": 7260
+    },
+    {
+      "epoch": 1.1753293994018268,
+      "grad_norm": 0.7092402577400208,
+      "learning_rate": 0.0002,
+      "loss": 0.7503,
+      "step": 7270
+    },
+    {
+      "epoch": 1.176946083582572,
+      "grad_norm": 0.6351709365844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1785627677633175,
+      "grad_norm": 0.6172189712524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7527,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1801794519440627,
+      "grad_norm": 0.6801714897155762,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7300
+    },
+    {
+      "epoch": 1.181796136124808,
+      "grad_norm": 0.6044712066650391,
+      "learning_rate": 0.0002,
+      "loss": 0.6941,
+      "step": 7310
+    },
+    {
+      "epoch": 1.1834128203055534,
+      "grad_norm": 0.7413212060928345,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 7320
+    },
+    {
+      "epoch": 1.1850295044862986,
+      "grad_norm": 0.5303856134414673,
+      "learning_rate": 0.0002,
+      "loss": 0.7396,
+      "step": 7330
+    },
+    {
+      "epoch": 1.1866461886670439,
+      "grad_norm": 0.5647098422050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6915,
+      "step": 7340
+    },
+    {
+      "epoch": 1.188262872847789,
+      "grad_norm": 0.7374135255813599,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 7350
+    },
+    {
+      "epoch": 1.1898795570285345,
+      "grad_norm": 0.5710089206695557,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1914962412092798,
+      "grad_norm": 0.6073619723320007,
+      "learning_rate": 0.0002,
+      "loss": 0.8289,
+      "step": 7370
+    },
+    {
+      "epoch": 1.193112925390025,
+      "grad_norm": 0.5899916887283325,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1947296095707705,
+      "grad_norm": 0.7762434482574463,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1963462937515157,
+      "grad_norm": 0.679949939250946,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 7400
+    },
+    {
+      "epoch": 1.197962977932261,
+      "grad_norm": 0.6106849312782288,
+      "learning_rate": 0.0002,
+      "loss": 0.7599,
+      "step": 7410
+    },
+    {
+      "epoch": 1.1995796621130062,
+      "grad_norm": 0.682461678981781,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2011963462937516,
+      "grad_norm": 0.6087017059326172,
+      "learning_rate": 0.0002,
+      "loss": 0.7741,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2028130304744968,
+      "grad_norm": 0.63739013671875,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 7440
+    },
+    {
+      "epoch": 1.204429714655242,
+      "grad_norm": 0.6154777407646179,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2060463988359873,
+      "grad_norm": 0.7491534948348999,
+      "learning_rate": 0.0002,
+      "loss": 0.7565,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2076630830167328,
+      "grad_norm": 0.6664797067642212,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 7470
+    },
+    {
+      "epoch": 1.209279767197478,
+      "grad_norm": 0.6660266518592834,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2108964513782232,
+      "grad_norm": 0.6972551345825195,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2125131355589684,
+      "grad_norm": 0.6157945990562439,
+      "learning_rate": 0.0002,
+      "loss": 0.7023,
+      "step": 7500
+    },
+    {
+      "epoch": 1.214129819739714,
+      "grad_norm": 0.5199310183525085,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2157465039204591,
+      "grad_norm": 0.577610433101654,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2173631881012044,
+      "grad_norm": 0.53652423620224,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2189798722819498,
+      "grad_norm": 0.6479050517082214,
+      "learning_rate": 0.0002,
+      "loss": 0.7393,
+      "step": 7540
+    },
+    {
+      "epoch": 1.220596556462695,
+      "grad_norm": 0.618748128414154,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2222132406434403,
+      "grad_norm": 0.6311424374580383,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2238299248241855,
+      "grad_norm": 0.6595825552940369,
+      "learning_rate": 0.0002,
+      "loss": 0.7272,
+      "step": 7570
+    },
+    {
+      "epoch": 1.225446609004931,
+      "grad_norm": 0.5198960900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2270632931856762,
+      "grad_norm": 0.578650712966919,
+      "learning_rate": 0.0002,
+      "loss": 0.674,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2286799773664214,
+      "grad_norm": 0.6080220937728882,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2302966615471669,
+      "grad_norm": 0.7050248384475708,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2319133457279121,
+      "grad_norm": 0.6652196049690247,
+      "learning_rate": 0.0002,
+      "loss": 0.7032,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2335300299086573,
+      "grad_norm": 0.7322776317596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7085,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2351467140894026,
+      "grad_norm": 0.4998728036880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2367633982701478,
+      "grad_norm": 0.6428788900375366,
+      "learning_rate": 0.0002,
+      "loss": 0.7214,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2383800824508933,
+      "grad_norm": 0.585242509841919,
+      "learning_rate": 0.0002,
+      "loss": 0.7699,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2399967666316385,
+      "grad_norm": 0.5211917757987976,
+      "learning_rate": 0.0002,
+      "loss": 0.7621,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2416134508123837,
+      "grad_norm": 0.6490384340286255,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2432301349931292,
+      "grad_norm": 0.6249763369560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2448468191738744,
+      "grad_norm": 0.71870356798172,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2464635033546196,
+      "grad_norm": 0.6761967539787292,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2480801875353649,
+      "grad_norm": 0.6500617265701294,
+      "learning_rate": 0.0002,
+      "loss": 0.7501,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2496968717161103,
+      "grad_norm": 0.8069869875907898,
+      "learning_rate": 0.0002,
+      "loss": 0.7903,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2513135558968556,
+      "grad_norm": 0.6044608950614929,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2529302400776008,
+      "grad_norm": 0.6573283076286316,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2545469242583462,
+      "grad_norm": 0.625430166721344,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2561636084390915,
+      "grad_norm": 0.5442022681236267,
+      "learning_rate": 0.0002,
+      "loss": 0.7041,
+      "step": 7770
+    },
+    {
+      "epoch": 1.2577802926198367,
+      "grad_norm": 0.6818386912345886,
+      "learning_rate": 0.0002,
+      "loss": 0.7172,
+      "step": 7780
+    },
+    {
+      "epoch": 1.259396976800582,
+      "grad_norm": 0.6381874084472656,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2610136609813272,
+      "grad_norm": 0.6269212961196899,
+      "learning_rate": 0.0002,
+      "loss": 0.6834,
+      "step": 7800
+    },
+    {
+      "epoch": 1.2626303451620726,
+      "grad_norm": 0.600121259689331,
+      "learning_rate": 0.0002,
+      "loss": 0.7821,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2642470293428179,
+      "grad_norm": 0.6337703466415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7761,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2658637135235633,
+      "grad_norm": 0.7234963774681091,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 7830
+    },
+    {
+      "epoch": 1.2674803977043085,
+      "grad_norm": 0.800184965133667,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2690970818850538,
+      "grad_norm": 0.7539464831352234,
+      "learning_rate": 0.0002,
+      "loss": 0.7426,
+      "step": 7850
+    },
+    {
+      "epoch": 1.270713766065799,
+      "grad_norm": 0.5493760704994202,
+      "learning_rate": 0.0002,
+      "loss": 0.7496,
+      "step": 7860
+    },
+    {
+      "epoch": 1.2723304502465442,
+      "grad_norm": 0.7477145791053772,
+      "learning_rate": 0.0002,
+      "loss": 0.7537,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2739471344272897,
+      "grad_norm": 0.6366362571716309,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 7880
+    },
+    {
+      "epoch": 1.275563818608035,
+      "grad_norm": 0.7419533729553223,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2771805027887801,
+      "grad_norm": 0.6141223311424255,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 7900
+    },
+    {
+      "epoch": 1.2787971869695256,
+      "grad_norm": 0.7522598505020142,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 7910
+    },
+    {
+      "epoch": 1.2804138711502708,
+      "grad_norm": 0.6935804486274719,
+      "learning_rate": 0.0002,
+      "loss": 0.7097,
+      "step": 7920
+    },
+    {
+      "epoch": 1.282030555331016,
+      "grad_norm": 0.7239290475845337,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 7930
+    },
+    {
+      "epoch": 1.2836472395117613,
+      "grad_norm": 0.8800187110900879,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 7940
+    },
+    {
+      "epoch": 1.2852639236925067,
+      "grad_norm": 0.540458083152771,
+      "learning_rate": 0.0002,
+      "loss": 0.6991,
+      "step": 7950
+    },
+    {
+      "epoch": 1.286880607873252,
+      "grad_norm": 0.6492934226989746,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 7960
+    },
+    {
+      "epoch": 1.2884972920539972,
+      "grad_norm": 0.6543959379196167,
+      "learning_rate": 0.0002,
+      "loss": 0.7742,
+      "step": 7970
+    },
+    {
+      "epoch": 1.2901139762347427,
+      "grad_norm": 0.5804705619812012,
+      "learning_rate": 0.0002,
+      "loss": 0.7316,
+      "step": 7980
+    },
+    {
+      "epoch": 1.291730660415488,
+      "grad_norm": 0.7074727416038513,
+      "learning_rate": 0.0002,
+      "loss": 0.796,
+      "step": 7990
+    },
+    {
+      "epoch": 1.2933473445962331,
+      "grad_norm": 0.5347974300384521,
+      "learning_rate": 0.0002,
+      "loss": 0.7034,
+      "step": 8000
+    },
+    {
+      "epoch": 1.2949640287769784,
+      "grad_norm": 0.6457298398017883,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 8010
+    },
+    {
+      "epoch": 1.2965807129577236,
+      "grad_norm": 0.6407219171524048,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8020
+    },
+    {
+      "epoch": 1.298197397138469,
+      "grad_norm": 0.828439474105835,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2998140813192143,
+      "grad_norm": 0.4840380549430847,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3014307654999595,
+      "grad_norm": 0.5921024680137634,
+      "learning_rate": 0.0002,
+      "loss": 0.7283,
+      "step": 8050
+    },
+    {
+      "epoch": 1.303047449680705,
+      "grad_norm": 0.6170315146446228,
+      "learning_rate": 0.0002,
+      "loss": 0.7477,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3046641338614502,
+      "grad_norm": 0.5374847054481506,
+      "learning_rate": 0.0002,
+      "loss": 0.7534,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3062808180421954,
+      "grad_norm": 0.545758068561554,
+      "learning_rate": 0.0002,
+      "loss": 0.7593,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3078975022229407,
+      "grad_norm": 0.55641770362854,
+      "learning_rate": 0.0002,
+      "loss": 0.7463,
+      "step": 8090
+    },
+    {
+      "epoch": 1.309514186403686,
+      "grad_norm": 0.6724897027015686,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3111308705844313,
+      "grad_norm": 0.6923972368240356,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3127475547651766,
+      "grad_norm": 0.5136841535568237,
+      "learning_rate": 0.0002,
+      "loss": 0.7149,
+      "step": 8120
+    },
+    {
+      "epoch": 1.314364238945922,
+      "grad_norm": 0.6766283512115479,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3159809231266673,
+      "grad_norm": 0.6283926367759705,
+      "learning_rate": 0.0002,
+      "loss": 0.7489,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3175976073074125,
+      "grad_norm": 0.644216001033783,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3192142914881577,
+      "grad_norm": 0.7827503085136414,
+      "learning_rate": 0.0002,
+      "loss": 0.7125,
+      "step": 8160
+    },
+    {
+      "epoch": 1.320830975668903,
+      "grad_norm": 0.6651390790939331,
+      "learning_rate": 0.0002,
+      "loss": 0.7271,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3224476598496484,
+      "grad_norm": 0.5547412633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7778,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3240643440303936,
+      "grad_norm": 0.6765179634094238,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 8190
+    },
+    {
+      "epoch": 1.325681028211139,
+      "grad_norm": 0.6822077035903931,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3272977123918843,
+      "grad_norm": 0.5941002368927002,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3289143965726296,
+      "grad_norm": 0.4850037097930908,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3305310807533748,
+      "grad_norm": 0.6162990927696228,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 8230
+    },
+    {
+      "epoch": 1.33214776493412,
+      "grad_norm": 0.6665613651275635,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3337644491148655,
+      "grad_norm": 0.618192732334137,
+      "learning_rate": 0.0002,
+      "loss": 0.759,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3353811332956107,
+      "grad_norm": 0.710418701171875,
+      "learning_rate": 0.0002,
+      "loss": 0.7532,
+      "step": 8260
+    },
+    {
+      "epoch": 1.336997817476356,
+      "grad_norm": 0.5109876990318298,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3386145016571014,
+      "grad_norm": 0.6791711449623108,
+      "learning_rate": 0.0002,
+      "loss": 0.7303,
+      "step": 8280
+    },
+    {
+      "epoch": 1.3402311858378466,
+      "grad_norm": 0.6836432814598083,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3418478700185918,
+      "grad_norm": 0.5579386353492737,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 8300
+    },
+    {
+      "epoch": 1.343464554199337,
+      "grad_norm": 0.6713546514511108,
+      "learning_rate": 0.0002,
+      "loss": 0.7377,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3450812383800825,
+      "grad_norm": 0.5353720188140869,
+      "learning_rate": 0.0002,
+      "loss": 0.7756,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3466979225608278,
+      "grad_norm": 0.5813682675361633,
+      "learning_rate": 0.0002,
+      "loss": 0.718,
+      "step": 8330
+    },
+    {
+      "epoch": 1.348314606741573,
+      "grad_norm": 0.8158791661262512,
+      "learning_rate": 0.0002,
+      "loss": 0.7294,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3499312909223184,
+      "grad_norm": 0.6193785071372986,
+      "learning_rate": 0.0002,
+      "loss": 0.6992,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3515479751030637,
+      "grad_norm": 0.6353939771652222,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 8360
+    },
+    {
+      "epoch": 1.353164659283809,
+      "grad_norm": 0.6925048232078552,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 8370
+    },
+    {
+      "epoch": 1.3547813434645541,
+      "grad_norm": 0.988264799118042,
+      "learning_rate": 0.0002,
+      "loss": 0.736,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3563980276452994,
+      "grad_norm": 0.6476002931594849,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3580147118260448,
+      "grad_norm": 0.7120398879051208,
+      "learning_rate": 0.0002,
+      "loss": 0.776,
+      "step": 8400
+    },
+    {
+      "epoch": 1.35963139600679,
+      "grad_norm": 0.9048416614532471,
+      "learning_rate": 0.0002,
+      "loss": 0.7368,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3612480801875353,
+      "grad_norm": 0.7000672817230225,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3628647643682807,
+      "grad_norm": 0.6015632152557373,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 8430
+    },
+    {
+      "epoch": 1.364481448549026,
+      "grad_norm": 0.612516462802887,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3660981327297712,
+      "grad_norm": 0.5969301462173462,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 8450
+    },
+    {
+      "epoch": 1.3677148169105164,
+      "grad_norm": 0.6730654239654541,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 8460
+    },
+    {
+      "epoch": 1.369331501091262,
+      "grad_norm": 0.6386392116546631,
+      "learning_rate": 0.0002,
+      "loss": 0.7465,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3709481852720071,
+      "grad_norm": 0.739544153213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3725648694527524,
+      "grad_norm": 0.6462782621383667,
+      "learning_rate": 0.0002,
+      "loss": 0.7892,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3741815536334978,
+      "grad_norm": 0.7346843481063843,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 8500
+    },
+    {
+      "epoch": 1.375798237814243,
+      "grad_norm": 0.6884821057319641,
+      "learning_rate": 0.0002,
+      "loss": 0.7634,
+      "step": 8510
+    },
+    {
+      "epoch": 1.3774149219949883,
+      "grad_norm": 0.6999333500862122,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3790316061757335,
+      "grad_norm": 0.5378713011741638,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 8530
+    },
+    {
+      "epoch": 1.3806482903564787,
+      "grad_norm": 0.5417906641960144,
+      "learning_rate": 0.0002,
+      "loss": 0.6797,
+      "step": 8540
+    },
+    {
+      "epoch": 1.3822649745372242,
+      "grad_norm": 0.6602526307106018,
+      "learning_rate": 0.0002,
+      "loss": 0.7499,
+      "step": 8550
+    },
+    {
+      "epoch": 1.3838816587179694,
+      "grad_norm": 0.7073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.7356,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3854983428987149,
+      "grad_norm": 0.5841707587242126,
+      "learning_rate": 0.0002,
+      "loss": 0.75,
+      "step": 8570
+    },
+    {
+      "epoch": 1.38711502707946,
+      "grad_norm": 0.7031095027923584,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 8580
+    },
+    {
+      "epoch": 1.3887317112602053,
+      "grad_norm": 0.5198570489883423,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 8590
+    },
+    {
+      "epoch": 1.3903483954409506,
+      "grad_norm": 0.7261320352554321,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 8600
+    },
+    {
+      "epoch": 1.3919650796216958,
+      "grad_norm": 0.5616350173950195,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3935817638024413,
+      "grad_norm": 0.5185914635658264,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 8620
+    },
+    {
+      "epoch": 1.3951984479831865,
+      "grad_norm": 0.5814694762229919,
+      "learning_rate": 0.0002,
+      "loss": 0.7456,
+      "step": 8630
+    },
+    {
+      "epoch": 1.3968151321639317,
+      "grad_norm": 0.6977371573448181,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 8640
+    },
+    {
+      "epoch": 1.3984318163446772,
+      "grad_norm": 0.6855689883232117,
+      "learning_rate": 0.0002,
+      "loss": 0.7574,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4000485005254224,
+      "grad_norm": 0.5414357781410217,
+      "learning_rate": 0.0002,
+      "loss": 0.7802,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4016651847061676,
+      "grad_norm": 0.6970012784004211,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4032818688869129,
+      "grad_norm": 0.526079535484314,
+      "learning_rate": 0.0002,
+      "loss": 0.7421,
+      "step": 8680
+    },
+    {
+      "epoch": 1.404898553067658,
+      "grad_norm": 0.758712887763977,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4065152372484035,
+      "grad_norm": 0.7118762731552124,
+      "learning_rate": 0.0002,
+      "loss": 0.7612,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4081319214291488,
+      "grad_norm": 0.5696909427642822,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4097486056098942,
+      "grad_norm": 0.7995436787605286,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4113652897906395,
+      "grad_norm": 0.7237521409988403,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 8730
+    },
+    {
+      "epoch": 1.4129819739713847,
+      "grad_norm": 0.744628369808197,
+      "learning_rate": 0.0002,
+      "loss": 0.7661,
+      "step": 8740
+    },
+    {
+      "epoch": 1.41459865815213,
+      "grad_norm": 0.6082926988601685,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4162153423328752,
+      "grad_norm": 0.5185243487358093,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4178320265136206,
+      "grad_norm": 0.5183082222938538,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4194487106943658,
+      "grad_norm": 0.7326041460037231,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 8780
+    },
+    {
+      "epoch": 1.421065394875111,
+      "grad_norm": 0.7174660563468933,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4226820790558565,
+      "grad_norm": 0.8080165982246399,
+      "learning_rate": 0.0002,
+      "loss": 0.7507,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4242987632366018,
+      "grad_norm": 0.5061507821083069,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 8810
+    },
+    {
+      "epoch": 1.425915447417347,
+      "grad_norm": 0.801602840423584,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4275321315980922,
+      "grad_norm": 0.6150273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.7287,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4291488157788377,
+      "grad_norm": 0.8786525726318359,
+      "learning_rate": 0.0002,
+      "loss": 0.7452,
+      "step": 8840
+    },
+    {
+      "epoch": 1.430765499959583,
+      "grad_norm": 0.6371538639068604,
+      "learning_rate": 0.0002,
+      "loss": 0.7257,
+      "step": 8850
+    },
+    {
+      "epoch": 1.4323821841403281,
+      "grad_norm": 0.6409295797348022,
+      "learning_rate": 0.0002,
+      "loss": 0.711,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4339988683210736,
+      "grad_norm": 0.6452359557151794,
+      "learning_rate": 0.0002,
+      "loss": 0.7891,
+      "step": 8870
+    },
+    {
+      "epoch": 1.4356155525018188,
+      "grad_norm": 0.5842334628105164,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 8880
+    },
+    {
+      "epoch": 1.437232236682564,
+      "grad_norm": 0.696761965751648,
+      "learning_rate": 0.0002,
+      "loss": 0.7446,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4388489208633093,
+      "grad_norm": 0.6384600400924683,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4404656050440545,
+      "grad_norm": 0.5981136560440063,
+      "learning_rate": 0.0002,
+      "loss": 0.7049,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4420822892248,
+      "grad_norm": 0.6355637907981873,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4436989734055452,
+      "grad_norm": 0.6374830603599548,
+      "learning_rate": 0.0002,
+      "loss": 0.7653,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4453156575862904,
+      "grad_norm": 0.559013307094574,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 8940
+    },
+    {
+      "epoch": 1.446932341767036,
+      "grad_norm": 0.7289170026779175,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4485490259477811,
+      "grad_norm": 0.8649206757545471,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4501657101285264,
+      "grad_norm": 0.7664689421653748,
+      "learning_rate": 0.0002,
+      "loss": 0.7745,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4517823943092716,
+      "grad_norm": 0.7109952569007874,
+      "learning_rate": 0.0002,
+      "loss": 0.7431,
+      "step": 8980
+    },
+    {
+      "epoch": 1.453399078490017,
+      "grad_norm": 0.6312844753265381,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4550157626707623,
+      "grad_norm": 0.6616617441177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7467,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4566324468515075,
+      "grad_norm": 0.7384068965911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7518,
+      "step": 9010
+    },
+    {
+      "epoch": 1.458249131032253,
+      "grad_norm": 0.6549670100212097,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9020
+    },
+    {
+      "epoch": 1.4598658152129982,
+      "grad_norm": 0.6254119277000427,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4614824993937434,
+      "grad_norm": 0.6806328892707825,
+      "learning_rate": 0.0002,
+      "loss": 0.7645,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4630991835744886,
+      "grad_norm": 0.6803115010261536,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4647158677552339,
+      "grad_norm": 0.48529282212257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7264,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4663325519359793,
+      "grad_norm": 0.5995030999183655,
+      "learning_rate": 0.0002,
+      "loss": 0.7542,
+      "step": 9070
+    },
+    {
+      "epoch": 1.4679492361167246,
+      "grad_norm": 0.6005427837371826,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9080
+    },
+    {
+      "epoch": 1.46956592029747,
+      "grad_norm": 0.718564510345459,
+      "learning_rate": 0.0002,
+      "loss": 0.7288,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4711826044782153,
+      "grad_norm": 0.7003577351570129,
+      "learning_rate": 0.0002,
+      "loss": 0.7089,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4727992886589605,
+      "grad_norm": 0.5888323783874512,
+      "learning_rate": 0.0002,
+      "loss": 0.8069,
+      "step": 9110
+    },
+    {
+      "epoch": 1.4744159728397057,
+      "grad_norm": 0.6417609453201294,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 9120
+    },
+    {
+      "epoch": 1.476032657020451,
+      "grad_norm": 0.572294294834137,
+      "learning_rate": 0.0002,
+      "loss": 0.7441,
+      "step": 9130
+    },
+    {
+      "epoch": 1.4776493412011964,
+      "grad_norm": 0.8200714588165283,
+      "learning_rate": 0.0002,
+      "loss": 0.8053,
+      "step": 9140
+    },
+    {
+      "epoch": 1.4792660253819416,
+      "grad_norm": 0.6343288421630859,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4808827095626869,
+      "grad_norm": 0.7017961144447327,
+      "learning_rate": 0.0002,
+      "loss": 0.7641,
+      "step": 9160
+    },
+    {
+      "epoch": 1.4824993937434323,
+      "grad_norm": 0.6202912926673889,
+      "learning_rate": 0.0002,
+      "loss": 0.7619,
+      "step": 9170
+    },
+    {
+      "epoch": 1.4841160779241775,
+      "grad_norm": 0.6677869558334351,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 9180
+    },
+    {
+      "epoch": 1.4857327621049228,
+      "grad_norm": 0.6052267551422119,
+      "learning_rate": 0.0002,
+      "loss": 0.7648,
+      "step": 9190
+    },
+    {
+      "epoch": 1.487349446285668,
+      "grad_norm": 0.6638872027397156,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 9200
+    },
+    {
+      "epoch": 1.4889661304664135,
+      "grad_norm": 0.6245523691177368,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 9210
+    },
+    {
+      "epoch": 1.4905828146471587,
+      "grad_norm": 0.5761767625808716,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 9220
+    },
+    {
+      "epoch": 1.492199498827904,
+      "grad_norm": 0.8175981640815735,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 9230
+    },
+    {
+      "epoch": 1.4938161830086494,
+      "grad_norm": 0.9144009947776794,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 9240
+    },
+    {
+      "epoch": 1.4954328671893946,
+      "grad_norm": 0.5742552876472473,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 9250
+    },
+    {
+      "epoch": 1.4970495513701398,
+      "grad_norm": 0.534534215927124,
+      "learning_rate": 0.0002,
+      "loss": 0.7418,
+      "step": 9260
+    },
+    {
+      "epoch": 1.498666235550885,
+      "grad_norm": 0.7836225032806396,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5002829197316303,
+      "grad_norm": 0.5292993187904358,
+      "learning_rate": 0.0002,
+      "loss": 0.7453,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5018996039123758,
+      "grad_norm": 0.8044071793556213,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9290
+    },
+    {
+      "epoch": 1.503516288093121,
+      "grad_norm": 0.6185805201530457,
+      "learning_rate": 0.0002,
+      "loss": 0.7229,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5051329722738664,
+      "grad_norm": 0.6093607544898987,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5067496564546117,
+      "grad_norm": 0.5891730189323425,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 9320
+    },
+    {
+      "epoch": 1.508366340635357,
+      "grad_norm": 0.6331129670143127,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5099830248161021,
+      "grad_norm": 0.7690958380699158,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5115997089968474,
+      "grad_norm": 0.6548877358436584,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5132163931775926,
+      "grad_norm": 0.6545143127441406,
+      "learning_rate": 0.0002,
+      "loss": 0.7408,
+      "step": 9360
+    },
+    {
+      "epoch": 1.514833077358338,
+      "grad_norm": 0.553247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.7432,
+      "step": 9370
+    },
+    {
+      "epoch": 1.5164497615390833,
+      "grad_norm": 0.8145074844360352,
+      "learning_rate": 0.0002,
+      "loss": 0.7265,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5180664457198287,
+      "grad_norm": 0.7636994123458862,
+      "learning_rate": 0.0002,
+      "loss": 0.7379,
+      "step": 9390
+    },
+    {
+      "epoch": 1.519683129900574,
+      "grad_norm": 0.6838982701301575,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5212998140813192,
+      "grad_norm": 0.8599441647529602,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5229164982620644,
+      "grad_norm": 0.7020329833030701,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5245331824428097,
+      "grad_norm": 0.6964772343635559,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5261498666235551,
+      "grad_norm": 0.6916600465774536,
+      "learning_rate": 0.0002,
+      "loss": 0.7168,
+      "step": 9440
+    },
+    {
+      "epoch": 1.5277665508043003,
+      "grad_norm": 0.7282621264457703,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5293832349850458,
+      "grad_norm": 0.5363983511924744,
+      "learning_rate": 0.0002,
+      "loss": 0.7628,
+      "step": 9460
+    },
+    {
+      "epoch": 1.530999919165791,
+      "grad_norm": 0.6184861063957214,
+      "learning_rate": 0.0002,
+      "loss": 0.7154,
+      "step": 9470
+    },
+    {
+      "epoch": 1.5326166033465363,
+      "grad_norm": 0.5991285443305969,
+      "learning_rate": 0.0002,
+      "loss": 0.7837,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5342332875272815,
+      "grad_norm": 0.8176587820053101,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5358499717080267,
+      "grad_norm": 0.6473721861839294,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 9500
+    },
+    {
+      "epoch": 1.5374666558887722,
+      "grad_norm": 0.7319952845573425,
+      "learning_rate": 0.0002,
+      "loss": 0.7632,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5390833400695174,
+      "grad_norm": 0.702900230884552,
+      "learning_rate": 0.0002,
+      "loss": 0.7706,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5407000242502629,
+      "grad_norm": 0.7971600294113159,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 9530
+    },
+    {
+      "epoch": 1.542316708431008,
+      "grad_norm": 0.6527525186538696,
+      "learning_rate": 0.0002,
+      "loss": 0.7352,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5439333926117533,
+      "grad_norm": 0.5791676044464111,
+      "learning_rate": 0.0002,
+      "loss": 0.7425,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5455500767924986,
+      "grad_norm": 0.5619390606880188,
+      "learning_rate": 0.0002,
+      "loss": 0.7585,
+      "step": 9560
+    },
+    {
+      "epoch": 1.5471667609732438,
+      "grad_norm": 0.5701689124107361,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 9570
+    },
+    {
+      "epoch": 1.548783445153989,
+      "grad_norm": 0.47549352049827576,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5504001293347345,
+      "grad_norm": 0.8730611205101013,
+      "learning_rate": 0.0002,
+      "loss": 0.7276,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5520168135154797,
+      "grad_norm": 0.6842091083526611,
+      "learning_rate": 0.0002,
+      "loss": 0.798,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5536334976962252,
+      "grad_norm": 0.6675129532814026,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5552501818769704,
+      "grad_norm": 0.8173956274986267,
+      "learning_rate": 0.0002,
+      "loss": 0.7954,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5568668660577156,
+      "grad_norm": 0.724947452545166,
+      "learning_rate": 0.0002,
+      "loss": 0.7535,
+      "step": 9630
+    },
+    {
+      "epoch": 1.5584835502384609,
+      "grad_norm": 0.6154758930206299,
+      "learning_rate": 0.0002,
+      "loss": 0.7738,
+      "step": 9640
+    },
+    {
+      "epoch": 1.560100234419206,
+      "grad_norm": 0.6072008013725281,
+      "learning_rate": 0.0002,
+      "loss": 0.7568,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5617169185999515,
+      "grad_norm": 0.659010648727417,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 9660
+    },
+    {
+      "epoch": 1.5633336027806968,
+      "grad_norm": 0.65857994556427,
+      "learning_rate": 0.0002,
+      "loss": 0.673,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5649502869614422,
+      "grad_norm": 0.5914267301559448,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5665669711421875,
+      "grad_norm": 0.6248020529747009,
+      "learning_rate": 0.0002,
+      "loss": 0.7414,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5681836553229327,
+      "grad_norm": 0.7147795557975769,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 9700
+    },
+    {
+      "epoch": 1.569800339503678,
+      "grad_norm": 0.7076232433319092,
+      "learning_rate": 0.0002,
+      "loss": 0.7335,
+      "step": 9710
+    },
+    {
+      "epoch": 1.5714170236844232,
+      "grad_norm": 0.6217400431632996,
+      "learning_rate": 0.0002,
+      "loss": 0.7413,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5730337078651684,
+      "grad_norm": 0.6709911227226257,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 9730
+    },
+    {
+      "epoch": 1.5746503920459138,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.7306,
+      "step": 9740
+    },
+    {
+      "epoch": 1.576267076226659,
+      "grad_norm": 0.6241145730018616,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 9750
+    },
+    {
+      "epoch": 1.5778837604074045,
+      "grad_norm": 0.4960934817790985,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 9760
+    },
+    {
+      "epoch": 1.5795004445881498,
+      "grad_norm": 0.6593309640884399,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 9770
+    },
+    {
+      "epoch": 1.581117128768895,
+      "grad_norm": 0.5814042091369629,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 9780
+    },
+    {
+      "epoch": 1.5827338129496402,
+      "grad_norm": 0.5936070680618286,
+      "learning_rate": 0.0002,
+      "loss": 0.7109,
+      "step": 9790
+    },
+    {
+      "epoch": 1.5843504971303854,
+      "grad_norm": 0.6454403400421143,
+      "learning_rate": 0.0002,
+      "loss": 0.7769,
+      "step": 9800
+    },
+    {
+      "epoch": 1.585967181311131,
+      "grad_norm": 0.7612107992172241,
+      "learning_rate": 0.0002,
+      "loss": 0.7677,
+      "step": 9810
+    },
+    {
+      "epoch": 1.5875838654918761,
+      "grad_norm": 0.6494482755661011,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5892005496726216,
+      "grad_norm": 0.7825694680213928,
+      "learning_rate": 0.0002,
+      "loss": 0.7569,
+      "step": 9830
+    },
+    {
+      "epoch": 1.5908172338533668,
+      "grad_norm": 0.6757757663726807,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 9840
+    },
+    {
+      "epoch": 1.592433918034112,
+      "grad_norm": 0.7105609178543091,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 9850
+    },
+    {
+      "epoch": 1.5940506022148573,
+      "grad_norm": 0.7596991062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 9860
+    },
+    {
+      "epoch": 1.5956672863956025,
+      "grad_norm": 0.5681525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7108,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5972839705763477,
+      "grad_norm": 0.6090980768203735,
+      "learning_rate": 0.0002,
+      "loss": 0.7811,
+      "step": 9880
+    },
+    {
+      "epoch": 1.5989006547570932,
+      "grad_norm": 0.6271613240242004,
+      "learning_rate": 0.0002,
+      "loss": 0.7339,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6005173389378387,
+      "grad_norm": 0.7656369805335999,
+      "learning_rate": 0.0002,
+      "loss": 0.7419,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6021340231185839,
+      "grad_norm": 0.7504446506500244,
+      "learning_rate": 0.0002,
+      "loss": 0.7336,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6037507072993291,
+      "grad_norm": 0.659656286239624,
+      "learning_rate": 0.0002,
+      "loss": 0.7479,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6053673914800743,
+      "grad_norm": 0.6006826162338257,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6069840756608196,
+      "grad_norm": 0.7872757911682129,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6086007598415648,
+      "grad_norm": 0.5545852780342102,
+      "learning_rate": 0.0002,
+      "loss": 0.768,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6102174440223103,
+      "grad_norm": 0.7429468631744385,
+      "learning_rate": 0.0002,
+      "loss": 0.8064,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6118341282030555,
+      "grad_norm": 0.6873556971549988,
+      "learning_rate": 0.0002,
+      "loss": 0.714,
+      "step": 9970
+    },
+    {
+      "epoch": 1.613450812383801,
+      "grad_norm": 0.5874287486076355,
+      "learning_rate": 0.0002,
+      "loss": 0.7324,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6150674965645462,
+      "grad_norm": 0.6039386987686157,
+      "learning_rate": 0.0002,
+      "loss": 0.7141,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6166841807452914,
+      "grad_norm": 0.6233575940132141,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6183008649260366,
+      "grad_norm": 0.7676448225975037,
+      "learning_rate": 0.0002,
+      "loss": 0.7602,
+      "step": 10010
+    },
+    {
+      "epoch": 1.6199175491067819,
+      "grad_norm": 0.6565698385238647,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6215342332875273,
+      "grad_norm": 0.6787590384483337,
+      "learning_rate": 0.0002,
+      "loss": 0.7104,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6231509174682726,
+      "grad_norm": 0.6137678027153015,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 10040
+    },
+    {
+      "epoch": 1.624767601649018,
+      "grad_norm": 0.5236800312995911,
+      "learning_rate": 0.0002,
+      "loss": 0.7646,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6263842858297632,
+      "grad_norm": 0.7626367807388306,
+      "learning_rate": 0.0002,
+      "loss": 0.7437,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6280009700105085,
+      "grad_norm": 0.5657260417938232,
+      "learning_rate": 0.0002,
+      "loss": 0.7273,
+      "step": 10070
+    },
+    {
+      "epoch": 1.6296176541912537,
+      "grad_norm": 0.4913991391658783,
+      "learning_rate": 0.0002,
+      "loss": 0.7354,
+      "step": 10080
+    },
+    {
+      "epoch": 1.631234338371999,
+      "grad_norm": 0.7715556621551514,
+      "learning_rate": 0.0002,
+      "loss": 0.7596,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6328510225527442,
+      "grad_norm": 0.6509000062942505,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6344677067334896,
+      "grad_norm": 0.6215850114822388,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6360843909142349,
+      "grad_norm": 0.6956844329833984,
+      "learning_rate": 0.0002,
+      "loss": 0.7705,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6377010750949803,
+      "grad_norm": 0.6111597418785095,
+      "learning_rate": 0.0002,
+      "loss": 0.7129,
+      "step": 10130
+    },
+    {
+      "epoch": 1.6393177592757255,
+      "grad_norm": 0.6518288850784302,
+      "learning_rate": 0.0002,
+      "loss": 0.6955,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6409344434564708,
+      "grad_norm": 0.6914522051811218,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10150
+    },
+    {
+      "epoch": 1.642551127637216,
+      "grad_norm": 0.63785719871521,
+      "learning_rate": 0.0002,
+      "loss": 0.7295,
+      "step": 10160
+    },
+    {
+      "epoch": 1.6441678118179612,
+      "grad_norm": 0.6379287838935852,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6457844959987067,
+      "grad_norm": 0.6793403029441833,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 10180
+    },
+    {
+      "epoch": 1.647401180179452,
+      "grad_norm": 0.6099132895469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7402,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6490178643601974,
+      "grad_norm": 0.5869854092597961,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6506345485409426,
+      "grad_norm": 0.7716999053955078,
+      "learning_rate": 0.0002,
+      "loss": 0.8308,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6522512327216878,
+      "grad_norm": 0.6854110360145569,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 10220
+    },
+    {
+      "epoch": 1.653867916902433,
+      "grad_norm": 0.6957170367240906,
+      "learning_rate": 0.0002,
+      "loss": 0.782,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6554846010831783,
+      "grad_norm": 0.6932903528213501,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6571012852639235,
+      "grad_norm": 0.7713165283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10250
+    },
+    {
+      "epoch": 1.658717969444669,
+      "grad_norm": 0.7455793619155884,
+      "learning_rate": 0.0002,
+      "loss": 0.7099,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6603346536254144,
+      "grad_norm": 0.5464168190956116,
+      "learning_rate": 0.0002,
+      "loss": 0.7524,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6619513378061597,
+      "grad_norm": 0.6782926321029663,
+      "learning_rate": 0.0002,
+      "loss": 0.7328,
+      "step": 10280
+    },
+    {
+      "epoch": 1.663568021986905,
+      "grad_norm": 0.7962649464607239,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6651847061676501,
+      "grad_norm": 0.6814526319503784,
+      "learning_rate": 0.0002,
+      "loss": 0.7142,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6668013903483954,
+      "grad_norm": 0.656895101070404,
+      "learning_rate": 0.0002,
+      "loss": 0.7285,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6684180745291406,
+      "grad_norm": 0.6085672378540039,
+      "learning_rate": 0.0002,
+      "loss": 0.7358,
+      "step": 10320
+    },
+    {
+      "epoch": 1.670034758709886,
+      "grad_norm": 0.585508406162262,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 10330
+    },
+    {
+      "epoch": 1.6716514428906313,
+      "grad_norm": 0.6930184364318848,
+      "learning_rate": 0.0002,
+      "loss": 0.7604,
+      "step": 10340
+    },
+    {
+      "epoch": 1.6732681270713767,
+      "grad_norm": 0.575663149356842,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10350
+    },
+    {
+      "epoch": 1.674884811252122,
+      "grad_norm": 0.582502543926239,
+      "learning_rate": 0.0002,
+      "loss": 0.7198,
+      "step": 10360
+    },
+    {
+      "epoch": 1.6765014954328672,
+      "grad_norm": 0.5668916702270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 10370
+    },
+    {
+      "epoch": 1.6781181796136124,
+      "grad_norm": 0.6070065498352051,
+      "learning_rate": 0.0002,
+      "loss": 0.7478,
+      "step": 10380
+    },
+    {
+      "epoch": 1.6797348637943577,
+      "grad_norm": 0.6141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.7939,
+      "step": 10390
+    },
+    {
+      "epoch": 1.6813515479751031,
+      "grad_norm": 0.8359124064445496,
+      "learning_rate": 0.0002,
+      "loss": 0.7573,
+      "step": 10400
+    },
+    {
+      "epoch": 1.6829682321558483,
+      "grad_norm": 0.5378185510635376,
+      "learning_rate": 0.0002,
+      "loss": 0.7488,
+      "step": 10410
+    },
+    {
+      "epoch": 1.6845849163365938,
+      "grad_norm": 0.6959536075592041,
+      "learning_rate": 0.0002,
+      "loss": 0.7588,
+      "step": 10420
+    },
+    {
+      "epoch": 1.686201600517339,
+      "grad_norm": 0.6514357328414917,
+      "learning_rate": 0.0002,
+      "loss": 0.7872,
+      "step": 10430
+    },
+    {
+      "epoch": 1.6878182846980843,
+      "grad_norm": 0.7706646919250488,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10440
+    },
+    {
+      "epoch": 1.6894349688788295,
+      "grad_norm": 0.6183337569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.7673,
+      "step": 10450
+    },
+    {
+      "epoch": 1.6910516530595747,
+      "grad_norm": 0.6123278141021729,
+      "learning_rate": 0.0002,
+      "loss": 0.7566,
+      "step": 10460
+    },
+    {
+      "epoch": 1.69266833724032,
+      "grad_norm": 0.6894851326942444,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 10470
+    },
+    {
+      "epoch": 1.6942850214210654,
+      "grad_norm": 0.7497312426567078,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 10480
+    },
+    {
+      "epoch": 1.6959017056018106,
+      "grad_norm": 0.5968214273452759,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10490
+    },
+    {
+      "epoch": 1.697518389782556,
+      "grad_norm": 0.6747927069664001,
+      "learning_rate": 0.0002,
+      "loss": 0.6793,
+      "step": 10500
+    },
+    {
+      "epoch": 1.6991350739633013,
+      "grad_norm": 0.5708310008049011,
+      "learning_rate": 0.0002,
+      "loss": 0.7415,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7007517581440466,
+      "grad_norm": 0.606526792049408,
+      "learning_rate": 0.0002,
+      "loss": 0.7385,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7023684423247918,
+      "grad_norm": 0.662011981010437,
+      "learning_rate": 0.0002,
+      "loss": 0.7204,
+      "step": 10530
+    },
+    {
+      "epoch": 1.703985126505537,
+      "grad_norm": 0.7583045363426208,
+      "learning_rate": 0.0002,
+      "loss": 0.7999,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7056018106862825,
+      "grad_norm": 0.721632182598114,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7072184948670277,
+      "grad_norm": 0.6107715368270874,
+      "learning_rate": 0.0002,
+      "loss": 0.7407,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7088351790477732,
+      "grad_norm": 0.6652471423149109,
+      "learning_rate": 0.0002,
+      "loss": 0.7519,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7104518632285184,
+      "grad_norm": 0.6308087110519409,
+      "learning_rate": 0.0002,
+      "loss": 0.7767,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7120685474092636,
+      "grad_norm": 0.5464386940002441,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7136852315900089,
+      "grad_norm": 0.6558911204338074,
+      "learning_rate": 0.0002,
+      "loss": 0.7063,
+      "step": 10600
+    },
+    {
+      "epoch": 1.715301915770754,
+      "grad_norm": 0.5665024518966675,
+      "learning_rate": 0.0002,
+      "loss": 0.7126,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7169185999514993,
+      "grad_norm": 0.7888094186782837,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7185352841322448,
+      "grad_norm": 0.7084909081459045,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 10630
+    },
+    {
+      "epoch": 1.7201519683129902,
+      "grad_norm": 0.7982324361801147,
+      "learning_rate": 0.0002,
+      "loss": 0.7557,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7217686524937355,
+      "grad_norm": 0.6418732404708862,
+      "learning_rate": 0.0002,
+      "loss": 0.7345,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7233853366744807,
+      "grad_norm": 0.7636681795120239,
+      "learning_rate": 0.0002,
+      "loss": 0.7734,
+      "step": 10660
+    },
+    {
+      "epoch": 1.725002020855226,
+      "grad_norm": 0.5646875500679016,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7266187050359711,
+      "grad_norm": 0.5231260657310486,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7282353892167164,
+      "grad_norm": 0.7635011672973633,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7298520733974618,
+      "grad_norm": 0.7518259286880493,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 10700
+    },
+    {
+      "epoch": 1.731468757578207,
+      "grad_norm": 0.7295602560043335,
+      "learning_rate": 0.0002,
+      "loss": 0.751,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7330854417589525,
+      "grad_norm": 0.6984632015228271,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7347021259396977,
+      "grad_norm": 0.6198219060897827,
+      "learning_rate": 0.0002,
+      "loss": 0.7921,
+      "step": 10730
+    },
+    {
+      "epoch": 1.736318810120443,
+      "grad_norm": 0.6957576274871826,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7379354943011882,
+      "grad_norm": 0.6430263519287109,
+      "learning_rate": 0.0002,
+      "loss": 0.7917,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7395521784819334,
+      "grad_norm": 0.6134995222091675,
+      "learning_rate": 0.0002,
+      "loss": 0.7156,
+      "step": 10760
+    },
+    {
+      "epoch": 1.741168862662679,
+      "grad_norm": 0.7209452986717224,
+      "learning_rate": 0.0002,
+      "loss": 0.7584,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7427855468434241,
+      "grad_norm": 0.6735447645187378,
+      "learning_rate": 0.0002,
+      "loss": 0.7528,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7444022310241696,
+      "grad_norm": 0.5605693459510803,
+      "learning_rate": 0.0002,
+      "loss": 0.756,
+      "step": 10790
+    },
+    {
+      "epoch": 1.7460189152049148,
+      "grad_norm": 0.6882363557815552,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 10800
+    },
+    {
+      "epoch": 1.74763559938566,
+      "grad_norm": 0.6386259198188782,
+      "learning_rate": 0.0002,
+      "loss": 0.7544,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7492522835664053,
+      "grad_norm": 0.6529015302658081,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7508689677471505,
+      "grad_norm": 0.5664082765579224,
+      "learning_rate": 0.0002,
+      "loss": 0.7219,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7524856519278957,
+      "grad_norm": 0.7532684206962585,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7541023361086412,
+      "grad_norm": 0.77171391248703,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7557190202893864,
+      "grad_norm": 0.7255431413650513,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7573357044701319,
+      "grad_norm": 0.763083279132843,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10870
+    },
+    {
+      "epoch": 1.758952388650877,
+      "grad_norm": 0.6042402982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7605690728316223,
+      "grad_norm": 0.7642518281936646,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 10890
+    },
+    {
+      "epoch": 1.7621857570123676,
+      "grad_norm": 0.6347904801368713,
+      "learning_rate": 0.0002,
+      "loss": 0.746,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7638024411931128,
+      "grad_norm": 0.5371627807617188,
+      "learning_rate": 0.0002,
+      "loss": 0.7458,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7654191253738583,
+      "grad_norm": 0.6840225458145142,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7670358095546035,
+      "grad_norm": 0.5288469195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.725,
+      "step": 10930
+    },
+    {
+      "epoch": 1.768652493735349,
+      "grad_norm": 0.69020676612854,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 10940
+    },
+    {
+      "epoch": 1.7702691779160942,
+      "grad_norm": 0.5943242311477661,
+      "learning_rate": 0.0002,
+      "loss": 0.7468,
+      "step": 10950
+    },
+    {
+      "epoch": 1.7718858620968394,
+      "grad_norm": 0.5616418123245239,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 10960
+    },
+    {
+      "epoch": 1.7735025462775846,
+      "grad_norm": 0.7209470868110657,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 10970
+    },
+    {
+      "epoch": 1.7751192304583299,
+      "grad_norm": 0.6657957434654236,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 10980
+    },
+    {
+      "epoch": 1.776735914639075,
+      "grad_norm": 0.6469064950942993,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 10990
+    },
+    {
+      "epoch": 1.7783525988198206,
+      "grad_norm": 0.6615678071975708,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 11000
+    },
+    {
+      "epoch": 1.779969283000566,
+      "grad_norm": 0.6722439527511597,
+      "learning_rate": 0.0002,
+      "loss": 0.747,
+      "step": 11010
+    },
+    {
+      "epoch": 1.7815859671813112,
+      "grad_norm": 0.634136974811554,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 11020
+    },
+    {
+      "epoch": 1.7832026513620565,
+      "grad_norm": 0.6024377346038818,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 11030
+    },
+    {
+      "epoch": 1.7848193355428017,
+      "grad_norm": 0.6909403800964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 11040
+    },
+    {
+      "epoch": 1.786436019723547,
+      "grad_norm": 0.7148767709732056,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 11050
+    },
+    {
+      "epoch": 1.7880527039042922,
+      "grad_norm": 0.7442979216575623,
+      "learning_rate": 0.0002,
+      "loss": 0.7145,
+      "step": 11060
+    },
+    {
+      "epoch": 1.7896693880850376,
+      "grad_norm": 0.6830431818962097,
+      "learning_rate": 0.0002,
+      "loss": 0.7215,
+      "step": 11070
+    },
+    {
+      "epoch": 1.7912860722657828,
+      "grad_norm": 0.9172667264938354,
+      "learning_rate": 0.0002,
+      "loss": 0.7625,
+      "step": 11080
+    },
+    {
+      "epoch": 1.7929027564465283,
+      "grad_norm": 0.6799490451812744,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 11090
+    },
+    {
+      "epoch": 1.7945194406272735,
+      "grad_norm": 0.7617024779319763,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 11100
+    },
+    {
+      "epoch": 1.7961361248080188,
+      "grad_norm": 0.7701810002326965,
+      "learning_rate": 0.0002,
+      "loss": 0.7586,
+      "step": 11110
+    },
+    {
+      "epoch": 1.797752808988764,
+      "grad_norm": 0.7454385757446289,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11120
+    },
+    {
+      "epoch": 1.7993694931695092,
+      "grad_norm": 0.6121436953544617,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8009861773502547,
+      "grad_norm": 0.6237571835517883,
+      "learning_rate": 0.0002,
+      "loss": 0.7305,
+      "step": 11140
+    },
+    {
+      "epoch": 1.802602861531,
+      "grad_norm": 0.6818515658378601,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 11150
+    },
+    {
+      "epoch": 1.8042195457117454,
+      "grad_norm": 0.7768308520317078,
+      "learning_rate": 0.0002,
+      "loss": 0.6876,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8058362298924906,
+      "grad_norm": 0.6875537633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 11170
+    },
+    {
+      "epoch": 1.8074529140732358,
+      "grad_norm": 0.7950584888458252,
+      "learning_rate": 0.0002,
+      "loss": 0.761,
+      "step": 11180
+    },
+    {
+      "epoch": 1.809069598253981,
+      "grad_norm": 0.8210248351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.7623,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8106862824347263,
+      "grad_norm": 0.6674110889434814,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8123029666154715,
+      "grad_norm": 0.6261674761772156,
+      "learning_rate": 0.0002,
+      "loss": 0.7663,
+      "step": 11210
+    },
+    {
+      "epoch": 1.813919650796217,
+      "grad_norm": 0.6484741568565369,
+      "learning_rate": 0.0002,
+      "loss": 0.7122,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8155363349769622,
+      "grad_norm": 0.6231244206428528,
+      "learning_rate": 0.0002,
+      "loss": 0.7718,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8171530191577077,
+      "grad_norm": 0.7243146896362305,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11240
+    },
+    {
+      "epoch": 1.818769703338453,
+      "grad_norm": 0.6776193380355835,
+      "learning_rate": 0.0002,
+      "loss": 0.7448,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8203863875191981,
+      "grad_norm": 0.5973618030548096,
+      "learning_rate": 0.0002,
+      "loss": 0.7317,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8220030716999434,
+      "grad_norm": 0.6451361179351807,
+      "learning_rate": 0.0002,
+      "loss": 0.7961,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8236197558806886,
+      "grad_norm": 0.5963068008422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 11280
+    },
+    {
+      "epoch": 1.825236440061434,
+      "grad_norm": 0.536902129650116,
+      "learning_rate": 0.0002,
+      "loss": 0.7466,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8268531242421793,
+      "grad_norm": 0.6993787288665771,
+      "learning_rate": 0.0002,
+      "loss": 0.708,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8284698084229247,
+      "grad_norm": 0.6135255098342896,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11310
+    },
+    {
+      "epoch": 1.83008649260367,
+      "grad_norm": 0.6057423949241638,
+      "learning_rate": 0.0002,
+      "loss": 0.7423,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8317031767844152,
+      "grad_norm": 0.6598812341690063,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8333198609651604,
+      "grad_norm": 0.6075948476791382,
+      "learning_rate": 0.0002,
+      "loss": 0.7278,
+      "step": 11340
+    },
+    {
+      "epoch": 1.8349365451459057,
+      "grad_norm": 0.7065447568893433,
+      "learning_rate": 0.0002,
+      "loss": 0.7846,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8365532293266509,
+      "grad_norm": 0.680526614189148,
+      "learning_rate": 0.0002,
+      "loss": 0.7365,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8381699135073963,
+      "grad_norm": 0.6356695294380188,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8397865976881416,
+      "grad_norm": 0.6399052143096924,
+      "learning_rate": 0.0002,
+      "loss": 0.721,
+      "step": 11380
+    },
+    {
+      "epoch": 1.841403281868887,
+      "grad_norm": 0.6125704050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 11390
+    },
+    {
+      "epoch": 1.8430199660496323,
+      "grad_norm": 0.7124643325805664,
+      "learning_rate": 0.0002,
+      "loss": 0.755,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8446366502303775,
+      "grad_norm": 0.6099604964256287,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8462533344111227,
+      "grad_norm": 0.7338208556175232,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 11420
+    },
+    {
+      "epoch": 1.847870018591868,
+      "grad_norm": 0.7534668445587158,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8494867027726134,
+      "grad_norm": 0.6135470271110535,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8511033869533586,
+      "grad_norm": 0.6229309439659119,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 11450
+    },
+    {
+      "epoch": 1.852720071134104,
+      "grad_norm": 0.706423282623291,
+      "learning_rate": 0.0002,
+      "loss": 0.7594,
+      "step": 11460
+    },
+    {
+      "epoch": 1.8543367553148493,
+      "grad_norm": 0.5460049510002136,
+      "learning_rate": 0.0002,
+      "loss": 0.7411,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8559534394955945,
+      "grad_norm": 0.6616711020469666,
+      "learning_rate": 0.0002,
+      "loss": 0.7416,
+      "step": 11480
+    },
+    {
+      "epoch": 1.8575701236763398,
+      "grad_norm": 0.6372783184051514,
+      "learning_rate": 0.0002,
+      "loss": 0.729,
+      "step": 11490
+    },
+    {
+      "epoch": 1.859186807857085,
+      "grad_norm": 0.7162668108940125,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8608034920378305,
+      "grad_norm": 0.6605209708213806,
+      "learning_rate": 0.0002,
+      "loss": 0.7747,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8624201762185757,
+      "grad_norm": 0.6933956742286682,
+      "learning_rate": 0.0002,
+      "loss": 0.7258,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8640368603993211,
+      "grad_norm": 0.6582090854644775,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8656535445800664,
+      "grad_norm": 0.6416500806808472,
+      "learning_rate": 0.0002,
+      "loss": 0.7313,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8672702287608116,
+      "grad_norm": 0.5434312224388123,
+      "learning_rate": 0.0002,
+      "loss": 0.7372,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8688869129415568,
+      "grad_norm": 0.6827567219734192,
+      "learning_rate": 0.0002,
+      "loss": 0.7635,
+      "step": 11560
+    },
+    {
+      "epoch": 1.870503597122302,
+      "grad_norm": 0.7354370951652527,
+      "learning_rate": 0.0002,
+      "loss": 0.7137,
+      "step": 11570
+    },
+    {
+      "epoch": 1.8721202813030473,
+      "grad_norm": 0.590372622013092,
+      "learning_rate": 0.0002,
+      "loss": 0.7526,
+      "step": 11580
+    },
+    {
+      "epoch": 1.8737369654837928,
+      "grad_norm": 0.853183925151825,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 11590
+    },
+    {
+      "epoch": 1.875353649664538,
+      "grad_norm": 0.822678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.7487,
+      "step": 11600
+    },
+    {
+      "epoch": 1.8769703338452834,
+      "grad_norm": 0.6591550707817078,
+      "learning_rate": 0.0002,
+      "loss": 0.7427,
+      "step": 11610
+    },
+    {
+      "epoch": 1.8785870180260287,
+      "grad_norm": 0.7475301623344421,
+      "learning_rate": 0.0002,
+      "loss": 0.7054,
+      "step": 11620
+    },
+    {
+      "epoch": 1.880203702206774,
+      "grad_norm": 0.6390765309333801,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 11630
+    },
+    {
+      "epoch": 1.8818203863875191,
+      "grad_norm": 0.6589758992195129,
+      "learning_rate": 0.0002,
+      "loss": 0.7531,
+      "step": 11640
+    },
+    {
+      "epoch": 1.8834370705682644,
+      "grad_norm": 0.6765508651733398,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 11650
+    },
+    {
+      "epoch": 1.8850537547490098,
+      "grad_norm": 0.6527857780456543,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 11660
+    },
+    {
+      "epoch": 1.886670438929755,
+      "grad_norm": 0.6642923951148987,
+      "learning_rate": 0.0002,
+      "loss": 0.7504,
+      "step": 11670
+    },
+    {
+      "epoch": 1.8882871231105005,
+      "grad_norm": 0.6945584416389465,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 11680
+    },
+    {
+      "epoch": 1.8899038072912457,
+      "grad_norm": 0.694018542766571,
+      "learning_rate": 0.0002,
+      "loss": 0.7711,
+      "step": 11690
+    },
+    {
+      "epoch": 1.891520491471991,
+      "grad_norm": 0.7237417101860046,
+      "learning_rate": 0.0002,
+      "loss": 0.7195,
+      "step": 11700
+    },
+    {
+      "epoch": 1.8931371756527362,
+      "grad_norm": 0.7401309609413147,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 11710
+    },
+    {
+      "epoch": 1.8947538598334814,
+      "grad_norm": 0.6537784337997437,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 11720
+    },
+    {
+      "epoch": 1.8963705440142267,
+      "grad_norm": 0.7398539185523987,
+      "learning_rate": 0.0002,
+      "loss": 0.793,
+      "step": 11730
+    },
+    {
+      "epoch": 1.8979872281949721,
+      "grad_norm": 0.6696075797080994,
+      "learning_rate": 0.0002,
+      "loss": 0.7561,
+      "step": 11740
+    },
+    {
+      "epoch": 1.8996039123757174,
+      "grad_norm": 0.6014142036437988,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9012205965564628,
+      "grad_norm": 0.7023524641990662,
+      "learning_rate": 0.0002,
+      "loss": 0.7714,
+      "step": 11760
+    },
+    {
+      "epoch": 1.902837280737208,
+      "grad_norm": 0.739973783493042,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9044539649179533,
+      "grad_norm": 0.5576770901679993,
+      "learning_rate": 0.0002,
+      "loss": 0.7848,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9060706490986985,
+      "grad_norm": 0.6907393932342529,
+      "learning_rate": 0.0002,
+      "loss": 0.7483,
+      "step": 11790
+    },
+    {
+      "epoch": 1.9076873332794437,
+      "grad_norm": 0.6934581995010376,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9093040174601892,
+      "grad_norm": 0.591774582862854,
+      "learning_rate": 0.0002,
+      "loss": 0.7199,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9109207016409344,
+      "grad_norm": 0.6249791383743286,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 11820
+    },
+    {
+      "epoch": 1.9125373858216799,
+      "grad_norm": 0.6755744218826294,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 11830
+    },
+    {
+      "epoch": 1.914154070002425,
+      "grad_norm": 0.7286285161972046,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9157707541831703,
+      "grad_norm": 0.7867850065231323,
+      "learning_rate": 0.0002,
+      "loss": 0.7509,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9173874383639156,
+      "grad_norm": 0.6283972859382629,
+      "learning_rate": 0.0002,
+      "loss": 0.735,
+      "step": 11860
+    },
+    {
+      "epoch": 1.9190041225446608,
+      "grad_norm": 0.605823814868927,
+      "learning_rate": 0.0002,
+      "loss": 0.7296,
+      "step": 11870
+    },
+    {
+      "epoch": 1.920620806725406,
+      "grad_norm": 0.5927976965904236,
+      "learning_rate": 0.0002,
+      "loss": 0.6598,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9222374909061515,
+      "grad_norm": 0.5974002480506897,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 11890
+    },
+    {
+      "epoch": 1.923854175086897,
+      "grad_norm": 0.7091866135597229,
+      "learning_rate": 0.0002,
+      "loss": 0.7843,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9254708592676422,
+      "grad_norm": 0.72496497631073,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9270875434483874,
+      "grad_norm": 0.6131896376609802,
+      "learning_rate": 0.0002,
+      "loss": 0.7153,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9287042276291326,
+      "grad_norm": 0.6556436419487,
+      "learning_rate": 0.0002,
+      "loss": 0.7228,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9303209118098779,
+      "grad_norm": 0.622932493686676,
+      "learning_rate": 0.0002,
+      "loss": 0.7319,
+      "step": 11940
+    },
+    {
+      "epoch": 1.931937595990623,
+      "grad_norm": 0.6618631482124329,
+      "learning_rate": 0.0002,
+      "loss": 0.7592,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9335542801713685,
+      "grad_norm": 0.630966305732727,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 11960
+    },
+    {
+      "epoch": 1.9351709643521138,
+      "grad_norm": 0.6336734890937805,
+      "learning_rate": 0.0002,
+      "loss": 0.6854,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9367876485328592,
+      "grad_norm": 0.655403196811676,
+      "learning_rate": 0.0002,
+      "loss": 0.7433,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9384043327136045,
+      "grad_norm": 0.5640574097633362,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 11990
+    },
+    {
+      "epoch": 1.9400210168943497,
+      "grad_norm": 0.6322951316833496,
+      "learning_rate": 0.0002,
+      "loss": 0.7289,
+      "step": 12000
+    },
+    {
+      "epoch": 1.941637701075095,
+      "grad_norm": 0.615703821182251,
+      "learning_rate": 0.0002,
+      "loss": 0.7627,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9432543852558402,
+      "grad_norm": 0.6487536430358887,
+      "learning_rate": 0.0002,
+      "loss": 0.786,
+      "step": 12020
+    },
+    {
+      "epoch": 1.9448710694365856,
+      "grad_norm": 0.9209630489349365,
+      "learning_rate": 0.0002,
+      "loss": 0.7435,
+      "step": 12030
+    },
+    {
+      "epoch": 1.9464877536173308,
+      "grad_norm": 0.67485511302948,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9481044377980763,
+      "grad_norm": 0.6831230521202087,
+      "learning_rate": 0.0002,
+      "loss": 0.7551,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9497211219788215,
+      "grad_norm": 0.6578302383422852,
+      "learning_rate": 0.0002,
+      "loss": 0.7546,
+      "step": 12060
+    },
+    {
+      "epoch": 1.9513378061595668,
+      "grad_norm": 0.9975938200950623,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12070
+    },
+    {
+      "epoch": 1.952954490340312,
+      "grad_norm": 0.6637365221977234,
+      "learning_rate": 0.0002,
+      "loss": 0.7952,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9545711745210572,
+      "grad_norm": 0.605707049369812,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9561878587018025,
+      "grad_norm": 0.6584440469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7768,
+      "step": 12100
+    },
+    {
+      "epoch": 1.957804542882548,
+      "grad_norm": 0.6070835590362549,
+      "learning_rate": 0.0002,
+      "loss": 0.7187,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9594212270632931,
+      "grad_norm": 0.7862601280212402,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9610379112440386,
+      "grad_norm": 0.8175255060195923,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 12130
+    },
+    {
+      "epoch": 1.9626545954247838,
+      "grad_norm": 0.5648472905158997,
+      "learning_rate": 0.0002,
+      "loss": 0.7242,
+      "step": 12140
+    },
+    {
+      "epoch": 1.964271279605529,
+      "grad_norm": 0.6591973304748535,
+      "learning_rate": 0.0002,
+      "loss": 0.7321,
+      "step": 12150
+    },
+    {
+      "epoch": 1.9658879637862743,
+      "grad_norm": 0.5960676074028015,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9675046479670195,
+      "grad_norm": 0.7272544503211975,
+      "learning_rate": 0.0002,
+      "loss": 0.7254,
+      "step": 12170
+    },
+    {
+      "epoch": 1.969121332147765,
+      "grad_norm": 0.7176699042320251,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 12180
+    },
+    {
+      "epoch": 1.9707380163285102,
+      "grad_norm": 0.6927123665809631,
+      "learning_rate": 0.0002,
+      "loss": 0.7525,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9723547005092557,
+      "grad_norm": 0.5536034107208252,
+      "learning_rate": 0.0002,
+      "loss": 0.7318,
+      "step": 12200
+    },
+    {
+      "epoch": 1.9739713846900009,
+      "grad_norm": 0.8348390460014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 12210
+    },
+    {
+      "epoch": 1.9755880688707461,
+      "grad_norm": 0.6591181755065918,
+      "learning_rate": 0.0002,
+      "loss": 0.7494,
+      "step": 12220
+    },
+    {
+      "epoch": 1.9772047530514913,
+      "grad_norm": 1.0624109506607056,
+      "learning_rate": 0.0002,
+      "loss": 0.763,
+      "step": 12230
+    },
+    {
+      "epoch": 1.9788214372322366,
+      "grad_norm": 0.9265586137771606,
+      "learning_rate": 0.0002,
+      "loss": 0.7541,
+      "step": 12240
+    },
+    {
+      "epoch": 1.9804381214129818,
+      "grad_norm": 0.5998196005821228,
+      "learning_rate": 0.0002,
+      "loss": 0.7533,
+      "step": 12250
+    },
+    {
+      "epoch": 1.9820548055937273,
+      "grad_norm": 0.6960851550102234,
+      "learning_rate": 0.0002,
+      "loss": 0.7225,
+      "step": 12260
+    },
+    {
+      "epoch": 1.9836714897744727,
+      "grad_norm": 0.7674502730369568,
+      "learning_rate": 0.0002,
+      "loss": 0.7398,
+      "step": 12270
+    },
+    {
+      "epoch": 1.985288173955218,
+      "grad_norm": 0.6407275795936584,
+      "learning_rate": 0.0002,
+      "loss": 0.7185,
+      "step": 12280
+    },
+    {
+      "epoch": 1.9869048581359632,
+      "grad_norm": 0.6673079133033752,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 12290
+    },
+    {
+      "epoch": 1.9885215423167084,
+      "grad_norm": 0.6989844441413879,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 12300
+    },
+    {
+      "epoch": 1.9901382264974536,
+      "grad_norm": 0.7564442157745361,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 12310
+    },
+    {
+      "epoch": 1.9917549106781989,
+      "grad_norm": 0.6385478973388672,
+      "learning_rate": 0.0002,
+      "loss": 0.7719,
+      "step": 12320
+    },
+    {
+      "epoch": 1.9933715948589443,
+      "grad_norm": 0.7193717956542969,
+      "learning_rate": 0.0002,
+      "loss": 0.7369,
+      "step": 12330
+    },
+    {
+      "epoch": 1.9949882790396896,
+      "grad_norm": 0.7987112402915955,
+      "learning_rate": 0.0002,
+      "loss": 0.7583,
+      "step": 12340
+    },
+    {
+      "epoch": 1.996604963220435,
+      "grad_norm": 0.7260826826095581,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 12350
+    },
+    {
+      "epoch": 1.9982216474011802,
+      "grad_norm": 0.7968255281448364,
+      "learning_rate": 0.0002,
+      "loss": 0.7505,
+      "step": 12360
+    },
+    {
+      "epoch": 1.9998383315819255,
+      "grad_norm": 0.6893062591552734,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1044032573699951,
+      "eval_runtime": 122.1508,
+      "eval_samples_per_second": 6.001,
+      "eval_steps_per_second": 0.753,
+      "step": 12371
+    },
+    {
+      "epoch": 2.0014550157626707,
+      "grad_norm": 0.7775409817695618,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 12380
+    },
+    {
+      "epoch": 2.003071699943416,
+      "grad_norm": 0.76218581199646,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 12390
+    },
+    {
+      "epoch": 2.004688384124161,
+      "grad_norm": 0.5677764415740967,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 12400
+    },
+    {
+      "epoch": 2.006305068304907,
+      "grad_norm": 0.808442234992981,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 12410
+    },
+    {
+      "epoch": 2.007921752485652,
+      "grad_norm": 0.7144765257835388,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 12420
+    },
+    {
+      "epoch": 2.0095384366663973,
+      "grad_norm": 0.6914031505584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 12430
+    },
+    {
+      "epoch": 2.0111551208471425,
+      "grad_norm": 0.7581454515457153,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 12440
+    },
+    {
+      "epoch": 2.0127718050278878,
+      "grad_norm": 0.8388504981994629,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 12450
+    },
+    {
+      "epoch": 2.014388489208633,
+      "grad_norm": 0.6716406941413879,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 12460
+    },
+    {
+      "epoch": 2.0160051733893782,
+      "grad_norm": 0.898902416229248,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 12470
+    },
+    {
+      "epoch": 2.0176218575701235,
+      "grad_norm": 0.6432679891586304,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 12480
+    },
+    {
+      "epoch": 2.019238541750869,
+      "grad_norm": 0.8021109104156494,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12490
+    },
+    {
+      "epoch": 2.0208552259316144,
+      "grad_norm": 0.7039216756820679,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 12500
+    },
+    {
+      "epoch": 2.0224719101123596,
+      "grad_norm": 0.646531879901886,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 12510
+    },
+    {
+      "epoch": 2.024088594293105,
+      "grad_norm": 0.783704400062561,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 12520
+    },
+    {
+      "epoch": 2.02570527847385,
+      "grad_norm": 0.8805046677589417,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12530
+    },
+    {
+      "epoch": 2.0273219626545953,
+      "grad_norm": 0.7289270758628845,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0289386468353405,
+      "grad_norm": 0.71653151512146,
+      "learning_rate": 0.0002,
+      "loss": 0.6663,
+      "step": 12550
+    },
+    {
+      "epoch": 2.030555331016086,
+      "grad_norm": 0.73281329870224,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 12560
+    },
+    {
+      "epoch": 2.0321720151968314,
+      "grad_norm": 0.6657090187072754,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 12570
+    },
+    {
+      "epoch": 2.0337886993775767,
+      "grad_norm": 0.8241133093833923,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 12580
+    },
+    {
+      "epoch": 2.035405383558322,
+      "grad_norm": 0.5834135413169861,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 12590
+    },
+    {
+      "epoch": 2.037022067739067,
+      "grad_norm": 0.84502112865448,
+      "learning_rate": 0.0002,
+      "loss": 0.6188,
+      "step": 12600
+    },
+    {
+      "epoch": 2.0386387519198124,
+      "grad_norm": 0.8952481746673584,
+      "learning_rate": 0.0002,
+      "loss": 0.6349,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0402554361005576,
+      "grad_norm": 0.7801461815834045,
+      "learning_rate": 0.0002,
+      "loss": 0.6923,
+      "step": 12620
+    },
+    {
+      "epoch": 2.041872120281303,
+      "grad_norm": 0.6788367033004761,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 12630
+    },
+    {
+      "epoch": 2.0434888044620485,
+      "grad_norm": 0.7241756319999695,
+      "learning_rate": 0.0002,
+      "loss": 0.6162,
+      "step": 12640
+    },
+    {
+      "epoch": 2.0451054886427937,
+      "grad_norm": 0.6933388113975525,
+      "learning_rate": 0.0002,
+      "loss": 0.655,
+      "step": 12650
+    },
+    {
+      "epoch": 2.046722172823539,
+      "grad_norm": 0.8029746413230896,
+      "learning_rate": 0.0002,
+      "loss": 0.6431,
+      "step": 12660
+    },
+    {
+      "epoch": 2.048338857004284,
+      "grad_norm": 0.946399986743927,
+      "learning_rate": 0.0002,
+      "loss": 0.7164,
+      "step": 12670
+    },
+    {
+      "epoch": 2.0499555411850294,
+      "grad_norm": 0.7072678804397583,
+      "learning_rate": 0.0002,
+      "loss": 0.638,
+      "step": 12680
+    },
+    {
+      "epoch": 2.0515722253657747,
+      "grad_norm": 0.6810618042945862,
+      "learning_rate": 0.0002,
+      "loss": 0.6487,
+      "step": 12690
+    },
+    {
+      "epoch": 2.05318890954652,
+      "grad_norm": 0.7661160230636597,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 12700
+    },
+    {
+      "epoch": 2.0548055937272656,
+      "grad_norm": 0.6350653767585754,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 12710
+    },
+    {
+      "epoch": 2.056422277908011,
+      "grad_norm": 0.861890971660614,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 12720
+    },
+    {
+      "epoch": 2.058038962088756,
+      "grad_norm": 0.6489875912666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 12730
+    },
+    {
+      "epoch": 2.0596556462695013,
+      "grad_norm": 0.8268506526947021,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 12740
+    },
+    {
+      "epoch": 2.0612723304502465,
+      "grad_norm": 0.607679545879364,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 12750
+    },
+    {
+      "epoch": 2.0628890146309917,
+      "grad_norm": 0.6754153370857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6649,
+      "step": 12760
+    },
+    {
+      "epoch": 2.064505698811737,
+      "grad_norm": 0.7263124585151672,
+      "learning_rate": 0.0002,
+      "loss": 0.6549,
+      "step": 12770
+    },
+    {
+      "epoch": 2.0661223829924826,
+      "grad_norm": 0.6986154317855835,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 12780
+    },
+    {
+      "epoch": 2.067739067173228,
+      "grad_norm": 0.7768576741218567,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 12790
+    },
+    {
+      "epoch": 2.069355751353973,
+      "grad_norm": 0.7546762824058533,
+      "learning_rate": 0.0002,
+      "loss": 0.677,
+      "step": 12800
+    },
+    {
+      "epoch": 2.0709724355347183,
+      "grad_norm": 0.7588880062103271,
+      "learning_rate": 0.0002,
+      "loss": 0.6485,
+      "step": 12810
+    },
+    {
+      "epoch": 2.0725891197154636,
+      "grad_norm": 0.7457242608070374,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 12820
+    },
+    {
+      "epoch": 2.074205803896209,
+      "grad_norm": 0.6983516812324524,
+      "learning_rate": 0.0002,
+      "loss": 0.6489,
+      "step": 12830
+    },
+    {
+      "epoch": 2.075822488076954,
+      "grad_norm": 0.7950928807258606,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 12840
+    },
+    {
+      "epoch": 2.0774391722576993,
+      "grad_norm": 0.9248087406158447,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 12850
+    },
+    {
+      "epoch": 2.079055856438445,
+      "grad_norm": 0.7229493260383606,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 12860
+    },
+    {
+      "epoch": 2.08067254061919,
+      "grad_norm": 0.5710847973823547,
+      "learning_rate": 0.0002,
+      "loss": 0.6702,
+      "step": 12870
+    },
+    {
+      "epoch": 2.0822892247999354,
+      "grad_norm": 0.9580423831939697,
+      "learning_rate": 0.0002,
+      "loss": 0.6974,
+      "step": 12880
+    },
+    {
+      "epoch": 2.0839059089806806,
+      "grad_norm": 0.7399665713310242,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 12890
+    },
+    {
+      "epoch": 2.085522593161426,
+      "grad_norm": 0.7981410622596741,
+      "learning_rate": 0.0002,
+      "loss": 0.6993,
+      "step": 12900
+    },
+    {
+      "epoch": 2.087139277342171,
+      "grad_norm": 0.870759904384613,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 12910
+    },
+    {
+      "epoch": 2.0887559615229163,
+      "grad_norm": 0.7001481652259827,
+      "learning_rate": 0.0002,
+      "loss": 0.7194,
+      "step": 12920
+    },
+    {
+      "epoch": 2.090372645703662,
+      "grad_norm": 0.6745418310165405,
+      "learning_rate": 0.0002,
+      "loss": 0.6383,
+      "step": 12930
+    },
+    {
+      "epoch": 2.0919893298844072,
+      "grad_norm": 0.7739067673683167,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 12940
+    },
+    {
+      "epoch": 2.0936060140651525,
+      "grad_norm": 0.6742934584617615,
+      "learning_rate": 0.0002,
+      "loss": 0.6856,
+      "step": 12950
+    },
+    {
+      "epoch": 2.0952226982458977,
+      "grad_norm": 0.7270349860191345,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 12960
+    },
+    {
+      "epoch": 2.096839382426643,
+      "grad_norm": 0.7150624394416809,
+      "learning_rate": 0.0002,
+      "loss": 0.6783,
+      "step": 12970
+    },
+    {
+      "epoch": 2.098456066607388,
+      "grad_norm": 0.7734767198562622,
+      "learning_rate": 0.0002,
+      "loss": 0.6093,
+      "step": 12980
+    },
+    {
+      "epoch": 2.1000727507881334,
+      "grad_norm": 0.7618662118911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 12990
+    },
+    {
+      "epoch": 2.101689434968879,
+      "grad_norm": 0.6557944416999817,
+      "learning_rate": 0.0002,
+      "loss": 0.6707,
+      "step": 13000
+    },
+    {
+      "epoch": 2.1033061191496243,
+      "grad_norm": 0.8786448240280151,
+      "learning_rate": 0.0002,
+      "loss": 0.7268,
+      "step": 13010
+    },
+    {
+      "epoch": 2.1049228033303695,
+      "grad_norm": 0.6878724098205566,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 13020
+    },
+    {
+      "epoch": 2.1065394875111147,
+      "grad_norm": 0.822318971157074,
+      "learning_rate": 0.0002,
+      "loss": 0.6824,
+      "step": 13030
+    },
+    {
+      "epoch": 2.10815617169186,
+      "grad_norm": 0.831468939781189,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 13040
+    },
+    {
+      "epoch": 2.109772855872605,
+      "grad_norm": 0.7699505686759949,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 13050
+    },
+    {
+      "epoch": 2.1113895400533504,
+      "grad_norm": 0.7559016346931458,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1130062242340957,
+      "grad_norm": 0.6942209601402283,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 13070
+    },
+    {
+      "epoch": 2.1146229084148414,
+      "grad_norm": 0.6098947525024414,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 13080
+    },
+    {
+      "epoch": 2.1162395925955866,
+      "grad_norm": 0.6499016284942627,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 13090
+    },
+    {
+      "epoch": 2.117856276776332,
+      "grad_norm": 0.7719953060150146,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 13100
+    },
+    {
+      "epoch": 2.119472960957077,
+      "grad_norm": 0.6708134412765503,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1210896451378223,
+      "grad_norm": 0.8119585514068604,
+      "learning_rate": 0.0002,
+      "loss": 0.6056,
+      "step": 13120
+    },
+    {
+      "epoch": 2.1227063293185675,
+      "grad_norm": 0.6947157979011536,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 13130
+    },
+    {
+      "epoch": 2.1243230134993127,
+      "grad_norm": 0.8831837773323059,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 13140
+    },
+    {
+      "epoch": 2.1259396976800584,
+      "grad_norm": 0.7266910672187805,
+      "learning_rate": 0.0002,
+      "loss": 0.6997,
+      "step": 13150
+    },
+    {
+      "epoch": 2.1275563818608036,
+      "grad_norm": 0.8864351511001587,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 13160
+    },
+    {
+      "epoch": 2.129173066041549,
+      "grad_norm": 0.8104248046875,
+      "learning_rate": 0.0002,
+      "loss": 0.6762,
+      "step": 13170
+    },
+    {
+      "epoch": 2.130789750222294,
+      "grad_norm": 0.6077079772949219,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 13180
+    },
+    {
+      "epoch": 2.1324064344030393,
+      "grad_norm": 0.6874213814735413,
+      "learning_rate": 0.0002,
+      "loss": 0.6572,
+      "step": 13190
+    },
+    {
+      "epoch": 2.1340231185837846,
+      "grad_norm": 0.7134367823600769,
+      "learning_rate": 0.0002,
+      "loss": 0.642,
+      "step": 13200
+    },
+    {
+      "epoch": 2.13563980276453,
+      "grad_norm": 0.6101235151290894,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 13210
+    },
+    {
+      "epoch": 2.137256486945275,
+      "grad_norm": 0.6042411923408508,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13220
+    },
+    {
+      "epoch": 2.1388731711260207,
+      "grad_norm": 0.914601743221283,
+      "learning_rate": 0.0002,
+      "loss": 0.7179,
+      "step": 13230
+    },
+    {
+      "epoch": 2.140489855306766,
+      "grad_norm": 0.7104284167289734,
+      "learning_rate": 0.0002,
+      "loss": 0.6513,
+      "step": 13240
+    },
+    {
+      "epoch": 2.142106539487511,
+      "grad_norm": 0.664395272731781,
+      "learning_rate": 0.0002,
+      "loss": 0.6607,
+      "step": 13250
+    },
+    {
+      "epoch": 2.1437232236682564,
+      "grad_norm": 0.6991241574287415,
+      "learning_rate": 0.0002,
+      "loss": 0.7211,
+      "step": 13260
+    },
+    {
+      "epoch": 2.1453399078490016,
+      "grad_norm": 0.5469560623168945,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 13270
+    },
+    {
+      "epoch": 2.146956592029747,
+      "grad_norm": 0.8454998135566711,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13280
+    },
+    {
+      "epoch": 2.148573276210492,
+      "grad_norm": 0.7088868618011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 13290
+    },
+    {
+      "epoch": 2.1501899603912378,
+      "grad_norm": 0.7002687454223633,
+      "learning_rate": 0.0002,
+      "loss": 0.6835,
+      "step": 13300
+    },
+    {
+      "epoch": 2.151806644571983,
+      "grad_norm": 0.7785214781761169,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 13310
+    },
+    {
+      "epoch": 2.1534233287527282,
+      "grad_norm": 0.8049132227897644,
+      "learning_rate": 0.0002,
+      "loss": 0.67,
+      "step": 13320
+    },
+    {
+      "epoch": 2.1550400129334735,
+      "grad_norm": 0.8062595129013062,
+      "learning_rate": 0.0002,
+      "loss": 0.6495,
+      "step": 13330
+    },
+    {
+      "epoch": 2.1566566971142187,
+      "grad_norm": 0.6208319067955017,
+      "learning_rate": 0.0002,
+      "loss": 0.6603,
+      "step": 13340
+    },
+    {
+      "epoch": 2.158273381294964,
+      "grad_norm": 0.7519655823707581,
+      "learning_rate": 0.0002,
+      "loss": 0.6584,
+      "step": 13350
+    },
+    {
+      "epoch": 2.159890065475709,
+      "grad_norm": 0.7645747065544128,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 13360
+    },
+    {
+      "epoch": 2.1615067496564544,
+      "grad_norm": 0.6847302913665771,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 13370
+    },
+    {
+      "epoch": 2.1631234338372,
+      "grad_norm": 0.8630441427230835,
+      "learning_rate": 0.0002,
+      "loss": 0.6903,
+      "step": 13380
+    },
+    {
+      "epoch": 2.1647401180179453,
+      "grad_norm": 0.7947702407836914,
+      "learning_rate": 0.0002,
+      "loss": 0.6742,
+      "step": 13390
+    },
+    {
+      "epoch": 2.1663568021986905,
+      "grad_norm": 0.6836977005004883,
+      "learning_rate": 0.0002,
+      "loss": 0.7206,
+      "step": 13400
+    },
+    {
+      "epoch": 2.1679734863794358,
+      "grad_norm": 0.7340566515922546,
+      "learning_rate": 0.0002,
+      "loss": 0.6304,
+      "step": 13410
+    },
+    {
+      "epoch": 2.169590170560181,
+      "grad_norm": 0.7075738906860352,
+      "learning_rate": 0.0002,
+      "loss": 0.6528,
+      "step": 13420
+    },
+    {
+      "epoch": 2.1712068547409262,
+      "grad_norm": 0.7080879807472229,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 13430
+    },
+    {
+      "epoch": 2.1728235389216715,
+      "grad_norm": 0.6218613386154175,
+      "learning_rate": 0.0002,
+      "loss": 0.6615,
+      "step": 13440
+    },
+    {
+      "epoch": 2.174440223102417,
+      "grad_norm": 0.8211479187011719,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 13450
+    },
+    {
+      "epoch": 2.1760569072831624,
+      "grad_norm": 0.864466667175293,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 13460
+    },
+    {
+      "epoch": 2.1776735914639076,
+      "grad_norm": 0.7943857908248901,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 13470
+    },
+    {
+      "epoch": 2.179290275644653,
+      "grad_norm": 0.78728187084198,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 13480
+    },
+    {
+      "epoch": 2.180906959825398,
+      "grad_norm": 0.697527289390564,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 13490
+    },
+    {
+      "epoch": 2.1825236440061433,
+      "grad_norm": 0.8205804228782654,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 13500
+    },
+    {
+      "epoch": 2.1841403281868885,
+      "grad_norm": 0.8709042072296143,
+      "learning_rate": 0.0002,
+      "loss": 0.7227,
+      "step": 13510
+    },
+    {
+      "epoch": 2.1857570123676338,
+      "grad_norm": 0.6228537559509277,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 13520
+    },
+    {
+      "epoch": 2.1873736965483794,
+      "grad_norm": 0.9566980004310608,
+      "learning_rate": 0.0002,
+      "loss": 0.7025,
+      "step": 13530
+    },
+    {
+      "epoch": 2.1889903807291247,
+      "grad_norm": 0.7128894329071045,
+      "learning_rate": 0.0002,
+      "loss": 0.6755,
+      "step": 13540
+    },
+    {
+      "epoch": 2.19060706490987,
+      "grad_norm": 0.6888654232025146,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 13550
+    },
+    {
+      "epoch": 2.192223749090615,
+      "grad_norm": 0.6444337368011475,
+      "learning_rate": 0.0002,
+      "loss": 0.6961,
+      "step": 13560
+    },
+    {
+      "epoch": 2.1938404332713604,
+      "grad_norm": 0.8008806705474854,
+      "learning_rate": 0.0002,
+      "loss": 0.656,
+      "step": 13570
+    },
+    {
+      "epoch": 2.1954571174521056,
+      "grad_norm": 0.8482748866081238,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 13580
+    },
+    {
+      "epoch": 2.197073801632851,
+      "grad_norm": 0.8584157228469849,
+      "learning_rate": 0.0002,
+      "loss": 0.7326,
+      "step": 13590
+    },
+    {
+      "epoch": 2.1986904858135965,
+      "grad_norm": 0.7513734698295593,
+      "learning_rate": 0.0002,
+      "loss": 0.7014,
+      "step": 13600
+    },
+    {
+      "epoch": 2.2003071699943417,
+      "grad_norm": 0.7864262461662292,
+      "learning_rate": 0.0002,
+      "loss": 0.6632,
+      "step": 13610
+    },
+    {
+      "epoch": 2.201923854175087,
+      "grad_norm": 0.8493645191192627,
+      "learning_rate": 0.0002,
+      "loss": 0.6879,
+      "step": 13620
+    },
+    {
+      "epoch": 2.203540538355832,
+      "grad_norm": 0.6902140974998474,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 13630
+    },
+    {
+      "epoch": 2.2051572225365774,
+      "grad_norm": 0.8711254596710205,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 13640
+    },
+    {
+      "epoch": 2.2067739067173227,
+      "grad_norm": 0.7832191586494446,
+      "learning_rate": 0.0002,
+      "loss": 0.6359,
+      "step": 13650
+    },
+    {
+      "epoch": 2.208390590898068,
+      "grad_norm": 0.5668176412582397,
+      "learning_rate": 0.0002,
+      "loss": 0.6723,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2100072750788136,
+      "grad_norm": 0.8648375272750854,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13670
+    },
+    {
+      "epoch": 2.211623959259559,
+      "grad_norm": 0.7643089890480042,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13680
+    },
+    {
+      "epoch": 2.213240643440304,
+      "grad_norm": 0.6293777823448181,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 13690
+    },
+    {
+      "epoch": 2.2148573276210493,
+      "grad_norm": 0.6459372639656067,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 13700
+    },
+    {
+      "epoch": 2.2164740118017945,
+      "grad_norm": 0.7060744166374207,
+      "learning_rate": 0.0002,
+      "loss": 0.6526,
+      "step": 13710
+    },
+    {
+      "epoch": 2.2180906959825397,
+      "grad_norm": 0.674109160900116,
+      "learning_rate": 0.0002,
+      "loss": 0.7101,
+      "step": 13720
+    },
+    {
+      "epoch": 2.219707380163285,
+      "grad_norm": 0.830392062664032,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 13730
+    },
+    {
+      "epoch": 2.2213240643440306,
+      "grad_norm": 0.6474477052688599,
+      "learning_rate": 0.0002,
+      "loss": 0.6733,
+      "step": 13740
+    },
+    {
+      "epoch": 2.222940748524776,
+      "grad_norm": 0.7037909626960754,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13750
+    },
+    {
+      "epoch": 2.224557432705521,
+      "grad_norm": 0.6554131507873535,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 13760
+    },
+    {
+      "epoch": 2.2261741168862663,
+      "grad_norm": 0.7822230458259583,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 13770
+    },
+    {
+      "epoch": 2.2277908010670116,
+      "grad_norm": 0.9082167744636536,
+      "learning_rate": 0.0002,
+      "loss": 0.6505,
+      "step": 13780
+    },
+    {
+      "epoch": 2.229407485247757,
+      "grad_norm": 0.7918276190757751,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 13790
+    },
+    {
+      "epoch": 2.231024169428502,
+      "grad_norm": 0.7354569435119629,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 13800
+    },
+    {
+      "epoch": 2.2326408536092472,
+      "grad_norm": 0.8265249133110046,
+      "learning_rate": 0.0002,
+      "loss": 0.6503,
+      "step": 13810
+    },
+    {
+      "epoch": 2.234257537789993,
+      "grad_norm": 0.6653847098350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6871,
+      "step": 13820
+    },
+    {
+      "epoch": 2.235874221970738,
+      "grad_norm": 0.7157923579216003,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 13830
+    },
+    {
+      "epoch": 2.2374909061514834,
+      "grad_norm": 0.7110323309898376,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 13840
+    },
+    {
+      "epoch": 2.2391075903322286,
+      "grad_norm": 0.7155357599258423,
+      "learning_rate": 0.0002,
+      "loss": 0.6913,
+      "step": 13850
+    },
+    {
+      "epoch": 2.240724274512974,
+      "grad_norm": 1.0177817344665527,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 13860
+    },
+    {
+      "epoch": 2.242340958693719,
+      "grad_norm": 0.7601948380470276,
+      "learning_rate": 0.0002,
+      "loss": 0.635,
+      "step": 13870
+    },
+    {
+      "epoch": 2.2439576428744643,
+      "grad_norm": 0.7628820538520813,
+      "learning_rate": 0.0002,
+      "loss": 0.6679,
+      "step": 13880
+    },
+    {
+      "epoch": 2.24557432705521,
+      "grad_norm": 0.7089297771453857,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 13890
+    },
+    {
+      "epoch": 2.247191011235955,
+      "grad_norm": 0.695178210735321,
+      "learning_rate": 0.0002,
+      "loss": 0.7236,
+      "step": 13900
+    },
+    {
+      "epoch": 2.2488076954167004,
+      "grad_norm": 0.7631948590278625,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 13910
+    },
+    {
+      "epoch": 2.2504243795974457,
+      "grad_norm": 0.8203101754188538,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 13920
+    },
+    {
+      "epoch": 2.252041063778191,
+      "grad_norm": 0.8099079728126526,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 13930
+    },
+    {
+      "epoch": 2.253657747958936,
+      "grad_norm": 0.6498546004295349,
+      "learning_rate": 0.0002,
+      "loss": 0.694,
+      "step": 13940
+    },
+    {
+      "epoch": 2.2552744321396814,
+      "grad_norm": 0.7797415256500244,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 13950
+    },
+    {
+      "epoch": 2.2568911163204266,
+      "grad_norm": 0.8254124522209167,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 13960
+    },
+    {
+      "epoch": 2.2585078005011723,
+      "grad_norm": 0.6327953338623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 13970
+    },
+    {
+      "epoch": 2.2601244846819175,
+      "grad_norm": 0.734194278717041,
+      "learning_rate": 0.0002,
+      "loss": 0.668,
+      "step": 13980
+    },
+    {
+      "epoch": 2.2617411688626627,
+      "grad_norm": 0.9014202952384949,
+      "learning_rate": 0.0002,
+      "loss": 0.6912,
+      "step": 13990
+    },
+    {
+      "epoch": 2.263357853043408,
+      "grad_norm": 0.7643631100654602,
+      "learning_rate": 0.0002,
+      "loss": 0.692,
+      "step": 14000
+    },
+    {
+      "epoch": 2.264974537224153,
+      "grad_norm": 0.8882834911346436,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 14010
+    },
+    {
+      "epoch": 2.2665912214048984,
+      "grad_norm": 0.7975873351097107,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14020
+    },
+    {
+      "epoch": 2.2682079055856437,
+      "grad_norm": 0.7765783071517944,
+      "learning_rate": 0.0002,
+      "loss": 0.7193,
+      "step": 14030
+    },
+    {
+      "epoch": 2.2698245897663893,
+      "grad_norm": 0.8846288323402405,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 14040
+    },
+    {
+      "epoch": 2.2714412739471346,
+      "grad_norm": 0.9006744027137756,
+      "learning_rate": 0.0002,
+      "loss": 0.6494,
+      "step": 14050
+    },
+    {
+      "epoch": 2.27305795812788,
+      "grad_norm": 0.7420173287391663,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 14060
+    },
+    {
+      "epoch": 2.274674642308625,
+      "grad_norm": 0.7956424951553345,
+      "learning_rate": 0.0002,
+      "loss": 0.7068,
+      "step": 14070
+    },
+    {
+      "epoch": 2.2762913264893703,
+      "grad_norm": 0.7783209085464478,
+      "learning_rate": 0.0002,
+      "loss": 0.6581,
+      "step": 14080
+    },
+    {
+      "epoch": 2.2779080106701155,
+      "grad_norm": 0.7597188949584961,
+      "learning_rate": 0.0002,
+      "loss": 0.7202,
+      "step": 14090
+    },
+    {
+      "epoch": 2.2795246948508607,
+      "grad_norm": 0.6718921661376953,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14100
+    },
+    {
+      "epoch": 2.281141379031606,
+      "grad_norm": 0.7528082132339478,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 14110
+    },
+    {
+      "epoch": 2.2827580632123516,
+      "grad_norm": 0.8379864692687988,
+      "learning_rate": 0.0002,
+      "loss": 0.7608,
+      "step": 14120
+    },
+    {
+      "epoch": 2.284374747393097,
+      "grad_norm": 0.748613715171814,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 14130
+    },
+    {
+      "epoch": 2.285991431573842,
+      "grad_norm": 0.7435423135757446,
+      "learning_rate": 0.0002,
+      "loss": 0.6641,
+      "step": 14140
+    },
+    {
+      "epoch": 2.2876081157545873,
+      "grad_norm": 0.7580803632736206,
+      "learning_rate": 0.0002,
+      "loss": 0.6849,
+      "step": 14150
+    },
+    {
+      "epoch": 2.2892247999353326,
+      "grad_norm": 0.6278321146965027,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 14160
+    },
+    {
+      "epoch": 2.290841484116078,
+      "grad_norm": 0.7663896083831787,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 14170
+    },
+    {
+      "epoch": 2.292458168296823,
+      "grad_norm": 0.9716812372207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 14180
+    },
+    {
+      "epoch": 2.2940748524775687,
+      "grad_norm": 0.8993458151817322,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14190
+    },
+    {
+      "epoch": 2.295691536658314,
+      "grad_norm": 0.6156117916107178,
+      "learning_rate": 0.0002,
+      "loss": 0.6172,
+      "step": 14200
+    },
+    {
+      "epoch": 2.297308220839059,
+      "grad_norm": 0.8911278247833252,
+      "learning_rate": 0.0002,
+      "loss": 0.6318,
+      "step": 14210
+    },
+    {
+      "epoch": 2.2989249050198044,
+      "grad_norm": 0.6422147154808044,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 14220
+    },
+    {
+      "epoch": 2.3005415892005496,
+      "grad_norm": 0.6866879463195801,
+      "learning_rate": 0.0002,
+      "loss": 0.6795,
+      "step": 14230
+    },
+    {
+      "epoch": 2.302158273381295,
+      "grad_norm": 0.9297130107879639,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 14240
+    },
+    {
+      "epoch": 2.30377495756204,
+      "grad_norm": 0.7501356601715088,
+      "learning_rate": 0.0002,
+      "loss": 0.6823,
+      "step": 14250
+    },
+    {
+      "epoch": 2.3053916417427853,
+      "grad_norm": 0.8363515138626099,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 14260
+    },
+    {
+      "epoch": 2.307008325923531,
+      "grad_norm": 0.9083868265151978,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3086250101042762,
+      "grad_norm": 0.7791516780853271,
+      "learning_rate": 0.0002,
+      "loss": 0.6862,
+      "step": 14280
+    },
+    {
+      "epoch": 2.3102416942850215,
+      "grad_norm": 0.8766953349113464,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14290
+    },
+    {
+      "epoch": 2.3118583784657667,
+      "grad_norm": 0.7916635274887085,
+      "learning_rate": 0.0002,
+      "loss": 0.6698,
+      "step": 14300
+    },
+    {
+      "epoch": 2.313475062646512,
+      "grad_norm": 0.627525269985199,
+      "learning_rate": 0.0002,
+      "loss": 0.6927,
+      "step": 14310
+    },
+    {
+      "epoch": 2.315091746827257,
+      "grad_norm": 0.8856783509254456,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 14320
+    },
+    {
+      "epoch": 2.316708431008003,
+      "grad_norm": 0.6758689284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 14330
+    },
+    {
+      "epoch": 2.318325115188748,
+      "grad_norm": 0.6428321003913879,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 14340
+    },
+    {
+      "epoch": 2.3199417993694933,
+      "grad_norm": 0.9032121300697327,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 14350
+    },
+    {
+      "epoch": 2.3215584835502385,
+      "grad_norm": 0.8035986423492432,
+      "learning_rate": 0.0002,
+      "loss": 0.6569,
+      "step": 14360
+    },
+    {
+      "epoch": 2.3231751677309838,
+      "grad_norm": 0.7974579334259033,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 14370
+    },
+    {
+      "epoch": 2.324791851911729,
+      "grad_norm": 0.8356034755706787,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 14380
+    },
+    {
+      "epoch": 2.326408536092474,
+      "grad_norm": 0.998760998249054,
+      "learning_rate": 0.0002,
+      "loss": 0.6623,
+      "step": 14390
+    },
+    {
+      "epoch": 2.3280252202732195,
+      "grad_norm": 0.6518142223358154,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 14400
+    },
+    {
+      "epoch": 2.3296419044539647,
+      "grad_norm": 0.7443506717681885,
+      "learning_rate": 0.0002,
+      "loss": 0.7146,
+      "step": 14410
+    },
+    {
+      "epoch": 2.3312585886347104,
+      "grad_norm": 0.8436172604560852,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14420
+    },
+    {
+      "epoch": 2.3328752728154556,
+      "grad_norm": 0.7411080598831177,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 14430
+    },
+    {
+      "epoch": 2.334491956996201,
+      "grad_norm": 0.8839048743247986,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 14440
+    },
+    {
+      "epoch": 2.336108641176946,
+      "grad_norm": 0.8360885977745056,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 14450
+    },
+    {
+      "epoch": 2.3377253253576913,
+      "grad_norm": 0.7608986496925354,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 14460
+    },
+    {
+      "epoch": 2.3393420095384365,
+      "grad_norm": 0.8179867267608643,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14470
+    },
+    {
+      "epoch": 2.340958693719182,
+      "grad_norm": 0.5989999771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14480
+    },
+    {
+      "epoch": 2.3425753778999274,
+      "grad_norm": 0.9450054168701172,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 14490
+    },
+    {
+      "epoch": 2.3441920620806727,
+      "grad_norm": 0.7885149717330933,
+      "learning_rate": 0.0002,
+      "loss": 0.6983,
+      "step": 14500
+    },
+    {
+      "epoch": 2.345808746261418,
+      "grad_norm": 0.8152616620063782,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 14510
+    },
+    {
+      "epoch": 2.347425430442163,
+      "grad_norm": 0.7193838953971863,
+      "learning_rate": 0.0002,
+      "loss": 0.6989,
+      "step": 14520
+    },
+    {
+      "epoch": 2.3490421146229084,
+      "grad_norm": 0.6701092720031738,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 14530
+    },
+    {
+      "epoch": 2.3506587988036536,
+      "grad_norm": 0.7529364228248596,
+      "learning_rate": 0.0002,
+      "loss": 0.6559,
+      "step": 14540
+    },
+    {
+      "epoch": 2.352275482984399,
+      "grad_norm": 0.6599733829498291,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 14550
+    },
+    {
+      "epoch": 2.353892167165144,
+      "grad_norm": 0.9502474069595337,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14560
+    },
+    {
+      "epoch": 2.3555088513458897,
+      "grad_norm": 0.7619650959968567,
+      "learning_rate": 0.0002,
+      "loss": 0.717,
+      "step": 14570
+    },
+    {
+      "epoch": 2.357125535526635,
+      "grad_norm": 0.9854652285575867,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14580
+    },
+    {
+      "epoch": 2.35874221970738,
+      "grad_norm": 0.727439284324646,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 14590
+    },
+    {
+      "epoch": 2.3603589038881254,
+      "grad_norm": 0.6994746327400208,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 14600
+    },
+    {
+      "epoch": 2.3619755880688706,
+      "grad_norm": 0.7117531299591064,
+      "learning_rate": 0.0002,
+      "loss": 0.6587,
+      "step": 14610
+    },
+    {
+      "epoch": 2.363592272249616,
+      "grad_norm": 0.6403067708015442,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 14620
+    },
+    {
+      "epoch": 2.3652089564303616,
+      "grad_norm": 0.8377841711044312,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 14630
+    },
+    {
+      "epoch": 2.366825640611107,
+      "grad_norm": 0.749171257019043,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 14640
+    },
+    {
+      "epoch": 2.368442324791852,
+      "grad_norm": 0.8418586254119873,
+      "learning_rate": 0.0002,
+      "loss": 0.6552,
+      "step": 14650
+    },
+    {
+      "epoch": 2.3700590089725972,
+      "grad_norm": 0.6178573369979858,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 14660
+    },
+    {
+      "epoch": 2.3716756931533425,
+      "grad_norm": 0.6368302702903748,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 14670
+    },
+    {
+      "epoch": 2.3732923773340877,
+      "grad_norm": 0.9122977256774902,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14680
+    },
+    {
+      "epoch": 2.374909061514833,
+      "grad_norm": 0.7086195349693298,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 14690
+    },
+    {
+      "epoch": 2.376525745695578,
+      "grad_norm": 0.7500800490379333,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 14700
+    },
+    {
+      "epoch": 2.378142429876324,
+      "grad_norm": 0.6634900569915771,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 14710
+    },
+    {
+      "epoch": 2.379759114057069,
+      "grad_norm": 0.839898407459259,
+      "learning_rate": 0.0002,
+      "loss": 0.6614,
+      "step": 14720
+    },
+    {
+      "epoch": 2.3813757982378143,
+      "grad_norm": 0.7578426003456116,
+      "learning_rate": 0.0002,
+      "loss": 0.6453,
+      "step": 14730
+    },
+    {
+      "epoch": 2.3829924824185595,
+      "grad_norm": 1.0213173627853394,
+      "learning_rate": 0.0002,
+      "loss": 0.7282,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3846091665993048,
+      "grad_norm": 0.7855949401855469,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 14750
+    },
+    {
+      "epoch": 2.38622585078005,
+      "grad_norm": 0.7224128842353821,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 14760
+    },
+    {
+      "epoch": 2.3878425349607952,
+      "grad_norm": 0.8040381669998169,
+      "learning_rate": 0.0002,
+      "loss": 0.7017,
+      "step": 14770
+    },
+    {
+      "epoch": 2.389459219141541,
+      "grad_norm": 0.7705281376838684,
+      "learning_rate": 0.0002,
+      "loss": 0.6799,
+      "step": 14780
+    },
+    {
+      "epoch": 2.391075903322286,
+      "grad_norm": 0.667966902256012,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 14790
+    },
+    {
+      "epoch": 2.3926925875030314,
+      "grad_norm": 0.6611011028289795,
+      "learning_rate": 0.0002,
+      "loss": 0.7061,
+      "step": 14800
+    },
+    {
+      "epoch": 2.3943092716837766,
+      "grad_norm": 0.6862651705741882,
+      "learning_rate": 0.0002,
+      "loss": 0.6527,
+      "step": 14810
+    },
+    {
+      "epoch": 2.395925955864522,
+      "grad_norm": 0.8086010217666626,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 14820
+    },
+    {
+      "epoch": 2.397542640045267,
+      "grad_norm": 0.7189689874649048,
+      "learning_rate": 0.0002,
+      "loss": 0.7189,
+      "step": 14830
+    },
+    {
+      "epoch": 2.3991593242260123,
+      "grad_norm": 0.6280009150505066,
+      "learning_rate": 0.0002,
+      "loss": 0.6709,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4007760084067575,
+      "grad_norm": 0.7826612591743469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 14850
+    },
+    {
+      "epoch": 2.402392692587503,
+      "grad_norm": 0.7681610584259033,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4040093767682484,
+      "grad_norm": 0.720966100692749,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 14870
+    },
+    {
+      "epoch": 2.4056260609489937,
+      "grad_norm": 0.8202250599861145,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 14880
+    },
+    {
+      "epoch": 2.407242745129739,
+      "grad_norm": 0.786212682723999,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 14890
+    },
+    {
+      "epoch": 2.408859429310484,
+      "grad_norm": 0.6647164821624756,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 14900
+    },
+    {
+      "epoch": 2.4104761134912294,
+      "grad_norm": 0.7566399574279785,
+      "learning_rate": 0.0002,
+      "loss": 0.6706,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4120927976719746,
+      "grad_norm": 0.748814582824707,
+      "learning_rate": 0.0002,
+      "loss": 0.7188,
+      "step": 14920
+    },
+    {
+      "epoch": 2.4137094818527203,
+      "grad_norm": 0.7624038457870483,
+      "learning_rate": 0.0002,
+      "loss": 0.6684,
+      "step": 14930
+    },
+    {
+      "epoch": 2.4153261660334655,
+      "grad_norm": 0.8267335295677185,
+      "learning_rate": 0.0002,
+      "loss": 0.6483,
+      "step": 14940
+    },
+    {
+      "epoch": 2.4169428502142107,
+      "grad_norm": 0.8785360455513,
+      "learning_rate": 0.0002,
+      "loss": 0.6612,
+      "step": 14950
+    },
+    {
+      "epoch": 2.418559534394956,
+      "grad_norm": 0.679887592792511,
+      "learning_rate": 0.0002,
+      "loss": 0.6718,
+      "step": 14960
+    },
+    {
+      "epoch": 2.420176218575701,
+      "grad_norm": 0.7218474745750427,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4217929027564464,
+      "grad_norm": 0.6342799663543701,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 14980
+    },
+    {
+      "epoch": 2.4234095869371917,
+      "grad_norm": 0.7098712921142578,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 14990
+    },
+    {
+      "epoch": 2.425026271117937,
+      "grad_norm": 0.7497431635856628,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4266429552986826,
+      "grad_norm": 0.934836208820343,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15010
+    },
+    {
+      "epoch": 2.428259639479428,
+      "grad_norm": 0.8430966734886169,
+      "learning_rate": 0.0002,
+      "loss": 0.7221,
+      "step": 15020
+    },
+    {
+      "epoch": 2.429876323660173,
+      "grad_norm": 0.7032104730606079,
+      "learning_rate": 0.0002,
+      "loss": 0.6985,
+      "step": 15030
+    },
+    {
+      "epoch": 2.4314930078409183,
+      "grad_norm": 0.7746111750602722,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 15040
+    },
+    {
+      "epoch": 2.4331096920216635,
+      "grad_norm": 0.7661406397819519,
+      "learning_rate": 0.0002,
+      "loss": 0.7177,
+      "step": 15050
+    },
+    {
+      "epoch": 2.4347263762024087,
+      "grad_norm": 0.6941645741462708,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 15060
+    },
+    {
+      "epoch": 2.436343060383154,
+      "grad_norm": 0.7487249374389648,
+      "learning_rate": 0.0002,
+      "loss": 0.6421,
+      "step": 15070
+    },
+    {
+      "epoch": 2.4379597445638996,
+      "grad_norm": 0.7639912962913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 15080
+    },
+    {
+      "epoch": 2.439576428744645,
+      "grad_norm": 0.7708953619003296,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 15090
+    },
+    {
+      "epoch": 2.44119311292539,
+      "grad_norm": 0.9135832190513611,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15100
+    },
+    {
+      "epoch": 2.4428097971061353,
+      "grad_norm": 0.8283005356788635,
+      "learning_rate": 0.0002,
+      "loss": 0.672,
+      "step": 15110
+    },
+    {
+      "epoch": 2.4444264812868806,
+      "grad_norm": 0.925299346446991,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 15120
+    },
+    {
+      "epoch": 2.446043165467626,
+      "grad_norm": 0.7013528943061829,
+      "learning_rate": 0.0002,
+      "loss": 0.687,
+      "step": 15130
+    },
+    {
+      "epoch": 2.447659849648371,
+      "grad_norm": 0.622303307056427,
+      "learning_rate": 0.0002,
+      "loss": 0.6842,
+      "step": 15140
+    },
+    {
+      "epoch": 2.4492765338291163,
+      "grad_norm": 0.876569390296936,
+      "learning_rate": 0.0002,
+      "loss": 0.6676,
+      "step": 15150
+    },
+    {
+      "epoch": 2.450893218009862,
+      "grad_norm": 0.6836351752281189,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 15160
+    },
+    {
+      "epoch": 2.452509902190607,
+      "grad_norm": 0.7886684536933899,
+      "learning_rate": 0.0002,
+      "loss": 0.6781,
+      "step": 15170
+    },
+    {
+      "epoch": 2.4541265863713524,
+      "grad_norm": 0.6647440791130066,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 15180
+    },
+    {
+      "epoch": 2.4557432705520976,
+      "grad_norm": 0.7477722764015198,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 15190
+    },
+    {
+      "epoch": 2.457359954732843,
+      "grad_norm": 0.8192033767700195,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 15200
+    },
+    {
+      "epoch": 2.458976638913588,
+      "grad_norm": 0.847537100315094,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 15210
+    },
+    {
+      "epoch": 2.4605933230943338,
+      "grad_norm": 0.9027776122093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 15220
+    },
+    {
+      "epoch": 2.462210007275079,
+      "grad_norm": 0.7217772006988525,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 15230
+    },
+    {
+      "epoch": 2.4638266914558242,
+      "grad_norm": 0.7994546294212341,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 15240
+    },
+    {
+      "epoch": 2.4654433756365695,
+      "grad_norm": 0.939916729927063,
+      "learning_rate": 0.0002,
+      "loss": 0.6828,
+      "step": 15250
+    },
+    {
+      "epoch": 2.4670600598173147,
+      "grad_norm": 1.0009053945541382,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15260
+    },
+    {
+      "epoch": 2.46867674399806,
+      "grad_norm": 0.625555694103241,
+      "learning_rate": 0.0002,
+      "loss": 0.643,
+      "step": 15270
+    },
+    {
+      "epoch": 2.470293428178805,
+      "grad_norm": 0.7924878597259521,
+      "learning_rate": 0.0002,
+      "loss": 0.688,
+      "step": 15280
+    },
+    {
+      "epoch": 2.4719101123595504,
+      "grad_norm": 0.8536689877510071,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 15290
+    },
+    {
+      "epoch": 2.4735267965402956,
+      "grad_norm": 0.8572589755058289,
+      "learning_rate": 0.0002,
+      "loss": 0.6924,
+      "step": 15300
+    },
+    {
+      "epoch": 2.4751434807210413,
+      "grad_norm": 0.773279070854187,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 15310
+    },
+    {
+      "epoch": 2.4767601649017865,
+      "grad_norm": 0.7708749771118164,
+      "learning_rate": 0.0002,
+      "loss": 0.6573,
+      "step": 15320
+    },
+    {
+      "epoch": 2.4783768490825318,
+      "grad_norm": 0.770905077457428,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 15330
+    },
+    {
+      "epoch": 2.479993533263277,
+      "grad_norm": 0.8238571882247925,
+      "learning_rate": 0.0002,
+      "loss": 0.6878,
+      "step": 15340
+    },
+    {
+      "epoch": 2.481610217444022,
+      "grad_norm": 0.7670477032661438,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15350
+    },
+    {
+      "epoch": 2.4832269016247674,
+      "grad_norm": 0.905036985874176,
+      "learning_rate": 0.0002,
+      "loss": 0.7759,
+      "step": 15360
+    },
+    {
+      "epoch": 2.484843585805513,
+      "grad_norm": 0.6672089695930481,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 15370
+    },
+    {
+      "epoch": 2.4864602699862584,
+      "grad_norm": 0.625095784664154,
+      "learning_rate": 0.0002,
+      "loss": 0.6722,
+      "step": 15380
+    },
+    {
+      "epoch": 2.4880769541670036,
+      "grad_norm": 0.679772675037384,
+      "learning_rate": 0.0002,
+      "loss": 0.6396,
+      "step": 15390
+    },
+    {
+      "epoch": 2.489693638347749,
+      "grad_norm": 0.711492121219635,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 15400
+    },
+    {
+      "epoch": 2.491310322528494,
+      "grad_norm": 0.876189112663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6966,
+      "step": 15410
+    },
+    {
+      "epoch": 2.4929270067092393,
+      "grad_norm": 0.7236915230751038,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 15420
+    },
+    {
+      "epoch": 2.4945436908899845,
+      "grad_norm": 0.6629832983016968,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 15430
+    },
+    {
+      "epoch": 2.4961603750707297,
+      "grad_norm": 0.9756859540939331,
+      "learning_rate": 0.0002,
+      "loss": 0.6669,
+      "step": 15440
+    },
+    {
+      "epoch": 2.4977770592514754,
+      "grad_norm": 0.6896940469741821,
+      "learning_rate": 0.0002,
+      "loss": 0.7559,
+      "step": 15450
+    },
+    {
+      "epoch": 2.4993937434322206,
+      "grad_norm": 0.7105149626731873,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 15460
+    },
+    {
+      "epoch": 2.501010427612966,
+      "grad_norm": 0.8374546766281128,
+      "learning_rate": 0.0002,
+      "loss": 0.6859,
+      "step": 15470
+    },
+    {
+      "epoch": 2.502627111793711,
+      "grad_norm": 0.7320070266723633,
+      "learning_rate": 0.0002,
+      "loss": 0.6512,
+      "step": 15480
+    },
+    {
+      "epoch": 2.5042437959744563,
+      "grad_norm": 0.8306367993354797,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 15490
+    },
+    {
+      "epoch": 2.5058604801552016,
+      "grad_norm": 0.7472721338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.7253,
+      "step": 15500
+    },
+    {
+      "epoch": 2.507477164335947,
+      "grad_norm": 0.6147692203521729,
+      "learning_rate": 0.0002,
+      "loss": 0.6699,
+      "step": 15510
+    },
+    {
+      "epoch": 2.5090938485166925,
+      "grad_norm": 0.7788505554199219,
+      "learning_rate": 0.0002,
+      "loss": 0.7158,
+      "step": 15520
+    },
+    {
+      "epoch": 2.5107105326974377,
+      "grad_norm": 0.8807527422904968,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 15530
+    },
+    {
+      "epoch": 2.512327216878183,
+      "grad_norm": 0.7521643042564392,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 15540
+    },
+    {
+      "epoch": 2.513943901058928,
+      "grad_norm": 0.6900225281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 15550
+    },
+    {
+      "epoch": 2.5155605852396734,
+      "grad_norm": 0.6601938605308533,
+      "learning_rate": 0.0002,
+      "loss": 0.6769,
+      "step": 15560
+    },
+    {
+      "epoch": 2.5171772694204186,
+      "grad_norm": 0.8179984092712402,
+      "learning_rate": 0.0002,
+      "loss": 0.6648,
+      "step": 15570
+    },
+    {
+      "epoch": 2.518793953601164,
+      "grad_norm": 0.792556881904602,
+      "learning_rate": 0.0002,
+      "loss": 0.7028,
+      "step": 15580
+    },
+    {
+      "epoch": 2.520410637781909,
+      "grad_norm": 0.7081938982009888,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 15590
+    },
+    {
+      "epoch": 2.5220273219626543,
+      "grad_norm": 0.8733121156692505,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 15600
+    },
+    {
+      "epoch": 2.5236440061434,
+      "grad_norm": 0.7980992794036865,
+      "learning_rate": 0.0002,
+      "loss": 0.6969,
+      "step": 15610
+    },
+    {
+      "epoch": 2.5252606903241452,
+      "grad_norm": 0.883664071559906,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 15620
+    },
+    {
+      "epoch": 2.5268773745048905,
+      "grad_norm": 0.6963341236114502,
+      "learning_rate": 0.0002,
+      "loss": 0.7022,
+      "step": 15630
+    },
+    {
+      "epoch": 2.5284940586856357,
+      "grad_norm": 0.6433573365211487,
+      "learning_rate": 0.0002,
+      "loss": 0.7334,
+      "step": 15640
+    },
+    {
+      "epoch": 2.530110742866381,
+      "grad_norm": 0.8538183569908142,
+      "learning_rate": 0.0002,
+      "loss": 0.6889,
+      "step": 15650
+    },
+    {
+      "epoch": 2.5317274270471266,
+      "grad_norm": 0.9748201370239258,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 15660
+    },
+    {
+      "epoch": 2.533344111227872,
+      "grad_norm": 0.7670575380325317,
+      "learning_rate": 0.0002,
+      "loss": 0.6765,
+      "step": 15670
+    },
+    {
+      "epoch": 2.534960795408617,
+      "grad_norm": 0.8738890290260315,
+      "learning_rate": 0.0002,
+      "loss": 0.6435,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5365774795893623,
+      "grad_norm": 0.8391636610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.6802,
+      "step": 15690
+    },
+    {
+      "epoch": 2.5381941637701075,
+      "grad_norm": 0.7239366769790649,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 15700
+    },
+    {
+      "epoch": 2.5398108479508528,
+      "grad_norm": 0.8498379588127136,
+      "learning_rate": 0.0002,
+      "loss": 0.7011,
+      "step": 15710
+    },
+    {
+      "epoch": 2.541427532131598,
+      "grad_norm": 0.8029484152793884,
+      "learning_rate": 0.0002,
+      "loss": 0.6998,
+      "step": 15720
+    },
+    {
+      "epoch": 2.5430442163123432,
+      "grad_norm": 1.0639333724975586,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 15730
+    },
+    {
+      "epoch": 2.5446609004930885,
+      "grad_norm": 0.6401297450065613,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 15740
+    },
+    {
+      "epoch": 2.5462775846738337,
+      "grad_norm": 0.7123814821243286,
+      "learning_rate": 0.0002,
+      "loss": 0.7196,
+      "step": 15750
+    },
+    {
+      "epoch": 2.5478942688545794,
+      "grad_norm": 0.7874974608421326,
+      "learning_rate": 0.0002,
+      "loss": 0.654,
+      "step": 15760
+    },
+    {
+      "epoch": 2.5495109530353246,
+      "grad_norm": 0.8046808838844299,
+      "learning_rate": 0.0002,
+      "loss": 0.6721,
+      "step": 15770
+    },
+    {
+      "epoch": 2.55112763721607,
+      "grad_norm": 0.7888661623001099,
+      "learning_rate": 0.0002,
+      "loss": 0.6665,
+      "step": 15780
+    },
+    {
+      "epoch": 2.552744321396815,
+      "grad_norm": 0.8445866107940674,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 15790
+    },
+    {
+      "epoch": 2.5543610055775603,
+      "grad_norm": 0.7475846409797668,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 15800
+    },
+    {
+      "epoch": 2.555977689758306,
+      "grad_norm": 0.7455102801322937,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 15810
+    },
+    {
+      "epoch": 2.557594373939051,
+      "grad_norm": 0.8226983547210693,
+      "learning_rate": 0.0002,
+      "loss": 0.6932,
+      "step": 15820
+    },
+    {
+      "epoch": 2.5592110581197964,
+      "grad_norm": 0.8920368552207947,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 15830
+    },
+    {
+      "epoch": 2.5608277423005417,
+      "grad_norm": 0.8413904905319214,
+      "learning_rate": 0.0002,
+      "loss": 0.6297,
+      "step": 15840
+    },
+    {
+      "epoch": 2.562444426481287,
+      "grad_norm": 0.8483649492263794,
+      "learning_rate": 0.0002,
+      "loss": 0.7106,
+      "step": 15850
+    },
+    {
+      "epoch": 2.564061110662032,
+      "grad_norm": 0.5923284292221069,
+      "learning_rate": 0.0002,
+      "loss": 0.6957,
+      "step": 15860
+    },
+    {
+      "epoch": 2.5656777948427774,
+      "grad_norm": 0.8518726229667664,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 15870
+    },
+    {
+      "epoch": 2.5672944790235226,
+      "grad_norm": 0.731235146522522,
+      "learning_rate": 0.0002,
+      "loss": 0.6362,
+      "step": 15880
+    },
+    {
+      "epoch": 2.568911163204268,
+      "grad_norm": 0.7517194151878357,
+      "learning_rate": 0.0002,
+      "loss": 0.7611,
+      "step": 15890
+    },
+    {
+      "epoch": 2.5705278473850135,
+      "grad_norm": 0.8378692269325256,
+      "learning_rate": 0.0002,
+      "loss": 0.6907,
+      "step": 15900
+    },
+    {
+      "epoch": 2.5721445315657587,
+      "grad_norm": 0.843701958656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 15910
+    },
+    {
+      "epoch": 2.573761215746504,
+      "grad_norm": 0.7254629731178284,
+      "learning_rate": 0.0002,
+      "loss": 0.6882,
+      "step": 15920
+    },
+    {
+      "epoch": 2.575377899927249,
+      "grad_norm": 0.8863335847854614,
+      "learning_rate": 0.0002,
+      "loss": 0.6872,
+      "step": 15930
+    },
+    {
+      "epoch": 2.5769945841079944,
+      "grad_norm": 0.7675097584724426,
+      "learning_rate": 0.0002,
+      "loss": 0.6813,
+      "step": 15940
+    },
+    {
+      "epoch": 2.5786112682887397,
+      "grad_norm": 0.82063889503479,
+      "learning_rate": 0.0002,
+      "loss": 0.7357,
+      "step": 15950
+    },
+    {
+      "epoch": 2.5802279524694853,
+      "grad_norm": 0.7729717493057251,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 15960
+    },
+    {
+      "epoch": 2.5818446366502306,
+      "grad_norm": 0.8301846981048584,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 15970
+    },
+    {
+      "epoch": 2.583461320830976,
+      "grad_norm": 0.7906861305236816,
+      "learning_rate": 0.0002,
+      "loss": 0.6897,
+      "step": 15980
+    },
+    {
+      "epoch": 2.585078005011721,
+      "grad_norm": 0.6749057173728943,
+      "learning_rate": 0.0002,
+      "loss": 0.7175,
+      "step": 15990
+    },
+    {
+      "epoch": 2.5866946891924663,
+      "grad_norm": 0.9386842846870422,
+      "learning_rate": 0.0002,
+      "loss": 0.7212,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5883113733732115,
+      "grad_norm": 0.7868891358375549,
+      "learning_rate": 0.0002,
+      "loss": 0.6934,
+      "step": 16010
+    },
+    {
+      "epoch": 2.5899280575539567,
+      "grad_norm": 0.8674671053886414,
+      "learning_rate": 0.0002,
+      "loss": 0.7036,
+      "step": 16020
+    },
+    {
+      "epoch": 2.591544741734702,
+      "grad_norm": 0.7043559551239014,
+      "learning_rate": 0.0002,
+      "loss": 0.7217,
+      "step": 16030
+    },
+    {
+      "epoch": 2.593161425915447,
+      "grad_norm": 0.5846083760261536,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 16040
+    },
+    {
+      "epoch": 2.594778110096193,
+      "grad_norm": 0.7323982119560242,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 16050
+    },
+    {
+      "epoch": 2.596394794276938,
+      "grad_norm": 0.9069556593894958,
+      "learning_rate": 0.0002,
+      "loss": 0.6794,
+      "step": 16060
+    },
+    {
+      "epoch": 2.5980114784576833,
+      "grad_norm": 0.7522736191749573,
+      "learning_rate": 0.0002,
+      "loss": 0.7076,
+      "step": 16070
+    },
+    {
+      "epoch": 2.5996281626384286,
+      "grad_norm": 0.8149648308753967,
+      "learning_rate": 0.0002,
+      "loss": 0.6477,
+      "step": 16080
+    },
+    {
+      "epoch": 2.601244846819174,
+      "grad_norm": 0.6214233040809631,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 16090
+    },
+    {
+      "epoch": 2.602861530999919,
+      "grad_norm": 0.6803743839263916,
+      "learning_rate": 0.0002,
+      "loss": 0.7307,
+      "step": 16100
+    },
+    {
+      "epoch": 2.6044782151806647,
+      "grad_norm": 0.7223997116088867,
+      "learning_rate": 0.0002,
+      "loss": 0.7244,
+      "step": 16110
+    },
+    {
+      "epoch": 2.60609489936141,
+      "grad_norm": 0.7324174642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.6867,
+      "step": 16120
+    },
+    {
+      "epoch": 2.607711583542155,
+      "grad_norm": 0.9594739675521851,
+      "learning_rate": 0.0002,
+      "loss": 0.7159,
+      "step": 16130
+    },
+    {
+      "epoch": 2.6093282677229004,
+      "grad_norm": 0.9485327005386353,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 16140
+    },
+    {
+      "epoch": 2.6109449519036456,
+      "grad_norm": 0.8449000120162964,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 16150
+    },
+    {
+      "epoch": 2.612561636084391,
+      "grad_norm": 0.8520140051841736,
+      "learning_rate": 0.0002,
+      "loss": 0.7152,
+      "step": 16160
+    },
+    {
+      "epoch": 2.614178320265136,
+      "grad_norm": 0.7456524968147278,
+      "learning_rate": 0.0002,
+      "loss": 0.6759,
+      "step": 16170
+    },
+    {
+      "epoch": 2.6157950044458813,
+      "grad_norm": 0.9912857413291931,
+      "learning_rate": 0.0002,
+      "loss": 0.6893,
+      "step": 16180
+    },
+    {
+      "epoch": 2.6174116886266265,
+      "grad_norm": 0.9001946449279785,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 16190
+    },
+    {
+      "epoch": 2.619028372807372,
+      "grad_norm": 0.6568667888641357,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16200
+    },
+    {
+      "epoch": 2.6206450569881174,
+      "grad_norm": 1.0248128175735474,
+      "learning_rate": 0.0002,
+      "loss": 0.7013,
+      "step": 16210
+    },
+    {
+      "epoch": 2.6222617411688627,
+      "grad_norm": 0.6509039998054504,
+      "learning_rate": 0.0002,
+      "loss": 0.7045,
+      "step": 16220
+    },
+    {
+      "epoch": 2.623878425349608,
+      "grad_norm": 0.7626351118087769,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 16230
+    },
+    {
+      "epoch": 2.625495109530353,
+      "grad_norm": 0.6938552260398865,
+      "learning_rate": 0.0002,
+      "loss": 0.6556,
+      "step": 16240
+    },
+    {
+      "epoch": 2.6271117937110984,
+      "grad_norm": 0.6434680819511414,
+      "learning_rate": 0.0002,
+      "loss": 0.65,
+      "step": 16250
+    },
+    {
+      "epoch": 2.628728477891844,
+      "grad_norm": 0.7111515998840332,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 16260
+    },
+    {
+      "epoch": 2.6303451620725893,
+      "grad_norm": 0.7712395787239075,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 16270
+    },
+    {
+      "epoch": 2.6319618462533345,
+      "grad_norm": 0.792209267616272,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 16280
+    },
+    {
+      "epoch": 2.6335785304340797,
+      "grad_norm": 0.6801066398620605,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 16290
+    },
+    {
+      "epoch": 2.635195214614825,
+      "grad_norm": 0.7802573442459106,
+      "learning_rate": 0.0002,
+      "loss": 0.73,
+      "step": 16300
+    },
+    {
+      "epoch": 2.63681189879557,
+      "grad_norm": 0.7742244601249695,
+      "learning_rate": 0.0002,
+      "loss": 0.7484,
+      "step": 16310
+    },
+    {
+      "epoch": 2.6384285829763154,
+      "grad_norm": 0.664184033870697,
+      "learning_rate": 0.0002,
+      "loss": 0.6524,
+      "step": 16320
+    },
+    {
+      "epoch": 2.6400452671570607,
+      "grad_norm": 0.9242228865623474,
+      "learning_rate": 0.0002,
+      "loss": 0.6442,
+      "step": 16330
+    },
+    {
+      "epoch": 2.641661951337806,
+      "grad_norm": 0.9661325216293335,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 16340
+    },
+    {
+      "epoch": 2.6432786355185516,
+      "grad_norm": 0.837526798248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6847,
+      "step": 16350
+    },
+    {
+      "epoch": 2.644895319699297,
+      "grad_norm": 1.1834373474121094,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 16360
+    },
+    {
+      "epoch": 2.646512003880042,
+      "grad_norm": 0.7467831373214722,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 16370
+    },
+    {
+      "epoch": 2.6481286880607873,
+      "grad_norm": 0.8627146482467651,
+      "learning_rate": 0.0002,
+      "loss": 0.6935,
+      "step": 16380
+    },
+    {
+      "epoch": 2.6497453722415325,
+      "grad_norm": 0.790447473526001,
+      "learning_rate": 0.0002,
+      "loss": 0.715,
+      "step": 16390
+    },
+    {
+      "epoch": 2.651362056422278,
+      "grad_norm": 0.8447365164756775,
+      "learning_rate": 0.0002,
+      "loss": 0.723,
+      "step": 16400
+    },
+    {
+      "epoch": 2.6529787406030234,
+      "grad_norm": 0.7831417918205261,
+      "learning_rate": 0.0002,
+      "loss": 0.6628,
+      "step": 16410
+    },
+    {
+      "epoch": 2.6545954247837686,
+      "grad_norm": 0.6837952136993408,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 16420
+    },
+    {
+      "epoch": 2.656212108964514,
+      "grad_norm": 0.7031801342964172,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 16430
+    },
+    {
+      "epoch": 2.657828793145259,
+      "grad_norm": 0.8963770866394043,
+      "learning_rate": 0.0002,
+      "loss": 0.7382,
+      "step": 16440
+    },
+    {
+      "epoch": 2.6594454773260043,
+      "grad_norm": 0.6852328181266785,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 16450
+    },
+    {
+      "epoch": 2.6610621615067496,
+      "grad_norm": 0.8069294095039368,
+      "learning_rate": 0.0002,
+      "loss": 0.6278,
+      "step": 16460
+    },
+    {
+      "epoch": 2.662678845687495,
+      "grad_norm": 0.7503686547279358,
+      "learning_rate": 0.0002,
+      "loss": 0.6939,
+      "step": 16470
+    },
+    {
+      "epoch": 2.66429552986824,
+      "grad_norm": 0.6430956125259399,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16480
+    },
+    {
+      "epoch": 2.6659122140489853,
+      "grad_norm": 0.7894312739372253,
+      "learning_rate": 0.0002,
+      "loss": 0.6863,
+      "step": 16490
+    },
+    {
+      "epoch": 2.667528898229731,
+      "grad_norm": 0.7277431488037109,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 16500
+    },
+    {
+      "epoch": 2.669145582410476,
+      "grad_norm": 0.6816153526306152,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 16510
+    },
+    {
+      "epoch": 2.6707622665912214,
+      "grad_norm": 0.8145235776901245,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 16520
+    },
+    {
+      "epoch": 2.6723789507719666,
+      "grad_norm": 0.8645890355110168,
+      "learning_rate": 0.0002,
+      "loss": 0.709,
+      "step": 16530
+    },
+    {
+      "epoch": 2.673995634952712,
+      "grad_norm": 0.704393208026886,
+      "learning_rate": 0.0002,
+      "loss": 0.6946,
+      "step": 16540
+    },
+    {
+      "epoch": 2.6756123191334575,
+      "grad_norm": 1.0120846033096313,
+      "learning_rate": 0.0002,
+      "loss": 0.6378,
+      "step": 16550
+    },
+    {
+      "epoch": 2.6772290033142028,
+      "grad_norm": 0.6919328570365906,
+      "learning_rate": 0.0002,
+      "loss": 0.7241,
+      "step": 16560
+    },
+    {
+      "epoch": 2.678845687494948,
+      "grad_norm": 0.6924574971199036,
+      "learning_rate": 0.0002,
+      "loss": 0.7098,
+      "step": 16570
+    },
+    {
+      "epoch": 2.6804623716756932,
+      "grad_norm": 0.9679301381111145,
+      "learning_rate": 0.0002,
+      "loss": 0.731,
+      "step": 16580
+    },
+    {
+      "epoch": 2.6820790558564385,
+      "grad_norm": 0.6810211539268494,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 16590
+    },
+    {
+      "epoch": 2.6836957400371837,
+      "grad_norm": 0.9730555415153503,
+      "learning_rate": 0.0002,
+      "loss": 0.6688,
+      "step": 16600
+    },
+    {
+      "epoch": 2.685312424217929,
+      "grad_norm": 0.7852821350097656,
+      "learning_rate": 0.0002,
+      "loss": 0.7344,
+      "step": 16610
+    },
+    {
+      "epoch": 2.686929108398674,
+      "grad_norm": 0.6059057116508484,
+      "learning_rate": 0.0002,
+      "loss": 0.6401,
+      "step": 16620
+    },
+    {
+      "epoch": 2.6885457925794194,
+      "grad_norm": 0.9395958781242371,
+      "learning_rate": 0.0002,
+      "loss": 0.6796,
+      "step": 16630
+    },
+    {
+      "epoch": 2.690162476760165,
+      "grad_norm": 0.7473729848861694,
+      "learning_rate": 0.0002,
+      "loss": 0.7174,
+      "step": 16640
+    },
+    {
+      "epoch": 2.6917791609409103,
+      "grad_norm": 0.765934407711029,
+      "learning_rate": 0.0002,
+      "loss": 0.7087,
+      "step": 16650
+    },
+    {
+      "epoch": 2.6933958451216555,
+      "grad_norm": 0.8496677279472351,
+      "learning_rate": 0.0002,
+      "loss": 0.707,
+      "step": 16660
+    },
+    {
+      "epoch": 2.6950125293024008,
+      "grad_norm": 0.7641879916191101,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 16670
+    },
+    {
+      "epoch": 2.696629213483146,
+      "grad_norm": 0.8471952676773071,
+      "learning_rate": 0.0002,
+      "loss": 0.6566,
+      "step": 16680
+    },
+    {
+      "epoch": 2.6982458976638912,
+      "grad_norm": 0.6946060657501221,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 16690
+    },
+    {
+      "epoch": 2.699862581844637,
+      "grad_norm": 0.7361312508583069,
+      "learning_rate": 0.0002,
+      "loss": 0.7027,
+      "step": 16700
+    },
+    {
+      "epoch": 2.701479266025382,
+      "grad_norm": 0.6605038046836853,
+      "learning_rate": 0.0002,
+      "loss": 0.6767,
+      "step": 16710
+    },
+    {
+      "epoch": 2.7030959502061274,
+      "grad_norm": 0.7164411544799805,
+      "learning_rate": 0.0002,
+      "loss": 0.6885,
+      "step": 16720
+    },
+    {
+      "epoch": 2.7047126343868726,
+      "grad_norm": 0.6496201157569885,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 16730
+    },
+    {
+      "epoch": 2.706329318567618,
+      "grad_norm": 0.7826663851737976,
+      "learning_rate": 0.0002,
+      "loss": 0.6942,
+      "step": 16740
+    },
+    {
+      "epoch": 2.707946002748363,
+      "grad_norm": 0.7639131546020508,
+      "learning_rate": 0.0002,
+      "loss": 0.6773,
+      "step": 16750
+    },
+    {
+      "epoch": 2.7095626869291083,
+      "grad_norm": 0.7976210713386536,
+      "learning_rate": 0.0002,
+      "loss": 0.69,
+      "step": 16760
+    },
+    {
+      "epoch": 2.7111793711098535,
+      "grad_norm": 0.6836577653884888,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 16770
+    },
+    {
+      "epoch": 2.7127960552905988,
+      "grad_norm": 0.8025202751159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 16780
+    },
+    {
+      "epoch": 2.7144127394713444,
+      "grad_norm": 0.7636463642120361,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7160294236520897,
+      "grad_norm": 0.7481677532196045,
+      "learning_rate": 0.0002,
+      "loss": 0.6227,
+      "step": 16800
+    },
+    {
+      "epoch": 2.717646107832835,
+      "grad_norm": 0.7566834688186646,
+      "learning_rate": 0.0002,
+      "loss": 0.6925,
+      "step": 16810
+    },
+    {
+      "epoch": 2.71926279201358,
+      "grad_norm": 0.7931267619132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6531,
+      "step": 16820
+    },
+    {
+      "epoch": 2.7208794761943254,
+      "grad_norm": 0.8811662197113037,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 16830
+    },
+    {
+      "epoch": 2.7224961603750706,
+      "grad_norm": 0.8561240434646606,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 16840
+    },
+    {
+      "epoch": 2.7241128445558163,
+      "grad_norm": 0.7121599316596985,
+      "learning_rate": 0.0002,
+      "loss": 0.7135,
+      "step": 16850
+    },
+    {
+      "epoch": 2.7257295287365615,
+      "grad_norm": 0.8066257238388062,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 16860
+    },
+    {
+      "epoch": 2.7273462129173067,
+      "grad_norm": 0.7699271440505981,
+      "learning_rate": 0.0002,
+      "loss": 0.6839,
+      "step": 16870
+    },
+    {
+      "epoch": 2.728962897098052,
+      "grad_norm": 1.1828432083129883,
+      "learning_rate": 0.0002,
+      "loss": 0.699,
+      "step": 16880
+    },
+    {
+      "epoch": 2.730579581278797,
+      "grad_norm": 0.9989302754402161,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 16890
+    },
+    {
+      "epoch": 2.7321962654595424,
+      "grad_norm": 0.8100560307502747,
+      "learning_rate": 0.0002,
+      "loss": 0.7015,
+      "step": 16900
+    },
+    {
+      "epoch": 2.7338129496402876,
+      "grad_norm": 0.8615233898162842,
+      "learning_rate": 0.0002,
+      "loss": 0.6851,
+      "step": 16910
+    },
+    {
+      "epoch": 2.735429633821033,
+      "grad_norm": 0.8633756041526794,
+      "learning_rate": 0.0002,
+      "loss": 0.6322,
+      "step": 16920
+    },
+    {
+      "epoch": 2.737046318001778,
+      "grad_norm": 0.7769348621368408,
+      "learning_rate": 0.0002,
+      "loss": 0.6488,
+      "step": 16930
+    },
+    {
+      "epoch": 2.738663002182524,
+      "grad_norm": 0.6943058371543884,
+      "learning_rate": 0.0002,
+      "loss": 0.6582,
+      "step": 16940
+    },
+    {
+      "epoch": 2.740279686363269,
+      "grad_norm": 0.8510736227035522,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 16950
+    },
+    {
+      "epoch": 2.7418963705440142,
+      "grad_norm": 0.7732602953910828,
+      "learning_rate": 0.0002,
+      "loss": 0.7275,
+      "step": 16960
+    },
+    {
+      "epoch": 2.7435130547247595,
+      "grad_norm": 0.5981788635253906,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 16970
+    },
+    {
+      "epoch": 2.7451297389055047,
+      "grad_norm": 0.7604416012763977,
+      "learning_rate": 0.0002,
+      "loss": 0.6777,
+      "step": 16980
+    },
+    {
+      "epoch": 2.74674642308625,
+      "grad_norm": 0.7377738356590271,
+      "learning_rate": 0.0002,
+      "loss": 0.6981,
+      "step": 16990
+    },
+    {
+      "epoch": 2.7483631072669956,
+      "grad_norm": 0.9400289058685303,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 17000
+    },
+    {
+      "epoch": 2.749979791447741,
+      "grad_norm": 0.6340599656105042,
+      "learning_rate": 0.0002,
+      "loss": 0.6952,
+      "step": 17010
+    },
+    {
+      "epoch": 2.751596475628486,
+      "grad_norm": 0.7297601103782654,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 17020
+    },
+    {
+      "epoch": 2.7532131598092313,
+      "grad_norm": 0.9479979872703552,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 17030
+    },
+    {
+      "epoch": 2.7548298439899765,
+      "grad_norm": 0.8461511135101318,
+      "learning_rate": 0.0002,
+      "loss": 0.691,
+      "step": 17040
+    },
+    {
+      "epoch": 2.7564465281707218,
+      "grad_norm": 0.7477551698684692,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17050
+    },
+    {
+      "epoch": 2.758063212351467,
+      "grad_norm": 1.019270420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 17060
+    },
+    {
+      "epoch": 2.7596798965322122,
+      "grad_norm": 0.7730235457420349,
+      "learning_rate": 0.0002,
+      "loss": 0.7119,
+      "step": 17070
+    },
+    {
+      "epoch": 2.7612965807129575,
+      "grad_norm": 0.8216866254806519,
+      "learning_rate": 0.0002,
+      "loss": 0.6886,
+      "step": 17080
+    },
+    {
+      "epoch": 2.762913264893703,
+      "grad_norm": 0.7235931754112244,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17090
+    },
+    {
+      "epoch": 2.7645299490744484,
+      "grad_norm": 0.7352296710014343,
+      "learning_rate": 0.0002,
+      "loss": 0.7031,
+      "step": 17100
+    },
+    {
+      "epoch": 2.7661466332551936,
+      "grad_norm": 0.8129373788833618,
+      "learning_rate": 0.0002,
+      "loss": 0.6951,
+      "step": 17110
+    },
+    {
+      "epoch": 2.767763317435939,
+      "grad_norm": 0.7387019991874695,
+      "learning_rate": 0.0002,
+      "loss": 0.6703,
+      "step": 17120
+    },
+    {
+      "epoch": 2.769380001616684,
+      "grad_norm": 0.9149190187454224,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 17130
+    },
+    {
+      "epoch": 2.7709966857974297,
+      "grad_norm": 0.7352971434593201,
+      "learning_rate": 0.0002,
+      "loss": 0.6038,
+      "step": 17140
+    },
+    {
+      "epoch": 2.772613369978175,
+      "grad_norm": 0.7903780341148376,
+      "learning_rate": 0.0002,
+      "loss": 0.6728,
+      "step": 17150
+    },
+    {
+      "epoch": 2.77423005415892,
+      "grad_norm": 0.8255927562713623,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17160
+    },
+    {
+      "epoch": 2.7758467383396654,
+      "grad_norm": 0.7235927581787109,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 17170
+    },
+    {
+      "epoch": 2.7774634225204107,
+      "grad_norm": 0.8281434774398804,
+      "learning_rate": 0.0002,
+      "loss": 0.7161,
+      "step": 17180
+    },
+    {
+      "epoch": 2.779080106701156,
+      "grad_norm": 0.7586921453475952,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 17190
+    },
+    {
+      "epoch": 2.780696790881901,
+      "grad_norm": 0.7161715030670166,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 17200
+    },
+    {
+      "epoch": 2.7823134750626464,
+      "grad_norm": 0.762868344783783,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 17210
+    },
+    {
+      "epoch": 2.7839301592433916,
+      "grad_norm": 0.9285483360290527,
+      "learning_rate": 0.0002,
+      "loss": 0.705,
+      "step": 17220
+    },
+    {
+      "epoch": 2.785546843424137,
+      "grad_norm": 0.6900462508201599,
+      "learning_rate": 0.0002,
+      "loss": 0.7084,
+      "step": 17230
+    },
+    {
+      "epoch": 2.7871635276048825,
+      "grad_norm": 0.780384361743927,
+      "learning_rate": 0.0002,
+      "loss": 0.6988,
+      "step": 17240
+    },
+    {
+      "epoch": 2.7887802117856277,
+      "grad_norm": 0.7580406665802002,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17250
+    },
+    {
+      "epoch": 2.790396895966373,
+      "grad_norm": 0.8145199418067932,
+      "learning_rate": 0.0002,
+      "loss": 0.6833,
+      "step": 17260
+    },
+    {
+      "epoch": 2.792013580147118,
+      "grad_norm": 0.9159596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.6909,
+      "step": 17270
+    },
+    {
+      "epoch": 2.7936302643278634,
+      "grad_norm": 0.9590014219284058,
+      "learning_rate": 0.0002,
+      "loss": 0.6008,
+      "step": 17280
+    },
+    {
+      "epoch": 2.795246948508609,
+      "grad_norm": 0.7603529691696167,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 17290
+    },
+    {
+      "epoch": 2.7968636326893543,
+      "grad_norm": 0.8039976358413696,
+      "learning_rate": 0.0002,
+      "loss": 0.7165,
+      "step": 17300
+    },
+    {
+      "epoch": 2.7984803168700996,
+      "grad_norm": 0.8364847302436829,
+      "learning_rate": 0.0002,
+      "loss": 0.7037,
+      "step": 17310
+    },
+    {
+      "epoch": 2.800097001050845,
+      "grad_norm": 0.8763046860694885,
+      "learning_rate": 0.0002,
+      "loss": 0.6749,
+      "step": 17320
+    },
+    {
+      "epoch": 2.80171368523159,
+      "grad_norm": 0.8409647941589355,
+      "learning_rate": 0.0002,
+      "loss": 0.6844,
+      "step": 17330
+    },
+    {
+      "epoch": 2.8033303694123353,
+      "grad_norm": 0.7649006247520447,
+      "learning_rate": 0.0002,
+      "loss": 0.6936,
+      "step": 17340
+    },
+    {
+      "epoch": 2.8049470535930805,
+      "grad_norm": 0.7970262169837952,
+      "learning_rate": 0.0002,
+      "loss": 0.7051,
+      "step": 17350
+    },
+    {
+      "epoch": 2.8065637377738257,
+      "grad_norm": 0.9088607430458069,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 17360
+    },
+    {
+      "epoch": 2.808180421954571,
+      "grad_norm": 0.6454846858978271,
+      "learning_rate": 0.0002,
+      "loss": 0.675,
+      "step": 17370
+    },
+    {
+      "epoch": 2.809797106135316,
+      "grad_norm": 0.7744787931442261,
+      "learning_rate": 0.0002,
+      "loss": 0.7069,
+      "step": 17380
+    },
+    {
+      "epoch": 2.811413790316062,
+      "grad_norm": 0.6678640842437744,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 17390
+    },
+    {
+      "epoch": 2.813030474496807,
+      "grad_norm": 0.772676944732666,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 17400
+    },
+    {
+      "epoch": 2.8146471586775523,
+      "grad_norm": 0.7088175415992737,
+      "learning_rate": 0.0002,
+      "loss": 0.7252,
+      "step": 17410
+    },
+    {
+      "epoch": 2.8162638428582976,
+      "grad_norm": 0.8280573487281799,
+      "learning_rate": 0.0002,
+      "loss": 0.7086,
+      "step": 17420
+    },
+    {
+      "epoch": 2.817880527039043,
+      "grad_norm": 0.6665388345718384,
+      "learning_rate": 0.0002,
+      "loss": 0.6732,
+      "step": 17430
+    },
+    {
+      "epoch": 2.8194972112197885,
+      "grad_norm": 0.6427883505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 17440
+    },
+    {
+      "epoch": 2.8211138954005337,
+      "grad_norm": 0.9697760343551636,
+      "learning_rate": 0.0002,
+      "loss": 0.6972,
+      "step": 17450
+    },
+    {
+      "epoch": 2.822730579581279,
+      "grad_norm": 0.7573966383934021,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 17460
+    },
+    {
+      "epoch": 2.824347263762024,
+      "grad_norm": 0.878688633441925,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 17470
+    },
+    {
+      "epoch": 2.8259639479427694,
+      "grad_norm": 0.7752242684364319,
+      "learning_rate": 0.0002,
+      "loss": 0.6666,
+      "step": 17480
+    },
+    {
+      "epoch": 2.8275806321235146,
+      "grad_norm": 0.6135398745536804,
+      "learning_rate": 0.0002,
+      "loss": 0.6638,
+      "step": 17490
+    },
+    {
+      "epoch": 2.82919731630426,
+      "grad_norm": 0.6924924850463867,
+      "learning_rate": 0.0002,
+      "loss": 0.6829,
+      "step": 17500
+    },
+    {
+      "epoch": 2.830814000485005,
+      "grad_norm": 0.7471627593040466,
+      "learning_rate": 0.0002,
+      "loss": 0.6731,
+      "step": 17510
+    },
+    {
+      "epoch": 2.8324306846657503,
+      "grad_norm": 0.7145499587059021,
+      "learning_rate": 0.0002,
+      "loss": 0.7016,
+      "step": 17520
+    },
+    {
+      "epoch": 2.834047368846496,
+      "grad_norm": 0.7415414452552795,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17530
+    },
+    {
+      "epoch": 2.8356640530272412,
+      "grad_norm": 0.7328441739082336,
+      "learning_rate": 0.0002,
+      "loss": 0.6811,
+      "step": 17540
+    },
+    {
+      "epoch": 2.8372807372079865,
+      "grad_norm": 0.8267839550971985,
+      "learning_rate": 0.0002,
+      "loss": 0.6866,
+      "step": 17550
+    },
+    {
+      "epoch": 2.8388974213887317,
+      "grad_norm": 0.8877885341644287,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 17560
+    },
+    {
+      "epoch": 2.840514105569477,
+      "grad_norm": 0.857138454914093,
+      "learning_rate": 0.0002,
+      "loss": 0.7136,
+      "step": 17570
+    },
+    {
+      "epoch": 2.842130789750222,
+      "grad_norm": 0.8470779657363892,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 17580
+    },
+    {
+      "epoch": 2.843747473930968,
+      "grad_norm": 0.8553254008293152,
+      "learning_rate": 0.0002,
+      "loss": 0.6976,
+      "step": 17590
+    },
+    {
+      "epoch": 2.845364158111713,
+      "grad_norm": 0.8033196926116943,
+      "learning_rate": 0.0002,
+      "loss": 0.7297,
+      "step": 17600
+    },
+    {
+      "epoch": 2.8469808422924583,
+      "grad_norm": 0.7949087023735046,
+      "learning_rate": 0.0002,
+      "loss": 0.7062,
+      "step": 17610
+    },
+    {
+      "epoch": 2.8485975264732035,
+      "grad_norm": 0.9241406321525574,
+      "learning_rate": 0.0002,
+      "loss": 0.651,
+      "step": 17620
+    },
+    {
+      "epoch": 2.8502142106539488,
+      "grad_norm": 0.7721285223960876,
+      "learning_rate": 0.0002,
+      "loss": 0.6601,
+      "step": 17630
+    },
+    {
+      "epoch": 2.851830894834694,
+      "grad_norm": 1.0246692895889282,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 17640
+    },
+    {
+      "epoch": 2.853447579015439,
+      "grad_norm": 0.9244589805603027,
+      "learning_rate": 0.0002,
+      "loss": 0.7007,
+      "step": 17650
+    },
+    {
+      "epoch": 2.8550642631961844,
+      "grad_norm": 0.7243508696556091,
+      "learning_rate": 0.0002,
+      "loss": 0.7274,
+      "step": 17660
+    },
+    {
+      "epoch": 2.8566809473769297,
+      "grad_norm": 0.8943371176719666,
+      "learning_rate": 0.0002,
+      "loss": 0.6471,
+      "step": 17670
+    },
+    {
+      "epoch": 2.8582976315576754,
+      "grad_norm": 0.6531758904457092,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 17680
+    },
+    {
+      "epoch": 2.8599143157384206,
+      "grad_norm": 0.8367000818252563,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 17690
+    },
+    {
+      "epoch": 2.861530999919166,
+      "grad_norm": 0.7868556380271912,
+      "learning_rate": 0.0002,
+      "loss": 0.6943,
+      "step": 17700
+    },
+    {
+      "epoch": 2.863147684099911,
+      "grad_norm": 0.7213859558105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6919,
+      "step": 17710
+    },
+    {
+      "epoch": 2.8647643682806563,
+      "grad_norm": 0.7383931279182434,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 17720
+    },
+    {
+      "epoch": 2.8663810524614015,
+      "grad_norm": 0.7566812634468079,
+      "learning_rate": 0.0002,
+      "loss": 0.6841,
+      "step": 17730
+    },
+    {
+      "epoch": 2.867997736642147,
+      "grad_norm": 0.6930373311042786,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 17740
+    },
+    {
+      "epoch": 2.8696144208228924,
+      "grad_norm": 0.7911090850830078,
+      "learning_rate": 0.0002,
+      "loss": 0.6764,
+      "step": 17750
+    },
+    {
+      "epoch": 2.8712311050036377,
+      "grad_norm": 0.8484548926353455,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 17760
+    },
+    {
+      "epoch": 2.872847789184383,
+      "grad_norm": 0.7647597193717957,
+      "learning_rate": 0.0002,
+      "loss": 0.6931,
+      "step": 17770
+    },
+    {
+      "epoch": 2.874464473365128,
+      "grad_norm": 0.8791151642799377,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 17780
+    },
+    {
+      "epoch": 2.8760811575458733,
+      "grad_norm": 0.7253178358078003,
+      "learning_rate": 0.0002,
+      "loss": 0.7078,
+      "step": 17790
+    },
+    {
+      "epoch": 2.8776978417266186,
+      "grad_norm": 0.7956077456474304,
+      "learning_rate": 0.0002,
+      "loss": 0.6474,
+      "step": 17800
+    },
+    {
+      "epoch": 2.879314525907364,
+      "grad_norm": 0.8657688498497009,
+      "learning_rate": 0.0002,
+      "loss": 0.6687,
+      "step": 17810
+    },
+    {
+      "epoch": 2.880931210088109,
+      "grad_norm": 0.7059141993522644,
+      "learning_rate": 0.0002,
+      "loss": 0.7171,
+      "step": 17820
+    },
+    {
+      "epoch": 2.8825478942688547,
+      "grad_norm": 0.8886896967887878,
+      "learning_rate": 0.0002,
+      "loss": 0.683,
+      "step": 17830
+    },
+    {
+      "epoch": 2.8841645784496,
+      "grad_norm": 0.821032702922821,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 17840
+    },
+    {
+      "epoch": 2.885781262630345,
+      "grad_norm": 0.7183963656425476,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 17850
+    },
+    {
+      "epoch": 2.8873979468110904,
+      "grad_norm": 0.6222899556159973,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 17860
+    },
+    {
+      "epoch": 2.8890146309918356,
+      "grad_norm": 0.8187434077262878,
+      "learning_rate": 0.0002,
+      "loss": 0.6626,
+      "step": 17870
+    },
+    {
+      "epoch": 2.890631315172581,
+      "grad_norm": 0.9838479161262512,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 17880
+    },
+    {
+      "epoch": 2.8922479993533265,
+      "grad_norm": 0.7567742466926575,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 17890
+    },
+    {
+      "epoch": 2.893864683534072,
+      "grad_norm": 0.6875903606414795,
+      "learning_rate": 0.0002,
+      "loss": 0.7073,
+      "step": 17900
+    },
+    {
+      "epoch": 2.895481367714817,
+      "grad_norm": 0.8043789267539978,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 17910
+    },
+    {
+      "epoch": 2.8970980518955622,
+      "grad_norm": 0.8062626719474792,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 17920
+    },
+    {
+      "epoch": 2.8987147360763075,
+      "grad_norm": 1.0251191854476929,
+      "learning_rate": 0.0002,
+      "loss": 0.7151,
+      "step": 17930
+    },
+    {
+      "epoch": 2.9003314202570527,
+      "grad_norm": 0.882253110408783,
+      "learning_rate": 0.0002,
+      "loss": 0.6605,
+      "step": 17940
+    },
+    {
+      "epoch": 2.901948104437798,
+      "grad_norm": 0.8683299422264099,
+      "learning_rate": 0.0002,
+      "loss": 0.6719,
+      "step": 17950
+    },
+    {
+      "epoch": 2.903564788618543,
+      "grad_norm": 0.7167282104492188,
+      "learning_rate": 0.0002,
+      "loss": 0.6896,
+      "step": 17960
+    },
+    {
+      "epoch": 2.9051814727992884,
+      "grad_norm": 0.7093694806098938,
+      "learning_rate": 0.0002,
+      "loss": 0.663,
+      "step": 17970
+    },
+    {
+      "epoch": 2.906798156980034,
+      "grad_norm": 0.8549879193305969,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 17980
+    },
+    {
+      "epoch": 2.9084148411607793,
+      "grad_norm": 0.6989606618881226,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 17990
+    },
+    {
+      "epoch": 2.9100315253415245,
+      "grad_norm": 0.9482976794242859,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 18000
+    },
+    {
+      "epoch": 2.9116482095222698,
+      "grad_norm": 0.7182440161705017,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 18010
+    },
+    {
+      "epoch": 2.913264893703015,
+      "grad_norm": 0.7732226252555847,
+      "learning_rate": 0.0002,
+      "loss": 0.6827,
+      "step": 18020
+    },
+    {
+      "epoch": 2.9148815778837607,
+      "grad_norm": 0.7936875224113464,
+      "learning_rate": 0.0002,
+      "loss": 0.7123,
+      "step": 18030
+    },
+    {
+      "epoch": 2.916498262064506,
+      "grad_norm": 0.8825615644454956,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 18040
+    },
+    {
+      "epoch": 2.918114946245251,
+      "grad_norm": 0.6778587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7139,
+      "step": 18050
+    },
+    {
+      "epoch": 2.9197316304259964,
+      "grad_norm": 0.7529265880584717,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 18060
+    },
+    {
+      "epoch": 2.9213483146067416,
+      "grad_norm": 0.7111883163452148,
+      "learning_rate": 0.0002,
+      "loss": 0.737,
+      "step": 18070
+    },
+    {
+      "epoch": 2.922964998787487,
+      "grad_norm": 0.7214767932891846,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 18080
+    },
+    {
+      "epoch": 2.924581682968232,
+      "grad_norm": 0.800417423248291,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 18090
+    },
+    {
+      "epoch": 2.9261983671489773,
+      "grad_norm": 1.248575210571289,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 18100
+    },
+    {
+      "epoch": 2.9278150513297225,
+      "grad_norm": 0.757788360118866,
+      "learning_rate": 0.0002,
+      "loss": 0.7004,
+      "step": 18110
+    },
+    {
+      "epoch": 2.9294317355104678,
+      "grad_norm": 1.0583995580673218,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 18120
+    },
+    {
+      "epoch": 2.9310484196912134,
+      "grad_norm": 0.8228777647018433,
+      "learning_rate": 0.0002,
+      "loss": 0.6365,
+      "step": 18130
+    },
+    {
+      "epoch": 2.9326651038719587,
+      "grad_norm": 0.8374035358428955,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 18140
+    },
+    {
+      "epoch": 2.934281788052704,
+      "grad_norm": 0.7976473569869995,
+      "learning_rate": 0.0002,
+      "loss": 0.6399,
+      "step": 18150
+    },
+    {
+      "epoch": 2.935898472233449,
+      "grad_norm": 0.8009907603263855,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 18160
+    },
+    {
+      "epoch": 2.9375151564141944,
+      "grad_norm": 0.835213303565979,
+      "learning_rate": 0.0002,
+      "loss": 0.7485,
+      "step": 18170
+    },
+    {
+      "epoch": 2.93913184059494,
+      "grad_norm": 0.7982219457626343,
+      "learning_rate": 0.0002,
+      "loss": 0.7376,
+      "step": 18180
+    },
+    {
+      "epoch": 2.9407485247756853,
+      "grad_norm": 0.7070978879928589,
+      "learning_rate": 0.0002,
+      "loss": 0.6348,
+      "step": 18190
+    },
+    {
+      "epoch": 2.9423652089564305,
+      "grad_norm": 0.8619440197944641,
+      "learning_rate": 0.0002,
+      "loss": 0.6608,
+      "step": 18200
+    },
+    {
+      "epoch": 2.9439818931371757,
+      "grad_norm": 0.6693987250328064,
+      "learning_rate": 0.0002,
+      "loss": 0.666,
+      "step": 18210
+    },
+    {
+      "epoch": 2.945598577317921,
+      "grad_norm": 0.6747021079063416,
+      "learning_rate": 0.0002,
+      "loss": 0.728,
+      "step": 18220
+    },
+    {
+      "epoch": 2.947215261498666,
+      "grad_norm": 0.860387921333313,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 18230
+    },
+    {
+      "epoch": 2.9488319456794114,
+      "grad_norm": 0.799976646900177,
+      "learning_rate": 0.0002,
+      "loss": 0.6945,
+      "step": 18240
+    },
+    {
+      "epoch": 2.9504486298601567,
+      "grad_norm": 0.7864769101142883,
+      "learning_rate": 0.0002,
+      "loss": 0.7243,
+      "step": 18250
+    },
+    {
+      "epoch": 2.952065314040902,
+      "grad_norm": 0.6713884472846985,
+      "learning_rate": 0.0002,
+      "loss": 0.6785,
+      "step": 18260
+    },
+    {
+      "epoch": 2.9536819982216476,
+      "grad_norm": 0.9031508564949036,
+      "learning_rate": 0.0002,
+      "loss": 0.7429,
+      "step": 18270
+    },
+    {
+      "epoch": 2.955298682402393,
+      "grad_norm": 0.7205073237419128,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 18280
+    },
+    {
+      "epoch": 2.956915366583138,
+      "grad_norm": 0.7746205925941467,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 18290
+    },
+    {
+      "epoch": 2.9585320507638833,
+      "grad_norm": 0.6533427834510803,
+      "learning_rate": 0.0002,
+      "loss": 0.6218,
+      "step": 18300
+    },
+    {
+      "epoch": 2.9601487349446285,
+      "grad_norm": 0.9083208441734314,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 18310
+    },
+    {
+      "epoch": 2.9617654191253737,
+      "grad_norm": 0.7446991801261902,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 18320
+    },
+    {
+      "epoch": 2.9633821033061194,
+      "grad_norm": 0.6514461636543274,
+      "learning_rate": 0.0002,
+      "loss": 0.6738,
+      "step": 18330
+    },
+    {
+      "epoch": 2.9649987874868646,
+      "grad_norm": 0.8580465912818909,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18340
+    },
+    {
+      "epoch": 2.96661547166761,
+      "grad_norm": 0.7074266076087952,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 18350
+    },
+    {
+      "epoch": 2.968232155848355,
+      "grad_norm": 0.899892270565033,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 18360
+    },
+    {
+      "epoch": 2.9698488400291003,
+      "grad_norm": 0.8217641711235046,
+      "learning_rate": 0.0002,
+      "loss": 0.7094,
+      "step": 18370
+    },
+    {
+      "epoch": 2.9714655242098456,
+      "grad_norm": 0.8611799478530884,
+      "learning_rate": 0.0002,
+      "loss": 0.6916,
+      "step": 18380
+    },
+    {
+      "epoch": 2.973082208390591,
+      "grad_norm": 0.6909302473068237,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 18390
+    },
+    {
+      "epoch": 2.974698892571336,
+      "grad_norm": 0.6554358005523682,
+      "learning_rate": 0.0002,
+      "loss": 0.7247,
+      "step": 18400
+    },
+    {
+      "epoch": 2.9763155767520812,
+      "grad_norm": 0.7803071737289429,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 18410
+    },
+    {
+      "epoch": 2.977932260932827,
+      "grad_norm": 0.7838954925537109,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 18420
+    },
+    {
+      "epoch": 2.979548945113572,
+      "grad_norm": 0.7098495364189148,
+      "learning_rate": 0.0002,
+      "loss": 0.6522,
+      "step": 18430
+    },
+    {
+      "epoch": 2.9811656292943174,
+      "grad_norm": 0.8981785774230957,
+      "learning_rate": 0.0002,
+      "loss": 0.739,
+      "step": 18440
+    },
+    {
+      "epoch": 2.9827823134750626,
+      "grad_norm": 0.7197171449661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6689,
+      "step": 18450
+    },
+    {
+      "epoch": 2.984398997655808,
+      "grad_norm": 0.793185293674469,
+      "learning_rate": 0.0002,
+      "loss": 0.706,
+      "step": 18460
+    },
+    {
+      "epoch": 2.986015681836553,
+      "grad_norm": 0.8531473875045776,
+      "learning_rate": 0.0002,
+      "loss": 0.7124,
+      "step": 18470
+    },
+    {
+      "epoch": 2.9876323660172988,
+      "grad_norm": 0.6627361178398132,
+      "learning_rate": 0.0002,
+      "loss": 0.6901,
+      "step": 18480
+    },
+    {
+      "epoch": 2.989249050198044,
+      "grad_norm": 0.5708155035972595,
+      "learning_rate": 0.0002,
+      "loss": 0.6591,
+      "step": 18490
+    },
+    {
+      "epoch": 2.990865734378789,
+      "grad_norm": 0.8227280378341675,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 18500
+    },
+    {
+      "epoch": 2.9924824185595345,
+      "grad_norm": 0.7102749943733215,
+      "learning_rate": 0.0002,
+      "loss": 0.6701,
+      "step": 18510
+    },
+    {
+      "epoch": 2.9940991027402797,
+      "grad_norm": 0.839485228061676,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 18520
+    },
+    {
+      "epoch": 2.995715786921025,
+      "grad_norm": 0.9038704037666321,
+      "learning_rate": 0.0002,
+      "loss": 0.6521,
+      "step": 18530
+    },
+    {
+      "epoch": 2.99733247110177,
+      "grad_norm": 0.8737510442733765,
+      "learning_rate": 0.0002,
+      "loss": 0.7186,
+      "step": 18540
+    },
+    {
+      "epoch": 2.9989491552825154,
+      "grad_norm": 0.7323142886161804,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 18550
+    },
+    {
+      "epoch": 2.9999191657909625,
+      "eval_loss": 1.1262480020523071,
+      "eval_runtime": 122.0868,
+      "eval_samples_per_second": 6.004,
+      "eval_steps_per_second": 0.754,
+      "step": 18556
+    },
+    {
+      "epoch": 3.000565839463261,
+      "grad_norm": 0.8465463519096375,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 18560
+    },
+    {
+      "epoch": 3.0021825236440063,
+      "grad_norm": 0.9134138822555542,
+      "learning_rate": 0.0002,
+      "loss": 0.6064,
+      "step": 18570
+    },
+    {
+      "epoch": 3.0037992078247515,
+      "grad_norm": 0.760715126991272,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 18580
+    },
+    {
+      "epoch": 3.0054158920054967,
+      "grad_norm": 0.9208743572235107,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18590
+    },
+    {
+      "epoch": 3.007032576186242,
+      "grad_norm": 0.9232364892959595,
+      "learning_rate": 0.0002,
+      "loss": 0.5731,
+      "step": 18600
+    },
+    {
+      "epoch": 3.008649260366987,
+      "grad_norm": 1.1881544589996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6299,
+      "step": 18610
+    },
+    {
+      "epoch": 3.0102659445477324,
+      "grad_norm": 0.9372987747192383,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 18620
+    },
+    {
+      "epoch": 3.0118826287284777,
+      "grad_norm": 0.6900241374969482,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 18630
+    },
+    {
+      "epoch": 3.0134993129092233,
+      "grad_norm": 0.8451071381568909,
+      "learning_rate": 0.0002,
+      "loss": 0.5256,
+      "step": 18640
+    },
+    {
+      "epoch": 3.0151159970899686,
+      "grad_norm": 0.7763112187385559,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 18650
+    },
+    {
+      "epoch": 3.016732681270714,
+      "grad_norm": 1.043653964996338,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 18660
+    },
+    {
+      "epoch": 3.018349365451459,
+      "grad_norm": 1.0170660018920898,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 18670
+    },
+    {
+      "epoch": 3.0199660496322043,
+      "grad_norm": 0.7534180283546448,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 18680
+    },
+    {
+      "epoch": 3.0215827338129495,
+      "grad_norm": 0.7507367730140686,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 18690
+    },
+    {
+      "epoch": 3.0231994179936947,
+      "grad_norm": 0.7861620187759399,
+      "learning_rate": 0.0002,
+      "loss": 0.6201,
+      "step": 18700
+    },
+    {
+      "epoch": 3.0248161021744404,
+      "grad_norm": 1.0580339431762695,
+      "learning_rate": 0.0002,
+      "loss": 0.5802,
+      "step": 18710
+    },
+    {
+      "epoch": 3.0264327863551856,
+      "grad_norm": 0.7542710900306702,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 18720
+    },
+    {
+      "epoch": 3.028049470535931,
+      "grad_norm": 0.8189544677734375,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 18730
+    },
+    {
+      "epoch": 3.029666154716676,
+      "grad_norm": 0.9126611351966858,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 18740
+    },
+    {
+      "epoch": 3.0312828388974213,
+      "grad_norm": 0.8891341686248779,
+      "learning_rate": 0.0002,
+      "loss": 0.6443,
+      "step": 18750
+    },
+    {
+      "epoch": 3.0328995230781666,
+      "grad_norm": 0.8419283032417297,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 18760
+    },
+    {
+      "epoch": 3.034516207258912,
+      "grad_norm": 0.8048048615455627,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18770
+    },
+    {
+      "epoch": 3.0361328914396575,
+      "grad_norm": 0.7820217609405518,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 18780
+    },
+    {
+      "epoch": 3.0377495756204027,
+      "grad_norm": 0.854721188545227,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 18790
+    },
+    {
+      "epoch": 3.039366259801148,
+      "grad_norm": 0.912092924118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5784,
+      "step": 18800
+    },
+    {
+      "epoch": 3.040982943981893,
+      "grad_norm": 0.6596226096153259,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 18810
+    },
+    {
+      "epoch": 3.0425996281626384,
+      "grad_norm": 0.6351348757743835,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 18820
+    },
+    {
+      "epoch": 3.0442163123433836,
+      "grad_norm": 0.778188943862915,
+      "learning_rate": 0.0002,
+      "loss": 0.5953,
+      "step": 18830
+    },
+    {
+      "epoch": 3.045832996524129,
+      "grad_norm": 0.68234783411026,
+      "learning_rate": 0.0002,
+      "loss": 0.602,
+      "step": 18840
+    },
+    {
+      "epoch": 3.047449680704874,
+      "grad_norm": 0.998628556728363,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 18850
+    },
+    {
+      "epoch": 3.0490663648856198,
+      "grad_norm": 0.7393841743469238,
+      "learning_rate": 0.0002,
+      "loss": 0.6231,
+      "step": 18860
+    },
+    {
+      "epoch": 3.050683049066365,
+      "grad_norm": 0.84438556432724,
+      "learning_rate": 0.0002,
+      "loss": 0.568,
+      "step": 18870
+    },
+    {
+      "epoch": 3.0522997332471102,
+      "grad_norm": 0.8857501745223999,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 18880
+    },
+    {
+      "epoch": 3.0539164174278555,
+      "grad_norm": 0.7208474278450012,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 18890
+    },
+    {
+      "epoch": 3.0555331016086007,
+      "grad_norm": 0.7135229110717773,
+      "learning_rate": 0.0002,
+      "loss": 0.5998,
+      "step": 18900
+    },
+    {
+      "epoch": 3.057149785789346,
+      "grad_norm": 0.9130001664161682,
+      "learning_rate": 0.0002,
+      "loss": 0.5575,
+      "step": 18910
+    },
+    {
+      "epoch": 3.058766469970091,
+      "grad_norm": 0.9001716375350952,
+      "learning_rate": 0.0002,
+      "loss": 0.5955,
+      "step": 18920
+    },
+    {
+      "epoch": 3.060383154150837,
+      "grad_norm": 0.8667559623718262,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 18930
+    },
+    {
+      "epoch": 3.061999838331582,
+      "grad_norm": 0.8943959474563599,
+      "learning_rate": 0.0002,
+      "loss": 0.5818,
+      "step": 18940
+    },
+    {
+      "epoch": 3.0636165225123273,
+      "grad_norm": 0.8298377990722656,
+      "learning_rate": 0.0002,
+      "loss": 0.5978,
+      "step": 18950
+    },
+    {
+      "epoch": 3.0652332066930725,
+      "grad_norm": 0.7935267686843872,
+      "learning_rate": 0.0002,
+      "loss": 0.5782,
+      "step": 18960
+    },
+    {
+      "epoch": 3.0668498908738178,
+      "grad_norm": 1.1506379842758179,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 18970
+    },
+    {
+      "epoch": 3.068466575054563,
+      "grad_norm": 0.7693049907684326,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 18980
+    },
+    {
+      "epoch": 3.0700832592353082,
+      "grad_norm": 0.8040135502815247,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 18990
+    },
+    {
+      "epoch": 3.0716999434160535,
+      "grad_norm": 0.828404426574707,
+      "learning_rate": 0.0002,
+      "loss": 0.5541,
+      "step": 19000
+    },
+    {
+      "epoch": 3.073316627596799,
+      "grad_norm": 0.8811164498329163,
+      "learning_rate": 0.0002,
+      "loss": 0.6048,
+      "step": 19010
+    },
+    {
+      "epoch": 3.0749333117775444,
+      "grad_norm": 1.036205768585205,
+      "learning_rate": 0.0002,
+      "loss": 0.5845,
+      "step": 19020
+    },
+    {
+      "epoch": 3.0765499959582896,
+      "grad_norm": 0.8857285976409912,
+      "learning_rate": 0.0002,
+      "loss": 0.5838,
+      "step": 19030
+    },
+    {
+      "epoch": 3.078166680139035,
+      "grad_norm": 0.8392079472541809,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19040
+    },
+    {
+      "epoch": 3.07978336431978,
+      "grad_norm": 1.0287401676177979,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 19050
+    },
+    {
+      "epoch": 3.0814000485005253,
+      "grad_norm": 1.0086315870285034,
+      "learning_rate": 0.0002,
+      "loss": 0.5964,
+      "step": 19060
+    },
+    {
+      "epoch": 3.0830167326812705,
+      "grad_norm": 0.9245324730873108,
+      "learning_rate": 0.0002,
+      "loss": 0.5567,
+      "step": 19070
+    },
+    {
+      "epoch": 3.084633416862016,
+      "grad_norm": 0.8680877089500427,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 19080
+    },
+    {
+      "epoch": 3.0862501010427614,
+      "grad_norm": 0.8814793825149536,
+      "learning_rate": 0.0002,
+      "loss": 0.5611,
+      "step": 19090
+    },
+    {
+      "epoch": 3.0878667852235067,
+      "grad_norm": 0.9234458208084106,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19100
+    },
+    {
+      "epoch": 3.089483469404252,
+      "grad_norm": 1.1291664838790894,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 19110
+    },
+    {
+      "epoch": 3.091100153584997,
+      "grad_norm": 0.9191402792930603,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 19120
+    },
+    {
+      "epoch": 3.0927168377657424,
+      "grad_norm": 0.7103154063224792,
+      "learning_rate": 0.0002,
+      "loss": 0.5856,
+      "step": 19130
+    },
+    {
+      "epoch": 3.0943335219464876,
+      "grad_norm": 0.9368883967399597,
+      "learning_rate": 0.0002,
+      "loss": 0.6479,
+      "step": 19140
+    },
+    {
+      "epoch": 3.095950206127233,
+      "grad_norm": 0.9676656723022461,
+      "learning_rate": 0.0002,
+      "loss": 0.6167,
+      "step": 19150
+    },
+    {
+      "epoch": 3.0975668903079785,
+      "grad_norm": 0.8739792704582214,
+      "learning_rate": 0.0002,
+      "loss": 0.5794,
+      "step": 19160
+    },
+    {
+      "epoch": 3.0991835744887237,
+      "grad_norm": 0.8530174493789673,
+      "learning_rate": 0.0002,
+      "loss": 0.6112,
+      "step": 19170
+    },
+    {
+      "epoch": 3.100800258669469,
+      "grad_norm": 0.794945478439331,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 19180
+    },
+    {
+      "epoch": 3.102416942850214,
+      "grad_norm": 0.9508888125419617,
+      "learning_rate": 0.0002,
+      "loss": 0.5928,
+      "step": 19190
+    },
+    {
+      "epoch": 3.1040336270309594,
+      "grad_norm": 1.0599955320358276,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 19200
+    },
+    {
+      "epoch": 3.1056503112117047,
+      "grad_norm": 1.0673625469207764,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 19210
+    },
+    {
+      "epoch": 3.10726699539245,
+      "grad_norm": 0.7739115953445435,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 19220
+    },
+    {
+      "epoch": 3.1088836795731956,
+      "grad_norm": 0.9884951114654541,
+      "learning_rate": 0.0002,
+      "loss": 0.6046,
+      "step": 19230
+    },
+    {
+      "epoch": 3.110500363753941,
+      "grad_norm": 0.862260103225708,
+      "learning_rate": 0.0002,
+      "loss": 0.5932,
+      "step": 19240
+    },
+    {
+      "epoch": 3.112117047934686,
+      "grad_norm": 0.7690284848213196,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 19250
+    },
+    {
+      "epoch": 3.1137337321154313,
+      "grad_norm": 0.8758958578109741,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 19260
+    },
+    {
+      "epoch": 3.1153504162961765,
+      "grad_norm": 1.0356395244598389,
+      "learning_rate": 0.0002,
+      "loss": 0.6136,
+      "step": 19270
+    },
+    {
+      "epoch": 3.1169671004769217,
+      "grad_norm": 0.6950937509536743,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 19280
+    },
+    {
+      "epoch": 3.118583784657667,
+      "grad_norm": 0.760998010635376,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 19290
+    },
+    {
+      "epoch": 3.1202004688384126,
+      "grad_norm": 0.9335789084434509,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 19300
+    },
+    {
+      "epoch": 3.121817153019158,
+      "grad_norm": 0.9636204242706299,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 19310
+    },
+    {
+      "epoch": 3.123433837199903,
+      "grad_norm": 1.0820997953414917,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 19320
+    },
+    {
+      "epoch": 3.1250505213806483,
+      "grad_norm": 0.7333487272262573,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 19330
+    },
+    {
+      "epoch": 3.1266672055613935,
+      "grad_norm": 1.0417509078979492,
+      "learning_rate": 0.0002,
+      "loss": 0.6178,
+      "step": 19340
+    },
+    {
+      "epoch": 3.128283889742139,
+      "grad_norm": 0.9267749190330505,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 19350
+    },
+    {
+      "epoch": 3.129900573922884,
+      "grad_norm": 0.777798593044281,
+      "learning_rate": 0.0002,
+      "loss": 0.6063,
+      "step": 19360
+    },
+    {
+      "epoch": 3.1315172581036297,
+      "grad_norm": 0.8425456881523132,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 19370
+    },
+    {
+      "epoch": 3.133133942284375,
+      "grad_norm": 0.9617102146148682,
+      "learning_rate": 0.0002,
+      "loss": 0.6042,
+      "step": 19380
+    },
+    {
+      "epoch": 3.13475062646512,
+      "grad_norm": 1.0052828788757324,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 19390
+    },
+    {
+      "epoch": 3.1363673106458654,
+      "grad_norm": 0.7637009024620056,
+      "learning_rate": 0.0002,
+      "loss": 0.5713,
+      "step": 19400
+    },
+    {
+      "epoch": 3.1379839948266106,
+      "grad_norm": 0.7958088517189026,
+      "learning_rate": 0.0002,
+      "loss": 0.5497,
+      "step": 19410
+    },
+    {
+      "epoch": 3.139600679007356,
+      "grad_norm": 0.9161727428436279,
+      "learning_rate": 0.0002,
+      "loss": 0.6283,
+      "step": 19420
+    },
+    {
+      "epoch": 3.141217363188101,
+      "grad_norm": 0.8402149677276611,
+      "learning_rate": 0.0002,
+      "loss": 0.5638,
+      "step": 19430
+    },
+    {
+      "epoch": 3.1428340473688463,
+      "grad_norm": 1.0056525468826294,
+      "learning_rate": 0.0002,
+      "loss": 0.5848,
+      "step": 19440
+    },
+    {
+      "epoch": 3.144450731549592,
+      "grad_norm": 1.0129190683364868,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 19450
+    },
+    {
+      "epoch": 3.146067415730337,
+      "grad_norm": 0.790825366973877,
+      "learning_rate": 0.0002,
+      "loss": 0.5808,
+      "step": 19460
+    },
+    {
+      "epoch": 3.1476840999110824,
+      "grad_norm": 1.441665530204773,
+      "learning_rate": 0.0002,
+      "loss": 0.5607,
+      "step": 19470
+    },
+    {
+      "epoch": 3.1493007840918277,
+      "grad_norm": 0.7846331596374512,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19480
+    },
+    {
+      "epoch": 3.150917468272573,
+      "grad_norm": 0.7915332913398743,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 19490
+    },
+    {
+      "epoch": 3.152534152453318,
+      "grad_norm": 0.933982253074646,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 19500
+    },
+    {
+      "epoch": 3.1541508366340634,
+      "grad_norm": 1.038408637046814,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 19510
+    },
+    {
+      "epoch": 3.155767520814809,
+      "grad_norm": 1.018935203552246,
+      "learning_rate": 0.0002,
+      "loss": 0.6271,
+      "step": 19520
+    },
+    {
+      "epoch": 3.1573842049955543,
+      "grad_norm": 0.9618112444877625,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 19530
+    },
+    {
+      "epoch": 3.1590008891762995,
+      "grad_norm": 0.8900452852249146,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 19540
+    },
+    {
+      "epoch": 3.1606175733570447,
+      "grad_norm": 0.8254160284996033,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 19550
+    },
+    {
+      "epoch": 3.16223425753779,
+      "grad_norm": 1.004376769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19560
+    },
+    {
+      "epoch": 3.163850941718535,
+      "grad_norm": 1.0490446090698242,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 19570
+    },
+    {
+      "epoch": 3.1654676258992804,
+      "grad_norm": 0.7387403845787048,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19580
+    },
+    {
+      "epoch": 3.1670843100800257,
+      "grad_norm": 0.7611538171768188,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 19590
+    },
+    {
+      "epoch": 3.1687009942607713,
+      "grad_norm": 0.8239886164665222,
+      "learning_rate": 0.0002,
+      "loss": 0.5937,
+      "step": 19600
+    },
+    {
+      "epoch": 3.1703176784415166,
+      "grad_norm": 0.9327243566513062,
+      "learning_rate": 0.0002,
+      "loss": 0.6068,
+      "step": 19610
+    },
+    {
+      "epoch": 3.171934362622262,
+      "grad_norm": 0.9662560224533081,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 19620
+    },
+    {
+      "epoch": 3.173551046803007,
+      "grad_norm": 0.9183341860771179,
+      "learning_rate": 0.0002,
+      "loss": 0.5988,
+      "step": 19630
+    },
+    {
+      "epoch": 3.1751677309837523,
+      "grad_norm": 0.875066876411438,
+      "learning_rate": 0.0002,
+      "loss": 0.5909,
+      "step": 19640
+    },
+    {
+      "epoch": 3.1767844151644975,
+      "grad_norm": 0.8567508459091187,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 19650
+    },
+    {
+      "epoch": 3.1784010993452427,
+      "grad_norm": 0.6805780529975891,
+      "learning_rate": 0.0002,
+      "loss": 0.5805,
+      "step": 19660
+    },
+    {
+      "epoch": 3.1800177835259884,
+      "grad_norm": 0.8776944279670715,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 19670
+    },
+    {
+      "epoch": 3.1816344677067336,
+      "grad_norm": 0.9036329984664917,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 19680
+    },
+    {
+      "epoch": 3.183251151887479,
+      "grad_norm": 0.8527372479438782,
+      "learning_rate": 0.0002,
+      "loss": 0.6238,
+      "step": 19690
+    },
+    {
+      "epoch": 3.184867836068224,
+      "grad_norm": 1.1045585870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 19700
+    },
+    {
+      "epoch": 3.1864845202489693,
+      "grad_norm": 0.9213830828666687,
+      "learning_rate": 0.0002,
+      "loss": 0.5491,
+      "step": 19710
+    },
+    {
+      "epoch": 3.1881012044297146,
+      "grad_norm": 0.8865814805030823,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 19720
+    },
+    {
+      "epoch": 3.18971788861046,
+      "grad_norm": 0.7939388751983643,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 19730
+    },
+    {
+      "epoch": 3.191334572791205,
+      "grad_norm": 0.6966729760169983,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 19740
+    },
+    {
+      "epoch": 3.1929512569719507,
+      "grad_norm": 0.8023673295974731,
+      "learning_rate": 0.0002,
+      "loss": 0.5839,
+      "step": 19750
+    },
+    {
+      "epoch": 3.194567941152696,
+      "grad_norm": 0.7992037534713745,
+      "learning_rate": 0.0002,
+      "loss": 0.6267,
+      "step": 19760
+    },
+    {
+      "epoch": 3.196184625333441,
+      "grad_norm": 0.7412247657775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 19770
+    },
+    {
+      "epoch": 3.1978013095141864,
+      "grad_norm": 0.9598729014396667,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 19780
+    },
+    {
+      "epoch": 3.1994179936949316,
+      "grad_norm": 0.8331366777420044,
+      "learning_rate": 0.0002,
+      "loss": 0.5685,
+      "step": 19790
+    },
+    {
+      "epoch": 3.201034677875677,
+      "grad_norm": 0.8939169645309448,
+      "learning_rate": 0.0002,
+      "loss": 0.6104,
+      "step": 19800
+    },
+    {
+      "epoch": 3.202651362056422,
+      "grad_norm": 0.9219734072685242,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 19810
+    },
+    {
+      "epoch": 3.2042680462371678,
+      "grad_norm": 0.869490385055542,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 19820
+    },
+    {
+      "epoch": 3.205884730417913,
+      "grad_norm": 0.8989706635475159,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 19830
+    },
+    {
+      "epoch": 3.2075014145986582,
+      "grad_norm": 0.8477165102958679,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 19840
+    },
+    {
+      "epoch": 3.2091180987794035,
+      "grad_norm": 0.8720678687095642,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 19850
+    },
+    {
+      "epoch": 3.2107347829601487,
+      "grad_norm": 0.861406683921814,
+      "learning_rate": 0.0002,
+      "loss": 0.5694,
+      "step": 19860
+    },
+    {
+      "epoch": 3.212351467140894,
+      "grad_norm": 0.8228686451911926,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 19870
+    },
+    {
+      "epoch": 3.213968151321639,
+      "grad_norm": 0.7936596870422363,
+      "learning_rate": 0.0002,
+      "loss": 0.625,
+      "step": 19880
+    },
+    {
+      "epoch": 3.2155848355023844,
+      "grad_norm": 1.097377896308899,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 19890
+    },
+    {
+      "epoch": 3.21720151968313,
+      "grad_norm": 0.9544782638549805,
+      "learning_rate": 0.0002,
+      "loss": 0.6725,
+      "step": 19900
+    },
+    {
+      "epoch": 3.2188182038638753,
+      "grad_norm": 0.8240751624107361,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 19910
+    },
+    {
+      "epoch": 3.2204348880446205,
+      "grad_norm": 0.8332096338272095,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 19920
+    },
+    {
+      "epoch": 3.2220515722253658,
+      "grad_norm": 1.0954567193984985,
+      "learning_rate": 0.0002,
+      "loss": 0.6274,
+      "step": 19930
+    },
+    {
+      "epoch": 3.223668256406111,
+      "grad_norm": 0.7790525555610657,
+      "learning_rate": 0.0002,
+      "loss": 0.652,
+      "step": 19940
+    },
+    {
+      "epoch": 3.225284940586856,
+      "grad_norm": 0.7966814041137695,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 19950
+    },
+    {
+      "epoch": 3.2269016247676015,
+      "grad_norm": 0.9751881957054138,
+      "learning_rate": 0.0002,
+      "loss": 0.5911,
+      "step": 19960
+    },
+    {
+      "epoch": 3.228518308948347,
+      "grad_norm": 0.9856047630310059,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 19970
+    },
+    {
+      "epoch": 3.2301349931290924,
+      "grad_norm": 1.3062353134155273,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 19980
+    },
+    {
+      "epoch": 3.2317516773098376,
+      "grad_norm": 0.9510692358016968,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 19990
+    },
+    {
+      "epoch": 3.233368361490583,
+      "grad_norm": 0.8630342483520508,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 20000
+    },
+    {
+      "epoch": 3.234985045671328,
+      "grad_norm": 0.8966519236564636,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20010
+    },
+    {
+      "epoch": 3.2366017298520733,
+      "grad_norm": 0.7093510627746582,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 20020
+    },
+    {
+      "epoch": 3.2382184140328185,
+      "grad_norm": 0.7771096229553223,
+      "learning_rate": 0.0002,
+      "loss": 0.585,
+      "step": 20030
+    },
+    {
+      "epoch": 3.2398350982135637,
+      "grad_norm": 0.841058075428009,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 20040
+    },
+    {
+      "epoch": 3.2414517823943094,
+      "grad_norm": 0.909712553024292,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 20050
+    },
+    {
+      "epoch": 3.2430684665750547,
+      "grad_norm": 0.8321019411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20060
+    },
+    {
+      "epoch": 3.2446851507558,
+      "grad_norm": 0.779901921749115,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 20070
+    },
+    {
+      "epoch": 3.246301834936545,
+      "grad_norm": 0.6249170303344727,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 20080
+    },
+    {
+      "epoch": 3.2479185191172903,
+      "grad_norm": 0.8000940680503845,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 20090
+    },
+    {
+      "epoch": 3.2495352032980356,
+      "grad_norm": 0.7627735137939453,
+      "learning_rate": 0.0002,
+      "loss": 0.6273,
+      "step": 20100
+    },
+    {
+      "epoch": 3.2511518874787813,
+      "grad_norm": 0.8780747056007385,
+      "learning_rate": 0.0002,
+      "loss": 0.6223,
+      "step": 20110
+    },
+    {
+      "epoch": 3.2527685716595265,
+      "grad_norm": 0.772037148475647,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 20120
+    },
+    {
+      "epoch": 3.2543852558402717,
+      "grad_norm": 1.0086580514907837,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 20130
+    },
+    {
+      "epoch": 3.256001940021017,
+      "grad_norm": 0.9360289573669434,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20140
+    },
+    {
+      "epoch": 3.257618624201762,
+      "grad_norm": 1.2099586725234985,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 20150
+    },
+    {
+      "epoch": 3.2592353083825074,
+      "grad_norm": 0.8368481397628784,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 20160
+    },
+    {
+      "epoch": 3.2608519925632526,
+      "grad_norm": 0.7391039133071899,
+      "learning_rate": 0.0002,
+      "loss": 0.5626,
+      "step": 20170
+    },
+    {
+      "epoch": 3.262468676743998,
+      "grad_norm": 0.9122273325920105,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 20180
+    },
+    {
+      "epoch": 3.264085360924743,
+      "grad_norm": 0.8502281904220581,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 20190
+    },
+    {
+      "epoch": 3.265702045105489,
+      "grad_norm": 1.0926852226257324,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 20200
+    },
+    {
+      "epoch": 3.267318729286234,
+      "grad_norm": 0.7902828454971313,
+      "learning_rate": 0.0002,
+      "loss": 0.6027,
+      "step": 20210
+    },
+    {
+      "epoch": 3.2689354134669792,
+      "grad_norm": 0.8724729418754578,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 20220
+    },
+    {
+      "epoch": 3.2705520976477245,
+      "grad_norm": 0.8469277024269104,
+      "learning_rate": 0.0002,
+      "loss": 0.6242,
+      "step": 20230
+    },
+    {
+      "epoch": 3.2721687818284697,
+      "grad_norm": 0.8865092992782593,
+      "learning_rate": 0.0002,
+      "loss": 0.644,
+      "step": 20240
+    },
+    {
+      "epoch": 3.273785466009215,
+      "grad_norm": 1.0979334115982056,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20250
+    },
+    {
+      "epoch": 3.2754021501899606,
+      "grad_norm": 1.0860793590545654,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 20260
+    },
+    {
+      "epoch": 3.277018834370706,
+      "grad_norm": 0.981745183467865,
+      "learning_rate": 0.0002,
+      "loss": 0.6105,
+      "step": 20270
+    },
+    {
+      "epoch": 3.278635518551451,
+      "grad_norm": 0.9155020713806152,
+      "learning_rate": 0.0002,
+      "loss": 0.627,
+      "step": 20280
+    },
+    {
+      "epoch": 3.2802522027321963,
+      "grad_norm": 0.8436718583106995,
+      "learning_rate": 0.0002,
+      "loss": 0.5899,
+      "step": 20290
+    },
+    {
+      "epoch": 3.2818688869129415,
+      "grad_norm": 1.0329409837722778,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 20300
+    },
+    {
+      "epoch": 3.2834855710936868,
+      "grad_norm": 0.9876394271850586,
+      "learning_rate": 0.0002,
+      "loss": 0.6,
+      "step": 20310
+    },
+    {
+      "epoch": 3.285102255274432,
+      "grad_norm": 0.8052917718887329,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 20320
+    },
+    {
+      "epoch": 3.2867189394551772,
+      "grad_norm": 0.8390680551528931,
+      "learning_rate": 0.0002,
+      "loss": 0.5949,
+      "step": 20330
+    },
+    {
+      "epoch": 3.288335623635923,
+      "grad_norm": 0.9515735507011414,
+      "learning_rate": 0.0002,
+      "loss": 0.6492,
+      "step": 20340
+    },
+    {
+      "epoch": 3.289952307816668,
+      "grad_norm": 0.8028870224952698,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 20350
+    },
+    {
+      "epoch": 3.2915689919974134,
+      "grad_norm": 0.862592339515686,
+      "learning_rate": 0.0002,
+      "loss": 0.634,
+      "step": 20360
+    },
+    {
+      "epoch": 3.2931856761781586,
+      "grad_norm": 0.7451621890068054,
+      "learning_rate": 0.0002,
+      "loss": 0.6345,
+      "step": 20370
+    },
+    {
+      "epoch": 3.294802360358904,
+      "grad_norm": 0.8966776728630066,
+      "learning_rate": 0.0002,
+      "loss": 0.6458,
+      "step": 20380
+    },
+    {
+      "epoch": 3.296419044539649,
+      "grad_norm": 0.9289216995239258,
+      "learning_rate": 0.0002,
+      "loss": 0.5967,
+      "step": 20390
+    },
+    {
+      "epoch": 3.2980357287203943,
+      "grad_norm": 0.9649626612663269,
+      "learning_rate": 0.0002,
+      "loss": 0.6599,
+      "step": 20400
+    },
+    {
+      "epoch": 3.29965241290114,
+      "grad_norm": 1.1953798532485962,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 20410
+    },
+    {
+      "epoch": 3.301269097081885,
+      "grad_norm": 0.8929083943367004,
+      "learning_rate": 0.0002,
+      "loss": 0.5997,
+      "step": 20420
+    },
+    {
+      "epoch": 3.3028857812626304,
+      "grad_norm": 0.8922014236450195,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 20430
+    },
+    {
+      "epoch": 3.3045024654433757,
+      "grad_norm": 0.9754860401153564,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 20440
+    },
+    {
+      "epoch": 3.306119149624121,
+      "grad_norm": 0.8873140215873718,
+      "learning_rate": 0.0002,
+      "loss": 0.5653,
+      "step": 20450
+    },
+    {
+      "epoch": 3.307735833804866,
+      "grad_norm": 0.857271671295166,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20460
+    },
+    {
+      "epoch": 3.3093525179856114,
+      "grad_norm": 0.9022141098976135,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 20470
+    },
+    {
+      "epoch": 3.3109692021663566,
+      "grad_norm": 0.8614798188209534,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 20480
+    },
+    {
+      "epoch": 3.3125858863471023,
+      "grad_norm": 0.8838164210319519,
+      "learning_rate": 0.0002,
+      "loss": 0.6254,
+      "step": 20490
+    },
+    {
+      "epoch": 3.3142025705278475,
+      "grad_norm": 0.8709736466407776,
+      "learning_rate": 0.0002,
+      "loss": 0.5849,
+      "step": 20500
+    },
+    {
+      "epoch": 3.3158192547085927,
+      "grad_norm": 0.9533300995826721,
+      "learning_rate": 0.0002,
+      "loss": 0.6146,
+      "step": 20510
+    },
+    {
+      "epoch": 3.317435938889338,
+      "grad_norm": 0.8259269595146179,
+      "learning_rate": 0.0002,
+      "loss": 0.6029,
+      "step": 20520
+    },
+    {
+      "epoch": 3.319052623070083,
+      "grad_norm": 0.8607608079910278,
+      "learning_rate": 0.0002,
+      "loss": 0.6268,
+      "step": 20530
+    },
+    {
+      "epoch": 3.3206693072508284,
+      "grad_norm": 1.0863020420074463,
+      "learning_rate": 0.0002,
+      "loss": 0.5676,
+      "step": 20540
+    },
+    {
+      "epoch": 3.3222859914315737,
+      "grad_norm": 1.011489987373352,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 20550
+    },
+    {
+      "epoch": 3.3239026756123193,
+      "grad_norm": 0.6952177882194519,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 20560
+    },
+    {
+      "epoch": 3.3255193597930646,
+      "grad_norm": 0.9638974070549011,
+      "learning_rate": 0.0002,
+      "loss": 0.6229,
+      "step": 20570
+    },
+    {
+      "epoch": 3.32713604397381,
+      "grad_norm": 1.0310138463974,
+      "learning_rate": 0.0002,
+      "loss": 0.5882,
+      "step": 20580
+    },
+    {
+      "epoch": 3.328752728154555,
+      "grad_norm": 0.9371318221092224,
+      "learning_rate": 0.0002,
+      "loss": 0.594,
+      "step": 20590
+    },
+    {
+      "epoch": 3.3303694123353003,
+      "grad_norm": 0.8756691813468933,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 20600
+    },
+    {
+      "epoch": 3.3319860965160455,
+      "grad_norm": 1.054175853729248,
+      "learning_rate": 0.0002,
+      "loss": 0.5994,
+      "step": 20610
+    },
+    {
+      "epoch": 3.3336027806967907,
+      "grad_norm": 0.9074128270149231,
+      "learning_rate": 0.0002,
+      "loss": 0.6169,
+      "step": 20620
+    },
+    {
+      "epoch": 3.335219464877536,
+      "grad_norm": 0.906900942325592,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 20630
+    },
+    {
+      "epoch": 3.3368361490582816,
+      "grad_norm": 0.8689333200454712,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 20640
+    },
+    {
+      "epoch": 3.338452833239027,
+      "grad_norm": 0.9889747500419617,
+      "learning_rate": 0.0002,
+      "loss": 0.6079,
+      "step": 20650
+    },
+    {
+      "epoch": 3.340069517419772,
+      "grad_norm": 1.0685805082321167,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 20660
+    },
+    {
+      "epoch": 3.3416862016005173,
+      "grad_norm": 0.7495010495185852,
+      "learning_rate": 0.0002,
+      "loss": 0.6091,
+      "step": 20670
+    },
+    {
+      "epoch": 3.3433028857812626,
+      "grad_norm": 0.8747848272323608,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 20680
+    },
+    {
+      "epoch": 3.344919569962008,
+      "grad_norm": 0.9762673377990723,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 20690
+    },
+    {
+      "epoch": 3.346536254142753,
+      "grad_norm": 1.0284489393234253,
+      "learning_rate": 0.0002,
+      "loss": 0.6784,
+      "step": 20700
+    },
+    {
+      "epoch": 3.3481529383234987,
+      "grad_norm": 0.7293812036514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 20710
+    },
+    {
+      "epoch": 3.349769622504244,
+      "grad_norm": 0.8330199122428894,
+      "learning_rate": 0.0002,
+      "loss": 0.609,
+      "step": 20720
+    },
+    {
+      "epoch": 3.351386306684989,
+      "grad_norm": 0.9808499217033386,
+      "learning_rate": 0.0002,
+      "loss": 0.5729,
+      "step": 20730
+    },
+    {
+      "epoch": 3.3530029908657344,
+      "grad_norm": 0.9508825540542603,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 20740
+    },
+    {
+      "epoch": 3.3546196750464796,
+      "grad_norm": 0.790483832359314,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 20750
+    },
+    {
+      "epoch": 3.356236359227225,
+      "grad_norm": 1.022793173789978,
+      "learning_rate": 0.0002,
+      "loss": 0.6327,
+      "step": 20760
+    },
+    {
+      "epoch": 3.35785304340797,
+      "grad_norm": 0.8318950533866882,
+      "learning_rate": 0.0002,
+      "loss": 0.6439,
+      "step": 20770
+    },
+    {
+      "epoch": 3.3594697275887153,
+      "grad_norm": 0.7980858087539673,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 20780
+    },
+    {
+      "epoch": 3.361086411769461,
+      "grad_norm": 0.8114802241325378,
+      "learning_rate": 0.0002,
+      "loss": 0.6746,
+      "step": 20790
+    },
+    {
+      "epoch": 3.3627030959502062,
+      "grad_norm": 0.8522519469261169,
+      "learning_rate": 0.0002,
+      "loss": 0.6017,
+      "step": 20800
+    },
+    {
+      "epoch": 3.3643197801309515,
+      "grad_norm": 0.9142431616783142,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 20810
+    },
+    {
+      "epoch": 3.3659364643116967,
+      "grad_norm": 0.771170437335968,
+      "learning_rate": 0.0002,
+      "loss": 0.6331,
+      "step": 20820
+    },
+    {
+      "epoch": 3.367553148492442,
+      "grad_norm": 1.0628231763839722,
+      "learning_rate": 0.0002,
+      "loss": 0.5879,
+      "step": 20830
+    },
+    {
+      "epoch": 3.369169832673187,
+      "grad_norm": 0.9384352564811707,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 20840
+    },
+    {
+      "epoch": 3.370786516853933,
+      "grad_norm": 1.1286591291427612,
+      "learning_rate": 0.0002,
+      "loss": 0.6292,
+      "step": 20850
+    },
+    {
+      "epoch": 3.372403201034678,
+      "grad_norm": 1.1349513530731201,
+      "learning_rate": 0.0002,
+      "loss": 0.5986,
+      "step": 20860
+    },
+    {
+      "epoch": 3.3740198852154233,
+      "grad_norm": 1.0127464532852173,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 20870
+    },
+    {
+      "epoch": 3.3756365693961685,
+      "grad_norm": 0.9111971855163574,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 20880
+    },
+    {
+      "epoch": 3.3772532535769137,
+      "grad_norm": 0.871356725692749,
+      "learning_rate": 0.0002,
+      "loss": 0.6101,
+      "step": 20890
+    },
+    {
+      "epoch": 3.378869937757659,
+      "grad_norm": 0.7774117588996887,
+      "learning_rate": 0.0002,
+      "loss": 0.5995,
+      "step": 20900
+    },
+    {
+      "epoch": 3.380486621938404,
+      "grad_norm": 1.0089964866638184,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 20910
+    },
+    {
+      "epoch": 3.3821033061191494,
+      "grad_norm": 0.7855867147445679,
+      "learning_rate": 0.0002,
+      "loss": 0.5908,
+      "step": 20920
+    },
+    {
+      "epoch": 3.3837199902998947,
+      "grad_norm": 1.3713710308074951,
+      "learning_rate": 0.0002,
+      "loss": 0.6373,
+      "step": 20930
+    },
+    {
+      "epoch": 3.3853366744806404,
+      "grad_norm": 0.8599116206169128,
+      "learning_rate": 0.0002,
+      "loss": 0.6627,
+      "step": 20940
+    },
+    {
+      "epoch": 3.3869533586613856,
+      "grad_norm": 0.9392673373222351,
+      "learning_rate": 0.0002,
+      "loss": 0.6224,
+      "step": 20950
+    },
+    {
+      "epoch": 3.388570042842131,
+      "grad_norm": 0.8764075040817261,
+      "learning_rate": 0.0002,
+      "loss": 0.5855,
+      "step": 20960
+    },
+    {
+      "epoch": 3.390186727022876,
+      "grad_norm": 0.8240136504173279,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 20970
+    },
+    {
+      "epoch": 3.3918034112036213,
+      "grad_norm": 1.0982369184494019,
+      "learning_rate": 0.0002,
+      "loss": 0.5783,
+      "step": 20980
+    },
+    {
+      "epoch": 3.3934200953843665,
+      "grad_norm": 1.0599013566970825,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 20990
+    },
+    {
+      "epoch": 3.395036779565112,
+      "grad_norm": 0.895438015460968,
+      "learning_rate": 0.0002,
+      "loss": 0.6356,
+      "step": 21000
+    },
+    {
+      "epoch": 3.3966534637458574,
+      "grad_norm": 0.6974841356277466,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 21010
+    },
+    {
+      "epoch": 3.3982701479266026,
+      "grad_norm": 0.9571719765663147,
+      "learning_rate": 0.0002,
+      "loss": 0.5704,
+      "step": 21020
+    },
+    {
+      "epoch": 3.399886832107348,
+      "grad_norm": 0.831912636756897,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 21030
+    },
+    {
+      "epoch": 3.401503516288093,
+      "grad_norm": 0.831936240196228,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 21040
+    },
+    {
+      "epoch": 3.4031202004688383,
+      "grad_norm": 0.7388373613357544,
+      "learning_rate": 0.0002,
+      "loss": 0.5857,
+      "step": 21050
+    },
+    {
+      "epoch": 3.4047368846495836,
+      "grad_norm": 0.938667356967926,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21060
+    },
+    {
+      "epoch": 3.406353568830329,
+      "grad_norm": 0.9202313423156738,
+      "learning_rate": 0.0002,
+      "loss": 0.6121,
+      "step": 21070
+    },
+    {
+      "epoch": 3.4079702530110745,
+      "grad_norm": 0.9888381958007812,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 21080
+    },
+    {
+      "epoch": 3.4095869371918197,
+      "grad_norm": 0.8526970744132996,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 21090
+    },
+    {
+      "epoch": 3.411203621372565,
+      "grad_norm": 0.7939383387565613,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 21100
+    },
+    {
+      "epoch": 3.41282030555331,
+      "grad_norm": 0.9986352920532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 21110
+    },
+    {
+      "epoch": 3.4144369897340554,
+      "grad_norm": 0.8895300030708313,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 21120
+    },
+    {
+      "epoch": 3.4160536739148006,
+      "grad_norm": 0.9559482932090759,
+      "learning_rate": 0.0002,
+      "loss": 0.6264,
+      "step": 21130
+    },
+    {
+      "epoch": 3.417670358095546,
+      "grad_norm": 0.8351506590843201,
+      "learning_rate": 0.0002,
+      "loss": 0.6491,
+      "step": 21140
+    },
+    {
+      "epoch": 3.4192870422762915,
+      "grad_norm": 0.8224456906318665,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 21150
+    },
+    {
+      "epoch": 3.4209037264570368,
+      "grad_norm": 1.0110299587249756,
+      "learning_rate": 0.0002,
+      "loss": 0.5871,
+      "step": 21160
+    },
+    {
+      "epoch": 3.422520410637782,
+      "grad_norm": 0.82564777135849,
+      "learning_rate": 0.0002,
+      "loss": 0.6116,
+      "step": 21170
+    },
+    {
+      "epoch": 3.4241370948185272,
+      "grad_norm": 1.004738688468933,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 21180
+    },
+    {
+      "epoch": 3.4257537789992725,
+      "grad_norm": 0.7545676827430725,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 21190
+    },
+    {
+      "epoch": 3.4273704631800177,
+      "grad_norm": 0.8918704390525818,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 21200
+    },
+    {
+      "epoch": 3.428987147360763,
+      "grad_norm": 0.8336876034736633,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 21210
+    },
+    {
+      "epoch": 3.430603831541508,
+      "grad_norm": 0.8928771018981934,
+      "learning_rate": 0.0002,
+      "loss": 0.5824,
+      "step": 21220
+    },
+    {
+      "epoch": 3.432220515722254,
+      "grad_norm": 0.7663705945014954,
+      "learning_rate": 0.0002,
+      "loss": 0.6468,
+      "step": 21230
+    },
+    {
+      "epoch": 3.433837199902999,
+      "grad_norm": 0.8392598628997803,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 21240
+    },
+    {
+      "epoch": 3.4354538840837443,
+      "grad_norm": 0.8819600343704224,
+      "learning_rate": 0.0002,
+      "loss": 0.5971,
+      "step": 21250
+    },
+    {
+      "epoch": 3.4370705682644895,
+      "grad_norm": 0.9124642014503479,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 21260
+    },
+    {
+      "epoch": 3.4386872524452348,
+      "grad_norm": 0.8329763412475586,
+      "learning_rate": 0.0002,
+      "loss": 0.5925,
+      "step": 21270
+    },
+    {
+      "epoch": 3.44030393662598,
+      "grad_norm": 0.9982839822769165,
+      "learning_rate": 0.0002,
+      "loss": 0.6541,
+      "step": 21280
+    },
+    {
+      "epoch": 3.4419206208067252,
+      "grad_norm": 0.9105954766273499,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 21290
+    },
+    {
+      "epoch": 3.443537304987471,
+      "grad_norm": 0.8182359337806702,
+      "learning_rate": 0.0002,
+      "loss": 0.6028,
+      "step": 21300
+    },
+    {
+      "epoch": 3.445153989168216,
+      "grad_norm": 1.0568904876708984,
+      "learning_rate": 0.0002,
+      "loss": 0.5991,
+      "step": 21310
+    },
+    {
+      "epoch": 3.4467706733489614,
+      "grad_norm": 0.968539834022522,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 21320
+    },
+    {
+      "epoch": 3.4483873575297066,
+      "grad_norm": 0.8774511218070984,
+      "learning_rate": 0.0002,
+      "loss": 0.6219,
+      "step": 21330
+    },
+    {
+      "epoch": 3.450004041710452,
+      "grad_norm": 0.7598156332969666,
+      "learning_rate": 0.0002,
+      "loss": 0.6438,
+      "step": 21340
+    },
+    {
+      "epoch": 3.451620725891197,
+      "grad_norm": 1.1012897491455078,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 21350
+    },
+    {
+      "epoch": 3.4532374100719423,
+      "grad_norm": 0.8040637373924255,
+      "learning_rate": 0.0002,
+      "loss": 0.6137,
+      "step": 21360
+    },
+    {
+      "epoch": 3.4548540942526875,
+      "grad_norm": 0.8497496247291565,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 21370
+    },
+    {
+      "epoch": 3.456470778433433,
+      "grad_norm": 0.8429915904998779,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 21380
+    },
+    {
+      "epoch": 3.4580874626141784,
+      "grad_norm": 0.8107112646102905,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 21390
+    },
+    {
+      "epoch": 3.4597041467949237,
+      "grad_norm": 1.00872004032135,
+      "learning_rate": 0.0002,
+      "loss": 0.6109,
+      "step": 21400
+    },
+    {
+      "epoch": 3.461320830975669,
+      "grad_norm": 0.8266542553901672,
+      "learning_rate": 0.0002,
+      "loss": 0.5712,
+      "step": 21410
+    },
+    {
+      "epoch": 3.462937515156414,
+      "grad_norm": 0.8972568511962891,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 21420
+    },
+    {
+      "epoch": 3.4645541993371594,
+      "grad_norm": 1.0781476497650146,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 21430
+    },
+    {
+      "epoch": 3.4661708835179046,
+      "grad_norm": 0.9571592807769775,
+      "learning_rate": 0.0002,
+      "loss": 0.6303,
+      "step": 21440
+    },
+    {
+      "epoch": 3.4677875676986503,
+      "grad_norm": 0.881547212600708,
+      "learning_rate": 0.0002,
+      "loss": 0.6309,
+      "step": 21450
+    },
+    {
+      "epoch": 3.4694042518793955,
+      "grad_norm": 0.6955338716506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 21460
+    },
+    {
+      "epoch": 3.4710209360601407,
+      "grad_norm": 0.901187539100647,
+      "learning_rate": 0.0002,
+      "loss": 0.6205,
+      "step": 21470
+    },
+    {
+      "epoch": 3.472637620240886,
+      "grad_norm": 0.7063511610031128,
+      "learning_rate": 0.0002,
+      "loss": 0.639,
+      "step": 21480
+    },
+    {
+      "epoch": 3.474254304421631,
+      "grad_norm": 0.8462792038917542,
+      "learning_rate": 0.0002,
+      "loss": 0.6154,
+      "step": 21490
+    },
+    {
+      "epoch": 3.4758709886023764,
+      "grad_norm": 1.1861060857772827,
+      "learning_rate": 0.0002,
+      "loss": 0.61,
+      "step": 21500
+    },
+    {
+      "epoch": 3.4774876727831217,
+      "grad_norm": 0.70503169298172,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 21510
+    },
+    {
+      "epoch": 3.479104356963867,
+      "grad_norm": 0.9650066494941711,
+      "learning_rate": 0.0002,
+      "loss": 0.6475,
+      "step": 21520
+    },
+    {
+      "epoch": 3.4807210411446126,
+      "grad_norm": 1.0266852378845215,
+      "learning_rate": 0.0002,
+      "loss": 0.6452,
+      "step": 21530
+    },
+    {
+      "epoch": 3.482337725325358,
+      "grad_norm": 0.956372857093811,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 21540
+    },
+    {
+      "epoch": 3.483954409506103,
+      "grad_norm": 0.8848432898521423,
+      "learning_rate": 0.0002,
+      "loss": 0.6667,
+      "step": 21550
+    },
+    {
+      "epoch": 3.4855710936868483,
+      "grad_norm": 1.0805351734161377,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 21560
+    },
+    {
+      "epoch": 3.4871877778675935,
+      "grad_norm": 0.9279725551605225,
+      "learning_rate": 0.0002,
+      "loss": 0.6958,
+      "step": 21570
+    },
+    {
+      "epoch": 3.4888044620483387,
+      "grad_norm": 0.9049562215805054,
+      "learning_rate": 0.0002,
+      "loss": 0.6354,
+      "step": 21580
+    },
+    {
+      "epoch": 3.4904211462290844,
+      "grad_norm": 0.9619429111480713,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 21590
+    },
+    {
+      "epoch": 3.4920378304098296,
+      "grad_norm": 0.8508906960487366,
+      "learning_rate": 0.0002,
+      "loss": 0.5927,
+      "step": 21600
+    },
+    {
+      "epoch": 3.493654514590575,
+      "grad_norm": 0.8692502379417419,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 21610
+    },
+    {
+      "epoch": 3.49527119877132,
+      "grad_norm": 0.8187332153320312,
+      "learning_rate": 0.0002,
+      "loss": 0.5878,
+      "step": 21620
+    },
+    {
+      "epoch": 3.4968878829520653,
+      "grad_norm": 1.145400047302246,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 21630
+    },
+    {
+      "epoch": 3.4985045671328105,
+      "grad_norm": 0.8281388282775879,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 21640
+    },
+    {
+      "epoch": 3.500121251313556,
+      "grad_norm": 0.82256019115448,
+      "learning_rate": 0.0002,
+      "loss": 0.6624,
+      "step": 21650
+    },
+    {
+      "epoch": 3.501737935494301,
+      "grad_norm": 0.9315484762191772,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 21660
+    },
+    {
+      "epoch": 3.5033546196750462,
+      "grad_norm": 0.7626111507415771,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 21670
+    },
+    {
+      "epoch": 3.504971303855792,
+      "grad_norm": 0.9275059103965759,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 21680
+    },
+    {
+      "epoch": 3.506587988036537,
+      "grad_norm": 0.7906724810600281,
+      "learning_rate": 0.0002,
+      "loss": 0.64,
+      "step": 21690
+    },
+    {
+      "epoch": 3.5082046722172824,
+      "grad_norm": 0.8289761543273926,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 21700
+    },
+    {
+      "epoch": 3.5098213563980276,
+      "grad_norm": 0.8316431045532227,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 21710
+    },
+    {
+      "epoch": 3.511438040578773,
+      "grad_norm": 1.0451812744140625,
+      "learning_rate": 0.0002,
+      "loss": 0.619,
+      "step": 21720
+    },
+    {
+      "epoch": 3.513054724759518,
+      "grad_norm": 0.928252637386322,
+      "learning_rate": 0.0002,
+      "loss": 0.632,
+      "step": 21730
+    },
+    {
+      "epoch": 3.5146714089402638,
+      "grad_norm": 0.7985895276069641,
+      "learning_rate": 0.0002,
+      "loss": 0.6062,
+      "step": 21740
+    },
+    {
+      "epoch": 3.516288093121009,
+      "grad_norm": 0.6740974187850952,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 21750
+    },
+    {
+      "epoch": 3.517904777301754,
+      "grad_norm": 0.8482223749160767,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 21760
+    },
+    {
+      "epoch": 3.5195214614824994,
+      "grad_norm": 0.889947772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.6277,
+      "step": 21770
+    },
+    {
+      "epoch": 3.5211381456632447,
+      "grad_norm": 0.8304598927497864,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 21780
+    },
+    {
+      "epoch": 3.52275482984399,
+      "grad_norm": 0.8002981543540955,
+      "learning_rate": 0.0002,
+      "loss": 0.6156,
+      "step": 21790
+    },
+    {
+      "epoch": 3.524371514024735,
+      "grad_norm": 0.8115083575248718,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21800
+    },
+    {
+      "epoch": 3.5259881982054804,
+      "grad_norm": 0.9715048670768738,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 21810
+    },
+    {
+      "epoch": 3.5276048823862256,
+      "grad_norm": 1.0910786390304565,
+      "learning_rate": 0.0002,
+      "loss": 0.6715,
+      "step": 21820
+    },
+    {
+      "epoch": 3.5292215665669713,
+      "grad_norm": 0.8438942432403564,
+      "learning_rate": 0.0002,
+      "loss": 0.6543,
+      "step": 21830
+    },
+    {
+      "epoch": 3.5308382507477165,
+      "grad_norm": 0.8813382983207703,
+      "learning_rate": 0.0002,
+      "loss": 0.6509,
+      "step": 21840
+    },
+    {
+      "epoch": 3.5324549349284617,
+      "grad_norm": 0.7092908024787903,
+      "learning_rate": 0.0002,
+      "loss": 0.6049,
+      "step": 21850
+    },
+    {
+      "epoch": 3.534071619109207,
+      "grad_norm": 0.8332187533378601,
+      "learning_rate": 0.0002,
+      "loss": 0.5678,
+      "step": 21860
+    },
+    {
+      "epoch": 3.535688303289952,
+      "grad_norm": 0.8958209156990051,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 21870
+    },
+    {
+      "epoch": 3.5373049874706974,
+      "grad_norm": 0.824138879776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 21880
+    },
+    {
+      "epoch": 3.538921671651443,
+      "grad_norm": 0.8375158309936523,
+      "learning_rate": 0.0002,
+      "loss": 0.6022,
+      "step": 21890
+    },
+    {
+      "epoch": 3.5405383558321883,
+      "grad_norm": 1.0274608135223389,
+      "learning_rate": 0.0002,
+      "loss": 0.6019,
+      "step": 21900
+    },
+    {
+      "epoch": 3.5421550400129336,
+      "grad_norm": 0.7088932394981384,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 21910
+    },
+    {
+      "epoch": 3.543771724193679,
+      "grad_norm": 0.8172445297241211,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 21920
+    },
+    {
+      "epoch": 3.545388408374424,
+      "grad_norm": 0.9904135465621948,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 21930
+    },
+    {
+      "epoch": 3.5470050925551693,
+      "grad_norm": 0.9900432229042053,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 21940
+    },
+    {
+      "epoch": 3.5486217767359145,
+      "grad_norm": 0.8963301181793213,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 21950
+    },
+    {
+      "epoch": 3.5502384609166597,
+      "grad_norm": 0.8551464676856995,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 21960
+    },
+    {
+      "epoch": 3.551855145097405,
+      "grad_norm": 1.0916603803634644,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 21970
+    },
+    {
+      "epoch": 3.5534718292781506,
+      "grad_norm": 0.841598391532898,
+      "learning_rate": 0.0002,
+      "loss": 0.6523,
+      "step": 21980
+    },
+    {
+      "epoch": 3.555088513458896,
+      "grad_norm": 0.8566757440567017,
+      "learning_rate": 0.0002,
+      "loss": 0.617,
+      "step": 21990
+    },
+    {
+      "epoch": 3.556705197639641,
+      "grad_norm": 1.0145052671432495,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 22000
+    },
+    {
+      "epoch": 3.5583218818203863,
+      "grad_norm": 0.9293754696846008,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 22010
+    },
+    {
+      "epoch": 3.5599385660011316,
+      "grad_norm": 0.9568536281585693,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 22020
+    },
+    {
+      "epoch": 3.5615552501818772,
+      "grad_norm": 0.8613139986991882,
+      "learning_rate": 0.0002,
+      "loss": 0.641,
+      "step": 22030
+    },
+    {
+      "epoch": 3.5631719343626225,
+      "grad_norm": 0.8179237246513367,
+      "learning_rate": 0.0002,
+      "loss": 0.6496,
+      "step": 22040
+    },
+    {
+      "epoch": 3.5647886185433677,
+      "grad_norm": 0.9059830904006958,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 22050
+    },
+    {
+      "epoch": 3.566405302724113,
+      "grad_norm": 1.0068252086639404,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 22060
+    },
+    {
+      "epoch": 3.568021986904858,
+      "grad_norm": 0.9682072997093201,
+      "learning_rate": 0.0002,
+      "loss": 0.6239,
+      "step": 22070
+    },
+    {
+      "epoch": 3.5696386710856034,
+      "grad_norm": 0.8514005541801453,
+      "learning_rate": 0.0002,
+      "loss": 0.6808,
+      "step": 22080
+    },
+    {
+      "epoch": 3.5712553552663486,
+      "grad_norm": 0.8327770829200745,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 22090
+    },
+    {
+      "epoch": 3.572872039447094,
+      "grad_norm": 1.024976372718811,
+      "learning_rate": 0.0002,
+      "loss": 0.5976,
+      "step": 22100
+    },
+    {
+      "epoch": 3.574488723627839,
+      "grad_norm": 0.7721174955368042,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 22110
+    },
+    {
+      "epoch": 3.5761054078085843,
+      "grad_norm": 1.0351054668426514,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 22120
+    },
+    {
+      "epoch": 3.57772209198933,
+      "grad_norm": 0.9680907130241394,
+      "learning_rate": 0.0002,
+      "loss": 0.6379,
+      "step": 22130
+    },
+    {
+      "epoch": 3.5793387761700752,
+      "grad_norm": 0.8016974925994873,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 22140
+    },
+    {
+      "epoch": 3.5809554603508205,
+      "grad_norm": 1.0109003782272339,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 22150
+    },
+    {
+      "epoch": 3.5825721445315657,
+      "grad_norm": 1.0473392009735107,
+      "learning_rate": 0.0002,
+      "loss": 0.6368,
+      "step": 22160
+    },
+    {
+      "epoch": 3.584188828712311,
+      "grad_norm": 0.8686613440513611,
+      "learning_rate": 0.0002,
+      "loss": 0.6353,
+      "step": 22170
+    },
+    {
+      "epoch": 3.5858055128930566,
+      "grad_norm": 0.869149923324585,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 22180
+    },
+    {
+      "epoch": 3.587422197073802,
+      "grad_norm": 0.9769062995910645,
+      "learning_rate": 0.0002,
+      "loss": 0.5895,
+      "step": 22190
+    },
+    {
+      "epoch": 3.589038881254547,
+      "grad_norm": 0.779636561870575,
+      "learning_rate": 0.0002,
+      "loss": 0.5939,
+      "step": 22200
+    },
+    {
+      "epoch": 3.5906555654352923,
+      "grad_norm": 0.9063841104507446,
+      "learning_rate": 0.0002,
+      "loss": 0.5875,
+      "step": 22210
+    },
+    {
+      "epoch": 3.5922722496160375,
+      "grad_norm": 0.9216037392616272,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 22220
+    },
+    {
+      "epoch": 3.5938889337967828,
+      "grad_norm": 1.0217336416244507,
+      "learning_rate": 0.0002,
+      "loss": 0.6484,
+      "step": 22230
+    },
+    {
+      "epoch": 3.595505617977528,
+      "grad_norm": 0.8513161540031433,
+      "learning_rate": 0.0002,
+      "loss": 0.6511,
+      "step": 22240
+    },
+    {
+      "epoch": 3.597122302158273,
+      "grad_norm": 0.8084813952445984,
+      "learning_rate": 0.0002,
+      "loss": 0.6301,
+      "step": 22250
+    },
+    {
+      "epoch": 3.5987389863390185,
+      "grad_norm": 0.8524802923202515,
+      "learning_rate": 0.0002,
+      "loss": 0.6197,
+      "step": 22260
+    },
+    {
+      "epoch": 3.600355670519764,
+      "grad_norm": 0.9356237649917603,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 22270
+    },
+    {
+      "epoch": 3.6019723547005094,
+      "grad_norm": 1.009600281715393,
+      "learning_rate": 0.0002,
+      "loss": 0.628,
+      "step": 22280
+    },
+    {
+      "epoch": 3.6035890388812546,
+      "grad_norm": 0.9900581240653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 22290
+    },
+    {
+      "epoch": 3.605205723062,
+      "grad_norm": 1.062495231628418,
+      "learning_rate": 0.0002,
+      "loss": 0.5725,
+      "step": 22300
+    },
+    {
+      "epoch": 3.606822407242745,
+      "grad_norm": 0.8832381367683411,
+      "learning_rate": 0.0002,
+      "loss": 0.607,
+      "step": 22310
+    },
+    {
+      "epoch": 3.6084390914234903,
+      "grad_norm": 0.9284297823905945,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 22320
+    },
+    {
+      "epoch": 3.610055775604236,
+      "grad_norm": 1.2381829023361206,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 22330
+    },
+    {
+      "epoch": 3.611672459784981,
+      "grad_norm": 0.929434597492218,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 22340
+    },
+    {
+      "epoch": 3.6132891439657264,
+      "grad_norm": 0.9714490175247192,
+      "learning_rate": 0.0002,
+      "loss": 0.6141,
+      "step": 22350
+    },
+    {
+      "epoch": 3.6149058281464717,
+      "grad_norm": 0.808014988899231,
+      "learning_rate": 0.0002,
+      "loss": 0.6861,
+      "step": 22360
+    },
+    {
+      "epoch": 3.616522512327217,
+      "grad_norm": 1.0364398956298828,
+      "learning_rate": 0.0002,
+      "loss": 0.6428,
+      "step": 22370
+    },
+    {
+      "epoch": 3.618139196507962,
+      "grad_norm": 0.7858489751815796,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22380
+    },
+    {
+      "epoch": 3.6197558806887074,
+      "grad_norm": 0.9920870065689087,
+      "learning_rate": 0.0002,
+      "loss": 0.6214,
+      "step": 22390
+    },
+    {
+      "epoch": 3.6213725648694526,
+      "grad_norm": 0.9183220863342285,
+      "learning_rate": 0.0002,
+      "loss": 0.6659,
+      "step": 22400
+    },
+    {
+      "epoch": 3.622989249050198,
+      "grad_norm": 0.9826246500015259,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22410
+    },
+    {
+      "epoch": 3.6246059332309435,
+      "grad_norm": 0.8632931113243103,
+      "learning_rate": 0.0002,
+      "loss": 0.6441,
+      "step": 22420
+    },
+    {
+      "epoch": 3.6262226174116887,
+      "grad_norm": 0.8468965291976929,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 22430
+    },
+    {
+      "epoch": 3.627839301592434,
+      "grad_norm": 0.8466871976852417,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 22440
+    },
+    {
+      "epoch": 3.629455985773179,
+      "grad_norm": 0.9501169919967651,
+      "learning_rate": 0.0002,
+      "loss": 0.5941,
+      "step": 22450
+    },
+    {
+      "epoch": 3.6310726699539244,
+      "grad_norm": 0.8906720876693726,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 22460
+    },
+    {
+      "epoch": 3.6326893541346696,
+      "grad_norm": 0.7400227189064026,
+      "learning_rate": 0.0002,
+      "loss": 0.6928,
+      "step": 22470
+    },
+    {
+      "epoch": 3.6343060383154153,
+      "grad_norm": 0.9756355881690979,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22480
+    },
+    {
+      "epoch": 3.6359227224961606,
+      "grad_norm": 0.7504993081092834,
+      "learning_rate": 0.0002,
+      "loss": 0.6203,
+      "step": 22490
+    },
+    {
+      "epoch": 3.637539406676906,
+      "grad_norm": 0.9270039200782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 22500
+    },
+    {
+      "epoch": 3.639156090857651,
+      "grad_norm": 0.8841686844825745,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 22510
+    },
+    {
+      "epoch": 3.6407727750383962,
+      "grad_norm": 0.8533213138580322,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 22520
+    },
+    {
+      "epoch": 3.6423894592191415,
+      "grad_norm": 1.0052043199539185,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 22530
+    },
+    {
+      "epoch": 3.6440061433998867,
+      "grad_norm": 1.0323461294174194,
+      "learning_rate": 0.0002,
+      "loss": 0.6363,
+      "step": 22540
+    },
+    {
+      "epoch": 3.645622827580632,
+      "grad_norm": 0.8654312491416931,
+      "learning_rate": 0.0002,
+      "loss": 0.6545,
+      "step": 22550
+    },
+    {
+      "epoch": 3.647239511761377,
+      "grad_norm": 0.6400038003921509,
+      "learning_rate": 0.0002,
+      "loss": 0.6155,
+      "step": 22560
+    },
+    {
+      "epoch": 3.648856195942123,
+      "grad_norm": 0.8061298727989197,
+      "learning_rate": 0.0002,
+      "loss": 0.5829,
+      "step": 22570
+    },
+    {
+      "epoch": 3.650472880122868,
+      "grad_norm": 0.9257854223251343,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 22580
+    },
+    {
+      "epoch": 3.6520895643036133,
+      "grad_norm": 0.8439396619796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6409,
+      "step": 22590
+    },
+    {
+      "epoch": 3.6537062484843585,
+      "grad_norm": 0.7764544486999512,
+      "learning_rate": 0.0002,
+      "loss": 0.5996,
+      "step": 22600
+    },
+    {
+      "epoch": 3.6553229326651038,
+      "grad_norm": 1.125451683998108,
+      "learning_rate": 0.0002,
+      "loss": 0.6434,
+      "step": 22610
+    },
+    {
+      "epoch": 3.656939616845849,
+      "grad_norm": 0.7523018717765808,
+      "learning_rate": 0.0002,
+      "loss": 0.6579,
+      "step": 22620
+    },
+    {
+      "epoch": 3.6585563010265947,
+      "grad_norm": 1.071026086807251,
+      "learning_rate": 0.0002,
+      "loss": 0.6476,
+      "step": 22630
+    },
+    {
+      "epoch": 3.66017298520734,
+      "grad_norm": 0.945791482925415,
+      "learning_rate": 0.0002,
+      "loss": 0.6459,
+      "step": 22640
+    },
+    {
+      "epoch": 3.661789669388085,
+      "grad_norm": 0.8001811504364014,
+      "learning_rate": 0.0002,
+      "loss": 0.659,
+      "step": 22650
+    },
+    {
+      "epoch": 3.6634063535688304,
+      "grad_norm": 0.9700816869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.6385,
+      "step": 22660
+    },
+    {
+      "epoch": 3.6650230377495756,
+      "grad_norm": 0.9053242206573486,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 22670
+    },
+    {
+      "epoch": 3.666639721930321,
+      "grad_norm": 0.944362461566925,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 22680
+    },
+    {
+      "epoch": 3.668256406111066,
+      "grad_norm": 1.067489504814148,
+      "learning_rate": 0.0002,
+      "loss": 0.6235,
+      "step": 22690
+    },
+    {
+      "epoch": 3.6698730902918113,
+      "grad_norm": 1.0984995365142822,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 22700
+    },
+    {
+      "epoch": 3.6714897744725565,
+      "grad_norm": 0.9336317777633667,
+      "learning_rate": 0.0002,
+      "loss": 0.6717,
+      "step": 22710
+    },
+    {
+      "epoch": 3.673106458653302,
+      "grad_norm": 0.9261918663978577,
+      "learning_rate": 0.0002,
+      "loss": 0.6195,
+      "step": 22720
+    },
+    {
+      "epoch": 3.6747231428340474,
+      "grad_norm": 0.8648008704185486,
+      "learning_rate": 0.0002,
+      "loss": 0.6332,
+      "step": 22730
+    },
+    {
+      "epoch": 3.6763398270147927,
+      "grad_norm": 0.7225083708763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6576,
+      "step": 22740
+    },
+    {
+      "epoch": 3.677956511195538,
+      "grad_norm": 0.9258282780647278,
+      "learning_rate": 0.0002,
+      "loss": 0.6406,
+      "step": 22750
+    },
+    {
+      "epoch": 3.679573195376283,
+      "grad_norm": 0.70876145362854,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 22760
+    },
+    {
+      "epoch": 3.681189879557029,
+      "grad_norm": 0.8780210018157959,
+      "learning_rate": 0.0002,
+      "loss": 0.6821,
+      "step": 22770
+    },
+    {
+      "epoch": 3.682806563737774,
+      "grad_norm": 0.8075440526008606,
+      "learning_rate": 0.0002,
+      "loss": 0.6036,
+      "step": 22780
+    },
+    {
+      "epoch": 3.6844232479185193,
+      "grad_norm": 0.8503130674362183,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22790
+    },
+    {
+      "epoch": 3.6860399320992645,
+      "grad_norm": 0.8413618206977844,
+      "learning_rate": 0.0002,
+      "loss": 0.6082,
+      "step": 22800
+    },
+    {
+      "epoch": 3.6876566162800097,
+      "grad_norm": 0.8675165176391602,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 22810
+    },
+    {
+      "epoch": 3.689273300460755,
+      "grad_norm": 0.8235884308815002,
+      "learning_rate": 0.0002,
+      "loss": 0.6157,
+      "step": 22820
+    },
+    {
+      "epoch": 3.6908899846415,
+      "grad_norm": 0.9477725625038147,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 22830
+    },
+    {
+      "epoch": 3.6925066688222454,
+      "grad_norm": 0.7883533835411072,
+      "learning_rate": 0.0002,
+      "loss": 0.6481,
+      "step": 22840
+    },
+    {
+      "epoch": 3.6941233530029907,
+      "grad_norm": 1.047913908958435,
+      "learning_rate": 0.0002,
+      "loss": 0.5872,
+      "step": 22850
+    },
+    {
+      "epoch": 3.695740037183736,
+      "grad_norm": 0.9171528816223145,
+      "learning_rate": 0.0002,
+      "loss": 0.6176,
+      "step": 22860
+    },
+    {
+      "epoch": 3.6973567213644816,
+      "grad_norm": 0.9338192343711853,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 22870
+    },
+    {
+      "epoch": 3.698973405545227,
+      "grad_norm": 0.8799443244934082,
+      "learning_rate": 0.0002,
+      "loss": 0.686,
+      "step": 22880
+    },
+    {
+      "epoch": 3.700590089725972,
+      "grad_norm": 0.8515434861183167,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 22890
+    },
+    {
+      "epoch": 3.7022067739067173,
+      "grad_norm": 0.7805591821670532,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 22900
+    },
+    {
+      "epoch": 3.7038234580874625,
+      "grad_norm": 0.8470911979675293,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 22910
+    },
+    {
+      "epoch": 3.705440142268208,
+      "grad_norm": 0.9452309012413025,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 22920
+    },
+    {
+      "epoch": 3.7070568264489534,
+      "grad_norm": 0.950243353843689,
+      "learning_rate": 0.0002,
+      "loss": 0.6529,
+      "step": 22930
+    },
+    {
+      "epoch": 3.7086735106296986,
+      "grad_norm": 0.7882499098777771,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 22940
+    },
+    {
+      "epoch": 3.710290194810444,
+      "grad_norm": 0.8307787775993347,
+      "learning_rate": 0.0002,
+      "loss": 0.6462,
+      "step": 22950
+    },
+    {
+      "epoch": 3.711906878991189,
+      "grad_norm": 1.0970630645751953,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 22960
+    },
+    {
+      "epoch": 3.7135235631719343,
+      "grad_norm": 0.8269566297531128,
+      "learning_rate": 0.0002,
+      "loss": 0.6281,
+      "step": 22970
+    },
+    {
+      "epoch": 3.7151402473526796,
+      "grad_norm": 0.8306704759597778,
+      "learning_rate": 0.0002,
+      "loss": 0.6561,
+      "step": 22980
+    },
+    {
+      "epoch": 3.716756931533425,
+      "grad_norm": 0.9710225462913513,
+      "learning_rate": 0.0002,
+      "loss": 0.6418,
+      "step": 22990
+    },
+    {
+      "epoch": 3.71837361571417,
+      "grad_norm": 0.8890530467033386,
+      "learning_rate": 0.0002,
+      "loss": 0.6639,
+      "step": 23000
+    },
+    {
+      "epoch": 3.7199902998949153,
+      "grad_norm": 0.883522629737854,
+      "learning_rate": 0.0002,
+      "loss": 0.6084,
+      "step": 23010
+    },
+    {
+      "epoch": 3.721606984075661,
+      "grad_norm": 0.8662652373313904,
+      "learning_rate": 0.0002,
+      "loss": 0.6183,
+      "step": 23020
+    },
+    {
+      "epoch": 3.723223668256406,
+      "grad_norm": 0.7228406667709351,
+      "learning_rate": 0.0002,
+      "loss": 0.6266,
+      "step": 23030
+    },
+    {
+      "epoch": 3.7248403524371514,
+      "grad_norm": 1.060792088508606,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23040
+    },
+    {
+      "epoch": 3.7264570366178966,
+      "grad_norm": 1.0119613409042358,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 23050
+    },
+    {
+      "epoch": 3.728073720798642,
+      "grad_norm": 0.9212996959686279,
+      "learning_rate": 0.0002,
+      "loss": 0.6466,
+      "step": 23060
+    },
+    {
+      "epoch": 3.7296904049793875,
+      "grad_norm": 0.925690233707428,
+      "learning_rate": 0.0002,
+      "loss": 0.6454,
+      "step": 23070
+    },
+    {
+      "epoch": 3.7313070891601328,
+      "grad_norm": 0.8323310613632202,
+      "learning_rate": 0.0002,
+      "loss": 0.615,
+      "step": 23080
+    },
+    {
+      "epoch": 3.732923773340878,
+      "grad_norm": 0.8966048955917358,
+      "learning_rate": 0.0002,
+      "loss": 0.679,
+      "step": 23090
+    },
+    {
+      "epoch": 3.7345404575216232,
+      "grad_norm": 0.8995837569236755,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23100
+    },
+    {
+      "epoch": 3.7361571417023685,
+      "grad_norm": 0.8748890161514282,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23110
+    },
+    {
+      "epoch": 3.7377738258831137,
+      "grad_norm": 0.7985540628433228,
+      "learning_rate": 0.0002,
+      "loss": 0.6246,
+      "step": 23120
+    },
+    {
+      "epoch": 3.739390510063859,
+      "grad_norm": 1.0240917205810547,
+      "learning_rate": 0.0002,
+      "loss": 0.6279,
+      "step": 23130
+    },
+    {
+      "epoch": 3.741007194244604,
+      "grad_norm": 0.9181789755821228,
+      "learning_rate": 0.0002,
+      "loss": 0.6747,
+      "step": 23140
+    },
+    {
+      "epoch": 3.7426238784253494,
+      "grad_norm": 0.8896583914756775,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 23150
+    },
+    {
+      "epoch": 3.744240562606095,
+      "grad_norm": 0.8635515570640564,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 23160
+    },
+    {
+      "epoch": 3.7458572467868403,
+      "grad_norm": 0.8873575329780579,
+      "learning_rate": 0.0002,
+      "loss": 0.6683,
+      "step": 23170
+    },
+    {
+      "epoch": 3.7474739309675855,
+      "grad_norm": 0.9807148575782776,
+      "learning_rate": 0.0002,
+      "loss": 0.6143,
+      "step": 23180
+    },
+    {
+      "epoch": 3.7490906151483308,
+      "grad_norm": 0.900477945804596,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 23190
+    },
+    {
+      "epoch": 3.750707299329076,
+      "grad_norm": 0.9379992485046387,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23200
+    },
+    {
+      "epoch": 3.752323983509821,
+      "grad_norm": 0.9649890661239624,
+      "learning_rate": 0.0002,
+      "loss": 0.6015,
+      "step": 23210
+    },
+    {
+      "epoch": 3.753940667690567,
+      "grad_norm": 0.824442446231842,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 23220
+    },
+    {
+      "epoch": 3.755557351871312,
+      "grad_norm": 0.8896150588989258,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 23230
+    },
+    {
+      "epoch": 3.7571740360520574,
+      "grad_norm": 0.751249372959137,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 23240
+    },
+    {
+      "epoch": 3.7587907202328026,
+      "grad_norm": 0.9392193555831909,
+      "learning_rate": 0.0002,
+      "loss": 0.629,
+      "step": 23250
+    },
+    {
+      "epoch": 3.760407404413548,
+      "grad_norm": 0.9284586310386658,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 23260
+    },
+    {
+      "epoch": 3.762024088594293,
+      "grad_norm": 0.7738175392150879,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23270
+    },
+    {
+      "epoch": 3.7636407727750383,
+      "grad_norm": 0.9252978563308716,
+      "learning_rate": 0.0002,
+      "loss": 0.6743,
+      "step": 23280
+    },
+    {
+      "epoch": 3.7652574569557835,
+      "grad_norm": 0.9501895904541016,
+      "learning_rate": 0.0002,
+      "loss": 0.5984,
+      "step": 23290
+    },
+    {
+      "epoch": 3.7668741411365287,
+      "grad_norm": 0.9416276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.6568,
+      "step": 23300
+    },
+    {
+      "epoch": 3.7684908253172744,
+      "grad_norm": 0.7076631784439087,
+      "learning_rate": 0.0002,
+      "loss": 0.6507,
+      "step": 23310
+    },
+    {
+      "epoch": 3.7701075094980196,
+      "grad_norm": 0.9864492416381836,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 23320
+    },
+    {
+      "epoch": 3.771724193678765,
+      "grad_norm": 0.8450456261634827,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 23330
+    },
+    {
+      "epoch": 3.77334087785951,
+      "grad_norm": 1.0768941640853882,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23340
+    },
+    {
+      "epoch": 3.7749575620402553,
+      "grad_norm": 0.9956819415092468,
+      "learning_rate": 0.0002,
+      "loss": 0.6408,
+      "step": 23350
+    },
+    {
+      "epoch": 3.7765742462210006,
+      "grad_norm": 0.9234658479690552,
+      "learning_rate": 0.0002,
+      "loss": 0.6464,
+      "step": 23360
+    },
+    {
+      "epoch": 3.7781909304017463,
+      "grad_norm": 1.0993858575820923,
+      "learning_rate": 0.0002,
+      "loss": 0.6542,
+      "step": 23370
+    },
+    {
+      "epoch": 3.7798076145824915,
+      "grad_norm": 0.923159658908844,
+      "learning_rate": 0.0002,
+      "loss": 0.6391,
+      "step": 23380
+    },
+    {
+      "epoch": 3.7814242987632367,
+      "grad_norm": 0.9311541318893433,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23390
+    },
+    {
+      "epoch": 3.783040982943982,
+      "grad_norm": 0.919681191444397,
+      "learning_rate": 0.0002,
+      "loss": 0.6535,
+      "step": 23400
+    },
+    {
+      "epoch": 3.784657667124727,
+      "grad_norm": 1.7406195402145386,
+      "learning_rate": 0.0002,
+      "loss": 0.6138,
+      "step": 23410
+    },
+    {
+      "epoch": 3.7862743513054724,
+      "grad_norm": 0.7789074182510376,
+      "learning_rate": 0.0002,
+      "loss": 0.657,
+      "step": 23420
+    },
+    {
+      "epoch": 3.7878910354862176,
+      "grad_norm": 0.8302814960479736,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23430
+    },
+    {
+      "epoch": 3.789507719666963,
+      "grad_norm": 0.8089349269866943,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23440
+    },
+    {
+      "epoch": 3.791124403847708,
+      "grad_norm": 0.9006284475326538,
+      "learning_rate": 0.0002,
+      "loss": 0.6682,
+      "step": 23450
+    },
+    {
+      "epoch": 3.7927410880284538,
+      "grad_norm": 0.8426766991615295,
+      "learning_rate": 0.0002,
+      "loss": 0.6335,
+      "step": 23460
+    },
+    {
+      "epoch": 3.794357772209199,
+      "grad_norm": 1.2576252222061157,
+      "learning_rate": 0.0002,
+      "loss": 0.6364,
+      "step": 23470
+    },
+    {
+      "epoch": 3.7959744563899442,
+      "grad_norm": 1.0307610034942627,
+      "learning_rate": 0.0002,
+      "loss": 0.6324,
+      "step": 23480
+    },
+    {
+      "epoch": 3.7975911405706895,
+      "grad_norm": 0.8525972962379456,
+      "learning_rate": 0.0002,
+      "loss": 0.6262,
+      "step": 23490
+    },
+    {
+      "epoch": 3.7992078247514347,
+      "grad_norm": 1.159039855003357,
+      "learning_rate": 0.0002,
+      "loss": 0.6757,
+      "step": 23500
+    },
+    {
+      "epoch": 3.80082450893218,
+      "grad_norm": 1.4193549156188965,
+      "learning_rate": 0.0002,
+      "loss": 0.6414,
+      "step": 23510
+    },
+    {
+      "epoch": 3.8024411931129256,
+      "grad_norm": 0.8245543837547302,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 23520
+    },
+    {
+      "epoch": 3.804057877293671,
+      "grad_norm": 0.8847230076789856,
+      "learning_rate": 0.0002,
+      "loss": 0.6417,
+      "step": 23530
+    },
+    {
+      "epoch": 3.805674561474416,
+      "grad_norm": 0.9574624300003052,
+      "learning_rate": 0.0002,
+      "loss": 0.6415,
+      "step": 23540
+    },
+    {
+      "epoch": 3.8072912456551613,
+      "grad_norm": 1.048020601272583,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 23550
+    },
+    {
+      "epoch": 3.8089079298359065,
+      "grad_norm": 0.8302255868911743,
+      "learning_rate": 0.0002,
+      "loss": 0.6497,
+      "step": 23560
+    },
+    {
+      "epoch": 3.8105246140166518,
+      "grad_norm": 0.8269215822219849,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 23570
+    },
+    {
+      "epoch": 3.812141298197397,
+      "grad_norm": 0.9375753402709961,
+      "learning_rate": 0.0002,
+      "loss": 0.6294,
+      "step": 23580
+    },
+    {
+      "epoch": 3.8137579823781422,
+      "grad_norm": 1.0234097242355347,
+      "learning_rate": 0.0002,
+      "loss": 0.6132,
+      "step": 23590
+    },
+    {
+      "epoch": 3.8153746665588875,
+      "grad_norm": 0.8978445529937744,
+      "learning_rate": 0.0002,
+      "loss": 0.6625,
+      "step": 23600
+    },
+    {
+      "epoch": 3.816991350739633,
+      "grad_norm": 0.7929515838623047,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 23610
+    },
+    {
+      "epoch": 3.8186080349203784,
+      "grad_norm": 1.3255881071090698,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23620
+    },
+    {
+      "epoch": 3.8202247191011236,
+      "grad_norm": 0.9188598990440369,
+      "learning_rate": 0.0002,
+      "loss": 0.5947,
+      "step": 23630
+    },
+    {
+      "epoch": 3.821841403281869,
+      "grad_norm": 0.8811675906181335,
+      "learning_rate": 0.0002,
+      "loss": 0.6152,
+      "step": 23640
+    },
+    {
+      "epoch": 3.823458087462614,
+      "grad_norm": 0.8061038255691528,
+      "learning_rate": 0.0002,
+      "loss": 0.6253,
+      "step": 23650
+    },
+    {
+      "epoch": 3.8250747716433597,
+      "grad_norm": 0.9975376129150391,
+      "learning_rate": 0.0002,
+      "loss": 0.6517,
+      "step": 23660
+    },
+    {
+      "epoch": 3.826691455824105,
+      "grad_norm": 0.8036105036735535,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 23670
+    },
+    {
+      "epoch": 3.82830814000485,
+      "grad_norm": 0.7401984333992004,
+      "learning_rate": 0.0002,
+      "loss": 0.6845,
+      "step": 23680
+    },
+    {
+      "epoch": 3.8299248241855954,
+      "grad_norm": 0.829753041267395,
+      "learning_rate": 0.0002,
+      "loss": 0.6423,
+      "step": 23690
+    },
+    {
+      "epoch": 3.8315415083663407,
+      "grad_norm": 0.8753240704536438,
+      "learning_rate": 0.0002,
+      "loss": 0.6611,
+      "step": 23700
+    },
+    {
+      "epoch": 3.833158192547086,
+      "grad_norm": 0.8157842755317688,
+      "learning_rate": 0.0002,
+      "loss": 0.6686,
+      "step": 23710
+    },
+    {
+      "epoch": 3.834774876727831,
+      "grad_norm": 0.6183798909187317,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 23720
+    },
+    {
+      "epoch": 3.8363915609085764,
+      "grad_norm": 0.9548442363739014,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 23730
+    },
+    {
+      "epoch": 3.8380082450893216,
+      "grad_norm": 0.8319669961929321,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23740
+    },
+    {
+      "epoch": 3.839624929270067,
+      "grad_norm": 0.9718693494796753,
+      "learning_rate": 0.0002,
+      "loss": 0.6585,
+      "step": 23750
+    },
+    {
+      "epoch": 3.8412416134508125,
+      "grad_norm": 0.8672235012054443,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 23760
+    },
+    {
+      "epoch": 3.8428582976315577,
+      "grad_norm": 1.1210707426071167,
+      "learning_rate": 0.0002,
+      "loss": 0.6774,
+      "step": 23770
+    },
+    {
+      "epoch": 3.844474981812303,
+      "grad_norm": 0.9177767634391785,
+      "learning_rate": 0.0002,
+      "loss": 0.5923,
+      "step": 23780
+    },
+    {
+      "epoch": 3.846091665993048,
+      "grad_norm": 0.8714171648025513,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 23790
+    },
+    {
+      "epoch": 3.8477083501737934,
+      "grad_norm": 1.1853246688842773,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 23800
+    },
+    {
+      "epoch": 3.849325034354539,
+      "grad_norm": 0.8091260194778442,
+      "learning_rate": 0.0002,
+      "loss": 0.6144,
+      "step": 23810
+    },
+    {
+      "epoch": 3.8509417185352843,
+      "grad_norm": 0.9710774421691895,
+      "learning_rate": 0.0002,
+      "loss": 0.658,
+      "step": 23820
+    },
+    {
+      "epoch": 3.8525584027160296,
+      "grad_norm": 0.7648707628250122,
+      "learning_rate": 0.0002,
+      "loss": 0.6151,
+      "step": 23830
+    },
+    {
+      "epoch": 3.854175086896775,
+      "grad_norm": 0.7809253931045532,
+      "learning_rate": 0.0002,
+      "loss": 0.6013,
+      "step": 23840
+    },
+    {
+      "epoch": 3.85579177107752,
+      "grad_norm": 0.8337951898574829,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 23850
+    },
+    {
+      "epoch": 3.8574084552582653,
+      "grad_norm": 0.9271913170814514,
+      "learning_rate": 0.0002,
+      "loss": 0.6456,
+      "step": 23860
+    },
+    {
+      "epoch": 3.8590251394390105,
+      "grad_norm": 0.985334038734436,
+      "learning_rate": 0.0002,
+      "loss": 0.6671,
+      "step": 23870
+    },
+    {
+      "epoch": 3.8606418236197557,
+      "grad_norm": 0.8458583354949951,
+      "learning_rate": 0.0002,
+      "loss": 0.6693,
+      "step": 23880
+    },
+    {
+      "epoch": 3.862258507800501,
+      "grad_norm": 1.015348196029663,
+      "learning_rate": 0.0002,
+      "loss": 0.6207,
+      "step": 23890
+    },
+    {
+      "epoch": 3.8638751919812466,
+      "grad_norm": 1.0121688842773438,
+      "learning_rate": 0.0002,
+      "loss": 0.649,
+      "step": 23900
+    },
+    {
+      "epoch": 3.865491876161992,
+      "grad_norm": 0.8883971571922302,
+      "learning_rate": 0.0002,
+      "loss": 0.5921,
+      "step": 23910
+    },
+    {
+      "epoch": 3.867108560342737,
+      "grad_norm": 1.028086543083191,
+      "learning_rate": 0.0002,
+      "loss": 0.6597,
+      "step": 23920
+    },
+    {
+      "epoch": 3.8687252445234823,
+      "grad_norm": 0.9645734429359436,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 23930
+    },
+    {
+      "epoch": 3.8703419287042276,
+      "grad_norm": 0.8235350251197815,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 23940
+    },
+    {
+      "epoch": 3.871958612884973,
+      "grad_norm": 1.0298916101455688,
+      "learning_rate": 0.0002,
+      "loss": 0.6387,
+      "step": 23950
+    },
+    {
+      "epoch": 3.8735752970657185,
+      "grad_norm": 1.0063377618789673,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 23960
+    },
+    {
+      "epoch": 3.8751919812464637,
+      "grad_norm": 0.9230626821517944,
+      "learning_rate": 0.0002,
+      "loss": 0.6234,
+      "step": 23970
+    },
+    {
+      "epoch": 3.876808665427209,
+      "grad_norm": 0.9243063926696777,
+      "learning_rate": 0.0002,
+      "loss": 0.6159,
+      "step": 23980
+    },
+    {
+      "epoch": 3.878425349607954,
+      "grad_norm": 1.0211291313171387,
+      "learning_rate": 0.0002,
+      "loss": 0.6035,
+      "step": 23990
+    },
+    {
+      "epoch": 3.8800420337886994,
+      "grad_norm": 0.7800535559654236,
+      "learning_rate": 0.0002,
+      "loss": 0.6351,
+      "step": 24000
+    },
+    {
+      "epoch": 3.8816587179694446,
+      "grad_norm": 0.7904248833656311,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 24010
+    },
+    {
+      "epoch": 3.88327540215019,
+      "grad_norm": 1.1975988149642944,
+      "learning_rate": 0.0002,
+      "loss": 0.6516,
+      "step": 24020
+    },
+    {
+      "epoch": 3.884892086330935,
+      "grad_norm": 1.0626593828201294,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 24030
+    },
+    {
+      "epoch": 3.8865087705116803,
+      "grad_norm": 0.9012193083763123,
+      "learning_rate": 0.0002,
+      "loss": 0.6115,
+      "step": 24040
+    },
+    {
+      "epoch": 3.888125454692426,
+      "grad_norm": 1.1159172058105469,
+      "learning_rate": 0.0002,
+      "loss": 0.6786,
+      "step": 24050
+    },
+    {
+      "epoch": 3.889742138873171,
+      "grad_norm": 1.276838779449463,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 24060
+    },
+    {
+      "epoch": 3.8913588230539164,
+      "grad_norm": 0.8467690348625183,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 24070
+    },
+    {
+      "epoch": 3.8929755072346617,
+      "grad_norm": 0.9862841963768005,
+      "learning_rate": 0.0002,
+      "loss": 0.6655,
+      "step": 24080
+    },
+    {
+      "epoch": 3.894592191415407,
+      "grad_norm": 0.7134621739387512,
+      "learning_rate": 0.0002,
+      "loss": 0.6098,
+      "step": 24090
+    },
+    {
+      "epoch": 3.896208875596152,
+      "grad_norm": 0.8178175091743469,
+      "learning_rate": 0.0002,
+      "loss": 0.618,
+      "step": 24100
+    },
+    {
+      "epoch": 3.897825559776898,
+      "grad_norm": 0.9229172468185425,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 24110
+    },
+    {
+      "epoch": 3.899442243957643,
+      "grad_norm": 1.0878316164016724,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 24120
+    },
+    {
+      "epoch": 3.9010589281383883,
+      "grad_norm": 0.971645712852478,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24130
+    },
+    {
+      "epoch": 3.9026756123191335,
+      "grad_norm": 0.8862188458442688,
+      "learning_rate": 0.0002,
+      "loss": 0.6228,
+      "step": 24140
+    },
+    {
+      "epoch": 3.9042922964998787,
+      "grad_norm": 0.9126982688903809,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 24150
+    },
+    {
+      "epoch": 3.905908980680624,
+      "grad_norm": 0.8833470940589905,
+      "learning_rate": 0.0002,
+      "loss": 0.6734,
+      "step": 24160
+    },
+    {
+      "epoch": 3.907525664861369,
+      "grad_norm": 0.8320947885513306,
+      "learning_rate": 0.0002,
+      "loss": 0.5832,
+      "step": 24170
+    },
+    {
+      "epoch": 3.9091423490421144,
+      "grad_norm": 0.9156602025032043,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 24180
+    },
+    {
+      "epoch": 3.9107590332228597,
+      "grad_norm": 1.029181957244873,
+      "learning_rate": 0.0002,
+      "loss": 0.6678,
+      "step": 24190
+    },
+    {
+      "epoch": 3.9123757174036053,
+      "grad_norm": 0.9052802324295044,
+      "learning_rate": 0.0002,
+      "loss": 0.6565,
+      "step": 24200
+    },
+    {
+      "epoch": 3.9139924015843506,
+      "grad_norm": 0.8847255110740662,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 24210
+    },
+    {
+      "epoch": 3.915609085765096,
+      "grad_norm": 0.9642062187194824,
+      "learning_rate": 0.0002,
+      "loss": 0.6343,
+      "step": 24220
+    },
+    {
+      "epoch": 3.917225769945841,
+      "grad_norm": 0.8629093766212463,
+      "learning_rate": 0.0002,
+      "loss": 0.6557,
+      "step": 24230
+    },
+    {
+      "epoch": 3.9188424541265863,
+      "grad_norm": 0.8674976825714111,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 24240
+    },
+    {
+      "epoch": 3.9204591383073315,
+      "grad_norm": 1.104846477508545,
+      "learning_rate": 0.0002,
+      "loss": 0.5874,
+      "step": 24250
+    },
+    {
+      "epoch": 3.922075822488077,
+      "grad_norm": 1.0874955654144287,
+      "learning_rate": 0.0002,
+      "loss": 0.6501,
+      "step": 24260
+    },
+    {
+      "epoch": 3.9236925066688224,
+      "grad_norm": 0.8689812421798706,
+      "learning_rate": 0.0002,
+      "loss": 0.6455,
+      "step": 24270
+    },
+    {
+      "epoch": 3.9253091908495676,
+      "grad_norm": 0.9724617004394531,
+      "learning_rate": 0.0002,
+      "loss": 0.5893,
+      "step": 24280
+    },
+    {
+      "epoch": 3.926925875030313,
+      "grad_norm": 0.9165538549423218,
+      "learning_rate": 0.0002,
+      "loss": 0.6616,
+      "step": 24290
+    },
+    {
+      "epoch": 3.928542559211058,
+      "grad_norm": 0.9307710528373718,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 24300
+    },
+    {
+      "epoch": 3.9301592433918033,
+      "grad_norm": 0.8589295148849487,
+      "learning_rate": 0.0002,
+      "loss": 0.6071,
+      "step": 24310
+    },
+    {
+      "epoch": 3.9317759275725486,
+      "grad_norm": 0.9151099920272827,
+      "learning_rate": 0.0002,
+      "loss": 0.6662,
+      "step": 24320
+    },
+    {
+      "epoch": 3.933392611753294,
+      "grad_norm": 0.9633517265319824,
+      "learning_rate": 0.0002,
+      "loss": 0.7075,
+      "step": 24330
+    },
+    {
+      "epoch": 3.935009295934039,
+      "grad_norm": 0.9521116018295288,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24340
+    },
+    {
+      "epoch": 3.9366259801147847,
+      "grad_norm": 0.8366776704788208,
+      "learning_rate": 0.0002,
+      "loss": 0.6457,
+      "step": 24350
+    },
+    {
+      "epoch": 3.93824266429553,
+      "grad_norm": 0.8972663283348083,
+      "learning_rate": 0.0002,
+      "loss": 0.6139,
+      "step": 24360
+    },
+    {
+      "epoch": 3.939859348476275,
+      "grad_norm": 0.8102919459342957,
+      "learning_rate": 0.0002,
+      "loss": 0.661,
+      "step": 24370
+    },
+    {
+      "epoch": 3.9414760326570204,
+      "grad_norm": 0.8189975023269653,
+      "learning_rate": 0.0002,
+      "loss": 0.6388,
+      "step": 24380
+    },
+    {
+      "epoch": 3.9430927168377656,
+      "grad_norm": 0.9569464921951294,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 24390
+    },
+    {
+      "epoch": 3.9447094010185113,
+      "grad_norm": 0.7459101676940918,
+      "learning_rate": 0.0002,
+      "loss": 0.6999,
+      "step": 24400
+    },
+    {
+      "epoch": 3.9463260851992565,
+      "grad_norm": 0.8536974787712097,
+      "learning_rate": 0.0002,
+      "loss": 0.6069,
+      "step": 24410
+    },
+    {
+      "epoch": 3.9479427693800018,
+      "grad_norm": 0.8763698935508728,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 24420
+    },
+    {
+      "epoch": 3.949559453560747,
+      "grad_norm": 0.9381106495857239,
+      "learning_rate": 0.0002,
+      "loss": 0.6478,
+      "step": 24430
+    },
+    {
+      "epoch": 3.9511761377414922,
+      "grad_norm": 0.934440016746521,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 24440
+    },
+    {
+      "epoch": 3.9527928219222375,
+      "grad_norm": 0.903918981552124,
+      "learning_rate": 0.0002,
+      "loss": 0.6393,
+      "step": 24450
+    },
+    {
+      "epoch": 3.9544095061029827,
+      "grad_norm": 0.8771953582763672,
+      "learning_rate": 0.0002,
+      "loss": 0.6175,
+      "step": 24460
+    },
+    {
+      "epoch": 3.956026190283728,
+      "grad_norm": 1.0375410318374634,
+      "learning_rate": 0.0002,
+      "loss": 0.6971,
+      "step": 24470
+    },
+    {
+      "epoch": 3.957642874464473,
+      "grad_norm": 0.9439185261726379,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 24480
+    },
+    {
+      "epoch": 3.9592595586452184,
+      "grad_norm": 0.935467004776001,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 24490
+    },
+    {
+      "epoch": 3.960876242825964,
+      "grad_norm": 0.6900772452354431,
+      "learning_rate": 0.0002,
+      "loss": 0.6437,
+      "step": 24500
+    },
+    {
+      "epoch": 3.9624929270067093,
+      "grad_norm": 1.0172916650772095,
+      "learning_rate": 0.0002,
+      "loss": 0.6445,
+      "step": 24510
+    },
+    {
+      "epoch": 3.9641096111874545,
+      "grad_norm": 0.9167046546936035,
+      "learning_rate": 0.0002,
+      "loss": 0.6308,
+      "step": 24520
+    },
+    {
+      "epoch": 3.9657262953681998,
+      "grad_norm": 0.7230527997016907,
+      "learning_rate": 0.0002,
+      "loss": 0.6519,
+      "step": 24530
+    },
+    {
+      "epoch": 3.967342979548945,
+      "grad_norm": 0.8980403542518616,
+      "learning_rate": 0.0002,
+      "loss": 0.6564,
+      "step": 24540
+    },
+    {
+      "epoch": 3.9689596637296907,
+      "grad_norm": 0.8555465936660767,
+      "learning_rate": 0.0002,
+      "loss": 0.6099,
+      "step": 24550
+    },
+    {
+      "epoch": 3.970576347910436,
+      "grad_norm": 0.7825445532798767,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 24560
+    },
+    {
+      "epoch": 3.972193032091181,
+      "grad_norm": 0.7273133993148804,
+      "learning_rate": 0.0002,
+      "loss": 0.604,
+      "step": 24570
+    },
+    {
+      "epoch": 3.9738097162719264,
+      "grad_norm": 0.9612047672271729,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 24580
+    },
+    {
+      "epoch": 3.9754264004526716,
+      "grad_norm": 0.9865460991859436,
+      "learning_rate": 0.0002,
+      "loss": 0.6426,
+      "step": 24590
+    },
+    {
+      "epoch": 3.977043084633417,
+      "grad_norm": 0.8638762831687927,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 24600
+    },
+    {
+      "epoch": 3.978659768814162,
+      "grad_norm": 1.0096198320388794,
+      "learning_rate": 0.0002,
+      "loss": 0.6097,
+      "step": 24610
+    },
+    {
+      "epoch": 3.9802764529949073,
+      "grad_norm": 0.8475532531738281,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 24620
+    },
+    {
+      "epoch": 3.9818931371756525,
+      "grad_norm": 0.9696195721626282,
+      "learning_rate": 0.0002,
+      "loss": 0.6711,
+      "step": 24630
+    },
+    {
+      "epoch": 3.9835098213563978,
+      "grad_norm": 0.7499843239784241,
+      "learning_rate": 0.0002,
+      "loss": 0.6446,
+      "step": 24640
+    },
+    {
+      "epoch": 3.9851265055371434,
+      "grad_norm": 0.8865424990653992,
+      "learning_rate": 0.0002,
+      "loss": 0.6054,
+      "step": 24650
+    },
+    {
+      "epoch": 3.9867431897178887,
+      "grad_norm": 0.8089959025382996,
+      "learning_rate": 0.0002,
+      "loss": 0.5975,
+      "step": 24660
+    },
+    {
+      "epoch": 3.988359873898634,
+      "grad_norm": 0.6946012377738953,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 24670
+    },
+    {
+      "epoch": 3.989976558079379,
+      "grad_norm": 0.7991759181022644,
+      "learning_rate": 0.0002,
+      "loss": 0.6329,
+      "step": 24680
+    },
+    {
+      "epoch": 3.9915932422601244,
+      "grad_norm": 0.8803931474685669,
+      "learning_rate": 0.0002,
+      "loss": 0.6449,
+      "step": 24690
+    },
+    {
+      "epoch": 3.99320992644087,
+      "grad_norm": 0.8848299980163574,
+      "learning_rate": 0.0002,
+      "loss": 0.7091,
+      "step": 24700
+    },
+    {
+      "epoch": 3.9948266106216153,
+      "grad_norm": 0.7448889017105103,
+      "learning_rate": 0.0002,
+      "loss": 0.6551,
+      "step": 24710
+    },
+    {
+      "epoch": 3.9964432948023605,
+      "grad_norm": 0.9361620545387268,
+      "learning_rate": 0.0002,
+      "loss": 0.6432,
+      "step": 24720
+    },
+    {
+      "epoch": 3.9980599789831057,
+      "grad_norm": 0.9958081245422363,
+      "learning_rate": 0.0002,
+      "loss": 0.5917,
+      "step": 24730
+    },
+    {
+      "epoch": 3.999676663163851,
+      "grad_norm": 1.026004672050476,
+      "learning_rate": 0.0002,
+      "loss": 0.6567,
+      "step": 24740
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.1524168252944946,
+      "eval_runtime": 122.1585,
+      "eval_samples_per_second": 6.0,
+      "eval_steps_per_second": 0.753,
+      "step": 24742
+    },
+    {
+      "epoch": 4.001293347344596,
+      "grad_norm": 1.0664808750152588,
+      "learning_rate": 0.0002,
+      "loss": 0.6057,
+      "step": 24750
+    },
+    {
+      "epoch": 4.002910031525341,
+      "grad_norm": 1.0113720893859863,
+      "learning_rate": 0.0002,
+      "loss": 0.5644,
+      "step": 24760
+    },
+    {
+      "epoch": 4.004526715706087,
+      "grad_norm": 0.991486668586731,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 24770
+    },
+    {
+      "epoch": 4.006143399886832,
+      "grad_norm": 0.951754629611969,
+      "learning_rate": 0.0002,
+      "loss": 0.508,
+      "step": 24780
+    },
+    {
+      "epoch": 4.007760084067577,
+      "grad_norm": 1.13059401512146,
+      "learning_rate": 0.0002,
+      "loss": 0.5314,
+      "step": 24790
+    },
+    {
+      "epoch": 4.009376768248322,
+      "grad_norm": 0.9343926310539246,
+      "learning_rate": 0.0002,
+      "loss": 0.5323,
+      "step": 24800
+    },
+    {
+      "epoch": 4.010993452429068,
+      "grad_norm": 1.0680590867996216,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 24810
+    },
+    {
+      "epoch": 4.012610136609814,
+      "grad_norm": 1.0022706985473633,
+      "learning_rate": 0.0002,
+      "loss": 0.513,
+      "step": 24820
+    },
+    {
+      "epoch": 4.014226820790559,
+      "grad_norm": 1.0285297632217407,
+      "learning_rate": 0.0002,
+      "loss": 0.543,
+      "step": 24830
+    },
+    {
+      "epoch": 4.015843504971304,
+      "grad_norm": 0.8347002863883972,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 24840
+    },
+    {
+      "epoch": 4.017460189152049,
+      "grad_norm": 0.9675396680831909,
+      "learning_rate": 0.0002,
+      "loss": 0.5655,
+      "step": 24850
+    },
+    {
+      "epoch": 4.019076873332795,
+      "grad_norm": 0.9238511323928833,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 24860
+    },
+    {
+      "epoch": 4.02069355751354,
+      "grad_norm": 1.1576941013336182,
+      "learning_rate": 0.0002,
+      "loss": 0.5327,
+      "step": 24870
+    },
+    {
+      "epoch": 4.022310241694285,
+      "grad_norm": 0.8583757281303406,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 24880
+    },
+    {
+      "epoch": 4.02392692587503,
+      "grad_norm": 0.9816817045211792,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 24890
+    },
+    {
+      "epoch": 4.0255436100557755,
+      "grad_norm": 0.955073893070221,
+      "learning_rate": 0.0002,
+      "loss": 0.5605,
+      "step": 24900
+    },
+    {
+      "epoch": 4.027160294236521,
+      "grad_norm": 1.1054974794387817,
+      "learning_rate": 0.0002,
+      "loss": 0.4896,
+      "step": 24910
+    },
+    {
+      "epoch": 4.028776978417266,
+      "grad_norm": 1.1240060329437256,
+      "learning_rate": 0.0002,
+      "loss": 0.5246,
+      "step": 24920
+    },
+    {
+      "epoch": 4.030393662598011,
+      "grad_norm": 0.9512825012207031,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 24930
+    },
+    {
+      "epoch": 4.0320103467787565,
+      "grad_norm": 0.85965496301651,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 24940
+    },
+    {
+      "epoch": 4.033627030959502,
+      "grad_norm": 0.9378061294555664,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 24950
+    },
+    {
+      "epoch": 4.035243715140247,
+      "grad_norm": 0.9655424356460571,
+      "learning_rate": 0.0002,
+      "loss": 0.5008,
+      "step": 24960
+    },
+    {
+      "epoch": 4.036860399320993,
+      "grad_norm": 1.1393707990646362,
+      "learning_rate": 0.0002,
+      "loss": 0.5538,
+      "step": 24970
+    },
+    {
+      "epoch": 4.038477083501738,
+      "grad_norm": 1.0220451354980469,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 24980
+    },
+    {
+      "epoch": 4.0400937676824835,
+      "grad_norm": 0.9785808324813843,
+      "learning_rate": 0.0002,
+      "loss": 0.5813,
+      "step": 24990
+    },
+    {
+      "epoch": 4.041710451863229,
+      "grad_norm": 1.0257649421691895,
+      "learning_rate": 0.0002,
+      "loss": 0.5153,
+      "step": 25000
+    },
+    {
+      "epoch": 4.043327136043974,
+      "grad_norm": 0.9737892150878906,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 25010
+    },
+    {
+      "epoch": 4.044943820224719,
+      "grad_norm": 0.7416959404945374,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 25020
+    },
+    {
+      "epoch": 4.046560504405464,
+      "grad_norm": 0.7909596562385559,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 25030
+    },
+    {
+      "epoch": 4.04817718858621,
+      "grad_norm": 0.8923130631446838,
+      "learning_rate": 0.0002,
+      "loss": 0.5265,
+      "step": 25040
+    },
+    {
+      "epoch": 4.049793872766955,
+      "grad_norm": 0.9044941663742065,
+      "learning_rate": 0.0002,
+      "loss": 0.5035,
+      "step": 25050
+    },
+    {
+      "epoch": 4.0514105569477,
+      "grad_norm": 0.866352379322052,
+      "learning_rate": 0.0002,
+      "loss": 0.5135,
+      "step": 25060
+    },
+    {
+      "epoch": 4.053027241128445,
+      "grad_norm": 1.544549822807312,
+      "learning_rate": 0.0002,
+      "loss": 0.5956,
+      "step": 25070
+    },
+    {
+      "epoch": 4.054643925309191,
+      "grad_norm": 0.8426995277404785,
+      "learning_rate": 0.0002,
+      "loss": 0.5418,
+      "step": 25080
+    },
+    {
+      "epoch": 4.056260609489936,
+      "grad_norm": 0.9797548651695251,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 25090
+    },
+    {
+      "epoch": 4.057877293670681,
+      "grad_norm": 0.8468434810638428,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 25100
+    },
+    {
+      "epoch": 4.059493977851426,
+      "grad_norm": 0.9294559955596924,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 25110
+    },
+    {
+      "epoch": 4.061110662032172,
+      "grad_norm": 0.9686688780784607,
+      "learning_rate": 0.0002,
+      "loss": 0.5295,
+      "step": 25120
+    },
+    {
+      "epoch": 4.062727346212918,
+      "grad_norm": 0.8042728304862976,
+      "learning_rate": 0.0002,
+      "loss": 0.5642,
+      "step": 25130
+    },
+    {
+      "epoch": 4.064344030393663,
+      "grad_norm": 1.165160894393921,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 25140
+    },
+    {
+      "epoch": 4.065960714574408,
+      "grad_norm": 1.2161961793899536,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 25150
+    },
+    {
+      "epoch": 4.067577398755153,
+      "grad_norm": 1.0762810707092285,
+      "learning_rate": 0.0002,
+      "loss": 0.5217,
+      "step": 25160
+    },
+    {
+      "epoch": 4.069194082935899,
+      "grad_norm": 0.7580869793891907,
+      "learning_rate": 0.0002,
+      "loss": 0.5886,
+      "step": 25170
+    },
+    {
+      "epoch": 4.070810767116644,
+      "grad_norm": 0.9630117416381836,
+      "learning_rate": 0.0002,
+      "loss": 0.5401,
+      "step": 25180
+    },
+    {
+      "epoch": 4.072427451297389,
+      "grad_norm": 0.9049716591835022,
+      "learning_rate": 0.0002,
+      "loss": 0.5378,
+      "step": 25190
+    },
+    {
+      "epoch": 4.074044135478134,
+      "grad_norm": 1.1536930799484253,
+      "learning_rate": 0.0002,
+      "loss": 0.5266,
+      "step": 25200
+    },
+    {
+      "epoch": 4.0756608196588795,
+      "grad_norm": 0.901461124420166,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 25210
+    },
+    {
+      "epoch": 4.077277503839625,
+      "grad_norm": 1.3318437337875366,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 25220
+    },
+    {
+      "epoch": 4.07889418802037,
+      "grad_norm": 0.8811455368995667,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 25230
+    },
+    {
+      "epoch": 4.080510872201115,
+      "grad_norm": 1.0564165115356445,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 25240
+    },
+    {
+      "epoch": 4.08212755638186,
+      "grad_norm": 1.1008027791976929,
+      "learning_rate": 0.0002,
+      "loss": 0.5472,
+      "step": 25250
+    },
+    {
+      "epoch": 4.083744240562606,
+      "grad_norm": 1.150097131729126,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 25260
+    },
+    {
+      "epoch": 4.085360924743352,
+      "grad_norm": 0.9339924454689026,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 25270
+    },
+    {
+      "epoch": 4.086977608924097,
+      "grad_norm": 1.0902045965194702,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 25280
+    },
+    {
+      "epoch": 4.088594293104842,
+      "grad_norm": 0.8483911156654358,
+      "learning_rate": 0.0002,
+      "loss": 0.5203,
+      "step": 25290
+    },
+    {
+      "epoch": 4.0902109772855875,
+      "grad_norm": 0.9477024674415588,
+      "learning_rate": 0.0002,
+      "loss": 0.5697,
+      "step": 25300
+    },
+    {
+      "epoch": 4.091827661466333,
+      "grad_norm": 0.9500215649604797,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 25310
+    },
+    {
+      "epoch": 4.093444345647078,
+      "grad_norm": 1.040468454360962,
+      "learning_rate": 0.0002,
+      "loss": 0.5045,
+      "step": 25320
+    },
+    {
+      "epoch": 4.095061029827823,
+      "grad_norm": 0.7457592487335205,
+      "learning_rate": 0.0002,
+      "loss": 0.5488,
+      "step": 25330
+    },
+    {
+      "epoch": 4.096677714008568,
+      "grad_norm": 1.2092097997665405,
+      "learning_rate": 0.0002,
+      "loss": 0.609,
+      "step": 25340
+    },
+    {
+      "epoch": 4.098294398189314,
+      "grad_norm": 0.9652107954025269,
+      "learning_rate": 0.0002,
+      "loss": 0.5174,
+      "step": 25350
+    },
+    {
+      "epoch": 4.099911082370059,
+      "grad_norm": 0.8464955687522888,
+      "learning_rate": 0.0002,
+      "loss": 0.5559,
+      "step": 25360
+    },
+    {
+      "epoch": 4.101527766550804,
+      "grad_norm": 0.875026285648346,
+      "learning_rate": 0.0002,
+      "loss": 0.5635,
+      "step": 25370
+    },
+    {
+      "epoch": 4.103144450731549,
+      "grad_norm": 0.9241740107536316,
+      "learning_rate": 0.0002,
+      "loss": 0.5774,
+      "step": 25380
+    },
+    {
+      "epoch": 4.1047611349122946,
+      "grad_norm": 0.9769546389579773,
+      "learning_rate": 0.0002,
+      "loss": 0.5578,
+      "step": 25390
+    },
+    {
+      "epoch": 4.10637781909304,
+      "grad_norm": 1.1501960754394531,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 25400
+    },
+    {
+      "epoch": 4.107994503273786,
+      "grad_norm": 0.9135243892669678,
+      "learning_rate": 0.0002,
+      "loss": 0.5241,
+      "step": 25410
+    },
+    {
+      "epoch": 4.109611187454531,
+      "grad_norm": 0.9905396103858948,
+      "learning_rate": 0.0002,
+      "loss": 0.5152,
+      "step": 25420
+    },
+    {
+      "epoch": 4.111227871635276,
+      "grad_norm": 0.9845104217529297,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 25430
+    },
+    {
+      "epoch": 4.112844555816022,
+      "grad_norm": 0.8326883912086487,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 25440
+    },
+    {
+      "epoch": 4.114461239996767,
+      "grad_norm": 0.9264556765556335,
+      "learning_rate": 0.0002,
+      "loss": 0.5312,
+      "step": 25450
+    },
+    {
+      "epoch": 4.116077924177512,
+      "grad_norm": 1.043080449104309,
+      "learning_rate": 0.0002,
+      "loss": 0.5968,
+      "step": 25460
+    },
+    {
+      "epoch": 4.117694608358257,
+      "grad_norm": 0.8533386588096619,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 25470
+    },
+    {
+      "epoch": 4.1193112925390025,
+      "grad_norm": 1.0133965015411377,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 25480
+    },
+    {
+      "epoch": 4.120927976719748,
+      "grad_norm": 0.7476310133934021,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 25490
+    },
+    {
+      "epoch": 4.122544660900493,
+      "grad_norm": 1.1247259378433228,
+      "learning_rate": 0.0002,
+      "loss": 0.5189,
+      "step": 25500
+    },
+    {
+      "epoch": 4.124161345081238,
+      "grad_norm": 1.0764678716659546,
+      "learning_rate": 0.0002,
+      "loss": 0.5751,
+      "step": 25510
+    },
+    {
+      "epoch": 4.1257780292619834,
+      "grad_norm": 0.7679798007011414,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 25520
+    },
+    {
+      "epoch": 4.127394713442729,
+      "grad_norm": 0.8877071142196655,
+      "learning_rate": 0.0002,
+      "loss": 0.5233,
+      "step": 25530
+    },
+    {
+      "epoch": 4.129011397623474,
+      "grad_norm": 1.0440239906311035,
+      "learning_rate": 0.0002,
+      "loss": 0.5769,
+      "step": 25540
+    },
+    {
+      "epoch": 4.130628081804219,
+      "grad_norm": 0.984145998954773,
+      "learning_rate": 0.0002,
+      "loss": 0.5723,
+      "step": 25550
+    },
+    {
+      "epoch": 4.132244765984965,
+      "grad_norm": 0.8667055368423462,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 25560
+    },
+    {
+      "epoch": 4.1338614501657105,
+      "grad_norm": 1.1300835609436035,
+      "learning_rate": 0.0002,
+      "loss": 0.5816,
+      "step": 25570
+    },
+    {
+      "epoch": 4.135478134346456,
+      "grad_norm": 0.9314348101615906,
+      "learning_rate": 0.0002,
+      "loss": 0.524,
+      "step": 25580
+    },
+    {
+      "epoch": 4.137094818527201,
+      "grad_norm": 0.7731879949569702,
+      "learning_rate": 0.0002,
+      "loss": 0.5283,
+      "step": 25590
+    },
+    {
+      "epoch": 4.138711502707946,
+      "grad_norm": 1.0080097913742065,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 25600
+    },
+    {
+      "epoch": 4.140328186888691,
+      "grad_norm": 1.2475038766860962,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 25610
+    },
+    {
+      "epoch": 4.141944871069437,
+      "grad_norm": 0.9912930727005005,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 25620
+    },
+    {
+      "epoch": 4.143561555250182,
+      "grad_norm": 0.9088651537895203,
+      "learning_rate": 0.0002,
+      "loss": 0.5624,
+      "step": 25630
+    },
+    {
+      "epoch": 4.145178239430927,
+      "grad_norm": 0.8940697312355042,
+      "learning_rate": 0.0002,
+      "loss": 0.5393,
+      "step": 25640
+    },
+    {
+      "epoch": 4.146794923611672,
+      "grad_norm": 1.0798203945159912,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 25650
+    },
+    {
+      "epoch": 4.148411607792418,
+      "grad_norm": 0.955172061920166,
+      "learning_rate": 0.0002,
+      "loss": 0.5987,
+      "step": 25660
+    },
+    {
+      "epoch": 4.150028291973163,
+      "grad_norm": 0.9692716002464294,
+      "learning_rate": 0.0002,
+      "loss": 0.569,
+      "step": 25670
+    },
+    {
+      "epoch": 4.151644976153908,
+      "grad_norm": 1.0813939571380615,
+      "learning_rate": 0.0002,
+      "loss": 0.5478,
+      "step": 25680
+    },
+    {
+      "epoch": 4.153261660334653,
+      "grad_norm": 1.135675072669983,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 25690
+    },
+    {
+      "epoch": 4.1548783445153985,
+      "grad_norm": 1.0392236709594727,
+      "learning_rate": 0.0002,
+      "loss": 0.5247,
+      "step": 25700
+    },
+    {
+      "epoch": 4.156495028696145,
+      "grad_norm": 0.9473116993904114,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 25710
+    },
+    {
+      "epoch": 4.15811171287689,
+      "grad_norm": 0.712493896484375,
+      "learning_rate": 0.0002,
+      "loss": 0.5339,
+      "step": 25720
+    },
+    {
+      "epoch": 4.159728397057635,
+      "grad_norm": 0.8724465370178223,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 25730
+    },
+    {
+      "epoch": 4.16134508123838,
+      "grad_norm": 0.9870015978813171,
+      "learning_rate": 0.0002,
+      "loss": 0.5325,
+      "step": 25740
+    },
+    {
+      "epoch": 4.1629617654191255,
+      "grad_norm": 1.025273084640503,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 25750
+    },
+    {
+      "epoch": 4.164578449599871,
+      "grad_norm": 0.9243090152740479,
+      "learning_rate": 0.0002,
+      "loss": 0.5223,
+      "step": 25760
+    },
+    {
+      "epoch": 4.166195133780616,
+      "grad_norm": 1.1656451225280762,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 25770
+    },
+    {
+      "epoch": 4.167811817961361,
+      "grad_norm": 0.936358630657196,
+      "learning_rate": 0.0002,
+      "loss": 0.5334,
+      "step": 25780
+    },
+    {
+      "epoch": 4.1694285021421065,
+      "grad_norm": 0.8618208169937134,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 25790
+    },
+    {
+      "epoch": 4.171045186322852,
+      "grad_norm": 0.8580600023269653,
+      "learning_rate": 0.0002,
+      "loss": 0.5186,
+      "step": 25800
+    },
+    {
+      "epoch": 4.172661870503597,
+      "grad_norm": 1.0128562450408936,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 25810
+    },
+    {
+      "epoch": 4.174278554684342,
+      "grad_norm": 0.854865312576294,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 25820
+    },
+    {
+      "epoch": 4.175895238865087,
+      "grad_norm": 1.235082745552063,
+      "learning_rate": 0.0002,
+      "loss": 0.5377,
+      "step": 25830
+    },
+    {
+      "epoch": 4.177511923045833,
+      "grad_norm": 0.9796220660209656,
+      "learning_rate": 0.0002,
+      "loss": 0.5614,
+      "step": 25840
+    },
+    {
+      "epoch": 4.179128607226578,
+      "grad_norm": 0.8922094702720642,
+      "learning_rate": 0.0002,
+      "loss": 0.5689,
+      "step": 25850
+    },
+    {
+      "epoch": 4.180745291407324,
+      "grad_norm": 0.9672530293464661,
+      "learning_rate": 0.0002,
+      "loss": 0.5806,
+      "step": 25860
+    },
+    {
+      "epoch": 4.182361975588069,
+      "grad_norm": 0.8662548661231995,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 25870
+    },
+    {
+      "epoch": 4.1839786597688144,
+      "grad_norm": 0.7938798069953918,
+      "learning_rate": 0.0002,
+      "loss": 0.5329,
+      "step": 25880
+    },
+    {
+      "epoch": 4.18559534394956,
+      "grad_norm": 1.0517958402633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5427,
+      "step": 25890
+    },
+    {
+      "epoch": 4.187212028130305,
+      "grad_norm": 0.8939275145530701,
+      "learning_rate": 0.0002,
+      "loss": 0.5147,
+      "step": 25900
+    },
+    {
+      "epoch": 4.18882871231105,
+      "grad_norm": 1.0296672582626343,
+      "learning_rate": 0.0002,
+      "loss": 0.5199,
+      "step": 25910
+    },
+    {
+      "epoch": 4.190445396491795,
+      "grad_norm": 0.8104017972946167,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 25920
+    },
+    {
+      "epoch": 4.192062080672541,
+      "grad_norm": 0.9984509944915771,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 25930
+    },
+    {
+      "epoch": 4.193678764853286,
+      "grad_norm": 0.9844784736633301,
+      "learning_rate": 0.0002,
+      "loss": 0.5356,
+      "step": 25940
+    },
+    {
+      "epoch": 4.195295449034031,
+      "grad_norm": 0.8168622255325317,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 25950
+    },
+    {
+      "epoch": 4.196912133214776,
+      "grad_norm": 1.0878913402557373,
+      "learning_rate": 0.0002,
+      "loss": 0.542,
+      "step": 25960
+    },
+    {
+      "epoch": 4.1985288173955215,
+      "grad_norm": 0.927126407623291,
+      "learning_rate": 0.0002,
+      "loss": 0.5414,
+      "step": 25970
+    },
+    {
+      "epoch": 4.200145501576267,
+      "grad_norm": 0.838586688041687,
+      "learning_rate": 0.0002,
+      "loss": 0.5794,
+      "step": 25980
+    },
+    {
+      "epoch": 4.201762185757012,
+      "grad_norm": 1.2572145462036133,
+      "learning_rate": 0.0002,
+      "loss": 0.5454,
+      "step": 25990
+    },
+    {
+      "epoch": 4.203378869937758,
+      "grad_norm": 1.0476740598678589,
+      "learning_rate": 0.0002,
+      "loss": 0.5048,
+      "step": 26000
+    },
+    {
+      "epoch": 4.204995554118503,
+      "grad_norm": 1.0873368978500366,
+      "learning_rate": 0.0002,
+      "loss": 0.5127,
+      "step": 26010
+    },
+    {
+      "epoch": 4.206612238299249,
+      "grad_norm": 1.2664896249771118,
+      "learning_rate": 0.0002,
+      "loss": 0.5679,
+      "step": 26020
+    },
+    {
+      "epoch": 4.208228922479994,
+      "grad_norm": 1.0312391519546509,
+      "learning_rate": 0.0002,
+      "loss": 0.5814,
+      "step": 26030
+    },
+    {
+      "epoch": 4.209845606660739,
+      "grad_norm": 1.0235042572021484,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 26040
+    },
+    {
+      "epoch": 4.211462290841484,
+      "grad_norm": 0.8882219195365906,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 26050
+    },
+    {
+      "epoch": 4.2130789750222295,
+      "grad_norm": 0.9115961790084839,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 26060
+    },
+    {
+      "epoch": 4.214695659202975,
+      "grad_norm": 1.0218228101730347,
+      "learning_rate": 0.0002,
+      "loss": 0.5455,
+      "step": 26070
+    },
+    {
+      "epoch": 4.21631234338372,
+      "grad_norm": 1.0802232027053833,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 26080
+    },
+    {
+      "epoch": 4.217929027564465,
+      "grad_norm": 1.1488053798675537,
+      "learning_rate": 0.0002,
+      "loss": 0.557,
+      "step": 26090
+    },
+    {
+      "epoch": 4.21954571174521,
+      "grad_norm": 1.0487725734710693,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 26100
+    },
+    {
+      "epoch": 4.221162395925956,
+      "grad_norm": 0.9131165742874146,
+      "learning_rate": 0.0002,
+      "loss": 0.5568,
+      "step": 26110
+    },
+    {
+      "epoch": 4.222779080106701,
+      "grad_norm": 0.9012845158576965,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 26120
+    },
+    {
+      "epoch": 4.224395764287446,
+      "grad_norm": 0.8389840126037598,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 26130
+    },
+    {
+      "epoch": 4.226012448468191,
+      "grad_norm": 0.8924660682678223,
+      "learning_rate": 0.0002,
+      "loss": 0.5268,
+      "step": 26140
+    },
+    {
+      "epoch": 4.2276291326489375,
+      "grad_norm": 0.8556463718414307,
+      "learning_rate": 0.0002,
+      "loss": 0.5715,
+      "step": 26150
+    },
+    {
+      "epoch": 4.229245816829683,
+      "grad_norm": 0.9643129110336304,
+      "learning_rate": 0.0002,
+      "loss": 0.5695,
+      "step": 26160
+    },
+    {
+      "epoch": 4.230862501010428,
+      "grad_norm": 0.9865712523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 26170
+    },
+    {
+      "epoch": 4.232479185191173,
+      "grad_norm": 1.152641773223877,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 26180
+    },
+    {
+      "epoch": 4.234095869371918,
+      "grad_norm": 0.9157698154449463,
+      "learning_rate": 0.0002,
+      "loss": 0.5632,
+      "step": 26190
+    },
+    {
+      "epoch": 4.235712553552664,
+      "grad_norm": 0.8418048620223999,
+      "learning_rate": 0.0002,
+      "loss": 0.5717,
+      "step": 26200
+    },
+    {
+      "epoch": 4.237329237733409,
+      "grad_norm": 0.9430168867111206,
+      "learning_rate": 0.0002,
+      "loss": 0.5624,
+      "step": 26210
+    },
+    {
+      "epoch": 4.238945921914154,
+      "grad_norm": 1.012582778930664,
+      "learning_rate": 0.0002,
+      "loss": 0.5574,
+      "step": 26220
+    },
+    {
+      "epoch": 4.240562606094899,
+      "grad_norm": 1.112619400024414,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 26230
+    },
+    {
+      "epoch": 4.2421792902756446,
+      "grad_norm": 0.9243621826171875,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 26240
+    },
+    {
+      "epoch": 4.24379597445639,
+      "grad_norm": 0.6977595686912537,
+      "learning_rate": 0.0002,
+      "loss": 0.569,
+      "step": 26250
+    },
+    {
+      "epoch": 4.245412658637135,
+      "grad_norm": 0.9600721597671509,
+      "learning_rate": 0.0002,
+      "loss": 0.5379,
+      "step": 26260
+    },
+    {
+      "epoch": 4.24702934281788,
+      "grad_norm": 0.882641613483429,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 26270
+    },
+    {
+      "epoch": 4.2486460269986255,
+      "grad_norm": 1.010920763015747,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 26280
+    },
+    {
+      "epoch": 4.250262711179371,
+      "grad_norm": 0.9289400577545166,
+      "learning_rate": 0.0002,
+      "loss": 0.5803,
+      "step": 26290
+    },
+    {
+      "epoch": 4.251879395360117,
+      "grad_norm": 1.137397289276123,
+      "learning_rate": 0.0002,
+      "loss": 0.541,
+      "step": 26300
+    },
+    {
+      "epoch": 4.253496079540862,
+      "grad_norm": 1.0136182308197021,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 26310
+    },
+    {
+      "epoch": 4.255112763721607,
+      "grad_norm": 0.9387356042861938,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 26320
+    },
+    {
+      "epoch": 4.2567294479023525,
+      "grad_norm": 1.1833957433700562,
+      "learning_rate": 0.0002,
+      "loss": 0.5948,
+      "step": 26330
+    },
+    {
+      "epoch": 4.258346132083098,
+      "grad_norm": 0.9415934681892395,
+      "learning_rate": 0.0002,
+      "loss": 0.5905,
+      "step": 26340
+    },
+    {
+      "epoch": 4.259962816263843,
+      "grad_norm": 0.8550165891647339,
+      "learning_rate": 0.0002,
+      "loss": 0.5539,
+      "step": 26350
+    },
+    {
+      "epoch": 4.261579500444588,
+      "grad_norm": 9.924622535705566,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 26360
+    },
+    {
+      "epoch": 4.2631961846253335,
+      "grad_norm": 1.0104902982711792,
+      "learning_rate": 0.0002,
+      "loss": 0.5689,
+      "step": 26370
+    },
+    {
+      "epoch": 4.264812868806079,
+      "grad_norm": 0.890794038772583,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 26380
+    },
+    {
+      "epoch": 4.266429552986824,
+      "grad_norm": 1.0560191869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.563,
+      "step": 26390
+    },
+    {
+      "epoch": 4.268046237167569,
+      "grad_norm": 1.0135581493377686,
+      "learning_rate": 0.0002,
+      "loss": 0.5119,
+      "step": 26400
+    },
+    {
+      "epoch": 4.269662921348314,
+      "grad_norm": 1.1304140090942383,
+      "learning_rate": 0.0002,
+      "loss": 0.5359,
+      "step": 26410
+    },
+    {
+      "epoch": 4.27127960552906,
+      "grad_norm": 0.9899303913116455,
+      "learning_rate": 0.0002,
+      "loss": 0.5615,
+      "step": 26420
+    },
+    {
+      "epoch": 4.272896289709805,
+      "grad_norm": 1.0505329370498657,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 26430
+    },
+    {
+      "epoch": 4.27451297389055,
+      "grad_norm": 0.9389396905899048,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 26440
+    },
+    {
+      "epoch": 4.276129658071296,
+      "grad_norm": 0.875328779220581,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 26450
+    },
+    {
+      "epoch": 4.277746342252041,
+      "grad_norm": 1.0689256191253662,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 26460
+    },
+    {
+      "epoch": 4.279363026432787,
+      "grad_norm": 0.9988957643508911,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 26470
+    },
+    {
+      "epoch": 4.280979710613532,
+      "grad_norm": 0.8721813559532166,
+      "learning_rate": 0.0002,
+      "loss": 0.5478,
+      "step": 26480
+    },
+    {
+      "epoch": 4.282596394794277,
+      "grad_norm": 1.100109577178955,
+      "learning_rate": 0.0002,
+      "loss": 0.5424,
+      "step": 26490
+    },
+    {
+      "epoch": 4.284213078975022,
+      "grad_norm": 1.1607271432876587,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 26500
+    },
+    {
+      "epoch": 4.285829763155768,
+      "grad_norm": 0.879088819026947,
+      "learning_rate": 0.0002,
+      "loss": 0.6287,
+      "step": 26510
+    },
+    {
+      "epoch": 4.287446447336513,
+      "grad_norm": 0.9891700744628906,
+      "learning_rate": 0.0002,
+      "loss": 0.573,
+      "step": 26520
+    },
+    {
+      "epoch": 4.289063131517258,
+      "grad_norm": 1.0831127166748047,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 26530
+    },
+    {
+      "epoch": 4.290679815698003,
+      "grad_norm": 1.4108285903930664,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 26540
+    },
+    {
+      "epoch": 4.2922964998787485,
+      "grad_norm": 1.0630289316177368,
+      "learning_rate": 0.0002,
+      "loss": 0.5888,
+      "step": 26550
+    },
+    {
+      "epoch": 4.293913184059494,
+      "grad_norm": 1.0854572057724,
+      "learning_rate": 0.0002,
+      "loss": 0.5817,
+      "step": 26560
+    },
+    {
+      "epoch": 4.295529868240239,
+      "grad_norm": 0.9561646580696106,
+      "learning_rate": 0.0002,
+      "loss": 0.5586,
+      "step": 26570
+    },
+    {
+      "epoch": 4.297146552420984,
+      "grad_norm": 0.9064981937408447,
+      "learning_rate": 0.0002,
+      "loss": 0.5674,
+      "step": 26580
+    },
+    {
+      "epoch": 4.298763236601729,
+      "grad_norm": 1.0082972049713135,
+      "learning_rate": 0.0002,
+      "loss": 0.5847,
+      "step": 26590
+    },
+    {
+      "epoch": 4.3003799207824756,
+      "grad_norm": 1.1613214015960693,
+      "learning_rate": 0.0002,
+      "loss": 0.5711,
+      "step": 26600
+    },
+    {
+      "epoch": 4.301996604963221,
+      "grad_norm": 0.9847695231437683,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 26610
+    },
+    {
+      "epoch": 4.303613289143966,
+      "grad_norm": 1.0980697870254517,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 26620
+    },
+    {
+      "epoch": 4.305229973324711,
+      "grad_norm": 0.8861175179481506,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 26630
+    },
+    {
+      "epoch": 4.3068466575054565,
+      "grad_norm": 0.8917363286018372,
+      "learning_rate": 0.0002,
+      "loss": 0.5716,
+      "step": 26640
+    },
+    {
+      "epoch": 4.308463341686202,
+      "grad_norm": 1.0458378791809082,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 26650
+    },
+    {
+      "epoch": 4.310080025866947,
+      "grad_norm": 1.4859240055084229,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 26660
+    },
+    {
+      "epoch": 4.311696710047692,
+      "grad_norm": 1.1376359462738037,
+      "learning_rate": 0.0002,
+      "loss": 0.5296,
+      "step": 26670
+    },
+    {
+      "epoch": 4.313313394228437,
+      "grad_norm": 0.991349995136261,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 26680
+    },
+    {
+      "epoch": 4.314930078409183,
+      "grad_norm": 0.9995543956756592,
+      "learning_rate": 0.0002,
+      "loss": 0.5338,
+      "step": 26690
+    },
+    {
+      "epoch": 4.316546762589928,
+      "grad_norm": 1.0515851974487305,
+      "learning_rate": 0.0002,
+      "loss": 0.5542,
+      "step": 26700
+    },
+    {
+      "epoch": 4.318163446770673,
+      "grad_norm": 1.008023977279663,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 26710
+    },
+    {
+      "epoch": 4.319780130951418,
+      "grad_norm": 1.0184582471847534,
+      "learning_rate": 0.0002,
+      "loss": 0.5506,
+      "step": 26720
+    },
+    {
+      "epoch": 4.321396815132164,
+      "grad_norm": 1.161071538925171,
+      "learning_rate": 0.0002,
+      "loss": 0.5828,
+      "step": 26730
+    },
+    {
+      "epoch": 4.323013499312909,
+      "grad_norm": 0.9580779671669006,
+      "learning_rate": 0.0002,
+      "loss": 0.5633,
+      "step": 26740
+    },
+    {
+      "epoch": 4.324630183493655,
+      "grad_norm": 1.0189911127090454,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 26750
+    },
+    {
+      "epoch": 4.3262468676744,
+      "grad_norm": 0.7484358549118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 26760
+    },
+    {
+      "epoch": 4.327863551855145,
+      "grad_norm": 1.0015908479690552,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 26770
+    },
+    {
+      "epoch": 4.329480236035891,
+      "grad_norm": 0.8972945809364319,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 26780
+    },
+    {
+      "epoch": 4.331096920216636,
+      "grad_norm": 1.01099693775177,
+      "learning_rate": 0.0002,
+      "loss": 0.5857,
+      "step": 26790
+    },
+    {
+      "epoch": 4.332713604397381,
+      "grad_norm": 0.846958339214325,
+      "learning_rate": 0.0002,
+      "loss": 0.5591,
+      "step": 26800
+    },
+    {
+      "epoch": 4.334330288578126,
+      "grad_norm": 1.0792603492736816,
+      "learning_rate": 0.0002,
+      "loss": 0.5547,
+      "step": 26810
+    },
+    {
+      "epoch": 4.3359469727588715,
+      "grad_norm": 1.0373345613479614,
+      "learning_rate": 0.0002,
+      "loss": 0.5747,
+      "step": 26820
+    },
+    {
+      "epoch": 4.337563656939617,
+      "grad_norm": 0.9779167771339417,
+      "learning_rate": 0.0002,
+      "loss": 0.558,
+      "step": 26830
+    },
+    {
+      "epoch": 4.339180341120362,
+      "grad_norm": 1.0235520601272583,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 26840
+    },
+    {
+      "epoch": 4.340797025301107,
+      "grad_norm": 1.04195237159729,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 26850
+    },
+    {
+      "epoch": 4.3424137094818525,
+      "grad_norm": 0.9479565620422363,
+      "learning_rate": 0.0002,
+      "loss": 0.5474,
+      "step": 26860
+    },
+    {
+      "epoch": 4.344030393662598,
+      "grad_norm": 0.9526172280311584,
+      "learning_rate": 0.0002,
+      "loss": 0.5646,
+      "step": 26870
+    },
+    {
+      "epoch": 4.345647077843343,
+      "grad_norm": 0.8571456074714661,
+      "learning_rate": 0.0002,
+      "loss": 0.521,
+      "step": 26880
+    },
+    {
+      "epoch": 4.347263762024088,
+      "grad_norm": 0.9475828409194946,
+      "learning_rate": 0.0002,
+      "loss": 0.5846,
+      "step": 26890
+    },
+    {
+      "epoch": 4.348880446204834,
+      "grad_norm": 1.0529576539993286,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 26900
+    },
+    {
+      "epoch": 4.3504971303855795,
+      "grad_norm": 0.9648140072822571,
+      "learning_rate": 0.0002,
+      "loss": 0.56,
+      "step": 26910
+    },
+    {
+      "epoch": 4.352113814566325,
+      "grad_norm": 1.0488841533660889,
+      "learning_rate": 0.0002,
+      "loss": 0.5162,
+      "step": 26920
+    },
+    {
+      "epoch": 4.35373049874707,
+      "grad_norm": 0.8771942257881165,
+      "learning_rate": 0.0002,
+      "loss": 0.5842,
+      "step": 26930
+    },
+    {
+      "epoch": 4.355347182927815,
+      "grad_norm": 0.9411202073097229,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 26940
+    },
+    {
+      "epoch": 4.35696386710856,
+      "grad_norm": 1.0997588634490967,
+      "learning_rate": 0.0002,
+      "loss": 0.6001,
+      "step": 26950
+    },
+    {
+      "epoch": 4.358580551289306,
+      "grad_norm": 0.968754768371582,
+      "learning_rate": 0.0002,
+      "loss": 0.5528,
+      "step": 26960
+    },
+    {
+      "epoch": 4.360197235470051,
+      "grad_norm": 0.9990773797035217,
+      "learning_rate": 0.0002,
+      "loss": 0.5881,
+      "step": 26970
+    },
+    {
+      "epoch": 4.361813919650796,
+      "grad_norm": 1.0210620164871216,
+      "learning_rate": 0.0002,
+      "loss": 0.5761,
+      "step": 26980
+    },
+    {
+      "epoch": 4.363430603831541,
+      "grad_norm": 0.855462908744812,
+      "learning_rate": 0.0002,
+      "loss": 0.5768,
+      "step": 26990
+    },
+    {
+      "epoch": 4.365047288012287,
+      "grad_norm": 0.9169660806655884,
+      "learning_rate": 0.0002,
+      "loss": 0.5493,
+      "step": 27000
+    },
+    {
+      "epoch": 4.366663972193032,
+      "grad_norm": 1.089629888534546,
+      "learning_rate": 0.0002,
+      "loss": 0.5697,
+      "step": 27010
+    },
+    {
+      "epoch": 4.368280656373777,
+      "grad_norm": 1.0932867527008057,
+      "learning_rate": 0.0002,
+      "loss": 0.5854,
+      "step": 27020
+    },
+    {
+      "epoch": 4.369897340554522,
+      "grad_norm": 0.9290956854820251,
+      "learning_rate": 0.0002,
+      "loss": 0.5656,
+      "step": 27030
+    },
+    {
+      "epoch": 4.3715140247352675,
+      "grad_norm": 1.2800624370574951,
+      "learning_rate": 0.0002,
+      "loss": 0.5727,
+      "step": 27040
+    },
+    {
+      "epoch": 4.373130708916014,
+      "grad_norm": 0.8993493318557739,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 27050
+    },
+    {
+      "epoch": 4.374747393096759,
+      "grad_norm": 1.1566431522369385,
+      "learning_rate": 0.0002,
+      "loss": 0.6232,
+      "step": 27060
+    },
+    {
+      "epoch": 4.376364077277504,
+      "grad_norm": 0.9479052424430847,
+      "learning_rate": 0.0002,
+      "loss": 0.5902,
+      "step": 27070
+    },
+    {
+      "epoch": 4.377980761458249,
+      "grad_norm": 1.0063648223876953,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 27080
+    },
+    {
+      "epoch": 4.379597445638995,
+      "grad_norm": 0.8342045545578003,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 27090
+    },
+    {
+      "epoch": 4.38121412981974,
+      "grad_norm": 1.1390739679336548,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 27100
+    },
+    {
+      "epoch": 4.382830814000485,
+      "grad_norm": 0.9547637104988098,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 27110
+    },
+    {
+      "epoch": 4.38444749818123,
+      "grad_norm": 1.0503804683685303,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 27120
+    },
+    {
+      "epoch": 4.3860641823619755,
+      "grad_norm": 0.9064017534255981,
+      "learning_rate": 0.0002,
+      "loss": 0.5787,
+      "step": 27130
+    },
+    {
+      "epoch": 4.387680866542721,
+      "grad_norm": 0.9382519125938416,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 27140
+    },
+    {
+      "epoch": 4.389297550723466,
+      "grad_norm": 1.0410341024398804,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 27150
+    },
+    {
+      "epoch": 4.390914234904211,
+      "grad_norm": 0.9218655824661255,
+      "learning_rate": 0.0002,
+      "loss": 0.6034,
+      "step": 27160
+    },
+    {
+      "epoch": 4.392530919084956,
+      "grad_norm": 0.8119737505912781,
+      "learning_rate": 0.0002,
+      "loss": 0.5204,
+      "step": 27170
+    },
+    {
+      "epoch": 4.394147603265702,
+      "grad_norm": 0.8584722876548767,
+      "learning_rate": 0.0002,
+      "loss": 0.5612,
+      "step": 27180
+    },
+    {
+      "epoch": 4.395764287446447,
+      "grad_norm": 0.9668293595314026,
+      "learning_rate": 0.0002,
+      "loss": 0.5772,
+      "step": 27190
+    },
+    {
+      "epoch": 4.397380971627193,
+      "grad_norm": 1.022334098815918,
+      "learning_rate": 0.0002,
+      "loss": 0.6009,
+      "step": 27200
+    },
+    {
+      "epoch": 4.398997655807938,
+      "grad_norm": 0.9553216099739075,
+      "learning_rate": 0.0002,
+      "loss": 0.5573,
+      "step": 27210
+    },
+    {
+      "epoch": 4.4006143399886835,
+      "grad_norm": 0.9282339215278625,
+      "learning_rate": 0.0002,
+      "loss": 0.5604,
+      "step": 27220
+    },
+    {
+      "epoch": 4.402231024169429,
+      "grad_norm": 1.0232292413711548,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 27230
+    },
+    {
+      "epoch": 4.403847708350174,
+      "grad_norm": 0.9915700554847717,
+      "learning_rate": 0.0002,
+      "loss": 0.6078,
+      "step": 27240
+    },
+    {
+      "epoch": 4.405464392530919,
+      "grad_norm": 1.0014961957931519,
+      "learning_rate": 0.0002,
+      "loss": 0.5778,
+      "step": 27250
+    },
+    {
+      "epoch": 4.407081076711664,
+      "grad_norm": 1.1172103881835938,
+      "learning_rate": 0.0002,
+      "loss": 0.5824,
+      "step": 27260
+    },
+    {
+      "epoch": 4.40869776089241,
+      "grad_norm": 0.8583093285560608,
+      "learning_rate": 0.0002,
+      "loss": 0.5286,
+      "step": 27270
+    },
+    {
+      "epoch": 4.410314445073155,
+      "grad_norm": 0.7609201669692993,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 27280
+    },
+    {
+      "epoch": 4.4119311292539,
+      "grad_norm": 1.0619351863861084,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 27290
+    },
+    {
+      "epoch": 4.413547813434645,
+      "grad_norm": 1.0177674293518066,
+      "learning_rate": 0.0002,
+      "loss": 0.5579,
+      "step": 27300
+    },
+    {
+      "epoch": 4.4151644976153905,
+      "grad_norm": 0.9921218156814575,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 27310
+    },
+    {
+      "epoch": 4.416781181796136,
+      "grad_norm": 1.126244306564331,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 27320
+    },
+    {
+      "epoch": 4.418397865976881,
+      "grad_norm": 1.0678540468215942,
+      "learning_rate": 0.0002,
+      "loss": 0.5743,
+      "step": 27330
+    },
+    {
+      "epoch": 4.420014550157627,
+      "grad_norm": 0.8705704212188721,
+      "learning_rate": 0.0002,
+      "loss": 0.5665,
+      "step": 27340
+    },
+    {
+      "epoch": 4.421631234338372,
+      "grad_norm": 1.272074818611145,
+      "learning_rate": 0.0002,
+      "loss": 0.5763,
+      "step": 27350
+    },
+    {
+      "epoch": 4.423247918519118,
+      "grad_norm": 0.8740444183349609,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 27360
+    },
+    {
+      "epoch": 4.424864602699863,
+      "grad_norm": 1.0584250688552856,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 27370
+    },
+    {
+      "epoch": 4.426481286880608,
+      "grad_norm": 1.059870719909668,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 27380
+    },
+    {
+      "epoch": 4.428097971061353,
+      "grad_norm": 1.072265863418579,
+      "learning_rate": 0.0002,
+      "loss": 0.5551,
+      "step": 27390
+    },
+    {
+      "epoch": 4.4297146552420985,
+      "grad_norm": 0.871481716632843,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 27400
+    },
+    {
+      "epoch": 4.431331339422844,
+      "grad_norm": 0.9555448293685913,
+      "learning_rate": 0.0002,
+      "loss": 0.5372,
+      "step": 27410
+    },
+    {
+      "epoch": 4.432948023603589,
+      "grad_norm": 1.0402292013168335,
+      "learning_rate": 0.0002,
+      "loss": 0.5593,
+      "step": 27420
+    },
+    {
+      "epoch": 4.434564707784334,
+      "grad_norm": 1.12587571144104,
+      "learning_rate": 0.0002,
+      "loss": 0.5532,
+      "step": 27430
+    },
+    {
+      "epoch": 4.436181391965079,
+      "grad_norm": 1.0783193111419678,
+      "learning_rate": 0.0002,
+      "loss": 0.5403,
+      "step": 27440
+    },
+    {
+      "epoch": 4.437798076145825,
+      "grad_norm": 1.024133563041687,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 27450
+    },
+    {
+      "epoch": 4.43941476032657,
+      "grad_norm": 0.9156768918037415,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 27460
+    },
+    {
+      "epoch": 4.441031444507315,
+      "grad_norm": 1.0215224027633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 27470
+    },
+    {
+      "epoch": 4.442648128688061,
+      "grad_norm": 1.082116961479187,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 27480
+    },
+    {
+      "epoch": 4.4442648128688065,
+      "grad_norm": 1.0412873029708862,
+      "learning_rate": 0.0002,
+      "loss": 0.6203,
+      "step": 27490
+    },
+    {
+      "epoch": 4.445881497049552,
+      "grad_norm": 1.0509289503097534,
+      "learning_rate": 0.0002,
+      "loss": 0.5939,
+      "step": 27500
+    },
+    {
+      "epoch": 4.447498181230297,
+      "grad_norm": 0.9291498064994812,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 27510
+    },
+    {
+      "epoch": 4.449114865411042,
+      "grad_norm": 0.970184326171875,
+      "learning_rate": 0.0002,
+      "loss": 0.5408,
+      "step": 27520
+    },
+    {
+      "epoch": 4.450731549591787,
+      "grad_norm": 0.8418883681297302,
+      "learning_rate": 0.0002,
+      "loss": 0.5705,
+      "step": 27530
+    },
+    {
+      "epoch": 4.452348233772533,
+      "grad_norm": 0.8823825120925903,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 27540
+    },
+    {
+      "epoch": 4.453964917953278,
+      "grad_norm": 1.1909019947052002,
+      "learning_rate": 0.0002,
+      "loss": 0.5867,
+      "step": 27550
+    },
+    {
+      "epoch": 4.455581602134023,
+      "grad_norm": 1.0317302942276,
+      "learning_rate": 0.0002,
+      "loss": 0.5685,
+      "step": 27560
+    },
+    {
+      "epoch": 4.457198286314768,
+      "grad_norm": 0.9977751970291138,
+      "learning_rate": 0.0002,
+      "loss": 0.5538,
+      "step": 27570
+    },
+    {
+      "epoch": 4.458814970495514,
+      "grad_norm": 0.8909519910812378,
+      "learning_rate": 0.0002,
+      "loss": 0.5628,
+      "step": 27580
+    },
+    {
+      "epoch": 4.460431654676259,
+      "grad_norm": 0.8653029799461365,
+      "learning_rate": 0.0002,
+      "loss": 0.6099,
+      "step": 27590
+    },
+    {
+      "epoch": 4.462048338857004,
+      "grad_norm": 1.0783653259277344,
+      "learning_rate": 0.0002,
+      "loss": 0.5622,
+      "step": 27600
+    },
+    {
+      "epoch": 4.463665023037749,
+      "grad_norm": 1.1235394477844238,
+      "learning_rate": 0.0002,
+      "loss": 0.579,
+      "step": 27610
+    },
+    {
+      "epoch": 4.4652817072184945,
+      "grad_norm": 0.9386643767356873,
+      "learning_rate": 0.0002,
+      "loss": 0.5545,
+      "step": 27620
+    },
+    {
+      "epoch": 4.466898391399241,
+      "grad_norm": 1.0605148077011108,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 27630
+    },
+    {
+      "epoch": 4.468515075579986,
+      "grad_norm": 1.1283893585205078,
+      "learning_rate": 0.0002,
+      "loss": 0.5886,
+      "step": 27640
+    },
+    {
+      "epoch": 4.470131759760731,
+      "grad_norm": 1.0583468675613403,
+      "learning_rate": 0.0002,
+      "loss": 0.5801,
+      "step": 27650
+    },
+    {
+      "epoch": 4.471748443941476,
+      "grad_norm": 0.9563992023468018,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 27660
+    },
+    {
+      "epoch": 4.4733651281222215,
+      "grad_norm": 1.100598931312561,
+      "learning_rate": 0.0002,
+      "loss": 0.5687,
+      "step": 27670
+    },
+    {
+      "epoch": 4.474981812302967,
+      "grad_norm": 0.9386957287788391,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 27680
+    },
+    {
+      "epoch": 4.476598496483712,
+      "grad_norm": 1.2946288585662842,
+      "learning_rate": 0.0002,
+      "loss": 0.6241,
+      "step": 27690
+    },
+    {
+      "epoch": 4.478215180664457,
+      "grad_norm": 1.0325199365615845,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 27700
+    },
+    {
+      "epoch": 4.4798318648452025,
+      "grad_norm": 1.0318928956985474,
+      "learning_rate": 0.0002,
+      "loss": 0.588,
+      "step": 27710
+    },
+    {
+      "epoch": 4.481448549025948,
+      "grad_norm": 0.8721024394035339,
+      "learning_rate": 0.0002,
+      "loss": 0.5656,
+      "step": 27720
+    },
+    {
+      "epoch": 4.483065233206693,
+      "grad_norm": 1.17376708984375,
+      "learning_rate": 0.0002,
+      "loss": 0.5421,
+      "step": 27730
+    },
+    {
+      "epoch": 4.484681917387438,
+      "grad_norm": 1.0926326513290405,
+      "learning_rate": 0.0002,
+      "loss": 0.5657,
+      "step": 27740
+    },
+    {
+      "epoch": 4.486298601568183,
+      "grad_norm": 0.9043852686882019,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 27750
+    },
+    {
+      "epoch": 4.487915285748929,
+      "grad_norm": 1.064600944519043,
+      "learning_rate": 0.0002,
+      "loss": 0.582,
+      "step": 27760
+    },
+    {
+      "epoch": 4.489531969929674,
+      "grad_norm": 0.7833460569381714,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 27770
+    },
+    {
+      "epoch": 4.49114865411042,
+      "grad_norm": 1.1073496341705322,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 27780
+    },
+    {
+      "epoch": 4.492765338291165,
+      "grad_norm": 1.0799397230148315,
+      "learning_rate": 0.0002,
+      "loss": 0.5577,
+      "step": 27790
+    },
+    {
+      "epoch": 4.49438202247191,
+      "grad_norm": 1.1062238216400146,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 27800
+    },
+    {
+      "epoch": 4.495998706652656,
+      "grad_norm": 1.0568242073059082,
+      "learning_rate": 0.0002,
+      "loss": 0.6126,
+      "step": 27810
+    },
+    {
+      "epoch": 4.497615390833401,
+      "grad_norm": 0.8861091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 27820
+    },
+    {
+      "epoch": 4.499232075014146,
+      "grad_norm": 1.2297543287277222,
+      "learning_rate": 0.0002,
+      "loss": 0.5858,
+      "step": 27830
+    },
+    {
+      "epoch": 4.500848759194891,
+      "grad_norm": 0.9600302577018738,
+      "learning_rate": 0.0002,
+      "loss": 0.5859,
+      "step": 27840
+    },
+    {
+      "epoch": 4.502465443375637,
+      "grad_norm": 1.057051181793213,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 27850
+    },
+    {
+      "epoch": 4.504082127556382,
+      "grad_norm": 0.9839690923690796,
+      "learning_rate": 0.0002,
+      "loss": 0.5788,
+      "step": 27860
+    },
+    {
+      "epoch": 4.505698811737127,
+      "grad_norm": 1.1479853391647339,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 27870
+    },
+    {
+      "epoch": 4.507315495917872,
+      "grad_norm": 1.0550768375396729,
+      "learning_rate": 0.0002,
+      "loss": 0.6039,
+      "step": 27880
+    },
+    {
+      "epoch": 4.5089321800986175,
+      "grad_norm": 0.898209273815155,
+      "learning_rate": 0.0002,
+      "loss": 0.563,
+      "step": 27890
+    },
+    {
+      "epoch": 4.510548864279363,
+      "grad_norm": 0.9460315108299255,
+      "learning_rate": 0.0002,
+      "loss": 0.5734,
+      "step": 27900
+    },
+    {
+      "epoch": 4.512165548460108,
+      "grad_norm": 0.9499884247779846,
+      "learning_rate": 0.0002,
+      "loss": 0.5702,
+      "step": 27910
+    },
+    {
+      "epoch": 4.513782232640853,
+      "grad_norm": 0.7801318764686584,
+      "learning_rate": 0.0002,
+      "loss": 0.5385,
+      "step": 27920
+    },
+    {
+      "epoch": 4.515398916821599,
+      "grad_norm": 0.9286966323852539,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 27930
+    },
+    {
+      "epoch": 4.517015601002345,
+      "grad_norm": 0.9539980292320251,
+      "learning_rate": 0.0002,
+      "loss": 0.5717,
+      "step": 27940
+    },
+    {
+      "epoch": 4.51863228518309,
+      "grad_norm": 1.1053401231765747,
+      "learning_rate": 0.0002,
+      "loss": 0.6073,
+      "step": 27950
+    },
+    {
+      "epoch": 4.520248969363835,
+      "grad_norm": 0.7535534501075745,
+      "learning_rate": 0.0002,
+      "loss": 0.6087,
+      "step": 27960
+    },
+    {
+      "epoch": 4.52186565354458,
+      "grad_norm": 1.076926589012146,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 27970
+    },
+    {
+      "epoch": 4.5234823377253255,
+      "grad_norm": 1.181935429573059,
+      "learning_rate": 0.0002,
+      "loss": 0.6028,
+      "step": 27980
+    },
+    {
+      "epoch": 4.525099021906071,
+      "grad_norm": 0.9293407201766968,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 27990
+    },
+    {
+      "epoch": 4.526715706086816,
+      "grad_norm": 0.8953009247779846,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 28000
+    },
+    {
+      "epoch": 4.528332390267561,
+      "grad_norm": 1.0850225687026978,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 28010
+    },
+    {
+      "epoch": 4.529949074448306,
+      "grad_norm": 0.9125663042068481,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 28020
+    },
+    {
+      "epoch": 4.531565758629052,
+      "grad_norm": 0.8745216727256775,
+      "learning_rate": 0.0002,
+      "loss": 0.5922,
+      "step": 28030
+    },
+    {
+      "epoch": 4.533182442809797,
+      "grad_norm": 1.0783463716506958,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 28040
+    },
+    {
+      "epoch": 4.534799126990542,
+      "grad_norm": 0.7513844966888428,
+      "learning_rate": 0.0002,
+      "loss": 0.5754,
+      "step": 28050
+    },
+    {
+      "epoch": 4.536415811171287,
+      "grad_norm": 1.0135776996612549,
+      "learning_rate": 0.0002,
+      "loss": 0.5608,
+      "step": 28060
+    },
+    {
+      "epoch": 4.538032495352033,
+      "grad_norm": 0.8886825442314148,
+      "learning_rate": 0.0002,
+      "loss": 0.5827,
+      "step": 28070
+    },
+    {
+      "epoch": 4.539649179532779,
+      "grad_norm": 0.8153995275497437,
+      "learning_rate": 0.0002,
+      "loss": 0.5605,
+      "step": 28080
+    },
+    {
+      "epoch": 4.541265863713524,
+      "grad_norm": 0.9853341579437256,
+      "learning_rate": 0.0002,
+      "loss": 0.6377,
+      "step": 28090
+    },
+    {
+      "epoch": 4.542882547894269,
+      "grad_norm": 0.9365800023078918,
+      "learning_rate": 0.0002,
+      "loss": 0.5957,
+      "step": 28100
+    },
+    {
+      "epoch": 4.544499232075014,
+      "grad_norm": 0.9765017628669739,
+      "learning_rate": 0.0002,
+      "loss": 0.5477,
+      "step": 28110
+    },
+    {
+      "epoch": 4.54611591625576,
+      "grad_norm": 0.9811279773712158,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 28120
+    },
+    {
+      "epoch": 4.547732600436505,
+      "grad_norm": 1.0387924909591675,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 28130
+    },
+    {
+      "epoch": 4.54934928461725,
+      "grad_norm": 1.0684878826141357,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 28140
+    },
+    {
+      "epoch": 4.550965968797995,
+      "grad_norm": 1.0000102519989014,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 28150
+    },
+    {
+      "epoch": 4.5525826529787405,
+      "grad_norm": 1.0717930793762207,
+      "learning_rate": 0.0002,
+      "loss": 0.5327,
+      "step": 28160
+    },
+    {
+      "epoch": 4.554199337159486,
+      "grad_norm": 0.990074634552002,
+      "learning_rate": 0.0002,
+      "loss": 0.5594,
+      "step": 28170
+    },
+    {
+      "epoch": 4.555816021340231,
+      "grad_norm": 0.8673754930496216,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 28180
+    },
+    {
+      "epoch": 4.557432705520976,
+      "grad_norm": 0.864247739315033,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 28190
+    },
+    {
+      "epoch": 4.5590493897017215,
+      "grad_norm": 0.8280200958251953,
+      "learning_rate": 0.0002,
+      "loss": 0.5516,
+      "step": 28200
+    },
+    {
+      "epoch": 4.560666073882467,
+      "grad_norm": 1.1312172412872314,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 28210
+    },
+    {
+      "epoch": 4.562282758063212,
+      "grad_norm": 0.9147403240203857,
+      "learning_rate": 0.0002,
+      "loss": 0.5776,
+      "step": 28220
+    },
+    {
+      "epoch": 4.563899442243958,
+      "grad_norm": 1.0321218967437744,
+      "learning_rate": 0.0002,
+      "loss": 0.5591,
+      "step": 28230
+    },
+    {
+      "epoch": 4.565516126424703,
+      "grad_norm": 1.168332815170288,
+      "learning_rate": 0.0002,
+      "loss": 0.5508,
+      "step": 28240
+    },
+    {
+      "epoch": 4.5671328106054485,
+      "grad_norm": 1.0067222118377686,
+      "learning_rate": 0.0002,
+      "loss": 0.5649,
+      "step": 28250
+    },
+    {
+      "epoch": 4.568749494786194,
+      "grad_norm": 1.0283393859863281,
+      "learning_rate": 0.0002,
+      "loss": 0.5853,
+      "step": 28260
+    },
+    {
+      "epoch": 4.570366178966939,
+      "grad_norm": 0.9912363886833191,
+      "learning_rate": 0.0002,
+      "loss": 0.5772,
+      "step": 28270
+    },
+    {
+      "epoch": 4.571982863147684,
+      "grad_norm": 1.108032464981079,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 28280
+    },
+    {
+      "epoch": 4.573599547328429,
+      "grad_norm": 0.8260078430175781,
+      "learning_rate": 0.0002,
+      "loss": 0.5529,
+      "step": 28290
+    },
+    {
+      "epoch": 4.575216231509175,
+      "grad_norm": 0.8946247100830078,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 28300
+    },
+    {
+      "epoch": 4.57683291568992,
+      "grad_norm": 0.8273587822914124,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 28310
+    },
+    {
+      "epoch": 4.578449599870665,
+      "grad_norm": 0.9040093421936035,
+      "learning_rate": 0.0002,
+      "loss": 0.6058,
+      "step": 28320
+    },
+    {
+      "epoch": 4.58006628405141,
+      "grad_norm": 0.8435290455818176,
+      "learning_rate": 0.0002,
+      "loss": 0.5521,
+      "step": 28330
+    },
+    {
+      "epoch": 4.581682968232156,
+      "grad_norm": 1.164088249206543,
+      "learning_rate": 0.0002,
+      "loss": 0.6086,
+      "step": 28340
+    },
+    {
+      "epoch": 4.583299652412901,
+      "grad_norm": 0.9861085414886475,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28350
+    },
+    {
+      "epoch": 4.584916336593646,
+      "grad_norm": 0.8892980813980103,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 28360
+    },
+    {
+      "epoch": 4.586533020774391,
+      "grad_norm": 1.240574836730957,
+      "learning_rate": 0.0002,
+      "loss": 0.598,
+      "step": 28370
+    },
+    {
+      "epoch": 4.588149704955137,
+      "grad_norm": 0.8669408559799194,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 28380
+    },
+    {
+      "epoch": 4.589766389135883,
+      "grad_norm": 0.9145985841751099,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28390
+    },
+    {
+      "epoch": 4.591383073316628,
+      "grad_norm": 0.8584614992141724,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 28400
+    },
+    {
+      "epoch": 4.592999757497373,
+      "grad_norm": 1.118829369544983,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 28410
+    },
+    {
+      "epoch": 4.594616441678118,
+      "grad_norm": 1.1411553621292114,
+      "learning_rate": 0.0002,
+      "loss": 0.5641,
+      "step": 28420
+    },
+    {
+      "epoch": 4.596233125858864,
+      "grad_norm": 0.9433278441429138,
+      "learning_rate": 0.0002,
+      "loss": 0.549,
+      "step": 28430
+    },
+    {
+      "epoch": 4.597849810039609,
+      "grad_norm": 0.816830039024353,
+      "learning_rate": 0.0002,
+      "loss": 0.5496,
+      "step": 28440
+    },
+    {
+      "epoch": 4.599466494220354,
+      "grad_norm": 1.2124968767166138,
+      "learning_rate": 0.0002,
+      "loss": 0.5543,
+      "step": 28450
+    },
+    {
+      "epoch": 4.601083178401099,
+      "grad_norm": 0.9658762216567993,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 28460
+    },
+    {
+      "epoch": 4.6026998625818445,
+      "grad_norm": 0.836100161075592,
+      "learning_rate": 0.0002,
+      "loss": 0.5902,
+      "step": 28470
+    },
+    {
+      "epoch": 4.60431654676259,
+      "grad_norm": 0.9989104270935059,
+      "learning_rate": 0.0002,
+      "loss": 0.5749,
+      "step": 28480
+    },
+    {
+      "epoch": 4.605933230943335,
+      "grad_norm": 1.1298956871032715,
+      "learning_rate": 0.0002,
+      "loss": 0.5616,
+      "step": 28490
+    },
+    {
+      "epoch": 4.60754991512408,
+      "grad_norm": 1.1731704473495483,
+      "learning_rate": 0.0002,
+      "loss": 0.5846,
+      "step": 28500
+    },
+    {
+      "epoch": 4.609166599304825,
+      "grad_norm": 0.9624714255332947,
+      "learning_rate": 0.0002,
+      "loss": 0.5816,
+      "step": 28510
+    },
+    {
+      "epoch": 4.610783283485571,
+      "grad_norm": 1.364073634147644,
+      "learning_rate": 0.0002,
+      "loss": 0.5868,
+      "step": 28520
+    },
+    {
+      "epoch": 4.612399967666317,
+      "grad_norm": 1.1827356815338135,
+      "learning_rate": 0.0002,
+      "loss": 0.6237,
+      "step": 28530
+    },
+    {
+      "epoch": 4.614016651847062,
+      "grad_norm": 0.6651531457901001,
+      "learning_rate": 0.0002,
+      "loss": 0.5643,
+      "step": 28540
+    },
+    {
+      "epoch": 4.615633336027807,
+      "grad_norm": 1.1640995740890503,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 28550
+    },
+    {
+      "epoch": 4.6172500202085525,
+      "grad_norm": 1.028918743133545,
+      "learning_rate": 0.0002,
+      "loss": 0.5995,
+      "step": 28560
+    },
+    {
+      "epoch": 4.618866704389298,
+      "grad_norm": 0.8252120614051819,
+      "learning_rate": 0.0002,
+      "loss": 0.5607,
+      "step": 28570
+    },
+    {
+      "epoch": 4.620483388570043,
+      "grad_norm": 1.3536735773086548,
+      "learning_rate": 0.0002,
+      "loss": 0.5769,
+      "step": 28580
+    },
+    {
+      "epoch": 4.622100072750788,
+      "grad_norm": 1.2146915197372437,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 28590
+    },
+    {
+      "epoch": 4.623716756931533,
+      "grad_norm": 1.0122549533843994,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 28600
+    },
+    {
+      "epoch": 4.625333441112279,
+      "grad_norm": 0.9977872967720032,
+      "learning_rate": 0.0002,
+      "loss": 0.6072,
+      "step": 28610
+    },
+    {
+      "epoch": 4.626950125293024,
+      "grad_norm": 1.0159751176834106,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 28620
+    },
+    {
+      "epoch": 4.628566809473769,
+      "grad_norm": 1.0028325319290161,
+      "learning_rate": 0.0002,
+      "loss": 0.5935,
+      "step": 28630
+    },
+    {
+      "epoch": 4.630183493654514,
+      "grad_norm": 0.901638388633728,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 28640
+    },
+    {
+      "epoch": 4.6318001778352595,
+      "grad_norm": 0.9450507164001465,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 28650
+    },
+    {
+      "epoch": 4.633416862016006,
+      "grad_norm": 0.9987545013427734,
+      "learning_rate": 0.0002,
+      "loss": 0.5972,
+      "step": 28660
+    },
+    {
+      "epoch": 4.63503354619675,
+      "grad_norm": 0.9574332237243652,
+      "learning_rate": 0.0002,
+      "loss": 0.5863,
+      "step": 28670
+    },
+    {
+      "epoch": 4.636650230377496,
+      "grad_norm": 1.2215653657913208,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 28680
+    },
+    {
+      "epoch": 4.638266914558241,
+      "grad_norm": 0.9798858761787415,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 28690
+    },
+    {
+      "epoch": 4.639883598738987,
+      "grad_norm": 1.0648466348648071,
+      "learning_rate": 0.0002,
+      "loss": 0.5773,
+      "step": 28700
+    },
+    {
+      "epoch": 4.641500282919732,
+      "grad_norm": 1.0606504678726196,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 28710
+    },
+    {
+      "epoch": 4.643116967100477,
+      "grad_norm": 1.0892442464828491,
+      "learning_rate": 0.0002,
+      "loss": 0.5801,
+      "step": 28720
+    },
+    {
+      "epoch": 4.644733651281222,
+      "grad_norm": 0.914391040802002,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 28730
+    },
+    {
+      "epoch": 4.6463503354619675,
+      "grad_norm": 0.9782370328903198,
+      "learning_rate": 0.0002,
+      "loss": 0.5439,
+      "step": 28740
+    },
+    {
+      "epoch": 4.647967019642713,
+      "grad_norm": 1.0344339609146118,
+      "learning_rate": 0.0002,
+      "loss": 0.6035,
+      "step": 28750
+    },
+    {
+      "epoch": 4.649583703823458,
+      "grad_norm": 1.0513931512832642,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 28760
+    },
+    {
+      "epoch": 4.651200388004203,
+      "grad_norm": 0.9711475968360901,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 28770
+    },
+    {
+      "epoch": 4.652817072184948,
+      "grad_norm": 0.977519690990448,
+      "learning_rate": 0.0002,
+      "loss": 0.5472,
+      "step": 28780
+    },
+    {
+      "epoch": 4.654433756365694,
+      "grad_norm": 0.9150224924087524,
+      "learning_rate": 0.0002,
+      "loss": 0.5826,
+      "step": 28790
+    },
+    {
+      "epoch": 4.656050440546439,
+      "grad_norm": 1.0973542928695679,
+      "learning_rate": 0.0002,
+      "loss": 0.5382,
+      "step": 28800
+    },
+    {
+      "epoch": 4.657667124727185,
+      "grad_norm": 0.944877564907074,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 28810
+    },
+    {
+      "epoch": 4.659283808907929,
+      "grad_norm": 0.9508748650550842,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 28820
+    },
+    {
+      "epoch": 4.6609004930886755,
+      "grad_norm": 0.9681721329689026,
+      "learning_rate": 0.0002,
+      "loss": 0.5537,
+      "step": 28830
+    },
+    {
+      "epoch": 4.662517177269421,
+      "grad_norm": 1.0214351415634155,
+      "learning_rate": 0.0002,
+      "loss": 0.592,
+      "step": 28840
+    },
+    {
+      "epoch": 4.664133861450166,
+      "grad_norm": 0.9748611450195312,
+      "learning_rate": 0.0002,
+      "loss": 0.6031,
+      "step": 28850
+    },
+    {
+      "epoch": 4.665750545630911,
+      "grad_norm": 0.8484147191047668,
+      "learning_rate": 0.0002,
+      "loss": 0.572,
+      "step": 28860
+    },
+    {
+      "epoch": 4.667367229811656,
+      "grad_norm": 1.1252986192703247,
+      "learning_rate": 0.0002,
+      "loss": 0.5699,
+      "step": 28870
+    },
+    {
+      "epoch": 4.668983913992402,
+      "grad_norm": 0.8706206679344177,
+      "learning_rate": 0.0002,
+      "loss": 0.5724,
+      "step": 28880
+    },
+    {
+      "epoch": 4.670600598173147,
+      "grad_norm": 1.1432424783706665,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 28890
+    },
+    {
+      "epoch": 4.672217282353892,
+      "grad_norm": 1.017029047012329,
+      "learning_rate": 0.0002,
+      "loss": 0.5675,
+      "step": 28900
+    },
+    {
+      "epoch": 4.673833966534637,
+      "grad_norm": 1.085597038269043,
+      "learning_rate": 0.0002,
+      "loss": 0.5831,
+      "step": 28910
+    },
+    {
+      "epoch": 4.675450650715383,
+      "grad_norm": 0.9275796413421631,
+      "learning_rate": 0.0002,
+      "loss": 0.5678,
+      "step": 28920
+    },
+    {
+      "epoch": 4.677067334896128,
+      "grad_norm": 0.9518964886665344,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 28930
+    },
+    {
+      "epoch": 4.678684019076873,
+      "grad_norm": 1.0352122783660889,
+      "learning_rate": 0.0002,
+      "loss": 0.6232,
+      "step": 28940
+    },
+    {
+      "epoch": 4.680300703257618,
+      "grad_norm": 1.090124249458313,
+      "learning_rate": 0.0002,
+      "loss": 0.5786,
+      "step": 28950
+    },
+    {
+      "epoch": 4.681917387438364,
+      "grad_norm": 0.8799563050270081,
+      "learning_rate": 0.0002,
+      "loss": 0.5728,
+      "step": 28960
+    },
+    {
+      "epoch": 4.683534071619109,
+      "grad_norm": 1.0929821729660034,
+      "learning_rate": 0.0002,
+      "loss": 0.5787,
+      "step": 28970
+    },
+    {
+      "epoch": 4.685150755799855,
+      "grad_norm": 0.903727650642395,
+      "learning_rate": 0.0002,
+      "loss": 0.6134,
+      "step": 28980
+    },
+    {
+      "epoch": 4.6867674399806,
+      "grad_norm": 0.9752424955368042,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 28990
+    },
+    {
+      "epoch": 4.688384124161345,
+      "grad_norm": 0.9351571202278137,
+      "learning_rate": 0.0002,
+      "loss": 0.5762,
+      "step": 29000
+    },
+    {
+      "epoch": 4.6900008083420905,
+      "grad_norm": 0.923877477645874,
+      "learning_rate": 0.0002,
+      "loss": 0.5811,
+      "step": 29010
+    },
+    {
+      "epoch": 4.691617492522836,
+      "grad_norm": 1.045389175415039,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 29020
+    },
+    {
+      "epoch": 4.693234176703581,
+      "grad_norm": 1.0200831890106201,
+      "learning_rate": 0.0002,
+      "loss": 0.584,
+      "step": 29030
+    },
+    {
+      "epoch": 4.694850860884326,
+      "grad_norm": 1.1499706506729126,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 29040
+    },
+    {
+      "epoch": 4.6964675450650715,
+      "grad_norm": 0.860118567943573,
+      "learning_rate": 0.0002,
+      "loss": 0.5745,
+      "step": 29050
+    },
+    {
+      "epoch": 4.698084229245817,
+      "grad_norm": 0.9774864315986633,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 29060
+    },
+    {
+      "epoch": 4.699700913426562,
+      "grad_norm": 1.0323210954666138,
+      "learning_rate": 0.0002,
+      "loss": 0.5765,
+      "step": 29070
+    },
+    {
+      "epoch": 4.701317597607307,
+      "grad_norm": 0.8492481112480164,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 29080
+    },
+    {
+      "epoch": 4.702934281788052,
+      "grad_norm": 1.131951093673706,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 29090
+    },
+    {
+      "epoch": 4.704550965968798,
+      "grad_norm": 0.8763113021850586,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 29100
+    },
+    {
+      "epoch": 4.706167650149544,
+      "grad_norm": 1.045028805732727,
+      "learning_rate": 0.0002,
+      "loss": 0.575,
+      "step": 29110
+    },
+    {
+      "epoch": 4.707784334330288,
+      "grad_norm": 0.9961401224136353,
+      "learning_rate": 0.0002,
+      "loss": 0.5548,
+      "step": 29120
+    },
+    {
+      "epoch": 4.709401018511034,
+      "grad_norm": 0.9282503724098206,
+      "learning_rate": 0.0002,
+      "loss": 0.559,
+      "step": 29130
+    },
+    {
+      "epoch": 4.711017702691779,
+      "grad_norm": 1.1418932676315308,
+      "learning_rate": 0.0002,
+      "loss": 0.5744,
+      "step": 29140
+    },
+    {
+      "epoch": 4.712634386872525,
+      "grad_norm": 0.9950099587440491,
+      "learning_rate": 0.0002,
+      "loss": 0.5394,
+      "step": 29150
+    },
+    {
+      "epoch": 4.71425107105327,
+      "grad_norm": 0.8304893374443054,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 29160
+    },
+    {
+      "epoch": 4.715867755234015,
+      "grad_norm": 1.115626335144043,
+      "learning_rate": 0.0002,
+      "loss": 0.6074,
+      "step": 29170
+    },
+    {
+      "epoch": 4.71748443941476,
+      "grad_norm": 1.079818606376648,
+      "learning_rate": 0.0002,
+      "loss": 0.6265,
+      "step": 29180
+    },
+    {
+      "epoch": 4.719101123595506,
+      "grad_norm": 1.1929082870483398,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 29190
+    },
+    {
+      "epoch": 4.720717807776251,
+      "grad_norm": 0.9621080756187439,
+      "learning_rate": 0.0002,
+      "loss": 0.5708,
+      "step": 29200
+    },
+    {
+      "epoch": 4.722334491956996,
+      "grad_norm": 0.8549222350120544,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 29210
+    },
+    {
+      "epoch": 4.723951176137741,
+      "grad_norm": 0.9341941475868225,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 29220
+    },
+    {
+      "epoch": 4.7255678603184865,
+      "grad_norm": 1.075406789779663,
+      "learning_rate": 0.0002,
+      "loss": 0.5436,
+      "step": 29230
+    },
+    {
+      "epoch": 4.727184544499232,
+      "grad_norm": 1.0859880447387695,
+      "learning_rate": 0.0002,
+      "loss": 0.576,
+      "step": 29240
+    },
+    {
+      "epoch": 4.728801228679977,
+      "grad_norm": 0.8475605249404907,
+      "learning_rate": 0.0002,
+      "loss": 0.5525,
+      "step": 29250
+    },
+    {
+      "epoch": 4.730417912860723,
+      "grad_norm": 0.9331845641136169,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 29260
+    },
+    {
+      "epoch": 4.7320345970414674,
+      "grad_norm": 0.9279314279556274,
+      "learning_rate": 0.0002,
+      "loss": 0.5901,
+      "step": 29270
+    },
+    {
+      "epoch": 4.733651281222214,
+      "grad_norm": 0.7803558707237244,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 29280
+    },
+    {
+      "epoch": 4.735267965402959,
+      "grad_norm": 1.0159329175949097,
+      "learning_rate": 0.0002,
+      "loss": 0.5968,
+      "step": 29290
+    },
+    {
+      "epoch": 4.736884649583704,
+      "grad_norm": 0.9448670744895935,
+      "learning_rate": 0.0002,
+      "loss": 0.5333,
+      "step": 29300
+    },
+    {
+      "epoch": 4.738501333764449,
+      "grad_norm": 1.0732197761535645,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 29310
+    },
+    {
+      "epoch": 4.7401180179451945,
+      "grad_norm": 0.901830792427063,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 29320
+    },
+    {
+      "epoch": 4.74173470212594,
+      "grad_norm": 0.9141789674758911,
+      "learning_rate": 0.0002,
+      "loss": 0.6105,
+      "step": 29330
+    },
+    {
+      "epoch": 4.743351386306685,
+      "grad_norm": 0.9733418226242065,
+      "learning_rate": 0.0002,
+      "loss": 0.5481,
+      "step": 29340
+    },
+    {
+      "epoch": 4.74496807048743,
+      "grad_norm": 0.909810483455658,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 29350
+    },
+    {
+      "epoch": 4.746584754668175,
+      "grad_norm": 0.909541666507721,
+      "learning_rate": 0.0002,
+      "loss": 0.5911,
+      "step": 29360
+    },
+    {
+      "epoch": 4.748201438848921,
+      "grad_norm": 0.9383015632629395,
+      "learning_rate": 0.0002,
+      "loss": 0.5579,
+      "step": 29370
+    },
+    {
+      "epoch": 4.749818123029666,
+      "grad_norm": 0.9275668263435364,
+      "learning_rate": 0.0002,
+      "loss": 0.5529,
+      "step": 29380
+    },
+    {
+      "epoch": 4.751434807210411,
+      "grad_norm": 1.1146225929260254,
+      "learning_rate": 0.0002,
+      "loss": 0.5623,
+      "step": 29390
+    },
+    {
+      "epoch": 4.753051491391156,
+      "grad_norm": 1.0062453746795654,
+      "learning_rate": 0.0002,
+      "loss": 0.6018,
+      "step": 29400
+    },
+    {
+      "epoch": 4.7546681755719025,
+      "grad_norm": 0.9451895952224731,
+      "learning_rate": 0.0002,
+      "loss": 0.5872,
+      "step": 29410
+    },
+    {
+      "epoch": 4.756284859752648,
+      "grad_norm": 0.870457649230957,
+      "learning_rate": 0.0002,
+      "loss": 0.5767,
+      "step": 29420
+    },
+    {
+      "epoch": 4.757901543933393,
+      "grad_norm": 1.0411282777786255,
+      "learning_rate": 0.0002,
+      "loss": 0.57,
+      "step": 29430
+    },
+    {
+      "epoch": 4.759518228114138,
+      "grad_norm": 1.1648986339569092,
+      "learning_rate": 0.0002,
+      "loss": 0.5688,
+      "step": 29440
+    },
+    {
+      "epoch": 4.761134912294883,
+      "grad_norm": 0.8999572992324829,
+      "learning_rate": 0.0002,
+      "loss": 0.5432,
+      "step": 29450
+    },
+    {
+      "epoch": 4.762751596475629,
+      "grad_norm": 0.9863559007644653,
+      "learning_rate": 0.0002,
+      "loss": 0.5667,
+      "step": 29460
+    },
+    {
+      "epoch": 4.764368280656374,
+      "grad_norm": 0.9676542282104492,
+      "learning_rate": 0.0002,
+      "loss": 0.5779,
+      "step": 29470
+    },
+    {
+      "epoch": 4.765984964837119,
+      "grad_norm": 1.004775047302246,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 29480
+    },
+    {
+      "epoch": 4.767601649017864,
+      "grad_norm": 1.0937515497207642,
+      "learning_rate": 0.0002,
+      "loss": 0.6044,
+      "step": 29490
+    },
+    {
+      "epoch": 4.7692183331986095,
+      "grad_norm": 0.9551598429679871,
+      "learning_rate": 0.0002,
+      "loss": 0.5433,
+      "step": 29500
+    },
+    {
+      "epoch": 4.770835017379355,
+      "grad_norm": 1.0757228136062622,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 29510
+    },
+    {
+      "epoch": 4.7724517015601,
+      "grad_norm": 1.0588841438293457,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 29520
+    },
+    {
+      "epoch": 4.774068385740845,
+      "grad_norm": 1.0744032859802246,
+      "learning_rate": 0.0002,
+      "loss": 0.5814,
+      "step": 29530
+    },
+    {
+      "epoch": 4.7756850699215905,
+      "grad_norm": 1.0066277980804443,
+      "learning_rate": 0.0002,
+      "loss": 0.5681,
+      "step": 29540
+    },
+    {
+      "epoch": 4.777301754102336,
+      "grad_norm": 1.082319736480713,
+      "learning_rate": 0.0002,
+      "loss": 0.545,
+      "step": 29550
+    },
+    {
+      "epoch": 4.778918438283082,
+      "grad_norm": 0.8252472877502441,
+      "learning_rate": 0.0002,
+      "loss": 0.5709,
+      "step": 29560
+    },
+    {
+      "epoch": 4.780535122463827,
+      "grad_norm": 0.9855340123176575,
+      "learning_rate": 0.0002,
+      "loss": 0.5666,
+      "step": 29570
+    },
+    {
+      "epoch": 4.782151806644572,
+      "grad_norm": 0.9991421699523926,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 29580
+    },
+    {
+      "epoch": 4.7837684908253175,
+      "grad_norm": 1.316841959953308,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 29590
+    },
+    {
+      "epoch": 4.785385175006063,
+      "grad_norm": 1.1513035297393799,
+      "learning_rate": 0.0002,
+      "loss": 0.6102,
+      "step": 29600
+    },
+    {
+      "epoch": 4.787001859186808,
+      "grad_norm": 0.9767683744430542,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 29610
+    },
+    {
+      "epoch": 4.788618543367553,
+      "grad_norm": 0.9786278605461121,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 29620
+    },
+    {
+      "epoch": 4.7902352275482984,
+      "grad_norm": 0.8004973530769348,
+      "learning_rate": 0.0002,
+      "loss": 0.6108,
+      "step": 29630
+    },
+    {
+      "epoch": 4.791851911729044,
+      "grad_norm": 1.0997767448425293,
+      "learning_rate": 0.0002,
+      "loss": 0.5932,
+      "step": 29640
+    },
+    {
+      "epoch": 4.793468595909789,
+      "grad_norm": 0.9752856492996216,
+      "learning_rate": 0.0002,
+      "loss": 0.5655,
+      "step": 29650
+    },
+    {
+      "epoch": 4.795085280090534,
+      "grad_norm": 1.0518392324447632,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 29660
+    },
+    {
+      "epoch": 4.796701964271279,
+      "grad_norm": 1.1050055027008057,
+      "learning_rate": 0.0002,
+      "loss": 0.6042,
+      "step": 29670
+    },
+    {
+      "epoch": 4.798318648452025,
+      "grad_norm": 0.9933857917785645,
+      "learning_rate": 0.0002,
+      "loss": 0.6089,
+      "step": 29680
+    },
+    {
+      "epoch": 4.79993533263277,
+      "grad_norm": 1.2804018259048462,
+      "learning_rate": 0.0002,
+      "loss": 0.6041,
+      "step": 29690
+    },
+    {
+      "epoch": 4.801552016813515,
+      "grad_norm": 1.0133371353149414,
+      "learning_rate": 0.0002,
+      "loss": 0.636,
+      "step": 29700
+    },
+    {
+      "epoch": 4.803168700994261,
+      "grad_norm": 1.080350637435913,
+      "learning_rate": 0.0002,
+      "loss": 0.5662,
+      "step": 29710
+    },
+    {
+      "epoch": 4.804785385175006,
+      "grad_norm": 0.9986529350280762,
+      "learning_rate": 0.0002,
+      "loss": 0.5603,
+      "step": 29720
+    },
+    {
+      "epoch": 4.806402069355752,
+      "grad_norm": 0.975665807723999,
+      "learning_rate": 0.0002,
+      "loss": 0.5894,
+      "step": 29730
+    },
+    {
+      "epoch": 4.808018753536497,
+      "grad_norm": 0.8458138704299927,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 29740
+    },
+    {
+      "epoch": 4.809635437717242,
+      "grad_norm": 0.99330073595047,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 29750
+    },
+    {
+      "epoch": 4.811252121897987,
+      "grad_norm": 0.898274302482605,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 29760
+    },
+    {
+      "epoch": 4.812868806078733,
+      "grad_norm": 1.0504480600357056,
+      "learning_rate": 0.0002,
+      "loss": 0.5842,
+      "step": 29770
+    },
+    {
+      "epoch": 4.814485490259478,
+      "grad_norm": 0.937919020652771,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 29780
+    },
+    {
+      "epoch": 4.816102174440223,
+      "grad_norm": 0.9593307971954346,
+      "learning_rate": 0.0002,
+      "loss": 0.5885,
+      "step": 29790
+    },
+    {
+      "epoch": 4.817718858620968,
+      "grad_norm": 0.9431198835372925,
+      "learning_rate": 0.0002,
+      "loss": 0.578,
+      "step": 29800
+    },
+    {
+      "epoch": 4.8193355428017135,
+      "grad_norm": 1.2729957103729248,
+      "learning_rate": 0.0002,
+      "loss": 0.5739,
+      "step": 29810
+    },
+    {
+      "epoch": 4.820952226982459,
+      "grad_norm": 0.8876838684082031,
+      "learning_rate": 0.0002,
+      "loss": 0.6124,
+      "step": 29820
+    },
+    {
+      "epoch": 4.822568911163204,
+      "grad_norm": 1.0185000896453857,
+      "learning_rate": 0.0002,
+      "loss": 0.5583,
+      "step": 29830
+    },
+    {
+      "epoch": 4.824185595343949,
+      "grad_norm": 1.064276099205017,
+      "learning_rate": 0.0002,
+      "loss": 0.5686,
+      "step": 29840
+    },
+    {
+      "epoch": 4.825802279524694,
+      "grad_norm": 0.9774803519248962,
+      "learning_rate": 0.0002,
+      "loss": 0.5698,
+      "step": 29850
+    },
+    {
+      "epoch": 4.8274189637054405,
+      "grad_norm": 1.131646990776062,
+      "learning_rate": 0.0002,
+      "loss": 0.5533,
+      "step": 29860
+    },
+    {
+      "epoch": 4.829035647886186,
+      "grad_norm": 1.081455945968628,
+      "learning_rate": 0.0002,
+      "loss": 0.6371,
+      "step": 29870
+    },
+    {
+      "epoch": 4.830652332066931,
+      "grad_norm": 0.990538477897644,
+      "learning_rate": 0.0002,
+      "loss": 0.5793,
+      "step": 29880
+    },
+    {
+      "epoch": 4.832269016247676,
+      "grad_norm": 0.9750600457191467,
+      "learning_rate": 0.0002,
+      "loss": 0.5833,
+      "step": 29890
+    },
+    {
+      "epoch": 4.8338857004284215,
+      "grad_norm": 1.0600621700286865,
+      "learning_rate": 0.0002,
+      "loss": 0.619,
+      "step": 29900
+    },
+    {
+      "epoch": 4.835502384609167,
+      "grad_norm": 0.9237320423126221,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 29910
+    },
+    {
+      "epoch": 4.837119068789912,
+      "grad_norm": 0.9739177227020264,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 29920
+    },
+    {
+      "epoch": 4.838735752970657,
+      "grad_norm": 1.128677248954773,
+      "learning_rate": 0.0002,
+      "loss": 0.587,
+      "step": 29930
+    },
+    {
+      "epoch": 4.840352437151402,
+      "grad_norm": 1.042604923248291,
+      "learning_rate": 0.0002,
+      "loss": 0.564,
+      "step": 29940
+    },
+    {
+      "epoch": 4.841969121332148,
+      "grad_norm": 0.849758505821228,
+      "learning_rate": 0.0002,
+      "loss": 0.5885,
+      "step": 29950
+    },
+    {
+      "epoch": 4.843585805512893,
+      "grad_norm": 1.2809888124465942,
+      "learning_rate": 0.0002,
+      "loss": 0.5952,
+      "step": 29960
+    },
+    {
+      "epoch": 4.845202489693638,
+      "grad_norm": 1.0177865028381348,
+      "learning_rate": 0.0002,
+      "loss": 0.5703,
+      "step": 29970
+    },
+    {
+      "epoch": 4.846819173874383,
+      "grad_norm": 1.0026639699935913,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 29980
+    },
+    {
+      "epoch": 4.8484358580551286,
+      "grad_norm": 0.9679505228996277,
+      "learning_rate": 0.0002,
+      "loss": 0.5897,
+      "step": 29990
+    },
+    {
+      "epoch": 4.850052542235874,
+      "grad_norm": 0.8939532041549683,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 30000
+    },
+    {
+      "epoch": 4.85166922641662,
+      "grad_norm": 0.9957457780838013,
+      "learning_rate": 0.0002,
+      "loss": 0.5852,
+      "step": 30010
+    },
+    {
+      "epoch": 4.853285910597365,
+      "grad_norm": 1.1646790504455566,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 30020
+    },
+    {
+      "epoch": 4.85490259477811,
+      "grad_norm": 0.8804680705070496,
+      "learning_rate": 0.0002,
+      "loss": 0.5711,
+      "step": 30030
+    },
+    {
+      "epoch": 4.856519278958856,
+      "grad_norm": 1.161970853805542,
+      "learning_rate": 0.0002,
+      "loss": 0.5397,
+      "step": 30040
+    },
+    {
+      "epoch": 4.858135963139601,
+      "grad_norm": 0.9081037640571594,
+      "learning_rate": 0.0002,
+      "loss": 0.5552,
+      "step": 30050
+    },
+    {
+      "epoch": 4.859752647320346,
+      "grad_norm": 0.9402848482131958,
+      "learning_rate": 0.0002,
+      "loss": 0.6024,
+      "step": 30060
+    },
+    {
+      "epoch": 4.861369331501091,
+      "grad_norm": 0.9023865461349487,
+      "learning_rate": 0.0002,
+      "loss": 0.6256,
+      "step": 30070
+    },
+    {
+      "epoch": 4.8629860156818365,
+      "grad_norm": 1.0173414945602417,
+      "learning_rate": 0.0002,
+      "loss": 0.5926,
+      "step": 30080
+    },
+    {
+      "epoch": 4.864602699862582,
+      "grad_norm": 1.084402322769165,
+      "learning_rate": 0.0002,
+      "loss": 0.6274,
+      "step": 30090
+    },
+    {
+      "epoch": 4.866219384043327,
+      "grad_norm": 0.9577937126159668,
+      "learning_rate": 0.0002,
+      "loss": 0.6311,
+      "step": 30100
+    },
+    {
+      "epoch": 4.867836068224072,
+      "grad_norm": 0.9807606935501099,
+      "learning_rate": 0.0002,
+      "loss": 0.5724,
+      "step": 30110
+    },
+    {
+      "epoch": 4.8694527524048175,
+      "grad_norm": 0.978784441947937,
+      "learning_rate": 0.0002,
+      "loss": 0.5786,
+      "step": 30120
+    },
+    {
+      "epoch": 4.871069436585563,
+      "grad_norm": 0.9762914776802063,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 30130
+    },
+    {
+      "epoch": 4.872686120766308,
+      "grad_norm": 0.9404871463775635,
+      "learning_rate": 0.0002,
+      "loss": 0.5892,
+      "step": 30140
+    },
+    {
+      "epoch": 4.874302804947053,
+      "grad_norm": 1.0069509744644165,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 30150
+    },
+    {
+      "epoch": 4.875919489127799,
+      "grad_norm": 1.1770923137664795,
+      "learning_rate": 0.0002,
+      "loss": 0.6225,
+      "step": 30160
+    },
+    {
+      "epoch": 4.8775361733085445,
+      "grad_norm": 1.021210789680481,
+      "learning_rate": 0.0002,
+      "loss": 0.5657,
+      "step": 30170
+    },
+    {
+      "epoch": 4.87915285748929,
+      "grad_norm": 0.8512648940086365,
+      "learning_rate": 0.0002,
+      "loss": 0.6033,
+      "step": 30180
+    },
+    {
+      "epoch": 4.880769541670035,
+      "grad_norm": 0.9345870018005371,
+      "learning_rate": 0.0002,
+      "loss": 0.5519,
+      "step": 30190
+    },
+    {
+      "epoch": 4.88238622585078,
+      "grad_norm": 1.0224418640136719,
+      "learning_rate": 0.0002,
+      "loss": 0.5682,
+      "step": 30200
+    },
+    {
+      "epoch": 4.884002910031525,
+      "grad_norm": 1.0316044092178345,
+      "learning_rate": 0.0002,
+      "loss": 0.5807,
+      "step": 30210
+    },
+    {
+      "epoch": 4.885619594212271,
+      "grad_norm": 1.102437973022461,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 30220
+    },
+    {
+      "epoch": 4.887236278393016,
+      "grad_norm": 1.0220023393630981,
+      "learning_rate": 0.0002,
+      "loss": 0.586,
+      "step": 30230
+    },
+    {
+      "epoch": 4.888852962573761,
+      "grad_norm": 1.0934523344039917,
+      "learning_rate": 0.0002,
+      "loss": 0.5781,
+      "step": 30240
+    },
+    {
+      "epoch": 4.890469646754506,
+      "grad_norm": 1.264630913734436,
+      "learning_rate": 0.0002,
+      "loss": 0.6313,
+      "step": 30250
+    },
+    {
+      "epoch": 4.892086330935252,
+      "grad_norm": 1.0999879837036133,
+      "learning_rate": 0.0002,
+      "loss": 0.5712,
+      "step": 30260
+    },
+    {
+      "epoch": 4.893703015115997,
+      "grad_norm": 0.9124550223350525,
+      "learning_rate": 0.0002,
+      "loss": 0.6413,
+      "step": 30270
+    },
+    {
+      "epoch": 4.895319699296742,
+      "grad_norm": 0.9853624105453491,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 30280
+    },
+    {
+      "epoch": 4.896936383477488,
+      "grad_norm": 1.0589802265167236,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 30290
+    },
+    {
+      "epoch": 4.8985530676582325,
+      "grad_norm": 0.8487226366996765,
+      "learning_rate": 0.0002,
+      "loss": 0.6129,
+      "step": 30300
+    },
+    {
+      "epoch": 4.900169751838979,
+      "grad_norm": 1.0212191343307495,
+      "learning_rate": 0.0002,
+      "loss": 0.5514,
+      "step": 30310
+    },
+    {
+      "epoch": 4.901786436019724,
+      "grad_norm": 1.0187491178512573,
+      "learning_rate": 0.0002,
+      "loss": 0.5896,
+      "step": 30320
+    },
+    {
+      "epoch": 4.903403120200469,
+      "grad_norm": 1.0013091564178467,
+      "learning_rate": 0.0002,
+      "loss": 0.5809,
+      "step": 30330
+    },
+    {
+      "epoch": 4.905019804381214,
+      "grad_norm": 1.0017542839050293,
+      "learning_rate": 0.0002,
+      "loss": 0.5658,
+      "step": 30340
+    },
+    {
+      "epoch": 4.9066364885619596,
+      "grad_norm": 0.9665151238441467,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 30350
+    },
+    {
+      "epoch": 4.908253172742705,
+      "grad_norm": 0.8774822950363159,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 30360
+    },
+    {
+      "epoch": 4.90986985692345,
+      "grad_norm": 0.9449850916862488,
+      "learning_rate": 0.0002,
+      "loss": 0.5771,
+      "step": 30370
+    },
+    {
+      "epoch": 4.911486541104195,
+      "grad_norm": 0.7368341088294983,
+      "learning_rate": 0.0002,
+      "loss": 0.58,
+      "step": 30380
+    },
+    {
+      "epoch": 4.9131032252849405,
+      "grad_norm": 0.9669167995452881,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 30390
+    },
+    {
+      "epoch": 4.914719909465686,
+      "grad_norm": 1.1227794885635376,
+      "learning_rate": 0.0002,
+      "loss": 0.6202,
+      "step": 30400
+    },
+    {
+      "epoch": 4.916336593646431,
+      "grad_norm": 0.9884361028671265,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 30410
+    },
+    {
+      "epoch": 4.917953277827176,
+      "grad_norm": 0.9949551224708557,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 30420
+    },
+    {
+      "epoch": 4.919569962007921,
+      "grad_norm": 0.9491621851921082,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 30430
+    },
+    {
+      "epoch": 4.9211866461886675,
+      "grad_norm": 0.78848797082901,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 30440
+    },
+    {
+      "epoch": 4.922803330369412,
+      "grad_norm": 1.0693835020065308,
+      "learning_rate": 0.0002,
+      "loss": 0.5561,
+      "step": 30450
+    },
+    {
+      "epoch": 4.924420014550158,
+      "grad_norm": 0.9573729634284973,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 30460
+    },
+    {
+      "epoch": 4.926036698730903,
+      "grad_norm": 0.9975152611732483,
+      "learning_rate": 0.0002,
+      "loss": 0.6084,
+      "step": 30470
+    },
+    {
+      "epoch": 4.9276533829116484,
+      "grad_norm": 0.8695693016052246,
+      "learning_rate": 0.0002,
+      "loss": 0.5969,
+      "step": 30480
+    },
+    {
+      "epoch": 4.929270067092394,
+      "grad_norm": 1.145394206047058,
+      "learning_rate": 0.0002,
+      "loss": 0.6144,
+      "step": 30490
+    },
+    {
+      "epoch": 4.930886751273139,
+      "grad_norm": 0.7668989896774292,
+      "learning_rate": 0.0002,
+      "loss": 0.5736,
+      "step": 30500
+    },
+    {
+      "epoch": 4.932503435453884,
+      "grad_norm": 0.9630151391029358,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 30510
+    },
+    {
+      "epoch": 4.934120119634629,
+      "grad_norm": 0.940705418586731,
+      "learning_rate": 0.0002,
+      "loss": 0.6461,
+      "step": 30520
+    },
+    {
+      "epoch": 4.935736803815375,
+      "grad_norm": 1.3243348598480225,
+      "learning_rate": 0.0002,
+      "loss": 0.6326,
+      "step": 30530
+    },
+    {
+      "epoch": 4.93735348799612,
+      "grad_norm": 1.004347801208496,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 30540
+    },
+    {
+      "epoch": 4.938970172176865,
+      "grad_norm": 0.8711541295051575,
+      "learning_rate": 0.0002,
+      "loss": 0.583,
+      "step": 30550
+    },
+    {
+      "epoch": 4.94058685635761,
+      "grad_norm": 0.8980631828308105,
+      "learning_rate": 0.0002,
+      "loss": 0.599,
+      "step": 30560
+    },
+    {
+      "epoch": 4.9422035405383555,
+      "grad_norm": 0.8388893604278564,
+      "learning_rate": 0.0002,
+      "loss": 0.6024,
+      "step": 30570
+    },
+    {
+      "epoch": 4.943820224719101,
+      "grad_norm": 1.0991183519363403,
+      "learning_rate": 0.0002,
+      "loss": 0.6189,
+      "step": 30580
+    },
+    {
+      "epoch": 4.945436908899847,
+      "grad_norm": 0.9731075763702393,
+      "learning_rate": 0.0002,
+      "loss": 0.5906,
+      "step": 30590
+    },
+    {
+      "epoch": 4.947053593080591,
+      "grad_norm": 1.3904452323913574,
+      "learning_rate": 0.0002,
+      "loss": 0.5883,
+      "step": 30600
+    },
+    {
+      "epoch": 4.948670277261337,
+      "grad_norm": 1.2489882707595825,
+      "learning_rate": 0.0002,
+      "loss": 0.5952,
+      "step": 30610
+    },
+    {
+      "epoch": 4.950286961442083,
+      "grad_norm": 1.240072250366211,
+      "learning_rate": 0.0002,
+      "loss": 0.5887,
+      "step": 30620
+    },
+    {
+      "epoch": 4.951903645622828,
+      "grad_norm": 0.9191411733627319,
+      "learning_rate": 0.0002,
+      "loss": 0.5762,
+      "step": 30630
+    },
+    {
+      "epoch": 4.953520329803573,
+      "grad_norm": 0.8888895511627197,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 30640
+    },
+    {
+      "epoch": 4.955137013984318,
+      "grad_norm": 0.9001450538635254,
+      "learning_rate": 0.0002,
+      "loss": 0.6594,
+      "step": 30650
+    },
+    {
+      "epoch": 4.9567536981650635,
+      "grad_norm": 1.053971767425537,
+      "learning_rate": 0.0002,
+      "loss": 0.6047,
+      "step": 30660
+    },
+    {
+      "epoch": 4.958370382345809,
+      "grad_norm": 1.2224042415618896,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 30670
+    },
+    {
+      "epoch": 4.959987066526554,
+      "grad_norm": 0.8855111598968506,
+      "learning_rate": 0.0002,
+      "loss": 0.6211,
+      "step": 30680
+    },
+    {
+      "epoch": 4.961603750707299,
+      "grad_norm": 0.9489575624465942,
+      "learning_rate": 0.0002,
+      "loss": 0.5764,
+      "step": 30690
+    },
+    {
+      "epoch": 4.963220434888044,
+      "grad_norm": 0.9635404944419861,
+      "learning_rate": 0.0002,
+      "loss": 0.5371,
+      "step": 30700
+    },
+    {
+      "epoch": 4.96483711906879,
+      "grad_norm": 1.1784121990203857,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 30710
+    },
+    {
+      "epoch": 4.966453803249535,
+      "grad_norm": 1.0059462785720825,
+      "learning_rate": 0.0002,
+      "loss": 0.5803,
+      "step": 30720
+    },
+    {
+      "epoch": 4.96807048743028,
+      "grad_norm": 0.9479738473892212,
+      "learning_rate": 0.0002,
+      "loss": 0.5759,
+      "step": 30730
+    },
+    {
+      "epoch": 4.969687171611026,
+      "grad_norm": 1.0624593496322632,
+      "learning_rate": 0.0002,
+      "loss": 0.584,
+      "step": 30740
+    },
+    {
+      "epoch": 4.971303855791771,
+      "grad_norm": 1.1429259777069092,
+      "learning_rate": 0.0002,
+      "loss": 0.6202,
+      "step": 30750
+    },
+    {
+      "epoch": 4.972920539972517,
+      "grad_norm": 0.9102491140365601,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 30760
+    },
+    {
+      "epoch": 4.974537224153262,
+      "grad_norm": 1.1262688636779785,
+      "learning_rate": 0.0002,
+      "loss": 0.6025,
+      "step": 30770
+    },
+    {
+      "epoch": 4.976153908334007,
+      "grad_norm": 1.1415393352508545,
+      "learning_rate": 0.0002,
+      "loss": 0.588,
+      "step": 30780
+    },
+    {
+      "epoch": 4.977770592514752,
+      "grad_norm": 1.083078384399414,
+      "learning_rate": 0.0002,
+      "loss": 0.5832,
+      "step": 30790
+    },
+    {
+      "epoch": 4.979387276695498,
+      "grad_norm": 0.964859127998352,
+      "learning_rate": 0.0002,
+      "loss": 0.6025,
+      "step": 30800
+    },
+    {
+      "epoch": 4.981003960876243,
+      "grad_norm": 0.8704743385314941,
+      "learning_rate": 0.0002,
+      "loss": 0.6095,
+      "step": 30810
+    },
+    {
+      "epoch": 4.982620645056988,
+      "grad_norm": 1.0714856386184692,
+      "learning_rate": 0.0002,
+      "loss": 0.5666,
+      "step": 30820
+    },
+    {
+      "epoch": 4.984237329237733,
+      "grad_norm": 0.6818771362304688,
+      "learning_rate": 0.0002,
+      "loss": 0.565,
+      "step": 30830
+    },
+    {
+      "epoch": 4.985854013418479,
+      "grad_norm": 1.0454156398773193,
+      "learning_rate": 0.0002,
+      "loss": 0.5999,
+      "step": 30840
+    },
+    {
+      "epoch": 4.987470697599224,
+      "grad_norm": 0.9410776495933533,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 30850
+    },
+    {
+      "epoch": 4.989087381779969,
+      "grad_norm": 1.0878902673721313,
+      "learning_rate": 0.0002,
+      "loss": 0.5899,
+      "step": 30860
+    },
+    {
+      "epoch": 4.990704065960714,
+      "grad_norm": 0.8916727304458618,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 30870
+    },
+    {
+      "epoch": 4.9923207501414595,
+      "grad_norm": 1.045776128768921,
+      "learning_rate": 0.0002,
+      "loss": 0.6066,
+      "step": 30880
+    },
+    {
+      "epoch": 4.993937434322206,
+      "grad_norm": 0.9861903786659241,
+      "learning_rate": 0.0002,
+      "loss": 0.5767,
+      "step": 30890
+    },
+    {
+      "epoch": 4.995554118502951,
+      "grad_norm": 0.9275050759315491,
+      "learning_rate": 0.0002,
+      "loss": 0.6192,
+      "step": 30900
+    },
+    {
+      "epoch": 4.997170802683696,
+      "grad_norm": 0.94013911485672,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 30910
+    },
+    {
+      "epoch": 4.998787486864441,
+      "grad_norm": 0.9771268367767334,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 30920
+    },
+    {
+      "epoch": 4.9999191657909625,
+      "eval_loss": 1.1968598365783691,
+      "eval_runtime": 122.2519,
+      "eval_samples_per_second": 5.996,
+      "eval_steps_per_second": 0.753,
+      "step": 30927
+    },
+    {
+      "epoch": 5.0004041710451865,
+      "grad_norm": 0.8021580576896667,
+      "learning_rate": 0.0002,
+      "loss": 0.5238,
+      "step": 30930
+    },
+    {
+      "epoch": 5.002020855225932,
+      "grad_norm": 1.0807327032089233,
+      "learning_rate": 0.0002,
+      "loss": 0.4984,
+      "step": 30940
+    },
+    {
+      "epoch": 5.003637539406677,
+      "grad_norm": 1.1638425588607788,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 30950
+    },
+    {
+      "epoch": 5.005254223587422,
+      "grad_norm": 1.1700230836868286,
+      "learning_rate": 0.0002,
+      "loss": 0.4621,
+      "step": 30960
+    },
+    {
+      "epoch": 5.0068709077681675,
+      "grad_norm": 0.9053420424461365,
+      "learning_rate": 0.0002,
+      "loss": 0.4657,
+      "step": 30970
+    },
+    {
+      "epoch": 5.008487591948913,
+      "grad_norm": 0.9226111769676208,
+      "learning_rate": 0.0002,
+      "loss": 0.4865,
+      "step": 30980
+    },
+    {
+      "epoch": 5.010104276129658,
+      "grad_norm": 1.238669514656067,
+      "learning_rate": 0.0002,
+      "loss": 0.5011,
+      "step": 30990
+    },
+    {
+      "epoch": 5.011720960310403,
+      "grad_norm": 1.0668327808380127,
+      "learning_rate": 0.0002,
+      "loss": 0.4754,
+      "step": 31000
+    },
+    {
+      "epoch": 5.013337644491148,
+      "grad_norm": 1.0903944969177246,
+      "learning_rate": 0.0002,
+      "loss": 0.5414,
+      "step": 31010
+    },
+    {
+      "epoch": 5.014954328671894,
+      "grad_norm": 1.0763911008834839,
+      "learning_rate": 0.0002,
+      "loss": 0.5117,
+      "step": 31020
+    },
+    {
+      "epoch": 5.016571012852639,
+      "grad_norm": 1.0108771324157715,
+      "learning_rate": 0.0002,
+      "loss": 0.4908,
+      "step": 31030
+    },
+    {
+      "epoch": 5.018187697033385,
+      "grad_norm": 0.8816103935241699,
+      "learning_rate": 0.0002,
+      "loss": 0.5052,
+      "step": 31040
+    },
+    {
+      "epoch": 5.01980438121413,
+      "grad_norm": 1.11434805393219,
+      "learning_rate": 0.0002,
+      "loss": 0.4985,
+      "step": 31050
+    },
+    {
+      "epoch": 5.021421065394875,
+      "grad_norm": 1.0727789402008057,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 31060
+    },
+    {
+      "epoch": 5.023037749575621,
+      "grad_norm": 1.1480379104614258,
+      "learning_rate": 0.0002,
+      "loss": 0.4938,
+      "step": 31070
+    },
+    {
+      "epoch": 5.024654433756366,
+      "grad_norm": 1.0913071632385254,
+      "learning_rate": 0.0002,
+      "loss": 0.491,
+      "step": 31080
+    },
+    {
+      "epoch": 5.026271117937111,
+      "grad_norm": 0.9891864657402039,
+      "learning_rate": 0.0002,
+      "loss": 0.4896,
+      "step": 31090
+    },
+    {
+      "epoch": 5.027887802117856,
+      "grad_norm": 0.9167473912239075,
+      "learning_rate": 0.0002,
+      "loss": 0.4965,
+      "step": 31100
+    },
+    {
+      "epoch": 5.029504486298602,
+      "grad_norm": 1.2259035110473633,
+      "learning_rate": 0.0002,
+      "loss": 0.5098,
+      "step": 31110
+    },
+    {
+      "epoch": 5.031121170479347,
+      "grad_norm": 1.1812787055969238,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 31120
+    },
+    {
+      "epoch": 5.032737854660092,
+      "grad_norm": 1.0890522003173828,
+      "learning_rate": 0.0002,
+      "loss": 0.4725,
+      "step": 31130
+    },
+    {
+      "epoch": 5.034354538840837,
+      "grad_norm": 1.0521091222763062,
+      "learning_rate": 0.0002,
+      "loss": 0.4768,
+      "step": 31140
+    },
+    {
+      "epoch": 5.0359712230215825,
+      "grad_norm": 1.1274569034576416,
+      "learning_rate": 0.0002,
+      "loss": 0.4718,
+      "step": 31150
+    },
+    {
+      "epoch": 5.037587907202328,
+      "grad_norm": 1.140974998474121,
+      "learning_rate": 0.0002,
+      "loss": 0.4604,
+      "step": 31160
+    },
+    {
+      "epoch": 5.039204591383073,
+      "grad_norm": 1.1215609312057495,
+      "learning_rate": 0.0002,
+      "loss": 0.5077,
+      "step": 31170
+    },
+    {
+      "epoch": 5.040821275563818,
+      "grad_norm": 1.0107218027114868,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 31180
+    },
+    {
+      "epoch": 5.042437959744564,
+      "grad_norm": 1.0198770761489868,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 31190
+    },
+    {
+      "epoch": 5.0440546439253096,
+      "grad_norm": 1.1613430976867676,
+      "learning_rate": 0.0002,
+      "loss": 0.5004,
+      "step": 31200
+    },
+    {
+      "epoch": 5.045671328106055,
+      "grad_norm": 0.8555458188056946,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 31210
+    },
+    {
+      "epoch": 5.0472880122868,
+      "grad_norm": 1.0235545635223389,
+      "learning_rate": 0.0002,
+      "loss": 0.4878,
+      "step": 31220
+    },
+    {
+      "epoch": 5.048904696467545,
+      "grad_norm": 1.0228750705718994,
+      "learning_rate": 0.0002,
+      "loss": 0.499,
+      "step": 31230
+    },
+    {
+      "epoch": 5.0505213806482905,
+      "grad_norm": 0.8216419816017151,
+      "learning_rate": 0.0002,
+      "loss": 0.4544,
+      "step": 31240
+    },
+    {
+      "epoch": 5.052138064829036,
+      "grad_norm": 0.925828218460083,
+      "learning_rate": 0.0002,
+      "loss": 0.4947,
+      "step": 31250
+    },
+    {
+      "epoch": 5.053754749009781,
+      "grad_norm": 0.9229369759559631,
+      "learning_rate": 0.0002,
+      "loss": 0.4835,
+      "step": 31260
+    },
+    {
+      "epoch": 5.055371433190526,
+      "grad_norm": 0.9531727433204651,
+      "learning_rate": 0.0002,
+      "loss": 0.5136,
+      "step": 31270
+    },
+    {
+      "epoch": 5.056988117371271,
+      "grad_norm": 0.7738548517227173,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 31280
+    },
+    {
+      "epoch": 5.058604801552017,
+      "grad_norm": 1.0551451444625854,
+      "learning_rate": 0.0002,
+      "loss": 0.5166,
+      "step": 31290
+    },
+    {
+      "epoch": 5.060221485732762,
+      "grad_norm": 0.9782299399375916,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 31300
+    },
+    {
+      "epoch": 5.061838169913507,
+      "grad_norm": 1.0220632553100586,
+      "learning_rate": 0.0002,
+      "loss": 0.4776,
+      "step": 31310
+    },
+    {
+      "epoch": 5.063454854094252,
+      "grad_norm": 0.9808892607688904,
+      "learning_rate": 0.0002,
+      "loss": 0.5117,
+      "step": 31320
+    },
+    {
+      "epoch": 5.065071538274998,
+      "grad_norm": 1.0662003755569458,
+      "learning_rate": 0.0002,
+      "loss": 0.501,
+      "step": 31330
+    },
+    {
+      "epoch": 5.066688222455744,
+      "grad_norm": 1.0036940574645996,
+      "learning_rate": 0.0002,
+      "loss": 0.4844,
+      "step": 31340
+    },
+    {
+      "epoch": 5.068304906636489,
+      "grad_norm": 1.1931052207946777,
+      "learning_rate": 0.0002,
+      "loss": 0.5299,
+      "step": 31350
+    },
+    {
+      "epoch": 5.069921590817234,
+      "grad_norm": 0.9370693564414978,
+      "learning_rate": 0.0002,
+      "loss": 0.4646,
+      "step": 31360
+    },
+    {
+      "epoch": 5.071538274997979,
+      "grad_norm": 0.9589039087295532,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 31370
+    },
+    {
+      "epoch": 5.073154959178725,
+      "grad_norm": 1.0052711963653564,
+      "learning_rate": 0.0002,
+      "loss": 0.4669,
+      "step": 31380
+    },
+    {
+      "epoch": 5.07477164335947,
+      "grad_norm": 0.9991368651390076,
+      "learning_rate": 0.0002,
+      "loss": 0.5283,
+      "step": 31390
+    },
+    {
+      "epoch": 5.076388327540215,
+      "grad_norm": 0.8539695739746094,
+      "learning_rate": 0.0002,
+      "loss": 0.4579,
+      "step": 31400
+    },
+    {
+      "epoch": 5.07800501172096,
+      "grad_norm": 1.048775553703308,
+      "learning_rate": 0.0002,
+      "loss": 0.4609,
+      "step": 31410
+    },
+    {
+      "epoch": 5.0796216959017055,
+      "grad_norm": 0.9983724355697632,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 31420
+    },
+    {
+      "epoch": 5.081238380082451,
+      "grad_norm": 1.0189813375473022,
+      "learning_rate": 0.0002,
+      "loss": 0.4594,
+      "step": 31430
+    },
+    {
+      "epoch": 5.082855064263196,
+      "grad_norm": 0.9781646728515625,
+      "learning_rate": 0.0002,
+      "loss": 0.5449,
+      "step": 31440
+    },
+    {
+      "epoch": 5.084471748443941,
+      "grad_norm": 0.9424566030502319,
+      "learning_rate": 0.0002,
+      "loss": 0.4698,
+      "step": 31450
+    },
+    {
+      "epoch": 5.0860884326246865,
+      "grad_norm": 1.0036484003067017,
+      "learning_rate": 0.0002,
+      "loss": 0.4768,
+      "step": 31460
+    },
+    {
+      "epoch": 5.087705116805432,
+      "grad_norm": 1.0983147621154785,
+      "learning_rate": 0.0002,
+      "loss": 0.487,
+      "step": 31470
+    },
+    {
+      "epoch": 5.089321800986177,
+      "grad_norm": 1.0856730937957764,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 31480
+    },
+    {
+      "epoch": 5.090938485166923,
+      "grad_norm": 1.2191699743270874,
+      "learning_rate": 0.0002,
+      "loss": 0.485,
+      "step": 31490
+    },
+    {
+      "epoch": 5.092555169347668,
+      "grad_norm": 0.939346194267273,
+      "learning_rate": 0.0002,
+      "loss": 0.4936,
+      "step": 31500
+    },
+    {
+      "epoch": 5.0941718535284135,
+      "grad_norm": 0.9730121493339539,
+      "learning_rate": 0.0002,
+      "loss": 0.5107,
+      "step": 31510
+    },
+    {
+      "epoch": 5.095788537709159,
+      "grad_norm": 0.923686146736145,
+      "learning_rate": 0.0002,
+      "loss": 0.4973,
+      "step": 31520
+    },
+    {
+      "epoch": 5.097405221889904,
+      "grad_norm": 1.1734349727630615,
+      "learning_rate": 0.0002,
+      "loss": 0.4906,
+      "step": 31530
+    },
+    {
+      "epoch": 5.099021906070649,
+      "grad_norm": 1.084509015083313,
+      "learning_rate": 0.0002,
+      "loss": 0.5165,
+      "step": 31540
+    },
+    {
+      "epoch": 5.100638590251394,
+      "grad_norm": 1.0144678354263306,
+      "learning_rate": 0.0002,
+      "loss": 0.5078,
+      "step": 31550
+    },
+    {
+      "epoch": 5.10225527443214,
+      "grad_norm": 0.9958019256591797,
+      "learning_rate": 0.0002,
+      "loss": 0.4719,
+      "step": 31560
+    },
+    {
+      "epoch": 5.103871958612885,
+      "grad_norm": 0.8900736570358276,
+      "learning_rate": 0.0002,
+      "loss": 0.4876,
+      "step": 31570
+    },
+    {
+      "epoch": 5.10548864279363,
+      "grad_norm": 1.0921649932861328,
+      "learning_rate": 0.0002,
+      "loss": 0.463,
+      "step": 31580
+    },
+    {
+      "epoch": 5.107105326974375,
+      "grad_norm": 1.1613792181015015,
+      "learning_rate": 0.0002,
+      "loss": 0.5148,
+      "step": 31590
+    },
+    {
+      "epoch": 5.108722011155121,
+      "grad_norm": 0.9211367964744568,
+      "learning_rate": 0.0002,
+      "loss": 0.5055,
+      "step": 31600
+    },
+    {
+      "epoch": 5.110338695335866,
+      "grad_norm": 1.3315813541412354,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 31610
+    },
+    {
+      "epoch": 5.111955379516611,
+      "grad_norm": 1.3765019178390503,
+      "learning_rate": 0.0002,
+      "loss": 0.5336,
+      "step": 31620
+    },
+    {
+      "epoch": 5.113572063697356,
+      "grad_norm": 1.070198893547058,
+      "learning_rate": 0.0002,
+      "loss": 0.4861,
+      "step": 31630
+    },
+    {
+      "epoch": 5.115188747878102,
+      "grad_norm": 0.947631299495697,
+      "learning_rate": 0.0002,
+      "loss": 0.5046,
+      "step": 31640
+    },
+    {
+      "epoch": 5.116805432058848,
+      "grad_norm": 1.0197371244430542,
+      "learning_rate": 0.0002,
+      "loss": 0.5297,
+      "step": 31650
+    },
+    {
+      "epoch": 5.118422116239593,
+      "grad_norm": 0.8647911548614502,
+      "learning_rate": 0.0002,
+      "loss": 0.5014,
+      "step": 31660
+    },
+    {
+      "epoch": 5.120038800420338,
+      "grad_norm": 0.8944075107574463,
+      "learning_rate": 0.0002,
+      "loss": 0.4705,
+      "step": 31670
+    },
+    {
+      "epoch": 5.121655484601083,
+      "grad_norm": 1.124497652053833,
+      "learning_rate": 0.0002,
+      "loss": 0.5175,
+      "step": 31680
+    },
+    {
+      "epoch": 5.123272168781829,
+      "grad_norm": 0.893131673336029,
+      "learning_rate": 0.0002,
+      "loss": 0.5109,
+      "step": 31690
+    },
+    {
+      "epoch": 5.124888852962574,
+      "grad_norm": 1.0122284889221191,
+      "learning_rate": 0.0002,
+      "loss": 0.4937,
+      "step": 31700
+    },
+    {
+      "epoch": 5.126505537143319,
+      "grad_norm": 0.9493719935417175,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 31710
+    },
+    {
+      "epoch": 5.128122221324064,
+      "grad_norm": 0.9700539112091064,
+      "learning_rate": 0.0002,
+      "loss": 0.5031,
+      "step": 31720
+    },
+    {
+      "epoch": 5.1297389055048095,
+      "grad_norm": 1.111677646636963,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 31730
+    },
+    {
+      "epoch": 5.131355589685555,
+      "grad_norm": 0.8204274773597717,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 31740
+    },
+    {
+      "epoch": 5.1329722738663,
+      "grad_norm": 1.1029267311096191,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 31750
+    },
+    {
+      "epoch": 5.134588958047045,
+      "grad_norm": 1.065575122833252,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 31760
+    },
+    {
+      "epoch": 5.13620564222779,
+      "grad_norm": 0.8208706974983215,
+      "learning_rate": 0.0002,
+      "loss": 0.502,
+      "step": 31770
+    },
+    {
+      "epoch": 5.137822326408536,
+      "grad_norm": 1.0520979166030884,
+      "learning_rate": 0.0002,
+      "loss": 0.5352,
+      "step": 31780
+    },
+    {
+      "epoch": 5.139439010589282,
+      "grad_norm": 0.8585538268089294,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 31790
+    },
+    {
+      "epoch": 5.141055694770027,
+      "grad_norm": 1.1491447687149048,
+      "learning_rate": 0.0002,
+      "loss": 0.5159,
+      "step": 31800
+    },
+    {
+      "epoch": 5.142672378950772,
+      "grad_norm": 0.9441081285476685,
+      "learning_rate": 0.0002,
+      "loss": 0.5157,
+      "step": 31810
+    },
+    {
+      "epoch": 5.1442890631315175,
+      "grad_norm": 1.4146889448165894,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 31820
+    },
+    {
+      "epoch": 5.145905747312263,
+      "grad_norm": 1.0326547622680664,
+      "learning_rate": 0.0002,
+      "loss": 0.5159,
+      "step": 31830
+    },
+    {
+      "epoch": 5.147522431493008,
+      "grad_norm": 0.9879202842712402,
+      "learning_rate": 0.0002,
+      "loss": 0.5348,
+      "step": 31840
+    },
+    {
+      "epoch": 5.149139115673753,
+      "grad_norm": 1.0374281406402588,
+      "learning_rate": 0.0002,
+      "loss": 0.5083,
+      "step": 31850
+    },
+    {
+      "epoch": 5.150755799854498,
+      "grad_norm": 1.181229591369629,
+      "learning_rate": 0.0002,
+      "loss": 0.4827,
+      "step": 31860
+    },
+    {
+      "epoch": 5.152372484035244,
+      "grad_norm": 1.2078537940979004,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 31870
+    },
+    {
+      "epoch": 5.153989168215989,
+      "grad_norm": 0.9599190354347229,
+      "learning_rate": 0.0002,
+      "loss": 0.5329,
+      "step": 31880
+    },
+    {
+      "epoch": 5.155605852396734,
+      "grad_norm": 1.0378568172454834,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 31890
+    },
+    {
+      "epoch": 5.157222536577479,
+      "grad_norm": 0.8746536374092102,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 31900
+    },
+    {
+      "epoch": 5.1588392207582245,
+      "grad_norm": 1.0232136249542236,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 31910
+    },
+    {
+      "epoch": 5.16045590493897,
+      "grad_norm": 0.9827565550804138,
+      "learning_rate": 0.0002,
+      "loss": 0.4844,
+      "step": 31920
+    },
+    {
+      "epoch": 5.162072589119716,
+      "grad_norm": 1.342657208442688,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 31930
+    },
+    {
+      "epoch": 5.163689273300461,
+      "grad_norm": 1.18390691280365,
+      "learning_rate": 0.0002,
+      "loss": 0.513,
+      "step": 31940
+    },
+    {
+      "epoch": 5.165305957481206,
+      "grad_norm": 0.996350109577179,
+      "learning_rate": 0.0002,
+      "loss": 0.5267,
+      "step": 31950
+    },
+    {
+      "epoch": 5.166922641661952,
+      "grad_norm": 0.9710391163825989,
+      "learning_rate": 0.0002,
+      "loss": 0.5063,
+      "step": 31960
+    },
+    {
+      "epoch": 5.168539325842697,
+      "grad_norm": 1.0264002084732056,
+      "learning_rate": 0.0002,
+      "loss": 0.5115,
+      "step": 31970
+    },
+    {
+      "epoch": 5.170156010023442,
+      "grad_norm": 1.0028311014175415,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 31980
+    },
+    {
+      "epoch": 5.171772694204187,
+      "grad_norm": 1.1078234910964966,
+      "learning_rate": 0.0002,
+      "loss": 0.5103,
+      "step": 31990
+    },
+    {
+      "epoch": 5.1733893783849325,
+      "grad_norm": 0.9659610390663147,
+      "learning_rate": 0.0002,
+      "loss": 0.495,
+      "step": 32000
+    },
+    {
+      "epoch": 5.175006062565678,
+      "grad_norm": 0.841986894607544,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 32010
+    },
+    {
+      "epoch": 5.176622746746423,
+      "grad_norm": 1.095332384109497,
+      "learning_rate": 0.0002,
+      "loss": 0.48,
+      "step": 32020
+    },
+    {
+      "epoch": 5.178239430927168,
+      "grad_norm": 1.1242377758026123,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 32030
+    },
+    {
+      "epoch": 5.179856115107913,
+      "grad_norm": 0.9872292280197144,
+      "learning_rate": 0.0002,
+      "loss": 0.5573,
+      "step": 32040
+    },
+    {
+      "epoch": 5.181472799288659,
+      "grad_norm": 0.936161994934082,
+      "learning_rate": 0.0002,
+      "loss": 0.48,
+      "step": 32050
+    },
+    {
+      "epoch": 5.183089483469404,
+      "grad_norm": 1.166100025177002,
+      "learning_rate": 0.0002,
+      "loss": 0.5093,
+      "step": 32060
+    },
+    {
+      "epoch": 5.184706167650149,
+      "grad_norm": 1.0764425992965698,
+      "learning_rate": 0.0002,
+      "loss": 0.5438,
+      "step": 32070
+    },
+    {
+      "epoch": 5.186322851830895,
+      "grad_norm": 1.0480051040649414,
+      "learning_rate": 0.0002,
+      "loss": 0.4843,
+      "step": 32080
+    },
+    {
+      "epoch": 5.1879395360116405,
+      "grad_norm": 1.0874916315078735,
+      "learning_rate": 0.0002,
+      "loss": 0.5386,
+      "step": 32090
+    },
+    {
+      "epoch": 5.189556220192386,
+      "grad_norm": 1.0817396640777588,
+      "learning_rate": 0.0002,
+      "loss": 0.4975,
+      "step": 32100
+    },
+    {
+      "epoch": 5.191172904373131,
+      "grad_norm": 1.054111361503601,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 32110
+    },
+    {
+      "epoch": 5.192789588553876,
+      "grad_norm": 0.9655823707580566,
+      "learning_rate": 0.0002,
+      "loss": 0.5229,
+      "step": 32120
+    },
+    {
+      "epoch": 5.194406272734621,
+      "grad_norm": 1.1384109258651733,
+      "learning_rate": 0.0002,
+      "loss": 0.5105,
+      "step": 32130
+    },
+    {
+      "epoch": 5.196022956915367,
+      "grad_norm": 1.0149348974227905,
+      "learning_rate": 0.0002,
+      "loss": 0.5073,
+      "step": 32140
+    },
+    {
+      "epoch": 5.197639641096112,
+      "grad_norm": 1.1084046363830566,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 32150
+    },
+    {
+      "epoch": 5.199256325276857,
+      "grad_norm": 1.1209309101104736,
+      "learning_rate": 0.0002,
+      "loss": 0.4936,
+      "step": 32160
+    },
+    {
+      "epoch": 5.200873009457602,
+      "grad_norm": 1.133089542388916,
+      "learning_rate": 0.0002,
+      "loss": 0.5101,
+      "step": 32170
+    },
+    {
+      "epoch": 5.202489693638348,
+      "grad_norm": 1.0893020629882812,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 32180
+    },
+    {
+      "epoch": 5.204106377819093,
+      "grad_norm": 0.90018630027771,
+      "learning_rate": 0.0002,
+      "loss": 0.4872,
+      "step": 32190
+    },
+    {
+      "epoch": 5.205723061999838,
+      "grad_norm": 0.977622926235199,
+      "learning_rate": 0.0002,
+      "loss": 0.4999,
+      "step": 32200
+    },
+    {
+      "epoch": 5.207339746180583,
+      "grad_norm": 1.2940177917480469,
+      "learning_rate": 0.0002,
+      "loss": 0.5028,
+      "step": 32210
+    },
+    {
+      "epoch": 5.2089564303613285,
+      "grad_norm": 1.2131710052490234,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 32220
+    },
+    {
+      "epoch": 5.210573114542075,
+      "grad_norm": 1.0234841108322144,
+      "learning_rate": 0.0002,
+      "loss": 0.5189,
+      "step": 32230
+    },
+    {
+      "epoch": 5.21218979872282,
+      "grad_norm": 1.157975435256958,
+      "learning_rate": 0.0002,
+      "loss": 0.5424,
+      "step": 32240
+    },
+    {
+      "epoch": 5.213806482903565,
+      "grad_norm": 1.0381282567977905,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 32250
+    },
+    {
+      "epoch": 5.21542316708431,
+      "grad_norm": 1.0125395059585571,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 32260
+    },
+    {
+      "epoch": 5.2170398512650555,
+      "grad_norm": 1.272691011428833,
+      "learning_rate": 0.0002,
+      "loss": 0.5216,
+      "step": 32270
+    },
+    {
+      "epoch": 5.218656535445801,
+      "grad_norm": 1.0061250925064087,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 32280
+    },
+    {
+      "epoch": 5.220273219626546,
+      "grad_norm": 0.9752234816551208,
+      "learning_rate": 0.0002,
+      "loss": 0.4739,
+      "step": 32290
+    },
+    {
+      "epoch": 5.221889903807291,
+      "grad_norm": 1.1193140745162964,
+      "learning_rate": 0.0002,
+      "loss": 0.5471,
+      "step": 32300
+    },
+    {
+      "epoch": 5.2235065879880365,
+      "grad_norm": 1.0126434564590454,
+      "learning_rate": 0.0002,
+      "loss": 0.4976,
+      "step": 32310
+    },
+    {
+      "epoch": 5.225123272168782,
+      "grad_norm": 1.4338394403457642,
+      "learning_rate": 0.0002,
+      "loss": 0.5257,
+      "step": 32320
+    },
+    {
+      "epoch": 5.226739956349527,
+      "grad_norm": 1.004101276397705,
+      "learning_rate": 0.0002,
+      "loss": 0.5235,
+      "step": 32330
+    },
+    {
+      "epoch": 5.228356640530272,
+      "grad_norm": 0.8744166493415833,
+      "learning_rate": 0.0002,
+      "loss": 0.5091,
+      "step": 32340
+    },
+    {
+      "epoch": 5.229973324711017,
+      "grad_norm": 1.0165376663208008,
+      "learning_rate": 0.0002,
+      "loss": 0.5388,
+      "step": 32350
+    },
+    {
+      "epoch": 5.231590008891763,
+      "grad_norm": 0.8635954260826111,
+      "learning_rate": 0.0002,
+      "loss": 0.5469,
+      "step": 32360
+    },
+    {
+      "epoch": 5.233206693072509,
+      "grad_norm": 1.1392399072647095,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 32370
+    },
+    {
+      "epoch": 5.234823377253254,
+      "grad_norm": 1.0202113389968872,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 32380
+    },
+    {
+      "epoch": 5.236440061433999,
+      "grad_norm": 1.0417983531951904,
+      "learning_rate": 0.0002,
+      "loss": 0.4983,
+      "step": 32390
+    },
+    {
+      "epoch": 5.238056745614744,
+      "grad_norm": 0.8729333877563477,
+      "learning_rate": 0.0002,
+      "loss": 0.507,
+      "step": 32400
+    },
+    {
+      "epoch": 5.23967342979549,
+      "grad_norm": 1.1626229286193848,
+      "learning_rate": 0.0002,
+      "loss": 0.5426,
+      "step": 32410
+    },
+    {
+      "epoch": 5.241290113976235,
+      "grad_norm": 0.9086161851882935,
+      "learning_rate": 0.0002,
+      "loss": 0.5355,
+      "step": 32420
+    },
+    {
+      "epoch": 5.24290679815698,
+      "grad_norm": 1.3999892473220825,
+      "learning_rate": 0.0002,
+      "loss": 0.4927,
+      "step": 32430
+    },
+    {
+      "epoch": 5.244523482337725,
+      "grad_norm": 1.0356311798095703,
+      "learning_rate": 0.0002,
+      "loss": 0.4795,
+      "step": 32440
+    },
+    {
+      "epoch": 5.246140166518471,
+      "grad_norm": 0.9655531644821167,
+      "learning_rate": 0.0002,
+      "loss": 0.5035,
+      "step": 32450
+    },
+    {
+      "epoch": 5.247756850699216,
+      "grad_norm": 1.0411828756332397,
+      "learning_rate": 0.0002,
+      "loss": 0.5166,
+      "step": 32460
+    },
+    {
+      "epoch": 5.249373534879961,
+      "grad_norm": 1.1199816465377808,
+      "learning_rate": 0.0002,
+      "loss": 0.5141,
+      "step": 32470
+    },
+    {
+      "epoch": 5.250990219060706,
+      "grad_norm": 1.260321855545044,
+      "learning_rate": 0.0002,
+      "loss": 0.4864,
+      "step": 32480
+    },
+    {
+      "epoch": 5.2526069032414515,
+      "grad_norm": 1.2950857877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.4893,
+      "step": 32490
+    },
+    {
+      "epoch": 5.254223587422197,
+      "grad_norm": 0.8982820510864258,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 32500
+    },
+    {
+      "epoch": 5.255840271602942,
+      "grad_norm": 0.8512987494468689,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 32510
+    },
+    {
+      "epoch": 5.257456955783688,
+      "grad_norm": 1.067443609237671,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 32520
+    },
+    {
+      "epoch": 5.259073639964433,
+      "grad_norm": 1.0957417488098145,
+      "learning_rate": 0.0002,
+      "loss": 0.4928,
+      "step": 32530
+    },
+    {
+      "epoch": 5.260690324145179,
+      "grad_norm": 1.4161807298660278,
+      "learning_rate": 0.0002,
+      "loss": 0.5169,
+      "step": 32540
+    },
+    {
+      "epoch": 5.262307008325924,
+      "grad_norm": 1.2264093160629272,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 32550
+    },
+    {
+      "epoch": 5.263923692506669,
+      "grad_norm": 1.0015931129455566,
+      "learning_rate": 0.0002,
+      "loss": 0.5221,
+      "step": 32560
+    },
+    {
+      "epoch": 5.265540376687414,
+      "grad_norm": 1.0743094682693481,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 32570
+    },
+    {
+      "epoch": 5.2671570608681595,
+      "grad_norm": 1.1386840343475342,
+      "learning_rate": 0.0002,
+      "loss": 0.5289,
+      "step": 32580
+    },
+    {
+      "epoch": 5.268773745048905,
+      "grad_norm": 1.0093860626220703,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 32590
+    },
+    {
+      "epoch": 5.27039042922965,
+      "grad_norm": 0.9593744874000549,
+      "learning_rate": 0.0002,
+      "loss": 0.5175,
+      "step": 32600
+    },
+    {
+      "epoch": 5.272007113410395,
+      "grad_norm": 1.146021842956543,
+      "learning_rate": 0.0002,
+      "loss": 0.528,
+      "step": 32610
+    },
+    {
+      "epoch": 5.27362379759114,
+      "grad_norm": 0.9579031467437744,
+      "learning_rate": 0.0002,
+      "loss": 0.4983,
+      "step": 32620
+    },
+    {
+      "epoch": 5.275240481771886,
+      "grad_norm": 1.0548793077468872,
+      "learning_rate": 0.0002,
+      "loss": 0.5376,
+      "step": 32630
+    },
+    {
+      "epoch": 5.276857165952631,
+      "grad_norm": 1.0380561351776123,
+      "learning_rate": 0.0002,
+      "loss": 0.5267,
+      "step": 32640
+    },
+    {
+      "epoch": 5.278473850133376,
+      "grad_norm": 1.2119969129562378,
+      "learning_rate": 0.0002,
+      "loss": 0.5182,
+      "step": 32650
+    },
+    {
+      "epoch": 5.280090534314121,
+      "grad_norm": 1.0507797002792358,
+      "learning_rate": 0.0002,
+      "loss": 0.5298,
+      "step": 32660
+    },
+    {
+      "epoch": 5.2817072184948675,
+      "grad_norm": 1.0185176134109497,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 32670
+    },
+    {
+      "epoch": 5.283323902675613,
+      "grad_norm": 1.2358098030090332,
+      "learning_rate": 0.0002,
+      "loss": 0.4904,
+      "step": 32680
+    },
+    {
+      "epoch": 5.284940586856358,
+      "grad_norm": 0.7937114238739014,
+      "learning_rate": 0.0002,
+      "loss": 0.5169,
+      "step": 32690
+    },
+    {
+      "epoch": 5.286557271037103,
+      "grad_norm": 0.9825124740600586,
+      "learning_rate": 0.0002,
+      "loss": 0.495,
+      "step": 32700
+    },
+    {
+      "epoch": 5.288173955217848,
+      "grad_norm": 1.2059301137924194,
+      "learning_rate": 0.0002,
+      "loss": 0.5149,
+      "step": 32710
+    },
+    {
+      "epoch": 5.289790639398594,
+      "grad_norm": 1.0828571319580078,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 32720
+    },
+    {
+      "epoch": 5.291407323579339,
+      "grad_norm": 1.0129735469818115,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 32730
+    },
+    {
+      "epoch": 5.293024007760084,
+      "grad_norm": 1.0591634511947632,
+      "learning_rate": 0.0002,
+      "loss": 0.5216,
+      "step": 32740
+    },
+    {
+      "epoch": 5.294640691940829,
+      "grad_norm": 0.9256815910339355,
+      "learning_rate": 0.0002,
+      "loss": 0.522,
+      "step": 32750
+    },
+    {
+      "epoch": 5.2962573761215745,
+      "grad_norm": 1.0928633213043213,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 32760
+    },
+    {
+      "epoch": 5.29787406030232,
+      "grad_norm": 0.9415594935417175,
+      "learning_rate": 0.0002,
+      "loss": 0.5093,
+      "step": 32770
+    },
+    {
+      "epoch": 5.299490744483065,
+      "grad_norm": 1.141316294670105,
+      "learning_rate": 0.0002,
+      "loss": 0.5252,
+      "step": 32780
+    },
+    {
+      "epoch": 5.30110742866381,
+      "grad_norm": 1.0646510124206543,
+      "learning_rate": 0.0002,
+      "loss": 0.4837,
+      "step": 32790
+    },
+    {
+      "epoch": 5.3027241128445555,
+      "grad_norm": 1.189661979675293,
+      "learning_rate": 0.0002,
+      "loss": 0.5547,
+      "step": 32800
+    },
+    {
+      "epoch": 5.304340797025301,
+      "grad_norm": 0.9568731188774109,
+      "learning_rate": 0.0002,
+      "loss": 0.5664,
+      "step": 32810
+    },
+    {
+      "epoch": 5.305957481206047,
+      "grad_norm": 1.1556824445724487,
+      "learning_rate": 0.0002,
+      "loss": 0.5344,
+      "step": 32820
+    },
+    {
+      "epoch": 5.307574165386792,
+      "grad_norm": 0.9353463649749756,
+      "learning_rate": 0.0002,
+      "loss": 0.4894,
+      "step": 32830
+    },
+    {
+      "epoch": 5.309190849567537,
+      "grad_norm": 1.1208295822143555,
+      "learning_rate": 0.0002,
+      "loss": 0.5052,
+      "step": 32840
+    },
+    {
+      "epoch": 5.3108075337482825,
+      "grad_norm": 1.0894153118133545,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 32850
+    },
+    {
+      "epoch": 5.312424217929028,
+      "grad_norm": 1.090329647064209,
+      "learning_rate": 0.0002,
+      "loss": 0.5046,
+      "step": 32860
+    },
+    {
+      "epoch": 5.314040902109773,
+      "grad_norm": 1.0781712532043457,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 32870
+    },
+    {
+      "epoch": 5.315657586290518,
+      "grad_norm": 1.1785295009613037,
+      "learning_rate": 0.0002,
+      "loss": 0.57,
+      "step": 32880
+    },
+    {
+      "epoch": 5.317274270471263,
+      "grad_norm": 1.0406851768493652,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 32890
+    },
+    {
+      "epoch": 5.318890954652009,
+      "grad_norm": 1.0982953310012817,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 32900
+    },
+    {
+      "epoch": 5.320507638832754,
+      "grad_norm": 1.2969383001327515,
+      "learning_rate": 0.0002,
+      "loss": 0.4944,
+      "step": 32910
+    },
+    {
+      "epoch": 5.322124323013499,
+      "grad_norm": 0.9687288999557495,
+      "learning_rate": 0.0002,
+      "loss": 0.4786,
+      "step": 32920
+    },
+    {
+      "epoch": 5.323741007194244,
+      "grad_norm": 1.136760950088501,
+      "learning_rate": 0.0002,
+      "loss": 0.5286,
+      "step": 32930
+    },
+    {
+      "epoch": 5.32535769137499,
+      "grad_norm": 1.3045495748519897,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 32940
+    },
+    {
+      "epoch": 5.326974375555735,
+      "grad_norm": 1.221675992012024,
+      "learning_rate": 0.0002,
+      "loss": 0.5413,
+      "step": 32950
+    },
+    {
+      "epoch": 5.32859105973648,
+      "grad_norm": 1.1380633115768433,
+      "learning_rate": 0.0002,
+      "loss": 0.4999,
+      "step": 32960
+    },
+    {
+      "epoch": 5.330207743917226,
+      "grad_norm": 1.1065956354141235,
+      "learning_rate": 0.0002,
+      "loss": 0.5037,
+      "step": 32970
+    },
+    {
+      "epoch": 5.331824428097971,
+      "grad_norm": 1.0187175273895264,
+      "learning_rate": 0.0002,
+      "loss": 0.4913,
+      "step": 32980
+    },
+    {
+      "epoch": 5.333441112278717,
+      "grad_norm": 0.9077118039131165,
+      "learning_rate": 0.0002,
+      "loss": 0.5234,
+      "step": 32990
+    },
+    {
+      "epoch": 5.335057796459462,
+      "grad_norm": 1.0092815160751343,
+      "learning_rate": 0.0002,
+      "loss": 0.5071,
+      "step": 33000
+    },
+    {
+      "epoch": 5.336674480640207,
+      "grad_norm": 1.0168777704238892,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 33010
+    },
+    {
+      "epoch": 5.338291164820952,
+      "grad_norm": 0.996161937713623,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 33020
+    },
+    {
+      "epoch": 5.339907849001698,
+      "grad_norm": 0.794463038444519,
+      "learning_rate": 0.0002,
+      "loss": 0.5024,
+      "step": 33030
+    },
+    {
+      "epoch": 5.341524533182443,
+      "grad_norm": 0.9750674962997437,
+      "learning_rate": 0.0002,
+      "loss": 0.5112,
+      "step": 33040
+    },
+    {
+      "epoch": 5.343141217363188,
+      "grad_norm": 1.2770029306411743,
+      "learning_rate": 0.0002,
+      "loss": 0.528,
+      "step": 33050
+    },
+    {
+      "epoch": 5.344757901543933,
+      "grad_norm": 1.1500186920166016,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 33060
+    },
+    {
+      "epoch": 5.3463745857246785,
+      "grad_norm": 1.0726377964019775,
+      "learning_rate": 0.0002,
+      "loss": 0.4906,
+      "step": 33070
+    },
+    {
+      "epoch": 5.347991269905424,
+      "grad_norm": 0.9314153790473938,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 33080
+    },
+    {
+      "epoch": 5.349607954086169,
+      "grad_norm": 1.344988465309143,
+      "learning_rate": 0.0002,
+      "loss": 0.5434,
+      "step": 33090
+    },
+    {
+      "epoch": 5.351224638266914,
+      "grad_norm": 0.863196611404419,
+      "learning_rate": 0.0002,
+      "loss": 0.4874,
+      "step": 33100
+    },
+    {
+      "epoch": 5.352841322447659,
+      "grad_norm": 1.128100037574768,
+      "learning_rate": 0.0002,
+      "loss": 0.534,
+      "step": 33110
+    },
+    {
+      "epoch": 5.3544580066284055,
+      "grad_norm": 1.1673583984375,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 33120
+    },
+    {
+      "epoch": 5.356074690809151,
+      "grad_norm": 0.9416789412498474,
+      "learning_rate": 0.0002,
+      "loss": 0.4787,
+      "step": 33130
+    },
+    {
+      "epoch": 5.357691374989896,
+      "grad_norm": 1.1855236291885376,
+      "learning_rate": 0.0002,
+      "loss": 0.5155,
+      "step": 33140
+    },
+    {
+      "epoch": 5.359308059170641,
+      "grad_norm": 1.0415170192718506,
+      "learning_rate": 0.0002,
+      "loss": 0.515,
+      "step": 33150
+    },
+    {
+      "epoch": 5.3609247433513865,
+      "grad_norm": 0.9953004121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.545,
+      "step": 33160
+    },
+    {
+      "epoch": 5.362541427532132,
+      "grad_norm": 0.96138596534729,
+      "learning_rate": 0.0002,
+      "loss": 0.5305,
+      "step": 33170
+    },
+    {
+      "epoch": 5.364158111712877,
+      "grad_norm": 1.341979742050171,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 33180
+    },
+    {
+      "epoch": 5.365774795893622,
+      "grad_norm": 1.0136911869049072,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 33190
+    },
+    {
+      "epoch": 5.367391480074367,
+      "grad_norm": 0.8685575127601624,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 33200
+    },
+    {
+      "epoch": 5.369008164255113,
+      "grad_norm": 0.8833574652671814,
+      "learning_rate": 0.0002,
+      "loss": 0.5146,
+      "step": 33210
+    },
+    {
+      "epoch": 5.370624848435858,
+      "grad_norm": 0.9123612642288208,
+      "learning_rate": 0.0002,
+      "loss": 0.4982,
+      "step": 33220
+    },
+    {
+      "epoch": 5.372241532616603,
+      "grad_norm": 1.2720599174499512,
+      "learning_rate": 0.0002,
+      "loss": 0.5047,
+      "step": 33230
+    },
+    {
+      "epoch": 5.373858216797348,
+      "grad_norm": 1.0596648454666138,
+      "learning_rate": 0.0002,
+      "loss": 0.5175,
+      "step": 33240
+    },
+    {
+      "epoch": 5.3754749009780936,
+      "grad_norm": 1.119701623916626,
+      "learning_rate": 0.0002,
+      "loss": 0.5284,
+      "step": 33250
+    },
+    {
+      "epoch": 5.377091585158839,
+      "grad_norm": 1.3000061511993408,
+      "learning_rate": 0.0002,
+      "loss": 0.5217,
+      "step": 33260
+    },
+    {
+      "epoch": 5.378708269339585,
+      "grad_norm": 1.083891749382019,
+      "learning_rate": 0.0002,
+      "loss": 0.5125,
+      "step": 33270
+    },
+    {
+      "epoch": 5.38032495352033,
+      "grad_norm": 0.9402718544006348,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 33280
+    },
+    {
+      "epoch": 5.381941637701075,
+      "grad_norm": 1.3376892805099487,
+      "learning_rate": 0.0002,
+      "loss": 0.5559,
+      "step": 33290
+    },
+    {
+      "epoch": 5.383558321881821,
+      "grad_norm": 1.1600074768066406,
+      "learning_rate": 0.0002,
+      "loss": 0.5193,
+      "step": 33300
+    },
+    {
+      "epoch": 5.385175006062566,
+      "grad_norm": 1.1449427604675293,
+      "learning_rate": 0.0002,
+      "loss": 0.4907,
+      "step": 33310
+    },
+    {
+      "epoch": 5.386791690243311,
+      "grad_norm": 1.3118891716003418,
+      "learning_rate": 0.0002,
+      "loss": 0.5449,
+      "step": 33320
+    },
+    {
+      "epoch": 5.388408374424056,
+      "grad_norm": 0.743449866771698,
+      "learning_rate": 0.0002,
+      "loss": 0.547,
+      "step": 33330
+    },
+    {
+      "epoch": 5.3900250586048015,
+      "grad_norm": 0.9358304142951965,
+      "learning_rate": 0.0002,
+      "loss": 0.5555,
+      "step": 33340
+    },
+    {
+      "epoch": 5.391641742785547,
+      "grad_norm": 1.0447142124176025,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 33350
+    },
+    {
+      "epoch": 5.393258426966292,
+      "grad_norm": 1.1088626384735107,
+      "learning_rate": 0.0002,
+      "loss": 0.5106,
+      "step": 33360
+    },
+    {
+      "epoch": 5.394875111147037,
+      "grad_norm": 1.1267958879470825,
+      "learning_rate": 0.0002,
+      "loss": 0.4929,
+      "step": 33370
+    },
+    {
+      "epoch": 5.3964917953277824,
+      "grad_norm": 0.9709370136260986,
+      "learning_rate": 0.0002,
+      "loss": 0.5165,
+      "step": 33380
+    },
+    {
+      "epoch": 5.398108479508528,
+      "grad_norm": 1.0939103364944458,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 33390
+    },
+    {
+      "epoch": 5.399725163689273,
+      "grad_norm": 0.9559304714202881,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 33400
+    },
+    {
+      "epoch": 5.401341847870018,
+      "grad_norm": 1.199580430984497,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 33410
+    },
+    {
+      "epoch": 5.402958532050764,
+      "grad_norm": 0.9097000360488892,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 33420
+    },
+    {
+      "epoch": 5.4045752162315095,
+      "grad_norm": 1.1940981149673462,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 33430
+    },
+    {
+      "epoch": 5.406191900412255,
+      "grad_norm": 1.0530916452407837,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 33440
+    },
+    {
+      "epoch": 5.407808584593,
+      "grad_norm": 1.0482549667358398,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 33450
+    },
+    {
+      "epoch": 5.409425268773745,
+      "grad_norm": 1.2524714469909668,
+      "learning_rate": 0.0002,
+      "loss": 0.501,
+      "step": 33460
+    },
+    {
+      "epoch": 5.41104195295449,
+      "grad_norm": 1.1091666221618652,
+      "learning_rate": 0.0002,
+      "loss": 0.5597,
+      "step": 33470
+    },
+    {
+      "epoch": 5.412658637135236,
+      "grad_norm": 0.9981587529182434,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 33480
+    },
+    {
+      "epoch": 5.414275321315981,
+      "grad_norm": 1.016681432723999,
+      "learning_rate": 0.0002,
+      "loss": 0.4977,
+      "step": 33490
+    },
+    {
+      "epoch": 5.415892005496726,
+      "grad_norm": 1.1456854343414307,
+      "learning_rate": 0.0002,
+      "loss": 0.5388,
+      "step": 33500
+    },
+    {
+      "epoch": 5.417508689677471,
+      "grad_norm": 1.1454259157180786,
+      "learning_rate": 0.0002,
+      "loss": 0.5292,
+      "step": 33510
+    },
+    {
+      "epoch": 5.419125373858217,
+      "grad_norm": 0.9858416318893433,
+      "learning_rate": 0.0002,
+      "loss": 0.5061,
+      "step": 33520
+    },
+    {
+      "epoch": 5.420742058038962,
+      "grad_norm": 0.9764766693115234,
+      "learning_rate": 0.0002,
+      "loss": 0.5139,
+      "step": 33530
+    },
+    {
+      "epoch": 5.422358742219707,
+      "grad_norm": 1.199920892715454,
+      "learning_rate": 0.0002,
+      "loss": 0.5518,
+      "step": 33540
+    },
+    {
+      "epoch": 5.423975426400452,
+      "grad_norm": 1.3107370138168335,
+      "learning_rate": 0.0002,
+      "loss": 0.5182,
+      "step": 33550
+    },
+    {
+      "epoch": 5.4255921105811975,
+      "grad_norm": 0.9637970328330994,
+      "learning_rate": 0.0002,
+      "loss": 0.5149,
+      "step": 33560
+    },
+    {
+      "epoch": 5.427208794761944,
+      "grad_norm": 1.023359775543213,
+      "learning_rate": 0.0002,
+      "loss": 0.526,
+      "step": 33570
+    },
+    {
+      "epoch": 5.428825478942689,
+      "grad_norm": 1.060417652130127,
+      "learning_rate": 0.0002,
+      "loss": 0.5206,
+      "step": 33580
+    },
+    {
+      "epoch": 5.430442163123434,
+      "grad_norm": 0.9971120953559875,
+      "learning_rate": 0.0002,
+      "loss": 0.5052,
+      "step": 33590
+    },
+    {
+      "epoch": 5.432058847304179,
+      "grad_norm": 0.9213743209838867,
+      "learning_rate": 0.0002,
+      "loss": 0.5044,
+      "step": 33600
+    },
+    {
+      "epoch": 5.4336755314849245,
+      "grad_norm": 1.1512309312820435,
+      "learning_rate": 0.0002,
+      "loss": 0.5714,
+      "step": 33610
+    },
+    {
+      "epoch": 5.43529221566567,
+      "grad_norm": 1.2198847532272339,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 33620
+    },
+    {
+      "epoch": 5.436908899846415,
+      "grad_norm": 1.0329595804214478,
+      "learning_rate": 0.0002,
+      "loss": 0.5237,
+      "step": 33630
+    },
+    {
+      "epoch": 5.43852558402716,
+      "grad_norm": 1.1075750589370728,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 33640
+    },
+    {
+      "epoch": 5.4401422682079055,
+      "grad_norm": 1.006342887878418,
+      "learning_rate": 0.0002,
+      "loss": 0.5295,
+      "step": 33650
+    },
+    {
+      "epoch": 5.441758952388651,
+      "grad_norm": 0.9179885983467102,
+      "learning_rate": 0.0002,
+      "loss": 0.5394,
+      "step": 33660
+    },
+    {
+      "epoch": 5.443375636569396,
+      "grad_norm": 1.2799493074417114,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 33670
+    },
+    {
+      "epoch": 5.444992320750141,
+      "grad_norm": 1.1153863668441772,
+      "learning_rate": 0.0002,
+      "loss": 0.5426,
+      "step": 33680
+    },
+    {
+      "epoch": 5.446609004930886,
+      "grad_norm": 1.0681028366088867,
+      "learning_rate": 0.0002,
+      "loss": 0.5087,
+      "step": 33690
+    },
+    {
+      "epoch": 5.448225689111632,
+      "grad_norm": 0.9788817167282104,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 33700
+    },
+    {
+      "epoch": 5.449842373292377,
+      "grad_norm": 0.8481608629226685,
+      "learning_rate": 0.0002,
+      "loss": 0.5308,
+      "step": 33710
+    },
+    {
+      "epoch": 5.451459057473123,
+      "grad_norm": 1.113756537437439,
+      "learning_rate": 0.0002,
+      "loss": 0.5225,
+      "step": 33720
+    },
+    {
+      "epoch": 5.453075741653868,
+      "grad_norm": 0.8425475358963013,
+      "learning_rate": 0.0002,
+      "loss": 0.5213,
+      "step": 33730
+    },
+    {
+      "epoch": 5.4546924258346134,
+      "grad_norm": 1.0852208137512207,
+      "learning_rate": 0.0002,
+      "loss": 0.571,
+      "step": 33740
+    },
+    {
+      "epoch": 5.456309110015359,
+      "grad_norm": 1.1664748191833496,
+      "learning_rate": 0.0002,
+      "loss": 0.5535,
+      "step": 33750
+    },
+    {
+      "epoch": 5.457925794196104,
+      "grad_norm": 1.217241644859314,
+      "learning_rate": 0.0002,
+      "loss": 0.5419,
+      "step": 33760
+    },
+    {
+      "epoch": 5.459542478376849,
+      "grad_norm": 1.1572928428649902,
+      "learning_rate": 0.0002,
+      "loss": 0.5351,
+      "step": 33770
+    },
+    {
+      "epoch": 5.461159162557594,
+      "grad_norm": 1.0437318086624146,
+      "learning_rate": 0.0002,
+      "loss": 0.5161,
+      "step": 33780
+    },
+    {
+      "epoch": 5.46277584673834,
+      "grad_norm": 0.9807571768760681,
+      "learning_rate": 0.0002,
+      "loss": 0.5266,
+      "step": 33790
+    },
+    {
+      "epoch": 5.464392530919085,
+      "grad_norm": 1.1436342000961304,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 33800
+    },
+    {
+      "epoch": 5.46600921509983,
+      "grad_norm": 1.1004794836044312,
+      "learning_rate": 0.0002,
+      "loss": 0.5338,
+      "step": 33810
+    },
+    {
+      "epoch": 5.467625899280575,
+      "grad_norm": 1.2130268812179565,
+      "learning_rate": 0.0002,
+      "loss": 0.4868,
+      "step": 33820
+    },
+    {
+      "epoch": 5.4692425834613205,
+      "grad_norm": 1.3154419660568237,
+      "learning_rate": 0.0002,
+      "loss": 0.516,
+      "step": 33830
+    },
+    {
+      "epoch": 5.470859267642066,
+      "grad_norm": 0.7934383749961853,
+      "learning_rate": 0.0002,
+      "loss": 0.4934,
+      "step": 33840
+    },
+    {
+      "epoch": 5.472475951822812,
+      "grad_norm": 0.7838410139083862,
+      "learning_rate": 0.0002,
+      "loss": 0.5133,
+      "step": 33850
+    },
+    {
+      "epoch": 5.474092636003557,
+      "grad_norm": 1.0415139198303223,
+      "learning_rate": 0.0002,
+      "loss": 0.4926,
+      "step": 33860
+    },
+    {
+      "epoch": 5.475709320184302,
+      "grad_norm": 0.9213164448738098,
+      "learning_rate": 0.0002,
+      "loss": 0.5323,
+      "step": 33870
+    },
+    {
+      "epoch": 5.477326004365048,
+      "grad_norm": 1.0364776849746704,
+      "learning_rate": 0.0002,
+      "loss": 0.5125,
+      "step": 33880
+    },
+    {
+      "epoch": 5.478942688545793,
+      "grad_norm": 0.9994072318077087,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 33890
+    },
+    {
+      "epoch": 5.480559372726538,
+      "grad_norm": 1.196730136871338,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 33900
+    },
+    {
+      "epoch": 5.482176056907283,
+      "grad_norm": 0.9955780506134033,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 33910
+    },
+    {
+      "epoch": 5.4837927410880285,
+      "grad_norm": 1.168188214302063,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 33920
+    },
+    {
+      "epoch": 5.485409425268774,
+      "grad_norm": 1.1816450357437134,
+      "learning_rate": 0.0002,
+      "loss": 0.5548,
+      "step": 33930
+    },
+    {
+      "epoch": 5.487026109449519,
+      "grad_norm": 1.079715609550476,
+      "learning_rate": 0.0002,
+      "loss": 0.5535,
+      "step": 33940
+    },
+    {
+      "epoch": 5.488642793630264,
+      "grad_norm": 1.153850793838501,
+      "learning_rate": 0.0002,
+      "loss": 0.5262,
+      "step": 33950
+    },
+    {
+      "epoch": 5.490259477811009,
+      "grad_norm": 1.0207297801971436,
+      "learning_rate": 0.0002,
+      "loss": 0.5248,
+      "step": 33960
+    },
+    {
+      "epoch": 5.491876161991755,
+      "grad_norm": 1.1290855407714844,
+      "learning_rate": 0.0002,
+      "loss": 0.5142,
+      "step": 33970
+    },
+    {
+      "epoch": 5.4934928461725,
+      "grad_norm": 1.068058967590332,
+      "learning_rate": 0.0002,
+      "loss": 0.5168,
+      "step": 33980
+    },
+    {
+      "epoch": 5.495109530353245,
+      "grad_norm": 0.9789979457855225,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 33990
+    },
+    {
+      "epoch": 5.496726214533991,
+      "grad_norm": 0.9696692824363708,
+      "learning_rate": 0.0002,
+      "loss": 0.5113,
+      "step": 34000
+    },
+    {
+      "epoch": 5.4983428987147365,
+      "grad_norm": 1.0539981126785278,
+      "learning_rate": 0.0002,
+      "loss": 0.5413,
+      "step": 34010
+    },
+    {
+      "epoch": 5.499959582895482,
+      "grad_norm": 1.0249929428100586,
+      "learning_rate": 0.0002,
+      "loss": 0.5783,
+      "step": 34020
+    },
+    {
+      "epoch": 5.501576267076227,
+      "grad_norm": 0.9577504992485046,
+      "learning_rate": 0.0002,
+      "loss": 0.4888,
+      "step": 34030
+    },
+    {
+      "epoch": 5.503192951256972,
+      "grad_norm": 1.0963513851165771,
+      "learning_rate": 0.0002,
+      "loss": 0.5291,
+      "step": 34040
+    },
+    {
+      "epoch": 5.504809635437717,
+      "grad_norm": 0.8339345455169678,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 34050
+    },
+    {
+      "epoch": 5.506426319618463,
+      "grad_norm": 1.0138782262802124,
+      "learning_rate": 0.0002,
+      "loss": 0.5191,
+      "step": 34060
+    },
+    {
+      "epoch": 5.508043003799208,
+      "grad_norm": 1.0180109739303589,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 34070
+    },
+    {
+      "epoch": 5.509659687979953,
+      "grad_norm": 1.2790818214416504,
+      "learning_rate": 0.0002,
+      "loss": 0.5083,
+      "step": 34080
+    },
+    {
+      "epoch": 5.511276372160698,
+      "grad_norm": 1.428247332572937,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 34090
+    },
+    {
+      "epoch": 5.5128930563414436,
+      "grad_norm": 1.0926059484481812,
+      "learning_rate": 0.0002,
+      "loss": 0.5291,
+      "step": 34100
+    },
+    {
+      "epoch": 5.514509740522189,
+      "grad_norm": 1.2353343963623047,
+      "learning_rate": 0.0002,
+      "loss": 0.5665,
+      "step": 34110
+    },
+    {
+      "epoch": 5.516126424702934,
+      "grad_norm": 0.935587465763092,
+      "learning_rate": 0.0002,
+      "loss": 0.5331,
+      "step": 34120
+    },
+    {
+      "epoch": 5.517743108883679,
+      "grad_norm": 0.9767586588859558,
+      "learning_rate": 0.0002,
+      "loss": 0.5512,
+      "step": 34130
+    },
+    {
+      "epoch": 5.5193597930644245,
+      "grad_norm": 1.1660610437393188,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 34140
+    },
+    {
+      "epoch": 5.520976477245171,
+      "grad_norm": 0.9828870892524719,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 34150
+    },
+    {
+      "epoch": 5.522593161425916,
+      "grad_norm": 1.0097278356552124,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 34160
+    },
+    {
+      "epoch": 5.524209845606661,
+      "grad_norm": 1.1766167879104614,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 34170
+    },
+    {
+      "epoch": 5.525826529787406,
+      "grad_norm": 0.982292115688324,
+      "learning_rate": 0.0002,
+      "loss": 0.5258,
+      "step": 34180
+    },
+    {
+      "epoch": 5.5274432139681515,
+      "grad_norm": 1.0744609832763672,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 34190
+    },
+    {
+      "epoch": 5.529059898148897,
+      "grad_norm": 1.3831160068511963,
+      "learning_rate": 0.0002,
+      "loss": 0.5469,
+      "step": 34200
+    },
+    {
+      "epoch": 5.530676582329642,
+      "grad_norm": 1.074771761894226,
+      "learning_rate": 0.0002,
+      "loss": 0.5819,
+      "step": 34210
+    },
+    {
+      "epoch": 5.532293266510387,
+      "grad_norm": 1.016652226448059,
+      "learning_rate": 0.0002,
+      "loss": 0.5399,
+      "step": 34220
+    },
+    {
+      "epoch": 5.5339099506911325,
+      "grad_norm": 1.2231552600860596,
+      "learning_rate": 0.0002,
+      "loss": 0.5158,
+      "step": 34230
+    },
+    {
+      "epoch": 5.535526634871878,
+      "grad_norm": 0.8051198720932007,
+      "learning_rate": 0.0002,
+      "loss": 0.5091,
+      "step": 34240
+    },
+    {
+      "epoch": 5.537143319052623,
+      "grad_norm": 1.1779674291610718,
+      "learning_rate": 0.0002,
+      "loss": 0.5583,
+      "step": 34250
+    },
+    {
+      "epoch": 5.538760003233368,
+      "grad_norm": 1.2468291521072388,
+      "learning_rate": 0.0002,
+      "loss": 0.5044,
+      "step": 34260
+    },
+    {
+      "epoch": 5.540376687414113,
+      "grad_norm": 1.14818274974823,
+      "learning_rate": 0.0002,
+      "loss": 0.523,
+      "step": 34270
+    },
+    {
+      "epoch": 5.541993371594859,
+      "grad_norm": 1.2362616062164307,
+      "learning_rate": 0.0002,
+      "loss": 0.5375,
+      "step": 34280
+    },
+    {
+      "epoch": 5.543610055775604,
+      "grad_norm": 1.0206977128982544,
+      "learning_rate": 0.0002,
+      "loss": 0.4996,
+      "step": 34290
+    },
+    {
+      "epoch": 5.54522673995635,
+      "grad_norm": 1.2018457651138306,
+      "learning_rate": 0.0002,
+      "loss": 0.5212,
+      "step": 34300
+    },
+    {
+      "epoch": 5.546843424137095,
+      "grad_norm": 1.0349043607711792,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 34310
+    },
+    {
+      "epoch": 5.54846010831784,
+      "grad_norm": 1.2022006511688232,
+      "learning_rate": 0.0002,
+      "loss": 0.5231,
+      "step": 34320
+    },
+    {
+      "epoch": 5.550076792498586,
+      "grad_norm": 1.0810624361038208,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 34330
+    },
+    {
+      "epoch": 5.551693476679331,
+      "grad_norm": 1.3297529220581055,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 34340
+    },
+    {
+      "epoch": 5.553310160860076,
+      "grad_norm": 0.9722549915313721,
+      "learning_rate": 0.0002,
+      "loss": 0.5321,
+      "step": 34350
+    },
+    {
+      "epoch": 5.554926845040821,
+      "grad_norm": 0.9903425574302673,
+      "learning_rate": 0.0002,
+      "loss": 0.4823,
+      "step": 34360
+    },
+    {
+      "epoch": 5.556543529221567,
+      "grad_norm": 0.9568067789077759,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 34370
+    },
+    {
+      "epoch": 5.558160213402312,
+      "grad_norm": 1.113870620727539,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 34380
+    },
+    {
+      "epoch": 5.559776897583057,
+      "grad_norm": 1.0557632446289062,
+      "learning_rate": 0.0002,
+      "loss": 0.5278,
+      "step": 34390
+    },
+    {
+      "epoch": 5.561393581763802,
+      "grad_norm": 0.9615673422813416,
+      "learning_rate": 0.0002,
+      "loss": 0.5501,
+      "step": 34400
+    },
+    {
+      "epoch": 5.5630102659445475,
+      "grad_norm": 0.9536027312278748,
+      "learning_rate": 0.0002,
+      "loss": 0.5066,
+      "step": 34410
+    },
+    {
+      "epoch": 5.564626950125293,
+      "grad_norm": 0.8808749318122864,
+      "learning_rate": 0.0002,
+      "loss": 0.4949,
+      "step": 34420
+    },
+    {
+      "epoch": 5.566243634306038,
+      "grad_norm": 1.286132574081421,
+      "learning_rate": 0.0002,
+      "loss": 0.5954,
+      "step": 34430
+    },
+    {
+      "epoch": 5.567860318486783,
+      "grad_norm": 1.259644865989685,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 34440
+    },
+    {
+      "epoch": 5.569477002667529,
+      "grad_norm": 0.9920216798782349,
+      "learning_rate": 0.0002,
+      "loss": 0.4922,
+      "step": 34450
+    },
+    {
+      "epoch": 5.5710936868482746,
+      "grad_norm": 1.182926893234253,
+      "learning_rate": 0.0002,
+      "loss": 0.5527,
+      "step": 34460
+    },
+    {
+      "epoch": 5.57271037102902,
+      "grad_norm": 1.1434749364852905,
+      "learning_rate": 0.0002,
+      "loss": 0.5185,
+      "step": 34470
+    },
+    {
+      "epoch": 5.574327055209765,
+      "grad_norm": 1.2420979738235474,
+      "learning_rate": 0.0002,
+      "loss": 0.5256,
+      "step": 34480
+    },
+    {
+      "epoch": 5.57594373939051,
+      "grad_norm": 0.9338384866714478,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 34490
+    },
+    {
+      "epoch": 5.5775604235712555,
+      "grad_norm": 1.0196425914764404,
+      "learning_rate": 0.0002,
+      "loss": 0.5634,
+      "step": 34500
+    },
+    {
+      "epoch": 5.579177107752001,
+      "grad_norm": 0.9586997032165527,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 34510
+    },
+    {
+      "epoch": 5.580793791932746,
+      "grad_norm": 1.2409086227416992,
+      "learning_rate": 0.0002,
+      "loss": 0.5336,
+      "step": 34520
+    },
+    {
+      "epoch": 5.582410476113491,
+      "grad_norm": 1.1483757495880127,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 34530
+    },
+    {
+      "epoch": 5.584027160294236,
+      "grad_norm": 1.1624305248260498,
+      "learning_rate": 0.0002,
+      "loss": 0.5325,
+      "step": 34540
+    },
+    {
+      "epoch": 5.585643844474982,
+      "grad_norm": 1.2635223865509033,
+      "learning_rate": 0.0002,
+      "loss": 0.5342,
+      "step": 34550
+    },
+    {
+      "epoch": 5.587260528655727,
+      "grad_norm": 0.9824051856994629,
+      "learning_rate": 0.0002,
+      "loss": 0.4924,
+      "step": 34560
+    },
+    {
+      "epoch": 5.588877212836472,
+      "grad_norm": 1.0858620405197144,
+      "learning_rate": 0.0002,
+      "loss": 0.5395,
+      "step": 34570
+    },
+    {
+      "epoch": 5.590493897017217,
+      "grad_norm": 1.1452655792236328,
+      "learning_rate": 0.0002,
+      "loss": 0.5459,
+      "step": 34580
+    },
+    {
+      "epoch": 5.592110581197963,
+      "grad_norm": 1.110610842704773,
+      "learning_rate": 0.0002,
+      "loss": 0.5746,
+      "step": 34590
+    },
+    {
+      "epoch": 5.593727265378709,
+      "grad_norm": 0.9976194500923157,
+      "learning_rate": 0.0002,
+      "loss": 0.5285,
+      "step": 34600
+    },
+    {
+      "epoch": 5.595343949559454,
+      "grad_norm": 1.0698920488357544,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 34610
+    },
+    {
+      "epoch": 5.596960633740199,
+      "grad_norm": 1.1505171060562134,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 34620
+    },
+    {
+      "epoch": 5.598577317920944,
+      "grad_norm": 1.1014643907546997,
+      "learning_rate": 0.0002,
+      "loss": 0.5471,
+      "step": 34630
+    },
+    {
+      "epoch": 5.60019400210169,
+      "grad_norm": 0.915595293045044,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 34640
+    },
+    {
+      "epoch": 5.601810686282435,
+      "grad_norm": 1.1856765747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 34650
+    },
+    {
+      "epoch": 5.60342737046318,
+      "grad_norm": 1.1357687711715698,
+      "learning_rate": 0.0002,
+      "loss": 0.5502,
+      "step": 34660
+    },
+    {
+      "epoch": 5.605044054643925,
+      "grad_norm": 1.0232492685317993,
+      "learning_rate": 0.0002,
+      "loss": 0.5034,
+      "step": 34670
+    },
+    {
+      "epoch": 5.6066607388246705,
+      "grad_norm": 0.9375017881393433,
+      "learning_rate": 0.0002,
+      "loss": 0.5357,
+      "step": 34680
+    },
+    {
+      "epoch": 5.608277423005416,
+      "grad_norm": 1.0796529054641724,
+      "learning_rate": 0.0002,
+      "loss": 0.5518,
+      "step": 34690
+    },
+    {
+      "epoch": 5.609894107186161,
+      "grad_norm": 1.1383336782455444,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 34700
+    },
+    {
+      "epoch": 5.611510791366906,
+      "grad_norm": 1.0248544216156006,
+      "learning_rate": 0.0002,
+      "loss": 0.5477,
+      "step": 34710
+    },
+    {
+      "epoch": 5.6131274755476515,
+      "grad_norm": 1.0986040830612183,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 34720
+    },
+    {
+      "epoch": 5.614744159728397,
+      "grad_norm": 1.2689568996429443,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 34730
+    },
+    {
+      "epoch": 5.616360843909142,
+      "grad_norm": 1.4044264554977417,
+      "learning_rate": 0.0002,
+      "loss": 0.5136,
+      "step": 34740
+    },
+    {
+      "epoch": 5.617977528089888,
+      "grad_norm": 1.2084474563598633,
+      "learning_rate": 0.0002,
+      "loss": 0.5699,
+      "step": 34750
+    },
+    {
+      "epoch": 5.619594212270633,
+      "grad_norm": 1.061248540878296,
+      "learning_rate": 0.0002,
+      "loss": 0.5377,
+      "step": 34760
+    },
+    {
+      "epoch": 5.6212108964513785,
+      "grad_norm": 1.0220764875411987,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 34770
+    },
+    {
+      "epoch": 5.622827580632124,
+      "grad_norm": 1.0859092473983765,
+      "learning_rate": 0.0002,
+      "loss": 0.54,
+      "step": 34780
+    },
+    {
+      "epoch": 5.624444264812869,
+      "grad_norm": 0.9049732089042664,
+      "learning_rate": 0.0002,
+      "loss": 0.5308,
+      "step": 34790
+    },
+    {
+      "epoch": 5.626060948993614,
+      "grad_norm": 1.2103937864303589,
+      "learning_rate": 0.0002,
+      "loss": 0.5433,
+      "step": 34800
+    },
+    {
+      "epoch": 5.627677633174359,
+      "grad_norm": 0.9854230284690857,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 34810
+    },
+    {
+      "epoch": 5.629294317355105,
+      "grad_norm": 0.9316635131835938,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 34820
+    },
+    {
+      "epoch": 5.63091100153585,
+      "grad_norm": 1.105296015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.5393,
+      "step": 34830
+    },
+    {
+      "epoch": 5.632527685716595,
+      "grad_norm": 0.993383526802063,
+      "learning_rate": 0.0002,
+      "loss": 0.5527,
+      "step": 34840
+    },
+    {
+      "epoch": 5.63414436989734,
+      "grad_norm": 1.1544116735458374,
+      "learning_rate": 0.0002,
+      "loss": 0.5375,
+      "step": 34850
+    },
+    {
+      "epoch": 5.635761054078086,
+      "grad_norm": 1.284475326538086,
+      "learning_rate": 0.0002,
+      "loss": 0.5448,
+      "step": 34860
+    },
+    {
+      "epoch": 5.637377738258831,
+      "grad_norm": 1.121997594833374,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 34870
+    },
+    {
+      "epoch": 5.638994422439576,
+      "grad_norm": 1.213040828704834,
+      "learning_rate": 0.0002,
+      "loss": 0.5335,
+      "step": 34880
+    },
+    {
+      "epoch": 5.640611106620321,
+      "grad_norm": 1.23222017288208,
+      "learning_rate": 0.0002,
+      "loss": 0.5623,
+      "step": 34890
+    },
+    {
+      "epoch": 5.642227790801067,
+      "grad_norm": 0.9793637990951538,
+      "learning_rate": 0.0002,
+      "loss": 0.5622,
+      "step": 34900
+    },
+    {
+      "epoch": 5.643844474981813,
+      "grad_norm": 1.38919997215271,
+      "learning_rate": 0.0002,
+      "loss": 0.5405,
+      "step": 34910
+    },
+    {
+      "epoch": 5.645461159162558,
+      "grad_norm": 0.8390951156616211,
+      "learning_rate": 0.0002,
+      "loss": 0.5007,
+      "step": 34920
+    },
+    {
+      "epoch": 5.647077843343303,
+      "grad_norm": 0.9465909004211426,
+      "learning_rate": 0.0002,
+      "loss": 0.5974,
+      "step": 34930
+    },
+    {
+      "epoch": 5.648694527524048,
+      "grad_norm": 1.066957712173462,
+      "learning_rate": 0.0002,
+      "loss": 0.5264,
+      "step": 34940
+    },
+    {
+      "epoch": 5.650311211704794,
+      "grad_norm": 0.9842154383659363,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 34950
+    },
+    {
+      "epoch": 5.651927895885539,
+      "grad_norm": 1.1766440868377686,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 34960
+    },
+    {
+      "epoch": 5.653544580066284,
+      "grad_norm": 0.9061306118965149,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 34970
+    },
+    {
+      "epoch": 5.655161264247029,
+      "grad_norm": 1.2941309213638306,
+      "learning_rate": 0.0002,
+      "loss": 0.5446,
+      "step": 34980
+    },
+    {
+      "epoch": 5.6567779484277745,
+      "grad_norm": 0.9741247892379761,
+      "learning_rate": 0.0002,
+      "loss": 0.5704,
+      "step": 34990
+    },
+    {
+      "epoch": 5.65839463260852,
+      "grad_norm": 1.0784187316894531,
+      "learning_rate": 0.0002,
+      "loss": 0.5152,
+      "step": 35000
+    },
+    {
+      "epoch": 5.660011316789265,
+      "grad_norm": 0.937889814376831,
+      "learning_rate": 0.0002,
+      "loss": 0.5363,
+      "step": 35010
+    },
+    {
+      "epoch": 5.66162800097001,
+      "grad_norm": 0.9667879939079285,
+      "learning_rate": 0.0002,
+      "loss": 0.5019,
+      "step": 35020
+    },
+    {
+      "epoch": 5.663244685150756,
+      "grad_norm": 1.0554876327514648,
+      "learning_rate": 0.0002,
+      "loss": 0.5209,
+      "step": 35030
+    },
+    {
+      "epoch": 5.664861369331501,
+      "grad_norm": 1.2030539512634277,
+      "learning_rate": 0.0002,
+      "loss": 0.523,
+      "step": 35040
+    },
+    {
+      "epoch": 5.666478053512247,
+      "grad_norm": 1.0849953889846802,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 35050
+    },
+    {
+      "epoch": 5.668094737692992,
+      "grad_norm": 1.1598973274230957,
+      "learning_rate": 0.0002,
+      "loss": 0.5747,
+      "step": 35060
+    },
+    {
+      "epoch": 5.669711421873737,
+      "grad_norm": 1.0233359336853027,
+      "learning_rate": 0.0002,
+      "loss": 0.5488,
+      "step": 35070
+    },
+    {
+      "epoch": 5.6713281060544825,
+      "grad_norm": 1.1124799251556396,
+      "learning_rate": 0.0002,
+      "loss": 0.5409,
+      "step": 35080
+    },
+    {
+      "epoch": 5.672944790235228,
+      "grad_norm": 1.2351475954055786,
+      "learning_rate": 0.0002,
+      "loss": 0.5578,
+      "step": 35090
+    },
+    {
+      "epoch": 5.674561474415973,
+      "grad_norm": 1.0240728855133057,
+      "learning_rate": 0.0002,
+      "loss": 0.5638,
+      "step": 35100
+    },
+    {
+      "epoch": 5.676178158596718,
+      "grad_norm": 1.0223692655563354,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 35110
+    },
+    {
+      "epoch": 5.677794842777463,
+      "grad_norm": 1.4569132328033447,
+      "learning_rate": 0.0002,
+      "loss": 0.524,
+      "step": 35120
+    },
+    {
+      "epoch": 5.679411526958209,
+      "grad_norm": 0.8983587026596069,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 35130
+    },
+    {
+      "epoch": 5.681028211138954,
+      "grad_norm": 1.0775383710861206,
+      "learning_rate": 0.0002,
+      "loss": 0.5439,
+      "step": 35140
+    },
+    {
+      "epoch": 5.682644895319699,
+      "grad_norm": 0.9800270795822144,
+      "learning_rate": 0.0002,
+      "loss": 0.5289,
+      "step": 35150
+    },
+    {
+      "epoch": 5.684261579500444,
+      "grad_norm": 0.9858237504959106,
+      "learning_rate": 0.0002,
+      "loss": 0.533,
+      "step": 35160
+    },
+    {
+      "epoch": 5.6858782636811895,
+      "grad_norm": 1.031087040901184,
+      "learning_rate": 0.0002,
+      "loss": 0.5671,
+      "step": 35170
+    },
+    {
+      "epoch": 5.687494947861936,
+      "grad_norm": 1.0294365882873535,
+      "learning_rate": 0.0002,
+      "loss": 0.5528,
+      "step": 35180
+    },
+    {
+      "epoch": 5.68911163204268,
+      "grad_norm": 1.108144760131836,
+      "learning_rate": 0.0002,
+      "loss": 0.5581,
+      "step": 35190
+    },
+    {
+      "epoch": 5.690728316223426,
+      "grad_norm": 1.0813100337982178,
+      "learning_rate": 0.0002,
+      "loss": 0.5373,
+      "step": 35200
+    },
+    {
+      "epoch": 5.692345000404171,
+      "grad_norm": 1.3146867752075195,
+      "learning_rate": 0.0002,
+      "loss": 0.5429,
+      "step": 35210
+    },
+    {
+      "epoch": 5.693961684584917,
+      "grad_norm": 1.16780424118042,
+      "learning_rate": 0.0002,
+      "loss": 0.5297,
+      "step": 35220
+    },
+    {
+      "epoch": 5.695578368765662,
+      "grad_norm": 0.9929125905036926,
+      "learning_rate": 0.0002,
+      "loss": 0.577,
+      "step": 35230
+    },
+    {
+      "epoch": 5.697195052946407,
+      "grad_norm": 0.9049441814422607,
+      "learning_rate": 0.0002,
+      "loss": 0.5441,
+      "step": 35240
+    },
+    {
+      "epoch": 5.698811737127152,
+      "grad_norm": 0.9768866300582886,
+      "learning_rate": 0.0002,
+      "loss": 0.5349,
+      "step": 35250
+    },
+    {
+      "epoch": 5.7004284213078975,
+      "grad_norm": 0.8306029438972473,
+      "learning_rate": 0.0002,
+      "loss": 0.542,
+      "step": 35260
+    },
+    {
+      "epoch": 5.702045105488643,
+      "grad_norm": 0.8417280316352844,
+      "learning_rate": 0.0002,
+      "loss": 0.4771,
+      "step": 35270
+    },
+    {
+      "epoch": 5.703661789669388,
+      "grad_norm": 0.9954485893249512,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 35280
+    },
+    {
+      "epoch": 5.705278473850133,
+      "grad_norm": 1.2417993545532227,
+      "learning_rate": 0.0002,
+      "loss": 0.5469,
+      "step": 35290
+    },
+    {
+      "epoch": 5.706895158030878,
+      "grad_norm": 1.1696544885635376,
+      "learning_rate": 0.0002,
+      "loss": 0.5275,
+      "step": 35300
+    },
+    {
+      "epoch": 5.708511842211624,
+      "grad_norm": 1.2424817085266113,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 35310
+    },
+    {
+      "epoch": 5.710128526392369,
+      "grad_norm": 1.1791106462478638,
+      "learning_rate": 0.0002,
+      "loss": 0.5595,
+      "step": 35320
+    },
+    {
+      "epoch": 5.711745210573115,
+      "grad_norm": 1.202181339263916,
+      "learning_rate": 0.0002,
+      "loss": 0.5076,
+      "step": 35330
+    },
+    {
+      "epoch": 5.713361894753859,
+      "grad_norm": 1.1006861925125122,
+      "learning_rate": 0.0002,
+      "loss": 0.5847,
+      "step": 35340
+    },
+    {
+      "epoch": 5.7149785789346055,
+      "grad_norm": 1.0918344259262085,
+      "learning_rate": 0.0002,
+      "loss": 0.5627,
+      "step": 35350
+    },
+    {
+      "epoch": 5.716595263115351,
+      "grad_norm": 1.0427305698394775,
+      "learning_rate": 0.0002,
+      "loss": 0.5677,
+      "step": 35360
+    },
+    {
+      "epoch": 5.718211947296096,
+      "grad_norm": 1.0818872451782227,
+      "learning_rate": 0.0002,
+      "loss": 0.5288,
+      "step": 35370
+    },
+    {
+      "epoch": 5.719828631476841,
+      "grad_norm": 1.186006784439087,
+      "learning_rate": 0.0002,
+      "loss": 0.5296,
+      "step": 35380
+    },
+    {
+      "epoch": 5.721445315657586,
+      "grad_norm": 1.2073674201965332,
+      "learning_rate": 0.0002,
+      "loss": 0.5507,
+      "step": 35390
+    },
+    {
+      "epoch": 5.723061999838332,
+      "grad_norm": 1.065338134765625,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 35400
+    },
+    {
+      "epoch": 5.724678684019077,
+      "grad_norm": 0.9448973536491394,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 35410
+    },
+    {
+      "epoch": 5.726295368199822,
+      "grad_norm": 1.1487499475479126,
+      "learning_rate": 0.0002,
+      "loss": 0.5276,
+      "step": 35420
+    },
+    {
+      "epoch": 5.727912052380567,
+      "grad_norm": 1.1334216594696045,
+      "learning_rate": 0.0002,
+      "loss": 0.5435,
+      "step": 35430
+    },
+    {
+      "epoch": 5.729528736561313,
+      "grad_norm": 1.1932826042175293,
+      "learning_rate": 0.0002,
+      "loss": 0.5074,
+      "step": 35440
+    },
+    {
+      "epoch": 5.731145420742058,
+      "grad_norm": 1.2615786790847778,
+      "learning_rate": 0.0002,
+      "loss": 0.5502,
+      "step": 35450
+    },
+    {
+      "epoch": 5.732762104922803,
+      "grad_norm": 1.2803694009780884,
+      "learning_rate": 0.0002,
+      "loss": 0.5612,
+      "step": 35460
+    },
+    {
+      "epoch": 5.734378789103548,
+      "grad_norm": 0.9271906614303589,
+      "learning_rate": 0.0002,
+      "loss": 0.5458,
+      "step": 35470
+    },
+    {
+      "epoch": 5.735995473284294,
+      "grad_norm": 1.0958917140960693,
+      "learning_rate": 0.0002,
+      "loss": 0.5342,
+      "step": 35480
+    },
+    {
+      "epoch": 5.737612157465039,
+      "grad_norm": 1.1072784662246704,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 35490
+    },
+    {
+      "epoch": 5.739228841645785,
+      "grad_norm": 1.1641002893447876,
+      "learning_rate": 0.0002,
+      "loss": 0.5683,
+      "step": 35500
+    },
+    {
+      "epoch": 5.74084552582653,
+      "grad_norm": 1.0246447324752808,
+      "learning_rate": 0.0002,
+      "loss": 0.5252,
+      "step": 35510
+    },
+    {
+      "epoch": 5.742462210007275,
+      "grad_norm": 1.032474398612976,
+      "learning_rate": 0.0002,
+      "loss": 0.55,
+      "step": 35520
+    },
+    {
+      "epoch": 5.7440788941880205,
+      "grad_norm": 1.1600854396820068,
+      "learning_rate": 0.0002,
+      "loss": 0.4965,
+      "step": 35530
+    },
+    {
+      "epoch": 5.745695578368766,
+      "grad_norm": 1.0686054229736328,
+      "learning_rate": 0.0002,
+      "loss": 0.5543,
+      "step": 35540
+    },
+    {
+      "epoch": 5.747312262549511,
+      "grad_norm": 1.2314637899398804,
+      "learning_rate": 0.0002,
+      "loss": 0.5706,
+      "step": 35550
+    },
+    {
+      "epoch": 5.748928946730256,
+      "grad_norm": 0.922134280204773,
+      "learning_rate": 0.0002,
+      "loss": 0.5492,
+      "step": 35560
+    },
+    {
+      "epoch": 5.7505456309110015,
+      "grad_norm": 0.933043360710144,
+      "learning_rate": 0.0002,
+      "loss": 0.5495,
+      "step": 35570
+    },
+    {
+      "epoch": 5.752162315091747,
+      "grad_norm": 1.1911931037902832,
+      "learning_rate": 0.0002,
+      "loss": 0.5007,
+      "step": 35580
+    },
+    {
+      "epoch": 5.753778999272492,
+      "grad_norm": 0.8984857797622681,
+      "learning_rate": 0.0002,
+      "loss": 0.5244,
+      "step": 35590
+    },
+    {
+      "epoch": 5.755395683453237,
+      "grad_norm": 0.9495107531547546,
+      "learning_rate": 0.0002,
+      "loss": 0.5493,
+      "step": 35600
+    },
+    {
+      "epoch": 5.757012367633982,
+      "grad_norm": 1.2805472612380981,
+      "learning_rate": 0.0002,
+      "loss": 0.5326,
+      "step": 35610
+    },
+    {
+      "epoch": 5.758629051814728,
+      "grad_norm": 1.1236625909805298,
+      "learning_rate": 0.0002,
+      "loss": 0.5276,
+      "step": 35620
+    },
+    {
+      "epoch": 5.760245735995474,
+      "grad_norm": 1.0552798509597778,
+      "learning_rate": 0.0002,
+      "loss": 0.6102,
+      "step": 35630
+    },
+    {
+      "epoch": 5.761862420176218,
+      "grad_norm": 1.119909644126892,
+      "learning_rate": 0.0002,
+      "loss": 0.5479,
+      "step": 35640
+    },
+    {
+      "epoch": 5.763479104356964,
+      "grad_norm": 0.8786116242408752,
+      "learning_rate": 0.0002,
+      "loss": 0.5282,
+      "step": 35650
+    },
+    {
+      "epoch": 5.765095788537709,
+      "grad_norm": 1.2417117357254028,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 35660
+    },
+    {
+      "epoch": 5.766712472718455,
+      "grad_norm": 1.255200982093811,
+      "learning_rate": 0.0002,
+      "loss": 0.537,
+      "step": 35670
+    },
+    {
+      "epoch": 5.7683291568992,
+      "grad_norm": 1.0611358880996704,
+      "learning_rate": 0.0002,
+      "loss": 0.5308,
+      "step": 35680
+    },
+    {
+      "epoch": 5.769945841079945,
+      "grad_norm": 1.1443911790847778,
+      "learning_rate": 0.0002,
+      "loss": 0.5614,
+      "step": 35690
+    },
+    {
+      "epoch": 5.77156252526069,
+      "grad_norm": 1.1437989473342896,
+      "learning_rate": 0.0002,
+      "loss": 0.5386,
+      "step": 35700
+    },
+    {
+      "epoch": 5.773179209441436,
+      "grad_norm": 1.1375046968460083,
+      "learning_rate": 0.0002,
+      "loss": 0.537,
+      "step": 35710
+    },
+    {
+      "epoch": 5.774795893622181,
+      "grad_norm": 1.0777729749679565,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 35720
+    },
+    {
+      "epoch": 5.776412577802926,
+      "grad_norm": 1.1160215139389038,
+      "learning_rate": 0.0002,
+      "loss": 0.5521,
+      "step": 35730
+    },
+    {
+      "epoch": 5.778029261983671,
+      "grad_norm": 1.1268514394760132,
+      "learning_rate": 0.0002,
+      "loss": 0.5569,
+      "step": 35740
+    },
+    {
+      "epoch": 5.7796459461644165,
+      "grad_norm": 1.2752262353897095,
+      "learning_rate": 0.0002,
+      "loss": 0.5311,
+      "step": 35750
+    },
+    {
+      "epoch": 5.781262630345162,
+      "grad_norm": 1.0416184663772583,
+      "learning_rate": 0.0002,
+      "loss": 0.5625,
+      "step": 35760
+    },
+    {
+      "epoch": 5.782879314525907,
+      "grad_norm": 1.0622444152832031,
+      "learning_rate": 0.0002,
+      "loss": 0.5438,
+      "step": 35770
+    },
+    {
+      "epoch": 5.784495998706653,
+      "grad_norm": 1.1217877864837646,
+      "learning_rate": 0.0002,
+      "loss": 0.5268,
+      "step": 35780
+    },
+    {
+      "epoch": 5.786112682887398,
+      "grad_norm": 0.9363139867782593,
+      "learning_rate": 0.0002,
+      "loss": 0.5225,
+      "step": 35790
+    },
+    {
+      "epoch": 5.787729367068144,
+      "grad_norm": 0.96628737449646,
+      "learning_rate": 0.0002,
+      "loss": 0.5524,
+      "step": 35800
+    },
+    {
+      "epoch": 5.789346051248889,
+      "grad_norm": 0.9572572112083435,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 35810
+    },
+    {
+      "epoch": 5.790962735429634,
+      "grad_norm": 0.938724935054779,
+      "learning_rate": 0.0002,
+      "loss": 0.5615,
+      "step": 35820
+    },
+    {
+      "epoch": 5.792579419610379,
+      "grad_norm": 1.3314417600631714,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 35830
+    },
+    {
+      "epoch": 5.7941961037911245,
+      "grad_norm": 1.0097602605819702,
+      "learning_rate": 0.0002,
+      "loss": 0.5441,
+      "step": 35840
+    },
+    {
+      "epoch": 5.79581278797187,
+      "grad_norm": 1.1265122890472412,
+      "learning_rate": 0.0002,
+      "loss": 0.591,
+      "step": 35850
+    },
+    {
+      "epoch": 5.797429472152615,
+      "grad_norm": 1.2191909551620483,
+      "learning_rate": 0.0002,
+      "loss": 0.5333,
+      "step": 35860
+    },
+    {
+      "epoch": 5.79904615633336,
+      "grad_norm": 0.9690808057785034,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 35870
+    },
+    {
+      "epoch": 5.800662840514105,
+      "grad_norm": 1.0871665477752686,
+      "learning_rate": 0.0002,
+      "loss": 0.5425,
+      "step": 35880
+    },
+    {
+      "epoch": 5.802279524694851,
+      "grad_norm": 1.1093597412109375,
+      "learning_rate": 0.0002,
+      "loss": 0.5602,
+      "step": 35890
+    },
+    {
+      "epoch": 5.803896208875596,
+      "grad_norm": 1.2434282302856445,
+      "learning_rate": 0.0002,
+      "loss": 0.5475,
+      "step": 35900
+    },
+    {
+      "epoch": 5.805512893056341,
+      "grad_norm": 1.2933623790740967,
+      "learning_rate": 0.0002,
+      "loss": 0.5288,
+      "step": 35910
+    },
+    {
+      "epoch": 5.807129577237086,
+      "grad_norm": 1.0005441904067993,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 35920
+    },
+    {
+      "epoch": 5.8087462614178325,
+      "grad_norm": 1.2373108863830566,
+      "learning_rate": 0.0002,
+      "loss": 0.5318,
+      "step": 35930
+    },
+    {
+      "epoch": 5.810362945598578,
+      "grad_norm": 1.2622692584991455,
+      "learning_rate": 0.0002,
+      "loss": 0.5413,
+      "step": 35940
+    },
+    {
+      "epoch": 5.811979629779323,
+      "grad_norm": 1.0112963914871216,
+      "learning_rate": 0.0002,
+      "loss": 0.5558,
+      "step": 35950
+    },
+    {
+      "epoch": 5.813596313960068,
+      "grad_norm": 1.050572395324707,
+      "learning_rate": 0.0002,
+      "loss": 0.5115,
+      "step": 35960
+    },
+    {
+      "epoch": 5.815212998140813,
+      "grad_norm": 0.9774560928344727,
+      "learning_rate": 0.0002,
+      "loss": 0.5288,
+      "step": 35970
+    },
+    {
+      "epoch": 5.816829682321559,
+      "grad_norm": 1.19438898563385,
+      "learning_rate": 0.0002,
+      "loss": 0.585,
+      "step": 35980
+    },
+    {
+      "epoch": 5.818446366502304,
+      "grad_norm": 1.0267130136489868,
+      "learning_rate": 0.0002,
+      "loss": 0.5798,
+      "step": 35990
+    },
+    {
+      "epoch": 5.820063050683049,
+      "grad_norm": 0.9813851714134216,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 36000
+    },
+    {
+      "epoch": 5.821679734863794,
+      "grad_norm": 0.9177457094192505,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 36010
+    },
+    {
+      "epoch": 5.8232964190445395,
+      "grad_norm": 1.0020731687545776,
+      "learning_rate": 0.0002,
+      "loss": 0.5453,
+      "step": 36020
+    },
+    {
+      "epoch": 5.824913103225285,
+      "grad_norm": 1.073222041130066,
+      "learning_rate": 0.0002,
+      "loss": 0.5646,
+      "step": 36030
+    },
+    {
+      "epoch": 5.82652978740603,
+      "grad_norm": 1.016337513923645,
+      "learning_rate": 0.0002,
+      "loss": 0.5539,
+      "step": 36040
+    },
+    {
+      "epoch": 5.828146471586775,
+      "grad_norm": 1.267364263534546,
+      "learning_rate": 0.0002,
+      "loss": 0.5592,
+      "step": 36050
+    },
+    {
+      "epoch": 5.8297631557675205,
+      "grad_norm": 1.2730127573013306,
+      "learning_rate": 0.0002,
+      "loss": 0.595,
+      "step": 36060
+    },
+    {
+      "epoch": 5.831379839948266,
+      "grad_norm": 1.108442783355713,
+      "learning_rate": 0.0002,
+      "loss": 0.5247,
+      "step": 36070
+    },
+    {
+      "epoch": 5.832996524129012,
+      "grad_norm": 1.198072075843811,
+      "learning_rate": 0.0002,
+      "loss": 0.5103,
+      "step": 36080
+    },
+    {
+      "epoch": 5.834613208309757,
+      "grad_norm": 1.0458786487579346,
+      "learning_rate": 0.0002,
+      "loss": 0.5479,
+      "step": 36090
+    },
+    {
+      "epoch": 5.836229892490502,
+      "grad_norm": 0.9096664786338806,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 36100
+    },
+    {
+      "epoch": 5.8378465766712475,
+      "grad_norm": 0.9957793951034546,
+      "learning_rate": 0.0002,
+      "loss": 0.5602,
+      "step": 36110
+    },
+    {
+      "epoch": 5.839463260851993,
+      "grad_norm": 1.3693058490753174,
+      "learning_rate": 0.0002,
+      "loss": 0.5799,
+      "step": 36120
+    },
+    {
+      "epoch": 5.841079945032738,
+      "grad_norm": 1.268608808517456,
+      "learning_rate": 0.0002,
+      "loss": 0.5425,
+      "step": 36130
+    },
+    {
+      "epoch": 5.842696629213483,
+      "grad_norm": 0.8516020178794861,
+      "learning_rate": 0.0002,
+      "loss": 0.5653,
+      "step": 36140
+    },
+    {
+      "epoch": 5.844313313394228,
+      "grad_norm": 0.90385502576828,
+      "learning_rate": 0.0002,
+      "loss": 0.5475,
+      "step": 36150
+    },
+    {
+      "epoch": 5.845929997574974,
+      "grad_norm": 1.0910571813583374,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 36160
+    },
+    {
+      "epoch": 5.847546681755719,
+      "grad_norm": 0.9417795538902283,
+      "learning_rate": 0.0002,
+      "loss": 0.555,
+      "step": 36170
+    },
+    {
+      "epoch": 5.849163365936464,
+      "grad_norm": 1.0027360916137695,
+      "learning_rate": 0.0002,
+      "loss": 0.5784,
+      "step": 36180
+    },
+    {
+      "epoch": 5.850780050117209,
+      "grad_norm": 1.1480516195297241,
+      "learning_rate": 0.0002,
+      "loss": 0.5423,
+      "step": 36190
+    },
+    {
+      "epoch": 5.852396734297955,
+      "grad_norm": 1.2431457042694092,
+      "learning_rate": 0.0002,
+      "loss": 0.5517,
+      "step": 36200
+    },
+    {
+      "epoch": 5.8540134184787,
+      "grad_norm": 1.091465950012207,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 36210
+    },
+    {
+      "epoch": 5.855630102659445,
+      "grad_norm": 0.9693930745124817,
+      "learning_rate": 0.0002,
+      "loss": 0.53,
+      "step": 36220
+    },
+    {
+      "epoch": 5.857246786840191,
+      "grad_norm": 0.9937465190887451,
+      "learning_rate": 0.0002,
+      "loss": 0.5453,
+      "step": 36230
+    },
+    {
+      "epoch": 5.858863471020936,
+      "grad_norm": 1.0731011629104614,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 36240
+    },
+    {
+      "epoch": 5.860480155201682,
+      "grad_norm": 1.0869048833847046,
+      "learning_rate": 0.0002,
+      "loss": 0.5687,
+      "step": 36250
+    },
+    {
+      "epoch": 5.862096839382427,
+      "grad_norm": 0.9226390719413757,
+      "learning_rate": 0.0002,
+      "loss": 0.5576,
+      "step": 36260
+    },
+    {
+      "epoch": 5.863713523563172,
+      "grad_norm": 1.1755430698394775,
+      "learning_rate": 0.0002,
+      "loss": 0.531,
+      "step": 36270
+    },
+    {
+      "epoch": 5.865330207743917,
+      "grad_norm": 0.8815974593162537,
+      "learning_rate": 0.0002,
+      "loss": 0.558,
+      "step": 36280
+    },
+    {
+      "epoch": 5.866946891924663,
+      "grad_norm": 1.3648751974105835,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 36290
+    },
+    {
+      "epoch": 5.868563576105408,
+      "grad_norm": 0.8729211091995239,
+      "learning_rate": 0.0002,
+      "loss": 0.536,
+      "step": 36300
+    },
+    {
+      "epoch": 5.870180260286153,
+      "grad_norm": 1.0870907306671143,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 36310
+    },
+    {
+      "epoch": 5.871796944466898,
+      "grad_norm": 1.1164259910583496,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 36320
+    },
+    {
+      "epoch": 5.8734136286476435,
+      "grad_norm": 1.1572535037994385,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 36330
+    },
+    {
+      "epoch": 5.875030312828389,
+      "grad_norm": 1.0456238985061646,
+      "learning_rate": 0.0002,
+      "loss": 0.5898,
+      "step": 36340
+    },
+    {
+      "epoch": 5.876646997009134,
+      "grad_norm": 1.1310722827911377,
+      "learning_rate": 0.0002,
+      "loss": 0.5008,
+      "step": 36350
+    },
+    {
+      "epoch": 5.878263681189879,
+      "grad_norm": 1.0004712343215942,
+      "learning_rate": 0.0002,
+      "loss": 0.5352,
+      "step": 36360
+    },
+    {
+      "epoch": 5.879880365370624,
+      "grad_norm": 1.0991777181625366,
+      "learning_rate": 0.0002,
+      "loss": 0.5632,
+      "step": 36370
+    },
+    {
+      "epoch": 5.8814970495513705,
+      "grad_norm": 1.2789239883422852,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 36380
+    },
+    {
+      "epoch": 5.883113733732116,
+      "grad_norm": 0.9524819850921631,
+      "learning_rate": 0.0002,
+      "loss": 0.56,
+      "step": 36390
+    },
+    {
+      "epoch": 5.884730417912861,
+      "grad_norm": 1.1115771532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.5701,
+      "step": 36400
+    },
+    {
+      "epoch": 5.886347102093606,
+      "grad_norm": 1.37419855594635,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 36410
+    },
+    {
+      "epoch": 5.8879637862743515,
+      "grad_norm": 1.1449527740478516,
+      "learning_rate": 0.0002,
+      "loss": 0.5675,
+      "step": 36420
+    },
+    {
+      "epoch": 5.889580470455097,
+      "grad_norm": 1.198046326637268,
+      "learning_rate": 0.0002,
+      "loss": 0.5255,
+      "step": 36430
+    },
+    {
+      "epoch": 5.891197154635842,
+      "grad_norm": 1.0180530548095703,
+      "learning_rate": 0.0002,
+      "loss": 0.5383,
+      "step": 36440
+    },
+    {
+      "epoch": 5.892813838816587,
+      "grad_norm": 1.0516417026519775,
+      "learning_rate": 0.0002,
+      "loss": 0.5319,
+      "step": 36450
+    },
+    {
+      "epoch": 5.894430522997332,
+      "grad_norm": 1.1658052206039429,
+      "learning_rate": 0.0002,
+      "loss": 0.5782,
+      "step": 36460
+    },
+    {
+      "epoch": 5.896047207178078,
+      "grad_norm": 1.190699577331543,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 36470
+    },
+    {
+      "epoch": 5.897663891358823,
+      "grad_norm": 1.1235495805740356,
+      "learning_rate": 0.0002,
+      "loss": 0.5451,
+      "step": 36480
+    },
+    {
+      "epoch": 5.899280575539568,
+      "grad_norm": 1.1926926374435425,
+      "learning_rate": 0.0002,
+      "loss": 0.5284,
+      "step": 36490
+    },
+    {
+      "epoch": 5.900897259720313,
+      "grad_norm": 1.1184662580490112,
+      "learning_rate": 0.0002,
+      "loss": 0.5686,
+      "step": 36500
+    },
+    {
+      "epoch": 5.9025139439010585,
+      "grad_norm": 1.000970721244812,
+      "learning_rate": 0.0002,
+      "loss": 0.5147,
+      "step": 36510
+    },
+    {
+      "epoch": 5.904130628081804,
+      "grad_norm": 1.0373306274414062,
+      "learning_rate": 0.0002,
+      "loss": 0.5351,
+      "step": 36520
+    },
+    {
+      "epoch": 5.90574731226255,
+      "grad_norm": 1.0840669870376587,
+      "learning_rate": 0.0002,
+      "loss": 0.535,
+      "step": 36530
+    },
+    {
+      "epoch": 5.907363996443295,
+      "grad_norm": 0.9908381104469299,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 36540
+    },
+    {
+      "epoch": 5.90898068062404,
+      "grad_norm": 1.0456029176712036,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 36550
+    },
+    {
+      "epoch": 5.910597364804786,
+      "grad_norm": 1.1381454467773438,
+      "learning_rate": 0.0002,
+      "loss": 0.5693,
+      "step": 36560
+    },
+    {
+      "epoch": 5.912214048985531,
+      "grad_norm": 0.9440900087356567,
+      "learning_rate": 0.0002,
+      "loss": 0.5473,
+      "step": 36570
+    },
+    {
+      "epoch": 5.913830733166276,
+      "grad_norm": 1.1674573421478271,
+      "learning_rate": 0.0002,
+      "loss": 0.5542,
+      "step": 36580
+    },
+    {
+      "epoch": 5.915447417347021,
+      "grad_norm": 1.1226966381072998,
+      "learning_rate": 0.0002,
+      "loss": 0.526,
+      "step": 36590
+    },
+    {
+      "epoch": 5.9170641015277665,
+      "grad_norm": 0.9696915745735168,
+      "learning_rate": 0.0002,
+      "loss": 0.6091,
+      "step": 36600
+    },
+    {
+      "epoch": 5.918680785708512,
+      "grad_norm": 0.9593005180358887,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 36610
+    },
+    {
+      "epoch": 5.920297469889257,
+      "grad_norm": 1.122169852256775,
+      "learning_rate": 0.0002,
+      "loss": 0.5536,
+      "step": 36620
+    },
+    {
+      "epoch": 5.921914154070002,
+      "grad_norm": 0.9923415780067444,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 36630
+    },
+    {
+      "epoch": 5.923530838250747,
+      "grad_norm": 1.063838005065918,
+      "learning_rate": 0.0002,
+      "loss": 0.5893,
+      "step": 36640
+    },
+    {
+      "epoch": 5.925147522431493,
+      "grad_norm": 0.9083505272865295,
+      "learning_rate": 0.0002,
+      "loss": 0.5799,
+      "step": 36650
+    },
+    {
+      "epoch": 5.926764206612239,
+      "grad_norm": 0.9439437985420227,
+      "learning_rate": 0.0002,
+      "loss": 0.5264,
+      "step": 36660
+    },
+    {
+      "epoch": 5.928380890792983,
+      "grad_norm": 0.9778534173965454,
+      "learning_rate": 0.0002,
+      "loss": 0.5891,
+      "step": 36670
+    },
+    {
+      "epoch": 5.929997574973729,
+      "grad_norm": 0.9723961353302002,
+      "learning_rate": 0.0002,
+      "loss": 0.566,
+      "step": 36680
+    },
+    {
+      "epoch": 5.9316142591544745,
+      "grad_norm": 1.162333607673645,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 36690
+    },
+    {
+      "epoch": 5.93323094333522,
+      "grad_norm": 1.2784897089004517,
+      "learning_rate": 0.0002,
+      "loss": 0.5771,
+      "step": 36700
+    },
+    {
+      "epoch": 5.934847627515965,
+      "grad_norm": 1.0924867391586304,
+      "learning_rate": 0.0002,
+      "loss": 0.5343,
+      "step": 36710
+    },
+    {
+      "epoch": 5.93646431169671,
+      "grad_norm": 1.046922206878662,
+      "learning_rate": 0.0002,
+      "loss": 0.5554,
+      "step": 36720
+    },
+    {
+      "epoch": 5.938080995877455,
+      "grad_norm": 0.8632535338401794,
+      "learning_rate": 0.0002,
+      "loss": 0.5476,
+      "step": 36730
+    },
+    {
+      "epoch": 5.939697680058201,
+      "grad_norm": 1.358762502670288,
+      "learning_rate": 0.0002,
+      "loss": 0.5456,
+      "step": 36740
+    },
+    {
+      "epoch": 5.941314364238946,
+      "grad_norm": 1.2058624029159546,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 36750
+    },
+    {
+      "epoch": 5.942931048419691,
+      "grad_norm": 1.1396408081054688,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 36760
+    },
+    {
+      "epoch": 5.944547732600436,
+      "grad_norm": 1.1510354280471802,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 36770
+    },
+    {
+      "epoch": 5.946164416781182,
+      "grad_norm": 1.1401607990264893,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 36780
+    },
+    {
+      "epoch": 5.947781100961927,
+      "grad_norm": 1.1871325969696045,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 36790
+    },
+    {
+      "epoch": 5.949397785142672,
+      "grad_norm": 0.9928333163261414,
+      "learning_rate": 0.0002,
+      "loss": 0.4945,
+      "step": 36800
+    },
+    {
+      "epoch": 5.951014469323418,
+      "grad_norm": 1.0549445152282715,
+      "learning_rate": 0.0002,
+      "loss": 0.5303,
+      "step": 36810
+    },
+    {
+      "epoch": 5.9526311535041625,
+      "grad_norm": 0.9791563749313354,
+      "learning_rate": 0.0002,
+      "loss": 0.5532,
+      "step": 36820
+    },
+    {
+      "epoch": 5.954247837684909,
+      "grad_norm": 1.1268441677093506,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 36830
+    },
+    {
+      "epoch": 5.955864521865654,
+      "grad_norm": 1.0533992052078247,
+      "learning_rate": 0.0002,
+      "loss": 0.5585,
+      "step": 36840
+    },
+    {
+      "epoch": 5.957481206046399,
+      "grad_norm": 1.023358941078186,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 36850
+    },
+    {
+      "epoch": 5.959097890227144,
+      "grad_norm": 1.2631961107254028,
+      "learning_rate": 0.0002,
+      "loss": 0.5557,
+      "step": 36860
+    },
+    {
+      "epoch": 5.9607145744078895,
+      "grad_norm": 0.9397698640823364,
+      "learning_rate": 0.0002,
+      "loss": 0.5662,
+      "step": 36870
+    },
+    {
+      "epoch": 5.962331258588635,
+      "grad_norm": 1.1678427457809448,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 36880
+    },
+    {
+      "epoch": 5.96394794276938,
+      "grad_norm": 1.1403759717941284,
+      "learning_rate": 0.0002,
+      "loss": 0.5435,
+      "step": 36890
+    },
+    {
+      "epoch": 5.965564626950125,
+      "grad_norm": 1.030572772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.5479,
+      "step": 36900
+    },
+    {
+      "epoch": 5.9671813111308705,
+      "grad_norm": 1.0992497205734253,
+      "learning_rate": 0.0002,
+      "loss": 0.5838,
+      "step": 36910
+    },
+    {
+      "epoch": 5.968797995311616,
+      "grad_norm": 1.075466275215149,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 36920
+    },
+    {
+      "epoch": 5.970414679492361,
+      "grad_norm": 1.0153694152832031,
+      "learning_rate": 0.0002,
+      "loss": 0.5739,
+      "step": 36930
+    },
+    {
+      "epoch": 5.972031363673106,
+      "grad_norm": 0.973193883895874,
+      "learning_rate": 0.0002,
+      "loss": 0.5672,
+      "step": 36940
+    },
+    {
+      "epoch": 5.973648047853851,
+      "grad_norm": 0.8294678926467896,
+      "learning_rate": 0.0002,
+      "loss": 0.5585,
+      "step": 36950
+    },
+    {
+      "epoch": 5.9752647320345975,
+      "grad_norm": 1.0048716068267822,
+      "learning_rate": 0.0002,
+      "loss": 0.5631,
+      "step": 36960
+    },
+    {
+      "epoch": 5.976881416215342,
+      "grad_norm": 0.9714070558547974,
+      "learning_rate": 0.0002,
+      "loss": 0.5471,
+      "step": 36970
+    },
+    {
+      "epoch": 5.978498100396088,
+      "grad_norm": 0.8667682409286499,
+      "learning_rate": 0.0002,
+      "loss": 0.5419,
+      "step": 36980
+    },
+    {
+      "epoch": 5.980114784576833,
+      "grad_norm": 1.0461409091949463,
+      "learning_rate": 0.0002,
+      "loss": 0.5474,
+      "step": 36990
+    },
+    {
+      "epoch": 5.981731468757578,
+      "grad_norm": 0.9229754209518433,
+      "learning_rate": 0.0002,
+      "loss": 0.5454,
+      "step": 37000
+    },
+    {
+      "epoch": 5.983348152938324,
+      "grad_norm": 1.0406876802444458,
+      "learning_rate": 0.0002,
+      "loss": 0.5599,
+      "step": 37010
+    },
+    {
+      "epoch": 5.984964837119069,
+      "grad_norm": 0.8993828296661377,
+      "learning_rate": 0.0002,
+      "loss": 0.5569,
+      "step": 37020
+    },
+    {
+      "epoch": 5.986581521299814,
+      "grad_norm": 1.2260479927062988,
+      "learning_rate": 0.0002,
+      "loss": 0.5611,
+      "step": 37030
+    },
+    {
+      "epoch": 5.988198205480559,
+      "grad_norm": 1.0107380151748657,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 37040
+    },
+    {
+      "epoch": 5.989814889661305,
+      "grad_norm": 1.0240139961242676,
+      "learning_rate": 0.0002,
+      "loss": 0.5639,
+      "step": 37050
+    },
+    {
+      "epoch": 5.99143157384205,
+      "grad_norm": 1.0185275077819824,
+      "learning_rate": 0.0002,
+      "loss": 0.5209,
+      "step": 37060
+    },
+    {
+      "epoch": 5.993048258022795,
+      "grad_norm": 1.1361802816390991,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 37070
+    },
+    {
+      "epoch": 5.99466494220354,
+      "grad_norm": 1.0395532846450806,
+      "learning_rate": 0.0002,
+      "loss": 0.5692,
+      "step": 37080
+    },
+    {
+      "epoch": 5.9962816263842855,
+      "grad_norm": 0.9463558197021484,
+      "learning_rate": 0.0002,
+      "loss": 0.594,
+      "step": 37090
+    },
+    {
+      "epoch": 5.997898310565031,
+      "grad_norm": 1.2066948413848877,
+      "learning_rate": 0.0002,
+      "loss": 0.5775,
+      "step": 37100
+    },
+    {
+      "epoch": 5.999514994745777,
+      "grad_norm": 0.9749386310577393,
+      "learning_rate": 0.0002,
+      "loss": 0.5356,
+      "step": 37110
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 1.2270219326019287,
+      "eval_runtime": 122.2047,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 37113
+    },
+    {
+      "epoch": 6.001131678926522,
+      "grad_norm": 0.9641092419624329,
+      "learning_rate": 0.0002,
+      "loss": 0.4855,
+      "step": 37120
+    },
+    {
+      "epoch": 6.002748363107267,
+      "grad_norm": 1.103379249572754,
+      "learning_rate": 0.0002,
+      "loss": 0.4112,
+      "step": 37130
+    },
+    {
+      "epoch": 6.004365047288013,
+      "grad_norm": 0.8381665349006653,
+      "learning_rate": 0.0002,
+      "loss": 0.4577,
+      "step": 37140
+    },
+    {
+      "epoch": 6.005981731468758,
+      "grad_norm": 1.245323896408081,
+      "learning_rate": 0.0002,
+      "loss": 0.4794,
+      "step": 37150
+    },
+    {
+      "epoch": 6.007598415649503,
+      "grad_norm": 1.3140289783477783,
+      "learning_rate": 0.0002,
+      "loss": 0.4503,
+      "step": 37160
+    },
+    {
+      "epoch": 6.009215099830248,
+      "grad_norm": 0.8479695916175842,
+      "learning_rate": 0.0002,
+      "loss": 0.4456,
+      "step": 37170
+    },
+    {
+      "epoch": 6.0108317840109935,
+      "grad_norm": 0.8841437101364136,
+      "learning_rate": 0.0002,
+      "loss": 0.4573,
+      "step": 37180
+    },
+    {
+      "epoch": 6.012448468191739,
+      "grad_norm": 0.8900154829025269,
+      "learning_rate": 0.0002,
+      "loss": 0.4565,
+      "step": 37190
+    },
+    {
+      "epoch": 6.014065152372484,
+      "grad_norm": 1.2753345966339111,
+      "learning_rate": 0.0002,
+      "loss": 0.457,
+      "step": 37200
+    },
+    {
+      "epoch": 6.015681836553229,
+      "grad_norm": 1.4625498056411743,
+      "learning_rate": 0.0002,
+      "loss": 0.4365,
+      "step": 37210
+    },
+    {
+      "epoch": 6.017298520733974,
+      "grad_norm": 0.7455034852027893,
+      "learning_rate": 0.0002,
+      "loss": 0.4252,
+      "step": 37220
+    },
+    {
+      "epoch": 6.01891520491472,
+      "grad_norm": 1.1658862829208374,
+      "learning_rate": 0.0002,
+      "loss": 0.4433,
+      "step": 37230
+    },
+    {
+      "epoch": 6.020531889095465,
+      "grad_norm": 0.9785751104354858,
+      "learning_rate": 0.0002,
+      "loss": 0.4499,
+      "step": 37240
+    },
+    {
+      "epoch": 6.02214857327621,
+      "grad_norm": 1.3193122148513794,
+      "learning_rate": 0.0002,
+      "loss": 0.4956,
+      "step": 37250
+    },
+    {
+      "epoch": 6.023765257456955,
+      "grad_norm": 1.038273572921753,
+      "learning_rate": 0.0002,
+      "loss": 0.4727,
+      "step": 37260
+    },
+    {
+      "epoch": 6.0253819416377015,
+      "grad_norm": 1.0550594329833984,
+      "learning_rate": 0.0002,
+      "loss": 0.4395,
+      "step": 37270
+    },
+    {
+      "epoch": 6.026998625818447,
+      "grad_norm": 0.9745930433273315,
+      "learning_rate": 0.0002,
+      "loss": 0.4767,
+      "step": 37280
+    },
+    {
+      "epoch": 6.028615309999192,
+      "grad_norm": 0.9273530840873718,
+      "learning_rate": 0.0002,
+      "loss": 0.4233,
+      "step": 37290
+    },
+    {
+      "epoch": 6.030231994179937,
+      "grad_norm": 1.3844057321548462,
+      "learning_rate": 0.0002,
+      "loss": 0.4195,
+      "step": 37300
+    },
+    {
+      "epoch": 6.031848678360682,
+      "grad_norm": 1.2058762311935425,
+      "learning_rate": 0.0002,
+      "loss": 0.4768,
+      "step": 37310
+    },
+    {
+      "epoch": 6.033465362541428,
+      "grad_norm": 1.242663025856018,
+      "learning_rate": 0.0002,
+      "loss": 0.4499,
+      "step": 37320
+    },
+    {
+      "epoch": 6.035082046722173,
+      "grad_norm": 1.3504270315170288,
+      "learning_rate": 0.0002,
+      "loss": 0.4597,
+      "step": 37330
+    },
+    {
+      "epoch": 6.036698730902918,
+      "grad_norm": 0.8734912276268005,
+      "learning_rate": 0.0002,
+      "loss": 0.4402,
+      "step": 37340
+    },
+    {
+      "epoch": 6.038315415083663,
+      "grad_norm": 1.0182311534881592,
+      "learning_rate": 0.0002,
+      "loss": 0.477,
+      "step": 37350
+    },
+    {
+      "epoch": 6.0399320992644085,
+      "grad_norm": 0.9898499846458435,
+      "learning_rate": 0.0002,
+      "loss": 0.4261,
+      "step": 37360
+    },
+    {
+      "epoch": 6.041548783445154,
+      "grad_norm": 1.0637860298156738,
+      "learning_rate": 0.0002,
+      "loss": 0.4459,
+      "step": 37370
+    },
+    {
+      "epoch": 6.043165467625899,
+      "grad_norm": 1.0099523067474365,
+      "learning_rate": 0.0002,
+      "loss": 0.4958,
+      "step": 37380
+    },
+    {
+      "epoch": 6.044782151806644,
+      "grad_norm": 1.1080750226974487,
+      "learning_rate": 0.0002,
+      "loss": 0.4459,
+      "step": 37390
+    },
+    {
+      "epoch": 6.0463988359873895,
+      "grad_norm": 1.2551289796829224,
+      "learning_rate": 0.0002,
+      "loss": 0.4473,
+      "step": 37400
+    },
+    {
+      "epoch": 6.048015520168136,
+      "grad_norm": 0.8959632515907288,
+      "learning_rate": 0.0002,
+      "loss": 0.468,
+      "step": 37410
+    },
+    {
+      "epoch": 6.049632204348881,
+      "grad_norm": 1.1748892068862915,
+      "learning_rate": 0.0002,
+      "loss": 0.4255,
+      "step": 37420
+    },
+    {
+      "epoch": 6.051248888529626,
+      "grad_norm": 1.3122745752334595,
+      "learning_rate": 0.0002,
+      "loss": 0.4458,
+      "step": 37430
+    },
+    {
+      "epoch": 6.052865572710371,
+      "grad_norm": 1.0227985382080078,
+      "learning_rate": 0.0002,
+      "loss": 0.4676,
+      "step": 37440
+    },
+    {
+      "epoch": 6.0544822568911165,
+      "grad_norm": 1.0380030870437622,
+      "learning_rate": 0.0002,
+      "loss": 0.4503,
+      "step": 37450
+    },
+    {
+      "epoch": 6.056098941071862,
+      "grad_norm": 0.8919622898101807,
+      "learning_rate": 0.0002,
+      "loss": 0.4686,
+      "step": 37460
+    },
+    {
+      "epoch": 6.057715625252607,
+      "grad_norm": 1.4554150104522705,
+      "learning_rate": 0.0002,
+      "loss": 0.4406,
+      "step": 37470
+    },
+    {
+      "epoch": 6.059332309433352,
+      "grad_norm": 1.2853292226791382,
+      "learning_rate": 0.0002,
+      "loss": 0.4688,
+      "step": 37480
+    },
+    {
+      "epoch": 6.0609489936140974,
+      "grad_norm": 1.2951840162277222,
+      "learning_rate": 0.0002,
+      "loss": 0.4489,
+      "step": 37490
+    },
+    {
+      "epoch": 6.062565677794843,
+      "grad_norm": 1.1750973463058472,
+      "learning_rate": 0.0002,
+      "loss": 0.4819,
+      "step": 37500
+    },
+    {
+      "epoch": 6.064182361975588,
+      "grad_norm": 0.9328424334526062,
+      "learning_rate": 0.0002,
+      "loss": 0.4574,
+      "step": 37510
+    },
+    {
+      "epoch": 6.065799046156333,
+      "grad_norm": 1.0353537797927856,
+      "learning_rate": 0.0002,
+      "loss": 0.4597,
+      "step": 37520
+    },
+    {
+      "epoch": 6.067415730337078,
+      "grad_norm": 1.1594274044036865,
+      "learning_rate": 0.0002,
+      "loss": 0.4407,
+      "step": 37530
+    },
+    {
+      "epoch": 6.069032414517824,
+      "grad_norm": 0.9034168124198914,
+      "learning_rate": 0.0002,
+      "loss": 0.4642,
+      "step": 37540
+    },
+    {
+      "epoch": 6.070649098698569,
+      "grad_norm": 1.068617820739746,
+      "learning_rate": 0.0002,
+      "loss": 0.4625,
+      "step": 37550
+    },
+    {
+      "epoch": 6.072265782879315,
+      "grad_norm": 1.0931321382522583,
+      "learning_rate": 0.0002,
+      "loss": 0.4378,
+      "step": 37560
+    },
+    {
+      "epoch": 6.07388246706006,
+      "grad_norm": 1.2542688846588135,
+      "learning_rate": 0.0002,
+      "loss": 0.4527,
+      "step": 37570
+    },
+    {
+      "epoch": 6.075499151240805,
+      "grad_norm": 1.273384928703308,
+      "learning_rate": 0.0002,
+      "loss": 0.4725,
+      "step": 37580
+    },
+    {
+      "epoch": 6.077115835421551,
+      "grad_norm": 1.4771400690078735,
+      "learning_rate": 0.0002,
+      "loss": 0.4928,
+      "step": 37590
+    },
+    {
+      "epoch": 6.078732519602296,
+      "grad_norm": 1.3751444816589355,
+      "learning_rate": 0.0002,
+      "loss": 0.461,
+      "step": 37600
+    },
+    {
+      "epoch": 6.080349203783041,
+      "grad_norm": 1.4532550573349,
+      "learning_rate": 0.0002,
+      "loss": 0.4602,
+      "step": 37610
+    },
+    {
+      "epoch": 6.081965887963786,
+      "grad_norm": 1.3175991773605347,
+      "learning_rate": 0.0002,
+      "loss": 0.4428,
+      "step": 37620
+    },
+    {
+      "epoch": 6.083582572144532,
+      "grad_norm": 1.0624970197677612,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 37630
+    },
+    {
+      "epoch": 6.085199256325277,
+      "grad_norm": 1.099715232849121,
+      "learning_rate": 0.0002,
+      "loss": 0.413,
+      "step": 37640
+    },
+    {
+      "epoch": 6.086815940506022,
+      "grad_norm": 1.0380114316940308,
+      "learning_rate": 0.0002,
+      "loss": 0.4528,
+      "step": 37650
+    },
+    {
+      "epoch": 6.088432624686767,
+      "grad_norm": 1.1136109828948975,
+      "learning_rate": 0.0002,
+      "loss": 0.4373,
+      "step": 37660
+    },
+    {
+      "epoch": 6.0900493088675125,
+      "grad_norm": 0.996498703956604,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 37670
+    },
+    {
+      "epoch": 6.091665993048258,
+      "grad_norm": 1.0552574396133423,
+      "learning_rate": 0.0002,
+      "loss": 0.4713,
+      "step": 37680
+    },
+    {
+      "epoch": 6.093282677229003,
+      "grad_norm": 1.4108527898788452,
+      "learning_rate": 0.0002,
+      "loss": 0.4414,
+      "step": 37690
+    },
+    {
+      "epoch": 6.094899361409748,
+      "grad_norm": 1.1323093175888062,
+      "learning_rate": 0.0002,
+      "loss": 0.4851,
+      "step": 37700
+    },
+    {
+      "epoch": 6.096516045590494,
+      "grad_norm": 0.9364377856254578,
+      "learning_rate": 0.0002,
+      "loss": 0.4455,
+      "step": 37710
+    },
+    {
+      "epoch": 6.0981327297712395,
+      "grad_norm": 1.1300561428070068,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 37720
+    },
+    {
+      "epoch": 6.099749413951985,
+      "grad_norm": 1.0616047382354736,
+      "learning_rate": 0.0002,
+      "loss": 0.4539,
+      "step": 37730
+    },
+    {
+      "epoch": 6.10136609813273,
+      "grad_norm": 1.1205905675888062,
+      "learning_rate": 0.0002,
+      "loss": 0.4516,
+      "step": 37740
+    },
+    {
+      "epoch": 6.102982782313475,
+      "grad_norm": 0.9592534303665161,
+      "learning_rate": 0.0002,
+      "loss": 0.4688,
+      "step": 37750
+    },
+    {
+      "epoch": 6.1045994664942205,
+      "grad_norm": 0.9797531962394714,
+      "learning_rate": 0.0002,
+      "loss": 0.4494,
+      "step": 37760
+    },
+    {
+      "epoch": 6.106216150674966,
+      "grad_norm": 1.093404769897461,
+      "learning_rate": 0.0002,
+      "loss": 0.4237,
+      "step": 37770
+    },
+    {
+      "epoch": 6.107832834855711,
+      "grad_norm": 1.2172642946243286,
+      "learning_rate": 0.0002,
+      "loss": 0.4691,
+      "step": 37780
+    },
+    {
+      "epoch": 6.109449519036456,
+      "grad_norm": 1.0467255115509033,
+      "learning_rate": 0.0002,
+      "loss": 0.4398,
+      "step": 37790
+    },
+    {
+      "epoch": 6.111066203217201,
+      "grad_norm": 1.159318208694458,
+      "learning_rate": 0.0002,
+      "loss": 0.4676,
+      "step": 37800
+    },
+    {
+      "epoch": 6.112682887397947,
+      "grad_norm": 1.0615603923797607,
+      "learning_rate": 0.0002,
+      "loss": 0.4539,
+      "step": 37810
+    },
+    {
+      "epoch": 6.114299571578692,
+      "grad_norm": 1.0542045831680298,
+      "learning_rate": 0.0002,
+      "loss": 0.4957,
+      "step": 37820
+    },
+    {
+      "epoch": 6.115916255759437,
+      "grad_norm": 0.8962697982788086,
+      "learning_rate": 0.0002,
+      "loss": 0.4512,
+      "step": 37830
+    },
+    {
+      "epoch": 6.117532939940182,
+      "grad_norm": 1.106352686882019,
+      "learning_rate": 0.0002,
+      "loss": 0.4519,
+      "step": 37840
+    },
+    {
+      "epoch": 6.1191496241209276,
+      "grad_norm": 1.1660276651382446,
+      "learning_rate": 0.0002,
+      "loss": 0.4421,
+      "step": 37850
+    },
+    {
+      "epoch": 6.120766308301674,
+      "grad_norm": 1.3524385690689087,
+      "learning_rate": 0.0002,
+      "loss": 0.4701,
+      "step": 37860
+    },
+    {
+      "epoch": 6.122382992482419,
+      "grad_norm": 1.1056050062179565,
+      "learning_rate": 0.0002,
+      "loss": 0.4684,
+      "step": 37870
+    },
+    {
+      "epoch": 6.123999676663164,
+      "grad_norm": 1.0772725343704224,
+      "learning_rate": 0.0002,
+      "loss": 0.4518,
+      "step": 37880
+    },
+    {
+      "epoch": 6.125616360843909,
+      "grad_norm": 1.1011115312576294,
+      "learning_rate": 0.0002,
+      "loss": 0.4356,
+      "step": 37890
+    },
+    {
+      "epoch": 6.127233045024655,
+      "grad_norm": 0.8952536582946777,
+      "learning_rate": 0.0002,
+      "loss": 0.4909,
+      "step": 37900
+    },
+    {
+      "epoch": 6.1288497292054,
+      "grad_norm": 1.244398593902588,
+      "learning_rate": 0.0002,
+      "loss": 0.4299,
+      "step": 37910
+    },
+    {
+      "epoch": 6.130466413386145,
+      "grad_norm": 0.9658283591270447,
+      "learning_rate": 0.0002,
+      "loss": 0.4764,
+      "step": 37920
+    },
+    {
+      "epoch": 6.13208309756689,
+      "grad_norm": 1.0649068355560303,
+      "learning_rate": 0.0002,
+      "loss": 0.4378,
+      "step": 37930
+    },
+    {
+      "epoch": 6.1336997817476355,
+      "grad_norm": 0.94698166847229,
+      "learning_rate": 0.0002,
+      "loss": 0.4638,
+      "step": 37940
+    },
+    {
+      "epoch": 6.135316465928381,
+      "grad_norm": 1.1450897455215454,
+      "learning_rate": 0.0002,
+      "loss": 0.488,
+      "step": 37950
+    },
+    {
+      "epoch": 6.136933150109126,
+      "grad_norm": 1.032482624053955,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 37960
+    },
+    {
+      "epoch": 6.138549834289871,
+      "grad_norm": 1.0993428230285645,
+      "learning_rate": 0.0002,
+      "loss": 0.4179,
+      "step": 37970
+    },
+    {
+      "epoch": 6.1401665184706165,
+      "grad_norm": 1.2907029390335083,
+      "learning_rate": 0.0002,
+      "loss": 0.4781,
+      "step": 37980
+    },
+    {
+      "epoch": 6.141783202651362,
+      "grad_norm": 1.1007903814315796,
+      "learning_rate": 0.0002,
+      "loss": 0.4671,
+      "step": 37990
+    },
+    {
+      "epoch": 6.143399886832107,
+      "grad_norm": 0.9286124110221863,
+      "learning_rate": 0.0002,
+      "loss": 0.4213,
+      "step": 38000
+    },
+    {
+      "epoch": 6.145016571012853,
+      "grad_norm": 1.1426366567611694,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 38010
+    },
+    {
+      "epoch": 6.146633255193598,
+      "grad_norm": 1.2608287334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 38020
+    },
+    {
+      "epoch": 6.1482499393743435,
+      "grad_norm": 1.1346837282180786,
+      "learning_rate": 0.0002,
+      "loss": 0.454,
+      "step": 38030
+    },
+    {
+      "epoch": 6.149866623555089,
+      "grad_norm": 1.144080400466919,
+      "learning_rate": 0.0002,
+      "loss": 0.4469,
+      "step": 38040
+    },
+    {
+      "epoch": 6.151483307735834,
+      "grad_norm": 1.3456705808639526,
+      "learning_rate": 0.0002,
+      "loss": 0.4515,
+      "step": 38050
+    },
+    {
+      "epoch": 6.153099991916579,
+      "grad_norm": 1.0517960786819458,
+      "learning_rate": 0.0002,
+      "loss": 0.4775,
+      "step": 38060
+    },
+    {
+      "epoch": 6.154716676097324,
+      "grad_norm": 1.1887445449829102,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 38070
+    },
+    {
+      "epoch": 6.15633336027807,
+      "grad_norm": 1.0449163913726807,
+      "learning_rate": 0.0002,
+      "loss": 0.4516,
+      "step": 38080
+    },
+    {
+      "epoch": 6.157950044458815,
+      "grad_norm": 1.3218743801116943,
+      "learning_rate": 0.0002,
+      "loss": 0.4808,
+      "step": 38090
+    },
+    {
+      "epoch": 6.15956672863956,
+      "grad_norm": 1.003208875656128,
+      "learning_rate": 0.0002,
+      "loss": 0.4632,
+      "step": 38100
+    },
+    {
+      "epoch": 6.161183412820305,
+      "grad_norm": 1.008623719215393,
+      "learning_rate": 0.0002,
+      "loss": 0.4978,
+      "step": 38110
+    },
+    {
+      "epoch": 6.162800097001051,
+      "grad_norm": 1.2122787237167358,
+      "learning_rate": 0.0002,
+      "loss": 0.4608,
+      "step": 38120
+    },
+    {
+      "epoch": 6.164416781181796,
+      "grad_norm": 1.253403902053833,
+      "learning_rate": 0.0002,
+      "loss": 0.4666,
+      "step": 38130
+    },
+    {
+      "epoch": 6.166033465362541,
+      "grad_norm": 1.2289724349975586,
+      "learning_rate": 0.0002,
+      "loss": 0.4778,
+      "step": 38140
+    },
+    {
+      "epoch": 6.167650149543286,
+      "grad_norm": 1.330694556236267,
+      "learning_rate": 0.0002,
+      "loss": 0.4774,
+      "step": 38150
+    },
+    {
+      "epoch": 6.169266833724032,
+      "grad_norm": 1.0946741104125977,
+      "learning_rate": 0.0002,
+      "loss": 0.4699,
+      "step": 38160
+    },
+    {
+      "epoch": 6.170883517904778,
+      "grad_norm": 1.0719934701919556,
+      "learning_rate": 0.0002,
+      "loss": 0.4816,
+      "step": 38170
+    },
+    {
+      "epoch": 6.172500202085523,
+      "grad_norm": 1.1142133474349976,
+      "learning_rate": 0.0002,
+      "loss": 0.4678,
+      "step": 38180
+    },
+    {
+      "epoch": 6.174116886266268,
+      "grad_norm": 1.1221938133239746,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 38190
+    },
+    {
+      "epoch": 6.175733570447013,
+      "grad_norm": 1.1391617059707642,
+      "learning_rate": 0.0002,
+      "loss": 0.4462,
+      "step": 38200
+    },
+    {
+      "epoch": 6.1773502546277586,
+      "grad_norm": 1.2263455390930176,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 38210
+    },
+    {
+      "epoch": 6.178966938808504,
+      "grad_norm": 1.0930434465408325,
+      "learning_rate": 0.0002,
+      "loss": 0.4633,
+      "step": 38220
+    },
+    {
+      "epoch": 6.180583622989249,
+      "grad_norm": 1.3489030599594116,
+      "learning_rate": 0.0002,
+      "loss": 0.4406,
+      "step": 38230
+    },
+    {
+      "epoch": 6.182200307169994,
+      "grad_norm": 1.1383486986160278,
+      "learning_rate": 0.0002,
+      "loss": 0.4994,
+      "step": 38240
+    },
+    {
+      "epoch": 6.1838169913507395,
+      "grad_norm": 1.2408897876739502,
+      "learning_rate": 0.0002,
+      "loss": 0.4851,
+      "step": 38250
+    },
+    {
+      "epoch": 6.185433675531485,
+      "grad_norm": 1.1436222791671753,
+      "learning_rate": 0.0002,
+      "loss": 0.4848,
+      "step": 38260
+    },
+    {
+      "epoch": 6.18705035971223,
+      "grad_norm": 1.370117425918579,
+      "learning_rate": 0.0002,
+      "loss": 0.4594,
+      "step": 38270
+    },
+    {
+      "epoch": 6.188667043892975,
+      "grad_norm": 0.8862423300743103,
+      "learning_rate": 0.0002,
+      "loss": 0.5023,
+      "step": 38280
+    },
+    {
+      "epoch": 6.19028372807372,
+      "grad_norm": 0.9603779315948486,
+      "learning_rate": 0.0002,
+      "loss": 0.4559,
+      "step": 38290
+    },
+    {
+      "epoch": 6.191900412254466,
+      "grad_norm": 1.389291524887085,
+      "learning_rate": 0.0002,
+      "loss": 0.4835,
+      "step": 38300
+    },
+    {
+      "epoch": 6.193517096435212,
+      "grad_norm": 1.0767031908035278,
+      "learning_rate": 0.0002,
+      "loss": 0.4435,
+      "step": 38310
+    },
+    {
+      "epoch": 6.195133780615957,
+      "grad_norm": 1.1800403594970703,
+      "learning_rate": 0.0002,
+      "loss": 0.4683,
+      "step": 38320
+    },
+    {
+      "epoch": 6.196750464796702,
+      "grad_norm": 0.997891366481781,
+      "learning_rate": 0.0002,
+      "loss": 0.4608,
+      "step": 38330
+    },
+    {
+      "epoch": 6.1983671489774474,
+      "grad_norm": 1.1201492547988892,
+      "learning_rate": 0.0002,
+      "loss": 0.4575,
+      "step": 38340
+    },
+    {
+      "epoch": 6.199983833158193,
+      "grad_norm": 0.9769026637077332,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 38350
+    },
+    {
+      "epoch": 6.201600517338938,
+      "grad_norm": 0.9447069764137268,
+      "learning_rate": 0.0002,
+      "loss": 0.4563,
+      "step": 38360
+    },
+    {
+      "epoch": 6.203217201519683,
+      "grad_norm": 1.0959235429763794,
+      "learning_rate": 0.0002,
+      "loss": 0.516,
+      "step": 38370
+    },
+    {
+      "epoch": 6.204833885700428,
+      "grad_norm": 1.2495406866073608,
+      "learning_rate": 0.0002,
+      "loss": 0.4688,
+      "step": 38380
+    },
+    {
+      "epoch": 6.206450569881174,
+      "grad_norm": 0.8589218258857727,
+      "learning_rate": 0.0002,
+      "loss": 0.4445,
+      "step": 38390
+    },
+    {
+      "epoch": 6.208067254061919,
+      "grad_norm": 0.959155797958374,
+      "learning_rate": 0.0002,
+      "loss": 0.4808,
+      "step": 38400
+    },
+    {
+      "epoch": 6.209683938242664,
+      "grad_norm": 1.0105533599853516,
+      "learning_rate": 0.0002,
+      "loss": 0.4622,
+      "step": 38410
+    },
+    {
+      "epoch": 6.211300622423409,
+      "grad_norm": 0.9824615120887756,
+      "learning_rate": 0.0002,
+      "loss": 0.4887,
+      "step": 38420
+    },
+    {
+      "epoch": 6.2129173066041545,
+      "grad_norm": 0.8616500496864319,
+      "learning_rate": 0.0002,
+      "loss": 0.4656,
+      "step": 38430
+    },
+    {
+      "epoch": 6.2145339907849,
+      "grad_norm": 1.2917758226394653,
+      "learning_rate": 0.0002,
+      "loss": 0.449,
+      "step": 38440
+    },
+    {
+      "epoch": 6.216150674965646,
+      "grad_norm": 1.0564531087875366,
+      "learning_rate": 0.0002,
+      "loss": 0.4201,
+      "step": 38450
+    },
+    {
+      "epoch": 6.217767359146391,
+      "grad_norm": 1.152331829071045,
+      "learning_rate": 0.0002,
+      "loss": 0.4849,
+      "step": 38460
+    },
+    {
+      "epoch": 6.219384043327136,
+      "grad_norm": 0.9152206778526306,
+      "learning_rate": 0.0002,
+      "loss": 0.4887,
+      "step": 38470
+    },
+    {
+      "epoch": 6.221000727507882,
+      "grad_norm": 0.9931167960166931,
+      "learning_rate": 0.0002,
+      "loss": 0.4686,
+      "step": 38480
+    },
+    {
+      "epoch": 6.222617411688627,
+      "grad_norm": 1.3248072862625122,
+      "learning_rate": 0.0002,
+      "loss": 0.4765,
+      "step": 38490
+    },
+    {
+      "epoch": 6.224234095869372,
+      "grad_norm": 1.3916507959365845,
+      "learning_rate": 0.0002,
+      "loss": 0.4636,
+      "step": 38500
+    },
+    {
+      "epoch": 6.225850780050117,
+      "grad_norm": 1.1775140762329102,
+      "learning_rate": 0.0002,
+      "loss": 0.506,
+      "step": 38510
+    },
+    {
+      "epoch": 6.2274674642308625,
+      "grad_norm": 1.1581059694290161,
+      "learning_rate": 0.0002,
+      "loss": 0.47,
+      "step": 38520
+    },
+    {
+      "epoch": 6.229084148411608,
+      "grad_norm": 1.359320878982544,
+      "learning_rate": 0.0002,
+      "loss": 0.4679,
+      "step": 38530
+    },
+    {
+      "epoch": 6.230700832592353,
+      "grad_norm": 1.185041904449463,
+      "learning_rate": 0.0002,
+      "loss": 0.4697,
+      "step": 38540
+    },
+    {
+      "epoch": 6.232317516773098,
+      "grad_norm": 1.1861097812652588,
+      "learning_rate": 0.0002,
+      "loss": 0.4815,
+      "step": 38550
+    },
+    {
+      "epoch": 6.233934200953843,
+      "grad_norm": 1.126990556716919,
+      "learning_rate": 0.0002,
+      "loss": 0.4925,
+      "step": 38560
+    },
+    {
+      "epoch": 6.235550885134589,
+      "grad_norm": 0.9744541049003601,
+      "learning_rate": 0.0002,
+      "loss": 0.4414,
+      "step": 38570
+    },
+    {
+      "epoch": 6.237167569315334,
+      "grad_norm": 1.1260887384414673,
+      "learning_rate": 0.0002,
+      "loss": 0.4577,
+      "step": 38580
+    },
+    {
+      "epoch": 6.238784253496079,
+      "grad_norm": 1.1290327310562134,
+      "learning_rate": 0.0002,
+      "loss": 0.4852,
+      "step": 38590
+    },
+    {
+      "epoch": 6.240400937676825,
+      "grad_norm": 1.0952879190444946,
+      "learning_rate": 0.0002,
+      "loss": 0.4805,
+      "step": 38600
+    },
+    {
+      "epoch": 6.2420176218575705,
+      "grad_norm": 1.1037684679031372,
+      "learning_rate": 0.0002,
+      "loss": 0.4436,
+      "step": 38610
+    },
+    {
+      "epoch": 6.243634306038316,
+      "grad_norm": 1.1356085538864136,
+      "learning_rate": 0.0002,
+      "loss": 0.466,
+      "step": 38620
+    },
+    {
+      "epoch": 6.245250990219061,
+      "grad_norm": 1.0677106380462646,
+      "learning_rate": 0.0002,
+      "loss": 0.5129,
+      "step": 38630
+    },
+    {
+      "epoch": 6.246867674399806,
+      "grad_norm": 1.1573411226272583,
+      "learning_rate": 0.0002,
+      "loss": 0.4907,
+      "step": 38640
+    },
+    {
+      "epoch": 6.248484358580551,
+      "grad_norm": 1.2707505226135254,
+      "learning_rate": 0.0002,
+      "loss": 0.5098,
+      "step": 38650
+    },
+    {
+      "epoch": 6.250101042761297,
+      "grad_norm": 1.0480109453201294,
+      "learning_rate": 0.0002,
+      "loss": 0.4926,
+      "step": 38660
+    },
+    {
+      "epoch": 6.251717726942042,
+      "grad_norm": 1.3668724298477173,
+      "learning_rate": 0.0002,
+      "loss": 0.4654,
+      "step": 38670
+    },
+    {
+      "epoch": 6.253334411122787,
+      "grad_norm": 1.217289686203003,
+      "learning_rate": 0.0002,
+      "loss": 0.5128,
+      "step": 38680
+    },
+    {
+      "epoch": 6.254951095303532,
+      "grad_norm": 1.2950236797332764,
+      "learning_rate": 0.0002,
+      "loss": 0.4621,
+      "step": 38690
+    },
+    {
+      "epoch": 6.256567779484278,
+      "grad_norm": 1.4506934881210327,
+      "learning_rate": 0.0002,
+      "loss": 0.5076,
+      "step": 38700
+    },
+    {
+      "epoch": 6.258184463665023,
+      "grad_norm": 1.1248667240142822,
+      "learning_rate": 0.0002,
+      "loss": 0.4803,
+      "step": 38710
+    },
+    {
+      "epoch": 6.259801147845768,
+      "grad_norm": 1.3384023904800415,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 38720
+    },
+    {
+      "epoch": 6.261417832026513,
+      "grad_norm": 1.128074288368225,
+      "learning_rate": 0.0002,
+      "loss": 0.473,
+      "step": 38730
+    },
+    {
+      "epoch": 6.263034516207259,
+      "grad_norm": 1.1169012784957886,
+      "learning_rate": 0.0002,
+      "loss": 0.4638,
+      "step": 38740
+    },
+    {
+      "epoch": 6.264651200388005,
+      "grad_norm": 1.195198893547058,
+      "learning_rate": 0.0002,
+      "loss": 0.4747,
+      "step": 38750
+    },
+    {
+      "epoch": 6.26626788456875,
+      "grad_norm": 1.2471518516540527,
+      "learning_rate": 0.0002,
+      "loss": 0.4906,
+      "step": 38760
+    },
+    {
+      "epoch": 6.267884568749495,
+      "grad_norm": 1.2646394968032837,
+      "learning_rate": 0.0002,
+      "loss": 0.4507,
+      "step": 38770
+    },
+    {
+      "epoch": 6.26950125293024,
+      "grad_norm": 1.0286450386047363,
+      "learning_rate": 0.0002,
+      "loss": 0.4934,
+      "step": 38780
+    },
+    {
+      "epoch": 6.2711179371109855,
+      "grad_norm": 1.2440695762634277,
+      "learning_rate": 0.0002,
+      "loss": 0.4787,
+      "step": 38790
+    },
+    {
+      "epoch": 6.272734621291731,
+      "grad_norm": 0.8941256403923035,
+      "learning_rate": 0.0002,
+      "loss": 0.4806,
+      "step": 38800
+    },
+    {
+      "epoch": 6.274351305472476,
+      "grad_norm": 1.0693447589874268,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 38810
+    },
+    {
+      "epoch": 6.275967989653221,
+      "grad_norm": 1.0936840772628784,
+      "learning_rate": 0.0002,
+      "loss": 0.4408,
+      "step": 38820
+    },
+    {
+      "epoch": 6.2775846738339665,
+      "grad_norm": 1.0961874723434448,
+      "learning_rate": 0.0002,
+      "loss": 0.4729,
+      "step": 38830
+    },
+    {
+      "epoch": 6.279201358014712,
+      "grad_norm": 1.1465433835983276,
+      "learning_rate": 0.0002,
+      "loss": 0.4504,
+      "step": 38840
+    },
+    {
+      "epoch": 6.280818042195457,
+      "grad_norm": 1.2987004518508911,
+      "learning_rate": 0.0002,
+      "loss": 0.4771,
+      "step": 38850
+    },
+    {
+      "epoch": 6.282434726376202,
+      "grad_norm": 1.1310304403305054,
+      "learning_rate": 0.0002,
+      "loss": 0.4945,
+      "step": 38860
+    },
+    {
+      "epoch": 6.284051410556947,
+      "grad_norm": 1.306538462638855,
+      "learning_rate": 0.0002,
+      "loss": 0.5346,
+      "step": 38870
+    },
+    {
+      "epoch": 6.285668094737693,
+      "grad_norm": 1.2405401468276978,
+      "learning_rate": 0.0002,
+      "loss": 0.4873,
+      "step": 38880
+    },
+    {
+      "epoch": 6.287284778918439,
+      "grad_norm": 1.0934767723083496,
+      "learning_rate": 0.0002,
+      "loss": 0.4929,
+      "step": 38890
+    },
+    {
+      "epoch": 6.288901463099184,
+      "grad_norm": 1.3370496034622192,
+      "learning_rate": 0.0002,
+      "loss": 0.4853,
+      "step": 38900
+    },
+    {
+      "epoch": 6.290518147279929,
+      "grad_norm": 1.0319404602050781,
+      "learning_rate": 0.0002,
+      "loss": 0.4892,
+      "step": 38910
+    },
+    {
+      "epoch": 6.292134831460674,
+      "grad_norm": 0.9734271168708801,
+      "learning_rate": 0.0002,
+      "loss": 0.4685,
+      "step": 38920
+    },
+    {
+      "epoch": 6.29375151564142,
+      "grad_norm": 1.0940454006195068,
+      "learning_rate": 0.0002,
+      "loss": 0.5085,
+      "step": 38930
+    },
+    {
+      "epoch": 6.295368199822165,
+      "grad_norm": 1.036500334739685,
+      "learning_rate": 0.0002,
+      "loss": 0.4985,
+      "step": 38940
+    },
+    {
+      "epoch": 6.29698488400291,
+      "grad_norm": 1.020308256149292,
+      "learning_rate": 0.0002,
+      "loss": 0.4878,
+      "step": 38950
+    },
+    {
+      "epoch": 6.298601568183655,
+      "grad_norm": 1.1416399478912354,
+      "learning_rate": 0.0002,
+      "loss": 0.4668,
+      "step": 38960
+    },
+    {
+      "epoch": 6.300218252364401,
+      "grad_norm": 1.2497479915618896,
+      "learning_rate": 0.0002,
+      "loss": 0.4727,
+      "step": 38970
+    },
+    {
+      "epoch": 6.301834936545146,
+      "grad_norm": 1.1692523956298828,
+      "learning_rate": 0.0002,
+      "loss": 0.4721,
+      "step": 38980
+    },
+    {
+      "epoch": 6.303451620725891,
+      "grad_norm": 1.0693109035491943,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 38990
+    },
+    {
+      "epoch": 6.305068304906636,
+      "grad_norm": 0.8883291482925415,
+      "learning_rate": 0.0002,
+      "loss": 0.4875,
+      "step": 39000
+    },
+    {
+      "epoch": 6.3066849890873815,
+      "grad_norm": 1.1445088386535645,
+      "learning_rate": 0.0002,
+      "loss": 0.5371,
+      "step": 39010
+    },
+    {
+      "epoch": 6.308301673268127,
+      "grad_norm": 1.226792335510254,
+      "learning_rate": 0.0002,
+      "loss": 0.5089,
+      "step": 39020
+    },
+    {
+      "epoch": 6.309918357448872,
+      "grad_norm": 1.0498932600021362,
+      "learning_rate": 0.0002,
+      "loss": 0.474,
+      "step": 39030
+    },
+    {
+      "epoch": 6.311535041629618,
+      "grad_norm": 1.0834535360336304,
+      "learning_rate": 0.0002,
+      "loss": 0.4964,
+      "step": 39040
+    },
+    {
+      "epoch": 6.313151725810363,
+      "grad_norm": 1.144666075706482,
+      "learning_rate": 0.0002,
+      "loss": 0.4733,
+      "step": 39050
+    },
+    {
+      "epoch": 6.3147684099911086,
+      "grad_norm": 1.1468489170074463,
+      "learning_rate": 0.0002,
+      "loss": 0.4784,
+      "step": 39060
+    },
+    {
+      "epoch": 6.316385094171854,
+      "grad_norm": 1.290949821472168,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 39070
+    },
+    {
+      "epoch": 6.318001778352599,
+      "grad_norm": 1.087868094444275,
+      "learning_rate": 0.0002,
+      "loss": 0.5002,
+      "step": 39080
+    },
+    {
+      "epoch": 6.319618462533344,
+      "grad_norm": 1.0156296491622925,
+      "learning_rate": 0.0002,
+      "loss": 0.4944,
+      "step": 39090
+    },
+    {
+      "epoch": 6.3212351467140895,
+      "grad_norm": 1.0805060863494873,
+      "learning_rate": 0.0002,
+      "loss": 0.5019,
+      "step": 39100
+    },
+    {
+      "epoch": 6.322851830894835,
+      "grad_norm": 0.9030579924583435,
+      "learning_rate": 0.0002,
+      "loss": 0.4598,
+      "step": 39110
+    },
+    {
+      "epoch": 6.32446851507558,
+      "grad_norm": 1.1488285064697266,
+      "learning_rate": 0.0002,
+      "loss": 0.4635,
+      "step": 39120
+    },
+    {
+      "epoch": 6.326085199256325,
+      "grad_norm": 1.2050796747207642,
+      "learning_rate": 0.0002,
+      "loss": 0.5368,
+      "step": 39130
+    },
+    {
+      "epoch": 6.32770188343707,
+      "grad_norm": 1.093451738357544,
+      "learning_rate": 0.0002,
+      "loss": 0.4854,
+      "step": 39140
+    },
+    {
+      "epoch": 6.329318567617816,
+      "grad_norm": 1.2046772241592407,
+      "learning_rate": 0.0002,
+      "loss": 0.5055,
+      "step": 39150
+    },
+    {
+      "epoch": 6.330935251798561,
+      "grad_norm": 1.045777678489685,
+      "learning_rate": 0.0002,
+      "loss": 0.4703,
+      "step": 39160
+    },
+    {
+      "epoch": 6.332551935979306,
+      "grad_norm": 1.2008492946624756,
+      "learning_rate": 0.0002,
+      "loss": 0.513,
+      "step": 39170
+    },
+    {
+      "epoch": 6.334168620160051,
+      "grad_norm": 1.0613869428634644,
+      "learning_rate": 0.0002,
+      "loss": 0.4909,
+      "step": 39180
+    },
+    {
+      "epoch": 6.3357853043407975,
+      "grad_norm": 1.058440089225769,
+      "learning_rate": 0.0002,
+      "loss": 0.4708,
+      "step": 39190
+    },
+    {
+      "epoch": 6.337401988521543,
+      "grad_norm": 1.195658802986145,
+      "learning_rate": 0.0002,
+      "loss": 0.4719,
+      "step": 39200
+    },
+    {
+      "epoch": 6.339018672702288,
+      "grad_norm": 1.1595174074172974,
+      "learning_rate": 0.0002,
+      "loss": 0.4901,
+      "step": 39210
+    },
+    {
+      "epoch": 6.340635356883033,
+      "grad_norm": 1.0674750804901123,
+      "learning_rate": 0.0002,
+      "loss": 0.4587,
+      "step": 39220
+    },
+    {
+      "epoch": 6.342252041063778,
+      "grad_norm": 1.3306758403778076,
+      "learning_rate": 0.0002,
+      "loss": 0.4801,
+      "step": 39230
+    },
+    {
+      "epoch": 6.343868725244524,
+      "grad_norm": 1.3582593202590942,
+      "learning_rate": 0.0002,
+      "loss": 0.4839,
+      "step": 39240
+    },
+    {
+      "epoch": 6.345485409425269,
+      "grad_norm": 1.2351572513580322,
+      "learning_rate": 0.0002,
+      "loss": 0.4964,
+      "step": 39250
+    },
+    {
+      "epoch": 6.347102093606014,
+      "grad_norm": 1.3623450994491577,
+      "learning_rate": 0.0002,
+      "loss": 0.4806,
+      "step": 39260
+    },
+    {
+      "epoch": 6.348718777786759,
+      "grad_norm": 1.201270580291748,
+      "learning_rate": 0.0002,
+      "loss": 0.466,
+      "step": 39270
+    },
+    {
+      "epoch": 6.3503354619675045,
+      "grad_norm": 0.9300584197044373,
+      "learning_rate": 0.0002,
+      "loss": 0.4899,
+      "step": 39280
+    },
+    {
+      "epoch": 6.35195214614825,
+      "grad_norm": 0.944525957107544,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 39290
+    },
+    {
+      "epoch": 6.353568830328995,
+      "grad_norm": 1.4263732433319092,
+      "learning_rate": 0.0002,
+      "loss": 0.4954,
+      "step": 39300
+    },
+    {
+      "epoch": 6.35518551450974,
+      "grad_norm": 1.392592191696167,
+      "learning_rate": 0.0002,
+      "loss": 0.4982,
+      "step": 39310
+    },
+    {
+      "epoch": 6.3568021986904855,
+      "grad_norm": 1.0753393173217773,
+      "learning_rate": 0.0002,
+      "loss": 0.4868,
+      "step": 39320
+    },
+    {
+      "epoch": 6.358418882871231,
+      "grad_norm": 1.0088151693344116,
+      "learning_rate": 0.0002,
+      "loss": 0.4896,
+      "step": 39330
+    },
+    {
+      "epoch": 6.360035567051977,
+      "grad_norm": 1.1784582138061523,
+      "learning_rate": 0.0002,
+      "loss": 0.4684,
+      "step": 39340
+    },
+    {
+      "epoch": 6.361652251232722,
+      "grad_norm": 1.020526647567749,
+      "learning_rate": 0.0002,
+      "loss": 0.4732,
+      "step": 39350
+    },
+    {
+      "epoch": 6.363268935413467,
+      "grad_norm": 1.1400747299194336,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 39360
+    },
+    {
+      "epoch": 6.3648856195942125,
+      "grad_norm": 0.9960665702819824,
+      "learning_rate": 0.0002,
+      "loss": 0.4976,
+      "step": 39370
+    },
+    {
+      "epoch": 6.366502303774958,
+      "grad_norm": 1.1547569036483765,
+      "learning_rate": 0.0002,
+      "loss": 0.483,
+      "step": 39380
+    },
+    {
+      "epoch": 6.368118987955703,
+      "grad_norm": 1.2180676460266113,
+      "learning_rate": 0.0002,
+      "loss": 0.4861,
+      "step": 39390
+    },
+    {
+      "epoch": 6.369735672136448,
+      "grad_norm": 1.1391799449920654,
+      "learning_rate": 0.0002,
+      "loss": 0.4805,
+      "step": 39400
+    },
+    {
+      "epoch": 6.371352356317193,
+      "grad_norm": 1.2893574237823486,
+      "learning_rate": 0.0002,
+      "loss": 0.5004,
+      "step": 39410
+    },
+    {
+      "epoch": 6.372969040497939,
+      "grad_norm": 1.192878246307373,
+      "learning_rate": 0.0002,
+      "loss": 0.4807,
+      "step": 39420
+    },
+    {
+      "epoch": 6.374585724678684,
+      "grad_norm": 0.9771704077720642,
+      "learning_rate": 0.0002,
+      "loss": 0.4637,
+      "step": 39430
+    },
+    {
+      "epoch": 6.376202408859429,
+      "grad_norm": 1.285387635231018,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 39440
+    },
+    {
+      "epoch": 6.377819093040174,
+      "grad_norm": 1.019957184791565,
+      "learning_rate": 0.0002,
+      "loss": 0.4593,
+      "step": 39450
+    },
+    {
+      "epoch": 6.37943577722092,
+      "grad_norm": 1.2002915143966675,
+      "learning_rate": 0.0002,
+      "loss": 0.473,
+      "step": 39460
+    },
+    {
+      "epoch": 6.381052461401665,
+      "grad_norm": 1.3285092115402222,
+      "learning_rate": 0.0002,
+      "loss": 0.5025,
+      "step": 39470
+    },
+    {
+      "epoch": 6.38266914558241,
+      "grad_norm": 1.097846269607544,
+      "learning_rate": 0.0002,
+      "loss": 0.4626,
+      "step": 39480
+    },
+    {
+      "epoch": 6.384285829763156,
+      "grad_norm": 0.9537988305091858,
+      "learning_rate": 0.0002,
+      "loss": 0.5109,
+      "step": 39490
+    },
+    {
+      "epoch": 6.385902513943901,
+      "grad_norm": 1.0350042581558228,
+      "learning_rate": 0.0002,
+      "loss": 0.4492,
+      "step": 39500
+    },
+    {
+      "epoch": 6.387519198124647,
+      "grad_norm": 0.9559133052825928,
+      "learning_rate": 0.0002,
+      "loss": 0.4824,
+      "step": 39510
+    },
+    {
+      "epoch": 6.389135882305392,
+      "grad_norm": 0.9615123271942139,
+      "learning_rate": 0.0002,
+      "loss": 0.5189,
+      "step": 39520
+    },
+    {
+      "epoch": 6.390752566486137,
+      "grad_norm": 1.0604504346847534,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 39530
+    },
+    {
+      "epoch": 6.392369250666882,
+      "grad_norm": 1.2460750341415405,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 39540
+    },
+    {
+      "epoch": 6.393985934847628,
+      "grad_norm": 1.1496477127075195,
+      "learning_rate": 0.0002,
+      "loss": 0.4929,
+      "step": 39550
+    },
+    {
+      "epoch": 6.395602619028373,
+      "grad_norm": 1.048043966293335,
+      "learning_rate": 0.0002,
+      "loss": 0.4872,
+      "step": 39560
+    },
+    {
+      "epoch": 6.397219303209118,
+      "grad_norm": 1.333539366722107,
+      "learning_rate": 0.0002,
+      "loss": 0.5231,
+      "step": 39570
+    },
+    {
+      "epoch": 6.398835987389863,
+      "grad_norm": 1.0605626106262207,
+      "learning_rate": 0.0002,
+      "loss": 0.4877,
+      "step": 39580
+    },
+    {
+      "epoch": 6.4004526715706085,
+      "grad_norm": 1.163220763206482,
+      "learning_rate": 0.0002,
+      "loss": 0.4643,
+      "step": 39590
+    },
+    {
+      "epoch": 6.402069355751354,
+      "grad_norm": 1.1878494024276733,
+      "learning_rate": 0.0002,
+      "loss": 0.4824,
+      "step": 39600
+    },
+    {
+      "epoch": 6.403686039932099,
+      "grad_norm": 1.4630796909332275,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 39610
+    },
+    {
+      "epoch": 6.405302724112844,
+      "grad_norm": 1.073255181312561,
+      "learning_rate": 0.0002,
+      "loss": 0.4985,
+      "step": 39620
+    },
+    {
+      "epoch": 6.406919408293589,
+      "grad_norm": 1.0538873672485352,
+      "learning_rate": 0.0002,
+      "loss": 0.5108,
+      "step": 39630
+    },
+    {
+      "epoch": 6.4085360924743355,
+      "grad_norm": 1.015525221824646,
+      "learning_rate": 0.0002,
+      "loss": 0.4801,
+      "step": 39640
+    },
+    {
+      "epoch": 6.410152776655081,
+      "grad_norm": 1.1454379558563232,
+      "learning_rate": 0.0002,
+      "loss": 0.4781,
+      "step": 39650
+    },
+    {
+      "epoch": 6.411769460835826,
+      "grad_norm": 1.2801800966262817,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 39660
+    },
+    {
+      "epoch": 6.413386145016571,
+      "grad_norm": 1.077579140663147,
+      "learning_rate": 0.0002,
+      "loss": 0.4804,
+      "step": 39670
+    },
+    {
+      "epoch": 6.4150028291973165,
+      "grad_norm": 1.376662015914917,
+      "learning_rate": 0.0002,
+      "loss": 0.51,
+      "step": 39680
+    },
+    {
+      "epoch": 6.416619513378062,
+      "grad_norm": 1.2064344882965088,
+      "learning_rate": 0.0002,
+      "loss": 0.4956,
+      "step": 39690
+    },
+    {
+      "epoch": 6.418236197558807,
+      "grad_norm": 1.0689115524291992,
+      "learning_rate": 0.0002,
+      "loss": 0.4762,
+      "step": 39700
+    },
+    {
+      "epoch": 6.419852881739552,
+      "grad_norm": 0.9997019171714783,
+      "learning_rate": 0.0002,
+      "loss": 0.4762,
+      "step": 39710
+    },
+    {
+      "epoch": 6.421469565920297,
+      "grad_norm": 1.2368080615997314,
+      "learning_rate": 0.0002,
+      "loss": 0.49,
+      "step": 39720
+    },
+    {
+      "epoch": 6.423086250101043,
+      "grad_norm": 1.2085820436477661,
+      "learning_rate": 0.0002,
+      "loss": 0.4774,
+      "step": 39730
+    },
+    {
+      "epoch": 6.424702934281788,
+      "grad_norm": 1.057246208190918,
+      "learning_rate": 0.0002,
+      "loss": 0.4671,
+      "step": 39740
+    },
+    {
+      "epoch": 6.426319618462533,
+      "grad_norm": 1.1311043500900269,
+      "learning_rate": 0.0002,
+      "loss": 0.5315,
+      "step": 39750
+    },
+    {
+      "epoch": 6.427936302643278,
+      "grad_norm": 1.2352231740951538,
+      "learning_rate": 0.0002,
+      "loss": 0.5171,
+      "step": 39760
+    },
+    {
+      "epoch": 6.4295529868240235,
+      "grad_norm": 0.953233540058136,
+      "learning_rate": 0.0002,
+      "loss": 0.466,
+      "step": 39770
+    },
+    {
+      "epoch": 6.431169671004769,
+      "grad_norm": 1.0632505416870117,
+      "learning_rate": 0.0002,
+      "loss": 0.4834,
+      "step": 39780
+    },
+    {
+      "epoch": 6.432786355185515,
+      "grad_norm": 1.0916751623153687,
+      "learning_rate": 0.0002,
+      "loss": 0.5053,
+      "step": 39790
+    },
+    {
+      "epoch": 6.43440303936626,
+      "grad_norm": 0.9732703566551208,
+      "learning_rate": 0.0002,
+      "loss": 0.4788,
+      "step": 39800
+    },
+    {
+      "epoch": 6.436019723547005,
+      "grad_norm": 1.1673705577850342,
+      "learning_rate": 0.0002,
+      "loss": 0.4982,
+      "step": 39810
+    },
+    {
+      "epoch": 6.437636407727751,
+      "grad_norm": 1.1049559116363525,
+      "learning_rate": 0.0002,
+      "loss": 0.4484,
+      "step": 39820
+    },
+    {
+      "epoch": 6.439253091908496,
+      "grad_norm": 1.345277190208435,
+      "learning_rate": 0.0002,
+      "loss": 0.4784,
+      "step": 39830
+    },
+    {
+      "epoch": 6.440869776089241,
+      "grad_norm": 1.1118950843811035,
+      "learning_rate": 0.0002,
+      "loss": 0.4716,
+      "step": 39840
+    },
+    {
+      "epoch": 6.442486460269986,
+      "grad_norm": 1.4872850179672241,
+      "learning_rate": 0.0002,
+      "loss": 0.5133,
+      "step": 39850
+    },
+    {
+      "epoch": 6.4441031444507315,
+      "grad_norm": 1.0763497352600098,
+      "learning_rate": 0.0002,
+      "loss": 0.4532,
+      "step": 39860
+    },
+    {
+      "epoch": 6.445719828631477,
+      "grad_norm": 0.9245555400848389,
+      "learning_rate": 0.0002,
+      "loss": 0.4572,
+      "step": 39870
+    },
+    {
+      "epoch": 6.447336512812222,
+      "grad_norm": 1.4154807329177856,
+      "learning_rate": 0.0002,
+      "loss": 0.4917,
+      "step": 39880
+    },
+    {
+      "epoch": 6.448953196992967,
+      "grad_norm": 1.0885124206542969,
+      "learning_rate": 0.0002,
+      "loss": 0.4852,
+      "step": 39890
+    },
+    {
+      "epoch": 6.450569881173712,
+      "grad_norm": 1.3989344835281372,
+      "learning_rate": 0.0002,
+      "loss": 0.5399,
+      "step": 39900
+    },
+    {
+      "epoch": 6.452186565354458,
+      "grad_norm": 0.9763124585151672,
+      "learning_rate": 0.0002,
+      "loss": 0.509,
+      "step": 39910
+    },
+    {
+      "epoch": 6.453803249535203,
+      "grad_norm": 1.135272741317749,
+      "learning_rate": 0.0002,
+      "loss": 0.5134,
+      "step": 39920
+    },
+    {
+      "epoch": 6.455419933715948,
+      "grad_norm": 1.1140081882476807,
+      "learning_rate": 0.0002,
+      "loss": 0.4941,
+      "step": 39930
+    },
+    {
+      "epoch": 6.457036617896694,
+      "grad_norm": 1.0992448329925537,
+      "learning_rate": 0.0002,
+      "loss": 0.5137,
+      "step": 39940
+    },
+    {
+      "epoch": 6.4586533020774395,
+      "grad_norm": 1.1658501625061035,
+      "learning_rate": 0.0002,
+      "loss": 0.4914,
+      "step": 39950
+    },
+    {
+      "epoch": 6.460269986258185,
+      "grad_norm": 1.1122797727584839,
+      "learning_rate": 0.0002,
+      "loss": 0.5036,
+      "step": 39960
+    },
+    {
+      "epoch": 6.46188667043893,
+      "grad_norm": 0.9664968252182007,
+      "learning_rate": 0.0002,
+      "loss": 0.5159,
+      "step": 39970
+    },
+    {
+      "epoch": 6.463503354619675,
+      "grad_norm": 1.2513965368270874,
+      "learning_rate": 0.0002,
+      "loss": 0.4989,
+      "step": 39980
+    },
+    {
+      "epoch": 6.46512003880042,
+      "grad_norm": 1.1198630332946777,
+      "learning_rate": 0.0002,
+      "loss": 0.4694,
+      "step": 39990
+    },
+    {
+      "epoch": 6.466736722981166,
+      "grad_norm": 0.8783249855041504,
+      "learning_rate": 0.0002,
+      "loss": 0.5023,
+      "step": 40000
+    },
+    {
+      "epoch": 6.468353407161911,
+      "grad_norm": 1.1313109397888184,
+      "learning_rate": 0.0002,
+      "loss": 0.4648,
+      "step": 40010
+    },
+    {
+      "epoch": 6.469970091342656,
+      "grad_norm": 1.0854487419128418,
+      "learning_rate": 0.0002,
+      "loss": 0.4965,
+      "step": 40020
+    },
+    {
+      "epoch": 6.471586775523401,
+      "grad_norm": 1.1738566160202026,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 40030
+    },
+    {
+      "epoch": 6.473203459704147,
+      "grad_norm": 0.9720084071159363,
+      "learning_rate": 0.0002,
+      "loss": 0.4947,
+      "step": 40040
+    },
+    {
+      "epoch": 6.474820143884892,
+      "grad_norm": 1.105618953704834,
+      "learning_rate": 0.0002,
+      "loss": 0.5218,
+      "step": 40050
+    },
+    {
+      "epoch": 6.476436828065637,
+      "grad_norm": 1.2007657289505005,
+      "learning_rate": 0.0002,
+      "loss": 0.4943,
+      "step": 40060
+    },
+    {
+      "epoch": 6.478053512246382,
+      "grad_norm": 1.088402509689331,
+      "learning_rate": 0.0002,
+      "loss": 0.4882,
+      "step": 40070
+    },
+    {
+      "epoch": 6.4796701964271275,
+      "grad_norm": 1.0775291919708252,
+      "learning_rate": 0.0002,
+      "loss": 0.504,
+      "step": 40080
+    },
+    {
+      "epoch": 6.481286880607874,
+      "grad_norm": 1.1018189191818237,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 40090
+    },
+    {
+      "epoch": 6.482903564788619,
+      "grad_norm": 1.1676557064056396,
+      "learning_rate": 0.0002,
+      "loss": 0.488,
+      "step": 40100
+    },
+    {
+      "epoch": 6.484520248969364,
+      "grad_norm": 0.9619805812835693,
+      "learning_rate": 0.0002,
+      "loss": 0.4818,
+      "step": 40110
+    },
+    {
+      "epoch": 6.486136933150109,
+      "grad_norm": 1.2408208847045898,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 40120
+    },
+    {
+      "epoch": 6.4877536173308545,
+      "grad_norm": 1.3488136529922485,
+      "learning_rate": 0.0002,
+      "loss": 0.4668,
+      "step": 40130
+    },
+    {
+      "epoch": 6.4893703015116,
+      "grad_norm": 0.9864488244056702,
+      "learning_rate": 0.0002,
+      "loss": 0.4774,
+      "step": 40140
+    },
+    {
+      "epoch": 6.490986985692345,
+      "grad_norm": 0.9437947273254395,
+      "learning_rate": 0.0002,
+      "loss": 0.4651,
+      "step": 40150
+    },
+    {
+      "epoch": 6.49260366987309,
+      "grad_norm": 1.2005455493927002,
+      "learning_rate": 0.0002,
+      "loss": 0.542,
+      "step": 40160
+    },
+    {
+      "epoch": 6.4942203540538355,
+      "grad_norm": 1.0796732902526855,
+      "learning_rate": 0.0002,
+      "loss": 0.4704,
+      "step": 40170
+    },
+    {
+      "epoch": 6.495837038234581,
+      "grad_norm": 1.1347825527191162,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 40180
+    },
+    {
+      "epoch": 6.497453722415326,
+      "grad_norm": 1.2311455011367798,
+      "learning_rate": 0.0002,
+      "loss": 0.5215,
+      "step": 40190
+    },
+    {
+      "epoch": 6.499070406596071,
+      "grad_norm": 1.068609356880188,
+      "learning_rate": 0.0002,
+      "loss": 0.5043,
+      "step": 40200
+    },
+    {
+      "epoch": 6.500687090776816,
+      "grad_norm": 1.196425437927246,
+      "learning_rate": 0.0002,
+      "loss": 0.4868,
+      "step": 40210
+    },
+    {
+      "epoch": 6.5023037749575625,
+      "grad_norm": 1.183927297592163,
+      "learning_rate": 0.0002,
+      "loss": 0.4881,
+      "step": 40220
+    },
+    {
+      "epoch": 6.503920459138307,
+      "grad_norm": 0.9099724292755127,
+      "learning_rate": 0.0002,
+      "loss": 0.4958,
+      "step": 40230
+    },
+    {
+      "epoch": 6.505537143319053,
+      "grad_norm": 0.9261038899421692,
+      "learning_rate": 0.0002,
+      "loss": 0.4816,
+      "step": 40240
+    },
+    {
+      "epoch": 6.507153827499798,
+      "grad_norm": 1.185491681098938,
+      "learning_rate": 0.0002,
+      "loss": 0.5151,
+      "step": 40250
+    },
+    {
+      "epoch": 6.508770511680543,
+      "grad_norm": 1.1866052150726318,
+      "learning_rate": 0.0002,
+      "loss": 0.4853,
+      "step": 40260
+    },
+    {
+      "epoch": 6.510387195861289,
+      "grad_norm": 1.1600912809371948,
+      "learning_rate": 0.0002,
+      "loss": 0.491,
+      "step": 40270
+    },
+    {
+      "epoch": 6.512003880042034,
+      "grad_norm": 0.9609426259994507,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 40280
+    },
+    {
+      "epoch": 6.513620564222779,
+      "grad_norm": 1.078864336013794,
+      "learning_rate": 0.0002,
+      "loss": 0.4794,
+      "step": 40290
+    },
+    {
+      "epoch": 6.515237248403524,
+      "grad_norm": 1.042761206626892,
+      "learning_rate": 0.0002,
+      "loss": 0.46,
+      "step": 40300
+    },
+    {
+      "epoch": 6.51685393258427,
+      "grad_norm": 0.9742481112480164,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 40310
+    },
+    {
+      "epoch": 6.518470616765015,
+      "grad_norm": 1.2544835805892944,
+      "learning_rate": 0.0002,
+      "loss": 0.5234,
+      "step": 40320
+    },
+    {
+      "epoch": 6.52008730094576,
+      "grad_norm": 1.3019760847091675,
+      "learning_rate": 0.0002,
+      "loss": 0.4815,
+      "step": 40330
+    },
+    {
+      "epoch": 6.521703985126505,
+      "grad_norm": 1.3196964263916016,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 40340
+    },
+    {
+      "epoch": 6.5233206693072505,
+      "grad_norm": 1.2795668840408325,
+      "learning_rate": 0.0002,
+      "loss": 0.4979,
+      "step": 40350
+    },
+    {
+      "epoch": 6.524937353487996,
+      "grad_norm": 1.1618940830230713,
+      "learning_rate": 0.0002,
+      "loss": 0.5075,
+      "step": 40360
+    },
+    {
+      "epoch": 6.526554037668742,
+      "grad_norm": 1.330543041229248,
+      "learning_rate": 0.0002,
+      "loss": 0.5081,
+      "step": 40370
+    },
+    {
+      "epoch": 6.528170721849486,
+      "grad_norm": 1.1946901082992554,
+      "learning_rate": 0.0002,
+      "loss": 0.5055,
+      "step": 40380
+    },
+    {
+      "epoch": 6.529787406030232,
+      "grad_norm": 1.1708201169967651,
+      "learning_rate": 0.0002,
+      "loss": 0.4518,
+      "step": 40390
+    },
+    {
+      "epoch": 6.531404090210978,
+      "grad_norm": 0.894036591053009,
+      "learning_rate": 0.0002,
+      "loss": 0.4556,
+      "step": 40400
+    },
+    {
+      "epoch": 6.533020774391723,
+      "grad_norm": 1.1199041604995728,
+      "learning_rate": 0.0002,
+      "loss": 0.4919,
+      "step": 40410
+    },
+    {
+      "epoch": 6.534637458572468,
+      "grad_norm": 1.180317759513855,
+      "learning_rate": 0.0002,
+      "loss": 0.471,
+      "step": 40420
+    },
+    {
+      "epoch": 6.536254142753213,
+      "grad_norm": 1.37367582321167,
+      "learning_rate": 0.0002,
+      "loss": 0.4914,
+      "step": 40430
+    },
+    {
+      "epoch": 6.5378708269339585,
+      "grad_norm": 1.134791612625122,
+      "learning_rate": 0.0002,
+      "loss": 0.4561,
+      "step": 40440
+    },
+    {
+      "epoch": 6.539487511114704,
+      "grad_norm": 1.1160204410552979,
+      "learning_rate": 0.0002,
+      "loss": 0.5337,
+      "step": 40450
+    },
+    {
+      "epoch": 6.541104195295449,
+      "grad_norm": 1.268347978591919,
+      "learning_rate": 0.0002,
+      "loss": 0.5299,
+      "step": 40460
+    },
+    {
+      "epoch": 6.542720879476194,
+      "grad_norm": 1.1424330472946167,
+      "learning_rate": 0.0002,
+      "loss": 0.5167,
+      "step": 40470
+    },
+    {
+      "epoch": 6.544337563656939,
+      "grad_norm": 1.3098465204238892,
+      "learning_rate": 0.0002,
+      "loss": 0.5114,
+      "step": 40480
+    },
+    {
+      "epoch": 6.545954247837685,
+      "grad_norm": 1.3439544439315796,
+      "learning_rate": 0.0002,
+      "loss": 0.4865,
+      "step": 40490
+    },
+    {
+      "epoch": 6.54757093201843,
+      "grad_norm": 1.2708452939987183,
+      "learning_rate": 0.0002,
+      "loss": 0.5183,
+      "step": 40500
+    },
+    {
+      "epoch": 6.549187616199175,
+      "grad_norm": 1.483680248260498,
+      "learning_rate": 0.0002,
+      "loss": 0.5099,
+      "step": 40510
+    },
+    {
+      "epoch": 6.550804300379921,
+      "grad_norm": 1.1697806119918823,
+      "learning_rate": 0.0002,
+      "loss": 0.4811,
+      "step": 40520
+    },
+    {
+      "epoch": 6.5524209845606665,
+      "grad_norm": 1.1665642261505127,
+      "learning_rate": 0.0002,
+      "loss": 0.4814,
+      "step": 40530
+    },
+    {
+      "epoch": 6.554037668741412,
+      "grad_norm": 1.1243325471878052,
+      "learning_rate": 0.0002,
+      "loss": 0.4985,
+      "step": 40540
+    },
+    {
+      "epoch": 6.555654352922157,
+      "grad_norm": 1.0277988910675049,
+      "learning_rate": 0.0002,
+      "loss": 0.4936,
+      "step": 40550
+    },
+    {
+      "epoch": 6.557271037102902,
+      "grad_norm": 1.1466810703277588,
+      "learning_rate": 0.0002,
+      "loss": 0.487,
+      "step": 40560
+    },
+    {
+      "epoch": 6.558887721283647,
+      "grad_norm": 1.1415363550186157,
+      "learning_rate": 0.0002,
+      "loss": 0.4851,
+      "step": 40570
+    },
+    {
+      "epoch": 6.560504405464393,
+      "grad_norm": 1.1923491954803467,
+      "learning_rate": 0.0002,
+      "loss": 0.4631,
+      "step": 40580
+    },
+    {
+      "epoch": 6.562121089645138,
+      "grad_norm": 0.9264549612998962,
+      "learning_rate": 0.0002,
+      "loss": 0.5071,
+      "step": 40590
+    },
+    {
+      "epoch": 6.563737773825883,
+      "grad_norm": 0.8810341954231262,
+      "learning_rate": 0.0002,
+      "loss": 0.466,
+      "step": 40600
+    },
+    {
+      "epoch": 6.565354458006628,
+      "grad_norm": 2.3296701908111572,
+      "learning_rate": 0.0002,
+      "loss": 0.5085,
+      "step": 40610
+    },
+    {
+      "epoch": 6.5669711421873735,
+      "grad_norm": 1.0865163803100586,
+      "learning_rate": 0.0002,
+      "loss": 0.5196,
+      "step": 40620
+    },
+    {
+      "epoch": 6.568587826368119,
+      "grad_norm": 0.9844607710838318,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 40630
+    },
+    {
+      "epoch": 6.570204510548864,
+      "grad_norm": 1.1686855554580688,
+      "learning_rate": 0.0002,
+      "loss": 0.5437,
+      "step": 40640
+    },
+    {
+      "epoch": 6.571821194729609,
+      "grad_norm": 1.016829252243042,
+      "learning_rate": 0.0002,
+      "loss": 0.5293,
+      "step": 40650
+    },
+    {
+      "epoch": 6.5734378789103545,
+      "grad_norm": 1.2789337635040283,
+      "learning_rate": 0.0002,
+      "loss": 0.5243,
+      "step": 40660
+    },
+    {
+      "epoch": 6.575054563091101,
+      "grad_norm": 1.0819072723388672,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 40670
+    },
+    {
+      "epoch": 6.576671247271846,
+      "grad_norm": 1.1478345394134521,
+      "learning_rate": 0.0002,
+      "loss": 0.5024,
+      "step": 40680
+    },
+    {
+      "epoch": 6.578287931452591,
+      "grad_norm": 0.7972208857536316,
+      "learning_rate": 0.0002,
+      "loss": 0.5282,
+      "step": 40690
+    },
+    {
+      "epoch": 6.579904615633336,
+      "grad_norm": 1.1481789350509644,
+      "learning_rate": 0.0002,
+      "loss": 0.4877,
+      "step": 40700
+    },
+    {
+      "epoch": 6.5815212998140815,
+      "grad_norm": 1.0921871662139893,
+      "learning_rate": 0.0002,
+      "loss": 0.5143,
+      "step": 40710
+    },
+    {
+      "epoch": 6.583137983994827,
+      "grad_norm": 1.0230315923690796,
+      "learning_rate": 0.0002,
+      "loss": 0.5441,
+      "step": 40720
+    },
+    {
+      "epoch": 6.584754668175572,
+      "grad_norm": 1.151049017906189,
+      "learning_rate": 0.0002,
+      "loss": 0.4734,
+      "step": 40730
+    },
+    {
+      "epoch": 6.586371352356317,
+      "grad_norm": 1.4016883373260498,
+      "learning_rate": 0.0002,
+      "loss": 0.4782,
+      "step": 40740
+    },
+    {
+      "epoch": 6.587988036537062,
+      "grad_norm": 1.2211825847625732,
+      "learning_rate": 0.0002,
+      "loss": 0.5195,
+      "step": 40750
+    },
+    {
+      "epoch": 6.589604720717808,
+      "grad_norm": 1.2803404331207275,
+      "learning_rate": 0.0002,
+      "loss": 0.4815,
+      "step": 40760
+    },
+    {
+      "epoch": 6.591221404898553,
+      "grad_norm": 1.1119942665100098,
+      "learning_rate": 0.0002,
+      "loss": 0.5329,
+      "step": 40770
+    },
+    {
+      "epoch": 6.592838089079298,
+      "grad_norm": 1.464650273323059,
+      "learning_rate": 0.0002,
+      "loss": 0.5135,
+      "step": 40780
+    },
+    {
+      "epoch": 6.594454773260043,
+      "grad_norm": 1.1751397848129272,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 40790
+    },
+    {
+      "epoch": 6.596071457440789,
+      "grad_norm": 1.0866316556930542,
+      "learning_rate": 0.0002,
+      "loss": 0.4772,
+      "step": 40800
+    },
+    {
+      "epoch": 6.597688141621534,
+      "grad_norm": 1.1733694076538086,
+      "learning_rate": 0.0002,
+      "loss": 0.5132,
+      "step": 40810
+    },
+    {
+      "epoch": 6.59930482580228,
+      "grad_norm": 1.184708833694458,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 40820
+    },
+    {
+      "epoch": 6.600921509983025,
+      "grad_norm": 1.406081199645996,
+      "learning_rate": 0.0002,
+      "loss": 0.4885,
+      "step": 40830
+    },
+    {
+      "epoch": 6.60253819416377,
+      "grad_norm": 0.9658212661743164,
+      "learning_rate": 0.0002,
+      "loss": 0.499,
+      "step": 40840
+    },
+    {
+      "epoch": 6.604154878344516,
+      "grad_norm": 1.1457678079605103,
+      "learning_rate": 0.0002,
+      "loss": 0.5113,
+      "step": 40850
+    },
+    {
+      "epoch": 6.605771562525261,
+      "grad_norm": 1.0487784147262573,
+      "learning_rate": 0.0002,
+      "loss": 0.4916,
+      "step": 40860
+    },
+    {
+      "epoch": 6.607388246706006,
+      "grad_norm": 0.9357177019119263,
+      "learning_rate": 0.0002,
+      "loss": 0.4682,
+      "step": 40870
+    },
+    {
+      "epoch": 6.609004930886751,
+      "grad_norm": 1.1479727029800415,
+      "learning_rate": 0.0002,
+      "loss": 0.4751,
+      "step": 40880
+    },
+    {
+      "epoch": 6.610621615067497,
+      "grad_norm": 1.3729329109191895,
+      "learning_rate": 0.0002,
+      "loss": 0.5493,
+      "step": 40890
+    },
+    {
+      "epoch": 6.612238299248242,
+      "grad_norm": 1.0085599422454834,
+      "learning_rate": 0.0002,
+      "loss": 0.4886,
+      "step": 40900
+    },
+    {
+      "epoch": 6.613854983428987,
+      "grad_norm": 1.2750911712646484,
+      "learning_rate": 0.0002,
+      "loss": 0.516,
+      "step": 40910
+    },
+    {
+      "epoch": 6.615471667609732,
+      "grad_norm": 1.1929547786712646,
+      "learning_rate": 0.0002,
+      "loss": 0.5342,
+      "step": 40920
+    },
+    {
+      "epoch": 6.6170883517904775,
+      "grad_norm": 1.0821375846862793,
+      "learning_rate": 0.0002,
+      "loss": 0.4919,
+      "step": 40930
+    },
+    {
+      "epoch": 6.618705035971223,
+      "grad_norm": 1.197347640991211,
+      "learning_rate": 0.0002,
+      "loss": 0.5057,
+      "step": 40940
+    },
+    {
+      "epoch": 6.620321720151968,
+      "grad_norm": 1.2074699401855469,
+      "learning_rate": 0.0002,
+      "loss": 0.492,
+      "step": 40950
+    },
+    {
+      "epoch": 6.621938404332713,
+      "grad_norm": 1.312009572982788,
+      "learning_rate": 0.0002,
+      "loss": 0.5089,
+      "step": 40960
+    },
+    {
+      "epoch": 6.623555088513459,
+      "grad_norm": 1.4381471872329712,
+      "learning_rate": 0.0002,
+      "loss": 0.5476,
+      "step": 40970
+    },
+    {
+      "epoch": 6.6251717726942045,
+      "grad_norm": 1.1574671268463135,
+      "learning_rate": 0.0002,
+      "loss": 0.4904,
+      "step": 40980
+    },
+    {
+      "epoch": 6.62678845687495,
+      "grad_norm": 0.885661780834198,
+      "learning_rate": 0.0002,
+      "loss": 0.531,
+      "step": 40990
+    },
+    {
+      "epoch": 6.628405141055695,
+      "grad_norm": 1.024571180343628,
+      "learning_rate": 0.0002,
+      "loss": 0.5145,
+      "step": 41000
+    },
+    {
+      "epoch": 6.63002182523644,
+      "grad_norm": 1.103437900543213,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 41010
+    },
+    {
+      "epoch": 6.6316385094171855,
+      "grad_norm": 1.122450828552246,
+      "learning_rate": 0.0002,
+      "loss": 0.4671,
+      "step": 41020
+    },
+    {
+      "epoch": 6.633255193597931,
+      "grad_norm": 1.2256295680999756,
+      "learning_rate": 0.0002,
+      "loss": 0.5134,
+      "step": 41030
+    },
+    {
+      "epoch": 6.634871877778676,
+      "grad_norm": 1.364594578742981,
+      "learning_rate": 0.0002,
+      "loss": 0.4908,
+      "step": 41040
+    },
+    {
+      "epoch": 6.636488561959421,
+      "grad_norm": 0.9550056457519531,
+      "learning_rate": 0.0002,
+      "loss": 0.4964,
+      "step": 41050
+    },
+    {
+      "epoch": 6.638105246140166,
+      "grad_norm": 1.3174707889556885,
+      "learning_rate": 0.0002,
+      "loss": 0.5028,
+      "step": 41060
+    },
+    {
+      "epoch": 6.639721930320912,
+      "grad_norm": 1.0835540294647217,
+      "learning_rate": 0.0002,
+      "loss": 0.4717,
+      "step": 41070
+    },
+    {
+      "epoch": 6.641338614501657,
+      "grad_norm": 1.1432770490646362,
+      "learning_rate": 0.0002,
+      "loss": 0.497,
+      "step": 41080
+    },
+    {
+      "epoch": 6.642955298682402,
+      "grad_norm": 1.2398556470870972,
+      "learning_rate": 0.0002,
+      "loss": 0.4903,
+      "step": 41090
+    },
+    {
+      "epoch": 6.644571982863147,
+      "grad_norm": 1.1147747039794922,
+      "learning_rate": 0.0002,
+      "loss": 0.4991,
+      "step": 41100
+    },
+    {
+      "epoch": 6.6461886670438926,
+      "grad_norm": 1.0730493068695068,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 41110
+    },
+    {
+      "epoch": 6.647805351224639,
+      "grad_norm": 1.3218451738357544,
+      "learning_rate": 0.0002,
+      "loss": 0.486,
+      "step": 41120
+    },
+    {
+      "epoch": 6.649422035405384,
+      "grad_norm": 1.3027331829071045,
+      "learning_rate": 0.0002,
+      "loss": 0.5276,
+      "step": 41130
+    },
+    {
+      "epoch": 6.651038719586129,
+      "grad_norm": 1.0280735492706299,
+      "learning_rate": 0.0002,
+      "loss": 0.5263,
+      "step": 41140
+    },
+    {
+      "epoch": 6.652655403766874,
+      "grad_norm": 1.109916090965271,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 41150
+    },
+    {
+      "epoch": 6.65427208794762,
+      "grad_norm": 1.078734040260315,
+      "learning_rate": 0.0002,
+      "loss": 0.5001,
+      "step": 41160
+    },
+    {
+      "epoch": 6.655888772128365,
+      "grad_norm": 1.1595654487609863,
+      "learning_rate": 0.0002,
+      "loss": 0.484,
+      "step": 41170
+    },
+    {
+      "epoch": 6.65750545630911,
+      "grad_norm": 1.1701031923294067,
+      "learning_rate": 0.0002,
+      "loss": 0.5101,
+      "step": 41180
+    },
+    {
+      "epoch": 6.659122140489855,
+      "grad_norm": 1.0424643754959106,
+      "learning_rate": 0.0002,
+      "loss": 0.5341,
+      "step": 41190
+    },
+    {
+      "epoch": 6.6607388246706005,
+      "grad_norm": 1.22880220413208,
+      "learning_rate": 0.0002,
+      "loss": 0.4863,
+      "step": 41200
+    },
+    {
+      "epoch": 6.662355508851346,
+      "grad_norm": 1.1907655000686646,
+      "learning_rate": 0.0002,
+      "loss": 0.4987,
+      "step": 41210
+    },
+    {
+      "epoch": 6.663972193032091,
+      "grad_norm": 1.0765007734298706,
+      "learning_rate": 0.0002,
+      "loss": 0.5343,
+      "step": 41220
+    },
+    {
+      "epoch": 6.665588877212836,
+      "grad_norm": 0.9994917511940002,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 41230
+    },
+    {
+      "epoch": 6.6672055613935814,
+      "grad_norm": 0.968578040599823,
+      "learning_rate": 0.0002,
+      "loss": 0.507,
+      "step": 41240
+    },
+    {
+      "epoch": 6.668822245574327,
+      "grad_norm": 1.0576032400131226,
+      "learning_rate": 0.0002,
+      "loss": 0.5068,
+      "step": 41250
+    },
+    {
+      "epoch": 6.670438929755072,
+      "grad_norm": 1.2183765172958374,
+      "learning_rate": 0.0002,
+      "loss": 0.486,
+      "step": 41260
+    },
+    {
+      "epoch": 6.672055613935818,
+      "grad_norm": 1.2548623085021973,
+      "learning_rate": 0.0002,
+      "loss": 0.4764,
+      "step": 41270
+    },
+    {
+      "epoch": 6.673672298116563,
+      "grad_norm": 1.0848388671875,
+      "learning_rate": 0.0002,
+      "loss": 0.5014,
+      "step": 41280
+    },
+    {
+      "epoch": 6.6752889822973085,
+      "grad_norm": 1.21421217918396,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 41290
+    },
+    {
+      "epoch": 6.676905666478054,
+      "grad_norm": 1.1453598737716675,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 41300
+    },
+    {
+      "epoch": 6.678522350658799,
+      "grad_norm": 1.2682722806930542,
+      "learning_rate": 0.0002,
+      "loss": 0.5033,
+      "step": 41310
+    },
+    {
+      "epoch": 6.680139034839544,
+      "grad_norm": 1.1659725904464722,
+      "learning_rate": 0.0002,
+      "loss": 0.5313,
+      "step": 41320
+    },
+    {
+      "epoch": 6.681755719020289,
+      "grad_norm": 1.36194908618927,
+      "learning_rate": 0.0002,
+      "loss": 0.5505,
+      "step": 41330
+    },
+    {
+      "epoch": 6.683372403201035,
+      "grad_norm": 1.1712592840194702,
+      "learning_rate": 0.0002,
+      "loss": 0.5127,
+      "step": 41340
+    },
+    {
+      "epoch": 6.68498908738178,
+      "grad_norm": 1.4168336391448975,
+      "learning_rate": 0.0002,
+      "loss": 0.5082,
+      "step": 41350
+    },
+    {
+      "epoch": 6.686605771562525,
+      "grad_norm": 1.0395328998565674,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 41360
+    },
+    {
+      "epoch": 6.68822245574327,
+      "grad_norm": 1.2511054277420044,
+      "learning_rate": 0.0002,
+      "loss": 0.5404,
+      "step": 41370
+    },
+    {
+      "epoch": 6.689839139924016,
+      "grad_norm": 1.0438542366027832,
+      "learning_rate": 0.0002,
+      "loss": 0.5027,
+      "step": 41380
+    },
+    {
+      "epoch": 6.691455824104761,
+      "grad_norm": 1.08684241771698,
+      "learning_rate": 0.0002,
+      "loss": 0.5069,
+      "step": 41390
+    },
+    {
+      "epoch": 6.693072508285506,
+      "grad_norm": 1.250788927078247,
+      "learning_rate": 0.0002,
+      "loss": 0.5224,
+      "step": 41400
+    },
+    {
+      "epoch": 6.694689192466251,
+      "grad_norm": 1.313890814781189,
+      "learning_rate": 0.0002,
+      "loss": 0.4921,
+      "step": 41410
+    },
+    {
+      "epoch": 6.696305876646997,
+      "grad_norm": 1.3218982219696045,
+      "learning_rate": 0.0002,
+      "loss": 0.5028,
+      "step": 41420
+    },
+    {
+      "epoch": 6.697922560827743,
+      "grad_norm": 1.0366582870483398,
+      "learning_rate": 0.0002,
+      "loss": 0.4851,
+      "step": 41430
+    },
+    {
+      "epoch": 6.699539245008488,
+      "grad_norm": 1.066121220588684,
+      "learning_rate": 0.0002,
+      "loss": 0.5103,
+      "step": 41440
+    },
+    {
+      "epoch": 6.701155929189233,
+      "grad_norm": 1.0239925384521484,
+      "learning_rate": 0.0002,
+      "loss": 0.4966,
+      "step": 41450
+    },
+    {
+      "epoch": 6.702772613369978,
+      "grad_norm": 0.9402176141738892,
+      "learning_rate": 0.0002,
+      "loss": 0.4767,
+      "step": 41460
+    },
+    {
+      "epoch": 6.7043892975507235,
+      "grad_norm": 1.391718864440918,
+      "learning_rate": 0.0002,
+      "loss": 0.5381,
+      "step": 41470
+    },
+    {
+      "epoch": 6.706005981731469,
+      "grad_norm": 1.215600609779358,
+      "learning_rate": 0.0002,
+      "loss": 0.512,
+      "step": 41480
+    },
+    {
+      "epoch": 6.707622665912214,
+      "grad_norm": 1.063722848892212,
+      "learning_rate": 0.0002,
+      "loss": 0.5219,
+      "step": 41490
+    },
+    {
+      "epoch": 6.709239350092959,
+      "grad_norm": 1.132149577140808,
+      "learning_rate": 0.0002,
+      "loss": 0.492,
+      "step": 41500
+    },
+    {
+      "epoch": 6.7108560342737045,
+      "grad_norm": 1.0302950143814087,
+      "learning_rate": 0.0002,
+      "loss": 0.4812,
+      "step": 41510
+    },
+    {
+      "epoch": 6.71247271845445,
+      "grad_norm": 1.5342752933502197,
+      "learning_rate": 0.0002,
+      "loss": 0.5141,
+      "step": 41520
+    },
+    {
+      "epoch": 6.714089402635195,
+      "grad_norm": 1.177137017250061,
+      "learning_rate": 0.0002,
+      "loss": 0.5123,
+      "step": 41530
+    },
+    {
+      "epoch": 6.71570608681594,
+      "grad_norm": 1.2335538864135742,
+      "learning_rate": 0.0002,
+      "loss": 0.5082,
+      "step": 41540
+    },
+    {
+      "epoch": 6.717322770996686,
+      "grad_norm": 1.140604853630066,
+      "learning_rate": 0.0002,
+      "loss": 0.4864,
+      "step": 41550
+    },
+    {
+      "epoch": 6.718939455177431,
+      "grad_norm": 1.3567465543746948,
+      "learning_rate": 0.0002,
+      "loss": 0.4888,
+      "step": 41560
+    },
+    {
+      "epoch": 6.720556139358177,
+      "grad_norm": 1.0693929195404053,
+      "learning_rate": 0.0002,
+      "loss": 0.5183,
+      "step": 41570
+    },
+    {
+      "epoch": 6.722172823538922,
+      "grad_norm": 1.1592605113983154,
+      "learning_rate": 0.0002,
+      "loss": 0.5131,
+      "step": 41580
+    },
+    {
+      "epoch": 6.723789507719667,
+      "grad_norm": 0.989006519317627,
+      "learning_rate": 0.0002,
+      "loss": 0.5476,
+      "step": 41590
+    },
+    {
+      "epoch": 6.7254061919004124,
+      "grad_norm": 1.04103422164917,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 41600
+    },
+    {
+      "epoch": 6.727022876081158,
+      "grad_norm": 1.1129004955291748,
+      "learning_rate": 0.0002,
+      "loss": 0.4823,
+      "step": 41610
+    },
+    {
+      "epoch": 6.728639560261903,
+      "grad_norm": 1.1473113298416138,
+      "learning_rate": 0.0002,
+      "loss": 0.5032,
+      "step": 41620
+    },
+    {
+      "epoch": 6.730256244442648,
+      "grad_norm": 1.348036527633667,
+      "learning_rate": 0.0002,
+      "loss": 0.5253,
+      "step": 41630
+    },
+    {
+      "epoch": 6.731872928623393,
+      "grad_norm": 1.259942650794983,
+      "learning_rate": 0.0002,
+      "loss": 0.4983,
+      "step": 41640
+    },
+    {
+      "epoch": 6.733489612804139,
+      "grad_norm": 1.0591514110565186,
+      "learning_rate": 0.0002,
+      "loss": 0.5182,
+      "step": 41650
+    },
+    {
+      "epoch": 6.735106296984884,
+      "grad_norm": 0.9737129211425781,
+      "learning_rate": 0.0002,
+      "loss": 0.4886,
+      "step": 41660
+    },
+    {
+      "epoch": 6.736722981165629,
+      "grad_norm": 1.2520451545715332,
+      "learning_rate": 0.0002,
+      "loss": 0.5051,
+      "step": 41670
+    },
+    {
+      "epoch": 6.738339665346374,
+      "grad_norm": 1.0555530786514282,
+      "learning_rate": 0.0002,
+      "loss": 0.5364,
+      "step": 41680
+    },
+    {
+      "epoch": 6.7399563495271195,
+      "grad_norm": 1.0025697946548462,
+      "learning_rate": 0.0002,
+      "loss": 0.4954,
+      "step": 41690
+    },
+    {
+      "epoch": 6.741573033707866,
+      "grad_norm": 1.1114100217819214,
+      "learning_rate": 0.0002,
+      "loss": 0.5485,
+      "step": 41700
+    },
+    {
+      "epoch": 6.74318971788861,
+      "grad_norm": 1.1537504196166992,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 41710
+    },
+    {
+      "epoch": 6.744806402069356,
+      "grad_norm": 1.037880539894104,
+      "learning_rate": 0.0002,
+      "loss": 0.5025,
+      "step": 41720
+    },
+    {
+      "epoch": 6.746423086250101,
+      "grad_norm": 1.0691965818405151,
+      "learning_rate": 0.0002,
+      "loss": 0.482,
+      "step": 41730
+    },
+    {
+      "epoch": 6.748039770430847,
+      "grad_norm": 1.376325011253357,
+      "learning_rate": 0.0002,
+      "loss": 0.5272,
+      "step": 41740
+    },
+    {
+      "epoch": 6.749656454611592,
+      "grad_norm": 1.4667129516601562,
+      "learning_rate": 0.0002,
+      "loss": 0.5484,
+      "step": 41750
+    },
+    {
+      "epoch": 6.751273138792337,
+      "grad_norm": 1.1517162322998047,
+      "learning_rate": 0.0002,
+      "loss": 0.5139,
+      "step": 41760
+    },
+    {
+      "epoch": 6.752889822973082,
+      "grad_norm": 1.1454511880874634,
+      "learning_rate": 0.0002,
+      "loss": 0.5523,
+      "step": 41770
+    },
+    {
+      "epoch": 6.7545065071538275,
+      "grad_norm": 1.6323128938674927,
+      "learning_rate": 0.0002,
+      "loss": 0.4664,
+      "step": 41780
+    },
+    {
+      "epoch": 6.756123191334573,
+      "grad_norm": 1.0951642990112305,
+      "learning_rate": 0.0002,
+      "loss": 0.5153,
+      "step": 41790
+    },
+    {
+      "epoch": 6.757739875515318,
+      "grad_norm": 1.0766983032226562,
+      "learning_rate": 0.0002,
+      "loss": 0.4998,
+      "step": 41800
+    },
+    {
+      "epoch": 6.759356559696063,
+      "grad_norm": 1.3472381830215454,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 41810
+    },
+    {
+      "epoch": 6.760973243876808,
+      "grad_norm": 1.0248444080352783,
+      "learning_rate": 0.0002,
+      "loss": 0.5172,
+      "step": 41820
+    },
+    {
+      "epoch": 6.762589928057554,
+      "grad_norm": 1.1276055574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 41830
+    },
+    {
+      "epoch": 6.764206612238299,
+      "grad_norm": 1.5398495197296143,
+      "learning_rate": 0.0002,
+      "loss": 0.5044,
+      "step": 41840
+    },
+    {
+      "epoch": 6.765823296419045,
+      "grad_norm": 1.1886497735977173,
+      "learning_rate": 0.0002,
+      "loss": 0.5097,
+      "step": 41850
+    },
+    {
+      "epoch": 6.767439980599789,
+      "grad_norm": 1.027198076248169,
+      "learning_rate": 0.0002,
+      "loss": 0.499,
+      "step": 41860
+    },
+    {
+      "epoch": 6.7690566647805355,
+      "grad_norm": 1.4644980430603027,
+      "learning_rate": 0.0002,
+      "loss": 0.5444,
+      "step": 41870
+    },
+    {
+      "epoch": 6.770673348961281,
+      "grad_norm": 0.9633586406707764,
+      "learning_rate": 0.0002,
+      "loss": 0.5009,
+      "step": 41880
+    },
+    {
+      "epoch": 6.772290033142026,
+      "grad_norm": 1.0895354747772217,
+      "learning_rate": 0.0002,
+      "loss": 0.484,
+      "step": 41890
+    },
+    {
+      "epoch": 6.773906717322771,
+      "grad_norm": 1.1887167692184448,
+      "learning_rate": 0.0002,
+      "loss": 0.5172,
+      "step": 41900
+    },
+    {
+      "epoch": 6.775523401503516,
+      "grad_norm": 1.3699820041656494,
+      "learning_rate": 0.0002,
+      "loss": 0.5399,
+      "step": 41910
+    },
+    {
+      "epoch": 6.777140085684262,
+      "grad_norm": 1.0266352891921997,
+      "learning_rate": 0.0002,
+      "loss": 0.5504,
+      "step": 41920
+    },
+    {
+      "epoch": 6.778756769865007,
+      "grad_norm": 1.0919075012207031,
+      "learning_rate": 0.0002,
+      "loss": 0.5105,
+      "step": 41930
+    },
+    {
+      "epoch": 6.780373454045752,
+      "grad_norm": 0.9839563369750977,
+      "learning_rate": 0.0002,
+      "loss": 0.4842,
+      "step": 41940
+    },
+    {
+      "epoch": 6.781990138226497,
+      "grad_norm": 1.2605451345443726,
+      "learning_rate": 0.0002,
+      "loss": 0.5081,
+      "step": 41950
+    },
+    {
+      "epoch": 6.7836068224072426,
+      "grad_norm": 0.9268672466278076,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 41960
+    },
+    {
+      "epoch": 6.785223506587988,
+      "grad_norm": 1.2002313137054443,
+      "learning_rate": 0.0002,
+      "loss": 0.4916,
+      "step": 41970
+    },
+    {
+      "epoch": 6.786840190768733,
+      "grad_norm": 1.2018438577651978,
+      "learning_rate": 0.0002,
+      "loss": 0.5467,
+      "step": 41980
+    },
+    {
+      "epoch": 6.788456874949478,
+      "grad_norm": 1.17646062374115,
+      "learning_rate": 0.0002,
+      "loss": 0.5491,
+      "step": 41990
+    },
+    {
+      "epoch": 6.790073559130224,
+      "grad_norm": 1.1080009937286377,
+      "learning_rate": 0.0002,
+      "loss": 0.5354,
+      "step": 42000
+    },
+    {
+      "epoch": 6.791690243310969,
+      "grad_norm": 1.1606498956680298,
+      "learning_rate": 0.0002,
+      "loss": 0.5384,
+      "step": 42010
+    },
+    {
+      "epoch": 6.793306927491715,
+      "grad_norm": 1.2484819889068604,
+      "learning_rate": 0.0002,
+      "loss": 0.4931,
+      "step": 42020
+    },
+    {
+      "epoch": 6.79492361167246,
+      "grad_norm": 1.1363215446472168,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 42030
+    },
+    {
+      "epoch": 6.796540295853205,
+      "grad_norm": 1.4469727277755737,
+      "learning_rate": 0.0002,
+      "loss": 0.5343,
+      "step": 42040
+    },
+    {
+      "epoch": 6.7981569800339505,
+      "grad_norm": 1.0617138147354126,
+      "learning_rate": 0.0002,
+      "loss": 0.5146,
+      "step": 42050
+    },
+    {
+      "epoch": 6.799773664214696,
+      "grad_norm": 1.1459330320358276,
+      "learning_rate": 0.0002,
+      "loss": 0.5188,
+      "step": 42060
+    },
+    {
+      "epoch": 6.801390348395441,
+      "grad_norm": 1.2095019817352295,
+      "learning_rate": 0.0002,
+      "loss": 0.5116,
+      "step": 42070
+    },
+    {
+      "epoch": 6.803007032576186,
+      "grad_norm": 1.3200831413269043,
+      "learning_rate": 0.0002,
+      "loss": 0.545,
+      "step": 42080
+    },
+    {
+      "epoch": 6.8046237167569315,
+      "grad_norm": 1.1633318662643433,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 42090
+    },
+    {
+      "epoch": 6.806240400937677,
+      "grad_norm": 0.8986614942550659,
+      "learning_rate": 0.0002,
+      "loss": 0.4938,
+      "step": 42100
+    },
+    {
+      "epoch": 6.807857085118422,
+      "grad_norm": 1.3705275058746338,
+      "learning_rate": 0.0002,
+      "loss": 0.559,
+      "step": 42110
+    },
+    {
+      "epoch": 6.809473769299167,
+      "grad_norm": 1.2418090105056763,
+      "learning_rate": 0.0002,
+      "loss": 0.5022,
+      "step": 42120
+    },
+    {
+      "epoch": 6.811090453479912,
+      "grad_norm": 1.0818954706192017,
+      "learning_rate": 0.0002,
+      "loss": 0.5014,
+      "step": 42130
+    },
+    {
+      "epoch": 6.812707137660658,
+      "grad_norm": 0.9293872117996216,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 42140
+    },
+    {
+      "epoch": 6.814323821841404,
+      "grad_norm": 0.9791894555091858,
+      "learning_rate": 0.0002,
+      "loss": 0.5009,
+      "step": 42150
+    },
+    {
+      "epoch": 6.815940506022149,
+      "grad_norm": 1.1956568956375122,
+      "learning_rate": 0.0002,
+      "loss": 0.5142,
+      "step": 42160
+    },
+    {
+      "epoch": 6.817557190202894,
+      "grad_norm": 0.9643568992614746,
+      "learning_rate": 0.0002,
+      "loss": 0.5126,
+      "step": 42170
+    },
+    {
+      "epoch": 6.819173874383639,
+      "grad_norm": 1.2499792575836182,
+      "learning_rate": 0.0002,
+      "loss": 0.5121,
+      "step": 42180
+    },
+    {
+      "epoch": 6.820790558564385,
+      "grad_norm": 1.1779413223266602,
+      "learning_rate": 0.0002,
+      "loss": 0.4942,
+      "step": 42190
+    },
+    {
+      "epoch": 6.82240724274513,
+      "grad_norm": 1.0570595264434814,
+      "learning_rate": 0.0002,
+      "loss": 0.498,
+      "step": 42200
+    },
+    {
+      "epoch": 6.824023926925875,
+      "grad_norm": 1.1393938064575195,
+      "learning_rate": 0.0002,
+      "loss": 0.4997,
+      "step": 42210
+    },
+    {
+      "epoch": 6.82564061110662,
+      "grad_norm": 1.152463436126709,
+      "learning_rate": 0.0002,
+      "loss": 0.4842,
+      "step": 42220
+    },
+    {
+      "epoch": 6.827257295287366,
+      "grad_norm": 1.3353025913238525,
+      "learning_rate": 0.0002,
+      "loss": 0.5234,
+      "step": 42230
+    },
+    {
+      "epoch": 6.828873979468111,
+      "grad_norm": 1.1719051599502563,
+      "learning_rate": 0.0002,
+      "loss": 0.539,
+      "step": 42240
+    },
+    {
+      "epoch": 6.830490663648856,
+      "grad_norm": 1.262141227722168,
+      "learning_rate": 0.0002,
+      "loss": 0.5139,
+      "step": 42250
+    },
+    {
+      "epoch": 6.832107347829601,
+      "grad_norm": 1.240899920463562,
+      "learning_rate": 0.0002,
+      "loss": 0.5021,
+      "step": 42260
+    },
+    {
+      "epoch": 6.8337240320103465,
+      "grad_norm": 1.0505269765853882,
+      "learning_rate": 0.0002,
+      "loss": 0.4961,
+      "step": 42270
+    },
+    {
+      "epoch": 6.835340716191092,
+      "grad_norm": 1.1556071043014526,
+      "learning_rate": 0.0002,
+      "loss": 0.4932,
+      "step": 42280
+    },
+    {
+      "epoch": 6.836957400371837,
+      "grad_norm": 1.1427719593048096,
+      "learning_rate": 0.0002,
+      "loss": 0.5461,
+      "step": 42290
+    },
+    {
+      "epoch": 6.838574084552583,
+      "grad_norm": 1.1540080308914185,
+      "learning_rate": 0.0002,
+      "loss": 0.5199,
+      "step": 42300
+    },
+    {
+      "epoch": 6.840190768733328,
+      "grad_norm": 1.0521200895309448,
+      "learning_rate": 0.0002,
+      "loss": 0.5269,
+      "step": 42310
+    },
+    {
+      "epoch": 6.8418074529140736,
+      "grad_norm": 1.0205531120300293,
+      "learning_rate": 0.0002,
+      "loss": 0.541,
+      "step": 42320
+    },
+    {
+      "epoch": 6.843424137094819,
+      "grad_norm": 1.0010193586349487,
+      "learning_rate": 0.0002,
+      "loss": 0.5225,
+      "step": 42330
+    },
+    {
+      "epoch": 6.845040821275564,
+      "grad_norm": 1.2138770818710327,
+      "learning_rate": 0.0002,
+      "loss": 0.5101,
+      "step": 42340
+    },
+    {
+      "epoch": 6.846657505456309,
+      "grad_norm": 1.3028651475906372,
+      "learning_rate": 0.0002,
+      "loss": 0.5452,
+      "step": 42350
+    },
+    {
+      "epoch": 6.8482741896370545,
+      "grad_norm": 1.0326353311538696,
+      "learning_rate": 0.0002,
+      "loss": 0.4894,
+      "step": 42360
+    },
+    {
+      "epoch": 6.8498908738178,
+      "grad_norm": 1.036085605621338,
+      "learning_rate": 0.0002,
+      "loss": 0.5285,
+      "step": 42370
+    },
+    {
+      "epoch": 6.851507557998545,
+      "grad_norm": 1.0575472116470337,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 42380
+    },
+    {
+      "epoch": 6.85312424217929,
+      "grad_norm": 1.1749629974365234,
+      "learning_rate": 0.0002,
+      "loss": 0.4997,
+      "step": 42390
+    },
+    {
+      "epoch": 6.854740926360035,
+      "grad_norm": 1.1747760772705078,
+      "learning_rate": 0.0002,
+      "loss": 0.4961,
+      "step": 42400
+    },
+    {
+      "epoch": 6.856357610540781,
+      "grad_norm": 1.1877071857452393,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 42410
+    },
+    {
+      "epoch": 6.857974294721526,
+      "grad_norm": 1.1209983825683594,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 42420
+    },
+    {
+      "epoch": 6.859590978902271,
+      "grad_norm": 1.2918205261230469,
+      "learning_rate": 0.0002,
+      "loss": 0.4939,
+      "step": 42430
+    },
+    {
+      "epoch": 6.861207663083016,
+      "grad_norm": 1.2443464994430542,
+      "learning_rate": 0.0002,
+      "loss": 0.5012,
+      "step": 42440
+    },
+    {
+      "epoch": 6.8628243472637624,
+      "grad_norm": 0.9336795210838318,
+      "learning_rate": 0.0002,
+      "loss": 0.5226,
+      "step": 42450
+    },
+    {
+      "epoch": 6.864441031444508,
+      "grad_norm": 1.2183542251586914,
+      "learning_rate": 0.0002,
+      "loss": 0.5108,
+      "step": 42460
+    },
+    {
+      "epoch": 6.866057715625253,
+      "grad_norm": 1.0071234703063965,
+      "learning_rate": 0.0002,
+      "loss": 0.5245,
+      "step": 42470
+    },
+    {
+      "epoch": 6.867674399805998,
+      "grad_norm": 1.2914012670516968,
+      "learning_rate": 0.0002,
+      "loss": 0.4753,
+      "step": 42480
+    },
+    {
+      "epoch": 6.869291083986743,
+      "grad_norm": 1.1050426959991455,
+      "learning_rate": 0.0002,
+      "loss": 0.4865,
+      "step": 42490
+    },
+    {
+      "epoch": 6.870907768167489,
+      "grad_norm": 1.1163811683654785,
+      "learning_rate": 0.0002,
+      "loss": 0.5243,
+      "step": 42500
+    },
+    {
+      "epoch": 6.872524452348234,
+      "grad_norm": 1.1575818061828613,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 42510
+    },
+    {
+      "epoch": 6.874141136528979,
+      "grad_norm": 1.11167311668396,
+      "learning_rate": 0.0002,
+      "loss": 0.5353,
+      "step": 42520
+    },
+    {
+      "epoch": 6.875757820709724,
+      "grad_norm": 1.0379102230072021,
+      "learning_rate": 0.0002,
+      "loss": 0.5141,
+      "step": 42530
+    },
+    {
+      "epoch": 6.8773745048904695,
+      "grad_norm": 1.2617160081863403,
+      "learning_rate": 0.0002,
+      "loss": 0.5355,
+      "step": 42540
+    },
+    {
+      "epoch": 6.878991189071215,
+      "grad_norm": 1.1749719381332397,
+      "learning_rate": 0.0002,
+      "loss": 0.4785,
+      "step": 42550
+    },
+    {
+      "epoch": 6.88060787325196,
+      "grad_norm": 1.2284821271896362,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 42560
+    },
+    {
+      "epoch": 6.882224557432705,
+      "grad_norm": 1.1917030811309814,
+      "learning_rate": 0.0002,
+      "loss": 0.5065,
+      "step": 42570
+    },
+    {
+      "epoch": 6.8838412416134505,
+      "grad_norm": 1.1943914890289307,
+      "learning_rate": 0.0002,
+      "loss": 0.5176,
+      "step": 42580
+    },
+    {
+      "epoch": 6.885457925794196,
+      "grad_norm": 1.2641394138336182,
+      "learning_rate": 0.0002,
+      "loss": 0.5072,
+      "step": 42590
+    },
+    {
+      "epoch": 6.887074609974942,
+      "grad_norm": 1.1280436515808105,
+      "learning_rate": 0.0002,
+      "loss": 0.5004,
+      "step": 42600
+    },
+    {
+      "epoch": 6.888691294155687,
+      "grad_norm": 0.9865449070930481,
+      "learning_rate": 0.0002,
+      "loss": 0.5328,
+      "step": 42610
+    },
+    {
+      "epoch": 6.890307978336432,
+      "grad_norm": 0.994987428188324,
+      "learning_rate": 0.0002,
+      "loss": 0.4953,
+      "step": 42620
+    },
+    {
+      "epoch": 6.8919246625171775,
+      "grad_norm": 0.9900388717651367,
+      "learning_rate": 0.0002,
+      "loss": 0.4805,
+      "step": 42630
+    },
+    {
+      "epoch": 6.893541346697923,
+      "grad_norm": 1.2992421388626099,
+      "learning_rate": 0.0002,
+      "loss": 0.5467,
+      "step": 42640
+    },
+    {
+      "epoch": 6.895158030878668,
+      "grad_norm": 1.0152487754821777,
+      "learning_rate": 0.0002,
+      "loss": 0.5017,
+      "step": 42650
+    },
+    {
+      "epoch": 6.896774715059413,
+      "grad_norm": 1.199453353881836,
+      "learning_rate": 0.0002,
+      "loss": 0.5043,
+      "step": 42660
+    },
+    {
+      "epoch": 6.898391399240158,
+      "grad_norm": 1.100630521774292,
+      "learning_rate": 0.0002,
+      "loss": 0.5106,
+      "step": 42670
+    },
+    {
+      "epoch": 6.900008083420904,
+      "grad_norm": 1.0489764213562012,
+      "learning_rate": 0.0002,
+      "loss": 0.503,
+      "step": 42680
+    },
+    {
+      "epoch": 6.901624767601649,
+      "grad_norm": 1.101407527923584,
+      "learning_rate": 0.0002,
+      "loss": 0.4634,
+      "step": 42690
+    },
+    {
+      "epoch": 6.903241451782394,
+      "grad_norm": 1.3130593299865723,
+      "learning_rate": 0.0002,
+      "loss": 0.5361,
+      "step": 42700
+    },
+    {
+      "epoch": 6.904858135963139,
+      "grad_norm": 0.9906072616577148,
+      "learning_rate": 0.0002,
+      "loss": 0.5119,
+      "step": 42710
+    },
+    {
+      "epoch": 6.906474820143885,
+      "grad_norm": 1.094502329826355,
+      "learning_rate": 0.0002,
+      "loss": 0.5146,
+      "step": 42720
+    },
+    {
+      "epoch": 6.90809150432463,
+      "grad_norm": 1.1025426387786865,
+      "learning_rate": 0.0002,
+      "loss": 0.5165,
+      "step": 42730
+    },
+    {
+      "epoch": 6.909708188505375,
+      "grad_norm": 1.0644042491912842,
+      "learning_rate": 0.0002,
+      "loss": 0.5463,
+      "step": 42740
+    },
+    {
+      "epoch": 6.911324872686121,
+      "grad_norm": 1.0709129571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.5024,
+      "step": 42750
+    },
+    {
+      "epoch": 6.912941556866866,
+      "grad_norm": 1.2445871829986572,
+      "learning_rate": 0.0002,
+      "loss": 0.5093,
+      "step": 42760
+    },
+    {
+      "epoch": 6.914558241047612,
+      "grad_norm": 1.020058035850525,
+      "learning_rate": 0.0002,
+      "loss": 0.5305,
+      "step": 42770
+    },
+    {
+      "epoch": 6.916174925228357,
+      "grad_norm": 0.9795091152191162,
+      "learning_rate": 0.0002,
+      "loss": 0.5382,
+      "step": 42780
+    },
+    {
+      "epoch": 6.917791609409102,
+      "grad_norm": 0.9369977116584778,
+      "learning_rate": 0.0002,
+      "loss": 0.5429,
+      "step": 42790
+    },
+    {
+      "epoch": 6.919408293589847,
+      "grad_norm": 1.0741904973983765,
+      "learning_rate": 0.0002,
+      "loss": 0.5444,
+      "step": 42800
+    },
+    {
+      "epoch": 6.921024977770593,
+      "grad_norm": 1.0702799558639526,
+      "learning_rate": 0.0002,
+      "loss": 0.5402,
+      "step": 42810
+    },
+    {
+      "epoch": 6.922641661951338,
+      "grad_norm": 1.0383983850479126,
+      "learning_rate": 0.0002,
+      "loss": 0.5291,
+      "step": 42820
+    },
+    {
+      "epoch": 6.924258346132083,
+      "grad_norm": 1.0761083364486694,
+      "learning_rate": 0.0002,
+      "loss": 0.5106,
+      "step": 42830
+    },
+    {
+      "epoch": 6.925875030312828,
+      "grad_norm": 1.2332350015640259,
+      "learning_rate": 0.0002,
+      "loss": 0.5726,
+      "step": 42840
+    },
+    {
+      "epoch": 6.9274917144935735,
+      "grad_norm": 1.3184348344802856,
+      "learning_rate": 0.0002,
+      "loss": 0.4996,
+      "step": 42850
+    },
+    {
+      "epoch": 6.929108398674319,
+      "grad_norm": 1.0586378574371338,
+      "learning_rate": 0.0002,
+      "loss": 0.5503,
+      "step": 42860
+    },
+    {
+      "epoch": 6.930725082855064,
+      "grad_norm": 1.2294201850891113,
+      "learning_rate": 0.0002,
+      "loss": 0.511,
+      "step": 42870
+    },
+    {
+      "epoch": 6.932341767035809,
+      "grad_norm": 1.3097991943359375,
+      "learning_rate": 0.0002,
+      "loss": 0.54,
+      "step": 42880
+    },
+    {
+      "epoch": 6.933958451216554,
+      "grad_norm": 0.9006873965263367,
+      "learning_rate": 0.0002,
+      "loss": 0.5228,
+      "step": 42890
+    },
+    {
+      "epoch": 6.9355751353973005,
+      "grad_norm": 1.265931248664856,
+      "learning_rate": 0.0002,
+      "loss": 0.4617,
+      "step": 42900
+    },
+    {
+      "epoch": 6.937191819578046,
+      "grad_norm": 1.1013522148132324,
+      "learning_rate": 0.0002,
+      "loss": 0.5029,
+      "step": 42910
+    },
+    {
+      "epoch": 6.938808503758791,
+      "grad_norm": 0.9910131692886353,
+      "learning_rate": 0.0002,
+      "loss": 0.5334,
+      "step": 42920
+    },
+    {
+      "epoch": 6.940425187939536,
+      "grad_norm": 1.102683424949646,
+      "learning_rate": 0.0002,
+      "loss": 0.5211,
+      "step": 42930
+    },
+    {
+      "epoch": 6.9420418721202815,
+      "grad_norm": 1.232961893081665,
+      "learning_rate": 0.0002,
+      "loss": 0.5588,
+      "step": 42940
+    },
+    {
+      "epoch": 6.943658556301027,
+      "grad_norm": 1.1714650392532349,
+      "learning_rate": 0.0002,
+      "loss": 0.5357,
+      "step": 42950
+    },
+    {
+      "epoch": 6.945275240481772,
+      "grad_norm": 1.1684318780899048,
+      "learning_rate": 0.0002,
+      "loss": 0.5232,
+      "step": 42960
+    },
+    {
+      "epoch": 6.946891924662517,
+      "grad_norm": 1.2074716091156006,
+      "learning_rate": 0.0002,
+      "loss": 0.5035,
+      "step": 42970
+    },
+    {
+      "epoch": 6.948508608843262,
+      "grad_norm": 1.2061275243759155,
+      "learning_rate": 0.0002,
+      "loss": 0.5111,
+      "step": 42980
+    },
+    {
+      "epoch": 6.950125293024008,
+      "grad_norm": 1.1216989755630493,
+      "learning_rate": 0.0002,
+      "loss": 0.5066,
+      "step": 42990
+    },
+    {
+      "epoch": 6.951741977204753,
+      "grad_norm": 1.304117202758789,
+      "learning_rate": 0.0002,
+      "loss": 0.4948,
+      "step": 43000
+    },
+    {
+      "epoch": 6.953358661385498,
+      "grad_norm": 1.2377972602844238,
+      "learning_rate": 0.0002,
+      "loss": 0.5684,
+      "step": 43010
+    },
+    {
+      "epoch": 6.954975345566243,
+      "grad_norm": 1.2332178354263306,
+      "learning_rate": 0.0002,
+      "loss": 0.4792,
+      "step": 43020
+    },
+    {
+      "epoch": 6.956592029746989,
+      "grad_norm": 1.1919599771499634,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 43030
+    },
+    {
+      "epoch": 6.958208713927734,
+      "grad_norm": 1.272700548171997,
+      "learning_rate": 0.0002,
+      "loss": 0.5352,
+      "step": 43040
+    },
+    {
+      "epoch": 6.95982539810848,
+      "grad_norm": 1.4377546310424805,
+      "learning_rate": 0.0002,
+      "loss": 0.5328,
+      "step": 43050
+    },
+    {
+      "epoch": 6.961442082289225,
+      "grad_norm": 1.2070353031158447,
+      "learning_rate": 0.0002,
+      "loss": 0.4894,
+      "step": 43060
+    },
+    {
+      "epoch": 6.96305876646997,
+      "grad_norm": 1.090205430984497,
+      "learning_rate": 0.0002,
+      "loss": 0.525,
+      "step": 43070
+    },
+    {
+      "epoch": 6.964675450650716,
+      "grad_norm": 1.1832911968231201,
+      "learning_rate": 0.0002,
+      "loss": 0.5255,
+      "step": 43080
+    },
+    {
+      "epoch": 6.966292134831461,
+      "grad_norm": 1.2921082973480225,
+      "learning_rate": 0.0002,
+      "loss": 0.5497,
+      "step": 43090
+    },
+    {
+      "epoch": 6.967908819012206,
+      "grad_norm": 1.4303096532821655,
+      "learning_rate": 0.0002,
+      "loss": 0.5527,
+      "step": 43100
+    },
+    {
+      "epoch": 6.969525503192951,
+      "grad_norm": 1.0788004398345947,
+      "learning_rate": 0.0002,
+      "loss": 0.4807,
+      "step": 43110
+    },
+    {
+      "epoch": 6.9711421873736965,
+      "grad_norm": 1.2192047834396362,
+      "learning_rate": 0.0002,
+      "loss": 0.5006,
+      "step": 43120
+    },
+    {
+      "epoch": 6.972758871554442,
+      "grad_norm": 1.0735143423080444,
+      "learning_rate": 0.0002,
+      "loss": 0.4714,
+      "step": 43130
+    },
+    {
+      "epoch": 6.974375555735187,
+      "grad_norm": 1.0317153930664062,
+      "learning_rate": 0.0002,
+      "loss": 0.5307,
+      "step": 43140
+    },
+    {
+      "epoch": 6.975992239915932,
+      "grad_norm": 1.0926798582077026,
+      "learning_rate": 0.0002,
+      "loss": 0.5154,
+      "step": 43150
+    },
+    {
+      "epoch": 6.977608924096677,
+      "grad_norm": 1.1660500764846802,
+      "learning_rate": 0.0002,
+      "loss": 0.4976,
+      "step": 43160
+    },
+    {
+      "epoch": 6.979225608277423,
+      "grad_norm": 1.3945232629776,
+      "learning_rate": 0.0002,
+      "loss": 0.5456,
+      "step": 43170
+    },
+    {
+      "epoch": 6.980842292458169,
+      "grad_norm": 1.2684587240219116,
+      "learning_rate": 0.0002,
+      "loss": 0.4979,
+      "step": 43180
+    },
+    {
+      "epoch": 6.982458976638913,
+      "grad_norm": 1.1574004888534546,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 43190
+    },
+    {
+      "epoch": 6.984075660819659,
+      "grad_norm": 1.2534198760986328,
+      "learning_rate": 0.0002,
+      "loss": 0.5629,
+      "step": 43200
+    },
+    {
+      "epoch": 6.9856923450004045,
+      "grad_norm": 1.135245442390442,
+      "learning_rate": 0.0002,
+      "loss": 0.5191,
+      "step": 43210
+    },
+    {
+      "epoch": 6.98730902918115,
+      "grad_norm": 1.3824104070663452,
+      "learning_rate": 0.0002,
+      "loss": 0.548,
+      "step": 43220
+    },
+    {
+      "epoch": 6.988925713361895,
+      "grad_norm": 1.2128452062606812,
+      "learning_rate": 0.0002,
+      "loss": 0.5294,
+      "step": 43230
+    },
+    {
+      "epoch": 6.99054239754264,
+      "grad_norm": 1.0795245170593262,
+      "learning_rate": 0.0002,
+      "loss": 0.505,
+      "step": 43240
+    },
+    {
+      "epoch": 6.992159081723385,
+      "grad_norm": 1.337353229522705,
+      "learning_rate": 0.0002,
+      "loss": 0.4889,
+      "step": 43250
+    },
+    {
+      "epoch": 6.993775765904131,
+      "grad_norm": 1.1731765270233154,
+      "learning_rate": 0.0002,
+      "loss": 0.4749,
+      "step": 43260
+    },
+    {
+      "epoch": 6.995392450084876,
+      "grad_norm": 1.0203192234039307,
+      "learning_rate": 0.0002,
+      "loss": 0.4897,
+      "step": 43270
+    },
+    {
+      "epoch": 6.997009134265621,
+      "grad_norm": 0.9261201620101929,
+      "learning_rate": 0.0002,
+      "loss": 0.5324,
+      "step": 43280
+    },
+    {
+      "epoch": 6.998625818446366,
+      "grad_norm": 1.107865810394287,
+      "learning_rate": 0.0002,
+      "loss": 0.5227,
+      "step": 43290
+    },
+    {
+      "epoch": 6.9999191657909625,
+      "eval_loss": 1.2679380178451538,
+      "eval_runtime": 122.202,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 43298
+    },
+    {
+      "epoch": 7.000242502627112,
+      "grad_norm": 0.9555306434631348,
+      "learning_rate": 0.0002,
+      "loss": 0.4651,
+      "step": 43300
+    },
+    {
+      "epoch": 7.001859186807857,
+      "grad_norm": 1.3280415534973145,
+      "learning_rate": 0.0002,
+      "loss": 0.4301,
+      "step": 43310
+    },
+    {
+      "epoch": 7.003475870988602,
+      "grad_norm": 1.5583289861679077,
+      "learning_rate": 0.0002,
+      "loss": 0.437,
+      "step": 43320
+    },
+    {
+      "epoch": 7.005092555169347,
+      "grad_norm": 1.0714443922042847,
+      "learning_rate": 0.0002,
+      "loss": 0.4532,
+      "step": 43330
+    },
+    {
+      "epoch": 7.006709239350093,
+      "grad_norm": 1.048075795173645,
+      "learning_rate": 0.0002,
+      "loss": 0.4048,
+      "step": 43340
+    },
+    {
+      "epoch": 7.008325923530839,
+      "grad_norm": 1.1053836345672607,
+      "learning_rate": 0.0002,
+      "loss": 0.4119,
+      "step": 43350
+    },
+    {
+      "epoch": 7.009942607711584,
+      "grad_norm": 0.8911725282669067,
+      "learning_rate": 0.0002,
+      "loss": 0.4352,
+      "step": 43360
+    },
+    {
+      "epoch": 7.011559291892329,
+      "grad_norm": 0.9404396414756775,
+      "learning_rate": 0.0002,
+      "loss": 0.4236,
+      "step": 43370
+    },
+    {
+      "epoch": 7.013175976073074,
+      "grad_norm": 1.152365803718567,
+      "learning_rate": 0.0002,
+      "loss": 0.4529,
+      "step": 43380
+    },
+    {
+      "epoch": 7.0147926602538195,
+      "grad_norm": 1.2118251323699951,
+      "learning_rate": 0.0002,
+      "loss": 0.4381,
+      "step": 43390
+    },
+    {
+      "epoch": 7.016409344434565,
+      "grad_norm": 1.2046295404434204,
+      "learning_rate": 0.0002,
+      "loss": 0.4404,
+      "step": 43400
+    },
+    {
+      "epoch": 7.01802602861531,
+      "grad_norm": 0.929465115070343,
+      "learning_rate": 0.0002,
+      "loss": 0.4177,
+      "step": 43410
+    },
+    {
+      "epoch": 7.019642712796055,
+      "grad_norm": 1.3720149993896484,
+      "learning_rate": 0.0002,
+      "loss": 0.4086,
+      "step": 43420
+    },
+    {
+      "epoch": 7.0212593969768005,
+      "grad_norm": 1.1316810846328735,
+      "learning_rate": 0.0002,
+      "loss": 0.4174,
+      "step": 43430
+    },
+    {
+      "epoch": 7.022876081157546,
+      "grad_norm": 1.0342087745666504,
+      "learning_rate": 0.0002,
+      "loss": 0.451,
+      "step": 43440
+    },
+    {
+      "epoch": 7.024492765338291,
+      "grad_norm": 1.1455655097961426,
+      "learning_rate": 0.0002,
+      "loss": 0.4084,
+      "step": 43450
+    },
+    {
+      "epoch": 7.026109449519036,
+      "grad_norm": 1.1308859586715698,
+      "learning_rate": 0.0002,
+      "loss": 0.4168,
+      "step": 43460
+    },
+    {
+      "epoch": 7.027726133699781,
+      "grad_norm": 1.0796722173690796,
+      "learning_rate": 0.0002,
+      "loss": 0.4099,
+      "step": 43470
+    },
+    {
+      "epoch": 7.029342817880527,
+      "grad_norm": 1.0031877756118774,
+      "learning_rate": 0.0002,
+      "loss": 0.4047,
+      "step": 43480
+    },
+    {
+      "epoch": 7.030959502061273,
+      "grad_norm": 1.2391340732574463,
+      "learning_rate": 0.0002,
+      "loss": 0.3968,
+      "step": 43490
+    },
+    {
+      "epoch": 7.032576186242018,
+      "grad_norm": 1.0807358026504517,
+      "learning_rate": 0.0002,
+      "loss": 0.4155,
+      "step": 43500
+    },
+    {
+      "epoch": 7.034192870422763,
+      "grad_norm": 1.230995535850525,
+      "learning_rate": 0.0002,
+      "loss": 0.4322,
+      "step": 43510
+    },
+    {
+      "epoch": 7.035809554603508,
+      "grad_norm": 1.2200509309768677,
+      "learning_rate": 0.0002,
+      "loss": 0.3971,
+      "step": 43520
+    },
+    {
+      "epoch": 7.037426238784254,
+      "grad_norm": 0.9785236120223999,
+      "learning_rate": 0.0002,
+      "loss": 0.4242,
+      "step": 43530
+    },
+    {
+      "epoch": 7.039042922964999,
+      "grad_norm": 1.0009595155715942,
+      "learning_rate": 0.0002,
+      "loss": 0.4173,
+      "step": 43540
+    },
+    {
+      "epoch": 7.040659607145744,
+      "grad_norm": 0.9783103466033936,
+      "learning_rate": 0.0002,
+      "loss": 0.4175,
+      "step": 43550
+    },
+    {
+      "epoch": 7.042276291326489,
+      "grad_norm": 1.1303530931472778,
+      "learning_rate": 0.0002,
+      "loss": 0.4307,
+      "step": 43560
+    },
+    {
+      "epoch": 7.043892975507235,
+      "grad_norm": 1.1768499612808228,
+      "learning_rate": 0.0002,
+      "loss": 0.4066,
+      "step": 43570
+    },
+    {
+      "epoch": 7.04550965968798,
+      "grad_norm": 1.1040459871292114,
+      "learning_rate": 0.0002,
+      "loss": 0.4492,
+      "step": 43580
+    },
+    {
+      "epoch": 7.047126343868725,
+      "grad_norm": 1.0673959255218506,
+      "learning_rate": 0.0002,
+      "loss": 0.4314,
+      "step": 43590
+    },
+    {
+      "epoch": 7.04874302804947,
+      "grad_norm": 1.1220765113830566,
+      "learning_rate": 0.0002,
+      "loss": 0.402,
+      "step": 43600
+    },
+    {
+      "epoch": 7.0503597122302155,
+      "grad_norm": 1.1746923923492432,
+      "learning_rate": 0.0002,
+      "loss": 0.4108,
+      "step": 43610
+    },
+    {
+      "epoch": 7.051976396410961,
+      "grad_norm": 1.2764517068862915,
+      "learning_rate": 0.0002,
+      "loss": 0.4618,
+      "step": 43620
+    },
+    {
+      "epoch": 7.053593080591706,
+      "grad_norm": 1.1180157661437988,
+      "learning_rate": 0.0002,
+      "loss": 0.4243,
+      "step": 43630
+    },
+    {
+      "epoch": 7.055209764772452,
+      "grad_norm": 1.3558318614959717,
+      "learning_rate": 0.0002,
+      "loss": 0.4593,
+      "step": 43640
+    },
+    {
+      "epoch": 7.056826448953197,
+      "grad_norm": 0.9804982542991638,
+      "learning_rate": 0.0002,
+      "loss": 0.4351,
+      "step": 43650
+    },
+    {
+      "epoch": 7.058443133133943,
+      "grad_norm": 1.106404423713684,
+      "learning_rate": 0.0002,
+      "loss": 0.4309,
+      "step": 43660
+    },
+    {
+      "epoch": 7.060059817314688,
+      "grad_norm": 0.9469243884086609,
+      "learning_rate": 0.0002,
+      "loss": 0.4183,
+      "step": 43670
+    },
+    {
+      "epoch": 7.061676501495433,
+      "grad_norm": 1.272987723350525,
+      "learning_rate": 0.0002,
+      "loss": 0.4335,
+      "step": 43680
+    },
+    {
+      "epoch": 7.063293185676178,
+      "grad_norm": 1.0536233186721802,
+      "learning_rate": 0.0002,
+      "loss": 0.4017,
+      "step": 43690
+    },
+    {
+      "epoch": 7.0649098698569235,
+      "grad_norm": 1.1730698347091675,
+      "learning_rate": 0.0002,
+      "loss": 0.4597,
+      "step": 43700
+    },
+    {
+      "epoch": 7.066526554037669,
+      "grad_norm": 1.150707483291626,
+      "learning_rate": 0.0002,
+      "loss": 0.4304,
+      "step": 43710
+    },
+    {
+      "epoch": 7.068143238218414,
+      "grad_norm": 1.4583828449249268,
+      "learning_rate": 0.0002,
+      "loss": 0.4136,
+      "step": 43720
+    },
+    {
+      "epoch": 7.069759922399159,
+      "grad_norm": 1.569705843925476,
+      "learning_rate": 0.0002,
+      "loss": 0.4385,
+      "step": 43730
+    },
+    {
+      "epoch": 7.071376606579904,
+      "grad_norm": 1.156192660331726,
+      "learning_rate": 0.0002,
+      "loss": 0.4051,
+      "step": 43740
+    },
+    {
+      "epoch": 7.07299329076065,
+      "grad_norm": 1.25005304813385,
+      "learning_rate": 0.0002,
+      "loss": 0.4375,
+      "step": 43750
+    },
+    {
+      "epoch": 7.074609974941395,
+      "grad_norm": 1.0468846559524536,
+      "learning_rate": 0.0002,
+      "loss": 0.4096,
+      "step": 43760
+    },
+    {
+      "epoch": 7.07622665912214,
+      "grad_norm": 1.2045108079910278,
+      "learning_rate": 0.0002,
+      "loss": 0.4253,
+      "step": 43770
+    },
+    {
+      "epoch": 7.077843343302886,
+      "grad_norm": 1.1341021060943604,
+      "learning_rate": 0.0002,
+      "loss": 0.4248,
+      "step": 43780
+    },
+    {
+      "epoch": 7.0794600274836315,
+      "grad_norm": 1.0712201595306396,
+      "learning_rate": 0.0002,
+      "loss": 0.394,
+      "step": 43790
+    },
+    {
+      "epoch": 7.081076711664377,
+      "grad_norm": 1.0421321392059326,
+      "learning_rate": 0.0002,
+      "loss": 0.4093,
+      "step": 43800
+    },
+    {
+      "epoch": 7.082693395845122,
+      "grad_norm": 1.2241183519363403,
+      "learning_rate": 0.0002,
+      "loss": 0.4317,
+      "step": 43810
+    },
+    {
+      "epoch": 7.084310080025867,
+      "grad_norm": 1.0945624113082886,
+      "learning_rate": 0.0002,
+      "loss": 0.4064,
+      "step": 43820
+    },
+    {
+      "epoch": 7.085926764206612,
+      "grad_norm": 1.2772969007492065,
+      "learning_rate": 0.0002,
+      "loss": 0.4049,
+      "step": 43830
+    },
+    {
+      "epoch": 7.087543448387358,
+      "grad_norm": 1.1715508699417114,
+      "learning_rate": 0.0002,
+      "loss": 0.4098,
+      "step": 43840
+    },
+    {
+      "epoch": 7.089160132568103,
+      "grad_norm": 1.1975586414337158,
+      "learning_rate": 0.0002,
+      "loss": 0.4756,
+      "step": 43850
+    },
+    {
+      "epoch": 7.090776816748848,
+      "grad_norm": 1.1673274040222168,
+      "learning_rate": 0.0002,
+      "loss": 0.4272,
+      "step": 43860
+    },
+    {
+      "epoch": 7.092393500929593,
+      "grad_norm": 1.096590518951416,
+      "learning_rate": 0.0002,
+      "loss": 0.4435,
+      "step": 43870
+    },
+    {
+      "epoch": 7.0940101851103385,
+      "grad_norm": 1.0174020528793335,
+      "learning_rate": 0.0002,
+      "loss": 0.4329,
+      "step": 43880
+    },
+    {
+      "epoch": 7.095626869291084,
+      "grad_norm": 1.0147380828857422,
+      "learning_rate": 0.0002,
+      "loss": 0.4307,
+      "step": 43890
+    },
+    {
+      "epoch": 7.097243553471829,
+      "grad_norm": 1.0056098699569702,
+      "learning_rate": 0.0002,
+      "loss": 0.4115,
+      "step": 43900
+    },
+    {
+      "epoch": 7.098860237652574,
+      "grad_norm": 1.4678083658218384,
+      "learning_rate": 0.0002,
+      "loss": 0.4181,
+      "step": 43910
+    },
+    {
+      "epoch": 7.1004769218333195,
+      "grad_norm": 1.3740565776824951,
+      "learning_rate": 0.0002,
+      "loss": 0.4404,
+      "step": 43920
+    },
+    {
+      "epoch": 7.102093606014066,
+      "grad_norm": 1.0279403924942017,
+      "learning_rate": 0.0002,
+      "loss": 0.4435,
+      "step": 43930
+    },
+    {
+      "epoch": 7.103710290194811,
+      "grad_norm": 1.186720371246338,
+      "learning_rate": 0.0002,
+      "loss": 0.4247,
+      "step": 43940
+    },
+    {
+      "epoch": 7.105326974375556,
+      "grad_norm": 1.3767904043197632,
+      "learning_rate": 0.0002,
+      "loss": 0.4001,
+      "step": 43950
+    },
+    {
+      "epoch": 7.106943658556301,
+      "grad_norm": 1.1637471914291382,
+      "learning_rate": 0.0002,
+      "loss": 0.4314,
+      "step": 43960
+    },
+    {
+      "epoch": 7.1085603427370465,
+      "grad_norm": 1.1860042810440063,
+      "learning_rate": 0.0002,
+      "loss": 0.3996,
+      "step": 43970
+    },
+    {
+      "epoch": 7.110177026917792,
+      "grad_norm": 1.080944538116455,
+      "learning_rate": 0.0002,
+      "loss": 0.4014,
+      "step": 43980
+    },
+    {
+      "epoch": 7.111793711098537,
+      "grad_norm": 1.0119353532791138,
+      "learning_rate": 0.0002,
+      "loss": 0.4152,
+      "step": 43990
+    },
+    {
+      "epoch": 7.113410395279282,
+      "grad_norm": 1.179388403892517,
+      "learning_rate": 0.0002,
+      "loss": 0.4354,
+      "step": 44000
+    },
+    {
+      "epoch": 7.115027079460027,
+      "grad_norm": 0.9202800989151001,
+      "learning_rate": 0.0002,
+      "loss": 0.4494,
+      "step": 44010
+    },
+    {
+      "epoch": 7.116643763640773,
+      "grad_norm": 1.142206072807312,
+      "learning_rate": 0.0002,
+      "loss": 0.4356,
+      "step": 44020
+    },
+    {
+      "epoch": 7.118260447821518,
+      "grad_norm": 1.17897367477417,
+      "learning_rate": 0.0002,
+      "loss": 0.4197,
+      "step": 44030
+    },
+    {
+      "epoch": 7.119877132002263,
+      "grad_norm": 1.238087773323059,
+      "learning_rate": 0.0002,
+      "loss": 0.4394,
+      "step": 44040
+    },
+    {
+      "epoch": 7.121493816183008,
+      "grad_norm": 1.5113195180892944,
+      "learning_rate": 0.0002,
+      "loss": 0.4844,
+      "step": 44050
+    },
+    {
+      "epoch": 7.123110500363754,
+      "grad_norm": 1.1819349527359009,
+      "learning_rate": 0.0002,
+      "loss": 0.4526,
+      "step": 44060
+    },
+    {
+      "epoch": 7.124727184544499,
+      "grad_norm": 1.1062556505203247,
+      "learning_rate": 0.0002,
+      "loss": 0.4071,
+      "step": 44070
+    },
+    {
+      "epoch": 7.126343868725245,
+      "grad_norm": 0.986954927444458,
+      "learning_rate": 0.0002,
+      "loss": 0.4282,
+      "step": 44080
+    },
+    {
+      "epoch": 7.12796055290599,
+      "grad_norm": 0.9641291499137878,
+      "learning_rate": 0.0002,
+      "loss": 0.4497,
+      "step": 44090
+    },
+    {
+      "epoch": 7.129577237086735,
+      "grad_norm": 0.9519979953765869,
+      "learning_rate": 0.0002,
+      "loss": 0.4348,
+      "step": 44100
+    },
+    {
+      "epoch": 7.131193921267481,
+      "grad_norm": 1.0477287769317627,
+      "learning_rate": 0.0002,
+      "loss": 0.4527,
+      "step": 44110
+    },
+    {
+      "epoch": 7.132810605448226,
+      "grad_norm": 0.9185389280319214,
+      "learning_rate": 0.0002,
+      "loss": 0.4168,
+      "step": 44120
+    },
+    {
+      "epoch": 7.134427289628971,
+      "grad_norm": 1.0224069356918335,
+      "learning_rate": 0.0002,
+      "loss": 0.4255,
+      "step": 44130
+    },
+    {
+      "epoch": 7.136043973809716,
+      "grad_norm": 1.0762630701065063,
+      "learning_rate": 0.0002,
+      "loss": 0.4598,
+      "step": 44140
+    },
+    {
+      "epoch": 7.137660657990462,
+      "grad_norm": 1.330917477607727,
+      "learning_rate": 0.0002,
+      "loss": 0.4308,
+      "step": 44150
+    },
+    {
+      "epoch": 7.139277342171207,
+      "grad_norm": 1.220115303993225,
+      "learning_rate": 0.0002,
+      "loss": 0.4548,
+      "step": 44160
+    },
+    {
+      "epoch": 7.140894026351952,
+      "grad_norm": 0.9959004521369934,
+      "learning_rate": 0.0002,
+      "loss": 0.4089,
+      "step": 44170
+    },
+    {
+      "epoch": 7.142510710532697,
+      "grad_norm": 1.272449016571045,
+      "learning_rate": 0.0002,
+      "loss": 0.4475,
+      "step": 44180
+    },
+    {
+      "epoch": 7.1441273947134425,
+      "grad_norm": 1.0696483850479126,
+      "learning_rate": 0.0002,
+      "loss": 0.4268,
+      "step": 44190
+    },
+    {
+      "epoch": 7.145744078894188,
+      "grad_norm": 1.347206711769104,
+      "learning_rate": 0.0002,
+      "loss": 0.4218,
+      "step": 44200
+    },
+    {
+      "epoch": 7.147360763074933,
+      "grad_norm": 1.1455401182174683,
+      "learning_rate": 0.0002,
+      "loss": 0.4652,
+      "step": 44210
+    },
+    {
+      "epoch": 7.148977447255678,
+      "grad_norm": 1.1443370580673218,
+      "learning_rate": 0.0002,
+      "loss": 0.4186,
+      "step": 44220
+    },
+    {
+      "epoch": 7.150594131436424,
+      "grad_norm": 1.0239921808242798,
+      "learning_rate": 0.0002,
+      "loss": 0.4669,
+      "step": 44230
+    },
+    {
+      "epoch": 7.1522108156171695,
+      "grad_norm": 1.1596333980560303,
+      "learning_rate": 0.0002,
+      "loss": 0.4601,
+      "step": 44240
+    },
+    {
+      "epoch": 7.153827499797915,
+      "grad_norm": 1.2471510171890259,
+      "learning_rate": 0.0002,
+      "loss": 0.44,
+      "step": 44250
+    },
+    {
+      "epoch": 7.15544418397866,
+      "grad_norm": 1.0713822841644287,
+      "learning_rate": 0.0002,
+      "loss": 0.426,
+      "step": 44260
+    },
+    {
+      "epoch": 7.157060868159405,
+      "grad_norm": 1.3523266315460205,
+      "learning_rate": 0.0002,
+      "loss": 0.4381,
+      "step": 44270
+    },
+    {
+      "epoch": 7.1586775523401505,
+      "grad_norm": 1.1620066165924072,
+      "learning_rate": 0.0002,
+      "loss": 0.4101,
+      "step": 44280
+    },
+    {
+      "epoch": 7.160294236520896,
+      "grad_norm": 1.2988619804382324,
+      "learning_rate": 0.0002,
+      "loss": 0.4195,
+      "step": 44290
+    },
+    {
+      "epoch": 7.161910920701641,
+      "grad_norm": 1.2527822256088257,
+      "learning_rate": 0.0002,
+      "loss": 0.4405,
+      "step": 44300
+    },
+    {
+      "epoch": 7.163527604882386,
+      "grad_norm": 1.2322553396224976,
+      "learning_rate": 0.0002,
+      "loss": 0.4813,
+      "step": 44310
+    },
+    {
+      "epoch": 7.165144289063131,
+      "grad_norm": 1.0497055053710938,
+      "learning_rate": 0.0002,
+      "loss": 0.4274,
+      "step": 44320
+    },
+    {
+      "epoch": 7.166760973243877,
+      "grad_norm": 1.1928341388702393,
+      "learning_rate": 0.0002,
+      "loss": 0.4236,
+      "step": 44330
+    },
+    {
+      "epoch": 7.168377657424622,
+      "grad_norm": 1.0016584396362305,
+      "learning_rate": 0.0002,
+      "loss": 0.4511,
+      "step": 44340
+    },
+    {
+      "epoch": 7.169994341605367,
+      "grad_norm": 1.0385509729385376,
+      "learning_rate": 0.0002,
+      "loss": 0.4566,
+      "step": 44350
+    },
+    {
+      "epoch": 7.171611025786112,
+      "grad_norm": 1.3217328786849976,
+      "learning_rate": 0.0002,
+      "loss": 0.4178,
+      "step": 44360
+    },
+    {
+      "epoch": 7.1732277099668575,
+      "grad_norm": 1.240696668624878,
+      "learning_rate": 0.0002,
+      "loss": 0.425,
+      "step": 44370
+    },
+    {
+      "epoch": 7.174844394147604,
+      "grad_norm": 1.1037760972976685,
+      "learning_rate": 0.0002,
+      "loss": 0.4572,
+      "step": 44380
+    },
+    {
+      "epoch": 7.176461078328349,
+      "grad_norm": 1.062762975692749,
+      "learning_rate": 0.0002,
+      "loss": 0.4525,
+      "step": 44390
+    },
+    {
+      "epoch": 7.178077762509094,
+      "grad_norm": 1.2859047651290894,
+      "learning_rate": 0.0002,
+      "loss": 0.4766,
+      "step": 44400
+    },
+    {
+      "epoch": 7.179694446689839,
+      "grad_norm": 1.1852408647537231,
+      "learning_rate": 0.0002,
+      "loss": 0.4511,
+      "step": 44410
+    },
+    {
+      "epoch": 7.181311130870585,
+      "grad_norm": 1.315587043762207,
+      "learning_rate": 0.0002,
+      "loss": 0.4386,
+      "step": 44420
+    },
+    {
+      "epoch": 7.18292781505133,
+      "grad_norm": 0.889542281627655,
+      "learning_rate": 0.0002,
+      "loss": 0.4491,
+      "step": 44430
+    },
+    {
+      "epoch": 7.184544499232075,
+      "grad_norm": 1.0123721361160278,
+      "learning_rate": 0.0002,
+      "loss": 0.4328,
+      "step": 44440
+    },
+    {
+      "epoch": 7.18616118341282,
+      "grad_norm": 1.0503462553024292,
+      "learning_rate": 0.0002,
+      "loss": 0.4096,
+      "step": 44450
+    },
+    {
+      "epoch": 7.1877778675935655,
+      "grad_norm": 1.338188886642456,
+      "learning_rate": 0.0002,
+      "loss": 0.44,
+      "step": 44460
+    },
+    {
+      "epoch": 7.189394551774311,
+      "grad_norm": 1.206543207168579,
+      "learning_rate": 0.0002,
+      "loss": 0.4451,
+      "step": 44470
+    },
+    {
+      "epoch": 7.191011235955056,
+      "grad_norm": 1.2013356685638428,
+      "learning_rate": 0.0002,
+      "loss": 0.4415,
+      "step": 44480
+    },
+    {
+      "epoch": 7.192627920135801,
+      "grad_norm": 1.1124168634414673,
+      "learning_rate": 0.0002,
+      "loss": 0.4291,
+      "step": 44490
+    },
+    {
+      "epoch": 7.194244604316546,
+      "grad_norm": 1.199379324913025,
+      "learning_rate": 0.0002,
+      "loss": 0.4182,
+      "step": 44500
+    },
+    {
+      "epoch": 7.195861288497292,
+      "grad_norm": 1.196746587753296,
+      "learning_rate": 0.0002,
+      "loss": 0.4525,
+      "step": 44510
+    },
+    {
+      "epoch": 7.197477972678037,
+      "grad_norm": 0.9684673547744751,
+      "learning_rate": 0.0002,
+      "loss": 0.4876,
+      "step": 44520
+    },
+    {
+      "epoch": 7.199094656858783,
+      "grad_norm": 1.5727651119232178,
+      "learning_rate": 0.0002,
+      "loss": 0.4403,
+      "step": 44530
+    },
+    {
+      "epoch": 7.200711341039528,
+      "grad_norm": 0.8371674418449402,
+      "learning_rate": 0.0002,
+      "loss": 0.4424,
+      "step": 44540
+    },
+    {
+      "epoch": 7.2023280252202735,
+      "grad_norm": 1.0343716144561768,
+      "learning_rate": 0.0002,
+      "loss": 0.4366,
+      "step": 44550
+    },
+    {
+      "epoch": 7.203944709401019,
+      "grad_norm": 1.1839478015899658,
+      "learning_rate": 0.0002,
+      "loss": 0.4557,
+      "step": 44560
+    },
+    {
+      "epoch": 7.205561393581764,
+      "grad_norm": 0.9466627836227417,
+      "learning_rate": 0.0002,
+      "loss": 0.4293,
+      "step": 44570
+    },
+    {
+      "epoch": 7.207178077762509,
+      "grad_norm": 1.1452360153198242,
+      "learning_rate": 0.0002,
+      "loss": 0.4651,
+      "step": 44580
+    },
+    {
+      "epoch": 7.208794761943254,
+      "grad_norm": 1.63698410987854,
+      "learning_rate": 0.0002,
+      "loss": 0.5037,
+      "step": 44590
+    },
+    {
+      "epoch": 7.210411446124,
+      "grad_norm": 1.1124789714813232,
+      "learning_rate": 0.0002,
+      "loss": 0.4212,
+      "step": 44600
+    },
+    {
+      "epoch": 7.212028130304745,
+      "grad_norm": 1.4233685731887817,
+      "learning_rate": 0.0002,
+      "loss": 0.4323,
+      "step": 44610
+    },
+    {
+      "epoch": 7.21364481448549,
+      "grad_norm": 1.302145004272461,
+      "learning_rate": 0.0002,
+      "loss": 0.4176,
+      "step": 44620
+    },
+    {
+      "epoch": 7.215261498666235,
+      "grad_norm": 1.2115466594696045,
+      "learning_rate": 0.0002,
+      "loss": 0.457,
+      "step": 44630
+    },
+    {
+      "epoch": 7.216878182846981,
+      "grad_norm": 1.0771325826644897,
+      "learning_rate": 0.0002,
+      "loss": 0.4419,
+      "step": 44640
+    },
+    {
+      "epoch": 7.218494867027726,
+      "grad_norm": 1.1603602170944214,
+      "learning_rate": 0.0002,
+      "loss": 0.4183,
+      "step": 44650
+    },
+    {
+      "epoch": 7.220111551208471,
+      "grad_norm": 1.4013969898223877,
+      "learning_rate": 0.0002,
+      "loss": 0.468,
+      "step": 44660
+    },
+    {
+      "epoch": 7.221728235389216,
+      "grad_norm": 1.2145777940750122,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 44670
+    },
+    {
+      "epoch": 7.223344919569962,
+      "grad_norm": 1.2084238529205322,
+      "learning_rate": 0.0002,
+      "loss": 0.4436,
+      "step": 44680
+    },
+    {
+      "epoch": 7.224961603750708,
+      "grad_norm": 1.1801965236663818,
+      "learning_rate": 0.0002,
+      "loss": 0.4423,
+      "step": 44690
+    },
+    {
+      "epoch": 7.226578287931453,
+      "grad_norm": 0.9561195969581604,
+      "learning_rate": 0.0002,
+      "loss": 0.426,
+      "step": 44700
+    },
+    {
+      "epoch": 7.228194972112198,
+      "grad_norm": 1.1857006549835205,
+      "learning_rate": 0.0002,
+      "loss": 0.4895,
+      "step": 44710
+    },
+    {
+      "epoch": 7.229811656292943,
+      "grad_norm": 1.1576673984527588,
+      "learning_rate": 0.0002,
+      "loss": 0.4382,
+      "step": 44720
+    },
+    {
+      "epoch": 7.2314283404736885,
+      "grad_norm": 1.3517892360687256,
+      "learning_rate": 0.0002,
+      "loss": 0.4717,
+      "step": 44730
+    },
+    {
+      "epoch": 7.233045024654434,
+      "grad_norm": 1.1489306688308716,
+      "learning_rate": 0.0002,
+      "loss": 0.4495,
+      "step": 44740
+    },
+    {
+      "epoch": 7.234661708835179,
+      "grad_norm": 1.0758644342422485,
+      "learning_rate": 0.0002,
+      "loss": 0.438,
+      "step": 44750
+    },
+    {
+      "epoch": 7.236278393015924,
+      "grad_norm": 1.1679041385650635,
+      "learning_rate": 0.0002,
+      "loss": 0.4431,
+      "step": 44760
+    },
+    {
+      "epoch": 7.2378950771966695,
+      "grad_norm": 1.1404961347579956,
+      "learning_rate": 0.0002,
+      "loss": 0.504,
+      "step": 44770
+    },
+    {
+      "epoch": 7.239511761377415,
+      "grad_norm": 1.2602572441101074,
+      "learning_rate": 0.0002,
+      "loss": 0.4499,
+      "step": 44780
+    },
+    {
+      "epoch": 7.24112844555816,
+      "grad_norm": 1.2912664413452148,
+      "learning_rate": 0.0002,
+      "loss": 0.4669,
+      "step": 44790
+    },
+    {
+      "epoch": 7.242745129738905,
+      "grad_norm": 1.340198278427124,
+      "learning_rate": 0.0002,
+      "loss": 0.4336,
+      "step": 44800
+    },
+    {
+      "epoch": 7.24436181391965,
+      "grad_norm": 1.0613332986831665,
+      "learning_rate": 0.0002,
+      "loss": 0.4336,
+      "step": 44810
+    },
+    {
+      "epoch": 7.2459784981003965,
+      "grad_norm": 1.1658564805984497,
+      "learning_rate": 0.0002,
+      "loss": 0.4433,
+      "step": 44820
+    },
+    {
+      "epoch": 7.247595182281142,
+      "grad_norm": 1.046440839767456,
+      "learning_rate": 0.0002,
+      "loss": 0.4532,
+      "step": 44830
+    },
+    {
+      "epoch": 7.249211866461887,
+      "grad_norm": 1.2335407733917236,
+      "learning_rate": 0.0002,
+      "loss": 0.4332,
+      "step": 44840
+    },
+    {
+      "epoch": 7.250828550642632,
+      "grad_norm": 1.3742769956588745,
+      "learning_rate": 0.0002,
+      "loss": 0.455,
+      "step": 44850
+    },
+    {
+      "epoch": 7.252445234823377,
+      "grad_norm": 1.1744071245193481,
+      "learning_rate": 0.0002,
+      "loss": 0.4297,
+      "step": 44860
+    },
+    {
+      "epoch": 7.254061919004123,
+      "grad_norm": 1.4268226623535156,
+      "learning_rate": 0.0002,
+      "loss": 0.4348,
+      "step": 44870
+    },
+    {
+      "epoch": 7.255678603184868,
+      "grad_norm": 1.1255686283111572,
+      "learning_rate": 0.0002,
+      "loss": 0.4485,
+      "step": 44880
+    },
+    {
+      "epoch": 7.257295287365613,
+      "grad_norm": 1.255053162574768,
+      "learning_rate": 0.0002,
+      "loss": 0.4264,
+      "step": 44890
+    },
+    {
+      "epoch": 7.258911971546358,
+      "grad_norm": 1.4957616329193115,
+      "learning_rate": 0.0002,
+      "loss": 0.455,
+      "step": 44900
+    },
+    {
+      "epoch": 7.260528655727104,
+      "grad_norm": 1.0546756982803345,
+      "learning_rate": 0.0002,
+      "loss": 0.4465,
+      "step": 44910
+    },
+    {
+      "epoch": 7.262145339907849,
+      "grad_norm": 1.4683036804199219,
+      "learning_rate": 0.0002,
+      "loss": 0.4802,
+      "step": 44920
+    },
+    {
+      "epoch": 7.263762024088594,
+      "grad_norm": 1.2027722597122192,
+      "learning_rate": 0.0002,
+      "loss": 0.4175,
+      "step": 44930
+    },
+    {
+      "epoch": 7.265378708269339,
+      "grad_norm": 1.277767539024353,
+      "learning_rate": 0.0002,
+      "loss": 0.4316,
+      "step": 44940
+    },
+    {
+      "epoch": 7.2669953924500845,
+      "grad_norm": 1.4894379377365112,
+      "learning_rate": 0.0002,
+      "loss": 0.4749,
+      "step": 44950
+    },
+    {
+      "epoch": 7.26861207663083,
+      "grad_norm": 1.0998231172561646,
+      "learning_rate": 0.0002,
+      "loss": 0.46,
+      "step": 44960
+    },
+    {
+      "epoch": 7.270228760811576,
+      "grad_norm": 1.3713536262512207,
+      "learning_rate": 0.0002,
+      "loss": 0.4414,
+      "step": 44970
+    },
+    {
+      "epoch": 7.271845444992321,
+      "grad_norm": 1.473396897315979,
+      "learning_rate": 0.0002,
+      "loss": 0.4431,
+      "step": 44980
+    },
+    {
+      "epoch": 7.273462129173066,
+      "grad_norm": 1.0893826484680176,
+      "learning_rate": 0.0002,
+      "loss": 0.4582,
+      "step": 44990
+    },
+    {
+      "epoch": 7.275078813353812,
+      "grad_norm": 1.4798463582992554,
+      "learning_rate": 0.0002,
+      "loss": 0.4297,
+      "step": 45000
+    },
+    {
+      "epoch": 7.276695497534557,
+      "grad_norm": 1.0536930561065674,
+      "learning_rate": 0.0002,
+      "loss": 0.4604,
+      "step": 45010
+    },
+    {
+      "epoch": 7.278312181715302,
+      "grad_norm": 1.064450740814209,
+      "learning_rate": 0.0002,
+      "loss": 0.4396,
+      "step": 45020
+    },
+    {
+      "epoch": 7.279928865896047,
+      "grad_norm": 1.3605865240097046,
+      "learning_rate": 0.0002,
+      "loss": 0.4445,
+      "step": 45030
+    },
+    {
+      "epoch": 7.2815455500767925,
+      "grad_norm": 1.1779286861419678,
+      "learning_rate": 0.0002,
+      "loss": 0.4123,
+      "step": 45040
+    },
+    {
+      "epoch": 7.283162234257538,
+      "grad_norm": 1.1568892002105713,
+      "learning_rate": 0.0002,
+      "loss": 0.4588,
+      "step": 45050
+    },
+    {
+      "epoch": 7.284778918438283,
+      "grad_norm": 1.0677175521850586,
+      "learning_rate": 0.0002,
+      "loss": 0.4557,
+      "step": 45060
+    },
+    {
+      "epoch": 7.286395602619028,
+      "grad_norm": 1.1939430236816406,
+      "learning_rate": 0.0002,
+      "loss": 0.4765,
+      "step": 45070
+    },
+    {
+      "epoch": 7.288012286799773,
+      "grad_norm": 1.0273144245147705,
+      "learning_rate": 0.0002,
+      "loss": 0.443,
+      "step": 45080
+    },
+    {
+      "epoch": 7.289628970980519,
+      "grad_norm": 1.358487844467163,
+      "learning_rate": 0.0002,
+      "loss": 0.4472,
+      "step": 45090
+    },
+    {
+      "epoch": 7.291245655161264,
+      "grad_norm": 1.2139160633087158,
+      "learning_rate": 0.0002,
+      "loss": 0.4207,
+      "step": 45100
+    },
+    {
+      "epoch": 7.29286233934201,
+      "grad_norm": 1.2484227418899536,
+      "learning_rate": 0.0002,
+      "loss": 0.4221,
+      "step": 45110
+    },
+    {
+      "epoch": 7.294479023522755,
+      "grad_norm": 1.2373738288879395,
+      "learning_rate": 0.0002,
+      "loss": 0.4351,
+      "step": 45120
+    },
+    {
+      "epoch": 7.2960957077035005,
+      "grad_norm": 1.3877158164978027,
+      "learning_rate": 0.0002,
+      "loss": 0.472,
+      "step": 45130
+    },
+    {
+      "epoch": 7.297712391884246,
+      "grad_norm": 1.1372028589248657,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 45140
+    },
+    {
+      "epoch": 7.299329076064991,
+      "grad_norm": 1.259987711906433,
+      "learning_rate": 0.0002,
+      "loss": 0.4465,
+      "step": 45150
+    },
+    {
+      "epoch": 7.300945760245736,
+      "grad_norm": 1.6501492261886597,
+      "learning_rate": 0.0002,
+      "loss": 0.4795,
+      "step": 45160
+    },
+    {
+      "epoch": 7.302562444426481,
+      "grad_norm": 1.5927983522415161,
+      "learning_rate": 0.0002,
+      "loss": 0.4441,
+      "step": 45170
+    },
+    {
+      "epoch": 7.304179128607227,
+      "grad_norm": 0.957084596157074,
+      "learning_rate": 0.0002,
+      "loss": 0.4513,
+      "step": 45180
+    },
+    {
+      "epoch": 7.305795812787972,
+      "grad_norm": 1.7777647972106934,
+      "learning_rate": 0.0002,
+      "loss": 0.4367,
+      "step": 45190
+    },
+    {
+      "epoch": 7.307412496968717,
+      "grad_norm": 1.1905052661895752,
+      "learning_rate": 0.0002,
+      "loss": 0.4365,
+      "step": 45200
+    },
+    {
+      "epoch": 7.309029181149462,
+      "grad_norm": 1.0944236516952515,
+      "learning_rate": 0.0002,
+      "loss": 0.4354,
+      "step": 45210
+    },
+    {
+      "epoch": 7.3106458653302075,
+      "grad_norm": 1.171034336090088,
+      "learning_rate": 0.0002,
+      "loss": 0.4558,
+      "step": 45220
+    },
+    {
+      "epoch": 7.312262549510953,
+      "grad_norm": 1.421743392944336,
+      "learning_rate": 0.0002,
+      "loss": 0.4518,
+      "step": 45230
+    },
+    {
+      "epoch": 7.313879233691698,
+      "grad_norm": 1.1282994747161865,
+      "learning_rate": 0.0002,
+      "loss": 0.4713,
+      "step": 45240
+    },
+    {
+      "epoch": 7.315495917872443,
+      "grad_norm": 1.0742822885513306,
+      "learning_rate": 0.0002,
+      "loss": 0.471,
+      "step": 45250
+    },
+    {
+      "epoch": 7.317112602053189,
+      "grad_norm": 1.2697997093200684,
+      "learning_rate": 0.0002,
+      "loss": 0.4887,
+      "step": 45260
+    },
+    {
+      "epoch": 7.318729286233935,
+      "grad_norm": 1.2066359519958496,
+      "learning_rate": 0.0002,
+      "loss": 0.461,
+      "step": 45270
+    },
+    {
+      "epoch": 7.32034597041468,
+      "grad_norm": 1.0044163465499878,
+      "learning_rate": 0.0002,
+      "loss": 0.463,
+      "step": 45280
+    },
+    {
+      "epoch": 7.321962654595425,
+      "grad_norm": 1.2365968227386475,
+      "learning_rate": 0.0002,
+      "loss": 0.4394,
+      "step": 45290
+    },
+    {
+      "epoch": 7.32357933877617,
+      "grad_norm": 1.0731542110443115,
+      "learning_rate": 0.0002,
+      "loss": 0.4305,
+      "step": 45300
+    },
+    {
+      "epoch": 7.3251960229569155,
+      "grad_norm": 1.1595830917358398,
+      "learning_rate": 0.0002,
+      "loss": 0.4744,
+      "step": 45310
+    },
+    {
+      "epoch": 7.326812707137661,
+      "grad_norm": 1.3445849418640137,
+      "learning_rate": 0.0002,
+      "loss": 0.4393,
+      "step": 45320
+    },
+    {
+      "epoch": 7.328429391318406,
+      "grad_norm": 1.3067926168441772,
+      "learning_rate": 0.0002,
+      "loss": 0.4288,
+      "step": 45330
+    },
+    {
+      "epoch": 7.330046075499151,
+      "grad_norm": 1.200667381286621,
+      "learning_rate": 0.0002,
+      "loss": 0.4569,
+      "step": 45340
+    },
+    {
+      "epoch": 7.3316627596798964,
+      "grad_norm": 0.9936319589614868,
+      "learning_rate": 0.0002,
+      "loss": 0.4449,
+      "step": 45350
+    },
+    {
+      "epoch": 7.333279443860642,
+      "grad_norm": 1.1291998624801636,
+      "learning_rate": 0.0002,
+      "loss": 0.4481,
+      "step": 45360
+    },
+    {
+      "epoch": 7.334896128041387,
+      "grad_norm": 1.3663034439086914,
+      "learning_rate": 0.0002,
+      "loss": 0.4643,
+      "step": 45370
+    },
+    {
+      "epoch": 7.336512812222132,
+      "grad_norm": 1.0762227773666382,
+      "learning_rate": 0.0002,
+      "loss": 0.4548,
+      "step": 45380
+    },
+    {
+      "epoch": 7.338129496402877,
+      "grad_norm": 0.9525768160820007,
+      "learning_rate": 0.0002,
+      "loss": 0.4495,
+      "step": 45390
+    },
+    {
+      "epoch": 7.339746180583623,
+      "grad_norm": 1.1143709421157837,
+      "learning_rate": 0.0002,
+      "loss": 0.472,
+      "step": 45400
+    },
+    {
+      "epoch": 7.341362864764369,
+      "grad_norm": 1.0711175203323364,
+      "learning_rate": 0.0002,
+      "loss": 0.4432,
+      "step": 45410
+    },
+    {
+      "epoch": 7.342979548945114,
+      "grad_norm": 1.2650856971740723,
+      "learning_rate": 0.0002,
+      "loss": 0.4603,
+      "step": 45420
+    },
+    {
+      "epoch": 7.344596233125859,
+      "grad_norm": 1.194861888885498,
+      "learning_rate": 0.0002,
+      "loss": 0.5021,
+      "step": 45430
+    },
+    {
+      "epoch": 7.346212917306604,
+      "grad_norm": 1.4936751127243042,
+      "learning_rate": 0.0002,
+      "loss": 0.467,
+      "step": 45440
+    },
+    {
+      "epoch": 7.34782960148735,
+      "grad_norm": 1.2938975095748901,
+      "learning_rate": 0.0002,
+      "loss": 0.4798,
+      "step": 45450
+    },
+    {
+      "epoch": 7.349446285668095,
+      "grad_norm": 1.2841941118240356,
+      "learning_rate": 0.0002,
+      "loss": 0.4589,
+      "step": 45460
+    },
+    {
+      "epoch": 7.35106296984884,
+      "grad_norm": 1.5376560688018799,
+      "learning_rate": 0.0002,
+      "loss": 0.4398,
+      "step": 45470
+    },
+    {
+      "epoch": 7.352679654029585,
+      "grad_norm": 1.1307156085968018,
+      "learning_rate": 0.0002,
+      "loss": 0.4583,
+      "step": 45480
+    },
+    {
+      "epoch": 7.354296338210331,
+      "grad_norm": 1.4883167743682861,
+      "learning_rate": 0.0002,
+      "loss": 0.4678,
+      "step": 45490
+    },
+    {
+      "epoch": 7.355913022391076,
+      "grad_norm": 1.0547393560409546,
+      "learning_rate": 0.0002,
+      "loss": 0.4966,
+      "step": 45500
+    },
+    {
+      "epoch": 7.357529706571821,
+      "grad_norm": 1.5476845502853394,
+      "learning_rate": 0.0002,
+      "loss": 0.4601,
+      "step": 45510
+    },
+    {
+      "epoch": 7.359146390752566,
+      "grad_norm": 1.1916698217391968,
+      "learning_rate": 0.0002,
+      "loss": 0.4466,
+      "step": 45520
+    },
+    {
+      "epoch": 7.3607630749333115,
+      "grad_norm": 1.238319754600525,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 45530
+    },
+    {
+      "epoch": 7.362379759114057,
+      "grad_norm": 1.4216728210449219,
+      "learning_rate": 0.0002,
+      "loss": 0.4818,
+      "step": 45540
+    },
+    {
+      "epoch": 7.363996443294802,
+      "grad_norm": 1.303995132446289,
+      "learning_rate": 0.0002,
+      "loss": 0.4828,
+      "step": 45550
+    },
+    {
+      "epoch": 7.365613127475548,
+      "grad_norm": 1.2453089952468872,
+      "learning_rate": 0.0002,
+      "loss": 0.464,
+      "step": 45560
+    },
+    {
+      "epoch": 7.367229811656293,
+      "grad_norm": 1.1971137523651123,
+      "learning_rate": 0.0002,
+      "loss": 0.4735,
+      "step": 45570
+    },
+    {
+      "epoch": 7.3688464958370385,
+      "grad_norm": 1.0801963806152344,
+      "learning_rate": 0.0002,
+      "loss": 0.4415,
+      "step": 45580
+    },
+    {
+      "epoch": 7.370463180017784,
+      "grad_norm": 1.1602367162704468,
+      "learning_rate": 0.0002,
+      "loss": 0.4946,
+      "step": 45590
+    },
+    {
+      "epoch": 7.372079864198529,
+      "grad_norm": 1.1623423099517822,
+      "learning_rate": 0.0002,
+      "loss": 0.45,
+      "step": 45600
+    },
+    {
+      "epoch": 7.373696548379274,
+      "grad_norm": 1.108467936515808,
+      "learning_rate": 0.0002,
+      "loss": 0.4648,
+      "step": 45610
+    },
+    {
+      "epoch": 7.3753132325600195,
+      "grad_norm": 1.087322473526001,
+      "learning_rate": 0.0002,
+      "loss": 0.4566,
+      "step": 45620
+    },
+    {
+      "epoch": 7.376929916740765,
+      "grad_norm": 1.0945587158203125,
+      "learning_rate": 0.0002,
+      "loss": 0.4505,
+      "step": 45630
+    },
+    {
+      "epoch": 7.37854660092151,
+      "grad_norm": 1.6565983295440674,
+      "learning_rate": 0.0002,
+      "loss": 0.4864,
+      "step": 45640
+    },
+    {
+      "epoch": 7.380163285102255,
+      "grad_norm": 1.1279444694519043,
+      "learning_rate": 0.0002,
+      "loss": 0.4491,
+      "step": 45650
+    },
+    {
+      "epoch": 7.381779969283,
+      "grad_norm": 1.0888527631759644,
+      "learning_rate": 0.0002,
+      "loss": 0.4606,
+      "step": 45660
+    },
+    {
+      "epoch": 7.383396653463746,
+      "grad_norm": 1.1114956140518188,
+      "learning_rate": 0.0002,
+      "loss": 0.429,
+      "step": 45670
+    },
+    {
+      "epoch": 7.385013337644491,
+      "grad_norm": 1.195497751235962,
+      "learning_rate": 0.0002,
+      "loss": 0.4726,
+      "step": 45680
+    },
+    {
+      "epoch": 7.386630021825236,
+      "grad_norm": 1.3111436367034912,
+      "learning_rate": 0.0002,
+      "loss": 0.4643,
+      "step": 45690
+    },
+    {
+      "epoch": 7.388246706005981,
+      "grad_norm": 1.22647225856781,
+      "learning_rate": 0.0002,
+      "loss": 0.4777,
+      "step": 45700
+    },
+    {
+      "epoch": 7.389863390186727,
+      "grad_norm": 0.9309225678443909,
+      "learning_rate": 0.0002,
+      "loss": 0.4877,
+      "step": 45710
+    },
+    {
+      "epoch": 7.391480074367473,
+      "grad_norm": 1.198773741722107,
+      "learning_rate": 0.0002,
+      "loss": 0.4789,
+      "step": 45720
+    },
+    {
+      "epoch": 7.393096758548218,
+      "grad_norm": 1.2208130359649658,
+      "learning_rate": 0.0002,
+      "loss": 0.4496,
+      "step": 45730
+    },
+    {
+      "epoch": 7.394713442728963,
+      "grad_norm": 1.0756449699401855,
+      "learning_rate": 0.0002,
+      "loss": 0.4614,
+      "step": 45740
+    },
+    {
+      "epoch": 7.396330126909708,
+      "grad_norm": 1.0117692947387695,
+      "learning_rate": 0.0002,
+      "loss": 0.4469,
+      "step": 45750
+    },
+    {
+      "epoch": 7.397946811090454,
+      "grad_norm": 1.1144468784332275,
+      "learning_rate": 0.0002,
+      "loss": 0.4217,
+      "step": 45760
+    },
+    {
+      "epoch": 7.399563495271199,
+      "grad_norm": 1.140549898147583,
+      "learning_rate": 0.0002,
+      "loss": 0.4737,
+      "step": 45770
+    },
+    {
+      "epoch": 7.401180179451944,
+      "grad_norm": 1.2335172891616821,
+      "learning_rate": 0.0002,
+      "loss": 0.453,
+      "step": 45780
+    },
+    {
+      "epoch": 7.402796863632689,
+      "grad_norm": 1.296393632888794,
+      "learning_rate": 0.0002,
+      "loss": 0.4501,
+      "step": 45790
+    },
+    {
+      "epoch": 7.4044135478134345,
+      "grad_norm": 1.2551302909851074,
+      "learning_rate": 0.0002,
+      "loss": 0.4716,
+      "step": 45800
+    },
+    {
+      "epoch": 7.40603023199418,
+      "grad_norm": 1.1909204721450806,
+      "learning_rate": 0.0002,
+      "loss": 0.451,
+      "step": 45810
+    },
+    {
+      "epoch": 7.407646916174925,
+      "grad_norm": 1.17038893699646,
+      "learning_rate": 0.0002,
+      "loss": 0.4296,
+      "step": 45820
+    },
+    {
+      "epoch": 7.40926360035567,
+      "grad_norm": 1.0033377408981323,
+      "learning_rate": 0.0002,
+      "loss": 0.4574,
+      "step": 45830
+    },
+    {
+      "epoch": 7.4108802845364155,
+      "grad_norm": 1.2957805395126343,
+      "learning_rate": 0.0002,
+      "loss": 0.4413,
+      "step": 45840
+    },
+    {
+      "epoch": 7.412496968717161,
+      "grad_norm": 1.347462773323059,
+      "learning_rate": 0.0002,
+      "loss": 0.4597,
+      "step": 45850
+    },
+    {
+      "epoch": 7.414113652897907,
+      "grad_norm": 1.3187026977539062,
+      "learning_rate": 0.0002,
+      "loss": 0.4499,
+      "step": 45860
+    },
+    {
+      "epoch": 7.415730337078652,
+      "grad_norm": 1.092236876487732,
+      "learning_rate": 0.0002,
+      "loss": 0.4624,
+      "step": 45870
+    },
+    {
+      "epoch": 7.417347021259397,
+      "grad_norm": 1.075634241104126,
+      "learning_rate": 0.0002,
+      "loss": 0.4636,
+      "step": 45880
+    },
+    {
+      "epoch": 7.4189637054401425,
+      "grad_norm": 1.0200046300888062,
+      "learning_rate": 0.0002,
+      "loss": 0.4465,
+      "step": 45890
+    },
+    {
+      "epoch": 7.420580389620888,
+      "grad_norm": 1.1419479846954346,
+      "learning_rate": 0.0002,
+      "loss": 0.47,
+      "step": 45900
+    },
+    {
+      "epoch": 7.422197073801633,
+      "grad_norm": 1.0798102617263794,
+      "learning_rate": 0.0002,
+      "loss": 0.4409,
+      "step": 45910
+    },
+    {
+      "epoch": 7.423813757982378,
+      "grad_norm": 0.9999571442604065,
+      "learning_rate": 0.0002,
+      "loss": 0.5173,
+      "step": 45920
+    },
+    {
+      "epoch": 7.425430442163123,
+      "grad_norm": 1.2220723628997803,
+      "learning_rate": 0.0002,
+      "loss": 0.4714,
+      "step": 45930
+    },
+    {
+      "epoch": 7.427047126343869,
+      "grad_norm": 1.1209388971328735,
+      "learning_rate": 0.0002,
+      "loss": 0.4844,
+      "step": 45940
+    },
+    {
+      "epoch": 7.428663810524614,
+      "grad_norm": 1.1198307275772095,
+      "learning_rate": 0.0002,
+      "loss": 0.4534,
+      "step": 45950
+    },
+    {
+      "epoch": 7.430280494705359,
+      "grad_norm": 1.0170516967773438,
+      "learning_rate": 0.0002,
+      "loss": 0.4486,
+      "step": 45960
+    },
+    {
+      "epoch": 7.431897178886104,
+      "grad_norm": 1.2963446378707886,
+      "learning_rate": 0.0002,
+      "loss": 0.488,
+      "step": 45970
+    },
+    {
+      "epoch": 7.43351386306685,
+      "grad_norm": 1.4202494621276855,
+      "learning_rate": 0.0002,
+      "loss": 0.4346,
+      "step": 45980
+    },
+    {
+      "epoch": 7.435130547247595,
+      "grad_norm": 1.066774845123291,
+      "learning_rate": 0.0002,
+      "loss": 0.4917,
+      "step": 45990
+    },
+    {
+      "epoch": 7.43674723142834,
+      "grad_norm": 1.2760428190231323,
+      "learning_rate": 0.0002,
+      "loss": 0.4897,
+      "step": 46000
+    },
+    {
+      "epoch": 7.438363915609086,
+      "grad_norm": 1.530720829963684,
+      "learning_rate": 0.0002,
+      "loss": 0.4562,
+      "step": 46010
+    },
+    {
+      "epoch": 7.439980599789831,
+      "grad_norm": 1.1914178133010864,
+      "learning_rate": 0.0002,
+      "loss": 0.4691,
+      "step": 46020
+    },
+    {
+      "epoch": 7.441597283970577,
+      "grad_norm": 1.466650128364563,
+      "learning_rate": 0.0002,
+      "loss": 0.5038,
+      "step": 46030
+    },
+    {
+      "epoch": 7.443213968151322,
+      "grad_norm": 1.1567928791046143,
+      "learning_rate": 0.0002,
+      "loss": 0.4673,
+      "step": 46040
+    },
+    {
+      "epoch": 7.444830652332067,
+      "grad_norm": 1.252336025238037,
+      "learning_rate": 0.0002,
+      "loss": 0.4778,
+      "step": 46050
+    },
+    {
+      "epoch": 7.446447336512812,
+      "grad_norm": 1.2095589637756348,
+      "learning_rate": 0.0002,
+      "loss": 0.4493,
+      "step": 46060
+    },
+    {
+      "epoch": 7.4480640206935576,
+      "grad_norm": 1.4075263738632202,
+      "learning_rate": 0.0002,
+      "loss": 0.4407,
+      "step": 46070
+    },
+    {
+      "epoch": 7.449680704874303,
+      "grad_norm": 1.2527226209640503,
+      "learning_rate": 0.0002,
+      "loss": 0.4328,
+      "step": 46080
+    },
+    {
+      "epoch": 7.451297389055048,
+      "grad_norm": 1.3044105768203735,
+      "learning_rate": 0.0002,
+      "loss": 0.4922,
+      "step": 46090
+    },
+    {
+      "epoch": 7.452914073235793,
+      "grad_norm": 1.2888941764831543,
+      "learning_rate": 0.0002,
+      "loss": 0.4465,
+      "step": 46100
+    },
+    {
+      "epoch": 7.4545307574165385,
+      "grad_norm": 1.3148317337036133,
+      "learning_rate": 0.0002,
+      "loss": 0.4414,
+      "step": 46110
+    },
+    {
+      "epoch": 7.456147441597284,
+      "grad_norm": 0.9526162147521973,
+      "learning_rate": 0.0002,
+      "loss": 0.4431,
+      "step": 46120
+    },
+    {
+      "epoch": 7.457764125778029,
+      "grad_norm": 1.2618519067764282,
+      "learning_rate": 0.0002,
+      "loss": 0.4422,
+      "step": 46130
+    },
+    {
+      "epoch": 7.459380809958774,
+      "grad_norm": 1.0392966270446777,
+      "learning_rate": 0.0002,
+      "loss": 0.4745,
+      "step": 46140
+    },
+    {
+      "epoch": 7.460997494139519,
+      "grad_norm": 1.3286794424057007,
+      "learning_rate": 0.0002,
+      "loss": 0.4589,
+      "step": 46150
+    },
+    {
+      "epoch": 7.4626141783202655,
+      "grad_norm": 1.2377561330795288,
+      "learning_rate": 0.0002,
+      "loss": 0.4762,
+      "step": 46160
+    },
+    {
+      "epoch": 7.464230862501011,
+      "grad_norm": 1.034134030342102,
+      "learning_rate": 0.0002,
+      "loss": 0.4119,
+      "step": 46170
+    },
+    {
+      "epoch": 7.465847546681756,
+      "grad_norm": 1.1719683408737183,
+      "learning_rate": 0.0002,
+      "loss": 0.4487,
+      "step": 46180
+    },
+    {
+      "epoch": 7.467464230862501,
+      "grad_norm": 1.182691216468811,
+      "learning_rate": 0.0002,
+      "loss": 0.4423,
+      "step": 46190
+    },
+    {
+      "epoch": 7.4690809150432464,
+      "grad_norm": 1.1898412704467773,
+      "learning_rate": 0.0002,
+      "loss": 0.4341,
+      "step": 46200
+    },
+    {
+      "epoch": 7.470697599223992,
+      "grad_norm": 1.0543978214263916,
+      "learning_rate": 0.0002,
+      "loss": 0.4753,
+      "step": 46210
+    },
+    {
+      "epoch": 7.472314283404737,
+      "grad_norm": 1.176971673965454,
+      "learning_rate": 0.0002,
+      "loss": 0.4673,
+      "step": 46220
+    },
+    {
+      "epoch": 7.473930967585482,
+      "grad_norm": 1.129456639289856,
+      "learning_rate": 0.0002,
+      "loss": 0.4598,
+      "step": 46230
+    },
+    {
+      "epoch": 7.475547651766227,
+      "grad_norm": 1.1782855987548828,
+      "learning_rate": 0.0002,
+      "loss": 0.4805,
+      "step": 46240
+    },
+    {
+      "epoch": 7.477164335946973,
+      "grad_norm": 1.1678800582885742,
+      "learning_rate": 0.0002,
+      "loss": 0.4979,
+      "step": 46250
+    },
+    {
+      "epoch": 7.478781020127718,
+      "grad_norm": 0.9768722653388977,
+      "learning_rate": 0.0002,
+      "loss": 0.4374,
+      "step": 46260
+    },
+    {
+      "epoch": 7.480397704308463,
+      "grad_norm": 1.3222670555114746,
+      "learning_rate": 0.0002,
+      "loss": 0.4683,
+      "step": 46270
+    },
+    {
+      "epoch": 7.482014388489208,
+      "grad_norm": 1.0573948621749878,
+      "learning_rate": 0.0002,
+      "loss": 0.459,
+      "step": 46280
+    },
+    {
+      "epoch": 7.4836310726699535,
+      "grad_norm": 1.3233898878097534,
+      "learning_rate": 0.0002,
+      "loss": 0.5019,
+      "step": 46290
+    },
+    {
+      "epoch": 7.485247756850699,
+      "grad_norm": 0.9695420265197754,
+      "learning_rate": 0.0002,
+      "loss": 0.4689,
+      "step": 46300
+    },
+    {
+      "epoch": 7.486864441031445,
+      "grad_norm": 1.2072020769119263,
+      "learning_rate": 0.0002,
+      "loss": 0.471,
+      "step": 46310
+    },
+    {
+      "epoch": 7.48848112521219,
+      "grad_norm": 1.2161253690719604,
+      "learning_rate": 0.0002,
+      "loss": 0.486,
+      "step": 46320
+    },
+    {
+      "epoch": 7.490097809392935,
+      "grad_norm": 1.185958743095398,
+      "learning_rate": 0.0002,
+      "loss": 0.4581,
+      "step": 46330
+    },
+    {
+      "epoch": 7.491714493573681,
+      "grad_norm": 1.3741549253463745,
+      "learning_rate": 0.0002,
+      "loss": 0.4617,
+      "step": 46340
+    },
+    {
+      "epoch": 7.493331177754426,
+      "grad_norm": 1.0586212873458862,
+      "learning_rate": 0.0002,
+      "loss": 0.4772,
+      "step": 46350
+    },
+    {
+      "epoch": 7.494947861935171,
+      "grad_norm": 1.2000513076782227,
+      "learning_rate": 0.0002,
+      "loss": 0.4644,
+      "step": 46360
+    },
+    {
+      "epoch": 7.496564546115916,
+      "grad_norm": 1.3326879739761353,
+      "learning_rate": 0.0002,
+      "loss": 0.4584,
+      "step": 46370
+    },
+    {
+      "epoch": 7.4981812302966615,
+      "grad_norm": 1.3452857732772827,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 46380
+    },
+    {
+      "epoch": 7.499797914477407,
+      "grad_norm": 1.2885284423828125,
+      "learning_rate": 0.0002,
+      "loss": 0.4747,
+      "step": 46390
+    },
+    {
+      "epoch": 7.501414598658152,
+      "grad_norm": 1.097342610359192,
+      "learning_rate": 0.0002,
+      "loss": 0.4648,
+      "step": 46400
+    },
+    {
+      "epoch": 7.503031282838897,
+      "grad_norm": 1.2342469692230225,
+      "learning_rate": 0.0002,
+      "loss": 0.4714,
+      "step": 46410
+    },
+    {
+      "epoch": 7.504647967019642,
+      "grad_norm": 1.0151721239089966,
+      "learning_rate": 0.0002,
+      "loss": 0.4421,
+      "step": 46420
+    },
+    {
+      "epoch": 7.506264651200388,
+      "grad_norm": 1.2487123012542725,
+      "learning_rate": 0.0002,
+      "loss": 0.4347,
+      "step": 46430
+    },
+    {
+      "epoch": 7.507881335381134,
+      "grad_norm": 0.9319046139717102,
+      "learning_rate": 0.0002,
+      "loss": 0.4768,
+      "step": 46440
+    },
+    {
+      "epoch": 7.509498019561878,
+      "grad_norm": 1.1362226009368896,
+      "learning_rate": 0.0002,
+      "loss": 0.4693,
+      "step": 46450
+    },
+    {
+      "epoch": 7.511114703742624,
+      "grad_norm": 1.2883973121643066,
+      "learning_rate": 0.0002,
+      "loss": 0.5007,
+      "step": 46460
+    },
+    {
+      "epoch": 7.5127313879233695,
+      "grad_norm": 1.0892037153244019,
+      "learning_rate": 0.0002,
+      "loss": 0.4455,
+      "step": 46470
+    },
+    {
+      "epoch": 7.514348072104115,
+      "grad_norm": 1.1870533227920532,
+      "learning_rate": 0.0002,
+      "loss": 0.4721,
+      "step": 46480
+    },
+    {
+      "epoch": 7.51596475628486,
+      "grad_norm": 1.2103877067565918,
+      "learning_rate": 0.0002,
+      "loss": 0.4824,
+      "step": 46490
+    },
+    {
+      "epoch": 7.517581440465605,
+      "grad_norm": 1.0980644226074219,
+      "learning_rate": 0.0002,
+      "loss": 0.4573,
+      "step": 46500
+    },
+    {
+      "epoch": 7.51919812464635,
+      "grad_norm": 1.4729726314544678,
+      "learning_rate": 0.0002,
+      "loss": 0.4759,
+      "step": 46510
+    },
+    {
+      "epoch": 7.520814808827096,
+      "grad_norm": 1.1808913946151733,
+      "learning_rate": 0.0002,
+      "loss": 0.4413,
+      "step": 46520
+    },
+    {
+      "epoch": 7.522431493007841,
+      "grad_norm": 1.2347747087478638,
+      "learning_rate": 0.0002,
+      "loss": 0.4278,
+      "step": 46530
+    },
+    {
+      "epoch": 7.524048177188586,
+      "grad_norm": 1.5921525955200195,
+      "learning_rate": 0.0002,
+      "loss": 0.4745,
+      "step": 46540
+    },
+    {
+      "epoch": 7.525664861369331,
+      "grad_norm": 1.1328861713409424,
+      "learning_rate": 0.0002,
+      "loss": 0.4525,
+      "step": 46550
+    },
+    {
+      "epoch": 7.527281545550077,
+      "grad_norm": 1.289947748184204,
+      "learning_rate": 0.0002,
+      "loss": 0.4771,
+      "step": 46560
+    },
+    {
+      "epoch": 7.528898229730822,
+      "grad_norm": 1.0198370218276978,
+      "learning_rate": 0.0002,
+      "loss": 0.4711,
+      "step": 46570
+    },
+    {
+      "epoch": 7.530514913911567,
+      "grad_norm": 1.3007137775421143,
+      "learning_rate": 0.0002,
+      "loss": 0.504,
+      "step": 46580
+    },
+    {
+      "epoch": 7.532131598092313,
+      "grad_norm": 1.2864280939102173,
+      "learning_rate": 0.0002,
+      "loss": 0.4496,
+      "step": 46590
+    },
+    {
+      "epoch": 7.5337482822730575,
+      "grad_norm": 1.1005513668060303,
+      "learning_rate": 0.0002,
+      "loss": 0.463,
+      "step": 46600
+    },
+    {
+      "epoch": 7.535364966453804,
+      "grad_norm": 0.9998318552970886,
+      "learning_rate": 0.0002,
+      "loss": 0.4426,
+      "step": 46610
+    },
+    {
+      "epoch": 7.536981650634549,
+      "grad_norm": 1.2042466402053833,
+      "learning_rate": 0.0002,
+      "loss": 0.4762,
+      "step": 46620
+    },
+    {
+      "epoch": 7.538598334815294,
+      "grad_norm": 1.3240692615509033,
+      "learning_rate": 0.0002,
+      "loss": 0.4685,
+      "step": 46630
+    },
+    {
+      "epoch": 7.540215018996039,
+      "grad_norm": 1.2145483493804932,
+      "learning_rate": 0.0002,
+      "loss": 0.4608,
+      "step": 46640
+    },
+    {
+      "epoch": 7.5418317031767845,
+      "grad_norm": 1.169691801071167,
+      "learning_rate": 0.0002,
+      "loss": 0.4608,
+      "step": 46650
+    },
+    {
+      "epoch": 7.54344838735753,
+      "grad_norm": 1.194045901298523,
+      "learning_rate": 0.0002,
+      "loss": 0.4527,
+      "step": 46660
+    },
+    {
+      "epoch": 7.545065071538275,
+      "grad_norm": 1.0481327772140503,
+      "learning_rate": 0.0002,
+      "loss": 0.4599,
+      "step": 46670
+    },
+    {
+      "epoch": 7.54668175571902,
+      "grad_norm": 1.0714460611343384,
+      "learning_rate": 0.0002,
+      "loss": 0.4729,
+      "step": 46680
+    },
+    {
+      "epoch": 7.5482984398997655,
+      "grad_norm": 1.1811443567276,
+      "learning_rate": 0.0002,
+      "loss": 0.4703,
+      "step": 46690
+    },
+    {
+      "epoch": 7.549915124080511,
+      "grad_norm": 1.2794281244277954,
+      "learning_rate": 0.0002,
+      "loss": 0.4628,
+      "step": 46700
+    },
+    {
+      "epoch": 7.551531808261256,
+      "grad_norm": 1.001287817955017,
+      "learning_rate": 0.0002,
+      "loss": 0.4659,
+      "step": 46710
+    },
+    {
+      "epoch": 7.553148492442001,
+      "grad_norm": 1.3598867654800415,
+      "learning_rate": 0.0002,
+      "loss": 0.4938,
+      "step": 46720
+    },
+    {
+      "epoch": 7.554765176622746,
+      "grad_norm": 1.206254482269287,
+      "learning_rate": 0.0002,
+      "loss": 0.4731,
+      "step": 46730
+    },
+    {
+      "epoch": 7.5563818608034925,
+      "grad_norm": 1.1095832586288452,
+      "learning_rate": 0.0002,
+      "loss": 0.4581,
+      "step": 46740
+    },
+    {
+      "epoch": 7.557998544984237,
+      "grad_norm": 1.3912206888198853,
+      "learning_rate": 0.0002,
+      "loss": 0.4625,
+      "step": 46750
+    },
+    {
+      "epoch": 7.559615229164983,
+      "grad_norm": 0.9883413314819336,
+      "learning_rate": 0.0002,
+      "loss": 0.4464,
+      "step": 46760
+    },
+    {
+      "epoch": 7.561231913345728,
+      "grad_norm": 1.0965087413787842,
+      "learning_rate": 0.0002,
+      "loss": 0.4535,
+      "step": 46770
+    },
+    {
+      "epoch": 7.562848597526473,
+      "grad_norm": 1.092261552810669,
+      "learning_rate": 0.0002,
+      "loss": 0.469,
+      "step": 46780
+    },
+    {
+      "epoch": 7.564465281707219,
+      "grad_norm": 1.0443673133850098,
+      "learning_rate": 0.0002,
+      "loss": 0.488,
+      "step": 46790
+    },
+    {
+      "epoch": 7.566081965887964,
+      "grad_norm": 1.2420614957809448,
+      "learning_rate": 0.0002,
+      "loss": 0.4875,
+      "step": 46800
+    },
+    {
+      "epoch": 7.567698650068709,
+      "grad_norm": 1.0510783195495605,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 46810
+    },
+    {
+      "epoch": 7.569315334249454,
+      "grad_norm": 1.0291800498962402,
+      "learning_rate": 0.0002,
+      "loss": 0.4541,
+      "step": 46820
+    },
+    {
+      "epoch": 7.5709320184302,
+      "grad_norm": 1.1784595251083374,
+      "learning_rate": 0.0002,
+      "loss": 0.4591,
+      "step": 46830
+    },
+    {
+      "epoch": 7.572548702610945,
+      "grad_norm": 1.0424436330795288,
+      "learning_rate": 0.0002,
+      "loss": 0.5154,
+      "step": 46840
+    },
+    {
+      "epoch": 7.57416538679169,
+      "grad_norm": 1.182131290435791,
+      "learning_rate": 0.0002,
+      "loss": 0.4612,
+      "step": 46850
+    },
+    {
+      "epoch": 7.575782070972435,
+      "grad_norm": 0.9917051792144775,
+      "learning_rate": 0.0002,
+      "loss": 0.446,
+      "step": 46860
+    },
+    {
+      "epoch": 7.5773987551531805,
+      "grad_norm": 1.1616078615188599,
+      "learning_rate": 0.0002,
+      "loss": 0.4428,
+      "step": 46870
+    },
+    {
+      "epoch": 7.579015439333926,
+      "grad_norm": 1.401071548461914,
+      "learning_rate": 0.0002,
+      "loss": 0.4769,
+      "step": 46880
+    },
+    {
+      "epoch": 7.580632123514672,
+      "grad_norm": 0.874487578868866,
+      "learning_rate": 0.0002,
+      "loss": 0.4635,
+      "step": 46890
+    },
+    {
+      "epoch": 7.582248807695416,
+      "grad_norm": 1.2511193752288818,
+      "learning_rate": 0.0002,
+      "loss": 0.4641,
+      "step": 46900
+    },
+    {
+      "epoch": 7.583865491876162,
+      "grad_norm": 1.7548277378082275,
+      "learning_rate": 0.0002,
+      "loss": 0.4715,
+      "step": 46910
+    },
+    {
+      "epoch": 7.5854821760569076,
+      "grad_norm": 1.349366545677185,
+      "learning_rate": 0.0002,
+      "loss": 0.4681,
+      "step": 46920
+    },
+    {
+      "epoch": 7.587098860237653,
+      "grad_norm": 1.0609583854675293,
+      "learning_rate": 0.0002,
+      "loss": 0.4819,
+      "step": 46930
+    },
+    {
+      "epoch": 7.588715544418398,
+      "grad_norm": 1.031512975692749,
+      "learning_rate": 0.0002,
+      "loss": 0.4498,
+      "step": 46940
+    },
+    {
+      "epoch": 7.590332228599143,
+      "grad_norm": 1.1440242528915405,
+      "learning_rate": 0.0002,
+      "loss": 0.4688,
+      "step": 46950
+    },
+    {
+      "epoch": 7.5919489127798885,
+      "grad_norm": 1.2762987613677979,
+      "learning_rate": 0.0002,
+      "loss": 0.4568,
+      "step": 46960
+    },
+    {
+      "epoch": 7.593565596960634,
+      "grad_norm": 1.167269229888916,
+      "learning_rate": 0.0002,
+      "loss": 0.4569,
+      "step": 46970
+    },
+    {
+      "epoch": 7.595182281141379,
+      "grad_norm": 1.131127953529358,
+      "learning_rate": 0.0002,
+      "loss": 0.461,
+      "step": 46980
+    },
+    {
+      "epoch": 7.596798965322124,
+      "grad_norm": 1.4527075290679932,
+      "learning_rate": 0.0002,
+      "loss": 0.4666,
+      "step": 46990
+    },
+    {
+      "epoch": 7.598415649502869,
+      "grad_norm": 1.330132007598877,
+      "learning_rate": 0.0002,
+      "loss": 0.4973,
+      "step": 47000
+    },
+    {
+      "epoch": 7.600032333683615,
+      "grad_norm": 1.4223501682281494,
+      "learning_rate": 0.0002,
+      "loss": 0.4969,
+      "step": 47010
+    },
+    {
+      "epoch": 7.60164901786436,
+      "grad_norm": 1.2045072317123413,
+      "learning_rate": 0.0002,
+      "loss": 0.4572,
+      "step": 47020
+    },
+    {
+      "epoch": 7.603265702045105,
+      "grad_norm": 1.1549896001815796,
+      "learning_rate": 0.0002,
+      "loss": 0.4666,
+      "step": 47030
+    },
+    {
+      "epoch": 7.604882386225851,
+      "grad_norm": 1.2221543788909912,
+      "learning_rate": 0.0002,
+      "loss": 0.4383,
+      "step": 47040
+    },
+    {
+      "epoch": 7.6064990704065965,
+      "grad_norm": 1.1171326637268066,
+      "learning_rate": 0.0002,
+      "loss": 0.4826,
+      "step": 47050
+    },
+    {
+      "epoch": 7.608115754587342,
+      "grad_norm": 1.073671817779541,
+      "learning_rate": 0.0002,
+      "loss": 0.4465,
+      "step": 47060
+    },
+    {
+      "epoch": 7.609732438768087,
+      "grad_norm": 1.2524123191833496,
+      "learning_rate": 0.0002,
+      "loss": 0.4623,
+      "step": 47070
+    },
+    {
+      "epoch": 7.611349122948832,
+      "grad_norm": 1.2015056610107422,
+      "learning_rate": 0.0002,
+      "loss": 0.4538,
+      "step": 47080
+    },
+    {
+      "epoch": 7.612965807129577,
+      "grad_norm": 1.2454534769058228,
+      "learning_rate": 0.0002,
+      "loss": 0.4871,
+      "step": 47090
+    },
+    {
+      "epoch": 7.614582491310323,
+      "grad_norm": 0.9815779328346252,
+      "learning_rate": 0.0002,
+      "loss": 0.5064,
+      "step": 47100
+    },
+    {
+      "epoch": 7.616199175491068,
+      "grad_norm": 1.1437602043151855,
+      "learning_rate": 0.0002,
+      "loss": 0.4841,
+      "step": 47110
+    },
+    {
+      "epoch": 7.617815859671813,
+      "grad_norm": 1.1004078388214111,
+      "learning_rate": 0.0002,
+      "loss": 0.453,
+      "step": 47120
+    },
+    {
+      "epoch": 7.619432543852558,
+      "grad_norm": 1.069453477859497,
+      "learning_rate": 0.0002,
+      "loss": 0.4552,
+      "step": 47130
+    },
+    {
+      "epoch": 7.6210492280333035,
+      "grad_norm": 1.1434191465377808,
+      "learning_rate": 0.0002,
+      "loss": 0.4627,
+      "step": 47140
+    },
+    {
+      "epoch": 7.622665912214049,
+      "grad_norm": 1.216845989227295,
+      "learning_rate": 0.0002,
+      "loss": 0.4882,
+      "step": 47150
+    },
+    {
+      "epoch": 7.624282596394794,
+      "grad_norm": 1.2302134037017822,
+      "learning_rate": 0.0002,
+      "loss": 0.481,
+      "step": 47160
+    },
+    {
+      "epoch": 7.625899280575539,
+      "grad_norm": 1.4284924268722534,
+      "learning_rate": 0.0002,
+      "loss": 0.4806,
+      "step": 47170
+    },
+    {
+      "epoch": 7.6275159647562845,
+      "grad_norm": 1.3359615802764893,
+      "learning_rate": 0.0002,
+      "loss": 0.4458,
+      "step": 47180
+    },
+    {
+      "epoch": 7.629132648937031,
+      "grad_norm": 1.0242379903793335,
+      "learning_rate": 0.0002,
+      "loss": 0.4842,
+      "step": 47190
+    },
+    {
+      "epoch": 7.630749333117776,
+      "grad_norm": 1.249513030052185,
+      "learning_rate": 0.0002,
+      "loss": 0.5137,
+      "step": 47200
+    },
+    {
+      "epoch": 7.632366017298521,
+      "grad_norm": 1.0881463289260864,
+      "learning_rate": 0.0002,
+      "loss": 0.4462,
+      "step": 47210
+    },
+    {
+      "epoch": 7.633982701479266,
+      "grad_norm": 1.2903773784637451,
+      "learning_rate": 0.0002,
+      "loss": 0.4864,
+      "step": 47220
+    },
+    {
+      "epoch": 7.6355993856600115,
+      "grad_norm": 1.1671710014343262,
+      "learning_rate": 0.0002,
+      "loss": 0.4729,
+      "step": 47230
+    },
+    {
+      "epoch": 7.637216069840757,
+      "grad_norm": 1.1960735321044922,
+      "learning_rate": 0.0002,
+      "loss": 0.4936,
+      "step": 47240
+    },
+    {
+      "epoch": 7.638832754021502,
+      "grad_norm": 1.2692298889160156,
+      "learning_rate": 0.0002,
+      "loss": 0.4884,
+      "step": 47250
+    },
+    {
+      "epoch": 7.640449438202247,
+      "grad_norm": 0.9812195301055908,
+      "learning_rate": 0.0002,
+      "loss": 0.423,
+      "step": 47260
+    },
+    {
+      "epoch": 7.642066122382992,
+      "grad_norm": 1.3986053466796875,
+      "learning_rate": 0.0002,
+      "loss": 0.4737,
+      "step": 47270
+    },
+    {
+      "epoch": 7.643682806563738,
+      "grad_norm": 1.2692067623138428,
+      "learning_rate": 0.0002,
+      "loss": 0.4834,
+      "step": 47280
+    },
+    {
+      "epoch": 7.645299490744483,
+      "grad_norm": 1.1185054779052734,
+      "learning_rate": 0.0002,
+      "loss": 0.4893,
+      "step": 47290
+    },
+    {
+      "epoch": 7.646916174925228,
+      "grad_norm": 1.2837327718734741,
+      "learning_rate": 0.0002,
+      "loss": 0.4828,
+      "step": 47300
+    },
+    {
+      "epoch": 7.648532859105973,
+      "grad_norm": 1.8518418073654175,
+      "learning_rate": 0.0002,
+      "loss": 0.4891,
+      "step": 47310
+    },
+    {
+      "epoch": 7.650149543286719,
+      "grad_norm": 0.9781302213668823,
+      "learning_rate": 0.0002,
+      "loss": 0.4626,
+      "step": 47320
+    },
+    {
+      "epoch": 7.651766227467464,
+      "grad_norm": 1.0777910947799683,
+      "learning_rate": 0.0002,
+      "loss": 0.501,
+      "step": 47330
+    },
+    {
+      "epoch": 7.65338291164821,
+      "grad_norm": 1.2031499147415161,
+      "learning_rate": 0.0002,
+      "loss": 0.4927,
+      "step": 47340
+    },
+    {
+      "epoch": 7.654999595828955,
+      "grad_norm": 1.14322829246521,
+      "learning_rate": 0.0002,
+      "loss": 0.439,
+      "step": 47350
+    },
+    {
+      "epoch": 7.6566162800097,
+      "grad_norm": 1.3211992979049683,
+      "learning_rate": 0.0002,
+      "loss": 0.4481,
+      "step": 47360
+    },
+    {
+      "epoch": 7.658232964190446,
+      "grad_norm": 1.3632899522781372,
+      "learning_rate": 0.0002,
+      "loss": 0.4462,
+      "step": 47370
+    },
+    {
+      "epoch": 7.659849648371191,
+      "grad_norm": 1.2593929767608643,
+      "learning_rate": 0.0002,
+      "loss": 0.4934,
+      "step": 47380
+    },
+    {
+      "epoch": 7.661466332551936,
+      "grad_norm": 1.442670464515686,
+      "learning_rate": 0.0002,
+      "loss": 0.4645,
+      "step": 47390
+    },
+    {
+      "epoch": 7.663083016732681,
+      "grad_norm": 1.2304763793945312,
+      "learning_rate": 0.0002,
+      "loss": 0.4584,
+      "step": 47400
+    },
+    {
+      "epoch": 7.664699700913427,
+      "grad_norm": 1.0182652473449707,
+      "learning_rate": 0.0002,
+      "loss": 0.464,
+      "step": 47410
+    },
+    {
+      "epoch": 7.666316385094172,
+      "grad_norm": 1.365441083908081,
+      "learning_rate": 0.0002,
+      "loss": 0.457,
+      "step": 47420
+    },
+    {
+      "epoch": 7.667933069274917,
+      "grad_norm": 1.1578556299209595,
+      "learning_rate": 0.0002,
+      "loss": 0.4787,
+      "step": 47430
+    },
+    {
+      "epoch": 7.669549753455662,
+      "grad_norm": 1.0346194505691528,
+      "learning_rate": 0.0002,
+      "loss": 0.486,
+      "step": 47440
+    },
+    {
+      "epoch": 7.6711664376364075,
+      "grad_norm": 1.2567378282546997,
+      "learning_rate": 0.0002,
+      "loss": 0.4703,
+      "step": 47450
+    },
+    {
+      "epoch": 7.672783121817153,
+      "grad_norm": 1.1669118404388428,
+      "learning_rate": 0.0002,
+      "loss": 0.4853,
+      "step": 47460
+    },
+    {
+      "epoch": 7.674399805997898,
+      "grad_norm": 1.0174756050109863,
+      "learning_rate": 0.0002,
+      "loss": 0.4869,
+      "step": 47470
+    },
+    {
+      "epoch": 7.676016490178643,
+      "grad_norm": 1.0962231159210205,
+      "learning_rate": 0.0002,
+      "loss": 0.4601,
+      "step": 47480
+    },
+    {
+      "epoch": 7.677633174359389,
+      "grad_norm": 1.1098674535751343,
+      "learning_rate": 0.0002,
+      "loss": 0.4866,
+      "step": 47490
+    },
+    {
+      "epoch": 7.6792498585401345,
+      "grad_norm": 1.1441160440444946,
+      "learning_rate": 0.0002,
+      "loss": 0.4682,
+      "step": 47500
+    },
+    {
+      "epoch": 7.68086654272088,
+      "grad_norm": 1.0473432540893555,
+      "learning_rate": 0.0002,
+      "loss": 0.4552,
+      "step": 47510
+    },
+    {
+      "epoch": 7.682483226901625,
+      "grad_norm": 1.2954738140106201,
+      "learning_rate": 0.0002,
+      "loss": 0.4771,
+      "step": 47520
+    },
+    {
+      "epoch": 7.68409991108237,
+      "grad_norm": 1.2931294441223145,
+      "learning_rate": 0.0002,
+      "loss": 0.5012,
+      "step": 47530
+    },
+    {
+      "epoch": 7.6857165952631155,
+      "grad_norm": 1.4005156755447388,
+      "learning_rate": 0.0002,
+      "loss": 0.4808,
+      "step": 47540
+    },
+    {
+      "epoch": 7.687333279443861,
+      "grad_norm": 1.0998929738998413,
+      "learning_rate": 0.0002,
+      "loss": 0.4847,
+      "step": 47550
+    },
+    {
+      "epoch": 7.688949963624606,
+      "grad_norm": 1.3478347063064575,
+      "learning_rate": 0.0002,
+      "loss": 0.4839,
+      "step": 47560
+    },
+    {
+      "epoch": 7.690566647805351,
+      "grad_norm": 1.2991969585418701,
+      "learning_rate": 0.0002,
+      "loss": 0.4918,
+      "step": 47570
+    },
+    {
+      "epoch": 7.692183331986096,
+      "grad_norm": 1.0892608165740967,
+      "learning_rate": 0.0002,
+      "loss": 0.4673,
+      "step": 47580
+    },
+    {
+      "epoch": 7.693800016166842,
+      "grad_norm": 1.2230998277664185,
+      "learning_rate": 0.0002,
+      "loss": 0.4937,
+      "step": 47590
+    },
+    {
+      "epoch": 7.695416700347587,
+      "grad_norm": 1.2635555267333984,
+      "learning_rate": 0.0002,
+      "loss": 0.5222,
+      "step": 47600
+    },
+    {
+      "epoch": 7.697033384528332,
+      "grad_norm": 1.1720705032348633,
+      "learning_rate": 0.0002,
+      "loss": 0.4883,
+      "step": 47610
+    },
+    {
+      "epoch": 7.698650068709077,
+      "grad_norm": 1.1134333610534668,
+      "learning_rate": 0.0002,
+      "loss": 0.452,
+      "step": 47620
+    },
+    {
+      "epoch": 7.7002667528898225,
+      "grad_norm": 1.2643009424209595,
+      "learning_rate": 0.0002,
+      "loss": 0.4859,
+      "step": 47630
+    },
+    {
+      "epoch": 7.701883437070569,
+      "grad_norm": 1.1145045757293701,
+      "learning_rate": 0.0002,
+      "loss": 0.4825,
+      "step": 47640
+    },
+    {
+      "epoch": 7.703500121251314,
+      "grad_norm": 1.1808549165725708,
+      "learning_rate": 0.0002,
+      "loss": 0.4735,
+      "step": 47650
+    },
+    {
+      "epoch": 7.705116805432059,
+      "grad_norm": 1.2996630668640137,
+      "learning_rate": 0.0002,
+      "loss": 0.4841,
+      "step": 47660
+    },
+    {
+      "epoch": 7.706733489612804,
+      "grad_norm": 1.2786413431167603,
+      "learning_rate": 0.0002,
+      "loss": 0.4712,
+      "step": 47670
+    },
+    {
+      "epoch": 7.70835017379355,
+      "grad_norm": 1.3245121240615845,
+      "learning_rate": 0.0002,
+      "loss": 0.4694,
+      "step": 47680
+    },
+    {
+      "epoch": 7.709966857974295,
+      "grad_norm": 1.4168202877044678,
+      "learning_rate": 0.0002,
+      "loss": 0.4467,
+      "step": 47690
+    },
+    {
+      "epoch": 7.71158354215504,
+      "grad_norm": 1.0354000329971313,
+      "learning_rate": 0.0002,
+      "loss": 0.5143,
+      "step": 47700
+    },
+    {
+      "epoch": 7.713200226335785,
+      "grad_norm": 0.9630362391471863,
+      "learning_rate": 0.0002,
+      "loss": 0.4703,
+      "step": 47710
+    },
+    {
+      "epoch": 7.7148169105165305,
+      "grad_norm": 1.1045806407928467,
+      "learning_rate": 0.0002,
+      "loss": 0.4996,
+      "step": 47720
+    },
+    {
+      "epoch": 7.716433594697276,
+      "grad_norm": 1.2403767108917236,
+      "learning_rate": 0.0002,
+      "loss": 0.4756,
+      "step": 47730
+    },
+    {
+      "epoch": 7.718050278878021,
+      "grad_norm": 0.9893410801887512,
+      "learning_rate": 0.0002,
+      "loss": 0.4658,
+      "step": 47740
+    },
+    {
+      "epoch": 7.719666963058766,
+      "grad_norm": 1.0749315023422241,
+      "learning_rate": 0.0002,
+      "loss": 0.4463,
+      "step": 47750
+    },
+    {
+      "epoch": 7.721283647239511,
+      "grad_norm": 1.2851510047912598,
+      "learning_rate": 0.0002,
+      "loss": 0.467,
+      "step": 47760
+    },
+    {
+      "epoch": 7.722900331420257,
+      "grad_norm": 1.2964261770248413,
+      "learning_rate": 0.0002,
+      "loss": 0.489,
+      "step": 47770
+    },
+    {
+      "epoch": 7.724517015601002,
+      "grad_norm": 1.0603861808776855,
+      "learning_rate": 0.0002,
+      "loss": 0.4702,
+      "step": 47780
+    },
+    {
+      "epoch": 7.726133699781748,
+      "grad_norm": 1.2728440761566162,
+      "learning_rate": 0.0002,
+      "loss": 0.478,
+      "step": 47790
+    },
+    {
+      "epoch": 7.727750383962493,
+      "grad_norm": 1.20509934425354,
+      "learning_rate": 0.0002,
+      "loss": 0.4746,
+      "step": 47800
+    },
+    {
+      "epoch": 7.7293670681432385,
+      "grad_norm": 1.397595763206482,
+      "learning_rate": 0.0002,
+      "loss": 0.4556,
+      "step": 47810
+    },
+    {
+      "epoch": 7.730983752323984,
+      "grad_norm": 1.2595560550689697,
+      "learning_rate": 0.0002,
+      "loss": 0.4736,
+      "step": 47820
+    },
+    {
+      "epoch": 7.732600436504729,
+      "grad_norm": 1.166074514389038,
+      "learning_rate": 0.0002,
+      "loss": 0.5061,
+      "step": 47830
+    },
+    {
+      "epoch": 7.734217120685474,
+      "grad_norm": 1.258192777633667,
+      "learning_rate": 0.0002,
+      "loss": 0.4907,
+      "step": 47840
+    },
+    {
+      "epoch": 7.735833804866219,
+      "grad_norm": 1.0394890308380127,
+      "learning_rate": 0.0002,
+      "loss": 0.5256,
+      "step": 47850
+    },
+    {
+      "epoch": 7.737450489046965,
+      "grad_norm": 1.2017768621444702,
+      "learning_rate": 0.0002,
+      "loss": 0.4863,
+      "step": 47860
+    },
+    {
+      "epoch": 7.73906717322771,
+      "grad_norm": 1.1070265769958496,
+      "learning_rate": 0.0002,
+      "loss": 0.4784,
+      "step": 47870
+    },
+    {
+      "epoch": 7.740683857408455,
+      "grad_norm": 1.0544345378875732,
+      "learning_rate": 0.0002,
+      "loss": 0.4616,
+      "step": 47880
+    },
+    {
+      "epoch": 7.7423005415892,
+      "grad_norm": 1.0194088220596313,
+      "learning_rate": 0.0002,
+      "loss": 0.4519,
+      "step": 47890
+    },
+    {
+      "epoch": 7.743917225769946,
+      "grad_norm": 1.3095234632492065,
+      "learning_rate": 0.0002,
+      "loss": 0.4758,
+      "step": 47900
+    },
+    {
+      "epoch": 7.745533909950691,
+      "grad_norm": 1.0579626560211182,
+      "learning_rate": 0.0002,
+      "loss": 0.4646,
+      "step": 47910
+    },
+    {
+      "epoch": 7.747150594131437,
+      "grad_norm": 1.012990951538086,
+      "learning_rate": 0.0002,
+      "loss": 0.4532,
+      "step": 47920
+    },
+    {
+      "epoch": 7.748767278312181,
+      "grad_norm": 1.485148549079895,
+      "learning_rate": 0.0002,
+      "loss": 0.4775,
+      "step": 47930
+    },
+    {
+      "epoch": 7.750383962492927,
+      "grad_norm": 1.3595696687698364,
+      "learning_rate": 0.0002,
+      "loss": 0.4892,
+      "step": 47940
+    },
+    {
+      "epoch": 7.752000646673673,
+      "grad_norm": 0.9945753216743469,
+      "learning_rate": 0.0002,
+      "loss": 0.4609,
+      "step": 47950
+    },
+    {
+      "epoch": 7.753617330854418,
+      "grad_norm": 1.2098956108093262,
+      "learning_rate": 0.0002,
+      "loss": 0.5138,
+      "step": 47960
+    },
+    {
+      "epoch": 7.755234015035163,
+      "grad_norm": 1.3056198358535767,
+      "learning_rate": 0.0002,
+      "loss": 0.4815,
+      "step": 47970
+    },
+    {
+      "epoch": 7.756850699215908,
+      "grad_norm": 1.2247772216796875,
+      "learning_rate": 0.0002,
+      "loss": 0.4761,
+      "step": 47980
+    },
+    {
+      "epoch": 7.7584673833966535,
+      "grad_norm": 1.397642970085144,
+      "learning_rate": 0.0002,
+      "loss": 0.5023,
+      "step": 47990
+    },
+    {
+      "epoch": 7.760084067577399,
+      "grad_norm": 1.2565888166427612,
+      "learning_rate": 0.0002,
+      "loss": 0.4901,
+      "step": 48000
+    },
+    {
+      "epoch": 7.761700751758144,
+      "grad_norm": 1.0065099000930786,
+      "learning_rate": 0.0002,
+      "loss": 0.469,
+      "step": 48010
+    },
+    {
+      "epoch": 7.763317435938889,
+      "grad_norm": 1.1466305255889893,
+      "learning_rate": 0.0002,
+      "loss": 0.4886,
+      "step": 48020
+    },
+    {
+      "epoch": 7.7649341201196345,
+      "grad_norm": 1.4492419958114624,
+      "learning_rate": 0.0002,
+      "loss": 0.4898,
+      "step": 48030
+    },
+    {
+      "epoch": 7.76655080430038,
+      "grad_norm": 1.0945932865142822,
+      "learning_rate": 0.0002,
+      "loss": 0.489,
+      "step": 48040
+    },
+    {
+      "epoch": 7.768167488481125,
+      "grad_norm": 1.1938602924346924,
+      "learning_rate": 0.0002,
+      "loss": 0.4968,
+      "step": 48050
+    },
+    {
+      "epoch": 7.76978417266187,
+      "grad_norm": 1.168890357017517,
+      "learning_rate": 0.0002,
+      "loss": 0.4497,
+      "step": 48060
+    },
+    {
+      "epoch": 7.771400856842616,
+      "grad_norm": 1.3134305477142334,
+      "learning_rate": 0.0002,
+      "loss": 0.4881,
+      "step": 48070
+    },
+    {
+      "epoch": 7.773017541023361,
+      "grad_norm": 1.044438123703003,
+      "learning_rate": 0.0002,
+      "loss": 0.5226,
+      "step": 48080
+    },
+    {
+      "epoch": 7.774634225204107,
+      "grad_norm": 1.1275628805160522,
+      "learning_rate": 0.0002,
+      "loss": 0.4497,
+      "step": 48090
+    },
+    {
+      "epoch": 7.776250909384852,
+      "grad_norm": 1.0877318382263184,
+      "learning_rate": 0.0002,
+      "loss": 0.5063,
+      "step": 48100
+    },
+    {
+      "epoch": 7.777867593565597,
+      "grad_norm": 1.4800893068313599,
+      "learning_rate": 0.0002,
+      "loss": 0.4795,
+      "step": 48110
+    },
+    {
+      "epoch": 7.779484277746342,
+      "grad_norm": 1.1495977640151978,
+      "learning_rate": 0.0002,
+      "loss": 0.4984,
+      "step": 48120
+    },
+    {
+      "epoch": 7.781100961927088,
+      "grad_norm": 1.2175556421279907,
+      "learning_rate": 0.0002,
+      "loss": 0.4662,
+      "step": 48130
+    },
+    {
+      "epoch": 7.782717646107833,
+      "grad_norm": 1.150556206703186,
+      "learning_rate": 0.0002,
+      "loss": 0.4935,
+      "step": 48140
+    },
+    {
+      "epoch": 7.784334330288578,
+      "grad_norm": 1.051145315170288,
+      "learning_rate": 0.0002,
+      "loss": 0.5039,
+      "step": 48150
+    },
+    {
+      "epoch": 7.785951014469323,
+      "grad_norm": 1.2842742204666138,
+      "learning_rate": 0.0002,
+      "loss": 0.4611,
+      "step": 48160
+    },
+    {
+      "epoch": 7.787567698650069,
+      "grad_norm": 1.2251030206680298,
+      "learning_rate": 0.0002,
+      "loss": 0.4589,
+      "step": 48170
+    },
+    {
+      "epoch": 7.789184382830814,
+      "grad_norm": 1.2809321880340576,
+      "learning_rate": 0.0002,
+      "loss": 0.4905,
+      "step": 48180
+    },
+    {
+      "epoch": 7.790801067011559,
+      "grad_norm": 1.005690336227417,
+      "learning_rate": 0.0002,
+      "loss": 0.4569,
+      "step": 48190
+    },
+    {
+      "epoch": 7.792417751192304,
+      "grad_norm": 1.325501561164856,
+      "learning_rate": 0.0002,
+      "loss": 0.4862,
+      "step": 48200
+    },
+    {
+      "epoch": 7.7940344353730495,
+      "grad_norm": 1.4551857709884644,
+      "learning_rate": 0.0002,
+      "loss": 0.4384,
+      "step": 48210
+    },
+    {
+      "epoch": 7.795651119553796,
+      "grad_norm": 1.3399626016616821,
+      "learning_rate": 0.0002,
+      "loss": 0.4696,
+      "step": 48220
+    },
+    {
+      "epoch": 7.79726780373454,
+      "grad_norm": 1.0379714965820312,
+      "learning_rate": 0.0002,
+      "loss": 0.4654,
+      "step": 48230
+    },
+    {
+      "epoch": 7.798884487915286,
+      "grad_norm": 0.9725802540779114,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 48240
+    },
+    {
+      "epoch": 7.800501172096031,
+      "grad_norm": 1.0202224254608154,
+      "learning_rate": 0.0002,
+      "loss": 0.4583,
+      "step": 48250
+    },
+    {
+      "epoch": 7.802117856276777,
+      "grad_norm": 0.9477742910385132,
+      "learning_rate": 0.0002,
+      "loss": 0.4792,
+      "step": 48260
+    },
+    {
+      "epoch": 7.803734540457522,
+      "grad_norm": 1.2726924419403076,
+      "learning_rate": 0.0002,
+      "loss": 0.4836,
+      "step": 48270
+    },
+    {
+      "epoch": 7.805351224638267,
+      "grad_norm": 1.453190565109253,
+      "learning_rate": 0.0002,
+      "loss": 0.494,
+      "step": 48280
+    },
+    {
+      "epoch": 7.806967908819012,
+      "grad_norm": 1.2806978225708008,
+      "learning_rate": 0.0002,
+      "loss": 0.4559,
+      "step": 48290
+    },
+    {
+      "epoch": 7.8085845929997575,
+      "grad_norm": 1.0897129774093628,
+      "learning_rate": 0.0002,
+      "loss": 0.4867,
+      "step": 48300
+    },
+    {
+      "epoch": 7.810201277180503,
+      "grad_norm": 1.381636381149292,
+      "learning_rate": 0.0002,
+      "loss": 0.4939,
+      "step": 48310
+    },
+    {
+      "epoch": 7.811817961361248,
+      "grad_norm": 0.9954851269721985,
+      "learning_rate": 0.0002,
+      "loss": 0.4797,
+      "step": 48320
+    },
+    {
+      "epoch": 7.813434645541993,
+      "grad_norm": 1.1756198406219482,
+      "learning_rate": 0.0002,
+      "loss": 0.4995,
+      "step": 48330
+    },
+    {
+      "epoch": 7.815051329722738,
+      "grad_norm": 1.2087817192077637,
+      "learning_rate": 0.0002,
+      "loss": 0.4904,
+      "step": 48340
+    },
+    {
+      "epoch": 7.816668013903484,
+      "grad_norm": 1.3075505495071411,
+      "learning_rate": 0.0002,
+      "loss": 0.4935,
+      "step": 48350
+    },
+    {
+      "epoch": 7.818284698084229,
+      "grad_norm": 1.1872076988220215,
+      "learning_rate": 0.0002,
+      "loss": 0.486,
+      "step": 48360
+    },
+    {
+      "epoch": 7.819901382264975,
+      "grad_norm": 1.2134783267974854,
+      "learning_rate": 0.0002,
+      "loss": 0.538,
+      "step": 48370
+    },
+    {
+      "epoch": 7.821518066445719,
+      "grad_norm": 1.28566312789917,
+      "learning_rate": 0.0002,
+      "loss": 0.4759,
+      "step": 48380
+    },
+    {
+      "epoch": 7.8231347506264655,
+      "grad_norm": 1.0578798055648804,
+      "learning_rate": 0.0002,
+      "loss": 0.4962,
+      "step": 48390
+    },
+    {
+      "epoch": 7.824751434807211,
+      "grad_norm": 1.1225441694259644,
+      "learning_rate": 0.0002,
+      "loss": 0.4924,
+      "step": 48400
+    },
+    {
+      "epoch": 7.826368118987956,
+      "grad_norm": 1.2029428482055664,
+      "learning_rate": 0.0002,
+      "loss": 0.5081,
+      "step": 48410
+    },
+    {
+      "epoch": 7.827984803168701,
+      "grad_norm": 1.252485990524292,
+      "learning_rate": 0.0002,
+      "loss": 0.4669,
+      "step": 48420
+    },
+    {
+      "epoch": 7.829601487349446,
+      "grad_norm": 1.1822574138641357,
+      "learning_rate": 0.0002,
+      "loss": 0.4932,
+      "step": 48430
+    },
+    {
+      "epoch": 7.831218171530192,
+      "grad_norm": 1.2428245544433594,
+      "learning_rate": 0.0002,
+      "loss": 0.4692,
+      "step": 48440
+    },
+    {
+      "epoch": 7.832834855710937,
+      "grad_norm": 1.0565894842147827,
+      "learning_rate": 0.0002,
+      "loss": 0.4796,
+      "step": 48450
+    },
+    {
+      "epoch": 7.834451539891682,
+      "grad_norm": 1.363452672958374,
+      "learning_rate": 0.0002,
+      "loss": 0.5016,
+      "step": 48460
+    },
+    {
+      "epoch": 7.836068224072427,
+      "grad_norm": 1.2436026334762573,
+      "learning_rate": 0.0002,
+      "loss": 0.463,
+      "step": 48470
+    },
+    {
+      "epoch": 7.8376849082531725,
+      "grad_norm": 1.2623029947280884,
+      "learning_rate": 0.0002,
+      "loss": 0.4794,
+      "step": 48480
+    },
+    {
+      "epoch": 7.839301592433918,
+      "grad_norm": 1.0942288637161255,
+      "learning_rate": 0.0002,
+      "loss": 0.4863,
+      "step": 48490
+    },
+    {
+      "epoch": 7.840918276614663,
+      "grad_norm": 1.1791462898254395,
+      "learning_rate": 0.0002,
+      "loss": 0.4723,
+      "step": 48500
+    },
+    {
+      "epoch": 7.842534960795408,
+      "grad_norm": 1.3342814445495605,
+      "learning_rate": 0.0002,
+      "loss": 0.501,
+      "step": 48510
+    },
+    {
+      "epoch": 7.844151644976154,
+      "grad_norm": 1.0511828660964966,
+      "learning_rate": 0.0002,
+      "loss": 0.4654,
+      "step": 48520
+    },
+    {
+      "epoch": 7.845768329156899,
+      "grad_norm": 1.48568594455719,
+      "learning_rate": 0.0002,
+      "loss": 0.5051,
+      "step": 48530
+    },
+    {
+      "epoch": 7.847385013337645,
+      "grad_norm": 1.296844720840454,
+      "learning_rate": 0.0002,
+      "loss": 0.4572,
+      "step": 48540
+    },
+    {
+      "epoch": 7.84900169751839,
+      "grad_norm": 1.3032835721969604,
+      "learning_rate": 0.0002,
+      "loss": 0.5216,
+      "step": 48550
+    },
+    {
+      "epoch": 7.850618381699135,
+      "grad_norm": 1.260769009590149,
+      "learning_rate": 0.0002,
+      "loss": 0.472,
+      "step": 48560
+    },
+    {
+      "epoch": 7.8522350658798805,
+      "grad_norm": 1.3309531211853027,
+      "learning_rate": 0.0002,
+      "loss": 0.5094,
+      "step": 48570
+    },
+    {
+      "epoch": 7.853851750060626,
+      "grad_norm": 1.1907469034194946,
+      "learning_rate": 0.0002,
+      "loss": 0.4759,
+      "step": 48580
+    },
+    {
+      "epoch": 7.855468434241371,
+      "grad_norm": 0.9690865874290466,
+      "learning_rate": 0.0002,
+      "loss": 0.485,
+      "step": 48590
+    },
+    {
+      "epoch": 7.857085118422116,
+      "grad_norm": 1.2417343854904175,
+      "learning_rate": 0.0002,
+      "loss": 0.4899,
+      "step": 48600
+    },
+    {
+      "epoch": 7.858701802602861,
+      "grad_norm": 1.1366082429885864,
+      "learning_rate": 0.0002,
+      "loss": 0.4888,
+      "step": 48610
+    },
+    {
+      "epoch": 7.860318486783607,
+      "grad_norm": 1.4737876653671265,
+      "learning_rate": 0.0002,
+      "loss": 0.4909,
+      "step": 48620
+    },
+    {
+      "epoch": 7.861935170964352,
+      "grad_norm": 1.3934144973754883,
+      "learning_rate": 0.0002,
+      "loss": 0.4476,
+      "step": 48630
+    },
+    {
+      "epoch": 7.863551855145097,
+      "grad_norm": 0.9997506737709045,
+      "learning_rate": 0.0002,
+      "loss": 0.5141,
+      "step": 48640
+    },
+    {
+      "epoch": 7.865168539325842,
+      "grad_norm": 1.3827011585235596,
+      "learning_rate": 0.0002,
+      "loss": 0.4766,
+      "step": 48650
+    },
+    {
+      "epoch": 7.866785223506588,
+      "grad_norm": 1.2811808586120605,
+      "learning_rate": 0.0002,
+      "loss": 0.4926,
+      "step": 48660
+    },
+    {
+      "epoch": 7.868401907687334,
+      "grad_norm": 1.394400715827942,
+      "learning_rate": 0.0002,
+      "loss": 0.4898,
+      "step": 48670
+    },
+    {
+      "epoch": 7.870018591868079,
+      "grad_norm": 1.5635628700256348,
+      "learning_rate": 0.0002,
+      "loss": 0.4889,
+      "step": 48680
+    },
+    {
+      "epoch": 7.871635276048824,
+      "grad_norm": 1.147349238395691,
+      "learning_rate": 0.0002,
+      "loss": 0.4822,
+      "step": 48690
+    },
+    {
+      "epoch": 7.873251960229569,
+      "grad_norm": 1.2417502403259277,
+      "learning_rate": 0.0002,
+      "loss": 0.485,
+      "step": 48700
+    },
+    {
+      "epoch": 7.874868644410315,
+      "grad_norm": 1.0380291938781738,
+      "learning_rate": 0.0002,
+      "loss": 0.4976,
+      "step": 48710
+    },
+    {
+      "epoch": 7.87648532859106,
+      "grad_norm": 1.2139482498168945,
+      "learning_rate": 0.0002,
+      "loss": 0.5119,
+      "step": 48720
+    },
+    {
+      "epoch": 7.878102012771805,
+      "grad_norm": 1.2833739519119263,
+      "learning_rate": 0.0002,
+      "loss": 0.497,
+      "step": 48730
+    },
+    {
+      "epoch": 7.87971869695255,
+      "grad_norm": 1.2405574321746826,
+      "learning_rate": 0.0002,
+      "loss": 0.4776,
+      "step": 48740
+    },
+    {
+      "epoch": 7.881335381133296,
+      "grad_norm": 1.1267465353012085,
+      "learning_rate": 0.0002,
+      "loss": 0.5085,
+      "step": 48750
+    },
+    {
+      "epoch": 7.882952065314041,
+      "grad_norm": 1.3052713871002197,
+      "learning_rate": 0.0002,
+      "loss": 0.4894,
+      "step": 48760
+    },
+    {
+      "epoch": 7.884568749494786,
+      "grad_norm": 1.0581550598144531,
+      "learning_rate": 0.0002,
+      "loss": 0.5077,
+      "step": 48770
+    },
+    {
+      "epoch": 7.886185433675531,
+      "grad_norm": 1.1074683666229248,
+      "learning_rate": 0.0002,
+      "loss": 0.4598,
+      "step": 48780
+    },
+    {
+      "epoch": 7.8878021178562765,
+      "grad_norm": 1.0812418460845947,
+      "learning_rate": 0.0002,
+      "loss": 0.4813,
+      "step": 48790
+    },
+    {
+      "epoch": 7.889418802037022,
+      "grad_norm": 1.3083902597427368,
+      "learning_rate": 0.0002,
+      "loss": 0.477,
+      "step": 48800
+    },
+    {
+      "epoch": 7.891035486217767,
+      "grad_norm": 1.457373023033142,
+      "learning_rate": 0.0002,
+      "loss": 0.4717,
+      "step": 48810
+    },
+    {
+      "epoch": 7.892652170398513,
+      "grad_norm": 1.048091173171997,
+      "learning_rate": 0.0002,
+      "loss": 0.4607,
+      "step": 48820
+    },
+    {
+      "epoch": 7.894268854579258,
+      "grad_norm": 1.1420985460281372,
+      "learning_rate": 0.0002,
+      "loss": 0.4848,
+      "step": 48830
+    },
+    {
+      "epoch": 7.8958855387600035,
+      "grad_norm": 1.0286061763763428,
+      "learning_rate": 0.0002,
+      "loss": 0.5019,
+      "step": 48840
+    },
+    {
+      "epoch": 7.897502222940749,
+      "grad_norm": 1.0361840724945068,
+      "learning_rate": 0.0002,
+      "loss": 0.5054,
+      "step": 48850
+    },
+    {
+      "epoch": 7.899118907121494,
+      "grad_norm": 1.1862726211547852,
+      "learning_rate": 0.0002,
+      "loss": 0.4908,
+      "step": 48860
+    },
+    {
+      "epoch": 7.900735591302239,
+      "grad_norm": 1.2256416082382202,
+      "learning_rate": 0.0002,
+      "loss": 0.4491,
+      "step": 48870
+    },
+    {
+      "epoch": 7.9023522754829845,
+      "grad_norm": 1.0664557218551636,
+      "learning_rate": 0.0002,
+      "loss": 0.4732,
+      "step": 48880
+    },
+    {
+      "epoch": 7.90396895966373,
+      "grad_norm": 1.3960802555084229,
+      "learning_rate": 0.0002,
+      "loss": 0.4741,
+      "step": 48890
+    },
+    {
+      "epoch": 7.905585643844475,
+      "grad_norm": 1.230430245399475,
+      "learning_rate": 0.0002,
+      "loss": 0.5061,
+      "step": 48900
+    },
+    {
+      "epoch": 7.90720232802522,
+      "grad_norm": 1.0949305295944214,
+      "learning_rate": 0.0002,
+      "loss": 0.4698,
+      "step": 48910
+    },
+    {
+      "epoch": 7.908819012205965,
+      "grad_norm": 1.4402074813842773,
+      "learning_rate": 0.0002,
+      "loss": 0.4964,
+      "step": 48920
+    },
+    {
+      "epoch": 7.910435696386711,
+      "grad_norm": 1.1064879894256592,
+      "learning_rate": 0.0002,
+      "loss": 0.5057,
+      "step": 48930
+    },
+    {
+      "epoch": 7.912052380567456,
+      "grad_norm": 0.9874461889266968,
+      "learning_rate": 0.0002,
+      "loss": 0.482,
+      "step": 48940
+    },
+    {
+      "epoch": 7.913669064748201,
+      "grad_norm": 1.2584952116012573,
+      "learning_rate": 0.0002,
+      "loss": 0.4851,
+      "step": 48950
+    },
+    {
+      "epoch": 7.915285748928946,
+      "grad_norm": 1.3016353845596313,
+      "learning_rate": 0.0002,
+      "loss": 0.4744,
+      "step": 48960
+    },
+    {
+      "epoch": 7.916902433109692,
+      "grad_norm": 1.104179859161377,
+      "learning_rate": 0.0002,
+      "loss": 0.4734,
+      "step": 48970
+    },
+    {
+      "epoch": 7.918519117290438,
+      "grad_norm": 1.26803457736969,
+      "learning_rate": 0.0002,
+      "loss": 0.5143,
+      "step": 48980
+    },
+    {
+      "epoch": 7.920135801471183,
+      "grad_norm": 1.0336869955062866,
+      "learning_rate": 0.0002,
+      "loss": 0.4808,
+      "step": 48990
+    },
+    {
+      "epoch": 7.921752485651928,
+      "grad_norm": 1.0630918741226196,
+      "learning_rate": 0.0002,
+      "loss": 0.4938,
+      "step": 49000
+    },
+    {
+      "epoch": 7.923369169832673,
+      "grad_norm": 1.2257622480392456,
+      "learning_rate": 0.0002,
+      "loss": 0.4988,
+      "step": 49010
+    },
+    {
+      "epoch": 7.924985854013419,
+      "grad_norm": 1.1722705364227295,
+      "learning_rate": 0.0002,
+      "loss": 0.5116,
+      "step": 49020
+    },
+    {
+      "epoch": 7.926602538194164,
+      "grad_norm": 1.4473323822021484,
+      "learning_rate": 0.0002,
+      "loss": 0.4737,
+      "step": 49030
+    },
+    {
+      "epoch": 7.928219222374909,
+      "grad_norm": 1.3780192136764526,
+      "learning_rate": 0.0002,
+      "loss": 0.5412,
+      "step": 49040
+    },
+    {
+      "epoch": 7.929835906555654,
+      "grad_norm": 1.253423810005188,
+      "learning_rate": 0.0002,
+      "loss": 0.4849,
+      "step": 49050
+    },
+    {
+      "epoch": 7.9314525907363995,
+      "grad_norm": 1.1733828783035278,
+      "learning_rate": 0.0002,
+      "loss": 0.5103,
+      "step": 49060
+    },
+    {
+      "epoch": 7.933069274917145,
+      "grad_norm": 1.249990701675415,
+      "learning_rate": 0.0002,
+      "loss": 0.4967,
+      "step": 49070
+    },
+    {
+      "epoch": 7.93468595909789,
+      "grad_norm": 1.4012458324432373,
+      "learning_rate": 0.0002,
+      "loss": 0.51,
+      "step": 49080
+    },
+    {
+      "epoch": 7.936302643278635,
+      "grad_norm": 1.268652319908142,
+      "learning_rate": 0.0002,
+      "loss": 0.4785,
+      "step": 49090
+    },
+    {
+      "epoch": 7.9379193274593804,
+      "grad_norm": 1.0469073057174683,
+      "learning_rate": 0.0002,
+      "loss": 0.5319,
+      "step": 49100
+    },
+    {
+      "epoch": 7.939536011640126,
+      "grad_norm": 1.3028813600540161,
+      "learning_rate": 0.0002,
+      "loss": 0.4836,
+      "step": 49110
+    },
+    {
+      "epoch": 7.941152695820872,
+      "grad_norm": 1.0998128652572632,
+      "learning_rate": 0.0002,
+      "loss": 0.4791,
+      "step": 49120
+    },
+    {
+      "epoch": 7.942769380001617,
+      "grad_norm": 1.300884485244751,
+      "learning_rate": 0.0002,
+      "loss": 0.5022,
+      "step": 49130
+    },
+    {
+      "epoch": 7.944386064182362,
+      "grad_norm": 1.257865071296692,
+      "learning_rate": 0.0002,
+      "loss": 0.5193,
+      "step": 49140
+    },
+    {
+      "epoch": 7.9460027483631075,
+      "grad_norm": 1.074731707572937,
+      "learning_rate": 0.0002,
+      "loss": 0.4755,
+      "step": 49150
+    },
+    {
+      "epoch": 7.947619432543853,
+      "grad_norm": 1.1055876016616821,
+      "learning_rate": 0.0002,
+      "loss": 0.4675,
+      "step": 49160
+    },
+    {
+      "epoch": 7.949236116724598,
+      "grad_norm": 1.1986541748046875,
+      "learning_rate": 0.0002,
+      "loss": 0.4801,
+      "step": 49170
+    },
+    {
+      "epoch": 7.950852800905343,
+      "grad_norm": 1.094555139541626,
+      "learning_rate": 0.0002,
+      "loss": 0.488,
+      "step": 49180
+    },
+    {
+      "epoch": 7.952469485086088,
+      "grad_norm": 1.2922005653381348,
+      "learning_rate": 0.0002,
+      "loss": 0.4974,
+      "step": 49190
+    },
+    {
+      "epoch": 7.954086169266834,
+      "grad_norm": 1.1557104587554932,
+      "learning_rate": 0.0002,
+      "loss": 0.4973,
+      "step": 49200
+    },
+    {
+      "epoch": 7.955702853447579,
+      "grad_norm": 1.2414908409118652,
+      "learning_rate": 0.0002,
+      "loss": 0.4806,
+      "step": 49210
+    },
+    {
+      "epoch": 7.957319537628324,
+      "grad_norm": 1.3606830835342407,
+      "learning_rate": 0.0002,
+      "loss": 0.4848,
+      "step": 49220
+    },
+    {
+      "epoch": 7.958936221809069,
+      "grad_norm": 0.9592481851577759,
+      "learning_rate": 0.0002,
+      "loss": 0.4981,
+      "step": 49230
+    },
+    {
+      "epoch": 7.960552905989815,
+      "grad_norm": 1.2130779027938843,
+      "learning_rate": 0.0002,
+      "loss": 0.4731,
+      "step": 49240
+    },
+    {
+      "epoch": 7.96216959017056,
+      "grad_norm": 1.1078767776489258,
+      "learning_rate": 0.0002,
+      "loss": 0.4529,
+      "step": 49250
+    },
+    {
+      "epoch": 7.963786274351305,
+      "grad_norm": 1.0684230327606201,
+      "learning_rate": 0.0002,
+      "loss": 0.4983,
+      "step": 49260
+    },
+    {
+      "epoch": 7.965402958532051,
+      "grad_norm": 1.1368396282196045,
+      "learning_rate": 0.0002,
+      "loss": 0.4832,
+      "step": 49270
+    },
+    {
+      "epoch": 7.967019642712796,
+      "grad_norm": 1.2161095142364502,
+      "learning_rate": 0.0002,
+      "loss": 0.5226,
+      "step": 49280
+    },
+    {
+      "epoch": 7.968636326893542,
+      "grad_norm": 1.2087634801864624,
+      "learning_rate": 0.0002,
+      "loss": 0.4938,
+      "step": 49290
+    },
+    {
+      "epoch": 7.970253011074287,
+      "grad_norm": 1.1078447103500366,
+      "learning_rate": 0.0002,
+      "loss": 0.4969,
+      "step": 49300
+    },
+    {
+      "epoch": 7.971869695255032,
+      "grad_norm": 1.3378221988677979,
+      "learning_rate": 0.0002,
+      "loss": 0.5333,
+      "step": 49310
+    },
+    {
+      "epoch": 7.973486379435777,
+      "grad_norm": 1.0475801229476929,
+      "learning_rate": 0.0002,
+      "loss": 0.4736,
+      "step": 49320
+    },
+    {
+      "epoch": 7.9751030636165225,
+      "grad_norm": 0.9948194622993469,
+      "learning_rate": 0.0002,
+      "loss": 0.4515,
+      "step": 49330
+    },
+    {
+      "epoch": 7.976719747797268,
+      "grad_norm": 1.06312894821167,
+      "learning_rate": 0.0002,
+      "loss": 0.4685,
+      "step": 49340
+    },
+    {
+      "epoch": 7.978336431978013,
+      "grad_norm": 1.4047085046768188,
+      "learning_rate": 0.0002,
+      "loss": 0.453,
+      "step": 49350
+    },
+    {
+      "epoch": 7.979953116158758,
+      "grad_norm": 1.086578130722046,
+      "learning_rate": 0.0002,
+      "loss": 0.5054,
+      "step": 49360
+    },
+    {
+      "epoch": 7.9815698003395035,
+      "grad_norm": 1.2896746397018433,
+      "learning_rate": 0.0002,
+      "loss": 0.5024,
+      "step": 49370
+    },
+    {
+      "epoch": 7.983186484520249,
+      "grad_norm": 1.260717511177063,
+      "learning_rate": 0.0002,
+      "loss": 0.5102,
+      "step": 49380
+    },
+    {
+      "epoch": 7.984803168700994,
+      "grad_norm": 1.4238426685333252,
+      "learning_rate": 0.0002,
+      "loss": 0.4836,
+      "step": 49390
+    },
+    {
+      "epoch": 7.986419852881739,
+      "grad_norm": 1.1800259351730347,
+      "learning_rate": 0.0002,
+      "loss": 0.4797,
+      "step": 49400
+    },
+    {
+      "epoch": 7.988036537062484,
+      "grad_norm": 1.128868579864502,
+      "learning_rate": 0.0002,
+      "loss": 0.4911,
+      "step": 49410
+    },
+    {
+      "epoch": 7.9896532212432305,
+      "grad_norm": 1.1832106113433838,
+      "learning_rate": 0.0002,
+      "loss": 0.4674,
+      "step": 49420
+    },
+    {
+      "epoch": 7.991269905423976,
+      "grad_norm": 1.1728334426879883,
+      "learning_rate": 0.0002,
+      "loss": 0.5016,
+      "step": 49430
+    },
+    {
+      "epoch": 7.992886589604721,
+      "grad_norm": 1.2403929233551025,
+      "learning_rate": 0.0002,
+      "loss": 0.4706,
+      "step": 49440
+    },
+    {
+      "epoch": 7.994503273785466,
+      "grad_norm": 1.245354175567627,
+      "learning_rate": 0.0002,
+      "loss": 0.4989,
+      "step": 49450
+    },
+    {
+      "epoch": 7.9961199579662114,
+      "grad_norm": 1.3526462316513062,
+      "learning_rate": 0.0002,
+      "loss": 0.5088,
+      "step": 49460
+    },
+    {
+      "epoch": 7.997736642146957,
+      "grad_norm": 1.2117315530776978,
+      "learning_rate": 0.0002,
+      "loss": 0.4703,
+      "step": 49470
+    },
+    {
+      "epoch": 7.999353326327702,
+      "grad_norm": 1.0393620729446411,
+      "learning_rate": 0.0002,
+      "loss": 0.4802,
+      "step": 49480
+    },
+    {
+      "epoch": 7.999353326327702,
+      "eval_loss": 1.292362093925476,
+      "eval_runtime": 122.1709,
+      "eval_samples_per_second": 6.0,
+      "eval_steps_per_second": 0.753,
+      "step": 49480
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 49480,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.2898236097927578e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/training_args.bin b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..155b12fa9acbc6e71dba75c92bfa79e152397ebf
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-49480/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28694d5564a2b5c7d6881d4ba2af103356aa22489d2c22768ebbe47283c0f4a1
+size 5560
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/README.md b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..835e31ab8469ee39ddc8b2b6b2143a8c66dad510
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/README.md
@@ -0,0 +1,202 @@
+---
+base_model: meta-llama/Meta-Llama-3-8B-Instruct
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.13.1
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/adapter_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..406c5a08dc4a2a33b52c62a482f98c217c417215
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/adapter_config.json
@@ -0,0 +1,29 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/adapter_model.safetensors b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..894306c21591f0f9fc1ad6edef9d7f95864e6e88
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6e6524f0f0d768c7962609d66799cf291b435521c1246c8b6c63ea99e47dce1
+size 109069176
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/optimizer.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b063d727e44d7b9a7a80ed77f4aacaf940f61605
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3898ae941d376d381d8f74c20011586558919a6034d4d4f1516c378d88476c35
+size 55532666
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/rng_state.pth b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4499ab3321a9b904d4df9aa714313a7423559597
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63e86de25b60f1f051d0aad81f1d4dae37811756184027f20eeb1d49d57860e4
+size 14244
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/scheduler.pt b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8ee6ca37eba996eef47c9431c9823aaa6ddcc0c6
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43eacc2532dd42a4d4cd9c082df2820848486282537958f00be686f4de55e6ba
+size 1064
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/special_tokens_map.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/special_tokens_map.json
@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/tokenizer.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..86a33946b0c77216d2cce91bb28c8fada4a5e80b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/tokenizer_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..061e40d9db3253624f86e8e364c15ef546527c9d
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/tokenizer_config.json
@@ -0,0 +1,2063 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/trainer_state.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..e5c06890b605df5b7d9a602659fcf0147712e156
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/trainer_state.json
@@ -0,0 +1,4367 @@
+{
+  "best_metric": 1.0871200561523438,
+  "best_model_checkpoint": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185",
+  "epoch": 0.9999191657909627,
+  "eval_steps": 10,
+  "global_step": 6185,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016166841807452913,
+      "grad_norm": 0.9894065856933594,
+      "learning_rate": 0.0002,
+      "loss": 1.6636,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032333683614905826,
+      "grad_norm": 1.7810699939727783,
+      "learning_rate": 0.0002,
+      "loss": 1.1528,
+      "step": 20
+    },
+    {
+      "epoch": 0.004850052542235874,
+      "grad_norm": 0.5969577431678772,
+      "learning_rate": 0.0002,
+      "loss": 0.9767,
+      "step": 30
+    },
+    {
+      "epoch": 0.006466736722981165,
+      "grad_norm": 0.6354120969772339,
+      "learning_rate": 0.0002,
+      "loss": 0.9772,
+      "step": 40
+    },
+    {
+      "epoch": 0.008083420903726457,
+      "grad_norm": 0.5604607462882996,
+      "learning_rate": 0.0002,
+      "loss": 0.8643,
+      "step": 50
+    },
+    {
+      "epoch": 0.009700105084471748,
+      "grad_norm": 0.4676193594932556,
+      "learning_rate": 0.0002,
+      "loss": 0.8841,
+      "step": 60
+    },
+    {
+      "epoch": 0.01131678926521704,
+      "grad_norm": 0.6099211573600769,
+      "learning_rate": 0.0002,
+      "loss": 0.9022,
+      "step": 70
+    },
+    {
+      "epoch": 0.01293347344596233,
+      "grad_norm": 0.48639994859695435,
+      "learning_rate": 0.0002,
+      "loss": 0.9133,
+      "step": 80
+    },
+    {
+      "epoch": 0.014550157626707623,
+      "grad_norm": 0.4904264509677887,
+      "learning_rate": 0.0002,
+      "loss": 0.8704,
+      "step": 90
+    },
+    {
+      "epoch": 0.016166841807452915,
+      "grad_norm": 2.8334362506866455,
+      "learning_rate": 0.0002,
+      "loss": 0.8855,
+      "step": 100
+    },
+    {
+      "epoch": 0.017783525988198205,
+      "grad_norm": 0.43221670389175415,
+      "learning_rate": 0.0002,
+      "loss": 0.8958,
+      "step": 110
+    },
+    {
+      "epoch": 0.019400210168943496,
+      "grad_norm": 0.42244166135787964,
+      "learning_rate": 0.0002,
+      "loss": 0.8412,
+      "step": 120
+    },
+    {
+      "epoch": 0.02101689434968879,
+      "grad_norm": 0.45363298058509827,
+      "learning_rate": 0.0002,
+      "loss": 0.8467,
+      "step": 130
+    },
+    {
+      "epoch": 0.02263357853043408,
+      "grad_norm": 0.44816508889198303,
+      "learning_rate": 0.0002,
+      "loss": 0.8641,
+      "step": 140
+    },
+    {
+      "epoch": 0.02425026271117937,
+      "grad_norm": 0.43308213353157043,
+      "learning_rate": 0.0002,
+      "loss": 0.8496,
+      "step": 150
+    },
+    {
+      "epoch": 0.02586694689192466,
+      "grad_norm": 0.4084763526916504,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 160
+    },
+    {
+      "epoch": 0.027483631072669955,
+      "grad_norm": 0.5363703966140747,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 170
+    },
+    {
+      "epoch": 0.029100315253415245,
+      "grad_norm": 0.4619699716567993,
+      "learning_rate": 0.0002,
+      "loss": 0.8558,
+      "step": 180
+    },
+    {
+      "epoch": 0.030716999434160536,
+      "grad_norm": 0.49069908261299133,
+      "learning_rate": 0.0002,
+      "loss": 0.8878,
+      "step": 190
+    },
+    {
+      "epoch": 0.03233368361490583,
+      "grad_norm": 0.4645835757255554,
+      "learning_rate": 0.0002,
+      "loss": 0.8867,
+      "step": 200
+    },
+    {
+      "epoch": 0.03395036779565112,
+      "grad_norm": 1.2411243915557861,
+      "learning_rate": 0.0002,
+      "loss": 0.8842,
+      "step": 210
+    },
+    {
+      "epoch": 0.03556705197639641,
+      "grad_norm": 0.5211851596832275,
+      "learning_rate": 0.0002,
+      "loss": 0.8245,
+      "step": 220
+    },
+    {
+      "epoch": 0.037183736157141704,
+      "grad_norm": 0.5253691673278809,
+      "learning_rate": 0.0002,
+      "loss": 0.8194,
+      "step": 230
+    },
+    {
+      "epoch": 0.03880042033788699,
+      "grad_norm": 0.4567478895187378,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 240
+    },
+    {
+      "epoch": 0.040417104518632285,
+      "grad_norm": 0.5472128391265869,
+      "learning_rate": 0.0002,
+      "loss": 0.838,
+      "step": 250
+    },
+    {
+      "epoch": 0.04203378869937758,
+      "grad_norm": 0.42978546023368835,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 260
+    },
+    {
+      "epoch": 0.043650472880122866,
+      "grad_norm": 0.601734459400177,
+      "learning_rate": 0.0002,
+      "loss": 0.8334,
+      "step": 270
+    },
+    {
+      "epoch": 0.04526715706086816,
+      "grad_norm": 0.4286513328552246,
+      "learning_rate": 0.0002,
+      "loss": 0.815,
+      "step": 280
+    },
+    {
+      "epoch": 0.046883841241613454,
+      "grad_norm": 0.5230861902236938,
+      "learning_rate": 0.0002,
+      "loss": 0.8758,
+      "step": 290
+    },
+    {
+      "epoch": 0.04850052542235874,
+      "grad_norm": 0.6504611968994141,
+      "learning_rate": 0.0002,
+      "loss": 0.8636,
+      "step": 300
+    },
+    {
+      "epoch": 0.050117209603104035,
+      "grad_norm": 0.43485215306282043,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 310
+    },
+    {
+      "epoch": 0.05173389378384932,
+      "grad_norm": 0.4717007875442505,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 320
+    },
+    {
+      "epoch": 0.053350577964594616,
+      "grad_norm": 0.4059787690639496,
+      "learning_rate": 0.0002,
+      "loss": 0.8469,
+      "step": 330
+    },
+    {
+      "epoch": 0.05496726214533991,
+      "grad_norm": 0.4366913437843323,
+      "learning_rate": 0.0002,
+      "loss": 0.8866,
+      "step": 340
+    },
+    {
+      "epoch": 0.0565839463260852,
+      "grad_norm": 0.4233848452568054,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 350
+    },
+    {
+      "epoch": 0.05820063050683049,
+      "grad_norm": 0.4209108352661133,
+      "learning_rate": 0.0002,
+      "loss": 0.8456,
+      "step": 360
+    },
+    {
+      "epoch": 0.059817314687575784,
+      "grad_norm": 0.41637396812438965,
+      "learning_rate": 0.0002,
+      "loss": 0.816,
+      "step": 370
+    },
+    {
+      "epoch": 0.06143399886832107,
+      "grad_norm": 0.46235376596450806,
+      "learning_rate": 0.0002,
+      "loss": 0.7976,
+      "step": 380
+    },
+    {
+      "epoch": 0.06305068304906636,
+      "grad_norm": 0.4013484716415405,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 390
+    },
+    {
+      "epoch": 0.06466736722981166,
+      "grad_norm": 0.47443896532058716,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 400
+    },
+    {
+      "epoch": 0.06628405141055695,
+      "grad_norm": 0.3942156434059143,
+      "learning_rate": 0.0002,
+      "loss": 0.8666,
+      "step": 410
+    },
+    {
+      "epoch": 0.06790073559130223,
+      "grad_norm": 0.4965320825576782,
+      "learning_rate": 0.0002,
+      "loss": 0.8402,
+      "step": 420
+    },
+    {
+      "epoch": 0.06951741977204753,
+      "grad_norm": 0.4304835796356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8317,
+      "step": 430
+    },
+    {
+      "epoch": 0.07113410395279282,
+      "grad_norm": 0.511726975440979,
+      "learning_rate": 0.0002,
+      "loss": 0.8528,
+      "step": 440
+    },
+    {
+      "epoch": 0.07275078813353811,
+      "grad_norm": 0.4040689170360565,
+      "learning_rate": 0.0002,
+      "loss": 0.8675,
+      "step": 450
+    },
+    {
+      "epoch": 0.07436747231428341,
+      "grad_norm": 0.5402171015739441,
+      "learning_rate": 0.0002,
+      "loss": 0.8788,
+      "step": 460
+    },
+    {
+      "epoch": 0.0759841564950287,
+      "grad_norm": 0.4174517095088959,
+      "learning_rate": 0.0002,
+      "loss": 0.8737,
+      "step": 470
+    },
+    {
+      "epoch": 0.07760084067577398,
+      "grad_norm": 0.4306182265281677,
+      "learning_rate": 0.0002,
+      "loss": 0.7605,
+      "step": 480
+    },
+    {
+      "epoch": 0.07921752485651928,
+      "grad_norm": 0.535210132598877,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 490
+    },
+    {
+      "epoch": 0.08083420903726457,
+      "grad_norm": 0.5339109897613525,
+      "learning_rate": 0.0002,
+      "loss": 0.7825,
+      "step": 500
+    },
+    {
+      "epoch": 0.08245089321800986,
+      "grad_norm": 0.45754891633987427,
+      "learning_rate": 0.0002,
+      "loss": 0.8985,
+      "step": 510
+    },
+    {
+      "epoch": 0.08406757739875516,
+      "grad_norm": 0.43820783495903015,
+      "learning_rate": 0.0002,
+      "loss": 0.8144,
+      "step": 520
+    },
+    {
+      "epoch": 0.08568426157950045,
+      "grad_norm": 0.4434749186038971,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 530
+    },
+    {
+      "epoch": 0.08730094576024573,
+      "grad_norm": 0.43111467361450195,
+      "learning_rate": 0.0002,
+      "loss": 0.7857,
+      "step": 540
+    },
+    {
+      "epoch": 0.08891762994099103,
+      "grad_norm": 0.4378940165042877,
+      "learning_rate": 0.0002,
+      "loss": 0.8418,
+      "step": 550
+    },
+    {
+      "epoch": 0.09053431412173632,
+      "grad_norm": 0.4772215187549591,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 560
+    },
+    {
+      "epoch": 0.09215099830248161,
+      "grad_norm": 0.6837629079818726,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 570
+    },
+    {
+      "epoch": 0.09376768248322691,
+      "grad_norm": 0.42241212725639343,
+      "learning_rate": 0.0002,
+      "loss": 0.8607,
+      "step": 580
+    },
+    {
+      "epoch": 0.0953843666639722,
+      "grad_norm": 0.5165936350822449,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 590
+    },
+    {
+      "epoch": 0.09700105084471748,
+      "grad_norm": 0.48737478256225586,
+      "learning_rate": 0.0002,
+      "loss": 0.8664,
+      "step": 600
+    },
+    {
+      "epoch": 0.09861773502546278,
+      "grad_norm": 0.47419852018356323,
+      "learning_rate": 0.0002,
+      "loss": 0.8806,
+      "step": 610
+    },
+    {
+      "epoch": 0.10023441920620807,
+      "grad_norm": 0.4975486099720001,
+      "learning_rate": 0.0002,
+      "loss": 0.8254,
+      "step": 620
+    },
+    {
+      "epoch": 0.10185110338695336,
+      "grad_norm": 0.49123844504356384,
+      "learning_rate": 0.0002,
+      "loss": 0.8548,
+      "step": 630
+    },
+    {
+      "epoch": 0.10346778756769864,
+      "grad_norm": 0.6288952827453613,
+      "learning_rate": 0.0002,
+      "loss": 0.8911,
+      "step": 640
+    },
+    {
+      "epoch": 0.10508447174844394,
+      "grad_norm": 0.4277345836162567,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 650
+    },
+    {
+      "epoch": 0.10670115592918923,
+      "grad_norm": 0.4021061956882477,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 660
+    },
+    {
+      "epoch": 0.10831784010993452,
+      "grad_norm": 0.3492237329483032,
+      "learning_rate": 0.0002,
+      "loss": 0.87,
+      "step": 670
+    },
+    {
+      "epoch": 0.10993452429067982,
+      "grad_norm": 0.4341012239456177,
+      "learning_rate": 0.0002,
+      "loss": 0.8698,
+      "step": 680
+    },
+    {
+      "epoch": 0.1115512084714251,
+      "grad_norm": 0.7296304106712341,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 690
+    },
+    {
+      "epoch": 0.1131678926521704,
+      "grad_norm": 0.397494912147522,
+      "learning_rate": 0.0002,
+      "loss": 0.8433,
+      "step": 700
+    },
+    {
+      "epoch": 0.1147845768329157,
+      "grad_norm": 0.396431028842926,
+      "learning_rate": 0.0002,
+      "loss": 0.827,
+      "step": 710
+    },
+    {
+      "epoch": 0.11640126101366098,
+      "grad_norm": 0.48842838406562805,
+      "learning_rate": 0.0002,
+      "loss": 0.8379,
+      "step": 720
+    },
+    {
+      "epoch": 0.11801794519440627,
+      "grad_norm": 0.46322616934776306,
+      "learning_rate": 0.0002,
+      "loss": 0.8238,
+      "step": 730
+    },
+    {
+      "epoch": 0.11963462937515157,
+      "grad_norm": 0.47990912199020386,
+      "learning_rate": 0.0002,
+      "loss": 0.8041,
+      "step": 740
+    },
+    {
+      "epoch": 0.12125131355589686,
+      "grad_norm": 0.4997142255306244,
+      "learning_rate": 0.0002,
+      "loss": 0.82,
+      "step": 750
+    },
+    {
+      "epoch": 0.12286799773664214,
+      "grad_norm": 0.4040526747703552,
+      "learning_rate": 0.0002,
+      "loss": 0.7702,
+      "step": 760
+    },
+    {
+      "epoch": 0.12448468191738744,
+      "grad_norm": 0.453095942735672,
+      "learning_rate": 0.0002,
+      "loss": 0.863,
+      "step": 770
+    },
+    {
+      "epoch": 0.12610136609813272,
+      "grad_norm": 0.4636971950531006,
+      "learning_rate": 0.0002,
+      "loss": 0.8792,
+      "step": 780
+    },
+    {
+      "epoch": 0.12771805027887803,
+      "grad_norm": 0.4279276132583618,
+      "learning_rate": 0.0002,
+      "loss": 0.8112,
+      "step": 790
+    },
+    {
+      "epoch": 0.12933473445962332,
+      "grad_norm": 0.46212655305862427,
+      "learning_rate": 0.0002,
+      "loss": 0.8711,
+      "step": 800
+    },
+    {
+      "epoch": 0.1309514186403686,
+      "grad_norm": 0.43127650022506714,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 810
+    },
+    {
+      "epoch": 0.1325681028211139,
+      "grad_norm": 0.4201301336288452,
+      "learning_rate": 0.0002,
+      "loss": 0.8476,
+      "step": 820
+    },
+    {
+      "epoch": 0.13418478700185918,
+      "grad_norm": 0.42583167552948,
+      "learning_rate": 0.0002,
+      "loss": 0.8078,
+      "step": 830
+    },
+    {
+      "epoch": 0.13580147118260447,
+      "grad_norm": 0.4535622000694275,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 840
+    },
+    {
+      "epoch": 0.13741815536334978,
+      "grad_norm": 0.4116036891937256,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 850
+    },
+    {
+      "epoch": 0.13903483954409507,
+      "grad_norm": 0.45997580885887146,
+      "learning_rate": 0.0002,
+      "loss": 0.8466,
+      "step": 860
+    },
+    {
+      "epoch": 0.14065152372484035,
+      "grad_norm": 0.4487837255001068,
+      "learning_rate": 0.0002,
+      "loss": 0.8917,
+      "step": 870
+    },
+    {
+      "epoch": 0.14226820790558564,
+      "grad_norm": 0.43650057911872864,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 880
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 0.5335358381271362,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 890
+    },
+    {
+      "epoch": 0.14550157626707622,
+      "grad_norm": 0.5989000201225281,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 900
+    },
+    {
+      "epoch": 0.14711826044782153,
+      "grad_norm": 0.517179012298584,
+      "learning_rate": 0.0002,
+      "loss": 0.8385,
+      "step": 910
+    },
+    {
+      "epoch": 0.14873494462856682,
+      "grad_norm": 0.44435232877731323,
+      "learning_rate": 0.0002,
+      "loss": 0.8255,
+      "step": 920
+    },
+    {
+      "epoch": 0.1503516288093121,
+      "grad_norm": 0.42635923624038696,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 930
+    },
+    {
+      "epoch": 0.1519683129900574,
+      "grad_norm": 0.49603334069252014,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 940
+    },
+    {
+      "epoch": 0.15358499717080268,
+      "grad_norm": 0.40639808773994446,
+      "learning_rate": 0.0002,
+      "loss": 0.8377,
+      "step": 950
+    },
+    {
+      "epoch": 0.15520168135154797,
+      "grad_norm": 0.4850759208202362,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 960
+    },
+    {
+      "epoch": 0.15681836553229328,
+      "grad_norm": 0.4427442252635956,
+      "learning_rate": 0.0002,
+      "loss": 0.846,
+      "step": 970
+    },
+    {
+      "epoch": 0.15843504971303857,
+      "grad_norm": 0.3760930001735687,
+      "learning_rate": 0.0002,
+      "loss": 0.8705,
+      "step": 980
+    },
+    {
+      "epoch": 0.16005173389378385,
+      "grad_norm": 0.4794144332408905,
+      "learning_rate": 0.0002,
+      "loss": 0.8644,
+      "step": 990
+    },
+    {
+      "epoch": 0.16166841807452914,
+      "grad_norm": 0.45828768610954285,
+      "learning_rate": 0.0002,
+      "loss": 0.8002,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16328510225527443,
+      "grad_norm": 0.6313053369522095,
+      "learning_rate": 0.0002,
+      "loss": 0.7658,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16490178643601971,
+      "grad_norm": 0.45041006803512573,
+      "learning_rate": 0.0002,
+      "loss": 0.8047,
+      "step": 1020
+    },
+    {
+      "epoch": 0.166518470616765,
+      "grad_norm": 0.441403865814209,
+      "learning_rate": 0.0002,
+      "loss": 0.8423,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16813515479751032,
+      "grad_norm": 0.8171296119689941,
+      "learning_rate": 0.0002,
+      "loss": 0.8475,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1697518389782556,
+      "grad_norm": 0.7137420773506165,
+      "learning_rate": 0.0002,
+      "loss": 0.845,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1713685231590009,
+      "grad_norm": 0.5236809849739075,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17298520733974618,
+      "grad_norm": 0.5021864175796509,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17460189152049146,
+      "grad_norm": 0.47347521781921387,
+      "learning_rate": 0.0002,
+      "loss": 0.8305,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17621857570123675,
+      "grad_norm": 0.4631653428077698,
+      "learning_rate": 0.0002,
+      "loss": 0.8105,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17783525988198207,
+      "grad_norm": 0.49169182777404785,
+      "learning_rate": 0.0002,
+      "loss": 0.8166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17945194406272735,
+      "grad_norm": 0.5019739270210266,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18106862824347264,
+      "grad_norm": 0.5100422501564026,
+      "learning_rate": 0.0002,
+      "loss": 0.8247,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18268531242421793,
+      "grad_norm": 0.3888324499130249,
+      "learning_rate": 0.0002,
+      "loss": 0.8142,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18430199660496321,
+      "grad_norm": 0.39765217900276184,
+      "learning_rate": 0.0002,
+      "loss": 0.8533,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1859186807857085,
+      "grad_norm": 0.47190186381340027,
+      "learning_rate": 0.0002,
+      "loss": 0.8541,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18753536496645382,
+      "grad_norm": 0.4464188814163208,
+      "learning_rate": 0.0002,
+      "loss": 0.8301,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1891520491471991,
+      "grad_norm": 0.5153930187225342,
+      "learning_rate": 0.0002,
+      "loss": 0.8341,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1907687333279444,
+      "grad_norm": 0.4779708683490753,
+      "learning_rate": 0.0002,
+      "loss": 0.8033,
+      "step": 1180
+    },
+    {
+      "epoch": 0.19238541750868968,
+      "grad_norm": 0.4834315776824951,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19400210168943496,
+      "grad_norm": 0.402357816696167,
+      "learning_rate": 0.0002,
+      "loss": 0.7721,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19561878587018025,
+      "grad_norm": 0.45899084210395813,
+      "learning_rate": 0.0002,
+      "loss": 0.7941,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19723547005092557,
+      "grad_norm": 0.5106529593467712,
+      "learning_rate": 0.0002,
+      "loss": 0.8353,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19885215423167085,
+      "grad_norm": 0.45261722803115845,
+      "learning_rate": 0.0002,
+      "loss": 0.7816,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20046883841241614,
+      "grad_norm": 0.4647127091884613,
+      "learning_rate": 0.0002,
+      "loss": 0.8068,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20208552259316143,
+      "grad_norm": 0.4849368929862976,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2037022067739067,
+      "grad_norm": 0.4518061578273773,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 1260
+    },
+    {
+      "epoch": 0.205318890954652,
+      "grad_norm": 0.49535325169563293,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2069355751353973,
+      "grad_norm": 0.4835205376148224,
+      "learning_rate": 0.0002,
+      "loss": 0.8348,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2085522593161426,
+      "grad_norm": 0.45308539271354675,
+      "learning_rate": 0.0002,
+      "loss": 0.8428,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2101689434968879,
+      "grad_norm": 0.5369905233383179,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21178562767763318,
+      "grad_norm": 0.5031622052192688,
+      "learning_rate": 0.0002,
+      "loss": 0.8676,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21340231185837846,
+      "grad_norm": 0.48010334372520447,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21501899603912375,
+      "grad_norm": 0.4905701279640198,
+      "learning_rate": 0.0002,
+      "loss": 0.806,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21663568021986904,
+      "grad_norm": 0.43531742691993713,
+      "learning_rate": 0.0002,
+      "loss": 0.7885,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21825236440061435,
+      "grad_norm": 0.44330692291259766,
+      "learning_rate": 0.0002,
+      "loss": 0.8191,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21986904858135964,
+      "grad_norm": 0.5384416580200195,
+      "learning_rate": 0.0002,
+      "loss": 0.8205,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22148573276210493,
+      "grad_norm": 0.4181833863258362,
+      "learning_rate": 0.0002,
+      "loss": 0.7726,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2231024169428502,
+      "grad_norm": 0.523833692073822,
+      "learning_rate": 0.0002,
+      "loss": 0.8311,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.5528736710548401,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2263357853043408,
+      "grad_norm": 0.43515023589134216,
+      "learning_rate": 0.0002,
+      "loss": 0.8079,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2279524694850861,
+      "grad_norm": 0.48809877038002014,
+      "learning_rate": 0.0002,
+      "loss": 0.8403,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2295691536658314,
+      "grad_norm": 0.43591251969337463,
+      "learning_rate": 0.0002,
+      "loss": 0.8165,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23118583784657668,
+      "grad_norm": 0.44625312089920044,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23280252202732196,
+      "grad_norm": 0.4390665292739868,
+      "learning_rate": 0.0002,
+      "loss": 0.8134,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23441920620806725,
+      "grad_norm": 0.48496049642562866,
+      "learning_rate": 0.0002,
+      "loss": 0.8465,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23603589038881254,
+      "grad_norm": 0.45919957756996155,
+      "learning_rate": 0.0002,
+      "loss": 0.775,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23765257456955785,
+      "grad_norm": 0.5471845865249634,
+      "learning_rate": 0.0002,
+      "loss": 0.8659,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23926925875030314,
+      "grad_norm": 0.47269317507743835,
+      "learning_rate": 0.0002,
+      "loss": 0.8164,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24088594293104842,
+      "grad_norm": 0.4930245578289032,
+      "learning_rate": 0.0002,
+      "loss": 0.854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2425026271117937,
+      "grad_norm": 0.5605630278587341,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 1500
+    },
+    {
+      "epoch": 0.244119311292539,
+      "grad_norm": 0.4435870945453644,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24573599547328429,
+      "grad_norm": 0.4941999912261963,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24735267965402957,
+      "grad_norm": 0.5100624561309814,
+      "learning_rate": 0.0002,
+      "loss": 0.8427,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2489693638347749,
+      "grad_norm": 0.4638267457485199,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25058604801552015,
+      "grad_norm": 0.5071570873260498,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 1550
+    },
+    {
+      "epoch": 0.25220273219626543,
+      "grad_norm": 0.4291319251060486,
+      "learning_rate": 0.0002,
+      "loss": 0.7724,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2538194163770108,
+      "grad_norm": 0.5388049483299255,
+      "learning_rate": 0.0002,
+      "loss": 0.7984,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25543610055775606,
+      "grad_norm": 0.5083683729171753,
+      "learning_rate": 0.0002,
+      "loss": 0.8176,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25705278473850135,
+      "grad_norm": 0.4824463725090027,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25866946891924664,
+      "grad_norm": 0.41177722811698914,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2602861530999919,
+      "grad_norm": 0.5656219124794006,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2619028372807372,
+      "grad_norm": 0.41063204407691956,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2635195214614825,
+      "grad_norm": 0.4897061288356781,
+      "learning_rate": 0.0002,
+      "loss": 0.7998,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2651362056422278,
+      "grad_norm": 0.4454376697540283,
+      "learning_rate": 0.0002,
+      "loss": 0.8198,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26675288982297307,
+      "grad_norm": 0.4355238378047943,
+      "learning_rate": 0.0002,
+      "loss": 0.8684,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26836957400371836,
+      "grad_norm": 0.458310067653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26998625818446365,
+      "grad_norm": 0.4752083718776703,
+      "learning_rate": 0.0002,
+      "loss": 0.7935,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27160294236520893,
+      "grad_norm": 0.4666106402873993,
+      "learning_rate": 0.0002,
+      "loss": 0.8267,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2732196265459543,
+      "grad_norm": 0.4213818609714508,
+      "learning_rate": 0.0002,
+      "loss": 0.8252,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27483631072669956,
+      "grad_norm": 0.5768913626670837,
+      "learning_rate": 0.0002,
+      "loss": 0.8559,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27645299490744485,
+      "grad_norm": 0.4209914803504944,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27806967908819014,
+      "grad_norm": 0.501909613609314,
+      "learning_rate": 0.0002,
+      "loss": 0.8167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2796863632689354,
+      "grad_norm": 0.5266261100769043,
+      "learning_rate": 0.0002,
+      "loss": 0.7832,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2813030474496807,
+      "grad_norm": 0.43806859850883484,
+      "learning_rate": 0.0002,
+      "loss": 0.8102,
+      "step": 1740
+    },
+    {
+      "epoch": 0.282919731630426,
+      "grad_norm": 0.46048814058303833,
+      "learning_rate": 0.0002,
+      "loss": 0.8157,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2845364158111713,
+      "grad_norm": 0.44972819089889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8596,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28615309999191657,
+      "grad_norm": 0.5114831328392029,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.47931742668151855,
+      "learning_rate": 0.0002,
+      "loss": 0.8361,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28938646835340714,
+      "grad_norm": 0.5092599987983704,
+      "learning_rate": 0.0002,
+      "loss": 0.8265,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29100315253415243,
+      "grad_norm": 0.37581443786621094,
+      "learning_rate": 0.0002,
+      "loss": 0.8506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2926198367148977,
+      "grad_norm": 0.47097381949424744,
+      "learning_rate": 0.0002,
+      "loss": 0.7932,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29423652089564306,
+      "grad_norm": 0.48300236463546753,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29585320507638835,
+      "grad_norm": 0.5600419640541077,
+      "learning_rate": 0.0002,
+      "loss": 0.8391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29746988925713364,
+      "grad_norm": 0.48555272817611694,
+      "learning_rate": 0.0002,
+      "loss": 0.8507,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2990865734378789,
+      "grad_norm": 0.3752668499946594,
+      "learning_rate": 0.0002,
+      "loss": 0.7657,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3007032576186242,
+      "grad_norm": 0.5328747034072876,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3023199417993695,
+      "grad_norm": 0.48716455698013306,
+      "learning_rate": 0.0002,
+      "loss": 0.8426,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3039366259801148,
+      "grad_norm": 0.5011493563652039,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30555331016086007,
+      "grad_norm": 0.46461427211761475,
+      "learning_rate": 0.0002,
+      "loss": 0.852,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30716999434160536,
+      "grad_norm": 0.36630210280418396,
+      "learning_rate": 0.0002,
+      "loss": 0.8478,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30878667852235064,
+      "grad_norm": 0.4217296242713928,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31040336270309593,
+      "grad_norm": 0.4394875466823578,
+      "learning_rate": 0.0002,
+      "loss": 0.8128,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3120200468838412,
+      "grad_norm": 0.6587965488433838,
+      "learning_rate": 0.0002,
+      "loss": 0.8471,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31363673106458656,
+      "grad_norm": 0.5469298958778381,
+      "learning_rate": 0.0002,
+      "loss": 0.8565,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31525341524533185,
+      "grad_norm": 0.4371595084667206,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31687009942607713,
+      "grad_norm": 0.4809541404247284,
+      "learning_rate": 0.0002,
+      "loss": 0.887,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3184867836068224,
+      "grad_norm": 0.6061086654663086,
+      "learning_rate": 0.0002,
+      "loss": 0.7855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3201034677875677,
+      "grad_norm": 0.5342657566070557,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 1980
+    },
+    {
+      "epoch": 0.321720151968313,
+      "grad_norm": 0.5057743787765503,
+      "learning_rate": 0.0002,
+      "loss": 0.7955,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3233368361490583,
+      "grad_norm": 0.528626024723053,
+      "learning_rate": 0.0002,
+      "loss": 0.7774,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32495352032980357,
+      "grad_norm": 0.46742770075798035,
+      "learning_rate": 0.0002,
+      "loss": 0.8845,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32657020451054886,
+      "grad_norm": 0.515101432800293,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32818688869129414,
+      "grad_norm": 0.41941216588020325,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32980357287203943,
+      "grad_norm": 0.49902522563934326,
+      "learning_rate": 0.0002,
+      "loss": 0.7637,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3314202570527847,
+      "grad_norm": 0.4120897650718689,
+      "learning_rate": 0.0002,
+      "loss": 0.7822,
+      "step": 2050
+    },
+    {
+      "epoch": 0.33303694123353,
+      "grad_norm": 0.45352041721343994,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33465362541427535,
+      "grad_norm": 0.523199737071991,
+      "learning_rate": 0.0002,
+      "loss": 0.7913,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33627030959502063,
+      "grad_norm": 0.4390358626842499,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3378869937757659,
+      "grad_norm": 0.6752901077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3395036779565112,
+      "grad_norm": 0.547821044921875,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3411203621372565,
+      "grad_norm": 0.5161308646202087,
+      "learning_rate": 0.0002,
+      "loss": 0.8561,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3427370463180018,
+      "grad_norm": 0.4565401077270508,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 2120
+    },
+    {
+      "epoch": 0.34435373049874707,
+      "grad_norm": 0.4666115939617157,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34597041467949236,
+      "grad_norm": 0.4090428352355957,
+      "learning_rate": 0.0002,
+      "loss": 0.8189,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34758709886023764,
+      "grad_norm": 0.510845422744751,
+      "learning_rate": 0.0002,
+      "loss": 0.8817,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34920378304098293,
+      "grad_norm": 0.42861923575401306,
+      "learning_rate": 0.0002,
+      "loss": 0.8398,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3508204672217282,
+      "grad_norm": 0.4476332664489746,
+      "learning_rate": 0.0002,
+      "loss": 0.7716,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3524371514024735,
+      "grad_norm": 0.6065791249275208,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2180
+    },
+    {
+      "epoch": 0.35405383558321885,
+      "grad_norm": 0.42335066199302673,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35567051976396413,
+      "grad_norm": 0.5094629526138306,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3572872039447094,
+      "grad_norm": 0.5476373434066772,
+      "learning_rate": 0.0002,
+      "loss": 0.7807,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3589038881254547,
+      "grad_norm": 0.3911719024181366,
+      "learning_rate": 0.0002,
+      "loss": 0.814,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3605205723062,
+      "grad_norm": 0.6599636077880859,
+      "learning_rate": 0.0002,
+      "loss": 0.8599,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3621372564869453,
+      "grad_norm": 0.40381914377212524,
+      "learning_rate": 0.0002,
+      "loss": 0.7482,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36375394066769057,
+      "grad_norm": 0.4433908462524414,
+      "learning_rate": 0.0002,
+      "loss": 0.7772,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36537062484843585,
+      "grad_norm": 0.578326940536499,
+      "learning_rate": 0.0002,
+      "loss": 0.8503,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36698730902918114,
+      "grad_norm": 0.5734784007072449,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36860399320992643,
+      "grad_norm": 0.45555487275123596,
+      "learning_rate": 0.0002,
+      "loss": 0.8193,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3702206773906717,
+      "grad_norm": 0.5666276216506958,
+      "learning_rate": 0.0002,
+      "loss": 0.7929,
+      "step": 2290
+    },
+    {
+      "epoch": 0.371837361571417,
+      "grad_norm": 0.5461117625236511,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3734540457521623,
+      "grad_norm": 0.6318911910057068,
+      "learning_rate": 0.0002,
+      "loss": 0.8204,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37507072993290763,
+      "grad_norm": 0.493263304233551,
+      "learning_rate": 0.0002,
+      "loss": 0.7964,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3766874141136529,
+      "grad_norm": 0.5888760089874268,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3783040982943982,
+      "grad_norm": 0.48671841621398926,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3799207824751435,
+      "grad_norm": 0.4385145306587219,
+      "learning_rate": 0.0002,
+      "loss": 0.8367,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3815374666558888,
+      "grad_norm": 0.5523318648338318,
+      "learning_rate": 0.0002,
+      "loss": 0.812,
+      "step": 2360
+    },
+    {
+      "epoch": 0.38315415083663407,
+      "grad_norm": 0.7308220267295837,
+      "learning_rate": 0.0002,
+      "loss": 0.8351,
+      "step": 2370
+    },
+    {
+      "epoch": 0.38477083501737935,
+      "grad_norm": 0.554214358329773,
+      "learning_rate": 0.0002,
+      "loss": 0.859,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38638751919812464,
+      "grad_norm": 0.5425800085067749,
+      "learning_rate": 0.0002,
+      "loss": 0.8146,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3880042033788699,
+      "grad_norm": 0.48811158537864685,
+      "learning_rate": 0.0002,
+      "loss": 0.8282,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3896208875596152,
+      "grad_norm": 0.49212366342544556,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3912375717403605,
+      "grad_norm": 0.5222218632698059,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3928542559211058,
+      "grad_norm": 0.4699819087982178,
+      "learning_rate": 0.0002,
+      "loss": 0.8182,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39447094010185113,
+      "grad_norm": 0.46153587102890015,
+      "learning_rate": 0.0002,
+      "loss": 0.7919,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3960876242825964,
+      "grad_norm": 0.4150611162185669,
+      "learning_rate": 0.0002,
+      "loss": 0.8111,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3977043084633417,
+      "grad_norm": 0.5799614787101746,
+      "learning_rate": 0.0002,
+      "loss": 0.8589,
+      "step": 2460
+    },
+    {
+      "epoch": 0.399320992644087,
+      "grad_norm": 0.56536865234375,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4009376768248323,
+      "grad_norm": 0.5451247096061707,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40255436100557757,
+      "grad_norm": 0.5914521217346191,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40417104518632285,
+      "grad_norm": 0.4428117275238037,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40578772936706814,
+      "grad_norm": 0.48580947518348694,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4074044135478134,
+      "grad_norm": 0.436734676361084,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4090210977285587,
+      "grad_norm": 0.5752223134040833,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 2530
+    },
+    {
+      "epoch": 0.410637781909304,
+      "grad_norm": 0.4271308183670044,
+      "learning_rate": 0.0002,
+      "loss": 0.8181,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4122544660900493,
+      "grad_norm": 0.46294718980789185,
+      "learning_rate": 0.0002,
+      "loss": 0.8058,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4138711502707946,
+      "grad_norm": 0.49407583475112915,
+      "learning_rate": 0.0002,
+      "loss": 0.8473,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4154878344515399,
+      "grad_norm": 0.4729035496711731,
+      "learning_rate": 0.0002,
+      "loss": 0.7881,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4171045186322852,
+      "grad_norm": 0.4129747152328491,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4187212028130305,
+      "grad_norm": 0.5684236288070679,
+      "learning_rate": 0.0002,
+      "loss": 0.7859,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4203378869937758,
+      "grad_norm": 0.4862157106399536,
+      "learning_rate": 0.0002,
+      "loss": 0.811,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42195457117452106,
+      "grad_norm": 0.46567976474761963,
+      "learning_rate": 0.0002,
+      "loss": 0.7582,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42357125535526635,
+      "grad_norm": 0.5710650682449341,
+      "learning_rate": 0.0002,
+      "loss": 0.7755,
+      "step": 2620
+    },
+    {
+      "epoch": 0.42518793953601164,
+      "grad_norm": 0.5660041570663452,
+      "learning_rate": 0.0002,
+      "loss": 0.8573,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4268046237167569,
+      "grad_norm": 0.47944375872612,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4284213078975022,
+      "grad_norm": 0.537223756313324,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4300379920782475,
+      "grad_norm": 0.41669997572898865,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 0.44727686047554016,
+      "learning_rate": 0.0002,
+      "loss": 0.7785,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4332713604397381,
+      "grad_norm": 0.5600888729095459,
+      "learning_rate": 0.0002,
+      "loss": 0.8241,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4348880446204834,
+      "grad_norm": 0.39820605516433716,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4365047288012287,
+      "grad_norm": 0.5637655854225159,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 2700
+    },
+    {
+      "epoch": 0.438121412981974,
+      "grad_norm": 0.6363666653633118,
+      "learning_rate": 0.0002,
+      "loss": 0.855,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4397380971627193,
+      "grad_norm": 0.5656129121780396,
+      "learning_rate": 0.0002,
+      "loss": 0.8468,
+      "step": 2720
+    },
+    {
+      "epoch": 0.44135478134346456,
+      "grad_norm": 0.5600156188011169,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 2730
+    },
+    {
+      "epoch": 0.44297146552420985,
+      "grad_norm": 0.5506579875946045,
+      "learning_rate": 0.0002,
+      "loss": 0.8405,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44458814970495514,
+      "grad_norm": 0.49878305196762085,
+      "learning_rate": 0.0002,
+      "loss": 0.7725,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4462048338857004,
+      "grad_norm": 0.4569213092327118,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4478215180664457,
+      "grad_norm": 0.6056680083274841,
+      "learning_rate": 0.0002,
+      "loss": 0.8028,
+      "step": 2770
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.44474557042121887,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4510548864279363,
+      "grad_norm": 0.46055394411087036,
+      "learning_rate": 0.0002,
+      "loss": 0.801,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4526715706086816,
+      "grad_norm": 0.4904133379459381,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45428825478942686,
+      "grad_norm": 0.5647031664848328,
+      "learning_rate": 0.0002,
+      "loss": 0.8829,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4559049389701722,
+      "grad_norm": 0.5759473443031311,
+      "learning_rate": 0.0002,
+      "loss": 0.8622,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4575216231509175,
+      "grad_norm": 0.5161895751953125,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4591383073316628,
+      "grad_norm": 0.4248254597187042,
+      "learning_rate": 0.0002,
+      "loss": 0.8045,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46075499151240806,
+      "grad_norm": 0.45395001769065857,
+      "learning_rate": 0.0002,
+      "loss": 0.7838,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46237167569315335,
+      "grad_norm": 0.5358697772026062,
+      "learning_rate": 0.0002,
+      "loss": 0.8208,
+      "step": 2860
+    },
+    {
+      "epoch": 0.46398835987389864,
+      "grad_norm": 0.5379165410995483,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4656050440546439,
+      "grad_norm": 0.4601989686489105,
+      "learning_rate": 0.0002,
+      "loss": 0.7403,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4672217282353892,
+      "grad_norm": 0.671115517616272,
+      "learning_rate": 0.0002,
+      "loss": 0.8523,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4688384124161345,
+      "grad_norm": 0.4425133168697357,
+      "learning_rate": 0.0002,
+      "loss": 0.8262,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4704550965968798,
+      "grad_norm": 0.5446155071258545,
+      "learning_rate": 0.0002,
+      "loss": 0.8178,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47207178077762507,
+      "grad_norm": 0.603306233882904,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 2920
+    },
+    {
+      "epoch": 0.47368846495837036,
+      "grad_norm": 0.5377997159957886,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4753051491391157,
+      "grad_norm": 0.4931027591228485,
+      "learning_rate": 0.0002,
+      "loss": 0.8075,
+      "step": 2940
+    },
+    {
+      "epoch": 0.476921833319861,
+      "grad_norm": 0.4711960256099701,
+      "learning_rate": 0.0002,
+      "loss": 0.8004,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4785385175006063,
+      "grad_norm": 0.5020492672920227,
+      "learning_rate": 0.0002,
+      "loss": 0.8121,
+      "step": 2960
+    },
+    {
+      "epoch": 0.48015520168135156,
+      "grad_norm": 0.5428946614265442,
+      "learning_rate": 0.0002,
+      "loss": 0.8221,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48177188586209685,
+      "grad_norm": 0.5294089317321777,
+      "learning_rate": 0.0002,
+      "loss": 0.7849,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48338857004284214,
+      "grad_norm": 0.648289144039154,
+      "learning_rate": 0.0002,
+      "loss": 0.8553,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4850052542235874,
+      "grad_norm": 0.47916680574417114,
+      "learning_rate": 0.0002,
+      "loss": 0.7874,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4866219384043327,
+      "grad_norm": 0.43849772214889526,
+      "learning_rate": 0.0002,
+      "loss": 0.8087,
+      "step": 3010
+    },
+    {
+      "epoch": 0.488238622585078,
+      "grad_norm": 0.47007861733436584,
+      "learning_rate": 0.0002,
+      "loss": 0.7662,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4898553067658233,
+      "grad_norm": 0.6314331293106079,
+      "learning_rate": 0.0002,
+      "loss": 0.757,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49147199094656857,
+      "grad_norm": 0.49211493134498596,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49308867512731386,
+      "grad_norm": 0.4537973403930664,
+      "learning_rate": 0.0002,
+      "loss": 0.8335,
+      "step": 3050
+    },
+    {
+      "epoch": 0.49470535930805914,
+      "grad_norm": 0.47326919436454773,
+      "learning_rate": 0.0002,
+      "loss": 0.8095,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4963220434888045,
+      "grad_norm": 0.525874137878418,
+      "learning_rate": 0.0002,
+      "loss": 0.8447,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4979387276695498,
+      "grad_norm": 0.6361091732978821,
+      "learning_rate": 0.0002,
+      "loss": 0.8339,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49955541185029506,
+      "grad_norm": 0.5850642919540405,
+      "learning_rate": 0.0002,
+      "loss": 0.821,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5011720960310403,
+      "grad_norm": 0.47299543023109436,
+      "learning_rate": 0.0002,
+      "loss": 0.8279,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5027887802117856,
+      "grad_norm": 0.473099946975708,
+      "learning_rate": 0.0002,
+      "loss": 0.8681,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5044054643925309,
+      "grad_norm": 0.48186397552490234,
+      "learning_rate": 0.0002,
+      "loss": 0.8223,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5060221485732762,
+      "grad_norm": 0.5015401840209961,
+      "learning_rate": 0.0002,
+      "loss": 0.8292,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5076388327540216,
+      "grad_norm": 0.5617750287055969,
+      "learning_rate": 0.0002,
+      "loss": 0.7692,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5092555169347668,
+      "grad_norm": 0.5169327259063721,
+      "learning_rate": 0.0002,
+      "loss": 0.8708,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5108722011155121,
+      "grad_norm": 0.545657753944397,
+      "learning_rate": 0.0002,
+      "loss": 0.7845,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5124888852962574,
+      "grad_norm": 0.512864351272583,
+      "learning_rate": 0.0002,
+      "loss": 0.799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5141055694770027,
+      "grad_norm": 0.4113546311855316,
+      "learning_rate": 0.0002,
+      "loss": 0.7794,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5157222536577479,
+      "grad_norm": 0.44532445073127747,
+      "learning_rate": 0.0002,
+      "loss": 0.8206,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5173389378384933,
+      "grad_norm": 0.5623497366905212,
+      "learning_rate": 0.0002,
+      "loss": 0.8213,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5189556220192385,
+      "grad_norm": 0.5084741115570068,
+      "learning_rate": 0.0002,
+      "loss": 0.7928,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5205723061999838,
+      "grad_norm": 0.5305403470993042,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5221889903807291,
+      "grad_norm": 0.4708254337310791,
+      "learning_rate": 0.0002,
+      "loss": 0.8139,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5238056745614744,
+      "grad_norm": 0.43827131390571594,
+      "learning_rate": 0.0002,
+      "loss": 0.7639,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5254223587422197,
+      "grad_norm": 0.5630002617835999,
+      "learning_rate": 0.0002,
+      "loss": 0.7993,
+      "step": 3250
+    },
+    {
+      "epoch": 0.527039042922965,
+      "grad_norm": 0.5010961890220642,
+      "learning_rate": 0.0002,
+      "loss": 0.7522,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5286557271037103,
+      "grad_norm": 0.6303122043609619,
+      "learning_rate": 0.0002,
+      "loss": 0.8374,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5302724112844556,
+      "grad_norm": 0.5107331275939941,
+      "learning_rate": 0.0002,
+      "loss": 0.7727,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5318890954652009,
+      "grad_norm": 0.5700443387031555,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5335057796459461,
+      "grad_norm": 0.46296367049217224,
+      "learning_rate": 0.0002,
+      "loss": 0.7776,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5351224638266915,
+      "grad_norm": 0.531568706035614,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5367391480074367,
+      "grad_norm": 0.4686741530895233,
+      "learning_rate": 0.0002,
+      "loss": 0.843,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5383558321881821,
+      "grad_norm": 0.5404331088066101,
+      "learning_rate": 0.0002,
+      "loss": 0.8104,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5399725163689273,
+      "grad_norm": 0.6368790864944458,
+      "learning_rate": 0.0002,
+      "loss": 0.7686,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5415892005496726,
+      "grad_norm": 0.42300888895988464,
+      "learning_rate": 0.0002,
+      "loss": 0.8514,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5432058847304179,
+      "grad_norm": 0.5362542867660522,
+      "learning_rate": 0.0002,
+      "loss": 0.8236,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5448225689111632,
+      "grad_norm": 0.497128963470459,
+      "learning_rate": 0.0002,
+      "loss": 0.858,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5464392530919085,
+      "grad_norm": 0.5006386041641235,
+      "learning_rate": 0.0002,
+      "loss": 0.8519,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5480559372726538,
+      "grad_norm": 0.44136837124824524,
+      "learning_rate": 0.0002,
+      "loss": 0.7867,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5496726214533991,
+      "grad_norm": 0.5897833108901978,
+      "learning_rate": 0.0002,
+      "loss": 0.773,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5512893056341444,
+      "grad_norm": 0.641075611114502,
+      "learning_rate": 0.0002,
+      "loss": 0.8895,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5529059898148897,
+      "grad_norm": 0.7251322269439697,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5545226739956349,
+      "grad_norm": 0.47411349415779114,
+      "learning_rate": 0.0002,
+      "loss": 0.7626,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5561393581763803,
+      "grad_norm": 0.4994310438632965,
+      "learning_rate": 0.0002,
+      "loss": 0.8196,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5577560423571255,
+      "grad_norm": 0.5814438462257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5593727265378708,
+      "grad_norm": 0.6278898119926453,
+      "learning_rate": 0.0002,
+      "loss": 0.8805,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5609894107186161,
+      "grad_norm": 0.46208274364471436,
+      "learning_rate": 0.0002,
+      "loss": 0.813,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5626060948993614,
+      "grad_norm": 0.5718930959701538,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5642227790801067,
+      "grad_norm": 0.48178744316101074,
+      "learning_rate": 0.0002,
+      "loss": 0.8152,
+      "step": 3490
+    },
+    {
+      "epoch": 0.565839463260852,
+      "grad_norm": 0.47336965799331665,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5674561474415973,
+      "grad_norm": 0.43442684412002563,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5690728316223426,
+      "grad_norm": 0.6463358998298645,
+      "learning_rate": 0.0002,
+      "loss": 0.7564,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5706895158030879,
+      "grad_norm": 0.5286486744880676,
+      "learning_rate": 0.0002,
+      "loss": 0.836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5723061999838331,
+      "grad_norm": 0.5405499935150146,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5739228841645785,
+      "grad_norm": 0.6654391884803772,
+      "learning_rate": 0.0002,
+      "loss": 0.7614,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.5081980228424072,
+      "learning_rate": 0.0002,
+      "loss": 0.7803,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5771562525260691,
+      "grad_norm": 0.48978179693222046,
+      "learning_rate": 0.0002,
+      "loss": 0.7753,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5787729367068143,
+      "grad_norm": 0.5840612053871155,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5803896208875596,
+      "grad_norm": 0.5235261917114258,
+      "learning_rate": 0.0002,
+      "loss": 0.8937,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5820063050683049,
+      "grad_norm": 0.5672075748443604,
+      "learning_rate": 0.0002,
+      "loss": 0.7894,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5836229892490502,
+      "grad_norm": 0.5613429546356201,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5852396734297954,
+      "grad_norm": 0.4032273590564728,
+      "learning_rate": 0.0002,
+      "loss": 0.8274,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5868563576105408,
+      "grad_norm": 0.49559324979782104,
+      "learning_rate": 0.0002,
+      "loss": 0.8421,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5884730417912861,
+      "grad_norm": 0.6895697712898254,
+      "learning_rate": 0.0002,
+      "loss": 0.8332,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5900897259720314,
+      "grad_norm": 0.4750136435031891,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5917064101527767,
+      "grad_norm": 0.5176819562911987,
+      "learning_rate": 0.0002,
+      "loss": 0.8219,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5933230943335219,
+      "grad_norm": 0.5817760229110718,
+      "learning_rate": 0.0002,
+      "loss": 0.8151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5949397785142673,
+      "grad_norm": 0.6064626574516296,
+      "learning_rate": 0.0002,
+      "loss": 0.7823,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5965564626950125,
+      "grad_norm": 0.6728700995445251,
+      "learning_rate": 0.0002,
+      "loss": 0.8422,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5981731468757578,
+      "grad_norm": 0.609305202960968,
+      "learning_rate": 0.0002,
+      "loss": 0.7679,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5997898310565031,
+      "grad_norm": 0.4615488350391388,
+      "learning_rate": 0.0002,
+      "loss": 0.8048,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6014065152372484,
+      "grad_norm": 2.0531179904937744,
+      "learning_rate": 0.0002,
+      "loss": 0.8214,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6030231994179936,
+      "grad_norm": 0.5091132521629333,
+      "learning_rate": 0.0002,
+      "loss": 0.8158,
+      "step": 3730
+    },
+    {
+      "epoch": 0.604639883598739,
+      "grad_norm": 0.5951124429702759,
+      "learning_rate": 0.0002,
+      "loss": 0.7833,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6062565677794842,
+      "grad_norm": 0.5870208144187927,
+      "learning_rate": 0.0002,
+      "loss": 0.7784,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6078732519602296,
+      "grad_norm": 0.6254619359970093,
+      "learning_rate": 0.0002,
+      "loss": 0.8044,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6094899361409749,
+      "grad_norm": 0.5577626824378967,
+      "learning_rate": 0.0002,
+      "loss": 0.7868,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6111066203217201,
+      "grad_norm": 0.5004405379295349,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6127233045024655,
+      "grad_norm": 0.5527383685112,
+      "learning_rate": 0.0002,
+      "loss": 0.8092,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6143399886832107,
+      "grad_norm": 0.49116113781929016,
+      "learning_rate": 0.0002,
+      "loss": 0.8036,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6159566728639561,
+      "grad_norm": 0.5299299359321594,
+      "learning_rate": 0.0002,
+      "loss": 0.8352,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6175733570447013,
+      "grad_norm": 0.464897483587265,
+      "learning_rate": 0.0002,
+      "loss": 0.7737,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6191900412254466,
+      "grad_norm": 0.6505740880966187,
+      "learning_rate": 0.0002,
+      "loss": 0.7923,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6208067254061919,
+      "grad_norm": 0.5512559413909912,
+      "learning_rate": 0.0002,
+      "loss": 0.8123,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6224234095869372,
+      "grad_norm": 0.49427518248558044,
+      "learning_rate": 0.0002,
+      "loss": 0.8856,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6240400937676824,
+      "grad_norm": 0.3839147090911865,
+      "learning_rate": 0.0002,
+      "loss": 0.7751,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6256567779484278,
+      "grad_norm": 0.5760218501091003,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6272734621291731,
+      "grad_norm": 0.7226507067680359,
+      "learning_rate": 0.0002,
+      "loss": 0.7836,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6288901463099184,
+      "grad_norm": 0.676781415939331,
+      "learning_rate": 0.0002,
+      "loss": 0.8244,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6305068304906637,
+      "grad_norm": 0.4284018278121948,
+      "learning_rate": 0.0002,
+      "loss": 0.8239,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6321235146714089,
+      "grad_norm": 0.5060628056526184,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6337401988521543,
+      "grad_norm": 0.5524522066116333,
+      "learning_rate": 0.0002,
+      "loss": 0.8089,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6353568830328995,
+      "grad_norm": 0.6099881529808044,
+      "learning_rate": 0.0002,
+      "loss": 0.8276,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6369735672136448,
+      "grad_norm": 0.43155938386917114,
+      "learning_rate": 0.0002,
+      "loss": 0.809,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6385902513943901,
+      "grad_norm": 0.6427084803581238,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6402069355751354,
+      "grad_norm": 0.541220486164093,
+      "learning_rate": 0.0002,
+      "loss": 0.8368,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6418236197558806,
+      "grad_norm": 0.5414294600486755,
+      "learning_rate": 0.0002,
+      "loss": 0.8539,
+      "step": 3970
+    },
+    {
+      "epoch": 0.643440303936626,
+      "grad_norm": 0.46344003081321716,
+      "learning_rate": 0.0002,
+      "loss": 0.7996,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6450569881173712,
+      "grad_norm": 0.45209285616874695,
+      "learning_rate": 0.0002,
+      "loss": 0.7474,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6466736722981166,
+      "grad_norm": 0.5417284369468689,
+      "learning_rate": 0.0002,
+      "loss": 0.8202,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6482903564788619,
+      "grad_norm": 0.7995685935020447,
+      "learning_rate": 0.0002,
+      "loss": 0.7563,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6499070406596071,
+      "grad_norm": 0.6384002566337585,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6515237248403525,
+      "grad_norm": 0.4472815692424774,
+      "learning_rate": 0.0002,
+      "loss": 0.732,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6531404090210977,
+      "grad_norm": 0.6834294199943542,
+      "learning_rate": 0.0002,
+      "loss": 0.8071,
+      "step": 4040
+    },
+    {
+      "epoch": 0.654757093201843,
+      "grad_norm": 0.4612339735031128,
+      "learning_rate": 0.0002,
+      "loss": 0.7812,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6563737773825883,
+      "grad_norm": 0.9266576170921326,
+      "learning_rate": 0.0002,
+      "loss": 0.8141,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6579904615633336,
+      "grad_norm": 0.4470861852169037,
+      "learning_rate": 0.0002,
+      "loss": 0.7991,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6596071457440789,
+      "grad_norm": 0.45544925332069397,
+      "learning_rate": 0.0002,
+      "loss": 0.8293,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6612238299248242,
+      "grad_norm": 0.6144481301307678,
+      "learning_rate": 0.0002,
+      "loss": 0.8455,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6628405141055694,
+      "grad_norm": 0.5936288237571716,
+      "learning_rate": 0.0002,
+      "loss": 0.7877,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6644571982863148,
+      "grad_norm": 0.4822963774204254,
+      "learning_rate": 0.0002,
+      "loss": 0.7617,
+      "step": 4110
+    },
+    {
+      "epoch": 0.66607388246706,
+      "grad_norm": 0.48432496190071106,
+      "learning_rate": 0.0002,
+      "loss": 0.7997,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6676905666478054,
+      "grad_norm": 0.4901607930660248,
+      "learning_rate": 0.0002,
+      "loss": 0.8404,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6693072508285507,
+      "grad_norm": 0.5018393397331238,
+      "learning_rate": 0.0002,
+      "loss": 0.8085,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6709239350092959,
+      "grad_norm": 0.6946378946304321,
+      "learning_rate": 0.0002,
+      "loss": 0.8065,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6725406191900413,
+      "grad_norm": 0.5997390747070312,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.6738849878311157,
+      "learning_rate": 0.0002,
+      "loss": 0.8268,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6757739875515318,
+      "grad_norm": 0.6110581159591675,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6773906717322771,
+      "grad_norm": 0.5703322291374207,
+      "learning_rate": 0.0002,
+      "loss": 0.8043,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6790073559130224,
+      "grad_norm": 0.4686066210269928,
+      "learning_rate": 0.0002,
+      "loss": 0.8099,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6806240400937676,
+      "grad_norm": 0.6394643783569336,
+      "learning_rate": 0.0002,
+      "loss": 0.8441,
+      "step": 4210
+    },
+    {
+      "epoch": 0.682240724274513,
+      "grad_norm": 0.5454841256141663,
+      "learning_rate": 0.0002,
+      "loss": 0.8011,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6838574084552582,
+      "grad_norm": 0.4859732985496521,
+      "learning_rate": 0.0002,
+      "loss": 0.8307,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6854740926360036,
+      "grad_norm": 0.5544065833091736,
+      "learning_rate": 0.0002,
+      "loss": 0.8161,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6870907768167488,
+      "grad_norm": 0.4902505576610565,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6887074609974941,
+      "grad_norm": 0.4768051505088806,
+      "learning_rate": 0.0002,
+      "loss": 0.7977,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6903241451782395,
+      "grad_norm": 0.49982190132141113,
+      "learning_rate": 0.0002,
+      "loss": 0.7539,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6919408293589847,
+      "grad_norm": 0.6351838111877441,
+      "learning_rate": 0.0002,
+      "loss": 0.7353,
+      "step": 4280
+    },
+    {
+      "epoch": 0.69355751353973,
+      "grad_norm": 0.5647561550140381,
+      "learning_rate": 0.0002,
+      "loss": 0.7664,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6951741977204753,
+      "grad_norm": 0.5340486764907837,
+      "learning_rate": 0.0002,
+      "loss": 0.7618,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6967908819012206,
+      "grad_norm": 0.5649092793464661,
+      "learning_rate": 0.0002,
+      "loss": 0.8526,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6984075660819659,
+      "grad_norm": 0.6183916926383972,
+      "learning_rate": 0.0002,
+      "loss": 0.8246,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7000242502627112,
+      "grad_norm": 0.6154509782791138,
+      "learning_rate": 0.0002,
+      "loss": 0.792,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7016409344434564,
+      "grad_norm": 0.5156264305114746,
+      "learning_rate": 0.0002,
+      "loss": 0.8397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7032576186242018,
+      "grad_norm": 0.562171459197998,
+      "learning_rate": 0.0002,
+      "loss": 0.8512,
+      "step": 4350
+    },
+    {
+      "epoch": 0.704874302804947,
+      "grad_norm": 0.4949502646923065,
+      "learning_rate": 0.0002,
+      "loss": 0.7882,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7064909869856923,
+      "grad_norm": 0.5171684622764587,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7081076711664377,
+      "grad_norm": 0.6198443174362183,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7097243553471829,
+      "grad_norm": 0.5802276134490967,
+      "learning_rate": 0.0002,
+      "loss": 0.7606,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7113410395279283,
+      "grad_norm": 0.41096967458724976,
+      "learning_rate": 0.0002,
+      "loss": 0.8797,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7129577237086735,
+      "grad_norm": 0.4397392272949219,
+      "learning_rate": 0.0002,
+      "loss": 0.805,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7145744078894188,
+      "grad_norm": 0.45228442549705505,
+      "learning_rate": 0.0002,
+      "loss": 0.7651,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7161910920701641,
+      "grad_norm": 0.4839673936367035,
+      "learning_rate": 0.0002,
+      "loss": 0.7938,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7178077762509094,
+      "grad_norm": 0.6140755414962769,
+      "learning_rate": 0.0002,
+      "loss": 0.8362,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 0.6841378808021545,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7210411446124,
+      "grad_norm": 0.6664239168167114,
+      "learning_rate": 0.0002,
+      "loss": 0.8177,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7226578287931452,
+      "grad_norm": 0.47552719712257385,
+      "learning_rate": 0.0002,
+      "loss": 0.7983,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7242745129738906,
+      "grad_norm": 0.6649776101112366,
+      "learning_rate": 0.0002,
+      "loss": 0.8982,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7258911971546358,
+      "grad_norm": 0.5159541964530945,
+      "learning_rate": 0.0002,
+      "loss": 0.8074,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7275078813353811,
+      "grad_norm": 0.6693112850189209,
+      "learning_rate": 0.0002,
+      "loss": 0.7786,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7291245655161265,
+      "grad_norm": 0.48870977759361267,
+      "learning_rate": 0.0002,
+      "loss": 0.8655,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7307412496968717,
+      "grad_norm": 0.4857887923717499,
+      "learning_rate": 0.0002,
+      "loss": 0.7337,
+      "step": 4520
+    },
+    {
+      "epoch": 0.732357933877617,
+      "grad_norm": 0.5515662431716919,
+      "learning_rate": 0.0002,
+      "loss": 0.8026,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7339746180583623,
+      "grad_norm": 0.6292222738265991,
+      "learning_rate": 0.0002,
+      "loss": 0.8031,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7355913022391076,
+      "grad_norm": 0.48265689611434937,
+      "learning_rate": 0.0002,
+      "loss": 0.7749,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7372079864198529,
+      "grad_norm": 0.8044266104698181,
+      "learning_rate": 0.0002,
+      "loss": 0.8499,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7388246706005982,
+      "grad_norm": 0.6111769676208496,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7404413547813434,
+      "grad_norm": 0.5229553580284119,
+      "learning_rate": 0.0002,
+      "loss": 0.7291,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7420580389620888,
+      "grad_norm": 0.6054152250289917,
+      "learning_rate": 0.0002,
+      "loss": 0.8038,
+      "step": 4590
+    },
+    {
+      "epoch": 0.743674723142834,
+      "grad_norm": 0.5574966669082642,
+      "learning_rate": 0.0002,
+      "loss": 0.8169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7452914073235793,
+      "grad_norm": 0.5395817160606384,
+      "learning_rate": 0.0002,
+      "loss": 0.8439,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7469080915043246,
+      "grad_norm": 0.7116472721099854,
+      "learning_rate": 0.0002,
+      "loss": 0.8495,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7485247756850699,
+      "grad_norm": 0.5618700981140137,
+      "learning_rate": 0.0002,
+      "loss": 0.7743,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7501414598658153,
+      "grad_norm": 0.5802770853042603,
+      "learning_rate": 0.0002,
+      "loss": 0.7744,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7517581440465605,
+      "grad_norm": 0.5690428018569946,
+      "learning_rate": 0.0002,
+      "loss": 0.7924,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7533748282273058,
+      "grad_norm": 0.4813360273838043,
+      "learning_rate": 0.0002,
+      "loss": 0.8017,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7549915124080511,
+      "grad_norm": 0.5434042811393738,
+      "learning_rate": 0.0002,
+      "loss": 0.8108,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7566081965887964,
+      "grad_norm": 0.5502099990844727,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7582248807695416,
+      "grad_norm": 0.6020621061325073,
+      "learning_rate": 0.0002,
+      "loss": 0.8598,
+      "step": 4690
+    },
+    {
+      "epoch": 0.759841564950287,
+      "grad_norm": 0.4922301471233368,
+      "learning_rate": 0.0002,
+      "loss": 0.7937,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7614582491310322,
+      "grad_norm": 0.6492828726768494,
+      "learning_rate": 0.0002,
+      "loss": 0.788,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7630749333117776,
+      "grad_norm": 0.4865580201148987,
+      "learning_rate": 0.0002,
+      "loss": 0.8313,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7646916174925228,
+      "grad_norm": 0.5971422791481018,
+      "learning_rate": 0.0002,
+      "loss": 0.7966,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7663083016732681,
+      "grad_norm": 0.6832674145698547,
+      "learning_rate": 0.0002,
+      "loss": 0.8298,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7679249858540134,
+      "grad_norm": 0.500908613204956,
+      "learning_rate": 0.0002,
+      "loss": 0.8156,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7695416700347587,
+      "grad_norm": 0.6112465858459473,
+      "learning_rate": 0.0002,
+      "loss": 0.8383,
+      "step": 4760
+    },
+    {
+      "epoch": 0.771158354215504,
+      "grad_norm": 0.5753506422042847,
+      "learning_rate": 0.0002,
+      "loss": 0.76,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7727750383962493,
+      "grad_norm": 0.6529405117034912,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7743917225769946,
+      "grad_norm": 0.5916843414306641,
+      "learning_rate": 0.0002,
+      "loss": 0.8171,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7760084067577399,
+      "grad_norm": 0.4821224510669708,
+      "learning_rate": 0.0002,
+      "loss": 0.83,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7776250909384852,
+      "grad_norm": 0.5532580018043518,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7792417751192304,
+      "grad_norm": 0.4604877233505249,
+      "learning_rate": 0.0002,
+      "loss": 0.7363,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7808584592999758,
+      "grad_norm": 0.5009613037109375,
+      "learning_rate": 0.0002,
+      "loss": 0.7506,
+      "step": 4830
+    },
+    {
+      "epoch": 0.782475143480721,
+      "grad_norm": 0.6448560357093811,
+      "learning_rate": 0.0002,
+      "loss": 0.7863,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7840918276614663,
+      "grad_norm": 0.44327953457832336,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7857085118422116,
+      "grad_norm": 0.5355411171913147,
+      "learning_rate": 0.0002,
+      "loss": 0.7925,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7873251960229569,
+      "grad_norm": 0.5635677576065063,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7889418802037023,
+      "grad_norm": 0.5417491793632507,
+      "learning_rate": 0.0002,
+      "loss": 0.7931,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7905585643844475,
+      "grad_norm": 0.4567430913448334,
+      "learning_rate": 0.0002,
+      "loss": 0.7819,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7921752485651928,
+      "grad_norm": 0.44651296734809875,
+      "learning_rate": 0.0002,
+      "loss": 0.8454,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7937919327459381,
+      "grad_norm": 0.5741217136383057,
+      "learning_rate": 0.0002,
+      "loss": 0.7959,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7954086169266834,
+      "grad_norm": 0.6605045199394226,
+      "learning_rate": 0.0002,
+      "loss": 0.8093,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7970253011074286,
+      "grad_norm": 0.5126531720161438,
+      "learning_rate": 0.0002,
+      "loss": 0.77,
+      "step": 4930
+    },
+    {
+      "epoch": 0.798641985288174,
+      "grad_norm": 0.513648271560669,
+      "learning_rate": 0.0002,
+      "loss": 0.7793,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8002586694689192,
+      "grad_norm": 0.5350404381752014,
+      "learning_rate": 0.0002,
+      "loss": 0.8314,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8018753536496646,
+      "grad_norm": 0.5731674432754517,
+      "learning_rate": 0.0002,
+      "loss": 0.7649,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8034920378304098,
+      "grad_norm": 0.5974258184432983,
+      "learning_rate": 0.0002,
+      "loss": 0.8572,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8051087220111551,
+      "grad_norm": 0.8774799704551697,
+      "learning_rate": 0.0002,
+      "loss": 0.7972,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8067254061919004,
+      "grad_norm": 0.5994430184364319,
+      "learning_rate": 0.0002,
+      "loss": 0.7899,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8083420903726457,
+      "grad_norm": 0.4894903004169464,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5000
+    },
+    {
+      "epoch": 0.809958774553391,
+      "grad_norm": 0.5218459367752075,
+      "learning_rate": 0.0002,
+      "loss": 0.78,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8115754587341363,
+      "grad_norm": 0.5232468843460083,
+      "learning_rate": 0.0002,
+      "loss": 0.817,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8131921429148816,
+      "grad_norm": 0.44358372688293457,
+      "learning_rate": 0.0002,
+      "loss": 0.7704,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8148088270956269,
+      "grad_norm": 0.6202037334442139,
+      "learning_rate": 0.0002,
+      "loss": 0.785,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8164255112763722,
+      "grad_norm": 0.7721474170684814,
+      "learning_rate": 0.0002,
+      "loss": 0.7351,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8180421954571174,
+      "grad_norm": 0.5568501353263855,
+      "learning_rate": 0.0002,
+      "loss": 0.8297,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8196588796378628,
+      "grad_norm": 0.49148809909820557,
+      "learning_rate": 0.0002,
+      "loss": 0.7733,
+      "step": 5070
+    },
+    {
+      "epoch": 0.821275563818608,
+      "grad_norm": 0.4956012964248657,
+      "learning_rate": 0.0002,
+      "loss": 0.8054,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8228922479993533,
+      "grad_norm": 0.6078833937644958,
+      "learning_rate": 0.0002,
+      "loss": 0.8201,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8245089321800986,
+      "grad_norm": 0.46906954050064087,
+      "learning_rate": 0.0002,
+      "loss": 0.828,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8261256163608439,
+      "grad_norm": 0.50812166929245,
+      "learning_rate": 0.0002,
+      "loss": 0.7703,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8277423005415891,
+      "grad_norm": 0.5319661498069763,
+      "learning_rate": 0.0002,
+      "loss": 0.8243,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8293589847223345,
+      "grad_norm": 0.4949689209461212,
+      "learning_rate": 0.0002,
+      "loss": 0.7798,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8309756689030798,
+      "grad_norm": 0.5151591300964355,
+      "learning_rate": 0.0002,
+      "loss": 0.7428,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8325923530838251,
+      "grad_norm": 0.5530214309692383,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8342090372645704,
+      "grad_norm": 0.6297410130500793,
+      "learning_rate": 0.0002,
+      "loss": 0.8251,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8358257214453156,
+      "grad_norm": 0.5466840267181396,
+      "learning_rate": 0.0002,
+      "loss": 0.8067,
+      "step": 5170
+    },
+    {
+      "epoch": 0.837442405626061,
+      "grad_norm": 0.652913510799408,
+      "learning_rate": 0.0002,
+      "loss": 0.7875,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8390590898068062,
+      "grad_norm": 0.5811293125152588,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8406757739875516,
+      "grad_norm": 0.5109550952911377,
+      "learning_rate": 0.0002,
+      "loss": 0.7412,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8422924581682968,
+      "grad_norm": 0.4551706612110138,
+      "learning_rate": 0.0002,
+      "loss": 0.8077,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8439091423490421,
+      "grad_norm": 0.5813754200935364,
+      "learning_rate": 0.0002,
+      "loss": 0.7827,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8455258265297874,
+      "grad_norm": 0.5856947898864746,
+      "learning_rate": 0.0002,
+      "loss": 0.802,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8471425107105327,
+      "grad_norm": 0.5482739210128784,
+      "learning_rate": 0.0002,
+      "loss": 0.7957,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8487591948912779,
+      "grad_norm": 0.49023720622062683,
+      "learning_rate": 0.0002,
+      "loss": 0.8295,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8503758790720233,
+      "grad_norm": 0.49472475051879883,
+      "learning_rate": 0.0002,
+      "loss": 0.8022,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8519925632527686,
+      "grad_norm": 0.5490226745605469,
+      "learning_rate": 0.0002,
+      "loss": 0.8001,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8536092474335139,
+      "grad_norm": 0.5340665578842163,
+      "learning_rate": 0.0002,
+      "loss": 0.8333,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8552259316142592,
+      "grad_norm": 0.5962483882904053,
+      "learning_rate": 0.0002,
+      "loss": 0.8277,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8568426157950044,
+      "grad_norm": 0.586358368396759,
+      "learning_rate": 0.0002,
+      "loss": 0.8765,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8584592999757498,
+      "grad_norm": 0.49120277166366577,
+      "learning_rate": 0.0002,
+      "loss": 0.7831,
+      "step": 5310
+    },
+    {
+      "epoch": 0.860075984156495,
+      "grad_norm": 0.5887332558631897,
+      "learning_rate": 0.0002,
+      "loss": 0.8162,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8616926683372403,
+      "grad_norm": 0.42496153712272644,
+      "learning_rate": 0.0002,
+      "loss": 0.7464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.5489874482154846,
+      "learning_rate": 0.0002,
+      "loss": 0.7905,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8649260366987309,
+      "grad_norm": 0.5850813984870911,
+      "learning_rate": 0.0002,
+      "loss": 0.7958,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8665427208794761,
+      "grad_norm": 0.517487108707428,
+      "learning_rate": 0.0002,
+      "loss": 0.7642,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8681594050602215,
+      "grad_norm": 0.5339142680168152,
+      "learning_rate": 0.0002,
+      "loss": 0.7801,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8697760892409668,
+      "grad_norm": 0.6236387491226196,
+      "learning_rate": 0.0002,
+      "loss": 0.818,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8713927734217121,
+      "grad_norm": 0.5752192735671997,
+      "learning_rate": 0.0002,
+      "loss": 0.7708,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8730094576024574,
+      "grad_norm": 0.6724614500999451,
+      "learning_rate": 0.0002,
+      "loss": 0.8542,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8746261417832026,
+      "grad_norm": 0.5280613303184509,
+      "learning_rate": 0.0002,
+      "loss": 0.7581,
+      "step": 5410
+    },
+    {
+      "epoch": 0.876242825963948,
+      "grad_norm": 0.44033288955688477,
+      "learning_rate": 0.0002,
+      "loss": 0.8231,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8778595101446932,
+      "grad_norm": 0.5199708342552185,
+      "learning_rate": 0.0002,
+      "loss": 0.8839,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8794761943254386,
+      "grad_norm": 0.46778348088264465,
+      "learning_rate": 0.0002,
+      "loss": 0.7852,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8810928785061838,
+      "grad_norm": 0.4657754898071289,
+      "learning_rate": 0.0002,
+      "loss": 0.7834,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8827095626869291,
+      "grad_norm": 0.5472902655601501,
+      "learning_rate": 0.0002,
+      "loss": 0.7799,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8843262468676744,
+      "grad_norm": 0.4876766800880432,
+      "learning_rate": 0.0002,
+      "loss": 0.8253,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8859429310484197,
+      "grad_norm": 0.5057248473167419,
+      "learning_rate": 0.0002,
+      "loss": 0.7906,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8875596152291649,
+      "grad_norm": 0.4637320637702942,
+      "learning_rate": 0.0002,
+      "loss": 0.8124,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8891762994099103,
+      "grad_norm": 0.471955806016922,
+      "learning_rate": 0.0002,
+      "loss": 0.781,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8907929835906556,
+      "grad_norm": 0.5209813714027405,
+      "learning_rate": 0.0002,
+      "loss": 0.8057,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8924096677714008,
+      "grad_norm": 0.6213834285736084,
+      "learning_rate": 0.0002,
+      "loss": 0.8106,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8940263519521462,
+      "grad_norm": 0.5215408205986023,
+      "learning_rate": 0.0002,
+      "loss": 0.7787,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8956430361328914,
+      "grad_norm": 0.580478310585022,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8972597203136368,
+      "grad_norm": 0.49102169275283813,
+      "learning_rate": 0.0002,
+      "loss": 0.8371,
+      "step": 5550
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.6043479442596436,
+      "learning_rate": 0.0002,
+      "loss": 0.7806,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9004930886751273,
+      "grad_norm": 0.5636463165283203,
+      "learning_rate": 0.0002,
+      "loss": 0.7754,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9021097728558726,
+      "grad_norm": 0.5620124340057373,
+      "learning_rate": 0.0002,
+      "loss": 0.8145,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9037264570366179,
+      "grad_norm": 0.5206354856491089,
+      "learning_rate": 0.0002,
+      "loss": 0.8083,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9053431412173631,
+      "grad_norm": 0.5798229575157166,
+      "learning_rate": 0.0002,
+      "loss": 0.8557,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9069598253981085,
+      "grad_norm": 0.6428212523460388,
+      "learning_rate": 0.0002,
+      "loss": 0.8097,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9085765095788537,
+      "grad_norm": 0.48064687848091125,
+      "learning_rate": 0.0002,
+      "loss": 0.7839,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9101931937595991,
+      "grad_norm": 0.6347860097885132,
+      "learning_rate": 0.0002,
+      "loss": 0.8343,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9118098779403444,
+      "grad_norm": 0.5353913307189941,
+      "learning_rate": 0.0002,
+      "loss": 0.851,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9134265621210896,
+      "grad_norm": 0.5323944091796875,
+      "learning_rate": 0.0002,
+      "loss": 0.7736,
+      "step": 5650
+    },
+    {
+      "epoch": 0.915043246301835,
+      "grad_norm": 0.5261843204498291,
+      "learning_rate": 0.0002,
+      "loss": 0.8393,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9166599304825802,
+      "grad_norm": 0.5451326966285706,
+      "learning_rate": 0.0002,
+      "loss": 0.7355,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9182766146633256,
+      "grad_norm": 0.5183324217796326,
+      "learning_rate": 0.0002,
+      "loss": 0.8012,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9198932988440708,
+      "grad_norm": 0.47229018807411194,
+      "learning_rate": 0.0002,
+      "loss": 0.7659,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9215099830248161,
+      "grad_norm": 0.49180513620376587,
+      "learning_rate": 0.0002,
+      "loss": 0.7757,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9231266672055614,
+      "grad_norm": 0.5419785380363464,
+      "learning_rate": 0.0002,
+      "loss": 0.8735,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9247433513863067,
+      "grad_norm": 0.5408698916435242,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9263600355670519,
+      "grad_norm": 0.5286232829093933,
+      "learning_rate": 0.0002,
+      "loss": 0.7701,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9279767197477973,
+      "grad_norm": 0.7539758086204529,
+      "learning_rate": 0.0002,
+      "loss": 0.8242,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9295934039285425,
+      "grad_norm": 0.5166944861412048,
+      "learning_rate": 0.0002,
+      "loss": 0.8118,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9312100881092878,
+      "grad_norm": 0.6601425409317017,
+      "learning_rate": 0.0002,
+      "loss": 0.783,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9328267722900332,
+      "grad_norm": 0.5029960870742798,
+      "learning_rate": 0.0002,
+      "loss": 0.7873,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9344434564707784,
+      "grad_norm": 0.4926645755767822,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9360601406515238,
+      "grad_norm": 0.5739615559577942,
+      "learning_rate": 0.0002,
+      "loss": 0.8174,
+      "step": 5790
+    },
+    {
+      "epoch": 0.937676824832269,
+      "grad_norm": 0.5058279037475586,
+      "learning_rate": 0.0002,
+      "loss": 0.8037,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9392935090130143,
+      "grad_norm": 0.5260962247848511,
+      "learning_rate": 0.0002,
+      "loss": 0.8537,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9409101931937596,
+      "grad_norm": 0.5768588185310364,
+      "learning_rate": 0.0002,
+      "loss": 0.7486,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9425268773745049,
+      "grad_norm": 0.5170126557350159,
+      "learning_rate": 0.0002,
+      "loss": 0.8215,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9441435615552501,
+      "grad_norm": 0.5745864510536194,
+      "learning_rate": 0.0002,
+      "loss": 0.7422,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9457602457359955,
+      "grad_norm": 0.5551357865333557,
+      "learning_rate": 0.0002,
+      "loss": 0.7824,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9473769299167407,
+      "grad_norm": 0.5776078701019287,
+      "learning_rate": 0.0002,
+      "loss": 0.8529,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9489936140974861,
+      "grad_norm": 0.5340062379837036,
+      "learning_rate": 0.0002,
+      "loss": 0.8527,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9506102982782314,
+      "grad_norm": 0.6447290182113647,
+      "learning_rate": 0.0002,
+      "loss": 0.8217,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9522269824589766,
+      "grad_norm": 0.5123815536499023,
+      "learning_rate": 0.0002,
+      "loss": 0.7945,
+      "step": 5890
+    },
+    {
+      "epoch": 0.953843666639722,
+      "grad_norm": 0.48547613620758057,
+      "learning_rate": 0.0002,
+      "loss": 0.8209,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9554603508204672,
+      "grad_norm": 0.5791414976119995,
+      "learning_rate": 0.0002,
+      "loss": 0.7896,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9570770350012126,
+      "grad_norm": 0.6195011734962463,
+      "learning_rate": 0.0002,
+      "loss": 0.8408,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9586937191819578,
+      "grad_norm": 0.6323803067207336,
+      "learning_rate": 0.0002,
+      "loss": 0.7805,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9603104033627031,
+      "grad_norm": 0.45552879571914673,
+      "learning_rate": 0.0002,
+      "loss": 0.8484,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9619270875434484,
+      "grad_norm": 0.5796473622322083,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9635437717241937,
+      "grad_norm": 0.647261381149292,
+      "learning_rate": 0.0002,
+      "loss": 0.7672,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9651604559049389,
+      "grad_norm": 0.5487682819366455,
+      "learning_rate": 0.0002,
+      "loss": 0.8086,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9667771400856843,
+      "grad_norm": 0.5743663907051086,
+      "learning_rate": 0.0002,
+      "loss": 0.7973,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9683938242664295,
+      "grad_norm": 0.5470591187477112,
+      "learning_rate": 0.0002,
+      "loss": 0.8153,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9700105084471748,
+      "grad_norm": 0.5901660323143005,
+      "learning_rate": 0.0002,
+      "loss": 0.8119,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9716271926279202,
+      "grad_norm": 0.6544759273529053,
+      "learning_rate": 0.0002,
+      "loss": 0.8147,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9732438768086654,
+      "grad_norm": 0.6288470029830933,
+      "learning_rate": 0.0002,
+      "loss": 0.7536,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9748605609894108,
+      "grad_norm": 0.673153817653656,
+      "learning_rate": 0.0002,
+      "loss": 0.7989,
+      "step": 6030
+    },
+    {
+      "epoch": 0.976477245170156,
+      "grad_norm": 0.42854753136634827,
+      "learning_rate": 0.0002,
+      "loss": 0.7556,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9780939293509013,
+      "grad_norm": 0.5227066278457642,
+      "learning_rate": 0.0002,
+      "loss": 0.8006,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9797106135316466,
+      "grad_norm": 0.5372416973114014,
+      "learning_rate": 0.0002,
+      "loss": 0.795,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9813272977123919,
+      "grad_norm": 0.6026402115821838,
+      "learning_rate": 0.0002,
+      "loss": 0.7591,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9829439818931371,
+      "grad_norm": 0.49547791481018066,
+      "learning_rate": 0.0002,
+      "loss": 0.8347,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9845606660738825,
+      "grad_norm": 0.4641951322555542,
+      "learning_rate": 0.0002,
+      "loss": 0.7722,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9861773502546277,
+      "grad_norm": 0.5818535089492798,
+      "learning_rate": 0.0002,
+      "loss": 0.8125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9877940344353731,
+      "grad_norm": 0.63955157995224,
+      "learning_rate": 0.0002,
+      "loss": 0.81,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9894107186161183,
+      "grad_norm": 0.5649438500404358,
+      "learning_rate": 0.0002,
+      "loss": 0.7547,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9910274027968636,
+      "grad_norm": 0.5290433168411255,
+      "learning_rate": 0.0002,
+      "loss": 0.7861,
+      "step": 6130
+    },
+    {
+      "epoch": 0.992644086977609,
+      "grad_norm": 0.6399374008178711,
+      "learning_rate": 0.0002,
+      "loss": 0.8109,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9942607711583542,
+      "grad_norm": 0.6736576557159424,
+      "learning_rate": 0.0002,
+      "loss": 0.8373,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9958774553390995,
+      "grad_norm": 0.515420138835907,
+      "learning_rate": 0.0002,
+      "loss": 0.7915,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9974941395198448,
+      "grad_norm": 0.562677800655365,
+      "learning_rate": 0.0002,
+      "loss": 0.8032,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9991108237005901,
+      "grad_norm": 0.7113858461380005,
+      "learning_rate": 0.0002,
+      "loss": 0.8187,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9999191657909627,
+      "eval_loss": 1.0871200561523438,
+      "eval_runtime": 122.2071,
+      "eval_samples_per_second": 5.998,
+      "eval_steps_per_second": 0.753,
+      "step": 6185
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 49480,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.8625109010454938e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/training_args.bin b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..155b12fa9acbc6e71dba75c92bfa79e152397ebf
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28694d5564a2b5c7d6881d4ba2af103356aa22489d2c22768ebbe47283c0f4a1
+size 5560
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/special_tokens_map.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43be96621d147110fb8a18b5776ec6e38516127
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/special_tokens_map.json
@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/tokenizer.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..86a33946b0c77216d2cce91bb28c8fada4a5e80b
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/tokenizer_config.json b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..061e40d9db3253624f86e8e364c15ef546527c9d
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/tokenizer_config.json
@@ -0,0 +1,2063 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/training_args.bin b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..155b12fa9acbc6e71dba75c92bfa79e152397ebf
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28694d5564a2b5c7d6881d4ba2af103356aa22489d2c22768ebbe47283c0f4a1
+size 5560
diff --git a/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/training_log.jsonl b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/training_log.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..bf2bb281c11c4ec2a30943e797dbce81ac45baed
--- /dev/null
+++ b/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/training_log.jsonl
@@ -0,0 +1,8 @@
+{"epoch": 0.9999191657909627, "step": 6185, "epoch_duration": 16870.920749664307, "total_accumulated_duration": 16870.920749664307, "gpu_info": {"GPU_0": "NVIDIA A100-PCIE-40GB"}, "memory_usage": {"avg_memory_usage": {"GPU_0": 7672.4541015625}, "peak_memory_usage": {"GPU_0": 9688.99365234375}, "avg_memory_reserved": {"GPU_0": 10406.0}, "peak_memory_reserved": {"GPU_0": 10406.0}, "total_memory": {"GPU_0": 40444.375}}, "best_checkpoint_path": "N/A", "params": {"epochs": 8, "batch_size": 1, "learning_rate": 0.0002, "gradient_accumulation_steps": 2, "warmup_ratio": 0.03, "max_grad_norm": 0.3, "lora_alpha": 32, "lora_dropout": 0.05, "lora_r": 64, "tf32": true, "seed": 42}, "log_history": [{"loss": 1.6636, "grad_norm": 0.9894065856933594, "learning_rate": 0.0002, "epoch": 0.0016166841807452913, "step": 10}, {"loss": 1.1528, "grad_norm": 1.7810699939727783, "learning_rate": 0.0002, "epoch": 0.0032333683614905826, "step": 20}, {"loss": 0.9767, "grad_norm": 0.5969577431678772, "learning_rate": 0.0002, "epoch": 0.004850052542235874, "step": 30}, {"loss": 0.9772, "grad_norm": 0.6354120969772339, "learning_rate": 0.0002, "epoch": 0.006466736722981165, "step": 40}, {"loss": 0.8643, "grad_norm": 0.5604607462882996, "learning_rate": 0.0002, "epoch": 0.008083420903726457, "step": 50}, {"loss": 0.8841, "grad_norm": 0.4676193594932556, "learning_rate": 0.0002, "epoch": 0.009700105084471748, "step": 60}, {"loss": 0.9022, "grad_norm": 0.6099211573600769, "learning_rate": 0.0002, "epoch": 0.01131678926521704, "step": 70}, {"loss": 0.9133, "grad_norm": 0.48639994859695435, "learning_rate": 0.0002, "epoch": 0.01293347344596233, "step": 80}, {"loss": 0.8704, "grad_norm": 0.4904264509677887, "learning_rate": 0.0002, "epoch": 0.014550157626707623, "step": 90}, {"loss": 0.8855, "grad_norm": 2.8334362506866455, "learning_rate": 0.0002, "epoch": 0.016166841807452915, "step": 100}, {"loss": 0.8958, "grad_norm": 0.43221670389175415, "learning_rate": 0.0002, "epoch": 0.017783525988198205, "step": 110}, {"loss": 0.8412, "grad_norm": 0.42244166135787964, "learning_rate": 0.0002, "epoch": 0.019400210168943496, "step": 120}, {"loss": 0.8467, "grad_norm": 0.45363298058509827, "learning_rate": 0.0002, "epoch": 0.02101689434968879, "step": 130}, {"loss": 0.8641, "grad_norm": 0.44816508889198303, "learning_rate": 0.0002, "epoch": 0.02263357853043408, "step": 140}, {"loss": 0.8496, "grad_norm": 0.43308213353157043, "learning_rate": 0.0002, "epoch": 0.02425026271117937, "step": 150}, {"loss": 0.8213, "grad_norm": 0.4084763526916504, "learning_rate": 0.0002, "epoch": 0.02586694689192466, "step": 160}, {"loss": 0.8343, "grad_norm": 0.5363703966140747, "learning_rate": 0.0002, "epoch": 0.027483631072669955, "step": 170}, {"loss": 0.8558, "grad_norm": 0.4619699716567993, "learning_rate": 0.0002, "epoch": 0.029100315253415245, "step": 180}, {"loss": 0.8878, "grad_norm": 0.49069908261299133, "learning_rate": 0.0002, "epoch": 0.030716999434160536, "step": 190}, {"loss": 0.8867, "grad_norm": 0.4645835757255554, "learning_rate": 0.0002, "epoch": 0.03233368361490583, "step": 200}, {"loss": 0.8842, "grad_norm": 1.2411243915557861, "learning_rate": 0.0002, "epoch": 0.03395036779565112, "step": 210}, {"loss": 0.8245, "grad_norm": 0.5211851596832275, "learning_rate": 0.0002, "epoch": 0.03556705197639641, "step": 220}, {"loss": 0.8194, "grad_norm": 0.5253691673278809, "learning_rate": 0.0002, "epoch": 0.037183736157141704, "step": 230}, {"loss": 0.8856, "grad_norm": 0.4567478895187378, "learning_rate": 0.0002, "epoch": 0.03880042033788699, "step": 240}, {"loss": 0.838, "grad_norm": 0.5472128391265869, "learning_rate": 0.0002, "epoch": 0.040417104518632285, "step": 250}, {"loss": 0.8201, "grad_norm": 0.42978546023368835, "learning_rate": 0.0002, "epoch": 0.04203378869937758, "step": 260}, {"loss": 0.8334, "grad_norm": 0.601734459400177, "learning_rate": 0.0002, "epoch": 0.043650472880122866, "step": 270}, {"loss": 0.815, "grad_norm": 0.4286513328552246, "learning_rate": 0.0002, "epoch": 0.04526715706086816, "step": 280}, {"loss": 0.8758, "grad_norm": 0.5230861902236938, "learning_rate": 0.0002, "epoch": 0.046883841241613454, "step": 290}, {"loss": 0.8636, "grad_norm": 0.6504611968994141, "learning_rate": 0.0002, "epoch": 0.04850052542235874, "step": 300}, {"loss": 0.8102, "grad_norm": 0.43485215306282043, "learning_rate": 0.0002, "epoch": 0.050117209603104035, "step": 310}, {"loss": 0.8221, "grad_norm": 0.4717007875442505, "learning_rate": 0.0002, "epoch": 0.05173389378384932, "step": 320}, {"loss": 0.8469, "grad_norm": 0.4059787690639496, "learning_rate": 0.0002, "epoch": 0.053350577964594616, "step": 330}, {"loss": 0.8866, "grad_norm": 0.4366913437843323, "learning_rate": 0.0002, "epoch": 0.05496726214533991, "step": 340}, {"loss": 0.7976, "grad_norm": 0.4233848452568054, "learning_rate": 0.0002, "epoch": 0.0565839463260852, "step": 350}, {"loss": 0.8456, "grad_norm": 0.4209108352661133, "learning_rate": 0.0002, "epoch": 0.05820063050683049, "step": 360}, {"loss": 0.816, "grad_norm": 0.41637396812438965, "learning_rate": 0.0002, "epoch": 0.059817314687575784, "step": 370}, {"loss": 0.7976, "grad_norm": 0.46235376596450806, "learning_rate": 0.0002, "epoch": 0.06143399886832107, "step": 380}, {"loss": 0.7966, "grad_norm": 0.4013484716415405, "learning_rate": 0.0002, "epoch": 0.06305068304906636, "step": 390}, {"loss": 0.8253, "grad_norm": 0.47443896532058716, "learning_rate": 0.0002, "epoch": 0.06466736722981166, "step": 400}, {"loss": 0.8666, "grad_norm": 0.3942156434059143, "learning_rate": 0.0002, "epoch": 0.06628405141055695, "step": 410}, {"loss": 0.8402, "grad_norm": 0.4965320825576782, "learning_rate": 0.0002, "epoch": 0.06790073559130223, "step": 420}, {"loss": 0.8317, "grad_norm": 0.4304835796356201, "learning_rate": 0.0002, "epoch": 0.06951741977204753, "step": 430}, {"loss": 0.8528, "grad_norm": 0.511726975440979, "learning_rate": 0.0002, "epoch": 0.07113410395279282, "step": 440}, {"loss": 0.8675, "grad_norm": 0.4040689170360565, "learning_rate": 0.0002, "epoch": 0.07275078813353811, "step": 450}, {"loss": 0.8788, "grad_norm": 0.5402171015739441, "learning_rate": 0.0002, "epoch": 0.07436747231428341, "step": 460}, {"loss": 0.8737, "grad_norm": 0.4174517095088959, "learning_rate": 0.0002, "epoch": 0.0759841564950287, "step": 470}, {"loss": 0.7605, "grad_norm": 0.4306182265281677, "learning_rate": 0.0002, "epoch": 0.07760084067577398, "step": 480}, {"loss": 0.799, "grad_norm": 0.535210132598877, "learning_rate": 0.0002, "epoch": 0.07921752485651928, "step": 490}, {"loss": 0.7825, "grad_norm": 0.5339109897613525, "learning_rate": 0.0002, "epoch": 0.08083420903726457, "step": 500}, {"loss": 0.8985, "grad_norm": 0.45754891633987427, "learning_rate": 0.0002, "epoch": 0.08245089321800986, "step": 510}, {"loss": 0.8144, "grad_norm": 0.43820783495903015, "learning_rate": 0.0002, "epoch": 0.08406757739875516, "step": 520}, {"loss": 0.8001, "grad_norm": 0.4434749186038971, "learning_rate": 0.0002, "epoch": 0.08568426157950045, "step": 530}, {"loss": 0.7857, "grad_norm": 0.43111467361450195, "learning_rate": 0.0002, "epoch": 0.08730094576024573, "step": 540}, {"loss": 0.8418, "grad_norm": 0.4378940165042877, "learning_rate": 0.0002, "epoch": 0.08891762994099103, "step": 550}, {"loss": 0.8361, "grad_norm": 0.4772215187549591, "learning_rate": 0.0002, "epoch": 0.09053431412173632, "step": 560}, {"loss": 0.8268, "grad_norm": 0.6837629079818726, "learning_rate": 0.0002, "epoch": 0.09215099830248161, "step": 570}, {"loss": 0.8607, "grad_norm": 0.42241212725639343, "learning_rate": 0.0002, "epoch": 0.09376768248322691, "step": 580}, {"loss": 0.852, "grad_norm": 0.5165936350822449, "learning_rate": 0.0002, "epoch": 0.0953843666639722, "step": 590}, {"loss": 0.8664, "grad_norm": 0.48737478256225586, "learning_rate": 0.0002, "epoch": 0.09700105084471748, "step": 600}, {"loss": 0.8806, "grad_norm": 0.47419852018356323, "learning_rate": 0.0002, "epoch": 0.09861773502546278, "step": 610}, {"loss": 0.8254, "grad_norm": 0.4975486099720001, "learning_rate": 0.0002, "epoch": 0.10023441920620807, "step": 620}, {"loss": 0.8548, "grad_norm": 0.49123844504356384, "learning_rate": 0.0002, "epoch": 0.10185110338695336, "step": 630}, {"loss": 0.8911, "grad_norm": 0.6288952827453613, "learning_rate": 0.0002, "epoch": 0.10346778756769864, "step": 640}, {"loss": 0.827, "grad_norm": 0.4277345836162567, "learning_rate": 0.0002, "epoch": 0.10508447174844394, "step": 650}, {"loss": 0.7996, "grad_norm": 0.4021061956882477, "learning_rate": 0.0002, "epoch": 0.10670115592918923, "step": 660}, {"loss": 0.87, "grad_norm": 0.3492237329483032, "learning_rate": 0.0002, "epoch": 0.10831784010993452, "step": 670}, {"loss": 0.8698, "grad_norm": 0.4341012239456177, "learning_rate": 0.0002, "epoch": 0.10993452429067982, "step": 680}, {"loss": 0.781, "grad_norm": 0.7296304106712341, "learning_rate": 0.0002, "epoch": 0.1115512084714251, "step": 690}, {"loss": 0.8433, "grad_norm": 0.397494912147522, "learning_rate": 0.0002, "epoch": 0.1131678926521704, "step": 700}, {"loss": 0.827, "grad_norm": 0.396431028842926, "learning_rate": 0.0002, "epoch": 0.1147845768329157, "step": 710}, {"loss": 0.8379, "grad_norm": 0.48842838406562805, "learning_rate": 0.0002, "epoch": 0.11640126101366098, "step": 720}, {"loss": 0.8238, "grad_norm": 0.46322616934776306, "learning_rate": 0.0002, "epoch": 0.11801794519440627, "step": 730}, {"loss": 0.8041, "grad_norm": 0.47990912199020386, "learning_rate": 0.0002, "epoch": 0.11963462937515157, "step": 740}, {"loss": 0.82, "grad_norm": 0.4997142255306244, "learning_rate": 0.0002, "epoch": 0.12125131355589686, "step": 750}, {"loss": 0.7702, "grad_norm": 0.4040526747703552, "learning_rate": 0.0002, "epoch": 0.12286799773664214, "step": 760}, {"loss": 0.863, "grad_norm": 0.453095942735672, "learning_rate": 0.0002, "epoch": 0.12448468191738744, "step": 770}, {"loss": 0.8792, "grad_norm": 0.4636971950531006, "learning_rate": 0.0002, "epoch": 0.12610136609813272, "step": 780}, {"loss": 0.8112, "grad_norm": 0.4279276132583618, "learning_rate": 0.0002, "epoch": 0.12771805027887803, "step": 790}, {"loss": 0.8711, "grad_norm": 0.46212655305862427, "learning_rate": 0.0002, "epoch": 0.12933473445962332, "step": 800}, {"loss": 0.8368, "grad_norm": 0.43127650022506714, "learning_rate": 0.0002, "epoch": 0.1309514186403686, "step": 810}, {"loss": 0.8476, "grad_norm": 0.4201301336288452, "learning_rate": 0.0002, "epoch": 0.1325681028211139, "step": 820}, {"loss": 0.8078, "grad_norm": 0.42583167552948, "learning_rate": 0.0002, "epoch": 0.13418478700185918, "step": 830}, {"loss": 0.8219, "grad_norm": 0.4535622000694275, "learning_rate": 0.0002, "epoch": 0.13580147118260447, "step": 840}, {"loss": 0.8423, "grad_norm": 0.4116036891937256, "learning_rate": 0.0002, "epoch": 0.13741815536334978, "step": 850}, {"loss": 0.8466, "grad_norm": 0.45997580885887146, "learning_rate": 0.0002, "epoch": 0.13903483954409507, "step": 860}, {"loss": 0.8917, "grad_norm": 0.4487837255001068, "learning_rate": 0.0002, "epoch": 0.14065152372484035, "step": 870}, {"loss": 0.8217, "grad_norm": 0.43650057911872864, "learning_rate": 0.0002, "epoch": 0.14226820790558564, "step": 880}, {"loss": 0.8178, "grad_norm": 0.5335358381271362, "learning_rate": 0.0002, "epoch": 0.14388489208633093, "step": 890}, {"loss": 0.7957, "grad_norm": 0.5989000201225281, "learning_rate": 0.0002, "epoch": 0.14550157626707622, "step": 900}, {"loss": 0.8385, "grad_norm": 0.517179012298584, "learning_rate": 0.0002, "epoch": 0.14711826044782153, "step": 910}, {"loss": 0.8255, "grad_norm": 0.44435232877731323, "learning_rate": 0.0002, "epoch": 0.14873494462856682, "step": 920}, {"loss": 0.8305, "grad_norm": 0.42635923624038696, "learning_rate": 0.0002, "epoch": 0.1503516288093121, "step": 930}, {"loss": 0.8043, "grad_norm": 0.49603334069252014, "learning_rate": 0.0002, "epoch": 0.1519683129900574, "step": 940}, {"loss": 0.8377, "grad_norm": 0.40639808773994446, "learning_rate": 0.0002, "epoch": 0.15358499717080268, "step": 950}, {"loss": 0.8529, "grad_norm": 0.4850759208202362, "learning_rate": 0.0002, "epoch": 0.15520168135154797, "step": 960}, {"loss": 0.846, "grad_norm": 0.4427442252635956, "learning_rate": 0.0002, "epoch": 0.15681836553229328, "step": 970}, {"loss": 0.8705, "grad_norm": 0.3760930001735687, "learning_rate": 0.0002, "epoch": 0.15843504971303857, "step": 980}, {"loss": 0.8644, "grad_norm": 0.4794144332408905, "learning_rate": 0.0002, "epoch": 0.16005173389378385, "step": 990}, {"loss": 0.8002, "grad_norm": 0.45828768610954285, "learning_rate": 0.0002, "epoch": 0.16166841807452914, "step": 1000}, {"loss": 0.7658, "grad_norm": 0.6313053369522095, "learning_rate": 0.0002, "epoch": 0.16328510225527443, "step": 1010}, {"loss": 0.8047, "grad_norm": 0.45041006803512573, "learning_rate": 0.0002, "epoch": 0.16490178643601971, "step": 1020}, {"loss": 0.8423, "grad_norm": 0.441403865814209, "learning_rate": 0.0002, "epoch": 0.166518470616765, "step": 1030}, {"loss": 0.8475, "grad_norm": 0.8171296119689941, "learning_rate": 0.0002, "epoch": 0.16813515479751032, "step": 1040}, {"loss": 0.845, "grad_norm": 0.7137420773506165, "learning_rate": 0.0002, "epoch": 0.1697518389782556, "step": 1050}, {"loss": 0.8213, "grad_norm": 0.5236809849739075, "learning_rate": 0.0002, "epoch": 0.1713685231590009, "step": 1060}, {"loss": 0.8265, "grad_norm": 0.5021864175796509, "learning_rate": 0.0002, "epoch": 0.17298520733974618, "step": 1070}, {"loss": 0.8305, "grad_norm": 0.47347521781921387, "learning_rate": 0.0002, "epoch": 0.17460189152049146, "step": 1080}, {"loss": 0.8105, "grad_norm": 0.4631653428077698, "learning_rate": 0.0002, "epoch": 0.17621857570123675, "step": 1090}, {"loss": 0.8166, "grad_norm": 0.49169182777404785, "learning_rate": 0.0002, "epoch": 0.17783525988198207, "step": 1100}, {"loss": 0.8012, "grad_norm": 0.5019739270210266, "learning_rate": 0.0002, "epoch": 0.17945194406272735, "step": 1110}, {"loss": 0.8247, "grad_norm": 0.5100422501564026, "learning_rate": 0.0002, "epoch": 0.18106862824347264, "step": 1120}, {"loss": 0.8142, "grad_norm": 0.3888324499130249, "learning_rate": 0.0002, "epoch": 0.18268531242421793, "step": 1130}, {"loss": 0.8533, "grad_norm": 0.39765217900276184, "learning_rate": 0.0002, "epoch": 0.18430199660496321, "step": 1140}, {"loss": 0.8541, "grad_norm": 0.47190186381340027, "learning_rate": 0.0002, "epoch": 0.1859186807857085, "step": 1150}, {"loss": 0.8301, "grad_norm": 0.4464188814163208, "learning_rate": 0.0002, "epoch": 0.18753536496645382, "step": 1160}, {"loss": 0.8341, "grad_norm": 0.5153930187225342, "learning_rate": 0.0002, "epoch": 0.1891520491471991, "step": 1170}, {"loss": 0.8033, "grad_norm": 0.4779708683490753, "learning_rate": 0.0002, "epoch": 0.1907687333279444, "step": 1180}, {"loss": 0.8187, "grad_norm": 0.4834315776824951, "learning_rate": 0.0002, "epoch": 0.19238541750868968, "step": 1190}, {"loss": 0.7721, "grad_norm": 0.402357816696167, "learning_rate": 0.0002, "epoch": 0.19400210168943496, "step": 1200}, {"loss": 0.7941, "grad_norm": 0.45899084210395813, "learning_rate": 0.0002, "epoch": 0.19561878587018025, "step": 1210}, {"loss": 0.8353, "grad_norm": 0.5106529593467712, "learning_rate": 0.0002, "epoch": 0.19723547005092557, "step": 1220}, {"loss": 0.7816, "grad_norm": 0.45261722803115845, "learning_rate": 0.0002, "epoch": 0.19885215423167085, "step": 1230}, {"loss": 0.8068, "grad_norm": 0.4647127091884613, "learning_rate": 0.0002, "epoch": 0.20046883841241614, "step": 1240}, {"loss": 0.8239, "grad_norm": 0.4849368929862976, "learning_rate": 0.0002, "epoch": 0.20208552259316143, "step": 1250}, {"loss": 0.8514, "grad_norm": 0.4518061578273773, "learning_rate": 0.0002, "epoch": 0.2037022067739067, "step": 1260}, {"loss": 0.8158, "grad_norm": 0.49535325169563293, "learning_rate": 0.0002, "epoch": 0.205318890954652, "step": 1270}, {"loss": 0.8348, "grad_norm": 0.4835205376148224, "learning_rate": 0.0002, "epoch": 0.2069355751353973, "step": 1280}, {"loss": 0.8428, "grad_norm": 0.45308539271354675, "learning_rate": 0.0002, "epoch": 0.2085522593161426, "step": 1290}, {"loss": 0.7993, "grad_norm": 0.5369905233383179, "learning_rate": 0.0002, "epoch": 0.2101689434968879, "step": 1300}, {"loss": 0.8676, "grad_norm": 0.5031622052192688, "learning_rate": 0.0002, "epoch": 0.21178562767763318, "step": 1310}, {"loss": 0.7686, "grad_norm": 0.48010334372520447, "learning_rate": 0.0002, "epoch": 0.21340231185837846, "step": 1320}, {"loss": 0.806, "grad_norm": 0.4905701279640198, "learning_rate": 0.0002, "epoch": 0.21501899603912375, "step": 1330}, {"loss": 0.7885, "grad_norm": 0.43531742691993713, "learning_rate": 0.0002, "epoch": 0.21663568021986904, "step": 1340}, {"loss": 0.8191, "grad_norm": 0.44330692291259766, "learning_rate": 0.0002, "epoch": 0.21825236440061435, "step": 1350}, {"loss": 0.8205, "grad_norm": 0.5384416580200195, "learning_rate": 0.0002, "epoch": 0.21986904858135964, "step": 1360}, {"loss": 0.7726, "grad_norm": 0.4181833863258362, "learning_rate": 0.0002, "epoch": 0.22148573276210493, "step": 1370}, {"loss": 0.8311, "grad_norm": 0.523833692073822, "learning_rate": 0.0002, "epoch": 0.2231024169428502, "step": 1380}, {"loss": 0.7913, "grad_norm": 0.5528736710548401, "learning_rate": 0.0002, "epoch": 0.2247191011235955, "step": 1390}, {"loss": 0.8079, "grad_norm": 0.43515023589134216, "learning_rate": 0.0002, "epoch": 0.2263357853043408, "step": 1400}, {"loss": 0.8403, "grad_norm": 0.48809877038002014, "learning_rate": 0.0002, "epoch": 0.2279524694850861, "step": 1410}, {"loss": 0.8165, "grad_norm": 0.43591251969337463, "learning_rate": 0.0002, "epoch": 0.2295691536658314, "step": 1420}, {"loss": 0.8147, "grad_norm": 0.44625312089920044, "learning_rate": 0.0002, "epoch": 0.23118583784657668, "step": 1430}, {"loss": 0.8134, "grad_norm": 0.4390665292739868, "learning_rate": 0.0002, "epoch": 0.23280252202732196, "step": 1440}, {"loss": 0.8465, "grad_norm": 0.48496049642562866, "learning_rate": 0.0002, "epoch": 0.23441920620806725, "step": 1450}, {"loss": 0.775, "grad_norm": 0.45919957756996155, "learning_rate": 0.0002, "epoch": 0.23603589038881254, "step": 1460}, {"loss": 0.8659, "grad_norm": 0.5471845865249634, "learning_rate": 0.0002, "epoch": 0.23765257456955785, "step": 1470}, {"loss": 0.8164, "grad_norm": 0.47269317507743835, "learning_rate": 0.0002, "epoch": 0.23926925875030314, "step": 1480}, {"loss": 0.854, "grad_norm": 0.4930245578289032, "learning_rate": 0.0002, "epoch": 0.24088594293104842, "step": 1490}, {"loss": 0.8139, "grad_norm": 0.5605630278587341, "learning_rate": 0.0002, "epoch": 0.2425026271117937, "step": 1500}, {"loss": 0.8125, "grad_norm": 0.4435870945453644, "learning_rate": 0.0002, "epoch": 0.244119311292539, "step": 1510}, {"loss": 0.8123, "grad_norm": 0.4941999912261963, "learning_rate": 0.0002, "epoch": 0.24573599547328429, "step": 1520}, {"loss": 0.8427, "grad_norm": 0.5100624561309814, "learning_rate": 0.0002, "epoch": 0.24735267965402957, "step": 1530}, {"loss": 0.8405, "grad_norm": 0.4638267457485199, "learning_rate": 0.0002, "epoch": 0.2489693638347749, "step": 1540}, {"loss": 0.81, "grad_norm": 0.5071570873260498, "learning_rate": 0.0002, "epoch": 0.25058604801552015, "step": 1550}, {"loss": 0.7724, "grad_norm": 0.4291319251060486, "learning_rate": 0.0002, "epoch": 0.25220273219626543, "step": 1560}, {"loss": 0.7984, "grad_norm": 0.5388049483299255, "learning_rate": 0.0002, "epoch": 0.2538194163770108, "step": 1570}, {"loss": 0.8176, "grad_norm": 0.5083683729171753, "learning_rate": 0.0002, "epoch": 0.25543610055775606, "step": 1580}, {"loss": 0.843, "grad_norm": 0.4824463725090027, "learning_rate": 0.0002, "epoch": 0.25705278473850135, "step": 1590}, {"loss": 0.7996, "grad_norm": 0.41177722811698914, "learning_rate": 0.0002, "epoch": 0.25866946891924664, "step": 1600}, {"loss": 0.7772, "grad_norm": 0.5656219124794006, "learning_rate": 0.0002, "epoch": 0.2602861530999919, "step": 1610}, {"loss": 0.7955, "grad_norm": 0.41063204407691956, "learning_rate": 0.0002, "epoch": 0.2619028372807372, "step": 1620}, {"loss": 0.7998, "grad_norm": 0.4897061288356781, "learning_rate": 0.0002, "epoch": 0.2635195214614825, "step": 1630}, {"loss": 0.8198, "grad_norm": 0.4454376697540283, "learning_rate": 0.0002, "epoch": 0.2651362056422278, "step": 1640}, {"loss": 0.8684, "grad_norm": 0.4355238378047943, "learning_rate": 0.0002, "epoch": 0.26675288982297307, "step": 1650}, {"loss": 0.7801, "grad_norm": 0.458310067653656, "learning_rate": 0.0002, "epoch": 0.26836957400371836, "step": 1660}, {"loss": 0.7935, "grad_norm": 0.4752083718776703, "learning_rate": 0.0002, "epoch": 0.26998625818446365, "step": 1670}, {"loss": 0.8267, "grad_norm": 0.4666106402873993, "learning_rate": 0.0002, "epoch": 0.27160294236520893, "step": 1680}, {"loss": 0.8252, "grad_norm": 0.4213818609714508, "learning_rate": 0.0002, "epoch": 0.2732196265459543, "step": 1690}, {"loss": 0.8559, "grad_norm": 0.5768913626670837, "learning_rate": 0.0002, "epoch": 0.27483631072669956, "step": 1700}, {"loss": 0.7931, "grad_norm": 0.4209914803504944, "learning_rate": 0.0002, "epoch": 0.27645299490744485, "step": 1710}, {"loss": 0.8167, "grad_norm": 0.501909613609314, "learning_rate": 0.0002, "epoch": 0.27806967908819014, "step": 1720}, {"loss": 0.7832, "grad_norm": 0.5266261100769043, "learning_rate": 0.0002, "epoch": 0.2796863632689354, "step": 1730}, {"loss": 0.8102, "grad_norm": 0.43806859850883484, "learning_rate": 0.0002, "epoch": 0.2813030474496807, "step": 1740}, {"loss": 0.8157, "grad_norm": 0.46048814058303833, "learning_rate": 0.0002, "epoch": 0.282919731630426, "step": 1750}, {"loss": 0.8596, "grad_norm": 0.44972819089889526, "learning_rate": 0.0002, "epoch": 0.2845364158111713, "step": 1760}, {"loss": 0.8421, "grad_norm": 0.5114831328392029, "learning_rate": 0.0002, "epoch": 0.28615309999191657, "step": 1770}, {"loss": 0.8361, "grad_norm": 0.47931742668151855, "learning_rate": 0.0002, "epoch": 0.28776978417266186, "step": 1780}, {"loss": 0.8265, "grad_norm": 0.5092599987983704, "learning_rate": 0.0002, "epoch": 0.28938646835340714, "step": 1790}, {"loss": 0.8506, "grad_norm": 0.37581443786621094, "learning_rate": 0.0002, "epoch": 0.29100315253415243, "step": 1800}, {"loss": 0.7932, "grad_norm": 0.47097381949424744, "learning_rate": 0.0002, "epoch": 0.2926198367148977, "step": 1810}, {"loss": 0.7787, "grad_norm": 0.48300236463546753, "learning_rate": 0.0002, "epoch": 0.29423652089564306, "step": 1820}, {"loss": 0.8391, "grad_norm": 0.5600419640541077, "learning_rate": 0.0002, "epoch": 0.29585320507638835, "step": 1830}, {"loss": 0.8507, "grad_norm": 0.48555272817611694, "learning_rate": 0.0002, "epoch": 0.29746988925713364, "step": 1840}, {"loss": 0.7657, "grad_norm": 0.3752668499946594, "learning_rate": 0.0002, "epoch": 0.2990865734378789, "step": 1850}, {"loss": 0.7915, "grad_norm": 0.5328747034072876, "learning_rate": 0.0002, "epoch": 0.3007032576186242, "step": 1860}, {"loss": 0.8426, "grad_norm": 0.48716455698013306, "learning_rate": 0.0002, "epoch": 0.3023199417993695, "step": 1870}, {"loss": 0.8335, "grad_norm": 0.5011493563652039, "learning_rate": 0.0002, "epoch": 0.3039366259801148, "step": 1880}, {"loss": 0.852, "grad_norm": 0.46461427211761475, "learning_rate": 0.0002, "epoch": 0.30555331016086007, "step": 1890}, {"loss": 0.8478, "grad_norm": 0.36630210280418396, "learning_rate": 0.0002, "epoch": 0.30716999434160536, "step": 1900}, {"loss": 0.8162, "grad_norm": 0.4217296242713928, "learning_rate": 0.0002, "epoch": 0.30878667852235064, "step": 1910}, {"loss": 0.8128, "grad_norm": 0.4394875466823578, "learning_rate": 0.0002, "epoch": 0.31040336270309593, "step": 1920}, {"loss": 0.8471, "grad_norm": 0.6587965488433838, "learning_rate": 0.0002, "epoch": 0.3120200468838412, "step": 1930}, {"loss": 0.8565, "grad_norm": 0.5469298958778381, "learning_rate": 0.0002, "epoch": 0.31363673106458656, "step": 1940}, {"loss": 0.8236, "grad_norm": 0.4371595084667206, "learning_rate": 0.0002, "epoch": 0.31525341524533185, "step": 1950}, {"loss": 0.887, "grad_norm": 0.4809541404247284, "learning_rate": 0.0002, "epoch": 0.31687009942607713, "step": 1960}, {"loss": 0.7855, "grad_norm": 0.6061086654663086, "learning_rate": 0.0002, "epoch": 0.3184867836068224, "step": 1970}, {"loss": 0.7679, "grad_norm": 0.5342657566070557, "learning_rate": 0.0002, "epoch": 0.3201034677875677, "step": 1980}, {"loss": 0.7955, "grad_norm": 0.5057743787765503, "learning_rate": 0.0002, "epoch": 0.321720151968313, "step": 1990}, {"loss": 0.7774, "grad_norm": 0.528626024723053, "learning_rate": 0.0002, "epoch": 0.3233368361490583, "step": 2000}, {"loss": 0.8845, "grad_norm": 0.46742770075798035, "learning_rate": 0.0002, "epoch": 0.32495352032980357, "step": 2010}, {"loss": 0.8484, "grad_norm": 0.515101432800293, "learning_rate": 0.0002, "epoch": 0.32657020451054886, "step": 2020}, {"loss": 0.8139, "grad_norm": 0.41941216588020325, "learning_rate": 0.0002, "epoch": 0.32818688869129414, "step": 2030}, {"loss": 0.7637, "grad_norm": 0.49902522563934326, "learning_rate": 0.0002, "epoch": 0.32980357287203943, "step": 2040}, {"loss": 0.7822, "grad_norm": 0.4120897650718689, "learning_rate": 0.0002, "epoch": 0.3314202570527847, "step": 2050}, {"loss": 0.8057, "grad_norm": 0.45352041721343994, "learning_rate": 0.0002, "epoch": 0.33303694123353, "step": 2060}, {"loss": 0.7913, "grad_norm": 0.523199737071991, "learning_rate": 0.0002, "epoch": 0.33465362541427535, "step": 2070}, {"loss": 0.8036, "grad_norm": 0.4390358626842499, "learning_rate": 0.0002, "epoch": 0.33627030959502063, "step": 2080}, {"loss": 0.8145, "grad_norm": 0.6752901077270508, "learning_rate": 0.0002, "epoch": 0.3378869937757659, "step": 2090}, {"loss": 0.7807, "grad_norm": 0.547821044921875, "learning_rate": 0.0002, "epoch": 0.3395036779565112, "step": 2100}, {"loss": 0.8561, "grad_norm": 0.5161308646202087, "learning_rate": 0.0002, "epoch": 0.3411203621372565, "step": 2110}, {"loss": 0.7697, "grad_norm": 0.4565401077270508, "learning_rate": 0.0002, "epoch": 0.3427370463180018, "step": 2120}, {"loss": 0.7964, "grad_norm": 0.4666115939617157, "learning_rate": 0.0002, "epoch": 0.34435373049874707, "step": 2130}, {"loss": 0.8189, "grad_norm": 0.4090428352355957, "learning_rate": 0.0002, "epoch": 0.34597041467949236, "step": 2140}, {"loss": 0.8817, "grad_norm": 0.510845422744751, "learning_rate": 0.0002, "epoch": 0.34758709886023764, "step": 2150}, {"loss": 0.8398, "grad_norm": 0.42861923575401306, "learning_rate": 0.0002, "epoch": 0.34920378304098293, "step": 2160}, {"loss": 0.7716, "grad_norm": 0.4476332664489746, "learning_rate": 0.0002, "epoch": 0.3508204672217282, "step": 2170}, {"loss": 0.7845, "grad_norm": 0.6065791249275208, "learning_rate": 0.0002, "epoch": 0.3524371514024735, "step": 2180}, {"loss": 0.8187, "grad_norm": 0.42335066199302673, "learning_rate": 0.0002, "epoch": 0.35405383558321885, "step": 2190}, {"loss": 0.8239, "grad_norm": 0.5094629526138306, "learning_rate": 0.0002, "epoch": 0.35567051976396413, "step": 2200}, {"loss": 0.7807, "grad_norm": 0.5476373434066772, "learning_rate": 0.0002, "epoch": 0.3572872039447094, "step": 2210}, {"loss": 0.814, "grad_norm": 0.3911719024181366, "learning_rate": 0.0002, "epoch": 0.3589038881254547, "step": 2220}, {"loss": 0.8599, "grad_norm": 0.6599636077880859, "learning_rate": 0.0002, "epoch": 0.3605205723062, "step": 2230}, {"loss": 0.7482, "grad_norm": 0.40381914377212524, "learning_rate": 0.0002, "epoch": 0.3621372564869453, "step": 2240}, {"loss": 0.7772, "grad_norm": 0.4433908462524414, "learning_rate": 0.0002, "epoch": 0.36375394066769057, "step": 2250}, {"loss": 0.8503, "grad_norm": 0.578326940536499, "learning_rate": 0.0002, "epoch": 0.36537062484843585, "step": 2260}, {"loss": 0.8178, "grad_norm": 0.5734784007072449, "learning_rate": 0.0002, "epoch": 0.36698730902918114, "step": 2270}, {"loss": 0.8193, "grad_norm": 0.45555487275123596, "learning_rate": 0.0002, "epoch": 0.36860399320992643, "step": 2280}, {"loss": 0.7929, "grad_norm": 0.5666276216506958, "learning_rate": 0.0002, "epoch": 0.3702206773906717, "step": 2290}, {"loss": 0.8292, "grad_norm": 0.5461117625236511, "learning_rate": 0.0002, "epoch": 0.371837361571417, "step": 2300}, {"loss": 0.8204, "grad_norm": 0.6318911910057068, "learning_rate": 0.0002, "epoch": 0.3734540457521623, "step": 2310}, {"loss": 0.7964, "grad_norm": 0.493263304233551, "learning_rate": 0.0002, "epoch": 0.37507072993290763, "step": 2320}, {"loss": 0.8339, "grad_norm": 0.5888760089874268, "learning_rate": 0.0002, "epoch": 0.3766874141136529, "step": 2330}, {"loss": 0.7737, "grad_norm": 0.48671841621398926, "learning_rate": 0.0002, "epoch": 0.3783040982943982, "step": 2340}, {"loss": 0.8367, "grad_norm": 0.4385145306587219, "learning_rate": 0.0002, "epoch": 0.3799207824751435, "step": 2350}, {"loss": 0.812, "grad_norm": 0.5523318648338318, "learning_rate": 0.0002, "epoch": 0.3815374666558888, "step": 2360}, {"loss": 0.8351, "grad_norm": 0.7308220267295837, "learning_rate": 0.0002, "epoch": 0.38315415083663407, "step": 2370}, {"loss": 0.859, "grad_norm": 0.554214358329773, "learning_rate": 0.0002, "epoch": 0.38477083501737935, "step": 2380}, {"loss": 0.8146, "grad_norm": 0.5425800085067749, "learning_rate": 0.0002, "epoch": 0.38638751919812464, "step": 2390}, {"loss": 0.8282, "grad_norm": 0.48811158537864685, "learning_rate": 0.0002, "epoch": 0.3880042033788699, "step": 2400}, {"loss": 0.8074, "grad_norm": 0.49212366342544556, "learning_rate": 0.0002, "epoch": 0.3896208875596152, "step": 2410}, {"loss": 0.7991, "grad_norm": 0.5222218632698059, "learning_rate": 0.0002, "epoch": 0.3912375717403605, "step": 2420}, {"loss": 0.8182, "grad_norm": 0.4699819087982178, "learning_rate": 0.0002, "epoch": 0.3928542559211058, "step": 2430}, {"loss": 0.7919, "grad_norm": 0.46153587102890015, "learning_rate": 0.0002, "epoch": 0.39447094010185113, "step": 2440}, {"loss": 0.8111, "grad_norm": 0.4150611162185669, "learning_rate": 0.0002, "epoch": 0.3960876242825964, "step": 2450}, {"loss": 0.8589, "grad_norm": 0.5799614787101746, "learning_rate": 0.0002, "epoch": 0.3977043084633417, "step": 2460}, {"loss": 0.8085, "grad_norm": 0.56536865234375, "learning_rate": 0.0002, "epoch": 0.399320992644087, "step": 2470}, {"loss": 0.8022, "grad_norm": 0.5451247096061707, "learning_rate": 0.0002, "epoch": 0.4009376768248323, "step": 2480}, {"loss": 0.8217, "grad_norm": 0.5914521217346191, "learning_rate": 0.0002, "epoch": 0.40255436100557757, "step": 2490}, {"loss": 0.7859, "grad_norm": 0.4428117275238037, "learning_rate": 0.0002, "epoch": 0.40417104518632285, "step": 2500}, {"loss": 0.8054, "grad_norm": 0.48580947518348694, "learning_rate": 0.0002, "epoch": 0.40578772936706814, "step": 2510}, {"loss": 0.8405, "grad_norm": 0.436734676361084, "learning_rate": 0.0002, "epoch": 0.4074044135478134, "step": 2520}, {"loss": 0.8209, "grad_norm": 0.5752223134040833, "learning_rate": 0.0002, "epoch": 0.4090210977285587, "step": 2530}, {"loss": 0.8181, "grad_norm": 0.4271308183670044, "learning_rate": 0.0002, "epoch": 0.410637781909304, "step": 2540}, {"loss": 0.8058, "grad_norm": 0.46294718980789185, "learning_rate": 0.0002, "epoch": 0.4122544660900493, "step": 2550}, {"loss": 0.8473, "grad_norm": 0.49407583475112915, "learning_rate": 0.0002, "epoch": 0.4138711502707946, "step": 2560}, {"loss": 0.7881, "grad_norm": 0.4729035496711731, "learning_rate": 0.0002, "epoch": 0.4154878344515399, "step": 2570}, {"loss": 0.7834, "grad_norm": 0.4129747152328491, "learning_rate": 0.0002, "epoch": 0.4171045186322852, "step": 2580}, {"loss": 0.7859, "grad_norm": 0.5684236288070679, "learning_rate": 0.0002, "epoch": 0.4187212028130305, "step": 2590}, {"loss": 0.811, "grad_norm": 0.4862157106399536, "learning_rate": 0.0002, "epoch": 0.4203378869937758, "step": 2600}, {"loss": 0.7582, "grad_norm": 0.46567976474761963, "learning_rate": 0.0002, "epoch": 0.42195457117452106, "step": 2610}, {"loss": 0.7755, "grad_norm": 0.5710650682449341, "learning_rate": 0.0002, "epoch": 0.42357125535526635, "step": 2620}, {"loss": 0.8573, "grad_norm": 0.5660041570663452, "learning_rate": 0.0002, "epoch": 0.42518793953601164, "step": 2630}, {"loss": 0.7812, "grad_norm": 0.47944375872612, "learning_rate": 0.0002, "epoch": 0.4268046237167569, "step": 2640}, {"loss": 0.7459, "grad_norm": 0.537223756313324, "learning_rate": 0.0002, "epoch": 0.4284213078975022, "step": 2650}, {"loss": 0.8246, "grad_norm": 0.41669997572898865, "learning_rate": 0.0002, "epoch": 0.4300379920782475, "step": 2660}, {"loss": 0.7785, "grad_norm": 0.44727686047554016, "learning_rate": 0.0002, "epoch": 0.4316546762589928, "step": 2670}, {"loss": 0.8241, "grad_norm": 0.5600888729095459, "learning_rate": 0.0002, "epoch": 0.4332713604397381, "step": 2680}, {"loss": 0.7708, "grad_norm": 0.39820605516433716, "learning_rate": 0.0002, "epoch": 0.4348880446204834, "step": 2690}, {"loss": 0.8202, "grad_norm": 0.5637655854225159, "learning_rate": 0.0002, "epoch": 0.4365047288012287, "step": 2700}, {"loss": 0.855, "grad_norm": 0.6363666653633118, "learning_rate": 0.0002, "epoch": 0.438121412981974, "step": 2710}, {"loss": 0.8468, "grad_norm": 0.5656129121780396, "learning_rate": 0.0002, "epoch": 0.4397380971627193, "step": 2720}, {"loss": 0.7845, "grad_norm": 0.5600156188011169, "learning_rate": 0.0002, "epoch": 0.44135478134346456, "step": 2730}, {"loss": 0.8405, "grad_norm": 0.5506579875946045, "learning_rate": 0.0002, "epoch": 0.44297146552420985, "step": 2740}, {"loss": 0.7725, "grad_norm": 0.49878305196762085, "learning_rate": 0.0002, "epoch": 0.44458814970495514, "step": 2750}, {"loss": 0.8292, "grad_norm": 0.4569213092327118, "learning_rate": 0.0002, "epoch": 0.4462048338857004, "step": 2760}, {"loss": 0.8028, "grad_norm": 0.6056680083274841, "learning_rate": 0.0002, "epoch": 0.4478215180664457, "step": 2770}, {"loss": 0.8242, "grad_norm": 0.44474557042121887, "learning_rate": 0.0002, "epoch": 0.449438202247191, "step": 2780}, {"loss": 0.801, "grad_norm": 0.46055394411087036, "learning_rate": 0.0002, "epoch": 0.4510548864279363, "step": 2790}, {"loss": 0.7521, "grad_norm": 0.4904133379459381, "learning_rate": 0.0002, "epoch": 0.4526715706086816, "step": 2800}, {"loss": 0.8829, "grad_norm": 0.5647031664848328, "learning_rate": 0.0002, "epoch": 0.45428825478942686, "step": 2810}, {"loss": 0.8622, "grad_norm": 0.5759473443031311, "learning_rate": 0.0002, "epoch": 0.4559049389701722, "step": 2820}, {"loss": 0.7812, "grad_norm": 0.5161895751953125, "learning_rate": 0.0002, "epoch": 0.4575216231509175, "step": 2830}, {"loss": 0.8045, "grad_norm": 0.4248254597187042, "learning_rate": 0.0002, "epoch": 0.4591383073316628, "step": 2840}, {"loss": 0.7838, "grad_norm": 0.45395001769065857, "learning_rate": 0.0002, "epoch": 0.46075499151240806, "step": 2850}, {"loss": 0.8208, "grad_norm": 0.5358697772026062, "learning_rate": 0.0002, "epoch": 0.46237167569315335, "step": 2860}, {"loss": 0.8147, "grad_norm": 0.5379165410995483, "learning_rate": 0.0002, "epoch": 0.46398835987389864, "step": 2870}, {"loss": 0.7403, "grad_norm": 0.4601989686489105, "learning_rate": 0.0002, "epoch": 0.4656050440546439, "step": 2880}, {"loss": 0.8523, "grad_norm": 0.671115517616272, "learning_rate": 0.0002, "epoch": 0.4672217282353892, "step": 2890}, {"loss": 0.8262, "grad_norm": 0.4425133168697357, "learning_rate": 0.0002, "epoch": 0.4688384124161345, "step": 2900}, {"loss": 0.8178, "grad_norm": 0.5446155071258545, "learning_rate": 0.0002, "epoch": 0.4704550965968798, "step": 2910}, {"loss": 0.8106, "grad_norm": 0.603306233882904, "learning_rate": 0.0002, "epoch": 0.47207178077762507, "step": 2920}, {"loss": 0.8044, "grad_norm": 0.5377997159957886, "learning_rate": 0.0002, "epoch": 0.47368846495837036, "step": 2930}, {"loss": 0.8075, "grad_norm": 0.4931027591228485, "learning_rate": 0.0002, "epoch": 0.4753051491391157, "step": 2940}, {"loss": 0.8004, "grad_norm": 0.4711960256099701, "learning_rate": 0.0002, "epoch": 0.476921833319861, "step": 2950}, {"loss": 0.8121, "grad_norm": 0.5020492672920227, "learning_rate": 0.0002, "epoch": 0.4785385175006063, "step": 2960}, {"loss": 0.8221, "grad_norm": 0.5428946614265442, "learning_rate": 0.0002, "epoch": 0.48015520168135156, "step": 2970}, {"loss": 0.7849, "grad_norm": 0.5294089317321777, "learning_rate": 0.0002, "epoch": 0.48177188586209685, "step": 2980}, {"loss": 0.8553, "grad_norm": 0.648289144039154, "learning_rate": 0.0002, "epoch": 0.48338857004284214, "step": 2990}, {"loss": 0.7874, "grad_norm": 0.47916680574417114, "learning_rate": 0.0002, "epoch": 0.4850052542235874, "step": 3000}, {"loss": 0.8087, "grad_norm": 0.43849772214889526, "learning_rate": 0.0002, "epoch": 0.4866219384043327, "step": 3010}, {"loss": 0.7662, "grad_norm": 0.47007861733436584, "learning_rate": 0.0002, "epoch": 0.488238622585078, "step": 3020}, {"loss": 0.757, "grad_norm": 0.6314331293106079, "learning_rate": 0.0002, "epoch": 0.4898553067658233, "step": 3030}, {"loss": 0.7863, "grad_norm": 0.49211493134498596, "learning_rate": 0.0002, "epoch": 0.49147199094656857, "step": 3040}, {"loss": 0.8335, "grad_norm": 0.4537973403930664, "learning_rate": 0.0002, "epoch": 0.49308867512731386, "step": 3050}, {"loss": 0.8095, "grad_norm": 0.47326919436454773, "learning_rate": 0.0002, "epoch": 0.49470535930805914, "step": 3060}, {"loss": 0.8447, "grad_norm": 0.525874137878418, "learning_rate": 0.0002, "epoch": 0.4963220434888045, "step": 3070}, {"loss": 0.8339, "grad_norm": 0.6361091732978821, "learning_rate": 0.0002, "epoch": 0.4979387276695498, "step": 3080}, {"loss": 0.821, "grad_norm": 0.5850642919540405, "learning_rate": 0.0002, "epoch": 0.49955541185029506, "step": 3090}, {"loss": 0.8279, "grad_norm": 0.47299543023109436, "learning_rate": 0.0002, "epoch": 0.5011720960310403, "step": 3100}, {"loss": 0.8681, "grad_norm": 0.473099946975708, "learning_rate": 0.0002, "epoch": 0.5027887802117856, "step": 3110}, {"loss": 0.8223, "grad_norm": 0.48186397552490234, "learning_rate": 0.0002, "epoch": 0.5044054643925309, "step": 3120}, {"loss": 0.8292, "grad_norm": 0.5015401840209961, "learning_rate": 0.0002, "epoch": 0.5060221485732762, "step": 3130}, {"loss": 0.7692, "grad_norm": 0.5617750287055969, "learning_rate": 0.0002, "epoch": 0.5076388327540216, "step": 3140}, {"loss": 0.8708, "grad_norm": 0.5169327259063721, "learning_rate": 0.0002, "epoch": 0.5092555169347668, "step": 3150}, {"loss": 0.7845, "grad_norm": 0.545657753944397, "learning_rate": 0.0002, "epoch": 0.5108722011155121, "step": 3160}, {"loss": 0.799, "grad_norm": 0.512864351272583, "learning_rate": 0.0002, "epoch": 0.5124888852962574, "step": 3170}, {"loss": 0.7794, "grad_norm": 0.4113546311855316, "learning_rate": 0.0002, "epoch": 0.5141055694770027, "step": 3180}, {"loss": 0.8206, "grad_norm": 0.44532445073127747, "learning_rate": 0.0002, "epoch": 0.5157222536577479, "step": 3190}, {"loss": 0.8213, "grad_norm": 0.5623497366905212, "learning_rate": 0.0002, "epoch": 0.5173389378384933, "step": 3200}, {"loss": 0.7928, "grad_norm": 0.5084741115570068, "learning_rate": 0.0002, "epoch": 0.5189556220192385, "step": 3210}, {"loss": 0.8174, "grad_norm": 0.5305403470993042, "learning_rate": 0.0002, "epoch": 0.5205723061999838, "step": 3220}, {"loss": 0.8139, "grad_norm": 0.4708254337310791, "learning_rate": 0.0002, "epoch": 0.5221889903807291, "step": 3230}, {"loss": 0.7639, "grad_norm": 0.43827131390571594, "learning_rate": 0.0002, "epoch": 0.5238056745614744, "step": 3240}, {"loss": 0.7993, "grad_norm": 0.5630002617835999, "learning_rate": 0.0002, "epoch": 0.5254223587422197, "step": 3250}, {"loss": 0.7522, "grad_norm": 0.5010961890220642, "learning_rate": 0.0002, "epoch": 0.527039042922965, "step": 3260}, {"loss": 0.8374, "grad_norm": 0.6303122043609619, "learning_rate": 0.0002, "epoch": 0.5286557271037103, "step": 3270}, {"loss": 0.7727, "grad_norm": 0.5107331275939941, "learning_rate": 0.0002, "epoch": 0.5302724112844556, "step": 3280}, {"loss": 0.8495, "grad_norm": 0.5700443387031555, "learning_rate": 0.0002, "epoch": 0.5318890954652009, "step": 3290}, {"loss": 0.7776, "grad_norm": 0.46296367049217224, "learning_rate": 0.0002, "epoch": 0.5335057796459461, "step": 3300}, {"loss": 0.7931, "grad_norm": 0.531568706035614, "learning_rate": 0.0002, "epoch": 0.5351224638266915, "step": 3310}, {"loss": 0.843, "grad_norm": 0.4686741530895233, "learning_rate": 0.0002, "epoch": 0.5367391480074367, "step": 3320}, {"loss": 0.8104, "grad_norm": 0.5404331088066101, "learning_rate": 0.0002, "epoch": 0.5383558321881821, "step": 3330}, {"loss": 0.7686, "grad_norm": 0.6368790864944458, "learning_rate": 0.0002, "epoch": 0.5399725163689273, "step": 3340}, {"loss": 0.8514, "grad_norm": 0.42300888895988464, "learning_rate": 0.0002, "epoch": 0.5415892005496726, "step": 3350}, {"loss": 0.8236, "grad_norm": 0.5362542867660522, "learning_rate": 0.0002, "epoch": 0.5432058847304179, "step": 3360}, {"loss": 0.858, "grad_norm": 0.497128963470459, "learning_rate": 0.0002, "epoch": 0.5448225689111632, "step": 3370}, {"loss": 0.8519, "grad_norm": 0.5006386041641235, "learning_rate": 0.0002, "epoch": 0.5464392530919085, "step": 3380}, {"loss": 0.7867, "grad_norm": 0.44136837124824524, "learning_rate": 0.0002, "epoch": 0.5480559372726538, "step": 3390}, {"loss": 0.773, "grad_norm": 0.5897833108901978, "learning_rate": 0.0002, "epoch": 0.5496726214533991, "step": 3400}, {"loss": 0.8895, "grad_norm": 0.641075611114502, "learning_rate": 0.0002, "epoch": 0.5512893056341444, "step": 3410}, {"loss": 0.7827, "grad_norm": 0.7251322269439697, "learning_rate": 0.0002, "epoch": 0.5529059898148897, "step": 3420}, {"loss": 0.7626, "grad_norm": 0.47411349415779114, "learning_rate": 0.0002, "epoch": 0.5545226739956349, "step": 3430}, {"loss": 0.8196, "grad_norm": 0.4994310438632965, "learning_rate": 0.0002, "epoch": 0.5561393581763803, "step": 3440}, {"loss": 0.7812, "grad_norm": 0.5814438462257385, "learning_rate": 0.0002, "epoch": 0.5577560423571255, "step": 3450}, {"loss": 0.8805, "grad_norm": 0.6278898119926453, "learning_rate": 0.0002, "epoch": 0.5593727265378708, "step": 3460}, {"loss": 0.813, "grad_norm": 0.46208274364471436, "learning_rate": 0.0002, "epoch": 0.5609894107186161, "step": 3470}, {"loss": 0.8295, "grad_norm": 0.5718930959701538, "learning_rate": 0.0002, "epoch": 0.5626060948993614, "step": 3480}, {"loss": 0.8152, "grad_norm": 0.48178744316101074, "learning_rate": 0.0002, "epoch": 0.5642227790801067, "step": 3490}, {"loss": 0.8244, "grad_norm": 0.47336965799331665, "learning_rate": 0.0002, "epoch": 0.565839463260852, "step": 3500}, {"loss": 0.8099, "grad_norm": 0.43442684412002563, "learning_rate": 0.0002, "epoch": 0.5674561474415973, "step": 3510}, {"loss": 0.7564, "grad_norm": 0.6463358998298645, "learning_rate": 0.0002, "epoch": 0.5690728316223426, "step": 3520}, {"loss": 0.836, "grad_norm": 0.5286486744880676, "learning_rate": 0.0002, "epoch": 0.5706895158030879, "step": 3530}, {"loss": 0.8421, "grad_norm": 0.5405499935150146, "learning_rate": 0.0002, "epoch": 0.5723061999838331, "step": 3540}, {"loss": 0.7614, "grad_norm": 0.6654391884803772, "learning_rate": 0.0002, "epoch": 0.5739228841645785, "step": 3550}, {"loss": 0.7803, "grad_norm": 0.5081980228424072, "learning_rate": 0.0002, "epoch": 0.5755395683453237, "step": 3560}, {"loss": 0.7753, "grad_norm": 0.48978179693222046, "learning_rate": 0.0002, "epoch": 0.5771562525260691, "step": 3570}, {"loss": 0.8151, "grad_norm": 0.5840612053871155, "learning_rate": 0.0002, "epoch": 0.5787729367068143, "step": 3580}, {"loss": 0.8937, "grad_norm": 0.5235261917114258, "learning_rate": 0.0002, "epoch": 0.5803896208875596, "step": 3590}, {"loss": 0.7894, "grad_norm": 0.5672075748443604, "learning_rate": 0.0002, "epoch": 0.5820063050683049, "step": 3600}, {"loss": 0.8347, "grad_norm": 0.5613429546356201, "learning_rate": 0.0002, "epoch": 0.5836229892490502, "step": 3610}, {"loss": 0.8274, "grad_norm": 0.4032273590564728, "learning_rate": 0.0002, "epoch": 0.5852396734297954, "step": 3620}, {"loss": 0.8421, "grad_norm": 0.49559324979782104, "learning_rate": 0.0002, "epoch": 0.5868563576105408, "step": 3630}, {"loss": 0.8332, "grad_norm": 0.6895697712898254, "learning_rate": 0.0002, "epoch": 0.5884730417912861, "step": 3640}, {"loss": 0.7877, "grad_norm": 0.4750136435031891, "learning_rate": 0.0002, "epoch": 0.5900897259720314, "step": 3650}, {"loss": 0.8219, "grad_norm": 0.5176819562911987, "learning_rate": 0.0002, "epoch": 0.5917064101527767, "step": 3660}, {"loss": 0.8151, "grad_norm": 0.5817760229110718, "learning_rate": 0.0002, "epoch": 0.5933230943335219, "step": 3670}, {"loss": 0.7823, "grad_norm": 0.6064626574516296, "learning_rate": 0.0002, "epoch": 0.5949397785142673, "step": 3680}, {"loss": 0.8422, "grad_norm": 0.6728700995445251, "learning_rate": 0.0002, "epoch": 0.5965564626950125, "step": 3690}, {"loss": 0.7679, "grad_norm": 0.609305202960968, "learning_rate": 0.0002, "epoch": 0.5981731468757578, "step": 3700}, {"loss": 0.8048, "grad_norm": 0.4615488350391388, "learning_rate": 0.0002, "epoch": 0.5997898310565031, "step": 3710}, {"loss": 0.8214, "grad_norm": 2.0531179904937744, "learning_rate": 0.0002, "epoch": 0.6014065152372484, "step": 3720}, {"loss": 0.8158, "grad_norm": 0.5091132521629333, "learning_rate": 0.0002, "epoch": 0.6030231994179936, "step": 3730}, {"loss": 0.7833, "grad_norm": 0.5951124429702759, "learning_rate": 0.0002, "epoch": 0.604639883598739, "step": 3740}, {"loss": 0.7784, "grad_norm": 0.5870208144187927, "learning_rate": 0.0002, "epoch": 0.6062565677794842, "step": 3750}, {"loss": 0.8044, "grad_norm": 0.6254619359970093, "learning_rate": 0.0002, "epoch": 0.6078732519602296, "step": 3760}, {"loss": 0.7868, "grad_norm": 0.5577626824378967, "learning_rate": 0.0002, "epoch": 0.6094899361409749, "step": 3770}, {"loss": 0.8108, "grad_norm": 0.5004405379295349, "learning_rate": 0.0002, "epoch": 0.6111066203217201, "step": 3780}, {"loss": 0.8092, "grad_norm": 0.5527383685112, "learning_rate": 0.0002, "epoch": 0.6127233045024655, "step": 3790}, {"loss": 0.8036, "grad_norm": 0.49116113781929016, "learning_rate": 0.0002, "epoch": 0.6143399886832107, "step": 3800}, {"loss": 0.8352, "grad_norm": 0.5299299359321594, "learning_rate": 0.0002, "epoch": 0.6159566728639561, "step": 3810}, {"loss": 0.7737, "grad_norm": 0.464897483587265, "learning_rate": 0.0002, "epoch": 0.6175733570447013, "step": 3820}, {"loss": 0.7923, "grad_norm": 0.6505740880966187, "learning_rate": 0.0002, "epoch": 0.6191900412254466, "step": 3830}, {"loss": 0.8123, "grad_norm": 0.5512559413909912, "learning_rate": 0.0002, "epoch": 0.6208067254061919, "step": 3840}, {"loss": 0.8856, "grad_norm": 0.49427518248558044, "learning_rate": 0.0002, "epoch": 0.6224234095869372, "step": 3850}, {"loss": 0.7751, "grad_norm": 0.3839147090911865, "learning_rate": 0.0002, "epoch": 0.6240400937676824, "step": 3860}, {"loss": 0.8006, "grad_norm": 0.5760218501091003, "learning_rate": 0.0002, "epoch": 0.6256567779484278, "step": 3870}, {"loss": 0.7836, "grad_norm": 0.7226507067680359, "learning_rate": 0.0002, "epoch": 0.6272734621291731, "step": 3880}, {"loss": 0.8244, "grad_norm": 0.676781415939331, "learning_rate": 0.0002, "epoch": 0.6288901463099184, "step": 3890}, {"loss": 0.8239, "grad_norm": 0.4284018278121948, "learning_rate": 0.0002, "epoch": 0.6305068304906637, "step": 3900}, {"loss": 0.7996, "grad_norm": 0.5060628056526184, "learning_rate": 0.0002, "epoch": 0.6321235146714089, "step": 3910}, {"loss": 0.8089, "grad_norm": 0.5524522066116333, "learning_rate": 0.0002, "epoch": 0.6337401988521543, "step": 3920}, {"loss": 0.8276, "grad_norm": 0.6099881529808044, "learning_rate": 0.0002, "epoch": 0.6353568830328995, "step": 3930}, {"loss": 0.809, "grad_norm": 0.43155938386917114, "learning_rate": 0.0002, "epoch": 0.6369735672136448, "step": 3940}, {"loss": 0.8404, "grad_norm": 0.6427084803581238, "learning_rate": 0.0002, "epoch": 0.6385902513943901, "step": 3950}, {"loss": 0.8368, "grad_norm": 0.541220486164093, "learning_rate": 0.0002, "epoch": 0.6402069355751354, "step": 3960}, {"loss": 0.8539, "grad_norm": 0.5414294600486755, "learning_rate": 0.0002, "epoch": 0.6418236197558806, "step": 3970}, {"loss": 0.7996, "grad_norm": 0.46344003081321716, "learning_rate": 0.0002, "epoch": 0.643440303936626, "step": 3980}, {"loss": 0.7474, "grad_norm": 0.45209285616874695, "learning_rate": 0.0002, "epoch": 0.6450569881173712, "step": 3990}, {"loss": 0.8202, "grad_norm": 0.5417284369468689, "learning_rate": 0.0002, "epoch": 0.6466736722981166, "step": 4000}, {"loss": 0.7563, "grad_norm": 0.7995685935020447, "learning_rate": 0.0002, "epoch": 0.6482903564788619, "step": 4010}, {"loss": 0.7812, "grad_norm": 0.6384002566337585, "learning_rate": 0.0002, "epoch": 0.6499070406596071, "step": 4020}, {"loss": 0.732, "grad_norm": 0.4472815692424774, "learning_rate": 0.0002, "epoch": 0.6515237248403525, "step": 4030}, {"loss": 0.8071, "grad_norm": 0.6834294199943542, "learning_rate": 0.0002, "epoch": 0.6531404090210977, "step": 4040}, {"loss": 0.7812, "grad_norm": 0.4612339735031128, "learning_rate": 0.0002, "epoch": 0.654757093201843, "step": 4050}, {"loss": 0.8141, "grad_norm": 0.9266576170921326, "learning_rate": 0.0002, "epoch": 0.6563737773825883, "step": 4060}, {"loss": 0.7991, "grad_norm": 0.4470861852169037, "learning_rate": 0.0002, "epoch": 0.6579904615633336, "step": 4070}, {"loss": 0.8293, "grad_norm": 0.45544925332069397, "learning_rate": 0.0002, "epoch": 0.6596071457440789, "step": 4080}, {"loss": 0.8455, "grad_norm": 0.6144481301307678, "learning_rate": 0.0002, "epoch": 0.6612238299248242, "step": 4090}, {"loss": 0.7877, "grad_norm": 0.5936288237571716, "learning_rate": 0.0002, "epoch": 0.6628405141055694, "step": 4100}, {"loss": 0.7617, "grad_norm": 0.4822963774204254, "learning_rate": 0.0002, "epoch": 0.6644571982863148, "step": 4110}, {"loss": 0.7997, "grad_norm": 0.48432496190071106, "learning_rate": 0.0002, "epoch": 0.66607388246706, "step": 4120}, {"loss": 0.8404, "grad_norm": 0.4901607930660248, "learning_rate": 0.0002, "epoch": 0.6676905666478054, "step": 4130}, {"loss": 0.8085, "grad_norm": 0.5018393397331238, "learning_rate": 0.0002, "epoch": 0.6693072508285507, "step": 4140}, {"loss": 0.8065, "grad_norm": 0.6946378946304321, "learning_rate": 0.0002, "epoch": 0.6709239350092959, "step": 4150}, {"loss": 0.8147, "grad_norm": 0.5997390747070312, "learning_rate": 0.0002, "epoch": 0.6725406191900413, "step": 4160}, {"loss": 0.8268, "grad_norm": 0.6738849878311157, "learning_rate": 0.0002, "epoch": 0.6741573033707865, "step": 4170}, {"loss": 0.7704, "grad_norm": 0.6110581159591675, "learning_rate": 0.0002, "epoch": 0.6757739875515318, "step": 4180}, {"loss": 0.8043, "grad_norm": 0.5703322291374207, "learning_rate": 0.0002, "epoch": 0.6773906717322771, "step": 4190}, {"loss": 0.8099, "grad_norm": 0.4686066210269928, "learning_rate": 0.0002, "epoch": 0.6790073559130224, "step": 4200}, {"loss": 0.8441, "grad_norm": 0.6394643783569336, "learning_rate": 0.0002, "epoch": 0.6806240400937676, "step": 4210}, {"loss": 0.8011, "grad_norm": 0.5454841256141663, "learning_rate": 0.0002, "epoch": 0.682240724274513, "step": 4220}, {"loss": 0.8307, "grad_norm": 0.4859732985496521, "learning_rate": 0.0002, "epoch": 0.6838574084552582, "step": 4230}, {"loss": 0.8161, "grad_norm": 0.5544065833091736, "learning_rate": 0.0002, "epoch": 0.6854740926360036, "step": 4240}, {"loss": 0.7839, "grad_norm": 0.4902505576610565, "learning_rate": 0.0002, "epoch": 0.6870907768167488, "step": 4250}, {"loss": 0.7977, "grad_norm": 0.4768051505088806, "learning_rate": 0.0002, "epoch": 0.6887074609974941, "step": 4260}, {"loss": 0.7539, "grad_norm": 0.49982190132141113, "learning_rate": 0.0002, "epoch": 0.6903241451782395, "step": 4270}, {"loss": 0.7353, "grad_norm": 0.6351838111877441, "learning_rate": 0.0002, "epoch": 0.6919408293589847, "step": 4280}, {"loss": 0.7664, "grad_norm": 0.5647561550140381, "learning_rate": 0.0002, "epoch": 0.69355751353973, "step": 4290}, {"loss": 0.7618, "grad_norm": 0.5340486764907837, "learning_rate": 0.0002, "epoch": 0.6951741977204753, "step": 4300}, {"loss": 0.8526, "grad_norm": 0.5649092793464661, "learning_rate": 0.0002, "epoch": 0.6967908819012206, "step": 4310}, {"loss": 0.8246, "grad_norm": 0.6183916926383972, "learning_rate": 0.0002, "epoch": 0.6984075660819659, "step": 4320}, {"loss": 0.792, "grad_norm": 0.6154509782791138, "learning_rate": 0.0002, "epoch": 0.7000242502627112, "step": 4330}, {"loss": 0.8397, "grad_norm": 0.5156264305114746, "learning_rate": 0.0002, "epoch": 0.7016409344434564, "step": 4340}, {"loss": 0.8512, "grad_norm": 0.562171459197998, "learning_rate": 0.0002, "epoch": 0.7032576186242018, "step": 4350}, {"loss": 0.7882, "grad_norm": 0.4949502646923065, "learning_rate": 0.0002, "epoch": 0.704874302804947, "step": 4360}, {"loss": 0.738, "grad_norm": 0.5171684622764587, "learning_rate": 0.0002, "epoch": 0.7064909869856923, "step": 4370}, {"loss": 0.8001, "grad_norm": 0.6198443174362183, "learning_rate": 0.0002, "epoch": 0.7081076711664377, "step": 4380}, {"loss": 0.7606, "grad_norm": 0.5802276134490967, "learning_rate": 0.0002, "epoch": 0.7097243553471829, "step": 4390}, {"loss": 0.8797, "grad_norm": 0.41096967458724976, "learning_rate": 0.0002, "epoch": 0.7113410395279283, "step": 4400}, {"loss": 0.805, "grad_norm": 0.4397392272949219, "learning_rate": 0.0002, "epoch": 0.7129577237086735, "step": 4410}, {"loss": 0.7651, "grad_norm": 0.45228442549705505, "learning_rate": 0.0002, "epoch": 0.7145744078894188, "step": 4420}, {"loss": 0.7938, "grad_norm": 0.4839673936367035, "learning_rate": 0.0002, "epoch": 0.7161910920701641, "step": 4430}, {"loss": 0.8362, "grad_norm": 0.6140755414962769, "learning_rate": 0.0002, "epoch": 0.7178077762509094, "step": 4440}, {"loss": 0.7722, "grad_norm": 0.6841378808021545, "learning_rate": 0.0002, "epoch": 0.7194244604316546, "step": 4450}, {"loss": 0.8177, "grad_norm": 0.6664239168167114, "learning_rate": 0.0002, "epoch": 0.7210411446124, "step": 4460}, {"loss": 0.7983, "grad_norm": 0.47552719712257385, "learning_rate": 0.0002, "epoch": 0.7226578287931452, "step": 4470}, {"loss": 0.8982, "grad_norm": 0.6649776101112366, "learning_rate": 0.0002, "epoch": 0.7242745129738906, "step": 4480}, {"loss": 0.8074, "grad_norm": 0.5159541964530945, "learning_rate": 0.0002, "epoch": 0.7258911971546358, "step": 4490}, {"loss": 0.7786, "grad_norm": 0.6693112850189209, "learning_rate": 0.0002, "epoch": 0.7275078813353811, "step": 4500}, {"loss": 0.8655, "grad_norm": 0.48870977759361267, "learning_rate": 0.0002, "epoch": 0.7291245655161265, "step": 4510}, {"loss": 0.7337, "grad_norm": 0.4857887923717499, "learning_rate": 0.0002, "epoch": 0.7307412496968717, "step": 4520}, {"loss": 0.8026, "grad_norm": 0.5515662431716919, "learning_rate": 0.0002, "epoch": 0.732357933877617, "step": 4530}, {"loss": 0.8031, "grad_norm": 0.6292222738265991, "learning_rate": 0.0002, "epoch": 0.7339746180583623, "step": 4540}, {"loss": 0.7749, "grad_norm": 0.48265689611434937, "learning_rate": 0.0002, "epoch": 0.7355913022391076, "step": 4550}, {"loss": 0.8499, "grad_norm": 0.8044266104698181, "learning_rate": 0.0002, "epoch": 0.7372079864198529, "step": 4560}, {"loss": 0.8162, "grad_norm": 0.6111769676208496, "learning_rate": 0.0002, "epoch": 0.7388246706005982, "step": 4570}, {"loss": 0.7291, "grad_norm": 0.5229553580284119, "learning_rate": 0.0002, "epoch": 0.7404413547813434, "step": 4580}, {"loss": 0.8038, "grad_norm": 0.6054152250289917, "learning_rate": 0.0002, "epoch": 0.7420580389620888, "step": 4590}, {"loss": 0.8169, "grad_norm": 0.5574966669082642, "learning_rate": 0.0002, "epoch": 0.743674723142834, "step": 4600}, {"loss": 0.8439, "grad_norm": 0.5395817160606384, "learning_rate": 0.0002, "epoch": 0.7452914073235793, "step": 4610}, {"loss": 0.8495, "grad_norm": 0.7116472721099854, "learning_rate": 0.0002, "epoch": 0.7469080915043246, "step": 4620}, {"loss": 0.7743, "grad_norm": 0.5618700981140137, "learning_rate": 0.0002, "epoch": 0.7485247756850699, "step": 4630}, {"loss": 0.7744, "grad_norm": 0.5802770853042603, "learning_rate": 0.0002, "epoch": 0.7501414598658153, "step": 4640}, {"loss": 0.7924, "grad_norm": 0.5690428018569946, "learning_rate": 0.0002, "epoch": 0.7517581440465605, "step": 4650}, {"loss": 0.8017, "grad_norm": 0.4813360273838043, "learning_rate": 0.0002, "epoch": 0.7533748282273058, "step": 4660}, {"loss": 0.8108, "grad_norm": 0.5434042811393738, "learning_rate": 0.0002, "epoch": 0.7549915124080511, "step": 4670}, {"loss": 0.7824, "grad_norm": 0.5502099990844727, "learning_rate": 0.0002, "epoch": 0.7566081965887964, "step": 4680}, {"loss": 0.8598, "grad_norm": 0.6020621061325073, "learning_rate": 0.0002, "epoch": 0.7582248807695416, "step": 4690}, {"loss": 0.7937, "grad_norm": 0.4922301471233368, "learning_rate": 0.0002, "epoch": 0.759841564950287, "step": 4700}, {"loss": 0.788, "grad_norm": 0.6492828726768494, "learning_rate": 0.0002, "epoch": 0.7614582491310322, "step": 4710}, {"loss": 0.8313, "grad_norm": 0.4865580201148987, "learning_rate": 0.0002, "epoch": 0.7630749333117776, "step": 4720}, {"loss": 0.7966, "grad_norm": 0.5971422791481018, "learning_rate": 0.0002, "epoch": 0.7646916174925228, "step": 4730}, {"loss": 0.8298, "grad_norm": 0.6832674145698547, "learning_rate": 0.0002, "epoch": 0.7663083016732681, "step": 4740}, {"loss": 0.8156, "grad_norm": 0.500908613204956, "learning_rate": 0.0002, "epoch": 0.7679249858540134, "step": 4750}, {"loss": 0.8383, "grad_norm": 0.6112465858459473, "learning_rate": 0.0002, "epoch": 0.7695416700347587, "step": 4760}, {"loss": 0.76, "grad_norm": 0.5753506422042847, "learning_rate": 0.0002, "epoch": 0.771158354215504, "step": 4770}, {"loss": 0.8297, "grad_norm": 0.6529405117034912, "learning_rate": 0.0002, "epoch": 0.7727750383962493, "step": 4780}, {"loss": 0.8171, "grad_norm": 0.5916843414306641, "learning_rate": 0.0002, "epoch": 0.7743917225769946, "step": 4790}, {"loss": 0.83, "grad_norm": 0.4821224510669708, "learning_rate": 0.0002, "epoch": 0.7760084067577399, "step": 4800}, {"loss": 0.7703, "grad_norm": 0.5532580018043518, "learning_rate": 0.0002, "epoch": 0.7776250909384852, "step": 4810}, {"loss": 0.7363, "grad_norm": 0.4604877233505249, "learning_rate": 0.0002, "epoch": 0.7792417751192304, "step": 4820}, {"loss": 0.7506, "grad_norm": 0.5009613037109375, "learning_rate": 0.0002, "epoch": 0.7808584592999758, "step": 4830}, {"loss": 0.7863, "grad_norm": 0.6448560357093811, "learning_rate": 0.0002, "epoch": 0.782475143480721, "step": 4840}, {"loss": 0.7957, "grad_norm": 0.44327953457832336, "learning_rate": 0.0002, "epoch": 0.7840918276614663, "step": 4850}, {"loss": 0.7925, "grad_norm": 0.5355411171913147, "learning_rate": 0.0002, "epoch": 0.7857085118422116, "step": 4860}, {"loss": 0.7754, "grad_norm": 0.5635677576065063, "learning_rate": 0.0002, "epoch": 0.7873251960229569, "step": 4870}, {"loss": 0.7931, "grad_norm": 0.5417491793632507, "learning_rate": 0.0002, "epoch": 0.7889418802037023, "step": 4880}, {"loss": 0.7819, "grad_norm": 0.4567430913448334, "learning_rate": 0.0002, "epoch": 0.7905585643844475, "step": 4890}, {"loss": 0.8454, "grad_norm": 0.44651296734809875, "learning_rate": 0.0002, "epoch": 0.7921752485651928, "step": 4900}, {"loss": 0.7959, "grad_norm": 0.5741217136383057, "learning_rate": 0.0002, "epoch": 0.7937919327459381, "step": 4910}, {"loss": 0.8093, "grad_norm": 0.6605045199394226, "learning_rate": 0.0002, "epoch": 0.7954086169266834, "step": 4920}, {"loss": 0.77, "grad_norm": 0.5126531720161438, "learning_rate": 0.0002, "epoch": 0.7970253011074286, "step": 4930}, {"loss": 0.7793, "grad_norm": 0.513648271560669, "learning_rate": 0.0002, "epoch": 0.798641985288174, "step": 4940}, {"loss": 0.8314, "grad_norm": 0.5350404381752014, "learning_rate": 0.0002, "epoch": 0.8002586694689192, "step": 4950}, {"loss": 0.7649, "grad_norm": 0.5731674432754517, "learning_rate": 0.0002, "epoch": 0.8018753536496646, "step": 4960}, {"loss": 0.8572, "grad_norm": 0.5974258184432983, "learning_rate": 0.0002, "epoch": 0.8034920378304098, "step": 4970}, {"loss": 0.7972, "grad_norm": 0.8774799704551697, "learning_rate": 0.0002, "epoch": 0.8051087220111551, "step": 4980}, {"loss": 0.7899, "grad_norm": 0.5994430184364319, "learning_rate": 0.0002, "epoch": 0.8067254061919004, "step": 4990}, {"loss": 0.7736, "grad_norm": 0.4894903004169464, "learning_rate": 0.0002, "epoch": 0.8083420903726457, "step": 5000}, {"loss": 0.78, "grad_norm": 0.5218459367752075, "learning_rate": 0.0002, "epoch": 0.809958774553391, "step": 5010}, {"loss": 0.817, "grad_norm": 0.5232468843460083, "learning_rate": 0.0002, "epoch": 0.8115754587341363, "step": 5020}, {"loss": 0.7704, "grad_norm": 0.44358372688293457, "learning_rate": 0.0002, "epoch": 0.8131921429148816, "step": 5030}, {"loss": 0.785, "grad_norm": 0.6202037334442139, "learning_rate": 0.0002, "epoch": 0.8148088270956269, "step": 5040}, {"loss": 0.7351, "grad_norm": 0.7721474170684814, "learning_rate": 0.0002, "epoch": 0.8164255112763722, "step": 5050}, {"loss": 0.8297, "grad_norm": 0.5568501353263855, "learning_rate": 0.0002, "epoch": 0.8180421954571174, "step": 5060}, {"loss": 0.7733, "grad_norm": 0.49148809909820557, "learning_rate": 0.0002, "epoch": 0.8196588796378628, "step": 5070}, {"loss": 0.8054, "grad_norm": 0.4956012964248657, "learning_rate": 0.0002, "epoch": 0.821275563818608, "step": 5080}, {"loss": 0.8201, "grad_norm": 0.6078833937644958, "learning_rate": 0.0002, "epoch": 0.8228922479993533, "step": 5090}, {"loss": 0.828, "grad_norm": 0.46906954050064087, "learning_rate": 0.0002, "epoch": 0.8245089321800986, "step": 5100}, {"loss": 0.7703, "grad_norm": 0.50812166929245, "learning_rate": 0.0002, "epoch": 0.8261256163608439, "step": 5110}, {"loss": 0.8243, "grad_norm": 0.5319661498069763, "learning_rate": 0.0002, "epoch": 0.8277423005415891, "step": 5120}, {"loss": 0.7798, "grad_norm": 0.4949689209461212, "learning_rate": 0.0002, "epoch": 0.8293589847223345, "step": 5130}, {"loss": 0.7428, "grad_norm": 0.5151591300964355, "learning_rate": 0.0002, "epoch": 0.8309756689030798, "step": 5140}, {"loss": 0.8147, "grad_norm": 0.5530214309692383, "learning_rate": 0.0002, "epoch": 0.8325923530838251, "step": 5150}, {"loss": 0.8251, "grad_norm": 0.6297410130500793, "learning_rate": 0.0002, "epoch": 0.8342090372645704, "step": 5160}, {"loss": 0.8067, "grad_norm": 0.5466840267181396, "learning_rate": 0.0002, "epoch": 0.8358257214453156, "step": 5170}, {"loss": 0.7875, "grad_norm": 0.652913510799408, "learning_rate": 0.0002, "epoch": 0.837442405626061, "step": 5180}, {"loss": 0.8295, "grad_norm": 0.5811293125152588, "learning_rate": 0.0002, "epoch": 0.8390590898068062, "step": 5190}, {"loss": 0.7412, "grad_norm": 0.5109550952911377, "learning_rate": 0.0002, "epoch": 0.8406757739875516, "step": 5200}, {"loss": 0.8077, "grad_norm": 0.4551706612110138, "learning_rate": 0.0002, "epoch": 0.8422924581682968, "step": 5210}, {"loss": 0.7827, "grad_norm": 0.5813754200935364, "learning_rate": 0.0002, "epoch": 0.8439091423490421, "step": 5220}, {"loss": 0.802, "grad_norm": 0.5856947898864746, "learning_rate": 0.0002, "epoch": 0.8455258265297874, "step": 5230}, {"loss": 0.7957, "grad_norm": 0.5482739210128784, "learning_rate": 0.0002, "epoch": 0.8471425107105327, "step": 5240}, {"loss": 0.8295, "grad_norm": 0.49023720622062683, "learning_rate": 0.0002, "epoch": 0.8487591948912779, "step": 5250}, {"loss": 0.8022, "grad_norm": 0.49472475051879883, "learning_rate": 0.0002, "epoch": 0.8503758790720233, "step": 5260}, {"loss": 0.8001, "grad_norm": 0.5490226745605469, "learning_rate": 0.0002, "epoch": 0.8519925632527686, "step": 5270}, {"loss": 0.8333, "grad_norm": 0.5340665578842163, "learning_rate": 0.0002, "epoch": 0.8536092474335139, "step": 5280}, {"loss": 0.8277, "grad_norm": 0.5962483882904053, "learning_rate": 0.0002, "epoch": 0.8552259316142592, "step": 5290}, {"loss": 0.8765, "grad_norm": 0.586358368396759, "learning_rate": 0.0002, "epoch": 0.8568426157950044, "step": 5300}, {"loss": 0.7831, "grad_norm": 0.49120277166366577, "learning_rate": 0.0002, "epoch": 0.8584592999757498, "step": 5310}, {"loss": 0.8162, "grad_norm": 0.5887332558631897, "learning_rate": 0.0002, "epoch": 0.860075984156495, "step": 5320}, {"loss": 0.7464, "grad_norm": 0.42496153712272644, "learning_rate": 0.0002, "epoch": 0.8616926683372403, "step": 5330}, {"loss": 0.7905, "grad_norm": 0.5489874482154846, "learning_rate": 0.0002, "epoch": 0.8633093525179856, "step": 5340}, {"loss": 0.7958, "grad_norm": 0.5850813984870911, "learning_rate": 0.0002, "epoch": 0.8649260366987309, "step": 5350}, {"loss": 0.7642, "grad_norm": 0.517487108707428, "learning_rate": 0.0002, "epoch": 0.8665427208794761, "step": 5360}, {"loss": 0.7801, "grad_norm": 0.5339142680168152, "learning_rate": 0.0002, "epoch": 0.8681594050602215, "step": 5370}, {"loss": 0.818, "grad_norm": 0.6236387491226196, "learning_rate": 0.0002, "epoch": 0.8697760892409668, "step": 5380}, {"loss": 0.7708, "grad_norm": 0.5752192735671997, "learning_rate": 0.0002, "epoch": 0.8713927734217121, "step": 5390}, {"loss": 0.8542, "grad_norm": 0.6724614500999451, "learning_rate": 0.0002, "epoch": 0.8730094576024574, "step": 5400}, {"loss": 0.7581, "grad_norm": 0.5280613303184509, "learning_rate": 0.0002, "epoch": 0.8746261417832026, "step": 5410}, {"loss": 0.8231, "grad_norm": 0.44033288955688477, "learning_rate": 0.0002, "epoch": 0.876242825963948, "step": 5420}, {"loss": 0.8839, "grad_norm": 0.5199708342552185, "learning_rate": 0.0002, "epoch": 0.8778595101446932, "step": 5430}, {"loss": 0.7852, "grad_norm": 0.46778348088264465, "learning_rate": 0.0002, "epoch": 0.8794761943254386, "step": 5440}, {"loss": 0.7834, "grad_norm": 0.4657754898071289, "learning_rate": 0.0002, "epoch": 0.8810928785061838, "step": 5450}, {"loss": 0.7799, "grad_norm": 0.5472902655601501, "learning_rate": 0.0002, "epoch": 0.8827095626869291, "step": 5460}, {"loss": 0.8253, "grad_norm": 0.4876766800880432, "learning_rate": 0.0002, "epoch": 0.8843262468676744, "step": 5470}, {"loss": 0.7906, "grad_norm": 0.5057248473167419, "learning_rate": 0.0002, "epoch": 0.8859429310484197, "step": 5480}, {"loss": 0.8124, "grad_norm": 0.4637320637702942, "learning_rate": 0.0002, "epoch": 0.8875596152291649, "step": 5490}, {"loss": 0.781, "grad_norm": 0.471955806016922, "learning_rate": 0.0002, "epoch": 0.8891762994099103, "step": 5500}, {"loss": 0.8057, "grad_norm": 0.5209813714027405, "learning_rate": 0.0002, "epoch": 0.8907929835906556, "step": 5510}, {"loss": 0.8106, "grad_norm": 0.6213834285736084, "learning_rate": 0.0002, "epoch": 0.8924096677714008, "step": 5520}, {"loss": 0.7787, "grad_norm": 0.5215408205986023, "learning_rate": 0.0002, "epoch": 0.8940263519521462, "step": 5530}, {"loss": 0.8174, "grad_norm": 0.580478310585022, "learning_rate": 0.0002, "epoch": 0.8956430361328914, "step": 5540}, {"loss": 0.8371, "grad_norm": 0.49102169275283813, "learning_rate": 0.0002, "epoch": 0.8972597203136368, "step": 5550}, {"loss": 0.7806, "grad_norm": 0.6043479442596436, "learning_rate": 0.0002, "epoch": 0.898876404494382, "step": 5560}, {"loss": 0.7754, "grad_norm": 0.5636463165283203, "learning_rate": 0.0002, "epoch": 0.9004930886751273, "step": 5570}, {"loss": 0.8145, "grad_norm": 0.5620124340057373, "learning_rate": 0.0002, "epoch": 0.9021097728558726, "step": 5580}, {"loss": 0.8083, "grad_norm": 0.5206354856491089, "learning_rate": 0.0002, "epoch": 0.9037264570366179, "step": 5590}, {"loss": 0.8557, "grad_norm": 0.5798229575157166, "learning_rate": 0.0002, "epoch": 0.9053431412173631, "step": 5600}, {"loss": 0.8097, "grad_norm": 0.6428212523460388, "learning_rate": 0.0002, "epoch": 0.9069598253981085, "step": 5610}, {"loss": 0.7839, "grad_norm": 0.48064687848091125, "learning_rate": 0.0002, "epoch": 0.9085765095788537, "step": 5620}, {"loss": 0.8343, "grad_norm": 0.6347860097885132, "learning_rate": 0.0002, "epoch": 0.9101931937595991, "step": 5630}, {"loss": 0.851, "grad_norm": 0.5353913307189941, "learning_rate": 0.0002, "epoch": 0.9118098779403444, "step": 5640}, {"loss": 0.7736, "grad_norm": 0.5323944091796875, "learning_rate": 0.0002, "epoch": 0.9134265621210896, "step": 5650}, {"loss": 0.8393, "grad_norm": 0.5261843204498291, "learning_rate": 0.0002, "epoch": 0.915043246301835, "step": 5660}, {"loss": 0.7355, "grad_norm": 0.5451326966285706, "learning_rate": 0.0002, "epoch": 0.9166599304825802, "step": 5670}, {"loss": 0.8012, "grad_norm": 0.5183324217796326, "learning_rate": 0.0002, "epoch": 0.9182766146633256, "step": 5680}, {"loss": 0.7659, "grad_norm": 0.47229018807411194, "learning_rate": 0.0002, "epoch": 0.9198932988440708, "step": 5690}, {"loss": 0.7757, "grad_norm": 0.49180513620376587, "learning_rate": 0.0002, "epoch": 0.9215099830248161, "step": 5700}, {"loss": 0.8735, "grad_norm": 0.5419785380363464, "learning_rate": 0.0002, "epoch": 0.9231266672055614, "step": 5710}, {"loss": 0.7378, "grad_norm": 0.5408698916435242, "learning_rate": 0.0002, "epoch": 0.9247433513863067, "step": 5720}, {"loss": 0.7701, "grad_norm": 0.5286232829093933, "learning_rate": 0.0002, "epoch": 0.9263600355670519, "step": 5730}, {"loss": 0.8242, "grad_norm": 0.7539758086204529, "learning_rate": 0.0002, "epoch": 0.9279767197477973, "step": 5740}, {"loss": 0.8118, "grad_norm": 0.5166944861412048, "learning_rate": 0.0002, "epoch": 0.9295934039285425, "step": 5750}, {"loss": 0.783, "grad_norm": 0.6601425409317017, "learning_rate": 0.0002, "epoch": 0.9312100881092878, "step": 5760}, {"loss": 0.7873, "grad_norm": 0.5029960870742798, "learning_rate": 0.0002, "epoch": 0.9328267722900332, "step": 5770}, {"loss": 0.7989, "grad_norm": 0.4926645755767822, "learning_rate": 0.0002, "epoch": 0.9344434564707784, "step": 5780}, {"loss": 0.8174, "grad_norm": 0.5739615559577942, "learning_rate": 0.0002, "epoch": 0.9360601406515238, "step": 5790}, {"loss": 0.8037, "grad_norm": 0.5058279037475586, "learning_rate": 0.0002, "epoch": 0.937676824832269, "step": 5800}, {"loss": 0.8537, "grad_norm": 0.5260962247848511, "learning_rate": 0.0002, "epoch": 0.9392935090130143, "step": 5810}, {"loss": 0.7486, "grad_norm": 0.5768588185310364, "learning_rate": 0.0002, "epoch": 0.9409101931937596, "step": 5820}, {"loss": 0.8215, "grad_norm": 0.5170126557350159, "learning_rate": 0.0002, "epoch": 0.9425268773745049, "step": 5830}, {"loss": 0.7422, "grad_norm": 0.5745864510536194, "learning_rate": 0.0002, "epoch": 0.9441435615552501, "step": 5840}, {"loss": 0.7824, "grad_norm": 0.5551357865333557, "learning_rate": 0.0002, "epoch": 0.9457602457359955, "step": 5850}, {"loss": 0.8529, "grad_norm": 0.5776078701019287, "learning_rate": 0.0002, "epoch": 0.9473769299167407, "step": 5860}, {"loss": 0.8527, "grad_norm": 0.5340062379837036, "learning_rate": 0.0002, "epoch": 0.9489936140974861, "step": 5870}, {"loss": 0.8217, "grad_norm": 0.6447290182113647, "learning_rate": 0.0002, "epoch": 0.9506102982782314, "step": 5880}, {"loss": 0.7945, "grad_norm": 0.5123815536499023, "learning_rate": 0.0002, "epoch": 0.9522269824589766, "step": 5890}, {"loss": 0.8209, "grad_norm": 0.48547613620758057, "learning_rate": 0.0002, "epoch": 0.953843666639722, "step": 5900}, {"loss": 0.7896, "grad_norm": 0.5791414976119995, "learning_rate": 0.0002, "epoch": 0.9554603508204672, "step": 5910}, {"loss": 0.8408, "grad_norm": 0.6195011734962463, "learning_rate": 0.0002, "epoch": 0.9570770350012126, "step": 5920}, {"loss": 0.7805, "grad_norm": 0.6323803067207336, "learning_rate": 0.0002, "epoch": 0.9586937191819578, "step": 5930}, {"loss": 0.8484, "grad_norm": 0.45552879571914673, "learning_rate": 0.0002, "epoch": 0.9603104033627031, "step": 5940}, {"loss": 0.7367, "grad_norm": 0.5796473622322083, "learning_rate": 0.0002, "epoch": 0.9619270875434484, "step": 5950}, {"loss": 0.7672, "grad_norm": 0.647261381149292, "learning_rate": 0.0002, "epoch": 0.9635437717241937, "step": 5960}, {"loss": 0.8086, "grad_norm": 0.5487682819366455, "learning_rate": 0.0002, "epoch": 0.9651604559049389, "step": 5970}, {"loss": 0.7973, "grad_norm": 0.5743663907051086, "learning_rate": 0.0002, "epoch": 0.9667771400856843, "step": 5980}, {"loss": 0.8153, "grad_norm": 0.5470591187477112, "learning_rate": 0.0002, "epoch": 0.9683938242664295, "step": 5990}, {"loss": 0.8119, "grad_norm": 0.5901660323143005, "learning_rate": 0.0002, "epoch": 0.9700105084471748, "step": 6000}, {"loss": 0.8147, "grad_norm": 0.6544759273529053, "learning_rate": 0.0002, "epoch": 0.9716271926279202, "step": 6010}, {"loss": 0.7536, "grad_norm": 0.6288470029830933, "learning_rate": 0.0002, "epoch": 0.9732438768086654, "step": 6020}, {"loss": 0.7989, "grad_norm": 0.673153817653656, "learning_rate": 0.0002, "epoch": 0.9748605609894108, "step": 6030}, {"loss": 0.7556, "grad_norm": 0.42854753136634827, "learning_rate": 0.0002, "epoch": 0.976477245170156, "step": 6040}, {"loss": 0.8006, "grad_norm": 0.5227066278457642, "learning_rate": 0.0002, "epoch": 0.9780939293509013, "step": 6050}, {"loss": 0.795, "grad_norm": 0.5372416973114014, "learning_rate": 0.0002, "epoch": 0.9797106135316466, "step": 6060}, {"loss": 0.7591, "grad_norm": 0.6026402115821838, "learning_rate": 0.0002, "epoch": 0.9813272977123919, "step": 6070}, {"loss": 0.8347, "grad_norm": 0.49547791481018066, "learning_rate": 0.0002, "epoch": 0.9829439818931371, "step": 6080}, {"loss": 0.7722, "grad_norm": 0.4641951322555542, "learning_rate": 0.0002, "epoch": 0.9845606660738825, "step": 6090}, {"loss": 0.8125, "grad_norm": 0.5818535089492798, "learning_rate": 0.0002, "epoch": 0.9861773502546277, "step": 6100}, {"loss": 0.81, "grad_norm": 0.63955157995224, "learning_rate": 0.0002, "epoch": 0.9877940344353731, "step": 6110}, {"loss": 0.7547, "grad_norm": 0.5649438500404358, "learning_rate": 0.0002, "epoch": 0.9894107186161183, "step": 6120}, {"loss": 0.7861, "grad_norm": 0.5290433168411255, "learning_rate": 0.0002, "epoch": 0.9910274027968636, "step": 6130}, {"loss": 0.8109, "grad_norm": 0.6399374008178711, "learning_rate": 0.0002, "epoch": 0.992644086977609, "step": 6140}, {"loss": 0.8373, "grad_norm": 0.6736576557159424, "learning_rate": 0.0002, "epoch": 0.9942607711583542, "step": 6150}, {"loss": 0.7915, "grad_norm": 0.515420138835907, "learning_rate": 0.0002, "epoch": 0.9958774553390995, "step": 6160}, {"loss": 0.8032, "grad_norm": 0.562677800655365, "learning_rate": 0.0002, "epoch": 0.9974941395198448, "step": 6170}, {"loss": 0.8187, "grad_norm": 0.7113858461380005, "learning_rate": 0.0002, "epoch": 0.9991108237005901, "step": 6180}]}
+{"epoch": 2.0, "step": 12371, "epoch_duration": 16880.00914287567, "total_accumulated_duration": 33750.92989253998, "gpu_info": {"GPU_0": "NVIDIA A100-PCIE-40GB"}, "memory_usage": {"avg_memory_usage": {"GPU_0": 7568.4541015625}, "peak_memory_usage": {"GPU_0": 13792.75537109375}, "avg_memory_reserved": {"GPU_0": 17416.0}, "peak_memory_reserved": {"GPU_0": 17416.0}, "total_memory": {"GPU_0": 40444.375}}, "best_checkpoint_path": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185", "params": {"epochs": 8, "batch_size": 1, "learning_rate": 0.0002, "gradient_accumulation_steps": 2, "warmup_ratio": 0.03, "max_grad_norm": 0.3, "lora_alpha": 32, "lora_dropout": 0.05, "lora_r": 64, "tf32": true, "seed": 42}, "log_history": [{"loss": 1.6636, "grad_norm": 0.9894065856933594, "learning_rate": 0.0002, "epoch": 0.0016166841807452913, "step": 10}, {"loss": 1.1528, "grad_norm": 1.7810699939727783, "learning_rate": 0.0002, "epoch": 0.0032333683614905826, "step": 20}, {"loss": 0.9767, "grad_norm": 0.5969577431678772, "learning_rate": 0.0002, "epoch": 0.004850052542235874, "step": 30}, {"loss": 0.9772, "grad_norm": 0.6354120969772339, "learning_rate": 0.0002, "epoch": 0.006466736722981165, "step": 40}, {"loss": 0.8643, "grad_norm": 0.5604607462882996, "learning_rate": 0.0002, "epoch": 0.008083420903726457, "step": 50}, {"loss": 0.8841, "grad_norm": 0.4676193594932556, "learning_rate": 0.0002, "epoch": 0.009700105084471748, "step": 60}, {"loss": 0.9022, "grad_norm": 0.6099211573600769, "learning_rate": 0.0002, "epoch": 0.01131678926521704, "step": 70}, {"loss": 0.9133, "grad_norm": 0.48639994859695435, "learning_rate": 0.0002, "epoch": 0.01293347344596233, "step": 80}, {"loss": 0.8704, "grad_norm": 0.4904264509677887, "learning_rate": 0.0002, "epoch": 0.014550157626707623, "step": 90}, {"loss": 0.8855, "grad_norm": 2.8334362506866455, "learning_rate": 0.0002, "epoch": 0.016166841807452915, "step": 100}, {"loss": 0.8958, "grad_norm": 0.43221670389175415, "learning_rate": 0.0002, "epoch": 0.017783525988198205, "step": 110}, {"loss": 0.8412, "grad_norm": 0.42244166135787964, "learning_rate": 0.0002, "epoch": 0.019400210168943496, "step": 120}, {"loss": 0.8467, "grad_norm": 0.45363298058509827, "learning_rate": 0.0002, "epoch": 0.02101689434968879, "step": 130}, {"loss": 0.8641, "grad_norm": 0.44816508889198303, "learning_rate": 0.0002, "epoch": 0.02263357853043408, "step": 140}, {"loss": 0.8496, "grad_norm": 0.43308213353157043, "learning_rate": 0.0002, "epoch": 0.02425026271117937, "step": 150}, {"loss": 0.8213, "grad_norm": 0.4084763526916504, "learning_rate": 0.0002, "epoch": 0.02586694689192466, "step": 160}, {"loss": 0.8343, "grad_norm": 0.5363703966140747, "learning_rate": 0.0002, "epoch": 0.027483631072669955, "step": 170}, {"loss": 0.8558, "grad_norm": 0.4619699716567993, "learning_rate": 0.0002, "epoch": 0.029100315253415245, "step": 180}, {"loss": 0.8878, "grad_norm": 0.49069908261299133, "learning_rate": 0.0002, "epoch": 0.030716999434160536, "step": 190}, {"loss": 0.8867, "grad_norm": 0.4645835757255554, "learning_rate": 0.0002, "epoch": 0.03233368361490583, "step": 200}, {"loss": 0.8842, "grad_norm": 1.2411243915557861, "learning_rate": 0.0002, "epoch": 0.03395036779565112, "step": 210}, {"loss": 0.8245, "grad_norm": 0.5211851596832275, "learning_rate": 0.0002, "epoch": 0.03556705197639641, "step": 220}, {"loss": 0.8194, "grad_norm": 0.5253691673278809, "learning_rate": 0.0002, "epoch": 0.037183736157141704, "step": 230}, {"loss": 0.8856, "grad_norm": 0.4567478895187378, "learning_rate": 0.0002, "epoch": 0.03880042033788699, "step": 240}, {"loss": 0.838, "grad_norm": 0.5472128391265869, "learning_rate": 0.0002, "epoch": 0.040417104518632285, "step": 250}, {"loss": 0.8201, "grad_norm": 0.42978546023368835, "learning_rate": 0.0002, "epoch": 0.04203378869937758, "step": 260}, {"loss": 0.8334, "grad_norm": 0.601734459400177, "learning_rate": 0.0002, "epoch": 0.043650472880122866, "step": 270}, {"loss": 0.815, "grad_norm": 0.4286513328552246, "learning_rate": 0.0002, "epoch": 0.04526715706086816, "step": 280}, {"loss": 0.8758, "grad_norm": 0.5230861902236938, "learning_rate": 0.0002, "epoch": 0.046883841241613454, "step": 290}, {"loss": 0.8636, "grad_norm": 0.6504611968994141, "learning_rate": 0.0002, "epoch": 0.04850052542235874, "step": 300}, {"loss": 0.8102, "grad_norm": 0.43485215306282043, "learning_rate": 0.0002, "epoch": 0.050117209603104035, "step": 310}, {"loss": 0.8221, "grad_norm": 0.4717007875442505, "learning_rate": 0.0002, "epoch": 0.05173389378384932, "step": 320}, {"loss": 0.8469, "grad_norm": 0.4059787690639496, "learning_rate": 0.0002, "epoch": 0.053350577964594616, "step": 330}, {"loss": 0.8866, "grad_norm": 0.4366913437843323, "learning_rate": 0.0002, "epoch": 0.05496726214533991, "step": 340}, {"loss": 0.7976, "grad_norm": 0.4233848452568054, "learning_rate": 0.0002, "epoch": 0.0565839463260852, "step": 350}, {"loss": 0.8456, "grad_norm": 0.4209108352661133, "learning_rate": 0.0002, "epoch": 0.05820063050683049, "step": 360}, {"loss": 0.816, "grad_norm": 0.41637396812438965, "learning_rate": 0.0002, "epoch": 0.059817314687575784, "step": 370}, {"loss": 0.7976, "grad_norm": 0.46235376596450806, "learning_rate": 0.0002, "epoch": 0.06143399886832107, "step": 380}, {"loss": 0.7966, "grad_norm": 0.4013484716415405, "learning_rate": 0.0002, "epoch": 0.06305068304906636, "step": 390}, {"loss": 0.8253, "grad_norm": 0.47443896532058716, "learning_rate": 0.0002, "epoch": 0.06466736722981166, "step": 400}, {"loss": 0.8666, "grad_norm": 0.3942156434059143, "learning_rate": 0.0002, "epoch": 0.06628405141055695, "step": 410}, {"loss": 0.8402, "grad_norm": 0.4965320825576782, "learning_rate": 0.0002, "epoch": 0.06790073559130223, "step": 420}, {"loss": 0.8317, "grad_norm": 0.4304835796356201, "learning_rate": 0.0002, "epoch": 0.06951741977204753, "step": 430}, {"loss": 0.8528, "grad_norm": 0.511726975440979, "learning_rate": 0.0002, "epoch": 0.07113410395279282, "step": 440}, {"loss": 0.8675, "grad_norm": 0.4040689170360565, "learning_rate": 0.0002, "epoch": 0.07275078813353811, "step": 450}, {"loss": 0.8788, "grad_norm": 0.5402171015739441, "learning_rate": 0.0002, "epoch": 0.07436747231428341, "step": 460}, {"loss": 0.8737, "grad_norm": 0.4174517095088959, "learning_rate": 0.0002, "epoch": 0.0759841564950287, "step": 470}, {"loss": 0.7605, "grad_norm": 0.4306182265281677, "learning_rate": 0.0002, "epoch": 0.07760084067577398, "step": 480}, {"loss": 0.799, "grad_norm": 0.535210132598877, "learning_rate": 0.0002, "epoch": 0.07921752485651928, "step": 490}, {"loss": 0.7825, "grad_norm": 0.5339109897613525, "learning_rate": 0.0002, "epoch": 0.08083420903726457, "step": 500}, {"loss": 0.8985, "grad_norm": 0.45754891633987427, "learning_rate": 0.0002, "epoch": 0.08245089321800986, "step": 510}, {"loss": 0.8144, "grad_norm": 0.43820783495903015, "learning_rate": 0.0002, "epoch": 0.08406757739875516, "step": 520}, {"loss": 0.8001, "grad_norm": 0.4434749186038971, "learning_rate": 0.0002, "epoch": 0.08568426157950045, "step": 530}, {"loss": 0.7857, "grad_norm": 0.43111467361450195, "learning_rate": 0.0002, "epoch": 0.08730094576024573, "step": 540}, {"loss": 0.8418, "grad_norm": 0.4378940165042877, "learning_rate": 0.0002, "epoch": 0.08891762994099103, "step": 550}, {"loss": 0.8361, "grad_norm": 0.4772215187549591, "learning_rate": 0.0002, "epoch": 0.09053431412173632, "step": 560}, {"loss": 0.8268, "grad_norm": 0.6837629079818726, "learning_rate": 0.0002, "epoch": 0.09215099830248161, "step": 570}, {"loss": 0.8607, "grad_norm": 0.42241212725639343, "learning_rate": 0.0002, "epoch": 0.09376768248322691, "step": 580}, {"loss": 0.852, "grad_norm": 0.5165936350822449, "learning_rate": 0.0002, "epoch": 0.0953843666639722, "step": 590}, {"loss": 0.8664, "grad_norm": 0.48737478256225586, "learning_rate": 0.0002, "epoch": 0.09700105084471748, "step": 600}, {"loss": 0.8806, "grad_norm": 0.47419852018356323, "learning_rate": 0.0002, "epoch": 0.09861773502546278, "step": 610}, {"loss": 0.8254, "grad_norm": 0.4975486099720001, "learning_rate": 0.0002, "epoch": 0.10023441920620807, "step": 620}, {"loss": 0.8548, "grad_norm": 0.49123844504356384, "learning_rate": 0.0002, "epoch": 0.10185110338695336, "step": 630}, {"loss": 0.8911, "grad_norm": 0.6288952827453613, "learning_rate": 0.0002, "epoch": 0.10346778756769864, "step": 640}, {"loss": 0.827, "grad_norm": 0.4277345836162567, "learning_rate": 0.0002, "epoch": 0.10508447174844394, "step": 650}, {"loss": 0.7996, "grad_norm": 0.4021061956882477, "learning_rate": 0.0002, "epoch": 0.10670115592918923, "step": 660}, {"loss": 0.87, "grad_norm": 0.3492237329483032, "learning_rate": 0.0002, "epoch": 0.10831784010993452, "step": 670}, {"loss": 0.8698, "grad_norm": 0.4341012239456177, "learning_rate": 0.0002, "epoch": 0.10993452429067982, "step": 680}, {"loss": 0.781, "grad_norm": 0.7296304106712341, "learning_rate": 0.0002, "epoch": 0.1115512084714251, "step": 690}, {"loss": 0.8433, "grad_norm": 0.397494912147522, "learning_rate": 0.0002, "epoch": 0.1131678926521704, "step": 700}, {"loss": 0.827, "grad_norm": 0.396431028842926, "learning_rate": 0.0002, "epoch": 0.1147845768329157, "step": 710}, {"loss": 0.8379, "grad_norm": 0.48842838406562805, "learning_rate": 0.0002, "epoch": 0.11640126101366098, "step": 720}, {"loss": 0.8238, "grad_norm": 0.46322616934776306, "learning_rate": 0.0002, "epoch": 0.11801794519440627, "step": 730}, {"loss": 0.8041, "grad_norm": 0.47990912199020386, "learning_rate": 0.0002, "epoch": 0.11963462937515157, "step": 740}, {"loss": 0.82, "grad_norm": 0.4997142255306244, "learning_rate": 0.0002, "epoch": 0.12125131355589686, "step": 750}, {"loss": 0.7702, "grad_norm": 0.4040526747703552, "learning_rate": 0.0002, "epoch": 0.12286799773664214, "step": 760}, {"loss": 0.863, "grad_norm": 0.453095942735672, "learning_rate": 0.0002, "epoch": 0.12448468191738744, "step": 770}, {"loss": 0.8792, "grad_norm": 0.4636971950531006, "learning_rate": 0.0002, "epoch": 0.12610136609813272, "step": 780}, {"loss": 0.8112, "grad_norm": 0.4279276132583618, "learning_rate": 0.0002, "epoch": 0.12771805027887803, "step": 790}, {"loss": 0.8711, "grad_norm": 0.46212655305862427, "learning_rate": 0.0002, "epoch": 0.12933473445962332, "step": 800}, {"loss": 0.8368, "grad_norm": 0.43127650022506714, "learning_rate": 0.0002, "epoch": 0.1309514186403686, "step": 810}, {"loss": 0.8476, "grad_norm": 0.4201301336288452, "learning_rate": 0.0002, "epoch": 0.1325681028211139, "step": 820}, {"loss": 0.8078, "grad_norm": 0.42583167552948, "learning_rate": 0.0002, "epoch": 0.13418478700185918, "step": 830}, {"loss": 0.8219, "grad_norm": 0.4535622000694275, "learning_rate": 0.0002, "epoch": 0.13580147118260447, "step": 840}, {"loss": 0.8423, "grad_norm": 0.4116036891937256, "learning_rate": 0.0002, "epoch": 0.13741815536334978, "step": 850}, {"loss": 0.8466, "grad_norm": 0.45997580885887146, "learning_rate": 0.0002, "epoch": 0.13903483954409507, "step": 860}, {"loss": 0.8917, "grad_norm": 0.4487837255001068, "learning_rate": 0.0002, "epoch": 0.14065152372484035, "step": 870}, {"loss": 0.8217, "grad_norm": 0.43650057911872864, "learning_rate": 0.0002, "epoch": 0.14226820790558564, "step": 880}, {"loss": 0.8178, "grad_norm": 0.5335358381271362, "learning_rate": 0.0002, "epoch": 0.14388489208633093, "step": 890}, {"loss": 0.7957, "grad_norm": 0.5989000201225281, "learning_rate": 0.0002, "epoch": 0.14550157626707622, "step": 900}, {"loss": 0.8385, "grad_norm": 0.517179012298584, "learning_rate": 0.0002, "epoch": 0.14711826044782153, "step": 910}, {"loss": 0.8255, "grad_norm": 0.44435232877731323, "learning_rate": 0.0002, "epoch": 0.14873494462856682, "step": 920}, {"loss": 0.8305, "grad_norm": 0.42635923624038696, "learning_rate": 0.0002, "epoch": 0.1503516288093121, "step": 930}, {"loss": 0.8043, "grad_norm": 0.49603334069252014, "learning_rate": 0.0002, "epoch": 0.1519683129900574, "step": 940}, {"loss": 0.8377, "grad_norm": 0.40639808773994446, "learning_rate": 0.0002, "epoch": 0.15358499717080268, "step": 950}, {"loss": 0.8529, "grad_norm": 0.4850759208202362, "learning_rate": 0.0002, "epoch": 0.15520168135154797, "step": 960}, {"loss": 0.846, "grad_norm": 0.4427442252635956, "learning_rate": 0.0002, "epoch": 0.15681836553229328, "step": 970}, {"loss": 0.8705, "grad_norm": 0.3760930001735687, "learning_rate": 0.0002, "epoch": 0.15843504971303857, "step": 980}, {"loss": 0.8644, "grad_norm": 0.4794144332408905, "learning_rate": 0.0002, "epoch": 0.16005173389378385, "step": 990}, {"loss": 0.8002, "grad_norm": 0.45828768610954285, "learning_rate": 0.0002, "epoch": 0.16166841807452914, "step": 1000}, {"loss": 0.7658, "grad_norm": 0.6313053369522095, "learning_rate": 0.0002, "epoch": 0.16328510225527443, "step": 1010}, {"loss": 0.8047, "grad_norm": 0.45041006803512573, "learning_rate": 0.0002, "epoch": 0.16490178643601971, "step": 1020}, {"loss": 0.8423, "grad_norm": 0.441403865814209, "learning_rate": 0.0002, "epoch": 0.166518470616765, "step": 1030}, {"loss": 0.8475, "grad_norm": 0.8171296119689941, "learning_rate": 0.0002, "epoch": 0.16813515479751032, "step": 1040}, {"loss": 0.845, "grad_norm": 0.7137420773506165, "learning_rate": 0.0002, "epoch": 0.1697518389782556, "step": 1050}, {"loss": 0.8213, "grad_norm": 0.5236809849739075, "learning_rate": 0.0002, "epoch": 0.1713685231590009, "step": 1060}, {"loss": 0.8265, "grad_norm": 0.5021864175796509, "learning_rate": 0.0002, "epoch": 0.17298520733974618, "step": 1070}, {"loss": 0.8305, "grad_norm": 0.47347521781921387, "learning_rate": 0.0002, "epoch": 0.17460189152049146, "step": 1080}, {"loss": 0.8105, "grad_norm": 0.4631653428077698, "learning_rate": 0.0002, "epoch": 0.17621857570123675, "step": 1090}, {"loss": 0.8166, "grad_norm": 0.49169182777404785, "learning_rate": 0.0002, "epoch": 0.17783525988198207, "step": 1100}, {"loss": 0.8012, "grad_norm": 0.5019739270210266, "learning_rate": 0.0002, "epoch": 0.17945194406272735, "step": 1110}, {"loss": 0.8247, "grad_norm": 0.5100422501564026, "learning_rate": 0.0002, "epoch": 0.18106862824347264, "step": 1120}, {"loss": 0.8142, "grad_norm": 0.3888324499130249, "learning_rate": 0.0002, "epoch": 0.18268531242421793, "step": 1130}, {"loss": 0.8533, "grad_norm": 0.39765217900276184, "learning_rate": 0.0002, "epoch": 0.18430199660496321, "step": 1140}, {"loss": 0.8541, "grad_norm": 0.47190186381340027, "learning_rate": 0.0002, "epoch": 0.1859186807857085, "step": 1150}, {"loss": 0.8301, "grad_norm": 0.4464188814163208, "learning_rate": 0.0002, "epoch": 0.18753536496645382, "step": 1160}, {"loss": 0.8341, "grad_norm": 0.5153930187225342, "learning_rate": 0.0002, "epoch": 0.1891520491471991, "step": 1170}, {"loss": 0.8033, "grad_norm": 0.4779708683490753, "learning_rate": 0.0002, "epoch": 0.1907687333279444, "step": 1180}, {"loss": 0.8187, "grad_norm": 0.4834315776824951, "learning_rate": 0.0002, "epoch": 0.19238541750868968, "step": 1190}, {"loss": 0.7721, "grad_norm": 0.402357816696167, "learning_rate": 0.0002, "epoch": 0.19400210168943496, "step": 1200}, {"loss": 0.7941, "grad_norm": 0.45899084210395813, "learning_rate": 0.0002, "epoch": 0.19561878587018025, "step": 1210}, {"loss": 0.8353, "grad_norm": 0.5106529593467712, "learning_rate": 0.0002, "epoch": 0.19723547005092557, "step": 1220}, {"loss": 0.7816, "grad_norm": 0.45261722803115845, "learning_rate": 0.0002, "epoch": 0.19885215423167085, "step": 1230}, {"loss": 0.8068, "grad_norm": 0.4647127091884613, "learning_rate": 0.0002, "epoch": 0.20046883841241614, "step": 1240}, {"loss": 0.8239, "grad_norm": 0.4849368929862976, "learning_rate": 0.0002, "epoch": 0.20208552259316143, "step": 1250}, {"loss": 0.8514, "grad_norm": 0.4518061578273773, "learning_rate": 0.0002, "epoch": 0.2037022067739067, "step": 1260}, {"loss": 0.8158, "grad_norm": 0.49535325169563293, "learning_rate": 0.0002, "epoch": 0.205318890954652, "step": 1270}, {"loss": 0.8348, "grad_norm": 0.4835205376148224, "learning_rate": 0.0002, "epoch": 0.2069355751353973, "step": 1280}, {"loss": 0.8428, "grad_norm": 0.45308539271354675, "learning_rate": 0.0002, "epoch": 0.2085522593161426, "step": 1290}, {"loss": 0.7993, "grad_norm": 0.5369905233383179, "learning_rate": 0.0002, "epoch": 0.2101689434968879, "step": 1300}, {"loss": 0.8676, "grad_norm": 0.5031622052192688, "learning_rate": 0.0002, "epoch": 0.21178562767763318, "step": 1310}, {"loss": 0.7686, "grad_norm": 0.48010334372520447, "learning_rate": 0.0002, "epoch": 0.21340231185837846, "step": 1320}, {"loss": 0.806, "grad_norm": 0.4905701279640198, "learning_rate": 0.0002, "epoch": 0.21501899603912375, "step": 1330}, {"loss": 0.7885, "grad_norm": 0.43531742691993713, "learning_rate": 0.0002, "epoch": 0.21663568021986904, "step": 1340}, {"loss": 0.8191, "grad_norm": 0.44330692291259766, "learning_rate": 0.0002, "epoch": 0.21825236440061435, "step": 1350}, {"loss": 0.8205, "grad_norm": 0.5384416580200195, "learning_rate": 0.0002, "epoch": 0.21986904858135964, "step": 1360}, {"loss": 0.7726, "grad_norm": 0.4181833863258362, "learning_rate": 0.0002, "epoch": 0.22148573276210493, "step": 1370}, {"loss": 0.8311, "grad_norm": 0.523833692073822, "learning_rate": 0.0002, "epoch": 0.2231024169428502, "step": 1380}, {"loss": 0.7913, "grad_norm": 0.5528736710548401, "learning_rate": 0.0002, "epoch": 0.2247191011235955, "step": 1390}, {"loss": 0.8079, "grad_norm": 0.43515023589134216, "learning_rate": 0.0002, "epoch": 0.2263357853043408, "step": 1400}, {"loss": 0.8403, "grad_norm": 0.48809877038002014, "learning_rate": 0.0002, "epoch": 0.2279524694850861, "step": 1410}, {"loss": 0.8165, "grad_norm": 0.43591251969337463, "learning_rate": 0.0002, "epoch": 0.2295691536658314, "step": 1420}, {"loss": 0.8147, "grad_norm": 0.44625312089920044, "learning_rate": 0.0002, "epoch": 0.23118583784657668, "step": 1430}, {"loss": 0.8134, "grad_norm": 0.4390665292739868, "learning_rate": 0.0002, "epoch": 0.23280252202732196, "step": 1440}, {"loss": 0.8465, "grad_norm": 0.48496049642562866, "learning_rate": 0.0002, "epoch": 0.23441920620806725, "step": 1450}, {"loss": 0.775, "grad_norm": 0.45919957756996155, "learning_rate": 0.0002, "epoch": 0.23603589038881254, "step": 1460}, {"loss": 0.8659, "grad_norm": 0.5471845865249634, "learning_rate": 0.0002, "epoch": 0.23765257456955785, "step": 1470}, {"loss": 0.8164, "grad_norm": 0.47269317507743835, "learning_rate": 0.0002, "epoch": 0.23926925875030314, "step": 1480}, {"loss": 0.854, "grad_norm": 0.4930245578289032, "learning_rate": 0.0002, "epoch": 0.24088594293104842, "step": 1490}, {"loss": 0.8139, "grad_norm": 0.5605630278587341, "learning_rate": 0.0002, "epoch": 0.2425026271117937, "step": 1500}, {"loss": 0.8125, "grad_norm": 0.4435870945453644, "learning_rate": 0.0002, "epoch": 0.244119311292539, "step": 1510}, {"loss": 0.8123, "grad_norm": 0.4941999912261963, "learning_rate": 0.0002, "epoch": 0.24573599547328429, "step": 1520}, {"loss": 0.8427, "grad_norm": 0.5100624561309814, "learning_rate": 0.0002, "epoch": 0.24735267965402957, "step": 1530}, {"loss": 0.8405, "grad_norm": 0.4638267457485199, "learning_rate": 0.0002, "epoch": 0.2489693638347749, "step": 1540}, {"loss": 0.81, "grad_norm": 0.5071570873260498, "learning_rate": 0.0002, "epoch": 0.25058604801552015, "step": 1550}, {"loss": 0.7724, "grad_norm": 0.4291319251060486, "learning_rate": 0.0002, "epoch": 0.25220273219626543, "step": 1560}, {"loss": 0.7984, "grad_norm": 0.5388049483299255, "learning_rate": 0.0002, "epoch": 0.2538194163770108, "step": 1570}, {"loss": 0.8176, "grad_norm": 0.5083683729171753, "learning_rate": 0.0002, "epoch": 0.25543610055775606, "step": 1580}, {"loss": 0.843, "grad_norm": 0.4824463725090027, "learning_rate": 0.0002, "epoch": 0.25705278473850135, "step": 1590}, {"loss": 0.7996, "grad_norm": 0.41177722811698914, "learning_rate": 0.0002, "epoch": 0.25866946891924664, "step": 1600}, {"loss": 0.7772, "grad_norm": 0.5656219124794006, "learning_rate": 0.0002, "epoch": 0.2602861530999919, "step": 1610}, {"loss": 0.7955, "grad_norm": 0.41063204407691956, "learning_rate": 0.0002, "epoch": 0.2619028372807372, "step": 1620}, {"loss": 0.7998, "grad_norm": 0.4897061288356781, "learning_rate": 0.0002, "epoch": 0.2635195214614825, "step": 1630}, {"loss": 0.8198, "grad_norm": 0.4454376697540283, "learning_rate": 0.0002, "epoch": 0.2651362056422278, "step": 1640}, {"loss": 0.8684, "grad_norm": 0.4355238378047943, "learning_rate": 0.0002, "epoch": 0.26675288982297307, "step": 1650}, {"loss": 0.7801, "grad_norm": 0.458310067653656, "learning_rate": 0.0002, "epoch": 0.26836957400371836, "step": 1660}, {"loss": 0.7935, "grad_norm": 0.4752083718776703, "learning_rate": 0.0002, "epoch": 0.26998625818446365, "step": 1670}, {"loss": 0.8267, "grad_norm": 0.4666106402873993, "learning_rate": 0.0002, "epoch": 0.27160294236520893, "step": 1680}, {"loss": 0.8252, "grad_norm": 0.4213818609714508, "learning_rate": 0.0002, "epoch": 0.2732196265459543, "step": 1690}, {"loss": 0.8559, "grad_norm": 0.5768913626670837, "learning_rate": 0.0002, "epoch": 0.27483631072669956, "step": 1700}, {"loss": 0.7931, "grad_norm": 0.4209914803504944, "learning_rate": 0.0002, "epoch": 0.27645299490744485, "step": 1710}, {"loss": 0.8167, "grad_norm": 0.501909613609314, "learning_rate": 0.0002, "epoch": 0.27806967908819014, "step": 1720}, {"loss": 0.7832, "grad_norm": 0.5266261100769043, "learning_rate": 0.0002, "epoch": 0.2796863632689354, "step": 1730}, {"loss": 0.8102, "grad_norm": 0.43806859850883484, "learning_rate": 0.0002, "epoch": 0.2813030474496807, "step": 1740}, {"loss": 0.8157, "grad_norm": 0.46048814058303833, "learning_rate": 0.0002, "epoch": 0.282919731630426, "step": 1750}, {"loss": 0.8596, "grad_norm": 0.44972819089889526, "learning_rate": 0.0002, "epoch": 0.2845364158111713, "step": 1760}, {"loss": 0.8421, "grad_norm": 0.5114831328392029, "learning_rate": 0.0002, "epoch": 0.28615309999191657, "step": 1770}, {"loss": 0.8361, "grad_norm": 0.47931742668151855, "learning_rate": 0.0002, "epoch": 0.28776978417266186, "step": 1780}, {"loss": 0.8265, "grad_norm": 0.5092599987983704, "learning_rate": 0.0002, "epoch": 0.28938646835340714, "step": 1790}, {"loss": 0.8506, "grad_norm": 0.37581443786621094, "learning_rate": 0.0002, "epoch": 0.29100315253415243, "step": 1800}, {"loss": 0.7932, "grad_norm": 0.47097381949424744, "learning_rate": 0.0002, "epoch": 0.2926198367148977, "step": 1810}, {"loss": 0.7787, "grad_norm": 0.48300236463546753, "learning_rate": 0.0002, "epoch": 0.29423652089564306, "step": 1820}, {"loss": 0.8391, "grad_norm": 0.5600419640541077, "learning_rate": 0.0002, "epoch": 0.29585320507638835, "step": 1830}, {"loss": 0.8507, "grad_norm": 0.48555272817611694, "learning_rate": 0.0002, "epoch": 0.29746988925713364, "step": 1840}, {"loss": 0.7657, "grad_norm": 0.3752668499946594, "learning_rate": 0.0002, "epoch": 0.2990865734378789, "step": 1850}, {"loss": 0.7915, "grad_norm": 0.5328747034072876, "learning_rate": 0.0002, "epoch": 0.3007032576186242, "step": 1860}, {"loss": 0.8426, "grad_norm": 0.48716455698013306, "learning_rate": 0.0002, "epoch": 0.3023199417993695, "step": 1870}, {"loss": 0.8335, "grad_norm": 0.5011493563652039, "learning_rate": 0.0002, "epoch": 0.3039366259801148, "step": 1880}, {"loss": 0.852, "grad_norm": 0.46461427211761475, "learning_rate": 0.0002, "epoch": 0.30555331016086007, "step": 1890}, {"loss": 0.8478, "grad_norm": 0.36630210280418396, "learning_rate": 0.0002, "epoch": 0.30716999434160536, "step": 1900}, {"loss": 0.8162, "grad_norm": 0.4217296242713928, "learning_rate": 0.0002, "epoch": 0.30878667852235064, "step": 1910}, {"loss": 0.8128, "grad_norm": 0.4394875466823578, "learning_rate": 0.0002, "epoch": 0.31040336270309593, "step": 1920}, {"loss": 0.8471, "grad_norm": 0.6587965488433838, "learning_rate": 0.0002, "epoch": 0.3120200468838412, "step": 1930}, {"loss": 0.8565, "grad_norm": 0.5469298958778381, "learning_rate": 0.0002, "epoch": 0.31363673106458656, "step": 1940}, {"loss": 0.8236, "grad_norm": 0.4371595084667206, "learning_rate": 0.0002, "epoch": 0.31525341524533185, "step": 1950}, {"loss": 0.887, "grad_norm": 0.4809541404247284, "learning_rate": 0.0002, "epoch": 0.31687009942607713, "step": 1960}, {"loss": 0.7855, "grad_norm": 0.6061086654663086, "learning_rate": 0.0002, "epoch": 0.3184867836068224, "step": 1970}, {"loss": 0.7679, "grad_norm": 0.5342657566070557, "learning_rate": 0.0002, "epoch": 0.3201034677875677, "step": 1980}, {"loss": 0.7955, "grad_norm": 0.5057743787765503, "learning_rate": 0.0002, "epoch": 0.321720151968313, "step": 1990}, {"loss": 0.7774, "grad_norm": 0.528626024723053, "learning_rate": 0.0002, "epoch": 0.3233368361490583, "step": 2000}, {"loss": 0.8845, "grad_norm": 0.46742770075798035, "learning_rate": 0.0002, "epoch": 0.32495352032980357, "step": 2010}, {"loss": 0.8484, "grad_norm": 0.515101432800293, "learning_rate": 0.0002, "epoch": 0.32657020451054886, "step": 2020}, {"loss": 0.8139, "grad_norm": 0.41941216588020325, "learning_rate": 0.0002, "epoch": 0.32818688869129414, "step": 2030}, {"loss": 0.7637, "grad_norm": 0.49902522563934326, "learning_rate": 0.0002, "epoch": 0.32980357287203943, "step": 2040}, {"loss": 0.7822, "grad_norm": 0.4120897650718689, "learning_rate": 0.0002, "epoch": 0.3314202570527847, "step": 2050}, {"loss": 0.8057, "grad_norm": 0.45352041721343994, "learning_rate": 0.0002, "epoch": 0.33303694123353, "step": 2060}, {"loss": 0.7913, "grad_norm": 0.523199737071991, "learning_rate": 0.0002, "epoch": 0.33465362541427535, "step": 2070}, {"loss": 0.8036, "grad_norm": 0.4390358626842499, "learning_rate": 0.0002, "epoch": 0.33627030959502063, "step": 2080}, {"loss": 0.8145, "grad_norm": 0.6752901077270508, "learning_rate": 0.0002, "epoch": 0.3378869937757659, "step": 2090}, {"loss": 0.7807, "grad_norm": 0.547821044921875, "learning_rate": 0.0002, "epoch": 0.3395036779565112, "step": 2100}, {"loss": 0.8561, "grad_norm": 0.5161308646202087, "learning_rate": 0.0002, "epoch": 0.3411203621372565, "step": 2110}, {"loss": 0.7697, "grad_norm": 0.4565401077270508, "learning_rate": 0.0002, "epoch": 0.3427370463180018, "step": 2120}, {"loss": 0.7964, "grad_norm": 0.4666115939617157, "learning_rate": 0.0002, "epoch": 0.34435373049874707, "step": 2130}, {"loss": 0.8189, "grad_norm": 0.4090428352355957, "learning_rate": 0.0002, "epoch": 0.34597041467949236, "step": 2140}, {"loss": 0.8817, "grad_norm": 0.510845422744751, "learning_rate": 0.0002, "epoch": 0.34758709886023764, "step": 2150}, {"loss": 0.8398, "grad_norm": 0.42861923575401306, "learning_rate": 0.0002, "epoch": 0.34920378304098293, "step": 2160}, {"loss": 0.7716, "grad_norm": 0.4476332664489746, "learning_rate": 0.0002, "epoch": 0.3508204672217282, "step": 2170}, {"loss": 0.7845, "grad_norm": 0.6065791249275208, "learning_rate": 0.0002, "epoch": 0.3524371514024735, "step": 2180}, {"loss": 0.8187, "grad_norm": 0.42335066199302673, "learning_rate": 0.0002, "epoch": 0.35405383558321885, "step": 2190}, {"loss": 0.8239, "grad_norm": 0.5094629526138306, "learning_rate": 0.0002, "epoch": 0.35567051976396413, "step": 2200}, {"loss": 0.7807, "grad_norm": 0.5476373434066772, "learning_rate": 0.0002, "epoch": 0.3572872039447094, "step": 2210}, {"loss": 0.814, "grad_norm": 0.3911719024181366, "learning_rate": 0.0002, "epoch": 0.3589038881254547, "step": 2220}, {"loss": 0.8599, "grad_norm": 0.6599636077880859, "learning_rate": 0.0002, "epoch": 0.3605205723062, "step": 2230}, {"loss": 0.7482, "grad_norm": 0.40381914377212524, "learning_rate": 0.0002, "epoch": 0.3621372564869453, "step": 2240}, {"loss": 0.7772, "grad_norm": 0.4433908462524414, "learning_rate": 0.0002, "epoch": 0.36375394066769057, "step": 2250}, {"loss": 0.8503, "grad_norm": 0.578326940536499, "learning_rate": 0.0002, "epoch": 0.36537062484843585, "step": 2260}, {"loss": 0.8178, "grad_norm": 0.5734784007072449, "learning_rate": 0.0002, "epoch": 0.36698730902918114, "step": 2270}, {"loss": 0.8193, "grad_norm": 0.45555487275123596, "learning_rate": 0.0002, "epoch": 0.36860399320992643, "step": 2280}, {"loss": 0.7929, "grad_norm": 0.5666276216506958, "learning_rate": 0.0002, "epoch": 0.3702206773906717, "step": 2290}, {"loss": 0.8292, "grad_norm": 0.5461117625236511, "learning_rate": 0.0002, "epoch": 0.371837361571417, "step": 2300}, {"loss": 0.8204, "grad_norm": 0.6318911910057068, "learning_rate": 0.0002, "epoch": 0.3734540457521623, "step": 2310}, {"loss": 0.7964, "grad_norm": 0.493263304233551, "learning_rate": 0.0002, "epoch": 0.37507072993290763, "step": 2320}, {"loss": 0.8339, "grad_norm": 0.5888760089874268, "learning_rate": 0.0002, "epoch": 0.3766874141136529, "step": 2330}, {"loss": 0.7737, "grad_norm": 0.48671841621398926, "learning_rate": 0.0002, "epoch": 0.3783040982943982, "step": 2340}, {"loss": 0.8367, "grad_norm": 0.4385145306587219, "learning_rate": 0.0002, "epoch": 0.3799207824751435, "step": 2350}, {"loss": 0.812, "grad_norm": 0.5523318648338318, "learning_rate": 0.0002, "epoch": 0.3815374666558888, "step": 2360}, {"loss": 0.8351, "grad_norm": 0.7308220267295837, "learning_rate": 0.0002, "epoch": 0.38315415083663407, "step": 2370}, {"loss": 0.859, "grad_norm": 0.554214358329773, "learning_rate": 0.0002, "epoch": 0.38477083501737935, "step": 2380}, {"loss": 0.8146, "grad_norm": 0.5425800085067749, "learning_rate": 0.0002, "epoch": 0.38638751919812464, "step": 2390}, {"loss": 0.8282, "grad_norm": 0.48811158537864685, "learning_rate": 0.0002, "epoch": 0.3880042033788699, "step": 2400}, {"loss": 0.8074, "grad_norm": 0.49212366342544556, "learning_rate": 0.0002, "epoch": 0.3896208875596152, "step": 2410}, {"loss": 0.7991, "grad_norm": 0.5222218632698059, "learning_rate": 0.0002, "epoch": 0.3912375717403605, "step": 2420}, {"loss": 0.8182, "grad_norm": 0.4699819087982178, "learning_rate": 0.0002, "epoch": 0.3928542559211058, "step": 2430}, {"loss": 0.7919, "grad_norm": 0.46153587102890015, "learning_rate": 0.0002, "epoch": 0.39447094010185113, "step": 2440}, {"loss": 0.8111, "grad_norm": 0.4150611162185669, "learning_rate": 0.0002, "epoch": 0.3960876242825964, "step": 2450}, {"loss": 0.8589, "grad_norm": 0.5799614787101746, "learning_rate": 0.0002, "epoch": 0.3977043084633417, "step": 2460}, {"loss": 0.8085, "grad_norm": 0.56536865234375, "learning_rate": 0.0002, "epoch": 0.399320992644087, "step": 2470}, {"loss": 0.8022, "grad_norm": 0.5451247096061707, "learning_rate": 0.0002, "epoch": 0.4009376768248323, "step": 2480}, {"loss": 0.8217, "grad_norm": 0.5914521217346191, "learning_rate": 0.0002, "epoch": 0.40255436100557757, "step": 2490}, {"loss": 0.7859, "grad_norm": 0.4428117275238037, "learning_rate": 0.0002, "epoch": 0.40417104518632285, "step": 2500}, {"loss": 0.8054, "grad_norm": 0.48580947518348694, "learning_rate": 0.0002, "epoch": 0.40578772936706814, "step": 2510}, {"loss": 0.8405, "grad_norm": 0.436734676361084, "learning_rate": 0.0002, "epoch": 0.4074044135478134, "step": 2520}, {"loss": 0.8209, "grad_norm": 0.5752223134040833, "learning_rate": 0.0002, "epoch": 0.4090210977285587, "step": 2530}, {"loss": 0.8181, "grad_norm": 0.4271308183670044, "learning_rate": 0.0002, "epoch": 0.410637781909304, "step": 2540}, {"loss": 0.8058, "grad_norm": 0.46294718980789185, "learning_rate": 0.0002, "epoch": 0.4122544660900493, "step": 2550}, {"loss": 0.8473, "grad_norm": 0.49407583475112915, "learning_rate": 0.0002, "epoch": 0.4138711502707946, "step": 2560}, {"loss": 0.7881, "grad_norm": 0.4729035496711731, "learning_rate": 0.0002, "epoch": 0.4154878344515399, "step": 2570}, {"loss": 0.7834, "grad_norm": 0.4129747152328491, "learning_rate": 0.0002, "epoch": 0.4171045186322852, "step": 2580}, {"loss": 0.7859, "grad_norm": 0.5684236288070679, "learning_rate": 0.0002, "epoch": 0.4187212028130305, "step": 2590}, {"loss": 0.811, "grad_norm": 0.4862157106399536, "learning_rate": 0.0002, "epoch": 0.4203378869937758, "step": 2600}, {"loss": 0.7582, "grad_norm": 0.46567976474761963, "learning_rate": 0.0002, "epoch": 0.42195457117452106, "step": 2610}, {"loss": 0.7755, "grad_norm": 0.5710650682449341, "learning_rate": 0.0002, "epoch": 0.42357125535526635, "step": 2620}, {"loss": 0.8573, "grad_norm": 0.5660041570663452, "learning_rate": 0.0002, "epoch": 0.42518793953601164, "step": 2630}, {"loss": 0.7812, "grad_norm": 0.47944375872612, "learning_rate": 0.0002, "epoch": 0.4268046237167569, "step": 2640}, {"loss": 0.7459, "grad_norm": 0.537223756313324, "learning_rate": 0.0002, "epoch": 0.4284213078975022, "step": 2650}, {"loss": 0.8246, "grad_norm": 0.41669997572898865, "learning_rate": 0.0002, "epoch": 0.4300379920782475, "step": 2660}, {"loss": 0.7785, "grad_norm": 0.44727686047554016, "learning_rate": 0.0002, "epoch": 0.4316546762589928, "step": 2670}, {"loss": 0.8241, "grad_norm": 0.5600888729095459, "learning_rate": 0.0002, "epoch": 0.4332713604397381, "step": 2680}, {"loss": 0.7708, "grad_norm": 0.39820605516433716, "learning_rate": 0.0002, "epoch": 0.4348880446204834, "step": 2690}, {"loss": 0.8202, "grad_norm": 0.5637655854225159, "learning_rate": 0.0002, "epoch": 0.4365047288012287, "step": 2700}, {"loss": 0.855, "grad_norm": 0.6363666653633118, "learning_rate": 0.0002, "epoch": 0.438121412981974, "step": 2710}, {"loss": 0.8468, "grad_norm": 0.5656129121780396, "learning_rate": 0.0002, "epoch": 0.4397380971627193, "step": 2720}, {"loss": 0.7845, "grad_norm": 0.5600156188011169, "learning_rate": 0.0002, "epoch": 0.44135478134346456, "step": 2730}, {"loss": 0.8405, "grad_norm": 0.5506579875946045, "learning_rate": 0.0002, "epoch": 0.44297146552420985, "step": 2740}, {"loss": 0.7725, "grad_norm": 0.49878305196762085, "learning_rate": 0.0002, "epoch": 0.44458814970495514, "step": 2750}, {"loss": 0.8292, "grad_norm": 0.4569213092327118, "learning_rate": 0.0002, "epoch": 0.4462048338857004, "step": 2760}, {"loss": 0.8028, "grad_norm": 0.6056680083274841, "learning_rate": 0.0002, "epoch": 0.4478215180664457, "step": 2770}, {"loss": 0.8242, "grad_norm": 0.44474557042121887, "learning_rate": 0.0002, "epoch": 0.449438202247191, "step": 2780}, {"loss": 0.801, "grad_norm": 0.46055394411087036, "learning_rate": 0.0002, "epoch": 0.4510548864279363, "step": 2790}, {"loss": 0.7521, "grad_norm": 0.4904133379459381, "learning_rate": 0.0002, "epoch": 0.4526715706086816, "step": 2800}, {"loss": 0.8829, "grad_norm": 0.5647031664848328, "learning_rate": 0.0002, "epoch": 0.45428825478942686, "step": 2810}, {"loss": 0.8622, "grad_norm": 0.5759473443031311, "learning_rate": 0.0002, "epoch": 0.4559049389701722, "step": 2820}, {"loss": 0.7812, "grad_norm": 0.5161895751953125, "learning_rate": 0.0002, "epoch": 0.4575216231509175, "step": 2830}, {"loss": 0.8045, "grad_norm": 0.4248254597187042, "learning_rate": 0.0002, "epoch": 0.4591383073316628, "step": 2840}, {"loss": 0.7838, "grad_norm": 0.45395001769065857, "learning_rate": 0.0002, "epoch": 0.46075499151240806, "step": 2850}, {"loss": 0.8208, "grad_norm": 0.5358697772026062, "learning_rate": 0.0002, "epoch": 0.46237167569315335, "step": 2860}, {"loss": 0.8147, "grad_norm": 0.5379165410995483, "learning_rate": 0.0002, "epoch": 0.46398835987389864, "step": 2870}, {"loss": 0.7403, "grad_norm": 0.4601989686489105, "learning_rate": 0.0002, "epoch": 0.4656050440546439, "step": 2880}, {"loss": 0.8523, "grad_norm": 0.671115517616272, "learning_rate": 0.0002, "epoch": 0.4672217282353892, "step": 2890}, {"loss": 0.8262, "grad_norm": 0.4425133168697357, "learning_rate": 0.0002, "epoch": 0.4688384124161345, "step": 2900}, {"loss": 0.8178, "grad_norm": 0.5446155071258545, "learning_rate": 0.0002, "epoch": 0.4704550965968798, "step": 2910}, {"loss": 0.8106, "grad_norm": 0.603306233882904, "learning_rate": 0.0002, "epoch": 0.47207178077762507, "step": 2920}, {"loss": 0.8044, "grad_norm": 0.5377997159957886, "learning_rate": 0.0002, "epoch": 0.47368846495837036, "step": 2930}, {"loss": 0.8075, "grad_norm": 0.4931027591228485, "learning_rate": 0.0002, "epoch": 0.4753051491391157, "step": 2940}, {"loss": 0.8004, "grad_norm": 0.4711960256099701, "learning_rate": 0.0002, "epoch": 0.476921833319861, "step": 2950}, {"loss": 0.8121, "grad_norm": 0.5020492672920227, "learning_rate": 0.0002, "epoch": 0.4785385175006063, "step": 2960}, {"loss": 0.8221, "grad_norm": 0.5428946614265442, "learning_rate": 0.0002, "epoch": 0.48015520168135156, "step": 2970}, {"loss": 0.7849, "grad_norm": 0.5294089317321777, "learning_rate": 0.0002, "epoch": 0.48177188586209685, "step": 2980}, {"loss": 0.8553, "grad_norm": 0.648289144039154, "learning_rate": 0.0002, "epoch": 0.48338857004284214, "step": 2990}, {"loss": 0.7874, "grad_norm": 0.47916680574417114, "learning_rate": 0.0002, "epoch": 0.4850052542235874, "step": 3000}, {"loss": 0.8087, "grad_norm": 0.43849772214889526, "learning_rate": 0.0002, "epoch": 0.4866219384043327, "step": 3010}, {"loss": 0.7662, "grad_norm": 0.47007861733436584, "learning_rate": 0.0002, "epoch": 0.488238622585078, "step": 3020}, {"loss": 0.757, "grad_norm": 0.6314331293106079, "learning_rate": 0.0002, "epoch": 0.4898553067658233, "step": 3030}, {"loss": 0.7863, "grad_norm": 0.49211493134498596, "learning_rate": 0.0002, "epoch": 0.49147199094656857, "step": 3040}, {"loss": 0.8335, "grad_norm": 0.4537973403930664, "learning_rate": 0.0002, "epoch": 0.49308867512731386, "step": 3050}, {"loss": 0.8095, "grad_norm": 0.47326919436454773, "learning_rate": 0.0002, "epoch": 0.49470535930805914, "step": 3060}, {"loss": 0.8447, "grad_norm": 0.525874137878418, "learning_rate": 0.0002, "epoch": 0.4963220434888045, "step": 3070}, {"loss": 0.8339, "grad_norm": 0.6361091732978821, "learning_rate": 0.0002, "epoch": 0.4979387276695498, "step": 3080}, {"loss": 0.821, "grad_norm": 0.5850642919540405, "learning_rate": 0.0002, "epoch": 0.49955541185029506, "step": 3090}, {"loss": 0.8279, "grad_norm": 0.47299543023109436, "learning_rate": 0.0002, "epoch": 0.5011720960310403, "step": 3100}, {"loss": 0.8681, "grad_norm": 0.473099946975708, "learning_rate": 0.0002, "epoch": 0.5027887802117856, "step": 3110}, {"loss": 0.8223, "grad_norm": 0.48186397552490234, "learning_rate": 0.0002, "epoch": 0.5044054643925309, "step": 3120}, {"loss": 0.8292, "grad_norm": 0.5015401840209961, "learning_rate": 0.0002, "epoch": 0.5060221485732762, "step": 3130}, {"loss": 0.7692, "grad_norm": 0.5617750287055969, "learning_rate": 0.0002, "epoch": 0.5076388327540216, "step": 3140}, {"loss": 0.8708, "grad_norm": 0.5169327259063721, "learning_rate": 0.0002, "epoch": 0.5092555169347668, "step": 3150}, {"loss": 0.7845, "grad_norm": 0.545657753944397, "learning_rate": 0.0002, "epoch": 0.5108722011155121, "step": 3160}, {"loss": 0.799, "grad_norm": 0.512864351272583, "learning_rate": 0.0002, "epoch": 0.5124888852962574, "step": 3170}, {"loss": 0.7794, "grad_norm": 0.4113546311855316, "learning_rate": 0.0002, "epoch": 0.5141055694770027, "step": 3180}, {"loss": 0.8206, "grad_norm": 0.44532445073127747, "learning_rate": 0.0002, "epoch": 0.5157222536577479, "step": 3190}, {"loss": 0.8213, "grad_norm": 0.5623497366905212, "learning_rate": 0.0002, "epoch": 0.5173389378384933, "step": 3200}, {"loss": 0.7928, "grad_norm": 0.5084741115570068, "learning_rate": 0.0002, "epoch": 0.5189556220192385, "step": 3210}, {"loss": 0.8174, "grad_norm": 0.5305403470993042, "learning_rate": 0.0002, "epoch": 0.5205723061999838, "step": 3220}, {"loss": 0.8139, "grad_norm": 0.4708254337310791, "learning_rate": 0.0002, "epoch": 0.5221889903807291, "step": 3230}, {"loss": 0.7639, "grad_norm": 0.43827131390571594, "learning_rate": 0.0002, "epoch": 0.5238056745614744, "step": 3240}, {"loss": 0.7993, "grad_norm": 0.5630002617835999, "learning_rate": 0.0002, "epoch": 0.5254223587422197, "step": 3250}, {"loss": 0.7522, "grad_norm": 0.5010961890220642, "learning_rate": 0.0002, "epoch": 0.527039042922965, "step": 3260}, {"loss": 0.8374, "grad_norm": 0.6303122043609619, "learning_rate": 0.0002, "epoch": 0.5286557271037103, "step": 3270}, {"loss": 0.7727, "grad_norm": 0.5107331275939941, "learning_rate": 0.0002, "epoch": 0.5302724112844556, "step": 3280}, {"loss": 0.8495, "grad_norm": 0.5700443387031555, "learning_rate": 0.0002, "epoch": 0.5318890954652009, "step": 3290}, {"loss": 0.7776, "grad_norm": 0.46296367049217224, "learning_rate": 0.0002, "epoch": 0.5335057796459461, "step": 3300}, {"loss": 0.7931, "grad_norm": 0.531568706035614, "learning_rate": 0.0002, "epoch": 0.5351224638266915, "step": 3310}, {"loss": 0.843, "grad_norm": 0.4686741530895233, "learning_rate": 0.0002, "epoch": 0.5367391480074367, "step": 3320}, {"loss": 0.8104, "grad_norm": 0.5404331088066101, "learning_rate": 0.0002, "epoch": 0.5383558321881821, "step": 3330}, {"loss": 0.7686, "grad_norm": 0.6368790864944458, "learning_rate": 0.0002, "epoch": 0.5399725163689273, "step": 3340}, {"loss": 0.8514, "grad_norm": 0.42300888895988464, "learning_rate": 0.0002, "epoch": 0.5415892005496726, "step": 3350}, {"loss": 0.8236, "grad_norm": 0.5362542867660522, "learning_rate": 0.0002, "epoch": 0.5432058847304179, "step": 3360}, {"loss": 0.858, "grad_norm": 0.497128963470459, "learning_rate": 0.0002, "epoch": 0.5448225689111632, "step": 3370}, {"loss": 0.8519, "grad_norm": 0.5006386041641235, "learning_rate": 0.0002, "epoch": 0.5464392530919085, "step": 3380}, {"loss": 0.7867, "grad_norm": 0.44136837124824524, "learning_rate": 0.0002, "epoch": 0.5480559372726538, "step": 3390}, {"loss": 0.773, "grad_norm": 0.5897833108901978, "learning_rate": 0.0002, "epoch": 0.5496726214533991, "step": 3400}, {"loss": 0.8895, "grad_norm": 0.641075611114502, "learning_rate": 0.0002, "epoch": 0.5512893056341444, "step": 3410}, {"loss": 0.7827, "grad_norm": 0.7251322269439697, "learning_rate": 0.0002, "epoch": 0.5529059898148897, "step": 3420}, {"loss": 0.7626, "grad_norm": 0.47411349415779114, "learning_rate": 0.0002, "epoch": 0.5545226739956349, "step": 3430}, {"loss": 0.8196, "grad_norm": 0.4994310438632965, "learning_rate": 0.0002, "epoch": 0.5561393581763803, "step": 3440}, {"loss": 0.7812, "grad_norm": 0.5814438462257385, "learning_rate": 0.0002, "epoch": 0.5577560423571255, "step": 3450}, {"loss": 0.8805, "grad_norm": 0.6278898119926453, "learning_rate": 0.0002, "epoch": 0.5593727265378708, "step": 3460}, {"loss": 0.813, "grad_norm": 0.46208274364471436, "learning_rate": 0.0002, "epoch": 0.5609894107186161, "step": 3470}, {"loss": 0.8295, "grad_norm": 0.5718930959701538, "learning_rate": 0.0002, "epoch": 0.5626060948993614, "step": 3480}, {"loss": 0.8152, "grad_norm": 0.48178744316101074, "learning_rate": 0.0002, "epoch": 0.5642227790801067, "step": 3490}, {"loss": 0.8244, "grad_norm": 0.47336965799331665, "learning_rate": 0.0002, "epoch": 0.565839463260852, "step": 3500}, {"loss": 0.8099, "grad_norm": 0.43442684412002563, "learning_rate": 0.0002, "epoch": 0.5674561474415973, "step": 3510}, {"loss": 0.7564, "grad_norm": 0.6463358998298645, "learning_rate": 0.0002, "epoch": 0.5690728316223426, "step": 3520}, {"loss": 0.836, "grad_norm": 0.5286486744880676, "learning_rate": 0.0002, "epoch": 0.5706895158030879, "step": 3530}, {"loss": 0.8421, "grad_norm": 0.5405499935150146, "learning_rate": 0.0002, "epoch": 0.5723061999838331, "step": 3540}, {"loss": 0.7614, "grad_norm": 0.6654391884803772, "learning_rate": 0.0002, "epoch": 0.5739228841645785, "step": 3550}, {"loss": 0.7803, "grad_norm": 0.5081980228424072, "learning_rate": 0.0002, "epoch": 0.5755395683453237, "step": 3560}, {"loss": 0.7753, "grad_norm": 0.48978179693222046, "learning_rate": 0.0002, "epoch": 0.5771562525260691, "step": 3570}, {"loss": 0.8151, "grad_norm": 0.5840612053871155, "learning_rate": 0.0002, "epoch": 0.5787729367068143, "step": 3580}, {"loss": 0.8937, "grad_norm": 0.5235261917114258, "learning_rate": 0.0002, "epoch": 0.5803896208875596, "step": 3590}, {"loss": 0.7894, "grad_norm": 0.5672075748443604, "learning_rate": 0.0002, "epoch": 0.5820063050683049, "step": 3600}, {"loss": 0.8347, "grad_norm": 0.5613429546356201, "learning_rate": 0.0002, "epoch": 0.5836229892490502, "step": 3610}, {"loss": 0.8274, "grad_norm": 0.4032273590564728, "learning_rate": 0.0002, "epoch": 0.5852396734297954, "step": 3620}, {"loss": 0.8421, "grad_norm": 0.49559324979782104, "learning_rate": 0.0002, "epoch": 0.5868563576105408, "step": 3630}, {"loss": 0.8332, "grad_norm": 0.6895697712898254, "learning_rate": 0.0002, "epoch": 0.5884730417912861, "step": 3640}, {"loss": 0.7877, "grad_norm": 0.4750136435031891, "learning_rate": 0.0002, "epoch": 0.5900897259720314, "step": 3650}, {"loss": 0.8219, "grad_norm": 0.5176819562911987, "learning_rate": 0.0002, "epoch": 0.5917064101527767, "step": 3660}, {"loss": 0.8151, "grad_norm": 0.5817760229110718, "learning_rate": 0.0002, "epoch": 0.5933230943335219, "step": 3670}, {"loss": 0.7823, "grad_norm": 0.6064626574516296, "learning_rate": 0.0002, "epoch": 0.5949397785142673, "step": 3680}, {"loss": 0.8422, "grad_norm": 0.6728700995445251, "learning_rate": 0.0002, "epoch": 0.5965564626950125, "step": 3690}, {"loss": 0.7679, "grad_norm": 0.609305202960968, "learning_rate": 0.0002, "epoch": 0.5981731468757578, "step": 3700}, {"loss": 0.8048, "grad_norm": 0.4615488350391388, "learning_rate": 0.0002, "epoch": 0.5997898310565031, "step": 3710}, {"loss": 0.8214, "grad_norm": 2.0531179904937744, "learning_rate": 0.0002, "epoch": 0.6014065152372484, "step": 3720}, {"loss": 0.8158, "grad_norm": 0.5091132521629333, "learning_rate": 0.0002, "epoch": 0.6030231994179936, "step": 3730}, {"loss": 0.7833, "grad_norm": 0.5951124429702759, "learning_rate": 0.0002, "epoch": 0.604639883598739, "step": 3740}, {"loss": 0.7784, "grad_norm": 0.5870208144187927, "learning_rate": 0.0002, "epoch": 0.6062565677794842, "step": 3750}, {"loss": 0.8044, "grad_norm": 0.6254619359970093, "learning_rate": 0.0002, "epoch": 0.6078732519602296, "step": 3760}, {"loss": 0.7868, "grad_norm": 0.5577626824378967, "learning_rate": 0.0002, "epoch": 0.6094899361409749, "step": 3770}, {"loss": 0.8108, "grad_norm": 0.5004405379295349, "learning_rate": 0.0002, "epoch": 0.6111066203217201, "step": 3780}, {"loss": 0.8092, "grad_norm": 0.5527383685112, "learning_rate": 0.0002, "epoch": 0.6127233045024655, "step": 3790}, {"loss": 0.8036, "grad_norm": 0.49116113781929016, "learning_rate": 0.0002, "epoch": 0.6143399886832107, "step": 3800}, {"loss": 0.8352, "grad_norm": 0.5299299359321594, "learning_rate": 0.0002, "epoch": 0.6159566728639561, "step": 3810}, {"loss": 0.7737, "grad_norm": 0.464897483587265, "learning_rate": 0.0002, "epoch": 0.6175733570447013, "step": 3820}, {"loss": 0.7923, "grad_norm": 0.6505740880966187, "learning_rate": 0.0002, "epoch": 0.6191900412254466, "step": 3830}, {"loss": 0.8123, "grad_norm": 0.5512559413909912, "learning_rate": 0.0002, "epoch": 0.6208067254061919, "step": 3840}, {"loss": 0.8856, "grad_norm": 0.49427518248558044, "learning_rate": 0.0002, "epoch": 0.6224234095869372, "step": 3850}, {"loss": 0.7751, "grad_norm": 0.3839147090911865, "learning_rate": 0.0002, "epoch": 0.6240400937676824, "step": 3860}, {"loss": 0.8006, "grad_norm": 0.5760218501091003, "learning_rate": 0.0002, "epoch": 0.6256567779484278, "step": 3870}, {"loss": 0.7836, "grad_norm": 0.7226507067680359, "learning_rate": 0.0002, "epoch": 0.6272734621291731, "step": 3880}, {"loss": 0.8244, "grad_norm": 0.676781415939331, "learning_rate": 0.0002, "epoch": 0.6288901463099184, "step": 3890}, {"loss": 0.8239, "grad_norm": 0.4284018278121948, "learning_rate": 0.0002, "epoch": 0.6305068304906637, "step": 3900}, {"loss": 0.7996, "grad_norm": 0.5060628056526184, "learning_rate": 0.0002, "epoch": 0.6321235146714089, "step": 3910}, {"loss": 0.8089, "grad_norm": 0.5524522066116333, "learning_rate": 0.0002, "epoch": 0.6337401988521543, "step": 3920}, {"loss": 0.8276, "grad_norm": 0.6099881529808044, "learning_rate": 0.0002, "epoch": 0.6353568830328995, "step": 3930}, {"loss": 0.809, "grad_norm": 0.43155938386917114, "learning_rate": 0.0002, "epoch": 0.6369735672136448, "step": 3940}, {"loss": 0.8404, "grad_norm": 0.6427084803581238, "learning_rate": 0.0002, "epoch": 0.6385902513943901, "step": 3950}, {"loss": 0.8368, "grad_norm": 0.541220486164093, "learning_rate": 0.0002, "epoch": 0.6402069355751354, "step": 3960}, {"loss": 0.8539, "grad_norm": 0.5414294600486755, "learning_rate": 0.0002, "epoch": 0.6418236197558806, "step": 3970}, {"loss": 0.7996, "grad_norm": 0.46344003081321716, "learning_rate": 0.0002, "epoch": 0.643440303936626, "step": 3980}, {"loss": 0.7474, "grad_norm": 0.45209285616874695, "learning_rate": 0.0002, "epoch": 0.6450569881173712, "step": 3990}, {"loss": 0.8202, "grad_norm": 0.5417284369468689, "learning_rate": 0.0002, "epoch": 0.6466736722981166, "step": 4000}, {"loss": 0.7563, "grad_norm": 0.7995685935020447, "learning_rate": 0.0002, "epoch": 0.6482903564788619, "step": 4010}, {"loss": 0.7812, "grad_norm": 0.6384002566337585, "learning_rate": 0.0002, "epoch": 0.6499070406596071, "step": 4020}, {"loss": 0.732, "grad_norm": 0.4472815692424774, "learning_rate": 0.0002, "epoch": 0.6515237248403525, "step": 4030}, {"loss": 0.8071, "grad_norm": 0.6834294199943542, "learning_rate": 0.0002, "epoch": 0.6531404090210977, "step": 4040}, {"loss": 0.7812, "grad_norm": 0.4612339735031128, "learning_rate": 0.0002, "epoch": 0.654757093201843, "step": 4050}, {"loss": 0.8141, "grad_norm": 0.9266576170921326, "learning_rate": 0.0002, "epoch": 0.6563737773825883, "step": 4060}, {"loss": 0.7991, "grad_norm": 0.4470861852169037, "learning_rate": 0.0002, "epoch": 0.6579904615633336, "step": 4070}, {"loss": 0.8293, "grad_norm": 0.45544925332069397, "learning_rate": 0.0002, "epoch": 0.6596071457440789, "step": 4080}, {"loss": 0.8455, "grad_norm": 0.6144481301307678, "learning_rate": 0.0002, "epoch": 0.6612238299248242, "step": 4090}, {"loss": 0.7877, "grad_norm": 0.5936288237571716, "learning_rate": 0.0002, "epoch": 0.6628405141055694, "step": 4100}, {"loss": 0.7617, "grad_norm": 0.4822963774204254, "learning_rate": 0.0002, "epoch": 0.6644571982863148, "step": 4110}, {"loss": 0.7997, "grad_norm": 0.48432496190071106, "learning_rate": 0.0002, "epoch": 0.66607388246706, "step": 4120}, {"loss": 0.8404, "grad_norm": 0.4901607930660248, "learning_rate": 0.0002, "epoch": 0.6676905666478054, "step": 4130}, {"loss": 0.8085, "grad_norm": 0.5018393397331238, "learning_rate": 0.0002, "epoch": 0.6693072508285507, "step": 4140}, {"loss": 0.8065, "grad_norm": 0.6946378946304321, "learning_rate": 0.0002, "epoch": 0.6709239350092959, "step": 4150}, {"loss": 0.8147, "grad_norm": 0.5997390747070312, "learning_rate": 0.0002, "epoch": 0.6725406191900413, "step": 4160}, {"loss": 0.8268, "grad_norm": 0.6738849878311157, "learning_rate": 0.0002, "epoch": 0.6741573033707865, "step": 4170}, {"loss": 0.7704, "grad_norm": 0.6110581159591675, "learning_rate": 0.0002, "epoch": 0.6757739875515318, "step": 4180}, {"loss": 0.8043, "grad_norm": 0.5703322291374207, "learning_rate": 0.0002, "epoch": 0.6773906717322771, "step": 4190}, {"loss": 0.8099, "grad_norm": 0.4686066210269928, "learning_rate": 0.0002, "epoch": 0.6790073559130224, "step": 4200}, {"loss": 0.8441, "grad_norm": 0.6394643783569336, "learning_rate": 0.0002, "epoch": 0.6806240400937676, "step": 4210}, {"loss": 0.8011, "grad_norm": 0.5454841256141663, "learning_rate": 0.0002, "epoch": 0.682240724274513, "step": 4220}, {"loss": 0.8307, "grad_norm": 0.4859732985496521, "learning_rate": 0.0002, "epoch": 0.6838574084552582, "step": 4230}, {"loss": 0.8161, "grad_norm": 0.5544065833091736, "learning_rate": 0.0002, "epoch": 0.6854740926360036, "step": 4240}, {"loss": 0.7839, "grad_norm": 0.4902505576610565, "learning_rate": 0.0002, "epoch": 0.6870907768167488, "step": 4250}, {"loss": 0.7977, "grad_norm": 0.4768051505088806, "learning_rate": 0.0002, "epoch": 0.6887074609974941, "step": 4260}, {"loss": 0.7539, "grad_norm": 0.49982190132141113, "learning_rate": 0.0002, "epoch": 0.6903241451782395, "step": 4270}, {"loss": 0.7353, "grad_norm": 0.6351838111877441, "learning_rate": 0.0002, "epoch": 0.6919408293589847, "step": 4280}, {"loss": 0.7664, "grad_norm": 0.5647561550140381, "learning_rate": 0.0002, "epoch": 0.69355751353973, "step": 4290}, {"loss": 0.7618, "grad_norm": 0.5340486764907837, "learning_rate": 0.0002, "epoch": 0.6951741977204753, "step": 4300}, {"loss": 0.8526, "grad_norm": 0.5649092793464661, "learning_rate": 0.0002, "epoch": 0.6967908819012206, "step": 4310}, {"loss": 0.8246, "grad_norm": 0.6183916926383972, "learning_rate": 0.0002, "epoch": 0.6984075660819659, "step": 4320}, {"loss": 0.792, "grad_norm": 0.6154509782791138, "learning_rate": 0.0002, "epoch": 0.7000242502627112, "step": 4330}, {"loss": 0.8397, "grad_norm": 0.5156264305114746, "learning_rate": 0.0002, "epoch": 0.7016409344434564, "step": 4340}, {"loss": 0.8512, "grad_norm": 0.562171459197998, "learning_rate": 0.0002, "epoch": 0.7032576186242018, "step": 4350}, {"loss": 0.7882, "grad_norm": 0.4949502646923065, "learning_rate": 0.0002, "epoch": 0.704874302804947, "step": 4360}, {"loss": 0.738, "grad_norm": 0.5171684622764587, "learning_rate": 0.0002, "epoch": 0.7064909869856923, "step": 4370}, {"loss": 0.8001, "grad_norm": 0.6198443174362183, "learning_rate": 0.0002, "epoch": 0.7081076711664377, "step": 4380}, {"loss": 0.7606, "grad_norm": 0.5802276134490967, "learning_rate": 0.0002, "epoch": 0.7097243553471829, "step": 4390}, {"loss": 0.8797, "grad_norm": 0.41096967458724976, "learning_rate": 0.0002, "epoch": 0.7113410395279283, "step": 4400}, {"loss": 0.805, "grad_norm": 0.4397392272949219, "learning_rate": 0.0002, "epoch": 0.7129577237086735, "step": 4410}, {"loss": 0.7651, "grad_norm": 0.45228442549705505, "learning_rate": 0.0002, "epoch": 0.7145744078894188, "step": 4420}, {"loss": 0.7938, "grad_norm": 0.4839673936367035, "learning_rate": 0.0002, "epoch": 0.7161910920701641, "step": 4430}, {"loss": 0.8362, "grad_norm": 0.6140755414962769, "learning_rate": 0.0002, "epoch": 0.7178077762509094, "step": 4440}, {"loss": 0.7722, "grad_norm": 0.6841378808021545, "learning_rate": 0.0002, "epoch": 0.7194244604316546, "step": 4450}, {"loss": 0.8177, "grad_norm": 0.6664239168167114, "learning_rate": 0.0002, "epoch": 0.7210411446124, "step": 4460}, {"loss": 0.7983, "grad_norm": 0.47552719712257385, "learning_rate": 0.0002, "epoch": 0.7226578287931452, "step": 4470}, {"loss": 0.8982, "grad_norm": 0.6649776101112366, "learning_rate": 0.0002, "epoch": 0.7242745129738906, "step": 4480}, {"loss": 0.8074, "grad_norm": 0.5159541964530945, "learning_rate": 0.0002, "epoch": 0.7258911971546358, "step": 4490}, {"loss": 0.7786, "grad_norm": 0.6693112850189209, "learning_rate": 0.0002, "epoch": 0.7275078813353811, "step": 4500}, {"loss": 0.8655, "grad_norm": 0.48870977759361267, "learning_rate": 0.0002, "epoch": 0.7291245655161265, "step": 4510}, {"loss": 0.7337, "grad_norm": 0.4857887923717499, "learning_rate": 0.0002, "epoch": 0.7307412496968717, "step": 4520}, {"loss": 0.8026, "grad_norm": 0.5515662431716919, "learning_rate": 0.0002, "epoch": 0.732357933877617, "step": 4530}, {"loss": 0.8031, "grad_norm": 0.6292222738265991, "learning_rate": 0.0002, "epoch": 0.7339746180583623, "step": 4540}, {"loss": 0.7749, "grad_norm": 0.48265689611434937, "learning_rate": 0.0002, "epoch": 0.7355913022391076, "step": 4550}, {"loss": 0.8499, "grad_norm": 0.8044266104698181, "learning_rate": 0.0002, "epoch": 0.7372079864198529, "step": 4560}, {"loss": 0.8162, "grad_norm": 0.6111769676208496, "learning_rate": 0.0002, "epoch": 0.7388246706005982, "step": 4570}, {"loss": 0.7291, "grad_norm": 0.5229553580284119, "learning_rate": 0.0002, "epoch": 0.7404413547813434, "step": 4580}, {"loss": 0.8038, "grad_norm": 0.6054152250289917, "learning_rate": 0.0002, "epoch": 0.7420580389620888, "step": 4590}, {"loss": 0.8169, "grad_norm": 0.5574966669082642, "learning_rate": 0.0002, "epoch": 0.743674723142834, "step": 4600}, {"loss": 0.8439, "grad_norm": 0.5395817160606384, "learning_rate": 0.0002, "epoch": 0.7452914073235793, "step": 4610}, {"loss": 0.8495, "grad_norm": 0.7116472721099854, "learning_rate": 0.0002, "epoch": 0.7469080915043246, "step": 4620}, {"loss": 0.7743, "grad_norm": 0.5618700981140137, "learning_rate": 0.0002, "epoch": 0.7485247756850699, "step": 4630}, {"loss": 0.7744, "grad_norm": 0.5802770853042603, "learning_rate": 0.0002, "epoch": 0.7501414598658153, "step": 4640}, {"loss": 0.7924, "grad_norm": 0.5690428018569946, "learning_rate": 0.0002, "epoch": 0.7517581440465605, "step": 4650}, {"loss": 0.8017, "grad_norm": 0.4813360273838043, "learning_rate": 0.0002, "epoch": 0.7533748282273058, "step": 4660}, {"loss": 0.8108, "grad_norm": 0.5434042811393738, "learning_rate": 0.0002, "epoch": 0.7549915124080511, "step": 4670}, {"loss": 0.7824, "grad_norm": 0.5502099990844727, "learning_rate": 0.0002, "epoch": 0.7566081965887964, "step": 4680}, {"loss": 0.8598, "grad_norm": 0.6020621061325073, "learning_rate": 0.0002, "epoch": 0.7582248807695416, "step": 4690}, {"loss": 0.7937, "grad_norm": 0.4922301471233368, "learning_rate": 0.0002, "epoch": 0.759841564950287, "step": 4700}, {"loss": 0.788, "grad_norm": 0.6492828726768494, "learning_rate": 0.0002, "epoch": 0.7614582491310322, "step": 4710}, {"loss": 0.8313, "grad_norm": 0.4865580201148987, "learning_rate": 0.0002, "epoch": 0.7630749333117776, "step": 4720}, {"loss": 0.7966, "grad_norm": 0.5971422791481018, "learning_rate": 0.0002, "epoch": 0.7646916174925228, "step": 4730}, {"loss": 0.8298, "grad_norm": 0.6832674145698547, "learning_rate": 0.0002, "epoch": 0.7663083016732681, "step": 4740}, {"loss": 0.8156, "grad_norm": 0.500908613204956, "learning_rate": 0.0002, "epoch": 0.7679249858540134, "step": 4750}, {"loss": 0.8383, "grad_norm": 0.6112465858459473, "learning_rate": 0.0002, "epoch": 0.7695416700347587, "step": 4760}, {"loss": 0.76, "grad_norm": 0.5753506422042847, "learning_rate": 0.0002, "epoch": 0.771158354215504, "step": 4770}, {"loss": 0.8297, "grad_norm": 0.6529405117034912, "learning_rate": 0.0002, "epoch": 0.7727750383962493, "step": 4780}, {"loss": 0.8171, "grad_norm": 0.5916843414306641, "learning_rate": 0.0002, "epoch": 0.7743917225769946, "step": 4790}, {"loss": 0.83, "grad_norm": 0.4821224510669708, "learning_rate": 0.0002, "epoch": 0.7760084067577399, "step": 4800}, {"loss": 0.7703, "grad_norm": 0.5532580018043518, "learning_rate": 0.0002, "epoch": 0.7776250909384852, "step": 4810}, {"loss": 0.7363, "grad_norm": 0.4604877233505249, "learning_rate": 0.0002, "epoch": 0.7792417751192304, "step": 4820}, {"loss": 0.7506, "grad_norm": 0.5009613037109375, "learning_rate": 0.0002, "epoch": 0.7808584592999758, "step": 4830}, {"loss": 0.7863, "grad_norm": 0.6448560357093811, "learning_rate": 0.0002, "epoch": 0.782475143480721, "step": 4840}, {"loss": 0.7957, "grad_norm": 0.44327953457832336, "learning_rate": 0.0002, "epoch": 0.7840918276614663, "step": 4850}, {"loss": 0.7925, "grad_norm": 0.5355411171913147, "learning_rate": 0.0002, "epoch": 0.7857085118422116, "step": 4860}, {"loss": 0.7754, "grad_norm": 0.5635677576065063, "learning_rate": 0.0002, "epoch": 0.7873251960229569, "step": 4870}, {"loss": 0.7931, "grad_norm": 0.5417491793632507, "learning_rate": 0.0002, "epoch": 0.7889418802037023, "step": 4880}, {"loss": 0.7819, "grad_norm": 0.4567430913448334, "learning_rate": 0.0002, "epoch": 0.7905585643844475, "step": 4890}, {"loss": 0.8454, "grad_norm": 0.44651296734809875, "learning_rate": 0.0002, "epoch": 0.7921752485651928, "step": 4900}, {"loss": 0.7959, "grad_norm": 0.5741217136383057, "learning_rate": 0.0002, "epoch": 0.7937919327459381, "step": 4910}, {"loss": 0.8093, "grad_norm": 0.6605045199394226, "learning_rate": 0.0002, "epoch": 0.7954086169266834, "step": 4920}, {"loss": 0.77, "grad_norm": 0.5126531720161438, "learning_rate": 0.0002, "epoch": 0.7970253011074286, "step": 4930}, {"loss": 0.7793, "grad_norm": 0.513648271560669, "learning_rate": 0.0002, "epoch": 0.798641985288174, "step": 4940}, {"loss": 0.8314, "grad_norm": 0.5350404381752014, "learning_rate": 0.0002, "epoch": 0.8002586694689192, "step": 4950}, {"loss": 0.7649, "grad_norm": 0.5731674432754517, "learning_rate": 0.0002, "epoch": 0.8018753536496646, "step": 4960}, {"loss": 0.8572, "grad_norm": 0.5974258184432983, "learning_rate": 0.0002, "epoch": 0.8034920378304098, "step": 4970}, {"loss": 0.7972, "grad_norm": 0.8774799704551697, "learning_rate": 0.0002, "epoch": 0.8051087220111551, "step": 4980}, {"loss": 0.7899, "grad_norm": 0.5994430184364319, "learning_rate": 0.0002, "epoch": 0.8067254061919004, "step": 4990}, {"loss": 0.7736, "grad_norm": 0.4894903004169464, "learning_rate": 0.0002, "epoch": 0.8083420903726457, "step": 5000}, {"loss": 0.78, "grad_norm": 0.5218459367752075, "learning_rate": 0.0002, "epoch": 0.809958774553391, "step": 5010}, {"loss": 0.817, "grad_norm": 0.5232468843460083, "learning_rate": 0.0002, "epoch": 0.8115754587341363, "step": 5020}, {"loss": 0.7704, "grad_norm": 0.44358372688293457, "learning_rate": 0.0002, "epoch": 0.8131921429148816, "step": 5030}, {"loss": 0.785, "grad_norm": 0.6202037334442139, "learning_rate": 0.0002, "epoch": 0.8148088270956269, "step": 5040}, {"loss": 0.7351, "grad_norm": 0.7721474170684814, "learning_rate": 0.0002, "epoch": 0.8164255112763722, "step": 5050}, {"loss": 0.8297, "grad_norm": 0.5568501353263855, "learning_rate": 0.0002, "epoch": 0.8180421954571174, "step": 5060}, {"loss": 0.7733, "grad_norm": 0.49148809909820557, "learning_rate": 0.0002, "epoch": 0.8196588796378628, "step": 5070}, {"loss": 0.8054, "grad_norm": 0.4956012964248657, "learning_rate": 0.0002, "epoch": 0.821275563818608, "step": 5080}, {"loss": 0.8201, "grad_norm": 0.6078833937644958, "learning_rate": 0.0002, "epoch": 0.8228922479993533, "step": 5090}, {"loss": 0.828, "grad_norm": 0.46906954050064087, "learning_rate": 0.0002, "epoch": 0.8245089321800986, "step": 5100}, {"loss": 0.7703, "grad_norm": 0.50812166929245, "learning_rate": 0.0002, "epoch": 0.8261256163608439, "step": 5110}, {"loss": 0.8243, "grad_norm": 0.5319661498069763, "learning_rate": 0.0002, "epoch": 0.8277423005415891, "step": 5120}, {"loss": 0.7798, "grad_norm": 0.4949689209461212, "learning_rate": 0.0002, "epoch": 0.8293589847223345, "step": 5130}, {"loss": 0.7428, "grad_norm": 0.5151591300964355, "learning_rate": 0.0002, "epoch": 0.8309756689030798, "step": 5140}, {"loss": 0.8147, "grad_norm": 0.5530214309692383, "learning_rate": 0.0002, "epoch": 0.8325923530838251, "step": 5150}, {"loss": 0.8251, "grad_norm": 0.6297410130500793, "learning_rate": 0.0002, "epoch": 0.8342090372645704, "step": 5160}, {"loss": 0.8067, "grad_norm": 0.5466840267181396, "learning_rate": 0.0002, "epoch": 0.8358257214453156, "step": 5170}, {"loss": 0.7875, "grad_norm": 0.652913510799408, "learning_rate": 0.0002, "epoch": 0.837442405626061, "step": 5180}, {"loss": 0.8295, "grad_norm": 0.5811293125152588, "learning_rate": 0.0002, "epoch": 0.8390590898068062, "step": 5190}, {"loss": 0.7412, "grad_norm": 0.5109550952911377, "learning_rate": 0.0002, "epoch": 0.8406757739875516, "step": 5200}, {"loss": 0.8077, "grad_norm": 0.4551706612110138, "learning_rate": 0.0002, "epoch": 0.8422924581682968, "step": 5210}, {"loss": 0.7827, "grad_norm": 0.5813754200935364, "learning_rate": 0.0002, "epoch": 0.8439091423490421, "step": 5220}, {"loss": 0.802, "grad_norm": 0.5856947898864746, "learning_rate": 0.0002, "epoch": 0.8455258265297874, "step": 5230}, {"loss": 0.7957, "grad_norm": 0.5482739210128784, "learning_rate": 0.0002, "epoch": 0.8471425107105327, "step": 5240}, {"loss": 0.8295, "grad_norm": 0.49023720622062683, "learning_rate": 0.0002, "epoch": 0.8487591948912779, "step": 5250}, {"loss": 0.8022, "grad_norm": 0.49472475051879883, "learning_rate": 0.0002, "epoch": 0.8503758790720233, "step": 5260}, {"loss": 0.8001, "grad_norm": 0.5490226745605469, "learning_rate": 0.0002, "epoch": 0.8519925632527686, "step": 5270}, {"loss": 0.8333, "grad_norm": 0.5340665578842163, "learning_rate": 0.0002, "epoch": 0.8536092474335139, "step": 5280}, {"loss": 0.8277, "grad_norm": 0.5962483882904053, "learning_rate": 0.0002, "epoch": 0.8552259316142592, "step": 5290}, {"loss": 0.8765, "grad_norm": 0.586358368396759, "learning_rate": 0.0002, "epoch": 0.8568426157950044, "step": 5300}, {"loss": 0.7831, "grad_norm": 0.49120277166366577, "learning_rate": 0.0002, "epoch": 0.8584592999757498, "step": 5310}, {"loss": 0.8162, "grad_norm": 0.5887332558631897, "learning_rate": 0.0002, "epoch": 0.860075984156495, "step": 5320}, {"loss": 0.7464, "grad_norm": 0.42496153712272644, "learning_rate": 0.0002, "epoch": 0.8616926683372403, "step": 5330}, {"loss": 0.7905, "grad_norm": 0.5489874482154846, "learning_rate": 0.0002, "epoch": 0.8633093525179856, "step": 5340}, {"loss": 0.7958, "grad_norm": 0.5850813984870911, "learning_rate": 0.0002, "epoch": 0.8649260366987309, "step": 5350}, {"loss": 0.7642, "grad_norm": 0.517487108707428, "learning_rate": 0.0002, "epoch": 0.8665427208794761, "step": 5360}, {"loss": 0.7801, "grad_norm": 0.5339142680168152, "learning_rate": 0.0002, "epoch": 0.8681594050602215, "step": 5370}, {"loss": 0.818, "grad_norm": 0.6236387491226196, "learning_rate": 0.0002, "epoch": 0.8697760892409668, "step": 5380}, {"loss": 0.7708, "grad_norm": 0.5752192735671997, "learning_rate": 0.0002, "epoch": 0.8713927734217121, "step": 5390}, {"loss": 0.8542, "grad_norm": 0.6724614500999451, "learning_rate": 0.0002, "epoch": 0.8730094576024574, "step": 5400}, {"loss": 0.7581, "grad_norm": 0.5280613303184509, "learning_rate": 0.0002, "epoch": 0.8746261417832026, "step": 5410}, {"loss": 0.8231, "grad_norm": 0.44033288955688477, "learning_rate": 0.0002, "epoch": 0.876242825963948, "step": 5420}, {"loss": 0.8839, "grad_norm": 0.5199708342552185, "learning_rate": 0.0002, "epoch": 0.8778595101446932, "step": 5430}, {"loss": 0.7852, "grad_norm": 0.46778348088264465, "learning_rate": 0.0002, "epoch": 0.8794761943254386, "step": 5440}, {"loss": 0.7834, "grad_norm": 0.4657754898071289, "learning_rate": 0.0002, "epoch": 0.8810928785061838, "step": 5450}, {"loss": 0.7799, "grad_norm": 0.5472902655601501, "learning_rate": 0.0002, "epoch": 0.8827095626869291, "step": 5460}, {"loss": 0.8253, "grad_norm": 0.4876766800880432, "learning_rate": 0.0002, "epoch": 0.8843262468676744, "step": 5470}, {"loss": 0.7906, "grad_norm": 0.5057248473167419, "learning_rate": 0.0002, "epoch": 0.8859429310484197, "step": 5480}, {"loss": 0.8124, "grad_norm": 0.4637320637702942, "learning_rate": 0.0002, "epoch": 0.8875596152291649, "step": 5490}, {"loss": 0.781, "grad_norm": 0.471955806016922, "learning_rate": 0.0002, "epoch": 0.8891762994099103, "step": 5500}, {"loss": 0.8057, "grad_norm": 0.5209813714027405, "learning_rate": 0.0002, "epoch": 0.8907929835906556, "step": 5510}, {"loss": 0.8106, "grad_norm": 0.6213834285736084, "learning_rate": 0.0002, "epoch": 0.8924096677714008, "step": 5520}, {"loss": 0.7787, "grad_norm": 0.5215408205986023, "learning_rate": 0.0002, "epoch": 0.8940263519521462, "step": 5530}, {"loss": 0.8174, "grad_norm": 0.580478310585022, "learning_rate": 0.0002, "epoch": 0.8956430361328914, "step": 5540}, {"loss": 0.8371, "grad_norm": 0.49102169275283813, "learning_rate": 0.0002, "epoch": 0.8972597203136368, "step": 5550}, {"loss": 0.7806, "grad_norm": 0.6043479442596436, "learning_rate": 0.0002, "epoch": 0.898876404494382, "step": 5560}, {"loss": 0.7754, "grad_norm": 0.5636463165283203, "learning_rate": 0.0002, "epoch": 0.9004930886751273, "step": 5570}, {"loss": 0.8145, "grad_norm": 0.5620124340057373, "learning_rate": 0.0002, "epoch": 0.9021097728558726, "step": 5580}, {"loss": 0.8083, "grad_norm": 0.5206354856491089, "learning_rate": 0.0002, "epoch": 0.9037264570366179, "step": 5590}, {"loss": 0.8557, "grad_norm": 0.5798229575157166, "learning_rate": 0.0002, "epoch": 0.9053431412173631, "step": 5600}, {"loss": 0.8097, "grad_norm": 0.6428212523460388, "learning_rate": 0.0002, "epoch": 0.9069598253981085, "step": 5610}, {"loss": 0.7839, "grad_norm": 0.48064687848091125, "learning_rate": 0.0002, "epoch": 0.9085765095788537, "step": 5620}, {"loss": 0.8343, "grad_norm": 0.6347860097885132, "learning_rate": 0.0002, "epoch": 0.9101931937595991, "step": 5630}, {"loss": 0.851, "grad_norm": 0.5353913307189941, "learning_rate": 0.0002, "epoch": 0.9118098779403444, "step": 5640}, {"loss": 0.7736, "grad_norm": 0.5323944091796875, "learning_rate": 0.0002, "epoch": 0.9134265621210896, "step": 5650}, {"loss": 0.8393, "grad_norm": 0.5261843204498291, "learning_rate": 0.0002, "epoch": 0.915043246301835, "step": 5660}, {"loss": 0.7355, "grad_norm": 0.5451326966285706, "learning_rate": 0.0002, "epoch": 0.9166599304825802, "step": 5670}, {"loss": 0.8012, "grad_norm": 0.5183324217796326, "learning_rate": 0.0002, "epoch": 0.9182766146633256, "step": 5680}, {"loss": 0.7659, "grad_norm": 0.47229018807411194, "learning_rate": 0.0002, "epoch": 0.9198932988440708, "step": 5690}, {"loss": 0.7757, "grad_norm": 0.49180513620376587, "learning_rate": 0.0002, "epoch": 0.9215099830248161, "step": 5700}, {"loss": 0.8735, "grad_norm": 0.5419785380363464, "learning_rate": 0.0002, "epoch": 0.9231266672055614, "step": 5710}, {"loss": 0.7378, "grad_norm": 0.5408698916435242, "learning_rate": 0.0002, "epoch": 0.9247433513863067, "step": 5720}, {"loss": 0.7701, "grad_norm": 0.5286232829093933, "learning_rate": 0.0002, "epoch": 0.9263600355670519, "step": 5730}, {"loss": 0.8242, "grad_norm": 0.7539758086204529, "learning_rate": 0.0002, "epoch": 0.9279767197477973, "step": 5740}, {"loss": 0.8118, "grad_norm": 0.5166944861412048, "learning_rate": 0.0002, "epoch": 0.9295934039285425, "step": 5750}, {"loss": 0.783, "grad_norm": 0.6601425409317017, "learning_rate": 0.0002, "epoch": 0.9312100881092878, "step": 5760}, {"loss": 0.7873, "grad_norm": 0.5029960870742798, "learning_rate": 0.0002, "epoch": 0.9328267722900332, "step": 5770}, {"loss": 0.7989, "grad_norm": 0.4926645755767822, "learning_rate": 0.0002, "epoch": 0.9344434564707784, "step": 5780}, {"loss": 0.8174, "grad_norm": 0.5739615559577942, "learning_rate": 0.0002, "epoch": 0.9360601406515238, "step": 5790}, {"loss": 0.8037, "grad_norm": 0.5058279037475586, "learning_rate": 0.0002, "epoch": 0.937676824832269, "step": 5800}, {"loss": 0.8537, "grad_norm": 0.5260962247848511, "learning_rate": 0.0002, "epoch": 0.9392935090130143, "step": 5810}, {"loss": 0.7486, "grad_norm": 0.5768588185310364, "learning_rate": 0.0002, "epoch": 0.9409101931937596, "step": 5820}, {"loss": 0.8215, "grad_norm": 0.5170126557350159, "learning_rate": 0.0002, "epoch": 0.9425268773745049, "step": 5830}, {"loss": 0.7422, "grad_norm": 0.5745864510536194, "learning_rate": 0.0002, "epoch": 0.9441435615552501, "step": 5840}, {"loss": 0.7824, "grad_norm": 0.5551357865333557, "learning_rate": 0.0002, "epoch": 0.9457602457359955, "step": 5850}, {"loss": 0.8529, "grad_norm": 0.5776078701019287, "learning_rate": 0.0002, "epoch": 0.9473769299167407, "step": 5860}, {"loss": 0.8527, "grad_norm": 0.5340062379837036, "learning_rate": 0.0002, "epoch": 0.9489936140974861, "step": 5870}, {"loss": 0.8217, "grad_norm": 0.6447290182113647, "learning_rate": 0.0002, "epoch": 0.9506102982782314, "step": 5880}, {"loss": 0.7945, "grad_norm": 0.5123815536499023, "learning_rate": 0.0002, "epoch": 0.9522269824589766, "step": 5890}, {"loss": 0.8209, "grad_norm": 0.48547613620758057, "learning_rate": 0.0002, "epoch": 0.953843666639722, "step": 5900}, {"loss": 0.7896, "grad_norm": 0.5791414976119995, "learning_rate": 0.0002, "epoch": 0.9554603508204672, "step": 5910}, {"loss": 0.8408, "grad_norm": 0.6195011734962463, "learning_rate": 0.0002, "epoch": 0.9570770350012126, "step": 5920}, {"loss": 0.7805, "grad_norm": 0.6323803067207336, "learning_rate": 0.0002, "epoch": 0.9586937191819578, "step": 5930}, {"loss": 0.8484, "grad_norm": 0.45552879571914673, "learning_rate": 0.0002, "epoch": 0.9603104033627031, "step": 5940}, {"loss": 0.7367, "grad_norm": 0.5796473622322083, "learning_rate": 0.0002, "epoch": 0.9619270875434484, "step": 5950}, {"loss": 0.7672, "grad_norm": 0.647261381149292, "learning_rate": 0.0002, "epoch": 0.9635437717241937, "step": 5960}, {"loss": 0.8086, "grad_norm": 0.5487682819366455, "learning_rate": 0.0002, "epoch": 0.9651604559049389, "step": 5970}, {"loss": 0.7973, "grad_norm": 0.5743663907051086, "learning_rate": 0.0002, "epoch": 0.9667771400856843, "step": 5980}, {"loss": 0.8153, "grad_norm": 0.5470591187477112, "learning_rate": 0.0002, "epoch": 0.9683938242664295, "step": 5990}, {"loss": 0.8119, "grad_norm": 0.5901660323143005, "learning_rate": 0.0002, "epoch": 0.9700105084471748, "step": 6000}, {"loss": 0.8147, "grad_norm": 0.6544759273529053, "learning_rate": 0.0002, "epoch": 0.9716271926279202, "step": 6010}, {"loss": 0.7536, "grad_norm": 0.6288470029830933, "learning_rate": 0.0002, "epoch": 0.9732438768086654, "step": 6020}, {"loss": 0.7989, "grad_norm": 0.673153817653656, "learning_rate": 0.0002, "epoch": 0.9748605609894108, "step": 6030}, {"loss": 0.7556, "grad_norm": 0.42854753136634827, "learning_rate": 0.0002, "epoch": 0.976477245170156, "step": 6040}, {"loss": 0.8006, "grad_norm": 0.5227066278457642, "learning_rate": 0.0002, "epoch": 0.9780939293509013, "step": 6050}, {"loss": 0.795, "grad_norm": 0.5372416973114014, "learning_rate": 0.0002, "epoch": 0.9797106135316466, "step": 6060}, {"loss": 0.7591, "grad_norm": 0.6026402115821838, "learning_rate": 0.0002, "epoch": 0.9813272977123919, "step": 6070}, {"loss": 0.8347, "grad_norm": 0.49547791481018066, "learning_rate": 0.0002, "epoch": 0.9829439818931371, "step": 6080}, {"loss": 0.7722, "grad_norm": 0.4641951322555542, "learning_rate": 0.0002, "epoch": 0.9845606660738825, "step": 6090}, {"loss": 0.8125, "grad_norm": 0.5818535089492798, "learning_rate": 0.0002, "epoch": 0.9861773502546277, "step": 6100}, {"loss": 0.81, "grad_norm": 0.63955157995224, "learning_rate": 0.0002, "epoch": 0.9877940344353731, "step": 6110}, {"loss": 0.7547, "grad_norm": 0.5649438500404358, "learning_rate": 0.0002, "epoch": 0.9894107186161183, "step": 6120}, {"loss": 0.7861, "grad_norm": 0.5290433168411255, "learning_rate": 0.0002, "epoch": 0.9910274027968636, "step": 6130}, {"loss": 0.8109, "grad_norm": 0.6399374008178711, "learning_rate": 0.0002, "epoch": 0.992644086977609, "step": 6140}, {"loss": 0.8373, "grad_norm": 0.6736576557159424, "learning_rate": 0.0002, "epoch": 0.9942607711583542, "step": 6150}, {"loss": 0.7915, "grad_norm": 0.515420138835907, "learning_rate": 0.0002, "epoch": 0.9958774553390995, "step": 6160}, {"loss": 0.8032, "grad_norm": 0.562677800655365, "learning_rate": 0.0002, "epoch": 0.9974941395198448, "step": 6170}, {"loss": 0.8187, "grad_norm": 0.7113858461380005, "learning_rate": 0.0002, "epoch": 0.9991108237005901, "step": 6180}, {"eval_loss": 1.0871200561523438, "eval_runtime": 122.2071, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 0.9999191657909627, "step": 6185}, {"loss": 0.7507, "grad_norm": 0.7111801505088806, "learning_rate": 0.0002, "epoch": 1.0007275078813354, "step": 6190}, {"loss": 0.6865, "grad_norm": 0.5402125716209412, "learning_rate": 0.0002, "epoch": 1.0023441920620806, "step": 6200}, {"loss": 0.7625, "grad_norm": 0.6098830103874207, "learning_rate": 0.0002, "epoch": 1.003960876242826, "step": 6210}, {"loss": 0.7631, "grad_norm": 0.5829983353614807, "learning_rate": 0.0002, "epoch": 1.0055775604235713, "step": 6220}, {"loss": 0.7188, "grad_norm": 0.5614621043205261, "learning_rate": 0.0002, "epoch": 1.0071942446043165, "step": 6230}, {"loss": 0.7505, "grad_norm": 0.5954238772392273, "learning_rate": 0.0002, "epoch": 1.0088109287850617, "step": 6240}, {"loss": 0.7448, "grad_norm": 0.6480574607849121, "learning_rate": 0.0002, "epoch": 1.0104276129658072, "step": 6250}, {"loss": 0.7514, "grad_norm": 0.6051128506660461, "learning_rate": 0.0002, "epoch": 1.0120442971465524, "step": 6260}, {"loss": 0.7237, "grad_norm": 0.6318870782852173, "learning_rate": 0.0002, "epoch": 1.0136609813272976, "step": 6270}, {"loss": 0.7178, "grad_norm": 0.5048980116844177, "learning_rate": 0.0002, "epoch": 1.015277665508043, "step": 6280}, {"loss": 0.7391, "grad_norm": 0.6346936225891113, "learning_rate": 0.0002, "epoch": 1.0168943496887883, "step": 6290}, {"loss": 0.7486, "grad_norm": 0.5711665749549866, "learning_rate": 0.0002, "epoch": 1.0185110338695336, "step": 6300}, {"loss": 0.6808, "grad_norm": 0.5175361037254333, "learning_rate": 0.0002, "epoch": 1.0201277180502788, "step": 6310}, {"loss": 0.7539, "grad_norm": 0.5360831618309021, "learning_rate": 0.0002, "epoch": 1.0217444022310243, "step": 6320}, {"loss": 0.7112, "grad_norm": 0.614675760269165, "learning_rate": 0.0002, "epoch": 1.0233610864117695, "step": 6330}, {"loss": 0.7748, "grad_norm": 0.5626118183135986, "learning_rate": 0.0002, "epoch": 1.0249777705925147, "step": 6340}, {"loss": 0.7375, "grad_norm": 0.574897289276123, "learning_rate": 0.0002, "epoch": 1.02659445477326, "step": 6350}, {"loss": 0.759, "grad_norm": 0.7185447812080383, "learning_rate": 0.0002, "epoch": 1.0282111389540054, "step": 6360}, {"loss": 0.703, "grad_norm": 0.6705799698829651, "learning_rate": 0.0002, "epoch": 1.0298278231347506, "step": 6370}, {"loss": 0.7139, "grad_norm": 0.6740428805351257, "learning_rate": 0.0002, "epoch": 1.0314445073154959, "step": 6380}, {"loss": 0.7252, "grad_norm": 0.663902759552002, "learning_rate": 0.0002, "epoch": 1.0330611914962413, "step": 6390}, {"loss": 0.7065, "grad_norm": 0.5029543042182922, "learning_rate": 0.0002, "epoch": 1.0346778756769865, "step": 6400}, {"loss": 0.711, "grad_norm": 0.7813863158226013, "learning_rate": 0.0002, "epoch": 1.0362945598577318, "step": 6410}, {"loss": 0.7433, "grad_norm": 0.5396282076835632, "learning_rate": 0.0002, "epoch": 1.037911244038477, "step": 6420}, {"loss": 0.7222, "grad_norm": 0.5253293514251709, "learning_rate": 0.0002, "epoch": 1.0395279282192225, "step": 6430}, {"loss": 0.715, "grad_norm": 0.7236770987510681, "learning_rate": 0.0002, "epoch": 1.0411446123999677, "step": 6440}, {"loss": 0.7259, "grad_norm": 0.5670917630195618, "learning_rate": 0.0002, "epoch": 1.042761296580713, "step": 6450}, {"loss": 0.7195, "grad_norm": 0.6031978726387024, "learning_rate": 0.0002, "epoch": 1.0443779807614582, "step": 6460}, {"loss": 0.7648, "grad_norm": 0.5309213399887085, "learning_rate": 0.0002, "epoch": 1.0459946649422036, "step": 6470}, {"loss": 0.7161, "grad_norm": 0.7114651799201965, "learning_rate": 0.0002, "epoch": 1.0476113491229488, "step": 6480}, {"loss": 0.7583, "grad_norm": 0.5591610670089722, "learning_rate": 0.0002, "epoch": 1.049228033303694, "step": 6490}, {"loss": 0.6645, "grad_norm": 0.5185961127281189, "learning_rate": 0.0002, "epoch": 1.0508447174844395, "step": 6500}, {"loss": 0.7654, "grad_norm": 0.6510552167892456, "learning_rate": 0.0002, "epoch": 1.0524614016651848, "step": 6510}, {"loss": 0.7057, "grad_norm": 0.6557928919792175, "learning_rate": 0.0002, "epoch": 1.05407808584593, "step": 6520}, {"loss": 0.8056, "grad_norm": 0.6973192691802979, "learning_rate": 0.0002, "epoch": 1.0556947700266752, "step": 6530}, {"loss": 0.6793, "grad_norm": 0.6226583123207092, "learning_rate": 0.0002, "epoch": 1.0573114542074207, "step": 6540}, {"loss": 0.7151, "grad_norm": 0.5633195638656616, "learning_rate": 0.0002, "epoch": 1.058928138388166, "step": 6550}, {"loss": 0.7082, "grad_norm": 0.7466658353805542, "learning_rate": 0.0002, "epoch": 1.0605448225689111, "step": 6560}, {"loss": 0.7059, "grad_norm": 0.6462772488594055, "learning_rate": 0.0002, "epoch": 1.0621615067496564, "step": 6570}, {"loss": 0.7046, "grad_norm": 0.5266856551170349, "learning_rate": 0.0002, "epoch": 1.0637781909304018, "step": 6580}, {"loss": 0.7157, "grad_norm": 0.534392774105072, "learning_rate": 0.0002, "epoch": 1.065394875111147, "step": 6590}, {"loss": 0.7115, "grad_norm": 0.7514177560806274, "learning_rate": 0.0002, "epoch": 1.0670115592918923, "step": 6600}, {"loss": 0.7545, "grad_norm": 0.7593035697937012, "learning_rate": 0.0002, "epoch": 1.0686282434726375, "step": 6610}, {"loss": 0.6836, "grad_norm": 0.5277858972549438, "learning_rate": 0.0002, "epoch": 1.070244927653383, "step": 6620}, {"loss": 0.7405, "grad_norm": 0.5573670268058777, "learning_rate": 0.0002, "epoch": 1.0718616118341282, "step": 6630}, {"loss": 0.6774, "grad_norm": 0.6802396774291992, "learning_rate": 0.0002, "epoch": 1.0734782960148734, "step": 6640}, {"loss": 0.723, "grad_norm": 0.7367215752601624, "learning_rate": 0.0002, "epoch": 1.0750949801956189, "step": 6650}, {"loss": 0.7429, "grad_norm": 0.5961891412734985, "learning_rate": 0.0002, "epoch": 1.0767116643763641, "step": 6660}, {"loss": 0.6791, "grad_norm": 0.5736313462257385, "learning_rate": 0.0002, "epoch": 1.0783283485571094, "step": 6670}, {"loss": 0.7178, "grad_norm": 0.619219183921814, "learning_rate": 0.0002, "epoch": 1.0799450327378546, "step": 6680}, {"loss": 0.7318, "grad_norm": 0.6214390993118286, "learning_rate": 0.0002, "epoch": 1.0815617169186, "step": 6690}, {"loss": 0.7554, "grad_norm": 0.564536988735199, "learning_rate": 0.0002, "epoch": 1.0831784010993453, "step": 6700}, {"loss": 0.7362, "grad_norm": 0.5838140249252319, "learning_rate": 0.0002, "epoch": 1.0847950852800905, "step": 6710}, {"loss": 0.739, "grad_norm": 0.7000553607940674, "learning_rate": 0.0002, "epoch": 1.0864117694608357, "step": 6720}, {"loss": 0.7369, "grad_norm": 0.7078263759613037, "learning_rate": 0.0002, "epoch": 1.0880284536415812, "step": 6730}, {"loss": 0.7654, "grad_norm": 0.8353848457336426, "learning_rate": 0.0002, "epoch": 1.0896451378223264, "step": 6740}, {"loss": 0.7015, "grad_norm": 0.5615518689155579, "learning_rate": 0.0002, "epoch": 1.0912618220030716, "step": 6750}, {"loss": 0.7396, "grad_norm": 0.5475581288337708, "learning_rate": 0.0002, "epoch": 1.0928785061838169, "step": 6760}, {"loss": 0.7652, "grad_norm": 0.5835978388786316, "learning_rate": 0.0002, "epoch": 1.0944951903645623, "step": 6770}, {"loss": 0.7541, "grad_norm": 0.5516105890274048, "learning_rate": 0.0002, "epoch": 1.0961118745453076, "step": 6780}, {"loss": 0.6842, "grad_norm": 0.5875251889228821, "learning_rate": 0.0002, "epoch": 1.0977285587260528, "step": 6790}, {"loss": 0.6903, "grad_norm": 0.7376947999000549, "learning_rate": 0.0002, "epoch": 1.0993452429067982, "step": 6800}, {"loss": 0.7512, "grad_norm": 0.5656165480613708, "learning_rate": 0.0002, "epoch": 1.1009619270875435, "step": 6810}, {"loss": 0.7409, "grad_norm": 0.6365954279899597, "learning_rate": 0.0002, "epoch": 1.1025786112682887, "step": 6820}, {"loss": 0.7392, "grad_norm": 0.5033080577850342, "learning_rate": 0.0002, "epoch": 1.104195295449034, "step": 6830}, {"loss": 0.6909, "grad_norm": 0.617396891117096, "learning_rate": 0.0002, "epoch": 1.1058119796297794, "step": 6840}, {"loss": 0.7006, "grad_norm": 0.6395374536514282, "learning_rate": 0.0002, "epoch": 1.1074286638105246, "step": 6850}, {"loss": 0.7335, "grad_norm": 0.6775295734405518, "learning_rate": 0.0002, "epoch": 1.1090453479912699, "step": 6860}, {"loss": 0.764, "grad_norm": 0.6655223965644836, "learning_rate": 0.0002, "epoch": 1.1106620321720153, "step": 6870}, {"loss": 0.7553, "grad_norm": 0.676655113697052, "learning_rate": 0.0002, "epoch": 1.1122787163527605, "step": 6880}, {"loss": 0.7342, "grad_norm": 0.6062718629837036, "learning_rate": 0.0002, "epoch": 1.1138954005335058, "step": 6890}, {"loss": 0.7446, "grad_norm": 0.590943455696106, "learning_rate": 0.0002, "epoch": 1.115512084714251, "step": 6900}, {"loss": 0.6705, "grad_norm": 0.6315317153930664, "learning_rate": 0.0002, "epoch": 1.1171287688949965, "step": 6910}, {"loss": 0.6912, "grad_norm": 0.47979024052619934, "learning_rate": 0.0002, "epoch": 1.1187454530757417, "step": 6920}, {"loss": 0.7002, "grad_norm": 0.647298276424408, "learning_rate": 0.0002, "epoch": 1.120362137256487, "step": 6930}, {"loss": 0.7502, "grad_norm": 0.7336484789848328, "learning_rate": 0.0002, "epoch": 1.1219788214372322, "step": 6940}, {"loss": 0.693, "grad_norm": 0.5071424245834351, "learning_rate": 0.0002, "epoch": 1.1235955056179776, "step": 6950}, {"loss": 0.7378, "grad_norm": 0.6527144312858582, "learning_rate": 0.0002, "epoch": 1.1252121897987228, "step": 6960}, {"loss": 0.7228, "grad_norm": 0.6935935020446777, "learning_rate": 0.0002, "epoch": 1.126828873979468, "step": 6970}, {"loss": 0.699, "grad_norm": 0.8026931881904602, "learning_rate": 0.0002, "epoch": 1.1284455581602133, "step": 6980}, {"loss": 0.7361, "grad_norm": 0.5210393667221069, "learning_rate": 0.0002, "epoch": 1.1300622423409588, "step": 6990}, {"loss": 0.7456, "grad_norm": 0.60475093126297, "learning_rate": 0.0002, "epoch": 1.131678926521704, "step": 7000}, {"loss": 0.7495, "grad_norm": 0.6417073607444763, "learning_rate": 0.0002, "epoch": 1.1332956107024492, "step": 7010}, {"loss": 0.7459, "grad_norm": 0.6732175946235657, "learning_rate": 0.0002, "epoch": 1.1349122948831947, "step": 7020}, {"loss": 0.7278, "grad_norm": 0.6719491481781006, "learning_rate": 0.0002, "epoch": 1.13652897906394, "step": 7030}, {"loss": 0.7694, "grad_norm": 0.5708295106887817, "learning_rate": 0.0002, "epoch": 1.1381456632446851, "step": 7040}, {"loss": 0.7823, "grad_norm": 0.7141719460487366, "learning_rate": 0.0002, "epoch": 1.1397623474254304, "step": 7050}, {"loss": 0.764, "grad_norm": 0.6187017560005188, "learning_rate": 0.0002, "epoch": 1.1413790316061758, "step": 7060}, {"loss": 0.7657, "grad_norm": 0.50581294298172, "learning_rate": 0.0002, "epoch": 1.142995715786921, "step": 7070}, {"loss": 0.7357, "grad_norm": 0.5620143413543701, "learning_rate": 0.0002, "epoch": 1.1446123999676663, "step": 7080}, {"loss": 0.7287, "grad_norm": 0.6231929659843445, "learning_rate": 0.0002, "epoch": 1.1462290841484115, "step": 7090}, {"loss": 0.7328, "grad_norm": 0.5775774121284485, "learning_rate": 0.0002, "epoch": 1.147845768329157, "step": 7100}, {"loss": 0.7728, "grad_norm": 0.6492809653282166, "learning_rate": 0.0002, "epoch": 1.1494624525099022, "step": 7110}, {"loss": 0.7545, "grad_norm": 0.6434972286224365, "learning_rate": 0.0002, "epoch": 1.1510791366906474, "step": 7120}, {"loss": 0.7374, "grad_norm": 0.6191812753677368, "learning_rate": 0.0002, "epoch": 1.1526958208713927, "step": 7130}, {"loss": 0.7276, "grad_norm": 0.6690331697463989, "learning_rate": 0.0002, "epoch": 1.1543125050521381, "step": 7140}, {"loss": 0.7704, "grad_norm": 0.5977938175201416, "learning_rate": 0.0002, "epoch": 1.1559291892328833, "step": 7150}, {"loss": 0.7251, "grad_norm": 0.6195854544639587, "learning_rate": 0.0002, "epoch": 1.1575458734136286, "step": 7160}, {"loss": 0.7249, "grad_norm": 0.5752048492431641, "learning_rate": 0.0002, "epoch": 1.159162557594374, "step": 7170}, {"loss": 0.7593, "grad_norm": 0.589081883430481, "learning_rate": 0.0002, "epoch": 1.1607792417751193, "step": 7180}, {"loss": 0.704, "grad_norm": 0.756996750831604, "learning_rate": 0.0002, "epoch": 1.1623959259558645, "step": 7190}, {"loss": 0.7404, "grad_norm": 0.7614967226982117, "learning_rate": 0.0002, "epoch": 1.1640126101366097, "step": 7200}, {"loss": 0.7867, "grad_norm": 0.6120437979698181, "learning_rate": 0.0002, "epoch": 1.1656292943173552, "step": 7210}, {"loss": 0.7384, "grad_norm": 0.6210004687309265, "learning_rate": 0.0002, "epoch": 1.1672459784981004, "step": 7220}, {"loss": 0.7251, "grad_norm": 0.6044116020202637, "learning_rate": 0.0002, "epoch": 1.1688626626788456, "step": 7230}, {"loss": 0.7361, "grad_norm": 0.5418457388877869, "learning_rate": 0.0002, "epoch": 1.170479346859591, "step": 7240}, {"loss": 0.6938, "grad_norm": 0.6413537263870239, "learning_rate": 0.0002, "epoch": 1.1720960310403363, "step": 7250}, {"loss": 0.6978, "grad_norm": 0.5777867436408997, "learning_rate": 0.0002, "epoch": 1.1737127152210816, "step": 7260}, {"loss": 0.7503, "grad_norm": 0.7092402577400208, "learning_rate": 0.0002, "epoch": 1.1753293994018268, "step": 7270}, {"loss": 0.7487, "grad_norm": 0.6351709365844727, "learning_rate": 0.0002, "epoch": 1.176946083582572, "step": 7280}, {"loss": 0.7527, "grad_norm": 0.6172189712524414, "learning_rate": 0.0002, "epoch": 1.1785627677633175, "step": 7290}, {"loss": 0.7319, "grad_norm": 0.6801714897155762, "learning_rate": 0.0002, "epoch": 1.1801794519440627, "step": 7300}, {"loss": 0.6941, "grad_norm": 0.6044712066650391, "learning_rate": 0.0002, "epoch": 1.181796136124808, "step": 7310}, {"loss": 0.6951, "grad_norm": 0.7413212060928345, "learning_rate": 0.0002, "epoch": 1.1834128203055534, "step": 7320}, {"loss": 0.7396, "grad_norm": 0.5303856134414673, "learning_rate": 0.0002, "epoch": 1.1850295044862986, "step": 7330}, {"loss": 0.6915, "grad_norm": 0.5647098422050476, "learning_rate": 0.0002, "epoch": 1.1866461886670439, "step": 7340}, {"loss": 0.7506, "grad_norm": 0.7374135255813599, "learning_rate": 0.0002, "epoch": 1.188262872847789, "step": 7350}, {"loss": 0.7041, "grad_norm": 0.5710089206695557, "learning_rate": 0.0002, "epoch": 1.1898795570285345, "step": 7360}, {"loss": 0.8289, "grad_norm": 0.6073619723320007, "learning_rate": 0.0002, "epoch": 1.1914962412092798, "step": 7370}, {"loss": 0.7722, "grad_norm": 0.5899916887283325, "learning_rate": 0.0002, "epoch": 1.193112925390025, "step": 7380}, {"loss": 0.756, "grad_norm": 0.7762434482574463, "learning_rate": 0.0002, "epoch": 1.1947296095707705, "step": 7390}, {"loss": 0.7319, "grad_norm": 0.679949939250946, "learning_rate": 0.0002, "epoch": 1.1963462937515157, "step": 7400}, {"loss": 0.7599, "grad_norm": 0.6106849312782288, "learning_rate": 0.0002, "epoch": 1.197962977932261, "step": 7410}, {"loss": 0.7648, "grad_norm": 0.682461678981781, "learning_rate": 0.0002, "epoch": 1.1995796621130062, "step": 7420}, {"loss": 0.7741, "grad_norm": 0.6087017059326172, "learning_rate": 0.0002, "epoch": 1.2011963462937516, "step": 7430}, {"loss": 0.7642, "grad_norm": 0.63739013671875, "learning_rate": 0.0002, "epoch": 1.2028130304744968, "step": 7440}, {"loss": 0.7611, "grad_norm": 0.6154777407646179, "learning_rate": 0.0002, "epoch": 1.204429714655242, "step": 7450}, {"loss": 0.7565, "grad_norm": 0.7491534948348999, "learning_rate": 0.0002, "epoch": 1.2060463988359873, "step": 7460}, {"loss": 0.698, "grad_norm": 0.6664797067642212, "learning_rate": 0.0002, "epoch": 1.2076630830167328, "step": 7470}, {"loss": 0.7456, "grad_norm": 0.6660266518592834, "learning_rate": 0.0002, "epoch": 1.209279767197478, "step": 7480}, {"loss": 0.714, "grad_norm": 0.6972551345825195, "learning_rate": 0.0002, "epoch": 1.2108964513782232, "step": 7490}, {"loss": 0.7023, "grad_norm": 0.6157945990562439, "learning_rate": 0.0002, "epoch": 1.2125131355589684, "step": 7500}, {"loss": 0.7326, "grad_norm": 0.5199310183525085, "learning_rate": 0.0002, "epoch": 1.214129819739714, "step": 7510}, {"loss": 0.7586, "grad_norm": 0.577610433101654, "learning_rate": 0.0002, "epoch": 1.2157465039204591, "step": 7520}, {"loss": 0.7179, "grad_norm": 0.53652423620224, "learning_rate": 0.0002, "epoch": 1.2173631881012044, "step": 7530}, {"loss": 0.7393, "grad_norm": 0.6479050517082214, "learning_rate": 0.0002, "epoch": 1.2189798722819498, "step": 7540}, {"loss": 0.7534, "grad_norm": 0.618748128414154, "learning_rate": 0.0002, "epoch": 1.220596556462695, "step": 7550}, {"loss": 0.6886, "grad_norm": 0.6311424374580383, "learning_rate": 0.0002, "epoch": 1.2222132406434403, "step": 7560}, {"loss": 0.7272, "grad_norm": 0.6595825552940369, "learning_rate": 0.0002, "epoch": 1.2238299248241855, "step": 7570}, {"loss": 0.7353, "grad_norm": 0.5198960900306702, "learning_rate": 0.0002, "epoch": 1.225446609004931, "step": 7580}, {"loss": 0.674, "grad_norm": 0.578650712966919, "learning_rate": 0.0002, "epoch": 1.2270632931856762, "step": 7590}, {"loss": 0.7507, "grad_norm": 0.6080220937728882, "learning_rate": 0.0002, "epoch": 1.2286799773664214, "step": 7600}, {"loss": 0.7733, "grad_norm": 0.7050248384475708, "learning_rate": 0.0002, "epoch": 1.2302966615471669, "step": 7610}, {"loss": 0.7032, "grad_norm": 0.6652196049690247, "learning_rate": 0.0002, "epoch": 1.2319133457279121, "step": 7620}, {"loss": 0.7085, "grad_norm": 0.7322776317596436, "learning_rate": 0.0002, "epoch": 1.2335300299086573, "step": 7630}, {"loss": 0.7402, "grad_norm": 0.4998728036880493, "learning_rate": 0.0002, "epoch": 1.2351467140894026, "step": 7640}, {"loss": 0.7214, "grad_norm": 0.6428788900375366, "learning_rate": 0.0002, "epoch": 1.2367633982701478, "step": 7650}, {"loss": 0.7699, "grad_norm": 0.585242509841919, "learning_rate": 0.0002, "epoch": 1.2383800824508933, "step": 7660}, {"loss": 0.7621, "grad_norm": 0.5211917757987976, "learning_rate": 0.0002, "epoch": 1.2399967666316385, "step": 7670}, {"loss": 0.746, "grad_norm": 0.6490384340286255, "learning_rate": 0.0002, "epoch": 1.2416134508123837, "step": 7680}, {"loss": 0.7186, "grad_norm": 0.6249763369560242, "learning_rate": 0.0002, "epoch": 1.2432301349931292, "step": 7690}, {"loss": 0.7761, "grad_norm": 0.71870356798172, "learning_rate": 0.0002, "epoch": 1.2448468191738744, "step": 7700}, {"loss": 0.7525, "grad_norm": 0.6761967539787292, "learning_rate": 0.0002, "epoch": 1.2464635033546196, "step": 7710}, {"loss": 0.7501, "grad_norm": 0.6500617265701294, "learning_rate": 0.0002, "epoch": 1.2480801875353649, "step": 7720}, {"loss": 0.7903, "grad_norm": 0.8069869875907898, "learning_rate": 0.0002, "epoch": 1.2496968717161103, "step": 7730}, {"loss": 0.6747, "grad_norm": 0.6044608950614929, "learning_rate": 0.0002, "epoch": 1.2513135558968556, "step": 7740}, {"loss": 0.6825, "grad_norm": 0.6573283076286316, "learning_rate": 0.0002, "epoch": 1.2529302400776008, "step": 7750}, {"loss": 0.7617, "grad_norm": 0.625430166721344, "learning_rate": 0.0002, "epoch": 1.2545469242583462, "step": 7760}, {"loss": 0.7041, "grad_norm": 0.5442022681236267, "learning_rate": 0.0002, "epoch": 1.2561636084390915, "step": 7770}, {"loss": 0.7172, "grad_norm": 0.6818386912345886, "learning_rate": 0.0002, "epoch": 1.2577802926198367, "step": 7780}, {"loss": 0.696, "grad_norm": 0.6381874084472656, "learning_rate": 0.0002, "epoch": 1.259396976800582, "step": 7790}, {"loss": 0.6834, "grad_norm": 0.6269212961196899, "learning_rate": 0.0002, "epoch": 1.2610136609813272, "step": 7800}, {"loss": 0.7821, "grad_norm": 0.600121259689331, "learning_rate": 0.0002, "epoch": 1.2626303451620726, "step": 7810}, {"loss": 0.7761, "grad_norm": 0.6337703466415405, "learning_rate": 0.0002, "epoch": 1.2642470293428179, "step": 7820}, {"loss": 0.732, "grad_norm": 0.7234963774681091, "learning_rate": 0.0002, "epoch": 1.2658637135235633, "step": 7830}, {"loss": 0.785, "grad_norm": 0.800184965133667, "learning_rate": 0.0002, "epoch": 1.2674803977043085, "step": 7840}, {"loss": 0.7426, "grad_norm": 0.7539464831352234, "learning_rate": 0.0002, "epoch": 1.2690970818850538, "step": 7850}, {"loss": 0.7496, "grad_norm": 0.5493760704994202, "learning_rate": 0.0002, "epoch": 1.270713766065799, "step": 7860}, {"loss": 0.7537, "grad_norm": 0.7477145791053772, "learning_rate": 0.0002, "epoch": 1.2723304502465442, "step": 7870}, {"loss": 0.7573, "grad_norm": 0.6366362571716309, "learning_rate": 0.0002, "epoch": 1.2739471344272897, "step": 7880}, {"loss": 0.7608, "grad_norm": 0.7419533729553223, "learning_rate": 0.0002, "epoch": 1.275563818608035, "step": 7890}, {"loss": 0.7873, "grad_norm": 0.6141223311424255, "learning_rate": 0.0002, "epoch": 1.2771805027887801, "step": 7900}, {"loss": 0.6916, "grad_norm": 0.7522598505020142, "learning_rate": 0.0002, "epoch": 1.2787971869695256, "step": 7910}, {"loss": 0.7097, "grad_norm": 0.6935804486274719, "learning_rate": 0.0002, "epoch": 1.2804138711502708, "step": 7920}, {"loss": 0.7185, "grad_norm": 0.7239290475845337, "learning_rate": 0.0002, "epoch": 1.282030555331016, "step": 7930}, {"loss": 0.7145, "grad_norm": 0.8800187110900879, "learning_rate": 0.0002, "epoch": 1.2836472395117613, "step": 7940}, {"loss": 0.6991, "grad_norm": 0.540458083152771, "learning_rate": 0.0002, "epoch": 1.2852639236925067, "step": 7950}, {"loss": 0.7139, "grad_norm": 0.6492934226989746, "learning_rate": 0.0002, "epoch": 1.286880607873252, "step": 7960}, {"loss": 0.7742, "grad_norm": 0.6543959379196167, "learning_rate": 0.0002, "epoch": 1.2884972920539972, "step": 7970}, {"loss": 0.7316, "grad_norm": 0.5804705619812012, "learning_rate": 0.0002, "epoch": 1.2901139762347427, "step": 7980}, {"loss": 0.796, "grad_norm": 0.7074727416038513, "learning_rate": 0.0002, "epoch": 1.291730660415488, "step": 7990}, {"loss": 0.7034, "grad_norm": 0.5347974300384521, "learning_rate": 0.0002, "epoch": 1.2933473445962331, "step": 8000}, {"loss": 0.738, "grad_norm": 0.6457298398017883, "learning_rate": 0.0002, "epoch": 1.2949640287769784, "step": 8010}, {"loss": 0.7634, "grad_norm": 0.6407219171524048, "learning_rate": 0.0002, "epoch": 1.2965807129577236, "step": 8020}, {"loss": 0.7506, "grad_norm": 0.828439474105835, "learning_rate": 0.0002, "epoch": 1.298197397138469, "step": 8030}, {"loss": 0.735, "grad_norm": 0.4840380549430847, "learning_rate": 0.0002, "epoch": 1.2998140813192143, "step": 8040}, {"loss": 0.7283, "grad_norm": 0.5921024680137634, "learning_rate": 0.0002, "epoch": 1.3014307654999595, "step": 8050}, {"loss": 0.7477, "grad_norm": 0.6170315146446228, "learning_rate": 0.0002, "epoch": 1.303047449680705, "step": 8060}, {"loss": 0.7534, "grad_norm": 0.5374847054481506, "learning_rate": 0.0002, "epoch": 1.3046641338614502, "step": 8070}, {"loss": 0.7593, "grad_norm": 0.545758068561554, "learning_rate": 0.0002, "epoch": 1.3062808180421954, "step": 8080}, {"loss": 0.7463, "grad_norm": 0.55641770362854, "learning_rate": 0.0002, "epoch": 1.3078975022229407, "step": 8090}, {"loss": 0.7594, "grad_norm": 0.6724897027015686, "learning_rate": 0.0002, "epoch": 1.309514186403686, "step": 8100}, {"loss": 0.7105, "grad_norm": 0.6923972368240356, "learning_rate": 0.0002, "epoch": 1.3111308705844313, "step": 8110}, {"loss": 0.7149, "grad_norm": 0.5136841535568237, "learning_rate": 0.0002, "epoch": 1.3127475547651766, "step": 8120}, {"loss": 0.7504, "grad_norm": 0.6766283512115479, "learning_rate": 0.0002, "epoch": 1.314364238945922, "step": 8130}, {"loss": 0.7489, "grad_norm": 0.6283926367759705, "learning_rate": 0.0002, "epoch": 1.3159809231266673, "step": 8140}, {"loss": 0.7459, "grad_norm": 0.644216001033783, "learning_rate": 0.0002, "epoch": 1.3175976073074125, "step": 8150}, {"loss": 0.7125, "grad_norm": 0.7827503085136414, "learning_rate": 0.0002, "epoch": 1.3192142914881577, "step": 8160}, {"loss": 0.7271, "grad_norm": 0.6651390790939331, "learning_rate": 0.0002, "epoch": 1.320830975668903, "step": 8170}, {"loss": 0.7778, "grad_norm": 0.5547412633895874, "learning_rate": 0.0002, "epoch": 1.3224476598496484, "step": 8180}, {"loss": 0.7402, "grad_norm": 0.6765179634094238, "learning_rate": 0.0002, "epoch": 1.3240643440303936, "step": 8190}, {"loss": 0.7106, "grad_norm": 0.6822077035903931, "learning_rate": 0.0002, "epoch": 1.325681028211139, "step": 8200}, {"loss": 0.7288, "grad_norm": 0.5941002368927002, "learning_rate": 0.0002, "epoch": 1.3272977123918843, "step": 8210}, {"loss": 0.7494, "grad_norm": 0.4850037097930908, "learning_rate": 0.0002, "epoch": 1.3289143965726296, "step": 8220}, {"loss": 0.7474, "grad_norm": 0.6162990927696228, "learning_rate": 0.0002, "epoch": 1.3305310807533748, "step": 8230}, {"loss": 0.7751, "grad_norm": 0.6665613651275635, "learning_rate": 0.0002, "epoch": 1.33214776493412, "step": 8240}, {"loss": 0.759, "grad_norm": 0.618192732334137, "learning_rate": 0.0002, "epoch": 1.3337644491148655, "step": 8250}, {"loss": 0.7532, "grad_norm": 0.710418701171875, "learning_rate": 0.0002, "epoch": 1.3353811332956107, "step": 8260}, {"loss": 0.7306, "grad_norm": 0.5109876990318298, "learning_rate": 0.0002, "epoch": 1.336997817476356, "step": 8270}, {"loss": 0.7303, "grad_norm": 0.6791711449623108, "learning_rate": 0.0002, "epoch": 1.3386145016571014, "step": 8280}, {"loss": 0.7594, "grad_norm": 0.6836432814598083, "learning_rate": 0.0002, "epoch": 1.3402311858378466, "step": 8290}, {"loss": 0.7594, "grad_norm": 0.5579386353492737, "learning_rate": 0.0002, "epoch": 1.3418478700185918, "step": 8300}, {"loss": 0.7377, "grad_norm": 0.6713546514511108, "learning_rate": 0.0002, "epoch": 1.343464554199337, "step": 8310}, {"loss": 0.7756, "grad_norm": 0.5353720188140869, "learning_rate": 0.0002, "epoch": 1.3450812383800825, "step": 8320}, {"loss": 0.718, "grad_norm": 0.5813682675361633, "learning_rate": 0.0002, "epoch": 1.3466979225608278, "step": 8330}, {"loss": 0.7294, "grad_norm": 0.8158791661262512, "learning_rate": 0.0002, "epoch": 1.348314606741573, "step": 8340}, {"loss": 0.6992, "grad_norm": 0.6193785071372986, "learning_rate": 0.0002, "epoch": 1.3499312909223184, "step": 8350}, {"loss": 0.7654, "grad_norm": 0.6353939771652222, "learning_rate": 0.0002, "epoch": 1.3515479751030637, "step": 8360}, {"loss": 0.7519, "grad_norm": 0.6925048232078552, "learning_rate": 0.0002, "epoch": 1.353164659283809, "step": 8370}, {"loss": 0.736, "grad_norm": 0.988264799118042, "learning_rate": 0.0002, "epoch": 1.3547813434645541, "step": 8380}, {"loss": 0.7744, "grad_norm": 0.6476002931594849, "learning_rate": 0.0002, "epoch": 1.3563980276452994, "step": 8390}, {"loss": 0.776, "grad_norm": 0.7120398879051208, "learning_rate": 0.0002, "epoch": 1.3580147118260448, "step": 8400}, {"loss": 0.7368, "grad_norm": 0.9048416614532471, "learning_rate": 0.0002, "epoch": 1.35963139600679, "step": 8410}, {"loss": 0.7544, "grad_norm": 0.7000672817230225, "learning_rate": 0.0002, "epoch": 1.3612480801875353, "step": 8420}, {"loss": 0.7358, "grad_norm": 0.6015632152557373, "learning_rate": 0.0002, "epoch": 1.3628647643682807, "step": 8430}, {"loss": 0.7298, "grad_norm": 0.612516462802887, "learning_rate": 0.0002, "epoch": 1.364481448549026, "step": 8440}, {"loss": 0.7055, "grad_norm": 0.5969301462173462, "learning_rate": 0.0002, "epoch": 1.3660981327297712, "step": 8450}, {"loss": 0.7754, "grad_norm": 0.6730654239654541, "learning_rate": 0.0002, "epoch": 1.3677148169105164, "step": 8460}, {"loss": 0.7465, "grad_norm": 0.6386392116546631, "learning_rate": 0.0002, "epoch": 1.369331501091262, "step": 8470}, {"loss": 0.7433, "grad_norm": 0.739544153213501, "learning_rate": 0.0002, "epoch": 1.3709481852720071, "step": 8480}, {"loss": 0.7892, "grad_norm": 0.6462782621383667, "learning_rate": 0.0002, "epoch": 1.3725648694527524, "step": 8490}, {"loss": 0.7302, "grad_norm": 0.7346843481063843, "learning_rate": 0.0002, "epoch": 1.3741815536334978, "step": 8500}, {"loss": 0.7634, "grad_norm": 0.6884821057319641, "learning_rate": 0.0002, "epoch": 1.375798237814243, "step": 8510}, {"loss": 0.7614, "grad_norm": 0.6999333500862122, "learning_rate": 0.0002, "epoch": 1.3774149219949883, "step": 8520}, {"loss": 0.729, "grad_norm": 0.5378713011741638, "learning_rate": 0.0002, "epoch": 1.3790316061757335, "step": 8530}, {"loss": 0.6797, "grad_norm": 0.5417906641960144, "learning_rate": 0.0002, "epoch": 1.3806482903564787, "step": 8540}, {"loss": 0.7499, "grad_norm": 0.6602526307106018, "learning_rate": 0.0002, "epoch": 1.3822649745372242, "step": 8550}, {"loss": 0.7356, "grad_norm": 0.7073674201965332, "learning_rate": 0.0002, "epoch": 1.3838816587179694, "step": 8560}, {"loss": 0.75, "grad_norm": 0.5841707587242126, "learning_rate": 0.0002, "epoch": 1.3854983428987149, "step": 8570}, {"loss": 0.732, "grad_norm": 0.7031095027923584, "learning_rate": 0.0002, "epoch": 1.38711502707946, "step": 8580}, {"loss": 0.7464, "grad_norm": 0.5198570489883423, "learning_rate": 0.0002, "epoch": 1.3887317112602053, "step": 8590}, {"loss": 0.7354, "grad_norm": 0.7261320352554321, "learning_rate": 0.0002, "epoch": 1.3903483954409506, "step": 8600}, {"loss": 0.7339, "grad_norm": 0.5616350173950195, "learning_rate": 0.0002, "epoch": 1.3919650796216958, "step": 8610}, {"loss": 0.7382, "grad_norm": 0.5185914635658264, "learning_rate": 0.0002, "epoch": 1.3935817638024413, "step": 8620}, {"loss": 0.7456, "grad_norm": 0.5814694762229919, "learning_rate": 0.0002, "epoch": 1.3951984479831865, "step": 8630}, {"loss": 0.7413, "grad_norm": 0.6977371573448181, "learning_rate": 0.0002, "epoch": 1.3968151321639317, "step": 8640}, {"loss": 0.7574, "grad_norm": 0.6855689883232117, "learning_rate": 0.0002, "epoch": 1.3984318163446772, "step": 8650}, {"loss": 0.7802, "grad_norm": 0.5414357781410217, "learning_rate": 0.0002, "epoch": 1.4000485005254224, "step": 8660}, {"loss": 0.7487, "grad_norm": 0.6970012784004211, "learning_rate": 0.0002, "epoch": 1.4016651847061676, "step": 8670}, {"loss": 0.7421, "grad_norm": 0.526079535484314, "learning_rate": 0.0002, "epoch": 1.4032818688869129, "step": 8680}, {"loss": 0.737, "grad_norm": 0.758712887763977, "learning_rate": 0.0002, "epoch": 1.404898553067658, "step": 8690}, {"loss": 0.7612, "grad_norm": 0.7118762731552124, "learning_rate": 0.0002, "epoch": 1.4065152372484035, "step": 8700}, {"loss": 0.7628, "grad_norm": 0.5696909427642822, "learning_rate": 0.0002, "epoch": 1.4081319214291488, "step": 8710}, {"loss": 0.7156, "grad_norm": 0.7995436787605286, "learning_rate": 0.0002, "epoch": 1.4097486056098942, "step": 8720}, {"loss": 0.7521, "grad_norm": 0.7237521409988403, "learning_rate": 0.0002, "epoch": 1.4113652897906395, "step": 8730}, {"loss": 0.7661, "grad_norm": 0.744628369808197, "learning_rate": 0.0002, "epoch": 1.4129819739713847, "step": 8740}, {"loss": 0.7073, "grad_norm": 0.6082926988601685, "learning_rate": 0.0002, "epoch": 1.41459865815213, "step": 8750}, {"loss": 0.7282, "grad_norm": 0.5185243487358093, "learning_rate": 0.0002, "epoch": 1.4162153423328752, "step": 8760}, {"loss": 0.7592, "grad_norm": 0.5183082222938538, "learning_rate": 0.0002, "epoch": 1.4178320265136206, "step": 8770}, {"loss": 0.7509, "grad_norm": 0.7326041460037231, "learning_rate": 0.0002, "epoch": 1.4194487106943658, "step": 8780}, {"loss": 0.7398, "grad_norm": 0.7174660563468933, "learning_rate": 0.0002, "epoch": 1.421065394875111, "step": 8790}, {"loss": 0.7507, "grad_norm": 0.8080165982246399, "learning_rate": 0.0002, "epoch": 1.4226820790558565, "step": 8800}, {"loss": 0.72, "grad_norm": 0.5061507821083069, "learning_rate": 0.0002, "epoch": 1.4242987632366018, "step": 8810}, {"loss": 0.7563, "grad_norm": 0.801602840423584, "learning_rate": 0.0002, "epoch": 1.425915447417347, "step": 8820}, {"loss": 0.7287, "grad_norm": 0.6150273084640503, "learning_rate": 0.0002, "epoch": 1.4275321315980922, "step": 8830}, {"loss": 0.7452, "grad_norm": 0.8786525726318359, "learning_rate": 0.0002, "epoch": 1.4291488157788377, "step": 8840}, {"loss": 0.7257, "grad_norm": 0.6371538639068604, "learning_rate": 0.0002, "epoch": 1.430765499959583, "step": 8850}, {"loss": 0.711, "grad_norm": 0.6409295797348022, "learning_rate": 0.0002, "epoch": 1.4323821841403281, "step": 8860}, {"loss": 0.7891, "grad_norm": 0.6452359557151794, "learning_rate": 0.0002, "epoch": 1.4339988683210736, "step": 8870}, {"loss": 0.7588, "grad_norm": 0.5842334628105164, "learning_rate": 0.0002, "epoch": 1.4356155525018188, "step": 8880}, {"loss": 0.7446, "grad_norm": 0.696761965751648, "learning_rate": 0.0002, "epoch": 1.437232236682564, "step": 8890}, {"loss": 0.7541, "grad_norm": 0.6384600400924683, "learning_rate": 0.0002, "epoch": 1.4388489208633093, "step": 8900}, {"loss": 0.7049, "grad_norm": 0.5981136560440063, "learning_rate": 0.0002, "epoch": 1.4404656050440545, "step": 8910}, {"loss": 0.795, "grad_norm": 0.6355637907981873, "learning_rate": 0.0002, "epoch": 1.4420822892248, "step": 8920}, {"loss": 0.7653, "grad_norm": 0.6374830603599548, "learning_rate": 0.0002, "epoch": 1.4436989734055452, "step": 8930}, {"loss": 0.8108, "grad_norm": 0.559013307094574, "learning_rate": 0.0002, "epoch": 1.4453156575862904, "step": 8940}, {"loss": 0.7045, "grad_norm": 0.7289170026779175, "learning_rate": 0.0002, "epoch": 1.446932341767036, "step": 8950}, {"loss": 0.7484, "grad_norm": 0.8649206757545471, "learning_rate": 0.0002, "epoch": 1.4485490259477811, "step": 8960}, {"loss": 0.7745, "grad_norm": 0.7664689421653748, "learning_rate": 0.0002, "epoch": 1.4501657101285264, "step": 8970}, {"loss": 0.7431, "grad_norm": 0.7109952569007874, "learning_rate": 0.0002, "epoch": 1.4517823943092716, "step": 8980}, {"loss": 0.7997, "grad_norm": 0.6312844753265381, "learning_rate": 0.0002, "epoch": 1.453399078490017, "step": 8990}, {"loss": 0.7467, "grad_norm": 0.6616617441177368, "learning_rate": 0.0002, "epoch": 1.4550157626707623, "step": 9000}, {"loss": 0.7518, "grad_norm": 0.7384068965911865, "learning_rate": 0.0002, "epoch": 1.4566324468515075, "step": 9010}, {"loss": 0.7483, "grad_norm": 0.6549670100212097, "learning_rate": 0.0002, "epoch": 1.458249131032253, "step": 9020}, {"loss": 0.7423, "grad_norm": 0.6254119277000427, "learning_rate": 0.0002, "epoch": 1.4598658152129982, "step": 9030}, {"loss": 0.7645, "grad_norm": 0.6806328892707825, "learning_rate": 0.0002, "epoch": 1.4614824993937434, "step": 9040}, {"loss": 0.7221, "grad_norm": 0.6803115010261536, "learning_rate": 0.0002, "epoch": 1.4630991835744886, "step": 9050}, {"loss": 0.7264, "grad_norm": 0.48529282212257385, "learning_rate": 0.0002, "epoch": 1.4647158677552339, "step": 9060}, {"loss": 0.7542, "grad_norm": 0.5995030999183655, "learning_rate": 0.0002, "epoch": 1.4663325519359793, "step": 9070}, {"loss": 0.7894, "grad_norm": 0.6005427837371826, "learning_rate": 0.0002, "epoch": 1.4679492361167246, "step": 9080}, {"loss": 0.7288, "grad_norm": 0.718564510345459, "learning_rate": 0.0002, "epoch": 1.46956592029747, "step": 9090}, {"loss": 0.7089, "grad_norm": 0.7003577351570129, "learning_rate": 0.0002, "epoch": 1.4711826044782153, "step": 9100}, {"loss": 0.8069, "grad_norm": 0.5888323783874512, "learning_rate": 0.0002, "epoch": 1.4727992886589605, "step": 9110}, {"loss": 0.7275, "grad_norm": 0.6417609453201294, "learning_rate": 0.0002, "epoch": 1.4744159728397057, "step": 9120}, {"loss": 0.7441, "grad_norm": 0.572294294834137, "learning_rate": 0.0002, "epoch": 1.476032657020451, "step": 9130}, {"loss": 0.8053, "grad_norm": 0.8200714588165283, "learning_rate": 0.0002, "epoch": 1.4776493412011964, "step": 9140}, {"loss": 0.7382, "grad_norm": 0.6343288421630859, "learning_rate": 0.0002, "epoch": 1.4792660253819416, "step": 9150}, {"loss": 0.7641, "grad_norm": 0.7017961144447327, "learning_rate": 0.0002, "epoch": 1.4808827095626869, "step": 9160}, {"loss": 0.7619, "grad_norm": 0.6202912926673889, "learning_rate": 0.0002, "epoch": 1.4824993937434323, "step": 9170}, {"loss": 0.7428, "grad_norm": 0.6677869558334351, "learning_rate": 0.0002, "epoch": 1.4841160779241775, "step": 9180}, {"loss": 0.7648, "grad_norm": 0.6052267551422119, "learning_rate": 0.0002, "epoch": 1.4857327621049228, "step": 9190}, {"loss": 0.7152, "grad_norm": 0.6638872027397156, "learning_rate": 0.0002, "epoch": 1.487349446285668, "step": 9200}, {"loss": 0.7448, "grad_norm": 0.6245523691177368, "learning_rate": 0.0002, "epoch": 1.4889661304664135, "step": 9210}, {"loss": 0.6958, "grad_norm": 0.5761767625808716, "learning_rate": 0.0002, "epoch": 1.4905828146471587, "step": 9220}, {"loss": 0.8012, "grad_norm": 0.8175981640815735, "learning_rate": 0.0002, "epoch": 1.492199498827904, "step": 9230}, {"loss": 0.683, "grad_norm": 0.9144009947776794, "learning_rate": 0.0002, "epoch": 1.4938161830086494, "step": 9240}, {"loss": 0.7623, "grad_norm": 0.5742552876472473, "learning_rate": 0.0002, "epoch": 1.4954328671893946, "step": 9250}, {"loss": 0.7418, "grad_norm": 0.534534215927124, "learning_rate": 0.0002, "epoch": 1.4970495513701398, "step": 9260}, {"loss": 0.7194, "grad_norm": 0.7836225032806396, "learning_rate": 0.0002, "epoch": 1.498666235550885, "step": 9270}, {"loss": 0.7453, "grad_norm": 0.5292993187904358, "learning_rate": 0.0002, "epoch": 1.5002829197316303, "step": 9280}, {"loss": 0.7168, "grad_norm": 0.8044071793556213, "learning_rate": 0.0002, "epoch": 1.5018996039123758, "step": 9290}, {"loss": 0.7229, "grad_norm": 0.6185805201530457, "learning_rate": 0.0002, "epoch": 1.503516288093121, "step": 9300}, {"loss": 0.684, "grad_norm": 0.6093607544898987, "learning_rate": 0.0002, "epoch": 1.5051329722738664, "step": 9310}, {"loss": 0.7973, "grad_norm": 0.5891730189323425, "learning_rate": 0.0002, "epoch": 1.5067496564546117, "step": 9320}, {"loss": 0.7474, "grad_norm": 0.6331129670143127, "learning_rate": 0.0002, "epoch": 1.508366340635357, "step": 9330}, {"loss": 0.7074, "grad_norm": 0.7690958380699158, "learning_rate": 0.0002, "epoch": 1.5099830248161021, "step": 9340}, {"loss": 0.672, "grad_norm": 0.6548877358436584, "learning_rate": 0.0002, "epoch": 1.5115997089968474, "step": 9350}, {"loss": 0.7408, "grad_norm": 0.6545143127441406, "learning_rate": 0.0002, "epoch": 1.5132163931775926, "step": 9360}, {"loss": 0.7432, "grad_norm": 0.553247332572937, "learning_rate": 0.0002, "epoch": 1.514833077358338, "step": 9370}, {"loss": 0.7265, "grad_norm": 0.8145074844360352, "learning_rate": 0.0002, "epoch": 1.5164497615390833, "step": 9380}, {"loss": 0.7379, "grad_norm": 0.7636994123458862, "learning_rate": 0.0002, "epoch": 1.5180664457198287, "step": 9390}, {"loss": 0.7413, "grad_norm": 0.6838982701301575, "learning_rate": 0.0002, "epoch": 1.519683129900574, "step": 9400}, {"loss": 0.7367, "grad_norm": 0.8599441647529602, "learning_rate": 0.0002, "epoch": 1.5212998140813192, "step": 9410}, {"loss": 0.7663, "grad_norm": 0.7020329833030701, "learning_rate": 0.0002, "epoch": 1.5229164982620644, "step": 9420}, {"loss": 0.7928, "grad_norm": 0.6964772343635559, "learning_rate": 0.0002, "epoch": 1.5245331824428097, "step": 9430}, {"loss": 0.7168, "grad_norm": 0.6916600465774536, "learning_rate": 0.0002, "epoch": 1.5261498666235551, "step": 9440}, {"loss": 0.7519, "grad_norm": 0.7282621264457703, "learning_rate": 0.0002, "epoch": 1.5277665508043003, "step": 9450}, {"loss": 0.7628, "grad_norm": 0.5363983511924744, "learning_rate": 0.0002, "epoch": 1.5293832349850458, "step": 9460}, {"loss": 0.7154, "grad_norm": 0.6184861063957214, "learning_rate": 0.0002, "epoch": 1.530999919165791, "step": 9470}, {"loss": 0.7837, "grad_norm": 0.5991285443305969, "learning_rate": 0.0002, "epoch": 1.5326166033465363, "step": 9480}, {"loss": 0.7827, "grad_norm": 0.8176587820053101, "learning_rate": 0.0002, "epoch": 1.5342332875272815, "step": 9490}, {"loss": 0.7415, "grad_norm": 0.6473721861839294, "learning_rate": 0.0002, "epoch": 1.5358499717080267, "step": 9500}, {"loss": 0.7632, "grad_norm": 0.7319952845573425, "learning_rate": 0.0002, "epoch": 1.5374666558887722, "step": 9510}, {"loss": 0.7706, "grad_norm": 0.702900230884552, "learning_rate": 0.0002, "epoch": 1.5390833400695174, "step": 9520}, {"loss": 0.7754, "grad_norm": 0.7971600294113159, "learning_rate": 0.0002, "epoch": 1.5407000242502629, "step": 9530}, {"loss": 0.7352, "grad_norm": 0.6527525186538696, "learning_rate": 0.0002, "epoch": 1.542316708431008, "step": 9540}, {"loss": 0.7425, "grad_norm": 0.5791676044464111, "learning_rate": 0.0002, "epoch": 1.5439333926117533, "step": 9550}, {"loss": 0.7585, "grad_norm": 0.5619390606880188, "learning_rate": 0.0002, "epoch": 1.5455500767924986, "step": 9560}, {"loss": 0.7894, "grad_norm": 0.5701689124107361, "learning_rate": 0.0002, "epoch": 1.5471667609732438, "step": 9570}, {"loss": 0.793, "grad_norm": 0.47549352049827576, "learning_rate": 0.0002, "epoch": 1.548783445153989, "step": 9580}, {"loss": 0.7276, "grad_norm": 0.8730611205101013, "learning_rate": 0.0002, "epoch": 1.5504001293347345, "step": 9590}, {"loss": 0.798, "grad_norm": 0.6842091083526611, "learning_rate": 0.0002, "epoch": 1.5520168135154797, "step": 9600}, {"loss": 0.7528, "grad_norm": 0.6675129532814026, "learning_rate": 0.0002, "epoch": 1.5536334976962252, "step": 9610}, {"loss": 0.7954, "grad_norm": 0.8173956274986267, "learning_rate": 0.0002, "epoch": 1.5552501818769704, "step": 9620}, {"loss": 0.7535, "grad_norm": 0.724947452545166, "learning_rate": 0.0002, "epoch": 1.5568668660577156, "step": 9630}, {"loss": 0.7738, "grad_norm": 0.6154758930206299, "learning_rate": 0.0002, "epoch": 1.5584835502384609, "step": 9640}, {"loss": 0.7568, "grad_norm": 0.6072008013725281, "learning_rate": 0.0002, "epoch": 1.560100234419206, "step": 9650}, {"loss": 0.7219, "grad_norm": 0.659010648727417, "learning_rate": 0.0002, "epoch": 1.5617169185999515, "step": 9660}, {"loss": 0.673, "grad_norm": 0.65857994556427, "learning_rate": 0.0002, "epoch": 1.5633336027806968, "step": 9670}, {"loss": 0.7156, "grad_norm": 0.5914267301559448, "learning_rate": 0.0002, "epoch": 1.5649502869614422, "step": 9680}, {"loss": 0.7414, "grad_norm": 0.6248020529747009, "learning_rate": 0.0002, "epoch": 1.5665669711421875, "step": 9690}, {"loss": 0.694, "grad_norm": 0.7147795557975769, "learning_rate": 0.0002, "epoch": 1.5681836553229327, "step": 9700}, {"loss": 0.7335, "grad_norm": 0.7076232433319092, "learning_rate": 0.0002, "epoch": 1.569800339503678, "step": 9710}, {"loss": 0.7413, "grad_norm": 0.6217400431632996, "learning_rate": 0.0002, "epoch": 1.5714170236844232, "step": 9720}, {"loss": 0.7296, "grad_norm": 0.6709911227226257, "learning_rate": 0.0002, "epoch": 1.5730337078651684, "step": 9730}, {"loss": 0.7306, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 1.5746503920459138, "step": 9740}, {"loss": 0.7242, "grad_norm": 0.6241145730018616, "learning_rate": 0.0002, "epoch": 1.576267076226659, "step": 9750}, {"loss": 0.7384, "grad_norm": 0.4960934817790985, "learning_rate": 0.0002, "epoch": 1.5778837604074045, "step": 9760}, {"loss": 0.725, "grad_norm": 0.6593309640884399, "learning_rate": 0.0002, "epoch": 1.5795004445881498, "step": 9770}, {"loss": 0.7531, "grad_norm": 0.5814042091369629, "learning_rate": 0.0002, "epoch": 1.581117128768895, "step": 9780}, {"loss": 0.7109, "grad_norm": 0.5936070680618286, "learning_rate": 0.0002, "epoch": 1.5827338129496402, "step": 9790}, {"loss": 0.7769, "grad_norm": 0.6454403400421143, "learning_rate": 0.0002, "epoch": 1.5843504971303854, "step": 9800}, {"loss": 0.7677, "grad_norm": 0.7612107992172241, "learning_rate": 0.0002, "epoch": 1.585967181311131, "step": 9810}, {"loss": 0.7649, "grad_norm": 0.6494482755661011, "learning_rate": 0.0002, "epoch": 1.5875838654918761, "step": 9820}, {"loss": 0.7569, "grad_norm": 0.7825694680213928, "learning_rate": 0.0002, "epoch": 1.5892005496726216, "step": 9830}, {"loss": 0.706, "grad_norm": 0.6757757663726807, "learning_rate": 0.0002, "epoch": 1.5908172338533668, "step": 9840}, {"loss": 0.7803, "grad_norm": 0.7105609178543091, "learning_rate": 0.0002, "epoch": 1.592433918034112, "step": 9850}, {"loss": 0.7925, "grad_norm": 0.7596991062164307, "learning_rate": 0.0002, "epoch": 1.5940506022148573, "step": 9860}, {"loss": 0.7108, "grad_norm": 0.5681525468826294, "learning_rate": 0.0002, "epoch": 1.5956672863956025, "step": 9870}, {"loss": 0.7811, "grad_norm": 0.6090980768203735, "learning_rate": 0.0002, "epoch": 1.5972839705763477, "step": 9880}, {"loss": 0.7339, "grad_norm": 0.6271613240242004, "learning_rate": 0.0002, "epoch": 1.5989006547570932, "step": 9890}, {"loss": 0.7419, "grad_norm": 0.7656369805335999, "learning_rate": 0.0002, "epoch": 1.6005173389378387, "step": 9900}, {"loss": 0.7336, "grad_norm": 0.7504446506500244, "learning_rate": 0.0002, "epoch": 1.6021340231185839, "step": 9910}, {"loss": 0.7479, "grad_norm": 0.659656286239624, "learning_rate": 0.0002, "epoch": 1.6037507072993291, "step": 9920}, {"loss": 0.7483, "grad_norm": 0.6006826162338257, "learning_rate": 0.0002, "epoch": 1.6053673914800743, "step": 9930}, {"loss": 0.732, "grad_norm": 0.7872757911682129, "learning_rate": 0.0002, "epoch": 1.6069840756608196, "step": 9940}, {"loss": 0.768, "grad_norm": 0.5545852780342102, "learning_rate": 0.0002, "epoch": 1.6086007598415648, "step": 9950}, {"loss": 0.8064, "grad_norm": 0.7429468631744385, "learning_rate": 0.0002, "epoch": 1.6102174440223103, "step": 9960}, {"loss": 0.714, "grad_norm": 0.6873556971549988, "learning_rate": 0.0002, "epoch": 1.6118341282030555, "step": 9970}, {"loss": 0.7324, "grad_norm": 0.5874287486076355, "learning_rate": 0.0002, "epoch": 1.613450812383801, "step": 9980}, {"loss": 0.7141, "grad_norm": 0.6039386987686157, "learning_rate": 0.0002, "epoch": 1.6150674965645462, "step": 9990}, {"loss": 0.6674, "grad_norm": 0.6233575940132141, "learning_rate": 0.0002, "epoch": 1.6166841807452914, "step": 10000}, {"loss": 0.7602, "grad_norm": 0.7676448225975037, "learning_rate": 0.0002, "epoch": 1.6183008649260366, "step": 10010}, {"loss": 0.7784, "grad_norm": 0.6565698385238647, "learning_rate": 0.0002, "epoch": 1.6199175491067819, "step": 10020}, {"loss": 0.7104, "grad_norm": 0.6787590384483337, "learning_rate": 0.0002, "epoch": 1.6215342332875273, "step": 10030}, {"loss": 0.7464, "grad_norm": 0.6137678027153015, "learning_rate": 0.0002, "epoch": 1.6231509174682726, "step": 10040}, {"loss": 0.7646, "grad_norm": 0.5236800312995911, "learning_rate": 0.0002, "epoch": 1.624767601649018, "step": 10050}, {"loss": 0.7437, "grad_norm": 0.7626367807388306, "learning_rate": 0.0002, "epoch": 1.6263842858297632, "step": 10060}, {"loss": 0.7273, "grad_norm": 0.5657260417938232, "learning_rate": 0.0002, "epoch": 1.6280009700105085, "step": 10070}, {"loss": 0.7354, "grad_norm": 0.4913991391658783, "learning_rate": 0.0002, "epoch": 1.6296176541912537, "step": 10080}, {"loss": 0.7596, "grad_norm": 0.7715556621551514, "learning_rate": 0.0002, "epoch": 1.631234338371999, "step": 10090}, {"loss": 0.7105, "grad_norm": 0.6509000062942505, "learning_rate": 0.0002, "epoch": 1.6328510225527442, "step": 10100}, {"loss": 0.7274, "grad_norm": 0.6215850114822388, "learning_rate": 0.0002, "epoch": 1.6344677067334896, "step": 10110}, {"loss": 0.7705, "grad_norm": 0.6956844329833984, "learning_rate": 0.0002, "epoch": 1.6360843909142349, "step": 10120}, {"loss": 0.7129, "grad_norm": 0.6111597418785095, "learning_rate": 0.0002, "epoch": 1.6377010750949803, "step": 10130}, {"loss": 0.6955, "grad_norm": 0.6518288850784302, "learning_rate": 0.0002, "epoch": 1.6393177592757255, "step": 10140}, {"loss": 0.731, "grad_norm": 0.6914522051811218, "learning_rate": 0.0002, "epoch": 1.6409344434564708, "step": 10150}, {"loss": 0.7295, "grad_norm": 0.63785719871521, "learning_rate": 0.0002, "epoch": 1.642551127637216, "step": 10160}, {"loss": 0.7355, "grad_norm": 0.6379287838935852, "learning_rate": 0.0002, "epoch": 1.6441678118179612, "step": 10170}, {"loss": 0.7359, "grad_norm": 0.6793403029441833, "learning_rate": 0.0002, "epoch": 1.6457844959987067, "step": 10180}, {"loss": 0.7402, "grad_norm": 0.6099132895469666, "learning_rate": 0.0002, "epoch": 1.647401180179452, "step": 10190}, {"loss": 0.7353, "grad_norm": 0.5869854092597961, "learning_rate": 0.0002, "epoch": 1.6490178643601974, "step": 10200}, {"loss": 0.8308, "grad_norm": 0.7716999053955078, "learning_rate": 0.0002, "epoch": 1.6506345485409426, "step": 10210}, {"loss": 0.7215, "grad_norm": 0.6854110360145569, "learning_rate": 0.0002, "epoch": 1.6522512327216878, "step": 10220}, {"loss": 0.782, "grad_norm": 0.6957170367240906, "learning_rate": 0.0002, "epoch": 1.653867916902433, "step": 10230}, {"loss": 0.7282, "grad_norm": 0.6932903528213501, "learning_rate": 0.0002, "epoch": 1.6554846010831783, "step": 10240}, {"loss": 0.7478, "grad_norm": 0.7713165283203125, "learning_rate": 0.0002, "epoch": 1.6571012852639235, "step": 10250}, {"loss": 0.7099, "grad_norm": 0.7455793619155884, "learning_rate": 0.0002, "epoch": 1.658717969444669, "step": 10260}, {"loss": 0.7524, "grad_norm": 0.5464168190956116, "learning_rate": 0.0002, "epoch": 1.6603346536254144, "step": 10270}, {"loss": 0.7328, "grad_norm": 0.6782926321029663, "learning_rate": 0.0002, "epoch": 1.6619513378061597, "step": 10280}, {"loss": 0.7801, "grad_norm": 0.7962649464607239, "learning_rate": 0.0002, "epoch": 1.663568021986905, "step": 10290}, {"loss": 0.7142, "grad_norm": 0.6814526319503784, "learning_rate": 0.0002, "epoch": 1.6651847061676501, "step": 10300}, {"loss": 0.7285, "grad_norm": 0.656895101070404, "learning_rate": 0.0002, "epoch": 1.6668013903483954, "step": 10310}, {"loss": 0.7358, "grad_norm": 0.6085672378540039, "learning_rate": 0.0002, "epoch": 1.6684180745291406, "step": 10320}, {"loss": 0.7074, "grad_norm": 0.585508406162262, "learning_rate": 0.0002, "epoch": 1.670034758709886, "step": 10330}, {"loss": 0.7604, "grad_norm": 0.6930184364318848, "learning_rate": 0.0002, "epoch": 1.6716514428906313, "step": 10340}, {"loss": 0.7169, "grad_norm": 0.575663149356842, "learning_rate": 0.0002, "epoch": 1.6732681270713767, "step": 10350}, {"loss": 0.7198, "grad_norm": 0.582502543926239, "learning_rate": 0.0002, "epoch": 1.674884811252122, "step": 10360}, {"loss": 0.7793, "grad_norm": 0.5668916702270508, "learning_rate": 0.0002, "epoch": 1.6765014954328672, "step": 10370}, {"loss": 0.7478, "grad_norm": 0.6070065498352051, "learning_rate": 0.0002, "epoch": 1.6781181796136124, "step": 10380}, {"loss": 0.7939, "grad_norm": 0.6141316294670105, "learning_rate": 0.0002, "epoch": 1.6797348637943577, "step": 10390}, {"loss": 0.7573, "grad_norm": 0.8359124064445496, "learning_rate": 0.0002, "epoch": 1.6813515479751031, "step": 10400}, {"loss": 0.7488, "grad_norm": 0.5378185510635376, "learning_rate": 0.0002, "epoch": 1.6829682321558483, "step": 10410}, {"loss": 0.7588, "grad_norm": 0.6959536075592041, "learning_rate": 0.0002, "epoch": 1.6845849163365938, "step": 10420}, {"loss": 0.7872, "grad_norm": 0.6514357328414917, "learning_rate": 0.0002, "epoch": 1.686201600517339, "step": 10430}, {"loss": 0.725, "grad_norm": 0.7706646919250488, "learning_rate": 0.0002, "epoch": 1.6878182846980843, "step": 10440}, {"loss": 0.7673, "grad_norm": 0.6183337569236755, "learning_rate": 0.0002, "epoch": 1.6894349688788295, "step": 10450}, {"loss": 0.7566, "grad_norm": 0.6123278141021729, "learning_rate": 0.0002, "epoch": 1.6910516530595747, "step": 10460}, {"loss": 0.7169, "grad_norm": 0.6894851326942444, "learning_rate": 0.0002, "epoch": 1.69266833724032, "step": 10470}, {"loss": 0.7435, "grad_norm": 0.7497312426567078, "learning_rate": 0.0002, "epoch": 1.6942850214210654, "step": 10480}, {"loss": 0.7544, "grad_norm": 0.5968214273452759, "learning_rate": 0.0002, "epoch": 1.6959017056018106, "step": 10490}, {"loss": 0.6793, "grad_norm": 0.6747927069664001, "learning_rate": 0.0002, "epoch": 1.697518389782556, "step": 10500}, {"loss": 0.7415, "grad_norm": 0.5708310008049011, "learning_rate": 0.0002, "epoch": 1.6991350739633013, "step": 10510}, {"loss": 0.7385, "grad_norm": 0.606526792049408, "learning_rate": 0.0002, "epoch": 1.7007517581440466, "step": 10520}, {"loss": 0.7204, "grad_norm": 0.662011981010437, "learning_rate": 0.0002, "epoch": 1.7023684423247918, "step": 10530}, {"loss": 0.7999, "grad_norm": 0.7583045363426208, "learning_rate": 0.0002, "epoch": 1.703985126505537, "step": 10540}, {"loss": 0.7563, "grad_norm": 0.721632182598114, "learning_rate": 0.0002, "epoch": 1.7056018106862825, "step": 10550}, {"loss": 0.7407, "grad_norm": 0.6107715368270874, "learning_rate": 0.0002, "epoch": 1.7072184948670277, "step": 10560}, {"loss": 0.7519, "grad_norm": 0.6652471423149109, "learning_rate": 0.0002, "epoch": 1.7088351790477732, "step": 10570}, {"loss": 0.7767, "grad_norm": 0.6308087110519409, "learning_rate": 0.0002, "epoch": 1.7104518632285184, "step": 10580}, {"loss": 0.7659, "grad_norm": 0.5464386940002441, "learning_rate": 0.0002, "epoch": 1.7120685474092636, "step": 10590}, {"loss": 0.7063, "grad_norm": 0.6558911204338074, "learning_rate": 0.0002, "epoch": 1.7136852315900089, "step": 10600}, {"loss": 0.7126, "grad_norm": 0.5665024518966675, "learning_rate": 0.0002, "epoch": 1.715301915770754, "step": 10610}, {"loss": 0.6958, "grad_norm": 0.7888094186782837, "learning_rate": 0.0002, "epoch": 1.7169185999514993, "step": 10620}, {"loss": 0.7785, "grad_norm": 0.7084909081459045, "learning_rate": 0.0002, "epoch": 1.7185352841322448, "step": 10630}, {"loss": 0.7557, "grad_norm": 0.7982324361801147, "learning_rate": 0.0002, "epoch": 1.7201519683129902, "step": 10640}, {"loss": 0.7345, "grad_norm": 0.6418732404708862, "learning_rate": 0.0002, "epoch": 1.7217686524937355, "step": 10650}, {"loss": 0.7734, "grad_norm": 0.7636681795120239, "learning_rate": 0.0002, "epoch": 1.7233853366744807, "step": 10660}, {"loss": 0.7541, "grad_norm": 0.5646875500679016, "learning_rate": 0.0002, "epoch": 1.725002020855226, "step": 10670}, {"loss": 0.7642, "grad_norm": 0.5231260657310486, "learning_rate": 0.0002, "epoch": 1.7266187050359711, "step": 10680}, {"loss": 0.7846, "grad_norm": 0.7635011672973633, "learning_rate": 0.0002, "epoch": 1.7282353892167164, "step": 10690}, {"loss": 0.7471, "grad_norm": 0.7518259286880493, "learning_rate": 0.0002, "epoch": 1.7298520733974618, "step": 10700}, {"loss": 0.751, "grad_norm": 0.7295602560043335, "learning_rate": 0.0002, "epoch": 1.731468757578207, "step": 10710}, {"loss": 0.731, "grad_norm": 0.6984632015228271, "learning_rate": 0.0002, "epoch": 1.7330854417589525, "step": 10720}, {"loss": 0.7921, "grad_norm": 0.6198219060897827, "learning_rate": 0.0002, "epoch": 1.7347021259396977, "step": 10730}, {"loss": 0.7642, "grad_norm": 0.6957576274871826, "learning_rate": 0.0002, "epoch": 1.736318810120443, "step": 10740}, {"loss": 0.7917, "grad_norm": 0.6430263519287109, "learning_rate": 0.0002, "epoch": 1.7379354943011882, "step": 10750}, {"loss": 0.7156, "grad_norm": 0.6134995222091675, "learning_rate": 0.0002, "epoch": 1.7395521784819334, "step": 10760}, {"loss": 0.7584, "grad_norm": 0.7209452986717224, "learning_rate": 0.0002, "epoch": 1.741168862662679, "step": 10770}, {"loss": 0.7528, "grad_norm": 0.6735447645187378, "learning_rate": 0.0002, "epoch": 1.7427855468434241, "step": 10780}, {"loss": 0.756, "grad_norm": 0.5605693459510803, "learning_rate": 0.0002, "epoch": 1.7444022310241696, "step": 10790}, {"loss": 0.7759, "grad_norm": 0.6882363557815552, "learning_rate": 0.0002, "epoch": 1.7460189152049148, "step": 10800}, {"loss": 0.7544, "grad_norm": 0.6386259198188782, "learning_rate": 0.0002, "epoch": 1.74763559938566, "step": 10810}, {"loss": 0.7697, "grad_norm": 0.6529015302658081, "learning_rate": 0.0002, "epoch": 1.7492522835664053, "step": 10820}, {"loss": 0.7219, "grad_norm": 0.5664082765579224, "learning_rate": 0.0002, "epoch": 1.7508689677471505, "step": 10830}, {"loss": 0.7586, "grad_norm": 0.7532684206962585, "learning_rate": 0.0002, "epoch": 1.7524856519278957, "step": 10840}, {"loss": 0.6919, "grad_norm": 0.77171391248703, "learning_rate": 0.0002, "epoch": 1.7541023361086412, "step": 10850}, {"loss": 0.785, "grad_norm": 0.7255431413650513, "learning_rate": 0.0002, "epoch": 1.7557190202893864, "step": 10860}, {"loss": 0.7458, "grad_norm": 0.763083279132843, "learning_rate": 0.0002, "epoch": 1.7573357044701319, "step": 10870}, {"loss": 0.7846, "grad_norm": 0.6042402982711792, "learning_rate": 0.0002, "epoch": 1.758952388650877, "step": 10880}, {"loss": 0.7027, "grad_norm": 0.7642518281936646, "learning_rate": 0.0002, "epoch": 1.7605690728316223, "step": 10890}, {"loss": 0.746, "grad_norm": 0.6347904801368713, "learning_rate": 0.0002, "epoch": 1.7621857570123676, "step": 10900}, {"loss": 0.7458, "grad_norm": 0.5371627807617188, "learning_rate": 0.0002, "epoch": 1.7638024411931128, "step": 10910}, {"loss": 0.7466, "grad_norm": 0.6840225458145142, "learning_rate": 0.0002, "epoch": 1.7654191253738583, "step": 10920}, {"loss": 0.725, "grad_norm": 0.5288469195365906, "learning_rate": 0.0002, "epoch": 1.7670358095546035, "step": 10930}, {"loss": 0.7863, "grad_norm": 0.69020676612854, "learning_rate": 0.0002, "epoch": 1.768652493735349, "step": 10940}, {"loss": 0.7468, "grad_norm": 0.5943242311477661, "learning_rate": 0.0002, "epoch": 1.7702691779160942, "step": 10950}, {"loss": 0.7244, "grad_norm": 0.5616418123245239, "learning_rate": 0.0002, "epoch": 1.7718858620968394, "step": 10960}, {"loss": 0.7137, "grad_norm": 0.7209470868110657, "learning_rate": 0.0002, "epoch": 1.7735025462775846, "step": 10970}, {"loss": 0.7459, "grad_norm": 0.6657957434654236, "learning_rate": 0.0002, "epoch": 1.7751192304583299, "step": 10980}, {"loss": 0.7076, "grad_norm": 0.6469064950942993, "learning_rate": 0.0002, "epoch": 1.776735914639075, "step": 10990}, {"loss": 0.7321, "grad_norm": 0.6615678071975708, "learning_rate": 0.0002, "epoch": 1.7783525988198206, "step": 11000}, {"loss": 0.747, "grad_norm": 0.6722439527511597, "learning_rate": 0.0002, "epoch": 1.779969283000566, "step": 11010}, {"loss": 0.7302, "grad_norm": 0.634136974811554, "learning_rate": 0.0002, "epoch": 1.7815859671813112, "step": 11020}, {"loss": 0.8105, "grad_norm": 0.6024377346038818, "learning_rate": 0.0002, "epoch": 1.7832026513620565, "step": 11030}, {"loss": 0.7855, "grad_norm": 0.6909403800964355, "learning_rate": 0.0002, "epoch": 1.7848193355428017, "step": 11040}, {"loss": 0.7471, "grad_norm": 0.7148767709732056, "learning_rate": 0.0002, "epoch": 1.786436019723547, "step": 11050}, {"loss": 0.7145, "grad_norm": 0.7442979216575623, "learning_rate": 0.0002, "epoch": 1.7880527039042922, "step": 11060}, {"loss": 0.7215, "grad_norm": 0.6830431818962097, "learning_rate": 0.0002, "epoch": 1.7896693880850376, "step": 11070}, {"loss": 0.7625, "grad_norm": 0.9172667264938354, "learning_rate": 0.0002, "epoch": 1.7912860722657828, "step": 11080}, {"loss": 0.76, "grad_norm": 0.6799490451812744, "learning_rate": 0.0002, "epoch": 1.7929027564465283, "step": 11090}, {"loss": 0.7716, "grad_norm": 0.7617024779319763, "learning_rate": 0.0002, "epoch": 1.7945194406272735, "step": 11100}, {"loss": 0.7586, "grad_norm": 0.7701810002326965, "learning_rate": 0.0002, "epoch": 1.7961361248080188, "step": 11110}, {"loss": 0.7843, "grad_norm": 0.7454385757446289, "learning_rate": 0.0002, "epoch": 1.797752808988764, "step": 11120}, {"loss": 0.7873, "grad_norm": 0.6121436953544617, "learning_rate": 0.0002, "epoch": 1.7993694931695092, "step": 11130}, {"loss": 0.7305, "grad_norm": 0.6237571835517883, "learning_rate": 0.0002, "epoch": 1.8009861773502547, "step": 11140}, {"loss": 0.6827, "grad_norm": 0.6818515658378601, "learning_rate": 0.0002, "epoch": 1.802602861531, "step": 11150}, {"loss": 0.6876, "grad_norm": 0.7768308520317078, "learning_rate": 0.0002, "epoch": 1.8042195457117454, "step": 11160}, {"loss": 0.7533, "grad_norm": 0.6875537633895874, "learning_rate": 0.0002, "epoch": 1.8058362298924906, "step": 11170}, {"loss": 0.761, "grad_norm": 0.7950584888458252, "learning_rate": 0.0002, "epoch": 1.8074529140732358, "step": 11180}, {"loss": 0.7623, "grad_norm": 0.8210248351097107, "learning_rate": 0.0002, "epoch": 1.809069598253981, "step": 11190}, {"loss": 0.7556, "grad_norm": 0.6674110889434814, "learning_rate": 0.0002, "epoch": 1.8106862824347263, "step": 11200}, {"loss": 0.7663, "grad_norm": 0.6261674761772156, "learning_rate": 0.0002, "epoch": 1.8123029666154715, "step": 11210}, {"loss": 0.7122, "grad_norm": 0.6484741568565369, "learning_rate": 0.0002, "epoch": 1.813919650796217, "step": 11220}, {"loss": 0.7718, "grad_norm": 0.6231244206428528, "learning_rate": 0.0002, "epoch": 1.8155363349769622, "step": 11230}, {"loss": 0.7152, "grad_norm": 0.7243146896362305, "learning_rate": 0.0002, "epoch": 1.8171530191577077, "step": 11240}, {"loss": 0.7448, "grad_norm": 0.6776193380355835, "learning_rate": 0.0002, "epoch": 1.818769703338453, "step": 11250}, {"loss": 0.7317, "grad_norm": 0.5973618030548096, "learning_rate": 0.0002, "epoch": 1.8203863875191981, "step": 11260}, {"loss": 0.7961, "grad_norm": 0.6451361179351807, "learning_rate": 0.0002, "epoch": 1.8220030716999434, "step": 11270}, {"loss": 0.7611, "grad_norm": 0.5963068008422852, "learning_rate": 0.0002, "epoch": 1.8236197558806886, "step": 11280}, {"loss": 0.7466, "grad_norm": 0.536902129650116, "learning_rate": 0.0002, "epoch": 1.825236440061434, "step": 11290}, {"loss": 0.708, "grad_norm": 0.6993787288665771, "learning_rate": 0.0002, "epoch": 1.8268531242421793, "step": 11300}, {"loss": 0.7153, "grad_norm": 0.6135255098342896, "learning_rate": 0.0002, "epoch": 1.8284698084229247, "step": 11310}, {"loss": 0.7423, "grad_norm": 0.6057423949241638, "learning_rate": 0.0002, "epoch": 1.83008649260367, "step": 11320}, {"loss": 0.735, "grad_norm": 0.6598812341690063, "learning_rate": 0.0002, "epoch": 1.8317031767844152, "step": 11330}, {"loss": 0.7278, "grad_norm": 0.6075948476791382, "learning_rate": 0.0002, "epoch": 1.8333198609651604, "step": 11340}, {"loss": 0.7846, "grad_norm": 0.7065447568893433, "learning_rate": 0.0002, "epoch": 1.8349365451459057, "step": 11350}, {"loss": 0.7365, "grad_norm": 0.680526614189148, "learning_rate": 0.0002, "epoch": 1.8365532293266509, "step": 11360}, {"loss": 0.7152, "grad_norm": 0.6356695294380188, "learning_rate": 0.0002, "epoch": 1.8381699135073963, "step": 11370}, {"loss": 0.721, "grad_norm": 0.6399052143096924, "learning_rate": 0.0002, "epoch": 1.8397865976881416, "step": 11380}, {"loss": 0.7618, "grad_norm": 0.6125704050064087, "learning_rate": 0.0002, "epoch": 1.841403281868887, "step": 11390}, {"loss": 0.755, "grad_norm": 0.7124643325805664, "learning_rate": 0.0002, "epoch": 1.8430199660496323, "step": 11400}, {"loss": 0.7972, "grad_norm": 0.6099604964256287, "learning_rate": 0.0002, "epoch": 1.8446366502303775, "step": 11410}, {"loss": 0.7187, "grad_norm": 0.7338208556175232, "learning_rate": 0.0002, "epoch": 1.8462533344111227, "step": 11420}, {"loss": 0.7007, "grad_norm": 0.7534668445587158, "learning_rate": 0.0002, "epoch": 1.847870018591868, "step": 11430}, {"loss": 0.7464, "grad_norm": 0.6135470271110535, "learning_rate": 0.0002, "epoch": 1.8494867027726134, "step": 11440}, {"loss": 0.7955, "grad_norm": 0.6229309439659119, "learning_rate": 0.0002, "epoch": 1.8511033869533586, "step": 11450}, {"loss": 0.7594, "grad_norm": 0.706423282623291, "learning_rate": 0.0002, "epoch": 1.852720071134104, "step": 11460}, {"loss": 0.7411, "grad_norm": 0.5460049510002136, "learning_rate": 0.0002, "epoch": 1.8543367553148493, "step": 11470}, {"loss": 0.7416, "grad_norm": 0.6616711020469666, "learning_rate": 0.0002, "epoch": 1.8559534394955945, "step": 11480}, {"loss": 0.729, "grad_norm": 0.6372783184051514, "learning_rate": 0.0002, "epoch": 1.8575701236763398, "step": 11490}, {"loss": 0.7333, "grad_norm": 0.7162668108940125, "learning_rate": 0.0002, "epoch": 1.859186807857085, "step": 11500}, {"loss": 0.7747, "grad_norm": 0.6605209708213806, "learning_rate": 0.0002, "epoch": 1.8608034920378305, "step": 11510}, {"loss": 0.7258, "grad_norm": 0.6933956742286682, "learning_rate": 0.0002, "epoch": 1.8624201762185757, "step": 11520}, {"loss": 0.7243, "grad_norm": 0.6582090854644775, "learning_rate": 0.0002, "epoch": 1.8640368603993211, "step": 11530}, {"loss": 0.7313, "grad_norm": 0.6416500806808472, "learning_rate": 0.0002, "epoch": 1.8656535445800664, "step": 11540}, {"loss": 0.7372, "grad_norm": 0.5434312224388123, "learning_rate": 0.0002, "epoch": 1.8672702287608116, "step": 11550}, {"loss": 0.7635, "grad_norm": 0.6827567219734192, "learning_rate": 0.0002, "epoch": 1.8688869129415568, "step": 11560}, {"loss": 0.7137, "grad_norm": 0.7354370951652527, "learning_rate": 0.0002, "epoch": 1.870503597122302, "step": 11570}, {"loss": 0.7526, "grad_norm": 0.590372622013092, "learning_rate": 0.0002, "epoch": 1.8721202813030473, "step": 11580}, {"loss": 0.731, "grad_norm": 0.853183925151825, "learning_rate": 0.0002, "epoch": 1.8737369654837928, "step": 11590}, {"loss": 0.7487, "grad_norm": 0.822678804397583, "learning_rate": 0.0002, "epoch": 1.875353649664538, "step": 11600}, {"loss": 0.7427, "grad_norm": 0.6591550707817078, "learning_rate": 0.0002, "epoch": 1.8769703338452834, "step": 11610}, {"loss": 0.7054, "grad_norm": 0.7475301623344421, "learning_rate": 0.0002, "epoch": 1.8785870180260287, "step": 11620}, {"loss": 0.811, "grad_norm": 0.6390765309333801, "learning_rate": 0.0002, "epoch": 1.880203702206774, "step": 11630}, {"loss": 0.7531, "grad_norm": 0.6589758992195129, "learning_rate": 0.0002, "epoch": 1.8818203863875191, "step": 11640}, {"loss": 0.7475, "grad_norm": 0.6765508651733398, "learning_rate": 0.0002, "epoch": 1.8834370705682644, "step": 11650}, {"loss": 0.738, "grad_norm": 0.6527857780456543, "learning_rate": 0.0002, "epoch": 1.8850537547490098, "step": 11660}, {"loss": 0.7504, "grad_norm": 0.6642923951148987, "learning_rate": 0.0002, "epoch": 1.886670438929755, "step": 11670}, {"loss": 0.7701, "grad_norm": 0.6945584416389465, "learning_rate": 0.0002, "epoch": 1.8882871231105005, "step": 11680}, {"loss": 0.7711, "grad_norm": 0.694018542766571, "learning_rate": 0.0002, "epoch": 1.8899038072912457, "step": 11690}, {"loss": 0.7195, "grad_norm": 0.7237417101860046, "learning_rate": 0.0002, "epoch": 1.891520491471991, "step": 11700}, {"loss": 0.7491, "grad_norm": 0.7401309609413147, "learning_rate": 0.0002, "epoch": 1.8931371756527362, "step": 11710}, {"loss": 0.805, "grad_norm": 0.6537784337997437, "learning_rate": 0.0002, "epoch": 1.8947538598334814, "step": 11720}, {"loss": 0.793, "grad_norm": 0.7398539185523987, "learning_rate": 0.0002, "epoch": 1.8963705440142267, "step": 11730}, {"loss": 0.7561, "grad_norm": 0.6696075797080994, "learning_rate": 0.0002, "epoch": 1.8979872281949721, "step": 11740}, {"loss": 0.7353, "grad_norm": 0.6014142036437988, "learning_rate": 0.0002, "epoch": 1.8996039123757174, "step": 11750}, {"loss": 0.7714, "grad_norm": 0.7023524641990662, "learning_rate": 0.0002, "epoch": 1.9012205965564628, "step": 11760}, {"loss": 0.7088, "grad_norm": 0.739973783493042, "learning_rate": 0.0002, "epoch": 1.902837280737208, "step": 11770}, {"loss": 0.7848, "grad_norm": 0.5576770901679993, "learning_rate": 0.0002, "epoch": 1.9044539649179533, "step": 11780}, {"loss": 0.7483, "grad_norm": 0.6907393932342529, "learning_rate": 0.0002, "epoch": 1.9060706490986985, "step": 11790}, {"loss": 0.7827, "grad_norm": 0.6934581995010376, "learning_rate": 0.0002, "epoch": 1.9076873332794437, "step": 11800}, {"loss": 0.7199, "grad_norm": 0.591774582862854, "learning_rate": 0.0002, "epoch": 1.9093040174601892, "step": 11810}, {"loss": 0.7333, "grad_norm": 0.6249791383743286, "learning_rate": 0.0002, "epoch": 1.9109207016409344, "step": 11820}, {"loss": 0.7581, "grad_norm": 0.6755744218826294, "learning_rate": 0.0002, "epoch": 1.9125373858216799, "step": 11830}, {"loss": 0.696, "grad_norm": 0.7286285161972046, "learning_rate": 0.0002, "epoch": 1.914154070002425, "step": 11840}, {"loss": 0.7509, "grad_norm": 0.7867850065231323, "learning_rate": 0.0002, "epoch": 1.9157707541831703, "step": 11850}, {"loss": 0.735, "grad_norm": 0.6283972859382629, "learning_rate": 0.0002, "epoch": 1.9173874383639156, "step": 11860}, {"loss": 0.7296, "grad_norm": 0.605823814868927, "learning_rate": 0.0002, "epoch": 1.9190041225446608, "step": 11870}, {"loss": 0.6598, "grad_norm": 0.5927976965904236, "learning_rate": 0.0002, "epoch": 1.920620806725406, "step": 11880}, {"loss": 0.7649, "grad_norm": 0.5974002480506897, "learning_rate": 0.0002, "epoch": 1.9222374909061515, "step": 11890}, {"loss": 0.7843, "grad_norm": 0.7091866135597229, "learning_rate": 0.0002, "epoch": 1.923854175086897, "step": 11900}, {"loss": 0.775, "grad_norm": 0.72496497631073, "learning_rate": 0.0002, "epoch": 1.9254708592676422, "step": 11910}, {"loss": 0.7153, "grad_norm": 0.6131896376609802, "learning_rate": 0.0002, "epoch": 1.9270875434483874, "step": 11920}, {"loss": 0.7228, "grad_norm": 0.6556436419487, "learning_rate": 0.0002, "epoch": 1.9287042276291326, "step": 11930}, {"loss": 0.7319, "grad_norm": 0.622932493686676, "learning_rate": 0.0002, "epoch": 1.9303209118098779, "step": 11940}, {"loss": 0.7592, "grad_norm": 0.6618631482124329, "learning_rate": 0.0002, "epoch": 1.931937595990623, "step": 11950}, {"loss": 0.8332, "grad_norm": 0.630966305732727, "learning_rate": 0.0002, "epoch": 1.9335542801713685, "step": 11960}, {"loss": 0.6854, "grad_norm": 0.6336734890937805, "learning_rate": 0.0002, "epoch": 1.9351709643521138, "step": 11970}, {"loss": 0.7433, "grad_norm": 0.655403196811676, "learning_rate": 0.0002, "epoch": 1.9367876485328592, "step": 11980}, {"loss": 0.7282, "grad_norm": 0.5640574097633362, "learning_rate": 0.0002, "epoch": 1.9384043327136045, "step": 11990}, {"loss": 0.7289, "grad_norm": 0.6322951316833496, "learning_rate": 0.0002, "epoch": 1.9400210168943497, "step": 12000}, {"loss": 0.7627, "grad_norm": 0.615703821182251, "learning_rate": 0.0002, "epoch": 1.941637701075095, "step": 12010}, {"loss": 0.786, "grad_norm": 0.6487536430358887, "learning_rate": 0.0002, "epoch": 1.9432543852558402, "step": 12020}, {"loss": 0.7435, "grad_norm": 0.9209630489349365, "learning_rate": 0.0002, "epoch": 1.9448710694365856, "step": 12030}, {"loss": 0.7274, "grad_norm": 0.67485511302948, "learning_rate": 0.0002, "epoch": 1.9464877536173308, "step": 12040}, {"loss": 0.7551, "grad_norm": 0.6831230521202087, "learning_rate": 0.0002, "epoch": 1.9481044377980763, "step": 12050}, {"loss": 0.7546, "grad_norm": 0.6578302383422852, "learning_rate": 0.0002, "epoch": 1.9497211219788215, "step": 12060}, {"loss": 0.6989, "grad_norm": 0.9975938200950623, "learning_rate": 0.0002, "epoch": 1.9513378061595668, "step": 12070}, {"loss": 0.7952, "grad_norm": 0.6637365221977234, "learning_rate": 0.0002, "epoch": 1.952954490340312, "step": 12080}, {"loss": 0.7482, "grad_norm": 0.605707049369812, "learning_rate": 0.0002, "epoch": 1.9545711745210572, "step": 12090}, {"loss": 0.7768, "grad_norm": 0.6584440469741821, "learning_rate": 0.0002, "epoch": 1.9561878587018025, "step": 12100}, {"loss": 0.7187, "grad_norm": 0.6070835590362549, "learning_rate": 0.0002, "epoch": 1.957804542882548, "step": 12110}, {"loss": 0.7491, "grad_norm": 0.7862601280212402, "learning_rate": 0.0002, "epoch": 1.9594212270632931, "step": 12120}, {"loss": 0.7972, "grad_norm": 0.8175255060195923, "learning_rate": 0.0002, "epoch": 1.9610379112440386, "step": 12130}, {"loss": 0.7242, "grad_norm": 0.5648472905158997, "learning_rate": 0.0002, "epoch": 1.9626545954247838, "step": 12140}, {"loss": 0.7321, "grad_norm": 0.6591973304748535, "learning_rate": 0.0002, "epoch": 1.964271279605529, "step": 12150}, {"loss": 0.739, "grad_norm": 0.5960676074028015, "learning_rate": 0.0002, "epoch": 1.9658879637862743, "step": 12160}, {"loss": 0.7254, "grad_norm": 0.7272544503211975, "learning_rate": 0.0002, "epoch": 1.9675046479670195, "step": 12170}, {"loss": 0.7376, "grad_norm": 0.7176699042320251, "learning_rate": 0.0002, "epoch": 1.969121332147765, "step": 12180}, {"loss": 0.7525, "grad_norm": 0.6927123665809631, "learning_rate": 0.0002, "epoch": 1.9707380163285102, "step": 12190}, {"loss": 0.7318, "grad_norm": 0.5536034107208252, "learning_rate": 0.0002, "epoch": 1.9723547005092557, "step": 12200}, {"loss": 0.7737, "grad_norm": 0.8348390460014343, "learning_rate": 0.0002, "epoch": 1.9739713846900009, "step": 12210}, {"loss": 0.7494, "grad_norm": 0.6591181755065918, "learning_rate": 0.0002, "epoch": 1.9755880688707461, "step": 12220}, {"loss": 0.763, "grad_norm": 1.0624109506607056, "learning_rate": 0.0002, "epoch": 1.9772047530514913, "step": 12230}, {"loss": 0.7541, "grad_norm": 0.9265586137771606, "learning_rate": 0.0002, "epoch": 1.9788214372322366, "step": 12240}, {"loss": 0.7533, "grad_norm": 0.5998196005821228, "learning_rate": 0.0002, "epoch": 1.9804381214129818, "step": 12250}, {"loss": 0.7225, "grad_norm": 0.6960851550102234, "learning_rate": 0.0002, "epoch": 1.9820548055937273, "step": 12260}, {"loss": 0.7398, "grad_norm": 0.7674502730369568, "learning_rate": 0.0002, "epoch": 1.9836714897744727, "step": 12270}, {"loss": 0.7185, "grad_norm": 0.6407275795936584, "learning_rate": 0.0002, "epoch": 1.985288173955218, "step": 12280}, {"loss": 0.7382, "grad_norm": 0.6673079133033752, "learning_rate": 0.0002, "epoch": 1.9869048581359632, "step": 12290}, {"loss": 0.7326, "grad_norm": 0.6989844441413879, "learning_rate": 0.0002, "epoch": 1.9885215423167084, "step": 12300}, {"loss": 0.7559, "grad_norm": 0.7564442157745361, "learning_rate": 0.0002, "epoch": 1.9901382264974536, "step": 12310}, {"loss": 0.7719, "grad_norm": 0.6385478973388672, "learning_rate": 0.0002, "epoch": 1.9917549106781989, "step": 12320}, {"loss": 0.7369, "grad_norm": 0.7193717956542969, "learning_rate": 0.0002, "epoch": 1.9933715948589443, "step": 12330}, {"loss": 0.7583, "grad_norm": 0.7987112402915955, "learning_rate": 0.0002, "epoch": 1.9949882790396896, "step": 12340}, {"loss": 0.7793, "grad_norm": 0.7260826826095581, "learning_rate": 0.0002, "epoch": 1.996604963220435, "step": 12350}, {"loss": 0.7505, "grad_norm": 0.7968255281448364, "learning_rate": 0.0002, "epoch": 1.9982216474011802, "step": 12360}, {"loss": 0.717, "grad_norm": 0.6893062591552734, "learning_rate": 0.0002, "epoch": 1.9998383315819255, "step": 12370}]}
+{"epoch": 2.9999191657909625, "step": 18556, "epoch_duration": 16900.518332481384, "total_accumulated_duration": 50651.44822502136, "gpu_info": {"GPU_0": "NVIDIA A100-PCIE-40GB"}, "memory_usage": {"avg_memory_usage": {"GPU_0": 7672.4541015625}, "peak_memory_usage": {"GPU_0": 13792.75537109375}, "avg_memory_reserved": {"GPU_0": 17416.0}, "peak_memory_reserved": {"GPU_0": 17416.0}, "total_memory": {"GPU_0": 40444.375}}, "best_checkpoint_path": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185", "params": {"epochs": 8, "batch_size": 1, "learning_rate": 0.0002, "gradient_accumulation_steps": 2, "warmup_ratio": 0.03, "max_grad_norm": 0.3, "lora_alpha": 32, "lora_dropout": 0.05, "lora_r": 64, "tf32": true, "seed": 42}, "log_history": [{"loss": 1.6636, "grad_norm": 0.9894065856933594, "learning_rate": 0.0002, "epoch": 0.0016166841807452913, "step": 10}, {"loss": 1.1528, "grad_norm": 1.7810699939727783, "learning_rate": 0.0002, "epoch": 0.0032333683614905826, "step": 20}, {"loss": 0.9767, "grad_norm": 0.5969577431678772, "learning_rate": 0.0002, "epoch": 0.004850052542235874, "step": 30}, {"loss": 0.9772, "grad_norm": 0.6354120969772339, "learning_rate": 0.0002, "epoch": 0.006466736722981165, "step": 40}, {"loss": 0.8643, "grad_norm": 0.5604607462882996, "learning_rate": 0.0002, "epoch": 0.008083420903726457, "step": 50}, {"loss": 0.8841, "grad_norm": 0.4676193594932556, "learning_rate": 0.0002, "epoch": 0.009700105084471748, "step": 60}, {"loss": 0.9022, "grad_norm": 0.6099211573600769, "learning_rate": 0.0002, "epoch": 0.01131678926521704, "step": 70}, {"loss": 0.9133, "grad_norm": 0.48639994859695435, "learning_rate": 0.0002, "epoch": 0.01293347344596233, "step": 80}, {"loss": 0.8704, "grad_norm": 0.4904264509677887, "learning_rate": 0.0002, "epoch": 0.014550157626707623, "step": 90}, {"loss": 0.8855, "grad_norm": 2.8334362506866455, "learning_rate": 0.0002, "epoch": 0.016166841807452915, "step": 100}, {"loss": 0.8958, "grad_norm": 0.43221670389175415, "learning_rate": 0.0002, "epoch": 0.017783525988198205, "step": 110}, {"loss": 0.8412, "grad_norm": 0.42244166135787964, "learning_rate": 0.0002, "epoch": 0.019400210168943496, "step": 120}, {"loss": 0.8467, "grad_norm": 0.45363298058509827, "learning_rate": 0.0002, "epoch": 0.02101689434968879, "step": 130}, {"loss": 0.8641, "grad_norm": 0.44816508889198303, "learning_rate": 0.0002, "epoch": 0.02263357853043408, "step": 140}, {"loss": 0.8496, "grad_norm": 0.43308213353157043, "learning_rate": 0.0002, "epoch": 0.02425026271117937, "step": 150}, {"loss": 0.8213, "grad_norm": 0.4084763526916504, "learning_rate": 0.0002, "epoch": 0.02586694689192466, "step": 160}, {"loss": 0.8343, "grad_norm": 0.5363703966140747, "learning_rate": 0.0002, "epoch": 0.027483631072669955, "step": 170}, {"loss": 0.8558, "grad_norm": 0.4619699716567993, "learning_rate": 0.0002, "epoch": 0.029100315253415245, "step": 180}, {"loss": 0.8878, "grad_norm": 0.49069908261299133, "learning_rate": 0.0002, "epoch": 0.030716999434160536, "step": 190}, {"loss": 0.8867, "grad_norm": 0.4645835757255554, "learning_rate": 0.0002, "epoch": 0.03233368361490583, "step": 200}, {"loss": 0.8842, "grad_norm": 1.2411243915557861, "learning_rate": 0.0002, "epoch": 0.03395036779565112, "step": 210}, {"loss": 0.8245, "grad_norm": 0.5211851596832275, "learning_rate": 0.0002, "epoch": 0.03556705197639641, "step": 220}, {"loss": 0.8194, "grad_norm": 0.5253691673278809, "learning_rate": 0.0002, "epoch": 0.037183736157141704, "step": 230}, {"loss": 0.8856, "grad_norm": 0.4567478895187378, "learning_rate": 0.0002, "epoch": 0.03880042033788699, "step": 240}, {"loss": 0.838, "grad_norm": 0.5472128391265869, "learning_rate": 0.0002, "epoch": 0.040417104518632285, "step": 250}, {"loss": 0.8201, "grad_norm": 0.42978546023368835, "learning_rate": 0.0002, "epoch": 0.04203378869937758, "step": 260}, {"loss": 0.8334, "grad_norm": 0.601734459400177, "learning_rate": 0.0002, "epoch": 0.043650472880122866, "step": 270}, {"loss": 0.815, "grad_norm": 0.4286513328552246, "learning_rate": 0.0002, "epoch": 0.04526715706086816, "step": 280}, {"loss": 0.8758, "grad_norm": 0.5230861902236938, "learning_rate": 0.0002, "epoch": 0.046883841241613454, "step": 290}, {"loss": 0.8636, "grad_norm": 0.6504611968994141, "learning_rate": 0.0002, "epoch": 0.04850052542235874, "step": 300}, {"loss": 0.8102, "grad_norm": 0.43485215306282043, "learning_rate": 0.0002, "epoch": 0.050117209603104035, "step": 310}, {"loss": 0.8221, "grad_norm": 0.4717007875442505, "learning_rate": 0.0002, "epoch": 0.05173389378384932, "step": 320}, {"loss": 0.8469, "grad_norm": 0.4059787690639496, "learning_rate": 0.0002, "epoch": 0.053350577964594616, "step": 330}, {"loss": 0.8866, "grad_norm": 0.4366913437843323, "learning_rate": 0.0002, "epoch": 0.05496726214533991, "step": 340}, {"loss": 0.7976, "grad_norm": 0.4233848452568054, "learning_rate": 0.0002, "epoch": 0.0565839463260852, "step": 350}, {"loss": 0.8456, "grad_norm": 0.4209108352661133, "learning_rate": 0.0002, "epoch": 0.05820063050683049, "step": 360}, {"loss": 0.816, "grad_norm": 0.41637396812438965, "learning_rate": 0.0002, "epoch": 0.059817314687575784, "step": 370}, {"loss": 0.7976, "grad_norm": 0.46235376596450806, "learning_rate": 0.0002, "epoch": 0.06143399886832107, "step": 380}, {"loss": 0.7966, "grad_norm": 0.4013484716415405, "learning_rate": 0.0002, "epoch": 0.06305068304906636, "step": 390}, {"loss": 0.8253, "grad_norm": 0.47443896532058716, "learning_rate": 0.0002, "epoch": 0.06466736722981166, "step": 400}, {"loss": 0.8666, "grad_norm": 0.3942156434059143, "learning_rate": 0.0002, "epoch": 0.06628405141055695, "step": 410}, {"loss": 0.8402, "grad_norm": 0.4965320825576782, "learning_rate": 0.0002, "epoch": 0.06790073559130223, "step": 420}, {"loss": 0.8317, "grad_norm": 0.4304835796356201, "learning_rate": 0.0002, "epoch": 0.06951741977204753, "step": 430}, {"loss": 0.8528, "grad_norm": 0.511726975440979, "learning_rate": 0.0002, "epoch": 0.07113410395279282, "step": 440}, {"loss": 0.8675, "grad_norm": 0.4040689170360565, "learning_rate": 0.0002, "epoch": 0.07275078813353811, "step": 450}, {"loss": 0.8788, "grad_norm": 0.5402171015739441, "learning_rate": 0.0002, "epoch": 0.07436747231428341, "step": 460}, {"loss": 0.8737, "grad_norm": 0.4174517095088959, "learning_rate": 0.0002, "epoch": 0.0759841564950287, "step": 470}, {"loss": 0.7605, "grad_norm": 0.4306182265281677, "learning_rate": 0.0002, "epoch": 0.07760084067577398, "step": 480}, {"loss": 0.799, "grad_norm": 0.535210132598877, "learning_rate": 0.0002, "epoch": 0.07921752485651928, "step": 490}, {"loss": 0.7825, "grad_norm": 0.5339109897613525, "learning_rate": 0.0002, "epoch": 0.08083420903726457, "step": 500}, {"loss": 0.8985, "grad_norm": 0.45754891633987427, "learning_rate": 0.0002, "epoch": 0.08245089321800986, "step": 510}, {"loss": 0.8144, "grad_norm": 0.43820783495903015, "learning_rate": 0.0002, "epoch": 0.08406757739875516, "step": 520}, {"loss": 0.8001, "grad_norm": 0.4434749186038971, "learning_rate": 0.0002, "epoch": 0.08568426157950045, "step": 530}, {"loss": 0.7857, "grad_norm": 0.43111467361450195, "learning_rate": 0.0002, "epoch": 0.08730094576024573, "step": 540}, {"loss": 0.8418, "grad_norm": 0.4378940165042877, "learning_rate": 0.0002, "epoch": 0.08891762994099103, "step": 550}, {"loss": 0.8361, "grad_norm": 0.4772215187549591, "learning_rate": 0.0002, "epoch": 0.09053431412173632, "step": 560}, {"loss": 0.8268, "grad_norm": 0.6837629079818726, "learning_rate": 0.0002, "epoch": 0.09215099830248161, "step": 570}, {"loss": 0.8607, "grad_norm": 0.42241212725639343, "learning_rate": 0.0002, "epoch": 0.09376768248322691, "step": 580}, {"loss": 0.852, "grad_norm": 0.5165936350822449, "learning_rate": 0.0002, "epoch": 0.0953843666639722, "step": 590}, {"loss": 0.8664, "grad_norm": 0.48737478256225586, "learning_rate": 0.0002, "epoch": 0.09700105084471748, "step": 600}, {"loss": 0.8806, "grad_norm": 0.47419852018356323, "learning_rate": 0.0002, "epoch": 0.09861773502546278, "step": 610}, {"loss": 0.8254, "grad_norm": 0.4975486099720001, "learning_rate": 0.0002, "epoch": 0.10023441920620807, "step": 620}, {"loss": 0.8548, "grad_norm": 0.49123844504356384, "learning_rate": 0.0002, "epoch": 0.10185110338695336, "step": 630}, {"loss": 0.8911, "grad_norm": 0.6288952827453613, "learning_rate": 0.0002, "epoch": 0.10346778756769864, "step": 640}, {"loss": 0.827, "grad_norm": 0.4277345836162567, "learning_rate": 0.0002, "epoch": 0.10508447174844394, "step": 650}, {"loss": 0.7996, "grad_norm": 0.4021061956882477, "learning_rate": 0.0002, "epoch": 0.10670115592918923, "step": 660}, {"loss": 0.87, "grad_norm": 0.3492237329483032, "learning_rate": 0.0002, "epoch": 0.10831784010993452, "step": 670}, {"loss": 0.8698, "grad_norm": 0.4341012239456177, "learning_rate": 0.0002, "epoch": 0.10993452429067982, "step": 680}, {"loss": 0.781, "grad_norm": 0.7296304106712341, "learning_rate": 0.0002, "epoch": 0.1115512084714251, "step": 690}, {"loss": 0.8433, "grad_norm": 0.397494912147522, "learning_rate": 0.0002, "epoch": 0.1131678926521704, "step": 700}, {"loss": 0.827, "grad_norm": 0.396431028842926, "learning_rate": 0.0002, "epoch": 0.1147845768329157, "step": 710}, {"loss": 0.8379, "grad_norm": 0.48842838406562805, "learning_rate": 0.0002, "epoch": 0.11640126101366098, "step": 720}, {"loss": 0.8238, "grad_norm": 0.46322616934776306, "learning_rate": 0.0002, "epoch": 0.11801794519440627, "step": 730}, {"loss": 0.8041, "grad_norm": 0.47990912199020386, "learning_rate": 0.0002, "epoch": 0.11963462937515157, "step": 740}, {"loss": 0.82, "grad_norm": 0.4997142255306244, "learning_rate": 0.0002, "epoch": 0.12125131355589686, "step": 750}, {"loss": 0.7702, "grad_norm": 0.4040526747703552, "learning_rate": 0.0002, "epoch": 0.12286799773664214, "step": 760}, {"loss": 0.863, "grad_norm": 0.453095942735672, "learning_rate": 0.0002, "epoch": 0.12448468191738744, "step": 770}, {"loss": 0.8792, "grad_norm": 0.4636971950531006, "learning_rate": 0.0002, "epoch": 0.12610136609813272, "step": 780}, {"loss": 0.8112, "grad_norm": 0.4279276132583618, "learning_rate": 0.0002, "epoch": 0.12771805027887803, "step": 790}, {"loss": 0.8711, "grad_norm": 0.46212655305862427, "learning_rate": 0.0002, "epoch": 0.12933473445962332, "step": 800}, {"loss": 0.8368, "grad_norm": 0.43127650022506714, "learning_rate": 0.0002, "epoch": 0.1309514186403686, "step": 810}, {"loss": 0.8476, "grad_norm": 0.4201301336288452, "learning_rate": 0.0002, "epoch": 0.1325681028211139, "step": 820}, {"loss": 0.8078, "grad_norm": 0.42583167552948, "learning_rate": 0.0002, "epoch": 0.13418478700185918, "step": 830}, {"loss": 0.8219, "grad_norm": 0.4535622000694275, "learning_rate": 0.0002, "epoch": 0.13580147118260447, "step": 840}, {"loss": 0.8423, "grad_norm": 0.4116036891937256, "learning_rate": 0.0002, "epoch": 0.13741815536334978, "step": 850}, {"loss": 0.8466, "grad_norm": 0.45997580885887146, "learning_rate": 0.0002, "epoch": 0.13903483954409507, "step": 860}, {"loss": 0.8917, "grad_norm": 0.4487837255001068, "learning_rate": 0.0002, "epoch": 0.14065152372484035, "step": 870}, {"loss": 0.8217, "grad_norm": 0.43650057911872864, "learning_rate": 0.0002, "epoch": 0.14226820790558564, "step": 880}, {"loss": 0.8178, "grad_norm": 0.5335358381271362, "learning_rate": 0.0002, "epoch": 0.14388489208633093, "step": 890}, {"loss": 0.7957, "grad_norm": 0.5989000201225281, "learning_rate": 0.0002, "epoch": 0.14550157626707622, "step": 900}, {"loss": 0.8385, "grad_norm": 0.517179012298584, "learning_rate": 0.0002, "epoch": 0.14711826044782153, "step": 910}, {"loss": 0.8255, "grad_norm": 0.44435232877731323, "learning_rate": 0.0002, "epoch": 0.14873494462856682, "step": 920}, {"loss": 0.8305, "grad_norm": 0.42635923624038696, "learning_rate": 0.0002, "epoch": 0.1503516288093121, "step": 930}, {"loss": 0.8043, "grad_norm": 0.49603334069252014, "learning_rate": 0.0002, "epoch": 0.1519683129900574, "step": 940}, {"loss": 0.8377, "grad_norm": 0.40639808773994446, "learning_rate": 0.0002, "epoch": 0.15358499717080268, "step": 950}, {"loss": 0.8529, "grad_norm": 0.4850759208202362, "learning_rate": 0.0002, "epoch": 0.15520168135154797, "step": 960}, {"loss": 0.846, "grad_norm": 0.4427442252635956, "learning_rate": 0.0002, "epoch": 0.15681836553229328, "step": 970}, {"loss": 0.8705, "grad_norm": 0.3760930001735687, "learning_rate": 0.0002, "epoch": 0.15843504971303857, "step": 980}, {"loss": 0.8644, "grad_norm": 0.4794144332408905, "learning_rate": 0.0002, "epoch": 0.16005173389378385, "step": 990}, {"loss": 0.8002, "grad_norm": 0.45828768610954285, "learning_rate": 0.0002, "epoch": 0.16166841807452914, "step": 1000}, {"loss": 0.7658, "grad_norm": 0.6313053369522095, "learning_rate": 0.0002, "epoch": 0.16328510225527443, "step": 1010}, {"loss": 0.8047, "grad_norm": 0.45041006803512573, "learning_rate": 0.0002, "epoch": 0.16490178643601971, "step": 1020}, {"loss": 0.8423, "grad_norm": 0.441403865814209, "learning_rate": 0.0002, "epoch": 0.166518470616765, "step": 1030}, {"loss": 0.8475, "grad_norm": 0.8171296119689941, "learning_rate": 0.0002, "epoch": 0.16813515479751032, "step": 1040}, {"loss": 0.845, "grad_norm": 0.7137420773506165, "learning_rate": 0.0002, "epoch": 0.1697518389782556, "step": 1050}, {"loss": 0.8213, "grad_norm": 0.5236809849739075, "learning_rate": 0.0002, "epoch": 0.1713685231590009, "step": 1060}, {"loss": 0.8265, "grad_norm": 0.5021864175796509, "learning_rate": 0.0002, "epoch": 0.17298520733974618, "step": 1070}, {"loss": 0.8305, "grad_norm": 0.47347521781921387, "learning_rate": 0.0002, "epoch": 0.17460189152049146, "step": 1080}, {"loss": 0.8105, "grad_norm": 0.4631653428077698, "learning_rate": 0.0002, "epoch": 0.17621857570123675, "step": 1090}, {"loss": 0.8166, "grad_norm": 0.49169182777404785, "learning_rate": 0.0002, "epoch": 0.17783525988198207, "step": 1100}, {"loss": 0.8012, "grad_norm": 0.5019739270210266, "learning_rate": 0.0002, "epoch": 0.17945194406272735, "step": 1110}, {"loss": 0.8247, "grad_norm": 0.5100422501564026, "learning_rate": 0.0002, "epoch": 0.18106862824347264, "step": 1120}, {"loss": 0.8142, "grad_norm": 0.3888324499130249, "learning_rate": 0.0002, "epoch": 0.18268531242421793, "step": 1130}, {"loss": 0.8533, "grad_norm": 0.39765217900276184, "learning_rate": 0.0002, "epoch": 0.18430199660496321, "step": 1140}, {"loss": 0.8541, "grad_norm": 0.47190186381340027, "learning_rate": 0.0002, "epoch": 0.1859186807857085, "step": 1150}, {"loss": 0.8301, "grad_norm": 0.4464188814163208, "learning_rate": 0.0002, "epoch": 0.18753536496645382, "step": 1160}, {"loss": 0.8341, "grad_norm": 0.5153930187225342, "learning_rate": 0.0002, "epoch": 0.1891520491471991, "step": 1170}, {"loss": 0.8033, "grad_norm": 0.4779708683490753, "learning_rate": 0.0002, "epoch": 0.1907687333279444, "step": 1180}, {"loss": 0.8187, "grad_norm": 0.4834315776824951, "learning_rate": 0.0002, "epoch": 0.19238541750868968, "step": 1190}, {"loss": 0.7721, "grad_norm": 0.402357816696167, "learning_rate": 0.0002, "epoch": 0.19400210168943496, "step": 1200}, {"loss": 0.7941, "grad_norm": 0.45899084210395813, "learning_rate": 0.0002, "epoch": 0.19561878587018025, "step": 1210}, {"loss": 0.8353, "grad_norm": 0.5106529593467712, "learning_rate": 0.0002, "epoch": 0.19723547005092557, "step": 1220}, {"loss": 0.7816, "grad_norm": 0.45261722803115845, "learning_rate": 0.0002, "epoch": 0.19885215423167085, "step": 1230}, {"loss": 0.8068, "grad_norm": 0.4647127091884613, "learning_rate": 0.0002, "epoch": 0.20046883841241614, "step": 1240}, {"loss": 0.8239, "grad_norm": 0.4849368929862976, "learning_rate": 0.0002, "epoch": 0.20208552259316143, "step": 1250}, {"loss": 0.8514, "grad_norm": 0.4518061578273773, "learning_rate": 0.0002, "epoch": 0.2037022067739067, "step": 1260}, {"loss": 0.8158, "grad_norm": 0.49535325169563293, "learning_rate": 0.0002, "epoch": 0.205318890954652, "step": 1270}, {"loss": 0.8348, "grad_norm": 0.4835205376148224, "learning_rate": 0.0002, "epoch": 0.2069355751353973, "step": 1280}, {"loss": 0.8428, "grad_norm": 0.45308539271354675, "learning_rate": 0.0002, "epoch": 0.2085522593161426, "step": 1290}, {"loss": 0.7993, "grad_norm": 0.5369905233383179, "learning_rate": 0.0002, "epoch": 0.2101689434968879, "step": 1300}, {"loss": 0.8676, "grad_norm": 0.5031622052192688, "learning_rate": 0.0002, "epoch": 0.21178562767763318, "step": 1310}, {"loss": 0.7686, "grad_norm": 0.48010334372520447, "learning_rate": 0.0002, "epoch": 0.21340231185837846, "step": 1320}, {"loss": 0.806, "grad_norm": 0.4905701279640198, "learning_rate": 0.0002, "epoch": 0.21501899603912375, "step": 1330}, {"loss": 0.7885, "grad_norm": 0.43531742691993713, "learning_rate": 0.0002, "epoch": 0.21663568021986904, "step": 1340}, {"loss": 0.8191, "grad_norm": 0.44330692291259766, "learning_rate": 0.0002, "epoch": 0.21825236440061435, "step": 1350}, {"loss": 0.8205, "grad_norm": 0.5384416580200195, "learning_rate": 0.0002, "epoch": 0.21986904858135964, "step": 1360}, {"loss": 0.7726, "grad_norm": 0.4181833863258362, "learning_rate": 0.0002, "epoch": 0.22148573276210493, "step": 1370}, {"loss": 0.8311, "grad_norm": 0.523833692073822, "learning_rate": 0.0002, "epoch": 0.2231024169428502, "step": 1380}, {"loss": 0.7913, "grad_norm": 0.5528736710548401, "learning_rate": 0.0002, "epoch": 0.2247191011235955, "step": 1390}, {"loss": 0.8079, "grad_norm": 0.43515023589134216, "learning_rate": 0.0002, "epoch": 0.2263357853043408, "step": 1400}, {"loss": 0.8403, "grad_norm": 0.48809877038002014, "learning_rate": 0.0002, "epoch": 0.2279524694850861, "step": 1410}, {"loss": 0.8165, "grad_norm": 0.43591251969337463, "learning_rate": 0.0002, "epoch": 0.2295691536658314, "step": 1420}, {"loss": 0.8147, "grad_norm": 0.44625312089920044, "learning_rate": 0.0002, "epoch": 0.23118583784657668, "step": 1430}, {"loss": 0.8134, "grad_norm": 0.4390665292739868, "learning_rate": 0.0002, "epoch": 0.23280252202732196, "step": 1440}, {"loss": 0.8465, "grad_norm": 0.48496049642562866, "learning_rate": 0.0002, "epoch": 0.23441920620806725, "step": 1450}, {"loss": 0.775, "grad_norm": 0.45919957756996155, "learning_rate": 0.0002, "epoch": 0.23603589038881254, "step": 1460}, {"loss": 0.8659, "grad_norm": 0.5471845865249634, "learning_rate": 0.0002, "epoch": 0.23765257456955785, "step": 1470}, {"loss": 0.8164, "grad_norm": 0.47269317507743835, "learning_rate": 0.0002, "epoch": 0.23926925875030314, "step": 1480}, {"loss": 0.854, "grad_norm": 0.4930245578289032, "learning_rate": 0.0002, "epoch": 0.24088594293104842, "step": 1490}, {"loss": 0.8139, "grad_norm": 0.5605630278587341, "learning_rate": 0.0002, "epoch": 0.2425026271117937, "step": 1500}, {"loss": 0.8125, "grad_norm": 0.4435870945453644, "learning_rate": 0.0002, "epoch": 0.244119311292539, "step": 1510}, {"loss": 0.8123, "grad_norm": 0.4941999912261963, "learning_rate": 0.0002, "epoch": 0.24573599547328429, "step": 1520}, {"loss": 0.8427, "grad_norm": 0.5100624561309814, "learning_rate": 0.0002, "epoch": 0.24735267965402957, "step": 1530}, {"loss": 0.8405, "grad_norm": 0.4638267457485199, "learning_rate": 0.0002, "epoch": 0.2489693638347749, "step": 1540}, {"loss": 0.81, "grad_norm": 0.5071570873260498, "learning_rate": 0.0002, "epoch": 0.25058604801552015, "step": 1550}, {"loss": 0.7724, "grad_norm": 0.4291319251060486, "learning_rate": 0.0002, "epoch": 0.25220273219626543, "step": 1560}, {"loss": 0.7984, "grad_norm": 0.5388049483299255, "learning_rate": 0.0002, "epoch": 0.2538194163770108, "step": 1570}, {"loss": 0.8176, "grad_norm": 0.5083683729171753, "learning_rate": 0.0002, "epoch": 0.25543610055775606, "step": 1580}, {"loss": 0.843, "grad_norm": 0.4824463725090027, "learning_rate": 0.0002, "epoch": 0.25705278473850135, "step": 1590}, {"loss": 0.7996, "grad_norm": 0.41177722811698914, "learning_rate": 0.0002, "epoch": 0.25866946891924664, "step": 1600}, {"loss": 0.7772, "grad_norm": 0.5656219124794006, "learning_rate": 0.0002, "epoch": 0.2602861530999919, "step": 1610}, {"loss": 0.7955, "grad_norm": 0.41063204407691956, "learning_rate": 0.0002, "epoch": 0.2619028372807372, "step": 1620}, {"loss": 0.7998, "grad_norm": 0.4897061288356781, "learning_rate": 0.0002, "epoch": 0.2635195214614825, "step": 1630}, {"loss": 0.8198, "grad_norm": 0.4454376697540283, "learning_rate": 0.0002, "epoch": 0.2651362056422278, "step": 1640}, {"loss": 0.8684, "grad_norm": 0.4355238378047943, "learning_rate": 0.0002, "epoch": 0.26675288982297307, "step": 1650}, {"loss": 0.7801, "grad_norm": 0.458310067653656, "learning_rate": 0.0002, "epoch": 0.26836957400371836, "step": 1660}, {"loss": 0.7935, "grad_norm": 0.4752083718776703, "learning_rate": 0.0002, "epoch": 0.26998625818446365, "step": 1670}, {"loss": 0.8267, "grad_norm": 0.4666106402873993, "learning_rate": 0.0002, "epoch": 0.27160294236520893, "step": 1680}, {"loss": 0.8252, "grad_norm": 0.4213818609714508, "learning_rate": 0.0002, "epoch": 0.2732196265459543, "step": 1690}, {"loss": 0.8559, "grad_norm": 0.5768913626670837, "learning_rate": 0.0002, "epoch": 0.27483631072669956, "step": 1700}, {"loss": 0.7931, "grad_norm": 0.4209914803504944, "learning_rate": 0.0002, "epoch": 0.27645299490744485, "step": 1710}, {"loss": 0.8167, "grad_norm": 0.501909613609314, "learning_rate": 0.0002, "epoch": 0.27806967908819014, "step": 1720}, {"loss": 0.7832, "grad_norm": 0.5266261100769043, "learning_rate": 0.0002, "epoch": 0.2796863632689354, "step": 1730}, {"loss": 0.8102, "grad_norm": 0.43806859850883484, "learning_rate": 0.0002, "epoch": 0.2813030474496807, "step": 1740}, {"loss": 0.8157, "grad_norm": 0.46048814058303833, "learning_rate": 0.0002, "epoch": 0.282919731630426, "step": 1750}, {"loss": 0.8596, "grad_norm": 0.44972819089889526, "learning_rate": 0.0002, "epoch": 0.2845364158111713, "step": 1760}, {"loss": 0.8421, "grad_norm": 0.5114831328392029, "learning_rate": 0.0002, "epoch": 0.28615309999191657, "step": 1770}, {"loss": 0.8361, "grad_norm": 0.47931742668151855, "learning_rate": 0.0002, "epoch": 0.28776978417266186, "step": 1780}, {"loss": 0.8265, "grad_norm": 0.5092599987983704, "learning_rate": 0.0002, "epoch": 0.28938646835340714, "step": 1790}, {"loss": 0.8506, "grad_norm": 0.37581443786621094, "learning_rate": 0.0002, "epoch": 0.29100315253415243, "step": 1800}, {"loss": 0.7932, "grad_norm": 0.47097381949424744, "learning_rate": 0.0002, "epoch": 0.2926198367148977, "step": 1810}, {"loss": 0.7787, "grad_norm": 0.48300236463546753, "learning_rate": 0.0002, "epoch": 0.29423652089564306, "step": 1820}, {"loss": 0.8391, "grad_norm": 0.5600419640541077, "learning_rate": 0.0002, "epoch": 0.29585320507638835, "step": 1830}, {"loss": 0.8507, "grad_norm": 0.48555272817611694, "learning_rate": 0.0002, "epoch": 0.29746988925713364, "step": 1840}, {"loss": 0.7657, "grad_norm": 0.3752668499946594, "learning_rate": 0.0002, "epoch": 0.2990865734378789, "step": 1850}, {"loss": 0.7915, "grad_norm": 0.5328747034072876, "learning_rate": 0.0002, "epoch": 0.3007032576186242, "step": 1860}, {"loss": 0.8426, "grad_norm": 0.48716455698013306, "learning_rate": 0.0002, "epoch": 0.3023199417993695, "step": 1870}, {"loss": 0.8335, "grad_norm": 0.5011493563652039, "learning_rate": 0.0002, "epoch": 0.3039366259801148, "step": 1880}, {"loss": 0.852, "grad_norm": 0.46461427211761475, "learning_rate": 0.0002, "epoch": 0.30555331016086007, "step": 1890}, {"loss": 0.8478, "grad_norm": 0.36630210280418396, "learning_rate": 0.0002, "epoch": 0.30716999434160536, "step": 1900}, {"loss": 0.8162, "grad_norm": 0.4217296242713928, "learning_rate": 0.0002, "epoch": 0.30878667852235064, "step": 1910}, {"loss": 0.8128, "grad_norm": 0.4394875466823578, "learning_rate": 0.0002, "epoch": 0.31040336270309593, "step": 1920}, {"loss": 0.8471, "grad_norm": 0.6587965488433838, "learning_rate": 0.0002, "epoch": 0.3120200468838412, "step": 1930}, {"loss": 0.8565, "grad_norm": 0.5469298958778381, "learning_rate": 0.0002, "epoch": 0.31363673106458656, "step": 1940}, {"loss": 0.8236, "grad_norm": 0.4371595084667206, "learning_rate": 0.0002, "epoch": 0.31525341524533185, "step": 1950}, {"loss": 0.887, "grad_norm": 0.4809541404247284, "learning_rate": 0.0002, "epoch": 0.31687009942607713, "step": 1960}, {"loss": 0.7855, "grad_norm": 0.6061086654663086, "learning_rate": 0.0002, "epoch": 0.3184867836068224, "step": 1970}, {"loss": 0.7679, "grad_norm": 0.5342657566070557, "learning_rate": 0.0002, "epoch": 0.3201034677875677, "step": 1980}, {"loss": 0.7955, "grad_norm": 0.5057743787765503, "learning_rate": 0.0002, "epoch": 0.321720151968313, "step": 1990}, {"loss": 0.7774, "grad_norm": 0.528626024723053, "learning_rate": 0.0002, "epoch": 0.3233368361490583, "step": 2000}, {"loss": 0.8845, "grad_norm": 0.46742770075798035, "learning_rate": 0.0002, "epoch": 0.32495352032980357, "step": 2010}, {"loss": 0.8484, "grad_norm": 0.515101432800293, "learning_rate": 0.0002, "epoch": 0.32657020451054886, "step": 2020}, {"loss": 0.8139, "grad_norm": 0.41941216588020325, "learning_rate": 0.0002, "epoch": 0.32818688869129414, "step": 2030}, {"loss": 0.7637, "grad_norm": 0.49902522563934326, "learning_rate": 0.0002, "epoch": 0.32980357287203943, "step": 2040}, {"loss": 0.7822, "grad_norm": 0.4120897650718689, "learning_rate": 0.0002, "epoch": 0.3314202570527847, "step": 2050}, {"loss": 0.8057, "grad_norm": 0.45352041721343994, "learning_rate": 0.0002, "epoch": 0.33303694123353, "step": 2060}, {"loss": 0.7913, "grad_norm": 0.523199737071991, "learning_rate": 0.0002, "epoch": 0.33465362541427535, "step": 2070}, {"loss": 0.8036, "grad_norm": 0.4390358626842499, "learning_rate": 0.0002, "epoch": 0.33627030959502063, "step": 2080}, {"loss": 0.8145, "grad_norm": 0.6752901077270508, "learning_rate": 0.0002, "epoch": 0.3378869937757659, "step": 2090}, {"loss": 0.7807, "grad_norm": 0.547821044921875, "learning_rate": 0.0002, "epoch": 0.3395036779565112, "step": 2100}, {"loss": 0.8561, "grad_norm": 0.5161308646202087, "learning_rate": 0.0002, "epoch": 0.3411203621372565, "step": 2110}, {"loss": 0.7697, "grad_norm": 0.4565401077270508, "learning_rate": 0.0002, "epoch": 0.3427370463180018, "step": 2120}, {"loss": 0.7964, "grad_norm": 0.4666115939617157, "learning_rate": 0.0002, "epoch": 0.34435373049874707, "step": 2130}, {"loss": 0.8189, "grad_norm": 0.4090428352355957, "learning_rate": 0.0002, "epoch": 0.34597041467949236, "step": 2140}, {"loss": 0.8817, "grad_norm": 0.510845422744751, "learning_rate": 0.0002, "epoch": 0.34758709886023764, "step": 2150}, {"loss": 0.8398, "grad_norm": 0.42861923575401306, "learning_rate": 0.0002, "epoch": 0.34920378304098293, "step": 2160}, {"loss": 0.7716, "grad_norm": 0.4476332664489746, "learning_rate": 0.0002, "epoch": 0.3508204672217282, "step": 2170}, {"loss": 0.7845, "grad_norm": 0.6065791249275208, "learning_rate": 0.0002, "epoch": 0.3524371514024735, "step": 2180}, {"loss": 0.8187, "grad_norm": 0.42335066199302673, "learning_rate": 0.0002, "epoch": 0.35405383558321885, "step": 2190}, {"loss": 0.8239, "grad_norm": 0.5094629526138306, "learning_rate": 0.0002, "epoch": 0.35567051976396413, "step": 2200}, {"loss": 0.7807, "grad_norm": 0.5476373434066772, "learning_rate": 0.0002, "epoch": 0.3572872039447094, "step": 2210}, {"loss": 0.814, "grad_norm": 0.3911719024181366, "learning_rate": 0.0002, "epoch": 0.3589038881254547, "step": 2220}, {"loss": 0.8599, "grad_norm": 0.6599636077880859, "learning_rate": 0.0002, "epoch": 0.3605205723062, "step": 2230}, {"loss": 0.7482, "grad_norm": 0.40381914377212524, "learning_rate": 0.0002, "epoch": 0.3621372564869453, "step": 2240}, {"loss": 0.7772, "grad_norm": 0.4433908462524414, "learning_rate": 0.0002, "epoch": 0.36375394066769057, "step": 2250}, {"loss": 0.8503, "grad_norm": 0.578326940536499, "learning_rate": 0.0002, "epoch": 0.36537062484843585, "step": 2260}, {"loss": 0.8178, "grad_norm": 0.5734784007072449, "learning_rate": 0.0002, "epoch": 0.36698730902918114, "step": 2270}, {"loss": 0.8193, "grad_norm": 0.45555487275123596, "learning_rate": 0.0002, "epoch": 0.36860399320992643, "step": 2280}, {"loss": 0.7929, "grad_norm": 0.5666276216506958, "learning_rate": 0.0002, "epoch": 0.3702206773906717, "step": 2290}, {"loss": 0.8292, "grad_norm": 0.5461117625236511, "learning_rate": 0.0002, "epoch": 0.371837361571417, "step": 2300}, {"loss": 0.8204, "grad_norm": 0.6318911910057068, "learning_rate": 0.0002, "epoch": 0.3734540457521623, "step": 2310}, {"loss": 0.7964, "grad_norm": 0.493263304233551, "learning_rate": 0.0002, "epoch": 0.37507072993290763, "step": 2320}, {"loss": 0.8339, "grad_norm": 0.5888760089874268, "learning_rate": 0.0002, "epoch": 0.3766874141136529, "step": 2330}, {"loss": 0.7737, "grad_norm": 0.48671841621398926, "learning_rate": 0.0002, "epoch": 0.3783040982943982, "step": 2340}, {"loss": 0.8367, "grad_norm": 0.4385145306587219, "learning_rate": 0.0002, "epoch": 0.3799207824751435, "step": 2350}, {"loss": 0.812, "grad_norm": 0.5523318648338318, "learning_rate": 0.0002, "epoch": 0.3815374666558888, "step": 2360}, {"loss": 0.8351, "grad_norm": 0.7308220267295837, "learning_rate": 0.0002, "epoch": 0.38315415083663407, "step": 2370}, {"loss": 0.859, "grad_norm": 0.554214358329773, "learning_rate": 0.0002, "epoch": 0.38477083501737935, "step": 2380}, {"loss": 0.8146, "grad_norm": 0.5425800085067749, "learning_rate": 0.0002, "epoch": 0.38638751919812464, "step": 2390}, {"loss": 0.8282, "grad_norm": 0.48811158537864685, "learning_rate": 0.0002, "epoch": 0.3880042033788699, "step": 2400}, {"loss": 0.8074, "grad_norm": 0.49212366342544556, "learning_rate": 0.0002, "epoch": 0.3896208875596152, "step": 2410}, {"loss": 0.7991, "grad_norm": 0.5222218632698059, "learning_rate": 0.0002, "epoch": 0.3912375717403605, "step": 2420}, {"loss": 0.8182, "grad_norm": 0.4699819087982178, "learning_rate": 0.0002, "epoch": 0.3928542559211058, "step": 2430}, {"loss": 0.7919, "grad_norm": 0.46153587102890015, "learning_rate": 0.0002, "epoch": 0.39447094010185113, "step": 2440}, {"loss": 0.8111, "grad_norm": 0.4150611162185669, "learning_rate": 0.0002, "epoch": 0.3960876242825964, "step": 2450}, {"loss": 0.8589, "grad_norm": 0.5799614787101746, "learning_rate": 0.0002, "epoch": 0.3977043084633417, "step": 2460}, {"loss": 0.8085, "grad_norm": 0.56536865234375, "learning_rate": 0.0002, "epoch": 0.399320992644087, "step": 2470}, {"loss": 0.8022, "grad_norm": 0.5451247096061707, "learning_rate": 0.0002, "epoch": 0.4009376768248323, "step": 2480}, {"loss": 0.8217, "grad_norm": 0.5914521217346191, "learning_rate": 0.0002, "epoch": 0.40255436100557757, "step": 2490}, {"loss": 0.7859, "grad_norm": 0.4428117275238037, "learning_rate": 0.0002, "epoch": 0.40417104518632285, "step": 2500}, {"loss": 0.8054, "grad_norm": 0.48580947518348694, "learning_rate": 0.0002, "epoch": 0.40578772936706814, "step": 2510}, {"loss": 0.8405, "grad_norm": 0.436734676361084, "learning_rate": 0.0002, "epoch": 0.4074044135478134, "step": 2520}, {"loss": 0.8209, "grad_norm": 0.5752223134040833, "learning_rate": 0.0002, "epoch": 0.4090210977285587, "step": 2530}, {"loss": 0.8181, "grad_norm": 0.4271308183670044, "learning_rate": 0.0002, "epoch": 0.410637781909304, "step": 2540}, {"loss": 0.8058, "grad_norm": 0.46294718980789185, "learning_rate": 0.0002, "epoch": 0.4122544660900493, "step": 2550}, {"loss": 0.8473, "grad_norm": 0.49407583475112915, "learning_rate": 0.0002, "epoch": 0.4138711502707946, "step": 2560}, {"loss": 0.7881, "grad_norm": 0.4729035496711731, "learning_rate": 0.0002, "epoch": 0.4154878344515399, "step": 2570}, {"loss": 0.7834, "grad_norm": 0.4129747152328491, "learning_rate": 0.0002, "epoch": 0.4171045186322852, "step": 2580}, {"loss": 0.7859, "grad_norm": 0.5684236288070679, "learning_rate": 0.0002, "epoch": 0.4187212028130305, "step": 2590}, {"loss": 0.811, "grad_norm": 0.4862157106399536, "learning_rate": 0.0002, "epoch": 0.4203378869937758, "step": 2600}, {"loss": 0.7582, "grad_norm": 0.46567976474761963, "learning_rate": 0.0002, "epoch": 0.42195457117452106, "step": 2610}, {"loss": 0.7755, "grad_norm": 0.5710650682449341, "learning_rate": 0.0002, "epoch": 0.42357125535526635, "step": 2620}, {"loss": 0.8573, "grad_norm": 0.5660041570663452, "learning_rate": 0.0002, "epoch": 0.42518793953601164, "step": 2630}, {"loss": 0.7812, "grad_norm": 0.47944375872612, "learning_rate": 0.0002, "epoch": 0.4268046237167569, "step": 2640}, {"loss": 0.7459, "grad_norm": 0.537223756313324, "learning_rate": 0.0002, "epoch": 0.4284213078975022, "step": 2650}, {"loss": 0.8246, "grad_norm": 0.41669997572898865, "learning_rate": 0.0002, "epoch": 0.4300379920782475, "step": 2660}, {"loss": 0.7785, "grad_norm": 0.44727686047554016, "learning_rate": 0.0002, "epoch": 0.4316546762589928, "step": 2670}, {"loss": 0.8241, "grad_norm": 0.5600888729095459, "learning_rate": 0.0002, "epoch": 0.4332713604397381, "step": 2680}, {"loss": 0.7708, "grad_norm": 0.39820605516433716, "learning_rate": 0.0002, "epoch": 0.4348880446204834, "step": 2690}, {"loss": 0.8202, "grad_norm": 0.5637655854225159, "learning_rate": 0.0002, "epoch": 0.4365047288012287, "step": 2700}, {"loss": 0.855, "grad_norm": 0.6363666653633118, "learning_rate": 0.0002, "epoch": 0.438121412981974, "step": 2710}, {"loss": 0.8468, "grad_norm": 0.5656129121780396, "learning_rate": 0.0002, "epoch": 0.4397380971627193, "step": 2720}, {"loss": 0.7845, "grad_norm": 0.5600156188011169, "learning_rate": 0.0002, "epoch": 0.44135478134346456, "step": 2730}, {"loss": 0.8405, "grad_norm": 0.5506579875946045, "learning_rate": 0.0002, "epoch": 0.44297146552420985, "step": 2740}, {"loss": 0.7725, "grad_norm": 0.49878305196762085, "learning_rate": 0.0002, "epoch": 0.44458814970495514, "step": 2750}, {"loss": 0.8292, "grad_norm": 0.4569213092327118, "learning_rate": 0.0002, "epoch": 0.4462048338857004, "step": 2760}, {"loss": 0.8028, "grad_norm": 0.6056680083274841, "learning_rate": 0.0002, "epoch": 0.4478215180664457, "step": 2770}, {"loss": 0.8242, "grad_norm": 0.44474557042121887, "learning_rate": 0.0002, "epoch": 0.449438202247191, "step": 2780}, {"loss": 0.801, "grad_norm": 0.46055394411087036, "learning_rate": 0.0002, "epoch": 0.4510548864279363, "step": 2790}, {"loss": 0.7521, "grad_norm": 0.4904133379459381, "learning_rate": 0.0002, "epoch": 0.4526715706086816, "step": 2800}, {"loss": 0.8829, "grad_norm": 0.5647031664848328, "learning_rate": 0.0002, "epoch": 0.45428825478942686, "step": 2810}, {"loss": 0.8622, "grad_norm": 0.5759473443031311, "learning_rate": 0.0002, "epoch": 0.4559049389701722, "step": 2820}, {"loss": 0.7812, "grad_norm": 0.5161895751953125, "learning_rate": 0.0002, "epoch": 0.4575216231509175, "step": 2830}, {"loss": 0.8045, "grad_norm": 0.4248254597187042, "learning_rate": 0.0002, "epoch": 0.4591383073316628, "step": 2840}, {"loss": 0.7838, "grad_norm": 0.45395001769065857, "learning_rate": 0.0002, "epoch": 0.46075499151240806, "step": 2850}, {"loss": 0.8208, "grad_norm": 0.5358697772026062, "learning_rate": 0.0002, "epoch": 0.46237167569315335, "step": 2860}, {"loss": 0.8147, "grad_norm": 0.5379165410995483, "learning_rate": 0.0002, "epoch": 0.46398835987389864, "step": 2870}, {"loss": 0.7403, "grad_norm": 0.4601989686489105, "learning_rate": 0.0002, "epoch": 0.4656050440546439, "step": 2880}, {"loss": 0.8523, "grad_norm": 0.671115517616272, "learning_rate": 0.0002, "epoch": 0.4672217282353892, "step": 2890}, {"loss": 0.8262, "grad_norm": 0.4425133168697357, "learning_rate": 0.0002, "epoch": 0.4688384124161345, "step": 2900}, {"loss": 0.8178, "grad_norm": 0.5446155071258545, "learning_rate": 0.0002, "epoch": 0.4704550965968798, "step": 2910}, {"loss": 0.8106, "grad_norm": 0.603306233882904, "learning_rate": 0.0002, "epoch": 0.47207178077762507, "step": 2920}, {"loss": 0.8044, "grad_norm": 0.5377997159957886, "learning_rate": 0.0002, "epoch": 0.47368846495837036, "step": 2930}, {"loss": 0.8075, "grad_norm": 0.4931027591228485, "learning_rate": 0.0002, "epoch": 0.4753051491391157, "step": 2940}, {"loss": 0.8004, "grad_norm": 0.4711960256099701, "learning_rate": 0.0002, "epoch": 0.476921833319861, "step": 2950}, {"loss": 0.8121, "grad_norm": 0.5020492672920227, "learning_rate": 0.0002, "epoch": 0.4785385175006063, "step": 2960}, {"loss": 0.8221, "grad_norm": 0.5428946614265442, "learning_rate": 0.0002, "epoch": 0.48015520168135156, "step": 2970}, {"loss": 0.7849, "grad_norm": 0.5294089317321777, "learning_rate": 0.0002, "epoch": 0.48177188586209685, "step": 2980}, {"loss": 0.8553, "grad_norm": 0.648289144039154, "learning_rate": 0.0002, "epoch": 0.48338857004284214, "step": 2990}, {"loss": 0.7874, "grad_norm": 0.47916680574417114, "learning_rate": 0.0002, "epoch": 0.4850052542235874, "step": 3000}, {"loss": 0.8087, "grad_norm": 0.43849772214889526, "learning_rate": 0.0002, "epoch": 0.4866219384043327, "step": 3010}, {"loss": 0.7662, "grad_norm": 0.47007861733436584, "learning_rate": 0.0002, "epoch": 0.488238622585078, "step": 3020}, {"loss": 0.757, "grad_norm": 0.6314331293106079, "learning_rate": 0.0002, "epoch": 0.4898553067658233, "step": 3030}, {"loss": 0.7863, "grad_norm": 0.49211493134498596, "learning_rate": 0.0002, "epoch": 0.49147199094656857, "step": 3040}, {"loss": 0.8335, "grad_norm": 0.4537973403930664, "learning_rate": 0.0002, "epoch": 0.49308867512731386, "step": 3050}, {"loss": 0.8095, "grad_norm": 0.47326919436454773, "learning_rate": 0.0002, "epoch": 0.49470535930805914, "step": 3060}, {"loss": 0.8447, "grad_norm": 0.525874137878418, "learning_rate": 0.0002, "epoch": 0.4963220434888045, "step": 3070}, {"loss": 0.8339, "grad_norm": 0.6361091732978821, "learning_rate": 0.0002, "epoch": 0.4979387276695498, "step": 3080}, {"loss": 0.821, "grad_norm": 0.5850642919540405, "learning_rate": 0.0002, "epoch": 0.49955541185029506, "step": 3090}, {"loss": 0.8279, "grad_norm": 0.47299543023109436, "learning_rate": 0.0002, "epoch": 0.5011720960310403, "step": 3100}, {"loss": 0.8681, "grad_norm": 0.473099946975708, "learning_rate": 0.0002, "epoch": 0.5027887802117856, "step": 3110}, {"loss": 0.8223, "grad_norm": 0.48186397552490234, "learning_rate": 0.0002, "epoch": 0.5044054643925309, "step": 3120}, {"loss": 0.8292, "grad_norm": 0.5015401840209961, "learning_rate": 0.0002, "epoch": 0.5060221485732762, "step": 3130}, {"loss": 0.7692, "grad_norm": 0.5617750287055969, "learning_rate": 0.0002, "epoch": 0.5076388327540216, "step": 3140}, {"loss": 0.8708, "grad_norm": 0.5169327259063721, "learning_rate": 0.0002, "epoch": 0.5092555169347668, "step": 3150}, {"loss": 0.7845, "grad_norm": 0.545657753944397, "learning_rate": 0.0002, "epoch": 0.5108722011155121, "step": 3160}, {"loss": 0.799, "grad_norm": 0.512864351272583, "learning_rate": 0.0002, "epoch": 0.5124888852962574, "step": 3170}, {"loss": 0.7794, "grad_norm": 0.4113546311855316, "learning_rate": 0.0002, "epoch": 0.5141055694770027, "step": 3180}, {"loss": 0.8206, "grad_norm": 0.44532445073127747, "learning_rate": 0.0002, "epoch": 0.5157222536577479, "step": 3190}, {"loss": 0.8213, "grad_norm": 0.5623497366905212, "learning_rate": 0.0002, "epoch": 0.5173389378384933, "step": 3200}, {"loss": 0.7928, "grad_norm": 0.5084741115570068, "learning_rate": 0.0002, "epoch": 0.5189556220192385, "step": 3210}, {"loss": 0.8174, "grad_norm": 0.5305403470993042, "learning_rate": 0.0002, "epoch": 0.5205723061999838, "step": 3220}, {"loss": 0.8139, "grad_norm": 0.4708254337310791, "learning_rate": 0.0002, "epoch": 0.5221889903807291, "step": 3230}, {"loss": 0.7639, "grad_norm": 0.43827131390571594, "learning_rate": 0.0002, "epoch": 0.5238056745614744, "step": 3240}, {"loss": 0.7993, "grad_norm": 0.5630002617835999, "learning_rate": 0.0002, "epoch": 0.5254223587422197, "step": 3250}, {"loss": 0.7522, "grad_norm": 0.5010961890220642, "learning_rate": 0.0002, "epoch": 0.527039042922965, "step": 3260}, {"loss": 0.8374, "grad_norm": 0.6303122043609619, "learning_rate": 0.0002, "epoch": 0.5286557271037103, "step": 3270}, {"loss": 0.7727, "grad_norm": 0.5107331275939941, "learning_rate": 0.0002, "epoch": 0.5302724112844556, "step": 3280}, {"loss": 0.8495, "grad_norm": 0.5700443387031555, "learning_rate": 0.0002, "epoch": 0.5318890954652009, "step": 3290}, {"loss": 0.7776, "grad_norm": 0.46296367049217224, "learning_rate": 0.0002, "epoch": 0.5335057796459461, "step": 3300}, {"loss": 0.7931, "grad_norm": 0.531568706035614, "learning_rate": 0.0002, "epoch": 0.5351224638266915, "step": 3310}, {"loss": 0.843, "grad_norm": 0.4686741530895233, "learning_rate": 0.0002, "epoch": 0.5367391480074367, "step": 3320}, {"loss": 0.8104, "grad_norm": 0.5404331088066101, "learning_rate": 0.0002, "epoch": 0.5383558321881821, "step": 3330}, {"loss": 0.7686, "grad_norm": 0.6368790864944458, "learning_rate": 0.0002, "epoch": 0.5399725163689273, "step": 3340}, {"loss": 0.8514, "grad_norm": 0.42300888895988464, "learning_rate": 0.0002, "epoch": 0.5415892005496726, "step": 3350}, {"loss": 0.8236, "grad_norm": 0.5362542867660522, "learning_rate": 0.0002, "epoch": 0.5432058847304179, "step": 3360}, {"loss": 0.858, "grad_norm": 0.497128963470459, "learning_rate": 0.0002, "epoch": 0.5448225689111632, "step": 3370}, {"loss": 0.8519, "grad_norm": 0.5006386041641235, "learning_rate": 0.0002, "epoch": 0.5464392530919085, "step": 3380}, {"loss": 0.7867, "grad_norm": 0.44136837124824524, "learning_rate": 0.0002, "epoch": 0.5480559372726538, "step": 3390}, {"loss": 0.773, "grad_norm": 0.5897833108901978, "learning_rate": 0.0002, "epoch": 0.5496726214533991, "step": 3400}, {"loss": 0.8895, "grad_norm": 0.641075611114502, "learning_rate": 0.0002, "epoch": 0.5512893056341444, "step": 3410}, {"loss": 0.7827, "grad_norm": 0.7251322269439697, "learning_rate": 0.0002, "epoch": 0.5529059898148897, "step": 3420}, {"loss": 0.7626, "grad_norm": 0.47411349415779114, "learning_rate": 0.0002, "epoch": 0.5545226739956349, "step": 3430}, {"loss": 0.8196, "grad_norm": 0.4994310438632965, "learning_rate": 0.0002, "epoch": 0.5561393581763803, "step": 3440}, {"loss": 0.7812, "grad_norm": 0.5814438462257385, "learning_rate": 0.0002, "epoch": 0.5577560423571255, "step": 3450}, {"loss": 0.8805, "grad_norm": 0.6278898119926453, "learning_rate": 0.0002, "epoch": 0.5593727265378708, "step": 3460}, {"loss": 0.813, "grad_norm": 0.46208274364471436, "learning_rate": 0.0002, "epoch": 0.5609894107186161, "step": 3470}, {"loss": 0.8295, "grad_norm": 0.5718930959701538, "learning_rate": 0.0002, "epoch": 0.5626060948993614, "step": 3480}, {"loss": 0.8152, "grad_norm": 0.48178744316101074, "learning_rate": 0.0002, "epoch": 0.5642227790801067, "step": 3490}, {"loss": 0.8244, "grad_norm": 0.47336965799331665, "learning_rate": 0.0002, "epoch": 0.565839463260852, "step": 3500}, {"loss": 0.8099, "grad_norm": 0.43442684412002563, "learning_rate": 0.0002, "epoch": 0.5674561474415973, "step": 3510}, {"loss": 0.7564, "grad_norm": 0.6463358998298645, "learning_rate": 0.0002, "epoch": 0.5690728316223426, "step": 3520}, {"loss": 0.836, "grad_norm": 0.5286486744880676, "learning_rate": 0.0002, "epoch": 0.5706895158030879, "step": 3530}, {"loss": 0.8421, "grad_norm": 0.5405499935150146, "learning_rate": 0.0002, "epoch": 0.5723061999838331, "step": 3540}, {"loss": 0.7614, "grad_norm": 0.6654391884803772, "learning_rate": 0.0002, "epoch": 0.5739228841645785, "step": 3550}, {"loss": 0.7803, "grad_norm": 0.5081980228424072, "learning_rate": 0.0002, "epoch": 0.5755395683453237, "step": 3560}, {"loss": 0.7753, "grad_norm": 0.48978179693222046, "learning_rate": 0.0002, "epoch": 0.5771562525260691, "step": 3570}, {"loss": 0.8151, "grad_norm": 0.5840612053871155, "learning_rate": 0.0002, "epoch": 0.5787729367068143, "step": 3580}, {"loss": 0.8937, "grad_norm": 0.5235261917114258, "learning_rate": 0.0002, "epoch": 0.5803896208875596, "step": 3590}, {"loss": 0.7894, "grad_norm": 0.5672075748443604, "learning_rate": 0.0002, "epoch": 0.5820063050683049, "step": 3600}, {"loss": 0.8347, "grad_norm": 0.5613429546356201, "learning_rate": 0.0002, "epoch": 0.5836229892490502, "step": 3610}, {"loss": 0.8274, "grad_norm": 0.4032273590564728, "learning_rate": 0.0002, "epoch": 0.5852396734297954, "step": 3620}, {"loss": 0.8421, "grad_norm": 0.49559324979782104, "learning_rate": 0.0002, "epoch": 0.5868563576105408, "step": 3630}, {"loss": 0.8332, "grad_norm": 0.6895697712898254, "learning_rate": 0.0002, "epoch": 0.5884730417912861, "step": 3640}, {"loss": 0.7877, "grad_norm": 0.4750136435031891, "learning_rate": 0.0002, "epoch": 0.5900897259720314, "step": 3650}, {"loss": 0.8219, "grad_norm": 0.5176819562911987, "learning_rate": 0.0002, "epoch": 0.5917064101527767, "step": 3660}, {"loss": 0.8151, "grad_norm": 0.5817760229110718, "learning_rate": 0.0002, "epoch": 0.5933230943335219, "step": 3670}, {"loss": 0.7823, "grad_norm": 0.6064626574516296, "learning_rate": 0.0002, "epoch": 0.5949397785142673, "step": 3680}, {"loss": 0.8422, "grad_norm": 0.6728700995445251, "learning_rate": 0.0002, "epoch": 0.5965564626950125, "step": 3690}, {"loss": 0.7679, "grad_norm": 0.609305202960968, "learning_rate": 0.0002, "epoch": 0.5981731468757578, "step": 3700}, {"loss": 0.8048, "grad_norm": 0.4615488350391388, "learning_rate": 0.0002, "epoch": 0.5997898310565031, "step": 3710}, {"loss": 0.8214, "grad_norm": 2.0531179904937744, "learning_rate": 0.0002, "epoch": 0.6014065152372484, "step": 3720}, {"loss": 0.8158, "grad_norm": 0.5091132521629333, "learning_rate": 0.0002, "epoch": 0.6030231994179936, "step": 3730}, {"loss": 0.7833, "grad_norm": 0.5951124429702759, "learning_rate": 0.0002, "epoch": 0.604639883598739, "step": 3740}, {"loss": 0.7784, "grad_norm": 0.5870208144187927, "learning_rate": 0.0002, "epoch": 0.6062565677794842, "step": 3750}, {"loss": 0.8044, "grad_norm": 0.6254619359970093, "learning_rate": 0.0002, "epoch": 0.6078732519602296, "step": 3760}, {"loss": 0.7868, "grad_norm": 0.5577626824378967, "learning_rate": 0.0002, "epoch": 0.6094899361409749, "step": 3770}, {"loss": 0.8108, "grad_norm": 0.5004405379295349, "learning_rate": 0.0002, "epoch": 0.6111066203217201, "step": 3780}, {"loss": 0.8092, "grad_norm": 0.5527383685112, "learning_rate": 0.0002, "epoch": 0.6127233045024655, "step": 3790}, {"loss": 0.8036, "grad_norm": 0.49116113781929016, "learning_rate": 0.0002, "epoch": 0.6143399886832107, "step": 3800}, {"loss": 0.8352, "grad_norm": 0.5299299359321594, "learning_rate": 0.0002, "epoch": 0.6159566728639561, "step": 3810}, {"loss": 0.7737, "grad_norm": 0.464897483587265, "learning_rate": 0.0002, "epoch": 0.6175733570447013, "step": 3820}, {"loss": 0.7923, "grad_norm": 0.6505740880966187, "learning_rate": 0.0002, "epoch": 0.6191900412254466, "step": 3830}, {"loss": 0.8123, "grad_norm": 0.5512559413909912, "learning_rate": 0.0002, "epoch": 0.6208067254061919, "step": 3840}, {"loss": 0.8856, "grad_norm": 0.49427518248558044, "learning_rate": 0.0002, "epoch": 0.6224234095869372, "step": 3850}, {"loss": 0.7751, "grad_norm": 0.3839147090911865, "learning_rate": 0.0002, "epoch": 0.6240400937676824, "step": 3860}, {"loss": 0.8006, "grad_norm": 0.5760218501091003, "learning_rate": 0.0002, "epoch": 0.6256567779484278, "step": 3870}, {"loss": 0.7836, "grad_norm": 0.7226507067680359, "learning_rate": 0.0002, "epoch": 0.6272734621291731, "step": 3880}, {"loss": 0.8244, "grad_norm": 0.676781415939331, "learning_rate": 0.0002, "epoch": 0.6288901463099184, "step": 3890}, {"loss": 0.8239, "grad_norm": 0.4284018278121948, "learning_rate": 0.0002, "epoch": 0.6305068304906637, "step": 3900}, {"loss": 0.7996, "grad_norm": 0.5060628056526184, "learning_rate": 0.0002, "epoch": 0.6321235146714089, "step": 3910}, {"loss": 0.8089, "grad_norm": 0.5524522066116333, "learning_rate": 0.0002, "epoch": 0.6337401988521543, "step": 3920}, {"loss": 0.8276, "grad_norm": 0.6099881529808044, "learning_rate": 0.0002, "epoch": 0.6353568830328995, "step": 3930}, {"loss": 0.809, "grad_norm": 0.43155938386917114, "learning_rate": 0.0002, "epoch": 0.6369735672136448, "step": 3940}, {"loss": 0.8404, "grad_norm": 0.6427084803581238, "learning_rate": 0.0002, "epoch": 0.6385902513943901, "step": 3950}, {"loss": 0.8368, "grad_norm": 0.541220486164093, "learning_rate": 0.0002, "epoch": 0.6402069355751354, "step": 3960}, {"loss": 0.8539, "grad_norm": 0.5414294600486755, "learning_rate": 0.0002, "epoch": 0.6418236197558806, "step": 3970}, {"loss": 0.7996, "grad_norm": 0.46344003081321716, "learning_rate": 0.0002, "epoch": 0.643440303936626, "step": 3980}, {"loss": 0.7474, "grad_norm": 0.45209285616874695, "learning_rate": 0.0002, "epoch": 0.6450569881173712, "step": 3990}, {"loss": 0.8202, "grad_norm": 0.5417284369468689, "learning_rate": 0.0002, "epoch": 0.6466736722981166, "step": 4000}, {"loss": 0.7563, "grad_norm": 0.7995685935020447, "learning_rate": 0.0002, "epoch": 0.6482903564788619, "step": 4010}, {"loss": 0.7812, "grad_norm": 0.6384002566337585, "learning_rate": 0.0002, "epoch": 0.6499070406596071, "step": 4020}, {"loss": 0.732, "grad_norm": 0.4472815692424774, "learning_rate": 0.0002, "epoch": 0.6515237248403525, "step": 4030}, {"loss": 0.8071, "grad_norm": 0.6834294199943542, "learning_rate": 0.0002, "epoch": 0.6531404090210977, "step": 4040}, {"loss": 0.7812, "grad_norm": 0.4612339735031128, "learning_rate": 0.0002, "epoch": 0.654757093201843, "step": 4050}, {"loss": 0.8141, "grad_norm": 0.9266576170921326, "learning_rate": 0.0002, "epoch": 0.6563737773825883, "step": 4060}, {"loss": 0.7991, "grad_norm": 0.4470861852169037, "learning_rate": 0.0002, "epoch": 0.6579904615633336, "step": 4070}, {"loss": 0.8293, "grad_norm": 0.45544925332069397, "learning_rate": 0.0002, "epoch": 0.6596071457440789, "step": 4080}, {"loss": 0.8455, "grad_norm": 0.6144481301307678, "learning_rate": 0.0002, "epoch": 0.6612238299248242, "step": 4090}, {"loss": 0.7877, "grad_norm": 0.5936288237571716, "learning_rate": 0.0002, "epoch": 0.6628405141055694, "step": 4100}, {"loss": 0.7617, "grad_norm": 0.4822963774204254, "learning_rate": 0.0002, "epoch": 0.6644571982863148, "step": 4110}, {"loss": 0.7997, "grad_norm": 0.48432496190071106, "learning_rate": 0.0002, "epoch": 0.66607388246706, "step": 4120}, {"loss": 0.8404, "grad_norm": 0.4901607930660248, "learning_rate": 0.0002, "epoch": 0.6676905666478054, "step": 4130}, {"loss": 0.8085, "grad_norm": 0.5018393397331238, "learning_rate": 0.0002, "epoch": 0.6693072508285507, "step": 4140}, {"loss": 0.8065, "grad_norm": 0.6946378946304321, "learning_rate": 0.0002, "epoch": 0.6709239350092959, "step": 4150}, {"loss": 0.8147, "grad_norm": 0.5997390747070312, "learning_rate": 0.0002, "epoch": 0.6725406191900413, "step": 4160}, {"loss": 0.8268, "grad_norm": 0.6738849878311157, "learning_rate": 0.0002, "epoch": 0.6741573033707865, "step": 4170}, {"loss": 0.7704, "grad_norm": 0.6110581159591675, "learning_rate": 0.0002, "epoch": 0.6757739875515318, "step": 4180}, {"loss": 0.8043, "grad_norm": 0.5703322291374207, "learning_rate": 0.0002, "epoch": 0.6773906717322771, "step": 4190}, {"loss": 0.8099, "grad_norm": 0.4686066210269928, "learning_rate": 0.0002, "epoch": 0.6790073559130224, "step": 4200}, {"loss": 0.8441, "grad_norm": 0.6394643783569336, "learning_rate": 0.0002, "epoch": 0.6806240400937676, "step": 4210}, {"loss": 0.8011, "grad_norm": 0.5454841256141663, "learning_rate": 0.0002, "epoch": 0.682240724274513, "step": 4220}, {"loss": 0.8307, "grad_norm": 0.4859732985496521, "learning_rate": 0.0002, "epoch": 0.6838574084552582, "step": 4230}, {"loss": 0.8161, "grad_norm": 0.5544065833091736, "learning_rate": 0.0002, "epoch": 0.6854740926360036, "step": 4240}, {"loss": 0.7839, "grad_norm": 0.4902505576610565, "learning_rate": 0.0002, "epoch": 0.6870907768167488, "step": 4250}, {"loss": 0.7977, "grad_norm": 0.4768051505088806, "learning_rate": 0.0002, "epoch": 0.6887074609974941, "step": 4260}, {"loss": 0.7539, "grad_norm": 0.49982190132141113, "learning_rate": 0.0002, "epoch": 0.6903241451782395, "step": 4270}, {"loss": 0.7353, "grad_norm": 0.6351838111877441, "learning_rate": 0.0002, "epoch": 0.6919408293589847, "step": 4280}, {"loss": 0.7664, "grad_norm": 0.5647561550140381, "learning_rate": 0.0002, "epoch": 0.69355751353973, "step": 4290}, {"loss": 0.7618, "grad_norm": 0.5340486764907837, "learning_rate": 0.0002, "epoch": 0.6951741977204753, "step": 4300}, {"loss": 0.8526, "grad_norm": 0.5649092793464661, "learning_rate": 0.0002, "epoch": 0.6967908819012206, "step": 4310}, {"loss": 0.8246, "grad_norm": 0.6183916926383972, "learning_rate": 0.0002, "epoch": 0.6984075660819659, "step": 4320}, {"loss": 0.792, "grad_norm": 0.6154509782791138, "learning_rate": 0.0002, "epoch": 0.7000242502627112, "step": 4330}, {"loss": 0.8397, "grad_norm": 0.5156264305114746, "learning_rate": 0.0002, "epoch": 0.7016409344434564, "step": 4340}, {"loss": 0.8512, "grad_norm": 0.562171459197998, "learning_rate": 0.0002, "epoch": 0.7032576186242018, "step": 4350}, {"loss": 0.7882, "grad_norm": 0.4949502646923065, "learning_rate": 0.0002, "epoch": 0.704874302804947, "step": 4360}, {"loss": 0.738, "grad_norm": 0.5171684622764587, "learning_rate": 0.0002, "epoch": 0.7064909869856923, "step": 4370}, {"loss": 0.8001, "grad_norm": 0.6198443174362183, "learning_rate": 0.0002, "epoch": 0.7081076711664377, "step": 4380}, {"loss": 0.7606, "grad_norm": 0.5802276134490967, "learning_rate": 0.0002, "epoch": 0.7097243553471829, "step": 4390}, {"loss": 0.8797, "grad_norm": 0.41096967458724976, "learning_rate": 0.0002, "epoch": 0.7113410395279283, "step": 4400}, {"loss": 0.805, "grad_norm": 0.4397392272949219, "learning_rate": 0.0002, "epoch": 0.7129577237086735, "step": 4410}, {"loss": 0.7651, "grad_norm": 0.45228442549705505, "learning_rate": 0.0002, "epoch": 0.7145744078894188, "step": 4420}, {"loss": 0.7938, "grad_norm": 0.4839673936367035, "learning_rate": 0.0002, "epoch": 0.7161910920701641, "step": 4430}, {"loss": 0.8362, "grad_norm": 0.6140755414962769, "learning_rate": 0.0002, "epoch": 0.7178077762509094, "step": 4440}, {"loss": 0.7722, "grad_norm": 0.6841378808021545, "learning_rate": 0.0002, "epoch": 0.7194244604316546, "step": 4450}, {"loss": 0.8177, "grad_norm": 0.6664239168167114, "learning_rate": 0.0002, "epoch": 0.7210411446124, "step": 4460}, {"loss": 0.7983, "grad_norm": 0.47552719712257385, "learning_rate": 0.0002, "epoch": 0.7226578287931452, "step": 4470}, {"loss": 0.8982, "grad_norm": 0.6649776101112366, "learning_rate": 0.0002, "epoch": 0.7242745129738906, "step": 4480}, {"loss": 0.8074, "grad_norm": 0.5159541964530945, "learning_rate": 0.0002, "epoch": 0.7258911971546358, "step": 4490}, {"loss": 0.7786, "grad_norm": 0.6693112850189209, "learning_rate": 0.0002, "epoch": 0.7275078813353811, "step": 4500}, {"loss": 0.8655, "grad_norm": 0.48870977759361267, "learning_rate": 0.0002, "epoch": 0.7291245655161265, "step": 4510}, {"loss": 0.7337, "grad_norm": 0.4857887923717499, "learning_rate": 0.0002, "epoch": 0.7307412496968717, "step": 4520}, {"loss": 0.8026, "grad_norm": 0.5515662431716919, "learning_rate": 0.0002, "epoch": 0.732357933877617, "step": 4530}, {"loss": 0.8031, "grad_norm": 0.6292222738265991, "learning_rate": 0.0002, "epoch": 0.7339746180583623, "step": 4540}, {"loss": 0.7749, "grad_norm": 0.48265689611434937, "learning_rate": 0.0002, "epoch": 0.7355913022391076, "step": 4550}, {"loss": 0.8499, "grad_norm": 0.8044266104698181, "learning_rate": 0.0002, "epoch": 0.7372079864198529, "step": 4560}, {"loss": 0.8162, "grad_norm": 0.6111769676208496, "learning_rate": 0.0002, "epoch": 0.7388246706005982, "step": 4570}, {"loss": 0.7291, "grad_norm": 0.5229553580284119, "learning_rate": 0.0002, "epoch": 0.7404413547813434, "step": 4580}, {"loss": 0.8038, "grad_norm": 0.6054152250289917, "learning_rate": 0.0002, "epoch": 0.7420580389620888, "step": 4590}, {"loss": 0.8169, "grad_norm": 0.5574966669082642, "learning_rate": 0.0002, "epoch": 0.743674723142834, "step": 4600}, {"loss": 0.8439, "grad_norm": 0.5395817160606384, "learning_rate": 0.0002, "epoch": 0.7452914073235793, "step": 4610}, {"loss": 0.8495, "grad_norm": 0.7116472721099854, "learning_rate": 0.0002, "epoch": 0.7469080915043246, "step": 4620}, {"loss": 0.7743, "grad_norm": 0.5618700981140137, "learning_rate": 0.0002, "epoch": 0.7485247756850699, "step": 4630}, {"loss": 0.7744, "grad_norm": 0.5802770853042603, "learning_rate": 0.0002, "epoch": 0.7501414598658153, "step": 4640}, {"loss": 0.7924, "grad_norm": 0.5690428018569946, "learning_rate": 0.0002, "epoch": 0.7517581440465605, "step": 4650}, {"loss": 0.8017, "grad_norm": 0.4813360273838043, "learning_rate": 0.0002, "epoch": 0.7533748282273058, "step": 4660}, {"loss": 0.8108, "grad_norm": 0.5434042811393738, "learning_rate": 0.0002, "epoch": 0.7549915124080511, "step": 4670}, {"loss": 0.7824, "grad_norm": 0.5502099990844727, "learning_rate": 0.0002, "epoch": 0.7566081965887964, "step": 4680}, {"loss": 0.8598, "grad_norm": 0.6020621061325073, "learning_rate": 0.0002, "epoch": 0.7582248807695416, "step": 4690}, {"loss": 0.7937, "grad_norm": 0.4922301471233368, "learning_rate": 0.0002, "epoch": 0.759841564950287, "step": 4700}, {"loss": 0.788, "grad_norm": 0.6492828726768494, "learning_rate": 0.0002, "epoch": 0.7614582491310322, "step": 4710}, {"loss": 0.8313, "grad_norm": 0.4865580201148987, "learning_rate": 0.0002, "epoch": 0.7630749333117776, "step": 4720}, {"loss": 0.7966, "grad_norm": 0.5971422791481018, "learning_rate": 0.0002, "epoch": 0.7646916174925228, "step": 4730}, {"loss": 0.8298, "grad_norm": 0.6832674145698547, "learning_rate": 0.0002, "epoch": 0.7663083016732681, "step": 4740}, {"loss": 0.8156, "grad_norm": 0.500908613204956, "learning_rate": 0.0002, "epoch": 0.7679249858540134, "step": 4750}, {"loss": 0.8383, "grad_norm": 0.6112465858459473, "learning_rate": 0.0002, "epoch": 0.7695416700347587, "step": 4760}, {"loss": 0.76, "grad_norm": 0.5753506422042847, "learning_rate": 0.0002, "epoch": 0.771158354215504, "step": 4770}, {"loss": 0.8297, "grad_norm": 0.6529405117034912, "learning_rate": 0.0002, "epoch": 0.7727750383962493, "step": 4780}, {"loss": 0.8171, "grad_norm": 0.5916843414306641, "learning_rate": 0.0002, "epoch": 0.7743917225769946, "step": 4790}, {"loss": 0.83, "grad_norm": 0.4821224510669708, "learning_rate": 0.0002, "epoch": 0.7760084067577399, "step": 4800}, {"loss": 0.7703, "grad_norm": 0.5532580018043518, "learning_rate": 0.0002, "epoch": 0.7776250909384852, "step": 4810}, {"loss": 0.7363, "grad_norm": 0.4604877233505249, "learning_rate": 0.0002, "epoch": 0.7792417751192304, "step": 4820}, {"loss": 0.7506, "grad_norm": 0.5009613037109375, "learning_rate": 0.0002, "epoch": 0.7808584592999758, "step": 4830}, {"loss": 0.7863, "grad_norm": 0.6448560357093811, "learning_rate": 0.0002, "epoch": 0.782475143480721, "step": 4840}, {"loss": 0.7957, "grad_norm": 0.44327953457832336, "learning_rate": 0.0002, "epoch": 0.7840918276614663, "step": 4850}, {"loss": 0.7925, "grad_norm": 0.5355411171913147, "learning_rate": 0.0002, "epoch": 0.7857085118422116, "step": 4860}, {"loss": 0.7754, "grad_norm": 0.5635677576065063, "learning_rate": 0.0002, "epoch": 0.7873251960229569, "step": 4870}, {"loss": 0.7931, "grad_norm": 0.5417491793632507, "learning_rate": 0.0002, "epoch": 0.7889418802037023, "step": 4880}, {"loss": 0.7819, "grad_norm": 0.4567430913448334, "learning_rate": 0.0002, "epoch": 0.7905585643844475, "step": 4890}, {"loss": 0.8454, "grad_norm": 0.44651296734809875, "learning_rate": 0.0002, "epoch": 0.7921752485651928, "step": 4900}, {"loss": 0.7959, "grad_norm": 0.5741217136383057, "learning_rate": 0.0002, "epoch": 0.7937919327459381, "step": 4910}, {"loss": 0.8093, "grad_norm": 0.6605045199394226, "learning_rate": 0.0002, "epoch": 0.7954086169266834, "step": 4920}, {"loss": 0.77, "grad_norm": 0.5126531720161438, "learning_rate": 0.0002, "epoch": 0.7970253011074286, "step": 4930}, {"loss": 0.7793, "grad_norm": 0.513648271560669, "learning_rate": 0.0002, "epoch": 0.798641985288174, "step": 4940}, {"loss": 0.8314, "grad_norm": 0.5350404381752014, "learning_rate": 0.0002, "epoch": 0.8002586694689192, "step": 4950}, {"loss": 0.7649, "grad_norm": 0.5731674432754517, "learning_rate": 0.0002, "epoch": 0.8018753536496646, "step": 4960}, {"loss": 0.8572, "grad_norm": 0.5974258184432983, "learning_rate": 0.0002, "epoch": 0.8034920378304098, "step": 4970}, {"loss": 0.7972, "grad_norm": 0.8774799704551697, "learning_rate": 0.0002, "epoch": 0.8051087220111551, "step": 4980}, {"loss": 0.7899, "grad_norm": 0.5994430184364319, "learning_rate": 0.0002, "epoch": 0.8067254061919004, "step": 4990}, {"loss": 0.7736, "grad_norm": 0.4894903004169464, "learning_rate": 0.0002, "epoch": 0.8083420903726457, "step": 5000}, {"loss": 0.78, "grad_norm": 0.5218459367752075, "learning_rate": 0.0002, "epoch": 0.809958774553391, "step": 5010}, {"loss": 0.817, "grad_norm": 0.5232468843460083, "learning_rate": 0.0002, "epoch": 0.8115754587341363, "step": 5020}, {"loss": 0.7704, "grad_norm": 0.44358372688293457, "learning_rate": 0.0002, "epoch": 0.8131921429148816, "step": 5030}, {"loss": 0.785, "grad_norm": 0.6202037334442139, "learning_rate": 0.0002, "epoch": 0.8148088270956269, "step": 5040}, {"loss": 0.7351, "grad_norm": 0.7721474170684814, "learning_rate": 0.0002, "epoch": 0.8164255112763722, "step": 5050}, {"loss": 0.8297, "grad_norm": 0.5568501353263855, "learning_rate": 0.0002, "epoch": 0.8180421954571174, "step": 5060}, {"loss": 0.7733, "grad_norm": 0.49148809909820557, "learning_rate": 0.0002, "epoch": 0.8196588796378628, "step": 5070}, {"loss": 0.8054, "grad_norm": 0.4956012964248657, "learning_rate": 0.0002, "epoch": 0.821275563818608, "step": 5080}, {"loss": 0.8201, "grad_norm": 0.6078833937644958, "learning_rate": 0.0002, "epoch": 0.8228922479993533, "step": 5090}, {"loss": 0.828, "grad_norm": 0.46906954050064087, "learning_rate": 0.0002, "epoch": 0.8245089321800986, "step": 5100}, {"loss": 0.7703, "grad_norm": 0.50812166929245, "learning_rate": 0.0002, "epoch": 0.8261256163608439, "step": 5110}, {"loss": 0.8243, "grad_norm": 0.5319661498069763, "learning_rate": 0.0002, "epoch": 0.8277423005415891, "step": 5120}, {"loss": 0.7798, "grad_norm": 0.4949689209461212, "learning_rate": 0.0002, "epoch": 0.8293589847223345, "step": 5130}, {"loss": 0.7428, "grad_norm": 0.5151591300964355, "learning_rate": 0.0002, "epoch": 0.8309756689030798, "step": 5140}, {"loss": 0.8147, "grad_norm": 0.5530214309692383, "learning_rate": 0.0002, "epoch": 0.8325923530838251, "step": 5150}, {"loss": 0.8251, "grad_norm": 0.6297410130500793, "learning_rate": 0.0002, "epoch": 0.8342090372645704, "step": 5160}, {"loss": 0.8067, "grad_norm": 0.5466840267181396, "learning_rate": 0.0002, "epoch": 0.8358257214453156, "step": 5170}, {"loss": 0.7875, "grad_norm": 0.652913510799408, "learning_rate": 0.0002, "epoch": 0.837442405626061, "step": 5180}, {"loss": 0.8295, "grad_norm": 0.5811293125152588, "learning_rate": 0.0002, "epoch": 0.8390590898068062, "step": 5190}, {"loss": 0.7412, "grad_norm": 0.5109550952911377, "learning_rate": 0.0002, "epoch": 0.8406757739875516, "step": 5200}, {"loss": 0.8077, "grad_norm": 0.4551706612110138, "learning_rate": 0.0002, "epoch": 0.8422924581682968, "step": 5210}, {"loss": 0.7827, "grad_norm": 0.5813754200935364, "learning_rate": 0.0002, "epoch": 0.8439091423490421, "step": 5220}, {"loss": 0.802, "grad_norm": 0.5856947898864746, "learning_rate": 0.0002, "epoch": 0.8455258265297874, "step": 5230}, {"loss": 0.7957, "grad_norm": 0.5482739210128784, "learning_rate": 0.0002, "epoch": 0.8471425107105327, "step": 5240}, {"loss": 0.8295, "grad_norm": 0.49023720622062683, "learning_rate": 0.0002, "epoch": 0.8487591948912779, "step": 5250}, {"loss": 0.8022, "grad_norm": 0.49472475051879883, "learning_rate": 0.0002, "epoch": 0.8503758790720233, "step": 5260}, {"loss": 0.8001, "grad_norm": 0.5490226745605469, "learning_rate": 0.0002, "epoch": 0.8519925632527686, "step": 5270}, {"loss": 0.8333, "grad_norm": 0.5340665578842163, "learning_rate": 0.0002, "epoch": 0.8536092474335139, "step": 5280}, {"loss": 0.8277, "grad_norm": 0.5962483882904053, "learning_rate": 0.0002, "epoch": 0.8552259316142592, "step": 5290}, {"loss": 0.8765, "grad_norm": 0.586358368396759, "learning_rate": 0.0002, "epoch": 0.8568426157950044, "step": 5300}, {"loss": 0.7831, "grad_norm": 0.49120277166366577, "learning_rate": 0.0002, "epoch": 0.8584592999757498, "step": 5310}, {"loss": 0.8162, "grad_norm": 0.5887332558631897, "learning_rate": 0.0002, "epoch": 0.860075984156495, "step": 5320}, {"loss": 0.7464, "grad_norm": 0.42496153712272644, "learning_rate": 0.0002, "epoch": 0.8616926683372403, "step": 5330}, {"loss": 0.7905, "grad_norm": 0.5489874482154846, "learning_rate": 0.0002, "epoch": 0.8633093525179856, "step": 5340}, {"loss": 0.7958, "grad_norm": 0.5850813984870911, "learning_rate": 0.0002, "epoch": 0.8649260366987309, "step": 5350}, {"loss": 0.7642, "grad_norm": 0.517487108707428, "learning_rate": 0.0002, "epoch": 0.8665427208794761, "step": 5360}, {"loss": 0.7801, "grad_norm": 0.5339142680168152, "learning_rate": 0.0002, "epoch": 0.8681594050602215, "step": 5370}, {"loss": 0.818, "grad_norm": 0.6236387491226196, "learning_rate": 0.0002, "epoch": 0.8697760892409668, "step": 5380}, {"loss": 0.7708, "grad_norm": 0.5752192735671997, "learning_rate": 0.0002, "epoch": 0.8713927734217121, "step": 5390}, {"loss": 0.8542, "grad_norm": 0.6724614500999451, "learning_rate": 0.0002, "epoch": 0.8730094576024574, "step": 5400}, {"loss": 0.7581, "grad_norm": 0.5280613303184509, "learning_rate": 0.0002, "epoch": 0.8746261417832026, "step": 5410}, {"loss": 0.8231, "grad_norm": 0.44033288955688477, "learning_rate": 0.0002, "epoch": 0.876242825963948, "step": 5420}, {"loss": 0.8839, "grad_norm": 0.5199708342552185, "learning_rate": 0.0002, "epoch": 0.8778595101446932, "step": 5430}, {"loss": 0.7852, "grad_norm": 0.46778348088264465, "learning_rate": 0.0002, "epoch": 0.8794761943254386, "step": 5440}, {"loss": 0.7834, "grad_norm": 0.4657754898071289, "learning_rate": 0.0002, "epoch": 0.8810928785061838, "step": 5450}, {"loss": 0.7799, "grad_norm": 0.5472902655601501, "learning_rate": 0.0002, "epoch": 0.8827095626869291, "step": 5460}, {"loss": 0.8253, "grad_norm": 0.4876766800880432, "learning_rate": 0.0002, "epoch": 0.8843262468676744, "step": 5470}, {"loss": 0.7906, "grad_norm": 0.5057248473167419, "learning_rate": 0.0002, "epoch": 0.8859429310484197, "step": 5480}, {"loss": 0.8124, "grad_norm": 0.4637320637702942, "learning_rate": 0.0002, "epoch": 0.8875596152291649, "step": 5490}, {"loss": 0.781, "grad_norm": 0.471955806016922, "learning_rate": 0.0002, "epoch": 0.8891762994099103, "step": 5500}, {"loss": 0.8057, "grad_norm": 0.5209813714027405, "learning_rate": 0.0002, "epoch": 0.8907929835906556, "step": 5510}, {"loss": 0.8106, "grad_norm": 0.6213834285736084, "learning_rate": 0.0002, "epoch": 0.8924096677714008, "step": 5520}, {"loss": 0.7787, "grad_norm": 0.5215408205986023, "learning_rate": 0.0002, "epoch": 0.8940263519521462, "step": 5530}, {"loss": 0.8174, "grad_norm": 0.580478310585022, "learning_rate": 0.0002, "epoch": 0.8956430361328914, "step": 5540}, {"loss": 0.8371, "grad_norm": 0.49102169275283813, "learning_rate": 0.0002, "epoch": 0.8972597203136368, "step": 5550}, {"loss": 0.7806, "grad_norm": 0.6043479442596436, "learning_rate": 0.0002, "epoch": 0.898876404494382, "step": 5560}, {"loss": 0.7754, "grad_norm": 0.5636463165283203, "learning_rate": 0.0002, "epoch": 0.9004930886751273, "step": 5570}, {"loss": 0.8145, "grad_norm": 0.5620124340057373, "learning_rate": 0.0002, "epoch": 0.9021097728558726, "step": 5580}, {"loss": 0.8083, "grad_norm": 0.5206354856491089, "learning_rate": 0.0002, "epoch": 0.9037264570366179, "step": 5590}, {"loss": 0.8557, "grad_norm": 0.5798229575157166, "learning_rate": 0.0002, "epoch": 0.9053431412173631, "step": 5600}, {"loss": 0.8097, "grad_norm": 0.6428212523460388, "learning_rate": 0.0002, "epoch": 0.9069598253981085, "step": 5610}, {"loss": 0.7839, "grad_norm": 0.48064687848091125, "learning_rate": 0.0002, "epoch": 0.9085765095788537, "step": 5620}, {"loss": 0.8343, "grad_norm": 0.6347860097885132, "learning_rate": 0.0002, "epoch": 0.9101931937595991, "step": 5630}, {"loss": 0.851, "grad_norm": 0.5353913307189941, "learning_rate": 0.0002, "epoch": 0.9118098779403444, "step": 5640}, {"loss": 0.7736, "grad_norm": 0.5323944091796875, "learning_rate": 0.0002, "epoch": 0.9134265621210896, "step": 5650}, {"loss": 0.8393, "grad_norm": 0.5261843204498291, "learning_rate": 0.0002, "epoch": 0.915043246301835, "step": 5660}, {"loss": 0.7355, "grad_norm": 0.5451326966285706, "learning_rate": 0.0002, "epoch": 0.9166599304825802, "step": 5670}, {"loss": 0.8012, "grad_norm": 0.5183324217796326, "learning_rate": 0.0002, "epoch": 0.9182766146633256, "step": 5680}, {"loss": 0.7659, "grad_norm": 0.47229018807411194, "learning_rate": 0.0002, "epoch": 0.9198932988440708, "step": 5690}, {"loss": 0.7757, "grad_norm": 0.49180513620376587, "learning_rate": 0.0002, "epoch": 0.9215099830248161, "step": 5700}, {"loss": 0.8735, "grad_norm": 0.5419785380363464, "learning_rate": 0.0002, "epoch": 0.9231266672055614, "step": 5710}, {"loss": 0.7378, "grad_norm": 0.5408698916435242, "learning_rate": 0.0002, "epoch": 0.9247433513863067, "step": 5720}, {"loss": 0.7701, "grad_norm": 0.5286232829093933, "learning_rate": 0.0002, "epoch": 0.9263600355670519, "step": 5730}, {"loss": 0.8242, "grad_norm": 0.7539758086204529, "learning_rate": 0.0002, "epoch": 0.9279767197477973, "step": 5740}, {"loss": 0.8118, "grad_norm": 0.5166944861412048, "learning_rate": 0.0002, "epoch": 0.9295934039285425, "step": 5750}, {"loss": 0.783, "grad_norm": 0.6601425409317017, "learning_rate": 0.0002, "epoch": 0.9312100881092878, "step": 5760}, {"loss": 0.7873, "grad_norm": 0.5029960870742798, "learning_rate": 0.0002, "epoch": 0.9328267722900332, "step": 5770}, {"loss": 0.7989, "grad_norm": 0.4926645755767822, "learning_rate": 0.0002, "epoch": 0.9344434564707784, "step": 5780}, {"loss": 0.8174, "grad_norm": 0.5739615559577942, "learning_rate": 0.0002, "epoch": 0.9360601406515238, "step": 5790}, {"loss": 0.8037, "grad_norm": 0.5058279037475586, "learning_rate": 0.0002, "epoch": 0.937676824832269, "step": 5800}, {"loss": 0.8537, "grad_norm": 0.5260962247848511, "learning_rate": 0.0002, "epoch": 0.9392935090130143, "step": 5810}, {"loss": 0.7486, "grad_norm": 0.5768588185310364, "learning_rate": 0.0002, "epoch": 0.9409101931937596, "step": 5820}, {"loss": 0.8215, "grad_norm": 0.5170126557350159, "learning_rate": 0.0002, "epoch": 0.9425268773745049, "step": 5830}, {"loss": 0.7422, "grad_norm": 0.5745864510536194, "learning_rate": 0.0002, "epoch": 0.9441435615552501, "step": 5840}, {"loss": 0.7824, "grad_norm": 0.5551357865333557, "learning_rate": 0.0002, "epoch": 0.9457602457359955, "step": 5850}, {"loss": 0.8529, "grad_norm": 0.5776078701019287, "learning_rate": 0.0002, "epoch": 0.9473769299167407, "step": 5860}, {"loss": 0.8527, "grad_norm": 0.5340062379837036, "learning_rate": 0.0002, "epoch": 0.9489936140974861, "step": 5870}, {"loss": 0.8217, "grad_norm": 0.6447290182113647, "learning_rate": 0.0002, "epoch": 0.9506102982782314, "step": 5880}, {"loss": 0.7945, "grad_norm": 0.5123815536499023, "learning_rate": 0.0002, "epoch": 0.9522269824589766, "step": 5890}, {"loss": 0.8209, "grad_norm": 0.48547613620758057, "learning_rate": 0.0002, "epoch": 0.953843666639722, "step": 5900}, {"loss": 0.7896, "grad_norm": 0.5791414976119995, "learning_rate": 0.0002, "epoch": 0.9554603508204672, "step": 5910}, {"loss": 0.8408, "grad_norm": 0.6195011734962463, "learning_rate": 0.0002, "epoch": 0.9570770350012126, "step": 5920}, {"loss": 0.7805, "grad_norm": 0.6323803067207336, "learning_rate": 0.0002, "epoch": 0.9586937191819578, "step": 5930}, {"loss": 0.8484, "grad_norm": 0.45552879571914673, "learning_rate": 0.0002, "epoch": 0.9603104033627031, "step": 5940}, {"loss": 0.7367, "grad_norm": 0.5796473622322083, "learning_rate": 0.0002, "epoch": 0.9619270875434484, "step": 5950}, {"loss": 0.7672, "grad_norm": 0.647261381149292, "learning_rate": 0.0002, "epoch": 0.9635437717241937, "step": 5960}, {"loss": 0.8086, "grad_norm": 0.5487682819366455, "learning_rate": 0.0002, "epoch": 0.9651604559049389, "step": 5970}, {"loss": 0.7973, "grad_norm": 0.5743663907051086, "learning_rate": 0.0002, "epoch": 0.9667771400856843, "step": 5980}, {"loss": 0.8153, "grad_norm": 0.5470591187477112, "learning_rate": 0.0002, "epoch": 0.9683938242664295, "step": 5990}, {"loss": 0.8119, "grad_norm": 0.5901660323143005, "learning_rate": 0.0002, "epoch": 0.9700105084471748, "step": 6000}, {"loss": 0.8147, "grad_norm": 0.6544759273529053, "learning_rate": 0.0002, "epoch": 0.9716271926279202, "step": 6010}, {"loss": 0.7536, "grad_norm": 0.6288470029830933, "learning_rate": 0.0002, "epoch": 0.9732438768086654, "step": 6020}, {"loss": 0.7989, "grad_norm": 0.673153817653656, "learning_rate": 0.0002, "epoch": 0.9748605609894108, "step": 6030}, {"loss": 0.7556, "grad_norm": 0.42854753136634827, "learning_rate": 0.0002, "epoch": 0.976477245170156, "step": 6040}, {"loss": 0.8006, "grad_norm": 0.5227066278457642, "learning_rate": 0.0002, "epoch": 0.9780939293509013, "step": 6050}, {"loss": 0.795, "grad_norm": 0.5372416973114014, "learning_rate": 0.0002, "epoch": 0.9797106135316466, "step": 6060}, {"loss": 0.7591, "grad_norm": 0.6026402115821838, "learning_rate": 0.0002, "epoch": 0.9813272977123919, "step": 6070}, {"loss": 0.8347, "grad_norm": 0.49547791481018066, "learning_rate": 0.0002, "epoch": 0.9829439818931371, "step": 6080}, {"loss": 0.7722, "grad_norm": 0.4641951322555542, "learning_rate": 0.0002, "epoch": 0.9845606660738825, "step": 6090}, {"loss": 0.8125, "grad_norm": 0.5818535089492798, "learning_rate": 0.0002, "epoch": 0.9861773502546277, "step": 6100}, {"loss": 0.81, "grad_norm": 0.63955157995224, "learning_rate": 0.0002, "epoch": 0.9877940344353731, "step": 6110}, {"loss": 0.7547, "grad_norm": 0.5649438500404358, "learning_rate": 0.0002, "epoch": 0.9894107186161183, "step": 6120}, {"loss": 0.7861, "grad_norm": 0.5290433168411255, "learning_rate": 0.0002, "epoch": 0.9910274027968636, "step": 6130}, {"loss": 0.8109, "grad_norm": 0.6399374008178711, "learning_rate": 0.0002, "epoch": 0.992644086977609, "step": 6140}, {"loss": 0.8373, "grad_norm": 0.6736576557159424, "learning_rate": 0.0002, "epoch": 0.9942607711583542, "step": 6150}, {"loss": 0.7915, "grad_norm": 0.515420138835907, "learning_rate": 0.0002, "epoch": 0.9958774553390995, "step": 6160}, {"loss": 0.8032, "grad_norm": 0.562677800655365, "learning_rate": 0.0002, "epoch": 0.9974941395198448, "step": 6170}, {"loss": 0.8187, "grad_norm": 0.7113858461380005, "learning_rate": 0.0002, "epoch": 0.9991108237005901, "step": 6180}, {"eval_loss": 1.0871200561523438, "eval_runtime": 122.2071, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 0.9999191657909627, "step": 6185}, {"loss": 0.7507, "grad_norm": 0.7111801505088806, "learning_rate": 0.0002, "epoch": 1.0007275078813354, "step": 6190}, {"loss": 0.6865, "grad_norm": 0.5402125716209412, "learning_rate": 0.0002, "epoch": 1.0023441920620806, "step": 6200}, {"loss": 0.7625, "grad_norm": 0.6098830103874207, "learning_rate": 0.0002, "epoch": 1.003960876242826, "step": 6210}, {"loss": 0.7631, "grad_norm": 0.5829983353614807, "learning_rate": 0.0002, "epoch": 1.0055775604235713, "step": 6220}, {"loss": 0.7188, "grad_norm": 0.5614621043205261, "learning_rate": 0.0002, "epoch": 1.0071942446043165, "step": 6230}, {"loss": 0.7505, "grad_norm": 0.5954238772392273, "learning_rate": 0.0002, "epoch": 1.0088109287850617, "step": 6240}, {"loss": 0.7448, "grad_norm": 0.6480574607849121, "learning_rate": 0.0002, "epoch": 1.0104276129658072, "step": 6250}, {"loss": 0.7514, "grad_norm": 0.6051128506660461, "learning_rate": 0.0002, "epoch": 1.0120442971465524, "step": 6260}, {"loss": 0.7237, "grad_norm": 0.6318870782852173, "learning_rate": 0.0002, "epoch": 1.0136609813272976, "step": 6270}, {"loss": 0.7178, "grad_norm": 0.5048980116844177, "learning_rate": 0.0002, "epoch": 1.015277665508043, "step": 6280}, {"loss": 0.7391, "grad_norm": 0.6346936225891113, "learning_rate": 0.0002, "epoch": 1.0168943496887883, "step": 6290}, {"loss": 0.7486, "grad_norm": 0.5711665749549866, "learning_rate": 0.0002, "epoch": 1.0185110338695336, "step": 6300}, {"loss": 0.6808, "grad_norm": 0.5175361037254333, "learning_rate": 0.0002, "epoch": 1.0201277180502788, "step": 6310}, {"loss": 0.7539, "grad_norm": 0.5360831618309021, "learning_rate": 0.0002, "epoch": 1.0217444022310243, "step": 6320}, {"loss": 0.7112, "grad_norm": 0.614675760269165, "learning_rate": 0.0002, "epoch": 1.0233610864117695, "step": 6330}, {"loss": 0.7748, "grad_norm": 0.5626118183135986, "learning_rate": 0.0002, "epoch": 1.0249777705925147, "step": 6340}, {"loss": 0.7375, "grad_norm": 0.574897289276123, "learning_rate": 0.0002, "epoch": 1.02659445477326, "step": 6350}, {"loss": 0.759, "grad_norm": 0.7185447812080383, "learning_rate": 0.0002, "epoch": 1.0282111389540054, "step": 6360}, {"loss": 0.703, "grad_norm": 0.6705799698829651, "learning_rate": 0.0002, "epoch": 1.0298278231347506, "step": 6370}, {"loss": 0.7139, "grad_norm": 0.6740428805351257, "learning_rate": 0.0002, "epoch": 1.0314445073154959, "step": 6380}, {"loss": 0.7252, "grad_norm": 0.663902759552002, "learning_rate": 0.0002, "epoch": 1.0330611914962413, "step": 6390}, {"loss": 0.7065, "grad_norm": 0.5029543042182922, "learning_rate": 0.0002, "epoch": 1.0346778756769865, "step": 6400}, {"loss": 0.711, "grad_norm": 0.7813863158226013, "learning_rate": 0.0002, "epoch": 1.0362945598577318, "step": 6410}, {"loss": 0.7433, "grad_norm": 0.5396282076835632, "learning_rate": 0.0002, "epoch": 1.037911244038477, "step": 6420}, {"loss": 0.7222, "grad_norm": 0.5253293514251709, "learning_rate": 0.0002, "epoch": 1.0395279282192225, "step": 6430}, {"loss": 0.715, "grad_norm": 0.7236770987510681, "learning_rate": 0.0002, "epoch": 1.0411446123999677, "step": 6440}, {"loss": 0.7259, "grad_norm": 0.5670917630195618, "learning_rate": 0.0002, "epoch": 1.042761296580713, "step": 6450}, {"loss": 0.7195, "grad_norm": 0.6031978726387024, "learning_rate": 0.0002, "epoch": 1.0443779807614582, "step": 6460}, {"loss": 0.7648, "grad_norm": 0.5309213399887085, "learning_rate": 0.0002, "epoch": 1.0459946649422036, "step": 6470}, {"loss": 0.7161, "grad_norm": 0.7114651799201965, "learning_rate": 0.0002, "epoch": 1.0476113491229488, "step": 6480}, {"loss": 0.7583, "grad_norm": 0.5591610670089722, "learning_rate": 0.0002, "epoch": 1.049228033303694, "step": 6490}, {"loss": 0.6645, "grad_norm": 0.5185961127281189, "learning_rate": 0.0002, "epoch": 1.0508447174844395, "step": 6500}, {"loss": 0.7654, "grad_norm": 0.6510552167892456, "learning_rate": 0.0002, "epoch": 1.0524614016651848, "step": 6510}, {"loss": 0.7057, "grad_norm": 0.6557928919792175, "learning_rate": 0.0002, "epoch": 1.05407808584593, "step": 6520}, {"loss": 0.8056, "grad_norm": 0.6973192691802979, "learning_rate": 0.0002, "epoch": 1.0556947700266752, "step": 6530}, {"loss": 0.6793, "grad_norm": 0.6226583123207092, "learning_rate": 0.0002, "epoch": 1.0573114542074207, "step": 6540}, {"loss": 0.7151, "grad_norm": 0.5633195638656616, "learning_rate": 0.0002, "epoch": 1.058928138388166, "step": 6550}, {"loss": 0.7082, "grad_norm": 0.7466658353805542, "learning_rate": 0.0002, "epoch": 1.0605448225689111, "step": 6560}, {"loss": 0.7059, "grad_norm": 0.6462772488594055, "learning_rate": 0.0002, "epoch": 1.0621615067496564, "step": 6570}, {"loss": 0.7046, "grad_norm": 0.5266856551170349, "learning_rate": 0.0002, "epoch": 1.0637781909304018, "step": 6580}, {"loss": 0.7157, "grad_norm": 0.534392774105072, "learning_rate": 0.0002, "epoch": 1.065394875111147, "step": 6590}, {"loss": 0.7115, "grad_norm": 0.7514177560806274, "learning_rate": 0.0002, "epoch": 1.0670115592918923, "step": 6600}, {"loss": 0.7545, "grad_norm": 0.7593035697937012, "learning_rate": 0.0002, "epoch": 1.0686282434726375, "step": 6610}, {"loss": 0.6836, "grad_norm": 0.5277858972549438, "learning_rate": 0.0002, "epoch": 1.070244927653383, "step": 6620}, {"loss": 0.7405, "grad_norm": 0.5573670268058777, "learning_rate": 0.0002, "epoch": 1.0718616118341282, "step": 6630}, {"loss": 0.6774, "grad_norm": 0.6802396774291992, "learning_rate": 0.0002, "epoch": 1.0734782960148734, "step": 6640}, {"loss": 0.723, "grad_norm": 0.7367215752601624, "learning_rate": 0.0002, "epoch": 1.0750949801956189, "step": 6650}, {"loss": 0.7429, "grad_norm": 0.5961891412734985, "learning_rate": 0.0002, "epoch": 1.0767116643763641, "step": 6660}, {"loss": 0.6791, "grad_norm": 0.5736313462257385, "learning_rate": 0.0002, "epoch": 1.0783283485571094, "step": 6670}, {"loss": 0.7178, "grad_norm": 0.619219183921814, "learning_rate": 0.0002, "epoch": 1.0799450327378546, "step": 6680}, {"loss": 0.7318, "grad_norm": 0.6214390993118286, "learning_rate": 0.0002, "epoch": 1.0815617169186, "step": 6690}, {"loss": 0.7554, "grad_norm": 0.564536988735199, "learning_rate": 0.0002, "epoch": 1.0831784010993453, "step": 6700}, {"loss": 0.7362, "grad_norm": 0.5838140249252319, "learning_rate": 0.0002, "epoch": 1.0847950852800905, "step": 6710}, {"loss": 0.739, "grad_norm": 0.7000553607940674, "learning_rate": 0.0002, "epoch": 1.0864117694608357, "step": 6720}, {"loss": 0.7369, "grad_norm": 0.7078263759613037, "learning_rate": 0.0002, "epoch": 1.0880284536415812, "step": 6730}, {"loss": 0.7654, "grad_norm": 0.8353848457336426, "learning_rate": 0.0002, "epoch": 1.0896451378223264, "step": 6740}, {"loss": 0.7015, "grad_norm": 0.5615518689155579, "learning_rate": 0.0002, "epoch": 1.0912618220030716, "step": 6750}, {"loss": 0.7396, "grad_norm": 0.5475581288337708, "learning_rate": 0.0002, "epoch": 1.0928785061838169, "step": 6760}, {"loss": 0.7652, "grad_norm": 0.5835978388786316, "learning_rate": 0.0002, "epoch": 1.0944951903645623, "step": 6770}, {"loss": 0.7541, "grad_norm": 0.5516105890274048, "learning_rate": 0.0002, "epoch": 1.0961118745453076, "step": 6780}, {"loss": 0.6842, "grad_norm": 0.5875251889228821, "learning_rate": 0.0002, "epoch": 1.0977285587260528, "step": 6790}, {"loss": 0.6903, "grad_norm": 0.7376947999000549, "learning_rate": 0.0002, "epoch": 1.0993452429067982, "step": 6800}, {"loss": 0.7512, "grad_norm": 0.5656165480613708, "learning_rate": 0.0002, "epoch": 1.1009619270875435, "step": 6810}, {"loss": 0.7409, "grad_norm": 0.6365954279899597, "learning_rate": 0.0002, "epoch": 1.1025786112682887, "step": 6820}, {"loss": 0.7392, "grad_norm": 0.5033080577850342, "learning_rate": 0.0002, "epoch": 1.104195295449034, "step": 6830}, {"loss": 0.6909, "grad_norm": 0.617396891117096, "learning_rate": 0.0002, "epoch": 1.1058119796297794, "step": 6840}, {"loss": 0.7006, "grad_norm": 0.6395374536514282, "learning_rate": 0.0002, "epoch": 1.1074286638105246, "step": 6850}, {"loss": 0.7335, "grad_norm": 0.6775295734405518, "learning_rate": 0.0002, "epoch": 1.1090453479912699, "step": 6860}, {"loss": 0.764, "grad_norm": 0.6655223965644836, "learning_rate": 0.0002, "epoch": 1.1106620321720153, "step": 6870}, {"loss": 0.7553, "grad_norm": 0.676655113697052, "learning_rate": 0.0002, "epoch": 1.1122787163527605, "step": 6880}, {"loss": 0.7342, "grad_norm": 0.6062718629837036, "learning_rate": 0.0002, "epoch": 1.1138954005335058, "step": 6890}, {"loss": 0.7446, "grad_norm": 0.590943455696106, "learning_rate": 0.0002, "epoch": 1.115512084714251, "step": 6900}, {"loss": 0.6705, "grad_norm": 0.6315317153930664, "learning_rate": 0.0002, "epoch": 1.1171287688949965, "step": 6910}, {"loss": 0.6912, "grad_norm": 0.47979024052619934, "learning_rate": 0.0002, "epoch": 1.1187454530757417, "step": 6920}, {"loss": 0.7002, "grad_norm": 0.647298276424408, "learning_rate": 0.0002, "epoch": 1.120362137256487, "step": 6930}, {"loss": 0.7502, "grad_norm": 0.7336484789848328, "learning_rate": 0.0002, "epoch": 1.1219788214372322, "step": 6940}, {"loss": 0.693, "grad_norm": 0.5071424245834351, "learning_rate": 0.0002, "epoch": 1.1235955056179776, "step": 6950}, {"loss": 0.7378, "grad_norm": 0.6527144312858582, "learning_rate": 0.0002, "epoch": 1.1252121897987228, "step": 6960}, {"loss": 0.7228, "grad_norm": 0.6935935020446777, "learning_rate": 0.0002, "epoch": 1.126828873979468, "step": 6970}, {"loss": 0.699, "grad_norm": 0.8026931881904602, "learning_rate": 0.0002, "epoch": 1.1284455581602133, "step": 6980}, {"loss": 0.7361, "grad_norm": 0.5210393667221069, "learning_rate": 0.0002, "epoch": 1.1300622423409588, "step": 6990}, {"loss": 0.7456, "grad_norm": 0.60475093126297, "learning_rate": 0.0002, "epoch": 1.131678926521704, "step": 7000}, {"loss": 0.7495, "grad_norm": 0.6417073607444763, "learning_rate": 0.0002, "epoch": 1.1332956107024492, "step": 7010}, {"loss": 0.7459, "grad_norm": 0.6732175946235657, "learning_rate": 0.0002, "epoch": 1.1349122948831947, "step": 7020}, {"loss": 0.7278, "grad_norm": 0.6719491481781006, "learning_rate": 0.0002, "epoch": 1.13652897906394, "step": 7030}, {"loss": 0.7694, "grad_norm": 0.5708295106887817, "learning_rate": 0.0002, "epoch": 1.1381456632446851, "step": 7040}, {"loss": 0.7823, "grad_norm": 0.7141719460487366, "learning_rate": 0.0002, "epoch": 1.1397623474254304, "step": 7050}, {"loss": 0.764, "grad_norm": 0.6187017560005188, "learning_rate": 0.0002, "epoch": 1.1413790316061758, "step": 7060}, {"loss": 0.7657, "grad_norm": 0.50581294298172, "learning_rate": 0.0002, "epoch": 1.142995715786921, "step": 7070}, {"loss": 0.7357, "grad_norm": 0.5620143413543701, "learning_rate": 0.0002, "epoch": 1.1446123999676663, "step": 7080}, {"loss": 0.7287, "grad_norm": 0.6231929659843445, "learning_rate": 0.0002, "epoch": 1.1462290841484115, "step": 7090}, {"loss": 0.7328, "grad_norm": 0.5775774121284485, "learning_rate": 0.0002, "epoch": 1.147845768329157, "step": 7100}, {"loss": 0.7728, "grad_norm": 0.6492809653282166, "learning_rate": 0.0002, "epoch": 1.1494624525099022, "step": 7110}, {"loss": 0.7545, "grad_norm": 0.6434972286224365, "learning_rate": 0.0002, "epoch": 1.1510791366906474, "step": 7120}, {"loss": 0.7374, "grad_norm": 0.6191812753677368, "learning_rate": 0.0002, "epoch": 1.1526958208713927, "step": 7130}, {"loss": 0.7276, "grad_norm": 0.6690331697463989, "learning_rate": 0.0002, "epoch": 1.1543125050521381, "step": 7140}, {"loss": 0.7704, "grad_norm": 0.5977938175201416, "learning_rate": 0.0002, "epoch": 1.1559291892328833, "step": 7150}, {"loss": 0.7251, "grad_norm": 0.6195854544639587, "learning_rate": 0.0002, "epoch": 1.1575458734136286, "step": 7160}, {"loss": 0.7249, "grad_norm": 0.5752048492431641, "learning_rate": 0.0002, "epoch": 1.159162557594374, "step": 7170}, {"loss": 0.7593, "grad_norm": 0.589081883430481, "learning_rate": 0.0002, "epoch": 1.1607792417751193, "step": 7180}, {"loss": 0.704, "grad_norm": 0.756996750831604, "learning_rate": 0.0002, "epoch": 1.1623959259558645, "step": 7190}, {"loss": 0.7404, "grad_norm": 0.7614967226982117, "learning_rate": 0.0002, "epoch": 1.1640126101366097, "step": 7200}, {"loss": 0.7867, "grad_norm": 0.6120437979698181, "learning_rate": 0.0002, "epoch": 1.1656292943173552, "step": 7210}, {"loss": 0.7384, "grad_norm": 0.6210004687309265, "learning_rate": 0.0002, "epoch": 1.1672459784981004, "step": 7220}, {"loss": 0.7251, "grad_norm": 0.6044116020202637, "learning_rate": 0.0002, "epoch": 1.1688626626788456, "step": 7230}, {"loss": 0.7361, "grad_norm": 0.5418457388877869, "learning_rate": 0.0002, "epoch": 1.170479346859591, "step": 7240}, {"loss": 0.6938, "grad_norm": 0.6413537263870239, "learning_rate": 0.0002, "epoch": 1.1720960310403363, "step": 7250}, {"loss": 0.6978, "grad_norm": 0.5777867436408997, "learning_rate": 0.0002, "epoch": 1.1737127152210816, "step": 7260}, {"loss": 0.7503, "grad_norm": 0.7092402577400208, "learning_rate": 0.0002, "epoch": 1.1753293994018268, "step": 7270}, {"loss": 0.7487, "grad_norm": 0.6351709365844727, "learning_rate": 0.0002, "epoch": 1.176946083582572, "step": 7280}, {"loss": 0.7527, "grad_norm": 0.6172189712524414, "learning_rate": 0.0002, "epoch": 1.1785627677633175, "step": 7290}, {"loss": 0.7319, "grad_norm": 0.6801714897155762, "learning_rate": 0.0002, "epoch": 1.1801794519440627, "step": 7300}, {"loss": 0.6941, "grad_norm": 0.6044712066650391, "learning_rate": 0.0002, "epoch": 1.181796136124808, "step": 7310}, {"loss": 0.6951, "grad_norm": 0.7413212060928345, "learning_rate": 0.0002, "epoch": 1.1834128203055534, "step": 7320}, {"loss": 0.7396, "grad_norm": 0.5303856134414673, "learning_rate": 0.0002, "epoch": 1.1850295044862986, "step": 7330}, {"loss": 0.6915, "grad_norm": 0.5647098422050476, "learning_rate": 0.0002, "epoch": 1.1866461886670439, "step": 7340}, {"loss": 0.7506, "grad_norm": 0.7374135255813599, "learning_rate": 0.0002, "epoch": 1.188262872847789, "step": 7350}, {"loss": 0.7041, "grad_norm": 0.5710089206695557, "learning_rate": 0.0002, "epoch": 1.1898795570285345, "step": 7360}, {"loss": 0.8289, "grad_norm": 0.6073619723320007, "learning_rate": 0.0002, "epoch": 1.1914962412092798, "step": 7370}, {"loss": 0.7722, "grad_norm": 0.5899916887283325, "learning_rate": 0.0002, "epoch": 1.193112925390025, "step": 7380}, {"loss": 0.756, "grad_norm": 0.7762434482574463, "learning_rate": 0.0002, "epoch": 1.1947296095707705, "step": 7390}, {"loss": 0.7319, "grad_norm": 0.679949939250946, "learning_rate": 0.0002, "epoch": 1.1963462937515157, "step": 7400}, {"loss": 0.7599, "grad_norm": 0.6106849312782288, "learning_rate": 0.0002, "epoch": 1.197962977932261, "step": 7410}, {"loss": 0.7648, "grad_norm": 0.682461678981781, "learning_rate": 0.0002, "epoch": 1.1995796621130062, "step": 7420}, {"loss": 0.7741, "grad_norm": 0.6087017059326172, "learning_rate": 0.0002, "epoch": 1.2011963462937516, "step": 7430}, {"loss": 0.7642, "grad_norm": 0.63739013671875, "learning_rate": 0.0002, "epoch": 1.2028130304744968, "step": 7440}, {"loss": 0.7611, "grad_norm": 0.6154777407646179, "learning_rate": 0.0002, "epoch": 1.204429714655242, "step": 7450}, {"loss": 0.7565, "grad_norm": 0.7491534948348999, "learning_rate": 0.0002, "epoch": 1.2060463988359873, "step": 7460}, {"loss": 0.698, "grad_norm": 0.6664797067642212, "learning_rate": 0.0002, "epoch": 1.2076630830167328, "step": 7470}, {"loss": 0.7456, "grad_norm": 0.6660266518592834, "learning_rate": 0.0002, "epoch": 1.209279767197478, "step": 7480}, {"loss": 0.714, "grad_norm": 0.6972551345825195, "learning_rate": 0.0002, "epoch": 1.2108964513782232, "step": 7490}, {"loss": 0.7023, "grad_norm": 0.6157945990562439, "learning_rate": 0.0002, "epoch": 1.2125131355589684, "step": 7500}, {"loss": 0.7326, "grad_norm": 0.5199310183525085, "learning_rate": 0.0002, "epoch": 1.214129819739714, "step": 7510}, {"loss": 0.7586, "grad_norm": 0.577610433101654, "learning_rate": 0.0002, "epoch": 1.2157465039204591, "step": 7520}, {"loss": 0.7179, "grad_norm": 0.53652423620224, "learning_rate": 0.0002, "epoch": 1.2173631881012044, "step": 7530}, {"loss": 0.7393, "grad_norm": 0.6479050517082214, "learning_rate": 0.0002, "epoch": 1.2189798722819498, "step": 7540}, {"loss": 0.7534, "grad_norm": 0.618748128414154, "learning_rate": 0.0002, "epoch": 1.220596556462695, "step": 7550}, {"loss": 0.6886, "grad_norm": 0.6311424374580383, "learning_rate": 0.0002, "epoch": 1.2222132406434403, "step": 7560}, {"loss": 0.7272, "grad_norm": 0.6595825552940369, "learning_rate": 0.0002, "epoch": 1.2238299248241855, "step": 7570}, {"loss": 0.7353, "grad_norm": 0.5198960900306702, "learning_rate": 0.0002, "epoch": 1.225446609004931, "step": 7580}, {"loss": 0.674, "grad_norm": 0.578650712966919, "learning_rate": 0.0002, "epoch": 1.2270632931856762, "step": 7590}, {"loss": 0.7507, "grad_norm": 0.6080220937728882, "learning_rate": 0.0002, "epoch": 1.2286799773664214, "step": 7600}, {"loss": 0.7733, "grad_norm": 0.7050248384475708, "learning_rate": 0.0002, "epoch": 1.2302966615471669, "step": 7610}, {"loss": 0.7032, "grad_norm": 0.6652196049690247, "learning_rate": 0.0002, "epoch": 1.2319133457279121, "step": 7620}, {"loss": 0.7085, "grad_norm": 0.7322776317596436, "learning_rate": 0.0002, "epoch": 1.2335300299086573, "step": 7630}, {"loss": 0.7402, "grad_norm": 0.4998728036880493, "learning_rate": 0.0002, "epoch": 1.2351467140894026, "step": 7640}, {"loss": 0.7214, "grad_norm": 0.6428788900375366, "learning_rate": 0.0002, "epoch": 1.2367633982701478, "step": 7650}, {"loss": 0.7699, "grad_norm": 0.585242509841919, "learning_rate": 0.0002, "epoch": 1.2383800824508933, "step": 7660}, {"loss": 0.7621, "grad_norm": 0.5211917757987976, "learning_rate": 0.0002, "epoch": 1.2399967666316385, "step": 7670}, {"loss": 0.746, "grad_norm": 0.6490384340286255, "learning_rate": 0.0002, "epoch": 1.2416134508123837, "step": 7680}, {"loss": 0.7186, "grad_norm": 0.6249763369560242, "learning_rate": 0.0002, "epoch": 1.2432301349931292, "step": 7690}, {"loss": 0.7761, "grad_norm": 0.71870356798172, "learning_rate": 0.0002, "epoch": 1.2448468191738744, "step": 7700}, {"loss": 0.7525, "grad_norm": 0.6761967539787292, "learning_rate": 0.0002, "epoch": 1.2464635033546196, "step": 7710}, {"loss": 0.7501, "grad_norm": 0.6500617265701294, "learning_rate": 0.0002, "epoch": 1.2480801875353649, "step": 7720}, {"loss": 0.7903, "grad_norm": 0.8069869875907898, "learning_rate": 0.0002, "epoch": 1.2496968717161103, "step": 7730}, {"loss": 0.6747, "grad_norm": 0.6044608950614929, "learning_rate": 0.0002, "epoch": 1.2513135558968556, "step": 7740}, {"loss": 0.6825, "grad_norm": 0.6573283076286316, "learning_rate": 0.0002, "epoch": 1.2529302400776008, "step": 7750}, {"loss": 0.7617, "grad_norm": 0.625430166721344, "learning_rate": 0.0002, "epoch": 1.2545469242583462, "step": 7760}, {"loss": 0.7041, "grad_norm": 0.5442022681236267, "learning_rate": 0.0002, "epoch": 1.2561636084390915, "step": 7770}, {"loss": 0.7172, "grad_norm": 0.6818386912345886, "learning_rate": 0.0002, "epoch": 1.2577802926198367, "step": 7780}, {"loss": 0.696, "grad_norm": 0.6381874084472656, "learning_rate": 0.0002, "epoch": 1.259396976800582, "step": 7790}, {"loss": 0.6834, "grad_norm": 0.6269212961196899, "learning_rate": 0.0002, "epoch": 1.2610136609813272, "step": 7800}, {"loss": 0.7821, "grad_norm": 0.600121259689331, "learning_rate": 0.0002, "epoch": 1.2626303451620726, "step": 7810}, {"loss": 0.7761, "grad_norm": 0.6337703466415405, "learning_rate": 0.0002, "epoch": 1.2642470293428179, "step": 7820}, {"loss": 0.732, "grad_norm": 0.7234963774681091, "learning_rate": 0.0002, "epoch": 1.2658637135235633, "step": 7830}, {"loss": 0.785, "grad_norm": 0.800184965133667, "learning_rate": 0.0002, "epoch": 1.2674803977043085, "step": 7840}, {"loss": 0.7426, "grad_norm": 0.7539464831352234, "learning_rate": 0.0002, "epoch": 1.2690970818850538, "step": 7850}, {"loss": 0.7496, "grad_norm": 0.5493760704994202, "learning_rate": 0.0002, "epoch": 1.270713766065799, "step": 7860}, {"loss": 0.7537, "grad_norm": 0.7477145791053772, "learning_rate": 0.0002, "epoch": 1.2723304502465442, "step": 7870}, {"loss": 0.7573, "grad_norm": 0.6366362571716309, "learning_rate": 0.0002, "epoch": 1.2739471344272897, "step": 7880}, {"loss": 0.7608, "grad_norm": 0.7419533729553223, "learning_rate": 0.0002, "epoch": 1.275563818608035, "step": 7890}, {"loss": 0.7873, "grad_norm": 0.6141223311424255, "learning_rate": 0.0002, "epoch": 1.2771805027887801, "step": 7900}, {"loss": 0.6916, "grad_norm": 0.7522598505020142, "learning_rate": 0.0002, "epoch": 1.2787971869695256, "step": 7910}, {"loss": 0.7097, "grad_norm": 0.6935804486274719, "learning_rate": 0.0002, "epoch": 1.2804138711502708, "step": 7920}, {"loss": 0.7185, "grad_norm": 0.7239290475845337, "learning_rate": 0.0002, "epoch": 1.282030555331016, "step": 7930}, {"loss": 0.7145, "grad_norm": 0.8800187110900879, "learning_rate": 0.0002, "epoch": 1.2836472395117613, "step": 7940}, {"loss": 0.6991, "grad_norm": 0.540458083152771, "learning_rate": 0.0002, "epoch": 1.2852639236925067, "step": 7950}, {"loss": 0.7139, "grad_norm": 0.6492934226989746, "learning_rate": 0.0002, "epoch": 1.286880607873252, "step": 7960}, {"loss": 0.7742, "grad_norm": 0.6543959379196167, "learning_rate": 0.0002, "epoch": 1.2884972920539972, "step": 7970}, {"loss": 0.7316, "grad_norm": 0.5804705619812012, "learning_rate": 0.0002, "epoch": 1.2901139762347427, "step": 7980}, {"loss": 0.796, "grad_norm": 0.7074727416038513, "learning_rate": 0.0002, "epoch": 1.291730660415488, "step": 7990}, {"loss": 0.7034, "grad_norm": 0.5347974300384521, "learning_rate": 0.0002, "epoch": 1.2933473445962331, "step": 8000}, {"loss": 0.738, "grad_norm": 0.6457298398017883, "learning_rate": 0.0002, "epoch": 1.2949640287769784, "step": 8010}, {"loss": 0.7634, "grad_norm": 0.6407219171524048, "learning_rate": 0.0002, "epoch": 1.2965807129577236, "step": 8020}, {"loss": 0.7506, "grad_norm": 0.828439474105835, "learning_rate": 0.0002, "epoch": 1.298197397138469, "step": 8030}, {"loss": 0.735, "grad_norm": 0.4840380549430847, "learning_rate": 0.0002, "epoch": 1.2998140813192143, "step": 8040}, {"loss": 0.7283, "grad_norm": 0.5921024680137634, "learning_rate": 0.0002, "epoch": 1.3014307654999595, "step": 8050}, {"loss": 0.7477, "grad_norm": 0.6170315146446228, "learning_rate": 0.0002, "epoch": 1.303047449680705, "step": 8060}, {"loss": 0.7534, "grad_norm": 0.5374847054481506, "learning_rate": 0.0002, "epoch": 1.3046641338614502, "step": 8070}, {"loss": 0.7593, "grad_norm": 0.545758068561554, "learning_rate": 0.0002, "epoch": 1.3062808180421954, "step": 8080}, {"loss": 0.7463, "grad_norm": 0.55641770362854, "learning_rate": 0.0002, "epoch": 1.3078975022229407, "step": 8090}, {"loss": 0.7594, "grad_norm": 0.6724897027015686, "learning_rate": 0.0002, "epoch": 1.309514186403686, "step": 8100}, {"loss": 0.7105, "grad_norm": 0.6923972368240356, "learning_rate": 0.0002, "epoch": 1.3111308705844313, "step": 8110}, {"loss": 0.7149, "grad_norm": 0.5136841535568237, "learning_rate": 0.0002, "epoch": 1.3127475547651766, "step": 8120}, {"loss": 0.7504, "grad_norm": 0.6766283512115479, "learning_rate": 0.0002, "epoch": 1.314364238945922, "step": 8130}, {"loss": 0.7489, "grad_norm": 0.6283926367759705, "learning_rate": 0.0002, "epoch": 1.3159809231266673, "step": 8140}, {"loss": 0.7459, "grad_norm": 0.644216001033783, "learning_rate": 0.0002, "epoch": 1.3175976073074125, "step": 8150}, {"loss": 0.7125, "grad_norm": 0.7827503085136414, "learning_rate": 0.0002, "epoch": 1.3192142914881577, "step": 8160}, {"loss": 0.7271, "grad_norm": 0.6651390790939331, "learning_rate": 0.0002, "epoch": 1.320830975668903, "step": 8170}, {"loss": 0.7778, "grad_norm": 0.5547412633895874, "learning_rate": 0.0002, "epoch": 1.3224476598496484, "step": 8180}, {"loss": 0.7402, "grad_norm": 0.6765179634094238, "learning_rate": 0.0002, "epoch": 1.3240643440303936, "step": 8190}, {"loss": 0.7106, "grad_norm": 0.6822077035903931, "learning_rate": 0.0002, "epoch": 1.325681028211139, "step": 8200}, {"loss": 0.7288, "grad_norm": 0.5941002368927002, "learning_rate": 0.0002, "epoch": 1.3272977123918843, "step": 8210}, {"loss": 0.7494, "grad_norm": 0.4850037097930908, "learning_rate": 0.0002, "epoch": 1.3289143965726296, "step": 8220}, {"loss": 0.7474, "grad_norm": 0.6162990927696228, "learning_rate": 0.0002, "epoch": 1.3305310807533748, "step": 8230}, {"loss": 0.7751, "grad_norm": 0.6665613651275635, "learning_rate": 0.0002, "epoch": 1.33214776493412, "step": 8240}, {"loss": 0.759, "grad_norm": 0.618192732334137, "learning_rate": 0.0002, "epoch": 1.3337644491148655, "step": 8250}, {"loss": 0.7532, "grad_norm": 0.710418701171875, "learning_rate": 0.0002, "epoch": 1.3353811332956107, "step": 8260}, {"loss": 0.7306, "grad_norm": 0.5109876990318298, "learning_rate": 0.0002, "epoch": 1.336997817476356, "step": 8270}, {"loss": 0.7303, "grad_norm": 0.6791711449623108, "learning_rate": 0.0002, "epoch": 1.3386145016571014, "step": 8280}, {"loss": 0.7594, "grad_norm": 0.6836432814598083, "learning_rate": 0.0002, "epoch": 1.3402311858378466, "step": 8290}, {"loss": 0.7594, "grad_norm": 0.5579386353492737, "learning_rate": 0.0002, "epoch": 1.3418478700185918, "step": 8300}, {"loss": 0.7377, "grad_norm": 0.6713546514511108, "learning_rate": 0.0002, "epoch": 1.343464554199337, "step": 8310}, {"loss": 0.7756, "grad_norm": 0.5353720188140869, "learning_rate": 0.0002, "epoch": 1.3450812383800825, "step": 8320}, {"loss": 0.718, "grad_norm": 0.5813682675361633, "learning_rate": 0.0002, "epoch": 1.3466979225608278, "step": 8330}, {"loss": 0.7294, "grad_norm": 0.8158791661262512, "learning_rate": 0.0002, "epoch": 1.348314606741573, "step": 8340}, {"loss": 0.6992, "grad_norm": 0.6193785071372986, "learning_rate": 0.0002, "epoch": 1.3499312909223184, "step": 8350}, {"loss": 0.7654, "grad_norm": 0.6353939771652222, "learning_rate": 0.0002, "epoch": 1.3515479751030637, "step": 8360}, {"loss": 0.7519, "grad_norm": 0.6925048232078552, "learning_rate": 0.0002, "epoch": 1.353164659283809, "step": 8370}, {"loss": 0.736, "grad_norm": 0.988264799118042, "learning_rate": 0.0002, "epoch": 1.3547813434645541, "step": 8380}, {"loss": 0.7744, "grad_norm": 0.6476002931594849, "learning_rate": 0.0002, "epoch": 1.3563980276452994, "step": 8390}, {"loss": 0.776, "grad_norm": 0.7120398879051208, "learning_rate": 0.0002, "epoch": 1.3580147118260448, "step": 8400}, {"loss": 0.7368, "grad_norm": 0.9048416614532471, "learning_rate": 0.0002, "epoch": 1.35963139600679, "step": 8410}, {"loss": 0.7544, "grad_norm": 0.7000672817230225, "learning_rate": 0.0002, "epoch": 1.3612480801875353, "step": 8420}, {"loss": 0.7358, "grad_norm": 0.6015632152557373, "learning_rate": 0.0002, "epoch": 1.3628647643682807, "step": 8430}, {"loss": 0.7298, "grad_norm": 0.612516462802887, "learning_rate": 0.0002, "epoch": 1.364481448549026, "step": 8440}, {"loss": 0.7055, "grad_norm": 0.5969301462173462, "learning_rate": 0.0002, "epoch": 1.3660981327297712, "step": 8450}, {"loss": 0.7754, "grad_norm": 0.6730654239654541, "learning_rate": 0.0002, "epoch": 1.3677148169105164, "step": 8460}, {"loss": 0.7465, "grad_norm": 0.6386392116546631, "learning_rate": 0.0002, "epoch": 1.369331501091262, "step": 8470}, {"loss": 0.7433, "grad_norm": 0.739544153213501, "learning_rate": 0.0002, "epoch": 1.3709481852720071, "step": 8480}, {"loss": 0.7892, "grad_norm": 0.6462782621383667, "learning_rate": 0.0002, "epoch": 1.3725648694527524, "step": 8490}, {"loss": 0.7302, "grad_norm": 0.7346843481063843, "learning_rate": 0.0002, "epoch": 1.3741815536334978, "step": 8500}, {"loss": 0.7634, "grad_norm": 0.6884821057319641, "learning_rate": 0.0002, "epoch": 1.375798237814243, "step": 8510}, {"loss": 0.7614, "grad_norm": 0.6999333500862122, "learning_rate": 0.0002, "epoch": 1.3774149219949883, "step": 8520}, {"loss": 0.729, "grad_norm": 0.5378713011741638, "learning_rate": 0.0002, "epoch": 1.3790316061757335, "step": 8530}, {"loss": 0.6797, "grad_norm": 0.5417906641960144, "learning_rate": 0.0002, "epoch": 1.3806482903564787, "step": 8540}, {"loss": 0.7499, "grad_norm": 0.6602526307106018, "learning_rate": 0.0002, "epoch": 1.3822649745372242, "step": 8550}, {"loss": 0.7356, "grad_norm": 0.7073674201965332, "learning_rate": 0.0002, "epoch": 1.3838816587179694, "step": 8560}, {"loss": 0.75, "grad_norm": 0.5841707587242126, "learning_rate": 0.0002, "epoch": 1.3854983428987149, "step": 8570}, {"loss": 0.732, "grad_norm": 0.7031095027923584, "learning_rate": 0.0002, "epoch": 1.38711502707946, "step": 8580}, {"loss": 0.7464, "grad_norm": 0.5198570489883423, "learning_rate": 0.0002, "epoch": 1.3887317112602053, "step": 8590}, {"loss": 0.7354, "grad_norm": 0.7261320352554321, "learning_rate": 0.0002, "epoch": 1.3903483954409506, "step": 8600}, {"loss": 0.7339, "grad_norm": 0.5616350173950195, "learning_rate": 0.0002, "epoch": 1.3919650796216958, "step": 8610}, {"loss": 0.7382, "grad_norm": 0.5185914635658264, "learning_rate": 0.0002, "epoch": 1.3935817638024413, "step": 8620}, {"loss": 0.7456, "grad_norm": 0.5814694762229919, "learning_rate": 0.0002, "epoch": 1.3951984479831865, "step": 8630}, {"loss": 0.7413, "grad_norm": 0.6977371573448181, "learning_rate": 0.0002, "epoch": 1.3968151321639317, "step": 8640}, {"loss": 0.7574, "grad_norm": 0.6855689883232117, "learning_rate": 0.0002, "epoch": 1.3984318163446772, "step": 8650}, {"loss": 0.7802, "grad_norm": 0.5414357781410217, "learning_rate": 0.0002, "epoch": 1.4000485005254224, "step": 8660}, {"loss": 0.7487, "grad_norm": 0.6970012784004211, "learning_rate": 0.0002, "epoch": 1.4016651847061676, "step": 8670}, {"loss": 0.7421, "grad_norm": 0.526079535484314, "learning_rate": 0.0002, "epoch": 1.4032818688869129, "step": 8680}, {"loss": 0.737, "grad_norm": 0.758712887763977, "learning_rate": 0.0002, "epoch": 1.404898553067658, "step": 8690}, {"loss": 0.7612, "grad_norm": 0.7118762731552124, "learning_rate": 0.0002, "epoch": 1.4065152372484035, "step": 8700}, {"loss": 0.7628, "grad_norm": 0.5696909427642822, "learning_rate": 0.0002, "epoch": 1.4081319214291488, "step": 8710}, {"loss": 0.7156, "grad_norm": 0.7995436787605286, "learning_rate": 0.0002, "epoch": 1.4097486056098942, "step": 8720}, {"loss": 0.7521, "grad_norm": 0.7237521409988403, "learning_rate": 0.0002, "epoch": 1.4113652897906395, "step": 8730}, {"loss": 0.7661, "grad_norm": 0.744628369808197, "learning_rate": 0.0002, "epoch": 1.4129819739713847, "step": 8740}, {"loss": 0.7073, "grad_norm": 0.6082926988601685, "learning_rate": 0.0002, "epoch": 1.41459865815213, "step": 8750}, {"loss": 0.7282, "grad_norm": 0.5185243487358093, "learning_rate": 0.0002, "epoch": 1.4162153423328752, "step": 8760}, {"loss": 0.7592, "grad_norm": 0.5183082222938538, "learning_rate": 0.0002, "epoch": 1.4178320265136206, "step": 8770}, {"loss": 0.7509, "grad_norm": 0.7326041460037231, "learning_rate": 0.0002, "epoch": 1.4194487106943658, "step": 8780}, {"loss": 0.7398, "grad_norm": 0.7174660563468933, "learning_rate": 0.0002, "epoch": 1.421065394875111, "step": 8790}, {"loss": 0.7507, "grad_norm": 0.8080165982246399, "learning_rate": 0.0002, "epoch": 1.4226820790558565, "step": 8800}, {"loss": 0.72, "grad_norm": 0.5061507821083069, "learning_rate": 0.0002, "epoch": 1.4242987632366018, "step": 8810}, {"loss": 0.7563, "grad_norm": 0.801602840423584, "learning_rate": 0.0002, "epoch": 1.425915447417347, "step": 8820}, {"loss": 0.7287, "grad_norm": 0.6150273084640503, "learning_rate": 0.0002, "epoch": 1.4275321315980922, "step": 8830}, {"loss": 0.7452, "grad_norm": 0.8786525726318359, "learning_rate": 0.0002, "epoch": 1.4291488157788377, "step": 8840}, {"loss": 0.7257, "grad_norm": 0.6371538639068604, "learning_rate": 0.0002, "epoch": 1.430765499959583, "step": 8850}, {"loss": 0.711, "grad_norm": 0.6409295797348022, "learning_rate": 0.0002, "epoch": 1.4323821841403281, "step": 8860}, {"loss": 0.7891, "grad_norm": 0.6452359557151794, "learning_rate": 0.0002, "epoch": 1.4339988683210736, "step": 8870}, {"loss": 0.7588, "grad_norm": 0.5842334628105164, "learning_rate": 0.0002, "epoch": 1.4356155525018188, "step": 8880}, {"loss": 0.7446, "grad_norm": 0.696761965751648, "learning_rate": 0.0002, "epoch": 1.437232236682564, "step": 8890}, {"loss": 0.7541, "grad_norm": 0.6384600400924683, "learning_rate": 0.0002, "epoch": 1.4388489208633093, "step": 8900}, {"loss": 0.7049, "grad_norm": 0.5981136560440063, "learning_rate": 0.0002, "epoch": 1.4404656050440545, "step": 8910}, {"loss": 0.795, "grad_norm": 0.6355637907981873, "learning_rate": 0.0002, "epoch": 1.4420822892248, "step": 8920}, {"loss": 0.7653, "grad_norm": 0.6374830603599548, "learning_rate": 0.0002, "epoch": 1.4436989734055452, "step": 8930}, {"loss": 0.8108, "grad_norm": 0.559013307094574, "learning_rate": 0.0002, "epoch": 1.4453156575862904, "step": 8940}, {"loss": 0.7045, "grad_norm": 0.7289170026779175, "learning_rate": 0.0002, "epoch": 1.446932341767036, "step": 8950}, {"loss": 0.7484, "grad_norm": 0.8649206757545471, "learning_rate": 0.0002, "epoch": 1.4485490259477811, "step": 8960}, {"loss": 0.7745, "grad_norm": 0.7664689421653748, "learning_rate": 0.0002, "epoch": 1.4501657101285264, "step": 8970}, {"loss": 0.7431, "grad_norm": 0.7109952569007874, "learning_rate": 0.0002, "epoch": 1.4517823943092716, "step": 8980}, {"loss": 0.7997, "grad_norm": 0.6312844753265381, "learning_rate": 0.0002, "epoch": 1.453399078490017, "step": 8990}, {"loss": 0.7467, "grad_norm": 0.6616617441177368, "learning_rate": 0.0002, "epoch": 1.4550157626707623, "step": 9000}, {"loss": 0.7518, "grad_norm": 0.7384068965911865, "learning_rate": 0.0002, "epoch": 1.4566324468515075, "step": 9010}, {"loss": 0.7483, "grad_norm": 0.6549670100212097, "learning_rate": 0.0002, "epoch": 1.458249131032253, "step": 9020}, {"loss": 0.7423, "grad_norm": 0.6254119277000427, "learning_rate": 0.0002, "epoch": 1.4598658152129982, "step": 9030}, {"loss": 0.7645, "grad_norm": 0.6806328892707825, "learning_rate": 0.0002, "epoch": 1.4614824993937434, "step": 9040}, {"loss": 0.7221, "grad_norm": 0.6803115010261536, "learning_rate": 0.0002, "epoch": 1.4630991835744886, "step": 9050}, {"loss": 0.7264, "grad_norm": 0.48529282212257385, "learning_rate": 0.0002, "epoch": 1.4647158677552339, "step": 9060}, {"loss": 0.7542, "grad_norm": 0.5995030999183655, "learning_rate": 0.0002, "epoch": 1.4663325519359793, "step": 9070}, {"loss": 0.7894, "grad_norm": 0.6005427837371826, "learning_rate": 0.0002, "epoch": 1.4679492361167246, "step": 9080}, {"loss": 0.7288, "grad_norm": 0.718564510345459, "learning_rate": 0.0002, "epoch": 1.46956592029747, "step": 9090}, {"loss": 0.7089, "grad_norm": 0.7003577351570129, "learning_rate": 0.0002, "epoch": 1.4711826044782153, "step": 9100}, {"loss": 0.8069, "grad_norm": 0.5888323783874512, "learning_rate": 0.0002, "epoch": 1.4727992886589605, "step": 9110}, {"loss": 0.7275, "grad_norm": 0.6417609453201294, "learning_rate": 0.0002, "epoch": 1.4744159728397057, "step": 9120}, {"loss": 0.7441, "grad_norm": 0.572294294834137, "learning_rate": 0.0002, "epoch": 1.476032657020451, "step": 9130}, {"loss": 0.8053, "grad_norm": 0.8200714588165283, "learning_rate": 0.0002, "epoch": 1.4776493412011964, "step": 9140}, {"loss": 0.7382, "grad_norm": 0.6343288421630859, "learning_rate": 0.0002, "epoch": 1.4792660253819416, "step": 9150}, {"loss": 0.7641, "grad_norm": 0.7017961144447327, "learning_rate": 0.0002, "epoch": 1.4808827095626869, "step": 9160}, {"loss": 0.7619, "grad_norm": 0.6202912926673889, "learning_rate": 0.0002, "epoch": 1.4824993937434323, "step": 9170}, {"loss": 0.7428, "grad_norm": 0.6677869558334351, "learning_rate": 0.0002, "epoch": 1.4841160779241775, "step": 9180}, {"loss": 0.7648, "grad_norm": 0.6052267551422119, "learning_rate": 0.0002, "epoch": 1.4857327621049228, "step": 9190}, {"loss": 0.7152, "grad_norm": 0.6638872027397156, "learning_rate": 0.0002, "epoch": 1.487349446285668, "step": 9200}, {"loss": 0.7448, "grad_norm": 0.6245523691177368, "learning_rate": 0.0002, "epoch": 1.4889661304664135, "step": 9210}, {"loss": 0.6958, "grad_norm": 0.5761767625808716, "learning_rate": 0.0002, "epoch": 1.4905828146471587, "step": 9220}, {"loss": 0.8012, "grad_norm": 0.8175981640815735, "learning_rate": 0.0002, "epoch": 1.492199498827904, "step": 9230}, {"loss": 0.683, "grad_norm": 0.9144009947776794, "learning_rate": 0.0002, "epoch": 1.4938161830086494, "step": 9240}, {"loss": 0.7623, "grad_norm": 0.5742552876472473, "learning_rate": 0.0002, "epoch": 1.4954328671893946, "step": 9250}, {"loss": 0.7418, "grad_norm": 0.534534215927124, "learning_rate": 0.0002, "epoch": 1.4970495513701398, "step": 9260}, {"loss": 0.7194, "grad_norm": 0.7836225032806396, "learning_rate": 0.0002, "epoch": 1.498666235550885, "step": 9270}, {"loss": 0.7453, "grad_norm": 0.5292993187904358, "learning_rate": 0.0002, "epoch": 1.5002829197316303, "step": 9280}, {"loss": 0.7168, "grad_norm": 0.8044071793556213, "learning_rate": 0.0002, "epoch": 1.5018996039123758, "step": 9290}, {"loss": 0.7229, "grad_norm": 0.6185805201530457, "learning_rate": 0.0002, "epoch": 1.503516288093121, "step": 9300}, {"loss": 0.684, "grad_norm": 0.6093607544898987, "learning_rate": 0.0002, "epoch": 1.5051329722738664, "step": 9310}, {"loss": 0.7973, "grad_norm": 0.5891730189323425, "learning_rate": 0.0002, "epoch": 1.5067496564546117, "step": 9320}, {"loss": 0.7474, "grad_norm": 0.6331129670143127, "learning_rate": 0.0002, "epoch": 1.508366340635357, "step": 9330}, {"loss": 0.7074, "grad_norm": 0.7690958380699158, "learning_rate": 0.0002, "epoch": 1.5099830248161021, "step": 9340}, {"loss": 0.672, "grad_norm": 0.6548877358436584, "learning_rate": 0.0002, "epoch": 1.5115997089968474, "step": 9350}, {"loss": 0.7408, "grad_norm": 0.6545143127441406, "learning_rate": 0.0002, "epoch": 1.5132163931775926, "step": 9360}, {"loss": 0.7432, "grad_norm": 0.553247332572937, "learning_rate": 0.0002, "epoch": 1.514833077358338, "step": 9370}, {"loss": 0.7265, "grad_norm": 0.8145074844360352, "learning_rate": 0.0002, "epoch": 1.5164497615390833, "step": 9380}, {"loss": 0.7379, "grad_norm": 0.7636994123458862, "learning_rate": 0.0002, "epoch": 1.5180664457198287, "step": 9390}, {"loss": 0.7413, "grad_norm": 0.6838982701301575, "learning_rate": 0.0002, "epoch": 1.519683129900574, "step": 9400}, {"loss": 0.7367, "grad_norm": 0.8599441647529602, "learning_rate": 0.0002, "epoch": 1.5212998140813192, "step": 9410}, {"loss": 0.7663, "grad_norm": 0.7020329833030701, "learning_rate": 0.0002, "epoch": 1.5229164982620644, "step": 9420}, {"loss": 0.7928, "grad_norm": 0.6964772343635559, "learning_rate": 0.0002, "epoch": 1.5245331824428097, "step": 9430}, {"loss": 0.7168, "grad_norm": 0.6916600465774536, "learning_rate": 0.0002, "epoch": 1.5261498666235551, "step": 9440}, {"loss": 0.7519, "grad_norm": 0.7282621264457703, "learning_rate": 0.0002, "epoch": 1.5277665508043003, "step": 9450}, {"loss": 0.7628, "grad_norm": 0.5363983511924744, "learning_rate": 0.0002, "epoch": 1.5293832349850458, "step": 9460}, {"loss": 0.7154, "grad_norm": 0.6184861063957214, "learning_rate": 0.0002, "epoch": 1.530999919165791, "step": 9470}, {"loss": 0.7837, "grad_norm": 0.5991285443305969, "learning_rate": 0.0002, "epoch": 1.5326166033465363, "step": 9480}, {"loss": 0.7827, "grad_norm": 0.8176587820053101, "learning_rate": 0.0002, "epoch": 1.5342332875272815, "step": 9490}, {"loss": 0.7415, "grad_norm": 0.6473721861839294, "learning_rate": 0.0002, "epoch": 1.5358499717080267, "step": 9500}, {"loss": 0.7632, "grad_norm": 0.7319952845573425, "learning_rate": 0.0002, "epoch": 1.5374666558887722, "step": 9510}, {"loss": 0.7706, "grad_norm": 0.702900230884552, "learning_rate": 0.0002, "epoch": 1.5390833400695174, "step": 9520}, {"loss": 0.7754, "grad_norm": 0.7971600294113159, "learning_rate": 0.0002, "epoch": 1.5407000242502629, "step": 9530}, {"loss": 0.7352, "grad_norm": 0.6527525186538696, "learning_rate": 0.0002, "epoch": 1.542316708431008, "step": 9540}, {"loss": 0.7425, "grad_norm": 0.5791676044464111, "learning_rate": 0.0002, "epoch": 1.5439333926117533, "step": 9550}, {"loss": 0.7585, "grad_norm": 0.5619390606880188, "learning_rate": 0.0002, "epoch": 1.5455500767924986, "step": 9560}, {"loss": 0.7894, "grad_norm": 0.5701689124107361, "learning_rate": 0.0002, "epoch": 1.5471667609732438, "step": 9570}, {"loss": 0.793, "grad_norm": 0.47549352049827576, "learning_rate": 0.0002, "epoch": 1.548783445153989, "step": 9580}, {"loss": 0.7276, "grad_norm": 0.8730611205101013, "learning_rate": 0.0002, "epoch": 1.5504001293347345, "step": 9590}, {"loss": 0.798, "grad_norm": 0.6842091083526611, "learning_rate": 0.0002, "epoch": 1.5520168135154797, "step": 9600}, {"loss": 0.7528, "grad_norm": 0.6675129532814026, "learning_rate": 0.0002, "epoch": 1.5536334976962252, "step": 9610}, {"loss": 0.7954, "grad_norm": 0.8173956274986267, "learning_rate": 0.0002, "epoch": 1.5552501818769704, "step": 9620}, {"loss": 0.7535, "grad_norm": 0.724947452545166, "learning_rate": 0.0002, "epoch": 1.5568668660577156, "step": 9630}, {"loss": 0.7738, "grad_norm": 0.6154758930206299, "learning_rate": 0.0002, "epoch": 1.5584835502384609, "step": 9640}, {"loss": 0.7568, "grad_norm": 0.6072008013725281, "learning_rate": 0.0002, "epoch": 1.560100234419206, "step": 9650}, {"loss": 0.7219, "grad_norm": 0.659010648727417, "learning_rate": 0.0002, "epoch": 1.5617169185999515, "step": 9660}, {"loss": 0.673, "grad_norm": 0.65857994556427, "learning_rate": 0.0002, "epoch": 1.5633336027806968, "step": 9670}, {"loss": 0.7156, "grad_norm": 0.5914267301559448, "learning_rate": 0.0002, "epoch": 1.5649502869614422, "step": 9680}, {"loss": 0.7414, "grad_norm": 0.6248020529747009, "learning_rate": 0.0002, "epoch": 1.5665669711421875, "step": 9690}, {"loss": 0.694, "grad_norm": 0.7147795557975769, "learning_rate": 0.0002, "epoch": 1.5681836553229327, "step": 9700}, {"loss": 0.7335, "grad_norm": 0.7076232433319092, "learning_rate": 0.0002, "epoch": 1.569800339503678, "step": 9710}, {"loss": 0.7413, "grad_norm": 0.6217400431632996, "learning_rate": 0.0002, "epoch": 1.5714170236844232, "step": 9720}, {"loss": 0.7296, "grad_norm": 0.6709911227226257, "learning_rate": 0.0002, "epoch": 1.5730337078651684, "step": 9730}, {"loss": 0.7306, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 1.5746503920459138, "step": 9740}, {"loss": 0.7242, "grad_norm": 0.6241145730018616, "learning_rate": 0.0002, "epoch": 1.576267076226659, "step": 9750}, {"loss": 0.7384, "grad_norm": 0.4960934817790985, "learning_rate": 0.0002, "epoch": 1.5778837604074045, "step": 9760}, {"loss": 0.725, "grad_norm": 0.6593309640884399, "learning_rate": 0.0002, "epoch": 1.5795004445881498, "step": 9770}, {"loss": 0.7531, "grad_norm": 0.5814042091369629, "learning_rate": 0.0002, "epoch": 1.581117128768895, "step": 9780}, {"loss": 0.7109, "grad_norm": 0.5936070680618286, "learning_rate": 0.0002, "epoch": 1.5827338129496402, "step": 9790}, {"loss": 0.7769, "grad_norm": 0.6454403400421143, "learning_rate": 0.0002, "epoch": 1.5843504971303854, "step": 9800}, {"loss": 0.7677, "grad_norm": 0.7612107992172241, "learning_rate": 0.0002, "epoch": 1.585967181311131, "step": 9810}, {"loss": 0.7649, "grad_norm": 0.6494482755661011, "learning_rate": 0.0002, "epoch": 1.5875838654918761, "step": 9820}, {"loss": 0.7569, "grad_norm": 0.7825694680213928, "learning_rate": 0.0002, "epoch": 1.5892005496726216, "step": 9830}, {"loss": 0.706, "grad_norm": 0.6757757663726807, "learning_rate": 0.0002, "epoch": 1.5908172338533668, "step": 9840}, {"loss": 0.7803, "grad_norm": 0.7105609178543091, "learning_rate": 0.0002, "epoch": 1.592433918034112, "step": 9850}, {"loss": 0.7925, "grad_norm": 0.7596991062164307, "learning_rate": 0.0002, "epoch": 1.5940506022148573, "step": 9860}, {"loss": 0.7108, "grad_norm": 0.5681525468826294, "learning_rate": 0.0002, "epoch": 1.5956672863956025, "step": 9870}, {"loss": 0.7811, "grad_norm": 0.6090980768203735, "learning_rate": 0.0002, "epoch": 1.5972839705763477, "step": 9880}, {"loss": 0.7339, "grad_norm": 0.6271613240242004, "learning_rate": 0.0002, "epoch": 1.5989006547570932, "step": 9890}, {"loss": 0.7419, "grad_norm": 0.7656369805335999, "learning_rate": 0.0002, "epoch": 1.6005173389378387, "step": 9900}, {"loss": 0.7336, "grad_norm": 0.7504446506500244, "learning_rate": 0.0002, "epoch": 1.6021340231185839, "step": 9910}, {"loss": 0.7479, "grad_norm": 0.659656286239624, "learning_rate": 0.0002, "epoch": 1.6037507072993291, "step": 9920}, {"loss": 0.7483, "grad_norm": 0.6006826162338257, "learning_rate": 0.0002, "epoch": 1.6053673914800743, "step": 9930}, {"loss": 0.732, "grad_norm": 0.7872757911682129, "learning_rate": 0.0002, "epoch": 1.6069840756608196, "step": 9940}, {"loss": 0.768, "grad_norm": 0.5545852780342102, "learning_rate": 0.0002, "epoch": 1.6086007598415648, "step": 9950}, {"loss": 0.8064, "grad_norm": 0.7429468631744385, "learning_rate": 0.0002, "epoch": 1.6102174440223103, "step": 9960}, {"loss": 0.714, "grad_norm": 0.6873556971549988, "learning_rate": 0.0002, "epoch": 1.6118341282030555, "step": 9970}, {"loss": 0.7324, "grad_norm": 0.5874287486076355, "learning_rate": 0.0002, "epoch": 1.613450812383801, "step": 9980}, {"loss": 0.7141, "grad_norm": 0.6039386987686157, "learning_rate": 0.0002, "epoch": 1.6150674965645462, "step": 9990}, {"loss": 0.6674, "grad_norm": 0.6233575940132141, "learning_rate": 0.0002, "epoch": 1.6166841807452914, "step": 10000}, {"loss": 0.7602, "grad_norm": 0.7676448225975037, "learning_rate": 0.0002, "epoch": 1.6183008649260366, "step": 10010}, {"loss": 0.7784, "grad_norm": 0.6565698385238647, "learning_rate": 0.0002, "epoch": 1.6199175491067819, "step": 10020}, {"loss": 0.7104, "grad_norm": 0.6787590384483337, "learning_rate": 0.0002, "epoch": 1.6215342332875273, "step": 10030}, {"loss": 0.7464, "grad_norm": 0.6137678027153015, "learning_rate": 0.0002, "epoch": 1.6231509174682726, "step": 10040}, {"loss": 0.7646, "grad_norm": 0.5236800312995911, "learning_rate": 0.0002, "epoch": 1.624767601649018, "step": 10050}, {"loss": 0.7437, "grad_norm": 0.7626367807388306, "learning_rate": 0.0002, "epoch": 1.6263842858297632, "step": 10060}, {"loss": 0.7273, "grad_norm": 0.5657260417938232, "learning_rate": 0.0002, "epoch": 1.6280009700105085, "step": 10070}, {"loss": 0.7354, "grad_norm": 0.4913991391658783, "learning_rate": 0.0002, "epoch": 1.6296176541912537, "step": 10080}, {"loss": 0.7596, "grad_norm": 0.7715556621551514, "learning_rate": 0.0002, "epoch": 1.631234338371999, "step": 10090}, {"loss": 0.7105, "grad_norm": 0.6509000062942505, "learning_rate": 0.0002, "epoch": 1.6328510225527442, "step": 10100}, {"loss": 0.7274, "grad_norm": 0.6215850114822388, "learning_rate": 0.0002, "epoch": 1.6344677067334896, "step": 10110}, {"loss": 0.7705, "grad_norm": 0.6956844329833984, "learning_rate": 0.0002, "epoch": 1.6360843909142349, "step": 10120}, {"loss": 0.7129, "grad_norm": 0.6111597418785095, "learning_rate": 0.0002, "epoch": 1.6377010750949803, "step": 10130}, {"loss": 0.6955, "grad_norm": 0.6518288850784302, "learning_rate": 0.0002, "epoch": 1.6393177592757255, "step": 10140}, {"loss": 0.731, "grad_norm": 0.6914522051811218, "learning_rate": 0.0002, "epoch": 1.6409344434564708, "step": 10150}, {"loss": 0.7295, "grad_norm": 0.63785719871521, "learning_rate": 0.0002, "epoch": 1.642551127637216, "step": 10160}, {"loss": 0.7355, "grad_norm": 0.6379287838935852, "learning_rate": 0.0002, "epoch": 1.6441678118179612, "step": 10170}, {"loss": 0.7359, "grad_norm": 0.6793403029441833, "learning_rate": 0.0002, "epoch": 1.6457844959987067, "step": 10180}, {"loss": 0.7402, "grad_norm": 0.6099132895469666, "learning_rate": 0.0002, "epoch": 1.647401180179452, "step": 10190}, {"loss": 0.7353, "grad_norm": 0.5869854092597961, "learning_rate": 0.0002, "epoch": 1.6490178643601974, "step": 10200}, {"loss": 0.8308, "grad_norm": 0.7716999053955078, "learning_rate": 0.0002, "epoch": 1.6506345485409426, "step": 10210}, {"loss": 0.7215, "grad_norm": 0.6854110360145569, "learning_rate": 0.0002, "epoch": 1.6522512327216878, "step": 10220}, {"loss": 0.782, "grad_norm": 0.6957170367240906, "learning_rate": 0.0002, "epoch": 1.653867916902433, "step": 10230}, {"loss": 0.7282, "grad_norm": 0.6932903528213501, "learning_rate": 0.0002, "epoch": 1.6554846010831783, "step": 10240}, {"loss": 0.7478, "grad_norm": 0.7713165283203125, "learning_rate": 0.0002, "epoch": 1.6571012852639235, "step": 10250}, {"loss": 0.7099, "grad_norm": 0.7455793619155884, "learning_rate": 0.0002, "epoch": 1.658717969444669, "step": 10260}, {"loss": 0.7524, "grad_norm": 0.5464168190956116, "learning_rate": 0.0002, "epoch": 1.6603346536254144, "step": 10270}, {"loss": 0.7328, "grad_norm": 0.6782926321029663, "learning_rate": 0.0002, "epoch": 1.6619513378061597, "step": 10280}, {"loss": 0.7801, "grad_norm": 0.7962649464607239, "learning_rate": 0.0002, "epoch": 1.663568021986905, "step": 10290}, {"loss": 0.7142, "grad_norm": 0.6814526319503784, "learning_rate": 0.0002, "epoch": 1.6651847061676501, "step": 10300}, {"loss": 0.7285, "grad_norm": 0.656895101070404, "learning_rate": 0.0002, "epoch": 1.6668013903483954, "step": 10310}, {"loss": 0.7358, "grad_norm": 0.6085672378540039, "learning_rate": 0.0002, "epoch": 1.6684180745291406, "step": 10320}, {"loss": 0.7074, "grad_norm": 0.585508406162262, "learning_rate": 0.0002, "epoch": 1.670034758709886, "step": 10330}, {"loss": 0.7604, "grad_norm": 0.6930184364318848, "learning_rate": 0.0002, "epoch": 1.6716514428906313, "step": 10340}, {"loss": 0.7169, "grad_norm": 0.575663149356842, "learning_rate": 0.0002, "epoch": 1.6732681270713767, "step": 10350}, {"loss": 0.7198, "grad_norm": 0.582502543926239, "learning_rate": 0.0002, "epoch": 1.674884811252122, "step": 10360}, {"loss": 0.7793, "grad_norm": 0.5668916702270508, "learning_rate": 0.0002, "epoch": 1.6765014954328672, "step": 10370}, {"loss": 0.7478, "grad_norm": 0.6070065498352051, "learning_rate": 0.0002, "epoch": 1.6781181796136124, "step": 10380}, {"loss": 0.7939, "grad_norm": 0.6141316294670105, "learning_rate": 0.0002, "epoch": 1.6797348637943577, "step": 10390}, {"loss": 0.7573, "grad_norm": 0.8359124064445496, "learning_rate": 0.0002, "epoch": 1.6813515479751031, "step": 10400}, {"loss": 0.7488, "grad_norm": 0.5378185510635376, "learning_rate": 0.0002, "epoch": 1.6829682321558483, "step": 10410}, {"loss": 0.7588, "grad_norm": 0.6959536075592041, "learning_rate": 0.0002, "epoch": 1.6845849163365938, "step": 10420}, {"loss": 0.7872, "grad_norm": 0.6514357328414917, "learning_rate": 0.0002, "epoch": 1.686201600517339, "step": 10430}, {"loss": 0.725, "grad_norm": 0.7706646919250488, "learning_rate": 0.0002, "epoch": 1.6878182846980843, "step": 10440}, {"loss": 0.7673, "grad_norm": 0.6183337569236755, "learning_rate": 0.0002, "epoch": 1.6894349688788295, "step": 10450}, {"loss": 0.7566, "grad_norm": 0.6123278141021729, "learning_rate": 0.0002, "epoch": 1.6910516530595747, "step": 10460}, {"loss": 0.7169, "grad_norm": 0.6894851326942444, "learning_rate": 0.0002, "epoch": 1.69266833724032, "step": 10470}, {"loss": 0.7435, "grad_norm": 0.7497312426567078, "learning_rate": 0.0002, "epoch": 1.6942850214210654, "step": 10480}, {"loss": 0.7544, "grad_norm": 0.5968214273452759, "learning_rate": 0.0002, "epoch": 1.6959017056018106, "step": 10490}, {"loss": 0.6793, "grad_norm": 0.6747927069664001, "learning_rate": 0.0002, "epoch": 1.697518389782556, "step": 10500}, {"loss": 0.7415, "grad_norm": 0.5708310008049011, "learning_rate": 0.0002, "epoch": 1.6991350739633013, "step": 10510}, {"loss": 0.7385, "grad_norm": 0.606526792049408, "learning_rate": 0.0002, "epoch": 1.7007517581440466, "step": 10520}, {"loss": 0.7204, "grad_norm": 0.662011981010437, "learning_rate": 0.0002, "epoch": 1.7023684423247918, "step": 10530}, {"loss": 0.7999, "grad_norm": 0.7583045363426208, "learning_rate": 0.0002, "epoch": 1.703985126505537, "step": 10540}, {"loss": 0.7563, "grad_norm": 0.721632182598114, "learning_rate": 0.0002, "epoch": 1.7056018106862825, "step": 10550}, {"loss": 0.7407, "grad_norm": 0.6107715368270874, "learning_rate": 0.0002, "epoch": 1.7072184948670277, "step": 10560}, {"loss": 0.7519, "grad_norm": 0.6652471423149109, "learning_rate": 0.0002, "epoch": 1.7088351790477732, "step": 10570}, {"loss": 0.7767, "grad_norm": 0.6308087110519409, "learning_rate": 0.0002, "epoch": 1.7104518632285184, "step": 10580}, {"loss": 0.7659, "grad_norm": 0.5464386940002441, "learning_rate": 0.0002, "epoch": 1.7120685474092636, "step": 10590}, {"loss": 0.7063, "grad_norm": 0.6558911204338074, "learning_rate": 0.0002, "epoch": 1.7136852315900089, "step": 10600}, {"loss": 0.7126, "grad_norm": 0.5665024518966675, "learning_rate": 0.0002, "epoch": 1.715301915770754, "step": 10610}, {"loss": 0.6958, "grad_norm": 0.7888094186782837, "learning_rate": 0.0002, "epoch": 1.7169185999514993, "step": 10620}, {"loss": 0.7785, "grad_norm": 0.7084909081459045, "learning_rate": 0.0002, "epoch": 1.7185352841322448, "step": 10630}, {"loss": 0.7557, "grad_norm": 0.7982324361801147, "learning_rate": 0.0002, "epoch": 1.7201519683129902, "step": 10640}, {"loss": 0.7345, "grad_norm": 0.6418732404708862, "learning_rate": 0.0002, "epoch": 1.7217686524937355, "step": 10650}, {"loss": 0.7734, "grad_norm": 0.7636681795120239, "learning_rate": 0.0002, "epoch": 1.7233853366744807, "step": 10660}, {"loss": 0.7541, "grad_norm": 0.5646875500679016, "learning_rate": 0.0002, "epoch": 1.725002020855226, "step": 10670}, {"loss": 0.7642, "grad_norm": 0.5231260657310486, "learning_rate": 0.0002, "epoch": 1.7266187050359711, "step": 10680}, {"loss": 0.7846, "grad_norm": 0.7635011672973633, "learning_rate": 0.0002, "epoch": 1.7282353892167164, "step": 10690}, {"loss": 0.7471, "grad_norm": 0.7518259286880493, "learning_rate": 0.0002, "epoch": 1.7298520733974618, "step": 10700}, {"loss": 0.751, "grad_norm": 0.7295602560043335, "learning_rate": 0.0002, "epoch": 1.731468757578207, "step": 10710}, {"loss": 0.731, "grad_norm": 0.6984632015228271, "learning_rate": 0.0002, "epoch": 1.7330854417589525, "step": 10720}, {"loss": 0.7921, "grad_norm": 0.6198219060897827, "learning_rate": 0.0002, "epoch": 1.7347021259396977, "step": 10730}, {"loss": 0.7642, "grad_norm": 0.6957576274871826, "learning_rate": 0.0002, "epoch": 1.736318810120443, "step": 10740}, {"loss": 0.7917, "grad_norm": 0.6430263519287109, "learning_rate": 0.0002, "epoch": 1.7379354943011882, "step": 10750}, {"loss": 0.7156, "grad_norm": 0.6134995222091675, "learning_rate": 0.0002, "epoch": 1.7395521784819334, "step": 10760}, {"loss": 0.7584, "grad_norm": 0.7209452986717224, "learning_rate": 0.0002, "epoch": 1.741168862662679, "step": 10770}, {"loss": 0.7528, "grad_norm": 0.6735447645187378, "learning_rate": 0.0002, "epoch": 1.7427855468434241, "step": 10780}, {"loss": 0.756, "grad_norm": 0.5605693459510803, "learning_rate": 0.0002, "epoch": 1.7444022310241696, "step": 10790}, {"loss": 0.7759, "grad_norm": 0.6882363557815552, "learning_rate": 0.0002, "epoch": 1.7460189152049148, "step": 10800}, {"loss": 0.7544, "grad_norm": 0.6386259198188782, "learning_rate": 0.0002, "epoch": 1.74763559938566, "step": 10810}, {"loss": 0.7697, "grad_norm": 0.6529015302658081, "learning_rate": 0.0002, "epoch": 1.7492522835664053, "step": 10820}, {"loss": 0.7219, "grad_norm": 0.5664082765579224, "learning_rate": 0.0002, "epoch": 1.7508689677471505, "step": 10830}, {"loss": 0.7586, "grad_norm": 0.7532684206962585, "learning_rate": 0.0002, "epoch": 1.7524856519278957, "step": 10840}, {"loss": 0.6919, "grad_norm": 0.77171391248703, "learning_rate": 0.0002, "epoch": 1.7541023361086412, "step": 10850}, {"loss": 0.785, "grad_norm": 0.7255431413650513, "learning_rate": 0.0002, "epoch": 1.7557190202893864, "step": 10860}, {"loss": 0.7458, "grad_norm": 0.763083279132843, "learning_rate": 0.0002, "epoch": 1.7573357044701319, "step": 10870}, {"loss": 0.7846, "grad_norm": 0.6042402982711792, "learning_rate": 0.0002, "epoch": 1.758952388650877, "step": 10880}, {"loss": 0.7027, "grad_norm": 0.7642518281936646, "learning_rate": 0.0002, "epoch": 1.7605690728316223, "step": 10890}, {"loss": 0.746, "grad_norm": 0.6347904801368713, "learning_rate": 0.0002, "epoch": 1.7621857570123676, "step": 10900}, {"loss": 0.7458, "grad_norm": 0.5371627807617188, "learning_rate": 0.0002, "epoch": 1.7638024411931128, "step": 10910}, {"loss": 0.7466, "grad_norm": 0.6840225458145142, "learning_rate": 0.0002, "epoch": 1.7654191253738583, "step": 10920}, {"loss": 0.725, "grad_norm": 0.5288469195365906, "learning_rate": 0.0002, "epoch": 1.7670358095546035, "step": 10930}, {"loss": 0.7863, "grad_norm": 0.69020676612854, "learning_rate": 0.0002, "epoch": 1.768652493735349, "step": 10940}, {"loss": 0.7468, "grad_norm": 0.5943242311477661, "learning_rate": 0.0002, "epoch": 1.7702691779160942, "step": 10950}, {"loss": 0.7244, "grad_norm": 0.5616418123245239, "learning_rate": 0.0002, "epoch": 1.7718858620968394, "step": 10960}, {"loss": 0.7137, "grad_norm": 0.7209470868110657, "learning_rate": 0.0002, "epoch": 1.7735025462775846, "step": 10970}, {"loss": 0.7459, "grad_norm": 0.6657957434654236, "learning_rate": 0.0002, "epoch": 1.7751192304583299, "step": 10980}, {"loss": 0.7076, "grad_norm": 0.6469064950942993, "learning_rate": 0.0002, "epoch": 1.776735914639075, "step": 10990}, {"loss": 0.7321, "grad_norm": 0.6615678071975708, "learning_rate": 0.0002, "epoch": 1.7783525988198206, "step": 11000}, {"loss": 0.747, "grad_norm": 0.6722439527511597, "learning_rate": 0.0002, "epoch": 1.779969283000566, "step": 11010}, {"loss": 0.7302, "grad_norm": 0.634136974811554, "learning_rate": 0.0002, "epoch": 1.7815859671813112, "step": 11020}, {"loss": 0.8105, "grad_norm": 0.6024377346038818, "learning_rate": 0.0002, "epoch": 1.7832026513620565, "step": 11030}, {"loss": 0.7855, "grad_norm": 0.6909403800964355, "learning_rate": 0.0002, "epoch": 1.7848193355428017, "step": 11040}, {"loss": 0.7471, "grad_norm": 0.7148767709732056, "learning_rate": 0.0002, "epoch": 1.786436019723547, "step": 11050}, {"loss": 0.7145, "grad_norm": 0.7442979216575623, "learning_rate": 0.0002, "epoch": 1.7880527039042922, "step": 11060}, {"loss": 0.7215, "grad_norm": 0.6830431818962097, "learning_rate": 0.0002, "epoch": 1.7896693880850376, "step": 11070}, {"loss": 0.7625, "grad_norm": 0.9172667264938354, "learning_rate": 0.0002, "epoch": 1.7912860722657828, "step": 11080}, {"loss": 0.76, "grad_norm": 0.6799490451812744, "learning_rate": 0.0002, "epoch": 1.7929027564465283, "step": 11090}, {"loss": 0.7716, "grad_norm": 0.7617024779319763, "learning_rate": 0.0002, "epoch": 1.7945194406272735, "step": 11100}, {"loss": 0.7586, "grad_norm": 0.7701810002326965, "learning_rate": 0.0002, "epoch": 1.7961361248080188, "step": 11110}, {"loss": 0.7843, "grad_norm": 0.7454385757446289, "learning_rate": 0.0002, "epoch": 1.797752808988764, "step": 11120}, {"loss": 0.7873, "grad_norm": 0.6121436953544617, "learning_rate": 0.0002, "epoch": 1.7993694931695092, "step": 11130}, {"loss": 0.7305, "grad_norm": 0.6237571835517883, "learning_rate": 0.0002, "epoch": 1.8009861773502547, "step": 11140}, {"loss": 0.6827, "grad_norm": 0.6818515658378601, "learning_rate": 0.0002, "epoch": 1.802602861531, "step": 11150}, {"loss": 0.6876, "grad_norm": 0.7768308520317078, "learning_rate": 0.0002, "epoch": 1.8042195457117454, "step": 11160}, {"loss": 0.7533, "grad_norm": 0.6875537633895874, "learning_rate": 0.0002, "epoch": 1.8058362298924906, "step": 11170}, {"loss": 0.761, "grad_norm": 0.7950584888458252, "learning_rate": 0.0002, "epoch": 1.8074529140732358, "step": 11180}, {"loss": 0.7623, "grad_norm": 0.8210248351097107, "learning_rate": 0.0002, "epoch": 1.809069598253981, "step": 11190}, {"loss": 0.7556, "grad_norm": 0.6674110889434814, "learning_rate": 0.0002, "epoch": 1.8106862824347263, "step": 11200}, {"loss": 0.7663, "grad_norm": 0.6261674761772156, "learning_rate": 0.0002, "epoch": 1.8123029666154715, "step": 11210}, {"loss": 0.7122, "grad_norm": 0.6484741568565369, "learning_rate": 0.0002, "epoch": 1.813919650796217, "step": 11220}, {"loss": 0.7718, "grad_norm": 0.6231244206428528, "learning_rate": 0.0002, "epoch": 1.8155363349769622, "step": 11230}, {"loss": 0.7152, "grad_norm": 0.7243146896362305, "learning_rate": 0.0002, "epoch": 1.8171530191577077, "step": 11240}, {"loss": 0.7448, "grad_norm": 0.6776193380355835, "learning_rate": 0.0002, "epoch": 1.818769703338453, "step": 11250}, {"loss": 0.7317, "grad_norm": 0.5973618030548096, "learning_rate": 0.0002, "epoch": 1.8203863875191981, "step": 11260}, {"loss": 0.7961, "grad_norm": 0.6451361179351807, "learning_rate": 0.0002, "epoch": 1.8220030716999434, "step": 11270}, {"loss": 0.7611, "grad_norm": 0.5963068008422852, "learning_rate": 0.0002, "epoch": 1.8236197558806886, "step": 11280}, {"loss": 0.7466, "grad_norm": 0.536902129650116, "learning_rate": 0.0002, "epoch": 1.825236440061434, "step": 11290}, {"loss": 0.708, "grad_norm": 0.6993787288665771, "learning_rate": 0.0002, "epoch": 1.8268531242421793, "step": 11300}, {"loss": 0.7153, "grad_norm": 0.6135255098342896, "learning_rate": 0.0002, "epoch": 1.8284698084229247, "step": 11310}, {"loss": 0.7423, "grad_norm": 0.6057423949241638, "learning_rate": 0.0002, "epoch": 1.83008649260367, "step": 11320}, {"loss": 0.735, "grad_norm": 0.6598812341690063, "learning_rate": 0.0002, "epoch": 1.8317031767844152, "step": 11330}, {"loss": 0.7278, "grad_norm": 0.6075948476791382, "learning_rate": 0.0002, "epoch": 1.8333198609651604, "step": 11340}, {"loss": 0.7846, "grad_norm": 0.7065447568893433, "learning_rate": 0.0002, "epoch": 1.8349365451459057, "step": 11350}, {"loss": 0.7365, "grad_norm": 0.680526614189148, "learning_rate": 0.0002, "epoch": 1.8365532293266509, "step": 11360}, {"loss": 0.7152, "grad_norm": 0.6356695294380188, "learning_rate": 0.0002, "epoch": 1.8381699135073963, "step": 11370}, {"loss": 0.721, "grad_norm": 0.6399052143096924, "learning_rate": 0.0002, "epoch": 1.8397865976881416, "step": 11380}, {"loss": 0.7618, "grad_norm": 0.6125704050064087, "learning_rate": 0.0002, "epoch": 1.841403281868887, "step": 11390}, {"loss": 0.755, "grad_norm": 0.7124643325805664, "learning_rate": 0.0002, "epoch": 1.8430199660496323, "step": 11400}, {"loss": 0.7972, "grad_norm": 0.6099604964256287, "learning_rate": 0.0002, "epoch": 1.8446366502303775, "step": 11410}, {"loss": 0.7187, "grad_norm": 0.7338208556175232, "learning_rate": 0.0002, "epoch": 1.8462533344111227, "step": 11420}, {"loss": 0.7007, "grad_norm": 0.7534668445587158, "learning_rate": 0.0002, "epoch": 1.847870018591868, "step": 11430}, {"loss": 0.7464, "grad_norm": 0.6135470271110535, "learning_rate": 0.0002, "epoch": 1.8494867027726134, "step": 11440}, {"loss": 0.7955, "grad_norm": 0.6229309439659119, "learning_rate": 0.0002, "epoch": 1.8511033869533586, "step": 11450}, {"loss": 0.7594, "grad_norm": 0.706423282623291, "learning_rate": 0.0002, "epoch": 1.852720071134104, "step": 11460}, {"loss": 0.7411, "grad_norm": 0.5460049510002136, "learning_rate": 0.0002, "epoch": 1.8543367553148493, "step": 11470}, {"loss": 0.7416, "grad_norm": 0.6616711020469666, "learning_rate": 0.0002, "epoch": 1.8559534394955945, "step": 11480}, {"loss": 0.729, "grad_norm": 0.6372783184051514, "learning_rate": 0.0002, "epoch": 1.8575701236763398, "step": 11490}, {"loss": 0.7333, "grad_norm": 0.7162668108940125, "learning_rate": 0.0002, "epoch": 1.859186807857085, "step": 11500}, {"loss": 0.7747, "grad_norm": 0.6605209708213806, "learning_rate": 0.0002, "epoch": 1.8608034920378305, "step": 11510}, {"loss": 0.7258, "grad_norm": 0.6933956742286682, "learning_rate": 0.0002, "epoch": 1.8624201762185757, "step": 11520}, {"loss": 0.7243, "grad_norm": 0.6582090854644775, "learning_rate": 0.0002, "epoch": 1.8640368603993211, "step": 11530}, {"loss": 0.7313, "grad_norm": 0.6416500806808472, "learning_rate": 0.0002, "epoch": 1.8656535445800664, "step": 11540}, {"loss": 0.7372, "grad_norm": 0.5434312224388123, "learning_rate": 0.0002, "epoch": 1.8672702287608116, "step": 11550}, {"loss": 0.7635, "grad_norm": 0.6827567219734192, "learning_rate": 0.0002, "epoch": 1.8688869129415568, "step": 11560}, {"loss": 0.7137, "grad_norm": 0.7354370951652527, "learning_rate": 0.0002, "epoch": 1.870503597122302, "step": 11570}, {"loss": 0.7526, "grad_norm": 0.590372622013092, "learning_rate": 0.0002, "epoch": 1.8721202813030473, "step": 11580}, {"loss": 0.731, "grad_norm": 0.853183925151825, "learning_rate": 0.0002, "epoch": 1.8737369654837928, "step": 11590}, {"loss": 0.7487, "grad_norm": 0.822678804397583, "learning_rate": 0.0002, "epoch": 1.875353649664538, "step": 11600}, {"loss": 0.7427, "grad_norm": 0.6591550707817078, "learning_rate": 0.0002, "epoch": 1.8769703338452834, "step": 11610}, {"loss": 0.7054, "grad_norm": 0.7475301623344421, "learning_rate": 0.0002, "epoch": 1.8785870180260287, "step": 11620}, {"loss": 0.811, "grad_norm": 0.6390765309333801, "learning_rate": 0.0002, "epoch": 1.880203702206774, "step": 11630}, {"loss": 0.7531, "grad_norm": 0.6589758992195129, "learning_rate": 0.0002, "epoch": 1.8818203863875191, "step": 11640}, {"loss": 0.7475, "grad_norm": 0.6765508651733398, "learning_rate": 0.0002, "epoch": 1.8834370705682644, "step": 11650}, {"loss": 0.738, "grad_norm": 0.6527857780456543, "learning_rate": 0.0002, "epoch": 1.8850537547490098, "step": 11660}, {"loss": 0.7504, "grad_norm": 0.6642923951148987, "learning_rate": 0.0002, "epoch": 1.886670438929755, "step": 11670}, {"loss": 0.7701, "grad_norm": 0.6945584416389465, "learning_rate": 0.0002, "epoch": 1.8882871231105005, "step": 11680}, {"loss": 0.7711, "grad_norm": 0.694018542766571, "learning_rate": 0.0002, "epoch": 1.8899038072912457, "step": 11690}, {"loss": 0.7195, "grad_norm": 0.7237417101860046, "learning_rate": 0.0002, "epoch": 1.891520491471991, "step": 11700}, {"loss": 0.7491, "grad_norm": 0.7401309609413147, "learning_rate": 0.0002, "epoch": 1.8931371756527362, "step": 11710}, {"loss": 0.805, "grad_norm": 0.6537784337997437, "learning_rate": 0.0002, "epoch": 1.8947538598334814, "step": 11720}, {"loss": 0.793, "grad_norm": 0.7398539185523987, "learning_rate": 0.0002, "epoch": 1.8963705440142267, "step": 11730}, {"loss": 0.7561, "grad_norm": 0.6696075797080994, "learning_rate": 0.0002, "epoch": 1.8979872281949721, "step": 11740}, {"loss": 0.7353, "grad_norm": 0.6014142036437988, "learning_rate": 0.0002, "epoch": 1.8996039123757174, "step": 11750}, {"loss": 0.7714, "grad_norm": 0.7023524641990662, "learning_rate": 0.0002, "epoch": 1.9012205965564628, "step": 11760}, {"loss": 0.7088, "grad_norm": 0.739973783493042, "learning_rate": 0.0002, "epoch": 1.902837280737208, "step": 11770}, {"loss": 0.7848, "grad_norm": 0.5576770901679993, "learning_rate": 0.0002, "epoch": 1.9044539649179533, "step": 11780}, {"loss": 0.7483, "grad_norm": 0.6907393932342529, "learning_rate": 0.0002, "epoch": 1.9060706490986985, "step": 11790}, {"loss": 0.7827, "grad_norm": 0.6934581995010376, "learning_rate": 0.0002, "epoch": 1.9076873332794437, "step": 11800}, {"loss": 0.7199, "grad_norm": 0.591774582862854, "learning_rate": 0.0002, "epoch": 1.9093040174601892, "step": 11810}, {"loss": 0.7333, "grad_norm": 0.6249791383743286, "learning_rate": 0.0002, "epoch": 1.9109207016409344, "step": 11820}, {"loss": 0.7581, "grad_norm": 0.6755744218826294, "learning_rate": 0.0002, "epoch": 1.9125373858216799, "step": 11830}, {"loss": 0.696, "grad_norm": 0.7286285161972046, "learning_rate": 0.0002, "epoch": 1.914154070002425, "step": 11840}, {"loss": 0.7509, "grad_norm": 0.7867850065231323, "learning_rate": 0.0002, "epoch": 1.9157707541831703, "step": 11850}, {"loss": 0.735, "grad_norm": 0.6283972859382629, "learning_rate": 0.0002, "epoch": 1.9173874383639156, "step": 11860}, {"loss": 0.7296, "grad_norm": 0.605823814868927, "learning_rate": 0.0002, "epoch": 1.9190041225446608, "step": 11870}, {"loss": 0.6598, "grad_norm": 0.5927976965904236, "learning_rate": 0.0002, "epoch": 1.920620806725406, "step": 11880}, {"loss": 0.7649, "grad_norm": 0.5974002480506897, "learning_rate": 0.0002, "epoch": 1.9222374909061515, "step": 11890}, {"loss": 0.7843, "grad_norm": 0.7091866135597229, "learning_rate": 0.0002, "epoch": 1.923854175086897, "step": 11900}, {"loss": 0.775, "grad_norm": 0.72496497631073, "learning_rate": 0.0002, "epoch": 1.9254708592676422, "step": 11910}, {"loss": 0.7153, "grad_norm": 0.6131896376609802, "learning_rate": 0.0002, "epoch": 1.9270875434483874, "step": 11920}, {"loss": 0.7228, "grad_norm": 0.6556436419487, "learning_rate": 0.0002, "epoch": 1.9287042276291326, "step": 11930}, {"loss": 0.7319, "grad_norm": 0.622932493686676, "learning_rate": 0.0002, "epoch": 1.9303209118098779, "step": 11940}, {"loss": 0.7592, "grad_norm": 0.6618631482124329, "learning_rate": 0.0002, "epoch": 1.931937595990623, "step": 11950}, {"loss": 0.8332, "grad_norm": 0.630966305732727, "learning_rate": 0.0002, "epoch": 1.9335542801713685, "step": 11960}, {"loss": 0.6854, "grad_norm": 0.6336734890937805, "learning_rate": 0.0002, "epoch": 1.9351709643521138, "step": 11970}, {"loss": 0.7433, "grad_norm": 0.655403196811676, "learning_rate": 0.0002, "epoch": 1.9367876485328592, "step": 11980}, {"loss": 0.7282, "grad_norm": 0.5640574097633362, "learning_rate": 0.0002, "epoch": 1.9384043327136045, "step": 11990}, {"loss": 0.7289, "grad_norm": 0.6322951316833496, "learning_rate": 0.0002, "epoch": 1.9400210168943497, "step": 12000}, {"loss": 0.7627, "grad_norm": 0.615703821182251, "learning_rate": 0.0002, "epoch": 1.941637701075095, "step": 12010}, {"loss": 0.786, "grad_norm": 0.6487536430358887, "learning_rate": 0.0002, "epoch": 1.9432543852558402, "step": 12020}, {"loss": 0.7435, "grad_norm": 0.9209630489349365, "learning_rate": 0.0002, "epoch": 1.9448710694365856, "step": 12030}, {"loss": 0.7274, "grad_norm": 0.67485511302948, "learning_rate": 0.0002, "epoch": 1.9464877536173308, "step": 12040}, {"loss": 0.7551, "grad_norm": 0.6831230521202087, "learning_rate": 0.0002, "epoch": 1.9481044377980763, "step": 12050}, {"loss": 0.7546, "grad_norm": 0.6578302383422852, "learning_rate": 0.0002, "epoch": 1.9497211219788215, "step": 12060}, {"loss": 0.6989, "grad_norm": 0.9975938200950623, "learning_rate": 0.0002, "epoch": 1.9513378061595668, "step": 12070}, {"loss": 0.7952, "grad_norm": 0.6637365221977234, "learning_rate": 0.0002, "epoch": 1.952954490340312, "step": 12080}, {"loss": 0.7482, "grad_norm": 0.605707049369812, "learning_rate": 0.0002, "epoch": 1.9545711745210572, "step": 12090}, {"loss": 0.7768, "grad_norm": 0.6584440469741821, "learning_rate": 0.0002, "epoch": 1.9561878587018025, "step": 12100}, {"loss": 0.7187, "grad_norm": 0.6070835590362549, "learning_rate": 0.0002, "epoch": 1.957804542882548, "step": 12110}, {"loss": 0.7491, "grad_norm": 0.7862601280212402, "learning_rate": 0.0002, "epoch": 1.9594212270632931, "step": 12120}, {"loss": 0.7972, "grad_norm": 0.8175255060195923, "learning_rate": 0.0002, "epoch": 1.9610379112440386, "step": 12130}, {"loss": 0.7242, "grad_norm": 0.5648472905158997, "learning_rate": 0.0002, "epoch": 1.9626545954247838, "step": 12140}, {"loss": 0.7321, "grad_norm": 0.6591973304748535, "learning_rate": 0.0002, "epoch": 1.964271279605529, "step": 12150}, {"loss": 0.739, "grad_norm": 0.5960676074028015, "learning_rate": 0.0002, "epoch": 1.9658879637862743, "step": 12160}, {"loss": 0.7254, "grad_norm": 0.7272544503211975, "learning_rate": 0.0002, "epoch": 1.9675046479670195, "step": 12170}, {"loss": 0.7376, "grad_norm": 0.7176699042320251, "learning_rate": 0.0002, "epoch": 1.969121332147765, "step": 12180}, {"loss": 0.7525, "grad_norm": 0.6927123665809631, "learning_rate": 0.0002, "epoch": 1.9707380163285102, "step": 12190}, {"loss": 0.7318, "grad_norm": 0.5536034107208252, "learning_rate": 0.0002, "epoch": 1.9723547005092557, "step": 12200}, {"loss": 0.7737, "grad_norm": 0.8348390460014343, "learning_rate": 0.0002, "epoch": 1.9739713846900009, "step": 12210}, {"loss": 0.7494, "grad_norm": 0.6591181755065918, "learning_rate": 0.0002, "epoch": 1.9755880688707461, "step": 12220}, {"loss": 0.763, "grad_norm": 1.0624109506607056, "learning_rate": 0.0002, "epoch": 1.9772047530514913, "step": 12230}, {"loss": 0.7541, "grad_norm": 0.9265586137771606, "learning_rate": 0.0002, "epoch": 1.9788214372322366, "step": 12240}, {"loss": 0.7533, "grad_norm": 0.5998196005821228, "learning_rate": 0.0002, "epoch": 1.9804381214129818, "step": 12250}, {"loss": 0.7225, "grad_norm": 0.6960851550102234, "learning_rate": 0.0002, "epoch": 1.9820548055937273, "step": 12260}, {"loss": 0.7398, "grad_norm": 0.7674502730369568, "learning_rate": 0.0002, "epoch": 1.9836714897744727, "step": 12270}, {"loss": 0.7185, "grad_norm": 0.6407275795936584, "learning_rate": 0.0002, "epoch": 1.985288173955218, "step": 12280}, {"loss": 0.7382, "grad_norm": 0.6673079133033752, "learning_rate": 0.0002, "epoch": 1.9869048581359632, "step": 12290}, {"loss": 0.7326, "grad_norm": 0.6989844441413879, "learning_rate": 0.0002, "epoch": 1.9885215423167084, "step": 12300}, {"loss": 0.7559, "grad_norm": 0.7564442157745361, "learning_rate": 0.0002, "epoch": 1.9901382264974536, "step": 12310}, {"loss": 0.7719, "grad_norm": 0.6385478973388672, "learning_rate": 0.0002, "epoch": 1.9917549106781989, "step": 12320}, {"loss": 0.7369, "grad_norm": 0.7193717956542969, "learning_rate": 0.0002, "epoch": 1.9933715948589443, "step": 12330}, {"loss": 0.7583, "grad_norm": 0.7987112402915955, "learning_rate": 0.0002, "epoch": 1.9949882790396896, "step": 12340}, {"loss": 0.7793, "grad_norm": 0.7260826826095581, "learning_rate": 0.0002, "epoch": 1.996604963220435, "step": 12350}, {"loss": 0.7505, "grad_norm": 0.7968255281448364, "learning_rate": 0.0002, "epoch": 1.9982216474011802, "step": 12360}, {"loss": 0.717, "grad_norm": 0.6893062591552734, "learning_rate": 0.0002, "epoch": 1.9998383315819255, "step": 12370}, {"eval_loss": 1.1044032573699951, "eval_runtime": 122.1508, "eval_samples_per_second": 6.001, "eval_steps_per_second": 0.753, "epoch": 2.0, "step": 12371}, {"loss": 0.6604, "grad_norm": 0.7775409817695618, "learning_rate": 0.0002, "epoch": 2.0014550157626707, "step": 12380}, {"loss": 0.6845, "grad_norm": 0.76218581199646, "learning_rate": 0.0002, "epoch": 2.003071699943416, "step": 12390}, {"loss": 0.6909, "grad_norm": 0.5677764415740967, "learning_rate": 0.0002, "epoch": 2.004688384124161, "step": 12400}, {"loss": 0.6584, "grad_norm": 0.808442234992981, "learning_rate": 0.0002, "epoch": 2.006305068304907, "step": 12410}, {"loss": 0.659, "grad_norm": 0.7144765257835388, "learning_rate": 0.0002, "epoch": 2.007921752485652, "step": 12420}, {"loss": 0.6666, "grad_norm": 0.6914031505584717, "learning_rate": 0.0002, "epoch": 2.0095384366663973, "step": 12430}, {"loss": 0.6596, "grad_norm": 0.7581454515457153, "learning_rate": 0.0002, "epoch": 2.0111551208471425, "step": 12440}, {"loss": 0.6785, "grad_norm": 0.8388504981994629, "learning_rate": 0.0002, "epoch": 2.0127718050278878, "step": 12450}, {"loss": 0.6942, "grad_norm": 0.6716406941413879, "learning_rate": 0.0002, "epoch": 2.014388489208633, "step": 12460}, {"loss": 0.6441, "grad_norm": 0.898902416229248, "learning_rate": 0.0002, "epoch": 2.0160051733893782, "step": 12470}, {"loss": 0.6655, "grad_norm": 0.6432679891586304, "learning_rate": 0.0002, "epoch": 2.0176218575701235, "step": 12480}, {"loss": 0.6521, "grad_norm": 0.8021109104156494, "learning_rate": 0.0002, "epoch": 2.019238541750869, "step": 12490}, {"loss": 0.6581, "grad_norm": 0.7039216756820679, "learning_rate": 0.0002, "epoch": 2.0208552259316144, "step": 12500}, {"loss": 0.6521, "grad_norm": 0.646531879901886, "learning_rate": 0.0002, "epoch": 2.0224719101123596, "step": 12510}, {"loss": 0.6302, "grad_norm": 0.783704400062561, "learning_rate": 0.0002, "epoch": 2.024088594293105, "step": 12520}, {"loss": 0.6288, "grad_norm": 0.8805046677589417, "learning_rate": 0.0002, "epoch": 2.02570527847385, "step": 12530}, {"loss": 0.6288, "grad_norm": 0.7289270758628845, "learning_rate": 0.0002, "epoch": 2.0273219626545953, "step": 12540}, {"loss": 0.6663, "grad_norm": 0.71653151512146, "learning_rate": 0.0002, "epoch": 2.0289386468353405, "step": 12550}, {"loss": 0.625, "grad_norm": 0.73281329870224, "learning_rate": 0.0002, "epoch": 2.030555331016086, "step": 12560}, {"loss": 0.6448, "grad_norm": 0.6657090187072754, "learning_rate": 0.0002, "epoch": 2.0321720151968314, "step": 12570}, {"loss": 0.6983, "grad_norm": 0.8241133093833923, "learning_rate": 0.0002, "epoch": 2.0337886993775767, "step": 12580}, {"loss": 0.6488, "grad_norm": 0.5834135413169861, "learning_rate": 0.0002, "epoch": 2.035405383558322, "step": 12590}, {"loss": 0.6188, "grad_norm": 0.84502112865448, "learning_rate": 0.0002, "epoch": 2.037022067739067, "step": 12600}, {"loss": 0.6349, "grad_norm": 0.8952481746673584, "learning_rate": 0.0002, "epoch": 2.0386387519198124, "step": 12610}, {"loss": 0.6923, "grad_norm": 0.7801461815834045, "learning_rate": 0.0002, "epoch": 2.0402554361005576, "step": 12620}, {"loss": 0.6176, "grad_norm": 0.6788367033004761, "learning_rate": 0.0002, "epoch": 2.041872120281303, "step": 12630}, {"loss": 0.6162, "grad_norm": 0.7241756319999695, "learning_rate": 0.0002, "epoch": 2.0434888044620485, "step": 12640}, {"loss": 0.655, "grad_norm": 0.6933388113975525, "learning_rate": 0.0002, "epoch": 2.0451054886427937, "step": 12650}, {"loss": 0.6431, "grad_norm": 0.8029746413230896, "learning_rate": 0.0002, "epoch": 2.046722172823539, "step": 12660}, {"loss": 0.7164, "grad_norm": 0.946399986743927, "learning_rate": 0.0002, "epoch": 2.048338857004284, "step": 12670}, {"loss": 0.638, "grad_norm": 0.7072678804397583, "learning_rate": 0.0002, "epoch": 2.0499555411850294, "step": 12680}, {"loss": 0.6487, "grad_norm": 0.6810618042945862, "learning_rate": 0.0002, "epoch": 2.0515722253657747, "step": 12690}, {"loss": 0.6554, "grad_norm": 0.7661160230636597, "learning_rate": 0.0002, "epoch": 2.05318890954652, "step": 12700}, {"loss": 0.6799, "grad_norm": 0.6350653767585754, "learning_rate": 0.0002, "epoch": 2.0548055937272656, "step": 12710}, {"loss": 0.6654, "grad_norm": 0.861890971660614, "learning_rate": 0.0002, "epoch": 2.056422277908011, "step": 12720}, {"loss": 0.6286, "grad_norm": 0.6489875912666321, "learning_rate": 0.0002, "epoch": 2.058038962088756, "step": 12730}, {"loss": 0.6811, "grad_norm": 0.8268506526947021, "learning_rate": 0.0002, "epoch": 2.0596556462695013, "step": 12740}, {"loss": 0.6524, "grad_norm": 0.607679545879364, "learning_rate": 0.0002, "epoch": 2.0612723304502465, "step": 12750}, {"loss": 0.6649, "grad_norm": 0.6754153370857239, "learning_rate": 0.0002, "epoch": 2.0628890146309917, "step": 12760}, {"loss": 0.6549, "grad_norm": 0.7263124585151672, "learning_rate": 0.0002, "epoch": 2.064505698811737, "step": 12770}, {"loss": 0.6189, "grad_norm": 0.6986154317855835, "learning_rate": 0.0002, "epoch": 2.0661223829924826, "step": 12780}, {"loss": 0.6723, "grad_norm": 0.7768576741218567, "learning_rate": 0.0002, "epoch": 2.067739067173228, "step": 12790}, {"loss": 0.677, "grad_norm": 0.7546762824058533, "learning_rate": 0.0002, "epoch": 2.069355751353973, "step": 12800}, {"loss": 0.6485, "grad_norm": 0.7588880062103271, "learning_rate": 0.0002, "epoch": 2.0709724355347183, "step": 12810}, {"loss": 0.6989, "grad_norm": 0.7457242608070374, "learning_rate": 0.0002, "epoch": 2.0725891197154636, "step": 12820}, {"loss": 0.6489, "grad_norm": 0.6983516812324524, "learning_rate": 0.0002, "epoch": 2.074205803896209, "step": 12830}, {"loss": 0.651, "grad_norm": 0.7950928807258606, "learning_rate": 0.0002, "epoch": 2.075822488076954, "step": 12840}, {"loss": 0.6603, "grad_norm": 0.9248087406158447, "learning_rate": 0.0002, "epoch": 2.0774391722576993, "step": 12850}, {"loss": 0.6847, "grad_norm": 0.7229493260383606, "learning_rate": 0.0002, "epoch": 2.079055856438445, "step": 12860}, {"loss": 0.6702, "grad_norm": 0.5710847973823547, "learning_rate": 0.0002, "epoch": 2.08067254061919, "step": 12870}, {"loss": 0.6974, "grad_norm": 0.9580423831939697, "learning_rate": 0.0002, "epoch": 2.0822892247999354, "step": 12880}, {"loss": 0.6341, "grad_norm": 0.7399665713310242, "learning_rate": 0.0002, "epoch": 2.0839059089806806, "step": 12890}, {"loss": 0.6993, "grad_norm": 0.7981410622596741, "learning_rate": 0.0002, "epoch": 2.085522593161426, "step": 12900}, {"loss": 0.6976, "grad_norm": 0.870759904384613, "learning_rate": 0.0002, "epoch": 2.087139277342171, "step": 12910}, {"loss": 0.7194, "grad_norm": 0.7001481652259827, "learning_rate": 0.0002, "epoch": 2.0887559615229163, "step": 12920}, {"loss": 0.6383, "grad_norm": 0.6745418310165405, "learning_rate": 0.0002, "epoch": 2.090372645703662, "step": 12930}, {"loss": 0.6519, "grad_norm": 0.7739067673683167, "learning_rate": 0.0002, "epoch": 2.0919893298844072, "step": 12940}, {"loss": 0.6856, "grad_norm": 0.6742934584617615, "learning_rate": 0.0002, "epoch": 2.0936060140651525, "step": 12950}, {"loss": 0.6279, "grad_norm": 0.7270349860191345, "learning_rate": 0.0002, "epoch": 2.0952226982458977, "step": 12960}, {"loss": 0.6783, "grad_norm": 0.7150624394416809, "learning_rate": 0.0002, "epoch": 2.096839382426643, "step": 12970}, {"loss": 0.6093, "grad_norm": 0.7734767198562622, "learning_rate": 0.0002, "epoch": 2.098456066607388, "step": 12980}, {"loss": 0.6534, "grad_norm": 0.7618662118911743, "learning_rate": 0.0002, "epoch": 2.1000727507881334, "step": 12990}, {"loss": 0.6707, "grad_norm": 0.6557944416999817, "learning_rate": 0.0002, "epoch": 2.101689434968879, "step": 13000}, {"loss": 0.7268, "grad_norm": 0.8786448240280151, "learning_rate": 0.0002, "epoch": 2.1033061191496243, "step": 13010}, {"loss": 0.6677, "grad_norm": 0.6878724098205566, "learning_rate": 0.0002, "epoch": 2.1049228033303695, "step": 13020}, {"loss": 0.6824, "grad_norm": 0.822318971157074, "learning_rate": 0.0002, "epoch": 2.1065394875111147, "step": 13030}, {"loss": 0.6228, "grad_norm": 0.831468939781189, "learning_rate": 0.0002, "epoch": 2.10815617169186, "step": 13040}, {"loss": 0.6511, "grad_norm": 0.7699505686759949, "learning_rate": 0.0002, "epoch": 2.109772855872605, "step": 13050}, {"loss": 0.6671, "grad_norm": 0.7559016346931458, "learning_rate": 0.0002, "epoch": 2.1113895400533504, "step": 13060}, {"loss": 0.6215, "grad_norm": 0.6942209601402283, "learning_rate": 0.0002, "epoch": 2.1130062242340957, "step": 13070}, {"loss": 0.6449, "grad_norm": 0.6098947525024414, "learning_rate": 0.0002, "epoch": 2.1146229084148414, "step": 13080}, {"loss": 0.7091, "grad_norm": 0.6499016284942627, "learning_rate": 0.0002, "epoch": 2.1162395925955866, "step": 13090}, {"loss": 0.6247, "grad_norm": 0.7719953060150146, "learning_rate": 0.0002, "epoch": 2.117856276776332, "step": 13100}, {"loss": 0.6064, "grad_norm": 0.6708134412765503, "learning_rate": 0.0002, "epoch": 2.119472960957077, "step": 13110}, {"loss": 0.6056, "grad_norm": 0.8119585514068604, "learning_rate": 0.0002, "epoch": 2.1210896451378223, "step": 13120}, {"loss": 0.6628, "grad_norm": 0.6947157979011536, "learning_rate": 0.0002, "epoch": 2.1227063293185675, "step": 13130}, {"loss": 0.6375, "grad_norm": 0.8831837773323059, "learning_rate": 0.0002, "epoch": 2.1243230134993127, "step": 13140}, {"loss": 0.6997, "grad_norm": 0.7266910672187805, "learning_rate": 0.0002, "epoch": 2.1259396976800584, "step": 13150}, {"loss": 0.6446, "grad_norm": 0.8864351511001587, "learning_rate": 0.0002, "epoch": 2.1275563818608036, "step": 13160}, {"loss": 0.6762, "grad_norm": 0.8104248046875, "learning_rate": 0.0002, "epoch": 2.129173066041549, "step": 13170}, {"loss": 0.6581, "grad_norm": 0.6077079772949219, "learning_rate": 0.0002, "epoch": 2.130789750222294, "step": 13180}, {"loss": 0.6572, "grad_norm": 0.6874213814735413, "learning_rate": 0.0002, "epoch": 2.1324064344030393, "step": 13190}, {"loss": 0.642, "grad_norm": 0.7134367823600769, "learning_rate": 0.0002, "epoch": 2.1340231185837846, "step": 13200}, {"loss": 0.7016, "grad_norm": 0.6101235151290894, "learning_rate": 0.0002, "epoch": 2.13563980276453, "step": 13210}, {"loss": 0.6529, "grad_norm": 0.6042411923408508, "learning_rate": 0.0002, "epoch": 2.137256486945275, "step": 13220}, {"loss": 0.7179, "grad_norm": 0.914601743221283, "learning_rate": 0.0002, "epoch": 2.1388731711260207, "step": 13230}, {"loss": 0.6513, "grad_norm": 0.7104284167289734, "learning_rate": 0.0002, "epoch": 2.140489855306766, "step": 13240}, {"loss": 0.6607, "grad_norm": 0.664395272731781, "learning_rate": 0.0002, "epoch": 2.142106539487511, "step": 13250}, {"loss": 0.7211, "grad_norm": 0.6991241574287415, "learning_rate": 0.0002, "epoch": 2.1437232236682564, "step": 13260}, {"loss": 0.6484, "grad_norm": 0.5469560623168945, "learning_rate": 0.0002, "epoch": 2.1453399078490016, "step": 13270}, {"loss": 0.6765, "grad_norm": 0.8454998135566711, "learning_rate": 0.0002, "epoch": 2.146956592029747, "step": 13280}, {"loss": 0.6683, "grad_norm": 0.7088868618011475, "learning_rate": 0.0002, "epoch": 2.148573276210492, "step": 13290}, {"loss": 0.6835, "grad_norm": 0.7002687454223633, "learning_rate": 0.0002, "epoch": 2.1501899603912378, "step": 13300}, {"loss": 0.6399, "grad_norm": 0.7785214781761169, "learning_rate": 0.0002, "epoch": 2.151806644571983, "step": 13310}, {"loss": 0.67, "grad_norm": 0.8049132227897644, "learning_rate": 0.0002, "epoch": 2.1534233287527282, "step": 13320}, {"loss": 0.6495, "grad_norm": 0.8062595129013062, "learning_rate": 0.0002, "epoch": 2.1550400129334735, "step": 13330}, {"loss": 0.6603, "grad_norm": 0.6208319067955017, "learning_rate": 0.0002, "epoch": 2.1566566971142187, "step": 13340}, {"loss": 0.6584, "grad_norm": 0.7519655823707581, "learning_rate": 0.0002, "epoch": 2.158273381294964, "step": 13350}, {"loss": 0.6457, "grad_norm": 0.7645747065544128, "learning_rate": 0.0002, "epoch": 2.159890065475709, "step": 13360}, {"loss": 0.645, "grad_norm": 0.6847302913665771, "learning_rate": 0.0002, "epoch": 2.1615067496564544, "step": 13370}, {"loss": 0.6903, "grad_norm": 0.8630441427230835, "learning_rate": 0.0002, "epoch": 2.1631234338372, "step": 13380}, {"loss": 0.6742, "grad_norm": 0.7947702407836914, "learning_rate": 0.0002, "epoch": 2.1647401180179453, "step": 13390}, {"loss": 0.7206, "grad_norm": 0.6836977005004883, "learning_rate": 0.0002, "epoch": 2.1663568021986905, "step": 13400}, {"loss": 0.6304, "grad_norm": 0.7340566515922546, "learning_rate": 0.0002, "epoch": 2.1679734863794358, "step": 13410}, {"loss": 0.6528, "grad_norm": 0.7075738906860352, "learning_rate": 0.0002, "epoch": 2.169590170560181, "step": 13420}, {"loss": 0.6585, "grad_norm": 0.7080879807472229, "learning_rate": 0.0002, "epoch": 2.1712068547409262, "step": 13430}, {"loss": 0.6615, "grad_norm": 0.6218613386154175, "learning_rate": 0.0002, "epoch": 2.1728235389216715, "step": 13440}, {"loss": 0.6488, "grad_norm": 0.8211479187011719, "learning_rate": 0.0002, "epoch": 2.174440223102417, "step": 13450}, {"loss": 0.6738, "grad_norm": 0.864466667175293, "learning_rate": 0.0002, "epoch": 2.1760569072831624, "step": 13460}, {"loss": 0.679, "grad_norm": 0.7943857908248901, "learning_rate": 0.0002, "epoch": 2.1776735914639076, "step": 13470}, {"loss": 0.6838, "grad_norm": 0.78728187084198, "learning_rate": 0.0002, "epoch": 2.179290275644653, "step": 13480}, {"loss": 0.6397, "grad_norm": 0.697527289390564, "learning_rate": 0.0002, "epoch": 2.180906959825398, "step": 13490}, {"loss": 0.669, "grad_norm": 0.8205804228782654, "learning_rate": 0.0002, "epoch": 2.1825236440061433, "step": 13500}, {"loss": 0.7227, "grad_norm": 0.8709042072296143, "learning_rate": 0.0002, "epoch": 2.1841403281868885, "step": 13510}, {"loss": 0.6313, "grad_norm": 0.6228537559509277, "learning_rate": 0.0002, "epoch": 2.1857570123676338, "step": 13520}, {"loss": 0.7025, "grad_norm": 0.9566980004310608, "learning_rate": 0.0002, "epoch": 2.1873736965483794, "step": 13530}, {"loss": 0.6755, "grad_norm": 0.7128894329071045, "learning_rate": 0.0002, "epoch": 2.1889903807291247, "step": 13540}, {"loss": 0.6827, "grad_norm": 0.6888654232025146, "learning_rate": 0.0002, "epoch": 2.19060706490987, "step": 13550}, {"loss": 0.6961, "grad_norm": 0.6444337368011475, "learning_rate": 0.0002, "epoch": 2.192223749090615, "step": 13560}, {"loss": 0.656, "grad_norm": 0.8008806705474854, "learning_rate": 0.0002, "epoch": 2.1938404332713604, "step": 13570}, {"loss": 0.7, "grad_norm": 0.8482748866081238, "learning_rate": 0.0002, "epoch": 2.1954571174521056, "step": 13580}, {"loss": 0.7326, "grad_norm": 0.8584157228469849, "learning_rate": 0.0002, "epoch": 2.197073801632851, "step": 13590}, {"loss": 0.7014, "grad_norm": 0.7513734698295593, "learning_rate": 0.0002, "epoch": 2.1986904858135965, "step": 13600}, {"loss": 0.6632, "grad_norm": 0.7864262461662292, "learning_rate": 0.0002, "epoch": 2.2003071699943417, "step": 13610}, {"loss": 0.6879, "grad_norm": 0.8493645191192627, "learning_rate": 0.0002, "epoch": 2.201923854175087, "step": 13620}, {"loss": 0.6617, "grad_norm": 0.6902140974998474, "learning_rate": 0.0002, "epoch": 2.203540538355832, "step": 13630}, {"loss": 0.6655, "grad_norm": 0.8711254596710205, "learning_rate": 0.0002, "epoch": 2.2051572225365774, "step": 13640}, {"loss": 0.6359, "grad_norm": 0.7832191586494446, "learning_rate": 0.0002, "epoch": 2.2067739067173227, "step": 13650}, {"loss": 0.6723, "grad_norm": 0.5668176412582397, "learning_rate": 0.0002, "epoch": 2.208390590898068, "step": 13660}, {"loss": 0.635, "grad_norm": 0.8648375272750854, "learning_rate": 0.0002, "epoch": 2.2100072750788136, "step": 13670}, {"loss": 0.653, "grad_norm": 0.7643089890480042, "learning_rate": 0.0002, "epoch": 2.211623959259559, "step": 13680}, {"loss": 0.6765, "grad_norm": 0.6293777823448181, "learning_rate": 0.0002, "epoch": 2.213240643440304, "step": 13690}, {"loss": 0.6842, "grad_norm": 0.6459372639656067, "learning_rate": 0.0002, "epoch": 2.2148573276210493, "step": 13700}, {"loss": 0.6526, "grad_norm": 0.7060744166374207, "learning_rate": 0.0002, "epoch": 2.2164740118017945, "step": 13710}, {"loss": 0.7101, "grad_norm": 0.674109160900116, "learning_rate": 0.0002, "epoch": 2.2180906959825397, "step": 13720}, {"loss": 0.6529, "grad_norm": 0.830392062664032, "learning_rate": 0.0002, "epoch": 2.219707380163285, "step": 13730}, {"loss": 0.6733, "grad_norm": 0.6474477052688599, "learning_rate": 0.0002, "epoch": 2.2213240643440306, "step": 13740}, {"loss": 0.6413, "grad_norm": 0.7037909626960754, "learning_rate": 0.0002, "epoch": 2.222940748524776, "step": 13750}, {"loss": 0.6417, "grad_norm": 0.6554131507873535, "learning_rate": 0.0002, "epoch": 2.224557432705521, "step": 13760}, {"loss": 0.6907, "grad_norm": 0.7822230458259583, "learning_rate": 0.0002, "epoch": 2.2261741168862663, "step": 13770}, {"loss": 0.6505, "grad_norm": 0.9082167744636536, "learning_rate": 0.0002, "epoch": 2.2277908010670116, "step": 13780}, {"loss": 0.6878, "grad_norm": 0.7918276190757751, "learning_rate": 0.0002, "epoch": 2.229407485247757, "step": 13790}, {"loss": 0.6669, "grad_norm": 0.7354569435119629, "learning_rate": 0.0002, "epoch": 2.231024169428502, "step": 13800}, {"loss": 0.6503, "grad_norm": 0.8265249133110046, "learning_rate": 0.0002, "epoch": 2.2326408536092472, "step": 13810}, {"loss": 0.6871, "grad_norm": 0.6653847098350525, "learning_rate": 0.0002, "epoch": 2.234257537789993, "step": 13820}, {"loss": 0.6413, "grad_norm": 0.7157923579216003, "learning_rate": 0.0002, "epoch": 2.235874221970738, "step": 13830}, {"loss": 0.6306, "grad_norm": 0.7110323309898376, "learning_rate": 0.0002, "epoch": 2.2374909061514834, "step": 13840}, {"loss": 0.6913, "grad_norm": 0.7155357599258423, "learning_rate": 0.0002, "epoch": 2.2391075903322286, "step": 13850}, {"loss": 0.6579, "grad_norm": 1.0177817344665527, "learning_rate": 0.0002, "epoch": 2.240724274512974, "step": 13860}, {"loss": 0.635, "grad_norm": 0.7601948380470276, "learning_rate": 0.0002, "epoch": 2.242340958693719, "step": 13870}, {"loss": 0.6679, "grad_norm": 0.7628820538520813, "learning_rate": 0.0002, "epoch": 2.2439576428744643, "step": 13880}, {"loss": 0.6805, "grad_norm": 0.7089297771453857, "learning_rate": 0.0002, "epoch": 2.24557432705521, "step": 13890}, {"loss": 0.7236, "grad_norm": 0.695178210735321, "learning_rate": 0.0002, "epoch": 2.247191011235955, "step": 13900}, {"loss": 0.7084, "grad_norm": 0.7631948590278625, "learning_rate": 0.0002, "epoch": 2.2488076954167004, "step": 13910}, {"loss": 0.685, "grad_norm": 0.8203101754188538, "learning_rate": 0.0002, "epoch": 2.2504243795974457, "step": 13920}, {"loss": 0.653, "grad_norm": 0.8099079728126526, "learning_rate": 0.0002, "epoch": 2.252041063778191, "step": 13930}, {"loss": 0.694, "grad_norm": 0.6498546004295349, "learning_rate": 0.0002, "epoch": 2.253657747958936, "step": 13940}, {"loss": 0.6684, "grad_norm": 0.7797415256500244, "learning_rate": 0.0002, "epoch": 2.2552744321396814, "step": 13950}, {"loss": 0.683, "grad_norm": 0.8254124522209167, "learning_rate": 0.0002, "epoch": 2.2568911163204266, "step": 13960}, {"loss": 0.6806, "grad_norm": 0.6327953338623047, "learning_rate": 0.0002, "epoch": 2.2585078005011723, "step": 13970}, {"loss": 0.668, "grad_norm": 0.734194278717041, "learning_rate": 0.0002, "epoch": 2.2601244846819175, "step": 13980}, {"loss": 0.6912, "grad_norm": 0.9014202952384949, "learning_rate": 0.0002, "epoch": 2.2617411688626627, "step": 13990}, {"loss": 0.692, "grad_norm": 0.7643631100654602, "learning_rate": 0.0002, "epoch": 2.263357853043408, "step": 14000}, {"loss": 0.6657, "grad_norm": 0.8882834911346436, "learning_rate": 0.0002, "epoch": 2.264974537224153, "step": 14010}, {"loss": 0.6453, "grad_norm": 0.7975873351097107, "learning_rate": 0.0002, "epoch": 2.2665912214048984, "step": 14020}, {"loss": 0.7193, "grad_norm": 0.7765783071517944, "learning_rate": 0.0002, "epoch": 2.2682079055856437, "step": 14030}, {"loss": 0.662, "grad_norm": 0.8846288323402405, "learning_rate": 0.0002, "epoch": 2.2698245897663893, "step": 14040}, {"loss": 0.6494, "grad_norm": 0.9006744027137756, "learning_rate": 0.0002, "epoch": 2.2714412739471346, "step": 14050}, {"loss": 0.6423, "grad_norm": 0.7420173287391663, "learning_rate": 0.0002, "epoch": 2.27305795812788, "step": 14060}, {"loss": 0.7068, "grad_norm": 0.7956424951553345, "learning_rate": 0.0002, "epoch": 2.274674642308625, "step": 14070}, {"loss": 0.6581, "grad_norm": 0.7783209085464478, "learning_rate": 0.0002, "epoch": 2.2762913264893703, "step": 14080}, {"loss": 0.7202, "grad_norm": 0.7597188949584961, "learning_rate": 0.0002, "epoch": 2.2779080106701155, "step": 14090}, {"loss": 0.6778, "grad_norm": 0.6718921661376953, "learning_rate": 0.0002, "epoch": 2.2795246948508607, "step": 14100}, {"loss": 0.632, "grad_norm": 0.7528082132339478, "learning_rate": 0.0002, "epoch": 2.281141379031606, "step": 14110}, {"loss": 0.7608, "grad_norm": 0.8379864692687988, "learning_rate": 0.0002, "epoch": 2.2827580632123516, "step": 14120}, {"loss": 0.6767, "grad_norm": 0.748613715171814, "learning_rate": 0.0002, "epoch": 2.284374747393097, "step": 14130}, {"loss": 0.6641, "grad_norm": 0.7435423135757446, "learning_rate": 0.0002, "epoch": 2.285991431573842, "step": 14140}, {"loss": 0.6849, "grad_norm": 0.7580803632736206, "learning_rate": 0.0002, "epoch": 2.2876081157545873, "step": 14150}, {"loss": 0.6604, "grad_norm": 0.6278321146965027, "learning_rate": 0.0002, "epoch": 2.2892247999353326, "step": 14160}, {"loss": 0.6573, "grad_norm": 0.7663896083831787, "learning_rate": 0.0002, "epoch": 2.290841484116078, "step": 14170}, {"loss": 0.6655, "grad_norm": 0.9716812372207642, "learning_rate": 0.0002, "epoch": 2.292458168296823, "step": 14180}, {"loss": 0.7067, "grad_norm": 0.8993458151817322, "learning_rate": 0.0002, "epoch": 2.2940748524775687, "step": 14190}, {"loss": 0.6172, "grad_norm": 0.6156117916107178, "learning_rate": 0.0002, "epoch": 2.295691536658314, "step": 14200}, {"loss": 0.6318, "grad_norm": 0.8911278247833252, "learning_rate": 0.0002, "epoch": 2.297308220839059, "step": 14210}, {"loss": 0.6364, "grad_norm": 0.6422147154808044, "learning_rate": 0.0002, "epoch": 2.2989249050198044, "step": 14220}, {"loss": 0.6795, "grad_norm": 0.6866879463195801, "learning_rate": 0.0002, "epoch": 2.3005415892005496, "step": 14230}, {"loss": 0.6907, "grad_norm": 0.9297130107879639, "learning_rate": 0.0002, "epoch": 2.302158273381295, "step": 14240}, {"loss": 0.6823, "grad_norm": 0.7501356601715088, "learning_rate": 0.0002, "epoch": 2.30377495756204, "step": 14250}, {"loss": 0.6414, "grad_norm": 0.8363515138626099, "learning_rate": 0.0002, "epoch": 2.3053916417427853, "step": 14260}, {"loss": 0.6362, "grad_norm": 0.9083868265151978, "learning_rate": 0.0002, "epoch": 2.307008325923531, "step": 14270}, {"loss": 0.6862, "grad_norm": 0.7791516780853271, "learning_rate": 0.0002, "epoch": 2.3086250101042762, "step": 14280}, {"loss": 0.6569, "grad_norm": 0.8766953349113464, "learning_rate": 0.0002, "epoch": 2.3102416942850215, "step": 14290}, {"loss": 0.6698, "grad_norm": 0.7916635274887085, "learning_rate": 0.0002, "epoch": 2.3118583784657667, "step": 14300}, {"loss": 0.6927, "grad_norm": 0.627525269985199, "learning_rate": 0.0002, "epoch": 2.313475062646512, "step": 14310}, {"loss": 0.6541, "grad_norm": 0.8856783509254456, "learning_rate": 0.0002, "epoch": 2.315091746827257, "step": 14320}, {"loss": 0.6806, "grad_norm": 0.6758689284324646, "learning_rate": 0.0002, "epoch": 2.316708431008003, "step": 14330}, {"loss": 0.6794, "grad_norm": 0.6428321003913879, "learning_rate": 0.0002, "epoch": 2.318325115188748, "step": 14340}, {"loss": 0.682, "grad_norm": 0.9032121300697327, "learning_rate": 0.0002, "epoch": 2.3199417993694933, "step": 14350}, {"loss": 0.6569, "grad_norm": 0.8035986423492432, "learning_rate": 0.0002, "epoch": 2.3215584835502385, "step": 14360}, {"loss": 0.7067, "grad_norm": 0.7974579334259033, "learning_rate": 0.0002, "epoch": 2.3231751677309838, "step": 14370}, {"loss": 0.6451, "grad_norm": 0.8356034755706787, "learning_rate": 0.0002, "epoch": 2.324791851911729, "step": 14380}, {"loss": 0.6623, "grad_norm": 0.998760998249054, "learning_rate": 0.0002, "epoch": 2.326408536092474, "step": 14390}, {"loss": 0.649, "grad_norm": 0.6518142223358154, "learning_rate": 0.0002, "epoch": 2.3280252202732195, "step": 14400}, {"loss": 0.7146, "grad_norm": 0.7443506717681885, "learning_rate": 0.0002, "epoch": 2.3296419044539647, "step": 14410}, {"loss": 0.648, "grad_norm": 0.8436172604560852, "learning_rate": 0.0002, "epoch": 2.3312585886347104, "step": 14420}, {"loss": 0.6585, "grad_norm": 0.7411080598831177, "learning_rate": 0.0002, "epoch": 2.3328752728154556, "step": 14430}, {"loss": 0.6781, "grad_norm": 0.8839048743247986, "learning_rate": 0.0002, "epoch": 2.334491956996201, "step": 14440}, {"loss": 0.6565, "grad_norm": 0.8360885977745056, "learning_rate": 0.0002, "epoch": 2.336108641176946, "step": 14450}, {"loss": 0.6662, "grad_norm": 0.7608986496925354, "learning_rate": 0.0002, "epoch": 2.3377253253576913, "step": 14460}, {"loss": 0.6685, "grad_norm": 0.8179867267608643, "learning_rate": 0.0002, "epoch": 2.3393420095384365, "step": 14470}, {"loss": 0.7055, "grad_norm": 0.5989999771118164, "learning_rate": 0.0002, "epoch": 2.340958693719182, "step": 14480}, {"loss": 0.644, "grad_norm": 0.9450054168701172, "learning_rate": 0.0002, "epoch": 2.3425753778999274, "step": 14490}, {"loss": 0.6983, "grad_norm": 0.7885149717330933, "learning_rate": 0.0002, "epoch": 2.3441920620806727, "step": 14500}, {"loss": 0.6819, "grad_norm": 0.8152616620063782, "learning_rate": 0.0002, "epoch": 2.345808746261418, "step": 14510}, {"loss": 0.6989, "grad_norm": 0.7193838953971863, "learning_rate": 0.0002, "epoch": 2.347425430442163, "step": 14520}, {"loss": 0.6594, "grad_norm": 0.6701092720031738, "learning_rate": 0.0002, "epoch": 2.3490421146229084, "step": 14530}, {"loss": 0.6559, "grad_norm": 0.7529364228248596, "learning_rate": 0.0002, "epoch": 2.3506587988036536, "step": 14540}, {"loss": 0.6306, "grad_norm": 0.6599733829498291, "learning_rate": 0.0002, "epoch": 2.352275482984399, "step": 14550}, {"loss": 0.706, "grad_norm": 0.9502474069595337, "learning_rate": 0.0002, "epoch": 2.353892167165144, "step": 14560}, {"loss": 0.717, "grad_norm": 0.7619650959968567, "learning_rate": 0.0002, "epoch": 2.3555088513458897, "step": 14570}, {"loss": 0.6684, "grad_norm": 0.9854652285575867, "learning_rate": 0.0002, "epoch": 2.357125535526635, "step": 14580}, {"loss": 0.6455, "grad_norm": 0.727439284324646, "learning_rate": 0.0002, "epoch": 2.35874221970738, "step": 14590}, {"loss": 0.6645, "grad_norm": 0.6994746327400208, "learning_rate": 0.0002, "epoch": 2.3603589038881254, "step": 14600}, {"loss": 0.6587, "grad_norm": 0.7117531299591064, "learning_rate": 0.0002, "epoch": 2.3619755880688706, "step": 14610}, {"loss": 0.6804, "grad_norm": 0.6403067708015442, "learning_rate": 0.0002, "epoch": 2.363592272249616, "step": 14620}, {"loss": 0.7055, "grad_norm": 0.8377841711044312, "learning_rate": 0.0002, "epoch": 2.3652089564303616, "step": 14630}, {"loss": 0.6778, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 2.366825640611107, "step": 14640}, {"loss": 0.6552, "grad_norm": 0.8418586254119873, "learning_rate": 0.0002, "epoch": 2.368442324791852, "step": 14650}, {"loss": 0.6685, "grad_norm": 0.6178573369979858, "learning_rate": 0.0002, "epoch": 2.3700590089725972, "step": 14660}, {"loss": 0.6774, "grad_norm": 0.6368302702903748, "learning_rate": 0.0002, "epoch": 2.3716756931533425, "step": 14670}, {"loss": 0.6136, "grad_norm": 0.9122977256774902, "learning_rate": 0.0002, "epoch": 2.3732923773340877, "step": 14680}, {"loss": 0.6675, "grad_norm": 0.7086195349693298, "learning_rate": 0.0002, "epoch": 2.374909061514833, "step": 14690}, {"loss": 0.6582, "grad_norm": 0.7500800490379333, "learning_rate": 0.0002, "epoch": 2.376525745695578, "step": 14700}, {"loss": 0.6792, "grad_norm": 0.6634900569915771, "learning_rate": 0.0002, "epoch": 2.378142429876324, "step": 14710}, {"loss": 0.6614, "grad_norm": 0.839898407459259, "learning_rate": 0.0002, "epoch": 2.379759114057069, "step": 14720}, {"loss": 0.6453, "grad_norm": 0.7578426003456116, "learning_rate": 0.0002, "epoch": 2.3813757982378143, "step": 14730}, {"loss": 0.7282, "grad_norm": 1.0213173627853394, "learning_rate": 0.0002, "epoch": 2.3829924824185595, "step": 14740}, {"loss": 0.6704, "grad_norm": 0.7855949401855469, "learning_rate": 0.0002, "epoch": 2.3846091665993048, "step": 14750}, {"loss": 0.6694, "grad_norm": 0.7224128842353821, "learning_rate": 0.0002, "epoch": 2.38622585078005, "step": 14760}, {"loss": 0.7017, "grad_norm": 0.8040381669998169, "learning_rate": 0.0002, "epoch": 2.3878425349607952, "step": 14770}, {"loss": 0.6799, "grad_norm": 0.7705281376838684, "learning_rate": 0.0002, "epoch": 2.389459219141541, "step": 14780}, {"loss": 0.6326, "grad_norm": 0.667966902256012, "learning_rate": 0.0002, "epoch": 2.391075903322286, "step": 14790}, {"loss": 0.7061, "grad_norm": 0.6611011028289795, "learning_rate": 0.0002, "epoch": 2.3926925875030314, "step": 14800}, {"loss": 0.6527, "grad_norm": 0.6862651705741882, "learning_rate": 0.0002, "epoch": 2.3943092716837766, "step": 14810}, {"loss": 0.6537, "grad_norm": 0.8086010217666626, "learning_rate": 0.0002, "epoch": 2.395925955864522, "step": 14820}, {"loss": 0.7189, "grad_norm": 0.7189689874649048, "learning_rate": 0.0002, "epoch": 2.397542640045267, "step": 14830}, {"loss": 0.6709, "grad_norm": 0.6280009150505066, "learning_rate": 0.0002, "epoch": 2.3991593242260123, "step": 14840}, {"loss": 0.706, "grad_norm": 0.7826612591743469, "learning_rate": 0.0002, "epoch": 2.4007760084067575, "step": 14850}, {"loss": 0.6738, "grad_norm": 0.7681610584259033, "learning_rate": 0.0002, "epoch": 2.402392692587503, "step": 14860}, {"loss": 0.636, "grad_norm": 0.720966100692749, "learning_rate": 0.0002, "epoch": 2.4040093767682484, "step": 14870}, {"loss": 0.6667, "grad_norm": 0.8202250599861145, "learning_rate": 0.0002, "epoch": 2.4056260609489937, "step": 14880}, {"loss": 0.6935, "grad_norm": 0.786212682723999, "learning_rate": 0.0002, "epoch": 2.407242745129739, "step": 14890}, {"loss": 0.6628, "grad_norm": 0.6647164821624756, "learning_rate": 0.0002, "epoch": 2.408859429310484, "step": 14900}, {"loss": 0.6706, "grad_norm": 0.7566399574279785, "learning_rate": 0.0002, "epoch": 2.4104761134912294, "step": 14910}, {"loss": 0.7188, "grad_norm": 0.748814582824707, "learning_rate": 0.0002, "epoch": 2.4120927976719746, "step": 14920}, {"loss": 0.6684, "grad_norm": 0.7624038457870483, "learning_rate": 0.0002, "epoch": 2.4137094818527203, "step": 14930}, {"loss": 0.6483, "grad_norm": 0.8267335295677185, "learning_rate": 0.0002, "epoch": 2.4153261660334655, "step": 14940}, {"loss": 0.6612, "grad_norm": 0.8785360455513, "learning_rate": 0.0002, "epoch": 2.4169428502142107, "step": 14950}, {"loss": 0.6718, "grad_norm": 0.679887592792511, "learning_rate": 0.0002, "epoch": 2.418559534394956, "step": 14960}, {"loss": 0.6136, "grad_norm": 0.7218474745750427, "learning_rate": 0.0002, "epoch": 2.420176218575701, "step": 14970}, {"loss": 0.648, "grad_norm": 0.6342799663543701, "learning_rate": 0.0002, "epoch": 2.4217929027564464, "step": 14980}, {"loss": 0.6617, "grad_norm": 0.7098712921142578, "learning_rate": 0.0002, "epoch": 2.4234095869371917, "step": 14990}, {"loss": 0.6942, "grad_norm": 0.7497431635856628, "learning_rate": 0.0002, "epoch": 2.425026271117937, "step": 15000}, {"loss": 0.6772, "grad_norm": 0.934836208820343, "learning_rate": 0.0002, "epoch": 2.4266429552986826, "step": 15010}, {"loss": 0.7221, "grad_norm": 0.8430966734886169, "learning_rate": 0.0002, "epoch": 2.428259639479428, "step": 15020}, {"loss": 0.6985, "grad_norm": 0.7032104730606079, "learning_rate": 0.0002, "epoch": 2.429876323660173, "step": 15030}, {"loss": 0.6715, "grad_norm": 0.7746111750602722, "learning_rate": 0.0002, "epoch": 2.4314930078409183, "step": 15040}, {"loss": 0.7177, "grad_norm": 0.7661406397819519, "learning_rate": 0.0002, "epoch": 2.4331096920216635, "step": 15050}, {"loss": 0.6517, "grad_norm": 0.6941645741462708, "learning_rate": 0.0002, "epoch": 2.4347263762024087, "step": 15060}, {"loss": 0.6421, "grad_norm": 0.7487249374389648, "learning_rate": 0.0002, "epoch": 2.436343060383154, "step": 15070}, {"loss": 0.6796, "grad_norm": 0.7639912962913513, "learning_rate": 0.0002, "epoch": 2.4379597445638996, "step": 15080}, {"loss": 0.7087, "grad_norm": 0.7708953619003296, "learning_rate": 0.0002, "epoch": 2.439576428744645, "step": 15090}, {"loss": 0.7065, "grad_norm": 0.9135832190513611, "learning_rate": 0.0002, "epoch": 2.44119311292539, "step": 15100}, {"loss": 0.672, "grad_norm": 0.8283005356788635, "learning_rate": 0.0002, "epoch": 2.4428097971061353, "step": 15110}, {"loss": 0.6551, "grad_norm": 0.925299346446991, "learning_rate": 0.0002, "epoch": 2.4444264812868806, "step": 15120}, {"loss": 0.687, "grad_norm": 0.7013528943061829, "learning_rate": 0.0002, "epoch": 2.446043165467626, "step": 15130}, {"loss": 0.6842, "grad_norm": 0.622303307056427, "learning_rate": 0.0002, "epoch": 2.447659849648371, "step": 15140}, {"loss": 0.6676, "grad_norm": 0.876569390296936, "learning_rate": 0.0002, "epoch": 2.4492765338291163, "step": 15150}, {"loss": 0.6463, "grad_norm": 0.6836351752281189, "learning_rate": 0.0002, "epoch": 2.450893218009862, "step": 15160}, {"loss": 0.6781, "grad_norm": 0.7886684536933899, "learning_rate": 0.0002, "epoch": 2.452509902190607, "step": 15170}, {"loss": 0.6794, "grad_norm": 0.6647440791130066, "learning_rate": 0.0002, "epoch": 2.4541265863713524, "step": 15180}, {"loss": 0.6353, "grad_norm": 0.7477722764015198, "learning_rate": 0.0002, "epoch": 2.4557432705520976, "step": 15190}, {"loss": 0.698, "grad_norm": 0.8192033767700195, "learning_rate": 0.0002, "epoch": 2.457359954732843, "step": 15200}, {"loss": 0.6735, "grad_norm": 0.847537100315094, "learning_rate": 0.0002, "epoch": 2.458976638913588, "step": 15210}, {"loss": 0.6962, "grad_norm": 0.9027776122093201, "learning_rate": 0.0002, "epoch": 2.4605933230943338, "step": 15220}, {"loss": 0.7084, "grad_norm": 0.7217772006988525, "learning_rate": 0.0002, "epoch": 2.462210007275079, "step": 15230}, {"loss": 0.691, "grad_norm": 0.7994546294212341, "learning_rate": 0.0002, "epoch": 2.4638266914558242, "step": 15240}, {"loss": 0.6828, "grad_norm": 0.939916729927063, "learning_rate": 0.0002, "epoch": 2.4654433756365695, "step": 15250}, {"loss": 0.6893, "grad_norm": 1.0009053945541382, "learning_rate": 0.0002, "epoch": 2.4670600598173147, "step": 15260}, {"loss": 0.643, "grad_norm": 0.625555694103241, "learning_rate": 0.0002, "epoch": 2.46867674399806, "step": 15270}, {"loss": 0.688, "grad_norm": 0.7924878597259521, "learning_rate": 0.0002, "epoch": 2.470293428178805, "step": 15280}, {"loss": 0.6789, "grad_norm": 0.8536689877510071, "learning_rate": 0.0002, "epoch": 2.4719101123595504, "step": 15290}, {"loss": 0.6924, "grad_norm": 0.8572589755058289, "learning_rate": 0.0002, "epoch": 2.4735267965402956, "step": 15300}, {"loss": 0.604, "grad_norm": 0.773279070854187, "learning_rate": 0.0002, "epoch": 2.4751434807210413, "step": 15310}, {"loss": 0.6573, "grad_norm": 0.7708749771118164, "learning_rate": 0.0002, "epoch": 2.4767601649017865, "step": 15320}, {"loss": 0.7065, "grad_norm": 0.770905077457428, "learning_rate": 0.0002, "epoch": 2.4783768490825318, "step": 15330}, {"loss": 0.6878, "grad_norm": 0.8238571882247925, "learning_rate": 0.0002, "epoch": 2.479993533263277, "step": 15340}, {"loss": 0.6772, "grad_norm": 0.7670477032661438, "learning_rate": 0.0002, "epoch": 2.481610217444022, "step": 15350}, {"loss": 0.7759, "grad_norm": 0.905036985874176, "learning_rate": 0.0002, "epoch": 2.4832269016247674, "step": 15360}, {"loss": 0.706, "grad_norm": 0.6672089695930481, "learning_rate": 0.0002, "epoch": 2.484843585805513, "step": 15370}, {"loss": 0.6722, "grad_norm": 0.625095784664154, "learning_rate": 0.0002, "epoch": 2.4864602699862584, "step": 15380}, {"loss": 0.6396, "grad_norm": 0.679772675037384, "learning_rate": 0.0002, "epoch": 2.4880769541670036, "step": 15390}, {"loss": 0.6778, "grad_norm": 0.711492121219635, "learning_rate": 0.0002, "epoch": 2.489693638347749, "step": 15400}, {"loss": 0.6966, "grad_norm": 0.876189112663269, "learning_rate": 0.0002, "epoch": 2.491310322528494, "step": 15410}, {"loss": 0.7307, "grad_norm": 0.7236915230751038, "learning_rate": 0.0002, "epoch": 2.4929270067092393, "step": 15420}, {"loss": 0.647, "grad_norm": 0.6629832983016968, "learning_rate": 0.0002, "epoch": 2.4945436908899845, "step": 15430}, {"loss": 0.6669, "grad_norm": 0.9756859540939331, "learning_rate": 0.0002, "epoch": 2.4961603750707297, "step": 15440}, {"loss": 0.7559, "grad_norm": 0.6896940469741821, "learning_rate": 0.0002, "epoch": 2.4977770592514754, "step": 15450}, {"loss": 0.6818, "grad_norm": 0.7105149626731873, "learning_rate": 0.0002, "epoch": 2.4993937434322206, "step": 15460}, {"loss": 0.6859, "grad_norm": 0.8374546766281128, "learning_rate": 0.0002, "epoch": 2.501010427612966, "step": 15470}, {"loss": 0.6512, "grad_norm": 0.7320070266723633, "learning_rate": 0.0002, "epoch": 2.502627111793711, "step": 15480}, {"loss": 0.685, "grad_norm": 0.8306367993354797, "learning_rate": 0.0002, "epoch": 2.5042437959744563, "step": 15490}, {"loss": 0.7253, "grad_norm": 0.7472721338272095, "learning_rate": 0.0002, "epoch": 2.5058604801552016, "step": 15500}, {"loss": 0.6699, "grad_norm": 0.6147692203521729, "learning_rate": 0.0002, "epoch": 2.507477164335947, "step": 15510}, {"loss": 0.7158, "grad_norm": 0.7788505554199219, "learning_rate": 0.0002, "epoch": 2.5090938485166925, "step": 15520}, {"loss": 0.6521, "grad_norm": 0.8807527422904968, "learning_rate": 0.0002, "epoch": 2.5107105326974377, "step": 15530}, {"loss": 0.6792, "grad_norm": 0.7521643042564392, "learning_rate": 0.0002, "epoch": 2.512327216878183, "step": 15540}, {"loss": 0.6772, "grad_norm": 0.6900225281715393, "learning_rate": 0.0002, "epoch": 2.513943901058928, "step": 15550}, {"loss": 0.6769, "grad_norm": 0.6601938605308533, "learning_rate": 0.0002, "epoch": 2.5155605852396734, "step": 15560}, {"loss": 0.6648, "grad_norm": 0.8179984092712402, "learning_rate": 0.0002, "epoch": 2.5171772694204186, "step": 15570}, {"loss": 0.7028, "grad_norm": 0.792556881904602, "learning_rate": 0.0002, "epoch": 2.518793953601164, "step": 15580}, {"loss": 0.6464, "grad_norm": 0.7081938982009888, "learning_rate": 0.0002, "epoch": 2.520410637781909, "step": 15590}, {"loss": 0.6691, "grad_norm": 0.8733121156692505, "learning_rate": 0.0002, "epoch": 2.5220273219626543, "step": 15600}, {"loss": 0.6969, "grad_norm": 0.7980992794036865, "learning_rate": 0.0002, "epoch": 2.5236440061434, "step": 15610}, {"loss": 0.7124, "grad_norm": 0.883664071559906, "learning_rate": 0.0002, "epoch": 2.5252606903241452, "step": 15620}, {"loss": 0.7022, "grad_norm": 0.6963341236114502, "learning_rate": 0.0002, "epoch": 2.5268773745048905, "step": 15630}, {"loss": 0.7334, "grad_norm": 0.6433573365211487, "learning_rate": 0.0002, "epoch": 2.5284940586856357, "step": 15640}, {"loss": 0.6889, "grad_norm": 0.8538183569908142, "learning_rate": 0.0002, "epoch": 2.530110742866381, "step": 15650}, {"loss": 0.6841, "grad_norm": 0.9748201370239258, "learning_rate": 0.0002, "epoch": 2.5317274270471266, "step": 15660}, {"loss": 0.6765, "grad_norm": 0.7670575380325317, "learning_rate": 0.0002, "epoch": 2.533344111227872, "step": 15670}, {"loss": 0.6435, "grad_norm": 0.8738890290260315, "learning_rate": 0.0002, "epoch": 2.534960795408617, "step": 15680}, {"loss": 0.6802, "grad_norm": 0.8391636610031128, "learning_rate": 0.0002, "epoch": 2.5365774795893623, "step": 15690}, {"loss": 0.6901, "grad_norm": 0.7239366769790649, "learning_rate": 0.0002, "epoch": 2.5381941637701075, "step": 15700}, {"loss": 0.7011, "grad_norm": 0.8498379588127136, "learning_rate": 0.0002, "epoch": 2.5398108479508528, "step": 15710}, {"loss": 0.6998, "grad_norm": 0.8029484152793884, "learning_rate": 0.0002, "epoch": 2.541427532131598, "step": 15720}, {"loss": 0.6678, "grad_norm": 1.0639333724975586, "learning_rate": 0.0002, "epoch": 2.5430442163123432, "step": 15730}, {"loss": 0.6341, "grad_norm": 0.6401297450065613, "learning_rate": 0.0002, "epoch": 2.5446609004930885, "step": 15740}, {"loss": 0.7196, "grad_norm": 0.7123814821243286, "learning_rate": 0.0002, "epoch": 2.5462775846738337, "step": 15750}, {"loss": 0.654, "grad_norm": 0.7874974608421326, "learning_rate": 0.0002, "epoch": 2.5478942688545794, "step": 15760}, {"loss": 0.6721, "grad_norm": 0.8046808838844299, "learning_rate": 0.0002, "epoch": 2.5495109530353246, "step": 15770}, {"loss": 0.6665, "grad_norm": 0.7888661623001099, "learning_rate": 0.0002, "epoch": 2.55112763721607, "step": 15780}, {"loss": 0.6893, "grad_norm": 0.8445866107940674, "learning_rate": 0.0002, "epoch": 2.552744321396815, "step": 15790}, {"loss": 0.6815, "grad_norm": 0.7475846409797668, "learning_rate": 0.0002, "epoch": 2.5543610055775603, "step": 15800}, {"loss": 0.6711, "grad_norm": 0.7455102801322937, "learning_rate": 0.0002, "epoch": 2.555977689758306, "step": 15810}, {"loss": 0.6932, "grad_norm": 0.8226983547210693, "learning_rate": 0.0002, "epoch": 2.557594373939051, "step": 15820}, {"loss": 0.651, "grad_norm": 0.8920368552207947, "learning_rate": 0.0002, "epoch": 2.5592110581197964, "step": 15830}, {"loss": 0.6297, "grad_norm": 0.8413904905319214, "learning_rate": 0.0002, "epoch": 2.5608277423005417, "step": 15840}, {"loss": 0.7106, "grad_norm": 0.8483649492263794, "learning_rate": 0.0002, "epoch": 2.562444426481287, "step": 15850}, {"loss": 0.6957, "grad_norm": 0.5923284292221069, "learning_rate": 0.0002, "epoch": 2.564061110662032, "step": 15860}, {"loss": 0.6847, "grad_norm": 0.8518726229667664, "learning_rate": 0.0002, "epoch": 2.5656777948427774, "step": 15870}, {"loss": 0.6362, "grad_norm": 0.731235146522522, "learning_rate": 0.0002, "epoch": 2.5672944790235226, "step": 15880}, {"loss": 0.7611, "grad_norm": 0.7517194151878357, "learning_rate": 0.0002, "epoch": 2.568911163204268, "step": 15890}, {"loss": 0.6907, "grad_norm": 0.8378692269325256, "learning_rate": 0.0002, "epoch": 2.5705278473850135, "step": 15900}, {"loss": 0.7055, "grad_norm": 0.843701958656311, "learning_rate": 0.0002, "epoch": 2.5721445315657587, "step": 15910}, {"loss": 0.6882, "grad_norm": 0.7254629731178284, "learning_rate": 0.0002, "epoch": 2.573761215746504, "step": 15920}, {"loss": 0.6872, "grad_norm": 0.8863335847854614, "learning_rate": 0.0002, "epoch": 2.575377899927249, "step": 15930}, {"loss": 0.6813, "grad_norm": 0.7675097584724426, "learning_rate": 0.0002, "epoch": 2.5769945841079944, "step": 15940}, {"loss": 0.7357, "grad_norm": 0.82063889503479, "learning_rate": 0.0002, "epoch": 2.5786112682887397, "step": 15950}, {"loss": 0.662, "grad_norm": 0.7729717493057251, "learning_rate": 0.0002, "epoch": 2.5802279524694853, "step": 15960}, {"loss": 0.633, "grad_norm": 0.8301846981048584, "learning_rate": 0.0002, "epoch": 2.5818446366502306, "step": 15970}, {"loss": 0.6897, "grad_norm": 0.7906861305236816, "learning_rate": 0.0002, "epoch": 2.583461320830976, "step": 15980}, {"loss": 0.7175, "grad_norm": 0.6749057173728943, "learning_rate": 0.0002, "epoch": 2.585078005011721, "step": 15990}, {"loss": 0.7212, "grad_norm": 0.9386842846870422, "learning_rate": 0.0002, "epoch": 2.5866946891924663, "step": 16000}, {"loss": 0.6934, "grad_norm": 0.7868891358375549, "learning_rate": 0.0002, "epoch": 2.5883113733732115, "step": 16010}, {"loss": 0.7036, "grad_norm": 0.8674671053886414, "learning_rate": 0.0002, "epoch": 2.5899280575539567, "step": 16020}, {"loss": 0.7217, "grad_norm": 0.7043559551239014, "learning_rate": 0.0002, "epoch": 2.591544741734702, "step": 16030}, {"loss": 0.6967, "grad_norm": 0.5846083760261536, "learning_rate": 0.0002, "epoch": 2.593161425915447, "step": 16040}, {"loss": 0.7322, "grad_norm": 0.7323982119560242, "learning_rate": 0.0002, "epoch": 2.594778110096193, "step": 16050}, {"loss": 0.6794, "grad_norm": 0.9069556593894958, "learning_rate": 0.0002, "epoch": 2.596394794276938, "step": 16060}, {"loss": 0.7076, "grad_norm": 0.7522736191749573, "learning_rate": 0.0002, "epoch": 2.5980114784576833, "step": 16070}, {"loss": 0.6477, "grad_norm": 0.8149648308753967, "learning_rate": 0.0002, "epoch": 2.5996281626384286, "step": 16080}, {"loss": 0.6664, "grad_norm": 0.6214233040809631, "learning_rate": 0.0002, "epoch": 2.601244846819174, "step": 16090}, {"loss": 0.7307, "grad_norm": 0.6803743839263916, "learning_rate": 0.0002, "epoch": 2.602861530999919, "step": 16100}, {"loss": 0.7244, "grad_norm": 0.7223997116088867, "learning_rate": 0.0002, "epoch": 2.6044782151806647, "step": 16110}, {"loss": 0.6867, "grad_norm": 0.7324174642562866, "learning_rate": 0.0002, "epoch": 2.60609489936141, "step": 16120}, {"loss": 0.7159, "grad_norm": 0.9594739675521851, "learning_rate": 0.0002, "epoch": 2.607711583542155, "step": 16130}, {"loss": 0.6451, "grad_norm": 0.9485327005386353, "learning_rate": 0.0002, "epoch": 2.6093282677229004, "step": 16140}, {"loss": 0.6815, "grad_norm": 0.8449000120162964, "learning_rate": 0.0002, "epoch": 2.6109449519036456, "step": 16150}, {"loss": 0.7152, "grad_norm": 0.8520140051841736, "learning_rate": 0.0002, "epoch": 2.612561636084391, "step": 16160}, {"loss": 0.6759, "grad_norm": 0.7456524968147278, "learning_rate": 0.0002, "epoch": 2.614178320265136, "step": 16170}, {"loss": 0.6893, "grad_norm": 0.9912857413291931, "learning_rate": 0.0002, "epoch": 2.6157950044458813, "step": 16180}, {"loss": 0.7243, "grad_norm": 0.9001946449279785, "learning_rate": 0.0002, "epoch": 2.6174116886266265, "step": 16190}, {"loss": 0.6825, "grad_norm": 0.6568667888641357, "learning_rate": 0.0002, "epoch": 2.619028372807372, "step": 16200}, {"loss": 0.7013, "grad_norm": 1.0248128175735474, "learning_rate": 0.0002, "epoch": 2.6206450569881174, "step": 16210}, {"loss": 0.7045, "grad_norm": 0.6509039998054504, "learning_rate": 0.0002, "epoch": 2.6222617411688627, "step": 16220}, {"loss": 0.72, "grad_norm": 0.7626351118087769, "learning_rate": 0.0002, "epoch": 2.623878425349608, "step": 16230}, {"loss": 0.6556, "grad_norm": 0.6938552260398865, "learning_rate": 0.0002, "epoch": 2.625495109530353, "step": 16240}, {"loss": 0.65, "grad_norm": 0.6434680819511414, "learning_rate": 0.0002, "epoch": 2.6271117937110984, "step": 16250}, {"loss": 0.6943, "grad_norm": 0.7111515998840332, "learning_rate": 0.0002, "epoch": 2.628728477891844, "step": 16260}, {"loss": 0.679, "grad_norm": 0.7712395787239075, "learning_rate": 0.0002, "epoch": 2.6303451620725893, "step": 16270}, {"loss": 0.6886, "grad_norm": 0.792209267616272, "learning_rate": 0.0002, "epoch": 2.6319618462533345, "step": 16280}, {"loss": 0.6554, "grad_norm": 0.6801066398620605, "learning_rate": 0.0002, "epoch": 2.6335785304340797, "step": 16290}, {"loss": 0.73, "grad_norm": 0.7802573442459106, "learning_rate": 0.0002, "epoch": 2.635195214614825, "step": 16300}, {"loss": 0.7484, "grad_norm": 0.7742244601249695, "learning_rate": 0.0002, "epoch": 2.63681189879557, "step": 16310}, {"loss": 0.6524, "grad_norm": 0.664184033870697, "learning_rate": 0.0002, "epoch": 2.6384285829763154, "step": 16320}, {"loss": 0.6442, "grad_norm": 0.9242228865623474, "learning_rate": 0.0002, "epoch": 2.6400452671570607, "step": 16330}, {"loss": 0.6792, "grad_norm": 0.9661325216293335, "learning_rate": 0.0002, "epoch": 2.641661951337806, "step": 16340}, {"loss": 0.6847, "grad_norm": 0.837526798248291, "learning_rate": 0.0002, "epoch": 2.6432786355185516, "step": 16350}, {"loss": 0.7686, "grad_norm": 1.1834373474121094, "learning_rate": 0.0002, "epoch": 2.644895319699297, "step": 16360}, {"loss": 0.6746, "grad_norm": 0.7467831373214722, "learning_rate": 0.0002, "epoch": 2.646512003880042, "step": 16370}, {"loss": 0.6935, "grad_norm": 0.8627146482467651, "learning_rate": 0.0002, "epoch": 2.6481286880607873, "step": 16380}, {"loss": 0.715, "grad_norm": 0.790447473526001, "learning_rate": 0.0002, "epoch": 2.6497453722415325, "step": 16390}, {"loss": 0.723, "grad_norm": 0.8447365164756775, "learning_rate": 0.0002, "epoch": 2.651362056422278, "step": 16400}, {"loss": 0.6628, "grad_norm": 0.7831417918205261, "learning_rate": 0.0002, "epoch": 2.6529787406030234, "step": 16410}, {"loss": 0.6691, "grad_norm": 0.6837952136993408, "learning_rate": 0.0002, "epoch": 2.6545954247837686, "step": 16420}, {"loss": 0.6139, "grad_norm": 0.7031801342964172, "learning_rate": 0.0002, "epoch": 2.656212108964514, "step": 16430}, {"loss": 0.7382, "grad_norm": 0.8963770866394043, "learning_rate": 0.0002, "epoch": 2.657828793145259, "step": 16440}, {"loss": 0.6439, "grad_norm": 0.6852328181266785, "learning_rate": 0.0002, "epoch": 2.6594454773260043, "step": 16450}, {"loss": 0.6278, "grad_norm": 0.8069294095039368, "learning_rate": 0.0002, "epoch": 2.6610621615067496, "step": 16460}, {"loss": 0.6939, "grad_norm": 0.7503686547279358, "learning_rate": 0.0002, "epoch": 2.662678845687495, "step": 16470}, {"loss": 0.6777, "grad_norm": 0.6430956125259399, "learning_rate": 0.0002, "epoch": 2.66429552986824, "step": 16480}, {"loss": 0.6863, "grad_norm": 0.7894312739372253, "learning_rate": 0.0002, "epoch": 2.6659122140489853, "step": 16490}, {"loss": 0.7165, "grad_norm": 0.7277431488037109, "learning_rate": 0.0002, "epoch": 2.667528898229731, "step": 16500}, {"loss": 0.6772, "grad_norm": 0.6816153526306152, "learning_rate": 0.0002, "epoch": 2.669145582410476, "step": 16510}, {"loss": 0.691, "grad_norm": 0.8145235776901245, "learning_rate": 0.0002, "epoch": 2.6707622665912214, "step": 16520}, {"loss": 0.709, "grad_norm": 0.8645890355110168, "learning_rate": 0.0002, "epoch": 2.6723789507719666, "step": 16530}, {"loss": 0.6946, "grad_norm": 0.704393208026886, "learning_rate": 0.0002, "epoch": 2.673995634952712, "step": 16540}, {"loss": 0.6378, "grad_norm": 1.0120846033096313, "learning_rate": 0.0002, "epoch": 2.6756123191334575, "step": 16550}, {"loss": 0.7241, "grad_norm": 0.6919328570365906, "learning_rate": 0.0002, "epoch": 2.6772290033142028, "step": 16560}, {"loss": 0.7098, "grad_norm": 0.6924574971199036, "learning_rate": 0.0002, "epoch": 2.678845687494948, "step": 16570}, {"loss": 0.731, "grad_norm": 0.9679301381111145, "learning_rate": 0.0002, "epoch": 2.6804623716756932, "step": 16580}, {"loss": 0.7124, "grad_norm": 0.6810211539268494, "learning_rate": 0.0002, "epoch": 2.6820790558564385, "step": 16590}, {"loss": 0.6688, "grad_norm": 0.9730555415153503, "learning_rate": 0.0002, "epoch": 2.6836957400371837, "step": 16600}, {"loss": 0.7344, "grad_norm": 0.7852821350097656, "learning_rate": 0.0002, "epoch": 2.685312424217929, "step": 16610}, {"loss": 0.6401, "grad_norm": 0.6059057116508484, "learning_rate": 0.0002, "epoch": 2.686929108398674, "step": 16620}, {"loss": 0.6796, "grad_norm": 0.9395958781242371, "learning_rate": 0.0002, "epoch": 2.6885457925794194, "step": 16630}, {"loss": 0.7174, "grad_norm": 0.7473729848861694, "learning_rate": 0.0002, "epoch": 2.690162476760165, "step": 16640}, {"loss": 0.7087, "grad_norm": 0.765934407711029, "learning_rate": 0.0002, "epoch": 2.6917791609409103, "step": 16650}, {"loss": 0.707, "grad_norm": 0.8496677279472351, "learning_rate": 0.0002, "epoch": 2.6933958451216555, "step": 16660}, {"loss": 0.7084, "grad_norm": 0.7641879916191101, "learning_rate": 0.0002, "epoch": 2.6950125293024008, "step": 16670}, {"loss": 0.6566, "grad_norm": 0.8471952676773071, "learning_rate": 0.0002, "epoch": 2.696629213483146, "step": 16680}, {"loss": 0.6635, "grad_norm": 0.6946060657501221, "learning_rate": 0.0002, "epoch": 2.6982458976638912, "step": 16690}, {"loss": 0.7027, "grad_norm": 0.7361312508583069, "learning_rate": 0.0002, "epoch": 2.699862581844637, "step": 16700}, {"loss": 0.6767, "grad_norm": 0.6605038046836853, "learning_rate": 0.0002, "epoch": 2.701479266025382, "step": 16710}, {"loss": 0.6885, "grad_norm": 0.7164411544799805, "learning_rate": 0.0002, "epoch": 2.7030959502061274, "step": 16720}, {"loss": 0.6736, "grad_norm": 0.6496201157569885, "learning_rate": 0.0002, "epoch": 2.7047126343868726, "step": 16730}, {"loss": 0.6942, "grad_norm": 0.7826663851737976, "learning_rate": 0.0002, "epoch": 2.706329318567618, "step": 16740}, {"loss": 0.6773, "grad_norm": 0.7639131546020508, "learning_rate": 0.0002, "epoch": 2.707946002748363, "step": 16750}, {"loss": 0.69, "grad_norm": 0.7976210713386536, "learning_rate": 0.0002, "epoch": 2.7095626869291083, "step": 16760}, {"loss": 0.6735, "grad_norm": 0.6836577653884888, "learning_rate": 0.0002, "epoch": 2.7111793711098535, "step": 16770}, {"loss": 0.6596, "grad_norm": 0.8025202751159668, "learning_rate": 0.0002, "epoch": 2.7127960552905988, "step": 16780}, {"loss": 0.6324, "grad_norm": 0.7636463642120361, "learning_rate": 0.0002, "epoch": 2.7144127394713444, "step": 16790}, {"loss": 0.6227, "grad_norm": 0.7481677532196045, "learning_rate": 0.0002, "epoch": 2.7160294236520897, "step": 16800}, {"loss": 0.6925, "grad_norm": 0.7566834688186646, "learning_rate": 0.0002, "epoch": 2.717646107832835, "step": 16810}, {"loss": 0.6531, "grad_norm": 0.7931267619132996, "learning_rate": 0.0002, "epoch": 2.71926279201358, "step": 16820}, {"loss": 0.6672, "grad_norm": 0.8811662197113037, "learning_rate": 0.0002, "epoch": 2.7208794761943254, "step": 16830}, {"loss": 0.6675, "grad_norm": 0.8561240434646606, "learning_rate": 0.0002, "epoch": 2.7224961603750706, "step": 16840}, {"loss": 0.7135, "grad_norm": 0.7121599316596985, "learning_rate": 0.0002, "epoch": 2.7241128445558163, "step": 16850}, {"loss": 0.6825, "grad_norm": 0.8066257238388062, "learning_rate": 0.0002, "epoch": 2.7257295287365615, "step": 16860}, {"loss": 0.6839, "grad_norm": 0.7699271440505981, "learning_rate": 0.0002, "epoch": 2.7273462129173067, "step": 16870}, {"loss": 0.699, "grad_norm": 1.1828432083129883, "learning_rate": 0.0002, "epoch": 2.728962897098052, "step": 16880}, {"loss": 0.6518, "grad_norm": 0.9989302754402161, "learning_rate": 0.0002, "epoch": 2.730579581278797, "step": 16890}, {"loss": 0.7015, "grad_norm": 0.8100560307502747, "learning_rate": 0.0002, "epoch": 2.7321962654595424, "step": 16900}, {"loss": 0.6851, "grad_norm": 0.8615233898162842, "learning_rate": 0.0002, "epoch": 2.7338129496402876, "step": 16910}, {"loss": 0.6322, "grad_norm": 0.8633756041526794, "learning_rate": 0.0002, "epoch": 2.735429633821033, "step": 16920}, {"loss": 0.6488, "grad_norm": 0.7769348621368408, "learning_rate": 0.0002, "epoch": 2.737046318001778, "step": 16930}, {"loss": 0.6582, "grad_norm": 0.6943058371543884, "learning_rate": 0.0002, "epoch": 2.738663002182524, "step": 16940}, {"loss": 0.6516, "grad_norm": 0.8510736227035522, "learning_rate": 0.0002, "epoch": 2.740279686363269, "step": 16950}, {"loss": 0.7275, "grad_norm": 0.7732602953910828, "learning_rate": 0.0002, "epoch": 2.7418963705440142, "step": 16960}, {"loss": 0.6553, "grad_norm": 0.5981788635253906, "learning_rate": 0.0002, "epoch": 2.7435130547247595, "step": 16970}, {"loss": 0.6777, "grad_norm": 0.7604416012763977, "learning_rate": 0.0002, "epoch": 2.7451297389055047, "step": 16980}, {"loss": 0.6981, "grad_norm": 0.7377738356590271, "learning_rate": 0.0002, "epoch": 2.74674642308625, "step": 16990}, {"loss": 0.6294, "grad_norm": 0.9400289058685303, "learning_rate": 0.0002, "epoch": 2.7483631072669956, "step": 17000}, {"loss": 0.6952, "grad_norm": 0.6340599656105042, "learning_rate": 0.0002, "epoch": 2.749979791447741, "step": 17010}, {"loss": 0.7222, "grad_norm": 0.7297601103782654, "learning_rate": 0.0002, "epoch": 2.751596475628486, "step": 17020}, {"loss": 0.6659, "grad_norm": 0.9479979872703552, "learning_rate": 0.0002, "epoch": 2.7532131598092313, "step": 17030}, {"loss": 0.691, "grad_norm": 0.8461511135101318, "learning_rate": 0.0002, "epoch": 2.7548298439899765, "step": 17040}, {"loss": 0.6764, "grad_norm": 0.7477551698684692, "learning_rate": 0.0002, "epoch": 2.7564465281707218, "step": 17050}, {"loss": 0.684, "grad_norm": 1.019270420074463, "learning_rate": 0.0002, "epoch": 2.758063212351467, "step": 17060}, {"loss": 0.7119, "grad_norm": 0.7730235457420349, "learning_rate": 0.0002, "epoch": 2.7596798965322122, "step": 17070}, {"loss": 0.6886, "grad_norm": 0.8216866254806519, "learning_rate": 0.0002, "epoch": 2.7612965807129575, "step": 17080}, {"loss": 0.6811, "grad_norm": 0.7235931754112244, "learning_rate": 0.0002, "epoch": 2.762913264893703, "step": 17090}, {"loss": 0.7031, "grad_norm": 0.7352296710014343, "learning_rate": 0.0002, "epoch": 2.7645299490744484, "step": 17100}, {"loss": 0.6951, "grad_norm": 0.8129373788833618, "learning_rate": 0.0002, "epoch": 2.7661466332551936, "step": 17110}, {"loss": 0.6703, "grad_norm": 0.7387019991874695, "learning_rate": 0.0002, "epoch": 2.767763317435939, "step": 17120}, {"loss": 0.6789, "grad_norm": 0.9149190187454224, "learning_rate": 0.0002, "epoch": 2.769380001616684, "step": 17130}, {"loss": 0.6038, "grad_norm": 0.7352971434593201, "learning_rate": 0.0002, "epoch": 2.7709966857974297, "step": 17140}, {"loss": 0.6728, "grad_norm": 0.7903780341148376, "learning_rate": 0.0002, "epoch": 2.772613369978175, "step": 17150}, {"loss": 0.6988, "grad_norm": 0.8255927562713623, "learning_rate": 0.0002, "epoch": 2.77423005415892, "step": 17160}, {"loss": 0.6694, "grad_norm": 0.7235927581787109, "learning_rate": 0.0002, "epoch": 2.7758467383396654, "step": 17170}, {"loss": 0.7161, "grad_norm": 0.8281434774398804, "learning_rate": 0.0002, "epoch": 2.7774634225204107, "step": 17180}, {"loss": 0.682, "grad_norm": 0.7586921453475952, "learning_rate": 0.0002, "epoch": 2.779080106701156, "step": 17190}, {"loss": 0.6427, "grad_norm": 0.7161715030670166, "learning_rate": 0.0002, "epoch": 2.780696790881901, "step": 17200}, {"loss": 0.6426, "grad_norm": 0.762868344783783, "learning_rate": 0.0002, "epoch": 2.7823134750626464, "step": 17210}, {"loss": 0.705, "grad_norm": 0.9285483360290527, "learning_rate": 0.0002, "epoch": 2.7839301592433916, "step": 17220}, {"loss": 0.7084, "grad_norm": 0.6900462508201599, "learning_rate": 0.0002, "epoch": 2.785546843424137, "step": 17230}, {"loss": 0.6988, "grad_norm": 0.780384361743927, "learning_rate": 0.0002, "epoch": 2.7871635276048825, "step": 17240}, {"loss": 0.7073, "grad_norm": 0.7580406665802002, "learning_rate": 0.0002, "epoch": 2.7887802117856277, "step": 17250}, {"loss": 0.6833, "grad_norm": 0.8145199418067932, "learning_rate": 0.0002, "epoch": 2.790396895966373, "step": 17260}, {"loss": 0.6909, "grad_norm": 0.9159596562385559, "learning_rate": 0.0002, "epoch": 2.792013580147118, "step": 17270}, {"loss": 0.6008, "grad_norm": 0.9590014219284058, "learning_rate": 0.0002, "epoch": 2.7936302643278634, "step": 17280}, {"loss": 0.6704, "grad_norm": 0.7603529691696167, "learning_rate": 0.0002, "epoch": 2.795246948508609, "step": 17290}, {"loss": 0.7165, "grad_norm": 0.8039976358413696, "learning_rate": 0.0002, "epoch": 2.7968636326893543, "step": 17300}, {"loss": 0.7037, "grad_norm": 0.8364847302436829, "learning_rate": 0.0002, "epoch": 2.7984803168700996, "step": 17310}, {"loss": 0.6749, "grad_norm": 0.8763046860694885, "learning_rate": 0.0002, "epoch": 2.800097001050845, "step": 17320}, {"loss": 0.6844, "grad_norm": 0.8409647941589355, "learning_rate": 0.0002, "epoch": 2.80171368523159, "step": 17330}, {"loss": 0.6936, "grad_norm": 0.7649006247520447, "learning_rate": 0.0002, "epoch": 2.8033303694123353, "step": 17340}, {"loss": 0.7051, "grad_norm": 0.7970262169837952, "learning_rate": 0.0002, "epoch": 2.8049470535930805, "step": 17350}, {"loss": 0.6533, "grad_norm": 0.9088607430458069, "learning_rate": 0.0002, "epoch": 2.8065637377738257, "step": 17360}, {"loss": 0.675, "grad_norm": 0.6454846858978271, "learning_rate": 0.0002, "epoch": 2.808180421954571, "step": 17370}, {"loss": 0.7069, "grad_norm": 0.7744787931442261, "learning_rate": 0.0002, "epoch": 2.809797106135316, "step": 17380}, {"loss": 0.6772, "grad_norm": 0.6678640842437744, "learning_rate": 0.0002, "epoch": 2.811413790316062, "step": 17390}, {"loss": 0.6784, "grad_norm": 0.772676944732666, "learning_rate": 0.0002, "epoch": 2.813030474496807, "step": 17400}, {"loss": 0.7252, "grad_norm": 0.7088175415992737, "learning_rate": 0.0002, "epoch": 2.8146471586775523, "step": 17410}, {"loss": 0.7086, "grad_norm": 0.8280573487281799, "learning_rate": 0.0002, "epoch": 2.8162638428582976, "step": 17420}, {"loss": 0.6732, "grad_norm": 0.6665388345718384, "learning_rate": 0.0002, "epoch": 2.817880527039043, "step": 17430}, {"loss": 0.6675, "grad_norm": 0.6427883505821228, "learning_rate": 0.0002, "epoch": 2.8194972112197885, "step": 17440}, {"loss": 0.6972, "grad_norm": 0.9697760343551636, "learning_rate": 0.0002, "epoch": 2.8211138954005337, "step": 17450}, {"loss": 0.6838, "grad_norm": 0.7573966383934021, "learning_rate": 0.0002, "epoch": 2.822730579581279, "step": 17460}, {"loss": 0.7243, "grad_norm": 0.878688633441925, "learning_rate": 0.0002, "epoch": 2.824347263762024, "step": 17470}, {"loss": 0.6666, "grad_norm": 0.7752242684364319, "learning_rate": 0.0002, "epoch": 2.8259639479427694, "step": 17480}, {"loss": 0.6638, "grad_norm": 0.6135398745536804, "learning_rate": 0.0002, "epoch": 2.8275806321235146, "step": 17490}, {"loss": 0.6829, "grad_norm": 0.6924924850463867, "learning_rate": 0.0002, "epoch": 2.82919731630426, "step": 17500}, {"loss": 0.6731, "grad_norm": 0.7471627593040466, "learning_rate": 0.0002, "epoch": 2.830814000485005, "step": 17510}, {"loss": 0.7016, "grad_norm": 0.7145499587059021, "learning_rate": 0.0002, "epoch": 2.8324306846657503, "step": 17520}, {"loss": 0.6787, "grad_norm": 0.7415414452552795, "learning_rate": 0.0002, "epoch": 2.834047368846496, "step": 17530}, {"loss": 0.6811, "grad_norm": 0.7328441739082336, "learning_rate": 0.0002, "epoch": 2.8356640530272412, "step": 17540}, {"loss": 0.6866, "grad_norm": 0.8267839550971985, "learning_rate": 0.0002, "epoch": 2.8372807372079865, "step": 17550}, {"loss": 0.6787, "grad_norm": 0.8877885341644287, "learning_rate": 0.0002, "epoch": 2.8388974213887317, "step": 17560}, {"loss": 0.7136, "grad_norm": 0.857138454914093, "learning_rate": 0.0002, "epoch": 2.840514105569477, "step": 17570}, {"loss": 0.6454, "grad_norm": 0.8470779657363892, "learning_rate": 0.0002, "epoch": 2.842130789750222, "step": 17580}, {"loss": 0.6976, "grad_norm": 0.8553254008293152, "learning_rate": 0.0002, "epoch": 2.843747473930968, "step": 17590}, {"loss": 0.7297, "grad_norm": 0.8033196926116943, "learning_rate": 0.0002, "epoch": 2.845364158111713, "step": 17600}, {"loss": 0.7062, "grad_norm": 0.7949087023735046, "learning_rate": 0.0002, "epoch": 2.8469808422924583, "step": 17610}, {"loss": 0.651, "grad_norm": 0.9241406321525574, "learning_rate": 0.0002, "epoch": 2.8485975264732035, "step": 17620}, {"loss": 0.6601, "grad_norm": 0.7721285223960876, "learning_rate": 0.0002, "epoch": 2.8502142106539488, "step": 17630}, {"loss": 0.6183, "grad_norm": 1.0246692895889282, "learning_rate": 0.0002, "epoch": 2.851830894834694, "step": 17640}, {"loss": 0.7007, "grad_norm": 0.9244589805603027, "learning_rate": 0.0002, "epoch": 2.853447579015439, "step": 17650}, {"loss": 0.7274, "grad_norm": 0.7243508696556091, "learning_rate": 0.0002, "epoch": 2.8550642631961844, "step": 17660}, {"loss": 0.6471, "grad_norm": 0.8943371176719666, "learning_rate": 0.0002, "epoch": 2.8566809473769297, "step": 17670}, {"loss": 0.686, "grad_norm": 0.6531758904457092, "learning_rate": 0.0002, "epoch": 2.8582976315576754, "step": 17680}, {"loss": 0.6253, "grad_norm": 0.8367000818252563, "learning_rate": 0.0002, "epoch": 2.8599143157384206, "step": 17690}, {"loss": 0.6943, "grad_norm": 0.7868556380271912, "learning_rate": 0.0002, "epoch": 2.861530999919166, "step": 17700}, {"loss": 0.6919, "grad_norm": 0.7213859558105469, "learning_rate": 0.0002, "epoch": 2.863147684099911, "step": 17710}, {"loss": 0.6657, "grad_norm": 0.7383931279182434, "learning_rate": 0.0002, "epoch": 2.8647643682806563, "step": 17720}, {"loss": 0.6841, "grad_norm": 0.7566812634468079, "learning_rate": 0.0002, "epoch": 2.8663810524614015, "step": 17730}, {"loss": 0.6449, "grad_norm": 0.6930373311042786, "learning_rate": 0.0002, "epoch": 2.867997736642147, "step": 17740}, {"loss": 0.6764, "grad_norm": 0.7911090850830078, "learning_rate": 0.0002, "epoch": 2.8696144208228924, "step": 17750}, {"loss": 0.6554, "grad_norm": 0.8484548926353455, "learning_rate": 0.0002, "epoch": 2.8712311050036377, "step": 17760}, {"loss": 0.6931, "grad_norm": 0.7647597193717957, "learning_rate": 0.0002, "epoch": 2.872847789184383, "step": 17770}, {"loss": 0.6945, "grad_norm": 0.8791151642799377, "learning_rate": 0.0002, "epoch": 2.874464473365128, "step": 17780}, {"loss": 0.7078, "grad_norm": 0.7253178358078003, "learning_rate": 0.0002, "epoch": 2.8760811575458733, "step": 17790}, {"loss": 0.6474, "grad_norm": 0.7956077456474304, "learning_rate": 0.0002, "epoch": 2.8776978417266186, "step": 17800}, {"loss": 0.6687, "grad_norm": 0.8657688498497009, "learning_rate": 0.0002, "epoch": 2.879314525907364, "step": 17810}, {"loss": 0.7171, "grad_norm": 0.7059141993522644, "learning_rate": 0.0002, "epoch": 2.880931210088109, "step": 17820}, {"loss": 0.683, "grad_norm": 0.8886896967887878, "learning_rate": 0.0002, "epoch": 2.8825478942688547, "step": 17830}, {"loss": 0.669, "grad_norm": 0.821032702922821, "learning_rate": 0.0002, "epoch": 2.8841645784496, "step": 17840}, {"loss": 0.6805, "grad_norm": 0.7183963656425476, "learning_rate": 0.0002, "epoch": 2.885781262630345, "step": 17850}, {"loss": 0.7088, "grad_norm": 0.6222899556159973, "learning_rate": 0.0002, "epoch": 2.8873979468110904, "step": 17860}, {"loss": 0.6626, "grad_norm": 0.8187434077262878, "learning_rate": 0.0002, "epoch": 2.8890146309918356, "step": 17870}, {"loss": 0.6815, "grad_norm": 0.9838479161262512, "learning_rate": 0.0002, "epoch": 2.890631315172581, "step": 17880}, {"loss": 0.6967, "grad_norm": 0.7567742466926575, "learning_rate": 0.0002, "epoch": 2.8922479993533265, "step": 17890}, {"loss": 0.7073, "grad_norm": 0.6875903606414795, "learning_rate": 0.0002, "epoch": 2.893864683534072, "step": 17900}, {"loss": 0.6415, "grad_norm": 0.8043789267539978, "learning_rate": 0.0002, "epoch": 2.895481367714817, "step": 17910}, {"loss": 0.6588, "grad_norm": 0.8062626719474792, "learning_rate": 0.0002, "epoch": 2.8970980518955622, "step": 17920}, {"loss": 0.7151, "grad_norm": 1.0251191854476929, "learning_rate": 0.0002, "epoch": 2.8987147360763075, "step": 17930}, {"loss": 0.6605, "grad_norm": 0.882253110408783, "learning_rate": 0.0002, "epoch": 2.9003314202570527, "step": 17940}, {"loss": 0.6719, "grad_norm": 0.8683299422264099, "learning_rate": 0.0002, "epoch": 2.901948104437798, "step": 17950}, {"loss": 0.6896, "grad_norm": 0.7167282104492188, "learning_rate": 0.0002, "epoch": 2.903564788618543, "step": 17960}, {"loss": 0.663, "grad_norm": 0.7093694806098938, "learning_rate": 0.0002, "epoch": 2.9051814727992884, "step": 17970}, {"loss": 0.6591, "grad_norm": 0.8549879193305969, "learning_rate": 0.0002, "epoch": 2.906798156980034, "step": 17980}, {"loss": 0.6962, "grad_norm": 0.6989606618881226, "learning_rate": 0.0002, "epoch": 2.9084148411607793, "step": 17990}, {"loss": 0.6635, "grad_norm": 0.9482976794242859, "learning_rate": 0.0002, "epoch": 2.9100315253415245, "step": 18000}, {"loss": 0.6586, "grad_norm": 0.7182440161705017, "learning_rate": 0.0002, "epoch": 2.9116482095222698, "step": 18010}, {"loss": 0.6827, "grad_norm": 0.7732226252555847, "learning_rate": 0.0002, "epoch": 2.913264893703015, "step": 18020}, {"loss": 0.7123, "grad_norm": 0.7936875224113464, "learning_rate": 0.0002, "epoch": 2.9148815778837607, "step": 18030}, {"loss": 0.6736, "grad_norm": 0.8825615644454956, "learning_rate": 0.0002, "epoch": 2.916498262064506, "step": 18040}, {"loss": 0.7139, "grad_norm": 0.6778587102890015, "learning_rate": 0.0002, "epoch": 2.918114946245251, "step": 18050}, {"loss": 0.6588, "grad_norm": 0.7529265880584717, "learning_rate": 0.0002, "epoch": 2.9197316304259964, "step": 18060}, {"loss": 0.737, "grad_norm": 0.7111883163452148, "learning_rate": 0.0002, "epoch": 2.9213483146067416, "step": 18070}, {"loss": 0.7475, "grad_norm": 0.7214767932891846, "learning_rate": 0.0002, "epoch": 2.922964998787487, "step": 18080}, {"loss": 0.6672, "grad_norm": 0.800417423248291, "learning_rate": 0.0002, "epoch": 2.924581682968232, "step": 18090}, {"loss": 0.6694, "grad_norm": 1.248575210571289, "learning_rate": 0.0002, "epoch": 2.9261983671489773, "step": 18100}, {"loss": 0.7004, "grad_norm": 0.757788360118866, "learning_rate": 0.0002, "epoch": 2.9278150513297225, "step": 18110}, {"loss": 0.6999, "grad_norm": 1.0583995580673218, "learning_rate": 0.0002, "epoch": 2.9294317355104678, "step": 18120}, {"loss": 0.6365, "grad_norm": 0.8228777647018433, "learning_rate": 0.0002, "epoch": 2.9310484196912134, "step": 18130}, {"loss": 0.6791, "grad_norm": 0.8374035358428955, "learning_rate": 0.0002, "epoch": 2.9326651038719587, "step": 18140}, {"loss": 0.6399, "grad_norm": 0.7976473569869995, "learning_rate": 0.0002, "epoch": 2.934281788052704, "step": 18150}, {"loss": 0.6585, "grad_norm": 0.8009907603263855, "learning_rate": 0.0002, "epoch": 2.935898472233449, "step": 18160}, {"loss": 0.7485, "grad_norm": 0.835213303565979, "learning_rate": 0.0002, "epoch": 2.9375151564141944, "step": 18170}, {"loss": 0.7376, "grad_norm": 0.7982219457626343, "learning_rate": 0.0002, "epoch": 2.93913184059494, "step": 18180}, {"loss": 0.6348, "grad_norm": 0.7070978879928589, "learning_rate": 0.0002, "epoch": 2.9407485247756853, "step": 18190}, {"loss": 0.6608, "grad_norm": 0.8619440197944641, "learning_rate": 0.0002, "epoch": 2.9423652089564305, "step": 18200}, {"loss": 0.666, "grad_norm": 0.6693987250328064, "learning_rate": 0.0002, "epoch": 2.9439818931371757, "step": 18210}, {"loss": 0.728, "grad_norm": 0.6747021079063416, "learning_rate": 0.0002, "epoch": 2.945598577317921, "step": 18220}, {"loss": 0.6686, "grad_norm": 0.860387921333313, "learning_rate": 0.0002, "epoch": 2.947215261498666, "step": 18230}, {"loss": 0.6945, "grad_norm": 0.799976646900177, "learning_rate": 0.0002, "epoch": 2.9488319456794114, "step": 18240}, {"loss": 0.7243, "grad_norm": 0.7864769101142883, "learning_rate": 0.0002, "epoch": 2.9504486298601567, "step": 18250}, {"loss": 0.6785, "grad_norm": 0.6713884472846985, "learning_rate": 0.0002, "epoch": 2.952065314040902, "step": 18260}, {"loss": 0.7429, "grad_norm": 0.9031508564949036, "learning_rate": 0.0002, "epoch": 2.9536819982216476, "step": 18270}, {"loss": 0.7055, "grad_norm": 0.7205073237419128, "learning_rate": 0.0002, "epoch": 2.955298682402393, "step": 18280}, {"loss": 0.7298, "grad_norm": 0.7746205925941467, "learning_rate": 0.0002, "epoch": 2.956915366583138, "step": 18290}, {"loss": 0.6218, "grad_norm": 0.6533427834510803, "learning_rate": 0.0002, "epoch": 2.9585320507638833, "step": 18300}, {"loss": 0.6674, "grad_norm": 0.9083208441734314, "learning_rate": 0.0002, "epoch": 2.9601487349446285, "step": 18310}, {"loss": 0.7359, "grad_norm": 0.7446991801261902, "learning_rate": 0.0002, "epoch": 2.9617654191253737, "step": 18320}, {"loss": 0.6738, "grad_norm": 0.6514461636543274, "learning_rate": 0.0002, "epoch": 2.9633821033061194, "step": 18330}, {"loss": 0.6677, "grad_norm": 0.8580465912818909, "learning_rate": 0.0002, "epoch": 2.9649987874868646, "step": 18340}, {"loss": 0.6971, "grad_norm": 0.7074266076087952, "learning_rate": 0.0002, "epoch": 2.96661547166761, "step": 18350}, {"loss": 0.6804, "grad_norm": 0.899892270565033, "learning_rate": 0.0002, "epoch": 2.968232155848355, "step": 18360}, {"loss": 0.7094, "grad_norm": 0.8217641711235046, "learning_rate": 0.0002, "epoch": 2.9698488400291003, "step": 18370}, {"loss": 0.6916, "grad_norm": 0.8611799478530884, "learning_rate": 0.0002, "epoch": 2.9714655242098456, "step": 18380}, {"loss": 0.6677, "grad_norm": 0.6909302473068237, "learning_rate": 0.0002, "epoch": 2.973082208390591, "step": 18390}, {"loss": 0.7247, "grad_norm": 0.6554358005523682, "learning_rate": 0.0002, "epoch": 2.974698892571336, "step": 18400}, {"loss": 0.6516, "grad_norm": 0.7803071737289429, "learning_rate": 0.0002, "epoch": 2.9763155767520812, "step": 18410}, {"loss": 0.7322, "grad_norm": 0.7838954925537109, "learning_rate": 0.0002, "epoch": 2.977932260932827, "step": 18420}, {"loss": 0.6522, "grad_norm": 0.7098495364189148, "learning_rate": 0.0002, "epoch": 2.979548945113572, "step": 18430}, {"loss": 0.739, "grad_norm": 0.8981785774230957, "learning_rate": 0.0002, "epoch": 2.9811656292943174, "step": 18440}, {"loss": 0.6689, "grad_norm": 0.7197171449661255, "learning_rate": 0.0002, "epoch": 2.9827823134750626, "step": 18450}, {"loss": 0.706, "grad_norm": 0.793185293674469, "learning_rate": 0.0002, "epoch": 2.984398997655808, "step": 18460}, {"loss": 0.7124, "grad_norm": 0.8531473875045776, "learning_rate": 0.0002, "epoch": 2.986015681836553, "step": 18470}, {"loss": 0.6901, "grad_norm": 0.6627361178398132, "learning_rate": 0.0002, "epoch": 2.9876323660172988, "step": 18480}, {"loss": 0.6591, "grad_norm": 0.5708155035972595, "learning_rate": 0.0002, "epoch": 2.989249050198044, "step": 18490}, {"loss": 0.6725, "grad_norm": 0.8227280378341675, "learning_rate": 0.0002, "epoch": 2.990865734378789, "step": 18500}, {"loss": 0.6701, "grad_norm": 0.7102749943733215, "learning_rate": 0.0002, "epoch": 2.9924824185595345, "step": 18510}, {"loss": 0.7091, "grad_norm": 0.839485228061676, "learning_rate": 0.0002, "epoch": 2.9940991027402797, "step": 18520}, {"loss": 0.6521, "grad_norm": 0.9038704037666321, "learning_rate": 0.0002, "epoch": 2.995715786921025, "step": 18530}, {"loss": 0.7186, "grad_norm": 0.8737510442733765, "learning_rate": 0.0002, "epoch": 2.99733247110177, "step": 18540}, {"loss": 0.6819, "grad_norm": 0.7323142886161804, "learning_rate": 0.0002, "epoch": 2.9989491552825154, "step": 18550}]}
+{"epoch": 4.0, "step": 24742, "epoch_duration": 16883.95777654648, "total_accumulated_duration": 67535.40600156784, "gpu_info": {"GPU_0": "NVIDIA A100-PCIE-40GB"}, "memory_usage": {"avg_memory_usage": {"GPU_0": 7568.4541015625}, "peak_memory_usage": {"GPU_0": 13792.75537109375}, "avg_memory_reserved": {"GPU_0": 17416.0}, "peak_memory_reserved": {"GPU_0": 17416.0}, "total_memory": {"GPU_0": 40444.375}}, "best_checkpoint_path": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185", "params": {"epochs": 8, "batch_size": 1, "learning_rate": 0.0002, "gradient_accumulation_steps": 2, "warmup_ratio": 0.03, "max_grad_norm": 0.3, "lora_alpha": 32, "lora_dropout": 0.05, "lora_r": 64, "tf32": true, "seed": 42}, "log_history": [{"loss": 1.6636, "grad_norm": 0.9894065856933594, "learning_rate": 0.0002, "epoch": 0.0016166841807452913, "step": 10}, {"loss": 1.1528, "grad_norm": 1.7810699939727783, "learning_rate": 0.0002, "epoch": 0.0032333683614905826, "step": 20}, {"loss": 0.9767, "grad_norm": 0.5969577431678772, "learning_rate": 0.0002, "epoch": 0.004850052542235874, "step": 30}, {"loss": 0.9772, "grad_norm": 0.6354120969772339, "learning_rate": 0.0002, "epoch": 0.006466736722981165, "step": 40}, {"loss": 0.8643, "grad_norm": 0.5604607462882996, "learning_rate": 0.0002, "epoch": 0.008083420903726457, "step": 50}, {"loss": 0.8841, "grad_norm": 0.4676193594932556, "learning_rate": 0.0002, "epoch": 0.009700105084471748, "step": 60}, {"loss": 0.9022, "grad_norm": 0.6099211573600769, "learning_rate": 0.0002, "epoch": 0.01131678926521704, "step": 70}, {"loss": 0.9133, "grad_norm": 0.48639994859695435, "learning_rate": 0.0002, "epoch": 0.01293347344596233, "step": 80}, {"loss": 0.8704, "grad_norm": 0.4904264509677887, "learning_rate": 0.0002, "epoch": 0.014550157626707623, "step": 90}, {"loss": 0.8855, "grad_norm": 2.8334362506866455, "learning_rate": 0.0002, "epoch": 0.016166841807452915, "step": 100}, {"loss": 0.8958, "grad_norm": 0.43221670389175415, "learning_rate": 0.0002, "epoch": 0.017783525988198205, "step": 110}, {"loss": 0.8412, "grad_norm": 0.42244166135787964, "learning_rate": 0.0002, "epoch": 0.019400210168943496, "step": 120}, {"loss": 0.8467, "grad_norm": 0.45363298058509827, "learning_rate": 0.0002, "epoch": 0.02101689434968879, "step": 130}, {"loss": 0.8641, "grad_norm": 0.44816508889198303, "learning_rate": 0.0002, "epoch": 0.02263357853043408, "step": 140}, {"loss": 0.8496, "grad_norm": 0.43308213353157043, "learning_rate": 0.0002, "epoch": 0.02425026271117937, "step": 150}, {"loss": 0.8213, "grad_norm": 0.4084763526916504, "learning_rate": 0.0002, "epoch": 0.02586694689192466, "step": 160}, {"loss": 0.8343, "grad_norm": 0.5363703966140747, "learning_rate": 0.0002, "epoch": 0.027483631072669955, "step": 170}, {"loss": 0.8558, "grad_norm": 0.4619699716567993, "learning_rate": 0.0002, "epoch": 0.029100315253415245, "step": 180}, {"loss": 0.8878, "grad_norm": 0.49069908261299133, "learning_rate": 0.0002, "epoch": 0.030716999434160536, "step": 190}, {"loss": 0.8867, "grad_norm": 0.4645835757255554, "learning_rate": 0.0002, "epoch": 0.03233368361490583, "step": 200}, {"loss": 0.8842, "grad_norm": 1.2411243915557861, "learning_rate": 0.0002, "epoch": 0.03395036779565112, "step": 210}, {"loss": 0.8245, "grad_norm": 0.5211851596832275, "learning_rate": 0.0002, "epoch": 0.03556705197639641, "step": 220}, {"loss": 0.8194, "grad_norm": 0.5253691673278809, "learning_rate": 0.0002, "epoch": 0.037183736157141704, "step": 230}, {"loss": 0.8856, "grad_norm": 0.4567478895187378, "learning_rate": 0.0002, "epoch": 0.03880042033788699, "step": 240}, {"loss": 0.838, "grad_norm": 0.5472128391265869, "learning_rate": 0.0002, "epoch": 0.040417104518632285, "step": 250}, {"loss": 0.8201, "grad_norm": 0.42978546023368835, "learning_rate": 0.0002, "epoch": 0.04203378869937758, "step": 260}, {"loss": 0.8334, "grad_norm": 0.601734459400177, "learning_rate": 0.0002, "epoch": 0.043650472880122866, "step": 270}, {"loss": 0.815, "grad_norm": 0.4286513328552246, "learning_rate": 0.0002, "epoch": 0.04526715706086816, "step": 280}, {"loss": 0.8758, "grad_norm": 0.5230861902236938, "learning_rate": 0.0002, "epoch": 0.046883841241613454, "step": 290}, {"loss": 0.8636, "grad_norm": 0.6504611968994141, "learning_rate": 0.0002, "epoch": 0.04850052542235874, "step": 300}, {"loss": 0.8102, "grad_norm": 0.43485215306282043, "learning_rate": 0.0002, "epoch": 0.050117209603104035, "step": 310}, {"loss": 0.8221, "grad_norm": 0.4717007875442505, "learning_rate": 0.0002, "epoch": 0.05173389378384932, "step": 320}, {"loss": 0.8469, "grad_norm": 0.4059787690639496, "learning_rate": 0.0002, "epoch": 0.053350577964594616, "step": 330}, {"loss": 0.8866, "grad_norm": 0.4366913437843323, "learning_rate": 0.0002, "epoch": 0.05496726214533991, "step": 340}, {"loss": 0.7976, "grad_norm": 0.4233848452568054, "learning_rate": 0.0002, "epoch": 0.0565839463260852, "step": 350}, {"loss": 0.8456, "grad_norm": 0.4209108352661133, "learning_rate": 0.0002, "epoch": 0.05820063050683049, "step": 360}, {"loss": 0.816, "grad_norm": 0.41637396812438965, "learning_rate": 0.0002, "epoch": 0.059817314687575784, "step": 370}, {"loss": 0.7976, "grad_norm": 0.46235376596450806, "learning_rate": 0.0002, "epoch": 0.06143399886832107, "step": 380}, {"loss": 0.7966, "grad_norm": 0.4013484716415405, "learning_rate": 0.0002, "epoch": 0.06305068304906636, "step": 390}, {"loss": 0.8253, "grad_norm": 0.47443896532058716, "learning_rate": 0.0002, "epoch": 0.06466736722981166, "step": 400}, {"loss": 0.8666, "grad_norm": 0.3942156434059143, "learning_rate": 0.0002, "epoch": 0.06628405141055695, "step": 410}, {"loss": 0.8402, "grad_norm": 0.4965320825576782, "learning_rate": 0.0002, "epoch": 0.06790073559130223, "step": 420}, {"loss": 0.8317, "grad_norm": 0.4304835796356201, "learning_rate": 0.0002, "epoch": 0.06951741977204753, "step": 430}, {"loss": 0.8528, "grad_norm": 0.511726975440979, "learning_rate": 0.0002, "epoch": 0.07113410395279282, "step": 440}, {"loss": 0.8675, "grad_norm": 0.4040689170360565, "learning_rate": 0.0002, "epoch": 0.07275078813353811, "step": 450}, {"loss": 0.8788, "grad_norm": 0.5402171015739441, "learning_rate": 0.0002, "epoch": 0.07436747231428341, "step": 460}, {"loss": 0.8737, "grad_norm": 0.4174517095088959, "learning_rate": 0.0002, "epoch": 0.0759841564950287, "step": 470}, {"loss": 0.7605, "grad_norm": 0.4306182265281677, "learning_rate": 0.0002, "epoch": 0.07760084067577398, "step": 480}, {"loss": 0.799, "grad_norm": 0.535210132598877, "learning_rate": 0.0002, "epoch": 0.07921752485651928, "step": 490}, {"loss": 0.7825, "grad_norm": 0.5339109897613525, "learning_rate": 0.0002, "epoch": 0.08083420903726457, "step": 500}, {"loss": 0.8985, "grad_norm": 0.45754891633987427, "learning_rate": 0.0002, "epoch": 0.08245089321800986, "step": 510}, {"loss": 0.8144, "grad_norm": 0.43820783495903015, "learning_rate": 0.0002, "epoch": 0.08406757739875516, "step": 520}, {"loss": 0.8001, "grad_norm": 0.4434749186038971, "learning_rate": 0.0002, "epoch": 0.08568426157950045, "step": 530}, {"loss": 0.7857, "grad_norm": 0.43111467361450195, "learning_rate": 0.0002, "epoch": 0.08730094576024573, "step": 540}, {"loss": 0.8418, "grad_norm": 0.4378940165042877, "learning_rate": 0.0002, "epoch": 0.08891762994099103, "step": 550}, {"loss": 0.8361, "grad_norm": 0.4772215187549591, "learning_rate": 0.0002, "epoch": 0.09053431412173632, "step": 560}, {"loss": 0.8268, "grad_norm": 0.6837629079818726, "learning_rate": 0.0002, "epoch": 0.09215099830248161, "step": 570}, {"loss": 0.8607, "grad_norm": 0.42241212725639343, "learning_rate": 0.0002, "epoch": 0.09376768248322691, "step": 580}, {"loss": 0.852, "grad_norm": 0.5165936350822449, "learning_rate": 0.0002, "epoch": 0.0953843666639722, "step": 590}, {"loss": 0.8664, "grad_norm": 0.48737478256225586, "learning_rate": 0.0002, "epoch": 0.09700105084471748, "step": 600}, {"loss": 0.8806, "grad_norm": 0.47419852018356323, "learning_rate": 0.0002, "epoch": 0.09861773502546278, "step": 610}, {"loss": 0.8254, "grad_norm": 0.4975486099720001, "learning_rate": 0.0002, "epoch": 0.10023441920620807, "step": 620}, {"loss": 0.8548, "grad_norm": 0.49123844504356384, "learning_rate": 0.0002, "epoch": 0.10185110338695336, "step": 630}, {"loss": 0.8911, "grad_norm": 0.6288952827453613, "learning_rate": 0.0002, "epoch": 0.10346778756769864, "step": 640}, {"loss": 0.827, "grad_norm": 0.4277345836162567, "learning_rate": 0.0002, "epoch": 0.10508447174844394, "step": 650}, {"loss": 0.7996, "grad_norm": 0.4021061956882477, "learning_rate": 0.0002, "epoch": 0.10670115592918923, "step": 660}, {"loss": 0.87, "grad_norm": 0.3492237329483032, "learning_rate": 0.0002, "epoch": 0.10831784010993452, "step": 670}, {"loss": 0.8698, "grad_norm": 0.4341012239456177, "learning_rate": 0.0002, "epoch": 0.10993452429067982, "step": 680}, {"loss": 0.781, "grad_norm": 0.7296304106712341, "learning_rate": 0.0002, "epoch": 0.1115512084714251, "step": 690}, {"loss": 0.8433, "grad_norm": 0.397494912147522, "learning_rate": 0.0002, "epoch": 0.1131678926521704, "step": 700}, {"loss": 0.827, "grad_norm": 0.396431028842926, "learning_rate": 0.0002, "epoch": 0.1147845768329157, "step": 710}, {"loss": 0.8379, "grad_norm": 0.48842838406562805, "learning_rate": 0.0002, "epoch": 0.11640126101366098, "step": 720}, {"loss": 0.8238, "grad_norm": 0.46322616934776306, "learning_rate": 0.0002, "epoch": 0.11801794519440627, "step": 730}, {"loss": 0.8041, "grad_norm": 0.47990912199020386, "learning_rate": 0.0002, "epoch": 0.11963462937515157, "step": 740}, {"loss": 0.82, "grad_norm": 0.4997142255306244, "learning_rate": 0.0002, "epoch": 0.12125131355589686, "step": 750}, {"loss": 0.7702, "grad_norm": 0.4040526747703552, "learning_rate": 0.0002, "epoch": 0.12286799773664214, "step": 760}, {"loss": 0.863, "grad_norm": 0.453095942735672, "learning_rate": 0.0002, "epoch": 0.12448468191738744, "step": 770}, {"loss": 0.8792, "grad_norm": 0.4636971950531006, "learning_rate": 0.0002, "epoch": 0.12610136609813272, "step": 780}, {"loss": 0.8112, "grad_norm": 0.4279276132583618, "learning_rate": 0.0002, "epoch": 0.12771805027887803, "step": 790}, {"loss": 0.8711, "grad_norm": 0.46212655305862427, "learning_rate": 0.0002, "epoch": 0.12933473445962332, "step": 800}, {"loss": 0.8368, "grad_norm": 0.43127650022506714, "learning_rate": 0.0002, "epoch": 0.1309514186403686, "step": 810}, {"loss": 0.8476, "grad_norm": 0.4201301336288452, "learning_rate": 0.0002, "epoch": 0.1325681028211139, "step": 820}, {"loss": 0.8078, "grad_norm": 0.42583167552948, "learning_rate": 0.0002, "epoch": 0.13418478700185918, "step": 830}, {"loss": 0.8219, "grad_norm": 0.4535622000694275, "learning_rate": 0.0002, "epoch": 0.13580147118260447, "step": 840}, {"loss": 0.8423, "grad_norm": 0.4116036891937256, "learning_rate": 0.0002, "epoch": 0.13741815536334978, "step": 850}, {"loss": 0.8466, "grad_norm": 0.45997580885887146, "learning_rate": 0.0002, "epoch": 0.13903483954409507, "step": 860}, {"loss": 0.8917, "grad_norm": 0.4487837255001068, "learning_rate": 0.0002, "epoch": 0.14065152372484035, "step": 870}, {"loss": 0.8217, "grad_norm": 0.43650057911872864, "learning_rate": 0.0002, "epoch": 0.14226820790558564, "step": 880}, {"loss": 0.8178, "grad_norm": 0.5335358381271362, "learning_rate": 0.0002, "epoch": 0.14388489208633093, "step": 890}, {"loss": 0.7957, "grad_norm": 0.5989000201225281, "learning_rate": 0.0002, "epoch": 0.14550157626707622, "step": 900}, {"loss": 0.8385, "grad_norm": 0.517179012298584, "learning_rate": 0.0002, "epoch": 0.14711826044782153, "step": 910}, {"loss": 0.8255, "grad_norm": 0.44435232877731323, "learning_rate": 0.0002, "epoch": 0.14873494462856682, "step": 920}, {"loss": 0.8305, "grad_norm": 0.42635923624038696, "learning_rate": 0.0002, "epoch": 0.1503516288093121, "step": 930}, {"loss": 0.8043, "grad_norm": 0.49603334069252014, "learning_rate": 0.0002, "epoch": 0.1519683129900574, "step": 940}, {"loss": 0.8377, "grad_norm": 0.40639808773994446, "learning_rate": 0.0002, "epoch": 0.15358499717080268, "step": 950}, {"loss": 0.8529, "grad_norm": 0.4850759208202362, "learning_rate": 0.0002, "epoch": 0.15520168135154797, "step": 960}, {"loss": 0.846, "grad_norm": 0.4427442252635956, "learning_rate": 0.0002, "epoch": 0.15681836553229328, "step": 970}, {"loss": 0.8705, "grad_norm": 0.3760930001735687, "learning_rate": 0.0002, "epoch": 0.15843504971303857, "step": 980}, {"loss": 0.8644, "grad_norm": 0.4794144332408905, "learning_rate": 0.0002, "epoch": 0.16005173389378385, "step": 990}, {"loss": 0.8002, "grad_norm": 0.45828768610954285, "learning_rate": 0.0002, "epoch": 0.16166841807452914, "step": 1000}, {"loss": 0.7658, "grad_norm": 0.6313053369522095, "learning_rate": 0.0002, "epoch": 0.16328510225527443, "step": 1010}, {"loss": 0.8047, "grad_norm": 0.45041006803512573, "learning_rate": 0.0002, "epoch": 0.16490178643601971, "step": 1020}, {"loss": 0.8423, "grad_norm": 0.441403865814209, "learning_rate": 0.0002, "epoch": 0.166518470616765, "step": 1030}, {"loss": 0.8475, "grad_norm": 0.8171296119689941, "learning_rate": 0.0002, "epoch": 0.16813515479751032, "step": 1040}, {"loss": 0.845, "grad_norm": 0.7137420773506165, "learning_rate": 0.0002, "epoch": 0.1697518389782556, "step": 1050}, {"loss": 0.8213, "grad_norm": 0.5236809849739075, "learning_rate": 0.0002, "epoch": 0.1713685231590009, "step": 1060}, {"loss": 0.8265, "grad_norm": 0.5021864175796509, "learning_rate": 0.0002, "epoch": 0.17298520733974618, "step": 1070}, {"loss": 0.8305, "grad_norm": 0.47347521781921387, "learning_rate": 0.0002, "epoch": 0.17460189152049146, "step": 1080}, {"loss": 0.8105, "grad_norm": 0.4631653428077698, "learning_rate": 0.0002, "epoch": 0.17621857570123675, "step": 1090}, {"loss": 0.8166, "grad_norm": 0.49169182777404785, "learning_rate": 0.0002, "epoch": 0.17783525988198207, "step": 1100}, {"loss": 0.8012, "grad_norm": 0.5019739270210266, "learning_rate": 0.0002, "epoch": 0.17945194406272735, "step": 1110}, {"loss": 0.8247, "grad_norm": 0.5100422501564026, "learning_rate": 0.0002, "epoch": 0.18106862824347264, "step": 1120}, {"loss": 0.8142, "grad_norm": 0.3888324499130249, "learning_rate": 0.0002, "epoch": 0.18268531242421793, "step": 1130}, {"loss": 0.8533, "grad_norm": 0.39765217900276184, "learning_rate": 0.0002, "epoch": 0.18430199660496321, "step": 1140}, {"loss": 0.8541, "grad_norm": 0.47190186381340027, "learning_rate": 0.0002, "epoch": 0.1859186807857085, "step": 1150}, {"loss": 0.8301, "grad_norm": 0.4464188814163208, "learning_rate": 0.0002, "epoch": 0.18753536496645382, "step": 1160}, {"loss": 0.8341, "grad_norm": 0.5153930187225342, "learning_rate": 0.0002, "epoch": 0.1891520491471991, "step": 1170}, {"loss": 0.8033, "grad_norm": 0.4779708683490753, "learning_rate": 0.0002, "epoch": 0.1907687333279444, "step": 1180}, {"loss": 0.8187, "grad_norm": 0.4834315776824951, "learning_rate": 0.0002, "epoch": 0.19238541750868968, "step": 1190}, {"loss": 0.7721, "grad_norm": 0.402357816696167, "learning_rate": 0.0002, "epoch": 0.19400210168943496, "step": 1200}, {"loss": 0.7941, "grad_norm": 0.45899084210395813, "learning_rate": 0.0002, "epoch": 0.19561878587018025, "step": 1210}, {"loss": 0.8353, "grad_norm": 0.5106529593467712, "learning_rate": 0.0002, "epoch": 0.19723547005092557, "step": 1220}, {"loss": 0.7816, "grad_norm": 0.45261722803115845, "learning_rate": 0.0002, "epoch": 0.19885215423167085, "step": 1230}, {"loss": 0.8068, "grad_norm": 0.4647127091884613, "learning_rate": 0.0002, "epoch": 0.20046883841241614, "step": 1240}, {"loss": 0.8239, "grad_norm": 0.4849368929862976, "learning_rate": 0.0002, "epoch": 0.20208552259316143, "step": 1250}, {"loss": 0.8514, "grad_norm": 0.4518061578273773, "learning_rate": 0.0002, "epoch": 0.2037022067739067, "step": 1260}, {"loss": 0.8158, "grad_norm": 0.49535325169563293, "learning_rate": 0.0002, "epoch": 0.205318890954652, "step": 1270}, {"loss": 0.8348, "grad_norm": 0.4835205376148224, "learning_rate": 0.0002, "epoch": 0.2069355751353973, "step": 1280}, {"loss": 0.8428, "grad_norm": 0.45308539271354675, "learning_rate": 0.0002, "epoch": 0.2085522593161426, "step": 1290}, {"loss": 0.7993, "grad_norm": 0.5369905233383179, "learning_rate": 0.0002, "epoch": 0.2101689434968879, "step": 1300}, {"loss": 0.8676, "grad_norm": 0.5031622052192688, "learning_rate": 0.0002, "epoch": 0.21178562767763318, "step": 1310}, {"loss": 0.7686, "grad_norm": 0.48010334372520447, "learning_rate": 0.0002, "epoch": 0.21340231185837846, "step": 1320}, {"loss": 0.806, "grad_norm": 0.4905701279640198, "learning_rate": 0.0002, "epoch": 0.21501899603912375, "step": 1330}, {"loss": 0.7885, "grad_norm": 0.43531742691993713, "learning_rate": 0.0002, "epoch": 0.21663568021986904, "step": 1340}, {"loss": 0.8191, "grad_norm": 0.44330692291259766, "learning_rate": 0.0002, "epoch": 0.21825236440061435, "step": 1350}, {"loss": 0.8205, "grad_norm": 0.5384416580200195, "learning_rate": 0.0002, "epoch": 0.21986904858135964, "step": 1360}, {"loss": 0.7726, "grad_norm": 0.4181833863258362, "learning_rate": 0.0002, "epoch": 0.22148573276210493, "step": 1370}, {"loss": 0.8311, "grad_norm": 0.523833692073822, "learning_rate": 0.0002, "epoch": 0.2231024169428502, "step": 1380}, {"loss": 0.7913, "grad_norm": 0.5528736710548401, "learning_rate": 0.0002, "epoch": 0.2247191011235955, "step": 1390}, {"loss": 0.8079, "grad_norm": 0.43515023589134216, "learning_rate": 0.0002, "epoch": 0.2263357853043408, "step": 1400}, {"loss": 0.8403, "grad_norm": 0.48809877038002014, "learning_rate": 0.0002, "epoch": 0.2279524694850861, "step": 1410}, {"loss": 0.8165, "grad_norm": 0.43591251969337463, "learning_rate": 0.0002, "epoch": 0.2295691536658314, "step": 1420}, {"loss": 0.8147, "grad_norm": 0.44625312089920044, "learning_rate": 0.0002, "epoch": 0.23118583784657668, "step": 1430}, {"loss": 0.8134, "grad_norm": 0.4390665292739868, "learning_rate": 0.0002, "epoch": 0.23280252202732196, "step": 1440}, {"loss": 0.8465, "grad_norm": 0.48496049642562866, "learning_rate": 0.0002, "epoch": 0.23441920620806725, "step": 1450}, {"loss": 0.775, "grad_norm": 0.45919957756996155, "learning_rate": 0.0002, "epoch": 0.23603589038881254, "step": 1460}, {"loss": 0.8659, "grad_norm": 0.5471845865249634, "learning_rate": 0.0002, "epoch": 0.23765257456955785, "step": 1470}, {"loss": 0.8164, "grad_norm": 0.47269317507743835, "learning_rate": 0.0002, "epoch": 0.23926925875030314, "step": 1480}, {"loss": 0.854, "grad_norm": 0.4930245578289032, "learning_rate": 0.0002, "epoch": 0.24088594293104842, "step": 1490}, {"loss": 0.8139, "grad_norm": 0.5605630278587341, "learning_rate": 0.0002, "epoch": 0.2425026271117937, "step": 1500}, {"loss": 0.8125, "grad_norm": 0.4435870945453644, "learning_rate": 0.0002, "epoch": 0.244119311292539, "step": 1510}, {"loss": 0.8123, "grad_norm": 0.4941999912261963, "learning_rate": 0.0002, "epoch": 0.24573599547328429, "step": 1520}, {"loss": 0.8427, "grad_norm": 0.5100624561309814, "learning_rate": 0.0002, "epoch": 0.24735267965402957, "step": 1530}, {"loss": 0.8405, "grad_norm": 0.4638267457485199, "learning_rate": 0.0002, "epoch": 0.2489693638347749, "step": 1540}, {"loss": 0.81, "grad_norm": 0.5071570873260498, "learning_rate": 0.0002, "epoch": 0.25058604801552015, "step": 1550}, {"loss": 0.7724, "grad_norm": 0.4291319251060486, "learning_rate": 0.0002, "epoch": 0.25220273219626543, "step": 1560}, {"loss": 0.7984, "grad_norm": 0.5388049483299255, "learning_rate": 0.0002, "epoch": 0.2538194163770108, "step": 1570}, {"loss": 0.8176, "grad_norm": 0.5083683729171753, "learning_rate": 0.0002, "epoch": 0.25543610055775606, "step": 1580}, {"loss": 0.843, "grad_norm": 0.4824463725090027, "learning_rate": 0.0002, "epoch": 0.25705278473850135, "step": 1590}, {"loss": 0.7996, "grad_norm": 0.41177722811698914, "learning_rate": 0.0002, "epoch": 0.25866946891924664, "step": 1600}, {"loss": 0.7772, "grad_norm": 0.5656219124794006, "learning_rate": 0.0002, "epoch": 0.2602861530999919, "step": 1610}, {"loss": 0.7955, "grad_norm": 0.41063204407691956, "learning_rate": 0.0002, "epoch": 0.2619028372807372, "step": 1620}, {"loss": 0.7998, "grad_norm": 0.4897061288356781, "learning_rate": 0.0002, "epoch": 0.2635195214614825, "step": 1630}, {"loss": 0.8198, "grad_norm": 0.4454376697540283, "learning_rate": 0.0002, "epoch": 0.2651362056422278, "step": 1640}, {"loss": 0.8684, "grad_norm": 0.4355238378047943, "learning_rate": 0.0002, "epoch": 0.26675288982297307, "step": 1650}, {"loss": 0.7801, "grad_norm": 0.458310067653656, "learning_rate": 0.0002, "epoch": 0.26836957400371836, "step": 1660}, {"loss": 0.7935, "grad_norm": 0.4752083718776703, "learning_rate": 0.0002, "epoch": 0.26998625818446365, "step": 1670}, {"loss": 0.8267, "grad_norm": 0.4666106402873993, "learning_rate": 0.0002, "epoch": 0.27160294236520893, "step": 1680}, {"loss": 0.8252, "grad_norm": 0.4213818609714508, "learning_rate": 0.0002, "epoch": 0.2732196265459543, "step": 1690}, {"loss": 0.8559, "grad_norm": 0.5768913626670837, "learning_rate": 0.0002, "epoch": 0.27483631072669956, "step": 1700}, {"loss": 0.7931, "grad_norm": 0.4209914803504944, "learning_rate": 0.0002, "epoch": 0.27645299490744485, "step": 1710}, {"loss": 0.8167, "grad_norm": 0.501909613609314, "learning_rate": 0.0002, "epoch": 0.27806967908819014, "step": 1720}, {"loss": 0.7832, "grad_norm": 0.5266261100769043, "learning_rate": 0.0002, "epoch": 0.2796863632689354, "step": 1730}, {"loss": 0.8102, "grad_norm": 0.43806859850883484, "learning_rate": 0.0002, "epoch": 0.2813030474496807, "step": 1740}, {"loss": 0.8157, "grad_norm": 0.46048814058303833, "learning_rate": 0.0002, "epoch": 0.282919731630426, "step": 1750}, {"loss": 0.8596, "grad_norm": 0.44972819089889526, "learning_rate": 0.0002, "epoch": 0.2845364158111713, "step": 1760}, {"loss": 0.8421, "grad_norm": 0.5114831328392029, "learning_rate": 0.0002, "epoch": 0.28615309999191657, "step": 1770}, {"loss": 0.8361, "grad_norm": 0.47931742668151855, "learning_rate": 0.0002, "epoch": 0.28776978417266186, "step": 1780}, {"loss": 0.8265, "grad_norm": 0.5092599987983704, "learning_rate": 0.0002, "epoch": 0.28938646835340714, "step": 1790}, {"loss": 0.8506, "grad_norm": 0.37581443786621094, "learning_rate": 0.0002, "epoch": 0.29100315253415243, "step": 1800}, {"loss": 0.7932, "grad_norm": 0.47097381949424744, "learning_rate": 0.0002, "epoch": 0.2926198367148977, "step": 1810}, {"loss": 0.7787, "grad_norm": 0.48300236463546753, "learning_rate": 0.0002, "epoch": 0.29423652089564306, "step": 1820}, {"loss": 0.8391, "grad_norm": 0.5600419640541077, "learning_rate": 0.0002, "epoch": 0.29585320507638835, "step": 1830}, {"loss": 0.8507, "grad_norm": 0.48555272817611694, "learning_rate": 0.0002, "epoch": 0.29746988925713364, "step": 1840}, {"loss": 0.7657, "grad_norm": 0.3752668499946594, "learning_rate": 0.0002, "epoch": 0.2990865734378789, "step": 1850}, {"loss": 0.7915, "grad_norm": 0.5328747034072876, "learning_rate": 0.0002, "epoch": 0.3007032576186242, "step": 1860}, {"loss": 0.8426, "grad_norm": 0.48716455698013306, "learning_rate": 0.0002, "epoch": 0.3023199417993695, "step": 1870}, {"loss": 0.8335, "grad_norm": 0.5011493563652039, "learning_rate": 0.0002, "epoch": 0.3039366259801148, "step": 1880}, {"loss": 0.852, "grad_norm": 0.46461427211761475, "learning_rate": 0.0002, "epoch": 0.30555331016086007, "step": 1890}, {"loss": 0.8478, "grad_norm": 0.36630210280418396, "learning_rate": 0.0002, "epoch": 0.30716999434160536, "step": 1900}, {"loss": 0.8162, "grad_norm": 0.4217296242713928, "learning_rate": 0.0002, "epoch": 0.30878667852235064, "step": 1910}, {"loss": 0.8128, "grad_norm": 0.4394875466823578, "learning_rate": 0.0002, "epoch": 0.31040336270309593, "step": 1920}, {"loss": 0.8471, "grad_norm": 0.6587965488433838, "learning_rate": 0.0002, "epoch": 0.3120200468838412, "step": 1930}, {"loss": 0.8565, "grad_norm": 0.5469298958778381, "learning_rate": 0.0002, "epoch": 0.31363673106458656, "step": 1940}, {"loss": 0.8236, "grad_norm": 0.4371595084667206, "learning_rate": 0.0002, "epoch": 0.31525341524533185, "step": 1950}, {"loss": 0.887, "grad_norm": 0.4809541404247284, "learning_rate": 0.0002, "epoch": 0.31687009942607713, "step": 1960}, {"loss": 0.7855, "grad_norm": 0.6061086654663086, "learning_rate": 0.0002, "epoch": 0.3184867836068224, "step": 1970}, {"loss": 0.7679, "grad_norm": 0.5342657566070557, "learning_rate": 0.0002, "epoch": 0.3201034677875677, "step": 1980}, {"loss": 0.7955, "grad_norm": 0.5057743787765503, "learning_rate": 0.0002, "epoch": 0.321720151968313, "step": 1990}, {"loss": 0.7774, "grad_norm": 0.528626024723053, "learning_rate": 0.0002, "epoch": 0.3233368361490583, "step": 2000}, {"loss": 0.8845, "grad_norm": 0.46742770075798035, "learning_rate": 0.0002, "epoch": 0.32495352032980357, "step": 2010}, {"loss": 0.8484, "grad_norm": 0.515101432800293, "learning_rate": 0.0002, "epoch": 0.32657020451054886, "step": 2020}, {"loss": 0.8139, "grad_norm": 0.41941216588020325, "learning_rate": 0.0002, "epoch": 0.32818688869129414, "step": 2030}, {"loss": 0.7637, "grad_norm": 0.49902522563934326, "learning_rate": 0.0002, "epoch": 0.32980357287203943, "step": 2040}, {"loss": 0.7822, "grad_norm": 0.4120897650718689, "learning_rate": 0.0002, "epoch": 0.3314202570527847, "step": 2050}, {"loss": 0.8057, "grad_norm": 0.45352041721343994, "learning_rate": 0.0002, "epoch": 0.33303694123353, "step": 2060}, {"loss": 0.7913, "grad_norm": 0.523199737071991, "learning_rate": 0.0002, "epoch": 0.33465362541427535, "step": 2070}, {"loss": 0.8036, "grad_norm": 0.4390358626842499, "learning_rate": 0.0002, "epoch": 0.33627030959502063, "step": 2080}, {"loss": 0.8145, "grad_norm": 0.6752901077270508, "learning_rate": 0.0002, "epoch": 0.3378869937757659, "step": 2090}, {"loss": 0.7807, "grad_norm": 0.547821044921875, "learning_rate": 0.0002, "epoch": 0.3395036779565112, "step": 2100}, {"loss": 0.8561, "grad_norm": 0.5161308646202087, "learning_rate": 0.0002, "epoch": 0.3411203621372565, "step": 2110}, {"loss": 0.7697, "grad_norm": 0.4565401077270508, "learning_rate": 0.0002, "epoch": 0.3427370463180018, "step": 2120}, {"loss": 0.7964, "grad_norm": 0.4666115939617157, "learning_rate": 0.0002, "epoch": 0.34435373049874707, "step": 2130}, {"loss": 0.8189, "grad_norm": 0.4090428352355957, "learning_rate": 0.0002, "epoch": 0.34597041467949236, "step": 2140}, {"loss": 0.8817, "grad_norm": 0.510845422744751, "learning_rate": 0.0002, "epoch": 0.34758709886023764, "step": 2150}, {"loss": 0.8398, "grad_norm": 0.42861923575401306, "learning_rate": 0.0002, "epoch": 0.34920378304098293, "step": 2160}, {"loss": 0.7716, "grad_norm": 0.4476332664489746, "learning_rate": 0.0002, "epoch": 0.3508204672217282, "step": 2170}, {"loss": 0.7845, "grad_norm": 0.6065791249275208, "learning_rate": 0.0002, "epoch": 0.3524371514024735, "step": 2180}, {"loss": 0.8187, "grad_norm": 0.42335066199302673, "learning_rate": 0.0002, "epoch": 0.35405383558321885, "step": 2190}, {"loss": 0.8239, "grad_norm": 0.5094629526138306, "learning_rate": 0.0002, "epoch": 0.35567051976396413, "step": 2200}, {"loss": 0.7807, "grad_norm": 0.5476373434066772, "learning_rate": 0.0002, "epoch": 0.3572872039447094, "step": 2210}, {"loss": 0.814, "grad_norm": 0.3911719024181366, "learning_rate": 0.0002, "epoch": 0.3589038881254547, "step": 2220}, {"loss": 0.8599, "grad_norm": 0.6599636077880859, "learning_rate": 0.0002, "epoch": 0.3605205723062, "step": 2230}, {"loss": 0.7482, "grad_norm": 0.40381914377212524, "learning_rate": 0.0002, "epoch": 0.3621372564869453, "step": 2240}, {"loss": 0.7772, "grad_norm": 0.4433908462524414, "learning_rate": 0.0002, "epoch": 0.36375394066769057, "step": 2250}, {"loss": 0.8503, "grad_norm": 0.578326940536499, "learning_rate": 0.0002, "epoch": 0.36537062484843585, "step": 2260}, {"loss": 0.8178, "grad_norm": 0.5734784007072449, "learning_rate": 0.0002, "epoch": 0.36698730902918114, "step": 2270}, {"loss": 0.8193, "grad_norm": 0.45555487275123596, "learning_rate": 0.0002, "epoch": 0.36860399320992643, "step": 2280}, {"loss": 0.7929, "grad_norm": 0.5666276216506958, "learning_rate": 0.0002, "epoch": 0.3702206773906717, "step": 2290}, {"loss": 0.8292, "grad_norm": 0.5461117625236511, "learning_rate": 0.0002, "epoch": 0.371837361571417, "step": 2300}, {"loss": 0.8204, "grad_norm": 0.6318911910057068, "learning_rate": 0.0002, "epoch": 0.3734540457521623, "step": 2310}, {"loss": 0.7964, "grad_norm": 0.493263304233551, "learning_rate": 0.0002, "epoch": 0.37507072993290763, "step": 2320}, {"loss": 0.8339, "grad_norm": 0.5888760089874268, "learning_rate": 0.0002, "epoch": 0.3766874141136529, "step": 2330}, {"loss": 0.7737, "grad_norm": 0.48671841621398926, "learning_rate": 0.0002, "epoch": 0.3783040982943982, "step": 2340}, {"loss": 0.8367, "grad_norm": 0.4385145306587219, "learning_rate": 0.0002, "epoch": 0.3799207824751435, "step": 2350}, {"loss": 0.812, "grad_norm": 0.5523318648338318, "learning_rate": 0.0002, "epoch": 0.3815374666558888, "step": 2360}, {"loss": 0.8351, "grad_norm": 0.7308220267295837, "learning_rate": 0.0002, "epoch": 0.38315415083663407, "step": 2370}, {"loss": 0.859, "grad_norm": 0.554214358329773, "learning_rate": 0.0002, "epoch": 0.38477083501737935, "step": 2380}, {"loss": 0.8146, "grad_norm": 0.5425800085067749, "learning_rate": 0.0002, "epoch": 0.38638751919812464, "step": 2390}, {"loss": 0.8282, "grad_norm": 0.48811158537864685, "learning_rate": 0.0002, "epoch": 0.3880042033788699, "step": 2400}, {"loss": 0.8074, "grad_norm": 0.49212366342544556, "learning_rate": 0.0002, "epoch": 0.3896208875596152, "step": 2410}, {"loss": 0.7991, "grad_norm": 0.5222218632698059, "learning_rate": 0.0002, "epoch": 0.3912375717403605, "step": 2420}, {"loss": 0.8182, "grad_norm": 0.4699819087982178, "learning_rate": 0.0002, "epoch": 0.3928542559211058, "step": 2430}, {"loss": 0.7919, "grad_norm": 0.46153587102890015, "learning_rate": 0.0002, "epoch": 0.39447094010185113, "step": 2440}, {"loss": 0.8111, "grad_norm": 0.4150611162185669, "learning_rate": 0.0002, "epoch": 0.3960876242825964, "step": 2450}, {"loss": 0.8589, "grad_norm": 0.5799614787101746, "learning_rate": 0.0002, "epoch": 0.3977043084633417, "step": 2460}, {"loss": 0.8085, "grad_norm": 0.56536865234375, "learning_rate": 0.0002, "epoch": 0.399320992644087, "step": 2470}, {"loss": 0.8022, "grad_norm": 0.5451247096061707, "learning_rate": 0.0002, "epoch": 0.4009376768248323, "step": 2480}, {"loss": 0.8217, "grad_norm": 0.5914521217346191, "learning_rate": 0.0002, "epoch": 0.40255436100557757, "step": 2490}, {"loss": 0.7859, "grad_norm": 0.4428117275238037, "learning_rate": 0.0002, "epoch": 0.40417104518632285, "step": 2500}, {"loss": 0.8054, "grad_norm": 0.48580947518348694, "learning_rate": 0.0002, "epoch": 0.40578772936706814, "step": 2510}, {"loss": 0.8405, "grad_norm": 0.436734676361084, "learning_rate": 0.0002, "epoch": 0.4074044135478134, "step": 2520}, {"loss": 0.8209, "grad_norm": 0.5752223134040833, "learning_rate": 0.0002, "epoch": 0.4090210977285587, "step": 2530}, {"loss": 0.8181, "grad_norm": 0.4271308183670044, "learning_rate": 0.0002, "epoch": 0.410637781909304, "step": 2540}, {"loss": 0.8058, "grad_norm": 0.46294718980789185, "learning_rate": 0.0002, "epoch": 0.4122544660900493, "step": 2550}, {"loss": 0.8473, "grad_norm": 0.49407583475112915, "learning_rate": 0.0002, "epoch": 0.4138711502707946, "step": 2560}, {"loss": 0.7881, "grad_norm": 0.4729035496711731, "learning_rate": 0.0002, "epoch": 0.4154878344515399, "step": 2570}, {"loss": 0.7834, "grad_norm": 0.4129747152328491, "learning_rate": 0.0002, "epoch": 0.4171045186322852, "step": 2580}, {"loss": 0.7859, "grad_norm": 0.5684236288070679, "learning_rate": 0.0002, "epoch": 0.4187212028130305, "step": 2590}, {"loss": 0.811, "grad_norm": 0.4862157106399536, "learning_rate": 0.0002, "epoch": 0.4203378869937758, "step": 2600}, {"loss": 0.7582, "grad_norm": 0.46567976474761963, "learning_rate": 0.0002, "epoch": 0.42195457117452106, "step": 2610}, {"loss": 0.7755, "grad_norm": 0.5710650682449341, "learning_rate": 0.0002, "epoch": 0.42357125535526635, "step": 2620}, {"loss": 0.8573, "grad_norm": 0.5660041570663452, "learning_rate": 0.0002, "epoch": 0.42518793953601164, "step": 2630}, {"loss": 0.7812, "grad_norm": 0.47944375872612, "learning_rate": 0.0002, "epoch": 0.4268046237167569, "step": 2640}, {"loss": 0.7459, "grad_norm": 0.537223756313324, "learning_rate": 0.0002, "epoch": 0.4284213078975022, "step": 2650}, {"loss": 0.8246, "grad_norm": 0.41669997572898865, "learning_rate": 0.0002, "epoch": 0.4300379920782475, "step": 2660}, {"loss": 0.7785, "grad_norm": 0.44727686047554016, "learning_rate": 0.0002, "epoch": 0.4316546762589928, "step": 2670}, {"loss": 0.8241, "grad_norm": 0.5600888729095459, "learning_rate": 0.0002, "epoch": 0.4332713604397381, "step": 2680}, {"loss": 0.7708, "grad_norm": 0.39820605516433716, "learning_rate": 0.0002, "epoch": 0.4348880446204834, "step": 2690}, {"loss": 0.8202, "grad_norm": 0.5637655854225159, "learning_rate": 0.0002, "epoch": 0.4365047288012287, "step": 2700}, {"loss": 0.855, "grad_norm": 0.6363666653633118, "learning_rate": 0.0002, "epoch": 0.438121412981974, "step": 2710}, {"loss": 0.8468, "grad_norm": 0.5656129121780396, "learning_rate": 0.0002, "epoch": 0.4397380971627193, "step": 2720}, {"loss": 0.7845, "grad_norm": 0.5600156188011169, "learning_rate": 0.0002, "epoch": 0.44135478134346456, "step": 2730}, {"loss": 0.8405, "grad_norm": 0.5506579875946045, "learning_rate": 0.0002, "epoch": 0.44297146552420985, "step": 2740}, {"loss": 0.7725, "grad_norm": 0.49878305196762085, "learning_rate": 0.0002, "epoch": 0.44458814970495514, "step": 2750}, {"loss": 0.8292, "grad_norm": 0.4569213092327118, "learning_rate": 0.0002, "epoch": 0.4462048338857004, "step": 2760}, {"loss": 0.8028, "grad_norm": 0.6056680083274841, "learning_rate": 0.0002, "epoch": 0.4478215180664457, "step": 2770}, {"loss": 0.8242, "grad_norm": 0.44474557042121887, "learning_rate": 0.0002, "epoch": 0.449438202247191, "step": 2780}, {"loss": 0.801, "grad_norm": 0.46055394411087036, "learning_rate": 0.0002, "epoch": 0.4510548864279363, "step": 2790}, {"loss": 0.7521, "grad_norm": 0.4904133379459381, "learning_rate": 0.0002, "epoch": 0.4526715706086816, "step": 2800}, {"loss": 0.8829, "grad_norm": 0.5647031664848328, "learning_rate": 0.0002, "epoch": 0.45428825478942686, "step": 2810}, {"loss": 0.8622, "grad_norm": 0.5759473443031311, "learning_rate": 0.0002, "epoch": 0.4559049389701722, "step": 2820}, {"loss": 0.7812, "grad_norm": 0.5161895751953125, "learning_rate": 0.0002, "epoch": 0.4575216231509175, "step": 2830}, {"loss": 0.8045, "grad_norm": 0.4248254597187042, "learning_rate": 0.0002, "epoch": 0.4591383073316628, "step": 2840}, {"loss": 0.7838, "grad_norm": 0.45395001769065857, "learning_rate": 0.0002, "epoch": 0.46075499151240806, "step": 2850}, {"loss": 0.8208, "grad_norm": 0.5358697772026062, "learning_rate": 0.0002, "epoch": 0.46237167569315335, "step": 2860}, {"loss": 0.8147, "grad_norm": 0.5379165410995483, "learning_rate": 0.0002, "epoch": 0.46398835987389864, "step": 2870}, {"loss": 0.7403, "grad_norm": 0.4601989686489105, "learning_rate": 0.0002, "epoch": 0.4656050440546439, "step": 2880}, {"loss": 0.8523, "grad_norm": 0.671115517616272, "learning_rate": 0.0002, "epoch": 0.4672217282353892, "step": 2890}, {"loss": 0.8262, "grad_norm": 0.4425133168697357, "learning_rate": 0.0002, "epoch": 0.4688384124161345, "step": 2900}, {"loss": 0.8178, "grad_norm": 0.5446155071258545, "learning_rate": 0.0002, "epoch": 0.4704550965968798, "step": 2910}, {"loss": 0.8106, "grad_norm": 0.603306233882904, "learning_rate": 0.0002, "epoch": 0.47207178077762507, "step": 2920}, {"loss": 0.8044, "grad_norm": 0.5377997159957886, "learning_rate": 0.0002, "epoch": 0.47368846495837036, "step": 2930}, {"loss": 0.8075, "grad_norm": 0.4931027591228485, "learning_rate": 0.0002, "epoch": 0.4753051491391157, "step": 2940}, {"loss": 0.8004, "grad_norm": 0.4711960256099701, "learning_rate": 0.0002, "epoch": 0.476921833319861, "step": 2950}, {"loss": 0.8121, "grad_norm": 0.5020492672920227, "learning_rate": 0.0002, "epoch": 0.4785385175006063, "step": 2960}, {"loss": 0.8221, "grad_norm": 0.5428946614265442, "learning_rate": 0.0002, "epoch": 0.48015520168135156, "step": 2970}, {"loss": 0.7849, "grad_norm": 0.5294089317321777, "learning_rate": 0.0002, "epoch": 0.48177188586209685, "step": 2980}, {"loss": 0.8553, "grad_norm": 0.648289144039154, "learning_rate": 0.0002, "epoch": 0.48338857004284214, "step": 2990}, {"loss": 0.7874, "grad_norm": 0.47916680574417114, "learning_rate": 0.0002, "epoch": 0.4850052542235874, "step": 3000}, {"loss": 0.8087, "grad_norm": 0.43849772214889526, "learning_rate": 0.0002, "epoch": 0.4866219384043327, "step": 3010}, {"loss": 0.7662, "grad_norm": 0.47007861733436584, "learning_rate": 0.0002, "epoch": 0.488238622585078, "step": 3020}, {"loss": 0.757, "grad_norm": 0.6314331293106079, "learning_rate": 0.0002, "epoch": 0.4898553067658233, "step": 3030}, {"loss": 0.7863, "grad_norm": 0.49211493134498596, "learning_rate": 0.0002, "epoch": 0.49147199094656857, "step": 3040}, {"loss": 0.8335, "grad_norm": 0.4537973403930664, "learning_rate": 0.0002, "epoch": 0.49308867512731386, "step": 3050}, {"loss": 0.8095, "grad_norm": 0.47326919436454773, "learning_rate": 0.0002, "epoch": 0.49470535930805914, "step": 3060}, {"loss": 0.8447, "grad_norm": 0.525874137878418, "learning_rate": 0.0002, "epoch": 0.4963220434888045, "step": 3070}, {"loss": 0.8339, "grad_norm": 0.6361091732978821, "learning_rate": 0.0002, "epoch": 0.4979387276695498, "step": 3080}, {"loss": 0.821, "grad_norm": 0.5850642919540405, "learning_rate": 0.0002, "epoch": 0.49955541185029506, "step": 3090}, {"loss": 0.8279, "grad_norm": 0.47299543023109436, "learning_rate": 0.0002, "epoch": 0.5011720960310403, "step": 3100}, {"loss": 0.8681, "grad_norm": 0.473099946975708, "learning_rate": 0.0002, "epoch": 0.5027887802117856, "step": 3110}, {"loss": 0.8223, "grad_norm": 0.48186397552490234, "learning_rate": 0.0002, "epoch": 0.5044054643925309, "step": 3120}, {"loss": 0.8292, "grad_norm": 0.5015401840209961, "learning_rate": 0.0002, "epoch": 0.5060221485732762, "step": 3130}, {"loss": 0.7692, "grad_norm": 0.5617750287055969, "learning_rate": 0.0002, "epoch": 0.5076388327540216, "step": 3140}, {"loss": 0.8708, "grad_norm": 0.5169327259063721, "learning_rate": 0.0002, "epoch": 0.5092555169347668, "step": 3150}, {"loss": 0.7845, "grad_norm": 0.545657753944397, "learning_rate": 0.0002, "epoch": 0.5108722011155121, "step": 3160}, {"loss": 0.799, "grad_norm": 0.512864351272583, "learning_rate": 0.0002, "epoch": 0.5124888852962574, "step": 3170}, {"loss": 0.7794, "grad_norm": 0.4113546311855316, "learning_rate": 0.0002, "epoch": 0.5141055694770027, "step": 3180}, {"loss": 0.8206, "grad_norm": 0.44532445073127747, "learning_rate": 0.0002, "epoch": 0.5157222536577479, "step": 3190}, {"loss": 0.8213, "grad_norm": 0.5623497366905212, "learning_rate": 0.0002, "epoch": 0.5173389378384933, "step": 3200}, {"loss": 0.7928, "grad_norm": 0.5084741115570068, "learning_rate": 0.0002, "epoch": 0.5189556220192385, "step": 3210}, {"loss": 0.8174, "grad_norm": 0.5305403470993042, "learning_rate": 0.0002, "epoch": 0.5205723061999838, "step": 3220}, {"loss": 0.8139, "grad_norm": 0.4708254337310791, "learning_rate": 0.0002, "epoch": 0.5221889903807291, "step": 3230}, {"loss": 0.7639, "grad_norm": 0.43827131390571594, "learning_rate": 0.0002, "epoch": 0.5238056745614744, "step": 3240}, {"loss": 0.7993, "grad_norm": 0.5630002617835999, "learning_rate": 0.0002, "epoch": 0.5254223587422197, "step": 3250}, {"loss": 0.7522, "grad_norm": 0.5010961890220642, "learning_rate": 0.0002, "epoch": 0.527039042922965, "step": 3260}, {"loss": 0.8374, "grad_norm": 0.6303122043609619, "learning_rate": 0.0002, "epoch": 0.5286557271037103, "step": 3270}, {"loss": 0.7727, "grad_norm": 0.5107331275939941, "learning_rate": 0.0002, "epoch": 0.5302724112844556, "step": 3280}, {"loss": 0.8495, "grad_norm": 0.5700443387031555, "learning_rate": 0.0002, "epoch": 0.5318890954652009, "step": 3290}, {"loss": 0.7776, "grad_norm": 0.46296367049217224, "learning_rate": 0.0002, "epoch": 0.5335057796459461, "step": 3300}, {"loss": 0.7931, "grad_norm": 0.531568706035614, "learning_rate": 0.0002, "epoch": 0.5351224638266915, "step": 3310}, {"loss": 0.843, "grad_norm": 0.4686741530895233, "learning_rate": 0.0002, "epoch": 0.5367391480074367, "step": 3320}, {"loss": 0.8104, "grad_norm": 0.5404331088066101, "learning_rate": 0.0002, "epoch": 0.5383558321881821, "step": 3330}, {"loss": 0.7686, "grad_norm": 0.6368790864944458, "learning_rate": 0.0002, "epoch": 0.5399725163689273, "step": 3340}, {"loss": 0.8514, "grad_norm": 0.42300888895988464, "learning_rate": 0.0002, "epoch": 0.5415892005496726, "step": 3350}, {"loss": 0.8236, "grad_norm": 0.5362542867660522, "learning_rate": 0.0002, "epoch": 0.5432058847304179, "step": 3360}, {"loss": 0.858, "grad_norm": 0.497128963470459, "learning_rate": 0.0002, "epoch": 0.5448225689111632, "step": 3370}, {"loss": 0.8519, "grad_norm": 0.5006386041641235, "learning_rate": 0.0002, "epoch": 0.5464392530919085, "step": 3380}, {"loss": 0.7867, "grad_norm": 0.44136837124824524, "learning_rate": 0.0002, "epoch": 0.5480559372726538, "step": 3390}, {"loss": 0.773, "grad_norm": 0.5897833108901978, "learning_rate": 0.0002, "epoch": 0.5496726214533991, "step": 3400}, {"loss": 0.8895, "grad_norm": 0.641075611114502, "learning_rate": 0.0002, "epoch": 0.5512893056341444, "step": 3410}, {"loss": 0.7827, "grad_norm": 0.7251322269439697, "learning_rate": 0.0002, "epoch": 0.5529059898148897, "step": 3420}, {"loss": 0.7626, "grad_norm": 0.47411349415779114, "learning_rate": 0.0002, "epoch": 0.5545226739956349, "step": 3430}, {"loss": 0.8196, "grad_norm": 0.4994310438632965, "learning_rate": 0.0002, "epoch": 0.5561393581763803, "step": 3440}, {"loss": 0.7812, "grad_norm": 0.5814438462257385, "learning_rate": 0.0002, "epoch": 0.5577560423571255, "step": 3450}, {"loss": 0.8805, "grad_norm": 0.6278898119926453, "learning_rate": 0.0002, "epoch": 0.5593727265378708, "step": 3460}, {"loss": 0.813, "grad_norm": 0.46208274364471436, "learning_rate": 0.0002, "epoch": 0.5609894107186161, "step": 3470}, {"loss": 0.8295, "grad_norm": 0.5718930959701538, "learning_rate": 0.0002, "epoch": 0.5626060948993614, "step": 3480}, {"loss": 0.8152, "grad_norm": 0.48178744316101074, "learning_rate": 0.0002, "epoch": 0.5642227790801067, "step": 3490}, {"loss": 0.8244, "grad_norm": 0.47336965799331665, "learning_rate": 0.0002, "epoch": 0.565839463260852, "step": 3500}, {"loss": 0.8099, "grad_norm": 0.43442684412002563, "learning_rate": 0.0002, "epoch": 0.5674561474415973, "step": 3510}, {"loss": 0.7564, "grad_norm": 0.6463358998298645, "learning_rate": 0.0002, "epoch": 0.5690728316223426, "step": 3520}, {"loss": 0.836, "grad_norm": 0.5286486744880676, "learning_rate": 0.0002, "epoch": 0.5706895158030879, "step": 3530}, {"loss": 0.8421, "grad_norm": 0.5405499935150146, "learning_rate": 0.0002, "epoch": 0.5723061999838331, "step": 3540}, {"loss": 0.7614, "grad_norm": 0.6654391884803772, "learning_rate": 0.0002, "epoch": 0.5739228841645785, "step": 3550}, {"loss": 0.7803, "grad_norm": 0.5081980228424072, "learning_rate": 0.0002, "epoch": 0.5755395683453237, "step": 3560}, {"loss": 0.7753, "grad_norm": 0.48978179693222046, "learning_rate": 0.0002, "epoch": 0.5771562525260691, "step": 3570}, {"loss": 0.8151, "grad_norm": 0.5840612053871155, "learning_rate": 0.0002, "epoch": 0.5787729367068143, "step": 3580}, {"loss": 0.8937, "grad_norm": 0.5235261917114258, "learning_rate": 0.0002, "epoch": 0.5803896208875596, "step": 3590}, {"loss": 0.7894, "grad_norm": 0.5672075748443604, "learning_rate": 0.0002, "epoch": 0.5820063050683049, "step": 3600}, {"loss": 0.8347, "grad_norm": 0.5613429546356201, "learning_rate": 0.0002, "epoch": 0.5836229892490502, "step": 3610}, {"loss": 0.8274, "grad_norm": 0.4032273590564728, "learning_rate": 0.0002, "epoch": 0.5852396734297954, "step": 3620}, {"loss": 0.8421, "grad_norm": 0.49559324979782104, "learning_rate": 0.0002, "epoch": 0.5868563576105408, "step": 3630}, {"loss": 0.8332, "grad_norm": 0.6895697712898254, "learning_rate": 0.0002, "epoch": 0.5884730417912861, "step": 3640}, {"loss": 0.7877, "grad_norm": 0.4750136435031891, "learning_rate": 0.0002, "epoch": 0.5900897259720314, "step": 3650}, {"loss": 0.8219, "grad_norm": 0.5176819562911987, "learning_rate": 0.0002, "epoch": 0.5917064101527767, "step": 3660}, {"loss": 0.8151, "grad_norm": 0.5817760229110718, "learning_rate": 0.0002, "epoch": 0.5933230943335219, "step": 3670}, {"loss": 0.7823, "grad_norm": 0.6064626574516296, "learning_rate": 0.0002, "epoch": 0.5949397785142673, "step": 3680}, {"loss": 0.8422, "grad_norm": 0.6728700995445251, "learning_rate": 0.0002, "epoch": 0.5965564626950125, "step": 3690}, {"loss": 0.7679, "grad_norm": 0.609305202960968, "learning_rate": 0.0002, "epoch": 0.5981731468757578, "step": 3700}, {"loss": 0.8048, "grad_norm": 0.4615488350391388, "learning_rate": 0.0002, "epoch": 0.5997898310565031, "step": 3710}, {"loss": 0.8214, "grad_norm": 2.0531179904937744, "learning_rate": 0.0002, "epoch": 0.6014065152372484, "step": 3720}, {"loss": 0.8158, "grad_norm": 0.5091132521629333, "learning_rate": 0.0002, "epoch": 0.6030231994179936, "step": 3730}, {"loss": 0.7833, "grad_norm": 0.5951124429702759, "learning_rate": 0.0002, "epoch": 0.604639883598739, "step": 3740}, {"loss": 0.7784, "grad_norm": 0.5870208144187927, "learning_rate": 0.0002, "epoch": 0.6062565677794842, "step": 3750}, {"loss": 0.8044, "grad_norm": 0.6254619359970093, "learning_rate": 0.0002, "epoch": 0.6078732519602296, "step": 3760}, {"loss": 0.7868, "grad_norm": 0.5577626824378967, "learning_rate": 0.0002, "epoch": 0.6094899361409749, "step": 3770}, {"loss": 0.8108, "grad_norm": 0.5004405379295349, "learning_rate": 0.0002, "epoch": 0.6111066203217201, "step": 3780}, {"loss": 0.8092, "grad_norm": 0.5527383685112, "learning_rate": 0.0002, "epoch": 0.6127233045024655, "step": 3790}, {"loss": 0.8036, "grad_norm": 0.49116113781929016, "learning_rate": 0.0002, "epoch": 0.6143399886832107, "step": 3800}, {"loss": 0.8352, "grad_norm": 0.5299299359321594, "learning_rate": 0.0002, "epoch": 0.6159566728639561, "step": 3810}, {"loss": 0.7737, "grad_norm": 0.464897483587265, "learning_rate": 0.0002, "epoch": 0.6175733570447013, "step": 3820}, {"loss": 0.7923, "grad_norm": 0.6505740880966187, "learning_rate": 0.0002, "epoch": 0.6191900412254466, "step": 3830}, {"loss": 0.8123, "grad_norm": 0.5512559413909912, "learning_rate": 0.0002, "epoch": 0.6208067254061919, "step": 3840}, {"loss": 0.8856, "grad_norm": 0.49427518248558044, "learning_rate": 0.0002, "epoch": 0.6224234095869372, "step": 3850}, {"loss": 0.7751, "grad_norm": 0.3839147090911865, "learning_rate": 0.0002, "epoch": 0.6240400937676824, "step": 3860}, {"loss": 0.8006, "grad_norm": 0.5760218501091003, "learning_rate": 0.0002, "epoch": 0.6256567779484278, "step": 3870}, {"loss": 0.7836, "grad_norm": 0.7226507067680359, "learning_rate": 0.0002, "epoch": 0.6272734621291731, "step": 3880}, {"loss": 0.8244, "grad_norm": 0.676781415939331, "learning_rate": 0.0002, "epoch": 0.6288901463099184, "step": 3890}, {"loss": 0.8239, "grad_norm": 0.4284018278121948, "learning_rate": 0.0002, "epoch": 0.6305068304906637, "step": 3900}, {"loss": 0.7996, "grad_norm": 0.5060628056526184, "learning_rate": 0.0002, "epoch": 0.6321235146714089, "step": 3910}, {"loss": 0.8089, "grad_norm": 0.5524522066116333, "learning_rate": 0.0002, "epoch": 0.6337401988521543, "step": 3920}, {"loss": 0.8276, "grad_norm": 0.6099881529808044, "learning_rate": 0.0002, "epoch": 0.6353568830328995, "step": 3930}, {"loss": 0.809, "grad_norm": 0.43155938386917114, "learning_rate": 0.0002, "epoch": 0.6369735672136448, "step": 3940}, {"loss": 0.8404, "grad_norm": 0.6427084803581238, "learning_rate": 0.0002, "epoch": 0.6385902513943901, "step": 3950}, {"loss": 0.8368, "grad_norm": 0.541220486164093, "learning_rate": 0.0002, "epoch": 0.6402069355751354, "step": 3960}, {"loss": 0.8539, "grad_norm": 0.5414294600486755, "learning_rate": 0.0002, "epoch": 0.6418236197558806, "step": 3970}, {"loss": 0.7996, "grad_norm": 0.46344003081321716, "learning_rate": 0.0002, "epoch": 0.643440303936626, "step": 3980}, {"loss": 0.7474, "grad_norm": 0.45209285616874695, "learning_rate": 0.0002, "epoch": 0.6450569881173712, "step": 3990}, {"loss": 0.8202, "grad_norm": 0.5417284369468689, "learning_rate": 0.0002, "epoch": 0.6466736722981166, "step": 4000}, {"loss": 0.7563, "grad_norm": 0.7995685935020447, "learning_rate": 0.0002, "epoch": 0.6482903564788619, "step": 4010}, {"loss": 0.7812, "grad_norm": 0.6384002566337585, "learning_rate": 0.0002, "epoch": 0.6499070406596071, "step": 4020}, {"loss": 0.732, "grad_norm": 0.4472815692424774, "learning_rate": 0.0002, "epoch": 0.6515237248403525, "step": 4030}, {"loss": 0.8071, "grad_norm": 0.6834294199943542, "learning_rate": 0.0002, "epoch": 0.6531404090210977, "step": 4040}, {"loss": 0.7812, "grad_norm": 0.4612339735031128, "learning_rate": 0.0002, "epoch": 0.654757093201843, "step": 4050}, {"loss": 0.8141, "grad_norm": 0.9266576170921326, "learning_rate": 0.0002, "epoch": 0.6563737773825883, "step": 4060}, {"loss": 0.7991, "grad_norm": 0.4470861852169037, "learning_rate": 0.0002, "epoch": 0.6579904615633336, "step": 4070}, {"loss": 0.8293, "grad_norm": 0.45544925332069397, "learning_rate": 0.0002, "epoch": 0.6596071457440789, "step": 4080}, {"loss": 0.8455, "grad_norm": 0.6144481301307678, "learning_rate": 0.0002, "epoch": 0.6612238299248242, "step": 4090}, {"loss": 0.7877, "grad_norm": 0.5936288237571716, "learning_rate": 0.0002, "epoch": 0.6628405141055694, "step": 4100}, {"loss": 0.7617, "grad_norm": 0.4822963774204254, "learning_rate": 0.0002, "epoch": 0.6644571982863148, "step": 4110}, {"loss": 0.7997, "grad_norm": 0.48432496190071106, "learning_rate": 0.0002, "epoch": 0.66607388246706, "step": 4120}, {"loss": 0.8404, "grad_norm": 0.4901607930660248, "learning_rate": 0.0002, "epoch": 0.6676905666478054, "step": 4130}, {"loss": 0.8085, "grad_norm": 0.5018393397331238, "learning_rate": 0.0002, "epoch": 0.6693072508285507, "step": 4140}, {"loss": 0.8065, "grad_norm": 0.6946378946304321, "learning_rate": 0.0002, "epoch": 0.6709239350092959, "step": 4150}, {"loss": 0.8147, "grad_norm": 0.5997390747070312, "learning_rate": 0.0002, "epoch": 0.6725406191900413, "step": 4160}, {"loss": 0.8268, "grad_norm": 0.6738849878311157, "learning_rate": 0.0002, "epoch": 0.6741573033707865, "step": 4170}, {"loss": 0.7704, "grad_norm": 0.6110581159591675, "learning_rate": 0.0002, "epoch": 0.6757739875515318, "step": 4180}, {"loss": 0.8043, "grad_norm": 0.5703322291374207, "learning_rate": 0.0002, "epoch": 0.6773906717322771, "step": 4190}, {"loss": 0.8099, "grad_norm": 0.4686066210269928, "learning_rate": 0.0002, "epoch": 0.6790073559130224, "step": 4200}, {"loss": 0.8441, "grad_norm": 0.6394643783569336, "learning_rate": 0.0002, "epoch": 0.6806240400937676, "step": 4210}, {"loss": 0.8011, "grad_norm": 0.5454841256141663, "learning_rate": 0.0002, "epoch": 0.682240724274513, "step": 4220}, {"loss": 0.8307, "grad_norm": 0.4859732985496521, "learning_rate": 0.0002, "epoch": 0.6838574084552582, "step": 4230}, {"loss": 0.8161, "grad_norm": 0.5544065833091736, "learning_rate": 0.0002, "epoch": 0.6854740926360036, "step": 4240}, {"loss": 0.7839, "grad_norm": 0.4902505576610565, "learning_rate": 0.0002, "epoch": 0.6870907768167488, "step": 4250}, {"loss": 0.7977, "grad_norm": 0.4768051505088806, "learning_rate": 0.0002, "epoch": 0.6887074609974941, "step": 4260}, {"loss": 0.7539, "grad_norm": 0.49982190132141113, "learning_rate": 0.0002, "epoch": 0.6903241451782395, "step": 4270}, {"loss": 0.7353, "grad_norm": 0.6351838111877441, "learning_rate": 0.0002, "epoch": 0.6919408293589847, "step": 4280}, {"loss": 0.7664, "grad_norm": 0.5647561550140381, "learning_rate": 0.0002, "epoch": 0.69355751353973, "step": 4290}, {"loss": 0.7618, "grad_norm": 0.5340486764907837, "learning_rate": 0.0002, "epoch": 0.6951741977204753, "step": 4300}, {"loss": 0.8526, "grad_norm": 0.5649092793464661, "learning_rate": 0.0002, "epoch": 0.6967908819012206, "step": 4310}, {"loss": 0.8246, "grad_norm": 0.6183916926383972, "learning_rate": 0.0002, "epoch": 0.6984075660819659, "step": 4320}, {"loss": 0.792, "grad_norm": 0.6154509782791138, "learning_rate": 0.0002, "epoch": 0.7000242502627112, "step": 4330}, {"loss": 0.8397, "grad_norm": 0.5156264305114746, "learning_rate": 0.0002, "epoch": 0.7016409344434564, "step": 4340}, {"loss": 0.8512, "grad_norm": 0.562171459197998, "learning_rate": 0.0002, "epoch": 0.7032576186242018, "step": 4350}, {"loss": 0.7882, "grad_norm": 0.4949502646923065, "learning_rate": 0.0002, "epoch": 0.704874302804947, "step": 4360}, {"loss": 0.738, "grad_norm": 0.5171684622764587, "learning_rate": 0.0002, "epoch": 0.7064909869856923, "step": 4370}, {"loss": 0.8001, "grad_norm": 0.6198443174362183, "learning_rate": 0.0002, "epoch": 0.7081076711664377, "step": 4380}, {"loss": 0.7606, "grad_norm": 0.5802276134490967, "learning_rate": 0.0002, "epoch": 0.7097243553471829, "step": 4390}, {"loss": 0.8797, "grad_norm": 0.41096967458724976, "learning_rate": 0.0002, "epoch": 0.7113410395279283, "step": 4400}, {"loss": 0.805, "grad_norm": 0.4397392272949219, "learning_rate": 0.0002, "epoch": 0.7129577237086735, "step": 4410}, {"loss": 0.7651, "grad_norm": 0.45228442549705505, "learning_rate": 0.0002, "epoch": 0.7145744078894188, "step": 4420}, {"loss": 0.7938, "grad_norm": 0.4839673936367035, "learning_rate": 0.0002, "epoch": 0.7161910920701641, "step": 4430}, {"loss": 0.8362, "grad_norm": 0.6140755414962769, "learning_rate": 0.0002, "epoch": 0.7178077762509094, "step": 4440}, {"loss": 0.7722, "grad_norm": 0.6841378808021545, "learning_rate": 0.0002, "epoch": 0.7194244604316546, "step": 4450}, {"loss": 0.8177, "grad_norm": 0.6664239168167114, "learning_rate": 0.0002, "epoch": 0.7210411446124, "step": 4460}, {"loss": 0.7983, "grad_norm": 0.47552719712257385, "learning_rate": 0.0002, "epoch": 0.7226578287931452, "step": 4470}, {"loss": 0.8982, "grad_norm": 0.6649776101112366, "learning_rate": 0.0002, "epoch": 0.7242745129738906, "step": 4480}, {"loss": 0.8074, "grad_norm": 0.5159541964530945, "learning_rate": 0.0002, "epoch": 0.7258911971546358, "step": 4490}, {"loss": 0.7786, "grad_norm": 0.6693112850189209, "learning_rate": 0.0002, "epoch": 0.7275078813353811, "step": 4500}, {"loss": 0.8655, "grad_norm": 0.48870977759361267, "learning_rate": 0.0002, "epoch": 0.7291245655161265, "step": 4510}, {"loss": 0.7337, "grad_norm": 0.4857887923717499, "learning_rate": 0.0002, "epoch": 0.7307412496968717, "step": 4520}, {"loss": 0.8026, "grad_norm": 0.5515662431716919, "learning_rate": 0.0002, "epoch": 0.732357933877617, "step": 4530}, {"loss": 0.8031, "grad_norm": 0.6292222738265991, "learning_rate": 0.0002, "epoch": 0.7339746180583623, "step": 4540}, {"loss": 0.7749, "grad_norm": 0.48265689611434937, "learning_rate": 0.0002, "epoch": 0.7355913022391076, "step": 4550}, {"loss": 0.8499, "grad_norm": 0.8044266104698181, "learning_rate": 0.0002, "epoch": 0.7372079864198529, "step": 4560}, {"loss": 0.8162, "grad_norm": 0.6111769676208496, "learning_rate": 0.0002, "epoch": 0.7388246706005982, "step": 4570}, {"loss": 0.7291, "grad_norm": 0.5229553580284119, "learning_rate": 0.0002, "epoch": 0.7404413547813434, "step": 4580}, {"loss": 0.8038, "grad_norm": 0.6054152250289917, "learning_rate": 0.0002, "epoch": 0.7420580389620888, "step": 4590}, {"loss": 0.8169, "grad_norm": 0.5574966669082642, "learning_rate": 0.0002, "epoch": 0.743674723142834, "step": 4600}, {"loss": 0.8439, "grad_norm": 0.5395817160606384, "learning_rate": 0.0002, "epoch": 0.7452914073235793, "step": 4610}, {"loss": 0.8495, "grad_norm": 0.7116472721099854, "learning_rate": 0.0002, "epoch": 0.7469080915043246, "step": 4620}, {"loss": 0.7743, "grad_norm": 0.5618700981140137, "learning_rate": 0.0002, "epoch": 0.7485247756850699, "step": 4630}, {"loss": 0.7744, "grad_norm": 0.5802770853042603, "learning_rate": 0.0002, "epoch": 0.7501414598658153, "step": 4640}, {"loss": 0.7924, "grad_norm": 0.5690428018569946, "learning_rate": 0.0002, "epoch": 0.7517581440465605, "step": 4650}, {"loss": 0.8017, "grad_norm": 0.4813360273838043, "learning_rate": 0.0002, "epoch": 0.7533748282273058, "step": 4660}, {"loss": 0.8108, "grad_norm": 0.5434042811393738, "learning_rate": 0.0002, "epoch": 0.7549915124080511, "step": 4670}, {"loss": 0.7824, "grad_norm": 0.5502099990844727, "learning_rate": 0.0002, "epoch": 0.7566081965887964, "step": 4680}, {"loss": 0.8598, "grad_norm": 0.6020621061325073, "learning_rate": 0.0002, "epoch": 0.7582248807695416, "step": 4690}, {"loss": 0.7937, "grad_norm": 0.4922301471233368, "learning_rate": 0.0002, "epoch": 0.759841564950287, "step": 4700}, {"loss": 0.788, "grad_norm": 0.6492828726768494, "learning_rate": 0.0002, "epoch": 0.7614582491310322, "step": 4710}, {"loss": 0.8313, "grad_norm": 0.4865580201148987, "learning_rate": 0.0002, "epoch": 0.7630749333117776, "step": 4720}, {"loss": 0.7966, "grad_norm": 0.5971422791481018, "learning_rate": 0.0002, "epoch": 0.7646916174925228, "step": 4730}, {"loss": 0.8298, "grad_norm": 0.6832674145698547, "learning_rate": 0.0002, "epoch": 0.7663083016732681, "step": 4740}, {"loss": 0.8156, "grad_norm": 0.500908613204956, "learning_rate": 0.0002, "epoch": 0.7679249858540134, "step": 4750}, {"loss": 0.8383, "grad_norm": 0.6112465858459473, "learning_rate": 0.0002, "epoch": 0.7695416700347587, "step": 4760}, {"loss": 0.76, "grad_norm": 0.5753506422042847, "learning_rate": 0.0002, "epoch": 0.771158354215504, "step": 4770}, {"loss": 0.8297, "grad_norm": 0.6529405117034912, "learning_rate": 0.0002, "epoch": 0.7727750383962493, "step": 4780}, {"loss": 0.8171, "grad_norm": 0.5916843414306641, "learning_rate": 0.0002, "epoch": 0.7743917225769946, "step": 4790}, {"loss": 0.83, "grad_norm": 0.4821224510669708, "learning_rate": 0.0002, "epoch": 0.7760084067577399, "step": 4800}, {"loss": 0.7703, "grad_norm": 0.5532580018043518, "learning_rate": 0.0002, "epoch": 0.7776250909384852, "step": 4810}, {"loss": 0.7363, "grad_norm": 0.4604877233505249, "learning_rate": 0.0002, "epoch": 0.7792417751192304, "step": 4820}, {"loss": 0.7506, "grad_norm": 0.5009613037109375, "learning_rate": 0.0002, "epoch": 0.7808584592999758, "step": 4830}, {"loss": 0.7863, "grad_norm": 0.6448560357093811, "learning_rate": 0.0002, "epoch": 0.782475143480721, "step": 4840}, {"loss": 0.7957, "grad_norm": 0.44327953457832336, "learning_rate": 0.0002, "epoch": 0.7840918276614663, "step": 4850}, {"loss": 0.7925, "grad_norm": 0.5355411171913147, "learning_rate": 0.0002, "epoch": 0.7857085118422116, "step": 4860}, {"loss": 0.7754, "grad_norm": 0.5635677576065063, "learning_rate": 0.0002, "epoch": 0.7873251960229569, "step": 4870}, {"loss": 0.7931, "grad_norm": 0.5417491793632507, "learning_rate": 0.0002, "epoch": 0.7889418802037023, "step": 4880}, {"loss": 0.7819, "grad_norm": 0.4567430913448334, "learning_rate": 0.0002, "epoch": 0.7905585643844475, "step": 4890}, {"loss": 0.8454, "grad_norm": 0.44651296734809875, "learning_rate": 0.0002, "epoch": 0.7921752485651928, "step": 4900}, {"loss": 0.7959, "grad_norm": 0.5741217136383057, "learning_rate": 0.0002, "epoch": 0.7937919327459381, "step": 4910}, {"loss": 0.8093, "grad_norm": 0.6605045199394226, "learning_rate": 0.0002, "epoch": 0.7954086169266834, "step": 4920}, {"loss": 0.77, "grad_norm": 0.5126531720161438, "learning_rate": 0.0002, "epoch": 0.7970253011074286, "step": 4930}, {"loss": 0.7793, "grad_norm": 0.513648271560669, "learning_rate": 0.0002, "epoch": 0.798641985288174, "step": 4940}, {"loss": 0.8314, "grad_norm": 0.5350404381752014, "learning_rate": 0.0002, "epoch": 0.8002586694689192, "step": 4950}, {"loss": 0.7649, "grad_norm": 0.5731674432754517, "learning_rate": 0.0002, "epoch": 0.8018753536496646, "step": 4960}, {"loss": 0.8572, "grad_norm": 0.5974258184432983, "learning_rate": 0.0002, "epoch": 0.8034920378304098, "step": 4970}, {"loss": 0.7972, "grad_norm": 0.8774799704551697, "learning_rate": 0.0002, "epoch": 0.8051087220111551, "step": 4980}, {"loss": 0.7899, "grad_norm": 0.5994430184364319, "learning_rate": 0.0002, "epoch": 0.8067254061919004, "step": 4990}, {"loss": 0.7736, "grad_norm": 0.4894903004169464, "learning_rate": 0.0002, "epoch": 0.8083420903726457, "step": 5000}, {"loss": 0.78, "grad_norm": 0.5218459367752075, "learning_rate": 0.0002, "epoch": 0.809958774553391, "step": 5010}, {"loss": 0.817, "grad_norm": 0.5232468843460083, "learning_rate": 0.0002, "epoch": 0.8115754587341363, "step": 5020}, {"loss": 0.7704, "grad_norm": 0.44358372688293457, "learning_rate": 0.0002, "epoch": 0.8131921429148816, "step": 5030}, {"loss": 0.785, "grad_norm": 0.6202037334442139, "learning_rate": 0.0002, "epoch": 0.8148088270956269, "step": 5040}, {"loss": 0.7351, "grad_norm": 0.7721474170684814, "learning_rate": 0.0002, "epoch": 0.8164255112763722, "step": 5050}, {"loss": 0.8297, "grad_norm": 0.5568501353263855, "learning_rate": 0.0002, "epoch": 0.8180421954571174, "step": 5060}, {"loss": 0.7733, "grad_norm": 0.49148809909820557, "learning_rate": 0.0002, "epoch": 0.8196588796378628, "step": 5070}, {"loss": 0.8054, "grad_norm": 0.4956012964248657, "learning_rate": 0.0002, "epoch": 0.821275563818608, "step": 5080}, {"loss": 0.8201, "grad_norm": 0.6078833937644958, "learning_rate": 0.0002, "epoch": 0.8228922479993533, "step": 5090}, {"loss": 0.828, "grad_norm": 0.46906954050064087, "learning_rate": 0.0002, "epoch": 0.8245089321800986, "step": 5100}, {"loss": 0.7703, "grad_norm": 0.50812166929245, "learning_rate": 0.0002, "epoch": 0.8261256163608439, "step": 5110}, {"loss": 0.8243, "grad_norm": 0.5319661498069763, "learning_rate": 0.0002, "epoch": 0.8277423005415891, "step": 5120}, {"loss": 0.7798, "grad_norm": 0.4949689209461212, "learning_rate": 0.0002, "epoch": 0.8293589847223345, "step": 5130}, {"loss": 0.7428, "grad_norm": 0.5151591300964355, "learning_rate": 0.0002, "epoch": 0.8309756689030798, "step": 5140}, {"loss": 0.8147, "grad_norm": 0.5530214309692383, "learning_rate": 0.0002, "epoch": 0.8325923530838251, "step": 5150}, {"loss": 0.8251, "grad_norm": 0.6297410130500793, "learning_rate": 0.0002, "epoch": 0.8342090372645704, "step": 5160}, {"loss": 0.8067, "grad_norm": 0.5466840267181396, "learning_rate": 0.0002, "epoch": 0.8358257214453156, "step": 5170}, {"loss": 0.7875, "grad_norm": 0.652913510799408, "learning_rate": 0.0002, "epoch": 0.837442405626061, "step": 5180}, {"loss": 0.8295, "grad_norm": 0.5811293125152588, "learning_rate": 0.0002, "epoch": 0.8390590898068062, "step": 5190}, {"loss": 0.7412, "grad_norm": 0.5109550952911377, "learning_rate": 0.0002, "epoch": 0.8406757739875516, "step": 5200}, {"loss": 0.8077, "grad_norm": 0.4551706612110138, "learning_rate": 0.0002, "epoch": 0.8422924581682968, "step": 5210}, {"loss": 0.7827, "grad_norm": 0.5813754200935364, "learning_rate": 0.0002, "epoch": 0.8439091423490421, "step": 5220}, {"loss": 0.802, "grad_norm": 0.5856947898864746, "learning_rate": 0.0002, "epoch": 0.8455258265297874, "step": 5230}, {"loss": 0.7957, "grad_norm": 0.5482739210128784, "learning_rate": 0.0002, "epoch": 0.8471425107105327, "step": 5240}, {"loss": 0.8295, "grad_norm": 0.49023720622062683, "learning_rate": 0.0002, "epoch": 0.8487591948912779, "step": 5250}, {"loss": 0.8022, "grad_norm": 0.49472475051879883, "learning_rate": 0.0002, "epoch": 0.8503758790720233, "step": 5260}, {"loss": 0.8001, "grad_norm": 0.5490226745605469, "learning_rate": 0.0002, "epoch": 0.8519925632527686, "step": 5270}, {"loss": 0.8333, "grad_norm": 0.5340665578842163, "learning_rate": 0.0002, "epoch": 0.8536092474335139, "step": 5280}, {"loss": 0.8277, "grad_norm": 0.5962483882904053, "learning_rate": 0.0002, "epoch": 0.8552259316142592, "step": 5290}, {"loss": 0.8765, "grad_norm": 0.586358368396759, "learning_rate": 0.0002, "epoch": 0.8568426157950044, "step": 5300}, {"loss": 0.7831, "grad_norm": 0.49120277166366577, "learning_rate": 0.0002, "epoch": 0.8584592999757498, "step": 5310}, {"loss": 0.8162, "grad_norm": 0.5887332558631897, "learning_rate": 0.0002, "epoch": 0.860075984156495, "step": 5320}, {"loss": 0.7464, "grad_norm": 0.42496153712272644, "learning_rate": 0.0002, "epoch": 0.8616926683372403, "step": 5330}, {"loss": 0.7905, "grad_norm": 0.5489874482154846, "learning_rate": 0.0002, "epoch": 0.8633093525179856, "step": 5340}, {"loss": 0.7958, "grad_norm": 0.5850813984870911, "learning_rate": 0.0002, "epoch": 0.8649260366987309, "step": 5350}, {"loss": 0.7642, "grad_norm": 0.517487108707428, "learning_rate": 0.0002, "epoch": 0.8665427208794761, "step": 5360}, {"loss": 0.7801, "grad_norm": 0.5339142680168152, "learning_rate": 0.0002, "epoch": 0.8681594050602215, "step": 5370}, {"loss": 0.818, "grad_norm": 0.6236387491226196, "learning_rate": 0.0002, "epoch": 0.8697760892409668, "step": 5380}, {"loss": 0.7708, "grad_norm": 0.5752192735671997, "learning_rate": 0.0002, "epoch": 0.8713927734217121, "step": 5390}, {"loss": 0.8542, "grad_norm": 0.6724614500999451, "learning_rate": 0.0002, "epoch": 0.8730094576024574, "step": 5400}, {"loss": 0.7581, "grad_norm": 0.5280613303184509, "learning_rate": 0.0002, "epoch": 0.8746261417832026, "step": 5410}, {"loss": 0.8231, "grad_norm": 0.44033288955688477, "learning_rate": 0.0002, "epoch": 0.876242825963948, "step": 5420}, {"loss": 0.8839, "grad_norm": 0.5199708342552185, "learning_rate": 0.0002, "epoch": 0.8778595101446932, "step": 5430}, {"loss": 0.7852, "grad_norm": 0.46778348088264465, "learning_rate": 0.0002, "epoch": 0.8794761943254386, "step": 5440}, {"loss": 0.7834, "grad_norm": 0.4657754898071289, "learning_rate": 0.0002, "epoch": 0.8810928785061838, "step": 5450}, {"loss": 0.7799, "grad_norm": 0.5472902655601501, "learning_rate": 0.0002, "epoch": 0.8827095626869291, "step": 5460}, {"loss": 0.8253, "grad_norm": 0.4876766800880432, "learning_rate": 0.0002, "epoch": 0.8843262468676744, "step": 5470}, {"loss": 0.7906, "grad_norm": 0.5057248473167419, "learning_rate": 0.0002, "epoch": 0.8859429310484197, "step": 5480}, {"loss": 0.8124, "grad_norm": 0.4637320637702942, "learning_rate": 0.0002, "epoch": 0.8875596152291649, "step": 5490}, {"loss": 0.781, "grad_norm": 0.471955806016922, "learning_rate": 0.0002, "epoch": 0.8891762994099103, "step": 5500}, {"loss": 0.8057, "grad_norm": 0.5209813714027405, "learning_rate": 0.0002, "epoch": 0.8907929835906556, "step": 5510}, {"loss": 0.8106, "grad_norm": 0.6213834285736084, "learning_rate": 0.0002, "epoch": 0.8924096677714008, "step": 5520}, {"loss": 0.7787, "grad_norm": 0.5215408205986023, "learning_rate": 0.0002, "epoch": 0.8940263519521462, "step": 5530}, {"loss": 0.8174, "grad_norm": 0.580478310585022, "learning_rate": 0.0002, "epoch": 0.8956430361328914, "step": 5540}, {"loss": 0.8371, "grad_norm": 0.49102169275283813, "learning_rate": 0.0002, "epoch": 0.8972597203136368, "step": 5550}, {"loss": 0.7806, "grad_norm": 0.6043479442596436, "learning_rate": 0.0002, "epoch": 0.898876404494382, "step": 5560}, {"loss": 0.7754, "grad_norm": 0.5636463165283203, "learning_rate": 0.0002, "epoch": 0.9004930886751273, "step": 5570}, {"loss": 0.8145, "grad_norm": 0.5620124340057373, "learning_rate": 0.0002, "epoch": 0.9021097728558726, "step": 5580}, {"loss": 0.8083, "grad_norm": 0.5206354856491089, "learning_rate": 0.0002, "epoch": 0.9037264570366179, "step": 5590}, {"loss": 0.8557, "grad_norm": 0.5798229575157166, "learning_rate": 0.0002, "epoch": 0.9053431412173631, "step": 5600}, {"loss": 0.8097, "grad_norm": 0.6428212523460388, "learning_rate": 0.0002, "epoch": 0.9069598253981085, "step": 5610}, {"loss": 0.7839, "grad_norm": 0.48064687848091125, "learning_rate": 0.0002, "epoch": 0.9085765095788537, "step": 5620}, {"loss": 0.8343, "grad_norm": 0.6347860097885132, "learning_rate": 0.0002, "epoch": 0.9101931937595991, "step": 5630}, {"loss": 0.851, "grad_norm": 0.5353913307189941, "learning_rate": 0.0002, "epoch": 0.9118098779403444, "step": 5640}, {"loss": 0.7736, "grad_norm": 0.5323944091796875, "learning_rate": 0.0002, "epoch": 0.9134265621210896, "step": 5650}, {"loss": 0.8393, "grad_norm": 0.5261843204498291, "learning_rate": 0.0002, "epoch": 0.915043246301835, "step": 5660}, {"loss": 0.7355, "grad_norm": 0.5451326966285706, "learning_rate": 0.0002, "epoch": 0.9166599304825802, "step": 5670}, {"loss": 0.8012, "grad_norm": 0.5183324217796326, "learning_rate": 0.0002, "epoch": 0.9182766146633256, "step": 5680}, {"loss": 0.7659, "grad_norm": 0.47229018807411194, "learning_rate": 0.0002, "epoch": 0.9198932988440708, "step": 5690}, {"loss": 0.7757, "grad_norm": 0.49180513620376587, "learning_rate": 0.0002, "epoch": 0.9215099830248161, "step": 5700}, {"loss": 0.8735, "grad_norm": 0.5419785380363464, "learning_rate": 0.0002, "epoch": 0.9231266672055614, "step": 5710}, {"loss": 0.7378, "grad_norm": 0.5408698916435242, "learning_rate": 0.0002, "epoch": 0.9247433513863067, "step": 5720}, {"loss": 0.7701, "grad_norm": 0.5286232829093933, "learning_rate": 0.0002, "epoch": 0.9263600355670519, "step": 5730}, {"loss": 0.8242, "grad_norm": 0.7539758086204529, "learning_rate": 0.0002, "epoch": 0.9279767197477973, "step": 5740}, {"loss": 0.8118, "grad_norm": 0.5166944861412048, "learning_rate": 0.0002, "epoch": 0.9295934039285425, "step": 5750}, {"loss": 0.783, "grad_norm": 0.6601425409317017, "learning_rate": 0.0002, "epoch": 0.9312100881092878, "step": 5760}, {"loss": 0.7873, "grad_norm": 0.5029960870742798, "learning_rate": 0.0002, "epoch": 0.9328267722900332, "step": 5770}, {"loss": 0.7989, "grad_norm": 0.4926645755767822, "learning_rate": 0.0002, "epoch": 0.9344434564707784, "step": 5780}, {"loss": 0.8174, "grad_norm": 0.5739615559577942, "learning_rate": 0.0002, "epoch": 0.9360601406515238, "step": 5790}, {"loss": 0.8037, "grad_norm": 0.5058279037475586, "learning_rate": 0.0002, "epoch": 0.937676824832269, "step": 5800}, {"loss": 0.8537, "grad_norm": 0.5260962247848511, "learning_rate": 0.0002, "epoch": 0.9392935090130143, "step": 5810}, {"loss": 0.7486, "grad_norm": 0.5768588185310364, "learning_rate": 0.0002, "epoch": 0.9409101931937596, "step": 5820}, {"loss": 0.8215, "grad_norm": 0.5170126557350159, "learning_rate": 0.0002, "epoch": 0.9425268773745049, "step": 5830}, {"loss": 0.7422, "grad_norm": 0.5745864510536194, "learning_rate": 0.0002, "epoch": 0.9441435615552501, "step": 5840}, {"loss": 0.7824, "grad_norm": 0.5551357865333557, "learning_rate": 0.0002, "epoch": 0.9457602457359955, "step": 5850}, {"loss": 0.8529, "grad_norm": 0.5776078701019287, "learning_rate": 0.0002, "epoch": 0.9473769299167407, "step": 5860}, {"loss": 0.8527, "grad_norm": 0.5340062379837036, "learning_rate": 0.0002, "epoch": 0.9489936140974861, "step": 5870}, {"loss": 0.8217, "grad_norm": 0.6447290182113647, "learning_rate": 0.0002, "epoch": 0.9506102982782314, "step": 5880}, {"loss": 0.7945, "grad_norm": 0.5123815536499023, "learning_rate": 0.0002, "epoch": 0.9522269824589766, "step": 5890}, {"loss": 0.8209, "grad_norm": 0.48547613620758057, "learning_rate": 0.0002, "epoch": 0.953843666639722, "step": 5900}, {"loss": 0.7896, "grad_norm": 0.5791414976119995, "learning_rate": 0.0002, "epoch": 0.9554603508204672, "step": 5910}, {"loss": 0.8408, "grad_norm": 0.6195011734962463, "learning_rate": 0.0002, "epoch": 0.9570770350012126, "step": 5920}, {"loss": 0.7805, "grad_norm": 0.6323803067207336, "learning_rate": 0.0002, "epoch": 0.9586937191819578, "step": 5930}, {"loss": 0.8484, "grad_norm": 0.45552879571914673, "learning_rate": 0.0002, "epoch": 0.9603104033627031, "step": 5940}, {"loss": 0.7367, "grad_norm": 0.5796473622322083, "learning_rate": 0.0002, "epoch": 0.9619270875434484, "step": 5950}, {"loss": 0.7672, "grad_norm": 0.647261381149292, "learning_rate": 0.0002, "epoch": 0.9635437717241937, "step": 5960}, {"loss": 0.8086, "grad_norm": 0.5487682819366455, "learning_rate": 0.0002, "epoch": 0.9651604559049389, "step": 5970}, {"loss": 0.7973, "grad_norm": 0.5743663907051086, "learning_rate": 0.0002, "epoch": 0.9667771400856843, "step": 5980}, {"loss": 0.8153, "grad_norm": 0.5470591187477112, "learning_rate": 0.0002, "epoch": 0.9683938242664295, "step": 5990}, {"loss": 0.8119, "grad_norm": 0.5901660323143005, "learning_rate": 0.0002, "epoch": 0.9700105084471748, "step": 6000}, {"loss": 0.8147, "grad_norm": 0.6544759273529053, "learning_rate": 0.0002, "epoch": 0.9716271926279202, "step": 6010}, {"loss": 0.7536, "grad_norm": 0.6288470029830933, "learning_rate": 0.0002, "epoch": 0.9732438768086654, "step": 6020}, {"loss": 0.7989, "grad_norm": 0.673153817653656, "learning_rate": 0.0002, "epoch": 0.9748605609894108, "step": 6030}, {"loss": 0.7556, "grad_norm": 0.42854753136634827, "learning_rate": 0.0002, "epoch": 0.976477245170156, "step": 6040}, {"loss": 0.8006, "grad_norm": 0.5227066278457642, "learning_rate": 0.0002, "epoch": 0.9780939293509013, "step": 6050}, {"loss": 0.795, "grad_norm": 0.5372416973114014, "learning_rate": 0.0002, "epoch": 0.9797106135316466, "step": 6060}, {"loss": 0.7591, "grad_norm": 0.6026402115821838, "learning_rate": 0.0002, "epoch": 0.9813272977123919, "step": 6070}, {"loss": 0.8347, "grad_norm": 0.49547791481018066, "learning_rate": 0.0002, "epoch": 0.9829439818931371, "step": 6080}, {"loss": 0.7722, "grad_norm": 0.4641951322555542, "learning_rate": 0.0002, "epoch": 0.9845606660738825, "step": 6090}, {"loss": 0.8125, "grad_norm": 0.5818535089492798, "learning_rate": 0.0002, "epoch": 0.9861773502546277, "step": 6100}, {"loss": 0.81, "grad_norm": 0.63955157995224, "learning_rate": 0.0002, "epoch": 0.9877940344353731, "step": 6110}, {"loss": 0.7547, "grad_norm": 0.5649438500404358, "learning_rate": 0.0002, "epoch": 0.9894107186161183, "step": 6120}, {"loss": 0.7861, "grad_norm": 0.5290433168411255, "learning_rate": 0.0002, "epoch": 0.9910274027968636, "step": 6130}, {"loss": 0.8109, "grad_norm": 0.6399374008178711, "learning_rate": 0.0002, "epoch": 0.992644086977609, "step": 6140}, {"loss": 0.8373, "grad_norm": 0.6736576557159424, "learning_rate": 0.0002, "epoch": 0.9942607711583542, "step": 6150}, {"loss": 0.7915, "grad_norm": 0.515420138835907, "learning_rate": 0.0002, "epoch": 0.9958774553390995, "step": 6160}, {"loss": 0.8032, "grad_norm": 0.562677800655365, "learning_rate": 0.0002, "epoch": 0.9974941395198448, "step": 6170}, {"loss": 0.8187, "grad_norm": 0.7113858461380005, "learning_rate": 0.0002, "epoch": 0.9991108237005901, "step": 6180}, {"eval_loss": 1.0871200561523438, "eval_runtime": 122.2071, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 0.9999191657909627, "step": 6185}, {"loss": 0.7507, "grad_norm": 0.7111801505088806, "learning_rate": 0.0002, "epoch": 1.0007275078813354, "step": 6190}, {"loss": 0.6865, "grad_norm": 0.5402125716209412, "learning_rate": 0.0002, "epoch": 1.0023441920620806, "step": 6200}, {"loss": 0.7625, "grad_norm": 0.6098830103874207, "learning_rate": 0.0002, "epoch": 1.003960876242826, "step": 6210}, {"loss": 0.7631, "grad_norm": 0.5829983353614807, "learning_rate": 0.0002, "epoch": 1.0055775604235713, "step": 6220}, {"loss": 0.7188, "grad_norm": 0.5614621043205261, "learning_rate": 0.0002, "epoch": 1.0071942446043165, "step": 6230}, {"loss": 0.7505, "grad_norm": 0.5954238772392273, "learning_rate": 0.0002, "epoch": 1.0088109287850617, "step": 6240}, {"loss": 0.7448, "grad_norm": 0.6480574607849121, "learning_rate": 0.0002, "epoch": 1.0104276129658072, "step": 6250}, {"loss": 0.7514, "grad_norm": 0.6051128506660461, "learning_rate": 0.0002, "epoch": 1.0120442971465524, "step": 6260}, {"loss": 0.7237, "grad_norm": 0.6318870782852173, "learning_rate": 0.0002, "epoch": 1.0136609813272976, "step": 6270}, {"loss": 0.7178, "grad_norm": 0.5048980116844177, "learning_rate": 0.0002, "epoch": 1.015277665508043, "step": 6280}, {"loss": 0.7391, "grad_norm": 0.6346936225891113, "learning_rate": 0.0002, "epoch": 1.0168943496887883, "step": 6290}, {"loss": 0.7486, "grad_norm": 0.5711665749549866, "learning_rate": 0.0002, "epoch": 1.0185110338695336, "step": 6300}, {"loss": 0.6808, "grad_norm": 0.5175361037254333, "learning_rate": 0.0002, "epoch": 1.0201277180502788, "step": 6310}, {"loss": 0.7539, "grad_norm": 0.5360831618309021, "learning_rate": 0.0002, "epoch": 1.0217444022310243, "step": 6320}, {"loss": 0.7112, "grad_norm": 0.614675760269165, "learning_rate": 0.0002, "epoch": 1.0233610864117695, "step": 6330}, {"loss": 0.7748, "grad_norm": 0.5626118183135986, "learning_rate": 0.0002, "epoch": 1.0249777705925147, "step": 6340}, {"loss": 0.7375, "grad_norm": 0.574897289276123, "learning_rate": 0.0002, "epoch": 1.02659445477326, "step": 6350}, {"loss": 0.759, "grad_norm": 0.7185447812080383, "learning_rate": 0.0002, "epoch": 1.0282111389540054, "step": 6360}, {"loss": 0.703, "grad_norm": 0.6705799698829651, "learning_rate": 0.0002, "epoch": 1.0298278231347506, "step": 6370}, {"loss": 0.7139, "grad_norm": 0.6740428805351257, "learning_rate": 0.0002, "epoch": 1.0314445073154959, "step": 6380}, {"loss": 0.7252, "grad_norm": 0.663902759552002, "learning_rate": 0.0002, "epoch": 1.0330611914962413, "step": 6390}, {"loss": 0.7065, "grad_norm": 0.5029543042182922, "learning_rate": 0.0002, "epoch": 1.0346778756769865, "step": 6400}, {"loss": 0.711, "grad_norm": 0.7813863158226013, "learning_rate": 0.0002, "epoch": 1.0362945598577318, "step": 6410}, {"loss": 0.7433, "grad_norm": 0.5396282076835632, "learning_rate": 0.0002, "epoch": 1.037911244038477, "step": 6420}, {"loss": 0.7222, "grad_norm": 0.5253293514251709, "learning_rate": 0.0002, "epoch": 1.0395279282192225, "step": 6430}, {"loss": 0.715, "grad_norm": 0.7236770987510681, "learning_rate": 0.0002, "epoch": 1.0411446123999677, "step": 6440}, {"loss": 0.7259, "grad_norm": 0.5670917630195618, "learning_rate": 0.0002, "epoch": 1.042761296580713, "step": 6450}, {"loss": 0.7195, "grad_norm": 0.6031978726387024, "learning_rate": 0.0002, "epoch": 1.0443779807614582, "step": 6460}, {"loss": 0.7648, "grad_norm": 0.5309213399887085, "learning_rate": 0.0002, "epoch": 1.0459946649422036, "step": 6470}, {"loss": 0.7161, "grad_norm": 0.7114651799201965, "learning_rate": 0.0002, "epoch": 1.0476113491229488, "step": 6480}, {"loss": 0.7583, "grad_norm": 0.5591610670089722, "learning_rate": 0.0002, "epoch": 1.049228033303694, "step": 6490}, {"loss": 0.6645, "grad_norm": 0.5185961127281189, "learning_rate": 0.0002, "epoch": 1.0508447174844395, "step": 6500}, {"loss": 0.7654, "grad_norm": 0.6510552167892456, "learning_rate": 0.0002, "epoch": 1.0524614016651848, "step": 6510}, {"loss": 0.7057, "grad_norm": 0.6557928919792175, "learning_rate": 0.0002, "epoch": 1.05407808584593, "step": 6520}, {"loss": 0.8056, "grad_norm": 0.6973192691802979, "learning_rate": 0.0002, "epoch": 1.0556947700266752, "step": 6530}, {"loss": 0.6793, "grad_norm": 0.6226583123207092, "learning_rate": 0.0002, "epoch": 1.0573114542074207, "step": 6540}, {"loss": 0.7151, "grad_norm": 0.5633195638656616, "learning_rate": 0.0002, "epoch": 1.058928138388166, "step": 6550}, {"loss": 0.7082, "grad_norm": 0.7466658353805542, "learning_rate": 0.0002, "epoch": 1.0605448225689111, "step": 6560}, {"loss": 0.7059, "grad_norm": 0.6462772488594055, "learning_rate": 0.0002, "epoch": 1.0621615067496564, "step": 6570}, {"loss": 0.7046, "grad_norm": 0.5266856551170349, "learning_rate": 0.0002, "epoch": 1.0637781909304018, "step": 6580}, {"loss": 0.7157, "grad_norm": 0.534392774105072, "learning_rate": 0.0002, "epoch": 1.065394875111147, "step": 6590}, {"loss": 0.7115, "grad_norm": 0.7514177560806274, "learning_rate": 0.0002, "epoch": 1.0670115592918923, "step": 6600}, {"loss": 0.7545, "grad_norm": 0.7593035697937012, "learning_rate": 0.0002, "epoch": 1.0686282434726375, "step": 6610}, {"loss": 0.6836, "grad_norm": 0.5277858972549438, "learning_rate": 0.0002, "epoch": 1.070244927653383, "step": 6620}, {"loss": 0.7405, "grad_norm": 0.5573670268058777, "learning_rate": 0.0002, "epoch": 1.0718616118341282, "step": 6630}, {"loss": 0.6774, "grad_norm": 0.6802396774291992, "learning_rate": 0.0002, "epoch": 1.0734782960148734, "step": 6640}, {"loss": 0.723, "grad_norm": 0.7367215752601624, "learning_rate": 0.0002, "epoch": 1.0750949801956189, "step": 6650}, {"loss": 0.7429, "grad_norm": 0.5961891412734985, "learning_rate": 0.0002, "epoch": 1.0767116643763641, "step": 6660}, {"loss": 0.6791, "grad_norm": 0.5736313462257385, "learning_rate": 0.0002, "epoch": 1.0783283485571094, "step": 6670}, {"loss": 0.7178, "grad_norm": 0.619219183921814, "learning_rate": 0.0002, "epoch": 1.0799450327378546, "step": 6680}, {"loss": 0.7318, "grad_norm": 0.6214390993118286, "learning_rate": 0.0002, "epoch": 1.0815617169186, "step": 6690}, {"loss": 0.7554, "grad_norm": 0.564536988735199, "learning_rate": 0.0002, "epoch": 1.0831784010993453, "step": 6700}, {"loss": 0.7362, "grad_norm": 0.5838140249252319, "learning_rate": 0.0002, "epoch": 1.0847950852800905, "step": 6710}, {"loss": 0.739, "grad_norm": 0.7000553607940674, "learning_rate": 0.0002, "epoch": 1.0864117694608357, "step": 6720}, {"loss": 0.7369, "grad_norm": 0.7078263759613037, "learning_rate": 0.0002, "epoch": 1.0880284536415812, "step": 6730}, {"loss": 0.7654, "grad_norm": 0.8353848457336426, "learning_rate": 0.0002, "epoch": 1.0896451378223264, "step": 6740}, {"loss": 0.7015, "grad_norm": 0.5615518689155579, "learning_rate": 0.0002, "epoch": 1.0912618220030716, "step": 6750}, {"loss": 0.7396, "grad_norm": 0.5475581288337708, "learning_rate": 0.0002, "epoch": 1.0928785061838169, "step": 6760}, {"loss": 0.7652, "grad_norm": 0.5835978388786316, "learning_rate": 0.0002, "epoch": 1.0944951903645623, "step": 6770}, {"loss": 0.7541, "grad_norm": 0.5516105890274048, "learning_rate": 0.0002, "epoch": 1.0961118745453076, "step": 6780}, {"loss": 0.6842, "grad_norm": 0.5875251889228821, "learning_rate": 0.0002, "epoch": 1.0977285587260528, "step": 6790}, {"loss": 0.6903, "grad_norm": 0.7376947999000549, "learning_rate": 0.0002, "epoch": 1.0993452429067982, "step": 6800}, {"loss": 0.7512, "grad_norm": 0.5656165480613708, "learning_rate": 0.0002, "epoch": 1.1009619270875435, "step": 6810}, {"loss": 0.7409, "grad_norm": 0.6365954279899597, "learning_rate": 0.0002, "epoch": 1.1025786112682887, "step": 6820}, {"loss": 0.7392, "grad_norm": 0.5033080577850342, "learning_rate": 0.0002, "epoch": 1.104195295449034, "step": 6830}, {"loss": 0.6909, "grad_norm": 0.617396891117096, "learning_rate": 0.0002, "epoch": 1.1058119796297794, "step": 6840}, {"loss": 0.7006, "grad_norm": 0.6395374536514282, "learning_rate": 0.0002, "epoch": 1.1074286638105246, "step": 6850}, {"loss": 0.7335, "grad_norm": 0.6775295734405518, "learning_rate": 0.0002, "epoch": 1.1090453479912699, "step": 6860}, {"loss": 0.764, "grad_norm": 0.6655223965644836, "learning_rate": 0.0002, "epoch": 1.1106620321720153, "step": 6870}, {"loss": 0.7553, "grad_norm": 0.676655113697052, "learning_rate": 0.0002, "epoch": 1.1122787163527605, "step": 6880}, {"loss": 0.7342, "grad_norm": 0.6062718629837036, "learning_rate": 0.0002, "epoch": 1.1138954005335058, "step": 6890}, {"loss": 0.7446, "grad_norm": 0.590943455696106, "learning_rate": 0.0002, "epoch": 1.115512084714251, "step": 6900}, {"loss": 0.6705, "grad_norm": 0.6315317153930664, "learning_rate": 0.0002, "epoch": 1.1171287688949965, "step": 6910}, {"loss": 0.6912, "grad_norm": 0.47979024052619934, "learning_rate": 0.0002, "epoch": 1.1187454530757417, "step": 6920}, {"loss": 0.7002, "grad_norm": 0.647298276424408, "learning_rate": 0.0002, "epoch": 1.120362137256487, "step": 6930}, {"loss": 0.7502, "grad_norm": 0.7336484789848328, "learning_rate": 0.0002, "epoch": 1.1219788214372322, "step": 6940}, {"loss": 0.693, "grad_norm": 0.5071424245834351, "learning_rate": 0.0002, "epoch": 1.1235955056179776, "step": 6950}, {"loss": 0.7378, "grad_norm": 0.6527144312858582, "learning_rate": 0.0002, "epoch": 1.1252121897987228, "step": 6960}, {"loss": 0.7228, "grad_norm": 0.6935935020446777, "learning_rate": 0.0002, "epoch": 1.126828873979468, "step": 6970}, {"loss": 0.699, "grad_norm": 0.8026931881904602, "learning_rate": 0.0002, "epoch": 1.1284455581602133, "step": 6980}, {"loss": 0.7361, "grad_norm": 0.5210393667221069, "learning_rate": 0.0002, "epoch": 1.1300622423409588, "step": 6990}, {"loss": 0.7456, "grad_norm": 0.60475093126297, "learning_rate": 0.0002, "epoch": 1.131678926521704, "step": 7000}, {"loss": 0.7495, "grad_norm": 0.6417073607444763, "learning_rate": 0.0002, "epoch": 1.1332956107024492, "step": 7010}, {"loss": 0.7459, "grad_norm": 0.6732175946235657, "learning_rate": 0.0002, "epoch": 1.1349122948831947, "step": 7020}, {"loss": 0.7278, "grad_norm": 0.6719491481781006, "learning_rate": 0.0002, "epoch": 1.13652897906394, "step": 7030}, {"loss": 0.7694, "grad_norm": 0.5708295106887817, "learning_rate": 0.0002, "epoch": 1.1381456632446851, "step": 7040}, {"loss": 0.7823, "grad_norm": 0.7141719460487366, "learning_rate": 0.0002, "epoch": 1.1397623474254304, "step": 7050}, {"loss": 0.764, "grad_norm": 0.6187017560005188, "learning_rate": 0.0002, "epoch": 1.1413790316061758, "step": 7060}, {"loss": 0.7657, "grad_norm": 0.50581294298172, "learning_rate": 0.0002, "epoch": 1.142995715786921, "step": 7070}, {"loss": 0.7357, "grad_norm": 0.5620143413543701, "learning_rate": 0.0002, "epoch": 1.1446123999676663, "step": 7080}, {"loss": 0.7287, "grad_norm": 0.6231929659843445, "learning_rate": 0.0002, "epoch": 1.1462290841484115, "step": 7090}, {"loss": 0.7328, "grad_norm": 0.5775774121284485, "learning_rate": 0.0002, "epoch": 1.147845768329157, "step": 7100}, {"loss": 0.7728, "grad_norm": 0.6492809653282166, "learning_rate": 0.0002, "epoch": 1.1494624525099022, "step": 7110}, {"loss": 0.7545, "grad_norm": 0.6434972286224365, "learning_rate": 0.0002, "epoch": 1.1510791366906474, "step": 7120}, {"loss": 0.7374, "grad_norm": 0.6191812753677368, "learning_rate": 0.0002, "epoch": 1.1526958208713927, "step": 7130}, {"loss": 0.7276, "grad_norm": 0.6690331697463989, "learning_rate": 0.0002, "epoch": 1.1543125050521381, "step": 7140}, {"loss": 0.7704, "grad_norm": 0.5977938175201416, "learning_rate": 0.0002, "epoch": 1.1559291892328833, "step": 7150}, {"loss": 0.7251, "grad_norm": 0.6195854544639587, "learning_rate": 0.0002, "epoch": 1.1575458734136286, "step": 7160}, {"loss": 0.7249, "grad_norm": 0.5752048492431641, "learning_rate": 0.0002, "epoch": 1.159162557594374, "step": 7170}, {"loss": 0.7593, "grad_norm": 0.589081883430481, "learning_rate": 0.0002, "epoch": 1.1607792417751193, "step": 7180}, {"loss": 0.704, "grad_norm": 0.756996750831604, "learning_rate": 0.0002, "epoch": 1.1623959259558645, "step": 7190}, {"loss": 0.7404, "grad_norm": 0.7614967226982117, "learning_rate": 0.0002, "epoch": 1.1640126101366097, "step": 7200}, {"loss": 0.7867, "grad_norm": 0.6120437979698181, "learning_rate": 0.0002, "epoch": 1.1656292943173552, "step": 7210}, {"loss": 0.7384, "grad_norm": 0.6210004687309265, "learning_rate": 0.0002, "epoch": 1.1672459784981004, "step": 7220}, {"loss": 0.7251, "grad_norm": 0.6044116020202637, "learning_rate": 0.0002, "epoch": 1.1688626626788456, "step": 7230}, {"loss": 0.7361, "grad_norm": 0.5418457388877869, "learning_rate": 0.0002, "epoch": 1.170479346859591, "step": 7240}, {"loss": 0.6938, "grad_norm": 0.6413537263870239, "learning_rate": 0.0002, "epoch": 1.1720960310403363, "step": 7250}, {"loss": 0.6978, "grad_norm": 0.5777867436408997, "learning_rate": 0.0002, "epoch": 1.1737127152210816, "step": 7260}, {"loss": 0.7503, "grad_norm": 0.7092402577400208, "learning_rate": 0.0002, "epoch": 1.1753293994018268, "step": 7270}, {"loss": 0.7487, "grad_norm": 0.6351709365844727, "learning_rate": 0.0002, "epoch": 1.176946083582572, "step": 7280}, {"loss": 0.7527, "grad_norm": 0.6172189712524414, "learning_rate": 0.0002, "epoch": 1.1785627677633175, "step": 7290}, {"loss": 0.7319, "grad_norm": 0.6801714897155762, "learning_rate": 0.0002, "epoch": 1.1801794519440627, "step": 7300}, {"loss": 0.6941, "grad_norm": 0.6044712066650391, "learning_rate": 0.0002, "epoch": 1.181796136124808, "step": 7310}, {"loss": 0.6951, "grad_norm": 0.7413212060928345, "learning_rate": 0.0002, "epoch": 1.1834128203055534, "step": 7320}, {"loss": 0.7396, "grad_norm": 0.5303856134414673, "learning_rate": 0.0002, "epoch": 1.1850295044862986, "step": 7330}, {"loss": 0.6915, "grad_norm": 0.5647098422050476, "learning_rate": 0.0002, "epoch": 1.1866461886670439, "step": 7340}, {"loss": 0.7506, "grad_norm": 0.7374135255813599, "learning_rate": 0.0002, "epoch": 1.188262872847789, "step": 7350}, {"loss": 0.7041, "grad_norm": 0.5710089206695557, "learning_rate": 0.0002, "epoch": 1.1898795570285345, "step": 7360}, {"loss": 0.8289, "grad_norm": 0.6073619723320007, "learning_rate": 0.0002, "epoch": 1.1914962412092798, "step": 7370}, {"loss": 0.7722, "grad_norm": 0.5899916887283325, "learning_rate": 0.0002, "epoch": 1.193112925390025, "step": 7380}, {"loss": 0.756, "grad_norm": 0.7762434482574463, "learning_rate": 0.0002, "epoch": 1.1947296095707705, "step": 7390}, {"loss": 0.7319, "grad_norm": 0.679949939250946, "learning_rate": 0.0002, "epoch": 1.1963462937515157, "step": 7400}, {"loss": 0.7599, "grad_norm": 0.6106849312782288, "learning_rate": 0.0002, "epoch": 1.197962977932261, "step": 7410}, {"loss": 0.7648, "grad_norm": 0.682461678981781, "learning_rate": 0.0002, "epoch": 1.1995796621130062, "step": 7420}, {"loss": 0.7741, "grad_norm": 0.6087017059326172, "learning_rate": 0.0002, "epoch": 1.2011963462937516, "step": 7430}, {"loss": 0.7642, "grad_norm": 0.63739013671875, "learning_rate": 0.0002, "epoch": 1.2028130304744968, "step": 7440}, {"loss": 0.7611, "grad_norm": 0.6154777407646179, "learning_rate": 0.0002, "epoch": 1.204429714655242, "step": 7450}, {"loss": 0.7565, "grad_norm": 0.7491534948348999, "learning_rate": 0.0002, "epoch": 1.2060463988359873, "step": 7460}, {"loss": 0.698, "grad_norm": 0.6664797067642212, "learning_rate": 0.0002, "epoch": 1.2076630830167328, "step": 7470}, {"loss": 0.7456, "grad_norm": 0.6660266518592834, "learning_rate": 0.0002, "epoch": 1.209279767197478, "step": 7480}, {"loss": 0.714, "grad_norm": 0.6972551345825195, "learning_rate": 0.0002, "epoch": 1.2108964513782232, "step": 7490}, {"loss": 0.7023, "grad_norm": 0.6157945990562439, "learning_rate": 0.0002, "epoch": 1.2125131355589684, "step": 7500}, {"loss": 0.7326, "grad_norm": 0.5199310183525085, "learning_rate": 0.0002, "epoch": 1.214129819739714, "step": 7510}, {"loss": 0.7586, "grad_norm": 0.577610433101654, "learning_rate": 0.0002, "epoch": 1.2157465039204591, "step": 7520}, {"loss": 0.7179, "grad_norm": 0.53652423620224, "learning_rate": 0.0002, "epoch": 1.2173631881012044, "step": 7530}, {"loss": 0.7393, "grad_norm": 0.6479050517082214, "learning_rate": 0.0002, "epoch": 1.2189798722819498, "step": 7540}, {"loss": 0.7534, "grad_norm": 0.618748128414154, "learning_rate": 0.0002, "epoch": 1.220596556462695, "step": 7550}, {"loss": 0.6886, "grad_norm": 0.6311424374580383, "learning_rate": 0.0002, "epoch": 1.2222132406434403, "step": 7560}, {"loss": 0.7272, "grad_norm": 0.6595825552940369, "learning_rate": 0.0002, "epoch": 1.2238299248241855, "step": 7570}, {"loss": 0.7353, "grad_norm": 0.5198960900306702, "learning_rate": 0.0002, "epoch": 1.225446609004931, "step": 7580}, {"loss": 0.674, "grad_norm": 0.578650712966919, "learning_rate": 0.0002, "epoch": 1.2270632931856762, "step": 7590}, {"loss": 0.7507, "grad_norm": 0.6080220937728882, "learning_rate": 0.0002, "epoch": 1.2286799773664214, "step": 7600}, {"loss": 0.7733, "grad_norm": 0.7050248384475708, "learning_rate": 0.0002, "epoch": 1.2302966615471669, "step": 7610}, {"loss": 0.7032, "grad_norm": 0.6652196049690247, "learning_rate": 0.0002, "epoch": 1.2319133457279121, "step": 7620}, {"loss": 0.7085, "grad_norm": 0.7322776317596436, "learning_rate": 0.0002, "epoch": 1.2335300299086573, "step": 7630}, {"loss": 0.7402, "grad_norm": 0.4998728036880493, "learning_rate": 0.0002, "epoch": 1.2351467140894026, "step": 7640}, {"loss": 0.7214, "grad_norm": 0.6428788900375366, "learning_rate": 0.0002, "epoch": 1.2367633982701478, "step": 7650}, {"loss": 0.7699, "grad_norm": 0.585242509841919, "learning_rate": 0.0002, "epoch": 1.2383800824508933, "step": 7660}, {"loss": 0.7621, "grad_norm": 0.5211917757987976, "learning_rate": 0.0002, "epoch": 1.2399967666316385, "step": 7670}, {"loss": 0.746, "grad_norm": 0.6490384340286255, "learning_rate": 0.0002, "epoch": 1.2416134508123837, "step": 7680}, {"loss": 0.7186, "grad_norm": 0.6249763369560242, "learning_rate": 0.0002, "epoch": 1.2432301349931292, "step": 7690}, {"loss": 0.7761, "grad_norm": 0.71870356798172, "learning_rate": 0.0002, "epoch": 1.2448468191738744, "step": 7700}, {"loss": 0.7525, "grad_norm": 0.6761967539787292, "learning_rate": 0.0002, "epoch": 1.2464635033546196, "step": 7710}, {"loss": 0.7501, "grad_norm": 0.6500617265701294, "learning_rate": 0.0002, "epoch": 1.2480801875353649, "step": 7720}, {"loss": 0.7903, "grad_norm": 0.8069869875907898, "learning_rate": 0.0002, "epoch": 1.2496968717161103, "step": 7730}, {"loss": 0.6747, "grad_norm": 0.6044608950614929, "learning_rate": 0.0002, "epoch": 1.2513135558968556, "step": 7740}, {"loss": 0.6825, "grad_norm": 0.6573283076286316, "learning_rate": 0.0002, "epoch": 1.2529302400776008, "step": 7750}, {"loss": 0.7617, "grad_norm": 0.625430166721344, "learning_rate": 0.0002, "epoch": 1.2545469242583462, "step": 7760}, {"loss": 0.7041, "grad_norm": 0.5442022681236267, "learning_rate": 0.0002, "epoch": 1.2561636084390915, "step": 7770}, {"loss": 0.7172, "grad_norm": 0.6818386912345886, "learning_rate": 0.0002, "epoch": 1.2577802926198367, "step": 7780}, {"loss": 0.696, "grad_norm": 0.6381874084472656, "learning_rate": 0.0002, "epoch": 1.259396976800582, "step": 7790}, {"loss": 0.6834, "grad_norm": 0.6269212961196899, "learning_rate": 0.0002, "epoch": 1.2610136609813272, "step": 7800}, {"loss": 0.7821, "grad_norm": 0.600121259689331, "learning_rate": 0.0002, "epoch": 1.2626303451620726, "step": 7810}, {"loss": 0.7761, "grad_norm": 0.6337703466415405, "learning_rate": 0.0002, "epoch": 1.2642470293428179, "step": 7820}, {"loss": 0.732, "grad_norm": 0.7234963774681091, "learning_rate": 0.0002, "epoch": 1.2658637135235633, "step": 7830}, {"loss": 0.785, "grad_norm": 0.800184965133667, "learning_rate": 0.0002, "epoch": 1.2674803977043085, "step": 7840}, {"loss": 0.7426, "grad_norm": 0.7539464831352234, "learning_rate": 0.0002, "epoch": 1.2690970818850538, "step": 7850}, {"loss": 0.7496, "grad_norm": 0.5493760704994202, "learning_rate": 0.0002, "epoch": 1.270713766065799, "step": 7860}, {"loss": 0.7537, "grad_norm": 0.7477145791053772, "learning_rate": 0.0002, "epoch": 1.2723304502465442, "step": 7870}, {"loss": 0.7573, "grad_norm": 0.6366362571716309, "learning_rate": 0.0002, "epoch": 1.2739471344272897, "step": 7880}, {"loss": 0.7608, "grad_norm": 0.7419533729553223, "learning_rate": 0.0002, "epoch": 1.275563818608035, "step": 7890}, {"loss": 0.7873, "grad_norm": 0.6141223311424255, "learning_rate": 0.0002, "epoch": 1.2771805027887801, "step": 7900}, {"loss": 0.6916, "grad_norm": 0.7522598505020142, "learning_rate": 0.0002, "epoch": 1.2787971869695256, "step": 7910}, {"loss": 0.7097, "grad_norm": 0.6935804486274719, "learning_rate": 0.0002, "epoch": 1.2804138711502708, "step": 7920}, {"loss": 0.7185, "grad_norm": 0.7239290475845337, "learning_rate": 0.0002, "epoch": 1.282030555331016, "step": 7930}, {"loss": 0.7145, "grad_norm": 0.8800187110900879, "learning_rate": 0.0002, "epoch": 1.2836472395117613, "step": 7940}, {"loss": 0.6991, "grad_norm": 0.540458083152771, "learning_rate": 0.0002, "epoch": 1.2852639236925067, "step": 7950}, {"loss": 0.7139, "grad_norm": 0.6492934226989746, "learning_rate": 0.0002, "epoch": 1.286880607873252, "step": 7960}, {"loss": 0.7742, "grad_norm": 0.6543959379196167, "learning_rate": 0.0002, "epoch": 1.2884972920539972, "step": 7970}, {"loss": 0.7316, "grad_norm": 0.5804705619812012, "learning_rate": 0.0002, "epoch": 1.2901139762347427, "step": 7980}, {"loss": 0.796, "grad_norm": 0.7074727416038513, "learning_rate": 0.0002, "epoch": 1.291730660415488, "step": 7990}, {"loss": 0.7034, "grad_norm": 0.5347974300384521, "learning_rate": 0.0002, "epoch": 1.2933473445962331, "step": 8000}, {"loss": 0.738, "grad_norm": 0.6457298398017883, "learning_rate": 0.0002, "epoch": 1.2949640287769784, "step": 8010}, {"loss": 0.7634, "grad_norm": 0.6407219171524048, "learning_rate": 0.0002, "epoch": 1.2965807129577236, "step": 8020}, {"loss": 0.7506, "grad_norm": 0.828439474105835, "learning_rate": 0.0002, "epoch": 1.298197397138469, "step": 8030}, {"loss": 0.735, "grad_norm": 0.4840380549430847, "learning_rate": 0.0002, "epoch": 1.2998140813192143, "step": 8040}, {"loss": 0.7283, "grad_norm": 0.5921024680137634, "learning_rate": 0.0002, "epoch": 1.3014307654999595, "step": 8050}, {"loss": 0.7477, "grad_norm": 0.6170315146446228, "learning_rate": 0.0002, "epoch": 1.303047449680705, "step": 8060}, {"loss": 0.7534, "grad_norm": 0.5374847054481506, "learning_rate": 0.0002, "epoch": 1.3046641338614502, "step": 8070}, {"loss": 0.7593, "grad_norm": 0.545758068561554, "learning_rate": 0.0002, "epoch": 1.3062808180421954, "step": 8080}, {"loss": 0.7463, "grad_norm": 0.55641770362854, "learning_rate": 0.0002, "epoch": 1.3078975022229407, "step": 8090}, {"loss": 0.7594, "grad_norm": 0.6724897027015686, "learning_rate": 0.0002, "epoch": 1.309514186403686, "step": 8100}, {"loss": 0.7105, "grad_norm": 0.6923972368240356, "learning_rate": 0.0002, "epoch": 1.3111308705844313, "step": 8110}, {"loss": 0.7149, "grad_norm": 0.5136841535568237, "learning_rate": 0.0002, "epoch": 1.3127475547651766, "step": 8120}, {"loss": 0.7504, "grad_norm": 0.6766283512115479, "learning_rate": 0.0002, "epoch": 1.314364238945922, "step": 8130}, {"loss": 0.7489, "grad_norm": 0.6283926367759705, "learning_rate": 0.0002, "epoch": 1.3159809231266673, "step": 8140}, {"loss": 0.7459, "grad_norm": 0.644216001033783, "learning_rate": 0.0002, "epoch": 1.3175976073074125, "step": 8150}, {"loss": 0.7125, "grad_norm": 0.7827503085136414, "learning_rate": 0.0002, "epoch": 1.3192142914881577, "step": 8160}, {"loss": 0.7271, "grad_norm": 0.6651390790939331, "learning_rate": 0.0002, "epoch": 1.320830975668903, "step": 8170}, {"loss": 0.7778, "grad_norm": 0.5547412633895874, "learning_rate": 0.0002, "epoch": 1.3224476598496484, "step": 8180}, {"loss": 0.7402, "grad_norm": 0.6765179634094238, "learning_rate": 0.0002, "epoch": 1.3240643440303936, "step": 8190}, {"loss": 0.7106, "grad_norm": 0.6822077035903931, "learning_rate": 0.0002, "epoch": 1.325681028211139, "step": 8200}, {"loss": 0.7288, "grad_norm": 0.5941002368927002, "learning_rate": 0.0002, "epoch": 1.3272977123918843, "step": 8210}, {"loss": 0.7494, "grad_norm": 0.4850037097930908, "learning_rate": 0.0002, "epoch": 1.3289143965726296, "step": 8220}, {"loss": 0.7474, "grad_norm": 0.6162990927696228, "learning_rate": 0.0002, "epoch": 1.3305310807533748, "step": 8230}, {"loss": 0.7751, "grad_norm": 0.6665613651275635, "learning_rate": 0.0002, "epoch": 1.33214776493412, "step": 8240}, {"loss": 0.759, "grad_norm": 0.618192732334137, "learning_rate": 0.0002, "epoch": 1.3337644491148655, "step": 8250}, {"loss": 0.7532, "grad_norm": 0.710418701171875, "learning_rate": 0.0002, "epoch": 1.3353811332956107, "step": 8260}, {"loss": 0.7306, "grad_norm": 0.5109876990318298, "learning_rate": 0.0002, "epoch": 1.336997817476356, "step": 8270}, {"loss": 0.7303, "grad_norm": 0.6791711449623108, "learning_rate": 0.0002, "epoch": 1.3386145016571014, "step": 8280}, {"loss": 0.7594, "grad_norm": 0.6836432814598083, "learning_rate": 0.0002, "epoch": 1.3402311858378466, "step": 8290}, {"loss": 0.7594, "grad_norm": 0.5579386353492737, "learning_rate": 0.0002, "epoch": 1.3418478700185918, "step": 8300}, {"loss": 0.7377, "grad_norm": 0.6713546514511108, "learning_rate": 0.0002, "epoch": 1.343464554199337, "step": 8310}, {"loss": 0.7756, "grad_norm": 0.5353720188140869, "learning_rate": 0.0002, "epoch": 1.3450812383800825, "step": 8320}, {"loss": 0.718, "grad_norm": 0.5813682675361633, "learning_rate": 0.0002, "epoch": 1.3466979225608278, "step": 8330}, {"loss": 0.7294, "grad_norm": 0.8158791661262512, "learning_rate": 0.0002, "epoch": 1.348314606741573, "step": 8340}, {"loss": 0.6992, "grad_norm": 0.6193785071372986, "learning_rate": 0.0002, "epoch": 1.3499312909223184, "step": 8350}, {"loss": 0.7654, "grad_norm": 0.6353939771652222, "learning_rate": 0.0002, "epoch": 1.3515479751030637, "step": 8360}, {"loss": 0.7519, "grad_norm": 0.6925048232078552, "learning_rate": 0.0002, "epoch": 1.353164659283809, "step": 8370}, {"loss": 0.736, "grad_norm": 0.988264799118042, "learning_rate": 0.0002, "epoch": 1.3547813434645541, "step": 8380}, {"loss": 0.7744, "grad_norm": 0.6476002931594849, "learning_rate": 0.0002, "epoch": 1.3563980276452994, "step": 8390}, {"loss": 0.776, "grad_norm": 0.7120398879051208, "learning_rate": 0.0002, "epoch": 1.3580147118260448, "step": 8400}, {"loss": 0.7368, "grad_norm": 0.9048416614532471, "learning_rate": 0.0002, "epoch": 1.35963139600679, "step": 8410}, {"loss": 0.7544, "grad_norm": 0.7000672817230225, "learning_rate": 0.0002, "epoch": 1.3612480801875353, "step": 8420}, {"loss": 0.7358, "grad_norm": 0.6015632152557373, "learning_rate": 0.0002, "epoch": 1.3628647643682807, "step": 8430}, {"loss": 0.7298, "grad_norm": 0.612516462802887, "learning_rate": 0.0002, "epoch": 1.364481448549026, "step": 8440}, {"loss": 0.7055, "grad_norm": 0.5969301462173462, "learning_rate": 0.0002, "epoch": 1.3660981327297712, "step": 8450}, {"loss": 0.7754, "grad_norm": 0.6730654239654541, "learning_rate": 0.0002, "epoch": 1.3677148169105164, "step": 8460}, {"loss": 0.7465, "grad_norm": 0.6386392116546631, "learning_rate": 0.0002, "epoch": 1.369331501091262, "step": 8470}, {"loss": 0.7433, "grad_norm": 0.739544153213501, "learning_rate": 0.0002, "epoch": 1.3709481852720071, "step": 8480}, {"loss": 0.7892, "grad_norm": 0.6462782621383667, "learning_rate": 0.0002, "epoch": 1.3725648694527524, "step": 8490}, {"loss": 0.7302, "grad_norm": 0.7346843481063843, "learning_rate": 0.0002, "epoch": 1.3741815536334978, "step": 8500}, {"loss": 0.7634, "grad_norm": 0.6884821057319641, "learning_rate": 0.0002, "epoch": 1.375798237814243, "step": 8510}, {"loss": 0.7614, "grad_norm": 0.6999333500862122, "learning_rate": 0.0002, "epoch": 1.3774149219949883, "step": 8520}, {"loss": 0.729, "grad_norm": 0.5378713011741638, "learning_rate": 0.0002, "epoch": 1.3790316061757335, "step": 8530}, {"loss": 0.6797, "grad_norm": 0.5417906641960144, "learning_rate": 0.0002, "epoch": 1.3806482903564787, "step": 8540}, {"loss": 0.7499, "grad_norm": 0.6602526307106018, "learning_rate": 0.0002, "epoch": 1.3822649745372242, "step": 8550}, {"loss": 0.7356, "grad_norm": 0.7073674201965332, "learning_rate": 0.0002, "epoch": 1.3838816587179694, "step": 8560}, {"loss": 0.75, "grad_norm": 0.5841707587242126, "learning_rate": 0.0002, "epoch": 1.3854983428987149, "step": 8570}, {"loss": 0.732, "grad_norm": 0.7031095027923584, "learning_rate": 0.0002, "epoch": 1.38711502707946, "step": 8580}, {"loss": 0.7464, "grad_norm": 0.5198570489883423, "learning_rate": 0.0002, "epoch": 1.3887317112602053, "step": 8590}, {"loss": 0.7354, "grad_norm": 0.7261320352554321, "learning_rate": 0.0002, "epoch": 1.3903483954409506, "step": 8600}, {"loss": 0.7339, "grad_norm": 0.5616350173950195, "learning_rate": 0.0002, "epoch": 1.3919650796216958, "step": 8610}, {"loss": 0.7382, "grad_norm": 0.5185914635658264, "learning_rate": 0.0002, "epoch": 1.3935817638024413, "step": 8620}, {"loss": 0.7456, "grad_norm": 0.5814694762229919, "learning_rate": 0.0002, "epoch": 1.3951984479831865, "step": 8630}, {"loss": 0.7413, "grad_norm": 0.6977371573448181, "learning_rate": 0.0002, "epoch": 1.3968151321639317, "step": 8640}, {"loss": 0.7574, "grad_norm": 0.6855689883232117, "learning_rate": 0.0002, "epoch": 1.3984318163446772, "step": 8650}, {"loss": 0.7802, "grad_norm": 0.5414357781410217, "learning_rate": 0.0002, "epoch": 1.4000485005254224, "step": 8660}, {"loss": 0.7487, "grad_norm": 0.6970012784004211, "learning_rate": 0.0002, "epoch": 1.4016651847061676, "step": 8670}, {"loss": 0.7421, "grad_norm": 0.526079535484314, "learning_rate": 0.0002, "epoch": 1.4032818688869129, "step": 8680}, {"loss": 0.737, "grad_norm": 0.758712887763977, "learning_rate": 0.0002, "epoch": 1.404898553067658, "step": 8690}, {"loss": 0.7612, "grad_norm": 0.7118762731552124, "learning_rate": 0.0002, "epoch": 1.4065152372484035, "step": 8700}, {"loss": 0.7628, "grad_norm": 0.5696909427642822, "learning_rate": 0.0002, "epoch": 1.4081319214291488, "step": 8710}, {"loss": 0.7156, "grad_norm": 0.7995436787605286, "learning_rate": 0.0002, "epoch": 1.4097486056098942, "step": 8720}, {"loss": 0.7521, "grad_norm": 0.7237521409988403, "learning_rate": 0.0002, "epoch": 1.4113652897906395, "step": 8730}, {"loss": 0.7661, "grad_norm": 0.744628369808197, "learning_rate": 0.0002, "epoch": 1.4129819739713847, "step": 8740}, {"loss": 0.7073, "grad_norm": 0.6082926988601685, "learning_rate": 0.0002, "epoch": 1.41459865815213, "step": 8750}, {"loss": 0.7282, "grad_norm": 0.5185243487358093, "learning_rate": 0.0002, "epoch": 1.4162153423328752, "step": 8760}, {"loss": 0.7592, "grad_norm": 0.5183082222938538, "learning_rate": 0.0002, "epoch": 1.4178320265136206, "step": 8770}, {"loss": 0.7509, "grad_norm": 0.7326041460037231, "learning_rate": 0.0002, "epoch": 1.4194487106943658, "step": 8780}, {"loss": 0.7398, "grad_norm": 0.7174660563468933, "learning_rate": 0.0002, "epoch": 1.421065394875111, "step": 8790}, {"loss": 0.7507, "grad_norm": 0.8080165982246399, "learning_rate": 0.0002, "epoch": 1.4226820790558565, "step": 8800}, {"loss": 0.72, "grad_norm": 0.5061507821083069, "learning_rate": 0.0002, "epoch": 1.4242987632366018, "step": 8810}, {"loss": 0.7563, "grad_norm": 0.801602840423584, "learning_rate": 0.0002, "epoch": 1.425915447417347, "step": 8820}, {"loss": 0.7287, "grad_norm": 0.6150273084640503, "learning_rate": 0.0002, "epoch": 1.4275321315980922, "step": 8830}, {"loss": 0.7452, "grad_norm": 0.8786525726318359, "learning_rate": 0.0002, "epoch": 1.4291488157788377, "step": 8840}, {"loss": 0.7257, "grad_norm": 0.6371538639068604, "learning_rate": 0.0002, "epoch": 1.430765499959583, "step": 8850}, {"loss": 0.711, "grad_norm": 0.6409295797348022, "learning_rate": 0.0002, "epoch": 1.4323821841403281, "step": 8860}, {"loss": 0.7891, "grad_norm": 0.6452359557151794, "learning_rate": 0.0002, "epoch": 1.4339988683210736, "step": 8870}, {"loss": 0.7588, "grad_norm": 0.5842334628105164, "learning_rate": 0.0002, "epoch": 1.4356155525018188, "step": 8880}, {"loss": 0.7446, "grad_norm": 0.696761965751648, "learning_rate": 0.0002, "epoch": 1.437232236682564, "step": 8890}, {"loss": 0.7541, "grad_norm": 0.6384600400924683, "learning_rate": 0.0002, "epoch": 1.4388489208633093, "step": 8900}, {"loss": 0.7049, "grad_norm": 0.5981136560440063, "learning_rate": 0.0002, "epoch": 1.4404656050440545, "step": 8910}, {"loss": 0.795, "grad_norm": 0.6355637907981873, "learning_rate": 0.0002, "epoch": 1.4420822892248, "step": 8920}, {"loss": 0.7653, "grad_norm": 0.6374830603599548, "learning_rate": 0.0002, "epoch": 1.4436989734055452, "step": 8930}, {"loss": 0.8108, "grad_norm": 0.559013307094574, "learning_rate": 0.0002, "epoch": 1.4453156575862904, "step": 8940}, {"loss": 0.7045, "grad_norm": 0.7289170026779175, "learning_rate": 0.0002, "epoch": 1.446932341767036, "step": 8950}, {"loss": 0.7484, "grad_norm": 0.8649206757545471, "learning_rate": 0.0002, "epoch": 1.4485490259477811, "step": 8960}, {"loss": 0.7745, "grad_norm": 0.7664689421653748, "learning_rate": 0.0002, "epoch": 1.4501657101285264, "step": 8970}, {"loss": 0.7431, "grad_norm": 0.7109952569007874, "learning_rate": 0.0002, "epoch": 1.4517823943092716, "step": 8980}, {"loss": 0.7997, "grad_norm": 0.6312844753265381, "learning_rate": 0.0002, "epoch": 1.453399078490017, "step": 8990}, {"loss": 0.7467, "grad_norm": 0.6616617441177368, "learning_rate": 0.0002, "epoch": 1.4550157626707623, "step": 9000}, {"loss": 0.7518, "grad_norm": 0.7384068965911865, "learning_rate": 0.0002, "epoch": 1.4566324468515075, "step": 9010}, {"loss": 0.7483, "grad_norm": 0.6549670100212097, "learning_rate": 0.0002, "epoch": 1.458249131032253, "step": 9020}, {"loss": 0.7423, "grad_norm": 0.6254119277000427, "learning_rate": 0.0002, "epoch": 1.4598658152129982, "step": 9030}, {"loss": 0.7645, "grad_norm": 0.6806328892707825, "learning_rate": 0.0002, "epoch": 1.4614824993937434, "step": 9040}, {"loss": 0.7221, "grad_norm": 0.6803115010261536, "learning_rate": 0.0002, "epoch": 1.4630991835744886, "step": 9050}, {"loss": 0.7264, "grad_norm": 0.48529282212257385, "learning_rate": 0.0002, "epoch": 1.4647158677552339, "step": 9060}, {"loss": 0.7542, "grad_norm": 0.5995030999183655, "learning_rate": 0.0002, "epoch": 1.4663325519359793, "step": 9070}, {"loss": 0.7894, "grad_norm": 0.6005427837371826, "learning_rate": 0.0002, "epoch": 1.4679492361167246, "step": 9080}, {"loss": 0.7288, "grad_norm": 0.718564510345459, "learning_rate": 0.0002, "epoch": 1.46956592029747, "step": 9090}, {"loss": 0.7089, "grad_norm": 0.7003577351570129, "learning_rate": 0.0002, "epoch": 1.4711826044782153, "step": 9100}, {"loss": 0.8069, "grad_norm": 0.5888323783874512, "learning_rate": 0.0002, "epoch": 1.4727992886589605, "step": 9110}, {"loss": 0.7275, "grad_norm": 0.6417609453201294, "learning_rate": 0.0002, "epoch": 1.4744159728397057, "step": 9120}, {"loss": 0.7441, "grad_norm": 0.572294294834137, "learning_rate": 0.0002, "epoch": 1.476032657020451, "step": 9130}, {"loss": 0.8053, "grad_norm": 0.8200714588165283, "learning_rate": 0.0002, "epoch": 1.4776493412011964, "step": 9140}, {"loss": 0.7382, "grad_norm": 0.6343288421630859, "learning_rate": 0.0002, "epoch": 1.4792660253819416, "step": 9150}, {"loss": 0.7641, "grad_norm": 0.7017961144447327, "learning_rate": 0.0002, "epoch": 1.4808827095626869, "step": 9160}, {"loss": 0.7619, "grad_norm": 0.6202912926673889, "learning_rate": 0.0002, "epoch": 1.4824993937434323, "step": 9170}, {"loss": 0.7428, "grad_norm": 0.6677869558334351, "learning_rate": 0.0002, "epoch": 1.4841160779241775, "step": 9180}, {"loss": 0.7648, "grad_norm": 0.6052267551422119, "learning_rate": 0.0002, "epoch": 1.4857327621049228, "step": 9190}, {"loss": 0.7152, "grad_norm": 0.6638872027397156, "learning_rate": 0.0002, "epoch": 1.487349446285668, "step": 9200}, {"loss": 0.7448, "grad_norm": 0.6245523691177368, "learning_rate": 0.0002, "epoch": 1.4889661304664135, "step": 9210}, {"loss": 0.6958, "grad_norm": 0.5761767625808716, "learning_rate": 0.0002, "epoch": 1.4905828146471587, "step": 9220}, {"loss": 0.8012, "grad_norm": 0.8175981640815735, "learning_rate": 0.0002, "epoch": 1.492199498827904, "step": 9230}, {"loss": 0.683, "grad_norm": 0.9144009947776794, "learning_rate": 0.0002, "epoch": 1.4938161830086494, "step": 9240}, {"loss": 0.7623, "grad_norm": 0.5742552876472473, "learning_rate": 0.0002, "epoch": 1.4954328671893946, "step": 9250}, {"loss": 0.7418, "grad_norm": 0.534534215927124, "learning_rate": 0.0002, "epoch": 1.4970495513701398, "step": 9260}, {"loss": 0.7194, "grad_norm": 0.7836225032806396, "learning_rate": 0.0002, "epoch": 1.498666235550885, "step": 9270}, {"loss": 0.7453, "grad_norm": 0.5292993187904358, "learning_rate": 0.0002, "epoch": 1.5002829197316303, "step": 9280}, {"loss": 0.7168, "grad_norm": 0.8044071793556213, "learning_rate": 0.0002, "epoch": 1.5018996039123758, "step": 9290}, {"loss": 0.7229, "grad_norm": 0.6185805201530457, "learning_rate": 0.0002, "epoch": 1.503516288093121, "step": 9300}, {"loss": 0.684, "grad_norm": 0.6093607544898987, "learning_rate": 0.0002, "epoch": 1.5051329722738664, "step": 9310}, {"loss": 0.7973, "grad_norm": 0.5891730189323425, "learning_rate": 0.0002, "epoch": 1.5067496564546117, "step": 9320}, {"loss": 0.7474, "grad_norm": 0.6331129670143127, "learning_rate": 0.0002, "epoch": 1.508366340635357, "step": 9330}, {"loss": 0.7074, "grad_norm": 0.7690958380699158, "learning_rate": 0.0002, "epoch": 1.5099830248161021, "step": 9340}, {"loss": 0.672, "grad_norm": 0.6548877358436584, "learning_rate": 0.0002, "epoch": 1.5115997089968474, "step": 9350}, {"loss": 0.7408, "grad_norm": 0.6545143127441406, "learning_rate": 0.0002, "epoch": 1.5132163931775926, "step": 9360}, {"loss": 0.7432, "grad_norm": 0.553247332572937, "learning_rate": 0.0002, "epoch": 1.514833077358338, "step": 9370}, {"loss": 0.7265, "grad_norm": 0.8145074844360352, "learning_rate": 0.0002, "epoch": 1.5164497615390833, "step": 9380}, {"loss": 0.7379, "grad_norm": 0.7636994123458862, "learning_rate": 0.0002, "epoch": 1.5180664457198287, "step": 9390}, {"loss": 0.7413, "grad_norm": 0.6838982701301575, "learning_rate": 0.0002, "epoch": 1.519683129900574, "step": 9400}, {"loss": 0.7367, "grad_norm": 0.8599441647529602, "learning_rate": 0.0002, "epoch": 1.5212998140813192, "step": 9410}, {"loss": 0.7663, "grad_norm": 0.7020329833030701, "learning_rate": 0.0002, "epoch": 1.5229164982620644, "step": 9420}, {"loss": 0.7928, "grad_norm": 0.6964772343635559, "learning_rate": 0.0002, "epoch": 1.5245331824428097, "step": 9430}, {"loss": 0.7168, "grad_norm": 0.6916600465774536, "learning_rate": 0.0002, "epoch": 1.5261498666235551, "step": 9440}, {"loss": 0.7519, "grad_norm": 0.7282621264457703, "learning_rate": 0.0002, "epoch": 1.5277665508043003, "step": 9450}, {"loss": 0.7628, "grad_norm": 0.5363983511924744, "learning_rate": 0.0002, "epoch": 1.5293832349850458, "step": 9460}, {"loss": 0.7154, "grad_norm": 0.6184861063957214, "learning_rate": 0.0002, "epoch": 1.530999919165791, "step": 9470}, {"loss": 0.7837, "grad_norm": 0.5991285443305969, "learning_rate": 0.0002, "epoch": 1.5326166033465363, "step": 9480}, {"loss": 0.7827, "grad_norm": 0.8176587820053101, "learning_rate": 0.0002, "epoch": 1.5342332875272815, "step": 9490}, {"loss": 0.7415, "grad_norm": 0.6473721861839294, "learning_rate": 0.0002, "epoch": 1.5358499717080267, "step": 9500}, {"loss": 0.7632, "grad_norm": 0.7319952845573425, "learning_rate": 0.0002, "epoch": 1.5374666558887722, "step": 9510}, {"loss": 0.7706, "grad_norm": 0.702900230884552, "learning_rate": 0.0002, "epoch": 1.5390833400695174, "step": 9520}, {"loss": 0.7754, "grad_norm": 0.7971600294113159, "learning_rate": 0.0002, "epoch": 1.5407000242502629, "step": 9530}, {"loss": 0.7352, "grad_norm": 0.6527525186538696, "learning_rate": 0.0002, "epoch": 1.542316708431008, "step": 9540}, {"loss": 0.7425, "grad_norm": 0.5791676044464111, "learning_rate": 0.0002, "epoch": 1.5439333926117533, "step": 9550}, {"loss": 0.7585, "grad_norm": 0.5619390606880188, "learning_rate": 0.0002, "epoch": 1.5455500767924986, "step": 9560}, {"loss": 0.7894, "grad_norm": 0.5701689124107361, "learning_rate": 0.0002, "epoch": 1.5471667609732438, "step": 9570}, {"loss": 0.793, "grad_norm": 0.47549352049827576, "learning_rate": 0.0002, "epoch": 1.548783445153989, "step": 9580}, {"loss": 0.7276, "grad_norm": 0.8730611205101013, "learning_rate": 0.0002, "epoch": 1.5504001293347345, "step": 9590}, {"loss": 0.798, "grad_norm": 0.6842091083526611, "learning_rate": 0.0002, "epoch": 1.5520168135154797, "step": 9600}, {"loss": 0.7528, "grad_norm": 0.6675129532814026, "learning_rate": 0.0002, "epoch": 1.5536334976962252, "step": 9610}, {"loss": 0.7954, "grad_norm": 0.8173956274986267, "learning_rate": 0.0002, "epoch": 1.5552501818769704, "step": 9620}, {"loss": 0.7535, "grad_norm": 0.724947452545166, "learning_rate": 0.0002, "epoch": 1.5568668660577156, "step": 9630}, {"loss": 0.7738, "grad_norm": 0.6154758930206299, "learning_rate": 0.0002, "epoch": 1.5584835502384609, "step": 9640}, {"loss": 0.7568, "grad_norm": 0.6072008013725281, "learning_rate": 0.0002, "epoch": 1.560100234419206, "step": 9650}, {"loss": 0.7219, "grad_norm": 0.659010648727417, "learning_rate": 0.0002, "epoch": 1.5617169185999515, "step": 9660}, {"loss": 0.673, "grad_norm": 0.65857994556427, "learning_rate": 0.0002, "epoch": 1.5633336027806968, "step": 9670}, {"loss": 0.7156, "grad_norm": 0.5914267301559448, "learning_rate": 0.0002, "epoch": 1.5649502869614422, "step": 9680}, {"loss": 0.7414, "grad_norm": 0.6248020529747009, "learning_rate": 0.0002, "epoch": 1.5665669711421875, "step": 9690}, {"loss": 0.694, "grad_norm": 0.7147795557975769, "learning_rate": 0.0002, "epoch": 1.5681836553229327, "step": 9700}, {"loss": 0.7335, "grad_norm": 0.7076232433319092, "learning_rate": 0.0002, "epoch": 1.569800339503678, "step": 9710}, {"loss": 0.7413, "grad_norm": 0.6217400431632996, "learning_rate": 0.0002, "epoch": 1.5714170236844232, "step": 9720}, {"loss": 0.7296, "grad_norm": 0.6709911227226257, "learning_rate": 0.0002, "epoch": 1.5730337078651684, "step": 9730}, {"loss": 0.7306, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 1.5746503920459138, "step": 9740}, {"loss": 0.7242, "grad_norm": 0.6241145730018616, "learning_rate": 0.0002, "epoch": 1.576267076226659, "step": 9750}, {"loss": 0.7384, "grad_norm": 0.4960934817790985, "learning_rate": 0.0002, "epoch": 1.5778837604074045, "step": 9760}, {"loss": 0.725, "grad_norm": 0.6593309640884399, "learning_rate": 0.0002, "epoch": 1.5795004445881498, "step": 9770}, {"loss": 0.7531, "grad_norm": 0.5814042091369629, "learning_rate": 0.0002, "epoch": 1.581117128768895, "step": 9780}, {"loss": 0.7109, "grad_norm": 0.5936070680618286, "learning_rate": 0.0002, "epoch": 1.5827338129496402, "step": 9790}, {"loss": 0.7769, "grad_norm": 0.6454403400421143, "learning_rate": 0.0002, "epoch": 1.5843504971303854, "step": 9800}, {"loss": 0.7677, "grad_norm": 0.7612107992172241, "learning_rate": 0.0002, "epoch": 1.585967181311131, "step": 9810}, {"loss": 0.7649, "grad_norm": 0.6494482755661011, "learning_rate": 0.0002, "epoch": 1.5875838654918761, "step": 9820}, {"loss": 0.7569, "grad_norm": 0.7825694680213928, "learning_rate": 0.0002, "epoch": 1.5892005496726216, "step": 9830}, {"loss": 0.706, "grad_norm": 0.6757757663726807, "learning_rate": 0.0002, "epoch": 1.5908172338533668, "step": 9840}, {"loss": 0.7803, "grad_norm": 0.7105609178543091, "learning_rate": 0.0002, "epoch": 1.592433918034112, "step": 9850}, {"loss": 0.7925, "grad_norm": 0.7596991062164307, "learning_rate": 0.0002, "epoch": 1.5940506022148573, "step": 9860}, {"loss": 0.7108, "grad_norm": 0.5681525468826294, "learning_rate": 0.0002, "epoch": 1.5956672863956025, "step": 9870}, {"loss": 0.7811, "grad_norm": 0.6090980768203735, "learning_rate": 0.0002, "epoch": 1.5972839705763477, "step": 9880}, {"loss": 0.7339, "grad_norm": 0.6271613240242004, "learning_rate": 0.0002, "epoch": 1.5989006547570932, "step": 9890}, {"loss": 0.7419, "grad_norm": 0.7656369805335999, "learning_rate": 0.0002, "epoch": 1.6005173389378387, "step": 9900}, {"loss": 0.7336, "grad_norm": 0.7504446506500244, "learning_rate": 0.0002, "epoch": 1.6021340231185839, "step": 9910}, {"loss": 0.7479, "grad_norm": 0.659656286239624, "learning_rate": 0.0002, "epoch": 1.6037507072993291, "step": 9920}, {"loss": 0.7483, "grad_norm": 0.6006826162338257, "learning_rate": 0.0002, "epoch": 1.6053673914800743, "step": 9930}, {"loss": 0.732, "grad_norm": 0.7872757911682129, "learning_rate": 0.0002, "epoch": 1.6069840756608196, "step": 9940}, {"loss": 0.768, "grad_norm": 0.5545852780342102, "learning_rate": 0.0002, "epoch": 1.6086007598415648, "step": 9950}, {"loss": 0.8064, "grad_norm": 0.7429468631744385, "learning_rate": 0.0002, "epoch": 1.6102174440223103, "step": 9960}, {"loss": 0.714, "grad_norm": 0.6873556971549988, "learning_rate": 0.0002, "epoch": 1.6118341282030555, "step": 9970}, {"loss": 0.7324, "grad_norm": 0.5874287486076355, "learning_rate": 0.0002, "epoch": 1.613450812383801, "step": 9980}, {"loss": 0.7141, "grad_norm": 0.6039386987686157, "learning_rate": 0.0002, "epoch": 1.6150674965645462, "step": 9990}, {"loss": 0.6674, "grad_norm": 0.6233575940132141, "learning_rate": 0.0002, "epoch": 1.6166841807452914, "step": 10000}, {"loss": 0.7602, "grad_norm": 0.7676448225975037, "learning_rate": 0.0002, "epoch": 1.6183008649260366, "step": 10010}, {"loss": 0.7784, "grad_norm": 0.6565698385238647, "learning_rate": 0.0002, "epoch": 1.6199175491067819, "step": 10020}, {"loss": 0.7104, "grad_norm": 0.6787590384483337, "learning_rate": 0.0002, "epoch": 1.6215342332875273, "step": 10030}, {"loss": 0.7464, "grad_norm": 0.6137678027153015, "learning_rate": 0.0002, "epoch": 1.6231509174682726, "step": 10040}, {"loss": 0.7646, "grad_norm": 0.5236800312995911, "learning_rate": 0.0002, "epoch": 1.624767601649018, "step": 10050}, {"loss": 0.7437, "grad_norm": 0.7626367807388306, "learning_rate": 0.0002, "epoch": 1.6263842858297632, "step": 10060}, {"loss": 0.7273, "grad_norm": 0.5657260417938232, "learning_rate": 0.0002, "epoch": 1.6280009700105085, "step": 10070}, {"loss": 0.7354, "grad_norm": 0.4913991391658783, "learning_rate": 0.0002, "epoch": 1.6296176541912537, "step": 10080}, {"loss": 0.7596, "grad_norm": 0.7715556621551514, "learning_rate": 0.0002, "epoch": 1.631234338371999, "step": 10090}, {"loss": 0.7105, "grad_norm": 0.6509000062942505, "learning_rate": 0.0002, "epoch": 1.6328510225527442, "step": 10100}, {"loss": 0.7274, "grad_norm": 0.6215850114822388, "learning_rate": 0.0002, "epoch": 1.6344677067334896, "step": 10110}, {"loss": 0.7705, "grad_norm": 0.6956844329833984, "learning_rate": 0.0002, "epoch": 1.6360843909142349, "step": 10120}, {"loss": 0.7129, "grad_norm": 0.6111597418785095, "learning_rate": 0.0002, "epoch": 1.6377010750949803, "step": 10130}, {"loss": 0.6955, "grad_norm": 0.6518288850784302, "learning_rate": 0.0002, "epoch": 1.6393177592757255, "step": 10140}, {"loss": 0.731, "grad_norm": 0.6914522051811218, "learning_rate": 0.0002, "epoch": 1.6409344434564708, "step": 10150}, {"loss": 0.7295, "grad_norm": 0.63785719871521, "learning_rate": 0.0002, "epoch": 1.642551127637216, "step": 10160}, {"loss": 0.7355, "grad_norm": 0.6379287838935852, "learning_rate": 0.0002, "epoch": 1.6441678118179612, "step": 10170}, {"loss": 0.7359, "grad_norm": 0.6793403029441833, "learning_rate": 0.0002, "epoch": 1.6457844959987067, "step": 10180}, {"loss": 0.7402, "grad_norm": 0.6099132895469666, "learning_rate": 0.0002, "epoch": 1.647401180179452, "step": 10190}, {"loss": 0.7353, "grad_norm": 0.5869854092597961, "learning_rate": 0.0002, "epoch": 1.6490178643601974, "step": 10200}, {"loss": 0.8308, "grad_norm": 0.7716999053955078, "learning_rate": 0.0002, "epoch": 1.6506345485409426, "step": 10210}, {"loss": 0.7215, "grad_norm": 0.6854110360145569, "learning_rate": 0.0002, "epoch": 1.6522512327216878, "step": 10220}, {"loss": 0.782, "grad_norm": 0.6957170367240906, "learning_rate": 0.0002, "epoch": 1.653867916902433, "step": 10230}, {"loss": 0.7282, "grad_norm": 0.6932903528213501, "learning_rate": 0.0002, "epoch": 1.6554846010831783, "step": 10240}, {"loss": 0.7478, "grad_norm": 0.7713165283203125, "learning_rate": 0.0002, "epoch": 1.6571012852639235, "step": 10250}, {"loss": 0.7099, "grad_norm": 0.7455793619155884, "learning_rate": 0.0002, "epoch": 1.658717969444669, "step": 10260}, {"loss": 0.7524, "grad_norm": 0.5464168190956116, "learning_rate": 0.0002, "epoch": 1.6603346536254144, "step": 10270}, {"loss": 0.7328, "grad_norm": 0.6782926321029663, "learning_rate": 0.0002, "epoch": 1.6619513378061597, "step": 10280}, {"loss": 0.7801, "grad_norm": 0.7962649464607239, "learning_rate": 0.0002, "epoch": 1.663568021986905, "step": 10290}, {"loss": 0.7142, "grad_norm": 0.6814526319503784, "learning_rate": 0.0002, "epoch": 1.6651847061676501, "step": 10300}, {"loss": 0.7285, "grad_norm": 0.656895101070404, "learning_rate": 0.0002, "epoch": 1.6668013903483954, "step": 10310}, {"loss": 0.7358, "grad_norm": 0.6085672378540039, "learning_rate": 0.0002, "epoch": 1.6684180745291406, "step": 10320}, {"loss": 0.7074, "grad_norm": 0.585508406162262, "learning_rate": 0.0002, "epoch": 1.670034758709886, "step": 10330}, {"loss": 0.7604, "grad_norm": 0.6930184364318848, "learning_rate": 0.0002, "epoch": 1.6716514428906313, "step": 10340}, {"loss": 0.7169, "grad_norm": 0.575663149356842, "learning_rate": 0.0002, "epoch": 1.6732681270713767, "step": 10350}, {"loss": 0.7198, "grad_norm": 0.582502543926239, "learning_rate": 0.0002, "epoch": 1.674884811252122, "step": 10360}, {"loss": 0.7793, "grad_norm": 0.5668916702270508, "learning_rate": 0.0002, "epoch": 1.6765014954328672, "step": 10370}, {"loss": 0.7478, "grad_norm": 0.6070065498352051, "learning_rate": 0.0002, "epoch": 1.6781181796136124, "step": 10380}, {"loss": 0.7939, "grad_norm": 0.6141316294670105, "learning_rate": 0.0002, "epoch": 1.6797348637943577, "step": 10390}, {"loss": 0.7573, "grad_norm": 0.8359124064445496, "learning_rate": 0.0002, "epoch": 1.6813515479751031, "step": 10400}, {"loss": 0.7488, "grad_norm": 0.5378185510635376, "learning_rate": 0.0002, "epoch": 1.6829682321558483, "step": 10410}, {"loss": 0.7588, "grad_norm": 0.6959536075592041, "learning_rate": 0.0002, "epoch": 1.6845849163365938, "step": 10420}, {"loss": 0.7872, "grad_norm": 0.6514357328414917, "learning_rate": 0.0002, "epoch": 1.686201600517339, "step": 10430}, {"loss": 0.725, "grad_norm": 0.7706646919250488, "learning_rate": 0.0002, "epoch": 1.6878182846980843, "step": 10440}, {"loss": 0.7673, "grad_norm": 0.6183337569236755, "learning_rate": 0.0002, "epoch": 1.6894349688788295, "step": 10450}, {"loss": 0.7566, "grad_norm": 0.6123278141021729, "learning_rate": 0.0002, "epoch": 1.6910516530595747, "step": 10460}, {"loss": 0.7169, "grad_norm": 0.6894851326942444, "learning_rate": 0.0002, "epoch": 1.69266833724032, "step": 10470}, {"loss": 0.7435, "grad_norm": 0.7497312426567078, "learning_rate": 0.0002, "epoch": 1.6942850214210654, "step": 10480}, {"loss": 0.7544, "grad_norm": 0.5968214273452759, "learning_rate": 0.0002, "epoch": 1.6959017056018106, "step": 10490}, {"loss": 0.6793, "grad_norm": 0.6747927069664001, "learning_rate": 0.0002, "epoch": 1.697518389782556, "step": 10500}, {"loss": 0.7415, "grad_norm": 0.5708310008049011, "learning_rate": 0.0002, "epoch": 1.6991350739633013, "step": 10510}, {"loss": 0.7385, "grad_norm": 0.606526792049408, "learning_rate": 0.0002, "epoch": 1.7007517581440466, "step": 10520}, {"loss": 0.7204, "grad_norm": 0.662011981010437, "learning_rate": 0.0002, "epoch": 1.7023684423247918, "step": 10530}, {"loss": 0.7999, "grad_norm": 0.7583045363426208, "learning_rate": 0.0002, "epoch": 1.703985126505537, "step": 10540}, {"loss": 0.7563, "grad_norm": 0.721632182598114, "learning_rate": 0.0002, "epoch": 1.7056018106862825, "step": 10550}, {"loss": 0.7407, "grad_norm": 0.6107715368270874, "learning_rate": 0.0002, "epoch": 1.7072184948670277, "step": 10560}, {"loss": 0.7519, "grad_norm": 0.6652471423149109, "learning_rate": 0.0002, "epoch": 1.7088351790477732, "step": 10570}, {"loss": 0.7767, "grad_norm": 0.6308087110519409, "learning_rate": 0.0002, "epoch": 1.7104518632285184, "step": 10580}, {"loss": 0.7659, "grad_norm": 0.5464386940002441, "learning_rate": 0.0002, "epoch": 1.7120685474092636, "step": 10590}, {"loss": 0.7063, "grad_norm": 0.6558911204338074, "learning_rate": 0.0002, "epoch": 1.7136852315900089, "step": 10600}, {"loss": 0.7126, "grad_norm": 0.5665024518966675, "learning_rate": 0.0002, "epoch": 1.715301915770754, "step": 10610}, {"loss": 0.6958, "grad_norm": 0.7888094186782837, "learning_rate": 0.0002, "epoch": 1.7169185999514993, "step": 10620}, {"loss": 0.7785, "grad_norm": 0.7084909081459045, "learning_rate": 0.0002, "epoch": 1.7185352841322448, "step": 10630}, {"loss": 0.7557, "grad_norm": 0.7982324361801147, "learning_rate": 0.0002, "epoch": 1.7201519683129902, "step": 10640}, {"loss": 0.7345, "grad_norm": 0.6418732404708862, "learning_rate": 0.0002, "epoch": 1.7217686524937355, "step": 10650}, {"loss": 0.7734, "grad_norm": 0.7636681795120239, "learning_rate": 0.0002, "epoch": 1.7233853366744807, "step": 10660}, {"loss": 0.7541, "grad_norm": 0.5646875500679016, "learning_rate": 0.0002, "epoch": 1.725002020855226, "step": 10670}, {"loss": 0.7642, "grad_norm": 0.5231260657310486, "learning_rate": 0.0002, "epoch": 1.7266187050359711, "step": 10680}, {"loss": 0.7846, "grad_norm": 0.7635011672973633, "learning_rate": 0.0002, "epoch": 1.7282353892167164, "step": 10690}, {"loss": 0.7471, "grad_norm": 0.7518259286880493, "learning_rate": 0.0002, "epoch": 1.7298520733974618, "step": 10700}, {"loss": 0.751, "grad_norm": 0.7295602560043335, "learning_rate": 0.0002, "epoch": 1.731468757578207, "step": 10710}, {"loss": 0.731, "grad_norm": 0.6984632015228271, "learning_rate": 0.0002, "epoch": 1.7330854417589525, "step": 10720}, {"loss": 0.7921, "grad_norm": 0.6198219060897827, "learning_rate": 0.0002, "epoch": 1.7347021259396977, "step": 10730}, {"loss": 0.7642, "grad_norm": 0.6957576274871826, "learning_rate": 0.0002, "epoch": 1.736318810120443, "step": 10740}, {"loss": 0.7917, "grad_norm": 0.6430263519287109, "learning_rate": 0.0002, "epoch": 1.7379354943011882, "step": 10750}, {"loss": 0.7156, "grad_norm": 0.6134995222091675, "learning_rate": 0.0002, "epoch": 1.7395521784819334, "step": 10760}, {"loss": 0.7584, "grad_norm": 0.7209452986717224, "learning_rate": 0.0002, "epoch": 1.741168862662679, "step": 10770}, {"loss": 0.7528, "grad_norm": 0.6735447645187378, "learning_rate": 0.0002, "epoch": 1.7427855468434241, "step": 10780}, {"loss": 0.756, "grad_norm": 0.5605693459510803, "learning_rate": 0.0002, "epoch": 1.7444022310241696, "step": 10790}, {"loss": 0.7759, "grad_norm": 0.6882363557815552, "learning_rate": 0.0002, "epoch": 1.7460189152049148, "step": 10800}, {"loss": 0.7544, "grad_norm": 0.6386259198188782, "learning_rate": 0.0002, "epoch": 1.74763559938566, "step": 10810}, {"loss": 0.7697, "grad_norm": 0.6529015302658081, "learning_rate": 0.0002, "epoch": 1.7492522835664053, "step": 10820}, {"loss": 0.7219, "grad_norm": 0.5664082765579224, "learning_rate": 0.0002, "epoch": 1.7508689677471505, "step": 10830}, {"loss": 0.7586, "grad_norm": 0.7532684206962585, "learning_rate": 0.0002, "epoch": 1.7524856519278957, "step": 10840}, {"loss": 0.6919, "grad_norm": 0.77171391248703, "learning_rate": 0.0002, "epoch": 1.7541023361086412, "step": 10850}, {"loss": 0.785, "grad_norm": 0.7255431413650513, "learning_rate": 0.0002, "epoch": 1.7557190202893864, "step": 10860}, {"loss": 0.7458, "grad_norm": 0.763083279132843, "learning_rate": 0.0002, "epoch": 1.7573357044701319, "step": 10870}, {"loss": 0.7846, "grad_norm": 0.6042402982711792, "learning_rate": 0.0002, "epoch": 1.758952388650877, "step": 10880}, {"loss": 0.7027, "grad_norm": 0.7642518281936646, "learning_rate": 0.0002, "epoch": 1.7605690728316223, "step": 10890}, {"loss": 0.746, "grad_norm": 0.6347904801368713, "learning_rate": 0.0002, "epoch": 1.7621857570123676, "step": 10900}, {"loss": 0.7458, "grad_norm": 0.5371627807617188, "learning_rate": 0.0002, "epoch": 1.7638024411931128, "step": 10910}, {"loss": 0.7466, "grad_norm": 0.6840225458145142, "learning_rate": 0.0002, "epoch": 1.7654191253738583, "step": 10920}, {"loss": 0.725, "grad_norm": 0.5288469195365906, "learning_rate": 0.0002, "epoch": 1.7670358095546035, "step": 10930}, {"loss": 0.7863, "grad_norm": 0.69020676612854, "learning_rate": 0.0002, "epoch": 1.768652493735349, "step": 10940}, {"loss": 0.7468, "grad_norm": 0.5943242311477661, "learning_rate": 0.0002, "epoch": 1.7702691779160942, "step": 10950}, {"loss": 0.7244, "grad_norm": 0.5616418123245239, "learning_rate": 0.0002, "epoch": 1.7718858620968394, "step": 10960}, {"loss": 0.7137, "grad_norm": 0.7209470868110657, "learning_rate": 0.0002, "epoch": 1.7735025462775846, "step": 10970}, {"loss": 0.7459, "grad_norm": 0.6657957434654236, "learning_rate": 0.0002, "epoch": 1.7751192304583299, "step": 10980}, {"loss": 0.7076, "grad_norm": 0.6469064950942993, "learning_rate": 0.0002, "epoch": 1.776735914639075, "step": 10990}, {"loss": 0.7321, "grad_norm": 0.6615678071975708, "learning_rate": 0.0002, "epoch": 1.7783525988198206, "step": 11000}, {"loss": 0.747, "grad_norm": 0.6722439527511597, "learning_rate": 0.0002, "epoch": 1.779969283000566, "step": 11010}, {"loss": 0.7302, "grad_norm": 0.634136974811554, "learning_rate": 0.0002, "epoch": 1.7815859671813112, "step": 11020}, {"loss": 0.8105, "grad_norm": 0.6024377346038818, "learning_rate": 0.0002, "epoch": 1.7832026513620565, "step": 11030}, {"loss": 0.7855, "grad_norm": 0.6909403800964355, "learning_rate": 0.0002, "epoch": 1.7848193355428017, "step": 11040}, {"loss": 0.7471, "grad_norm": 0.7148767709732056, "learning_rate": 0.0002, "epoch": 1.786436019723547, "step": 11050}, {"loss": 0.7145, "grad_norm": 0.7442979216575623, "learning_rate": 0.0002, "epoch": 1.7880527039042922, "step": 11060}, {"loss": 0.7215, "grad_norm": 0.6830431818962097, "learning_rate": 0.0002, "epoch": 1.7896693880850376, "step": 11070}, {"loss": 0.7625, "grad_norm": 0.9172667264938354, "learning_rate": 0.0002, "epoch": 1.7912860722657828, "step": 11080}, {"loss": 0.76, "grad_norm": 0.6799490451812744, "learning_rate": 0.0002, "epoch": 1.7929027564465283, "step": 11090}, {"loss": 0.7716, "grad_norm": 0.7617024779319763, "learning_rate": 0.0002, "epoch": 1.7945194406272735, "step": 11100}, {"loss": 0.7586, "grad_norm": 0.7701810002326965, "learning_rate": 0.0002, "epoch": 1.7961361248080188, "step": 11110}, {"loss": 0.7843, "grad_norm": 0.7454385757446289, "learning_rate": 0.0002, "epoch": 1.797752808988764, "step": 11120}, {"loss": 0.7873, "grad_norm": 0.6121436953544617, "learning_rate": 0.0002, "epoch": 1.7993694931695092, "step": 11130}, {"loss": 0.7305, "grad_norm": 0.6237571835517883, "learning_rate": 0.0002, "epoch": 1.8009861773502547, "step": 11140}, {"loss": 0.6827, "grad_norm": 0.6818515658378601, "learning_rate": 0.0002, "epoch": 1.802602861531, "step": 11150}, {"loss": 0.6876, "grad_norm": 0.7768308520317078, "learning_rate": 0.0002, "epoch": 1.8042195457117454, "step": 11160}, {"loss": 0.7533, "grad_norm": 0.6875537633895874, "learning_rate": 0.0002, "epoch": 1.8058362298924906, "step": 11170}, {"loss": 0.761, "grad_norm": 0.7950584888458252, "learning_rate": 0.0002, "epoch": 1.8074529140732358, "step": 11180}, {"loss": 0.7623, "grad_norm": 0.8210248351097107, "learning_rate": 0.0002, "epoch": 1.809069598253981, "step": 11190}, {"loss": 0.7556, "grad_norm": 0.6674110889434814, "learning_rate": 0.0002, "epoch": 1.8106862824347263, "step": 11200}, {"loss": 0.7663, "grad_norm": 0.6261674761772156, "learning_rate": 0.0002, "epoch": 1.8123029666154715, "step": 11210}, {"loss": 0.7122, "grad_norm": 0.6484741568565369, "learning_rate": 0.0002, "epoch": 1.813919650796217, "step": 11220}, {"loss": 0.7718, "grad_norm": 0.6231244206428528, "learning_rate": 0.0002, "epoch": 1.8155363349769622, "step": 11230}, {"loss": 0.7152, "grad_norm": 0.7243146896362305, "learning_rate": 0.0002, "epoch": 1.8171530191577077, "step": 11240}, {"loss": 0.7448, "grad_norm": 0.6776193380355835, "learning_rate": 0.0002, "epoch": 1.818769703338453, "step": 11250}, {"loss": 0.7317, "grad_norm": 0.5973618030548096, "learning_rate": 0.0002, "epoch": 1.8203863875191981, "step": 11260}, {"loss": 0.7961, "grad_norm": 0.6451361179351807, "learning_rate": 0.0002, "epoch": 1.8220030716999434, "step": 11270}, {"loss": 0.7611, "grad_norm": 0.5963068008422852, "learning_rate": 0.0002, "epoch": 1.8236197558806886, "step": 11280}, {"loss": 0.7466, "grad_norm": 0.536902129650116, "learning_rate": 0.0002, "epoch": 1.825236440061434, "step": 11290}, {"loss": 0.708, "grad_norm": 0.6993787288665771, "learning_rate": 0.0002, "epoch": 1.8268531242421793, "step": 11300}, {"loss": 0.7153, "grad_norm": 0.6135255098342896, "learning_rate": 0.0002, "epoch": 1.8284698084229247, "step": 11310}, {"loss": 0.7423, "grad_norm": 0.6057423949241638, "learning_rate": 0.0002, "epoch": 1.83008649260367, "step": 11320}, {"loss": 0.735, "grad_norm": 0.6598812341690063, "learning_rate": 0.0002, "epoch": 1.8317031767844152, "step": 11330}, {"loss": 0.7278, "grad_norm": 0.6075948476791382, "learning_rate": 0.0002, "epoch": 1.8333198609651604, "step": 11340}, {"loss": 0.7846, "grad_norm": 0.7065447568893433, "learning_rate": 0.0002, "epoch": 1.8349365451459057, "step": 11350}, {"loss": 0.7365, "grad_norm": 0.680526614189148, "learning_rate": 0.0002, "epoch": 1.8365532293266509, "step": 11360}, {"loss": 0.7152, "grad_norm": 0.6356695294380188, "learning_rate": 0.0002, "epoch": 1.8381699135073963, "step": 11370}, {"loss": 0.721, "grad_norm": 0.6399052143096924, "learning_rate": 0.0002, "epoch": 1.8397865976881416, "step": 11380}, {"loss": 0.7618, "grad_norm": 0.6125704050064087, "learning_rate": 0.0002, "epoch": 1.841403281868887, "step": 11390}, {"loss": 0.755, "grad_norm": 0.7124643325805664, "learning_rate": 0.0002, "epoch": 1.8430199660496323, "step": 11400}, {"loss": 0.7972, "grad_norm": 0.6099604964256287, "learning_rate": 0.0002, "epoch": 1.8446366502303775, "step": 11410}, {"loss": 0.7187, "grad_norm": 0.7338208556175232, "learning_rate": 0.0002, "epoch": 1.8462533344111227, "step": 11420}, {"loss": 0.7007, "grad_norm": 0.7534668445587158, "learning_rate": 0.0002, "epoch": 1.847870018591868, "step": 11430}, {"loss": 0.7464, "grad_norm": 0.6135470271110535, "learning_rate": 0.0002, "epoch": 1.8494867027726134, "step": 11440}, {"loss": 0.7955, "grad_norm": 0.6229309439659119, "learning_rate": 0.0002, "epoch": 1.8511033869533586, "step": 11450}, {"loss": 0.7594, "grad_norm": 0.706423282623291, "learning_rate": 0.0002, "epoch": 1.852720071134104, "step": 11460}, {"loss": 0.7411, "grad_norm": 0.5460049510002136, "learning_rate": 0.0002, "epoch": 1.8543367553148493, "step": 11470}, {"loss": 0.7416, "grad_norm": 0.6616711020469666, "learning_rate": 0.0002, "epoch": 1.8559534394955945, "step": 11480}, {"loss": 0.729, "grad_norm": 0.6372783184051514, "learning_rate": 0.0002, "epoch": 1.8575701236763398, "step": 11490}, {"loss": 0.7333, "grad_norm": 0.7162668108940125, "learning_rate": 0.0002, "epoch": 1.859186807857085, "step": 11500}, {"loss": 0.7747, "grad_norm": 0.6605209708213806, "learning_rate": 0.0002, "epoch": 1.8608034920378305, "step": 11510}, {"loss": 0.7258, "grad_norm": 0.6933956742286682, "learning_rate": 0.0002, "epoch": 1.8624201762185757, "step": 11520}, {"loss": 0.7243, "grad_norm": 0.6582090854644775, "learning_rate": 0.0002, "epoch": 1.8640368603993211, "step": 11530}, {"loss": 0.7313, "grad_norm": 0.6416500806808472, "learning_rate": 0.0002, "epoch": 1.8656535445800664, "step": 11540}, {"loss": 0.7372, "grad_norm": 0.5434312224388123, "learning_rate": 0.0002, "epoch": 1.8672702287608116, "step": 11550}, {"loss": 0.7635, "grad_norm": 0.6827567219734192, "learning_rate": 0.0002, "epoch": 1.8688869129415568, "step": 11560}, {"loss": 0.7137, "grad_norm": 0.7354370951652527, "learning_rate": 0.0002, "epoch": 1.870503597122302, "step": 11570}, {"loss": 0.7526, "grad_norm": 0.590372622013092, "learning_rate": 0.0002, "epoch": 1.8721202813030473, "step": 11580}, {"loss": 0.731, "grad_norm": 0.853183925151825, "learning_rate": 0.0002, "epoch": 1.8737369654837928, "step": 11590}, {"loss": 0.7487, "grad_norm": 0.822678804397583, "learning_rate": 0.0002, "epoch": 1.875353649664538, "step": 11600}, {"loss": 0.7427, "grad_norm": 0.6591550707817078, "learning_rate": 0.0002, "epoch": 1.8769703338452834, "step": 11610}, {"loss": 0.7054, "grad_norm": 0.7475301623344421, "learning_rate": 0.0002, "epoch": 1.8785870180260287, "step": 11620}, {"loss": 0.811, "grad_norm": 0.6390765309333801, "learning_rate": 0.0002, "epoch": 1.880203702206774, "step": 11630}, {"loss": 0.7531, "grad_norm": 0.6589758992195129, "learning_rate": 0.0002, "epoch": 1.8818203863875191, "step": 11640}, {"loss": 0.7475, "grad_norm": 0.6765508651733398, "learning_rate": 0.0002, "epoch": 1.8834370705682644, "step": 11650}, {"loss": 0.738, "grad_norm": 0.6527857780456543, "learning_rate": 0.0002, "epoch": 1.8850537547490098, "step": 11660}, {"loss": 0.7504, "grad_norm": 0.6642923951148987, "learning_rate": 0.0002, "epoch": 1.886670438929755, "step": 11670}, {"loss": 0.7701, "grad_norm": 0.6945584416389465, "learning_rate": 0.0002, "epoch": 1.8882871231105005, "step": 11680}, {"loss": 0.7711, "grad_norm": 0.694018542766571, "learning_rate": 0.0002, "epoch": 1.8899038072912457, "step": 11690}, {"loss": 0.7195, "grad_norm": 0.7237417101860046, "learning_rate": 0.0002, "epoch": 1.891520491471991, "step": 11700}, {"loss": 0.7491, "grad_norm": 0.7401309609413147, "learning_rate": 0.0002, "epoch": 1.8931371756527362, "step": 11710}, {"loss": 0.805, "grad_norm": 0.6537784337997437, "learning_rate": 0.0002, "epoch": 1.8947538598334814, "step": 11720}, {"loss": 0.793, "grad_norm": 0.7398539185523987, "learning_rate": 0.0002, "epoch": 1.8963705440142267, "step": 11730}, {"loss": 0.7561, "grad_norm": 0.6696075797080994, "learning_rate": 0.0002, "epoch": 1.8979872281949721, "step": 11740}, {"loss": 0.7353, "grad_norm": 0.6014142036437988, "learning_rate": 0.0002, "epoch": 1.8996039123757174, "step": 11750}, {"loss": 0.7714, "grad_norm": 0.7023524641990662, "learning_rate": 0.0002, "epoch": 1.9012205965564628, "step": 11760}, {"loss": 0.7088, "grad_norm": 0.739973783493042, "learning_rate": 0.0002, "epoch": 1.902837280737208, "step": 11770}, {"loss": 0.7848, "grad_norm": 0.5576770901679993, "learning_rate": 0.0002, "epoch": 1.9044539649179533, "step": 11780}, {"loss": 0.7483, "grad_norm": 0.6907393932342529, "learning_rate": 0.0002, "epoch": 1.9060706490986985, "step": 11790}, {"loss": 0.7827, "grad_norm": 0.6934581995010376, "learning_rate": 0.0002, "epoch": 1.9076873332794437, "step": 11800}, {"loss": 0.7199, "grad_norm": 0.591774582862854, "learning_rate": 0.0002, "epoch": 1.9093040174601892, "step": 11810}, {"loss": 0.7333, "grad_norm": 0.6249791383743286, "learning_rate": 0.0002, "epoch": 1.9109207016409344, "step": 11820}, {"loss": 0.7581, "grad_norm": 0.6755744218826294, "learning_rate": 0.0002, "epoch": 1.9125373858216799, "step": 11830}, {"loss": 0.696, "grad_norm": 0.7286285161972046, "learning_rate": 0.0002, "epoch": 1.914154070002425, "step": 11840}, {"loss": 0.7509, "grad_norm": 0.7867850065231323, "learning_rate": 0.0002, "epoch": 1.9157707541831703, "step": 11850}, {"loss": 0.735, "grad_norm": 0.6283972859382629, "learning_rate": 0.0002, "epoch": 1.9173874383639156, "step": 11860}, {"loss": 0.7296, "grad_norm": 0.605823814868927, "learning_rate": 0.0002, "epoch": 1.9190041225446608, "step": 11870}, {"loss": 0.6598, "grad_norm": 0.5927976965904236, "learning_rate": 0.0002, "epoch": 1.920620806725406, "step": 11880}, {"loss": 0.7649, "grad_norm": 0.5974002480506897, "learning_rate": 0.0002, "epoch": 1.9222374909061515, "step": 11890}, {"loss": 0.7843, "grad_norm": 0.7091866135597229, "learning_rate": 0.0002, "epoch": 1.923854175086897, "step": 11900}, {"loss": 0.775, "grad_norm": 0.72496497631073, "learning_rate": 0.0002, "epoch": 1.9254708592676422, "step": 11910}, {"loss": 0.7153, "grad_norm": 0.6131896376609802, "learning_rate": 0.0002, "epoch": 1.9270875434483874, "step": 11920}, {"loss": 0.7228, "grad_norm": 0.6556436419487, "learning_rate": 0.0002, "epoch": 1.9287042276291326, "step": 11930}, {"loss": 0.7319, "grad_norm": 0.622932493686676, "learning_rate": 0.0002, "epoch": 1.9303209118098779, "step": 11940}, {"loss": 0.7592, "grad_norm": 0.6618631482124329, "learning_rate": 0.0002, "epoch": 1.931937595990623, "step": 11950}, {"loss": 0.8332, "grad_norm": 0.630966305732727, "learning_rate": 0.0002, "epoch": 1.9335542801713685, "step": 11960}, {"loss": 0.6854, "grad_norm": 0.6336734890937805, "learning_rate": 0.0002, "epoch": 1.9351709643521138, "step": 11970}, {"loss": 0.7433, "grad_norm": 0.655403196811676, "learning_rate": 0.0002, "epoch": 1.9367876485328592, "step": 11980}, {"loss": 0.7282, "grad_norm": 0.5640574097633362, "learning_rate": 0.0002, "epoch": 1.9384043327136045, "step": 11990}, {"loss": 0.7289, "grad_norm": 0.6322951316833496, "learning_rate": 0.0002, "epoch": 1.9400210168943497, "step": 12000}, {"loss": 0.7627, "grad_norm": 0.615703821182251, "learning_rate": 0.0002, "epoch": 1.941637701075095, "step": 12010}, {"loss": 0.786, "grad_norm": 0.6487536430358887, "learning_rate": 0.0002, "epoch": 1.9432543852558402, "step": 12020}, {"loss": 0.7435, "grad_norm": 0.9209630489349365, "learning_rate": 0.0002, "epoch": 1.9448710694365856, "step": 12030}, {"loss": 0.7274, "grad_norm": 0.67485511302948, "learning_rate": 0.0002, "epoch": 1.9464877536173308, "step": 12040}, {"loss": 0.7551, "grad_norm": 0.6831230521202087, "learning_rate": 0.0002, "epoch": 1.9481044377980763, "step": 12050}, {"loss": 0.7546, "grad_norm": 0.6578302383422852, "learning_rate": 0.0002, "epoch": 1.9497211219788215, "step": 12060}, {"loss": 0.6989, "grad_norm": 0.9975938200950623, "learning_rate": 0.0002, "epoch": 1.9513378061595668, "step": 12070}, {"loss": 0.7952, "grad_norm": 0.6637365221977234, "learning_rate": 0.0002, "epoch": 1.952954490340312, "step": 12080}, {"loss": 0.7482, "grad_norm": 0.605707049369812, "learning_rate": 0.0002, "epoch": 1.9545711745210572, "step": 12090}, {"loss": 0.7768, "grad_norm": 0.6584440469741821, "learning_rate": 0.0002, "epoch": 1.9561878587018025, "step": 12100}, {"loss": 0.7187, "grad_norm": 0.6070835590362549, "learning_rate": 0.0002, "epoch": 1.957804542882548, "step": 12110}, {"loss": 0.7491, "grad_norm": 0.7862601280212402, "learning_rate": 0.0002, "epoch": 1.9594212270632931, "step": 12120}, {"loss": 0.7972, "grad_norm": 0.8175255060195923, "learning_rate": 0.0002, "epoch": 1.9610379112440386, "step": 12130}, {"loss": 0.7242, "grad_norm": 0.5648472905158997, "learning_rate": 0.0002, "epoch": 1.9626545954247838, "step": 12140}, {"loss": 0.7321, "grad_norm": 0.6591973304748535, "learning_rate": 0.0002, "epoch": 1.964271279605529, "step": 12150}, {"loss": 0.739, "grad_norm": 0.5960676074028015, "learning_rate": 0.0002, "epoch": 1.9658879637862743, "step": 12160}, {"loss": 0.7254, "grad_norm": 0.7272544503211975, "learning_rate": 0.0002, "epoch": 1.9675046479670195, "step": 12170}, {"loss": 0.7376, "grad_norm": 0.7176699042320251, "learning_rate": 0.0002, "epoch": 1.969121332147765, "step": 12180}, {"loss": 0.7525, "grad_norm": 0.6927123665809631, "learning_rate": 0.0002, "epoch": 1.9707380163285102, "step": 12190}, {"loss": 0.7318, "grad_norm": 0.5536034107208252, "learning_rate": 0.0002, "epoch": 1.9723547005092557, "step": 12200}, {"loss": 0.7737, "grad_norm": 0.8348390460014343, "learning_rate": 0.0002, "epoch": 1.9739713846900009, "step": 12210}, {"loss": 0.7494, "grad_norm": 0.6591181755065918, "learning_rate": 0.0002, "epoch": 1.9755880688707461, "step": 12220}, {"loss": 0.763, "grad_norm": 1.0624109506607056, "learning_rate": 0.0002, "epoch": 1.9772047530514913, "step": 12230}, {"loss": 0.7541, "grad_norm": 0.9265586137771606, "learning_rate": 0.0002, "epoch": 1.9788214372322366, "step": 12240}, {"loss": 0.7533, "grad_norm": 0.5998196005821228, "learning_rate": 0.0002, "epoch": 1.9804381214129818, "step": 12250}, {"loss": 0.7225, "grad_norm": 0.6960851550102234, "learning_rate": 0.0002, "epoch": 1.9820548055937273, "step": 12260}, {"loss": 0.7398, "grad_norm": 0.7674502730369568, "learning_rate": 0.0002, "epoch": 1.9836714897744727, "step": 12270}, {"loss": 0.7185, "grad_norm": 0.6407275795936584, "learning_rate": 0.0002, "epoch": 1.985288173955218, "step": 12280}, {"loss": 0.7382, "grad_norm": 0.6673079133033752, "learning_rate": 0.0002, "epoch": 1.9869048581359632, "step": 12290}, {"loss": 0.7326, "grad_norm": 0.6989844441413879, "learning_rate": 0.0002, "epoch": 1.9885215423167084, "step": 12300}, {"loss": 0.7559, "grad_norm": 0.7564442157745361, "learning_rate": 0.0002, "epoch": 1.9901382264974536, "step": 12310}, {"loss": 0.7719, "grad_norm": 0.6385478973388672, "learning_rate": 0.0002, "epoch": 1.9917549106781989, "step": 12320}, {"loss": 0.7369, "grad_norm": 0.7193717956542969, "learning_rate": 0.0002, "epoch": 1.9933715948589443, "step": 12330}, {"loss": 0.7583, "grad_norm": 0.7987112402915955, "learning_rate": 0.0002, "epoch": 1.9949882790396896, "step": 12340}, {"loss": 0.7793, "grad_norm": 0.7260826826095581, "learning_rate": 0.0002, "epoch": 1.996604963220435, "step": 12350}, {"loss": 0.7505, "grad_norm": 0.7968255281448364, "learning_rate": 0.0002, "epoch": 1.9982216474011802, "step": 12360}, {"loss": 0.717, "grad_norm": 0.6893062591552734, "learning_rate": 0.0002, "epoch": 1.9998383315819255, "step": 12370}, {"eval_loss": 1.1044032573699951, "eval_runtime": 122.1508, "eval_samples_per_second": 6.001, "eval_steps_per_second": 0.753, "epoch": 2.0, "step": 12371}, {"loss": 0.6604, "grad_norm": 0.7775409817695618, "learning_rate": 0.0002, "epoch": 2.0014550157626707, "step": 12380}, {"loss": 0.6845, "grad_norm": 0.76218581199646, "learning_rate": 0.0002, "epoch": 2.003071699943416, "step": 12390}, {"loss": 0.6909, "grad_norm": 0.5677764415740967, "learning_rate": 0.0002, "epoch": 2.004688384124161, "step": 12400}, {"loss": 0.6584, "grad_norm": 0.808442234992981, "learning_rate": 0.0002, "epoch": 2.006305068304907, "step": 12410}, {"loss": 0.659, "grad_norm": 0.7144765257835388, "learning_rate": 0.0002, "epoch": 2.007921752485652, "step": 12420}, {"loss": 0.6666, "grad_norm": 0.6914031505584717, "learning_rate": 0.0002, "epoch": 2.0095384366663973, "step": 12430}, {"loss": 0.6596, "grad_norm": 0.7581454515457153, "learning_rate": 0.0002, "epoch": 2.0111551208471425, "step": 12440}, {"loss": 0.6785, "grad_norm": 0.8388504981994629, "learning_rate": 0.0002, "epoch": 2.0127718050278878, "step": 12450}, {"loss": 0.6942, "grad_norm": 0.6716406941413879, "learning_rate": 0.0002, "epoch": 2.014388489208633, "step": 12460}, {"loss": 0.6441, "grad_norm": 0.898902416229248, "learning_rate": 0.0002, "epoch": 2.0160051733893782, "step": 12470}, {"loss": 0.6655, "grad_norm": 0.6432679891586304, "learning_rate": 0.0002, "epoch": 2.0176218575701235, "step": 12480}, {"loss": 0.6521, "grad_norm": 0.8021109104156494, "learning_rate": 0.0002, "epoch": 2.019238541750869, "step": 12490}, {"loss": 0.6581, "grad_norm": 0.7039216756820679, "learning_rate": 0.0002, "epoch": 2.0208552259316144, "step": 12500}, {"loss": 0.6521, "grad_norm": 0.646531879901886, "learning_rate": 0.0002, "epoch": 2.0224719101123596, "step": 12510}, {"loss": 0.6302, "grad_norm": 0.783704400062561, "learning_rate": 0.0002, "epoch": 2.024088594293105, "step": 12520}, {"loss": 0.6288, "grad_norm": 0.8805046677589417, "learning_rate": 0.0002, "epoch": 2.02570527847385, "step": 12530}, {"loss": 0.6288, "grad_norm": 0.7289270758628845, "learning_rate": 0.0002, "epoch": 2.0273219626545953, "step": 12540}, {"loss": 0.6663, "grad_norm": 0.71653151512146, "learning_rate": 0.0002, "epoch": 2.0289386468353405, "step": 12550}, {"loss": 0.625, "grad_norm": 0.73281329870224, "learning_rate": 0.0002, "epoch": 2.030555331016086, "step": 12560}, {"loss": 0.6448, "grad_norm": 0.6657090187072754, "learning_rate": 0.0002, "epoch": 2.0321720151968314, "step": 12570}, {"loss": 0.6983, "grad_norm": 0.8241133093833923, "learning_rate": 0.0002, "epoch": 2.0337886993775767, "step": 12580}, {"loss": 0.6488, "grad_norm": 0.5834135413169861, "learning_rate": 0.0002, "epoch": 2.035405383558322, "step": 12590}, {"loss": 0.6188, "grad_norm": 0.84502112865448, "learning_rate": 0.0002, "epoch": 2.037022067739067, "step": 12600}, {"loss": 0.6349, "grad_norm": 0.8952481746673584, "learning_rate": 0.0002, "epoch": 2.0386387519198124, "step": 12610}, {"loss": 0.6923, "grad_norm": 0.7801461815834045, "learning_rate": 0.0002, "epoch": 2.0402554361005576, "step": 12620}, {"loss": 0.6176, "grad_norm": 0.6788367033004761, "learning_rate": 0.0002, "epoch": 2.041872120281303, "step": 12630}, {"loss": 0.6162, "grad_norm": 0.7241756319999695, "learning_rate": 0.0002, "epoch": 2.0434888044620485, "step": 12640}, {"loss": 0.655, "grad_norm": 0.6933388113975525, "learning_rate": 0.0002, "epoch": 2.0451054886427937, "step": 12650}, {"loss": 0.6431, "grad_norm": 0.8029746413230896, "learning_rate": 0.0002, "epoch": 2.046722172823539, "step": 12660}, {"loss": 0.7164, "grad_norm": 0.946399986743927, "learning_rate": 0.0002, "epoch": 2.048338857004284, "step": 12670}, {"loss": 0.638, "grad_norm": 0.7072678804397583, "learning_rate": 0.0002, "epoch": 2.0499555411850294, "step": 12680}, {"loss": 0.6487, "grad_norm": 0.6810618042945862, "learning_rate": 0.0002, "epoch": 2.0515722253657747, "step": 12690}, {"loss": 0.6554, "grad_norm": 0.7661160230636597, "learning_rate": 0.0002, "epoch": 2.05318890954652, "step": 12700}, {"loss": 0.6799, "grad_norm": 0.6350653767585754, "learning_rate": 0.0002, "epoch": 2.0548055937272656, "step": 12710}, {"loss": 0.6654, "grad_norm": 0.861890971660614, "learning_rate": 0.0002, "epoch": 2.056422277908011, "step": 12720}, {"loss": 0.6286, "grad_norm": 0.6489875912666321, "learning_rate": 0.0002, "epoch": 2.058038962088756, "step": 12730}, {"loss": 0.6811, "grad_norm": 0.8268506526947021, "learning_rate": 0.0002, "epoch": 2.0596556462695013, "step": 12740}, {"loss": 0.6524, "grad_norm": 0.607679545879364, "learning_rate": 0.0002, "epoch": 2.0612723304502465, "step": 12750}, {"loss": 0.6649, "grad_norm": 0.6754153370857239, "learning_rate": 0.0002, "epoch": 2.0628890146309917, "step": 12760}, {"loss": 0.6549, "grad_norm": 0.7263124585151672, "learning_rate": 0.0002, "epoch": 2.064505698811737, "step": 12770}, {"loss": 0.6189, "grad_norm": 0.6986154317855835, "learning_rate": 0.0002, "epoch": 2.0661223829924826, "step": 12780}, {"loss": 0.6723, "grad_norm": 0.7768576741218567, "learning_rate": 0.0002, "epoch": 2.067739067173228, "step": 12790}, {"loss": 0.677, "grad_norm": 0.7546762824058533, "learning_rate": 0.0002, "epoch": 2.069355751353973, "step": 12800}, {"loss": 0.6485, "grad_norm": 0.7588880062103271, "learning_rate": 0.0002, "epoch": 2.0709724355347183, "step": 12810}, {"loss": 0.6989, "grad_norm": 0.7457242608070374, "learning_rate": 0.0002, "epoch": 2.0725891197154636, "step": 12820}, {"loss": 0.6489, "grad_norm": 0.6983516812324524, "learning_rate": 0.0002, "epoch": 2.074205803896209, "step": 12830}, {"loss": 0.651, "grad_norm": 0.7950928807258606, "learning_rate": 0.0002, "epoch": 2.075822488076954, "step": 12840}, {"loss": 0.6603, "grad_norm": 0.9248087406158447, "learning_rate": 0.0002, "epoch": 2.0774391722576993, "step": 12850}, {"loss": 0.6847, "grad_norm": 0.7229493260383606, "learning_rate": 0.0002, "epoch": 2.079055856438445, "step": 12860}, {"loss": 0.6702, "grad_norm": 0.5710847973823547, "learning_rate": 0.0002, "epoch": 2.08067254061919, "step": 12870}, {"loss": 0.6974, "grad_norm": 0.9580423831939697, "learning_rate": 0.0002, "epoch": 2.0822892247999354, "step": 12880}, {"loss": 0.6341, "grad_norm": 0.7399665713310242, "learning_rate": 0.0002, "epoch": 2.0839059089806806, "step": 12890}, {"loss": 0.6993, "grad_norm": 0.7981410622596741, "learning_rate": 0.0002, "epoch": 2.085522593161426, "step": 12900}, {"loss": 0.6976, "grad_norm": 0.870759904384613, "learning_rate": 0.0002, "epoch": 2.087139277342171, "step": 12910}, {"loss": 0.7194, "grad_norm": 0.7001481652259827, "learning_rate": 0.0002, "epoch": 2.0887559615229163, "step": 12920}, {"loss": 0.6383, "grad_norm": 0.6745418310165405, "learning_rate": 0.0002, "epoch": 2.090372645703662, "step": 12930}, {"loss": 0.6519, "grad_norm": 0.7739067673683167, "learning_rate": 0.0002, "epoch": 2.0919893298844072, "step": 12940}, {"loss": 0.6856, "grad_norm": 0.6742934584617615, "learning_rate": 0.0002, "epoch": 2.0936060140651525, "step": 12950}, {"loss": 0.6279, "grad_norm": 0.7270349860191345, "learning_rate": 0.0002, "epoch": 2.0952226982458977, "step": 12960}, {"loss": 0.6783, "grad_norm": 0.7150624394416809, "learning_rate": 0.0002, "epoch": 2.096839382426643, "step": 12970}, {"loss": 0.6093, "grad_norm": 0.7734767198562622, "learning_rate": 0.0002, "epoch": 2.098456066607388, "step": 12980}, {"loss": 0.6534, "grad_norm": 0.7618662118911743, "learning_rate": 0.0002, "epoch": 2.1000727507881334, "step": 12990}, {"loss": 0.6707, "grad_norm": 0.6557944416999817, "learning_rate": 0.0002, "epoch": 2.101689434968879, "step": 13000}, {"loss": 0.7268, "grad_norm": 0.8786448240280151, "learning_rate": 0.0002, "epoch": 2.1033061191496243, "step": 13010}, {"loss": 0.6677, "grad_norm": 0.6878724098205566, "learning_rate": 0.0002, "epoch": 2.1049228033303695, "step": 13020}, {"loss": 0.6824, "grad_norm": 0.822318971157074, "learning_rate": 0.0002, "epoch": 2.1065394875111147, "step": 13030}, {"loss": 0.6228, "grad_norm": 0.831468939781189, "learning_rate": 0.0002, "epoch": 2.10815617169186, "step": 13040}, {"loss": 0.6511, "grad_norm": 0.7699505686759949, "learning_rate": 0.0002, "epoch": 2.109772855872605, "step": 13050}, {"loss": 0.6671, "grad_norm": 0.7559016346931458, "learning_rate": 0.0002, "epoch": 2.1113895400533504, "step": 13060}, {"loss": 0.6215, "grad_norm": 0.6942209601402283, "learning_rate": 0.0002, "epoch": 2.1130062242340957, "step": 13070}, {"loss": 0.6449, "grad_norm": 0.6098947525024414, "learning_rate": 0.0002, "epoch": 2.1146229084148414, "step": 13080}, {"loss": 0.7091, "grad_norm": 0.6499016284942627, "learning_rate": 0.0002, "epoch": 2.1162395925955866, "step": 13090}, {"loss": 0.6247, "grad_norm": 0.7719953060150146, "learning_rate": 0.0002, "epoch": 2.117856276776332, "step": 13100}, {"loss": 0.6064, "grad_norm": 0.6708134412765503, "learning_rate": 0.0002, "epoch": 2.119472960957077, "step": 13110}, {"loss": 0.6056, "grad_norm": 0.8119585514068604, "learning_rate": 0.0002, "epoch": 2.1210896451378223, "step": 13120}, {"loss": 0.6628, "grad_norm": 0.6947157979011536, "learning_rate": 0.0002, "epoch": 2.1227063293185675, "step": 13130}, {"loss": 0.6375, "grad_norm": 0.8831837773323059, "learning_rate": 0.0002, "epoch": 2.1243230134993127, "step": 13140}, {"loss": 0.6997, "grad_norm": 0.7266910672187805, "learning_rate": 0.0002, "epoch": 2.1259396976800584, "step": 13150}, {"loss": 0.6446, "grad_norm": 0.8864351511001587, "learning_rate": 0.0002, "epoch": 2.1275563818608036, "step": 13160}, {"loss": 0.6762, "grad_norm": 0.8104248046875, "learning_rate": 0.0002, "epoch": 2.129173066041549, "step": 13170}, {"loss": 0.6581, "grad_norm": 0.6077079772949219, "learning_rate": 0.0002, "epoch": 2.130789750222294, "step": 13180}, {"loss": 0.6572, "grad_norm": 0.6874213814735413, "learning_rate": 0.0002, "epoch": 2.1324064344030393, "step": 13190}, {"loss": 0.642, "grad_norm": 0.7134367823600769, "learning_rate": 0.0002, "epoch": 2.1340231185837846, "step": 13200}, {"loss": 0.7016, "grad_norm": 0.6101235151290894, "learning_rate": 0.0002, "epoch": 2.13563980276453, "step": 13210}, {"loss": 0.6529, "grad_norm": 0.6042411923408508, "learning_rate": 0.0002, "epoch": 2.137256486945275, "step": 13220}, {"loss": 0.7179, "grad_norm": 0.914601743221283, "learning_rate": 0.0002, "epoch": 2.1388731711260207, "step": 13230}, {"loss": 0.6513, "grad_norm": 0.7104284167289734, "learning_rate": 0.0002, "epoch": 2.140489855306766, "step": 13240}, {"loss": 0.6607, "grad_norm": 0.664395272731781, "learning_rate": 0.0002, "epoch": 2.142106539487511, "step": 13250}, {"loss": 0.7211, "grad_norm": 0.6991241574287415, "learning_rate": 0.0002, "epoch": 2.1437232236682564, "step": 13260}, {"loss": 0.6484, "grad_norm": 0.5469560623168945, "learning_rate": 0.0002, "epoch": 2.1453399078490016, "step": 13270}, {"loss": 0.6765, "grad_norm": 0.8454998135566711, "learning_rate": 0.0002, "epoch": 2.146956592029747, "step": 13280}, {"loss": 0.6683, "grad_norm": 0.7088868618011475, "learning_rate": 0.0002, "epoch": 2.148573276210492, "step": 13290}, {"loss": 0.6835, "grad_norm": 0.7002687454223633, "learning_rate": 0.0002, "epoch": 2.1501899603912378, "step": 13300}, {"loss": 0.6399, "grad_norm": 0.7785214781761169, "learning_rate": 0.0002, "epoch": 2.151806644571983, "step": 13310}, {"loss": 0.67, "grad_norm": 0.8049132227897644, "learning_rate": 0.0002, "epoch": 2.1534233287527282, "step": 13320}, {"loss": 0.6495, "grad_norm": 0.8062595129013062, "learning_rate": 0.0002, "epoch": 2.1550400129334735, "step": 13330}, {"loss": 0.6603, "grad_norm": 0.6208319067955017, "learning_rate": 0.0002, "epoch": 2.1566566971142187, "step": 13340}, {"loss": 0.6584, "grad_norm": 0.7519655823707581, "learning_rate": 0.0002, "epoch": 2.158273381294964, "step": 13350}, {"loss": 0.6457, "grad_norm": 0.7645747065544128, "learning_rate": 0.0002, "epoch": 2.159890065475709, "step": 13360}, {"loss": 0.645, "grad_norm": 0.6847302913665771, "learning_rate": 0.0002, "epoch": 2.1615067496564544, "step": 13370}, {"loss": 0.6903, "grad_norm": 0.8630441427230835, "learning_rate": 0.0002, "epoch": 2.1631234338372, "step": 13380}, {"loss": 0.6742, "grad_norm": 0.7947702407836914, "learning_rate": 0.0002, "epoch": 2.1647401180179453, "step": 13390}, {"loss": 0.7206, "grad_norm": 0.6836977005004883, "learning_rate": 0.0002, "epoch": 2.1663568021986905, "step": 13400}, {"loss": 0.6304, "grad_norm": 0.7340566515922546, "learning_rate": 0.0002, "epoch": 2.1679734863794358, "step": 13410}, {"loss": 0.6528, "grad_norm": 0.7075738906860352, "learning_rate": 0.0002, "epoch": 2.169590170560181, "step": 13420}, {"loss": 0.6585, "grad_norm": 0.7080879807472229, "learning_rate": 0.0002, "epoch": 2.1712068547409262, "step": 13430}, {"loss": 0.6615, "grad_norm": 0.6218613386154175, "learning_rate": 0.0002, "epoch": 2.1728235389216715, "step": 13440}, {"loss": 0.6488, "grad_norm": 0.8211479187011719, "learning_rate": 0.0002, "epoch": 2.174440223102417, "step": 13450}, {"loss": 0.6738, "grad_norm": 0.864466667175293, "learning_rate": 0.0002, "epoch": 2.1760569072831624, "step": 13460}, {"loss": 0.679, "grad_norm": 0.7943857908248901, "learning_rate": 0.0002, "epoch": 2.1776735914639076, "step": 13470}, {"loss": 0.6838, "grad_norm": 0.78728187084198, "learning_rate": 0.0002, "epoch": 2.179290275644653, "step": 13480}, {"loss": 0.6397, "grad_norm": 0.697527289390564, "learning_rate": 0.0002, "epoch": 2.180906959825398, "step": 13490}, {"loss": 0.669, "grad_norm": 0.8205804228782654, "learning_rate": 0.0002, "epoch": 2.1825236440061433, "step": 13500}, {"loss": 0.7227, "grad_norm": 0.8709042072296143, "learning_rate": 0.0002, "epoch": 2.1841403281868885, "step": 13510}, {"loss": 0.6313, "grad_norm": 0.6228537559509277, "learning_rate": 0.0002, "epoch": 2.1857570123676338, "step": 13520}, {"loss": 0.7025, "grad_norm": 0.9566980004310608, "learning_rate": 0.0002, "epoch": 2.1873736965483794, "step": 13530}, {"loss": 0.6755, "grad_norm": 0.7128894329071045, "learning_rate": 0.0002, "epoch": 2.1889903807291247, "step": 13540}, {"loss": 0.6827, "grad_norm": 0.6888654232025146, "learning_rate": 0.0002, "epoch": 2.19060706490987, "step": 13550}, {"loss": 0.6961, "grad_norm": 0.6444337368011475, "learning_rate": 0.0002, "epoch": 2.192223749090615, "step": 13560}, {"loss": 0.656, "grad_norm": 0.8008806705474854, "learning_rate": 0.0002, "epoch": 2.1938404332713604, "step": 13570}, {"loss": 0.7, "grad_norm": 0.8482748866081238, "learning_rate": 0.0002, "epoch": 2.1954571174521056, "step": 13580}, {"loss": 0.7326, "grad_norm": 0.8584157228469849, "learning_rate": 0.0002, "epoch": 2.197073801632851, "step": 13590}, {"loss": 0.7014, "grad_norm": 0.7513734698295593, "learning_rate": 0.0002, "epoch": 2.1986904858135965, "step": 13600}, {"loss": 0.6632, "grad_norm": 0.7864262461662292, "learning_rate": 0.0002, "epoch": 2.2003071699943417, "step": 13610}, {"loss": 0.6879, "grad_norm": 0.8493645191192627, "learning_rate": 0.0002, "epoch": 2.201923854175087, "step": 13620}, {"loss": 0.6617, "grad_norm": 0.6902140974998474, "learning_rate": 0.0002, "epoch": 2.203540538355832, "step": 13630}, {"loss": 0.6655, "grad_norm": 0.8711254596710205, "learning_rate": 0.0002, "epoch": 2.2051572225365774, "step": 13640}, {"loss": 0.6359, "grad_norm": 0.7832191586494446, "learning_rate": 0.0002, "epoch": 2.2067739067173227, "step": 13650}, {"loss": 0.6723, "grad_norm": 0.5668176412582397, "learning_rate": 0.0002, "epoch": 2.208390590898068, "step": 13660}, {"loss": 0.635, "grad_norm": 0.8648375272750854, "learning_rate": 0.0002, "epoch": 2.2100072750788136, "step": 13670}, {"loss": 0.653, "grad_norm": 0.7643089890480042, "learning_rate": 0.0002, "epoch": 2.211623959259559, "step": 13680}, {"loss": 0.6765, "grad_norm": 0.6293777823448181, "learning_rate": 0.0002, "epoch": 2.213240643440304, "step": 13690}, {"loss": 0.6842, "grad_norm": 0.6459372639656067, "learning_rate": 0.0002, "epoch": 2.2148573276210493, "step": 13700}, {"loss": 0.6526, "grad_norm": 0.7060744166374207, "learning_rate": 0.0002, "epoch": 2.2164740118017945, "step": 13710}, {"loss": 0.7101, "grad_norm": 0.674109160900116, "learning_rate": 0.0002, "epoch": 2.2180906959825397, "step": 13720}, {"loss": 0.6529, "grad_norm": 0.830392062664032, "learning_rate": 0.0002, "epoch": 2.219707380163285, "step": 13730}, {"loss": 0.6733, "grad_norm": 0.6474477052688599, "learning_rate": 0.0002, "epoch": 2.2213240643440306, "step": 13740}, {"loss": 0.6413, "grad_norm": 0.7037909626960754, "learning_rate": 0.0002, "epoch": 2.222940748524776, "step": 13750}, {"loss": 0.6417, "grad_norm": 0.6554131507873535, "learning_rate": 0.0002, "epoch": 2.224557432705521, "step": 13760}, {"loss": 0.6907, "grad_norm": 0.7822230458259583, "learning_rate": 0.0002, "epoch": 2.2261741168862663, "step": 13770}, {"loss": 0.6505, "grad_norm": 0.9082167744636536, "learning_rate": 0.0002, "epoch": 2.2277908010670116, "step": 13780}, {"loss": 0.6878, "grad_norm": 0.7918276190757751, "learning_rate": 0.0002, "epoch": 2.229407485247757, "step": 13790}, {"loss": 0.6669, "grad_norm": 0.7354569435119629, "learning_rate": 0.0002, "epoch": 2.231024169428502, "step": 13800}, {"loss": 0.6503, "grad_norm": 0.8265249133110046, "learning_rate": 0.0002, "epoch": 2.2326408536092472, "step": 13810}, {"loss": 0.6871, "grad_norm": 0.6653847098350525, "learning_rate": 0.0002, "epoch": 2.234257537789993, "step": 13820}, {"loss": 0.6413, "grad_norm": 0.7157923579216003, "learning_rate": 0.0002, "epoch": 2.235874221970738, "step": 13830}, {"loss": 0.6306, "grad_norm": 0.7110323309898376, "learning_rate": 0.0002, "epoch": 2.2374909061514834, "step": 13840}, {"loss": 0.6913, "grad_norm": 0.7155357599258423, "learning_rate": 0.0002, "epoch": 2.2391075903322286, "step": 13850}, {"loss": 0.6579, "grad_norm": 1.0177817344665527, "learning_rate": 0.0002, "epoch": 2.240724274512974, "step": 13860}, {"loss": 0.635, "grad_norm": 0.7601948380470276, "learning_rate": 0.0002, "epoch": 2.242340958693719, "step": 13870}, {"loss": 0.6679, "grad_norm": 0.7628820538520813, "learning_rate": 0.0002, "epoch": 2.2439576428744643, "step": 13880}, {"loss": 0.6805, "grad_norm": 0.7089297771453857, "learning_rate": 0.0002, "epoch": 2.24557432705521, "step": 13890}, {"loss": 0.7236, "grad_norm": 0.695178210735321, "learning_rate": 0.0002, "epoch": 2.247191011235955, "step": 13900}, {"loss": 0.7084, "grad_norm": 0.7631948590278625, "learning_rate": 0.0002, "epoch": 2.2488076954167004, "step": 13910}, {"loss": 0.685, "grad_norm": 0.8203101754188538, "learning_rate": 0.0002, "epoch": 2.2504243795974457, "step": 13920}, {"loss": 0.653, "grad_norm": 0.8099079728126526, "learning_rate": 0.0002, "epoch": 2.252041063778191, "step": 13930}, {"loss": 0.694, "grad_norm": 0.6498546004295349, "learning_rate": 0.0002, "epoch": 2.253657747958936, "step": 13940}, {"loss": 0.6684, "grad_norm": 0.7797415256500244, "learning_rate": 0.0002, "epoch": 2.2552744321396814, "step": 13950}, {"loss": 0.683, "grad_norm": 0.8254124522209167, "learning_rate": 0.0002, "epoch": 2.2568911163204266, "step": 13960}, {"loss": 0.6806, "grad_norm": 0.6327953338623047, "learning_rate": 0.0002, "epoch": 2.2585078005011723, "step": 13970}, {"loss": 0.668, "grad_norm": 0.734194278717041, "learning_rate": 0.0002, "epoch": 2.2601244846819175, "step": 13980}, {"loss": 0.6912, "grad_norm": 0.9014202952384949, "learning_rate": 0.0002, "epoch": 2.2617411688626627, "step": 13990}, {"loss": 0.692, "grad_norm": 0.7643631100654602, "learning_rate": 0.0002, "epoch": 2.263357853043408, "step": 14000}, {"loss": 0.6657, "grad_norm": 0.8882834911346436, "learning_rate": 0.0002, "epoch": 2.264974537224153, "step": 14010}, {"loss": 0.6453, "grad_norm": 0.7975873351097107, "learning_rate": 0.0002, "epoch": 2.2665912214048984, "step": 14020}, {"loss": 0.7193, "grad_norm": 0.7765783071517944, "learning_rate": 0.0002, "epoch": 2.2682079055856437, "step": 14030}, {"loss": 0.662, "grad_norm": 0.8846288323402405, "learning_rate": 0.0002, "epoch": 2.2698245897663893, "step": 14040}, {"loss": 0.6494, "grad_norm": 0.9006744027137756, "learning_rate": 0.0002, "epoch": 2.2714412739471346, "step": 14050}, {"loss": 0.6423, "grad_norm": 0.7420173287391663, "learning_rate": 0.0002, "epoch": 2.27305795812788, "step": 14060}, {"loss": 0.7068, "grad_norm": 0.7956424951553345, "learning_rate": 0.0002, "epoch": 2.274674642308625, "step": 14070}, {"loss": 0.6581, "grad_norm": 0.7783209085464478, "learning_rate": 0.0002, "epoch": 2.2762913264893703, "step": 14080}, {"loss": 0.7202, "grad_norm": 0.7597188949584961, "learning_rate": 0.0002, "epoch": 2.2779080106701155, "step": 14090}, {"loss": 0.6778, "grad_norm": 0.6718921661376953, "learning_rate": 0.0002, "epoch": 2.2795246948508607, "step": 14100}, {"loss": 0.632, "grad_norm": 0.7528082132339478, "learning_rate": 0.0002, "epoch": 2.281141379031606, "step": 14110}, {"loss": 0.7608, "grad_norm": 0.8379864692687988, "learning_rate": 0.0002, "epoch": 2.2827580632123516, "step": 14120}, {"loss": 0.6767, "grad_norm": 0.748613715171814, "learning_rate": 0.0002, "epoch": 2.284374747393097, "step": 14130}, {"loss": 0.6641, "grad_norm": 0.7435423135757446, "learning_rate": 0.0002, "epoch": 2.285991431573842, "step": 14140}, {"loss": 0.6849, "grad_norm": 0.7580803632736206, "learning_rate": 0.0002, "epoch": 2.2876081157545873, "step": 14150}, {"loss": 0.6604, "grad_norm": 0.6278321146965027, "learning_rate": 0.0002, "epoch": 2.2892247999353326, "step": 14160}, {"loss": 0.6573, "grad_norm": 0.7663896083831787, "learning_rate": 0.0002, "epoch": 2.290841484116078, "step": 14170}, {"loss": 0.6655, "grad_norm": 0.9716812372207642, "learning_rate": 0.0002, "epoch": 2.292458168296823, "step": 14180}, {"loss": 0.7067, "grad_norm": 0.8993458151817322, "learning_rate": 0.0002, "epoch": 2.2940748524775687, "step": 14190}, {"loss": 0.6172, "grad_norm": 0.6156117916107178, "learning_rate": 0.0002, "epoch": 2.295691536658314, "step": 14200}, {"loss": 0.6318, "grad_norm": 0.8911278247833252, "learning_rate": 0.0002, "epoch": 2.297308220839059, "step": 14210}, {"loss": 0.6364, "grad_norm": 0.6422147154808044, "learning_rate": 0.0002, "epoch": 2.2989249050198044, "step": 14220}, {"loss": 0.6795, "grad_norm": 0.6866879463195801, "learning_rate": 0.0002, "epoch": 2.3005415892005496, "step": 14230}, {"loss": 0.6907, "grad_norm": 0.9297130107879639, "learning_rate": 0.0002, "epoch": 2.302158273381295, "step": 14240}, {"loss": 0.6823, "grad_norm": 0.7501356601715088, "learning_rate": 0.0002, "epoch": 2.30377495756204, "step": 14250}, {"loss": 0.6414, "grad_norm": 0.8363515138626099, "learning_rate": 0.0002, "epoch": 2.3053916417427853, "step": 14260}, {"loss": 0.6362, "grad_norm": 0.9083868265151978, "learning_rate": 0.0002, "epoch": 2.307008325923531, "step": 14270}, {"loss": 0.6862, "grad_norm": 0.7791516780853271, "learning_rate": 0.0002, "epoch": 2.3086250101042762, "step": 14280}, {"loss": 0.6569, "grad_norm": 0.8766953349113464, "learning_rate": 0.0002, "epoch": 2.3102416942850215, "step": 14290}, {"loss": 0.6698, "grad_norm": 0.7916635274887085, "learning_rate": 0.0002, "epoch": 2.3118583784657667, "step": 14300}, {"loss": 0.6927, "grad_norm": 0.627525269985199, "learning_rate": 0.0002, "epoch": 2.313475062646512, "step": 14310}, {"loss": 0.6541, "grad_norm": 0.8856783509254456, "learning_rate": 0.0002, "epoch": 2.315091746827257, "step": 14320}, {"loss": 0.6806, "grad_norm": 0.6758689284324646, "learning_rate": 0.0002, "epoch": 2.316708431008003, "step": 14330}, {"loss": 0.6794, "grad_norm": 0.6428321003913879, "learning_rate": 0.0002, "epoch": 2.318325115188748, "step": 14340}, {"loss": 0.682, "grad_norm": 0.9032121300697327, "learning_rate": 0.0002, "epoch": 2.3199417993694933, "step": 14350}, {"loss": 0.6569, "grad_norm": 0.8035986423492432, "learning_rate": 0.0002, "epoch": 2.3215584835502385, "step": 14360}, {"loss": 0.7067, "grad_norm": 0.7974579334259033, "learning_rate": 0.0002, "epoch": 2.3231751677309838, "step": 14370}, {"loss": 0.6451, "grad_norm": 0.8356034755706787, "learning_rate": 0.0002, "epoch": 2.324791851911729, "step": 14380}, {"loss": 0.6623, "grad_norm": 0.998760998249054, "learning_rate": 0.0002, "epoch": 2.326408536092474, "step": 14390}, {"loss": 0.649, "grad_norm": 0.6518142223358154, "learning_rate": 0.0002, "epoch": 2.3280252202732195, "step": 14400}, {"loss": 0.7146, "grad_norm": 0.7443506717681885, "learning_rate": 0.0002, "epoch": 2.3296419044539647, "step": 14410}, {"loss": 0.648, "grad_norm": 0.8436172604560852, "learning_rate": 0.0002, "epoch": 2.3312585886347104, "step": 14420}, {"loss": 0.6585, "grad_norm": 0.7411080598831177, "learning_rate": 0.0002, "epoch": 2.3328752728154556, "step": 14430}, {"loss": 0.6781, "grad_norm": 0.8839048743247986, "learning_rate": 0.0002, "epoch": 2.334491956996201, "step": 14440}, {"loss": 0.6565, "grad_norm": 0.8360885977745056, "learning_rate": 0.0002, "epoch": 2.336108641176946, "step": 14450}, {"loss": 0.6662, "grad_norm": 0.7608986496925354, "learning_rate": 0.0002, "epoch": 2.3377253253576913, "step": 14460}, {"loss": 0.6685, "grad_norm": 0.8179867267608643, "learning_rate": 0.0002, "epoch": 2.3393420095384365, "step": 14470}, {"loss": 0.7055, "grad_norm": 0.5989999771118164, "learning_rate": 0.0002, "epoch": 2.340958693719182, "step": 14480}, {"loss": 0.644, "grad_norm": 0.9450054168701172, "learning_rate": 0.0002, "epoch": 2.3425753778999274, "step": 14490}, {"loss": 0.6983, "grad_norm": 0.7885149717330933, "learning_rate": 0.0002, "epoch": 2.3441920620806727, "step": 14500}, {"loss": 0.6819, "grad_norm": 0.8152616620063782, "learning_rate": 0.0002, "epoch": 2.345808746261418, "step": 14510}, {"loss": 0.6989, "grad_norm": 0.7193838953971863, "learning_rate": 0.0002, "epoch": 2.347425430442163, "step": 14520}, {"loss": 0.6594, "grad_norm": 0.6701092720031738, "learning_rate": 0.0002, "epoch": 2.3490421146229084, "step": 14530}, {"loss": 0.6559, "grad_norm": 0.7529364228248596, "learning_rate": 0.0002, "epoch": 2.3506587988036536, "step": 14540}, {"loss": 0.6306, "grad_norm": 0.6599733829498291, "learning_rate": 0.0002, "epoch": 2.352275482984399, "step": 14550}, {"loss": 0.706, "grad_norm": 0.9502474069595337, "learning_rate": 0.0002, "epoch": 2.353892167165144, "step": 14560}, {"loss": 0.717, "grad_norm": 0.7619650959968567, "learning_rate": 0.0002, "epoch": 2.3555088513458897, "step": 14570}, {"loss": 0.6684, "grad_norm": 0.9854652285575867, "learning_rate": 0.0002, "epoch": 2.357125535526635, "step": 14580}, {"loss": 0.6455, "grad_norm": 0.727439284324646, "learning_rate": 0.0002, "epoch": 2.35874221970738, "step": 14590}, {"loss": 0.6645, "grad_norm": 0.6994746327400208, "learning_rate": 0.0002, "epoch": 2.3603589038881254, "step": 14600}, {"loss": 0.6587, "grad_norm": 0.7117531299591064, "learning_rate": 0.0002, "epoch": 2.3619755880688706, "step": 14610}, {"loss": 0.6804, "grad_norm": 0.6403067708015442, "learning_rate": 0.0002, "epoch": 2.363592272249616, "step": 14620}, {"loss": 0.7055, "grad_norm": 0.8377841711044312, "learning_rate": 0.0002, "epoch": 2.3652089564303616, "step": 14630}, {"loss": 0.6778, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 2.366825640611107, "step": 14640}, {"loss": 0.6552, "grad_norm": 0.8418586254119873, "learning_rate": 0.0002, "epoch": 2.368442324791852, "step": 14650}, {"loss": 0.6685, "grad_norm": 0.6178573369979858, "learning_rate": 0.0002, "epoch": 2.3700590089725972, "step": 14660}, {"loss": 0.6774, "grad_norm": 0.6368302702903748, "learning_rate": 0.0002, "epoch": 2.3716756931533425, "step": 14670}, {"loss": 0.6136, "grad_norm": 0.9122977256774902, "learning_rate": 0.0002, "epoch": 2.3732923773340877, "step": 14680}, {"loss": 0.6675, "grad_norm": 0.7086195349693298, "learning_rate": 0.0002, "epoch": 2.374909061514833, "step": 14690}, {"loss": 0.6582, "grad_norm": 0.7500800490379333, "learning_rate": 0.0002, "epoch": 2.376525745695578, "step": 14700}, {"loss": 0.6792, "grad_norm": 0.6634900569915771, "learning_rate": 0.0002, "epoch": 2.378142429876324, "step": 14710}, {"loss": 0.6614, "grad_norm": 0.839898407459259, "learning_rate": 0.0002, "epoch": 2.379759114057069, "step": 14720}, {"loss": 0.6453, "grad_norm": 0.7578426003456116, "learning_rate": 0.0002, "epoch": 2.3813757982378143, "step": 14730}, {"loss": 0.7282, "grad_norm": 1.0213173627853394, "learning_rate": 0.0002, "epoch": 2.3829924824185595, "step": 14740}, {"loss": 0.6704, "grad_norm": 0.7855949401855469, "learning_rate": 0.0002, "epoch": 2.3846091665993048, "step": 14750}, {"loss": 0.6694, "grad_norm": 0.7224128842353821, "learning_rate": 0.0002, "epoch": 2.38622585078005, "step": 14760}, {"loss": 0.7017, "grad_norm": 0.8040381669998169, "learning_rate": 0.0002, "epoch": 2.3878425349607952, "step": 14770}, {"loss": 0.6799, "grad_norm": 0.7705281376838684, "learning_rate": 0.0002, "epoch": 2.389459219141541, "step": 14780}, {"loss": 0.6326, "grad_norm": 0.667966902256012, "learning_rate": 0.0002, "epoch": 2.391075903322286, "step": 14790}, {"loss": 0.7061, "grad_norm": 0.6611011028289795, "learning_rate": 0.0002, "epoch": 2.3926925875030314, "step": 14800}, {"loss": 0.6527, "grad_norm": 0.6862651705741882, "learning_rate": 0.0002, "epoch": 2.3943092716837766, "step": 14810}, {"loss": 0.6537, "grad_norm": 0.8086010217666626, "learning_rate": 0.0002, "epoch": 2.395925955864522, "step": 14820}, {"loss": 0.7189, "grad_norm": 0.7189689874649048, "learning_rate": 0.0002, "epoch": 2.397542640045267, "step": 14830}, {"loss": 0.6709, "grad_norm": 0.6280009150505066, "learning_rate": 0.0002, "epoch": 2.3991593242260123, "step": 14840}, {"loss": 0.706, "grad_norm": 0.7826612591743469, "learning_rate": 0.0002, "epoch": 2.4007760084067575, "step": 14850}, {"loss": 0.6738, "grad_norm": 0.7681610584259033, "learning_rate": 0.0002, "epoch": 2.402392692587503, "step": 14860}, {"loss": 0.636, "grad_norm": 0.720966100692749, "learning_rate": 0.0002, "epoch": 2.4040093767682484, "step": 14870}, {"loss": 0.6667, "grad_norm": 0.8202250599861145, "learning_rate": 0.0002, "epoch": 2.4056260609489937, "step": 14880}, {"loss": 0.6935, "grad_norm": 0.786212682723999, "learning_rate": 0.0002, "epoch": 2.407242745129739, "step": 14890}, {"loss": 0.6628, "grad_norm": 0.6647164821624756, "learning_rate": 0.0002, "epoch": 2.408859429310484, "step": 14900}, {"loss": 0.6706, "grad_norm": 0.7566399574279785, "learning_rate": 0.0002, "epoch": 2.4104761134912294, "step": 14910}, {"loss": 0.7188, "grad_norm": 0.748814582824707, "learning_rate": 0.0002, "epoch": 2.4120927976719746, "step": 14920}, {"loss": 0.6684, "grad_norm": 0.7624038457870483, "learning_rate": 0.0002, "epoch": 2.4137094818527203, "step": 14930}, {"loss": 0.6483, "grad_norm": 0.8267335295677185, "learning_rate": 0.0002, "epoch": 2.4153261660334655, "step": 14940}, {"loss": 0.6612, "grad_norm": 0.8785360455513, "learning_rate": 0.0002, "epoch": 2.4169428502142107, "step": 14950}, {"loss": 0.6718, "grad_norm": 0.679887592792511, "learning_rate": 0.0002, "epoch": 2.418559534394956, "step": 14960}, {"loss": 0.6136, "grad_norm": 0.7218474745750427, "learning_rate": 0.0002, "epoch": 2.420176218575701, "step": 14970}, {"loss": 0.648, "grad_norm": 0.6342799663543701, "learning_rate": 0.0002, "epoch": 2.4217929027564464, "step": 14980}, {"loss": 0.6617, "grad_norm": 0.7098712921142578, "learning_rate": 0.0002, "epoch": 2.4234095869371917, "step": 14990}, {"loss": 0.6942, "grad_norm": 0.7497431635856628, "learning_rate": 0.0002, "epoch": 2.425026271117937, "step": 15000}, {"loss": 0.6772, "grad_norm": 0.934836208820343, "learning_rate": 0.0002, "epoch": 2.4266429552986826, "step": 15010}, {"loss": 0.7221, "grad_norm": 0.8430966734886169, "learning_rate": 0.0002, "epoch": 2.428259639479428, "step": 15020}, {"loss": 0.6985, "grad_norm": 0.7032104730606079, "learning_rate": 0.0002, "epoch": 2.429876323660173, "step": 15030}, {"loss": 0.6715, "grad_norm": 0.7746111750602722, "learning_rate": 0.0002, "epoch": 2.4314930078409183, "step": 15040}, {"loss": 0.7177, "grad_norm": 0.7661406397819519, "learning_rate": 0.0002, "epoch": 2.4331096920216635, "step": 15050}, {"loss": 0.6517, "grad_norm": 0.6941645741462708, "learning_rate": 0.0002, "epoch": 2.4347263762024087, "step": 15060}, {"loss": 0.6421, "grad_norm": 0.7487249374389648, "learning_rate": 0.0002, "epoch": 2.436343060383154, "step": 15070}, {"loss": 0.6796, "grad_norm": 0.7639912962913513, "learning_rate": 0.0002, "epoch": 2.4379597445638996, "step": 15080}, {"loss": 0.7087, "grad_norm": 0.7708953619003296, "learning_rate": 0.0002, "epoch": 2.439576428744645, "step": 15090}, {"loss": 0.7065, "grad_norm": 0.9135832190513611, "learning_rate": 0.0002, "epoch": 2.44119311292539, "step": 15100}, {"loss": 0.672, "grad_norm": 0.8283005356788635, "learning_rate": 0.0002, "epoch": 2.4428097971061353, "step": 15110}, {"loss": 0.6551, "grad_norm": 0.925299346446991, "learning_rate": 0.0002, "epoch": 2.4444264812868806, "step": 15120}, {"loss": 0.687, "grad_norm": 0.7013528943061829, "learning_rate": 0.0002, "epoch": 2.446043165467626, "step": 15130}, {"loss": 0.6842, "grad_norm": 0.622303307056427, "learning_rate": 0.0002, "epoch": 2.447659849648371, "step": 15140}, {"loss": 0.6676, "grad_norm": 0.876569390296936, "learning_rate": 0.0002, "epoch": 2.4492765338291163, "step": 15150}, {"loss": 0.6463, "grad_norm": 0.6836351752281189, "learning_rate": 0.0002, "epoch": 2.450893218009862, "step": 15160}, {"loss": 0.6781, "grad_norm": 0.7886684536933899, "learning_rate": 0.0002, "epoch": 2.452509902190607, "step": 15170}, {"loss": 0.6794, "grad_norm": 0.6647440791130066, "learning_rate": 0.0002, "epoch": 2.4541265863713524, "step": 15180}, {"loss": 0.6353, "grad_norm": 0.7477722764015198, "learning_rate": 0.0002, "epoch": 2.4557432705520976, "step": 15190}, {"loss": 0.698, "grad_norm": 0.8192033767700195, "learning_rate": 0.0002, "epoch": 2.457359954732843, "step": 15200}, {"loss": 0.6735, "grad_norm": 0.847537100315094, "learning_rate": 0.0002, "epoch": 2.458976638913588, "step": 15210}, {"loss": 0.6962, "grad_norm": 0.9027776122093201, "learning_rate": 0.0002, "epoch": 2.4605933230943338, "step": 15220}, {"loss": 0.7084, "grad_norm": 0.7217772006988525, "learning_rate": 0.0002, "epoch": 2.462210007275079, "step": 15230}, {"loss": 0.691, "grad_norm": 0.7994546294212341, "learning_rate": 0.0002, "epoch": 2.4638266914558242, "step": 15240}, {"loss": 0.6828, "grad_norm": 0.939916729927063, "learning_rate": 0.0002, "epoch": 2.4654433756365695, "step": 15250}, {"loss": 0.6893, "grad_norm": 1.0009053945541382, "learning_rate": 0.0002, "epoch": 2.4670600598173147, "step": 15260}, {"loss": 0.643, "grad_norm": 0.625555694103241, "learning_rate": 0.0002, "epoch": 2.46867674399806, "step": 15270}, {"loss": 0.688, "grad_norm": 0.7924878597259521, "learning_rate": 0.0002, "epoch": 2.470293428178805, "step": 15280}, {"loss": 0.6789, "grad_norm": 0.8536689877510071, "learning_rate": 0.0002, "epoch": 2.4719101123595504, "step": 15290}, {"loss": 0.6924, "grad_norm": 0.8572589755058289, "learning_rate": 0.0002, "epoch": 2.4735267965402956, "step": 15300}, {"loss": 0.604, "grad_norm": 0.773279070854187, "learning_rate": 0.0002, "epoch": 2.4751434807210413, "step": 15310}, {"loss": 0.6573, "grad_norm": 0.7708749771118164, "learning_rate": 0.0002, "epoch": 2.4767601649017865, "step": 15320}, {"loss": 0.7065, "grad_norm": 0.770905077457428, "learning_rate": 0.0002, "epoch": 2.4783768490825318, "step": 15330}, {"loss": 0.6878, "grad_norm": 0.8238571882247925, "learning_rate": 0.0002, "epoch": 2.479993533263277, "step": 15340}, {"loss": 0.6772, "grad_norm": 0.7670477032661438, "learning_rate": 0.0002, "epoch": 2.481610217444022, "step": 15350}, {"loss": 0.7759, "grad_norm": 0.905036985874176, "learning_rate": 0.0002, "epoch": 2.4832269016247674, "step": 15360}, {"loss": 0.706, "grad_norm": 0.6672089695930481, "learning_rate": 0.0002, "epoch": 2.484843585805513, "step": 15370}, {"loss": 0.6722, "grad_norm": 0.625095784664154, "learning_rate": 0.0002, "epoch": 2.4864602699862584, "step": 15380}, {"loss": 0.6396, "grad_norm": 0.679772675037384, "learning_rate": 0.0002, "epoch": 2.4880769541670036, "step": 15390}, {"loss": 0.6778, "grad_norm": 0.711492121219635, "learning_rate": 0.0002, "epoch": 2.489693638347749, "step": 15400}, {"loss": 0.6966, "grad_norm": 0.876189112663269, "learning_rate": 0.0002, "epoch": 2.491310322528494, "step": 15410}, {"loss": 0.7307, "grad_norm": 0.7236915230751038, "learning_rate": 0.0002, "epoch": 2.4929270067092393, "step": 15420}, {"loss": 0.647, "grad_norm": 0.6629832983016968, "learning_rate": 0.0002, "epoch": 2.4945436908899845, "step": 15430}, {"loss": 0.6669, "grad_norm": 0.9756859540939331, "learning_rate": 0.0002, "epoch": 2.4961603750707297, "step": 15440}, {"loss": 0.7559, "grad_norm": 0.6896940469741821, "learning_rate": 0.0002, "epoch": 2.4977770592514754, "step": 15450}, {"loss": 0.6818, "grad_norm": 0.7105149626731873, "learning_rate": 0.0002, "epoch": 2.4993937434322206, "step": 15460}, {"loss": 0.6859, "grad_norm": 0.8374546766281128, "learning_rate": 0.0002, "epoch": 2.501010427612966, "step": 15470}, {"loss": 0.6512, "grad_norm": 0.7320070266723633, "learning_rate": 0.0002, "epoch": 2.502627111793711, "step": 15480}, {"loss": 0.685, "grad_norm": 0.8306367993354797, "learning_rate": 0.0002, "epoch": 2.5042437959744563, "step": 15490}, {"loss": 0.7253, "grad_norm": 0.7472721338272095, "learning_rate": 0.0002, "epoch": 2.5058604801552016, "step": 15500}, {"loss": 0.6699, "grad_norm": 0.6147692203521729, "learning_rate": 0.0002, "epoch": 2.507477164335947, "step": 15510}, {"loss": 0.7158, "grad_norm": 0.7788505554199219, "learning_rate": 0.0002, "epoch": 2.5090938485166925, "step": 15520}, {"loss": 0.6521, "grad_norm": 0.8807527422904968, "learning_rate": 0.0002, "epoch": 2.5107105326974377, "step": 15530}, {"loss": 0.6792, "grad_norm": 0.7521643042564392, "learning_rate": 0.0002, "epoch": 2.512327216878183, "step": 15540}, {"loss": 0.6772, "grad_norm": 0.6900225281715393, "learning_rate": 0.0002, "epoch": 2.513943901058928, "step": 15550}, {"loss": 0.6769, "grad_norm": 0.6601938605308533, "learning_rate": 0.0002, "epoch": 2.5155605852396734, "step": 15560}, {"loss": 0.6648, "grad_norm": 0.8179984092712402, "learning_rate": 0.0002, "epoch": 2.5171772694204186, "step": 15570}, {"loss": 0.7028, "grad_norm": 0.792556881904602, "learning_rate": 0.0002, "epoch": 2.518793953601164, "step": 15580}, {"loss": 0.6464, "grad_norm": 0.7081938982009888, "learning_rate": 0.0002, "epoch": 2.520410637781909, "step": 15590}, {"loss": 0.6691, "grad_norm": 0.8733121156692505, "learning_rate": 0.0002, "epoch": 2.5220273219626543, "step": 15600}, {"loss": 0.6969, "grad_norm": 0.7980992794036865, "learning_rate": 0.0002, "epoch": 2.5236440061434, "step": 15610}, {"loss": 0.7124, "grad_norm": 0.883664071559906, "learning_rate": 0.0002, "epoch": 2.5252606903241452, "step": 15620}, {"loss": 0.7022, "grad_norm": 0.6963341236114502, "learning_rate": 0.0002, "epoch": 2.5268773745048905, "step": 15630}, {"loss": 0.7334, "grad_norm": 0.6433573365211487, "learning_rate": 0.0002, "epoch": 2.5284940586856357, "step": 15640}, {"loss": 0.6889, "grad_norm": 0.8538183569908142, "learning_rate": 0.0002, "epoch": 2.530110742866381, "step": 15650}, {"loss": 0.6841, "grad_norm": 0.9748201370239258, "learning_rate": 0.0002, "epoch": 2.5317274270471266, "step": 15660}, {"loss": 0.6765, "grad_norm": 0.7670575380325317, "learning_rate": 0.0002, "epoch": 2.533344111227872, "step": 15670}, {"loss": 0.6435, "grad_norm": 0.8738890290260315, "learning_rate": 0.0002, "epoch": 2.534960795408617, "step": 15680}, {"loss": 0.6802, "grad_norm": 0.8391636610031128, "learning_rate": 0.0002, "epoch": 2.5365774795893623, "step": 15690}, {"loss": 0.6901, "grad_norm": 0.7239366769790649, "learning_rate": 0.0002, "epoch": 2.5381941637701075, "step": 15700}, {"loss": 0.7011, "grad_norm": 0.8498379588127136, "learning_rate": 0.0002, "epoch": 2.5398108479508528, "step": 15710}, {"loss": 0.6998, "grad_norm": 0.8029484152793884, "learning_rate": 0.0002, "epoch": 2.541427532131598, "step": 15720}, {"loss": 0.6678, "grad_norm": 1.0639333724975586, "learning_rate": 0.0002, "epoch": 2.5430442163123432, "step": 15730}, {"loss": 0.6341, "grad_norm": 0.6401297450065613, "learning_rate": 0.0002, "epoch": 2.5446609004930885, "step": 15740}, {"loss": 0.7196, "grad_norm": 0.7123814821243286, "learning_rate": 0.0002, "epoch": 2.5462775846738337, "step": 15750}, {"loss": 0.654, "grad_norm": 0.7874974608421326, "learning_rate": 0.0002, "epoch": 2.5478942688545794, "step": 15760}, {"loss": 0.6721, "grad_norm": 0.8046808838844299, "learning_rate": 0.0002, "epoch": 2.5495109530353246, "step": 15770}, {"loss": 0.6665, "grad_norm": 0.7888661623001099, "learning_rate": 0.0002, "epoch": 2.55112763721607, "step": 15780}, {"loss": 0.6893, "grad_norm": 0.8445866107940674, "learning_rate": 0.0002, "epoch": 2.552744321396815, "step": 15790}, {"loss": 0.6815, "grad_norm": 0.7475846409797668, "learning_rate": 0.0002, "epoch": 2.5543610055775603, "step": 15800}, {"loss": 0.6711, "grad_norm": 0.7455102801322937, "learning_rate": 0.0002, "epoch": 2.555977689758306, "step": 15810}, {"loss": 0.6932, "grad_norm": 0.8226983547210693, "learning_rate": 0.0002, "epoch": 2.557594373939051, "step": 15820}, {"loss": 0.651, "grad_norm": 0.8920368552207947, "learning_rate": 0.0002, "epoch": 2.5592110581197964, "step": 15830}, {"loss": 0.6297, "grad_norm": 0.8413904905319214, "learning_rate": 0.0002, "epoch": 2.5608277423005417, "step": 15840}, {"loss": 0.7106, "grad_norm": 0.8483649492263794, "learning_rate": 0.0002, "epoch": 2.562444426481287, "step": 15850}, {"loss": 0.6957, "grad_norm": 0.5923284292221069, "learning_rate": 0.0002, "epoch": 2.564061110662032, "step": 15860}, {"loss": 0.6847, "grad_norm": 0.8518726229667664, "learning_rate": 0.0002, "epoch": 2.5656777948427774, "step": 15870}, {"loss": 0.6362, "grad_norm": 0.731235146522522, "learning_rate": 0.0002, "epoch": 2.5672944790235226, "step": 15880}, {"loss": 0.7611, "grad_norm": 0.7517194151878357, "learning_rate": 0.0002, "epoch": 2.568911163204268, "step": 15890}, {"loss": 0.6907, "grad_norm": 0.8378692269325256, "learning_rate": 0.0002, "epoch": 2.5705278473850135, "step": 15900}, {"loss": 0.7055, "grad_norm": 0.843701958656311, "learning_rate": 0.0002, "epoch": 2.5721445315657587, "step": 15910}, {"loss": 0.6882, "grad_norm": 0.7254629731178284, "learning_rate": 0.0002, "epoch": 2.573761215746504, "step": 15920}, {"loss": 0.6872, "grad_norm": 0.8863335847854614, "learning_rate": 0.0002, "epoch": 2.575377899927249, "step": 15930}, {"loss": 0.6813, "grad_norm": 0.7675097584724426, "learning_rate": 0.0002, "epoch": 2.5769945841079944, "step": 15940}, {"loss": 0.7357, "grad_norm": 0.82063889503479, "learning_rate": 0.0002, "epoch": 2.5786112682887397, "step": 15950}, {"loss": 0.662, "grad_norm": 0.7729717493057251, "learning_rate": 0.0002, "epoch": 2.5802279524694853, "step": 15960}, {"loss": 0.633, "grad_norm": 0.8301846981048584, "learning_rate": 0.0002, "epoch": 2.5818446366502306, "step": 15970}, {"loss": 0.6897, "grad_norm": 0.7906861305236816, "learning_rate": 0.0002, "epoch": 2.583461320830976, "step": 15980}, {"loss": 0.7175, "grad_norm": 0.6749057173728943, "learning_rate": 0.0002, "epoch": 2.585078005011721, "step": 15990}, {"loss": 0.7212, "grad_norm": 0.9386842846870422, "learning_rate": 0.0002, "epoch": 2.5866946891924663, "step": 16000}, {"loss": 0.6934, "grad_norm": 0.7868891358375549, "learning_rate": 0.0002, "epoch": 2.5883113733732115, "step": 16010}, {"loss": 0.7036, "grad_norm": 0.8674671053886414, "learning_rate": 0.0002, "epoch": 2.5899280575539567, "step": 16020}, {"loss": 0.7217, "grad_norm": 0.7043559551239014, "learning_rate": 0.0002, "epoch": 2.591544741734702, "step": 16030}, {"loss": 0.6967, "grad_norm": 0.5846083760261536, "learning_rate": 0.0002, "epoch": 2.593161425915447, "step": 16040}, {"loss": 0.7322, "grad_norm": 0.7323982119560242, "learning_rate": 0.0002, "epoch": 2.594778110096193, "step": 16050}, {"loss": 0.6794, "grad_norm": 0.9069556593894958, "learning_rate": 0.0002, "epoch": 2.596394794276938, "step": 16060}, {"loss": 0.7076, "grad_norm": 0.7522736191749573, "learning_rate": 0.0002, "epoch": 2.5980114784576833, "step": 16070}, {"loss": 0.6477, "grad_norm": 0.8149648308753967, "learning_rate": 0.0002, "epoch": 2.5996281626384286, "step": 16080}, {"loss": 0.6664, "grad_norm": 0.6214233040809631, "learning_rate": 0.0002, "epoch": 2.601244846819174, "step": 16090}, {"loss": 0.7307, "grad_norm": 0.6803743839263916, "learning_rate": 0.0002, "epoch": 2.602861530999919, "step": 16100}, {"loss": 0.7244, "grad_norm": 0.7223997116088867, "learning_rate": 0.0002, "epoch": 2.6044782151806647, "step": 16110}, {"loss": 0.6867, "grad_norm": 0.7324174642562866, "learning_rate": 0.0002, "epoch": 2.60609489936141, "step": 16120}, {"loss": 0.7159, "grad_norm": 0.9594739675521851, "learning_rate": 0.0002, "epoch": 2.607711583542155, "step": 16130}, {"loss": 0.6451, "grad_norm": 0.9485327005386353, "learning_rate": 0.0002, "epoch": 2.6093282677229004, "step": 16140}, {"loss": 0.6815, "grad_norm": 0.8449000120162964, "learning_rate": 0.0002, "epoch": 2.6109449519036456, "step": 16150}, {"loss": 0.7152, "grad_norm": 0.8520140051841736, "learning_rate": 0.0002, "epoch": 2.612561636084391, "step": 16160}, {"loss": 0.6759, "grad_norm": 0.7456524968147278, "learning_rate": 0.0002, "epoch": 2.614178320265136, "step": 16170}, {"loss": 0.6893, "grad_norm": 0.9912857413291931, "learning_rate": 0.0002, "epoch": 2.6157950044458813, "step": 16180}, {"loss": 0.7243, "grad_norm": 0.9001946449279785, "learning_rate": 0.0002, "epoch": 2.6174116886266265, "step": 16190}, {"loss": 0.6825, "grad_norm": 0.6568667888641357, "learning_rate": 0.0002, "epoch": 2.619028372807372, "step": 16200}, {"loss": 0.7013, "grad_norm": 1.0248128175735474, "learning_rate": 0.0002, "epoch": 2.6206450569881174, "step": 16210}, {"loss": 0.7045, "grad_norm": 0.6509039998054504, "learning_rate": 0.0002, "epoch": 2.6222617411688627, "step": 16220}, {"loss": 0.72, "grad_norm": 0.7626351118087769, "learning_rate": 0.0002, "epoch": 2.623878425349608, "step": 16230}, {"loss": 0.6556, "grad_norm": 0.6938552260398865, "learning_rate": 0.0002, "epoch": 2.625495109530353, "step": 16240}, {"loss": 0.65, "grad_norm": 0.6434680819511414, "learning_rate": 0.0002, "epoch": 2.6271117937110984, "step": 16250}, {"loss": 0.6943, "grad_norm": 0.7111515998840332, "learning_rate": 0.0002, "epoch": 2.628728477891844, "step": 16260}, {"loss": 0.679, "grad_norm": 0.7712395787239075, "learning_rate": 0.0002, "epoch": 2.6303451620725893, "step": 16270}, {"loss": 0.6886, "grad_norm": 0.792209267616272, "learning_rate": 0.0002, "epoch": 2.6319618462533345, "step": 16280}, {"loss": 0.6554, "grad_norm": 0.6801066398620605, "learning_rate": 0.0002, "epoch": 2.6335785304340797, "step": 16290}, {"loss": 0.73, "grad_norm": 0.7802573442459106, "learning_rate": 0.0002, "epoch": 2.635195214614825, "step": 16300}, {"loss": 0.7484, "grad_norm": 0.7742244601249695, "learning_rate": 0.0002, "epoch": 2.63681189879557, "step": 16310}, {"loss": 0.6524, "grad_norm": 0.664184033870697, "learning_rate": 0.0002, "epoch": 2.6384285829763154, "step": 16320}, {"loss": 0.6442, "grad_norm": 0.9242228865623474, "learning_rate": 0.0002, "epoch": 2.6400452671570607, "step": 16330}, {"loss": 0.6792, "grad_norm": 0.9661325216293335, "learning_rate": 0.0002, "epoch": 2.641661951337806, "step": 16340}, {"loss": 0.6847, "grad_norm": 0.837526798248291, "learning_rate": 0.0002, "epoch": 2.6432786355185516, "step": 16350}, {"loss": 0.7686, "grad_norm": 1.1834373474121094, "learning_rate": 0.0002, "epoch": 2.644895319699297, "step": 16360}, {"loss": 0.6746, "grad_norm": 0.7467831373214722, "learning_rate": 0.0002, "epoch": 2.646512003880042, "step": 16370}, {"loss": 0.6935, "grad_norm": 0.8627146482467651, "learning_rate": 0.0002, "epoch": 2.6481286880607873, "step": 16380}, {"loss": 0.715, "grad_norm": 0.790447473526001, "learning_rate": 0.0002, "epoch": 2.6497453722415325, "step": 16390}, {"loss": 0.723, "grad_norm": 0.8447365164756775, "learning_rate": 0.0002, "epoch": 2.651362056422278, "step": 16400}, {"loss": 0.6628, "grad_norm": 0.7831417918205261, "learning_rate": 0.0002, "epoch": 2.6529787406030234, "step": 16410}, {"loss": 0.6691, "grad_norm": 0.6837952136993408, "learning_rate": 0.0002, "epoch": 2.6545954247837686, "step": 16420}, {"loss": 0.6139, "grad_norm": 0.7031801342964172, "learning_rate": 0.0002, "epoch": 2.656212108964514, "step": 16430}, {"loss": 0.7382, "grad_norm": 0.8963770866394043, "learning_rate": 0.0002, "epoch": 2.657828793145259, "step": 16440}, {"loss": 0.6439, "grad_norm": 0.6852328181266785, "learning_rate": 0.0002, "epoch": 2.6594454773260043, "step": 16450}, {"loss": 0.6278, "grad_norm": 0.8069294095039368, "learning_rate": 0.0002, "epoch": 2.6610621615067496, "step": 16460}, {"loss": 0.6939, "grad_norm": 0.7503686547279358, "learning_rate": 0.0002, "epoch": 2.662678845687495, "step": 16470}, {"loss": 0.6777, "grad_norm": 0.6430956125259399, "learning_rate": 0.0002, "epoch": 2.66429552986824, "step": 16480}, {"loss": 0.6863, "grad_norm": 0.7894312739372253, "learning_rate": 0.0002, "epoch": 2.6659122140489853, "step": 16490}, {"loss": 0.7165, "grad_norm": 0.7277431488037109, "learning_rate": 0.0002, "epoch": 2.667528898229731, "step": 16500}, {"loss": 0.6772, "grad_norm": 0.6816153526306152, "learning_rate": 0.0002, "epoch": 2.669145582410476, "step": 16510}, {"loss": 0.691, "grad_norm": 0.8145235776901245, "learning_rate": 0.0002, "epoch": 2.6707622665912214, "step": 16520}, {"loss": 0.709, "grad_norm": 0.8645890355110168, "learning_rate": 0.0002, "epoch": 2.6723789507719666, "step": 16530}, {"loss": 0.6946, "grad_norm": 0.704393208026886, "learning_rate": 0.0002, "epoch": 2.673995634952712, "step": 16540}, {"loss": 0.6378, "grad_norm": 1.0120846033096313, "learning_rate": 0.0002, "epoch": 2.6756123191334575, "step": 16550}, {"loss": 0.7241, "grad_norm": 0.6919328570365906, "learning_rate": 0.0002, "epoch": 2.6772290033142028, "step": 16560}, {"loss": 0.7098, "grad_norm": 0.6924574971199036, "learning_rate": 0.0002, "epoch": 2.678845687494948, "step": 16570}, {"loss": 0.731, "grad_norm": 0.9679301381111145, "learning_rate": 0.0002, "epoch": 2.6804623716756932, "step": 16580}, {"loss": 0.7124, "grad_norm": 0.6810211539268494, "learning_rate": 0.0002, "epoch": 2.6820790558564385, "step": 16590}, {"loss": 0.6688, "grad_norm": 0.9730555415153503, "learning_rate": 0.0002, "epoch": 2.6836957400371837, "step": 16600}, {"loss": 0.7344, "grad_norm": 0.7852821350097656, "learning_rate": 0.0002, "epoch": 2.685312424217929, "step": 16610}, {"loss": 0.6401, "grad_norm": 0.6059057116508484, "learning_rate": 0.0002, "epoch": 2.686929108398674, "step": 16620}, {"loss": 0.6796, "grad_norm": 0.9395958781242371, "learning_rate": 0.0002, "epoch": 2.6885457925794194, "step": 16630}, {"loss": 0.7174, "grad_norm": 0.7473729848861694, "learning_rate": 0.0002, "epoch": 2.690162476760165, "step": 16640}, {"loss": 0.7087, "grad_norm": 0.765934407711029, "learning_rate": 0.0002, "epoch": 2.6917791609409103, "step": 16650}, {"loss": 0.707, "grad_norm": 0.8496677279472351, "learning_rate": 0.0002, "epoch": 2.6933958451216555, "step": 16660}, {"loss": 0.7084, "grad_norm": 0.7641879916191101, "learning_rate": 0.0002, "epoch": 2.6950125293024008, "step": 16670}, {"loss": 0.6566, "grad_norm": 0.8471952676773071, "learning_rate": 0.0002, "epoch": 2.696629213483146, "step": 16680}, {"loss": 0.6635, "grad_norm": 0.6946060657501221, "learning_rate": 0.0002, "epoch": 2.6982458976638912, "step": 16690}, {"loss": 0.7027, "grad_norm": 0.7361312508583069, "learning_rate": 0.0002, "epoch": 2.699862581844637, "step": 16700}, {"loss": 0.6767, "grad_norm": 0.6605038046836853, "learning_rate": 0.0002, "epoch": 2.701479266025382, "step": 16710}, {"loss": 0.6885, "grad_norm": 0.7164411544799805, "learning_rate": 0.0002, "epoch": 2.7030959502061274, "step": 16720}, {"loss": 0.6736, "grad_norm": 0.6496201157569885, "learning_rate": 0.0002, "epoch": 2.7047126343868726, "step": 16730}, {"loss": 0.6942, "grad_norm": 0.7826663851737976, "learning_rate": 0.0002, "epoch": 2.706329318567618, "step": 16740}, {"loss": 0.6773, "grad_norm": 0.7639131546020508, "learning_rate": 0.0002, "epoch": 2.707946002748363, "step": 16750}, {"loss": 0.69, "grad_norm": 0.7976210713386536, "learning_rate": 0.0002, "epoch": 2.7095626869291083, "step": 16760}, {"loss": 0.6735, "grad_norm": 0.6836577653884888, "learning_rate": 0.0002, "epoch": 2.7111793711098535, "step": 16770}, {"loss": 0.6596, "grad_norm": 0.8025202751159668, "learning_rate": 0.0002, "epoch": 2.7127960552905988, "step": 16780}, {"loss": 0.6324, "grad_norm": 0.7636463642120361, "learning_rate": 0.0002, "epoch": 2.7144127394713444, "step": 16790}, {"loss": 0.6227, "grad_norm": 0.7481677532196045, "learning_rate": 0.0002, "epoch": 2.7160294236520897, "step": 16800}, {"loss": 0.6925, "grad_norm": 0.7566834688186646, "learning_rate": 0.0002, "epoch": 2.717646107832835, "step": 16810}, {"loss": 0.6531, "grad_norm": 0.7931267619132996, "learning_rate": 0.0002, "epoch": 2.71926279201358, "step": 16820}, {"loss": 0.6672, "grad_norm": 0.8811662197113037, "learning_rate": 0.0002, "epoch": 2.7208794761943254, "step": 16830}, {"loss": 0.6675, "grad_norm": 0.8561240434646606, "learning_rate": 0.0002, "epoch": 2.7224961603750706, "step": 16840}, {"loss": 0.7135, "grad_norm": 0.7121599316596985, "learning_rate": 0.0002, "epoch": 2.7241128445558163, "step": 16850}, {"loss": 0.6825, "grad_norm": 0.8066257238388062, "learning_rate": 0.0002, "epoch": 2.7257295287365615, "step": 16860}, {"loss": 0.6839, "grad_norm": 0.7699271440505981, "learning_rate": 0.0002, "epoch": 2.7273462129173067, "step": 16870}, {"loss": 0.699, "grad_norm": 1.1828432083129883, "learning_rate": 0.0002, "epoch": 2.728962897098052, "step": 16880}, {"loss": 0.6518, "grad_norm": 0.9989302754402161, "learning_rate": 0.0002, "epoch": 2.730579581278797, "step": 16890}, {"loss": 0.7015, "grad_norm": 0.8100560307502747, "learning_rate": 0.0002, "epoch": 2.7321962654595424, "step": 16900}, {"loss": 0.6851, "grad_norm": 0.8615233898162842, "learning_rate": 0.0002, "epoch": 2.7338129496402876, "step": 16910}, {"loss": 0.6322, "grad_norm": 0.8633756041526794, "learning_rate": 0.0002, "epoch": 2.735429633821033, "step": 16920}, {"loss": 0.6488, "grad_norm": 0.7769348621368408, "learning_rate": 0.0002, "epoch": 2.737046318001778, "step": 16930}, {"loss": 0.6582, "grad_norm": 0.6943058371543884, "learning_rate": 0.0002, "epoch": 2.738663002182524, "step": 16940}, {"loss": 0.6516, "grad_norm": 0.8510736227035522, "learning_rate": 0.0002, "epoch": 2.740279686363269, "step": 16950}, {"loss": 0.7275, "grad_norm": 0.7732602953910828, "learning_rate": 0.0002, "epoch": 2.7418963705440142, "step": 16960}, {"loss": 0.6553, "grad_norm": 0.5981788635253906, "learning_rate": 0.0002, "epoch": 2.7435130547247595, "step": 16970}, {"loss": 0.6777, "grad_norm": 0.7604416012763977, "learning_rate": 0.0002, "epoch": 2.7451297389055047, "step": 16980}, {"loss": 0.6981, "grad_norm": 0.7377738356590271, "learning_rate": 0.0002, "epoch": 2.74674642308625, "step": 16990}, {"loss": 0.6294, "grad_norm": 0.9400289058685303, "learning_rate": 0.0002, "epoch": 2.7483631072669956, "step": 17000}, {"loss": 0.6952, "grad_norm": 0.6340599656105042, "learning_rate": 0.0002, "epoch": 2.749979791447741, "step": 17010}, {"loss": 0.7222, "grad_norm": 0.7297601103782654, "learning_rate": 0.0002, "epoch": 2.751596475628486, "step": 17020}, {"loss": 0.6659, "grad_norm": 0.9479979872703552, "learning_rate": 0.0002, "epoch": 2.7532131598092313, "step": 17030}, {"loss": 0.691, "grad_norm": 0.8461511135101318, "learning_rate": 0.0002, "epoch": 2.7548298439899765, "step": 17040}, {"loss": 0.6764, "grad_norm": 0.7477551698684692, "learning_rate": 0.0002, "epoch": 2.7564465281707218, "step": 17050}, {"loss": 0.684, "grad_norm": 1.019270420074463, "learning_rate": 0.0002, "epoch": 2.758063212351467, "step": 17060}, {"loss": 0.7119, "grad_norm": 0.7730235457420349, "learning_rate": 0.0002, "epoch": 2.7596798965322122, "step": 17070}, {"loss": 0.6886, "grad_norm": 0.8216866254806519, "learning_rate": 0.0002, "epoch": 2.7612965807129575, "step": 17080}, {"loss": 0.6811, "grad_norm": 0.7235931754112244, "learning_rate": 0.0002, "epoch": 2.762913264893703, "step": 17090}, {"loss": 0.7031, "grad_norm": 0.7352296710014343, "learning_rate": 0.0002, "epoch": 2.7645299490744484, "step": 17100}, {"loss": 0.6951, "grad_norm": 0.8129373788833618, "learning_rate": 0.0002, "epoch": 2.7661466332551936, "step": 17110}, {"loss": 0.6703, "grad_norm": 0.7387019991874695, "learning_rate": 0.0002, "epoch": 2.767763317435939, "step": 17120}, {"loss": 0.6789, "grad_norm": 0.9149190187454224, "learning_rate": 0.0002, "epoch": 2.769380001616684, "step": 17130}, {"loss": 0.6038, "grad_norm": 0.7352971434593201, "learning_rate": 0.0002, "epoch": 2.7709966857974297, "step": 17140}, {"loss": 0.6728, "grad_norm": 0.7903780341148376, "learning_rate": 0.0002, "epoch": 2.772613369978175, "step": 17150}, {"loss": 0.6988, "grad_norm": 0.8255927562713623, "learning_rate": 0.0002, "epoch": 2.77423005415892, "step": 17160}, {"loss": 0.6694, "grad_norm": 0.7235927581787109, "learning_rate": 0.0002, "epoch": 2.7758467383396654, "step": 17170}, {"loss": 0.7161, "grad_norm": 0.8281434774398804, "learning_rate": 0.0002, "epoch": 2.7774634225204107, "step": 17180}, {"loss": 0.682, "grad_norm": 0.7586921453475952, "learning_rate": 0.0002, "epoch": 2.779080106701156, "step": 17190}, {"loss": 0.6427, "grad_norm": 0.7161715030670166, "learning_rate": 0.0002, "epoch": 2.780696790881901, "step": 17200}, {"loss": 0.6426, "grad_norm": 0.762868344783783, "learning_rate": 0.0002, "epoch": 2.7823134750626464, "step": 17210}, {"loss": 0.705, "grad_norm": 0.9285483360290527, "learning_rate": 0.0002, "epoch": 2.7839301592433916, "step": 17220}, {"loss": 0.7084, "grad_norm": 0.6900462508201599, "learning_rate": 0.0002, "epoch": 2.785546843424137, "step": 17230}, {"loss": 0.6988, "grad_norm": 0.780384361743927, "learning_rate": 0.0002, "epoch": 2.7871635276048825, "step": 17240}, {"loss": 0.7073, "grad_norm": 0.7580406665802002, "learning_rate": 0.0002, "epoch": 2.7887802117856277, "step": 17250}, {"loss": 0.6833, "grad_norm": 0.8145199418067932, "learning_rate": 0.0002, "epoch": 2.790396895966373, "step": 17260}, {"loss": 0.6909, "grad_norm": 0.9159596562385559, "learning_rate": 0.0002, "epoch": 2.792013580147118, "step": 17270}, {"loss": 0.6008, "grad_norm": 0.9590014219284058, "learning_rate": 0.0002, "epoch": 2.7936302643278634, "step": 17280}, {"loss": 0.6704, "grad_norm": 0.7603529691696167, "learning_rate": 0.0002, "epoch": 2.795246948508609, "step": 17290}, {"loss": 0.7165, "grad_norm": 0.8039976358413696, "learning_rate": 0.0002, "epoch": 2.7968636326893543, "step": 17300}, {"loss": 0.7037, "grad_norm": 0.8364847302436829, "learning_rate": 0.0002, "epoch": 2.7984803168700996, "step": 17310}, {"loss": 0.6749, "grad_norm": 0.8763046860694885, "learning_rate": 0.0002, "epoch": 2.800097001050845, "step": 17320}, {"loss": 0.6844, "grad_norm": 0.8409647941589355, "learning_rate": 0.0002, "epoch": 2.80171368523159, "step": 17330}, {"loss": 0.6936, "grad_norm": 0.7649006247520447, "learning_rate": 0.0002, "epoch": 2.8033303694123353, "step": 17340}, {"loss": 0.7051, "grad_norm": 0.7970262169837952, "learning_rate": 0.0002, "epoch": 2.8049470535930805, "step": 17350}, {"loss": 0.6533, "grad_norm": 0.9088607430458069, "learning_rate": 0.0002, "epoch": 2.8065637377738257, "step": 17360}, {"loss": 0.675, "grad_norm": 0.6454846858978271, "learning_rate": 0.0002, "epoch": 2.808180421954571, "step": 17370}, {"loss": 0.7069, "grad_norm": 0.7744787931442261, "learning_rate": 0.0002, "epoch": 2.809797106135316, "step": 17380}, {"loss": 0.6772, "grad_norm": 0.6678640842437744, "learning_rate": 0.0002, "epoch": 2.811413790316062, "step": 17390}, {"loss": 0.6784, "grad_norm": 0.772676944732666, "learning_rate": 0.0002, "epoch": 2.813030474496807, "step": 17400}, {"loss": 0.7252, "grad_norm": 0.7088175415992737, "learning_rate": 0.0002, "epoch": 2.8146471586775523, "step": 17410}, {"loss": 0.7086, "grad_norm": 0.8280573487281799, "learning_rate": 0.0002, "epoch": 2.8162638428582976, "step": 17420}, {"loss": 0.6732, "grad_norm": 0.6665388345718384, "learning_rate": 0.0002, "epoch": 2.817880527039043, "step": 17430}, {"loss": 0.6675, "grad_norm": 0.6427883505821228, "learning_rate": 0.0002, "epoch": 2.8194972112197885, "step": 17440}, {"loss": 0.6972, "grad_norm": 0.9697760343551636, "learning_rate": 0.0002, "epoch": 2.8211138954005337, "step": 17450}, {"loss": 0.6838, "grad_norm": 0.7573966383934021, "learning_rate": 0.0002, "epoch": 2.822730579581279, "step": 17460}, {"loss": 0.7243, "grad_norm": 0.878688633441925, "learning_rate": 0.0002, "epoch": 2.824347263762024, "step": 17470}, {"loss": 0.6666, "grad_norm": 0.7752242684364319, "learning_rate": 0.0002, "epoch": 2.8259639479427694, "step": 17480}, {"loss": 0.6638, "grad_norm": 0.6135398745536804, "learning_rate": 0.0002, "epoch": 2.8275806321235146, "step": 17490}, {"loss": 0.6829, "grad_norm": 0.6924924850463867, "learning_rate": 0.0002, "epoch": 2.82919731630426, "step": 17500}, {"loss": 0.6731, "grad_norm": 0.7471627593040466, "learning_rate": 0.0002, "epoch": 2.830814000485005, "step": 17510}, {"loss": 0.7016, "grad_norm": 0.7145499587059021, "learning_rate": 0.0002, "epoch": 2.8324306846657503, "step": 17520}, {"loss": 0.6787, "grad_norm": 0.7415414452552795, "learning_rate": 0.0002, "epoch": 2.834047368846496, "step": 17530}, {"loss": 0.6811, "grad_norm": 0.7328441739082336, "learning_rate": 0.0002, "epoch": 2.8356640530272412, "step": 17540}, {"loss": 0.6866, "grad_norm": 0.8267839550971985, "learning_rate": 0.0002, "epoch": 2.8372807372079865, "step": 17550}, {"loss": 0.6787, "grad_norm": 0.8877885341644287, "learning_rate": 0.0002, "epoch": 2.8388974213887317, "step": 17560}, {"loss": 0.7136, "grad_norm": 0.857138454914093, "learning_rate": 0.0002, "epoch": 2.840514105569477, "step": 17570}, {"loss": 0.6454, "grad_norm": 0.8470779657363892, "learning_rate": 0.0002, "epoch": 2.842130789750222, "step": 17580}, {"loss": 0.6976, "grad_norm": 0.8553254008293152, "learning_rate": 0.0002, "epoch": 2.843747473930968, "step": 17590}, {"loss": 0.7297, "grad_norm": 0.8033196926116943, "learning_rate": 0.0002, "epoch": 2.845364158111713, "step": 17600}, {"loss": 0.7062, "grad_norm": 0.7949087023735046, "learning_rate": 0.0002, "epoch": 2.8469808422924583, "step": 17610}, {"loss": 0.651, "grad_norm": 0.9241406321525574, "learning_rate": 0.0002, "epoch": 2.8485975264732035, "step": 17620}, {"loss": 0.6601, "grad_norm": 0.7721285223960876, "learning_rate": 0.0002, "epoch": 2.8502142106539488, "step": 17630}, {"loss": 0.6183, "grad_norm": 1.0246692895889282, "learning_rate": 0.0002, "epoch": 2.851830894834694, "step": 17640}, {"loss": 0.7007, "grad_norm": 0.9244589805603027, "learning_rate": 0.0002, "epoch": 2.853447579015439, "step": 17650}, {"loss": 0.7274, "grad_norm": 0.7243508696556091, "learning_rate": 0.0002, "epoch": 2.8550642631961844, "step": 17660}, {"loss": 0.6471, "grad_norm": 0.8943371176719666, "learning_rate": 0.0002, "epoch": 2.8566809473769297, "step": 17670}, {"loss": 0.686, "grad_norm": 0.6531758904457092, "learning_rate": 0.0002, "epoch": 2.8582976315576754, "step": 17680}, {"loss": 0.6253, "grad_norm": 0.8367000818252563, "learning_rate": 0.0002, "epoch": 2.8599143157384206, "step": 17690}, {"loss": 0.6943, "grad_norm": 0.7868556380271912, "learning_rate": 0.0002, "epoch": 2.861530999919166, "step": 17700}, {"loss": 0.6919, "grad_norm": 0.7213859558105469, "learning_rate": 0.0002, "epoch": 2.863147684099911, "step": 17710}, {"loss": 0.6657, "grad_norm": 0.7383931279182434, "learning_rate": 0.0002, "epoch": 2.8647643682806563, "step": 17720}, {"loss": 0.6841, "grad_norm": 0.7566812634468079, "learning_rate": 0.0002, "epoch": 2.8663810524614015, "step": 17730}, {"loss": 0.6449, "grad_norm": 0.6930373311042786, "learning_rate": 0.0002, "epoch": 2.867997736642147, "step": 17740}, {"loss": 0.6764, "grad_norm": 0.7911090850830078, "learning_rate": 0.0002, "epoch": 2.8696144208228924, "step": 17750}, {"loss": 0.6554, "grad_norm": 0.8484548926353455, "learning_rate": 0.0002, "epoch": 2.8712311050036377, "step": 17760}, {"loss": 0.6931, "grad_norm": 0.7647597193717957, "learning_rate": 0.0002, "epoch": 2.872847789184383, "step": 17770}, {"loss": 0.6945, "grad_norm": 0.8791151642799377, "learning_rate": 0.0002, "epoch": 2.874464473365128, "step": 17780}, {"loss": 0.7078, "grad_norm": 0.7253178358078003, "learning_rate": 0.0002, "epoch": 2.8760811575458733, "step": 17790}, {"loss": 0.6474, "grad_norm": 0.7956077456474304, "learning_rate": 0.0002, "epoch": 2.8776978417266186, "step": 17800}, {"loss": 0.6687, "grad_norm": 0.8657688498497009, "learning_rate": 0.0002, "epoch": 2.879314525907364, "step": 17810}, {"loss": 0.7171, "grad_norm": 0.7059141993522644, "learning_rate": 0.0002, "epoch": 2.880931210088109, "step": 17820}, {"loss": 0.683, "grad_norm": 0.8886896967887878, "learning_rate": 0.0002, "epoch": 2.8825478942688547, "step": 17830}, {"loss": 0.669, "grad_norm": 0.821032702922821, "learning_rate": 0.0002, "epoch": 2.8841645784496, "step": 17840}, {"loss": 0.6805, "grad_norm": 0.7183963656425476, "learning_rate": 0.0002, "epoch": 2.885781262630345, "step": 17850}, {"loss": 0.7088, "grad_norm": 0.6222899556159973, "learning_rate": 0.0002, "epoch": 2.8873979468110904, "step": 17860}, {"loss": 0.6626, "grad_norm": 0.8187434077262878, "learning_rate": 0.0002, "epoch": 2.8890146309918356, "step": 17870}, {"loss": 0.6815, "grad_norm": 0.9838479161262512, "learning_rate": 0.0002, "epoch": 2.890631315172581, "step": 17880}, {"loss": 0.6967, "grad_norm": 0.7567742466926575, "learning_rate": 0.0002, "epoch": 2.8922479993533265, "step": 17890}, {"loss": 0.7073, "grad_norm": 0.6875903606414795, "learning_rate": 0.0002, "epoch": 2.893864683534072, "step": 17900}, {"loss": 0.6415, "grad_norm": 0.8043789267539978, "learning_rate": 0.0002, "epoch": 2.895481367714817, "step": 17910}, {"loss": 0.6588, "grad_norm": 0.8062626719474792, "learning_rate": 0.0002, "epoch": 2.8970980518955622, "step": 17920}, {"loss": 0.7151, "grad_norm": 1.0251191854476929, "learning_rate": 0.0002, "epoch": 2.8987147360763075, "step": 17930}, {"loss": 0.6605, "grad_norm": 0.882253110408783, "learning_rate": 0.0002, "epoch": 2.9003314202570527, "step": 17940}, {"loss": 0.6719, "grad_norm": 0.8683299422264099, "learning_rate": 0.0002, "epoch": 2.901948104437798, "step": 17950}, {"loss": 0.6896, "grad_norm": 0.7167282104492188, "learning_rate": 0.0002, "epoch": 2.903564788618543, "step": 17960}, {"loss": 0.663, "grad_norm": 0.7093694806098938, "learning_rate": 0.0002, "epoch": 2.9051814727992884, "step": 17970}, {"loss": 0.6591, "grad_norm": 0.8549879193305969, "learning_rate": 0.0002, "epoch": 2.906798156980034, "step": 17980}, {"loss": 0.6962, "grad_norm": 0.6989606618881226, "learning_rate": 0.0002, "epoch": 2.9084148411607793, "step": 17990}, {"loss": 0.6635, "grad_norm": 0.9482976794242859, "learning_rate": 0.0002, "epoch": 2.9100315253415245, "step": 18000}, {"loss": 0.6586, "grad_norm": 0.7182440161705017, "learning_rate": 0.0002, "epoch": 2.9116482095222698, "step": 18010}, {"loss": 0.6827, "grad_norm": 0.7732226252555847, "learning_rate": 0.0002, "epoch": 2.913264893703015, "step": 18020}, {"loss": 0.7123, "grad_norm": 0.7936875224113464, "learning_rate": 0.0002, "epoch": 2.9148815778837607, "step": 18030}, {"loss": 0.6736, "grad_norm": 0.8825615644454956, "learning_rate": 0.0002, "epoch": 2.916498262064506, "step": 18040}, {"loss": 0.7139, "grad_norm": 0.6778587102890015, "learning_rate": 0.0002, "epoch": 2.918114946245251, "step": 18050}, {"loss": 0.6588, "grad_norm": 0.7529265880584717, "learning_rate": 0.0002, "epoch": 2.9197316304259964, "step": 18060}, {"loss": 0.737, "grad_norm": 0.7111883163452148, "learning_rate": 0.0002, "epoch": 2.9213483146067416, "step": 18070}, {"loss": 0.7475, "grad_norm": 0.7214767932891846, "learning_rate": 0.0002, "epoch": 2.922964998787487, "step": 18080}, {"loss": 0.6672, "grad_norm": 0.800417423248291, "learning_rate": 0.0002, "epoch": 2.924581682968232, "step": 18090}, {"loss": 0.6694, "grad_norm": 1.248575210571289, "learning_rate": 0.0002, "epoch": 2.9261983671489773, "step": 18100}, {"loss": 0.7004, "grad_norm": 0.757788360118866, "learning_rate": 0.0002, "epoch": 2.9278150513297225, "step": 18110}, {"loss": 0.6999, "grad_norm": 1.0583995580673218, "learning_rate": 0.0002, "epoch": 2.9294317355104678, "step": 18120}, {"loss": 0.6365, "grad_norm": 0.8228777647018433, "learning_rate": 0.0002, "epoch": 2.9310484196912134, "step": 18130}, {"loss": 0.6791, "grad_norm": 0.8374035358428955, "learning_rate": 0.0002, "epoch": 2.9326651038719587, "step": 18140}, {"loss": 0.6399, "grad_norm": 0.7976473569869995, "learning_rate": 0.0002, "epoch": 2.934281788052704, "step": 18150}, {"loss": 0.6585, "grad_norm": 0.8009907603263855, "learning_rate": 0.0002, "epoch": 2.935898472233449, "step": 18160}, {"loss": 0.7485, "grad_norm": 0.835213303565979, "learning_rate": 0.0002, "epoch": 2.9375151564141944, "step": 18170}, {"loss": 0.7376, "grad_norm": 0.7982219457626343, "learning_rate": 0.0002, "epoch": 2.93913184059494, "step": 18180}, {"loss": 0.6348, "grad_norm": 0.7070978879928589, "learning_rate": 0.0002, "epoch": 2.9407485247756853, "step": 18190}, {"loss": 0.6608, "grad_norm": 0.8619440197944641, "learning_rate": 0.0002, "epoch": 2.9423652089564305, "step": 18200}, {"loss": 0.666, "grad_norm": 0.6693987250328064, "learning_rate": 0.0002, "epoch": 2.9439818931371757, "step": 18210}, {"loss": 0.728, "grad_norm": 0.6747021079063416, "learning_rate": 0.0002, "epoch": 2.945598577317921, "step": 18220}, {"loss": 0.6686, "grad_norm": 0.860387921333313, "learning_rate": 0.0002, "epoch": 2.947215261498666, "step": 18230}, {"loss": 0.6945, "grad_norm": 0.799976646900177, "learning_rate": 0.0002, "epoch": 2.9488319456794114, "step": 18240}, {"loss": 0.7243, "grad_norm": 0.7864769101142883, "learning_rate": 0.0002, "epoch": 2.9504486298601567, "step": 18250}, {"loss": 0.6785, "grad_norm": 0.6713884472846985, "learning_rate": 0.0002, "epoch": 2.952065314040902, "step": 18260}, {"loss": 0.7429, "grad_norm": 0.9031508564949036, "learning_rate": 0.0002, "epoch": 2.9536819982216476, "step": 18270}, {"loss": 0.7055, "grad_norm": 0.7205073237419128, "learning_rate": 0.0002, "epoch": 2.955298682402393, "step": 18280}, {"loss": 0.7298, "grad_norm": 0.7746205925941467, "learning_rate": 0.0002, "epoch": 2.956915366583138, "step": 18290}, {"loss": 0.6218, "grad_norm": 0.6533427834510803, "learning_rate": 0.0002, "epoch": 2.9585320507638833, "step": 18300}, {"loss": 0.6674, "grad_norm": 0.9083208441734314, "learning_rate": 0.0002, "epoch": 2.9601487349446285, "step": 18310}, {"loss": 0.7359, "grad_norm": 0.7446991801261902, "learning_rate": 0.0002, "epoch": 2.9617654191253737, "step": 18320}, {"loss": 0.6738, "grad_norm": 0.6514461636543274, "learning_rate": 0.0002, "epoch": 2.9633821033061194, "step": 18330}, {"loss": 0.6677, "grad_norm": 0.8580465912818909, "learning_rate": 0.0002, "epoch": 2.9649987874868646, "step": 18340}, {"loss": 0.6971, "grad_norm": 0.7074266076087952, "learning_rate": 0.0002, "epoch": 2.96661547166761, "step": 18350}, {"loss": 0.6804, "grad_norm": 0.899892270565033, "learning_rate": 0.0002, "epoch": 2.968232155848355, "step": 18360}, {"loss": 0.7094, "grad_norm": 0.8217641711235046, "learning_rate": 0.0002, "epoch": 2.9698488400291003, "step": 18370}, {"loss": 0.6916, "grad_norm": 0.8611799478530884, "learning_rate": 0.0002, "epoch": 2.9714655242098456, "step": 18380}, {"loss": 0.6677, "grad_norm": 0.6909302473068237, "learning_rate": 0.0002, "epoch": 2.973082208390591, "step": 18390}, {"loss": 0.7247, "grad_norm": 0.6554358005523682, "learning_rate": 0.0002, "epoch": 2.974698892571336, "step": 18400}, {"loss": 0.6516, "grad_norm": 0.7803071737289429, "learning_rate": 0.0002, "epoch": 2.9763155767520812, "step": 18410}, {"loss": 0.7322, "grad_norm": 0.7838954925537109, "learning_rate": 0.0002, "epoch": 2.977932260932827, "step": 18420}, {"loss": 0.6522, "grad_norm": 0.7098495364189148, "learning_rate": 0.0002, "epoch": 2.979548945113572, "step": 18430}, {"loss": 0.739, "grad_norm": 0.8981785774230957, "learning_rate": 0.0002, "epoch": 2.9811656292943174, "step": 18440}, {"loss": 0.6689, "grad_norm": 0.7197171449661255, "learning_rate": 0.0002, "epoch": 2.9827823134750626, "step": 18450}, {"loss": 0.706, "grad_norm": 0.793185293674469, "learning_rate": 0.0002, "epoch": 2.984398997655808, "step": 18460}, {"loss": 0.7124, "grad_norm": 0.8531473875045776, "learning_rate": 0.0002, "epoch": 2.986015681836553, "step": 18470}, {"loss": 0.6901, "grad_norm": 0.6627361178398132, "learning_rate": 0.0002, "epoch": 2.9876323660172988, "step": 18480}, {"loss": 0.6591, "grad_norm": 0.5708155035972595, "learning_rate": 0.0002, "epoch": 2.989249050198044, "step": 18490}, {"loss": 0.6725, "grad_norm": 0.8227280378341675, "learning_rate": 0.0002, "epoch": 2.990865734378789, "step": 18500}, {"loss": 0.6701, "grad_norm": 0.7102749943733215, "learning_rate": 0.0002, "epoch": 2.9924824185595345, "step": 18510}, {"loss": 0.7091, "grad_norm": 0.839485228061676, "learning_rate": 0.0002, "epoch": 2.9940991027402797, "step": 18520}, {"loss": 0.6521, "grad_norm": 0.9038704037666321, "learning_rate": 0.0002, "epoch": 2.995715786921025, "step": 18530}, {"loss": 0.7186, "grad_norm": 0.8737510442733765, "learning_rate": 0.0002, "epoch": 2.99733247110177, "step": 18540}, {"loss": 0.6819, "grad_norm": 0.7323142886161804, "learning_rate": 0.0002, "epoch": 2.9989491552825154, "step": 18550}, {"eval_loss": 1.1262480020523071, "eval_runtime": 122.0868, "eval_samples_per_second": 6.004, "eval_steps_per_second": 0.754, "epoch": 2.9999191657909625, "step": 18556}, {"loss": 0.6337, "grad_norm": 0.8465463519096375, "learning_rate": 0.0002, "epoch": 3.000565839463261, "step": 18560}, {"loss": 0.6064, "grad_norm": 0.9134138822555542, "learning_rate": 0.0002, "epoch": 3.0021825236440063, "step": 18570}, {"loss": 0.5804, "grad_norm": 0.760715126991272, "learning_rate": 0.0002, "epoch": 3.0037992078247515, "step": 18580}, {"loss": 0.5571, "grad_norm": 0.9208743572235107, "learning_rate": 0.0002, "epoch": 3.0054158920054967, "step": 18590}, {"loss": 0.5731, "grad_norm": 0.9232364892959595, "learning_rate": 0.0002, "epoch": 3.007032576186242, "step": 18600}, {"loss": 0.6299, "grad_norm": 1.1881544589996338, "learning_rate": 0.0002, "epoch": 3.008649260366987, "step": 18610}, {"loss": 0.5482, "grad_norm": 0.9372987747192383, "learning_rate": 0.0002, "epoch": 3.0102659445477324, "step": 18620}, {"loss": 0.5709, "grad_norm": 0.6900241374969482, "learning_rate": 0.0002, "epoch": 3.0118826287284777, "step": 18630}, {"loss": 0.5256, "grad_norm": 0.8451071381568909, "learning_rate": 0.0002, "epoch": 3.0134993129092233, "step": 18640}, {"loss": 0.5916, "grad_norm": 0.7763112187385559, "learning_rate": 0.0002, "epoch": 3.0151159970899686, "step": 18650}, {"loss": 0.6095, "grad_norm": 1.043653964996338, "learning_rate": 0.0002, "epoch": 3.016732681270714, "step": 18660}, {"loss": 0.6228, "grad_norm": 1.0170660018920898, "learning_rate": 0.0002, "epoch": 3.018349365451459, "step": 18670}, {"loss": 0.5671, "grad_norm": 0.7534180283546448, "learning_rate": 0.0002, "epoch": 3.0199660496322043, "step": 18680}, {"loss": 0.6015, "grad_norm": 0.7507367730140686, "learning_rate": 0.0002, "epoch": 3.0215827338129495, "step": 18690}, {"loss": 0.6201, "grad_norm": 0.7861620187759399, "learning_rate": 0.0002, "epoch": 3.0231994179936947, "step": 18700}, {"loss": 0.5802, "grad_norm": 1.0580339431762695, "learning_rate": 0.0002, "epoch": 3.0248161021744404, "step": 18710}, {"loss": 0.5975, "grad_norm": 0.7542710900306702, "learning_rate": 0.0002, "epoch": 3.0264327863551856, "step": 18720}, {"loss": 0.5695, "grad_norm": 0.8189544677734375, "learning_rate": 0.0002, "epoch": 3.028049470535931, "step": 18730}, {"loss": 0.6109, "grad_norm": 0.9126611351966858, "learning_rate": 0.0002, "epoch": 3.029666154716676, "step": 18740}, {"loss": 0.6443, "grad_norm": 0.8891341686248779, "learning_rate": 0.0002, "epoch": 3.0312828388974213, "step": 18750}, {"loss": 0.6207, "grad_norm": 0.8419283032417297, "learning_rate": 0.0002, "epoch": 3.0328995230781666, "step": 18760}, {"loss": 0.5818, "grad_norm": 0.8048048615455627, "learning_rate": 0.0002, "epoch": 3.034516207258912, "step": 18770}, {"loss": 0.6381, "grad_norm": 0.7820217609405518, "learning_rate": 0.0002, "epoch": 3.0361328914396575, "step": 18780}, {"loss": 0.5843, "grad_norm": 0.854721188545227, "learning_rate": 0.0002, "epoch": 3.0377495756204027, "step": 18790}, {"loss": 0.5784, "grad_norm": 0.912092924118042, "learning_rate": 0.0002, "epoch": 3.039366259801148, "step": 18800}, {"loss": 0.5734, "grad_norm": 0.6596226096153259, "learning_rate": 0.0002, "epoch": 3.040982943981893, "step": 18810}, {"loss": 0.5969, "grad_norm": 0.6351348757743835, "learning_rate": 0.0002, "epoch": 3.0425996281626384, "step": 18820}, {"loss": 0.5953, "grad_norm": 0.778188943862915, "learning_rate": 0.0002, "epoch": 3.0442163123433836, "step": 18830}, {"loss": 0.602, "grad_norm": 0.68234783411026, "learning_rate": 0.0002, "epoch": 3.045832996524129, "step": 18840}, {"loss": 0.5785, "grad_norm": 0.998628556728363, "learning_rate": 0.0002, "epoch": 3.047449680704874, "step": 18850}, {"loss": 0.6231, "grad_norm": 0.7393841743469238, "learning_rate": 0.0002, "epoch": 3.0490663648856198, "step": 18860}, {"loss": 0.568, "grad_norm": 0.84438556432724, "learning_rate": 0.0002, "epoch": 3.050683049066365, "step": 18870}, {"loss": 0.6205, "grad_norm": 0.8857501745223999, "learning_rate": 0.0002, "epoch": 3.0522997332471102, "step": 18880}, {"loss": 0.6335, "grad_norm": 0.7208474278450012, "learning_rate": 0.0002, "epoch": 3.0539164174278555, "step": 18890}, {"loss": 0.5998, "grad_norm": 0.7135229110717773, "learning_rate": 0.0002, "epoch": 3.0555331016086007, "step": 18900}, {"loss": 0.5575, "grad_norm": 0.9130001664161682, "learning_rate": 0.0002, "epoch": 3.057149785789346, "step": 18910}, {"loss": 0.5955, "grad_norm": 0.9001716375350952, "learning_rate": 0.0002, "epoch": 3.058766469970091, "step": 18920}, {"loss": 0.6052, "grad_norm": 0.8667559623718262, "learning_rate": 0.0002, "epoch": 3.060383154150837, "step": 18930}, {"loss": 0.5818, "grad_norm": 0.8943959474563599, "learning_rate": 0.0002, "epoch": 3.061999838331582, "step": 18940}, {"loss": 0.5978, "grad_norm": 0.8298377990722656, "learning_rate": 0.0002, "epoch": 3.0636165225123273, "step": 18950}, {"loss": 0.5782, "grad_norm": 0.7935267686843872, "learning_rate": 0.0002, "epoch": 3.0652332066930725, "step": 18960}, {"loss": 0.6434, "grad_norm": 1.1506379842758179, "learning_rate": 0.0002, "epoch": 3.0668498908738178, "step": 18970}, {"loss": 0.5571, "grad_norm": 0.7693049907684326, "learning_rate": 0.0002, "epoch": 3.068466575054563, "step": 18980}, {"loss": 0.5971, "grad_norm": 0.8040135502815247, "learning_rate": 0.0002, "epoch": 3.0700832592353082, "step": 18990}, {"loss": 0.5541, "grad_norm": 0.828404426574707, "learning_rate": 0.0002, "epoch": 3.0716999434160535, "step": 19000}, {"loss": 0.6048, "grad_norm": 0.8811164498329163, "learning_rate": 0.0002, "epoch": 3.073316627596799, "step": 19010}, {"loss": 0.5845, "grad_norm": 1.036205768585205, "learning_rate": 0.0002, "epoch": 3.0749333117775444, "step": 19020}, {"loss": 0.5838, "grad_norm": 0.8857285976409912, "learning_rate": 0.0002, "epoch": 3.0765499959582896, "step": 19030}, {"loss": 0.592, "grad_norm": 0.8392079472541809, "learning_rate": 0.0002, "epoch": 3.078166680139035, "step": 19040}, {"loss": 0.5927, "grad_norm": 1.0287401676177979, "learning_rate": 0.0002, "epoch": 3.07978336431978, "step": 19050}, {"loss": 0.5964, "grad_norm": 1.0086315870285034, "learning_rate": 0.0002, "epoch": 3.0814000485005253, "step": 19060}, {"loss": 0.5567, "grad_norm": 0.9245324730873108, "learning_rate": 0.0002, "epoch": 3.0830167326812705, "step": 19070}, {"loss": 0.5797, "grad_norm": 0.8680877089500427, "learning_rate": 0.0002, "epoch": 3.084633416862016, "step": 19080}, {"loss": 0.5611, "grad_norm": 0.8814793825149536, "learning_rate": 0.0002, "epoch": 3.0862501010427614, "step": 19090}, {"loss": 0.6051, "grad_norm": 0.9234458208084106, "learning_rate": 0.0002, "epoch": 3.0878667852235067, "step": 19100}, {"loss": 0.6209, "grad_norm": 1.1291664838790894, "learning_rate": 0.0002, "epoch": 3.089483469404252, "step": 19110}, {"loss": 0.5695, "grad_norm": 0.9191402792930603, "learning_rate": 0.0002, "epoch": 3.091100153584997, "step": 19120}, {"loss": 0.5856, "grad_norm": 0.7103154063224792, "learning_rate": 0.0002, "epoch": 3.0927168377657424, "step": 19130}, {"loss": 0.6479, "grad_norm": 0.9368883967399597, "learning_rate": 0.0002, "epoch": 3.0943335219464876, "step": 19140}, {"loss": 0.6167, "grad_norm": 0.9676656723022461, "learning_rate": 0.0002, "epoch": 3.095950206127233, "step": 19150}, {"loss": 0.5794, "grad_norm": 0.8739792704582214, "learning_rate": 0.0002, "epoch": 3.0975668903079785, "step": 19160}, {"loss": 0.6112, "grad_norm": 0.8530174493789673, "learning_rate": 0.0002, "epoch": 3.0991835744887237, "step": 19170}, {"loss": 0.6568, "grad_norm": 0.794945478439331, "learning_rate": 0.0002, "epoch": 3.100800258669469, "step": 19180}, {"loss": 0.5928, "grad_norm": 0.9508888125419617, "learning_rate": 0.0002, "epoch": 3.102416942850214, "step": 19190}, {"loss": 0.5757, "grad_norm": 1.0599955320358276, "learning_rate": 0.0002, "epoch": 3.1040336270309594, "step": 19200}, {"loss": 0.6151, "grad_norm": 1.0673625469207764, "learning_rate": 0.0002, "epoch": 3.1056503112117047, "step": 19210}, {"loss": 0.6043, "grad_norm": 0.7739115953445435, "learning_rate": 0.0002, "epoch": 3.10726699539245, "step": 19220}, {"loss": 0.6046, "grad_norm": 0.9884951114654541, "learning_rate": 0.0002, "epoch": 3.1088836795731956, "step": 19230}, {"loss": 0.5932, "grad_norm": 0.862260103225708, "learning_rate": 0.0002, "epoch": 3.110500363753941, "step": 19240}, {"loss": 0.6098, "grad_norm": 0.7690284848213196, "learning_rate": 0.0002, "epoch": 3.112117047934686, "step": 19250}, {"loss": 0.5791, "grad_norm": 0.8758958578109741, "learning_rate": 0.0002, "epoch": 3.1137337321154313, "step": 19260}, {"loss": 0.6136, "grad_norm": 1.0356395244598389, "learning_rate": 0.0002, "epoch": 3.1153504162961765, "step": 19270}, {"loss": 0.6159, "grad_norm": 0.6950937509536743, "learning_rate": 0.0002, "epoch": 3.1169671004769217, "step": 19280}, {"loss": 0.592, "grad_norm": 0.760998010635376, "learning_rate": 0.0002, "epoch": 3.118583784657667, "step": 19290}, {"loss": 0.575, "grad_norm": 0.9335789084434509, "learning_rate": 0.0002, "epoch": 3.1202004688384126, "step": 19300}, {"loss": 0.6139, "grad_norm": 0.9636204242706299, "learning_rate": 0.0002, "epoch": 3.121817153019158, "step": 19310}, {"loss": 0.6001, "grad_norm": 1.0820997953414917, "learning_rate": 0.0002, "epoch": 3.123433837199903, "step": 19320}, {"loss": 0.6542, "grad_norm": 0.7333487272262573, "learning_rate": 0.0002, "epoch": 3.1250505213806483, "step": 19330}, {"loss": 0.6178, "grad_norm": 1.0417509078979492, "learning_rate": 0.0002, "epoch": 3.1266672055613935, "step": 19340}, {"loss": 0.603, "grad_norm": 0.9267749190330505, "learning_rate": 0.0002, "epoch": 3.128283889742139, "step": 19350}, {"loss": 0.6063, "grad_norm": 0.777798593044281, "learning_rate": 0.0002, "epoch": 3.129900573922884, "step": 19360}, {"loss": 0.5913, "grad_norm": 0.8425456881523132, "learning_rate": 0.0002, "epoch": 3.1315172581036297, "step": 19370}, {"loss": 0.6042, "grad_norm": 0.9617102146148682, "learning_rate": 0.0002, "epoch": 3.133133942284375, "step": 19380}, {"loss": 0.633, "grad_norm": 1.0052828788757324, "learning_rate": 0.0002, "epoch": 3.13475062646512, "step": 19390}, {"loss": 0.5713, "grad_norm": 0.7637009024620056, "learning_rate": 0.0002, "epoch": 3.1363673106458654, "step": 19400}, {"loss": 0.5497, "grad_norm": 0.7958088517189026, "learning_rate": 0.0002, "epoch": 3.1379839948266106, "step": 19410}, {"loss": 0.6283, "grad_norm": 0.9161727428436279, "learning_rate": 0.0002, "epoch": 3.139600679007356, "step": 19420}, {"loss": 0.5638, "grad_norm": 0.8402149677276611, "learning_rate": 0.0002, "epoch": 3.141217363188101, "step": 19430}, {"loss": 0.5848, "grad_norm": 1.0056525468826294, "learning_rate": 0.0002, "epoch": 3.1428340473688463, "step": 19440}, {"loss": 0.5954, "grad_norm": 1.0129190683364868, "learning_rate": 0.0002, "epoch": 3.144450731549592, "step": 19450}, {"loss": 0.5808, "grad_norm": 0.790825366973877, "learning_rate": 0.0002, "epoch": 3.146067415730337, "step": 19460}, {"loss": 0.5607, "grad_norm": 1.441665530204773, "learning_rate": 0.0002, "epoch": 3.1476840999110824, "step": 19470}, {"loss": 0.5785, "grad_norm": 0.7846331596374512, "learning_rate": 0.0002, "epoch": 3.1493007840918277, "step": 19480}, {"loss": 0.5892, "grad_norm": 0.7915332913398743, "learning_rate": 0.0002, "epoch": 3.150917468272573, "step": 19490}, {"loss": 0.5759, "grad_norm": 0.933982253074646, "learning_rate": 0.0002, "epoch": 3.152534152453318, "step": 19500}, {"loss": 0.6206, "grad_norm": 1.038408637046814, "learning_rate": 0.0002, "epoch": 3.1541508366340634, "step": 19510}, {"loss": 0.6271, "grad_norm": 1.018935203552246, "learning_rate": 0.0002, "epoch": 3.155767520814809, "step": 19520}, {"loss": 0.6173, "grad_norm": 0.9618112444877625, "learning_rate": 0.0002, "epoch": 3.1573842049955543, "step": 19530}, {"loss": 0.5972, "grad_norm": 0.8900452852249146, "learning_rate": 0.0002, "epoch": 3.1590008891762995, "step": 19540}, {"loss": 0.5925, "grad_norm": 0.8254160284996033, "learning_rate": 0.0002, "epoch": 3.1606175733570447, "step": 19550}, {"loss": 0.625, "grad_norm": 1.004376769065857, "learning_rate": 0.0002, "epoch": 3.16223425753779, "step": 19560}, {"loss": 0.5775, "grad_norm": 1.0490446090698242, "learning_rate": 0.0002, "epoch": 3.163850941718535, "step": 19570}, {"loss": 0.5986, "grad_norm": 0.7387403845787048, "learning_rate": 0.0002, "epoch": 3.1654676258992804, "step": 19580}, {"loss": 0.5898, "grad_norm": 0.7611538171768188, "learning_rate": 0.0002, "epoch": 3.1670843100800257, "step": 19590}, {"loss": 0.5937, "grad_norm": 0.8239886164665222, "learning_rate": 0.0002, "epoch": 3.1687009942607713, "step": 19600}, {"loss": 0.6068, "grad_norm": 0.9327243566513062, "learning_rate": 0.0002, "epoch": 3.1703176784415166, "step": 19610}, {"loss": 0.572, "grad_norm": 0.9662560224533081, "learning_rate": 0.0002, "epoch": 3.171934362622262, "step": 19620}, {"loss": 0.5988, "grad_norm": 0.9183341860771179, "learning_rate": 0.0002, "epoch": 3.173551046803007, "step": 19630}, {"loss": 0.5909, "grad_norm": 0.875066876411438, "learning_rate": 0.0002, "epoch": 3.1751677309837523, "step": 19640}, {"loss": 0.5956, "grad_norm": 0.8567508459091187, "learning_rate": 0.0002, "epoch": 3.1767844151644975, "step": 19650}, {"loss": 0.5805, "grad_norm": 0.6805780529975891, "learning_rate": 0.0002, "epoch": 3.1784010993452427, "step": 19660}, {"loss": 0.6204, "grad_norm": 0.8776944279670715, "learning_rate": 0.0002, "epoch": 3.1800177835259884, "step": 19670}, {"loss": 0.6108, "grad_norm": 0.9036329984664917, "learning_rate": 0.0002, "epoch": 3.1816344677067336, "step": 19680}, {"loss": 0.6238, "grad_norm": 0.8527372479438782, "learning_rate": 0.0002, "epoch": 3.183251151887479, "step": 19690}, {"loss": 0.6089, "grad_norm": 1.1045585870742798, "learning_rate": 0.0002, "epoch": 3.184867836068224, "step": 19700}, {"loss": 0.5491, "grad_norm": 0.9213830828666687, "learning_rate": 0.0002, "epoch": 3.1864845202489693, "step": 19710}, {"loss": 0.618, "grad_norm": 0.8865814805030823, "learning_rate": 0.0002, "epoch": 3.1881012044297146, "step": 19720}, {"loss": 0.5785, "grad_norm": 0.7939388751983643, "learning_rate": 0.0002, "epoch": 3.18971788861046, "step": 19730}, {"loss": 0.5682, "grad_norm": 0.6966729760169983, "learning_rate": 0.0002, "epoch": 3.191334572791205, "step": 19740}, {"loss": 0.5839, "grad_norm": 0.8023673295974731, "learning_rate": 0.0002, "epoch": 3.1929512569719507, "step": 19750}, {"loss": 0.6267, "grad_norm": 0.7992037534713745, "learning_rate": 0.0002, "epoch": 3.194567941152696, "step": 19760}, {"loss": 0.6141, "grad_norm": 0.7412247657775879, "learning_rate": 0.0002, "epoch": 3.196184625333441, "step": 19770}, {"loss": 0.6179, "grad_norm": 0.9598729014396667, "learning_rate": 0.0002, "epoch": 3.1978013095141864, "step": 19780}, {"loss": 0.5685, "grad_norm": 0.8331366777420044, "learning_rate": 0.0002, "epoch": 3.1994179936949316, "step": 19790}, {"loss": 0.6104, "grad_norm": 0.8939169645309448, "learning_rate": 0.0002, "epoch": 3.201034677875677, "step": 19800}, {"loss": 0.6147, "grad_norm": 0.9219734072685242, "learning_rate": 0.0002, "epoch": 3.202651362056422, "step": 19810}, {"loss": 0.6051, "grad_norm": 0.869490385055542, "learning_rate": 0.0002, "epoch": 3.2042680462371678, "step": 19820}, {"loss": 0.5946, "grad_norm": 0.8989706635475159, "learning_rate": 0.0002, "epoch": 3.205884730417913, "step": 19830}, {"loss": 0.5866, "grad_norm": 0.8477165102958679, "learning_rate": 0.0002, "epoch": 3.2075014145986582, "step": 19840}, {"loss": 0.6176, "grad_norm": 0.8720678687095642, "learning_rate": 0.0002, "epoch": 3.2091180987794035, "step": 19850}, {"loss": 0.5694, "grad_norm": 0.861406683921814, "learning_rate": 0.0002, "epoch": 3.2107347829601487, "step": 19860}, {"loss": 0.6264, "grad_norm": 0.8228686451911926, "learning_rate": 0.0002, "epoch": 3.212351467140894, "step": 19870}, {"loss": 0.625, "grad_norm": 0.7936596870422363, "learning_rate": 0.0002, "epoch": 3.213968151321639, "step": 19880}, {"loss": 0.5698, "grad_norm": 1.097377896308899, "learning_rate": 0.0002, "epoch": 3.2155848355023844, "step": 19890}, {"loss": 0.6725, "grad_norm": 0.9544782638549805, "learning_rate": 0.0002, "epoch": 3.21720151968313, "step": 19900}, {"loss": 0.6022, "grad_norm": 0.8240751624107361, "learning_rate": 0.0002, "epoch": 3.2188182038638753, "step": 19910}, {"loss": 0.5659, "grad_norm": 0.8332096338272095, "learning_rate": 0.0002, "epoch": 3.2204348880446205, "step": 19920}, {"loss": 0.6274, "grad_norm": 1.0954567193984985, "learning_rate": 0.0002, "epoch": 3.2220515722253658, "step": 19930}, {"loss": 0.652, "grad_norm": 0.7790525555610657, "learning_rate": 0.0002, "epoch": 3.223668256406111, "step": 19940}, {"loss": 0.5986, "grad_norm": 0.7966814041137695, "learning_rate": 0.0002, "epoch": 3.225284940586856, "step": 19950}, {"loss": 0.5911, "grad_norm": 0.9751881957054138, "learning_rate": 0.0002, "epoch": 3.2269016247676015, "step": 19960}, {"loss": 0.6071, "grad_norm": 0.9856047630310059, "learning_rate": 0.0002, "epoch": 3.228518308948347, "step": 19970}, {"loss": 0.5837, "grad_norm": 1.3062353134155273, "learning_rate": 0.0002, "epoch": 3.2301349931290924, "step": 19980}, {"loss": 0.6588, "grad_norm": 0.9510692358016968, "learning_rate": 0.0002, "epoch": 3.2317516773098376, "step": 19990}, {"loss": 0.6264, "grad_norm": 0.8630342483520508, "learning_rate": 0.0002, "epoch": 3.233368361490583, "step": 20000}, {"loss": 0.6073, "grad_norm": 0.8966519236564636, "learning_rate": 0.0002, "epoch": 3.234985045671328, "step": 20010}, {"loss": 0.612, "grad_norm": 0.7093510627746582, "learning_rate": 0.0002, "epoch": 3.2366017298520733, "step": 20020}, {"loss": 0.585, "grad_norm": 0.7771096229553223, "learning_rate": 0.0002, "epoch": 3.2382184140328185, "step": 20030}, {"loss": 0.5821, "grad_norm": 0.841058075428009, "learning_rate": 0.0002, "epoch": 3.2398350982135637, "step": 20040}, {"loss": 0.6519, "grad_norm": 0.909712553024292, "learning_rate": 0.0002, "epoch": 3.2414517823943094, "step": 20050}, {"loss": 0.6089, "grad_norm": 0.8321019411087036, "learning_rate": 0.0002, "epoch": 3.2430684665750547, "step": 20060}, {"loss": 0.6115, "grad_norm": 0.779901921749115, "learning_rate": 0.0002, "epoch": 3.2446851507558, "step": 20070}, {"loss": 0.6107, "grad_norm": 0.6249170303344727, "learning_rate": 0.0002, "epoch": 3.246301834936545, "step": 20080}, {"loss": 0.603, "grad_norm": 0.8000940680503845, "learning_rate": 0.0002, "epoch": 3.2479185191172903, "step": 20090}, {"loss": 0.6273, "grad_norm": 0.7627735137939453, "learning_rate": 0.0002, "epoch": 3.2495352032980356, "step": 20100}, {"loss": 0.6223, "grad_norm": 0.8780747056007385, "learning_rate": 0.0002, "epoch": 3.2511518874787813, "step": 20110}, {"loss": 0.5969, "grad_norm": 0.772037148475647, "learning_rate": 0.0002, "epoch": 3.2527685716595265, "step": 20120}, {"loss": 0.5843, "grad_norm": 1.0086580514907837, "learning_rate": 0.0002, "epoch": 3.2543852558402717, "step": 20130}, {"loss": 0.5777, "grad_norm": 0.9360289573669434, "learning_rate": 0.0002, "epoch": 3.256001940021017, "step": 20140}, {"loss": 0.5777, "grad_norm": 1.2099586725234985, "learning_rate": 0.0002, "epoch": 3.257618624201762, "step": 20150}, {"loss": 0.624, "grad_norm": 0.8368481397628784, "learning_rate": 0.0002, "epoch": 3.2592353083825074, "step": 20160}, {"loss": 0.5626, "grad_norm": 0.7391039133071899, "learning_rate": 0.0002, "epoch": 3.2608519925632526, "step": 20170}, {"loss": 0.6041, "grad_norm": 0.9122273325920105, "learning_rate": 0.0002, "epoch": 3.262468676743998, "step": 20180}, {"loss": 0.5868, "grad_norm": 0.8502281904220581, "learning_rate": 0.0002, "epoch": 3.264085360924743, "step": 20190}, {"loss": 0.5841, "grad_norm": 1.0926852226257324, "learning_rate": 0.0002, "epoch": 3.265702045105489, "step": 20200}, {"loss": 0.6027, "grad_norm": 0.7902828454971313, "learning_rate": 0.0002, "epoch": 3.267318729286234, "step": 20210}, {"loss": 0.6089, "grad_norm": 0.8724729418754578, "learning_rate": 0.0002, "epoch": 3.2689354134669792, "step": 20220}, {"loss": 0.6242, "grad_norm": 0.8469277024269104, "learning_rate": 0.0002, "epoch": 3.2705520976477245, "step": 20230}, {"loss": 0.644, "grad_norm": 0.8865092992782593, "learning_rate": 0.0002, "epoch": 3.2721687818284697, "step": 20240}, {"loss": 0.6464, "grad_norm": 1.0979334115982056, "learning_rate": 0.0002, "epoch": 3.273785466009215, "step": 20250}, {"loss": 0.647, "grad_norm": 1.0860793590545654, "learning_rate": 0.0002, "epoch": 3.2754021501899606, "step": 20260}, {"loss": 0.6105, "grad_norm": 0.981745183467865, "learning_rate": 0.0002, "epoch": 3.277018834370706, "step": 20270}, {"loss": 0.627, "grad_norm": 0.9155020713806152, "learning_rate": 0.0002, "epoch": 3.278635518551451, "step": 20280}, {"loss": 0.5899, "grad_norm": 0.8436718583106995, "learning_rate": 0.0002, "epoch": 3.2802522027321963, "step": 20290}, {"loss": 0.6371, "grad_norm": 1.0329409837722778, "learning_rate": 0.0002, "epoch": 3.2818688869129415, "step": 20300}, {"loss": 0.6, "grad_norm": 0.9876394271850586, "learning_rate": 0.0002, "epoch": 3.2834855710936868, "step": 20310}, {"loss": 0.5463, "grad_norm": 0.8052917718887329, "learning_rate": 0.0002, "epoch": 3.285102255274432, "step": 20320}, {"loss": 0.5949, "grad_norm": 0.8390680551528931, "learning_rate": 0.0002, "epoch": 3.2867189394551772, "step": 20330}, {"loss": 0.6492, "grad_norm": 0.9515735507011414, "learning_rate": 0.0002, "epoch": 3.288335623635923, "step": 20340}, {"loss": 0.596, "grad_norm": 0.8028870224952698, "learning_rate": 0.0002, "epoch": 3.289952307816668, "step": 20350}, {"loss": 0.634, "grad_norm": 0.862592339515686, "learning_rate": 0.0002, "epoch": 3.2915689919974134, "step": 20360}, {"loss": 0.6345, "grad_norm": 0.7451621890068054, "learning_rate": 0.0002, "epoch": 3.2931856761781586, "step": 20370}, {"loss": 0.6458, "grad_norm": 0.8966776728630066, "learning_rate": 0.0002, "epoch": 3.294802360358904, "step": 20380}, {"loss": 0.5967, "grad_norm": 0.9289216995239258, "learning_rate": 0.0002, "epoch": 3.296419044539649, "step": 20390}, {"loss": 0.6599, "grad_norm": 0.9649626612663269, "learning_rate": 0.0002, "epoch": 3.2980357287203943, "step": 20400}, {"loss": 0.5781, "grad_norm": 1.1953798532485962, "learning_rate": 0.0002, "epoch": 3.29965241290114, "step": 20410}, {"loss": 0.5997, "grad_norm": 0.8929083943367004, "learning_rate": 0.0002, "epoch": 3.301269097081885, "step": 20420}, {"loss": 0.597, "grad_norm": 0.8922014236450195, "learning_rate": 0.0002, "epoch": 3.3028857812626304, "step": 20430}, {"loss": 0.5766, "grad_norm": 0.9754860401153564, "learning_rate": 0.0002, "epoch": 3.3045024654433757, "step": 20440}, {"loss": 0.5653, "grad_norm": 0.8873140215873718, "learning_rate": 0.0002, "epoch": 3.306119149624121, "step": 20450}, {"loss": 0.6138, "grad_norm": 0.857271671295166, "learning_rate": 0.0002, "epoch": 3.307735833804866, "step": 20460}, {"loss": 0.633, "grad_norm": 0.9022141098976135, "learning_rate": 0.0002, "epoch": 3.3093525179856114, "step": 20470}, {"loss": 0.6654, "grad_norm": 0.8614798188209534, "learning_rate": 0.0002, "epoch": 3.3109692021663566, "step": 20480}, {"loss": 0.6254, "grad_norm": 0.8838164210319519, "learning_rate": 0.0002, "epoch": 3.3125858863471023, "step": 20490}, {"loss": 0.5849, "grad_norm": 0.8709736466407776, "learning_rate": 0.0002, "epoch": 3.3142025705278475, "step": 20500}, {"loss": 0.6146, "grad_norm": 0.9533300995826721, "learning_rate": 0.0002, "epoch": 3.3158192547085927, "step": 20510}, {"loss": 0.6029, "grad_norm": 0.8259269595146179, "learning_rate": 0.0002, "epoch": 3.317435938889338, "step": 20520}, {"loss": 0.6268, "grad_norm": 0.8607608079910278, "learning_rate": 0.0002, "epoch": 3.319052623070083, "step": 20530}, {"loss": 0.5676, "grad_norm": 1.0863020420074463, "learning_rate": 0.0002, "epoch": 3.3206693072508284, "step": 20540}, {"loss": 0.6412, "grad_norm": 1.011489987373352, "learning_rate": 0.0002, "epoch": 3.3222859914315737, "step": 20550}, {"loss": 0.6247, "grad_norm": 0.6952177882194519, "learning_rate": 0.0002, "epoch": 3.3239026756123193, "step": 20560}, {"loss": 0.6229, "grad_norm": 0.9638974070549011, "learning_rate": 0.0002, "epoch": 3.3255193597930646, "step": 20570}, {"loss": 0.5882, "grad_norm": 1.0310138463974, "learning_rate": 0.0002, "epoch": 3.32713604397381, "step": 20580}, {"loss": 0.594, "grad_norm": 0.9371318221092224, "learning_rate": 0.0002, "epoch": 3.328752728154555, "step": 20590}, {"loss": 0.6137, "grad_norm": 0.8756691813468933, "learning_rate": 0.0002, "epoch": 3.3303694123353003, "step": 20600}, {"loss": 0.5994, "grad_norm": 1.054175853729248, "learning_rate": 0.0002, "epoch": 3.3319860965160455, "step": 20610}, {"loss": 0.6169, "grad_norm": 0.9074128270149231, "learning_rate": 0.0002, "epoch": 3.3336027806967907, "step": 20620}, {"loss": 0.6138, "grad_norm": 0.906900942325592, "learning_rate": 0.0002, "epoch": 3.335219464877536, "step": 20630}, {"loss": 0.571, "grad_norm": 0.8689333200454712, "learning_rate": 0.0002, "epoch": 3.3368361490582816, "step": 20640}, {"loss": 0.6079, "grad_norm": 0.9889747500419617, "learning_rate": 0.0002, "epoch": 3.338452833239027, "step": 20650}, {"loss": 0.6073, "grad_norm": 1.0685805082321167, "learning_rate": 0.0002, "epoch": 3.340069517419772, "step": 20660}, {"loss": 0.6091, "grad_norm": 0.7495010495185852, "learning_rate": 0.0002, "epoch": 3.3416862016005173, "step": 20670}, {"loss": 0.5883, "grad_norm": 0.8747848272323608, "learning_rate": 0.0002, "epoch": 3.3433028857812626, "step": 20680}, {"loss": 0.604, "grad_norm": 0.9762673377990723, "learning_rate": 0.0002, "epoch": 3.344919569962008, "step": 20690}, {"loss": 0.6784, "grad_norm": 1.0284489393234253, "learning_rate": 0.0002, "epoch": 3.346536254142753, "step": 20700}, {"loss": 0.6464, "grad_norm": 0.7293812036514282, "learning_rate": 0.0002, "epoch": 3.3481529383234987, "step": 20710}, {"loss": 0.609, "grad_norm": 0.8330199122428894, "learning_rate": 0.0002, "epoch": 3.349769622504244, "step": 20720}, {"loss": 0.5729, "grad_norm": 0.9808499217033386, "learning_rate": 0.0002, "epoch": 3.351386306684989, "step": 20730}, {"loss": 0.6315, "grad_norm": 0.9508825540542603, "learning_rate": 0.0002, "epoch": 3.3530029908657344, "step": 20740}, {"loss": 0.5965, "grad_norm": 0.790483832359314, "learning_rate": 0.0002, "epoch": 3.3546196750464796, "step": 20750}, {"loss": 0.6327, "grad_norm": 1.022793173789978, "learning_rate": 0.0002, "epoch": 3.356236359227225, "step": 20760}, {"loss": 0.6439, "grad_norm": 0.8318950533866882, "learning_rate": 0.0002, "epoch": 3.35785304340797, "step": 20770}, {"loss": 0.6037, "grad_norm": 0.7980858087539673, "learning_rate": 0.0002, "epoch": 3.3594697275887153, "step": 20780}, {"loss": 0.6746, "grad_norm": 0.8114802241325378, "learning_rate": 0.0002, "epoch": 3.361086411769461, "step": 20790}, {"loss": 0.6017, "grad_norm": 0.8522519469261169, "learning_rate": 0.0002, "epoch": 3.3627030959502062, "step": 20800}, {"loss": 0.5864, "grad_norm": 0.9142431616783142, "learning_rate": 0.0002, "epoch": 3.3643197801309515, "step": 20810}, {"loss": 0.6331, "grad_norm": 0.771170437335968, "learning_rate": 0.0002, "epoch": 3.3659364643116967, "step": 20820}, {"loss": 0.5879, "grad_norm": 1.0628231763839722, "learning_rate": 0.0002, "epoch": 3.367553148492442, "step": 20830}, {"loss": 0.6533, "grad_norm": 0.9384352564811707, "learning_rate": 0.0002, "epoch": 3.369169832673187, "step": 20840}, {"loss": 0.6292, "grad_norm": 1.1286591291427612, "learning_rate": 0.0002, "epoch": 3.370786516853933, "step": 20850}, {"loss": 0.5986, "grad_norm": 1.1349513530731201, "learning_rate": 0.0002, "epoch": 3.372403201034678, "step": 20860}, {"loss": 0.6413, "grad_norm": 1.0127464532852173, "learning_rate": 0.0002, "epoch": 3.3740198852154233, "step": 20870}, {"loss": 0.6414, "grad_norm": 0.9111971855163574, "learning_rate": 0.0002, "epoch": 3.3756365693961685, "step": 20880}, {"loss": 0.6101, "grad_norm": 0.871356725692749, "learning_rate": 0.0002, "epoch": 3.3772532535769137, "step": 20890}, {"loss": 0.5995, "grad_norm": 0.7774117588996887, "learning_rate": 0.0002, "epoch": 3.378869937757659, "step": 20900}, {"loss": 0.6062, "grad_norm": 1.0089964866638184, "learning_rate": 0.0002, "epoch": 3.380486621938404, "step": 20910}, {"loss": 0.5908, "grad_norm": 0.7855867147445679, "learning_rate": 0.0002, "epoch": 3.3821033061191494, "step": 20920}, {"loss": 0.6373, "grad_norm": 1.3713710308074951, "learning_rate": 0.0002, "epoch": 3.3837199902998947, "step": 20930}, {"loss": 0.6627, "grad_norm": 0.8599116206169128, "learning_rate": 0.0002, "epoch": 3.3853366744806404, "step": 20940}, {"loss": 0.6224, "grad_norm": 0.9392673373222351, "learning_rate": 0.0002, "epoch": 3.3869533586613856, "step": 20950}, {"loss": 0.5855, "grad_norm": 0.8764075040817261, "learning_rate": 0.0002, "epoch": 3.388570042842131, "step": 20960}, {"loss": 0.5734, "grad_norm": 0.8240136504173279, "learning_rate": 0.0002, "epoch": 3.390186727022876, "step": 20970}, {"loss": 0.5783, "grad_norm": 1.0982369184494019, "learning_rate": 0.0002, "epoch": 3.3918034112036213, "step": 20980}, {"loss": 0.5451, "grad_norm": 1.0599013566970825, "learning_rate": 0.0002, "epoch": 3.3934200953843665, "step": 20990}, {"loss": 0.6356, "grad_norm": 0.895438015460968, "learning_rate": 0.0002, "epoch": 3.395036779565112, "step": 21000}, {"loss": 0.6065, "grad_norm": 0.6974841356277466, "learning_rate": 0.0002, "epoch": 3.3966534637458574, "step": 21010}, {"loss": 0.5704, "grad_norm": 0.9571719765663147, "learning_rate": 0.0002, "epoch": 3.3982701479266026, "step": 21020}, {"loss": 0.679, "grad_norm": 0.831912636756897, "learning_rate": 0.0002, "epoch": 3.399886832107348, "step": 21030}, {"loss": 0.6051, "grad_norm": 0.831936240196228, "learning_rate": 0.0002, "epoch": 3.401503516288093, "step": 21040}, {"loss": 0.5857, "grad_norm": 0.7388373613357544, "learning_rate": 0.0002, "epoch": 3.4031202004688383, "step": 21050}, {"loss": 0.6245, "grad_norm": 0.938667356967926, "learning_rate": 0.0002, "epoch": 3.4047368846495836, "step": 21060}, {"loss": 0.6121, "grad_norm": 0.9202313423156738, "learning_rate": 0.0002, "epoch": 3.406353568830329, "step": 21070}, {"loss": 0.6388, "grad_norm": 0.9888381958007812, "learning_rate": 0.0002, "epoch": 3.4079702530110745, "step": 21080}, {"loss": 0.6245, "grad_norm": 0.8526970744132996, "learning_rate": 0.0002, "epoch": 3.4095869371918197, "step": 21090}, {"loss": 0.5914, "grad_norm": 0.7939383387565613, "learning_rate": 0.0002, "epoch": 3.411203621372565, "step": 21100}, {"loss": 0.6066, "grad_norm": 0.9986352920532227, "learning_rate": 0.0002, "epoch": 3.41282030555331, "step": 21110}, {"loss": 0.5947, "grad_norm": 0.8895300030708313, "learning_rate": 0.0002, "epoch": 3.4144369897340554, "step": 21120}, {"loss": 0.6264, "grad_norm": 0.9559482932090759, "learning_rate": 0.0002, "epoch": 3.4160536739148006, "step": 21130}, {"loss": 0.6491, "grad_norm": 0.8351506590843201, "learning_rate": 0.0002, "epoch": 3.417670358095546, "step": 21140}, {"loss": 0.567, "grad_norm": 0.8224456906318665, "learning_rate": 0.0002, "epoch": 3.4192870422762915, "step": 21150}, {"loss": 0.5871, "grad_norm": 1.0110299587249756, "learning_rate": 0.0002, "epoch": 3.4209037264570368, "step": 21160}, {"loss": 0.6116, "grad_norm": 0.82564777135849, "learning_rate": 0.0002, "epoch": 3.422520410637782, "step": 21170}, {"loss": 0.595, "grad_norm": 1.004738688468933, "learning_rate": 0.0002, "epoch": 3.4241370948185272, "step": 21180}, {"loss": 0.6286, "grad_norm": 0.7545676827430725, "learning_rate": 0.0002, "epoch": 3.4257537789992725, "step": 21190}, {"loss": 0.5868, "grad_norm": 0.8918704390525818, "learning_rate": 0.0002, "epoch": 3.4273704631800177, "step": 21200}, {"loss": 0.6542, "grad_norm": 0.8336876034736633, "learning_rate": 0.0002, "epoch": 3.428987147360763, "step": 21210}, {"loss": 0.5824, "grad_norm": 0.8928771018981934, "learning_rate": 0.0002, "epoch": 3.430603831541508, "step": 21220}, {"loss": 0.6468, "grad_norm": 0.7663705945014954, "learning_rate": 0.0002, "epoch": 3.432220515722254, "step": 21230}, {"loss": 0.6693, "grad_norm": 0.8392598628997803, "learning_rate": 0.0002, "epoch": 3.433837199902999, "step": 21240}, {"loss": 0.5971, "grad_norm": 0.8819600343704224, "learning_rate": 0.0002, "epoch": 3.4354538840837443, "step": 21250}, {"loss": 0.6791, "grad_norm": 0.9124642014503479, "learning_rate": 0.0002, "epoch": 3.4370705682644895, "step": 21260}, {"loss": 0.5925, "grad_norm": 0.8329763412475586, "learning_rate": 0.0002, "epoch": 3.4386872524452348, "step": 21270}, {"loss": 0.6541, "grad_norm": 0.9982839822769165, "learning_rate": 0.0002, "epoch": 3.44030393662598, "step": 21280}, {"loss": 0.6441, "grad_norm": 0.9105954766273499, "learning_rate": 0.0002, "epoch": 3.4419206208067252, "step": 21290}, {"loss": 0.6028, "grad_norm": 0.8182359337806702, "learning_rate": 0.0002, "epoch": 3.443537304987471, "step": 21300}, {"loss": 0.5991, "grad_norm": 1.0568904876708984, "learning_rate": 0.0002, "epoch": 3.445153989168216, "step": 21310}, {"loss": 0.6117, "grad_norm": 0.968539834022522, "learning_rate": 0.0002, "epoch": 3.4467706733489614, "step": 21320}, {"loss": 0.6219, "grad_norm": 0.8774511218070984, "learning_rate": 0.0002, "epoch": 3.4483873575297066, "step": 21330}, {"loss": 0.6438, "grad_norm": 0.7598156332969666, "learning_rate": 0.0002, "epoch": 3.450004041710452, "step": 21340}, {"loss": 0.6033, "grad_norm": 1.1012897491455078, "learning_rate": 0.0002, "epoch": 3.451620725891197, "step": 21350}, {"loss": 0.6137, "grad_norm": 0.8040637373924255, "learning_rate": 0.0002, "epoch": 3.4532374100719423, "step": 21360}, {"loss": 0.6173, "grad_norm": 0.8497496247291565, "learning_rate": 0.0002, "epoch": 3.4548540942526875, "step": 21370}, {"loss": 0.6005, "grad_norm": 0.8429915904998779, "learning_rate": 0.0002, "epoch": 3.456470778433433, "step": 21380}, {"loss": 0.6182, "grad_norm": 0.8107112646102905, "learning_rate": 0.0002, "epoch": 3.4580874626141784, "step": 21390}, {"loss": 0.6109, "grad_norm": 1.00872004032135, "learning_rate": 0.0002, "epoch": 3.4597041467949237, "step": 21400}, {"loss": 0.5712, "grad_norm": 0.8266542553901672, "learning_rate": 0.0002, "epoch": 3.461320830975669, "step": 21410}, {"loss": 0.6457, "grad_norm": 0.8972568511962891, "learning_rate": 0.0002, "epoch": 3.462937515156414, "step": 21420}, {"loss": 0.6081, "grad_norm": 1.0781476497650146, "learning_rate": 0.0002, "epoch": 3.4645541993371594, "step": 21430}, {"loss": 0.6303, "grad_norm": 0.9571592807769775, "learning_rate": 0.0002, "epoch": 3.4661708835179046, "step": 21440}, {"loss": 0.6309, "grad_norm": 0.881547212600708, "learning_rate": 0.0002, "epoch": 3.4677875676986503, "step": 21450}, {"loss": 0.6076, "grad_norm": 0.6955338716506958, "learning_rate": 0.0002, "epoch": 3.4694042518793955, "step": 21460}, {"loss": 0.6205, "grad_norm": 0.901187539100647, "learning_rate": 0.0002, "epoch": 3.4710209360601407, "step": 21470}, {"loss": 0.639, "grad_norm": 0.7063511610031128, "learning_rate": 0.0002, "epoch": 3.472637620240886, "step": 21480}, {"loss": 0.6154, "grad_norm": 0.8462792038917542, "learning_rate": 0.0002, "epoch": 3.474254304421631, "step": 21490}, {"loss": 0.61, "grad_norm": 1.1861060857772827, "learning_rate": 0.0002, "epoch": 3.4758709886023764, "step": 21500}, {"loss": 0.6586, "grad_norm": 0.70503169298172, "learning_rate": 0.0002, "epoch": 3.4774876727831217, "step": 21510}, {"loss": 0.6475, "grad_norm": 0.9650066494941711, "learning_rate": 0.0002, "epoch": 3.479104356963867, "step": 21520}, {"loss": 0.6452, "grad_norm": 1.0266852378845215, "learning_rate": 0.0002, "epoch": 3.4807210411446126, "step": 21530}, {"loss": 0.6553, "grad_norm": 0.956372857093811, "learning_rate": 0.0002, "epoch": 3.482337725325358, "step": 21540}, {"loss": 0.6667, "grad_norm": 0.8848432898521423, "learning_rate": 0.0002, "epoch": 3.483954409506103, "step": 21550}, {"loss": 0.6375, "grad_norm": 1.0805351734161377, "learning_rate": 0.0002, "epoch": 3.4855710936868483, "step": 21560}, {"loss": 0.6958, "grad_norm": 0.9279725551605225, "learning_rate": 0.0002, "epoch": 3.4871877778675935, "step": 21570}, {"loss": 0.6354, "grad_norm": 0.9049562215805054, "learning_rate": 0.0002, "epoch": 3.4888044620483387, "step": 21580}, {"loss": 0.6071, "grad_norm": 0.9619429111480713, "learning_rate": 0.0002, "epoch": 3.4904211462290844, "step": 21590}, {"loss": 0.5927, "grad_norm": 0.8508906960487366, "learning_rate": 0.0002, "epoch": 3.4920378304098296, "step": 21600}, {"loss": 0.6115, "grad_norm": 0.8692502379417419, "learning_rate": 0.0002, "epoch": 3.493654514590575, "step": 21610}, {"loss": 0.5878, "grad_norm": 0.8187332153320312, "learning_rate": 0.0002, "epoch": 3.49527119877132, "step": 21620}, {"loss": 0.5874, "grad_norm": 1.145400047302246, "learning_rate": 0.0002, "epoch": 3.4968878829520653, "step": 21630}, {"loss": 0.6313, "grad_norm": 0.8281388282775879, "learning_rate": 0.0002, "epoch": 3.4985045671328105, "step": 21640}, {"loss": 0.6624, "grad_norm": 0.82256019115448, "learning_rate": 0.0002, "epoch": 3.500121251313556, "step": 21650}, {"loss": 0.6346, "grad_norm": 0.9315484762191772, "learning_rate": 0.0002, "epoch": 3.501737935494301, "step": 21660}, {"loss": 0.6086, "grad_norm": 0.7626111507415771, "learning_rate": 0.0002, "epoch": 3.5033546196750462, "step": 21670}, {"loss": 0.6177, "grad_norm": 0.9275059103965759, "learning_rate": 0.0002, "epoch": 3.504971303855792, "step": 21680}, {"loss": 0.64, "grad_norm": 0.7906724810600281, "learning_rate": 0.0002, "epoch": 3.506587988036537, "step": 21690}, {"loss": 0.6015, "grad_norm": 0.8289761543273926, "learning_rate": 0.0002, "epoch": 3.5082046722172824, "step": 21700}, {"loss": 0.6246, "grad_norm": 0.8316431045532227, "learning_rate": 0.0002, "epoch": 3.5098213563980276, "step": 21710}, {"loss": 0.619, "grad_norm": 1.0451812744140625, "learning_rate": 0.0002, "epoch": 3.511438040578773, "step": 21720}, {"loss": 0.632, "grad_norm": 0.928252637386322, "learning_rate": 0.0002, "epoch": 3.513054724759518, "step": 21730}, {"loss": 0.6062, "grad_norm": 0.7985895276069641, "learning_rate": 0.0002, "epoch": 3.5146714089402638, "step": 21740}, {"loss": 0.6463, "grad_norm": 0.6740974187850952, "learning_rate": 0.0002, "epoch": 3.516288093121009, "step": 21750}, {"loss": 0.6138, "grad_norm": 0.8482223749160767, "learning_rate": 0.0002, "epoch": 3.517904777301754, "step": 21760}, {"loss": 0.6277, "grad_norm": 0.889947772026062, "learning_rate": 0.0002, "epoch": 3.5195214614824994, "step": 21770}, {"loss": 0.6174, "grad_norm": 0.8304598927497864, "learning_rate": 0.0002, "epoch": 3.5211381456632447, "step": 21780}, {"loss": 0.6156, "grad_norm": 0.8002981543540955, "learning_rate": 0.0002, "epoch": 3.52275482984399, "step": 21790}, {"loss": 0.5896, "grad_norm": 0.8115083575248718, "learning_rate": 0.0002, "epoch": 3.524371514024735, "step": 21800}, {"loss": 0.6041, "grad_norm": 0.9715048670768738, "learning_rate": 0.0002, "epoch": 3.5259881982054804, "step": 21810}, {"loss": 0.6715, "grad_norm": 1.0910786390304565, "learning_rate": 0.0002, "epoch": 3.5276048823862256, "step": 21820}, {"loss": 0.6543, "grad_norm": 0.8438942432403564, "learning_rate": 0.0002, "epoch": 3.5292215665669713, "step": 21830}, {"loss": 0.6509, "grad_norm": 0.8813382983207703, "learning_rate": 0.0002, "epoch": 3.5308382507477165, "step": 21840}, {"loss": 0.6049, "grad_norm": 0.7092908024787903, "learning_rate": 0.0002, "epoch": 3.5324549349284617, "step": 21850}, {"loss": 0.5678, "grad_norm": 0.8332187533378601, "learning_rate": 0.0002, "epoch": 3.534071619109207, "step": 21860}, {"loss": 0.5896, "grad_norm": 0.8958209156990051, "learning_rate": 0.0002, "epoch": 3.535688303289952, "step": 21870}, {"loss": 0.6476, "grad_norm": 0.824138879776001, "learning_rate": 0.0002, "epoch": 3.5373049874706974, "step": 21880}, {"loss": 0.6022, "grad_norm": 0.8375158309936523, "learning_rate": 0.0002, "epoch": 3.538921671651443, "step": 21890}, {"loss": 0.6019, "grad_norm": 1.0274608135223389, "learning_rate": 0.0002, "epoch": 3.5405383558321883, "step": 21900}, {"loss": 0.6194, "grad_norm": 0.7088932394981384, "learning_rate": 0.0002, "epoch": 3.5421550400129336, "step": 21910}, {"loss": 0.6554, "grad_norm": 0.8172445297241211, "learning_rate": 0.0002, "epoch": 3.543771724193679, "step": 21920}, {"loss": 0.6711, "grad_norm": 0.9904135465621948, "learning_rate": 0.0002, "epoch": 3.545388408374424, "step": 21930}, {"loss": 0.6001, "grad_norm": 0.9900432229042053, "learning_rate": 0.0002, "epoch": 3.5470050925551693, "step": 21940}, {"loss": 0.6195, "grad_norm": 0.8963301181793213, "learning_rate": 0.0002, "epoch": 3.5486217767359145, "step": 21950}, {"loss": 0.5972, "grad_norm": 0.8551464676856995, "learning_rate": 0.0002, "epoch": 3.5502384609166597, "step": 21960}, {"loss": 0.6206, "grad_norm": 1.0916603803634644, "learning_rate": 0.0002, "epoch": 3.551855145097405, "step": 21970}, {"loss": 0.6523, "grad_norm": 0.841598391532898, "learning_rate": 0.0002, "epoch": 3.5534718292781506, "step": 21980}, {"loss": 0.617, "grad_norm": 0.8566757440567017, "learning_rate": 0.0002, "epoch": 3.555088513458896, "step": 21990}, {"loss": 0.6192, "grad_norm": 1.0145052671432495, "learning_rate": 0.0002, "epoch": 3.556705197639641, "step": 22000}, {"loss": 0.6173, "grad_norm": 0.9293754696846008, "learning_rate": 0.0002, "epoch": 3.5583218818203863, "step": 22010}, {"loss": 0.612, "grad_norm": 0.9568536281585693, "learning_rate": 0.0002, "epoch": 3.5599385660011316, "step": 22020}, {"loss": 0.641, "grad_norm": 0.8613139986991882, "learning_rate": 0.0002, "epoch": 3.5615552501818772, "step": 22030}, {"loss": 0.6496, "grad_norm": 0.8179237246513367, "learning_rate": 0.0002, "epoch": 3.5631719343626225, "step": 22040}, {"loss": 0.574, "grad_norm": 0.9059830904006958, "learning_rate": 0.0002, "epoch": 3.5647886185433677, "step": 22050}, {"loss": 0.6448, "grad_norm": 1.0068252086639404, "learning_rate": 0.0002, "epoch": 3.566405302724113, "step": 22060}, {"loss": 0.6239, "grad_norm": 0.9682072997093201, "learning_rate": 0.0002, "epoch": 3.568021986904858, "step": 22070}, {"loss": 0.6808, "grad_norm": 0.8514005541801453, "learning_rate": 0.0002, "epoch": 3.5696386710856034, "step": 22080}, {"loss": 0.5956, "grad_norm": 0.8327770829200745, "learning_rate": 0.0002, "epoch": 3.5712553552663486, "step": 22090}, {"loss": 0.5976, "grad_norm": 1.024976372718811, "learning_rate": 0.0002, "epoch": 3.572872039447094, "step": 22100}, {"loss": 0.624, "grad_norm": 0.7721174955368042, "learning_rate": 0.0002, "epoch": 3.574488723627839, "step": 22110}, {"loss": 0.5896, "grad_norm": 1.0351054668426514, "learning_rate": 0.0002, "epoch": 3.5761054078085843, "step": 22120}, {"loss": 0.6379, "grad_norm": 0.9680907130241394, "learning_rate": 0.0002, "epoch": 3.57772209198933, "step": 22130}, {"loss": 0.6194, "grad_norm": 0.8016974925994873, "learning_rate": 0.0002, "epoch": 3.5793387761700752, "step": 22140}, {"loss": 0.6387, "grad_norm": 1.0109003782272339, "learning_rate": 0.0002, "epoch": 3.5809554603508205, "step": 22150}, {"loss": 0.6368, "grad_norm": 1.0473392009735107, "learning_rate": 0.0002, "epoch": 3.5825721445315657, "step": 22160}, {"loss": 0.6353, "grad_norm": 0.8686613440513611, "learning_rate": 0.0002, "epoch": 3.584188828712311, "step": 22170}, {"loss": 0.5791, "grad_norm": 0.869149923324585, "learning_rate": 0.0002, "epoch": 3.5858055128930566, "step": 22180}, {"loss": 0.5895, "grad_norm": 0.9769062995910645, "learning_rate": 0.0002, "epoch": 3.587422197073802, "step": 22190}, {"loss": 0.5939, "grad_norm": 0.779636561870575, "learning_rate": 0.0002, "epoch": 3.589038881254547, "step": 22200}, {"loss": 0.5875, "grad_norm": 0.9063841104507446, "learning_rate": 0.0002, "epoch": 3.5906555654352923, "step": 22210}, {"loss": 0.5671, "grad_norm": 0.9216037392616272, "learning_rate": 0.0002, "epoch": 3.5922722496160375, "step": 22220}, {"loss": 0.6484, "grad_norm": 1.0217336416244507, "learning_rate": 0.0002, "epoch": 3.5938889337967828, "step": 22230}, {"loss": 0.6511, "grad_norm": 0.8513161540031433, "learning_rate": 0.0002, "epoch": 3.595505617977528, "step": 22240}, {"loss": 0.6301, "grad_norm": 0.8084813952445984, "learning_rate": 0.0002, "epoch": 3.597122302158273, "step": 22250}, {"loss": 0.6197, "grad_norm": 0.8524802923202515, "learning_rate": 0.0002, "epoch": 3.5987389863390185, "step": 22260}, {"loss": 0.5599, "grad_norm": 0.9356237649917603, "learning_rate": 0.0002, "epoch": 3.600355670519764, "step": 22270}, {"loss": 0.628, "grad_norm": 1.009600281715393, "learning_rate": 0.0002, "epoch": 3.6019723547005094, "step": 22280}, {"loss": 0.6179, "grad_norm": 0.9900581240653992, "learning_rate": 0.0002, "epoch": 3.6035890388812546, "step": 22290}, {"loss": 0.5725, "grad_norm": 1.062495231628418, "learning_rate": 0.0002, "epoch": 3.605205723062, "step": 22300}, {"loss": 0.607, "grad_norm": 0.8832381367683411, "learning_rate": 0.0002, "epoch": 3.606822407242745, "step": 22310}, {"loss": 0.6215, "grad_norm": 0.9284297823905945, "learning_rate": 0.0002, "epoch": 3.6084390914234903, "step": 22320}, {"loss": 0.685, "grad_norm": 1.2381829023361206, "learning_rate": 0.0002, "epoch": 3.610055775604236, "step": 22330}, {"loss": 0.6181, "grad_norm": 0.929434597492218, "learning_rate": 0.0002, "epoch": 3.611672459784981, "step": 22340}, {"loss": 0.6141, "grad_norm": 0.9714490175247192, "learning_rate": 0.0002, "epoch": 3.6132891439657264, "step": 22350}, {"loss": 0.6861, "grad_norm": 0.808014988899231, "learning_rate": 0.0002, "epoch": 3.6149058281464717, "step": 22360}, {"loss": 0.6428, "grad_norm": 1.0364398956298828, "learning_rate": 0.0002, "epoch": 3.616522512327217, "step": 22370}, {"loss": 0.6337, "grad_norm": 0.7858489751815796, "learning_rate": 0.0002, "epoch": 3.618139196507962, "step": 22380}, {"loss": 0.6214, "grad_norm": 0.9920870065689087, "learning_rate": 0.0002, "epoch": 3.6197558806887074, "step": 22390}, {"loss": 0.6659, "grad_norm": 0.9183220863342285, "learning_rate": 0.0002, "epoch": 3.6213725648694526, "step": 22400}, {"loss": 0.6036, "grad_norm": 0.9826246500015259, "learning_rate": 0.0002, "epoch": 3.622989249050198, "step": 22410}, {"loss": 0.6441, "grad_norm": 0.8632931113243103, "learning_rate": 0.0002, "epoch": 3.6246059332309435, "step": 22420}, {"loss": 0.6124, "grad_norm": 0.8468965291976929, "learning_rate": 0.0002, "epoch": 3.6262226174116887, "step": 22430}, {"loss": 0.6328, "grad_norm": 0.8466871976852417, "learning_rate": 0.0002, "epoch": 3.627839301592434, "step": 22440}, {"loss": 0.5941, "grad_norm": 0.9501169919967651, "learning_rate": 0.0002, "epoch": 3.629455985773179, "step": 22450}, {"loss": 0.6069, "grad_norm": 0.8906720876693726, "learning_rate": 0.0002, "epoch": 3.6310726699539244, "step": 22460}, {"loss": 0.6928, "grad_norm": 0.7400227189064026, "learning_rate": 0.0002, "epoch": 3.6326893541346696, "step": 22470}, {"loss": 0.6337, "grad_norm": 0.9756355881690979, "learning_rate": 0.0002, "epoch": 3.6343060383154153, "step": 22480}, {"loss": 0.6203, "grad_norm": 0.7504993081092834, "learning_rate": 0.0002, "epoch": 3.6359227224961606, "step": 22490}, {"loss": 0.6302, "grad_norm": 0.9270039200782776, "learning_rate": 0.0002, "epoch": 3.637539406676906, "step": 22500}, {"loss": 0.6026, "grad_norm": 0.8841686844825745, "learning_rate": 0.0002, "epoch": 3.639156090857651, "step": 22510}, {"loss": 0.6098, "grad_norm": 0.8533213138580322, "learning_rate": 0.0002, "epoch": 3.6407727750383962, "step": 22520}, {"loss": 0.6412, "grad_norm": 1.0052043199539185, "learning_rate": 0.0002, "epoch": 3.6423894592191415, "step": 22530}, {"loss": 0.6363, "grad_norm": 1.0323461294174194, "learning_rate": 0.0002, "epoch": 3.6440061433998867, "step": 22540}, {"loss": 0.6545, "grad_norm": 0.8654312491416931, "learning_rate": 0.0002, "epoch": 3.645622827580632, "step": 22550}, {"loss": 0.6155, "grad_norm": 0.6400038003921509, "learning_rate": 0.0002, "epoch": 3.647239511761377, "step": 22560}, {"loss": 0.5829, "grad_norm": 0.8061298727989197, "learning_rate": 0.0002, "epoch": 3.648856195942123, "step": 22570}, {"loss": 0.6388, "grad_norm": 0.9257854223251343, "learning_rate": 0.0002, "epoch": 3.650472880122868, "step": 22580}, {"loss": 0.6409, "grad_norm": 0.8439396619796753, "learning_rate": 0.0002, "epoch": 3.6520895643036133, "step": 22590}, {"loss": 0.5996, "grad_norm": 0.7764544486999512, "learning_rate": 0.0002, "epoch": 3.6537062484843585, "step": 22600}, {"loss": 0.6434, "grad_norm": 1.125451683998108, "learning_rate": 0.0002, "epoch": 3.6553229326651038, "step": 22610}, {"loss": 0.6579, "grad_norm": 0.7523018717765808, "learning_rate": 0.0002, "epoch": 3.656939616845849, "step": 22620}, {"loss": 0.6476, "grad_norm": 1.071026086807251, "learning_rate": 0.0002, "epoch": 3.6585563010265947, "step": 22630}, {"loss": 0.6459, "grad_norm": 0.945791482925415, "learning_rate": 0.0002, "epoch": 3.66017298520734, "step": 22640}, {"loss": 0.659, "grad_norm": 0.8001811504364014, "learning_rate": 0.0002, "epoch": 3.661789669388085, "step": 22650}, {"loss": 0.6385, "grad_norm": 0.9700816869735718, "learning_rate": 0.0002, "epoch": 3.6634063535688304, "step": 22660}, {"loss": 0.6337, "grad_norm": 0.9053242206573486, "learning_rate": 0.0002, "epoch": 3.6650230377495756, "step": 22670}, {"loss": 0.6335, "grad_norm": 0.944362461566925, "learning_rate": 0.0002, "epoch": 3.666639721930321, "step": 22680}, {"loss": 0.6235, "grad_norm": 1.067489504814148, "learning_rate": 0.0002, "epoch": 3.668256406111066, "step": 22690}, {"loss": 0.698, "grad_norm": 1.0984995365142822, "learning_rate": 0.0002, "epoch": 3.6698730902918113, "step": 22700}, {"loss": 0.6717, "grad_norm": 0.9336317777633667, "learning_rate": 0.0002, "epoch": 3.6714897744725565, "step": 22710}, {"loss": 0.6195, "grad_norm": 0.9261918663978577, "learning_rate": 0.0002, "epoch": 3.673106458653302, "step": 22720}, {"loss": 0.6332, "grad_norm": 0.8648008704185486, "learning_rate": 0.0002, "epoch": 3.6747231428340474, "step": 22730}, {"loss": 0.6576, "grad_norm": 0.7225083708763123, "learning_rate": 0.0002, "epoch": 3.6763398270147927, "step": 22740}, {"loss": 0.6406, "grad_norm": 0.9258282780647278, "learning_rate": 0.0002, "epoch": 3.677956511195538, "step": 22750}, {"loss": 0.6397, "grad_norm": 0.70876145362854, "learning_rate": 0.0002, "epoch": 3.679573195376283, "step": 22760}, {"loss": 0.6821, "grad_norm": 0.8780210018157959, "learning_rate": 0.0002, "epoch": 3.681189879557029, "step": 22770}, {"loss": 0.6036, "grad_norm": 0.8075440526008606, "learning_rate": 0.0002, "epoch": 3.682806563737774, "step": 22780}, {"loss": 0.6561, "grad_norm": 0.8503130674362183, "learning_rate": 0.0002, "epoch": 3.6844232479185193, "step": 22790}, {"loss": 0.6082, "grad_norm": 0.8413618206977844, "learning_rate": 0.0002, "epoch": 3.6860399320992645, "step": 22800}, {"loss": 0.614, "grad_norm": 0.8675165176391602, "learning_rate": 0.0002, "epoch": 3.6876566162800097, "step": 22810}, {"loss": 0.6157, "grad_norm": 0.8235884308815002, "learning_rate": 0.0002, "epoch": 3.689273300460755, "step": 22820}, {"loss": 0.5708, "grad_norm": 0.9477725625038147, "learning_rate": 0.0002, "epoch": 3.6908899846415, "step": 22830}, {"loss": 0.6481, "grad_norm": 0.7883533835411072, "learning_rate": 0.0002, "epoch": 3.6925066688222454, "step": 22840}, {"loss": 0.5872, "grad_norm": 1.047913908958435, "learning_rate": 0.0002, "epoch": 3.6941233530029907, "step": 22850}, {"loss": 0.6176, "grad_norm": 0.9171528816223145, "learning_rate": 0.0002, "epoch": 3.695740037183736, "step": 22860}, {"loss": 0.6204, "grad_norm": 0.9338192343711853, "learning_rate": 0.0002, "epoch": 3.6973567213644816, "step": 22870}, {"loss": 0.686, "grad_norm": 0.8799443244934082, "learning_rate": 0.0002, "epoch": 3.698973405545227, "step": 22880}, {"loss": 0.6206, "grad_norm": 0.8515434861183167, "learning_rate": 0.0002, "epoch": 3.700590089725972, "step": 22890}, {"loss": 0.5954, "grad_norm": 0.7805591821670532, "learning_rate": 0.0002, "epoch": 3.7022067739067173, "step": 22900}, {"loss": 0.6108, "grad_norm": 0.8470911979675293, "learning_rate": 0.0002, "epoch": 3.7038234580874625, "step": 22910}, {"loss": 0.6557, "grad_norm": 0.9452309012413025, "learning_rate": 0.0002, "epoch": 3.705440142268208, "step": 22920}, {"loss": 0.6529, "grad_norm": 0.950243353843689, "learning_rate": 0.0002, "epoch": 3.7070568264489534, "step": 22930}, {"loss": 0.6364, "grad_norm": 0.7882499098777771, "learning_rate": 0.0002, "epoch": 3.7086735106296986, "step": 22940}, {"loss": 0.6462, "grad_norm": 0.8307787775993347, "learning_rate": 0.0002, "epoch": 3.710290194810444, "step": 22950}, {"loss": 0.6371, "grad_norm": 1.0970630645751953, "learning_rate": 0.0002, "epoch": 3.711906878991189, "step": 22960}, {"loss": 0.6281, "grad_norm": 0.8269566297531128, "learning_rate": 0.0002, "epoch": 3.7135235631719343, "step": 22970}, {"loss": 0.6561, "grad_norm": 0.8306704759597778, "learning_rate": 0.0002, "epoch": 3.7151402473526796, "step": 22980}, {"loss": 0.6418, "grad_norm": 0.9710225462913513, "learning_rate": 0.0002, "epoch": 3.716756931533425, "step": 22990}, {"loss": 0.6639, "grad_norm": 0.8890530467033386, "learning_rate": 0.0002, "epoch": 3.71837361571417, "step": 23000}, {"loss": 0.6084, "grad_norm": 0.883522629737854, "learning_rate": 0.0002, "epoch": 3.7199902998949153, "step": 23010}, {"loss": 0.6183, "grad_norm": 0.8662652373313904, "learning_rate": 0.0002, "epoch": 3.721606984075661, "step": 23020}, {"loss": 0.6266, "grad_norm": 0.7228406667709351, "learning_rate": 0.0002, "epoch": 3.723223668256406, "step": 23030}, {"loss": 0.6417, "grad_norm": 1.060792088508606, "learning_rate": 0.0002, "epoch": 3.7248403524371514, "step": 23040}, {"loss": 0.6346, "grad_norm": 1.0119613409042358, "learning_rate": 0.0002, "epoch": 3.7264570366178966, "step": 23050}, {"loss": 0.6466, "grad_norm": 0.9212996959686279, "learning_rate": 0.0002, "epoch": 3.728073720798642, "step": 23060}, {"loss": 0.6454, "grad_norm": 0.925690233707428, "learning_rate": 0.0002, "epoch": 3.7296904049793875, "step": 23070}, {"loss": 0.615, "grad_norm": 0.8323310613632202, "learning_rate": 0.0002, "epoch": 3.7313070891601328, "step": 23080}, {"loss": 0.679, "grad_norm": 0.8966048955917358, "learning_rate": 0.0002, "epoch": 3.732923773340878, "step": 23090}, {"loss": 0.6151, "grad_norm": 0.8995837569236755, "learning_rate": 0.0002, "epoch": 3.7345404575216232, "step": 23100}, {"loss": 0.6143, "grad_norm": 0.8748890161514282, "learning_rate": 0.0002, "epoch": 3.7361571417023685, "step": 23110}, {"loss": 0.6246, "grad_norm": 0.7985540628433228, "learning_rate": 0.0002, "epoch": 3.7377738258831137, "step": 23120}, {"loss": 0.6279, "grad_norm": 1.0240917205810547, "learning_rate": 0.0002, "epoch": 3.739390510063859, "step": 23130}, {"loss": 0.6747, "grad_norm": 0.9181789755821228, "learning_rate": 0.0002, "epoch": 3.741007194244604, "step": 23140}, {"loss": 0.6026, "grad_norm": 0.8896583914756775, "learning_rate": 0.0002, "epoch": 3.7426238784253494, "step": 23150}, {"loss": 0.5972, "grad_norm": 0.8635515570640564, "learning_rate": 0.0002, "epoch": 3.744240562606095, "step": 23160}, {"loss": 0.6683, "grad_norm": 0.8873575329780579, "learning_rate": 0.0002, "epoch": 3.7458572467868403, "step": 23170}, {"loss": 0.6143, "grad_norm": 0.9807148575782776, "learning_rate": 0.0002, "epoch": 3.7474739309675855, "step": 23180}, {"loss": 0.6381, "grad_norm": 0.900477945804596, "learning_rate": 0.0002, "epoch": 3.7490906151483308, "step": 23190}, {"loss": 0.6542, "grad_norm": 0.9379992485046387, "learning_rate": 0.0002, "epoch": 3.750707299329076, "step": 23200}, {"loss": 0.6015, "grad_norm": 0.9649890661239624, "learning_rate": 0.0002, "epoch": 3.752323983509821, "step": 23210}, {"loss": 0.6735, "grad_norm": 0.824442446231842, "learning_rate": 0.0002, "epoch": 3.753940667690567, "step": 23220}, {"loss": 0.5992, "grad_norm": 0.8896150588989258, "learning_rate": 0.0002, "epoch": 3.755557351871312, "step": 23230}, {"loss": 0.6081, "grad_norm": 0.751249372959137, "learning_rate": 0.0002, "epoch": 3.7571740360520574, "step": 23240}, {"loss": 0.629, "grad_norm": 0.9392193555831909, "learning_rate": 0.0002, "epoch": 3.7587907202328026, "step": 23250}, {"loss": 0.6209, "grad_norm": 0.9284586310386658, "learning_rate": 0.0002, "epoch": 3.760407404413548, "step": 23260}, {"loss": 0.6414, "grad_norm": 0.7738175392150879, "learning_rate": 0.0002, "epoch": 3.762024088594293, "step": 23270}, {"loss": 0.6743, "grad_norm": 0.9252978563308716, "learning_rate": 0.0002, "epoch": 3.7636407727750383, "step": 23280}, {"loss": 0.5984, "grad_norm": 0.9501895904541016, "learning_rate": 0.0002, "epoch": 3.7652574569557835, "step": 23290}, {"loss": 0.6568, "grad_norm": 0.9416276216506958, "learning_rate": 0.0002, "epoch": 3.7668741411365287, "step": 23300}, {"loss": 0.6507, "grad_norm": 0.7076631784439087, "learning_rate": 0.0002, "epoch": 3.7684908253172744, "step": 23310}, {"loss": 0.6329, "grad_norm": 0.9864492416381836, "learning_rate": 0.0002, "epoch": 3.7701075094980196, "step": 23320}, {"loss": 0.6537, "grad_norm": 0.8450456261634827, "learning_rate": 0.0002, "epoch": 3.771724193678765, "step": 23330}, {"loss": 0.658, "grad_norm": 1.0768941640853882, "learning_rate": 0.0002, "epoch": 3.77334087785951, "step": 23340}, {"loss": 0.6408, "grad_norm": 0.9956819415092468, "learning_rate": 0.0002, "epoch": 3.7749575620402553, "step": 23350}, {"loss": 0.6464, "grad_norm": 0.9234658479690552, "learning_rate": 0.0002, "epoch": 3.7765742462210006, "step": 23360}, {"loss": 0.6542, "grad_norm": 1.0993858575820923, "learning_rate": 0.0002, "epoch": 3.7781909304017463, "step": 23370}, {"loss": 0.6391, "grad_norm": 0.923159658908844, "learning_rate": 0.0002, "epoch": 3.7798076145824915, "step": 23380}, {"loss": 0.6625, "grad_norm": 0.9311541318893433, "learning_rate": 0.0002, "epoch": 3.7814242987632367, "step": 23390}, {"loss": 0.6535, "grad_norm": 0.919681191444397, "learning_rate": 0.0002, "epoch": 3.783040982943982, "step": 23400}, {"loss": 0.6138, "grad_norm": 1.7406195402145386, "learning_rate": 0.0002, "epoch": 3.784657667124727, "step": 23410}, {"loss": 0.657, "grad_norm": 0.7789074182510376, "learning_rate": 0.0002, "epoch": 3.7862743513054724, "step": 23420}, {"loss": 0.658, "grad_norm": 0.8302814960479736, "learning_rate": 0.0002, "epoch": 3.7878910354862176, "step": 23430}, {"loss": 0.649, "grad_norm": 0.8089349269866943, "learning_rate": 0.0002, "epoch": 3.789507719666963, "step": 23440}, {"loss": 0.6682, "grad_norm": 0.9006284475326538, "learning_rate": 0.0002, "epoch": 3.791124403847708, "step": 23450}, {"loss": 0.6335, "grad_norm": 0.8426766991615295, "learning_rate": 0.0002, "epoch": 3.7927410880284538, "step": 23460}, {"loss": 0.6364, "grad_norm": 1.2576252222061157, "learning_rate": 0.0002, "epoch": 3.794357772209199, "step": 23470}, {"loss": 0.6324, "grad_norm": 1.0307610034942627, "learning_rate": 0.0002, "epoch": 3.7959744563899442, "step": 23480}, {"loss": 0.6262, "grad_norm": 0.8525972962379456, "learning_rate": 0.0002, "epoch": 3.7975911405706895, "step": 23490}, {"loss": 0.6757, "grad_norm": 1.159039855003357, "learning_rate": 0.0002, "epoch": 3.7992078247514347, "step": 23500}, {"loss": 0.6414, "grad_norm": 1.4193549156188965, "learning_rate": 0.0002, "epoch": 3.80082450893218, "step": 23510}, {"loss": 0.6413, "grad_norm": 0.8245543837547302, "learning_rate": 0.0002, "epoch": 3.8024411931129256, "step": 23520}, {"loss": 0.6417, "grad_norm": 0.8847230076789856, "learning_rate": 0.0002, "epoch": 3.804057877293671, "step": 23530}, {"loss": 0.6415, "grad_norm": 0.9574624300003052, "learning_rate": 0.0002, "epoch": 3.805674561474416, "step": 23540}, {"loss": 0.5765, "grad_norm": 1.048020601272583, "learning_rate": 0.0002, "epoch": 3.8072912456551613, "step": 23550}, {"loss": 0.6497, "grad_norm": 0.8302255868911743, "learning_rate": 0.0002, "epoch": 3.8089079298359065, "step": 23560}, {"loss": 0.6534, "grad_norm": 0.8269215822219849, "learning_rate": 0.0002, "epoch": 3.8105246140166518, "step": 23570}, {"loss": 0.6294, "grad_norm": 0.9375753402709961, "learning_rate": 0.0002, "epoch": 3.812141298197397, "step": 23580}, {"loss": 0.6132, "grad_norm": 1.0234097242355347, "learning_rate": 0.0002, "epoch": 3.8137579823781422, "step": 23590}, {"loss": 0.6625, "grad_norm": 0.8978445529937744, "learning_rate": 0.0002, "epoch": 3.8153746665588875, "step": 23600}, {"loss": 0.6315, "grad_norm": 0.7929515838623047, "learning_rate": 0.0002, "epoch": 3.816991350739633, "step": 23610}, {"loss": 0.6387, "grad_norm": 1.3255881071090698, "learning_rate": 0.0002, "epoch": 3.8186080349203784, "step": 23620}, {"loss": 0.5947, "grad_norm": 0.9188598990440369, "learning_rate": 0.0002, "epoch": 3.8202247191011236, "step": 23630}, {"loss": 0.6152, "grad_norm": 0.8811675906181335, "learning_rate": 0.0002, "epoch": 3.821841403281869, "step": 23640}, {"loss": 0.6253, "grad_norm": 0.8061038255691528, "learning_rate": 0.0002, "epoch": 3.823458087462614, "step": 23650}, {"loss": 0.6517, "grad_norm": 0.9975376129150391, "learning_rate": 0.0002, "epoch": 3.8250747716433597, "step": 23660}, {"loss": 0.6288, "grad_norm": 0.8036105036735535, "learning_rate": 0.0002, "epoch": 3.826691455824105, "step": 23670}, {"loss": 0.6845, "grad_norm": 0.7401984333992004, "learning_rate": 0.0002, "epoch": 3.82830814000485, "step": 23680}, {"loss": 0.6423, "grad_norm": 0.829753041267395, "learning_rate": 0.0002, "epoch": 3.8299248241855954, "step": 23690}, {"loss": 0.6611, "grad_norm": 0.8753240704536438, "learning_rate": 0.0002, "epoch": 3.8315415083663407, "step": 23700}, {"loss": 0.6686, "grad_norm": 0.8157842755317688, "learning_rate": 0.0002, "epoch": 3.833158192547086, "step": 23710}, {"loss": 0.6181, "grad_norm": 0.6183798909187317, "learning_rate": 0.0002, "epoch": 3.834774876727831, "step": 23720}, {"loss": 0.5965, "grad_norm": 0.9548442363739014, "learning_rate": 0.0002, "epoch": 3.8363915609085764, "step": 23730}, {"loss": 0.6456, "grad_norm": 0.8319669961929321, "learning_rate": 0.0002, "epoch": 3.8380082450893216, "step": 23740}, {"loss": 0.6585, "grad_norm": 0.9718693494796753, "learning_rate": 0.0002, "epoch": 3.839624929270067, "step": 23750}, {"loss": 0.6518, "grad_norm": 0.8672235012054443, "learning_rate": 0.0002, "epoch": 3.8412416134508125, "step": 23760}, {"loss": 0.6774, "grad_norm": 1.1210707426071167, "learning_rate": 0.0002, "epoch": 3.8428582976315577, "step": 23770}, {"loss": 0.5923, "grad_norm": 0.9177767634391785, "learning_rate": 0.0002, "epoch": 3.844474981812303, "step": 23780}, {"loss": 0.6286, "grad_norm": 0.8714171648025513, "learning_rate": 0.0002, "epoch": 3.846091665993048, "step": 23790}, {"loss": 0.6302, "grad_norm": 1.1853246688842773, "learning_rate": 0.0002, "epoch": 3.8477083501737934, "step": 23800}, {"loss": 0.6144, "grad_norm": 0.8091260194778442, "learning_rate": 0.0002, "epoch": 3.849325034354539, "step": 23810}, {"loss": 0.658, "grad_norm": 0.9710774421691895, "learning_rate": 0.0002, "epoch": 3.8509417185352843, "step": 23820}, {"loss": 0.6151, "grad_norm": 0.7648707628250122, "learning_rate": 0.0002, "epoch": 3.8525584027160296, "step": 23830}, {"loss": 0.6013, "grad_norm": 0.7809253931045532, "learning_rate": 0.0002, "epoch": 3.854175086896775, "step": 23840}, {"loss": 0.6006, "grad_norm": 0.8337951898574829, "learning_rate": 0.0002, "epoch": 3.85579177107752, "step": 23850}, {"loss": 0.6456, "grad_norm": 0.9271913170814514, "learning_rate": 0.0002, "epoch": 3.8574084552582653, "step": 23860}, {"loss": 0.6671, "grad_norm": 0.985334038734436, "learning_rate": 0.0002, "epoch": 3.8590251394390105, "step": 23870}, {"loss": 0.6693, "grad_norm": 0.8458583354949951, "learning_rate": 0.0002, "epoch": 3.8606418236197557, "step": 23880}, {"loss": 0.6207, "grad_norm": 1.015348196029663, "learning_rate": 0.0002, "epoch": 3.862258507800501, "step": 23890}, {"loss": 0.649, "grad_norm": 1.0121688842773438, "learning_rate": 0.0002, "epoch": 3.8638751919812466, "step": 23900}, {"loss": 0.5921, "grad_norm": 0.8883971571922302, "learning_rate": 0.0002, "epoch": 3.865491876161992, "step": 23910}, {"loss": 0.6597, "grad_norm": 1.028086543083191, "learning_rate": 0.0002, "epoch": 3.867108560342737, "step": 23920}, {"loss": 0.6654, "grad_norm": 0.9645734429359436, "learning_rate": 0.0002, "epoch": 3.8687252445234823, "step": 23930}, {"loss": 0.6328, "grad_norm": 0.8235350251197815, "learning_rate": 0.0002, "epoch": 3.8703419287042276, "step": 23940}, {"loss": 0.6387, "grad_norm": 1.0298916101455688, "learning_rate": 0.0002, "epoch": 3.871958612884973, "step": 23950}, {"loss": 0.5966, "grad_norm": 1.0063377618789673, "learning_rate": 0.0002, "epoch": 3.8735752970657185, "step": 23960}, {"loss": 0.6234, "grad_norm": 0.9230626821517944, "learning_rate": 0.0002, "epoch": 3.8751919812464637, "step": 23970}, {"loss": 0.6159, "grad_norm": 0.9243063926696777, "learning_rate": 0.0002, "epoch": 3.876808665427209, "step": 23980}, {"loss": 0.6035, "grad_norm": 1.0211291313171387, "learning_rate": 0.0002, "epoch": 3.878425349607954, "step": 23990}, {"loss": 0.6351, "grad_norm": 0.7800535559654236, "learning_rate": 0.0002, "epoch": 3.8800420337886994, "step": 24000}, {"loss": 0.7, "grad_norm": 0.7904248833656311, "learning_rate": 0.0002, "epoch": 3.8816587179694446, "step": 24010}, {"loss": 0.6516, "grad_norm": 1.1975988149642944, "learning_rate": 0.0002, "epoch": 3.88327540215019, "step": 24020}, {"loss": 0.6006, "grad_norm": 1.0626593828201294, "learning_rate": 0.0002, "epoch": 3.884892086330935, "step": 24030}, {"loss": 0.6115, "grad_norm": 0.9012193083763123, "learning_rate": 0.0002, "epoch": 3.8865087705116803, "step": 24040}, {"loss": 0.6786, "grad_norm": 1.1159172058105469, "learning_rate": 0.0002, "epoch": 3.888125454692426, "step": 24050}, {"loss": 0.6635, "grad_norm": 1.276838779449463, "learning_rate": 0.0002, "epoch": 3.889742138873171, "step": 24060}, {"loss": 0.5985, "grad_norm": 0.8467690348625183, "learning_rate": 0.0002, "epoch": 3.8913588230539164, "step": 24070}, {"loss": 0.6655, "grad_norm": 0.9862841963768005, "learning_rate": 0.0002, "epoch": 3.8929755072346617, "step": 24080}, {"loss": 0.6098, "grad_norm": 0.7134621739387512, "learning_rate": 0.0002, "epoch": 3.894592191415407, "step": 24090}, {"loss": 0.618, "grad_norm": 0.8178175091743469, "learning_rate": 0.0002, "epoch": 3.896208875596152, "step": 24100}, {"loss": 0.6147, "grad_norm": 0.9229172468185425, "learning_rate": 0.0002, "epoch": 3.897825559776898, "step": 24110}, {"loss": 0.6554, "grad_norm": 1.0878316164016724, "learning_rate": 0.0002, "epoch": 3.899442243957643, "step": 24120}, {"loss": 0.6616, "grad_norm": 0.971645712852478, "learning_rate": 0.0002, "epoch": 3.9010589281383883, "step": 24130}, {"loss": 0.6228, "grad_norm": 0.8862188458442688, "learning_rate": 0.0002, "epoch": 3.9026756123191335, "step": 24140}, {"loss": 0.6192, "grad_norm": 0.9126982688903809, "learning_rate": 0.0002, "epoch": 3.9042922964998787, "step": 24150}, {"loss": 0.6734, "grad_norm": 0.8833470940589905, "learning_rate": 0.0002, "epoch": 3.905908980680624, "step": 24160}, {"loss": 0.5832, "grad_norm": 0.8320947885513306, "learning_rate": 0.0002, "epoch": 3.907525664861369, "step": 24170}, {"loss": 0.6247, "grad_norm": 0.9156602025032043, "learning_rate": 0.0002, "epoch": 3.9091423490421144, "step": 24180}, {"loss": 0.6678, "grad_norm": 1.029181957244873, "learning_rate": 0.0002, "epoch": 3.9107590332228597, "step": 24190}, {"loss": 0.6565, "grad_norm": 0.9052802324295044, "learning_rate": 0.0002, "epoch": 3.9123757174036053, "step": 24200}, {"loss": 0.6346, "grad_norm": 0.8847255110740662, "learning_rate": 0.0002, "epoch": 3.9139924015843506, "step": 24210}, {"loss": 0.6343, "grad_norm": 0.9642062187194824, "learning_rate": 0.0002, "epoch": 3.915609085765096, "step": 24220}, {"loss": 0.6557, "grad_norm": 0.8629093766212463, "learning_rate": 0.0002, "epoch": 3.917225769945841, "step": 24230}, {"loss": 0.6086, "grad_norm": 0.8674976825714111, "learning_rate": 0.0002, "epoch": 3.9188424541265863, "step": 24240}, {"loss": 0.5874, "grad_norm": 1.104846477508545, "learning_rate": 0.0002, "epoch": 3.9204591383073315, "step": 24250}, {"loss": 0.6501, "grad_norm": 1.0874955654144287, "learning_rate": 0.0002, "epoch": 3.922075822488077, "step": 24260}, {"loss": 0.6455, "grad_norm": 0.8689812421798706, "learning_rate": 0.0002, "epoch": 3.9236925066688224, "step": 24270}, {"loss": 0.5893, "grad_norm": 0.9724617004394531, "learning_rate": 0.0002, "epoch": 3.9253091908495676, "step": 24280}, {"loss": 0.6616, "grad_norm": 0.9165538549423218, "learning_rate": 0.0002, "epoch": 3.926925875030313, "step": 24290}, {"loss": 0.645, "grad_norm": 0.9307710528373718, "learning_rate": 0.0002, "epoch": 3.928542559211058, "step": 24300}, {"loss": 0.6071, "grad_norm": 0.8589295148849487, "learning_rate": 0.0002, "epoch": 3.9301592433918033, "step": 24310}, {"loss": 0.6662, "grad_norm": 0.9151099920272827, "learning_rate": 0.0002, "epoch": 3.9317759275725486, "step": 24320}, {"loss": 0.7075, "grad_norm": 0.9633517265319824, "learning_rate": 0.0002, "epoch": 3.933392611753294, "step": 24330}, {"loss": 0.6432, "grad_norm": 0.9521116018295288, "learning_rate": 0.0002, "epoch": 3.935009295934039, "step": 24340}, {"loss": 0.6457, "grad_norm": 0.8366776704788208, "learning_rate": 0.0002, "epoch": 3.9366259801147847, "step": 24350}, {"loss": 0.6139, "grad_norm": 0.8972663283348083, "learning_rate": 0.0002, "epoch": 3.93824266429553, "step": 24360}, {"loss": 0.661, "grad_norm": 0.8102919459342957, "learning_rate": 0.0002, "epoch": 3.939859348476275, "step": 24370}, {"loss": 0.6388, "grad_norm": 0.8189975023269653, "learning_rate": 0.0002, "epoch": 3.9414760326570204, "step": 24380}, {"loss": 0.6818, "grad_norm": 0.9569464921951294, "learning_rate": 0.0002, "epoch": 3.9430927168377656, "step": 24390}, {"loss": 0.6999, "grad_norm": 0.7459101676940918, "learning_rate": 0.0002, "epoch": 3.9447094010185113, "step": 24400}, {"loss": 0.6069, "grad_norm": 0.8536974787712097, "learning_rate": 0.0002, "epoch": 3.9463260851992565, "step": 24410}, {"loss": 0.5683, "grad_norm": 0.8763698935508728, "learning_rate": 0.0002, "epoch": 3.9479427693800018, "step": 24420}, {"loss": 0.6478, "grad_norm": 0.9381106495857239, "learning_rate": 0.0002, "epoch": 3.949559453560747, "step": 24430}, {"loss": 0.6371, "grad_norm": 0.934440016746521, "learning_rate": 0.0002, "epoch": 3.9511761377414922, "step": 24440}, {"loss": 0.6393, "grad_norm": 0.903918981552124, "learning_rate": 0.0002, "epoch": 3.9527928219222375, "step": 24450}, {"loss": 0.6175, "grad_norm": 0.8771953582763672, "learning_rate": 0.0002, "epoch": 3.9544095061029827, "step": 24460}, {"loss": 0.6971, "grad_norm": 1.0375410318374634, "learning_rate": 0.0002, "epoch": 3.956026190283728, "step": 24470}, {"loss": 0.6313, "grad_norm": 0.9439185261726379, "learning_rate": 0.0002, "epoch": 3.957642874464473, "step": 24480}, {"loss": 0.6076, "grad_norm": 0.935467004776001, "learning_rate": 0.0002, "epoch": 3.9592595586452184, "step": 24490}, {"loss": 0.6437, "grad_norm": 0.6900772452354431, "learning_rate": 0.0002, "epoch": 3.960876242825964, "step": 24500}, {"loss": 0.6445, "grad_norm": 1.0172916650772095, "learning_rate": 0.0002, "epoch": 3.9624929270067093, "step": 24510}, {"loss": 0.6308, "grad_norm": 0.9167046546936035, "learning_rate": 0.0002, "epoch": 3.9641096111874545, "step": 24520}, {"loss": 0.6519, "grad_norm": 0.7230527997016907, "learning_rate": 0.0002, "epoch": 3.9657262953681998, "step": 24530}, {"loss": 0.6564, "grad_norm": 0.8980403542518616, "learning_rate": 0.0002, "epoch": 3.967342979548945, "step": 24540}, {"loss": 0.6099, "grad_norm": 0.8555465936660767, "learning_rate": 0.0002, "epoch": 3.9689596637296907, "step": 24550}, {"loss": 0.6617, "grad_norm": 0.7825445532798767, "learning_rate": 0.0002, "epoch": 3.970576347910436, "step": 24560}, {"loss": 0.604, "grad_norm": 0.7273133993148804, "learning_rate": 0.0002, "epoch": 3.972193032091181, "step": 24570}, {"loss": 0.6427, "grad_norm": 0.9612047672271729, "learning_rate": 0.0002, "epoch": 3.9738097162719264, "step": 24580}, {"loss": 0.6426, "grad_norm": 0.9865460991859436, "learning_rate": 0.0002, "epoch": 3.9754264004526716, "step": 24590}, {"loss": 0.6052, "grad_norm": 0.8638762831687927, "learning_rate": 0.0002, "epoch": 3.977043084633417, "step": 24600}, {"loss": 0.6097, "grad_norm": 1.0096198320388794, "learning_rate": 0.0002, "epoch": 3.978659768814162, "step": 24610}, {"loss": 0.6664, "grad_norm": 0.8475532531738281, "learning_rate": 0.0002, "epoch": 3.9802764529949073, "step": 24620}, {"loss": 0.6711, "grad_norm": 0.9696195721626282, "learning_rate": 0.0002, "epoch": 3.9818931371756525, "step": 24630}, {"loss": 0.6446, "grad_norm": 0.7499843239784241, "learning_rate": 0.0002, "epoch": 3.9835098213563978, "step": 24640}, {"loss": 0.6054, "grad_norm": 0.8865424990653992, "learning_rate": 0.0002, "epoch": 3.9851265055371434, "step": 24650}, {"loss": 0.5975, "grad_norm": 0.8089959025382996, "learning_rate": 0.0002, "epoch": 3.9867431897178887, "step": 24660}, {"loss": 0.6677, "grad_norm": 0.6946012377738953, "learning_rate": 0.0002, "epoch": 3.988359873898634, "step": 24670}, {"loss": 0.6329, "grad_norm": 0.7991759181022644, "learning_rate": 0.0002, "epoch": 3.989976558079379, "step": 24680}, {"loss": 0.6449, "grad_norm": 0.8803931474685669, "learning_rate": 0.0002, "epoch": 3.9915932422601244, "step": 24690}, {"loss": 0.7091, "grad_norm": 0.8848299980163574, "learning_rate": 0.0002, "epoch": 3.99320992644087, "step": 24700}, {"loss": 0.6551, "grad_norm": 0.7448889017105103, "learning_rate": 0.0002, "epoch": 3.9948266106216153, "step": 24710}, {"loss": 0.6432, "grad_norm": 0.9361620545387268, "learning_rate": 0.0002, "epoch": 3.9964432948023605, "step": 24720}, {"loss": 0.5917, "grad_norm": 0.9958081245422363, "learning_rate": 0.0002, "epoch": 3.9980599789831057, "step": 24730}, {"loss": 0.6567, "grad_norm": 1.026004672050476, "learning_rate": 0.0002, "epoch": 3.999676663163851, "step": 24740}]}
+{"epoch": 4.9999191657909625, "step": 30927, "epoch_duration": 16879.647474765778, "total_accumulated_duration": 84415.05347633362, "gpu_info": {"GPU_0": "NVIDIA A100-PCIE-40GB"}, "memory_usage": {"avg_memory_usage": {"GPU_0": 7672.4541015625}, "peak_memory_usage": {"GPU_0": 13792.75537109375}, "avg_memory_reserved": {"GPU_0": 17416.0}, "peak_memory_reserved": {"GPU_0": 17416.0}, "total_memory": {"GPU_0": 40444.375}}, "best_checkpoint_path": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185", "params": {"epochs": 8, "batch_size": 1, "learning_rate": 0.0002, "gradient_accumulation_steps": 2, "warmup_ratio": 0.03, "max_grad_norm": 0.3, "lora_alpha": 32, "lora_dropout": 0.05, "lora_r": 64, "tf32": true, "seed": 42}, "log_history": [{"loss": 1.6636, "grad_norm": 0.9894065856933594, "learning_rate": 0.0002, "epoch": 0.0016166841807452913, "step": 10}, {"loss": 1.1528, "grad_norm": 1.7810699939727783, "learning_rate": 0.0002, "epoch": 0.0032333683614905826, "step": 20}, {"loss": 0.9767, "grad_norm": 0.5969577431678772, "learning_rate": 0.0002, "epoch": 0.004850052542235874, "step": 30}, {"loss": 0.9772, "grad_norm": 0.6354120969772339, "learning_rate": 0.0002, "epoch": 0.006466736722981165, "step": 40}, {"loss": 0.8643, "grad_norm": 0.5604607462882996, "learning_rate": 0.0002, "epoch": 0.008083420903726457, "step": 50}, {"loss": 0.8841, "grad_norm": 0.4676193594932556, "learning_rate": 0.0002, "epoch": 0.009700105084471748, "step": 60}, {"loss": 0.9022, "grad_norm": 0.6099211573600769, "learning_rate": 0.0002, "epoch": 0.01131678926521704, "step": 70}, {"loss": 0.9133, "grad_norm": 0.48639994859695435, "learning_rate": 0.0002, "epoch": 0.01293347344596233, "step": 80}, {"loss": 0.8704, "grad_norm": 0.4904264509677887, "learning_rate": 0.0002, "epoch": 0.014550157626707623, "step": 90}, {"loss": 0.8855, "grad_norm": 2.8334362506866455, "learning_rate": 0.0002, "epoch": 0.016166841807452915, "step": 100}, {"loss": 0.8958, "grad_norm": 0.43221670389175415, "learning_rate": 0.0002, "epoch": 0.017783525988198205, "step": 110}, {"loss": 0.8412, "grad_norm": 0.42244166135787964, "learning_rate": 0.0002, "epoch": 0.019400210168943496, "step": 120}, {"loss": 0.8467, "grad_norm": 0.45363298058509827, "learning_rate": 0.0002, "epoch": 0.02101689434968879, "step": 130}, {"loss": 0.8641, "grad_norm": 0.44816508889198303, "learning_rate": 0.0002, "epoch": 0.02263357853043408, "step": 140}, {"loss": 0.8496, "grad_norm": 0.43308213353157043, "learning_rate": 0.0002, "epoch": 0.02425026271117937, "step": 150}, {"loss": 0.8213, "grad_norm": 0.4084763526916504, "learning_rate": 0.0002, "epoch": 0.02586694689192466, "step": 160}, {"loss": 0.8343, "grad_norm": 0.5363703966140747, "learning_rate": 0.0002, "epoch": 0.027483631072669955, "step": 170}, {"loss": 0.8558, "grad_norm": 0.4619699716567993, "learning_rate": 0.0002, "epoch": 0.029100315253415245, "step": 180}, {"loss": 0.8878, "grad_norm": 0.49069908261299133, "learning_rate": 0.0002, "epoch": 0.030716999434160536, "step": 190}, {"loss": 0.8867, "grad_norm": 0.4645835757255554, "learning_rate": 0.0002, "epoch": 0.03233368361490583, "step": 200}, {"loss": 0.8842, "grad_norm": 1.2411243915557861, "learning_rate": 0.0002, "epoch": 0.03395036779565112, "step": 210}, {"loss": 0.8245, "grad_norm": 0.5211851596832275, "learning_rate": 0.0002, "epoch": 0.03556705197639641, "step": 220}, {"loss": 0.8194, "grad_norm": 0.5253691673278809, "learning_rate": 0.0002, "epoch": 0.037183736157141704, "step": 230}, {"loss": 0.8856, "grad_norm": 0.4567478895187378, "learning_rate": 0.0002, "epoch": 0.03880042033788699, "step": 240}, {"loss": 0.838, "grad_norm": 0.5472128391265869, "learning_rate": 0.0002, "epoch": 0.040417104518632285, "step": 250}, {"loss": 0.8201, "grad_norm": 0.42978546023368835, "learning_rate": 0.0002, "epoch": 0.04203378869937758, "step": 260}, {"loss": 0.8334, "grad_norm": 0.601734459400177, "learning_rate": 0.0002, "epoch": 0.043650472880122866, "step": 270}, {"loss": 0.815, "grad_norm": 0.4286513328552246, "learning_rate": 0.0002, "epoch": 0.04526715706086816, "step": 280}, {"loss": 0.8758, "grad_norm": 0.5230861902236938, "learning_rate": 0.0002, "epoch": 0.046883841241613454, "step": 290}, {"loss": 0.8636, "grad_norm": 0.6504611968994141, "learning_rate": 0.0002, "epoch": 0.04850052542235874, "step": 300}, {"loss": 0.8102, "grad_norm": 0.43485215306282043, "learning_rate": 0.0002, "epoch": 0.050117209603104035, "step": 310}, {"loss": 0.8221, "grad_norm": 0.4717007875442505, "learning_rate": 0.0002, "epoch": 0.05173389378384932, "step": 320}, {"loss": 0.8469, "grad_norm": 0.4059787690639496, "learning_rate": 0.0002, "epoch": 0.053350577964594616, "step": 330}, {"loss": 0.8866, "grad_norm": 0.4366913437843323, "learning_rate": 0.0002, "epoch": 0.05496726214533991, "step": 340}, {"loss": 0.7976, "grad_norm": 0.4233848452568054, "learning_rate": 0.0002, "epoch": 0.0565839463260852, "step": 350}, {"loss": 0.8456, "grad_norm": 0.4209108352661133, "learning_rate": 0.0002, "epoch": 0.05820063050683049, "step": 360}, {"loss": 0.816, "grad_norm": 0.41637396812438965, "learning_rate": 0.0002, "epoch": 0.059817314687575784, "step": 370}, {"loss": 0.7976, "grad_norm": 0.46235376596450806, "learning_rate": 0.0002, "epoch": 0.06143399886832107, "step": 380}, {"loss": 0.7966, "grad_norm": 0.4013484716415405, "learning_rate": 0.0002, "epoch": 0.06305068304906636, "step": 390}, {"loss": 0.8253, "grad_norm": 0.47443896532058716, "learning_rate": 0.0002, "epoch": 0.06466736722981166, "step": 400}, {"loss": 0.8666, "grad_norm": 0.3942156434059143, "learning_rate": 0.0002, "epoch": 0.06628405141055695, "step": 410}, {"loss": 0.8402, "grad_norm": 0.4965320825576782, "learning_rate": 0.0002, "epoch": 0.06790073559130223, "step": 420}, {"loss": 0.8317, "grad_norm": 0.4304835796356201, "learning_rate": 0.0002, "epoch": 0.06951741977204753, "step": 430}, {"loss": 0.8528, "grad_norm": 0.511726975440979, "learning_rate": 0.0002, "epoch": 0.07113410395279282, "step": 440}, {"loss": 0.8675, "grad_norm": 0.4040689170360565, "learning_rate": 0.0002, "epoch": 0.07275078813353811, "step": 450}, {"loss": 0.8788, "grad_norm": 0.5402171015739441, "learning_rate": 0.0002, "epoch": 0.07436747231428341, "step": 460}, {"loss": 0.8737, "grad_norm": 0.4174517095088959, "learning_rate": 0.0002, "epoch": 0.0759841564950287, "step": 470}, {"loss": 0.7605, "grad_norm": 0.4306182265281677, "learning_rate": 0.0002, "epoch": 0.07760084067577398, "step": 480}, {"loss": 0.799, "grad_norm": 0.535210132598877, "learning_rate": 0.0002, "epoch": 0.07921752485651928, "step": 490}, {"loss": 0.7825, "grad_norm": 0.5339109897613525, "learning_rate": 0.0002, "epoch": 0.08083420903726457, "step": 500}, {"loss": 0.8985, "grad_norm": 0.45754891633987427, "learning_rate": 0.0002, "epoch": 0.08245089321800986, "step": 510}, {"loss": 0.8144, "grad_norm": 0.43820783495903015, "learning_rate": 0.0002, "epoch": 0.08406757739875516, "step": 520}, {"loss": 0.8001, "grad_norm": 0.4434749186038971, "learning_rate": 0.0002, "epoch": 0.08568426157950045, "step": 530}, {"loss": 0.7857, "grad_norm": 0.43111467361450195, "learning_rate": 0.0002, "epoch": 0.08730094576024573, "step": 540}, {"loss": 0.8418, "grad_norm": 0.4378940165042877, "learning_rate": 0.0002, "epoch": 0.08891762994099103, "step": 550}, {"loss": 0.8361, "grad_norm": 0.4772215187549591, "learning_rate": 0.0002, "epoch": 0.09053431412173632, "step": 560}, {"loss": 0.8268, "grad_norm": 0.6837629079818726, "learning_rate": 0.0002, "epoch": 0.09215099830248161, "step": 570}, {"loss": 0.8607, "grad_norm": 0.42241212725639343, "learning_rate": 0.0002, "epoch": 0.09376768248322691, "step": 580}, {"loss": 0.852, "grad_norm": 0.5165936350822449, "learning_rate": 0.0002, "epoch": 0.0953843666639722, "step": 590}, {"loss": 0.8664, "grad_norm": 0.48737478256225586, "learning_rate": 0.0002, "epoch": 0.09700105084471748, "step": 600}, {"loss": 0.8806, "grad_norm": 0.47419852018356323, "learning_rate": 0.0002, "epoch": 0.09861773502546278, "step": 610}, {"loss": 0.8254, "grad_norm": 0.4975486099720001, "learning_rate": 0.0002, "epoch": 0.10023441920620807, "step": 620}, {"loss": 0.8548, "grad_norm": 0.49123844504356384, "learning_rate": 0.0002, "epoch": 0.10185110338695336, "step": 630}, {"loss": 0.8911, "grad_norm": 0.6288952827453613, "learning_rate": 0.0002, "epoch": 0.10346778756769864, "step": 640}, {"loss": 0.827, "grad_norm": 0.4277345836162567, "learning_rate": 0.0002, "epoch": 0.10508447174844394, "step": 650}, {"loss": 0.7996, "grad_norm": 0.4021061956882477, "learning_rate": 0.0002, "epoch": 0.10670115592918923, "step": 660}, {"loss": 0.87, "grad_norm": 0.3492237329483032, "learning_rate": 0.0002, "epoch": 0.10831784010993452, "step": 670}, {"loss": 0.8698, "grad_norm": 0.4341012239456177, "learning_rate": 0.0002, "epoch": 0.10993452429067982, "step": 680}, {"loss": 0.781, "grad_norm": 0.7296304106712341, "learning_rate": 0.0002, "epoch": 0.1115512084714251, "step": 690}, {"loss": 0.8433, "grad_norm": 0.397494912147522, "learning_rate": 0.0002, "epoch": 0.1131678926521704, "step": 700}, {"loss": 0.827, "grad_norm": 0.396431028842926, "learning_rate": 0.0002, "epoch": 0.1147845768329157, "step": 710}, {"loss": 0.8379, "grad_norm": 0.48842838406562805, "learning_rate": 0.0002, "epoch": 0.11640126101366098, "step": 720}, {"loss": 0.8238, "grad_norm": 0.46322616934776306, "learning_rate": 0.0002, "epoch": 0.11801794519440627, "step": 730}, {"loss": 0.8041, "grad_norm": 0.47990912199020386, "learning_rate": 0.0002, "epoch": 0.11963462937515157, "step": 740}, {"loss": 0.82, "grad_norm": 0.4997142255306244, "learning_rate": 0.0002, "epoch": 0.12125131355589686, "step": 750}, {"loss": 0.7702, "grad_norm": 0.4040526747703552, "learning_rate": 0.0002, "epoch": 0.12286799773664214, "step": 760}, {"loss": 0.863, "grad_norm": 0.453095942735672, "learning_rate": 0.0002, "epoch": 0.12448468191738744, "step": 770}, {"loss": 0.8792, "grad_norm": 0.4636971950531006, "learning_rate": 0.0002, "epoch": 0.12610136609813272, "step": 780}, {"loss": 0.8112, "grad_norm": 0.4279276132583618, "learning_rate": 0.0002, "epoch": 0.12771805027887803, "step": 790}, {"loss": 0.8711, "grad_norm": 0.46212655305862427, "learning_rate": 0.0002, "epoch": 0.12933473445962332, "step": 800}, {"loss": 0.8368, "grad_norm": 0.43127650022506714, "learning_rate": 0.0002, "epoch": 0.1309514186403686, "step": 810}, {"loss": 0.8476, "grad_norm": 0.4201301336288452, "learning_rate": 0.0002, "epoch": 0.1325681028211139, "step": 820}, {"loss": 0.8078, "grad_norm": 0.42583167552948, "learning_rate": 0.0002, "epoch": 0.13418478700185918, "step": 830}, {"loss": 0.8219, "grad_norm": 0.4535622000694275, "learning_rate": 0.0002, "epoch": 0.13580147118260447, "step": 840}, {"loss": 0.8423, "grad_norm": 0.4116036891937256, "learning_rate": 0.0002, "epoch": 0.13741815536334978, "step": 850}, {"loss": 0.8466, "grad_norm": 0.45997580885887146, "learning_rate": 0.0002, "epoch": 0.13903483954409507, "step": 860}, {"loss": 0.8917, "grad_norm": 0.4487837255001068, "learning_rate": 0.0002, "epoch": 0.14065152372484035, "step": 870}, {"loss": 0.8217, "grad_norm": 0.43650057911872864, "learning_rate": 0.0002, "epoch": 0.14226820790558564, "step": 880}, {"loss": 0.8178, "grad_norm": 0.5335358381271362, "learning_rate": 0.0002, "epoch": 0.14388489208633093, "step": 890}, {"loss": 0.7957, "grad_norm": 0.5989000201225281, "learning_rate": 0.0002, "epoch": 0.14550157626707622, "step": 900}, {"loss": 0.8385, "grad_norm": 0.517179012298584, "learning_rate": 0.0002, "epoch": 0.14711826044782153, "step": 910}, {"loss": 0.8255, "grad_norm": 0.44435232877731323, "learning_rate": 0.0002, "epoch": 0.14873494462856682, "step": 920}, {"loss": 0.8305, "grad_norm": 0.42635923624038696, "learning_rate": 0.0002, "epoch": 0.1503516288093121, "step": 930}, {"loss": 0.8043, "grad_norm": 0.49603334069252014, "learning_rate": 0.0002, "epoch": 0.1519683129900574, "step": 940}, {"loss": 0.8377, "grad_norm": 0.40639808773994446, "learning_rate": 0.0002, "epoch": 0.15358499717080268, "step": 950}, {"loss": 0.8529, "grad_norm": 0.4850759208202362, "learning_rate": 0.0002, "epoch": 0.15520168135154797, "step": 960}, {"loss": 0.846, "grad_norm": 0.4427442252635956, "learning_rate": 0.0002, "epoch": 0.15681836553229328, "step": 970}, {"loss": 0.8705, "grad_norm": 0.3760930001735687, "learning_rate": 0.0002, "epoch": 0.15843504971303857, "step": 980}, {"loss": 0.8644, "grad_norm": 0.4794144332408905, "learning_rate": 0.0002, "epoch": 0.16005173389378385, "step": 990}, {"loss": 0.8002, "grad_norm": 0.45828768610954285, "learning_rate": 0.0002, "epoch": 0.16166841807452914, "step": 1000}, {"loss": 0.7658, "grad_norm": 0.6313053369522095, "learning_rate": 0.0002, "epoch": 0.16328510225527443, "step": 1010}, {"loss": 0.8047, "grad_norm": 0.45041006803512573, "learning_rate": 0.0002, "epoch": 0.16490178643601971, "step": 1020}, {"loss": 0.8423, "grad_norm": 0.441403865814209, "learning_rate": 0.0002, "epoch": 0.166518470616765, "step": 1030}, {"loss": 0.8475, "grad_norm": 0.8171296119689941, "learning_rate": 0.0002, "epoch": 0.16813515479751032, "step": 1040}, {"loss": 0.845, "grad_norm": 0.7137420773506165, "learning_rate": 0.0002, "epoch": 0.1697518389782556, "step": 1050}, {"loss": 0.8213, "grad_norm": 0.5236809849739075, "learning_rate": 0.0002, "epoch": 0.1713685231590009, "step": 1060}, {"loss": 0.8265, "grad_norm": 0.5021864175796509, "learning_rate": 0.0002, "epoch": 0.17298520733974618, "step": 1070}, {"loss": 0.8305, "grad_norm": 0.47347521781921387, "learning_rate": 0.0002, "epoch": 0.17460189152049146, "step": 1080}, {"loss": 0.8105, "grad_norm": 0.4631653428077698, "learning_rate": 0.0002, "epoch": 0.17621857570123675, "step": 1090}, {"loss": 0.8166, "grad_norm": 0.49169182777404785, "learning_rate": 0.0002, "epoch": 0.17783525988198207, "step": 1100}, {"loss": 0.8012, "grad_norm": 0.5019739270210266, "learning_rate": 0.0002, "epoch": 0.17945194406272735, "step": 1110}, {"loss": 0.8247, "grad_norm": 0.5100422501564026, "learning_rate": 0.0002, "epoch": 0.18106862824347264, "step": 1120}, {"loss": 0.8142, "grad_norm": 0.3888324499130249, "learning_rate": 0.0002, "epoch": 0.18268531242421793, "step": 1130}, {"loss": 0.8533, "grad_norm": 0.39765217900276184, "learning_rate": 0.0002, "epoch": 0.18430199660496321, "step": 1140}, {"loss": 0.8541, "grad_norm": 0.47190186381340027, "learning_rate": 0.0002, "epoch": 0.1859186807857085, "step": 1150}, {"loss": 0.8301, "grad_norm": 0.4464188814163208, "learning_rate": 0.0002, "epoch": 0.18753536496645382, "step": 1160}, {"loss": 0.8341, "grad_norm": 0.5153930187225342, "learning_rate": 0.0002, "epoch": 0.1891520491471991, "step": 1170}, {"loss": 0.8033, "grad_norm": 0.4779708683490753, "learning_rate": 0.0002, "epoch": 0.1907687333279444, "step": 1180}, {"loss": 0.8187, "grad_norm": 0.4834315776824951, "learning_rate": 0.0002, "epoch": 0.19238541750868968, "step": 1190}, {"loss": 0.7721, "grad_norm": 0.402357816696167, "learning_rate": 0.0002, "epoch": 0.19400210168943496, "step": 1200}, {"loss": 0.7941, "grad_norm": 0.45899084210395813, "learning_rate": 0.0002, "epoch": 0.19561878587018025, "step": 1210}, {"loss": 0.8353, "grad_norm": 0.5106529593467712, "learning_rate": 0.0002, "epoch": 0.19723547005092557, "step": 1220}, {"loss": 0.7816, "grad_norm": 0.45261722803115845, "learning_rate": 0.0002, "epoch": 0.19885215423167085, "step": 1230}, {"loss": 0.8068, "grad_norm": 0.4647127091884613, "learning_rate": 0.0002, "epoch": 0.20046883841241614, "step": 1240}, {"loss": 0.8239, "grad_norm": 0.4849368929862976, "learning_rate": 0.0002, "epoch": 0.20208552259316143, "step": 1250}, {"loss": 0.8514, "grad_norm": 0.4518061578273773, "learning_rate": 0.0002, "epoch": 0.2037022067739067, "step": 1260}, {"loss": 0.8158, "grad_norm": 0.49535325169563293, "learning_rate": 0.0002, "epoch": 0.205318890954652, "step": 1270}, {"loss": 0.8348, "grad_norm": 0.4835205376148224, "learning_rate": 0.0002, "epoch": 0.2069355751353973, "step": 1280}, {"loss": 0.8428, "grad_norm": 0.45308539271354675, "learning_rate": 0.0002, "epoch": 0.2085522593161426, "step": 1290}, {"loss": 0.7993, "grad_norm": 0.5369905233383179, "learning_rate": 0.0002, "epoch": 0.2101689434968879, "step": 1300}, {"loss": 0.8676, "grad_norm": 0.5031622052192688, "learning_rate": 0.0002, "epoch": 0.21178562767763318, "step": 1310}, {"loss": 0.7686, "grad_norm": 0.48010334372520447, "learning_rate": 0.0002, "epoch": 0.21340231185837846, "step": 1320}, {"loss": 0.806, "grad_norm": 0.4905701279640198, "learning_rate": 0.0002, "epoch": 0.21501899603912375, "step": 1330}, {"loss": 0.7885, "grad_norm": 0.43531742691993713, "learning_rate": 0.0002, "epoch": 0.21663568021986904, "step": 1340}, {"loss": 0.8191, "grad_norm": 0.44330692291259766, "learning_rate": 0.0002, "epoch": 0.21825236440061435, "step": 1350}, {"loss": 0.8205, "grad_norm": 0.5384416580200195, "learning_rate": 0.0002, "epoch": 0.21986904858135964, "step": 1360}, {"loss": 0.7726, "grad_norm": 0.4181833863258362, "learning_rate": 0.0002, "epoch": 0.22148573276210493, "step": 1370}, {"loss": 0.8311, "grad_norm": 0.523833692073822, "learning_rate": 0.0002, "epoch": 0.2231024169428502, "step": 1380}, {"loss": 0.7913, "grad_norm": 0.5528736710548401, "learning_rate": 0.0002, "epoch": 0.2247191011235955, "step": 1390}, {"loss": 0.8079, "grad_norm": 0.43515023589134216, "learning_rate": 0.0002, "epoch": 0.2263357853043408, "step": 1400}, {"loss": 0.8403, "grad_norm": 0.48809877038002014, "learning_rate": 0.0002, "epoch": 0.2279524694850861, "step": 1410}, {"loss": 0.8165, "grad_norm": 0.43591251969337463, "learning_rate": 0.0002, "epoch": 0.2295691536658314, "step": 1420}, {"loss": 0.8147, "grad_norm": 0.44625312089920044, "learning_rate": 0.0002, "epoch": 0.23118583784657668, "step": 1430}, {"loss": 0.8134, "grad_norm": 0.4390665292739868, "learning_rate": 0.0002, "epoch": 0.23280252202732196, "step": 1440}, {"loss": 0.8465, "grad_norm": 0.48496049642562866, "learning_rate": 0.0002, "epoch": 0.23441920620806725, "step": 1450}, {"loss": 0.775, "grad_norm": 0.45919957756996155, "learning_rate": 0.0002, "epoch": 0.23603589038881254, "step": 1460}, {"loss": 0.8659, "grad_norm": 0.5471845865249634, "learning_rate": 0.0002, "epoch": 0.23765257456955785, "step": 1470}, {"loss": 0.8164, "grad_norm": 0.47269317507743835, "learning_rate": 0.0002, "epoch": 0.23926925875030314, "step": 1480}, {"loss": 0.854, "grad_norm": 0.4930245578289032, "learning_rate": 0.0002, "epoch": 0.24088594293104842, "step": 1490}, {"loss": 0.8139, "grad_norm": 0.5605630278587341, "learning_rate": 0.0002, "epoch": 0.2425026271117937, "step": 1500}, {"loss": 0.8125, "grad_norm": 0.4435870945453644, "learning_rate": 0.0002, "epoch": 0.244119311292539, "step": 1510}, {"loss": 0.8123, "grad_norm": 0.4941999912261963, "learning_rate": 0.0002, "epoch": 0.24573599547328429, "step": 1520}, {"loss": 0.8427, "grad_norm": 0.5100624561309814, "learning_rate": 0.0002, "epoch": 0.24735267965402957, "step": 1530}, {"loss": 0.8405, "grad_norm": 0.4638267457485199, "learning_rate": 0.0002, "epoch": 0.2489693638347749, "step": 1540}, {"loss": 0.81, "grad_norm": 0.5071570873260498, "learning_rate": 0.0002, "epoch": 0.25058604801552015, "step": 1550}, {"loss": 0.7724, "grad_norm": 0.4291319251060486, "learning_rate": 0.0002, "epoch": 0.25220273219626543, "step": 1560}, {"loss": 0.7984, "grad_norm": 0.5388049483299255, "learning_rate": 0.0002, "epoch": 0.2538194163770108, "step": 1570}, {"loss": 0.8176, "grad_norm": 0.5083683729171753, "learning_rate": 0.0002, "epoch": 0.25543610055775606, "step": 1580}, {"loss": 0.843, "grad_norm": 0.4824463725090027, "learning_rate": 0.0002, "epoch": 0.25705278473850135, "step": 1590}, {"loss": 0.7996, "grad_norm": 0.41177722811698914, "learning_rate": 0.0002, "epoch": 0.25866946891924664, "step": 1600}, {"loss": 0.7772, "grad_norm": 0.5656219124794006, "learning_rate": 0.0002, "epoch": 0.2602861530999919, "step": 1610}, {"loss": 0.7955, "grad_norm": 0.41063204407691956, "learning_rate": 0.0002, "epoch": 0.2619028372807372, "step": 1620}, {"loss": 0.7998, "grad_norm": 0.4897061288356781, "learning_rate": 0.0002, "epoch": 0.2635195214614825, "step": 1630}, {"loss": 0.8198, "grad_norm": 0.4454376697540283, "learning_rate": 0.0002, "epoch": 0.2651362056422278, "step": 1640}, {"loss": 0.8684, "grad_norm": 0.4355238378047943, "learning_rate": 0.0002, "epoch": 0.26675288982297307, "step": 1650}, {"loss": 0.7801, "grad_norm": 0.458310067653656, "learning_rate": 0.0002, "epoch": 0.26836957400371836, "step": 1660}, {"loss": 0.7935, "grad_norm": 0.4752083718776703, "learning_rate": 0.0002, "epoch": 0.26998625818446365, "step": 1670}, {"loss": 0.8267, "grad_norm": 0.4666106402873993, "learning_rate": 0.0002, "epoch": 0.27160294236520893, "step": 1680}, {"loss": 0.8252, "grad_norm": 0.4213818609714508, "learning_rate": 0.0002, "epoch": 0.2732196265459543, "step": 1690}, {"loss": 0.8559, "grad_norm": 0.5768913626670837, "learning_rate": 0.0002, "epoch": 0.27483631072669956, "step": 1700}, {"loss": 0.7931, "grad_norm": 0.4209914803504944, "learning_rate": 0.0002, "epoch": 0.27645299490744485, "step": 1710}, {"loss": 0.8167, "grad_norm": 0.501909613609314, "learning_rate": 0.0002, "epoch": 0.27806967908819014, "step": 1720}, {"loss": 0.7832, "grad_norm": 0.5266261100769043, "learning_rate": 0.0002, "epoch": 0.2796863632689354, "step": 1730}, {"loss": 0.8102, "grad_norm": 0.43806859850883484, "learning_rate": 0.0002, "epoch": 0.2813030474496807, "step": 1740}, {"loss": 0.8157, "grad_norm": 0.46048814058303833, "learning_rate": 0.0002, "epoch": 0.282919731630426, "step": 1750}, {"loss": 0.8596, "grad_norm": 0.44972819089889526, "learning_rate": 0.0002, "epoch": 0.2845364158111713, "step": 1760}, {"loss": 0.8421, "grad_norm": 0.5114831328392029, "learning_rate": 0.0002, "epoch": 0.28615309999191657, "step": 1770}, {"loss": 0.8361, "grad_norm": 0.47931742668151855, "learning_rate": 0.0002, "epoch": 0.28776978417266186, "step": 1780}, {"loss": 0.8265, "grad_norm": 0.5092599987983704, "learning_rate": 0.0002, "epoch": 0.28938646835340714, "step": 1790}, {"loss": 0.8506, "grad_norm": 0.37581443786621094, "learning_rate": 0.0002, "epoch": 0.29100315253415243, "step": 1800}, {"loss": 0.7932, "grad_norm": 0.47097381949424744, "learning_rate": 0.0002, "epoch": 0.2926198367148977, "step": 1810}, {"loss": 0.7787, "grad_norm": 0.48300236463546753, "learning_rate": 0.0002, "epoch": 0.29423652089564306, "step": 1820}, {"loss": 0.8391, "grad_norm": 0.5600419640541077, "learning_rate": 0.0002, "epoch": 0.29585320507638835, "step": 1830}, {"loss": 0.8507, "grad_norm": 0.48555272817611694, "learning_rate": 0.0002, "epoch": 0.29746988925713364, "step": 1840}, {"loss": 0.7657, "grad_norm": 0.3752668499946594, "learning_rate": 0.0002, "epoch": 0.2990865734378789, "step": 1850}, {"loss": 0.7915, "grad_norm": 0.5328747034072876, "learning_rate": 0.0002, "epoch": 0.3007032576186242, "step": 1860}, {"loss": 0.8426, "grad_norm": 0.48716455698013306, "learning_rate": 0.0002, "epoch": 0.3023199417993695, "step": 1870}, {"loss": 0.8335, "grad_norm": 0.5011493563652039, "learning_rate": 0.0002, "epoch": 0.3039366259801148, "step": 1880}, {"loss": 0.852, "grad_norm": 0.46461427211761475, "learning_rate": 0.0002, "epoch": 0.30555331016086007, "step": 1890}, {"loss": 0.8478, "grad_norm": 0.36630210280418396, "learning_rate": 0.0002, "epoch": 0.30716999434160536, "step": 1900}, {"loss": 0.8162, "grad_norm": 0.4217296242713928, "learning_rate": 0.0002, "epoch": 0.30878667852235064, "step": 1910}, {"loss": 0.8128, "grad_norm": 0.4394875466823578, "learning_rate": 0.0002, "epoch": 0.31040336270309593, "step": 1920}, {"loss": 0.8471, "grad_norm": 0.6587965488433838, "learning_rate": 0.0002, "epoch": 0.3120200468838412, "step": 1930}, {"loss": 0.8565, "grad_norm": 0.5469298958778381, "learning_rate": 0.0002, "epoch": 0.31363673106458656, "step": 1940}, {"loss": 0.8236, "grad_norm": 0.4371595084667206, "learning_rate": 0.0002, "epoch": 0.31525341524533185, "step": 1950}, {"loss": 0.887, "grad_norm": 0.4809541404247284, "learning_rate": 0.0002, "epoch": 0.31687009942607713, "step": 1960}, {"loss": 0.7855, "grad_norm": 0.6061086654663086, "learning_rate": 0.0002, "epoch": 0.3184867836068224, "step": 1970}, {"loss": 0.7679, "grad_norm": 0.5342657566070557, "learning_rate": 0.0002, "epoch": 0.3201034677875677, "step": 1980}, {"loss": 0.7955, "grad_norm": 0.5057743787765503, "learning_rate": 0.0002, "epoch": 0.321720151968313, "step": 1990}, {"loss": 0.7774, "grad_norm": 0.528626024723053, "learning_rate": 0.0002, "epoch": 0.3233368361490583, "step": 2000}, {"loss": 0.8845, "grad_norm": 0.46742770075798035, "learning_rate": 0.0002, "epoch": 0.32495352032980357, "step": 2010}, {"loss": 0.8484, "grad_norm": 0.515101432800293, "learning_rate": 0.0002, "epoch": 0.32657020451054886, "step": 2020}, {"loss": 0.8139, "grad_norm": 0.41941216588020325, "learning_rate": 0.0002, "epoch": 0.32818688869129414, "step": 2030}, {"loss": 0.7637, "grad_norm": 0.49902522563934326, "learning_rate": 0.0002, "epoch": 0.32980357287203943, "step": 2040}, {"loss": 0.7822, "grad_norm": 0.4120897650718689, "learning_rate": 0.0002, "epoch": 0.3314202570527847, "step": 2050}, {"loss": 0.8057, "grad_norm": 0.45352041721343994, "learning_rate": 0.0002, "epoch": 0.33303694123353, "step": 2060}, {"loss": 0.7913, "grad_norm": 0.523199737071991, "learning_rate": 0.0002, "epoch": 0.33465362541427535, "step": 2070}, {"loss": 0.8036, "grad_norm": 0.4390358626842499, "learning_rate": 0.0002, "epoch": 0.33627030959502063, "step": 2080}, {"loss": 0.8145, "grad_norm": 0.6752901077270508, "learning_rate": 0.0002, "epoch": 0.3378869937757659, "step": 2090}, {"loss": 0.7807, "grad_norm": 0.547821044921875, "learning_rate": 0.0002, "epoch": 0.3395036779565112, "step": 2100}, {"loss": 0.8561, "grad_norm": 0.5161308646202087, "learning_rate": 0.0002, "epoch": 0.3411203621372565, "step": 2110}, {"loss": 0.7697, "grad_norm": 0.4565401077270508, "learning_rate": 0.0002, "epoch": 0.3427370463180018, "step": 2120}, {"loss": 0.7964, "grad_norm": 0.4666115939617157, "learning_rate": 0.0002, "epoch": 0.34435373049874707, "step": 2130}, {"loss": 0.8189, "grad_norm": 0.4090428352355957, "learning_rate": 0.0002, "epoch": 0.34597041467949236, "step": 2140}, {"loss": 0.8817, "grad_norm": 0.510845422744751, "learning_rate": 0.0002, "epoch": 0.34758709886023764, "step": 2150}, {"loss": 0.8398, "grad_norm": 0.42861923575401306, "learning_rate": 0.0002, "epoch": 0.34920378304098293, "step": 2160}, {"loss": 0.7716, "grad_norm": 0.4476332664489746, "learning_rate": 0.0002, "epoch": 0.3508204672217282, "step": 2170}, {"loss": 0.7845, "grad_norm": 0.6065791249275208, "learning_rate": 0.0002, "epoch": 0.3524371514024735, "step": 2180}, {"loss": 0.8187, "grad_norm": 0.42335066199302673, "learning_rate": 0.0002, "epoch": 0.35405383558321885, "step": 2190}, {"loss": 0.8239, "grad_norm": 0.5094629526138306, "learning_rate": 0.0002, "epoch": 0.35567051976396413, "step": 2200}, {"loss": 0.7807, "grad_norm": 0.5476373434066772, "learning_rate": 0.0002, "epoch": 0.3572872039447094, "step": 2210}, {"loss": 0.814, "grad_norm": 0.3911719024181366, "learning_rate": 0.0002, "epoch": 0.3589038881254547, "step": 2220}, {"loss": 0.8599, "grad_norm": 0.6599636077880859, "learning_rate": 0.0002, "epoch": 0.3605205723062, "step": 2230}, {"loss": 0.7482, "grad_norm": 0.40381914377212524, "learning_rate": 0.0002, "epoch": 0.3621372564869453, "step": 2240}, {"loss": 0.7772, "grad_norm": 0.4433908462524414, "learning_rate": 0.0002, "epoch": 0.36375394066769057, "step": 2250}, {"loss": 0.8503, "grad_norm": 0.578326940536499, "learning_rate": 0.0002, "epoch": 0.36537062484843585, "step": 2260}, {"loss": 0.8178, "grad_norm": 0.5734784007072449, "learning_rate": 0.0002, "epoch": 0.36698730902918114, "step": 2270}, {"loss": 0.8193, "grad_norm": 0.45555487275123596, "learning_rate": 0.0002, "epoch": 0.36860399320992643, "step": 2280}, {"loss": 0.7929, "grad_norm": 0.5666276216506958, "learning_rate": 0.0002, "epoch": 0.3702206773906717, "step": 2290}, {"loss": 0.8292, "grad_norm": 0.5461117625236511, "learning_rate": 0.0002, "epoch": 0.371837361571417, "step": 2300}, {"loss": 0.8204, "grad_norm": 0.6318911910057068, "learning_rate": 0.0002, "epoch": 0.3734540457521623, "step": 2310}, {"loss": 0.7964, "grad_norm": 0.493263304233551, "learning_rate": 0.0002, "epoch": 0.37507072993290763, "step": 2320}, {"loss": 0.8339, "grad_norm": 0.5888760089874268, "learning_rate": 0.0002, "epoch": 0.3766874141136529, "step": 2330}, {"loss": 0.7737, "grad_norm": 0.48671841621398926, "learning_rate": 0.0002, "epoch": 0.3783040982943982, "step": 2340}, {"loss": 0.8367, "grad_norm": 0.4385145306587219, "learning_rate": 0.0002, "epoch": 0.3799207824751435, "step": 2350}, {"loss": 0.812, "grad_norm": 0.5523318648338318, "learning_rate": 0.0002, "epoch": 0.3815374666558888, "step": 2360}, {"loss": 0.8351, "grad_norm": 0.7308220267295837, "learning_rate": 0.0002, "epoch": 0.38315415083663407, "step": 2370}, {"loss": 0.859, "grad_norm": 0.554214358329773, "learning_rate": 0.0002, "epoch": 0.38477083501737935, "step": 2380}, {"loss": 0.8146, "grad_norm": 0.5425800085067749, "learning_rate": 0.0002, "epoch": 0.38638751919812464, "step": 2390}, {"loss": 0.8282, "grad_norm": 0.48811158537864685, "learning_rate": 0.0002, "epoch": 0.3880042033788699, "step": 2400}, {"loss": 0.8074, "grad_norm": 0.49212366342544556, "learning_rate": 0.0002, "epoch": 0.3896208875596152, "step": 2410}, {"loss": 0.7991, "grad_norm": 0.5222218632698059, "learning_rate": 0.0002, "epoch": 0.3912375717403605, "step": 2420}, {"loss": 0.8182, "grad_norm": 0.4699819087982178, "learning_rate": 0.0002, "epoch": 0.3928542559211058, "step": 2430}, {"loss": 0.7919, "grad_norm": 0.46153587102890015, "learning_rate": 0.0002, "epoch": 0.39447094010185113, "step": 2440}, {"loss": 0.8111, "grad_norm": 0.4150611162185669, "learning_rate": 0.0002, "epoch": 0.3960876242825964, "step": 2450}, {"loss": 0.8589, "grad_norm": 0.5799614787101746, "learning_rate": 0.0002, "epoch": 0.3977043084633417, "step": 2460}, {"loss": 0.8085, "grad_norm": 0.56536865234375, "learning_rate": 0.0002, "epoch": 0.399320992644087, "step": 2470}, {"loss": 0.8022, "grad_norm": 0.5451247096061707, "learning_rate": 0.0002, "epoch": 0.4009376768248323, "step": 2480}, {"loss": 0.8217, "grad_norm": 0.5914521217346191, "learning_rate": 0.0002, "epoch": 0.40255436100557757, "step": 2490}, {"loss": 0.7859, "grad_norm": 0.4428117275238037, "learning_rate": 0.0002, "epoch": 0.40417104518632285, "step": 2500}, {"loss": 0.8054, "grad_norm": 0.48580947518348694, "learning_rate": 0.0002, "epoch": 0.40578772936706814, "step": 2510}, {"loss": 0.8405, "grad_norm": 0.436734676361084, "learning_rate": 0.0002, "epoch": 0.4074044135478134, "step": 2520}, {"loss": 0.8209, "grad_norm": 0.5752223134040833, "learning_rate": 0.0002, "epoch": 0.4090210977285587, "step": 2530}, {"loss": 0.8181, "grad_norm": 0.4271308183670044, "learning_rate": 0.0002, "epoch": 0.410637781909304, "step": 2540}, {"loss": 0.8058, "grad_norm": 0.46294718980789185, "learning_rate": 0.0002, "epoch": 0.4122544660900493, "step": 2550}, {"loss": 0.8473, "grad_norm": 0.49407583475112915, "learning_rate": 0.0002, "epoch": 0.4138711502707946, "step": 2560}, {"loss": 0.7881, "grad_norm": 0.4729035496711731, "learning_rate": 0.0002, "epoch": 0.4154878344515399, "step": 2570}, {"loss": 0.7834, "grad_norm": 0.4129747152328491, "learning_rate": 0.0002, "epoch": 0.4171045186322852, "step": 2580}, {"loss": 0.7859, "grad_norm": 0.5684236288070679, "learning_rate": 0.0002, "epoch": 0.4187212028130305, "step": 2590}, {"loss": 0.811, "grad_norm": 0.4862157106399536, "learning_rate": 0.0002, "epoch": 0.4203378869937758, "step": 2600}, {"loss": 0.7582, "grad_norm": 0.46567976474761963, "learning_rate": 0.0002, "epoch": 0.42195457117452106, "step": 2610}, {"loss": 0.7755, "grad_norm": 0.5710650682449341, "learning_rate": 0.0002, "epoch": 0.42357125535526635, "step": 2620}, {"loss": 0.8573, "grad_norm": 0.5660041570663452, "learning_rate": 0.0002, "epoch": 0.42518793953601164, "step": 2630}, {"loss": 0.7812, "grad_norm": 0.47944375872612, "learning_rate": 0.0002, "epoch": 0.4268046237167569, "step": 2640}, {"loss": 0.7459, "grad_norm": 0.537223756313324, "learning_rate": 0.0002, "epoch": 0.4284213078975022, "step": 2650}, {"loss": 0.8246, "grad_norm": 0.41669997572898865, "learning_rate": 0.0002, "epoch": 0.4300379920782475, "step": 2660}, {"loss": 0.7785, "grad_norm": 0.44727686047554016, "learning_rate": 0.0002, "epoch": 0.4316546762589928, "step": 2670}, {"loss": 0.8241, "grad_norm": 0.5600888729095459, "learning_rate": 0.0002, "epoch": 0.4332713604397381, "step": 2680}, {"loss": 0.7708, "grad_norm": 0.39820605516433716, "learning_rate": 0.0002, "epoch": 0.4348880446204834, "step": 2690}, {"loss": 0.8202, "grad_norm": 0.5637655854225159, "learning_rate": 0.0002, "epoch": 0.4365047288012287, "step": 2700}, {"loss": 0.855, "grad_norm": 0.6363666653633118, "learning_rate": 0.0002, "epoch": 0.438121412981974, "step": 2710}, {"loss": 0.8468, "grad_norm": 0.5656129121780396, "learning_rate": 0.0002, "epoch": 0.4397380971627193, "step": 2720}, {"loss": 0.7845, "grad_norm": 0.5600156188011169, "learning_rate": 0.0002, "epoch": 0.44135478134346456, "step": 2730}, {"loss": 0.8405, "grad_norm": 0.5506579875946045, "learning_rate": 0.0002, "epoch": 0.44297146552420985, "step": 2740}, {"loss": 0.7725, "grad_norm": 0.49878305196762085, "learning_rate": 0.0002, "epoch": 0.44458814970495514, "step": 2750}, {"loss": 0.8292, "grad_norm": 0.4569213092327118, "learning_rate": 0.0002, "epoch": 0.4462048338857004, "step": 2760}, {"loss": 0.8028, "grad_norm": 0.6056680083274841, "learning_rate": 0.0002, "epoch": 0.4478215180664457, "step": 2770}, {"loss": 0.8242, "grad_norm": 0.44474557042121887, "learning_rate": 0.0002, "epoch": 0.449438202247191, "step": 2780}, {"loss": 0.801, "grad_norm": 0.46055394411087036, "learning_rate": 0.0002, "epoch": 0.4510548864279363, "step": 2790}, {"loss": 0.7521, "grad_norm": 0.4904133379459381, "learning_rate": 0.0002, "epoch": 0.4526715706086816, "step": 2800}, {"loss": 0.8829, "grad_norm": 0.5647031664848328, "learning_rate": 0.0002, "epoch": 0.45428825478942686, "step": 2810}, {"loss": 0.8622, "grad_norm": 0.5759473443031311, "learning_rate": 0.0002, "epoch": 0.4559049389701722, "step": 2820}, {"loss": 0.7812, "grad_norm": 0.5161895751953125, "learning_rate": 0.0002, "epoch": 0.4575216231509175, "step": 2830}, {"loss": 0.8045, "grad_norm": 0.4248254597187042, "learning_rate": 0.0002, "epoch": 0.4591383073316628, "step": 2840}, {"loss": 0.7838, "grad_norm": 0.45395001769065857, "learning_rate": 0.0002, "epoch": 0.46075499151240806, "step": 2850}, {"loss": 0.8208, "grad_norm": 0.5358697772026062, "learning_rate": 0.0002, "epoch": 0.46237167569315335, "step": 2860}, {"loss": 0.8147, "grad_norm": 0.5379165410995483, "learning_rate": 0.0002, "epoch": 0.46398835987389864, "step": 2870}, {"loss": 0.7403, "grad_norm": 0.4601989686489105, "learning_rate": 0.0002, "epoch": 0.4656050440546439, "step": 2880}, {"loss": 0.8523, "grad_norm": 0.671115517616272, "learning_rate": 0.0002, "epoch": 0.4672217282353892, "step": 2890}, {"loss": 0.8262, "grad_norm": 0.4425133168697357, "learning_rate": 0.0002, "epoch": 0.4688384124161345, "step": 2900}, {"loss": 0.8178, "grad_norm": 0.5446155071258545, "learning_rate": 0.0002, "epoch": 0.4704550965968798, "step": 2910}, {"loss": 0.8106, "grad_norm": 0.603306233882904, "learning_rate": 0.0002, "epoch": 0.47207178077762507, "step": 2920}, {"loss": 0.8044, "grad_norm": 0.5377997159957886, "learning_rate": 0.0002, "epoch": 0.47368846495837036, "step": 2930}, {"loss": 0.8075, "grad_norm": 0.4931027591228485, "learning_rate": 0.0002, "epoch": 0.4753051491391157, "step": 2940}, {"loss": 0.8004, "grad_norm": 0.4711960256099701, "learning_rate": 0.0002, "epoch": 0.476921833319861, "step": 2950}, {"loss": 0.8121, "grad_norm": 0.5020492672920227, "learning_rate": 0.0002, "epoch": 0.4785385175006063, "step": 2960}, {"loss": 0.8221, "grad_norm": 0.5428946614265442, "learning_rate": 0.0002, "epoch": 0.48015520168135156, "step": 2970}, {"loss": 0.7849, "grad_norm": 0.5294089317321777, "learning_rate": 0.0002, "epoch": 0.48177188586209685, "step": 2980}, {"loss": 0.8553, "grad_norm": 0.648289144039154, "learning_rate": 0.0002, "epoch": 0.48338857004284214, "step": 2990}, {"loss": 0.7874, "grad_norm": 0.47916680574417114, "learning_rate": 0.0002, "epoch": 0.4850052542235874, "step": 3000}, {"loss": 0.8087, "grad_norm": 0.43849772214889526, "learning_rate": 0.0002, "epoch": 0.4866219384043327, "step": 3010}, {"loss": 0.7662, "grad_norm": 0.47007861733436584, "learning_rate": 0.0002, "epoch": 0.488238622585078, "step": 3020}, {"loss": 0.757, "grad_norm": 0.6314331293106079, "learning_rate": 0.0002, "epoch": 0.4898553067658233, "step": 3030}, {"loss": 0.7863, "grad_norm": 0.49211493134498596, "learning_rate": 0.0002, "epoch": 0.49147199094656857, "step": 3040}, {"loss": 0.8335, "grad_norm": 0.4537973403930664, "learning_rate": 0.0002, "epoch": 0.49308867512731386, "step": 3050}, {"loss": 0.8095, "grad_norm": 0.47326919436454773, "learning_rate": 0.0002, "epoch": 0.49470535930805914, "step": 3060}, {"loss": 0.8447, "grad_norm": 0.525874137878418, "learning_rate": 0.0002, "epoch": 0.4963220434888045, "step": 3070}, {"loss": 0.8339, "grad_norm": 0.6361091732978821, "learning_rate": 0.0002, "epoch": 0.4979387276695498, "step": 3080}, {"loss": 0.821, "grad_norm": 0.5850642919540405, "learning_rate": 0.0002, "epoch": 0.49955541185029506, "step": 3090}, {"loss": 0.8279, "grad_norm": 0.47299543023109436, "learning_rate": 0.0002, "epoch": 0.5011720960310403, "step": 3100}, {"loss": 0.8681, "grad_norm": 0.473099946975708, "learning_rate": 0.0002, "epoch": 0.5027887802117856, "step": 3110}, {"loss": 0.8223, "grad_norm": 0.48186397552490234, "learning_rate": 0.0002, "epoch": 0.5044054643925309, "step": 3120}, {"loss": 0.8292, "grad_norm": 0.5015401840209961, "learning_rate": 0.0002, "epoch": 0.5060221485732762, "step": 3130}, {"loss": 0.7692, "grad_norm": 0.5617750287055969, "learning_rate": 0.0002, "epoch": 0.5076388327540216, "step": 3140}, {"loss": 0.8708, "grad_norm": 0.5169327259063721, "learning_rate": 0.0002, "epoch": 0.5092555169347668, "step": 3150}, {"loss": 0.7845, "grad_norm": 0.545657753944397, "learning_rate": 0.0002, "epoch": 0.5108722011155121, "step": 3160}, {"loss": 0.799, "grad_norm": 0.512864351272583, "learning_rate": 0.0002, "epoch": 0.5124888852962574, "step": 3170}, {"loss": 0.7794, "grad_norm": 0.4113546311855316, "learning_rate": 0.0002, "epoch": 0.5141055694770027, "step": 3180}, {"loss": 0.8206, "grad_norm": 0.44532445073127747, "learning_rate": 0.0002, "epoch": 0.5157222536577479, "step": 3190}, {"loss": 0.8213, "grad_norm": 0.5623497366905212, "learning_rate": 0.0002, "epoch": 0.5173389378384933, "step": 3200}, {"loss": 0.7928, "grad_norm": 0.5084741115570068, "learning_rate": 0.0002, "epoch": 0.5189556220192385, "step": 3210}, {"loss": 0.8174, "grad_norm": 0.5305403470993042, "learning_rate": 0.0002, "epoch": 0.5205723061999838, "step": 3220}, {"loss": 0.8139, "grad_norm": 0.4708254337310791, "learning_rate": 0.0002, "epoch": 0.5221889903807291, "step": 3230}, {"loss": 0.7639, "grad_norm": 0.43827131390571594, "learning_rate": 0.0002, "epoch": 0.5238056745614744, "step": 3240}, {"loss": 0.7993, "grad_norm": 0.5630002617835999, "learning_rate": 0.0002, "epoch": 0.5254223587422197, "step": 3250}, {"loss": 0.7522, "grad_norm": 0.5010961890220642, "learning_rate": 0.0002, "epoch": 0.527039042922965, "step": 3260}, {"loss": 0.8374, "grad_norm": 0.6303122043609619, "learning_rate": 0.0002, "epoch": 0.5286557271037103, "step": 3270}, {"loss": 0.7727, "grad_norm": 0.5107331275939941, "learning_rate": 0.0002, "epoch": 0.5302724112844556, "step": 3280}, {"loss": 0.8495, "grad_norm": 0.5700443387031555, "learning_rate": 0.0002, "epoch": 0.5318890954652009, "step": 3290}, {"loss": 0.7776, "grad_norm": 0.46296367049217224, "learning_rate": 0.0002, "epoch": 0.5335057796459461, "step": 3300}, {"loss": 0.7931, "grad_norm": 0.531568706035614, "learning_rate": 0.0002, "epoch": 0.5351224638266915, "step": 3310}, {"loss": 0.843, "grad_norm": 0.4686741530895233, "learning_rate": 0.0002, "epoch": 0.5367391480074367, "step": 3320}, {"loss": 0.8104, "grad_norm": 0.5404331088066101, "learning_rate": 0.0002, "epoch": 0.5383558321881821, "step": 3330}, {"loss": 0.7686, "grad_norm": 0.6368790864944458, "learning_rate": 0.0002, "epoch": 0.5399725163689273, "step": 3340}, {"loss": 0.8514, "grad_norm": 0.42300888895988464, "learning_rate": 0.0002, "epoch": 0.5415892005496726, "step": 3350}, {"loss": 0.8236, "grad_norm": 0.5362542867660522, "learning_rate": 0.0002, "epoch": 0.5432058847304179, "step": 3360}, {"loss": 0.858, "grad_norm": 0.497128963470459, "learning_rate": 0.0002, "epoch": 0.5448225689111632, "step": 3370}, {"loss": 0.8519, "grad_norm": 0.5006386041641235, "learning_rate": 0.0002, "epoch": 0.5464392530919085, "step": 3380}, {"loss": 0.7867, "grad_norm": 0.44136837124824524, "learning_rate": 0.0002, "epoch": 0.5480559372726538, "step": 3390}, {"loss": 0.773, "grad_norm": 0.5897833108901978, "learning_rate": 0.0002, "epoch": 0.5496726214533991, "step": 3400}, {"loss": 0.8895, "grad_norm": 0.641075611114502, "learning_rate": 0.0002, "epoch": 0.5512893056341444, "step": 3410}, {"loss": 0.7827, "grad_norm": 0.7251322269439697, "learning_rate": 0.0002, "epoch": 0.5529059898148897, "step": 3420}, {"loss": 0.7626, "grad_norm": 0.47411349415779114, "learning_rate": 0.0002, "epoch": 0.5545226739956349, "step": 3430}, {"loss": 0.8196, "grad_norm": 0.4994310438632965, "learning_rate": 0.0002, "epoch": 0.5561393581763803, "step": 3440}, {"loss": 0.7812, "grad_norm": 0.5814438462257385, "learning_rate": 0.0002, "epoch": 0.5577560423571255, "step": 3450}, {"loss": 0.8805, "grad_norm": 0.6278898119926453, "learning_rate": 0.0002, "epoch": 0.5593727265378708, "step": 3460}, {"loss": 0.813, "grad_norm": 0.46208274364471436, "learning_rate": 0.0002, "epoch": 0.5609894107186161, "step": 3470}, {"loss": 0.8295, "grad_norm": 0.5718930959701538, "learning_rate": 0.0002, "epoch": 0.5626060948993614, "step": 3480}, {"loss": 0.8152, "grad_norm": 0.48178744316101074, "learning_rate": 0.0002, "epoch": 0.5642227790801067, "step": 3490}, {"loss": 0.8244, "grad_norm": 0.47336965799331665, "learning_rate": 0.0002, "epoch": 0.565839463260852, "step": 3500}, {"loss": 0.8099, "grad_norm": 0.43442684412002563, "learning_rate": 0.0002, "epoch": 0.5674561474415973, "step": 3510}, {"loss": 0.7564, "grad_norm": 0.6463358998298645, "learning_rate": 0.0002, "epoch": 0.5690728316223426, "step": 3520}, {"loss": 0.836, "grad_norm": 0.5286486744880676, "learning_rate": 0.0002, "epoch": 0.5706895158030879, "step": 3530}, {"loss": 0.8421, "grad_norm": 0.5405499935150146, "learning_rate": 0.0002, "epoch": 0.5723061999838331, "step": 3540}, {"loss": 0.7614, "grad_norm": 0.6654391884803772, "learning_rate": 0.0002, "epoch": 0.5739228841645785, "step": 3550}, {"loss": 0.7803, "grad_norm": 0.5081980228424072, "learning_rate": 0.0002, "epoch": 0.5755395683453237, "step": 3560}, {"loss": 0.7753, "grad_norm": 0.48978179693222046, "learning_rate": 0.0002, "epoch": 0.5771562525260691, "step": 3570}, {"loss": 0.8151, "grad_norm": 0.5840612053871155, "learning_rate": 0.0002, "epoch": 0.5787729367068143, "step": 3580}, {"loss": 0.8937, "grad_norm": 0.5235261917114258, "learning_rate": 0.0002, "epoch": 0.5803896208875596, "step": 3590}, {"loss": 0.7894, "grad_norm": 0.5672075748443604, "learning_rate": 0.0002, "epoch": 0.5820063050683049, "step": 3600}, {"loss": 0.8347, "grad_norm": 0.5613429546356201, "learning_rate": 0.0002, "epoch": 0.5836229892490502, "step": 3610}, {"loss": 0.8274, "grad_norm": 0.4032273590564728, "learning_rate": 0.0002, "epoch": 0.5852396734297954, "step": 3620}, {"loss": 0.8421, "grad_norm": 0.49559324979782104, "learning_rate": 0.0002, "epoch": 0.5868563576105408, "step": 3630}, {"loss": 0.8332, "grad_norm": 0.6895697712898254, "learning_rate": 0.0002, "epoch": 0.5884730417912861, "step": 3640}, {"loss": 0.7877, "grad_norm": 0.4750136435031891, "learning_rate": 0.0002, "epoch": 0.5900897259720314, "step": 3650}, {"loss": 0.8219, "grad_norm": 0.5176819562911987, "learning_rate": 0.0002, "epoch": 0.5917064101527767, "step": 3660}, {"loss": 0.8151, "grad_norm": 0.5817760229110718, "learning_rate": 0.0002, "epoch": 0.5933230943335219, "step": 3670}, {"loss": 0.7823, "grad_norm": 0.6064626574516296, "learning_rate": 0.0002, "epoch": 0.5949397785142673, "step": 3680}, {"loss": 0.8422, "grad_norm": 0.6728700995445251, "learning_rate": 0.0002, "epoch": 0.5965564626950125, "step": 3690}, {"loss": 0.7679, "grad_norm": 0.609305202960968, "learning_rate": 0.0002, "epoch": 0.5981731468757578, "step": 3700}, {"loss": 0.8048, "grad_norm": 0.4615488350391388, "learning_rate": 0.0002, "epoch": 0.5997898310565031, "step": 3710}, {"loss": 0.8214, "grad_norm": 2.0531179904937744, "learning_rate": 0.0002, "epoch": 0.6014065152372484, "step": 3720}, {"loss": 0.8158, "grad_norm": 0.5091132521629333, "learning_rate": 0.0002, "epoch": 0.6030231994179936, "step": 3730}, {"loss": 0.7833, "grad_norm": 0.5951124429702759, "learning_rate": 0.0002, "epoch": 0.604639883598739, "step": 3740}, {"loss": 0.7784, "grad_norm": 0.5870208144187927, "learning_rate": 0.0002, "epoch": 0.6062565677794842, "step": 3750}, {"loss": 0.8044, "grad_norm": 0.6254619359970093, "learning_rate": 0.0002, "epoch": 0.6078732519602296, "step": 3760}, {"loss": 0.7868, "grad_norm": 0.5577626824378967, "learning_rate": 0.0002, "epoch": 0.6094899361409749, "step": 3770}, {"loss": 0.8108, "grad_norm": 0.5004405379295349, "learning_rate": 0.0002, "epoch": 0.6111066203217201, "step": 3780}, {"loss": 0.8092, "grad_norm": 0.5527383685112, "learning_rate": 0.0002, "epoch": 0.6127233045024655, "step": 3790}, {"loss": 0.8036, "grad_norm": 0.49116113781929016, "learning_rate": 0.0002, "epoch": 0.6143399886832107, "step": 3800}, {"loss": 0.8352, "grad_norm": 0.5299299359321594, "learning_rate": 0.0002, "epoch": 0.6159566728639561, "step": 3810}, {"loss": 0.7737, "grad_norm": 0.464897483587265, "learning_rate": 0.0002, "epoch": 0.6175733570447013, "step": 3820}, {"loss": 0.7923, "grad_norm": 0.6505740880966187, "learning_rate": 0.0002, "epoch": 0.6191900412254466, "step": 3830}, {"loss": 0.8123, "grad_norm": 0.5512559413909912, "learning_rate": 0.0002, "epoch": 0.6208067254061919, "step": 3840}, {"loss": 0.8856, "grad_norm": 0.49427518248558044, "learning_rate": 0.0002, "epoch": 0.6224234095869372, "step": 3850}, {"loss": 0.7751, "grad_norm": 0.3839147090911865, "learning_rate": 0.0002, "epoch": 0.6240400937676824, "step": 3860}, {"loss": 0.8006, "grad_norm": 0.5760218501091003, "learning_rate": 0.0002, "epoch": 0.6256567779484278, "step": 3870}, {"loss": 0.7836, "grad_norm": 0.7226507067680359, "learning_rate": 0.0002, "epoch": 0.6272734621291731, "step": 3880}, {"loss": 0.8244, "grad_norm": 0.676781415939331, "learning_rate": 0.0002, "epoch": 0.6288901463099184, "step": 3890}, {"loss": 0.8239, "grad_norm": 0.4284018278121948, "learning_rate": 0.0002, "epoch": 0.6305068304906637, "step": 3900}, {"loss": 0.7996, "grad_norm": 0.5060628056526184, "learning_rate": 0.0002, "epoch": 0.6321235146714089, "step": 3910}, {"loss": 0.8089, "grad_norm": 0.5524522066116333, "learning_rate": 0.0002, "epoch": 0.6337401988521543, "step": 3920}, {"loss": 0.8276, "grad_norm": 0.6099881529808044, "learning_rate": 0.0002, "epoch": 0.6353568830328995, "step": 3930}, {"loss": 0.809, "grad_norm": 0.43155938386917114, "learning_rate": 0.0002, "epoch": 0.6369735672136448, "step": 3940}, {"loss": 0.8404, "grad_norm": 0.6427084803581238, "learning_rate": 0.0002, "epoch": 0.6385902513943901, "step": 3950}, {"loss": 0.8368, "grad_norm": 0.541220486164093, "learning_rate": 0.0002, "epoch": 0.6402069355751354, "step": 3960}, {"loss": 0.8539, "grad_norm": 0.5414294600486755, "learning_rate": 0.0002, "epoch": 0.6418236197558806, "step": 3970}, {"loss": 0.7996, "grad_norm": 0.46344003081321716, "learning_rate": 0.0002, "epoch": 0.643440303936626, "step": 3980}, {"loss": 0.7474, "grad_norm": 0.45209285616874695, "learning_rate": 0.0002, "epoch": 0.6450569881173712, "step": 3990}, {"loss": 0.8202, "grad_norm": 0.5417284369468689, "learning_rate": 0.0002, "epoch": 0.6466736722981166, "step": 4000}, {"loss": 0.7563, "grad_norm": 0.7995685935020447, "learning_rate": 0.0002, "epoch": 0.6482903564788619, "step": 4010}, {"loss": 0.7812, "grad_norm": 0.6384002566337585, "learning_rate": 0.0002, "epoch": 0.6499070406596071, "step": 4020}, {"loss": 0.732, "grad_norm": 0.4472815692424774, "learning_rate": 0.0002, "epoch": 0.6515237248403525, "step": 4030}, {"loss": 0.8071, "grad_norm": 0.6834294199943542, "learning_rate": 0.0002, "epoch": 0.6531404090210977, "step": 4040}, {"loss": 0.7812, "grad_norm": 0.4612339735031128, "learning_rate": 0.0002, "epoch": 0.654757093201843, "step": 4050}, {"loss": 0.8141, "grad_norm": 0.9266576170921326, "learning_rate": 0.0002, "epoch": 0.6563737773825883, "step": 4060}, {"loss": 0.7991, "grad_norm": 0.4470861852169037, "learning_rate": 0.0002, "epoch": 0.6579904615633336, "step": 4070}, {"loss": 0.8293, "grad_norm": 0.45544925332069397, "learning_rate": 0.0002, "epoch": 0.6596071457440789, "step": 4080}, {"loss": 0.8455, "grad_norm": 0.6144481301307678, "learning_rate": 0.0002, "epoch": 0.6612238299248242, "step": 4090}, {"loss": 0.7877, "grad_norm": 0.5936288237571716, "learning_rate": 0.0002, "epoch": 0.6628405141055694, "step": 4100}, {"loss": 0.7617, "grad_norm": 0.4822963774204254, "learning_rate": 0.0002, "epoch": 0.6644571982863148, "step": 4110}, {"loss": 0.7997, "grad_norm": 0.48432496190071106, "learning_rate": 0.0002, "epoch": 0.66607388246706, "step": 4120}, {"loss": 0.8404, "grad_norm": 0.4901607930660248, "learning_rate": 0.0002, "epoch": 0.6676905666478054, "step": 4130}, {"loss": 0.8085, "grad_norm": 0.5018393397331238, "learning_rate": 0.0002, "epoch": 0.6693072508285507, "step": 4140}, {"loss": 0.8065, "grad_norm": 0.6946378946304321, "learning_rate": 0.0002, "epoch": 0.6709239350092959, "step": 4150}, {"loss": 0.8147, "grad_norm": 0.5997390747070312, "learning_rate": 0.0002, "epoch": 0.6725406191900413, "step": 4160}, {"loss": 0.8268, "grad_norm": 0.6738849878311157, "learning_rate": 0.0002, "epoch": 0.6741573033707865, "step": 4170}, {"loss": 0.7704, "grad_norm": 0.6110581159591675, "learning_rate": 0.0002, "epoch": 0.6757739875515318, "step": 4180}, {"loss": 0.8043, "grad_norm": 0.5703322291374207, "learning_rate": 0.0002, "epoch": 0.6773906717322771, "step": 4190}, {"loss": 0.8099, "grad_norm": 0.4686066210269928, "learning_rate": 0.0002, "epoch": 0.6790073559130224, "step": 4200}, {"loss": 0.8441, "grad_norm": 0.6394643783569336, "learning_rate": 0.0002, "epoch": 0.6806240400937676, "step": 4210}, {"loss": 0.8011, "grad_norm": 0.5454841256141663, "learning_rate": 0.0002, "epoch": 0.682240724274513, "step": 4220}, {"loss": 0.8307, "grad_norm": 0.4859732985496521, "learning_rate": 0.0002, "epoch": 0.6838574084552582, "step": 4230}, {"loss": 0.8161, "grad_norm": 0.5544065833091736, "learning_rate": 0.0002, "epoch": 0.6854740926360036, "step": 4240}, {"loss": 0.7839, "grad_norm": 0.4902505576610565, "learning_rate": 0.0002, "epoch": 0.6870907768167488, "step": 4250}, {"loss": 0.7977, "grad_norm": 0.4768051505088806, "learning_rate": 0.0002, "epoch": 0.6887074609974941, "step": 4260}, {"loss": 0.7539, "grad_norm": 0.49982190132141113, "learning_rate": 0.0002, "epoch": 0.6903241451782395, "step": 4270}, {"loss": 0.7353, "grad_norm": 0.6351838111877441, "learning_rate": 0.0002, "epoch": 0.6919408293589847, "step": 4280}, {"loss": 0.7664, "grad_norm": 0.5647561550140381, "learning_rate": 0.0002, "epoch": 0.69355751353973, "step": 4290}, {"loss": 0.7618, "grad_norm": 0.5340486764907837, "learning_rate": 0.0002, "epoch": 0.6951741977204753, "step": 4300}, {"loss": 0.8526, "grad_norm": 0.5649092793464661, "learning_rate": 0.0002, "epoch": 0.6967908819012206, "step": 4310}, {"loss": 0.8246, "grad_norm": 0.6183916926383972, "learning_rate": 0.0002, "epoch": 0.6984075660819659, "step": 4320}, {"loss": 0.792, "grad_norm": 0.6154509782791138, "learning_rate": 0.0002, "epoch": 0.7000242502627112, "step": 4330}, {"loss": 0.8397, "grad_norm": 0.5156264305114746, "learning_rate": 0.0002, "epoch": 0.7016409344434564, "step": 4340}, {"loss": 0.8512, "grad_norm": 0.562171459197998, "learning_rate": 0.0002, "epoch": 0.7032576186242018, "step": 4350}, {"loss": 0.7882, "grad_norm": 0.4949502646923065, "learning_rate": 0.0002, "epoch": 0.704874302804947, "step": 4360}, {"loss": 0.738, "grad_norm": 0.5171684622764587, "learning_rate": 0.0002, "epoch": 0.7064909869856923, "step": 4370}, {"loss": 0.8001, "grad_norm": 0.6198443174362183, "learning_rate": 0.0002, "epoch": 0.7081076711664377, "step": 4380}, {"loss": 0.7606, "grad_norm": 0.5802276134490967, "learning_rate": 0.0002, "epoch": 0.7097243553471829, "step": 4390}, {"loss": 0.8797, "grad_norm": 0.41096967458724976, "learning_rate": 0.0002, "epoch": 0.7113410395279283, "step": 4400}, {"loss": 0.805, "grad_norm": 0.4397392272949219, "learning_rate": 0.0002, "epoch": 0.7129577237086735, "step": 4410}, {"loss": 0.7651, "grad_norm": 0.45228442549705505, "learning_rate": 0.0002, "epoch": 0.7145744078894188, "step": 4420}, {"loss": 0.7938, "grad_norm": 0.4839673936367035, "learning_rate": 0.0002, "epoch": 0.7161910920701641, "step": 4430}, {"loss": 0.8362, "grad_norm": 0.6140755414962769, "learning_rate": 0.0002, "epoch": 0.7178077762509094, "step": 4440}, {"loss": 0.7722, "grad_norm": 0.6841378808021545, "learning_rate": 0.0002, "epoch": 0.7194244604316546, "step": 4450}, {"loss": 0.8177, "grad_norm": 0.6664239168167114, "learning_rate": 0.0002, "epoch": 0.7210411446124, "step": 4460}, {"loss": 0.7983, "grad_norm": 0.47552719712257385, "learning_rate": 0.0002, "epoch": 0.7226578287931452, "step": 4470}, {"loss": 0.8982, "grad_norm": 0.6649776101112366, "learning_rate": 0.0002, "epoch": 0.7242745129738906, "step": 4480}, {"loss": 0.8074, "grad_norm": 0.5159541964530945, "learning_rate": 0.0002, "epoch": 0.7258911971546358, "step": 4490}, {"loss": 0.7786, "grad_norm": 0.6693112850189209, "learning_rate": 0.0002, "epoch": 0.7275078813353811, "step": 4500}, {"loss": 0.8655, "grad_norm": 0.48870977759361267, "learning_rate": 0.0002, "epoch": 0.7291245655161265, "step": 4510}, {"loss": 0.7337, "grad_norm": 0.4857887923717499, "learning_rate": 0.0002, "epoch": 0.7307412496968717, "step": 4520}, {"loss": 0.8026, "grad_norm": 0.5515662431716919, "learning_rate": 0.0002, "epoch": 0.732357933877617, "step": 4530}, {"loss": 0.8031, "grad_norm": 0.6292222738265991, "learning_rate": 0.0002, "epoch": 0.7339746180583623, "step": 4540}, {"loss": 0.7749, "grad_norm": 0.48265689611434937, "learning_rate": 0.0002, "epoch": 0.7355913022391076, "step": 4550}, {"loss": 0.8499, "grad_norm": 0.8044266104698181, "learning_rate": 0.0002, "epoch": 0.7372079864198529, "step": 4560}, {"loss": 0.8162, "grad_norm": 0.6111769676208496, "learning_rate": 0.0002, "epoch": 0.7388246706005982, "step": 4570}, {"loss": 0.7291, "grad_norm": 0.5229553580284119, "learning_rate": 0.0002, "epoch": 0.7404413547813434, "step": 4580}, {"loss": 0.8038, "grad_norm": 0.6054152250289917, "learning_rate": 0.0002, "epoch": 0.7420580389620888, "step": 4590}, {"loss": 0.8169, "grad_norm": 0.5574966669082642, "learning_rate": 0.0002, "epoch": 0.743674723142834, "step": 4600}, {"loss": 0.8439, "grad_norm": 0.5395817160606384, "learning_rate": 0.0002, "epoch": 0.7452914073235793, "step": 4610}, {"loss": 0.8495, "grad_norm": 0.7116472721099854, "learning_rate": 0.0002, "epoch": 0.7469080915043246, "step": 4620}, {"loss": 0.7743, "grad_norm": 0.5618700981140137, "learning_rate": 0.0002, "epoch": 0.7485247756850699, "step": 4630}, {"loss": 0.7744, "grad_norm": 0.5802770853042603, "learning_rate": 0.0002, "epoch": 0.7501414598658153, "step": 4640}, {"loss": 0.7924, "grad_norm": 0.5690428018569946, "learning_rate": 0.0002, "epoch": 0.7517581440465605, "step": 4650}, {"loss": 0.8017, "grad_norm": 0.4813360273838043, "learning_rate": 0.0002, "epoch": 0.7533748282273058, "step": 4660}, {"loss": 0.8108, "grad_norm": 0.5434042811393738, "learning_rate": 0.0002, "epoch": 0.7549915124080511, "step": 4670}, {"loss": 0.7824, "grad_norm": 0.5502099990844727, "learning_rate": 0.0002, "epoch": 0.7566081965887964, "step": 4680}, {"loss": 0.8598, "grad_norm": 0.6020621061325073, "learning_rate": 0.0002, "epoch": 0.7582248807695416, "step": 4690}, {"loss": 0.7937, "grad_norm": 0.4922301471233368, "learning_rate": 0.0002, "epoch": 0.759841564950287, "step": 4700}, {"loss": 0.788, "grad_norm": 0.6492828726768494, "learning_rate": 0.0002, "epoch": 0.7614582491310322, "step": 4710}, {"loss": 0.8313, "grad_norm": 0.4865580201148987, "learning_rate": 0.0002, "epoch": 0.7630749333117776, "step": 4720}, {"loss": 0.7966, "grad_norm": 0.5971422791481018, "learning_rate": 0.0002, "epoch": 0.7646916174925228, "step": 4730}, {"loss": 0.8298, "grad_norm": 0.6832674145698547, "learning_rate": 0.0002, "epoch": 0.7663083016732681, "step": 4740}, {"loss": 0.8156, "grad_norm": 0.500908613204956, "learning_rate": 0.0002, "epoch": 0.7679249858540134, "step": 4750}, {"loss": 0.8383, "grad_norm": 0.6112465858459473, "learning_rate": 0.0002, "epoch": 0.7695416700347587, "step": 4760}, {"loss": 0.76, "grad_norm": 0.5753506422042847, "learning_rate": 0.0002, "epoch": 0.771158354215504, "step": 4770}, {"loss": 0.8297, "grad_norm": 0.6529405117034912, "learning_rate": 0.0002, "epoch": 0.7727750383962493, "step": 4780}, {"loss": 0.8171, "grad_norm": 0.5916843414306641, "learning_rate": 0.0002, "epoch": 0.7743917225769946, "step": 4790}, {"loss": 0.83, "grad_norm": 0.4821224510669708, "learning_rate": 0.0002, "epoch": 0.7760084067577399, "step": 4800}, {"loss": 0.7703, "grad_norm": 0.5532580018043518, "learning_rate": 0.0002, "epoch": 0.7776250909384852, "step": 4810}, {"loss": 0.7363, "grad_norm": 0.4604877233505249, "learning_rate": 0.0002, "epoch": 0.7792417751192304, "step": 4820}, {"loss": 0.7506, "grad_norm": 0.5009613037109375, "learning_rate": 0.0002, "epoch": 0.7808584592999758, "step": 4830}, {"loss": 0.7863, "grad_norm": 0.6448560357093811, "learning_rate": 0.0002, "epoch": 0.782475143480721, "step": 4840}, {"loss": 0.7957, "grad_norm": 0.44327953457832336, "learning_rate": 0.0002, "epoch": 0.7840918276614663, "step": 4850}, {"loss": 0.7925, "grad_norm": 0.5355411171913147, "learning_rate": 0.0002, "epoch": 0.7857085118422116, "step": 4860}, {"loss": 0.7754, "grad_norm": 0.5635677576065063, "learning_rate": 0.0002, "epoch": 0.7873251960229569, "step": 4870}, {"loss": 0.7931, "grad_norm": 0.5417491793632507, "learning_rate": 0.0002, "epoch": 0.7889418802037023, "step": 4880}, {"loss": 0.7819, "grad_norm": 0.4567430913448334, "learning_rate": 0.0002, "epoch": 0.7905585643844475, "step": 4890}, {"loss": 0.8454, "grad_norm": 0.44651296734809875, "learning_rate": 0.0002, "epoch": 0.7921752485651928, "step": 4900}, {"loss": 0.7959, "grad_norm": 0.5741217136383057, "learning_rate": 0.0002, "epoch": 0.7937919327459381, "step": 4910}, {"loss": 0.8093, "grad_norm": 0.6605045199394226, "learning_rate": 0.0002, "epoch": 0.7954086169266834, "step": 4920}, {"loss": 0.77, "grad_norm": 0.5126531720161438, "learning_rate": 0.0002, "epoch": 0.7970253011074286, "step": 4930}, {"loss": 0.7793, "grad_norm": 0.513648271560669, "learning_rate": 0.0002, "epoch": 0.798641985288174, "step": 4940}, {"loss": 0.8314, "grad_norm": 0.5350404381752014, "learning_rate": 0.0002, "epoch": 0.8002586694689192, "step": 4950}, {"loss": 0.7649, "grad_norm": 0.5731674432754517, "learning_rate": 0.0002, "epoch": 0.8018753536496646, "step": 4960}, {"loss": 0.8572, "grad_norm": 0.5974258184432983, "learning_rate": 0.0002, "epoch": 0.8034920378304098, "step": 4970}, {"loss": 0.7972, "grad_norm": 0.8774799704551697, "learning_rate": 0.0002, "epoch": 0.8051087220111551, "step": 4980}, {"loss": 0.7899, "grad_norm": 0.5994430184364319, "learning_rate": 0.0002, "epoch": 0.8067254061919004, "step": 4990}, {"loss": 0.7736, "grad_norm": 0.4894903004169464, "learning_rate": 0.0002, "epoch": 0.8083420903726457, "step": 5000}, {"loss": 0.78, "grad_norm": 0.5218459367752075, "learning_rate": 0.0002, "epoch": 0.809958774553391, "step": 5010}, {"loss": 0.817, "grad_norm": 0.5232468843460083, "learning_rate": 0.0002, "epoch": 0.8115754587341363, "step": 5020}, {"loss": 0.7704, "grad_norm": 0.44358372688293457, "learning_rate": 0.0002, "epoch": 0.8131921429148816, "step": 5030}, {"loss": 0.785, "grad_norm": 0.6202037334442139, "learning_rate": 0.0002, "epoch": 0.8148088270956269, "step": 5040}, {"loss": 0.7351, "grad_norm": 0.7721474170684814, "learning_rate": 0.0002, "epoch": 0.8164255112763722, "step": 5050}, {"loss": 0.8297, "grad_norm": 0.5568501353263855, "learning_rate": 0.0002, "epoch": 0.8180421954571174, "step": 5060}, {"loss": 0.7733, "grad_norm": 0.49148809909820557, "learning_rate": 0.0002, "epoch": 0.8196588796378628, "step": 5070}, {"loss": 0.8054, "grad_norm": 0.4956012964248657, "learning_rate": 0.0002, "epoch": 0.821275563818608, "step": 5080}, {"loss": 0.8201, "grad_norm": 0.6078833937644958, "learning_rate": 0.0002, "epoch": 0.8228922479993533, "step": 5090}, {"loss": 0.828, "grad_norm": 0.46906954050064087, "learning_rate": 0.0002, "epoch": 0.8245089321800986, "step": 5100}, {"loss": 0.7703, "grad_norm": 0.50812166929245, "learning_rate": 0.0002, "epoch": 0.8261256163608439, "step": 5110}, {"loss": 0.8243, "grad_norm": 0.5319661498069763, "learning_rate": 0.0002, "epoch": 0.8277423005415891, "step": 5120}, {"loss": 0.7798, "grad_norm": 0.4949689209461212, "learning_rate": 0.0002, "epoch": 0.8293589847223345, "step": 5130}, {"loss": 0.7428, "grad_norm": 0.5151591300964355, "learning_rate": 0.0002, "epoch": 0.8309756689030798, "step": 5140}, {"loss": 0.8147, "grad_norm": 0.5530214309692383, "learning_rate": 0.0002, "epoch": 0.8325923530838251, "step": 5150}, {"loss": 0.8251, "grad_norm": 0.6297410130500793, "learning_rate": 0.0002, "epoch": 0.8342090372645704, "step": 5160}, {"loss": 0.8067, "grad_norm": 0.5466840267181396, "learning_rate": 0.0002, "epoch": 0.8358257214453156, "step": 5170}, {"loss": 0.7875, "grad_norm": 0.652913510799408, "learning_rate": 0.0002, "epoch": 0.837442405626061, "step": 5180}, {"loss": 0.8295, "grad_norm": 0.5811293125152588, "learning_rate": 0.0002, "epoch": 0.8390590898068062, "step": 5190}, {"loss": 0.7412, "grad_norm": 0.5109550952911377, "learning_rate": 0.0002, "epoch": 0.8406757739875516, "step": 5200}, {"loss": 0.8077, "grad_norm": 0.4551706612110138, "learning_rate": 0.0002, "epoch": 0.8422924581682968, "step": 5210}, {"loss": 0.7827, "grad_norm": 0.5813754200935364, "learning_rate": 0.0002, "epoch": 0.8439091423490421, "step": 5220}, {"loss": 0.802, "grad_norm": 0.5856947898864746, "learning_rate": 0.0002, "epoch": 0.8455258265297874, "step": 5230}, {"loss": 0.7957, "grad_norm": 0.5482739210128784, "learning_rate": 0.0002, "epoch": 0.8471425107105327, "step": 5240}, {"loss": 0.8295, "grad_norm": 0.49023720622062683, "learning_rate": 0.0002, "epoch": 0.8487591948912779, "step": 5250}, {"loss": 0.8022, "grad_norm": 0.49472475051879883, "learning_rate": 0.0002, "epoch": 0.8503758790720233, "step": 5260}, {"loss": 0.8001, "grad_norm": 0.5490226745605469, "learning_rate": 0.0002, "epoch": 0.8519925632527686, "step": 5270}, {"loss": 0.8333, "grad_norm": 0.5340665578842163, "learning_rate": 0.0002, "epoch": 0.8536092474335139, "step": 5280}, {"loss": 0.8277, "grad_norm": 0.5962483882904053, "learning_rate": 0.0002, "epoch": 0.8552259316142592, "step": 5290}, {"loss": 0.8765, "grad_norm": 0.586358368396759, "learning_rate": 0.0002, "epoch": 0.8568426157950044, "step": 5300}, {"loss": 0.7831, "grad_norm": 0.49120277166366577, "learning_rate": 0.0002, "epoch": 0.8584592999757498, "step": 5310}, {"loss": 0.8162, "grad_norm": 0.5887332558631897, "learning_rate": 0.0002, "epoch": 0.860075984156495, "step": 5320}, {"loss": 0.7464, "grad_norm": 0.42496153712272644, "learning_rate": 0.0002, "epoch": 0.8616926683372403, "step": 5330}, {"loss": 0.7905, "grad_norm": 0.5489874482154846, "learning_rate": 0.0002, "epoch": 0.8633093525179856, "step": 5340}, {"loss": 0.7958, "grad_norm": 0.5850813984870911, "learning_rate": 0.0002, "epoch": 0.8649260366987309, "step": 5350}, {"loss": 0.7642, "grad_norm": 0.517487108707428, "learning_rate": 0.0002, "epoch": 0.8665427208794761, "step": 5360}, {"loss": 0.7801, "grad_norm": 0.5339142680168152, "learning_rate": 0.0002, "epoch": 0.8681594050602215, "step": 5370}, {"loss": 0.818, "grad_norm": 0.6236387491226196, "learning_rate": 0.0002, "epoch": 0.8697760892409668, "step": 5380}, {"loss": 0.7708, "grad_norm": 0.5752192735671997, "learning_rate": 0.0002, "epoch": 0.8713927734217121, "step": 5390}, {"loss": 0.8542, "grad_norm": 0.6724614500999451, "learning_rate": 0.0002, "epoch": 0.8730094576024574, "step": 5400}, {"loss": 0.7581, "grad_norm": 0.5280613303184509, "learning_rate": 0.0002, "epoch": 0.8746261417832026, "step": 5410}, {"loss": 0.8231, "grad_norm": 0.44033288955688477, "learning_rate": 0.0002, "epoch": 0.876242825963948, "step": 5420}, {"loss": 0.8839, "grad_norm": 0.5199708342552185, "learning_rate": 0.0002, "epoch": 0.8778595101446932, "step": 5430}, {"loss": 0.7852, "grad_norm": 0.46778348088264465, "learning_rate": 0.0002, "epoch": 0.8794761943254386, "step": 5440}, {"loss": 0.7834, "grad_norm": 0.4657754898071289, "learning_rate": 0.0002, "epoch": 0.8810928785061838, "step": 5450}, {"loss": 0.7799, "grad_norm": 0.5472902655601501, "learning_rate": 0.0002, "epoch": 0.8827095626869291, "step": 5460}, {"loss": 0.8253, "grad_norm": 0.4876766800880432, "learning_rate": 0.0002, "epoch": 0.8843262468676744, "step": 5470}, {"loss": 0.7906, "grad_norm": 0.5057248473167419, "learning_rate": 0.0002, "epoch": 0.8859429310484197, "step": 5480}, {"loss": 0.8124, "grad_norm": 0.4637320637702942, "learning_rate": 0.0002, "epoch": 0.8875596152291649, "step": 5490}, {"loss": 0.781, "grad_norm": 0.471955806016922, "learning_rate": 0.0002, "epoch": 0.8891762994099103, "step": 5500}, {"loss": 0.8057, "grad_norm": 0.5209813714027405, "learning_rate": 0.0002, "epoch": 0.8907929835906556, "step": 5510}, {"loss": 0.8106, "grad_norm": 0.6213834285736084, "learning_rate": 0.0002, "epoch": 0.8924096677714008, "step": 5520}, {"loss": 0.7787, "grad_norm": 0.5215408205986023, "learning_rate": 0.0002, "epoch": 0.8940263519521462, "step": 5530}, {"loss": 0.8174, "grad_norm": 0.580478310585022, "learning_rate": 0.0002, "epoch": 0.8956430361328914, "step": 5540}, {"loss": 0.8371, "grad_norm": 0.49102169275283813, "learning_rate": 0.0002, "epoch": 0.8972597203136368, "step": 5550}, {"loss": 0.7806, "grad_norm": 0.6043479442596436, "learning_rate": 0.0002, "epoch": 0.898876404494382, "step": 5560}, {"loss": 0.7754, "grad_norm": 0.5636463165283203, "learning_rate": 0.0002, "epoch": 0.9004930886751273, "step": 5570}, {"loss": 0.8145, "grad_norm": 0.5620124340057373, "learning_rate": 0.0002, "epoch": 0.9021097728558726, "step": 5580}, {"loss": 0.8083, "grad_norm": 0.5206354856491089, "learning_rate": 0.0002, "epoch": 0.9037264570366179, "step": 5590}, {"loss": 0.8557, "grad_norm": 0.5798229575157166, "learning_rate": 0.0002, "epoch": 0.9053431412173631, "step": 5600}, {"loss": 0.8097, "grad_norm": 0.6428212523460388, "learning_rate": 0.0002, "epoch": 0.9069598253981085, "step": 5610}, {"loss": 0.7839, "grad_norm": 0.48064687848091125, "learning_rate": 0.0002, "epoch": 0.9085765095788537, "step": 5620}, {"loss": 0.8343, "grad_norm": 0.6347860097885132, "learning_rate": 0.0002, "epoch": 0.9101931937595991, "step": 5630}, {"loss": 0.851, "grad_norm": 0.5353913307189941, "learning_rate": 0.0002, "epoch": 0.9118098779403444, "step": 5640}, {"loss": 0.7736, "grad_norm": 0.5323944091796875, "learning_rate": 0.0002, "epoch": 0.9134265621210896, "step": 5650}, {"loss": 0.8393, "grad_norm": 0.5261843204498291, "learning_rate": 0.0002, "epoch": 0.915043246301835, "step": 5660}, {"loss": 0.7355, "grad_norm": 0.5451326966285706, "learning_rate": 0.0002, "epoch": 0.9166599304825802, "step": 5670}, {"loss": 0.8012, "grad_norm": 0.5183324217796326, "learning_rate": 0.0002, "epoch": 0.9182766146633256, "step": 5680}, {"loss": 0.7659, "grad_norm": 0.47229018807411194, "learning_rate": 0.0002, "epoch": 0.9198932988440708, "step": 5690}, {"loss": 0.7757, "grad_norm": 0.49180513620376587, "learning_rate": 0.0002, "epoch": 0.9215099830248161, "step": 5700}, {"loss": 0.8735, "grad_norm": 0.5419785380363464, "learning_rate": 0.0002, "epoch": 0.9231266672055614, "step": 5710}, {"loss": 0.7378, "grad_norm": 0.5408698916435242, "learning_rate": 0.0002, "epoch": 0.9247433513863067, "step": 5720}, {"loss": 0.7701, "grad_norm": 0.5286232829093933, "learning_rate": 0.0002, "epoch": 0.9263600355670519, "step": 5730}, {"loss": 0.8242, "grad_norm": 0.7539758086204529, "learning_rate": 0.0002, "epoch": 0.9279767197477973, "step": 5740}, {"loss": 0.8118, "grad_norm": 0.5166944861412048, "learning_rate": 0.0002, "epoch": 0.9295934039285425, "step": 5750}, {"loss": 0.783, "grad_norm": 0.6601425409317017, "learning_rate": 0.0002, "epoch": 0.9312100881092878, "step": 5760}, {"loss": 0.7873, "grad_norm": 0.5029960870742798, "learning_rate": 0.0002, "epoch": 0.9328267722900332, "step": 5770}, {"loss": 0.7989, "grad_norm": 0.4926645755767822, "learning_rate": 0.0002, "epoch": 0.9344434564707784, "step": 5780}, {"loss": 0.8174, "grad_norm": 0.5739615559577942, "learning_rate": 0.0002, "epoch": 0.9360601406515238, "step": 5790}, {"loss": 0.8037, "grad_norm": 0.5058279037475586, "learning_rate": 0.0002, "epoch": 0.937676824832269, "step": 5800}, {"loss": 0.8537, "grad_norm": 0.5260962247848511, "learning_rate": 0.0002, "epoch": 0.9392935090130143, "step": 5810}, {"loss": 0.7486, "grad_norm": 0.5768588185310364, "learning_rate": 0.0002, "epoch": 0.9409101931937596, "step": 5820}, {"loss": 0.8215, "grad_norm": 0.5170126557350159, "learning_rate": 0.0002, "epoch": 0.9425268773745049, "step": 5830}, {"loss": 0.7422, "grad_norm": 0.5745864510536194, "learning_rate": 0.0002, "epoch": 0.9441435615552501, "step": 5840}, {"loss": 0.7824, "grad_norm": 0.5551357865333557, "learning_rate": 0.0002, "epoch": 0.9457602457359955, "step": 5850}, {"loss": 0.8529, "grad_norm": 0.5776078701019287, "learning_rate": 0.0002, "epoch": 0.9473769299167407, "step": 5860}, {"loss": 0.8527, "grad_norm": 0.5340062379837036, "learning_rate": 0.0002, "epoch": 0.9489936140974861, "step": 5870}, {"loss": 0.8217, "grad_norm": 0.6447290182113647, "learning_rate": 0.0002, "epoch": 0.9506102982782314, "step": 5880}, {"loss": 0.7945, "grad_norm": 0.5123815536499023, "learning_rate": 0.0002, "epoch": 0.9522269824589766, "step": 5890}, {"loss": 0.8209, "grad_norm": 0.48547613620758057, "learning_rate": 0.0002, "epoch": 0.953843666639722, "step": 5900}, {"loss": 0.7896, "grad_norm": 0.5791414976119995, "learning_rate": 0.0002, "epoch": 0.9554603508204672, "step": 5910}, {"loss": 0.8408, "grad_norm": 0.6195011734962463, "learning_rate": 0.0002, "epoch": 0.9570770350012126, "step": 5920}, {"loss": 0.7805, "grad_norm": 0.6323803067207336, "learning_rate": 0.0002, "epoch": 0.9586937191819578, "step": 5930}, {"loss": 0.8484, "grad_norm": 0.45552879571914673, "learning_rate": 0.0002, "epoch": 0.9603104033627031, "step": 5940}, {"loss": 0.7367, "grad_norm": 0.5796473622322083, "learning_rate": 0.0002, "epoch": 0.9619270875434484, "step": 5950}, {"loss": 0.7672, "grad_norm": 0.647261381149292, "learning_rate": 0.0002, "epoch": 0.9635437717241937, "step": 5960}, {"loss": 0.8086, "grad_norm": 0.5487682819366455, "learning_rate": 0.0002, "epoch": 0.9651604559049389, "step": 5970}, {"loss": 0.7973, "grad_norm": 0.5743663907051086, "learning_rate": 0.0002, "epoch": 0.9667771400856843, "step": 5980}, {"loss": 0.8153, "grad_norm": 0.5470591187477112, "learning_rate": 0.0002, "epoch": 0.9683938242664295, "step": 5990}, {"loss": 0.8119, "grad_norm": 0.5901660323143005, "learning_rate": 0.0002, "epoch": 0.9700105084471748, "step": 6000}, {"loss": 0.8147, "grad_norm": 0.6544759273529053, "learning_rate": 0.0002, "epoch": 0.9716271926279202, "step": 6010}, {"loss": 0.7536, "grad_norm": 0.6288470029830933, "learning_rate": 0.0002, "epoch": 0.9732438768086654, "step": 6020}, {"loss": 0.7989, "grad_norm": 0.673153817653656, "learning_rate": 0.0002, "epoch": 0.9748605609894108, "step": 6030}, {"loss": 0.7556, "grad_norm": 0.42854753136634827, "learning_rate": 0.0002, "epoch": 0.976477245170156, "step": 6040}, {"loss": 0.8006, "grad_norm": 0.5227066278457642, "learning_rate": 0.0002, "epoch": 0.9780939293509013, "step": 6050}, {"loss": 0.795, "grad_norm": 0.5372416973114014, "learning_rate": 0.0002, "epoch": 0.9797106135316466, "step": 6060}, {"loss": 0.7591, "grad_norm": 0.6026402115821838, "learning_rate": 0.0002, "epoch": 0.9813272977123919, "step": 6070}, {"loss": 0.8347, "grad_norm": 0.49547791481018066, "learning_rate": 0.0002, "epoch": 0.9829439818931371, "step": 6080}, {"loss": 0.7722, "grad_norm": 0.4641951322555542, "learning_rate": 0.0002, "epoch": 0.9845606660738825, "step": 6090}, {"loss": 0.8125, "grad_norm": 0.5818535089492798, "learning_rate": 0.0002, "epoch": 0.9861773502546277, "step": 6100}, {"loss": 0.81, "grad_norm": 0.63955157995224, "learning_rate": 0.0002, "epoch": 0.9877940344353731, "step": 6110}, {"loss": 0.7547, "grad_norm": 0.5649438500404358, "learning_rate": 0.0002, "epoch": 0.9894107186161183, "step": 6120}, {"loss": 0.7861, "grad_norm": 0.5290433168411255, "learning_rate": 0.0002, "epoch": 0.9910274027968636, "step": 6130}, {"loss": 0.8109, "grad_norm": 0.6399374008178711, "learning_rate": 0.0002, "epoch": 0.992644086977609, "step": 6140}, {"loss": 0.8373, "grad_norm": 0.6736576557159424, "learning_rate": 0.0002, "epoch": 0.9942607711583542, "step": 6150}, {"loss": 0.7915, "grad_norm": 0.515420138835907, "learning_rate": 0.0002, "epoch": 0.9958774553390995, "step": 6160}, {"loss": 0.8032, "grad_norm": 0.562677800655365, "learning_rate": 0.0002, "epoch": 0.9974941395198448, "step": 6170}, {"loss": 0.8187, "grad_norm": 0.7113858461380005, "learning_rate": 0.0002, "epoch": 0.9991108237005901, "step": 6180}, {"eval_loss": 1.0871200561523438, "eval_runtime": 122.2071, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 0.9999191657909627, "step": 6185}, {"loss": 0.7507, "grad_norm": 0.7111801505088806, "learning_rate": 0.0002, "epoch": 1.0007275078813354, "step": 6190}, {"loss": 0.6865, "grad_norm": 0.5402125716209412, "learning_rate": 0.0002, "epoch": 1.0023441920620806, "step": 6200}, {"loss": 0.7625, "grad_norm": 0.6098830103874207, "learning_rate": 0.0002, "epoch": 1.003960876242826, "step": 6210}, {"loss": 0.7631, "grad_norm": 0.5829983353614807, "learning_rate": 0.0002, "epoch": 1.0055775604235713, "step": 6220}, {"loss": 0.7188, "grad_norm": 0.5614621043205261, "learning_rate": 0.0002, "epoch": 1.0071942446043165, "step": 6230}, {"loss": 0.7505, "grad_norm": 0.5954238772392273, "learning_rate": 0.0002, "epoch": 1.0088109287850617, "step": 6240}, {"loss": 0.7448, "grad_norm": 0.6480574607849121, "learning_rate": 0.0002, "epoch": 1.0104276129658072, "step": 6250}, {"loss": 0.7514, "grad_norm": 0.6051128506660461, "learning_rate": 0.0002, "epoch": 1.0120442971465524, "step": 6260}, {"loss": 0.7237, "grad_norm": 0.6318870782852173, "learning_rate": 0.0002, "epoch": 1.0136609813272976, "step": 6270}, {"loss": 0.7178, "grad_norm": 0.5048980116844177, "learning_rate": 0.0002, "epoch": 1.015277665508043, "step": 6280}, {"loss": 0.7391, "grad_norm": 0.6346936225891113, "learning_rate": 0.0002, "epoch": 1.0168943496887883, "step": 6290}, {"loss": 0.7486, "grad_norm": 0.5711665749549866, "learning_rate": 0.0002, "epoch": 1.0185110338695336, "step": 6300}, {"loss": 0.6808, "grad_norm": 0.5175361037254333, "learning_rate": 0.0002, "epoch": 1.0201277180502788, "step": 6310}, {"loss": 0.7539, "grad_norm": 0.5360831618309021, "learning_rate": 0.0002, "epoch": 1.0217444022310243, "step": 6320}, {"loss": 0.7112, "grad_norm": 0.614675760269165, "learning_rate": 0.0002, "epoch": 1.0233610864117695, "step": 6330}, {"loss": 0.7748, "grad_norm": 0.5626118183135986, "learning_rate": 0.0002, "epoch": 1.0249777705925147, "step": 6340}, {"loss": 0.7375, "grad_norm": 0.574897289276123, "learning_rate": 0.0002, "epoch": 1.02659445477326, "step": 6350}, {"loss": 0.759, "grad_norm": 0.7185447812080383, "learning_rate": 0.0002, "epoch": 1.0282111389540054, "step": 6360}, {"loss": 0.703, "grad_norm": 0.6705799698829651, "learning_rate": 0.0002, "epoch": 1.0298278231347506, "step": 6370}, {"loss": 0.7139, "grad_norm": 0.6740428805351257, "learning_rate": 0.0002, "epoch": 1.0314445073154959, "step": 6380}, {"loss": 0.7252, "grad_norm": 0.663902759552002, "learning_rate": 0.0002, "epoch": 1.0330611914962413, "step": 6390}, {"loss": 0.7065, "grad_norm": 0.5029543042182922, "learning_rate": 0.0002, "epoch": 1.0346778756769865, "step": 6400}, {"loss": 0.711, "grad_norm": 0.7813863158226013, "learning_rate": 0.0002, "epoch": 1.0362945598577318, "step": 6410}, {"loss": 0.7433, "grad_norm": 0.5396282076835632, "learning_rate": 0.0002, "epoch": 1.037911244038477, "step": 6420}, {"loss": 0.7222, "grad_norm": 0.5253293514251709, "learning_rate": 0.0002, "epoch": 1.0395279282192225, "step": 6430}, {"loss": 0.715, "grad_norm": 0.7236770987510681, "learning_rate": 0.0002, "epoch": 1.0411446123999677, "step": 6440}, {"loss": 0.7259, "grad_norm": 0.5670917630195618, "learning_rate": 0.0002, "epoch": 1.042761296580713, "step": 6450}, {"loss": 0.7195, "grad_norm": 0.6031978726387024, "learning_rate": 0.0002, "epoch": 1.0443779807614582, "step": 6460}, {"loss": 0.7648, "grad_norm": 0.5309213399887085, "learning_rate": 0.0002, "epoch": 1.0459946649422036, "step": 6470}, {"loss": 0.7161, "grad_norm": 0.7114651799201965, "learning_rate": 0.0002, "epoch": 1.0476113491229488, "step": 6480}, {"loss": 0.7583, "grad_norm": 0.5591610670089722, "learning_rate": 0.0002, "epoch": 1.049228033303694, "step": 6490}, {"loss": 0.6645, "grad_norm": 0.5185961127281189, "learning_rate": 0.0002, "epoch": 1.0508447174844395, "step": 6500}, {"loss": 0.7654, "grad_norm": 0.6510552167892456, "learning_rate": 0.0002, "epoch": 1.0524614016651848, "step": 6510}, {"loss": 0.7057, "grad_norm": 0.6557928919792175, "learning_rate": 0.0002, "epoch": 1.05407808584593, "step": 6520}, {"loss": 0.8056, "grad_norm": 0.6973192691802979, "learning_rate": 0.0002, "epoch": 1.0556947700266752, "step": 6530}, {"loss": 0.6793, "grad_norm": 0.6226583123207092, "learning_rate": 0.0002, "epoch": 1.0573114542074207, "step": 6540}, {"loss": 0.7151, "grad_norm": 0.5633195638656616, "learning_rate": 0.0002, "epoch": 1.058928138388166, "step": 6550}, {"loss": 0.7082, "grad_norm": 0.7466658353805542, "learning_rate": 0.0002, "epoch": 1.0605448225689111, "step": 6560}, {"loss": 0.7059, "grad_norm": 0.6462772488594055, "learning_rate": 0.0002, "epoch": 1.0621615067496564, "step": 6570}, {"loss": 0.7046, "grad_norm": 0.5266856551170349, "learning_rate": 0.0002, "epoch": 1.0637781909304018, "step": 6580}, {"loss": 0.7157, "grad_norm": 0.534392774105072, "learning_rate": 0.0002, "epoch": 1.065394875111147, "step": 6590}, {"loss": 0.7115, "grad_norm": 0.7514177560806274, "learning_rate": 0.0002, "epoch": 1.0670115592918923, "step": 6600}, {"loss": 0.7545, "grad_norm": 0.7593035697937012, "learning_rate": 0.0002, "epoch": 1.0686282434726375, "step": 6610}, {"loss": 0.6836, "grad_norm": 0.5277858972549438, "learning_rate": 0.0002, "epoch": 1.070244927653383, "step": 6620}, {"loss": 0.7405, "grad_norm": 0.5573670268058777, "learning_rate": 0.0002, "epoch": 1.0718616118341282, "step": 6630}, {"loss": 0.6774, "grad_norm": 0.6802396774291992, "learning_rate": 0.0002, "epoch": 1.0734782960148734, "step": 6640}, {"loss": 0.723, "grad_norm": 0.7367215752601624, "learning_rate": 0.0002, "epoch": 1.0750949801956189, "step": 6650}, {"loss": 0.7429, "grad_norm": 0.5961891412734985, "learning_rate": 0.0002, "epoch": 1.0767116643763641, "step": 6660}, {"loss": 0.6791, "grad_norm": 0.5736313462257385, "learning_rate": 0.0002, "epoch": 1.0783283485571094, "step": 6670}, {"loss": 0.7178, "grad_norm": 0.619219183921814, "learning_rate": 0.0002, "epoch": 1.0799450327378546, "step": 6680}, {"loss": 0.7318, "grad_norm": 0.6214390993118286, "learning_rate": 0.0002, "epoch": 1.0815617169186, "step": 6690}, {"loss": 0.7554, "grad_norm": 0.564536988735199, "learning_rate": 0.0002, "epoch": 1.0831784010993453, "step": 6700}, {"loss": 0.7362, "grad_norm": 0.5838140249252319, "learning_rate": 0.0002, "epoch": 1.0847950852800905, "step": 6710}, {"loss": 0.739, "grad_norm": 0.7000553607940674, "learning_rate": 0.0002, "epoch": 1.0864117694608357, "step": 6720}, {"loss": 0.7369, "grad_norm": 0.7078263759613037, "learning_rate": 0.0002, "epoch": 1.0880284536415812, "step": 6730}, {"loss": 0.7654, "grad_norm": 0.8353848457336426, "learning_rate": 0.0002, "epoch": 1.0896451378223264, "step": 6740}, {"loss": 0.7015, "grad_norm": 0.5615518689155579, "learning_rate": 0.0002, "epoch": 1.0912618220030716, "step": 6750}, {"loss": 0.7396, "grad_norm": 0.5475581288337708, "learning_rate": 0.0002, "epoch": 1.0928785061838169, "step": 6760}, {"loss": 0.7652, "grad_norm": 0.5835978388786316, "learning_rate": 0.0002, "epoch": 1.0944951903645623, "step": 6770}, {"loss": 0.7541, "grad_norm": 0.5516105890274048, "learning_rate": 0.0002, "epoch": 1.0961118745453076, "step": 6780}, {"loss": 0.6842, "grad_norm": 0.5875251889228821, "learning_rate": 0.0002, "epoch": 1.0977285587260528, "step": 6790}, {"loss": 0.6903, "grad_norm": 0.7376947999000549, "learning_rate": 0.0002, "epoch": 1.0993452429067982, "step": 6800}, {"loss": 0.7512, "grad_norm": 0.5656165480613708, "learning_rate": 0.0002, "epoch": 1.1009619270875435, "step": 6810}, {"loss": 0.7409, "grad_norm": 0.6365954279899597, "learning_rate": 0.0002, "epoch": 1.1025786112682887, "step": 6820}, {"loss": 0.7392, "grad_norm": 0.5033080577850342, "learning_rate": 0.0002, "epoch": 1.104195295449034, "step": 6830}, {"loss": 0.6909, "grad_norm": 0.617396891117096, "learning_rate": 0.0002, "epoch": 1.1058119796297794, "step": 6840}, {"loss": 0.7006, "grad_norm": 0.6395374536514282, "learning_rate": 0.0002, "epoch": 1.1074286638105246, "step": 6850}, {"loss": 0.7335, "grad_norm": 0.6775295734405518, "learning_rate": 0.0002, "epoch": 1.1090453479912699, "step": 6860}, {"loss": 0.764, "grad_norm": 0.6655223965644836, "learning_rate": 0.0002, "epoch": 1.1106620321720153, "step": 6870}, {"loss": 0.7553, "grad_norm": 0.676655113697052, "learning_rate": 0.0002, "epoch": 1.1122787163527605, "step": 6880}, {"loss": 0.7342, "grad_norm": 0.6062718629837036, "learning_rate": 0.0002, "epoch": 1.1138954005335058, "step": 6890}, {"loss": 0.7446, "grad_norm": 0.590943455696106, "learning_rate": 0.0002, "epoch": 1.115512084714251, "step": 6900}, {"loss": 0.6705, "grad_norm": 0.6315317153930664, "learning_rate": 0.0002, "epoch": 1.1171287688949965, "step": 6910}, {"loss": 0.6912, "grad_norm": 0.47979024052619934, "learning_rate": 0.0002, "epoch": 1.1187454530757417, "step": 6920}, {"loss": 0.7002, "grad_norm": 0.647298276424408, "learning_rate": 0.0002, "epoch": 1.120362137256487, "step": 6930}, {"loss": 0.7502, "grad_norm": 0.7336484789848328, "learning_rate": 0.0002, "epoch": 1.1219788214372322, "step": 6940}, {"loss": 0.693, "grad_norm": 0.5071424245834351, "learning_rate": 0.0002, "epoch": 1.1235955056179776, "step": 6950}, {"loss": 0.7378, "grad_norm": 0.6527144312858582, "learning_rate": 0.0002, "epoch": 1.1252121897987228, "step": 6960}, {"loss": 0.7228, "grad_norm": 0.6935935020446777, "learning_rate": 0.0002, "epoch": 1.126828873979468, "step": 6970}, {"loss": 0.699, "grad_norm": 0.8026931881904602, "learning_rate": 0.0002, "epoch": 1.1284455581602133, "step": 6980}, {"loss": 0.7361, "grad_norm": 0.5210393667221069, "learning_rate": 0.0002, "epoch": 1.1300622423409588, "step": 6990}, {"loss": 0.7456, "grad_norm": 0.60475093126297, "learning_rate": 0.0002, "epoch": 1.131678926521704, "step": 7000}, {"loss": 0.7495, "grad_norm": 0.6417073607444763, "learning_rate": 0.0002, "epoch": 1.1332956107024492, "step": 7010}, {"loss": 0.7459, "grad_norm": 0.6732175946235657, "learning_rate": 0.0002, "epoch": 1.1349122948831947, "step": 7020}, {"loss": 0.7278, "grad_norm": 0.6719491481781006, "learning_rate": 0.0002, "epoch": 1.13652897906394, "step": 7030}, {"loss": 0.7694, "grad_norm": 0.5708295106887817, "learning_rate": 0.0002, "epoch": 1.1381456632446851, "step": 7040}, {"loss": 0.7823, "grad_norm": 0.7141719460487366, "learning_rate": 0.0002, "epoch": 1.1397623474254304, "step": 7050}, {"loss": 0.764, "grad_norm": 0.6187017560005188, "learning_rate": 0.0002, "epoch": 1.1413790316061758, "step": 7060}, {"loss": 0.7657, "grad_norm": 0.50581294298172, "learning_rate": 0.0002, "epoch": 1.142995715786921, "step": 7070}, {"loss": 0.7357, "grad_norm": 0.5620143413543701, "learning_rate": 0.0002, "epoch": 1.1446123999676663, "step": 7080}, {"loss": 0.7287, "grad_norm": 0.6231929659843445, "learning_rate": 0.0002, "epoch": 1.1462290841484115, "step": 7090}, {"loss": 0.7328, "grad_norm": 0.5775774121284485, "learning_rate": 0.0002, "epoch": 1.147845768329157, "step": 7100}, {"loss": 0.7728, "grad_norm": 0.6492809653282166, "learning_rate": 0.0002, "epoch": 1.1494624525099022, "step": 7110}, {"loss": 0.7545, "grad_norm": 0.6434972286224365, "learning_rate": 0.0002, "epoch": 1.1510791366906474, "step": 7120}, {"loss": 0.7374, "grad_norm": 0.6191812753677368, "learning_rate": 0.0002, "epoch": 1.1526958208713927, "step": 7130}, {"loss": 0.7276, "grad_norm": 0.6690331697463989, "learning_rate": 0.0002, "epoch": 1.1543125050521381, "step": 7140}, {"loss": 0.7704, "grad_norm": 0.5977938175201416, "learning_rate": 0.0002, "epoch": 1.1559291892328833, "step": 7150}, {"loss": 0.7251, "grad_norm": 0.6195854544639587, "learning_rate": 0.0002, "epoch": 1.1575458734136286, "step": 7160}, {"loss": 0.7249, "grad_norm": 0.5752048492431641, "learning_rate": 0.0002, "epoch": 1.159162557594374, "step": 7170}, {"loss": 0.7593, "grad_norm": 0.589081883430481, "learning_rate": 0.0002, "epoch": 1.1607792417751193, "step": 7180}, {"loss": 0.704, "grad_norm": 0.756996750831604, "learning_rate": 0.0002, "epoch": 1.1623959259558645, "step": 7190}, {"loss": 0.7404, "grad_norm": 0.7614967226982117, "learning_rate": 0.0002, "epoch": 1.1640126101366097, "step": 7200}, {"loss": 0.7867, "grad_norm": 0.6120437979698181, "learning_rate": 0.0002, "epoch": 1.1656292943173552, "step": 7210}, {"loss": 0.7384, "grad_norm": 0.6210004687309265, "learning_rate": 0.0002, "epoch": 1.1672459784981004, "step": 7220}, {"loss": 0.7251, "grad_norm": 0.6044116020202637, "learning_rate": 0.0002, "epoch": 1.1688626626788456, "step": 7230}, {"loss": 0.7361, "grad_norm": 0.5418457388877869, "learning_rate": 0.0002, "epoch": 1.170479346859591, "step": 7240}, {"loss": 0.6938, "grad_norm": 0.6413537263870239, "learning_rate": 0.0002, "epoch": 1.1720960310403363, "step": 7250}, {"loss": 0.6978, "grad_norm": 0.5777867436408997, "learning_rate": 0.0002, "epoch": 1.1737127152210816, "step": 7260}, {"loss": 0.7503, "grad_norm": 0.7092402577400208, "learning_rate": 0.0002, "epoch": 1.1753293994018268, "step": 7270}, {"loss": 0.7487, "grad_norm": 0.6351709365844727, "learning_rate": 0.0002, "epoch": 1.176946083582572, "step": 7280}, {"loss": 0.7527, "grad_norm": 0.6172189712524414, "learning_rate": 0.0002, "epoch": 1.1785627677633175, "step": 7290}, {"loss": 0.7319, "grad_norm": 0.6801714897155762, "learning_rate": 0.0002, "epoch": 1.1801794519440627, "step": 7300}, {"loss": 0.6941, "grad_norm": 0.6044712066650391, "learning_rate": 0.0002, "epoch": 1.181796136124808, "step": 7310}, {"loss": 0.6951, "grad_norm": 0.7413212060928345, "learning_rate": 0.0002, "epoch": 1.1834128203055534, "step": 7320}, {"loss": 0.7396, "grad_norm": 0.5303856134414673, "learning_rate": 0.0002, "epoch": 1.1850295044862986, "step": 7330}, {"loss": 0.6915, "grad_norm": 0.5647098422050476, "learning_rate": 0.0002, "epoch": 1.1866461886670439, "step": 7340}, {"loss": 0.7506, "grad_norm": 0.7374135255813599, "learning_rate": 0.0002, "epoch": 1.188262872847789, "step": 7350}, {"loss": 0.7041, "grad_norm": 0.5710089206695557, "learning_rate": 0.0002, "epoch": 1.1898795570285345, "step": 7360}, {"loss": 0.8289, "grad_norm": 0.6073619723320007, "learning_rate": 0.0002, "epoch": 1.1914962412092798, "step": 7370}, {"loss": 0.7722, "grad_norm": 0.5899916887283325, "learning_rate": 0.0002, "epoch": 1.193112925390025, "step": 7380}, {"loss": 0.756, "grad_norm": 0.7762434482574463, "learning_rate": 0.0002, "epoch": 1.1947296095707705, "step": 7390}, {"loss": 0.7319, "grad_norm": 0.679949939250946, "learning_rate": 0.0002, "epoch": 1.1963462937515157, "step": 7400}, {"loss": 0.7599, "grad_norm": 0.6106849312782288, "learning_rate": 0.0002, "epoch": 1.197962977932261, "step": 7410}, {"loss": 0.7648, "grad_norm": 0.682461678981781, "learning_rate": 0.0002, "epoch": 1.1995796621130062, "step": 7420}, {"loss": 0.7741, "grad_norm": 0.6087017059326172, "learning_rate": 0.0002, "epoch": 1.2011963462937516, "step": 7430}, {"loss": 0.7642, "grad_norm": 0.63739013671875, "learning_rate": 0.0002, "epoch": 1.2028130304744968, "step": 7440}, {"loss": 0.7611, "grad_norm": 0.6154777407646179, "learning_rate": 0.0002, "epoch": 1.204429714655242, "step": 7450}, {"loss": 0.7565, "grad_norm": 0.7491534948348999, "learning_rate": 0.0002, "epoch": 1.2060463988359873, "step": 7460}, {"loss": 0.698, "grad_norm": 0.6664797067642212, "learning_rate": 0.0002, "epoch": 1.2076630830167328, "step": 7470}, {"loss": 0.7456, "grad_norm": 0.6660266518592834, "learning_rate": 0.0002, "epoch": 1.209279767197478, "step": 7480}, {"loss": 0.714, "grad_norm": 0.6972551345825195, "learning_rate": 0.0002, "epoch": 1.2108964513782232, "step": 7490}, {"loss": 0.7023, "grad_norm": 0.6157945990562439, "learning_rate": 0.0002, "epoch": 1.2125131355589684, "step": 7500}, {"loss": 0.7326, "grad_norm": 0.5199310183525085, "learning_rate": 0.0002, "epoch": 1.214129819739714, "step": 7510}, {"loss": 0.7586, "grad_norm": 0.577610433101654, "learning_rate": 0.0002, "epoch": 1.2157465039204591, "step": 7520}, {"loss": 0.7179, "grad_norm": 0.53652423620224, "learning_rate": 0.0002, "epoch": 1.2173631881012044, "step": 7530}, {"loss": 0.7393, "grad_norm": 0.6479050517082214, "learning_rate": 0.0002, "epoch": 1.2189798722819498, "step": 7540}, {"loss": 0.7534, "grad_norm": 0.618748128414154, "learning_rate": 0.0002, "epoch": 1.220596556462695, "step": 7550}, {"loss": 0.6886, "grad_norm": 0.6311424374580383, "learning_rate": 0.0002, "epoch": 1.2222132406434403, "step": 7560}, {"loss": 0.7272, "grad_norm": 0.6595825552940369, "learning_rate": 0.0002, "epoch": 1.2238299248241855, "step": 7570}, {"loss": 0.7353, "grad_norm": 0.5198960900306702, "learning_rate": 0.0002, "epoch": 1.225446609004931, "step": 7580}, {"loss": 0.674, "grad_norm": 0.578650712966919, "learning_rate": 0.0002, "epoch": 1.2270632931856762, "step": 7590}, {"loss": 0.7507, "grad_norm": 0.6080220937728882, "learning_rate": 0.0002, "epoch": 1.2286799773664214, "step": 7600}, {"loss": 0.7733, "grad_norm": 0.7050248384475708, "learning_rate": 0.0002, "epoch": 1.2302966615471669, "step": 7610}, {"loss": 0.7032, "grad_norm": 0.6652196049690247, "learning_rate": 0.0002, "epoch": 1.2319133457279121, "step": 7620}, {"loss": 0.7085, "grad_norm": 0.7322776317596436, "learning_rate": 0.0002, "epoch": 1.2335300299086573, "step": 7630}, {"loss": 0.7402, "grad_norm": 0.4998728036880493, "learning_rate": 0.0002, "epoch": 1.2351467140894026, "step": 7640}, {"loss": 0.7214, "grad_norm": 0.6428788900375366, "learning_rate": 0.0002, "epoch": 1.2367633982701478, "step": 7650}, {"loss": 0.7699, "grad_norm": 0.585242509841919, "learning_rate": 0.0002, "epoch": 1.2383800824508933, "step": 7660}, {"loss": 0.7621, "grad_norm": 0.5211917757987976, "learning_rate": 0.0002, "epoch": 1.2399967666316385, "step": 7670}, {"loss": 0.746, "grad_norm": 0.6490384340286255, "learning_rate": 0.0002, "epoch": 1.2416134508123837, "step": 7680}, {"loss": 0.7186, "grad_norm": 0.6249763369560242, "learning_rate": 0.0002, "epoch": 1.2432301349931292, "step": 7690}, {"loss": 0.7761, "grad_norm": 0.71870356798172, "learning_rate": 0.0002, "epoch": 1.2448468191738744, "step": 7700}, {"loss": 0.7525, "grad_norm": 0.6761967539787292, "learning_rate": 0.0002, "epoch": 1.2464635033546196, "step": 7710}, {"loss": 0.7501, "grad_norm": 0.6500617265701294, "learning_rate": 0.0002, "epoch": 1.2480801875353649, "step": 7720}, {"loss": 0.7903, "grad_norm": 0.8069869875907898, "learning_rate": 0.0002, "epoch": 1.2496968717161103, "step": 7730}, {"loss": 0.6747, "grad_norm": 0.6044608950614929, "learning_rate": 0.0002, "epoch": 1.2513135558968556, "step": 7740}, {"loss": 0.6825, "grad_norm": 0.6573283076286316, "learning_rate": 0.0002, "epoch": 1.2529302400776008, "step": 7750}, {"loss": 0.7617, "grad_norm": 0.625430166721344, "learning_rate": 0.0002, "epoch": 1.2545469242583462, "step": 7760}, {"loss": 0.7041, "grad_norm": 0.5442022681236267, "learning_rate": 0.0002, "epoch": 1.2561636084390915, "step": 7770}, {"loss": 0.7172, "grad_norm": 0.6818386912345886, "learning_rate": 0.0002, "epoch": 1.2577802926198367, "step": 7780}, {"loss": 0.696, "grad_norm": 0.6381874084472656, "learning_rate": 0.0002, "epoch": 1.259396976800582, "step": 7790}, {"loss": 0.6834, "grad_norm": 0.6269212961196899, "learning_rate": 0.0002, "epoch": 1.2610136609813272, "step": 7800}, {"loss": 0.7821, "grad_norm": 0.600121259689331, "learning_rate": 0.0002, "epoch": 1.2626303451620726, "step": 7810}, {"loss": 0.7761, "grad_norm": 0.6337703466415405, "learning_rate": 0.0002, "epoch": 1.2642470293428179, "step": 7820}, {"loss": 0.732, "grad_norm": 0.7234963774681091, "learning_rate": 0.0002, "epoch": 1.2658637135235633, "step": 7830}, {"loss": 0.785, "grad_norm": 0.800184965133667, "learning_rate": 0.0002, "epoch": 1.2674803977043085, "step": 7840}, {"loss": 0.7426, "grad_norm": 0.7539464831352234, "learning_rate": 0.0002, "epoch": 1.2690970818850538, "step": 7850}, {"loss": 0.7496, "grad_norm": 0.5493760704994202, "learning_rate": 0.0002, "epoch": 1.270713766065799, "step": 7860}, {"loss": 0.7537, "grad_norm": 0.7477145791053772, "learning_rate": 0.0002, "epoch": 1.2723304502465442, "step": 7870}, {"loss": 0.7573, "grad_norm": 0.6366362571716309, "learning_rate": 0.0002, "epoch": 1.2739471344272897, "step": 7880}, {"loss": 0.7608, "grad_norm": 0.7419533729553223, "learning_rate": 0.0002, "epoch": 1.275563818608035, "step": 7890}, {"loss": 0.7873, "grad_norm": 0.6141223311424255, "learning_rate": 0.0002, "epoch": 1.2771805027887801, "step": 7900}, {"loss": 0.6916, "grad_norm": 0.7522598505020142, "learning_rate": 0.0002, "epoch": 1.2787971869695256, "step": 7910}, {"loss": 0.7097, "grad_norm": 0.6935804486274719, "learning_rate": 0.0002, "epoch": 1.2804138711502708, "step": 7920}, {"loss": 0.7185, "grad_norm": 0.7239290475845337, "learning_rate": 0.0002, "epoch": 1.282030555331016, "step": 7930}, {"loss": 0.7145, "grad_norm": 0.8800187110900879, "learning_rate": 0.0002, "epoch": 1.2836472395117613, "step": 7940}, {"loss": 0.6991, "grad_norm": 0.540458083152771, "learning_rate": 0.0002, "epoch": 1.2852639236925067, "step": 7950}, {"loss": 0.7139, "grad_norm": 0.6492934226989746, "learning_rate": 0.0002, "epoch": 1.286880607873252, "step": 7960}, {"loss": 0.7742, "grad_norm": 0.6543959379196167, "learning_rate": 0.0002, "epoch": 1.2884972920539972, "step": 7970}, {"loss": 0.7316, "grad_norm": 0.5804705619812012, "learning_rate": 0.0002, "epoch": 1.2901139762347427, "step": 7980}, {"loss": 0.796, "grad_norm": 0.7074727416038513, "learning_rate": 0.0002, "epoch": 1.291730660415488, "step": 7990}, {"loss": 0.7034, "grad_norm": 0.5347974300384521, "learning_rate": 0.0002, "epoch": 1.2933473445962331, "step": 8000}, {"loss": 0.738, "grad_norm": 0.6457298398017883, "learning_rate": 0.0002, "epoch": 1.2949640287769784, "step": 8010}, {"loss": 0.7634, "grad_norm": 0.6407219171524048, "learning_rate": 0.0002, "epoch": 1.2965807129577236, "step": 8020}, {"loss": 0.7506, "grad_norm": 0.828439474105835, "learning_rate": 0.0002, "epoch": 1.298197397138469, "step": 8030}, {"loss": 0.735, "grad_norm": 0.4840380549430847, "learning_rate": 0.0002, "epoch": 1.2998140813192143, "step": 8040}, {"loss": 0.7283, "grad_norm": 0.5921024680137634, "learning_rate": 0.0002, "epoch": 1.3014307654999595, "step": 8050}, {"loss": 0.7477, "grad_norm": 0.6170315146446228, "learning_rate": 0.0002, "epoch": 1.303047449680705, "step": 8060}, {"loss": 0.7534, "grad_norm": 0.5374847054481506, "learning_rate": 0.0002, "epoch": 1.3046641338614502, "step": 8070}, {"loss": 0.7593, "grad_norm": 0.545758068561554, "learning_rate": 0.0002, "epoch": 1.3062808180421954, "step": 8080}, {"loss": 0.7463, "grad_norm": 0.55641770362854, "learning_rate": 0.0002, "epoch": 1.3078975022229407, "step": 8090}, {"loss": 0.7594, "grad_norm": 0.6724897027015686, "learning_rate": 0.0002, "epoch": 1.309514186403686, "step": 8100}, {"loss": 0.7105, "grad_norm": 0.6923972368240356, "learning_rate": 0.0002, "epoch": 1.3111308705844313, "step": 8110}, {"loss": 0.7149, "grad_norm": 0.5136841535568237, "learning_rate": 0.0002, "epoch": 1.3127475547651766, "step": 8120}, {"loss": 0.7504, "grad_norm": 0.6766283512115479, "learning_rate": 0.0002, "epoch": 1.314364238945922, "step": 8130}, {"loss": 0.7489, "grad_norm": 0.6283926367759705, "learning_rate": 0.0002, "epoch": 1.3159809231266673, "step": 8140}, {"loss": 0.7459, "grad_norm": 0.644216001033783, "learning_rate": 0.0002, "epoch": 1.3175976073074125, "step": 8150}, {"loss": 0.7125, "grad_norm": 0.7827503085136414, "learning_rate": 0.0002, "epoch": 1.3192142914881577, "step": 8160}, {"loss": 0.7271, "grad_norm": 0.6651390790939331, "learning_rate": 0.0002, "epoch": 1.320830975668903, "step": 8170}, {"loss": 0.7778, "grad_norm": 0.5547412633895874, "learning_rate": 0.0002, "epoch": 1.3224476598496484, "step": 8180}, {"loss": 0.7402, "grad_norm": 0.6765179634094238, "learning_rate": 0.0002, "epoch": 1.3240643440303936, "step": 8190}, {"loss": 0.7106, "grad_norm": 0.6822077035903931, "learning_rate": 0.0002, "epoch": 1.325681028211139, "step": 8200}, {"loss": 0.7288, "grad_norm": 0.5941002368927002, "learning_rate": 0.0002, "epoch": 1.3272977123918843, "step": 8210}, {"loss": 0.7494, "grad_norm": 0.4850037097930908, "learning_rate": 0.0002, "epoch": 1.3289143965726296, "step": 8220}, {"loss": 0.7474, "grad_norm": 0.6162990927696228, "learning_rate": 0.0002, "epoch": 1.3305310807533748, "step": 8230}, {"loss": 0.7751, "grad_norm": 0.6665613651275635, "learning_rate": 0.0002, "epoch": 1.33214776493412, "step": 8240}, {"loss": 0.759, "grad_norm": 0.618192732334137, "learning_rate": 0.0002, "epoch": 1.3337644491148655, "step": 8250}, {"loss": 0.7532, "grad_norm": 0.710418701171875, "learning_rate": 0.0002, "epoch": 1.3353811332956107, "step": 8260}, {"loss": 0.7306, "grad_norm": 0.5109876990318298, "learning_rate": 0.0002, "epoch": 1.336997817476356, "step": 8270}, {"loss": 0.7303, "grad_norm": 0.6791711449623108, "learning_rate": 0.0002, "epoch": 1.3386145016571014, "step": 8280}, {"loss": 0.7594, "grad_norm": 0.6836432814598083, "learning_rate": 0.0002, "epoch": 1.3402311858378466, "step": 8290}, {"loss": 0.7594, "grad_norm": 0.5579386353492737, "learning_rate": 0.0002, "epoch": 1.3418478700185918, "step": 8300}, {"loss": 0.7377, "grad_norm": 0.6713546514511108, "learning_rate": 0.0002, "epoch": 1.343464554199337, "step": 8310}, {"loss": 0.7756, "grad_norm": 0.5353720188140869, "learning_rate": 0.0002, "epoch": 1.3450812383800825, "step": 8320}, {"loss": 0.718, "grad_norm": 0.5813682675361633, "learning_rate": 0.0002, "epoch": 1.3466979225608278, "step": 8330}, {"loss": 0.7294, "grad_norm": 0.8158791661262512, "learning_rate": 0.0002, "epoch": 1.348314606741573, "step": 8340}, {"loss": 0.6992, "grad_norm": 0.6193785071372986, "learning_rate": 0.0002, "epoch": 1.3499312909223184, "step": 8350}, {"loss": 0.7654, "grad_norm": 0.6353939771652222, "learning_rate": 0.0002, "epoch": 1.3515479751030637, "step": 8360}, {"loss": 0.7519, "grad_norm": 0.6925048232078552, "learning_rate": 0.0002, "epoch": 1.353164659283809, "step": 8370}, {"loss": 0.736, "grad_norm": 0.988264799118042, "learning_rate": 0.0002, "epoch": 1.3547813434645541, "step": 8380}, {"loss": 0.7744, "grad_norm": 0.6476002931594849, "learning_rate": 0.0002, "epoch": 1.3563980276452994, "step": 8390}, {"loss": 0.776, "grad_norm": 0.7120398879051208, "learning_rate": 0.0002, "epoch": 1.3580147118260448, "step": 8400}, {"loss": 0.7368, "grad_norm": 0.9048416614532471, "learning_rate": 0.0002, "epoch": 1.35963139600679, "step": 8410}, {"loss": 0.7544, "grad_norm": 0.7000672817230225, "learning_rate": 0.0002, "epoch": 1.3612480801875353, "step": 8420}, {"loss": 0.7358, "grad_norm": 0.6015632152557373, "learning_rate": 0.0002, "epoch": 1.3628647643682807, "step": 8430}, {"loss": 0.7298, "grad_norm": 0.612516462802887, "learning_rate": 0.0002, "epoch": 1.364481448549026, "step": 8440}, {"loss": 0.7055, "grad_norm": 0.5969301462173462, "learning_rate": 0.0002, "epoch": 1.3660981327297712, "step": 8450}, {"loss": 0.7754, "grad_norm": 0.6730654239654541, "learning_rate": 0.0002, "epoch": 1.3677148169105164, "step": 8460}, {"loss": 0.7465, "grad_norm": 0.6386392116546631, "learning_rate": 0.0002, "epoch": 1.369331501091262, "step": 8470}, {"loss": 0.7433, "grad_norm": 0.739544153213501, "learning_rate": 0.0002, "epoch": 1.3709481852720071, "step": 8480}, {"loss": 0.7892, "grad_norm": 0.6462782621383667, "learning_rate": 0.0002, "epoch": 1.3725648694527524, "step": 8490}, {"loss": 0.7302, "grad_norm": 0.7346843481063843, "learning_rate": 0.0002, "epoch": 1.3741815536334978, "step": 8500}, {"loss": 0.7634, "grad_norm": 0.6884821057319641, "learning_rate": 0.0002, "epoch": 1.375798237814243, "step": 8510}, {"loss": 0.7614, "grad_norm": 0.6999333500862122, "learning_rate": 0.0002, "epoch": 1.3774149219949883, "step": 8520}, {"loss": 0.729, "grad_norm": 0.5378713011741638, "learning_rate": 0.0002, "epoch": 1.3790316061757335, "step": 8530}, {"loss": 0.6797, "grad_norm": 0.5417906641960144, "learning_rate": 0.0002, "epoch": 1.3806482903564787, "step": 8540}, {"loss": 0.7499, "grad_norm": 0.6602526307106018, "learning_rate": 0.0002, "epoch": 1.3822649745372242, "step": 8550}, {"loss": 0.7356, "grad_norm": 0.7073674201965332, "learning_rate": 0.0002, "epoch": 1.3838816587179694, "step": 8560}, {"loss": 0.75, "grad_norm": 0.5841707587242126, "learning_rate": 0.0002, "epoch": 1.3854983428987149, "step": 8570}, {"loss": 0.732, "grad_norm": 0.7031095027923584, "learning_rate": 0.0002, "epoch": 1.38711502707946, "step": 8580}, {"loss": 0.7464, "grad_norm": 0.5198570489883423, "learning_rate": 0.0002, "epoch": 1.3887317112602053, "step": 8590}, {"loss": 0.7354, "grad_norm": 0.7261320352554321, "learning_rate": 0.0002, "epoch": 1.3903483954409506, "step": 8600}, {"loss": 0.7339, "grad_norm": 0.5616350173950195, "learning_rate": 0.0002, "epoch": 1.3919650796216958, "step": 8610}, {"loss": 0.7382, "grad_norm": 0.5185914635658264, "learning_rate": 0.0002, "epoch": 1.3935817638024413, "step": 8620}, {"loss": 0.7456, "grad_norm": 0.5814694762229919, "learning_rate": 0.0002, "epoch": 1.3951984479831865, "step": 8630}, {"loss": 0.7413, "grad_norm": 0.6977371573448181, "learning_rate": 0.0002, "epoch": 1.3968151321639317, "step": 8640}, {"loss": 0.7574, "grad_norm": 0.6855689883232117, "learning_rate": 0.0002, "epoch": 1.3984318163446772, "step": 8650}, {"loss": 0.7802, "grad_norm": 0.5414357781410217, "learning_rate": 0.0002, "epoch": 1.4000485005254224, "step": 8660}, {"loss": 0.7487, "grad_norm": 0.6970012784004211, "learning_rate": 0.0002, "epoch": 1.4016651847061676, "step": 8670}, {"loss": 0.7421, "grad_norm": 0.526079535484314, "learning_rate": 0.0002, "epoch": 1.4032818688869129, "step": 8680}, {"loss": 0.737, "grad_norm": 0.758712887763977, "learning_rate": 0.0002, "epoch": 1.404898553067658, "step": 8690}, {"loss": 0.7612, "grad_norm": 0.7118762731552124, "learning_rate": 0.0002, "epoch": 1.4065152372484035, "step": 8700}, {"loss": 0.7628, "grad_norm": 0.5696909427642822, "learning_rate": 0.0002, "epoch": 1.4081319214291488, "step": 8710}, {"loss": 0.7156, "grad_norm": 0.7995436787605286, "learning_rate": 0.0002, "epoch": 1.4097486056098942, "step": 8720}, {"loss": 0.7521, "grad_norm": 0.7237521409988403, "learning_rate": 0.0002, "epoch": 1.4113652897906395, "step": 8730}, {"loss": 0.7661, "grad_norm": 0.744628369808197, "learning_rate": 0.0002, "epoch": 1.4129819739713847, "step": 8740}, {"loss": 0.7073, "grad_norm": 0.6082926988601685, "learning_rate": 0.0002, "epoch": 1.41459865815213, "step": 8750}, {"loss": 0.7282, "grad_norm": 0.5185243487358093, "learning_rate": 0.0002, "epoch": 1.4162153423328752, "step": 8760}, {"loss": 0.7592, "grad_norm": 0.5183082222938538, "learning_rate": 0.0002, "epoch": 1.4178320265136206, "step": 8770}, {"loss": 0.7509, "grad_norm": 0.7326041460037231, "learning_rate": 0.0002, "epoch": 1.4194487106943658, "step": 8780}, {"loss": 0.7398, "grad_norm": 0.7174660563468933, "learning_rate": 0.0002, "epoch": 1.421065394875111, "step": 8790}, {"loss": 0.7507, "grad_norm": 0.8080165982246399, "learning_rate": 0.0002, "epoch": 1.4226820790558565, "step": 8800}, {"loss": 0.72, "grad_norm": 0.5061507821083069, "learning_rate": 0.0002, "epoch": 1.4242987632366018, "step": 8810}, {"loss": 0.7563, "grad_norm": 0.801602840423584, "learning_rate": 0.0002, "epoch": 1.425915447417347, "step": 8820}, {"loss": 0.7287, "grad_norm": 0.6150273084640503, "learning_rate": 0.0002, "epoch": 1.4275321315980922, "step": 8830}, {"loss": 0.7452, "grad_norm": 0.8786525726318359, "learning_rate": 0.0002, "epoch": 1.4291488157788377, "step": 8840}, {"loss": 0.7257, "grad_norm": 0.6371538639068604, "learning_rate": 0.0002, "epoch": 1.430765499959583, "step": 8850}, {"loss": 0.711, "grad_norm": 0.6409295797348022, "learning_rate": 0.0002, "epoch": 1.4323821841403281, "step": 8860}, {"loss": 0.7891, "grad_norm": 0.6452359557151794, "learning_rate": 0.0002, "epoch": 1.4339988683210736, "step": 8870}, {"loss": 0.7588, "grad_norm": 0.5842334628105164, "learning_rate": 0.0002, "epoch": 1.4356155525018188, "step": 8880}, {"loss": 0.7446, "grad_norm": 0.696761965751648, "learning_rate": 0.0002, "epoch": 1.437232236682564, "step": 8890}, {"loss": 0.7541, "grad_norm": 0.6384600400924683, "learning_rate": 0.0002, "epoch": 1.4388489208633093, "step": 8900}, {"loss": 0.7049, "grad_norm": 0.5981136560440063, "learning_rate": 0.0002, "epoch": 1.4404656050440545, "step": 8910}, {"loss": 0.795, "grad_norm": 0.6355637907981873, "learning_rate": 0.0002, "epoch": 1.4420822892248, "step": 8920}, {"loss": 0.7653, "grad_norm": 0.6374830603599548, "learning_rate": 0.0002, "epoch": 1.4436989734055452, "step": 8930}, {"loss": 0.8108, "grad_norm": 0.559013307094574, "learning_rate": 0.0002, "epoch": 1.4453156575862904, "step": 8940}, {"loss": 0.7045, "grad_norm": 0.7289170026779175, "learning_rate": 0.0002, "epoch": 1.446932341767036, "step": 8950}, {"loss": 0.7484, "grad_norm": 0.8649206757545471, "learning_rate": 0.0002, "epoch": 1.4485490259477811, "step": 8960}, {"loss": 0.7745, "grad_norm": 0.7664689421653748, "learning_rate": 0.0002, "epoch": 1.4501657101285264, "step": 8970}, {"loss": 0.7431, "grad_norm": 0.7109952569007874, "learning_rate": 0.0002, "epoch": 1.4517823943092716, "step": 8980}, {"loss": 0.7997, "grad_norm": 0.6312844753265381, "learning_rate": 0.0002, "epoch": 1.453399078490017, "step": 8990}, {"loss": 0.7467, "grad_norm": 0.6616617441177368, "learning_rate": 0.0002, "epoch": 1.4550157626707623, "step": 9000}, {"loss": 0.7518, "grad_norm": 0.7384068965911865, "learning_rate": 0.0002, "epoch": 1.4566324468515075, "step": 9010}, {"loss": 0.7483, "grad_norm": 0.6549670100212097, "learning_rate": 0.0002, "epoch": 1.458249131032253, "step": 9020}, {"loss": 0.7423, "grad_norm": 0.6254119277000427, "learning_rate": 0.0002, "epoch": 1.4598658152129982, "step": 9030}, {"loss": 0.7645, "grad_norm": 0.6806328892707825, "learning_rate": 0.0002, "epoch": 1.4614824993937434, "step": 9040}, {"loss": 0.7221, "grad_norm": 0.6803115010261536, "learning_rate": 0.0002, "epoch": 1.4630991835744886, "step": 9050}, {"loss": 0.7264, "grad_norm": 0.48529282212257385, "learning_rate": 0.0002, "epoch": 1.4647158677552339, "step": 9060}, {"loss": 0.7542, "grad_norm": 0.5995030999183655, "learning_rate": 0.0002, "epoch": 1.4663325519359793, "step": 9070}, {"loss": 0.7894, "grad_norm": 0.6005427837371826, "learning_rate": 0.0002, "epoch": 1.4679492361167246, "step": 9080}, {"loss": 0.7288, "grad_norm": 0.718564510345459, "learning_rate": 0.0002, "epoch": 1.46956592029747, "step": 9090}, {"loss": 0.7089, "grad_norm": 0.7003577351570129, "learning_rate": 0.0002, "epoch": 1.4711826044782153, "step": 9100}, {"loss": 0.8069, "grad_norm": 0.5888323783874512, "learning_rate": 0.0002, "epoch": 1.4727992886589605, "step": 9110}, {"loss": 0.7275, "grad_norm": 0.6417609453201294, "learning_rate": 0.0002, "epoch": 1.4744159728397057, "step": 9120}, {"loss": 0.7441, "grad_norm": 0.572294294834137, "learning_rate": 0.0002, "epoch": 1.476032657020451, "step": 9130}, {"loss": 0.8053, "grad_norm": 0.8200714588165283, "learning_rate": 0.0002, "epoch": 1.4776493412011964, "step": 9140}, {"loss": 0.7382, "grad_norm": 0.6343288421630859, "learning_rate": 0.0002, "epoch": 1.4792660253819416, "step": 9150}, {"loss": 0.7641, "grad_norm": 0.7017961144447327, "learning_rate": 0.0002, "epoch": 1.4808827095626869, "step": 9160}, {"loss": 0.7619, "grad_norm": 0.6202912926673889, "learning_rate": 0.0002, "epoch": 1.4824993937434323, "step": 9170}, {"loss": 0.7428, "grad_norm": 0.6677869558334351, "learning_rate": 0.0002, "epoch": 1.4841160779241775, "step": 9180}, {"loss": 0.7648, "grad_norm": 0.6052267551422119, "learning_rate": 0.0002, "epoch": 1.4857327621049228, "step": 9190}, {"loss": 0.7152, "grad_norm": 0.6638872027397156, "learning_rate": 0.0002, "epoch": 1.487349446285668, "step": 9200}, {"loss": 0.7448, "grad_norm": 0.6245523691177368, "learning_rate": 0.0002, "epoch": 1.4889661304664135, "step": 9210}, {"loss": 0.6958, "grad_norm": 0.5761767625808716, "learning_rate": 0.0002, "epoch": 1.4905828146471587, "step": 9220}, {"loss": 0.8012, "grad_norm": 0.8175981640815735, "learning_rate": 0.0002, "epoch": 1.492199498827904, "step": 9230}, {"loss": 0.683, "grad_norm": 0.9144009947776794, "learning_rate": 0.0002, "epoch": 1.4938161830086494, "step": 9240}, {"loss": 0.7623, "grad_norm": 0.5742552876472473, "learning_rate": 0.0002, "epoch": 1.4954328671893946, "step": 9250}, {"loss": 0.7418, "grad_norm": 0.534534215927124, "learning_rate": 0.0002, "epoch": 1.4970495513701398, "step": 9260}, {"loss": 0.7194, "grad_norm": 0.7836225032806396, "learning_rate": 0.0002, "epoch": 1.498666235550885, "step": 9270}, {"loss": 0.7453, "grad_norm": 0.5292993187904358, "learning_rate": 0.0002, "epoch": 1.5002829197316303, "step": 9280}, {"loss": 0.7168, "grad_norm": 0.8044071793556213, "learning_rate": 0.0002, "epoch": 1.5018996039123758, "step": 9290}, {"loss": 0.7229, "grad_norm": 0.6185805201530457, "learning_rate": 0.0002, "epoch": 1.503516288093121, "step": 9300}, {"loss": 0.684, "grad_norm": 0.6093607544898987, "learning_rate": 0.0002, "epoch": 1.5051329722738664, "step": 9310}, {"loss": 0.7973, "grad_norm": 0.5891730189323425, "learning_rate": 0.0002, "epoch": 1.5067496564546117, "step": 9320}, {"loss": 0.7474, "grad_norm": 0.6331129670143127, "learning_rate": 0.0002, "epoch": 1.508366340635357, "step": 9330}, {"loss": 0.7074, "grad_norm": 0.7690958380699158, "learning_rate": 0.0002, "epoch": 1.5099830248161021, "step": 9340}, {"loss": 0.672, "grad_norm": 0.6548877358436584, "learning_rate": 0.0002, "epoch": 1.5115997089968474, "step": 9350}, {"loss": 0.7408, "grad_norm": 0.6545143127441406, "learning_rate": 0.0002, "epoch": 1.5132163931775926, "step": 9360}, {"loss": 0.7432, "grad_norm": 0.553247332572937, "learning_rate": 0.0002, "epoch": 1.514833077358338, "step": 9370}, {"loss": 0.7265, "grad_norm": 0.8145074844360352, "learning_rate": 0.0002, "epoch": 1.5164497615390833, "step": 9380}, {"loss": 0.7379, "grad_norm": 0.7636994123458862, "learning_rate": 0.0002, "epoch": 1.5180664457198287, "step": 9390}, {"loss": 0.7413, "grad_norm": 0.6838982701301575, "learning_rate": 0.0002, "epoch": 1.519683129900574, "step": 9400}, {"loss": 0.7367, "grad_norm": 0.8599441647529602, "learning_rate": 0.0002, "epoch": 1.5212998140813192, "step": 9410}, {"loss": 0.7663, "grad_norm": 0.7020329833030701, "learning_rate": 0.0002, "epoch": 1.5229164982620644, "step": 9420}, {"loss": 0.7928, "grad_norm": 0.6964772343635559, "learning_rate": 0.0002, "epoch": 1.5245331824428097, "step": 9430}, {"loss": 0.7168, "grad_norm": 0.6916600465774536, "learning_rate": 0.0002, "epoch": 1.5261498666235551, "step": 9440}, {"loss": 0.7519, "grad_norm": 0.7282621264457703, "learning_rate": 0.0002, "epoch": 1.5277665508043003, "step": 9450}, {"loss": 0.7628, "grad_norm": 0.5363983511924744, "learning_rate": 0.0002, "epoch": 1.5293832349850458, "step": 9460}, {"loss": 0.7154, "grad_norm": 0.6184861063957214, "learning_rate": 0.0002, "epoch": 1.530999919165791, "step": 9470}, {"loss": 0.7837, "grad_norm": 0.5991285443305969, "learning_rate": 0.0002, "epoch": 1.5326166033465363, "step": 9480}, {"loss": 0.7827, "grad_norm": 0.8176587820053101, "learning_rate": 0.0002, "epoch": 1.5342332875272815, "step": 9490}, {"loss": 0.7415, "grad_norm": 0.6473721861839294, "learning_rate": 0.0002, "epoch": 1.5358499717080267, "step": 9500}, {"loss": 0.7632, "grad_norm": 0.7319952845573425, "learning_rate": 0.0002, "epoch": 1.5374666558887722, "step": 9510}, {"loss": 0.7706, "grad_norm": 0.702900230884552, "learning_rate": 0.0002, "epoch": 1.5390833400695174, "step": 9520}, {"loss": 0.7754, "grad_norm": 0.7971600294113159, "learning_rate": 0.0002, "epoch": 1.5407000242502629, "step": 9530}, {"loss": 0.7352, "grad_norm": 0.6527525186538696, "learning_rate": 0.0002, "epoch": 1.542316708431008, "step": 9540}, {"loss": 0.7425, "grad_norm": 0.5791676044464111, "learning_rate": 0.0002, "epoch": 1.5439333926117533, "step": 9550}, {"loss": 0.7585, "grad_norm": 0.5619390606880188, "learning_rate": 0.0002, "epoch": 1.5455500767924986, "step": 9560}, {"loss": 0.7894, "grad_norm": 0.5701689124107361, "learning_rate": 0.0002, "epoch": 1.5471667609732438, "step": 9570}, {"loss": 0.793, "grad_norm": 0.47549352049827576, "learning_rate": 0.0002, "epoch": 1.548783445153989, "step": 9580}, {"loss": 0.7276, "grad_norm": 0.8730611205101013, "learning_rate": 0.0002, "epoch": 1.5504001293347345, "step": 9590}, {"loss": 0.798, "grad_norm": 0.6842091083526611, "learning_rate": 0.0002, "epoch": 1.5520168135154797, "step": 9600}, {"loss": 0.7528, "grad_norm": 0.6675129532814026, "learning_rate": 0.0002, "epoch": 1.5536334976962252, "step": 9610}, {"loss": 0.7954, "grad_norm": 0.8173956274986267, "learning_rate": 0.0002, "epoch": 1.5552501818769704, "step": 9620}, {"loss": 0.7535, "grad_norm": 0.724947452545166, "learning_rate": 0.0002, "epoch": 1.5568668660577156, "step": 9630}, {"loss": 0.7738, "grad_norm": 0.6154758930206299, "learning_rate": 0.0002, "epoch": 1.5584835502384609, "step": 9640}, {"loss": 0.7568, "grad_norm": 0.6072008013725281, "learning_rate": 0.0002, "epoch": 1.560100234419206, "step": 9650}, {"loss": 0.7219, "grad_norm": 0.659010648727417, "learning_rate": 0.0002, "epoch": 1.5617169185999515, "step": 9660}, {"loss": 0.673, "grad_norm": 0.65857994556427, "learning_rate": 0.0002, "epoch": 1.5633336027806968, "step": 9670}, {"loss": 0.7156, "grad_norm": 0.5914267301559448, "learning_rate": 0.0002, "epoch": 1.5649502869614422, "step": 9680}, {"loss": 0.7414, "grad_norm": 0.6248020529747009, "learning_rate": 0.0002, "epoch": 1.5665669711421875, "step": 9690}, {"loss": 0.694, "grad_norm": 0.7147795557975769, "learning_rate": 0.0002, "epoch": 1.5681836553229327, "step": 9700}, {"loss": 0.7335, "grad_norm": 0.7076232433319092, "learning_rate": 0.0002, "epoch": 1.569800339503678, "step": 9710}, {"loss": 0.7413, "grad_norm": 0.6217400431632996, "learning_rate": 0.0002, "epoch": 1.5714170236844232, "step": 9720}, {"loss": 0.7296, "grad_norm": 0.6709911227226257, "learning_rate": 0.0002, "epoch": 1.5730337078651684, "step": 9730}, {"loss": 0.7306, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 1.5746503920459138, "step": 9740}, {"loss": 0.7242, "grad_norm": 0.6241145730018616, "learning_rate": 0.0002, "epoch": 1.576267076226659, "step": 9750}, {"loss": 0.7384, "grad_norm": 0.4960934817790985, "learning_rate": 0.0002, "epoch": 1.5778837604074045, "step": 9760}, {"loss": 0.725, "grad_norm": 0.6593309640884399, "learning_rate": 0.0002, "epoch": 1.5795004445881498, "step": 9770}, {"loss": 0.7531, "grad_norm": 0.5814042091369629, "learning_rate": 0.0002, "epoch": 1.581117128768895, "step": 9780}, {"loss": 0.7109, "grad_norm": 0.5936070680618286, "learning_rate": 0.0002, "epoch": 1.5827338129496402, "step": 9790}, {"loss": 0.7769, "grad_norm": 0.6454403400421143, "learning_rate": 0.0002, "epoch": 1.5843504971303854, "step": 9800}, {"loss": 0.7677, "grad_norm": 0.7612107992172241, "learning_rate": 0.0002, "epoch": 1.585967181311131, "step": 9810}, {"loss": 0.7649, "grad_norm": 0.6494482755661011, "learning_rate": 0.0002, "epoch": 1.5875838654918761, "step": 9820}, {"loss": 0.7569, "grad_norm": 0.7825694680213928, "learning_rate": 0.0002, "epoch": 1.5892005496726216, "step": 9830}, {"loss": 0.706, "grad_norm": 0.6757757663726807, "learning_rate": 0.0002, "epoch": 1.5908172338533668, "step": 9840}, {"loss": 0.7803, "grad_norm": 0.7105609178543091, "learning_rate": 0.0002, "epoch": 1.592433918034112, "step": 9850}, {"loss": 0.7925, "grad_norm": 0.7596991062164307, "learning_rate": 0.0002, "epoch": 1.5940506022148573, "step": 9860}, {"loss": 0.7108, "grad_norm": 0.5681525468826294, "learning_rate": 0.0002, "epoch": 1.5956672863956025, "step": 9870}, {"loss": 0.7811, "grad_norm": 0.6090980768203735, "learning_rate": 0.0002, "epoch": 1.5972839705763477, "step": 9880}, {"loss": 0.7339, "grad_norm": 0.6271613240242004, "learning_rate": 0.0002, "epoch": 1.5989006547570932, "step": 9890}, {"loss": 0.7419, "grad_norm": 0.7656369805335999, "learning_rate": 0.0002, "epoch": 1.6005173389378387, "step": 9900}, {"loss": 0.7336, "grad_norm": 0.7504446506500244, "learning_rate": 0.0002, "epoch": 1.6021340231185839, "step": 9910}, {"loss": 0.7479, "grad_norm": 0.659656286239624, "learning_rate": 0.0002, "epoch": 1.6037507072993291, "step": 9920}, {"loss": 0.7483, "grad_norm": 0.6006826162338257, "learning_rate": 0.0002, "epoch": 1.6053673914800743, "step": 9930}, {"loss": 0.732, "grad_norm": 0.7872757911682129, "learning_rate": 0.0002, "epoch": 1.6069840756608196, "step": 9940}, {"loss": 0.768, "grad_norm": 0.5545852780342102, "learning_rate": 0.0002, "epoch": 1.6086007598415648, "step": 9950}, {"loss": 0.8064, "grad_norm": 0.7429468631744385, "learning_rate": 0.0002, "epoch": 1.6102174440223103, "step": 9960}, {"loss": 0.714, "grad_norm": 0.6873556971549988, "learning_rate": 0.0002, "epoch": 1.6118341282030555, "step": 9970}, {"loss": 0.7324, "grad_norm": 0.5874287486076355, "learning_rate": 0.0002, "epoch": 1.613450812383801, "step": 9980}, {"loss": 0.7141, "grad_norm": 0.6039386987686157, "learning_rate": 0.0002, "epoch": 1.6150674965645462, "step": 9990}, {"loss": 0.6674, "grad_norm": 0.6233575940132141, "learning_rate": 0.0002, "epoch": 1.6166841807452914, "step": 10000}, {"loss": 0.7602, "grad_norm": 0.7676448225975037, "learning_rate": 0.0002, "epoch": 1.6183008649260366, "step": 10010}, {"loss": 0.7784, "grad_norm": 0.6565698385238647, "learning_rate": 0.0002, "epoch": 1.6199175491067819, "step": 10020}, {"loss": 0.7104, "grad_norm": 0.6787590384483337, "learning_rate": 0.0002, "epoch": 1.6215342332875273, "step": 10030}, {"loss": 0.7464, "grad_norm": 0.6137678027153015, "learning_rate": 0.0002, "epoch": 1.6231509174682726, "step": 10040}, {"loss": 0.7646, "grad_norm": 0.5236800312995911, "learning_rate": 0.0002, "epoch": 1.624767601649018, "step": 10050}, {"loss": 0.7437, "grad_norm": 0.7626367807388306, "learning_rate": 0.0002, "epoch": 1.6263842858297632, "step": 10060}, {"loss": 0.7273, "grad_norm": 0.5657260417938232, "learning_rate": 0.0002, "epoch": 1.6280009700105085, "step": 10070}, {"loss": 0.7354, "grad_norm": 0.4913991391658783, "learning_rate": 0.0002, "epoch": 1.6296176541912537, "step": 10080}, {"loss": 0.7596, "grad_norm": 0.7715556621551514, "learning_rate": 0.0002, "epoch": 1.631234338371999, "step": 10090}, {"loss": 0.7105, "grad_norm": 0.6509000062942505, "learning_rate": 0.0002, "epoch": 1.6328510225527442, "step": 10100}, {"loss": 0.7274, "grad_norm": 0.6215850114822388, "learning_rate": 0.0002, "epoch": 1.6344677067334896, "step": 10110}, {"loss": 0.7705, "grad_norm": 0.6956844329833984, "learning_rate": 0.0002, "epoch": 1.6360843909142349, "step": 10120}, {"loss": 0.7129, "grad_norm": 0.6111597418785095, "learning_rate": 0.0002, "epoch": 1.6377010750949803, "step": 10130}, {"loss": 0.6955, "grad_norm": 0.6518288850784302, "learning_rate": 0.0002, "epoch": 1.6393177592757255, "step": 10140}, {"loss": 0.731, "grad_norm": 0.6914522051811218, "learning_rate": 0.0002, "epoch": 1.6409344434564708, "step": 10150}, {"loss": 0.7295, "grad_norm": 0.63785719871521, "learning_rate": 0.0002, "epoch": 1.642551127637216, "step": 10160}, {"loss": 0.7355, "grad_norm": 0.6379287838935852, "learning_rate": 0.0002, "epoch": 1.6441678118179612, "step": 10170}, {"loss": 0.7359, "grad_norm": 0.6793403029441833, "learning_rate": 0.0002, "epoch": 1.6457844959987067, "step": 10180}, {"loss": 0.7402, "grad_norm": 0.6099132895469666, "learning_rate": 0.0002, "epoch": 1.647401180179452, "step": 10190}, {"loss": 0.7353, "grad_norm": 0.5869854092597961, "learning_rate": 0.0002, "epoch": 1.6490178643601974, "step": 10200}, {"loss": 0.8308, "grad_norm": 0.7716999053955078, "learning_rate": 0.0002, "epoch": 1.6506345485409426, "step": 10210}, {"loss": 0.7215, "grad_norm": 0.6854110360145569, "learning_rate": 0.0002, "epoch": 1.6522512327216878, "step": 10220}, {"loss": 0.782, "grad_norm": 0.6957170367240906, "learning_rate": 0.0002, "epoch": 1.653867916902433, "step": 10230}, {"loss": 0.7282, "grad_norm": 0.6932903528213501, "learning_rate": 0.0002, "epoch": 1.6554846010831783, "step": 10240}, {"loss": 0.7478, "grad_norm": 0.7713165283203125, "learning_rate": 0.0002, "epoch": 1.6571012852639235, "step": 10250}, {"loss": 0.7099, "grad_norm": 0.7455793619155884, "learning_rate": 0.0002, "epoch": 1.658717969444669, "step": 10260}, {"loss": 0.7524, "grad_norm": 0.5464168190956116, "learning_rate": 0.0002, "epoch": 1.6603346536254144, "step": 10270}, {"loss": 0.7328, "grad_norm": 0.6782926321029663, "learning_rate": 0.0002, "epoch": 1.6619513378061597, "step": 10280}, {"loss": 0.7801, "grad_norm": 0.7962649464607239, "learning_rate": 0.0002, "epoch": 1.663568021986905, "step": 10290}, {"loss": 0.7142, "grad_norm": 0.6814526319503784, "learning_rate": 0.0002, "epoch": 1.6651847061676501, "step": 10300}, {"loss": 0.7285, "grad_norm": 0.656895101070404, "learning_rate": 0.0002, "epoch": 1.6668013903483954, "step": 10310}, {"loss": 0.7358, "grad_norm": 0.6085672378540039, "learning_rate": 0.0002, "epoch": 1.6684180745291406, "step": 10320}, {"loss": 0.7074, "grad_norm": 0.585508406162262, "learning_rate": 0.0002, "epoch": 1.670034758709886, "step": 10330}, {"loss": 0.7604, "grad_norm": 0.6930184364318848, "learning_rate": 0.0002, "epoch": 1.6716514428906313, "step": 10340}, {"loss": 0.7169, "grad_norm": 0.575663149356842, "learning_rate": 0.0002, "epoch": 1.6732681270713767, "step": 10350}, {"loss": 0.7198, "grad_norm": 0.582502543926239, "learning_rate": 0.0002, "epoch": 1.674884811252122, "step": 10360}, {"loss": 0.7793, "grad_norm": 0.5668916702270508, "learning_rate": 0.0002, "epoch": 1.6765014954328672, "step": 10370}, {"loss": 0.7478, "grad_norm": 0.6070065498352051, "learning_rate": 0.0002, "epoch": 1.6781181796136124, "step": 10380}, {"loss": 0.7939, "grad_norm": 0.6141316294670105, "learning_rate": 0.0002, "epoch": 1.6797348637943577, "step": 10390}, {"loss": 0.7573, "grad_norm": 0.8359124064445496, "learning_rate": 0.0002, "epoch": 1.6813515479751031, "step": 10400}, {"loss": 0.7488, "grad_norm": 0.5378185510635376, "learning_rate": 0.0002, "epoch": 1.6829682321558483, "step": 10410}, {"loss": 0.7588, "grad_norm": 0.6959536075592041, "learning_rate": 0.0002, "epoch": 1.6845849163365938, "step": 10420}, {"loss": 0.7872, "grad_norm": 0.6514357328414917, "learning_rate": 0.0002, "epoch": 1.686201600517339, "step": 10430}, {"loss": 0.725, "grad_norm": 0.7706646919250488, "learning_rate": 0.0002, "epoch": 1.6878182846980843, "step": 10440}, {"loss": 0.7673, "grad_norm": 0.6183337569236755, "learning_rate": 0.0002, "epoch": 1.6894349688788295, "step": 10450}, {"loss": 0.7566, "grad_norm": 0.6123278141021729, "learning_rate": 0.0002, "epoch": 1.6910516530595747, "step": 10460}, {"loss": 0.7169, "grad_norm": 0.6894851326942444, "learning_rate": 0.0002, "epoch": 1.69266833724032, "step": 10470}, {"loss": 0.7435, "grad_norm": 0.7497312426567078, "learning_rate": 0.0002, "epoch": 1.6942850214210654, "step": 10480}, {"loss": 0.7544, "grad_norm": 0.5968214273452759, "learning_rate": 0.0002, "epoch": 1.6959017056018106, "step": 10490}, {"loss": 0.6793, "grad_norm": 0.6747927069664001, "learning_rate": 0.0002, "epoch": 1.697518389782556, "step": 10500}, {"loss": 0.7415, "grad_norm": 0.5708310008049011, "learning_rate": 0.0002, "epoch": 1.6991350739633013, "step": 10510}, {"loss": 0.7385, "grad_norm": 0.606526792049408, "learning_rate": 0.0002, "epoch": 1.7007517581440466, "step": 10520}, {"loss": 0.7204, "grad_norm": 0.662011981010437, "learning_rate": 0.0002, "epoch": 1.7023684423247918, "step": 10530}, {"loss": 0.7999, "grad_norm": 0.7583045363426208, "learning_rate": 0.0002, "epoch": 1.703985126505537, "step": 10540}, {"loss": 0.7563, "grad_norm": 0.721632182598114, "learning_rate": 0.0002, "epoch": 1.7056018106862825, "step": 10550}, {"loss": 0.7407, "grad_norm": 0.6107715368270874, "learning_rate": 0.0002, "epoch": 1.7072184948670277, "step": 10560}, {"loss": 0.7519, "grad_norm": 0.6652471423149109, "learning_rate": 0.0002, "epoch": 1.7088351790477732, "step": 10570}, {"loss": 0.7767, "grad_norm": 0.6308087110519409, "learning_rate": 0.0002, "epoch": 1.7104518632285184, "step": 10580}, {"loss": 0.7659, "grad_norm": 0.5464386940002441, "learning_rate": 0.0002, "epoch": 1.7120685474092636, "step": 10590}, {"loss": 0.7063, "grad_norm": 0.6558911204338074, "learning_rate": 0.0002, "epoch": 1.7136852315900089, "step": 10600}, {"loss": 0.7126, "grad_norm": 0.5665024518966675, "learning_rate": 0.0002, "epoch": 1.715301915770754, "step": 10610}, {"loss": 0.6958, "grad_norm": 0.7888094186782837, "learning_rate": 0.0002, "epoch": 1.7169185999514993, "step": 10620}, {"loss": 0.7785, "grad_norm": 0.7084909081459045, "learning_rate": 0.0002, "epoch": 1.7185352841322448, "step": 10630}, {"loss": 0.7557, "grad_norm": 0.7982324361801147, "learning_rate": 0.0002, "epoch": 1.7201519683129902, "step": 10640}, {"loss": 0.7345, "grad_norm": 0.6418732404708862, "learning_rate": 0.0002, "epoch": 1.7217686524937355, "step": 10650}, {"loss": 0.7734, "grad_norm": 0.7636681795120239, "learning_rate": 0.0002, "epoch": 1.7233853366744807, "step": 10660}, {"loss": 0.7541, "grad_norm": 0.5646875500679016, "learning_rate": 0.0002, "epoch": 1.725002020855226, "step": 10670}, {"loss": 0.7642, "grad_norm": 0.5231260657310486, "learning_rate": 0.0002, "epoch": 1.7266187050359711, "step": 10680}, {"loss": 0.7846, "grad_norm": 0.7635011672973633, "learning_rate": 0.0002, "epoch": 1.7282353892167164, "step": 10690}, {"loss": 0.7471, "grad_norm": 0.7518259286880493, "learning_rate": 0.0002, "epoch": 1.7298520733974618, "step": 10700}, {"loss": 0.751, "grad_norm": 0.7295602560043335, "learning_rate": 0.0002, "epoch": 1.731468757578207, "step": 10710}, {"loss": 0.731, "grad_norm": 0.6984632015228271, "learning_rate": 0.0002, "epoch": 1.7330854417589525, "step": 10720}, {"loss": 0.7921, "grad_norm": 0.6198219060897827, "learning_rate": 0.0002, "epoch": 1.7347021259396977, "step": 10730}, {"loss": 0.7642, "grad_norm": 0.6957576274871826, "learning_rate": 0.0002, "epoch": 1.736318810120443, "step": 10740}, {"loss": 0.7917, "grad_norm": 0.6430263519287109, "learning_rate": 0.0002, "epoch": 1.7379354943011882, "step": 10750}, {"loss": 0.7156, "grad_norm": 0.6134995222091675, "learning_rate": 0.0002, "epoch": 1.7395521784819334, "step": 10760}, {"loss": 0.7584, "grad_norm": 0.7209452986717224, "learning_rate": 0.0002, "epoch": 1.741168862662679, "step": 10770}, {"loss": 0.7528, "grad_norm": 0.6735447645187378, "learning_rate": 0.0002, "epoch": 1.7427855468434241, "step": 10780}, {"loss": 0.756, "grad_norm": 0.5605693459510803, "learning_rate": 0.0002, "epoch": 1.7444022310241696, "step": 10790}, {"loss": 0.7759, "grad_norm": 0.6882363557815552, "learning_rate": 0.0002, "epoch": 1.7460189152049148, "step": 10800}, {"loss": 0.7544, "grad_norm": 0.6386259198188782, "learning_rate": 0.0002, "epoch": 1.74763559938566, "step": 10810}, {"loss": 0.7697, "grad_norm": 0.6529015302658081, "learning_rate": 0.0002, "epoch": 1.7492522835664053, "step": 10820}, {"loss": 0.7219, "grad_norm": 0.5664082765579224, "learning_rate": 0.0002, "epoch": 1.7508689677471505, "step": 10830}, {"loss": 0.7586, "grad_norm": 0.7532684206962585, "learning_rate": 0.0002, "epoch": 1.7524856519278957, "step": 10840}, {"loss": 0.6919, "grad_norm": 0.77171391248703, "learning_rate": 0.0002, "epoch": 1.7541023361086412, "step": 10850}, {"loss": 0.785, "grad_norm": 0.7255431413650513, "learning_rate": 0.0002, "epoch": 1.7557190202893864, "step": 10860}, {"loss": 0.7458, "grad_norm": 0.763083279132843, "learning_rate": 0.0002, "epoch": 1.7573357044701319, "step": 10870}, {"loss": 0.7846, "grad_norm": 0.6042402982711792, "learning_rate": 0.0002, "epoch": 1.758952388650877, "step": 10880}, {"loss": 0.7027, "grad_norm": 0.7642518281936646, "learning_rate": 0.0002, "epoch": 1.7605690728316223, "step": 10890}, {"loss": 0.746, "grad_norm": 0.6347904801368713, "learning_rate": 0.0002, "epoch": 1.7621857570123676, "step": 10900}, {"loss": 0.7458, "grad_norm": 0.5371627807617188, "learning_rate": 0.0002, "epoch": 1.7638024411931128, "step": 10910}, {"loss": 0.7466, "grad_norm": 0.6840225458145142, "learning_rate": 0.0002, "epoch": 1.7654191253738583, "step": 10920}, {"loss": 0.725, "grad_norm": 0.5288469195365906, "learning_rate": 0.0002, "epoch": 1.7670358095546035, "step": 10930}, {"loss": 0.7863, "grad_norm": 0.69020676612854, "learning_rate": 0.0002, "epoch": 1.768652493735349, "step": 10940}, {"loss": 0.7468, "grad_norm": 0.5943242311477661, "learning_rate": 0.0002, "epoch": 1.7702691779160942, "step": 10950}, {"loss": 0.7244, "grad_norm": 0.5616418123245239, "learning_rate": 0.0002, "epoch": 1.7718858620968394, "step": 10960}, {"loss": 0.7137, "grad_norm": 0.7209470868110657, "learning_rate": 0.0002, "epoch": 1.7735025462775846, "step": 10970}, {"loss": 0.7459, "grad_norm": 0.6657957434654236, "learning_rate": 0.0002, "epoch": 1.7751192304583299, "step": 10980}, {"loss": 0.7076, "grad_norm": 0.6469064950942993, "learning_rate": 0.0002, "epoch": 1.776735914639075, "step": 10990}, {"loss": 0.7321, "grad_norm": 0.6615678071975708, "learning_rate": 0.0002, "epoch": 1.7783525988198206, "step": 11000}, {"loss": 0.747, "grad_norm": 0.6722439527511597, "learning_rate": 0.0002, "epoch": 1.779969283000566, "step": 11010}, {"loss": 0.7302, "grad_norm": 0.634136974811554, "learning_rate": 0.0002, "epoch": 1.7815859671813112, "step": 11020}, {"loss": 0.8105, "grad_norm": 0.6024377346038818, "learning_rate": 0.0002, "epoch": 1.7832026513620565, "step": 11030}, {"loss": 0.7855, "grad_norm": 0.6909403800964355, "learning_rate": 0.0002, "epoch": 1.7848193355428017, "step": 11040}, {"loss": 0.7471, "grad_norm": 0.7148767709732056, "learning_rate": 0.0002, "epoch": 1.786436019723547, "step": 11050}, {"loss": 0.7145, "grad_norm": 0.7442979216575623, "learning_rate": 0.0002, "epoch": 1.7880527039042922, "step": 11060}, {"loss": 0.7215, "grad_norm": 0.6830431818962097, "learning_rate": 0.0002, "epoch": 1.7896693880850376, "step": 11070}, {"loss": 0.7625, "grad_norm": 0.9172667264938354, "learning_rate": 0.0002, "epoch": 1.7912860722657828, "step": 11080}, {"loss": 0.76, "grad_norm": 0.6799490451812744, "learning_rate": 0.0002, "epoch": 1.7929027564465283, "step": 11090}, {"loss": 0.7716, "grad_norm": 0.7617024779319763, "learning_rate": 0.0002, "epoch": 1.7945194406272735, "step": 11100}, {"loss": 0.7586, "grad_norm": 0.7701810002326965, "learning_rate": 0.0002, "epoch": 1.7961361248080188, "step": 11110}, {"loss": 0.7843, "grad_norm": 0.7454385757446289, "learning_rate": 0.0002, "epoch": 1.797752808988764, "step": 11120}, {"loss": 0.7873, "grad_norm": 0.6121436953544617, "learning_rate": 0.0002, "epoch": 1.7993694931695092, "step": 11130}, {"loss": 0.7305, "grad_norm": 0.6237571835517883, "learning_rate": 0.0002, "epoch": 1.8009861773502547, "step": 11140}, {"loss": 0.6827, "grad_norm": 0.6818515658378601, "learning_rate": 0.0002, "epoch": 1.802602861531, "step": 11150}, {"loss": 0.6876, "grad_norm": 0.7768308520317078, "learning_rate": 0.0002, "epoch": 1.8042195457117454, "step": 11160}, {"loss": 0.7533, "grad_norm": 0.6875537633895874, "learning_rate": 0.0002, "epoch": 1.8058362298924906, "step": 11170}, {"loss": 0.761, "grad_norm": 0.7950584888458252, "learning_rate": 0.0002, "epoch": 1.8074529140732358, "step": 11180}, {"loss": 0.7623, "grad_norm": 0.8210248351097107, "learning_rate": 0.0002, "epoch": 1.809069598253981, "step": 11190}, {"loss": 0.7556, "grad_norm": 0.6674110889434814, "learning_rate": 0.0002, "epoch": 1.8106862824347263, "step": 11200}, {"loss": 0.7663, "grad_norm": 0.6261674761772156, "learning_rate": 0.0002, "epoch": 1.8123029666154715, "step": 11210}, {"loss": 0.7122, "grad_norm": 0.6484741568565369, "learning_rate": 0.0002, "epoch": 1.813919650796217, "step": 11220}, {"loss": 0.7718, "grad_norm": 0.6231244206428528, "learning_rate": 0.0002, "epoch": 1.8155363349769622, "step": 11230}, {"loss": 0.7152, "grad_norm": 0.7243146896362305, "learning_rate": 0.0002, "epoch": 1.8171530191577077, "step": 11240}, {"loss": 0.7448, "grad_norm": 0.6776193380355835, "learning_rate": 0.0002, "epoch": 1.818769703338453, "step": 11250}, {"loss": 0.7317, "grad_norm": 0.5973618030548096, "learning_rate": 0.0002, "epoch": 1.8203863875191981, "step": 11260}, {"loss": 0.7961, "grad_norm": 0.6451361179351807, "learning_rate": 0.0002, "epoch": 1.8220030716999434, "step": 11270}, {"loss": 0.7611, "grad_norm": 0.5963068008422852, "learning_rate": 0.0002, "epoch": 1.8236197558806886, "step": 11280}, {"loss": 0.7466, "grad_norm": 0.536902129650116, "learning_rate": 0.0002, "epoch": 1.825236440061434, "step": 11290}, {"loss": 0.708, "grad_norm": 0.6993787288665771, "learning_rate": 0.0002, "epoch": 1.8268531242421793, "step": 11300}, {"loss": 0.7153, "grad_norm": 0.6135255098342896, "learning_rate": 0.0002, "epoch": 1.8284698084229247, "step": 11310}, {"loss": 0.7423, "grad_norm": 0.6057423949241638, "learning_rate": 0.0002, "epoch": 1.83008649260367, "step": 11320}, {"loss": 0.735, "grad_norm": 0.6598812341690063, "learning_rate": 0.0002, "epoch": 1.8317031767844152, "step": 11330}, {"loss": 0.7278, "grad_norm": 0.6075948476791382, "learning_rate": 0.0002, "epoch": 1.8333198609651604, "step": 11340}, {"loss": 0.7846, "grad_norm": 0.7065447568893433, "learning_rate": 0.0002, "epoch": 1.8349365451459057, "step": 11350}, {"loss": 0.7365, "grad_norm": 0.680526614189148, "learning_rate": 0.0002, "epoch": 1.8365532293266509, "step": 11360}, {"loss": 0.7152, "grad_norm": 0.6356695294380188, "learning_rate": 0.0002, "epoch": 1.8381699135073963, "step": 11370}, {"loss": 0.721, "grad_norm": 0.6399052143096924, "learning_rate": 0.0002, "epoch": 1.8397865976881416, "step": 11380}, {"loss": 0.7618, "grad_norm": 0.6125704050064087, "learning_rate": 0.0002, "epoch": 1.841403281868887, "step": 11390}, {"loss": 0.755, "grad_norm": 0.7124643325805664, "learning_rate": 0.0002, "epoch": 1.8430199660496323, "step": 11400}, {"loss": 0.7972, "grad_norm": 0.6099604964256287, "learning_rate": 0.0002, "epoch": 1.8446366502303775, "step": 11410}, {"loss": 0.7187, "grad_norm": 0.7338208556175232, "learning_rate": 0.0002, "epoch": 1.8462533344111227, "step": 11420}, {"loss": 0.7007, "grad_norm": 0.7534668445587158, "learning_rate": 0.0002, "epoch": 1.847870018591868, "step": 11430}, {"loss": 0.7464, "grad_norm": 0.6135470271110535, "learning_rate": 0.0002, "epoch": 1.8494867027726134, "step": 11440}, {"loss": 0.7955, "grad_norm": 0.6229309439659119, "learning_rate": 0.0002, "epoch": 1.8511033869533586, "step": 11450}, {"loss": 0.7594, "grad_norm": 0.706423282623291, "learning_rate": 0.0002, "epoch": 1.852720071134104, "step": 11460}, {"loss": 0.7411, "grad_norm": 0.5460049510002136, "learning_rate": 0.0002, "epoch": 1.8543367553148493, "step": 11470}, {"loss": 0.7416, "grad_norm": 0.6616711020469666, "learning_rate": 0.0002, "epoch": 1.8559534394955945, "step": 11480}, {"loss": 0.729, "grad_norm": 0.6372783184051514, "learning_rate": 0.0002, "epoch": 1.8575701236763398, "step": 11490}, {"loss": 0.7333, "grad_norm": 0.7162668108940125, "learning_rate": 0.0002, "epoch": 1.859186807857085, "step": 11500}, {"loss": 0.7747, "grad_norm": 0.6605209708213806, "learning_rate": 0.0002, "epoch": 1.8608034920378305, "step": 11510}, {"loss": 0.7258, "grad_norm": 0.6933956742286682, "learning_rate": 0.0002, "epoch": 1.8624201762185757, "step": 11520}, {"loss": 0.7243, "grad_norm": 0.6582090854644775, "learning_rate": 0.0002, "epoch": 1.8640368603993211, "step": 11530}, {"loss": 0.7313, "grad_norm": 0.6416500806808472, "learning_rate": 0.0002, "epoch": 1.8656535445800664, "step": 11540}, {"loss": 0.7372, "grad_norm": 0.5434312224388123, "learning_rate": 0.0002, "epoch": 1.8672702287608116, "step": 11550}, {"loss": 0.7635, "grad_norm": 0.6827567219734192, "learning_rate": 0.0002, "epoch": 1.8688869129415568, "step": 11560}, {"loss": 0.7137, "grad_norm": 0.7354370951652527, "learning_rate": 0.0002, "epoch": 1.870503597122302, "step": 11570}, {"loss": 0.7526, "grad_norm": 0.590372622013092, "learning_rate": 0.0002, "epoch": 1.8721202813030473, "step": 11580}, {"loss": 0.731, "grad_norm": 0.853183925151825, "learning_rate": 0.0002, "epoch": 1.8737369654837928, "step": 11590}, {"loss": 0.7487, "grad_norm": 0.822678804397583, "learning_rate": 0.0002, "epoch": 1.875353649664538, "step": 11600}, {"loss": 0.7427, "grad_norm": 0.6591550707817078, "learning_rate": 0.0002, "epoch": 1.8769703338452834, "step": 11610}, {"loss": 0.7054, "grad_norm": 0.7475301623344421, "learning_rate": 0.0002, "epoch": 1.8785870180260287, "step": 11620}, {"loss": 0.811, "grad_norm": 0.6390765309333801, "learning_rate": 0.0002, "epoch": 1.880203702206774, "step": 11630}, {"loss": 0.7531, "grad_norm": 0.6589758992195129, "learning_rate": 0.0002, "epoch": 1.8818203863875191, "step": 11640}, {"loss": 0.7475, "grad_norm": 0.6765508651733398, "learning_rate": 0.0002, "epoch": 1.8834370705682644, "step": 11650}, {"loss": 0.738, "grad_norm": 0.6527857780456543, "learning_rate": 0.0002, "epoch": 1.8850537547490098, "step": 11660}, {"loss": 0.7504, "grad_norm": 0.6642923951148987, "learning_rate": 0.0002, "epoch": 1.886670438929755, "step": 11670}, {"loss": 0.7701, "grad_norm": 0.6945584416389465, "learning_rate": 0.0002, "epoch": 1.8882871231105005, "step": 11680}, {"loss": 0.7711, "grad_norm": 0.694018542766571, "learning_rate": 0.0002, "epoch": 1.8899038072912457, "step": 11690}, {"loss": 0.7195, "grad_norm": 0.7237417101860046, "learning_rate": 0.0002, "epoch": 1.891520491471991, "step": 11700}, {"loss": 0.7491, "grad_norm": 0.7401309609413147, "learning_rate": 0.0002, "epoch": 1.8931371756527362, "step": 11710}, {"loss": 0.805, "grad_norm": 0.6537784337997437, "learning_rate": 0.0002, "epoch": 1.8947538598334814, "step": 11720}, {"loss": 0.793, "grad_norm": 0.7398539185523987, "learning_rate": 0.0002, "epoch": 1.8963705440142267, "step": 11730}, {"loss": 0.7561, "grad_norm": 0.6696075797080994, "learning_rate": 0.0002, "epoch": 1.8979872281949721, "step": 11740}, {"loss": 0.7353, "grad_norm": 0.6014142036437988, "learning_rate": 0.0002, "epoch": 1.8996039123757174, "step": 11750}, {"loss": 0.7714, "grad_norm": 0.7023524641990662, "learning_rate": 0.0002, "epoch": 1.9012205965564628, "step": 11760}, {"loss": 0.7088, "grad_norm": 0.739973783493042, "learning_rate": 0.0002, "epoch": 1.902837280737208, "step": 11770}, {"loss": 0.7848, "grad_norm": 0.5576770901679993, "learning_rate": 0.0002, "epoch": 1.9044539649179533, "step": 11780}, {"loss": 0.7483, "grad_norm": 0.6907393932342529, "learning_rate": 0.0002, "epoch": 1.9060706490986985, "step": 11790}, {"loss": 0.7827, "grad_norm": 0.6934581995010376, "learning_rate": 0.0002, "epoch": 1.9076873332794437, "step": 11800}, {"loss": 0.7199, "grad_norm": 0.591774582862854, "learning_rate": 0.0002, "epoch": 1.9093040174601892, "step": 11810}, {"loss": 0.7333, "grad_norm": 0.6249791383743286, "learning_rate": 0.0002, "epoch": 1.9109207016409344, "step": 11820}, {"loss": 0.7581, "grad_norm": 0.6755744218826294, "learning_rate": 0.0002, "epoch": 1.9125373858216799, "step": 11830}, {"loss": 0.696, "grad_norm": 0.7286285161972046, "learning_rate": 0.0002, "epoch": 1.914154070002425, "step": 11840}, {"loss": 0.7509, "grad_norm": 0.7867850065231323, "learning_rate": 0.0002, "epoch": 1.9157707541831703, "step": 11850}, {"loss": 0.735, "grad_norm": 0.6283972859382629, "learning_rate": 0.0002, "epoch": 1.9173874383639156, "step": 11860}, {"loss": 0.7296, "grad_norm": 0.605823814868927, "learning_rate": 0.0002, "epoch": 1.9190041225446608, "step": 11870}, {"loss": 0.6598, "grad_norm": 0.5927976965904236, "learning_rate": 0.0002, "epoch": 1.920620806725406, "step": 11880}, {"loss": 0.7649, "grad_norm": 0.5974002480506897, "learning_rate": 0.0002, "epoch": 1.9222374909061515, "step": 11890}, {"loss": 0.7843, "grad_norm": 0.7091866135597229, "learning_rate": 0.0002, "epoch": 1.923854175086897, "step": 11900}, {"loss": 0.775, "grad_norm": 0.72496497631073, "learning_rate": 0.0002, "epoch": 1.9254708592676422, "step": 11910}, {"loss": 0.7153, "grad_norm": 0.6131896376609802, "learning_rate": 0.0002, "epoch": 1.9270875434483874, "step": 11920}, {"loss": 0.7228, "grad_norm": 0.6556436419487, "learning_rate": 0.0002, "epoch": 1.9287042276291326, "step": 11930}, {"loss": 0.7319, "grad_norm": 0.622932493686676, "learning_rate": 0.0002, "epoch": 1.9303209118098779, "step": 11940}, {"loss": 0.7592, "grad_norm": 0.6618631482124329, "learning_rate": 0.0002, "epoch": 1.931937595990623, "step": 11950}, {"loss": 0.8332, "grad_norm": 0.630966305732727, "learning_rate": 0.0002, "epoch": 1.9335542801713685, "step": 11960}, {"loss": 0.6854, "grad_norm": 0.6336734890937805, "learning_rate": 0.0002, "epoch": 1.9351709643521138, "step": 11970}, {"loss": 0.7433, "grad_norm": 0.655403196811676, "learning_rate": 0.0002, "epoch": 1.9367876485328592, "step": 11980}, {"loss": 0.7282, "grad_norm": 0.5640574097633362, "learning_rate": 0.0002, "epoch": 1.9384043327136045, "step": 11990}, {"loss": 0.7289, "grad_norm": 0.6322951316833496, "learning_rate": 0.0002, "epoch": 1.9400210168943497, "step": 12000}, {"loss": 0.7627, "grad_norm": 0.615703821182251, "learning_rate": 0.0002, "epoch": 1.941637701075095, "step": 12010}, {"loss": 0.786, "grad_norm": 0.6487536430358887, "learning_rate": 0.0002, "epoch": 1.9432543852558402, "step": 12020}, {"loss": 0.7435, "grad_norm": 0.9209630489349365, "learning_rate": 0.0002, "epoch": 1.9448710694365856, "step": 12030}, {"loss": 0.7274, "grad_norm": 0.67485511302948, "learning_rate": 0.0002, "epoch": 1.9464877536173308, "step": 12040}, {"loss": 0.7551, "grad_norm": 0.6831230521202087, "learning_rate": 0.0002, "epoch": 1.9481044377980763, "step": 12050}, {"loss": 0.7546, "grad_norm": 0.6578302383422852, "learning_rate": 0.0002, "epoch": 1.9497211219788215, "step": 12060}, {"loss": 0.6989, "grad_norm": 0.9975938200950623, "learning_rate": 0.0002, "epoch": 1.9513378061595668, "step": 12070}, {"loss": 0.7952, "grad_norm": 0.6637365221977234, "learning_rate": 0.0002, "epoch": 1.952954490340312, "step": 12080}, {"loss": 0.7482, "grad_norm": 0.605707049369812, "learning_rate": 0.0002, "epoch": 1.9545711745210572, "step": 12090}, {"loss": 0.7768, "grad_norm": 0.6584440469741821, "learning_rate": 0.0002, "epoch": 1.9561878587018025, "step": 12100}, {"loss": 0.7187, "grad_norm": 0.6070835590362549, "learning_rate": 0.0002, "epoch": 1.957804542882548, "step": 12110}, {"loss": 0.7491, "grad_norm": 0.7862601280212402, "learning_rate": 0.0002, "epoch": 1.9594212270632931, "step": 12120}, {"loss": 0.7972, "grad_norm": 0.8175255060195923, "learning_rate": 0.0002, "epoch": 1.9610379112440386, "step": 12130}, {"loss": 0.7242, "grad_norm": 0.5648472905158997, "learning_rate": 0.0002, "epoch": 1.9626545954247838, "step": 12140}, {"loss": 0.7321, "grad_norm": 0.6591973304748535, "learning_rate": 0.0002, "epoch": 1.964271279605529, "step": 12150}, {"loss": 0.739, "grad_norm": 0.5960676074028015, "learning_rate": 0.0002, "epoch": 1.9658879637862743, "step": 12160}, {"loss": 0.7254, "grad_norm": 0.7272544503211975, "learning_rate": 0.0002, "epoch": 1.9675046479670195, "step": 12170}, {"loss": 0.7376, "grad_norm": 0.7176699042320251, "learning_rate": 0.0002, "epoch": 1.969121332147765, "step": 12180}, {"loss": 0.7525, "grad_norm": 0.6927123665809631, "learning_rate": 0.0002, "epoch": 1.9707380163285102, "step": 12190}, {"loss": 0.7318, "grad_norm": 0.5536034107208252, "learning_rate": 0.0002, "epoch": 1.9723547005092557, "step": 12200}, {"loss": 0.7737, "grad_norm": 0.8348390460014343, "learning_rate": 0.0002, "epoch": 1.9739713846900009, "step": 12210}, {"loss": 0.7494, "grad_norm": 0.6591181755065918, "learning_rate": 0.0002, "epoch": 1.9755880688707461, "step": 12220}, {"loss": 0.763, "grad_norm": 1.0624109506607056, "learning_rate": 0.0002, "epoch": 1.9772047530514913, "step": 12230}, {"loss": 0.7541, "grad_norm": 0.9265586137771606, "learning_rate": 0.0002, "epoch": 1.9788214372322366, "step": 12240}, {"loss": 0.7533, "grad_norm": 0.5998196005821228, "learning_rate": 0.0002, "epoch": 1.9804381214129818, "step": 12250}, {"loss": 0.7225, "grad_norm": 0.6960851550102234, "learning_rate": 0.0002, "epoch": 1.9820548055937273, "step": 12260}, {"loss": 0.7398, "grad_norm": 0.7674502730369568, "learning_rate": 0.0002, "epoch": 1.9836714897744727, "step": 12270}, {"loss": 0.7185, "grad_norm": 0.6407275795936584, "learning_rate": 0.0002, "epoch": 1.985288173955218, "step": 12280}, {"loss": 0.7382, "grad_norm": 0.6673079133033752, "learning_rate": 0.0002, "epoch": 1.9869048581359632, "step": 12290}, {"loss": 0.7326, "grad_norm": 0.6989844441413879, "learning_rate": 0.0002, "epoch": 1.9885215423167084, "step": 12300}, {"loss": 0.7559, "grad_norm": 0.7564442157745361, "learning_rate": 0.0002, "epoch": 1.9901382264974536, "step": 12310}, {"loss": 0.7719, "grad_norm": 0.6385478973388672, "learning_rate": 0.0002, "epoch": 1.9917549106781989, "step": 12320}, {"loss": 0.7369, "grad_norm": 0.7193717956542969, "learning_rate": 0.0002, "epoch": 1.9933715948589443, "step": 12330}, {"loss": 0.7583, "grad_norm": 0.7987112402915955, "learning_rate": 0.0002, "epoch": 1.9949882790396896, "step": 12340}, {"loss": 0.7793, "grad_norm": 0.7260826826095581, "learning_rate": 0.0002, "epoch": 1.996604963220435, "step": 12350}, {"loss": 0.7505, "grad_norm": 0.7968255281448364, "learning_rate": 0.0002, "epoch": 1.9982216474011802, "step": 12360}, {"loss": 0.717, "grad_norm": 0.6893062591552734, "learning_rate": 0.0002, "epoch": 1.9998383315819255, "step": 12370}, {"eval_loss": 1.1044032573699951, "eval_runtime": 122.1508, "eval_samples_per_second": 6.001, "eval_steps_per_second": 0.753, "epoch": 2.0, "step": 12371}, {"loss": 0.6604, "grad_norm": 0.7775409817695618, "learning_rate": 0.0002, "epoch": 2.0014550157626707, "step": 12380}, {"loss": 0.6845, "grad_norm": 0.76218581199646, "learning_rate": 0.0002, "epoch": 2.003071699943416, "step": 12390}, {"loss": 0.6909, "grad_norm": 0.5677764415740967, "learning_rate": 0.0002, "epoch": 2.004688384124161, "step": 12400}, {"loss": 0.6584, "grad_norm": 0.808442234992981, "learning_rate": 0.0002, "epoch": 2.006305068304907, "step": 12410}, {"loss": 0.659, "grad_norm": 0.7144765257835388, "learning_rate": 0.0002, "epoch": 2.007921752485652, "step": 12420}, {"loss": 0.6666, "grad_norm": 0.6914031505584717, "learning_rate": 0.0002, "epoch": 2.0095384366663973, "step": 12430}, {"loss": 0.6596, "grad_norm": 0.7581454515457153, "learning_rate": 0.0002, "epoch": 2.0111551208471425, "step": 12440}, {"loss": 0.6785, "grad_norm": 0.8388504981994629, "learning_rate": 0.0002, "epoch": 2.0127718050278878, "step": 12450}, {"loss": 0.6942, "grad_norm": 0.6716406941413879, "learning_rate": 0.0002, "epoch": 2.014388489208633, "step": 12460}, {"loss": 0.6441, "grad_norm": 0.898902416229248, "learning_rate": 0.0002, "epoch": 2.0160051733893782, "step": 12470}, {"loss": 0.6655, "grad_norm": 0.6432679891586304, "learning_rate": 0.0002, "epoch": 2.0176218575701235, "step": 12480}, {"loss": 0.6521, "grad_norm": 0.8021109104156494, "learning_rate": 0.0002, "epoch": 2.019238541750869, "step": 12490}, {"loss": 0.6581, "grad_norm": 0.7039216756820679, "learning_rate": 0.0002, "epoch": 2.0208552259316144, "step": 12500}, {"loss": 0.6521, "grad_norm": 0.646531879901886, "learning_rate": 0.0002, "epoch": 2.0224719101123596, "step": 12510}, {"loss": 0.6302, "grad_norm": 0.783704400062561, "learning_rate": 0.0002, "epoch": 2.024088594293105, "step": 12520}, {"loss": 0.6288, "grad_norm": 0.8805046677589417, "learning_rate": 0.0002, "epoch": 2.02570527847385, "step": 12530}, {"loss": 0.6288, "grad_norm": 0.7289270758628845, "learning_rate": 0.0002, "epoch": 2.0273219626545953, "step": 12540}, {"loss": 0.6663, "grad_norm": 0.71653151512146, "learning_rate": 0.0002, "epoch": 2.0289386468353405, "step": 12550}, {"loss": 0.625, "grad_norm": 0.73281329870224, "learning_rate": 0.0002, "epoch": 2.030555331016086, "step": 12560}, {"loss": 0.6448, "grad_norm": 0.6657090187072754, "learning_rate": 0.0002, "epoch": 2.0321720151968314, "step": 12570}, {"loss": 0.6983, "grad_norm": 0.8241133093833923, "learning_rate": 0.0002, "epoch": 2.0337886993775767, "step": 12580}, {"loss": 0.6488, "grad_norm": 0.5834135413169861, "learning_rate": 0.0002, "epoch": 2.035405383558322, "step": 12590}, {"loss": 0.6188, "grad_norm": 0.84502112865448, "learning_rate": 0.0002, "epoch": 2.037022067739067, "step": 12600}, {"loss": 0.6349, "grad_norm": 0.8952481746673584, "learning_rate": 0.0002, "epoch": 2.0386387519198124, "step": 12610}, {"loss": 0.6923, "grad_norm": 0.7801461815834045, "learning_rate": 0.0002, "epoch": 2.0402554361005576, "step": 12620}, {"loss": 0.6176, "grad_norm": 0.6788367033004761, "learning_rate": 0.0002, "epoch": 2.041872120281303, "step": 12630}, {"loss": 0.6162, "grad_norm": 0.7241756319999695, "learning_rate": 0.0002, "epoch": 2.0434888044620485, "step": 12640}, {"loss": 0.655, "grad_norm": 0.6933388113975525, "learning_rate": 0.0002, "epoch": 2.0451054886427937, "step": 12650}, {"loss": 0.6431, "grad_norm": 0.8029746413230896, "learning_rate": 0.0002, "epoch": 2.046722172823539, "step": 12660}, {"loss": 0.7164, "grad_norm": 0.946399986743927, "learning_rate": 0.0002, "epoch": 2.048338857004284, "step": 12670}, {"loss": 0.638, "grad_norm": 0.7072678804397583, "learning_rate": 0.0002, "epoch": 2.0499555411850294, "step": 12680}, {"loss": 0.6487, "grad_norm": 0.6810618042945862, "learning_rate": 0.0002, "epoch": 2.0515722253657747, "step": 12690}, {"loss": 0.6554, "grad_norm": 0.7661160230636597, "learning_rate": 0.0002, "epoch": 2.05318890954652, "step": 12700}, {"loss": 0.6799, "grad_norm": 0.6350653767585754, "learning_rate": 0.0002, "epoch": 2.0548055937272656, "step": 12710}, {"loss": 0.6654, "grad_norm": 0.861890971660614, "learning_rate": 0.0002, "epoch": 2.056422277908011, "step": 12720}, {"loss": 0.6286, "grad_norm": 0.6489875912666321, "learning_rate": 0.0002, "epoch": 2.058038962088756, "step": 12730}, {"loss": 0.6811, "grad_norm": 0.8268506526947021, "learning_rate": 0.0002, "epoch": 2.0596556462695013, "step": 12740}, {"loss": 0.6524, "grad_norm": 0.607679545879364, "learning_rate": 0.0002, "epoch": 2.0612723304502465, "step": 12750}, {"loss": 0.6649, "grad_norm": 0.6754153370857239, "learning_rate": 0.0002, "epoch": 2.0628890146309917, "step": 12760}, {"loss": 0.6549, "grad_norm": 0.7263124585151672, "learning_rate": 0.0002, "epoch": 2.064505698811737, "step": 12770}, {"loss": 0.6189, "grad_norm": 0.6986154317855835, "learning_rate": 0.0002, "epoch": 2.0661223829924826, "step": 12780}, {"loss": 0.6723, "grad_norm": 0.7768576741218567, "learning_rate": 0.0002, "epoch": 2.067739067173228, "step": 12790}, {"loss": 0.677, "grad_norm": 0.7546762824058533, "learning_rate": 0.0002, "epoch": 2.069355751353973, "step": 12800}, {"loss": 0.6485, "grad_norm": 0.7588880062103271, "learning_rate": 0.0002, "epoch": 2.0709724355347183, "step": 12810}, {"loss": 0.6989, "grad_norm": 0.7457242608070374, "learning_rate": 0.0002, "epoch": 2.0725891197154636, "step": 12820}, {"loss": 0.6489, "grad_norm": 0.6983516812324524, "learning_rate": 0.0002, "epoch": 2.074205803896209, "step": 12830}, {"loss": 0.651, "grad_norm": 0.7950928807258606, "learning_rate": 0.0002, "epoch": 2.075822488076954, "step": 12840}, {"loss": 0.6603, "grad_norm": 0.9248087406158447, "learning_rate": 0.0002, "epoch": 2.0774391722576993, "step": 12850}, {"loss": 0.6847, "grad_norm": 0.7229493260383606, "learning_rate": 0.0002, "epoch": 2.079055856438445, "step": 12860}, {"loss": 0.6702, "grad_norm": 0.5710847973823547, "learning_rate": 0.0002, "epoch": 2.08067254061919, "step": 12870}, {"loss": 0.6974, "grad_norm": 0.9580423831939697, "learning_rate": 0.0002, "epoch": 2.0822892247999354, "step": 12880}, {"loss": 0.6341, "grad_norm": 0.7399665713310242, "learning_rate": 0.0002, "epoch": 2.0839059089806806, "step": 12890}, {"loss": 0.6993, "grad_norm": 0.7981410622596741, "learning_rate": 0.0002, "epoch": 2.085522593161426, "step": 12900}, {"loss": 0.6976, "grad_norm": 0.870759904384613, "learning_rate": 0.0002, "epoch": 2.087139277342171, "step": 12910}, {"loss": 0.7194, "grad_norm": 0.7001481652259827, "learning_rate": 0.0002, "epoch": 2.0887559615229163, "step": 12920}, {"loss": 0.6383, "grad_norm": 0.6745418310165405, "learning_rate": 0.0002, "epoch": 2.090372645703662, "step": 12930}, {"loss": 0.6519, "grad_norm": 0.7739067673683167, "learning_rate": 0.0002, "epoch": 2.0919893298844072, "step": 12940}, {"loss": 0.6856, "grad_norm": 0.6742934584617615, "learning_rate": 0.0002, "epoch": 2.0936060140651525, "step": 12950}, {"loss": 0.6279, "grad_norm": 0.7270349860191345, "learning_rate": 0.0002, "epoch": 2.0952226982458977, "step": 12960}, {"loss": 0.6783, "grad_norm": 0.7150624394416809, "learning_rate": 0.0002, "epoch": 2.096839382426643, "step": 12970}, {"loss": 0.6093, "grad_norm": 0.7734767198562622, "learning_rate": 0.0002, "epoch": 2.098456066607388, "step": 12980}, {"loss": 0.6534, "grad_norm": 0.7618662118911743, "learning_rate": 0.0002, "epoch": 2.1000727507881334, "step": 12990}, {"loss": 0.6707, "grad_norm": 0.6557944416999817, "learning_rate": 0.0002, "epoch": 2.101689434968879, "step": 13000}, {"loss": 0.7268, "grad_norm": 0.8786448240280151, "learning_rate": 0.0002, "epoch": 2.1033061191496243, "step": 13010}, {"loss": 0.6677, "grad_norm": 0.6878724098205566, "learning_rate": 0.0002, "epoch": 2.1049228033303695, "step": 13020}, {"loss": 0.6824, "grad_norm": 0.822318971157074, "learning_rate": 0.0002, "epoch": 2.1065394875111147, "step": 13030}, {"loss": 0.6228, "grad_norm": 0.831468939781189, "learning_rate": 0.0002, "epoch": 2.10815617169186, "step": 13040}, {"loss": 0.6511, "grad_norm": 0.7699505686759949, "learning_rate": 0.0002, "epoch": 2.109772855872605, "step": 13050}, {"loss": 0.6671, "grad_norm": 0.7559016346931458, "learning_rate": 0.0002, "epoch": 2.1113895400533504, "step": 13060}, {"loss": 0.6215, "grad_norm": 0.6942209601402283, "learning_rate": 0.0002, "epoch": 2.1130062242340957, "step": 13070}, {"loss": 0.6449, "grad_norm": 0.6098947525024414, "learning_rate": 0.0002, "epoch": 2.1146229084148414, "step": 13080}, {"loss": 0.7091, "grad_norm": 0.6499016284942627, "learning_rate": 0.0002, "epoch": 2.1162395925955866, "step": 13090}, {"loss": 0.6247, "grad_norm": 0.7719953060150146, "learning_rate": 0.0002, "epoch": 2.117856276776332, "step": 13100}, {"loss": 0.6064, "grad_norm": 0.6708134412765503, "learning_rate": 0.0002, "epoch": 2.119472960957077, "step": 13110}, {"loss": 0.6056, "grad_norm": 0.8119585514068604, "learning_rate": 0.0002, "epoch": 2.1210896451378223, "step": 13120}, {"loss": 0.6628, "grad_norm": 0.6947157979011536, "learning_rate": 0.0002, "epoch": 2.1227063293185675, "step": 13130}, {"loss": 0.6375, "grad_norm": 0.8831837773323059, "learning_rate": 0.0002, "epoch": 2.1243230134993127, "step": 13140}, {"loss": 0.6997, "grad_norm": 0.7266910672187805, "learning_rate": 0.0002, "epoch": 2.1259396976800584, "step": 13150}, {"loss": 0.6446, "grad_norm": 0.8864351511001587, "learning_rate": 0.0002, "epoch": 2.1275563818608036, "step": 13160}, {"loss": 0.6762, "grad_norm": 0.8104248046875, "learning_rate": 0.0002, "epoch": 2.129173066041549, "step": 13170}, {"loss": 0.6581, "grad_norm": 0.6077079772949219, "learning_rate": 0.0002, "epoch": 2.130789750222294, "step": 13180}, {"loss": 0.6572, "grad_norm": 0.6874213814735413, "learning_rate": 0.0002, "epoch": 2.1324064344030393, "step": 13190}, {"loss": 0.642, "grad_norm": 0.7134367823600769, "learning_rate": 0.0002, "epoch": 2.1340231185837846, "step": 13200}, {"loss": 0.7016, "grad_norm": 0.6101235151290894, "learning_rate": 0.0002, "epoch": 2.13563980276453, "step": 13210}, {"loss": 0.6529, "grad_norm": 0.6042411923408508, "learning_rate": 0.0002, "epoch": 2.137256486945275, "step": 13220}, {"loss": 0.7179, "grad_norm": 0.914601743221283, "learning_rate": 0.0002, "epoch": 2.1388731711260207, "step": 13230}, {"loss": 0.6513, "grad_norm": 0.7104284167289734, "learning_rate": 0.0002, "epoch": 2.140489855306766, "step": 13240}, {"loss": 0.6607, "grad_norm": 0.664395272731781, "learning_rate": 0.0002, "epoch": 2.142106539487511, "step": 13250}, {"loss": 0.7211, "grad_norm": 0.6991241574287415, "learning_rate": 0.0002, "epoch": 2.1437232236682564, "step": 13260}, {"loss": 0.6484, "grad_norm": 0.5469560623168945, "learning_rate": 0.0002, "epoch": 2.1453399078490016, "step": 13270}, {"loss": 0.6765, "grad_norm": 0.8454998135566711, "learning_rate": 0.0002, "epoch": 2.146956592029747, "step": 13280}, {"loss": 0.6683, "grad_norm": 0.7088868618011475, "learning_rate": 0.0002, "epoch": 2.148573276210492, "step": 13290}, {"loss": 0.6835, "grad_norm": 0.7002687454223633, "learning_rate": 0.0002, "epoch": 2.1501899603912378, "step": 13300}, {"loss": 0.6399, "grad_norm": 0.7785214781761169, "learning_rate": 0.0002, "epoch": 2.151806644571983, "step": 13310}, {"loss": 0.67, "grad_norm": 0.8049132227897644, "learning_rate": 0.0002, "epoch": 2.1534233287527282, "step": 13320}, {"loss": 0.6495, "grad_norm": 0.8062595129013062, "learning_rate": 0.0002, "epoch": 2.1550400129334735, "step": 13330}, {"loss": 0.6603, "grad_norm": 0.6208319067955017, "learning_rate": 0.0002, "epoch": 2.1566566971142187, "step": 13340}, {"loss": 0.6584, "grad_norm": 0.7519655823707581, "learning_rate": 0.0002, "epoch": 2.158273381294964, "step": 13350}, {"loss": 0.6457, "grad_norm": 0.7645747065544128, "learning_rate": 0.0002, "epoch": 2.159890065475709, "step": 13360}, {"loss": 0.645, "grad_norm": 0.6847302913665771, "learning_rate": 0.0002, "epoch": 2.1615067496564544, "step": 13370}, {"loss": 0.6903, "grad_norm": 0.8630441427230835, "learning_rate": 0.0002, "epoch": 2.1631234338372, "step": 13380}, {"loss": 0.6742, "grad_norm": 0.7947702407836914, "learning_rate": 0.0002, "epoch": 2.1647401180179453, "step": 13390}, {"loss": 0.7206, "grad_norm": 0.6836977005004883, "learning_rate": 0.0002, "epoch": 2.1663568021986905, "step": 13400}, {"loss": 0.6304, "grad_norm": 0.7340566515922546, "learning_rate": 0.0002, "epoch": 2.1679734863794358, "step": 13410}, {"loss": 0.6528, "grad_norm": 0.7075738906860352, "learning_rate": 0.0002, "epoch": 2.169590170560181, "step": 13420}, {"loss": 0.6585, "grad_norm": 0.7080879807472229, "learning_rate": 0.0002, "epoch": 2.1712068547409262, "step": 13430}, {"loss": 0.6615, "grad_norm": 0.6218613386154175, "learning_rate": 0.0002, "epoch": 2.1728235389216715, "step": 13440}, {"loss": 0.6488, "grad_norm": 0.8211479187011719, "learning_rate": 0.0002, "epoch": 2.174440223102417, "step": 13450}, {"loss": 0.6738, "grad_norm": 0.864466667175293, "learning_rate": 0.0002, "epoch": 2.1760569072831624, "step": 13460}, {"loss": 0.679, "grad_norm": 0.7943857908248901, "learning_rate": 0.0002, "epoch": 2.1776735914639076, "step": 13470}, {"loss": 0.6838, "grad_norm": 0.78728187084198, "learning_rate": 0.0002, "epoch": 2.179290275644653, "step": 13480}, {"loss": 0.6397, "grad_norm": 0.697527289390564, "learning_rate": 0.0002, "epoch": 2.180906959825398, "step": 13490}, {"loss": 0.669, "grad_norm": 0.8205804228782654, "learning_rate": 0.0002, "epoch": 2.1825236440061433, "step": 13500}, {"loss": 0.7227, "grad_norm": 0.8709042072296143, "learning_rate": 0.0002, "epoch": 2.1841403281868885, "step": 13510}, {"loss": 0.6313, "grad_norm": 0.6228537559509277, "learning_rate": 0.0002, "epoch": 2.1857570123676338, "step": 13520}, {"loss": 0.7025, "grad_norm": 0.9566980004310608, "learning_rate": 0.0002, "epoch": 2.1873736965483794, "step": 13530}, {"loss": 0.6755, "grad_norm": 0.7128894329071045, "learning_rate": 0.0002, "epoch": 2.1889903807291247, "step": 13540}, {"loss": 0.6827, "grad_norm": 0.6888654232025146, "learning_rate": 0.0002, "epoch": 2.19060706490987, "step": 13550}, {"loss": 0.6961, "grad_norm": 0.6444337368011475, "learning_rate": 0.0002, "epoch": 2.192223749090615, "step": 13560}, {"loss": 0.656, "grad_norm": 0.8008806705474854, "learning_rate": 0.0002, "epoch": 2.1938404332713604, "step": 13570}, {"loss": 0.7, "grad_norm": 0.8482748866081238, "learning_rate": 0.0002, "epoch": 2.1954571174521056, "step": 13580}, {"loss": 0.7326, "grad_norm": 0.8584157228469849, "learning_rate": 0.0002, "epoch": 2.197073801632851, "step": 13590}, {"loss": 0.7014, "grad_norm": 0.7513734698295593, "learning_rate": 0.0002, "epoch": 2.1986904858135965, "step": 13600}, {"loss": 0.6632, "grad_norm": 0.7864262461662292, "learning_rate": 0.0002, "epoch": 2.2003071699943417, "step": 13610}, {"loss": 0.6879, "grad_norm": 0.8493645191192627, "learning_rate": 0.0002, "epoch": 2.201923854175087, "step": 13620}, {"loss": 0.6617, "grad_norm": 0.6902140974998474, "learning_rate": 0.0002, "epoch": 2.203540538355832, "step": 13630}, {"loss": 0.6655, "grad_norm": 0.8711254596710205, "learning_rate": 0.0002, "epoch": 2.2051572225365774, "step": 13640}, {"loss": 0.6359, "grad_norm": 0.7832191586494446, "learning_rate": 0.0002, "epoch": 2.2067739067173227, "step": 13650}, {"loss": 0.6723, "grad_norm": 0.5668176412582397, "learning_rate": 0.0002, "epoch": 2.208390590898068, "step": 13660}, {"loss": 0.635, "grad_norm": 0.8648375272750854, "learning_rate": 0.0002, "epoch": 2.2100072750788136, "step": 13670}, {"loss": 0.653, "grad_norm": 0.7643089890480042, "learning_rate": 0.0002, "epoch": 2.211623959259559, "step": 13680}, {"loss": 0.6765, "grad_norm": 0.6293777823448181, "learning_rate": 0.0002, "epoch": 2.213240643440304, "step": 13690}, {"loss": 0.6842, "grad_norm": 0.6459372639656067, "learning_rate": 0.0002, "epoch": 2.2148573276210493, "step": 13700}, {"loss": 0.6526, "grad_norm": 0.7060744166374207, "learning_rate": 0.0002, "epoch": 2.2164740118017945, "step": 13710}, {"loss": 0.7101, "grad_norm": 0.674109160900116, "learning_rate": 0.0002, "epoch": 2.2180906959825397, "step": 13720}, {"loss": 0.6529, "grad_norm": 0.830392062664032, "learning_rate": 0.0002, "epoch": 2.219707380163285, "step": 13730}, {"loss": 0.6733, "grad_norm": 0.6474477052688599, "learning_rate": 0.0002, "epoch": 2.2213240643440306, "step": 13740}, {"loss": 0.6413, "grad_norm": 0.7037909626960754, "learning_rate": 0.0002, "epoch": 2.222940748524776, "step": 13750}, {"loss": 0.6417, "grad_norm": 0.6554131507873535, "learning_rate": 0.0002, "epoch": 2.224557432705521, "step": 13760}, {"loss": 0.6907, "grad_norm": 0.7822230458259583, "learning_rate": 0.0002, "epoch": 2.2261741168862663, "step": 13770}, {"loss": 0.6505, "grad_norm": 0.9082167744636536, "learning_rate": 0.0002, "epoch": 2.2277908010670116, "step": 13780}, {"loss": 0.6878, "grad_norm": 0.7918276190757751, "learning_rate": 0.0002, "epoch": 2.229407485247757, "step": 13790}, {"loss": 0.6669, "grad_norm": 0.7354569435119629, "learning_rate": 0.0002, "epoch": 2.231024169428502, "step": 13800}, {"loss": 0.6503, "grad_norm": 0.8265249133110046, "learning_rate": 0.0002, "epoch": 2.2326408536092472, "step": 13810}, {"loss": 0.6871, "grad_norm": 0.6653847098350525, "learning_rate": 0.0002, "epoch": 2.234257537789993, "step": 13820}, {"loss": 0.6413, "grad_norm": 0.7157923579216003, "learning_rate": 0.0002, "epoch": 2.235874221970738, "step": 13830}, {"loss": 0.6306, "grad_norm": 0.7110323309898376, "learning_rate": 0.0002, "epoch": 2.2374909061514834, "step": 13840}, {"loss": 0.6913, "grad_norm": 0.7155357599258423, "learning_rate": 0.0002, "epoch": 2.2391075903322286, "step": 13850}, {"loss": 0.6579, "grad_norm": 1.0177817344665527, "learning_rate": 0.0002, "epoch": 2.240724274512974, "step": 13860}, {"loss": 0.635, "grad_norm": 0.7601948380470276, "learning_rate": 0.0002, "epoch": 2.242340958693719, "step": 13870}, {"loss": 0.6679, "grad_norm": 0.7628820538520813, "learning_rate": 0.0002, "epoch": 2.2439576428744643, "step": 13880}, {"loss": 0.6805, "grad_norm": 0.7089297771453857, "learning_rate": 0.0002, "epoch": 2.24557432705521, "step": 13890}, {"loss": 0.7236, "grad_norm": 0.695178210735321, "learning_rate": 0.0002, "epoch": 2.247191011235955, "step": 13900}, {"loss": 0.7084, "grad_norm": 0.7631948590278625, "learning_rate": 0.0002, "epoch": 2.2488076954167004, "step": 13910}, {"loss": 0.685, "grad_norm": 0.8203101754188538, "learning_rate": 0.0002, "epoch": 2.2504243795974457, "step": 13920}, {"loss": 0.653, "grad_norm": 0.8099079728126526, "learning_rate": 0.0002, "epoch": 2.252041063778191, "step": 13930}, {"loss": 0.694, "grad_norm": 0.6498546004295349, "learning_rate": 0.0002, "epoch": 2.253657747958936, "step": 13940}, {"loss": 0.6684, "grad_norm": 0.7797415256500244, "learning_rate": 0.0002, "epoch": 2.2552744321396814, "step": 13950}, {"loss": 0.683, "grad_norm": 0.8254124522209167, "learning_rate": 0.0002, "epoch": 2.2568911163204266, "step": 13960}, {"loss": 0.6806, "grad_norm": 0.6327953338623047, "learning_rate": 0.0002, "epoch": 2.2585078005011723, "step": 13970}, {"loss": 0.668, "grad_norm": 0.734194278717041, "learning_rate": 0.0002, "epoch": 2.2601244846819175, "step": 13980}, {"loss": 0.6912, "grad_norm": 0.9014202952384949, "learning_rate": 0.0002, "epoch": 2.2617411688626627, "step": 13990}, {"loss": 0.692, "grad_norm": 0.7643631100654602, "learning_rate": 0.0002, "epoch": 2.263357853043408, "step": 14000}, {"loss": 0.6657, "grad_norm": 0.8882834911346436, "learning_rate": 0.0002, "epoch": 2.264974537224153, "step": 14010}, {"loss": 0.6453, "grad_norm": 0.7975873351097107, "learning_rate": 0.0002, "epoch": 2.2665912214048984, "step": 14020}, {"loss": 0.7193, "grad_norm": 0.7765783071517944, "learning_rate": 0.0002, "epoch": 2.2682079055856437, "step": 14030}, {"loss": 0.662, "grad_norm": 0.8846288323402405, "learning_rate": 0.0002, "epoch": 2.2698245897663893, "step": 14040}, {"loss": 0.6494, "grad_norm": 0.9006744027137756, "learning_rate": 0.0002, "epoch": 2.2714412739471346, "step": 14050}, {"loss": 0.6423, "grad_norm": 0.7420173287391663, "learning_rate": 0.0002, "epoch": 2.27305795812788, "step": 14060}, {"loss": 0.7068, "grad_norm": 0.7956424951553345, "learning_rate": 0.0002, "epoch": 2.274674642308625, "step": 14070}, {"loss": 0.6581, "grad_norm": 0.7783209085464478, "learning_rate": 0.0002, "epoch": 2.2762913264893703, "step": 14080}, {"loss": 0.7202, "grad_norm": 0.7597188949584961, "learning_rate": 0.0002, "epoch": 2.2779080106701155, "step": 14090}, {"loss": 0.6778, "grad_norm": 0.6718921661376953, "learning_rate": 0.0002, "epoch": 2.2795246948508607, "step": 14100}, {"loss": 0.632, "grad_norm": 0.7528082132339478, "learning_rate": 0.0002, "epoch": 2.281141379031606, "step": 14110}, {"loss": 0.7608, "grad_norm": 0.8379864692687988, "learning_rate": 0.0002, "epoch": 2.2827580632123516, "step": 14120}, {"loss": 0.6767, "grad_norm": 0.748613715171814, "learning_rate": 0.0002, "epoch": 2.284374747393097, "step": 14130}, {"loss": 0.6641, "grad_norm": 0.7435423135757446, "learning_rate": 0.0002, "epoch": 2.285991431573842, "step": 14140}, {"loss": 0.6849, "grad_norm": 0.7580803632736206, "learning_rate": 0.0002, "epoch": 2.2876081157545873, "step": 14150}, {"loss": 0.6604, "grad_norm": 0.6278321146965027, "learning_rate": 0.0002, "epoch": 2.2892247999353326, "step": 14160}, {"loss": 0.6573, "grad_norm": 0.7663896083831787, "learning_rate": 0.0002, "epoch": 2.290841484116078, "step": 14170}, {"loss": 0.6655, "grad_norm": 0.9716812372207642, "learning_rate": 0.0002, "epoch": 2.292458168296823, "step": 14180}, {"loss": 0.7067, "grad_norm": 0.8993458151817322, "learning_rate": 0.0002, "epoch": 2.2940748524775687, "step": 14190}, {"loss": 0.6172, "grad_norm": 0.6156117916107178, "learning_rate": 0.0002, "epoch": 2.295691536658314, "step": 14200}, {"loss": 0.6318, "grad_norm": 0.8911278247833252, "learning_rate": 0.0002, "epoch": 2.297308220839059, "step": 14210}, {"loss": 0.6364, "grad_norm": 0.6422147154808044, "learning_rate": 0.0002, "epoch": 2.2989249050198044, "step": 14220}, {"loss": 0.6795, "grad_norm": 0.6866879463195801, "learning_rate": 0.0002, "epoch": 2.3005415892005496, "step": 14230}, {"loss": 0.6907, "grad_norm": 0.9297130107879639, "learning_rate": 0.0002, "epoch": 2.302158273381295, "step": 14240}, {"loss": 0.6823, "grad_norm": 0.7501356601715088, "learning_rate": 0.0002, "epoch": 2.30377495756204, "step": 14250}, {"loss": 0.6414, "grad_norm": 0.8363515138626099, "learning_rate": 0.0002, "epoch": 2.3053916417427853, "step": 14260}, {"loss": 0.6362, "grad_norm": 0.9083868265151978, "learning_rate": 0.0002, "epoch": 2.307008325923531, "step": 14270}, {"loss": 0.6862, "grad_norm": 0.7791516780853271, "learning_rate": 0.0002, "epoch": 2.3086250101042762, "step": 14280}, {"loss": 0.6569, "grad_norm": 0.8766953349113464, "learning_rate": 0.0002, "epoch": 2.3102416942850215, "step": 14290}, {"loss": 0.6698, "grad_norm": 0.7916635274887085, "learning_rate": 0.0002, "epoch": 2.3118583784657667, "step": 14300}, {"loss": 0.6927, "grad_norm": 0.627525269985199, "learning_rate": 0.0002, "epoch": 2.313475062646512, "step": 14310}, {"loss": 0.6541, "grad_norm": 0.8856783509254456, "learning_rate": 0.0002, "epoch": 2.315091746827257, "step": 14320}, {"loss": 0.6806, "grad_norm": 0.6758689284324646, "learning_rate": 0.0002, "epoch": 2.316708431008003, "step": 14330}, {"loss": 0.6794, "grad_norm": 0.6428321003913879, "learning_rate": 0.0002, "epoch": 2.318325115188748, "step": 14340}, {"loss": 0.682, "grad_norm": 0.9032121300697327, "learning_rate": 0.0002, "epoch": 2.3199417993694933, "step": 14350}, {"loss": 0.6569, "grad_norm": 0.8035986423492432, "learning_rate": 0.0002, "epoch": 2.3215584835502385, "step": 14360}, {"loss": 0.7067, "grad_norm": 0.7974579334259033, "learning_rate": 0.0002, "epoch": 2.3231751677309838, "step": 14370}, {"loss": 0.6451, "grad_norm": 0.8356034755706787, "learning_rate": 0.0002, "epoch": 2.324791851911729, "step": 14380}, {"loss": 0.6623, "grad_norm": 0.998760998249054, "learning_rate": 0.0002, "epoch": 2.326408536092474, "step": 14390}, {"loss": 0.649, "grad_norm": 0.6518142223358154, "learning_rate": 0.0002, "epoch": 2.3280252202732195, "step": 14400}, {"loss": 0.7146, "grad_norm": 0.7443506717681885, "learning_rate": 0.0002, "epoch": 2.3296419044539647, "step": 14410}, {"loss": 0.648, "grad_norm": 0.8436172604560852, "learning_rate": 0.0002, "epoch": 2.3312585886347104, "step": 14420}, {"loss": 0.6585, "grad_norm": 0.7411080598831177, "learning_rate": 0.0002, "epoch": 2.3328752728154556, "step": 14430}, {"loss": 0.6781, "grad_norm": 0.8839048743247986, "learning_rate": 0.0002, "epoch": 2.334491956996201, "step": 14440}, {"loss": 0.6565, "grad_norm": 0.8360885977745056, "learning_rate": 0.0002, "epoch": 2.336108641176946, "step": 14450}, {"loss": 0.6662, "grad_norm": 0.7608986496925354, "learning_rate": 0.0002, "epoch": 2.3377253253576913, "step": 14460}, {"loss": 0.6685, "grad_norm": 0.8179867267608643, "learning_rate": 0.0002, "epoch": 2.3393420095384365, "step": 14470}, {"loss": 0.7055, "grad_norm": 0.5989999771118164, "learning_rate": 0.0002, "epoch": 2.340958693719182, "step": 14480}, {"loss": 0.644, "grad_norm": 0.9450054168701172, "learning_rate": 0.0002, "epoch": 2.3425753778999274, "step": 14490}, {"loss": 0.6983, "grad_norm": 0.7885149717330933, "learning_rate": 0.0002, "epoch": 2.3441920620806727, "step": 14500}, {"loss": 0.6819, "grad_norm": 0.8152616620063782, "learning_rate": 0.0002, "epoch": 2.345808746261418, "step": 14510}, {"loss": 0.6989, "grad_norm": 0.7193838953971863, "learning_rate": 0.0002, "epoch": 2.347425430442163, "step": 14520}, {"loss": 0.6594, "grad_norm": 0.6701092720031738, "learning_rate": 0.0002, "epoch": 2.3490421146229084, "step": 14530}, {"loss": 0.6559, "grad_norm": 0.7529364228248596, "learning_rate": 0.0002, "epoch": 2.3506587988036536, "step": 14540}, {"loss": 0.6306, "grad_norm": 0.6599733829498291, "learning_rate": 0.0002, "epoch": 2.352275482984399, "step": 14550}, {"loss": 0.706, "grad_norm": 0.9502474069595337, "learning_rate": 0.0002, "epoch": 2.353892167165144, "step": 14560}, {"loss": 0.717, "grad_norm": 0.7619650959968567, "learning_rate": 0.0002, "epoch": 2.3555088513458897, "step": 14570}, {"loss": 0.6684, "grad_norm": 0.9854652285575867, "learning_rate": 0.0002, "epoch": 2.357125535526635, "step": 14580}, {"loss": 0.6455, "grad_norm": 0.727439284324646, "learning_rate": 0.0002, "epoch": 2.35874221970738, "step": 14590}, {"loss": 0.6645, "grad_norm": 0.6994746327400208, "learning_rate": 0.0002, "epoch": 2.3603589038881254, "step": 14600}, {"loss": 0.6587, "grad_norm": 0.7117531299591064, "learning_rate": 0.0002, "epoch": 2.3619755880688706, "step": 14610}, {"loss": 0.6804, "grad_norm": 0.6403067708015442, "learning_rate": 0.0002, "epoch": 2.363592272249616, "step": 14620}, {"loss": 0.7055, "grad_norm": 0.8377841711044312, "learning_rate": 0.0002, "epoch": 2.3652089564303616, "step": 14630}, {"loss": 0.6778, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 2.366825640611107, "step": 14640}, {"loss": 0.6552, "grad_norm": 0.8418586254119873, "learning_rate": 0.0002, "epoch": 2.368442324791852, "step": 14650}, {"loss": 0.6685, "grad_norm": 0.6178573369979858, "learning_rate": 0.0002, "epoch": 2.3700590089725972, "step": 14660}, {"loss": 0.6774, "grad_norm": 0.6368302702903748, "learning_rate": 0.0002, "epoch": 2.3716756931533425, "step": 14670}, {"loss": 0.6136, "grad_norm": 0.9122977256774902, "learning_rate": 0.0002, "epoch": 2.3732923773340877, "step": 14680}, {"loss": 0.6675, "grad_norm": 0.7086195349693298, "learning_rate": 0.0002, "epoch": 2.374909061514833, "step": 14690}, {"loss": 0.6582, "grad_norm": 0.7500800490379333, "learning_rate": 0.0002, "epoch": 2.376525745695578, "step": 14700}, {"loss": 0.6792, "grad_norm": 0.6634900569915771, "learning_rate": 0.0002, "epoch": 2.378142429876324, "step": 14710}, {"loss": 0.6614, "grad_norm": 0.839898407459259, "learning_rate": 0.0002, "epoch": 2.379759114057069, "step": 14720}, {"loss": 0.6453, "grad_norm": 0.7578426003456116, "learning_rate": 0.0002, "epoch": 2.3813757982378143, "step": 14730}, {"loss": 0.7282, "grad_norm": 1.0213173627853394, "learning_rate": 0.0002, "epoch": 2.3829924824185595, "step": 14740}, {"loss": 0.6704, "grad_norm": 0.7855949401855469, "learning_rate": 0.0002, "epoch": 2.3846091665993048, "step": 14750}, {"loss": 0.6694, "grad_norm": 0.7224128842353821, "learning_rate": 0.0002, "epoch": 2.38622585078005, "step": 14760}, {"loss": 0.7017, "grad_norm": 0.8040381669998169, "learning_rate": 0.0002, "epoch": 2.3878425349607952, "step": 14770}, {"loss": 0.6799, "grad_norm": 0.7705281376838684, "learning_rate": 0.0002, "epoch": 2.389459219141541, "step": 14780}, {"loss": 0.6326, "grad_norm": 0.667966902256012, "learning_rate": 0.0002, "epoch": 2.391075903322286, "step": 14790}, {"loss": 0.7061, "grad_norm": 0.6611011028289795, "learning_rate": 0.0002, "epoch": 2.3926925875030314, "step": 14800}, {"loss": 0.6527, "grad_norm": 0.6862651705741882, "learning_rate": 0.0002, "epoch": 2.3943092716837766, "step": 14810}, {"loss": 0.6537, "grad_norm": 0.8086010217666626, "learning_rate": 0.0002, "epoch": 2.395925955864522, "step": 14820}, {"loss": 0.7189, "grad_norm": 0.7189689874649048, "learning_rate": 0.0002, "epoch": 2.397542640045267, "step": 14830}, {"loss": 0.6709, "grad_norm": 0.6280009150505066, "learning_rate": 0.0002, "epoch": 2.3991593242260123, "step": 14840}, {"loss": 0.706, "grad_norm": 0.7826612591743469, "learning_rate": 0.0002, "epoch": 2.4007760084067575, "step": 14850}, {"loss": 0.6738, "grad_norm": 0.7681610584259033, "learning_rate": 0.0002, "epoch": 2.402392692587503, "step": 14860}, {"loss": 0.636, "grad_norm": 0.720966100692749, "learning_rate": 0.0002, "epoch": 2.4040093767682484, "step": 14870}, {"loss": 0.6667, "grad_norm": 0.8202250599861145, "learning_rate": 0.0002, "epoch": 2.4056260609489937, "step": 14880}, {"loss": 0.6935, "grad_norm": 0.786212682723999, "learning_rate": 0.0002, "epoch": 2.407242745129739, "step": 14890}, {"loss": 0.6628, "grad_norm": 0.6647164821624756, "learning_rate": 0.0002, "epoch": 2.408859429310484, "step": 14900}, {"loss": 0.6706, "grad_norm": 0.7566399574279785, "learning_rate": 0.0002, "epoch": 2.4104761134912294, "step": 14910}, {"loss": 0.7188, "grad_norm": 0.748814582824707, "learning_rate": 0.0002, "epoch": 2.4120927976719746, "step": 14920}, {"loss": 0.6684, "grad_norm": 0.7624038457870483, "learning_rate": 0.0002, "epoch": 2.4137094818527203, "step": 14930}, {"loss": 0.6483, "grad_norm": 0.8267335295677185, "learning_rate": 0.0002, "epoch": 2.4153261660334655, "step": 14940}, {"loss": 0.6612, "grad_norm": 0.8785360455513, "learning_rate": 0.0002, "epoch": 2.4169428502142107, "step": 14950}, {"loss": 0.6718, "grad_norm": 0.679887592792511, "learning_rate": 0.0002, "epoch": 2.418559534394956, "step": 14960}, {"loss": 0.6136, "grad_norm": 0.7218474745750427, "learning_rate": 0.0002, "epoch": 2.420176218575701, "step": 14970}, {"loss": 0.648, "grad_norm": 0.6342799663543701, "learning_rate": 0.0002, "epoch": 2.4217929027564464, "step": 14980}, {"loss": 0.6617, "grad_norm": 0.7098712921142578, "learning_rate": 0.0002, "epoch": 2.4234095869371917, "step": 14990}, {"loss": 0.6942, "grad_norm": 0.7497431635856628, "learning_rate": 0.0002, "epoch": 2.425026271117937, "step": 15000}, {"loss": 0.6772, "grad_norm": 0.934836208820343, "learning_rate": 0.0002, "epoch": 2.4266429552986826, "step": 15010}, {"loss": 0.7221, "grad_norm": 0.8430966734886169, "learning_rate": 0.0002, "epoch": 2.428259639479428, "step": 15020}, {"loss": 0.6985, "grad_norm": 0.7032104730606079, "learning_rate": 0.0002, "epoch": 2.429876323660173, "step": 15030}, {"loss": 0.6715, "grad_norm": 0.7746111750602722, "learning_rate": 0.0002, "epoch": 2.4314930078409183, "step": 15040}, {"loss": 0.7177, "grad_norm": 0.7661406397819519, "learning_rate": 0.0002, "epoch": 2.4331096920216635, "step": 15050}, {"loss": 0.6517, "grad_norm": 0.6941645741462708, "learning_rate": 0.0002, "epoch": 2.4347263762024087, "step": 15060}, {"loss": 0.6421, "grad_norm": 0.7487249374389648, "learning_rate": 0.0002, "epoch": 2.436343060383154, "step": 15070}, {"loss": 0.6796, "grad_norm": 0.7639912962913513, "learning_rate": 0.0002, "epoch": 2.4379597445638996, "step": 15080}, {"loss": 0.7087, "grad_norm": 0.7708953619003296, "learning_rate": 0.0002, "epoch": 2.439576428744645, "step": 15090}, {"loss": 0.7065, "grad_norm": 0.9135832190513611, "learning_rate": 0.0002, "epoch": 2.44119311292539, "step": 15100}, {"loss": 0.672, "grad_norm": 0.8283005356788635, "learning_rate": 0.0002, "epoch": 2.4428097971061353, "step": 15110}, {"loss": 0.6551, "grad_norm": 0.925299346446991, "learning_rate": 0.0002, "epoch": 2.4444264812868806, "step": 15120}, {"loss": 0.687, "grad_norm": 0.7013528943061829, "learning_rate": 0.0002, "epoch": 2.446043165467626, "step": 15130}, {"loss": 0.6842, "grad_norm": 0.622303307056427, "learning_rate": 0.0002, "epoch": 2.447659849648371, "step": 15140}, {"loss": 0.6676, "grad_norm": 0.876569390296936, "learning_rate": 0.0002, "epoch": 2.4492765338291163, "step": 15150}, {"loss": 0.6463, "grad_norm": 0.6836351752281189, "learning_rate": 0.0002, "epoch": 2.450893218009862, "step": 15160}, {"loss": 0.6781, "grad_norm": 0.7886684536933899, "learning_rate": 0.0002, "epoch": 2.452509902190607, "step": 15170}, {"loss": 0.6794, "grad_norm": 0.6647440791130066, "learning_rate": 0.0002, "epoch": 2.4541265863713524, "step": 15180}, {"loss": 0.6353, "grad_norm": 0.7477722764015198, "learning_rate": 0.0002, "epoch": 2.4557432705520976, "step": 15190}, {"loss": 0.698, "grad_norm": 0.8192033767700195, "learning_rate": 0.0002, "epoch": 2.457359954732843, "step": 15200}, {"loss": 0.6735, "grad_norm": 0.847537100315094, "learning_rate": 0.0002, "epoch": 2.458976638913588, "step": 15210}, {"loss": 0.6962, "grad_norm": 0.9027776122093201, "learning_rate": 0.0002, "epoch": 2.4605933230943338, "step": 15220}, {"loss": 0.7084, "grad_norm": 0.7217772006988525, "learning_rate": 0.0002, "epoch": 2.462210007275079, "step": 15230}, {"loss": 0.691, "grad_norm": 0.7994546294212341, "learning_rate": 0.0002, "epoch": 2.4638266914558242, "step": 15240}, {"loss": 0.6828, "grad_norm": 0.939916729927063, "learning_rate": 0.0002, "epoch": 2.4654433756365695, "step": 15250}, {"loss": 0.6893, "grad_norm": 1.0009053945541382, "learning_rate": 0.0002, "epoch": 2.4670600598173147, "step": 15260}, {"loss": 0.643, "grad_norm": 0.625555694103241, "learning_rate": 0.0002, "epoch": 2.46867674399806, "step": 15270}, {"loss": 0.688, "grad_norm": 0.7924878597259521, "learning_rate": 0.0002, "epoch": 2.470293428178805, "step": 15280}, {"loss": 0.6789, "grad_norm": 0.8536689877510071, "learning_rate": 0.0002, "epoch": 2.4719101123595504, "step": 15290}, {"loss": 0.6924, "grad_norm": 0.8572589755058289, "learning_rate": 0.0002, "epoch": 2.4735267965402956, "step": 15300}, {"loss": 0.604, "grad_norm": 0.773279070854187, "learning_rate": 0.0002, "epoch": 2.4751434807210413, "step": 15310}, {"loss": 0.6573, "grad_norm": 0.7708749771118164, "learning_rate": 0.0002, "epoch": 2.4767601649017865, "step": 15320}, {"loss": 0.7065, "grad_norm": 0.770905077457428, "learning_rate": 0.0002, "epoch": 2.4783768490825318, "step": 15330}, {"loss": 0.6878, "grad_norm": 0.8238571882247925, "learning_rate": 0.0002, "epoch": 2.479993533263277, "step": 15340}, {"loss": 0.6772, "grad_norm": 0.7670477032661438, "learning_rate": 0.0002, "epoch": 2.481610217444022, "step": 15350}, {"loss": 0.7759, "grad_norm": 0.905036985874176, "learning_rate": 0.0002, "epoch": 2.4832269016247674, "step": 15360}, {"loss": 0.706, "grad_norm": 0.6672089695930481, "learning_rate": 0.0002, "epoch": 2.484843585805513, "step": 15370}, {"loss": 0.6722, "grad_norm": 0.625095784664154, "learning_rate": 0.0002, "epoch": 2.4864602699862584, "step": 15380}, {"loss": 0.6396, "grad_norm": 0.679772675037384, "learning_rate": 0.0002, "epoch": 2.4880769541670036, "step": 15390}, {"loss": 0.6778, "grad_norm": 0.711492121219635, "learning_rate": 0.0002, "epoch": 2.489693638347749, "step": 15400}, {"loss": 0.6966, "grad_norm": 0.876189112663269, "learning_rate": 0.0002, "epoch": 2.491310322528494, "step": 15410}, {"loss": 0.7307, "grad_norm": 0.7236915230751038, "learning_rate": 0.0002, "epoch": 2.4929270067092393, "step": 15420}, {"loss": 0.647, "grad_norm": 0.6629832983016968, "learning_rate": 0.0002, "epoch": 2.4945436908899845, "step": 15430}, {"loss": 0.6669, "grad_norm": 0.9756859540939331, "learning_rate": 0.0002, "epoch": 2.4961603750707297, "step": 15440}, {"loss": 0.7559, "grad_norm": 0.6896940469741821, "learning_rate": 0.0002, "epoch": 2.4977770592514754, "step": 15450}, {"loss": 0.6818, "grad_norm": 0.7105149626731873, "learning_rate": 0.0002, "epoch": 2.4993937434322206, "step": 15460}, {"loss": 0.6859, "grad_norm": 0.8374546766281128, "learning_rate": 0.0002, "epoch": 2.501010427612966, "step": 15470}, {"loss": 0.6512, "grad_norm": 0.7320070266723633, "learning_rate": 0.0002, "epoch": 2.502627111793711, "step": 15480}, {"loss": 0.685, "grad_norm": 0.8306367993354797, "learning_rate": 0.0002, "epoch": 2.5042437959744563, "step": 15490}, {"loss": 0.7253, "grad_norm": 0.7472721338272095, "learning_rate": 0.0002, "epoch": 2.5058604801552016, "step": 15500}, {"loss": 0.6699, "grad_norm": 0.6147692203521729, "learning_rate": 0.0002, "epoch": 2.507477164335947, "step": 15510}, {"loss": 0.7158, "grad_norm": 0.7788505554199219, "learning_rate": 0.0002, "epoch": 2.5090938485166925, "step": 15520}, {"loss": 0.6521, "grad_norm": 0.8807527422904968, "learning_rate": 0.0002, "epoch": 2.5107105326974377, "step": 15530}, {"loss": 0.6792, "grad_norm": 0.7521643042564392, "learning_rate": 0.0002, "epoch": 2.512327216878183, "step": 15540}, {"loss": 0.6772, "grad_norm": 0.6900225281715393, "learning_rate": 0.0002, "epoch": 2.513943901058928, "step": 15550}, {"loss": 0.6769, "grad_norm": 0.6601938605308533, "learning_rate": 0.0002, "epoch": 2.5155605852396734, "step": 15560}, {"loss": 0.6648, "grad_norm": 0.8179984092712402, "learning_rate": 0.0002, "epoch": 2.5171772694204186, "step": 15570}, {"loss": 0.7028, "grad_norm": 0.792556881904602, "learning_rate": 0.0002, "epoch": 2.518793953601164, "step": 15580}, {"loss": 0.6464, "grad_norm": 0.7081938982009888, "learning_rate": 0.0002, "epoch": 2.520410637781909, "step": 15590}, {"loss": 0.6691, "grad_norm": 0.8733121156692505, "learning_rate": 0.0002, "epoch": 2.5220273219626543, "step": 15600}, {"loss": 0.6969, "grad_norm": 0.7980992794036865, "learning_rate": 0.0002, "epoch": 2.5236440061434, "step": 15610}, {"loss": 0.7124, "grad_norm": 0.883664071559906, "learning_rate": 0.0002, "epoch": 2.5252606903241452, "step": 15620}, {"loss": 0.7022, "grad_norm": 0.6963341236114502, "learning_rate": 0.0002, "epoch": 2.5268773745048905, "step": 15630}, {"loss": 0.7334, "grad_norm": 0.6433573365211487, "learning_rate": 0.0002, "epoch": 2.5284940586856357, "step": 15640}, {"loss": 0.6889, "grad_norm": 0.8538183569908142, "learning_rate": 0.0002, "epoch": 2.530110742866381, "step": 15650}, {"loss": 0.6841, "grad_norm": 0.9748201370239258, "learning_rate": 0.0002, "epoch": 2.5317274270471266, "step": 15660}, {"loss": 0.6765, "grad_norm": 0.7670575380325317, "learning_rate": 0.0002, "epoch": 2.533344111227872, "step": 15670}, {"loss": 0.6435, "grad_norm": 0.8738890290260315, "learning_rate": 0.0002, "epoch": 2.534960795408617, "step": 15680}, {"loss": 0.6802, "grad_norm": 0.8391636610031128, "learning_rate": 0.0002, "epoch": 2.5365774795893623, "step": 15690}, {"loss": 0.6901, "grad_norm": 0.7239366769790649, "learning_rate": 0.0002, "epoch": 2.5381941637701075, "step": 15700}, {"loss": 0.7011, "grad_norm": 0.8498379588127136, "learning_rate": 0.0002, "epoch": 2.5398108479508528, "step": 15710}, {"loss": 0.6998, "grad_norm": 0.8029484152793884, "learning_rate": 0.0002, "epoch": 2.541427532131598, "step": 15720}, {"loss": 0.6678, "grad_norm": 1.0639333724975586, "learning_rate": 0.0002, "epoch": 2.5430442163123432, "step": 15730}, {"loss": 0.6341, "grad_norm": 0.6401297450065613, "learning_rate": 0.0002, "epoch": 2.5446609004930885, "step": 15740}, {"loss": 0.7196, "grad_norm": 0.7123814821243286, "learning_rate": 0.0002, "epoch": 2.5462775846738337, "step": 15750}, {"loss": 0.654, "grad_norm": 0.7874974608421326, "learning_rate": 0.0002, "epoch": 2.5478942688545794, "step": 15760}, {"loss": 0.6721, "grad_norm": 0.8046808838844299, "learning_rate": 0.0002, "epoch": 2.5495109530353246, "step": 15770}, {"loss": 0.6665, "grad_norm": 0.7888661623001099, "learning_rate": 0.0002, "epoch": 2.55112763721607, "step": 15780}, {"loss": 0.6893, "grad_norm": 0.8445866107940674, "learning_rate": 0.0002, "epoch": 2.552744321396815, "step": 15790}, {"loss": 0.6815, "grad_norm": 0.7475846409797668, "learning_rate": 0.0002, "epoch": 2.5543610055775603, "step": 15800}, {"loss": 0.6711, "grad_norm": 0.7455102801322937, "learning_rate": 0.0002, "epoch": 2.555977689758306, "step": 15810}, {"loss": 0.6932, "grad_norm": 0.8226983547210693, "learning_rate": 0.0002, "epoch": 2.557594373939051, "step": 15820}, {"loss": 0.651, "grad_norm": 0.8920368552207947, "learning_rate": 0.0002, "epoch": 2.5592110581197964, "step": 15830}, {"loss": 0.6297, "grad_norm": 0.8413904905319214, "learning_rate": 0.0002, "epoch": 2.5608277423005417, "step": 15840}, {"loss": 0.7106, "grad_norm": 0.8483649492263794, "learning_rate": 0.0002, "epoch": 2.562444426481287, "step": 15850}, {"loss": 0.6957, "grad_norm": 0.5923284292221069, "learning_rate": 0.0002, "epoch": 2.564061110662032, "step": 15860}, {"loss": 0.6847, "grad_norm": 0.8518726229667664, "learning_rate": 0.0002, "epoch": 2.5656777948427774, "step": 15870}, {"loss": 0.6362, "grad_norm": 0.731235146522522, "learning_rate": 0.0002, "epoch": 2.5672944790235226, "step": 15880}, {"loss": 0.7611, "grad_norm": 0.7517194151878357, "learning_rate": 0.0002, "epoch": 2.568911163204268, "step": 15890}, {"loss": 0.6907, "grad_norm": 0.8378692269325256, "learning_rate": 0.0002, "epoch": 2.5705278473850135, "step": 15900}, {"loss": 0.7055, "grad_norm": 0.843701958656311, "learning_rate": 0.0002, "epoch": 2.5721445315657587, "step": 15910}, {"loss": 0.6882, "grad_norm": 0.7254629731178284, "learning_rate": 0.0002, "epoch": 2.573761215746504, "step": 15920}, {"loss": 0.6872, "grad_norm": 0.8863335847854614, "learning_rate": 0.0002, "epoch": 2.575377899927249, "step": 15930}, {"loss": 0.6813, "grad_norm": 0.7675097584724426, "learning_rate": 0.0002, "epoch": 2.5769945841079944, "step": 15940}, {"loss": 0.7357, "grad_norm": 0.82063889503479, "learning_rate": 0.0002, "epoch": 2.5786112682887397, "step": 15950}, {"loss": 0.662, "grad_norm": 0.7729717493057251, "learning_rate": 0.0002, "epoch": 2.5802279524694853, "step": 15960}, {"loss": 0.633, "grad_norm": 0.8301846981048584, "learning_rate": 0.0002, "epoch": 2.5818446366502306, "step": 15970}, {"loss": 0.6897, "grad_norm": 0.7906861305236816, "learning_rate": 0.0002, "epoch": 2.583461320830976, "step": 15980}, {"loss": 0.7175, "grad_norm": 0.6749057173728943, "learning_rate": 0.0002, "epoch": 2.585078005011721, "step": 15990}, {"loss": 0.7212, "grad_norm": 0.9386842846870422, "learning_rate": 0.0002, "epoch": 2.5866946891924663, "step": 16000}, {"loss": 0.6934, "grad_norm": 0.7868891358375549, "learning_rate": 0.0002, "epoch": 2.5883113733732115, "step": 16010}, {"loss": 0.7036, "grad_norm": 0.8674671053886414, "learning_rate": 0.0002, "epoch": 2.5899280575539567, "step": 16020}, {"loss": 0.7217, "grad_norm": 0.7043559551239014, "learning_rate": 0.0002, "epoch": 2.591544741734702, "step": 16030}, {"loss": 0.6967, "grad_norm": 0.5846083760261536, "learning_rate": 0.0002, "epoch": 2.593161425915447, "step": 16040}, {"loss": 0.7322, "grad_norm": 0.7323982119560242, "learning_rate": 0.0002, "epoch": 2.594778110096193, "step": 16050}, {"loss": 0.6794, "grad_norm": 0.9069556593894958, "learning_rate": 0.0002, "epoch": 2.596394794276938, "step": 16060}, {"loss": 0.7076, "grad_norm": 0.7522736191749573, "learning_rate": 0.0002, "epoch": 2.5980114784576833, "step": 16070}, {"loss": 0.6477, "grad_norm": 0.8149648308753967, "learning_rate": 0.0002, "epoch": 2.5996281626384286, "step": 16080}, {"loss": 0.6664, "grad_norm": 0.6214233040809631, "learning_rate": 0.0002, "epoch": 2.601244846819174, "step": 16090}, {"loss": 0.7307, "grad_norm": 0.6803743839263916, "learning_rate": 0.0002, "epoch": 2.602861530999919, "step": 16100}, {"loss": 0.7244, "grad_norm": 0.7223997116088867, "learning_rate": 0.0002, "epoch": 2.6044782151806647, "step": 16110}, {"loss": 0.6867, "grad_norm": 0.7324174642562866, "learning_rate": 0.0002, "epoch": 2.60609489936141, "step": 16120}, {"loss": 0.7159, "grad_norm": 0.9594739675521851, "learning_rate": 0.0002, "epoch": 2.607711583542155, "step": 16130}, {"loss": 0.6451, "grad_norm": 0.9485327005386353, "learning_rate": 0.0002, "epoch": 2.6093282677229004, "step": 16140}, {"loss": 0.6815, "grad_norm": 0.8449000120162964, "learning_rate": 0.0002, "epoch": 2.6109449519036456, "step": 16150}, {"loss": 0.7152, "grad_norm": 0.8520140051841736, "learning_rate": 0.0002, "epoch": 2.612561636084391, "step": 16160}, {"loss": 0.6759, "grad_norm": 0.7456524968147278, "learning_rate": 0.0002, "epoch": 2.614178320265136, "step": 16170}, {"loss": 0.6893, "grad_norm": 0.9912857413291931, "learning_rate": 0.0002, "epoch": 2.6157950044458813, "step": 16180}, {"loss": 0.7243, "grad_norm": 0.9001946449279785, "learning_rate": 0.0002, "epoch": 2.6174116886266265, "step": 16190}, {"loss": 0.6825, "grad_norm": 0.6568667888641357, "learning_rate": 0.0002, "epoch": 2.619028372807372, "step": 16200}, {"loss": 0.7013, "grad_norm": 1.0248128175735474, "learning_rate": 0.0002, "epoch": 2.6206450569881174, "step": 16210}, {"loss": 0.7045, "grad_norm": 0.6509039998054504, "learning_rate": 0.0002, "epoch": 2.6222617411688627, "step": 16220}, {"loss": 0.72, "grad_norm": 0.7626351118087769, "learning_rate": 0.0002, "epoch": 2.623878425349608, "step": 16230}, {"loss": 0.6556, "grad_norm": 0.6938552260398865, "learning_rate": 0.0002, "epoch": 2.625495109530353, "step": 16240}, {"loss": 0.65, "grad_norm": 0.6434680819511414, "learning_rate": 0.0002, "epoch": 2.6271117937110984, "step": 16250}, {"loss": 0.6943, "grad_norm": 0.7111515998840332, "learning_rate": 0.0002, "epoch": 2.628728477891844, "step": 16260}, {"loss": 0.679, "grad_norm": 0.7712395787239075, "learning_rate": 0.0002, "epoch": 2.6303451620725893, "step": 16270}, {"loss": 0.6886, "grad_norm": 0.792209267616272, "learning_rate": 0.0002, "epoch": 2.6319618462533345, "step": 16280}, {"loss": 0.6554, "grad_norm": 0.6801066398620605, "learning_rate": 0.0002, "epoch": 2.6335785304340797, "step": 16290}, {"loss": 0.73, "grad_norm": 0.7802573442459106, "learning_rate": 0.0002, "epoch": 2.635195214614825, "step": 16300}, {"loss": 0.7484, "grad_norm": 0.7742244601249695, "learning_rate": 0.0002, "epoch": 2.63681189879557, "step": 16310}, {"loss": 0.6524, "grad_norm": 0.664184033870697, "learning_rate": 0.0002, "epoch": 2.6384285829763154, "step": 16320}, {"loss": 0.6442, "grad_norm": 0.9242228865623474, "learning_rate": 0.0002, "epoch": 2.6400452671570607, "step": 16330}, {"loss": 0.6792, "grad_norm": 0.9661325216293335, "learning_rate": 0.0002, "epoch": 2.641661951337806, "step": 16340}, {"loss": 0.6847, "grad_norm": 0.837526798248291, "learning_rate": 0.0002, "epoch": 2.6432786355185516, "step": 16350}, {"loss": 0.7686, "grad_norm": 1.1834373474121094, "learning_rate": 0.0002, "epoch": 2.644895319699297, "step": 16360}, {"loss": 0.6746, "grad_norm": 0.7467831373214722, "learning_rate": 0.0002, "epoch": 2.646512003880042, "step": 16370}, {"loss": 0.6935, "grad_norm": 0.8627146482467651, "learning_rate": 0.0002, "epoch": 2.6481286880607873, "step": 16380}, {"loss": 0.715, "grad_norm": 0.790447473526001, "learning_rate": 0.0002, "epoch": 2.6497453722415325, "step": 16390}, {"loss": 0.723, "grad_norm": 0.8447365164756775, "learning_rate": 0.0002, "epoch": 2.651362056422278, "step": 16400}, {"loss": 0.6628, "grad_norm": 0.7831417918205261, "learning_rate": 0.0002, "epoch": 2.6529787406030234, "step": 16410}, {"loss": 0.6691, "grad_norm": 0.6837952136993408, "learning_rate": 0.0002, "epoch": 2.6545954247837686, "step": 16420}, {"loss": 0.6139, "grad_norm": 0.7031801342964172, "learning_rate": 0.0002, "epoch": 2.656212108964514, "step": 16430}, {"loss": 0.7382, "grad_norm": 0.8963770866394043, "learning_rate": 0.0002, "epoch": 2.657828793145259, "step": 16440}, {"loss": 0.6439, "grad_norm": 0.6852328181266785, "learning_rate": 0.0002, "epoch": 2.6594454773260043, "step": 16450}, {"loss": 0.6278, "grad_norm": 0.8069294095039368, "learning_rate": 0.0002, "epoch": 2.6610621615067496, "step": 16460}, {"loss": 0.6939, "grad_norm": 0.7503686547279358, "learning_rate": 0.0002, "epoch": 2.662678845687495, "step": 16470}, {"loss": 0.6777, "grad_norm": 0.6430956125259399, "learning_rate": 0.0002, "epoch": 2.66429552986824, "step": 16480}, {"loss": 0.6863, "grad_norm": 0.7894312739372253, "learning_rate": 0.0002, "epoch": 2.6659122140489853, "step": 16490}, {"loss": 0.7165, "grad_norm": 0.7277431488037109, "learning_rate": 0.0002, "epoch": 2.667528898229731, "step": 16500}, {"loss": 0.6772, "grad_norm": 0.6816153526306152, "learning_rate": 0.0002, "epoch": 2.669145582410476, "step": 16510}, {"loss": 0.691, "grad_norm": 0.8145235776901245, "learning_rate": 0.0002, "epoch": 2.6707622665912214, "step": 16520}, {"loss": 0.709, "grad_norm": 0.8645890355110168, "learning_rate": 0.0002, "epoch": 2.6723789507719666, "step": 16530}, {"loss": 0.6946, "grad_norm": 0.704393208026886, "learning_rate": 0.0002, "epoch": 2.673995634952712, "step": 16540}, {"loss": 0.6378, "grad_norm": 1.0120846033096313, "learning_rate": 0.0002, "epoch": 2.6756123191334575, "step": 16550}, {"loss": 0.7241, "grad_norm": 0.6919328570365906, "learning_rate": 0.0002, "epoch": 2.6772290033142028, "step": 16560}, {"loss": 0.7098, "grad_norm": 0.6924574971199036, "learning_rate": 0.0002, "epoch": 2.678845687494948, "step": 16570}, {"loss": 0.731, "grad_norm": 0.9679301381111145, "learning_rate": 0.0002, "epoch": 2.6804623716756932, "step": 16580}, {"loss": 0.7124, "grad_norm": 0.6810211539268494, "learning_rate": 0.0002, "epoch": 2.6820790558564385, "step": 16590}, {"loss": 0.6688, "grad_norm": 0.9730555415153503, "learning_rate": 0.0002, "epoch": 2.6836957400371837, "step": 16600}, {"loss": 0.7344, "grad_norm": 0.7852821350097656, "learning_rate": 0.0002, "epoch": 2.685312424217929, "step": 16610}, {"loss": 0.6401, "grad_norm": 0.6059057116508484, "learning_rate": 0.0002, "epoch": 2.686929108398674, "step": 16620}, {"loss": 0.6796, "grad_norm": 0.9395958781242371, "learning_rate": 0.0002, "epoch": 2.6885457925794194, "step": 16630}, {"loss": 0.7174, "grad_norm": 0.7473729848861694, "learning_rate": 0.0002, "epoch": 2.690162476760165, "step": 16640}, {"loss": 0.7087, "grad_norm": 0.765934407711029, "learning_rate": 0.0002, "epoch": 2.6917791609409103, "step": 16650}, {"loss": 0.707, "grad_norm": 0.8496677279472351, "learning_rate": 0.0002, "epoch": 2.6933958451216555, "step": 16660}, {"loss": 0.7084, "grad_norm": 0.7641879916191101, "learning_rate": 0.0002, "epoch": 2.6950125293024008, "step": 16670}, {"loss": 0.6566, "grad_norm": 0.8471952676773071, "learning_rate": 0.0002, "epoch": 2.696629213483146, "step": 16680}, {"loss": 0.6635, "grad_norm": 0.6946060657501221, "learning_rate": 0.0002, "epoch": 2.6982458976638912, "step": 16690}, {"loss": 0.7027, "grad_norm": 0.7361312508583069, "learning_rate": 0.0002, "epoch": 2.699862581844637, "step": 16700}, {"loss": 0.6767, "grad_norm": 0.6605038046836853, "learning_rate": 0.0002, "epoch": 2.701479266025382, "step": 16710}, {"loss": 0.6885, "grad_norm": 0.7164411544799805, "learning_rate": 0.0002, "epoch": 2.7030959502061274, "step": 16720}, {"loss": 0.6736, "grad_norm": 0.6496201157569885, "learning_rate": 0.0002, "epoch": 2.7047126343868726, "step": 16730}, {"loss": 0.6942, "grad_norm": 0.7826663851737976, "learning_rate": 0.0002, "epoch": 2.706329318567618, "step": 16740}, {"loss": 0.6773, "grad_norm": 0.7639131546020508, "learning_rate": 0.0002, "epoch": 2.707946002748363, "step": 16750}, {"loss": 0.69, "grad_norm": 0.7976210713386536, "learning_rate": 0.0002, "epoch": 2.7095626869291083, "step": 16760}, {"loss": 0.6735, "grad_norm": 0.6836577653884888, "learning_rate": 0.0002, "epoch": 2.7111793711098535, "step": 16770}, {"loss": 0.6596, "grad_norm": 0.8025202751159668, "learning_rate": 0.0002, "epoch": 2.7127960552905988, "step": 16780}, {"loss": 0.6324, "grad_norm": 0.7636463642120361, "learning_rate": 0.0002, "epoch": 2.7144127394713444, "step": 16790}, {"loss": 0.6227, "grad_norm": 0.7481677532196045, "learning_rate": 0.0002, "epoch": 2.7160294236520897, "step": 16800}, {"loss": 0.6925, "grad_norm": 0.7566834688186646, "learning_rate": 0.0002, "epoch": 2.717646107832835, "step": 16810}, {"loss": 0.6531, "grad_norm": 0.7931267619132996, "learning_rate": 0.0002, "epoch": 2.71926279201358, "step": 16820}, {"loss": 0.6672, "grad_norm": 0.8811662197113037, "learning_rate": 0.0002, "epoch": 2.7208794761943254, "step": 16830}, {"loss": 0.6675, "grad_norm": 0.8561240434646606, "learning_rate": 0.0002, "epoch": 2.7224961603750706, "step": 16840}, {"loss": 0.7135, "grad_norm": 0.7121599316596985, "learning_rate": 0.0002, "epoch": 2.7241128445558163, "step": 16850}, {"loss": 0.6825, "grad_norm": 0.8066257238388062, "learning_rate": 0.0002, "epoch": 2.7257295287365615, "step": 16860}, {"loss": 0.6839, "grad_norm": 0.7699271440505981, "learning_rate": 0.0002, "epoch": 2.7273462129173067, "step": 16870}, {"loss": 0.699, "grad_norm": 1.1828432083129883, "learning_rate": 0.0002, "epoch": 2.728962897098052, "step": 16880}, {"loss": 0.6518, "grad_norm": 0.9989302754402161, "learning_rate": 0.0002, "epoch": 2.730579581278797, "step": 16890}, {"loss": 0.7015, "grad_norm": 0.8100560307502747, "learning_rate": 0.0002, "epoch": 2.7321962654595424, "step": 16900}, {"loss": 0.6851, "grad_norm": 0.8615233898162842, "learning_rate": 0.0002, "epoch": 2.7338129496402876, "step": 16910}, {"loss": 0.6322, "grad_norm": 0.8633756041526794, "learning_rate": 0.0002, "epoch": 2.735429633821033, "step": 16920}, {"loss": 0.6488, "grad_norm": 0.7769348621368408, "learning_rate": 0.0002, "epoch": 2.737046318001778, "step": 16930}, {"loss": 0.6582, "grad_norm": 0.6943058371543884, "learning_rate": 0.0002, "epoch": 2.738663002182524, "step": 16940}, {"loss": 0.6516, "grad_norm": 0.8510736227035522, "learning_rate": 0.0002, "epoch": 2.740279686363269, "step": 16950}, {"loss": 0.7275, "grad_norm": 0.7732602953910828, "learning_rate": 0.0002, "epoch": 2.7418963705440142, "step": 16960}, {"loss": 0.6553, "grad_norm": 0.5981788635253906, "learning_rate": 0.0002, "epoch": 2.7435130547247595, "step": 16970}, {"loss": 0.6777, "grad_norm": 0.7604416012763977, "learning_rate": 0.0002, "epoch": 2.7451297389055047, "step": 16980}, {"loss": 0.6981, "grad_norm": 0.7377738356590271, "learning_rate": 0.0002, "epoch": 2.74674642308625, "step": 16990}, {"loss": 0.6294, "grad_norm": 0.9400289058685303, "learning_rate": 0.0002, "epoch": 2.7483631072669956, "step": 17000}, {"loss": 0.6952, "grad_norm": 0.6340599656105042, "learning_rate": 0.0002, "epoch": 2.749979791447741, "step": 17010}, {"loss": 0.7222, "grad_norm": 0.7297601103782654, "learning_rate": 0.0002, "epoch": 2.751596475628486, "step": 17020}, {"loss": 0.6659, "grad_norm": 0.9479979872703552, "learning_rate": 0.0002, "epoch": 2.7532131598092313, "step": 17030}, {"loss": 0.691, "grad_norm": 0.8461511135101318, "learning_rate": 0.0002, "epoch": 2.7548298439899765, "step": 17040}, {"loss": 0.6764, "grad_norm": 0.7477551698684692, "learning_rate": 0.0002, "epoch": 2.7564465281707218, "step": 17050}, {"loss": 0.684, "grad_norm": 1.019270420074463, "learning_rate": 0.0002, "epoch": 2.758063212351467, "step": 17060}, {"loss": 0.7119, "grad_norm": 0.7730235457420349, "learning_rate": 0.0002, "epoch": 2.7596798965322122, "step": 17070}, {"loss": 0.6886, "grad_norm": 0.8216866254806519, "learning_rate": 0.0002, "epoch": 2.7612965807129575, "step": 17080}, {"loss": 0.6811, "grad_norm": 0.7235931754112244, "learning_rate": 0.0002, "epoch": 2.762913264893703, "step": 17090}, {"loss": 0.7031, "grad_norm": 0.7352296710014343, "learning_rate": 0.0002, "epoch": 2.7645299490744484, "step": 17100}, {"loss": 0.6951, "grad_norm": 0.8129373788833618, "learning_rate": 0.0002, "epoch": 2.7661466332551936, "step": 17110}, {"loss": 0.6703, "grad_norm": 0.7387019991874695, "learning_rate": 0.0002, "epoch": 2.767763317435939, "step": 17120}, {"loss": 0.6789, "grad_norm": 0.9149190187454224, "learning_rate": 0.0002, "epoch": 2.769380001616684, "step": 17130}, {"loss": 0.6038, "grad_norm": 0.7352971434593201, "learning_rate": 0.0002, "epoch": 2.7709966857974297, "step": 17140}, {"loss": 0.6728, "grad_norm": 0.7903780341148376, "learning_rate": 0.0002, "epoch": 2.772613369978175, "step": 17150}, {"loss": 0.6988, "grad_norm": 0.8255927562713623, "learning_rate": 0.0002, "epoch": 2.77423005415892, "step": 17160}, {"loss": 0.6694, "grad_norm": 0.7235927581787109, "learning_rate": 0.0002, "epoch": 2.7758467383396654, "step": 17170}, {"loss": 0.7161, "grad_norm": 0.8281434774398804, "learning_rate": 0.0002, "epoch": 2.7774634225204107, "step": 17180}, {"loss": 0.682, "grad_norm": 0.7586921453475952, "learning_rate": 0.0002, "epoch": 2.779080106701156, "step": 17190}, {"loss": 0.6427, "grad_norm": 0.7161715030670166, "learning_rate": 0.0002, "epoch": 2.780696790881901, "step": 17200}, {"loss": 0.6426, "grad_norm": 0.762868344783783, "learning_rate": 0.0002, "epoch": 2.7823134750626464, "step": 17210}, {"loss": 0.705, "grad_norm": 0.9285483360290527, "learning_rate": 0.0002, "epoch": 2.7839301592433916, "step": 17220}, {"loss": 0.7084, "grad_norm": 0.6900462508201599, "learning_rate": 0.0002, "epoch": 2.785546843424137, "step": 17230}, {"loss": 0.6988, "grad_norm": 0.780384361743927, "learning_rate": 0.0002, "epoch": 2.7871635276048825, "step": 17240}, {"loss": 0.7073, "grad_norm": 0.7580406665802002, "learning_rate": 0.0002, "epoch": 2.7887802117856277, "step": 17250}, {"loss": 0.6833, "grad_norm": 0.8145199418067932, "learning_rate": 0.0002, "epoch": 2.790396895966373, "step": 17260}, {"loss": 0.6909, "grad_norm": 0.9159596562385559, "learning_rate": 0.0002, "epoch": 2.792013580147118, "step": 17270}, {"loss": 0.6008, "grad_norm": 0.9590014219284058, "learning_rate": 0.0002, "epoch": 2.7936302643278634, "step": 17280}, {"loss": 0.6704, "grad_norm": 0.7603529691696167, "learning_rate": 0.0002, "epoch": 2.795246948508609, "step": 17290}, {"loss": 0.7165, "grad_norm": 0.8039976358413696, "learning_rate": 0.0002, "epoch": 2.7968636326893543, "step": 17300}, {"loss": 0.7037, "grad_norm": 0.8364847302436829, "learning_rate": 0.0002, "epoch": 2.7984803168700996, "step": 17310}, {"loss": 0.6749, "grad_norm": 0.8763046860694885, "learning_rate": 0.0002, "epoch": 2.800097001050845, "step": 17320}, {"loss": 0.6844, "grad_norm": 0.8409647941589355, "learning_rate": 0.0002, "epoch": 2.80171368523159, "step": 17330}, {"loss": 0.6936, "grad_norm": 0.7649006247520447, "learning_rate": 0.0002, "epoch": 2.8033303694123353, "step": 17340}, {"loss": 0.7051, "grad_norm": 0.7970262169837952, "learning_rate": 0.0002, "epoch": 2.8049470535930805, "step": 17350}, {"loss": 0.6533, "grad_norm": 0.9088607430458069, "learning_rate": 0.0002, "epoch": 2.8065637377738257, "step": 17360}, {"loss": 0.675, "grad_norm": 0.6454846858978271, "learning_rate": 0.0002, "epoch": 2.808180421954571, "step": 17370}, {"loss": 0.7069, "grad_norm": 0.7744787931442261, "learning_rate": 0.0002, "epoch": 2.809797106135316, "step": 17380}, {"loss": 0.6772, "grad_norm": 0.6678640842437744, "learning_rate": 0.0002, "epoch": 2.811413790316062, "step": 17390}, {"loss": 0.6784, "grad_norm": 0.772676944732666, "learning_rate": 0.0002, "epoch": 2.813030474496807, "step": 17400}, {"loss": 0.7252, "grad_norm": 0.7088175415992737, "learning_rate": 0.0002, "epoch": 2.8146471586775523, "step": 17410}, {"loss": 0.7086, "grad_norm": 0.8280573487281799, "learning_rate": 0.0002, "epoch": 2.8162638428582976, "step": 17420}, {"loss": 0.6732, "grad_norm": 0.6665388345718384, "learning_rate": 0.0002, "epoch": 2.817880527039043, "step": 17430}, {"loss": 0.6675, "grad_norm": 0.6427883505821228, "learning_rate": 0.0002, "epoch": 2.8194972112197885, "step": 17440}, {"loss": 0.6972, "grad_norm": 0.9697760343551636, "learning_rate": 0.0002, "epoch": 2.8211138954005337, "step": 17450}, {"loss": 0.6838, "grad_norm": 0.7573966383934021, "learning_rate": 0.0002, "epoch": 2.822730579581279, "step": 17460}, {"loss": 0.7243, "grad_norm": 0.878688633441925, "learning_rate": 0.0002, "epoch": 2.824347263762024, "step": 17470}, {"loss": 0.6666, "grad_norm": 0.7752242684364319, "learning_rate": 0.0002, "epoch": 2.8259639479427694, "step": 17480}, {"loss": 0.6638, "grad_norm": 0.6135398745536804, "learning_rate": 0.0002, "epoch": 2.8275806321235146, "step": 17490}, {"loss": 0.6829, "grad_norm": 0.6924924850463867, "learning_rate": 0.0002, "epoch": 2.82919731630426, "step": 17500}, {"loss": 0.6731, "grad_norm": 0.7471627593040466, "learning_rate": 0.0002, "epoch": 2.830814000485005, "step": 17510}, {"loss": 0.7016, "grad_norm": 0.7145499587059021, "learning_rate": 0.0002, "epoch": 2.8324306846657503, "step": 17520}, {"loss": 0.6787, "grad_norm": 0.7415414452552795, "learning_rate": 0.0002, "epoch": 2.834047368846496, "step": 17530}, {"loss": 0.6811, "grad_norm": 0.7328441739082336, "learning_rate": 0.0002, "epoch": 2.8356640530272412, "step": 17540}, {"loss": 0.6866, "grad_norm": 0.8267839550971985, "learning_rate": 0.0002, "epoch": 2.8372807372079865, "step": 17550}, {"loss": 0.6787, "grad_norm": 0.8877885341644287, "learning_rate": 0.0002, "epoch": 2.8388974213887317, "step": 17560}, {"loss": 0.7136, "grad_norm": 0.857138454914093, "learning_rate": 0.0002, "epoch": 2.840514105569477, "step": 17570}, {"loss": 0.6454, "grad_norm": 0.8470779657363892, "learning_rate": 0.0002, "epoch": 2.842130789750222, "step": 17580}, {"loss": 0.6976, "grad_norm": 0.8553254008293152, "learning_rate": 0.0002, "epoch": 2.843747473930968, "step": 17590}, {"loss": 0.7297, "grad_norm": 0.8033196926116943, "learning_rate": 0.0002, "epoch": 2.845364158111713, "step": 17600}, {"loss": 0.7062, "grad_norm": 0.7949087023735046, "learning_rate": 0.0002, "epoch": 2.8469808422924583, "step": 17610}, {"loss": 0.651, "grad_norm": 0.9241406321525574, "learning_rate": 0.0002, "epoch": 2.8485975264732035, "step": 17620}, {"loss": 0.6601, "grad_norm": 0.7721285223960876, "learning_rate": 0.0002, "epoch": 2.8502142106539488, "step": 17630}, {"loss": 0.6183, "grad_norm": 1.0246692895889282, "learning_rate": 0.0002, "epoch": 2.851830894834694, "step": 17640}, {"loss": 0.7007, "grad_norm": 0.9244589805603027, "learning_rate": 0.0002, "epoch": 2.853447579015439, "step": 17650}, {"loss": 0.7274, "grad_norm": 0.7243508696556091, "learning_rate": 0.0002, "epoch": 2.8550642631961844, "step": 17660}, {"loss": 0.6471, "grad_norm": 0.8943371176719666, "learning_rate": 0.0002, "epoch": 2.8566809473769297, "step": 17670}, {"loss": 0.686, "grad_norm": 0.6531758904457092, "learning_rate": 0.0002, "epoch": 2.8582976315576754, "step": 17680}, {"loss": 0.6253, "grad_norm": 0.8367000818252563, "learning_rate": 0.0002, "epoch": 2.8599143157384206, "step": 17690}, {"loss": 0.6943, "grad_norm": 0.7868556380271912, "learning_rate": 0.0002, "epoch": 2.861530999919166, "step": 17700}, {"loss": 0.6919, "grad_norm": 0.7213859558105469, "learning_rate": 0.0002, "epoch": 2.863147684099911, "step": 17710}, {"loss": 0.6657, "grad_norm": 0.7383931279182434, "learning_rate": 0.0002, "epoch": 2.8647643682806563, "step": 17720}, {"loss": 0.6841, "grad_norm": 0.7566812634468079, "learning_rate": 0.0002, "epoch": 2.8663810524614015, "step": 17730}, {"loss": 0.6449, "grad_norm": 0.6930373311042786, "learning_rate": 0.0002, "epoch": 2.867997736642147, "step": 17740}, {"loss": 0.6764, "grad_norm": 0.7911090850830078, "learning_rate": 0.0002, "epoch": 2.8696144208228924, "step": 17750}, {"loss": 0.6554, "grad_norm": 0.8484548926353455, "learning_rate": 0.0002, "epoch": 2.8712311050036377, "step": 17760}, {"loss": 0.6931, "grad_norm": 0.7647597193717957, "learning_rate": 0.0002, "epoch": 2.872847789184383, "step": 17770}, {"loss": 0.6945, "grad_norm": 0.8791151642799377, "learning_rate": 0.0002, "epoch": 2.874464473365128, "step": 17780}, {"loss": 0.7078, "grad_norm": 0.7253178358078003, "learning_rate": 0.0002, "epoch": 2.8760811575458733, "step": 17790}, {"loss": 0.6474, "grad_norm": 0.7956077456474304, "learning_rate": 0.0002, "epoch": 2.8776978417266186, "step": 17800}, {"loss": 0.6687, "grad_norm": 0.8657688498497009, "learning_rate": 0.0002, "epoch": 2.879314525907364, "step": 17810}, {"loss": 0.7171, "grad_norm": 0.7059141993522644, "learning_rate": 0.0002, "epoch": 2.880931210088109, "step": 17820}, {"loss": 0.683, "grad_norm": 0.8886896967887878, "learning_rate": 0.0002, "epoch": 2.8825478942688547, "step": 17830}, {"loss": 0.669, "grad_norm": 0.821032702922821, "learning_rate": 0.0002, "epoch": 2.8841645784496, "step": 17840}, {"loss": 0.6805, "grad_norm": 0.7183963656425476, "learning_rate": 0.0002, "epoch": 2.885781262630345, "step": 17850}, {"loss": 0.7088, "grad_norm": 0.6222899556159973, "learning_rate": 0.0002, "epoch": 2.8873979468110904, "step": 17860}, {"loss": 0.6626, "grad_norm": 0.8187434077262878, "learning_rate": 0.0002, "epoch": 2.8890146309918356, "step": 17870}, {"loss": 0.6815, "grad_norm": 0.9838479161262512, "learning_rate": 0.0002, "epoch": 2.890631315172581, "step": 17880}, {"loss": 0.6967, "grad_norm": 0.7567742466926575, "learning_rate": 0.0002, "epoch": 2.8922479993533265, "step": 17890}, {"loss": 0.7073, "grad_norm": 0.6875903606414795, "learning_rate": 0.0002, "epoch": 2.893864683534072, "step": 17900}, {"loss": 0.6415, "grad_norm": 0.8043789267539978, "learning_rate": 0.0002, "epoch": 2.895481367714817, "step": 17910}, {"loss": 0.6588, "grad_norm": 0.8062626719474792, "learning_rate": 0.0002, "epoch": 2.8970980518955622, "step": 17920}, {"loss": 0.7151, "grad_norm": 1.0251191854476929, "learning_rate": 0.0002, "epoch": 2.8987147360763075, "step": 17930}, {"loss": 0.6605, "grad_norm": 0.882253110408783, "learning_rate": 0.0002, "epoch": 2.9003314202570527, "step": 17940}, {"loss": 0.6719, "grad_norm": 0.8683299422264099, "learning_rate": 0.0002, "epoch": 2.901948104437798, "step": 17950}, {"loss": 0.6896, "grad_norm": 0.7167282104492188, "learning_rate": 0.0002, "epoch": 2.903564788618543, "step": 17960}, {"loss": 0.663, "grad_norm": 0.7093694806098938, "learning_rate": 0.0002, "epoch": 2.9051814727992884, "step": 17970}, {"loss": 0.6591, "grad_norm": 0.8549879193305969, "learning_rate": 0.0002, "epoch": 2.906798156980034, "step": 17980}, {"loss": 0.6962, "grad_norm": 0.6989606618881226, "learning_rate": 0.0002, "epoch": 2.9084148411607793, "step": 17990}, {"loss": 0.6635, "grad_norm": 0.9482976794242859, "learning_rate": 0.0002, "epoch": 2.9100315253415245, "step": 18000}, {"loss": 0.6586, "grad_norm": 0.7182440161705017, "learning_rate": 0.0002, "epoch": 2.9116482095222698, "step": 18010}, {"loss": 0.6827, "grad_norm": 0.7732226252555847, "learning_rate": 0.0002, "epoch": 2.913264893703015, "step": 18020}, {"loss": 0.7123, "grad_norm": 0.7936875224113464, "learning_rate": 0.0002, "epoch": 2.9148815778837607, "step": 18030}, {"loss": 0.6736, "grad_norm": 0.8825615644454956, "learning_rate": 0.0002, "epoch": 2.916498262064506, "step": 18040}, {"loss": 0.7139, "grad_norm": 0.6778587102890015, "learning_rate": 0.0002, "epoch": 2.918114946245251, "step": 18050}, {"loss": 0.6588, "grad_norm": 0.7529265880584717, "learning_rate": 0.0002, "epoch": 2.9197316304259964, "step": 18060}, {"loss": 0.737, "grad_norm": 0.7111883163452148, "learning_rate": 0.0002, "epoch": 2.9213483146067416, "step": 18070}, {"loss": 0.7475, "grad_norm": 0.7214767932891846, "learning_rate": 0.0002, "epoch": 2.922964998787487, "step": 18080}, {"loss": 0.6672, "grad_norm": 0.800417423248291, "learning_rate": 0.0002, "epoch": 2.924581682968232, "step": 18090}, {"loss": 0.6694, "grad_norm": 1.248575210571289, "learning_rate": 0.0002, "epoch": 2.9261983671489773, "step": 18100}, {"loss": 0.7004, "grad_norm": 0.757788360118866, "learning_rate": 0.0002, "epoch": 2.9278150513297225, "step": 18110}, {"loss": 0.6999, "grad_norm": 1.0583995580673218, "learning_rate": 0.0002, "epoch": 2.9294317355104678, "step": 18120}, {"loss": 0.6365, "grad_norm": 0.8228777647018433, "learning_rate": 0.0002, "epoch": 2.9310484196912134, "step": 18130}, {"loss": 0.6791, "grad_norm": 0.8374035358428955, "learning_rate": 0.0002, "epoch": 2.9326651038719587, "step": 18140}, {"loss": 0.6399, "grad_norm": 0.7976473569869995, "learning_rate": 0.0002, "epoch": 2.934281788052704, "step": 18150}, {"loss": 0.6585, "grad_norm": 0.8009907603263855, "learning_rate": 0.0002, "epoch": 2.935898472233449, "step": 18160}, {"loss": 0.7485, "grad_norm": 0.835213303565979, "learning_rate": 0.0002, "epoch": 2.9375151564141944, "step": 18170}, {"loss": 0.7376, "grad_norm": 0.7982219457626343, "learning_rate": 0.0002, "epoch": 2.93913184059494, "step": 18180}, {"loss": 0.6348, "grad_norm": 0.7070978879928589, "learning_rate": 0.0002, "epoch": 2.9407485247756853, "step": 18190}, {"loss": 0.6608, "grad_norm": 0.8619440197944641, "learning_rate": 0.0002, "epoch": 2.9423652089564305, "step": 18200}, {"loss": 0.666, "grad_norm": 0.6693987250328064, "learning_rate": 0.0002, "epoch": 2.9439818931371757, "step": 18210}, {"loss": 0.728, "grad_norm": 0.6747021079063416, "learning_rate": 0.0002, "epoch": 2.945598577317921, "step": 18220}, {"loss": 0.6686, "grad_norm": 0.860387921333313, "learning_rate": 0.0002, "epoch": 2.947215261498666, "step": 18230}, {"loss": 0.6945, "grad_norm": 0.799976646900177, "learning_rate": 0.0002, "epoch": 2.9488319456794114, "step": 18240}, {"loss": 0.7243, "grad_norm": 0.7864769101142883, "learning_rate": 0.0002, "epoch": 2.9504486298601567, "step": 18250}, {"loss": 0.6785, "grad_norm": 0.6713884472846985, "learning_rate": 0.0002, "epoch": 2.952065314040902, "step": 18260}, {"loss": 0.7429, "grad_norm": 0.9031508564949036, "learning_rate": 0.0002, "epoch": 2.9536819982216476, "step": 18270}, {"loss": 0.7055, "grad_norm": 0.7205073237419128, "learning_rate": 0.0002, "epoch": 2.955298682402393, "step": 18280}, {"loss": 0.7298, "grad_norm": 0.7746205925941467, "learning_rate": 0.0002, "epoch": 2.956915366583138, "step": 18290}, {"loss": 0.6218, "grad_norm": 0.6533427834510803, "learning_rate": 0.0002, "epoch": 2.9585320507638833, "step": 18300}, {"loss": 0.6674, "grad_norm": 0.9083208441734314, "learning_rate": 0.0002, "epoch": 2.9601487349446285, "step": 18310}, {"loss": 0.7359, "grad_norm": 0.7446991801261902, "learning_rate": 0.0002, "epoch": 2.9617654191253737, "step": 18320}, {"loss": 0.6738, "grad_norm": 0.6514461636543274, "learning_rate": 0.0002, "epoch": 2.9633821033061194, "step": 18330}, {"loss": 0.6677, "grad_norm": 0.8580465912818909, "learning_rate": 0.0002, "epoch": 2.9649987874868646, "step": 18340}, {"loss": 0.6971, "grad_norm": 0.7074266076087952, "learning_rate": 0.0002, "epoch": 2.96661547166761, "step": 18350}, {"loss": 0.6804, "grad_norm": 0.899892270565033, "learning_rate": 0.0002, "epoch": 2.968232155848355, "step": 18360}, {"loss": 0.7094, "grad_norm": 0.8217641711235046, "learning_rate": 0.0002, "epoch": 2.9698488400291003, "step": 18370}, {"loss": 0.6916, "grad_norm": 0.8611799478530884, "learning_rate": 0.0002, "epoch": 2.9714655242098456, "step": 18380}, {"loss": 0.6677, "grad_norm": 0.6909302473068237, "learning_rate": 0.0002, "epoch": 2.973082208390591, "step": 18390}, {"loss": 0.7247, "grad_norm": 0.6554358005523682, "learning_rate": 0.0002, "epoch": 2.974698892571336, "step": 18400}, {"loss": 0.6516, "grad_norm": 0.7803071737289429, "learning_rate": 0.0002, "epoch": 2.9763155767520812, "step": 18410}, {"loss": 0.7322, "grad_norm": 0.7838954925537109, "learning_rate": 0.0002, "epoch": 2.977932260932827, "step": 18420}, {"loss": 0.6522, "grad_norm": 0.7098495364189148, "learning_rate": 0.0002, "epoch": 2.979548945113572, "step": 18430}, {"loss": 0.739, "grad_norm": 0.8981785774230957, "learning_rate": 0.0002, "epoch": 2.9811656292943174, "step": 18440}, {"loss": 0.6689, "grad_norm": 0.7197171449661255, "learning_rate": 0.0002, "epoch": 2.9827823134750626, "step": 18450}, {"loss": 0.706, "grad_norm": 0.793185293674469, "learning_rate": 0.0002, "epoch": 2.984398997655808, "step": 18460}, {"loss": 0.7124, "grad_norm": 0.8531473875045776, "learning_rate": 0.0002, "epoch": 2.986015681836553, "step": 18470}, {"loss": 0.6901, "grad_norm": 0.6627361178398132, "learning_rate": 0.0002, "epoch": 2.9876323660172988, "step": 18480}, {"loss": 0.6591, "grad_norm": 0.5708155035972595, "learning_rate": 0.0002, "epoch": 2.989249050198044, "step": 18490}, {"loss": 0.6725, "grad_norm": 0.8227280378341675, "learning_rate": 0.0002, "epoch": 2.990865734378789, "step": 18500}, {"loss": 0.6701, "grad_norm": 0.7102749943733215, "learning_rate": 0.0002, "epoch": 2.9924824185595345, "step": 18510}, {"loss": 0.7091, "grad_norm": 0.839485228061676, "learning_rate": 0.0002, "epoch": 2.9940991027402797, "step": 18520}, {"loss": 0.6521, "grad_norm": 0.9038704037666321, "learning_rate": 0.0002, "epoch": 2.995715786921025, "step": 18530}, {"loss": 0.7186, "grad_norm": 0.8737510442733765, "learning_rate": 0.0002, "epoch": 2.99733247110177, "step": 18540}, {"loss": 0.6819, "grad_norm": 0.7323142886161804, "learning_rate": 0.0002, "epoch": 2.9989491552825154, "step": 18550}, {"eval_loss": 1.1262480020523071, "eval_runtime": 122.0868, "eval_samples_per_second": 6.004, "eval_steps_per_second": 0.754, "epoch": 2.9999191657909625, "step": 18556}, {"loss": 0.6337, "grad_norm": 0.8465463519096375, "learning_rate": 0.0002, "epoch": 3.000565839463261, "step": 18560}, {"loss": 0.6064, "grad_norm": 0.9134138822555542, "learning_rate": 0.0002, "epoch": 3.0021825236440063, "step": 18570}, {"loss": 0.5804, "grad_norm": 0.760715126991272, "learning_rate": 0.0002, "epoch": 3.0037992078247515, "step": 18580}, {"loss": 0.5571, "grad_norm": 0.9208743572235107, "learning_rate": 0.0002, "epoch": 3.0054158920054967, "step": 18590}, {"loss": 0.5731, "grad_norm": 0.9232364892959595, "learning_rate": 0.0002, "epoch": 3.007032576186242, "step": 18600}, {"loss": 0.6299, "grad_norm": 1.1881544589996338, "learning_rate": 0.0002, "epoch": 3.008649260366987, "step": 18610}, {"loss": 0.5482, "grad_norm": 0.9372987747192383, "learning_rate": 0.0002, "epoch": 3.0102659445477324, "step": 18620}, {"loss": 0.5709, "grad_norm": 0.6900241374969482, "learning_rate": 0.0002, "epoch": 3.0118826287284777, "step": 18630}, {"loss": 0.5256, "grad_norm": 0.8451071381568909, "learning_rate": 0.0002, "epoch": 3.0134993129092233, "step": 18640}, {"loss": 0.5916, "grad_norm": 0.7763112187385559, "learning_rate": 0.0002, "epoch": 3.0151159970899686, "step": 18650}, {"loss": 0.6095, "grad_norm": 1.043653964996338, "learning_rate": 0.0002, "epoch": 3.016732681270714, "step": 18660}, {"loss": 0.6228, "grad_norm": 1.0170660018920898, "learning_rate": 0.0002, "epoch": 3.018349365451459, "step": 18670}, {"loss": 0.5671, "grad_norm": 0.7534180283546448, "learning_rate": 0.0002, "epoch": 3.0199660496322043, "step": 18680}, {"loss": 0.6015, "grad_norm": 0.7507367730140686, "learning_rate": 0.0002, "epoch": 3.0215827338129495, "step": 18690}, {"loss": 0.6201, "grad_norm": 0.7861620187759399, "learning_rate": 0.0002, "epoch": 3.0231994179936947, "step": 18700}, {"loss": 0.5802, "grad_norm": 1.0580339431762695, "learning_rate": 0.0002, "epoch": 3.0248161021744404, "step": 18710}, {"loss": 0.5975, "grad_norm": 0.7542710900306702, "learning_rate": 0.0002, "epoch": 3.0264327863551856, "step": 18720}, {"loss": 0.5695, "grad_norm": 0.8189544677734375, "learning_rate": 0.0002, "epoch": 3.028049470535931, "step": 18730}, {"loss": 0.6109, "grad_norm": 0.9126611351966858, "learning_rate": 0.0002, "epoch": 3.029666154716676, "step": 18740}, {"loss": 0.6443, "grad_norm": 0.8891341686248779, "learning_rate": 0.0002, "epoch": 3.0312828388974213, "step": 18750}, {"loss": 0.6207, "grad_norm": 0.8419283032417297, "learning_rate": 0.0002, "epoch": 3.0328995230781666, "step": 18760}, {"loss": 0.5818, "grad_norm": 0.8048048615455627, "learning_rate": 0.0002, "epoch": 3.034516207258912, "step": 18770}, {"loss": 0.6381, "grad_norm": 0.7820217609405518, "learning_rate": 0.0002, "epoch": 3.0361328914396575, "step": 18780}, {"loss": 0.5843, "grad_norm": 0.854721188545227, "learning_rate": 0.0002, "epoch": 3.0377495756204027, "step": 18790}, {"loss": 0.5784, "grad_norm": 0.912092924118042, "learning_rate": 0.0002, "epoch": 3.039366259801148, "step": 18800}, {"loss": 0.5734, "grad_norm": 0.6596226096153259, "learning_rate": 0.0002, "epoch": 3.040982943981893, "step": 18810}, {"loss": 0.5969, "grad_norm": 0.6351348757743835, "learning_rate": 0.0002, "epoch": 3.0425996281626384, "step": 18820}, {"loss": 0.5953, "grad_norm": 0.778188943862915, "learning_rate": 0.0002, "epoch": 3.0442163123433836, "step": 18830}, {"loss": 0.602, "grad_norm": 0.68234783411026, "learning_rate": 0.0002, "epoch": 3.045832996524129, "step": 18840}, {"loss": 0.5785, "grad_norm": 0.998628556728363, "learning_rate": 0.0002, "epoch": 3.047449680704874, "step": 18850}, {"loss": 0.6231, "grad_norm": 0.7393841743469238, "learning_rate": 0.0002, "epoch": 3.0490663648856198, "step": 18860}, {"loss": 0.568, "grad_norm": 0.84438556432724, "learning_rate": 0.0002, "epoch": 3.050683049066365, "step": 18870}, {"loss": 0.6205, "grad_norm": 0.8857501745223999, "learning_rate": 0.0002, "epoch": 3.0522997332471102, "step": 18880}, {"loss": 0.6335, "grad_norm": 0.7208474278450012, "learning_rate": 0.0002, "epoch": 3.0539164174278555, "step": 18890}, {"loss": 0.5998, "grad_norm": 0.7135229110717773, "learning_rate": 0.0002, "epoch": 3.0555331016086007, "step": 18900}, {"loss": 0.5575, "grad_norm": 0.9130001664161682, "learning_rate": 0.0002, "epoch": 3.057149785789346, "step": 18910}, {"loss": 0.5955, "grad_norm": 0.9001716375350952, "learning_rate": 0.0002, "epoch": 3.058766469970091, "step": 18920}, {"loss": 0.6052, "grad_norm": 0.8667559623718262, "learning_rate": 0.0002, "epoch": 3.060383154150837, "step": 18930}, {"loss": 0.5818, "grad_norm": 0.8943959474563599, "learning_rate": 0.0002, "epoch": 3.061999838331582, "step": 18940}, {"loss": 0.5978, "grad_norm": 0.8298377990722656, "learning_rate": 0.0002, "epoch": 3.0636165225123273, "step": 18950}, {"loss": 0.5782, "grad_norm": 0.7935267686843872, "learning_rate": 0.0002, "epoch": 3.0652332066930725, "step": 18960}, {"loss": 0.6434, "grad_norm": 1.1506379842758179, "learning_rate": 0.0002, "epoch": 3.0668498908738178, "step": 18970}, {"loss": 0.5571, "grad_norm": 0.7693049907684326, "learning_rate": 0.0002, "epoch": 3.068466575054563, "step": 18980}, {"loss": 0.5971, "grad_norm": 0.8040135502815247, "learning_rate": 0.0002, "epoch": 3.0700832592353082, "step": 18990}, {"loss": 0.5541, "grad_norm": 0.828404426574707, "learning_rate": 0.0002, "epoch": 3.0716999434160535, "step": 19000}, {"loss": 0.6048, "grad_norm": 0.8811164498329163, "learning_rate": 0.0002, "epoch": 3.073316627596799, "step": 19010}, {"loss": 0.5845, "grad_norm": 1.036205768585205, "learning_rate": 0.0002, "epoch": 3.0749333117775444, "step": 19020}, {"loss": 0.5838, "grad_norm": 0.8857285976409912, "learning_rate": 0.0002, "epoch": 3.0765499959582896, "step": 19030}, {"loss": 0.592, "grad_norm": 0.8392079472541809, "learning_rate": 0.0002, "epoch": 3.078166680139035, "step": 19040}, {"loss": 0.5927, "grad_norm": 1.0287401676177979, "learning_rate": 0.0002, "epoch": 3.07978336431978, "step": 19050}, {"loss": 0.5964, "grad_norm": 1.0086315870285034, "learning_rate": 0.0002, "epoch": 3.0814000485005253, "step": 19060}, {"loss": 0.5567, "grad_norm": 0.9245324730873108, "learning_rate": 0.0002, "epoch": 3.0830167326812705, "step": 19070}, {"loss": 0.5797, "grad_norm": 0.8680877089500427, "learning_rate": 0.0002, "epoch": 3.084633416862016, "step": 19080}, {"loss": 0.5611, "grad_norm": 0.8814793825149536, "learning_rate": 0.0002, "epoch": 3.0862501010427614, "step": 19090}, {"loss": 0.6051, "grad_norm": 0.9234458208084106, "learning_rate": 0.0002, "epoch": 3.0878667852235067, "step": 19100}, {"loss": 0.6209, "grad_norm": 1.1291664838790894, "learning_rate": 0.0002, "epoch": 3.089483469404252, "step": 19110}, {"loss": 0.5695, "grad_norm": 0.9191402792930603, "learning_rate": 0.0002, "epoch": 3.091100153584997, "step": 19120}, {"loss": 0.5856, "grad_norm": 0.7103154063224792, "learning_rate": 0.0002, "epoch": 3.0927168377657424, "step": 19130}, {"loss": 0.6479, "grad_norm": 0.9368883967399597, "learning_rate": 0.0002, "epoch": 3.0943335219464876, "step": 19140}, {"loss": 0.6167, "grad_norm": 0.9676656723022461, "learning_rate": 0.0002, "epoch": 3.095950206127233, "step": 19150}, {"loss": 0.5794, "grad_norm": 0.8739792704582214, "learning_rate": 0.0002, "epoch": 3.0975668903079785, "step": 19160}, {"loss": 0.6112, "grad_norm": 0.8530174493789673, "learning_rate": 0.0002, "epoch": 3.0991835744887237, "step": 19170}, {"loss": 0.6568, "grad_norm": 0.794945478439331, "learning_rate": 0.0002, "epoch": 3.100800258669469, "step": 19180}, {"loss": 0.5928, "grad_norm": 0.9508888125419617, "learning_rate": 0.0002, "epoch": 3.102416942850214, "step": 19190}, {"loss": 0.5757, "grad_norm": 1.0599955320358276, "learning_rate": 0.0002, "epoch": 3.1040336270309594, "step": 19200}, {"loss": 0.6151, "grad_norm": 1.0673625469207764, "learning_rate": 0.0002, "epoch": 3.1056503112117047, "step": 19210}, {"loss": 0.6043, "grad_norm": 0.7739115953445435, "learning_rate": 0.0002, "epoch": 3.10726699539245, "step": 19220}, {"loss": 0.6046, "grad_norm": 0.9884951114654541, "learning_rate": 0.0002, "epoch": 3.1088836795731956, "step": 19230}, {"loss": 0.5932, "grad_norm": 0.862260103225708, "learning_rate": 0.0002, "epoch": 3.110500363753941, "step": 19240}, {"loss": 0.6098, "grad_norm": 0.7690284848213196, "learning_rate": 0.0002, "epoch": 3.112117047934686, "step": 19250}, {"loss": 0.5791, "grad_norm": 0.8758958578109741, "learning_rate": 0.0002, "epoch": 3.1137337321154313, "step": 19260}, {"loss": 0.6136, "grad_norm": 1.0356395244598389, "learning_rate": 0.0002, "epoch": 3.1153504162961765, "step": 19270}, {"loss": 0.6159, "grad_norm": 0.6950937509536743, "learning_rate": 0.0002, "epoch": 3.1169671004769217, "step": 19280}, {"loss": 0.592, "grad_norm": 0.760998010635376, "learning_rate": 0.0002, "epoch": 3.118583784657667, "step": 19290}, {"loss": 0.575, "grad_norm": 0.9335789084434509, "learning_rate": 0.0002, "epoch": 3.1202004688384126, "step": 19300}, {"loss": 0.6139, "grad_norm": 0.9636204242706299, "learning_rate": 0.0002, "epoch": 3.121817153019158, "step": 19310}, {"loss": 0.6001, "grad_norm": 1.0820997953414917, "learning_rate": 0.0002, "epoch": 3.123433837199903, "step": 19320}, {"loss": 0.6542, "grad_norm": 0.7333487272262573, "learning_rate": 0.0002, "epoch": 3.1250505213806483, "step": 19330}, {"loss": 0.6178, "grad_norm": 1.0417509078979492, "learning_rate": 0.0002, "epoch": 3.1266672055613935, "step": 19340}, {"loss": 0.603, "grad_norm": 0.9267749190330505, "learning_rate": 0.0002, "epoch": 3.128283889742139, "step": 19350}, {"loss": 0.6063, "grad_norm": 0.777798593044281, "learning_rate": 0.0002, "epoch": 3.129900573922884, "step": 19360}, {"loss": 0.5913, "grad_norm": 0.8425456881523132, "learning_rate": 0.0002, "epoch": 3.1315172581036297, "step": 19370}, {"loss": 0.6042, "grad_norm": 0.9617102146148682, "learning_rate": 0.0002, "epoch": 3.133133942284375, "step": 19380}, {"loss": 0.633, "grad_norm": 1.0052828788757324, "learning_rate": 0.0002, "epoch": 3.13475062646512, "step": 19390}, {"loss": 0.5713, "grad_norm": 0.7637009024620056, "learning_rate": 0.0002, "epoch": 3.1363673106458654, "step": 19400}, {"loss": 0.5497, "grad_norm": 0.7958088517189026, "learning_rate": 0.0002, "epoch": 3.1379839948266106, "step": 19410}, {"loss": 0.6283, "grad_norm": 0.9161727428436279, "learning_rate": 0.0002, "epoch": 3.139600679007356, "step": 19420}, {"loss": 0.5638, "grad_norm": 0.8402149677276611, "learning_rate": 0.0002, "epoch": 3.141217363188101, "step": 19430}, {"loss": 0.5848, "grad_norm": 1.0056525468826294, "learning_rate": 0.0002, "epoch": 3.1428340473688463, "step": 19440}, {"loss": 0.5954, "grad_norm": 1.0129190683364868, "learning_rate": 0.0002, "epoch": 3.144450731549592, "step": 19450}, {"loss": 0.5808, "grad_norm": 0.790825366973877, "learning_rate": 0.0002, "epoch": 3.146067415730337, "step": 19460}, {"loss": 0.5607, "grad_norm": 1.441665530204773, "learning_rate": 0.0002, "epoch": 3.1476840999110824, "step": 19470}, {"loss": 0.5785, "grad_norm": 0.7846331596374512, "learning_rate": 0.0002, "epoch": 3.1493007840918277, "step": 19480}, {"loss": 0.5892, "grad_norm": 0.7915332913398743, "learning_rate": 0.0002, "epoch": 3.150917468272573, "step": 19490}, {"loss": 0.5759, "grad_norm": 0.933982253074646, "learning_rate": 0.0002, "epoch": 3.152534152453318, "step": 19500}, {"loss": 0.6206, "grad_norm": 1.038408637046814, "learning_rate": 0.0002, "epoch": 3.1541508366340634, "step": 19510}, {"loss": 0.6271, "grad_norm": 1.018935203552246, "learning_rate": 0.0002, "epoch": 3.155767520814809, "step": 19520}, {"loss": 0.6173, "grad_norm": 0.9618112444877625, "learning_rate": 0.0002, "epoch": 3.1573842049955543, "step": 19530}, {"loss": 0.5972, "grad_norm": 0.8900452852249146, "learning_rate": 0.0002, "epoch": 3.1590008891762995, "step": 19540}, {"loss": 0.5925, "grad_norm": 0.8254160284996033, "learning_rate": 0.0002, "epoch": 3.1606175733570447, "step": 19550}, {"loss": 0.625, "grad_norm": 1.004376769065857, "learning_rate": 0.0002, "epoch": 3.16223425753779, "step": 19560}, {"loss": 0.5775, "grad_norm": 1.0490446090698242, "learning_rate": 0.0002, "epoch": 3.163850941718535, "step": 19570}, {"loss": 0.5986, "grad_norm": 0.7387403845787048, "learning_rate": 0.0002, "epoch": 3.1654676258992804, "step": 19580}, {"loss": 0.5898, "grad_norm": 0.7611538171768188, "learning_rate": 0.0002, "epoch": 3.1670843100800257, "step": 19590}, {"loss": 0.5937, "grad_norm": 0.8239886164665222, "learning_rate": 0.0002, "epoch": 3.1687009942607713, "step": 19600}, {"loss": 0.6068, "grad_norm": 0.9327243566513062, "learning_rate": 0.0002, "epoch": 3.1703176784415166, "step": 19610}, {"loss": 0.572, "grad_norm": 0.9662560224533081, "learning_rate": 0.0002, "epoch": 3.171934362622262, "step": 19620}, {"loss": 0.5988, "grad_norm": 0.9183341860771179, "learning_rate": 0.0002, "epoch": 3.173551046803007, "step": 19630}, {"loss": 0.5909, "grad_norm": 0.875066876411438, "learning_rate": 0.0002, "epoch": 3.1751677309837523, "step": 19640}, {"loss": 0.5956, "grad_norm": 0.8567508459091187, "learning_rate": 0.0002, "epoch": 3.1767844151644975, "step": 19650}, {"loss": 0.5805, "grad_norm": 0.6805780529975891, "learning_rate": 0.0002, "epoch": 3.1784010993452427, "step": 19660}, {"loss": 0.6204, "grad_norm": 0.8776944279670715, "learning_rate": 0.0002, "epoch": 3.1800177835259884, "step": 19670}, {"loss": 0.6108, "grad_norm": 0.9036329984664917, "learning_rate": 0.0002, "epoch": 3.1816344677067336, "step": 19680}, {"loss": 0.6238, "grad_norm": 0.8527372479438782, "learning_rate": 0.0002, "epoch": 3.183251151887479, "step": 19690}, {"loss": 0.6089, "grad_norm": 1.1045585870742798, "learning_rate": 0.0002, "epoch": 3.184867836068224, "step": 19700}, {"loss": 0.5491, "grad_norm": 0.9213830828666687, "learning_rate": 0.0002, "epoch": 3.1864845202489693, "step": 19710}, {"loss": 0.618, "grad_norm": 0.8865814805030823, "learning_rate": 0.0002, "epoch": 3.1881012044297146, "step": 19720}, {"loss": 0.5785, "grad_norm": 0.7939388751983643, "learning_rate": 0.0002, "epoch": 3.18971788861046, "step": 19730}, {"loss": 0.5682, "grad_norm": 0.6966729760169983, "learning_rate": 0.0002, "epoch": 3.191334572791205, "step": 19740}, {"loss": 0.5839, "grad_norm": 0.8023673295974731, "learning_rate": 0.0002, "epoch": 3.1929512569719507, "step": 19750}, {"loss": 0.6267, "grad_norm": 0.7992037534713745, "learning_rate": 0.0002, "epoch": 3.194567941152696, "step": 19760}, {"loss": 0.6141, "grad_norm": 0.7412247657775879, "learning_rate": 0.0002, "epoch": 3.196184625333441, "step": 19770}, {"loss": 0.6179, "grad_norm": 0.9598729014396667, "learning_rate": 0.0002, "epoch": 3.1978013095141864, "step": 19780}, {"loss": 0.5685, "grad_norm": 0.8331366777420044, "learning_rate": 0.0002, "epoch": 3.1994179936949316, "step": 19790}, {"loss": 0.6104, "grad_norm": 0.8939169645309448, "learning_rate": 0.0002, "epoch": 3.201034677875677, "step": 19800}, {"loss": 0.6147, "grad_norm": 0.9219734072685242, "learning_rate": 0.0002, "epoch": 3.202651362056422, "step": 19810}, {"loss": 0.6051, "grad_norm": 0.869490385055542, "learning_rate": 0.0002, "epoch": 3.2042680462371678, "step": 19820}, {"loss": 0.5946, "grad_norm": 0.8989706635475159, "learning_rate": 0.0002, "epoch": 3.205884730417913, "step": 19830}, {"loss": 0.5866, "grad_norm": 0.8477165102958679, "learning_rate": 0.0002, "epoch": 3.2075014145986582, "step": 19840}, {"loss": 0.6176, "grad_norm": 0.8720678687095642, "learning_rate": 0.0002, "epoch": 3.2091180987794035, "step": 19850}, {"loss": 0.5694, "grad_norm": 0.861406683921814, "learning_rate": 0.0002, "epoch": 3.2107347829601487, "step": 19860}, {"loss": 0.6264, "grad_norm": 0.8228686451911926, "learning_rate": 0.0002, "epoch": 3.212351467140894, "step": 19870}, {"loss": 0.625, "grad_norm": 0.7936596870422363, "learning_rate": 0.0002, "epoch": 3.213968151321639, "step": 19880}, {"loss": 0.5698, "grad_norm": 1.097377896308899, "learning_rate": 0.0002, "epoch": 3.2155848355023844, "step": 19890}, {"loss": 0.6725, "grad_norm": 0.9544782638549805, "learning_rate": 0.0002, "epoch": 3.21720151968313, "step": 19900}, {"loss": 0.6022, "grad_norm": 0.8240751624107361, "learning_rate": 0.0002, "epoch": 3.2188182038638753, "step": 19910}, {"loss": 0.5659, "grad_norm": 0.8332096338272095, "learning_rate": 0.0002, "epoch": 3.2204348880446205, "step": 19920}, {"loss": 0.6274, "grad_norm": 1.0954567193984985, "learning_rate": 0.0002, "epoch": 3.2220515722253658, "step": 19930}, {"loss": 0.652, "grad_norm": 0.7790525555610657, "learning_rate": 0.0002, "epoch": 3.223668256406111, "step": 19940}, {"loss": 0.5986, "grad_norm": 0.7966814041137695, "learning_rate": 0.0002, "epoch": 3.225284940586856, "step": 19950}, {"loss": 0.5911, "grad_norm": 0.9751881957054138, "learning_rate": 0.0002, "epoch": 3.2269016247676015, "step": 19960}, {"loss": 0.6071, "grad_norm": 0.9856047630310059, "learning_rate": 0.0002, "epoch": 3.228518308948347, "step": 19970}, {"loss": 0.5837, "grad_norm": 1.3062353134155273, "learning_rate": 0.0002, "epoch": 3.2301349931290924, "step": 19980}, {"loss": 0.6588, "grad_norm": 0.9510692358016968, "learning_rate": 0.0002, "epoch": 3.2317516773098376, "step": 19990}, {"loss": 0.6264, "grad_norm": 0.8630342483520508, "learning_rate": 0.0002, "epoch": 3.233368361490583, "step": 20000}, {"loss": 0.6073, "grad_norm": 0.8966519236564636, "learning_rate": 0.0002, "epoch": 3.234985045671328, "step": 20010}, {"loss": 0.612, "grad_norm": 0.7093510627746582, "learning_rate": 0.0002, "epoch": 3.2366017298520733, "step": 20020}, {"loss": 0.585, "grad_norm": 0.7771096229553223, "learning_rate": 0.0002, "epoch": 3.2382184140328185, "step": 20030}, {"loss": 0.5821, "grad_norm": 0.841058075428009, "learning_rate": 0.0002, "epoch": 3.2398350982135637, "step": 20040}, {"loss": 0.6519, "grad_norm": 0.909712553024292, "learning_rate": 0.0002, "epoch": 3.2414517823943094, "step": 20050}, {"loss": 0.6089, "grad_norm": 0.8321019411087036, "learning_rate": 0.0002, "epoch": 3.2430684665750547, "step": 20060}, {"loss": 0.6115, "grad_norm": 0.779901921749115, "learning_rate": 0.0002, "epoch": 3.2446851507558, "step": 20070}, {"loss": 0.6107, "grad_norm": 0.6249170303344727, "learning_rate": 0.0002, "epoch": 3.246301834936545, "step": 20080}, {"loss": 0.603, "grad_norm": 0.8000940680503845, "learning_rate": 0.0002, "epoch": 3.2479185191172903, "step": 20090}, {"loss": 0.6273, "grad_norm": 0.7627735137939453, "learning_rate": 0.0002, "epoch": 3.2495352032980356, "step": 20100}, {"loss": 0.6223, "grad_norm": 0.8780747056007385, "learning_rate": 0.0002, "epoch": 3.2511518874787813, "step": 20110}, {"loss": 0.5969, "grad_norm": 0.772037148475647, "learning_rate": 0.0002, "epoch": 3.2527685716595265, "step": 20120}, {"loss": 0.5843, "grad_norm": 1.0086580514907837, "learning_rate": 0.0002, "epoch": 3.2543852558402717, "step": 20130}, {"loss": 0.5777, "grad_norm": 0.9360289573669434, "learning_rate": 0.0002, "epoch": 3.256001940021017, "step": 20140}, {"loss": 0.5777, "grad_norm": 1.2099586725234985, "learning_rate": 0.0002, "epoch": 3.257618624201762, "step": 20150}, {"loss": 0.624, "grad_norm": 0.8368481397628784, "learning_rate": 0.0002, "epoch": 3.2592353083825074, "step": 20160}, {"loss": 0.5626, "grad_norm": 0.7391039133071899, "learning_rate": 0.0002, "epoch": 3.2608519925632526, "step": 20170}, {"loss": 0.6041, "grad_norm": 0.9122273325920105, "learning_rate": 0.0002, "epoch": 3.262468676743998, "step": 20180}, {"loss": 0.5868, "grad_norm": 0.8502281904220581, "learning_rate": 0.0002, "epoch": 3.264085360924743, "step": 20190}, {"loss": 0.5841, "grad_norm": 1.0926852226257324, "learning_rate": 0.0002, "epoch": 3.265702045105489, "step": 20200}, {"loss": 0.6027, "grad_norm": 0.7902828454971313, "learning_rate": 0.0002, "epoch": 3.267318729286234, "step": 20210}, {"loss": 0.6089, "grad_norm": 0.8724729418754578, "learning_rate": 0.0002, "epoch": 3.2689354134669792, "step": 20220}, {"loss": 0.6242, "grad_norm": 0.8469277024269104, "learning_rate": 0.0002, "epoch": 3.2705520976477245, "step": 20230}, {"loss": 0.644, "grad_norm": 0.8865092992782593, "learning_rate": 0.0002, "epoch": 3.2721687818284697, "step": 20240}, {"loss": 0.6464, "grad_norm": 1.0979334115982056, "learning_rate": 0.0002, "epoch": 3.273785466009215, "step": 20250}, {"loss": 0.647, "grad_norm": 1.0860793590545654, "learning_rate": 0.0002, "epoch": 3.2754021501899606, "step": 20260}, {"loss": 0.6105, "grad_norm": 0.981745183467865, "learning_rate": 0.0002, "epoch": 3.277018834370706, "step": 20270}, {"loss": 0.627, "grad_norm": 0.9155020713806152, "learning_rate": 0.0002, "epoch": 3.278635518551451, "step": 20280}, {"loss": 0.5899, "grad_norm": 0.8436718583106995, "learning_rate": 0.0002, "epoch": 3.2802522027321963, "step": 20290}, {"loss": 0.6371, "grad_norm": 1.0329409837722778, "learning_rate": 0.0002, "epoch": 3.2818688869129415, "step": 20300}, {"loss": 0.6, "grad_norm": 0.9876394271850586, "learning_rate": 0.0002, "epoch": 3.2834855710936868, "step": 20310}, {"loss": 0.5463, "grad_norm": 0.8052917718887329, "learning_rate": 0.0002, "epoch": 3.285102255274432, "step": 20320}, {"loss": 0.5949, "grad_norm": 0.8390680551528931, "learning_rate": 0.0002, "epoch": 3.2867189394551772, "step": 20330}, {"loss": 0.6492, "grad_norm": 0.9515735507011414, "learning_rate": 0.0002, "epoch": 3.288335623635923, "step": 20340}, {"loss": 0.596, "grad_norm": 0.8028870224952698, "learning_rate": 0.0002, "epoch": 3.289952307816668, "step": 20350}, {"loss": 0.634, "grad_norm": 0.862592339515686, "learning_rate": 0.0002, "epoch": 3.2915689919974134, "step": 20360}, {"loss": 0.6345, "grad_norm": 0.7451621890068054, "learning_rate": 0.0002, "epoch": 3.2931856761781586, "step": 20370}, {"loss": 0.6458, "grad_norm": 0.8966776728630066, "learning_rate": 0.0002, "epoch": 3.294802360358904, "step": 20380}, {"loss": 0.5967, "grad_norm": 0.9289216995239258, "learning_rate": 0.0002, "epoch": 3.296419044539649, "step": 20390}, {"loss": 0.6599, "grad_norm": 0.9649626612663269, "learning_rate": 0.0002, "epoch": 3.2980357287203943, "step": 20400}, {"loss": 0.5781, "grad_norm": 1.1953798532485962, "learning_rate": 0.0002, "epoch": 3.29965241290114, "step": 20410}, {"loss": 0.5997, "grad_norm": 0.8929083943367004, "learning_rate": 0.0002, "epoch": 3.301269097081885, "step": 20420}, {"loss": 0.597, "grad_norm": 0.8922014236450195, "learning_rate": 0.0002, "epoch": 3.3028857812626304, "step": 20430}, {"loss": 0.5766, "grad_norm": 0.9754860401153564, "learning_rate": 0.0002, "epoch": 3.3045024654433757, "step": 20440}, {"loss": 0.5653, "grad_norm": 0.8873140215873718, "learning_rate": 0.0002, "epoch": 3.306119149624121, "step": 20450}, {"loss": 0.6138, "grad_norm": 0.857271671295166, "learning_rate": 0.0002, "epoch": 3.307735833804866, "step": 20460}, {"loss": 0.633, "grad_norm": 0.9022141098976135, "learning_rate": 0.0002, "epoch": 3.3093525179856114, "step": 20470}, {"loss": 0.6654, "grad_norm": 0.8614798188209534, "learning_rate": 0.0002, "epoch": 3.3109692021663566, "step": 20480}, {"loss": 0.6254, "grad_norm": 0.8838164210319519, "learning_rate": 0.0002, "epoch": 3.3125858863471023, "step": 20490}, {"loss": 0.5849, "grad_norm": 0.8709736466407776, "learning_rate": 0.0002, "epoch": 3.3142025705278475, "step": 20500}, {"loss": 0.6146, "grad_norm": 0.9533300995826721, "learning_rate": 0.0002, "epoch": 3.3158192547085927, "step": 20510}, {"loss": 0.6029, "grad_norm": 0.8259269595146179, "learning_rate": 0.0002, "epoch": 3.317435938889338, "step": 20520}, {"loss": 0.6268, "grad_norm": 0.8607608079910278, "learning_rate": 0.0002, "epoch": 3.319052623070083, "step": 20530}, {"loss": 0.5676, "grad_norm": 1.0863020420074463, "learning_rate": 0.0002, "epoch": 3.3206693072508284, "step": 20540}, {"loss": 0.6412, "grad_norm": 1.011489987373352, "learning_rate": 0.0002, "epoch": 3.3222859914315737, "step": 20550}, {"loss": 0.6247, "grad_norm": 0.6952177882194519, "learning_rate": 0.0002, "epoch": 3.3239026756123193, "step": 20560}, {"loss": 0.6229, "grad_norm": 0.9638974070549011, "learning_rate": 0.0002, "epoch": 3.3255193597930646, "step": 20570}, {"loss": 0.5882, "grad_norm": 1.0310138463974, "learning_rate": 0.0002, "epoch": 3.32713604397381, "step": 20580}, {"loss": 0.594, "grad_norm": 0.9371318221092224, "learning_rate": 0.0002, "epoch": 3.328752728154555, "step": 20590}, {"loss": 0.6137, "grad_norm": 0.8756691813468933, "learning_rate": 0.0002, "epoch": 3.3303694123353003, "step": 20600}, {"loss": 0.5994, "grad_norm": 1.054175853729248, "learning_rate": 0.0002, "epoch": 3.3319860965160455, "step": 20610}, {"loss": 0.6169, "grad_norm": 0.9074128270149231, "learning_rate": 0.0002, "epoch": 3.3336027806967907, "step": 20620}, {"loss": 0.6138, "grad_norm": 0.906900942325592, "learning_rate": 0.0002, "epoch": 3.335219464877536, "step": 20630}, {"loss": 0.571, "grad_norm": 0.8689333200454712, "learning_rate": 0.0002, "epoch": 3.3368361490582816, "step": 20640}, {"loss": 0.6079, "grad_norm": 0.9889747500419617, "learning_rate": 0.0002, "epoch": 3.338452833239027, "step": 20650}, {"loss": 0.6073, "grad_norm": 1.0685805082321167, "learning_rate": 0.0002, "epoch": 3.340069517419772, "step": 20660}, {"loss": 0.6091, "grad_norm": 0.7495010495185852, "learning_rate": 0.0002, "epoch": 3.3416862016005173, "step": 20670}, {"loss": 0.5883, "grad_norm": 0.8747848272323608, "learning_rate": 0.0002, "epoch": 3.3433028857812626, "step": 20680}, {"loss": 0.604, "grad_norm": 0.9762673377990723, "learning_rate": 0.0002, "epoch": 3.344919569962008, "step": 20690}, {"loss": 0.6784, "grad_norm": 1.0284489393234253, "learning_rate": 0.0002, "epoch": 3.346536254142753, "step": 20700}, {"loss": 0.6464, "grad_norm": 0.7293812036514282, "learning_rate": 0.0002, "epoch": 3.3481529383234987, "step": 20710}, {"loss": 0.609, "grad_norm": 0.8330199122428894, "learning_rate": 0.0002, "epoch": 3.349769622504244, "step": 20720}, {"loss": 0.5729, "grad_norm": 0.9808499217033386, "learning_rate": 0.0002, "epoch": 3.351386306684989, "step": 20730}, {"loss": 0.6315, "grad_norm": 0.9508825540542603, "learning_rate": 0.0002, "epoch": 3.3530029908657344, "step": 20740}, {"loss": 0.5965, "grad_norm": 0.790483832359314, "learning_rate": 0.0002, "epoch": 3.3546196750464796, "step": 20750}, {"loss": 0.6327, "grad_norm": 1.022793173789978, "learning_rate": 0.0002, "epoch": 3.356236359227225, "step": 20760}, {"loss": 0.6439, "grad_norm": 0.8318950533866882, "learning_rate": 0.0002, "epoch": 3.35785304340797, "step": 20770}, {"loss": 0.6037, "grad_norm": 0.7980858087539673, "learning_rate": 0.0002, "epoch": 3.3594697275887153, "step": 20780}, {"loss": 0.6746, "grad_norm": 0.8114802241325378, "learning_rate": 0.0002, "epoch": 3.361086411769461, "step": 20790}, {"loss": 0.6017, "grad_norm": 0.8522519469261169, "learning_rate": 0.0002, "epoch": 3.3627030959502062, "step": 20800}, {"loss": 0.5864, "grad_norm": 0.9142431616783142, "learning_rate": 0.0002, "epoch": 3.3643197801309515, "step": 20810}, {"loss": 0.6331, "grad_norm": 0.771170437335968, "learning_rate": 0.0002, "epoch": 3.3659364643116967, "step": 20820}, {"loss": 0.5879, "grad_norm": 1.0628231763839722, "learning_rate": 0.0002, "epoch": 3.367553148492442, "step": 20830}, {"loss": 0.6533, "grad_norm": 0.9384352564811707, "learning_rate": 0.0002, "epoch": 3.369169832673187, "step": 20840}, {"loss": 0.6292, "grad_norm": 1.1286591291427612, "learning_rate": 0.0002, "epoch": 3.370786516853933, "step": 20850}, {"loss": 0.5986, "grad_norm": 1.1349513530731201, "learning_rate": 0.0002, "epoch": 3.372403201034678, "step": 20860}, {"loss": 0.6413, "grad_norm": 1.0127464532852173, "learning_rate": 0.0002, "epoch": 3.3740198852154233, "step": 20870}, {"loss": 0.6414, "grad_norm": 0.9111971855163574, "learning_rate": 0.0002, "epoch": 3.3756365693961685, "step": 20880}, {"loss": 0.6101, "grad_norm": 0.871356725692749, "learning_rate": 0.0002, "epoch": 3.3772532535769137, "step": 20890}, {"loss": 0.5995, "grad_norm": 0.7774117588996887, "learning_rate": 0.0002, "epoch": 3.378869937757659, "step": 20900}, {"loss": 0.6062, "grad_norm": 1.0089964866638184, "learning_rate": 0.0002, "epoch": 3.380486621938404, "step": 20910}, {"loss": 0.5908, "grad_norm": 0.7855867147445679, "learning_rate": 0.0002, "epoch": 3.3821033061191494, "step": 20920}, {"loss": 0.6373, "grad_norm": 1.3713710308074951, "learning_rate": 0.0002, "epoch": 3.3837199902998947, "step": 20930}, {"loss": 0.6627, "grad_norm": 0.8599116206169128, "learning_rate": 0.0002, "epoch": 3.3853366744806404, "step": 20940}, {"loss": 0.6224, "grad_norm": 0.9392673373222351, "learning_rate": 0.0002, "epoch": 3.3869533586613856, "step": 20950}, {"loss": 0.5855, "grad_norm": 0.8764075040817261, "learning_rate": 0.0002, "epoch": 3.388570042842131, "step": 20960}, {"loss": 0.5734, "grad_norm": 0.8240136504173279, "learning_rate": 0.0002, "epoch": 3.390186727022876, "step": 20970}, {"loss": 0.5783, "grad_norm": 1.0982369184494019, "learning_rate": 0.0002, "epoch": 3.3918034112036213, "step": 20980}, {"loss": 0.5451, "grad_norm": 1.0599013566970825, "learning_rate": 0.0002, "epoch": 3.3934200953843665, "step": 20990}, {"loss": 0.6356, "grad_norm": 0.895438015460968, "learning_rate": 0.0002, "epoch": 3.395036779565112, "step": 21000}, {"loss": 0.6065, "grad_norm": 0.6974841356277466, "learning_rate": 0.0002, "epoch": 3.3966534637458574, "step": 21010}, {"loss": 0.5704, "grad_norm": 0.9571719765663147, "learning_rate": 0.0002, "epoch": 3.3982701479266026, "step": 21020}, {"loss": 0.679, "grad_norm": 0.831912636756897, "learning_rate": 0.0002, "epoch": 3.399886832107348, "step": 21030}, {"loss": 0.6051, "grad_norm": 0.831936240196228, "learning_rate": 0.0002, "epoch": 3.401503516288093, "step": 21040}, {"loss": 0.5857, "grad_norm": 0.7388373613357544, "learning_rate": 0.0002, "epoch": 3.4031202004688383, "step": 21050}, {"loss": 0.6245, "grad_norm": 0.938667356967926, "learning_rate": 0.0002, "epoch": 3.4047368846495836, "step": 21060}, {"loss": 0.6121, "grad_norm": 0.9202313423156738, "learning_rate": 0.0002, "epoch": 3.406353568830329, "step": 21070}, {"loss": 0.6388, "grad_norm": 0.9888381958007812, "learning_rate": 0.0002, "epoch": 3.4079702530110745, "step": 21080}, {"loss": 0.6245, "grad_norm": 0.8526970744132996, "learning_rate": 0.0002, "epoch": 3.4095869371918197, "step": 21090}, {"loss": 0.5914, "grad_norm": 0.7939383387565613, "learning_rate": 0.0002, "epoch": 3.411203621372565, "step": 21100}, {"loss": 0.6066, "grad_norm": 0.9986352920532227, "learning_rate": 0.0002, "epoch": 3.41282030555331, "step": 21110}, {"loss": 0.5947, "grad_norm": 0.8895300030708313, "learning_rate": 0.0002, "epoch": 3.4144369897340554, "step": 21120}, {"loss": 0.6264, "grad_norm": 0.9559482932090759, "learning_rate": 0.0002, "epoch": 3.4160536739148006, "step": 21130}, {"loss": 0.6491, "grad_norm": 0.8351506590843201, "learning_rate": 0.0002, "epoch": 3.417670358095546, "step": 21140}, {"loss": 0.567, "grad_norm": 0.8224456906318665, "learning_rate": 0.0002, "epoch": 3.4192870422762915, "step": 21150}, {"loss": 0.5871, "grad_norm": 1.0110299587249756, "learning_rate": 0.0002, "epoch": 3.4209037264570368, "step": 21160}, {"loss": 0.6116, "grad_norm": 0.82564777135849, "learning_rate": 0.0002, "epoch": 3.422520410637782, "step": 21170}, {"loss": 0.595, "grad_norm": 1.004738688468933, "learning_rate": 0.0002, "epoch": 3.4241370948185272, "step": 21180}, {"loss": 0.6286, "grad_norm": 0.7545676827430725, "learning_rate": 0.0002, "epoch": 3.4257537789992725, "step": 21190}, {"loss": 0.5868, "grad_norm": 0.8918704390525818, "learning_rate": 0.0002, "epoch": 3.4273704631800177, "step": 21200}, {"loss": 0.6542, "grad_norm": 0.8336876034736633, "learning_rate": 0.0002, "epoch": 3.428987147360763, "step": 21210}, {"loss": 0.5824, "grad_norm": 0.8928771018981934, "learning_rate": 0.0002, "epoch": 3.430603831541508, "step": 21220}, {"loss": 0.6468, "grad_norm": 0.7663705945014954, "learning_rate": 0.0002, "epoch": 3.432220515722254, "step": 21230}, {"loss": 0.6693, "grad_norm": 0.8392598628997803, "learning_rate": 0.0002, "epoch": 3.433837199902999, "step": 21240}, {"loss": 0.5971, "grad_norm": 0.8819600343704224, "learning_rate": 0.0002, "epoch": 3.4354538840837443, "step": 21250}, {"loss": 0.6791, "grad_norm": 0.9124642014503479, "learning_rate": 0.0002, "epoch": 3.4370705682644895, "step": 21260}, {"loss": 0.5925, "grad_norm": 0.8329763412475586, "learning_rate": 0.0002, "epoch": 3.4386872524452348, "step": 21270}, {"loss": 0.6541, "grad_norm": 0.9982839822769165, "learning_rate": 0.0002, "epoch": 3.44030393662598, "step": 21280}, {"loss": 0.6441, "grad_norm": 0.9105954766273499, "learning_rate": 0.0002, "epoch": 3.4419206208067252, "step": 21290}, {"loss": 0.6028, "grad_norm": 0.8182359337806702, "learning_rate": 0.0002, "epoch": 3.443537304987471, "step": 21300}, {"loss": 0.5991, "grad_norm": 1.0568904876708984, "learning_rate": 0.0002, "epoch": 3.445153989168216, "step": 21310}, {"loss": 0.6117, "grad_norm": 0.968539834022522, "learning_rate": 0.0002, "epoch": 3.4467706733489614, "step": 21320}, {"loss": 0.6219, "grad_norm": 0.8774511218070984, "learning_rate": 0.0002, "epoch": 3.4483873575297066, "step": 21330}, {"loss": 0.6438, "grad_norm": 0.7598156332969666, "learning_rate": 0.0002, "epoch": 3.450004041710452, "step": 21340}, {"loss": 0.6033, "grad_norm": 1.1012897491455078, "learning_rate": 0.0002, "epoch": 3.451620725891197, "step": 21350}, {"loss": 0.6137, "grad_norm": 0.8040637373924255, "learning_rate": 0.0002, "epoch": 3.4532374100719423, "step": 21360}, {"loss": 0.6173, "grad_norm": 0.8497496247291565, "learning_rate": 0.0002, "epoch": 3.4548540942526875, "step": 21370}, {"loss": 0.6005, "grad_norm": 0.8429915904998779, "learning_rate": 0.0002, "epoch": 3.456470778433433, "step": 21380}, {"loss": 0.6182, "grad_norm": 0.8107112646102905, "learning_rate": 0.0002, "epoch": 3.4580874626141784, "step": 21390}, {"loss": 0.6109, "grad_norm": 1.00872004032135, "learning_rate": 0.0002, "epoch": 3.4597041467949237, "step": 21400}, {"loss": 0.5712, "grad_norm": 0.8266542553901672, "learning_rate": 0.0002, "epoch": 3.461320830975669, "step": 21410}, {"loss": 0.6457, "grad_norm": 0.8972568511962891, "learning_rate": 0.0002, "epoch": 3.462937515156414, "step": 21420}, {"loss": 0.6081, "grad_norm": 1.0781476497650146, "learning_rate": 0.0002, "epoch": 3.4645541993371594, "step": 21430}, {"loss": 0.6303, "grad_norm": 0.9571592807769775, "learning_rate": 0.0002, "epoch": 3.4661708835179046, "step": 21440}, {"loss": 0.6309, "grad_norm": 0.881547212600708, "learning_rate": 0.0002, "epoch": 3.4677875676986503, "step": 21450}, {"loss": 0.6076, "grad_norm": 0.6955338716506958, "learning_rate": 0.0002, "epoch": 3.4694042518793955, "step": 21460}, {"loss": 0.6205, "grad_norm": 0.901187539100647, "learning_rate": 0.0002, "epoch": 3.4710209360601407, "step": 21470}, {"loss": 0.639, "grad_norm": 0.7063511610031128, "learning_rate": 0.0002, "epoch": 3.472637620240886, "step": 21480}, {"loss": 0.6154, "grad_norm": 0.8462792038917542, "learning_rate": 0.0002, "epoch": 3.474254304421631, "step": 21490}, {"loss": 0.61, "grad_norm": 1.1861060857772827, "learning_rate": 0.0002, "epoch": 3.4758709886023764, "step": 21500}, {"loss": 0.6586, "grad_norm": 0.70503169298172, "learning_rate": 0.0002, "epoch": 3.4774876727831217, "step": 21510}, {"loss": 0.6475, "grad_norm": 0.9650066494941711, "learning_rate": 0.0002, "epoch": 3.479104356963867, "step": 21520}, {"loss": 0.6452, "grad_norm": 1.0266852378845215, "learning_rate": 0.0002, "epoch": 3.4807210411446126, "step": 21530}, {"loss": 0.6553, "grad_norm": 0.956372857093811, "learning_rate": 0.0002, "epoch": 3.482337725325358, "step": 21540}, {"loss": 0.6667, "grad_norm": 0.8848432898521423, "learning_rate": 0.0002, "epoch": 3.483954409506103, "step": 21550}, {"loss": 0.6375, "grad_norm": 1.0805351734161377, "learning_rate": 0.0002, "epoch": 3.4855710936868483, "step": 21560}, {"loss": 0.6958, "grad_norm": 0.9279725551605225, "learning_rate": 0.0002, "epoch": 3.4871877778675935, "step": 21570}, {"loss": 0.6354, "grad_norm": 0.9049562215805054, "learning_rate": 0.0002, "epoch": 3.4888044620483387, "step": 21580}, {"loss": 0.6071, "grad_norm": 0.9619429111480713, "learning_rate": 0.0002, "epoch": 3.4904211462290844, "step": 21590}, {"loss": 0.5927, "grad_norm": 0.8508906960487366, "learning_rate": 0.0002, "epoch": 3.4920378304098296, "step": 21600}, {"loss": 0.6115, "grad_norm": 0.8692502379417419, "learning_rate": 0.0002, "epoch": 3.493654514590575, "step": 21610}, {"loss": 0.5878, "grad_norm": 0.8187332153320312, "learning_rate": 0.0002, "epoch": 3.49527119877132, "step": 21620}, {"loss": 0.5874, "grad_norm": 1.145400047302246, "learning_rate": 0.0002, "epoch": 3.4968878829520653, "step": 21630}, {"loss": 0.6313, "grad_norm": 0.8281388282775879, "learning_rate": 0.0002, "epoch": 3.4985045671328105, "step": 21640}, {"loss": 0.6624, "grad_norm": 0.82256019115448, "learning_rate": 0.0002, "epoch": 3.500121251313556, "step": 21650}, {"loss": 0.6346, "grad_norm": 0.9315484762191772, "learning_rate": 0.0002, "epoch": 3.501737935494301, "step": 21660}, {"loss": 0.6086, "grad_norm": 0.7626111507415771, "learning_rate": 0.0002, "epoch": 3.5033546196750462, "step": 21670}, {"loss": 0.6177, "grad_norm": 0.9275059103965759, "learning_rate": 0.0002, "epoch": 3.504971303855792, "step": 21680}, {"loss": 0.64, "grad_norm": 0.7906724810600281, "learning_rate": 0.0002, "epoch": 3.506587988036537, "step": 21690}, {"loss": 0.6015, "grad_norm": 0.8289761543273926, "learning_rate": 0.0002, "epoch": 3.5082046722172824, "step": 21700}, {"loss": 0.6246, "grad_norm": 0.8316431045532227, "learning_rate": 0.0002, "epoch": 3.5098213563980276, "step": 21710}, {"loss": 0.619, "grad_norm": 1.0451812744140625, "learning_rate": 0.0002, "epoch": 3.511438040578773, "step": 21720}, {"loss": 0.632, "grad_norm": 0.928252637386322, "learning_rate": 0.0002, "epoch": 3.513054724759518, "step": 21730}, {"loss": 0.6062, "grad_norm": 0.7985895276069641, "learning_rate": 0.0002, "epoch": 3.5146714089402638, "step": 21740}, {"loss": 0.6463, "grad_norm": 0.6740974187850952, "learning_rate": 0.0002, "epoch": 3.516288093121009, "step": 21750}, {"loss": 0.6138, "grad_norm": 0.8482223749160767, "learning_rate": 0.0002, "epoch": 3.517904777301754, "step": 21760}, {"loss": 0.6277, "grad_norm": 0.889947772026062, "learning_rate": 0.0002, "epoch": 3.5195214614824994, "step": 21770}, {"loss": 0.6174, "grad_norm": 0.8304598927497864, "learning_rate": 0.0002, "epoch": 3.5211381456632447, "step": 21780}, {"loss": 0.6156, "grad_norm": 0.8002981543540955, "learning_rate": 0.0002, "epoch": 3.52275482984399, "step": 21790}, {"loss": 0.5896, "grad_norm": 0.8115083575248718, "learning_rate": 0.0002, "epoch": 3.524371514024735, "step": 21800}, {"loss": 0.6041, "grad_norm": 0.9715048670768738, "learning_rate": 0.0002, "epoch": 3.5259881982054804, "step": 21810}, {"loss": 0.6715, "grad_norm": 1.0910786390304565, "learning_rate": 0.0002, "epoch": 3.5276048823862256, "step": 21820}, {"loss": 0.6543, "grad_norm": 0.8438942432403564, "learning_rate": 0.0002, "epoch": 3.5292215665669713, "step": 21830}, {"loss": 0.6509, "grad_norm": 0.8813382983207703, "learning_rate": 0.0002, "epoch": 3.5308382507477165, "step": 21840}, {"loss": 0.6049, "grad_norm": 0.7092908024787903, "learning_rate": 0.0002, "epoch": 3.5324549349284617, "step": 21850}, {"loss": 0.5678, "grad_norm": 0.8332187533378601, "learning_rate": 0.0002, "epoch": 3.534071619109207, "step": 21860}, {"loss": 0.5896, "grad_norm": 0.8958209156990051, "learning_rate": 0.0002, "epoch": 3.535688303289952, "step": 21870}, {"loss": 0.6476, "grad_norm": 0.824138879776001, "learning_rate": 0.0002, "epoch": 3.5373049874706974, "step": 21880}, {"loss": 0.6022, "grad_norm": 0.8375158309936523, "learning_rate": 0.0002, "epoch": 3.538921671651443, "step": 21890}, {"loss": 0.6019, "grad_norm": 1.0274608135223389, "learning_rate": 0.0002, "epoch": 3.5405383558321883, "step": 21900}, {"loss": 0.6194, "grad_norm": 0.7088932394981384, "learning_rate": 0.0002, "epoch": 3.5421550400129336, "step": 21910}, {"loss": 0.6554, "grad_norm": 0.8172445297241211, "learning_rate": 0.0002, "epoch": 3.543771724193679, "step": 21920}, {"loss": 0.6711, "grad_norm": 0.9904135465621948, "learning_rate": 0.0002, "epoch": 3.545388408374424, "step": 21930}, {"loss": 0.6001, "grad_norm": 0.9900432229042053, "learning_rate": 0.0002, "epoch": 3.5470050925551693, "step": 21940}, {"loss": 0.6195, "grad_norm": 0.8963301181793213, "learning_rate": 0.0002, "epoch": 3.5486217767359145, "step": 21950}, {"loss": 0.5972, "grad_norm": 0.8551464676856995, "learning_rate": 0.0002, "epoch": 3.5502384609166597, "step": 21960}, {"loss": 0.6206, "grad_norm": 1.0916603803634644, "learning_rate": 0.0002, "epoch": 3.551855145097405, "step": 21970}, {"loss": 0.6523, "grad_norm": 0.841598391532898, "learning_rate": 0.0002, "epoch": 3.5534718292781506, "step": 21980}, {"loss": 0.617, "grad_norm": 0.8566757440567017, "learning_rate": 0.0002, "epoch": 3.555088513458896, "step": 21990}, {"loss": 0.6192, "grad_norm": 1.0145052671432495, "learning_rate": 0.0002, "epoch": 3.556705197639641, "step": 22000}, {"loss": 0.6173, "grad_norm": 0.9293754696846008, "learning_rate": 0.0002, "epoch": 3.5583218818203863, "step": 22010}, {"loss": 0.612, "grad_norm": 0.9568536281585693, "learning_rate": 0.0002, "epoch": 3.5599385660011316, "step": 22020}, {"loss": 0.641, "grad_norm": 0.8613139986991882, "learning_rate": 0.0002, "epoch": 3.5615552501818772, "step": 22030}, {"loss": 0.6496, "grad_norm": 0.8179237246513367, "learning_rate": 0.0002, "epoch": 3.5631719343626225, "step": 22040}, {"loss": 0.574, "grad_norm": 0.9059830904006958, "learning_rate": 0.0002, "epoch": 3.5647886185433677, "step": 22050}, {"loss": 0.6448, "grad_norm": 1.0068252086639404, "learning_rate": 0.0002, "epoch": 3.566405302724113, "step": 22060}, {"loss": 0.6239, "grad_norm": 0.9682072997093201, "learning_rate": 0.0002, "epoch": 3.568021986904858, "step": 22070}, {"loss": 0.6808, "grad_norm": 0.8514005541801453, "learning_rate": 0.0002, "epoch": 3.5696386710856034, "step": 22080}, {"loss": 0.5956, "grad_norm": 0.8327770829200745, "learning_rate": 0.0002, "epoch": 3.5712553552663486, "step": 22090}, {"loss": 0.5976, "grad_norm": 1.024976372718811, "learning_rate": 0.0002, "epoch": 3.572872039447094, "step": 22100}, {"loss": 0.624, "grad_norm": 0.7721174955368042, "learning_rate": 0.0002, "epoch": 3.574488723627839, "step": 22110}, {"loss": 0.5896, "grad_norm": 1.0351054668426514, "learning_rate": 0.0002, "epoch": 3.5761054078085843, "step": 22120}, {"loss": 0.6379, "grad_norm": 0.9680907130241394, "learning_rate": 0.0002, "epoch": 3.57772209198933, "step": 22130}, {"loss": 0.6194, "grad_norm": 0.8016974925994873, "learning_rate": 0.0002, "epoch": 3.5793387761700752, "step": 22140}, {"loss": 0.6387, "grad_norm": 1.0109003782272339, "learning_rate": 0.0002, "epoch": 3.5809554603508205, "step": 22150}, {"loss": 0.6368, "grad_norm": 1.0473392009735107, "learning_rate": 0.0002, "epoch": 3.5825721445315657, "step": 22160}, {"loss": 0.6353, "grad_norm": 0.8686613440513611, "learning_rate": 0.0002, "epoch": 3.584188828712311, "step": 22170}, {"loss": 0.5791, "grad_norm": 0.869149923324585, "learning_rate": 0.0002, "epoch": 3.5858055128930566, "step": 22180}, {"loss": 0.5895, "grad_norm": 0.9769062995910645, "learning_rate": 0.0002, "epoch": 3.587422197073802, "step": 22190}, {"loss": 0.5939, "grad_norm": 0.779636561870575, "learning_rate": 0.0002, "epoch": 3.589038881254547, "step": 22200}, {"loss": 0.5875, "grad_norm": 0.9063841104507446, "learning_rate": 0.0002, "epoch": 3.5906555654352923, "step": 22210}, {"loss": 0.5671, "grad_norm": 0.9216037392616272, "learning_rate": 0.0002, "epoch": 3.5922722496160375, "step": 22220}, {"loss": 0.6484, "grad_norm": 1.0217336416244507, "learning_rate": 0.0002, "epoch": 3.5938889337967828, "step": 22230}, {"loss": 0.6511, "grad_norm": 0.8513161540031433, "learning_rate": 0.0002, "epoch": 3.595505617977528, "step": 22240}, {"loss": 0.6301, "grad_norm": 0.8084813952445984, "learning_rate": 0.0002, "epoch": 3.597122302158273, "step": 22250}, {"loss": 0.6197, "grad_norm": 0.8524802923202515, "learning_rate": 0.0002, "epoch": 3.5987389863390185, "step": 22260}, {"loss": 0.5599, "grad_norm": 0.9356237649917603, "learning_rate": 0.0002, "epoch": 3.600355670519764, "step": 22270}, {"loss": 0.628, "grad_norm": 1.009600281715393, "learning_rate": 0.0002, "epoch": 3.6019723547005094, "step": 22280}, {"loss": 0.6179, "grad_norm": 0.9900581240653992, "learning_rate": 0.0002, "epoch": 3.6035890388812546, "step": 22290}, {"loss": 0.5725, "grad_norm": 1.062495231628418, "learning_rate": 0.0002, "epoch": 3.605205723062, "step": 22300}, {"loss": 0.607, "grad_norm": 0.8832381367683411, "learning_rate": 0.0002, "epoch": 3.606822407242745, "step": 22310}, {"loss": 0.6215, "grad_norm": 0.9284297823905945, "learning_rate": 0.0002, "epoch": 3.6084390914234903, "step": 22320}, {"loss": 0.685, "grad_norm": 1.2381829023361206, "learning_rate": 0.0002, "epoch": 3.610055775604236, "step": 22330}, {"loss": 0.6181, "grad_norm": 0.929434597492218, "learning_rate": 0.0002, "epoch": 3.611672459784981, "step": 22340}, {"loss": 0.6141, "grad_norm": 0.9714490175247192, "learning_rate": 0.0002, "epoch": 3.6132891439657264, "step": 22350}, {"loss": 0.6861, "grad_norm": 0.808014988899231, "learning_rate": 0.0002, "epoch": 3.6149058281464717, "step": 22360}, {"loss": 0.6428, "grad_norm": 1.0364398956298828, "learning_rate": 0.0002, "epoch": 3.616522512327217, "step": 22370}, {"loss": 0.6337, "grad_norm": 0.7858489751815796, "learning_rate": 0.0002, "epoch": 3.618139196507962, "step": 22380}, {"loss": 0.6214, "grad_norm": 0.9920870065689087, "learning_rate": 0.0002, "epoch": 3.6197558806887074, "step": 22390}, {"loss": 0.6659, "grad_norm": 0.9183220863342285, "learning_rate": 0.0002, "epoch": 3.6213725648694526, "step": 22400}, {"loss": 0.6036, "grad_norm": 0.9826246500015259, "learning_rate": 0.0002, "epoch": 3.622989249050198, "step": 22410}, {"loss": 0.6441, "grad_norm": 0.8632931113243103, "learning_rate": 0.0002, "epoch": 3.6246059332309435, "step": 22420}, {"loss": 0.6124, "grad_norm": 0.8468965291976929, "learning_rate": 0.0002, "epoch": 3.6262226174116887, "step": 22430}, {"loss": 0.6328, "grad_norm": 0.8466871976852417, "learning_rate": 0.0002, "epoch": 3.627839301592434, "step": 22440}, {"loss": 0.5941, "grad_norm": 0.9501169919967651, "learning_rate": 0.0002, "epoch": 3.629455985773179, "step": 22450}, {"loss": 0.6069, "grad_norm": 0.8906720876693726, "learning_rate": 0.0002, "epoch": 3.6310726699539244, "step": 22460}, {"loss": 0.6928, "grad_norm": 0.7400227189064026, "learning_rate": 0.0002, "epoch": 3.6326893541346696, "step": 22470}, {"loss": 0.6337, "grad_norm": 0.9756355881690979, "learning_rate": 0.0002, "epoch": 3.6343060383154153, "step": 22480}, {"loss": 0.6203, "grad_norm": 0.7504993081092834, "learning_rate": 0.0002, "epoch": 3.6359227224961606, "step": 22490}, {"loss": 0.6302, "grad_norm": 0.9270039200782776, "learning_rate": 0.0002, "epoch": 3.637539406676906, "step": 22500}, {"loss": 0.6026, "grad_norm": 0.8841686844825745, "learning_rate": 0.0002, "epoch": 3.639156090857651, "step": 22510}, {"loss": 0.6098, "grad_norm": 0.8533213138580322, "learning_rate": 0.0002, "epoch": 3.6407727750383962, "step": 22520}, {"loss": 0.6412, "grad_norm": 1.0052043199539185, "learning_rate": 0.0002, "epoch": 3.6423894592191415, "step": 22530}, {"loss": 0.6363, "grad_norm": 1.0323461294174194, "learning_rate": 0.0002, "epoch": 3.6440061433998867, "step": 22540}, {"loss": 0.6545, "grad_norm": 0.8654312491416931, "learning_rate": 0.0002, "epoch": 3.645622827580632, "step": 22550}, {"loss": 0.6155, "grad_norm": 0.6400038003921509, "learning_rate": 0.0002, "epoch": 3.647239511761377, "step": 22560}, {"loss": 0.5829, "grad_norm": 0.8061298727989197, "learning_rate": 0.0002, "epoch": 3.648856195942123, "step": 22570}, {"loss": 0.6388, "grad_norm": 0.9257854223251343, "learning_rate": 0.0002, "epoch": 3.650472880122868, "step": 22580}, {"loss": 0.6409, "grad_norm": 0.8439396619796753, "learning_rate": 0.0002, "epoch": 3.6520895643036133, "step": 22590}, {"loss": 0.5996, "grad_norm": 0.7764544486999512, "learning_rate": 0.0002, "epoch": 3.6537062484843585, "step": 22600}, {"loss": 0.6434, "grad_norm": 1.125451683998108, "learning_rate": 0.0002, "epoch": 3.6553229326651038, "step": 22610}, {"loss": 0.6579, "grad_norm": 0.7523018717765808, "learning_rate": 0.0002, "epoch": 3.656939616845849, "step": 22620}, {"loss": 0.6476, "grad_norm": 1.071026086807251, "learning_rate": 0.0002, "epoch": 3.6585563010265947, "step": 22630}, {"loss": 0.6459, "grad_norm": 0.945791482925415, "learning_rate": 0.0002, "epoch": 3.66017298520734, "step": 22640}, {"loss": 0.659, "grad_norm": 0.8001811504364014, "learning_rate": 0.0002, "epoch": 3.661789669388085, "step": 22650}, {"loss": 0.6385, "grad_norm": 0.9700816869735718, "learning_rate": 0.0002, "epoch": 3.6634063535688304, "step": 22660}, {"loss": 0.6337, "grad_norm": 0.9053242206573486, "learning_rate": 0.0002, "epoch": 3.6650230377495756, "step": 22670}, {"loss": 0.6335, "grad_norm": 0.944362461566925, "learning_rate": 0.0002, "epoch": 3.666639721930321, "step": 22680}, {"loss": 0.6235, "grad_norm": 1.067489504814148, "learning_rate": 0.0002, "epoch": 3.668256406111066, "step": 22690}, {"loss": 0.698, "grad_norm": 1.0984995365142822, "learning_rate": 0.0002, "epoch": 3.6698730902918113, "step": 22700}, {"loss": 0.6717, "grad_norm": 0.9336317777633667, "learning_rate": 0.0002, "epoch": 3.6714897744725565, "step": 22710}, {"loss": 0.6195, "grad_norm": 0.9261918663978577, "learning_rate": 0.0002, "epoch": 3.673106458653302, "step": 22720}, {"loss": 0.6332, "grad_norm": 0.8648008704185486, "learning_rate": 0.0002, "epoch": 3.6747231428340474, "step": 22730}, {"loss": 0.6576, "grad_norm": 0.7225083708763123, "learning_rate": 0.0002, "epoch": 3.6763398270147927, "step": 22740}, {"loss": 0.6406, "grad_norm": 0.9258282780647278, "learning_rate": 0.0002, "epoch": 3.677956511195538, "step": 22750}, {"loss": 0.6397, "grad_norm": 0.70876145362854, "learning_rate": 0.0002, "epoch": 3.679573195376283, "step": 22760}, {"loss": 0.6821, "grad_norm": 0.8780210018157959, "learning_rate": 0.0002, "epoch": 3.681189879557029, "step": 22770}, {"loss": 0.6036, "grad_norm": 0.8075440526008606, "learning_rate": 0.0002, "epoch": 3.682806563737774, "step": 22780}, {"loss": 0.6561, "grad_norm": 0.8503130674362183, "learning_rate": 0.0002, "epoch": 3.6844232479185193, "step": 22790}, {"loss": 0.6082, "grad_norm": 0.8413618206977844, "learning_rate": 0.0002, "epoch": 3.6860399320992645, "step": 22800}, {"loss": 0.614, "grad_norm": 0.8675165176391602, "learning_rate": 0.0002, "epoch": 3.6876566162800097, "step": 22810}, {"loss": 0.6157, "grad_norm": 0.8235884308815002, "learning_rate": 0.0002, "epoch": 3.689273300460755, "step": 22820}, {"loss": 0.5708, "grad_norm": 0.9477725625038147, "learning_rate": 0.0002, "epoch": 3.6908899846415, "step": 22830}, {"loss": 0.6481, "grad_norm": 0.7883533835411072, "learning_rate": 0.0002, "epoch": 3.6925066688222454, "step": 22840}, {"loss": 0.5872, "grad_norm": 1.047913908958435, "learning_rate": 0.0002, "epoch": 3.6941233530029907, "step": 22850}, {"loss": 0.6176, "grad_norm": 0.9171528816223145, "learning_rate": 0.0002, "epoch": 3.695740037183736, "step": 22860}, {"loss": 0.6204, "grad_norm": 0.9338192343711853, "learning_rate": 0.0002, "epoch": 3.6973567213644816, "step": 22870}, {"loss": 0.686, "grad_norm": 0.8799443244934082, "learning_rate": 0.0002, "epoch": 3.698973405545227, "step": 22880}, {"loss": 0.6206, "grad_norm": 0.8515434861183167, "learning_rate": 0.0002, "epoch": 3.700590089725972, "step": 22890}, {"loss": 0.5954, "grad_norm": 0.7805591821670532, "learning_rate": 0.0002, "epoch": 3.7022067739067173, "step": 22900}, {"loss": 0.6108, "grad_norm": 0.8470911979675293, "learning_rate": 0.0002, "epoch": 3.7038234580874625, "step": 22910}, {"loss": 0.6557, "grad_norm": 0.9452309012413025, "learning_rate": 0.0002, "epoch": 3.705440142268208, "step": 22920}, {"loss": 0.6529, "grad_norm": 0.950243353843689, "learning_rate": 0.0002, "epoch": 3.7070568264489534, "step": 22930}, {"loss": 0.6364, "grad_norm": 0.7882499098777771, "learning_rate": 0.0002, "epoch": 3.7086735106296986, "step": 22940}, {"loss": 0.6462, "grad_norm": 0.8307787775993347, "learning_rate": 0.0002, "epoch": 3.710290194810444, "step": 22950}, {"loss": 0.6371, "grad_norm": 1.0970630645751953, "learning_rate": 0.0002, "epoch": 3.711906878991189, "step": 22960}, {"loss": 0.6281, "grad_norm": 0.8269566297531128, "learning_rate": 0.0002, "epoch": 3.7135235631719343, "step": 22970}, {"loss": 0.6561, "grad_norm": 0.8306704759597778, "learning_rate": 0.0002, "epoch": 3.7151402473526796, "step": 22980}, {"loss": 0.6418, "grad_norm": 0.9710225462913513, "learning_rate": 0.0002, "epoch": 3.716756931533425, "step": 22990}, {"loss": 0.6639, "grad_norm": 0.8890530467033386, "learning_rate": 0.0002, "epoch": 3.71837361571417, "step": 23000}, {"loss": 0.6084, "grad_norm": 0.883522629737854, "learning_rate": 0.0002, "epoch": 3.7199902998949153, "step": 23010}, {"loss": 0.6183, "grad_norm": 0.8662652373313904, "learning_rate": 0.0002, "epoch": 3.721606984075661, "step": 23020}, {"loss": 0.6266, "grad_norm": 0.7228406667709351, "learning_rate": 0.0002, "epoch": 3.723223668256406, "step": 23030}, {"loss": 0.6417, "grad_norm": 1.060792088508606, "learning_rate": 0.0002, "epoch": 3.7248403524371514, "step": 23040}, {"loss": 0.6346, "grad_norm": 1.0119613409042358, "learning_rate": 0.0002, "epoch": 3.7264570366178966, "step": 23050}, {"loss": 0.6466, "grad_norm": 0.9212996959686279, "learning_rate": 0.0002, "epoch": 3.728073720798642, "step": 23060}, {"loss": 0.6454, "grad_norm": 0.925690233707428, "learning_rate": 0.0002, "epoch": 3.7296904049793875, "step": 23070}, {"loss": 0.615, "grad_norm": 0.8323310613632202, "learning_rate": 0.0002, "epoch": 3.7313070891601328, "step": 23080}, {"loss": 0.679, "grad_norm": 0.8966048955917358, "learning_rate": 0.0002, "epoch": 3.732923773340878, "step": 23090}, {"loss": 0.6151, "grad_norm": 0.8995837569236755, "learning_rate": 0.0002, "epoch": 3.7345404575216232, "step": 23100}, {"loss": 0.6143, "grad_norm": 0.8748890161514282, "learning_rate": 0.0002, "epoch": 3.7361571417023685, "step": 23110}, {"loss": 0.6246, "grad_norm": 0.7985540628433228, "learning_rate": 0.0002, "epoch": 3.7377738258831137, "step": 23120}, {"loss": 0.6279, "grad_norm": 1.0240917205810547, "learning_rate": 0.0002, "epoch": 3.739390510063859, "step": 23130}, {"loss": 0.6747, "grad_norm": 0.9181789755821228, "learning_rate": 0.0002, "epoch": 3.741007194244604, "step": 23140}, {"loss": 0.6026, "grad_norm": 0.8896583914756775, "learning_rate": 0.0002, "epoch": 3.7426238784253494, "step": 23150}, {"loss": 0.5972, "grad_norm": 0.8635515570640564, "learning_rate": 0.0002, "epoch": 3.744240562606095, "step": 23160}, {"loss": 0.6683, "grad_norm": 0.8873575329780579, "learning_rate": 0.0002, "epoch": 3.7458572467868403, "step": 23170}, {"loss": 0.6143, "grad_norm": 0.9807148575782776, "learning_rate": 0.0002, "epoch": 3.7474739309675855, "step": 23180}, {"loss": 0.6381, "grad_norm": 0.900477945804596, "learning_rate": 0.0002, "epoch": 3.7490906151483308, "step": 23190}, {"loss": 0.6542, "grad_norm": 0.9379992485046387, "learning_rate": 0.0002, "epoch": 3.750707299329076, "step": 23200}, {"loss": 0.6015, "grad_norm": 0.9649890661239624, "learning_rate": 0.0002, "epoch": 3.752323983509821, "step": 23210}, {"loss": 0.6735, "grad_norm": 0.824442446231842, "learning_rate": 0.0002, "epoch": 3.753940667690567, "step": 23220}, {"loss": 0.5992, "grad_norm": 0.8896150588989258, "learning_rate": 0.0002, "epoch": 3.755557351871312, "step": 23230}, {"loss": 0.6081, "grad_norm": 0.751249372959137, "learning_rate": 0.0002, "epoch": 3.7571740360520574, "step": 23240}, {"loss": 0.629, "grad_norm": 0.9392193555831909, "learning_rate": 0.0002, "epoch": 3.7587907202328026, "step": 23250}, {"loss": 0.6209, "grad_norm": 0.9284586310386658, "learning_rate": 0.0002, "epoch": 3.760407404413548, "step": 23260}, {"loss": 0.6414, "grad_norm": 0.7738175392150879, "learning_rate": 0.0002, "epoch": 3.762024088594293, "step": 23270}, {"loss": 0.6743, "grad_norm": 0.9252978563308716, "learning_rate": 0.0002, "epoch": 3.7636407727750383, "step": 23280}, {"loss": 0.5984, "grad_norm": 0.9501895904541016, "learning_rate": 0.0002, "epoch": 3.7652574569557835, "step": 23290}, {"loss": 0.6568, "grad_norm": 0.9416276216506958, "learning_rate": 0.0002, "epoch": 3.7668741411365287, "step": 23300}, {"loss": 0.6507, "grad_norm": 0.7076631784439087, "learning_rate": 0.0002, "epoch": 3.7684908253172744, "step": 23310}, {"loss": 0.6329, "grad_norm": 0.9864492416381836, "learning_rate": 0.0002, "epoch": 3.7701075094980196, "step": 23320}, {"loss": 0.6537, "grad_norm": 0.8450456261634827, "learning_rate": 0.0002, "epoch": 3.771724193678765, "step": 23330}, {"loss": 0.658, "grad_norm": 1.0768941640853882, "learning_rate": 0.0002, "epoch": 3.77334087785951, "step": 23340}, {"loss": 0.6408, "grad_norm": 0.9956819415092468, "learning_rate": 0.0002, "epoch": 3.7749575620402553, "step": 23350}, {"loss": 0.6464, "grad_norm": 0.9234658479690552, "learning_rate": 0.0002, "epoch": 3.7765742462210006, "step": 23360}, {"loss": 0.6542, "grad_norm": 1.0993858575820923, "learning_rate": 0.0002, "epoch": 3.7781909304017463, "step": 23370}, {"loss": 0.6391, "grad_norm": 0.923159658908844, "learning_rate": 0.0002, "epoch": 3.7798076145824915, "step": 23380}, {"loss": 0.6625, "grad_norm": 0.9311541318893433, "learning_rate": 0.0002, "epoch": 3.7814242987632367, "step": 23390}, {"loss": 0.6535, "grad_norm": 0.919681191444397, "learning_rate": 0.0002, "epoch": 3.783040982943982, "step": 23400}, {"loss": 0.6138, "grad_norm": 1.7406195402145386, "learning_rate": 0.0002, "epoch": 3.784657667124727, "step": 23410}, {"loss": 0.657, "grad_norm": 0.7789074182510376, "learning_rate": 0.0002, "epoch": 3.7862743513054724, "step": 23420}, {"loss": 0.658, "grad_norm": 0.8302814960479736, "learning_rate": 0.0002, "epoch": 3.7878910354862176, "step": 23430}, {"loss": 0.649, "grad_norm": 0.8089349269866943, "learning_rate": 0.0002, "epoch": 3.789507719666963, "step": 23440}, {"loss": 0.6682, "grad_norm": 0.9006284475326538, "learning_rate": 0.0002, "epoch": 3.791124403847708, "step": 23450}, {"loss": 0.6335, "grad_norm": 0.8426766991615295, "learning_rate": 0.0002, "epoch": 3.7927410880284538, "step": 23460}, {"loss": 0.6364, "grad_norm": 1.2576252222061157, "learning_rate": 0.0002, "epoch": 3.794357772209199, "step": 23470}, {"loss": 0.6324, "grad_norm": 1.0307610034942627, "learning_rate": 0.0002, "epoch": 3.7959744563899442, "step": 23480}, {"loss": 0.6262, "grad_norm": 0.8525972962379456, "learning_rate": 0.0002, "epoch": 3.7975911405706895, "step": 23490}, {"loss": 0.6757, "grad_norm": 1.159039855003357, "learning_rate": 0.0002, "epoch": 3.7992078247514347, "step": 23500}, {"loss": 0.6414, "grad_norm": 1.4193549156188965, "learning_rate": 0.0002, "epoch": 3.80082450893218, "step": 23510}, {"loss": 0.6413, "grad_norm": 0.8245543837547302, "learning_rate": 0.0002, "epoch": 3.8024411931129256, "step": 23520}, {"loss": 0.6417, "grad_norm": 0.8847230076789856, "learning_rate": 0.0002, "epoch": 3.804057877293671, "step": 23530}, {"loss": 0.6415, "grad_norm": 0.9574624300003052, "learning_rate": 0.0002, "epoch": 3.805674561474416, "step": 23540}, {"loss": 0.5765, "grad_norm": 1.048020601272583, "learning_rate": 0.0002, "epoch": 3.8072912456551613, "step": 23550}, {"loss": 0.6497, "grad_norm": 0.8302255868911743, "learning_rate": 0.0002, "epoch": 3.8089079298359065, "step": 23560}, {"loss": 0.6534, "grad_norm": 0.8269215822219849, "learning_rate": 0.0002, "epoch": 3.8105246140166518, "step": 23570}, {"loss": 0.6294, "grad_norm": 0.9375753402709961, "learning_rate": 0.0002, "epoch": 3.812141298197397, "step": 23580}, {"loss": 0.6132, "grad_norm": 1.0234097242355347, "learning_rate": 0.0002, "epoch": 3.8137579823781422, "step": 23590}, {"loss": 0.6625, "grad_norm": 0.8978445529937744, "learning_rate": 0.0002, "epoch": 3.8153746665588875, "step": 23600}, {"loss": 0.6315, "grad_norm": 0.7929515838623047, "learning_rate": 0.0002, "epoch": 3.816991350739633, "step": 23610}, {"loss": 0.6387, "grad_norm": 1.3255881071090698, "learning_rate": 0.0002, "epoch": 3.8186080349203784, "step": 23620}, {"loss": 0.5947, "grad_norm": 0.9188598990440369, "learning_rate": 0.0002, "epoch": 3.8202247191011236, "step": 23630}, {"loss": 0.6152, "grad_norm": 0.8811675906181335, "learning_rate": 0.0002, "epoch": 3.821841403281869, "step": 23640}, {"loss": 0.6253, "grad_norm": 0.8061038255691528, "learning_rate": 0.0002, "epoch": 3.823458087462614, "step": 23650}, {"loss": 0.6517, "grad_norm": 0.9975376129150391, "learning_rate": 0.0002, "epoch": 3.8250747716433597, "step": 23660}, {"loss": 0.6288, "grad_norm": 0.8036105036735535, "learning_rate": 0.0002, "epoch": 3.826691455824105, "step": 23670}, {"loss": 0.6845, "grad_norm": 0.7401984333992004, "learning_rate": 0.0002, "epoch": 3.82830814000485, "step": 23680}, {"loss": 0.6423, "grad_norm": 0.829753041267395, "learning_rate": 0.0002, "epoch": 3.8299248241855954, "step": 23690}, {"loss": 0.6611, "grad_norm": 0.8753240704536438, "learning_rate": 0.0002, "epoch": 3.8315415083663407, "step": 23700}, {"loss": 0.6686, "grad_norm": 0.8157842755317688, "learning_rate": 0.0002, "epoch": 3.833158192547086, "step": 23710}, {"loss": 0.6181, "grad_norm": 0.6183798909187317, "learning_rate": 0.0002, "epoch": 3.834774876727831, "step": 23720}, {"loss": 0.5965, "grad_norm": 0.9548442363739014, "learning_rate": 0.0002, "epoch": 3.8363915609085764, "step": 23730}, {"loss": 0.6456, "grad_norm": 0.8319669961929321, "learning_rate": 0.0002, "epoch": 3.8380082450893216, "step": 23740}, {"loss": 0.6585, "grad_norm": 0.9718693494796753, "learning_rate": 0.0002, "epoch": 3.839624929270067, "step": 23750}, {"loss": 0.6518, "grad_norm": 0.8672235012054443, "learning_rate": 0.0002, "epoch": 3.8412416134508125, "step": 23760}, {"loss": 0.6774, "grad_norm": 1.1210707426071167, "learning_rate": 0.0002, "epoch": 3.8428582976315577, "step": 23770}, {"loss": 0.5923, "grad_norm": 0.9177767634391785, "learning_rate": 0.0002, "epoch": 3.844474981812303, "step": 23780}, {"loss": 0.6286, "grad_norm": 0.8714171648025513, "learning_rate": 0.0002, "epoch": 3.846091665993048, "step": 23790}, {"loss": 0.6302, "grad_norm": 1.1853246688842773, "learning_rate": 0.0002, "epoch": 3.8477083501737934, "step": 23800}, {"loss": 0.6144, "grad_norm": 0.8091260194778442, "learning_rate": 0.0002, "epoch": 3.849325034354539, "step": 23810}, {"loss": 0.658, "grad_norm": 0.9710774421691895, "learning_rate": 0.0002, "epoch": 3.8509417185352843, "step": 23820}, {"loss": 0.6151, "grad_norm": 0.7648707628250122, "learning_rate": 0.0002, "epoch": 3.8525584027160296, "step": 23830}, {"loss": 0.6013, "grad_norm": 0.7809253931045532, "learning_rate": 0.0002, "epoch": 3.854175086896775, "step": 23840}, {"loss": 0.6006, "grad_norm": 0.8337951898574829, "learning_rate": 0.0002, "epoch": 3.85579177107752, "step": 23850}, {"loss": 0.6456, "grad_norm": 0.9271913170814514, "learning_rate": 0.0002, "epoch": 3.8574084552582653, "step": 23860}, {"loss": 0.6671, "grad_norm": 0.985334038734436, "learning_rate": 0.0002, "epoch": 3.8590251394390105, "step": 23870}, {"loss": 0.6693, "grad_norm": 0.8458583354949951, "learning_rate": 0.0002, "epoch": 3.8606418236197557, "step": 23880}, {"loss": 0.6207, "grad_norm": 1.015348196029663, "learning_rate": 0.0002, "epoch": 3.862258507800501, "step": 23890}, {"loss": 0.649, "grad_norm": 1.0121688842773438, "learning_rate": 0.0002, "epoch": 3.8638751919812466, "step": 23900}, {"loss": 0.5921, "grad_norm": 0.8883971571922302, "learning_rate": 0.0002, "epoch": 3.865491876161992, "step": 23910}, {"loss": 0.6597, "grad_norm": 1.028086543083191, "learning_rate": 0.0002, "epoch": 3.867108560342737, "step": 23920}, {"loss": 0.6654, "grad_norm": 0.9645734429359436, "learning_rate": 0.0002, "epoch": 3.8687252445234823, "step": 23930}, {"loss": 0.6328, "grad_norm": 0.8235350251197815, "learning_rate": 0.0002, "epoch": 3.8703419287042276, "step": 23940}, {"loss": 0.6387, "grad_norm": 1.0298916101455688, "learning_rate": 0.0002, "epoch": 3.871958612884973, "step": 23950}, {"loss": 0.5966, "grad_norm": 1.0063377618789673, "learning_rate": 0.0002, "epoch": 3.8735752970657185, "step": 23960}, {"loss": 0.6234, "grad_norm": 0.9230626821517944, "learning_rate": 0.0002, "epoch": 3.8751919812464637, "step": 23970}, {"loss": 0.6159, "grad_norm": 0.9243063926696777, "learning_rate": 0.0002, "epoch": 3.876808665427209, "step": 23980}, {"loss": 0.6035, "grad_norm": 1.0211291313171387, "learning_rate": 0.0002, "epoch": 3.878425349607954, "step": 23990}, {"loss": 0.6351, "grad_norm": 0.7800535559654236, "learning_rate": 0.0002, "epoch": 3.8800420337886994, "step": 24000}, {"loss": 0.7, "grad_norm": 0.7904248833656311, "learning_rate": 0.0002, "epoch": 3.8816587179694446, "step": 24010}, {"loss": 0.6516, "grad_norm": 1.1975988149642944, "learning_rate": 0.0002, "epoch": 3.88327540215019, "step": 24020}, {"loss": 0.6006, "grad_norm": 1.0626593828201294, "learning_rate": 0.0002, "epoch": 3.884892086330935, "step": 24030}, {"loss": 0.6115, "grad_norm": 0.9012193083763123, "learning_rate": 0.0002, "epoch": 3.8865087705116803, "step": 24040}, {"loss": 0.6786, "grad_norm": 1.1159172058105469, "learning_rate": 0.0002, "epoch": 3.888125454692426, "step": 24050}, {"loss": 0.6635, "grad_norm": 1.276838779449463, "learning_rate": 0.0002, "epoch": 3.889742138873171, "step": 24060}, {"loss": 0.5985, "grad_norm": 0.8467690348625183, "learning_rate": 0.0002, "epoch": 3.8913588230539164, "step": 24070}, {"loss": 0.6655, "grad_norm": 0.9862841963768005, "learning_rate": 0.0002, "epoch": 3.8929755072346617, "step": 24080}, {"loss": 0.6098, "grad_norm": 0.7134621739387512, "learning_rate": 0.0002, "epoch": 3.894592191415407, "step": 24090}, {"loss": 0.618, "grad_norm": 0.8178175091743469, "learning_rate": 0.0002, "epoch": 3.896208875596152, "step": 24100}, {"loss": 0.6147, "grad_norm": 0.9229172468185425, "learning_rate": 0.0002, "epoch": 3.897825559776898, "step": 24110}, {"loss": 0.6554, "grad_norm": 1.0878316164016724, "learning_rate": 0.0002, "epoch": 3.899442243957643, "step": 24120}, {"loss": 0.6616, "grad_norm": 0.971645712852478, "learning_rate": 0.0002, "epoch": 3.9010589281383883, "step": 24130}, {"loss": 0.6228, "grad_norm": 0.8862188458442688, "learning_rate": 0.0002, "epoch": 3.9026756123191335, "step": 24140}, {"loss": 0.6192, "grad_norm": 0.9126982688903809, "learning_rate": 0.0002, "epoch": 3.9042922964998787, "step": 24150}, {"loss": 0.6734, "grad_norm": 0.8833470940589905, "learning_rate": 0.0002, "epoch": 3.905908980680624, "step": 24160}, {"loss": 0.5832, "grad_norm": 0.8320947885513306, "learning_rate": 0.0002, "epoch": 3.907525664861369, "step": 24170}, {"loss": 0.6247, "grad_norm": 0.9156602025032043, "learning_rate": 0.0002, "epoch": 3.9091423490421144, "step": 24180}, {"loss": 0.6678, "grad_norm": 1.029181957244873, "learning_rate": 0.0002, "epoch": 3.9107590332228597, "step": 24190}, {"loss": 0.6565, "grad_norm": 0.9052802324295044, "learning_rate": 0.0002, "epoch": 3.9123757174036053, "step": 24200}, {"loss": 0.6346, "grad_norm": 0.8847255110740662, "learning_rate": 0.0002, "epoch": 3.9139924015843506, "step": 24210}, {"loss": 0.6343, "grad_norm": 0.9642062187194824, "learning_rate": 0.0002, "epoch": 3.915609085765096, "step": 24220}, {"loss": 0.6557, "grad_norm": 0.8629093766212463, "learning_rate": 0.0002, "epoch": 3.917225769945841, "step": 24230}, {"loss": 0.6086, "grad_norm": 0.8674976825714111, "learning_rate": 0.0002, "epoch": 3.9188424541265863, "step": 24240}, {"loss": 0.5874, "grad_norm": 1.104846477508545, "learning_rate": 0.0002, "epoch": 3.9204591383073315, "step": 24250}, {"loss": 0.6501, "grad_norm": 1.0874955654144287, "learning_rate": 0.0002, "epoch": 3.922075822488077, "step": 24260}, {"loss": 0.6455, "grad_norm": 0.8689812421798706, "learning_rate": 0.0002, "epoch": 3.9236925066688224, "step": 24270}, {"loss": 0.5893, "grad_norm": 0.9724617004394531, "learning_rate": 0.0002, "epoch": 3.9253091908495676, "step": 24280}, {"loss": 0.6616, "grad_norm": 0.9165538549423218, "learning_rate": 0.0002, "epoch": 3.926925875030313, "step": 24290}, {"loss": 0.645, "grad_norm": 0.9307710528373718, "learning_rate": 0.0002, "epoch": 3.928542559211058, "step": 24300}, {"loss": 0.6071, "grad_norm": 0.8589295148849487, "learning_rate": 0.0002, "epoch": 3.9301592433918033, "step": 24310}, {"loss": 0.6662, "grad_norm": 0.9151099920272827, "learning_rate": 0.0002, "epoch": 3.9317759275725486, "step": 24320}, {"loss": 0.7075, "grad_norm": 0.9633517265319824, "learning_rate": 0.0002, "epoch": 3.933392611753294, "step": 24330}, {"loss": 0.6432, "grad_norm": 0.9521116018295288, "learning_rate": 0.0002, "epoch": 3.935009295934039, "step": 24340}, {"loss": 0.6457, "grad_norm": 0.8366776704788208, "learning_rate": 0.0002, "epoch": 3.9366259801147847, "step": 24350}, {"loss": 0.6139, "grad_norm": 0.8972663283348083, "learning_rate": 0.0002, "epoch": 3.93824266429553, "step": 24360}, {"loss": 0.661, "grad_norm": 0.8102919459342957, "learning_rate": 0.0002, "epoch": 3.939859348476275, "step": 24370}, {"loss": 0.6388, "grad_norm": 0.8189975023269653, "learning_rate": 0.0002, "epoch": 3.9414760326570204, "step": 24380}, {"loss": 0.6818, "grad_norm": 0.9569464921951294, "learning_rate": 0.0002, "epoch": 3.9430927168377656, "step": 24390}, {"loss": 0.6999, "grad_norm": 0.7459101676940918, "learning_rate": 0.0002, "epoch": 3.9447094010185113, "step": 24400}, {"loss": 0.6069, "grad_norm": 0.8536974787712097, "learning_rate": 0.0002, "epoch": 3.9463260851992565, "step": 24410}, {"loss": 0.5683, "grad_norm": 0.8763698935508728, "learning_rate": 0.0002, "epoch": 3.9479427693800018, "step": 24420}, {"loss": 0.6478, "grad_norm": 0.9381106495857239, "learning_rate": 0.0002, "epoch": 3.949559453560747, "step": 24430}, {"loss": 0.6371, "grad_norm": 0.934440016746521, "learning_rate": 0.0002, "epoch": 3.9511761377414922, "step": 24440}, {"loss": 0.6393, "grad_norm": 0.903918981552124, "learning_rate": 0.0002, "epoch": 3.9527928219222375, "step": 24450}, {"loss": 0.6175, "grad_norm": 0.8771953582763672, "learning_rate": 0.0002, "epoch": 3.9544095061029827, "step": 24460}, {"loss": 0.6971, "grad_norm": 1.0375410318374634, "learning_rate": 0.0002, "epoch": 3.956026190283728, "step": 24470}, {"loss": 0.6313, "grad_norm": 0.9439185261726379, "learning_rate": 0.0002, "epoch": 3.957642874464473, "step": 24480}, {"loss": 0.6076, "grad_norm": 0.935467004776001, "learning_rate": 0.0002, "epoch": 3.9592595586452184, "step": 24490}, {"loss": 0.6437, "grad_norm": 0.6900772452354431, "learning_rate": 0.0002, "epoch": 3.960876242825964, "step": 24500}, {"loss": 0.6445, "grad_norm": 1.0172916650772095, "learning_rate": 0.0002, "epoch": 3.9624929270067093, "step": 24510}, {"loss": 0.6308, "grad_norm": 0.9167046546936035, "learning_rate": 0.0002, "epoch": 3.9641096111874545, "step": 24520}, {"loss": 0.6519, "grad_norm": 0.7230527997016907, "learning_rate": 0.0002, "epoch": 3.9657262953681998, "step": 24530}, {"loss": 0.6564, "grad_norm": 0.8980403542518616, "learning_rate": 0.0002, "epoch": 3.967342979548945, "step": 24540}, {"loss": 0.6099, "grad_norm": 0.8555465936660767, "learning_rate": 0.0002, "epoch": 3.9689596637296907, "step": 24550}, {"loss": 0.6617, "grad_norm": 0.7825445532798767, "learning_rate": 0.0002, "epoch": 3.970576347910436, "step": 24560}, {"loss": 0.604, "grad_norm": 0.7273133993148804, "learning_rate": 0.0002, "epoch": 3.972193032091181, "step": 24570}, {"loss": 0.6427, "grad_norm": 0.9612047672271729, "learning_rate": 0.0002, "epoch": 3.9738097162719264, "step": 24580}, {"loss": 0.6426, "grad_norm": 0.9865460991859436, "learning_rate": 0.0002, "epoch": 3.9754264004526716, "step": 24590}, {"loss": 0.6052, "grad_norm": 0.8638762831687927, "learning_rate": 0.0002, "epoch": 3.977043084633417, "step": 24600}, {"loss": 0.6097, "grad_norm": 1.0096198320388794, "learning_rate": 0.0002, "epoch": 3.978659768814162, "step": 24610}, {"loss": 0.6664, "grad_norm": 0.8475532531738281, "learning_rate": 0.0002, "epoch": 3.9802764529949073, "step": 24620}, {"loss": 0.6711, "grad_norm": 0.9696195721626282, "learning_rate": 0.0002, "epoch": 3.9818931371756525, "step": 24630}, {"loss": 0.6446, "grad_norm": 0.7499843239784241, "learning_rate": 0.0002, "epoch": 3.9835098213563978, "step": 24640}, {"loss": 0.6054, "grad_norm": 0.8865424990653992, "learning_rate": 0.0002, "epoch": 3.9851265055371434, "step": 24650}, {"loss": 0.5975, "grad_norm": 0.8089959025382996, "learning_rate": 0.0002, "epoch": 3.9867431897178887, "step": 24660}, {"loss": 0.6677, "grad_norm": 0.6946012377738953, "learning_rate": 0.0002, "epoch": 3.988359873898634, "step": 24670}, {"loss": 0.6329, "grad_norm": 0.7991759181022644, "learning_rate": 0.0002, "epoch": 3.989976558079379, "step": 24680}, {"loss": 0.6449, "grad_norm": 0.8803931474685669, "learning_rate": 0.0002, "epoch": 3.9915932422601244, "step": 24690}, {"loss": 0.7091, "grad_norm": 0.8848299980163574, "learning_rate": 0.0002, "epoch": 3.99320992644087, "step": 24700}, {"loss": 0.6551, "grad_norm": 0.7448889017105103, "learning_rate": 0.0002, "epoch": 3.9948266106216153, "step": 24710}, {"loss": 0.6432, "grad_norm": 0.9361620545387268, "learning_rate": 0.0002, "epoch": 3.9964432948023605, "step": 24720}, {"loss": 0.5917, "grad_norm": 0.9958081245422363, "learning_rate": 0.0002, "epoch": 3.9980599789831057, "step": 24730}, {"loss": 0.6567, "grad_norm": 1.026004672050476, "learning_rate": 0.0002, "epoch": 3.999676663163851, "step": 24740}, {"eval_loss": 1.1524168252944946, "eval_runtime": 122.1585, "eval_samples_per_second": 6.0, "eval_steps_per_second": 0.753, "epoch": 4.0, "step": 24742}, {"loss": 0.6057, "grad_norm": 1.0664808750152588, "learning_rate": 0.0002, "epoch": 4.001293347344596, "step": 24750}, {"loss": 0.5644, "grad_norm": 1.0113720893859863, "learning_rate": 0.0002, "epoch": 4.002910031525341, "step": 24760}, {"loss": 0.5628, "grad_norm": 0.991486668586731, "learning_rate": 0.0002, "epoch": 4.004526715706087, "step": 24770}, {"loss": 0.508, "grad_norm": 0.951754629611969, "learning_rate": 0.0002, "epoch": 4.006143399886832, "step": 24780}, {"loss": 0.5314, "grad_norm": 1.13059401512146, "learning_rate": 0.0002, "epoch": 4.007760084067577, "step": 24790}, {"loss": 0.5323, "grad_norm": 0.9343926310539246, "learning_rate": 0.0002, "epoch": 4.009376768248322, "step": 24800}, {"loss": 0.5161, "grad_norm": 1.0680590867996216, "learning_rate": 0.0002, "epoch": 4.010993452429068, "step": 24810}, {"loss": 0.513, "grad_norm": 1.0022706985473633, "learning_rate": 0.0002, "epoch": 4.012610136609814, "step": 24820}, {"loss": 0.543, "grad_norm": 1.0285297632217407, "learning_rate": 0.0002, "epoch": 4.014226820790559, "step": 24830}, {"loss": 0.5311, "grad_norm": 0.8347002863883972, "learning_rate": 0.0002, "epoch": 4.015843504971304, "step": 24840}, {"loss": 0.5655, "grad_norm": 0.9675396680831909, "learning_rate": 0.0002, "epoch": 4.017460189152049, "step": 24850}, {"loss": 0.5625, "grad_norm": 0.9238511323928833, "learning_rate": 0.0002, "epoch": 4.019076873332795, "step": 24860}, {"loss": 0.5327, "grad_norm": 1.1576941013336182, "learning_rate": 0.0002, "epoch": 4.02069355751354, "step": 24870}, {"loss": 0.5533, "grad_norm": 0.8583757281303406, "learning_rate": 0.0002, "epoch": 4.022310241694285, "step": 24880}, {"loss": 0.5483, "grad_norm": 0.9816817045211792, "learning_rate": 0.0002, "epoch": 4.02392692587503, "step": 24890}, {"loss": 0.5605, "grad_norm": 0.955073893070221, "learning_rate": 0.0002, "epoch": 4.0255436100557755, "step": 24900}, {"loss": 0.4896, "grad_norm": 1.1054974794387817, "learning_rate": 0.0002, "epoch": 4.027160294236521, "step": 24910}, {"loss": 0.5246, "grad_norm": 1.1240060329437256, "learning_rate": 0.0002, "epoch": 4.028776978417266, "step": 24920}, {"loss": 0.5451, "grad_norm": 0.9512825012207031, "learning_rate": 0.0002, "epoch": 4.030393662598011, "step": 24930}, {"loss": 0.5584, "grad_norm": 0.85965496301651, "learning_rate": 0.0002, "epoch": 4.0320103467787565, "step": 24940}, {"loss": 0.5564, "grad_norm": 0.9378061294555664, "learning_rate": 0.0002, "epoch": 4.033627030959502, "step": 24950}, {"loss": 0.5008, "grad_norm": 0.9655424356460571, "learning_rate": 0.0002, "epoch": 4.035243715140247, "step": 24960}, {"loss": 0.5538, "grad_norm": 1.1393707990646362, "learning_rate": 0.0002, "epoch": 4.036860399320993, "step": 24970}, {"loss": 0.5785, "grad_norm": 1.0220451354980469, "learning_rate": 0.0002, "epoch": 4.038477083501738, "step": 24980}, {"loss": 0.5813, "grad_norm": 0.9785808324813843, "learning_rate": 0.0002, "epoch": 4.0400937676824835, "step": 24990}, {"loss": 0.5153, "grad_norm": 1.0257649421691895, "learning_rate": 0.0002, "epoch": 4.041710451863229, "step": 25000}, {"loss": 0.5658, "grad_norm": 0.9737892150878906, "learning_rate": 0.0002, "epoch": 4.043327136043974, "step": 25010}, {"loss": 0.5515, "grad_norm": 0.7416959404945374, "learning_rate": 0.0002, "epoch": 4.044943820224719, "step": 25020}, {"loss": 0.5372, "grad_norm": 0.7909596562385559, "learning_rate": 0.0002, "epoch": 4.046560504405464, "step": 25030}, {"loss": 0.5265, "grad_norm": 0.8923130631446838, "learning_rate": 0.0002, "epoch": 4.04817718858621, "step": 25040}, {"loss": 0.5035, "grad_norm": 0.9044941663742065, "learning_rate": 0.0002, "epoch": 4.049793872766955, "step": 25050}, {"loss": 0.5135, "grad_norm": 0.866352379322052, "learning_rate": 0.0002, "epoch": 4.0514105569477, "step": 25060}, {"loss": 0.5956, "grad_norm": 1.544549822807312, "learning_rate": 0.0002, "epoch": 4.053027241128445, "step": 25070}, {"loss": 0.5418, "grad_norm": 0.8426995277404785, "learning_rate": 0.0002, "epoch": 4.054643925309191, "step": 25080}, {"loss": 0.5537, "grad_norm": 0.9797548651695251, "learning_rate": 0.0002, "epoch": 4.056260609489936, "step": 25090}, {"loss": 0.55, "grad_norm": 0.8468434810638428, "learning_rate": 0.0002, "epoch": 4.057877293670681, "step": 25100}, {"loss": 0.5242, "grad_norm": 0.9294559955596924, "learning_rate": 0.0002, "epoch": 4.059493977851426, "step": 25110}, {"loss": 0.5295, "grad_norm": 0.9686688780784607, "learning_rate": 0.0002, "epoch": 4.061110662032172, "step": 25120}, {"loss": 0.5642, "grad_norm": 0.8042728304862976, "learning_rate": 0.0002, "epoch": 4.062727346212918, "step": 25130}, {"loss": 0.548, "grad_norm": 1.165160894393921, "learning_rate": 0.0002, "epoch": 4.064344030393663, "step": 25140}, {"loss": 0.5473, "grad_norm": 1.2161961793899536, "learning_rate": 0.0002, "epoch": 4.065960714574408, "step": 25150}, {"loss": 0.5217, "grad_norm": 1.0762810707092285, "learning_rate": 0.0002, "epoch": 4.067577398755153, "step": 25160}, {"loss": 0.5886, "grad_norm": 0.7580869793891907, "learning_rate": 0.0002, "epoch": 4.069194082935899, "step": 25170}, {"loss": 0.5401, "grad_norm": 0.9630117416381836, "learning_rate": 0.0002, "epoch": 4.070810767116644, "step": 25180}, {"loss": 0.5378, "grad_norm": 0.9049716591835022, "learning_rate": 0.0002, "epoch": 4.072427451297389, "step": 25190}, {"loss": 0.5266, "grad_norm": 1.1536930799484253, "learning_rate": 0.0002, "epoch": 4.074044135478134, "step": 25200}, {"loss": 0.5523, "grad_norm": 0.901461124420166, "learning_rate": 0.0002, "epoch": 4.0756608196588795, "step": 25210}, {"loss": 0.5132, "grad_norm": 1.3318437337875366, "learning_rate": 0.0002, "epoch": 4.077277503839625, "step": 25220}, {"loss": 0.5317, "grad_norm": 0.8811455368995667, "learning_rate": 0.0002, "epoch": 4.07889418802037, "step": 25230}, {"loss": 0.5798, "grad_norm": 1.0564165115356445, "learning_rate": 0.0002, "epoch": 4.080510872201115, "step": 25240}, {"loss": 0.5472, "grad_norm": 1.1008027791976929, "learning_rate": 0.0002, "epoch": 4.08212755638186, "step": 25250}, {"loss": 0.5195, "grad_norm": 1.150097131729126, "learning_rate": 0.0002, "epoch": 4.083744240562606, "step": 25260}, {"loss": 0.5321, "grad_norm": 0.9339924454689026, "learning_rate": 0.0002, "epoch": 4.085360924743352, "step": 25270}, {"loss": 0.5597, "grad_norm": 1.0902045965194702, "learning_rate": 0.0002, "epoch": 4.086977608924097, "step": 25280}, {"loss": 0.5203, "grad_norm": 0.8483911156654358, "learning_rate": 0.0002, "epoch": 4.088594293104842, "step": 25290}, {"loss": 0.5697, "grad_norm": 0.9477024674415588, "learning_rate": 0.0002, "epoch": 4.0902109772855875, "step": 25300}, {"loss": 0.5384, "grad_norm": 0.9500215649604797, "learning_rate": 0.0002, "epoch": 4.091827661466333, "step": 25310}, {"loss": 0.5045, "grad_norm": 1.040468454360962, "learning_rate": 0.0002, "epoch": 4.093444345647078, "step": 25320}, {"loss": 0.5488, "grad_norm": 0.7457592487335205, "learning_rate": 0.0002, "epoch": 4.095061029827823, "step": 25330}, {"loss": 0.609, "grad_norm": 1.2092097997665405, "learning_rate": 0.0002, "epoch": 4.096677714008568, "step": 25340}, {"loss": 0.5174, "grad_norm": 0.9652107954025269, "learning_rate": 0.0002, "epoch": 4.098294398189314, "step": 25350}, {"loss": 0.5559, "grad_norm": 0.8464955687522888, "learning_rate": 0.0002, "epoch": 4.099911082370059, "step": 25360}, {"loss": 0.5635, "grad_norm": 0.875026285648346, "learning_rate": 0.0002, "epoch": 4.101527766550804, "step": 25370}, {"loss": 0.5774, "grad_norm": 0.9241740107536316, "learning_rate": 0.0002, "epoch": 4.103144450731549, "step": 25380}, {"loss": 0.5578, "grad_norm": 0.9769546389579773, "learning_rate": 0.0002, "epoch": 4.1047611349122946, "step": 25390}, {"loss": 0.567, "grad_norm": 1.1501960754394531, "learning_rate": 0.0002, "epoch": 4.10637781909304, "step": 25400}, {"loss": 0.5241, "grad_norm": 0.9135243892669678, "learning_rate": 0.0002, "epoch": 4.107994503273786, "step": 25410}, {"loss": 0.5152, "grad_norm": 0.9905396103858948, "learning_rate": 0.0002, "epoch": 4.109611187454531, "step": 25420}, {"loss": 0.5064, "grad_norm": 0.9845104217529297, "learning_rate": 0.0002, "epoch": 4.111227871635276, "step": 25430}, {"loss": 0.5029, "grad_norm": 0.8326883912086487, "learning_rate": 0.0002, "epoch": 4.112844555816022, "step": 25440}, {"loss": 0.5312, "grad_norm": 0.9264556765556335, "learning_rate": 0.0002, "epoch": 4.114461239996767, "step": 25450}, {"loss": 0.5968, "grad_norm": 1.043080449104309, "learning_rate": 0.0002, "epoch": 4.116077924177512, "step": 25460}, {"loss": 0.5773, "grad_norm": 0.8533386588096619, "learning_rate": 0.0002, "epoch": 4.117694608358257, "step": 25470}, {"loss": 0.5584, "grad_norm": 1.0133965015411377, "learning_rate": 0.0002, "epoch": 4.1193112925390025, "step": 25480}, {"loss": 0.566, "grad_norm": 0.7476310133934021, "learning_rate": 0.0002, "epoch": 4.120927976719748, "step": 25490}, {"loss": 0.5189, "grad_norm": 1.1247259378433228, "learning_rate": 0.0002, "epoch": 4.122544660900493, "step": 25500}, {"loss": 0.5751, "grad_norm": 1.0764678716659546, "learning_rate": 0.0002, "epoch": 4.124161345081238, "step": 25510}, {"loss": 0.5391, "grad_norm": 0.7679798007011414, "learning_rate": 0.0002, "epoch": 4.1257780292619834, "step": 25520}, {"loss": 0.5233, "grad_norm": 0.8877071142196655, "learning_rate": 0.0002, "epoch": 4.127394713442729, "step": 25530}, {"loss": 0.5769, "grad_norm": 1.0440239906311035, "learning_rate": 0.0002, "epoch": 4.129011397623474, "step": 25540}, {"loss": 0.5723, "grad_norm": 0.984145998954773, "learning_rate": 0.0002, "epoch": 4.130628081804219, "step": 25550}, {"loss": 0.5741, "grad_norm": 0.8667055368423462, "learning_rate": 0.0002, "epoch": 4.132244765984965, "step": 25560}, {"loss": 0.5816, "grad_norm": 1.1300835609436035, "learning_rate": 0.0002, "epoch": 4.1338614501657105, "step": 25570}, {"loss": 0.524, "grad_norm": 0.9314348101615906, "learning_rate": 0.0002, "epoch": 4.135478134346456, "step": 25580}, {"loss": 0.5283, "grad_norm": 0.7731879949569702, "learning_rate": 0.0002, "epoch": 4.137094818527201, "step": 25590}, {"loss": 0.5307, "grad_norm": 1.0080097913742065, "learning_rate": 0.0002, "epoch": 4.138711502707946, "step": 25600}, {"loss": 0.5759, "grad_norm": 1.2475038766860962, "learning_rate": 0.0002, "epoch": 4.140328186888691, "step": 25610}, {"loss": 0.55, "grad_norm": 0.9912930727005005, "learning_rate": 0.0002, "epoch": 4.141944871069437, "step": 25620}, {"loss": 0.5624, "grad_norm": 0.9088651537895203, "learning_rate": 0.0002, "epoch": 4.143561555250182, "step": 25630}, {"loss": 0.5393, "grad_norm": 0.8940697312355042, "learning_rate": 0.0002, "epoch": 4.145178239430927, "step": 25640}, {"loss": 0.5341, "grad_norm": 1.0798203945159912, "learning_rate": 0.0002, "epoch": 4.146794923611672, "step": 25650}, {"loss": 0.5987, "grad_norm": 0.955172061920166, "learning_rate": 0.0002, "epoch": 4.148411607792418, "step": 25660}, {"loss": 0.569, "grad_norm": 0.9692716002464294, "learning_rate": 0.0002, "epoch": 4.150028291973163, "step": 25670}, {"loss": 0.5478, "grad_norm": 1.0813939571380615, "learning_rate": 0.0002, "epoch": 4.151644976153908, "step": 25680}, {"loss": 0.5383, "grad_norm": 1.135675072669983, "learning_rate": 0.0002, "epoch": 4.153261660334653, "step": 25690}, {"loss": 0.5247, "grad_norm": 1.0392236709594727, "learning_rate": 0.0002, "epoch": 4.1548783445153985, "step": 25700}, {"loss": 0.5204, "grad_norm": 0.9473116993904114, "learning_rate": 0.0002, "epoch": 4.156495028696145, "step": 25710}, {"loss": 0.5339, "grad_norm": 0.712493896484375, "learning_rate": 0.0002, "epoch": 4.15811171287689, "step": 25720}, {"loss": 0.5781, "grad_norm": 0.8724465370178223, "learning_rate": 0.0002, "epoch": 4.159728397057635, "step": 25730}, {"loss": 0.5325, "grad_norm": 0.9870015978813171, "learning_rate": 0.0002, "epoch": 4.16134508123838, "step": 25740}, {"loss": 0.5503, "grad_norm": 1.025273084640503, "learning_rate": 0.0002, "epoch": 4.1629617654191255, "step": 25750}, {"loss": 0.5223, "grad_norm": 0.9243090152740479, "learning_rate": 0.0002, "epoch": 4.164578449599871, "step": 25760}, {"loss": 0.5177, "grad_norm": 1.1656451225280762, "learning_rate": 0.0002, "epoch": 4.166195133780616, "step": 25770}, {"loss": 0.5334, "grad_norm": 0.936358630657196, "learning_rate": 0.0002, "epoch": 4.167811817961361, "step": 25780}, {"loss": 0.5236, "grad_norm": 0.8618208169937134, "learning_rate": 0.0002, "epoch": 4.1694285021421065, "step": 25790}, {"loss": 0.5186, "grad_norm": 0.8580600023269653, "learning_rate": 0.0002, "epoch": 4.171045186322852, "step": 25800}, {"loss": 0.5212, "grad_norm": 1.0128562450408936, "learning_rate": 0.0002, "epoch": 4.172661870503597, "step": 25810}, {"loss": 0.5404, "grad_norm": 0.854865312576294, "learning_rate": 0.0002, "epoch": 4.174278554684342, "step": 25820}, {"loss": 0.5377, "grad_norm": 1.235082745552063, "learning_rate": 0.0002, "epoch": 4.175895238865087, "step": 25830}, {"loss": 0.5614, "grad_norm": 0.9796220660209656, "learning_rate": 0.0002, "epoch": 4.177511923045833, "step": 25840}, {"loss": 0.5689, "grad_norm": 0.8922094702720642, "learning_rate": 0.0002, "epoch": 4.179128607226578, "step": 25850}, {"loss": 0.5806, "grad_norm": 0.9672530293464661, "learning_rate": 0.0002, "epoch": 4.180745291407324, "step": 25860}, {"loss": 0.5074, "grad_norm": 0.8662548661231995, "learning_rate": 0.0002, "epoch": 4.182361975588069, "step": 25870}, {"loss": 0.5329, "grad_norm": 0.7938798069953918, "learning_rate": 0.0002, "epoch": 4.1839786597688144, "step": 25880}, {"loss": 0.5427, "grad_norm": 1.0517958402633667, "learning_rate": 0.0002, "epoch": 4.18559534394956, "step": 25890}, {"loss": 0.5147, "grad_norm": 0.8939275145530701, "learning_rate": 0.0002, "epoch": 4.187212028130305, "step": 25900}, {"loss": 0.5199, "grad_norm": 1.0296672582626343, "learning_rate": 0.0002, "epoch": 4.18882871231105, "step": 25910}, {"loss": 0.5522, "grad_norm": 0.8104017972946167, "learning_rate": 0.0002, "epoch": 4.190445396491795, "step": 25920}, {"loss": 0.596, "grad_norm": 0.9984509944915771, "learning_rate": 0.0002, "epoch": 4.192062080672541, "step": 25930}, {"loss": 0.5356, "grad_norm": 0.9844784736633301, "learning_rate": 0.0002, "epoch": 4.193678764853286, "step": 25940}, {"loss": 0.5198, "grad_norm": 0.8168622255325317, "learning_rate": 0.0002, "epoch": 4.195295449034031, "step": 25950}, {"loss": 0.542, "grad_norm": 1.0878913402557373, "learning_rate": 0.0002, "epoch": 4.196912133214776, "step": 25960}, {"loss": 0.5414, "grad_norm": 0.927126407623291, "learning_rate": 0.0002, "epoch": 4.1985288173955215, "step": 25970}, {"loss": 0.5794, "grad_norm": 0.838586688041687, "learning_rate": 0.0002, "epoch": 4.200145501576267, "step": 25980}, {"loss": 0.5454, "grad_norm": 1.2572145462036133, "learning_rate": 0.0002, "epoch": 4.201762185757012, "step": 25990}, {"loss": 0.5048, "grad_norm": 1.0476740598678589, "learning_rate": 0.0002, "epoch": 4.203378869937758, "step": 26000}, {"loss": 0.5127, "grad_norm": 1.0873368978500366, "learning_rate": 0.0002, "epoch": 4.204995554118503, "step": 26010}, {"loss": 0.5679, "grad_norm": 1.2664896249771118, "learning_rate": 0.0002, "epoch": 4.206612238299249, "step": 26020}, {"loss": 0.5814, "grad_norm": 1.0312391519546509, "learning_rate": 0.0002, "epoch": 4.208228922479994, "step": 26030}, {"loss": 0.571, "grad_norm": 1.0235042572021484, "learning_rate": 0.0002, "epoch": 4.209845606660739, "step": 26040}, {"loss": 0.5766, "grad_norm": 0.8882219195365906, "learning_rate": 0.0002, "epoch": 4.211462290841484, "step": 26050}, {"loss": 0.5557, "grad_norm": 0.9115961790084839, "learning_rate": 0.0002, "epoch": 4.2130789750222295, "step": 26060}, {"loss": 0.5455, "grad_norm": 1.0218228101730347, "learning_rate": 0.0002, "epoch": 4.214695659202975, "step": 26070}, {"loss": 0.5462, "grad_norm": 1.0802232027053833, "learning_rate": 0.0002, "epoch": 4.21631234338372, "step": 26080}, {"loss": 0.557, "grad_norm": 1.1488053798675537, "learning_rate": 0.0002, "epoch": 4.217929027564465, "step": 26090}, {"loss": 0.52, "grad_norm": 1.0487725734710693, "learning_rate": 0.0002, "epoch": 4.21954571174521, "step": 26100}, {"loss": 0.5568, "grad_norm": 0.9131165742874146, "learning_rate": 0.0002, "epoch": 4.221162395925956, "step": 26110}, {"loss": 0.5206, "grad_norm": 0.9012845158576965, "learning_rate": 0.0002, "epoch": 4.222779080106701, "step": 26120}, {"loss": 0.561, "grad_norm": 0.8389840126037598, "learning_rate": 0.0002, "epoch": 4.224395764287446, "step": 26130}, {"loss": 0.5268, "grad_norm": 0.8924660682678223, "learning_rate": 0.0002, "epoch": 4.226012448468191, "step": 26140}, {"loss": 0.5715, "grad_norm": 0.8556463718414307, "learning_rate": 0.0002, "epoch": 4.2276291326489375, "step": 26150}, {"loss": 0.5695, "grad_norm": 0.9643129110336304, "learning_rate": 0.0002, "epoch": 4.229245816829683, "step": 26160}, {"loss": 0.5321, "grad_norm": 0.9865712523460388, "learning_rate": 0.0002, "epoch": 4.230862501010428, "step": 26170}, {"loss": 0.5406, "grad_norm": 1.152641773223877, "learning_rate": 0.0002, "epoch": 4.232479185191173, "step": 26180}, {"loss": 0.5632, "grad_norm": 0.9157698154449463, "learning_rate": 0.0002, "epoch": 4.234095869371918, "step": 26190}, {"loss": 0.5717, "grad_norm": 0.8418048620223999, "learning_rate": 0.0002, "epoch": 4.235712553552664, "step": 26200}, {"loss": 0.5624, "grad_norm": 0.9430168867111206, "learning_rate": 0.0002, "epoch": 4.237329237733409, "step": 26210}, {"loss": 0.5574, "grad_norm": 1.012582778930664, "learning_rate": 0.0002, "epoch": 4.238945921914154, "step": 26220}, {"loss": 0.5693, "grad_norm": 1.112619400024414, "learning_rate": 0.0002, "epoch": 4.240562606094899, "step": 26230}, {"loss": 0.6037, "grad_norm": 0.9243621826171875, "learning_rate": 0.0002, "epoch": 4.2421792902756446, "step": 26240}, {"loss": 0.569, "grad_norm": 0.6977595686912537, "learning_rate": 0.0002, "epoch": 4.24379597445639, "step": 26250}, {"loss": 0.5379, "grad_norm": 0.9600721597671509, "learning_rate": 0.0002, "epoch": 4.245412658637135, "step": 26260}, {"loss": 0.5658, "grad_norm": 0.882641613483429, "learning_rate": 0.0002, "epoch": 4.24702934281788, "step": 26270}, {"loss": 0.55, "grad_norm": 1.010920763015747, "learning_rate": 0.0002, "epoch": 4.2486460269986255, "step": 26280}, {"loss": 0.5803, "grad_norm": 0.9289400577545166, "learning_rate": 0.0002, "epoch": 4.250262711179371, "step": 26290}, {"loss": 0.541, "grad_norm": 1.137397289276123, "learning_rate": 0.0002, "epoch": 4.251879395360117, "step": 26300}, {"loss": 0.5204, "grad_norm": 1.0136182308197021, "learning_rate": 0.0002, "epoch": 4.253496079540862, "step": 26310}, {"loss": 0.5708, "grad_norm": 0.9387356042861938, "learning_rate": 0.0002, "epoch": 4.255112763721607, "step": 26320}, {"loss": 0.5948, "grad_norm": 1.1833957433700562, "learning_rate": 0.0002, "epoch": 4.2567294479023525, "step": 26330}, {"loss": 0.5905, "grad_norm": 0.9415934681892395, "learning_rate": 0.0002, "epoch": 4.258346132083098, "step": 26340}, {"loss": 0.5539, "grad_norm": 0.8550165891647339, "learning_rate": 0.0002, "epoch": 4.259962816263843, "step": 26350}, {"loss": 0.555, "grad_norm": 9.924622535705566, "learning_rate": 0.0002, "epoch": 4.261579500444588, "step": 26360}, {"loss": 0.5689, "grad_norm": 1.0104902982711792, "learning_rate": 0.0002, "epoch": 4.2631961846253335, "step": 26370}, {"loss": 0.5698, "grad_norm": 0.890794038772583, "learning_rate": 0.0002, "epoch": 4.264812868806079, "step": 26380}, {"loss": 0.563, "grad_norm": 1.0560191869735718, "learning_rate": 0.0002, "epoch": 4.266429552986824, "step": 26390}, {"loss": 0.5119, "grad_norm": 1.0135581493377686, "learning_rate": 0.0002, "epoch": 4.268046237167569, "step": 26400}, {"loss": 0.5359, "grad_norm": 1.1304140090942383, "learning_rate": 0.0002, "epoch": 4.269662921348314, "step": 26410}, {"loss": 0.5615, "grad_norm": 0.9899303913116455, "learning_rate": 0.0002, "epoch": 4.27127960552906, "step": 26420}, {"loss": 0.5815, "grad_norm": 1.0505329370498657, "learning_rate": 0.0002, "epoch": 4.272896289709805, "step": 26430}, {"loss": 0.5384, "grad_norm": 0.9389396905899048, "learning_rate": 0.0002, "epoch": 4.27451297389055, "step": 26440}, {"loss": 0.5558, "grad_norm": 0.875328779220581, "learning_rate": 0.0002, "epoch": 4.276129658071296, "step": 26450}, {"loss": 0.5601, "grad_norm": 1.0689256191253662, "learning_rate": 0.0002, "epoch": 4.277746342252041, "step": 26460}, {"loss": 0.546, "grad_norm": 0.9988957643508911, "learning_rate": 0.0002, "epoch": 4.279363026432787, "step": 26470}, {"loss": 0.5478, "grad_norm": 0.8721813559532166, "learning_rate": 0.0002, "epoch": 4.280979710613532, "step": 26480}, {"loss": 0.5424, "grad_norm": 1.100109577178955, "learning_rate": 0.0002, "epoch": 4.282596394794277, "step": 26490}, {"loss": 0.572, "grad_norm": 1.1607271432876587, "learning_rate": 0.0002, "epoch": 4.284213078975022, "step": 26500}, {"loss": 0.6287, "grad_norm": 0.879088819026947, "learning_rate": 0.0002, "epoch": 4.285829763155768, "step": 26510}, {"loss": 0.573, "grad_norm": 0.9891700744628906, "learning_rate": 0.0002, "epoch": 4.287446447336513, "step": 26520}, {"loss": 0.6018, "grad_norm": 1.0831127166748047, "learning_rate": 0.0002, "epoch": 4.289063131517258, "step": 26530}, {"loss": 0.5693, "grad_norm": 1.4108285903930664, "learning_rate": 0.0002, "epoch": 4.290679815698003, "step": 26540}, {"loss": 0.5888, "grad_norm": 1.0630289316177368, "learning_rate": 0.0002, "epoch": 4.2922964998787485, "step": 26550}, {"loss": 0.5817, "grad_norm": 1.0854572057724, "learning_rate": 0.0002, "epoch": 4.293913184059494, "step": 26560}, {"loss": 0.5586, "grad_norm": 0.9561646580696106, "learning_rate": 0.0002, "epoch": 4.295529868240239, "step": 26570}, {"loss": 0.5674, "grad_norm": 0.9064981937408447, "learning_rate": 0.0002, "epoch": 4.297146552420984, "step": 26580}, {"loss": 0.5847, "grad_norm": 1.0082972049713135, "learning_rate": 0.0002, "epoch": 4.298763236601729, "step": 26590}, {"loss": 0.5711, "grad_norm": 1.1613214015960693, "learning_rate": 0.0002, "epoch": 4.3003799207824756, "step": 26600}, {"loss": 0.551, "grad_norm": 0.9847695231437683, "learning_rate": 0.0002, "epoch": 4.301996604963221, "step": 26610}, {"loss": 0.6089, "grad_norm": 1.0980697870254517, "learning_rate": 0.0002, "epoch": 4.303613289143966, "step": 26620}, {"loss": 0.5797, "grad_norm": 0.8861175179481506, "learning_rate": 0.0002, "epoch": 4.305229973324711, "step": 26630}, {"loss": 0.5716, "grad_norm": 0.8917363286018372, "learning_rate": 0.0002, "epoch": 4.3068466575054565, "step": 26640}, {"loss": 0.5892, "grad_norm": 1.0458378791809082, "learning_rate": 0.0002, "epoch": 4.308463341686202, "step": 26650}, {"loss": 0.5883, "grad_norm": 1.4859240055084229, "learning_rate": 0.0002, "epoch": 4.310080025866947, "step": 26660}, {"loss": 0.5296, "grad_norm": 1.1376359462738037, "learning_rate": 0.0002, "epoch": 4.311696710047692, "step": 26670}, {"loss": 0.5671, "grad_norm": 0.991349995136261, "learning_rate": 0.0002, "epoch": 4.313313394228437, "step": 26680}, {"loss": 0.5338, "grad_norm": 0.9995543956756592, "learning_rate": 0.0002, "epoch": 4.314930078409183, "step": 26690}, {"loss": 0.5542, "grad_norm": 1.0515851974487305, "learning_rate": 0.0002, "epoch": 4.316546762589928, "step": 26700}, {"loss": 0.5473, "grad_norm": 1.008023977279663, "learning_rate": 0.0002, "epoch": 4.318163446770673, "step": 26710}, {"loss": 0.5506, "grad_norm": 1.0184582471847534, "learning_rate": 0.0002, "epoch": 4.319780130951418, "step": 26720}, {"loss": 0.5828, "grad_norm": 1.161071538925171, "learning_rate": 0.0002, "epoch": 4.321396815132164, "step": 26730}, {"loss": 0.5633, "grad_norm": 0.9580779671669006, "learning_rate": 0.0002, "epoch": 4.323013499312909, "step": 26740}, {"loss": 0.5785, "grad_norm": 1.0189911127090454, "learning_rate": 0.0002, "epoch": 4.324630183493655, "step": 26750}, {"loss": 0.5237, "grad_norm": 0.7484358549118042, "learning_rate": 0.0002, "epoch": 4.3262468676744, "step": 26760}, {"loss": 0.5728, "grad_norm": 1.0015908479690552, "learning_rate": 0.0002, "epoch": 4.327863551855145, "step": 26770}, {"loss": 0.5597, "grad_norm": 0.8972945809364319, "learning_rate": 0.0002, "epoch": 4.329480236035891, "step": 26780}, {"loss": 0.5857, "grad_norm": 1.01099693775177, "learning_rate": 0.0002, "epoch": 4.331096920216636, "step": 26790}, {"loss": 0.5591, "grad_norm": 0.846958339214325, "learning_rate": 0.0002, "epoch": 4.332713604397381, "step": 26800}, {"loss": 0.5547, "grad_norm": 1.0792603492736816, "learning_rate": 0.0002, "epoch": 4.334330288578126, "step": 26810}, {"loss": 0.5747, "grad_norm": 1.0373345613479614, "learning_rate": 0.0002, "epoch": 4.3359469727588715, "step": 26820}, {"loss": 0.558, "grad_norm": 0.9779167771339417, "learning_rate": 0.0002, "epoch": 4.337563656939617, "step": 26830}, {"loss": 0.5821, "grad_norm": 1.0235520601272583, "learning_rate": 0.0002, "epoch": 4.339180341120362, "step": 26840}, {"loss": 0.5843, "grad_norm": 1.04195237159729, "learning_rate": 0.0002, "epoch": 4.340797025301107, "step": 26850}, {"loss": 0.5474, "grad_norm": 0.9479565620422363, "learning_rate": 0.0002, "epoch": 4.3424137094818525, "step": 26860}, {"loss": 0.5646, "grad_norm": 0.9526172280311584, "learning_rate": 0.0002, "epoch": 4.344030393662598, "step": 26870}, {"loss": 0.521, "grad_norm": 0.8571456074714661, "learning_rate": 0.0002, "epoch": 4.345647077843343, "step": 26880}, {"loss": 0.5846, "grad_norm": 0.9475828409194946, "learning_rate": 0.0002, "epoch": 4.347263762024088, "step": 26890}, {"loss": 0.5815, "grad_norm": 1.0529576539993286, "learning_rate": 0.0002, "epoch": 4.348880446204834, "step": 26900}, {"loss": 0.56, "grad_norm": 0.9648140072822571, "learning_rate": 0.0002, "epoch": 4.3504971303855795, "step": 26910}, {"loss": 0.5162, "grad_norm": 1.0488841533660889, "learning_rate": 0.0002, "epoch": 4.352113814566325, "step": 26920}, {"loss": 0.5842, "grad_norm": 0.8771942257881165, "learning_rate": 0.0002, "epoch": 4.35373049874707, "step": 26930}, {"loss": 0.5966, "grad_norm": 0.9411202073097229, "learning_rate": 0.0002, "epoch": 4.355347182927815, "step": 26940}, {"loss": 0.6001, "grad_norm": 1.0997588634490967, "learning_rate": 0.0002, "epoch": 4.35696386710856, "step": 26950}, {"loss": 0.5528, "grad_norm": 0.968754768371582, "learning_rate": 0.0002, "epoch": 4.358580551289306, "step": 26960}, {"loss": 0.5881, "grad_norm": 0.9990773797035217, "learning_rate": 0.0002, "epoch": 4.360197235470051, "step": 26970}, {"loss": 0.5761, "grad_norm": 1.0210620164871216, "learning_rate": 0.0002, "epoch": 4.361813919650796, "step": 26980}, {"loss": 0.5768, "grad_norm": 0.855462908744812, "learning_rate": 0.0002, "epoch": 4.363430603831541, "step": 26990}, {"loss": 0.5493, "grad_norm": 0.9169660806655884, "learning_rate": 0.0002, "epoch": 4.365047288012287, "step": 27000}, {"loss": 0.5697, "grad_norm": 1.089629888534546, "learning_rate": 0.0002, "epoch": 4.366663972193032, "step": 27010}, {"loss": 0.5854, "grad_norm": 1.0932867527008057, "learning_rate": 0.0002, "epoch": 4.368280656373777, "step": 27020}, {"loss": 0.5656, "grad_norm": 0.9290956854820251, "learning_rate": 0.0002, "epoch": 4.369897340554522, "step": 27030}, {"loss": 0.5727, "grad_norm": 1.2800624370574951, "learning_rate": 0.0002, "epoch": 4.3715140247352675, "step": 27040}, {"loss": 0.5837, "grad_norm": 0.8993493318557739, "learning_rate": 0.0002, "epoch": 4.373130708916014, "step": 27050}, {"loss": 0.6232, "grad_norm": 1.1566431522369385, "learning_rate": 0.0002, "epoch": 4.374747393096759, "step": 27060}, {"loss": 0.5902, "grad_norm": 0.9479052424430847, "learning_rate": 0.0002, "epoch": 4.376364077277504, "step": 27070}, {"loss": 0.6189, "grad_norm": 1.0063648223876953, "learning_rate": 0.0002, "epoch": 4.377980761458249, "step": 27080}, {"loss": 0.561, "grad_norm": 0.8342045545578003, "learning_rate": 0.0002, "epoch": 4.379597445638995, "step": 27090}, {"loss": 0.5515, "grad_norm": 1.1390739679336548, "learning_rate": 0.0002, "epoch": 4.38121412981974, "step": 27100}, {"loss": 0.5372, "grad_norm": 0.9547637104988098, "learning_rate": 0.0002, "epoch": 4.382830814000485, "step": 27110}, {"loss": 0.5728, "grad_norm": 1.0503804683685303, "learning_rate": 0.0002, "epoch": 4.38444749818123, "step": 27120}, {"loss": 0.5787, "grad_norm": 0.9064017534255981, "learning_rate": 0.0002, "epoch": 4.3860641823619755, "step": 27130}, {"loss": 0.5798, "grad_norm": 0.9382519125938416, "learning_rate": 0.0002, "epoch": 4.387680866542721, "step": 27140}, {"loss": 0.5791, "grad_norm": 1.0410341024398804, "learning_rate": 0.0002, "epoch": 4.389297550723466, "step": 27150}, {"loss": 0.6034, "grad_norm": 0.9218655824661255, "learning_rate": 0.0002, "epoch": 4.390914234904211, "step": 27160}, {"loss": 0.5204, "grad_norm": 0.8119737505912781, "learning_rate": 0.0002, "epoch": 4.392530919084956, "step": 27170}, {"loss": 0.5612, "grad_norm": 0.8584722876548767, "learning_rate": 0.0002, "epoch": 4.394147603265702, "step": 27180}, {"loss": 0.5772, "grad_norm": 0.9668293595314026, "learning_rate": 0.0002, "epoch": 4.395764287446447, "step": 27190}, {"loss": 0.6009, "grad_norm": 1.022334098815918, "learning_rate": 0.0002, "epoch": 4.397380971627193, "step": 27200}, {"loss": 0.5573, "grad_norm": 0.9553216099739075, "learning_rate": 0.0002, "epoch": 4.398997655807938, "step": 27210}, {"loss": 0.5604, "grad_norm": 0.9282339215278625, "learning_rate": 0.0002, "epoch": 4.4006143399886835, "step": 27220}, {"loss": 0.5599, "grad_norm": 1.0232292413711548, "learning_rate": 0.0002, "epoch": 4.402231024169429, "step": 27230}, {"loss": 0.6078, "grad_norm": 0.9915700554847717, "learning_rate": 0.0002, "epoch": 4.403847708350174, "step": 27240}, {"loss": 0.5778, "grad_norm": 1.0014961957931519, "learning_rate": 0.0002, "epoch": 4.405464392530919, "step": 27250}, {"loss": 0.5824, "grad_norm": 1.1172103881835938, "learning_rate": 0.0002, "epoch": 4.407081076711664, "step": 27260}, {"loss": 0.5286, "grad_norm": 0.8583093285560608, "learning_rate": 0.0002, "epoch": 4.40869776089241, "step": 27270}, {"loss": 0.5507, "grad_norm": 0.7609201669692993, "learning_rate": 0.0002, "epoch": 4.410314445073155, "step": 27280}, {"loss": 0.575, "grad_norm": 1.0619351863861084, "learning_rate": 0.0002, "epoch": 4.4119311292539, "step": 27290}, {"loss": 0.5579, "grad_norm": 1.0177674293518066, "learning_rate": 0.0002, "epoch": 4.413547813434645, "step": 27300}, {"loss": 0.5628, "grad_norm": 0.9921218156814575, "learning_rate": 0.0002, "epoch": 4.4151644976153905, "step": 27310}, {"loss": 0.6018, "grad_norm": 1.126244306564331, "learning_rate": 0.0002, "epoch": 4.416781181796136, "step": 27320}, {"loss": 0.5743, "grad_norm": 1.0678540468215942, "learning_rate": 0.0002, "epoch": 4.418397865976881, "step": 27330}, {"loss": 0.5665, "grad_norm": 0.8705704212188721, "learning_rate": 0.0002, "epoch": 4.420014550157627, "step": 27340}, {"loss": 0.5763, "grad_norm": 1.272074818611145, "learning_rate": 0.0002, "epoch": 4.421631234338372, "step": 27350}, {"loss": 0.561, "grad_norm": 0.8740444183349609, "learning_rate": 0.0002, "epoch": 4.423247918519118, "step": 27360}, {"loss": 0.5492, "grad_norm": 1.0584250688552856, "learning_rate": 0.0002, "epoch": 4.424864602699863, "step": 27370}, {"loss": 0.589, "grad_norm": 1.059870719909668, "learning_rate": 0.0002, "epoch": 4.426481286880608, "step": 27380}, {"loss": 0.5551, "grad_norm": 1.072265863418579, "learning_rate": 0.0002, "epoch": 4.428097971061353, "step": 27390}, {"loss": 0.5584, "grad_norm": 0.871481716632843, "learning_rate": 0.0002, "epoch": 4.4297146552420985, "step": 27400}, {"loss": 0.5372, "grad_norm": 0.9555448293685913, "learning_rate": 0.0002, "epoch": 4.431331339422844, "step": 27410}, {"loss": 0.5593, "grad_norm": 1.0402292013168335, "learning_rate": 0.0002, "epoch": 4.432948023603589, "step": 27420}, {"loss": 0.5532, "grad_norm": 1.12587571144104, "learning_rate": 0.0002, "epoch": 4.434564707784334, "step": 27430}, {"loss": 0.5403, "grad_norm": 1.0783193111419678, "learning_rate": 0.0002, "epoch": 4.436181391965079, "step": 27440}, {"loss": 0.5313, "grad_norm": 1.024133563041687, "learning_rate": 0.0002, "epoch": 4.437798076145825, "step": 27450}, {"loss": 0.5621, "grad_norm": 0.9156768918037415, "learning_rate": 0.0002, "epoch": 4.43941476032657, "step": 27460}, {"loss": 0.5307, "grad_norm": 1.0215224027633667, "learning_rate": 0.0002, "epoch": 4.441031444507315, "step": 27470}, {"loss": 0.5188, "grad_norm": 1.082116961479187, "learning_rate": 0.0002, "epoch": 4.442648128688061, "step": 27480}, {"loss": 0.6203, "grad_norm": 1.0412873029708862, "learning_rate": 0.0002, "epoch": 4.4442648128688065, "step": 27490}, {"loss": 0.5939, "grad_norm": 1.0509289503097534, "learning_rate": 0.0002, "epoch": 4.445881497049552, "step": 27500}, {"loss": 0.5503, "grad_norm": 0.9291498064994812, "learning_rate": 0.0002, "epoch": 4.447498181230297, "step": 27510}, {"loss": 0.5408, "grad_norm": 0.970184326171875, "learning_rate": 0.0002, "epoch": 4.449114865411042, "step": 27520}, {"loss": 0.5705, "grad_norm": 0.8418883681297302, "learning_rate": 0.0002, "epoch": 4.450731549591787, "step": 27530}, {"loss": 0.5124, "grad_norm": 0.8823825120925903, "learning_rate": 0.0002, "epoch": 4.452348233772533, "step": 27540}, {"loss": 0.5867, "grad_norm": 1.1909019947052002, "learning_rate": 0.0002, "epoch": 4.453964917953278, "step": 27550}, {"loss": 0.5685, "grad_norm": 1.0317302942276, "learning_rate": 0.0002, "epoch": 4.455581602134023, "step": 27560}, {"loss": 0.5538, "grad_norm": 0.9977751970291138, "learning_rate": 0.0002, "epoch": 4.457198286314768, "step": 27570}, {"loss": 0.5628, "grad_norm": 0.8909519910812378, "learning_rate": 0.0002, "epoch": 4.458814970495514, "step": 27580}, {"loss": 0.6099, "grad_norm": 0.8653029799461365, "learning_rate": 0.0002, "epoch": 4.460431654676259, "step": 27590}, {"loss": 0.5622, "grad_norm": 1.0783653259277344, "learning_rate": 0.0002, "epoch": 4.462048338857004, "step": 27600}, {"loss": 0.579, "grad_norm": 1.1235394477844238, "learning_rate": 0.0002, "epoch": 4.463665023037749, "step": 27610}, {"loss": 0.5545, "grad_norm": 0.9386643767356873, "learning_rate": 0.0002, "epoch": 4.4652817072184945, "step": 27620}, {"loss": 0.5554, "grad_norm": 1.0605148077011108, "learning_rate": 0.0002, "epoch": 4.466898391399241, "step": 27630}, {"loss": 0.5886, "grad_norm": 1.1283893585205078, "learning_rate": 0.0002, "epoch": 4.468515075579986, "step": 27640}, {"loss": 0.5801, "grad_norm": 1.0583468675613403, "learning_rate": 0.0002, "epoch": 4.470131759760731, "step": 27650}, {"loss": 0.5601, "grad_norm": 0.9563992023468018, "learning_rate": 0.0002, "epoch": 4.471748443941476, "step": 27660}, {"loss": 0.5687, "grad_norm": 1.100598931312561, "learning_rate": 0.0002, "epoch": 4.4733651281222215, "step": 27670}, {"loss": 0.589, "grad_norm": 0.9386957287788391, "learning_rate": 0.0002, "epoch": 4.474981812302967, "step": 27680}, {"loss": 0.6241, "grad_norm": 1.2946288585662842, "learning_rate": 0.0002, "epoch": 4.476598496483712, "step": 27690}, {"loss": 0.6075, "grad_norm": 1.0325199365615845, "learning_rate": 0.0002, "epoch": 4.478215180664457, "step": 27700}, {"loss": 0.588, "grad_norm": 1.0318928956985474, "learning_rate": 0.0002, "epoch": 4.4798318648452025, "step": 27710}, {"loss": 0.5656, "grad_norm": 0.8721024394035339, "learning_rate": 0.0002, "epoch": 4.481448549025948, "step": 27720}, {"loss": 0.5421, "grad_norm": 1.17376708984375, "learning_rate": 0.0002, "epoch": 4.483065233206693, "step": 27730}, {"loss": 0.5657, "grad_norm": 1.0926326513290405, "learning_rate": 0.0002, "epoch": 4.484681917387438, "step": 27740}, {"loss": 0.5514, "grad_norm": 0.9043852686882019, "learning_rate": 0.0002, "epoch": 4.486298601568183, "step": 27750}, {"loss": 0.582, "grad_norm": 1.064600944519043, "learning_rate": 0.0002, "epoch": 4.487915285748929, "step": 27760}, {"loss": 0.6108, "grad_norm": 0.7833460569381714, "learning_rate": 0.0002, "epoch": 4.489531969929674, "step": 27770}, {"loss": 0.5985, "grad_norm": 1.1073496341705322, "learning_rate": 0.0002, "epoch": 4.49114865411042, "step": 27780}, {"loss": 0.5577, "grad_norm": 1.0799397230148315, "learning_rate": 0.0002, "epoch": 4.492765338291165, "step": 27790}, {"loss": 0.5601, "grad_norm": 1.1062238216400146, "learning_rate": 0.0002, "epoch": 4.49438202247191, "step": 27800}, {"loss": 0.6126, "grad_norm": 1.0568242073059082, "learning_rate": 0.0002, "epoch": 4.495998706652656, "step": 27810}, {"loss": 0.5913, "grad_norm": 0.8861091732978821, "learning_rate": 0.0002, "epoch": 4.497615390833401, "step": 27820}, {"loss": 0.5858, "grad_norm": 1.2297543287277222, "learning_rate": 0.0002, "epoch": 4.499232075014146, "step": 27830}, {"loss": 0.5859, "grad_norm": 0.9600302577018738, "learning_rate": 0.0002, "epoch": 4.500848759194891, "step": 27840}, {"loss": 0.6124, "grad_norm": 1.057051181793213, "learning_rate": 0.0002, "epoch": 4.502465443375637, "step": 27850}, {"loss": 0.5788, "grad_norm": 0.9839690923690796, "learning_rate": 0.0002, "epoch": 4.504082127556382, "step": 27860}, {"loss": 0.555, "grad_norm": 1.1479853391647339, "learning_rate": 0.0002, "epoch": 4.505698811737127, "step": 27870}, {"loss": 0.6039, "grad_norm": 1.0550768375396729, "learning_rate": 0.0002, "epoch": 4.507315495917872, "step": 27880}, {"loss": 0.563, "grad_norm": 0.898209273815155, "learning_rate": 0.0002, "epoch": 4.5089321800986175, "step": 27890}, {"loss": 0.5734, "grad_norm": 0.9460315108299255, "learning_rate": 0.0002, "epoch": 4.510548864279363, "step": 27900}, {"loss": 0.5702, "grad_norm": 0.9499884247779846, "learning_rate": 0.0002, "epoch": 4.512165548460108, "step": 27910}, {"loss": 0.5385, "grad_norm": 0.7801318764686584, "learning_rate": 0.0002, "epoch": 4.513782232640853, "step": 27920}, {"loss": 0.5391, "grad_norm": 0.9286966323852539, "learning_rate": 0.0002, "epoch": 4.515398916821599, "step": 27930}, {"loss": 0.5717, "grad_norm": 0.9539980292320251, "learning_rate": 0.0002, "epoch": 4.517015601002345, "step": 27940}, {"loss": 0.6073, "grad_norm": 1.1053401231765747, "learning_rate": 0.0002, "epoch": 4.51863228518309, "step": 27950}, {"loss": 0.6087, "grad_norm": 0.7535534501075745, "learning_rate": 0.0002, "epoch": 4.520248969363835, "step": 27960}, {"loss": 0.5701, "grad_norm": 1.076926589012146, "learning_rate": 0.0002, "epoch": 4.52186565354458, "step": 27970}, {"loss": 0.6028, "grad_norm": 1.181935429573059, "learning_rate": 0.0002, "epoch": 4.5234823377253255, "step": 27980}, {"loss": 0.6033, "grad_norm": 0.9293407201766968, "learning_rate": 0.0002, "epoch": 4.525099021906071, "step": 27990}, {"loss": 0.5815, "grad_norm": 0.8953009247779846, "learning_rate": 0.0002, "epoch": 4.526715706086816, "step": 28000}, {"loss": 0.5564, "grad_norm": 1.0850225687026978, "learning_rate": 0.0002, "epoch": 4.528332390267561, "step": 28010}, {"loss": 0.5459, "grad_norm": 0.9125663042068481, "learning_rate": 0.0002, "epoch": 4.529949074448306, "step": 28020}, {"loss": 0.5922, "grad_norm": 0.8745216727256775, "learning_rate": 0.0002, "epoch": 4.531565758629052, "step": 28030}, {"loss": 0.567, "grad_norm": 1.0783463716506958, "learning_rate": 0.0002, "epoch": 4.533182442809797, "step": 28040}, {"loss": 0.5754, "grad_norm": 0.7513844966888428, "learning_rate": 0.0002, "epoch": 4.534799126990542, "step": 28050}, {"loss": 0.5608, "grad_norm": 1.0135776996612549, "learning_rate": 0.0002, "epoch": 4.536415811171287, "step": 28060}, {"loss": 0.5827, "grad_norm": 0.8886825442314148, "learning_rate": 0.0002, "epoch": 4.538032495352033, "step": 28070}, {"loss": 0.5605, "grad_norm": 0.8153995275497437, "learning_rate": 0.0002, "epoch": 4.539649179532779, "step": 28080}, {"loss": 0.6377, "grad_norm": 0.9853341579437256, "learning_rate": 0.0002, "epoch": 4.541265863713524, "step": 28090}, {"loss": 0.5957, "grad_norm": 0.9365800023078918, "learning_rate": 0.0002, "epoch": 4.542882547894269, "step": 28100}, {"loss": 0.5477, "grad_norm": 0.9765017628669739, "learning_rate": 0.0002, "epoch": 4.544499232075014, "step": 28110}, {"loss": 0.6185, "grad_norm": 0.9811279773712158, "learning_rate": 0.0002, "epoch": 4.54611591625576, "step": 28120}, {"loss": 0.6095, "grad_norm": 1.0387924909591675, "learning_rate": 0.0002, "epoch": 4.547732600436505, "step": 28130}, {"loss": 0.6534, "grad_norm": 1.0684878826141357, "learning_rate": 0.0002, "epoch": 4.54934928461725, "step": 28140}, {"loss": 0.5701, "grad_norm": 1.0000102519989014, "learning_rate": 0.0002, "epoch": 4.550965968797995, "step": 28150}, {"loss": 0.5327, "grad_norm": 1.0717930793762207, "learning_rate": 0.0002, "epoch": 4.5525826529787405, "step": 28160}, {"loss": 0.5594, "grad_norm": 0.990074634552002, "learning_rate": 0.0002, "epoch": 4.554199337159486, "step": 28170}, {"loss": 0.5452, "grad_norm": 0.8673754930496216, "learning_rate": 0.0002, "epoch": 4.555816021340231, "step": 28180}, {"loss": 0.5773, "grad_norm": 0.864247739315033, "learning_rate": 0.0002, "epoch": 4.557432705520976, "step": 28190}, {"loss": 0.5516, "grad_norm": 0.8280200958251953, "learning_rate": 0.0002, "epoch": 4.5590493897017215, "step": 28200}, {"loss": 0.5709, "grad_norm": 1.1312172412872314, "learning_rate": 0.0002, "epoch": 4.560666073882467, "step": 28210}, {"loss": 0.5776, "grad_norm": 0.9147403240203857, "learning_rate": 0.0002, "epoch": 4.562282758063212, "step": 28220}, {"loss": 0.5591, "grad_norm": 1.0321218967437744, "learning_rate": 0.0002, "epoch": 4.563899442243958, "step": 28230}, {"loss": 0.5508, "grad_norm": 1.168332815170288, "learning_rate": 0.0002, "epoch": 4.565516126424703, "step": 28240}, {"loss": 0.5649, "grad_norm": 1.0067222118377686, "learning_rate": 0.0002, "epoch": 4.5671328106054485, "step": 28250}, {"loss": 0.5853, "grad_norm": 1.0283393859863281, "learning_rate": 0.0002, "epoch": 4.568749494786194, "step": 28260}, {"loss": 0.5772, "grad_norm": 0.9912363886833191, "learning_rate": 0.0002, "epoch": 4.570366178966939, "step": 28270}, {"loss": 0.5757, "grad_norm": 1.108032464981079, "learning_rate": 0.0002, "epoch": 4.571982863147684, "step": 28280}, {"loss": 0.5529, "grad_norm": 0.8260078430175781, "learning_rate": 0.0002, "epoch": 4.573599547328429, "step": 28290}, {"loss": 0.5625, "grad_norm": 0.8946247100830078, "learning_rate": 0.0002, "epoch": 4.575216231509175, "step": 28300}, {"loss": 0.5533, "grad_norm": 0.8273587822914124, "learning_rate": 0.0002, "epoch": 4.57683291568992, "step": 28310}, {"loss": 0.6058, "grad_norm": 0.9040093421936035, "learning_rate": 0.0002, "epoch": 4.578449599870665, "step": 28320}, {"loss": 0.5521, "grad_norm": 0.8435290455818176, "learning_rate": 0.0002, "epoch": 4.58006628405141, "step": 28330}, {"loss": 0.6086, "grad_norm": 1.164088249206543, "learning_rate": 0.0002, "epoch": 4.581682968232156, "step": 28340}, {"loss": 0.5603, "grad_norm": 0.9861085414886475, "learning_rate": 0.0002, "epoch": 4.583299652412901, "step": 28350}, {"loss": 0.5701, "grad_norm": 0.8892980813980103, "learning_rate": 0.0002, "epoch": 4.584916336593646, "step": 28360}, {"loss": 0.598, "grad_norm": 1.240574836730957, "learning_rate": 0.0002, "epoch": 4.586533020774391, "step": 28370}, {"loss": 0.5797, "grad_norm": 0.8669408559799194, "learning_rate": 0.0002, "epoch": 4.588149704955137, "step": 28380}, {"loss": 0.5603, "grad_norm": 0.9145985841751099, "learning_rate": 0.0002, "epoch": 4.589766389135883, "step": 28390}, {"loss": 0.5765, "grad_norm": 0.8584614992141724, "learning_rate": 0.0002, "epoch": 4.591383073316628, "step": 28400}, {"loss": 0.5898, "grad_norm": 1.118829369544983, "learning_rate": 0.0002, "epoch": 4.592999757497373, "step": 28410}, {"loss": 0.5641, "grad_norm": 1.1411553621292114, "learning_rate": 0.0002, "epoch": 4.594616441678118, "step": 28420}, {"loss": 0.549, "grad_norm": 0.9433278441429138, "learning_rate": 0.0002, "epoch": 4.596233125858864, "step": 28430}, {"loss": 0.5496, "grad_norm": 0.816830039024353, "learning_rate": 0.0002, "epoch": 4.597849810039609, "step": 28440}, {"loss": 0.5543, "grad_norm": 1.2124968767166138, "learning_rate": 0.0002, "epoch": 4.599466494220354, "step": 28450}, {"loss": 0.5759, "grad_norm": 0.9658762216567993, "learning_rate": 0.0002, "epoch": 4.601083178401099, "step": 28460}, {"loss": 0.5902, "grad_norm": 0.836100161075592, "learning_rate": 0.0002, "epoch": 4.6026998625818445, "step": 28470}, {"loss": 0.5749, "grad_norm": 0.9989104270935059, "learning_rate": 0.0002, "epoch": 4.60431654676259, "step": 28480}, {"loss": 0.5616, "grad_norm": 1.1298956871032715, "learning_rate": 0.0002, "epoch": 4.605933230943335, "step": 28490}, {"loss": 0.5846, "grad_norm": 1.1731704473495483, "learning_rate": 0.0002, "epoch": 4.60754991512408, "step": 28500}, {"loss": 0.5816, "grad_norm": 0.9624714255332947, "learning_rate": 0.0002, "epoch": 4.609166599304825, "step": 28510}, {"loss": 0.5868, "grad_norm": 1.364073634147644, "learning_rate": 0.0002, "epoch": 4.610783283485571, "step": 28520}, {"loss": 0.6237, "grad_norm": 1.1827356815338135, "learning_rate": 0.0002, "epoch": 4.612399967666317, "step": 28530}, {"loss": 0.5643, "grad_norm": 0.6651531457901001, "learning_rate": 0.0002, "epoch": 4.614016651847062, "step": 28540}, {"loss": 0.6051, "grad_norm": 1.1640995740890503, "learning_rate": 0.0002, "epoch": 4.615633336027807, "step": 28550}, {"loss": 0.5995, "grad_norm": 1.028918743133545, "learning_rate": 0.0002, "epoch": 4.6172500202085525, "step": 28560}, {"loss": 0.5607, "grad_norm": 0.8252120614051819, "learning_rate": 0.0002, "epoch": 4.618866704389298, "step": 28570}, {"loss": 0.5769, "grad_norm": 1.3536735773086548, "learning_rate": 0.0002, "epoch": 4.620483388570043, "step": 28580}, {"loss": 0.6006, "grad_norm": 1.2146915197372437, "learning_rate": 0.0002, "epoch": 4.622100072750788, "step": 28590}, {"loss": 0.5503, "grad_norm": 1.0122549533843994, "learning_rate": 0.0002, "epoch": 4.623716756931533, "step": 28600}, {"loss": 0.6072, "grad_norm": 0.9977872967720032, "learning_rate": 0.0002, "epoch": 4.625333441112279, "step": 28610}, {"loss": 0.5669, "grad_norm": 1.0159751176834106, "learning_rate": 0.0002, "epoch": 4.626950125293024, "step": 28620}, {"loss": 0.5935, "grad_norm": 1.0028325319290161, "learning_rate": 0.0002, "epoch": 4.628566809473769, "step": 28630}, {"loss": 0.5515, "grad_norm": 0.901638388633728, "learning_rate": 0.0002, "epoch": 4.630183493654514, "step": 28640}, {"loss": 0.595, "grad_norm": 0.9450507164001465, "learning_rate": 0.0002, "epoch": 4.6318001778352595, "step": 28650}, {"loss": 0.5972, "grad_norm": 0.9987545013427734, "learning_rate": 0.0002, "epoch": 4.633416862016006, "step": 28660}, {"loss": 0.5863, "grad_norm": 0.9574332237243652, "learning_rate": 0.0002, "epoch": 4.63503354619675, "step": 28670}, {"loss": 0.5804, "grad_norm": 1.2215653657913208, "learning_rate": 0.0002, "epoch": 4.636650230377496, "step": 28680}, {"loss": 0.5798, "grad_norm": 0.9798858761787415, "learning_rate": 0.0002, "epoch": 4.638266914558241, "step": 28690}, {"loss": 0.5773, "grad_norm": 1.0648466348648071, "learning_rate": 0.0002, "epoch": 4.639883598738987, "step": 28700}, {"loss": 0.6108, "grad_norm": 1.0606504678726196, "learning_rate": 0.0002, "epoch": 4.641500282919732, "step": 28710}, {"loss": 0.5801, "grad_norm": 1.0892442464828491, "learning_rate": 0.0002, "epoch": 4.643116967100477, "step": 28720}, {"loss": 0.5492, "grad_norm": 0.914391040802002, "learning_rate": 0.0002, "epoch": 4.644733651281222, "step": 28730}, {"loss": 0.5439, "grad_norm": 0.9782370328903198, "learning_rate": 0.0002, "epoch": 4.6463503354619675, "step": 28740}, {"loss": 0.6035, "grad_norm": 1.0344339609146118, "learning_rate": 0.0002, "epoch": 4.647967019642713, "step": 28750}, {"loss": 0.5775, "grad_norm": 1.0513931512832642, "learning_rate": 0.0002, "epoch": 4.649583703823458, "step": 28760}, {"loss": 0.546, "grad_norm": 0.9711475968360901, "learning_rate": 0.0002, "epoch": 4.651200388004203, "step": 28770}, {"loss": 0.5472, "grad_norm": 0.977519690990448, "learning_rate": 0.0002, "epoch": 4.652817072184948, "step": 28780}, {"loss": 0.5826, "grad_norm": 0.9150224924087524, "learning_rate": 0.0002, "epoch": 4.654433756365694, "step": 28790}, {"loss": 0.5382, "grad_norm": 1.0973542928695679, "learning_rate": 0.0002, "epoch": 4.656050440546439, "step": 28800}, {"loss": 0.6147, "grad_norm": 0.944877564907074, "learning_rate": 0.0002, "epoch": 4.657667124727185, "step": 28810}, {"loss": 0.5537, "grad_norm": 0.9508748650550842, "learning_rate": 0.0002, "epoch": 4.659283808907929, "step": 28820}, {"loss": 0.5537, "grad_norm": 0.9681721329689026, "learning_rate": 0.0002, "epoch": 4.6609004930886755, "step": 28830}, {"loss": 0.592, "grad_norm": 1.0214351415634155, "learning_rate": 0.0002, "epoch": 4.662517177269421, "step": 28840}, {"loss": 0.6031, "grad_norm": 0.9748611450195312, "learning_rate": 0.0002, "epoch": 4.664133861450166, "step": 28850}, {"loss": 0.572, "grad_norm": 0.8484147191047668, "learning_rate": 0.0002, "epoch": 4.665750545630911, "step": 28860}, {"loss": 0.5699, "grad_norm": 1.1252986192703247, "learning_rate": 0.0002, "epoch": 4.667367229811656, "step": 28870}, {"loss": 0.5724, "grad_norm": 0.8706206679344177, "learning_rate": 0.0002, "epoch": 4.668983913992402, "step": 28880}, {"loss": 0.6002, "grad_norm": 1.1432424783706665, "learning_rate": 0.0002, "epoch": 4.670600598173147, "step": 28890}, {"loss": 0.5675, "grad_norm": 1.017029047012329, "learning_rate": 0.0002, "epoch": 4.672217282353892, "step": 28900}, {"loss": 0.5831, "grad_norm": 1.085597038269043, "learning_rate": 0.0002, "epoch": 4.673833966534637, "step": 28910}, {"loss": 0.5678, "grad_norm": 0.9275796413421631, "learning_rate": 0.0002, "epoch": 4.675450650715383, "step": 28920}, {"loss": 0.5603, "grad_norm": 0.9518964886665344, "learning_rate": 0.0002, "epoch": 4.677067334896128, "step": 28930}, {"loss": 0.6232, "grad_norm": 1.0352122783660889, "learning_rate": 0.0002, "epoch": 4.678684019076873, "step": 28940}, {"loss": 0.5786, "grad_norm": 1.090124249458313, "learning_rate": 0.0002, "epoch": 4.680300703257618, "step": 28950}, {"loss": 0.5728, "grad_norm": 0.8799563050270081, "learning_rate": 0.0002, "epoch": 4.681917387438364, "step": 28960}, {"loss": 0.5787, "grad_norm": 1.0929821729660034, "learning_rate": 0.0002, "epoch": 4.683534071619109, "step": 28970}, {"loss": 0.6134, "grad_norm": 0.903727650642395, "learning_rate": 0.0002, "epoch": 4.685150755799855, "step": 28980}, {"loss": 0.5522, "grad_norm": 0.9752424955368042, "learning_rate": 0.0002, "epoch": 4.6867674399806, "step": 28990}, {"loss": 0.5762, "grad_norm": 0.9351571202278137, "learning_rate": 0.0002, "epoch": 4.688384124161345, "step": 29000}, {"loss": 0.5811, "grad_norm": 0.923877477645874, "learning_rate": 0.0002, "epoch": 4.6900008083420905, "step": 29010}, {"loss": 0.5682, "grad_norm": 1.045389175415039, "learning_rate": 0.0002, "epoch": 4.691617492522836, "step": 29020}, {"loss": 0.584, "grad_norm": 1.0200831890106201, "learning_rate": 0.0002, "epoch": 4.693234176703581, "step": 29030}, {"loss": 0.5514, "grad_norm": 1.1499706506729126, "learning_rate": 0.0002, "epoch": 4.694850860884326, "step": 29040}, {"loss": 0.5745, "grad_norm": 0.860118567943573, "learning_rate": 0.0002, "epoch": 4.6964675450650715, "step": 29050}, {"loss": 0.5741, "grad_norm": 0.9774864315986633, "learning_rate": 0.0002, "epoch": 4.698084229245817, "step": 29060}, {"loss": 0.5765, "grad_norm": 1.0323210954666138, "learning_rate": 0.0002, "epoch": 4.699700913426562, "step": 29070}, {"loss": 0.5452, "grad_norm": 0.8492481112480164, "learning_rate": 0.0002, "epoch": 4.701317597607307, "step": 29080}, {"loss": 0.5985, "grad_norm": 1.131951093673706, "learning_rate": 0.0002, "epoch": 4.702934281788052, "step": 29090}, {"loss": 0.6412, "grad_norm": 0.8763113021850586, "learning_rate": 0.0002, "epoch": 4.704550965968798, "step": 29100}, {"loss": 0.575, "grad_norm": 1.045028805732727, "learning_rate": 0.0002, "epoch": 4.706167650149544, "step": 29110}, {"loss": 0.5548, "grad_norm": 0.9961401224136353, "learning_rate": 0.0002, "epoch": 4.707784334330288, "step": 29120}, {"loss": 0.559, "grad_norm": 0.9282503724098206, "learning_rate": 0.0002, "epoch": 4.709401018511034, "step": 29130}, {"loss": 0.5744, "grad_norm": 1.1418932676315308, "learning_rate": 0.0002, "epoch": 4.711017702691779, "step": 29140}, {"loss": 0.5394, "grad_norm": 0.9950099587440491, "learning_rate": 0.0002, "epoch": 4.712634386872525, "step": 29150}, {"loss": 0.6177, "grad_norm": 0.8304893374443054, "learning_rate": 0.0002, "epoch": 4.71425107105327, "step": 29160}, {"loss": 0.6074, "grad_norm": 1.115626335144043, "learning_rate": 0.0002, "epoch": 4.715867755234015, "step": 29170}, {"loss": 0.6265, "grad_norm": 1.079818606376648, "learning_rate": 0.0002, "epoch": 4.71748443941476, "step": 29180}, {"loss": 0.561, "grad_norm": 1.1929082870483398, "learning_rate": 0.0002, "epoch": 4.719101123595506, "step": 29190}, {"loss": 0.5708, "grad_norm": 0.9621080756187439, "learning_rate": 0.0002, "epoch": 4.720717807776251, "step": 29200}, {"loss": 0.546, "grad_norm": 0.8549222350120544, "learning_rate": 0.0002, "epoch": 4.722334491956996, "step": 29210}, {"loss": 0.5775, "grad_norm": 0.9341941475868225, "learning_rate": 0.0002, "epoch": 4.723951176137741, "step": 29220}, {"loss": 0.5436, "grad_norm": 1.075406789779663, "learning_rate": 0.0002, "epoch": 4.7255678603184865, "step": 29230}, {"loss": 0.576, "grad_norm": 1.0859880447387695, "learning_rate": 0.0002, "epoch": 4.727184544499232, "step": 29240}, {"loss": 0.5525, "grad_norm": 0.8475605249404907, "learning_rate": 0.0002, "epoch": 4.728801228679977, "step": 29250}, {"loss": 0.5659, "grad_norm": 0.9331845641136169, "learning_rate": 0.0002, "epoch": 4.730417912860723, "step": 29260}, {"loss": 0.5901, "grad_norm": 0.9279314279556274, "learning_rate": 0.0002, "epoch": 4.7320345970414674, "step": 29270}, {"loss": 0.597, "grad_norm": 0.7803558707237244, "learning_rate": 0.0002, "epoch": 4.733651281222214, "step": 29280}, {"loss": 0.5968, "grad_norm": 1.0159329175949097, "learning_rate": 0.0002, "epoch": 4.735267965402959, "step": 29290}, {"loss": 0.5333, "grad_norm": 0.9448670744895935, "learning_rate": 0.0002, "epoch": 4.736884649583704, "step": 29300}, {"loss": 0.574, "grad_norm": 1.0732197761535645, "learning_rate": 0.0002, "epoch": 4.738501333764449, "step": 29310}, {"loss": 0.6066, "grad_norm": 0.901830792427063, "learning_rate": 0.0002, "epoch": 4.7401180179451945, "step": 29320}, {"loss": 0.6105, "grad_norm": 0.9141789674758911, "learning_rate": 0.0002, "epoch": 4.74173470212594, "step": 29330}, {"loss": 0.5481, "grad_norm": 0.9733418226242065, "learning_rate": 0.0002, "epoch": 4.743351386306685, "step": 29340}, {"loss": 0.612, "grad_norm": 0.909810483455658, "learning_rate": 0.0002, "epoch": 4.74496807048743, "step": 29350}, {"loss": 0.5911, "grad_norm": 0.909541666507721, "learning_rate": 0.0002, "epoch": 4.746584754668175, "step": 29360}, {"loss": 0.5579, "grad_norm": 0.9383015632629395, "learning_rate": 0.0002, "epoch": 4.748201438848921, "step": 29370}, {"loss": 0.5529, "grad_norm": 0.9275668263435364, "learning_rate": 0.0002, "epoch": 4.749818123029666, "step": 29380}, {"loss": 0.5623, "grad_norm": 1.1146225929260254, "learning_rate": 0.0002, "epoch": 4.751434807210411, "step": 29390}, {"loss": 0.6018, "grad_norm": 1.0062453746795654, "learning_rate": 0.0002, "epoch": 4.753051491391156, "step": 29400}, {"loss": 0.5872, "grad_norm": 0.9451895952224731, "learning_rate": 0.0002, "epoch": 4.7546681755719025, "step": 29410}, {"loss": 0.5767, "grad_norm": 0.870457649230957, "learning_rate": 0.0002, "epoch": 4.756284859752648, "step": 29420}, {"loss": 0.57, "grad_norm": 1.0411282777786255, "learning_rate": 0.0002, "epoch": 4.757901543933393, "step": 29430}, {"loss": 0.5688, "grad_norm": 1.1648986339569092, "learning_rate": 0.0002, "epoch": 4.759518228114138, "step": 29440}, {"loss": 0.5432, "grad_norm": 0.8999572992324829, "learning_rate": 0.0002, "epoch": 4.761134912294883, "step": 29450}, {"loss": 0.5667, "grad_norm": 0.9863559007644653, "learning_rate": 0.0002, "epoch": 4.762751596475629, "step": 29460}, {"loss": 0.5779, "grad_norm": 0.9676542282104492, "learning_rate": 0.0002, "epoch": 4.764368280656374, "step": 29470}, {"loss": 0.6075, "grad_norm": 1.004775047302246, "learning_rate": 0.0002, "epoch": 4.765984964837119, "step": 29480}, {"loss": 0.6044, "grad_norm": 1.0937515497207642, "learning_rate": 0.0002, "epoch": 4.767601649017864, "step": 29490}, {"loss": 0.5433, "grad_norm": 0.9551598429679871, "learning_rate": 0.0002, "epoch": 4.7692183331986095, "step": 29500}, {"loss": 0.5609, "grad_norm": 1.0757228136062622, "learning_rate": 0.0002, "epoch": 4.770835017379355, "step": 29510}, {"loss": 0.567, "grad_norm": 1.0588841438293457, "learning_rate": 0.0002, "epoch": 4.7724517015601, "step": 29520}, {"loss": 0.5814, "grad_norm": 1.0744032859802246, "learning_rate": 0.0002, "epoch": 4.774068385740845, "step": 29530}, {"loss": 0.5681, "grad_norm": 1.0066277980804443, "learning_rate": 0.0002, "epoch": 4.7756850699215905, "step": 29540}, {"loss": 0.545, "grad_norm": 1.082319736480713, "learning_rate": 0.0002, "epoch": 4.777301754102336, "step": 29550}, {"loss": 0.5709, "grad_norm": 0.8252472877502441, "learning_rate": 0.0002, "epoch": 4.778918438283082, "step": 29560}, {"loss": 0.5666, "grad_norm": 0.9855340123176575, "learning_rate": 0.0002, "epoch": 4.780535122463827, "step": 29570}, {"loss": 0.6117, "grad_norm": 0.9991421699523926, "learning_rate": 0.0002, "epoch": 4.782151806644572, "step": 29580}, {"loss": 0.5966, "grad_norm": 1.316841959953308, "learning_rate": 0.0002, "epoch": 4.7837684908253175, "step": 29590}, {"loss": 0.6102, "grad_norm": 1.1513035297393799, "learning_rate": 0.0002, "epoch": 4.785385175006063, "step": 29600}, {"loss": 0.5785, "grad_norm": 0.9767683744430542, "learning_rate": 0.0002, "epoch": 4.787001859186808, "step": 29610}, {"loss": 0.6037, "grad_norm": 0.9786278605461121, "learning_rate": 0.0002, "epoch": 4.788618543367553, "step": 29620}, {"loss": 0.6108, "grad_norm": 0.8004973530769348, "learning_rate": 0.0002, "epoch": 4.7902352275482984, "step": 29630}, {"loss": 0.5932, "grad_norm": 1.0997767448425293, "learning_rate": 0.0002, "epoch": 4.791851911729044, "step": 29640}, {"loss": 0.5655, "grad_norm": 0.9752856492996216, "learning_rate": 0.0002, "epoch": 4.793468595909789, "step": 29650}, {"loss": 0.5916, "grad_norm": 1.0518392324447632, "learning_rate": 0.0002, "epoch": 4.795085280090534, "step": 29660}, {"loss": 0.6042, "grad_norm": 1.1050055027008057, "learning_rate": 0.0002, "epoch": 4.796701964271279, "step": 29670}, {"loss": 0.6089, "grad_norm": 0.9933857917785645, "learning_rate": 0.0002, "epoch": 4.798318648452025, "step": 29680}, {"loss": 0.6041, "grad_norm": 1.2804018259048462, "learning_rate": 0.0002, "epoch": 4.79993533263277, "step": 29690}, {"loss": 0.636, "grad_norm": 1.0133371353149414, "learning_rate": 0.0002, "epoch": 4.801552016813515, "step": 29700}, {"loss": 0.5662, "grad_norm": 1.080350637435913, "learning_rate": 0.0002, "epoch": 4.803168700994261, "step": 29710}, {"loss": 0.5603, "grad_norm": 0.9986529350280762, "learning_rate": 0.0002, "epoch": 4.804785385175006, "step": 29720}, {"loss": 0.5894, "grad_norm": 0.975665807723999, "learning_rate": 0.0002, "epoch": 4.806402069355752, "step": 29730}, {"loss": 0.6328, "grad_norm": 0.8458138704299927, "learning_rate": 0.0002, "epoch": 4.808018753536497, "step": 29740}, {"loss": 0.5837, "grad_norm": 0.99330073595047, "learning_rate": 0.0002, "epoch": 4.809635437717242, "step": 29750}, {"loss": 0.5507, "grad_norm": 0.898274302482605, "learning_rate": 0.0002, "epoch": 4.811252121897987, "step": 29760}, {"loss": 0.5842, "grad_norm": 1.0504480600357056, "learning_rate": 0.0002, "epoch": 4.812868806078733, "step": 29770}, {"loss": 0.5821, "grad_norm": 0.937919020652771, "learning_rate": 0.0002, "epoch": 4.814485490259478, "step": 29780}, {"loss": 0.5885, "grad_norm": 0.9593307971954346, "learning_rate": 0.0002, "epoch": 4.816102174440223, "step": 29790}, {"loss": 0.578, "grad_norm": 0.9431198835372925, "learning_rate": 0.0002, "epoch": 4.817718858620968, "step": 29800}, {"loss": 0.5739, "grad_norm": 1.2729957103729248, "learning_rate": 0.0002, "epoch": 4.8193355428017135, "step": 29810}, {"loss": 0.6124, "grad_norm": 0.8876838684082031, "learning_rate": 0.0002, "epoch": 4.820952226982459, "step": 29820}, {"loss": 0.5583, "grad_norm": 1.0185000896453857, "learning_rate": 0.0002, "epoch": 4.822568911163204, "step": 29830}, {"loss": 0.5686, "grad_norm": 1.064276099205017, "learning_rate": 0.0002, "epoch": 4.824185595343949, "step": 29840}, {"loss": 0.5698, "grad_norm": 0.9774803519248962, "learning_rate": 0.0002, "epoch": 4.825802279524694, "step": 29850}, {"loss": 0.5533, "grad_norm": 1.131646990776062, "learning_rate": 0.0002, "epoch": 4.8274189637054405, "step": 29860}, {"loss": 0.6371, "grad_norm": 1.081455945968628, "learning_rate": 0.0002, "epoch": 4.829035647886186, "step": 29870}, {"loss": 0.5793, "grad_norm": 0.990538477897644, "learning_rate": 0.0002, "epoch": 4.830652332066931, "step": 29880}, {"loss": 0.5833, "grad_norm": 0.9750600457191467, "learning_rate": 0.0002, "epoch": 4.832269016247676, "step": 29890}, {"loss": 0.619, "grad_norm": 1.0600621700286865, "learning_rate": 0.0002, "epoch": 4.8338857004284215, "step": 29900}, {"loss": 0.5841, "grad_norm": 0.9237320423126221, "learning_rate": 0.0002, "epoch": 4.835502384609167, "step": 29910}, {"loss": 0.5513, "grad_norm": 0.9739177227020264, "learning_rate": 0.0002, "epoch": 4.837119068789912, "step": 29920}, {"loss": 0.587, "grad_norm": 1.128677248954773, "learning_rate": 0.0002, "epoch": 4.838735752970657, "step": 29930}, {"loss": 0.564, "grad_norm": 1.042604923248291, "learning_rate": 0.0002, "epoch": 4.840352437151402, "step": 29940}, {"loss": 0.5885, "grad_norm": 0.849758505821228, "learning_rate": 0.0002, "epoch": 4.841969121332148, "step": 29950}, {"loss": 0.5952, "grad_norm": 1.2809888124465942, "learning_rate": 0.0002, "epoch": 4.843585805512893, "step": 29960}, {"loss": 0.5703, "grad_norm": 1.0177865028381348, "learning_rate": 0.0002, "epoch": 4.845202489693638, "step": 29970}, {"loss": 0.5946, "grad_norm": 1.0026639699935913, "learning_rate": 0.0002, "epoch": 4.846819173874383, "step": 29980}, {"loss": 0.5897, "grad_norm": 0.9679505228996277, "learning_rate": 0.0002, "epoch": 4.8484358580551286, "step": 29990}, {"loss": 0.5621, "grad_norm": 0.8939532041549683, "learning_rate": 0.0002, "epoch": 4.850052542235874, "step": 30000}, {"loss": 0.5852, "grad_norm": 0.9957457780838013, "learning_rate": 0.0002, "epoch": 4.85166922641662, "step": 30010}, {"loss": 0.6117, "grad_norm": 1.1646790504455566, "learning_rate": 0.0002, "epoch": 4.853285910597365, "step": 30020}, {"loss": 0.5711, "grad_norm": 0.8804680705070496, "learning_rate": 0.0002, "epoch": 4.85490259477811, "step": 30030}, {"loss": 0.5397, "grad_norm": 1.161970853805542, "learning_rate": 0.0002, "epoch": 4.856519278958856, "step": 30040}, {"loss": 0.5552, "grad_norm": 0.9081037640571594, "learning_rate": 0.0002, "epoch": 4.858135963139601, "step": 30050}, {"loss": 0.6024, "grad_norm": 0.9402848482131958, "learning_rate": 0.0002, "epoch": 4.859752647320346, "step": 30060}, {"loss": 0.6256, "grad_norm": 0.9023865461349487, "learning_rate": 0.0002, "epoch": 4.861369331501091, "step": 30070}, {"loss": 0.5926, "grad_norm": 1.0173414945602417, "learning_rate": 0.0002, "epoch": 4.8629860156818365, "step": 30080}, {"loss": 0.6274, "grad_norm": 1.084402322769165, "learning_rate": 0.0002, "epoch": 4.864602699862582, "step": 30090}, {"loss": 0.6311, "grad_norm": 0.9577937126159668, "learning_rate": 0.0002, "epoch": 4.866219384043327, "step": 30100}, {"loss": 0.5724, "grad_norm": 0.9807606935501099, "learning_rate": 0.0002, "epoch": 4.867836068224072, "step": 30110}, {"loss": 0.5786, "grad_norm": 0.978784441947937, "learning_rate": 0.0002, "epoch": 4.8694527524048175, "step": 30120}, {"loss": 0.6194, "grad_norm": 0.9762914776802063, "learning_rate": 0.0002, "epoch": 4.871069436585563, "step": 30130}, {"loss": 0.5892, "grad_norm": 0.9404871463775635, "learning_rate": 0.0002, "epoch": 4.872686120766308, "step": 30140}, {"loss": 0.6182, "grad_norm": 1.0069509744644165, "learning_rate": 0.0002, "epoch": 4.874302804947053, "step": 30150}, {"loss": 0.6225, "grad_norm": 1.1770923137664795, "learning_rate": 0.0002, "epoch": 4.875919489127799, "step": 30160}, {"loss": 0.5657, "grad_norm": 1.021210789680481, "learning_rate": 0.0002, "epoch": 4.8775361733085445, "step": 30170}, {"loss": 0.6033, "grad_norm": 0.8512648940086365, "learning_rate": 0.0002, "epoch": 4.87915285748929, "step": 30180}, {"loss": 0.5519, "grad_norm": 0.9345870018005371, "learning_rate": 0.0002, "epoch": 4.880769541670035, "step": 30190}, {"loss": 0.5682, "grad_norm": 1.0224418640136719, "learning_rate": 0.0002, "epoch": 4.88238622585078, "step": 30200}, {"loss": 0.5807, "grad_norm": 1.0316044092178345, "learning_rate": 0.0002, "epoch": 4.884002910031525, "step": 30210}, {"loss": 0.6065, "grad_norm": 1.102437973022461, "learning_rate": 0.0002, "epoch": 4.885619594212271, "step": 30220}, {"loss": 0.586, "grad_norm": 1.0220023393630981, "learning_rate": 0.0002, "epoch": 4.887236278393016, "step": 30230}, {"loss": 0.5781, "grad_norm": 1.0934523344039917, "learning_rate": 0.0002, "epoch": 4.888852962573761, "step": 30240}, {"loss": 0.6313, "grad_norm": 1.264630913734436, "learning_rate": 0.0002, "epoch": 4.890469646754506, "step": 30250}, {"loss": 0.5712, "grad_norm": 1.0999879837036133, "learning_rate": 0.0002, "epoch": 4.892086330935252, "step": 30260}, {"loss": 0.6413, "grad_norm": 0.9124550223350525, "learning_rate": 0.0002, "epoch": 4.893703015115997, "step": 30270}, {"loss": 0.596, "grad_norm": 0.9853624105453491, "learning_rate": 0.0002, "epoch": 4.895319699296742, "step": 30280}, {"loss": 0.595, "grad_norm": 1.0589802265167236, "learning_rate": 0.0002, "epoch": 4.896936383477488, "step": 30290}, {"loss": 0.6129, "grad_norm": 0.8487226366996765, "learning_rate": 0.0002, "epoch": 4.8985530676582325, "step": 30300}, {"loss": 0.5514, "grad_norm": 1.0212191343307495, "learning_rate": 0.0002, "epoch": 4.900169751838979, "step": 30310}, {"loss": 0.5896, "grad_norm": 1.0187491178512573, "learning_rate": 0.0002, "epoch": 4.901786436019724, "step": 30320}, {"loss": 0.5809, "grad_norm": 1.0013091564178467, "learning_rate": 0.0002, "epoch": 4.903403120200469, "step": 30330}, {"loss": 0.5658, "grad_norm": 1.0017542839050293, "learning_rate": 0.0002, "epoch": 4.905019804381214, "step": 30340}, {"loss": 0.6002, "grad_norm": 0.9665151238441467, "learning_rate": 0.0002, "epoch": 4.9066364885619596, "step": 30350}, {"loss": 0.5864, "grad_norm": 0.8774822950363159, "learning_rate": 0.0002, "epoch": 4.908253172742705, "step": 30360}, {"loss": 0.5771, "grad_norm": 0.9449850916862488, "learning_rate": 0.0002, "epoch": 4.90986985692345, "step": 30370}, {"loss": 0.58, "grad_norm": 0.7368341088294983, "learning_rate": 0.0002, "epoch": 4.911486541104195, "step": 30380}, {"loss": 0.5992, "grad_norm": 0.9669167995452881, "learning_rate": 0.0002, "epoch": 4.9131032252849405, "step": 30390}, {"loss": 0.6202, "grad_norm": 1.1227794885635376, "learning_rate": 0.0002, "epoch": 4.914719909465686, "step": 30400}, {"loss": 0.6181, "grad_norm": 0.9884361028671265, "learning_rate": 0.0002, "epoch": 4.916336593646431, "step": 30410}, {"loss": 0.6185, "grad_norm": 0.9949551224708557, "learning_rate": 0.0002, "epoch": 4.917953277827176, "step": 30420}, {"loss": 0.5866, "grad_norm": 0.9491621851921082, "learning_rate": 0.0002, "epoch": 4.919569962007921, "step": 30430}, {"loss": 0.6005, "grad_norm": 0.78848797082901, "learning_rate": 0.0002, "epoch": 4.9211866461886675, "step": 30440}, {"loss": 0.5561, "grad_norm": 1.0693835020065308, "learning_rate": 0.0002, "epoch": 4.922803330369412, "step": 30450}, {"loss": 0.566, "grad_norm": 0.9573729634284973, "learning_rate": 0.0002, "epoch": 4.924420014550158, "step": 30460}, {"loss": 0.6084, "grad_norm": 0.9975152611732483, "learning_rate": 0.0002, "epoch": 4.926036698730903, "step": 30470}, {"loss": 0.5969, "grad_norm": 0.8695693016052246, "learning_rate": 0.0002, "epoch": 4.9276533829116484, "step": 30480}, {"loss": 0.6144, "grad_norm": 1.145394206047058, "learning_rate": 0.0002, "epoch": 4.929270067092394, "step": 30490}, {"loss": 0.5736, "grad_norm": 0.7668989896774292, "learning_rate": 0.0002, "epoch": 4.930886751273139, "step": 30500}, {"loss": 0.6052, "grad_norm": 0.9630151391029358, "learning_rate": 0.0002, "epoch": 4.932503435453884, "step": 30510}, {"loss": 0.6461, "grad_norm": 0.940705418586731, "learning_rate": 0.0002, "epoch": 4.934120119634629, "step": 30520}, {"loss": 0.6326, "grad_norm": 1.3243348598480225, "learning_rate": 0.0002, "epoch": 4.935736803815375, "step": 30530}, {"loss": 0.6174, "grad_norm": 1.004347801208496, "learning_rate": 0.0002, "epoch": 4.93735348799612, "step": 30540}, {"loss": 0.583, "grad_norm": 0.8711541295051575, "learning_rate": 0.0002, "epoch": 4.938970172176865, "step": 30550}, {"loss": 0.599, "grad_norm": 0.8980631828308105, "learning_rate": 0.0002, "epoch": 4.94058685635761, "step": 30560}, {"loss": 0.6024, "grad_norm": 0.8388893604278564, "learning_rate": 0.0002, "epoch": 4.9422035405383555, "step": 30570}, {"loss": 0.6189, "grad_norm": 1.0991183519363403, "learning_rate": 0.0002, "epoch": 4.943820224719101, "step": 30580}, {"loss": 0.5906, "grad_norm": 0.9731075763702393, "learning_rate": 0.0002, "epoch": 4.945436908899847, "step": 30590}, {"loss": 0.5883, "grad_norm": 1.3904452323913574, "learning_rate": 0.0002, "epoch": 4.947053593080591, "step": 30600}, {"loss": 0.5952, "grad_norm": 1.2489882707595825, "learning_rate": 0.0002, "epoch": 4.948670277261337, "step": 30610}, {"loss": 0.5887, "grad_norm": 1.240072250366211, "learning_rate": 0.0002, "epoch": 4.950286961442083, "step": 30620}, {"loss": 0.5762, "grad_norm": 0.9191411733627319, "learning_rate": 0.0002, "epoch": 4.951903645622828, "step": 30630}, {"loss": 0.5597, "grad_norm": 0.8888895511627197, "learning_rate": 0.0002, "epoch": 4.953520329803573, "step": 30640}, {"loss": 0.6594, "grad_norm": 0.9001450538635254, "learning_rate": 0.0002, "epoch": 4.955137013984318, "step": 30650}, {"loss": 0.6047, "grad_norm": 1.053971767425537, "learning_rate": 0.0002, "epoch": 4.9567536981650635, "step": 30660}, {"loss": 0.6107, "grad_norm": 1.2224042415618896, "learning_rate": 0.0002, "epoch": 4.958370382345809, "step": 30670}, {"loss": 0.6211, "grad_norm": 0.8855111598968506, "learning_rate": 0.0002, "epoch": 4.959987066526554, "step": 30680}, {"loss": 0.5764, "grad_norm": 0.9489575624465942, "learning_rate": 0.0002, "epoch": 4.961603750707299, "step": 30690}, {"loss": 0.5371, "grad_norm": 0.9635404944419861, "learning_rate": 0.0002, "epoch": 4.963220434888044, "step": 30700}, {"loss": 0.6043, "grad_norm": 1.1784121990203857, "learning_rate": 0.0002, "epoch": 4.96483711906879, "step": 30710}, {"loss": 0.5803, "grad_norm": 1.0059462785720825, "learning_rate": 0.0002, "epoch": 4.966453803249535, "step": 30720}, {"loss": 0.5759, "grad_norm": 0.9479738473892212, "learning_rate": 0.0002, "epoch": 4.96807048743028, "step": 30730}, {"loss": 0.584, "grad_norm": 1.0624593496322632, "learning_rate": 0.0002, "epoch": 4.969687171611026, "step": 30740}, {"loss": 0.6202, "grad_norm": 1.1429259777069092, "learning_rate": 0.0002, "epoch": 4.971303855791771, "step": 30750}, {"loss": 0.6174, "grad_norm": 0.9102491140365601, "learning_rate": 0.0002, "epoch": 4.972920539972517, "step": 30760}, {"loss": 0.6025, "grad_norm": 1.1262688636779785, "learning_rate": 0.0002, "epoch": 4.974537224153262, "step": 30770}, {"loss": 0.588, "grad_norm": 1.1415393352508545, "learning_rate": 0.0002, "epoch": 4.976153908334007, "step": 30780}, {"loss": 0.5832, "grad_norm": 1.083078384399414, "learning_rate": 0.0002, "epoch": 4.977770592514752, "step": 30790}, {"loss": 0.6025, "grad_norm": 0.964859127998352, "learning_rate": 0.0002, "epoch": 4.979387276695498, "step": 30800}, {"loss": 0.6095, "grad_norm": 0.8704743385314941, "learning_rate": 0.0002, "epoch": 4.981003960876243, "step": 30810}, {"loss": 0.5666, "grad_norm": 1.0714856386184692, "learning_rate": 0.0002, "epoch": 4.982620645056988, "step": 30820}, {"loss": 0.565, "grad_norm": 0.6818771362304688, "learning_rate": 0.0002, "epoch": 4.984237329237733, "step": 30830}, {"loss": 0.5999, "grad_norm": 1.0454156398773193, "learning_rate": 0.0002, "epoch": 4.985854013418479, "step": 30840}, {"loss": 0.5683, "grad_norm": 0.9410776495933533, "learning_rate": 0.0002, "epoch": 4.987470697599224, "step": 30850}, {"loss": 0.5899, "grad_norm": 1.0878902673721313, "learning_rate": 0.0002, "epoch": 4.989087381779969, "step": 30860}, {"loss": 0.5914, "grad_norm": 0.8916727304458618, "learning_rate": 0.0002, "epoch": 4.990704065960714, "step": 30870}, {"loss": 0.6066, "grad_norm": 1.045776128768921, "learning_rate": 0.0002, "epoch": 4.9923207501414595, "step": 30880}, {"loss": 0.5767, "grad_norm": 0.9861903786659241, "learning_rate": 0.0002, "epoch": 4.993937434322206, "step": 30890}, {"loss": 0.6192, "grad_norm": 0.9275050759315491, "learning_rate": 0.0002, "epoch": 4.995554118502951, "step": 30900}, {"loss": 0.6181, "grad_norm": 0.94013911485672, "learning_rate": 0.0002, "epoch": 4.997170802683696, "step": 30910}, {"loss": 0.614, "grad_norm": 0.9771268367767334, "learning_rate": 0.0002, "epoch": 4.998787486864441, "step": 30920}]}
+{"epoch": 6.0, "step": 37113, "epoch_duration": 16916.863001823425, "total_accumulated_duration": 101331.91647815704, "gpu_info": {"GPU_0": "NVIDIA A100-PCIE-40GB"}, "memory_usage": {"avg_memory_usage": {"GPU_0": 7568.4541015625}, "peak_memory_usage": {"GPU_0": 13792.75537109375}, "avg_memory_reserved": {"GPU_0": 17416.0}, "peak_memory_reserved": {"GPU_0": 17416.0}, "total_memory": {"GPU_0": 40444.375}}, "best_checkpoint_path": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185", "params": {"epochs": 8, "batch_size": 1, "learning_rate": 0.0002, "gradient_accumulation_steps": 2, "warmup_ratio": 0.03, "max_grad_norm": 0.3, "lora_alpha": 32, "lora_dropout": 0.05, "lora_r": 64, "tf32": true, "seed": 42}, "log_history": [{"loss": 1.6636, "grad_norm": 0.9894065856933594, "learning_rate": 0.0002, "epoch": 0.0016166841807452913, "step": 10}, {"loss": 1.1528, "grad_norm": 1.7810699939727783, "learning_rate": 0.0002, "epoch": 0.0032333683614905826, "step": 20}, {"loss": 0.9767, "grad_norm": 0.5969577431678772, "learning_rate": 0.0002, "epoch": 0.004850052542235874, "step": 30}, {"loss": 0.9772, "grad_norm": 0.6354120969772339, "learning_rate": 0.0002, "epoch": 0.006466736722981165, "step": 40}, {"loss": 0.8643, "grad_norm": 0.5604607462882996, "learning_rate": 0.0002, "epoch": 0.008083420903726457, "step": 50}, {"loss": 0.8841, "grad_norm": 0.4676193594932556, "learning_rate": 0.0002, "epoch": 0.009700105084471748, "step": 60}, {"loss": 0.9022, "grad_norm": 0.6099211573600769, "learning_rate": 0.0002, "epoch": 0.01131678926521704, "step": 70}, {"loss": 0.9133, "grad_norm": 0.48639994859695435, "learning_rate": 0.0002, "epoch": 0.01293347344596233, "step": 80}, {"loss": 0.8704, "grad_norm": 0.4904264509677887, "learning_rate": 0.0002, "epoch": 0.014550157626707623, "step": 90}, {"loss": 0.8855, "grad_norm": 2.8334362506866455, "learning_rate": 0.0002, "epoch": 0.016166841807452915, "step": 100}, {"loss": 0.8958, "grad_norm": 0.43221670389175415, "learning_rate": 0.0002, "epoch": 0.017783525988198205, "step": 110}, {"loss": 0.8412, "grad_norm": 0.42244166135787964, "learning_rate": 0.0002, "epoch": 0.019400210168943496, "step": 120}, {"loss": 0.8467, "grad_norm": 0.45363298058509827, "learning_rate": 0.0002, "epoch": 0.02101689434968879, "step": 130}, {"loss": 0.8641, "grad_norm": 0.44816508889198303, "learning_rate": 0.0002, "epoch": 0.02263357853043408, "step": 140}, {"loss": 0.8496, "grad_norm": 0.43308213353157043, "learning_rate": 0.0002, "epoch": 0.02425026271117937, "step": 150}, {"loss": 0.8213, "grad_norm": 0.4084763526916504, "learning_rate": 0.0002, "epoch": 0.02586694689192466, "step": 160}, {"loss": 0.8343, "grad_norm": 0.5363703966140747, "learning_rate": 0.0002, "epoch": 0.027483631072669955, "step": 170}, {"loss": 0.8558, "grad_norm": 0.4619699716567993, "learning_rate": 0.0002, "epoch": 0.029100315253415245, "step": 180}, {"loss": 0.8878, "grad_norm": 0.49069908261299133, "learning_rate": 0.0002, "epoch": 0.030716999434160536, "step": 190}, {"loss": 0.8867, "grad_norm": 0.4645835757255554, "learning_rate": 0.0002, "epoch": 0.03233368361490583, "step": 200}, {"loss": 0.8842, "grad_norm": 1.2411243915557861, "learning_rate": 0.0002, "epoch": 0.03395036779565112, "step": 210}, {"loss": 0.8245, "grad_norm": 0.5211851596832275, "learning_rate": 0.0002, "epoch": 0.03556705197639641, "step": 220}, {"loss": 0.8194, "grad_norm": 0.5253691673278809, "learning_rate": 0.0002, "epoch": 0.037183736157141704, "step": 230}, {"loss": 0.8856, "grad_norm": 0.4567478895187378, "learning_rate": 0.0002, "epoch": 0.03880042033788699, "step": 240}, {"loss": 0.838, "grad_norm": 0.5472128391265869, "learning_rate": 0.0002, "epoch": 0.040417104518632285, "step": 250}, {"loss": 0.8201, "grad_norm": 0.42978546023368835, "learning_rate": 0.0002, "epoch": 0.04203378869937758, "step": 260}, {"loss": 0.8334, "grad_norm": 0.601734459400177, "learning_rate": 0.0002, "epoch": 0.043650472880122866, "step": 270}, {"loss": 0.815, "grad_norm": 0.4286513328552246, "learning_rate": 0.0002, "epoch": 0.04526715706086816, "step": 280}, {"loss": 0.8758, "grad_norm": 0.5230861902236938, "learning_rate": 0.0002, "epoch": 0.046883841241613454, "step": 290}, {"loss": 0.8636, "grad_norm": 0.6504611968994141, "learning_rate": 0.0002, "epoch": 0.04850052542235874, "step": 300}, {"loss": 0.8102, "grad_norm": 0.43485215306282043, "learning_rate": 0.0002, "epoch": 0.050117209603104035, "step": 310}, {"loss": 0.8221, "grad_norm": 0.4717007875442505, "learning_rate": 0.0002, "epoch": 0.05173389378384932, "step": 320}, {"loss": 0.8469, "grad_norm": 0.4059787690639496, "learning_rate": 0.0002, "epoch": 0.053350577964594616, "step": 330}, {"loss": 0.8866, "grad_norm": 0.4366913437843323, "learning_rate": 0.0002, "epoch": 0.05496726214533991, "step": 340}, {"loss": 0.7976, "grad_norm": 0.4233848452568054, "learning_rate": 0.0002, "epoch": 0.0565839463260852, "step": 350}, {"loss": 0.8456, "grad_norm": 0.4209108352661133, "learning_rate": 0.0002, "epoch": 0.05820063050683049, "step": 360}, {"loss": 0.816, "grad_norm": 0.41637396812438965, "learning_rate": 0.0002, "epoch": 0.059817314687575784, "step": 370}, {"loss": 0.7976, "grad_norm": 0.46235376596450806, "learning_rate": 0.0002, "epoch": 0.06143399886832107, "step": 380}, {"loss": 0.7966, "grad_norm": 0.4013484716415405, "learning_rate": 0.0002, "epoch": 0.06305068304906636, "step": 390}, {"loss": 0.8253, "grad_norm": 0.47443896532058716, "learning_rate": 0.0002, "epoch": 0.06466736722981166, "step": 400}, {"loss": 0.8666, "grad_norm": 0.3942156434059143, "learning_rate": 0.0002, "epoch": 0.06628405141055695, "step": 410}, {"loss": 0.8402, "grad_norm": 0.4965320825576782, "learning_rate": 0.0002, "epoch": 0.06790073559130223, "step": 420}, {"loss": 0.8317, "grad_norm": 0.4304835796356201, "learning_rate": 0.0002, "epoch": 0.06951741977204753, "step": 430}, {"loss": 0.8528, "grad_norm": 0.511726975440979, "learning_rate": 0.0002, "epoch": 0.07113410395279282, "step": 440}, {"loss": 0.8675, "grad_norm": 0.4040689170360565, "learning_rate": 0.0002, "epoch": 0.07275078813353811, "step": 450}, {"loss": 0.8788, "grad_norm": 0.5402171015739441, "learning_rate": 0.0002, "epoch": 0.07436747231428341, "step": 460}, {"loss": 0.8737, "grad_norm": 0.4174517095088959, "learning_rate": 0.0002, "epoch": 0.0759841564950287, "step": 470}, {"loss": 0.7605, "grad_norm": 0.4306182265281677, "learning_rate": 0.0002, "epoch": 0.07760084067577398, "step": 480}, {"loss": 0.799, "grad_norm": 0.535210132598877, "learning_rate": 0.0002, "epoch": 0.07921752485651928, "step": 490}, {"loss": 0.7825, "grad_norm": 0.5339109897613525, "learning_rate": 0.0002, "epoch": 0.08083420903726457, "step": 500}, {"loss": 0.8985, "grad_norm": 0.45754891633987427, "learning_rate": 0.0002, "epoch": 0.08245089321800986, "step": 510}, {"loss": 0.8144, "grad_norm": 0.43820783495903015, "learning_rate": 0.0002, "epoch": 0.08406757739875516, "step": 520}, {"loss": 0.8001, "grad_norm": 0.4434749186038971, "learning_rate": 0.0002, "epoch": 0.08568426157950045, "step": 530}, {"loss": 0.7857, "grad_norm": 0.43111467361450195, "learning_rate": 0.0002, "epoch": 0.08730094576024573, "step": 540}, {"loss": 0.8418, "grad_norm": 0.4378940165042877, "learning_rate": 0.0002, "epoch": 0.08891762994099103, "step": 550}, {"loss": 0.8361, "grad_norm": 0.4772215187549591, "learning_rate": 0.0002, "epoch": 0.09053431412173632, "step": 560}, {"loss": 0.8268, "grad_norm": 0.6837629079818726, "learning_rate": 0.0002, "epoch": 0.09215099830248161, "step": 570}, {"loss": 0.8607, "grad_norm": 0.42241212725639343, "learning_rate": 0.0002, "epoch": 0.09376768248322691, "step": 580}, {"loss": 0.852, "grad_norm": 0.5165936350822449, "learning_rate": 0.0002, "epoch": 0.0953843666639722, "step": 590}, {"loss": 0.8664, "grad_norm": 0.48737478256225586, "learning_rate": 0.0002, "epoch": 0.09700105084471748, "step": 600}, {"loss": 0.8806, "grad_norm": 0.47419852018356323, "learning_rate": 0.0002, "epoch": 0.09861773502546278, "step": 610}, {"loss": 0.8254, "grad_norm": 0.4975486099720001, "learning_rate": 0.0002, "epoch": 0.10023441920620807, "step": 620}, {"loss": 0.8548, "grad_norm": 0.49123844504356384, "learning_rate": 0.0002, "epoch": 0.10185110338695336, "step": 630}, {"loss": 0.8911, "grad_norm": 0.6288952827453613, "learning_rate": 0.0002, "epoch": 0.10346778756769864, "step": 640}, {"loss": 0.827, "grad_norm": 0.4277345836162567, "learning_rate": 0.0002, "epoch": 0.10508447174844394, "step": 650}, {"loss": 0.7996, "grad_norm": 0.4021061956882477, "learning_rate": 0.0002, "epoch": 0.10670115592918923, "step": 660}, {"loss": 0.87, "grad_norm": 0.3492237329483032, "learning_rate": 0.0002, "epoch": 0.10831784010993452, "step": 670}, {"loss": 0.8698, "grad_norm": 0.4341012239456177, "learning_rate": 0.0002, "epoch": 0.10993452429067982, "step": 680}, {"loss": 0.781, "grad_norm": 0.7296304106712341, "learning_rate": 0.0002, "epoch": 0.1115512084714251, "step": 690}, {"loss": 0.8433, "grad_norm": 0.397494912147522, "learning_rate": 0.0002, "epoch": 0.1131678926521704, "step": 700}, {"loss": 0.827, "grad_norm": 0.396431028842926, "learning_rate": 0.0002, "epoch": 0.1147845768329157, "step": 710}, {"loss": 0.8379, "grad_norm": 0.48842838406562805, "learning_rate": 0.0002, "epoch": 0.11640126101366098, "step": 720}, {"loss": 0.8238, "grad_norm": 0.46322616934776306, "learning_rate": 0.0002, "epoch": 0.11801794519440627, "step": 730}, {"loss": 0.8041, "grad_norm": 0.47990912199020386, "learning_rate": 0.0002, "epoch": 0.11963462937515157, "step": 740}, {"loss": 0.82, "grad_norm": 0.4997142255306244, "learning_rate": 0.0002, "epoch": 0.12125131355589686, "step": 750}, {"loss": 0.7702, "grad_norm": 0.4040526747703552, "learning_rate": 0.0002, "epoch": 0.12286799773664214, "step": 760}, {"loss": 0.863, "grad_norm": 0.453095942735672, "learning_rate": 0.0002, "epoch": 0.12448468191738744, "step": 770}, {"loss": 0.8792, "grad_norm": 0.4636971950531006, "learning_rate": 0.0002, "epoch": 0.12610136609813272, "step": 780}, {"loss": 0.8112, "grad_norm": 0.4279276132583618, "learning_rate": 0.0002, "epoch": 0.12771805027887803, "step": 790}, {"loss": 0.8711, "grad_norm": 0.46212655305862427, "learning_rate": 0.0002, "epoch": 0.12933473445962332, "step": 800}, {"loss": 0.8368, "grad_norm": 0.43127650022506714, "learning_rate": 0.0002, "epoch": 0.1309514186403686, "step": 810}, {"loss": 0.8476, "grad_norm": 0.4201301336288452, "learning_rate": 0.0002, "epoch": 0.1325681028211139, "step": 820}, {"loss": 0.8078, "grad_norm": 0.42583167552948, "learning_rate": 0.0002, "epoch": 0.13418478700185918, "step": 830}, {"loss": 0.8219, "grad_norm": 0.4535622000694275, "learning_rate": 0.0002, "epoch": 0.13580147118260447, "step": 840}, {"loss": 0.8423, "grad_norm": 0.4116036891937256, "learning_rate": 0.0002, "epoch": 0.13741815536334978, "step": 850}, {"loss": 0.8466, "grad_norm": 0.45997580885887146, "learning_rate": 0.0002, "epoch": 0.13903483954409507, "step": 860}, {"loss": 0.8917, "grad_norm": 0.4487837255001068, "learning_rate": 0.0002, "epoch": 0.14065152372484035, "step": 870}, {"loss": 0.8217, "grad_norm": 0.43650057911872864, "learning_rate": 0.0002, "epoch": 0.14226820790558564, "step": 880}, {"loss": 0.8178, "grad_norm": 0.5335358381271362, "learning_rate": 0.0002, "epoch": 0.14388489208633093, "step": 890}, {"loss": 0.7957, "grad_norm": 0.5989000201225281, "learning_rate": 0.0002, "epoch": 0.14550157626707622, "step": 900}, {"loss": 0.8385, "grad_norm": 0.517179012298584, "learning_rate": 0.0002, "epoch": 0.14711826044782153, "step": 910}, {"loss": 0.8255, "grad_norm": 0.44435232877731323, "learning_rate": 0.0002, "epoch": 0.14873494462856682, "step": 920}, {"loss": 0.8305, "grad_norm": 0.42635923624038696, "learning_rate": 0.0002, "epoch": 0.1503516288093121, "step": 930}, {"loss": 0.8043, "grad_norm": 0.49603334069252014, "learning_rate": 0.0002, "epoch": 0.1519683129900574, "step": 940}, {"loss": 0.8377, "grad_norm": 0.40639808773994446, "learning_rate": 0.0002, "epoch": 0.15358499717080268, "step": 950}, {"loss": 0.8529, "grad_norm": 0.4850759208202362, "learning_rate": 0.0002, "epoch": 0.15520168135154797, "step": 960}, {"loss": 0.846, "grad_norm": 0.4427442252635956, "learning_rate": 0.0002, "epoch": 0.15681836553229328, "step": 970}, {"loss": 0.8705, "grad_norm": 0.3760930001735687, "learning_rate": 0.0002, "epoch": 0.15843504971303857, "step": 980}, {"loss": 0.8644, "grad_norm": 0.4794144332408905, "learning_rate": 0.0002, "epoch": 0.16005173389378385, "step": 990}, {"loss": 0.8002, "grad_norm": 0.45828768610954285, "learning_rate": 0.0002, "epoch": 0.16166841807452914, "step": 1000}, {"loss": 0.7658, "grad_norm": 0.6313053369522095, "learning_rate": 0.0002, "epoch": 0.16328510225527443, "step": 1010}, {"loss": 0.8047, "grad_norm": 0.45041006803512573, "learning_rate": 0.0002, "epoch": 0.16490178643601971, "step": 1020}, {"loss": 0.8423, "grad_norm": 0.441403865814209, "learning_rate": 0.0002, "epoch": 0.166518470616765, "step": 1030}, {"loss": 0.8475, "grad_norm": 0.8171296119689941, "learning_rate": 0.0002, "epoch": 0.16813515479751032, "step": 1040}, {"loss": 0.845, "grad_norm": 0.7137420773506165, "learning_rate": 0.0002, "epoch": 0.1697518389782556, "step": 1050}, {"loss": 0.8213, "grad_norm": 0.5236809849739075, "learning_rate": 0.0002, "epoch": 0.1713685231590009, "step": 1060}, {"loss": 0.8265, "grad_norm": 0.5021864175796509, "learning_rate": 0.0002, "epoch": 0.17298520733974618, "step": 1070}, {"loss": 0.8305, "grad_norm": 0.47347521781921387, "learning_rate": 0.0002, "epoch": 0.17460189152049146, "step": 1080}, {"loss": 0.8105, "grad_norm": 0.4631653428077698, "learning_rate": 0.0002, "epoch": 0.17621857570123675, "step": 1090}, {"loss": 0.8166, "grad_norm": 0.49169182777404785, "learning_rate": 0.0002, "epoch": 0.17783525988198207, "step": 1100}, {"loss": 0.8012, "grad_norm": 0.5019739270210266, "learning_rate": 0.0002, "epoch": 0.17945194406272735, "step": 1110}, {"loss": 0.8247, "grad_norm": 0.5100422501564026, "learning_rate": 0.0002, "epoch": 0.18106862824347264, "step": 1120}, {"loss": 0.8142, "grad_norm": 0.3888324499130249, "learning_rate": 0.0002, "epoch": 0.18268531242421793, "step": 1130}, {"loss": 0.8533, "grad_norm": 0.39765217900276184, "learning_rate": 0.0002, "epoch": 0.18430199660496321, "step": 1140}, {"loss": 0.8541, "grad_norm": 0.47190186381340027, "learning_rate": 0.0002, "epoch": 0.1859186807857085, "step": 1150}, {"loss": 0.8301, "grad_norm": 0.4464188814163208, "learning_rate": 0.0002, "epoch": 0.18753536496645382, "step": 1160}, {"loss": 0.8341, "grad_norm": 0.5153930187225342, "learning_rate": 0.0002, "epoch": 0.1891520491471991, "step": 1170}, {"loss": 0.8033, "grad_norm": 0.4779708683490753, "learning_rate": 0.0002, "epoch": 0.1907687333279444, "step": 1180}, {"loss": 0.8187, "grad_norm": 0.4834315776824951, "learning_rate": 0.0002, "epoch": 0.19238541750868968, "step": 1190}, {"loss": 0.7721, "grad_norm": 0.402357816696167, "learning_rate": 0.0002, "epoch": 0.19400210168943496, "step": 1200}, {"loss": 0.7941, "grad_norm": 0.45899084210395813, "learning_rate": 0.0002, "epoch": 0.19561878587018025, "step": 1210}, {"loss": 0.8353, "grad_norm": 0.5106529593467712, "learning_rate": 0.0002, "epoch": 0.19723547005092557, "step": 1220}, {"loss": 0.7816, "grad_norm": 0.45261722803115845, "learning_rate": 0.0002, "epoch": 0.19885215423167085, "step": 1230}, {"loss": 0.8068, "grad_norm": 0.4647127091884613, "learning_rate": 0.0002, "epoch": 0.20046883841241614, "step": 1240}, {"loss": 0.8239, "grad_norm": 0.4849368929862976, "learning_rate": 0.0002, "epoch": 0.20208552259316143, "step": 1250}, {"loss": 0.8514, "grad_norm": 0.4518061578273773, "learning_rate": 0.0002, "epoch": 0.2037022067739067, "step": 1260}, {"loss": 0.8158, "grad_norm": 0.49535325169563293, "learning_rate": 0.0002, "epoch": 0.205318890954652, "step": 1270}, {"loss": 0.8348, "grad_norm": 0.4835205376148224, "learning_rate": 0.0002, "epoch": 0.2069355751353973, "step": 1280}, {"loss": 0.8428, "grad_norm": 0.45308539271354675, "learning_rate": 0.0002, "epoch": 0.2085522593161426, "step": 1290}, {"loss": 0.7993, "grad_norm": 0.5369905233383179, "learning_rate": 0.0002, "epoch": 0.2101689434968879, "step": 1300}, {"loss": 0.8676, "grad_norm": 0.5031622052192688, "learning_rate": 0.0002, "epoch": 0.21178562767763318, "step": 1310}, {"loss": 0.7686, "grad_norm": 0.48010334372520447, "learning_rate": 0.0002, "epoch": 0.21340231185837846, "step": 1320}, {"loss": 0.806, "grad_norm": 0.4905701279640198, "learning_rate": 0.0002, "epoch": 0.21501899603912375, "step": 1330}, {"loss": 0.7885, "grad_norm": 0.43531742691993713, "learning_rate": 0.0002, "epoch": 0.21663568021986904, "step": 1340}, {"loss": 0.8191, "grad_norm": 0.44330692291259766, "learning_rate": 0.0002, "epoch": 0.21825236440061435, "step": 1350}, {"loss": 0.8205, "grad_norm": 0.5384416580200195, "learning_rate": 0.0002, "epoch": 0.21986904858135964, "step": 1360}, {"loss": 0.7726, "grad_norm": 0.4181833863258362, "learning_rate": 0.0002, "epoch": 0.22148573276210493, "step": 1370}, {"loss": 0.8311, "grad_norm": 0.523833692073822, "learning_rate": 0.0002, "epoch": 0.2231024169428502, "step": 1380}, {"loss": 0.7913, "grad_norm": 0.5528736710548401, "learning_rate": 0.0002, "epoch": 0.2247191011235955, "step": 1390}, {"loss": 0.8079, "grad_norm": 0.43515023589134216, "learning_rate": 0.0002, "epoch": 0.2263357853043408, "step": 1400}, {"loss": 0.8403, "grad_norm": 0.48809877038002014, "learning_rate": 0.0002, "epoch": 0.2279524694850861, "step": 1410}, {"loss": 0.8165, "grad_norm": 0.43591251969337463, "learning_rate": 0.0002, "epoch": 0.2295691536658314, "step": 1420}, {"loss": 0.8147, "grad_norm": 0.44625312089920044, "learning_rate": 0.0002, "epoch": 0.23118583784657668, "step": 1430}, {"loss": 0.8134, "grad_norm": 0.4390665292739868, "learning_rate": 0.0002, "epoch": 0.23280252202732196, "step": 1440}, {"loss": 0.8465, "grad_norm": 0.48496049642562866, "learning_rate": 0.0002, "epoch": 0.23441920620806725, "step": 1450}, {"loss": 0.775, "grad_norm": 0.45919957756996155, "learning_rate": 0.0002, "epoch": 0.23603589038881254, "step": 1460}, {"loss": 0.8659, "grad_norm": 0.5471845865249634, "learning_rate": 0.0002, "epoch": 0.23765257456955785, "step": 1470}, {"loss": 0.8164, "grad_norm": 0.47269317507743835, "learning_rate": 0.0002, "epoch": 0.23926925875030314, "step": 1480}, {"loss": 0.854, "grad_norm": 0.4930245578289032, "learning_rate": 0.0002, "epoch": 0.24088594293104842, "step": 1490}, {"loss": 0.8139, "grad_norm": 0.5605630278587341, "learning_rate": 0.0002, "epoch": 0.2425026271117937, "step": 1500}, {"loss": 0.8125, "grad_norm": 0.4435870945453644, "learning_rate": 0.0002, "epoch": 0.244119311292539, "step": 1510}, {"loss": 0.8123, "grad_norm": 0.4941999912261963, "learning_rate": 0.0002, "epoch": 0.24573599547328429, "step": 1520}, {"loss": 0.8427, "grad_norm": 0.5100624561309814, "learning_rate": 0.0002, "epoch": 0.24735267965402957, "step": 1530}, {"loss": 0.8405, "grad_norm": 0.4638267457485199, "learning_rate": 0.0002, "epoch": 0.2489693638347749, "step": 1540}, {"loss": 0.81, "grad_norm": 0.5071570873260498, "learning_rate": 0.0002, "epoch": 0.25058604801552015, "step": 1550}, {"loss": 0.7724, "grad_norm": 0.4291319251060486, "learning_rate": 0.0002, "epoch": 0.25220273219626543, "step": 1560}, {"loss": 0.7984, "grad_norm": 0.5388049483299255, "learning_rate": 0.0002, "epoch": 0.2538194163770108, "step": 1570}, {"loss": 0.8176, "grad_norm": 0.5083683729171753, "learning_rate": 0.0002, "epoch": 0.25543610055775606, "step": 1580}, {"loss": 0.843, "grad_norm": 0.4824463725090027, "learning_rate": 0.0002, "epoch": 0.25705278473850135, "step": 1590}, {"loss": 0.7996, "grad_norm": 0.41177722811698914, "learning_rate": 0.0002, "epoch": 0.25866946891924664, "step": 1600}, {"loss": 0.7772, "grad_norm": 0.5656219124794006, "learning_rate": 0.0002, "epoch": 0.2602861530999919, "step": 1610}, {"loss": 0.7955, "grad_norm": 0.41063204407691956, "learning_rate": 0.0002, "epoch": 0.2619028372807372, "step": 1620}, {"loss": 0.7998, "grad_norm": 0.4897061288356781, "learning_rate": 0.0002, "epoch": 0.2635195214614825, "step": 1630}, {"loss": 0.8198, "grad_norm": 0.4454376697540283, "learning_rate": 0.0002, "epoch": 0.2651362056422278, "step": 1640}, {"loss": 0.8684, "grad_norm": 0.4355238378047943, "learning_rate": 0.0002, "epoch": 0.26675288982297307, "step": 1650}, {"loss": 0.7801, "grad_norm": 0.458310067653656, "learning_rate": 0.0002, "epoch": 0.26836957400371836, "step": 1660}, {"loss": 0.7935, "grad_norm": 0.4752083718776703, "learning_rate": 0.0002, "epoch": 0.26998625818446365, "step": 1670}, {"loss": 0.8267, "grad_norm": 0.4666106402873993, "learning_rate": 0.0002, "epoch": 0.27160294236520893, "step": 1680}, {"loss": 0.8252, "grad_norm": 0.4213818609714508, "learning_rate": 0.0002, "epoch": 0.2732196265459543, "step": 1690}, {"loss": 0.8559, "grad_norm": 0.5768913626670837, "learning_rate": 0.0002, "epoch": 0.27483631072669956, "step": 1700}, {"loss": 0.7931, "grad_norm": 0.4209914803504944, "learning_rate": 0.0002, "epoch": 0.27645299490744485, "step": 1710}, {"loss": 0.8167, "grad_norm": 0.501909613609314, "learning_rate": 0.0002, "epoch": 0.27806967908819014, "step": 1720}, {"loss": 0.7832, "grad_norm": 0.5266261100769043, "learning_rate": 0.0002, "epoch": 0.2796863632689354, "step": 1730}, {"loss": 0.8102, "grad_norm": 0.43806859850883484, "learning_rate": 0.0002, "epoch": 0.2813030474496807, "step": 1740}, {"loss": 0.8157, "grad_norm": 0.46048814058303833, "learning_rate": 0.0002, "epoch": 0.282919731630426, "step": 1750}, {"loss": 0.8596, "grad_norm": 0.44972819089889526, "learning_rate": 0.0002, "epoch": 0.2845364158111713, "step": 1760}, {"loss": 0.8421, "grad_norm": 0.5114831328392029, "learning_rate": 0.0002, "epoch": 0.28615309999191657, "step": 1770}, {"loss": 0.8361, "grad_norm": 0.47931742668151855, "learning_rate": 0.0002, "epoch": 0.28776978417266186, "step": 1780}, {"loss": 0.8265, "grad_norm": 0.5092599987983704, "learning_rate": 0.0002, "epoch": 0.28938646835340714, "step": 1790}, {"loss": 0.8506, "grad_norm": 0.37581443786621094, "learning_rate": 0.0002, "epoch": 0.29100315253415243, "step": 1800}, {"loss": 0.7932, "grad_norm": 0.47097381949424744, "learning_rate": 0.0002, "epoch": 0.2926198367148977, "step": 1810}, {"loss": 0.7787, "grad_norm": 0.48300236463546753, "learning_rate": 0.0002, "epoch": 0.29423652089564306, "step": 1820}, {"loss": 0.8391, "grad_norm": 0.5600419640541077, "learning_rate": 0.0002, "epoch": 0.29585320507638835, "step": 1830}, {"loss": 0.8507, "grad_norm": 0.48555272817611694, "learning_rate": 0.0002, "epoch": 0.29746988925713364, "step": 1840}, {"loss": 0.7657, "grad_norm": 0.3752668499946594, "learning_rate": 0.0002, "epoch": 0.2990865734378789, "step": 1850}, {"loss": 0.7915, "grad_norm": 0.5328747034072876, "learning_rate": 0.0002, "epoch": 0.3007032576186242, "step": 1860}, {"loss": 0.8426, "grad_norm": 0.48716455698013306, "learning_rate": 0.0002, "epoch": 0.3023199417993695, "step": 1870}, {"loss": 0.8335, "grad_norm": 0.5011493563652039, "learning_rate": 0.0002, "epoch": 0.3039366259801148, "step": 1880}, {"loss": 0.852, "grad_norm": 0.46461427211761475, "learning_rate": 0.0002, "epoch": 0.30555331016086007, "step": 1890}, {"loss": 0.8478, "grad_norm": 0.36630210280418396, "learning_rate": 0.0002, "epoch": 0.30716999434160536, "step": 1900}, {"loss": 0.8162, "grad_norm": 0.4217296242713928, "learning_rate": 0.0002, "epoch": 0.30878667852235064, "step": 1910}, {"loss": 0.8128, "grad_norm": 0.4394875466823578, "learning_rate": 0.0002, "epoch": 0.31040336270309593, "step": 1920}, {"loss": 0.8471, "grad_norm": 0.6587965488433838, "learning_rate": 0.0002, "epoch": 0.3120200468838412, "step": 1930}, {"loss": 0.8565, "grad_norm": 0.5469298958778381, "learning_rate": 0.0002, "epoch": 0.31363673106458656, "step": 1940}, {"loss": 0.8236, "grad_norm": 0.4371595084667206, "learning_rate": 0.0002, "epoch": 0.31525341524533185, "step": 1950}, {"loss": 0.887, "grad_norm": 0.4809541404247284, "learning_rate": 0.0002, "epoch": 0.31687009942607713, "step": 1960}, {"loss": 0.7855, "grad_norm": 0.6061086654663086, "learning_rate": 0.0002, "epoch": 0.3184867836068224, "step": 1970}, {"loss": 0.7679, "grad_norm": 0.5342657566070557, "learning_rate": 0.0002, "epoch": 0.3201034677875677, "step": 1980}, {"loss": 0.7955, "grad_norm": 0.5057743787765503, "learning_rate": 0.0002, "epoch": 0.321720151968313, "step": 1990}, {"loss": 0.7774, "grad_norm": 0.528626024723053, "learning_rate": 0.0002, "epoch": 0.3233368361490583, "step": 2000}, {"loss": 0.8845, "grad_norm": 0.46742770075798035, "learning_rate": 0.0002, "epoch": 0.32495352032980357, "step": 2010}, {"loss": 0.8484, "grad_norm": 0.515101432800293, "learning_rate": 0.0002, "epoch": 0.32657020451054886, "step": 2020}, {"loss": 0.8139, "grad_norm": 0.41941216588020325, "learning_rate": 0.0002, "epoch": 0.32818688869129414, "step": 2030}, {"loss": 0.7637, "grad_norm": 0.49902522563934326, "learning_rate": 0.0002, "epoch": 0.32980357287203943, "step": 2040}, {"loss": 0.7822, "grad_norm": 0.4120897650718689, "learning_rate": 0.0002, "epoch": 0.3314202570527847, "step": 2050}, {"loss": 0.8057, "grad_norm": 0.45352041721343994, "learning_rate": 0.0002, "epoch": 0.33303694123353, "step": 2060}, {"loss": 0.7913, "grad_norm": 0.523199737071991, "learning_rate": 0.0002, "epoch": 0.33465362541427535, "step": 2070}, {"loss": 0.8036, "grad_norm": 0.4390358626842499, "learning_rate": 0.0002, "epoch": 0.33627030959502063, "step": 2080}, {"loss": 0.8145, "grad_norm": 0.6752901077270508, "learning_rate": 0.0002, "epoch": 0.3378869937757659, "step": 2090}, {"loss": 0.7807, "grad_norm": 0.547821044921875, "learning_rate": 0.0002, "epoch": 0.3395036779565112, "step": 2100}, {"loss": 0.8561, "grad_norm": 0.5161308646202087, "learning_rate": 0.0002, "epoch": 0.3411203621372565, "step": 2110}, {"loss": 0.7697, "grad_norm": 0.4565401077270508, "learning_rate": 0.0002, "epoch": 0.3427370463180018, "step": 2120}, {"loss": 0.7964, "grad_norm": 0.4666115939617157, "learning_rate": 0.0002, "epoch": 0.34435373049874707, "step": 2130}, {"loss": 0.8189, "grad_norm": 0.4090428352355957, "learning_rate": 0.0002, "epoch": 0.34597041467949236, "step": 2140}, {"loss": 0.8817, "grad_norm": 0.510845422744751, "learning_rate": 0.0002, "epoch": 0.34758709886023764, "step": 2150}, {"loss": 0.8398, "grad_norm": 0.42861923575401306, "learning_rate": 0.0002, "epoch": 0.34920378304098293, "step": 2160}, {"loss": 0.7716, "grad_norm": 0.4476332664489746, "learning_rate": 0.0002, "epoch": 0.3508204672217282, "step": 2170}, {"loss": 0.7845, "grad_norm": 0.6065791249275208, "learning_rate": 0.0002, "epoch": 0.3524371514024735, "step": 2180}, {"loss": 0.8187, "grad_norm": 0.42335066199302673, "learning_rate": 0.0002, "epoch": 0.35405383558321885, "step": 2190}, {"loss": 0.8239, "grad_norm": 0.5094629526138306, "learning_rate": 0.0002, "epoch": 0.35567051976396413, "step": 2200}, {"loss": 0.7807, "grad_norm": 0.5476373434066772, "learning_rate": 0.0002, "epoch": 0.3572872039447094, "step": 2210}, {"loss": 0.814, "grad_norm": 0.3911719024181366, "learning_rate": 0.0002, "epoch": 0.3589038881254547, "step": 2220}, {"loss": 0.8599, "grad_norm": 0.6599636077880859, "learning_rate": 0.0002, "epoch": 0.3605205723062, "step": 2230}, {"loss": 0.7482, "grad_norm": 0.40381914377212524, "learning_rate": 0.0002, "epoch": 0.3621372564869453, "step": 2240}, {"loss": 0.7772, "grad_norm": 0.4433908462524414, "learning_rate": 0.0002, "epoch": 0.36375394066769057, "step": 2250}, {"loss": 0.8503, "grad_norm": 0.578326940536499, "learning_rate": 0.0002, "epoch": 0.36537062484843585, "step": 2260}, {"loss": 0.8178, "grad_norm": 0.5734784007072449, "learning_rate": 0.0002, "epoch": 0.36698730902918114, "step": 2270}, {"loss": 0.8193, "grad_norm": 0.45555487275123596, "learning_rate": 0.0002, "epoch": 0.36860399320992643, "step": 2280}, {"loss": 0.7929, "grad_norm": 0.5666276216506958, "learning_rate": 0.0002, "epoch": 0.3702206773906717, "step": 2290}, {"loss": 0.8292, "grad_norm": 0.5461117625236511, "learning_rate": 0.0002, "epoch": 0.371837361571417, "step": 2300}, {"loss": 0.8204, "grad_norm": 0.6318911910057068, "learning_rate": 0.0002, "epoch": 0.3734540457521623, "step": 2310}, {"loss": 0.7964, "grad_norm": 0.493263304233551, "learning_rate": 0.0002, "epoch": 0.37507072993290763, "step": 2320}, {"loss": 0.8339, "grad_norm": 0.5888760089874268, "learning_rate": 0.0002, "epoch": 0.3766874141136529, "step": 2330}, {"loss": 0.7737, "grad_norm": 0.48671841621398926, "learning_rate": 0.0002, "epoch": 0.3783040982943982, "step": 2340}, {"loss": 0.8367, "grad_norm": 0.4385145306587219, "learning_rate": 0.0002, "epoch": 0.3799207824751435, "step": 2350}, {"loss": 0.812, "grad_norm": 0.5523318648338318, "learning_rate": 0.0002, "epoch": 0.3815374666558888, "step": 2360}, {"loss": 0.8351, "grad_norm": 0.7308220267295837, "learning_rate": 0.0002, "epoch": 0.38315415083663407, "step": 2370}, {"loss": 0.859, "grad_norm": 0.554214358329773, "learning_rate": 0.0002, "epoch": 0.38477083501737935, "step": 2380}, {"loss": 0.8146, "grad_norm": 0.5425800085067749, "learning_rate": 0.0002, "epoch": 0.38638751919812464, "step": 2390}, {"loss": 0.8282, "grad_norm": 0.48811158537864685, "learning_rate": 0.0002, "epoch": 0.3880042033788699, "step": 2400}, {"loss": 0.8074, "grad_norm": 0.49212366342544556, "learning_rate": 0.0002, "epoch": 0.3896208875596152, "step": 2410}, {"loss": 0.7991, "grad_norm": 0.5222218632698059, "learning_rate": 0.0002, "epoch": 0.3912375717403605, "step": 2420}, {"loss": 0.8182, "grad_norm": 0.4699819087982178, "learning_rate": 0.0002, "epoch": 0.3928542559211058, "step": 2430}, {"loss": 0.7919, "grad_norm": 0.46153587102890015, "learning_rate": 0.0002, "epoch": 0.39447094010185113, "step": 2440}, {"loss": 0.8111, "grad_norm": 0.4150611162185669, "learning_rate": 0.0002, "epoch": 0.3960876242825964, "step": 2450}, {"loss": 0.8589, "grad_norm": 0.5799614787101746, "learning_rate": 0.0002, "epoch": 0.3977043084633417, "step": 2460}, {"loss": 0.8085, "grad_norm": 0.56536865234375, "learning_rate": 0.0002, "epoch": 0.399320992644087, "step": 2470}, {"loss": 0.8022, "grad_norm": 0.5451247096061707, "learning_rate": 0.0002, "epoch": 0.4009376768248323, "step": 2480}, {"loss": 0.8217, "grad_norm": 0.5914521217346191, "learning_rate": 0.0002, "epoch": 0.40255436100557757, "step": 2490}, {"loss": 0.7859, "grad_norm": 0.4428117275238037, "learning_rate": 0.0002, "epoch": 0.40417104518632285, "step": 2500}, {"loss": 0.8054, "grad_norm": 0.48580947518348694, "learning_rate": 0.0002, "epoch": 0.40578772936706814, "step": 2510}, {"loss": 0.8405, "grad_norm": 0.436734676361084, "learning_rate": 0.0002, "epoch": 0.4074044135478134, "step": 2520}, {"loss": 0.8209, "grad_norm": 0.5752223134040833, "learning_rate": 0.0002, "epoch": 0.4090210977285587, "step": 2530}, {"loss": 0.8181, "grad_norm": 0.4271308183670044, "learning_rate": 0.0002, "epoch": 0.410637781909304, "step": 2540}, {"loss": 0.8058, "grad_norm": 0.46294718980789185, "learning_rate": 0.0002, "epoch": 0.4122544660900493, "step": 2550}, {"loss": 0.8473, "grad_norm": 0.49407583475112915, "learning_rate": 0.0002, "epoch": 0.4138711502707946, "step": 2560}, {"loss": 0.7881, "grad_norm": 0.4729035496711731, "learning_rate": 0.0002, "epoch": 0.4154878344515399, "step": 2570}, {"loss": 0.7834, "grad_norm": 0.4129747152328491, "learning_rate": 0.0002, "epoch": 0.4171045186322852, "step": 2580}, {"loss": 0.7859, "grad_norm": 0.5684236288070679, "learning_rate": 0.0002, "epoch": 0.4187212028130305, "step": 2590}, {"loss": 0.811, "grad_norm": 0.4862157106399536, "learning_rate": 0.0002, "epoch": 0.4203378869937758, "step": 2600}, {"loss": 0.7582, "grad_norm": 0.46567976474761963, "learning_rate": 0.0002, "epoch": 0.42195457117452106, "step": 2610}, {"loss": 0.7755, "grad_norm": 0.5710650682449341, "learning_rate": 0.0002, "epoch": 0.42357125535526635, "step": 2620}, {"loss": 0.8573, "grad_norm": 0.5660041570663452, "learning_rate": 0.0002, "epoch": 0.42518793953601164, "step": 2630}, {"loss": 0.7812, "grad_norm": 0.47944375872612, "learning_rate": 0.0002, "epoch": 0.4268046237167569, "step": 2640}, {"loss": 0.7459, "grad_norm": 0.537223756313324, "learning_rate": 0.0002, "epoch": 0.4284213078975022, "step": 2650}, {"loss": 0.8246, "grad_norm": 0.41669997572898865, "learning_rate": 0.0002, "epoch": 0.4300379920782475, "step": 2660}, {"loss": 0.7785, "grad_norm": 0.44727686047554016, "learning_rate": 0.0002, "epoch": 0.4316546762589928, "step": 2670}, {"loss": 0.8241, "grad_norm": 0.5600888729095459, "learning_rate": 0.0002, "epoch": 0.4332713604397381, "step": 2680}, {"loss": 0.7708, "grad_norm": 0.39820605516433716, "learning_rate": 0.0002, "epoch": 0.4348880446204834, "step": 2690}, {"loss": 0.8202, "grad_norm": 0.5637655854225159, "learning_rate": 0.0002, "epoch": 0.4365047288012287, "step": 2700}, {"loss": 0.855, "grad_norm": 0.6363666653633118, "learning_rate": 0.0002, "epoch": 0.438121412981974, "step": 2710}, {"loss": 0.8468, "grad_norm": 0.5656129121780396, "learning_rate": 0.0002, "epoch": 0.4397380971627193, "step": 2720}, {"loss": 0.7845, "grad_norm": 0.5600156188011169, "learning_rate": 0.0002, "epoch": 0.44135478134346456, "step": 2730}, {"loss": 0.8405, "grad_norm": 0.5506579875946045, "learning_rate": 0.0002, "epoch": 0.44297146552420985, "step": 2740}, {"loss": 0.7725, "grad_norm": 0.49878305196762085, "learning_rate": 0.0002, "epoch": 0.44458814970495514, "step": 2750}, {"loss": 0.8292, "grad_norm": 0.4569213092327118, "learning_rate": 0.0002, "epoch": 0.4462048338857004, "step": 2760}, {"loss": 0.8028, "grad_norm": 0.6056680083274841, "learning_rate": 0.0002, "epoch": 0.4478215180664457, "step": 2770}, {"loss": 0.8242, "grad_norm": 0.44474557042121887, "learning_rate": 0.0002, "epoch": 0.449438202247191, "step": 2780}, {"loss": 0.801, "grad_norm": 0.46055394411087036, "learning_rate": 0.0002, "epoch": 0.4510548864279363, "step": 2790}, {"loss": 0.7521, "grad_norm": 0.4904133379459381, "learning_rate": 0.0002, "epoch": 0.4526715706086816, "step": 2800}, {"loss": 0.8829, "grad_norm": 0.5647031664848328, "learning_rate": 0.0002, "epoch": 0.45428825478942686, "step": 2810}, {"loss": 0.8622, "grad_norm": 0.5759473443031311, "learning_rate": 0.0002, "epoch": 0.4559049389701722, "step": 2820}, {"loss": 0.7812, "grad_norm": 0.5161895751953125, "learning_rate": 0.0002, "epoch": 0.4575216231509175, "step": 2830}, {"loss": 0.8045, "grad_norm": 0.4248254597187042, "learning_rate": 0.0002, "epoch": 0.4591383073316628, "step": 2840}, {"loss": 0.7838, "grad_norm": 0.45395001769065857, "learning_rate": 0.0002, "epoch": 0.46075499151240806, "step": 2850}, {"loss": 0.8208, "grad_norm": 0.5358697772026062, "learning_rate": 0.0002, "epoch": 0.46237167569315335, "step": 2860}, {"loss": 0.8147, "grad_norm": 0.5379165410995483, "learning_rate": 0.0002, "epoch": 0.46398835987389864, "step": 2870}, {"loss": 0.7403, "grad_norm": 0.4601989686489105, "learning_rate": 0.0002, "epoch": 0.4656050440546439, "step": 2880}, {"loss": 0.8523, "grad_norm": 0.671115517616272, "learning_rate": 0.0002, "epoch": 0.4672217282353892, "step": 2890}, {"loss": 0.8262, "grad_norm": 0.4425133168697357, "learning_rate": 0.0002, "epoch": 0.4688384124161345, "step": 2900}, {"loss": 0.8178, "grad_norm": 0.5446155071258545, "learning_rate": 0.0002, "epoch": 0.4704550965968798, "step": 2910}, {"loss": 0.8106, "grad_norm": 0.603306233882904, "learning_rate": 0.0002, "epoch": 0.47207178077762507, "step": 2920}, {"loss": 0.8044, "grad_norm": 0.5377997159957886, "learning_rate": 0.0002, "epoch": 0.47368846495837036, "step": 2930}, {"loss": 0.8075, "grad_norm": 0.4931027591228485, "learning_rate": 0.0002, "epoch": 0.4753051491391157, "step": 2940}, {"loss": 0.8004, "grad_norm": 0.4711960256099701, "learning_rate": 0.0002, "epoch": 0.476921833319861, "step": 2950}, {"loss": 0.8121, "grad_norm": 0.5020492672920227, "learning_rate": 0.0002, "epoch": 0.4785385175006063, "step": 2960}, {"loss": 0.8221, "grad_norm": 0.5428946614265442, "learning_rate": 0.0002, "epoch": 0.48015520168135156, "step": 2970}, {"loss": 0.7849, "grad_norm": 0.5294089317321777, "learning_rate": 0.0002, "epoch": 0.48177188586209685, "step": 2980}, {"loss": 0.8553, "grad_norm": 0.648289144039154, "learning_rate": 0.0002, "epoch": 0.48338857004284214, "step": 2990}, {"loss": 0.7874, "grad_norm": 0.47916680574417114, "learning_rate": 0.0002, "epoch": 0.4850052542235874, "step": 3000}, {"loss": 0.8087, "grad_norm": 0.43849772214889526, "learning_rate": 0.0002, "epoch": 0.4866219384043327, "step": 3010}, {"loss": 0.7662, "grad_norm": 0.47007861733436584, "learning_rate": 0.0002, "epoch": 0.488238622585078, "step": 3020}, {"loss": 0.757, "grad_norm": 0.6314331293106079, "learning_rate": 0.0002, "epoch": 0.4898553067658233, "step": 3030}, {"loss": 0.7863, "grad_norm": 0.49211493134498596, "learning_rate": 0.0002, "epoch": 0.49147199094656857, "step": 3040}, {"loss": 0.8335, "grad_norm": 0.4537973403930664, "learning_rate": 0.0002, "epoch": 0.49308867512731386, "step": 3050}, {"loss": 0.8095, "grad_norm": 0.47326919436454773, "learning_rate": 0.0002, "epoch": 0.49470535930805914, "step": 3060}, {"loss": 0.8447, "grad_norm": 0.525874137878418, "learning_rate": 0.0002, "epoch": 0.4963220434888045, "step": 3070}, {"loss": 0.8339, "grad_norm": 0.6361091732978821, "learning_rate": 0.0002, "epoch": 0.4979387276695498, "step": 3080}, {"loss": 0.821, "grad_norm": 0.5850642919540405, "learning_rate": 0.0002, "epoch": 0.49955541185029506, "step": 3090}, {"loss": 0.8279, "grad_norm": 0.47299543023109436, "learning_rate": 0.0002, "epoch": 0.5011720960310403, "step": 3100}, {"loss": 0.8681, "grad_norm": 0.473099946975708, "learning_rate": 0.0002, "epoch": 0.5027887802117856, "step": 3110}, {"loss": 0.8223, "grad_norm": 0.48186397552490234, "learning_rate": 0.0002, "epoch": 0.5044054643925309, "step": 3120}, {"loss": 0.8292, "grad_norm": 0.5015401840209961, "learning_rate": 0.0002, "epoch": 0.5060221485732762, "step": 3130}, {"loss": 0.7692, "grad_norm": 0.5617750287055969, "learning_rate": 0.0002, "epoch": 0.5076388327540216, "step": 3140}, {"loss": 0.8708, "grad_norm": 0.5169327259063721, "learning_rate": 0.0002, "epoch": 0.5092555169347668, "step": 3150}, {"loss": 0.7845, "grad_norm": 0.545657753944397, "learning_rate": 0.0002, "epoch": 0.5108722011155121, "step": 3160}, {"loss": 0.799, "grad_norm": 0.512864351272583, "learning_rate": 0.0002, "epoch": 0.5124888852962574, "step": 3170}, {"loss": 0.7794, "grad_norm": 0.4113546311855316, "learning_rate": 0.0002, "epoch": 0.5141055694770027, "step": 3180}, {"loss": 0.8206, "grad_norm": 0.44532445073127747, "learning_rate": 0.0002, "epoch": 0.5157222536577479, "step": 3190}, {"loss": 0.8213, "grad_norm": 0.5623497366905212, "learning_rate": 0.0002, "epoch": 0.5173389378384933, "step": 3200}, {"loss": 0.7928, "grad_norm": 0.5084741115570068, "learning_rate": 0.0002, "epoch": 0.5189556220192385, "step": 3210}, {"loss": 0.8174, "grad_norm": 0.5305403470993042, "learning_rate": 0.0002, "epoch": 0.5205723061999838, "step": 3220}, {"loss": 0.8139, "grad_norm": 0.4708254337310791, "learning_rate": 0.0002, "epoch": 0.5221889903807291, "step": 3230}, {"loss": 0.7639, "grad_norm": 0.43827131390571594, "learning_rate": 0.0002, "epoch": 0.5238056745614744, "step": 3240}, {"loss": 0.7993, "grad_norm": 0.5630002617835999, "learning_rate": 0.0002, "epoch": 0.5254223587422197, "step": 3250}, {"loss": 0.7522, "grad_norm": 0.5010961890220642, "learning_rate": 0.0002, "epoch": 0.527039042922965, "step": 3260}, {"loss": 0.8374, "grad_norm": 0.6303122043609619, "learning_rate": 0.0002, "epoch": 0.5286557271037103, "step": 3270}, {"loss": 0.7727, "grad_norm": 0.5107331275939941, "learning_rate": 0.0002, "epoch": 0.5302724112844556, "step": 3280}, {"loss": 0.8495, "grad_norm": 0.5700443387031555, "learning_rate": 0.0002, "epoch": 0.5318890954652009, "step": 3290}, {"loss": 0.7776, "grad_norm": 0.46296367049217224, "learning_rate": 0.0002, "epoch": 0.5335057796459461, "step": 3300}, {"loss": 0.7931, "grad_norm": 0.531568706035614, "learning_rate": 0.0002, "epoch": 0.5351224638266915, "step": 3310}, {"loss": 0.843, "grad_norm": 0.4686741530895233, "learning_rate": 0.0002, "epoch": 0.5367391480074367, "step": 3320}, {"loss": 0.8104, "grad_norm": 0.5404331088066101, "learning_rate": 0.0002, "epoch": 0.5383558321881821, "step": 3330}, {"loss": 0.7686, "grad_norm": 0.6368790864944458, "learning_rate": 0.0002, "epoch": 0.5399725163689273, "step": 3340}, {"loss": 0.8514, "grad_norm": 0.42300888895988464, "learning_rate": 0.0002, "epoch": 0.5415892005496726, "step": 3350}, {"loss": 0.8236, "grad_norm": 0.5362542867660522, "learning_rate": 0.0002, "epoch": 0.5432058847304179, "step": 3360}, {"loss": 0.858, "grad_norm": 0.497128963470459, "learning_rate": 0.0002, "epoch": 0.5448225689111632, "step": 3370}, {"loss": 0.8519, "grad_norm": 0.5006386041641235, "learning_rate": 0.0002, "epoch": 0.5464392530919085, "step": 3380}, {"loss": 0.7867, "grad_norm": 0.44136837124824524, "learning_rate": 0.0002, "epoch": 0.5480559372726538, "step": 3390}, {"loss": 0.773, "grad_norm": 0.5897833108901978, "learning_rate": 0.0002, "epoch": 0.5496726214533991, "step": 3400}, {"loss": 0.8895, "grad_norm": 0.641075611114502, "learning_rate": 0.0002, "epoch": 0.5512893056341444, "step": 3410}, {"loss": 0.7827, "grad_norm": 0.7251322269439697, "learning_rate": 0.0002, "epoch": 0.5529059898148897, "step": 3420}, {"loss": 0.7626, "grad_norm": 0.47411349415779114, "learning_rate": 0.0002, "epoch": 0.5545226739956349, "step": 3430}, {"loss": 0.8196, "grad_norm": 0.4994310438632965, "learning_rate": 0.0002, "epoch": 0.5561393581763803, "step": 3440}, {"loss": 0.7812, "grad_norm": 0.5814438462257385, "learning_rate": 0.0002, "epoch": 0.5577560423571255, "step": 3450}, {"loss": 0.8805, "grad_norm": 0.6278898119926453, "learning_rate": 0.0002, "epoch": 0.5593727265378708, "step": 3460}, {"loss": 0.813, "grad_norm": 0.46208274364471436, "learning_rate": 0.0002, "epoch": 0.5609894107186161, "step": 3470}, {"loss": 0.8295, "grad_norm": 0.5718930959701538, "learning_rate": 0.0002, "epoch": 0.5626060948993614, "step": 3480}, {"loss": 0.8152, "grad_norm": 0.48178744316101074, "learning_rate": 0.0002, "epoch": 0.5642227790801067, "step": 3490}, {"loss": 0.8244, "grad_norm": 0.47336965799331665, "learning_rate": 0.0002, "epoch": 0.565839463260852, "step": 3500}, {"loss": 0.8099, "grad_norm": 0.43442684412002563, "learning_rate": 0.0002, "epoch": 0.5674561474415973, "step": 3510}, {"loss": 0.7564, "grad_norm": 0.6463358998298645, "learning_rate": 0.0002, "epoch": 0.5690728316223426, "step": 3520}, {"loss": 0.836, "grad_norm": 0.5286486744880676, "learning_rate": 0.0002, "epoch": 0.5706895158030879, "step": 3530}, {"loss": 0.8421, "grad_norm": 0.5405499935150146, "learning_rate": 0.0002, "epoch": 0.5723061999838331, "step": 3540}, {"loss": 0.7614, "grad_norm": 0.6654391884803772, "learning_rate": 0.0002, "epoch": 0.5739228841645785, "step": 3550}, {"loss": 0.7803, "grad_norm": 0.5081980228424072, "learning_rate": 0.0002, "epoch": 0.5755395683453237, "step": 3560}, {"loss": 0.7753, "grad_norm": 0.48978179693222046, "learning_rate": 0.0002, "epoch": 0.5771562525260691, "step": 3570}, {"loss": 0.8151, "grad_norm": 0.5840612053871155, "learning_rate": 0.0002, "epoch": 0.5787729367068143, "step": 3580}, {"loss": 0.8937, "grad_norm": 0.5235261917114258, "learning_rate": 0.0002, "epoch": 0.5803896208875596, "step": 3590}, {"loss": 0.7894, "grad_norm": 0.5672075748443604, "learning_rate": 0.0002, "epoch": 0.5820063050683049, "step": 3600}, {"loss": 0.8347, "grad_norm": 0.5613429546356201, "learning_rate": 0.0002, "epoch": 0.5836229892490502, "step": 3610}, {"loss": 0.8274, "grad_norm": 0.4032273590564728, "learning_rate": 0.0002, "epoch": 0.5852396734297954, "step": 3620}, {"loss": 0.8421, "grad_norm": 0.49559324979782104, "learning_rate": 0.0002, "epoch": 0.5868563576105408, "step": 3630}, {"loss": 0.8332, "grad_norm": 0.6895697712898254, "learning_rate": 0.0002, "epoch": 0.5884730417912861, "step": 3640}, {"loss": 0.7877, "grad_norm": 0.4750136435031891, "learning_rate": 0.0002, "epoch": 0.5900897259720314, "step": 3650}, {"loss": 0.8219, "grad_norm": 0.5176819562911987, "learning_rate": 0.0002, "epoch": 0.5917064101527767, "step": 3660}, {"loss": 0.8151, "grad_norm": 0.5817760229110718, "learning_rate": 0.0002, "epoch": 0.5933230943335219, "step": 3670}, {"loss": 0.7823, "grad_norm": 0.6064626574516296, "learning_rate": 0.0002, "epoch": 0.5949397785142673, "step": 3680}, {"loss": 0.8422, "grad_norm": 0.6728700995445251, "learning_rate": 0.0002, "epoch": 0.5965564626950125, "step": 3690}, {"loss": 0.7679, "grad_norm": 0.609305202960968, "learning_rate": 0.0002, "epoch": 0.5981731468757578, "step": 3700}, {"loss": 0.8048, "grad_norm": 0.4615488350391388, "learning_rate": 0.0002, "epoch": 0.5997898310565031, "step": 3710}, {"loss": 0.8214, "grad_norm": 2.0531179904937744, "learning_rate": 0.0002, "epoch": 0.6014065152372484, "step": 3720}, {"loss": 0.8158, "grad_norm": 0.5091132521629333, "learning_rate": 0.0002, "epoch": 0.6030231994179936, "step": 3730}, {"loss": 0.7833, "grad_norm": 0.5951124429702759, "learning_rate": 0.0002, "epoch": 0.604639883598739, "step": 3740}, {"loss": 0.7784, "grad_norm": 0.5870208144187927, "learning_rate": 0.0002, "epoch": 0.6062565677794842, "step": 3750}, {"loss": 0.8044, "grad_norm": 0.6254619359970093, "learning_rate": 0.0002, "epoch": 0.6078732519602296, "step": 3760}, {"loss": 0.7868, "grad_norm": 0.5577626824378967, "learning_rate": 0.0002, "epoch": 0.6094899361409749, "step": 3770}, {"loss": 0.8108, "grad_norm": 0.5004405379295349, "learning_rate": 0.0002, "epoch": 0.6111066203217201, "step": 3780}, {"loss": 0.8092, "grad_norm": 0.5527383685112, "learning_rate": 0.0002, "epoch": 0.6127233045024655, "step": 3790}, {"loss": 0.8036, "grad_norm": 0.49116113781929016, "learning_rate": 0.0002, "epoch": 0.6143399886832107, "step": 3800}, {"loss": 0.8352, "grad_norm": 0.5299299359321594, "learning_rate": 0.0002, "epoch": 0.6159566728639561, "step": 3810}, {"loss": 0.7737, "grad_norm": 0.464897483587265, "learning_rate": 0.0002, "epoch": 0.6175733570447013, "step": 3820}, {"loss": 0.7923, "grad_norm": 0.6505740880966187, "learning_rate": 0.0002, "epoch": 0.6191900412254466, "step": 3830}, {"loss": 0.8123, "grad_norm": 0.5512559413909912, "learning_rate": 0.0002, "epoch": 0.6208067254061919, "step": 3840}, {"loss": 0.8856, "grad_norm": 0.49427518248558044, "learning_rate": 0.0002, "epoch": 0.6224234095869372, "step": 3850}, {"loss": 0.7751, "grad_norm": 0.3839147090911865, "learning_rate": 0.0002, "epoch": 0.6240400937676824, "step": 3860}, {"loss": 0.8006, "grad_norm": 0.5760218501091003, "learning_rate": 0.0002, "epoch": 0.6256567779484278, "step": 3870}, {"loss": 0.7836, "grad_norm": 0.7226507067680359, "learning_rate": 0.0002, "epoch": 0.6272734621291731, "step": 3880}, {"loss": 0.8244, "grad_norm": 0.676781415939331, "learning_rate": 0.0002, "epoch": 0.6288901463099184, "step": 3890}, {"loss": 0.8239, "grad_norm": 0.4284018278121948, "learning_rate": 0.0002, "epoch": 0.6305068304906637, "step": 3900}, {"loss": 0.7996, "grad_norm": 0.5060628056526184, "learning_rate": 0.0002, "epoch": 0.6321235146714089, "step": 3910}, {"loss": 0.8089, "grad_norm": 0.5524522066116333, "learning_rate": 0.0002, "epoch": 0.6337401988521543, "step": 3920}, {"loss": 0.8276, "grad_norm": 0.6099881529808044, "learning_rate": 0.0002, "epoch": 0.6353568830328995, "step": 3930}, {"loss": 0.809, "grad_norm": 0.43155938386917114, "learning_rate": 0.0002, "epoch": 0.6369735672136448, "step": 3940}, {"loss": 0.8404, "grad_norm": 0.6427084803581238, "learning_rate": 0.0002, "epoch": 0.6385902513943901, "step": 3950}, {"loss": 0.8368, "grad_norm": 0.541220486164093, "learning_rate": 0.0002, "epoch": 0.6402069355751354, "step": 3960}, {"loss": 0.8539, "grad_norm": 0.5414294600486755, "learning_rate": 0.0002, "epoch": 0.6418236197558806, "step": 3970}, {"loss": 0.7996, "grad_norm": 0.46344003081321716, "learning_rate": 0.0002, "epoch": 0.643440303936626, "step": 3980}, {"loss": 0.7474, "grad_norm": 0.45209285616874695, "learning_rate": 0.0002, "epoch": 0.6450569881173712, "step": 3990}, {"loss": 0.8202, "grad_norm": 0.5417284369468689, "learning_rate": 0.0002, "epoch": 0.6466736722981166, "step": 4000}, {"loss": 0.7563, "grad_norm": 0.7995685935020447, "learning_rate": 0.0002, "epoch": 0.6482903564788619, "step": 4010}, {"loss": 0.7812, "grad_norm": 0.6384002566337585, "learning_rate": 0.0002, "epoch": 0.6499070406596071, "step": 4020}, {"loss": 0.732, "grad_norm": 0.4472815692424774, "learning_rate": 0.0002, "epoch": 0.6515237248403525, "step": 4030}, {"loss": 0.8071, "grad_norm": 0.6834294199943542, "learning_rate": 0.0002, "epoch": 0.6531404090210977, "step": 4040}, {"loss": 0.7812, "grad_norm": 0.4612339735031128, "learning_rate": 0.0002, "epoch": 0.654757093201843, "step": 4050}, {"loss": 0.8141, "grad_norm": 0.9266576170921326, "learning_rate": 0.0002, "epoch": 0.6563737773825883, "step": 4060}, {"loss": 0.7991, "grad_norm": 0.4470861852169037, "learning_rate": 0.0002, "epoch": 0.6579904615633336, "step": 4070}, {"loss": 0.8293, "grad_norm": 0.45544925332069397, "learning_rate": 0.0002, "epoch": 0.6596071457440789, "step": 4080}, {"loss": 0.8455, "grad_norm": 0.6144481301307678, "learning_rate": 0.0002, "epoch": 0.6612238299248242, "step": 4090}, {"loss": 0.7877, "grad_norm": 0.5936288237571716, "learning_rate": 0.0002, "epoch": 0.6628405141055694, "step": 4100}, {"loss": 0.7617, "grad_norm": 0.4822963774204254, "learning_rate": 0.0002, "epoch": 0.6644571982863148, "step": 4110}, {"loss": 0.7997, "grad_norm": 0.48432496190071106, "learning_rate": 0.0002, "epoch": 0.66607388246706, "step": 4120}, {"loss": 0.8404, "grad_norm": 0.4901607930660248, "learning_rate": 0.0002, "epoch": 0.6676905666478054, "step": 4130}, {"loss": 0.8085, "grad_norm": 0.5018393397331238, "learning_rate": 0.0002, "epoch": 0.6693072508285507, "step": 4140}, {"loss": 0.8065, "grad_norm": 0.6946378946304321, "learning_rate": 0.0002, "epoch": 0.6709239350092959, "step": 4150}, {"loss": 0.8147, "grad_norm": 0.5997390747070312, "learning_rate": 0.0002, "epoch": 0.6725406191900413, "step": 4160}, {"loss": 0.8268, "grad_norm": 0.6738849878311157, "learning_rate": 0.0002, "epoch": 0.6741573033707865, "step": 4170}, {"loss": 0.7704, "grad_norm": 0.6110581159591675, "learning_rate": 0.0002, "epoch": 0.6757739875515318, "step": 4180}, {"loss": 0.8043, "grad_norm": 0.5703322291374207, "learning_rate": 0.0002, "epoch": 0.6773906717322771, "step": 4190}, {"loss": 0.8099, "grad_norm": 0.4686066210269928, "learning_rate": 0.0002, "epoch": 0.6790073559130224, "step": 4200}, {"loss": 0.8441, "grad_norm": 0.6394643783569336, "learning_rate": 0.0002, "epoch": 0.6806240400937676, "step": 4210}, {"loss": 0.8011, "grad_norm": 0.5454841256141663, "learning_rate": 0.0002, "epoch": 0.682240724274513, "step": 4220}, {"loss": 0.8307, "grad_norm": 0.4859732985496521, "learning_rate": 0.0002, "epoch": 0.6838574084552582, "step": 4230}, {"loss": 0.8161, "grad_norm": 0.5544065833091736, "learning_rate": 0.0002, "epoch": 0.6854740926360036, "step": 4240}, {"loss": 0.7839, "grad_norm": 0.4902505576610565, "learning_rate": 0.0002, "epoch": 0.6870907768167488, "step": 4250}, {"loss": 0.7977, "grad_norm": 0.4768051505088806, "learning_rate": 0.0002, "epoch": 0.6887074609974941, "step": 4260}, {"loss": 0.7539, "grad_norm": 0.49982190132141113, "learning_rate": 0.0002, "epoch": 0.6903241451782395, "step": 4270}, {"loss": 0.7353, "grad_norm": 0.6351838111877441, "learning_rate": 0.0002, "epoch": 0.6919408293589847, "step": 4280}, {"loss": 0.7664, "grad_norm": 0.5647561550140381, "learning_rate": 0.0002, "epoch": 0.69355751353973, "step": 4290}, {"loss": 0.7618, "grad_norm": 0.5340486764907837, "learning_rate": 0.0002, "epoch": 0.6951741977204753, "step": 4300}, {"loss": 0.8526, "grad_norm": 0.5649092793464661, "learning_rate": 0.0002, "epoch": 0.6967908819012206, "step": 4310}, {"loss": 0.8246, "grad_norm": 0.6183916926383972, "learning_rate": 0.0002, "epoch": 0.6984075660819659, "step": 4320}, {"loss": 0.792, "grad_norm": 0.6154509782791138, "learning_rate": 0.0002, "epoch": 0.7000242502627112, "step": 4330}, {"loss": 0.8397, "grad_norm": 0.5156264305114746, "learning_rate": 0.0002, "epoch": 0.7016409344434564, "step": 4340}, {"loss": 0.8512, "grad_norm": 0.562171459197998, "learning_rate": 0.0002, "epoch": 0.7032576186242018, "step": 4350}, {"loss": 0.7882, "grad_norm": 0.4949502646923065, "learning_rate": 0.0002, "epoch": 0.704874302804947, "step": 4360}, {"loss": 0.738, "grad_norm": 0.5171684622764587, "learning_rate": 0.0002, "epoch": 0.7064909869856923, "step": 4370}, {"loss": 0.8001, "grad_norm": 0.6198443174362183, "learning_rate": 0.0002, "epoch": 0.7081076711664377, "step": 4380}, {"loss": 0.7606, "grad_norm": 0.5802276134490967, "learning_rate": 0.0002, "epoch": 0.7097243553471829, "step": 4390}, {"loss": 0.8797, "grad_norm": 0.41096967458724976, "learning_rate": 0.0002, "epoch": 0.7113410395279283, "step": 4400}, {"loss": 0.805, "grad_norm": 0.4397392272949219, "learning_rate": 0.0002, "epoch": 0.7129577237086735, "step": 4410}, {"loss": 0.7651, "grad_norm": 0.45228442549705505, "learning_rate": 0.0002, "epoch": 0.7145744078894188, "step": 4420}, {"loss": 0.7938, "grad_norm": 0.4839673936367035, "learning_rate": 0.0002, "epoch": 0.7161910920701641, "step": 4430}, {"loss": 0.8362, "grad_norm": 0.6140755414962769, "learning_rate": 0.0002, "epoch": 0.7178077762509094, "step": 4440}, {"loss": 0.7722, "grad_norm": 0.6841378808021545, "learning_rate": 0.0002, "epoch": 0.7194244604316546, "step": 4450}, {"loss": 0.8177, "grad_norm": 0.6664239168167114, "learning_rate": 0.0002, "epoch": 0.7210411446124, "step": 4460}, {"loss": 0.7983, "grad_norm": 0.47552719712257385, "learning_rate": 0.0002, "epoch": 0.7226578287931452, "step": 4470}, {"loss": 0.8982, "grad_norm": 0.6649776101112366, "learning_rate": 0.0002, "epoch": 0.7242745129738906, "step": 4480}, {"loss": 0.8074, "grad_norm": 0.5159541964530945, "learning_rate": 0.0002, "epoch": 0.7258911971546358, "step": 4490}, {"loss": 0.7786, "grad_norm": 0.6693112850189209, "learning_rate": 0.0002, "epoch": 0.7275078813353811, "step": 4500}, {"loss": 0.8655, "grad_norm": 0.48870977759361267, "learning_rate": 0.0002, "epoch": 0.7291245655161265, "step": 4510}, {"loss": 0.7337, "grad_norm": 0.4857887923717499, "learning_rate": 0.0002, "epoch": 0.7307412496968717, "step": 4520}, {"loss": 0.8026, "grad_norm": 0.5515662431716919, "learning_rate": 0.0002, "epoch": 0.732357933877617, "step": 4530}, {"loss": 0.8031, "grad_norm": 0.6292222738265991, "learning_rate": 0.0002, "epoch": 0.7339746180583623, "step": 4540}, {"loss": 0.7749, "grad_norm": 0.48265689611434937, "learning_rate": 0.0002, "epoch": 0.7355913022391076, "step": 4550}, {"loss": 0.8499, "grad_norm": 0.8044266104698181, "learning_rate": 0.0002, "epoch": 0.7372079864198529, "step": 4560}, {"loss": 0.8162, "grad_norm": 0.6111769676208496, "learning_rate": 0.0002, "epoch": 0.7388246706005982, "step": 4570}, {"loss": 0.7291, "grad_norm": 0.5229553580284119, "learning_rate": 0.0002, "epoch": 0.7404413547813434, "step": 4580}, {"loss": 0.8038, "grad_norm": 0.6054152250289917, "learning_rate": 0.0002, "epoch": 0.7420580389620888, "step": 4590}, {"loss": 0.8169, "grad_norm": 0.5574966669082642, "learning_rate": 0.0002, "epoch": 0.743674723142834, "step": 4600}, {"loss": 0.8439, "grad_norm": 0.5395817160606384, "learning_rate": 0.0002, "epoch": 0.7452914073235793, "step": 4610}, {"loss": 0.8495, "grad_norm": 0.7116472721099854, "learning_rate": 0.0002, "epoch": 0.7469080915043246, "step": 4620}, {"loss": 0.7743, "grad_norm": 0.5618700981140137, "learning_rate": 0.0002, "epoch": 0.7485247756850699, "step": 4630}, {"loss": 0.7744, "grad_norm": 0.5802770853042603, "learning_rate": 0.0002, "epoch": 0.7501414598658153, "step": 4640}, {"loss": 0.7924, "grad_norm": 0.5690428018569946, "learning_rate": 0.0002, "epoch": 0.7517581440465605, "step": 4650}, {"loss": 0.8017, "grad_norm": 0.4813360273838043, "learning_rate": 0.0002, "epoch": 0.7533748282273058, "step": 4660}, {"loss": 0.8108, "grad_norm": 0.5434042811393738, "learning_rate": 0.0002, "epoch": 0.7549915124080511, "step": 4670}, {"loss": 0.7824, "grad_norm": 0.5502099990844727, "learning_rate": 0.0002, "epoch": 0.7566081965887964, "step": 4680}, {"loss": 0.8598, "grad_norm": 0.6020621061325073, "learning_rate": 0.0002, "epoch": 0.7582248807695416, "step": 4690}, {"loss": 0.7937, "grad_norm": 0.4922301471233368, "learning_rate": 0.0002, "epoch": 0.759841564950287, "step": 4700}, {"loss": 0.788, "grad_norm": 0.6492828726768494, "learning_rate": 0.0002, "epoch": 0.7614582491310322, "step": 4710}, {"loss": 0.8313, "grad_norm": 0.4865580201148987, "learning_rate": 0.0002, "epoch": 0.7630749333117776, "step": 4720}, {"loss": 0.7966, "grad_norm": 0.5971422791481018, "learning_rate": 0.0002, "epoch": 0.7646916174925228, "step": 4730}, {"loss": 0.8298, "grad_norm": 0.6832674145698547, "learning_rate": 0.0002, "epoch": 0.7663083016732681, "step": 4740}, {"loss": 0.8156, "grad_norm": 0.500908613204956, "learning_rate": 0.0002, "epoch": 0.7679249858540134, "step": 4750}, {"loss": 0.8383, "grad_norm": 0.6112465858459473, "learning_rate": 0.0002, "epoch": 0.7695416700347587, "step": 4760}, {"loss": 0.76, "grad_norm": 0.5753506422042847, "learning_rate": 0.0002, "epoch": 0.771158354215504, "step": 4770}, {"loss": 0.8297, "grad_norm": 0.6529405117034912, "learning_rate": 0.0002, "epoch": 0.7727750383962493, "step": 4780}, {"loss": 0.8171, "grad_norm": 0.5916843414306641, "learning_rate": 0.0002, "epoch": 0.7743917225769946, "step": 4790}, {"loss": 0.83, "grad_norm": 0.4821224510669708, "learning_rate": 0.0002, "epoch": 0.7760084067577399, "step": 4800}, {"loss": 0.7703, "grad_norm": 0.5532580018043518, "learning_rate": 0.0002, "epoch": 0.7776250909384852, "step": 4810}, {"loss": 0.7363, "grad_norm": 0.4604877233505249, "learning_rate": 0.0002, "epoch": 0.7792417751192304, "step": 4820}, {"loss": 0.7506, "grad_norm": 0.5009613037109375, "learning_rate": 0.0002, "epoch": 0.7808584592999758, "step": 4830}, {"loss": 0.7863, "grad_norm": 0.6448560357093811, "learning_rate": 0.0002, "epoch": 0.782475143480721, "step": 4840}, {"loss": 0.7957, "grad_norm": 0.44327953457832336, "learning_rate": 0.0002, "epoch": 0.7840918276614663, "step": 4850}, {"loss": 0.7925, "grad_norm": 0.5355411171913147, "learning_rate": 0.0002, "epoch": 0.7857085118422116, "step": 4860}, {"loss": 0.7754, "grad_norm": 0.5635677576065063, "learning_rate": 0.0002, "epoch": 0.7873251960229569, "step": 4870}, {"loss": 0.7931, "grad_norm": 0.5417491793632507, "learning_rate": 0.0002, "epoch": 0.7889418802037023, "step": 4880}, {"loss": 0.7819, "grad_norm": 0.4567430913448334, "learning_rate": 0.0002, "epoch": 0.7905585643844475, "step": 4890}, {"loss": 0.8454, "grad_norm": 0.44651296734809875, "learning_rate": 0.0002, "epoch": 0.7921752485651928, "step": 4900}, {"loss": 0.7959, "grad_norm": 0.5741217136383057, "learning_rate": 0.0002, "epoch": 0.7937919327459381, "step": 4910}, {"loss": 0.8093, "grad_norm": 0.6605045199394226, "learning_rate": 0.0002, "epoch": 0.7954086169266834, "step": 4920}, {"loss": 0.77, "grad_norm": 0.5126531720161438, "learning_rate": 0.0002, "epoch": 0.7970253011074286, "step": 4930}, {"loss": 0.7793, "grad_norm": 0.513648271560669, "learning_rate": 0.0002, "epoch": 0.798641985288174, "step": 4940}, {"loss": 0.8314, "grad_norm": 0.5350404381752014, "learning_rate": 0.0002, "epoch": 0.8002586694689192, "step": 4950}, {"loss": 0.7649, "grad_norm": 0.5731674432754517, "learning_rate": 0.0002, "epoch": 0.8018753536496646, "step": 4960}, {"loss": 0.8572, "grad_norm": 0.5974258184432983, "learning_rate": 0.0002, "epoch": 0.8034920378304098, "step": 4970}, {"loss": 0.7972, "grad_norm": 0.8774799704551697, "learning_rate": 0.0002, "epoch": 0.8051087220111551, "step": 4980}, {"loss": 0.7899, "grad_norm": 0.5994430184364319, "learning_rate": 0.0002, "epoch": 0.8067254061919004, "step": 4990}, {"loss": 0.7736, "grad_norm": 0.4894903004169464, "learning_rate": 0.0002, "epoch": 0.8083420903726457, "step": 5000}, {"loss": 0.78, "grad_norm": 0.5218459367752075, "learning_rate": 0.0002, "epoch": 0.809958774553391, "step": 5010}, {"loss": 0.817, "grad_norm": 0.5232468843460083, "learning_rate": 0.0002, "epoch": 0.8115754587341363, "step": 5020}, {"loss": 0.7704, "grad_norm": 0.44358372688293457, "learning_rate": 0.0002, "epoch": 0.8131921429148816, "step": 5030}, {"loss": 0.785, "grad_norm": 0.6202037334442139, "learning_rate": 0.0002, "epoch": 0.8148088270956269, "step": 5040}, {"loss": 0.7351, "grad_norm": 0.7721474170684814, "learning_rate": 0.0002, "epoch": 0.8164255112763722, "step": 5050}, {"loss": 0.8297, "grad_norm": 0.5568501353263855, "learning_rate": 0.0002, "epoch": 0.8180421954571174, "step": 5060}, {"loss": 0.7733, "grad_norm": 0.49148809909820557, "learning_rate": 0.0002, "epoch": 0.8196588796378628, "step": 5070}, {"loss": 0.8054, "grad_norm": 0.4956012964248657, "learning_rate": 0.0002, "epoch": 0.821275563818608, "step": 5080}, {"loss": 0.8201, "grad_norm": 0.6078833937644958, "learning_rate": 0.0002, "epoch": 0.8228922479993533, "step": 5090}, {"loss": 0.828, "grad_norm": 0.46906954050064087, "learning_rate": 0.0002, "epoch": 0.8245089321800986, "step": 5100}, {"loss": 0.7703, "grad_norm": 0.50812166929245, "learning_rate": 0.0002, "epoch": 0.8261256163608439, "step": 5110}, {"loss": 0.8243, "grad_norm": 0.5319661498069763, "learning_rate": 0.0002, "epoch": 0.8277423005415891, "step": 5120}, {"loss": 0.7798, "grad_norm": 0.4949689209461212, "learning_rate": 0.0002, "epoch": 0.8293589847223345, "step": 5130}, {"loss": 0.7428, "grad_norm": 0.5151591300964355, "learning_rate": 0.0002, "epoch": 0.8309756689030798, "step": 5140}, {"loss": 0.8147, "grad_norm": 0.5530214309692383, "learning_rate": 0.0002, "epoch": 0.8325923530838251, "step": 5150}, {"loss": 0.8251, "grad_norm": 0.6297410130500793, "learning_rate": 0.0002, "epoch": 0.8342090372645704, "step": 5160}, {"loss": 0.8067, "grad_norm": 0.5466840267181396, "learning_rate": 0.0002, "epoch": 0.8358257214453156, "step": 5170}, {"loss": 0.7875, "grad_norm": 0.652913510799408, "learning_rate": 0.0002, "epoch": 0.837442405626061, "step": 5180}, {"loss": 0.8295, "grad_norm": 0.5811293125152588, "learning_rate": 0.0002, "epoch": 0.8390590898068062, "step": 5190}, {"loss": 0.7412, "grad_norm": 0.5109550952911377, "learning_rate": 0.0002, "epoch": 0.8406757739875516, "step": 5200}, {"loss": 0.8077, "grad_norm": 0.4551706612110138, "learning_rate": 0.0002, "epoch": 0.8422924581682968, "step": 5210}, {"loss": 0.7827, "grad_norm": 0.5813754200935364, "learning_rate": 0.0002, "epoch": 0.8439091423490421, "step": 5220}, {"loss": 0.802, "grad_norm": 0.5856947898864746, "learning_rate": 0.0002, "epoch": 0.8455258265297874, "step": 5230}, {"loss": 0.7957, "grad_norm": 0.5482739210128784, "learning_rate": 0.0002, "epoch": 0.8471425107105327, "step": 5240}, {"loss": 0.8295, "grad_norm": 0.49023720622062683, "learning_rate": 0.0002, "epoch": 0.8487591948912779, "step": 5250}, {"loss": 0.8022, "grad_norm": 0.49472475051879883, "learning_rate": 0.0002, "epoch": 0.8503758790720233, "step": 5260}, {"loss": 0.8001, "grad_norm": 0.5490226745605469, "learning_rate": 0.0002, "epoch": 0.8519925632527686, "step": 5270}, {"loss": 0.8333, "grad_norm": 0.5340665578842163, "learning_rate": 0.0002, "epoch": 0.8536092474335139, "step": 5280}, {"loss": 0.8277, "grad_norm": 0.5962483882904053, "learning_rate": 0.0002, "epoch": 0.8552259316142592, "step": 5290}, {"loss": 0.8765, "grad_norm": 0.586358368396759, "learning_rate": 0.0002, "epoch": 0.8568426157950044, "step": 5300}, {"loss": 0.7831, "grad_norm": 0.49120277166366577, "learning_rate": 0.0002, "epoch": 0.8584592999757498, "step": 5310}, {"loss": 0.8162, "grad_norm": 0.5887332558631897, "learning_rate": 0.0002, "epoch": 0.860075984156495, "step": 5320}, {"loss": 0.7464, "grad_norm": 0.42496153712272644, "learning_rate": 0.0002, "epoch": 0.8616926683372403, "step": 5330}, {"loss": 0.7905, "grad_norm": 0.5489874482154846, "learning_rate": 0.0002, "epoch": 0.8633093525179856, "step": 5340}, {"loss": 0.7958, "grad_norm": 0.5850813984870911, "learning_rate": 0.0002, "epoch": 0.8649260366987309, "step": 5350}, {"loss": 0.7642, "grad_norm": 0.517487108707428, "learning_rate": 0.0002, "epoch": 0.8665427208794761, "step": 5360}, {"loss": 0.7801, "grad_norm": 0.5339142680168152, "learning_rate": 0.0002, "epoch": 0.8681594050602215, "step": 5370}, {"loss": 0.818, "grad_norm": 0.6236387491226196, "learning_rate": 0.0002, "epoch": 0.8697760892409668, "step": 5380}, {"loss": 0.7708, "grad_norm": 0.5752192735671997, "learning_rate": 0.0002, "epoch": 0.8713927734217121, "step": 5390}, {"loss": 0.8542, "grad_norm": 0.6724614500999451, "learning_rate": 0.0002, "epoch": 0.8730094576024574, "step": 5400}, {"loss": 0.7581, "grad_norm": 0.5280613303184509, "learning_rate": 0.0002, "epoch": 0.8746261417832026, "step": 5410}, {"loss": 0.8231, "grad_norm": 0.44033288955688477, "learning_rate": 0.0002, "epoch": 0.876242825963948, "step": 5420}, {"loss": 0.8839, "grad_norm": 0.5199708342552185, "learning_rate": 0.0002, "epoch": 0.8778595101446932, "step": 5430}, {"loss": 0.7852, "grad_norm": 0.46778348088264465, "learning_rate": 0.0002, "epoch": 0.8794761943254386, "step": 5440}, {"loss": 0.7834, "grad_norm": 0.4657754898071289, "learning_rate": 0.0002, "epoch": 0.8810928785061838, "step": 5450}, {"loss": 0.7799, "grad_norm": 0.5472902655601501, "learning_rate": 0.0002, "epoch": 0.8827095626869291, "step": 5460}, {"loss": 0.8253, "grad_norm": 0.4876766800880432, "learning_rate": 0.0002, "epoch": 0.8843262468676744, "step": 5470}, {"loss": 0.7906, "grad_norm": 0.5057248473167419, "learning_rate": 0.0002, "epoch": 0.8859429310484197, "step": 5480}, {"loss": 0.8124, "grad_norm": 0.4637320637702942, "learning_rate": 0.0002, "epoch": 0.8875596152291649, "step": 5490}, {"loss": 0.781, "grad_norm": 0.471955806016922, "learning_rate": 0.0002, "epoch": 0.8891762994099103, "step": 5500}, {"loss": 0.8057, "grad_norm": 0.5209813714027405, "learning_rate": 0.0002, "epoch": 0.8907929835906556, "step": 5510}, {"loss": 0.8106, "grad_norm": 0.6213834285736084, "learning_rate": 0.0002, "epoch": 0.8924096677714008, "step": 5520}, {"loss": 0.7787, "grad_norm": 0.5215408205986023, "learning_rate": 0.0002, "epoch": 0.8940263519521462, "step": 5530}, {"loss": 0.8174, "grad_norm": 0.580478310585022, "learning_rate": 0.0002, "epoch": 0.8956430361328914, "step": 5540}, {"loss": 0.8371, "grad_norm": 0.49102169275283813, "learning_rate": 0.0002, "epoch": 0.8972597203136368, "step": 5550}, {"loss": 0.7806, "grad_norm": 0.6043479442596436, "learning_rate": 0.0002, "epoch": 0.898876404494382, "step": 5560}, {"loss": 0.7754, "grad_norm": 0.5636463165283203, "learning_rate": 0.0002, "epoch": 0.9004930886751273, "step": 5570}, {"loss": 0.8145, "grad_norm": 0.5620124340057373, "learning_rate": 0.0002, "epoch": 0.9021097728558726, "step": 5580}, {"loss": 0.8083, "grad_norm": 0.5206354856491089, "learning_rate": 0.0002, "epoch": 0.9037264570366179, "step": 5590}, {"loss": 0.8557, "grad_norm": 0.5798229575157166, "learning_rate": 0.0002, "epoch": 0.9053431412173631, "step": 5600}, {"loss": 0.8097, "grad_norm": 0.6428212523460388, "learning_rate": 0.0002, "epoch": 0.9069598253981085, "step": 5610}, {"loss": 0.7839, "grad_norm": 0.48064687848091125, "learning_rate": 0.0002, "epoch": 0.9085765095788537, "step": 5620}, {"loss": 0.8343, "grad_norm": 0.6347860097885132, "learning_rate": 0.0002, "epoch": 0.9101931937595991, "step": 5630}, {"loss": 0.851, "grad_norm": 0.5353913307189941, "learning_rate": 0.0002, "epoch": 0.9118098779403444, "step": 5640}, {"loss": 0.7736, "grad_norm": 0.5323944091796875, "learning_rate": 0.0002, "epoch": 0.9134265621210896, "step": 5650}, {"loss": 0.8393, "grad_norm": 0.5261843204498291, "learning_rate": 0.0002, "epoch": 0.915043246301835, "step": 5660}, {"loss": 0.7355, "grad_norm": 0.5451326966285706, "learning_rate": 0.0002, "epoch": 0.9166599304825802, "step": 5670}, {"loss": 0.8012, "grad_norm": 0.5183324217796326, "learning_rate": 0.0002, "epoch": 0.9182766146633256, "step": 5680}, {"loss": 0.7659, "grad_norm": 0.47229018807411194, "learning_rate": 0.0002, "epoch": 0.9198932988440708, "step": 5690}, {"loss": 0.7757, "grad_norm": 0.49180513620376587, "learning_rate": 0.0002, "epoch": 0.9215099830248161, "step": 5700}, {"loss": 0.8735, "grad_norm": 0.5419785380363464, "learning_rate": 0.0002, "epoch": 0.9231266672055614, "step": 5710}, {"loss": 0.7378, "grad_norm": 0.5408698916435242, "learning_rate": 0.0002, "epoch": 0.9247433513863067, "step": 5720}, {"loss": 0.7701, "grad_norm": 0.5286232829093933, "learning_rate": 0.0002, "epoch": 0.9263600355670519, "step": 5730}, {"loss": 0.8242, "grad_norm": 0.7539758086204529, "learning_rate": 0.0002, "epoch": 0.9279767197477973, "step": 5740}, {"loss": 0.8118, "grad_norm": 0.5166944861412048, "learning_rate": 0.0002, "epoch": 0.9295934039285425, "step": 5750}, {"loss": 0.783, "grad_norm": 0.6601425409317017, "learning_rate": 0.0002, "epoch": 0.9312100881092878, "step": 5760}, {"loss": 0.7873, "grad_norm": 0.5029960870742798, "learning_rate": 0.0002, "epoch": 0.9328267722900332, "step": 5770}, {"loss": 0.7989, "grad_norm": 0.4926645755767822, "learning_rate": 0.0002, "epoch": 0.9344434564707784, "step": 5780}, {"loss": 0.8174, "grad_norm": 0.5739615559577942, "learning_rate": 0.0002, "epoch": 0.9360601406515238, "step": 5790}, {"loss": 0.8037, "grad_norm": 0.5058279037475586, "learning_rate": 0.0002, "epoch": 0.937676824832269, "step": 5800}, {"loss": 0.8537, "grad_norm": 0.5260962247848511, "learning_rate": 0.0002, "epoch": 0.9392935090130143, "step": 5810}, {"loss": 0.7486, "grad_norm": 0.5768588185310364, "learning_rate": 0.0002, "epoch": 0.9409101931937596, "step": 5820}, {"loss": 0.8215, "grad_norm": 0.5170126557350159, "learning_rate": 0.0002, "epoch": 0.9425268773745049, "step": 5830}, {"loss": 0.7422, "grad_norm": 0.5745864510536194, "learning_rate": 0.0002, "epoch": 0.9441435615552501, "step": 5840}, {"loss": 0.7824, "grad_norm": 0.5551357865333557, "learning_rate": 0.0002, "epoch": 0.9457602457359955, "step": 5850}, {"loss": 0.8529, "grad_norm": 0.5776078701019287, "learning_rate": 0.0002, "epoch": 0.9473769299167407, "step": 5860}, {"loss": 0.8527, "grad_norm": 0.5340062379837036, "learning_rate": 0.0002, "epoch": 0.9489936140974861, "step": 5870}, {"loss": 0.8217, "grad_norm": 0.6447290182113647, "learning_rate": 0.0002, "epoch": 0.9506102982782314, "step": 5880}, {"loss": 0.7945, "grad_norm": 0.5123815536499023, "learning_rate": 0.0002, "epoch": 0.9522269824589766, "step": 5890}, {"loss": 0.8209, "grad_norm": 0.48547613620758057, "learning_rate": 0.0002, "epoch": 0.953843666639722, "step": 5900}, {"loss": 0.7896, "grad_norm": 0.5791414976119995, "learning_rate": 0.0002, "epoch": 0.9554603508204672, "step": 5910}, {"loss": 0.8408, "grad_norm": 0.6195011734962463, "learning_rate": 0.0002, "epoch": 0.9570770350012126, "step": 5920}, {"loss": 0.7805, "grad_norm": 0.6323803067207336, "learning_rate": 0.0002, "epoch": 0.9586937191819578, "step": 5930}, {"loss": 0.8484, "grad_norm": 0.45552879571914673, "learning_rate": 0.0002, "epoch": 0.9603104033627031, "step": 5940}, {"loss": 0.7367, "grad_norm": 0.5796473622322083, "learning_rate": 0.0002, "epoch": 0.9619270875434484, "step": 5950}, {"loss": 0.7672, "grad_norm": 0.647261381149292, "learning_rate": 0.0002, "epoch": 0.9635437717241937, "step": 5960}, {"loss": 0.8086, "grad_norm": 0.5487682819366455, "learning_rate": 0.0002, "epoch": 0.9651604559049389, "step": 5970}, {"loss": 0.7973, "grad_norm": 0.5743663907051086, "learning_rate": 0.0002, "epoch": 0.9667771400856843, "step": 5980}, {"loss": 0.8153, "grad_norm": 0.5470591187477112, "learning_rate": 0.0002, "epoch": 0.9683938242664295, "step": 5990}, {"loss": 0.8119, "grad_norm": 0.5901660323143005, "learning_rate": 0.0002, "epoch": 0.9700105084471748, "step": 6000}, {"loss": 0.8147, "grad_norm": 0.6544759273529053, "learning_rate": 0.0002, "epoch": 0.9716271926279202, "step": 6010}, {"loss": 0.7536, "grad_norm": 0.6288470029830933, "learning_rate": 0.0002, "epoch": 0.9732438768086654, "step": 6020}, {"loss": 0.7989, "grad_norm": 0.673153817653656, "learning_rate": 0.0002, "epoch": 0.9748605609894108, "step": 6030}, {"loss": 0.7556, "grad_norm": 0.42854753136634827, "learning_rate": 0.0002, "epoch": 0.976477245170156, "step": 6040}, {"loss": 0.8006, "grad_norm": 0.5227066278457642, "learning_rate": 0.0002, "epoch": 0.9780939293509013, "step": 6050}, {"loss": 0.795, "grad_norm": 0.5372416973114014, "learning_rate": 0.0002, "epoch": 0.9797106135316466, "step": 6060}, {"loss": 0.7591, "grad_norm": 0.6026402115821838, "learning_rate": 0.0002, "epoch": 0.9813272977123919, "step": 6070}, {"loss": 0.8347, "grad_norm": 0.49547791481018066, "learning_rate": 0.0002, "epoch": 0.9829439818931371, "step": 6080}, {"loss": 0.7722, "grad_norm": 0.4641951322555542, "learning_rate": 0.0002, "epoch": 0.9845606660738825, "step": 6090}, {"loss": 0.8125, "grad_norm": 0.5818535089492798, "learning_rate": 0.0002, "epoch": 0.9861773502546277, "step": 6100}, {"loss": 0.81, "grad_norm": 0.63955157995224, "learning_rate": 0.0002, "epoch": 0.9877940344353731, "step": 6110}, {"loss": 0.7547, "grad_norm": 0.5649438500404358, "learning_rate": 0.0002, "epoch": 0.9894107186161183, "step": 6120}, {"loss": 0.7861, "grad_norm": 0.5290433168411255, "learning_rate": 0.0002, "epoch": 0.9910274027968636, "step": 6130}, {"loss": 0.8109, "grad_norm": 0.6399374008178711, "learning_rate": 0.0002, "epoch": 0.992644086977609, "step": 6140}, {"loss": 0.8373, "grad_norm": 0.6736576557159424, "learning_rate": 0.0002, "epoch": 0.9942607711583542, "step": 6150}, {"loss": 0.7915, "grad_norm": 0.515420138835907, "learning_rate": 0.0002, "epoch": 0.9958774553390995, "step": 6160}, {"loss": 0.8032, "grad_norm": 0.562677800655365, "learning_rate": 0.0002, "epoch": 0.9974941395198448, "step": 6170}, {"loss": 0.8187, "grad_norm": 0.7113858461380005, "learning_rate": 0.0002, "epoch": 0.9991108237005901, "step": 6180}, {"eval_loss": 1.0871200561523438, "eval_runtime": 122.2071, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 0.9999191657909627, "step": 6185}, {"loss": 0.7507, "grad_norm": 0.7111801505088806, "learning_rate": 0.0002, "epoch": 1.0007275078813354, "step": 6190}, {"loss": 0.6865, "grad_norm": 0.5402125716209412, "learning_rate": 0.0002, "epoch": 1.0023441920620806, "step": 6200}, {"loss": 0.7625, "grad_norm": 0.6098830103874207, "learning_rate": 0.0002, "epoch": 1.003960876242826, "step": 6210}, {"loss": 0.7631, "grad_norm": 0.5829983353614807, "learning_rate": 0.0002, "epoch": 1.0055775604235713, "step": 6220}, {"loss": 0.7188, "grad_norm": 0.5614621043205261, "learning_rate": 0.0002, "epoch": 1.0071942446043165, "step": 6230}, {"loss": 0.7505, "grad_norm": 0.5954238772392273, "learning_rate": 0.0002, "epoch": 1.0088109287850617, "step": 6240}, {"loss": 0.7448, "grad_norm": 0.6480574607849121, "learning_rate": 0.0002, "epoch": 1.0104276129658072, "step": 6250}, {"loss": 0.7514, "grad_norm": 0.6051128506660461, "learning_rate": 0.0002, "epoch": 1.0120442971465524, "step": 6260}, {"loss": 0.7237, "grad_norm": 0.6318870782852173, "learning_rate": 0.0002, "epoch": 1.0136609813272976, "step": 6270}, {"loss": 0.7178, "grad_norm": 0.5048980116844177, "learning_rate": 0.0002, "epoch": 1.015277665508043, "step": 6280}, {"loss": 0.7391, "grad_norm": 0.6346936225891113, "learning_rate": 0.0002, "epoch": 1.0168943496887883, "step": 6290}, {"loss": 0.7486, "grad_norm": 0.5711665749549866, "learning_rate": 0.0002, "epoch": 1.0185110338695336, "step": 6300}, {"loss": 0.6808, "grad_norm": 0.5175361037254333, "learning_rate": 0.0002, "epoch": 1.0201277180502788, "step": 6310}, {"loss": 0.7539, "grad_norm": 0.5360831618309021, "learning_rate": 0.0002, "epoch": 1.0217444022310243, "step": 6320}, {"loss": 0.7112, "grad_norm": 0.614675760269165, "learning_rate": 0.0002, "epoch": 1.0233610864117695, "step": 6330}, {"loss": 0.7748, "grad_norm": 0.5626118183135986, "learning_rate": 0.0002, "epoch": 1.0249777705925147, "step": 6340}, {"loss": 0.7375, "grad_norm": 0.574897289276123, "learning_rate": 0.0002, "epoch": 1.02659445477326, "step": 6350}, {"loss": 0.759, "grad_norm": 0.7185447812080383, "learning_rate": 0.0002, "epoch": 1.0282111389540054, "step": 6360}, {"loss": 0.703, "grad_norm": 0.6705799698829651, "learning_rate": 0.0002, "epoch": 1.0298278231347506, "step": 6370}, {"loss": 0.7139, "grad_norm": 0.6740428805351257, "learning_rate": 0.0002, "epoch": 1.0314445073154959, "step": 6380}, {"loss": 0.7252, "grad_norm": 0.663902759552002, "learning_rate": 0.0002, "epoch": 1.0330611914962413, "step": 6390}, {"loss": 0.7065, "grad_norm": 0.5029543042182922, "learning_rate": 0.0002, "epoch": 1.0346778756769865, "step": 6400}, {"loss": 0.711, "grad_norm": 0.7813863158226013, "learning_rate": 0.0002, "epoch": 1.0362945598577318, "step": 6410}, {"loss": 0.7433, "grad_norm": 0.5396282076835632, "learning_rate": 0.0002, "epoch": 1.037911244038477, "step": 6420}, {"loss": 0.7222, "grad_norm": 0.5253293514251709, "learning_rate": 0.0002, "epoch": 1.0395279282192225, "step": 6430}, {"loss": 0.715, "grad_norm": 0.7236770987510681, "learning_rate": 0.0002, "epoch": 1.0411446123999677, "step": 6440}, {"loss": 0.7259, "grad_norm": 0.5670917630195618, "learning_rate": 0.0002, "epoch": 1.042761296580713, "step": 6450}, {"loss": 0.7195, "grad_norm": 0.6031978726387024, "learning_rate": 0.0002, "epoch": 1.0443779807614582, "step": 6460}, {"loss": 0.7648, "grad_norm": 0.5309213399887085, "learning_rate": 0.0002, "epoch": 1.0459946649422036, "step": 6470}, {"loss": 0.7161, "grad_norm": 0.7114651799201965, "learning_rate": 0.0002, "epoch": 1.0476113491229488, "step": 6480}, {"loss": 0.7583, "grad_norm": 0.5591610670089722, "learning_rate": 0.0002, "epoch": 1.049228033303694, "step": 6490}, {"loss": 0.6645, "grad_norm": 0.5185961127281189, "learning_rate": 0.0002, "epoch": 1.0508447174844395, "step": 6500}, {"loss": 0.7654, "grad_norm": 0.6510552167892456, "learning_rate": 0.0002, "epoch": 1.0524614016651848, "step": 6510}, {"loss": 0.7057, "grad_norm": 0.6557928919792175, "learning_rate": 0.0002, "epoch": 1.05407808584593, "step": 6520}, {"loss": 0.8056, "grad_norm": 0.6973192691802979, "learning_rate": 0.0002, "epoch": 1.0556947700266752, "step": 6530}, {"loss": 0.6793, "grad_norm": 0.6226583123207092, "learning_rate": 0.0002, "epoch": 1.0573114542074207, "step": 6540}, {"loss": 0.7151, "grad_norm": 0.5633195638656616, "learning_rate": 0.0002, "epoch": 1.058928138388166, "step": 6550}, {"loss": 0.7082, "grad_norm": 0.7466658353805542, "learning_rate": 0.0002, "epoch": 1.0605448225689111, "step": 6560}, {"loss": 0.7059, "grad_norm": 0.6462772488594055, "learning_rate": 0.0002, "epoch": 1.0621615067496564, "step": 6570}, {"loss": 0.7046, "grad_norm": 0.5266856551170349, "learning_rate": 0.0002, "epoch": 1.0637781909304018, "step": 6580}, {"loss": 0.7157, "grad_norm": 0.534392774105072, "learning_rate": 0.0002, "epoch": 1.065394875111147, "step": 6590}, {"loss": 0.7115, "grad_norm": 0.7514177560806274, "learning_rate": 0.0002, "epoch": 1.0670115592918923, "step": 6600}, {"loss": 0.7545, "grad_norm": 0.7593035697937012, "learning_rate": 0.0002, "epoch": 1.0686282434726375, "step": 6610}, {"loss": 0.6836, "grad_norm": 0.5277858972549438, "learning_rate": 0.0002, "epoch": 1.070244927653383, "step": 6620}, {"loss": 0.7405, "grad_norm": 0.5573670268058777, "learning_rate": 0.0002, "epoch": 1.0718616118341282, "step": 6630}, {"loss": 0.6774, "grad_norm": 0.6802396774291992, "learning_rate": 0.0002, "epoch": 1.0734782960148734, "step": 6640}, {"loss": 0.723, "grad_norm": 0.7367215752601624, "learning_rate": 0.0002, "epoch": 1.0750949801956189, "step": 6650}, {"loss": 0.7429, "grad_norm": 0.5961891412734985, "learning_rate": 0.0002, "epoch": 1.0767116643763641, "step": 6660}, {"loss": 0.6791, "grad_norm": 0.5736313462257385, "learning_rate": 0.0002, "epoch": 1.0783283485571094, "step": 6670}, {"loss": 0.7178, "grad_norm": 0.619219183921814, "learning_rate": 0.0002, "epoch": 1.0799450327378546, "step": 6680}, {"loss": 0.7318, "grad_norm": 0.6214390993118286, "learning_rate": 0.0002, "epoch": 1.0815617169186, "step": 6690}, {"loss": 0.7554, "grad_norm": 0.564536988735199, "learning_rate": 0.0002, "epoch": 1.0831784010993453, "step": 6700}, {"loss": 0.7362, "grad_norm": 0.5838140249252319, "learning_rate": 0.0002, "epoch": 1.0847950852800905, "step": 6710}, {"loss": 0.739, "grad_norm": 0.7000553607940674, "learning_rate": 0.0002, "epoch": 1.0864117694608357, "step": 6720}, {"loss": 0.7369, "grad_norm": 0.7078263759613037, "learning_rate": 0.0002, "epoch": 1.0880284536415812, "step": 6730}, {"loss": 0.7654, "grad_norm": 0.8353848457336426, "learning_rate": 0.0002, "epoch": 1.0896451378223264, "step": 6740}, {"loss": 0.7015, "grad_norm": 0.5615518689155579, "learning_rate": 0.0002, "epoch": 1.0912618220030716, "step": 6750}, {"loss": 0.7396, "grad_norm": 0.5475581288337708, "learning_rate": 0.0002, "epoch": 1.0928785061838169, "step": 6760}, {"loss": 0.7652, "grad_norm": 0.5835978388786316, "learning_rate": 0.0002, "epoch": 1.0944951903645623, "step": 6770}, {"loss": 0.7541, "grad_norm": 0.5516105890274048, "learning_rate": 0.0002, "epoch": 1.0961118745453076, "step": 6780}, {"loss": 0.6842, "grad_norm": 0.5875251889228821, "learning_rate": 0.0002, "epoch": 1.0977285587260528, "step": 6790}, {"loss": 0.6903, "grad_norm": 0.7376947999000549, "learning_rate": 0.0002, "epoch": 1.0993452429067982, "step": 6800}, {"loss": 0.7512, "grad_norm": 0.5656165480613708, "learning_rate": 0.0002, "epoch": 1.1009619270875435, "step": 6810}, {"loss": 0.7409, "grad_norm": 0.6365954279899597, "learning_rate": 0.0002, "epoch": 1.1025786112682887, "step": 6820}, {"loss": 0.7392, "grad_norm": 0.5033080577850342, "learning_rate": 0.0002, "epoch": 1.104195295449034, "step": 6830}, {"loss": 0.6909, "grad_norm": 0.617396891117096, "learning_rate": 0.0002, "epoch": 1.1058119796297794, "step": 6840}, {"loss": 0.7006, "grad_norm": 0.6395374536514282, "learning_rate": 0.0002, "epoch": 1.1074286638105246, "step": 6850}, {"loss": 0.7335, "grad_norm": 0.6775295734405518, "learning_rate": 0.0002, "epoch": 1.1090453479912699, "step": 6860}, {"loss": 0.764, "grad_norm": 0.6655223965644836, "learning_rate": 0.0002, "epoch": 1.1106620321720153, "step": 6870}, {"loss": 0.7553, "grad_norm": 0.676655113697052, "learning_rate": 0.0002, "epoch": 1.1122787163527605, "step": 6880}, {"loss": 0.7342, "grad_norm": 0.6062718629837036, "learning_rate": 0.0002, "epoch": 1.1138954005335058, "step": 6890}, {"loss": 0.7446, "grad_norm": 0.590943455696106, "learning_rate": 0.0002, "epoch": 1.115512084714251, "step": 6900}, {"loss": 0.6705, "grad_norm": 0.6315317153930664, "learning_rate": 0.0002, "epoch": 1.1171287688949965, "step": 6910}, {"loss": 0.6912, "grad_norm": 0.47979024052619934, "learning_rate": 0.0002, "epoch": 1.1187454530757417, "step": 6920}, {"loss": 0.7002, "grad_norm": 0.647298276424408, "learning_rate": 0.0002, "epoch": 1.120362137256487, "step": 6930}, {"loss": 0.7502, "grad_norm": 0.7336484789848328, "learning_rate": 0.0002, "epoch": 1.1219788214372322, "step": 6940}, {"loss": 0.693, "grad_norm": 0.5071424245834351, "learning_rate": 0.0002, "epoch": 1.1235955056179776, "step": 6950}, {"loss": 0.7378, "grad_norm": 0.6527144312858582, "learning_rate": 0.0002, "epoch": 1.1252121897987228, "step": 6960}, {"loss": 0.7228, "grad_norm": 0.6935935020446777, "learning_rate": 0.0002, "epoch": 1.126828873979468, "step": 6970}, {"loss": 0.699, "grad_norm": 0.8026931881904602, "learning_rate": 0.0002, "epoch": 1.1284455581602133, "step": 6980}, {"loss": 0.7361, "grad_norm": 0.5210393667221069, "learning_rate": 0.0002, "epoch": 1.1300622423409588, "step": 6990}, {"loss": 0.7456, "grad_norm": 0.60475093126297, "learning_rate": 0.0002, "epoch": 1.131678926521704, "step": 7000}, {"loss": 0.7495, "grad_norm": 0.6417073607444763, "learning_rate": 0.0002, "epoch": 1.1332956107024492, "step": 7010}, {"loss": 0.7459, "grad_norm": 0.6732175946235657, "learning_rate": 0.0002, "epoch": 1.1349122948831947, "step": 7020}, {"loss": 0.7278, "grad_norm": 0.6719491481781006, "learning_rate": 0.0002, "epoch": 1.13652897906394, "step": 7030}, {"loss": 0.7694, "grad_norm": 0.5708295106887817, "learning_rate": 0.0002, "epoch": 1.1381456632446851, "step": 7040}, {"loss": 0.7823, "grad_norm": 0.7141719460487366, "learning_rate": 0.0002, "epoch": 1.1397623474254304, "step": 7050}, {"loss": 0.764, "grad_norm": 0.6187017560005188, "learning_rate": 0.0002, "epoch": 1.1413790316061758, "step": 7060}, {"loss": 0.7657, "grad_norm": 0.50581294298172, "learning_rate": 0.0002, "epoch": 1.142995715786921, "step": 7070}, {"loss": 0.7357, "grad_norm": 0.5620143413543701, "learning_rate": 0.0002, "epoch": 1.1446123999676663, "step": 7080}, {"loss": 0.7287, "grad_norm": 0.6231929659843445, "learning_rate": 0.0002, "epoch": 1.1462290841484115, "step": 7090}, {"loss": 0.7328, "grad_norm": 0.5775774121284485, "learning_rate": 0.0002, "epoch": 1.147845768329157, "step": 7100}, {"loss": 0.7728, "grad_norm": 0.6492809653282166, "learning_rate": 0.0002, "epoch": 1.1494624525099022, "step": 7110}, {"loss": 0.7545, "grad_norm": 0.6434972286224365, "learning_rate": 0.0002, "epoch": 1.1510791366906474, "step": 7120}, {"loss": 0.7374, "grad_norm": 0.6191812753677368, "learning_rate": 0.0002, "epoch": 1.1526958208713927, "step": 7130}, {"loss": 0.7276, "grad_norm": 0.6690331697463989, "learning_rate": 0.0002, "epoch": 1.1543125050521381, "step": 7140}, {"loss": 0.7704, "grad_norm": 0.5977938175201416, "learning_rate": 0.0002, "epoch": 1.1559291892328833, "step": 7150}, {"loss": 0.7251, "grad_norm": 0.6195854544639587, "learning_rate": 0.0002, "epoch": 1.1575458734136286, "step": 7160}, {"loss": 0.7249, "grad_norm": 0.5752048492431641, "learning_rate": 0.0002, "epoch": 1.159162557594374, "step": 7170}, {"loss": 0.7593, "grad_norm": 0.589081883430481, "learning_rate": 0.0002, "epoch": 1.1607792417751193, "step": 7180}, {"loss": 0.704, "grad_norm": 0.756996750831604, "learning_rate": 0.0002, "epoch": 1.1623959259558645, "step": 7190}, {"loss": 0.7404, "grad_norm": 0.7614967226982117, "learning_rate": 0.0002, "epoch": 1.1640126101366097, "step": 7200}, {"loss": 0.7867, "grad_norm": 0.6120437979698181, "learning_rate": 0.0002, "epoch": 1.1656292943173552, "step": 7210}, {"loss": 0.7384, "grad_norm": 0.6210004687309265, "learning_rate": 0.0002, "epoch": 1.1672459784981004, "step": 7220}, {"loss": 0.7251, "grad_norm": 0.6044116020202637, "learning_rate": 0.0002, "epoch": 1.1688626626788456, "step": 7230}, {"loss": 0.7361, "grad_norm": 0.5418457388877869, "learning_rate": 0.0002, "epoch": 1.170479346859591, "step": 7240}, {"loss": 0.6938, "grad_norm": 0.6413537263870239, "learning_rate": 0.0002, "epoch": 1.1720960310403363, "step": 7250}, {"loss": 0.6978, "grad_norm": 0.5777867436408997, "learning_rate": 0.0002, "epoch": 1.1737127152210816, "step": 7260}, {"loss": 0.7503, "grad_norm": 0.7092402577400208, "learning_rate": 0.0002, "epoch": 1.1753293994018268, "step": 7270}, {"loss": 0.7487, "grad_norm": 0.6351709365844727, "learning_rate": 0.0002, "epoch": 1.176946083582572, "step": 7280}, {"loss": 0.7527, "grad_norm": 0.6172189712524414, "learning_rate": 0.0002, "epoch": 1.1785627677633175, "step": 7290}, {"loss": 0.7319, "grad_norm": 0.6801714897155762, "learning_rate": 0.0002, "epoch": 1.1801794519440627, "step": 7300}, {"loss": 0.6941, "grad_norm": 0.6044712066650391, "learning_rate": 0.0002, "epoch": 1.181796136124808, "step": 7310}, {"loss": 0.6951, "grad_norm": 0.7413212060928345, "learning_rate": 0.0002, "epoch": 1.1834128203055534, "step": 7320}, {"loss": 0.7396, "grad_norm": 0.5303856134414673, "learning_rate": 0.0002, "epoch": 1.1850295044862986, "step": 7330}, {"loss": 0.6915, "grad_norm": 0.5647098422050476, "learning_rate": 0.0002, "epoch": 1.1866461886670439, "step": 7340}, {"loss": 0.7506, "grad_norm": 0.7374135255813599, "learning_rate": 0.0002, "epoch": 1.188262872847789, "step": 7350}, {"loss": 0.7041, "grad_norm": 0.5710089206695557, "learning_rate": 0.0002, "epoch": 1.1898795570285345, "step": 7360}, {"loss": 0.8289, "grad_norm": 0.6073619723320007, "learning_rate": 0.0002, "epoch": 1.1914962412092798, "step": 7370}, {"loss": 0.7722, "grad_norm": 0.5899916887283325, "learning_rate": 0.0002, "epoch": 1.193112925390025, "step": 7380}, {"loss": 0.756, "grad_norm": 0.7762434482574463, "learning_rate": 0.0002, "epoch": 1.1947296095707705, "step": 7390}, {"loss": 0.7319, "grad_norm": 0.679949939250946, "learning_rate": 0.0002, "epoch": 1.1963462937515157, "step": 7400}, {"loss": 0.7599, "grad_norm": 0.6106849312782288, "learning_rate": 0.0002, "epoch": 1.197962977932261, "step": 7410}, {"loss": 0.7648, "grad_norm": 0.682461678981781, "learning_rate": 0.0002, "epoch": 1.1995796621130062, "step": 7420}, {"loss": 0.7741, "grad_norm": 0.6087017059326172, "learning_rate": 0.0002, "epoch": 1.2011963462937516, "step": 7430}, {"loss": 0.7642, "grad_norm": 0.63739013671875, "learning_rate": 0.0002, "epoch": 1.2028130304744968, "step": 7440}, {"loss": 0.7611, "grad_norm": 0.6154777407646179, "learning_rate": 0.0002, "epoch": 1.204429714655242, "step": 7450}, {"loss": 0.7565, "grad_norm": 0.7491534948348999, "learning_rate": 0.0002, "epoch": 1.2060463988359873, "step": 7460}, {"loss": 0.698, "grad_norm": 0.6664797067642212, "learning_rate": 0.0002, "epoch": 1.2076630830167328, "step": 7470}, {"loss": 0.7456, "grad_norm": 0.6660266518592834, "learning_rate": 0.0002, "epoch": 1.209279767197478, "step": 7480}, {"loss": 0.714, "grad_norm": 0.6972551345825195, "learning_rate": 0.0002, "epoch": 1.2108964513782232, "step": 7490}, {"loss": 0.7023, "grad_norm": 0.6157945990562439, "learning_rate": 0.0002, "epoch": 1.2125131355589684, "step": 7500}, {"loss": 0.7326, "grad_norm": 0.5199310183525085, "learning_rate": 0.0002, "epoch": 1.214129819739714, "step": 7510}, {"loss": 0.7586, "grad_norm": 0.577610433101654, "learning_rate": 0.0002, "epoch": 1.2157465039204591, "step": 7520}, {"loss": 0.7179, "grad_norm": 0.53652423620224, "learning_rate": 0.0002, "epoch": 1.2173631881012044, "step": 7530}, {"loss": 0.7393, "grad_norm": 0.6479050517082214, "learning_rate": 0.0002, "epoch": 1.2189798722819498, "step": 7540}, {"loss": 0.7534, "grad_norm": 0.618748128414154, "learning_rate": 0.0002, "epoch": 1.220596556462695, "step": 7550}, {"loss": 0.6886, "grad_norm": 0.6311424374580383, "learning_rate": 0.0002, "epoch": 1.2222132406434403, "step": 7560}, {"loss": 0.7272, "grad_norm": 0.6595825552940369, "learning_rate": 0.0002, "epoch": 1.2238299248241855, "step": 7570}, {"loss": 0.7353, "grad_norm": 0.5198960900306702, "learning_rate": 0.0002, "epoch": 1.225446609004931, "step": 7580}, {"loss": 0.674, "grad_norm": 0.578650712966919, "learning_rate": 0.0002, "epoch": 1.2270632931856762, "step": 7590}, {"loss": 0.7507, "grad_norm": 0.6080220937728882, "learning_rate": 0.0002, "epoch": 1.2286799773664214, "step": 7600}, {"loss": 0.7733, "grad_norm": 0.7050248384475708, "learning_rate": 0.0002, "epoch": 1.2302966615471669, "step": 7610}, {"loss": 0.7032, "grad_norm": 0.6652196049690247, "learning_rate": 0.0002, "epoch": 1.2319133457279121, "step": 7620}, {"loss": 0.7085, "grad_norm": 0.7322776317596436, "learning_rate": 0.0002, "epoch": 1.2335300299086573, "step": 7630}, {"loss": 0.7402, "grad_norm": 0.4998728036880493, "learning_rate": 0.0002, "epoch": 1.2351467140894026, "step": 7640}, {"loss": 0.7214, "grad_norm": 0.6428788900375366, "learning_rate": 0.0002, "epoch": 1.2367633982701478, "step": 7650}, {"loss": 0.7699, "grad_norm": 0.585242509841919, "learning_rate": 0.0002, "epoch": 1.2383800824508933, "step": 7660}, {"loss": 0.7621, "grad_norm": 0.5211917757987976, "learning_rate": 0.0002, "epoch": 1.2399967666316385, "step": 7670}, {"loss": 0.746, "grad_norm": 0.6490384340286255, "learning_rate": 0.0002, "epoch": 1.2416134508123837, "step": 7680}, {"loss": 0.7186, "grad_norm": 0.6249763369560242, "learning_rate": 0.0002, "epoch": 1.2432301349931292, "step": 7690}, {"loss": 0.7761, "grad_norm": 0.71870356798172, "learning_rate": 0.0002, "epoch": 1.2448468191738744, "step": 7700}, {"loss": 0.7525, "grad_norm": 0.6761967539787292, "learning_rate": 0.0002, "epoch": 1.2464635033546196, "step": 7710}, {"loss": 0.7501, "grad_norm": 0.6500617265701294, "learning_rate": 0.0002, "epoch": 1.2480801875353649, "step": 7720}, {"loss": 0.7903, "grad_norm": 0.8069869875907898, "learning_rate": 0.0002, "epoch": 1.2496968717161103, "step": 7730}, {"loss": 0.6747, "grad_norm": 0.6044608950614929, "learning_rate": 0.0002, "epoch": 1.2513135558968556, "step": 7740}, {"loss": 0.6825, "grad_norm": 0.6573283076286316, "learning_rate": 0.0002, "epoch": 1.2529302400776008, "step": 7750}, {"loss": 0.7617, "grad_norm": 0.625430166721344, "learning_rate": 0.0002, "epoch": 1.2545469242583462, "step": 7760}, {"loss": 0.7041, "grad_norm": 0.5442022681236267, "learning_rate": 0.0002, "epoch": 1.2561636084390915, "step": 7770}, {"loss": 0.7172, "grad_norm": 0.6818386912345886, "learning_rate": 0.0002, "epoch": 1.2577802926198367, "step": 7780}, {"loss": 0.696, "grad_norm": 0.6381874084472656, "learning_rate": 0.0002, "epoch": 1.259396976800582, "step": 7790}, {"loss": 0.6834, "grad_norm": 0.6269212961196899, "learning_rate": 0.0002, "epoch": 1.2610136609813272, "step": 7800}, {"loss": 0.7821, "grad_norm": 0.600121259689331, "learning_rate": 0.0002, "epoch": 1.2626303451620726, "step": 7810}, {"loss": 0.7761, "grad_norm": 0.6337703466415405, "learning_rate": 0.0002, "epoch": 1.2642470293428179, "step": 7820}, {"loss": 0.732, "grad_norm": 0.7234963774681091, "learning_rate": 0.0002, "epoch": 1.2658637135235633, "step": 7830}, {"loss": 0.785, "grad_norm": 0.800184965133667, "learning_rate": 0.0002, "epoch": 1.2674803977043085, "step": 7840}, {"loss": 0.7426, "grad_norm": 0.7539464831352234, "learning_rate": 0.0002, "epoch": 1.2690970818850538, "step": 7850}, {"loss": 0.7496, "grad_norm": 0.5493760704994202, "learning_rate": 0.0002, "epoch": 1.270713766065799, "step": 7860}, {"loss": 0.7537, "grad_norm": 0.7477145791053772, "learning_rate": 0.0002, "epoch": 1.2723304502465442, "step": 7870}, {"loss": 0.7573, "grad_norm": 0.6366362571716309, "learning_rate": 0.0002, "epoch": 1.2739471344272897, "step": 7880}, {"loss": 0.7608, "grad_norm": 0.7419533729553223, "learning_rate": 0.0002, "epoch": 1.275563818608035, "step": 7890}, {"loss": 0.7873, "grad_norm": 0.6141223311424255, "learning_rate": 0.0002, "epoch": 1.2771805027887801, "step": 7900}, {"loss": 0.6916, "grad_norm": 0.7522598505020142, "learning_rate": 0.0002, "epoch": 1.2787971869695256, "step": 7910}, {"loss": 0.7097, "grad_norm": 0.6935804486274719, "learning_rate": 0.0002, "epoch": 1.2804138711502708, "step": 7920}, {"loss": 0.7185, "grad_norm": 0.7239290475845337, "learning_rate": 0.0002, "epoch": 1.282030555331016, "step": 7930}, {"loss": 0.7145, "grad_norm": 0.8800187110900879, "learning_rate": 0.0002, "epoch": 1.2836472395117613, "step": 7940}, {"loss": 0.6991, "grad_norm": 0.540458083152771, "learning_rate": 0.0002, "epoch": 1.2852639236925067, "step": 7950}, {"loss": 0.7139, "grad_norm": 0.6492934226989746, "learning_rate": 0.0002, "epoch": 1.286880607873252, "step": 7960}, {"loss": 0.7742, "grad_norm": 0.6543959379196167, "learning_rate": 0.0002, "epoch": 1.2884972920539972, "step": 7970}, {"loss": 0.7316, "grad_norm": 0.5804705619812012, "learning_rate": 0.0002, "epoch": 1.2901139762347427, "step": 7980}, {"loss": 0.796, "grad_norm": 0.7074727416038513, "learning_rate": 0.0002, "epoch": 1.291730660415488, "step": 7990}, {"loss": 0.7034, "grad_norm": 0.5347974300384521, "learning_rate": 0.0002, "epoch": 1.2933473445962331, "step": 8000}, {"loss": 0.738, "grad_norm": 0.6457298398017883, "learning_rate": 0.0002, "epoch": 1.2949640287769784, "step": 8010}, {"loss": 0.7634, "grad_norm": 0.6407219171524048, "learning_rate": 0.0002, "epoch": 1.2965807129577236, "step": 8020}, {"loss": 0.7506, "grad_norm": 0.828439474105835, "learning_rate": 0.0002, "epoch": 1.298197397138469, "step": 8030}, {"loss": 0.735, "grad_norm": 0.4840380549430847, "learning_rate": 0.0002, "epoch": 1.2998140813192143, "step": 8040}, {"loss": 0.7283, "grad_norm": 0.5921024680137634, "learning_rate": 0.0002, "epoch": 1.3014307654999595, "step": 8050}, {"loss": 0.7477, "grad_norm": 0.6170315146446228, "learning_rate": 0.0002, "epoch": 1.303047449680705, "step": 8060}, {"loss": 0.7534, "grad_norm": 0.5374847054481506, "learning_rate": 0.0002, "epoch": 1.3046641338614502, "step": 8070}, {"loss": 0.7593, "grad_norm": 0.545758068561554, "learning_rate": 0.0002, "epoch": 1.3062808180421954, "step": 8080}, {"loss": 0.7463, "grad_norm": 0.55641770362854, "learning_rate": 0.0002, "epoch": 1.3078975022229407, "step": 8090}, {"loss": 0.7594, "grad_norm": 0.6724897027015686, "learning_rate": 0.0002, "epoch": 1.309514186403686, "step": 8100}, {"loss": 0.7105, "grad_norm": 0.6923972368240356, "learning_rate": 0.0002, "epoch": 1.3111308705844313, "step": 8110}, {"loss": 0.7149, "grad_norm": 0.5136841535568237, "learning_rate": 0.0002, "epoch": 1.3127475547651766, "step": 8120}, {"loss": 0.7504, "grad_norm": 0.6766283512115479, "learning_rate": 0.0002, "epoch": 1.314364238945922, "step": 8130}, {"loss": 0.7489, "grad_norm": 0.6283926367759705, "learning_rate": 0.0002, "epoch": 1.3159809231266673, "step": 8140}, {"loss": 0.7459, "grad_norm": 0.644216001033783, "learning_rate": 0.0002, "epoch": 1.3175976073074125, "step": 8150}, {"loss": 0.7125, "grad_norm": 0.7827503085136414, "learning_rate": 0.0002, "epoch": 1.3192142914881577, "step": 8160}, {"loss": 0.7271, "grad_norm": 0.6651390790939331, "learning_rate": 0.0002, "epoch": 1.320830975668903, "step": 8170}, {"loss": 0.7778, "grad_norm": 0.5547412633895874, "learning_rate": 0.0002, "epoch": 1.3224476598496484, "step": 8180}, {"loss": 0.7402, "grad_norm": 0.6765179634094238, "learning_rate": 0.0002, "epoch": 1.3240643440303936, "step": 8190}, {"loss": 0.7106, "grad_norm": 0.6822077035903931, "learning_rate": 0.0002, "epoch": 1.325681028211139, "step": 8200}, {"loss": 0.7288, "grad_norm": 0.5941002368927002, "learning_rate": 0.0002, "epoch": 1.3272977123918843, "step": 8210}, {"loss": 0.7494, "grad_norm": 0.4850037097930908, "learning_rate": 0.0002, "epoch": 1.3289143965726296, "step": 8220}, {"loss": 0.7474, "grad_norm": 0.6162990927696228, "learning_rate": 0.0002, "epoch": 1.3305310807533748, "step": 8230}, {"loss": 0.7751, "grad_norm": 0.6665613651275635, "learning_rate": 0.0002, "epoch": 1.33214776493412, "step": 8240}, {"loss": 0.759, "grad_norm": 0.618192732334137, "learning_rate": 0.0002, "epoch": 1.3337644491148655, "step": 8250}, {"loss": 0.7532, "grad_norm": 0.710418701171875, "learning_rate": 0.0002, "epoch": 1.3353811332956107, "step": 8260}, {"loss": 0.7306, "grad_norm": 0.5109876990318298, "learning_rate": 0.0002, "epoch": 1.336997817476356, "step": 8270}, {"loss": 0.7303, "grad_norm": 0.6791711449623108, "learning_rate": 0.0002, "epoch": 1.3386145016571014, "step": 8280}, {"loss": 0.7594, "grad_norm": 0.6836432814598083, "learning_rate": 0.0002, "epoch": 1.3402311858378466, "step": 8290}, {"loss": 0.7594, "grad_norm": 0.5579386353492737, "learning_rate": 0.0002, "epoch": 1.3418478700185918, "step": 8300}, {"loss": 0.7377, "grad_norm": 0.6713546514511108, "learning_rate": 0.0002, "epoch": 1.343464554199337, "step": 8310}, {"loss": 0.7756, "grad_norm": 0.5353720188140869, "learning_rate": 0.0002, "epoch": 1.3450812383800825, "step": 8320}, {"loss": 0.718, "grad_norm": 0.5813682675361633, "learning_rate": 0.0002, "epoch": 1.3466979225608278, "step": 8330}, {"loss": 0.7294, "grad_norm": 0.8158791661262512, "learning_rate": 0.0002, "epoch": 1.348314606741573, "step": 8340}, {"loss": 0.6992, "grad_norm": 0.6193785071372986, "learning_rate": 0.0002, "epoch": 1.3499312909223184, "step": 8350}, {"loss": 0.7654, "grad_norm": 0.6353939771652222, "learning_rate": 0.0002, "epoch": 1.3515479751030637, "step": 8360}, {"loss": 0.7519, "grad_norm": 0.6925048232078552, "learning_rate": 0.0002, "epoch": 1.353164659283809, "step": 8370}, {"loss": 0.736, "grad_norm": 0.988264799118042, "learning_rate": 0.0002, "epoch": 1.3547813434645541, "step": 8380}, {"loss": 0.7744, "grad_norm": 0.6476002931594849, "learning_rate": 0.0002, "epoch": 1.3563980276452994, "step": 8390}, {"loss": 0.776, "grad_norm": 0.7120398879051208, "learning_rate": 0.0002, "epoch": 1.3580147118260448, "step": 8400}, {"loss": 0.7368, "grad_norm": 0.9048416614532471, "learning_rate": 0.0002, "epoch": 1.35963139600679, "step": 8410}, {"loss": 0.7544, "grad_norm": 0.7000672817230225, "learning_rate": 0.0002, "epoch": 1.3612480801875353, "step": 8420}, {"loss": 0.7358, "grad_norm": 0.6015632152557373, "learning_rate": 0.0002, "epoch": 1.3628647643682807, "step": 8430}, {"loss": 0.7298, "grad_norm": 0.612516462802887, "learning_rate": 0.0002, "epoch": 1.364481448549026, "step": 8440}, {"loss": 0.7055, "grad_norm": 0.5969301462173462, "learning_rate": 0.0002, "epoch": 1.3660981327297712, "step": 8450}, {"loss": 0.7754, "grad_norm": 0.6730654239654541, "learning_rate": 0.0002, "epoch": 1.3677148169105164, "step": 8460}, {"loss": 0.7465, "grad_norm": 0.6386392116546631, "learning_rate": 0.0002, "epoch": 1.369331501091262, "step": 8470}, {"loss": 0.7433, "grad_norm": 0.739544153213501, "learning_rate": 0.0002, "epoch": 1.3709481852720071, "step": 8480}, {"loss": 0.7892, "grad_norm": 0.6462782621383667, "learning_rate": 0.0002, "epoch": 1.3725648694527524, "step": 8490}, {"loss": 0.7302, "grad_norm": 0.7346843481063843, "learning_rate": 0.0002, "epoch": 1.3741815536334978, "step": 8500}, {"loss": 0.7634, "grad_norm": 0.6884821057319641, "learning_rate": 0.0002, "epoch": 1.375798237814243, "step": 8510}, {"loss": 0.7614, "grad_norm": 0.6999333500862122, "learning_rate": 0.0002, "epoch": 1.3774149219949883, "step": 8520}, {"loss": 0.729, "grad_norm": 0.5378713011741638, "learning_rate": 0.0002, "epoch": 1.3790316061757335, "step": 8530}, {"loss": 0.6797, "grad_norm": 0.5417906641960144, "learning_rate": 0.0002, "epoch": 1.3806482903564787, "step": 8540}, {"loss": 0.7499, "grad_norm": 0.6602526307106018, "learning_rate": 0.0002, "epoch": 1.3822649745372242, "step": 8550}, {"loss": 0.7356, "grad_norm": 0.7073674201965332, "learning_rate": 0.0002, "epoch": 1.3838816587179694, "step": 8560}, {"loss": 0.75, "grad_norm": 0.5841707587242126, "learning_rate": 0.0002, "epoch": 1.3854983428987149, "step": 8570}, {"loss": 0.732, "grad_norm": 0.7031095027923584, "learning_rate": 0.0002, "epoch": 1.38711502707946, "step": 8580}, {"loss": 0.7464, "grad_norm": 0.5198570489883423, "learning_rate": 0.0002, "epoch": 1.3887317112602053, "step": 8590}, {"loss": 0.7354, "grad_norm": 0.7261320352554321, "learning_rate": 0.0002, "epoch": 1.3903483954409506, "step": 8600}, {"loss": 0.7339, "grad_norm": 0.5616350173950195, "learning_rate": 0.0002, "epoch": 1.3919650796216958, "step": 8610}, {"loss": 0.7382, "grad_norm": 0.5185914635658264, "learning_rate": 0.0002, "epoch": 1.3935817638024413, "step": 8620}, {"loss": 0.7456, "grad_norm": 0.5814694762229919, "learning_rate": 0.0002, "epoch": 1.3951984479831865, "step": 8630}, {"loss": 0.7413, "grad_norm": 0.6977371573448181, "learning_rate": 0.0002, "epoch": 1.3968151321639317, "step": 8640}, {"loss": 0.7574, "grad_norm": 0.6855689883232117, "learning_rate": 0.0002, "epoch": 1.3984318163446772, "step": 8650}, {"loss": 0.7802, "grad_norm": 0.5414357781410217, "learning_rate": 0.0002, "epoch": 1.4000485005254224, "step": 8660}, {"loss": 0.7487, "grad_norm": 0.6970012784004211, "learning_rate": 0.0002, "epoch": 1.4016651847061676, "step": 8670}, {"loss": 0.7421, "grad_norm": 0.526079535484314, "learning_rate": 0.0002, "epoch": 1.4032818688869129, "step": 8680}, {"loss": 0.737, "grad_norm": 0.758712887763977, "learning_rate": 0.0002, "epoch": 1.404898553067658, "step": 8690}, {"loss": 0.7612, "grad_norm": 0.7118762731552124, "learning_rate": 0.0002, "epoch": 1.4065152372484035, "step": 8700}, {"loss": 0.7628, "grad_norm": 0.5696909427642822, "learning_rate": 0.0002, "epoch": 1.4081319214291488, "step": 8710}, {"loss": 0.7156, "grad_norm": 0.7995436787605286, "learning_rate": 0.0002, "epoch": 1.4097486056098942, "step": 8720}, {"loss": 0.7521, "grad_norm": 0.7237521409988403, "learning_rate": 0.0002, "epoch": 1.4113652897906395, "step": 8730}, {"loss": 0.7661, "grad_norm": 0.744628369808197, "learning_rate": 0.0002, "epoch": 1.4129819739713847, "step": 8740}, {"loss": 0.7073, "grad_norm": 0.6082926988601685, "learning_rate": 0.0002, "epoch": 1.41459865815213, "step": 8750}, {"loss": 0.7282, "grad_norm": 0.5185243487358093, "learning_rate": 0.0002, "epoch": 1.4162153423328752, "step": 8760}, {"loss": 0.7592, "grad_norm": 0.5183082222938538, "learning_rate": 0.0002, "epoch": 1.4178320265136206, "step": 8770}, {"loss": 0.7509, "grad_norm": 0.7326041460037231, "learning_rate": 0.0002, "epoch": 1.4194487106943658, "step": 8780}, {"loss": 0.7398, "grad_norm": 0.7174660563468933, "learning_rate": 0.0002, "epoch": 1.421065394875111, "step": 8790}, {"loss": 0.7507, "grad_norm": 0.8080165982246399, "learning_rate": 0.0002, "epoch": 1.4226820790558565, "step": 8800}, {"loss": 0.72, "grad_norm": 0.5061507821083069, "learning_rate": 0.0002, "epoch": 1.4242987632366018, "step": 8810}, {"loss": 0.7563, "grad_norm": 0.801602840423584, "learning_rate": 0.0002, "epoch": 1.425915447417347, "step": 8820}, {"loss": 0.7287, "grad_norm": 0.6150273084640503, "learning_rate": 0.0002, "epoch": 1.4275321315980922, "step": 8830}, {"loss": 0.7452, "grad_norm": 0.8786525726318359, "learning_rate": 0.0002, "epoch": 1.4291488157788377, "step": 8840}, {"loss": 0.7257, "grad_norm": 0.6371538639068604, "learning_rate": 0.0002, "epoch": 1.430765499959583, "step": 8850}, {"loss": 0.711, "grad_norm": 0.6409295797348022, "learning_rate": 0.0002, "epoch": 1.4323821841403281, "step": 8860}, {"loss": 0.7891, "grad_norm": 0.6452359557151794, "learning_rate": 0.0002, "epoch": 1.4339988683210736, "step": 8870}, {"loss": 0.7588, "grad_norm": 0.5842334628105164, "learning_rate": 0.0002, "epoch": 1.4356155525018188, "step": 8880}, {"loss": 0.7446, "grad_norm": 0.696761965751648, "learning_rate": 0.0002, "epoch": 1.437232236682564, "step": 8890}, {"loss": 0.7541, "grad_norm": 0.6384600400924683, "learning_rate": 0.0002, "epoch": 1.4388489208633093, "step": 8900}, {"loss": 0.7049, "grad_norm": 0.5981136560440063, "learning_rate": 0.0002, "epoch": 1.4404656050440545, "step": 8910}, {"loss": 0.795, "grad_norm": 0.6355637907981873, "learning_rate": 0.0002, "epoch": 1.4420822892248, "step": 8920}, {"loss": 0.7653, "grad_norm": 0.6374830603599548, "learning_rate": 0.0002, "epoch": 1.4436989734055452, "step": 8930}, {"loss": 0.8108, "grad_norm": 0.559013307094574, "learning_rate": 0.0002, "epoch": 1.4453156575862904, "step": 8940}, {"loss": 0.7045, "grad_norm": 0.7289170026779175, "learning_rate": 0.0002, "epoch": 1.446932341767036, "step": 8950}, {"loss": 0.7484, "grad_norm": 0.8649206757545471, "learning_rate": 0.0002, "epoch": 1.4485490259477811, "step": 8960}, {"loss": 0.7745, "grad_norm": 0.7664689421653748, "learning_rate": 0.0002, "epoch": 1.4501657101285264, "step": 8970}, {"loss": 0.7431, "grad_norm": 0.7109952569007874, "learning_rate": 0.0002, "epoch": 1.4517823943092716, "step": 8980}, {"loss": 0.7997, "grad_norm": 0.6312844753265381, "learning_rate": 0.0002, "epoch": 1.453399078490017, "step": 8990}, {"loss": 0.7467, "grad_norm": 0.6616617441177368, "learning_rate": 0.0002, "epoch": 1.4550157626707623, "step": 9000}, {"loss": 0.7518, "grad_norm": 0.7384068965911865, "learning_rate": 0.0002, "epoch": 1.4566324468515075, "step": 9010}, {"loss": 0.7483, "grad_norm": 0.6549670100212097, "learning_rate": 0.0002, "epoch": 1.458249131032253, "step": 9020}, {"loss": 0.7423, "grad_norm": 0.6254119277000427, "learning_rate": 0.0002, "epoch": 1.4598658152129982, "step": 9030}, {"loss": 0.7645, "grad_norm": 0.6806328892707825, "learning_rate": 0.0002, "epoch": 1.4614824993937434, "step": 9040}, {"loss": 0.7221, "grad_norm": 0.6803115010261536, "learning_rate": 0.0002, "epoch": 1.4630991835744886, "step": 9050}, {"loss": 0.7264, "grad_norm": 0.48529282212257385, "learning_rate": 0.0002, "epoch": 1.4647158677552339, "step": 9060}, {"loss": 0.7542, "grad_norm": 0.5995030999183655, "learning_rate": 0.0002, "epoch": 1.4663325519359793, "step": 9070}, {"loss": 0.7894, "grad_norm": 0.6005427837371826, "learning_rate": 0.0002, "epoch": 1.4679492361167246, "step": 9080}, {"loss": 0.7288, "grad_norm": 0.718564510345459, "learning_rate": 0.0002, "epoch": 1.46956592029747, "step": 9090}, {"loss": 0.7089, "grad_norm": 0.7003577351570129, "learning_rate": 0.0002, "epoch": 1.4711826044782153, "step": 9100}, {"loss": 0.8069, "grad_norm": 0.5888323783874512, "learning_rate": 0.0002, "epoch": 1.4727992886589605, "step": 9110}, {"loss": 0.7275, "grad_norm": 0.6417609453201294, "learning_rate": 0.0002, "epoch": 1.4744159728397057, "step": 9120}, {"loss": 0.7441, "grad_norm": 0.572294294834137, "learning_rate": 0.0002, "epoch": 1.476032657020451, "step": 9130}, {"loss": 0.8053, "grad_norm": 0.8200714588165283, "learning_rate": 0.0002, "epoch": 1.4776493412011964, "step": 9140}, {"loss": 0.7382, "grad_norm": 0.6343288421630859, "learning_rate": 0.0002, "epoch": 1.4792660253819416, "step": 9150}, {"loss": 0.7641, "grad_norm": 0.7017961144447327, "learning_rate": 0.0002, "epoch": 1.4808827095626869, "step": 9160}, {"loss": 0.7619, "grad_norm": 0.6202912926673889, "learning_rate": 0.0002, "epoch": 1.4824993937434323, "step": 9170}, {"loss": 0.7428, "grad_norm": 0.6677869558334351, "learning_rate": 0.0002, "epoch": 1.4841160779241775, "step": 9180}, {"loss": 0.7648, "grad_norm": 0.6052267551422119, "learning_rate": 0.0002, "epoch": 1.4857327621049228, "step": 9190}, {"loss": 0.7152, "grad_norm": 0.6638872027397156, "learning_rate": 0.0002, "epoch": 1.487349446285668, "step": 9200}, {"loss": 0.7448, "grad_norm": 0.6245523691177368, "learning_rate": 0.0002, "epoch": 1.4889661304664135, "step": 9210}, {"loss": 0.6958, "grad_norm": 0.5761767625808716, "learning_rate": 0.0002, "epoch": 1.4905828146471587, "step": 9220}, {"loss": 0.8012, "grad_norm": 0.8175981640815735, "learning_rate": 0.0002, "epoch": 1.492199498827904, "step": 9230}, {"loss": 0.683, "grad_norm": 0.9144009947776794, "learning_rate": 0.0002, "epoch": 1.4938161830086494, "step": 9240}, {"loss": 0.7623, "grad_norm": 0.5742552876472473, "learning_rate": 0.0002, "epoch": 1.4954328671893946, "step": 9250}, {"loss": 0.7418, "grad_norm": 0.534534215927124, "learning_rate": 0.0002, "epoch": 1.4970495513701398, "step": 9260}, {"loss": 0.7194, "grad_norm": 0.7836225032806396, "learning_rate": 0.0002, "epoch": 1.498666235550885, "step": 9270}, {"loss": 0.7453, "grad_norm": 0.5292993187904358, "learning_rate": 0.0002, "epoch": 1.5002829197316303, "step": 9280}, {"loss": 0.7168, "grad_norm": 0.8044071793556213, "learning_rate": 0.0002, "epoch": 1.5018996039123758, "step": 9290}, {"loss": 0.7229, "grad_norm": 0.6185805201530457, "learning_rate": 0.0002, "epoch": 1.503516288093121, "step": 9300}, {"loss": 0.684, "grad_norm": 0.6093607544898987, "learning_rate": 0.0002, "epoch": 1.5051329722738664, "step": 9310}, {"loss": 0.7973, "grad_norm": 0.5891730189323425, "learning_rate": 0.0002, "epoch": 1.5067496564546117, "step": 9320}, {"loss": 0.7474, "grad_norm": 0.6331129670143127, "learning_rate": 0.0002, "epoch": 1.508366340635357, "step": 9330}, {"loss": 0.7074, "grad_norm": 0.7690958380699158, "learning_rate": 0.0002, "epoch": 1.5099830248161021, "step": 9340}, {"loss": 0.672, "grad_norm": 0.6548877358436584, "learning_rate": 0.0002, "epoch": 1.5115997089968474, "step": 9350}, {"loss": 0.7408, "grad_norm": 0.6545143127441406, "learning_rate": 0.0002, "epoch": 1.5132163931775926, "step": 9360}, {"loss": 0.7432, "grad_norm": 0.553247332572937, "learning_rate": 0.0002, "epoch": 1.514833077358338, "step": 9370}, {"loss": 0.7265, "grad_norm": 0.8145074844360352, "learning_rate": 0.0002, "epoch": 1.5164497615390833, "step": 9380}, {"loss": 0.7379, "grad_norm": 0.7636994123458862, "learning_rate": 0.0002, "epoch": 1.5180664457198287, "step": 9390}, {"loss": 0.7413, "grad_norm": 0.6838982701301575, "learning_rate": 0.0002, "epoch": 1.519683129900574, "step": 9400}, {"loss": 0.7367, "grad_norm": 0.8599441647529602, "learning_rate": 0.0002, "epoch": 1.5212998140813192, "step": 9410}, {"loss": 0.7663, "grad_norm": 0.7020329833030701, "learning_rate": 0.0002, "epoch": 1.5229164982620644, "step": 9420}, {"loss": 0.7928, "grad_norm": 0.6964772343635559, "learning_rate": 0.0002, "epoch": 1.5245331824428097, "step": 9430}, {"loss": 0.7168, "grad_norm": 0.6916600465774536, "learning_rate": 0.0002, "epoch": 1.5261498666235551, "step": 9440}, {"loss": 0.7519, "grad_norm": 0.7282621264457703, "learning_rate": 0.0002, "epoch": 1.5277665508043003, "step": 9450}, {"loss": 0.7628, "grad_norm": 0.5363983511924744, "learning_rate": 0.0002, "epoch": 1.5293832349850458, "step": 9460}, {"loss": 0.7154, "grad_norm": 0.6184861063957214, "learning_rate": 0.0002, "epoch": 1.530999919165791, "step": 9470}, {"loss": 0.7837, "grad_norm": 0.5991285443305969, "learning_rate": 0.0002, "epoch": 1.5326166033465363, "step": 9480}, {"loss": 0.7827, "grad_norm": 0.8176587820053101, "learning_rate": 0.0002, "epoch": 1.5342332875272815, "step": 9490}, {"loss": 0.7415, "grad_norm": 0.6473721861839294, "learning_rate": 0.0002, "epoch": 1.5358499717080267, "step": 9500}, {"loss": 0.7632, "grad_norm": 0.7319952845573425, "learning_rate": 0.0002, "epoch": 1.5374666558887722, "step": 9510}, {"loss": 0.7706, "grad_norm": 0.702900230884552, "learning_rate": 0.0002, "epoch": 1.5390833400695174, "step": 9520}, {"loss": 0.7754, "grad_norm": 0.7971600294113159, "learning_rate": 0.0002, "epoch": 1.5407000242502629, "step": 9530}, {"loss": 0.7352, "grad_norm": 0.6527525186538696, "learning_rate": 0.0002, "epoch": 1.542316708431008, "step": 9540}, {"loss": 0.7425, "grad_norm": 0.5791676044464111, "learning_rate": 0.0002, "epoch": 1.5439333926117533, "step": 9550}, {"loss": 0.7585, "grad_norm": 0.5619390606880188, "learning_rate": 0.0002, "epoch": 1.5455500767924986, "step": 9560}, {"loss": 0.7894, "grad_norm": 0.5701689124107361, "learning_rate": 0.0002, "epoch": 1.5471667609732438, "step": 9570}, {"loss": 0.793, "grad_norm": 0.47549352049827576, "learning_rate": 0.0002, "epoch": 1.548783445153989, "step": 9580}, {"loss": 0.7276, "grad_norm": 0.8730611205101013, "learning_rate": 0.0002, "epoch": 1.5504001293347345, "step": 9590}, {"loss": 0.798, "grad_norm": 0.6842091083526611, "learning_rate": 0.0002, "epoch": 1.5520168135154797, "step": 9600}, {"loss": 0.7528, "grad_norm": 0.6675129532814026, "learning_rate": 0.0002, "epoch": 1.5536334976962252, "step": 9610}, {"loss": 0.7954, "grad_norm": 0.8173956274986267, "learning_rate": 0.0002, "epoch": 1.5552501818769704, "step": 9620}, {"loss": 0.7535, "grad_norm": 0.724947452545166, "learning_rate": 0.0002, "epoch": 1.5568668660577156, "step": 9630}, {"loss": 0.7738, "grad_norm": 0.6154758930206299, "learning_rate": 0.0002, "epoch": 1.5584835502384609, "step": 9640}, {"loss": 0.7568, "grad_norm": 0.6072008013725281, "learning_rate": 0.0002, "epoch": 1.560100234419206, "step": 9650}, {"loss": 0.7219, "grad_norm": 0.659010648727417, "learning_rate": 0.0002, "epoch": 1.5617169185999515, "step": 9660}, {"loss": 0.673, "grad_norm": 0.65857994556427, "learning_rate": 0.0002, "epoch": 1.5633336027806968, "step": 9670}, {"loss": 0.7156, "grad_norm": 0.5914267301559448, "learning_rate": 0.0002, "epoch": 1.5649502869614422, "step": 9680}, {"loss": 0.7414, "grad_norm": 0.6248020529747009, "learning_rate": 0.0002, "epoch": 1.5665669711421875, "step": 9690}, {"loss": 0.694, "grad_norm": 0.7147795557975769, "learning_rate": 0.0002, "epoch": 1.5681836553229327, "step": 9700}, {"loss": 0.7335, "grad_norm": 0.7076232433319092, "learning_rate": 0.0002, "epoch": 1.569800339503678, "step": 9710}, {"loss": 0.7413, "grad_norm": 0.6217400431632996, "learning_rate": 0.0002, "epoch": 1.5714170236844232, "step": 9720}, {"loss": 0.7296, "grad_norm": 0.6709911227226257, "learning_rate": 0.0002, "epoch": 1.5730337078651684, "step": 9730}, {"loss": 0.7306, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 1.5746503920459138, "step": 9740}, {"loss": 0.7242, "grad_norm": 0.6241145730018616, "learning_rate": 0.0002, "epoch": 1.576267076226659, "step": 9750}, {"loss": 0.7384, "grad_norm": 0.4960934817790985, "learning_rate": 0.0002, "epoch": 1.5778837604074045, "step": 9760}, {"loss": 0.725, "grad_norm": 0.6593309640884399, "learning_rate": 0.0002, "epoch": 1.5795004445881498, "step": 9770}, {"loss": 0.7531, "grad_norm": 0.5814042091369629, "learning_rate": 0.0002, "epoch": 1.581117128768895, "step": 9780}, {"loss": 0.7109, "grad_norm": 0.5936070680618286, "learning_rate": 0.0002, "epoch": 1.5827338129496402, "step": 9790}, {"loss": 0.7769, "grad_norm": 0.6454403400421143, "learning_rate": 0.0002, "epoch": 1.5843504971303854, "step": 9800}, {"loss": 0.7677, "grad_norm": 0.7612107992172241, "learning_rate": 0.0002, "epoch": 1.585967181311131, "step": 9810}, {"loss": 0.7649, "grad_norm": 0.6494482755661011, "learning_rate": 0.0002, "epoch": 1.5875838654918761, "step": 9820}, {"loss": 0.7569, "grad_norm": 0.7825694680213928, "learning_rate": 0.0002, "epoch": 1.5892005496726216, "step": 9830}, {"loss": 0.706, "grad_norm": 0.6757757663726807, "learning_rate": 0.0002, "epoch": 1.5908172338533668, "step": 9840}, {"loss": 0.7803, "grad_norm": 0.7105609178543091, "learning_rate": 0.0002, "epoch": 1.592433918034112, "step": 9850}, {"loss": 0.7925, "grad_norm": 0.7596991062164307, "learning_rate": 0.0002, "epoch": 1.5940506022148573, "step": 9860}, {"loss": 0.7108, "grad_norm": 0.5681525468826294, "learning_rate": 0.0002, "epoch": 1.5956672863956025, "step": 9870}, {"loss": 0.7811, "grad_norm": 0.6090980768203735, "learning_rate": 0.0002, "epoch": 1.5972839705763477, "step": 9880}, {"loss": 0.7339, "grad_norm": 0.6271613240242004, "learning_rate": 0.0002, "epoch": 1.5989006547570932, "step": 9890}, {"loss": 0.7419, "grad_norm": 0.7656369805335999, "learning_rate": 0.0002, "epoch": 1.6005173389378387, "step": 9900}, {"loss": 0.7336, "grad_norm": 0.7504446506500244, "learning_rate": 0.0002, "epoch": 1.6021340231185839, "step": 9910}, {"loss": 0.7479, "grad_norm": 0.659656286239624, "learning_rate": 0.0002, "epoch": 1.6037507072993291, "step": 9920}, {"loss": 0.7483, "grad_norm": 0.6006826162338257, "learning_rate": 0.0002, "epoch": 1.6053673914800743, "step": 9930}, {"loss": 0.732, "grad_norm": 0.7872757911682129, "learning_rate": 0.0002, "epoch": 1.6069840756608196, "step": 9940}, {"loss": 0.768, "grad_norm": 0.5545852780342102, "learning_rate": 0.0002, "epoch": 1.6086007598415648, "step": 9950}, {"loss": 0.8064, "grad_norm": 0.7429468631744385, "learning_rate": 0.0002, "epoch": 1.6102174440223103, "step": 9960}, {"loss": 0.714, "grad_norm": 0.6873556971549988, "learning_rate": 0.0002, "epoch": 1.6118341282030555, "step": 9970}, {"loss": 0.7324, "grad_norm": 0.5874287486076355, "learning_rate": 0.0002, "epoch": 1.613450812383801, "step": 9980}, {"loss": 0.7141, "grad_norm": 0.6039386987686157, "learning_rate": 0.0002, "epoch": 1.6150674965645462, "step": 9990}, {"loss": 0.6674, "grad_norm": 0.6233575940132141, "learning_rate": 0.0002, "epoch": 1.6166841807452914, "step": 10000}, {"loss": 0.7602, "grad_norm": 0.7676448225975037, "learning_rate": 0.0002, "epoch": 1.6183008649260366, "step": 10010}, {"loss": 0.7784, "grad_norm": 0.6565698385238647, "learning_rate": 0.0002, "epoch": 1.6199175491067819, "step": 10020}, {"loss": 0.7104, "grad_norm": 0.6787590384483337, "learning_rate": 0.0002, "epoch": 1.6215342332875273, "step": 10030}, {"loss": 0.7464, "grad_norm": 0.6137678027153015, "learning_rate": 0.0002, "epoch": 1.6231509174682726, "step": 10040}, {"loss": 0.7646, "grad_norm": 0.5236800312995911, "learning_rate": 0.0002, "epoch": 1.624767601649018, "step": 10050}, {"loss": 0.7437, "grad_norm": 0.7626367807388306, "learning_rate": 0.0002, "epoch": 1.6263842858297632, "step": 10060}, {"loss": 0.7273, "grad_norm": 0.5657260417938232, "learning_rate": 0.0002, "epoch": 1.6280009700105085, "step": 10070}, {"loss": 0.7354, "grad_norm": 0.4913991391658783, "learning_rate": 0.0002, "epoch": 1.6296176541912537, "step": 10080}, {"loss": 0.7596, "grad_norm": 0.7715556621551514, "learning_rate": 0.0002, "epoch": 1.631234338371999, "step": 10090}, {"loss": 0.7105, "grad_norm": 0.6509000062942505, "learning_rate": 0.0002, "epoch": 1.6328510225527442, "step": 10100}, {"loss": 0.7274, "grad_norm": 0.6215850114822388, "learning_rate": 0.0002, "epoch": 1.6344677067334896, "step": 10110}, {"loss": 0.7705, "grad_norm": 0.6956844329833984, "learning_rate": 0.0002, "epoch": 1.6360843909142349, "step": 10120}, {"loss": 0.7129, "grad_norm": 0.6111597418785095, "learning_rate": 0.0002, "epoch": 1.6377010750949803, "step": 10130}, {"loss": 0.6955, "grad_norm": 0.6518288850784302, "learning_rate": 0.0002, "epoch": 1.6393177592757255, "step": 10140}, {"loss": 0.731, "grad_norm": 0.6914522051811218, "learning_rate": 0.0002, "epoch": 1.6409344434564708, "step": 10150}, {"loss": 0.7295, "grad_norm": 0.63785719871521, "learning_rate": 0.0002, "epoch": 1.642551127637216, "step": 10160}, {"loss": 0.7355, "grad_norm": 0.6379287838935852, "learning_rate": 0.0002, "epoch": 1.6441678118179612, "step": 10170}, {"loss": 0.7359, "grad_norm": 0.6793403029441833, "learning_rate": 0.0002, "epoch": 1.6457844959987067, "step": 10180}, {"loss": 0.7402, "grad_norm": 0.6099132895469666, "learning_rate": 0.0002, "epoch": 1.647401180179452, "step": 10190}, {"loss": 0.7353, "grad_norm": 0.5869854092597961, "learning_rate": 0.0002, "epoch": 1.6490178643601974, "step": 10200}, {"loss": 0.8308, "grad_norm": 0.7716999053955078, "learning_rate": 0.0002, "epoch": 1.6506345485409426, "step": 10210}, {"loss": 0.7215, "grad_norm": 0.6854110360145569, "learning_rate": 0.0002, "epoch": 1.6522512327216878, "step": 10220}, {"loss": 0.782, "grad_norm": 0.6957170367240906, "learning_rate": 0.0002, "epoch": 1.653867916902433, "step": 10230}, {"loss": 0.7282, "grad_norm": 0.6932903528213501, "learning_rate": 0.0002, "epoch": 1.6554846010831783, "step": 10240}, {"loss": 0.7478, "grad_norm": 0.7713165283203125, "learning_rate": 0.0002, "epoch": 1.6571012852639235, "step": 10250}, {"loss": 0.7099, "grad_norm": 0.7455793619155884, "learning_rate": 0.0002, "epoch": 1.658717969444669, "step": 10260}, {"loss": 0.7524, "grad_norm": 0.5464168190956116, "learning_rate": 0.0002, "epoch": 1.6603346536254144, "step": 10270}, {"loss": 0.7328, "grad_norm": 0.6782926321029663, "learning_rate": 0.0002, "epoch": 1.6619513378061597, "step": 10280}, {"loss": 0.7801, "grad_norm": 0.7962649464607239, "learning_rate": 0.0002, "epoch": 1.663568021986905, "step": 10290}, {"loss": 0.7142, "grad_norm": 0.6814526319503784, "learning_rate": 0.0002, "epoch": 1.6651847061676501, "step": 10300}, {"loss": 0.7285, "grad_norm": 0.656895101070404, "learning_rate": 0.0002, "epoch": 1.6668013903483954, "step": 10310}, {"loss": 0.7358, "grad_norm": 0.6085672378540039, "learning_rate": 0.0002, "epoch": 1.6684180745291406, "step": 10320}, {"loss": 0.7074, "grad_norm": 0.585508406162262, "learning_rate": 0.0002, "epoch": 1.670034758709886, "step": 10330}, {"loss": 0.7604, "grad_norm": 0.6930184364318848, "learning_rate": 0.0002, "epoch": 1.6716514428906313, "step": 10340}, {"loss": 0.7169, "grad_norm": 0.575663149356842, "learning_rate": 0.0002, "epoch": 1.6732681270713767, "step": 10350}, {"loss": 0.7198, "grad_norm": 0.582502543926239, "learning_rate": 0.0002, "epoch": 1.674884811252122, "step": 10360}, {"loss": 0.7793, "grad_norm": 0.5668916702270508, "learning_rate": 0.0002, "epoch": 1.6765014954328672, "step": 10370}, {"loss": 0.7478, "grad_norm": 0.6070065498352051, "learning_rate": 0.0002, "epoch": 1.6781181796136124, "step": 10380}, {"loss": 0.7939, "grad_norm": 0.6141316294670105, "learning_rate": 0.0002, "epoch": 1.6797348637943577, "step": 10390}, {"loss": 0.7573, "grad_norm": 0.8359124064445496, "learning_rate": 0.0002, "epoch": 1.6813515479751031, "step": 10400}, {"loss": 0.7488, "grad_norm": 0.5378185510635376, "learning_rate": 0.0002, "epoch": 1.6829682321558483, "step": 10410}, {"loss": 0.7588, "grad_norm": 0.6959536075592041, "learning_rate": 0.0002, "epoch": 1.6845849163365938, "step": 10420}, {"loss": 0.7872, "grad_norm": 0.6514357328414917, "learning_rate": 0.0002, "epoch": 1.686201600517339, "step": 10430}, {"loss": 0.725, "grad_norm": 0.7706646919250488, "learning_rate": 0.0002, "epoch": 1.6878182846980843, "step": 10440}, {"loss": 0.7673, "grad_norm": 0.6183337569236755, "learning_rate": 0.0002, "epoch": 1.6894349688788295, "step": 10450}, {"loss": 0.7566, "grad_norm": 0.6123278141021729, "learning_rate": 0.0002, "epoch": 1.6910516530595747, "step": 10460}, {"loss": 0.7169, "grad_norm": 0.6894851326942444, "learning_rate": 0.0002, "epoch": 1.69266833724032, "step": 10470}, {"loss": 0.7435, "grad_norm": 0.7497312426567078, "learning_rate": 0.0002, "epoch": 1.6942850214210654, "step": 10480}, {"loss": 0.7544, "grad_norm": 0.5968214273452759, "learning_rate": 0.0002, "epoch": 1.6959017056018106, "step": 10490}, {"loss": 0.6793, "grad_norm": 0.6747927069664001, "learning_rate": 0.0002, "epoch": 1.697518389782556, "step": 10500}, {"loss": 0.7415, "grad_norm": 0.5708310008049011, "learning_rate": 0.0002, "epoch": 1.6991350739633013, "step": 10510}, {"loss": 0.7385, "grad_norm": 0.606526792049408, "learning_rate": 0.0002, "epoch": 1.7007517581440466, "step": 10520}, {"loss": 0.7204, "grad_norm": 0.662011981010437, "learning_rate": 0.0002, "epoch": 1.7023684423247918, "step": 10530}, {"loss": 0.7999, "grad_norm": 0.7583045363426208, "learning_rate": 0.0002, "epoch": 1.703985126505537, "step": 10540}, {"loss": 0.7563, "grad_norm": 0.721632182598114, "learning_rate": 0.0002, "epoch": 1.7056018106862825, "step": 10550}, {"loss": 0.7407, "grad_norm": 0.6107715368270874, "learning_rate": 0.0002, "epoch": 1.7072184948670277, "step": 10560}, {"loss": 0.7519, "grad_norm": 0.6652471423149109, "learning_rate": 0.0002, "epoch": 1.7088351790477732, "step": 10570}, {"loss": 0.7767, "grad_norm": 0.6308087110519409, "learning_rate": 0.0002, "epoch": 1.7104518632285184, "step": 10580}, {"loss": 0.7659, "grad_norm": 0.5464386940002441, "learning_rate": 0.0002, "epoch": 1.7120685474092636, "step": 10590}, {"loss": 0.7063, "grad_norm": 0.6558911204338074, "learning_rate": 0.0002, "epoch": 1.7136852315900089, "step": 10600}, {"loss": 0.7126, "grad_norm": 0.5665024518966675, "learning_rate": 0.0002, "epoch": 1.715301915770754, "step": 10610}, {"loss": 0.6958, "grad_norm": 0.7888094186782837, "learning_rate": 0.0002, "epoch": 1.7169185999514993, "step": 10620}, {"loss": 0.7785, "grad_norm": 0.7084909081459045, "learning_rate": 0.0002, "epoch": 1.7185352841322448, "step": 10630}, {"loss": 0.7557, "grad_norm": 0.7982324361801147, "learning_rate": 0.0002, "epoch": 1.7201519683129902, "step": 10640}, {"loss": 0.7345, "grad_norm": 0.6418732404708862, "learning_rate": 0.0002, "epoch": 1.7217686524937355, "step": 10650}, {"loss": 0.7734, "grad_norm": 0.7636681795120239, "learning_rate": 0.0002, "epoch": 1.7233853366744807, "step": 10660}, {"loss": 0.7541, "grad_norm": 0.5646875500679016, "learning_rate": 0.0002, "epoch": 1.725002020855226, "step": 10670}, {"loss": 0.7642, "grad_norm": 0.5231260657310486, "learning_rate": 0.0002, "epoch": 1.7266187050359711, "step": 10680}, {"loss": 0.7846, "grad_norm": 0.7635011672973633, "learning_rate": 0.0002, "epoch": 1.7282353892167164, "step": 10690}, {"loss": 0.7471, "grad_norm": 0.7518259286880493, "learning_rate": 0.0002, "epoch": 1.7298520733974618, "step": 10700}, {"loss": 0.751, "grad_norm": 0.7295602560043335, "learning_rate": 0.0002, "epoch": 1.731468757578207, "step": 10710}, {"loss": 0.731, "grad_norm": 0.6984632015228271, "learning_rate": 0.0002, "epoch": 1.7330854417589525, "step": 10720}, {"loss": 0.7921, "grad_norm": 0.6198219060897827, "learning_rate": 0.0002, "epoch": 1.7347021259396977, "step": 10730}, {"loss": 0.7642, "grad_norm": 0.6957576274871826, "learning_rate": 0.0002, "epoch": 1.736318810120443, "step": 10740}, {"loss": 0.7917, "grad_norm": 0.6430263519287109, "learning_rate": 0.0002, "epoch": 1.7379354943011882, "step": 10750}, {"loss": 0.7156, "grad_norm": 0.6134995222091675, "learning_rate": 0.0002, "epoch": 1.7395521784819334, "step": 10760}, {"loss": 0.7584, "grad_norm": 0.7209452986717224, "learning_rate": 0.0002, "epoch": 1.741168862662679, "step": 10770}, {"loss": 0.7528, "grad_norm": 0.6735447645187378, "learning_rate": 0.0002, "epoch": 1.7427855468434241, "step": 10780}, {"loss": 0.756, "grad_norm": 0.5605693459510803, "learning_rate": 0.0002, "epoch": 1.7444022310241696, "step": 10790}, {"loss": 0.7759, "grad_norm": 0.6882363557815552, "learning_rate": 0.0002, "epoch": 1.7460189152049148, "step": 10800}, {"loss": 0.7544, "grad_norm": 0.6386259198188782, "learning_rate": 0.0002, "epoch": 1.74763559938566, "step": 10810}, {"loss": 0.7697, "grad_norm": 0.6529015302658081, "learning_rate": 0.0002, "epoch": 1.7492522835664053, "step": 10820}, {"loss": 0.7219, "grad_norm": 0.5664082765579224, "learning_rate": 0.0002, "epoch": 1.7508689677471505, "step": 10830}, {"loss": 0.7586, "grad_norm": 0.7532684206962585, "learning_rate": 0.0002, "epoch": 1.7524856519278957, "step": 10840}, {"loss": 0.6919, "grad_norm": 0.77171391248703, "learning_rate": 0.0002, "epoch": 1.7541023361086412, "step": 10850}, {"loss": 0.785, "grad_norm": 0.7255431413650513, "learning_rate": 0.0002, "epoch": 1.7557190202893864, "step": 10860}, {"loss": 0.7458, "grad_norm": 0.763083279132843, "learning_rate": 0.0002, "epoch": 1.7573357044701319, "step": 10870}, {"loss": 0.7846, "grad_norm": 0.6042402982711792, "learning_rate": 0.0002, "epoch": 1.758952388650877, "step": 10880}, {"loss": 0.7027, "grad_norm": 0.7642518281936646, "learning_rate": 0.0002, "epoch": 1.7605690728316223, "step": 10890}, {"loss": 0.746, "grad_norm": 0.6347904801368713, "learning_rate": 0.0002, "epoch": 1.7621857570123676, "step": 10900}, {"loss": 0.7458, "grad_norm": 0.5371627807617188, "learning_rate": 0.0002, "epoch": 1.7638024411931128, "step": 10910}, {"loss": 0.7466, "grad_norm": 0.6840225458145142, "learning_rate": 0.0002, "epoch": 1.7654191253738583, "step": 10920}, {"loss": 0.725, "grad_norm": 0.5288469195365906, "learning_rate": 0.0002, "epoch": 1.7670358095546035, "step": 10930}, {"loss": 0.7863, "grad_norm": 0.69020676612854, "learning_rate": 0.0002, "epoch": 1.768652493735349, "step": 10940}, {"loss": 0.7468, "grad_norm": 0.5943242311477661, "learning_rate": 0.0002, "epoch": 1.7702691779160942, "step": 10950}, {"loss": 0.7244, "grad_norm": 0.5616418123245239, "learning_rate": 0.0002, "epoch": 1.7718858620968394, "step": 10960}, {"loss": 0.7137, "grad_norm": 0.7209470868110657, "learning_rate": 0.0002, "epoch": 1.7735025462775846, "step": 10970}, {"loss": 0.7459, "grad_norm": 0.6657957434654236, "learning_rate": 0.0002, "epoch": 1.7751192304583299, "step": 10980}, {"loss": 0.7076, "grad_norm": 0.6469064950942993, "learning_rate": 0.0002, "epoch": 1.776735914639075, "step": 10990}, {"loss": 0.7321, "grad_norm": 0.6615678071975708, "learning_rate": 0.0002, "epoch": 1.7783525988198206, "step": 11000}, {"loss": 0.747, "grad_norm": 0.6722439527511597, "learning_rate": 0.0002, "epoch": 1.779969283000566, "step": 11010}, {"loss": 0.7302, "grad_norm": 0.634136974811554, "learning_rate": 0.0002, "epoch": 1.7815859671813112, "step": 11020}, {"loss": 0.8105, "grad_norm": 0.6024377346038818, "learning_rate": 0.0002, "epoch": 1.7832026513620565, "step": 11030}, {"loss": 0.7855, "grad_norm": 0.6909403800964355, "learning_rate": 0.0002, "epoch": 1.7848193355428017, "step": 11040}, {"loss": 0.7471, "grad_norm": 0.7148767709732056, "learning_rate": 0.0002, "epoch": 1.786436019723547, "step": 11050}, {"loss": 0.7145, "grad_norm": 0.7442979216575623, "learning_rate": 0.0002, "epoch": 1.7880527039042922, "step": 11060}, {"loss": 0.7215, "grad_norm": 0.6830431818962097, "learning_rate": 0.0002, "epoch": 1.7896693880850376, "step": 11070}, {"loss": 0.7625, "grad_norm": 0.9172667264938354, "learning_rate": 0.0002, "epoch": 1.7912860722657828, "step": 11080}, {"loss": 0.76, "grad_norm": 0.6799490451812744, "learning_rate": 0.0002, "epoch": 1.7929027564465283, "step": 11090}, {"loss": 0.7716, "grad_norm": 0.7617024779319763, "learning_rate": 0.0002, "epoch": 1.7945194406272735, "step": 11100}, {"loss": 0.7586, "grad_norm": 0.7701810002326965, "learning_rate": 0.0002, "epoch": 1.7961361248080188, "step": 11110}, {"loss": 0.7843, "grad_norm": 0.7454385757446289, "learning_rate": 0.0002, "epoch": 1.797752808988764, "step": 11120}, {"loss": 0.7873, "grad_norm": 0.6121436953544617, "learning_rate": 0.0002, "epoch": 1.7993694931695092, "step": 11130}, {"loss": 0.7305, "grad_norm": 0.6237571835517883, "learning_rate": 0.0002, "epoch": 1.8009861773502547, "step": 11140}, {"loss": 0.6827, "grad_norm": 0.6818515658378601, "learning_rate": 0.0002, "epoch": 1.802602861531, "step": 11150}, {"loss": 0.6876, "grad_norm": 0.7768308520317078, "learning_rate": 0.0002, "epoch": 1.8042195457117454, "step": 11160}, {"loss": 0.7533, "grad_norm": 0.6875537633895874, "learning_rate": 0.0002, "epoch": 1.8058362298924906, "step": 11170}, {"loss": 0.761, "grad_norm": 0.7950584888458252, "learning_rate": 0.0002, "epoch": 1.8074529140732358, "step": 11180}, {"loss": 0.7623, "grad_norm": 0.8210248351097107, "learning_rate": 0.0002, "epoch": 1.809069598253981, "step": 11190}, {"loss": 0.7556, "grad_norm": 0.6674110889434814, "learning_rate": 0.0002, "epoch": 1.8106862824347263, "step": 11200}, {"loss": 0.7663, "grad_norm": 0.6261674761772156, "learning_rate": 0.0002, "epoch": 1.8123029666154715, "step": 11210}, {"loss": 0.7122, "grad_norm": 0.6484741568565369, "learning_rate": 0.0002, "epoch": 1.813919650796217, "step": 11220}, {"loss": 0.7718, "grad_norm": 0.6231244206428528, "learning_rate": 0.0002, "epoch": 1.8155363349769622, "step": 11230}, {"loss": 0.7152, "grad_norm": 0.7243146896362305, "learning_rate": 0.0002, "epoch": 1.8171530191577077, "step": 11240}, {"loss": 0.7448, "grad_norm": 0.6776193380355835, "learning_rate": 0.0002, "epoch": 1.818769703338453, "step": 11250}, {"loss": 0.7317, "grad_norm": 0.5973618030548096, "learning_rate": 0.0002, "epoch": 1.8203863875191981, "step": 11260}, {"loss": 0.7961, "grad_norm": 0.6451361179351807, "learning_rate": 0.0002, "epoch": 1.8220030716999434, "step": 11270}, {"loss": 0.7611, "grad_norm": 0.5963068008422852, "learning_rate": 0.0002, "epoch": 1.8236197558806886, "step": 11280}, {"loss": 0.7466, "grad_norm": 0.536902129650116, "learning_rate": 0.0002, "epoch": 1.825236440061434, "step": 11290}, {"loss": 0.708, "grad_norm": 0.6993787288665771, "learning_rate": 0.0002, "epoch": 1.8268531242421793, "step": 11300}, {"loss": 0.7153, "grad_norm": 0.6135255098342896, "learning_rate": 0.0002, "epoch": 1.8284698084229247, "step": 11310}, {"loss": 0.7423, "grad_norm": 0.6057423949241638, "learning_rate": 0.0002, "epoch": 1.83008649260367, "step": 11320}, {"loss": 0.735, "grad_norm": 0.6598812341690063, "learning_rate": 0.0002, "epoch": 1.8317031767844152, "step": 11330}, {"loss": 0.7278, "grad_norm": 0.6075948476791382, "learning_rate": 0.0002, "epoch": 1.8333198609651604, "step": 11340}, {"loss": 0.7846, "grad_norm": 0.7065447568893433, "learning_rate": 0.0002, "epoch": 1.8349365451459057, "step": 11350}, {"loss": 0.7365, "grad_norm": 0.680526614189148, "learning_rate": 0.0002, "epoch": 1.8365532293266509, "step": 11360}, {"loss": 0.7152, "grad_norm": 0.6356695294380188, "learning_rate": 0.0002, "epoch": 1.8381699135073963, "step": 11370}, {"loss": 0.721, "grad_norm": 0.6399052143096924, "learning_rate": 0.0002, "epoch": 1.8397865976881416, "step": 11380}, {"loss": 0.7618, "grad_norm": 0.6125704050064087, "learning_rate": 0.0002, "epoch": 1.841403281868887, "step": 11390}, {"loss": 0.755, "grad_norm": 0.7124643325805664, "learning_rate": 0.0002, "epoch": 1.8430199660496323, "step": 11400}, {"loss": 0.7972, "grad_norm": 0.6099604964256287, "learning_rate": 0.0002, "epoch": 1.8446366502303775, "step": 11410}, {"loss": 0.7187, "grad_norm": 0.7338208556175232, "learning_rate": 0.0002, "epoch": 1.8462533344111227, "step": 11420}, {"loss": 0.7007, "grad_norm": 0.7534668445587158, "learning_rate": 0.0002, "epoch": 1.847870018591868, "step": 11430}, {"loss": 0.7464, "grad_norm": 0.6135470271110535, "learning_rate": 0.0002, "epoch": 1.8494867027726134, "step": 11440}, {"loss": 0.7955, "grad_norm": 0.6229309439659119, "learning_rate": 0.0002, "epoch": 1.8511033869533586, "step": 11450}, {"loss": 0.7594, "grad_norm": 0.706423282623291, "learning_rate": 0.0002, "epoch": 1.852720071134104, "step": 11460}, {"loss": 0.7411, "grad_norm": 0.5460049510002136, "learning_rate": 0.0002, "epoch": 1.8543367553148493, "step": 11470}, {"loss": 0.7416, "grad_norm": 0.6616711020469666, "learning_rate": 0.0002, "epoch": 1.8559534394955945, "step": 11480}, {"loss": 0.729, "grad_norm": 0.6372783184051514, "learning_rate": 0.0002, "epoch": 1.8575701236763398, "step": 11490}, {"loss": 0.7333, "grad_norm": 0.7162668108940125, "learning_rate": 0.0002, "epoch": 1.859186807857085, "step": 11500}, {"loss": 0.7747, "grad_norm": 0.6605209708213806, "learning_rate": 0.0002, "epoch": 1.8608034920378305, "step": 11510}, {"loss": 0.7258, "grad_norm": 0.6933956742286682, "learning_rate": 0.0002, "epoch": 1.8624201762185757, "step": 11520}, {"loss": 0.7243, "grad_norm": 0.6582090854644775, "learning_rate": 0.0002, "epoch": 1.8640368603993211, "step": 11530}, {"loss": 0.7313, "grad_norm": 0.6416500806808472, "learning_rate": 0.0002, "epoch": 1.8656535445800664, "step": 11540}, {"loss": 0.7372, "grad_norm": 0.5434312224388123, "learning_rate": 0.0002, "epoch": 1.8672702287608116, "step": 11550}, {"loss": 0.7635, "grad_norm": 0.6827567219734192, "learning_rate": 0.0002, "epoch": 1.8688869129415568, "step": 11560}, {"loss": 0.7137, "grad_norm": 0.7354370951652527, "learning_rate": 0.0002, "epoch": 1.870503597122302, "step": 11570}, {"loss": 0.7526, "grad_norm": 0.590372622013092, "learning_rate": 0.0002, "epoch": 1.8721202813030473, "step": 11580}, {"loss": 0.731, "grad_norm": 0.853183925151825, "learning_rate": 0.0002, "epoch": 1.8737369654837928, "step": 11590}, {"loss": 0.7487, "grad_norm": 0.822678804397583, "learning_rate": 0.0002, "epoch": 1.875353649664538, "step": 11600}, {"loss": 0.7427, "grad_norm": 0.6591550707817078, "learning_rate": 0.0002, "epoch": 1.8769703338452834, "step": 11610}, {"loss": 0.7054, "grad_norm": 0.7475301623344421, "learning_rate": 0.0002, "epoch": 1.8785870180260287, "step": 11620}, {"loss": 0.811, "grad_norm": 0.6390765309333801, "learning_rate": 0.0002, "epoch": 1.880203702206774, "step": 11630}, {"loss": 0.7531, "grad_norm": 0.6589758992195129, "learning_rate": 0.0002, "epoch": 1.8818203863875191, "step": 11640}, {"loss": 0.7475, "grad_norm": 0.6765508651733398, "learning_rate": 0.0002, "epoch": 1.8834370705682644, "step": 11650}, {"loss": 0.738, "grad_norm": 0.6527857780456543, "learning_rate": 0.0002, "epoch": 1.8850537547490098, "step": 11660}, {"loss": 0.7504, "grad_norm": 0.6642923951148987, "learning_rate": 0.0002, "epoch": 1.886670438929755, "step": 11670}, {"loss": 0.7701, "grad_norm": 0.6945584416389465, "learning_rate": 0.0002, "epoch": 1.8882871231105005, "step": 11680}, {"loss": 0.7711, "grad_norm": 0.694018542766571, "learning_rate": 0.0002, "epoch": 1.8899038072912457, "step": 11690}, {"loss": 0.7195, "grad_norm": 0.7237417101860046, "learning_rate": 0.0002, "epoch": 1.891520491471991, "step": 11700}, {"loss": 0.7491, "grad_norm": 0.7401309609413147, "learning_rate": 0.0002, "epoch": 1.8931371756527362, "step": 11710}, {"loss": 0.805, "grad_norm": 0.6537784337997437, "learning_rate": 0.0002, "epoch": 1.8947538598334814, "step": 11720}, {"loss": 0.793, "grad_norm": 0.7398539185523987, "learning_rate": 0.0002, "epoch": 1.8963705440142267, "step": 11730}, {"loss": 0.7561, "grad_norm": 0.6696075797080994, "learning_rate": 0.0002, "epoch": 1.8979872281949721, "step": 11740}, {"loss": 0.7353, "grad_norm": 0.6014142036437988, "learning_rate": 0.0002, "epoch": 1.8996039123757174, "step": 11750}, {"loss": 0.7714, "grad_norm": 0.7023524641990662, "learning_rate": 0.0002, "epoch": 1.9012205965564628, "step": 11760}, {"loss": 0.7088, "grad_norm": 0.739973783493042, "learning_rate": 0.0002, "epoch": 1.902837280737208, "step": 11770}, {"loss": 0.7848, "grad_norm": 0.5576770901679993, "learning_rate": 0.0002, "epoch": 1.9044539649179533, "step": 11780}, {"loss": 0.7483, "grad_norm": 0.6907393932342529, "learning_rate": 0.0002, "epoch": 1.9060706490986985, "step": 11790}, {"loss": 0.7827, "grad_norm": 0.6934581995010376, "learning_rate": 0.0002, "epoch": 1.9076873332794437, "step": 11800}, {"loss": 0.7199, "grad_norm": 0.591774582862854, "learning_rate": 0.0002, "epoch": 1.9093040174601892, "step": 11810}, {"loss": 0.7333, "grad_norm": 0.6249791383743286, "learning_rate": 0.0002, "epoch": 1.9109207016409344, "step": 11820}, {"loss": 0.7581, "grad_norm": 0.6755744218826294, "learning_rate": 0.0002, "epoch": 1.9125373858216799, "step": 11830}, {"loss": 0.696, "grad_norm": 0.7286285161972046, "learning_rate": 0.0002, "epoch": 1.914154070002425, "step": 11840}, {"loss": 0.7509, "grad_norm": 0.7867850065231323, "learning_rate": 0.0002, "epoch": 1.9157707541831703, "step": 11850}, {"loss": 0.735, "grad_norm": 0.6283972859382629, "learning_rate": 0.0002, "epoch": 1.9173874383639156, "step": 11860}, {"loss": 0.7296, "grad_norm": 0.605823814868927, "learning_rate": 0.0002, "epoch": 1.9190041225446608, "step": 11870}, {"loss": 0.6598, "grad_norm": 0.5927976965904236, "learning_rate": 0.0002, "epoch": 1.920620806725406, "step": 11880}, {"loss": 0.7649, "grad_norm": 0.5974002480506897, "learning_rate": 0.0002, "epoch": 1.9222374909061515, "step": 11890}, {"loss": 0.7843, "grad_norm": 0.7091866135597229, "learning_rate": 0.0002, "epoch": 1.923854175086897, "step": 11900}, {"loss": 0.775, "grad_norm": 0.72496497631073, "learning_rate": 0.0002, "epoch": 1.9254708592676422, "step": 11910}, {"loss": 0.7153, "grad_norm": 0.6131896376609802, "learning_rate": 0.0002, "epoch": 1.9270875434483874, "step": 11920}, {"loss": 0.7228, "grad_norm": 0.6556436419487, "learning_rate": 0.0002, "epoch": 1.9287042276291326, "step": 11930}, {"loss": 0.7319, "grad_norm": 0.622932493686676, "learning_rate": 0.0002, "epoch": 1.9303209118098779, "step": 11940}, {"loss": 0.7592, "grad_norm": 0.6618631482124329, "learning_rate": 0.0002, "epoch": 1.931937595990623, "step": 11950}, {"loss": 0.8332, "grad_norm": 0.630966305732727, "learning_rate": 0.0002, "epoch": 1.9335542801713685, "step": 11960}, {"loss": 0.6854, "grad_norm": 0.6336734890937805, "learning_rate": 0.0002, "epoch": 1.9351709643521138, "step": 11970}, {"loss": 0.7433, "grad_norm": 0.655403196811676, "learning_rate": 0.0002, "epoch": 1.9367876485328592, "step": 11980}, {"loss": 0.7282, "grad_norm": 0.5640574097633362, "learning_rate": 0.0002, "epoch": 1.9384043327136045, "step": 11990}, {"loss": 0.7289, "grad_norm": 0.6322951316833496, "learning_rate": 0.0002, "epoch": 1.9400210168943497, "step": 12000}, {"loss": 0.7627, "grad_norm": 0.615703821182251, "learning_rate": 0.0002, "epoch": 1.941637701075095, "step": 12010}, {"loss": 0.786, "grad_norm": 0.6487536430358887, "learning_rate": 0.0002, "epoch": 1.9432543852558402, "step": 12020}, {"loss": 0.7435, "grad_norm": 0.9209630489349365, "learning_rate": 0.0002, "epoch": 1.9448710694365856, "step": 12030}, {"loss": 0.7274, "grad_norm": 0.67485511302948, "learning_rate": 0.0002, "epoch": 1.9464877536173308, "step": 12040}, {"loss": 0.7551, "grad_norm": 0.6831230521202087, "learning_rate": 0.0002, "epoch": 1.9481044377980763, "step": 12050}, {"loss": 0.7546, "grad_norm": 0.6578302383422852, "learning_rate": 0.0002, "epoch": 1.9497211219788215, "step": 12060}, {"loss": 0.6989, "grad_norm": 0.9975938200950623, "learning_rate": 0.0002, "epoch": 1.9513378061595668, "step": 12070}, {"loss": 0.7952, "grad_norm": 0.6637365221977234, "learning_rate": 0.0002, "epoch": 1.952954490340312, "step": 12080}, {"loss": 0.7482, "grad_norm": 0.605707049369812, "learning_rate": 0.0002, "epoch": 1.9545711745210572, "step": 12090}, {"loss": 0.7768, "grad_norm": 0.6584440469741821, "learning_rate": 0.0002, "epoch": 1.9561878587018025, "step": 12100}, {"loss": 0.7187, "grad_norm": 0.6070835590362549, "learning_rate": 0.0002, "epoch": 1.957804542882548, "step": 12110}, {"loss": 0.7491, "grad_norm": 0.7862601280212402, "learning_rate": 0.0002, "epoch": 1.9594212270632931, "step": 12120}, {"loss": 0.7972, "grad_norm": 0.8175255060195923, "learning_rate": 0.0002, "epoch": 1.9610379112440386, "step": 12130}, {"loss": 0.7242, "grad_norm": 0.5648472905158997, "learning_rate": 0.0002, "epoch": 1.9626545954247838, "step": 12140}, {"loss": 0.7321, "grad_norm": 0.6591973304748535, "learning_rate": 0.0002, "epoch": 1.964271279605529, "step": 12150}, {"loss": 0.739, "grad_norm": 0.5960676074028015, "learning_rate": 0.0002, "epoch": 1.9658879637862743, "step": 12160}, {"loss": 0.7254, "grad_norm": 0.7272544503211975, "learning_rate": 0.0002, "epoch": 1.9675046479670195, "step": 12170}, {"loss": 0.7376, "grad_norm": 0.7176699042320251, "learning_rate": 0.0002, "epoch": 1.969121332147765, "step": 12180}, {"loss": 0.7525, "grad_norm": 0.6927123665809631, "learning_rate": 0.0002, "epoch": 1.9707380163285102, "step": 12190}, {"loss": 0.7318, "grad_norm": 0.5536034107208252, "learning_rate": 0.0002, "epoch": 1.9723547005092557, "step": 12200}, {"loss": 0.7737, "grad_norm": 0.8348390460014343, "learning_rate": 0.0002, "epoch": 1.9739713846900009, "step": 12210}, {"loss": 0.7494, "grad_norm": 0.6591181755065918, "learning_rate": 0.0002, "epoch": 1.9755880688707461, "step": 12220}, {"loss": 0.763, "grad_norm": 1.0624109506607056, "learning_rate": 0.0002, "epoch": 1.9772047530514913, "step": 12230}, {"loss": 0.7541, "grad_norm": 0.9265586137771606, "learning_rate": 0.0002, "epoch": 1.9788214372322366, "step": 12240}, {"loss": 0.7533, "grad_norm": 0.5998196005821228, "learning_rate": 0.0002, "epoch": 1.9804381214129818, "step": 12250}, {"loss": 0.7225, "grad_norm": 0.6960851550102234, "learning_rate": 0.0002, "epoch": 1.9820548055937273, "step": 12260}, {"loss": 0.7398, "grad_norm": 0.7674502730369568, "learning_rate": 0.0002, "epoch": 1.9836714897744727, "step": 12270}, {"loss": 0.7185, "grad_norm": 0.6407275795936584, "learning_rate": 0.0002, "epoch": 1.985288173955218, "step": 12280}, {"loss": 0.7382, "grad_norm": 0.6673079133033752, "learning_rate": 0.0002, "epoch": 1.9869048581359632, "step": 12290}, {"loss": 0.7326, "grad_norm": 0.6989844441413879, "learning_rate": 0.0002, "epoch": 1.9885215423167084, "step": 12300}, {"loss": 0.7559, "grad_norm": 0.7564442157745361, "learning_rate": 0.0002, "epoch": 1.9901382264974536, "step": 12310}, {"loss": 0.7719, "grad_norm": 0.6385478973388672, "learning_rate": 0.0002, "epoch": 1.9917549106781989, "step": 12320}, {"loss": 0.7369, "grad_norm": 0.7193717956542969, "learning_rate": 0.0002, "epoch": 1.9933715948589443, "step": 12330}, {"loss": 0.7583, "grad_norm": 0.7987112402915955, "learning_rate": 0.0002, "epoch": 1.9949882790396896, "step": 12340}, {"loss": 0.7793, "grad_norm": 0.7260826826095581, "learning_rate": 0.0002, "epoch": 1.996604963220435, "step": 12350}, {"loss": 0.7505, "grad_norm": 0.7968255281448364, "learning_rate": 0.0002, "epoch": 1.9982216474011802, "step": 12360}, {"loss": 0.717, "grad_norm": 0.6893062591552734, "learning_rate": 0.0002, "epoch": 1.9998383315819255, "step": 12370}, {"eval_loss": 1.1044032573699951, "eval_runtime": 122.1508, "eval_samples_per_second": 6.001, "eval_steps_per_second": 0.753, "epoch": 2.0, "step": 12371}, {"loss": 0.6604, "grad_norm": 0.7775409817695618, "learning_rate": 0.0002, "epoch": 2.0014550157626707, "step": 12380}, {"loss": 0.6845, "grad_norm": 0.76218581199646, "learning_rate": 0.0002, "epoch": 2.003071699943416, "step": 12390}, {"loss": 0.6909, "grad_norm": 0.5677764415740967, "learning_rate": 0.0002, "epoch": 2.004688384124161, "step": 12400}, {"loss": 0.6584, "grad_norm": 0.808442234992981, "learning_rate": 0.0002, "epoch": 2.006305068304907, "step": 12410}, {"loss": 0.659, "grad_norm": 0.7144765257835388, "learning_rate": 0.0002, "epoch": 2.007921752485652, "step": 12420}, {"loss": 0.6666, "grad_norm": 0.6914031505584717, "learning_rate": 0.0002, "epoch": 2.0095384366663973, "step": 12430}, {"loss": 0.6596, "grad_norm": 0.7581454515457153, "learning_rate": 0.0002, "epoch": 2.0111551208471425, "step": 12440}, {"loss": 0.6785, "grad_norm": 0.8388504981994629, "learning_rate": 0.0002, "epoch": 2.0127718050278878, "step": 12450}, {"loss": 0.6942, "grad_norm": 0.6716406941413879, "learning_rate": 0.0002, "epoch": 2.014388489208633, "step": 12460}, {"loss": 0.6441, "grad_norm": 0.898902416229248, "learning_rate": 0.0002, "epoch": 2.0160051733893782, "step": 12470}, {"loss": 0.6655, "grad_norm": 0.6432679891586304, "learning_rate": 0.0002, "epoch": 2.0176218575701235, "step": 12480}, {"loss": 0.6521, "grad_norm": 0.8021109104156494, "learning_rate": 0.0002, "epoch": 2.019238541750869, "step": 12490}, {"loss": 0.6581, "grad_norm": 0.7039216756820679, "learning_rate": 0.0002, "epoch": 2.0208552259316144, "step": 12500}, {"loss": 0.6521, "grad_norm": 0.646531879901886, "learning_rate": 0.0002, "epoch": 2.0224719101123596, "step": 12510}, {"loss": 0.6302, "grad_norm": 0.783704400062561, "learning_rate": 0.0002, "epoch": 2.024088594293105, "step": 12520}, {"loss": 0.6288, "grad_norm": 0.8805046677589417, "learning_rate": 0.0002, "epoch": 2.02570527847385, "step": 12530}, {"loss": 0.6288, "grad_norm": 0.7289270758628845, "learning_rate": 0.0002, "epoch": 2.0273219626545953, "step": 12540}, {"loss": 0.6663, "grad_norm": 0.71653151512146, "learning_rate": 0.0002, "epoch": 2.0289386468353405, "step": 12550}, {"loss": 0.625, "grad_norm": 0.73281329870224, "learning_rate": 0.0002, "epoch": 2.030555331016086, "step": 12560}, {"loss": 0.6448, "grad_norm": 0.6657090187072754, "learning_rate": 0.0002, "epoch": 2.0321720151968314, "step": 12570}, {"loss": 0.6983, "grad_norm": 0.8241133093833923, "learning_rate": 0.0002, "epoch": 2.0337886993775767, "step": 12580}, {"loss": 0.6488, "grad_norm": 0.5834135413169861, "learning_rate": 0.0002, "epoch": 2.035405383558322, "step": 12590}, {"loss": 0.6188, "grad_norm": 0.84502112865448, "learning_rate": 0.0002, "epoch": 2.037022067739067, "step": 12600}, {"loss": 0.6349, "grad_norm": 0.8952481746673584, "learning_rate": 0.0002, "epoch": 2.0386387519198124, "step": 12610}, {"loss": 0.6923, "grad_norm": 0.7801461815834045, "learning_rate": 0.0002, "epoch": 2.0402554361005576, "step": 12620}, {"loss": 0.6176, "grad_norm": 0.6788367033004761, "learning_rate": 0.0002, "epoch": 2.041872120281303, "step": 12630}, {"loss": 0.6162, "grad_norm": 0.7241756319999695, "learning_rate": 0.0002, "epoch": 2.0434888044620485, "step": 12640}, {"loss": 0.655, "grad_norm": 0.6933388113975525, "learning_rate": 0.0002, "epoch": 2.0451054886427937, "step": 12650}, {"loss": 0.6431, "grad_norm": 0.8029746413230896, "learning_rate": 0.0002, "epoch": 2.046722172823539, "step": 12660}, {"loss": 0.7164, "grad_norm": 0.946399986743927, "learning_rate": 0.0002, "epoch": 2.048338857004284, "step": 12670}, {"loss": 0.638, "grad_norm": 0.7072678804397583, "learning_rate": 0.0002, "epoch": 2.0499555411850294, "step": 12680}, {"loss": 0.6487, "grad_norm": 0.6810618042945862, "learning_rate": 0.0002, "epoch": 2.0515722253657747, "step": 12690}, {"loss": 0.6554, "grad_norm": 0.7661160230636597, "learning_rate": 0.0002, "epoch": 2.05318890954652, "step": 12700}, {"loss": 0.6799, "grad_norm": 0.6350653767585754, "learning_rate": 0.0002, "epoch": 2.0548055937272656, "step": 12710}, {"loss": 0.6654, "grad_norm": 0.861890971660614, "learning_rate": 0.0002, "epoch": 2.056422277908011, "step": 12720}, {"loss": 0.6286, "grad_norm": 0.6489875912666321, "learning_rate": 0.0002, "epoch": 2.058038962088756, "step": 12730}, {"loss": 0.6811, "grad_norm": 0.8268506526947021, "learning_rate": 0.0002, "epoch": 2.0596556462695013, "step": 12740}, {"loss": 0.6524, "grad_norm": 0.607679545879364, "learning_rate": 0.0002, "epoch": 2.0612723304502465, "step": 12750}, {"loss": 0.6649, "grad_norm": 0.6754153370857239, "learning_rate": 0.0002, "epoch": 2.0628890146309917, "step": 12760}, {"loss": 0.6549, "grad_norm": 0.7263124585151672, "learning_rate": 0.0002, "epoch": 2.064505698811737, "step": 12770}, {"loss": 0.6189, "grad_norm": 0.6986154317855835, "learning_rate": 0.0002, "epoch": 2.0661223829924826, "step": 12780}, {"loss": 0.6723, "grad_norm": 0.7768576741218567, "learning_rate": 0.0002, "epoch": 2.067739067173228, "step": 12790}, {"loss": 0.677, "grad_norm": 0.7546762824058533, "learning_rate": 0.0002, "epoch": 2.069355751353973, "step": 12800}, {"loss": 0.6485, "grad_norm": 0.7588880062103271, "learning_rate": 0.0002, "epoch": 2.0709724355347183, "step": 12810}, {"loss": 0.6989, "grad_norm": 0.7457242608070374, "learning_rate": 0.0002, "epoch": 2.0725891197154636, "step": 12820}, {"loss": 0.6489, "grad_norm": 0.6983516812324524, "learning_rate": 0.0002, "epoch": 2.074205803896209, "step": 12830}, {"loss": 0.651, "grad_norm": 0.7950928807258606, "learning_rate": 0.0002, "epoch": 2.075822488076954, "step": 12840}, {"loss": 0.6603, "grad_norm": 0.9248087406158447, "learning_rate": 0.0002, "epoch": 2.0774391722576993, "step": 12850}, {"loss": 0.6847, "grad_norm": 0.7229493260383606, "learning_rate": 0.0002, "epoch": 2.079055856438445, "step": 12860}, {"loss": 0.6702, "grad_norm": 0.5710847973823547, "learning_rate": 0.0002, "epoch": 2.08067254061919, "step": 12870}, {"loss": 0.6974, "grad_norm": 0.9580423831939697, "learning_rate": 0.0002, "epoch": 2.0822892247999354, "step": 12880}, {"loss": 0.6341, "grad_norm": 0.7399665713310242, "learning_rate": 0.0002, "epoch": 2.0839059089806806, "step": 12890}, {"loss": 0.6993, "grad_norm": 0.7981410622596741, "learning_rate": 0.0002, "epoch": 2.085522593161426, "step": 12900}, {"loss": 0.6976, "grad_norm": 0.870759904384613, "learning_rate": 0.0002, "epoch": 2.087139277342171, "step": 12910}, {"loss": 0.7194, "grad_norm": 0.7001481652259827, "learning_rate": 0.0002, "epoch": 2.0887559615229163, "step": 12920}, {"loss": 0.6383, "grad_norm": 0.6745418310165405, "learning_rate": 0.0002, "epoch": 2.090372645703662, "step": 12930}, {"loss": 0.6519, "grad_norm": 0.7739067673683167, "learning_rate": 0.0002, "epoch": 2.0919893298844072, "step": 12940}, {"loss": 0.6856, "grad_norm": 0.6742934584617615, "learning_rate": 0.0002, "epoch": 2.0936060140651525, "step": 12950}, {"loss": 0.6279, "grad_norm": 0.7270349860191345, "learning_rate": 0.0002, "epoch": 2.0952226982458977, "step": 12960}, {"loss": 0.6783, "grad_norm": 0.7150624394416809, "learning_rate": 0.0002, "epoch": 2.096839382426643, "step": 12970}, {"loss": 0.6093, "grad_norm": 0.7734767198562622, "learning_rate": 0.0002, "epoch": 2.098456066607388, "step": 12980}, {"loss": 0.6534, "grad_norm": 0.7618662118911743, "learning_rate": 0.0002, "epoch": 2.1000727507881334, "step": 12990}, {"loss": 0.6707, "grad_norm": 0.6557944416999817, "learning_rate": 0.0002, "epoch": 2.101689434968879, "step": 13000}, {"loss": 0.7268, "grad_norm": 0.8786448240280151, "learning_rate": 0.0002, "epoch": 2.1033061191496243, "step": 13010}, {"loss": 0.6677, "grad_norm": 0.6878724098205566, "learning_rate": 0.0002, "epoch": 2.1049228033303695, "step": 13020}, {"loss": 0.6824, "grad_norm": 0.822318971157074, "learning_rate": 0.0002, "epoch": 2.1065394875111147, "step": 13030}, {"loss": 0.6228, "grad_norm": 0.831468939781189, "learning_rate": 0.0002, "epoch": 2.10815617169186, "step": 13040}, {"loss": 0.6511, "grad_norm": 0.7699505686759949, "learning_rate": 0.0002, "epoch": 2.109772855872605, "step": 13050}, {"loss": 0.6671, "grad_norm": 0.7559016346931458, "learning_rate": 0.0002, "epoch": 2.1113895400533504, "step": 13060}, {"loss": 0.6215, "grad_norm": 0.6942209601402283, "learning_rate": 0.0002, "epoch": 2.1130062242340957, "step": 13070}, {"loss": 0.6449, "grad_norm": 0.6098947525024414, "learning_rate": 0.0002, "epoch": 2.1146229084148414, "step": 13080}, {"loss": 0.7091, "grad_norm": 0.6499016284942627, "learning_rate": 0.0002, "epoch": 2.1162395925955866, "step": 13090}, {"loss": 0.6247, "grad_norm": 0.7719953060150146, "learning_rate": 0.0002, "epoch": 2.117856276776332, "step": 13100}, {"loss": 0.6064, "grad_norm": 0.6708134412765503, "learning_rate": 0.0002, "epoch": 2.119472960957077, "step": 13110}, {"loss": 0.6056, "grad_norm": 0.8119585514068604, "learning_rate": 0.0002, "epoch": 2.1210896451378223, "step": 13120}, {"loss": 0.6628, "grad_norm": 0.6947157979011536, "learning_rate": 0.0002, "epoch": 2.1227063293185675, "step": 13130}, {"loss": 0.6375, "grad_norm": 0.8831837773323059, "learning_rate": 0.0002, "epoch": 2.1243230134993127, "step": 13140}, {"loss": 0.6997, "grad_norm": 0.7266910672187805, "learning_rate": 0.0002, "epoch": 2.1259396976800584, "step": 13150}, {"loss": 0.6446, "grad_norm": 0.8864351511001587, "learning_rate": 0.0002, "epoch": 2.1275563818608036, "step": 13160}, {"loss": 0.6762, "grad_norm": 0.8104248046875, "learning_rate": 0.0002, "epoch": 2.129173066041549, "step": 13170}, {"loss": 0.6581, "grad_norm": 0.6077079772949219, "learning_rate": 0.0002, "epoch": 2.130789750222294, "step": 13180}, {"loss": 0.6572, "grad_norm": 0.6874213814735413, "learning_rate": 0.0002, "epoch": 2.1324064344030393, "step": 13190}, {"loss": 0.642, "grad_norm": 0.7134367823600769, "learning_rate": 0.0002, "epoch": 2.1340231185837846, "step": 13200}, {"loss": 0.7016, "grad_norm": 0.6101235151290894, "learning_rate": 0.0002, "epoch": 2.13563980276453, "step": 13210}, {"loss": 0.6529, "grad_norm": 0.6042411923408508, "learning_rate": 0.0002, "epoch": 2.137256486945275, "step": 13220}, {"loss": 0.7179, "grad_norm": 0.914601743221283, "learning_rate": 0.0002, "epoch": 2.1388731711260207, "step": 13230}, {"loss": 0.6513, "grad_norm": 0.7104284167289734, "learning_rate": 0.0002, "epoch": 2.140489855306766, "step": 13240}, {"loss": 0.6607, "grad_norm": 0.664395272731781, "learning_rate": 0.0002, "epoch": 2.142106539487511, "step": 13250}, {"loss": 0.7211, "grad_norm": 0.6991241574287415, "learning_rate": 0.0002, "epoch": 2.1437232236682564, "step": 13260}, {"loss": 0.6484, "grad_norm": 0.5469560623168945, "learning_rate": 0.0002, "epoch": 2.1453399078490016, "step": 13270}, {"loss": 0.6765, "grad_norm": 0.8454998135566711, "learning_rate": 0.0002, "epoch": 2.146956592029747, "step": 13280}, {"loss": 0.6683, "grad_norm": 0.7088868618011475, "learning_rate": 0.0002, "epoch": 2.148573276210492, "step": 13290}, {"loss": 0.6835, "grad_norm": 0.7002687454223633, "learning_rate": 0.0002, "epoch": 2.1501899603912378, "step": 13300}, {"loss": 0.6399, "grad_norm": 0.7785214781761169, "learning_rate": 0.0002, "epoch": 2.151806644571983, "step": 13310}, {"loss": 0.67, "grad_norm": 0.8049132227897644, "learning_rate": 0.0002, "epoch": 2.1534233287527282, "step": 13320}, {"loss": 0.6495, "grad_norm": 0.8062595129013062, "learning_rate": 0.0002, "epoch": 2.1550400129334735, "step": 13330}, {"loss": 0.6603, "grad_norm": 0.6208319067955017, "learning_rate": 0.0002, "epoch": 2.1566566971142187, "step": 13340}, {"loss": 0.6584, "grad_norm": 0.7519655823707581, "learning_rate": 0.0002, "epoch": 2.158273381294964, "step": 13350}, {"loss": 0.6457, "grad_norm": 0.7645747065544128, "learning_rate": 0.0002, "epoch": 2.159890065475709, "step": 13360}, {"loss": 0.645, "grad_norm": 0.6847302913665771, "learning_rate": 0.0002, "epoch": 2.1615067496564544, "step": 13370}, {"loss": 0.6903, "grad_norm": 0.8630441427230835, "learning_rate": 0.0002, "epoch": 2.1631234338372, "step": 13380}, {"loss": 0.6742, "grad_norm": 0.7947702407836914, "learning_rate": 0.0002, "epoch": 2.1647401180179453, "step": 13390}, {"loss": 0.7206, "grad_norm": 0.6836977005004883, "learning_rate": 0.0002, "epoch": 2.1663568021986905, "step": 13400}, {"loss": 0.6304, "grad_norm": 0.7340566515922546, "learning_rate": 0.0002, "epoch": 2.1679734863794358, "step": 13410}, {"loss": 0.6528, "grad_norm": 0.7075738906860352, "learning_rate": 0.0002, "epoch": 2.169590170560181, "step": 13420}, {"loss": 0.6585, "grad_norm": 0.7080879807472229, "learning_rate": 0.0002, "epoch": 2.1712068547409262, "step": 13430}, {"loss": 0.6615, "grad_norm": 0.6218613386154175, "learning_rate": 0.0002, "epoch": 2.1728235389216715, "step": 13440}, {"loss": 0.6488, "grad_norm": 0.8211479187011719, "learning_rate": 0.0002, "epoch": 2.174440223102417, "step": 13450}, {"loss": 0.6738, "grad_norm": 0.864466667175293, "learning_rate": 0.0002, "epoch": 2.1760569072831624, "step": 13460}, {"loss": 0.679, "grad_norm": 0.7943857908248901, "learning_rate": 0.0002, "epoch": 2.1776735914639076, "step": 13470}, {"loss": 0.6838, "grad_norm": 0.78728187084198, "learning_rate": 0.0002, "epoch": 2.179290275644653, "step": 13480}, {"loss": 0.6397, "grad_norm": 0.697527289390564, "learning_rate": 0.0002, "epoch": 2.180906959825398, "step": 13490}, {"loss": 0.669, "grad_norm": 0.8205804228782654, "learning_rate": 0.0002, "epoch": 2.1825236440061433, "step": 13500}, {"loss": 0.7227, "grad_norm": 0.8709042072296143, "learning_rate": 0.0002, "epoch": 2.1841403281868885, "step": 13510}, {"loss": 0.6313, "grad_norm": 0.6228537559509277, "learning_rate": 0.0002, "epoch": 2.1857570123676338, "step": 13520}, {"loss": 0.7025, "grad_norm": 0.9566980004310608, "learning_rate": 0.0002, "epoch": 2.1873736965483794, "step": 13530}, {"loss": 0.6755, "grad_norm": 0.7128894329071045, "learning_rate": 0.0002, "epoch": 2.1889903807291247, "step": 13540}, {"loss": 0.6827, "grad_norm": 0.6888654232025146, "learning_rate": 0.0002, "epoch": 2.19060706490987, "step": 13550}, {"loss": 0.6961, "grad_norm": 0.6444337368011475, "learning_rate": 0.0002, "epoch": 2.192223749090615, "step": 13560}, {"loss": 0.656, "grad_norm": 0.8008806705474854, "learning_rate": 0.0002, "epoch": 2.1938404332713604, "step": 13570}, {"loss": 0.7, "grad_norm": 0.8482748866081238, "learning_rate": 0.0002, "epoch": 2.1954571174521056, "step": 13580}, {"loss": 0.7326, "grad_norm": 0.8584157228469849, "learning_rate": 0.0002, "epoch": 2.197073801632851, "step": 13590}, {"loss": 0.7014, "grad_norm": 0.7513734698295593, "learning_rate": 0.0002, "epoch": 2.1986904858135965, "step": 13600}, {"loss": 0.6632, "grad_norm": 0.7864262461662292, "learning_rate": 0.0002, "epoch": 2.2003071699943417, "step": 13610}, {"loss": 0.6879, "grad_norm": 0.8493645191192627, "learning_rate": 0.0002, "epoch": 2.201923854175087, "step": 13620}, {"loss": 0.6617, "grad_norm": 0.6902140974998474, "learning_rate": 0.0002, "epoch": 2.203540538355832, "step": 13630}, {"loss": 0.6655, "grad_norm": 0.8711254596710205, "learning_rate": 0.0002, "epoch": 2.2051572225365774, "step": 13640}, {"loss": 0.6359, "grad_norm": 0.7832191586494446, "learning_rate": 0.0002, "epoch": 2.2067739067173227, "step": 13650}, {"loss": 0.6723, "grad_norm": 0.5668176412582397, "learning_rate": 0.0002, "epoch": 2.208390590898068, "step": 13660}, {"loss": 0.635, "grad_norm": 0.8648375272750854, "learning_rate": 0.0002, "epoch": 2.2100072750788136, "step": 13670}, {"loss": 0.653, "grad_norm": 0.7643089890480042, "learning_rate": 0.0002, "epoch": 2.211623959259559, "step": 13680}, {"loss": 0.6765, "grad_norm": 0.6293777823448181, "learning_rate": 0.0002, "epoch": 2.213240643440304, "step": 13690}, {"loss": 0.6842, "grad_norm": 0.6459372639656067, "learning_rate": 0.0002, "epoch": 2.2148573276210493, "step": 13700}, {"loss": 0.6526, "grad_norm": 0.7060744166374207, "learning_rate": 0.0002, "epoch": 2.2164740118017945, "step": 13710}, {"loss": 0.7101, "grad_norm": 0.674109160900116, "learning_rate": 0.0002, "epoch": 2.2180906959825397, "step": 13720}, {"loss": 0.6529, "grad_norm": 0.830392062664032, "learning_rate": 0.0002, "epoch": 2.219707380163285, "step": 13730}, {"loss": 0.6733, "grad_norm": 0.6474477052688599, "learning_rate": 0.0002, "epoch": 2.2213240643440306, "step": 13740}, {"loss": 0.6413, "grad_norm": 0.7037909626960754, "learning_rate": 0.0002, "epoch": 2.222940748524776, "step": 13750}, {"loss": 0.6417, "grad_norm": 0.6554131507873535, "learning_rate": 0.0002, "epoch": 2.224557432705521, "step": 13760}, {"loss": 0.6907, "grad_norm": 0.7822230458259583, "learning_rate": 0.0002, "epoch": 2.2261741168862663, "step": 13770}, {"loss": 0.6505, "grad_norm": 0.9082167744636536, "learning_rate": 0.0002, "epoch": 2.2277908010670116, "step": 13780}, {"loss": 0.6878, "grad_norm": 0.7918276190757751, "learning_rate": 0.0002, "epoch": 2.229407485247757, "step": 13790}, {"loss": 0.6669, "grad_norm": 0.7354569435119629, "learning_rate": 0.0002, "epoch": 2.231024169428502, "step": 13800}, {"loss": 0.6503, "grad_norm": 0.8265249133110046, "learning_rate": 0.0002, "epoch": 2.2326408536092472, "step": 13810}, {"loss": 0.6871, "grad_norm": 0.6653847098350525, "learning_rate": 0.0002, "epoch": 2.234257537789993, "step": 13820}, {"loss": 0.6413, "grad_norm": 0.7157923579216003, "learning_rate": 0.0002, "epoch": 2.235874221970738, "step": 13830}, {"loss": 0.6306, "grad_norm": 0.7110323309898376, "learning_rate": 0.0002, "epoch": 2.2374909061514834, "step": 13840}, {"loss": 0.6913, "grad_norm": 0.7155357599258423, "learning_rate": 0.0002, "epoch": 2.2391075903322286, "step": 13850}, {"loss": 0.6579, "grad_norm": 1.0177817344665527, "learning_rate": 0.0002, "epoch": 2.240724274512974, "step": 13860}, {"loss": 0.635, "grad_norm": 0.7601948380470276, "learning_rate": 0.0002, "epoch": 2.242340958693719, "step": 13870}, {"loss": 0.6679, "grad_norm": 0.7628820538520813, "learning_rate": 0.0002, "epoch": 2.2439576428744643, "step": 13880}, {"loss": 0.6805, "grad_norm": 0.7089297771453857, "learning_rate": 0.0002, "epoch": 2.24557432705521, "step": 13890}, {"loss": 0.7236, "grad_norm": 0.695178210735321, "learning_rate": 0.0002, "epoch": 2.247191011235955, "step": 13900}, {"loss": 0.7084, "grad_norm": 0.7631948590278625, "learning_rate": 0.0002, "epoch": 2.2488076954167004, "step": 13910}, {"loss": 0.685, "grad_norm": 0.8203101754188538, "learning_rate": 0.0002, "epoch": 2.2504243795974457, "step": 13920}, {"loss": 0.653, "grad_norm": 0.8099079728126526, "learning_rate": 0.0002, "epoch": 2.252041063778191, "step": 13930}, {"loss": 0.694, "grad_norm": 0.6498546004295349, "learning_rate": 0.0002, "epoch": 2.253657747958936, "step": 13940}, {"loss": 0.6684, "grad_norm": 0.7797415256500244, "learning_rate": 0.0002, "epoch": 2.2552744321396814, "step": 13950}, {"loss": 0.683, "grad_norm": 0.8254124522209167, "learning_rate": 0.0002, "epoch": 2.2568911163204266, "step": 13960}, {"loss": 0.6806, "grad_norm": 0.6327953338623047, "learning_rate": 0.0002, "epoch": 2.2585078005011723, "step": 13970}, {"loss": 0.668, "grad_norm": 0.734194278717041, "learning_rate": 0.0002, "epoch": 2.2601244846819175, "step": 13980}, {"loss": 0.6912, "grad_norm": 0.9014202952384949, "learning_rate": 0.0002, "epoch": 2.2617411688626627, "step": 13990}, {"loss": 0.692, "grad_norm": 0.7643631100654602, "learning_rate": 0.0002, "epoch": 2.263357853043408, "step": 14000}, {"loss": 0.6657, "grad_norm": 0.8882834911346436, "learning_rate": 0.0002, "epoch": 2.264974537224153, "step": 14010}, {"loss": 0.6453, "grad_norm": 0.7975873351097107, "learning_rate": 0.0002, "epoch": 2.2665912214048984, "step": 14020}, {"loss": 0.7193, "grad_norm": 0.7765783071517944, "learning_rate": 0.0002, "epoch": 2.2682079055856437, "step": 14030}, {"loss": 0.662, "grad_norm": 0.8846288323402405, "learning_rate": 0.0002, "epoch": 2.2698245897663893, "step": 14040}, {"loss": 0.6494, "grad_norm": 0.9006744027137756, "learning_rate": 0.0002, "epoch": 2.2714412739471346, "step": 14050}, {"loss": 0.6423, "grad_norm": 0.7420173287391663, "learning_rate": 0.0002, "epoch": 2.27305795812788, "step": 14060}, {"loss": 0.7068, "grad_norm": 0.7956424951553345, "learning_rate": 0.0002, "epoch": 2.274674642308625, "step": 14070}, {"loss": 0.6581, "grad_norm": 0.7783209085464478, "learning_rate": 0.0002, "epoch": 2.2762913264893703, "step": 14080}, {"loss": 0.7202, "grad_norm": 0.7597188949584961, "learning_rate": 0.0002, "epoch": 2.2779080106701155, "step": 14090}, {"loss": 0.6778, "grad_norm": 0.6718921661376953, "learning_rate": 0.0002, "epoch": 2.2795246948508607, "step": 14100}, {"loss": 0.632, "grad_norm": 0.7528082132339478, "learning_rate": 0.0002, "epoch": 2.281141379031606, "step": 14110}, {"loss": 0.7608, "grad_norm": 0.8379864692687988, "learning_rate": 0.0002, "epoch": 2.2827580632123516, "step": 14120}, {"loss": 0.6767, "grad_norm": 0.748613715171814, "learning_rate": 0.0002, "epoch": 2.284374747393097, "step": 14130}, {"loss": 0.6641, "grad_norm": 0.7435423135757446, "learning_rate": 0.0002, "epoch": 2.285991431573842, "step": 14140}, {"loss": 0.6849, "grad_norm": 0.7580803632736206, "learning_rate": 0.0002, "epoch": 2.2876081157545873, "step": 14150}, {"loss": 0.6604, "grad_norm": 0.6278321146965027, "learning_rate": 0.0002, "epoch": 2.2892247999353326, "step": 14160}, {"loss": 0.6573, "grad_norm": 0.7663896083831787, "learning_rate": 0.0002, "epoch": 2.290841484116078, "step": 14170}, {"loss": 0.6655, "grad_norm": 0.9716812372207642, "learning_rate": 0.0002, "epoch": 2.292458168296823, "step": 14180}, {"loss": 0.7067, "grad_norm": 0.8993458151817322, "learning_rate": 0.0002, "epoch": 2.2940748524775687, "step": 14190}, {"loss": 0.6172, "grad_norm": 0.6156117916107178, "learning_rate": 0.0002, "epoch": 2.295691536658314, "step": 14200}, {"loss": 0.6318, "grad_norm": 0.8911278247833252, "learning_rate": 0.0002, "epoch": 2.297308220839059, "step": 14210}, {"loss": 0.6364, "grad_norm": 0.6422147154808044, "learning_rate": 0.0002, "epoch": 2.2989249050198044, "step": 14220}, {"loss": 0.6795, "grad_norm": 0.6866879463195801, "learning_rate": 0.0002, "epoch": 2.3005415892005496, "step": 14230}, {"loss": 0.6907, "grad_norm": 0.9297130107879639, "learning_rate": 0.0002, "epoch": 2.302158273381295, "step": 14240}, {"loss": 0.6823, "grad_norm": 0.7501356601715088, "learning_rate": 0.0002, "epoch": 2.30377495756204, "step": 14250}, {"loss": 0.6414, "grad_norm": 0.8363515138626099, "learning_rate": 0.0002, "epoch": 2.3053916417427853, "step": 14260}, {"loss": 0.6362, "grad_norm": 0.9083868265151978, "learning_rate": 0.0002, "epoch": 2.307008325923531, "step": 14270}, {"loss": 0.6862, "grad_norm": 0.7791516780853271, "learning_rate": 0.0002, "epoch": 2.3086250101042762, "step": 14280}, {"loss": 0.6569, "grad_norm": 0.8766953349113464, "learning_rate": 0.0002, "epoch": 2.3102416942850215, "step": 14290}, {"loss": 0.6698, "grad_norm": 0.7916635274887085, "learning_rate": 0.0002, "epoch": 2.3118583784657667, "step": 14300}, {"loss": 0.6927, "grad_norm": 0.627525269985199, "learning_rate": 0.0002, "epoch": 2.313475062646512, "step": 14310}, {"loss": 0.6541, "grad_norm": 0.8856783509254456, "learning_rate": 0.0002, "epoch": 2.315091746827257, "step": 14320}, {"loss": 0.6806, "grad_norm": 0.6758689284324646, "learning_rate": 0.0002, "epoch": 2.316708431008003, "step": 14330}, {"loss": 0.6794, "grad_norm": 0.6428321003913879, "learning_rate": 0.0002, "epoch": 2.318325115188748, "step": 14340}, {"loss": 0.682, "grad_norm": 0.9032121300697327, "learning_rate": 0.0002, "epoch": 2.3199417993694933, "step": 14350}, {"loss": 0.6569, "grad_norm": 0.8035986423492432, "learning_rate": 0.0002, "epoch": 2.3215584835502385, "step": 14360}, {"loss": 0.7067, "grad_norm": 0.7974579334259033, "learning_rate": 0.0002, "epoch": 2.3231751677309838, "step": 14370}, {"loss": 0.6451, "grad_norm": 0.8356034755706787, "learning_rate": 0.0002, "epoch": 2.324791851911729, "step": 14380}, {"loss": 0.6623, "grad_norm": 0.998760998249054, "learning_rate": 0.0002, "epoch": 2.326408536092474, "step": 14390}, {"loss": 0.649, "grad_norm": 0.6518142223358154, "learning_rate": 0.0002, "epoch": 2.3280252202732195, "step": 14400}, {"loss": 0.7146, "grad_norm": 0.7443506717681885, "learning_rate": 0.0002, "epoch": 2.3296419044539647, "step": 14410}, {"loss": 0.648, "grad_norm": 0.8436172604560852, "learning_rate": 0.0002, "epoch": 2.3312585886347104, "step": 14420}, {"loss": 0.6585, "grad_norm": 0.7411080598831177, "learning_rate": 0.0002, "epoch": 2.3328752728154556, "step": 14430}, {"loss": 0.6781, "grad_norm": 0.8839048743247986, "learning_rate": 0.0002, "epoch": 2.334491956996201, "step": 14440}, {"loss": 0.6565, "grad_norm": 0.8360885977745056, "learning_rate": 0.0002, "epoch": 2.336108641176946, "step": 14450}, {"loss": 0.6662, "grad_norm": 0.7608986496925354, "learning_rate": 0.0002, "epoch": 2.3377253253576913, "step": 14460}, {"loss": 0.6685, "grad_norm": 0.8179867267608643, "learning_rate": 0.0002, "epoch": 2.3393420095384365, "step": 14470}, {"loss": 0.7055, "grad_norm": 0.5989999771118164, "learning_rate": 0.0002, "epoch": 2.340958693719182, "step": 14480}, {"loss": 0.644, "grad_norm": 0.9450054168701172, "learning_rate": 0.0002, "epoch": 2.3425753778999274, "step": 14490}, {"loss": 0.6983, "grad_norm": 0.7885149717330933, "learning_rate": 0.0002, "epoch": 2.3441920620806727, "step": 14500}, {"loss": 0.6819, "grad_norm": 0.8152616620063782, "learning_rate": 0.0002, "epoch": 2.345808746261418, "step": 14510}, {"loss": 0.6989, "grad_norm": 0.7193838953971863, "learning_rate": 0.0002, "epoch": 2.347425430442163, "step": 14520}, {"loss": 0.6594, "grad_norm": 0.6701092720031738, "learning_rate": 0.0002, "epoch": 2.3490421146229084, "step": 14530}, {"loss": 0.6559, "grad_norm": 0.7529364228248596, "learning_rate": 0.0002, "epoch": 2.3506587988036536, "step": 14540}, {"loss": 0.6306, "grad_norm": 0.6599733829498291, "learning_rate": 0.0002, "epoch": 2.352275482984399, "step": 14550}, {"loss": 0.706, "grad_norm": 0.9502474069595337, "learning_rate": 0.0002, "epoch": 2.353892167165144, "step": 14560}, {"loss": 0.717, "grad_norm": 0.7619650959968567, "learning_rate": 0.0002, "epoch": 2.3555088513458897, "step": 14570}, {"loss": 0.6684, "grad_norm": 0.9854652285575867, "learning_rate": 0.0002, "epoch": 2.357125535526635, "step": 14580}, {"loss": 0.6455, "grad_norm": 0.727439284324646, "learning_rate": 0.0002, "epoch": 2.35874221970738, "step": 14590}, {"loss": 0.6645, "grad_norm": 0.6994746327400208, "learning_rate": 0.0002, "epoch": 2.3603589038881254, "step": 14600}, {"loss": 0.6587, "grad_norm": 0.7117531299591064, "learning_rate": 0.0002, "epoch": 2.3619755880688706, "step": 14610}, {"loss": 0.6804, "grad_norm": 0.6403067708015442, "learning_rate": 0.0002, "epoch": 2.363592272249616, "step": 14620}, {"loss": 0.7055, "grad_norm": 0.8377841711044312, "learning_rate": 0.0002, "epoch": 2.3652089564303616, "step": 14630}, {"loss": 0.6778, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 2.366825640611107, "step": 14640}, {"loss": 0.6552, "grad_norm": 0.8418586254119873, "learning_rate": 0.0002, "epoch": 2.368442324791852, "step": 14650}, {"loss": 0.6685, "grad_norm": 0.6178573369979858, "learning_rate": 0.0002, "epoch": 2.3700590089725972, "step": 14660}, {"loss": 0.6774, "grad_norm": 0.6368302702903748, "learning_rate": 0.0002, "epoch": 2.3716756931533425, "step": 14670}, {"loss": 0.6136, "grad_norm": 0.9122977256774902, "learning_rate": 0.0002, "epoch": 2.3732923773340877, "step": 14680}, {"loss": 0.6675, "grad_norm": 0.7086195349693298, "learning_rate": 0.0002, "epoch": 2.374909061514833, "step": 14690}, {"loss": 0.6582, "grad_norm": 0.7500800490379333, "learning_rate": 0.0002, "epoch": 2.376525745695578, "step": 14700}, {"loss": 0.6792, "grad_norm": 0.6634900569915771, "learning_rate": 0.0002, "epoch": 2.378142429876324, "step": 14710}, {"loss": 0.6614, "grad_norm": 0.839898407459259, "learning_rate": 0.0002, "epoch": 2.379759114057069, "step": 14720}, {"loss": 0.6453, "grad_norm": 0.7578426003456116, "learning_rate": 0.0002, "epoch": 2.3813757982378143, "step": 14730}, {"loss": 0.7282, "grad_norm": 1.0213173627853394, "learning_rate": 0.0002, "epoch": 2.3829924824185595, "step": 14740}, {"loss": 0.6704, "grad_norm": 0.7855949401855469, "learning_rate": 0.0002, "epoch": 2.3846091665993048, "step": 14750}, {"loss": 0.6694, "grad_norm": 0.7224128842353821, "learning_rate": 0.0002, "epoch": 2.38622585078005, "step": 14760}, {"loss": 0.7017, "grad_norm": 0.8040381669998169, "learning_rate": 0.0002, "epoch": 2.3878425349607952, "step": 14770}, {"loss": 0.6799, "grad_norm": 0.7705281376838684, "learning_rate": 0.0002, "epoch": 2.389459219141541, "step": 14780}, {"loss": 0.6326, "grad_norm": 0.667966902256012, "learning_rate": 0.0002, "epoch": 2.391075903322286, "step": 14790}, {"loss": 0.7061, "grad_norm": 0.6611011028289795, "learning_rate": 0.0002, "epoch": 2.3926925875030314, "step": 14800}, {"loss": 0.6527, "grad_norm": 0.6862651705741882, "learning_rate": 0.0002, "epoch": 2.3943092716837766, "step": 14810}, {"loss": 0.6537, "grad_norm": 0.8086010217666626, "learning_rate": 0.0002, "epoch": 2.395925955864522, "step": 14820}, {"loss": 0.7189, "grad_norm": 0.7189689874649048, "learning_rate": 0.0002, "epoch": 2.397542640045267, "step": 14830}, {"loss": 0.6709, "grad_norm": 0.6280009150505066, "learning_rate": 0.0002, "epoch": 2.3991593242260123, "step": 14840}, {"loss": 0.706, "grad_norm": 0.7826612591743469, "learning_rate": 0.0002, "epoch": 2.4007760084067575, "step": 14850}, {"loss": 0.6738, "grad_norm": 0.7681610584259033, "learning_rate": 0.0002, "epoch": 2.402392692587503, "step": 14860}, {"loss": 0.636, "grad_norm": 0.720966100692749, "learning_rate": 0.0002, "epoch": 2.4040093767682484, "step": 14870}, {"loss": 0.6667, "grad_norm": 0.8202250599861145, "learning_rate": 0.0002, "epoch": 2.4056260609489937, "step": 14880}, {"loss": 0.6935, "grad_norm": 0.786212682723999, "learning_rate": 0.0002, "epoch": 2.407242745129739, "step": 14890}, {"loss": 0.6628, "grad_norm": 0.6647164821624756, "learning_rate": 0.0002, "epoch": 2.408859429310484, "step": 14900}, {"loss": 0.6706, "grad_norm": 0.7566399574279785, "learning_rate": 0.0002, "epoch": 2.4104761134912294, "step": 14910}, {"loss": 0.7188, "grad_norm": 0.748814582824707, "learning_rate": 0.0002, "epoch": 2.4120927976719746, "step": 14920}, {"loss": 0.6684, "grad_norm": 0.7624038457870483, "learning_rate": 0.0002, "epoch": 2.4137094818527203, "step": 14930}, {"loss": 0.6483, "grad_norm": 0.8267335295677185, "learning_rate": 0.0002, "epoch": 2.4153261660334655, "step": 14940}, {"loss": 0.6612, "grad_norm": 0.8785360455513, "learning_rate": 0.0002, "epoch": 2.4169428502142107, "step": 14950}, {"loss": 0.6718, "grad_norm": 0.679887592792511, "learning_rate": 0.0002, "epoch": 2.418559534394956, "step": 14960}, {"loss": 0.6136, "grad_norm": 0.7218474745750427, "learning_rate": 0.0002, "epoch": 2.420176218575701, "step": 14970}, {"loss": 0.648, "grad_norm": 0.6342799663543701, "learning_rate": 0.0002, "epoch": 2.4217929027564464, "step": 14980}, {"loss": 0.6617, "grad_norm": 0.7098712921142578, "learning_rate": 0.0002, "epoch": 2.4234095869371917, "step": 14990}, {"loss": 0.6942, "grad_norm": 0.7497431635856628, "learning_rate": 0.0002, "epoch": 2.425026271117937, "step": 15000}, {"loss": 0.6772, "grad_norm": 0.934836208820343, "learning_rate": 0.0002, "epoch": 2.4266429552986826, "step": 15010}, {"loss": 0.7221, "grad_norm": 0.8430966734886169, "learning_rate": 0.0002, "epoch": 2.428259639479428, "step": 15020}, {"loss": 0.6985, "grad_norm": 0.7032104730606079, "learning_rate": 0.0002, "epoch": 2.429876323660173, "step": 15030}, {"loss": 0.6715, "grad_norm": 0.7746111750602722, "learning_rate": 0.0002, "epoch": 2.4314930078409183, "step": 15040}, {"loss": 0.7177, "grad_norm": 0.7661406397819519, "learning_rate": 0.0002, "epoch": 2.4331096920216635, "step": 15050}, {"loss": 0.6517, "grad_norm": 0.6941645741462708, "learning_rate": 0.0002, "epoch": 2.4347263762024087, "step": 15060}, {"loss": 0.6421, "grad_norm": 0.7487249374389648, "learning_rate": 0.0002, "epoch": 2.436343060383154, "step": 15070}, {"loss": 0.6796, "grad_norm": 0.7639912962913513, "learning_rate": 0.0002, "epoch": 2.4379597445638996, "step": 15080}, {"loss": 0.7087, "grad_norm": 0.7708953619003296, "learning_rate": 0.0002, "epoch": 2.439576428744645, "step": 15090}, {"loss": 0.7065, "grad_norm": 0.9135832190513611, "learning_rate": 0.0002, "epoch": 2.44119311292539, "step": 15100}, {"loss": 0.672, "grad_norm": 0.8283005356788635, "learning_rate": 0.0002, "epoch": 2.4428097971061353, "step": 15110}, {"loss": 0.6551, "grad_norm": 0.925299346446991, "learning_rate": 0.0002, "epoch": 2.4444264812868806, "step": 15120}, {"loss": 0.687, "grad_norm": 0.7013528943061829, "learning_rate": 0.0002, "epoch": 2.446043165467626, "step": 15130}, {"loss": 0.6842, "grad_norm": 0.622303307056427, "learning_rate": 0.0002, "epoch": 2.447659849648371, "step": 15140}, {"loss": 0.6676, "grad_norm": 0.876569390296936, "learning_rate": 0.0002, "epoch": 2.4492765338291163, "step": 15150}, {"loss": 0.6463, "grad_norm": 0.6836351752281189, "learning_rate": 0.0002, "epoch": 2.450893218009862, "step": 15160}, {"loss": 0.6781, "grad_norm": 0.7886684536933899, "learning_rate": 0.0002, "epoch": 2.452509902190607, "step": 15170}, {"loss": 0.6794, "grad_norm": 0.6647440791130066, "learning_rate": 0.0002, "epoch": 2.4541265863713524, "step": 15180}, {"loss": 0.6353, "grad_norm": 0.7477722764015198, "learning_rate": 0.0002, "epoch": 2.4557432705520976, "step": 15190}, {"loss": 0.698, "grad_norm": 0.8192033767700195, "learning_rate": 0.0002, "epoch": 2.457359954732843, "step": 15200}, {"loss": 0.6735, "grad_norm": 0.847537100315094, "learning_rate": 0.0002, "epoch": 2.458976638913588, "step": 15210}, {"loss": 0.6962, "grad_norm": 0.9027776122093201, "learning_rate": 0.0002, "epoch": 2.4605933230943338, "step": 15220}, {"loss": 0.7084, "grad_norm": 0.7217772006988525, "learning_rate": 0.0002, "epoch": 2.462210007275079, "step": 15230}, {"loss": 0.691, "grad_norm": 0.7994546294212341, "learning_rate": 0.0002, "epoch": 2.4638266914558242, "step": 15240}, {"loss": 0.6828, "grad_norm": 0.939916729927063, "learning_rate": 0.0002, "epoch": 2.4654433756365695, "step": 15250}, {"loss": 0.6893, "grad_norm": 1.0009053945541382, "learning_rate": 0.0002, "epoch": 2.4670600598173147, "step": 15260}, {"loss": 0.643, "grad_norm": 0.625555694103241, "learning_rate": 0.0002, "epoch": 2.46867674399806, "step": 15270}, {"loss": 0.688, "grad_norm": 0.7924878597259521, "learning_rate": 0.0002, "epoch": 2.470293428178805, "step": 15280}, {"loss": 0.6789, "grad_norm": 0.8536689877510071, "learning_rate": 0.0002, "epoch": 2.4719101123595504, "step": 15290}, {"loss": 0.6924, "grad_norm": 0.8572589755058289, "learning_rate": 0.0002, "epoch": 2.4735267965402956, "step": 15300}, {"loss": 0.604, "grad_norm": 0.773279070854187, "learning_rate": 0.0002, "epoch": 2.4751434807210413, "step": 15310}, {"loss": 0.6573, "grad_norm": 0.7708749771118164, "learning_rate": 0.0002, "epoch": 2.4767601649017865, "step": 15320}, {"loss": 0.7065, "grad_norm": 0.770905077457428, "learning_rate": 0.0002, "epoch": 2.4783768490825318, "step": 15330}, {"loss": 0.6878, "grad_norm": 0.8238571882247925, "learning_rate": 0.0002, "epoch": 2.479993533263277, "step": 15340}, {"loss": 0.6772, "grad_norm": 0.7670477032661438, "learning_rate": 0.0002, "epoch": 2.481610217444022, "step": 15350}, {"loss": 0.7759, "grad_norm": 0.905036985874176, "learning_rate": 0.0002, "epoch": 2.4832269016247674, "step": 15360}, {"loss": 0.706, "grad_norm": 0.6672089695930481, "learning_rate": 0.0002, "epoch": 2.484843585805513, "step": 15370}, {"loss": 0.6722, "grad_norm": 0.625095784664154, "learning_rate": 0.0002, "epoch": 2.4864602699862584, "step": 15380}, {"loss": 0.6396, "grad_norm": 0.679772675037384, "learning_rate": 0.0002, "epoch": 2.4880769541670036, "step": 15390}, {"loss": 0.6778, "grad_norm": 0.711492121219635, "learning_rate": 0.0002, "epoch": 2.489693638347749, "step": 15400}, {"loss": 0.6966, "grad_norm": 0.876189112663269, "learning_rate": 0.0002, "epoch": 2.491310322528494, "step": 15410}, {"loss": 0.7307, "grad_norm": 0.7236915230751038, "learning_rate": 0.0002, "epoch": 2.4929270067092393, "step": 15420}, {"loss": 0.647, "grad_norm": 0.6629832983016968, "learning_rate": 0.0002, "epoch": 2.4945436908899845, "step": 15430}, {"loss": 0.6669, "grad_norm": 0.9756859540939331, "learning_rate": 0.0002, "epoch": 2.4961603750707297, "step": 15440}, {"loss": 0.7559, "grad_norm": 0.6896940469741821, "learning_rate": 0.0002, "epoch": 2.4977770592514754, "step": 15450}, {"loss": 0.6818, "grad_norm": 0.7105149626731873, "learning_rate": 0.0002, "epoch": 2.4993937434322206, "step": 15460}, {"loss": 0.6859, "grad_norm": 0.8374546766281128, "learning_rate": 0.0002, "epoch": 2.501010427612966, "step": 15470}, {"loss": 0.6512, "grad_norm": 0.7320070266723633, "learning_rate": 0.0002, "epoch": 2.502627111793711, "step": 15480}, {"loss": 0.685, "grad_norm": 0.8306367993354797, "learning_rate": 0.0002, "epoch": 2.5042437959744563, "step": 15490}, {"loss": 0.7253, "grad_norm": 0.7472721338272095, "learning_rate": 0.0002, "epoch": 2.5058604801552016, "step": 15500}, {"loss": 0.6699, "grad_norm": 0.6147692203521729, "learning_rate": 0.0002, "epoch": 2.507477164335947, "step": 15510}, {"loss": 0.7158, "grad_norm": 0.7788505554199219, "learning_rate": 0.0002, "epoch": 2.5090938485166925, "step": 15520}, {"loss": 0.6521, "grad_norm": 0.8807527422904968, "learning_rate": 0.0002, "epoch": 2.5107105326974377, "step": 15530}, {"loss": 0.6792, "grad_norm": 0.7521643042564392, "learning_rate": 0.0002, "epoch": 2.512327216878183, "step": 15540}, {"loss": 0.6772, "grad_norm": 0.6900225281715393, "learning_rate": 0.0002, "epoch": 2.513943901058928, "step": 15550}, {"loss": 0.6769, "grad_norm": 0.6601938605308533, "learning_rate": 0.0002, "epoch": 2.5155605852396734, "step": 15560}, {"loss": 0.6648, "grad_norm": 0.8179984092712402, "learning_rate": 0.0002, "epoch": 2.5171772694204186, "step": 15570}, {"loss": 0.7028, "grad_norm": 0.792556881904602, "learning_rate": 0.0002, "epoch": 2.518793953601164, "step": 15580}, {"loss": 0.6464, "grad_norm": 0.7081938982009888, "learning_rate": 0.0002, "epoch": 2.520410637781909, "step": 15590}, {"loss": 0.6691, "grad_norm": 0.8733121156692505, "learning_rate": 0.0002, "epoch": 2.5220273219626543, "step": 15600}, {"loss": 0.6969, "grad_norm": 0.7980992794036865, "learning_rate": 0.0002, "epoch": 2.5236440061434, "step": 15610}, {"loss": 0.7124, "grad_norm": 0.883664071559906, "learning_rate": 0.0002, "epoch": 2.5252606903241452, "step": 15620}, {"loss": 0.7022, "grad_norm": 0.6963341236114502, "learning_rate": 0.0002, "epoch": 2.5268773745048905, "step": 15630}, {"loss": 0.7334, "grad_norm": 0.6433573365211487, "learning_rate": 0.0002, "epoch": 2.5284940586856357, "step": 15640}, {"loss": 0.6889, "grad_norm": 0.8538183569908142, "learning_rate": 0.0002, "epoch": 2.530110742866381, "step": 15650}, {"loss": 0.6841, "grad_norm": 0.9748201370239258, "learning_rate": 0.0002, "epoch": 2.5317274270471266, "step": 15660}, {"loss": 0.6765, "grad_norm": 0.7670575380325317, "learning_rate": 0.0002, "epoch": 2.533344111227872, "step": 15670}, {"loss": 0.6435, "grad_norm": 0.8738890290260315, "learning_rate": 0.0002, "epoch": 2.534960795408617, "step": 15680}, {"loss": 0.6802, "grad_norm": 0.8391636610031128, "learning_rate": 0.0002, "epoch": 2.5365774795893623, "step": 15690}, {"loss": 0.6901, "grad_norm": 0.7239366769790649, "learning_rate": 0.0002, "epoch": 2.5381941637701075, "step": 15700}, {"loss": 0.7011, "grad_norm": 0.8498379588127136, "learning_rate": 0.0002, "epoch": 2.5398108479508528, "step": 15710}, {"loss": 0.6998, "grad_norm": 0.8029484152793884, "learning_rate": 0.0002, "epoch": 2.541427532131598, "step": 15720}, {"loss": 0.6678, "grad_norm": 1.0639333724975586, "learning_rate": 0.0002, "epoch": 2.5430442163123432, "step": 15730}, {"loss": 0.6341, "grad_norm": 0.6401297450065613, "learning_rate": 0.0002, "epoch": 2.5446609004930885, "step": 15740}, {"loss": 0.7196, "grad_norm": 0.7123814821243286, "learning_rate": 0.0002, "epoch": 2.5462775846738337, "step": 15750}, {"loss": 0.654, "grad_norm": 0.7874974608421326, "learning_rate": 0.0002, "epoch": 2.5478942688545794, "step": 15760}, {"loss": 0.6721, "grad_norm": 0.8046808838844299, "learning_rate": 0.0002, "epoch": 2.5495109530353246, "step": 15770}, {"loss": 0.6665, "grad_norm": 0.7888661623001099, "learning_rate": 0.0002, "epoch": 2.55112763721607, "step": 15780}, {"loss": 0.6893, "grad_norm": 0.8445866107940674, "learning_rate": 0.0002, "epoch": 2.552744321396815, "step": 15790}, {"loss": 0.6815, "grad_norm": 0.7475846409797668, "learning_rate": 0.0002, "epoch": 2.5543610055775603, "step": 15800}, {"loss": 0.6711, "grad_norm": 0.7455102801322937, "learning_rate": 0.0002, "epoch": 2.555977689758306, "step": 15810}, {"loss": 0.6932, "grad_norm": 0.8226983547210693, "learning_rate": 0.0002, "epoch": 2.557594373939051, "step": 15820}, {"loss": 0.651, "grad_norm": 0.8920368552207947, "learning_rate": 0.0002, "epoch": 2.5592110581197964, "step": 15830}, {"loss": 0.6297, "grad_norm": 0.8413904905319214, "learning_rate": 0.0002, "epoch": 2.5608277423005417, "step": 15840}, {"loss": 0.7106, "grad_norm": 0.8483649492263794, "learning_rate": 0.0002, "epoch": 2.562444426481287, "step": 15850}, {"loss": 0.6957, "grad_norm": 0.5923284292221069, "learning_rate": 0.0002, "epoch": 2.564061110662032, "step": 15860}, {"loss": 0.6847, "grad_norm": 0.8518726229667664, "learning_rate": 0.0002, "epoch": 2.5656777948427774, "step": 15870}, {"loss": 0.6362, "grad_norm": 0.731235146522522, "learning_rate": 0.0002, "epoch": 2.5672944790235226, "step": 15880}, {"loss": 0.7611, "grad_norm": 0.7517194151878357, "learning_rate": 0.0002, "epoch": 2.568911163204268, "step": 15890}, {"loss": 0.6907, "grad_norm": 0.8378692269325256, "learning_rate": 0.0002, "epoch": 2.5705278473850135, "step": 15900}, {"loss": 0.7055, "grad_norm": 0.843701958656311, "learning_rate": 0.0002, "epoch": 2.5721445315657587, "step": 15910}, {"loss": 0.6882, "grad_norm": 0.7254629731178284, "learning_rate": 0.0002, "epoch": 2.573761215746504, "step": 15920}, {"loss": 0.6872, "grad_norm": 0.8863335847854614, "learning_rate": 0.0002, "epoch": 2.575377899927249, "step": 15930}, {"loss": 0.6813, "grad_norm": 0.7675097584724426, "learning_rate": 0.0002, "epoch": 2.5769945841079944, "step": 15940}, {"loss": 0.7357, "grad_norm": 0.82063889503479, "learning_rate": 0.0002, "epoch": 2.5786112682887397, "step": 15950}, {"loss": 0.662, "grad_norm": 0.7729717493057251, "learning_rate": 0.0002, "epoch": 2.5802279524694853, "step": 15960}, {"loss": 0.633, "grad_norm": 0.8301846981048584, "learning_rate": 0.0002, "epoch": 2.5818446366502306, "step": 15970}, {"loss": 0.6897, "grad_norm": 0.7906861305236816, "learning_rate": 0.0002, "epoch": 2.583461320830976, "step": 15980}, {"loss": 0.7175, "grad_norm": 0.6749057173728943, "learning_rate": 0.0002, "epoch": 2.585078005011721, "step": 15990}, {"loss": 0.7212, "grad_norm": 0.9386842846870422, "learning_rate": 0.0002, "epoch": 2.5866946891924663, "step": 16000}, {"loss": 0.6934, "grad_norm": 0.7868891358375549, "learning_rate": 0.0002, "epoch": 2.5883113733732115, "step": 16010}, {"loss": 0.7036, "grad_norm": 0.8674671053886414, "learning_rate": 0.0002, "epoch": 2.5899280575539567, "step": 16020}, {"loss": 0.7217, "grad_norm": 0.7043559551239014, "learning_rate": 0.0002, "epoch": 2.591544741734702, "step": 16030}, {"loss": 0.6967, "grad_norm": 0.5846083760261536, "learning_rate": 0.0002, "epoch": 2.593161425915447, "step": 16040}, {"loss": 0.7322, "grad_norm": 0.7323982119560242, "learning_rate": 0.0002, "epoch": 2.594778110096193, "step": 16050}, {"loss": 0.6794, "grad_norm": 0.9069556593894958, "learning_rate": 0.0002, "epoch": 2.596394794276938, "step": 16060}, {"loss": 0.7076, "grad_norm": 0.7522736191749573, "learning_rate": 0.0002, "epoch": 2.5980114784576833, "step": 16070}, {"loss": 0.6477, "grad_norm": 0.8149648308753967, "learning_rate": 0.0002, "epoch": 2.5996281626384286, "step": 16080}, {"loss": 0.6664, "grad_norm": 0.6214233040809631, "learning_rate": 0.0002, "epoch": 2.601244846819174, "step": 16090}, {"loss": 0.7307, "grad_norm": 0.6803743839263916, "learning_rate": 0.0002, "epoch": 2.602861530999919, "step": 16100}, {"loss": 0.7244, "grad_norm": 0.7223997116088867, "learning_rate": 0.0002, "epoch": 2.6044782151806647, "step": 16110}, {"loss": 0.6867, "grad_norm": 0.7324174642562866, "learning_rate": 0.0002, "epoch": 2.60609489936141, "step": 16120}, {"loss": 0.7159, "grad_norm": 0.9594739675521851, "learning_rate": 0.0002, "epoch": 2.607711583542155, "step": 16130}, {"loss": 0.6451, "grad_norm": 0.9485327005386353, "learning_rate": 0.0002, "epoch": 2.6093282677229004, "step": 16140}, {"loss": 0.6815, "grad_norm": 0.8449000120162964, "learning_rate": 0.0002, "epoch": 2.6109449519036456, "step": 16150}, {"loss": 0.7152, "grad_norm": 0.8520140051841736, "learning_rate": 0.0002, "epoch": 2.612561636084391, "step": 16160}, {"loss": 0.6759, "grad_norm": 0.7456524968147278, "learning_rate": 0.0002, "epoch": 2.614178320265136, "step": 16170}, {"loss": 0.6893, "grad_norm": 0.9912857413291931, "learning_rate": 0.0002, "epoch": 2.6157950044458813, "step": 16180}, {"loss": 0.7243, "grad_norm": 0.9001946449279785, "learning_rate": 0.0002, "epoch": 2.6174116886266265, "step": 16190}, {"loss": 0.6825, "grad_norm": 0.6568667888641357, "learning_rate": 0.0002, "epoch": 2.619028372807372, "step": 16200}, {"loss": 0.7013, "grad_norm": 1.0248128175735474, "learning_rate": 0.0002, "epoch": 2.6206450569881174, "step": 16210}, {"loss": 0.7045, "grad_norm": 0.6509039998054504, "learning_rate": 0.0002, "epoch": 2.6222617411688627, "step": 16220}, {"loss": 0.72, "grad_norm": 0.7626351118087769, "learning_rate": 0.0002, "epoch": 2.623878425349608, "step": 16230}, {"loss": 0.6556, "grad_norm": 0.6938552260398865, "learning_rate": 0.0002, "epoch": 2.625495109530353, "step": 16240}, {"loss": 0.65, "grad_norm": 0.6434680819511414, "learning_rate": 0.0002, "epoch": 2.6271117937110984, "step": 16250}, {"loss": 0.6943, "grad_norm": 0.7111515998840332, "learning_rate": 0.0002, "epoch": 2.628728477891844, "step": 16260}, {"loss": 0.679, "grad_norm": 0.7712395787239075, "learning_rate": 0.0002, "epoch": 2.6303451620725893, "step": 16270}, {"loss": 0.6886, "grad_norm": 0.792209267616272, "learning_rate": 0.0002, "epoch": 2.6319618462533345, "step": 16280}, {"loss": 0.6554, "grad_norm": 0.6801066398620605, "learning_rate": 0.0002, "epoch": 2.6335785304340797, "step": 16290}, {"loss": 0.73, "grad_norm": 0.7802573442459106, "learning_rate": 0.0002, "epoch": 2.635195214614825, "step": 16300}, {"loss": 0.7484, "grad_norm": 0.7742244601249695, "learning_rate": 0.0002, "epoch": 2.63681189879557, "step": 16310}, {"loss": 0.6524, "grad_norm": 0.664184033870697, "learning_rate": 0.0002, "epoch": 2.6384285829763154, "step": 16320}, {"loss": 0.6442, "grad_norm": 0.9242228865623474, "learning_rate": 0.0002, "epoch": 2.6400452671570607, "step": 16330}, {"loss": 0.6792, "grad_norm": 0.9661325216293335, "learning_rate": 0.0002, "epoch": 2.641661951337806, "step": 16340}, {"loss": 0.6847, "grad_norm": 0.837526798248291, "learning_rate": 0.0002, "epoch": 2.6432786355185516, "step": 16350}, {"loss": 0.7686, "grad_norm": 1.1834373474121094, "learning_rate": 0.0002, "epoch": 2.644895319699297, "step": 16360}, {"loss": 0.6746, "grad_norm": 0.7467831373214722, "learning_rate": 0.0002, "epoch": 2.646512003880042, "step": 16370}, {"loss": 0.6935, "grad_norm": 0.8627146482467651, "learning_rate": 0.0002, "epoch": 2.6481286880607873, "step": 16380}, {"loss": 0.715, "grad_norm": 0.790447473526001, "learning_rate": 0.0002, "epoch": 2.6497453722415325, "step": 16390}, {"loss": 0.723, "grad_norm": 0.8447365164756775, "learning_rate": 0.0002, "epoch": 2.651362056422278, "step": 16400}, {"loss": 0.6628, "grad_norm": 0.7831417918205261, "learning_rate": 0.0002, "epoch": 2.6529787406030234, "step": 16410}, {"loss": 0.6691, "grad_norm": 0.6837952136993408, "learning_rate": 0.0002, "epoch": 2.6545954247837686, "step": 16420}, {"loss": 0.6139, "grad_norm": 0.7031801342964172, "learning_rate": 0.0002, "epoch": 2.656212108964514, "step": 16430}, {"loss": 0.7382, "grad_norm": 0.8963770866394043, "learning_rate": 0.0002, "epoch": 2.657828793145259, "step": 16440}, {"loss": 0.6439, "grad_norm": 0.6852328181266785, "learning_rate": 0.0002, "epoch": 2.6594454773260043, "step": 16450}, {"loss": 0.6278, "grad_norm": 0.8069294095039368, "learning_rate": 0.0002, "epoch": 2.6610621615067496, "step": 16460}, {"loss": 0.6939, "grad_norm": 0.7503686547279358, "learning_rate": 0.0002, "epoch": 2.662678845687495, "step": 16470}, {"loss": 0.6777, "grad_norm": 0.6430956125259399, "learning_rate": 0.0002, "epoch": 2.66429552986824, "step": 16480}, {"loss": 0.6863, "grad_norm": 0.7894312739372253, "learning_rate": 0.0002, "epoch": 2.6659122140489853, "step": 16490}, {"loss": 0.7165, "grad_norm": 0.7277431488037109, "learning_rate": 0.0002, "epoch": 2.667528898229731, "step": 16500}, {"loss": 0.6772, "grad_norm": 0.6816153526306152, "learning_rate": 0.0002, "epoch": 2.669145582410476, "step": 16510}, {"loss": 0.691, "grad_norm": 0.8145235776901245, "learning_rate": 0.0002, "epoch": 2.6707622665912214, "step": 16520}, {"loss": 0.709, "grad_norm": 0.8645890355110168, "learning_rate": 0.0002, "epoch": 2.6723789507719666, "step": 16530}, {"loss": 0.6946, "grad_norm": 0.704393208026886, "learning_rate": 0.0002, "epoch": 2.673995634952712, "step": 16540}, {"loss": 0.6378, "grad_norm": 1.0120846033096313, "learning_rate": 0.0002, "epoch": 2.6756123191334575, "step": 16550}, {"loss": 0.7241, "grad_norm": 0.6919328570365906, "learning_rate": 0.0002, "epoch": 2.6772290033142028, "step": 16560}, {"loss": 0.7098, "grad_norm": 0.6924574971199036, "learning_rate": 0.0002, "epoch": 2.678845687494948, "step": 16570}, {"loss": 0.731, "grad_norm": 0.9679301381111145, "learning_rate": 0.0002, "epoch": 2.6804623716756932, "step": 16580}, {"loss": 0.7124, "grad_norm": 0.6810211539268494, "learning_rate": 0.0002, "epoch": 2.6820790558564385, "step": 16590}, {"loss": 0.6688, "grad_norm": 0.9730555415153503, "learning_rate": 0.0002, "epoch": 2.6836957400371837, "step": 16600}, {"loss": 0.7344, "grad_norm": 0.7852821350097656, "learning_rate": 0.0002, "epoch": 2.685312424217929, "step": 16610}, {"loss": 0.6401, "grad_norm": 0.6059057116508484, "learning_rate": 0.0002, "epoch": 2.686929108398674, "step": 16620}, {"loss": 0.6796, "grad_norm": 0.9395958781242371, "learning_rate": 0.0002, "epoch": 2.6885457925794194, "step": 16630}, {"loss": 0.7174, "grad_norm": 0.7473729848861694, "learning_rate": 0.0002, "epoch": 2.690162476760165, "step": 16640}, {"loss": 0.7087, "grad_norm": 0.765934407711029, "learning_rate": 0.0002, "epoch": 2.6917791609409103, "step": 16650}, {"loss": 0.707, "grad_norm": 0.8496677279472351, "learning_rate": 0.0002, "epoch": 2.6933958451216555, "step": 16660}, {"loss": 0.7084, "grad_norm": 0.7641879916191101, "learning_rate": 0.0002, "epoch": 2.6950125293024008, "step": 16670}, {"loss": 0.6566, "grad_norm": 0.8471952676773071, "learning_rate": 0.0002, "epoch": 2.696629213483146, "step": 16680}, {"loss": 0.6635, "grad_norm": 0.6946060657501221, "learning_rate": 0.0002, "epoch": 2.6982458976638912, "step": 16690}, {"loss": 0.7027, "grad_norm": 0.7361312508583069, "learning_rate": 0.0002, "epoch": 2.699862581844637, "step": 16700}, {"loss": 0.6767, "grad_norm": 0.6605038046836853, "learning_rate": 0.0002, "epoch": 2.701479266025382, "step": 16710}, {"loss": 0.6885, "grad_norm": 0.7164411544799805, "learning_rate": 0.0002, "epoch": 2.7030959502061274, "step": 16720}, {"loss": 0.6736, "grad_norm": 0.6496201157569885, "learning_rate": 0.0002, "epoch": 2.7047126343868726, "step": 16730}, {"loss": 0.6942, "grad_norm": 0.7826663851737976, "learning_rate": 0.0002, "epoch": 2.706329318567618, "step": 16740}, {"loss": 0.6773, "grad_norm": 0.7639131546020508, "learning_rate": 0.0002, "epoch": 2.707946002748363, "step": 16750}, {"loss": 0.69, "grad_norm": 0.7976210713386536, "learning_rate": 0.0002, "epoch": 2.7095626869291083, "step": 16760}, {"loss": 0.6735, "grad_norm": 0.6836577653884888, "learning_rate": 0.0002, "epoch": 2.7111793711098535, "step": 16770}, {"loss": 0.6596, "grad_norm": 0.8025202751159668, "learning_rate": 0.0002, "epoch": 2.7127960552905988, "step": 16780}, {"loss": 0.6324, "grad_norm": 0.7636463642120361, "learning_rate": 0.0002, "epoch": 2.7144127394713444, "step": 16790}, {"loss": 0.6227, "grad_norm": 0.7481677532196045, "learning_rate": 0.0002, "epoch": 2.7160294236520897, "step": 16800}, {"loss": 0.6925, "grad_norm": 0.7566834688186646, "learning_rate": 0.0002, "epoch": 2.717646107832835, "step": 16810}, {"loss": 0.6531, "grad_norm": 0.7931267619132996, "learning_rate": 0.0002, "epoch": 2.71926279201358, "step": 16820}, {"loss": 0.6672, "grad_norm": 0.8811662197113037, "learning_rate": 0.0002, "epoch": 2.7208794761943254, "step": 16830}, {"loss": 0.6675, "grad_norm": 0.8561240434646606, "learning_rate": 0.0002, "epoch": 2.7224961603750706, "step": 16840}, {"loss": 0.7135, "grad_norm": 0.7121599316596985, "learning_rate": 0.0002, "epoch": 2.7241128445558163, "step": 16850}, {"loss": 0.6825, "grad_norm": 0.8066257238388062, "learning_rate": 0.0002, "epoch": 2.7257295287365615, "step": 16860}, {"loss": 0.6839, "grad_norm": 0.7699271440505981, "learning_rate": 0.0002, "epoch": 2.7273462129173067, "step": 16870}, {"loss": 0.699, "grad_norm": 1.1828432083129883, "learning_rate": 0.0002, "epoch": 2.728962897098052, "step": 16880}, {"loss": 0.6518, "grad_norm": 0.9989302754402161, "learning_rate": 0.0002, "epoch": 2.730579581278797, "step": 16890}, {"loss": 0.7015, "grad_norm": 0.8100560307502747, "learning_rate": 0.0002, "epoch": 2.7321962654595424, "step": 16900}, {"loss": 0.6851, "grad_norm": 0.8615233898162842, "learning_rate": 0.0002, "epoch": 2.7338129496402876, "step": 16910}, {"loss": 0.6322, "grad_norm": 0.8633756041526794, "learning_rate": 0.0002, "epoch": 2.735429633821033, "step": 16920}, {"loss": 0.6488, "grad_norm": 0.7769348621368408, "learning_rate": 0.0002, "epoch": 2.737046318001778, "step": 16930}, {"loss": 0.6582, "grad_norm": 0.6943058371543884, "learning_rate": 0.0002, "epoch": 2.738663002182524, "step": 16940}, {"loss": 0.6516, "grad_norm": 0.8510736227035522, "learning_rate": 0.0002, "epoch": 2.740279686363269, "step": 16950}, {"loss": 0.7275, "grad_norm": 0.7732602953910828, "learning_rate": 0.0002, "epoch": 2.7418963705440142, "step": 16960}, {"loss": 0.6553, "grad_norm": 0.5981788635253906, "learning_rate": 0.0002, "epoch": 2.7435130547247595, "step": 16970}, {"loss": 0.6777, "grad_norm": 0.7604416012763977, "learning_rate": 0.0002, "epoch": 2.7451297389055047, "step": 16980}, {"loss": 0.6981, "grad_norm": 0.7377738356590271, "learning_rate": 0.0002, "epoch": 2.74674642308625, "step": 16990}, {"loss": 0.6294, "grad_norm": 0.9400289058685303, "learning_rate": 0.0002, "epoch": 2.7483631072669956, "step": 17000}, {"loss": 0.6952, "grad_norm": 0.6340599656105042, "learning_rate": 0.0002, "epoch": 2.749979791447741, "step": 17010}, {"loss": 0.7222, "grad_norm": 0.7297601103782654, "learning_rate": 0.0002, "epoch": 2.751596475628486, "step": 17020}, {"loss": 0.6659, "grad_norm": 0.9479979872703552, "learning_rate": 0.0002, "epoch": 2.7532131598092313, "step": 17030}, {"loss": 0.691, "grad_norm": 0.8461511135101318, "learning_rate": 0.0002, "epoch": 2.7548298439899765, "step": 17040}, {"loss": 0.6764, "grad_norm": 0.7477551698684692, "learning_rate": 0.0002, "epoch": 2.7564465281707218, "step": 17050}, {"loss": 0.684, "grad_norm": 1.019270420074463, "learning_rate": 0.0002, "epoch": 2.758063212351467, "step": 17060}, {"loss": 0.7119, "grad_norm": 0.7730235457420349, "learning_rate": 0.0002, "epoch": 2.7596798965322122, "step": 17070}, {"loss": 0.6886, "grad_norm": 0.8216866254806519, "learning_rate": 0.0002, "epoch": 2.7612965807129575, "step": 17080}, {"loss": 0.6811, "grad_norm": 0.7235931754112244, "learning_rate": 0.0002, "epoch": 2.762913264893703, "step": 17090}, {"loss": 0.7031, "grad_norm": 0.7352296710014343, "learning_rate": 0.0002, "epoch": 2.7645299490744484, "step": 17100}, {"loss": 0.6951, "grad_norm": 0.8129373788833618, "learning_rate": 0.0002, "epoch": 2.7661466332551936, "step": 17110}, {"loss": 0.6703, "grad_norm": 0.7387019991874695, "learning_rate": 0.0002, "epoch": 2.767763317435939, "step": 17120}, {"loss": 0.6789, "grad_norm": 0.9149190187454224, "learning_rate": 0.0002, "epoch": 2.769380001616684, "step": 17130}, {"loss": 0.6038, "grad_norm": 0.7352971434593201, "learning_rate": 0.0002, "epoch": 2.7709966857974297, "step": 17140}, {"loss": 0.6728, "grad_norm": 0.7903780341148376, "learning_rate": 0.0002, "epoch": 2.772613369978175, "step": 17150}, {"loss": 0.6988, "grad_norm": 0.8255927562713623, "learning_rate": 0.0002, "epoch": 2.77423005415892, "step": 17160}, {"loss": 0.6694, "grad_norm": 0.7235927581787109, "learning_rate": 0.0002, "epoch": 2.7758467383396654, "step": 17170}, {"loss": 0.7161, "grad_norm": 0.8281434774398804, "learning_rate": 0.0002, "epoch": 2.7774634225204107, "step": 17180}, {"loss": 0.682, "grad_norm": 0.7586921453475952, "learning_rate": 0.0002, "epoch": 2.779080106701156, "step": 17190}, {"loss": 0.6427, "grad_norm": 0.7161715030670166, "learning_rate": 0.0002, "epoch": 2.780696790881901, "step": 17200}, {"loss": 0.6426, "grad_norm": 0.762868344783783, "learning_rate": 0.0002, "epoch": 2.7823134750626464, "step": 17210}, {"loss": 0.705, "grad_norm": 0.9285483360290527, "learning_rate": 0.0002, "epoch": 2.7839301592433916, "step": 17220}, {"loss": 0.7084, "grad_norm": 0.6900462508201599, "learning_rate": 0.0002, "epoch": 2.785546843424137, "step": 17230}, {"loss": 0.6988, "grad_norm": 0.780384361743927, "learning_rate": 0.0002, "epoch": 2.7871635276048825, "step": 17240}, {"loss": 0.7073, "grad_norm": 0.7580406665802002, "learning_rate": 0.0002, "epoch": 2.7887802117856277, "step": 17250}, {"loss": 0.6833, "grad_norm": 0.8145199418067932, "learning_rate": 0.0002, "epoch": 2.790396895966373, "step": 17260}, {"loss": 0.6909, "grad_norm": 0.9159596562385559, "learning_rate": 0.0002, "epoch": 2.792013580147118, "step": 17270}, {"loss": 0.6008, "grad_norm": 0.9590014219284058, "learning_rate": 0.0002, "epoch": 2.7936302643278634, "step": 17280}, {"loss": 0.6704, "grad_norm": 0.7603529691696167, "learning_rate": 0.0002, "epoch": 2.795246948508609, "step": 17290}, {"loss": 0.7165, "grad_norm": 0.8039976358413696, "learning_rate": 0.0002, "epoch": 2.7968636326893543, "step": 17300}, {"loss": 0.7037, "grad_norm": 0.8364847302436829, "learning_rate": 0.0002, "epoch": 2.7984803168700996, "step": 17310}, {"loss": 0.6749, "grad_norm": 0.8763046860694885, "learning_rate": 0.0002, "epoch": 2.800097001050845, "step": 17320}, {"loss": 0.6844, "grad_norm": 0.8409647941589355, "learning_rate": 0.0002, "epoch": 2.80171368523159, "step": 17330}, {"loss": 0.6936, "grad_norm": 0.7649006247520447, "learning_rate": 0.0002, "epoch": 2.8033303694123353, "step": 17340}, {"loss": 0.7051, "grad_norm": 0.7970262169837952, "learning_rate": 0.0002, "epoch": 2.8049470535930805, "step": 17350}, {"loss": 0.6533, "grad_norm": 0.9088607430458069, "learning_rate": 0.0002, "epoch": 2.8065637377738257, "step": 17360}, {"loss": 0.675, "grad_norm": 0.6454846858978271, "learning_rate": 0.0002, "epoch": 2.808180421954571, "step": 17370}, {"loss": 0.7069, "grad_norm": 0.7744787931442261, "learning_rate": 0.0002, "epoch": 2.809797106135316, "step": 17380}, {"loss": 0.6772, "grad_norm": 0.6678640842437744, "learning_rate": 0.0002, "epoch": 2.811413790316062, "step": 17390}, {"loss": 0.6784, "grad_norm": 0.772676944732666, "learning_rate": 0.0002, "epoch": 2.813030474496807, "step": 17400}, {"loss": 0.7252, "grad_norm": 0.7088175415992737, "learning_rate": 0.0002, "epoch": 2.8146471586775523, "step": 17410}, {"loss": 0.7086, "grad_norm": 0.8280573487281799, "learning_rate": 0.0002, "epoch": 2.8162638428582976, "step": 17420}, {"loss": 0.6732, "grad_norm": 0.6665388345718384, "learning_rate": 0.0002, "epoch": 2.817880527039043, "step": 17430}, {"loss": 0.6675, "grad_norm": 0.6427883505821228, "learning_rate": 0.0002, "epoch": 2.8194972112197885, "step": 17440}, {"loss": 0.6972, "grad_norm": 0.9697760343551636, "learning_rate": 0.0002, "epoch": 2.8211138954005337, "step": 17450}, {"loss": 0.6838, "grad_norm": 0.7573966383934021, "learning_rate": 0.0002, "epoch": 2.822730579581279, "step": 17460}, {"loss": 0.7243, "grad_norm": 0.878688633441925, "learning_rate": 0.0002, "epoch": 2.824347263762024, "step": 17470}, {"loss": 0.6666, "grad_norm": 0.7752242684364319, "learning_rate": 0.0002, "epoch": 2.8259639479427694, "step": 17480}, {"loss": 0.6638, "grad_norm": 0.6135398745536804, "learning_rate": 0.0002, "epoch": 2.8275806321235146, "step": 17490}, {"loss": 0.6829, "grad_norm": 0.6924924850463867, "learning_rate": 0.0002, "epoch": 2.82919731630426, "step": 17500}, {"loss": 0.6731, "grad_norm": 0.7471627593040466, "learning_rate": 0.0002, "epoch": 2.830814000485005, "step": 17510}, {"loss": 0.7016, "grad_norm": 0.7145499587059021, "learning_rate": 0.0002, "epoch": 2.8324306846657503, "step": 17520}, {"loss": 0.6787, "grad_norm": 0.7415414452552795, "learning_rate": 0.0002, "epoch": 2.834047368846496, "step": 17530}, {"loss": 0.6811, "grad_norm": 0.7328441739082336, "learning_rate": 0.0002, "epoch": 2.8356640530272412, "step": 17540}, {"loss": 0.6866, "grad_norm": 0.8267839550971985, "learning_rate": 0.0002, "epoch": 2.8372807372079865, "step": 17550}, {"loss": 0.6787, "grad_norm": 0.8877885341644287, "learning_rate": 0.0002, "epoch": 2.8388974213887317, "step": 17560}, {"loss": 0.7136, "grad_norm": 0.857138454914093, "learning_rate": 0.0002, "epoch": 2.840514105569477, "step": 17570}, {"loss": 0.6454, "grad_norm": 0.8470779657363892, "learning_rate": 0.0002, "epoch": 2.842130789750222, "step": 17580}, {"loss": 0.6976, "grad_norm": 0.8553254008293152, "learning_rate": 0.0002, "epoch": 2.843747473930968, "step": 17590}, {"loss": 0.7297, "grad_norm": 0.8033196926116943, "learning_rate": 0.0002, "epoch": 2.845364158111713, "step": 17600}, {"loss": 0.7062, "grad_norm": 0.7949087023735046, "learning_rate": 0.0002, "epoch": 2.8469808422924583, "step": 17610}, {"loss": 0.651, "grad_norm": 0.9241406321525574, "learning_rate": 0.0002, "epoch": 2.8485975264732035, "step": 17620}, {"loss": 0.6601, "grad_norm": 0.7721285223960876, "learning_rate": 0.0002, "epoch": 2.8502142106539488, "step": 17630}, {"loss": 0.6183, "grad_norm": 1.0246692895889282, "learning_rate": 0.0002, "epoch": 2.851830894834694, "step": 17640}, {"loss": 0.7007, "grad_norm": 0.9244589805603027, "learning_rate": 0.0002, "epoch": 2.853447579015439, "step": 17650}, {"loss": 0.7274, "grad_norm": 0.7243508696556091, "learning_rate": 0.0002, "epoch": 2.8550642631961844, "step": 17660}, {"loss": 0.6471, "grad_norm": 0.8943371176719666, "learning_rate": 0.0002, "epoch": 2.8566809473769297, "step": 17670}, {"loss": 0.686, "grad_norm": 0.6531758904457092, "learning_rate": 0.0002, "epoch": 2.8582976315576754, "step": 17680}, {"loss": 0.6253, "grad_norm": 0.8367000818252563, "learning_rate": 0.0002, "epoch": 2.8599143157384206, "step": 17690}, {"loss": 0.6943, "grad_norm": 0.7868556380271912, "learning_rate": 0.0002, "epoch": 2.861530999919166, "step": 17700}, {"loss": 0.6919, "grad_norm": 0.7213859558105469, "learning_rate": 0.0002, "epoch": 2.863147684099911, "step": 17710}, {"loss": 0.6657, "grad_norm": 0.7383931279182434, "learning_rate": 0.0002, "epoch": 2.8647643682806563, "step": 17720}, {"loss": 0.6841, "grad_norm": 0.7566812634468079, "learning_rate": 0.0002, "epoch": 2.8663810524614015, "step": 17730}, {"loss": 0.6449, "grad_norm": 0.6930373311042786, "learning_rate": 0.0002, "epoch": 2.867997736642147, "step": 17740}, {"loss": 0.6764, "grad_norm": 0.7911090850830078, "learning_rate": 0.0002, "epoch": 2.8696144208228924, "step": 17750}, {"loss": 0.6554, "grad_norm": 0.8484548926353455, "learning_rate": 0.0002, "epoch": 2.8712311050036377, "step": 17760}, {"loss": 0.6931, "grad_norm": 0.7647597193717957, "learning_rate": 0.0002, "epoch": 2.872847789184383, "step": 17770}, {"loss": 0.6945, "grad_norm": 0.8791151642799377, "learning_rate": 0.0002, "epoch": 2.874464473365128, "step": 17780}, {"loss": 0.7078, "grad_norm": 0.7253178358078003, "learning_rate": 0.0002, "epoch": 2.8760811575458733, "step": 17790}, {"loss": 0.6474, "grad_norm": 0.7956077456474304, "learning_rate": 0.0002, "epoch": 2.8776978417266186, "step": 17800}, {"loss": 0.6687, "grad_norm": 0.8657688498497009, "learning_rate": 0.0002, "epoch": 2.879314525907364, "step": 17810}, {"loss": 0.7171, "grad_norm": 0.7059141993522644, "learning_rate": 0.0002, "epoch": 2.880931210088109, "step": 17820}, {"loss": 0.683, "grad_norm": 0.8886896967887878, "learning_rate": 0.0002, "epoch": 2.8825478942688547, "step": 17830}, {"loss": 0.669, "grad_norm": 0.821032702922821, "learning_rate": 0.0002, "epoch": 2.8841645784496, "step": 17840}, {"loss": 0.6805, "grad_norm": 0.7183963656425476, "learning_rate": 0.0002, "epoch": 2.885781262630345, "step": 17850}, {"loss": 0.7088, "grad_norm": 0.6222899556159973, "learning_rate": 0.0002, "epoch": 2.8873979468110904, "step": 17860}, {"loss": 0.6626, "grad_norm": 0.8187434077262878, "learning_rate": 0.0002, "epoch": 2.8890146309918356, "step": 17870}, {"loss": 0.6815, "grad_norm": 0.9838479161262512, "learning_rate": 0.0002, "epoch": 2.890631315172581, "step": 17880}, {"loss": 0.6967, "grad_norm": 0.7567742466926575, "learning_rate": 0.0002, "epoch": 2.8922479993533265, "step": 17890}, {"loss": 0.7073, "grad_norm": 0.6875903606414795, "learning_rate": 0.0002, "epoch": 2.893864683534072, "step": 17900}, {"loss": 0.6415, "grad_norm": 0.8043789267539978, "learning_rate": 0.0002, "epoch": 2.895481367714817, "step": 17910}, {"loss": 0.6588, "grad_norm": 0.8062626719474792, "learning_rate": 0.0002, "epoch": 2.8970980518955622, "step": 17920}, {"loss": 0.7151, "grad_norm": 1.0251191854476929, "learning_rate": 0.0002, "epoch": 2.8987147360763075, "step": 17930}, {"loss": 0.6605, "grad_norm": 0.882253110408783, "learning_rate": 0.0002, "epoch": 2.9003314202570527, "step": 17940}, {"loss": 0.6719, "grad_norm": 0.8683299422264099, "learning_rate": 0.0002, "epoch": 2.901948104437798, "step": 17950}, {"loss": 0.6896, "grad_norm": 0.7167282104492188, "learning_rate": 0.0002, "epoch": 2.903564788618543, "step": 17960}, {"loss": 0.663, "grad_norm": 0.7093694806098938, "learning_rate": 0.0002, "epoch": 2.9051814727992884, "step": 17970}, {"loss": 0.6591, "grad_norm": 0.8549879193305969, "learning_rate": 0.0002, "epoch": 2.906798156980034, "step": 17980}, {"loss": 0.6962, "grad_norm": 0.6989606618881226, "learning_rate": 0.0002, "epoch": 2.9084148411607793, "step": 17990}, {"loss": 0.6635, "grad_norm": 0.9482976794242859, "learning_rate": 0.0002, "epoch": 2.9100315253415245, "step": 18000}, {"loss": 0.6586, "grad_norm": 0.7182440161705017, "learning_rate": 0.0002, "epoch": 2.9116482095222698, "step": 18010}, {"loss": 0.6827, "grad_norm": 0.7732226252555847, "learning_rate": 0.0002, "epoch": 2.913264893703015, "step": 18020}, {"loss": 0.7123, "grad_norm": 0.7936875224113464, "learning_rate": 0.0002, "epoch": 2.9148815778837607, "step": 18030}, {"loss": 0.6736, "grad_norm": 0.8825615644454956, "learning_rate": 0.0002, "epoch": 2.916498262064506, "step": 18040}, {"loss": 0.7139, "grad_norm": 0.6778587102890015, "learning_rate": 0.0002, "epoch": 2.918114946245251, "step": 18050}, {"loss": 0.6588, "grad_norm": 0.7529265880584717, "learning_rate": 0.0002, "epoch": 2.9197316304259964, "step": 18060}, {"loss": 0.737, "grad_norm": 0.7111883163452148, "learning_rate": 0.0002, "epoch": 2.9213483146067416, "step": 18070}, {"loss": 0.7475, "grad_norm": 0.7214767932891846, "learning_rate": 0.0002, "epoch": 2.922964998787487, "step": 18080}, {"loss": 0.6672, "grad_norm": 0.800417423248291, "learning_rate": 0.0002, "epoch": 2.924581682968232, "step": 18090}, {"loss": 0.6694, "grad_norm": 1.248575210571289, "learning_rate": 0.0002, "epoch": 2.9261983671489773, "step": 18100}, {"loss": 0.7004, "grad_norm": 0.757788360118866, "learning_rate": 0.0002, "epoch": 2.9278150513297225, "step": 18110}, {"loss": 0.6999, "grad_norm": 1.0583995580673218, "learning_rate": 0.0002, "epoch": 2.9294317355104678, "step": 18120}, {"loss": 0.6365, "grad_norm": 0.8228777647018433, "learning_rate": 0.0002, "epoch": 2.9310484196912134, "step": 18130}, {"loss": 0.6791, "grad_norm": 0.8374035358428955, "learning_rate": 0.0002, "epoch": 2.9326651038719587, "step": 18140}, {"loss": 0.6399, "grad_norm": 0.7976473569869995, "learning_rate": 0.0002, "epoch": 2.934281788052704, "step": 18150}, {"loss": 0.6585, "grad_norm": 0.8009907603263855, "learning_rate": 0.0002, "epoch": 2.935898472233449, "step": 18160}, {"loss": 0.7485, "grad_norm": 0.835213303565979, "learning_rate": 0.0002, "epoch": 2.9375151564141944, "step": 18170}, {"loss": 0.7376, "grad_norm": 0.7982219457626343, "learning_rate": 0.0002, "epoch": 2.93913184059494, "step": 18180}, {"loss": 0.6348, "grad_norm": 0.7070978879928589, "learning_rate": 0.0002, "epoch": 2.9407485247756853, "step": 18190}, {"loss": 0.6608, "grad_norm": 0.8619440197944641, "learning_rate": 0.0002, "epoch": 2.9423652089564305, "step": 18200}, {"loss": 0.666, "grad_norm": 0.6693987250328064, "learning_rate": 0.0002, "epoch": 2.9439818931371757, "step": 18210}, {"loss": 0.728, "grad_norm": 0.6747021079063416, "learning_rate": 0.0002, "epoch": 2.945598577317921, "step": 18220}, {"loss": 0.6686, "grad_norm": 0.860387921333313, "learning_rate": 0.0002, "epoch": 2.947215261498666, "step": 18230}, {"loss": 0.6945, "grad_norm": 0.799976646900177, "learning_rate": 0.0002, "epoch": 2.9488319456794114, "step": 18240}, {"loss": 0.7243, "grad_norm": 0.7864769101142883, "learning_rate": 0.0002, "epoch": 2.9504486298601567, "step": 18250}, {"loss": 0.6785, "grad_norm": 0.6713884472846985, "learning_rate": 0.0002, "epoch": 2.952065314040902, "step": 18260}, {"loss": 0.7429, "grad_norm": 0.9031508564949036, "learning_rate": 0.0002, "epoch": 2.9536819982216476, "step": 18270}, {"loss": 0.7055, "grad_norm": 0.7205073237419128, "learning_rate": 0.0002, "epoch": 2.955298682402393, "step": 18280}, {"loss": 0.7298, "grad_norm": 0.7746205925941467, "learning_rate": 0.0002, "epoch": 2.956915366583138, "step": 18290}, {"loss": 0.6218, "grad_norm": 0.6533427834510803, "learning_rate": 0.0002, "epoch": 2.9585320507638833, "step": 18300}, {"loss": 0.6674, "grad_norm": 0.9083208441734314, "learning_rate": 0.0002, "epoch": 2.9601487349446285, "step": 18310}, {"loss": 0.7359, "grad_norm": 0.7446991801261902, "learning_rate": 0.0002, "epoch": 2.9617654191253737, "step": 18320}, {"loss": 0.6738, "grad_norm": 0.6514461636543274, "learning_rate": 0.0002, "epoch": 2.9633821033061194, "step": 18330}, {"loss": 0.6677, "grad_norm": 0.8580465912818909, "learning_rate": 0.0002, "epoch": 2.9649987874868646, "step": 18340}, {"loss": 0.6971, "grad_norm": 0.7074266076087952, "learning_rate": 0.0002, "epoch": 2.96661547166761, "step": 18350}, {"loss": 0.6804, "grad_norm": 0.899892270565033, "learning_rate": 0.0002, "epoch": 2.968232155848355, "step": 18360}, {"loss": 0.7094, "grad_norm": 0.8217641711235046, "learning_rate": 0.0002, "epoch": 2.9698488400291003, "step": 18370}, {"loss": 0.6916, "grad_norm": 0.8611799478530884, "learning_rate": 0.0002, "epoch": 2.9714655242098456, "step": 18380}, {"loss": 0.6677, "grad_norm": 0.6909302473068237, "learning_rate": 0.0002, "epoch": 2.973082208390591, "step": 18390}, {"loss": 0.7247, "grad_norm": 0.6554358005523682, "learning_rate": 0.0002, "epoch": 2.974698892571336, "step": 18400}, {"loss": 0.6516, "grad_norm": 0.7803071737289429, "learning_rate": 0.0002, "epoch": 2.9763155767520812, "step": 18410}, {"loss": 0.7322, "grad_norm": 0.7838954925537109, "learning_rate": 0.0002, "epoch": 2.977932260932827, "step": 18420}, {"loss": 0.6522, "grad_norm": 0.7098495364189148, "learning_rate": 0.0002, "epoch": 2.979548945113572, "step": 18430}, {"loss": 0.739, "grad_norm": 0.8981785774230957, "learning_rate": 0.0002, "epoch": 2.9811656292943174, "step": 18440}, {"loss": 0.6689, "grad_norm": 0.7197171449661255, "learning_rate": 0.0002, "epoch": 2.9827823134750626, "step": 18450}, {"loss": 0.706, "grad_norm": 0.793185293674469, "learning_rate": 0.0002, "epoch": 2.984398997655808, "step": 18460}, {"loss": 0.7124, "grad_norm": 0.8531473875045776, "learning_rate": 0.0002, "epoch": 2.986015681836553, "step": 18470}, {"loss": 0.6901, "grad_norm": 0.6627361178398132, "learning_rate": 0.0002, "epoch": 2.9876323660172988, "step": 18480}, {"loss": 0.6591, "grad_norm": 0.5708155035972595, "learning_rate": 0.0002, "epoch": 2.989249050198044, "step": 18490}, {"loss": 0.6725, "grad_norm": 0.8227280378341675, "learning_rate": 0.0002, "epoch": 2.990865734378789, "step": 18500}, {"loss": 0.6701, "grad_norm": 0.7102749943733215, "learning_rate": 0.0002, "epoch": 2.9924824185595345, "step": 18510}, {"loss": 0.7091, "grad_norm": 0.839485228061676, "learning_rate": 0.0002, "epoch": 2.9940991027402797, "step": 18520}, {"loss": 0.6521, "grad_norm": 0.9038704037666321, "learning_rate": 0.0002, "epoch": 2.995715786921025, "step": 18530}, {"loss": 0.7186, "grad_norm": 0.8737510442733765, "learning_rate": 0.0002, "epoch": 2.99733247110177, "step": 18540}, {"loss": 0.6819, "grad_norm": 0.7323142886161804, "learning_rate": 0.0002, "epoch": 2.9989491552825154, "step": 18550}, {"eval_loss": 1.1262480020523071, "eval_runtime": 122.0868, "eval_samples_per_second": 6.004, "eval_steps_per_second": 0.754, "epoch": 2.9999191657909625, "step": 18556}, {"loss": 0.6337, "grad_norm": 0.8465463519096375, "learning_rate": 0.0002, "epoch": 3.000565839463261, "step": 18560}, {"loss": 0.6064, "grad_norm": 0.9134138822555542, "learning_rate": 0.0002, "epoch": 3.0021825236440063, "step": 18570}, {"loss": 0.5804, "grad_norm": 0.760715126991272, "learning_rate": 0.0002, "epoch": 3.0037992078247515, "step": 18580}, {"loss": 0.5571, "grad_norm": 0.9208743572235107, "learning_rate": 0.0002, "epoch": 3.0054158920054967, "step": 18590}, {"loss": 0.5731, "grad_norm": 0.9232364892959595, "learning_rate": 0.0002, "epoch": 3.007032576186242, "step": 18600}, {"loss": 0.6299, "grad_norm": 1.1881544589996338, "learning_rate": 0.0002, "epoch": 3.008649260366987, "step": 18610}, {"loss": 0.5482, "grad_norm": 0.9372987747192383, "learning_rate": 0.0002, "epoch": 3.0102659445477324, "step": 18620}, {"loss": 0.5709, "grad_norm": 0.6900241374969482, "learning_rate": 0.0002, "epoch": 3.0118826287284777, "step": 18630}, {"loss": 0.5256, "grad_norm": 0.8451071381568909, "learning_rate": 0.0002, "epoch": 3.0134993129092233, "step": 18640}, {"loss": 0.5916, "grad_norm": 0.7763112187385559, "learning_rate": 0.0002, "epoch": 3.0151159970899686, "step": 18650}, {"loss": 0.6095, "grad_norm": 1.043653964996338, "learning_rate": 0.0002, "epoch": 3.016732681270714, "step": 18660}, {"loss": 0.6228, "grad_norm": 1.0170660018920898, "learning_rate": 0.0002, "epoch": 3.018349365451459, "step": 18670}, {"loss": 0.5671, "grad_norm": 0.7534180283546448, "learning_rate": 0.0002, "epoch": 3.0199660496322043, "step": 18680}, {"loss": 0.6015, "grad_norm": 0.7507367730140686, "learning_rate": 0.0002, "epoch": 3.0215827338129495, "step": 18690}, {"loss": 0.6201, "grad_norm": 0.7861620187759399, "learning_rate": 0.0002, "epoch": 3.0231994179936947, "step": 18700}, {"loss": 0.5802, "grad_norm": 1.0580339431762695, "learning_rate": 0.0002, "epoch": 3.0248161021744404, "step": 18710}, {"loss": 0.5975, "grad_norm": 0.7542710900306702, "learning_rate": 0.0002, "epoch": 3.0264327863551856, "step": 18720}, {"loss": 0.5695, "grad_norm": 0.8189544677734375, "learning_rate": 0.0002, "epoch": 3.028049470535931, "step": 18730}, {"loss": 0.6109, "grad_norm": 0.9126611351966858, "learning_rate": 0.0002, "epoch": 3.029666154716676, "step": 18740}, {"loss": 0.6443, "grad_norm": 0.8891341686248779, "learning_rate": 0.0002, "epoch": 3.0312828388974213, "step": 18750}, {"loss": 0.6207, "grad_norm": 0.8419283032417297, "learning_rate": 0.0002, "epoch": 3.0328995230781666, "step": 18760}, {"loss": 0.5818, "grad_norm": 0.8048048615455627, "learning_rate": 0.0002, "epoch": 3.034516207258912, "step": 18770}, {"loss": 0.6381, "grad_norm": 0.7820217609405518, "learning_rate": 0.0002, "epoch": 3.0361328914396575, "step": 18780}, {"loss": 0.5843, "grad_norm": 0.854721188545227, "learning_rate": 0.0002, "epoch": 3.0377495756204027, "step": 18790}, {"loss": 0.5784, "grad_norm": 0.912092924118042, "learning_rate": 0.0002, "epoch": 3.039366259801148, "step": 18800}, {"loss": 0.5734, "grad_norm": 0.6596226096153259, "learning_rate": 0.0002, "epoch": 3.040982943981893, "step": 18810}, {"loss": 0.5969, "grad_norm": 0.6351348757743835, "learning_rate": 0.0002, "epoch": 3.0425996281626384, "step": 18820}, {"loss": 0.5953, "grad_norm": 0.778188943862915, "learning_rate": 0.0002, "epoch": 3.0442163123433836, "step": 18830}, {"loss": 0.602, "grad_norm": 0.68234783411026, "learning_rate": 0.0002, "epoch": 3.045832996524129, "step": 18840}, {"loss": 0.5785, "grad_norm": 0.998628556728363, "learning_rate": 0.0002, "epoch": 3.047449680704874, "step": 18850}, {"loss": 0.6231, "grad_norm": 0.7393841743469238, "learning_rate": 0.0002, "epoch": 3.0490663648856198, "step": 18860}, {"loss": 0.568, "grad_norm": 0.84438556432724, "learning_rate": 0.0002, "epoch": 3.050683049066365, "step": 18870}, {"loss": 0.6205, "grad_norm": 0.8857501745223999, "learning_rate": 0.0002, "epoch": 3.0522997332471102, "step": 18880}, {"loss": 0.6335, "grad_norm": 0.7208474278450012, "learning_rate": 0.0002, "epoch": 3.0539164174278555, "step": 18890}, {"loss": 0.5998, "grad_norm": 0.7135229110717773, "learning_rate": 0.0002, "epoch": 3.0555331016086007, "step": 18900}, {"loss": 0.5575, "grad_norm": 0.9130001664161682, "learning_rate": 0.0002, "epoch": 3.057149785789346, "step": 18910}, {"loss": 0.5955, "grad_norm": 0.9001716375350952, "learning_rate": 0.0002, "epoch": 3.058766469970091, "step": 18920}, {"loss": 0.6052, "grad_norm": 0.8667559623718262, "learning_rate": 0.0002, "epoch": 3.060383154150837, "step": 18930}, {"loss": 0.5818, "grad_norm": 0.8943959474563599, "learning_rate": 0.0002, "epoch": 3.061999838331582, "step": 18940}, {"loss": 0.5978, "grad_norm": 0.8298377990722656, "learning_rate": 0.0002, "epoch": 3.0636165225123273, "step": 18950}, {"loss": 0.5782, "grad_norm": 0.7935267686843872, "learning_rate": 0.0002, "epoch": 3.0652332066930725, "step": 18960}, {"loss": 0.6434, "grad_norm": 1.1506379842758179, "learning_rate": 0.0002, "epoch": 3.0668498908738178, "step": 18970}, {"loss": 0.5571, "grad_norm": 0.7693049907684326, "learning_rate": 0.0002, "epoch": 3.068466575054563, "step": 18980}, {"loss": 0.5971, "grad_norm": 0.8040135502815247, "learning_rate": 0.0002, "epoch": 3.0700832592353082, "step": 18990}, {"loss": 0.5541, "grad_norm": 0.828404426574707, "learning_rate": 0.0002, "epoch": 3.0716999434160535, "step": 19000}, {"loss": 0.6048, "grad_norm": 0.8811164498329163, "learning_rate": 0.0002, "epoch": 3.073316627596799, "step": 19010}, {"loss": 0.5845, "grad_norm": 1.036205768585205, "learning_rate": 0.0002, "epoch": 3.0749333117775444, "step": 19020}, {"loss": 0.5838, "grad_norm": 0.8857285976409912, "learning_rate": 0.0002, "epoch": 3.0765499959582896, "step": 19030}, {"loss": 0.592, "grad_norm": 0.8392079472541809, "learning_rate": 0.0002, "epoch": 3.078166680139035, "step": 19040}, {"loss": 0.5927, "grad_norm": 1.0287401676177979, "learning_rate": 0.0002, "epoch": 3.07978336431978, "step": 19050}, {"loss": 0.5964, "grad_norm": 1.0086315870285034, "learning_rate": 0.0002, "epoch": 3.0814000485005253, "step": 19060}, {"loss": 0.5567, "grad_norm": 0.9245324730873108, "learning_rate": 0.0002, "epoch": 3.0830167326812705, "step": 19070}, {"loss": 0.5797, "grad_norm": 0.8680877089500427, "learning_rate": 0.0002, "epoch": 3.084633416862016, "step": 19080}, {"loss": 0.5611, "grad_norm": 0.8814793825149536, "learning_rate": 0.0002, "epoch": 3.0862501010427614, "step": 19090}, {"loss": 0.6051, "grad_norm": 0.9234458208084106, "learning_rate": 0.0002, "epoch": 3.0878667852235067, "step": 19100}, {"loss": 0.6209, "grad_norm": 1.1291664838790894, "learning_rate": 0.0002, "epoch": 3.089483469404252, "step": 19110}, {"loss": 0.5695, "grad_norm": 0.9191402792930603, "learning_rate": 0.0002, "epoch": 3.091100153584997, "step": 19120}, {"loss": 0.5856, "grad_norm": 0.7103154063224792, "learning_rate": 0.0002, "epoch": 3.0927168377657424, "step": 19130}, {"loss": 0.6479, "grad_norm": 0.9368883967399597, "learning_rate": 0.0002, "epoch": 3.0943335219464876, "step": 19140}, {"loss": 0.6167, "grad_norm": 0.9676656723022461, "learning_rate": 0.0002, "epoch": 3.095950206127233, "step": 19150}, {"loss": 0.5794, "grad_norm": 0.8739792704582214, "learning_rate": 0.0002, "epoch": 3.0975668903079785, "step": 19160}, {"loss": 0.6112, "grad_norm": 0.8530174493789673, "learning_rate": 0.0002, "epoch": 3.0991835744887237, "step": 19170}, {"loss": 0.6568, "grad_norm": 0.794945478439331, "learning_rate": 0.0002, "epoch": 3.100800258669469, "step": 19180}, {"loss": 0.5928, "grad_norm": 0.9508888125419617, "learning_rate": 0.0002, "epoch": 3.102416942850214, "step": 19190}, {"loss": 0.5757, "grad_norm": 1.0599955320358276, "learning_rate": 0.0002, "epoch": 3.1040336270309594, "step": 19200}, {"loss": 0.6151, "grad_norm": 1.0673625469207764, "learning_rate": 0.0002, "epoch": 3.1056503112117047, "step": 19210}, {"loss": 0.6043, "grad_norm": 0.7739115953445435, "learning_rate": 0.0002, "epoch": 3.10726699539245, "step": 19220}, {"loss": 0.6046, "grad_norm": 0.9884951114654541, "learning_rate": 0.0002, "epoch": 3.1088836795731956, "step": 19230}, {"loss": 0.5932, "grad_norm": 0.862260103225708, "learning_rate": 0.0002, "epoch": 3.110500363753941, "step": 19240}, {"loss": 0.6098, "grad_norm": 0.7690284848213196, "learning_rate": 0.0002, "epoch": 3.112117047934686, "step": 19250}, {"loss": 0.5791, "grad_norm": 0.8758958578109741, "learning_rate": 0.0002, "epoch": 3.1137337321154313, "step": 19260}, {"loss": 0.6136, "grad_norm": 1.0356395244598389, "learning_rate": 0.0002, "epoch": 3.1153504162961765, "step": 19270}, {"loss": 0.6159, "grad_norm": 0.6950937509536743, "learning_rate": 0.0002, "epoch": 3.1169671004769217, "step": 19280}, {"loss": 0.592, "grad_norm": 0.760998010635376, "learning_rate": 0.0002, "epoch": 3.118583784657667, "step": 19290}, {"loss": 0.575, "grad_norm": 0.9335789084434509, "learning_rate": 0.0002, "epoch": 3.1202004688384126, "step": 19300}, {"loss": 0.6139, "grad_norm": 0.9636204242706299, "learning_rate": 0.0002, "epoch": 3.121817153019158, "step": 19310}, {"loss": 0.6001, "grad_norm": 1.0820997953414917, "learning_rate": 0.0002, "epoch": 3.123433837199903, "step": 19320}, {"loss": 0.6542, "grad_norm": 0.7333487272262573, "learning_rate": 0.0002, "epoch": 3.1250505213806483, "step": 19330}, {"loss": 0.6178, "grad_norm": 1.0417509078979492, "learning_rate": 0.0002, "epoch": 3.1266672055613935, "step": 19340}, {"loss": 0.603, "grad_norm": 0.9267749190330505, "learning_rate": 0.0002, "epoch": 3.128283889742139, "step": 19350}, {"loss": 0.6063, "grad_norm": 0.777798593044281, "learning_rate": 0.0002, "epoch": 3.129900573922884, "step": 19360}, {"loss": 0.5913, "grad_norm": 0.8425456881523132, "learning_rate": 0.0002, "epoch": 3.1315172581036297, "step": 19370}, {"loss": 0.6042, "grad_norm": 0.9617102146148682, "learning_rate": 0.0002, "epoch": 3.133133942284375, "step": 19380}, {"loss": 0.633, "grad_norm": 1.0052828788757324, "learning_rate": 0.0002, "epoch": 3.13475062646512, "step": 19390}, {"loss": 0.5713, "grad_norm": 0.7637009024620056, "learning_rate": 0.0002, "epoch": 3.1363673106458654, "step": 19400}, {"loss": 0.5497, "grad_norm": 0.7958088517189026, "learning_rate": 0.0002, "epoch": 3.1379839948266106, "step": 19410}, {"loss": 0.6283, "grad_norm": 0.9161727428436279, "learning_rate": 0.0002, "epoch": 3.139600679007356, "step": 19420}, {"loss": 0.5638, "grad_norm": 0.8402149677276611, "learning_rate": 0.0002, "epoch": 3.141217363188101, "step": 19430}, {"loss": 0.5848, "grad_norm": 1.0056525468826294, "learning_rate": 0.0002, "epoch": 3.1428340473688463, "step": 19440}, {"loss": 0.5954, "grad_norm": 1.0129190683364868, "learning_rate": 0.0002, "epoch": 3.144450731549592, "step": 19450}, {"loss": 0.5808, "grad_norm": 0.790825366973877, "learning_rate": 0.0002, "epoch": 3.146067415730337, "step": 19460}, {"loss": 0.5607, "grad_norm": 1.441665530204773, "learning_rate": 0.0002, "epoch": 3.1476840999110824, "step": 19470}, {"loss": 0.5785, "grad_norm": 0.7846331596374512, "learning_rate": 0.0002, "epoch": 3.1493007840918277, "step": 19480}, {"loss": 0.5892, "grad_norm": 0.7915332913398743, "learning_rate": 0.0002, "epoch": 3.150917468272573, "step": 19490}, {"loss": 0.5759, "grad_norm": 0.933982253074646, "learning_rate": 0.0002, "epoch": 3.152534152453318, "step": 19500}, {"loss": 0.6206, "grad_norm": 1.038408637046814, "learning_rate": 0.0002, "epoch": 3.1541508366340634, "step": 19510}, {"loss": 0.6271, "grad_norm": 1.018935203552246, "learning_rate": 0.0002, "epoch": 3.155767520814809, "step": 19520}, {"loss": 0.6173, "grad_norm": 0.9618112444877625, "learning_rate": 0.0002, "epoch": 3.1573842049955543, "step": 19530}, {"loss": 0.5972, "grad_norm": 0.8900452852249146, "learning_rate": 0.0002, "epoch": 3.1590008891762995, "step": 19540}, {"loss": 0.5925, "grad_norm": 0.8254160284996033, "learning_rate": 0.0002, "epoch": 3.1606175733570447, "step": 19550}, {"loss": 0.625, "grad_norm": 1.004376769065857, "learning_rate": 0.0002, "epoch": 3.16223425753779, "step": 19560}, {"loss": 0.5775, "grad_norm": 1.0490446090698242, "learning_rate": 0.0002, "epoch": 3.163850941718535, "step": 19570}, {"loss": 0.5986, "grad_norm": 0.7387403845787048, "learning_rate": 0.0002, "epoch": 3.1654676258992804, "step": 19580}, {"loss": 0.5898, "grad_norm": 0.7611538171768188, "learning_rate": 0.0002, "epoch": 3.1670843100800257, "step": 19590}, {"loss": 0.5937, "grad_norm": 0.8239886164665222, "learning_rate": 0.0002, "epoch": 3.1687009942607713, "step": 19600}, {"loss": 0.6068, "grad_norm": 0.9327243566513062, "learning_rate": 0.0002, "epoch": 3.1703176784415166, "step": 19610}, {"loss": 0.572, "grad_norm": 0.9662560224533081, "learning_rate": 0.0002, "epoch": 3.171934362622262, "step": 19620}, {"loss": 0.5988, "grad_norm": 0.9183341860771179, "learning_rate": 0.0002, "epoch": 3.173551046803007, "step": 19630}, {"loss": 0.5909, "grad_norm": 0.875066876411438, "learning_rate": 0.0002, "epoch": 3.1751677309837523, "step": 19640}, {"loss": 0.5956, "grad_norm": 0.8567508459091187, "learning_rate": 0.0002, "epoch": 3.1767844151644975, "step": 19650}, {"loss": 0.5805, "grad_norm": 0.6805780529975891, "learning_rate": 0.0002, "epoch": 3.1784010993452427, "step": 19660}, {"loss": 0.6204, "grad_norm": 0.8776944279670715, "learning_rate": 0.0002, "epoch": 3.1800177835259884, "step": 19670}, {"loss": 0.6108, "grad_norm": 0.9036329984664917, "learning_rate": 0.0002, "epoch": 3.1816344677067336, "step": 19680}, {"loss": 0.6238, "grad_norm": 0.8527372479438782, "learning_rate": 0.0002, "epoch": 3.183251151887479, "step": 19690}, {"loss": 0.6089, "grad_norm": 1.1045585870742798, "learning_rate": 0.0002, "epoch": 3.184867836068224, "step": 19700}, {"loss": 0.5491, "grad_norm": 0.9213830828666687, "learning_rate": 0.0002, "epoch": 3.1864845202489693, "step": 19710}, {"loss": 0.618, "grad_norm": 0.8865814805030823, "learning_rate": 0.0002, "epoch": 3.1881012044297146, "step": 19720}, {"loss": 0.5785, "grad_norm": 0.7939388751983643, "learning_rate": 0.0002, "epoch": 3.18971788861046, "step": 19730}, {"loss": 0.5682, "grad_norm": 0.6966729760169983, "learning_rate": 0.0002, "epoch": 3.191334572791205, "step": 19740}, {"loss": 0.5839, "grad_norm": 0.8023673295974731, "learning_rate": 0.0002, "epoch": 3.1929512569719507, "step": 19750}, {"loss": 0.6267, "grad_norm": 0.7992037534713745, "learning_rate": 0.0002, "epoch": 3.194567941152696, "step": 19760}, {"loss": 0.6141, "grad_norm": 0.7412247657775879, "learning_rate": 0.0002, "epoch": 3.196184625333441, "step": 19770}, {"loss": 0.6179, "grad_norm": 0.9598729014396667, "learning_rate": 0.0002, "epoch": 3.1978013095141864, "step": 19780}, {"loss": 0.5685, "grad_norm": 0.8331366777420044, "learning_rate": 0.0002, "epoch": 3.1994179936949316, "step": 19790}, {"loss": 0.6104, "grad_norm": 0.8939169645309448, "learning_rate": 0.0002, "epoch": 3.201034677875677, "step": 19800}, {"loss": 0.6147, "grad_norm": 0.9219734072685242, "learning_rate": 0.0002, "epoch": 3.202651362056422, "step": 19810}, {"loss": 0.6051, "grad_norm": 0.869490385055542, "learning_rate": 0.0002, "epoch": 3.2042680462371678, "step": 19820}, {"loss": 0.5946, "grad_norm": 0.8989706635475159, "learning_rate": 0.0002, "epoch": 3.205884730417913, "step": 19830}, {"loss": 0.5866, "grad_norm": 0.8477165102958679, "learning_rate": 0.0002, "epoch": 3.2075014145986582, "step": 19840}, {"loss": 0.6176, "grad_norm": 0.8720678687095642, "learning_rate": 0.0002, "epoch": 3.2091180987794035, "step": 19850}, {"loss": 0.5694, "grad_norm": 0.861406683921814, "learning_rate": 0.0002, "epoch": 3.2107347829601487, "step": 19860}, {"loss": 0.6264, "grad_norm": 0.8228686451911926, "learning_rate": 0.0002, "epoch": 3.212351467140894, "step": 19870}, {"loss": 0.625, "grad_norm": 0.7936596870422363, "learning_rate": 0.0002, "epoch": 3.213968151321639, "step": 19880}, {"loss": 0.5698, "grad_norm": 1.097377896308899, "learning_rate": 0.0002, "epoch": 3.2155848355023844, "step": 19890}, {"loss": 0.6725, "grad_norm": 0.9544782638549805, "learning_rate": 0.0002, "epoch": 3.21720151968313, "step": 19900}, {"loss": 0.6022, "grad_norm": 0.8240751624107361, "learning_rate": 0.0002, "epoch": 3.2188182038638753, "step": 19910}, {"loss": 0.5659, "grad_norm": 0.8332096338272095, "learning_rate": 0.0002, "epoch": 3.2204348880446205, "step": 19920}, {"loss": 0.6274, "grad_norm": 1.0954567193984985, "learning_rate": 0.0002, "epoch": 3.2220515722253658, "step": 19930}, {"loss": 0.652, "grad_norm": 0.7790525555610657, "learning_rate": 0.0002, "epoch": 3.223668256406111, "step": 19940}, {"loss": 0.5986, "grad_norm": 0.7966814041137695, "learning_rate": 0.0002, "epoch": 3.225284940586856, "step": 19950}, {"loss": 0.5911, "grad_norm": 0.9751881957054138, "learning_rate": 0.0002, "epoch": 3.2269016247676015, "step": 19960}, {"loss": 0.6071, "grad_norm": 0.9856047630310059, "learning_rate": 0.0002, "epoch": 3.228518308948347, "step": 19970}, {"loss": 0.5837, "grad_norm": 1.3062353134155273, "learning_rate": 0.0002, "epoch": 3.2301349931290924, "step": 19980}, {"loss": 0.6588, "grad_norm": 0.9510692358016968, "learning_rate": 0.0002, "epoch": 3.2317516773098376, "step": 19990}, {"loss": 0.6264, "grad_norm": 0.8630342483520508, "learning_rate": 0.0002, "epoch": 3.233368361490583, "step": 20000}, {"loss": 0.6073, "grad_norm": 0.8966519236564636, "learning_rate": 0.0002, "epoch": 3.234985045671328, "step": 20010}, {"loss": 0.612, "grad_norm": 0.7093510627746582, "learning_rate": 0.0002, "epoch": 3.2366017298520733, "step": 20020}, {"loss": 0.585, "grad_norm": 0.7771096229553223, "learning_rate": 0.0002, "epoch": 3.2382184140328185, "step": 20030}, {"loss": 0.5821, "grad_norm": 0.841058075428009, "learning_rate": 0.0002, "epoch": 3.2398350982135637, "step": 20040}, {"loss": 0.6519, "grad_norm": 0.909712553024292, "learning_rate": 0.0002, "epoch": 3.2414517823943094, "step": 20050}, {"loss": 0.6089, "grad_norm": 0.8321019411087036, "learning_rate": 0.0002, "epoch": 3.2430684665750547, "step": 20060}, {"loss": 0.6115, "grad_norm": 0.779901921749115, "learning_rate": 0.0002, "epoch": 3.2446851507558, "step": 20070}, {"loss": 0.6107, "grad_norm": 0.6249170303344727, "learning_rate": 0.0002, "epoch": 3.246301834936545, "step": 20080}, {"loss": 0.603, "grad_norm": 0.8000940680503845, "learning_rate": 0.0002, "epoch": 3.2479185191172903, "step": 20090}, {"loss": 0.6273, "grad_norm": 0.7627735137939453, "learning_rate": 0.0002, "epoch": 3.2495352032980356, "step": 20100}, {"loss": 0.6223, "grad_norm": 0.8780747056007385, "learning_rate": 0.0002, "epoch": 3.2511518874787813, "step": 20110}, {"loss": 0.5969, "grad_norm": 0.772037148475647, "learning_rate": 0.0002, "epoch": 3.2527685716595265, "step": 20120}, {"loss": 0.5843, "grad_norm": 1.0086580514907837, "learning_rate": 0.0002, "epoch": 3.2543852558402717, "step": 20130}, {"loss": 0.5777, "grad_norm": 0.9360289573669434, "learning_rate": 0.0002, "epoch": 3.256001940021017, "step": 20140}, {"loss": 0.5777, "grad_norm": 1.2099586725234985, "learning_rate": 0.0002, "epoch": 3.257618624201762, "step": 20150}, {"loss": 0.624, "grad_norm": 0.8368481397628784, "learning_rate": 0.0002, "epoch": 3.2592353083825074, "step": 20160}, {"loss": 0.5626, "grad_norm": 0.7391039133071899, "learning_rate": 0.0002, "epoch": 3.2608519925632526, "step": 20170}, {"loss": 0.6041, "grad_norm": 0.9122273325920105, "learning_rate": 0.0002, "epoch": 3.262468676743998, "step": 20180}, {"loss": 0.5868, "grad_norm": 0.8502281904220581, "learning_rate": 0.0002, "epoch": 3.264085360924743, "step": 20190}, {"loss": 0.5841, "grad_norm": 1.0926852226257324, "learning_rate": 0.0002, "epoch": 3.265702045105489, "step": 20200}, {"loss": 0.6027, "grad_norm": 0.7902828454971313, "learning_rate": 0.0002, "epoch": 3.267318729286234, "step": 20210}, {"loss": 0.6089, "grad_norm": 0.8724729418754578, "learning_rate": 0.0002, "epoch": 3.2689354134669792, "step": 20220}, {"loss": 0.6242, "grad_norm": 0.8469277024269104, "learning_rate": 0.0002, "epoch": 3.2705520976477245, "step": 20230}, {"loss": 0.644, "grad_norm": 0.8865092992782593, "learning_rate": 0.0002, "epoch": 3.2721687818284697, "step": 20240}, {"loss": 0.6464, "grad_norm": 1.0979334115982056, "learning_rate": 0.0002, "epoch": 3.273785466009215, "step": 20250}, {"loss": 0.647, "grad_norm": 1.0860793590545654, "learning_rate": 0.0002, "epoch": 3.2754021501899606, "step": 20260}, {"loss": 0.6105, "grad_norm": 0.981745183467865, "learning_rate": 0.0002, "epoch": 3.277018834370706, "step": 20270}, {"loss": 0.627, "grad_norm": 0.9155020713806152, "learning_rate": 0.0002, "epoch": 3.278635518551451, "step": 20280}, {"loss": 0.5899, "grad_norm": 0.8436718583106995, "learning_rate": 0.0002, "epoch": 3.2802522027321963, "step": 20290}, {"loss": 0.6371, "grad_norm": 1.0329409837722778, "learning_rate": 0.0002, "epoch": 3.2818688869129415, "step": 20300}, {"loss": 0.6, "grad_norm": 0.9876394271850586, "learning_rate": 0.0002, "epoch": 3.2834855710936868, "step": 20310}, {"loss": 0.5463, "grad_norm": 0.8052917718887329, "learning_rate": 0.0002, "epoch": 3.285102255274432, "step": 20320}, {"loss": 0.5949, "grad_norm": 0.8390680551528931, "learning_rate": 0.0002, "epoch": 3.2867189394551772, "step": 20330}, {"loss": 0.6492, "grad_norm": 0.9515735507011414, "learning_rate": 0.0002, "epoch": 3.288335623635923, "step": 20340}, {"loss": 0.596, "grad_norm": 0.8028870224952698, "learning_rate": 0.0002, "epoch": 3.289952307816668, "step": 20350}, {"loss": 0.634, "grad_norm": 0.862592339515686, "learning_rate": 0.0002, "epoch": 3.2915689919974134, "step": 20360}, {"loss": 0.6345, "grad_norm": 0.7451621890068054, "learning_rate": 0.0002, "epoch": 3.2931856761781586, "step": 20370}, {"loss": 0.6458, "grad_norm": 0.8966776728630066, "learning_rate": 0.0002, "epoch": 3.294802360358904, "step": 20380}, {"loss": 0.5967, "grad_norm": 0.9289216995239258, "learning_rate": 0.0002, "epoch": 3.296419044539649, "step": 20390}, {"loss": 0.6599, "grad_norm": 0.9649626612663269, "learning_rate": 0.0002, "epoch": 3.2980357287203943, "step": 20400}, {"loss": 0.5781, "grad_norm": 1.1953798532485962, "learning_rate": 0.0002, "epoch": 3.29965241290114, "step": 20410}, {"loss": 0.5997, "grad_norm": 0.8929083943367004, "learning_rate": 0.0002, "epoch": 3.301269097081885, "step": 20420}, {"loss": 0.597, "grad_norm": 0.8922014236450195, "learning_rate": 0.0002, "epoch": 3.3028857812626304, "step": 20430}, {"loss": 0.5766, "grad_norm": 0.9754860401153564, "learning_rate": 0.0002, "epoch": 3.3045024654433757, "step": 20440}, {"loss": 0.5653, "grad_norm": 0.8873140215873718, "learning_rate": 0.0002, "epoch": 3.306119149624121, "step": 20450}, {"loss": 0.6138, "grad_norm": 0.857271671295166, "learning_rate": 0.0002, "epoch": 3.307735833804866, "step": 20460}, {"loss": 0.633, "grad_norm": 0.9022141098976135, "learning_rate": 0.0002, "epoch": 3.3093525179856114, "step": 20470}, {"loss": 0.6654, "grad_norm": 0.8614798188209534, "learning_rate": 0.0002, "epoch": 3.3109692021663566, "step": 20480}, {"loss": 0.6254, "grad_norm": 0.8838164210319519, "learning_rate": 0.0002, "epoch": 3.3125858863471023, "step": 20490}, {"loss": 0.5849, "grad_norm": 0.8709736466407776, "learning_rate": 0.0002, "epoch": 3.3142025705278475, "step": 20500}, {"loss": 0.6146, "grad_norm": 0.9533300995826721, "learning_rate": 0.0002, "epoch": 3.3158192547085927, "step": 20510}, {"loss": 0.6029, "grad_norm": 0.8259269595146179, "learning_rate": 0.0002, "epoch": 3.317435938889338, "step": 20520}, {"loss": 0.6268, "grad_norm": 0.8607608079910278, "learning_rate": 0.0002, "epoch": 3.319052623070083, "step": 20530}, {"loss": 0.5676, "grad_norm": 1.0863020420074463, "learning_rate": 0.0002, "epoch": 3.3206693072508284, "step": 20540}, {"loss": 0.6412, "grad_norm": 1.011489987373352, "learning_rate": 0.0002, "epoch": 3.3222859914315737, "step": 20550}, {"loss": 0.6247, "grad_norm": 0.6952177882194519, "learning_rate": 0.0002, "epoch": 3.3239026756123193, "step": 20560}, {"loss": 0.6229, "grad_norm": 0.9638974070549011, "learning_rate": 0.0002, "epoch": 3.3255193597930646, "step": 20570}, {"loss": 0.5882, "grad_norm": 1.0310138463974, "learning_rate": 0.0002, "epoch": 3.32713604397381, "step": 20580}, {"loss": 0.594, "grad_norm": 0.9371318221092224, "learning_rate": 0.0002, "epoch": 3.328752728154555, "step": 20590}, {"loss": 0.6137, "grad_norm": 0.8756691813468933, "learning_rate": 0.0002, "epoch": 3.3303694123353003, "step": 20600}, {"loss": 0.5994, "grad_norm": 1.054175853729248, "learning_rate": 0.0002, "epoch": 3.3319860965160455, "step": 20610}, {"loss": 0.6169, "grad_norm": 0.9074128270149231, "learning_rate": 0.0002, "epoch": 3.3336027806967907, "step": 20620}, {"loss": 0.6138, "grad_norm": 0.906900942325592, "learning_rate": 0.0002, "epoch": 3.335219464877536, "step": 20630}, {"loss": 0.571, "grad_norm": 0.8689333200454712, "learning_rate": 0.0002, "epoch": 3.3368361490582816, "step": 20640}, {"loss": 0.6079, "grad_norm": 0.9889747500419617, "learning_rate": 0.0002, "epoch": 3.338452833239027, "step": 20650}, {"loss": 0.6073, "grad_norm": 1.0685805082321167, "learning_rate": 0.0002, "epoch": 3.340069517419772, "step": 20660}, {"loss": 0.6091, "grad_norm": 0.7495010495185852, "learning_rate": 0.0002, "epoch": 3.3416862016005173, "step": 20670}, {"loss": 0.5883, "grad_norm": 0.8747848272323608, "learning_rate": 0.0002, "epoch": 3.3433028857812626, "step": 20680}, {"loss": 0.604, "grad_norm": 0.9762673377990723, "learning_rate": 0.0002, "epoch": 3.344919569962008, "step": 20690}, {"loss": 0.6784, "grad_norm": 1.0284489393234253, "learning_rate": 0.0002, "epoch": 3.346536254142753, "step": 20700}, {"loss": 0.6464, "grad_norm": 0.7293812036514282, "learning_rate": 0.0002, "epoch": 3.3481529383234987, "step": 20710}, {"loss": 0.609, "grad_norm": 0.8330199122428894, "learning_rate": 0.0002, "epoch": 3.349769622504244, "step": 20720}, {"loss": 0.5729, "grad_norm": 0.9808499217033386, "learning_rate": 0.0002, "epoch": 3.351386306684989, "step": 20730}, {"loss": 0.6315, "grad_norm": 0.9508825540542603, "learning_rate": 0.0002, "epoch": 3.3530029908657344, "step": 20740}, {"loss": 0.5965, "grad_norm": 0.790483832359314, "learning_rate": 0.0002, "epoch": 3.3546196750464796, "step": 20750}, {"loss": 0.6327, "grad_norm": 1.022793173789978, "learning_rate": 0.0002, "epoch": 3.356236359227225, "step": 20760}, {"loss": 0.6439, "grad_norm": 0.8318950533866882, "learning_rate": 0.0002, "epoch": 3.35785304340797, "step": 20770}, {"loss": 0.6037, "grad_norm": 0.7980858087539673, "learning_rate": 0.0002, "epoch": 3.3594697275887153, "step": 20780}, {"loss": 0.6746, "grad_norm": 0.8114802241325378, "learning_rate": 0.0002, "epoch": 3.361086411769461, "step": 20790}, {"loss": 0.6017, "grad_norm": 0.8522519469261169, "learning_rate": 0.0002, "epoch": 3.3627030959502062, "step": 20800}, {"loss": 0.5864, "grad_norm": 0.9142431616783142, "learning_rate": 0.0002, "epoch": 3.3643197801309515, "step": 20810}, {"loss": 0.6331, "grad_norm": 0.771170437335968, "learning_rate": 0.0002, "epoch": 3.3659364643116967, "step": 20820}, {"loss": 0.5879, "grad_norm": 1.0628231763839722, "learning_rate": 0.0002, "epoch": 3.367553148492442, "step": 20830}, {"loss": 0.6533, "grad_norm": 0.9384352564811707, "learning_rate": 0.0002, "epoch": 3.369169832673187, "step": 20840}, {"loss": 0.6292, "grad_norm": 1.1286591291427612, "learning_rate": 0.0002, "epoch": 3.370786516853933, "step": 20850}, {"loss": 0.5986, "grad_norm": 1.1349513530731201, "learning_rate": 0.0002, "epoch": 3.372403201034678, "step": 20860}, {"loss": 0.6413, "grad_norm": 1.0127464532852173, "learning_rate": 0.0002, "epoch": 3.3740198852154233, "step": 20870}, {"loss": 0.6414, "grad_norm": 0.9111971855163574, "learning_rate": 0.0002, "epoch": 3.3756365693961685, "step": 20880}, {"loss": 0.6101, "grad_norm": 0.871356725692749, "learning_rate": 0.0002, "epoch": 3.3772532535769137, "step": 20890}, {"loss": 0.5995, "grad_norm": 0.7774117588996887, "learning_rate": 0.0002, "epoch": 3.378869937757659, "step": 20900}, {"loss": 0.6062, "grad_norm": 1.0089964866638184, "learning_rate": 0.0002, "epoch": 3.380486621938404, "step": 20910}, {"loss": 0.5908, "grad_norm": 0.7855867147445679, "learning_rate": 0.0002, "epoch": 3.3821033061191494, "step": 20920}, {"loss": 0.6373, "grad_norm": 1.3713710308074951, "learning_rate": 0.0002, "epoch": 3.3837199902998947, "step": 20930}, {"loss": 0.6627, "grad_norm": 0.8599116206169128, "learning_rate": 0.0002, "epoch": 3.3853366744806404, "step": 20940}, {"loss": 0.6224, "grad_norm": 0.9392673373222351, "learning_rate": 0.0002, "epoch": 3.3869533586613856, "step": 20950}, {"loss": 0.5855, "grad_norm": 0.8764075040817261, "learning_rate": 0.0002, "epoch": 3.388570042842131, "step": 20960}, {"loss": 0.5734, "grad_norm": 0.8240136504173279, "learning_rate": 0.0002, "epoch": 3.390186727022876, "step": 20970}, {"loss": 0.5783, "grad_norm": 1.0982369184494019, "learning_rate": 0.0002, "epoch": 3.3918034112036213, "step": 20980}, {"loss": 0.5451, "grad_norm": 1.0599013566970825, "learning_rate": 0.0002, "epoch": 3.3934200953843665, "step": 20990}, {"loss": 0.6356, "grad_norm": 0.895438015460968, "learning_rate": 0.0002, "epoch": 3.395036779565112, "step": 21000}, {"loss": 0.6065, "grad_norm": 0.6974841356277466, "learning_rate": 0.0002, "epoch": 3.3966534637458574, "step": 21010}, {"loss": 0.5704, "grad_norm": 0.9571719765663147, "learning_rate": 0.0002, "epoch": 3.3982701479266026, "step": 21020}, {"loss": 0.679, "grad_norm": 0.831912636756897, "learning_rate": 0.0002, "epoch": 3.399886832107348, "step": 21030}, {"loss": 0.6051, "grad_norm": 0.831936240196228, "learning_rate": 0.0002, "epoch": 3.401503516288093, "step": 21040}, {"loss": 0.5857, "grad_norm": 0.7388373613357544, "learning_rate": 0.0002, "epoch": 3.4031202004688383, "step": 21050}, {"loss": 0.6245, "grad_norm": 0.938667356967926, "learning_rate": 0.0002, "epoch": 3.4047368846495836, "step": 21060}, {"loss": 0.6121, "grad_norm": 0.9202313423156738, "learning_rate": 0.0002, "epoch": 3.406353568830329, "step": 21070}, {"loss": 0.6388, "grad_norm": 0.9888381958007812, "learning_rate": 0.0002, "epoch": 3.4079702530110745, "step": 21080}, {"loss": 0.6245, "grad_norm": 0.8526970744132996, "learning_rate": 0.0002, "epoch": 3.4095869371918197, "step": 21090}, {"loss": 0.5914, "grad_norm": 0.7939383387565613, "learning_rate": 0.0002, "epoch": 3.411203621372565, "step": 21100}, {"loss": 0.6066, "grad_norm": 0.9986352920532227, "learning_rate": 0.0002, "epoch": 3.41282030555331, "step": 21110}, {"loss": 0.5947, "grad_norm": 0.8895300030708313, "learning_rate": 0.0002, "epoch": 3.4144369897340554, "step": 21120}, {"loss": 0.6264, "grad_norm": 0.9559482932090759, "learning_rate": 0.0002, "epoch": 3.4160536739148006, "step": 21130}, {"loss": 0.6491, "grad_norm": 0.8351506590843201, "learning_rate": 0.0002, "epoch": 3.417670358095546, "step": 21140}, {"loss": 0.567, "grad_norm": 0.8224456906318665, "learning_rate": 0.0002, "epoch": 3.4192870422762915, "step": 21150}, {"loss": 0.5871, "grad_norm": 1.0110299587249756, "learning_rate": 0.0002, "epoch": 3.4209037264570368, "step": 21160}, {"loss": 0.6116, "grad_norm": 0.82564777135849, "learning_rate": 0.0002, "epoch": 3.422520410637782, "step": 21170}, {"loss": 0.595, "grad_norm": 1.004738688468933, "learning_rate": 0.0002, "epoch": 3.4241370948185272, "step": 21180}, {"loss": 0.6286, "grad_norm": 0.7545676827430725, "learning_rate": 0.0002, "epoch": 3.4257537789992725, "step": 21190}, {"loss": 0.5868, "grad_norm": 0.8918704390525818, "learning_rate": 0.0002, "epoch": 3.4273704631800177, "step": 21200}, {"loss": 0.6542, "grad_norm": 0.8336876034736633, "learning_rate": 0.0002, "epoch": 3.428987147360763, "step": 21210}, {"loss": 0.5824, "grad_norm": 0.8928771018981934, "learning_rate": 0.0002, "epoch": 3.430603831541508, "step": 21220}, {"loss": 0.6468, "grad_norm": 0.7663705945014954, "learning_rate": 0.0002, "epoch": 3.432220515722254, "step": 21230}, {"loss": 0.6693, "grad_norm": 0.8392598628997803, "learning_rate": 0.0002, "epoch": 3.433837199902999, "step": 21240}, {"loss": 0.5971, "grad_norm": 0.8819600343704224, "learning_rate": 0.0002, "epoch": 3.4354538840837443, "step": 21250}, {"loss": 0.6791, "grad_norm": 0.9124642014503479, "learning_rate": 0.0002, "epoch": 3.4370705682644895, "step": 21260}, {"loss": 0.5925, "grad_norm": 0.8329763412475586, "learning_rate": 0.0002, "epoch": 3.4386872524452348, "step": 21270}, {"loss": 0.6541, "grad_norm": 0.9982839822769165, "learning_rate": 0.0002, "epoch": 3.44030393662598, "step": 21280}, {"loss": 0.6441, "grad_norm": 0.9105954766273499, "learning_rate": 0.0002, "epoch": 3.4419206208067252, "step": 21290}, {"loss": 0.6028, "grad_norm": 0.8182359337806702, "learning_rate": 0.0002, "epoch": 3.443537304987471, "step": 21300}, {"loss": 0.5991, "grad_norm": 1.0568904876708984, "learning_rate": 0.0002, "epoch": 3.445153989168216, "step": 21310}, {"loss": 0.6117, "grad_norm": 0.968539834022522, "learning_rate": 0.0002, "epoch": 3.4467706733489614, "step": 21320}, {"loss": 0.6219, "grad_norm": 0.8774511218070984, "learning_rate": 0.0002, "epoch": 3.4483873575297066, "step": 21330}, {"loss": 0.6438, "grad_norm": 0.7598156332969666, "learning_rate": 0.0002, "epoch": 3.450004041710452, "step": 21340}, {"loss": 0.6033, "grad_norm": 1.1012897491455078, "learning_rate": 0.0002, "epoch": 3.451620725891197, "step": 21350}, {"loss": 0.6137, "grad_norm": 0.8040637373924255, "learning_rate": 0.0002, "epoch": 3.4532374100719423, "step": 21360}, {"loss": 0.6173, "grad_norm": 0.8497496247291565, "learning_rate": 0.0002, "epoch": 3.4548540942526875, "step": 21370}, {"loss": 0.6005, "grad_norm": 0.8429915904998779, "learning_rate": 0.0002, "epoch": 3.456470778433433, "step": 21380}, {"loss": 0.6182, "grad_norm": 0.8107112646102905, "learning_rate": 0.0002, "epoch": 3.4580874626141784, "step": 21390}, {"loss": 0.6109, "grad_norm": 1.00872004032135, "learning_rate": 0.0002, "epoch": 3.4597041467949237, "step": 21400}, {"loss": 0.5712, "grad_norm": 0.8266542553901672, "learning_rate": 0.0002, "epoch": 3.461320830975669, "step": 21410}, {"loss": 0.6457, "grad_norm": 0.8972568511962891, "learning_rate": 0.0002, "epoch": 3.462937515156414, "step": 21420}, {"loss": 0.6081, "grad_norm": 1.0781476497650146, "learning_rate": 0.0002, "epoch": 3.4645541993371594, "step": 21430}, {"loss": 0.6303, "grad_norm": 0.9571592807769775, "learning_rate": 0.0002, "epoch": 3.4661708835179046, "step": 21440}, {"loss": 0.6309, "grad_norm": 0.881547212600708, "learning_rate": 0.0002, "epoch": 3.4677875676986503, "step": 21450}, {"loss": 0.6076, "grad_norm": 0.6955338716506958, "learning_rate": 0.0002, "epoch": 3.4694042518793955, "step": 21460}, {"loss": 0.6205, "grad_norm": 0.901187539100647, "learning_rate": 0.0002, "epoch": 3.4710209360601407, "step": 21470}, {"loss": 0.639, "grad_norm": 0.7063511610031128, "learning_rate": 0.0002, "epoch": 3.472637620240886, "step": 21480}, {"loss": 0.6154, "grad_norm": 0.8462792038917542, "learning_rate": 0.0002, "epoch": 3.474254304421631, "step": 21490}, {"loss": 0.61, "grad_norm": 1.1861060857772827, "learning_rate": 0.0002, "epoch": 3.4758709886023764, "step": 21500}, {"loss": 0.6586, "grad_norm": 0.70503169298172, "learning_rate": 0.0002, "epoch": 3.4774876727831217, "step": 21510}, {"loss": 0.6475, "grad_norm": 0.9650066494941711, "learning_rate": 0.0002, "epoch": 3.479104356963867, "step": 21520}, {"loss": 0.6452, "grad_norm": 1.0266852378845215, "learning_rate": 0.0002, "epoch": 3.4807210411446126, "step": 21530}, {"loss": 0.6553, "grad_norm": 0.956372857093811, "learning_rate": 0.0002, "epoch": 3.482337725325358, "step": 21540}, {"loss": 0.6667, "grad_norm": 0.8848432898521423, "learning_rate": 0.0002, "epoch": 3.483954409506103, "step": 21550}, {"loss": 0.6375, "grad_norm": 1.0805351734161377, "learning_rate": 0.0002, "epoch": 3.4855710936868483, "step": 21560}, {"loss": 0.6958, "grad_norm": 0.9279725551605225, "learning_rate": 0.0002, "epoch": 3.4871877778675935, "step": 21570}, {"loss": 0.6354, "grad_norm": 0.9049562215805054, "learning_rate": 0.0002, "epoch": 3.4888044620483387, "step": 21580}, {"loss": 0.6071, "grad_norm": 0.9619429111480713, "learning_rate": 0.0002, "epoch": 3.4904211462290844, "step": 21590}, {"loss": 0.5927, "grad_norm": 0.8508906960487366, "learning_rate": 0.0002, "epoch": 3.4920378304098296, "step": 21600}, {"loss": 0.6115, "grad_norm": 0.8692502379417419, "learning_rate": 0.0002, "epoch": 3.493654514590575, "step": 21610}, {"loss": 0.5878, "grad_norm": 0.8187332153320312, "learning_rate": 0.0002, "epoch": 3.49527119877132, "step": 21620}, {"loss": 0.5874, "grad_norm": 1.145400047302246, "learning_rate": 0.0002, "epoch": 3.4968878829520653, "step": 21630}, {"loss": 0.6313, "grad_norm": 0.8281388282775879, "learning_rate": 0.0002, "epoch": 3.4985045671328105, "step": 21640}, {"loss": 0.6624, "grad_norm": 0.82256019115448, "learning_rate": 0.0002, "epoch": 3.500121251313556, "step": 21650}, {"loss": 0.6346, "grad_norm": 0.9315484762191772, "learning_rate": 0.0002, "epoch": 3.501737935494301, "step": 21660}, {"loss": 0.6086, "grad_norm": 0.7626111507415771, "learning_rate": 0.0002, "epoch": 3.5033546196750462, "step": 21670}, {"loss": 0.6177, "grad_norm": 0.9275059103965759, "learning_rate": 0.0002, "epoch": 3.504971303855792, "step": 21680}, {"loss": 0.64, "grad_norm": 0.7906724810600281, "learning_rate": 0.0002, "epoch": 3.506587988036537, "step": 21690}, {"loss": 0.6015, "grad_norm": 0.8289761543273926, "learning_rate": 0.0002, "epoch": 3.5082046722172824, "step": 21700}, {"loss": 0.6246, "grad_norm": 0.8316431045532227, "learning_rate": 0.0002, "epoch": 3.5098213563980276, "step": 21710}, {"loss": 0.619, "grad_norm": 1.0451812744140625, "learning_rate": 0.0002, "epoch": 3.511438040578773, "step": 21720}, {"loss": 0.632, "grad_norm": 0.928252637386322, "learning_rate": 0.0002, "epoch": 3.513054724759518, "step": 21730}, {"loss": 0.6062, "grad_norm": 0.7985895276069641, "learning_rate": 0.0002, "epoch": 3.5146714089402638, "step": 21740}, {"loss": 0.6463, "grad_norm": 0.6740974187850952, "learning_rate": 0.0002, "epoch": 3.516288093121009, "step": 21750}, {"loss": 0.6138, "grad_norm": 0.8482223749160767, "learning_rate": 0.0002, "epoch": 3.517904777301754, "step": 21760}, {"loss": 0.6277, "grad_norm": 0.889947772026062, "learning_rate": 0.0002, "epoch": 3.5195214614824994, "step": 21770}, {"loss": 0.6174, "grad_norm": 0.8304598927497864, "learning_rate": 0.0002, "epoch": 3.5211381456632447, "step": 21780}, {"loss": 0.6156, "grad_norm": 0.8002981543540955, "learning_rate": 0.0002, "epoch": 3.52275482984399, "step": 21790}, {"loss": 0.5896, "grad_norm": 0.8115083575248718, "learning_rate": 0.0002, "epoch": 3.524371514024735, "step": 21800}, {"loss": 0.6041, "grad_norm": 0.9715048670768738, "learning_rate": 0.0002, "epoch": 3.5259881982054804, "step": 21810}, {"loss": 0.6715, "grad_norm": 1.0910786390304565, "learning_rate": 0.0002, "epoch": 3.5276048823862256, "step": 21820}, {"loss": 0.6543, "grad_norm": 0.8438942432403564, "learning_rate": 0.0002, "epoch": 3.5292215665669713, "step": 21830}, {"loss": 0.6509, "grad_norm": 0.8813382983207703, "learning_rate": 0.0002, "epoch": 3.5308382507477165, "step": 21840}, {"loss": 0.6049, "grad_norm": 0.7092908024787903, "learning_rate": 0.0002, "epoch": 3.5324549349284617, "step": 21850}, {"loss": 0.5678, "grad_norm": 0.8332187533378601, "learning_rate": 0.0002, "epoch": 3.534071619109207, "step": 21860}, {"loss": 0.5896, "grad_norm": 0.8958209156990051, "learning_rate": 0.0002, "epoch": 3.535688303289952, "step": 21870}, {"loss": 0.6476, "grad_norm": 0.824138879776001, "learning_rate": 0.0002, "epoch": 3.5373049874706974, "step": 21880}, {"loss": 0.6022, "grad_norm": 0.8375158309936523, "learning_rate": 0.0002, "epoch": 3.538921671651443, "step": 21890}, {"loss": 0.6019, "grad_norm": 1.0274608135223389, "learning_rate": 0.0002, "epoch": 3.5405383558321883, "step": 21900}, {"loss": 0.6194, "grad_norm": 0.7088932394981384, "learning_rate": 0.0002, "epoch": 3.5421550400129336, "step": 21910}, {"loss": 0.6554, "grad_norm": 0.8172445297241211, "learning_rate": 0.0002, "epoch": 3.543771724193679, "step": 21920}, {"loss": 0.6711, "grad_norm": 0.9904135465621948, "learning_rate": 0.0002, "epoch": 3.545388408374424, "step": 21930}, {"loss": 0.6001, "grad_norm": 0.9900432229042053, "learning_rate": 0.0002, "epoch": 3.5470050925551693, "step": 21940}, {"loss": 0.6195, "grad_norm": 0.8963301181793213, "learning_rate": 0.0002, "epoch": 3.5486217767359145, "step": 21950}, {"loss": 0.5972, "grad_norm": 0.8551464676856995, "learning_rate": 0.0002, "epoch": 3.5502384609166597, "step": 21960}, {"loss": 0.6206, "grad_norm": 1.0916603803634644, "learning_rate": 0.0002, "epoch": 3.551855145097405, "step": 21970}, {"loss": 0.6523, "grad_norm": 0.841598391532898, "learning_rate": 0.0002, "epoch": 3.5534718292781506, "step": 21980}, {"loss": 0.617, "grad_norm": 0.8566757440567017, "learning_rate": 0.0002, "epoch": 3.555088513458896, "step": 21990}, {"loss": 0.6192, "grad_norm": 1.0145052671432495, "learning_rate": 0.0002, "epoch": 3.556705197639641, "step": 22000}, {"loss": 0.6173, "grad_norm": 0.9293754696846008, "learning_rate": 0.0002, "epoch": 3.5583218818203863, "step": 22010}, {"loss": 0.612, "grad_norm": 0.9568536281585693, "learning_rate": 0.0002, "epoch": 3.5599385660011316, "step": 22020}, {"loss": 0.641, "grad_norm": 0.8613139986991882, "learning_rate": 0.0002, "epoch": 3.5615552501818772, "step": 22030}, {"loss": 0.6496, "grad_norm": 0.8179237246513367, "learning_rate": 0.0002, "epoch": 3.5631719343626225, "step": 22040}, {"loss": 0.574, "grad_norm": 0.9059830904006958, "learning_rate": 0.0002, "epoch": 3.5647886185433677, "step": 22050}, {"loss": 0.6448, "grad_norm": 1.0068252086639404, "learning_rate": 0.0002, "epoch": 3.566405302724113, "step": 22060}, {"loss": 0.6239, "grad_norm": 0.9682072997093201, "learning_rate": 0.0002, "epoch": 3.568021986904858, "step": 22070}, {"loss": 0.6808, "grad_norm": 0.8514005541801453, "learning_rate": 0.0002, "epoch": 3.5696386710856034, "step": 22080}, {"loss": 0.5956, "grad_norm": 0.8327770829200745, "learning_rate": 0.0002, "epoch": 3.5712553552663486, "step": 22090}, {"loss": 0.5976, "grad_norm": 1.024976372718811, "learning_rate": 0.0002, "epoch": 3.572872039447094, "step": 22100}, {"loss": 0.624, "grad_norm": 0.7721174955368042, "learning_rate": 0.0002, "epoch": 3.574488723627839, "step": 22110}, {"loss": 0.5896, "grad_norm": 1.0351054668426514, "learning_rate": 0.0002, "epoch": 3.5761054078085843, "step": 22120}, {"loss": 0.6379, "grad_norm": 0.9680907130241394, "learning_rate": 0.0002, "epoch": 3.57772209198933, "step": 22130}, {"loss": 0.6194, "grad_norm": 0.8016974925994873, "learning_rate": 0.0002, "epoch": 3.5793387761700752, "step": 22140}, {"loss": 0.6387, "grad_norm": 1.0109003782272339, "learning_rate": 0.0002, "epoch": 3.5809554603508205, "step": 22150}, {"loss": 0.6368, "grad_norm": 1.0473392009735107, "learning_rate": 0.0002, "epoch": 3.5825721445315657, "step": 22160}, {"loss": 0.6353, "grad_norm": 0.8686613440513611, "learning_rate": 0.0002, "epoch": 3.584188828712311, "step": 22170}, {"loss": 0.5791, "grad_norm": 0.869149923324585, "learning_rate": 0.0002, "epoch": 3.5858055128930566, "step": 22180}, {"loss": 0.5895, "grad_norm": 0.9769062995910645, "learning_rate": 0.0002, "epoch": 3.587422197073802, "step": 22190}, {"loss": 0.5939, "grad_norm": 0.779636561870575, "learning_rate": 0.0002, "epoch": 3.589038881254547, "step": 22200}, {"loss": 0.5875, "grad_norm": 0.9063841104507446, "learning_rate": 0.0002, "epoch": 3.5906555654352923, "step": 22210}, {"loss": 0.5671, "grad_norm": 0.9216037392616272, "learning_rate": 0.0002, "epoch": 3.5922722496160375, "step": 22220}, {"loss": 0.6484, "grad_norm": 1.0217336416244507, "learning_rate": 0.0002, "epoch": 3.5938889337967828, "step": 22230}, {"loss": 0.6511, "grad_norm": 0.8513161540031433, "learning_rate": 0.0002, "epoch": 3.595505617977528, "step": 22240}, {"loss": 0.6301, "grad_norm": 0.8084813952445984, "learning_rate": 0.0002, "epoch": 3.597122302158273, "step": 22250}, {"loss": 0.6197, "grad_norm": 0.8524802923202515, "learning_rate": 0.0002, "epoch": 3.5987389863390185, "step": 22260}, {"loss": 0.5599, "grad_norm": 0.9356237649917603, "learning_rate": 0.0002, "epoch": 3.600355670519764, "step": 22270}, {"loss": 0.628, "grad_norm": 1.009600281715393, "learning_rate": 0.0002, "epoch": 3.6019723547005094, "step": 22280}, {"loss": 0.6179, "grad_norm": 0.9900581240653992, "learning_rate": 0.0002, "epoch": 3.6035890388812546, "step": 22290}, {"loss": 0.5725, "grad_norm": 1.062495231628418, "learning_rate": 0.0002, "epoch": 3.605205723062, "step": 22300}, {"loss": 0.607, "grad_norm": 0.8832381367683411, "learning_rate": 0.0002, "epoch": 3.606822407242745, "step": 22310}, {"loss": 0.6215, "grad_norm": 0.9284297823905945, "learning_rate": 0.0002, "epoch": 3.6084390914234903, "step": 22320}, {"loss": 0.685, "grad_norm": 1.2381829023361206, "learning_rate": 0.0002, "epoch": 3.610055775604236, "step": 22330}, {"loss": 0.6181, "grad_norm": 0.929434597492218, "learning_rate": 0.0002, "epoch": 3.611672459784981, "step": 22340}, {"loss": 0.6141, "grad_norm": 0.9714490175247192, "learning_rate": 0.0002, "epoch": 3.6132891439657264, "step": 22350}, {"loss": 0.6861, "grad_norm": 0.808014988899231, "learning_rate": 0.0002, "epoch": 3.6149058281464717, "step": 22360}, {"loss": 0.6428, "grad_norm": 1.0364398956298828, "learning_rate": 0.0002, "epoch": 3.616522512327217, "step": 22370}, {"loss": 0.6337, "grad_norm": 0.7858489751815796, "learning_rate": 0.0002, "epoch": 3.618139196507962, "step": 22380}, {"loss": 0.6214, "grad_norm": 0.9920870065689087, "learning_rate": 0.0002, "epoch": 3.6197558806887074, "step": 22390}, {"loss": 0.6659, "grad_norm": 0.9183220863342285, "learning_rate": 0.0002, "epoch": 3.6213725648694526, "step": 22400}, {"loss": 0.6036, "grad_norm": 0.9826246500015259, "learning_rate": 0.0002, "epoch": 3.622989249050198, "step": 22410}, {"loss": 0.6441, "grad_norm": 0.8632931113243103, "learning_rate": 0.0002, "epoch": 3.6246059332309435, "step": 22420}, {"loss": 0.6124, "grad_norm": 0.8468965291976929, "learning_rate": 0.0002, "epoch": 3.6262226174116887, "step": 22430}, {"loss": 0.6328, "grad_norm": 0.8466871976852417, "learning_rate": 0.0002, "epoch": 3.627839301592434, "step": 22440}, {"loss": 0.5941, "grad_norm": 0.9501169919967651, "learning_rate": 0.0002, "epoch": 3.629455985773179, "step": 22450}, {"loss": 0.6069, "grad_norm": 0.8906720876693726, "learning_rate": 0.0002, "epoch": 3.6310726699539244, "step": 22460}, {"loss": 0.6928, "grad_norm": 0.7400227189064026, "learning_rate": 0.0002, "epoch": 3.6326893541346696, "step": 22470}, {"loss": 0.6337, "grad_norm": 0.9756355881690979, "learning_rate": 0.0002, "epoch": 3.6343060383154153, "step": 22480}, {"loss": 0.6203, "grad_norm": 0.7504993081092834, "learning_rate": 0.0002, "epoch": 3.6359227224961606, "step": 22490}, {"loss": 0.6302, "grad_norm": 0.9270039200782776, "learning_rate": 0.0002, "epoch": 3.637539406676906, "step": 22500}, {"loss": 0.6026, "grad_norm": 0.8841686844825745, "learning_rate": 0.0002, "epoch": 3.639156090857651, "step": 22510}, {"loss": 0.6098, "grad_norm": 0.8533213138580322, "learning_rate": 0.0002, "epoch": 3.6407727750383962, "step": 22520}, {"loss": 0.6412, "grad_norm": 1.0052043199539185, "learning_rate": 0.0002, "epoch": 3.6423894592191415, "step": 22530}, {"loss": 0.6363, "grad_norm": 1.0323461294174194, "learning_rate": 0.0002, "epoch": 3.6440061433998867, "step": 22540}, {"loss": 0.6545, "grad_norm": 0.8654312491416931, "learning_rate": 0.0002, "epoch": 3.645622827580632, "step": 22550}, {"loss": 0.6155, "grad_norm": 0.6400038003921509, "learning_rate": 0.0002, "epoch": 3.647239511761377, "step": 22560}, {"loss": 0.5829, "grad_norm": 0.8061298727989197, "learning_rate": 0.0002, "epoch": 3.648856195942123, "step": 22570}, {"loss": 0.6388, "grad_norm": 0.9257854223251343, "learning_rate": 0.0002, "epoch": 3.650472880122868, "step": 22580}, {"loss": 0.6409, "grad_norm": 0.8439396619796753, "learning_rate": 0.0002, "epoch": 3.6520895643036133, "step": 22590}, {"loss": 0.5996, "grad_norm": 0.7764544486999512, "learning_rate": 0.0002, "epoch": 3.6537062484843585, "step": 22600}, {"loss": 0.6434, "grad_norm": 1.125451683998108, "learning_rate": 0.0002, "epoch": 3.6553229326651038, "step": 22610}, {"loss": 0.6579, "grad_norm": 0.7523018717765808, "learning_rate": 0.0002, "epoch": 3.656939616845849, "step": 22620}, {"loss": 0.6476, "grad_norm": 1.071026086807251, "learning_rate": 0.0002, "epoch": 3.6585563010265947, "step": 22630}, {"loss": 0.6459, "grad_norm": 0.945791482925415, "learning_rate": 0.0002, "epoch": 3.66017298520734, "step": 22640}, {"loss": 0.659, "grad_norm": 0.8001811504364014, "learning_rate": 0.0002, "epoch": 3.661789669388085, "step": 22650}, {"loss": 0.6385, "grad_norm": 0.9700816869735718, "learning_rate": 0.0002, "epoch": 3.6634063535688304, "step": 22660}, {"loss": 0.6337, "grad_norm": 0.9053242206573486, "learning_rate": 0.0002, "epoch": 3.6650230377495756, "step": 22670}, {"loss": 0.6335, "grad_norm": 0.944362461566925, "learning_rate": 0.0002, "epoch": 3.666639721930321, "step": 22680}, {"loss": 0.6235, "grad_norm": 1.067489504814148, "learning_rate": 0.0002, "epoch": 3.668256406111066, "step": 22690}, {"loss": 0.698, "grad_norm": 1.0984995365142822, "learning_rate": 0.0002, "epoch": 3.6698730902918113, "step": 22700}, {"loss": 0.6717, "grad_norm": 0.9336317777633667, "learning_rate": 0.0002, "epoch": 3.6714897744725565, "step": 22710}, {"loss": 0.6195, "grad_norm": 0.9261918663978577, "learning_rate": 0.0002, "epoch": 3.673106458653302, "step": 22720}, {"loss": 0.6332, "grad_norm": 0.8648008704185486, "learning_rate": 0.0002, "epoch": 3.6747231428340474, "step": 22730}, {"loss": 0.6576, "grad_norm": 0.7225083708763123, "learning_rate": 0.0002, "epoch": 3.6763398270147927, "step": 22740}, {"loss": 0.6406, "grad_norm": 0.9258282780647278, "learning_rate": 0.0002, "epoch": 3.677956511195538, "step": 22750}, {"loss": 0.6397, "grad_norm": 0.70876145362854, "learning_rate": 0.0002, "epoch": 3.679573195376283, "step": 22760}, {"loss": 0.6821, "grad_norm": 0.8780210018157959, "learning_rate": 0.0002, "epoch": 3.681189879557029, "step": 22770}, {"loss": 0.6036, "grad_norm": 0.8075440526008606, "learning_rate": 0.0002, "epoch": 3.682806563737774, "step": 22780}, {"loss": 0.6561, "grad_norm": 0.8503130674362183, "learning_rate": 0.0002, "epoch": 3.6844232479185193, "step": 22790}, {"loss": 0.6082, "grad_norm": 0.8413618206977844, "learning_rate": 0.0002, "epoch": 3.6860399320992645, "step": 22800}, {"loss": 0.614, "grad_norm": 0.8675165176391602, "learning_rate": 0.0002, "epoch": 3.6876566162800097, "step": 22810}, {"loss": 0.6157, "grad_norm": 0.8235884308815002, "learning_rate": 0.0002, "epoch": 3.689273300460755, "step": 22820}, {"loss": 0.5708, "grad_norm": 0.9477725625038147, "learning_rate": 0.0002, "epoch": 3.6908899846415, "step": 22830}, {"loss": 0.6481, "grad_norm": 0.7883533835411072, "learning_rate": 0.0002, "epoch": 3.6925066688222454, "step": 22840}, {"loss": 0.5872, "grad_norm": 1.047913908958435, "learning_rate": 0.0002, "epoch": 3.6941233530029907, "step": 22850}, {"loss": 0.6176, "grad_norm": 0.9171528816223145, "learning_rate": 0.0002, "epoch": 3.695740037183736, "step": 22860}, {"loss": 0.6204, "grad_norm": 0.9338192343711853, "learning_rate": 0.0002, "epoch": 3.6973567213644816, "step": 22870}, {"loss": 0.686, "grad_norm": 0.8799443244934082, "learning_rate": 0.0002, "epoch": 3.698973405545227, "step": 22880}, {"loss": 0.6206, "grad_norm": 0.8515434861183167, "learning_rate": 0.0002, "epoch": 3.700590089725972, "step": 22890}, {"loss": 0.5954, "grad_norm": 0.7805591821670532, "learning_rate": 0.0002, "epoch": 3.7022067739067173, "step": 22900}, {"loss": 0.6108, "grad_norm": 0.8470911979675293, "learning_rate": 0.0002, "epoch": 3.7038234580874625, "step": 22910}, {"loss": 0.6557, "grad_norm": 0.9452309012413025, "learning_rate": 0.0002, "epoch": 3.705440142268208, "step": 22920}, {"loss": 0.6529, "grad_norm": 0.950243353843689, "learning_rate": 0.0002, "epoch": 3.7070568264489534, "step": 22930}, {"loss": 0.6364, "grad_norm": 0.7882499098777771, "learning_rate": 0.0002, "epoch": 3.7086735106296986, "step": 22940}, {"loss": 0.6462, "grad_norm": 0.8307787775993347, "learning_rate": 0.0002, "epoch": 3.710290194810444, "step": 22950}, {"loss": 0.6371, "grad_norm": 1.0970630645751953, "learning_rate": 0.0002, "epoch": 3.711906878991189, "step": 22960}, {"loss": 0.6281, "grad_norm": 0.8269566297531128, "learning_rate": 0.0002, "epoch": 3.7135235631719343, "step": 22970}, {"loss": 0.6561, "grad_norm": 0.8306704759597778, "learning_rate": 0.0002, "epoch": 3.7151402473526796, "step": 22980}, {"loss": 0.6418, "grad_norm": 0.9710225462913513, "learning_rate": 0.0002, "epoch": 3.716756931533425, "step": 22990}, {"loss": 0.6639, "grad_norm": 0.8890530467033386, "learning_rate": 0.0002, "epoch": 3.71837361571417, "step": 23000}, {"loss": 0.6084, "grad_norm": 0.883522629737854, "learning_rate": 0.0002, "epoch": 3.7199902998949153, "step": 23010}, {"loss": 0.6183, "grad_norm": 0.8662652373313904, "learning_rate": 0.0002, "epoch": 3.721606984075661, "step": 23020}, {"loss": 0.6266, "grad_norm": 0.7228406667709351, "learning_rate": 0.0002, "epoch": 3.723223668256406, "step": 23030}, {"loss": 0.6417, "grad_norm": 1.060792088508606, "learning_rate": 0.0002, "epoch": 3.7248403524371514, "step": 23040}, {"loss": 0.6346, "grad_norm": 1.0119613409042358, "learning_rate": 0.0002, "epoch": 3.7264570366178966, "step": 23050}, {"loss": 0.6466, "grad_norm": 0.9212996959686279, "learning_rate": 0.0002, "epoch": 3.728073720798642, "step": 23060}, {"loss": 0.6454, "grad_norm": 0.925690233707428, "learning_rate": 0.0002, "epoch": 3.7296904049793875, "step": 23070}, {"loss": 0.615, "grad_norm": 0.8323310613632202, "learning_rate": 0.0002, "epoch": 3.7313070891601328, "step": 23080}, {"loss": 0.679, "grad_norm": 0.8966048955917358, "learning_rate": 0.0002, "epoch": 3.732923773340878, "step": 23090}, {"loss": 0.6151, "grad_norm": 0.8995837569236755, "learning_rate": 0.0002, "epoch": 3.7345404575216232, "step": 23100}, {"loss": 0.6143, "grad_norm": 0.8748890161514282, "learning_rate": 0.0002, "epoch": 3.7361571417023685, "step": 23110}, {"loss": 0.6246, "grad_norm": 0.7985540628433228, "learning_rate": 0.0002, "epoch": 3.7377738258831137, "step": 23120}, {"loss": 0.6279, "grad_norm": 1.0240917205810547, "learning_rate": 0.0002, "epoch": 3.739390510063859, "step": 23130}, {"loss": 0.6747, "grad_norm": 0.9181789755821228, "learning_rate": 0.0002, "epoch": 3.741007194244604, "step": 23140}, {"loss": 0.6026, "grad_norm": 0.8896583914756775, "learning_rate": 0.0002, "epoch": 3.7426238784253494, "step": 23150}, {"loss": 0.5972, "grad_norm": 0.8635515570640564, "learning_rate": 0.0002, "epoch": 3.744240562606095, "step": 23160}, {"loss": 0.6683, "grad_norm": 0.8873575329780579, "learning_rate": 0.0002, "epoch": 3.7458572467868403, "step": 23170}, {"loss": 0.6143, "grad_norm": 0.9807148575782776, "learning_rate": 0.0002, "epoch": 3.7474739309675855, "step": 23180}, {"loss": 0.6381, "grad_norm": 0.900477945804596, "learning_rate": 0.0002, "epoch": 3.7490906151483308, "step": 23190}, {"loss": 0.6542, "grad_norm": 0.9379992485046387, "learning_rate": 0.0002, "epoch": 3.750707299329076, "step": 23200}, {"loss": 0.6015, "grad_norm": 0.9649890661239624, "learning_rate": 0.0002, "epoch": 3.752323983509821, "step": 23210}, {"loss": 0.6735, "grad_norm": 0.824442446231842, "learning_rate": 0.0002, "epoch": 3.753940667690567, "step": 23220}, {"loss": 0.5992, "grad_norm": 0.8896150588989258, "learning_rate": 0.0002, "epoch": 3.755557351871312, "step": 23230}, {"loss": 0.6081, "grad_norm": 0.751249372959137, "learning_rate": 0.0002, "epoch": 3.7571740360520574, "step": 23240}, {"loss": 0.629, "grad_norm": 0.9392193555831909, "learning_rate": 0.0002, "epoch": 3.7587907202328026, "step": 23250}, {"loss": 0.6209, "grad_norm": 0.9284586310386658, "learning_rate": 0.0002, "epoch": 3.760407404413548, "step": 23260}, {"loss": 0.6414, "grad_norm": 0.7738175392150879, "learning_rate": 0.0002, "epoch": 3.762024088594293, "step": 23270}, {"loss": 0.6743, "grad_norm": 0.9252978563308716, "learning_rate": 0.0002, "epoch": 3.7636407727750383, "step": 23280}, {"loss": 0.5984, "grad_norm": 0.9501895904541016, "learning_rate": 0.0002, "epoch": 3.7652574569557835, "step": 23290}, {"loss": 0.6568, "grad_norm": 0.9416276216506958, "learning_rate": 0.0002, "epoch": 3.7668741411365287, "step": 23300}, {"loss": 0.6507, "grad_norm": 0.7076631784439087, "learning_rate": 0.0002, "epoch": 3.7684908253172744, "step": 23310}, {"loss": 0.6329, "grad_norm": 0.9864492416381836, "learning_rate": 0.0002, "epoch": 3.7701075094980196, "step": 23320}, {"loss": 0.6537, "grad_norm": 0.8450456261634827, "learning_rate": 0.0002, "epoch": 3.771724193678765, "step": 23330}, {"loss": 0.658, "grad_norm": 1.0768941640853882, "learning_rate": 0.0002, "epoch": 3.77334087785951, "step": 23340}, {"loss": 0.6408, "grad_norm": 0.9956819415092468, "learning_rate": 0.0002, "epoch": 3.7749575620402553, "step": 23350}, {"loss": 0.6464, "grad_norm": 0.9234658479690552, "learning_rate": 0.0002, "epoch": 3.7765742462210006, "step": 23360}, {"loss": 0.6542, "grad_norm": 1.0993858575820923, "learning_rate": 0.0002, "epoch": 3.7781909304017463, "step": 23370}, {"loss": 0.6391, "grad_norm": 0.923159658908844, "learning_rate": 0.0002, "epoch": 3.7798076145824915, "step": 23380}, {"loss": 0.6625, "grad_norm": 0.9311541318893433, "learning_rate": 0.0002, "epoch": 3.7814242987632367, "step": 23390}, {"loss": 0.6535, "grad_norm": 0.919681191444397, "learning_rate": 0.0002, "epoch": 3.783040982943982, "step": 23400}, {"loss": 0.6138, "grad_norm": 1.7406195402145386, "learning_rate": 0.0002, "epoch": 3.784657667124727, "step": 23410}, {"loss": 0.657, "grad_norm": 0.7789074182510376, "learning_rate": 0.0002, "epoch": 3.7862743513054724, "step": 23420}, {"loss": 0.658, "grad_norm": 0.8302814960479736, "learning_rate": 0.0002, "epoch": 3.7878910354862176, "step": 23430}, {"loss": 0.649, "grad_norm": 0.8089349269866943, "learning_rate": 0.0002, "epoch": 3.789507719666963, "step": 23440}, {"loss": 0.6682, "grad_norm": 0.9006284475326538, "learning_rate": 0.0002, "epoch": 3.791124403847708, "step": 23450}, {"loss": 0.6335, "grad_norm": 0.8426766991615295, "learning_rate": 0.0002, "epoch": 3.7927410880284538, "step": 23460}, {"loss": 0.6364, "grad_norm": 1.2576252222061157, "learning_rate": 0.0002, "epoch": 3.794357772209199, "step": 23470}, {"loss": 0.6324, "grad_norm": 1.0307610034942627, "learning_rate": 0.0002, "epoch": 3.7959744563899442, "step": 23480}, {"loss": 0.6262, "grad_norm": 0.8525972962379456, "learning_rate": 0.0002, "epoch": 3.7975911405706895, "step": 23490}, {"loss": 0.6757, "grad_norm": 1.159039855003357, "learning_rate": 0.0002, "epoch": 3.7992078247514347, "step": 23500}, {"loss": 0.6414, "grad_norm": 1.4193549156188965, "learning_rate": 0.0002, "epoch": 3.80082450893218, "step": 23510}, {"loss": 0.6413, "grad_norm": 0.8245543837547302, "learning_rate": 0.0002, "epoch": 3.8024411931129256, "step": 23520}, {"loss": 0.6417, "grad_norm": 0.8847230076789856, "learning_rate": 0.0002, "epoch": 3.804057877293671, "step": 23530}, {"loss": 0.6415, "grad_norm": 0.9574624300003052, "learning_rate": 0.0002, "epoch": 3.805674561474416, "step": 23540}, {"loss": 0.5765, "grad_norm": 1.048020601272583, "learning_rate": 0.0002, "epoch": 3.8072912456551613, "step": 23550}, {"loss": 0.6497, "grad_norm": 0.8302255868911743, "learning_rate": 0.0002, "epoch": 3.8089079298359065, "step": 23560}, {"loss": 0.6534, "grad_norm": 0.8269215822219849, "learning_rate": 0.0002, "epoch": 3.8105246140166518, "step": 23570}, {"loss": 0.6294, "grad_norm": 0.9375753402709961, "learning_rate": 0.0002, "epoch": 3.812141298197397, "step": 23580}, {"loss": 0.6132, "grad_norm": 1.0234097242355347, "learning_rate": 0.0002, "epoch": 3.8137579823781422, "step": 23590}, {"loss": 0.6625, "grad_norm": 0.8978445529937744, "learning_rate": 0.0002, "epoch": 3.8153746665588875, "step": 23600}, {"loss": 0.6315, "grad_norm": 0.7929515838623047, "learning_rate": 0.0002, "epoch": 3.816991350739633, "step": 23610}, {"loss": 0.6387, "grad_norm": 1.3255881071090698, "learning_rate": 0.0002, "epoch": 3.8186080349203784, "step": 23620}, {"loss": 0.5947, "grad_norm": 0.9188598990440369, "learning_rate": 0.0002, "epoch": 3.8202247191011236, "step": 23630}, {"loss": 0.6152, "grad_norm": 0.8811675906181335, "learning_rate": 0.0002, "epoch": 3.821841403281869, "step": 23640}, {"loss": 0.6253, "grad_norm": 0.8061038255691528, "learning_rate": 0.0002, "epoch": 3.823458087462614, "step": 23650}, {"loss": 0.6517, "grad_norm": 0.9975376129150391, "learning_rate": 0.0002, "epoch": 3.8250747716433597, "step": 23660}, {"loss": 0.6288, "grad_norm": 0.8036105036735535, "learning_rate": 0.0002, "epoch": 3.826691455824105, "step": 23670}, {"loss": 0.6845, "grad_norm": 0.7401984333992004, "learning_rate": 0.0002, "epoch": 3.82830814000485, "step": 23680}, {"loss": 0.6423, "grad_norm": 0.829753041267395, "learning_rate": 0.0002, "epoch": 3.8299248241855954, "step": 23690}, {"loss": 0.6611, "grad_norm": 0.8753240704536438, "learning_rate": 0.0002, "epoch": 3.8315415083663407, "step": 23700}, {"loss": 0.6686, "grad_norm": 0.8157842755317688, "learning_rate": 0.0002, "epoch": 3.833158192547086, "step": 23710}, {"loss": 0.6181, "grad_norm": 0.6183798909187317, "learning_rate": 0.0002, "epoch": 3.834774876727831, "step": 23720}, {"loss": 0.5965, "grad_norm": 0.9548442363739014, "learning_rate": 0.0002, "epoch": 3.8363915609085764, "step": 23730}, {"loss": 0.6456, "grad_norm": 0.8319669961929321, "learning_rate": 0.0002, "epoch": 3.8380082450893216, "step": 23740}, {"loss": 0.6585, "grad_norm": 0.9718693494796753, "learning_rate": 0.0002, "epoch": 3.839624929270067, "step": 23750}, {"loss": 0.6518, "grad_norm": 0.8672235012054443, "learning_rate": 0.0002, "epoch": 3.8412416134508125, "step": 23760}, {"loss": 0.6774, "grad_norm": 1.1210707426071167, "learning_rate": 0.0002, "epoch": 3.8428582976315577, "step": 23770}, {"loss": 0.5923, "grad_norm": 0.9177767634391785, "learning_rate": 0.0002, "epoch": 3.844474981812303, "step": 23780}, {"loss": 0.6286, "grad_norm": 0.8714171648025513, "learning_rate": 0.0002, "epoch": 3.846091665993048, "step": 23790}, {"loss": 0.6302, "grad_norm": 1.1853246688842773, "learning_rate": 0.0002, "epoch": 3.8477083501737934, "step": 23800}, {"loss": 0.6144, "grad_norm": 0.8091260194778442, "learning_rate": 0.0002, "epoch": 3.849325034354539, "step": 23810}, {"loss": 0.658, "grad_norm": 0.9710774421691895, "learning_rate": 0.0002, "epoch": 3.8509417185352843, "step": 23820}, {"loss": 0.6151, "grad_norm": 0.7648707628250122, "learning_rate": 0.0002, "epoch": 3.8525584027160296, "step": 23830}, {"loss": 0.6013, "grad_norm": 0.7809253931045532, "learning_rate": 0.0002, "epoch": 3.854175086896775, "step": 23840}, {"loss": 0.6006, "grad_norm": 0.8337951898574829, "learning_rate": 0.0002, "epoch": 3.85579177107752, "step": 23850}, {"loss": 0.6456, "grad_norm": 0.9271913170814514, "learning_rate": 0.0002, "epoch": 3.8574084552582653, "step": 23860}, {"loss": 0.6671, "grad_norm": 0.985334038734436, "learning_rate": 0.0002, "epoch": 3.8590251394390105, "step": 23870}, {"loss": 0.6693, "grad_norm": 0.8458583354949951, "learning_rate": 0.0002, "epoch": 3.8606418236197557, "step": 23880}, {"loss": 0.6207, "grad_norm": 1.015348196029663, "learning_rate": 0.0002, "epoch": 3.862258507800501, "step": 23890}, {"loss": 0.649, "grad_norm": 1.0121688842773438, "learning_rate": 0.0002, "epoch": 3.8638751919812466, "step": 23900}, {"loss": 0.5921, "grad_norm": 0.8883971571922302, "learning_rate": 0.0002, "epoch": 3.865491876161992, "step": 23910}, {"loss": 0.6597, "grad_norm": 1.028086543083191, "learning_rate": 0.0002, "epoch": 3.867108560342737, "step": 23920}, {"loss": 0.6654, "grad_norm": 0.9645734429359436, "learning_rate": 0.0002, "epoch": 3.8687252445234823, "step": 23930}, {"loss": 0.6328, "grad_norm": 0.8235350251197815, "learning_rate": 0.0002, "epoch": 3.8703419287042276, "step": 23940}, {"loss": 0.6387, "grad_norm": 1.0298916101455688, "learning_rate": 0.0002, "epoch": 3.871958612884973, "step": 23950}, {"loss": 0.5966, "grad_norm": 1.0063377618789673, "learning_rate": 0.0002, "epoch": 3.8735752970657185, "step": 23960}, {"loss": 0.6234, "grad_norm": 0.9230626821517944, "learning_rate": 0.0002, "epoch": 3.8751919812464637, "step": 23970}, {"loss": 0.6159, "grad_norm": 0.9243063926696777, "learning_rate": 0.0002, "epoch": 3.876808665427209, "step": 23980}, {"loss": 0.6035, "grad_norm": 1.0211291313171387, "learning_rate": 0.0002, "epoch": 3.878425349607954, "step": 23990}, {"loss": 0.6351, "grad_norm": 0.7800535559654236, "learning_rate": 0.0002, "epoch": 3.8800420337886994, "step": 24000}, {"loss": 0.7, "grad_norm": 0.7904248833656311, "learning_rate": 0.0002, "epoch": 3.8816587179694446, "step": 24010}, {"loss": 0.6516, "grad_norm": 1.1975988149642944, "learning_rate": 0.0002, "epoch": 3.88327540215019, "step": 24020}, {"loss": 0.6006, "grad_norm": 1.0626593828201294, "learning_rate": 0.0002, "epoch": 3.884892086330935, "step": 24030}, {"loss": 0.6115, "grad_norm": 0.9012193083763123, "learning_rate": 0.0002, "epoch": 3.8865087705116803, "step": 24040}, {"loss": 0.6786, "grad_norm": 1.1159172058105469, "learning_rate": 0.0002, "epoch": 3.888125454692426, "step": 24050}, {"loss": 0.6635, "grad_norm": 1.276838779449463, "learning_rate": 0.0002, "epoch": 3.889742138873171, "step": 24060}, {"loss": 0.5985, "grad_norm": 0.8467690348625183, "learning_rate": 0.0002, "epoch": 3.8913588230539164, "step": 24070}, {"loss": 0.6655, "grad_norm": 0.9862841963768005, "learning_rate": 0.0002, "epoch": 3.8929755072346617, "step": 24080}, {"loss": 0.6098, "grad_norm": 0.7134621739387512, "learning_rate": 0.0002, "epoch": 3.894592191415407, "step": 24090}, {"loss": 0.618, "grad_norm": 0.8178175091743469, "learning_rate": 0.0002, "epoch": 3.896208875596152, "step": 24100}, {"loss": 0.6147, "grad_norm": 0.9229172468185425, "learning_rate": 0.0002, "epoch": 3.897825559776898, "step": 24110}, {"loss": 0.6554, "grad_norm": 1.0878316164016724, "learning_rate": 0.0002, "epoch": 3.899442243957643, "step": 24120}, {"loss": 0.6616, "grad_norm": 0.971645712852478, "learning_rate": 0.0002, "epoch": 3.9010589281383883, "step": 24130}, {"loss": 0.6228, "grad_norm": 0.8862188458442688, "learning_rate": 0.0002, "epoch": 3.9026756123191335, "step": 24140}, {"loss": 0.6192, "grad_norm": 0.9126982688903809, "learning_rate": 0.0002, "epoch": 3.9042922964998787, "step": 24150}, {"loss": 0.6734, "grad_norm": 0.8833470940589905, "learning_rate": 0.0002, "epoch": 3.905908980680624, "step": 24160}, {"loss": 0.5832, "grad_norm": 0.8320947885513306, "learning_rate": 0.0002, "epoch": 3.907525664861369, "step": 24170}, {"loss": 0.6247, "grad_norm": 0.9156602025032043, "learning_rate": 0.0002, "epoch": 3.9091423490421144, "step": 24180}, {"loss": 0.6678, "grad_norm": 1.029181957244873, "learning_rate": 0.0002, "epoch": 3.9107590332228597, "step": 24190}, {"loss": 0.6565, "grad_norm": 0.9052802324295044, "learning_rate": 0.0002, "epoch": 3.9123757174036053, "step": 24200}, {"loss": 0.6346, "grad_norm": 0.8847255110740662, "learning_rate": 0.0002, "epoch": 3.9139924015843506, "step": 24210}, {"loss": 0.6343, "grad_norm": 0.9642062187194824, "learning_rate": 0.0002, "epoch": 3.915609085765096, "step": 24220}, {"loss": 0.6557, "grad_norm": 0.8629093766212463, "learning_rate": 0.0002, "epoch": 3.917225769945841, "step": 24230}, {"loss": 0.6086, "grad_norm": 0.8674976825714111, "learning_rate": 0.0002, "epoch": 3.9188424541265863, "step": 24240}, {"loss": 0.5874, "grad_norm": 1.104846477508545, "learning_rate": 0.0002, "epoch": 3.9204591383073315, "step": 24250}, {"loss": 0.6501, "grad_norm": 1.0874955654144287, "learning_rate": 0.0002, "epoch": 3.922075822488077, "step": 24260}, {"loss": 0.6455, "grad_norm": 0.8689812421798706, "learning_rate": 0.0002, "epoch": 3.9236925066688224, "step": 24270}, {"loss": 0.5893, "grad_norm": 0.9724617004394531, "learning_rate": 0.0002, "epoch": 3.9253091908495676, "step": 24280}, {"loss": 0.6616, "grad_norm": 0.9165538549423218, "learning_rate": 0.0002, "epoch": 3.926925875030313, "step": 24290}, {"loss": 0.645, "grad_norm": 0.9307710528373718, "learning_rate": 0.0002, "epoch": 3.928542559211058, "step": 24300}, {"loss": 0.6071, "grad_norm": 0.8589295148849487, "learning_rate": 0.0002, "epoch": 3.9301592433918033, "step": 24310}, {"loss": 0.6662, "grad_norm": 0.9151099920272827, "learning_rate": 0.0002, "epoch": 3.9317759275725486, "step": 24320}, {"loss": 0.7075, "grad_norm": 0.9633517265319824, "learning_rate": 0.0002, "epoch": 3.933392611753294, "step": 24330}, {"loss": 0.6432, "grad_norm": 0.9521116018295288, "learning_rate": 0.0002, "epoch": 3.935009295934039, "step": 24340}, {"loss": 0.6457, "grad_norm": 0.8366776704788208, "learning_rate": 0.0002, "epoch": 3.9366259801147847, "step": 24350}, {"loss": 0.6139, "grad_norm": 0.8972663283348083, "learning_rate": 0.0002, "epoch": 3.93824266429553, "step": 24360}, {"loss": 0.661, "grad_norm": 0.8102919459342957, "learning_rate": 0.0002, "epoch": 3.939859348476275, "step": 24370}, {"loss": 0.6388, "grad_norm": 0.8189975023269653, "learning_rate": 0.0002, "epoch": 3.9414760326570204, "step": 24380}, {"loss": 0.6818, "grad_norm": 0.9569464921951294, "learning_rate": 0.0002, "epoch": 3.9430927168377656, "step": 24390}, {"loss": 0.6999, "grad_norm": 0.7459101676940918, "learning_rate": 0.0002, "epoch": 3.9447094010185113, "step": 24400}, {"loss": 0.6069, "grad_norm": 0.8536974787712097, "learning_rate": 0.0002, "epoch": 3.9463260851992565, "step": 24410}, {"loss": 0.5683, "grad_norm": 0.8763698935508728, "learning_rate": 0.0002, "epoch": 3.9479427693800018, "step": 24420}, {"loss": 0.6478, "grad_norm": 0.9381106495857239, "learning_rate": 0.0002, "epoch": 3.949559453560747, "step": 24430}, {"loss": 0.6371, "grad_norm": 0.934440016746521, "learning_rate": 0.0002, "epoch": 3.9511761377414922, "step": 24440}, {"loss": 0.6393, "grad_norm": 0.903918981552124, "learning_rate": 0.0002, "epoch": 3.9527928219222375, "step": 24450}, {"loss": 0.6175, "grad_norm": 0.8771953582763672, "learning_rate": 0.0002, "epoch": 3.9544095061029827, "step": 24460}, {"loss": 0.6971, "grad_norm": 1.0375410318374634, "learning_rate": 0.0002, "epoch": 3.956026190283728, "step": 24470}, {"loss": 0.6313, "grad_norm": 0.9439185261726379, "learning_rate": 0.0002, "epoch": 3.957642874464473, "step": 24480}, {"loss": 0.6076, "grad_norm": 0.935467004776001, "learning_rate": 0.0002, "epoch": 3.9592595586452184, "step": 24490}, {"loss": 0.6437, "grad_norm": 0.6900772452354431, "learning_rate": 0.0002, "epoch": 3.960876242825964, "step": 24500}, {"loss": 0.6445, "grad_norm": 1.0172916650772095, "learning_rate": 0.0002, "epoch": 3.9624929270067093, "step": 24510}, {"loss": 0.6308, "grad_norm": 0.9167046546936035, "learning_rate": 0.0002, "epoch": 3.9641096111874545, "step": 24520}, {"loss": 0.6519, "grad_norm": 0.7230527997016907, "learning_rate": 0.0002, "epoch": 3.9657262953681998, "step": 24530}, {"loss": 0.6564, "grad_norm": 0.8980403542518616, "learning_rate": 0.0002, "epoch": 3.967342979548945, "step": 24540}, {"loss": 0.6099, "grad_norm": 0.8555465936660767, "learning_rate": 0.0002, "epoch": 3.9689596637296907, "step": 24550}, {"loss": 0.6617, "grad_norm": 0.7825445532798767, "learning_rate": 0.0002, "epoch": 3.970576347910436, "step": 24560}, {"loss": 0.604, "grad_norm": 0.7273133993148804, "learning_rate": 0.0002, "epoch": 3.972193032091181, "step": 24570}, {"loss": 0.6427, "grad_norm": 0.9612047672271729, "learning_rate": 0.0002, "epoch": 3.9738097162719264, "step": 24580}, {"loss": 0.6426, "grad_norm": 0.9865460991859436, "learning_rate": 0.0002, "epoch": 3.9754264004526716, "step": 24590}, {"loss": 0.6052, "grad_norm": 0.8638762831687927, "learning_rate": 0.0002, "epoch": 3.977043084633417, "step": 24600}, {"loss": 0.6097, "grad_norm": 1.0096198320388794, "learning_rate": 0.0002, "epoch": 3.978659768814162, "step": 24610}, {"loss": 0.6664, "grad_norm": 0.8475532531738281, "learning_rate": 0.0002, "epoch": 3.9802764529949073, "step": 24620}, {"loss": 0.6711, "grad_norm": 0.9696195721626282, "learning_rate": 0.0002, "epoch": 3.9818931371756525, "step": 24630}, {"loss": 0.6446, "grad_norm": 0.7499843239784241, "learning_rate": 0.0002, "epoch": 3.9835098213563978, "step": 24640}, {"loss": 0.6054, "grad_norm": 0.8865424990653992, "learning_rate": 0.0002, "epoch": 3.9851265055371434, "step": 24650}, {"loss": 0.5975, "grad_norm": 0.8089959025382996, "learning_rate": 0.0002, "epoch": 3.9867431897178887, "step": 24660}, {"loss": 0.6677, "grad_norm": 0.6946012377738953, "learning_rate": 0.0002, "epoch": 3.988359873898634, "step": 24670}, {"loss": 0.6329, "grad_norm": 0.7991759181022644, "learning_rate": 0.0002, "epoch": 3.989976558079379, "step": 24680}, {"loss": 0.6449, "grad_norm": 0.8803931474685669, "learning_rate": 0.0002, "epoch": 3.9915932422601244, "step": 24690}, {"loss": 0.7091, "grad_norm": 0.8848299980163574, "learning_rate": 0.0002, "epoch": 3.99320992644087, "step": 24700}, {"loss": 0.6551, "grad_norm": 0.7448889017105103, "learning_rate": 0.0002, "epoch": 3.9948266106216153, "step": 24710}, {"loss": 0.6432, "grad_norm": 0.9361620545387268, "learning_rate": 0.0002, "epoch": 3.9964432948023605, "step": 24720}, {"loss": 0.5917, "grad_norm": 0.9958081245422363, "learning_rate": 0.0002, "epoch": 3.9980599789831057, "step": 24730}, {"loss": 0.6567, "grad_norm": 1.026004672050476, "learning_rate": 0.0002, "epoch": 3.999676663163851, "step": 24740}, {"eval_loss": 1.1524168252944946, "eval_runtime": 122.1585, "eval_samples_per_second": 6.0, "eval_steps_per_second": 0.753, "epoch": 4.0, "step": 24742}, {"loss": 0.6057, "grad_norm": 1.0664808750152588, "learning_rate": 0.0002, "epoch": 4.001293347344596, "step": 24750}, {"loss": 0.5644, "grad_norm": 1.0113720893859863, "learning_rate": 0.0002, "epoch": 4.002910031525341, "step": 24760}, {"loss": 0.5628, "grad_norm": 0.991486668586731, "learning_rate": 0.0002, "epoch": 4.004526715706087, "step": 24770}, {"loss": 0.508, "grad_norm": 0.951754629611969, "learning_rate": 0.0002, "epoch": 4.006143399886832, "step": 24780}, {"loss": 0.5314, "grad_norm": 1.13059401512146, "learning_rate": 0.0002, "epoch": 4.007760084067577, "step": 24790}, {"loss": 0.5323, "grad_norm": 0.9343926310539246, "learning_rate": 0.0002, "epoch": 4.009376768248322, "step": 24800}, {"loss": 0.5161, "grad_norm": 1.0680590867996216, "learning_rate": 0.0002, "epoch": 4.010993452429068, "step": 24810}, {"loss": 0.513, "grad_norm": 1.0022706985473633, "learning_rate": 0.0002, "epoch": 4.012610136609814, "step": 24820}, {"loss": 0.543, "grad_norm": 1.0285297632217407, "learning_rate": 0.0002, "epoch": 4.014226820790559, "step": 24830}, {"loss": 0.5311, "grad_norm": 0.8347002863883972, "learning_rate": 0.0002, "epoch": 4.015843504971304, "step": 24840}, {"loss": 0.5655, "grad_norm": 0.9675396680831909, "learning_rate": 0.0002, "epoch": 4.017460189152049, "step": 24850}, {"loss": 0.5625, "grad_norm": 0.9238511323928833, "learning_rate": 0.0002, "epoch": 4.019076873332795, "step": 24860}, {"loss": 0.5327, "grad_norm": 1.1576941013336182, "learning_rate": 0.0002, "epoch": 4.02069355751354, "step": 24870}, {"loss": 0.5533, "grad_norm": 0.8583757281303406, "learning_rate": 0.0002, "epoch": 4.022310241694285, "step": 24880}, {"loss": 0.5483, "grad_norm": 0.9816817045211792, "learning_rate": 0.0002, "epoch": 4.02392692587503, "step": 24890}, {"loss": 0.5605, "grad_norm": 0.955073893070221, "learning_rate": 0.0002, "epoch": 4.0255436100557755, "step": 24900}, {"loss": 0.4896, "grad_norm": 1.1054974794387817, "learning_rate": 0.0002, "epoch": 4.027160294236521, "step": 24910}, {"loss": 0.5246, "grad_norm": 1.1240060329437256, "learning_rate": 0.0002, "epoch": 4.028776978417266, "step": 24920}, {"loss": 0.5451, "grad_norm": 0.9512825012207031, "learning_rate": 0.0002, "epoch": 4.030393662598011, "step": 24930}, {"loss": 0.5584, "grad_norm": 0.85965496301651, "learning_rate": 0.0002, "epoch": 4.0320103467787565, "step": 24940}, {"loss": 0.5564, "grad_norm": 0.9378061294555664, "learning_rate": 0.0002, "epoch": 4.033627030959502, "step": 24950}, {"loss": 0.5008, "grad_norm": 0.9655424356460571, "learning_rate": 0.0002, "epoch": 4.035243715140247, "step": 24960}, {"loss": 0.5538, "grad_norm": 1.1393707990646362, "learning_rate": 0.0002, "epoch": 4.036860399320993, "step": 24970}, {"loss": 0.5785, "grad_norm": 1.0220451354980469, "learning_rate": 0.0002, "epoch": 4.038477083501738, "step": 24980}, {"loss": 0.5813, "grad_norm": 0.9785808324813843, "learning_rate": 0.0002, "epoch": 4.0400937676824835, "step": 24990}, {"loss": 0.5153, "grad_norm": 1.0257649421691895, "learning_rate": 0.0002, "epoch": 4.041710451863229, "step": 25000}, {"loss": 0.5658, "grad_norm": 0.9737892150878906, "learning_rate": 0.0002, "epoch": 4.043327136043974, "step": 25010}, {"loss": 0.5515, "grad_norm": 0.7416959404945374, "learning_rate": 0.0002, "epoch": 4.044943820224719, "step": 25020}, {"loss": 0.5372, "grad_norm": 0.7909596562385559, "learning_rate": 0.0002, "epoch": 4.046560504405464, "step": 25030}, {"loss": 0.5265, "grad_norm": 0.8923130631446838, "learning_rate": 0.0002, "epoch": 4.04817718858621, "step": 25040}, {"loss": 0.5035, "grad_norm": 0.9044941663742065, "learning_rate": 0.0002, "epoch": 4.049793872766955, "step": 25050}, {"loss": 0.5135, "grad_norm": 0.866352379322052, "learning_rate": 0.0002, "epoch": 4.0514105569477, "step": 25060}, {"loss": 0.5956, "grad_norm": 1.544549822807312, "learning_rate": 0.0002, "epoch": 4.053027241128445, "step": 25070}, {"loss": 0.5418, "grad_norm": 0.8426995277404785, "learning_rate": 0.0002, "epoch": 4.054643925309191, "step": 25080}, {"loss": 0.5537, "grad_norm": 0.9797548651695251, "learning_rate": 0.0002, "epoch": 4.056260609489936, "step": 25090}, {"loss": 0.55, "grad_norm": 0.8468434810638428, "learning_rate": 0.0002, "epoch": 4.057877293670681, "step": 25100}, {"loss": 0.5242, "grad_norm": 0.9294559955596924, "learning_rate": 0.0002, "epoch": 4.059493977851426, "step": 25110}, {"loss": 0.5295, "grad_norm": 0.9686688780784607, "learning_rate": 0.0002, "epoch": 4.061110662032172, "step": 25120}, {"loss": 0.5642, "grad_norm": 0.8042728304862976, "learning_rate": 0.0002, "epoch": 4.062727346212918, "step": 25130}, {"loss": 0.548, "grad_norm": 1.165160894393921, "learning_rate": 0.0002, "epoch": 4.064344030393663, "step": 25140}, {"loss": 0.5473, "grad_norm": 1.2161961793899536, "learning_rate": 0.0002, "epoch": 4.065960714574408, "step": 25150}, {"loss": 0.5217, "grad_norm": 1.0762810707092285, "learning_rate": 0.0002, "epoch": 4.067577398755153, "step": 25160}, {"loss": 0.5886, "grad_norm": 0.7580869793891907, "learning_rate": 0.0002, "epoch": 4.069194082935899, "step": 25170}, {"loss": 0.5401, "grad_norm": 0.9630117416381836, "learning_rate": 0.0002, "epoch": 4.070810767116644, "step": 25180}, {"loss": 0.5378, "grad_norm": 0.9049716591835022, "learning_rate": 0.0002, "epoch": 4.072427451297389, "step": 25190}, {"loss": 0.5266, "grad_norm": 1.1536930799484253, "learning_rate": 0.0002, "epoch": 4.074044135478134, "step": 25200}, {"loss": 0.5523, "grad_norm": 0.901461124420166, "learning_rate": 0.0002, "epoch": 4.0756608196588795, "step": 25210}, {"loss": 0.5132, "grad_norm": 1.3318437337875366, "learning_rate": 0.0002, "epoch": 4.077277503839625, "step": 25220}, {"loss": 0.5317, "grad_norm": 0.8811455368995667, "learning_rate": 0.0002, "epoch": 4.07889418802037, "step": 25230}, {"loss": 0.5798, "grad_norm": 1.0564165115356445, "learning_rate": 0.0002, "epoch": 4.080510872201115, "step": 25240}, {"loss": 0.5472, "grad_norm": 1.1008027791976929, "learning_rate": 0.0002, "epoch": 4.08212755638186, "step": 25250}, {"loss": 0.5195, "grad_norm": 1.150097131729126, "learning_rate": 0.0002, "epoch": 4.083744240562606, "step": 25260}, {"loss": 0.5321, "grad_norm": 0.9339924454689026, "learning_rate": 0.0002, "epoch": 4.085360924743352, "step": 25270}, {"loss": 0.5597, "grad_norm": 1.0902045965194702, "learning_rate": 0.0002, "epoch": 4.086977608924097, "step": 25280}, {"loss": 0.5203, "grad_norm": 0.8483911156654358, "learning_rate": 0.0002, "epoch": 4.088594293104842, "step": 25290}, {"loss": 0.5697, "grad_norm": 0.9477024674415588, "learning_rate": 0.0002, "epoch": 4.0902109772855875, "step": 25300}, {"loss": 0.5384, "grad_norm": 0.9500215649604797, "learning_rate": 0.0002, "epoch": 4.091827661466333, "step": 25310}, {"loss": 0.5045, "grad_norm": 1.040468454360962, "learning_rate": 0.0002, "epoch": 4.093444345647078, "step": 25320}, {"loss": 0.5488, "grad_norm": 0.7457592487335205, "learning_rate": 0.0002, "epoch": 4.095061029827823, "step": 25330}, {"loss": 0.609, "grad_norm": 1.2092097997665405, "learning_rate": 0.0002, "epoch": 4.096677714008568, "step": 25340}, {"loss": 0.5174, "grad_norm": 0.9652107954025269, "learning_rate": 0.0002, "epoch": 4.098294398189314, "step": 25350}, {"loss": 0.5559, "grad_norm": 0.8464955687522888, "learning_rate": 0.0002, "epoch": 4.099911082370059, "step": 25360}, {"loss": 0.5635, "grad_norm": 0.875026285648346, "learning_rate": 0.0002, "epoch": 4.101527766550804, "step": 25370}, {"loss": 0.5774, "grad_norm": 0.9241740107536316, "learning_rate": 0.0002, "epoch": 4.103144450731549, "step": 25380}, {"loss": 0.5578, "grad_norm": 0.9769546389579773, "learning_rate": 0.0002, "epoch": 4.1047611349122946, "step": 25390}, {"loss": 0.567, "grad_norm": 1.1501960754394531, "learning_rate": 0.0002, "epoch": 4.10637781909304, "step": 25400}, {"loss": 0.5241, "grad_norm": 0.9135243892669678, "learning_rate": 0.0002, "epoch": 4.107994503273786, "step": 25410}, {"loss": 0.5152, "grad_norm": 0.9905396103858948, "learning_rate": 0.0002, "epoch": 4.109611187454531, "step": 25420}, {"loss": 0.5064, "grad_norm": 0.9845104217529297, "learning_rate": 0.0002, "epoch": 4.111227871635276, "step": 25430}, {"loss": 0.5029, "grad_norm": 0.8326883912086487, "learning_rate": 0.0002, "epoch": 4.112844555816022, "step": 25440}, {"loss": 0.5312, "grad_norm": 0.9264556765556335, "learning_rate": 0.0002, "epoch": 4.114461239996767, "step": 25450}, {"loss": 0.5968, "grad_norm": 1.043080449104309, "learning_rate": 0.0002, "epoch": 4.116077924177512, "step": 25460}, {"loss": 0.5773, "grad_norm": 0.8533386588096619, "learning_rate": 0.0002, "epoch": 4.117694608358257, "step": 25470}, {"loss": 0.5584, "grad_norm": 1.0133965015411377, "learning_rate": 0.0002, "epoch": 4.1193112925390025, "step": 25480}, {"loss": 0.566, "grad_norm": 0.7476310133934021, "learning_rate": 0.0002, "epoch": 4.120927976719748, "step": 25490}, {"loss": 0.5189, "grad_norm": 1.1247259378433228, "learning_rate": 0.0002, "epoch": 4.122544660900493, "step": 25500}, {"loss": 0.5751, "grad_norm": 1.0764678716659546, "learning_rate": 0.0002, "epoch": 4.124161345081238, "step": 25510}, {"loss": 0.5391, "grad_norm": 0.7679798007011414, "learning_rate": 0.0002, "epoch": 4.1257780292619834, "step": 25520}, {"loss": 0.5233, "grad_norm": 0.8877071142196655, "learning_rate": 0.0002, "epoch": 4.127394713442729, "step": 25530}, {"loss": 0.5769, "grad_norm": 1.0440239906311035, "learning_rate": 0.0002, "epoch": 4.129011397623474, "step": 25540}, {"loss": 0.5723, "grad_norm": 0.984145998954773, "learning_rate": 0.0002, "epoch": 4.130628081804219, "step": 25550}, {"loss": 0.5741, "grad_norm": 0.8667055368423462, "learning_rate": 0.0002, "epoch": 4.132244765984965, "step": 25560}, {"loss": 0.5816, "grad_norm": 1.1300835609436035, "learning_rate": 0.0002, "epoch": 4.1338614501657105, "step": 25570}, {"loss": 0.524, "grad_norm": 0.9314348101615906, "learning_rate": 0.0002, "epoch": 4.135478134346456, "step": 25580}, {"loss": 0.5283, "grad_norm": 0.7731879949569702, "learning_rate": 0.0002, "epoch": 4.137094818527201, "step": 25590}, {"loss": 0.5307, "grad_norm": 1.0080097913742065, "learning_rate": 0.0002, "epoch": 4.138711502707946, "step": 25600}, {"loss": 0.5759, "grad_norm": 1.2475038766860962, "learning_rate": 0.0002, "epoch": 4.140328186888691, "step": 25610}, {"loss": 0.55, "grad_norm": 0.9912930727005005, "learning_rate": 0.0002, "epoch": 4.141944871069437, "step": 25620}, {"loss": 0.5624, "grad_norm": 0.9088651537895203, "learning_rate": 0.0002, "epoch": 4.143561555250182, "step": 25630}, {"loss": 0.5393, "grad_norm": 0.8940697312355042, "learning_rate": 0.0002, "epoch": 4.145178239430927, "step": 25640}, {"loss": 0.5341, "grad_norm": 1.0798203945159912, "learning_rate": 0.0002, "epoch": 4.146794923611672, "step": 25650}, {"loss": 0.5987, "grad_norm": 0.955172061920166, "learning_rate": 0.0002, "epoch": 4.148411607792418, "step": 25660}, {"loss": 0.569, "grad_norm": 0.9692716002464294, "learning_rate": 0.0002, "epoch": 4.150028291973163, "step": 25670}, {"loss": 0.5478, "grad_norm": 1.0813939571380615, "learning_rate": 0.0002, "epoch": 4.151644976153908, "step": 25680}, {"loss": 0.5383, "grad_norm": 1.135675072669983, "learning_rate": 0.0002, "epoch": 4.153261660334653, "step": 25690}, {"loss": 0.5247, "grad_norm": 1.0392236709594727, "learning_rate": 0.0002, "epoch": 4.1548783445153985, "step": 25700}, {"loss": 0.5204, "grad_norm": 0.9473116993904114, "learning_rate": 0.0002, "epoch": 4.156495028696145, "step": 25710}, {"loss": 0.5339, "grad_norm": 0.712493896484375, "learning_rate": 0.0002, "epoch": 4.15811171287689, "step": 25720}, {"loss": 0.5781, "grad_norm": 0.8724465370178223, "learning_rate": 0.0002, "epoch": 4.159728397057635, "step": 25730}, {"loss": 0.5325, "grad_norm": 0.9870015978813171, "learning_rate": 0.0002, "epoch": 4.16134508123838, "step": 25740}, {"loss": 0.5503, "grad_norm": 1.025273084640503, "learning_rate": 0.0002, "epoch": 4.1629617654191255, "step": 25750}, {"loss": 0.5223, "grad_norm": 0.9243090152740479, "learning_rate": 0.0002, "epoch": 4.164578449599871, "step": 25760}, {"loss": 0.5177, "grad_norm": 1.1656451225280762, "learning_rate": 0.0002, "epoch": 4.166195133780616, "step": 25770}, {"loss": 0.5334, "grad_norm": 0.936358630657196, "learning_rate": 0.0002, "epoch": 4.167811817961361, "step": 25780}, {"loss": 0.5236, "grad_norm": 0.8618208169937134, "learning_rate": 0.0002, "epoch": 4.1694285021421065, "step": 25790}, {"loss": 0.5186, "grad_norm": 0.8580600023269653, "learning_rate": 0.0002, "epoch": 4.171045186322852, "step": 25800}, {"loss": 0.5212, "grad_norm": 1.0128562450408936, "learning_rate": 0.0002, "epoch": 4.172661870503597, "step": 25810}, {"loss": 0.5404, "grad_norm": 0.854865312576294, "learning_rate": 0.0002, "epoch": 4.174278554684342, "step": 25820}, {"loss": 0.5377, "grad_norm": 1.235082745552063, "learning_rate": 0.0002, "epoch": 4.175895238865087, "step": 25830}, {"loss": 0.5614, "grad_norm": 0.9796220660209656, "learning_rate": 0.0002, "epoch": 4.177511923045833, "step": 25840}, {"loss": 0.5689, "grad_norm": 0.8922094702720642, "learning_rate": 0.0002, "epoch": 4.179128607226578, "step": 25850}, {"loss": 0.5806, "grad_norm": 0.9672530293464661, "learning_rate": 0.0002, "epoch": 4.180745291407324, "step": 25860}, {"loss": 0.5074, "grad_norm": 0.8662548661231995, "learning_rate": 0.0002, "epoch": 4.182361975588069, "step": 25870}, {"loss": 0.5329, "grad_norm": 0.7938798069953918, "learning_rate": 0.0002, "epoch": 4.1839786597688144, "step": 25880}, {"loss": 0.5427, "grad_norm": 1.0517958402633667, "learning_rate": 0.0002, "epoch": 4.18559534394956, "step": 25890}, {"loss": 0.5147, "grad_norm": 0.8939275145530701, "learning_rate": 0.0002, "epoch": 4.187212028130305, "step": 25900}, {"loss": 0.5199, "grad_norm": 1.0296672582626343, "learning_rate": 0.0002, "epoch": 4.18882871231105, "step": 25910}, {"loss": 0.5522, "grad_norm": 0.8104017972946167, "learning_rate": 0.0002, "epoch": 4.190445396491795, "step": 25920}, {"loss": 0.596, "grad_norm": 0.9984509944915771, "learning_rate": 0.0002, "epoch": 4.192062080672541, "step": 25930}, {"loss": 0.5356, "grad_norm": 0.9844784736633301, "learning_rate": 0.0002, "epoch": 4.193678764853286, "step": 25940}, {"loss": 0.5198, "grad_norm": 0.8168622255325317, "learning_rate": 0.0002, "epoch": 4.195295449034031, "step": 25950}, {"loss": 0.542, "grad_norm": 1.0878913402557373, "learning_rate": 0.0002, "epoch": 4.196912133214776, "step": 25960}, {"loss": 0.5414, "grad_norm": 0.927126407623291, "learning_rate": 0.0002, "epoch": 4.1985288173955215, "step": 25970}, {"loss": 0.5794, "grad_norm": 0.838586688041687, "learning_rate": 0.0002, "epoch": 4.200145501576267, "step": 25980}, {"loss": 0.5454, "grad_norm": 1.2572145462036133, "learning_rate": 0.0002, "epoch": 4.201762185757012, "step": 25990}, {"loss": 0.5048, "grad_norm": 1.0476740598678589, "learning_rate": 0.0002, "epoch": 4.203378869937758, "step": 26000}, {"loss": 0.5127, "grad_norm": 1.0873368978500366, "learning_rate": 0.0002, "epoch": 4.204995554118503, "step": 26010}, {"loss": 0.5679, "grad_norm": 1.2664896249771118, "learning_rate": 0.0002, "epoch": 4.206612238299249, "step": 26020}, {"loss": 0.5814, "grad_norm": 1.0312391519546509, "learning_rate": 0.0002, "epoch": 4.208228922479994, "step": 26030}, {"loss": 0.571, "grad_norm": 1.0235042572021484, "learning_rate": 0.0002, "epoch": 4.209845606660739, "step": 26040}, {"loss": 0.5766, "grad_norm": 0.8882219195365906, "learning_rate": 0.0002, "epoch": 4.211462290841484, "step": 26050}, {"loss": 0.5557, "grad_norm": 0.9115961790084839, "learning_rate": 0.0002, "epoch": 4.2130789750222295, "step": 26060}, {"loss": 0.5455, "grad_norm": 1.0218228101730347, "learning_rate": 0.0002, "epoch": 4.214695659202975, "step": 26070}, {"loss": 0.5462, "grad_norm": 1.0802232027053833, "learning_rate": 0.0002, "epoch": 4.21631234338372, "step": 26080}, {"loss": 0.557, "grad_norm": 1.1488053798675537, "learning_rate": 0.0002, "epoch": 4.217929027564465, "step": 26090}, {"loss": 0.52, "grad_norm": 1.0487725734710693, "learning_rate": 0.0002, "epoch": 4.21954571174521, "step": 26100}, {"loss": 0.5568, "grad_norm": 0.9131165742874146, "learning_rate": 0.0002, "epoch": 4.221162395925956, "step": 26110}, {"loss": 0.5206, "grad_norm": 0.9012845158576965, "learning_rate": 0.0002, "epoch": 4.222779080106701, "step": 26120}, {"loss": 0.561, "grad_norm": 0.8389840126037598, "learning_rate": 0.0002, "epoch": 4.224395764287446, "step": 26130}, {"loss": 0.5268, "grad_norm": 0.8924660682678223, "learning_rate": 0.0002, "epoch": 4.226012448468191, "step": 26140}, {"loss": 0.5715, "grad_norm": 0.8556463718414307, "learning_rate": 0.0002, "epoch": 4.2276291326489375, "step": 26150}, {"loss": 0.5695, "grad_norm": 0.9643129110336304, "learning_rate": 0.0002, "epoch": 4.229245816829683, "step": 26160}, {"loss": 0.5321, "grad_norm": 0.9865712523460388, "learning_rate": 0.0002, "epoch": 4.230862501010428, "step": 26170}, {"loss": 0.5406, "grad_norm": 1.152641773223877, "learning_rate": 0.0002, "epoch": 4.232479185191173, "step": 26180}, {"loss": 0.5632, "grad_norm": 0.9157698154449463, "learning_rate": 0.0002, "epoch": 4.234095869371918, "step": 26190}, {"loss": 0.5717, "grad_norm": 0.8418048620223999, "learning_rate": 0.0002, "epoch": 4.235712553552664, "step": 26200}, {"loss": 0.5624, "grad_norm": 0.9430168867111206, "learning_rate": 0.0002, "epoch": 4.237329237733409, "step": 26210}, {"loss": 0.5574, "grad_norm": 1.012582778930664, "learning_rate": 0.0002, "epoch": 4.238945921914154, "step": 26220}, {"loss": 0.5693, "grad_norm": 1.112619400024414, "learning_rate": 0.0002, "epoch": 4.240562606094899, "step": 26230}, {"loss": 0.6037, "grad_norm": 0.9243621826171875, "learning_rate": 0.0002, "epoch": 4.2421792902756446, "step": 26240}, {"loss": 0.569, "grad_norm": 0.6977595686912537, "learning_rate": 0.0002, "epoch": 4.24379597445639, "step": 26250}, {"loss": 0.5379, "grad_norm": 0.9600721597671509, "learning_rate": 0.0002, "epoch": 4.245412658637135, "step": 26260}, {"loss": 0.5658, "grad_norm": 0.882641613483429, "learning_rate": 0.0002, "epoch": 4.24702934281788, "step": 26270}, {"loss": 0.55, "grad_norm": 1.010920763015747, "learning_rate": 0.0002, "epoch": 4.2486460269986255, "step": 26280}, {"loss": 0.5803, "grad_norm": 0.9289400577545166, "learning_rate": 0.0002, "epoch": 4.250262711179371, "step": 26290}, {"loss": 0.541, "grad_norm": 1.137397289276123, "learning_rate": 0.0002, "epoch": 4.251879395360117, "step": 26300}, {"loss": 0.5204, "grad_norm": 1.0136182308197021, "learning_rate": 0.0002, "epoch": 4.253496079540862, "step": 26310}, {"loss": 0.5708, "grad_norm": 0.9387356042861938, "learning_rate": 0.0002, "epoch": 4.255112763721607, "step": 26320}, {"loss": 0.5948, "grad_norm": 1.1833957433700562, "learning_rate": 0.0002, "epoch": 4.2567294479023525, "step": 26330}, {"loss": 0.5905, "grad_norm": 0.9415934681892395, "learning_rate": 0.0002, "epoch": 4.258346132083098, "step": 26340}, {"loss": 0.5539, "grad_norm": 0.8550165891647339, "learning_rate": 0.0002, "epoch": 4.259962816263843, "step": 26350}, {"loss": 0.555, "grad_norm": 9.924622535705566, "learning_rate": 0.0002, "epoch": 4.261579500444588, "step": 26360}, {"loss": 0.5689, "grad_norm": 1.0104902982711792, "learning_rate": 0.0002, "epoch": 4.2631961846253335, "step": 26370}, {"loss": 0.5698, "grad_norm": 0.890794038772583, "learning_rate": 0.0002, "epoch": 4.264812868806079, "step": 26380}, {"loss": 0.563, "grad_norm": 1.0560191869735718, "learning_rate": 0.0002, "epoch": 4.266429552986824, "step": 26390}, {"loss": 0.5119, "grad_norm": 1.0135581493377686, "learning_rate": 0.0002, "epoch": 4.268046237167569, "step": 26400}, {"loss": 0.5359, "grad_norm": 1.1304140090942383, "learning_rate": 0.0002, "epoch": 4.269662921348314, "step": 26410}, {"loss": 0.5615, "grad_norm": 0.9899303913116455, "learning_rate": 0.0002, "epoch": 4.27127960552906, "step": 26420}, {"loss": 0.5815, "grad_norm": 1.0505329370498657, "learning_rate": 0.0002, "epoch": 4.272896289709805, "step": 26430}, {"loss": 0.5384, "grad_norm": 0.9389396905899048, "learning_rate": 0.0002, "epoch": 4.27451297389055, "step": 26440}, {"loss": 0.5558, "grad_norm": 0.875328779220581, "learning_rate": 0.0002, "epoch": 4.276129658071296, "step": 26450}, {"loss": 0.5601, "grad_norm": 1.0689256191253662, "learning_rate": 0.0002, "epoch": 4.277746342252041, "step": 26460}, {"loss": 0.546, "grad_norm": 0.9988957643508911, "learning_rate": 0.0002, "epoch": 4.279363026432787, "step": 26470}, {"loss": 0.5478, "grad_norm": 0.8721813559532166, "learning_rate": 0.0002, "epoch": 4.280979710613532, "step": 26480}, {"loss": 0.5424, "grad_norm": 1.100109577178955, "learning_rate": 0.0002, "epoch": 4.282596394794277, "step": 26490}, {"loss": 0.572, "grad_norm": 1.1607271432876587, "learning_rate": 0.0002, "epoch": 4.284213078975022, "step": 26500}, {"loss": 0.6287, "grad_norm": 0.879088819026947, "learning_rate": 0.0002, "epoch": 4.285829763155768, "step": 26510}, {"loss": 0.573, "grad_norm": 0.9891700744628906, "learning_rate": 0.0002, "epoch": 4.287446447336513, "step": 26520}, {"loss": 0.6018, "grad_norm": 1.0831127166748047, "learning_rate": 0.0002, "epoch": 4.289063131517258, "step": 26530}, {"loss": 0.5693, "grad_norm": 1.4108285903930664, "learning_rate": 0.0002, "epoch": 4.290679815698003, "step": 26540}, {"loss": 0.5888, "grad_norm": 1.0630289316177368, "learning_rate": 0.0002, "epoch": 4.2922964998787485, "step": 26550}, {"loss": 0.5817, "grad_norm": 1.0854572057724, "learning_rate": 0.0002, "epoch": 4.293913184059494, "step": 26560}, {"loss": 0.5586, "grad_norm": 0.9561646580696106, "learning_rate": 0.0002, "epoch": 4.295529868240239, "step": 26570}, {"loss": 0.5674, "grad_norm": 0.9064981937408447, "learning_rate": 0.0002, "epoch": 4.297146552420984, "step": 26580}, {"loss": 0.5847, "grad_norm": 1.0082972049713135, "learning_rate": 0.0002, "epoch": 4.298763236601729, "step": 26590}, {"loss": 0.5711, "grad_norm": 1.1613214015960693, "learning_rate": 0.0002, "epoch": 4.3003799207824756, "step": 26600}, {"loss": 0.551, "grad_norm": 0.9847695231437683, "learning_rate": 0.0002, "epoch": 4.301996604963221, "step": 26610}, {"loss": 0.6089, "grad_norm": 1.0980697870254517, "learning_rate": 0.0002, "epoch": 4.303613289143966, "step": 26620}, {"loss": 0.5797, "grad_norm": 0.8861175179481506, "learning_rate": 0.0002, "epoch": 4.305229973324711, "step": 26630}, {"loss": 0.5716, "grad_norm": 0.8917363286018372, "learning_rate": 0.0002, "epoch": 4.3068466575054565, "step": 26640}, {"loss": 0.5892, "grad_norm": 1.0458378791809082, "learning_rate": 0.0002, "epoch": 4.308463341686202, "step": 26650}, {"loss": 0.5883, "grad_norm": 1.4859240055084229, "learning_rate": 0.0002, "epoch": 4.310080025866947, "step": 26660}, {"loss": 0.5296, "grad_norm": 1.1376359462738037, "learning_rate": 0.0002, "epoch": 4.311696710047692, "step": 26670}, {"loss": 0.5671, "grad_norm": 0.991349995136261, "learning_rate": 0.0002, "epoch": 4.313313394228437, "step": 26680}, {"loss": 0.5338, "grad_norm": 0.9995543956756592, "learning_rate": 0.0002, "epoch": 4.314930078409183, "step": 26690}, {"loss": 0.5542, "grad_norm": 1.0515851974487305, "learning_rate": 0.0002, "epoch": 4.316546762589928, "step": 26700}, {"loss": 0.5473, "grad_norm": 1.008023977279663, "learning_rate": 0.0002, "epoch": 4.318163446770673, "step": 26710}, {"loss": 0.5506, "grad_norm": 1.0184582471847534, "learning_rate": 0.0002, "epoch": 4.319780130951418, "step": 26720}, {"loss": 0.5828, "grad_norm": 1.161071538925171, "learning_rate": 0.0002, "epoch": 4.321396815132164, "step": 26730}, {"loss": 0.5633, "grad_norm": 0.9580779671669006, "learning_rate": 0.0002, "epoch": 4.323013499312909, "step": 26740}, {"loss": 0.5785, "grad_norm": 1.0189911127090454, "learning_rate": 0.0002, "epoch": 4.324630183493655, "step": 26750}, {"loss": 0.5237, "grad_norm": 0.7484358549118042, "learning_rate": 0.0002, "epoch": 4.3262468676744, "step": 26760}, {"loss": 0.5728, "grad_norm": 1.0015908479690552, "learning_rate": 0.0002, "epoch": 4.327863551855145, "step": 26770}, {"loss": 0.5597, "grad_norm": 0.8972945809364319, "learning_rate": 0.0002, "epoch": 4.329480236035891, "step": 26780}, {"loss": 0.5857, "grad_norm": 1.01099693775177, "learning_rate": 0.0002, "epoch": 4.331096920216636, "step": 26790}, {"loss": 0.5591, "grad_norm": 0.846958339214325, "learning_rate": 0.0002, "epoch": 4.332713604397381, "step": 26800}, {"loss": 0.5547, "grad_norm": 1.0792603492736816, "learning_rate": 0.0002, "epoch": 4.334330288578126, "step": 26810}, {"loss": 0.5747, "grad_norm": 1.0373345613479614, "learning_rate": 0.0002, "epoch": 4.3359469727588715, "step": 26820}, {"loss": 0.558, "grad_norm": 0.9779167771339417, "learning_rate": 0.0002, "epoch": 4.337563656939617, "step": 26830}, {"loss": 0.5821, "grad_norm": 1.0235520601272583, "learning_rate": 0.0002, "epoch": 4.339180341120362, "step": 26840}, {"loss": 0.5843, "grad_norm": 1.04195237159729, "learning_rate": 0.0002, "epoch": 4.340797025301107, "step": 26850}, {"loss": 0.5474, "grad_norm": 0.9479565620422363, "learning_rate": 0.0002, "epoch": 4.3424137094818525, "step": 26860}, {"loss": 0.5646, "grad_norm": 0.9526172280311584, "learning_rate": 0.0002, "epoch": 4.344030393662598, "step": 26870}, {"loss": 0.521, "grad_norm": 0.8571456074714661, "learning_rate": 0.0002, "epoch": 4.345647077843343, "step": 26880}, {"loss": 0.5846, "grad_norm": 0.9475828409194946, "learning_rate": 0.0002, "epoch": 4.347263762024088, "step": 26890}, {"loss": 0.5815, "grad_norm": 1.0529576539993286, "learning_rate": 0.0002, "epoch": 4.348880446204834, "step": 26900}, {"loss": 0.56, "grad_norm": 0.9648140072822571, "learning_rate": 0.0002, "epoch": 4.3504971303855795, "step": 26910}, {"loss": 0.5162, "grad_norm": 1.0488841533660889, "learning_rate": 0.0002, "epoch": 4.352113814566325, "step": 26920}, {"loss": 0.5842, "grad_norm": 0.8771942257881165, "learning_rate": 0.0002, "epoch": 4.35373049874707, "step": 26930}, {"loss": 0.5966, "grad_norm": 0.9411202073097229, "learning_rate": 0.0002, "epoch": 4.355347182927815, "step": 26940}, {"loss": 0.6001, "grad_norm": 1.0997588634490967, "learning_rate": 0.0002, "epoch": 4.35696386710856, "step": 26950}, {"loss": 0.5528, "grad_norm": 0.968754768371582, "learning_rate": 0.0002, "epoch": 4.358580551289306, "step": 26960}, {"loss": 0.5881, "grad_norm": 0.9990773797035217, "learning_rate": 0.0002, "epoch": 4.360197235470051, "step": 26970}, {"loss": 0.5761, "grad_norm": 1.0210620164871216, "learning_rate": 0.0002, "epoch": 4.361813919650796, "step": 26980}, {"loss": 0.5768, "grad_norm": 0.855462908744812, "learning_rate": 0.0002, "epoch": 4.363430603831541, "step": 26990}, {"loss": 0.5493, "grad_norm": 0.9169660806655884, "learning_rate": 0.0002, "epoch": 4.365047288012287, "step": 27000}, {"loss": 0.5697, "grad_norm": 1.089629888534546, "learning_rate": 0.0002, "epoch": 4.366663972193032, "step": 27010}, {"loss": 0.5854, "grad_norm": 1.0932867527008057, "learning_rate": 0.0002, "epoch": 4.368280656373777, "step": 27020}, {"loss": 0.5656, "grad_norm": 0.9290956854820251, "learning_rate": 0.0002, "epoch": 4.369897340554522, "step": 27030}, {"loss": 0.5727, "grad_norm": 1.2800624370574951, "learning_rate": 0.0002, "epoch": 4.3715140247352675, "step": 27040}, {"loss": 0.5837, "grad_norm": 0.8993493318557739, "learning_rate": 0.0002, "epoch": 4.373130708916014, "step": 27050}, {"loss": 0.6232, "grad_norm": 1.1566431522369385, "learning_rate": 0.0002, "epoch": 4.374747393096759, "step": 27060}, {"loss": 0.5902, "grad_norm": 0.9479052424430847, "learning_rate": 0.0002, "epoch": 4.376364077277504, "step": 27070}, {"loss": 0.6189, "grad_norm": 1.0063648223876953, "learning_rate": 0.0002, "epoch": 4.377980761458249, "step": 27080}, {"loss": 0.561, "grad_norm": 0.8342045545578003, "learning_rate": 0.0002, "epoch": 4.379597445638995, "step": 27090}, {"loss": 0.5515, "grad_norm": 1.1390739679336548, "learning_rate": 0.0002, "epoch": 4.38121412981974, "step": 27100}, {"loss": 0.5372, "grad_norm": 0.9547637104988098, "learning_rate": 0.0002, "epoch": 4.382830814000485, "step": 27110}, {"loss": 0.5728, "grad_norm": 1.0503804683685303, "learning_rate": 0.0002, "epoch": 4.38444749818123, "step": 27120}, {"loss": 0.5787, "grad_norm": 0.9064017534255981, "learning_rate": 0.0002, "epoch": 4.3860641823619755, "step": 27130}, {"loss": 0.5798, "grad_norm": 0.9382519125938416, "learning_rate": 0.0002, "epoch": 4.387680866542721, "step": 27140}, {"loss": 0.5791, "grad_norm": 1.0410341024398804, "learning_rate": 0.0002, "epoch": 4.389297550723466, "step": 27150}, {"loss": 0.6034, "grad_norm": 0.9218655824661255, "learning_rate": 0.0002, "epoch": 4.390914234904211, "step": 27160}, {"loss": 0.5204, "grad_norm": 0.8119737505912781, "learning_rate": 0.0002, "epoch": 4.392530919084956, "step": 27170}, {"loss": 0.5612, "grad_norm": 0.8584722876548767, "learning_rate": 0.0002, "epoch": 4.394147603265702, "step": 27180}, {"loss": 0.5772, "grad_norm": 0.9668293595314026, "learning_rate": 0.0002, "epoch": 4.395764287446447, "step": 27190}, {"loss": 0.6009, "grad_norm": 1.022334098815918, "learning_rate": 0.0002, "epoch": 4.397380971627193, "step": 27200}, {"loss": 0.5573, "grad_norm": 0.9553216099739075, "learning_rate": 0.0002, "epoch": 4.398997655807938, "step": 27210}, {"loss": 0.5604, "grad_norm": 0.9282339215278625, "learning_rate": 0.0002, "epoch": 4.4006143399886835, "step": 27220}, {"loss": 0.5599, "grad_norm": 1.0232292413711548, "learning_rate": 0.0002, "epoch": 4.402231024169429, "step": 27230}, {"loss": 0.6078, "grad_norm": 0.9915700554847717, "learning_rate": 0.0002, "epoch": 4.403847708350174, "step": 27240}, {"loss": 0.5778, "grad_norm": 1.0014961957931519, "learning_rate": 0.0002, "epoch": 4.405464392530919, "step": 27250}, {"loss": 0.5824, "grad_norm": 1.1172103881835938, "learning_rate": 0.0002, "epoch": 4.407081076711664, "step": 27260}, {"loss": 0.5286, "grad_norm": 0.8583093285560608, "learning_rate": 0.0002, "epoch": 4.40869776089241, "step": 27270}, {"loss": 0.5507, "grad_norm": 0.7609201669692993, "learning_rate": 0.0002, "epoch": 4.410314445073155, "step": 27280}, {"loss": 0.575, "grad_norm": 1.0619351863861084, "learning_rate": 0.0002, "epoch": 4.4119311292539, "step": 27290}, {"loss": 0.5579, "grad_norm": 1.0177674293518066, "learning_rate": 0.0002, "epoch": 4.413547813434645, "step": 27300}, {"loss": 0.5628, "grad_norm": 0.9921218156814575, "learning_rate": 0.0002, "epoch": 4.4151644976153905, "step": 27310}, {"loss": 0.6018, "grad_norm": 1.126244306564331, "learning_rate": 0.0002, "epoch": 4.416781181796136, "step": 27320}, {"loss": 0.5743, "grad_norm": 1.0678540468215942, "learning_rate": 0.0002, "epoch": 4.418397865976881, "step": 27330}, {"loss": 0.5665, "grad_norm": 0.8705704212188721, "learning_rate": 0.0002, "epoch": 4.420014550157627, "step": 27340}, {"loss": 0.5763, "grad_norm": 1.272074818611145, "learning_rate": 0.0002, "epoch": 4.421631234338372, "step": 27350}, {"loss": 0.561, "grad_norm": 0.8740444183349609, "learning_rate": 0.0002, "epoch": 4.423247918519118, "step": 27360}, {"loss": 0.5492, "grad_norm": 1.0584250688552856, "learning_rate": 0.0002, "epoch": 4.424864602699863, "step": 27370}, {"loss": 0.589, "grad_norm": 1.059870719909668, "learning_rate": 0.0002, "epoch": 4.426481286880608, "step": 27380}, {"loss": 0.5551, "grad_norm": 1.072265863418579, "learning_rate": 0.0002, "epoch": 4.428097971061353, "step": 27390}, {"loss": 0.5584, "grad_norm": 0.871481716632843, "learning_rate": 0.0002, "epoch": 4.4297146552420985, "step": 27400}, {"loss": 0.5372, "grad_norm": 0.9555448293685913, "learning_rate": 0.0002, "epoch": 4.431331339422844, "step": 27410}, {"loss": 0.5593, "grad_norm": 1.0402292013168335, "learning_rate": 0.0002, "epoch": 4.432948023603589, "step": 27420}, {"loss": 0.5532, "grad_norm": 1.12587571144104, "learning_rate": 0.0002, "epoch": 4.434564707784334, "step": 27430}, {"loss": 0.5403, "grad_norm": 1.0783193111419678, "learning_rate": 0.0002, "epoch": 4.436181391965079, "step": 27440}, {"loss": 0.5313, "grad_norm": 1.024133563041687, "learning_rate": 0.0002, "epoch": 4.437798076145825, "step": 27450}, {"loss": 0.5621, "grad_norm": 0.9156768918037415, "learning_rate": 0.0002, "epoch": 4.43941476032657, "step": 27460}, {"loss": 0.5307, "grad_norm": 1.0215224027633667, "learning_rate": 0.0002, "epoch": 4.441031444507315, "step": 27470}, {"loss": 0.5188, "grad_norm": 1.082116961479187, "learning_rate": 0.0002, "epoch": 4.442648128688061, "step": 27480}, {"loss": 0.6203, "grad_norm": 1.0412873029708862, "learning_rate": 0.0002, "epoch": 4.4442648128688065, "step": 27490}, {"loss": 0.5939, "grad_norm": 1.0509289503097534, "learning_rate": 0.0002, "epoch": 4.445881497049552, "step": 27500}, {"loss": 0.5503, "grad_norm": 0.9291498064994812, "learning_rate": 0.0002, "epoch": 4.447498181230297, "step": 27510}, {"loss": 0.5408, "grad_norm": 0.970184326171875, "learning_rate": 0.0002, "epoch": 4.449114865411042, "step": 27520}, {"loss": 0.5705, "grad_norm": 0.8418883681297302, "learning_rate": 0.0002, "epoch": 4.450731549591787, "step": 27530}, {"loss": 0.5124, "grad_norm": 0.8823825120925903, "learning_rate": 0.0002, "epoch": 4.452348233772533, "step": 27540}, {"loss": 0.5867, "grad_norm": 1.1909019947052002, "learning_rate": 0.0002, "epoch": 4.453964917953278, "step": 27550}, {"loss": 0.5685, "grad_norm": 1.0317302942276, "learning_rate": 0.0002, "epoch": 4.455581602134023, "step": 27560}, {"loss": 0.5538, "grad_norm": 0.9977751970291138, "learning_rate": 0.0002, "epoch": 4.457198286314768, "step": 27570}, {"loss": 0.5628, "grad_norm": 0.8909519910812378, "learning_rate": 0.0002, "epoch": 4.458814970495514, "step": 27580}, {"loss": 0.6099, "grad_norm": 0.8653029799461365, "learning_rate": 0.0002, "epoch": 4.460431654676259, "step": 27590}, {"loss": 0.5622, "grad_norm": 1.0783653259277344, "learning_rate": 0.0002, "epoch": 4.462048338857004, "step": 27600}, {"loss": 0.579, "grad_norm": 1.1235394477844238, "learning_rate": 0.0002, "epoch": 4.463665023037749, "step": 27610}, {"loss": 0.5545, "grad_norm": 0.9386643767356873, "learning_rate": 0.0002, "epoch": 4.4652817072184945, "step": 27620}, {"loss": 0.5554, "grad_norm": 1.0605148077011108, "learning_rate": 0.0002, "epoch": 4.466898391399241, "step": 27630}, {"loss": 0.5886, "grad_norm": 1.1283893585205078, "learning_rate": 0.0002, "epoch": 4.468515075579986, "step": 27640}, {"loss": 0.5801, "grad_norm": 1.0583468675613403, "learning_rate": 0.0002, "epoch": 4.470131759760731, "step": 27650}, {"loss": 0.5601, "grad_norm": 0.9563992023468018, "learning_rate": 0.0002, "epoch": 4.471748443941476, "step": 27660}, {"loss": 0.5687, "grad_norm": 1.100598931312561, "learning_rate": 0.0002, "epoch": 4.4733651281222215, "step": 27670}, {"loss": 0.589, "grad_norm": 0.9386957287788391, "learning_rate": 0.0002, "epoch": 4.474981812302967, "step": 27680}, {"loss": 0.6241, "grad_norm": 1.2946288585662842, "learning_rate": 0.0002, "epoch": 4.476598496483712, "step": 27690}, {"loss": 0.6075, "grad_norm": 1.0325199365615845, "learning_rate": 0.0002, "epoch": 4.478215180664457, "step": 27700}, {"loss": 0.588, "grad_norm": 1.0318928956985474, "learning_rate": 0.0002, "epoch": 4.4798318648452025, "step": 27710}, {"loss": 0.5656, "grad_norm": 0.8721024394035339, "learning_rate": 0.0002, "epoch": 4.481448549025948, "step": 27720}, {"loss": 0.5421, "grad_norm": 1.17376708984375, "learning_rate": 0.0002, "epoch": 4.483065233206693, "step": 27730}, {"loss": 0.5657, "grad_norm": 1.0926326513290405, "learning_rate": 0.0002, "epoch": 4.484681917387438, "step": 27740}, {"loss": 0.5514, "grad_norm": 0.9043852686882019, "learning_rate": 0.0002, "epoch": 4.486298601568183, "step": 27750}, {"loss": 0.582, "grad_norm": 1.064600944519043, "learning_rate": 0.0002, "epoch": 4.487915285748929, "step": 27760}, {"loss": 0.6108, "grad_norm": 0.7833460569381714, "learning_rate": 0.0002, "epoch": 4.489531969929674, "step": 27770}, {"loss": 0.5985, "grad_norm": 1.1073496341705322, "learning_rate": 0.0002, "epoch": 4.49114865411042, "step": 27780}, {"loss": 0.5577, "grad_norm": 1.0799397230148315, "learning_rate": 0.0002, "epoch": 4.492765338291165, "step": 27790}, {"loss": 0.5601, "grad_norm": 1.1062238216400146, "learning_rate": 0.0002, "epoch": 4.49438202247191, "step": 27800}, {"loss": 0.6126, "grad_norm": 1.0568242073059082, "learning_rate": 0.0002, "epoch": 4.495998706652656, "step": 27810}, {"loss": 0.5913, "grad_norm": 0.8861091732978821, "learning_rate": 0.0002, "epoch": 4.497615390833401, "step": 27820}, {"loss": 0.5858, "grad_norm": 1.2297543287277222, "learning_rate": 0.0002, "epoch": 4.499232075014146, "step": 27830}, {"loss": 0.5859, "grad_norm": 0.9600302577018738, "learning_rate": 0.0002, "epoch": 4.500848759194891, "step": 27840}, {"loss": 0.6124, "grad_norm": 1.057051181793213, "learning_rate": 0.0002, "epoch": 4.502465443375637, "step": 27850}, {"loss": 0.5788, "grad_norm": 0.9839690923690796, "learning_rate": 0.0002, "epoch": 4.504082127556382, "step": 27860}, {"loss": 0.555, "grad_norm": 1.1479853391647339, "learning_rate": 0.0002, "epoch": 4.505698811737127, "step": 27870}, {"loss": 0.6039, "grad_norm": 1.0550768375396729, "learning_rate": 0.0002, "epoch": 4.507315495917872, "step": 27880}, {"loss": 0.563, "grad_norm": 0.898209273815155, "learning_rate": 0.0002, "epoch": 4.5089321800986175, "step": 27890}, {"loss": 0.5734, "grad_norm": 0.9460315108299255, "learning_rate": 0.0002, "epoch": 4.510548864279363, "step": 27900}, {"loss": 0.5702, "grad_norm": 0.9499884247779846, "learning_rate": 0.0002, "epoch": 4.512165548460108, "step": 27910}, {"loss": 0.5385, "grad_norm": 0.7801318764686584, "learning_rate": 0.0002, "epoch": 4.513782232640853, "step": 27920}, {"loss": 0.5391, "grad_norm": 0.9286966323852539, "learning_rate": 0.0002, "epoch": 4.515398916821599, "step": 27930}, {"loss": 0.5717, "grad_norm": 0.9539980292320251, "learning_rate": 0.0002, "epoch": 4.517015601002345, "step": 27940}, {"loss": 0.6073, "grad_norm": 1.1053401231765747, "learning_rate": 0.0002, "epoch": 4.51863228518309, "step": 27950}, {"loss": 0.6087, "grad_norm": 0.7535534501075745, "learning_rate": 0.0002, "epoch": 4.520248969363835, "step": 27960}, {"loss": 0.5701, "grad_norm": 1.076926589012146, "learning_rate": 0.0002, "epoch": 4.52186565354458, "step": 27970}, {"loss": 0.6028, "grad_norm": 1.181935429573059, "learning_rate": 0.0002, "epoch": 4.5234823377253255, "step": 27980}, {"loss": 0.6033, "grad_norm": 0.9293407201766968, "learning_rate": 0.0002, "epoch": 4.525099021906071, "step": 27990}, {"loss": 0.5815, "grad_norm": 0.8953009247779846, "learning_rate": 0.0002, "epoch": 4.526715706086816, "step": 28000}, {"loss": 0.5564, "grad_norm": 1.0850225687026978, "learning_rate": 0.0002, "epoch": 4.528332390267561, "step": 28010}, {"loss": 0.5459, "grad_norm": 0.9125663042068481, "learning_rate": 0.0002, "epoch": 4.529949074448306, "step": 28020}, {"loss": 0.5922, "grad_norm": 0.8745216727256775, "learning_rate": 0.0002, "epoch": 4.531565758629052, "step": 28030}, {"loss": 0.567, "grad_norm": 1.0783463716506958, "learning_rate": 0.0002, "epoch": 4.533182442809797, "step": 28040}, {"loss": 0.5754, "grad_norm": 0.7513844966888428, "learning_rate": 0.0002, "epoch": 4.534799126990542, "step": 28050}, {"loss": 0.5608, "grad_norm": 1.0135776996612549, "learning_rate": 0.0002, "epoch": 4.536415811171287, "step": 28060}, {"loss": 0.5827, "grad_norm": 0.8886825442314148, "learning_rate": 0.0002, "epoch": 4.538032495352033, "step": 28070}, {"loss": 0.5605, "grad_norm": 0.8153995275497437, "learning_rate": 0.0002, "epoch": 4.539649179532779, "step": 28080}, {"loss": 0.6377, "grad_norm": 0.9853341579437256, "learning_rate": 0.0002, "epoch": 4.541265863713524, "step": 28090}, {"loss": 0.5957, "grad_norm": 0.9365800023078918, "learning_rate": 0.0002, "epoch": 4.542882547894269, "step": 28100}, {"loss": 0.5477, "grad_norm": 0.9765017628669739, "learning_rate": 0.0002, "epoch": 4.544499232075014, "step": 28110}, {"loss": 0.6185, "grad_norm": 0.9811279773712158, "learning_rate": 0.0002, "epoch": 4.54611591625576, "step": 28120}, {"loss": 0.6095, "grad_norm": 1.0387924909591675, "learning_rate": 0.0002, "epoch": 4.547732600436505, "step": 28130}, {"loss": 0.6534, "grad_norm": 1.0684878826141357, "learning_rate": 0.0002, "epoch": 4.54934928461725, "step": 28140}, {"loss": 0.5701, "grad_norm": 1.0000102519989014, "learning_rate": 0.0002, "epoch": 4.550965968797995, "step": 28150}, {"loss": 0.5327, "grad_norm": 1.0717930793762207, "learning_rate": 0.0002, "epoch": 4.5525826529787405, "step": 28160}, {"loss": 0.5594, "grad_norm": 0.990074634552002, "learning_rate": 0.0002, "epoch": 4.554199337159486, "step": 28170}, {"loss": 0.5452, "grad_norm": 0.8673754930496216, "learning_rate": 0.0002, "epoch": 4.555816021340231, "step": 28180}, {"loss": 0.5773, "grad_norm": 0.864247739315033, "learning_rate": 0.0002, "epoch": 4.557432705520976, "step": 28190}, {"loss": 0.5516, "grad_norm": 0.8280200958251953, "learning_rate": 0.0002, "epoch": 4.5590493897017215, "step": 28200}, {"loss": 0.5709, "grad_norm": 1.1312172412872314, "learning_rate": 0.0002, "epoch": 4.560666073882467, "step": 28210}, {"loss": 0.5776, "grad_norm": 0.9147403240203857, "learning_rate": 0.0002, "epoch": 4.562282758063212, "step": 28220}, {"loss": 0.5591, "grad_norm": 1.0321218967437744, "learning_rate": 0.0002, "epoch": 4.563899442243958, "step": 28230}, {"loss": 0.5508, "grad_norm": 1.168332815170288, "learning_rate": 0.0002, "epoch": 4.565516126424703, "step": 28240}, {"loss": 0.5649, "grad_norm": 1.0067222118377686, "learning_rate": 0.0002, "epoch": 4.5671328106054485, "step": 28250}, {"loss": 0.5853, "grad_norm": 1.0283393859863281, "learning_rate": 0.0002, "epoch": 4.568749494786194, "step": 28260}, {"loss": 0.5772, "grad_norm": 0.9912363886833191, "learning_rate": 0.0002, "epoch": 4.570366178966939, "step": 28270}, {"loss": 0.5757, "grad_norm": 1.108032464981079, "learning_rate": 0.0002, "epoch": 4.571982863147684, "step": 28280}, {"loss": 0.5529, "grad_norm": 0.8260078430175781, "learning_rate": 0.0002, "epoch": 4.573599547328429, "step": 28290}, {"loss": 0.5625, "grad_norm": 0.8946247100830078, "learning_rate": 0.0002, "epoch": 4.575216231509175, "step": 28300}, {"loss": 0.5533, "grad_norm": 0.8273587822914124, "learning_rate": 0.0002, "epoch": 4.57683291568992, "step": 28310}, {"loss": 0.6058, "grad_norm": 0.9040093421936035, "learning_rate": 0.0002, "epoch": 4.578449599870665, "step": 28320}, {"loss": 0.5521, "grad_norm": 0.8435290455818176, "learning_rate": 0.0002, "epoch": 4.58006628405141, "step": 28330}, {"loss": 0.6086, "grad_norm": 1.164088249206543, "learning_rate": 0.0002, "epoch": 4.581682968232156, "step": 28340}, {"loss": 0.5603, "grad_norm": 0.9861085414886475, "learning_rate": 0.0002, "epoch": 4.583299652412901, "step": 28350}, {"loss": 0.5701, "grad_norm": 0.8892980813980103, "learning_rate": 0.0002, "epoch": 4.584916336593646, "step": 28360}, {"loss": 0.598, "grad_norm": 1.240574836730957, "learning_rate": 0.0002, "epoch": 4.586533020774391, "step": 28370}, {"loss": 0.5797, "grad_norm": 0.8669408559799194, "learning_rate": 0.0002, "epoch": 4.588149704955137, "step": 28380}, {"loss": 0.5603, "grad_norm": 0.9145985841751099, "learning_rate": 0.0002, "epoch": 4.589766389135883, "step": 28390}, {"loss": 0.5765, "grad_norm": 0.8584614992141724, "learning_rate": 0.0002, "epoch": 4.591383073316628, "step": 28400}, {"loss": 0.5898, "grad_norm": 1.118829369544983, "learning_rate": 0.0002, "epoch": 4.592999757497373, "step": 28410}, {"loss": 0.5641, "grad_norm": 1.1411553621292114, "learning_rate": 0.0002, "epoch": 4.594616441678118, "step": 28420}, {"loss": 0.549, "grad_norm": 0.9433278441429138, "learning_rate": 0.0002, "epoch": 4.596233125858864, "step": 28430}, {"loss": 0.5496, "grad_norm": 0.816830039024353, "learning_rate": 0.0002, "epoch": 4.597849810039609, "step": 28440}, {"loss": 0.5543, "grad_norm": 1.2124968767166138, "learning_rate": 0.0002, "epoch": 4.599466494220354, "step": 28450}, {"loss": 0.5759, "grad_norm": 0.9658762216567993, "learning_rate": 0.0002, "epoch": 4.601083178401099, "step": 28460}, {"loss": 0.5902, "grad_norm": 0.836100161075592, "learning_rate": 0.0002, "epoch": 4.6026998625818445, "step": 28470}, {"loss": 0.5749, "grad_norm": 0.9989104270935059, "learning_rate": 0.0002, "epoch": 4.60431654676259, "step": 28480}, {"loss": 0.5616, "grad_norm": 1.1298956871032715, "learning_rate": 0.0002, "epoch": 4.605933230943335, "step": 28490}, {"loss": 0.5846, "grad_norm": 1.1731704473495483, "learning_rate": 0.0002, "epoch": 4.60754991512408, "step": 28500}, {"loss": 0.5816, "grad_norm": 0.9624714255332947, "learning_rate": 0.0002, "epoch": 4.609166599304825, "step": 28510}, {"loss": 0.5868, "grad_norm": 1.364073634147644, "learning_rate": 0.0002, "epoch": 4.610783283485571, "step": 28520}, {"loss": 0.6237, "grad_norm": 1.1827356815338135, "learning_rate": 0.0002, "epoch": 4.612399967666317, "step": 28530}, {"loss": 0.5643, "grad_norm": 0.6651531457901001, "learning_rate": 0.0002, "epoch": 4.614016651847062, "step": 28540}, {"loss": 0.6051, "grad_norm": 1.1640995740890503, "learning_rate": 0.0002, "epoch": 4.615633336027807, "step": 28550}, {"loss": 0.5995, "grad_norm": 1.028918743133545, "learning_rate": 0.0002, "epoch": 4.6172500202085525, "step": 28560}, {"loss": 0.5607, "grad_norm": 0.8252120614051819, "learning_rate": 0.0002, "epoch": 4.618866704389298, "step": 28570}, {"loss": 0.5769, "grad_norm": 1.3536735773086548, "learning_rate": 0.0002, "epoch": 4.620483388570043, "step": 28580}, {"loss": 0.6006, "grad_norm": 1.2146915197372437, "learning_rate": 0.0002, "epoch": 4.622100072750788, "step": 28590}, {"loss": 0.5503, "grad_norm": 1.0122549533843994, "learning_rate": 0.0002, "epoch": 4.623716756931533, "step": 28600}, {"loss": 0.6072, "grad_norm": 0.9977872967720032, "learning_rate": 0.0002, "epoch": 4.625333441112279, "step": 28610}, {"loss": 0.5669, "grad_norm": 1.0159751176834106, "learning_rate": 0.0002, "epoch": 4.626950125293024, "step": 28620}, {"loss": 0.5935, "grad_norm": 1.0028325319290161, "learning_rate": 0.0002, "epoch": 4.628566809473769, "step": 28630}, {"loss": 0.5515, "grad_norm": 0.901638388633728, "learning_rate": 0.0002, "epoch": 4.630183493654514, "step": 28640}, {"loss": 0.595, "grad_norm": 0.9450507164001465, "learning_rate": 0.0002, "epoch": 4.6318001778352595, "step": 28650}, {"loss": 0.5972, "grad_norm": 0.9987545013427734, "learning_rate": 0.0002, "epoch": 4.633416862016006, "step": 28660}, {"loss": 0.5863, "grad_norm": 0.9574332237243652, "learning_rate": 0.0002, "epoch": 4.63503354619675, "step": 28670}, {"loss": 0.5804, "grad_norm": 1.2215653657913208, "learning_rate": 0.0002, "epoch": 4.636650230377496, "step": 28680}, {"loss": 0.5798, "grad_norm": 0.9798858761787415, "learning_rate": 0.0002, "epoch": 4.638266914558241, "step": 28690}, {"loss": 0.5773, "grad_norm": 1.0648466348648071, "learning_rate": 0.0002, "epoch": 4.639883598738987, "step": 28700}, {"loss": 0.6108, "grad_norm": 1.0606504678726196, "learning_rate": 0.0002, "epoch": 4.641500282919732, "step": 28710}, {"loss": 0.5801, "grad_norm": 1.0892442464828491, "learning_rate": 0.0002, "epoch": 4.643116967100477, "step": 28720}, {"loss": 0.5492, "grad_norm": 0.914391040802002, "learning_rate": 0.0002, "epoch": 4.644733651281222, "step": 28730}, {"loss": 0.5439, "grad_norm": 0.9782370328903198, "learning_rate": 0.0002, "epoch": 4.6463503354619675, "step": 28740}, {"loss": 0.6035, "grad_norm": 1.0344339609146118, "learning_rate": 0.0002, "epoch": 4.647967019642713, "step": 28750}, {"loss": 0.5775, "grad_norm": 1.0513931512832642, "learning_rate": 0.0002, "epoch": 4.649583703823458, "step": 28760}, {"loss": 0.546, "grad_norm": 0.9711475968360901, "learning_rate": 0.0002, "epoch": 4.651200388004203, "step": 28770}, {"loss": 0.5472, "grad_norm": 0.977519690990448, "learning_rate": 0.0002, "epoch": 4.652817072184948, "step": 28780}, {"loss": 0.5826, "grad_norm": 0.9150224924087524, "learning_rate": 0.0002, "epoch": 4.654433756365694, "step": 28790}, {"loss": 0.5382, "grad_norm": 1.0973542928695679, "learning_rate": 0.0002, "epoch": 4.656050440546439, "step": 28800}, {"loss": 0.6147, "grad_norm": 0.944877564907074, "learning_rate": 0.0002, "epoch": 4.657667124727185, "step": 28810}, {"loss": 0.5537, "grad_norm": 0.9508748650550842, "learning_rate": 0.0002, "epoch": 4.659283808907929, "step": 28820}, {"loss": 0.5537, "grad_norm": 0.9681721329689026, "learning_rate": 0.0002, "epoch": 4.6609004930886755, "step": 28830}, {"loss": 0.592, "grad_norm": 1.0214351415634155, "learning_rate": 0.0002, "epoch": 4.662517177269421, "step": 28840}, {"loss": 0.6031, "grad_norm": 0.9748611450195312, "learning_rate": 0.0002, "epoch": 4.664133861450166, "step": 28850}, {"loss": 0.572, "grad_norm": 0.8484147191047668, "learning_rate": 0.0002, "epoch": 4.665750545630911, "step": 28860}, {"loss": 0.5699, "grad_norm": 1.1252986192703247, "learning_rate": 0.0002, "epoch": 4.667367229811656, "step": 28870}, {"loss": 0.5724, "grad_norm": 0.8706206679344177, "learning_rate": 0.0002, "epoch": 4.668983913992402, "step": 28880}, {"loss": 0.6002, "grad_norm": 1.1432424783706665, "learning_rate": 0.0002, "epoch": 4.670600598173147, "step": 28890}, {"loss": 0.5675, "grad_norm": 1.017029047012329, "learning_rate": 0.0002, "epoch": 4.672217282353892, "step": 28900}, {"loss": 0.5831, "grad_norm": 1.085597038269043, "learning_rate": 0.0002, "epoch": 4.673833966534637, "step": 28910}, {"loss": 0.5678, "grad_norm": 0.9275796413421631, "learning_rate": 0.0002, "epoch": 4.675450650715383, "step": 28920}, {"loss": 0.5603, "grad_norm": 0.9518964886665344, "learning_rate": 0.0002, "epoch": 4.677067334896128, "step": 28930}, {"loss": 0.6232, "grad_norm": 1.0352122783660889, "learning_rate": 0.0002, "epoch": 4.678684019076873, "step": 28940}, {"loss": 0.5786, "grad_norm": 1.090124249458313, "learning_rate": 0.0002, "epoch": 4.680300703257618, "step": 28950}, {"loss": 0.5728, "grad_norm": 0.8799563050270081, "learning_rate": 0.0002, "epoch": 4.681917387438364, "step": 28960}, {"loss": 0.5787, "grad_norm": 1.0929821729660034, "learning_rate": 0.0002, "epoch": 4.683534071619109, "step": 28970}, {"loss": 0.6134, "grad_norm": 0.903727650642395, "learning_rate": 0.0002, "epoch": 4.685150755799855, "step": 28980}, {"loss": 0.5522, "grad_norm": 0.9752424955368042, "learning_rate": 0.0002, "epoch": 4.6867674399806, "step": 28990}, {"loss": 0.5762, "grad_norm": 0.9351571202278137, "learning_rate": 0.0002, "epoch": 4.688384124161345, "step": 29000}, {"loss": 0.5811, "grad_norm": 0.923877477645874, "learning_rate": 0.0002, "epoch": 4.6900008083420905, "step": 29010}, {"loss": 0.5682, "grad_norm": 1.045389175415039, "learning_rate": 0.0002, "epoch": 4.691617492522836, "step": 29020}, {"loss": 0.584, "grad_norm": 1.0200831890106201, "learning_rate": 0.0002, "epoch": 4.693234176703581, "step": 29030}, {"loss": 0.5514, "grad_norm": 1.1499706506729126, "learning_rate": 0.0002, "epoch": 4.694850860884326, "step": 29040}, {"loss": 0.5745, "grad_norm": 0.860118567943573, "learning_rate": 0.0002, "epoch": 4.6964675450650715, "step": 29050}, {"loss": 0.5741, "grad_norm": 0.9774864315986633, "learning_rate": 0.0002, "epoch": 4.698084229245817, "step": 29060}, {"loss": 0.5765, "grad_norm": 1.0323210954666138, "learning_rate": 0.0002, "epoch": 4.699700913426562, "step": 29070}, {"loss": 0.5452, "grad_norm": 0.8492481112480164, "learning_rate": 0.0002, "epoch": 4.701317597607307, "step": 29080}, {"loss": 0.5985, "grad_norm": 1.131951093673706, "learning_rate": 0.0002, "epoch": 4.702934281788052, "step": 29090}, {"loss": 0.6412, "grad_norm": 0.8763113021850586, "learning_rate": 0.0002, "epoch": 4.704550965968798, "step": 29100}, {"loss": 0.575, "grad_norm": 1.045028805732727, "learning_rate": 0.0002, "epoch": 4.706167650149544, "step": 29110}, {"loss": 0.5548, "grad_norm": 0.9961401224136353, "learning_rate": 0.0002, "epoch": 4.707784334330288, "step": 29120}, {"loss": 0.559, "grad_norm": 0.9282503724098206, "learning_rate": 0.0002, "epoch": 4.709401018511034, "step": 29130}, {"loss": 0.5744, "grad_norm": 1.1418932676315308, "learning_rate": 0.0002, "epoch": 4.711017702691779, "step": 29140}, {"loss": 0.5394, "grad_norm": 0.9950099587440491, "learning_rate": 0.0002, "epoch": 4.712634386872525, "step": 29150}, {"loss": 0.6177, "grad_norm": 0.8304893374443054, "learning_rate": 0.0002, "epoch": 4.71425107105327, "step": 29160}, {"loss": 0.6074, "grad_norm": 1.115626335144043, "learning_rate": 0.0002, "epoch": 4.715867755234015, "step": 29170}, {"loss": 0.6265, "grad_norm": 1.079818606376648, "learning_rate": 0.0002, "epoch": 4.71748443941476, "step": 29180}, {"loss": 0.561, "grad_norm": 1.1929082870483398, "learning_rate": 0.0002, "epoch": 4.719101123595506, "step": 29190}, {"loss": 0.5708, "grad_norm": 0.9621080756187439, "learning_rate": 0.0002, "epoch": 4.720717807776251, "step": 29200}, {"loss": 0.546, "grad_norm": 0.8549222350120544, "learning_rate": 0.0002, "epoch": 4.722334491956996, "step": 29210}, {"loss": 0.5775, "grad_norm": 0.9341941475868225, "learning_rate": 0.0002, "epoch": 4.723951176137741, "step": 29220}, {"loss": 0.5436, "grad_norm": 1.075406789779663, "learning_rate": 0.0002, "epoch": 4.7255678603184865, "step": 29230}, {"loss": 0.576, "grad_norm": 1.0859880447387695, "learning_rate": 0.0002, "epoch": 4.727184544499232, "step": 29240}, {"loss": 0.5525, "grad_norm": 0.8475605249404907, "learning_rate": 0.0002, "epoch": 4.728801228679977, "step": 29250}, {"loss": 0.5659, "grad_norm": 0.9331845641136169, "learning_rate": 0.0002, "epoch": 4.730417912860723, "step": 29260}, {"loss": 0.5901, "grad_norm": 0.9279314279556274, "learning_rate": 0.0002, "epoch": 4.7320345970414674, "step": 29270}, {"loss": 0.597, "grad_norm": 0.7803558707237244, "learning_rate": 0.0002, "epoch": 4.733651281222214, "step": 29280}, {"loss": 0.5968, "grad_norm": 1.0159329175949097, "learning_rate": 0.0002, "epoch": 4.735267965402959, "step": 29290}, {"loss": 0.5333, "grad_norm": 0.9448670744895935, "learning_rate": 0.0002, "epoch": 4.736884649583704, "step": 29300}, {"loss": 0.574, "grad_norm": 1.0732197761535645, "learning_rate": 0.0002, "epoch": 4.738501333764449, "step": 29310}, {"loss": 0.6066, "grad_norm": 0.901830792427063, "learning_rate": 0.0002, "epoch": 4.7401180179451945, "step": 29320}, {"loss": 0.6105, "grad_norm": 0.9141789674758911, "learning_rate": 0.0002, "epoch": 4.74173470212594, "step": 29330}, {"loss": 0.5481, "grad_norm": 0.9733418226242065, "learning_rate": 0.0002, "epoch": 4.743351386306685, "step": 29340}, {"loss": 0.612, "grad_norm": 0.909810483455658, "learning_rate": 0.0002, "epoch": 4.74496807048743, "step": 29350}, {"loss": 0.5911, "grad_norm": 0.909541666507721, "learning_rate": 0.0002, "epoch": 4.746584754668175, "step": 29360}, {"loss": 0.5579, "grad_norm": 0.9383015632629395, "learning_rate": 0.0002, "epoch": 4.748201438848921, "step": 29370}, {"loss": 0.5529, "grad_norm": 0.9275668263435364, "learning_rate": 0.0002, "epoch": 4.749818123029666, "step": 29380}, {"loss": 0.5623, "grad_norm": 1.1146225929260254, "learning_rate": 0.0002, "epoch": 4.751434807210411, "step": 29390}, {"loss": 0.6018, "grad_norm": 1.0062453746795654, "learning_rate": 0.0002, "epoch": 4.753051491391156, "step": 29400}, {"loss": 0.5872, "grad_norm": 0.9451895952224731, "learning_rate": 0.0002, "epoch": 4.7546681755719025, "step": 29410}, {"loss": 0.5767, "grad_norm": 0.870457649230957, "learning_rate": 0.0002, "epoch": 4.756284859752648, "step": 29420}, {"loss": 0.57, "grad_norm": 1.0411282777786255, "learning_rate": 0.0002, "epoch": 4.757901543933393, "step": 29430}, {"loss": 0.5688, "grad_norm": 1.1648986339569092, "learning_rate": 0.0002, "epoch": 4.759518228114138, "step": 29440}, {"loss": 0.5432, "grad_norm": 0.8999572992324829, "learning_rate": 0.0002, "epoch": 4.761134912294883, "step": 29450}, {"loss": 0.5667, "grad_norm": 0.9863559007644653, "learning_rate": 0.0002, "epoch": 4.762751596475629, "step": 29460}, {"loss": 0.5779, "grad_norm": 0.9676542282104492, "learning_rate": 0.0002, "epoch": 4.764368280656374, "step": 29470}, {"loss": 0.6075, "grad_norm": 1.004775047302246, "learning_rate": 0.0002, "epoch": 4.765984964837119, "step": 29480}, {"loss": 0.6044, "grad_norm": 1.0937515497207642, "learning_rate": 0.0002, "epoch": 4.767601649017864, "step": 29490}, {"loss": 0.5433, "grad_norm": 0.9551598429679871, "learning_rate": 0.0002, "epoch": 4.7692183331986095, "step": 29500}, {"loss": 0.5609, "grad_norm": 1.0757228136062622, "learning_rate": 0.0002, "epoch": 4.770835017379355, "step": 29510}, {"loss": 0.567, "grad_norm": 1.0588841438293457, "learning_rate": 0.0002, "epoch": 4.7724517015601, "step": 29520}, {"loss": 0.5814, "grad_norm": 1.0744032859802246, "learning_rate": 0.0002, "epoch": 4.774068385740845, "step": 29530}, {"loss": 0.5681, "grad_norm": 1.0066277980804443, "learning_rate": 0.0002, "epoch": 4.7756850699215905, "step": 29540}, {"loss": 0.545, "grad_norm": 1.082319736480713, "learning_rate": 0.0002, "epoch": 4.777301754102336, "step": 29550}, {"loss": 0.5709, "grad_norm": 0.8252472877502441, "learning_rate": 0.0002, "epoch": 4.778918438283082, "step": 29560}, {"loss": 0.5666, "grad_norm": 0.9855340123176575, "learning_rate": 0.0002, "epoch": 4.780535122463827, "step": 29570}, {"loss": 0.6117, "grad_norm": 0.9991421699523926, "learning_rate": 0.0002, "epoch": 4.782151806644572, "step": 29580}, {"loss": 0.5966, "grad_norm": 1.316841959953308, "learning_rate": 0.0002, "epoch": 4.7837684908253175, "step": 29590}, {"loss": 0.6102, "grad_norm": 1.1513035297393799, "learning_rate": 0.0002, "epoch": 4.785385175006063, "step": 29600}, {"loss": 0.5785, "grad_norm": 0.9767683744430542, "learning_rate": 0.0002, "epoch": 4.787001859186808, "step": 29610}, {"loss": 0.6037, "grad_norm": 0.9786278605461121, "learning_rate": 0.0002, "epoch": 4.788618543367553, "step": 29620}, {"loss": 0.6108, "grad_norm": 0.8004973530769348, "learning_rate": 0.0002, "epoch": 4.7902352275482984, "step": 29630}, {"loss": 0.5932, "grad_norm": 1.0997767448425293, "learning_rate": 0.0002, "epoch": 4.791851911729044, "step": 29640}, {"loss": 0.5655, "grad_norm": 0.9752856492996216, "learning_rate": 0.0002, "epoch": 4.793468595909789, "step": 29650}, {"loss": 0.5916, "grad_norm": 1.0518392324447632, "learning_rate": 0.0002, "epoch": 4.795085280090534, "step": 29660}, {"loss": 0.6042, "grad_norm": 1.1050055027008057, "learning_rate": 0.0002, "epoch": 4.796701964271279, "step": 29670}, {"loss": 0.6089, "grad_norm": 0.9933857917785645, "learning_rate": 0.0002, "epoch": 4.798318648452025, "step": 29680}, {"loss": 0.6041, "grad_norm": 1.2804018259048462, "learning_rate": 0.0002, "epoch": 4.79993533263277, "step": 29690}, {"loss": 0.636, "grad_norm": 1.0133371353149414, "learning_rate": 0.0002, "epoch": 4.801552016813515, "step": 29700}, {"loss": 0.5662, "grad_norm": 1.080350637435913, "learning_rate": 0.0002, "epoch": 4.803168700994261, "step": 29710}, {"loss": 0.5603, "grad_norm": 0.9986529350280762, "learning_rate": 0.0002, "epoch": 4.804785385175006, "step": 29720}, {"loss": 0.5894, "grad_norm": 0.975665807723999, "learning_rate": 0.0002, "epoch": 4.806402069355752, "step": 29730}, {"loss": 0.6328, "grad_norm": 0.8458138704299927, "learning_rate": 0.0002, "epoch": 4.808018753536497, "step": 29740}, {"loss": 0.5837, "grad_norm": 0.99330073595047, "learning_rate": 0.0002, "epoch": 4.809635437717242, "step": 29750}, {"loss": 0.5507, "grad_norm": 0.898274302482605, "learning_rate": 0.0002, "epoch": 4.811252121897987, "step": 29760}, {"loss": 0.5842, "grad_norm": 1.0504480600357056, "learning_rate": 0.0002, "epoch": 4.812868806078733, "step": 29770}, {"loss": 0.5821, "grad_norm": 0.937919020652771, "learning_rate": 0.0002, "epoch": 4.814485490259478, "step": 29780}, {"loss": 0.5885, "grad_norm": 0.9593307971954346, "learning_rate": 0.0002, "epoch": 4.816102174440223, "step": 29790}, {"loss": 0.578, "grad_norm": 0.9431198835372925, "learning_rate": 0.0002, "epoch": 4.817718858620968, "step": 29800}, {"loss": 0.5739, "grad_norm": 1.2729957103729248, "learning_rate": 0.0002, "epoch": 4.8193355428017135, "step": 29810}, {"loss": 0.6124, "grad_norm": 0.8876838684082031, "learning_rate": 0.0002, "epoch": 4.820952226982459, "step": 29820}, {"loss": 0.5583, "grad_norm": 1.0185000896453857, "learning_rate": 0.0002, "epoch": 4.822568911163204, "step": 29830}, {"loss": 0.5686, "grad_norm": 1.064276099205017, "learning_rate": 0.0002, "epoch": 4.824185595343949, "step": 29840}, {"loss": 0.5698, "grad_norm": 0.9774803519248962, "learning_rate": 0.0002, "epoch": 4.825802279524694, "step": 29850}, {"loss": 0.5533, "grad_norm": 1.131646990776062, "learning_rate": 0.0002, "epoch": 4.8274189637054405, "step": 29860}, {"loss": 0.6371, "grad_norm": 1.081455945968628, "learning_rate": 0.0002, "epoch": 4.829035647886186, "step": 29870}, {"loss": 0.5793, "grad_norm": 0.990538477897644, "learning_rate": 0.0002, "epoch": 4.830652332066931, "step": 29880}, {"loss": 0.5833, "grad_norm": 0.9750600457191467, "learning_rate": 0.0002, "epoch": 4.832269016247676, "step": 29890}, {"loss": 0.619, "grad_norm": 1.0600621700286865, "learning_rate": 0.0002, "epoch": 4.8338857004284215, "step": 29900}, {"loss": 0.5841, "grad_norm": 0.9237320423126221, "learning_rate": 0.0002, "epoch": 4.835502384609167, "step": 29910}, {"loss": 0.5513, "grad_norm": 0.9739177227020264, "learning_rate": 0.0002, "epoch": 4.837119068789912, "step": 29920}, {"loss": 0.587, "grad_norm": 1.128677248954773, "learning_rate": 0.0002, "epoch": 4.838735752970657, "step": 29930}, {"loss": 0.564, "grad_norm": 1.042604923248291, "learning_rate": 0.0002, "epoch": 4.840352437151402, "step": 29940}, {"loss": 0.5885, "grad_norm": 0.849758505821228, "learning_rate": 0.0002, "epoch": 4.841969121332148, "step": 29950}, {"loss": 0.5952, "grad_norm": 1.2809888124465942, "learning_rate": 0.0002, "epoch": 4.843585805512893, "step": 29960}, {"loss": 0.5703, "grad_norm": 1.0177865028381348, "learning_rate": 0.0002, "epoch": 4.845202489693638, "step": 29970}, {"loss": 0.5946, "grad_norm": 1.0026639699935913, "learning_rate": 0.0002, "epoch": 4.846819173874383, "step": 29980}, {"loss": 0.5897, "grad_norm": 0.9679505228996277, "learning_rate": 0.0002, "epoch": 4.8484358580551286, "step": 29990}, {"loss": 0.5621, "grad_norm": 0.8939532041549683, "learning_rate": 0.0002, "epoch": 4.850052542235874, "step": 30000}, {"loss": 0.5852, "grad_norm": 0.9957457780838013, "learning_rate": 0.0002, "epoch": 4.85166922641662, "step": 30010}, {"loss": 0.6117, "grad_norm": 1.1646790504455566, "learning_rate": 0.0002, "epoch": 4.853285910597365, "step": 30020}, {"loss": 0.5711, "grad_norm": 0.8804680705070496, "learning_rate": 0.0002, "epoch": 4.85490259477811, "step": 30030}, {"loss": 0.5397, "grad_norm": 1.161970853805542, "learning_rate": 0.0002, "epoch": 4.856519278958856, "step": 30040}, {"loss": 0.5552, "grad_norm": 0.9081037640571594, "learning_rate": 0.0002, "epoch": 4.858135963139601, "step": 30050}, {"loss": 0.6024, "grad_norm": 0.9402848482131958, "learning_rate": 0.0002, "epoch": 4.859752647320346, "step": 30060}, {"loss": 0.6256, "grad_norm": 0.9023865461349487, "learning_rate": 0.0002, "epoch": 4.861369331501091, "step": 30070}, {"loss": 0.5926, "grad_norm": 1.0173414945602417, "learning_rate": 0.0002, "epoch": 4.8629860156818365, "step": 30080}, {"loss": 0.6274, "grad_norm": 1.084402322769165, "learning_rate": 0.0002, "epoch": 4.864602699862582, "step": 30090}, {"loss": 0.6311, "grad_norm": 0.9577937126159668, "learning_rate": 0.0002, "epoch": 4.866219384043327, "step": 30100}, {"loss": 0.5724, "grad_norm": 0.9807606935501099, "learning_rate": 0.0002, "epoch": 4.867836068224072, "step": 30110}, {"loss": 0.5786, "grad_norm": 0.978784441947937, "learning_rate": 0.0002, "epoch": 4.8694527524048175, "step": 30120}, {"loss": 0.6194, "grad_norm": 0.9762914776802063, "learning_rate": 0.0002, "epoch": 4.871069436585563, "step": 30130}, {"loss": 0.5892, "grad_norm": 0.9404871463775635, "learning_rate": 0.0002, "epoch": 4.872686120766308, "step": 30140}, {"loss": 0.6182, "grad_norm": 1.0069509744644165, "learning_rate": 0.0002, "epoch": 4.874302804947053, "step": 30150}, {"loss": 0.6225, "grad_norm": 1.1770923137664795, "learning_rate": 0.0002, "epoch": 4.875919489127799, "step": 30160}, {"loss": 0.5657, "grad_norm": 1.021210789680481, "learning_rate": 0.0002, "epoch": 4.8775361733085445, "step": 30170}, {"loss": 0.6033, "grad_norm": 0.8512648940086365, "learning_rate": 0.0002, "epoch": 4.87915285748929, "step": 30180}, {"loss": 0.5519, "grad_norm": 0.9345870018005371, "learning_rate": 0.0002, "epoch": 4.880769541670035, "step": 30190}, {"loss": 0.5682, "grad_norm": 1.0224418640136719, "learning_rate": 0.0002, "epoch": 4.88238622585078, "step": 30200}, {"loss": 0.5807, "grad_norm": 1.0316044092178345, "learning_rate": 0.0002, "epoch": 4.884002910031525, "step": 30210}, {"loss": 0.6065, "grad_norm": 1.102437973022461, "learning_rate": 0.0002, "epoch": 4.885619594212271, "step": 30220}, {"loss": 0.586, "grad_norm": 1.0220023393630981, "learning_rate": 0.0002, "epoch": 4.887236278393016, "step": 30230}, {"loss": 0.5781, "grad_norm": 1.0934523344039917, "learning_rate": 0.0002, "epoch": 4.888852962573761, "step": 30240}, {"loss": 0.6313, "grad_norm": 1.264630913734436, "learning_rate": 0.0002, "epoch": 4.890469646754506, "step": 30250}, {"loss": 0.5712, "grad_norm": 1.0999879837036133, "learning_rate": 0.0002, "epoch": 4.892086330935252, "step": 30260}, {"loss": 0.6413, "grad_norm": 0.9124550223350525, "learning_rate": 0.0002, "epoch": 4.893703015115997, "step": 30270}, {"loss": 0.596, "grad_norm": 0.9853624105453491, "learning_rate": 0.0002, "epoch": 4.895319699296742, "step": 30280}, {"loss": 0.595, "grad_norm": 1.0589802265167236, "learning_rate": 0.0002, "epoch": 4.896936383477488, "step": 30290}, {"loss": 0.6129, "grad_norm": 0.8487226366996765, "learning_rate": 0.0002, "epoch": 4.8985530676582325, "step": 30300}, {"loss": 0.5514, "grad_norm": 1.0212191343307495, "learning_rate": 0.0002, "epoch": 4.900169751838979, "step": 30310}, {"loss": 0.5896, "grad_norm": 1.0187491178512573, "learning_rate": 0.0002, "epoch": 4.901786436019724, "step": 30320}, {"loss": 0.5809, "grad_norm": 1.0013091564178467, "learning_rate": 0.0002, "epoch": 4.903403120200469, "step": 30330}, {"loss": 0.5658, "grad_norm": 1.0017542839050293, "learning_rate": 0.0002, "epoch": 4.905019804381214, "step": 30340}, {"loss": 0.6002, "grad_norm": 0.9665151238441467, "learning_rate": 0.0002, "epoch": 4.9066364885619596, "step": 30350}, {"loss": 0.5864, "grad_norm": 0.8774822950363159, "learning_rate": 0.0002, "epoch": 4.908253172742705, "step": 30360}, {"loss": 0.5771, "grad_norm": 0.9449850916862488, "learning_rate": 0.0002, "epoch": 4.90986985692345, "step": 30370}, {"loss": 0.58, "grad_norm": 0.7368341088294983, "learning_rate": 0.0002, "epoch": 4.911486541104195, "step": 30380}, {"loss": 0.5992, "grad_norm": 0.9669167995452881, "learning_rate": 0.0002, "epoch": 4.9131032252849405, "step": 30390}, {"loss": 0.6202, "grad_norm": 1.1227794885635376, "learning_rate": 0.0002, "epoch": 4.914719909465686, "step": 30400}, {"loss": 0.6181, "grad_norm": 0.9884361028671265, "learning_rate": 0.0002, "epoch": 4.916336593646431, "step": 30410}, {"loss": 0.6185, "grad_norm": 0.9949551224708557, "learning_rate": 0.0002, "epoch": 4.917953277827176, "step": 30420}, {"loss": 0.5866, "grad_norm": 0.9491621851921082, "learning_rate": 0.0002, "epoch": 4.919569962007921, "step": 30430}, {"loss": 0.6005, "grad_norm": 0.78848797082901, "learning_rate": 0.0002, "epoch": 4.9211866461886675, "step": 30440}, {"loss": 0.5561, "grad_norm": 1.0693835020065308, "learning_rate": 0.0002, "epoch": 4.922803330369412, "step": 30450}, {"loss": 0.566, "grad_norm": 0.9573729634284973, "learning_rate": 0.0002, "epoch": 4.924420014550158, "step": 30460}, {"loss": 0.6084, "grad_norm": 0.9975152611732483, "learning_rate": 0.0002, "epoch": 4.926036698730903, "step": 30470}, {"loss": 0.5969, "grad_norm": 0.8695693016052246, "learning_rate": 0.0002, "epoch": 4.9276533829116484, "step": 30480}, {"loss": 0.6144, "grad_norm": 1.145394206047058, "learning_rate": 0.0002, "epoch": 4.929270067092394, "step": 30490}, {"loss": 0.5736, "grad_norm": 0.7668989896774292, "learning_rate": 0.0002, "epoch": 4.930886751273139, "step": 30500}, {"loss": 0.6052, "grad_norm": 0.9630151391029358, "learning_rate": 0.0002, "epoch": 4.932503435453884, "step": 30510}, {"loss": 0.6461, "grad_norm": 0.940705418586731, "learning_rate": 0.0002, "epoch": 4.934120119634629, "step": 30520}, {"loss": 0.6326, "grad_norm": 1.3243348598480225, "learning_rate": 0.0002, "epoch": 4.935736803815375, "step": 30530}, {"loss": 0.6174, "grad_norm": 1.004347801208496, "learning_rate": 0.0002, "epoch": 4.93735348799612, "step": 30540}, {"loss": 0.583, "grad_norm": 0.8711541295051575, "learning_rate": 0.0002, "epoch": 4.938970172176865, "step": 30550}, {"loss": 0.599, "grad_norm": 0.8980631828308105, "learning_rate": 0.0002, "epoch": 4.94058685635761, "step": 30560}, {"loss": 0.6024, "grad_norm": 0.8388893604278564, "learning_rate": 0.0002, "epoch": 4.9422035405383555, "step": 30570}, {"loss": 0.6189, "grad_norm": 1.0991183519363403, "learning_rate": 0.0002, "epoch": 4.943820224719101, "step": 30580}, {"loss": 0.5906, "grad_norm": 0.9731075763702393, "learning_rate": 0.0002, "epoch": 4.945436908899847, "step": 30590}, {"loss": 0.5883, "grad_norm": 1.3904452323913574, "learning_rate": 0.0002, "epoch": 4.947053593080591, "step": 30600}, {"loss": 0.5952, "grad_norm": 1.2489882707595825, "learning_rate": 0.0002, "epoch": 4.948670277261337, "step": 30610}, {"loss": 0.5887, "grad_norm": 1.240072250366211, "learning_rate": 0.0002, "epoch": 4.950286961442083, "step": 30620}, {"loss": 0.5762, "grad_norm": 0.9191411733627319, "learning_rate": 0.0002, "epoch": 4.951903645622828, "step": 30630}, {"loss": 0.5597, "grad_norm": 0.8888895511627197, "learning_rate": 0.0002, "epoch": 4.953520329803573, "step": 30640}, {"loss": 0.6594, "grad_norm": 0.9001450538635254, "learning_rate": 0.0002, "epoch": 4.955137013984318, "step": 30650}, {"loss": 0.6047, "grad_norm": 1.053971767425537, "learning_rate": 0.0002, "epoch": 4.9567536981650635, "step": 30660}, {"loss": 0.6107, "grad_norm": 1.2224042415618896, "learning_rate": 0.0002, "epoch": 4.958370382345809, "step": 30670}, {"loss": 0.6211, "grad_norm": 0.8855111598968506, "learning_rate": 0.0002, "epoch": 4.959987066526554, "step": 30680}, {"loss": 0.5764, "grad_norm": 0.9489575624465942, "learning_rate": 0.0002, "epoch": 4.961603750707299, "step": 30690}, {"loss": 0.5371, "grad_norm": 0.9635404944419861, "learning_rate": 0.0002, "epoch": 4.963220434888044, "step": 30700}, {"loss": 0.6043, "grad_norm": 1.1784121990203857, "learning_rate": 0.0002, "epoch": 4.96483711906879, "step": 30710}, {"loss": 0.5803, "grad_norm": 1.0059462785720825, "learning_rate": 0.0002, "epoch": 4.966453803249535, "step": 30720}, {"loss": 0.5759, "grad_norm": 0.9479738473892212, "learning_rate": 0.0002, "epoch": 4.96807048743028, "step": 30730}, {"loss": 0.584, "grad_norm": 1.0624593496322632, "learning_rate": 0.0002, "epoch": 4.969687171611026, "step": 30740}, {"loss": 0.6202, "grad_norm": 1.1429259777069092, "learning_rate": 0.0002, "epoch": 4.971303855791771, "step": 30750}, {"loss": 0.6174, "grad_norm": 0.9102491140365601, "learning_rate": 0.0002, "epoch": 4.972920539972517, "step": 30760}, {"loss": 0.6025, "grad_norm": 1.1262688636779785, "learning_rate": 0.0002, "epoch": 4.974537224153262, "step": 30770}, {"loss": 0.588, "grad_norm": 1.1415393352508545, "learning_rate": 0.0002, "epoch": 4.976153908334007, "step": 30780}, {"loss": 0.5832, "grad_norm": 1.083078384399414, "learning_rate": 0.0002, "epoch": 4.977770592514752, "step": 30790}, {"loss": 0.6025, "grad_norm": 0.964859127998352, "learning_rate": 0.0002, "epoch": 4.979387276695498, "step": 30800}, {"loss": 0.6095, "grad_norm": 0.8704743385314941, "learning_rate": 0.0002, "epoch": 4.981003960876243, "step": 30810}, {"loss": 0.5666, "grad_norm": 1.0714856386184692, "learning_rate": 0.0002, "epoch": 4.982620645056988, "step": 30820}, {"loss": 0.565, "grad_norm": 0.6818771362304688, "learning_rate": 0.0002, "epoch": 4.984237329237733, "step": 30830}, {"loss": 0.5999, "grad_norm": 1.0454156398773193, "learning_rate": 0.0002, "epoch": 4.985854013418479, "step": 30840}, {"loss": 0.5683, "grad_norm": 0.9410776495933533, "learning_rate": 0.0002, "epoch": 4.987470697599224, "step": 30850}, {"loss": 0.5899, "grad_norm": 1.0878902673721313, "learning_rate": 0.0002, "epoch": 4.989087381779969, "step": 30860}, {"loss": 0.5914, "grad_norm": 0.8916727304458618, "learning_rate": 0.0002, "epoch": 4.990704065960714, "step": 30870}, {"loss": 0.6066, "grad_norm": 1.045776128768921, "learning_rate": 0.0002, "epoch": 4.9923207501414595, "step": 30880}, {"loss": 0.5767, "grad_norm": 0.9861903786659241, "learning_rate": 0.0002, "epoch": 4.993937434322206, "step": 30890}, {"loss": 0.6192, "grad_norm": 0.9275050759315491, "learning_rate": 0.0002, "epoch": 4.995554118502951, "step": 30900}, {"loss": 0.6181, "grad_norm": 0.94013911485672, "learning_rate": 0.0002, "epoch": 4.997170802683696, "step": 30910}, {"loss": 0.614, "grad_norm": 0.9771268367767334, "learning_rate": 0.0002, "epoch": 4.998787486864441, "step": 30920}, {"eval_loss": 1.1968598365783691, "eval_runtime": 122.2519, "eval_samples_per_second": 5.996, "eval_steps_per_second": 0.753, "epoch": 4.9999191657909625, "step": 30927}, {"loss": 0.5238, "grad_norm": 0.8021580576896667, "learning_rate": 0.0002, "epoch": 5.0004041710451865, "step": 30930}, {"loss": 0.4984, "grad_norm": 1.0807327032089233, "learning_rate": 0.0002, "epoch": 5.002020855225932, "step": 30940}, {"loss": 0.514, "grad_norm": 1.1638425588607788, "learning_rate": 0.0002, "epoch": 5.003637539406677, "step": 30950}, {"loss": 0.4621, "grad_norm": 1.1700230836868286, "learning_rate": 0.0002, "epoch": 5.005254223587422, "step": 30960}, {"loss": 0.4657, "grad_norm": 0.9053420424461365, "learning_rate": 0.0002, "epoch": 5.0068709077681675, "step": 30970}, {"loss": 0.4865, "grad_norm": 0.9226111769676208, "learning_rate": 0.0002, "epoch": 5.008487591948913, "step": 30980}, {"loss": 0.5011, "grad_norm": 1.238669514656067, "learning_rate": 0.0002, "epoch": 5.010104276129658, "step": 30990}, {"loss": 0.4754, "grad_norm": 1.0668327808380127, "learning_rate": 0.0002, "epoch": 5.011720960310403, "step": 31000}, {"loss": 0.5414, "grad_norm": 1.0903944969177246, "learning_rate": 0.0002, "epoch": 5.013337644491148, "step": 31010}, {"loss": 0.5117, "grad_norm": 1.0763911008834839, "learning_rate": 0.0002, "epoch": 5.014954328671894, "step": 31020}, {"loss": 0.4908, "grad_norm": 1.0108771324157715, "learning_rate": 0.0002, "epoch": 5.016571012852639, "step": 31030}, {"loss": 0.5052, "grad_norm": 0.8816103935241699, "learning_rate": 0.0002, "epoch": 5.018187697033385, "step": 31040}, {"loss": 0.4985, "grad_norm": 1.11434805393219, "learning_rate": 0.0002, "epoch": 5.01980438121413, "step": 31050}, {"loss": 0.5074, "grad_norm": 1.0727789402008057, "learning_rate": 0.0002, "epoch": 5.021421065394875, "step": 31060}, {"loss": 0.4938, "grad_norm": 1.1480379104614258, "learning_rate": 0.0002, "epoch": 5.023037749575621, "step": 31070}, {"loss": 0.491, "grad_norm": 1.0913071632385254, "learning_rate": 0.0002, "epoch": 5.024654433756366, "step": 31080}, {"loss": 0.4896, "grad_norm": 0.9891864657402039, "learning_rate": 0.0002, "epoch": 5.026271117937111, "step": 31090}, {"loss": 0.4965, "grad_norm": 0.9167473912239075, "learning_rate": 0.0002, "epoch": 5.027887802117856, "step": 31100}, {"loss": 0.5098, "grad_norm": 1.2259035110473633, "learning_rate": 0.0002, "epoch": 5.029504486298602, "step": 31110}, {"loss": 0.5206, "grad_norm": 1.1812787055969238, "learning_rate": 0.0002, "epoch": 5.031121170479347, "step": 31120}, {"loss": 0.4725, "grad_norm": 1.0890522003173828, "learning_rate": 0.0002, "epoch": 5.032737854660092, "step": 31130}, {"loss": 0.4768, "grad_norm": 1.0521091222763062, "learning_rate": 0.0002, "epoch": 5.034354538840837, "step": 31140}, {"loss": 0.4718, "grad_norm": 1.1274569034576416, "learning_rate": 0.0002, "epoch": 5.0359712230215825, "step": 31150}, {"loss": 0.4604, "grad_norm": 1.140974998474121, "learning_rate": 0.0002, "epoch": 5.037587907202328, "step": 31160}, {"loss": 0.5077, "grad_norm": 1.1215609312057495, "learning_rate": 0.0002, "epoch": 5.039204591383073, "step": 31170}, {"loss": 0.4746, "grad_norm": 1.0107218027114868, "learning_rate": 0.0002, "epoch": 5.040821275563818, "step": 31180}, {"loss": 0.5126, "grad_norm": 1.0198770761489868, "learning_rate": 0.0002, "epoch": 5.042437959744564, "step": 31190}, {"loss": 0.5004, "grad_norm": 1.1613430976867676, "learning_rate": 0.0002, "epoch": 5.0440546439253096, "step": 31200}, {"loss": 0.5181, "grad_norm": 0.8555458188056946, "learning_rate": 0.0002, "epoch": 5.045671328106055, "step": 31210}, {"loss": 0.4878, "grad_norm": 1.0235545635223389, "learning_rate": 0.0002, "epoch": 5.0472880122868, "step": 31220}, {"loss": 0.499, "grad_norm": 1.0228750705718994, "learning_rate": 0.0002, "epoch": 5.048904696467545, "step": 31230}, {"loss": 0.4544, "grad_norm": 0.8216419816017151, "learning_rate": 0.0002, "epoch": 5.0505213806482905, "step": 31240}, {"loss": 0.4947, "grad_norm": 0.925828218460083, "learning_rate": 0.0002, "epoch": 5.052138064829036, "step": 31250}, {"loss": 0.4835, "grad_norm": 0.9229369759559631, "learning_rate": 0.0002, "epoch": 5.053754749009781, "step": 31260}, {"loss": 0.5136, "grad_norm": 0.9531727433204651, "learning_rate": 0.0002, "epoch": 5.055371433190526, "step": 31270}, {"loss": 0.5161, "grad_norm": 0.7738548517227173, "learning_rate": 0.0002, "epoch": 5.056988117371271, "step": 31280}, {"loss": 0.5166, "grad_norm": 1.0551451444625854, "learning_rate": 0.0002, "epoch": 5.058604801552017, "step": 31290}, {"loss": 0.4953, "grad_norm": 0.9782299399375916, "learning_rate": 0.0002, "epoch": 5.060221485732762, "step": 31300}, {"loss": 0.4776, "grad_norm": 1.0220632553100586, "learning_rate": 0.0002, "epoch": 5.061838169913507, "step": 31310}, {"loss": 0.5117, "grad_norm": 0.9808892607688904, "learning_rate": 0.0002, "epoch": 5.063454854094252, "step": 31320}, {"loss": 0.501, "grad_norm": 1.0662003755569458, "learning_rate": 0.0002, "epoch": 5.065071538274998, "step": 31330}, {"loss": 0.4844, "grad_norm": 1.0036940574645996, "learning_rate": 0.0002, "epoch": 5.066688222455744, "step": 31340}, {"loss": 0.5299, "grad_norm": 1.1931052207946777, "learning_rate": 0.0002, "epoch": 5.068304906636489, "step": 31350}, {"loss": 0.4646, "grad_norm": 0.9370693564414978, "learning_rate": 0.0002, "epoch": 5.069921590817234, "step": 31360}, {"loss": 0.5274, "grad_norm": 0.9589039087295532, "learning_rate": 0.0002, "epoch": 5.071538274997979, "step": 31370}, {"loss": 0.4669, "grad_norm": 1.0052711963653564, "learning_rate": 0.0002, "epoch": 5.073154959178725, "step": 31380}, {"loss": 0.5283, "grad_norm": 0.9991368651390076, "learning_rate": 0.0002, "epoch": 5.07477164335947, "step": 31390}, {"loss": 0.4579, "grad_norm": 0.8539695739746094, "learning_rate": 0.0002, "epoch": 5.076388327540215, "step": 31400}, {"loss": 0.4609, "grad_norm": 1.048775553703308, "learning_rate": 0.0002, "epoch": 5.07800501172096, "step": 31410}, {"loss": 0.4915, "grad_norm": 0.9983724355697632, "learning_rate": 0.0002, "epoch": 5.0796216959017055, "step": 31420}, {"loss": 0.4594, "grad_norm": 1.0189813375473022, "learning_rate": 0.0002, "epoch": 5.081238380082451, "step": 31430}, {"loss": 0.5449, "grad_norm": 0.9781646728515625, "learning_rate": 0.0002, "epoch": 5.082855064263196, "step": 31440}, {"loss": 0.4698, "grad_norm": 0.9424566030502319, "learning_rate": 0.0002, "epoch": 5.084471748443941, "step": 31450}, {"loss": 0.4768, "grad_norm": 1.0036484003067017, "learning_rate": 0.0002, "epoch": 5.0860884326246865, "step": 31460}, {"loss": 0.487, "grad_norm": 1.0983147621154785, "learning_rate": 0.0002, "epoch": 5.087705116805432, "step": 31470}, {"loss": 0.5236, "grad_norm": 1.0856730937957764, "learning_rate": 0.0002, "epoch": 5.089321800986177, "step": 31480}, {"loss": 0.485, "grad_norm": 1.2191699743270874, "learning_rate": 0.0002, "epoch": 5.090938485166923, "step": 31490}, {"loss": 0.4936, "grad_norm": 0.939346194267273, "learning_rate": 0.0002, "epoch": 5.092555169347668, "step": 31500}, {"loss": 0.5107, "grad_norm": 0.9730121493339539, "learning_rate": 0.0002, "epoch": 5.0941718535284135, "step": 31510}, {"loss": 0.4973, "grad_norm": 0.923686146736145, "learning_rate": 0.0002, "epoch": 5.095788537709159, "step": 31520}, {"loss": 0.4906, "grad_norm": 1.1734349727630615, "learning_rate": 0.0002, "epoch": 5.097405221889904, "step": 31530}, {"loss": 0.5165, "grad_norm": 1.084509015083313, "learning_rate": 0.0002, "epoch": 5.099021906070649, "step": 31540}, {"loss": 0.5078, "grad_norm": 1.0144678354263306, "learning_rate": 0.0002, "epoch": 5.100638590251394, "step": 31550}, {"loss": 0.4719, "grad_norm": 0.9958019256591797, "learning_rate": 0.0002, "epoch": 5.10225527443214, "step": 31560}, {"loss": 0.4876, "grad_norm": 0.8900736570358276, "learning_rate": 0.0002, "epoch": 5.103871958612885, "step": 31570}, {"loss": 0.463, "grad_norm": 1.0921649932861328, "learning_rate": 0.0002, "epoch": 5.10548864279363, "step": 31580}, {"loss": 0.5148, "grad_norm": 1.1613792181015015, "learning_rate": 0.0002, "epoch": 5.107105326974375, "step": 31590}, {"loss": 0.5055, "grad_norm": 0.9211367964744568, "learning_rate": 0.0002, "epoch": 5.108722011155121, "step": 31600}, {"loss": 0.5364, "grad_norm": 1.3315813541412354, "learning_rate": 0.0002, "epoch": 5.110338695335866, "step": 31610}, {"loss": 0.5336, "grad_norm": 1.3765019178390503, "learning_rate": 0.0002, "epoch": 5.111955379516611, "step": 31620}, {"loss": 0.4861, "grad_norm": 1.070198893547058, "learning_rate": 0.0002, "epoch": 5.113572063697356, "step": 31630}, {"loss": 0.5046, "grad_norm": 0.947631299495697, "learning_rate": 0.0002, "epoch": 5.115188747878102, "step": 31640}, {"loss": 0.5297, "grad_norm": 1.0197371244430542, "learning_rate": 0.0002, "epoch": 5.116805432058848, "step": 31650}, {"loss": 0.5014, "grad_norm": 0.8647911548614502, "learning_rate": 0.0002, "epoch": 5.118422116239593, "step": 31660}, {"loss": 0.4705, "grad_norm": 0.8944075107574463, "learning_rate": 0.0002, "epoch": 5.120038800420338, "step": 31670}, {"loss": 0.5175, "grad_norm": 1.124497652053833, "learning_rate": 0.0002, "epoch": 5.121655484601083, "step": 31680}, {"loss": 0.5109, "grad_norm": 0.893131673336029, "learning_rate": 0.0002, "epoch": 5.123272168781829, "step": 31690}, {"loss": 0.4937, "grad_norm": 1.0122284889221191, "learning_rate": 0.0002, "epoch": 5.124888852962574, "step": 31700}, {"loss": 0.5522, "grad_norm": 0.9493719935417175, "learning_rate": 0.0002, "epoch": 5.126505537143319, "step": 31710}, {"loss": 0.5031, "grad_norm": 0.9700539112091064, "learning_rate": 0.0002, "epoch": 5.128122221324064, "step": 31720}, {"loss": 0.5126, "grad_norm": 1.111677646636963, "learning_rate": 0.0002, "epoch": 5.1297389055048095, "step": 31730}, {"loss": 0.5272, "grad_norm": 0.8204274773597717, "learning_rate": 0.0002, "epoch": 5.131355589685555, "step": 31740}, {"loss": 0.5029, "grad_norm": 1.1029267311096191, "learning_rate": 0.0002, "epoch": 5.1329722738663, "step": 31750}, {"loss": 0.505, "grad_norm": 1.065575122833252, "learning_rate": 0.0002, "epoch": 5.134588958047045, "step": 31760}, {"loss": 0.502, "grad_norm": 0.8208706974983215, "learning_rate": 0.0002, "epoch": 5.13620564222779, "step": 31770}, {"loss": 0.5352, "grad_norm": 1.0520979166030884, "learning_rate": 0.0002, "epoch": 5.137822326408536, "step": 31780}, {"loss": 0.4911, "grad_norm": 0.8585538268089294, "learning_rate": 0.0002, "epoch": 5.139439010589282, "step": 31790}, {"loss": 0.5159, "grad_norm": 1.1491447687149048, "learning_rate": 0.0002, "epoch": 5.141055694770027, "step": 31800}, {"loss": 0.5157, "grad_norm": 0.9441081285476685, "learning_rate": 0.0002, "epoch": 5.142672378950772, "step": 31810}, {"loss": 0.5383, "grad_norm": 1.4146889448165894, "learning_rate": 0.0002, "epoch": 5.1442890631315175, "step": 31820}, {"loss": 0.5159, "grad_norm": 1.0326547622680664, "learning_rate": 0.0002, "epoch": 5.145905747312263, "step": 31830}, {"loss": 0.5348, "grad_norm": 0.9879202842712402, "learning_rate": 0.0002, "epoch": 5.147522431493008, "step": 31840}, {"loss": 0.5083, "grad_norm": 1.0374281406402588, "learning_rate": 0.0002, "epoch": 5.149139115673753, "step": 31850}, {"loss": 0.4827, "grad_norm": 1.181229591369629, "learning_rate": 0.0002, "epoch": 5.150755799854498, "step": 31860}, {"loss": 0.5313, "grad_norm": 1.2078537940979004, "learning_rate": 0.0002, "epoch": 5.152372484035244, "step": 31870}, {"loss": 0.5329, "grad_norm": 0.9599190354347229, "learning_rate": 0.0002, "epoch": 5.153989168215989, "step": 31880}, {"loss": 0.4953, "grad_norm": 1.0378568172454834, "learning_rate": 0.0002, "epoch": 5.155605852396734, "step": 31890}, {"loss": 0.5069, "grad_norm": 0.8746536374092102, "learning_rate": 0.0002, "epoch": 5.157222536577479, "step": 31900}, {"loss": 0.5272, "grad_norm": 1.0232136249542236, "learning_rate": 0.0002, "epoch": 5.1588392207582245, "step": 31910}, {"loss": 0.4844, "grad_norm": 0.9827565550804138, "learning_rate": 0.0002, "epoch": 5.16045590493897, "step": 31920}, {"loss": 0.5029, "grad_norm": 1.342657208442688, "learning_rate": 0.0002, "epoch": 5.162072589119716, "step": 31930}, {"loss": 0.513, "grad_norm": 1.18390691280365, "learning_rate": 0.0002, "epoch": 5.163689273300461, "step": 31940}, {"loss": 0.5267, "grad_norm": 0.996350109577179, "learning_rate": 0.0002, "epoch": 5.165305957481206, "step": 31950}, {"loss": 0.5063, "grad_norm": 0.9710391163825989, "learning_rate": 0.0002, "epoch": 5.166922641661952, "step": 31960}, {"loss": 0.5115, "grad_norm": 1.0264002084732056, "learning_rate": 0.0002, "epoch": 5.168539325842697, "step": 31970}, {"loss": 0.4972, "grad_norm": 1.0028311014175415, "learning_rate": 0.0002, "epoch": 5.170156010023442, "step": 31980}, {"loss": 0.5103, "grad_norm": 1.1078234910964966, "learning_rate": 0.0002, "epoch": 5.171772694204187, "step": 31990}, {"loss": 0.495, "grad_norm": 0.9659610390663147, "learning_rate": 0.0002, "epoch": 5.1733893783849325, "step": 32000}, {"loss": 0.5114, "grad_norm": 0.841986894607544, "learning_rate": 0.0002, "epoch": 5.175006062565678, "step": 32010}, {"loss": 0.48, "grad_norm": 1.095332384109497, "learning_rate": 0.0002, "epoch": 5.176622746746423, "step": 32020}, {"loss": 0.4741, "grad_norm": 1.1242377758026123, "learning_rate": 0.0002, "epoch": 5.178239430927168, "step": 32030}, {"loss": 0.5573, "grad_norm": 0.9872292280197144, "learning_rate": 0.0002, "epoch": 5.179856115107913, "step": 32040}, {"loss": 0.48, "grad_norm": 0.936161994934082, "learning_rate": 0.0002, "epoch": 5.181472799288659, "step": 32050}, {"loss": 0.5093, "grad_norm": 1.166100025177002, "learning_rate": 0.0002, "epoch": 5.183089483469404, "step": 32060}, {"loss": 0.5438, "grad_norm": 1.0764425992965698, "learning_rate": 0.0002, "epoch": 5.184706167650149, "step": 32070}, {"loss": 0.4843, "grad_norm": 1.0480051040649414, "learning_rate": 0.0002, "epoch": 5.186322851830895, "step": 32080}, {"loss": 0.5386, "grad_norm": 1.0874916315078735, "learning_rate": 0.0002, "epoch": 5.1879395360116405, "step": 32090}, {"loss": 0.4975, "grad_norm": 1.0817396640777588, "learning_rate": 0.0002, "epoch": 5.189556220192386, "step": 32100}, {"loss": 0.5177, "grad_norm": 1.054111361503601, "learning_rate": 0.0002, "epoch": 5.191172904373131, "step": 32110}, {"loss": 0.5229, "grad_norm": 0.9655823707580566, "learning_rate": 0.0002, "epoch": 5.192789588553876, "step": 32120}, {"loss": 0.5105, "grad_norm": 1.1384109258651733, "learning_rate": 0.0002, "epoch": 5.194406272734621, "step": 32130}, {"loss": 0.5073, "grad_norm": 1.0149348974227905, "learning_rate": 0.0002, "epoch": 5.196022956915367, "step": 32140}, {"loss": 0.5293, "grad_norm": 1.1084046363830566, "learning_rate": 0.0002, "epoch": 5.197639641096112, "step": 32150}, {"loss": 0.4936, "grad_norm": 1.1209309101104736, "learning_rate": 0.0002, "epoch": 5.199256325276857, "step": 32160}, {"loss": 0.5101, "grad_norm": 1.133089542388916, "learning_rate": 0.0002, "epoch": 5.200873009457602, "step": 32170}, {"loss": 0.5242, "grad_norm": 1.0893020629882812, "learning_rate": 0.0002, "epoch": 5.202489693638348, "step": 32180}, {"loss": 0.4872, "grad_norm": 0.90018630027771, "learning_rate": 0.0002, "epoch": 5.204106377819093, "step": 32190}, {"loss": 0.4999, "grad_norm": 0.977622926235199, "learning_rate": 0.0002, "epoch": 5.205723061999838, "step": 32200}, {"loss": 0.5028, "grad_norm": 1.2940177917480469, "learning_rate": 0.0002, "epoch": 5.207339746180583, "step": 32210}, {"loss": 0.5396, "grad_norm": 1.2131710052490234, "learning_rate": 0.0002, "epoch": 5.2089564303613285, "step": 32220}, {"loss": 0.5189, "grad_norm": 1.0234841108322144, "learning_rate": 0.0002, "epoch": 5.210573114542075, "step": 32230}, {"loss": 0.5424, "grad_norm": 1.157975435256958, "learning_rate": 0.0002, "epoch": 5.21218979872282, "step": 32240}, {"loss": 0.5396, "grad_norm": 1.0381282567977905, "learning_rate": 0.0002, "epoch": 5.213806482903565, "step": 32250}, {"loss": 0.5192, "grad_norm": 1.0125395059585571, "learning_rate": 0.0002, "epoch": 5.21542316708431, "step": 32260}, {"loss": 0.5216, "grad_norm": 1.272691011428833, "learning_rate": 0.0002, "epoch": 5.2170398512650555, "step": 32270}, {"loss": 0.52, "grad_norm": 1.0061250925064087, "learning_rate": 0.0002, "epoch": 5.218656535445801, "step": 32280}, {"loss": 0.4739, "grad_norm": 0.9752234816551208, "learning_rate": 0.0002, "epoch": 5.220273219626546, "step": 32290}, {"loss": 0.5471, "grad_norm": 1.1193140745162964, "learning_rate": 0.0002, "epoch": 5.221889903807291, "step": 32300}, {"loss": 0.4976, "grad_norm": 1.0126434564590454, "learning_rate": 0.0002, "epoch": 5.2235065879880365, "step": 32310}, {"loss": 0.5257, "grad_norm": 1.4338394403457642, "learning_rate": 0.0002, "epoch": 5.225123272168782, "step": 32320}, {"loss": 0.5235, "grad_norm": 1.004101276397705, "learning_rate": 0.0002, "epoch": 5.226739956349527, "step": 32330}, {"loss": 0.5091, "grad_norm": 0.8744166493415833, "learning_rate": 0.0002, "epoch": 5.228356640530272, "step": 32340}, {"loss": 0.5388, "grad_norm": 1.0165376663208008, "learning_rate": 0.0002, "epoch": 5.229973324711017, "step": 32350}, {"loss": 0.5469, "grad_norm": 0.8635954260826111, "learning_rate": 0.0002, "epoch": 5.231590008891763, "step": 32360}, {"loss": 0.5609, "grad_norm": 1.1392399072647095, "learning_rate": 0.0002, "epoch": 5.233206693072509, "step": 32370}, {"loss": 0.5173, "grad_norm": 1.0202113389968872, "learning_rate": 0.0002, "epoch": 5.234823377253254, "step": 32380}, {"loss": 0.4983, "grad_norm": 1.0417983531951904, "learning_rate": 0.0002, "epoch": 5.236440061433999, "step": 32390}, {"loss": 0.507, "grad_norm": 0.8729333877563477, "learning_rate": 0.0002, "epoch": 5.238056745614744, "step": 32400}, {"loss": 0.5426, "grad_norm": 1.1626229286193848, "learning_rate": 0.0002, "epoch": 5.23967342979549, "step": 32410}, {"loss": 0.5355, "grad_norm": 0.9086161851882935, "learning_rate": 0.0002, "epoch": 5.241290113976235, "step": 32420}, {"loss": 0.4927, "grad_norm": 1.3999892473220825, "learning_rate": 0.0002, "epoch": 5.24290679815698, "step": 32430}, {"loss": 0.4795, "grad_norm": 1.0356311798095703, "learning_rate": 0.0002, "epoch": 5.244523482337725, "step": 32440}, {"loss": 0.5035, "grad_norm": 0.9655531644821167, "learning_rate": 0.0002, "epoch": 5.246140166518471, "step": 32450}, {"loss": 0.5166, "grad_norm": 1.0411828756332397, "learning_rate": 0.0002, "epoch": 5.247756850699216, "step": 32460}, {"loss": 0.5141, "grad_norm": 1.1199816465377808, "learning_rate": 0.0002, "epoch": 5.249373534879961, "step": 32470}, {"loss": 0.4864, "grad_norm": 1.260321855545044, "learning_rate": 0.0002, "epoch": 5.250990219060706, "step": 32480}, {"loss": 0.4893, "grad_norm": 1.2950857877731323, "learning_rate": 0.0002, "epoch": 5.2526069032414515, "step": 32490}, {"loss": 0.4952, "grad_norm": 0.8982820510864258, "learning_rate": 0.0002, "epoch": 5.254223587422197, "step": 32500}, {"loss": 0.5138, "grad_norm": 0.8512987494468689, "learning_rate": 0.0002, "epoch": 5.255840271602942, "step": 32510}, {"loss": 0.5341, "grad_norm": 1.067443609237671, "learning_rate": 0.0002, "epoch": 5.257456955783688, "step": 32520}, {"loss": 0.4928, "grad_norm": 1.0957417488098145, "learning_rate": 0.0002, "epoch": 5.259073639964433, "step": 32530}, {"loss": 0.5169, "grad_norm": 1.4161807298660278, "learning_rate": 0.0002, "epoch": 5.260690324145179, "step": 32540}, {"loss": 0.5599, "grad_norm": 1.2264093160629272, "learning_rate": 0.0002, "epoch": 5.262307008325924, "step": 32550}, {"loss": 0.5221, "grad_norm": 1.0015931129455566, "learning_rate": 0.0002, "epoch": 5.263923692506669, "step": 32560}, {"loss": 0.5253, "grad_norm": 1.0743094682693481, "learning_rate": 0.0002, "epoch": 5.265540376687414, "step": 32570}, {"loss": 0.5289, "grad_norm": 1.1386840343475342, "learning_rate": 0.0002, "epoch": 5.2671570608681595, "step": 32580}, {"loss": 0.5315, "grad_norm": 1.0093860626220703, "learning_rate": 0.0002, "epoch": 5.268773745048905, "step": 32590}, {"loss": 0.5175, "grad_norm": 0.9593744874000549, "learning_rate": 0.0002, "epoch": 5.27039042922965, "step": 32600}, {"loss": 0.528, "grad_norm": 1.146021842956543, "learning_rate": 0.0002, "epoch": 5.272007113410395, "step": 32610}, {"loss": 0.4983, "grad_norm": 0.9579031467437744, "learning_rate": 0.0002, "epoch": 5.27362379759114, "step": 32620}, {"loss": 0.5376, "grad_norm": 1.0548793077468872, "learning_rate": 0.0002, "epoch": 5.275240481771886, "step": 32630}, {"loss": 0.5267, "grad_norm": 1.0380561351776123, "learning_rate": 0.0002, "epoch": 5.276857165952631, "step": 32640}, {"loss": 0.5182, "grad_norm": 1.2119969129562378, "learning_rate": 0.0002, "epoch": 5.278473850133376, "step": 32650}, {"loss": 0.5298, "grad_norm": 1.0507797002792358, "learning_rate": 0.0002, "epoch": 5.280090534314121, "step": 32660}, {"loss": 0.5253, "grad_norm": 1.0185176134109497, "learning_rate": 0.0002, "epoch": 5.2817072184948675, "step": 32670}, {"loss": 0.4904, "grad_norm": 1.2358098030090332, "learning_rate": 0.0002, "epoch": 5.283323902675613, "step": 32680}, {"loss": 0.5169, "grad_norm": 0.7937114238739014, "learning_rate": 0.0002, "epoch": 5.284940586856358, "step": 32690}, {"loss": 0.495, "grad_norm": 0.9825124740600586, "learning_rate": 0.0002, "epoch": 5.286557271037103, "step": 32700}, {"loss": 0.5149, "grad_norm": 1.2059301137924194, "learning_rate": 0.0002, "epoch": 5.288173955217848, "step": 32710}, {"loss": 0.5272, "grad_norm": 1.0828571319580078, "learning_rate": 0.0002, "epoch": 5.289790639398594, "step": 32720}, {"loss": 0.5383, "grad_norm": 1.0129735469818115, "learning_rate": 0.0002, "epoch": 5.291407323579339, "step": 32730}, {"loss": 0.5216, "grad_norm": 1.0591634511947632, "learning_rate": 0.0002, "epoch": 5.293024007760084, "step": 32740}, {"loss": 0.522, "grad_norm": 0.9256815910339355, "learning_rate": 0.0002, "epoch": 5.294640691940829, "step": 32750}, {"loss": 0.5396, "grad_norm": 1.0928633213043213, "learning_rate": 0.0002, "epoch": 5.2962573761215745, "step": 32760}, {"loss": 0.5093, "grad_norm": 0.9415594935417175, "learning_rate": 0.0002, "epoch": 5.29787406030232, "step": 32770}, {"loss": 0.5252, "grad_norm": 1.141316294670105, "learning_rate": 0.0002, "epoch": 5.299490744483065, "step": 32780}, {"loss": 0.4837, "grad_norm": 1.0646510124206543, "learning_rate": 0.0002, "epoch": 5.30110742866381, "step": 32790}, {"loss": 0.5547, "grad_norm": 1.189661979675293, "learning_rate": 0.0002, "epoch": 5.3027241128445555, "step": 32800}, {"loss": 0.5664, "grad_norm": 0.9568731188774109, "learning_rate": 0.0002, "epoch": 5.304340797025301, "step": 32810}, {"loss": 0.5344, "grad_norm": 1.1556824445724487, "learning_rate": 0.0002, "epoch": 5.305957481206047, "step": 32820}, {"loss": 0.4894, "grad_norm": 0.9353463649749756, "learning_rate": 0.0002, "epoch": 5.307574165386792, "step": 32830}, {"loss": 0.5052, "grad_norm": 1.1208295822143555, "learning_rate": 0.0002, "epoch": 5.309190849567537, "step": 32840}, {"loss": 0.5126, "grad_norm": 1.0894153118133545, "learning_rate": 0.0002, "epoch": 5.3108075337482825, "step": 32850}, {"loss": 0.5046, "grad_norm": 1.090329647064209, "learning_rate": 0.0002, "epoch": 5.312424217929028, "step": 32860}, {"loss": 0.5237, "grad_norm": 1.0781712532043457, "learning_rate": 0.0002, "epoch": 5.314040902109773, "step": 32870}, {"loss": 0.57, "grad_norm": 1.1785295009613037, "learning_rate": 0.0002, "epoch": 5.315657586290518, "step": 32880}, {"loss": 0.4953, "grad_norm": 1.0406851768493652, "learning_rate": 0.0002, "epoch": 5.317274270471263, "step": 32890}, {"loss": 0.514, "grad_norm": 1.0982953310012817, "learning_rate": 0.0002, "epoch": 5.318890954652009, "step": 32900}, {"loss": 0.4944, "grad_norm": 1.2969383001327515, "learning_rate": 0.0002, "epoch": 5.320507638832754, "step": 32910}, {"loss": 0.4786, "grad_norm": 0.9687288999557495, "learning_rate": 0.0002, "epoch": 5.322124323013499, "step": 32920}, {"loss": 0.5286, "grad_norm": 1.136760950088501, "learning_rate": 0.0002, "epoch": 5.323741007194244, "step": 32930}, {"loss": 0.5321, "grad_norm": 1.3045495748519897, "learning_rate": 0.0002, "epoch": 5.32535769137499, "step": 32940}, {"loss": 0.5413, "grad_norm": 1.221675992012024, "learning_rate": 0.0002, "epoch": 5.326974375555735, "step": 32950}, {"loss": 0.4999, "grad_norm": 1.1380633115768433, "learning_rate": 0.0002, "epoch": 5.32859105973648, "step": 32960}, {"loss": 0.5037, "grad_norm": 1.1065956354141235, "learning_rate": 0.0002, "epoch": 5.330207743917226, "step": 32970}, {"loss": 0.4913, "grad_norm": 1.0187175273895264, "learning_rate": 0.0002, "epoch": 5.331824428097971, "step": 32980}, {"loss": 0.5234, "grad_norm": 0.9077118039131165, "learning_rate": 0.0002, "epoch": 5.333441112278717, "step": 32990}, {"loss": 0.5071, "grad_norm": 1.0092815160751343, "learning_rate": 0.0002, "epoch": 5.335057796459462, "step": 33000}, {"loss": 0.498, "grad_norm": 1.0168777704238892, "learning_rate": 0.0002, "epoch": 5.336674480640207, "step": 33010}, {"loss": 0.4952, "grad_norm": 0.996161937713623, "learning_rate": 0.0002, "epoch": 5.338291164820952, "step": 33020}, {"loss": 0.5024, "grad_norm": 0.794463038444519, "learning_rate": 0.0002, "epoch": 5.339907849001698, "step": 33030}, {"loss": 0.5112, "grad_norm": 0.9750674962997437, "learning_rate": 0.0002, "epoch": 5.341524533182443, "step": 33040}, {"loss": 0.528, "grad_norm": 1.2770029306411743, "learning_rate": 0.0002, "epoch": 5.343141217363188, "step": 33050}, {"loss": 0.52, "grad_norm": 1.1500186920166016, "learning_rate": 0.0002, "epoch": 5.344757901543933, "step": 33060}, {"loss": 0.4906, "grad_norm": 1.0726377964019775, "learning_rate": 0.0002, "epoch": 5.3463745857246785, "step": 33070}, {"loss": 0.5212, "grad_norm": 0.9314153790473938, "learning_rate": 0.0002, "epoch": 5.347991269905424, "step": 33080}, {"loss": 0.5434, "grad_norm": 1.344988465309143, "learning_rate": 0.0002, "epoch": 5.349607954086169, "step": 33090}, {"loss": 0.4874, "grad_norm": 0.863196611404419, "learning_rate": 0.0002, "epoch": 5.351224638266914, "step": 33100}, {"loss": 0.534, "grad_norm": 1.128100037574768, "learning_rate": 0.0002, "epoch": 5.352841322447659, "step": 33110}, {"loss": 0.5293, "grad_norm": 1.1673583984375, "learning_rate": 0.0002, "epoch": 5.3544580066284055, "step": 33120}, {"loss": 0.4787, "grad_norm": 0.9416789412498474, "learning_rate": 0.0002, "epoch": 5.356074690809151, "step": 33130}, {"loss": 0.5155, "grad_norm": 1.1855236291885376, "learning_rate": 0.0002, "epoch": 5.357691374989896, "step": 33140}, {"loss": 0.515, "grad_norm": 1.0415170192718506, "learning_rate": 0.0002, "epoch": 5.359308059170641, "step": 33150}, {"loss": 0.545, "grad_norm": 0.9953004121780396, "learning_rate": 0.0002, "epoch": 5.3609247433513865, "step": 33160}, {"loss": 0.5305, "grad_norm": 0.96138596534729, "learning_rate": 0.0002, "epoch": 5.362541427532132, "step": 33170}, {"loss": 0.5064, "grad_norm": 1.341979742050171, "learning_rate": 0.0002, "epoch": 5.364158111712877, "step": 33180}, {"loss": 0.4986, "grad_norm": 1.0136911869049072, "learning_rate": 0.0002, "epoch": 5.365774795893622, "step": 33190}, {"loss": 0.5459, "grad_norm": 0.8685575127601624, "learning_rate": 0.0002, "epoch": 5.367391480074367, "step": 33200}, {"loss": 0.5146, "grad_norm": 0.8833574652671814, "learning_rate": 0.0002, "epoch": 5.369008164255113, "step": 33210}, {"loss": 0.4982, "grad_norm": 0.9123612642288208, "learning_rate": 0.0002, "epoch": 5.370624848435858, "step": 33220}, {"loss": 0.5047, "grad_norm": 1.2720599174499512, "learning_rate": 0.0002, "epoch": 5.372241532616603, "step": 33230}, {"loss": 0.5175, "grad_norm": 1.0596648454666138, "learning_rate": 0.0002, "epoch": 5.373858216797348, "step": 33240}, {"loss": 0.5284, "grad_norm": 1.119701623916626, "learning_rate": 0.0002, "epoch": 5.3754749009780936, "step": 33250}, {"loss": 0.5217, "grad_norm": 1.3000061511993408, "learning_rate": 0.0002, "epoch": 5.377091585158839, "step": 33260}, {"loss": 0.5125, "grad_norm": 1.083891749382019, "learning_rate": 0.0002, "epoch": 5.378708269339585, "step": 33270}, {"loss": 0.5065, "grad_norm": 0.9402718544006348, "learning_rate": 0.0002, "epoch": 5.38032495352033, "step": 33280}, {"loss": 0.5559, "grad_norm": 1.3376892805099487, "learning_rate": 0.0002, "epoch": 5.381941637701075, "step": 33290}, {"loss": 0.5193, "grad_norm": 1.1600074768066406, "learning_rate": 0.0002, "epoch": 5.383558321881821, "step": 33300}, {"loss": 0.4907, "grad_norm": 1.1449427604675293, "learning_rate": 0.0002, "epoch": 5.385175006062566, "step": 33310}, {"loss": 0.5449, "grad_norm": 1.3118891716003418, "learning_rate": 0.0002, "epoch": 5.386791690243311, "step": 33320}, {"loss": 0.547, "grad_norm": 0.743449866771698, "learning_rate": 0.0002, "epoch": 5.388408374424056, "step": 33330}, {"loss": 0.5555, "grad_norm": 0.9358304142951965, "learning_rate": 0.0002, "epoch": 5.3900250586048015, "step": 33340}, {"loss": 0.5558, "grad_norm": 1.0447142124176025, "learning_rate": 0.0002, "epoch": 5.391641742785547, "step": 33350}, {"loss": 0.5106, "grad_norm": 1.1088626384735107, "learning_rate": 0.0002, "epoch": 5.393258426966292, "step": 33360}, {"loss": 0.4929, "grad_norm": 1.1267958879470825, "learning_rate": 0.0002, "epoch": 5.394875111147037, "step": 33370}, {"loss": 0.5165, "grad_norm": 0.9709370136260986, "learning_rate": 0.0002, "epoch": 5.3964917953277824, "step": 33380}, {"loss": 0.5206, "grad_norm": 1.0939103364944458, "learning_rate": 0.0002, "epoch": 5.398108479508528, "step": 33390}, {"loss": 0.5177, "grad_norm": 0.9559304714202881, "learning_rate": 0.0002, "epoch": 5.399725163689273, "step": 33400}, {"loss": 0.5064, "grad_norm": 1.199580430984497, "learning_rate": 0.0002, "epoch": 5.401341847870018, "step": 33410}, {"loss": 0.52, "grad_norm": 0.9097000360488892, "learning_rate": 0.0002, "epoch": 5.402958532050764, "step": 33420}, {"loss": 0.514, "grad_norm": 1.1940981149673462, "learning_rate": 0.0002, "epoch": 5.4045752162315095, "step": 33430}, {"loss": 0.5069, "grad_norm": 1.0530916452407837, "learning_rate": 0.0002, "epoch": 5.406191900412255, "step": 33440}, {"loss": 0.5482, "grad_norm": 1.0482549667358398, "learning_rate": 0.0002, "epoch": 5.407808584593, "step": 33450}, {"loss": 0.501, "grad_norm": 1.2524714469909668, "learning_rate": 0.0002, "epoch": 5.409425268773745, "step": 33460}, {"loss": 0.5597, "grad_norm": 1.1091666221618652, "learning_rate": 0.0002, "epoch": 5.41104195295449, "step": 33470}, {"loss": 0.546, "grad_norm": 0.9981587529182434, "learning_rate": 0.0002, "epoch": 5.412658637135236, "step": 33480}, {"loss": 0.4977, "grad_norm": 1.016681432723999, "learning_rate": 0.0002, "epoch": 5.414275321315981, "step": 33490}, {"loss": 0.5388, "grad_norm": 1.1456854343414307, "learning_rate": 0.0002, "epoch": 5.415892005496726, "step": 33500}, {"loss": 0.5292, "grad_norm": 1.1454259157180786, "learning_rate": 0.0002, "epoch": 5.417508689677471, "step": 33510}, {"loss": 0.5061, "grad_norm": 0.9858416318893433, "learning_rate": 0.0002, "epoch": 5.419125373858217, "step": 33520}, {"loss": 0.5139, "grad_norm": 0.9764766693115234, "learning_rate": 0.0002, "epoch": 5.420742058038962, "step": 33530}, {"loss": 0.5518, "grad_norm": 1.199920892715454, "learning_rate": 0.0002, "epoch": 5.422358742219707, "step": 33540}, {"loss": 0.5182, "grad_norm": 1.3107370138168335, "learning_rate": 0.0002, "epoch": 5.423975426400452, "step": 33550}, {"loss": 0.5149, "grad_norm": 0.9637970328330994, "learning_rate": 0.0002, "epoch": 5.4255921105811975, "step": 33560}, {"loss": 0.526, "grad_norm": 1.023359775543213, "learning_rate": 0.0002, "epoch": 5.427208794761944, "step": 33570}, {"loss": 0.5206, "grad_norm": 1.060417652130127, "learning_rate": 0.0002, "epoch": 5.428825478942689, "step": 33580}, {"loss": 0.5052, "grad_norm": 0.9971120953559875, "learning_rate": 0.0002, "epoch": 5.430442163123434, "step": 33590}, {"loss": 0.5044, "grad_norm": 0.9213743209838867, "learning_rate": 0.0002, "epoch": 5.432058847304179, "step": 33600}, {"loss": 0.5714, "grad_norm": 1.1512309312820435, "learning_rate": 0.0002, "epoch": 5.4336755314849245, "step": 33610}, {"loss": 0.5317, "grad_norm": 1.2198847532272339, "learning_rate": 0.0002, "epoch": 5.43529221566567, "step": 33620}, {"loss": 0.5237, "grad_norm": 1.0329595804214478, "learning_rate": 0.0002, "epoch": 5.436908899846415, "step": 33630}, {"loss": 0.5364, "grad_norm": 1.1075750589370728, "learning_rate": 0.0002, "epoch": 5.43852558402716, "step": 33640}, {"loss": 0.5295, "grad_norm": 1.006342887878418, "learning_rate": 0.0002, "epoch": 5.4401422682079055, "step": 33650}, {"loss": 0.5394, "grad_norm": 0.9179885983467102, "learning_rate": 0.0002, "epoch": 5.441758952388651, "step": 33660}, {"loss": 0.5124, "grad_norm": 1.2799493074417114, "learning_rate": 0.0002, "epoch": 5.443375636569396, "step": 33670}, {"loss": 0.5426, "grad_norm": 1.1153863668441772, "learning_rate": 0.0002, "epoch": 5.444992320750141, "step": 33680}, {"loss": 0.5087, "grad_norm": 1.0681028366088867, "learning_rate": 0.0002, "epoch": 5.446609004930886, "step": 33690}, {"loss": 0.5272, "grad_norm": 0.9788817167282104, "learning_rate": 0.0002, "epoch": 5.448225689111632, "step": 33700}, {"loss": 0.5308, "grad_norm": 0.8481608629226685, "learning_rate": 0.0002, "epoch": 5.449842373292377, "step": 33710}, {"loss": 0.5225, "grad_norm": 1.113756537437439, "learning_rate": 0.0002, "epoch": 5.451459057473123, "step": 33720}, {"loss": 0.5213, "grad_norm": 0.8425475358963013, "learning_rate": 0.0002, "epoch": 5.453075741653868, "step": 33730}, {"loss": 0.571, "grad_norm": 1.0852208137512207, "learning_rate": 0.0002, "epoch": 5.4546924258346134, "step": 33740}, {"loss": 0.5535, "grad_norm": 1.1664748191833496, "learning_rate": 0.0002, "epoch": 5.456309110015359, "step": 33750}, {"loss": 0.5419, "grad_norm": 1.217241644859314, "learning_rate": 0.0002, "epoch": 5.457925794196104, "step": 33760}, {"loss": 0.5351, "grad_norm": 1.1572928428649902, "learning_rate": 0.0002, "epoch": 5.459542478376849, "step": 33770}, {"loss": 0.5161, "grad_norm": 1.0437318086624146, "learning_rate": 0.0002, "epoch": 5.461159162557594, "step": 33780}, {"loss": 0.5266, "grad_norm": 0.9807571768760681, "learning_rate": 0.0002, "epoch": 5.46277584673834, "step": 33790}, {"loss": 0.5384, "grad_norm": 1.1436342000961304, "learning_rate": 0.0002, "epoch": 5.464392530919085, "step": 33800}, {"loss": 0.5338, "grad_norm": 1.1004794836044312, "learning_rate": 0.0002, "epoch": 5.46600921509983, "step": 33810}, {"loss": 0.4868, "grad_norm": 1.2130268812179565, "learning_rate": 0.0002, "epoch": 5.467625899280575, "step": 33820}, {"loss": 0.516, "grad_norm": 1.3154419660568237, "learning_rate": 0.0002, "epoch": 5.4692425834613205, "step": 33830}, {"loss": 0.4934, "grad_norm": 0.7934383749961853, "learning_rate": 0.0002, "epoch": 5.470859267642066, "step": 33840}, {"loss": 0.5133, "grad_norm": 0.7838410139083862, "learning_rate": 0.0002, "epoch": 5.472475951822812, "step": 33850}, {"loss": 0.4926, "grad_norm": 1.0415139198303223, "learning_rate": 0.0002, "epoch": 5.474092636003557, "step": 33860}, {"loss": 0.5323, "grad_norm": 0.9213164448738098, "learning_rate": 0.0002, "epoch": 5.475709320184302, "step": 33870}, {"loss": 0.5125, "grad_norm": 1.0364776849746704, "learning_rate": 0.0002, "epoch": 5.477326004365048, "step": 33880}, {"loss": 0.5212, "grad_norm": 0.9994072318077087, "learning_rate": 0.0002, "epoch": 5.478942688545793, "step": 33890}, {"loss": 0.5396, "grad_norm": 1.196730136871338, "learning_rate": 0.0002, "epoch": 5.480559372726538, "step": 33900}, {"loss": 0.538, "grad_norm": 0.9955780506134033, "learning_rate": 0.0002, "epoch": 5.482176056907283, "step": 33910}, {"loss": 0.5307, "grad_norm": 1.168188214302063, "learning_rate": 0.0002, "epoch": 5.4837927410880285, "step": 33920}, {"loss": 0.5548, "grad_norm": 1.1816450357437134, "learning_rate": 0.0002, "epoch": 5.485409425268774, "step": 33930}, {"loss": 0.5535, "grad_norm": 1.079715609550476, "learning_rate": 0.0002, "epoch": 5.487026109449519, "step": 33940}, {"loss": 0.5262, "grad_norm": 1.153850793838501, "learning_rate": 0.0002, "epoch": 5.488642793630264, "step": 33950}, {"loss": 0.5248, "grad_norm": 1.0207297801971436, "learning_rate": 0.0002, "epoch": 5.490259477811009, "step": 33960}, {"loss": 0.5142, "grad_norm": 1.1290855407714844, "learning_rate": 0.0002, "epoch": 5.491876161991755, "step": 33970}, {"loss": 0.5168, "grad_norm": 1.068058967590332, "learning_rate": 0.0002, "epoch": 5.4934928461725, "step": 33980}, {"loss": 0.5317, "grad_norm": 0.9789979457855225, "learning_rate": 0.0002, "epoch": 5.495109530353245, "step": 33990}, {"loss": 0.5113, "grad_norm": 0.9696692824363708, "learning_rate": 0.0002, "epoch": 5.496726214533991, "step": 34000}, {"loss": 0.5413, "grad_norm": 1.0539981126785278, "learning_rate": 0.0002, "epoch": 5.4983428987147365, "step": 34010}, {"loss": 0.5783, "grad_norm": 1.0249929428100586, "learning_rate": 0.0002, "epoch": 5.499959582895482, "step": 34020}, {"loss": 0.4888, "grad_norm": 0.9577504992485046, "learning_rate": 0.0002, "epoch": 5.501576267076227, "step": 34030}, {"loss": 0.5291, "grad_norm": 1.0963513851165771, "learning_rate": 0.0002, "epoch": 5.503192951256972, "step": 34040}, {"loss": 0.5315, "grad_norm": 0.8339345455169678, "learning_rate": 0.0002, "epoch": 5.504809635437717, "step": 34050}, {"loss": 0.5191, "grad_norm": 1.0138782262802124, "learning_rate": 0.0002, "epoch": 5.506426319618463, "step": 34060}, {"loss": 0.5463, "grad_norm": 1.0180109739303589, "learning_rate": 0.0002, "epoch": 5.508043003799208, "step": 34070}, {"loss": 0.5083, "grad_norm": 1.2790818214416504, "learning_rate": 0.0002, "epoch": 5.509659687979953, "step": 34080}, {"loss": 0.5195, "grad_norm": 1.428247332572937, "learning_rate": 0.0002, "epoch": 5.511276372160698, "step": 34090}, {"loss": 0.5291, "grad_norm": 1.0926059484481812, "learning_rate": 0.0002, "epoch": 5.5128930563414436, "step": 34100}, {"loss": 0.5665, "grad_norm": 1.2353343963623047, "learning_rate": 0.0002, "epoch": 5.514509740522189, "step": 34110}, {"loss": 0.5331, "grad_norm": 0.935587465763092, "learning_rate": 0.0002, "epoch": 5.516126424702934, "step": 34120}, {"loss": 0.5512, "grad_norm": 0.9767586588859558, "learning_rate": 0.0002, "epoch": 5.517743108883679, "step": 34130}, {"loss": 0.5315, "grad_norm": 1.1660610437393188, "learning_rate": 0.0002, "epoch": 5.5193597930644245, "step": 34140}, {"loss": 0.52, "grad_norm": 0.9828870892524719, "learning_rate": 0.0002, "epoch": 5.520976477245171, "step": 34150}, {"loss": 0.5198, "grad_norm": 1.0097278356552124, "learning_rate": 0.0002, "epoch": 5.522593161425916, "step": 34160}, {"loss": 0.5293, "grad_norm": 1.1766167879104614, "learning_rate": 0.0002, "epoch": 5.524209845606661, "step": 34170}, {"loss": 0.5258, "grad_norm": 0.982292115688324, "learning_rate": 0.0002, "epoch": 5.525826529787406, "step": 34180}, {"loss": 0.5114, "grad_norm": 1.0744609832763672, "learning_rate": 0.0002, "epoch": 5.5274432139681515, "step": 34190}, {"loss": 0.5469, "grad_norm": 1.3831160068511963, "learning_rate": 0.0002, "epoch": 5.529059898148897, "step": 34200}, {"loss": 0.5819, "grad_norm": 1.074771761894226, "learning_rate": 0.0002, "epoch": 5.530676582329642, "step": 34210}, {"loss": 0.5399, "grad_norm": 1.016652226448059, "learning_rate": 0.0002, "epoch": 5.532293266510387, "step": 34220}, {"loss": 0.5158, "grad_norm": 1.2231552600860596, "learning_rate": 0.0002, "epoch": 5.5339099506911325, "step": 34230}, {"loss": 0.5091, "grad_norm": 0.8051198720932007, "learning_rate": 0.0002, "epoch": 5.535526634871878, "step": 34240}, {"loss": 0.5583, "grad_norm": 1.1779674291610718, "learning_rate": 0.0002, "epoch": 5.537143319052623, "step": 34250}, {"loss": 0.5044, "grad_norm": 1.2468291521072388, "learning_rate": 0.0002, "epoch": 5.538760003233368, "step": 34260}, {"loss": 0.523, "grad_norm": 1.14818274974823, "learning_rate": 0.0002, "epoch": 5.540376687414113, "step": 34270}, {"loss": 0.5375, "grad_norm": 1.2362616062164307, "learning_rate": 0.0002, "epoch": 5.541993371594859, "step": 34280}, {"loss": 0.4996, "grad_norm": 1.0206977128982544, "learning_rate": 0.0002, "epoch": 5.543610055775604, "step": 34290}, {"loss": 0.5212, "grad_norm": 1.2018457651138306, "learning_rate": 0.0002, "epoch": 5.54522673995635, "step": 34300}, {"loss": 0.5462, "grad_norm": 1.0349043607711792, "learning_rate": 0.0002, "epoch": 5.546843424137095, "step": 34310}, {"loss": 0.5231, "grad_norm": 1.2022006511688232, "learning_rate": 0.0002, "epoch": 5.54846010831784, "step": 34320}, {"loss": 0.5173, "grad_norm": 1.0810624361038208, "learning_rate": 0.0002, "epoch": 5.550076792498586, "step": 34330}, {"loss": 0.5821, "grad_norm": 1.3297529220581055, "learning_rate": 0.0002, "epoch": 5.551693476679331, "step": 34340}, {"loss": 0.5321, "grad_norm": 0.9722549915313721, "learning_rate": 0.0002, "epoch": 5.553310160860076, "step": 34350}, {"loss": 0.4823, "grad_norm": 0.9903425574302673, "learning_rate": 0.0002, "epoch": 5.554926845040821, "step": 34360}, {"loss": 0.5601, "grad_norm": 0.9568067789077759, "learning_rate": 0.0002, "epoch": 5.556543529221567, "step": 34370}, {"loss": 0.5242, "grad_norm": 1.113870620727539, "learning_rate": 0.0002, "epoch": 5.558160213402312, "step": 34380}, {"loss": 0.5278, "grad_norm": 1.0557632446289062, "learning_rate": 0.0002, "epoch": 5.559776897583057, "step": 34390}, {"loss": 0.5501, "grad_norm": 0.9615673422813416, "learning_rate": 0.0002, "epoch": 5.561393581763802, "step": 34400}, {"loss": 0.5066, "grad_norm": 0.9536027312278748, "learning_rate": 0.0002, "epoch": 5.5630102659445475, "step": 34410}, {"loss": 0.4949, "grad_norm": 0.8808749318122864, "learning_rate": 0.0002, "epoch": 5.564626950125293, "step": 34420}, {"loss": 0.5954, "grad_norm": 1.286132574081421, "learning_rate": 0.0002, "epoch": 5.566243634306038, "step": 34430}, {"loss": 0.5507, "grad_norm": 1.259644865989685, "learning_rate": 0.0002, "epoch": 5.567860318486783, "step": 34440}, {"loss": 0.4922, "grad_norm": 0.9920216798782349, "learning_rate": 0.0002, "epoch": 5.569477002667529, "step": 34450}, {"loss": 0.5527, "grad_norm": 1.182926893234253, "learning_rate": 0.0002, "epoch": 5.5710936868482746, "step": 34460}, {"loss": 0.5185, "grad_norm": 1.1434749364852905, "learning_rate": 0.0002, "epoch": 5.57271037102902, "step": 34470}, {"loss": 0.5256, "grad_norm": 1.2420979738235474, "learning_rate": 0.0002, "epoch": 5.574327055209765, "step": 34480}, {"loss": 0.5039, "grad_norm": 0.9338384866714478, "learning_rate": 0.0002, "epoch": 5.57594373939051, "step": 34490}, {"loss": 0.5634, "grad_norm": 1.0196425914764404, "learning_rate": 0.0002, "epoch": 5.5775604235712555, "step": 34500}, {"loss": 0.5132, "grad_norm": 0.9586997032165527, "learning_rate": 0.0002, "epoch": 5.579177107752001, "step": 34510}, {"loss": 0.5336, "grad_norm": 1.2409086227416992, "learning_rate": 0.0002, "epoch": 5.580793791932746, "step": 34520}, {"loss": 0.5364, "grad_norm": 1.1483757495880127, "learning_rate": 0.0002, "epoch": 5.582410476113491, "step": 34530}, {"loss": 0.5325, "grad_norm": 1.1624305248260498, "learning_rate": 0.0002, "epoch": 5.584027160294236, "step": 34540}, {"loss": 0.5342, "grad_norm": 1.2635223865509033, "learning_rate": 0.0002, "epoch": 5.585643844474982, "step": 34550}, {"loss": 0.4924, "grad_norm": 0.9824051856994629, "learning_rate": 0.0002, "epoch": 5.587260528655727, "step": 34560}, {"loss": 0.5395, "grad_norm": 1.0858620405197144, "learning_rate": 0.0002, "epoch": 5.588877212836472, "step": 34570}, {"loss": 0.5459, "grad_norm": 1.1452655792236328, "learning_rate": 0.0002, "epoch": 5.590493897017217, "step": 34580}, {"loss": 0.5746, "grad_norm": 1.110610842704773, "learning_rate": 0.0002, "epoch": 5.592110581197963, "step": 34590}, {"loss": 0.5285, "grad_norm": 0.9976194500923157, "learning_rate": 0.0002, "epoch": 5.593727265378709, "step": 34600}, {"loss": 0.548, "grad_norm": 1.0698920488357544, "learning_rate": 0.0002, "epoch": 5.595343949559454, "step": 34610}, {"loss": 0.5311, "grad_norm": 1.1505171060562134, "learning_rate": 0.0002, "epoch": 5.596960633740199, "step": 34620}, {"loss": 0.5471, "grad_norm": 1.1014643907546997, "learning_rate": 0.0002, "epoch": 5.598577317920944, "step": 34630}, {"loss": 0.55, "grad_norm": 0.915595293045044, "learning_rate": 0.0002, "epoch": 5.60019400210169, "step": 34640}, {"loss": 0.5821, "grad_norm": 1.1856765747070312, "learning_rate": 0.0002, "epoch": 5.601810686282435, "step": 34650}, {"loss": 0.5502, "grad_norm": 1.1357687711715698, "learning_rate": 0.0002, "epoch": 5.60342737046318, "step": 34660}, {"loss": 0.5034, "grad_norm": 1.0232492685317993, "learning_rate": 0.0002, "epoch": 5.605044054643925, "step": 34670}, {"loss": 0.5357, "grad_norm": 0.9375017881393433, "learning_rate": 0.0002, "epoch": 5.6066607388246705, "step": 34680}, {"loss": 0.5518, "grad_norm": 1.0796529054641724, "learning_rate": 0.0002, "epoch": 5.608277423005416, "step": 34690}, {"loss": 0.5173, "grad_norm": 1.1383336782455444, "learning_rate": 0.0002, "epoch": 5.609894107186161, "step": 34700}, {"loss": 0.5477, "grad_norm": 1.0248544216156006, "learning_rate": 0.0002, "epoch": 5.611510791366906, "step": 34710}, {"loss": 0.5669, "grad_norm": 1.0986040830612183, "learning_rate": 0.0002, "epoch": 5.6131274755476515, "step": 34720}, {"loss": 0.5188, "grad_norm": 1.2689568996429443, "learning_rate": 0.0002, "epoch": 5.614744159728397, "step": 34730}, {"loss": 0.5136, "grad_norm": 1.4044264554977417, "learning_rate": 0.0002, "epoch": 5.616360843909142, "step": 34740}, {"loss": 0.5699, "grad_norm": 1.2084474563598633, "learning_rate": 0.0002, "epoch": 5.617977528089888, "step": 34750}, {"loss": 0.5377, "grad_norm": 1.061248540878296, "learning_rate": 0.0002, "epoch": 5.619594212270633, "step": 34760}, {"loss": 0.5669, "grad_norm": 1.0220764875411987, "learning_rate": 0.0002, "epoch": 5.6212108964513785, "step": 34770}, {"loss": 0.54, "grad_norm": 1.0859092473983765, "learning_rate": 0.0002, "epoch": 5.622827580632124, "step": 34780}, {"loss": 0.5308, "grad_norm": 0.9049732089042664, "learning_rate": 0.0002, "epoch": 5.624444264812869, "step": 34790}, {"loss": 0.5433, "grad_norm": 1.2103937864303589, "learning_rate": 0.0002, "epoch": 5.626060948993614, "step": 34800}, {"loss": 0.5513, "grad_norm": 0.9854230284690857, "learning_rate": 0.0002, "epoch": 5.627677633174359, "step": 34810}, {"loss": 0.5274, "grad_norm": 0.9316635131835938, "learning_rate": 0.0002, "epoch": 5.629294317355105, "step": 34820}, {"loss": 0.5393, "grad_norm": 1.105296015739441, "learning_rate": 0.0002, "epoch": 5.63091100153585, "step": 34830}, {"loss": 0.5527, "grad_norm": 0.993383526802063, "learning_rate": 0.0002, "epoch": 5.632527685716595, "step": 34840}, {"loss": 0.5375, "grad_norm": 1.1544116735458374, "learning_rate": 0.0002, "epoch": 5.63414436989734, "step": 34850}, {"loss": 0.5448, "grad_norm": 1.284475326538086, "learning_rate": 0.0002, "epoch": 5.635761054078086, "step": 34860}, {"loss": 0.5069, "grad_norm": 1.121997594833374, "learning_rate": 0.0002, "epoch": 5.637377738258831, "step": 34870}, {"loss": 0.5335, "grad_norm": 1.213040828704834, "learning_rate": 0.0002, "epoch": 5.638994422439576, "step": 34880}, {"loss": 0.5623, "grad_norm": 1.23222017288208, "learning_rate": 0.0002, "epoch": 5.640611106620321, "step": 34890}, {"loss": 0.5622, "grad_norm": 0.9793637990951538, "learning_rate": 0.0002, "epoch": 5.642227790801067, "step": 34900}, {"loss": 0.5405, "grad_norm": 1.38919997215271, "learning_rate": 0.0002, "epoch": 5.643844474981813, "step": 34910}, {"loss": 0.5007, "grad_norm": 0.8390951156616211, "learning_rate": 0.0002, "epoch": 5.645461159162558, "step": 34920}, {"loss": 0.5974, "grad_norm": 0.9465909004211426, "learning_rate": 0.0002, "epoch": 5.647077843343303, "step": 34930}, {"loss": 0.5264, "grad_norm": 1.066957712173462, "learning_rate": 0.0002, "epoch": 5.648694527524048, "step": 34940}, {"loss": 0.5513, "grad_norm": 0.9842154383659363, "learning_rate": 0.0002, "epoch": 5.650311211704794, "step": 34950}, {"loss": 0.567, "grad_norm": 1.1766440868377686, "learning_rate": 0.0002, "epoch": 5.651927895885539, "step": 34960}, {"loss": 0.5462, "grad_norm": 0.9061306118965149, "learning_rate": 0.0002, "epoch": 5.653544580066284, "step": 34970}, {"loss": 0.5446, "grad_norm": 1.2941309213638306, "learning_rate": 0.0002, "epoch": 5.655161264247029, "step": 34980}, {"loss": 0.5704, "grad_norm": 0.9741247892379761, "learning_rate": 0.0002, "epoch": 5.6567779484277745, "step": 34990}, {"loss": 0.5152, "grad_norm": 1.0784187316894531, "learning_rate": 0.0002, "epoch": 5.65839463260852, "step": 35000}, {"loss": 0.5363, "grad_norm": 0.937889814376831, "learning_rate": 0.0002, "epoch": 5.660011316789265, "step": 35010}, {"loss": 0.5019, "grad_norm": 0.9667879939079285, "learning_rate": 0.0002, "epoch": 5.66162800097001, "step": 35020}, {"loss": 0.5209, "grad_norm": 1.0554876327514648, "learning_rate": 0.0002, "epoch": 5.663244685150756, "step": 35030}, {"loss": 0.523, "grad_norm": 1.2030539512634277, "learning_rate": 0.0002, "epoch": 5.664861369331501, "step": 35040}, {"loss": 0.5406, "grad_norm": 1.0849953889846802, "learning_rate": 0.0002, "epoch": 5.666478053512247, "step": 35050}, {"loss": 0.5747, "grad_norm": 1.1598973274230957, "learning_rate": 0.0002, "epoch": 5.668094737692992, "step": 35060}, {"loss": 0.5488, "grad_norm": 1.0233359336853027, "learning_rate": 0.0002, "epoch": 5.669711421873737, "step": 35070}, {"loss": 0.5409, "grad_norm": 1.1124799251556396, "learning_rate": 0.0002, "epoch": 5.6713281060544825, "step": 35080}, {"loss": 0.5578, "grad_norm": 1.2351475954055786, "learning_rate": 0.0002, "epoch": 5.672944790235228, "step": 35090}, {"loss": 0.5638, "grad_norm": 1.0240728855133057, "learning_rate": 0.0002, "epoch": 5.674561474415973, "step": 35100}, {"loss": 0.5192, "grad_norm": 1.0223692655563354, "learning_rate": 0.0002, "epoch": 5.676178158596718, "step": 35110}, {"loss": 0.524, "grad_norm": 1.4569132328033447, "learning_rate": 0.0002, "epoch": 5.677794842777463, "step": 35120}, {"loss": 0.555, "grad_norm": 0.8983587026596069, "learning_rate": 0.0002, "epoch": 5.679411526958209, "step": 35130}, {"loss": 0.5439, "grad_norm": 1.0775383710861206, "learning_rate": 0.0002, "epoch": 5.681028211138954, "step": 35140}, {"loss": 0.5289, "grad_norm": 0.9800270795822144, "learning_rate": 0.0002, "epoch": 5.682644895319699, "step": 35150}, {"loss": 0.533, "grad_norm": 0.9858237504959106, "learning_rate": 0.0002, "epoch": 5.684261579500444, "step": 35160}, {"loss": 0.5671, "grad_norm": 1.031087040901184, "learning_rate": 0.0002, "epoch": 5.6858782636811895, "step": 35170}, {"loss": 0.5528, "grad_norm": 1.0294365882873535, "learning_rate": 0.0002, "epoch": 5.687494947861936, "step": 35180}, {"loss": 0.5581, "grad_norm": 1.108144760131836, "learning_rate": 0.0002, "epoch": 5.68911163204268, "step": 35190}, {"loss": 0.5373, "grad_norm": 1.0813100337982178, "learning_rate": 0.0002, "epoch": 5.690728316223426, "step": 35200}, {"loss": 0.5429, "grad_norm": 1.3146867752075195, "learning_rate": 0.0002, "epoch": 5.692345000404171, "step": 35210}, {"loss": 0.5297, "grad_norm": 1.16780424118042, "learning_rate": 0.0002, "epoch": 5.693961684584917, "step": 35220}, {"loss": 0.577, "grad_norm": 0.9929125905036926, "learning_rate": 0.0002, "epoch": 5.695578368765662, "step": 35230}, {"loss": 0.5441, "grad_norm": 0.9049441814422607, "learning_rate": 0.0002, "epoch": 5.697195052946407, "step": 35240}, {"loss": 0.5349, "grad_norm": 0.9768866300582886, "learning_rate": 0.0002, "epoch": 5.698811737127152, "step": 35250}, {"loss": 0.542, "grad_norm": 0.8306029438972473, "learning_rate": 0.0002, "epoch": 5.7004284213078975, "step": 35260}, {"loss": 0.4771, "grad_norm": 0.8417280316352844, "learning_rate": 0.0002, "epoch": 5.702045105488643, "step": 35270}, {"loss": 0.574, "grad_norm": 0.9954485893249512, "learning_rate": 0.0002, "epoch": 5.703661789669388, "step": 35280}, {"loss": 0.5469, "grad_norm": 1.2417993545532227, "learning_rate": 0.0002, "epoch": 5.705278473850133, "step": 35290}, {"loss": 0.5275, "grad_norm": 1.1696544885635376, "learning_rate": 0.0002, "epoch": 5.706895158030878, "step": 35300}, {"loss": 0.5188, "grad_norm": 1.2424817085266113, "learning_rate": 0.0002, "epoch": 5.708511842211624, "step": 35310}, {"loss": 0.5595, "grad_norm": 1.1791106462478638, "learning_rate": 0.0002, "epoch": 5.710128526392369, "step": 35320}, {"loss": 0.5076, "grad_norm": 1.202181339263916, "learning_rate": 0.0002, "epoch": 5.711745210573115, "step": 35330}, {"loss": 0.5847, "grad_norm": 1.1006861925125122, "learning_rate": 0.0002, "epoch": 5.713361894753859, "step": 35340}, {"loss": 0.5627, "grad_norm": 1.0918344259262085, "learning_rate": 0.0002, "epoch": 5.7149785789346055, "step": 35350}, {"loss": 0.5677, "grad_norm": 1.0427305698394775, "learning_rate": 0.0002, "epoch": 5.716595263115351, "step": 35360}, {"loss": 0.5288, "grad_norm": 1.0818872451782227, "learning_rate": 0.0002, "epoch": 5.718211947296096, "step": 35370}, {"loss": 0.5296, "grad_norm": 1.186006784439087, "learning_rate": 0.0002, "epoch": 5.719828631476841, "step": 35380}, {"loss": 0.5507, "grad_norm": 1.2073674201965332, "learning_rate": 0.0002, "epoch": 5.721445315657586, "step": 35390}, {"loss": 0.5483, "grad_norm": 1.065338134765625, "learning_rate": 0.0002, "epoch": 5.723061999838332, "step": 35400}, {"loss": 0.5195, "grad_norm": 0.9448973536491394, "learning_rate": 0.0002, "epoch": 5.724678684019077, "step": 35410}, {"loss": 0.5276, "grad_norm": 1.1487499475479126, "learning_rate": 0.0002, "epoch": 5.726295368199822, "step": 35420}, {"loss": 0.5435, "grad_norm": 1.1334216594696045, "learning_rate": 0.0002, "epoch": 5.727912052380567, "step": 35430}, {"loss": 0.5074, "grad_norm": 1.1932826042175293, "learning_rate": 0.0002, "epoch": 5.729528736561313, "step": 35440}, {"loss": 0.5502, "grad_norm": 1.2615786790847778, "learning_rate": 0.0002, "epoch": 5.731145420742058, "step": 35450}, {"loss": 0.5612, "grad_norm": 1.2803694009780884, "learning_rate": 0.0002, "epoch": 5.732762104922803, "step": 35460}, {"loss": 0.5458, "grad_norm": 0.9271906614303589, "learning_rate": 0.0002, "epoch": 5.734378789103548, "step": 35470}, {"loss": 0.5342, "grad_norm": 1.0958917140960693, "learning_rate": 0.0002, "epoch": 5.735995473284294, "step": 35480}, {"loss": 0.538, "grad_norm": 1.1072784662246704, "learning_rate": 0.0002, "epoch": 5.737612157465039, "step": 35490}, {"loss": 0.5683, "grad_norm": 1.1641002893447876, "learning_rate": 0.0002, "epoch": 5.739228841645785, "step": 35500}, {"loss": 0.5252, "grad_norm": 1.0246447324752808, "learning_rate": 0.0002, "epoch": 5.74084552582653, "step": 35510}, {"loss": 0.55, "grad_norm": 1.032474398612976, "learning_rate": 0.0002, "epoch": 5.742462210007275, "step": 35520}, {"loss": 0.4965, "grad_norm": 1.1600854396820068, "learning_rate": 0.0002, "epoch": 5.7440788941880205, "step": 35530}, {"loss": 0.5543, "grad_norm": 1.0686054229736328, "learning_rate": 0.0002, "epoch": 5.745695578368766, "step": 35540}, {"loss": 0.5706, "grad_norm": 1.2314637899398804, "learning_rate": 0.0002, "epoch": 5.747312262549511, "step": 35550}, {"loss": 0.5492, "grad_norm": 0.922134280204773, "learning_rate": 0.0002, "epoch": 5.748928946730256, "step": 35560}, {"loss": 0.5495, "grad_norm": 0.933043360710144, "learning_rate": 0.0002, "epoch": 5.7505456309110015, "step": 35570}, {"loss": 0.5007, "grad_norm": 1.1911931037902832, "learning_rate": 0.0002, "epoch": 5.752162315091747, "step": 35580}, {"loss": 0.5244, "grad_norm": 0.8984857797622681, "learning_rate": 0.0002, "epoch": 5.753778999272492, "step": 35590}, {"loss": 0.5493, "grad_norm": 0.9495107531547546, "learning_rate": 0.0002, "epoch": 5.755395683453237, "step": 35600}, {"loss": 0.5326, "grad_norm": 1.2805472612380981, "learning_rate": 0.0002, "epoch": 5.757012367633982, "step": 35610}, {"loss": 0.5276, "grad_norm": 1.1236625909805298, "learning_rate": 0.0002, "epoch": 5.758629051814728, "step": 35620}, {"loss": 0.6102, "grad_norm": 1.0552798509597778, "learning_rate": 0.0002, "epoch": 5.760245735995474, "step": 35630}, {"loss": 0.5479, "grad_norm": 1.119909644126892, "learning_rate": 0.0002, "epoch": 5.761862420176218, "step": 35640}, {"loss": 0.5282, "grad_norm": 0.8786116242408752, "learning_rate": 0.0002, "epoch": 5.763479104356964, "step": 35650}, {"loss": 0.5406, "grad_norm": 1.2417117357254028, "learning_rate": 0.0002, "epoch": 5.765095788537709, "step": 35660}, {"loss": 0.537, "grad_norm": 1.255200982093811, "learning_rate": 0.0002, "epoch": 5.766712472718455, "step": 35670}, {"loss": 0.5308, "grad_norm": 1.0611358880996704, "learning_rate": 0.0002, "epoch": 5.7683291568992, "step": 35680}, {"loss": 0.5614, "grad_norm": 1.1443911790847778, "learning_rate": 0.0002, "epoch": 5.769945841079945, "step": 35690}, {"loss": 0.5386, "grad_norm": 1.1437989473342896, "learning_rate": 0.0002, "epoch": 5.77156252526069, "step": 35700}, {"loss": 0.537, "grad_norm": 1.1375046968460083, "learning_rate": 0.0002, "epoch": 5.773179209441436, "step": 35710}, {"loss": 0.5198, "grad_norm": 1.0777729749679565, "learning_rate": 0.0002, "epoch": 5.774795893622181, "step": 35720}, {"loss": 0.5521, "grad_norm": 1.1160215139389038, "learning_rate": 0.0002, "epoch": 5.776412577802926, "step": 35730}, {"loss": 0.5569, "grad_norm": 1.1268514394760132, "learning_rate": 0.0002, "epoch": 5.778029261983671, "step": 35740}, {"loss": 0.5311, "grad_norm": 1.2752262353897095, "learning_rate": 0.0002, "epoch": 5.7796459461644165, "step": 35750}, {"loss": 0.5625, "grad_norm": 1.0416184663772583, "learning_rate": 0.0002, "epoch": 5.781262630345162, "step": 35760}, {"loss": 0.5438, "grad_norm": 1.0622444152832031, "learning_rate": 0.0002, "epoch": 5.782879314525907, "step": 35770}, {"loss": 0.5268, "grad_norm": 1.1217877864837646, "learning_rate": 0.0002, "epoch": 5.784495998706653, "step": 35780}, {"loss": 0.5225, "grad_norm": 0.9363139867782593, "learning_rate": 0.0002, "epoch": 5.786112682887398, "step": 35790}, {"loss": 0.5524, "grad_norm": 0.96628737449646, "learning_rate": 0.0002, "epoch": 5.787729367068144, "step": 35800}, {"loss": 0.52, "grad_norm": 0.9572572112083435, "learning_rate": 0.0002, "epoch": 5.789346051248889, "step": 35810}, {"loss": 0.5615, "grad_norm": 0.938724935054779, "learning_rate": 0.0002, "epoch": 5.790962735429634, "step": 35820}, {"loss": 0.5391, "grad_norm": 1.3314417600631714, "learning_rate": 0.0002, "epoch": 5.792579419610379, "step": 35830}, {"loss": 0.5441, "grad_norm": 1.0097602605819702, "learning_rate": 0.0002, "epoch": 5.7941961037911245, "step": 35840}, {"loss": 0.591, "grad_norm": 1.1265122890472412, "learning_rate": 0.0002, "epoch": 5.79581278797187, "step": 35850}, {"loss": 0.5333, "grad_norm": 1.2191909551620483, "learning_rate": 0.0002, "epoch": 5.797429472152615, "step": 35860}, {"loss": 0.5274, "grad_norm": 0.9690808057785034, "learning_rate": 0.0002, "epoch": 5.79904615633336, "step": 35870}, {"loss": 0.5425, "grad_norm": 1.0871665477752686, "learning_rate": 0.0002, "epoch": 5.800662840514105, "step": 35880}, {"loss": 0.5602, "grad_norm": 1.1093597412109375, "learning_rate": 0.0002, "epoch": 5.802279524694851, "step": 35890}, {"loss": 0.5475, "grad_norm": 1.2434282302856445, "learning_rate": 0.0002, "epoch": 5.803896208875596, "step": 35900}, {"loss": 0.5288, "grad_norm": 1.2933623790740967, "learning_rate": 0.0002, "epoch": 5.805512893056341, "step": 35910}, {"loss": 0.5554, "grad_norm": 1.0005441904067993, "learning_rate": 0.0002, "epoch": 5.807129577237086, "step": 35920}, {"loss": 0.5318, "grad_norm": 1.2373108863830566, "learning_rate": 0.0002, "epoch": 5.8087462614178325, "step": 35930}, {"loss": 0.5413, "grad_norm": 1.2622692584991455, "learning_rate": 0.0002, "epoch": 5.810362945598578, "step": 35940}, {"loss": 0.5558, "grad_norm": 1.0112963914871216, "learning_rate": 0.0002, "epoch": 5.811979629779323, "step": 35950}, {"loss": 0.5115, "grad_norm": 1.050572395324707, "learning_rate": 0.0002, "epoch": 5.813596313960068, "step": 35960}, {"loss": 0.5288, "grad_norm": 0.9774560928344727, "learning_rate": 0.0002, "epoch": 5.815212998140813, "step": 35970}, {"loss": 0.585, "grad_norm": 1.19438898563385, "learning_rate": 0.0002, "epoch": 5.816829682321559, "step": 35980}, {"loss": 0.5798, "grad_norm": 1.0267130136489868, "learning_rate": 0.0002, "epoch": 5.818446366502304, "step": 35990}, {"loss": 0.5126, "grad_norm": 0.9813851714134216, "learning_rate": 0.0002, "epoch": 5.820063050683049, "step": 36000}, {"loss": 0.5138, "grad_norm": 0.9177457094192505, "learning_rate": 0.0002, "epoch": 5.821679734863794, "step": 36010}, {"loss": 0.5453, "grad_norm": 1.0020731687545776, "learning_rate": 0.0002, "epoch": 5.8232964190445395, "step": 36020}, {"loss": 0.5646, "grad_norm": 1.073222041130066, "learning_rate": 0.0002, "epoch": 5.824913103225285, "step": 36030}, {"loss": 0.5539, "grad_norm": 1.016337513923645, "learning_rate": 0.0002, "epoch": 5.82652978740603, "step": 36040}, {"loss": 0.5592, "grad_norm": 1.267364263534546, "learning_rate": 0.0002, "epoch": 5.828146471586775, "step": 36050}, {"loss": 0.595, "grad_norm": 1.2730127573013306, "learning_rate": 0.0002, "epoch": 5.8297631557675205, "step": 36060}, {"loss": 0.5247, "grad_norm": 1.108442783355713, "learning_rate": 0.0002, "epoch": 5.831379839948266, "step": 36070}, {"loss": 0.5103, "grad_norm": 1.198072075843811, "learning_rate": 0.0002, "epoch": 5.832996524129012, "step": 36080}, {"loss": 0.5479, "grad_norm": 1.0458786487579346, "learning_rate": 0.0002, "epoch": 5.834613208309757, "step": 36090}, {"loss": 0.5564, "grad_norm": 0.9096664786338806, "learning_rate": 0.0002, "epoch": 5.836229892490502, "step": 36100}, {"loss": 0.5602, "grad_norm": 0.9957793951034546, "learning_rate": 0.0002, "epoch": 5.8378465766712475, "step": 36110}, {"loss": 0.5799, "grad_norm": 1.3693058490753174, "learning_rate": 0.0002, "epoch": 5.839463260851993, "step": 36120}, {"loss": 0.5425, "grad_norm": 1.268608808517456, "learning_rate": 0.0002, "epoch": 5.841079945032738, "step": 36130}, {"loss": 0.5653, "grad_norm": 0.8516020178794861, "learning_rate": 0.0002, "epoch": 5.842696629213483, "step": 36140}, {"loss": 0.5475, "grad_norm": 0.90385502576828, "learning_rate": 0.0002, "epoch": 5.844313313394228, "step": 36150}, {"loss": 0.5274, "grad_norm": 1.0910571813583374, "learning_rate": 0.0002, "epoch": 5.845929997574974, "step": 36160}, {"loss": 0.555, "grad_norm": 0.9417795538902283, "learning_rate": 0.0002, "epoch": 5.847546681755719, "step": 36170}, {"loss": 0.5784, "grad_norm": 1.0027360916137695, "learning_rate": 0.0002, "epoch": 5.849163365936464, "step": 36180}, {"loss": 0.5423, "grad_norm": 1.1480516195297241, "learning_rate": 0.0002, "epoch": 5.850780050117209, "step": 36190}, {"loss": 0.5517, "grad_norm": 1.2431457042694092, "learning_rate": 0.0002, "epoch": 5.852396734297955, "step": 36200}, {"loss": 0.5404, "grad_norm": 1.091465950012207, "learning_rate": 0.0002, "epoch": 5.8540134184787, "step": 36210}, {"loss": 0.53, "grad_norm": 0.9693930745124817, "learning_rate": 0.0002, "epoch": 5.855630102659445, "step": 36220}, {"loss": 0.5453, "grad_norm": 0.9937465190887451, "learning_rate": 0.0002, "epoch": 5.857246786840191, "step": 36230}, {"loss": 0.5621, "grad_norm": 1.0731011629104614, "learning_rate": 0.0002, "epoch": 5.858863471020936, "step": 36240}, {"loss": 0.5687, "grad_norm": 1.0869048833847046, "learning_rate": 0.0002, "epoch": 5.860480155201682, "step": 36250}, {"loss": 0.5576, "grad_norm": 0.9226390719413757, "learning_rate": 0.0002, "epoch": 5.862096839382427, "step": 36260}, {"loss": 0.531, "grad_norm": 1.1755430698394775, "learning_rate": 0.0002, "epoch": 5.863713523563172, "step": 36270}, {"loss": 0.558, "grad_norm": 0.8815974593162537, "learning_rate": 0.0002, "epoch": 5.865330207743917, "step": 36280}, {"loss": 0.5065, "grad_norm": 1.3648751974105835, "learning_rate": 0.0002, "epoch": 5.866946891924663, "step": 36290}, {"loss": 0.536, "grad_norm": 0.8729211091995239, "learning_rate": 0.0002, "epoch": 5.868563576105408, "step": 36300}, {"loss": 0.5192, "grad_norm": 1.0870907306671143, "learning_rate": 0.0002, "epoch": 5.870180260286153, "step": 36310}, {"loss": 0.5609, "grad_norm": 1.1164259910583496, "learning_rate": 0.0002, "epoch": 5.871796944466898, "step": 36320}, {"loss": 0.551, "grad_norm": 1.1572535037994385, "learning_rate": 0.0002, "epoch": 5.8734136286476435, "step": 36330}, {"loss": 0.5898, "grad_norm": 1.0456238985061646, "learning_rate": 0.0002, "epoch": 5.875030312828389, "step": 36340}, {"loss": 0.5008, "grad_norm": 1.1310722827911377, "learning_rate": 0.0002, "epoch": 5.876646997009134, "step": 36350}, {"loss": 0.5352, "grad_norm": 1.0004712343215942, "learning_rate": 0.0002, "epoch": 5.878263681189879, "step": 36360}, {"loss": 0.5632, "grad_norm": 1.0991777181625366, "learning_rate": 0.0002, "epoch": 5.879880365370624, "step": 36370}, {"loss": 0.5815, "grad_norm": 1.2789239883422852, "learning_rate": 0.0002, "epoch": 5.8814970495513705, "step": 36380}, {"loss": 0.56, "grad_norm": 0.9524819850921631, "learning_rate": 0.0002, "epoch": 5.883113733732116, "step": 36390}, {"loss": 0.5701, "grad_norm": 1.1115771532058716, "learning_rate": 0.0002, "epoch": 5.884730417912861, "step": 36400}, {"loss": 0.5463, "grad_norm": 1.37419855594635, "learning_rate": 0.0002, "epoch": 5.886347102093606, "step": 36410}, {"loss": 0.5675, "grad_norm": 1.1449527740478516, "learning_rate": 0.0002, "epoch": 5.8879637862743515, "step": 36420}, {"loss": 0.5255, "grad_norm": 1.198046326637268, "learning_rate": 0.0002, "epoch": 5.889580470455097, "step": 36430}, {"loss": 0.5383, "grad_norm": 1.0180530548095703, "learning_rate": 0.0002, "epoch": 5.891197154635842, "step": 36440}, {"loss": 0.5319, "grad_norm": 1.0516417026519775, "learning_rate": 0.0002, "epoch": 5.892813838816587, "step": 36450}, {"loss": 0.5782, "grad_norm": 1.1658052206039429, "learning_rate": 0.0002, "epoch": 5.894430522997332, "step": 36460}, {"loss": 0.5864, "grad_norm": 1.190699577331543, "learning_rate": 0.0002, "epoch": 5.896047207178078, "step": 36470}, {"loss": 0.5451, "grad_norm": 1.1235495805740356, "learning_rate": 0.0002, "epoch": 5.897663891358823, "step": 36480}, {"loss": 0.5284, "grad_norm": 1.1926926374435425, "learning_rate": 0.0002, "epoch": 5.899280575539568, "step": 36490}, {"loss": 0.5686, "grad_norm": 1.1184662580490112, "learning_rate": 0.0002, "epoch": 5.900897259720313, "step": 36500}, {"loss": 0.5147, "grad_norm": 1.000970721244812, "learning_rate": 0.0002, "epoch": 5.9025139439010585, "step": 36510}, {"loss": 0.5351, "grad_norm": 1.0373306274414062, "learning_rate": 0.0002, "epoch": 5.904130628081804, "step": 36520}, {"loss": 0.535, "grad_norm": 1.0840669870376587, "learning_rate": 0.0002, "epoch": 5.90574731226255, "step": 36530}, {"loss": 0.538, "grad_norm": 0.9908381104469299, "learning_rate": 0.0002, "epoch": 5.907363996443295, "step": 36540}, {"loss": 0.5313, "grad_norm": 1.0456029176712036, "learning_rate": 0.0002, "epoch": 5.90898068062404, "step": 36550}, {"loss": 0.5693, "grad_norm": 1.1381454467773438, "learning_rate": 0.0002, "epoch": 5.910597364804786, "step": 36560}, {"loss": 0.5473, "grad_norm": 0.9440900087356567, "learning_rate": 0.0002, "epoch": 5.912214048985531, "step": 36570}, {"loss": 0.5542, "grad_norm": 1.1674573421478271, "learning_rate": 0.0002, "epoch": 5.913830733166276, "step": 36580}, {"loss": 0.526, "grad_norm": 1.1226966381072998, "learning_rate": 0.0002, "epoch": 5.915447417347021, "step": 36590}, {"loss": 0.6091, "grad_norm": 0.9696915745735168, "learning_rate": 0.0002, "epoch": 5.9170641015277665, "step": 36600}, {"loss": 0.5523, "grad_norm": 0.9593005180358887, "learning_rate": 0.0002, "epoch": 5.918680785708512, "step": 36610}, {"loss": 0.5536, "grad_norm": 1.122169852256775, "learning_rate": 0.0002, "epoch": 5.920297469889257, "step": 36620}, {"loss": 0.5039, "grad_norm": 0.9923415780067444, "learning_rate": 0.0002, "epoch": 5.921914154070002, "step": 36630}, {"loss": 0.5893, "grad_norm": 1.063838005065918, "learning_rate": 0.0002, "epoch": 5.923530838250747, "step": 36640}, {"loss": 0.5799, "grad_norm": 0.9083505272865295, "learning_rate": 0.0002, "epoch": 5.925147522431493, "step": 36650}, {"loss": 0.5264, "grad_norm": 0.9439437985420227, "learning_rate": 0.0002, "epoch": 5.926764206612239, "step": 36660}, {"loss": 0.5891, "grad_norm": 0.9778534173965454, "learning_rate": 0.0002, "epoch": 5.928380890792983, "step": 36670}, {"loss": 0.566, "grad_norm": 0.9723961353302002, "learning_rate": 0.0002, "epoch": 5.929997574973729, "step": 36680}, {"loss": 0.5741, "grad_norm": 1.162333607673645, "learning_rate": 0.0002, "epoch": 5.9316142591544745, "step": 36690}, {"loss": 0.5771, "grad_norm": 1.2784897089004517, "learning_rate": 0.0002, "epoch": 5.93323094333522, "step": 36700}, {"loss": 0.5343, "grad_norm": 1.0924867391586304, "learning_rate": 0.0002, "epoch": 5.934847627515965, "step": 36710}, {"loss": 0.5554, "grad_norm": 1.046922206878662, "learning_rate": 0.0002, "epoch": 5.93646431169671, "step": 36720}, {"loss": 0.5476, "grad_norm": 0.8632535338401794, "learning_rate": 0.0002, "epoch": 5.938080995877455, "step": 36730}, {"loss": 0.5456, "grad_norm": 1.358762502670288, "learning_rate": 0.0002, "epoch": 5.939697680058201, "step": 36740}, {"loss": 0.551, "grad_norm": 1.2058624029159546, "learning_rate": 0.0002, "epoch": 5.941314364238946, "step": 36750}, {"loss": 0.5462, "grad_norm": 1.1396408081054688, "learning_rate": 0.0002, "epoch": 5.942931048419691, "step": 36760}, {"loss": 0.5483, "grad_norm": 1.1510354280471802, "learning_rate": 0.0002, "epoch": 5.944547732600436, "step": 36770}, {"loss": 0.5659, "grad_norm": 1.1401607990264893, "learning_rate": 0.0002, "epoch": 5.946164416781182, "step": 36780}, {"loss": 0.5557, "grad_norm": 1.1871325969696045, "learning_rate": 0.0002, "epoch": 5.947781100961927, "step": 36790}, {"loss": 0.4945, "grad_norm": 0.9928333163261414, "learning_rate": 0.0002, "epoch": 5.949397785142672, "step": 36800}, {"loss": 0.5303, "grad_norm": 1.0549445152282715, "learning_rate": 0.0002, "epoch": 5.951014469323418, "step": 36810}, {"loss": 0.5532, "grad_norm": 0.9791563749313354, "learning_rate": 0.0002, "epoch": 5.9526311535041625, "step": 36820}, {"loss": 0.5317, "grad_norm": 1.1268441677093506, "learning_rate": 0.0002, "epoch": 5.954247837684909, "step": 36830}, {"loss": 0.5585, "grad_norm": 1.0533992052078247, "learning_rate": 0.0002, "epoch": 5.955864521865654, "step": 36840}, {"loss": 0.4972, "grad_norm": 1.023358941078186, "learning_rate": 0.0002, "epoch": 5.957481206046399, "step": 36850}, {"loss": 0.5557, "grad_norm": 1.2631961107254028, "learning_rate": 0.0002, "epoch": 5.959097890227144, "step": 36860}, {"loss": 0.5662, "grad_norm": 0.9397698640823364, "learning_rate": 0.0002, "epoch": 5.9607145744078895, "step": 36870}, {"loss": 0.5775, "grad_norm": 1.1678427457809448, "learning_rate": 0.0002, "epoch": 5.962331258588635, "step": 36880}, {"loss": 0.5435, "grad_norm": 1.1403759717941284, "learning_rate": 0.0002, "epoch": 5.96394794276938, "step": 36890}, {"loss": 0.5479, "grad_norm": 1.030572772026062, "learning_rate": 0.0002, "epoch": 5.965564626950125, "step": 36900}, {"loss": 0.5838, "grad_norm": 1.0992497205734253, "learning_rate": 0.0002, "epoch": 5.9671813111308705, "step": 36910}, {"loss": 0.5452, "grad_norm": 1.075466275215149, "learning_rate": 0.0002, "epoch": 5.968797995311616, "step": 36920}, {"loss": 0.5739, "grad_norm": 1.0153694152832031, "learning_rate": 0.0002, "epoch": 5.970414679492361, "step": 36930}, {"loss": 0.5672, "grad_norm": 0.973193883895874, "learning_rate": 0.0002, "epoch": 5.972031363673106, "step": 36940}, {"loss": 0.5585, "grad_norm": 0.8294678926467896, "learning_rate": 0.0002, "epoch": 5.973648047853851, "step": 36950}, {"loss": 0.5631, "grad_norm": 1.0048716068267822, "learning_rate": 0.0002, "epoch": 5.9752647320345975, "step": 36960}, {"loss": 0.5471, "grad_norm": 0.9714070558547974, "learning_rate": 0.0002, "epoch": 5.976881416215342, "step": 36970}, {"loss": 0.5419, "grad_norm": 0.8667682409286499, "learning_rate": 0.0002, "epoch": 5.978498100396088, "step": 36980}, {"loss": 0.5474, "grad_norm": 1.0461409091949463, "learning_rate": 0.0002, "epoch": 5.980114784576833, "step": 36990}, {"loss": 0.5454, "grad_norm": 0.9229754209518433, "learning_rate": 0.0002, "epoch": 5.981731468757578, "step": 37000}, {"loss": 0.5599, "grad_norm": 1.0406876802444458, "learning_rate": 0.0002, "epoch": 5.983348152938324, "step": 37010}, {"loss": 0.5569, "grad_norm": 0.8993828296661377, "learning_rate": 0.0002, "epoch": 5.984964837119069, "step": 37020}, {"loss": 0.5611, "grad_norm": 1.2260479927062988, "learning_rate": 0.0002, "epoch": 5.986581521299814, "step": 37030}, {"loss": 0.5523, "grad_norm": 1.0107380151748657, "learning_rate": 0.0002, "epoch": 5.988198205480559, "step": 37040}, {"loss": 0.5639, "grad_norm": 1.0240139961242676, "learning_rate": 0.0002, "epoch": 5.989814889661305, "step": 37050}, {"loss": 0.5209, "grad_norm": 1.0185275077819824, "learning_rate": 0.0002, "epoch": 5.99143157384205, "step": 37060}, {"loss": 0.5114, "grad_norm": 1.1361802816390991, "learning_rate": 0.0002, "epoch": 5.993048258022795, "step": 37070}, {"loss": 0.5692, "grad_norm": 1.0395532846450806, "learning_rate": 0.0002, "epoch": 5.99466494220354, "step": 37080}, {"loss": 0.594, "grad_norm": 0.9463558197021484, "learning_rate": 0.0002, "epoch": 5.9962816263842855, "step": 37090}, {"loss": 0.5775, "grad_norm": 1.2066948413848877, "learning_rate": 0.0002, "epoch": 5.997898310565031, "step": 37100}, {"loss": 0.5356, "grad_norm": 0.9749386310577393, "learning_rate": 0.0002, "epoch": 5.999514994745777, "step": 37110}]}
+{"epoch": 6.9999191657909625, "step": 43298, "epoch_duration": 16911.370942354202, "total_accumulated_duration": 118243.28742051125, "gpu_info": {"GPU_0": "NVIDIA A100-PCIE-40GB"}, "memory_usage": {"avg_memory_usage": {"GPU_0": 7672.4541015625}, "peak_memory_usage": {"GPU_0": 13792.75537109375}, "avg_memory_reserved": {"GPU_0": 17416.0}, "peak_memory_reserved": {"GPU_0": 17416.0}, "total_memory": {"GPU_0": 40444.375}}, "best_checkpoint_path": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185", "params": {"epochs": 8, "batch_size": 1, "learning_rate": 0.0002, "gradient_accumulation_steps": 2, "warmup_ratio": 0.03, "max_grad_norm": 0.3, "lora_alpha": 32, "lora_dropout": 0.05, "lora_r": 64, "tf32": true, "seed": 42}, "log_history": [{"loss": 1.6636, "grad_norm": 0.9894065856933594, "learning_rate": 0.0002, "epoch": 0.0016166841807452913, "step": 10}, {"loss": 1.1528, "grad_norm": 1.7810699939727783, "learning_rate": 0.0002, "epoch": 0.0032333683614905826, "step": 20}, {"loss": 0.9767, "grad_norm": 0.5969577431678772, "learning_rate": 0.0002, "epoch": 0.004850052542235874, "step": 30}, {"loss": 0.9772, "grad_norm": 0.6354120969772339, "learning_rate": 0.0002, "epoch": 0.006466736722981165, "step": 40}, {"loss": 0.8643, "grad_norm": 0.5604607462882996, "learning_rate": 0.0002, "epoch": 0.008083420903726457, "step": 50}, {"loss": 0.8841, "grad_norm": 0.4676193594932556, "learning_rate": 0.0002, "epoch": 0.009700105084471748, "step": 60}, {"loss": 0.9022, "grad_norm": 0.6099211573600769, "learning_rate": 0.0002, "epoch": 0.01131678926521704, "step": 70}, {"loss": 0.9133, "grad_norm": 0.48639994859695435, "learning_rate": 0.0002, "epoch": 0.01293347344596233, "step": 80}, {"loss": 0.8704, "grad_norm": 0.4904264509677887, "learning_rate": 0.0002, "epoch": 0.014550157626707623, "step": 90}, {"loss": 0.8855, "grad_norm": 2.8334362506866455, "learning_rate": 0.0002, "epoch": 0.016166841807452915, "step": 100}, {"loss": 0.8958, "grad_norm": 0.43221670389175415, "learning_rate": 0.0002, "epoch": 0.017783525988198205, "step": 110}, {"loss": 0.8412, "grad_norm": 0.42244166135787964, "learning_rate": 0.0002, "epoch": 0.019400210168943496, "step": 120}, {"loss": 0.8467, "grad_norm": 0.45363298058509827, "learning_rate": 0.0002, "epoch": 0.02101689434968879, "step": 130}, {"loss": 0.8641, "grad_norm": 0.44816508889198303, "learning_rate": 0.0002, "epoch": 0.02263357853043408, "step": 140}, {"loss": 0.8496, "grad_norm": 0.43308213353157043, "learning_rate": 0.0002, "epoch": 0.02425026271117937, "step": 150}, {"loss": 0.8213, "grad_norm": 0.4084763526916504, "learning_rate": 0.0002, "epoch": 0.02586694689192466, "step": 160}, {"loss": 0.8343, "grad_norm": 0.5363703966140747, "learning_rate": 0.0002, "epoch": 0.027483631072669955, "step": 170}, {"loss": 0.8558, "grad_norm": 0.4619699716567993, "learning_rate": 0.0002, "epoch": 0.029100315253415245, "step": 180}, {"loss": 0.8878, "grad_norm": 0.49069908261299133, "learning_rate": 0.0002, "epoch": 0.030716999434160536, "step": 190}, {"loss": 0.8867, "grad_norm": 0.4645835757255554, "learning_rate": 0.0002, "epoch": 0.03233368361490583, "step": 200}, {"loss": 0.8842, "grad_norm": 1.2411243915557861, "learning_rate": 0.0002, "epoch": 0.03395036779565112, "step": 210}, {"loss": 0.8245, "grad_norm": 0.5211851596832275, "learning_rate": 0.0002, "epoch": 0.03556705197639641, "step": 220}, {"loss": 0.8194, "grad_norm": 0.5253691673278809, "learning_rate": 0.0002, "epoch": 0.037183736157141704, "step": 230}, {"loss": 0.8856, "grad_norm": 0.4567478895187378, "learning_rate": 0.0002, "epoch": 0.03880042033788699, "step": 240}, {"loss": 0.838, "grad_norm": 0.5472128391265869, "learning_rate": 0.0002, "epoch": 0.040417104518632285, "step": 250}, {"loss": 0.8201, "grad_norm": 0.42978546023368835, "learning_rate": 0.0002, "epoch": 0.04203378869937758, "step": 260}, {"loss": 0.8334, "grad_norm": 0.601734459400177, "learning_rate": 0.0002, "epoch": 0.043650472880122866, "step": 270}, {"loss": 0.815, "grad_norm": 0.4286513328552246, "learning_rate": 0.0002, "epoch": 0.04526715706086816, "step": 280}, {"loss": 0.8758, "grad_norm": 0.5230861902236938, "learning_rate": 0.0002, "epoch": 0.046883841241613454, "step": 290}, {"loss": 0.8636, "grad_norm": 0.6504611968994141, "learning_rate": 0.0002, "epoch": 0.04850052542235874, "step": 300}, {"loss": 0.8102, "grad_norm": 0.43485215306282043, "learning_rate": 0.0002, "epoch": 0.050117209603104035, "step": 310}, {"loss": 0.8221, "grad_norm": 0.4717007875442505, "learning_rate": 0.0002, "epoch": 0.05173389378384932, "step": 320}, {"loss": 0.8469, "grad_norm": 0.4059787690639496, "learning_rate": 0.0002, "epoch": 0.053350577964594616, "step": 330}, {"loss": 0.8866, "grad_norm": 0.4366913437843323, "learning_rate": 0.0002, "epoch": 0.05496726214533991, "step": 340}, {"loss": 0.7976, "grad_norm": 0.4233848452568054, "learning_rate": 0.0002, "epoch": 0.0565839463260852, "step": 350}, {"loss": 0.8456, "grad_norm": 0.4209108352661133, "learning_rate": 0.0002, "epoch": 0.05820063050683049, "step": 360}, {"loss": 0.816, "grad_norm": 0.41637396812438965, "learning_rate": 0.0002, "epoch": 0.059817314687575784, "step": 370}, {"loss": 0.7976, "grad_norm": 0.46235376596450806, "learning_rate": 0.0002, "epoch": 0.06143399886832107, "step": 380}, {"loss": 0.7966, "grad_norm": 0.4013484716415405, "learning_rate": 0.0002, "epoch": 0.06305068304906636, "step": 390}, {"loss": 0.8253, "grad_norm": 0.47443896532058716, "learning_rate": 0.0002, "epoch": 0.06466736722981166, "step": 400}, {"loss": 0.8666, "grad_norm": 0.3942156434059143, "learning_rate": 0.0002, "epoch": 0.06628405141055695, "step": 410}, {"loss": 0.8402, "grad_norm": 0.4965320825576782, "learning_rate": 0.0002, "epoch": 0.06790073559130223, "step": 420}, {"loss": 0.8317, "grad_norm": 0.4304835796356201, "learning_rate": 0.0002, "epoch": 0.06951741977204753, "step": 430}, {"loss": 0.8528, "grad_norm": 0.511726975440979, "learning_rate": 0.0002, "epoch": 0.07113410395279282, "step": 440}, {"loss": 0.8675, "grad_norm": 0.4040689170360565, "learning_rate": 0.0002, "epoch": 0.07275078813353811, "step": 450}, {"loss": 0.8788, "grad_norm": 0.5402171015739441, "learning_rate": 0.0002, "epoch": 0.07436747231428341, "step": 460}, {"loss": 0.8737, "grad_norm": 0.4174517095088959, "learning_rate": 0.0002, "epoch": 0.0759841564950287, "step": 470}, {"loss": 0.7605, "grad_norm": 0.4306182265281677, "learning_rate": 0.0002, "epoch": 0.07760084067577398, "step": 480}, {"loss": 0.799, "grad_norm": 0.535210132598877, "learning_rate": 0.0002, "epoch": 0.07921752485651928, "step": 490}, {"loss": 0.7825, "grad_norm": 0.5339109897613525, "learning_rate": 0.0002, "epoch": 0.08083420903726457, "step": 500}, {"loss": 0.8985, "grad_norm": 0.45754891633987427, "learning_rate": 0.0002, "epoch": 0.08245089321800986, "step": 510}, {"loss": 0.8144, "grad_norm": 0.43820783495903015, "learning_rate": 0.0002, "epoch": 0.08406757739875516, "step": 520}, {"loss": 0.8001, "grad_norm": 0.4434749186038971, "learning_rate": 0.0002, "epoch": 0.08568426157950045, "step": 530}, {"loss": 0.7857, "grad_norm": 0.43111467361450195, "learning_rate": 0.0002, "epoch": 0.08730094576024573, "step": 540}, {"loss": 0.8418, "grad_norm": 0.4378940165042877, "learning_rate": 0.0002, "epoch": 0.08891762994099103, "step": 550}, {"loss": 0.8361, "grad_norm": 0.4772215187549591, "learning_rate": 0.0002, "epoch": 0.09053431412173632, "step": 560}, {"loss": 0.8268, "grad_norm": 0.6837629079818726, "learning_rate": 0.0002, "epoch": 0.09215099830248161, "step": 570}, {"loss": 0.8607, "grad_norm": 0.42241212725639343, "learning_rate": 0.0002, "epoch": 0.09376768248322691, "step": 580}, {"loss": 0.852, "grad_norm": 0.5165936350822449, "learning_rate": 0.0002, "epoch": 0.0953843666639722, "step": 590}, {"loss": 0.8664, "grad_norm": 0.48737478256225586, "learning_rate": 0.0002, "epoch": 0.09700105084471748, "step": 600}, {"loss": 0.8806, "grad_norm": 0.47419852018356323, "learning_rate": 0.0002, "epoch": 0.09861773502546278, "step": 610}, {"loss": 0.8254, "grad_norm": 0.4975486099720001, "learning_rate": 0.0002, "epoch": 0.10023441920620807, "step": 620}, {"loss": 0.8548, "grad_norm": 0.49123844504356384, "learning_rate": 0.0002, "epoch": 0.10185110338695336, "step": 630}, {"loss": 0.8911, "grad_norm": 0.6288952827453613, "learning_rate": 0.0002, "epoch": 0.10346778756769864, "step": 640}, {"loss": 0.827, "grad_norm": 0.4277345836162567, "learning_rate": 0.0002, "epoch": 0.10508447174844394, "step": 650}, {"loss": 0.7996, "grad_norm": 0.4021061956882477, "learning_rate": 0.0002, "epoch": 0.10670115592918923, "step": 660}, {"loss": 0.87, "grad_norm": 0.3492237329483032, "learning_rate": 0.0002, "epoch": 0.10831784010993452, "step": 670}, {"loss": 0.8698, "grad_norm": 0.4341012239456177, "learning_rate": 0.0002, "epoch": 0.10993452429067982, "step": 680}, {"loss": 0.781, "grad_norm": 0.7296304106712341, "learning_rate": 0.0002, "epoch": 0.1115512084714251, "step": 690}, {"loss": 0.8433, "grad_norm": 0.397494912147522, "learning_rate": 0.0002, "epoch": 0.1131678926521704, "step": 700}, {"loss": 0.827, "grad_norm": 0.396431028842926, "learning_rate": 0.0002, "epoch": 0.1147845768329157, "step": 710}, {"loss": 0.8379, "grad_norm": 0.48842838406562805, "learning_rate": 0.0002, "epoch": 0.11640126101366098, "step": 720}, {"loss": 0.8238, "grad_norm": 0.46322616934776306, "learning_rate": 0.0002, "epoch": 0.11801794519440627, "step": 730}, {"loss": 0.8041, "grad_norm": 0.47990912199020386, "learning_rate": 0.0002, "epoch": 0.11963462937515157, "step": 740}, {"loss": 0.82, "grad_norm": 0.4997142255306244, "learning_rate": 0.0002, "epoch": 0.12125131355589686, "step": 750}, {"loss": 0.7702, "grad_norm": 0.4040526747703552, "learning_rate": 0.0002, "epoch": 0.12286799773664214, "step": 760}, {"loss": 0.863, "grad_norm": 0.453095942735672, "learning_rate": 0.0002, "epoch": 0.12448468191738744, "step": 770}, {"loss": 0.8792, "grad_norm": 0.4636971950531006, "learning_rate": 0.0002, "epoch": 0.12610136609813272, "step": 780}, {"loss": 0.8112, "grad_norm": 0.4279276132583618, "learning_rate": 0.0002, "epoch": 0.12771805027887803, "step": 790}, {"loss": 0.8711, "grad_norm": 0.46212655305862427, "learning_rate": 0.0002, "epoch": 0.12933473445962332, "step": 800}, {"loss": 0.8368, "grad_norm": 0.43127650022506714, "learning_rate": 0.0002, "epoch": 0.1309514186403686, "step": 810}, {"loss": 0.8476, "grad_norm": 0.4201301336288452, "learning_rate": 0.0002, "epoch": 0.1325681028211139, "step": 820}, {"loss": 0.8078, "grad_norm": 0.42583167552948, "learning_rate": 0.0002, "epoch": 0.13418478700185918, "step": 830}, {"loss": 0.8219, "grad_norm": 0.4535622000694275, "learning_rate": 0.0002, "epoch": 0.13580147118260447, "step": 840}, {"loss": 0.8423, "grad_norm": 0.4116036891937256, "learning_rate": 0.0002, "epoch": 0.13741815536334978, "step": 850}, {"loss": 0.8466, "grad_norm": 0.45997580885887146, "learning_rate": 0.0002, "epoch": 0.13903483954409507, "step": 860}, {"loss": 0.8917, "grad_norm": 0.4487837255001068, "learning_rate": 0.0002, "epoch": 0.14065152372484035, "step": 870}, {"loss": 0.8217, "grad_norm": 0.43650057911872864, "learning_rate": 0.0002, "epoch": 0.14226820790558564, "step": 880}, {"loss": 0.8178, "grad_norm": 0.5335358381271362, "learning_rate": 0.0002, "epoch": 0.14388489208633093, "step": 890}, {"loss": 0.7957, "grad_norm": 0.5989000201225281, "learning_rate": 0.0002, "epoch": 0.14550157626707622, "step": 900}, {"loss": 0.8385, "grad_norm": 0.517179012298584, "learning_rate": 0.0002, "epoch": 0.14711826044782153, "step": 910}, {"loss": 0.8255, "grad_norm": 0.44435232877731323, "learning_rate": 0.0002, "epoch": 0.14873494462856682, "step": 920}, {"loss": 0.8305, "grad_norm": 0.42635923624038696, "learning_rate": 0.0002, "epoch": 0.1503516288093121, "step": 930}, {"loss": 0.8043, "grad_norm": 0.49603334069252014, "learning_rate": 0.0002, "epoch": 0.1519683129900574, "step": 940}, {"loss": 0.8377, "grad_norm": 0.40639808773994446, "learning_rate": 0.0002, "epoch": 0.15358499717080268, "step": 950}, {"loss": 0.8529, "grad_norm": 0.4850759208202362, "learning_rate": 0.0002, "epoch": 0.15520168135154797, "step": 960}, {"loss": 0.846, "grad_norm": 0.4427442252635956, "learning_rate": 0.0002, "epoch": 0.15681836553229328, "step": 970}, {"loss": 0.8705, "grad_norm": 0.3760930001735687, "learning_rate": 0.0002, "epoch": 0.15843504971303857, "step": 980}, {"loss": 0.8644, "grad_norm": 0.4794144332408905, "learning_rate": 0.0002, "epoch": 0.16005173389378385, "step": 990}, {"loss": 0.8002, "grad_norm": 0.45828768610954285, "learning_rate": 0.0002, "epoch": 0.16166841807452914, "step": 1000}, {"loss": 0.7658, "grad_norm": 0.6313053369522095, "learning_rate": 0.0002, "epoch": 0.16328510225527443, "step": 1010}, {"loss": 0.8047, "grad_norm": 0.45041006803512573, "learning_rate": 0.0002, "epoch": 0.16490178643601971, "step": 1020}, {"loss": 0.8423, "grad_norm": 0.441403865814209, "learning_rate": 0.0002, "epoch": 0.166518470616765, "step": 1030}, {"loss": 0.8475, "grad_norm": 0.8171296119689941, "learning_rate": 0.0002, "epoch": 0.16813515479751032, "step": 1040}, {"loss": 0.845, "grad_norm": 0.7137420773506165, "learning_rate": 0.0002, "epoch": 0.1697518389782556, "step": 1050}, {"loss": 0.8213, "grad_norm": 0.5236809849739075, "learning_rate": 0.0002, "epoch": 0.1713685231590009, "step": 1060}, {"loss": 0.8265, "grad_norm": 0.5021864175796509, "learning_rate": 0.0002, "epoch": 0.17298520733974618, "step": 1070}, {"loss": 0.8305, "grad_norm": 0.47347521781921387, "learning_rate": 0.0002, "epoch": 0.17460189152049146, "step": 1080}, {"loss": 0.8105, "grad_norm": 0.4631653428077698, "learning_rate": 0.0002, "epoch": 0.17621857570123675, "step": 1090}, {"loss": 0.8166, "grad_norm": 0.49169182777404785, "learning_rate": 0.0002, "epoch": 0.17783525988198207, "step": 1100}, {"loss": 0.8012, "grad_norm": 0.5019739270210266, "learning_rate": 0.0002, "epoch": 0.17945194406272735, "step": 1110}, {"loss": 0.8247, "grad_norm": 0.5100422501564026, "learning_rate": 0.0002, "epoch": 0.18106862824347264, "step": 1120}, {"loss": 0.8142, "grad_norm": 0.3888324499130249, "learning_rate": 0.0002, "epoch": 0.18268531242421793, "step": 1130}, {"loss": 0.8533, "grad_norm": 0.39765217900276184, "learning_rate": 0.0002, "epoch": 0.18430199660496321, "step": 1140}, {"loss": 0.8541, "grad_norm": 0.47190186381340027, "learning_rate": 0.0002, "epoch": 0.1859186807857085, "step": 1150}, {"loss": 0.8301, "grad_norm": 0.4464188814163208, "learning_rate": 0.0002, "epoch": 0.18753536496645382, "step": 1160}, {"loss": 0.8341, "grad_norm": 0.5153930187225342, "learning_rate": 0.0002, "epoch": 0.1891520491471991, "step": 1170}, {"loss": 0.8033, "grad_norm": 0.4779708683490753, "learning_rate": 0.0002, "epoch": 0.1907687333279444, "step": 1180}, {"loss": 0.8187, "grad_norm": 0.4834315776824951, "learning_rate": 0.0002, "epoch": 0.19238541750868968, "step": 1190}, {"loss": 0.7721, "grad_norm": 0.402357816696167, "learning_rate": 0.0002, "epoch": 0.19400210168943496, "step": 1200}, {"loss": 0.7941, "grad_norm": 0.45899084210395813, "learning_rate": 0.0002, "epoch": 0.19561878587018025, "step": 1210}, {"loss": 0.8353, "grad_norm": 0.5106529593467712, "learning_rate": 0.0002, "epoch": 0.19723547005092557, "step": 1220}, {"loss": 0.7816, "grad_norm": 0.45261722803115845, "learning_rate": 0.0002, "epoch": 0.19885215423167085, "step": 1230}, {"loss": 0.8068, "grad_norm": 0.4647127091884613, "learning_rate": 0.0002, "epoch": 0.20046883841241614, "step": 1240}, {"loss": 0.8239, "grad_norm": 0.4849368929862976, "learning_rate": 0.0002, "epoch": 0.20208552259316143, "step": 1250}, {"loss": 0.8514, "grad_norm": 0.4518061578273773, "learning_rate": 0.0002, "epoch": 0.2037022067739067, "step": 1260}, {"loss": 0.8158, "grad_norm": 0.49535325169563293, "learning_rate": 0.0002, "epoch": 0.205318890954652, "step": 1270}, {"loss": 0.8348, "grad_norm": 0.4835205376148224, "learning_rate": 0.0002, "epoch": 0.2069355751353973, "step": 1280}, {"loss": 0.8428, "grad_norm": 0.45308539271354675, "learning_rate": 0.0002, "epoch": 0.2085522593161426, "step": 1290}, {"loss": 0.7993, "grad_norm": 0.5369905233383179, "learning_rate": 0.0002, "epoch": 0.2101689434968879, "step": 1300}, {"loss": 0.8676, "grad_norm": 0.5031622052192688, "learning_rate": 0.0002, "epoch": 0.21178562767763318, "step": 1310}, {"loss": 0.7686, "grad_norm": 0.48010334372520447, "learning_rate": 0.0002, "epoch": 0.21340231185837846, "step": 1320}, {"loss": 0.806, "grad_norm": 0.4905701279640198, "learning_rate": 0.0002, "epoch": 0.21501899603912375, "step": 1330}, {"loss": 0.7885, "grad_norm": 0.43531742691993713, "learning_rate": 0.0002, "epoch": 0.21663568021986904, "step": 1340}, {"loss": 0.8191, "grad_norm": 0.44330692291259766, "learning_rate": 0.0002, "epoch": 0.21825236440061435, "step": 1350}, {"loss": 0.8205, "grad_norm": 0.5384416580200195, "learning_rate": 0.0002, "epoch": 0.21986904858135964, "step": 1360}, {"loss": 0.7726, "grad_norm": 0.4181833863258362, "learning_rate": 0.0002, "epoch": 0.22148573276210493, "step": 1370}, {"loss": 0.8311, "grad_norm": 0.523833692073822, "learning_rate": 0.0002, "epoch": 0.2231024169428502, "step": 1380}, {"loss": 0.7913, "grad_norm": 0.5528736710548401, "learning_rate": 0.0002, "epoch": 0.2247191011235955, "step": 1390}, {"loss": 0.8079, "grad_norm": 0.43515023589134216, "learning_rate": 0.0002, "epoch": 0.2263357853043408, "step": 1400}, {"loss": 0.8403, "grad_norm": 0.48809877038002014, "learning_rate": 0.0002, "epoch": 0.2279524694850861, "step": 1410}, {"loss": 0.8165, "grad_norm": 0.43591251969337463, "learning_rate": 0.0002, "epoch": 0.2295691536658314, "step": 1420}, {"loss": 0.8147, "grad_norm": 0.44625312089920044, "learning_rate": 0.0002, "epoch": 0.23118583784657668, "step": 1430}, {"loss": 0.8134, "grad_norm": 0.4390665292739868, "learning_rate": 0.0002, "epoch": 0.23280252202732196, "step": 1440}, {"loss": 0.8465, "grad_norm": 0.48496049642562866, "learning_rate": 0.0002, "epoch": 0.23441920620806725, "step": 1450}, {"loss": 0.775, "grad_norm": 0.45919957756996155, "learning_rate": 0.0002, "epoch": 0.23603589038881254, "step": 1460}, {"loss": 0.8659, "grad_norm": 0.5471845865249634, "learning_rate": 0.0002, "epoch": 0.23765257456955785, "step": 1470}, {"loss": 0.8164, "grad_norm": 0.47269317507743835, "learning_rate": 0.0002, "epoch": 0.23926925875030314, "step": 1480}, {"loss": 0.854, "grad_norm": 0.4930245578289032, "learning_rate": 0.0002, "epoch": 0.24088594293104842, "step": 1490}, {"loss": 0.8139, "grad_norm": 0.5605630278587341, "learning_rate": 0.0002, "epoch": 0.2425026271117937, "step": 1500}, {"loss": 0.8125, "grad_norm": 0.4435870945453644, "learning_rate": 0.0002, "epoch": 0.244119311292539, "step": 1510}, {"loss": 0.8123, "grad_norm": 0.4941999912261963, "learning_rate": 0.0002, "epoch": 0.24573599547328429, "step": 1520}, {"loss": 0.8427, "grad_norm": 0.5100624561309814, "learning_rate": 0.0002, "epoch": 0.24735267965402957, "step": 1530}, {"loss": 0.8405, "grad_norm": 0.4638267457485199, "learning_rate": 0.0002, "epoch": 0.2489693638347749, "step": 1540}, {"loss": 0.81, "grad_norm": 0.5071570873260498, "learning_rate": 0.0002, "epoch": 0.25058604801552015, "step": 1550}, {"loss": 0.7724, "grad_norm": 0.4291319251060486, "learning_rate": 0.0002, "epoch": 0.25220273219626543, "step": 1560}, {"loss": 0.7984, "grad_norm": 0.5388049483299255, "learning_rate": 0.0002, "epoch": 0.2538194163770108, "step": 1570}, {"loss": 0.8176, "grad_norm": 0.5083683729171753, "learning_rate": 0.0002, "epoch": 0.25543610055775606, "step": 1580}, {"loss": 0.843, "grad_norm": 0.4824463725090027, "learning_rate": 0.0002, "epoch": 0.25705278473850135, "step": 1590}, {"loss": 0.7996, "grad_norm": 0.41177722811698914, "learning_rate": 0.0002, "epoch": 0.25866946891924664, "step": 1600}, {"loss": 0.7772, "grad_norm": 0.5656219124794006, "learning_rate": 0.0002, "epoch": 0.2602861530999919, "step": 1610}, {"loss": 0.7955, "grad_norm": 0.41063204407691956, "learning_rate": 0.0002, "epoch": 0.2619028372807372, "step": 1620}, {"loss": 0.7998, "grad_norm": 0.4897061288356781, "learning_rate": 0.0002, "epoch": 0.2635195214614825, "step": 1630}, {"loss": 0.8198, "grad_norm": 0.4454376697540283, "learning_rate": 0.0002, "epoch": 0.2651362056422278, "step": 1640}, {"loss": 0.8684, "grad_norm": 0.4355238378047943, "learning_rate": 0.0002, "epoch": 0.26675288982297307, "step": 1650}, {"loss": 0.7801, "grad_norm": 0.458310067653656, "learning_rate": 0.0002, "epoch": 0.26836957400371836, "step": 1660}, {"loss": 0.7935, "grad_norm": 0.4752083718776703, "learning_rate": 0.0002, "epoch": 0.26998625818446365, "step": 1670}, {"loss": 0.8267, "grad_norm": 0.4666106402873993, "learning_rate": 0.0002, "epoch": 0.27160294236520893, "step": 1680}, {"loss": 0.8252, "grad_norm": 0.4213818609714508, "learning_rate": 0.0002, "epoch": 0.2732196265459543, "step": 1690}, {"loss": 0.8559, "grad_norm": 0.5768913626670837, "learning_rate": 0.0002, "epoch": 0.27483631072669956, "step": 1700}, {"loss": 0.7931, "grad_norm": 0.4209914803504944, "learning_rate": 0.0002, "epoch": 0.27645299490744485, "step": 1710}, {"loss": 0.8167, "grad_norm": 0.501909613609314, "learning_rate": 0.0002, "epoch": 0.27806967908819014, "step": 1720}, {"loss": 0.7832, "grad_norm": 0.5266261100769043, "learning_rate": 0.0002, "epoch": 0.2796863632689354, "step": 1730}, {"loss": 0.8102, "grad_norm": 0.43806859850883484, "learning_rate": 0.0002, "epoch": 0.2813030474496807, "step": 1740}, {"loss": 0.8157, "grad_norm": 0.46048814058303833, "learning_rate": 0.0002, "epoch": 0.282919731630426, "step": 1750}, {"loss": 0.8596, "grad_norm": 0.44972819089889526, "learning_rate": 0.0002, "epoch": 0.2845364158111713, "step": 1760}, {"loss": 0.8421, "grad_norm": 0.5114831328392029, "learning_rate": 0.0002, "epoch": 0.28615309999191657, "step": 1770}, {"loss": 0.8361, "grad_norm": 0.47931742668151855, "learning_rate": 0.0002, "epoch": 0.28776978417266186, "step": 1780}, {"loss": 0.8265, "grad_norm": 0.5092599987983704, "learning_rate": 0.0002, "epoch": 0.28938646835340714, "step": 1790}, {"loss": 0.8506, "grad_norm": 0.37581443786621094, "learning_rate": 0.0002, "epoch": 0.29100315253415243, "step": 1800}, {"loss": 0.7932, "grad_norm": 0.47097381949424744, "learning_rate": 0.0002, "epoch": 0.2926198367148977, "step": 1810}, {"loss": 0.7787, "grad_norm": 0.48300236463546753, "learning_rate": 0.0002, "epoch": 0.29423652089564306, "step": 1820}, {"loss": 0.8391, "grad_norm": 0.5600419640541077, "learning_rate": 0.0002, "epoch": 0.29585320507638835, "step": 1830}, {"loss": 0.8507, "grad_norm": 0.48555272817611694, "learning_rate": 0.0002, "epoch": 0.29746988925713364, "step": 1840}, {"loss": 0.7657, "grad_norm": 0.3752668499946594, "learning_rate": 0.0002, "epoch": 0.2990865734378789, "step": 1850}, {"loss": 0.7915, "grad_norm": 0.5328747034072876, "learning_rate": 0.0002, "epoch": 0.3007032576186242, "step": 1860}, {"loss": 0.8426, "grad_norm": 0.48716455698013306, "learning_rate": 0.0002, "epoch": 0.3023199417993695, "step": 1870}, {"loss": 0.8335, "grad_norm": 0.5011493563652039, "learning_rate": 0.0002, "epoch": 0.3039366259801148, "step": 1880}, {"loss": 0.852, "grad_norm": 0.46461427211761475, "learning_rate": 0.0002, "epoch": 0.30555331016086007, "step": 1890}, {"loss": 0.8478, "grad_norm": 0.36630210280418396, "learning_rate": 0.0002, "epoch": 0.30716999434160536, "step": 1900}, {"loss": 0.8162, "grad_norm": 0.4217296242713928, "learning_rate": 0.0002, "epoch": 0.30878667852235064, "step": 1910}, {"loss": 0.8128, "grad_norm": 0.4394875466823578, "learning_rate": 0.0002, "epoch": 0.31040336270309593, "step": 1920}, {"loss": 0.8471, "grad_norm": 0.6587965488433838, "learning_rate": 0.0002, "epoch": 0.3120200468838412, "step": 1930}, {"loss": 0.8565, "grad_norm": 0.5469298958778381, "learning_rate": 0.0002, "epoch": 0.31363673106458656, "step": 1940}, {"loss": 0.8236, "grad_norm": 0.4371595084667206, "learning_rate": 0.0002, "epoch": 0.31525341524533185, "step": 1950}, {"loss": 0.887, "grad_norm": 0.4809541404247284, "learning_rate": 0.0002, "epoch": 0.31687009942607713, "step": 1960}, {"loss": 0.7855, "grad_norm": 0.6061086654663086, "learning_rate": 0.0002, "epoch": 0.3184867836068224, "step": 1970}, {"loss": 0.7679, "grad_norm": 0.5342657566070557, "learning_rate": 0.0002, "epoch": 0.3201034677875677, "step": 1980}, {"loss": 0.7955, "grad_norm": 0.5057743787765503, "learning_rate": 0.0002, "epoch": 0.321720151968313, "step": 1990}, {"loss": 0.7774, "grad_norm": 0.528626024723053, "learning_rate": 0.0002, "epoch": 0.3233368361490583, "step": 2000}, {"loss": 0.8845, "grad_norm": 0.46742770075798035, "learning_rate": 0.0002, "epoch": 0.32495352032980357, "step": 2010}, {"loss": 0.8484, "grad_norm": 0.515101432800293, "learning_rate": 0.0002, "epoch": 0.32657020451054886, "step": 2020}, {"loss": 0.8139, "grad_norm": 0.41941216588020325, "learning_rate": 0.0002, "epoch": 0.32818688869129414, "step": 2030}, {"loss": 0.7637, "grad_norm": 0.49902522563934326, "learning_rate": 0.0002, "epoch": 0.32980357287203943, "step": 2040}, {"loss": 0.7822, "grad_norm": 0.4120897650718689, "learning_rate": 0.0002, "epoch": 0.3314202570527847, "step": 2050}, {"loss": 0.8057, "grad_norm": 0.45352041721343994, "learning_rate": 0.0002, "epoch": 0.33303694123353, "step": 2060}, {"loss": 0.7913, "grad_norm": 0.523199737071991, "learning_rate": 0.0002, "epoch": 0.33465362541427535, "step": 2070}, {"loss": 0.8036, "grad_norm": 0.4390358626842499, "learning_rate": 0.0002, "epoch": 0.33627030959502063, "step": 2080}, {"loss": 0.8145, "grad_norm": 0.6752901077270508, "learning_rate": 0.0002, "epoch": 0.3378869937757659, "step": 2090}, {"loss": 0.7807, "grad_norm": 0.547821044921875, "learning_rate": 0.0002, "epoch": 0.3395036779565112, "step": 2100}, {"loss": 0.8561, "grad_norm": 0.5161308646202087, "learning_rate": 0.0002, "epoch": 0.3411203621372565, "step": 2110}, {"loss": 0.7697, "grad_norm": 0.4565401077270508, "learning_rate": 0.0002, "epoch": 0.3427370463180018, "step": 2120}, {"loss": 0.7964, "grad_norm": 0.4666115939617157, "learning_rate": 0.0002, "epoch": 0.34435373049874707, "step": 2130}, {"loss": 0.8189, "grad_norm": 0.4090428352355957, "learning_rate": 0.0002, "epoch": 0.34597041467949236, "step": 2140}, {"loss": 0.8817, "grad_norm": 0.510845422744751, "learning_rate": 0.0002, "epoch": 0.34758709886023764, "step": 2150}, {"loss": 0.8398, "grad_norm": 0.42861923575401306, "learning_rate": 0.0002, "epoch": 0.34920378304098293, "step": 2160}, {"loss": 0.7716, "grad_norm": 0.4476332664489746, "learning_rate": 0.0002, "epoch": 0.3508204672217282, "step": 2170}, {"loss": 0.7845, "grad_norm": 0.6065791249275208, "learning_rate": 0.0002, "epoch": 0.3524371514024735, "step": 2180}, {"loss": 0.8187, "grad_norm": 0.42335066199302673, "learning_rate": 0.0002, "epoch": 0.35405383558321885, "step": 2190}, {"loss": 0.8239, "grad_norm": 0.5094629526138306, "learning_rate": 0.0002, "epoch": 0.35567051976396413, "step": 2200}, {"loss": 0.7807, "grad_norm": 0.5476373434066772, "learning_rate": 0.0002, "epoch": 0.3572872039447094, "step": 2210}, {"loss": 0.814, "grad_norm": 0.3911719024181366, "learning_rate": 0.0002, "epoch": 0.3589038881254547, "step": 2220}, {"loss": 0.8599, "grad_norm": 0.6599636077880859, "learning_rate": 0.0002, "epoch": 0.3605205723062, "step": 2230}, {"loss": 0.7482, "grad_norm": 0.40381914377212524, "learning_rate": 0.0002, "epoch": 0.3621372564869453, "step": 2240}, {"loss": 0.7772, "grad_norm": 0.4433908462524414, "learning_rate": 0.0002, "epoch": 0.36375394066769057, "step": 2250}, {"loss": 0.8503, "grad_norm": 0.578326940536499, "learning_rate": 0.0002, "epoch": 0.36537062484843585, "step": 2260}, {"loss": 0.8178, "grad_norm": 0.5734784007072449, "learning_rate": 0.0002, "epoch": 0.36698730902918114, "step": 2270}, {"loss": 0.8193, "grad_norm": 0.45555487275123596, "learning_rate": 0.0002, "epoch": 0.36860399320992643, "step": 2280}, {"loss": 0.7929, "grad_norm": 0.5666276216506958, "learning_rate": 0.0002, "epoch": 0.3702206773906717, "step": 2290}, {"loss": 0.8292, "grad_norm": 0.5461117625236511, "learning_rate": 0.0002, "epoch": 0.371837361571417, "step": 2300}, {"loss": 0.8204, "grad_norm": 0.6318911910057068, "learning_rate": 0.0002, "epoch": 0.3734540457521623, "step": 2310}, {"loss": 0.7964, "grad_norm": 0.493263304233551, "learning_rate": 0.0002, "epoch": 0.37507072993290763, "step": 2320}, {"loss": 0.8339, "grad_norm": 0.5888760089874268, "learning_rate": 0.0002, "epoch": 0.3766874141136529, "step": 2330}, {"loss": 0.7737, "grad_norm": 0.48671841621398926, "learning_rate": 0.0002, "epoch": 0.3783040982943982, "step": 2340}, {"loss": 0.8367, "grad_norm": 0.4385145306587219, "learning_rate": 0.0002, "epoch": 0.3799207824751435, "step": 2350}, {"loss": 0.812, "grad_norm": 0.5523318648338318, "learning_rate": 0.0002, "epoch": 0.3815374666558888, "step": 2360}, {"loss": 0.8351, "grad_norm": 0.7308220267295837, "learning_rate": 0.0002, "epoch": 0.38315415083663407, "step": 2370}, {"loss": 0.859, "grad_norm": 0.554214358329773, "learning_rate": 0.0002, "epoch": 0.38477083501737935, "step": 2380}, {"loss": 0.8146, "grad_norm": 0.5425800085067749, "learning_rate": 0.0002, "epoch": 0.38638751919812464, "step": 2390}, {"loss": 0.8282, "grad_norm": 0.48811158537864685, "learning_rate": 0.0002, "epoch": 0.3880042033788699, "step": 2400}, {"loss": 0.8074, "grad_norm": 0.49212366342544556, "learning_rate": 0.0002, "epoch": 0.3896208875596152, "step": 2410}, {"loss": 0.7991, "grad_norm": 0.5222218632698059, "learning_rate": 0.0002, "epoch": 0.3912375717403605, "step": 2420}, {"loss": 0.8182, "grad_norm": 0.4699819087982178, "learning_rate": 0.0002, "epoch": 0.3928542559211058, "step": 2430}, {"loss": 0.7919, "grad_norm": 0.46153587102890015, "learning_rate": 0.0002, "epoch": 0.39447094010185113, "step": 2440}, {"loss": 0.8111, "grad_norm": 0.4150611162185669, "learning_rate": 0.0002, "epoch": 0.3960876242825964, "step": 2450}, {"loss": 0.8589, "grad_norm": 0.5799614787101746, "learning_rate": 0.0002, "epoch": 0.3977043084633417, "step": 2460}, {"loss": 0.8085, "grad_norm": 0.56536865234375, "learning_rate": 0.0002, "epoch": 0.399320992644087, "step": 2470}, {"loss": 0.8022, "grad_norm": 0.5451247096061707, "learning_rate": 0.0002, "epoch": 0.4009376768248323, "step": 2480}, {"loss": 0.8217, "grad_norm": 0.5914521217346191, "learning_rate": 0.0002, "epoch": 0.40255436100557757, "step": 2490}, {"loss": 0.7859, "grad_norm": 0.4428117275238037, "learning_rate": 0.0002, "epoch": 0.40417104518632285, "step": 2500}, {"loss": 0.8054, "grad_norm": 0.48580947518348694, "learning_rate": 0.0002, "epoch": 0.40578772936706814, "step": 2510}, {"loss": 0.8405, "grad_norm": 0.436734676361084, "learning_rate": 0.0002, "epoch": 0.4074044135478134, "step": 2520}, {"loss": 0.8209, "grad_norm": 0.5752223134040833, "learning_rate": 0.0002, "epoch": 0.4090210977285587, "step": 2530}, {"loss": 0.8181, "grad_norm": 0.4271308183670044, "learning_rate": 0.0002, "epoch": 0.410637781909304, "step": 2540}, {"loss": 0.8058, "grad_norm": 0.46294718980789185, "learning_rate": 0.0002, "epoch": 0.4122544660900493, "step": 2550}, {"loss": 0.8473, "grad_norm": 0.49407583475112915, "learning_rate": 0.0002, "epoch": 0.4138711502707946, "step": 2560}, {"loss": 0.7881, "grad_norm": 0.4729035496711731, "learning_rate": 0.0002, "epoch": 0.4154878344515399, "step": 2570}, {"loss": 0.7834, "grad_norm": 0.4129747152328491, "learning_rate": 0.0002, "epoch": 0.4171045186322852, "step": 2580}, {"loss": 0.7859, "grad_norm": 0.5684236288070679, "learning_rate": 0.0002, "epoch": 0.4187212028130305, "step": 2590}, {"loss": 0.811, "grad_norm": 0.4862157106399536, "learning_rate": 0.0002, "epoch": 0.4203378869937758, "step": 2600}, {"loss": 0.7582, "grad_norm": 0.46567976474761963, "learning_rate": 0.0002, "epoch": 0.42195457117452106, "step": 2610}, {"loss": 0.7755, "grad_norm": 0.5710650682449341, "learning_rate": 0.0002, "epoch": 0.42357125535526635, "step": 2620}, {"loss": 0.8573, "grad_norm": 0.5660041570663452, "learning_rate": 0.0002, "epoch": 0.42518793953601164, "step": 2630}, {"loss": 0.7812, "grad_norm": 0.47944375872612, "learning_rate": 0.0002, "epoch": 0.4268046237167569, "step": 2640}, {"loss": 0.7459, "grad_norm": 0.537223756313324, "learning_rate": 0.0002, "epoch": 0.4284213078975022, "step": 2650}, {"loss": 0.8246, "grad_norm": 0.41669997572898865, "learning_rate": 0.0002, "epoch": 0.4300379920782475, "step": 2660}, {"loss": 0.7785, "grad_norm": 0.44727686047554016, "learning_rate": 0.0002, "epoch": 0.4316546762589928, "step": 2670}, {"loss": 0.8241, "grad_norm": 0.5600888729095459, "learning_rate": 0.0002, "epoch": 0.4332713604397381, "step": 2680}, {"loss": 0.7708, "grad_norm": 0.39820605516433716, "learning_rate": 0.0002, "epoch": 0.4348880446204834, "step": 2690}, {"loss": 0.8202, "grad_norm": 0.5637655854225159, "learning_rate": 0.0002, "epoch": 0.4365047288012287, "step": 2700}, {"loss": 0.855, "grad_norm": 0.6363666653633118, "learning_rate": 0.0002, "epoch": 0.438121412981974, "step": 2710}, {"loss": 0.8468, "grad_norm": 0.5656129121780396, "learning_rate": 0.0002, "epoch": 0.4397380971627193, "step": 2720}, {"loss": 0.7845, "grad_norm": 0.5600156188011169, "learning_rate": 0.0002, "epoch": 0.44135478134346456, "step": 2730}, {"loss": 0.8405, "grad_norm": 0.5506579875946045, "learning_rate": 0.0002, "epoch": 0.44297146552420985, "step": 2740}, {"loss": 0.7725, "grad_norm": 0.49878305196762085, "learning_rate": 0.0002, "epoch": 0.44458814970495514, "step": 2750}, {"loss": 0.8292, "grad_norm": 0.4569213092327118, "learning_rate": 0.0002, "epoch": 0.4462048338857004, "step": 2760}, {"loss": 0.8028, "grad_norm": 0.6056680083274841, "learning_rate": 0.0002, "epoch": 0.4478215180664457, "step": 2770}, {"loss": 0.8242, "grad_norm": 0.44474557042121887, "learning_rate": 0.0002, "epoch": 0.449438202247191, "step": 2780}, {"loss": 0.801, "grad_norm": 0.46055394411087036, "learning_rate": 0.0002, "epoch": 0.4510548864279363, "step": 2790}, {"loss": 0.7521, "grad_norm": 0.4904133379459381, "learning_rate": 0.0002, "epoch": 0.4526715706086816, "step": 2800}, {"loss": 0.8829, "grad_norm": 0.5647031664848328, "learning_rate": 0.0002, "epoch": 0.45428825478942686, "step": 2810}, {"loss": 0.8622, "grad_norm": 0.5759473443031311, "learning_rate": 0.0002, "epoch": 0.4559049389701722, "step": 2820}, {"loss": 0.7812, "grad_norm": 0.5161895751953125, "learning_rate": 0.0002, "epoch": 0.4575216231509175, "step": 2830}, {"loss": 0.8045, "grad_norm": 0.4248254597187042, "learning_rate": 0.0002, "epoch": 0.4591383073316628, "step": 2840}, {"loss": 0.7838, "grad_norm": 0.45395001769065857, "learning_rate": 0.0002, "epoch": 0.46075499151240806, "step": 2850}, {"loss": 0.8208, "grad_norm": 0.5358697772026062, "learning_rate": 0.0002, "epoch": 0.46237167569315335, "step": 2860}, {"loss": 0.8147, "grad_norm": 0.5379165410995483, "learning_rate": 0.0002, "epoch": 0.46398835987389864, "step": 2870}, {"loss": 0.7403, "grad_norm": 0.4601989686489105, "learning_rate": 0.0002, "epoch": 0.4656050440546439, "step": 2880}, {"loss": 0.8523, "grad_norm": 0.671115517616272, "learning_rate": 0.0002, "epoch": 0.4672217282353892, "step": 2890}, {"loss": 0.8262, "grad_norm": 0.4425133168697357, "learning_rate": 0.0002, "epoch": 0.4688384124161345, "step": 2900}, {"loss": 0.8178, "grad_norm": 0.5446155071258545, "learning_rate": 0.0002, "epoch": 0.4704550965968798, "step": 2910}, {"loss": 0.8106, "grad_norm": 0.603306233882904, "learning_rate": 0.0002, "epoch": 0.47207178077762507, "step": 2920}, {"loss": 0.8044, "grad_norm": 0.5377997159957886, "learning_rate": 0.0002, "epoch": 0.47368846495837036, "step": 2930}, {"loss": 0.8075, "grad_norm": 0.4931027591228485, "learning_rate": 0.0002, "epoch": 0.4753051491391157, "step": 2940}, {"loss": 0.8004, "grad_norm": 0.4711960256099701, "learning_rate": 0.0002, "epoch": 0.476921833319861, "step": 2950}, {"loss": 0.8121, "grad_norm": 0.5020492672920227, "learning_rate": 0.0002, "epoch": 0.4785385175006063, "step": 2960}, {"loss": 0.8221, "grad_norm": 0.5428946614265442, "learning_rate": 0.0002, "epoch": 0.48015520168135156, "step": 2970}, {"loss": 0.7849, "grad_norm": 0.5294089317321777, "learning_rate": 0.0002, "epoch": 0.48177188586209685, "step": 2980}, {"loss": 0.8553, "grad_norm": 0.648289144039154, "learning_rate": 0.0002, "epoch": 0.48338857004284214, "step": 2990}, {"loss": 0.7874, "grad_norm": 0.47916680574417114, "learning_rate": 0.0002, "epoch": 0.4850052542235874, "step": 3000}, {"loss": 0.8087, "grad_norm": 0.43849772214889526, "learning_rate": 0.0002, "epoch": 0.4866219384043327, "step": 3010}, {"loss": 0.7662, "grad_norm": 0.47007861733436584, "learning_rate": 0.0002, "epoch": 0.488238622585078, "step": 3020}, {"loss": 0.757, "grad_norm": 0.6314331293106079, "learning_rate": 0.0002, "epoch": 0.4898553067658233, "step": 3030}, {"loss": 0.7863, "grad_norm": 0.49211493134498596, "learning_rate": 0.0002, "epoch": 0.49147199094656857, "step": 3040}, {"loss": 0.8335, "grad_norm": 0.4537973403930664, "learning_rate": 0.0002, "epoch": 0.49308867512731386, "step": 3050}, {"loss": 0.8095, "grad_norm": 0.47326919436454773, "learning_rate": 0.0002, "epoch": 0.49470535930805914, "step": 3060}, {"loss": 0.8447, "grad_norm": 0.525874137878418, "learning_rate": 0.0002, "epoch": 0.4963220434888045, "step": 3070}, {"loss": 0.8339, "grad_norm": 0.6361091732978821, "learning_rate": 0.0002, "epoch": 0.4979387276695498, "step": 3080}, {"loss": 0.821, "grad_norm": 0.5850642919540405, "learning_rate": 0.0002, "epoch": 0.49955541185029506, "step": 3090}, {"loss": 0.8279, "grad_norm": 0.47299543023109436, "learning_rate": 0.0002, "epoch": 0.5011720960310403, "step": 3100}, {"loss": 0.8681, "grad_norm": 0.473099946975708, "learning_rate": 0.0002, "epoch": 0.5027887802117856, "step": 3110}, {"loss": 0.8223, "grad_norm": 0.48186397552490234, "learning_rate": 0.0002, "epoch": 0.5044054643925309, "step": 3120}, {"loss": 0.8292, "grad_norm": 0.5015401840209961, "learning_rate": 0.0002, "epoch": 0.5060221485732762, "step": 3130}, {"loss": 0.7692, "grad_norm": 0.5617750287055969, "learning_rate": 0.0002, "epoch": 0.5076388327540216, "step": 3140}, {"loss": 0.8708, "grad_norm": 0.5169327259063721, "learning_rate": 0.0002, "epoch": 0.5092555169347668, "step": 3150}, {"loss": 0.7845, "grad_norm": 0.545657753944397, "learning_rate": 0.0002, "epoch": 0.5108722011155121, "step": 3160}, {"loss": 0.799, "grad_norm": 0.512864351272583, "learning_rate": 0.0002, "epoch": 0.5124888852962574, "step": 3170}, {"loss": 0.7794, "grad_norm": 0.4113546311855316, "learning_rate": 0.0002, "epoch": 0.5141055694770027, "step": 3180}, {"loss": 0.8206, "grad_norm": 0.44532445073127747, "learning_rate": 0.0002, "epoch": 0.5157222536577479, "step": 3190}, {"loss": 0.8213, "grad_norm": 0.5623497366905212, "learning_rate": 0.0002, "epoch": 0.5173389378384933, "step": 3200}, {"loss": 0.7928, "grad_norm": 0.5084741115570068, "learning_rate": 0.0002, "epoch": 0.5189556220192385, "step": 3210}, {"loss": 0.8174, "grad_norm": 0.5305403470993042, "learning_rate": 0.0002, "epoch": 0.5205723061999838, "step": 3220}, {"loss": 0.8139, "grad_norm": 0.4708254337310791, "learning_rate": 0.0002, "epoch": 0.5221889903807291, "step": 3230}, {"loss": 0.7639, "grad_norm": 0.43827131390571594, "learning_rate": 0.0002, "epoch": 0.5238056745614744, "step": 3240}, {"loss": 0.7993, "grad_norm": 0.5630002617835999, "learning_rate": 0.0002, "epoch": 0.5254223587422197, "step": 3250}, {"loss": 0.7522, "grad_norm": 0.5010961890220642, "learning_rate": 0.0002, "epoch": 0.527039042922965, "step": 3260}, {"loss": 0.8374, "grad_norm": 0.6303122043609619, "learning_rate": 0.0002, "epoch": 0.5286557271037103, "step": 3270}, {"loss": 0.7727, "grad_norm": 0.5107331275939941, "learning_rate": 0.0002, "epoch": 0.5302724112844556, "step": 3280}, {"loss": 0.8495, "grad_norm": 0.5700443387031555, "learning_rate": 0.0002, "epoch": 0.5318890954652009, "step": 3290}, {"loss": 0.7776, "grad_norm": 0.46296367049217224, "learning_rate": 0.0002, "epoch": 0.5335057796459461, "step": 3300}, {"loss": 0.7931, "grad_norm": 0.531568706035614, "learning_rate": 0.0002, "epoch": 0.5351224638266915, "step": 3310}, {"loss": 0.843, "grad_norm": 0.4686741530895233, "learning_rate": 0.0002, "epoch": 0.5367391480074367, "step": 3320}, {"loss": 0.8104, "grad_norm": 0.5404331088066101, "learning_rate": 0.0002, "epoch": 0.5383558321881821, "step": 3330}, {"loss": 0.7686, "grad_norm": 0.6368790864944458, "learning_rate": 0.0002, "epoch": 0.5399725163689273, "step": 3340}, {"loss": 0.8514, "grad_norm": 0.42300888895988464, "learning_rate": 0.0002, "epoch": 0.5415892005496726, "step": 3350}, {"loss": 0.8236, "grad_norm": 0.5362542867660522, "learning_rate": 0.0002, "epoch": 0.5432058847304179, "step": 3360}, {"loss": 0.858, "grad_norm": 0.497128963470459, "learning_rate": 0.0002, "epoch": 0.5448225689111632, "step": 3370}, {"loss": 0.8519, "grad_norm": 0.5006386041641235, "learning_rate": 0.0002, "epoch": 0.5464392530919085, "step": 3380}, {"loss": 0.7867, "grad_norm": 0.44136837124824524, "learning_rate": 0.0002, "epoch": 0.5480559372726538, "step": 3390}, {"loss": 0.773, "grad_norm": 0.5897833108901978, "learning_rate": 0.0002, "epoch": 0.5496726214533991, "step": 3400}, {"loss": 0.8895, "grad_norm": 0.641075611114502, "learning_rate": 0.0002, "epoch": 0.5512893056341444, "step": 3410}, {"loss": 0.7827, "grad_norm": 0.7251322269439697, "learning_rate": 0.0002, "epoch": 0.5529059898148897, "step": 3420}, {"loss": 0.7626, "grad_norm": 0.47411349415779114, "learning_rate": 0.0002, "epoch": 0.5545226739956349, "step": 3430}, {"loss": 0.8196, "grad_norm": 0.4994310438632965, "learning_rate": 0.0002, "epoch": 0.5561393581763803, "step": 3440}, {"loss": 0.7812, "grad_norm": 0.5814438462257385, "learning_rate": 0.0002, "epoch": 0.5577560423571255, "step": 3450}, {"loss": 0.8805, "grad_norm": 0.6278898119926453, "learning_rate": 0.0002, "epoch": 0.5593727265378708, "step": 3460}, {"loss": 0.813, "grad_norm": 0.46208274364471436, "learning_rate": 0.0002, "epoch": 0.5609894107186161, "step": 3470}, {"loss": 0.8295, "grad_norm": 0.5718930959701538, "learning_rate": 0.0002, "epoch": 0.5626060948993614, "step": 3480}, {"loss": 0.8152, "grad_norm": 0.48178744316101074, "learning_rate": 0.0002, "epoch": 0.5642227790801067, "step": 3490}, {"loss": 0.8244, "grad_norm": 0.47336965799331665, "learning_rate": 0.0002, "epoch": 0.565839463260852, "step": 3500}, {"loss": 0.8099, "grad_norm": 0.43442684412002563, "learning_rate": 0.0002, "epoch": 0.5674561474415973, "step": 3510}, {"loss": 0.7564, "grad_norm": 0.6463358998298645, "learning_rate": 0.0002, "epoch": 0.5690728316223426, "step": 3520}, {"loss": 0.836, "grad_norm": 0.5286486744880676, "learning_rate": 0.0002, "epoch": 0.5706895158030879, "step": 3530}, {"loss": 0.8421, "grad_norm": 0.5405499935150146, "learning_rate": 0.0002, "epoch": 0.5723061999838331, "step": 3540}, {"loss": 0.7614, "grad_norm": 0.6654391884803772, "learning_rate": 0.0002, "epoch": 0.5739228841645785, "step": 3550}, {"loss": 0.7803, "grad_norm": 0.5081980228424072, "learning_rate": 0.0002, "epoch": 0.5755395683453237, "step": 3560}, {"loss": 0.7753, "grad_norm": 0.48978179693222046, "learning_rate": 0.0002, "epoch": 0.5771562525260691, "step": 3570}, {"loss": 0.8151, "grad_norm": 0.5840612053871155, "learning_rate": 0.0002, "epoch": 0.5787729367068143, "step": 3580}, {"loss": 0.8937, "grad_norm": 0.5235261917114258, "learning_rate": 0.0002, "epoch": 0.5803896208875596, "step": 3590}, {"loss": 0.7894, "grad_norm": 0.5672075748443604, "learning_rate": 0.0002, "epoch": 0.5820063050683049, "step": 3600}, {"loss": 0.8347, "grad_norm": 0.5613429546356201, "learning_rate": 0.0002, "epoch": 0.5836229892490502, "step": 3610}, {"loss": 0.8274, "grad_norm": 0.4032273590564728, "learning_rate": 0.0002, "epoch": 0.5852396734297954, "step": 3620}, {"loss": 0.8421, "grad_norm": 0.49559324979782104, "learning_rate": 0.0002, "epoch": 0.5868563576105408, "step": 3630}, {"loss": 0.8332, "grad_norm": 0.6895697712898254, "learning_rate": 0.0002, "epoch": 0.5884730417912861, "step": 3640}, {"loss": 0.7877, "grad_norm": 0.4750136435031891, "learning_rate": 0.0002, "epoch": 0.5900897259720314, "step": 3650}, {"loss": 0.8219, "grad_norm": 0.5176819562911987, "learning_rate": 0.0002, "epoch": 0.5917064101527767, "step": 3660}, {"loss": 0.8151, "grad_norm": 0.5817760229110718, "learning_rate": 0.0002, "epoch": 0.5933230943335219, "step": 3670}, {"loss": 0.7823, "grad_norm": 0.6064626574516296, "learning_rate": 0.0002, "epoch": 0.5949397785142673, "step": 3680}, {"loss": 0.8422, "grad_norm": 0.6728700995445251, "learning_rate": 0.0002, "epoch": 0.5965564626950125, "step": 3690}, {"loss": 0.7679, "grad_norm": 0.609305202960968, "learning_rate": 0.0002, "epoch": 0.5981731468757578, "step": 3700}, {"loss": 0.8048, "grad_norm": 0.4615488350391388, "learning_rate": 0.0002, "epoch": 0.5997898310565031, "step": 3710}, {"loss": 0.8214, "grad_norm": 2.0531179904937744, "learning_rate": 0.0002, "epoch": 0.6014065152372484, "step": 3720}, {"loss": 0.8158, "grad_norm": 0.5091132521629333, "learning_rate": 0.0002, "epoch": 0.6030231994179936, "step": 3730}, {"loss": 0.7833, "grad_norm": 0.5951124429702759, "learning_rate": 0.0002, "epoch": 0.604639883598739, "step": 3740}, {"loss": 0.7784, "grad_norm": 0.5870208144187927, "learning_rate": 0.0002, "epoch": 0.6062565677794842, "step": 3750}, {"loss": 0.8044, "grad_norm": 0.6254619359970093, "learning_rate": 0.0002, "epoch": 0.6078732519602296, "step": 3760}, {"loss": 0.7868, "grad_norm": 0.5577626824378967, "learning_rate": 0.0002, "epoch": 0.6094899361409749, "step": 3770}, {"loss": 0.8108, "grad_norm": 0.5004405379295349, "learning_rate": 0.0002, "epoch": 0.6111066203217201, "step": 3780}, {"loss": 0.8092, "grad_norm": 0.5527383685112, "learning_rate": 0.0002, "epoch": 0.6127233045024655, "step": 3790}, {"loss": 0.8036, "grad_norm": 0.49116113781929016, "learning_rate": 0.0002, "epoch": 0.6143399886832107, "step": 3800}, {"loss": 0.8352, "grad_norm": 0.5299299359321594, "learning_rate": 0.0002, "epoch": 0.6159566728639561, "step": 3810}, {"loss": 0.7737, "grad_norm": 0.464897483587265, "learning_rate": 0.0002, "epoch": 0.6175733570447013, "step": 3820}, {"loss": 0.7923, "grad_norm": 0.6505740880966187, "learning_rate": 0.0002, "epoch": 0.6191900412254466, "step": 3830}, {"loss": 0.8123, "grad_norm": 0.5512559413909912, "learning_rate": 0.0002, "epoch": 0.6208067254061919, "step": 3840}, {"loss": 0.8856, "grad_norm": 0.49427518248558044, "learning_rate": 0.0002, "epoch": 0.6224234095869372, "step": 3850}, {"loss": 0.7751, "grad_norm": 0.3839147090911865, "learning_rate": 0.0002, "epoch": 0.6240400937676824, "step": 3860}, {"loss": 0.8006, "grad_norm": 0.5760218501091003, "learning_rate": 0.0002, "epoch": 0.6256567779484278, "step": 3870}, {"loss": 0.7836, "grad_norm": 0.7226507067680359, "learning_rate": 0.0002, "epoch": 0.6272734621291731, "step": 3880}, {"loss": 0.8244, "grad_norm": 0.676781415939331, "learning_rate": 0.0002, "epoch": 0.6288901463099184, "step": 3890}, {"loss": 0.8239, "grad_norm": 0.4284018278121948, "learning_rate": 0.0002, "epoch": 0.6305068304906637, "step": 3900}, {"loss": 0.7996, "grad_norm": 0.5060628056526184, "learning_rate": 0.0002, "epoch": 0.6321235146714089, "step": 3910}, {"loss": 0.8089, "grad_norm": 0.5524522066116333, "learning_rate": 0.0002, "epoch": 0.6337401988521543, "step": 3920}, {"loss": 0.8276, "grad_norm": 0.6099881529808044, "learning_rate": 0.0002, "epoch": 0.6353568830328995, "step": 3930}, {"loss": 0.809, "grad_norm": 0.43155938386917114, "learning_rate": 0.0002, "epoch": 0.6369735672136448, "step": 3940}, {"loss": 0.8404, "grad_norm": 0.6427084803581238, "learning_rate": 0.0002, "epoch": 0.6385902513943901, "step": 3950}, {"loss": 0.8368, "grad_norm": 0.541220486164093, "learning_rate": 0.0002, "epoch": 0.6402069355751354, "step": 3960}, {"loss": 0.8539, "grad_norm": 0.5414294600486755, "learning_rate": 0.0002, "epoch": 0.6418236197558806, "step": 3970}, {"loss": 0.7996, "grad_norm": 0.46344003081321716, "learning_rate": 0.0002, "epoch": 0.643440303936626, "step": 3980}, {"loss": 0.7474, "grad_norm": 0.45209285616874695, "learning_rate": 0.0002, "epoch": 0.6450569881173712, "step": 3990}, {"loss": 0.8202, "grad_norm": 0.5417284369468689, "learning_rate": 0.0002, "epoch": 0.6466736722981166, "step": 4000}, {"loss": 0.7563, "grad_norm": 0.7995685935020447, "learning_rate": 0.0002, "epoch": 0.6482903564788619, "step": 4010}, {"loss": 0.7812, "grad_norm": 0.6384002566337585, "learning_rate": 0.0002, "epoch": 0.6499070406596071, "step": 4020}, {"loss": 0.732, "grad_norm": 0.4472815692424774, "learning_rate": 0.0002, "epoch": 0.6515237248403525, "step": 4030}, {"loss": 0.8071, "grad_norm": 0.6834294199943542, "learning_rate": 0.0002, "epoch": 0.6531404090210977, "step": 4040}, {"loss": 0.7812, "grad_norm": 0.4612339735031128, "learning_rate": 0.0002, "epoch": 0.654757093201843, "step": 4050}, {"loss": 0.8141, "grad_norm": 0.9266576170921326, "learning_rate": 0.0002, "epoch": 0.6563737773825883, "step": 4060}, {"loss": 0.7991, "grad_norm": 0.4470861852169037, "learning_rate": 0.0002, "epoch": 0.6579904615633336, "step": 4070}, {"loss": 0.8293, "grad_norm": 0.45544925332069397, "learning_rate": 0.0002, "epoch": 0.6596071457440789, "step": 4080}, {"loss": 0.8455, "grad_norm": 0.6144481301307678, "learning_rate": 0.0002, "epoch": 0.6612238299248242, "step": 4090}, {"loss": 0.7877, "grad_norm": 0.5936288237571716, "learning_rate": 0.0002, "epoch": 0.6628405141055694, "step": 4100}, {"loss": 0.7617, "grad_norm": 0.4822963774204254, "learning_rate": 0.0002, "epoch": 0.6644571982863148, "step": 4110}, {"loss": 0.7997, "grad_norm": 0.48432496190071106, "learning_rate": 0.0002, "epoch": 0.66607388246706, "step": 4120}, {"loss": 0.8404, "grad_norm": 0.4901607930660248, "learning_rate": 0.0002, "epoch": 0.6676905666478054, "step": 4130}, {"loss": 0.8085, "grad_norm": 0.5018393397331238, "learning_rate": 0.0002, "epoch": 0.6693072508285507, "step": 4140}, {"loss": 0.8065, "grad_norm": 0.6946378946304321, "learning_rate": 0.0002, "epoch": 0.6709239350092959, "step": 4150}, {"loss": 0.8147, "grad_norm": 0.5997390747070312, "learning_rate": 0.0002, "epoch": 0.6725406191900413, "step": 4160}, {"loss": 0.8268, "grad_norm": 0.6738849878311157, "learning_rate": 0.0002, "epoch": 0.6741573033707865, "step": 4170}, {"loss": 0.7704, "grad_norm": 0.6110581159591675, "learning_rate": 0.0002, "epoch": 0.6757739875515318, "step": 4180}, {"loss": 0.8043, "grad_norm": 0.5703322291374207, "learning_rate": 0.0002, "epoch": 0.6773906717322771, "step": 4190}, {"loss": 0.8099, "grad_norm": 0.4686066210269928, "learning_rate": 0.0002, "epoch": 0.6790073559130224, "step": 4200}, {"loss": 0.8441, "grad_norm": 0.6394643783569336, "learning_rate": 0.0002, "epoch": 0.6806240400937676, "step": 4210}, {"loss": 0.8011, "grad_norm": 0.5454841256141663, "learning_rate": 0.0002, "epoch": 0.682240724274513, "step": 4220}, {"loss": 0.8307, "grad_norm": 0.4859732985496521, "learning_rate": 0.0002, "epoch": 0.6838574084552582, "step": 4230}, {"loss": 0.8161, "grad_norm": 0.5544065833091736, "learning_rate": 0.0002, "epoch": 0.6854740926360036, "step": 4240}, {"loss": 0.7839, "grad_norm": 0.4902505576610565, "learning_rate": 0.0002, "epoch": 0.6870907768167488, "step": 4250}, {"loss": 0.7977, "grad_norm": 0.4768051505088806, "learning_rate": 0.0002, "epoch": 0.6887074609974941, "step": 4260}, {"loss": 0.7539, "grad_norm": 0.49982190132141113, "learning_rate": 0.0002, "epoch": 0.6903241451782395, "step": 4270}, {"loss": 0.7353, "grad_norm": 0.6351838111877441, "learning_rate": 0.0002, "epoch": 0.6919408293589847, "step": 4280}, {"loss": 0.7664, "grad_norm": 0.5647561550140381, "learning_rate": 0.0002, "epoch": 0.69355751353973, "step": 4290}, {"loss": 0.7618, "grad_norm": 0.5340486764907837, "learning_rate": 0.0002, "epoch": 0.6951741977204753, "step": 4300}, {"loss": 0.8526, "grad_norm": 0.5649092793464661, "learning_rate": 0.0002, "epoch": 0.6967908819012206, "step": 4310}, {"loss": 0.8246, "grad_norm": 0.6183916926383972, "learning_rate": 0.0002, "epoch": 0.6984075660819659, "step": 4320}, {"loss": 0.792, "grad_norm": 0.6154509782791138, "learning_rate": 0.0002, "epoch": 0.7000242502627112, "step": 4330}, {"loss": 0.8397, "grad_norm": 0.5156264305114746, "learning_rate": 0.0002, "epoch": 0.7016409344434564, "step": 4340}, {"loss": 0.8512, "grad_norm": 0.562171459197998, "learning_rate": 0.0002, "epoch": 0.7032576186242018, "step": 4350}, {"loss": 0.7882, "grad_norm": 0.4949502646923065, "learning_rate": 0.0002, "epoch": 0.704874302804947, "step": 4360}, {"loss": 0.738, "grad_norm": 0.5171684622764587, "learning_rate": 0.0002, "epoch": 0.7064909869856923, "step": 4370}, {"loss": 0.8001, "grad_norm": 0.6198443174362183, "learning_rate": 0.0002, "epoch": 0.7081076711664377, "step": 4380}, {"loss": 0.7606, "grad_norm": 0.5802276134490967, "learning_rate": 0.0002, "epoch": 0.7097243553471829, "step": 4390}, {"loss": 0.8797, "grad_norm": 0.41096967458724976, "learning_rate": 0.0002, "epoch": 0.7113410395279283, "step": 4400}, {"loss": 0.805, "grad_norm": 0.4397392272949219, "learning_rate": 0.0002, "epoch": 0.7129577237086735, "step": 4410}, {"loss": 0.7651, "grad_norm": 0.45228442549705505, "learning_rate": 0.0002, "epoch": 0.7145744078894188, "step": 4420}, {"loss": 0.7938, "grad_norm": 0.4839673936367035, "learning_rate": 0.0002, "epoch": 0.7161910920701641, "step": 4430}, {"loss": 0.8362, "grad_norm": 0.6140755414962769, "learning_rate": 0.0002, "epoch": 0.7178077762509094, "step": 4440}, {"loss": 0.7722, "grad_norm": 0.6841378808021545, "learning_rate": 0.0002, "epoch": 0.7194244604316546, "step": 4450}, {"loss": 0.8177, "grad_norm": 0.6664239168167114, "learning_rate": 0.0002, "epoch": 0.7210411446124, "step": 4460}, {"loss": 0.7983, "grad_norm": 0.47552719712257385, "learning_rate": 0.0002, "epoch": 0.7226578287931452, "step": 4470}, {"loss": 0.8982, "grad_norm": 0.6649776101112366, "learning_rate": 0.0002, "epoch": 0.7242745129738906, "step": 4480}, {"loss": 0.8074, "grad_norm": 0.5159541964530945, "learning_rate": 0.0002, "epoch": 0.7258911971546358, "step": 4490}, {"loss": 0.7786, "grad_norm": 0.6693112850189209, "learning_rate": 0.0002, "epoch": 0.7275078813353811, "step": 4500}, {"loss": 0.8655, "grad_norm": 0.48870977759361267, "learning_rate": 0.0002, "epoch": 0.7291245655161265, "step": 4510}, {"loss": 0.7337, "grad_norm": 0.4857887923717499, "learning_rate": 0.0002, "epoch": 0.7307412496968717, "step": 4520}, {"loss": 0.8026, "grad_norm": 0.5515662431716919, "learning_rate": 0.0002, "epoch": 0.732357933877617, "step": 4530}, {"loss": 0.8031, "grad_norm": 0.6292222738265991, "learning_rate": 0.0002, "epoch": 0.7339746180583623, "step": 4540}, {"loss": 0.7749, "grad_norm": 0.48265689611434937, "learning_rate": 0.0002, "epoch": 0.7355913022391076, "step": 4550}, {"loss": 0.8499, "grad_norm": 0.8044266104698181, "learning_rate": 0.0002, "epoch": 0.7372079864198529, "step": 4560}, {"loss": 0.8162, "grad_norm": 0.6111769676208496, "learning_rate": 0.0002, "epoch": 0.7388246706005982, "step": 4570}, {"loss": 0.7291, "grad_norm": 0.5229553580284119, "learning_rate": 0.0002, "epoch": 0.7404413547813434, "step": 4580}, {"loss": 0.8038, "grad_norm": 0.6054152250289917, "learning_rate": 0.0002, "epoch": 0.7420580389620888, "step": 4590}, {"loss": 0.8169, "grad_norm": 0.5574966669082642, "learning_rate": 0.0002, "epoch": 0.743674723142834, "step": 4600}, {"loss": 0.8439, "grad_norm": 0.5395817160606384, "learning_rate": 0.0002, "epoch": 0.7452914073235793, "step": 4610}, {"loss": 0.8495, "grad_norm": 0.7116472721099854, "learning_rate": 0.0002, "epoch": 0.7469080915043246, "step": 4620}, {"loss": 0.7743, "grad_norm": 0.5618700981140137, "learning_rate": 0.0002, "epoch": 0.7485247756850699, "step": 4630}, {"loss": 0.7744, "grad_norm": 0.5802770853042603, "learning_rate": 0.0002, "epoch": 0.7501414598658153, "step": 4640}, {"loss": 0.7924, "grad_norm": 0.5690428018569946, "learning_rate": 0.0002, "epoch": 0.7517581440465605, "step": 4650}, {"loss": 0.8017, "grad_norm": 0.4813360273838043, "learning_rate": 0.0002, "epoch": 0.7533748282273058, "step": 4660}, {"loss": 0.8108, "grad_norm": 0.5434042811393738, "learning_rate": 0.0002, "epoch": 0.7549915124080511, "step": 4670}, {"loss": 0.7824, "grad_norm": 0.5502099990844727, "learning_rate": 0.0002, "epoch": 0.7566081965887964, "step": 4680}, {"loss": 0.8598, "grad_norm": 0.6020621061325073, "learning_rate": 0.0002, "epoch": 0.7582248807695416, "step": 4690}, {"loss": 0.7937, "grad_norm": 0.4922301471233368, "learning_rate": 0.0002, "epoch": 0.759841564950287, "step": 4700}, {"loss": 0.788, "grad_norm": 0.6492828726768494, "learning_rate": 0.0002, "epoch": 0.7614582491310322, "step": 4710}, {"loss": 0.8313, "grad_norm": 0.4865580201148987, "learning_rate": 0.0002, "epoch": 0.7630749333117776, "step": 4720}, {"loss": 0.7966, "grad_norm": 0.5971422791481018, "learning_rate": 0.0002, "epoch": 0.7646916174925228, "step": 4730}, {"loss": 0.8298, "grad_norm": 0.6832674145698547, "learning_rate": 0.0002, "epoch": 0.7663083016732681, "step": 4740}, {"loss": 0.8156, "grad_norm": 0.500908613204956, "learning_rate": 0.0002, "epoch": 0.7679249858540134, "step": 4750}, {"loss": 0.8383, "grad_norm": 0.6112465858459473, "learning_rate": 0.0002, "epoch": 0.7695416700347587, "step": 4760}, {"loss": 0.76, "grad_norm": 0.5753506422042847, "learning_rate": 0.0002, "epoch": 0.771158354215504, "step": 4770}, {"loss": 0.8297, "grad_norm": 0.6529405117034912, "learning_rate": 0.0002, "epoch": 0.7727750383962493, "step": 4780}, {"loss": 0.8171, "grad_norm": 0.5916843414306641, "learning_rate": 0.0002, "epoch": 0.7743917225769946, "step": 4790}, {"loss": 0.83, "grad_norm": 0.4821224510669708, "learning_rate": 0.0002, "epoch": 0.7760084067577399, "step": 4800}, {"loss": 0.7703, "grad_norm": 0.5532580018043518, "learning_rate": 0.0002, "epoch": 0.7776250909384852, "step": 4810}, {"loss": 0.7363, "grad_norm": 0.4604877233505249, "learning_rate": 0.0002, "epoch": 0.7792417751192304, "step": 4820}, {"loss": 0.7506, "grad_norm": 0.5009613037109375, "learning_rate": 0.0002, "epoch": 0.7808584592999758, "step": 4830}, {"loss": 0.7863, "grad_norm": 0.6448560357093811, "learning_rate": 0.0002, "epoch": 0.782475143480721, "step": 4840}, {"loss": 0.7957, "grad_norm": 0.44327953457832336, "learning_rate": 0.0002, "epoch": 0.7840918276614663, "step": 4850}, {"loss": 0.7925, "grad_norm": 0.5355411171913147, "learning_rate": 0.0002, "epoch": 0.7857085118422116, "step": 4860}, {"loss": 0.7754, "grad_norm": 0.5635677576065063, "learning_rate": 0.0002, "epoch": 0.7873251960229569, "step": 4870}, {"loss": 0.7931, "grad_norm": 0.5417491793632507, "learning_rate": 0.0002, "epoch": 0.7889418802037023, "step": 4880}, {"loss": 0.7819, "grad_norm": 0.4567430913448334, "learning_rate": 0.0002, "epoch": 0.7905585643844475, "step": 4890}, {"loss": 0.8454, "grad_norm": 0.44651296734809875, "learning_rate": 0.0002, "epoch": 0.7921752485651928, "step": 4900}, {"loss": 0.7959, "grad_norm": 0.5741217136383057, "learning_rate": 0.0002, "epoch": 0.7937919327459381, "step": 4910}, {"loss": 0.8093, "grad_norm": 0.6605045199394226, "learning_rate": 0.0002, "epoch": 0.7954086169266834, "step": 4920}, {"loss": 0.77, "grad_norm": 0.5126531720161438, "learning_rate": 0.0002, "epoch": 0.7970253011074286, "step": 4930}, {"loss": 0.7793, "grad_norm": 0.513648271560669, "learning_rate": 0.0002, "epoch": 0.798641985288174, "step": 4940}, {"loss": 0.8314, "grad_norm": 0.5350404381752014, "learning_rate": 0.0002, "epoch": 0.8002586694689192, "step": 4950}, {"loss": 0.7649, "grad_norm": 0.5731674432754517, "learning_rate": 0.0002, "epoch": 0.8018753536496646, "step": 4960}, {"loss": 0.8572, "grad_norm": 0.5974258184432983, "learning_rate": 0.0002, "epoch": 0.8034920378304098, "step": 4970}, {"loss": 0.7972, "grad_norm": 0.8774799704551697, "learning_rate": 0.0002, "epoch": 0.8051087220111551, "step": 4980}, {"loss": 0.7899, "grad_norm": 0.5994430184364319, "learning_rate": 0.0002, "epoch": 0.8067254061919004, "step": 4990}, {"loss": 0.7736, "grad_norm": 0.4894903004169464, "learning_rate": 0.0002, "epoch": 0.8083420903726457, "step": 5000}, {"loss": 0.78, "grad_norm": 0.5218459367752075, "learning_rate": 0.0002, "epoch": 0.809958774553391, "step": 5010}, {"loss": 0.817, "grad_norm": 0.5232468843460083, "learning_rate": 0.0002, "epoch": 0.8115754587341363, "step": 5020}, {"loss": 0.7704, "grad_norm": 0.44358372688293457, "learning_rate": 0.0002, "epoch": 0.8131921429148816, "step": 5030}, {"loss": 0.785, "grad_norm": 0.6202037334442139, "learning_rate": 0.0002, "epoch": 0.8148088270956269, "step": 5040}, {"loss": 0.7351, "grad_norm": 0.7721474170684814, "learning_rate": 0.0002, "epoch": 0.8164255112763722, "step": 5050}, {"loss": 0.8297, "grad_norm": 0.5568501353263855, "learning_rate": 0.0002, "epoch": 0.8180421954571174, "step": 5060}, {"loss": 0.7733, "grad_norm": 0.49148809909820557, "learning_rate": 0.0002, "epoch": 0.8196588796378628, "step": 5070}, {"loss": 0.8054, "grad_norm": 0.4956012964248657, "learning_rate": 0.0002, "epoch": 0.821275563818608, "step": 5080}, {"loss": 0.8201, "grad_norm": 0.6078833937644958, "learning_rate": 0.0002, "epoch": 0.8228922479993533, "step": 5090}, {"loss": 0.828, "grad_norm": 0.46906954050064087, "learning_rate": 0.0002, "epoch": 0.8245089321800986, "step": 5100}, {"loss": 0.7703, "grad_norm": 0.50812166929245, "learning_rate": 0.0002, "epoch": 0.8261256163608439, "step": 5110}, {"loss": 0.8243, "grad_norm": 0.5319661498069763, "learning_rate": 0.0002, "epoch": 0.8277423005415891, "step": 5120}, {"loss": 0.7798, "grad_norm": 0.4949689209461212, "learning_rate": 0.0002, "epoch": 0.8293589847223345, "step": 5130}, {"loss": 0.7428, "grad_norm": 0.5151591300964355, "learning_rate": 0.0002, "epoch": 0.8309756689030798, "step": 5140}, {"loss": 0.8147, "grad_norm": 0.5530214309692383, "learning_rate": 0.0002, "epoch": 0.8325923530838251, "step": 5150}, {"loss": 0.8251, "grad_norm": 0.6297410130500793, "learning_rate": 0.0002, "epoch": 0.8342090372645704, "step": 5160}, {"loss": 0.8067, "grad_norm": 0.5466840267181396, "learning_rate": 0.0002, "epoch": 0.8358257214453156, "step": 5170}, {"loss": 0.7875, "grad_norm": 0.652913510799408, "learning_rate": 0.0002, "epoch": 0.837442405626061, "step": 5180}, {"loss": 0.8295, "grad_norm": 0.5811293125152588, "learning_rate": 0.0002, "epoch": 0.8390590898068062, "step": 5190}, {"loss": 0.7412, "grad_norm": 0.5109550952911377, "learning_rate": 0.0002, "epoch": 0.8406757739875516, "step": 5200}, {"loss": 0.8077, "grad_norm": 0.4551706612110138, "learning_rate": 0.0002, "epoch": 0.8422924581682968, "step": 5210}, {"loss": 0.7827, "grad_norm": 0.5813754200935364, "learning_rate": 0.0002, "epoch": 0.8439091423490421, "step": 5220}, {"loss": 0.802, "grad_norm": 0.5856947898864746, "learning_rate": 0.0002, "epoch": 0.8455258265297874, "step": 5230}, {"loss": 0.7957, "grad_norm": 0.5482739210128784, "learning_rate": 0.0002, "epoch": 0.8471425107105327, "step": 5240}, {"loss": 0.8295, "grad_norm": 0.49023720622062683, "learning_rate": 0.0002, "epoch": 0.8487591948912779, "step": 5250}, {"loss": 0.8022, "grad_norm": 0.49472475051879883, "learning_rate": 0.0002, "epoch": 0.8503758790720233, "step": 5260}, {"loss": 0.8001, "grad_norm": 0.5490226745605469, "learning_rate": 0.0002, "epoch": 0.8519925632527686, "step": 5270}, {"loss": 0.8333, "grad_norm": 0.5340665578842163, "learning_rate": 0.0002, "epoch": 0.8536092474335139, "step": 5280}, {"loss": 0.8277, "grad_norm": 0.5962483882904053, "learning_rate": 0.0002, "epoch": 0.8552259316142592, "step": 5290}, {"loss": 0.8765, "grad_norm": 0.586358368396759, "learning_rate": 0.0002, "epoch": 0.8568426157950044, "step": 5300}, {"loss": 0.7831, "grad_norm": 0.49120277166366577, "learning_rate": 0.0002, "epoch": 0.8584592999757498, "step": 5310}, {"loss": 0.8162, "grad_norm": 0.5887332558631897, "learning_rate": 0.0002, "epoch": 0.860075984156495, "step": 5320}, {"loss": 0.7464, "grad_norm": 0.42496153712272644, "learning_rate": 0.0002, "epoch": 0.8616926683372403, "step": 5330}, {"loss": 0.7905, "grad_norm": 0.5489874482154846, "learning_rate": 0.0002, "epoch": 0.8633093525179856, "step": 5340}, {"loss": 0.7958, "grad_norm": 0.5850813984870911, "learning_rate": 0.0002, "epoch": 0.8649260366987309, "step": 5350}, {"loss": 0.7642, "grad_norm": 0.517487108707428, "learning_rate": 0.0002, "epoch": 0.8665427208794761, "step": 5360}, {"loss": 0.7801, "grad_norm": 0.5339142680168152, "learning_rate": 0.0002, "epoch": 0.8681594050602215, "step": 5370}, {"loss": 0.818, "grad_norm": 0.6236387491226196, "learning_rate": 0.0002, "epoch": 0.8697760892409668, "step": 5380}, {"loss": 0.7708, "grad_norm": 0.5752192735671997, "learning_rate": 0.0002, "epoch": 0.8713927734217121, "step": 5390}, {"loss": 0.8542, "grad_norm": 0.6724614500999451, "learning_rate": 0.0002, "epoch": 0.8730094576024574, "step": 5400}, {"loss": 0.7581, "grad_norm": 0.5280613303184509, "learning_rate": 0.0002, "epoch": 0.8746261417832026, "step": 5410}, {"loss": 0.8231, "grad_norm": 0.44033288955688477, "learning_rate": 0.0002, "epoch": 0.876242825963948, "step": 5420}, {"loss": 0.8839, "grad_norm": 0.5199708342552185, "learning_rate": 0.0002, "epoch": 0.8778595101446932, "step": 5430}, {"loss": 0.7852, "grad_norm": 0.46778348088264465, "learning_rate": 0.0002, "epoch": 0.8794761943254386, "step": 5440}, {"loss": 0.7834, "grad_norm": 0.4657754898071289, "learning_rate": 0.0002, "epoch": 0.8810928785061838, "step": 5450}, {"loss": 0.7799, "grad_norm": 0.5472902655601501, "learning_rate": 0.0002, "epoch": 0.8827095626869291, "step": 5460}, {"loss": 0.8253, "grad_norm": 0.4876766800880432, "learning_rate": 0.0002, "epoch": 0.8843262468676744, "step": 5470}, {"loss": 0.7906, "grad_norm": 0.5057248473167419, "learning_rate": 0.0002, "epoch": 0.8859429310484197, "step": 5480}, {"loss": 0.8124, "grad_norm": 0.4637320637702942, "learning_rate": 0.0002, "epoch": 0.8875596152291649, "step": 5490}, {"loss": 0.781, "grad_norm": 0.471955806016922, "learning_rate": 0.0002, "epoch": 0.8891762994099103, "step": 5500}, {"loss": 0.8057, "grad_norm": 0.5209813714027405, "learning_rate": 0.0002, "epoch": 0.8907929835906556, "step": 5510}, {"loss": 0.8106, "grad_norm": 0.6213834285736084, "learning_rate": 0.0002, "epoch": 0.8924096677714008, "step": 5520}, {"loss": 0.7787, "grad_norm": 0.5215408205986023, "learning_rate": 0.0002, "epoch": 0.8940263519521462, "step": 5530}, {"loss": 0.8174, "grad_norm": 0.580478310585022, "learning_rate": 0.0002, "epoch": 0.8956430361328914, "step": 5540}, {"loss": 0.8371, "grad_norm": 0.49102169275283813, "learning_rate": 0.0002, "epoch": 0.8972597203136368, "step": 5550}, {"loss": 0.7806, "grad_norm": 0.6043479442596436, "learning_rate": 0.0002, "epoch": 0.898876404494382, "step": 5560}, {"loss": 0.7754, "grad_norm": 0.5636463165283203, "learning_rate": 0.0002, "epoch": 0.9004930886751273, "step": 5570}, {"loss": 0.8145, "grad_norm": 0.5620124340057373, "learning_rate": 0.0002, "epoch": 0.9021097728558726, "step": 5580}, {"loss": 0.8083, "grad_norm": 0.5206354856491089, "learning_rate": 0.0002, "epoch": 0.9037264570366179, "step": 5590}, {"loss": 0.8557, "grad_norm": 0.5798229575157166, "learning_rate": 0.0002, "epoch": 0.9053431412173631, "step": 5600}, {"loss": 0.8097, "grad_norm": 0.6428212523460388, "learning_rate": 0.0002, "epoch": 0.9069598253981085, "step": 5610}, {"loss": 0.7839, "grad_norm": 0.48064687848091125, "learning_rate": 0.0002, "epoch": 0.9085765095788537, "step": 5620}, {"loss": 0.8343, "grad_norm": 0.6347860097885132, "learning_rate": 0.0002, "epoch": 0.9101931937595991, "step": 5630}, {"loss": 0.851, "grad_norm": 0.5353913307189941, "learning_rate": 0.0002, "epoch": 0.9118098779403444, "step": 5640}, {"loss": 0.7736, "grad_norm": 0.5323944091796875, "learning_rate": 0.0002, "epoch": 0.9134265621210896, "step": 5650}, {"loss": 0.8393, "grad_norm": 0.5261843204498291, "learning_rate": 0.0002, "epoch": 0.915043246301835, "step": 5660}, {"loss": 0.7355, "grad_norm": 0.5451326966285706, "learning_rate": 0.0002, "epoch": 0.9166599304825802, "step": 5670}, {"loss": 0.8012, "grad_norm": 0.5183324217796326, "learning_rate": 0.0002, "epoch": 0.9182766146633256, "step": 5680}, {"loss": 0.7659, "grad_norm": 0.47229018807411194, "learning_rate": 0.0002, "epoch": 0.9198932988440708, "step": 5690}, {"loss": 0.7757, "grad_norm": 0.49180513620376587, "learning_rate": 0.0002, "epoch": 0.9215099830248161, "step": 5700}, {"loss": 0.8735, "grad_norm": 0.5419785380363464, "learning_rate": 0.0002, "epoch": 0.9231266672055614, "step": 5710}, {"loss": 0.7378, "grad_norm": 0.5408698916435242, "learning_rate": 0.0002, "epoch": 0.9247433513863067, "step": 5720}, {"loss": 0.7701, "grad_norm": 0.5286232829093933, "learning_rate": 0.0002, "epoch": 0.9263600355670519, "step": 5730}, {"loss": 0.8242, "grad_norm": 0.7539758086204529, "learning_rate": 0.0002, "epoch": 0.9279767197477973, "step": 5740}, {"loss": 0.8118, "grad_norm": 0.5166944861412048, "learning_rate": 0.0002, "epoch": 0.9295934039285425, "step": 5750}, {"loss": 0.783, "grad_norm": 0.6601425409317017, "learning_rate": 0.0002, "epoch": 0.9312100881092878, "step": 5760}, {"loss": 0.7873, "grad_norm": 0.5029960870742798, "learning_rate": 0.0002, "epoch": 0.9328267722900332, "step": 5770}, {"loss": 0.7989, "grad_norm": 0.4926645755767822, "learning_rate": 0.0002, "epoch": 0.9344434564707784, "step": 5780}, {"loss": 0.8174, "grad_norm": 0.5739615559577942, "learning_rate": 0.0002, "epoch": 0.9360601406515238, "step": 5790}, {"loss": 0.8037, "grad_norm": 0.5058279037475586, "learning_rate": 0.0002, "epoch": 0.937676824832269, "step": 5800}, {"loss": 0.8537, "grad_norm": 0.5260962247848511, "learning_rate": 0.0002, "epoch": 0.9392935090130143, "step": 5810}, {"loss": 0.7486, "grad_norm": 0.5768588185310364, "learning_rate": 0.0002, "epoch": 0.9409101931937596, "step": 5820}, {"loss": 0.8215, "grad_norm": 0.5170126557350159, "learning_rate": 0.0002, "epoch": 0.9425268773745049, "step": 5830}, {"loss": 0.7422, "grad_norm": 0.5745864510536194, "learning_rate": 0.0002, "epoch": 0.9441435615552501, "step": 5840}, {"loss": 0.7824, "grad_norm": 0.5551357865333557, "learning_rate": 0.0002, "epoch": 0.9457602457359955, "step": 5850}, {"loss": 0.8529, "grad_norm": 0.5776078701019287, "learning_rate": 0.0002, "epoch": 0.9473769299167407, "step": 5860}, {"loss": 0.8527, "grad_norm": 0.5340062379837036, "learning_rate": 0.0002, "epoch": 0.9489936140974861, "step": 5870}, {"loss": 0.8217, "grad_norm": 0.6447290182113647, "learning_rate": 0.0002, "epoch": 0.9506102982782314, "step": 5880}, {"loss": 0.7945, "grad_norm": 0.5123815536499023, "learning_rate": 0.0002, "epoch": 0.9522269824589766, "step": 5890}, {"loss": 0.8209, "grad_norm": 0.48547613620758057, "learning_rate": 0.0002, "epoch": 0.953843666639722, "step": 5900}, {"loss": 0.7896, "grad_norm": 0.5791414976119995, "learning_rate": 0.0002, "epoch": 0.9554603508204672, "step": 5910}, {"loss": 0.8408, "grad_norm": 0.6195011734962463, "learning_rate": 0.0002, "epoch": 0.9570770350012126, "step": 5920}, {"loss": 0.7805, "grad_norm": 0.6323803067207336, "learning_rate": 0.0002, "epoch": 0.9586937191819578, "step": 5930}, {"loss": 0.8484, "grad_norm": 0.45552879571914673, "learning_rate": 0.0002, "epoch": 0.9603104033627031, "step": 5940}, {"loss": 0.7367, "grad_norm": 0.5796473622322083, "learning_rate": 0.0002, "epoch": 0.9619270875434484, "step": 5950}, {"loss": 0.7672, "grad_norm": 0.647261381149292, "learning_rate": 0.0002, "epoch": 0.9635437717241937, "step": 5960}, {"loss": 0.8086, "grad_norm": 0.5487682819366455, "learning_rate": 0.0002, "epoch": 0.9651604559049389, "step": 5970}, {"loss": 0.7973, "grad_norm": 0.5743663907051086, "learning_rate": 0.0002, "epoch": 0.9667771400856843, "step": 5980}, {"loss": 0.8153, "grad_norm": 0.5470591187477112, "learning_rate": 0.0002, "epoch": 0.9683938242664295, "step": 5990}, {"loss": 0.8119, "grad_norm": 0.5901660323143005, "learning_rate": 0.0002, "epoch": 0.9700105084471748, "step": 6000}, {"loss": 0.8147, "grad_norm": 0.6544759273529053, "learning_rate": 0.0002, "epoch": 0.9716271926279202, "step": 6010}, {"loss": 0.7536, "grad_norm": 0.6288470029830933, "learning_rate": 0.0002, "epoch": 0.9732438768086654, "step": 6020}, {"loss": 0.7989, "grad_norm": 0.673153817653656, "learning_rate": 0.0002, "epoch": 0.9748605609894108, "step": 6030}, {"loss": 0.7556, "grad_norm": 0.42854753136634827, "learning_rate": 0.0002, "epoch": 0.976477245170156, "step": 6040}, {"loss": 0.8006, "grad_norm": 0.5227066278457642, "learning_rate": 0.0002, "epoch": 0.9780939293509013, "step": 6050}, {"loss": 0.795, "grad_norm": 0.5372416973114014, "learning_rate": 0.0002, "epoch": 0.9797106135316466, "step": 6060}, {"loss": 0.7591, "grad_norm": 0.6026402115821838, "learning_rate": 0.0002, "epoch": 0.9813272977123919, "step": 6070}, {"loss": 0.8347, "grad_norm": 0.49547791481018066, "learning_rate": 0.0002, "epoch": 0.9829439818931371, "step": 6080}, {"loss": 0.7722, "grad_norm": 0.4641951322555542, "learning_rate": 0.0002, "epoch": 0.9845606660738825, "step": 6090}, {"loss": 0.8125, "grad_norm": 0.5818535089492798, "learning_rate": 0.0002, "epoch": 0.9861773502546277, "step": 6100}, {"loss": 0.81, "grad_norm": 0.63955157995224, "learning_rate": 0.0002, "epoch": 0.9877940344353731, "step": 6110}, {"loss": 0.7547, "grad_norm": 0.5649438500404358, "learning_rate": 0.0002, "epoch": 0.9894107186161183, "step": 6120}, {"loss": 0.7861, "grad_norm": 0.5290433168411255, "learning_rate": 0.0002, "epoch": 0.9910274027968636, "step": 6130}, {"loss": 0.8109, "grad_norm": 0.6399374008178711, "learning_rate": 0.0002, "epoch": 0.992644086977609, "step": 6140}, {"loss": 0.8373, "grad_norm": 0.6736576557159424, "learning_rate": 0.0002, "epoch": 0.9942607711583542, "step": 6150}, {"loss": 0.7915, "grad_norm": 0.515420138835907, "learning_rate": 0.0002, "epoch": 0.9958774553390995, "step": 6160}, {"loss": 0.8032, "grad_norm": 0.562677800655365, "learning_rate": 0.0002, "epoch": 0.9974941395198448, "step": 6170}, {"loss": 0.8187, "grad_norm": 0.7113858461380005, "learning_rate": 0.0002, "epoch": 0.9991108237005901, "step": 6180}, {"eval_loss": 1.0871200561523438, "eval_runtime": 122.2071, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 0.9999191657909627, "step": 6185}, {"loss": 0.7507, "grad_norm": 0.7111801505088806, "learning_rate": 0.0002, "epoch": 1.0007275078813354, "step": 6190}, {"loss": 0.6865, "grad_norm": 0.5402125716209412, "learning_rate": 0.0002, "epoch": 1.0023441920620806, "step": 6200}, {"loss": 0.7625, "grad_norm": 0.6098830103874207, "learning_rate": 0.0002, "epoch": 1.003960876242826, "step": 6210}, {"loss": 0.7631, "grad_norm": 0.5829983353614807, "learning_rate": 0.0002, "epoch": 1.0055775604235713, "step": 6220}, {"loss": 0.7188, "grad_norm": 0.5614621043205261, "learning_rate": 0.0002, "epoch": 1.0071942446043165, "step": 6230}, {"loss": 0.7505, "grad_norm": 0.5954238772392273, "learning_rate": 0.0002, "epoch": 1.0088109287850617, "step": 6240}, {"loss": 0.7448, "grad_norm": 0.6480574607849121, "learning_rate": 0.0002, "epoch": 1.0104276129658072, "step": 6250}, {"loss": 0.7514, "grad_norm": 0.6051128506660461, "learning_rate": 0.0002, "epoch": 1.0120442971465524, "step": 6260}, {"loss": 0.7237, "grad_norm": 0.6318870782852173, "learning_rate": 0.0002, "epoch": 1.0136609813272976, "step": 6270}, {"loss": 0.7178, "grad_norm": 0.5048980116844177, "learning_rate": 0.0002, "epoch": 1.015277665508043, "step": 6280}, {"loss": 0.7391, "grad_norm": 0.6346936225891113, "learning_rate": 0.0002, "epoch": 1.0168943496887883, "step": 6290}, {"loss": 0.7486, "grad_norm": 0.5711665749549866, "learning_rate": 0.0002, "epoch": 1.0185110338695336, "step": 6300}, {"loss": 0.6808, "grad_norm": 0.5175361037254333, "learning_rate": 0.0002, "epoch": 1.0201277180502788, "step": 6310}, {"loss": 0.7539, "grad_norm": 0.5360831618309021, "learning_rate": 0.0002, "epoch": 1.0217444022310243, "step": 6320}, {"loss": 0.7112, "grad_norm": 0.614675760269165, "learning_rate": 0.0002, "epoch": 1.0233610864117695, "step": 6330}, {"loss": 0.7748, "grad_norm": 0.5626118183135986, "learning_rate": 0.0002, "epoch": 1.0249777705925147, "step": 6340}, {"loss": 0.7375, "grad_norm": 0.574897289276123, "learning_rate": 0.0002, "epoch": 1.02659445477326, "step": 6350}, {"loss": 0.759, "grad_norm": 0.7185447812080383, "learning_rate": 0.0002, "epoch": 1.0282111389540054, "step": 6360}, {"loss": 0.703, "grad_norm": 0.6705799698829651, "learning_rate": 0.0002, "epoch": 1.0298278231347506, "step": 6370}, {"loss": 0.7139, "grad_norm": 0.6740428805351257, "learning_rate": 0.0002, "epoch": 1.0314445073154959, "step": 6380}, {"loss": 0.7252, "grad_norm": 0.663902759552002, "learning_rate": 0.0002, "epoch": 1.0330611914962413, "step": 6390}, {"loss": 0.7065, "grad_norm": 0.5029543042182922, "learning_rate": 0.0002, "epoch": 1.0346778756769865, "step": 6400}, {"loss": 0.711, "grad_norm": 0.7813863158226013, "learning_rate": 0.0002, "epoch": 1.0362945598577318, "step": 6410}, {"loss": 0.7433, "grad_norm": 0.5396282076835632, "learning_rate": 0.0002, "epoch": 1.037911244038477, "step": 6420}, {"loss": 0.7222, "grad_norm": 0.5253293514251709, "learning_rate": 0.0002, "epoch": 1.0395279282192225, "step": 6430}, {"loss": 0.715, "grad_norm": 0.7236770987510681, "learning_rate": 0.0002, "epoch": 1.0411446123999677, "step": 6440}, {"loss": 0.7259, "grad_norm": 0.5670917630195618, "learning_rate": 0.0002, "epoch": 1.042761296580713, "step": 6450}, {"loss": 0.7195, "grad_norm": 0.6031978726387024, "learning_rate": 0.0002, "epoch": 1.0443779807614582, "step": 6460}, {"loss": 0.7648, "grad_norm": 0.5309213399887085, "learning_rate": 0.0002, "epoch": 1.0459946649422036, "step": 6470}, {"loss": 0.7161, "grad_norm": 0.7114651799201965, "learning_rate": 0.0002, "epoch": 1.0476113491229488, "step": 6480}, {"loss": 0.7583, "grad_norm": 0.5591610670089722, "learning_rate": 0.0002, "epoch": 1.049228033303694, "step": 6490}, {"loss": 0.6645, "grad_norm": 0.5185961127281189, "learning_rate": 0.0002, "epoch": 1.0508447174844395, "step": 6500}, {"loss": 0.7654, "grad_norm": 0.6510552167892456, "learning_rate": 0.0002, "epoch": 1.0524614016651848, "step": 6510}, {"loss": 0.7057, "grad_norm": 0.6557928919792175, "learning_rate": 0.0002, "epoch": 1.05407808584593, "step": 6520}, {"loss": 0.8056, "grad_norm": 0.6973192691802979, "learning_rate": 0.0002, "epoch": 1.0556947700266752, "step": 6530}, {"loss": 0.6793, "grad_norm": 0.6226583123207092, "learning_rate": 0.0002, "epoch": 1.0573114542074207, "step": 6540}, {"loss": 0.7151, "grad_norm": 0.5633195638656616, "learning_rate": 0.0002, "epoch": 1.058928138388166, "step": 6550}, {"loss": 0.7082, "grad_norm": 0.7466658353805542, "learning_rate": 0.0002, "epoch": 1.0605448225689111, "step": 6560}, {"loss": 0.7059, "grad_norm": 0.6462772488594055, "learning_rate": 0.0002, "epoch": 1.0621615067496564, "step": 6570}, {"loss": 0.7046, "grad_norm": 0.5266856551170349, "learning_rate": 0.0002, "epoch": 1.0637781909304018, "step": 6580}, {"loss": 0.7157, "grad_norm": 0.534392774105072, "learning_rate": 0.0002, "epoch": 1.065394875111147, "step": 6590}, {"loss": 0.7115, "grad_norm": 0.7514177560806274, "learning_rate": 0.0002, "epoch": 1.0670115592918923, "step": 6600}, {"loss": 0.7545, "grad_norm": 0.7593035697937012, "learning_rate": 0.0002, "epoch": 1.0686282434726375, "step": 6610}, {"loss": 0.6836, "grad_norm": 0.5277858972549438, "learning_rate": 0.0002, "epoch": 1.070244927653383, "step": 6620}, {"loss": 0.7405, "grad_norm": 0.5573670268058777, "learning_rate": 0.0002, "epoch": 1.0718616118341282, "step": 6630}, {"loss": 0.6774, "grad_norm": 0.6802396774291992, "learning_rate": 0.0002, "epoch": 1.0734782960148734, "step": 6640}, {"loss": 0.723, "grad_norm": 0.7367215752601624, "learning_rate": 0.0002, "epoch": 1.0750949801956189, "step": 6650}, {"loss": 0.7429, "grad_norm": 0.5961891412734985, "learning_rate": 0.0002, "epoch": 1.0767116643763641, "step": 6660}, {"loss": 0.6791, "grad_norm": 0.5736313462257385, "learning_rate": 0.0002, "epoch": 1.0783283485571094, "step": 6670}, {"loss": 0.7178, "grad_norm": 0.619219183921814, "learning_rate": 0.0002, "epoch": 1.0799450327378546, "step": 6680}, {"loss": 0.7318, "grad_norm": 0.6214390993118286, "learning_rate": 0.0002, "epoch": 1.0815617169186, "step": 6690}, {"loss": 0.7554, "grad_norm": 0.564536988735199, "learning_rate": 0.0002, "epoch": 1.0831784010993453, "step": 6700}, {"loss": 0.7362, "grad_norm": 0.5838140249252319, "learning_rate": 0.0002, "epoch": 1.0847950852800905, "step": 6710}, {"loss": 0.739, "grad_norm": 0.7000553607940674, "learning_rate": 0.0002, "epoch": 1.0864117694608357, "step": 6720}, {"loss": 0.7369, "grad_norm": 0.7078263759613037, "learning_rate": 0.0002, "epoch": 1.0880284536415812, "step": 6730}, {"loss": 0.7654, "grad_norm": 0.8353848457336426, "learning_rate": 0.0002, "epoch": 1.0896451378223264, "step": 6740}, {"loss": 0.7015, "grad_norm": 0.5615518689155579, "learning_rate": 0.0002, "epoch": 1.0912618220030716, "step": 6750}, {"loss": 0.7396, "grad_norm": 0.5475581288337708, "learning_rate": 0.0002, "epoch": 1.0928785061838169, "step": 6760}, {"loss": 0.7652, "grad_norm": 0.5835978388786316, "learning_rate": 0.0002, "epoch": 1.0944951903645623, "step": 6770}, {"loss": 0.7541, "grad_norm": 0.5516105890274048, "learning_rate": 0.0002, "epoch": 1.0961118745453076, "step": 6780}, {"loss": 0.6842, "grad_norm": 0.5875251889228821, "learning_rate": 0.0002, "epoch": 1.0977285587260528, "step": 6790}, {"loss": 0.6903, "grad_norm": 0.7376947999000549, "learning_rate": 0.0002, "epoch": 1.0993452429067982, "step": 6800}, {"loss": 0.7512, "grad_norm": 0.5656165480613708, "learning_rate": 0.0002, "epoch": 1.1009619270875435, "step": 6810}, {"loss": 0.7409, "grad_norm": 0.6365954279899597, "learning_rate": 0.0002, "epoch": 1.1025786112682887, "step": 6820}, {"loss": 0.7392, "grad_norm": 0.5033080577850342, "learning_rate": 0.0002, "epoch": 1.104195295449034, "step": 6830}, {"loss": 0.6909, "grad_norm": 0.617396891117096, "learning_rate": 0.0002, "epoch": 1.1058119796297794, "step": 6840}, {"loss": 0.7006, "grad_norm": 0.6395374536514282, "learning_rate": 0.0002, "epoch": 1.1074286638105246, "step": 6850}, {"loss": 0.7335, "grad_norm": 0.6775295734405518, "learning_rate": 0.0002, "epoch": 1.1090453479912699, "step": 6860}, {"loss": 0.764, "grad_norm": 0.6655223965644836, "learning_rate": 0.0002, "epoch": 1.1106620321720153, "step": 6870}, {"loss": 0.7553, "grad_norm": 0.676655113697052, "learning_rate": 0.0002, "epoch": 1.1122787163527605, "step": 6880}, {"loss": 0.7342, "grad_norm": 0.6062718629837036, "learning_rate": 0.0002, "epoch": 1.1138954005335058, "step": 6890}, {"loss": 0.7446, "grad_norm": 0.590943455696106, "learning_rate": 0.0002, "epoch": 1.115512084714251, "step": 6900}, {"loss": 0.6705, "grad_norm": 0.6315317153930664, "learning_rate": 0.0002, "epoch": 1.1171287688949965, "step": 6910}, {"loss": 0.6912, "grad_norm": 0.47979024052619934, "learning_rate": 0.0002, "epoch": 1.1187454530757417, "step": 6920}, {"loss": 0.7002, "grad_norm": 0.647298276424408, "learning_rate": 0.0002, "epoch": 1.120362137256487, "step": 6930}, {"loss": 0.7502, "grad_norm": 0.7336484789848328, "learning_rate": 0.0002, "epoch": 1.1219788214372322, "step": 6940}, {"loss": 0.693, "grad_norm": 0.5071424245834351, "learning_rate": 0.0002, "epoch": 1.1235955056179776, "step": 6950}, {"loss": 0.7378, "grad_norm": 0.6527144312858582, "learning_rate": 0.0002, "epoch": 1.1252121897987228, "step": 6960}, {"loss": 0.7228, "grad_norm": 0.6935935020446777, "learning_rate": 0.0002, "epoch": 1.126828873979468, "step": 6970}, {"loss": 0.699, "grad_norm": 0.8026931881904602, "learning_rate": 0.0002, "epoch": 1.1284455581602133, "step": 6980}, {"loss": 0.7361, "grad_norm": 0.5210393667221069, "learning_rate": 0.0002, "epoch": 1.1300622423409588, "step": 6990}, {"loss": 0.7456, "grad_norm": 0.60475093126297, "learning_rate": 0.0002, "epoch": 1.131678926521704, "step": 7000}, {"loss": 0.7495, "grad_norm": 0.6417073607444763, "learning_rate": 0.0002, "epoch": 1.1332956107024492, "step": 7010}, {"loss": 0.7459, "grad_norm": 0.6732175946235657, "learning_rate": 0.0002, "epoch": 1.1349122948831947, "step": 7020}, {"loss": 0.7278, "grad_norm": 0.6719491481781006, "learning_rate": 0.0002, "epoch": 1.13652897906394, "step": 7030}, {"loss": 0.7694, "grad_norm": 0.5708295106887817, "learning_rate": 0.0002, "epoch": 1.1381456632446851, "step": 7040}, {"loss": 0.7823, "grad_norm": 0.7141719460487366, "learning_rate": 0.0002, "epoch": 1.1397623474254304, "step": 7050}, {"loss": 0.764, "grad_norm": 0.6187017560005188, "learning_rate": 0.0002, "epoch": 1.1413790316061758, "step": 7060}, {"loss": 0.7657, "grad_norm": 0.50581294298172, "learning_rate": 0.0002, "epoch": 1.142995715786921, "step": 7070}, {"loss": 0.7357, "grad_norm": 0.5620143413543701, "learning_rate": 0.0002, "epoch": 1.1446123999676663, "step": 7080}, {"loss": 0.7287, "grad_norm": 0.6231929659843445, "learning_rate": 0.0002, "epoch": 1.1462290841484115, "step": 7090}, {"loss": 0.7328, "grad_norm": 0.5775774121284485, "learning_rate": 0.0002, "epoch": 1.147845768329157, "step": 7100}, {"loss": 0.7728, "grad_norm": 0.6492809653282166, "learning_rate": 0.0002, "epoch": 1.1494624525099022, "step": 7110}, {"loss": 0.7545, "grad_norm": 0.6434972286224365, "learning_rate": 0.0002, "epoch": 1.1510791366906474, "step": 7120}, {"loss": 0.7374, "grad_norm": 0.6191812753677368, "learning_rate": 0.0002, "epoch": 1.1526958208713927, "step": 7130}, {"loss": 0.7276, "grad_norm": 0.6690331697463989, "learning_rate": 0.0002, "epoch": 1.1543125050521381, "step": 7140}, {"loss": 0.7704, "grad_norm": 0.5977938175201416, "learning_rate": 0.0002, "epoch": 1.1559291892328833, "step": 7150}, {"loss": 0.7251, "grad_norm": 0.6195854544639587, "learning_rate": 0.0002, "epoch": 1.1575458734136286, "step": 7160}, {"loss": 0.7249, "grad_norm": 0.5752048492431641, "learning_rate": 0.0002, "epoch": 1.159162557594374, "step": 7170}, {"loss": 0.7593, "grad_norm": 0.589081883430481, "learning_rate": 0.0002, "epoch": 1.1607792417751193, "step": 7180}, {"loss": 0.704, "grad_norm": 0.756996750831604, "learning_rate": 0.0002, "epoch": 1.1623959259558645, "step": 7190}, {"loss": 0.7404, "grad_norm": 0.7614967226982117, "learning_rate": 0.0002, "epoch": 1.1640126101366097, "step": 7200}, {"loss": 0.7867, "grad_norm": 0.6120437979698181, "learning_rate": 0.0002, "epoch": 1.1656292943173552, "step": 7210}, {"loss": 0.7384, "grad_norm": 0.6210004687309265, "learning_rate": 0.0002, "epoch": 1.1672459784981004, "step": 7220}, {"loss": 0.7251, "grad_norm": 0.6044116020202637, "learning_rate": 0.0002, "epoch": 1.1688626626788456, "step": 7230}, {"loss": 0.7361, "grad_norm": 0.5418457388877869, "learning_rate": 0.0002, "epoch": 1.170479346859591, "step": 7240}, {"loss": 0.6938, "grad_norm": 0.6413537263870239, "learning_rate": 0.0002, "epoch": 1.1720960310403363, "step": 7250}, {"loss": 0.6978, "grad_norm": 0.5777867436408997, "learning_rate": 0.0002, "epoch": 1.1737127152210816, "step": 7260}, {"loss": 0.7503, "grad_norm": 0.7092402577400208, "learning_rate": 0.0002, "epoch": 1.1753293994018268, "step": 7270}, {"loss": 0.7487, "grad_norm": 0.6351709365844727, "learning_rate": 0.0002, "epoch": 1.176946083582572, "step": 7280}, {"loss": 0.7527, "grad_norm": 0.6172189712524414, "learning_rate": 0.0002, "epoch": 1.1785627677633175, "step": 7290}, {"loss": 0.7319, "grad_norm": 0.6801714897155762, "learning_rate": 0.0002, "epoch": 1.1801794519440627, "step": 7300}, {"loss": 0.6941, "grad_norm": 0.6044712066650391, "learning_rate": 0.0002, "epoch": 1.181796136124808, "step": 7310}, {"loss": 0.6951, "grad_norm": 0.7413212060928345, "learning_rate": 0.0002, "epoch": 1.1834128203055534, "step": 7320}, {"loss": 0.7396, "grad_norm": 0.5303856134414673, "learning_rate": 0.0002, "epoch": 1.1850295044862986, "step": 7330}, {"loss": 0.6915, "grad_norm": 0.5647098422050476, "learning_rate": 0.0002, "epoch": 1.1866461886670439, "step": 7340}, {"loss": 0.7506, "grad_norm": 0.7374135255813599, "learning_rate": 0.0002, "epoch": 1.188262872847789, "step": 7350}, {"loss": 0.7041, "grad_norm": 0.5710089206695557, "learning_rate": 0.0002, "epoch": 1.1898795570285345, "step": 7360}, {"loss": 0.8289, "grad_norm": 0.6073619723320007, "learning_rate": 0.0002, "epoch": 1.1914962412092798, "step": 7370}, {"loss": 0.7722, "grad_norm": 0.5899916887283325, "learning_rate": 0.0002, "epoch": 1.193112925390025, "step": 7380}, {"loss": 0.756, "grad_norm": 0.7762434482574463, "learning_rate": 0.0002, "epoch": 1.1947296095707705, "step": 7390}, {"loss": 0.7319, "grad_norm": 0.679949939250946, "learning_rate": 0.0002, "epoch": 1.1963462937515157, "step": 7400}, {"loss": 0.7599, "grad_norm": 0.6106849312782288, "learning_rate": 0.0002, "epoch": 1.197962977932261, "step": 7410}, {"loss": 0.7648, "grad_norm": 0.682461678981781, "learning_rate": 0.0002, "epoch": 1.1995796621130062, "step": 7420}, {"loss": 0.7741, "grad_norm": 0.6087017059326172, "learning_rate": 0.0002, "epoch": 1.2011963462937516, "step": 7430}, {"loss": 0.7642, "grad_norm": 0.63739013671875, "learning_rate": 0.0002, "epoch": 1.2028130304744968, "step": 7440}, {"loss": 0.7611, "grad_norm": 0.6154777407646179, "learning_rate": 0.0002, "epoch": 1.204429714655242, "step": 7450}, {"loss": 0.7565, "grad_norm": 0.7491534948348999, "learning_rate": 0.0002, "epoch": 1.2060463988359873, "step": 7460}, {"loss": 0.698, "grad_norm": 0.6664797067642212, "learning_rate": 0.0002, "epoch": 1.2076630830167328, "step": 7470}, {"loss": 0.7456, "grad_norm": 0.6660266518592834, "learning_rate": 0.0002, "epoch": 1.209279767197478, "step": 7480}, {"loss": 0.714, "grad_norm": 0.6972551345825195, "learning_rate": 0.0002, "epoch": 1.2108964513782232, "step": 7490}, {"loss": 0.7023, "grad_norm": 0.6157945990562439, "learning_rate": 0.0002, "epoch": 1.2125131355589684, "step": 7500}, {"loss": 0.7326, "grad_norm": 0.5199310183525085, "learning_rate": 0.0002, "epoch": 1.214129819739714, "step": 7510}, {"loss": 0.7586, "grad_norm": 0.577610433101654, "learning_rate": 0.0002, "epoch": 1.2157465039204591, "step": 7520}, {"loss": 0.7179, "grad_norm": 0.53652423620224, "learning_rate": 0.0002, "epoch": 1.2173631881012044, "step": 7530}, {"loss": 0.7393, "grad_norm": 0.6479050517082214, "learning_rate": 0.0002, "epoch": 1.2189798722819498, "step": 7540}, {"loss": 0.7534, "grad_norm": 0.618748128414154, "learning_rate": 0.0002, "epoch": 1.220596556462695, "step": 7550}, {"loss": 0.6886, "grad_norm": 0.6311424374580383, "learning_rate": 0.0002, "epoch": 1.2222132406434403, "step": 7560}, {"loss": 0.7272, "grad_norm": 0.6595825552940369, "learning_rate": 0.0002, "epoch": 1.2238299248241855, "step": 7570}, {"loss": 0.7353, "grad_norm": 0.5198960900306702, "learning_rate": 0.0002, "epoch": 1.225446609004931, "step": 7580}, {"loss": 0.674, "grad_norm": 0.578650712966919, "learning_rate": 0.0002, "epoch": 1.2270632931856762, "step": 7590}, {"loss": 0.7507, "grad_norm": 0.6080220937728882, "learning_rate": 0.0002, "epoch": 1.2286799773664214, "step": 7600}, {"loss": 0.7733, "grad_norm": 0.7050248384475708, "learning_rate": 0.0002, "epoch": 1.2302966615471669, "step": 7610}, {"loss": 0.7032, "grad_norm": 0.6652196049690247, "learning_rate": 0.0002, "epoch": 1.2319133457279121, "step": 7620}, {"loss": 0.7085, "grad_norm": 0.7322776317596436, "learning_rate": 0.0002, "epoch": 1.2335300299086573, "step": 7630}, {"loss": 0.7402, "grad_norm": 0.4998728036880493, "learning_rate": 0.0002, "epoch": 1.2351467140894026, "step": 7640}, {"loss": 0.7214, "grad_norm": 0.6428788900375366, "learning_rate": 0.0002, "epoch": 1.2367633982701478, "step": 7650}, {"loss": 0.7699, "grad_norm": 0.585242509841919, "learning_rate": 0.0002, "epoch": 1.2383800824508933, "step": 7660}, {"loss": 0.7621, "grad_norm": 0.5211917757987976, "learning_rate": 0.0002, "epoch": 1.2399967666316385, "step": 7670}, {"loss": 0.746, "grad_norm": 0.6490384340286255, "learning_rate": 0.0002, "epoch": 1.2416134508123837, "step": 7680}, {"loss": 0.7186, "grad_norm": 0.6249763369560242, "learning_rate": 0.0002, "epoch": 1.2432301349931292, "step": 7690}, {"loss": 0.7761, "grad_norm": 0.71870356798172, "learning_rate": 0.0002, "epoch": 1.2448468191738744, "step": 7700}, {"loss": 0.7525, "grad_norm": 0.6761967539787292, "learning_rate": 0.0002, "epoch": 1.2464635033546196, "step": 7710}, {"loss": 0.7501, "grad_norm": 0.6500617265701294, "learning_rate": 0.0002, "epoch": 1.2480801875353649, "step": 7720}, {"loss": 0.7903, "grad_norm": 0.8069869875907898, "learning_rate": 0.0002, "epoch": 1.2496968717161103, "step": 7730}, {"loss": 0.6747, "grad_norm": 0.6044608950614929, "learning_rate": 0.0002, "epoch": 1.2513135558968556, "step": 7740}, {"loss": 0.6825, "grad_norm": 0.6573283076286316, "learning_rate": 0.0002, "epoch": 1.2529302400776008, "step": 7750}, {"loss": 0.7617, "grad_norm": 0.625430166721344, "learning_rate": 0.0002, "epoch": 1.2545469242583462, "step": 7760}, {"loss": 0.7041, "grad_norm": 0.5442022681236267, "learning_rate": 0.0002, "epoch": 1.2561636084390915, "step": 7770}, {"loss": 0.7172, "grad_norm": 0.6818386912345886, "learning_rate": 0.0002, "epoch": 1.2577802926198367, "step": 7780}, {"loss": 0.696, "grad_norm": 0.6381874084472656, "learning_rate": 0.0002, "epoch": 1.259396976800582, "step": 7790}, {"loss": 0.6834, "grad_norm": 0.6269212961196899, "learning_rate": 0.0002, "epoch": 1.2610136609813272, "step": 7800}, {"loss": 0.7821, "grad_norm": 0.600121259689331, "learning_rate": 0.0002, "epoch": 1.2626303451620726, "step": 7810}, {"loss": 0.7761, "grad_norm": 0.6337703466415405, "learning_rate": 0.0002, "epoch": 1.2642470293428179, "step": 7820}, {"loss": 0.732, "grad_norm": 0.7234963774681091, "learning_rate": 0.0002, "epoch": 1.2658637135235633, "step": 7830}, {"loss": 0.785, "grad_norm": 0.800184965133667, "learning_rate": 0.0002, "epoch": 1.2674803977043085, "step": 7840}, {"loss": 0.7426, "grad_norm": 0.7539464831352234, "learning_rate": 0.0002, "epoch": 1.2690970818850538, "step": 7850}, {"loss": 0.7496, "grad_norm": 0.5493760704994202, "learning_rate": 0.0002, "epoch": 1.270713766065799, "step": 7860}, {"loss": 0.7537, "grad_norm": 0.7477145791053772, "learning_rate": 0.0002, "epoch": 1.2723304502465442, "step": 7870}, {"loss": 0.7573, "grad_norm": 0.6366362571716309, "learning_rate": 0.0002, "epoch": 1.2739471344272897, "step": 7880}, {"loss": 0.7608, "grad_norm": 0.7419533729553223, "learning_rate": 0.0002, "epoch": 1.275563818608035, "step": 7890}, {"loss": 0.7873, "grad_norm": 0.6141223311424255, "learning_rate": 0.0002, "epoch": 1.2771805027887801, "step": 7900}, {"loss": 0.6916, "grad_norm": 0.7522598505020142, "learning_rate": 0.0002, "epoch": 1.2787971869695256, "step": 7910}, {"loss": 0.7097, "grad_norm": 0.6935804486274719, "learning_rate": 0.0002, "epoch": 1.2804138711502708, "step": 7920}, {"loss": 0.7185, "grad_norm": 0.7239290475845337, "learning_rate": 0.0002, "epoch": 1.282030555331016, "step": 7930}, {"loss": 0.7145, "grad_norm": 0.8800187110900879, "learning_rate": 0.0002, "epoch": 1.2836472395117613, "step": 7940}, {"loss": 0.6991, "grad_norm": 0.540458083152771, "learning_rate": 0.0002, "epoch": 1.2852639236925067, "step": 7950}, {"loss": 0.7139, "grad_norm": 0.6492934226989746, "learning_rate": 0.0002, "epoch": 1.286880607873252, "step": 7960}, {"loss": 0.7742, "grad_norm": 0.6543959379196167, "learning_rate": 0.0002, "epoch": 1.2884972920539972, "step": 7970}, {"loss": 0.7316, "grad_norm": 0.5804705619812012, "learning_rate": 0.0002, "epoch": 1.2901139762347427, "step": 7980}, {"loss": 0.796, "grad_norm": 0.7074727416038513, "learning_rate": 0.0002, "epoch": 1.291730660415488, "step": 7990}, {"loss": 0.7034, "grad_norm": 0.5347974300384521, "learning_rate": 0.0002, "epoch": 1.2933473445962331, "step": 8000}, {"loss": 0.738, "grad_norm": 0.6457298398017883, "learning_rate": 0.0002, "epoch": 1.2949640287769784, "step": 8010}, {"loss": 0.7634, "grad_norm": 0.6407219171524048, "learning_rate": 0.0002, "epoch": 1.2965807129577236, "step": 8020}, {"loss": 0.7506, "grad_norm": 0.828439474105835, "learning_rate": 0.0002, "epoch": 1.298197397138469, "step": 8030}, {"loss": 0.735, "grad_norm": 0.4840380549430847, "learning_rate": 0.0002, "epoch": 1.2998140813192143, "step": 8040}, {"loss": 0.7283, "grad_norm": 0.5921024680137634, "learning_rate": 0.0002, "epoch": 1.3014307654999595, "step": 8050}, {"loss": 0.7477, "grad_norm": 0.6170315146446228, "learning_rate": 0.0002, "epoch": 1.303047449680705, "step": 8060}, {"loss": 0.7534, "grad_norm": 0.5374847054481506, "learning_rate": 0.0002, "epoch": 1.3046641338614502, "step": 8070}, {"loss": 0.7593, "grad_norm": 0.545758068561554, "learning_rate": 0.0002, "epoch": 1.3062808180421954, "step": 8080}, {"loss": 0.7463, "grad_norm": 0.55641770362854, "learning_rate": 0.0002, "epoch": 1.3078975022229407, "step": 8090}, {"loss": 0.7594, "grad_norm": 0.6724897027015686, "learning_rate": 0.0002, "epoch": 1.309514186403686, "step": 8100}, {"loss": 0.7105, "grad_norm": 0.6923972368240356, "learning_rate": 0.0002, "epoch": 1.3111308705844313, "step": 8110}, {"loss": 0.7149, "grad_norm": 0.5136841535568237, "learning_rate": 0.0002, "epoch": 1.3127475547651766, "step": 8120}, {"loss": 0.7504, "grad_norm": 0.6766283512115479, "learning_rate": 0.0002, "epoch": 1.314364238945922, "step": 8130}, {"loss": 0.7489, "grad_norm": 0.6283926367759705, "learning_rate": 0.0002, "epoch": 1.3159809231266673, "step": 8140}, {"loss": 0.7459, "grad_norm": 0.644216001033783, "learning_rate": 0.0002, "epoch": 1.3175976073074125, "step": 8150}, {"loss": 0.7125, "grad_norm": 0.7827503085136414, "learning_rate": 0.0002, "epoch": 1.3192142914881577, "step": 8160}, {"loss": 0.7271, "grad_norm": 0.6651390790939331, "learning_rate": 0.0002, "epoch": 1.320830975668903, "step": 8170}, {"loss": 0.7778, "grad_norm": 0.5547412633895874, "learning_rate": 0.0002, "epoch": 1.3224476598496484, "step": 8180}, {"loss": 0.7402, "grad_norm": 0.6765179634094238, "learning_rate": 0.0002, "epoch": 1.3240643440303936, "step": 8190}, {"loss": 0.7106, "grad_norm": 0.6822077035903931, "learning_rate": 0.0002, "epoch": 1.325681028211139, "step": 8200}, {"loss": 0.7288, "grad_norm": 0.5941002368927002, "learning_rate": 0.0002, "epoch": 1.3272977123918843, "step": 8210}, {"loss": 0.7494, "grad_norm": 0.4850037097930908, "learning_rate": 0.0002, "epoch": 1.3289143965726296, "step": 8220}, {"loss": 0.7474, "grad_norm": 0.6162990927696228, "learning_rate": 0.0002, "epoch": 1.3305310807533748, "step": 8230}, {"loss": 0.7751, "grad_norm": 0.6665613651275635, "learning_rate": 0.0002, "epoch": 1.33214776493412, "step": 8240}, {"loss": 0.759, "grad_norm": 0.618192732334137, "learning_rate": 0.0002, "epoch": 1.3337644491148655, "step": 8250}, {"loss": 0.7532, "grad_norm": 0.710418701171875, "learning_rate": 0.0002, "epoch": 1.3353811332956107, "step": 8260}, {"loss": 0.7306, "grad_norm": 0.5109876990318298, "learning_rate": 0.0002, "epoch": 1.336997817476356, "step": 8270}, {"loss": 0.7303, "grad_norm": 0.6791711449623108, "learning_rate": 0.0002, "epoch": 1.3386145016571014, "step": 8280}, {"loss": 0.7594, "grad_norm": 0.6836432814598083, "learning_rate": 0.0002, "epoch": 1.3402311858378466, "step": 8290}, {"loss": 0.7594, "grad_norm": 0.5579386353492737, "learning_rate": 0.0002, "epoch": 1.3418478700185918, "step": 8300}, {"loss": 0.7377, "grad_norm": 0.6713546514511108, "learning_rate": 0.0002, "epoch": 1.343464554199337, "step": 8310}, {"loss": 0.7756, "grad_norm": 0.5353720188140869, "learning_rate": 0.0002, "epoch": 1.3450812383800825, "step": 8320}, {"loss": 0.718, "grad_norm": 0.5813682675361633, "learning_rate": 0.0002, "epoch": 1.3466979225608278, "step": 8330}, {"loss": 0.7294, "grad_norm": 0.8158791661262512, "learning_rate": 0.0002, "epoch": 1.348314606741573, "step": 8340}, {"loss": 0.6992, "grad_norm": 0.6193785071372986, "learning_rate": 0.0002, "epoch": 1.3499312909223184, "step": 8350}, {"loss": 0.7654, "grad_norm": 0.6353939771652222, "learning_rate": 0.0002, "epoch": 1.3515479751030637, "step": 8360}, {"loss": 0.7519, "grad_norm": 0.6925048232078552, "learning_rate": 0.0002, "epoch": 1.353164659283809, "step": 8370}, {"loss": 0.736, "grad_norm": 0.988264799118042, "learning_rate": 0.0002, "epoch": 1.3547813434645541, "step": 8380}, {"loss": 0.7744, "grad_norm": 0.6476002931594849, "learning_rate": 0.0002, "epoch": 1.3563980276452994, "step": 8390}, {"loss": 0.776, "grad_norm": 0.7120398879051208, "learning_rate": 0.0002, "epoch": 1.3580147118260448, "step": 8400}, {"loss": 0.7368, "grad_norm": 0.9048416614532471, "learning_rate": 0.0002, "epoch": 1.35963139600679, "step": 8410}, {"loss": 0.7544, "grad_norm": 0.7000672817230225, "learning_rate": 0.0002, "epoch": 1.3612480801875353, "step": 8420}, {"loss": 0.7358, "grad_norm": 0.6015632152557373, "learning_rate": 0.0002, "epoch": 1.3628647643682807, "step": 8430}, {"loss": 0.7298, "grad_norm": 0.612516462802887, "learning_rate": 0.0002, "epoch": 1.364481448549026, "step": 8440}, {"loss": 0.7055, "grad_norm": 0.5969301462173462, "learning_rate": 0.0002, "epoch": 1.3660981327297712, "step": 8450}, {"loss": 0.7754, "grad_norm": 0.6730654239654541, "learning_rate": 0.0002, "epoch": 1.3677148169105164, "step": 8460}, {"loss": 0.7465, "grad_norm": 0.6386392116546631, "learning_rate": 0.0002, "epoch": 1.369331501091262, "step": 8470}, {"loss": 0.7433, "grad_norm": 0.739544153213501, "learning_rate": 0.0002, "epoch": 1.3709481852720071, "step": 8480}, {"loss": 0.7892, "grad_norm": 0.6462782621383667, "learning_rate": 0.0002, "epoch": 1.3725648694527524, "step": 8490}, {"loss": 0.7302, "grad_norm": 0.7346843481063843, "learning_rate": 0.0002, "epoch": 1.3741815536334978, "step": 8500}, {"loss": 0.7634, "grad_norm": 0.6884821057319641, "learning_rate": 0.0002, "epoch": 1.375798237814243, "step": 8510}, {"loss": 0.7614, "grad_norm": 0.6999333500862122, "learning_rate": 0.0002, "epoch": 1.3774149219949883, "step": 8520}, {"loss": 0.729, "grad_norm": 0.5378713011741638, "learning_rate": 0.0002, "epoch": 1.3790316061757335, "step": 8530}, {"loss": 0.6797, "grad_norm": 0.5417906641960144, "learning_rate": 0.0002, "epoch": 1.3806482903564787, "step": 8540}, {"loss": 0.7499, "grad_norm": 0.6602526307106018, "learning_rate": 0.0002, "epoch": 1.3822649745372242, "step": 8550}, {"loss": 0.7356, "grad_norm": 0.7073674201965332, "learning_rate": 0.0002, "epoch": 1.3838816587179694, "step": 8560}, {"loss": 0.75, "grad_norm": 0.5841707587242126, "learning_rate": 0.0002, "epoch": 1.3854983428987149, "step": 8570}, {"loss": 0.732, "grad_norm": 0.7031095027923584, "learning_rate": 0.0002, "epoch": 1.38711502707946, "step": 8580}, {"loss": 0.7464, "grad_norm": 0.5198570489883423, "learning_rate": 0.0002, "epoch": 1.3887317112602053, "step": 8590}, {"loss": 0.7354, "grad_norm": 0.7261320352554321, "learning_rate": 0.0002, "epoch": 1.3903483954409506, "step": 8600}, {"loss": 0.7339, "grad_norm": 0.5616350173950195, "learning_rate": 0.0002, "epoch": 1.3919650796216958, "step": 8610}, {"loss": 0.7382, "grad_norm": 0.5185914635658264, "learning_rate": 0.0002, "epoch": 1.3935817638024413, "step": 8620}, {"loss": 0.7456, "grad_norm": 0.5814694762229919, "learning_rate": 0.0002, "epoch": 1.3951984479831865, "step": 8630}, {"loss": 0.7413, "grad_norm": 0.6977371573448181, "learning_rate": 0.0002, "epoch": 1.3968151321639317, "step": 8640}, {"loss": 0.7574, "grad_norm": 0.6855689883232117, "learning_rate": 0.0002, "epoch": 1.3984318163446772, "step": 8650}, {"loss": 0.7802, "grad_norm": 0.5414357781410217, "learning_rate": 0.0002, "epoch": 1.4000485005254224, "step": 8660}, {"loss": 0.7487, "grad_norm": 0.6970012784004211, "learning_rate": 0.0002, "epoch": 1.4016651847061676, "step": 8670}, {"loss": 0.7421, "grad_norm": 0.526079535484314, "learning_rate": 0.0002, "epoch": 1.4032818688869129, "step": 8680}, {"loss": 0.737, "grad_norm": 0.758712887763977, "learning_rate": 0.0002, "epoch": 1.404898553067658, "step": 8690}, {"loss": 0.7612, "grad_norm": 0.7118762731552124, "learning_rate": 0.0002, "epoch": 1.4065152372484035, "step": 8700}, {"loss": 0.7628, "grad_norm": 0.5696909427642822, "learning_rate": 0.0002, "epoch": 1.4081319214291488, "step": 8710}, {"loss": 0.7156, "grad_norm": 0.7995436787605286, "learning_rate": 0.0002, "epoch": 1.4097486056098942, "step": 8720}, {"loss": 0.7521, "grad_norm": 0.7237521409988403, "learning_rate": 0.0002, "epoch": 1.4113652897906395, "step": 8730}, {"loss": 0.7661, "grad_norm": 0.744628369808197, "learning_rate": 0.0002, "epoch": 1.4129819739713847, "step": 8740}, {"loss": 0.7073, "grad_norm": 0.6082926988601685, "learning_rate": 0.0002, "epoch": 1.41459865815213, "step": 8750}, {"loss": 0.7282, "grad_norm": 0.5185243487358093, "learning_rate": 0.0002, "epoch": 1.4162153423328752, "step": 8760}, {"loss": 0.7592, "grad_norm": 0.5183082222938538, "learning_rate": 0.0002, "epoch": 1.4178320265136206, "step": 8770}, {"loss": 0.7509, "grad_norm": 0.7326041460037231, "learning_rate": 0.0002, "epoch": 1.4194487106943658, "step": 8780}, {"loss": 0.7398, "grad_norm": 0.7174660563468933, "learning_rate": 0.0002, "epoch": 1.421065394875111, "step": 8790}, {"loss": 0.7507, "grad_norm": 0.8080165982246399, "learning_rate": 0.0002, "epoch": 1.4226820790558565, "step": 8800}, {"loss": 0.72, "grad_norm": 0.5061507821083069, "learning_rate": 0.0002, "epoch": 1.4242987632366018, "step": 8810}, {"loss": 0.7563, "grad_norm": 0.801602840423584, "learning_rate": 0.0002, "epoch": 1.425915447417347, "step": 8820}, {"loss": 0.7287, "grad_norm": 0.6150273084640503, "learning_rate": 0.0002, "epoch": 1.4275321315980922, "step": 8830}, {"loss": 0.7452, "grad_norm": 0.8786525726318359, "learning_rate": 0.0002, "epoch": 1.4291488157788377, "step": 8840}, {"loss": 0.7257, "grad_norm": 0.6371538639068604, "learning_rate": 0.0002, "epoch": 1.430765499959583, "step": 8850}, {"loss": 0.711, "grad_norm": 0.6409295797348022, "learning_rate": 0.0002, "epoch": 1.4323821841403281, "step": 8860}, {"loss": 0.7891, "grad_norm": 0.6452359557151794, "learning_rate": 0.0002, "epoch": 1.4339988683210736, "step": 8870}, {"loss": 0.7588, "grad_norm": 0.5842334628105164, "learning_rate": 0.0002, "epoch": 1.4356155525018188, "step": 8880}, {"loss": 0.7446, "grad_norm": 0.696761965751648, "learning_rate": 0.0002, "epoch": 1.437232236682564, "step": 8890}, {"loss": 0.7541, "grad_norm": 0.6384600400924683, "learning_rate": 0.0002, "epoch": 1.4388489208633093, "step": 8900}, {"loss": 0.7049, "grad_norm": 0.5981136560440063, "learning_rate": 0.0002, "epoch": 1.4404656050440545, "step": 8910}, {"loss": 0.795, "grad_norm": 0.6355637907981873, "learning_rate": 0.0002, "epoch": 1.4420822892248, "step": 8920}, {"loss": 0.7653, "grad_norm": 0.6374830603599548, "learning_rate": 0.0002, "epoch": 1.4436989734055452, "step": 8930}, {"loss": 0.8108, "grad_norm": 0.559013307094574, "learning_rate": 0.0002, "epoch": 1.4453156575862904, "step": 8940}, {"loss": 0.7045, "grad_norm": 0.7289170026779175, "learning_rate": 0.0002, "epoch": 1.446932341767036, "step": 8950}, {"loss": 0.7484, "grad_norm": 0.8649206757545471, "learning_rate": 0.0002, "epoch": 1.4485490259477811, "step": 8960}, {"loss": 0.7745, "grad_norm": 0.7664689421653748, "learning_rate": 0.0002, "epoch": 1.4501657101285264, "step": 8970}, {"loss": 0.7431, "grad_norm": 0.7109952569007874, "learning_rate": 0.0002, "epoch": 1.4517823943092716, "step": 8980}, {"loss": 0.7997, "grad_norm": 0.6312844753265381, "learning_rate": 0.0002, "epoch": 1.453399078490017, "step": 8990}, {"loss": 0.7467, "grad_norm": 0.6616617441177368, "learning_rate": 0.0002, "epoch": 1.4550157626707623, "step": 9000}, {"loss": 0.7518, "grad_norm": 0.7384068965911865, "learning_rate": 0.0002, "epoch": 1.4566324468515075, "step": 9010}, {"loss": 0.7483, "grad_norm": 0.6549670100212097, "learning_rate": 0.0002, "epoch": 1.458249131032253, "step": 9020}, {"loss": 0.7423, "grad_norm": 0.6254119277000427, "learning_rate": 0.0002, "epoch": 1.4598658152129982, "step": 9030}, {"loss": 0.7645, "grad_norm": 0.6806328892707825, "learning_rate": 0.0002, "epoch": 1.4614824993937434, "step": 9040}, {"loss": 0.7221, "grad_norm": 0.6803115010261536, "learning_rate": 0.0002, "epoch": 1.4630991835744886, "step": 9050}, {"loss": 0.7264, "grad_norm": 0.48529282212257385, "learning_rate": 0.0002, "epoch": 1.4647158677552339, "step": 9060}, {"loss": 0.7542, "grad_norm": 0.5995030999183655, "learning_rate": 0.0002, "epoch": 1.4663325519359793, "step": 9070}, {"loss": 0.7894, "grad_norm": 0.6005427837371826, "learning_rate": 0.0002, "epoch": 1.4679492361167246, "step": 9080}, {"loss": 0.7288, "grad_norm": 0.718564510345459, "learning_rate": 0.0002, "epoch": 1.46956592029747, "step": 9090}, {"loss": 0.7089, "grad_norm": 0.7003577351570129, "learning_rate": 0.0002, "epoch": 1.4711826044782153, "step": 9100}, {"loss": 0.8069, "grad_norm": 0.5888323783874512, "learning_rate": 0.0002, "epoch": 1.4727992886589605, "step": 9110}, {"loss": 0.7275, "grad_norm": 0.6417609453201294, "learning_rate": 0.0002, "epoch": 1.4744159728397057, "step": 9120}, {"loss": 0.7441, "grad_norm": 0.572294294834137, "learning_rate": 0.0002, "epoch": 1.476032657020451, "step": 9130}, {"loss": 0.8053, "grad_norm": 0.8200714588165283, "learning_rate": 0.0002, "epoch": 1.4776493412011964, "step": 9140}, {"loss": 0.7382, "grad_norm": 0.6343288421630859, "learning_rate": 0.0002, "epoch": 1.4792660253819416, "step": 9150}, {"loss": 0.7641, "grad_norm": 0.7017961144447327, "learning_rate": 0.0002, "epoch": 1.4808827095626869, "step": 9160}, {"loss": 0.7619, "grad_norm": 0.6202912926673889, "learning_rate": 0.0002, "epoch": 1.4824993937434323, "step": 9170}, {"loss": 0.7428, "grad_norm": 0.6677869558334351, "learning_rate": 0.0002, "epoch": 1.4841160779241775, "step": 9180}, {"loss": 0.7648, "grad_norm": 0.6052267551422119, "learning_rate": 0.0002, "epoch": 1.4857327621049228, "step": 9190}, {"loss": 0.7152, "grad_norm": 0.6638872027397156, "learning_rate": 0.0002, "epoch": 1.487349446285668, "step": 9200}, {"loss": 0.7448, "grad_norm": 0.6245523691177368, "learning_rate": 0.0002, "epoch": 1.4889661304664135, "step": 9210}, {"loss": 0.6958, "grad_norm": 0.5761767625808716, "learning_rate": 0.0002, "epoch": 1.4905828146471587, "step": 9220}, {"loss": 0.8012, "grad_norm": 0.8175981640815735, "learning_rate": 0.0002, "epoch": 1.492199498827904, "step": 9230}, {"loss": 0.683, "grad_norm": 0.9144009947776794, "learning_rate": 0.0002, "epoch": 1.4938161830086494, "step": 9240}, {"loss": 0.7623, "grad_norm": 0.5742552876472473, "learning_rate": 0.0002, "epoch": 1.4954328671893946, "step": 9250}, {"loss": 0.7418, "grad_norm": 0.534534215927124, "learning_rate": 0.0002, "epoch": 1.4970495513701398, "step": 9260}, {"loss": 0.7194, "grad_norm": 0.7836225032806396, "learning_rate": 0.0002, "epoch": 1.498666235550885, "step": 9270}, {"loss": 0.7453, "grad_norm": 0.5292993187904358, "learning_rate": 0.0002, "epoch": 1.5002829197316303, "step": 9280}, {"loss": 0.7168, "grad_norm": 0.8044071793556213, "learning_rate": 0.0002, "epoch": 1.5018996039123758, "step": 9290}, {"loss": 0.7229, "grad_norm": 0.6185805201530457, "learning_rate": 0.0002, "epoch": 1.503516288093121, "step": 9300}, {"loss": 0.684, "grad_norm": 0.6093607544898987, "learning_rate": 0.0002, "epoch": 1.5051329722738664, "step": 9310}, {"loss": 0.7973, "grad_norm": 0.5891730189323425, "learning_rate": 0.0002, "epoch": 1.5067496564546117, "step": 9320}, {"loss": 0.7474, "grad_norm": 0.6331129670143127, "learning_rate": 0.0002, "epoch": 1.508366340635357, "step": 9330}, {"loss": 0.7074, "grad_norm": 0.7690958380699158, "learning_rate": 0.0002, "epoch": 1.5099830248161021, "step": 9340}, {"loss": 0.672, "grad_norm": 0.6548877358436584, "learning_rate": 0.0002, "epoch": 1.5115997089968474, "step": 9350}, {"loss": 0.7408, "grad_norm": 0.6545143127441406, "learning_rate": 0.0002, "epoch": 1.5132163931775926, "step": 9360}, {"loss": 0.7432, "grad_norm": 0.553247332572937, "learning_rate": 0.0002, "epoch": 1.514833077358338, "step": 9370}, {"loss": 0.7265, "grad_norm": 0.8145074844360352, "learning_rate": 0.0002, "epoch": 1.5164497615390833, "step": 9380}, {"loss": 0.7379, "grad_norm": 0.7636994123458862, "learning_rate": 0.0002, "epoch": 1.5180664457198287, "step": 9390}, {"loss": 0.7413, "grad_norm": 0.6838982701301575, "learning_rate": 0.0002, "epoch": 1.519683129900574, "step": 9400}, {"loss": 0.7367, "grad_norm": 0.8599441647529602, "learning_rate": 0.0002, "epoch": 1.5212998140813192, "step": 9410}, {"loss": 0.7663, "grad_norm": 0.7020329833030701, "learning_rate": 0.0002, "epoch": 1.5229164982620644, "step": 9420}, {"loss": 0.7928, "grad_norm": 0.6964772343635559, "learning_rate": 0.0002, "epoch": 1.5245331824428097, "step": 9430}, {"loss": 0.7168, "grad_norm": 0.6916600465774536, "learning_rate": 0.0002, "epoch": 1.5261498666235551, "step": 9440}, {"loss": 0.7519, "grad_norm": 0.7282621264457703, "learning_rate": 0.0002, "epoch": 1.5277665508043003, "step": 9450}, {"loss": 0.7628, "grad_norm": 0.5363983511924744, "learning_rate": 0.0002, "epoch": 1.5293832349850458, "step": 9460}, {"loss": 0.7154, "grad_norm": 0.6184861063957214, "learning_rate": 0.0002, "epoch": 1.530999919165791, "step": 9470}, {"loss": 0.7837, "grad_norm": 0.5991285443305969, "learning_rate": 0.0002, "epoch": 1.5326166033465363, "step": 9480}, {"loss": 0.7827, "grad_norm": 0.8176587820053101, "learning_rate": 0.0002, "epoch": 1.5342332875272815, "step": 9490}, {"loss": 0.7415, "grad_norm": 0.6473721861839294, "learning_rate": 0.0002, "epoch": 1.5358499717080267, "step": 9500}, {"loss": 0.7632, "grad_norm": 0.7319952845573425, "learning_rate": 0.0002, "epoch": 1.5374666558887722, "step": 9510}, {"loss": 0.7706, "grad_norm": 0.702900230884552, "learning_rate": 0.0002, "epoch": 1.5390833400695174, "step": 9520}, {"loss": 0.7754, "grad_norm": 0.7971600294113159, "learning_rate": 0.0002, "epoch": 1.5407000242502629, "step": 9530}, {"loss": 0.7352, "grad_norm": 0.6527525186538696, "learning_rate": 0.0002, "epoch": 1.542316708431008, "step": 9540}, {"loss": 0.7425, "grad_norm": 0.5791676044464111, "learning_rate": 0.0002, "epoch": 1.5439333926117533, "step": 9550}, {"loss": 0.7585, "grad_norm": 0.5619390606880188, "learning_rate": 0.0002, "epoch": 1.5455500767924986, "step": 9560}, {"loss": 0.7894, "grad_norm": 0.5701689124107361, "learning_rate": 0.0002, "epoch": 1.5471667609732438, "step": 9570}, {"loss": 0.793, "grad_norm": 0.47549352049827576, "learning_rate": 0.0002, "epoch": 1.548783445153989, "step": 9580}, {"loss": 0.7276, "grad_norm": 0.8730611205101013, "learning_rate": 0.0002, "epoch": 1.5504001293347345, "step": 9590}, {"loss": 0.798, "grad_norm": 0.6842091083526611, "learning_rate": 0.0002, "epoch": 1.5520168135154797, "step": 9600}, {"loss": 0.7528, "grad_norm": 0.6675129532814026, "learning_rate": 0.0002, "epoch": 1.5536334976962252, "step": 9610}, {"loss": 0.7954, "grad_norm": 0.8173956274986267, "learning_rate": 0.0002, "epoch": 1.5552501818769704, "step": 9620}, {"loss": 0.7535, "grad_norm": 0.724947452545166, "learning_rate": 0.0002, "epoch": 1.5568668660577156, "step": 9630}, {"loss": 0.7738, "grad_norm": 0.6154758930206299, "learning_rate": 0.0002, "epoch": 1.5584835502384609, "step": 9640}, {"loss": 0.7568, "grad_norm": 0.6072008013725281, "learning_rate": 0.0002, "epoch": 1.560100234419206, "step": 9650}, {"loss": 0.7219, "grad_norm": 0.659010648727417, "learning_rate": 0.0002, "epoch": 1.5617169185999515, "step": 9660}, {"loss": 0.673, "grad_norm": 0.65857994556427, "learning_rate": 0.0002, "epoch": 1.5633336027806968, "step": 9670}, {"loss": 0.7156, "grad_norm": 0.5914267301559448, "learning_rate": 0.0002, "epoch": 1.5649502869614422, "step": 9680}, {"loss": 0.7414, "grad_norm": 0.6248020529747009, "learning_rate": 0.0002, "epoch": 1.5665669711421875, "step": 9690}, {"loss": 0.694, "grad_norm": 0.7147795557975769, "learning_rate": 0.0002, "epoch": 1.5681836553229327, "step": 9700}, {"loss": 0.7335, "grad_norm": 0.7076232433319092, "learning_rate": 0.0002, "epoch": 1.569800339503678, "step": 9710}, {"loss": 0.7413, "grad_norm": 0.6217400431632996, "learning_rate": 0.0002, "epoch": 1.5714170236844232, "step": 9720}, {"loss": 0.7296, "grad_norm": 0.6709911227226257, "learning_rate": 0.0002, "epoch": 1.5730337078651684, "step": 9730}, {"loss": 0.7306, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 1.5746503920459138, "step": 9740}, {"loss": 0.7242, "grad_norm": 0.6241145730018616, "learning_rate": 0.0002, "epoch": 1.576267076226659, "step": 9750}, {"loss": 0.7384, "grad_norm": 0.4960934817790985, "learning_rate": 0.0002, "epoch": 1.5778837604074045, "step": 9760}, {"loss": 0.725, "grad_norm": 0.6593309640884399, "learning_rate": 0.0002, "epoch": 1.5795004445881498, "step": 9770}, {"loss": 0.7531, "grad_norm": 0.5814042091369629, "learning_rate": 0.0002, "epoch": 1.581117128768895, "step": 9780}, {"loss": 0.7109, "grad_norm": 0.5936070680618286, "learning_rate": 0.0002, "epoch": 1.5827338129496402, "step": 9790}, {"loss": 0.7769, "grad_norm": 0.6454403400421143, "learning_rate": 0.0002, "epoch": 1.5843504971303854, "step": 9800}, {"loss": 0.7677, "grad_norm": 0.7612107992172241, "learning_rate": 0.0002, "epoch": 1.585967181311131, "step": 9810}, {"loss": 0.7649, "grad_norm": 0.6494482755661011, "learning_rate": 0.0002, "epoch": 1.5875838654918761, "step": 9820}, {"loss": 0.7569, "grad_norm": 0.7825694680213928, "learning_rate": 0.0002, "epoch": 1.5892005496726216, "step": 9830}, {"loss": 0.706, "grad_norm": 0.6757757663726807, "learning_rate": 0.0002, "epoch": 1.5908172338533668, "step": 9840}, {"loss": 0.7803, "grad_norm": 0.7105609178543091, "learning_rate": 0.0002, "epoch": 1.592433918034112, "step": 9850}, {"loss": 0.7925, "grad_norm": 0.7596991062164307, "learning_rate": 0.0002, "epoch": 1.5940506022148573, "step": 9860}, {"loss": 0.7108, "grad_norm": 0.5681525468826294, "learning_rate": 0.0002, "epoch": 1.5956672863956025, "step": 9870}, {"loss": 0.7811, "grad_norm": 0.6090980768203735, "learning_rate": 0.0002, "epoch": 1.5972839705763477, "step": 9880}, {"loss": 0.7339, "grad_norm": 0.6271613240242004, "learning_rate": 0.0002, "epoch": 1.5989006547570932, "step": 9890}, {"loss": 0.7419, "grad_norm": 0.7656369805335999, "learning_rate": 0.0002, "epoch": 1.6005173389378387, "step": 9900}, {"loss": 0.7336, "grad_norm": 0.7504446506500244, "learning_rate": 0.0002, "epoch": 1.6021340231185839, "step": 9910}, {"loss": 0.7479, "grad_norm": 0.659656286239624, "learning_rate": 0.0002, "epoch": 1.6037507072993291, "step": 9920}, {"loss": 0.7483, "grad_norm": 0.6006826162338257, "learning_rate": 0.0002, "epoch": 1.6053673914800743, "step": 9930}, {"loss": 0.732, "grad_norm": 0.7872757911682129, "learning_rate": 0.0002, "epoch": 1.6069840756608196, "step": 9940}, {"loss": 0.768, "grad_norm": 0.5545852780342102, "learning_rate": 0.0002, "epoch": 1.6086007598415648, "step": 9950}, {"loss": 0.8064, "grad_norm": 0.7429468631744385, "learning_rate": 0.0002, "epoch": 1.6102174440223103, "step": 9960}, {"loss": 0.714, "grad_norm": 0.6873556971549988, "learning_rate": 0.0002, "epoch": 1.6118341282030555, "step": 9970}, {"loss": 0.7324, "grad_norm": 0.5874287486076355, "learning_rate": 0.0002, "epoch": 1.613450812383801, "step": 9980}, {"loss": 0.7141, "grad_norm": 0.6039386987686157, "learning_rate": 0.0002, "epoch": 1.6150674965645462, "step": 9990}, {"loss": 0.6674, "grad_norm": 0.6233575940132141, "learning_rate": 0.0002, "epoch": 1.6166841807452914, "step": 10000}, {"loss": 0.7602, "grad_norm": 0.7676448225975037, "learning_rate": 0.0002, "epoch": 1.6183008649260366, "step": 10010}, {"loss": 0.7784, "grad_norm": 0.6565698385238647, "learning_rate": 0.0002, "epoch": 1.6199175491067819, "step": 10020}, {"loss": 0.7104, "grad_norm": 0.6787590384483337, "learning_rate": 0.0002, "epoch": 1.6215342332875273, "step": 10030}, {"loss": 0.7464, "grad_norm": 0.6137678027153015, "learning_rate": 0.0002, "epoch": 1.6231509174682726, "step": 10040}, {"loss": 0.7646, "grad_norm": 0.5236800312995911, "learning_rate": 0.0002, "epoch": 1.624767601649018, "step": 10050}, {"loss": 0.7437, "grad_norm": 0.7626367807388306, "learning_rate": 0.0002, "epoch": 1.6263842858297632, "step": 10060}, {"loss": 0.7273, "grad_norm": 0.5657260417938232, "learning_rate": 0.0002, "epoch": 1.6280009700105085, "step": 10070}, {"loss": 0.7354, "grad_norm": 0.4913991391658783, "learning_rate": 0.0002, "epoch": 1.6296176541912537, "step": 10080}, {"loss": 0.7596, "grad_norm": 0.7715556621551514, "learning_rate": 0.0002, "epoch": 1.631234338371999, "step": 10090}, {"loss": 0.7105, "grad_norm": 0.6509000062942505, "learning_rate": 0.0002, "epoch": 1.6328510225527442, "step": 10100}, {"loss": 0.7274, "grad_norm": 0.6215850114822388, "learning_rate": 0.0002, "epoch": 1.6344677067334896, "step": 10110}, {"loss": 0.7705, "grad_norm": 0.6956844329833984, "learning_rate": 0.0002, "epoch": 1.6360843909142349, "step": 10120}, {"loss": 0.7129, "grad_norm": 0.6111597418785095, "learning_rate": 0.0002, "epoch": 1.6377010750949803, "step": 10130}, {"loss": 0.6955, "grad_norm": 0.6518288850784302, "learning_rate": 0.0002, "epoch": 1.6393177592757255, "step": 10140}, {"loss": 0.731, "grad_norm": 0.6914522051811218, "learning_rate": 0.0002, "epoch": 1.6409344434564708, "step": 10150}, {"loss": 0.7295, "grad_norm": 0.63785719871521, "learning_rate": 0.0002, "epoch": 1.642551127637216, "step": 10160}, {"loss": 0.7355, "grad_norm": 0.6379287838935852, "learning_rate": 0.0002, "epoch": 1.6441678118179612, "step": 10170}, {"loss": 0.7359, "grad_norm": 0.6793403029441833, "learning_rate": 0.0002, "epoch": 1.6457844959987067, "step": 10180}, {"loss": 0.7402, "grad_norm": 0.6099132895469666, "learning_rate": 0.0002, "epoch": 1.647401180179452, "step": 10190}, {"loss": 0.7353, "grad_norm": 0.5869854092597961, "learning_rate": 0.0002, "epoch": 1.6490178643601974, "step": 10200}, {"loss": 0.8308, "grad_norm": 0.7716999053955078, "learning_rate": 0.0002, "epoch": 1.6506345485409426, "step": 10210}, {"loss": 0.7215, "grad_norm": 0.6854110360145569, "learning_rate": 0.0002, "epoch": 1.6522512327216878, "step": 10220}, {"loss": 0.782, "grad_norm": 0.6957170367240906, "learning_rate": 0.0002, "epoch": 1.653867916902433, "step": 10230}, {"loss": 0.7282, "grad_norm": 0.6932903528213501, "learning_rate": 0.0002, "epoch": 1.6554846010831783, "step": 10240}, {"loss": 0.7478, "grad_norm": 0.7713165283203125, "learning_rate": 0.0002, "epoch": 1.6571012852639235, "step": 10250}, {"loss": 0.7099, "grad_norm": 0.7455793619155884, "learning_rate": 0.0002, "epoch": 1.658717969444669, "step": 10260}, {"loss": 0.7524, "grad_norm": 0.5464168190956116, "learning_rate": 0.0002, "epoch": 1.6603346536254144, "step": 10270}, {"loss": 0.7328, "grad_norm": 0.6782926321029663, "learning_rate": 0.0002, "epoch": 1.6619513378061597, "step": 10280}, {"loss": 0.7801, "grad_norm": 0.7962649464607239, "learning_rate": 0.0002, "epoch": 1.663568021986905, "step": 10290}, {"loss": 0.7142, "grad_norm": 0.6814526319503784, "learning_rate": 0.0002, "epoch": 1.6651847061676501, "step": 10300}, {"loss": 0.7285, "grad_norm": 0.656895101070404, "learning_rate": 0.0002, "epoch": 1.6668013903483954, "step": 10310}, {"loss": 0.7358, "grad_norm": 0.6085672378540039, "learning_rate": 0.0002, "epoch": 1.6684180745291406, "step": 10320}, {"loss": 0.7074, "grad_norm": 0.585508406162262, "learning_rate": 0.0002, "epoch": 1.670034758709886, "step": 10330}, {"loss": 0.7604, "grad_norm": 0.6930184364318848, "learning_rate": 0.0002, "epoch": 1.6716514428906313, "step": 10340}, {"loss": 0.7169, "grad_norm": 0.575663149356842, "learning_rate": 0.0002, "epoch": 1.6732681270713767, "step": 10350}, {"loss": 0.7198, "grad_norm": 0.582502543926239, "learning_rate": 0.0002, "epoch": 1.674884811252122, "step": 10360}, {"loss": 0.7793, "grad_norm": 0.5668916702270508, "learning_rate": 0.0002, "epoch": 1.6765014954328672, "step": 10370}, {"loss": 0.7478, "grad_norm": 0.6070065498352051, "learning_rate": 0.0002, "epoch": 1.6781181796136124, "step": 10380}, {"loss": 0.7939, "grad_norm": 0.6141316294670105, "learning_rate": 0.0002, "epoch": 1.6797348637943577, "step": 10390}, {"loss": 0.7573, "grad_norm": 0.8359124064445496, "learning_rate": 0.0002, "epoch": 1.6813515479751031, "step": 10400}, {"loss": 0.7488, "grad_norm": 0.5378185510635376, "learning_rate": 0.0002, "epoch": 1.6829682321558483, "step": 10410}, {"loss": 0.7588, "grad_norm": 0.6959536075592041, "learning_rate": 0.0002, "epoch": 1.6845849163365938, "step": 10420}, {"loss": 0.7872, "grad_norm": 0.6514357328414917, "learning_rate": 0.0002, "epoch": 1.686201600517339, "step": 10430}, {"loss": 0.725, "grad_norm": 0.7706646919250488, "learning_rate": 0.0002, "epoch": 1.6878182846980843, "step": 10440}, {"loss": 0.7673, "grad_norm": 0.6183337569236755, "learning_rate": 0.0002, "epoch": 1.6894349688788295, "step": 10450}, {"loss": 0.7566, "grad_norm": 0.6123278141021729, "learning_rate": 0.0002, "epoch": 1.6910516530595747, "step": 10460}, {"loss": 0.7169, "grad_norm": 0.6894851326942444, "learning_rate": 0.0002, "epoch": 1.69266833724032, "step": 10470}, {"loss": 0.7435, "grad_norm": 0.7497312426567078, "learning_rate": 0.0002, "epoch": 1.6942850214210654, "step": 10480}, {"loss": 0.7544, "grad_norm": 0.5968214273452759, "learning_rate": 0.0002, "epoch": 1.6959017056018106, "step": 10490}, {"loss": 0.6793, "grad_norm": 0.6747927069664001, "learning_rate": 0.0002, "epoch": 1.697518389782556, "step": 10500}, {"loss": 0.7415, "grad_norm": 0.5708310008049011, "learning_rate": 0.0002, "epoch": 1.6991350739633013, "step": 10510}, {"loss": 0.7385, "grad_norm": 0.606526792049408, "learning_rate": 0.0002, "epoch": 1.7007517581440466, "step": 10520}, {"loss": 0.7204, "grad_norm": 0.662011981010437, "learning_rate": 0.0002, "epoch": 1.7023684423247918, "step": 10530}, {"loss": 0.7999, "grad_norm": 0.7583045363426208, "learning_rate": 0.0002, "epoch": 1.703985126505537, "step": 10540}, {"loss": 0.7563, "grad_norm": 0.721632182598114, "learning_rate": 0.0002, "epoch": 1.7056018106862825, "step": 10550}, {"loss": 0.7407, "grad_norm": 0.6107715368270874, "learning_rate": 0.0002, "epoch": 1.7072184948670277, "step": 10560}, {"loss": 0.7519, "grad_norm": 0.6652471423149109, "learning_rate": 0.0002, "epoch": 1.7088351790477732, "step": 10570}, {"loss": 0.7767, "grad_norm": 0.6308087110519409, "learning_rate": 0.0002, "epoch": 1.7104518632285184, "step": 10580}, {"loss": 0.7659, "grad_norm": 0.5464386940002441, "learning_rate": 0.0002, "epoch": 1.7120685474092636, "step": 10590}, {"loss": 0.7063, "grad_norm": 0.6558911204338074, "learning_rate": 0.0002, "epoch": 1.7136852315900089, "step": 10600}, {"loss": 0.7126, "grad_norm": 0.5665024518966675, "learning_rate": 0.0002, "epoch": 1.715301915770754, "step": 10610}, {"loss": 0.6958, "grad_norm": 0.7888094186782837, "learning_rate": 0.0002, "epoch": 1.7169185999514993, "step": 10620}, {"loss": 0.7785, "grad_norm": 0.7084909081459045, "learning_rate": 0.0002, "epoch": 1.7185352841322448, "step": 10630}, {"loss": 0.7557, "grad_norm": 0.7982324361801147, "learning_rate": 0.0002, "epoch": 1.7201519683129902, "step": 10640}, {"loss": 0.7345, "grad_norm": 0.6418732404708862, "learning_rate": 0.0002, "epoch": 1.7217686524937355, "step": 10650}, {"loss": 0.7734, "grad_norm": 0.7636681795120239, "learning_rate": 0.0002, "epoch": 1.7233853366744807, "step": 10660}, {"loss": 0.7541, "grad_norm": 0.5646875500679016, "learning_rate": 0.0002, "epoch": 1.725002020855226, "step": 10670}, {"loss": 0.7642, "grad_norm": 0.5231260657310486, "learning_rate": 0.0002, "epoch": 1.7266187050359711, "step": 10680}, {"loss": 0.7846, "grad_norm": 0.7635011672973633, "learning_rate": 0.0002, "epoch": 1.7282353892167164, "step": 10690}, {"loss": 0.7471, "grad_norm": 0.7518259286880493, "learning_rate": 0.0002, "epoch": 1.7298520733974618, "step": 10700}, {"loss": 0.751, "grad_norm": 0.7295602560043335, "learning_rate": 0.0002, "epoch": 1.731468757578207, "step": 10710}, {"loss": 0.731, "grad_norm": 0.6984632015228271, "learning_rate": 0.0002, "epoch": 1.7330854417589525, "step": 10720}, {"loss": 0.7921, "grad_norm": 0.6198219060897827, "learning_rate": 0.0002, "epoch": 1.7347021259396977, "step": 10730}, {"loss": 0.7642, "grad_norm": 0.6957576274871826, "learning_rate": 0.0002, "epoch": 1.736318810120443, "step": 10740}, {"loss": 0.7917, "grad_norm": 0.6430263519287109, "learning_rate": 0.0002, "epoch": 1.7379354943011882, "step": 10750}, {"loss": 0.7156, "grad_norm": 0.6134995222091675, "learning_rate": 0.0002, "epoch": 1.7395521784819334, "step": 10760}, {"loss": 0.7584, "grad_norm": 0.7209452986717224, "learning_rate": 0.0002, "epoch": 1.741168862662679, "step": 10770}, {"loss": 0.7528, "grad_norm": 0.6735447645187378, "learning_rate": 0.0002, "epoch": 1.7427855468434241, "step": 10780}, {"loss": 0.756, "grad_norm": 0.5605693459510803, "learning_rate": 0.0002, "epoch": 1.7444022310241696, "step": 10790}, {"loss": 0.7759, "grad_norm": 0.6882363557815552, "learning_rate": 0.0002, "epoch": 1.7460189152049148, "step": 10800}, {"loss": 0.7544, "grad_norm": 0.6386259198188782, "learning_rate": 0.0002, "epoch": 1.74763559938566, "step": 10810}, {"loss": 0.7697, "grad_norm": 0.6529015302658081, "learning_rate": 0.0002, "epoch": 1.7492522835664053, "step": 10820}, {"loss": 0.7219, "grad_norm": 0.5664082765579224, "learning_rate": 0.0002, "epoch": 1.7508689677471505, "step": 10830}, {"loss": 0.7586, "grad_norm": 0.7532684206962585, "learning_rate": 0.0002, "epoch": 1.7524856519278957, "step": 10840}, {"loss": 0.6919, "grad_norm": 0.77171391248703, "learning_rate": 0.0002, "epoch": 1.7541023361086412, "step": 10850}, {"loss": 0.785, "grad_norm": 0.7255431413650513, "learning_rate": 0.0002, "epoch": 1.7557190202893864, "step": 10860}, {"loss": 0.7458, "grad_norm": 0.763083279132843, "learning_rate": 0.0002, "epoch": 1.7573357044701319, "step": 10870}, {"loss": 0.7846, "grad_norm": 0.6042402982711792, "learning_rate": 0.0002, "epoch": 1.758952388650877, "step": 10880}, {"loss": 0.7027, "grad_norm": 0.7642518281936646, "learning_rate": 0.0002, "epoch": 1.7605690728316223, "step": 10890}, {"loss": 0.746, "grad_norm": 0.6347904801368713, "learning_rate": 0.0002, "epoch": 1.7621857570123676, "step": 10900}, {"loss": 0.7458, "grad_norm": 0.5371627807617188, "learning_rate": 0.0002, "epoch": 1.7638024411931128, "step": 10910}, {"loss": 0.7466, "grad_norm": 0.6840225458145142, "learning_rate": 0.0002, "epoch": 1.7654191253738583, "step": 10920}, {"loss": 0.725, "grad_norm": 0.5288469195365906, "learning_rate": 0.0002, "epoch": 1.7670358095546035, "step": 10930}, {"loss": 0.7863, "grad_norm": 0.69020676612854, "learning_rate": 0.0002, "epoch": 1.768652493735349, "step": 10940}, {"loss": 0.7468, "grad_norm": 0.5943242311477661, "learning_rate": 0.0002, "epoch": 1.7702691779160942, "step": 10950}, {"loss": 0.7244, "grad_norm": 0.5616418123245239, "learning_rate": 0.0002, "epoch": 1.7718858620968394, "step": 10960}, {"loss": 0.7137, "grad_norm": 0.7209470868110657, "learning_rate": 0.0002, "epoch": 1.7735025462775846, "step": 10970}, {"loss": 0.7459, "grad_norm": 0.6657957434654236, "learning_rate": 0.0002, "epoch": 1.7751192304583299, "step": 10980}, {"loss": 0.7076, "grad_norm": 0.6469064950942993, "learning_rate": 0.0002, "epoch": 1.776735914639075, "step": 10990}, {"loss": 0.7321, "grad_norm": 0.6615678071975708, "learning_rate": 0.0002, "epoch": 1.7783525988198206, "step": 11000}, {"loss": 0.747, "grad_norm": 0.6722439527511597, "learning_rate": 0.0002, "epoch": 1.779969283000566, "step": 11010}, {"loss": 0.7302, "grad_norm": 0.634136974811554, "learning_rate": 0.0002, "epoch": 1.7815859671813112, "step": 11020}, {"loss": 0.8105, "grad_norm": 0.6024377346038818, "learning_rate": 0.0002, "epoch": 1.7832026513620565, "step": 11030}, {"loss": 0.7855, "grad_norm": 0.6909403800964355, "learning_rate": 0.0002, "epoch": 1.7848193355428017, "step": 11040}, {"loss": 0.7471, "grad_norm": 0.7148767709732056, "learning_rate": 0.0002, "epoch": 1.786436019723547, "step": 11050}, {"loss": 0.7145, "grad_norm": 0.7442979216575623, "learning_rate": 0.0002, "epoch": 1.7880527039042922, "step": 11060}, {"loss": 0.7215, "grad_norm": 0.6830431818962097, "learning_rate": 0.0002, "epoch": 1.7896693880850376, "step": 11070}, {"loss": 0.7625, "grad_norm": 0.9172667264938354, "learning_rate": 0.0002, "epoch": 1.7912860722657828, "step": 11080}, {"loss": 0.76, "grad_norm": 0.6799490451812744, "learning_rate": 0.0002, "epoch": 1.7929027564465283, "step": 11090}, {"loss": 0.7716, "grad_norm": 0.7617024779319763, "learning_rate": 0.0002, "epoch": 1.7945194406272735, "step": 11100}, {"loss": 0.7586, "grad_norm": 0.7701810002326965, "learning_rate": 0.0002, "epoch": 1.7961361248080188, "step": 11110}, {"loss": 0.7843, "grad_norm": 0.7454385757446289, "learning_rate": 0.0002, "epoch": 1.797752808988764, "step": 11120}, {"loss": 0.7873, "grad_norm": 0.6121436953544617, "learning_rate": 0.0002, "epoch": 1.7993694931695092, "step": 11130}, {"loss": 0.7305, "grad_norm": 0.6237571835517883, "learning_rate": 0.0002, "epoch": 1.8009861773502547, "step": 11140}, {"loss": 0.6827, "grad_norm": 0.6818515658378601, "learning_rate": 0.0002, "epoch": 1.802602861531, "step": 11150}, {"loss": 0.6876, "grad_norm": 0.7768308520317078, "learning_rate": 0.0002, "epoch": 1.8042195457117454, "step": 11160}, {"loss": 0.7533, "grad_norm": 0.6875537633895874, "learning_rate": 0.0002, "epoch": 1.8058362298924906, "step": 11170}, {"loss": 0.761, "grad_norm": 0.7950584888458252, "learning_rate": 0.0002, "epoch": 1.8074529140732358, "step": 11180}, {"loss": 0.7623, "grad_norm": 0.8210248351097107, "learning_rate": 0.0002, "epoch": 1.809069598253981, "step": 11190}, {"loss": 0.7556, "grad_norm": 0.6674110889434814, "learning_rate": 0.0002, "epoch": 1.8106862824347263, "step": 11200}, {"loss": 0.7663, "grad_norm": 0.6261674761772156, "learning_rate": 0.0002, "epoch": 1.8123029666154715, "step": 11210}, {"loss": 0.7122, "grad_norm": 0.6484741568565369, "learning_rate": 0.0002, "epoch": 1.813919650796217, "step": 11220}, {"loss": 0.7718, "grad_norm": 0.6231244206428528, "learning_rate": 0.0002, "epoch": 1.8155363349769622, "step": 11230}, {"loss": 0.7152, "grad_norm": 0.7243146896362305, "learning_rate": 0.0002, "epoch": 1.8171530191577077, "step": 11240}, {"loss": 0.7448, "grad_norm": 0.6776193380355835, "learning_rate": 0.0002, "epoch": 1.818769703338453, "step": 11250}, {"loss": 0.7317, "grad_norm": 0.5973618030548096, "learning_rate": 0.0002, "epoch": 1.8203863875191981, "step": 11260}, {"loss": 0.7961, "grad_norm": 0.6451361179351807, "learning_rate": 0.0002, "epoch": 1.8220030716999434, "step": 11270}, {"loss": 0.7611, "grad_norm": 0.5963068008422852, "learning_rate": 0.0002, "epoch": 1.8236197558806886, "step": 11280}, {"loss": 0.7466, "grad_norm": 0.536902129650116, "learning_rate": 0.0002, "epoch": 1.825236440061434, "step": 11290}, {"loss": 0.708, "grad_norm": 0.6993787288665771, "learning_rate": 0.0002, "epoch": 1.8268531242421793, "step": 11300}, {"loss": 0.7153, "grad_norm": 0.6135255098342896, "learning_rate": 0.0002, "epoch": 1.8284698084229247, "step": 11310}, {"loss": 0.7423, "grad_norm": 0.6057423949241638, "learning_rate": 0.0002, "epoch": 1.83008649260367, "step": 11320}, {"loss": 0.735, "grad_norm": 0.6598812341690063, "learning_rate": 0.0002, "epoch": 1.8317031767844152, "step": 11330}, {"loss": 0.7278, "grad_norm": 0.6075948476791382, "learning_rate": 0.0002, "epoch": 1.8333198609651604, "step": 11340}, {"loss": 0.7846, "grad_norm": 0.7065447568893433, "learning_rate": 0.0002, "epoch": 1.8349365451459057, "step": 11350}, {"loss": 0.7365, "grad_norm": 0.680526614189148, "learning_rate": 0.0002, "epoch": 1.8365532293266509, "step": 11360}, {"loss": 0.7152, "grad_norm": 0.6356695294380188, "learning_rate": 0.0002, "epoch": 1.8381699135073963, "step": 11370}, {"loss": 0.721, "grad_norm": 0.6399052143096924, "learning_rate": 0.0002, "epoch": 1.8397865976881416, "step": 11380}, {"loss": 0.7618, "grad_norm": 0.6125704050064087, "learning_rate": 0.0002, "epoch": 1.841403281868887, "step": 11390}, {"loss": 0.755, "grad_norm": 0.7124643325805664, "learning_rate": 0.0002, "epoch": 1.8430199660496323, "step": 11400}, {"loss": 0.7972, "grad_norm": 0.6099604964256287, "learning_rate": 0.0002, "epoch": 1.8446366502303775, "step": 11410}, {"loss": 0.7187, "grad_norm": 0.7338208556175232, "learning_rate": 0.0002, "epoch": 1.8462533344111227, "step": 11420}, {"loss": 0.7007, "grad_norm": 0.7534668445587158, "learning_rate": 0.0002, "epoch": 1.847870018591868, "step": 11430}, {"loss": 0.7464, "grad_norm": 0.6135470271110535, "learning_rate": 0.0002, "epoch": 1.8494867027726134, "step": 11440}, {"loss": 0.7955, "grad_norm": 0.6229309439659119, "learning_rate": 0.0002, "epoch": 1.8511033869533586, "step": 11450}, {"loss": 0.7594, "grad_norm": 0.706423282623291, "learning_rate": 0.0002, "epoch": 1.852720071134104, "step": 11460}, {"loss": 0.7411, "grad_norm": 0.5460049510002136, "learning_rate": 0.0002, "epoch": 1.8543367553148493, "step": 11470}, {"loss": 0.7416, "grad_norm": 0.6616711020469666, "learning_rate": 0.0002, "epoch": 1.8559534394955945, "step": 11480}, {"loss": 0.729, "grad_norm": 0.6372783184051514, "learning_rate": 0.0002, "epoch": 1.8575701236763398, "step": 11490}, {"loss": 0.7333, "grad_norm": 0.7162668108940125, "learning_rate": 0.0002, "epoch": 1.859186807857085, "step": 11500}, {"loss": 0.7747, "grad_norm": 0.6605209708213806, "learning_rate": 0.0002, "epoch": 1.8608034920378305, "step": 11510}, {"loss": 0.7258, "grad_norm": 0.6933956742286682, "learning_rate": 0.0002, "epoch": 1.8624201762185757, "step": 11520}, {"loss": 0.7243, "grad_norm": 0.6582090854644775, "learning_rate": 0.0002, "epoch": 1.8640368603993211, "step": 11530}, {"loss": 0.7313, "grad_norm": 0.6416500806808472, "learning_rate": 0.0002, "epoch": 1.8656535445800664, "step": 11540}, {"loss": 0.7372, "grad_norm": 0.5434312224388123, "learning_rate": 0.0002, "epoch": 1.8672702287608116, "step": 11550}, {"loss": 0.7635, "grad_norm": 0.6827567219734192, "learning_rate": 0.0002, "epoch": 1.8688869129415568, "step": 11560}, {"loss": 0.7137, "grad_norm": 0.7354370951652527, "learning_rate": 0.0002, "epoch": 1.870503597122302, "step": 11570}, {"loss": 0.7526, "grad_norm": 0.590372622013092, "learning_rate": 0.0002, "epoch": 1.8721202813030473, "step": 11580}, {"loss": 0.731, "grad_norm": 0.853183925151825, "learning_rate": 0.0002, "epoch": 1.8737369654837928, "step": 11590}, {"loss": 0.7487, "grad_norm": 0.822678804397583, "learning_rate": 0.0002, "epoch": 1.875353649664538, "step": 11600}, {"loss": 0.7427, "grad_norm": 0.6591550707817078, "learning_rate": 0.0002, "epoch": 1.8769703338452834, "step": 11610}, {"loss": 0.7054, "grad_norm": 0.7475301623344421, "learning_rate": 0.0002, "epoch": 1.8785870180260287, "step": 11620}, {"loss": 0.811, "grad_norm": 0.6390765309333801, "learning_rate": 0.0002, "epoch": 1.880203702206774, "step": 11630}, {"loss": 0.7531, "grad_norm": 0.6589758992195129, "learning_rate": 0.0002, "epoch": 1.8818203863875191, "step": 11640}, {"loss": 0.7475, "grad_norm": 0.6765508651733398, "learning_rate": 0.0002, "epoch": 1.8834370705682644, "step": 11650}, {"loss": 0.738, "grad_norm": 0.6527857780456543, "learning_rate": 0.0002, "epoch": 1.8850537547490098, "step": 11660}, {"loss": 0.7504, "grad_norm": 0.6642923951148987, "learning_rate": 0.0002, "epoch": 1.886670438929755, "step": 11670}, {"loss": 0.7701, "grad_norm": 0.6945584416389465, "learning_rate": 0.0002, "epoch": 1.8882871231105005, "step": 11680}, {"loss": 0.7711, "grad_norm": 0.694018542766571, "learning_rate": 0.0002, "epoch": 1.8899038072912457, "step": 11690}, {"loss": 0.7195, "grad_norm": 0.7237417101860046, "learning_rate": 0.0002, "epoch": 1.891520491471991, "step": 11700}, {"loss": 0.7491, "grad_norm": 0.7401309609413147, "learning_rate": 0.0002, "epoch": 1.8931371756527362, "step": 11710}, {"loss": 0.805, "grad_norm": 0.6537784337997437, "learning_rate": 0.0002, "epoch": 1.8947538598334814, "step": 11720}, {"loss": 0.793, "grad_norm": 0.7398539185523987, "learning_rate": 0.0002, "epoch": 1.8963705440142267, "step": 11730}, {"loss": 0.7561, "grad_norm": 0.6696075797080994, "learning_rate": 0.0002, "epoch": 1.8979872281949721, "step": 11740}, {"loss": 0.7353, "grad_norm": 0.6014142036437988, "learning_rate": 0.0002, "epoch": 1.8996039123757174, "step": 11750}, {"loss": 0.7714, "grad_norm": 0.7023524641990662, "learning_rate": 0.0002, "epoch": 1.9012205965564628, "step": 11760}, {"loss": 0.7088, "grad_norm": 0.739973783493042, "learning_rate": 0.0002, "epoch": 1.902837280737208, "step": 11770}, {"loss": 0.7848, "grad_norm": 0.5576770901679993, "learning_rate": 0.0002, "epoch": 1.9044539649179533, "step": 11780}, {"loss": 0.7483, "grad_norm": 0.6907393932342529, "learning_rate": 0.0002, "epoch": 1.9060706490986985, "step": 11790}, {"loss": 0.7827, "grad_norm": 0.6934581995010376, "learning_rate": 0.0002, "epoch": 1.9076873332794437, "step": 11800}, {"loss": 0.7199, "grad_norm": 0.591774582862854, "learning_rate": 0.0002, "epoch": 1.9093040174601892, "step": 11810}, {"loss": 0.7333, "grad_norm": 0.6249791383743286, "learning_rate": 0.0002, "epoch": 1.9109207016409344, "step": 11820}, {"loss": 0.7581, "grad_norm": 0.6755744218826294, "learning_rate": 0.0002, "epoch": 1.9125373858216799, "step": 11830}, {"loss": 0.696, "grad_norm": 0.7286285161972046, "learning_rate": 0.0002, "epoch": 1.914154070002425, "step": 11840}, {"loss": 0.7509, "grad_norm": 0.7867850065231323, "learning_rate": 0.0002, "epoch": 1.9157707541831703, "step": 11850}, {"loss": 0.735, "grad_norm": 0.6283972859382629, "learning_rate": 0.0002, "epoch": 1.9173874383639156, "step": 11860}, {"loss": 0.7296, "grad_norm": 0.605823814868927, "learning_rate": 0.0002, "epoch": 1.9190041225446608, "step": 11870}, {"loss": 0.6598, "grad_norm": 0.5927976965904236, "learning_rate": 0.0002, "epoch": 1.920620806725406, "step": 11880}, {"loss": 0.7649, "grad_norm": 0.5974002480506897, "learning_rate": 0.0002, "epoch": 1.9222374909061515, "step": 11890}, {"loss": 0.7843, "grad_norm": 0.7091866135597229, "learning_rate": 0.0002, "epoch": 1.923854175086897, "step": 11900}, {"loss": 0.775, "grad_norm": 0.72496497631073, "learning_rate": 0.0002, "epoch": 1.9254708592676422, "step": 11910}, {"loss": 0.7153, "grad_norm": 0.6131896376609802, "learning_rate": 0.0002, "epoch": 1.9270875434483874, "step": 11920}, {"loss": 0.7228, "grad_norm": 0.6556436419487, "learning_rate": 0.0002, "epoch": 1.9287042276291326, "step": 11930}, {"loss": 0.7319, "grad_norm": 0.622932493686676, "learning_rate": 0.0002, "epoch": 1.9303209118098779, "step": 11940}, {"loss": 0.7592, "grad_norm": 0.6618631482124329, "learning_rate": 0.0002, "epoch": 1.931937595990623, "step": 11950}, {"loss": 0.8332, "grad_norm": 0.630966305732727, "learning_rate": 0.0002, "epoch": 1.9335542801713685, "step": 11960}, {"loss": 0.6854, "grad_norm": 0.6336734890937805, "learning_rate": 0.0002, "epoch": 1.9351709643521138, "step": 11970}, {"loss": 0.7433, "grad_norm": 0.655403196811676, "learning_rate": 0.0002, "epoch": 1.9367876485328592, "step": 11980}, {"loss": 0.7282, "grad_norm": 0.5640574097633362, "learning_rate": 0.0002, "epoch": 1.9384043327136045, "step": 11990}, {"loss": 0.7289, "grad_norm": 0.6322951316833496, "learning_rate": 0.0002, "epoch": 1.9400210168943497, "step": 12000}, {"loss": 0.7627, "grad_norm": 0.615703821182251, "learning_rate": 0.0002, "epoch": 1.941637701075095, "step": 12010}, {"loss": 0.786, "grad_norm": 0.6487536430358887, "learning_rate": 0.0002, "epoch": 1.9432543852558402, "step": 12020}, {"loss": 0.7435, "grad_norm": 0.9209630489349365, "learning_rate": 0.0002, "epoch": 1.9448710694365856, "step": 12030}, {"loss": 0.7274, "grad_norm": 0.67485511302948, "learning_rate": 0.0002, "epoch": 1.9464877536173308, "step": 12040}, {"loss": 0.7551, "grad_norm": 0.6831230521202087, "learning_rate": 0.0002, "epoch": 1.9481044377980763, "step": 12050}, {"loss": 0.7546, "grad_norm": 0.6578302383422852, "learning_rate": 0.0002, "epoch": 1.9497211219788215, "step": 12060}, {"loss": 0.6989, "grad_norm": 0.9975938200950623, "learning_rate": 0.0002, "epoch": 1.9513378061595668, "step": 12070}, {"loss": 0.7952, "grad_norm": 0.6637365221977234, "learning_rate": 0.0002, "epoch": 1.952954490340312, "step": 12080}, {"loss": 0.7482, "grad_norm": 0.605707049369812, "learning_rate": 0.0002, "epoch": 1.9545711745210572, "step": 12090}, {"loss": 0.7768, "grad_norm": 0.6584440469741821, "learning_rate": 0.0002, "epoch": 1.9561878587018025, "step": 12100}, {"loss": 0.7187, "grad_norm": 0.6070835590362549, "learning_rate": 0.0002, "epoch": 1.957804542882548, "step": 12110}, {"loss": 0.7491, "grad_norm": 0.7862601280212402, "learning_rate": 0.0002, "epoch": 1.9594212270632931, "step": 12120}, {"loss": 0.7972, "grad_norm": 0.8175255060195923, "learning_rate": 0.0002, "epoch": 1.9610379112440386, "step": 12130}, {"loss": 0.7242, "grad_norm": 0.5648472905158997, "learning_rate": 0.0002, "epoch": 1.9626545954247838, "step": 12140}, {"loss": 0.7321, "grad_norm": 0.6591973304748535, "learning_rate": 0.0002, "epoch": 1.964271279605529, "step": 12150}, {"loss": 0.739, "grad_norm": 0.5960676074028015, "learning_rate": 0.0002, "epoch": 1.9658879637862743, "step": 12160}, {"loss": 0.7254, "grad_norm": 0.7272544503211975, "learning_rate": 0.0002, "epoch": 1.9675046479670195, "step": 12170}, {"loss": 0.7376, "grad_norm": 0.7176699042320251, "learning_rate": 0.0002, "epoch": 1.969121332147765, "step": 12180}, {"loss": 0.7525, "grad_norm": 0.6927123665809631, "learning_rate": 0.0002, "epoch": 1.9707380163285102, "step": 12190}, {"loss": 0.7318, "grad_norm": 0.5536034107208252, "learning_rate": 0.0002, "epoch": 1.9723547005092557, "step": 12200}, {"loss": 0.7737, "grad_norm": 0.8348390460014343, "learning_rate": 0.0002, "epoch": 1.9739713846900009, "step": 12210}, {"loss": 0.7494, "grad_norm": 0.6591181755065918, "learning_rate": 0.0002, "epoch": 1.9755880688707461, "step": 12220}, {"loss": 0.763, "grad_norm": 1.0624109506607056, "learning_rate": 0.0002, "epoch": 1.9772047530514913, "step": 12230}, {"loss": 0.7541, "grad_norm": 0.9265586137771606, "learning_rate": 0.0002, "epoch": 1.9788214372322366, "step": 12240}, {"loss": 0.7533, "grad_norm": 0.5998196005821228, "learning_rate": 0.0002, "epoch": 1.9804381214129818, "step": 12250}, {"loss": 0.7225, "grad_norm": 0.6960851550102234, "learning_rate": 0.0002, "epoch": 1.9820548055937273, "step": 12260}, {"loss": 0.7398, "grad_norm": 0.7674502730369568, "learning_rate": 0.0002, "epoch": 1.9836714897744727, "step": 12270}, {"loss": 0.7185, "grad_norm": 0.6407275795936584, "learning_rate": 0.0002, "epoch": 1.985288173955218, "step": 12280}, {"loss": 0.7382, "grad_norm": 0.6673079133033752, "learning_rate": 0.0002, "epoch": 1.9869048581359632, "step": 12290}, {"loss": 0.7326, "grad_norm": 0.6989844441413879, "learning_rate": 0.0002, "epoch": 1.9885215423167084, "step": 12300}, {"loss": 0.7559, "grad_norm": 0.7564442157745361, "learning_rate": 0.0002, "epoch": 1.9901382264974536, "step": 12310}, {"loss": 0.7719, "grad_norm": 0.6385478973388672, "learning_rate": 0.0002, "epoch": 1.9917549106781989, "step": 12320}, {"loss": 0.7369, "grad_norm": 0.7193717956542969, "learning_rate": 0.0002, "epoch": 1.9933715948589443, "step": 12330}, {"loss": 0.7583, "grad_norm": 0.7987112402915955, "learning_rate": 0.0002, "epoch": 1.9949882790396896, "step": 12340}, {"loss": 0.7793, "grad_norm": 0.7260826826095581, "learning_rate": 0.0002, "epoch": 1.996604963220435, "step": 12350}, {"loss": 0.7505, "grad_norm": 0.7968255281448364, "learning_rate": 0.0002, "epoch": 1.9982216474011802, "step": 12360}, {"loss": 0.717, "grad_norm": 0.6893062591552734, "learning_rate": 0.0002, "epoch": 1.9998383315819255, "step": 12370}, {"eval_loss": 1.1044032573699951, "eval_runtime": 122.1508, "eval_samples_per_second": 6.001, "eval_steps_per_second": 0.753, "epoch": 2.0, "step": 12371}, {"loss": 0.6604, "grad_norm": 0.7775409817695618, "learning_rate": 0.0002, "epoch": 2.0014550157626707, "step": 12380}, {"loss": 0.6845, "grad_norm": 0.76218581199646, "learning_rate": 0.0002, "epoch": 2.003071699943416, "step": 12390}, {"loss": 0.6909, "grad_norm": 0.5677764415740967, "learning_rate": 0.0002, "epoch": 2.004688384124161, "step": 12400}, {"loss": 0.6584, "grad_norm": 0.808442234992981, "learning_rate": 0.0002, "epoch": 2.006305068304907, "step": 12410}, {"loss": 0.659, "grad_norm": 0.7144765257835388, "learning_rate": 0.0002, "epoch": 2.007921752485652, "step": 12420}, {"loss": 0.6666, "grad_norm": 0.6914031505584717, "learning_rate": 0.0002, "epoch": 2.0095384366663973, "step": 12430}, {"loss": 0.6596, "grad_norm": 0.7581454515457153, "learning_rate": 0.0002, "epoch": 2.0111551208471425, "step": 12440}, {"loss": 0.6785, "grad_norm": 0.8388504981994629, "learning_rate": 0.0002, "epoch": 2.0127718050278878, "step": 12450}, {"loss": 0.6942, "grad_norm": 0.6716406941413879, "learning_rate": 0.0002, "epoch": 2.014388489208633, "step": 12460}, {"loss": 0.6441, "grad_norm": 0.898902416229248, "learning_rate": 0.0002, "epoch": 2.0160051733893782, "step": 12470}, {"loss": 0.6655, "grad_norm": 0.6432679891586304, "learning_rate": 0.0002, "epoch": 2.0176218575701235, "step": 12480}, {"loss": 0.6521, "grad_norm": 0.8021109104156494, "learning_rate": 0.0002, "epoch": 2.019238541750869, "step": 12490}, {"loss": 0.6581, "grad_norm": 0.7039216756820679, "learning_rate": 0.0002, "epoch": 2.0208552259316144, "step": 12500}, {"loss": 0.6521, "grad_norm": 0.646531879901886, "learning_rate": 0.0002, "epoch": 2.0224719101123596, "step": 12510}, {"loss": 0.6302, "grad_norm": 0.783704400062561, "learning_rate": 0.0002, "epoch": 2.024088594293105, "step": 12520}, {"loss": 0.6288, "grad_norm": 0.8805046677589417, "learning_rate": 0.0002, "epoch": 2.02570527847385, "step": 12530}, {"loss": 0.6288, "grad_norm": 0.7289270758628845, "learning_rate": 0.0002, "epoch": 2.0273219626545953, "step": 12540}, {"loss": 0.6663, "grad_norm": 0.71653151512146, "learning_rate": 0.0002, "epoch": 2.0289386468353405, "step": 12550}, {"loss": 0.625, "grad_norm": 0.73281329870224, "learning_rate": 0.0002, "epoch": 2.030555331016086, "step": 12560}, {"loss": 0.6448, "grad_norm": 0.6657090187072754, "learning_rate": 0.0002, "epoch": 2.0321720151968314, "step": 12570}, {"loss": 0.6983, "grad_norm": 0.8241133093833923, "learning_rate": 0.0002, "epoch": 2.0337886993775767, "step": 12580}, {"loss": 0.6488, "grad_norm": 0.5834135413169861, "learning_rate": 0.0002, "epoch": 2.035405383558322, "step": 12590}, {"loss": 0.6188, "grad_norm": 0.84502112865448, "learning_rate": 0.0002, "epoch": 2.037022067739067, "step": 12600}, {"loss": 0.6349, "grad_norm": 0.8952481746673584, "learning_rate": 0.0002, "epoch": 2.0386387519198124, "step": 12610}, {"loss": 0.6923, "grad_norm": 0.7801461815834045, "learning_rate": 0.0002, "epoch": 2.0402554361005576, "step": 12620}, {"loss": 0.6176, "grad_norm": 0.6788367033004761, "learning_rate": 0.0002, "epoch": 2.041872120281303, "step": 12630}, {"loss": 0.6162, "grad_norm": 0.7241756319999695, "learning_rate": 0.0002, "epoch": 2.0434888044620485, "step": 12640}, {"loss": 0.655, "grad_norm": 0.6933388113975525, "learning_rate": 0.0002, "epoch": 2.0451054886427937, "step": 12650}, {"loss": 0.6431, "grad_norm": 0.8029746413230896, "learning_rate": 0.0002, "epoch": 2.046722172823539, "step": 12660}, {"loss": 0.7164, "grad_norm": 0.946399986743927, "learning_rate": 0.0002, "epoch": 2.048338857004284, "step": 12670}, {"loss": 0.638, "grad_norm": 0.7072678804397583, "learning_rate": 0.0002, "epoch": 2.0499555411850294, "step": 12680}, {"loss": 0.6487, "grad_norm": 0.6810618042945862, "learning_rate": 0.0002, "epoch": 2.0515722253657747, "step": 12690}, {"loss": 0.6554, "grad_norm": 0.7661160230636597, "learning_rate": 0.0002, "epoch": 2.05318890954652, "step": 12700}, {"loss": 0.6799, "grad_norm": 0.6350653767585754, "learning_rate": 0.0002, "epoch": 2.0548055937272656, "step": 12710}, {"loss": 0.6654, "grad_norm": 0.861890971660614, "learning_rate": 0.0002, "epoch": 2.056422277908011, "step": 12720}, {"loss": 0.6286, "grad_norm": 0.6489875912666321, "learning_rate": 0.0002, "epoch": 2.058038962088756, "step": 12730}, {"loss": 0.6811, "grad_norm": 0.8268506526947021, "learning_rate": 0.0002, "epoch": 2.0596556462695013, "step": 12740}, {"loss": 0.6524, "grad_norm": 0.607679545879364, "learning_rate": 0.0002, "epoch": 2.0612723304502465, "step": 12750}, {"loss": 0.6649, "grad_norm": 0.6754153370857239, "learning_rate": 0.0002, "epoch": 2.0628890146309917, "step": 12760}, {"loss": 0.6549, "grad_norm": 0.7263124585151672, "learning_rate": 0.0002, "epoch": 2.064505698811737, "step": 12770}, {"loss": 0.6189, "grad_norm": 0.6986154317855835, "learning_rate": 0.0002, "epoch": 2.0661223829924826, "step": 12780}, {"loss": 0.6723, "grad_norm": 0.7768576741218567, "learning_rate": 0.0002, "epoch": 2.067739067173228, "step": 12790}, {"loss": 0.677, "grad_norm": 0.7546762824058533, "learning_rate": 0.0002, "epoch": 2.069355751353973, "step": 12800}, {"loss": 0.6485, "grad_norm": 0.7588880062103271, "learning_rate": 0.0002, "epoch": 2.0709724355347183, "step": 12810}, {"loss": 0.6989, "grad_norm": 0.7457242608070374, "learning_rate": 0.0002, "epoch": 2.0725891197154636, "step": 12820}, {"loss": 0.6489, "grad_norm": 0.6983516812324524, "learning_rate": 0.0002, "epoch": 2.074205803896209, "step": 12830}, {"loss": 0.651, "grad_norm": 0.7950928807258606, "learning_rate": 0.0002, "epoch": 2.075822488076954, "step": 12840}, {"loss": 0.6603, "grad_norm": 0.9248087406158447, "learning_rate": 0.0002, "epoch": 2.0774391722576993, "step": 12850}, {"loss": 0.6847, "grad_norm": 0.7229493260383606, "learning_rate": 0.0002, "epoch": 2.079055856438445, "step": 12860}, {"loss": 0.6702, "grad_norm": 0.5710847973823547, "learning_rate": 0.0002, "epoch": 2.08067254061919, "step": 12870}, {"loss": 0.6974, "grad_norm": 0.9580423831939697, "learning_rate": 0.0002, "epoch": 2.0822892247999354, "step": 12880}, {"loss": 0.6341, "grad_norm": 0.7399665713310242, "learning_rate": 0.0002, "epoch": 2.0839059089806806, "step": 12890}, {"loss": 0.6993, "grad_norm": 0.7981410622596741, "learning_rate": 0.0002, "epoch": 2.085522593161426, "step": 12900}, {"loss": 0.6976, "grad_norm": 0.870759904384613, "learning_rate": 0.0002, "epoch": 2.087139277342171, "step": 12910}, {"loss": 0.7194, "grad_norm": 0.7001481652259827, "learning_rate": 0.0002, "epoch": 2.0887559615229163, "step": 12920}, {"loss": 0.6383, "grad_norm": 0.6745418310165405, "learning_rate": 0.0002, "epoch": 2.090372645703662, "step": 12930}, {"loss": 0.6519, "grad_norm": 0.7739067673683167, "learning_rate": 0.0002, "epoch": 2.0919893298844072, "step": 12940}, {"loss": 0.6856, "grad_norm": 0.6742934584617615, "learning_rate": 0.0002, "epoch": 2.0936060140651525, "step": 12950}, {"loss": 0.6279, "grad_norm": 0.7270349860191345, "learning_rate": 0.0002, "epoch": 2.0952226982458977, "step": 12960}, {"loss": 0.6783, "grad_norm": 0.7150624394416809, "learning_rate": 0.0002, "epoch": 2.096839382426643, "step": 12970}, {"loss": 0.6093, "grad_norm": 0.7734767198562622, "learning_rate": 0.0002, "epoch": 2.098456066607388, "step": 12980}, {"loss": 0.6534, "grad_norm": 0.7618662118911743, "learning_rate": 0.0002, "epoch": 2.1000727507881334, "step": 12990}, {"loss": 0.6707, "grad_norm": 0.6557944416999817, "learning_rate": 0.0002, "epoch": 2.101689434968879, "step": 13000}, {"loss": 0.7268, "grad_norm": 0.8786448240280151, "learning_rate": 0.0002, "epoch": 2.1033061191496243, "step": 13010}, {"loss": 0.6677, "grad_norm": 0.6878724098205566, "learning_rate": 0.0002, "epoch": 2.1049228033303695, "step": 13020}, {"loss": 0.6824, "grad_norm": 0.822318971157074, "learning_rate": 0.0002, "epoch": 2.1065394875111147, "step": 13030}, {"loss": 0.6228, "grad_norm": 0.831468939781189, "learning_rate": 0.0002, "epoch": 2.10815617169186, "step": 13040}, {"loss": 0.6511, "grad_norm": 0.7699505686759949, "learning_rate": 0.0002, "epoch": 2.109772855872605, "step": 13050}, {"loss": 0.6671, "grad_norm": 0.7559016346931458, "learning_rate": 0.0002, "epoch": 2.1113895400533504, "step": 13060}, {"loss": 0.6215, "grad_norm": 0.6942209601402283, "learning_rate": 0.0002, "epoch": 2.1130062242340957, "step": 13070}, {"loss": 0.6449, "grad_norm": 0.6098947525024414, "learning_rate": 0.0002, "epoch": 2.1146229084148414, "step": 13080}, {"loss": 0.7091, "grad_norm": 0.6499016284942627, "learning_rate": 0.0002, "epoch": 2.1162395925955866, "step": 13090}, {"loss": 0.6247, "grad_norm": 0.7719953060150146, "learning_rate": 0.0002, "epoch": 2.117856276776332, "step": 13100}, {"loss": 0.6064, "grad_norm": 0.6708134412765503, "learning_rate": 0.0002, "epoch": 2.119472960957077, "step": 13110}, {"loss": 0.6056, "grad_norm": 0.8119585514068604, "learning_rate": 0.0002, "epoch": 2.1210896451378223, "step": 13120}, {"loss": 0.6628, "grad_norm": 0.6947157979011536, "learning_rate": 0.0002, "epoch": 2.1227063293185675, "step": 13130}, {"loss": 0.6375, "grad_norm": 0.8831837773323059, "learning_rate": 0.0002, "epoch": 2.1243230134993127, "step": 13140}, {"loss": 0.6997, "grad_norm": 0.7266910672187805, "learning_rate": 0.0002, "epoch": 2.1259396976800584, "step": 13150}, {"loss": 0.6446, "grad_norm": 0.8864351511001587, "learning_rate": 0.0002, "epoch": 2.1275563818608036, "step": 13160}, {"loss": 0.6762, "grad_norm": 0.8104248046875, "learning_rate": 0.0002, "epoch": 2.129173066041549, "step": 13170}, {"loss": 0.6581, "grad_norm": 0.6077079772949219, "learning_rate": 0.0002, "epoch": 2.130789750222294, "step": 13180}, {"loss": 0.6572, "grad_norm": 0.6874213814735413, "learning_rate": 0.0002, "epoch": 2.1324064344030393, "step": 13190}, {"loss": 0.642, "grad_norm": 0.7134367823600769, "learning_rate": 0.0002, "epoch": 2.1340231185837846, "step": 13200}, {"loss": 0.7016, "grad_norm": 0.6101235151290894, "learning_rate": 0.0002, "epoch": 2.13563980276453, "step": 13210}, {"loss": 0.6529, "grad_norm": 0.6042411923408508, "learning_rate": 0.0002, "epoch": 2.137256486945275, "step": 13220}, {"loss": 0.7179, "grad_norm": 0.914601743221283, "learning_rate": 0.0002, "epoch": 2.1388731711260207, "step": 13230}, {"loss": 0.6513, "grad_norm": 0.7104284167289734, "learning_rate": 0.0002, "epoch": 2.140489855306766, "step": 13240}, {"loss": 0.6607, "grad_norm": 0.664395272731781, "learning_rate": 0.0002, "epoch": 2.142106539487511, "step": 13250}, {"loss": 0.7211, "grad_norm": 0.6991241574287415, "learning_rate": 0.0002, "epoch": 2.1437232236682564, "step": 13260}, {"loss": 0.6484, "grad_norm": 0.5469560623168945, "learning_rate": 0.0002, "epoch": 2.1453399078490016, "step": 13270}, {"loss": 0.6765, "grad_norm": 0.8454998135566711, "learning_rate": 0.0002, "epoch": 2.146956592029747, "step": 13280}, {"loss": 0.6683, "grad_norm": 0.7088868618011475, "learning_rate": 0.0002, "epoch": 2.148573276210492, "step": 13290}, {"loss": 0.6835, "grad_norm": 0.7002687454223633, "learning_rate": 0.0002, "epoch": 2.1501899603912378, "step": 13300}, {"loss": 0.6399, "grad_norm": 0.7785214781761169, "learning_rate": 0.0002, "epoch": 2.151806644571983, "step": 13310}, {"loss": 0.67, "grad_norm": 0.8049132227897644, "learning_rate": 0.0002, "epoch": 2.1534233287527282, "step": 13320}, {"loss": 0.6495, "grad_norm": 0.8062595129013062, "learning_rate": 0.0002, "epoch": 2.1550400129334735, "step": 13330}, {"loss": 0.6603, "grad_norm": 0.6208319067955017, "learning_rate": 0.0002, "epoch": 2.1566566971142187, "step": 13340}, {"loss": 0.6584, "grad_norm": 0.7519655823707581, "learning_rate": 0.0002, "epoch": 2.158273381294964, "step": 13350}, {"loss": 0.6457, "grad_norm": 0.7645747065544128, "learning_rate": 0.0002, "epoch": 2.159890065475709, "step": 13360}, {"loss": 0.645, "grad_norm": 0.6847302913665771, "learning_rate": 0.0002, "epoch": 2.1615067496564544, "step": 13370}, {"loss": 0.6903, "grad_norm": 0.8630441427230835, "learning_rate": 0.0002, "epoch": 2.1631234338372, "step": 13380}, {"loss": 0.6742, "grad_norm": 0.7947702407836914, "learning_rate": 0.0002, "epoch": 2.1647401180179453, "step": 13390}, {"loss": 0.7206, "grad_norm": 0.6836977005004883, "learning_rate": 0.0002, "epoch": 2.1663568021986905, "step": 13400}, {"loss": 0.6304, "grad_norm": 0.7340566515922546, "learning_rate": 0.0002, "epoch": 2.1679734863794358, "step": 13410}, {"loss": 0.6528, "grad_norm": 0.7075738906860352, "learning_rate": 0.0002, "epoch": 2.169590170560181, "step": 13420}, {"loss": 0.6585, "grad_norm": 0.7080879807472229, "learning_rate": 0.0002, "epoch": 2.1712068547409262, "step": 13430}, {"loss": 0.6615, "grad_norm": 0.6218613386154175, "learning_rate": 0.0002, "epoch": 2.1728235389216715, "step": 13440}, {"loss": 0.6488, "grad_norm": 0.8211479187011719, "learning_rate": 0.0002, "epoch": 2.174440223102417, "step": 13450}, {"loss": 0.6738, "grad_norm": 0.864466667175293, "learning_rate": 0.0002, "epoch": 2.1760569072831624, "step": 13460}, {"loss": 0.679, "grad_norm": 0.7943857908248901, "learning_rate": 0.0002, "epoch": 2.1776735914639076, "step": 13470}, {"loss": 0.6838, "grad_norm": 0.78728187084198, "learning_rate": 0.0002, "epoch": 2.179290275644653, "step": 13480}, {"loss": 0.6397, "grad_norm": 0.697527289390564, "learning_rate": 0.0002, "epoch": 2.180906959825398, "step": 13490}, {"loss": 0.669, "grad_norm": 0.8205804228782654, "learning_rate": 0.0002, "epoch": 2.1825236440061433, "step": 13500}, {"loss": 0.7227, "grad_norm": 0.8709042072296143, "learning_rate": 0.0002, "epoch": 2.1841403281868885, "step": 13510}, {"loss": 0.6313, "grad_norm": 0.6228537559509277, "learning_rate": 0.0002, "epoch": 2.1857570123676338, "step": 13520}, {"loss": 0.7025, "grad_norm": 0.9566980004310608, "learning_rate": 0.0002, "epoch": 2.1873736965483794, "step": 13530}, {"loss": 0.6755, "grad_norm": 0.7128894329071045, "learning_rate": 0.0002, "epoch": 2.1889903807291247, "step": 13540}, {"loss": 0.6827, "grad_norm": 0.6888654232025146, "learning_rate": 0.0002, "epoch": 2.19060706490987, "step": 13550}, {"loss": 0.6961, "grad_norm": 0.6444337368011475, "learning_rate": 0.0002, "epoch": 2.192223749090615, "step": 13560}, {"loss": 0.656, "grad_norm": 0.8008806705474854, "learning_rate": 0.0002, "epoch": 2.1938404332713604, "step": 13570}, {"loss": 0.7, "grad_norm": 0.8482748866081238, "learning_rate": 0.0002, "epoch": 2.1954571174521056, "step": 13580}, {"loss": 0.7326, "grad_norm": 0.8584157228469849, "learning_rate": 0.0002, "epoch": 2.197073801632851, "step": 13590}, {"loss": 0.7014, "grad_norm": 0.7513734698295593, "learning_rate": 0.0002, "epoch": 2.1986904858135965, "step": 13600}, {"loss": 0.6632, "grad_norm": 0.7864262461662292, "learning_rate": 0.0002, "epoch": 2.2003071699943417, "step": 13610}, {"loss": 0.6879, "grad_norm": 0.8493645191192627, "learning_rate": 0.0002, "epoch": 2.201923854175087, "step": 13620}, {"loss": 0.6617, "grad_norm": 0.6902140974998474, "learning_rate": 0.0002, "epoch": 2.203540538355832, "step": 13630}, {"loss": 0.6655, "grad_norm": 0.8711254596710205, "learning_rate": 0.0002, "epoch": 2.2051572225365774, "step": 13640}, {"loss": 0.6359, "grad_norm": 0.7832191586494446, "learning_rate": 0.0002, "epoch": 2.2067739067173227, "step": 13650}, {"loss": 0.6723, "grad_norm": 0.5668176412582397, "learning_rate": 0.0002, "epoch": 2.208390590898068, "step": 13660}, {"loss": 0.635, "grad_norm": 0.8648375272750854, "learning_rate": 0.0002, "epoch": 2.2100072750788136, "step": 13670}, {"loss": 0.653, "grad_norm": 0.7643089890480042, "learning_rate": 0.0002, "epoch": 2.211623959259559, "step": 13680}, {"loss": 0.6765, "grad_norm": 0.6293777823448181, "learning_rate": 0.0002, "epoch": 2.213240643440304, "step": 13690}, {"loss": 0.6842, "grad_norm": 0.6459372639656067, "learning_rate": 0.0002, "epoch": 2.2148573276210493, "step": 13700}, {"loss": 0.6526, "grad_norm": 0.7060744166374207, "learning_rate": 0.0002, "epoch": 2.2164740118017945, "step": 13710}, {"loss": 0.7101, "grad_norm": 0.674109160900116, "learning_rate": 0.0002, "epoch": 2.2180906959825397, "step": 13720}, {"loss": 0.6529, "grad_norm": 0.830392062664032, "learning_rate": 0.0002, "epoch": 2.219707380163285, "step": 13730}, {"loss": 0.6733, "grad_norm": 0.6474477052688599, "learning_rate": 0.0002, "epoch": 2.2213240643440306, "step": 13740}, {"loss": 0.6413, "grad_norm": 0.7037909626960754, "learning_rate": 0.0002, "epoch": 2.222940748524776, "step": 13750}, {"loss": 0.6417, "grad_norm": 0.6554131507873535, "learning_rate": 0.0002, "epoch": 2.224557432705521, "step": 13760}, {"loss": 0.6907, "grad_norm": 0.7822230458259583, "learning_rate": 0.0002, "epoch": 2.2261741168862663, "step": 13770}, {"loss": 0.6505, "grad_norm": 0.9082167744636536, "learning_rate": 0.0002, "epoch": 2.2277908010670116, "step": 13780}, {"loss": 0.6878, "grad_norm": 0.7918276190757751, "learning_rate": 0.0002, "epoch": 2.229407485247757, "step": 13790}, {"loss": 0.6669, "grad_norm": 0.7354569435119629, "learning_rate": 0.0002, "epoch": 2.231024169428502, "step": 13800}, {"loss": 0.6503, "grad_norm": 0.8265249133110046, "learning_rate": 0.0002, "epoch": 2.2326408536092472, "step": 13810}, {"loss": 0.6871, "grad_norm": 0.6653847098350525, "learning_rate": 0.0002, "epoch": 2.234257537789993, "step": 13820}, {"loss": 0.6413, "grad_norm": 0.7157923579216003, "learning_rate": 0.0002, "epoch": 2.235874221970738, "step": 13830}, {"loss": 0.6306, "grad_norm": 0.7110323309898376, "learning_rate": 0.0002, "epoch": 2.2374909061514834, "step": 13840}, {"loss": 0.6913, "grad_norm": 0.7155357599258423, "learning_rate": 0.0002, "epoch": 2.2391075903322286, "step": 13850}, {"loss": 0.6579, "grad_norm": 1.0177817344665527, "learning_rate": 0.0002, "epoch": 2.240724274512974, "step": 13860}, {"loss": 0.635, "grad_norm": 0.7601948380470276, "learning_rate": 0.0002, "epoch": 2.242340958693719, "step": 13870}, {"loss": 0.6679, "grad_norm": 0.7628820538520813, "learning_rate": 0.0002, "epoch": 2.2439576428744643, "step": 13880}, {"loss": 0.6805, "grad_norm": 0.7089297771453857, "learning_rate": 0.0002, "epoch": 2.24557432705521, "step": 13890}, {"loss": 0.7236, "grad_norm": 0.695178210735321, "learning_rate": 0.0002, "epoch": 2.247191011235955, "step": 13900}, {"loss": 0.7084, "grad_norm": 0.7631948590278625, "learning_rate": 0.0002, "epoch": 2.2488076954167004, "step": 13910}, {"loss": 0.685, "grad_norm": 0.8203101754188538, "learning_rate": 0.0002, "epoch": 2.2504243795974457, "step": 13920}, {"loss": 0.653, "grad_norm": 0.8099079728126526, "learning_rate": 0.0002, "epoch": 2.252041063778191, "step": 13930}, {"loss": 0.694, "grad_norm": 0.6498546004295349, "learning_rate": 0.0002, "epoch": 2.253657747958936, "step": 13940}, {"loss": 0.6684, "grad_norm": 0.7797415256500244, "learning_rate": 0.0002, "epoch": 2.2552744321396814, "step": 13950}, {"loss": 0.683, "grad_norm": 0.8254124522209167, "learning_rate": 0.0002, "epoch": 2.2568911163204266, "step": 13960}, {"loss": 0.6806, "grad_norm": 0.6327953338623047, "learning_rate": 0.0002, "epoch": 2.2585078005011723, "step": 13970}, {"loss": 0.668, "grad_norm": 0.734194278717041, "learning_rate": 0.0002, "epoch": 2.2601244846819175, "step": 13980}, {"loss": 0.6912, "grad_norm": 0.9014202952384949, "learning_rate": 0.0002, "epoch": 2.2617411688626627, "step": 13990}, {"loss": 0.692, "grad_norm": 0.7643631100654602, "learning_rate": 0.0002, "epoch": 2.263357853043408, "step": 14000}, {"loss": 0.6657, "grad_norm": 0.8882834911346436, "learning_rate": 0.0002, "epoch": 2.264974537224153, "step": 14010}, {"loss": 0.6453, "grad_norm": 0.7975873351097107, "learning_rate": 0.0002, "epoch": 2.2665912214048984, "step": 14020}, {"loss": 0.7193, "grad_norm": 0.7765783071517944, "learning_rate": 0.0002, "epoch": 2.2682079055856437, "step": 14030}, {"loss": 0.662, "grad_norm": 0.8846288323402405, "learning_rate": 0.0002, "epoch": 2.2698245897663893, "step": 14040}, {"loss": 0.6494, "grad_norm": 0.9006744027137756, "learning_rate": 0.0002, "epoch": 2.2714412739471346, "step": 14050}, {"loss": 0.6423, "grad_norm": 0.7420173287391663, "learning_rate": 0.0002, "epoch": 2.27305795812788, "step": 14060}, {"loss": 0.7068, "grad_norm": 0.7956424951553345, "learning_rate": 0.0002, "epoch": 2.274674642308625, "step": 14070}, {"loss": 0.6581, "grad_norm": 0.7783209085464478, "learning_rate": 0.0002, "epoch": 2.2762913264893703, "step": 14080}, {"loss": 0.7202, "grad_norm": 0.7597188949584961, "learning_rate": 0.0002, "epoch": 2.2779080106701155, "step": 14090}, {"loss": 0.6778, "grad_norm": 0.6718921661376953, "learning_rate": 0.0002, "epoch": 2.2795246948508607, "step": 14100}, {"loss": 0.632, "grad_norm": 0.7528082132339478, "learning_rate": 0.0002, "epoch": 2.281141379031606, "step": 14110}, {"loss": 0.7608, "grad_norm": 0.8379864692687988, "learning_rate": 0.0002, "epoch": 2.2827580632123516, "step": 14120}, {"loss": 0.6767, "grad_norm": 0.748613715171814, "learning_rate": 0.0002, "epoch": 2.284374747393097, "step": 14130}, {"loss": 0.6641, "grad_norm": 0.7435423135757446, "learning_rate": 0.0002, "epoch": 2.285991431573842, "step": 14140}, {"loss": 0.6849, "grad_norm": 0.7580803632736206, "learning_rate": 0.0002, "epoch": 2.2876081157545873, "step": 14150}, {"loss": 0.6604, "grad_norm": 0.6278321146965027, "learning_rate": 0.0002, "epoch": 2.2892247999353326, "step": 14160}, {"loss": 0.6573, "grad_norm": 0.7663896083831787, "learning_rate": 0.0002, "epoch": 2.290841484116078, "step": 14170}, {"loss": 0.6655, "grad_norm": 0.9716812372207642, "learning_rate": 0.0002, "epoch": 2.292458168296823, "step": 14180}, {"loss": 0.7067, "grad_norm": 0.8993458151817322, "learning_rate": 0.0002, "epoch": 2.2940748524775687, "step": 14190}, {"loss": 0.6172, "grad_norm": 0.6156117916107178, "learning_rate": 0.0002, "epoch": 2.295691536658314, "step": 14200}, {"loss": 0.6318, "grad_norm": 0.8911278247833252, "learning_rate": 0.0002, "epoch": 2.297308220839059, "step": 14210}, {"loss": 0.6364, "grad_norm": 0.6422147154808044, "learning_rate": 0.0002, "epoch": 2.2989249050198044, "step": 14220}, {"loss": 0.6795, "grad_norm": 0.6866879463195801, "learning_rate": 0.0002, "epoch": 2.3005415892005496, "step": 14230}, {"loss": 0.6907, "grad_norm": 0.9297130107879639, "learning_rate": 0.0002, "epoch": 2.302158273381295, "step": 14240}, {"loss": 0.6823, "grad_norm": 0.7501356601715088, "learning_rate": 0.0002, "epoch": 2.30377495756204, "step": 14250}, {"loss": 0.6414, "grad_norm": 0.8363515138626099, "learning_rate": 0.0002, "epoch": 2.3053916417427853, "step": 14260}, {"loss": 0.6362, "grad_norm": 0.9083868265151978, "learning_rate": 0.0002, "epoch": 2.307008325923531, "step": 14270}, {"loss": 0.6862, "grad_norm": 0.7791516780853271, "learning_rate": 0.0002, "epoch": 2.3086250101042762, "step": 14280}, {"loss": 0.6569, "grad_norm": 0.8766953349113464, "learning_rate": 0.0002, "epoch": 2.3102416942850215, "step": 14290}, {"loss": 0.6698, "grad_norm": 0.7916635274887085, "learning_rate": 0.0002, "epoch": 2.3118583784657667, "step": 14300}, {"loss": 0.6927, "grad_norm": 0.627525269985199, "learning_rate": 0.0002, "epoch": 2.313475062646512, "step": 14310}, {"loss": 0.6541, "grad_norm": 0.8856783509254456, "learning_rate": 0.0002, "epoch": 2.315091746827257, "step": 14320}, {"loss": 0.6806, "grad_norm": 0.6758689284324646, "learning_rate": 0.0002, "epoch": 2.316708431008003, "step": 14330}, {"loss": 0.6794, "grad_norm": 0.6428321003913879, "learning_rate": 0.0002, "epoch": 2.318325115188748, "step": 14340}, {"loss": 0.682, "grad_norm": 0.9032121300697327, "learning_rate": 0.0002, "epoch": 2.3199417993694933, "step": 14350}, {"loss": 0.6569, "grad_norm": 0.8035986423492432, "learning_rate": 0.0002, "epoch": 2.3215584835502385, "step": 14360}, {"loss": 0.7067, "grad_norm": 0.7974579334259033, "learning_rate": 0.0002, "epoch": 2.3231751677309838, "step": 14370}, {"loss": 0.6451, "grad_norm": 0.8356034755706787, "learning_rate": 0.0002, "epoch": 2.324791851911729, "step": 14380}, {"loss": 0.6623, "grad_norm": 0.998760998249054, "learning_rate": 0.0002, "epoch": 2.326408536092474, "step": 14390}, {"loss": 0.649, "grad_norm": 0.6518142223358154, "learning_rate": 0.0002, "epoch": 2.3280252202732195, "step": 14400}, {"loss": 0.7146, "grad_norm": 0.7443506717681885, "learning_rate": 0.0002, "epoch": 2.3296419044539647, "step": 14410}, {"loss": 0.648, "grad_norm": 0.8436172604560852, "learning_rate": 0.0002, "epoch": 2.3312585886347104, "step": 14420}, {"loss": 0.6585, "grad_norm": 0.7411080598831177, "learning_rate": 0.0002, "epoch": 2.3328752728154556, "step": 14430}, {"loss": 0.6781, "grad_norm": 0.8839048743247986, "learning_rate": 0.0002, "epoch": 2.334491956996201, "step": 14440}, {"loss": 0.6565, "grad_norm": 0.8360885977745056, "learning_rate": 0.0002, "epoch": 2.336108641176946, "step": 14450}, {"loss": 0.6662, "grad_norm": 0.7608986496925354, "learning_rate": 0.0002, "epoch": 2.3377253253576913, "step": 14460}, {"loss": 0.6685, "grad_norm": 0.8179867267608643, "learning_rate": 0.0002, "epoch": 2.3393420095384365, "step": 14470}, {"loss": 0.7055, "grad_norm": 0.5989999771118164, "learning_rate": 0.0002, "epoch": 2.340958693719182, "step": 14480}, {"loss": 0.644, "grad_norm": 0.9450054168701172, "learning_rate": 0.0002, "epoch": 2.3425753778999274, "step": 14490}, {"loss": 0.6983, "grad_norm": 0.7885149717330933, "learning_rate": 0.0002, "epoch": 2.3441920620806727, "step": 14500}, {"loss": 0.6819, "grad_norm": 0.8152616620063782, "learning_rate": 0.0002, "epoch": 2.345808746261418, "step": 14510}, {"loss": 0.6989, "grad_norm": 0.7193838953971863, "learning_rate": 0.0002, "epoch": 2.347425430442163, "step": 14520}, {"loss": 0.6594, "grad_norm": 0.6701092720031738, "learning_rate": 0.0002, "epoch": 2.3490421146229084, "step": 14530}, {"loss": 0.6559, "grad_norm": 0.7529364228248596, "learning_rate": 0.0002, "epoch": 2.3506587988036536, "step": 14540}, {"loss": 0.6306, "grad_norm": 0.6599733829498291, "learning_rate": 0.0002, "epoch": 2.352275482984399, "step": 14550}, {"loss": 0.706, "grad_norm": 0.9502474069595337, "learning_rate": 0.0002, "epoch": 2.353892167165144, "step": 14560}, {"loss": 0.717, "grad_norm": 0.7619650959968567, "learning_rate": 0.0002, "epoch": 2.3555088513458897, "step": 14570}, {"loss": 0.6684, "grad_norm": 0.9854652285575867, "learning_rate": 0.0002, "epoch": 2.357125535526635, "step": 14580}, {"loss": 0.6455, "grad_norm": 0.727439284324646, "learning_rate": 0.0002, "epoch": 2.35874221970738, "step": 14590}, {"loss": 0.6645, "grad_norm": 0.6994746327400208, "learning_rate": 0.0002, "epoch": 2.3603589038881254, "step": 14600}, {"loss": 0.6587, "grad_norm": 0.7117531299591064, "learning_rate": 0.0002, "epoch": 2.3619755880688706, "step": 14610}, {"loss": 0.6804, "grad_norm": 0.6403067708015442, "learning_rate": 0.0002, "epoch": 2.363592272249616, "step": 14620}, {"loss": 0.7055, "grad_norm": 0.8377841711044312, "learning_rate": 0.0002, "epoch": 2.3652089564303616, "step": 14630}, {"loss": 0.6778, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 2.366825640611107, "step": 14640}, {"loss": 0.6552, "grad_norm": 0.8418586254119873, "learning_rate": 0.0002, "epoch": 2.368442324791852, "step": 14650}, {"loss": 0.6685, "grad_norm": 0.6178573369979858, "learning_rate": 0.0002, "epoch": 2.3700590089725972, "step": 14660}, {"loss": 0.6774, "grad_norm": 0.6368302702903748, "learning_rate": 0.0002, "epoch": 2.3716756931533425, "step": 14670}, {"loss": 0.6136, "grad_norm": 0.9122977256774902, "learning_rate": 0.0002, "epoch": 2.3732923773340877, "step": 14680}, {"loss": 0.6675, "grad_norm": 0.7086195349693298, "learning_rate": 0.0002, "epoch": 2.374909061514833, "step": 14690}, {"loss": 0.6582, "grad_norm": 0.7500800490379333, "learning_rate": 0.0002, "epoch": 2.376525745695578, "step": 14700}, {"loss": 0.6792, "grad_norm": 0.6634900569915771, "learning_rate": 0.0002, "epoch": 2.378142429876324, "step": 14710}, {"loss": 0.6614, "grad_norm": 0.839898407459259, "learning_rate": 0.0002, "epoch": 2.379759114057069, "step": 14720}, {"loss": 0.6453, "grad_norm": 0.7578426003456116, "learning_rate": 0.0002, "epoch": 2.3813757982378143, "step": 14730}, {"loss": 0.7282, "grad_norm": 1.0213173627853394, "learning_rate": 0.0002, "epoch": 2.3829924824185595, "step": 14740}, {"loss": 0.6704, "grad_norm": 0.7855949401855469, "learning_rate": 0.0002, "epoch": 2.3846091665993048, "step": 14750}, {"loss": 0.6694, "grad_norm": 0.7224128842353821, "learning_rate": 0.0002, "epoch": 2.38622585078005, "step": 14760}, {"loss": 0.7017, "grad_norm": 0.8040381669998169, "learning_rate": 0.0002, "epoch": 2.3878425349607952, "step": 14770}, {"loss": 0.6799, "grad_norm": 0.7705281376838684, "learning_rate": 0.0002, "epoch": 2.389459219141541, "step": 14780}, {"loss": 0.6326, "grad_norm": 0.667966902256012, "learning_rate": 0.0002, "epoch": 2.391075903322286, "step": 14790}, {"loss": 0.7061, "grad_norm": 0.6611011028289795, "learning_rate": 0.0002, "epoch": 2.3926925875030314, "step": 14800}, {"loss": 0.6527, "grad_norm": 0.6862651705741882, "learning_rate": 0.0002, "epoch": 2.3943092716837766, "step": 14810}, {"loss": 0.6537, "grad_norm": 0.8086010217666626, "learning_rate": 0.0002, "epoch": 2.395925955864522, "step": 14820}, {"loss": 0.7189, "grad_norm": 0.7189689874649048, "learning_rate": 0.0002, "epoch": 2.397542640045267, "step": 14830}, {"loss": 0.6709, "grad_norm": 0.6280009150505066, "learning_rate": 0.0002, "epoch": 2.3991593242260123, "step": 14840}, {"loss": 0.706, "grad_norm": 0.7826612591743469, "learning_rate": 0.0002, "epoch": 2.4007760084067575, "step": 14850}, {"loss": 0.6738, "grad_norm": 0.7681610584259033, "learning_rate": 0.0002, "epoch": 2.402392692587503, "step": 14860}, {"loss": 0.636, "grad_norm": 0.720966100692749, "learning_rate": 0.0002, "epoch": 2.4040093767682484, "step": 14870}, {"loss": 0.6667, "grad_norm": 0.8202250599861145, "learning_rate": 0.0002, "epoch": 2.4056260609489937, "step": 14880}, {"loss": 0.6935, "grad_norm": 0.786212682723999, "learning_rate": 0.0002, "epoch": 2.407242745129739, "step": 14890}, {"loss": 0.6628, "grad_norm": 0.6647164821624756, "learning_rate": 0.0002, "epoch": 2.408859429310484, "step": 14900}, {"loss": 0.6706, "grad_norm": 0.7566399574279785, "learning_rate": 0.0002, "epoch": 2.4104761134912294, "step": 14910}, {"loss": 0.7188, "grad_norm": 0.748814582824707, "learning_rate": 0.0002, "epoch": 2.4120927976719746, "step": 14920}, {"loss": 0.6684, "grad_norm": 0.7624038457870483, "learning_rate": 0.0002, "epoch": 2.4137094818527203, "step": 14930}, {"loss": 0.6483, "grad_norm": 0.8267335295677185, "learning_rate": 0.0002, "epoch": 2.4153261660334655, "step": 14940}, {"loss": 0.6612, "grad_norm": 0.8785360455513, "learning_rate": 0.0002, "epoch": 2.4169428502142107, "step": 14950}, {"loss": 0.6718, "grad_norm": 0.679887592792511, "learning_rate": 0.0002, "epoch": 2.418559534394956, "step": 14960}, {"loss": 0.6136, "grad_norm": 0.7218474745750427, "learning_rate": 0.0002, "epoch": 2.420176218575701, "step": 14970}, {"loss": 0.648, "grad_norm": 0.6342799663543701, "learning_rate": 0.0002, "epoch": 2.4217929027564464, "step": 14980}, {"loss": 0.6617, "grad_norm": 0.7098712921142578, "learning_rate": 0.0002, "epoch": 2.4234095869371917, "step": 14990}, {"loss": 0.6942, "grad_norm": 0.7497431635856628, "learning_rate": 0.0002, "epoch": 2.425026271117937, "step": 15000}, {"loss": 0.6772, "grad_norm": 0.934836208820343, "learning_rate": 0.0002, "epoch": 2.4266429552986826, "step": 15010}, {"loss": 0.7221, "grad_norm": 0.8430966734886169, "learning_rate": 0.0002, "epoch": 2.428259639479428, "step": 15020}, {"loss": 0.6985, "grad_norm": 0.7032104730606079, "learning_rate": 0.0002, "epoch": 2.429876323660173, "step": 15030}, {"loss": 0.6715, "grad_norm": 0.7746111750602722, "learning_rate": 0.0002, "epoch": 2.4314930078409183, "step": 15040}, {"loss": 0.7177, "grad_norm": 0.7661406397819519, "learning_rate": 0.0002, "epoch": 2.4331096920216635, "step": 15050}, {"loss": 0.6517, "grad_norm": 0.6941645741462708, "learning_rate": 0.0002, "epoch": 2.4347263762024087, "step": 15060}, {"loss": 0.6421, "grad_norm": 0.7487249374389648, "learning_rate": 0.0002, "epoch": 2.436343060383154, "step": 15070}, {"loss": 0.6796, "grad_norm": 0.7639912962913513, "learning_rate": 0.0002, "epoch": 2.4379597445638996, "step": 15080}, {"loss": 0.7087, "grad_norm": 0.7708953619003296, "learning_rate": 0.0002, "epoch": 2.439576428744645, "step": 15090}, {"loss": 0.7065, "grad_norm": 0.9135832190513611, "learning_rate": 0.0002, "epoch": 2.44119311292539, "step": 15100}, {"loss": 0.672, "grad_norm": 0.8283005356788635, "learning_rate": 0.0002, "epoch": 2.4428097971061353, "step": 15110}, {"loss": 0.6551, "grad_norm": 0.925299346446991, "learning_rate": 0.0002, "epoch": 2.4444264812868806, "step": 15120}, {"loss": 0.687, "grad_norm": 0.7013528943061829, "learning_rate": 0.0002, "epoch": 2.446043165467626, "step": 15130}, {"loss": 0.6842, "grad_norm": 0.622303307056427, "learning_rate": 0.0002, "epoch": 2.447659849648371, "step": 15140}, {"loss": 0.6676, "grad_norm": 0.876569390296936, "learning_rate": 0.0002, "epoch": 2.4492765338291163, "step": 15150}, {"loss": 0.6463, "grad_norm": 0.6836351752281189, "learning_rate": 0.0002, "epoch": 2.450893218009862, "step": 15160}, {"loss": 0.6781, "grad_norm": 0.7886684536933899, "learning_rate": 0.0002, "epoch": 2.452509902190607, "step": 15170}, {"loss": 0.6794, "grad_norm": 0.6647440791130066, "learning_rate": 0.0002, "epoch": 2.4541265863713524, "step": 15180}, {"loss": 0.6353, "grad_norm": 0.7477722764015198, "learning_rate": 0.0002, "epoch": 2.4557432705520976, "step": 15190}, {"loss": 0.698, "grad_norm": 0.8192033767700195, "learning_rate": 0.0002, "epoch": 2.457359954732843, "step": 15200}, {"loss": 0.6735, "grad_norm": 0.847537100315094, "learning_rate": 0.0002, "epoch": 2.458976638913588, "step": 15210}, {"loss": 0.6962, "grad_norm": 0.9027776122093201, "learning_rate": 0.0002, "epoch": 2.4605933230943338, "step": 15220}, {"loss": 0.7084, "grad_norm": 0.7217772006988525, "learning_rate": 0.0002, "epoch": 2.462210007275079, "step": 15230}, {"loss": 0.691, "grad_norm": 0.7994546294212341, "learning_rate": 0.0002, "epoch": 2.4638266914558242, "step": 15240}, {"loss": 0.6828, "grad_norm": 0.939916729927063, "learning_rate": 0.0002, "epoch": 2.4654433756365695, "step": 15250}, {"loss": 0.6893, "grad_norm": 1.0009053945541382, "learning_rate": 0.0002, "epoch": 2.4670600598173147, "step": 15260}, {"loss": 0.643, "grad_norm": 0.625555694103241, "learning_rate": 0.0002, "epoch": 2.46867674399806, "step": 15270}, {"loss": 0.688, "grad_norm": 0.7924878597259521, "learning_rate": 0.0002, "epoch": 2.470293428178805, "step": 15280}, {"loss": 0.6789, "grad_norm": 0.8536689877510071, "learning_rate": 0.0002, "epoch": 2.4719101123595504, "step": 15290}, {"loss": 0.6924, "grad_norm": 0.8572589755058289, "learning_rate": 0.0002, "epoch": 2.4735267965402956, "step": 15300}, {"loss": 0.604, "grad_norm": 0.773279070854187, "learning_rate": 0.0002, "epoch": 2.4751434807210413, "step": 15310}, {"loss": 0.6573, "grad_norm": 0.7708749771118164, "learning_rate": 0.0002, "epoch": 2.4767601649017865, "step": 15320}, {"loss": 0.7065, "grad_norm": 0.770905077457428, "learning_rate": 0.0002, "epoch": 2.4783768490825318, "step": 15330}, {"loss": 0.6878, "grad_norm": 0.8238571882247925, "learning_rate": 0.0002, "epoch": 2.479993533263277, "step": 15340}, {"loss": 0.6772, "grad_norm": 0.7670477032661438, "learning_rate": 0.0002, "epoch": 2.481610217444022, "step": 15350}, {"loss": 0.7759, "grad_norm": 0.905036985874176, "learning_rate": 0.0002, "epoch": 2.4832269016247674, "step": 15360}, {"loss": 0.706, "grad_norm": 0.6672089695930481, "learning_rate": 0.0002, "epoch": 2.484843585805513, "step": 15370}, {"loss": 0.6722, "grad_norm": 0.625095784664154, "learning_rate": 0.0002, "epoch": 2.4864602699862584, "step": 15380}, {"loss": 0.6396, "grad_norm": 0.679772675037384, "learning_rate": 0.0002, "epoch": 2.4880769541670036, "step": 15390}, {"loss": 0.6778, "grad_norm": 0.711492121219635, "learning_rate": 0.0002, "epoch": 2.489693638347749, "step": 15400}, {"loss": 0.6966, "grad_norm": 0.876189112663269, "learning_rate": 0.0002, "epoch": 2.491310322528494, "step": 15410}, {"loss": 0.7307, "grad_norm": 0.7236915230751038, "learning_rate": 0.0002, "epoch": 2.4929270067092393, "step": 15420}, {"loss": 0.647, "grad_norm": 0.6629832983016968, "learning_rate": 0.0002, "epoch": 2.4945436908899845, "step": 15430}, {"loss": 0.6669, "grad_norm": 0.9756859540939331, "learning_rate": 0.0002, "epoch": 2.4961603750707297, "step": 15440}, {"loss": 0.7559, "grad_norm": 0.6896940469741821, "learning_rate": 0.0002, "epoch": 2.4977770592514754, "step": 15450}, {"loss": 0.6818, "grad_norm": 0.7105149626731873, "learning_rate": 0.0002, "epoch": 2.4993937434322206, "step": 15460}, {"loss": 0.6859, "grad_norm": 0.8374546766281128, "learning_rate": 0.0002, "epoch": 2.501010427612966, "step": 15470}, {"loss": 0.6512, "grad_norm": 0.7320070266723633, "learning_rate": 0.0002, "epoch": 2.502627111793711, "step": 15480}, {"loss": 0.685, "grad_norm": 0.8306367993354797, "learning_rate": 0.0002, "epoch": 2.5042437959744563, "step": 15490}, {"loss": 0.7253, "grad_norm": 0.7472721338272095, "learning_rate": 0.0002, "epoch": 2.5058604801552016, "step": 15500}, {"loss": 0.6699, "grad_norm": 0.6147692203521729, "learning_rate": 0.0002, "epoch": 2.507477164335947, "step": 15510}, {"loss": 0.7158, "grad_norm": 0.7788505554199219, "learning_rate": 0.0002, "epoch": 2.5090938485166925, "step": 15520}, {"loss": 0.6521, "grad_norm": 0.8807527422904968, "learning_rate": 0.0002, "epoch": 2.5107105326974377, "step": 15530}, {"loss": 0.6792, "grad_norm": 0.7521643042564392, "learning_rate": 0.0002, "epoch": 2.512327216878183, "step": 15540}, {"loss": 0.6772, "grad_norm": 0.6900225281715393, "learning_rate": 0.0002, "epoch": 2.513943901058928, "step": 15550}, {"loss": 0.6769, "grad_norm": 0.6601938605308533, "learning_rate": 0.0002, "epoch": 2.5155605852396734, "step": 15560}, {"loss": 0.6648, "grad_norm": 0.8179984092712402, "learning_rate": 0.0002, "epoch": 2.5171772694204186, "step": 15570}, {"loss": 0.7028, "grad_norm": 0.792556881904602, "learning_rate": 0.0002, "epoch": 2.518793953601164, "step": 15580}, {"loss": 0.6464, "grad_norm": 0.7081938982009888, "learning_rate": 0.0002, "epoch": 2.520410637781909, "step": 15590}, {"loss": 0.6691, "grad_norm": 0.8733121156692505, "learning_rate": 0.0002, "epoch": 2.5220273219626543, "step": 15600}, {"loss": 0.6969, "grad_norm": 0.7980992794036865, "learning_rate": 0.0002, "epoch": 2.5236440061434, "step": 15610}, {"loss": 0.7124, "grad_norm": 0.883664071559906, "learning_rate": 0.0002, "epoch": 2.5252606903241452, "step": 15620}, {"loss": 0.7022, "grad_norm": 0.6963341236114502, "learning_rate": 0.0002, "epoch": 2.5268773745048905, "step": 15630}, {"loss": 0.7334, "grad_norm": 0.6433573365211487, "learning_rate": 0.0002, "epoch": 2.5284940586856357, "step": 15640}, {"loss": 0.6889, "grad_norm": 0.8538183569908142, "learning_rate": 0.0002, "epoch": 2.530110742866381, "step": 15650}, {"loss": 0.6841, "grad_norm": 0.9748201370239258, "learning_rate": 0.0002, "epoch": 2.5317274270471266, "step": 15660}, {"loss": 0.6765, "grad_norm": 0.7670575380325317, "learning_rate": 0.0002, "epoch": 2.533344111227872, "step": 15670}, {"loss": 0.6435, "grad_norm": 0.8738890290260315, "learning_rate": 0.0002, "epoch": 2.534960795408617, "step": 15680}, {"loss": 0.6802, "grad_norm": 0.8391636610031128, "learning_rate": 0.0002, "epoch": 2.5365774795893623, "step": 15690}, {"loss": 0.6901, "grad_norm": 0.7239366769790649, "learning_rate": 0.0002, "epoch": 2.5381941637701075, "step": 15700}, {"loss": 0.7011, "grad_norm": 0.8498379588127136, "learning_rate": 0.0002, "epoch": 2.5398108479508528, "step": 15710}, {"loss": 0.6998, "grad_norm": 0.8029484152793884, "learning_rate": 0.0002, "epoch": 2.541427532131598, "step": 15720}, {"loss": 0.6678, "grad_norm": 1.0639333724975586, "learning_rate": 0.0002, "epoch": 2.5430442163123432, "step": 15730}, {"loss": 0.6341, "grad_norm": 0.6401297450065613, "learning_rate": 0.0002, "epoch": 2.5446609004930885, "step": 15740}, {"loss": 0.7196, "grad_norm": 0.7123814821243286, "learning_rate": 0.0002, "epoch": 2.5462775846738337, "step": 15750}, {"loss": 0.654, "grad_norm": 0.7874974608421326, "learning_rate": 0.0002, "epoch": 2.5478942688545794, "step": 15760}, {"loss": 0.6721, "grad_norm": 0.8046808838844299, "learning_rate": 0.0002, "epoch": 2.5495109530353246, "step": 15770}, {"loss": 0.6665, "grad_norm": 0.7888661623001099, "learning_rate": 0.0002, "epoch": 2.55112763721607, "step": 15780}, {"loss": 0.6893, "grad_norm": 0.8445866107940674, "learning_rate": 0.0002, "epoch": 2.552744321396815, "step": 15790}, {"loss": 0.6815, "grad_norm": 0.7475846409797668, "learning_rate": 0.0002, "epoch": 2.5543610055775603, "step": 15800}, {"loss": 0.6711, "grad_norm": 0.7455102801322937, "learning_rate": 0.0002, "epoch": 2.555977689758306, "step": 15810}, {"loss": 0.6932, "grad_norm": 0.8226983547210693, "learning_rate": 0.0002, "epoch": 2.557594373939051, "step": 15820}, {"loss": 0.651, "grad_norm": 0.8920368552207947, "learning_rate": 0.0002, "epoch": 2.5592110581197964, "step": 15830}, {"loss": 0.6297, "grad_norm": 0.8413904905319214, "learning_rate": 0.0002, "epoch": 2.5608277423005417, "step": 15840}, {"loss": 0.7106, "grad_norm": 0.8483649492263794, "learning_rate": 0.0002, "epoch": 2.562444426481287, "step": 15850}, {"loss": 0.6957, "grad_norm": 0.5923284292221069, "learning_rate": 0.0002, "epoch": 2.564061110662032, "step": 15860}, {"loss": 0.6847, "grad_norm": 0.8518726229667664, "learning_rate": 0.0002, "epoch": 2.5656777948427774, "step": 15870}, {"loss": 0.6362, "grad_norm": 0.731235146522522, "learning_rate": 0.0002, "epoch": 2.5672944790235226, "step": 15880}, {"loss": 0.7611, "grad_norm": 0.7517194151878357, "learning_rate": 0.0002, "epoch": 2.568911163204268, "step": 15890}, {"loss": 0.6907, "grad_norm": 0.8378692269325256, "learning_rate": 0.0002, "epoch": 2.5705278473850135, "step": 15900}, {"loss": 0.7055, "grad_norm": 0.843701958656311, "learning_rate": 0.0002, "epoch": 2.5721445315657587, "step": 15910}, {"loss": 0.6882, "grad_norm": 0.7254629731178284, "learning_rate": 0.0002, "epoch": 2.573761215746504, "step": 15920}, {"loss": 0.6872, "grad_norm": 0.8863335847854614, "learning_rate": 0.0002, "epoch": 2.575377899927249, "step": 15930}, {"loss": 0.6813, "grad_norm": 0.7675097584724426, "learning_rate": 0.0002, "epoch": 2.5769945841079944, "step": 15940}, {"loss": 0.7357, "grad_norm": 0.82063889503479, "learning_rate": 0.0002, "epoch": 2.5786112682887397, "step": 15950}, {"loss": 0.662, "grad_norm": 0.7729717493057251, "learning_rate": 0.0002, "epoch": 2.5802279524694853, "step": 15960}, {"loss": 0.633, "grad_norm": 0.8301846981048584, "learning_rate": 0.0002, "epoch": 2.5818446366502306, "step": 15970}, {"loss": 0.6897, "grad_norm": 0.7906861305236816, "learning_rate": 0.0002, "epoch": 2.583461320830976, "step": 15980}, {"loss": 0.7175, "grad_norm": 0.6749057173728943, "learning_rate": 0.0002, "epoch": 2.585078005011721, "step": 15990}, {"loss": 0.7212, "grad_norm": 0.9386842846870422, "learning_rate": 0.0002, "epoch": 2.5866946891924663, "step": 16000}, {"loss": 0.6934, "grad_norm": 0.7868891358375549, "learning_rate": 0.0002, "epoch": 2.5883113733732115, "step": 16010}, {"loss": 0.7036, "grad_norm": 0.8674671053886414, "learning_rate": 0.0002, "epoch": 2.5899280575539567, "step": 16020}, {"loss": 0.7217, "grad_norm": 0.7043559551239014, "learning_rate": 0.0002, "epoch": 2.591544741734702, "step": 16030}, {"loss": 0.6967, "grad_norm": 0.5846083760261536, "learning_rate": 0.0002, "epoch": 2.593161425915447, "step": 16040}, {"loss": 0.7322, "grad_norm": 0.7323982119560242, "learning_rate": 0.0002, "epoch": 2.594778110096193, "step": 16050}, {"loss": 0.6794, "grad_norm": 0.9069556593894958, "learning_rate": 0.0002, "epoch": 2.596394794276938, "step": 16060}, {"loss": 0.7076, "grad_norm": 0.7522736191749573, "learning_rate": 0.0002, "epoch": 2.5980114784576833, "step": 16070}, {"loss": 0.6477, "grad_norm": 0.8149648308753967, "learning_rate": 0.0002, "epoch": 2.5996281626384286, "step": 16080}, {"loss": 0.6664, "grad_norm": 0.6214233040809631, "learning_rate": 0.0002, "epoch": 2.601244846819174, "step": 16090}, {"loss": 0.7307, "grad_norm": 0.6803743839263916, "learning_rate": 0.0002, "epoch": 2.602861530999919, "step": 16100}, {"loss": 0.7244, "grad_norm": 0.7223997116088867, "learning_rate": 0.0002, "epoch": 2.6044782151806647, "step": 16110}, {"loss": 0.6867, "grad_norm": 0.7324174642562866, "learning_rate": 0.0002, "epoch": 2.60609489936141, "step": 16120}, {"loss": 0.7159, "grad_norm": 0.9594739675521851, "learning_rate": 0.0002, "epoch": 2.607711583542155, "step": 16130}, {"loss": 0.6451, "grad_norm": 0.9485327005386353, "learning_rate": 0.0002, "epoch": 2.6093282677229004, "step": 16140}, {"loss": 0.6815, "grad_norm": 0.8449000120162964, "learning_rate": 0.0002, "epoch": 2.6109449519036456, "step": 16150}, {"loss": 0.7152, "grad_norm": 0.8520140051841736, "learning_rate": 0.0002, "epoch": 2.612561636084391, "step": 16160}, {"loss": 0.6759, "grad_norm": 0.7456524968147278, "learning_rate": 0.0002, "epoch": 2.614178320265136, "step": 16170}, {"loss": 0.6893, "grad_norm": 0.9912857413291931, "learning_rate": 0.0002, "epoch": 2.6157950044458813, "step": 16180}, {"loss": 0.7243, "grad_norm": 0.9001946449279785, "learning_rate": 0.0002, "epoch": 2.6174116886266265, "step": 16190}, {"loss": 0.6825, "grad_norm": 0.6568667888641357, "learning_rate": 0.0002, "epoch": 2.619028372807372, "step": 16200}, {"loss": 0.7013, "grad_norm": 1.0248128175735474, "learning_rate": 0.0002, "epoch": 2.6206450569881174, "step": 16210}, {"loss": 0.7045, "grad_norm": 0.6509039998054504, "learning_rate": 0.0002, "epoch": 2.6222617411688627, "step": 16220}, {"loss": 0.72, "grad_norm": 0.7626351118087769, "learning_rate": 0.0002, "epoch": 2.623878425349608, "step": 16230}, {"loss": 0.6556, "grad_norm": 0.6938552260398865, "learning_rate": 0.0002, "epoch": 2.625495109530353, "step": 16240}, {"loss": 0.65, "grad_norm": 0.6434680819511414, "learning_rate": 0.0002, "epoch": 2.6271117937110984, "step": 16250}, {"loss": 0.6943, "grad_norm": 0.7111515998840332, "learning_rate": 0.0002, "epoch": 2.628728477891844, "step": 16260}, {"loss": 0.679, "grad_norm": 0.7712395787239075, "learning_rate": 0.0002, "epoch": 2.6303451620725893, "step": 16270}, {"loss": 0.6886, "grad_norm": 0.792209267616272, "learning_rate": 0.0002, "epoch": 2.6319618462533345, "step": 16280}, {"loss": 0.6554, "grad_norm": 0.6801066398620605, "learning_rate": 0.0002, "epoch": 2.6335785304340797, "step": 16290}, {"loss": 0.73, "grad_norm": 0.7802573442459106, "learning_rate": 0.0002, "epoch": 2.635195214614825, "step": 16300}, {"loss": 0.7484, "grad_norm": 0.7742244601249695, "learning_rate": 0.0002, "epoch": 2.63681189879557, "step": 16310}, {"loss": 0.6524, "grad_norm": 0.664184033870697, "learning_rate": 0.0002, "epoch": 2.6384285829763154, "step": 16320}, {"loss": 0.6442, "grad_norm": 0.9242228865623474, "learning_rate": 0.0002, "epoch": 2.6400452671570607, "step": 16330}, {"loss": 0.6792, "grad_norm": 0.9661325216293335, "learning_rate": 0.0002, "epoch": 2.641661951337806, "step": 16340}, {"loss": 0.6847, "grad_norm": 0.837526798248291, "learning_rate": 0.0002, "epoch": 2.6432786355185516, "step": 16350}, {"loss": 0.7686, "grad_norm": 1.1834373474121094, "learning_rate": 0.0002, "epoch": 2.644895319699297, "step": 16360}, {"loss": 0.6746, "grad_norm": 0.7467831373214722, "learning_rate": 0.0002, "epoch": 2.646512003880042, "step": 16370}, {"loss": 0.6935, "grad_norm": 0.8627146482467651, "learning_rate": 0.0002, "epoch": 2.6481286880607873, "step": 16380}, {"loss": 0.715, "grad_norm": 0.790447473526001, "learning_rate": 0.0002, "epoch": 2.6497453722415325, "step": 16390}, {"loss": 0.723, "grad_norm": 0.8447365164756775, "learning_rate": 0.0002, "epoch": 2.651362056422278, "step": 16400}, {"loss": 0.6628, "grad_norm": 0.7831417918205261, "learning_rate": 0.0002, "epoch": 2.6529787406030234, "step": 16410}, {"loss": 0.6691, "grad_norm": 0.6837952136993408, "learning_rate": 0.0002, "epoch": 2.6545954247837686, "step": 16420}, {"loss": 0.6139, "grad_norm": 0.7031801342964172, "learning_rate": 0.0002, "epoch": 2.656212108964514, "step": 16430}, {"loss": 0.7382, "grad_norm": 0.8963770866394043, "learning_rate": 0.0002, "epoch": 2.657828793145259, "step": 16440}, {"loss": 0.6439, "grad_norm": 0.6852328181266785, "learning_rate": 0.0002, "epoch": 2.6594454773260043, "step": 16450}, {"loss": 0.6278, "grad_norm": 0.8069294095039368, "learning_rate": 0.0002, "epoch": 2.6610621615067496, "step": 16460}, {"loss": 0.6939, "grad_norm": 0.7503686547279358, "learning_rate": 0.0002, "epoch": 2.662678845687495, "step": 16470}, {"loss": 0.6777, "grad_norm": 0.6430956125259399, "learning_rate": 0.0002, "epoch": 2.66429552986824, "step": 16480}, {"loss": 0.6863, "grad_norm": 0.7894312739372253, "learning_rate": 0.0002, "epoch": 2.6659122140489853, "step": 16490}, {"loss": 0.7165, "grad_norm": 0.7277431488037109, "learning_rate": 0.0002, "epoch": 2.667528898229731, "step": 16500}, {"loss": 0.6772, "grad_norm": 0.6816153526306152, "learning_rate": 0.0002, "epoch": 2.669145582410476, "step": 16510}, {"loss": 0.691, "grad_norm": 0.8145235776901245, "learning_rate": 0.0002, "epoch": 2.6707622665912214, "step": 16520}, {"loss": 0.709, "grad_norm": 0.8645890355110168, "learning_rate": 0.0002, "epoch": 2.6723789507719666, "step": 16530}, {"loss": 0.6946, "grad_norm": 0.704393208026886, "learning_rate": 0.0002, "epoch": 2.673995634952712, "step": 16540}, {"loss": 0.6378, "grad_norm": 1.0120846033096313, "learning_rate": 0.0002, "epoch": 2.6756123191334575, "step": 16550}, {"loss": 0.7241, "grad_norm": 0.6919328570365906, "learning_rate": 0.0002, "epoch": 2.6772290033142028, "step": 16560}, {"loss": 0.7098, "grad_norm": 0.6924574971199036, "learning_rate": 0.0002, "epoch": 2.678845687494948, "step": 16570}, {"loss": 0.731, "grad_norm": 0.9679301381111145, "learning_rate": 0.0002, "epoch": 2.6804623716756932, "step": 16580}, {"loss": 0.7124, "grad_norm": 0.6810211539268494, "learning_rate": 0.0002, "epoch": 2.6820790558564385, "step": 16590}, {"loss": 0.6688, "grad_norm": 0.9730555415153503, "learning_rate": 0.0002, "epoch": 2.6836957400371837, "step": 16600}, {"loss": 0.7344, "grad_norm": 0.7852821350097656, "learning_rate": 0.0002, "epoch": 2.685312424217929, "step": 16610}, {"loss": 0.6401, "grad_norm": 0.6059057116508484, "learning_rate": 0.0002, "epoch": 2.686929108398674, "step": 16620}, {"loss": 0.6796, "grad_norm": 0.9395958781242371, "learning_rate": 0.0002, "epoch": 2.6885457925794194, "step": 16630}, {"loss": 0.7174, "grad_norm": 0.7473729848861694, "learning_rate": 0.0002, "epoch": 2.690162476760165, "step": 16640}, {"loss": 0.7087, "grad_norm": 0.765934407711029, "learning_rate": 0.0002, "epoch": 2.6917791609409103, "step": 16650}, {"loss": 0.707, "grad_norm": 0.8496677279472351, "learning_rate": 0.0002, "epoch": 2.6933958451216555, "step": 16660}, {"loss": 0.7084, "grad_norm": 0.7641879916191101, "learning_rate": 0.0002, "epoch": 2.6950125293024008, "step": 16670}, {"loss": 0.6566, "grad_norm": 0.8471952676773071, "learning_rate": 0.0002, "epoch": 2.696629213483146, "step": 16680}, {"loss": 0.6635, "grad_norm": 0.6946060657501221, "learning_rate": 0.0002, "epoch": 2.6982458976638912, "step": 16690}, {"loss": 0.7027, "grad_norm": 0.7361312508583069, "learning_rate": 0.0002, "epoch": 2.699862581844637, "step": 16700}, {"loss": 0.6767, "grad_norm": 0.6605038046836853, "learning_rate": 0.0002, "epoch": 2.701479266025382, "step": 16710}, {"loss": 0.6885, "grad_norm": 0.7164411544799805, "learning_rate": 0.0002, "epoch": 2.7030959502061274, "step": 16720}, {"loss": 0.6736, "grad_norm": 0.6496201157569885, "learning_rate": 0.0002, "epoch": 2.7047126343868726, "step": 16730}, {"loss": 0.6942, "grad_norm": 0.7826663851737976, "learning_rate": 0.0002, "epoch": 2.706329318567618, "step": 16740}, {"loss": 0.6773, "grad_norm": 0.7639131546020508, "learning_rate": 0.0002, "epoch": 2.707946002748363, "step": 16750}, {"loss": 0.69, "grad_norm": 0.7976210713386536, "learning_rate": 0.0002, "epoch": 2.7095626869291083, "step": 16760}, {"loss": 0.6735, "grad_norm": 0.6836577653884888, "learning_rate": 0.0002, "epoch": 2.7111793711098535, "step": 16770}, {"loss": 0.6596, "grad_norm": 0.8025202751159668, "learning_rate": 0.0002, "epoch": 2.7127960552905988, "step": 16780}, {"loss": 0.6324, "grad_norm": 0.7636463642120361, "learning_rate": 0.0002, "epoch": 2.7144127394713444, "step": 16790}, {"loss": 0.6227, "grad_norm": 0.7481677532196045, "learning_rate": 0.0002, "epoch": 2.7160294236520897, "step": 16800}, {"loss": 0.6925, "grad_norm": 0.7566834688186646, "learning_rate": 0.0002, "epoch": 2.717646107832835, "step": 16810}, {"loss": 0.6531, "grad_norm": 0.7931267619132996, "learning_rate": 0.0002, "epoch": 2.71926279201358, "step": 16820}, {"loss": 0.6672, "grad_norm": 0.8811662197113037, "learning_rate": 0.0002, "epoch": 2.7208794761943254, "step": 16830}, {"loss": 0.6675, "grad_norm": 0.8561240434646606, "learning_rate": 0.0002, "epoch": 2.7224961603750706, "step": 16840}, {"loss": 0.7135, "grad_norm": 0.7121599316596985, "learning_rate": 0.0002, "epoch": 2.7241128445558163, "step": 16850}, {"loss": 0.6825, "grad_norm": 0.8066257238388062, "learning_rate": 0.0002, "epoch": 2.7257295287365615, "step": 16860}, {"loss": 0.6839, "grad_norm": 0.7699271440505981, "learning_rate": 0.0002, "epoch": 2.7273462129173067, "step": 16870}, {"loss": 0.699, "grad_norm": 1.1828432083129883, "learning_rate": 0.0002, "epoch": 2.728962897098052, "step": 16880}, {"loss": 0.6518, "grad_norm": 0.9989302754402161, "learning_rate": 0.0002, "epoch": 2.730579581278797, "step": 16890}, {"loss": 0.7015, "grad_norm": 0.8100560307502747, "learning_rate": 0.0002, "epoch": 2.7321962654595424, "step": 16900}, {"loss": 0.6851, "grad_norm": 0.8615233898162842, "learning_rate": 0.0002, "epoch": 2.7338129496402876, "step": 16910}, {"loss": 0.6322, "grad_norm": 0.8633756041526794, "learning_rate": 0.0002, "epoch": 2.735429633821033, "step": 16920}, {"loss": 0.6488, "grad_norm": 0.7769348621368408, "learning_rate": 0.0002, "epoch": 2.737046318001778, "step": 16930}, {"loss": 0.6582, "grad_norm": 0.6943058371543884, "learning_rate": 0.0002, "epoch": 2.738663002182524, "step": 16940}, {"loss": 0.6516, "grad_norm": 0.8510736227035522, "learning_rate": 0.0002, "epoch": 2.740279686363269, "step": 16950}, {"loss": 0.7275, "grad_norm": 0.7732602953910828, "learning_rate": 0.0002, "epoch": 2.7418963705440142, "step": 16960}, {"loss": 0.6553, "grad_norm": 0.5981788635253906, "learning_rate": 0.0002, "epoch": 2.7435130547247595, "step": 16970}, {"loss": 0.6777, "grad_norm": 0.7604416012763977, "learning_rate": 0.0002, "epoch": 2.7451297389055047, "step": 16980}, {"loss": 0.6981, "grad_norm": 0.7377738356590271, "learning_rate": 0.0002, "epoch": 2.74674642308625, "step": 16990}, {"loss": 0.6294, "grad_norm": 0.9400289058685303, "learning_rate": 0.0002, "epoch": 2.7483631072669956, "step": 17000}, {"loss": 0.6952, "grad_norm": 0.6340599656105042, "learning_rate": 0.0002, "epoch": 2.749979791447741, "step": 17010}, {"loss": 0.7222, "grad_norm": 0.7297601103782654, "learning_rate": 0.0002, "epoch": 2.751596475628486, "step": 17020}, {"loss": 0.6659, "grad_norm": 0.9479979872703552, "learning_rate": 0.0002, "epoch": 2.7532131598092313, "step": 17030}, {"loss": 0.691, "grad_norm": 0.8461511135101318, "learning_rate": 0.0002, "epoch": 2.7548298439899765, "step": 17040}, {"loss": 0.6764, "grad_norm": 0.7477551698684692, "learning_rate": 0.0002, "epoch": 2.7564465281707218, "step": 17050}, {"loss": 0.684, "grad_norm": 1.019270420074463, "learning_rate": 0.0002, "epoch": 2.758063212351467, "step": 17060}, {"loss": 0.7119, "grad_norm": 0.7730235457420349, "learning_rate": 0.0002, "epoch": 2.7596798965322122, "step": 17070}, {"loss": 0.6886, "grad_norm": 0.8216866254806519, "learning_rate": 0.0002, "epoch": 2.7612965807129575, "step": 17080}, {"loss": 0.6811, "grad_norm": 0.7235931754112244, "learning_rate": 0.0002, "epoch": 2.762913264893703, "step": 17090}, {"loss": 0.7031, "grad_norm": 0.7352296710014343, "learning_rate": 0.0002, "epoch": 2.7645299490744484, "step": 17100}, {"loss": 0.6951, "grad_norm": 0.8129373788833618, "learning_rate": 0.0002, "epoch": 2.7661466332551936, "step": 17110}, {"loss": 0.6703, "grad_norm": 0.7387019991874695, "learning_rate": 0.0002, "epoch": 2.767763317435939, "step": 17120}, {"loss": 0.6789, "grad_norm": 0.9149190187454224, "learning_rate": 0.0002, "epoch": 2.769380001616684, "step": 17130}, {"loss": 0.6038, "grad_norm": 0.7352971434593201, "learning_rate": 0.0002, "epoch": 2.7709966857974297, "step": 17140}, {"loss": 0.6728, "grad_norm": 0.7903780341148376, "learning_rate": 0.0002, "epoch": 2.772613369978175, "step": 17150}, {"loss": 0.6988, "grad_norm": 0.8255927562713623, "learning_rate": 0.0002, "epoch": 2.77423005415892, "step": 17160}, {"loss": 0.6694, "grad_norm": 0.7235927581787109, "learning_rate": 0.0002, "epoch": 2.7758467383396654, "step": 17170}, {"loss": 0.7161, "grad_norm": 0.8281434774398804, "learning_rate": 0.0002, "epoch": 2.7774634225204107, "step": 17180}, {"loss": 0.682, "grad_norm": 0.7586921453475952, "learning_rate": 0.0002, "epoch": 2.779080106701156, "step": 17190}, {"loss": 0.6427, "grad_norm": 0.7161715030670166, "learning_rate": 0.0002, "epoch": 2.780696790881901, "step": 17200}, {"loss": 0.6426, "grad_norm": 0.762868344783783, "learning_rate": 0.0002, "epoch": 2.7823134750626464, "step": 17210}, {"loss": 0.705, "grad_norm": 0.9285483360290527, "learning_rate": 0.0002, "epoch": 2.7839301592433916, "step": 17220}, {"loss": 0.7084, "grad_norm": 0.6900462508201599, "learning_rate": 0.0002, "epoch": 2.785546843424137, "step": 17230}, {"loss": 0.6988, "grad_norm": 0.780384361743927, "learning_rate": 0.0002, "epoch": 2.7871635276048825, "step": 17240}, {"loss": 0.7073, "grad_norm": 0.7580406665802002, "learning_rate": 0.0002, "epoch": 2.7887802117856277, "step": 17250}, {"loss": 0.6833, "grad_norm": 0.8145199418067932, "learning_rate": 0.0002, "epoch": 2.790396895966373, "step": 17260}, {"loss": 0.6909, "grad_norm": 0.9159596562385559, "learning_rate": 0.0002, "epoch": 2.792013580147118, "step": 17270}, {"loss": 0.6008, "grad_norm": 0.9590014219284058, "learning_rate": 0.0002, "epoch": 2.7936302643278634, "step": 17280}, {"loss": 0.6704, "grad_norm": 0.7603529691696167, "learning_rate": 0.0002, "epoch": 2.795246948508609, "step": 17290}, {"loss": 0.7165, "grad_norm": 0.8039976358413696, "learning_rate": 0.0002, "epoch": 2.7968636326893543, "step": 17300}, {"loss": 0.7037, "grad_norm": 0.8364847302436829, "learning_rate": 0.0002, "epoch": 2.7984803168700996, "step": 17310}, {"loss": 0.6749, "grad_norm": 0.8763046860694885, "learning_rate": 0.0002, "epoch": 2.800097001050845, "step": 17320}, {"loss": 0.6844, "grad_norm": 0.8409647941589355, "learning_rate": 0.0002, "epoch": 2.80171368523159, "step": 17330}, {"loss": 0.6936, "grad_norm": 0.7649006247520447, "learning_rate": 0.0002, "epoch": 2.8033303694123353, "step": 17340}, {"loss": 0.7051, "grad_norm": 0.7970262169837952, "learning_rate": 0.0002, "epoch": 2.8049470535930805, "step": 17350}, {"loss": 0.6533, "grad_norm": 0.9088607430458069, "learning_rate": 0.0002, "epoch": 2.8065637377738257, "step": 17360}, {"loss": 0.675, "grad_norm": 0.6454846858978271, "learning_rate": 0.0002, "epoch": 2.808180421954571, "step": 17370}, {"loss": 0.7069, "grad_norm": 0.7744787931442261, "learning_rate": 0.0002, "epoch": 2.809797106135316, "step": 17380}, {"loss": 0.6772, "grad_norm": 0.6678640842437744, "learning_rate": 0.0002, "epoch": 2.811413790316062, "step": 17390}, {"loss": 0.6784, "grad_norm": 0.772676944732666, "learning_rate": 0.0002, "epoch": 2.813030474496807, "step": 17400}, {"loss": 0.7252, "grad_norm": 0.7088175415992737, "learning_rate": 0.0002, "epoch": 2.8146471586775523, "step": 17410}, {"loss": 0.7086, "grad_norm": 0.8280573487281799, "learning_rate": 0.0002, "epoch": 2.8162638428582976, "step": 17420}, {"loss": 0.6732, "grad_norm": 0.6665388345718384, "learning_rate": 0.0002, "epoch": 2.817880527039043, "step": 17430}, {"loss": 0.6675, "grad_norm": 0.6427883505821228, "learning_rate": 0.0002, "epoch": 2.8194972112197885, "step": 17440}, {"loss": 0.6972, "grad_norm": 0.9697760343551636, "learning_rate": 0.0002, "epoch": 2.8211138954005337, "step": 17450}, {"loss": 0.6838, "grad_norm": 0.7573966383934021, "learning_rate": 0.0002, "epoch": 2.822730579581279, "step": 17460}, {"loss": 0.7243, "grad_norm": 0.878688633441925, "learning_rate": 0.0002, "epoch": 2.824347263762024, "step": 17470}, {"loss": 0.6666, "grad_norm": 0.7752242684364319, "learning_rate": 0.0002, "epoch": 2.8259639479427694, "step": 17480}, {"loss": 0.6638, "grad_norm": 0.6135398745536804, "learning_rate": 0.0002, "epoch": 2.8275806321235146, "step": 17490}, {"loss": 0.6829, "grad_norm": 0.6924924850463867, "learning_rate": 0.0002, "epoch": 2.82919731630426, "step": 17500}, {"loss": 0.6731, "grad_norm": 0.7471627593040466, "learning_rate": 0.0002, "epoch": 2.830814000485005, "step": 17510}, {"loss": 0.7016, "grad_norm": 0.7145499587059021, "learning_rate": 0.0002, "epoch": 2.8324306846657503, "step": 17520}, {"loss": 0.6787, "grad_norm": 0.7415414452552795, "learning_rate": 0.0002, "epoch": 2.834047368846496, "step": 17530}, {"loss": 0.6811, "grad_norm": 0.7328441739082336, "learning_rate": 0.0002, "epoch": 2.8356640530272412, "step": 17540}, {"loss": 0.6866, "grad_norm": 0.8267839550971985, "learning_rate": 0.0002, "epoch": 2.8372807372079865, "step": 17550}, {"loss": 0.6787, "grad_norm": 0.8877885341644287, "learning_rate": 0.0002, "epoch": 2.8388974213887317, "step": 17560}, {"loss": 0.7136, "grad_norm": 0.857138454914093, "learning_rate": 0.0002, "epoch": 2.840514105569477, "step": 17570}, {"loss": 0.6454, "grad_norm": 0.8470779657363892, "learning_rate": 0.0002, "epoch": 2.842130789750222, "step": 17580}, {"loss": 0.6976, "grad_norm": 0.8553254008293152, "learning_rate": 0.0002, "epoch": 2.843747473930968, "step": 17590}, {"loss": 0.7297, "grad_norm": 0.8033196926116943, "learning_rate": 0.0002, "epoch": 2.845364158111713, "step": 17600}, {"loss": 0.7062, "grad_norm": 0.7949087023735046, "learning_rate": 0.0002, "epoch": 2.8469808422924583, "step": 17610}, {"loss": 0.651, "grad_norm": 0.9241406321525574, "learning_rate": 0.0002, "epoch": 2.8485975264732035, "step": 17620}, {"loss": 0.6601, "grad_norm": 0.7721285223960876, "learning_rate": 0.0002, "epoch": 2.8502142106539488, "step": 17630}, {"loss": 0.6183, "grad_norm": 1.0246692895889282, "learning_rate": 0.0002, "epoch": 2.851830894834694, "step": 17640}, {"loss": 0.7007, "grad_norm": 0.9244589805603027, "learning_rate": 0.0002, "epoch": 2.853447579015439, "step": 17650}, {"loss": 0.7274, "grad_norm": 0.7243508696556091, "learning_rate": 0.0002, "epoch": 2.8550642631961844, "step": 17660}, {"loss": 0.6471, "grad_norm": 0.8943371176719666, "learning_rate": 0.0002, "epoch": 2.8566809473769297, "step": 17670}, {"loss": 0.686, "grad_norm": 0.6531758904457092, "learning_rate": 0.0002, "epoch": 2.8582976315576754, "step": 17680}, {"loss": 0.6253, "grad_norm": 0.8367000818252563, "learning_rate": 0.0002, "epoch": 2.8599143157384206, "step": 17690}, {"loss": 0.6943, "grad_norm": 0.7868556380271912, "learning_rate": 0.0002, "epoch": 2.861530999919166, "step": 17700}, {"loss": 0.6919, "grad_norm": 0.7213859558105469, "learning_rate": 0.0002, "epoch": 2.863147684099911, "step": 17710}, {"loss": 0.6657, "grad_norm": 0.7383931279182434, "learning_rate": 0.0002, "epoch": 2.8647643682806563, "step": 17720}, {"loss": 0.6841, "grad_norm": 0.7566812634468079, "learning_rate": 0.0002, "epoch": 2.8663810524614015, "step": 17730}, {"loss": 0.6449, "grad_norm": 0.6930373311042786, "learning_rate": 0.0002, "epoch": 2.867997736642147, "step": 17740}, {"loss": 0.6764, "grad_norm": 0.7911090850830078, "learning_rate": 0.0002, "epoch": 2.8696144208228924, "step": 17750}, {"loss": 0.6554, "grad_norm": 0.8484548926353455, "learning_rate": 0.0002, "epoch": 2.8712311050036377, "step": 17760}, {"loss": 0.6931, "grad_norm": 0.7647597193717957, "learning_rate": 0.0002, "epoch": 2.872847789184383, "step": 17770}, {"loss": 0.6945, "grad_norm": 0.8791151642799377, "learning_rate": 0.0002, "epoch": 2.874464473365128, "step": 17780}, {"loss": 0.7078, "grad_norm": 0.7253178358078003, "learning_rate": 0.0002, "epoch": 2.8760811575458733, "step": 17790}, {"loss": 0.6474, "grad_norm": 0.7956077456474304, "learning_rate": 0.0002, "epoch": 2.8776978417266186, "step": 17800}, {"loss": 0.6687, "grad_norm": 0.8657688498497009, "learning_rate": 0.0002, "epoch": 2.879314525907364, "step": 17810}, {"loss": 0.7171, "grad_norm": 0.7059141993522644, "learning_rate": 0.0002, "epoch": 2.880931210088109, "step": 17820}, {"loss": 0.683, "grad_norm": 0.8886896967887878, "learning_rate": 0.0002, "epoch": 2.8825478942688547, "step": 17830}, {"loss": 0.669, "grad_norm": 0.821032702922821, "learning_rate": 0.0002, "epoch": 2.8841645784496, "step": 17840}, {"loss": 0.6805, "grad_norm": 0.7183963656425476, "learning_rate": 0.0002, "epoch": 2.885781262630345, "step": 17850}, {"loss": 0.7088, "grad_norm": 0.6222899556159973, "learning_rate": 0.0002, "epoch": 2.8873979468110904, "step": 17860}, {"loss": 0.6626, "grad_norm": 0.8187434077262878, "learning_rate": 0.0002, "epoch": 2.8890146309918356, "step": 17870}, {"loss": 0.6815, "grad_norm": 0.9838479161262512, "learning_rate": 0.0002, "epoch": 2.890631315172581, "step": 17880}, {"loss": 0.6967, "grad_norm": 0.7567742466926575, "learning_rate": 0.0002, "epoch": 2.8922479993533265, "step": 17890}, {"loss": 0.7073, "grad_norm": 0.6875903606414795, "learning_rate": 0.0002, "epoch": 2.893864683534072, "step": 17900}, {"loss": 0.6415, "grad_norm": 0.8043789267539978, "learning_rate": 0.0002, "epoch": 2.895481367714817, "step": 17910}, {"loss": 0.6588, "grad_norm": 0.8062626719474792, "learning_rate": 0.0002, "epoch": 2.8970980518955622, "step": 17920}, {"loss": 0.7151, "grad_norm": 1.0251191854476929, "learning_rate": 0.0002, "epoch": 2.8987147360763075, "step": 17930}, {"loss": 0.6605, "grad_norm": 0.882253110408783, "learning_rate": 0.0002, "epoch": 2.9003314202570527, "step": 17940}, {"loss": 0.6719, "grad_norm": 0.8683299422264099, "learning_rate": 0.0002, "epoch": 2.901948104437798, "step": 17950}, {"loss": 0.6896, "grad_norm": 0.7167282104492188, "learning_rate": 0.0002, "epoch": 2.903564788618543, "step": 17960}, {"loss": 0.663, "grad_norm": 0.7093694806098938, "learning_rate": 0.0002, "epoch": 2.9051814727992884, "step": 17970}, {"loss": 0.6591, "grad_norm": 0.8549879193305969, "learning_rate": 0.0002, "epoch": 2.906798156980034, "step": 17980}, {"loss": 0.6962, "grad_norm": 0.6989606618881226, "learning_rate": 0.0002, "epoch": 2.9084148411607793, "step": 17990}, {"loss": 0.6635, "grad_norm": 0.9482976794242859, "learning_rate": 0.0002, "epoch": 2.9100315253415245, "step": 18000}, {"loss": 0.6586, "grad_norm": 0.7182440161705017, "learning_rate": 0.0002, "epoch": 2.9116482095222698, "step": 18010}, {"loss": 0.6827, "grad_norm": 0.7732226252555847, "learning_rate": 0.0002, "epoch": 2.913264893703015, "step": 18020}, {"loss": 0.7123, "grad_norm": 0.7936875224113464, "learning_rate": 0.0002, "epoch": 2.9148815778837607, "step": 18030}, {"loss": 0.6736, "grad_norm": 0.8825615644454956, "learning_rate": 0.0002, "epoch": 2.916498262064506, "step": 18040}, {"loss": 0.7139, "grad_norm": 0.6778587102890015, "learning_rate": 0.0002, "epoch": 2.918114946245251, "step": 18050}, {"loss": 0.6588, "grad_norm": 0.7529265880584717, "learning_rate": 0.0002, "epoch": 2.9197316304259964, "step": 18060}, {"loss": 0.737, "grad_norm": 0.7111883163452148, "learning_rate": 0.0002, "epoch": 2.9213483146067416, "step": 18070}, {"loss": 0.7475, "grad_norm": 0.7214767932891846, "learning_rate": 0.0002, "epoch": 2.922964998787487, "step": 18080}, {"loss": 0.6672, "grad_norm": 0.800417423248291, "learning_rate": 0.0002, "epoch": 2.924581682968232, "step": 18090}, {"loss": 0.6694, "grad_norm": 1.248575210571289, "learning_rate": 0.0002, "epoch": 2.9261983671489773, "step": 18100}, {"loss": 0.7004, "grad_norm": 0.757788360118866, "learning_rate": 0.0002, "epoch": 2.9278150513297225, "step": 18110}, {"loss": 0.6999, "grad_norm": 1.0583995580673218, "learning_rate": 0.0002, "epoch": 2.9294317355104678, "step": 18120}, {"loss": 0.6365, "grad_norm": 0.8228777647018433, "learning_rate": 0.0002, "epoch": 2.9310484196912134, "step": 18130}, {"loss": 0.6791, "grad_norm": 0.8374035358428955, "learning_rate": 0.0002, "epoch": 2.9326651038719587, "step": 18140}, {"loss": 0.6399, "grad_norm": 0.7976473569869995, "learning_rate": 0.0002, "epoch": 2.934281788052704, "step": 18150}, {"loss": 0.6585, "grad_norm": 0.8009907603263855, "learning_rate": 0.0002, "epoch": 2.935898472233449, "step": 18160}, {"loss": 0.7485, "grad_norm": 0.835213303565979, "learning_rate": 0.0002, "epoch": 2.9375151564141944, "step": 18170}, {"loss": 0.7376, "grad_norm": 0.7982219457626343, "learning_rate": 0.0002, "epoch": 2.93913184059494, "step": 18180}, {"loss": 0.6348, "grad_norm": 0.7070978879928589, "learning_rate": 0.0002, "epoch": 2.9407485247756853, "step": 18190}, {"loss": 0.6608, "grad_norm": 0.8619440197944641, "learning_rate": 0.0002, "epoch": 2.9423652089564305, "step": 18200}, {"loss": 0.666, "grad_norm": 0.6693987250328064, "learning_rate": 0.0002, "epoch": 2.9439818931371757, "step": 18210}, {"loss": 0.728, "grad_norm": 0.6747021079063416, "learning_rate": 0.0002, "epoch": 2.945598577317921, "step": 18220}, {"loss": 0.6686, "grad_norm": 0.860387921333313, "learning_rate": 0.0002, "epoch": 2.947215261498666, "step": 18230}, {"loss": 0.6945, "grad_norm": 0.799976646900177, "learning_rate": 0.0002, "epoch": 2.9488319456794114, "step": 18240}, {"loss": 0.7243, "grad_norm": 0.7864769101142883, "learning_rate": 0.0002, "epoch": 2.9504486298601567, "step": 18250}, {"loss": 0.6785, "grad_norm": 0.6713884472846985, "learning_rate": 0.0002, "epoch": 2.952065314040902, "step": 18260}, {"loss": 0.7429, "grad_norm": 0.9031508564949036, "learning_rate": 0.0002, "epoch": 2.9536819982216476, "step": 18270}, {"loss": 0.7055, "grad_norm": 0.7205073237419128, "learning_rate": 0.0002, "epoch": 2.955298682402393, "step": 18280}, {"loss": 0.7298, "grad_norm": 0.7746205925941467, "learning_rate": 0.0002, "epoch": 2.956915366583138, "step": 18290}, {"loss": 0.6218, "grad_norm": 0.6533427834510803, "learning_rate": 0.0002, "epoch": 2.9585320507638833, "step": 18300}, {"loss": 0.6674, "grad_norm": 0.9083208441734314, "learning_rate": 0.0002, "epoch": 2.9601487349446285, "step": 18310}, {"loss": 0.7359, "grad_norm": 0.7446991801261902, "learning_rate": 0.0002, "epoch": 2.9617654191253737, "step": 18320}, {"loss": 0.6738, "grad_norm": 0.6514461636543274, "learning_rate": 0.0002, "epoch": 2.9633821033061194, "step": 18330}, {"loss": 0.6677, "grad_norm": 0.8580465912818909, "learning_rate": 0.0002, "epoch": 2.9649987874868646, "step": 18340}, {"loss": 0.6971, "grad_norm": 0.7074266076087952, "learning_rate": 0.0002, "epoch": 2.96661547166761, "step": 18350}, {"loss": 0.6804, "grad_norm": 0.899892270565033, "learning_rate": 0.0002, "epoch": 2.968232155848355, "step": 18360}, {"loss": 0.7094, "grad_norm": 0.8217641711235046, "learning_rate": 0.0002, "epoch": 2.9698488400291003, "step": 18370}, {"loss": 0.6916, "grad_norm": 0.8611799478530884, "learning_rate": 0.0002, "epoch": 2.9714655242098456, "step": 18380}, {"loss": 0.6677, "grad_norm": 0.6909302473068237, "learning_rate": 0.0002, "epoch": 2.973082208390591, "step": 18390}, {"loss": 0.7247, "grad_norm": 0.6554358005523682, "learning_rate": 0.0002, "epoch": 2.974698892571336, "step": 18400}, {"loss": 0.6516, "grad_norm": 0.7803071737289429, "learning_rate": 0.0002, "epoch": 2.9763155767520812, "step": 18410}, {"loss": 0.7322, "grad_norm": 0.7838954925537109, "learning_rate": 0.0002, "epoch": 2.977932260932827, "step": 18420}, {"loss": 0.6522, "grad_norm": 0.7098495364189148, "learning_rate": 0.0002, "epoch": 2.979548945113572, "step": 18430}, {"loss": 0.739, "grad_norm": 0.8981785774230957, "learning_rate": 0.0002, "epoch": 2.9811656292943174, "step": 18440}, {"loss": 0.6689, "grad_norm": 0.7197171449661255, "learning_rate": 0.0002, "epoch": 2.9827823134750626, "step": 18450}, {"loss": 0.706, "grad_norm": 0.793185293674469, "learning_rate": 0.0002, "epoch": 2.984398997655808, "step": 18460}, {"loss": 0.7124, "grad_norm": 0.8531473875045776, "learning_rate": 0.0002, "epoch": 2.986015681836553, "step": 18470}, {"loss": 0.6901, "grad_norm": 0.6627361178398132, "learning_rate": 0.0002, "epoch": 2.9876323660172988, "step": 18480}, {"loss": 0.6591, "grad_norm": 0.5708155035972595, "learning_rate": 0.0002, "epoch": 2.989249050198044, "step": 18490}, {"loss": 0.6725, "grad_norm": 0.8227280378341675, "learning_rate": 0.0002, "epoch": 2.990865734378789, "step": 18500}, {"loss": 0.6701, "grad_norm": 0.7102749943733215, "learning_rate": 0.0002, "epoch": 2.9924824185595345, "step": 18510}, {"loss": 0.7091, "grad_norm": 0.839485228061676, "learning_rate": 0.0002, "epoch": 2.9940991027402797, "step": 18520}, {"loss": 0.6521, "grad_norm": 0.9038704037666321, "learning_rate": 0.0002, "epoch": 2.995715786921025, "step": 18530}, {"loss": 0.7186, "grad_norm": 0.8737510442733765, "learning_rate": 0.0002, "epoch": 2.99733247110177, "step": 18540}, {"loss": 0.6819, "grad_norm": 0.7323142886161804, "learning_rate": 0.0002, "epoch": 2.9989491552825154, "step": 18550}, {"eval_loss": 1.1262480020523071, "eval_runtime": 122.0868, "eval_samples_per_second": 6.004, "eval_steps_per_second": 0.754, "epoch": 2.9999191657909625, "step": 18556}, {"loss": 0.6337, "grad_norm": 0.8465463519096375, "learning_rate": 0.0002, "epoch": 3.000565839463261, "step": 18560}, {"loss": 0.6064, "grad_norm": 0.9134138822555542, "learning_rate": 0.0002, "epoch": 3.0021825236440063, "step": 18570}, {"loss": 0.5804, "grad_norm": 0.760715126991272, "learning_rate": 0.0002, "epoch": 3.0037992078247515, "step": 18580}, {"loss": 0.5571, "grad_norm": 0.9208743572235107, "learning_rate": 0.0002, "epoch": 3.0054158920054967, "step": 18590}, {"loss": 0.5731, "grad_norm": 0.9232364892959595, "learning_rate": 0.0002, "epoch": 3.007032576186242, "step": 18600}, {"loss": 0.6299, "grad_norm": 1.1881544589996338, "learning_rate": 0.0002, "epoch": 3.008649260366987, "step": 18610}, {"loss": 0.5482, "grad_norm": 0.9372987747192383, "learning_rate": 0.0002, "epoch": 3.0102659445477324, "step": 18620}, {"loss": 0.5709, "grad_norm": 0.6900241374969482, "learning_rate": 0.0002, "epoch": 3.0118826287284777, "step": 18630}, {"loss": 0.5256, "grad_norm": 0.8451071381568909, "learning_rate": 0.0002, "epoch": 3.0134993129092233, "step": 18640}, {"loss": 0.5916, "grad_norm": 0.7763112187385559, "learning_rate": 0.0002, "epoch": 3.0151159970899686, "step": 18650}, {"loss": 0.6095, "grad_norm": 1.043653964996338, "learning_rate": 0.0002, "epoch": 3.016732681270714, "step": 18660}, {"loss": 0.6228, "grad_norm": 1.0170660018920898, "learning_rate": 0.0002, "epoch": 3.018349365451459, "step": 18670}, {"loss": 0.5671, "grad_norm": 0.7534180283546448, "learning_rate": 0.0002, "epoch": 3.0199660496322043, "step": 18680}, {"loss": 0.6015, "grad_norm": 0.7507367730140686, "learning_rate": 0.0002, "epoch": 3.0215827338129495, "step": 18690}, {"loss": 0.6201, "grad_norm": 0.7861620187759399, "learning_rate": 0.0002, "epoch": 3.0231994179936947, "step": 18700}, {"loss": 0.5802, "grad_norm": 1.0580339431762695, "learning_rate": 0.0002, "epoch": 3.0248161021744404, "step": 18710}, {"loss": 0.5975, "grad_norm": 0.7542710900306702, "learning_rate": 0.0002, "epoch": 3.0264327863551856, "step": 18720}, {"loss": 0.5695, "grad_norm": 0.8189544677734375, "learning_rate": 0.0002, "epoch": 3.028049470535931, "step": 18730}, {"loss": 0.6109, "grad_norm": 0.9126611351966858, "learning_rate": 0.0002, "epoch": 3.029666154716676, "step": 18740}, {"loss": 0.6443, "grad_norm": 0.8891341686248779, "learning_rate": 0.0002, "epoch": 3.0312828388974213, "step": 18750}, {"loss": 0.6207, "grad_norm": 0.8419283032417297, "learning_rate": 0.0002, "epoch": 3.0328995230781666, "step": 18760}, {"loss": 0.5818, "grad_norm": 0.8048048615455627, "learning_rate": 0.0002, "epoch": 3.034516207258912, "step": 18770}, {"loss": 0.6381, "grad_norm": 0.7820217609405518, "learning_rate": 0.0002, "epoch": 3.0361328914396575, "step": 18780}, {"loss": 0.5843, "grad_norm": 0.854721188545227, "learning_rate": 0.0002, "epoch": 3.0377495756204027, "step": 18790}, {"loss": 0.5784, "grad_norm": 0.912092924118042, "learning_rate": 0.0002, "epoch": 3.039366259801148, "step": 18800}, {"loss": 0.5734, "grad_norm": 0.6596226096153259, "learning_rate": 0.0002, "epoch": 3.040982943981893, "step": 18810}, {"loss": 0.5969, "grad_norm": 0.6351348757743835, "learning_rate": 0.0002, "epoch": 3.0425996281626384, "step": 18820}, {"loss": 0.5953, "grad_norm": 0.778188943862915, "learning_rate": 0.0002, "epoch": 3.0442163123433836, "step": 18830}, {"loss": 0.602, "grad_norm": 0.68234783411026, "learning_rate": 0.0002, "epoch": 3.045832996524129, "step": 18840}, {"loss": 0.5785, "grad_norm": 0.998628556728363, "learning_rate": 0.0002, "epoch": 3.047449680704874, "step": 18850}, {"loss": 0.6231, "grad_norm": 0.7393841743469238, "learning_rate": 0.0002, "epoch": 3.0490663648856198, "step": 18860}, {"loss": 0.568, "grad_norm": 0.84438556432724, "learning_rate": 0.0002, "epoch": 3.050683049066365, "step": 18870}, {"loss": 0.6205, "grad_norm": 0.8857501745223999, "learning_rate": 0.0002, "epoch": 3.0522997332471102, "step": 18880}, {"loss": 0.6335, "grad_norm": 0.7208474278450012, "learning_rate": 0.0002, "epoch": 3.0539164174278555, "step": 18890}, {"loss": 0.5998, "grad_norm": 0.7135229110717773, "learning_rate": 0.0002, "epoch": 3.0555331016086007, "step": 18900}, {"loss": 0.5575, "grad_norm": 0.9130001664161682, "learning_rate": 0.0002, "epoch": 3.057149785789346, "step": 18910}, {"loss": 0.5955, "grad_norm": 0.9001716375350952, "learning_rate": 0.0002, "epoch": 3.058766469970091, "step": 18920}, {"loss": 0.6052, "grad_norm": 0.8667559623718262, "learning_rate": 0.0002, "epoch": 3.060383154150837, "step": 18930}, {"loss": 0.5818, "grad_norm": 0.8943959474563599, "learning_rate": 0.0002, "epoch": 3.061999838331582, "step": 18940}, {"loss": 0.5978, "grad_norm": 0.8298377990722656, "learning_rate": 0.0002, "epoch": 3.0636165225123273, "step": 18950}, {"loss": 0.5782, "grad_norm": 0.7935267686843872, "learning_rate": 0.0002, "epoch": 3.0652332066930725, "step": 18960}, {"loss": 0.6434, "grad_norm": 1.1506379842758179, "learning_rate": 0.0002, "epoch": 3.0668498908738178, "step": 18970}, {"loss": 0.5571, "grad_norm": 0.7693049907684326, "learning_rate": 0.0002, "epoch": 3.068466575054563, "step": 18980}, {"loss": 0.5971, "grad_norm": 0.8040135502815247, "learning_rate": 0.0002, "epoch": 3.0700832592353082, "step": 18990}, {"loss": 0.5541, "grad_norm": 0.828404426574707, "learning_rate": 0.0002, "epoch": 3.0716999434160535, "step": 19000}, {"loss": 0.6048, "grad_norm": 0.8811164498329163, "learning_rate": 0.0002, "epoch": 3.073316627596799, "step": 19010}, {"loss": 0.5845, "grad_norm": 1.036205768585205, "learning_rate": 0.0002, "epoch": 3.0749333117775444, "step": 19020}, {"loss": 0.5838, "grad_norm": 0.8857285976409912, "learning_rate": 0.0002, "epoch": 3.0765499959582896, "step": 19030}, {"loss": 0.592, "grad_norm": 0.8392079472541809, "learning_rate": 0.0002, "epoch": 3.078166680139035, "step": 19040}, {"loss": 0.5927, "grad_norm": 1.0287401676177979, "learning_rate": 0.0002, "epoch": 3.07978336431978, "step": 19050}, {"loss": 0.5964, "grad_norm": 1.0086315870285034, "learning_rate": 0.0002, "epoch": 3.0814000485005253, "step": 19060}, {"loss": 0.5567, "grad_norm": 0.9245324730873108, "learning_rate": 0.0002, "epoch": 3.0830167326812705, "step": 19070}, {"loss": 0.5797, "grad_norm": 0.8680877089500427, "learning_rate": 0.0002, "epoch": 3.084633416862016, "step": 19080}, {"loss": 0.5611, "grad_norm": 0.8814793825149536, "learning_rate": 0.0002, "epoch": 3.0862501010427614, "step": 19090}, {"loss": 0.6051, "grad_norm": 0.9234458208084106, "learning_rate": 0.0002, "epoch": 3.0878667852235067, "step": 19100}, {"loss": 0.6209, "grad_norm": 1.1291664838790894, "learning_rate": 0.0002, "epoch": 3.089483469404252, "step": 19110}, {"loss": 0.5695, "grad_norm": 0.9191402792930603, "learning_rate": 0.0002, "epoch": 3.091100153584997, "step": 19120}, {"loss": 0.5856, "grad_norm": 0.7103154063224792, "learning_rate": 0.0002, "epoch": 3.0927168377657424, "step": 19130}, {"loss": 0.6479, "grad_norm": 0.9368883967399597, "learning_rate": 0.0002, "epoch": 3.0943335219464876, "step": 19140}, {"loss": 0.6167, "grad_norm": 0.9676656723022461, "learning_rate": 0.0002, "epoch": 3.095950206127233, "step": 19150}, {"loss": 0.5794, "grad_norm": 0.8739792704582214, "learning_rate": 0.0002, "epoch": 3.0975668903079785, "step": 19160}, {"loss": 0.6112, "grad_norm": 0.8530174493789673, "learning_rate": 0.0002, "epoch": 3.0991835744887237, "step": 19170}, {"loss": 0.6568, "grad_norm": 0.794945478439331, "learning_rate": 0.0002, "epoch": 3.100800258669469, "step": 19180}, {"loss": 0.5928, "grad_norm": 0.9508888125419617, "learning_rate": 0.0002, "epoch": 3.102416942850214, "step": 19190}, {"loss": 0.5757, "grad_norm": 1.0599955320358276, "learning_rate": 0.0002, "epoch": 3.1040336270309594, "step": 19200}, {"loss": 0.6151, "grad_norm": 1.0673625469207764, "learning_rate": 0.0002, "epoch": 3.1056503112117047, "step": 19210}, {"loss": 0.6043, "grad_norm": 0.7739115953445435, "learning_rate": 0.0002, "epoch": 3.10726699539245, "step": 19220}, {"loss": 0.6046, "grad_norm": 0.9884951114654541, "learning_rate": 0.0002, "epoch": 3.1088836795731956, "step": 19230}, {"loss": 0.5932, "grad_norm": 0.862260103225708, "learning_rate": 0.0002, "epoch": 3.110500363753941, "step": 19240}, {"loss": 0.6098, "grad_norm": 0.7690284848213196, "learning_rate": 0.0002, "epoch": 3.112117047934686, "step": 19250}, {"loss": 0.5791, "grad_norm": 0.8758958578109741, "learning_rate": 0.0002, "epoch": 3.1137337321154313, "step": 19260}, {"loss": 0.6136, "grad_norm": 1.0356395244598389, "learning_rate": 0.0002, "epoch": 3.1153504162961765, "step": 19270}, {"loss": 0.6159, "grad_norm": 0.6950937509536743, "learning_rate": 0.0002, "epoch": 3.1169671004769217, "step": 19280}, {"loss": 0.592, "grad_norm": 0.760998010635376, "learning_rate": 0.0002, "epoch": 3.118583784657667, "step": 19290}, {"loss": 0.575, "grad_norm": 0.9335789084434509, "learning_rate": 0.0002, "epoch": 3.1202004688384126, "step": 19300}, {"loss": 0.6139, "grad_norm": 0.9636204242706299, "learning_rate": 0.0002, "epoch": 3.121817153019158, "step": 19310}, {"loss": 0.6001, "grad_norm": 1.0820997953414917, "learning_rate": 0.0002, "epoch": 3.123433837199903, "step": 19320}, {"loss": 0.6542, "grad_norm": 0.7333487272262573, "learning_rate": 0.0002, "epoch": 3.1250505213806483, "step": 19330}, {"loss": 0.6178, "grad_norm": 1.0417509078979492, "learning_rate": 0.0002, "epoch": 3.1266672055613935, "step": 19340}, {"loss": 0.603, "grad_norm": 0.9267749190330505, "learning_rate": 0.0002, "epoch": 3.128283889742139, "step": 19350}, {"loss": 0.6063, "grad_norm": 0.777798593044281, "learning_rate": 0.0002, "epoch": 3.129900573922884, "step": 19360}, {"loss": 0.5913, "grad_norm": 0.8425456881523132, "learning_rate": 0.0002, "epoch": 3.1315172581036297, "step": 19370}, {"loss": 0.6042, "grad_norm": 0.9617102146148682, "learning_rate": 0.0002, "epoch": 3.133133942284375, "step": 19380}, {"loss": 0.633, "grad_norm": 1.0052828788757324, "learning_rate": 0.0002, "epoch": 3.13475062646512, "step": 19390}, {"loss": 0.5713, "grad_norm": 0.7637009024620056, "learning_rate": 0.0002, "epoch": 3.1363673106458654, "step": 19400}, {"loss": 0.5497, "grad_norm": 0.7958088517189026, "learning_rate": 0.0002, "epoch": 3.1379839948266106, "step": 19410}, {"loss": 0.6283, "grad_norm": 0.9161727428436279, "learning_rate": 0.0002, "epoch": 3.139600679007356, "step": 19420}, {"loss": 0.5638, "grad_norm": 0.8402149677276611, "learning_rate": 0.0002, "epoch": 3.141217363188101, "step": 19430}, {"loss": 0.5848, "grad_norm": 1.0056525468826294, "learning_rate": 0.0002, "epoch": 3.1428340473688463, "step": 19440}, {"loss": 0.5954, "grad_norm": 1.0129190683364868, "learning_rate": 0.0002, "epoch": 3.144450731549592, "step": 19450}, {"loss": 0.5808, "grad_norm": 0.790825366973877, "learning_rate": 0.0002, "epoch": 3.146067415730337, "step": 19460}, {"loss": 0.5607, "grad_norm": 1.441665530204773, "learning_rate": 0.0002, "epoch": 3.1476840999110824, "step": 19470}, {"loss": 0.5785, "grad_norm": 0.7846331596374512, "learning_rate": 0.0002, "epoch": 3.1493007840918277, "step": 19480}, {"loss": 0.5892, "grad_norm": 0.7915332913398743, "learning_rate": 0.0002, "epoch": 3.150917468272573, "step": 19490}, {"loss": 0.5759, "grad_norm": 0.933982253074646, "learning_rate": 0.0002, "epoch": 3.152534152453318, "step": 19500}, {"loss": 0.6206, "grad_norm": 1.038408637046814, "learning_rate": 0.0002, "epoch": 3.1541508366340634, "step": 19510}, {"loss": 0.6271, "grad_norm": 1.018935203552246, "learning_rate": 0.0002, "epoch": 3.155767520814809, "step": 19520}, {"loss": 0.6173, "grad_norm": 0.9618112444877625, "learning_rate": 0.0002, "epoch": 3.1573842049955543, "step": 19530}, {"loss": 0.5972, "grad_norm": 0.8900452852249146, "learning_rate": 0.0002, "epoch": 3.1590008891762995, "step": 19540}, {"loss": 0.5925, "grad_norm": 0.8254160284996033, "learning_rate": 0.0002, "epoch": 3.1606175733570447, "step": 19550}, {"loss": 0.625, "grad_norm": 1.004376769065857, "learning_rate": 0.0002, "epoch": 3.16223425753779, "step": 19560}, {"loss": 0.5775, "grad_norm": 1.0490446090698242, "learning_rate": 0.0002, "epoch": 3.163850941718535, "step": 19570}, {"loss": 0.5986, "grad_norm": 0.7387403845787048, "learning_rate": 0.0002, "epoch": 3.1654676258992804, "step": 19580}, {"loss": 0.5898, "grad_norm": 0.7611538171768188, "learning_rate": 0.0002, "epoch": 3.1670843100800257, "step": 19590}, {"loss": 0.5937, "grad_norm": 0.8239886164665222, "learning_rate": 0.0002, "epoch": 3.1687009942607713, "step": 19600}, {"loss": 0.6068, "grad_norm": 0.9327243566513062, "learning_rate": 0.0002, "epoch": 3.1703176784415166, "step": 19610}, {"loss": 0.572, "grad_norm": 0.9662560224533081, "learning_rate": 0.0002, "epoch": 3.171934362622262, "step": 19620}, {"loss": 0.5988, "grad_norm": 0.9183341860771179, "learning_rate": 0.0002, "epoch": 3.173551046803007, "step": 19630}, {"loss": 0.5909, "grad_norm": 0.875066876411438, "learning_rate": 0.0002, "epoch": 3.1751677309837523, "step": 19640}, {"loss": 0.5956, "grad_norm": 0.8567508459091187, "learning_rate": 0.0002, "epoch": 3.1767844151644975, "step": 19650}, {"loss": 0.5805, "grad_norm": 0.6805780529975891, "learning_rate": 0.0002, "epoch": 3.1784010993452427, "step": 19660}, {"loss": 0.6204, "grad_norm": 0.8776944279670715, "learning_rate": 0.0002, "epoch": 3.1800177835259884, "step": 19670}, {"loss": 0.6108, "grad_norm": 0.9036329984664917, "learning_rate": 0.0002, "epoch": 3.1816344677067336, "step": 19680}, {"loss": 0.6238, "grad_norm": 0.8527372479438782, "learning_rate": 0.0002, "epoch": 3.183251151887479, "step": 19690}, {"loss": 0.6089, "grad_norm": 1.1045585870742798, "learning_rate": 0.0002, "epoch": 3.184867836068224, "step": 19700}, {"loss": 0.5491, "grad_norm": 0.9213830828666687, "learning_rate": 0.0002, "epoch": 3.1864845202489693, "step": 19710}, {"loss": 0.618, "grad_norm": 0.8865814805030823, "learning_rate": 0.0002, "epoch": 3.1881012044297146, "step": 19720}, {"loss": 0.5785, "grad_norm": 0.7939388751983643, "learning_rate": 0.0002, "epoch": 3.18971788861046, "step": 19730}, {"loss": 0.5682, "grad_norm": 0.6966729760169983, "learning_rate": 0.0002, "epoch": 3.191334572791205, "step": 19740}, {"loss": 0.5839, "grad_norm": 0.8023673295974731, "learning_rate": 0.0002, "epoch": 3.1929512569719507, "step": 19750}, {"loss": 0.6267, "grad_norm": 0.7992037534713745, "learning_rate": 0.0002, "epoch": 3.194567941152696, "step": 19760}, {"loss": 0.6141, "grad_norm": 0.7412247657775879, "learning_rate": 0.0002, "epoch": 3.196184625333441, "step": 19770}, {"loss": 0.6179, "grad_norm": 0.9598729014396667, "learning_rate": 0.0002, "epoch": 3.1978013095141864, "step": 19780}, {"loss": 0.5685, "grad_norm": 0.8331366777420044, "learning_rate": 0.0002, "epoch": 3.1994179936949316, "step": 19790}, {"loss": 0.6104, "grad_norm": 0.8939169645309448, "learning_rate": 0.0002, "epoch": 3.201034677875677, "step": 19800}, {"loss": 0.6147, "grad_norm": 0.9219734072685242, "learning_rate": 0.0002, "epoch": 3.202651362056422, "step": 19810}, {"loss": 0.6051, "grad_norm": 0.869490385055542, "learning_rate": 0.0002, "epoch": 3.2042680462371678, "step": 19820}, {"loss": 0.5946, "grad_norm": 0.8989706635475159, "learning_rate": 0.0002, "epoch": 3.205884730417913, "step": 19830}, {"loss": 0.5866, "grad_norm": 0.8477165102958679, "learning_rate": 0.0002, "epoch": 3.2075014145986582, "step": 19840}, {"loss": 0.6176, "grad_norm": 0.8720678687095642, "learning_rate": 0.0002, "epoch": 3.2091180987794035, "step": 19850}, {"loss": 0.5694, "grad_norm": 0.861406683921814, "learning_rate": 0.0002, "epoch": 3.2107347829601487, "step": 19860}, {"loss": 0.6264, "grad_norm": 0.8228686451911926, "learning_rate": 0.0002, "epoch": 3.212351467140894, "step": 19870}, {"loss": 0.625, "grad_norm": 0.7936596870422363, "learning_rate": 0.0002, "epoch": 3.213968151321639, "step": 19880}, {"loss": 0.5698, "grad_norm": 1.097377896308899, "learning_rate": 0.0002, "epoch": 3.2155848355023844, "step": 19890}, {"loss": 0.6725, "grad_norm": 0.9544782638549805, "learning_rate": 0.0002, "epoch": 3.21720151968313, "step": 19900}, {"loss": 0.6022, "grad_norm": 0.8240751624107361, "learning_rate": 0.0002, "epoch": 3.2188182038638753, "step": 19910}, {"loss": 0.5659, "grad_norm": 0.8332096338272095, "learning_rate": 0.0002, "epoch": 3.2204348880446205, "step": 19920}, {"loss": 0.6274, "grad_norm": 1.0954567193984985, "learning_rate": 0.0002, "epoch": 3.2220515722253658, "step": 19930}, {"loss": 0.652, "grad_norm": 0.7790525555610657, "learning_rate": 0.0002, "epoch": 3.223668256406111, "step": 19940}, {"loss": 0.5986, "grad_norm": 0.7966814041137695, "learning_rate": 0.0002, "epoch": 3.225284940586856, "step": 19950}, {"loss": 0.5911, "grad_norm": 0.9751881957054138, "learning_rate": 0.0002, "epoch": 3.2269016247676015, "step": 19960}, {"loss": 0.6071, "grad_norm": 0.9856047630310059, "learning_rate": 0.0002, "epoch": 3.228518308948347, "step": 19970}, {"loss": 0.5837, "grad_norm": 1.3062353134155273, "learning_rate": 0.0002, "epoch": 3.2301349931290924, "step": 19980}, {"loss": 0.6588, "grad_norm": 0.9510692358016968, "learning_rate": 0.0002, "epoch": 3.2317516773098376, "step": 19990}, {"loss": 0.6264, "grad_norm": 0.8630342483520508, "learning_rate": 0.0002, "epoch": 3.233368361490583, "step": 20000}, {"loss": 0.6073, "grad_norm": 0.8966519236564636, "learning_rate": 0.0002, "epoch": 3.234985045671328, "step": 20010}, {"loss": 0.612, "grad_norm": 0.7093510627746582, "learning_rate": 0.0002, "epoch": 3.2366017298520733, "step": 20020}, {"loss": 0.585, "grad_norm": 0.7771096229553223, "learning_rate": 0.0002, "epoch": 3.2382184140328185, "step": 20030}, {"loss": 0.5821, "grad_norm": 0.841058075428009, "learning_rate": 0.0002, "epoch": 3.2398350982135637, "step": 20040}, {"loss": 0.6519, "grad_norm": 0.909712553024292, "learning_rate": 0.0002, "epoch": 3.2414517823943094, "step": 20050}, {"loss": 0.6089, "grad_norm": 0.8321019411087036, "learning_rate": 0.0002, "epoch": 3.2430684665750547, "step": 20060}, {"loss": 0.6115, "grad_norm": 0.779901921749115, "learning_rate": 0.0002, "epoch": 3.2446851507558, "step": 20070}, {"loss": 0.6107, "grad_norm": 0.6249170303344727, "learning_rate": 0.0002, "epoch": 3.246301834936545, "step": 20080}, {"loss": 0.603, "grad_norm": 0.8000940680503845, "learning_rate": 0.0002, "epoch": 3.2479185191172903, "step": 20090}, {"loss": 0.6273, "grad_norm": 0.7627735137939453, "learning_rate": 0.0002, "epoch": 3.2495352032980356, "step": 20100}, {"loss": 0.6223, "grad_norm": 0.8780747056007385, "learning_rate": 0.0002, "epoch": 3.2511518874787813, "step": 20110}, {"loss": 0.5969, "grad_norm": 0.772037148475647, "learning_rate": 0.0002, "epoch": 3.2527685716595265, "step": 20120}, {"loss": 0.5843, "grad_norm": 1.0086580514907837, "learning_rate": 0.0002, "epoch": 3.2543852558402717, "step": 20130}, {"loss": 0.5777, "grad_norm": 0.9360289573669434, "learning_rate": 0.0002, "epoch": 3.256001940021017, "step": 20140}, {"loss": 0.5777, "grad_norm": 1.2099586725234985, "learning_rate": 0.0002, "epoch": 3.257618624201762, "step": 20150}, {"loss": 0.624, "grad_norm": 0.8368481397628784, "learning_rate": 0.0002, "epoch": 3.2592353083825074, "step": 20160}, {"loss": 0.5626, "grad_norm": 0.7391039133071899, "learning_rate": 0.0002, "epoch": 3.2608519925632526, "step": 20170}, {"loss": 0.6041, "grad_norm": 0.9122273325920105, "learning_rate": 0.0002, "epoch": 3.262468676743998, "step": 20180}, {"loss": 0.5868, "grad_norm": 0.8502281904220581, "learning_rate": 0.0002, "epoch": 3.264085360924743, "step": 20190}, {"loss": 0.5841, "grad_norm": 1.0926852226257324, "learning_rate": 0.0002, "epoch": 3.265702045105489, "step": 20200}, {"loss": 0.6027, "grad_norm": 0.7902828454971313, "learning_rate": 0.0002, "epoch": 3.267318729286234, "step": 20210}, {"loss": 0.6089, "grad_norm": 0.8724729418754578, "learning_rate": 0.0002, "epoch": 3.2689354134669792, "step": 20220}, {"loss": 0.6242, "grad_norm": 0.8469277024269104, "learning_rate": 0.0002, "epoch": 3.2705520976477245, "step": 20230}, {"loss": 0.644, "grad_norm": 0.8865092992782593, "learning_rate": 0.0002, "epoch": 3.2721687818284697, "step": 20240}, {"loss": 0.6464, "grad_norm": 1.0979334115982056, "learning_rate": 0.0002, "epoch": 3.273785466009215, "step": 20250}, {"loss": 0.647, "grad_norm": 1.0860793590545654, "learning_rate": 0.0002, "epoch": 3.2754021501899606, "step": 20260}, {"loss": 0.6105, "grad_norm": 0.981745183467865, "learning_rate": 0.0002, "epoch": 3.277018834370706, "step": 20270}, {"loss": 0.627, "grad_norm": 0.9155020713806152, "learning_rate": 0.0002, "epoch": 3.278635518551451, "step": 20280}, {"loss": 0.5899, "grad_norm": 0.8436718583106995, "learning_rate": 0.0002, "epoch": 3.2802522027321963, "step": 20290}, {"loss": 0.6371, "grad_norm": 1.0329409837722778, "learning_rate": 0.0002, "epoch": 3.2818688869129415, "step": 20300}, {"loss": 0.6, "grad_norm": 0.9876394271850586, "learning_rate": 0.0002, "epoch": 3.2834855710936868, "step": 20310}, {"loss": 0.5463, "grad_norm": 0.8052917718887329, "learning_rate": 0.0002, "epoch": 3.285102255274432, "step": 20320}, {"loss": 0.5949, "grad_norm": 0.8390680551528931, "learning_rate": 0.0002, "epoch": 3.2867189394551772, "step": 20330}, {"loss": 0.6492, "grad_norm": 0.9515735507011414, "learning_rate": 0.0002, "epoch": 3.288335623635923, "step": 20340}, {"loss": 0.596, "grad_norm": 0.8028870224952698, "learning_rate": 0.0002, "epoch": 3.289952307816668, "step": 20350}, {"loss": 0.634, "grad_norm": 0.862592339515686, "learning_rate": 0.0002, "epoch": 3.2915689919974134, "step": 20360}, {"loss": 0.6345, "grad_norm": 0.7451621890068054, "learning_rate": 0.0002, "epoch": 3.2931856761781586, "step": 20370}, {"loss": 0.6458, "grad_norm": 0.8966776728630066, "learning_rate": 0.0002, "epoch": 3.294802360358904, "step": 20380}, {"loss": 0.5967, "grad_norm": 0.9289216995239258, "learning_rate": 0.0002, "epoch": 3.296419044539649, "step": 20390}, {"loss": 0.6599, "grad_norm": 0.9649626612663269, "learning_rate": 0.0002, "epoch": 3.2980357287203943, "step": 20400}, {"loss": 0.5781, "grad_norm": 1.1953798532485962, "learning_rate": 0.0002, "epoch": 3.29965241290114, "step": 20410}, {"loss": 0.5997, "grad_norm": 0.8929083943367004, "learning_rate": 0.0002, "epoch": 3.301269097081885, "step": 20420}, {"loss": 0.597, "grad_norm": 0.8922014236450195, "learning_rate": 0.0002, "epoch": 3.3028857812626304, "step": 20430}, {"loss": 0.5766, "grad_norm": 0.9754860401153564, "learning_rate": 0.0002, "epoch": 3.3045024654433757, "step": 20440}, {"loss": 0.5653, "grad_norm": 0.8873140215873718, "learning_rate": 0.0002, "epoch": 3.306119149624121, "step": 20450}, {"loss": 0.6138, "grad_norm": 0.857271671295166, "learning_rate": 0.0002, "epoch": 3.307735833804866, "step": 20460}, {"loss": 0.633, "grad_norm": 0.9022141098976135, "learning_rate": 0.0002, "epoch": 3.3093525179856114, "step": 20470}, {"loss": 0.6654, "grad_norm": 0.8614798188209534, "learning_rate": 0.0002, "epoch": 3.3109692021663566, "step": 20480}, {"loss": 0.6254, "grad_norm": 0.8838164210319519, "learning_rate": 0.0002, "epoch": 3.3125858863471023, "step": 20490}, {"loss": 0.5849, "grad_norm": 0.8709736466407776, "learning_rate": 0.0002, "epoch": 3.3142025705278475, "step": 20500}, {"loss": 0.6146, "grad_norm": 0.9533300995826721, "learning_rate": 0.0002, "epoch": 3.3158192547085927, "step": 20510}, {"loss": 0.6029, "grad_norm": 0.8259269595146179, "learning_rate": 0.0002, "epoch": 3.317435938889338, "step": 20520}, {"loss": 0.6268, "grad_norm": 0.8607608079910278, "learning_rate": 0.0002, "epoch": 3.319052623070083, "step": 20530}, {"loss": 0.5676, "grad_norm": 1.0863020420074463, "learning_rate": 0.0002, "epoch": 3.3206693072508284, "step": 20540}, {"loss": 0.6412, "grad_norm": 1.011489987373352, "learning_rate": 0.0002, "epoch": 3.3222859914315737, "step": 20550}, {"loss": 0.6247, "grad_norm": 0.6952177882194519, "learning_rate": 0.0002, "epoch": 3.3239026756123193, "step": 20560}, {"loss": 0.6229, "grad_norm": 0.9638974070549011, "learning_rate": 0.0002, "epoch": 3.3255193597930646, "step": 20570}, {"loss": 0.5882, "grad_norm": 1.0310138463974, "learning_rate": 0.0002, "epoch": 3.32713604397381, "step": 20580}, {"loss": 0.594, "grad_norm": 0.9371318221092224, "learning_rate": 0.0002, "epoch": 3.328752728154555, "step": 20590}, {"loss": 0.6137, "grad_norm": 0.8756691813468933, "learning_rate": 0.0002, "epoch": 3.3303694123353003, "step": 20600}, {"loss": 0.5994, "grad_norm": 1.054175853729248, "learning_rate": 0.0002, "epoch": 3.3319860965160455, "step": 20610}, {"loss": 0.6169, "grad_norm": 0.9074128270149231, "learning_rate": 0.0002, "epoch": 3.3336027806967907, "step": 20620}, {"loss": 0.6138, "grad_norm": 0.906900942325592, "learning_rate": 0.0002, "epoch": 3.335219464877536, "step": 20630}, {"loss": 0.571, "grad_norm": 0.8689333200454712, "learning_rate": 0.0002, "epoch": 3.3368361490582816, "step": 20640}, {"loss": 0.6079, "grad_norm": 0.9889747500419617, "learning_rate": 0.0002, "epoch": 3.338452833239027, "step": 20650}, {"loss": 0.6073, "grad_norm": 1.0685805082321167, "learning_rate": 0.0002, "epoch": 3.340069517419772, "step": 20660}, {"loss": 0.6091, "grad_norm": 0.7495010495185852, "learning_rate": 0.0002, "epoch": 3.3416862016005173, "step": 20670}, {"loss": 0.5883, "grad_norm": 0.8747848272323608, "learning_rate": 0.0002, "epoch": 3.3433028857812626, "step": 20680}, {"loss": 0.604, "grad_norm": 0.9762673377990723, "learning_rate": 0.0002, "epoch": 3.344919569962008, "step": 20690}, {"loss": 0.6784, "grad_norm": 1.0284489393234253, "learning_rate": 0.0002, "epoch": 3.346536254142753, "step": 20700}, {"loss": 0.6464, "grad_norm": 0.7293812036514282, "learning_rate": 0.0002, "epoch": 3.3481529383234987, "step": 20710}, {"loss": 0.609, "grad_norm": 0.8330199122428894, "learning_rate": 0.0002, "epoch": 3.349769622504244, "step": 20720}, {"loss": 0.5729, "grad_norm": 0.9808499217033386, "learning_rate": 0.0002, "epoch": 3.351386306684989, "step": 20730}, {"loss": 0.6315, "grad_norm": 0.9508825540542603, "learning_rate": 0.0002, "epoch": 3.3530029908657344, "step": 20740}, {"loss": 0.5965, "grad_norm": 0.790483832359314, "learning_rate": 0.0002, "epoch": 3.3546196750464796, "step": 20750}, {"loss": 0.6327, "grad_norm": 1.022793173789978, "learning_rate": 0.0002, "epoch": 3.356236359227225, "step": 20760}, {"loss": 0.6439, "grad_norm": 0.8318950533866882, "learning_rate": 0.0002, "epoch": 3.35785304340797, "step": 20770}, {"loss": 0.6037, "grad_norm": 0.7980858087539673, "learning_rate": 0.0002, "epoch": 3.3594697275887153, "step": 20780}, {"loss": 0.6746, "grad_norm": 0.8114802241325378, "learning_rate": 0.0002, "epoch": 3.361086411769461, "step": 20790}, {"loss": 0.6017, "grad_norm": 0.8522519469261169, "learning_rate": 0.0002, "epoch": 3.3627030959502062, "step": 20800}, {"loss": 0.5864, "grad_norm": 0.9142431616783142, "learning_rate": 0.0002, "epoch": 3.3643197801309515, "step": 20810}, {"loss": 0.6331, "grad_norm": 0.771170437335968, "learning_rate": 0.0002, "epoch": 3.3659364643116967, "step": 20820}, {"loss": 0.5879, "grad_norm": 1.0628231763839722, "learning_rate": 0.0002, "epoch": 3.367553148492442, "step": 20830}, {"loss": 0.6533, "grad_norm": 0.9384352564811707, "learning_rate": 0.0002, "epoch": 3.369169832673187, "step": 20840}, {"loss": 0.6292, "grad_norm": 1.1286591291427612, "learning_rate": 0.0002, "epoch": 3.370786516853933, "step": 20850}, {"loss": 0.5986, "grad_norm": 1.1349513530731201, "learning_rate": 0.0002, "epoch": 3.372403201034678, "step": 20860}, {"loss": 0.6413, "grad_norm": 1.0127464532852173, "learning_rate": 0.0002, "epoch": 3.3740198852154233, "step": 20870}, {"loss": 0.6414, "grad_norm": 0.9111971855163574, "learning_rate": 0.0002, "epoch": 3.3756365693961685, "step": 20880}, {"loss": 0.6101, "grad_norm": 0.871356725692749, "learning_rate": 0.0002, "epoch": 3.3772532535769137, "step": 20890}, {"loss": 0.5995, "grad_norm": 0.7774117588996887, "learning_rate": 0.0002, "epoch": 3.378869937757659, "step": 20900}, {"loss": 0.6062, "grad_norm": 1.0089964866638184, "learning_rate": 0.0002, "epoch": 3.380486621938404, "step": 20910}, {"loss": 0.5908, "grad_norm": 0.7855867147445679, "learning_rate": 0.0002, "epoch": 3.3821033061191494, "step": 20920}, {"loss": 0.6373, "grad_norm": 1.3713710308074951, "learning_rate": 0.0002, "epoch": 3.3837199902998947, "step": 20930}, {"loss": 0.6627, "grad_norm": 0.8599116206169128, "learning_rate": 0.0002, "epoch": 3.3853366744806404, "step": 20940}, {"loss": 0.6224, "grad_norm": 0.9392673373222351, "learning_rate": 0.0002, "epoch": 3.3869533586613856, "step": 20950}, {"loss": 0.5855, "grad_norm": 0.8764075040817261, "learning_rate": 0.0002, "epoch": 3.388570042842131, "step": 20960}, {"loss": 0.5734, "grad_norm": 0.8240136504173279, "learning_rate": 0.0002, "epoch": 3.390186727022876, "step": 20970}, {"loss": 0.5783, "grad_norm": 1.0982369184494019, "learning_rate": 0.0002, "epoch": 3.3918034112036213, "step": 20980}, {"loss": 0.5451, "grad_norm": 1.0599013566970825, "learning_rate": 0.0002, "epoch": 3.3934200953843665, "step": 20990}, {"loss": 0.6356, "grad_norm": 0.895438015460968, "learning_rate": 0.0002, "epoch": 3.395036779565112, "step": 21000}, {"loss": 0.6065, "grad_norm": 0.6974841356277466, "learning_rate": 0.0002, "epoch": 3.3966534637458574, "step": 21010}, {"loss": 0.5704, "grad_norm": 0.9571719765663147, "learning_rate": 0.0002, "epoch": 3.3982701479266026, "step": 21020}, {"loss": 0.679, "grad_norm": 0.831912636756897, "learning_rate": 0.0002, "epoch": 3.399886832107348, "step": 21030}, {"loss": 0.6051, "grad_norm": 0.831936240196228, "learning_rate": 0.0002, "epoch": 3.401503516288093, "step": 21040}, {"loss": 0.5857, "grad_norm": 0.7388373613357544, "learning_rate": 0.0002, "epoch": 3.4031202004688383, "step": 21050}, {"loss": 0.6245, "grad_norm": 0.938667356967926, "learning_rate": 0.0002, "epoch": 3.4047368846495836, "step": 21060}, {"loss": 0.6121, "grad_norm": 0.9202313423156738, "learning_rate": 0.0002, "epoch": 3.406353568830329, "step": 21070}, {"loss": 0.6388, "grad_norm": 0.9888381958007812, "learning_rate": 0.0002, "epoch": 3.4079702530110745, "step": 21080}, {"loss": 0.6245, "grad_norm": 0.8526970744132996, "learning_rate": 0.0002, "epoch": 3.4095869371918197, "step": 21090}, {"loss": 0.5914, "grad_norm": 0.7939383387565613, "learning_rate": 0.0002, "epoch": 3.411203621372565, "step": 21100}, {"loss": 0.6066, "grad_norm": 0.9986352920532227, "learning_rate": 0.0002, "epoch": 3.41282030555331, "step": 21110}, {"loss": 0.5947, "grad_norm": 0.8895300030708313, "learning_rate": 0.0002, "epoch": 3.4144369897340554, "step": 21120}, {"loss": 0.6264, "grad_norm": 0.9559482932090759, "learning_rate": 0.0002, "epoch": 3.4160536739148006, "step": 21130}, {"loss": 0.6491, "grad_norm": 0.8351506590843201, "learning_rate": 0.0002, "epoch": 3.417670358095546, "step": 21140}, {"loss": 0.567, "grad_norm": 0.8224456906318665, "learning_rate": 0.0002, "epoch": 3.4192870422762915, "step": 21150}, {"loss": 0.5871, "grad_norm": 1.0110299587249756, "learning_rate": 0.0002, "epoch": 3.4209037264570368, "step": 21160}, {"loss": 0.6116, "grad_norm": 0.82564777135849, "learning_rate": 0.0002, "epoch": 3.422520410637782, "step": 21170}, {"loss": 0.595, "grad_norm": 1.004738688468933, "learning_rate": 0.0002, "epoch": 3.4241370948185272, "step": 21180}, {"loss": 0.6286, "grad_norm": 0.7545676827430725, "learning_rate": 0.0002, "epoch": 3.4257537789992725, "step": 21190}, {"loss": 0.5868, "grad_norm": 0.8918704390525818, "learning_rate": 0.0002, "epoch": 3.4273704631800177, "step": 21200}, {"loss": 0.6542, "grad_norm": 0.8336876034736633, "learning_rate": 0.0002, "epoch": 3.428987147360763, "step": 21210}, {"loss": 0.5824, "grad_norm": 0.8928771018981934, "learning_rate": 0.0002, "epoch": 3.430603831541508, "step": 21220}, {"loss": 0.6468, "grad_norm": 0.7663705945014954, "learning_rate": 0.0002, "epoch": 3.432220515722254, "step": 21230}, {"loss": 0.6693, "grad_norm": 0.8392598628997803, "learning_rate": 0.0002, "epoch": 3.433837199902999, "step": 21240}, {"loss": 0.5971, "grad_norm": 0.8819600343704224, "learning_rate": 0.0002, "epoch": 3.4354538840837443, "step": 21250}, {"loss": 0.6791, "grad_norm": 0.9124642014503479, "learning_rate": 0.0002, "epoch": 3.4370705682644895, "step": 21260}, {"loss": 0.5925, "grad_norm": 0.8329763412475586, "learning_rate": 0.0002, "epoch": 3.4386872524452348, "step": 21270}, {"loss": 0.6541, "grad_norm": 0.9982839822769165, "learning_rate": 0.0002, "epoch": 3.44030393662598, "step": 21280}, {"loss": 0.6441, "grad_norm": 0.9105954766273499, "learning_rate": 0.0002, "epoch": 3.4419206208067252, "step": 21290}, {"loss": 0.6028, "grad_norm": 0.8182359337806702, "learning_rate": 0.0002, "epoch": 3.443537304987471, "step": 21300}, {"loss": 0.5991, "grad_norm": 1.0568904876708984, "learning_rate": 0.0002, "epoch": 3.445153989168216, "step": 21310}, {"loss": 0.6117, "grad_norm": 0.968539834022522, "learning_rate": 0.0002, "epoch": 3.4467706733489614, "step": 21320}, {"loss": 0.6219, "grad_norm": 0.8774511218070984, "learning_rate": 0.0002, "epoch": 3.4483873575297066, "step": 21330}, {"loss": 0.6438, "grad_norm": 0.7598156332969666, "learning_rate": 0.0002, "epoch": 3.450004041710452, "step": 21340}, {"loss": 0.6033, "grad_norm": 1.1012897491455078, "learning_rate": 0.0002, "epoch": 3.451620725891197, "step": 21350}, {"loss": 0.6137, "grad_norm": 0.8040637373924255, "learning_rate": 0.0002, "epoch": 3.4532374100719423, "step": 21360}, {"loss": 0.6173, "grad_norm": 0.8497496247291565, "learning_rate": 0.0002, "epoch": 3.4548540942526875, "step": 21370}, {"loss": 0.6005, "grad_norm": 0.8429915904998779, "learning_rate": 0.0002, "epoch": 3.456470778433433, "step": 21380}, {"loss": 0.6182, "grad_norm": 0.8107112646102905, "learning_rate": 0.0002, "epoch": 3.4580874626141784, "step": 21390}, {"loss": 0.6109, "grad_norm": 1.00872004032135, "learning_rate": 0.0002, "epoch": 3.4597041467949237, "step": 21400}, {"loss": 0.5712, "grad_norm": 0.8266542553901672, "learning_rate": 0.0002, "epoch": 3.461320830975669, "step": 21410}, {"loss": 0.6457, "grad_norm": 0.8972568511962891, "learning_rate": 0.0002, "epoch": 3.462937515156414, "step": 21420}, {"loss": 0.6081, "grad_norm": 1.0781476497650146, "learning_rate": 0.0002, "epoch": 3.4645541993371594, "step": 21430}, {"loss": 0.6303, "grad_norm": 0.9571592807769775, "learning_rate": 0.0002, "epoch": 3.4661708835179046, "step": 21440}, {"loss": 0.6309, "grad_norm": 0.881547212600708, "learning_rate": 0.0002, "epoch": 3.4677875676986503, "step": 21450}, {"loss": 0.6076, "grad_norm": 0.6955338716506958, "learning_rate": 0.0002, "epoch": 3.4694042518793955, "step": 21460}, {"loss": 0.6205, "grad_norm": 0.901187539100647, "learning_rate": 0.0002, "epoch": 3.4710209360601407, "step": 21470}, {"loss": 0.639, "grad_norm": 0.7063511610031128, "learning_rate": 0.0002, "epoch": 3.472637620240886, "step": 21480}, {"loss": 0.6154, "grad_norm": 0.8462792038917542, "learning_rate": 0.0002, "epoch": 3.474254304421631, "step": 21490}, {"loss": 0.61, "grad_norm": 1.1861060857772827, "learning_rate": 0.0002, "epoch": 3.4758709886023764, "step": 21500}, {"loss": 0.6586, "grad_norm": 0.70503169298172, "learning_rate": 0.0002, "epoch": 3.4774876727831217, "step": 21510}, {"loss": 0.6475, "grad_norm": 0.9650066494941711, "learning_rate": 0.0002, "epoch": 3.479104356963867, "step": 21520}, {"loss": 0.6452, "grad_norm": 1.0266852378845215, "learning_rate": 0.0002, "epoch": 3.4807210411446126, "step": 21530}, {"loss": 0.6553, "grad_norm": 0.956372857093811, "learning_rate": 0.0002, "epoch": 3.482337725325358, "step": 21540}, {"loss": 0.6667, "grad_norm": 0.8848432898521423, "learning_rate": 0.0002, "epoch": 3.483954409506103, "step": 21550}, {"loss": 0.6375, "grad_norm": 1.0805351734161377, "learning_rate": 0.0002, "epoch": 3.4855710936868483, "step": 21560}, {"loss": 0.6958, "grad_norm": 0.9279725551605225, "learning_rate": 0.0002, "epoch": 3.4871877778675935, "step": 21570}, {"loss": 0.6354, "grad_norm": 0.9049562215805054, "learning_rate": 0.0002, "epoch": 3.4888044620483387, "step": 21580}, {"loss": 0.6071, "grad_norm": 0.9619429111480713, "learning_rate": 0.0002, "epoch": 3.4904211462290844, "step": 21590}, {"loss": 0.5927, "grad_norm": 0.8508906960487366, "learning_rate": 0.0002, "epoch": 3.4920378304098296, "step": 21600}, {"loss": 0.6115, "grad_norm": 0.8692502379417419, "learning_rate": 0.0002, "epoch": 3.493654514590575, "step": 21610}, {"loss": 0.5878, "grad_norm": 0.8187332153320312, "learning_rate": 0.0002, "epoch": 3.49527119877132, "step": 21620}, {"loss": 0.5874, "grad_norm": 1.145400047302246, "learning_rate": 0.0002, "epoch": 3.4968878829520653, "step": 21630}, {"loss": 0.6313, "grad_norm": 0.8281388282775879, "learning_rate": 0.0002, "epoch": 3.4985045671328105, "step": 21640}, {"loss": 0.6624, "grad_norm": 0.82256019115448, "learning_rate": 0.0002, "epoch": 3.500121251313556, "step": 21650}, {"loss": 0.6346, "grad_norm": 0.9315484762191772, "learning_rate": 0.0002, "epoch": 3.501737935494301, "step": 21660}, {"loss": 0.6086, "grad_norm": 0.7626111507415771, "learning_rate": 0.0002, "epoch": 3.5033546196750462, "step": 21670}, {"loss": 0.6177, "grad_norm": 0.9275059103965759, "learning_rate": 0.0002, "epoch": 3.504971303855792, "step": 21680}, {"loss": 0.64, "grad_norm": 0.7906724810600281, "learning_rate": 0.0002, "epoch": 3.506587988036537, "step": 21690}, {"loss": 0.6015, "grad_norm": 0.8289761543273926, "learning_rate": 0.0002, "epoch": 3.5082046722172824, "step": 21700}, {"loss": 0.6246, "grad_norm": 0.8316431045532227, "learning_rate": 0.0002, "epoch": 3.5098213563980276, "step": 21710}, {"loss": 0.619, "grad_norm": 1.0451812744140625, "learning_rate": 0.0002, "epoch": 3.511438040578773, "step": 21720}, {"loss": 0.632, "grad_norm": 0.928252637386322, "learning_rate": 0.0002, "epoch": 3.513054724759518, "step": 21730}, {"loss": 0.6062, "grad_norm": 0.7985895276069641, "learning_rate": 0.0002, "epoch": 3.5146714089402638, "step": 21740}, {"loss": 0.6463, "grad_norm": 0.6740974187850952, "learning_rate": 0.0002, "epoch": 3.516288093121009, "step": 21750}, {"loss": 0.6138, "grad_norm": 0.8482223749160767, "learning_rate": 0.0002, "epoch": 3.517904777301754, "step": 21760}, {"loss": 0.6277, "grad_norm": 0.889947772026062, "learning_rate": 0.0002, "epoch": 3.5195214614824994, "step": 21770}, {"loss": 0.6174, "grad_norm": 0.8304598927497864, "learning_rate": 0.0002, "epoch": 3.5211381456632447, "step": 21780}, {"loss": 0.6156, "grad_norm": 0.8002981543540955, "learning_rate": 0.0002, "epoch": 3.52275482984399, "step": 21790}, {"loss": 0.5896, "grad_norm": 0.8115083575248718, "learning_rate": 0.0002, "epoch": 3.524371514024735, "step": 21800}, {"loss": 0.6041, "grad_norm": 0.9715048670768738, "learning_rate": 0.0002, "epoch": 3.5259881982054804, "step": 21810}, {"loss": 0.6715, "grad_norm": 1.0910786390304565, "learning_rate": 0.0002, "epoch": 3.5276048823862256, "step": 21820}, {"loss": 0.6543, "grad_norm": 0.8438942432403564, "learning_rate": 0.0002, "epoch": 3.5292215665669713, "step": 21830}, {"loss": 0.6509, "grad_norm": 0.8813382983207703, "learning_rate": 0.0002, "epoch": 3.5308382507477165, "step": 21840}, {"loss": 0.6049, "grad_norm": 0.7092908024787903, "learning_rate": 0.0002, "epoch": 3.5324549349284617, "step": 21850}, {"loss": 0.5678, "grad_norm": 0.8332187533378601, "learning_rate": 0.0002, "epoch": 3.534071619109207, "step": 21860}, {"loss": 0.5896, "grad_norm": 0.8958209156990051, "learning_rate": 0.0002, "epoch": 3.535688303289952, "step": 21870}, {"loss": 0.6476, "grad_norm": 0.824138879776001, "learning_rate": 0.0002, "epoch": 3.5373049874706974, "step": 21880}, {"loss": 0.6022, "grad_norm": 0.8375158309936523, "learning_rate": 0.0002, "epoch": 3.538921671651443, "step": 21890}, {"loss": 0.6019, "grad_norm": 1.0274608135223389, "learning_rate": 0.0002, "epoch": 3.5405383558321883, "step": 21900}, {"loss": 0.6194, "grad_norm": 0.7088932394981384, "learning_rate": 0.0002, "epoch": 3.5421550400129336, "step": 21910}, {"loss": 0.6554, "grad_norm": 0.8172445297241211, "learning_rate": 0.0002, "epoch": 3.543771724193679, "step": 21920}, {"loss": 0.6711, "grad_norm": 0.9904135465621948, "learning_rate": 0.0002, "epoch": 3.545388408374424, "step": 21930}, {"loss": 0.6001, "grad_norm": 0.9900432229042053, "learning_rate": 0.0002, "epoch": 3.5470050925551693, "step": 21940}, {"loss": 0.6195, "grad_norm": 0.8963301181793213, "learning_rate": 0.0002, "epoch": 3.5486217767359145, "step": 21950}, {"loss": 0.5972, "grad_norm": 0.8551464676856995, "learning_rate": 0.0002, "epoch": 3.5502384609166597, "step": 21960}, {"loss": 0.6206, "grad_norm": 1.0916603803634644, "learning_rate": 0.0002, "epoch": 3.551855145097405, "step": 21970}, {"loss": 0.6523, "grad_norm": 0.841598391532898, "learning_rate": 0.0002, "epoch": 3.5534718292781506, "step": 21980}, {"loss": 0.617, "grad_norm": 0.8566757440567017, "learning_rate": 0.0002, "epoch": 3.555088513458896, "step": 21990}, {"loss": 0.6192, "grad_norm": 1.0145052671432495, "learning_rate": 0.0002, "epoch": 3.556705197639641, "step": 22000}, {"loss": 0.6173, "grad_norm": 0.9293754696846008, "learning_rate": 0.0002, "epoch": 3.5583218818203863, "step": 22010}, {"loss": 0.612, "grad_norm": 0.9568536281585693, "learning_rate": 0.0002, "epoch": 3.5599385660011316, "step": 22020}, {"loss": 0.641, "grad_norm": 0.8613139986991882, "learning_rate": 0.0002, "epoch": 3.5615552501818772, "step": 22030}, {"loss": 0.6496, "grad_norm": 0.8179237246513367, "learning_rate": 0.0002, "epoch": 3.5631719343626225, "step": 22040}, {"loss": 0.574, "grad_norm": 0.9059830904006958, "learning_rate": 0.0002, "epoch": 3.5647886185433677, "step": 22050}, {"loss": 0.6448, "grad_norm": 1.0068252086639404, "learning_rate": 0.0002, "epoch": 3.566405302724113, "step": 22060}, {"loss": 0.6239, "grad_norm": 0.9682072997093201, "learning_rate": 0.0002, "epoch": 3.568021986904858, "step": 22070}, {"loss": 0.6808, "grad_norm": 0.8514005541801453, "learning_rate": 0.0002, "epoch": 3.5696386710856034, "step": 22080}, {"loss": 0.5956, "grad_norm": 0.8327770829200745, "learning_rate": 0.0002, "epoch": 3.5712553552663486, "step": 22090}, {"loss": 0.5976, "grad_norm": 1.024976372718811, "learning_rate": 0.0002, "epoch": 3.572872039447094, "step": 22100}, {"loss": 0.624, "grad_norm": 0.7721174955368042, "learning_rate": 0.0002, "epoch": 3.574488723627839, "step": 22110}, {"loss": 0.5896, "grad_norm": 1.0351054668426514, "learning_rate": 0.0002, "epoch": 3.5761054078085843, "step": 22120}, {"loss": 0.6379, "grad_norm": 0.9680907130241394, "learning_rate": 0.0002, "epoch": 3.57772209198933, "step": 22130}, {"loss": 0.6194, "grad_norm": 0.8016974925994873, "learning_rate": 0.0002, "epoch": 3.5793387761700752, "step": 22140}, {"loss": 0.6387, "grad_norm": 1.0109003782272339, "learning_rate": 0.0002, "epoch": 3.5809554603508205, "step": 22150}, {"loss": 0.6368, "grad_norm": 1.0473392009735107, "learning_rate": 0.0002, "epoch": 3.5825721445315657, "step": 22160}, {"loss": 0.6353, "grad_norm": 0.8686613440513611, "learning_rate": 0.0002, "epoch": 3.584188828712311, "step": 22170}, {"loss": 0.5791, "grad_norm": 0.869149923324585, "learning_rate": 0.0002, "epoch": 3.5858055128930566, "step": 22180}, {"loss": 0.5895, "grad_norm": 0.9769062995910645, "learning_rate": 0.0002, "epoch": 3.587422197073802, "step": 22190}, {"loss": 0.5939, "grad_norm": 0.779636561870575, "learning_rate": 0.0002, "epoch": 3.589038881254547, "step": 22200}, {"loss": 0.5875, "grad_norm": 0.9063841104507446, "learning_rate": 0.0002, "epoch": 3.5906555654352923, "step": 22210}, {"loss": 0.5671, "grad_norm": 0.9216037392616272, "learning_rate": 0.0002, "epoch": 3.5922722496160375, "step": 22220}, {"loss": 0.6484, "grad_norm": 1.0217336416244507, "learning_rate": 0.0002, "epoch": 3.5938889337967828, "step": 22230}, {"loss": 0.6511, "grad_norm": 0.8513161540031433, "learning_rate": 0.0002, "epoch": 3.595505617977528, "step": 22240}, {"loss": 0.6301, "grad_norm": 0.8084813952445984, "learning_rate": 0.0002, "epoch": 3.597122302158273, "step": 22250}, {"loss": 0.6197, "grad_norm": 0.8524802923202515, "learning_rate": 0.0002, "epoch": 3.5987389863390185, "step": 22260}, {"loss": 0.5599, "grad_norm": 0.9356237649917603, "learning_rate": 0.0002, "epoch": 3.600355670519764, "step": 22270}, {"loss": 0.628, "grad_norm": 1.009600281715393, "learning_rate": 0.0002, "epoch": 3.6019723547005094, "step": 22280}, {"loss": 0.6179, "grad_norm": 0.9900581240653992, "learning_rate": 0.0002, "epoch": 3.6035890388812546, "step": 22290}, {"loss": 0.5725, "grad_norm": 1.062495231628418, "learning_rate": 0.0002, "epoch": 3.605205723062, "step": 22300}, {"loss": 0.607, "grad_norm": 0.8832381367683411, "learning_rate": 0.0002, "epoch": 3.606822407242745, "step": 22310}, {"loss": 0.6215, "grad_norm": 0.9284297823905945, "learning_rate": 0.0002, "epoch": 3.6084390914234903, "step": 22320}, {"loss": 0.685, "grad_norm": 1.2381829023361206, "learning_rate": 0.0002, "epoch": 3.610055775604236, "step": 22330}, {"loss": 0.6181, "grad_norm": 0.929434597492218, "learning_rate": 0.0002, "epoch": 3.611672459784981, "step": 22340}, {"loss": 0.6141, "grad_norm": 0.9714490175247192, "learning_rate": 0.0002, "epoch": 3.6132891439657264, "step": 22350}, {"loss": 0.6861, "grad_norm": 0.808014988899231, "learning_rate": 0.0002, "epoch": 3.6149058281464717, "step": 22360}, {"loss": 0.6428, "grad_norm": 1.0364398956298828, "learning_rate": 0.0002, "epoch": 3.616522512327217, "step": 22370}, {"loss": 0.6337, "grad_norm": 0.7858489751815796, "learning_rate": 0.0002, "epoch": 3.618139196507962, "step": 22380}, {"loss": 0.6214, "grad_norm": 0.9920870065689087, "learning_rate": 0.0002, "epoch": 3.6197558806887074, "step": 22390}, {"loss": 0.6659, "grad_norm": 0.9183220863342285, "learning_rate": 0.0002, "epoch": 3.6213725648694526, "step": 22400}, {"loss": 0.6036, "grad_norm": 0.9826246500015259, "learning_rate": 0.0002, "epoch": 3.622989249050198, "step": 22410}, {"loss": 0.6441, "grad_norm": 0.8632931113243103, "learning_rate": 0.0002, "epoch": 3.6246059332309435, "step": 22420}, {"loss": 0.6124, "grad_norm": 0.8468965291976929, "learning_rate": 0.0002, "epoch": 3.6262226174116887, "step": 22430}, {"loss": 0.6328, "grad_norm": 0.8466871976852417, "learning_rate": 0.0002, "epoch": 3.627839301592434, "step": 22440}, {"loss": 0.5941, "grad_norm": 0.9501169919967651, "learning_rate": 0.0002, "epoch": 3.629455985773179, "step": 22450}, {"loss": 0.6069, "grad_norm": 0.8906720876693726, "learning_rate": 0.0002, "epoch": 3.6310726699539244, "step": 22460}, {"loss": 0.6928, "grad_norm": 0.7400227189064026, "learning_rate": 0.0002, "epoch": 3.6326893541346696, "step": 22470}, {"loss": 0.6337, "grad_norm": 0.9756355881690979, "learning_rate": 0.0002, "epoch": 3.6343060383154153, "step": 22480}, {"loss": 0.6203, "grad_norm": 0.7504993081092834, "learning_rate": 0.0002, "epoch": 3.6359227224961606, "step": 22490}, {"loss": 0.6302, "grad_norm": 0.9270039200782776, "learning_rate": 0.0002, "epoch": 3.637539406676906, "step": 22500}, {"loss": 0.6026, "grad_norm": 0.8841686844825745, "learning_rate": 0.0002, "epoch": 3.639156090857651, "step": 22510}, {"loss": 0.6098, "grad_norm": 0.8533213138580322, "learning_rate": 0.0002, "epoch": 3.6407727750383962, "step": 22520}, {"loss": 0.6412, "grad_norm": 1.0052043199539185, "learning_rate": 0.0002, "epoch": 3.6423894592191415, "step": 22530}, {"loss": 0.6363, "grad_norm": 1.0323461294174194, "learning_rate": 0.0002, "epoch": 3.6440061433998867, "step": 22540}, {"loss": 0.6545, "grad_norm": 0.8654312491416931, "learning_rate": 0.0002, "epoch": 3.645622827580632, "step": 22550}, {"loss": 0.6155, "grad_norm": 0.6400038003921509, "learning_rate": 0.0002, "epoch": 3.647239511761377, "step": 22560}, {"loss": 0.5829, "grad_norm": 0.8061298727989197, "learning_rate": 0.0002, "epoch": 3.648856195942123, "step": 22570}, {"loss": 0.6388, "grad_norm": 0.9257854223251343, "learning_rate": 0.0002, "epoch": 3.650472880122868, "step": 22580}, {"loss": 0.6409, "grad_norm": 0.8439396619796753, "learning_rate": 0.0002, "epoch": 3.6520895643036133, "step": 22590}, {"loss": 0.5996, "grad_norm": 0.7764544486999512, "learning_rate": 0.0002, "epoch": 3.6537062484843585, "step": 22600}, {"loss": 0.6434, "grad_norm": 1.125451683998108, "learning_rate": 0.0002, "epoch": 3.6553229326651038, "step": 22610}, {"loss": 0.6579, "grad_norm": 0.7523018717765808, "learning_rate": 0.0002, "epoch": 3.656939616845849, "step": 22620}, {"loss": 0.6476, "grad_norm": 1.071026086807251, "learning_rate": 0.0002, "epoch": 3.6585563010265947, "step": 22630}, {"loss": 0.6459, "grad_norm": 0.945791482925415, "learning_rate": 0.0002, "epoch": 3.66017298520734, "step": 22640}, {"loss": 0.659, "grad_norm": 0.8001811504364014, "learning_rate": 0.0002, "epoch": 3.661789669388085, "step": 22650}, {"loss": 0.6385, "grad_norm": 0.9700816869735718, "learning_rate": 0.0002, "epoch": 3.6634063535688304, "step": 22660}, {"loss": 0.6337, "grad_norm": 0.9053242206573486, "learning_rate": 0.0002, "epoch": 3.6650230377495756, "step": 22670}, {"loss": 0.6335, "grad_norm": 0.944362461566925, "learning_rate": 0.0002, "epoch": 3.666639721930321, "step": 22680}, {"loss": 0.6235, "grad_norm": 1.067489504814148, "learning_rate": 0.0002, "epoch": 3.668256406111066, "step": 22690}, {"loss": 0.698, "grad_norm": 1.0984995365142822, "learning_rate": 0.0002, "epoch": 3.6698730902918113, "step": 22700}, {"loss": 0.6717, "grad_norm": 0.9336317777633667, "learning_rate": 0.0002, "epoch": 3.6714897744725565, "step": 22710}, {"loss": 0.6195, "grad_norm": 0.9261918663978577, "learning_rate": 0.0002, "epoch": 3.673106458653302, "step": 22720}, {"loss": 0.6332, "grad_norm": 0.8648008704185486, "learning_rate": 0.0002, "epoch": 3.6747231428340474, "step": 22730}, {"loss": 0.6576, "grad_norm": 0.7225083708763123, "learning_rate": 0.0002, "epoch": 3.6763398270147927, "step": 22740}, {"loss": 0.6406, "grad_norm": 0.9258282780647278, "learning_rate": 0.0002, "epoch": 3.677956511195538, "step": 22750}, {"loss": 0.6397, "grad_norm": 0.70876145362854, "learning_rate": 0.0002, "epoch": 3.679573195376283, "step": 22760}, {"loss": 0.6821, "grad_norm": 0.8780210018157959, "learning_rate": 0.0002, "epoch": 3.681189879557029, "step": 22770}, {"loss": 0.6036, "grad_norm": 0.8075440526008606, "learning_rate": 0.0002, "epoch": 3.682806563737774, "step": 22780}, {"loss": 0.6561, "grad_norm": 0.8503130674362183, "learning_rate": 0.0002, "epoch": 3.6844232479185193, "step": 22790}, {"loss": 0.6082, "grad_norm": 0.8413618206977844, "learning_rate": 0.0002, "epoch": 3.6860399320992645, "step": 22800}, {"loss": 0.614, "grad_norm": 0.8675165176391602, "learning_rate": 0.0002, "epoch": 3.6876566162800097, "step": 22810}, {"loss": 0.6157, "grad_norm": 0.8235884308815002, "learning_rate": 0.0002, "epoch": 3.689273300460755, "step": 22820}, {"loss": 0.5708, "grad_norm": 0.9477725625038147, "learning_rate": 0.0002, "epoch": 3.6908899846415, "step": 22830}, {"loss": 0.6481, "grad_norm": 0.7883533835411072, "learning_rate": 0.0002, "epoch": 3.6925066688222454, "step": 22840}, {"loss": 0.5872, "grad_norm": 1.047913908958435, "learning_rate": 0.0002, "epoch": 3.6941233530029907, "step": 22850}, {"loss": 0.6176, "grad_norm": 0.9171528816223145, "learning_rate": 0.0002, "epoch": 3.695740037183736, "step": 22860}, {"loss": 0.6204, "grad_norm": 0.9338192343711853, "learning_rate": 0.0002, "epoch": 3.6973567213644816, "step": 22870}, {"loss": 0.686, "grad_norm": 0.8799443244934082, "learning_rate": 0.0002, "epoch": 3.698973405545227, "step": 22880}, {"loss": 0.6206, "grad_norm": 0.8515434861183167, "learning_rate": 0.0002, "epoch": 3.700590089725972, "step": 22890}, {"loss": 0.5954, "grad_norm": 0.7805591821670532, "learning_rate": 0.0002, "epoch": 3.7022067739067173, "step": 22900}, {"loss": 0.6108, "grad_norm": 0.8470911979675293, "learning_rate": 0.0002, "epoch": 3.7038234580874625, "step": 22910}, {"loss": 0.6557, "grad_norm": 0.9452309012413025, "learning_rate": 0.0002, "epoch": 3.705440142268208, "step": 22920}, {"loss": 0.6529, "grad_norm": 0.950243353843689, "learning_rate": 0.0002, "epoch": 3.7070568264489534, "step": 22930}, {"loss": 0.6364, "grad_norm": 0.7882499098777771, "learning_rate": 0.0002, "epoch": 3.7086735106296986, "step": 22940}, {"loss": 0.6462, "grad_norm": 0.8307787775993347, "learning_rate": 0.0002, "epoch": 3.710290194810444, "step": 22950}, {"loss": 0.6371, "grad_norm": 1.0970630645751953, "learning_rate": 0.0002, "epoch": 3.711906878991189, "step": 22960}, {"loss": 0.6281, "grad_norm": 0.8269566297531128, "learning_rate": 0.0002, "epoch": 3.7135235631719343, "step": 22970}, {"loss": 0.6561, "grad_norm": 0.8306704759597778, "learning_rate": 0.0002, "epoch": 3.7151402473526796, "step": 22980}, {"loss": 0.6418, "grad_norm": 0.9710225462913513, "learning_rate": 0.0002, "epoch": 3.716756931533425, "step": 22990}, {"loss": 0.6639, "grad_norm": 0.8890530467033386, "learning_rate": 0.0002, "epoch": 3.71837361571417, "step": 23000}, {"loss": 0.6084, "grad_norm": 0.883522629737854, "learning_rate": 0.0002, "epoch": 3.7199902998949153, "step": 23010}, {"loss": 0.6183, "grad_norm": 0.8662652373313904, "learning_rate": 0.0002, "epoch": 3.721606984075661, "step": 23020}, {"loss": 0.6266, "grad_norm": 0.7228406667709351, "learning_rate": 0.0002, "epoch": 3.723223668256406, "step": 23030}, {"loss": 0.6417, "grad_norm": 1.060792088508606, "learning_rate": 0.0002, "epoch": 3.7248403524371514, "step": 23040}, {"loss": 0.6346, "grad_norm": 1.0119613409042358, "learning_rate": 0.0002, "epoch": 3.7264570366178966, "step": 23050}, {"loss": 0.6466, "grad_norm": 0.9212996959686279, "learning_rate": 0.0002, "epoch": 3.728073720798642, "step": 23060}, {"loss": 0.6454, "grad_norm": 0.925690233707428, "learning_rate": 0.0002, "epoch": 3.7296904049793875, "step": 23070}, {"loss": 0.615, "grad_norm": 0.8323310613632202, "learning_rate": 0.0002, "epoch": 3.7313070891601328, "step": 23080}, {"loss": 0.679, "grad_norm": 0.8966048955917358, "learning_rate": 0.0002, "epoch": 3.732923773340878, "step": 23090}, {"loss": 0.6151, "grad_norm": 0.8995837569236755, "learning_rate": 0.0002, "epoch": 3.7345404575216232, "step": 23100}, {"loss": 0.6143, "grad_norm": 0.8748890161514282, "learning_rate": 0.0002, "epoch": 3.7361571417023685, "step": 23110}, {"loss": 0.6246, "grad_norm": 0.7985540628433228, "learning_rate": 0.0002, "epoch": 3.7377738258831137, "step": 23120}, {"loss": 0.6279, "grad_norm": 1.0240917205810547, "learning_rate": 0.0002, "epoch": 3.739390510063859, "step": 23130}, {"loss": 0.6747, "grad_norm": 0.9181789755821228, "learning_rate": 0.0002, "epoch": 3.741007194244604, "step": 23140}, {"loss": 0.6026, "grad_norm": 0.8896583914756775, "learning_rate": 0.0002, "epoch": 3.7426238784253494, "step": 23150}, {"loss": 0.5972, "grad_norm": 0.8635515570640564, "learning_rate": 0.0002, "epoch": 3.744240562606095, "step": 23160}, {"loss": 0.6683, "grad_norm": 0.8873575329780579, "learning_rate": 0.0002, "epoch": 3.7458572467868403, "step": 23170}, {"loss": 0.6143, "grad_norm": 0.9807148575782776, "learning_rate": 0.0002, "epoch": 3.7474739309675855, "step": 23180}, {"loss": 0.6381, "grad_norm": 0.900477945804596, "learning_rate": 0.0002, "epoch": 3.7490906151483308, "step": 23190}, {"loss": 0.6542, "grad_norm": 0.9379992485046387, "learning_rate": 0.0002, "epoch": 3.750707299329076, "step": 23200}, {"loss": 0.6015, "grad_norm": 0.9649890661239624, "learning_rate": 0.0002, "epoch": 3.752323983509821, "step": 23210}, {"loss": 0.6735, "grad_norm": 0.824442446231842, "learning_rate": 0.0002, "epoch": 3.753940667690567, "step": 23220}, {"loss": 0.5992, "grad_norm": 0.8896150588989258, "learning_rate": 0.0002, "epoch": 3.755557351871312, "step": 23230}, {"loss": 0.6081, "grad_norm": 0.751249372959137, "learning_rate": 0.0002, "epoch": 3.7571740360520574, "step": 23240}, {"loss": 0.629, "grad_norm": 0.9392193555831909, "learning_rate": 0.0002, "epoch": 3.7587907202328026, "step": 23250}, {"loss": 0.6209, "grad_norm": 0.9284586310386658, "learning_rate": 0.0002, "epoch": 3.760407404413548, "step": 23260}, {"loss": 0.6414, "grad_norm": 0.7738175392150879, "learning_rate": 0.0002, "epoch": 3.762024088594293, "step": 23270}, {"loss": 0.6743, "grad_norm": 0.9252978563308716, "learning_rate": 0.0002, "epoch": 3.7636407727750383, "step": 23280}, {"loss": 0.5984, "grad_norm": 0.9501895904541016, "learning_rate": 0.0002, "epoch": 3.7652574569557835, "step": 23290}, {"loss": 0.6568, "grad_norm": 0.9416276216506958, "learning_rate": 0.0002, "epoch": 3.7668741411365287, "step": 23300}, {"loss": 0.6507, "grad_norm": 0.7076631784439087, "learning_rate": 0.0002, "epoch": 3.7684908253172744, "step": 23310}, {"loss": 0.6329, "grad_norm": 0.9864492416381836, "learning_rate": 0.0002, "epoch": 3.7701075094980196, "step": 23320}, {"loss": 0.6537, "grad_norm": 0.8450456261634827, "learning_rate": 0.0002, "epoch": 3.771724193678765, "step": 23330}, {"loss": 0.658, "grad_norm": 1.0768941640853882, "learning_rate": 0.0002, "epoch": 3.77334087785951, "step": 23340}, {"loss": 0.6408, "grad_norm": 0.9956819415092468, "learning_rate": 0.0002, "epoch": 3.7749575620402553, "step": 23350}, {"loss": 0.6464, "grad_norm": 0.9234658479690552, "learning_rate": 0.0002, "epoch": 3.7765742462210006, "step": 23360}, {"loss": 0.6542, "grad_norm": 1.0993858575820923, "learning_rate": 0.0002, "epoch": 3.7781909304017463, "step": 23370}, {"loss": 0.6391, "grad_norm": 0.923159658908844, "learning_rate": 0.0002, "epoch": 3.7798076145824915, "step": 23380}, {"loss": 0.6625, "grad_norm": 0.9311541318893433, "learning_rate": 0.0002, "epoch": 3.7814242987632367, "step": 23390}, {"loss": 0.6535, "grad_norm": 0.919681191444397, "learning_rate": 0.0002, "epoch": 3.783040982943982, "step": 23400}, {"loss": 0.6138, "grad_norm": 1.7406195402145386, "learning_rate": 0.0002, "epoch": 3.784657667124727, "step": 23410}, {"loss": 0.657, "grad_norm": 0.7789074182510376, "learning_rate": 0.0002, "epoch": 3.7862743513054724, "step": 23420}, {"loss": 0.658, "grad_norm": 0.8302814960479736, "learning_rate": 0.0002, "epoch": 3.7878910354862176, "step": 23430}, {"loss": 0.649, "grad_norm": 0.8089349269866943, "learning_rate": 0.0002, "epoch": 3.789507719666963, "step": 23440}, {"loss": 0.6682, "grad_norm": 0.9006284475326538, "learning_rate": 0.0002, "epoch": 3.791124403847708, "step": 23450}, {"loss": 0.6335, "grad_norm": 0.8426766991615295, "learning_rate": 0.0002, "epoch": 3.7927410880284538, "step": 23460}, {"loss": 0.6364, "grad_norm": 1.2576252222061157, "learning_rate": 0.0002, "epoch": 3.794357772209199, "step": 23470}, {"loss": 0.6324, "grad_norm": 1.0307610034942627, "learning_rate": 0.0002, "epoch": 3.7959744563899442, "step": 23480}, {"loss": 0.6262, "grad_norm": 0.8525972962379456, "learning_rate": 0.0002, "epoch": 3.7975911405706895, "step": 23490}, {"loss": 0.6757, "grad_norm": 1.159039855003357, "learning_rate": 0.0002, "epoch": 3.7992078247514347, "step": 23500}, {"loss": 0.6414, "grad_norm": 1.4193549156188965, "learning_rate": 0.0002, "epoch": 3.80082450893218, "step": 23510}, {"loss": 0.6413, "grad_norm": 0.8245543837547302, "learning_rate": 0.0002, "epoch": 3.8024411931129256, "step": 23520}, {"loss": 0.6417, "grad_norm": 0.8847230076789856, "learning_rate": 0.0002, "epoch": 3.804057877293671, "step": 23530}, {"loss": 0.6415, "grad_norm": 0.9574624300003052, "learning_rate": 0.0002, "epoch": 3.805674561474416, "step": 23540}, {"loss": 0.5765, "grad_norm": 1.048020601272583, "learning_rate": 0.0002, "epoch": 3.8072912456551613, "step": 23550}, {"loss": 0.6497, "grad_norm": 0.8302255868911743, "learning_rate": 0.0002, "epoch": 3.8089079298359065, "step": 23560}, {"loss": 0.6534, "grad_norm": 0.8269215822219849, "learning_rate": 0.0002, "epoch": 3.8105246140166518, "step": 23570}, {"loss": 0.6294, "grad_norm": 0.9375753402709961, "learning_rate": 0.0002, "epoch": 3.812141298197397, "step": 23580}, {"loss": 0.6132, "grad_norm": 1.0234097242355347, "learning_rate": 0.0002, "epoch": 3.8137579823781422, "step": 23590}, {"loss": 0.6625, "grad_norm": 0.8978445529937744, "learning_rate": 0.0002, "epoch": 3.8153746665588875, "step": 23600}, {"loss": 0.6315, "grad_norm": 0.7929515838623047, "learning_rate": 0.0002, "epoch": 3.816991350739633, "step": 23610}, {"loss": 0.6387, "grad_norm": 1.3255881071090698, "learning_rate": 0.0002, "epoch": 3.8186080349203784, "step": 23620}, {"loss": 0.5947, "grad_norm": 0.9188598990440369, "learning_rate": 0.0002, "epoch": 3.8202247191011236, "step": 23630}, {"loss": 0.6152, "grad_norm": 0.8811675906181335, "learning_rate": 0.0002, "epoch": 3.821841403281869, "step": 23640}, {"loss": 0.6253, "grad_norm": 0.8061038255691528, "learning_rate": 0.0002, "epoch": 3.823458087462614, "step": 23650}, {"loss": 0.6517, "grad_norm": 0.9975376129150391, "learning_rate": 0.0002, "epoch": 3.8250747716433597, "step": 23660}, {"loss": 0.6288, "grad_norm": 0.8036105036735535, "learning_rate": 0.0002, "epoch": 3.826691455824105, "step": 23670}, {"loss": 0.6845, "grad_norm": 0.7401984333992004, "learning_rate": 0.0002, "epoch": 3.82830814000485, "step": 23680}, {"loss": 0.6423, "grad_norm": 0.829753041267395, "learning_rate": 0.0002, "epoch": 3.8299248241855954, "step": 23690}, {"loss": 0.6611, "grad_norm": 0.8753240704536438, "learning_rate": 0.0002, "epoch": 3.8315415083663407, "step": 23700}, {"loss": 0.6686, "grad_norm": 0.8157842755317688, "learning_rate": 0.0002, "epoch": 3.833158192547086, "step": 23710}, {"loss": 0.6181, "grad_norm": 0.6183798909187317, "learning_rate": 0.0002, "epoch": 3.834774876727831, "step": 23720}, {"loss": 0.5965, "grad_norm": 0.9548442363739014, "learning_rate": 0.0002, "epoch": 3.8363915609085764, "step": 23730}, {"loss": 0.6456, "grad_norm": 0.8319669961929321, "learning_rate": 0.0002, "epoch": 3.8380082450893216, "step": 23740}, {"loss": 0.6585, "grad_norm": 0.9718693494796753, "learning_rate": 0.0002, "epoch": 3.839624929270067, "step": 23750}, {"loss": 0.6518, "grad_norm": 0.8672235012054443, "learning_rate": 0.0002, "epoch": 3.8412416134508125, "step": 23760}, {"loss": 0.6774, "grad_norm": 1.1210707426071167, "learning_rate": 0.0002, "epoch": 3.8428582976315577, "step": 23770}, {"loss": 0.5923, "grad_norm": 0.9177767634391785, "learning_rate": 0.0002, "epoch": 3.844474981812303, "step": 23780}, {"loss": 0.6286, "grad_norm": 0.8714171648025513, "learning_rate": 0.0002, "epoch": 3.846091665993048, "step": 23790}, {"loss": 0.6302, "grad_norm": 1.1853246688842773, "learning_rate": 0.0002, "epoch": 3.8477083501737934, "step": 23800}, {"loss": 0.6144, "grad_norm": 0.8091260194778442, "learning_rate": 0.0002, "epoch": 3.849325034354539, "step": 23810}, {"loss": 0.658, "grad_norm": 0.9710774421691895, "learning_rate": 0.0002, "epoch": 3.8509417185352843, "step": 23820}, {"loss": 0.6151, "grad_norm": 0.7648707628250122, "learning_rate": 0.0002, "epoch": 3.8525584027160296, "step": 23830}, {"loss": 0.6013, "grad_norm": 0.7809253931045532, "learning_rate": 0.0002, "epoch": 3.854175086896775, "step": 23840}, {"loss": 0.6006, "grad_norm": 0.8337951898574829, "learning_rate": 0.0002, "epoch": 3.85579177107752, "step": 23850}, {"loss": 0.6456, "grad_norm": 0.9271913170814514, "learning_rate": 0.0002, "epoch": 3.8574084552582653, "step": 23860}, {"loss": 0.6671, "grad_norm": 0.985334038734436, "learning_rate": 0.0002, "epoch": 3.8590251394390105, "step": 23870}, {"loss": 0.6693, "grad_norm": 0.8458583354949951, "learning_rate": 0.0002, "epoch": 3.8606418236197557, "step": 23880}, {"loss": 0.6207, "grad_norm": 1.015348196029663, "learning_rate": 0.0002, "epoch": 3.862258507800501, "step": 23890}, {"loss": 0.649, "grad_norm": 1.0121688842773438, "learning_rate": 0.0002, "epoch": 3.8638751919812466, "step": 23900}, {"loss": 0.5921, "grad_norm": 0.8883971571922302, "learning_rate": 0.0002, "epoch": 3.865491876161992, "step": 23910}, {"loss": 0.6597, "grad_norm": 1.028086543083191, "learning_rate": 0.0002, "epoch": 3.867108560342737, "step": 23920}, {"loss": 0.6654, "grad_norm": 0.9645734429359436, "learning_rate": 0.0002, "epoch": 3.8687252445234823, "step": 23930}, {"loss": 0.6328, "grad_norm": 0.8235350251197815, "learning_rate": 0.0002, "epoch": 3.8703419287042276, "step": 23940}, {"loss": 0.6387, "grad_norm": 1.0298916101455688, "learning_rate": 0.0002, "epoch": 3.871958612884973, "step": 23950}, {"loss": 0.5966, "grad_norm": 1.0063377618789673, "learning_rate": 0.0002, "epoch": 3.8735752970657185, "step": 23960}, {"loss": 0.6234, "grad_norm": 0.9230626821517944, "learning_rate": 0.0002, "epoch": 3.8751919812464637, "step": 23970}, {"loss": 0.6159, "grad_norm": 0.9243063926696777, "learning_rate": 0.0002, "epoch": 3.876808665427209, "step": 23980}, {"loss": 0.6035, "grad_norm": 1.0211291313171387, "learning_rate": 0.0002, "epoch": 3.878425349607954, "step": 23990}, {"loss": 0.6351, "grad_norm": 0.7800535559654236, "learning_rate": 0.0002, "epoch": 3.8800420337886994, "step": 24000}, {"loss": 0.7, "grad_norm": 0.7904248833656311, "learning_rate": 0.0002, "epoch": 3.8816587179694446, "step": 24010}, {"loss": 0.6516, "grad_norm": 1.1975988149642944, "learning_rate": 0.0002, "epoch": 3.88327540215019, "step": 24020}, {"loss": 0.6006, "grad_norm": 1.0626593828201294, "learning_rate": 0.0002, "epoch": 3.884892086330935, "step": 24030}, {"loss": 0.6115, "grad_norm": 0.9012193083763123, "learning_rate": 0.0002, "epoch": 3.8865087705116803, "step": 24040}, {"loss": 0.6786, "grad_norm": 1.1159172058105469, "learning_rate": 0.0002, "epoch": 3.888125454692426, "step": 24050}, {"loss": 0.6635, "grad_norm": 1.276838779449463, "learning_rate": 0.0002, "epoch": 3.889742138873171, "step": 24060}, {"loss": 0.5985, "grad_norm": 0.8467690348625183, "learning_rate": 0.0002, "epoch": 3.8913588230539164, "step": 24070}, {"loss": 0.6655, "grad_norm": 0.9862841963768005, "learning_rate": 0.0002, "epoch": 3.8929755072346617, "step": 24080}, {"loss": 0.6098, "grad_norm": 0.7134621739387512, "learning_rate": 0.0002, "epoch": 3.894592191415407, "step": 24090}, {"loss": 0.618, "grad_norm": 0.8178175091743469, "learning_rate": 0.0002, "epoch": 3.896208875596152, "step": 24100}, {"loss": 0.6147, "grad_norm": 0.9229172468185425, "learning_rate": 0.0002, "epoch": 3.897825559776898, "step": 24110}, {"loss": 0.6554, "grad_norm": 1.0878316164016724, "learning_rate": 0.0002, "epoch": 3.899442243957643, "step": 24120}, {"loss": 0.6616, "grad_norm": 0.971645712852478, "learning_rate": 0.0002, "epoch": 3.9010589281383883, "step": 24130}, {"loss": 0.6228, "grad_norm": 0.8862188458442688, "learning_rate": 0.0002, "epoch": 3.9026756123191335, "step": 24140}, {"loss": 0.6192, "grad_norm": 0.9126982688903809, "learning_rate": 0.0002, "epoch": 3.9042922964998787, "step": 24150}, {"loss": 0.6734, "grad_norm": 0.8833470940589905, "learning_rate": 0.0002, "epoch": 3.905908980680624, "step": 24160}, {"loss": 0.5832, "grad_norm": 0.8320947885513306, "learning_rate": 0.0002, "epoch": 3.907525664861369, "step": 24170}, {"loss": 0.6247, "grad_norm": 0.9156602025032043, "learning_rate": 0.0002, "epoch": 3.9091423490421144, "step": 24180}, {"loss": 0.6678, "grad_norm": 1.029181957244873, "learning_rate": 0.0002, "epoch": 3.9107590332228597, "step": 24190}, {"loss": 0.6565, "grad_norm": 0.9052802324295044, "learning_rate": 0.0002, "epoch": 3.9123757174036053, "step": 24200}, {"loss": 0.6346, "grad_norm": 0.8847255110740662, "learning_rate": 0.0002, "epoch": 3.9139924015843506, "step": 24210}, {"loss": 0.6343, "grad_norm": 0.9642062187194824, "learning_rate": 0.0002, "epoch": 3.915609085765096, "step": 24220}, {"loss": 0.6557, "grad_norm": 0.8629093766212463, "learning_rate": 0.0002, "epoch": 3.917225769945841, "step": 24230}, {"loss": 0.6086, "grad_norm": 0.8674976825714111, "learning_rate": 0.0002, "epoch": 3.9188424541265863, "step": 24240}, {"loss": 0.5874, "grad_norm": 1.104846477508545, "learning_rate": 0.0002, "epoch": 3.9204591383073315, "step": 24250}, {"loss": 0.6501, "grad_norm": 1.0874955654144287, "learning_rate": 0.0002, "epoch": 3.922075822488077, "step": 24260}, {"loss": 0.6455, "grad_norm": 0.8689812421798706, "learning_rate": 0.0002, "epoch": 3.9236925066688224, "step": 24270}, {"loss": 0.5893, "grad_norm": 0.9724617004394531, "learning_rate": 0.0002, "epoch": 3.9253091908495676, "step": 24280}, {"loss": 0.6616, "grad_norm": 0.9165538549423218, "learning_rate": 0.0002, "epoch": 3.926925875030313, "step": 24290}, {"loss": 0.645, "grad_norm": 0.9307710528373718, "learning_rate": 0.0002, "epoch": 3.928542559211058, "step": 24300}, {"loss": 0.6071, "grad_norm": 0.8589295148849487, "learning_rate": 0.0002, "epoch": 3.9301592433918033, "step": 24310}, {"loss": 0.6662, "grad_norm": 0.9151099920272827, "learning_rate": 0.0002, "epoch": 3.9317759275725486, "step": 24320}, {"loss": 0.7075, "grad_norm": 0.9633517265319824, "learning_rate": 0.0002, "epoch": 3.933392611753294, "step": 24330}, {"loss": 0.6432, "grad_norm": 0.9521116018295288, "learning_rate": 0.0002, "epoch": 3.935009295934039, "step": 24340}, {"loss": 0.6457, "grad_norm": 0.8366776704788208, "learning_rate": 0.0002, "epoch": 3.9366259801147847, "step": 24350}, {"loss": 0.6139, "grad_norm": 0.8972663283348083, "learning_rate": 0.0002, "epoch": 3.93824266429553, "step": 24360}, {"loss": 0.661, "grad_norm": 0.8102919459342957, "learning_rate": 0.0002, "epoch": 3.939859348476275, "step": 24370}, {"loss": 0.6388, "grad_norm": 0.8189975023269653, "learning_rate": 0.0002, "epoch": 3.9414760326570204, "step": 24380}, {"loss": 0.6818, "grad_norm": 0.9569464921951294, "learning_rate": 0.0002, "epoch": 3.9430927168377656, "step": 24390}, {"loss": 0.6999, "grad_norm": 0.7459101676940918, "learning_rate": 0.0002, "epoch": 3.9447094010185113, "step": 24400}, {"loss": 0.6069, "grad_norm": 0.8536974787712097, "learning_rate": 0.0002, "epoch": 3.9463260851992565, "step": 24410}, {"loss": 0.5683, "grad_norm": 0.8763698935508728, "learning_rate": 0.0002, "epoch": 3.9479427693800018, "step": 24420}, {"loss": 0.6478, "grad_norm": 0.9381106495857239, "learning_rate": 0.0002, "epoch": 3.949559453560747, "step": 24430}, {"loss": 0.6371, "grad_norm": 0.934440016746521, "learning_rate": 0.0002, "epoch": 3.9511761377414922, "step": 24440}, {"loss": 0.6393, "grad_norm": 0.903918981552124, "learning_rate": 0.0002, "epoch": 3.9527928219222375, "step": 24450}, {"loss": 0.6175, "grad_norm": 0.8771953582763672, "learning_rate": 0.0002, "epoch": 3.9544095061029827, "step": 24460}, {"loss": 0.6971, "grad_norm": 1.0375410318374634, "learning_rate": 0.0002, "epoch": 3.956026190283728, "step": 24470}, {"loss": 0.6313, "grad_norm": 0.9439185261726379, "learning_rate": 0.0002, "epoch": 3.957642874464473, "step": 24480}, {"loss": 0.6076, "grad_norm": 0.935467004776001, "learning_rate": 0.0002, "epoch": 3.9592595586452184, "step": 24490}, {"loss": 0.6437, "grad_norm": 0.6900772452354431, "learning_rate": 0.0002, "epoch": 3.960876242825964, "step": 24500}, {"loss": 0.6445, "grad_norm": 1.0172916650772095, "learning_rate": 0.0002, "epoch": 3.9624929270067093, "step": 24510}, {"loss": 0.6308, "grad_norm": 0.9167046546936035, "learning_rate": 0.0002, "epoch": 3.9641096111874545, "step": 24520}, {"loss": 0.6519, "grad_norm": 0.7230527997016907, "learning_rate": 0.0002, "epoch": 3.9657262953681998, "step": 24530}, {"loss": 0.6564, "grad_norm": 0.8980403542518616, "learning_rate": 0.0002, "epoch": 3.967342979548945, "step": 24540}, {"loss": 0.6099, "grad_norm": 0.8555465936660767, "learning_rate": 0.0002, "epoch": 3.9689596637296907, "step": 24550}, {"loss": 0.6617, "grad_norm": 0.7825445532798767, "learning_rate": 0.0002, "epoch": 3.970576347910436, "step": 24560}, {"loss": 0.604, "grad_norm": 0.7273133993148804, "learning_rate": 0.0002, "epoch": 3.972193032091181, "step": 24570}, {"loss": 0.6427, "grad_norm": 0.9612047672271729, "learning_rate": 0.0002, "epoch": 3.9738097162719264, "step": 24580}, {"loss": 0.6426, "grad_norm": 0.9865460991859436, "learning_rate": 0.0002, "epoch": 3.9754264004526716, "step": 24590}, {"loss": 0.6052, "grad_norm": 0.8638762831687927, "learning_rate": 0.0002, "epoch": 3.977043084633417, "step": 24600}, {"loss": 0.6097, "grad_norm": 1.0096198320388794, "learning_rate": 0.0002, "epoch": 3.978659768814162, "step": 24610}, {"loss": 0.6664, "grad_norm": 0.8475532531738281, "learning_rate": 0.0002, "epoch": 3.9802764529949073, "step": 24620}, {"loss": 0.6711, "grad_norm": 0.9696195721626282, "learning_rate": 0.0002, "epoch": 3.9818931371756525, "step": 24630}, {"loss": 0.6446, "grad_norm": 0.7499843239784241, "learning_rate": 0.0002, "epoch": 3.9835098213563978, "step": 24640}, {"loss": 0.6054, "grad_norm": 0.8865424990653992, "learning_rate": 0.0002, "epoch": 3.9851265055371434, "step": 24650}, {"loss": 0.5975, "grad_norm": 0.8089959025382996, "learning_rate": 0.0002, "epoch": 3.9867431897178887, "step": 24660}, {"loss": 0.6677, "grad_norm": 0.6946012377738953, "learning_rate": 0.0002, "epoch": 3.988359873898634, "step": 24670}, {"loss": 0.6329, "grad_norm": 0.7991759181022644, "learning_rate": 0.0002, "epoch": 3.989976558079379, "step": 24680}, {"loss": 0.6449, "grad_norm": 0.8803931474685669, "learning_rate": 0.0002, "epoch": 3.9915932422601244, "step": 24690}, {"loss": 0.7091, "grad_norm": 0.8848299980163574, "learning_rate": 0.0002, "epoch": 3.99320992644087, "step": 24700}, {"loss": 0.6551, "grad_norm": 0.7448889017105103, "learning_rate": 0.0002, "epoch": 3.9948266106216153, "step": 24710}, {"loss": 0.6432, "grad_norm": 0.9361620545387268, "learning_rate": 0.0002, "epoch": 3.9964432948023605, "step": 24720}, {"loss": 0.5917, "grad_norm": 0.9958081245422363, "learning_rate": 0.0002, "epoch": 3.9980599789831057, "step": 24730}, {"loss": 0.6567, "grad_norm": 1.026004672050476, "learning_rate": 0.0002, "epoch": 3.999676663163851, "step": 24740}, {"eval_loss": 1.1524168252944946, "eval_runtime": 122.1585, "eval_samples_per_second": 6.0, "eval_steps_per_second": 0.753, "epoch": 4.0, "step": 24742}, {"loss": 0.6057, "grad_norm": 1.0664808750152588, "learning_rate": 0.0002, "epoch": 4.001293347344596, "step": 24750}, {"loss": 0.5644, "grad_norm": 1.0113720893859863, "learning_rate": 0.0002, "epoch": 4.002910031525341, "step": 24760}, {"loss": 0.5628, "grad_norm": 0.991486668586731, "learning_rate": 0.0002, "epoch": 4.004526715706087, "step": 24770}, {"loss": 0.508, "grad_norm": 0.951754629611969, "learning_rate": 0.0002, "epoch": 4.006143399886832, "step": 24780}, {"loss": 0.5314, "grad_norm": 1.13059401512146, "learning_rate": 0.0002, "epoch": 4.007760084067577, "step": 24790}, {"loss": 0.5323, "grad_norm": 0.9343926310539246, "learning_rate": 0.0002, "epoch": 4.009376768248322, "step": 24800}, {"loss": 0.5161, "grad_norm": 1.0680590867996216, "learning_rate": 0.0002, "epoch": 4.010993452429068, "step": 24810}, {"loss": 0.513, "grad_norm": 1.0022706985473633, "learning_rate": 0.0002, "epoch": 4.012610136609814, "step": 24820}, {"loss": 0.543, "grad_norm": 1.0285297632217407, "learning_rate": 0.0002, "epoch": 4.014226820790559, "step": 24830}, {"loss": 0.5311, "grad_norm": 0.8347002863883972, "learning_rate": 0.0002, "epoch": 4.015843504971304, "step": 24840}, {"loss": 0.5655, "grad_norm": 0.9675396680831909, "learning_rate": 0.0002, "epoch": 4.017460189152049, "step": 24850}, {"loss": 0.5625, "grad_norm": 0.9238511323928833, "learning_rate": 0.0002, "epoch": 4.019076873332795, "step": 24860}, {"loss": 0.5327, "grad_norm": 1.1576941013336182, "learning_rate": 0.0002, "epoch": 4.02069355751354, "step": 24870}, {"loss": 0.5533, "grad_norm": 0.8583757281303406, "learning_rate": 0.0002, "epoch": 4.022310241694285, "step": 24880}, {"loss": 0.5483, "grad_norm": 0.9816817045211792, "learning_rate": 0.0002, "epoch": 4.02392692587503, "step": 24890}, {"loss": 0.5605, "grad_norm": 0.955073893070221, "learning_rate": 0.0002, "epoch": 4.0255436100557755, "step": 24900}, {"loss": 0.4896, "grad_norm": 1.1054974794387817, "learning_rate": 0.0002, "epoch": 4.027160294236521, "step": 24910}, {"loss": 0.5246, "grad_norm": 1.1240060329437256, "learning_rate": 0.0002, "epoch": 4.028776978417266, "step": 24920}, {"loss": 0.5451, "grad_norm": 0.9512825012207031, "learning_rate": 0.0002, "epoch": 4.030393662598011, "step": 24930}, {"loss": 0.5584, "grad_norm": 0.85965496301651, "learning_rate": 0.0002, "epoch": 4.0320103467787565, "step": 24940}, {"loss": 0.5564, "grad_norm": 0.9378061294555664, "learning_rate": 0.0002, "epoch": 4.033627030959502, "step": 24950}, {"loss": 0.5008, "grad_norm": 0.9655424356460571, "learning_rate": 0.0002, "epoch": 4.035243715140247, "step": 24960}, {"loss": 0.5538, "grad_norm": 1.1393707990646362, "learning_rate": 0.0002, "epoch": 4.036860399320993, "step": 24970}, {"loss": 0.5785, "grad_norm": 1.0220451354980469, "learning_rate": 0.0002, "epoch": 4.038477083501738, "step": 24980}, {"loss": 0.5813, "grad_norm": 0.9785808324813843, "learning_rate": 0.0002, "epoch": 4.0400937676824835, "step": 24990}, {"loss": 0.5153, "grad_norm": 1.0257649421691895, "learning_rate": 0.0002, "epoch": 4.041710451863229, "step": 25000}, {"loss": 0.5658, "grad_norm": 0.9737892150878906, "learning_rate": 0.0002, "epoch": 4.043327136043974, "step": 25010}, {"loss": 0.5515, "grad_norm": 0.7416959404945374, "learning_rate": 0.0002, "epoch": 4.044943820224719, "step": 25020}, {"loss": 0.5372, "grad_norm": 0.7909596562385559, "learning_rate": 0.0002, "epoch": 4.046560504405464, "step": 25030}, {"loss": 0.5265, "grad_norm": 0.8923130631446838, "learning_rate": 0.0002, "epoch": 4.04817718858621, "step": 25040}, {"loss": 0.5035, "grad_norm": 0.9044941663742065, "learning_rate": 0.0002, "epoch": 4.049793872766955, "step": 25050}, {"loss": 0.5135, "grad_norm": 0.866352379322052, "learning_rate": 0.0002, "epoch": 4.0514105569477, "step": 25060}, {"loss": 0.5956, "grad_norm": 1.544549822807312, "learning_rate": 0.0002, "epoch": 4.053027241128445, "step": 25070}, {"loss": 0.5418, "grad_norm": 0.8426995277404785, "learning_rate": 0.0002, "epoch": 4.054643925309191, "step": 25080}, {"loss": 0.5537, "grad_norm": 0.9797548651695251, "learning_rate": 0.0002, "epoch": 4.056260609489936, "step": 25090}, {"loss": 0.55, "grad_norm": 0.8468434810638428, "learning_rate": 0.0002, "epoch": 4.057877293670681, "step": 25100}, {"loss": 0.5242, "grad_norm": 0.9294559955596924, "learning_rate": 0.0002, "epoch": 4.059493977851426, "step": 25110}, {"loss": 0.5295, "grad_norm": 0.9686688780784607, "learning_rate": 0.0002, "epoch": 4.061110662032172, "step": 25120}, {"loss": 0.5642, "grad_norm": 0.8042728304862976, "learning_rate": 0.0002, "epoch": 4.062727346212918, "step": 25130}, {"loss": 0.548, "grad_norm": 1.165160894393921, "learning_rate": 0.0002, "epoch": 4.064344030393663, "step": 25140}, {"loss": 0.5473, "grad_norm": 1.2161961793899536, "learning_rate": 0.0002, "epoch": 4.065960714574408, "step": 25150}, {"loss": 0.5217, "grad_norm": 1.0762810707092285, "learning_rate": 0.0002, "epoch": 4.067577398755153, "step": 25160}, {"loss": 0.5886, "grad_norm": 0.7580869793891907, "learning_rate": 0.0002, "epoch": 4.069194082935899, "step": 25170}, {"loss": 0.5401, "grad_norm": 0.9630117416381836, "learning_rate": 0.0002, "epoch": 4.070810767116644, "step": 25180}, {"loss": 0.5378, "grad_norm": 0.9049716591835022, "learning_rate": 0.0002, "epoch": 4.072427451297389, "step": 25190}, {"loss": 0.5266, "grad_norm": 1.1536930799484253, "learning_rate": 0.0002, "epoch": 4.074044135478134, "step": 25200}, {"loss": 0.5523, "grad_norm": 0.901461124420166, "learning_rate": 0.0002, "epoch": 4.0756608196588795, "step": 25210}, {"loss": 0.5132, "grad_norm": 1.3318437337875366, "learning_rate": 0.0002, "epoch": 4.077277503839625, "step": 25220}, {"loss": 0.5317, "grad_norm": 0.8811455368995667, "learning_rate": 0.0002, "epoch": 4.07889418802037, "step": 25230}, {"loss": 0.5798, "grad_norm": 1.0564165115356445, "learning_rate": 0.0002, "epoch": 4.080510872201115, "step": 25240}, {"loss": 0.5472, "grad_norm": 1.1008027791976929, "learning_rate": 0.0002, "epoch": 4.08212755638186, "step": 25250}, {"loss": 0.5195, "grad_norm": 1.150097131729126, "learning_rate": 0.0002, "epoch": 4.083744240562606, "step": 25260}, {"loss": 0.5321, "grad_norm": 0.9339924454689026, "learning_rate": 0.0002, "epoch": 4.085360924743352, "step": 25270}, {"loss": 0.5597, "grad_norm": 1.0902045965194702, "learning_rate": 0.0002, "epoch": 4.086977608924097, "step": 25280}, {"loss": 0.5203, "grad_norm": 0.8483911156654358, "learning_rate": 0.0002, "epoch": 4.088594293104842, "step": 25290}, {"loss": 0.5697, "grad_norm": 0.9477024674415588, "learning_rate": 0.0002, "epoch": 4.0902109772855875, "step": 25300}, {"loss": 0.5384, "grad_norm": 0.9500215649604797, "learning_rate": 0.0002, "epoch": 4.091827661466333, "step": 25310}, {"loss": 0.5045, "grad_norm": 1.040468454360962, "learning_rate": 0.0002, "epoch": 4.093444345647078, "step": 25320}, {"loss": 0.5488, "grad_norm": 0.7457592487335205, "learning_rate": 0.0002, "epoch": 4.095061029827823, "step": 25330}, {"loss": 0.609, "grad_norm": 1.2092097997665405, "learning_rate": 0.0002, "epoch": 4.096677714008568, "step": 25340}, {"loss": 0.5174, "grad_norm": 0.9652107954025269, "learning_rate": 0.0002, "epoch": 4.098294398189314, "step": 25350}, {"loss": 0.5559, "grad_norm": 0.8464955687522888, "learning_rate": 0.0002, "epoch": 4.099911082370059, "step": 25360}, {"loss": 0.5635, "grad_norm": 0.875026285648346, "learning_rate": 0.0002, "epoch": 4.101527766550804, "step": 25370}, {"loss": 0.5774, "grad_norm": 0.9241740107536316, "learning_rate": 0.0002, "epoch": 4.103144450731549, "step": 25380}, {"loss": 0.5578, "grad_norm": 0.9769546389579773, "learning_rate": 0.0002, "epoch": 4.1047611349122946, "step": 25390}, {"loss": 0.567, "grad_norm": 1.1501960754394531, "learning_rate": 0.0002, "epoch": 4.10637781909304, "step": 25400}, {"loss": 0.5241, "grad_norm": 0.9135243892669678, "learning_rate": 0.0002, "epoch": 4.107994503273786, "step": 25410}, {"loss": 0.5152, "grad_norm": 0.9905396103858948, "learning_rate": 0.0002, "epoch": 4.109611187454531, "step": 25420}, {"loss": 0.5064, "grad_norm": 0.9845104217529297, "learning_rate": 0.0002, "epoch": 4.111227871635276, "step": 25430}, {"loss": 0.5029, "grad_norm": 0.8326883912086487, "learning_rate": 0.0002, "epoch": 4.112844555816022, "step": 25440}, {"loss": 0.5312, "grad_norm": 0.9264556765556335, "learning_rate": 0.0002, "epoch": 4.114461239996767, "step": 25450}, {"loss": 0.5968, "grad_norm": 1.043080449104309, "learning_rate": 0.0002, "epoch": 4.116077924177512, "step": 25460}, {"loss": 0.5773, "grad_norm": 0.8533386588096619, "learning_rate": 0.0002, "epoch": 4.117694608358257, "step": 25470}, {"loss": 0.5584, "grad_norm": 1.0133965015411377, "learning_rate": 0.0002, "epoch": 4.1193112925390025, "step": 25480}, {"loss": 0.566, "grad_norm": 0.7476310133934021, "learning_rate": 0.0002, "epoch": 4.120927976719748, "step": 25490}, {"loss": 0.5189, "grad_norm": 1.1247259378433228, "learning_rate": 0.0002, "epoch": 4.122544660900493, "step": 25500}, {"loss": 0.5751, "grad_norm": 1.0764678716659546, "learning_rate": 0.0002, "epoch": 4.124161345081238, "step": 25510}, {"loss": 0.5391, "grad_norm": 0.7679798007011414, "learning_rate": 0.0002, "epoch": 4.1257780292619834, "step": 25520}, {"loss": 0.5233, "grad_norm": 0.8877071142196655, "learning_rate": 0.0002, "epoch": 4.127394713442729, "step": 25530}, {"loss": 0.5769, "grad_norm": 1.0440239906311035, "learning_rate": 0.0002, "epoch": 4.129011397623474, "step": 25540}, {"loss": 0.5723, "grad_norm": 0.984145998954773, "learning_rate": 0.0002, "epoch": 4.130628081804219, "step": 25550}, {"loss": 0.5741, "grad_norm": 0.8667055368423462, "learning_rate": 0.0002, "epoch": 4.132244765984965, "step": 25560}, {"loss": 0.5816, "grad_norm": 1.1300835609436035, "learning_rate": 0.0002, "epoch": 4.1338614501657105, "step": 25570}, {"loss": 0.524, "grad_norm": 0.9314348101615906, "learning_rate": 0.0002, "epoch": 4.135478134346456, "step": 25580}, {"loss": 0.5283, "grad_norm": 0.7731879949569702, "learning_rate": 0.0002, "epoch": 4.137094818527201, "step": 25590}, {"loss": 0.5307, "grad_norm": 1.0080097913742065, "learning_rate": 0.0002, "epoch": 4.138711502707946, "step": 25600}, {"loss": 0.5759, "grad_norm": 1.2475038766860962, "learning_rate": 0.0002, "epoch": 4.140328186888691, "step": 25610}, {"loss": 0.55, "grad_norm": 0.9912930727005005, "learning_rate": 0.0002, "epoch": 4.141944871069437, "step": 25620}, {"loss": 0.5624, "grad_norm": 0.9088651537895203, "learning_rate": 0.0002, "epoch": 4.143561555250182, "step": 25630}, {"loss": 0.5393, "grad_norm": 0.8940697312355042, "learning_rate": 0.0002, "epoch": 4.145178239430927, "step": 25640}, {"loss": 0.5341, "grad_norm": 1.0798203945159912, "learning_rate": 0.0002, "epoch": 4.146794923611672, "step": 25650}, {"loss": 0.5987, "grad_norm": 0.955172061920166, "learning_rate": 0.0002, "epoch": 4.148411607792418, "step": 25660}, {"loss": 0.569, "grad_norm": 0.9692716002464294, "learning_rate": 0.0002, "epoch": 4.150028291973163, "step": 25670}, {"loss": 0.5478, "grad_norm": 1.0813939571380615, "learning_rate": 0.0002, "epoch": 4.151644976153908, "step": 25680}, {"loss": 0.5383, "grad_norm": 1.135675072669983, "learning_rate": 0.0002, "epoch": 4.153261660334653, "step": 25690}, {"loss": 0.5247, "grad_norm": 1.0392236709594727, "learning_rate": 0.0002, "epoch": 4.1548783445153985, "step": 25700}, {"loss": 0.5204, "grad_norm": 0.9473116993904114, "learning_rate": 0.0002, "epoch": 4.156495028696145, "step": 25710}, {"loss": 0.5339, "grad_norm": 0.712493896484375, "learning_rate": 0.0002, "epoch": 4.15811171287689, "step": 25720}, {"loss": 0.5781, "grad_norm": 0.8724465370178223, "learning_rate": 0.0002, "epoch": 4.159728397057635, "step": 25730}, {"loss": 0.5325, "grad_norm": 0.9870015978813171, "learning_rate": 0.0002, "epoch": 4.16134508123838, "step": 25740}, {"loss": 0.5503, "grad_norm": 1.025273084640503, "learning_rate": 0.0002, "epoch": 4.1629617654191255, "step": 25750}, {"loss": 0.5223, "grad_norm": 0.9243090152740479, "learning_rate": 0.0002, "epoch": 4.164578449599871, "step": 25760}, {"loss": 0.5177, "grad_norm": 1.1656451225280762, "learning_rate": 0.0002, "epoch": 4.166195133780616, "step": 25770}, {"loss": 0.5334, "grad_norm": 0.936358630657196, "learning_rate": 0.0002, "epoch": 4.167811817961361, "step": 25780}, {"loss": 0.5236, "grad_norm": 0.8618208169937134, "learning_rate": 0.0002, "epoch": 4.1694285021421065, "step": 25790}, {"loss": 0.5186, "grad_norm": 0.8580600023269653, "learning_rate": 0.0002, "epoch": 4.171045186322852, "step": 25800}, {"loss": 0.5212, "grad_norm": 1.0128562450408936, "learning_rate": 0.0002, "epoch": 4.172661870503597, "step": 25810}, {"loss": 0.5404, "grad_norm": 0.854865312576294, "learning_rate": 0.0002, "epoch": 4.174278554684342, "step": 25820}, {"loss": 0.5377, "grad_norm": 1.235082745552063, "learning_rate": 0.0002, "epoch": 4.175895238865087, "step": 25830}, {"loss": 0.5614, "grad_norm": 0.9796220660209656, "learning_rate": 0.0002, "epoch": 4.177511923045833, "step": 25840}, {"loss": 0.5689, "grad_norm": 0.8922094702720642, "learning_rate": 0.0002, "epoch": 4.179128607226578, "step": 25850}, {"loss": 0.5806, "grad_norm": 0.9672530293464661, "learning_rate": 0.0002, "epoch": 4.180745291407324, "step": 25860}, {"loss": 0.5074, "grad_norm": 0.8662548661231995, "learning_rate": 0.0002, "epoch": 4.182361975588069, "step": 25870}, {"loss": 0.5329, "grad_norm": 0.7938798069953918, "learning_rate": 0.0002, "epoch": 4.1839786597688144, "step": 25880}, {"loss": 0.5427, "grad_norm": 1.0517958402633667, "learning_rate": 0.0002, "epoch": 4.18559534394956, "step": 25890}, {"loss": 0.5147, "grad_norm": 0.8939275145530701, "learning_rate": 0.0002, "epoch": 4.187212028130305, "step": 25900}, {"loss": 0.5199, "grad_norm": 1.0296672582626343, "learning_rate": 0.0002, "epoch": 4.18882871231105, "step": 25910}, {"loss": 0.5522, "grad_norm": 0.8104017972946167, "learning_rate": 0.0002, "epoch": 4.190445396491795, "step": 25920}, {"loss": 0.596, "grad_norm": 0.9984509944915771, "learning_rate": 0.0002, "epoch": 4.192062080672541, "step": 25930}, {"loss": 0.5356, "grad_norm": 0.9844784736633301, "learning_rate": 0.0002, "epoch": 4.193678764853286, "step": 25940}, {"loss": 0.5198, "grad_norm": 0.8168622255325317, "learning_rate": 0.0002, "epoch": 4.195295449034031, "step": 25950}, {"loss": 0.542, "grad_norm": 1.0878913402557373, "learning_rate": 0.0002, "epoch": 4.196912133214776, "step": 25960}, {"loss": 0.5414, "grad_norm": 0.927126407623291, "learning_rate": 0.0002, "epoch": 4.1985288173955215, "step": 25970}, {"loss": 0.5794, "grad_norm": 0.838586688041687, "learning_rate": 0.0002, "epoch": 4.200145501576267, "step": 25980}, {"loss": 0.5454, "grad_norm": 1.2572145462036133, "learning_rate": 0.0002, "epoch": 4.201762185757012, "step": 25990}, {"loss": 0.5048, "grad_norm": 1.0476740598678589, "learning_rate": 0.0002, "epoch": 4.203378869937758, "step": 26000}, {"loss": 0.5127, "grad_norm": 1.0873368978500366, "learning_rate": 0.0002, "epoch": 4.204995554118503, "step": 26010}, {"loss": 0.5679, "grad_norm": 1.2664896249771118, "learning_rate": 0.0002, "epoch": 4.206612238299249, "step": 26020}, {"loss": 0.5814, "grad_norm": 1.0312391519546509, "learning_rate": 0.0002, "epoch": 4.208228922479994, "step": 26030}, {"loss": 0.571, "grad_norm": 1.0235042572021484, "learning_rate": 0.0002, "epoch": 4.209845606660739, "step": 26040}, {"loss": 0.5766, "grad_norm": 0.8882219195365906, "learning_rate": 0.0002, "epoch": 4.211462290841484, "step": 26050}, {"loss": 0.5557, "grad_norm": 0.9115961790084839, "learning_rate": 0.0002, "epoch": 4.2130789750222295, "step": 26060}, {"loss": 0.5455, "grad_norm": 1.0218228101730347, "learning_rate": 0.0002, "epoch": 4.214695659202975, "step": 26070}, {"loss": 0.5462, "grad_norm": 1.0802232027053833, "learning_rate": 0.0002, "epoch": 4.21631234338372, "step": 26080}, {"loss": 0.557, "grad_norm": 1.1488053798675537, "learning_rate": 0.0002, "epoch": 4.217929027564465, "step": 26090}, {"loss": 0.52, "grad_norm": 1.0487725734710693, "learning_rate": 0.0002, "epoch": 4.21954571174521, "step": 26100}, {"loss": 0.5568, "grad_norm": 0.9131165742874146, "learning_rate": 0.0002, "epoch": 4.221162395925956, "step": 26110}, {"loss": 0.5206, "grad_norm": 0.9012845158576965, "learning_rate": 0.0002, "epoch": 4.222779080106701, "step": 26120}, {"loss": 0.561, "grad_norm": 0.8389840126037598, "learning_rate": 0.0002, "epoch": 4.224395764287446, "step": 26130}, {"loss": 0.5268, "grad_norm": 0.8924660682678223, "learning_rate": 0.0002, "epoch": 4.226012448468191, "step": 26140}, {"loss": 0.5715, "grad_norm": 0.8556463718414307, "learning_rate": 0.0002, "epoch": 4.2276291326489375, "step": 26150}, {"loss": 0.5695, "grad_norm": 0.9643129110336304, "learning_rate": 0.0002, "epoch": 4.229245816829683, "step": 26160}, {"loss": 0.5321, "grad_norm": 0.9865712523460388, "learning_rate": 0.0002, "epoch": 4.230862501010428, "step": 26170}, {"loss": 0.5406, "grad_norm": 1.152641773223877, "learning_rate": 0.0002, "epoch": 4.232479185191173, "step": 26180}, {"loss": 0.5632, "grad_norm": 0.9157698154449463, "learning_rate": 0.0002, "epoch": 4.234095869371918, "step": 26190}, {"loss": 0.5717, "grad_norm": 0.8418048620223999, "learning_rate": 0.0002, "epoch": 4.235712553552664, "step": 26200}, {"loss": 0.5624, "grad_norm": 0.9430168867111206, "learning_rate": 0.0002, "epoch": 4.237329237733409, "step": 26210}, {"loss": 0.5574, "grad_norm": 1.012582778930664, "learning_rate": 0.0002, "epoch": 4.238945921914154, "step": 26220}, {"loss": 0.5693, "grad_norm": 1.112619400024414, "learning_rate": 0.0002, "epoch": 4.240562606094899, "step": 26230}, {"loss": 0.6037, "grad_norm": 0.9243621826171875, "learning_rate": 0.0002, "epoch": 4.2421792902756446, "step": 26240}, {"loss": 0.569, "grad_norm": 0.6977595686912537, "learning_rate": 0.0002, "epoch": 4.24379597445639, "step": 26250}, {"loss": 0.5379, "grad_norm": 0.9600721597671509, "learning_rate": 0.0002, "epoch": 4.245412658637135, "step": 26260}, {"loss": 0.5658, "grad_norm": 0.882641613483429, "learning_rate": 0.0002, "epoch": 4.24702934281788, "step": 26270}, {"loss": 0.55, "grad_norm": 1.010920763015747, "learning_rate": 0.0002, "epoch": 4.2486460269986255, "step": 26280}, {"loss": 0.5803, "grad_norm": 0.9289400577545166, "learning_rate": 0.0002, "epoch": 4.250262711179371, "step": 26290}, {"loss": 0.541, "grad_norm": 1.137397289276123, "learning_rate": 0.0002, "epoch": 4.251879395360117, "step": 26300}, {"loss": 0.5204, "grad_norm": 1.0136182308197021, "learning_rate": 0.0002, "epoch": 4.253496079540862, "step": 26310}, {"loss": 0.5708, "grad_norm": 0.9387356042861938, "learning_rate": 0.0002, "epoch": 4.255112763721607, "step": 26320}, {"loss": 0.5948, "grad_norm": 1.1833957433700562, "learning_rate": 0.0002, "epoch": 4.2567294479023525, "step": 26330}, {"loss": 0.5905, "grad_norm": 0.9415934681892395, "learning_rate": 0.0002, "epoch": 4.258346132083098, "step": 26340}, {"loss": 0.5539, "grad_norm": 0.8550165891647339, "learning_rate": 0.0002, "epoch": 4.259962816263843, "step": 26350}, {"loss": 0.555, "grad_norm": 9.924622535705566, "learning_rate": 0.0002, "epoch": 4.261579500444588, "step": 26360}, {"loss": 0.5689, "grad_norm": 1.0104902982711792, "learning_rate": 0.0002, "epoch": 4.2631961846253335, "step": 26370}, {"loss": 0.5698, "grad_norm": 0.890794038772583, "learning_rate": 0.0002, "epoch": 4.264812868806079, "step": 26380}, {"loss": 0.563, "grad_norm": 1.0560191869735718, "learning_rate": 0.0002, "epoch": 4.266429552986824, "step": 26390}, {"loss": 0.5119, "grad_norm": 1.0135581493377686, "learning_rate": 0.0002, "epoch": 4.268046237167569, "step": 26400}, {"loss": 0.5359, "grad_norm": 1.1304140090942383, "learning_rate": 0.0002, "epoch": 4.269662921348314, "step": 26410}, {"loss": 0.5615, "grad_norm": 0.9899303913116455, "learning_rate": 0.0002, "epoch": 4.27127960552906, "step": 26420}, {"loss": 0.5815, "grad_norm": 1.0505329370498657, "learning_rate": 0.0002, "epoch": 4.272896289709805, "step": 26430}, {"loss": 0.5384, "grad_norm": 0.9389396905899048, "learning_rate": 0.0002, "epoch": 4.27451297389055, "step": 26440}, {"loss": 0.5558, "grad_norm": 0.875328779220581, "learning_rate": 0.0002, "epoch": 4.276129658071296, "step": 26450}, {"loss": 0.5601, "grad_norm": 1.0689256191253662, "learning_rate": 0.0002, "epoch": 4.277746342252041, "step": 26460}, {"loss": 0.546, "grad_norm": 0.9988957643508911, "learning_rate": 0.0002, "epoch": 4.279363026432787, "step": 26470}, {"loss": 0.5478, "grad_norm": 0.8721813559532166, "learning_rate": 0.0002, "epoch": 4.280979710613532, "step": 26480}, {"loss": 0.5424, "grad_norm": 1.100109577178955, "learning_rate": 0.0002, "epoch": 4.282596394794277, "step": 26490}, {"loss": 0.572, "grad_norm": 1.1607271432876587, "learning_rate": 0.0002, "epoch": 4.284213078975022, "step": 26500}, {"loss": 0.6287, "grad_norm": 0.879088819026947, "learning_rate": 0.0002, "epoch": 4.285829763155768, "step": 26510}, {"loss": 0.573, "grad_norm": 0.9891700744628906, "learning_rate": 0.0002, "epoch": 4.287446447336513, "step": 26520}, {"loss": 0.6018, "grad_norm": 1.0831127166748047, "learning_rate": 0.0002, "epoch": 4.289063131517258, "step": 26530}, {"loss": 0.5693, "grad_norm": 1.4108285903930664, "learning_rate": 0.0002, "epoch": 4.290679815698003, "step": 26540}, {"loss": 0.5888, "grad_norm": 1.0630289316177368, "learning_rate": 0.0002, "epoch": 4.2922964998787485, "step": 26550}, {"loss": 0.5817, "grad_norm": 1.0854572057724, "learning_rate": 0.0002, "epoch": 4.293913184059494, "step": 26560}, {"loss": 0.5586, "grad_norm": 0.9561646580696106, "learning_rate": 0.0002, "epoch": 4.295529868240239, "step": 26570}, {"loss": 0.5674, "grad_norm": 0.9064981937408447, "learning_rate": 0.0002, "epoch": 4.297146552420984, "step": 26580}, {"loss": 0.5847, "grad_norm": 1.0082972049713135, "learning_rate": 0.0002, "epoch": 4.298763236601729, "step": 26590}, {"loss": 0.5711, "grad_norm": 1.1613214015960693, "learning_rate": 0.0002, "epoch": 4.3003799207824756, "step": 26600}, {"loss": 0.551, "grad_norm": 0.9847695231437683, "learning_rate": 0.0002, "epoch": 4.301996604963221, "step": 26610}, {"loss": 0.6089, "grad_norm": 1.0980697870254517, "learning_rate": 0.0002, "epoch": 4.303613289143966, "step": 26620}, {"loss": 0.5797, "grad_norm": 0.8861175179481506, "learning_rate": 0.0002, "epoch": 4.305229973324711, "step": 26630}, {"loss": 0.5716, "grad_norm": 0.8917363286018372, "learning_rate": 0.0002, "epoch": 4.3068466575054565, "step": 26640}, {"loss": 0.5892, "grad_norm": 1.0458378791809082, "learning_rate": 0.0002, "epoch": 4.308463341686202, "step": 26650}, {"loss": 0.5883, "grad_norm": 1.4859240055084229, "learning_rate": 0.0002, "epoch": 4.310080025866947, "step": 26660}, {"loss": 0.5296, "grad_norm": 1.1376359462738037, "learning_rate": 0.0002, "epoch": 4.311696710047692, "step": 26670}, {"loss": 0.5671, "grad_norm": 0.991349995136261, "learning_rate": 0.0002, "epoch": 4.313313394228437, "step": 26680}, {"loss": 0.5338, "grad_norm": 0.9995543956756592, "learning_rate": 0.0002, "epoch": 4.314930078409183, "step": 26690}, {"loss": 0.5542, "grad_norm": 1.0515851974487305, "learning_rate": 0.0002, "epoch": 4.316546762589928, "step": 26700}, {"loss": 0.5473, "grad_norm": 1.008023977279663, "learning_rate": 0.0002, "epoch": 4.318163446770673, "step": 26710}, {"loss": 0.5506, "grad_norm": 1.0184582471847534, "learning_rate": 0.0002, "epoch": 4.319780130951418, "step": 26720}, {"loss": 0.5828, "grad_norm": 1.161071538925171, "learning_rate": 0.0002, "epoch": 4.321396815132164, "step": 26730}, {"loss": 0.5633, "grad_norm": 0.9580779671669006, "learning_rate": 0.0002, "epoch": 4.323013499312909, "step": 26740}, {"loss": 0.5785, "grad_norm": 1.0189911127090454, "learning_rate": 0.0002, "epoch": 4.324630183493655, "step": 26750}, {"loss": 0.5237, "grad_norm": 0.7484358549118042, "learning_rate": 0.0002, "epoch": 4.3262468676744, "step": 26760}, {"loss": 0.5728, "grad_norm": 1.0015908479690552, "learning_rate": 0.0002, "epoch": 4.327863551855145, "step": 26770}, {"loss": 0.5597, "grad_norm": 0.8972945809364319, "learning_rate": 0.0002, "epoch": 4.329480236035891, "step": 26780}, {"loss": 0.5857, "grad_norm": 1.01099693775177, "learning_rate": 0.0002, "epoch": 4.331096920216636, "step": 26790}, {"loss": 0.5591, "grad_norm": 0.846958339214325, "learning_rate": 0.0002, "epoch": 4.332713604397381, "step": 26800}, {"loss": 0.5547, "grad_norm": 1.0792603492736816, "learning_rate": 0.0002, "epoch": 4.334330288578126, "step": 26810}, {"loss": 0.5747, "grad_norm": 1.0373345613479614, "learning_rate": 0.0002, "epoch": 4.3359469727588715, "step": 26820}, {"loss": 0.558, "grad_norm": 0.9779167771339417, "learning_rate": 0.0002, "epoch": 4.337563656939617, "step": 26830}, {"loss": 0.5821, "grad_norm": 1.0235520601272583, "learning_rate": 0.0002, "epoch": 4.339180341120362, "step": 26840}, {"loss": 0.5843, "grad_norm": 1.04195237159729, "learning_rate": 0.0002, "epoch": 4.340797025301107, "step": 26850}, {"loss": 0.5474, "grad_norm": 0.9479565620422363, "learning_rate": 0.0002, "epoch": 4.3424137094818525, "step": 26860}, {"loss": 0.5646, "grad_norm": 0.9526172280311584, "learning_rate": 0.0002, "epoch": 4.344030393662598, "step": 26870}, {"loss": 0.521, "grad_norm": 0.8571456074714661, "learning_rate": 0.0002, "epoch": 4.345647077843343, "step": 26880}, {"loss": 0.5846, "grad_norm": 0.9475828409194946, "learning_rate": 0.0002, "epoch": 4.347263762024088, "step": 26890}, {"loss": 0.5815, "grad_norm": 1.0529576539993286, "learning_rate": 0.0002, "epoch": 4.348880446204834, "step": 26900}, {"loss": 0.56, "grad_norm": 0.9648140072822571, "learning_rate": 0.0002, "epoch": 4.3504971303855795, "step": 26910}, {"loss": 0.5162, "grad_norm": 1.0488841533660889, "learning_rate": 0.0002, "epoch": 4.352113814566325, "step": 26920}, {"loss": 0.5842, "grad_norm": 0.8771942257881165, "learning_rate": 0.0002, "epoch": 4.35373049874707, "step": 26930}, {"loss": 0.5966, "grad_norm": 0.9411202073097229, "learning_rate": 0.0002, "epoch": 4.355347182927815, "step": 26940}, {"loss": 0.6001, "grad_norm": 1.0997588634490967, "learning_rate": 0.0002, "epoch": 4.35696386710856, "step": 26950}, {"loss": 0.5528, "grad_norm": 0.968754768371582, "learning_rate": 0.0002, "epoch": 4.358580551289306, "step": 26960}, {"loss": 0.5881, "grad_norm": 0.9990773797035217, "learning_rate": 0.0002, "epoch": 4.360197235470051, "step": 26970}, {"loss": 0.5761, "grad_norm": 1.0210620164871216, "learning_rate": 0.0002, "epoch": 4.361813919650796, "step": 26980}, {"loss": 0.5768, "grad_norm": 0.855462908744812, "learning_rate": 0.0002, "epoch": 4.363430603831541, "step": 26990}, {"loss": 0.5493, "grad_norm": 0.9169660806655884, "learning_rate": 0.0002, "epoch": 4.365047288012287, "step": 27000}, {"loss": 0.5697, "grad_norm": 1.089629888534546, "learning_rate": 0.0002, "epoch": 4.366663972193032, "step": 27010}, {"loss": 0.5854, "grad_norm": 1.0932867527008057, "learning_rate": 0.0002, "epoch": 4.368280656373777, "step": 27020}, {"loss": 0.5656, "grad_norm": 0.9290956854820251, "learning_rate": 0.0002, "epoch": 4.369897340554522, "step": 27030}, {"loss": 0.5727, "grad_norm": 1.2800624370574951, "learning_rate": 0.0002, "epoch": 4.3715140247352675, "step": 27040}, {"loss": 0.5837, "grad_norm": 0.8993493318557739, "learning_rate": 0.0002, "epoch": 4.373130708916014, "step": 27050}, {"loss": 0.6232, "grad_norm": 1.1566431522369385, "learning_rate": 0.0002, "epoch": 4.374747393096759, "step": 27060}, {"loss": 0.5902, "grad_norm": 0.9479052424430847, "learning_rate": 0.0002, "epoch": 4.376364077277504, "step": 27070}, {"loss": 0.6189, "grad_norm": 1.0063648223876953, "learning_rate": 0.0002, "epoch": 4.377980761458249, "step": 27080}, {"loss": 0.561, "grad_norm": 0.8342045545578003, "learning_rate": 0.0002, "epoch": 4.379597445638995, "step": 27090}, {"loss": 0.5515, "grad_norm": 1.1390739679336548, "learning_rate": 0.0002, "epoch": 4.38121412981974, "step": 27100}, {"loss": 0.5372, "grad_norm": 0.9547637104988098, "learning_rate": 0.0002, "epoch": 4.382830814000485, "step": 27110}, {"loss": 0.5728, "grad_norm": 1.0503804683685303, "learning_rate": 0.0002, "epoch": 4.38444749818123, "step": 27120}, {"loss": 0.5787, "grad_norm": 0.9064017534255981, "learning_rate": 0.0002, "epoch": 4.3860641823619755, "step": 27130}, {"loss": 0.5798, "grad_norm": 0.9382519125938416, "learning_rate": 0.0002, "epoch": 4.387680866542721, "step": 27140}, {"loss": 0.5791, "grad_norm": 1.0410341024398804, "learning_rate": 0.0002, "epoch": 4.389297550723466, "step": 27150}, {"loss": 0.6034, "grad_norm": 0.9218655824661255, "learning_rate": 0.0002, "epoch": 4.390914234904211, "step": 27160}, {"loss": 0.5204, "grad_norm": 0.8119737505912781, "learning_rate": 0.0002, "epoch": 4.392530919084956, "step": 27170}, {"loss": 0.5612, "grad_norm": 0.8584722876548767, "learning_rate": 0.0002, "epoch": 4.394147603265702, "step": 27180}, {"loss": 0.5772, "grad_norm": 0.9668293595314026, "learning_rate": 0.0002, "epoch": 4.395764287446447, "step": 27190}, {"loss": 0.6009, "grad_norm": 1.022334098815918, "learning_rate": 0.0002, "epoch": 4.397380971627193, "step": 27200}, {"loss": 0.5573, "grad_norm": 0.9553216099739075, "learning_rate": 0.0002, "epoch": 4.398997655807938, "step": 27210}, {"loss": 0.5604, "grad_norm": 0.9282339215278625, "learning_rate": 0.0002, "epoch": 4.4006143399886835, "step": 27220}, {"loss": 0.5599, "grad_norm": 1.0232292413711548, "learning_rate": 0.0002, "epoch": 4.402231024169429, "step": 27230}, {"loss": 0.6078, "grad_norm": 0.9915700554847717, "learning_rate": 0.0002, "epoch": 4.403847708350174, "step": 27240}, {"loss": 0.5778, "grad_norm": 1.0014961957931519, "learning_rate": 0.0002, "epoch": 4.405464392530919, "step": 27250}, {"loss": 0.5824, "grad_norm": 1.1172103881835938, "learning_rate": 0.0002, "epoch": 4.407081076711664, "step": 27260}, {"loss": 0.5286, "grad_norm": 0.8583093285560608, "learning_rate": 0.0002, "epoch": 4.40869776089241, "step": 27270}, {"loss": 0.5507, "grad_norm": 0.7609201669692993, "learning_rate": 0.0002, "epoch": 4.410314445073155, "step": 27280}, {"loss": 0.575, "grad_norm": 1.0619351863861084, "learning_rate": 0.0002, "epoch": 4.4119311292539, "step": 27290}, {"loss": 0.5579, "grad_norm": 1.0177674293518066, "learning_rate": 0.0002, "epoch": 4.413547813434645, "step": 27300}, {"loss": 0.5628, "grad_norm": 0.9921218156814575, "learning_rate": 0.0002, "epoch": 4.4151644976153905, "step": 27310}, {"loss": 0.6018, "grad_norm": 1.126244306564331, "learning_rate": 0.0002, "epoch": 4.416781181796136, "step": 27320}, {"loss": 0.5743, "grad_norm": 1.0678540468215942, "learning_rate": 0.0002, "epoch": 4.418397865976881, "step": 27330}, {"loss": 0.5665, "grad_norm": 0.8705704212188721, "learning_rate": 0.0002, "epoch": 4.420014550157627, "step": 27340}, {"loss": 0.5763, "grad_norm": 1.272074818611145, "learning_rate": 0.0002, "epoch": 4.421631234338372, "step": 27350}, {"loss": 0.561, "grad_norm": 0.8740444183349609, "learning_rate": 0.0002, "epoch": 4.423247918519118, "step": 27360}, {"loss": 0.5492, "grad_norm": 1.0584250688552856, "learning_rate": 0.0002, "epoch": 4.424864602699863, "step": 27370}, {"loss": 0.589, "grad_norm": 1.059870719909668, "learning_rate": 0.0002, "epoch": 4.426481286880608, "step": 27380}, {"loss": 0.5551, "grad_norm": 1.072265863418579, "learning_rate": 0.0002, "epoch": 4.428097971061353, "step": 27390}, {"loss": 0.5584, "grad_norm": 0.871481716632843, "learning_rate": 0.0002, "epoch": 4.4297146552420985, "step": 27400}, {"loss": 0.5372, "grad_norm": 0.9555448293685913, "learning_rate": 0.0002, "epoch": 4.431331339422844, "step": 27410}, {"loss": 0.5593, "grad_norm": 1.0402292013168335, "learning_rate": 0.0002, "epoch": 4.432948023603589, "step": 27420}, {"loss": 0.5532, "grad_norm": 1.12587571144104, "learning_rate": 0.0002, "epoch": 4.434564707784334, "step": 27430}, {"loss": 0.5403, "grad_norm": 1.0783193111419678, "learning_rate": 0.0002, "epoch": 4.436181391965079, "step": 27440}, {"loss": 0.5313, "grad_norm": 1.024133563041687, "learning_rate": 0.0002, "epoch": 4.437798076145825, "step": 27450}, {"loss": 0.5621, "grad_norm": 0.9156768918037415, "learning_rate": 0.0002, "epoch": 4.43941476032657, "step": 27460}, {"loss": 0.5307, "grad_norm": 1.0215224027633667, "learning_rate": 0.0002, "epoch": 4.441031444507315, "step": 27470}, {"loss": 0.5188, "grad_norm": 1.082116961479187, "learning_rate": 0.0002, "epoch": 4.442648128688061, "step": 27480}, {"loss": 0.6203, "grad_norm": 1.0412873029708862, "learning_rate": 0.0002, "epoch": 4.4442648128688065, "step": 27490}, {"loss": 0.5939, "grad_norm": 1.0509289503097534, "learning_rate": 0.0002, "epoch": 4.445881497049552, "step": 27500}, {"loss": 0.5503, "grad_norm": 0.9291498064994812, "learning_rate": 0.0002, "epoch": 4.447498181230297, "step": 27510}, {"loss": 0.5408, "grad_norm": 0.970184326171875, "learning_rate": 0.0002, "epoch": 4.449114865411042, "step": 27520}, {"loss": 0.5705, "grad_norm": 0.8418883681297302, "learning_rate": 0.0002, "epoch": 4.450731549591787, "step": 27530}, {"loss": 0.5124, "grad_norm": 0.8823825120925903, "learning_rate": 0.0002, "epoch": 4.452348233772533, "step": 27540}, {"loss": 0.5867, "grad_norm": 1.1909019947052002, "learning_rate": 0.0002, "epoch": 4.453964917953278, "step": 27550}, {"loss": 0.5685, "grad_norm": 1.0317302942276, "learning_rate": 0.0002, "epoch": 4.455581602134023, "step": 27560}, {"loss": 0.5538, "grad_norm": 0.9977751970291138, "learning_rate": 0.0002, "epoch": 4.457198286314768, "step": 27570}, {"loss": 0.5628, "grad_norm": 0.8909519910812378, "learning_rate": 0.0002, "epoch": 4.458814970495514, "step": 27580}, {"loss": 0.6099, "grad_norm": 0.8653029799461365, "learning_rate": 0.0002, "epoch": 4.460431654676259, "step": 27590}, {"loss": 0.5622, "grad_norm": 1.0783653259277344, "learning_rate": 0.0002, "epoch": 4.462048338857004, "step": 27600}, {"loss": 0.579, "grad_norm": 1.1235394477844238, "learning_rate": 0.0002, "epoch": 4.463665023037749, "step": 27610}, {"loss": 0.5545, "grad_norm": 0.9386643767356873, "learning_rate": 0.0002, "epoch": 4.4652817072184945, "step": 27620}, {"loss": 0.5554, "grad_norm": 1.0605148077011108, "learning_rate": 0.0002, "epoch": 4.466898391399241, "step": 27630}, {"loss": 0.5886, "grad_norm": 1.1283893585205078, "learning_rate": 0.0002, "epoch": 4.468515075579986, "step": 27640}, {"loss": 0.5801, "grad_norm": 1.0583468675613403, "learning_rate": 0.0002, "epoch": 4.470131759760731, "step": 27650}, {"loss": 0.5601, "grad_norm": 0.9563992023468018, "learning_rate": 0.0002, "epoch": 4.471748443941476, "step": 27660}, {"loss": 0.5687, "grad_norm": 1.100598931312561, "learning_rate": 0.0002, "epoch": 4.4733651281222215, "step": 27670}, {"loss": 0.589, "grad_norm": 0.9386957287788391, "learning_rate": 0.0002, "epoch": 4.474981812302967, "step": 27680}, {"loss": 0.6241, "grad_norm": 1.2946288585662842, "learning_rate": 0.0002, "epoch": 4.476598496483712, "step": 27690}, {"loss": 0.6075, "grad_norm": 1.0325199365615845, "learning_rate": 0.0002, "epoch": 4.478215180664457, "step": 27700}, {"loss": 0.588, "grad_norm": 1.0318928956985474, "learning_rate": 0.0002, "epoch": 4.4798318648452025, "step": 27710}, {"loss": 0.5656, "grad_norm": 0.8721024394035339, "learning_rate": 0.0002, "epoch": 4.481448549025948, "step": 27720}, {"loss": 0.5421, "grad_norm": 1.17376708984375, "learning_rate": 0.0002, "epoch": 4.483065233206693, "step": 27730}, {"loss": 0.5657, "grad_norm": 1.0926326513290405, "learning_rate": 0.0002, "epoch": 4.484681917387438, "step": 27740}, {"loss": 0.5514, "grad_norm": 0.9043852686882019, "learning_rate": 0.0002, "epoch": 4.486298601568183, "step": 27750}, {"loss": 0.582, "grad_norm": 1.064600944519043, "learning_rate": 0.0002, "epoch": 4.487915285748929, "step": 27760}, {"loss": 0.6108, "grad_norm": 0.7833460569381714, "learning_rate": 0.0002, "epoch": 4.489531969929674, "step": 27770}, {"loss": 0.5985, "grad_norm": 1.1073496341705322, "learning_rate": 0.0002, "epoch": 4.49114865411042, "step": 27780}, {"loss": 0.5577, "grad_norm": 1.0799397230148315, "learning_rate": 0.0002, "epoch": 4.492765338291165, "step": 27790}, {"loss": 0.5601, "grad_norm": 1.1062238216400146, "learning_rate": 0.0002, "epoch": 4.49438202247191, "step": 27800}, {"loss": 0.6126, "grad_norm": 1.0568242073059082, "learning_rate": 0.0002, "epoch": 4.495998706652656, "step": 27810}, {"loss": 0.5913, "grad_norm": 0.8861091732978821, "learning_rate": 0.0002, "epoch": 4.497615390833401, "step": 27820}, {"loss": 0.5858, "grad_norm": 1.2297543287277222, "learning_rate": 0.0002, "epoch": 4.499232075014146, "step": 27830}, {"loss": 0.5859, "grad_norm": 0.9600302577018738, "learning_rate": 0.0002, "epoch": 4.500848759194891, "step": 27840}, {"loss": 0.6124, "grad_norm": 1.057051181793213, "learning_rate": 0.0002, "epoch": 4.502465443375637, "step": 27850}, {"loss": 0.5788, "grad_norm": 0.9839690923690796, "learning_rate": 0.0002, "epoch": 4.504082127556382, "step": 27860}, {"loss": 0.555, "grad_norm": 1.1479853391647339, "learning_rate": 0.0002, "epoch": 4.505698811737127, "step": 27870}, {"loss": 0.6039, "grad_norm": 1.0550768375396729, "learning_rate": 0.0002, "epoch": 4.507315495917872, "step": 27880}, {"loss": 0.563, "grad_norm": 0.898209273815155, "learning_rate": 0.0002, "epoch": 4.5089321800986175, "step": 27890}, {"loss": 0.5734, "grad_norm": 0.9460315108299255, "learning_rate": 0.0002, "epoch": 4.510548864279363, "step": 27900}, {"loss": 0.5702, "grad_norm": 0.9499884247779846, "learning_rate": 0.0002, "epoch": 4.512165548460108, "step": 27910}, {"loss": 0.5385, "grad_norm": 0.7801318764686584, "learning_rate": 0.0002, "epoch": 4.513782232640853, "step": 27920}, {"loss": 0.5391, "grad_norm": 0.9286966323852539, "learning_rate": 0.0002, "epoch": 4.515398916821599, "step": 27930}, {"loss": 0.5717, "grad_norm": 0.9539980292320251, "learning_rate": 0.0002, "epoch": 4.517015601002345, "step": 27940}, {"loss": 0.6073, "grad_norm": 1.1053401231765747, "learning_rate": 0.0002, "epoch": 4.51863228518309, "step": 27950}, {"loss": 0.6087, "grad_norm": 0.7535534501075745, "learning_rate": 0.0002, "epoch": 4.520248969363835, "step": 27960}, {"loss": 0.5701, "grad_norm": 1.076926589012146, "learning_rate": 0.0002, "epoch": 4.52186565354458, "step": 27970}, {"loss": 0.6028, "grad_norm": 1.181935429573059, "learning_rate": 0.0002, "epoch": 4.5234823377253255, "step": 27980}, {"loss": 0.6033, "grad_norm": 0.9293407201766968, "learning_rate": 0.0002, "epoch": 4.525099021906071, "step": 27990}, {"loss": 0.5815, "grad_norm": 0.8953009247779846, "learning_rate": 0.0002, "epoch": 4.526715706086816, "step": 28000}, {"loss": 0.5564, "grad_norm": 1.0850225687026978, "learning_rate": 0.0002, "epoch": 4.528332390267561, "step": 28010}, {"loss": 0.5459, "grad_norm": 0.9125663042068481, "learning_rate": 0.0002, "epoch": 4.529949074448306, "step": 28020}, {"loss": 0.5922, "grad_norm": 0.8745216727256775, "learning_rate": 0.0002, "epoch": 4.531565758629052, "step": 28030}, {"loss": 0.567, "grad_norm": 1.0783463716506958, "learning_rate": 0.0002, "epoch": 4.533182442809797, "step": 28040}, {"loss": 0.5754, "grad_norm": 0.7513844966888428, "learning_rate": 0.0002, "epoch": 4.534799126990542, "step": 28050}, {"loss": 0.5608, "grad_norm": 1.0135776996612549, "learning_rate": 0.0002, "epoch": 4.536415811171287, "step": 28060}, {"loss": 0.5827, "grad_norm": 0.8886825442314148, "learning_rate": 0.0002, "epoch": 4.538032495352033, "step": 28070}, {"loss": 0.5605, "grad_norm": 0.8153995275497437, "learning_rate": 0.0002, "epoch": 4.539649179532779, "step": 28080}, {"loss": 0.6377, "grad_norm": 0.9853341579437256, "learning_rate": 0.0002, "epoch": 4.541265863713524, "step": 28090}, {"loss": 0.5957, "grad_norm": 0.9365800023078918, "learning_rate": 0.0002, "epoch": 4.542882547894269, "step": 28100}, {"loss": 0.5477, "grad_norm": 0.9765017628669739, "learning_rate": 0.0002, "epoch": 4.544499232075014, "step": 28110}, {"loss": 0.6185, "grad_norm": 0.9811279773712158, "learning_rate": 0.0002, "epoch": 4.54611591625576, "step": 28120}, {"loss": 0.6095, "grad_norm": 1.0387924909591675, "learning_rate": 0.0002, "epoch": 4.547732600436505, "step": 28130}, {"loss": 0.6534, "grad_norm": 1.0684878826141357, "learning_rate": 0.0002, "epoch": 4.54934928461725, "step": 28140}, {"loss": 0.5701, "grad_norm": 1.0000102519989014, "learning_rate": 0.0002, "epoch": 4.550965968797995, "step": 28150}, {"loss": 0.5327, "grad_norm": 1.0717930793762207, "learning_rate": 0.0002, "epoch": 4.5525826529787405, "step": 28160}, {"loss": 0.5594, "grad_norm": 0.990074634552002, "learning_rate": 0.0002, "epoch": 4.554199337159486, "step": 28170}, {"loss": 0.5452, "grad_norm": 0.8673754930496216, "learning_rate": 0.0002, "epoch": 4.555816021340231, "step": 28180}, {"loss": 0.5773, "grad_norm": 0.864247739315033, "learning_rate": 0.0002, "epoch": 4.557432705520976, "step": 28190}, {"loss": 0.5516, "grad_norm": 0.8280200958251953, "learning_rate": 0.0002, "epoch": 4.5590493897017215, "step": 28200}, {"loss": 0.5709, "grad_norm": 1.1312172412872314, "learning_rate": 0.0002, "epoch": 4.560666073882467, "step": 28210}, {"loss": 0.5776, "grad_norm": 0.9147403240203857, "learning_rate": 0.0002, "epoch": 4.562282758063212, "step": 28220}, {"loss": 0.5591, "grad_norm": 1.0321218967437744, "learning_rate": 0.0002, "epoch": 4.563899442243958, "step": 28230}, {"loss": 0.5508, "grad_norm": 1.168332815170288, "learning_rate": 0.0002, "epoch": 4.565516126424703, "step": 28240}, {"loss": 0.5649, "grad_norm": 1.0067222118377686, "learning_rate": 0.0002, "epoch": 4.5671328106054485, "step": 28250}, {"loss": 0.5853, "grad_norm": 1.0283393859863281, "learning_rate": 0.0002, "epoch": 4.568749494786194, "step": 28260}, {"loss": 0.5772, "grad_norm": 0.9912363886833191, "learning_rate": 0.0002, "epoch": 4.570366178966939, "step": 28270}, {"loss": 0.5757, "grad_norm": 1.108032464981079, "learning_rate": 0.0002, "epoch": 4.571982863147684, "step": 28280}, {"loss": 0.5529, "grad_norm": 0.8260078430175781, "learning_rate": 0.0002, "epoch": 4.573599547328429, "step": 28290}, {"loss": 0.5625, "grad_norm": 0.8946247100830078, "learning_rate": 0.0002, "epoch": 4.575216231509175, "step": 28300}, {"loss": 0.5533, "grad_norm": 0.8273587822914124, "learning_rate": 0.0002, "epoch": 4.57683291568992, "step": 28310}, {"loss": 0.6058, "grad_norm": 0.9040093421936035, "learning_rate": 0.0002, "epoch": 4.578449599870665, "step": 28320}, {"loss": 0.5521, "grad_norm": 0.8435290455818176, "learning_rate": 0.0002, "epoch": 4.58006628405141, "step": 28330}, {"loss": 0.6086, "grad_norm": 1.164088249206543, "learning_rate": 0.0002, "epoch": 4.581682968232156, "step": 28340}, {"loss": 0.5603, "grad_norm": 0.9861085414886475, "learning_rate": 0.0002, "epoch": 4.583299652412901, "step": 28350}, {"loss": 0.5701, "grad_norm": 0.8892980813980103, "learning_rate": 0.0002, "epoch": 4.584916336593646, "step": 28360}, {"loss": 0.598, "grad_norm": 1.240574836730957, "learning_rate": 0.0002, "epoch": 4.586533020774391, "step": 28370}, {"loss": 0.5797, "grad_norm": 0.8669408559799194, "learning_rate": 0.0002, "epoch": 4.588149704955137, "step": 28380}, {"loss": 0.5603, "grad_norm": 0.9145985841751099, "learning_rate": 0.0002, "epoch": 4.589766389135883, "step": 28390}, {"loss": 0.5765, "grad_norm": 0.8584614992141724, "learning_rate": 0.0002, "epoch": 4.591383073316628, "step": 28400}, {"loss": 0.5898, "grad_norm": 1.118829369544983, "learning_rate": 0.0002, "epoch": 4.592999757497373, "step": 28410}, {"loss": 0.5641, "grad_norm": 1.1411553621292114, "learning_rate": 0.0002, "epoch": 4.594616441678118, "step": 28420}, {"loss": 0.549, "grad_norm": 0.9433278441429138, "learning_rate": 0.0002, "epoch": 4.596233125858864, "step": 28430}, {"loss": 0.5496, "grad_norm": 0.816830039024353, "learning_rate": 0.0002, "epoch": 4.597849810039609, "step": 28440}, {"loss": 0.5543, "grad_norm": 1.2124968767166138, "learning_rate": 0.0002, "epoch": 4.599466494220354, "step": 28450}, {"loss": 0.5759, "grad_norm": 0.9658762216567993, "learning_rate": 0.0002, "epoch": 4.601083178401099, "step": 28460}, {"loss": 0.5902, "grad_norm": 0.836100161075592, "learning_rate": 0.0002, "epoch": 4.6026998625818445, "step": 28470}, {"loss": 0.5749, "grad_norm": 0.9989104270935059, "learning_rate": 0.0002, "epoch": 4.60431654676259, "step": 28480}, {"loss": 0.5616, "grad_norm": 1.1298956871032715, "learning_rate": 0.0002, "epoch": 4.605933230943335, "step": 28490}, {"loss": 0.5846, "grad_norm": 1.1731704473495483, "learning_rate": 0.0002, "epoch": 4.60754991512408, "step": 28500}, {"loss": 0.5816, "grad_norm": 0.9624714255332947, "learning_rate": 0.0002, "epoch": 4.609166599304825, "step": 28510}, {"loss": 0.5868, "grad_norm": 1.364073634147644, "learning_rate": 0.0002, "epoch": 4.610783283485571, "step": 28520}, {"loss": 0.6237, "grad_norm": 1.1827356815338135, "learning_rate": 0.0002, "epoch": 4.612399967666317, "step": 28530}, {"loss": 0.5643, "grad_norm": 0.6651531457901001, "learning_rate": 0.0002, "epoch": 4.614016651847062, "step": 28540}, {"loss": 0.6051, "grad_norm": 1.1640995740890503, "learning_rate": 0.0002, "epoch": 4.615633336027807, "step": 28550}, {"loss": 0.5995, "grad_norm": 1.028918743133545, "learning_rate": 0.0002, "epoch": 4.6172500202085525, "step": 28560}, {"loss": 0.5607, "grad_norm": 0.8252120614051819, "learning_rate": 0.0002, "epoch": 4.618866704389298, "step": 28570}, {"loss": 0.5769, "grad_norm": 1.3536735773086548, "learning_rate": 0.0002, "epoch": 4.620483388570043, "step": 28580}, {"loss": 0.6006, "grad_norm": 1.2146915197372437, "learning_rate": 0.0002, "epoch": 4.622100072750788, "step": 28590}, {"loss": 0.5503, "grad_norm": 1.0122549533843994, "learning_rate": 0.0002, "epoch": 4.623716756931533, "step": 28600}, {"loss": 0.6072, "grad_norm": 0.9977872967720032, "learning_rate": 0.0002, "epoch": 4.625333441112279, "step": 28610}, {"loss": 0.5669, "grad_norm": 1.0159751176834106, "learning_rate": 0.0002, "epoch": 4.626950125293024, "step": 28620}, {"loss": 0.5935, "grad_norm": 1.0028325319290161, "learning_rate": 0.0002, "epoch": 4.628566809473769, "step": 28630}, {"loss": 0.5515, "grad_norm": 0.901638388633728, "learning_rate": 0.0002, "epoch": 4.630183493654514, "step": 28640}, {"loss": 0.595, "grad_norm": 0.9450507164001465, "learning_rate": 0.0002, "epoch": 4.6318001778352595, "step": 28650}, {"loss": 0.5972, "grad_norm": 0.9987545013427734, "learning_rate": 0.0002, "epoch": 4.633416862016006, "step": 28660}, {"loss": 0.5863, "grad_norm": 0.9574332237243652, "learning_rate": 0.0002, "epoch": 4.63503354619675, "step": 28670}, {"loss": 0.5804, "grad_norm": 1.2215653657913208, "learning_rate": 0.0002, "epoch": 4.636650230377496, "step": 28680}, {"loss": 0.5798, "grad_norm": 0.9798858761787415, "learning_rate": 0.0002, "epoch": 4.638266914558241, "step": 28690}, {"loss": 0.5773, "grad_norm": 1.0648466348648071, "learning_rate": 0.0002, "epoch": 4.639883598738987, "step": 28700}, {"loss": 0.6108, "grad_norm": 1.0606504678726196, "learning_rate": 0.0002, "epoch": 4.641500282919732, "step": 28710}, {"loss": 0.5801, "grad_norm": 1.0892442464828491, "learning_rate": 0.0002, "epoch": 4.643116967100477, "step": 28720}, {"loss": 0.5492, "grad_norm": 0.914391040802002, "learning_rate": 0.0002, "epoch": 4.644733651281222, "step": 28730}, {"loss": 0.5439, "grad_norm": 0.9782370328903198, "learning_rate": 0.0002, "epoch": 4.6463503354619675, "step": 28740}, {"loss": 0.6035, "grad_norm": 1.0344339609146118, "learning_rate": 0.0002, "epoch": 4.647967019642713, "step": 28750}, {"loss": 0.5775, "grad_norm": 1.0513931512832642, "learning_rate": 0.0002, "epoch": 4.649583703823458, "step": 28760}, {"loss": 0.546, "grad_norm": 0.9711475968360901, "learning_rate": 0.0002, "epoch": 4.651200388004203, "step": 28770}, {"loss": 0.5472, "grad_norm": 0.977519690990448, "learning_rate": 0.0002, "epoch": 4.652817072184948, "step": 28780}, {"loss": 0.5826, "grad_norm": 0.9150224924087524, "learning_rate": 0.0002, "epoch": 4.654433756365694, "step": 28790}, {"loss": 0.5382, "grad_norm": 1.0973542928695679, "learning_rate": 0.0002, "epoch": 4.656050440546439, "step": 28800}, {"loss": 0.6147, "grad_norm": 0.944877564907074, "learning_rate": 0.0002, "epoch": 4.657667124727185, "step": 28810}, {"loss": 0.5537, "grad_norm": 0.9508748650550842, "learning_rate": 0.0002, "epoch": 4.659283808907929, "step": 28820}, {"loss": 0.5537, "grad_norm": 0.9681721329689026, "learning_rate": 0.0002, "epoch": 4.6609004930886755, "step": 28830}, {"loss": 0.592, "grad_norm": 1.0214351415634155, "learning_rate": 0.0002, "epoch": 4.662517177269421, "step": 28840}, {"loss": 0.6031, "grad_norm": 0.9748611450195312, "learning_rate": 0.0002, "epoch": 4.664133861450166, "step": 28850}, {"loss": 0.572, "grad_norm": 0.8484147191047668, "learning_rate": 0.0002, "epoch": 4.665750545630911, "step": 28860}, {"loss": 0.5699, "grad_norm": 1.1252986192703247, "learning_rate": 0.0002, "epoch": 4.667367229811656, "step": 28870}, {"loss": 0.5724, "grad_norm": 0.8706206679344177, "learning_rate": 0.0002, "epoch": 4.668983913992402, "step": 28880}, {"loss": 0.6002, "grad_norm": 1.1432424783706665, "learning_rate": 0.0002, "epoch": 4.670600598173147, "step": 28890}, {"loss": 0.5675, "grad_norm": 1.017029047012329, "learning_rate": 0.0002, "epoch": 4.672217282353892, "step": 28900}, {"loss": 0.5831, "grad_norm": 1.085597038269043, "learning_rate": 0.0002, "epoch": 4.673833966534637, "step": 28910}, {"loss": 0.5678, "grad_norm": 0.9275796413421631, "learning_rate": 0.0002, "epoch": 4.675450650715383, "step": 28920}, {"loss": 0.5603, "grad_norm": 0.9518964886665344, "learning_rate": 0.0002, "epoch": 4.677067334896128, "step": 28930}, {"loss": 0.6232, "grad_norm": 1.0352122783660889, "learning_rate": 0.0002, "epoch": 4.678684019076873, "step": 28940}, {"loss": 0.5786, "grad_norm": 1.090124249458313, "learning_rate": 0.0002, "epoch": 4.680300703257618, "step": 28950}, {"loss": 0.5728, "grad_norm": 0.8799563050270081, "learning_rate": 0.0002, "epoch": 4.681917387438364, "step": 28960}, {"loss": 0.5787, "grad_norm": 1.0929821729660034, "learning_rate": 0.0002, "epoch": 4.683534071619109, "step": 28970}, {"loss": 0.6134, "grad_norm": 0.903727650642395, "learning_rate": 0.0002, "epoch": 4.685150755799855, "step": 28980}, {"loss": 0.5522, "grad_norm": 0.9752424955368042, "learning_rate": 0.0002, "epoch": 4.6867674399806, "step": 28990}, {"loss": 0.5762, "grad_norm": 0.9351571202278137, "learning_rate": 0.0002, "epoch": 4.688384124161345, "step": 29000}, {"loss": 0.5811, "grad_norm": 0.923877477645874, "learning_rate": 0.0002, "epoch": 4.6900008083420905, "step": 29010}, {"loss": 0.5682, "grad_norm": 1.045389175415039, "learning_rate": 0.0002, "epoch": 4.691617492522836, "step": 29020}, {"loss": 0.584, "grad_norm": 1.0200831890106201, "learning_rate": 0.0002, "epoch": 4.693234176703581, "step": 29030}, {"loss": 0.5514, "grad_norm": 1.1499706506729126, "learning_rate": 0.0002, "epoch": 4.694850860884326, "step": 29040}, {"loss": 0.5745, "grad_norm": 0.860118567943573, "learning_rate": 0.0002, "epoch": 4.6964675450650715, "step": 29050}, {"loss": 0.5741, "grad_norm": 0.9774864315986633, "learning_rate": 0.0002, "epoch": 4.698084229245817, "step": 29060}, {"loss": 0.5765, "grad_norm": 1.0323210954666138, "learning_rate": 0.0002, "epoch": 4.699700913426562, "step": 29070}, {"loss": 0.5452, "grad_norm": 0.8492481112480164, "learning_rate": 0.0002, "epoch": 4.701317597607307, "step": 29080}, {"loss": 0.5985, "grad_norm": 1.131951093673706, "learning_rate": 0.0002, "epoch": 4.702934281788052, "step": 29090}, {"loss": 0.6412, "grad_norm": 0.8763113021850586, "learning_rate": 0.0002, "epoch": 4.704550965968798, "step": 29100}, {"loss": 0.575, "grad_norm": 1.045028805732727, "learning_rate": 0.0002, "epoch": 4.706167650149544, "step": 29110}, {"loss": 0.5548, "grad_norm": 0.9961401224136353, "learning_rate": 0.0002, "epoch": 4.707784334330288, "step": 29120}, {"loss": 0.559, "grad_norm": 0.9282503724098206, "learning_rate": 0.0002, "epoch": 4.709401018511034, "step": 29130}, {"loss": 0.5744, "grad_norm": 1.1418932676315308, "learning_rate": 0.0002, "epoch": 4.711017702691779, "step": 29140}, {"loss": 0.5394, "grad_norm": 0.9950099587440491, "learning_rate": 0.0002, "epoch": 4.712634386872525, "step": 29150}, {"loss": 0.6177, "grad_norm": 0.8304893374443054, "learning_rate": 0.0002, "epoch": 4.71425107105327, "step": 29160}, {"loss": 0.6074, "grad_norm": 1.115626335144043, "learning_rate": 0.0002, "epoch": 4.715867755234015, "step": 29170}, {"loss": 0.6265, "grad_norm": 1.079818606376648, "learning_rate": 0.0002, "epoch": 4.71748443941476, "step": 29180}, {"loss": 0.561, "grad_norm": 1.1929082870483398, "learning_rate": 0.0002, "epoch": 4.719101123595506, "step": 29190}, {"loss": 0.5708, "grad_norm": 0.9621080756187439, "learning_rate": 0.0002, "epoch": 4.720717807776251, "step": 29200}, {"loss": 0.546, "grad_norm": 0.8549222350120544, "learning_rate": 0.0002, "epoch": 4.722334491956996, "step": 29210}, {"loss": 0.5775, "grad_norm": 0.9341941475868225, "learning_rate": 0.0002, "epoch": 4.723951176137741, "step": 29220}, {"loss": 0.5436, "grad_norm": 1.075406789779663, "learning_rate": 0.0002, "epoch": 4.7255678603184865, "step": 29230}, {"loss": 0.576, "grad_norm": 1.0859880447387695, "learning_rate": 0.0002, "epoch": 4.727184544499232, "step": 29240}, {"loss": 0.5525, "grad_norm": 0.8475605249404907, "learning_rate": 0.0002, "epoch": 4.728801228679977, "step": 29250}, {"loss": 0.5659, "grad_norm": 0.9331845641136169, "learning_rate": 0.0002, "epoch": 4.730417912860723, "step": 29260}, {"loss": 0.5901, "grad_norm": 0.9279314279556274, "learning_rate": 0.0002, "epoch": 4.7320345970414674, "step": 29270}, {"loss": 0.597, "grad_norm": 0.7803558707237244, "learning_rate": 0.0002, "epoch": 4.733651281222214, "step": 29280}, {"loss": 0.5968, "grad_norm": 1.0159329175949097, "learning_rate": 0.0002, "epoch": 4.735267965402959, "step": 29290}, {"loss": 0.5333, "grad_norm": 0.9448670744895935, "learning_rate": 0.0002, "epoch": 4.736884649583704, "step": 29300}, {"loss": 0.574, "grad_norm": 1.0732197761535645, "learning_rate": 0.0002, "epoch": 4.738501333764449, "step": 29310}, {"loss": 0.6066, "grad_norm": 0.901830792427063, "learning_rate": 0.0002, "epoch": 4.7401180179451945, "step": 29320}, {"loss": 0.6105, "grad_norm": 0.9141789674758911, "learning_rate": 0.0002, "epoch": 4.74173470212594, "step": 29330}, {"loss": 0.5481, "grad_norm": 0.9733418226242065, "learning_rate": 0.0002, "epoch": 4.743351386306685, "step": 29340}, {"loss": 0.612, "grad_norm": 0.909810483455658, "learning_rate": 0.0002, "epoch": 4.74496807048743, "step": 29350}, {"loss": 0.5911, "grad_norm": 0.909541666507721, "learning_rate": 0.0002, "epoch": 4.746584754668175, "step": 29360}, {"loss": 0.5579, "grad_norm": 0.9383015632629395, "learning_rate": 0.0002, "epoch": 4.748201438848921, "step": 29370}, {"loss": 0.5529, "grad_norm": 0.9275668263435364, "learning_rate": 0.0002, "epoch": 4.749818123029666, "step": 29380}, {"loss": 0.5623, "grad_norm": 1.1146225929260254, "learning_rate": 0.0002, "epoch": 4.751434807210411, "step": 29390}, {"loss": 0.6018, "grad_norm": 1.0062453746795654, "learning_rate": 0.0002, "epoch": 4.753051491391156, "step": 29400}, {"loss": 0.5872, "grad_norm": 0.9451895952224731, "learning_rate": 0.0002, "epoch": 4.7546681755719025, "step": 29410}, {"loss": 0.5767, "grad_norm": 0.870457649230957, "learning_rate": 0.0002, "epoch": 4.756284859752648, "step": 29420}, {"loss": 0.57, "grad_norm": 1.0411282777786255, "learning_rate": 0.0002, "epoch": 4.757901543933393, "step": 29430}, {"loss": 0.5688, "grad_norm": 1.1648986339569092, "learning_rate": 0.0002, "epoch": 4.759518228114138, "step": 29440}, {"loss": 0.5432, "grad_norm": 0.8999572992324829, "learning_rate": 0.0002, "epoch": 4.761134912294883, "step": 29450}, {"loss": 0.5667, "grad_norm": 0.9863559007644653, "learning_rate": 0.0002, "epoch": 4.762751596475629, "step": 29460}, {"loss": 0.5779, "grad_norm": 0.9676542282104492, "learning_rate": 0.0002, "epoch": 4.764368280656374, "step": 29470}, {"loss": 0.6075, "grad_norm": 1.004775047302246, "learning_rate": 0.0002, "epoch": 4.765984964837119, "step": 29480}, {"loss": 0.6044, "grad_norm": 1.0937515497207642, "learning_rate": 0.0002, "epoch": 4.767601649017864, "step": 29490}, {"loss": 0.5433, "grad_norm": 0.9551598429679871, "learning_rate": 0.0002, "epoch": 4.7692183331986095, "step": 29500}, {"loss": 0.5609, "grad_norm": 1.0757228136062622, "learning_rate": 0.0002, "epoch": 4.770835017379355, "step": 29510}, {"loss": 0.567, "grad_norm": 1.0588841438293457, "learning_rate": 0.0002, "epoch": 4.7724517015601, "step": 29520}, {"loss": 0.5814, "grad_norm": 1.0744032859802246, "learning_rate": 0.0002, "epoch": 4.774068385740845, "step": 29530}, {"loss": 0.5681, "grad_norm": 1.0066277980804443, "learning_rate": 0.0002, "epoch": 4.7756850699215905, "step": 29540}, {"loss": 0.545, "grad_norm": 1.082319736480713, "learning_rate": 0.0002, "epoch": 4.777301754102336, "step": 29550}, {"loss": 0.5709, "grad_norm": 0.8252472877502441, "learning_rate": 0.0002, "epoch": 4.778918438283082, "step": 29560}, {"loss": 0.5666, "grad_norm": 0.9855340123176575, "learning_rate": 0.0002, "epoch": 4.780535122463827, "step": 29570}, {"loss": 0.6117, "grad_norm": 0.9991421699523926, "learning_rate": 0.0002, "epoch": 4.782151806644572, "step": 29580}, {"loss": 0.5966, "grad_norm": 1.316841959953308, "learning_rate": 0.0002, "epoch": 4.7837684908253175, "step": 29590}, {"loss": 0.6102, "grad_norm": 1.1513035297393799, "learning_rate": 0.0002, "epoch": 4.785385175006063, "step": 29600}, {"loss": 0.5785, "grad_norm": 0.9767683744430542, "learning_rate": 0.0002, "epoch": 4.787001859186808, "step": 29610}, {"loss": 0.6037, "grad_norm": 0.9786278605461121, "learning_rate": 0.0002, "epoch": 4.788618543367553, "step": 29620}, {"loss": 0.6108, "grad_norm": 0.8004973530769348, "learning_rate": 0.0002, "epoch": 4.7902352275482984, "step": 29630}, {"loss": 0.5932, "grad_norm": 1.0997767448425293, "learning_rate": 0.0002, "epoch": 4.791851911729044, "step": 29640}, {"loss": 0.5655, "grad_norm": 0.9752856492996216, "learning_rate": 0.0002, "epoch": 4.793468595909789, "step": 29650}, {"loss": 0.5916, "grad_norm": 1.0518392324447632, "learning_rate": 0.0002, "epoch": 4.795085280090534, "step": 29660}, {"loss": 0.6042, "grad_norm": 1.1050055027008057, "learning_rate": 0.0002, "epoch": 4.796701964271279, "step": 29670}, {"loss": 0.6089, "grad_norm": 0.9933857917785645, "learning_rate": 0.0002, "epoch": 4.798318648452025, "step": 29680}, {"loss": 0.6041, "grad_norm": 1.2804018259048462, "learning_rate": 0.0002, "epoch": 4.79993533263277, "step": 29690}, {"loss": 0.636, "grad_norm": 1.0133371353149414, "learning_rate": 0.0002, "epoch": 4.801552016813515, "step": 29700}, {"loss": 0.5662, "grad_norm": 1.080350637435913, "learning_rate": 0.0002, "epoch": 4.803168700994261, "step": 29710}, {"loss": 0.5603, "grad_norm": 0.9986529350280762, "learning_rate": 0.0002, "epoch": 4.804785385175006, "step": 29720}, {"loss": 0.5894, "grad_norm": 0.975665807723999, "learning_rate": 0.0002, "epoch": 4.806402069355752, "step": 29730}, {"loss": 0.6328, "grad_norm": 0.8458138704299927, "learning_rate": 0.0002, "epoch": 4.808018753536497, "step": 29740}, {"loss": 0.5837, "grad_norm": 0.99330073595047, "learning_rate": 0.0002, "epoch": 4.809635437717242, "step": 29750}, {"loss": 0.5507, "grad_norm": 0.898274302482605, "learning_rate": 0.0002, "epoch": 4.811252121897987, "step": 29760}, {"loss": 0.5842, "grad_norm": 1.0504480600357056, "learning_rate": 0.0002, "epoch": 4.812868806078733, "step": 29770}, {"loss": 0.5821, "grad_norm": 0.937919020652771, "learning_rate": 0.0002, "epoch": 4.814485490259478, "step": 29780}, {"loss": 0.5885, "grad_norm": 0.9593307971954346, "learning_rate": 0.0002, "epoch": 4.816102174440223, "step": 29790}, {"loss": 0.578, "grad_norm": 0.9431198835372925, "learning_rate": 0.0002, "epoch": 4.817718858620968, "step": 29800}, {"loss": 0.5739, "grad_norm": 1.2729957103729248, "learning_rate": 0.0002, "epoch": 4.8193355428017135, "step": 29810}, {"loss": 0.6124, "grad_norm": 0.8876838684082031, "learning_rate": 0.0002, "epoch": 4.820952226982459, "step": 29820}, {"loss": 0.5583, "grad_norm": 1.0185000896453857, "learning_rate": 0.0002, "epoch": 4.822568911163204, "step": 29830}, {"loss": 0.5686, "grad_norm": 1.064276099205017, "learning_rate": 0.0002, "epoch": 4.824185595343949, "step": 29840}, {"loss": 0.5698, "grad_norm": 0.9774803519248962, "learning_rate": 0.0002, "epoch": 4.825802279524694, "step": 29850}, {"loss": 0.5533, "grad_norm": 1.131646990776062, "learning_rate": 0.0002, "epoch": 4.8274189637054405, "step": 29860}, {"loss": 0.6371, "grad_norm": 1.081455945968628, "learning_rate": 0.0002, "epoch": 4.829035647886186, "step": 29870}, {"loss": 0.5793, "grad_norm": 0.990538477897644, "learning_rate": 0.0002, "epoch": 4.830652332066931, "step": 29880}, {"loss": 0.5833, "grad_norm": 0.9750600457191467, "learning_rate": 0.0002, "epoch": 4.832269016247676, "step": 29890}, {"loss": 0.619, "grad_norm": 1.0600621700286865, "learning_rate": 0.0002, "epoch": 4.8338857004284215, "step": 29900}, {"loss": 0.5841, "grad_norm": 0.9237320423126221, "learning_rate": 0.0002, "epoch": 4.835502384609167, "step": 29910}, {"loss": 0.5513, "grad_norm": 0.9739177227020264, "learning_rate": 0.0002, "epoch": 4.837119068789912, "step": 29920}, {"loss": 0.587, "grad_norm": 1.128677248954773, "learning_rate": 0.0002, "epoch": 4.838735752970657, "step": 29930}, {"loss": 0.564, "grad_norm": 1.042604923248291, "learning_rate": 0.0002, "epoch": 4.840352437151402, "step": 29940}, {"loss": 0.5885, "grad_norm": 0.849758505821228, "learning_rate": 0.0002, "epoch": 4.841969121332148, "step": 29950}, {"loss": 0.5952, "grad_norm": 1.2809888124465942, "learning_rate": 0.0002, "epoch": 4.843585805512893, "step": 29960}, {"loss": 0.5703, "grad_norm": 1.0177865028381348, "learning_rate": 0.0002, "epoch": 4.845202489693638, "step": 29970}, {"loss": 0.5946, "grad_norm": 1.0026639699935913, "learning_rate": 0.0002, "epoch": 4.846819173874383, "step": 29980}, {"loss": 0.5897, "grad_norm": 0.9679505228996277, "learning_rate": 0.0002, "epoch": 4.8484358580551286, "step": 29990}, {"loss": 0.5621, "grad_norm": 0.8939532041549683, "learning_rate": 0.0002, "epoch": 4.850052542235874, "step": 30000}, {"loss": 0.5852, "grad_norm": 0.9957457780838013, "learning_rate": 0.0002, "epoch": 4.85166922641662, "step": 30010}, {"loss": 0.6117, "grad_norm": 1.1646790504455566, "learning_rate": 0.0002, "epoch": 4.853285910597365, "step": 30020}, {"loss": 0.5711, "grad_norm": 0.8804680705070496, "learning_rate": 0.0002, "epoch": 4.85490259477811, "step": 30030}, {"loss": 0.5397, "grad_norm": 1.161970853805542, "learning_rate": 0.0002, "epoch": 4.856519278958856, "step": 30040}, {"loss": 0.5552, "grad_norm": 0.9081037640571594, "learning_rate": 0.0002, "epoch": 4.858135963139601, "step": 30050}, {"loss": 0.6024, "grad_norm": 0.9402848482131958, "learning_rate": 0.0002, "epoch": 4.859752647320346, "step": 30060}, {"loss": 0.6256, "grad_norm": 0.9023865461349487, "learning_rate": 0.0002, "epoch": 4.861369331501091, "step": 30070}, {"loss": 0.5926, "grad_norm": 1.0173414945602417, "learning_rate": 0.0002, "epoch": 4.8629860156818365, "step": 30080}, {"loss": 0.6274, "grad_norm": 1.084402322769165, "learning_rate": 0.0002, "epoch": 4.864602699862582, "step": 30090}, {"loss": 0.6311, "grad_norm": 0.9577937126159668, "learning_rate": 0.0002, "epoch": 4.866219384043327, "step": 30100}, {"loss": 0.5724, "grad_norm": 0.9807606935501099, "learning_rate": 0.0002, "epoch": 4.867836068224072, "step": 30110}, {"loss": 0.5786, "grad_norm": 0.978784441947937, "learning_rate": 0.0002, "epoch": 4.8694527524048175, "step": 30120}, {"loss": 0.6194, "grad_norm": 0.9762914776802063, "learning_rate": 0.0002, "epoch": 4.871069436585563, "step": 30130}, {"loss": 0.5892, "grad_norm": 0.9404871463775635, "learning_rate": 0.0002, "epoch": 4.872686120766308, "step": 30140}, {"loss": 0.6182, "grad_norm": 1.0069509744644165, "learning_rate": 0.0002, "epoch": 4.874302804947053, "step": 30150}, {"loss": 0.6225, "grad_norm": 1.1770923137664795, "learning_rate": 0.0002, "epoch": 4.875919489127799, "step": 30160}, {"loss": 0.5657, "grad_norm": 1.021210789680481, "learning_rate": 0.0002, "epoch": 4.8775361733085445, "step": 30170}, {"loss": 0.6033, "grad_norm": 0.8512648940086365, "learning_rate": 0.0002, "epoch": 4.87915285748929, "step": 30180}, {"loss": 0.5519, "grad_norm": 0.9345870018005371, "learning_rate": 0.0002, "epoch": 4.880769541670035, "step": 30190}, {"loss": 0.5682, "grad_norm": 1.0224418640136719, "learning_rate": 0.0002, "epoch": 4.88238622585078, "step": 30200}, {"loss": 0.5807, "grad_norm": 1.0316044092178345, "learning_rate": 0.0002, "epoch": 4.884002910031525, "step": 30210}, {"loss": 0.6065, "grad_norm": 1.102437973022461, "learning_rate": 0.0002, "epoch": 4.885619594212271, "step": 30220}, {"loss": 0.586, "grad_norm": 1.0220023393630981, "learning_rate": 0.0002, "epoch": 4.887236278393016, "step": 30230}, {"loss": 0.5781, "grad_norm": 1.0934523344039917, "learning_rate": 0.0002, "epoch": 4.888852962573761, "step": 30240}, {"loss": 0.6313, "grad_norm": 1.264630913734436, "learning_rate": 0.0002, "epoch": 4.890469646754506, "step": 30250}, {"loss": 0.5712, "grad_norm": 1.0999879837036133, "learning_rate": 0.0002, "epoch": 4.892086330935252, "step": 30260}, {"loss": 0.6413, "grad_norm": 0.9124550223350525, "learning_rate": 0.0002, "epoch": 4.893703015115997, "step": 30270}, {"loss": 0.596, "grad_norm": 0.9853624105453491, "learning_rate": 0.0002, "epoch": 4.895319699296742, "step": 30280}, {"loss": 0.595, "grad_norm": 1.0589802265167236, "learning_rate": 0.0002, "epoch": 4.896936383477488, "step": 30290}, {"loss": 0.6129, "grad_norm": 0.8487226366996765, "learning_rate": 0.0002, "epoch": 4.8985530676582325, "step": 30300}, {"loss": 0.5514, "grad_norm": 1.0212191343307495, "learning_rate": 0.0002, "epoch": 4.900169751838979, "step": 30310}, {"loss": 0.5896, "grad_norm": 1.0187491178512573, "learning_rate": 0.0002, "epoch": 4.901786436019724, "step": 30320}, {"loss": 0.5809, "grad_norm": 1.0013091564178467, "learning_rate": 0.0002, "epoch": 4.903403120200469, "step": 30330}, {"loss": 0.5658, "grad_norm": 1.0017542839050293, "learning_rate": 0.0002, "epoch": 4.905019804381214, "step": 30340}, {"loss": 0.6002, "grad_norm": 0.9665151238441467, "learning_rate": 0.0002, "epoch": 4.9066364885619596, "step": 30350}, {"loss": 0.5864, "grad_norm": 0.8774822950363159, "learning_rate": 0.0002, "epoch": 4.908253172742705, "step": 30360}, {"loss": 0.5771, "grad_norm": 0.9449850916862488, "learning_rate": 0.0002, "epoch": 4.90986985692345, "step": 30370}, {"loss": 0.58, "grad_norm": 0.7368341088294983, "learning_rate": 0.0002, "epoch": 4.911486541104195, "step": 30380}, {"loss": 0.5992, "grad_norm": 0.9669167995452881, "learning_rate": 0.0002, "epoch": 4.9131032252849405, "step": 30390}, {"loss": 0.6202, "grad_norm": 1.1227794885635376, "learning_rate": 0.0002, "epoch": 4.914719909465686, "step": 30400}, {"loss": 0.6181, "grad_norm": 0.9884361028671265, "learning_rate": 0.0002, "epoch": 4.916336593646431, "step": 30410}, {"loss": 0.6185, "grad_norm": 0.9949551224708557, "learning_rate": 0.0002, "epoch": 4.917953277827176, "step": 30420}, {"loss": 0.5866, "grad_norm": 0.9491621851921082, "learning_rate": 0.0002, "epoch": 4.919569962007921, "step": 30430}, {"loss": 0.6005, "grad_norm": 0.78848797082901, "learning_rate": 0.0002, "epoch": 4.9211866461886675, "step": 30440}, {"loss": 0.5561, "grad_norm": 1.0693835020065308, "learning_rate": 0.0002, "epoch": 4.922803330369412, "step": 30450}, {"loss": 0.566, "grad_norm": 0.9573729634284973, "learning_rate": 0.0002, "epoch": 4.924420014550158, "step": 30460}, {"loss": 0.6084, "grad_norm": 0.9975152611732483, "learning_rate": 0.0002, "epoch": 4.926036698730903, "step": 30470}, {"loss": 0.5969, "grad_norm": 0.8695693016052246, "learning_rate": 0.0002, "epoch": 4.9276533829116484, "step": 30480}, {"loss": 0.6144, "grad_norm": 1.145394206047058, "learning_rate": 0.0002, "epoch": 4.929270067092394, "step": 30490}, {"loss": 0.5736, "grad_norm": 0.7668989896774292, "learning_rate": 0.0002, "epoch": 4.930886751273139, "step": 30500}, {"loss": 0.6052, "grad_norm": 0.9630151391029358, "learning_rate": 0.0002, "epoch": 4.932503435453884, "step": 30510}, {"loss": 0.6461, "grad_norm": 0.940705418586731, "learning_rate": 0.0002, "epoch": 4.934120119634629, "step": 30520}, {"loss": 0.6326, "grad_norm": 1.3243348598480225, "learning_rate": 0.0002, "epoch": 4.935736803815375, "step": 30530}, {"loss": 0.6174, "grad_norm": 1.004347801208496, "learning_rate": 0.0002, "epoch": 4.93735348799612, "step": 30540}, {"loss": 0.583, "grad_norm": 0.8711541295051575, "learning_rate": 0.0002, "epoch": 4.938970172176865, "step": 30550}, {"loss": 0.599, "grad_norm": 0.8980631828308105, "learning_rate": 0.0002, "epoch": 4.94058685635761, "step": 30560}, {"loss": 0.6024, "grad_norm": 0.8388893604278564, "learning_rate": 0.0002, "epoch": 4.9422035405383555, "step": 30570}, {"loss": 0.6189, "grad_norm": 1.0991183519363403, "learning_rate": 0.0002, "epoch": 4.943820224719101, "step": 30580}, {"loss": 0.5906, "grad_norm": 0.9731075763702393, "learning_rate": 0.0002, "epoch": 4.945436908899847, "step": 30590}, {"loss": 0.5883, "grad_norm": 1.3904452323913574, "learning_rate": 0.0002, "epoch": 4.947053593080591, "step": 30600}, {"loss": 0.5952, "grad_norm": 1.2489882707595825, "learning_rate": 0.0002, "epoch": 4.948670277261337, "step": 30610}, {"loss": 0.5887, "grad_norm": 1.240072250366211, "learning_rate": 0.0002, "epoch": 4.950286961442083, "step": 30620}, {"loss": 0.5762, "grad_norm": 0.9191411733627319, "learning_rate": 0.0002, "epoch": 4.951903645622828, "step": 30630}, {"loss": 0.5597, "grad_norm": 0.8888895511627197, "learning_rate": 0.0002, "epoch": 4.953520329803573, "step": 30640}, {"loss": 0.6594, "grad_norm": 0.9001450538635254, "learning_rate": 0.0002, "epoch": 4.955137013984318, "step": 30650}, {"loss": 0.6047, "grad_norm": 1.053971767425537, "learning_rate": 0.0002, "epoch": 4.9567536981650635, "step": 30660}, {"loss": 0.6107, "grad_norm": 1.2224042415618896, "learning_rate": 0.0002, "epoch": 4.958370382345809, "step": 30670}, {"loss": 0.6211, "grad_norm": 0.8855111598968506, "learning_rate": 0.0002, "epoch": 4.959987066526554, "step": 30680}, {"loss": 0.5764, "grad_norm": 0.9489575624465942, "learning_rate": 0.0002, "epoch": 4.961603750707299, "step": 30690}, {"loss": 0.5371, "grad_norm": 0.9635404944419861, "learning_rate": 0.0002, "epoch": 4.963220434888044, "step": 30700}, {"loss": 0.6043, "grad_norm": 1.1784121990203857, "learning_rate": 0.0002, "epoch": 4.96483711906879, "step": 30710}, {"loss": 0.5803, "grad_norm": 1.0059462785720825, "learning_rate": 0.0002, "epoch": 4.966453803249535, "step": 30720}, {"loss": 0.5759, "grad_norm": 0.9479738473892212, "learning_rate": 0.0002, "epoch": 4.96807048743028, "step": 30730}, {"loss": 0.584, "grad_norm": 1.0624593496322632, "learning_rate": 0.0002, "epoch": 4.969687171611026, "step": 30740}, {"loss": 0.6202, "grad_norm": 1.1429259777069092, "learning_rate": 0.0002, "epoch": 4.971303855791771, "step": 30750}, {"loss": 0.6174, "grad_norm": 0.9102491140365601, "learning_rate": 0.0002, "epoch": 4.972920539972517, "step": 30760}, {"loss": 0.6025, "grad_norm": 1.1262688636779785, "learning_rate": 0.0002, "epoch": 4.974537224153262, "step": 30770}, {"loss": 0.588, "grad_norm": 1.1415393352508545, "learning_rate": 0.0002, "epoch": 4.976153908334007, "step": 30780}, {"loss": 0.5832, "grad_norm": 1.083078384399414, "learning_rate": 0.0002, "epoch": 4.977770592514752, "step": 30790}, {"loss": 0.6025, "grad_norm": 0.964859127998352, "learning_rate": 0.0002, "epoch": 4.979387276695498, "step": 30800}, {"loss": 0.6095, "grad_norm": 0.8704743385314941, "learning_rate": 0.0002, "epoch": 4.981003960876243, "step": 30810}, {"loss": 0.5666, "grad_norm": 1.0714856386184692, "learning_rate": 0.0002, "epoch": 4.982620645056988, "step": 30820}, {"loss": 0.565, "grad_norm": 0.6818771362304688, "learning_rate": 0.0002, "epoch": 4.984237329237733, "step": 30830}, {"loss": 0.5999, "grad_norm": 1.0454156398773193, "learning_rate": 0.0002, "epoch": 4.985854013418479, "step": 30840}, {"loss": 0.5683, "grad_norm": 0.9410776495933533, "learning_rate": 0.0002, "epoch": 4.987470697599224, "step": 30850}, {"loss": 0.5899, "grad_norm": 1.0878902673721313, "learning_rate": 0.0002, "epoch": 4.989087381779969, "step": 30860}, {"loss": 0.5914, "grad_norm": 0.8916727304458618, "learning_rate": 0.0002, "epoch": 4.990704065960714, "step": 30870}, {"loss": 0.6066, "grad_norm": 1.045776128768921, "learning_rate": 0.0002, "epoch": 4.9923207501414595, "step": 30880}, {"loss": 0.5767, "grad_norm": 0.9861903786659241, "learning_rate": 0.0002, "epoch": 4.993937434322206, "step": 30890}, {"loss": 0.6192, "grad_norm": 0.9275050759315491, "learning_rate": 0.0002, "epoch": 4.995554118502951, "step": 30900}, {"loss": 0.6181, "grad_norm": 0.94013911485672, "learning_rate": 0.0002, "epoch": 4.997170802683696, "step": 30910}, {"loss": 0.614, "grad_norm": 0.9771268367767334, "learning_rate": 0.0002, "epoch": 4.998787486864441, "step": 30920}, {"eval_loss": 1.1968598365783691, "eval_runtime": 122.2519, "eval_samples_per_second": 5.996, "eval_steps_per_second": 0.753, "epoch": 4.9999191657909625, "step": 30927}, {"loss": 0.5238, "grad_norm": 0.8021580576896667, "learning_rate": 0.0002, "epoch": 5.0004041710451865, "step": 30930}, {"loss": 0.4984, "grad_norm": 1.0807327032089233, "learning_rate": 0.0002, "epoch": 5.002020855225932, "step": 30940}, {"loss": 0.514, "grad_norm": 1.1638425588607788, "learning_rate": 0.0002, "epoch": 5.003637539406677, "step": 30950}, {"loss": 0.4621, "grad_norm": 1.1700230836868286, "learning_rate": 0.0002, "epoch": 5.005254223587422, "step": 30960}, {"loss": 0.4657, "grad_norm": 0.9053420424461365, "learning_rate": 0.0002, "epoch": 5.0068709077681675, "step": 30970}, {"loss": 0.4865, "grad_norm": 0.9226111769676208, "learning_rate": 0.0002, "epoch": 5.008487591948913, "step": 30980}, {"loss": 0.5011, "grad_norm": 1.238669514656067, "learning_rate": 0.0002, "epoch": 5.010104276129658, "step": 30990}, {"loss": 0.4754, "grad_norm": 1.0668327808380127, "learning_rate": 0.0002, "epoch": 5.011720960310403, "step": 31000}, {"loss": 0.5414, "grad_norm": 1.0903944969177246, "learning_rate": 0.0002, "epoch": 5.013337644491148, "step": 31010}, {"loss": 0.5117, "grad_norm": 1.0763911008834839, "learning_rate": 0.0002, "epoch": 5.014954328671894, "step": 31020}, {"loss": 0.4908, "grad_norm": 1.0108771324157715, "learning_rate": 0.0002, "epoch": 5.016571012852639, "step": 31030}, {"loss": 0.5052, "grad_norm": 0.8816103935241699, "learning_rate": 0.0002, "epoch": 5.018187697033385, "step": 31040}, {"loss": 0.4985, "grad_norm": 1.11434805393219, "learning_rate": 0.0002, "epoch": 5.01980438121413, "step": 31050}, {"loss": 0.5074, "grad_norm": 1.0727789402008057, "learning_rate": 0.0002, "epoch": 5.021421065394875, "step": 31060}, {"loss": 0.4938, "grad_norm": 1.1480379104614258, "learning_rate": 0.0002, "epoch": 5.023037749575621, "step": 31070}, {"loss": 0.491, "grad_norm": 1.0913071632385254, "learning_rate": 0.0002, "epoch": 5.024654433756366, "step": 31080}, {"loss": 0.4896, "grad_norm": 0.9891864657402039, "learning_rate": 0.0002, "epoch": 5.026271117937111, "step": 31090}, {"loss": 0.4965, "grad_norm": 0.9167473912239075, "learning_rate": 0.0002, "epoch": 5.027887802117856, "step": 31100}, {"loss": 0.5098, "grad_norm": 1.2259035110473633, "learning_rate": 0.0002, "epoch": 5.029504486298602, "step": 31110}, {"loss": 0.5206, "grad_norm": 1.1812787055969238, "learning_rate": 0.0002, "epoch": 5.031121170479347, "step": 31120}, {"loss": 0.4725, "grad_norm": 1.0890522003173828, "learning_rate": 0.0002, "epoch": 5.032737854660092, "step": 31130}, {"loss": 0.4768, "grad_norm": 1.0521091222763062, "learning_rate": 0.0002, "epoch": 5.034354538840837, "step": 31140}, {"loss": 0.4718, "grad_norm": 1.1274569034576416, "learning_rate": 0.0002, "epoch": 5.0359712230215825, "step": 31150}, {"loss": 0.4604, "grad_norm": 1.140974998474121, "learning_rate": 0.0002, "epoch": 5.037587907202328, "step": 31160}, {"loss": 0.5077, "grad_norm": 1.1215609312057495, "learning_rate": 0.0002, "epoch": 5.039204591383073, "step": 31170}, {"loss": 0.4746, "grad_norm": 1.0107218027114868, "learning_rate": 0.0002, "epoch": 5.040821275563818, "step": 31180}, {"loss": 0.5126, "grad_norm": 1.0198770761489868, "learning_rate": 0.0002, "epoch": 5.042437959744564, "step": 31190}, {"loss": 0.5004, "grad_norm": 1.1613430976867676, "learning_rate": 0.0002, "epoch": 5.0440546439253096, "step": 31200}, {"loss": 0.5181, "grad_norm": 0.8555458188056946, "learning_rate": 0.0002, "epoch": 5.045671328106055, "step": 31210}, {"loss": 0.4878, "grad_norm": 1.0235545635223389, "learning_rate": 0.0002, "epoch": 5.0472880122868, "step": 31220}, {"loss": 0.499, "grad_norm": 1.0228750705718994, "learning_rate": 0.0002, "epoch": 5.048904696467545, "step": 31230}, {"loss": 0.4544, "grad_norm": 0.8216419816017151, "learning_rate": 0.0002, "epoch": 5.0505213806482905, "step": 31240}, {"loss": 0.4947, "grad_norm": 0.925828218460083, "learning_rate": 0.0002, "epoch": 5.052138064829036, "step": 31250}, {"loss": 0.4835, "grad_norm": 0.9229369759559631, "learning_rate": 0.0002, "epoch": 5.053754749009781, "step": 31260}, {"loss": 0.5136, "grad_norm": 0.9531727433204651, "learning_rate": 0.0002, "epoch": 5.055371433190526, "step": 31270}, {"loss": 0.5161, "grad_norm": 0.7738548517227173, "learning_rate": 0.0002, "epoch": 5.056988117371271, "step": 31280}, {"loss": 0.5166, "grad_norm": 1.0551451444625854, "learning_rate": 0.0002, "epoch": 5.058604801552017, "step": 31290}, {"loss": 0.4953, "grad_norm": 0.9782299399375916, "learning_rate": 0.0002, "epoch": 5.060221485732762, "step": 31300}, {"loss": 0.4776, "grad_norm": 1.0220632553100586, "learning_rate": 0.0002, "epoch": 5.061838169913507, "step": 31310}, {"loss": 0.5117, "grad_norm": 0.9808892607688904, "learning_rate": 0.0002, "epoch": 5.063454854094252, "step": 31320}, {"loss": 0.501, "grad_norm": 1.0662003755569458, "learning_rate": 0.0002, "epoch": 5.065071538274998, "step": 31330}, {"loss": 0.4844, "grad_norm": 1.0036940574645996, "learning_rate": 0.0002, "epoch": 5.066688222455744, "step": 31340}, {"loss": 0.5299, "grad_norm": 1.1931052207946777, "learning_rate": 0.0002, "epoch": 5.068304906636489, "step": 31350}, {"loss": 0.4646, "grad_norm": 0.9370693564414978, "learning_rate": 0.0002, "epoch": 5.069921590817234, "step": 31360}, {"loss": 0.5274, "grad_norm": 0.9589039087295532, "learning_rate": 0.0002, "epoch": 5.071538274997979, "step": 31370}, {"loss": 0.4669, "grad_norm": 1.0052711963653564, "learning_rate": 0.0002, "epoch": 5.073154959178725, "step": 31380}, {"loss": 0.5283, "grad_norm": 0.9991368651390076, "learning_rate": 0.0002, "epoch": 5.07477164335947, "step": 31390}, {"loss": 0.4579, "grad_norm": 0.8539695739746094, "learning_rate": 0.0002, "epoch": 5.076388327540215, "step": 31400}, {"loss": 0.4609, "grad_norm": 1.048775553703308, "learning_rate": 0.0002, "epoch": 5.07800501172096, "step": 31410}, {"loss": 0.4915, "grad_norm": 0.9983724355697632, "learning_rate": 0.0002, "epoch": 5.0796216959017055, "step": 31420}, {"loss": 0.4594, "grad_norm": 1.0189813375473022, "learning_rate": 0.0002, "epoch": 5.081238380082451, "step": 31430}, {"loss": 0.5449, "grad_norm": 0.9781646728515625, "learning_rate": 0.0002, "epoch": 5.082855064263196, "step": 31440}, {"loss": 0.4698, "grad_norm": 0.9424566030502319, "learning_rate": 0.0002, "epoch": 5.084471748443941, "step": 31450}, {"loss": 0.4768, "grad_norm": 1.0036484003067017, "learning_rate": 0.0002, "epoch": 5.0860884326246865, "step": 31460}, {"loss": 0.487, "grad_norm": 1.0983147621154785, "learning_rate": 0.0002, "epoch": 5.087705116805432, "step": 31470}, {"loss": 0.5236, "grad_norm": 1.0856730937957764, "learning_rate": 0.0002, "epoch": 5.089321800986177, "step": 31480}, {"loss": 0.485, "grad_norm": 1.2191699743270874, "learning_rate": 0.0002, "epoch": 5.090938485166923, "step": 31490}, {"loss": 0.4936, "grad_norm": 0.939346194267273, "learning_rate": 0.0002, "epoch": 5.092555169347668, "step": 31500}, {"loss": 0.5107, "grad_norm": 0.9730121493339539, "learning_rate": 0.0002, "epoch": 5.0941718535284135, "step": 31510}, {"loss": 0.4973, "grad_norm": 0.923686146736145, "learning_rate": 0.0002, "epoch": 5.095788537709159, "step": 31520}, {"loss": 0.4906, "grad_norm": 1.1734349727630615, "learning_rate": 0.0002, "epoch": 5.097405221889904, "step": 31530}, {"loss": 0.5165, "grad_norm": 1.084509015083313, "learning_rate": 0.0002, "epoch": 5.099021906070649, "step": 31540}, {"loss": 0.5078, "grad_norm": 1.0144678354263306, "learning_rate": 0.0002, "epoch": 5.100638590251394, "step": 31550}, {"loss": 0.4719, "grad_norm": 0.9958019256591797, "learning_rate": 0.0002, "epoch": 5.10225527443214, "step": 31560}, {"loss": 0.4876, "grad_norm": 0.8900736570358276, "learning_rate": 0.0002, "epoch": 5.103871958612885, "step": 31570}, {"loss": 0.463, "grad_norm": 1.0921649932861328, "learning_rate": 0.0002, "epoch": 5.10548864279363, "step": 31580}, {"loss": 0.5148, "grad_norm": 1.1613792181015015, "learning_rate": 0.0002, "epoch": 5.107105326974375, "step": 31590}, {"loss": 0.5055, "grad_norm": 0.9211367964744568, "learning_rate": 0.0002, "epoch": 5.108722011155121, "step": 31600}, {"loss": 0.5364, "grad_norm": 1.3315813541412354, "learning_rate": 0.0002, "epoch": 5.110338695335866, "step": 31610}, {"loss": 0.5336, "grad_norm": 1.3765019178390503, "learning_rate": 0.0002, "epoch": 5.111955379516611, "step": 31620}, {"loss": 0.4861, "grad_norm": 1.070198893547058, "learning_rate": 0.0002, "epoch": 5.113572063697356, "step": 31630}, {"loss": 0.5046, "grad_norm": 0.947631299495697, "learning_rate": 0.0002, "epoch": 5.115188747878102, "step": 31640}, {"loss": 0.5297, "grad_norm": 1.0197371244430542, "learning_rate": 0.0002, "epoch": 5.116805432058848, "step": 31650}, {"loss": 0.5014, "grad_norm": 0.8647911548614502, "learning_rate": 0.0002, "epoch": 5.118422116239593, "step": 31660}, {"loss": 0.4705, "grad_norm": 0.8944075107574463, "learning_rate": 0.0002, "epoch": 5.120038800420338, "step": 31670}, {"loss": 0.5175, "grad_norm": 1.124497652053833, "learning_rate": 0.0002, "epoch": 5.121655484601083, "step": 31680}, {"loss": 0.5109, "grad_norm": 0.893131673336029, "learning_rate": 0.0002, "epoch": 5.123272168781829, "step": 31690}, {"loss": 0.4937, "grad_norm": 1.0122284889221191, "learning_rate": 0.0002, "epoch": 5.124888852962574, "step": 31700}, {"loss": 0.5522, "grad_norm": 0.9493719935417175, "learning_rate": 0.0002, "epoch": 5.126505537143319, "step": 31710}, {"loss": 0.5031, "grad_norm": 0.9700539112091064, "learning_rate": 0.0002, "epoch": 5.128122221324064, "step": 31720}, {"loss": 0.5126, "grad_norm": 1.111677646636963, "learning_rate": 0.0002, "epoch": 5.1297389055048095, "step": 31730}, {"loss": 0.5272, "grad_norm": 0.8204274773597717, "learning_rate": 0.0002, "epoch": 5.131355589685555, "step": 31740}, {"loss": 0.5029, "grad_norm": 1.1029267311096191, "learning_rate": 0.0002, "epoch": 5.1329722738663, "step": 31750}, {"loss": 0.505, "grad_norm": 1.065575122833252, "learning_rate": 0.0002, "epoch": 5.134588958047045, "step": 31760}, {"loss": 0.502, "grad_norm": 0.8208706974983215, "learning_rate": 0.0002, "epoch": 5.13620564222779, "step": 31770}, {"loss": 0.5352, "grad_norm": 1.0520979166030884, "learning_rate": 0.0002, "epoch": 5.137822326408536, "step": 31780}, {"loss": 0.4911, "grad_norm": 0.8585538268089294, "learning_rate": 0.0002, "epoch": 5.139439010589282, "step": 31790}, {"loss": 0.5159, "grad_norm": 1.1491447687149048, "learning_rate": 0.0002, "epoch": 5.141055694770027, "step": 31800}, {"loss": 0.5157, "grad_norm": 0.9441081285476685, "learning_rate": 0.0002, "epoch": 5.142672378950772, "step": 31810}, {"loss": 0.5383, "grad_norm": 1.4146889448165894, "learning_rate": 0.0002, "epoch": 5.1442890631315175, "step": 31820}, {"loss": 0.5159, "grad_norm": 1.0326547622680664, "learning_rate": 0.0002, "epoch": 5.145905747312263, "step": 31830}, {"loss": 0.5348, "grad_norm": 0.9879202842712402, "learning_rate": 0.0002, "epoch": 5.147522431493008, "step": 31840}, {"loss": 0.5083, "grad_norm": 1.0374281406402588, "learning_rate": 0.0002, "epoch": 5.149139115673753, "step": 31850}, {"loss": 0.4827, "grad_norm": 1.181229591369629, "learning_rate": 0.0002, "epoch": 5.150755799854498, "step": 31860}, {"loss": 0.5313, "grad_norm": 1.2078537940979004, "learning_rate": 0.0002, "epoch": 5.152372484035244, "step": 31870}, {"loss": 0.5329, "grad_norm": 0.9599190354347229, "learning_rate": 0.0002, "epoch": 5.153989168215989, "step": 31880}, {"loss": 0.4953, "grad_norm": 1.0378568172454834, "learning_rate": 0.0002, "epoch": 5.155605852396734, "step": 31890}, {"loss": 0.5069, "grad_norm": 0.8746536374092102, "learning_rate": 0.0002, "epoch": 5.157222536577479, "step": 31900}, {"loss": 0.5272, "grad_norm": 1.0232136249542236, "learning_rate": 0.0002, "epoch": 5.1588392207582245, "step": 31910}, {"loss": 0.4844, "grad_norm": 0.9827565550804138, "learning_rate": 0.0002, "epoch": 5.16045590493897, "step": 31920}, {"loss": 0.5029, "grad_norm": 1.342657208442688, "learning_rate": 0.0002, "epoch": 5.162072589119716, "step": 31930}, {"loss": 0.513, "grad_norm": 1.18390691280365, "learning_rate": 0.0002, "epoch": 5.163689273300461, "step": 31940}, {"loss": 0.5267, "grad_norm": 0.996350109577179, "learning_rate": 0.0002, "epoch": 5.165305957481206, "step": 31950}, {"loss": 0.5063, "grad_norm": 0.9710391163825989, "learning_rate": 0.0002, "epoch": 5.166922641661952, "step": 31960}, {"loss": 0.5115, "grad_norm": 1.0264002084732056, "learning_rate": 0.0002, "epoch": 5.168539325842697, "step": 31970}, {"loss": 0.4972, "grad_norm": 1.0028311014175415, "learning_rate": 0.0002, "epoch": 5.170156010023442, "step": 31980}, {"loss": 0.5103, "grad_norm": 1.1078234910964966, "learning_rate": 0.0002, "epoch": 5.171772694204187, "step": 31990}, {"loss": 0.495, "grad_norm": 0.9659610390663147, "learning_rate": 0.0002, "epoch": 5.1733893783849325, "step": 32000}, {"loss": 0.5114, "grad_norm": 0.841986894607544, "learning_rate": 0.0002, "epoch": 5.175006062565678, "step": 32010}, {"loss": 0.48, "grad_norm": 1.095332384109497, "learning_rate": 0.0002, "epoch": 5.176622746746423, "step": 32020}, {"loss": 0.4741, "grad_norm": 1.1242377758026123, "learning_rate": 0.0002, "epoch": 5.178239430927168, "step": 32030}, {"loss": 0.5573, "grad_norm": 0.9872292280197144, "learning_rate": 0.0002, "epoch": 5.179856115107913, "step": 32040}, {"loss": 0.48, "grad_norm": 0.936161994934082, "learning_rate": 0.0002, "epoch": 5.181472799288659, "step": 32050}, {"loss": 0.5093, "grad_norm": 1.166100025177002, "learning_rate": 0.0002, "epoch": 5.183089483469404, "step": 32060}, {"loss": 0.5438, "grad_norm": 1.0764425992965698, "learning_rate": 0.0002, "epoch": 5.184706167650149, "step": 32070}, {"loss": 0.4843, "grad_norm": 1.0480051040649414, "learning_rate": 0.0002, "epoch": 5.186322851830895, "step": 32080}, {"loss": 0.5386, "grad_norm": 1.0874916315078735, "learning_rate": 0.0002, "epoch": 5.1879395360116405, "step": 32090}, {"loss": 0.4975, "grad_norm": 1.0817396640777588, "learning_rate": 0.0002, "epoch": 5.189556220192386, "step": 32100}, {"loss": 0.5177, "grad_norm": 1.054111361503601, "learning_rate": 0.0002, "epoch": 5.191172904373131, "step": 32110}, {"loss": 0.5229, "grad_norm": 0.9655823707580566, "learning_rate": 0.0002, "epoch": 5.192789588553876, "step": 32120}, {"loss": 0.5105, "grad_norm": 1.1384109258651733, "learning_rate": 0.0002, "epoch": 5.194406272734621, "step": 32130}, {"loss": 0.5073, "grad_norm": 1.0149348974227905, "learning_rate": 0.0002, "epoch": 5.196022956915367, "step": 32140}, {"loss": 0.5293, "grad_norm": 1.1084046363830566, "learning_rate": 0.0002, "epoch": 5.197639641096112, "step": 32150}, {"loss": 0.4936, "grad_norm": 1.1209309101104736, "learning_rate": 0.0002, "epoch": 5.199256325276857, "step": 32160}, {"loss": 0.5101, "grad_norm": 1.133089542388916, "learning_rate": 0.0002, "epoch": 5.200873009457602, "step": 32170}, {"loss": 0.5242, "grad_norm": 1.0893020629882812, "learning_rate": 0.0002, "epoch": 5.202489693638348, "step": 32180}, {"loss": 0.4872, "grad_norm": 0.90018630027771, "learning_rate": 0.0002, "epoch": 5.204106377819093, "step": 32190}, {"loss": 0.4999, "grad_norm": 0.977622926235199, "learning_rate": 0.0002, "epoch": 5.205723061999838, "step": 32200}, {"loss": 0.5028, "grad_norm": 1.2940177917480469, "learning_rate": 0.0002, "epoch": 5.207339746180583, "step": 32210}, {"loss": 0.5396, "grad_norm": 1.2131710052490234, "learning_rate": 0.0002, "epoch": 5.2089564303613285, "step": 32220}, {"loss": 0.5189, "grad_norm": 1.0234841108322144, "learning_rate": 0.0002, "epoch": 5.210573114542075, "step": 32230}, {"loss": 0.5424, "grad_norm": 1.157975435256958, "learning_rate": 0.0002, "epoch": 5.21218979872282, "step": 32240}, {"loss": 0.5396, "grad_norm": 1.0381282567977905, "learning_rate": 0.0002, "epoch": 5.213806482903565, "step": 32250}, {"loss": 0.5192, "grad_norm": 1.0125395059585571, "learning_rate": 0.0002, "epoch": 5.21542316708431, "step": 32260}, {"loss": 0.5216, "grad_norm": 1.272691011428833, "learning_rate": 0.0002, "epoch": 5.2170398512650555, "step": 32270}, {"loss": 0.52, "grad_norm": 1.0061250925064087, "learning_rate": 0.0002, "epoch": 5.218656535445801, "step": 32280}, {"loss": 0.4739, "grad_norm": 0.9752234816551208, "learning_rate": 0.0002, "epoch": 5.220273219626546, "step": 32290}, {"loss": 0.5471, "grad_norm": 1.1193140745162964, "learning_rate": 0.0002, "epoch": 5.221889903807291, "step": 32300}, {"loss": 0.4976, "grad_norm": 1.0126434564590454, "learning_rate": 0.0002, "epoch": 5.2235065879880365, "step": 32310}, {"loss": 0.5257, "grad_norm": 1.4338394403457642, "learning_rate": 0.0002, "epoch": 5.225123272168782, "step": 32320}, {"loss": 0.5235, "grad_norm": 1.004101276397705, "learning_rate": 0.0002, "epoch": 5.226739956349527, "step": 32330}, {"loss": 0.5091, "grad_norm": 0.8744166493415833, "learning_rate": 0.0002, "epoch": 5.228356640530272, "step": 32340}, {"loss": 0.5388, "grad_norm": 1.0165376663208008, "learning_rate": 0.0002, "epoch": 5.229973324711017, "step": 32350}, {"loss": 0.5469, "grad_norm": 0.8635954260826111, "learning_rate": 0.0002, "epoch": 5.231590008891763, "step": 32360}, {"loss": 0.5609, "grad_norm": 1.1392399072647095, "learning_rate": 0.0002, "epoch": 5.233206693072509, "step": 32370}, {"loss": 0.5173, "grad_norm": 1.0202113389968872, "learning_rate": 0.0002, "epoch": 5.234823377253254, "step": 32380}, {"loss": 0.4983, "grad_norm": 1.0417983531951904, "learning_rate": 0.0002, "epoch": 5.236440061433999, "step": 32390}, {"loss": 0.507, "grad_norm": 0.8729333877563477, "learning_rate": 0.0002, "epoch": 5.238056745614744, "step": 32400}, {"loss": 0.5426, "grad_norm": 1.1626229286193848, "learning_rate": 0.0002, "epoch": 5.23967342979549, "step": 32410}, {"loss": 0.5355, "grad_norm": 0.9086161851882935, "learning_rate": 0.0002, "epoch": 5.241290113976235, "step": 32420}, {"loss": 0.4927, "grad_norm": 1.3999892473220825, "learning_rate": 0.0002, "epoch": 5.24290679815698, "step": 32430}, {"loss": 0.4795, "grad_norm": 1.0356311798095703, "learning_rate": 0.0002, "epoch": 5.244523482337725, "step": 32440}, {"loss": 0.5035, "grad_norm": 0.9655531644821167, "learning_rate": 0.0002, "epoch": 5.246140166518471, "step": 32450}, {"loss": 0.5166, "grad_norm": 1.0411828756332397, "learning_rate": 0.0002, "epoch": 5.247756850699216, "step": 32460}, {"loss": 0.5141, "grad_norm": 1.1199816465377808, "learning_rate": 0.0002, "epoch": 5.249373534879961, "step": 32470}, {"loss": 0.4864, "grad_norm": 1.260321855545044, "learning_rate": 0.0002, "epoch": 5.250990219060706, "step": 32480}, {"loss": 0.4893, "grad_norm": 1.2950857877731323, "learning_rate": 0.0002, "epoch": 5.2526069032414515, "step": 32490}, {"loss": 0.4952, "grad_norm": 0.8982820510864258, "learning_rate": 0.0002, "epoch": 5.254223587422197, "step": 32500}, {"loss": 0.5138, "grad_norm": 0.8512987494468689, "learning_rate": 0.0002, "epoch": 5.255840271602942, "step": 32510}, {"loss": 0.5341, "grad_norm": 1.067443609237671, "learning_rate": 0.0002, "epoch": 5.257456955783688, "step": 32520}, {"loss": 0.4928, "grad_norm": 1.0957417488098145, "learning_rate": 0.0002, "epoch": 5.259073639964433, "step": 32530}, {"loss": 0.5169, "grad_norm": 1.4161807298660278, "learning_rate": 0.0002, "epoch": 5.260690324145179, "step": 32540}, {"loss": 0.5599, "grad_norm": 1.2264093160629272, "learning_rate": 0.0002, "epoch": 5.262307008325924, "step": 32550}, {"loss": 0.5221, "grad_norm": 1.0015931129455566, "learning_rate": 0.0002, "epoch": 5.263923692506669, "step": 32560}, {"loss": 0.5253, "grad_norm": 1.0743094682693481, "learning_rate": 0.0002, "epoch": 5.265540376687414, "step": 32570}, {"loss": 0.5289, "grad_norm": 1.1386840343475342, "learning_rate": 0.0002, "epoch": 5.2671570608681595, "step": 32580}, {"loss": 0.5315, "grad_norm": 1.0093860626220703, "learning_rate": 0.0002, "epoch": 5.268773745048905, "step": 32590}, {"loss": 0.5175, "grad_norm": 0.9593744874000549, "learning_rate": 0.0002, "epoch": 5.27039042922965, "step": 32600}, {"loss": 0.528, "grad_norm": 1.146021842956543, "learning_rate": 0.0002, "epoch": 5.272007113410395, "step": 32610}, {"loss": 0.4983, "grad_norm": 0.9579031467437744, "learning_rate": 0.0002, "epoch": 5.27362379759114, "step": 32620}, {"loss": 0.5376, "grad_norm": 1.0548793077468872, "learning_rate": 0.0002, "epoch": 5.275240481771886, "step": 32630}, {"loss": 0.5267, "grad_norm": 1.0380561351776123, "learning_rate": 0.0002, "epoch": 5.276857165952631, "step": 32640}, {"loss": 0.5182, "grad_norm": 1.2119969129562378, "learning_rate": 0.0002, "epoch": 5.278473850133376, "step": 32650}, {"loss": 0.5298, "grad_norm": 1.0507797002792358, "learning_rate": 0.0002, "epoch": 5.280090534314121, "step": 32660}, {"loss": 0.5253, "grad_norm": 1.0185176134109497, "learning_rate": 0.0002, "epoch": 5.2817072184948675, "step": 32670}, {"loss": 0.4904, "grad_norm": 1.2358098030090332, "learning_rate": 0.0002, "epoch": 5.283323902675613, "step": 32680}, {"loss": 0.5169, "grad_norm": 0.7937114238739014, "learning_rate": 0.0002, "epoch": 5.284940586856358, "step": 32690}, {"loss": 0.495, "grad_norm": 0.9825124740600586, "learning_rate": 0.0002, "epoch": 5.286557271037103, "step": 32700}, {"loss": 0.5149, "grad_norm": 1.2059301137924194, "learning_rate": 0.0002, "epoch": 5.288173955217848, "step": 32710}, {"loss": 0.5272, "grad_norm": 1.0828571319580078, "learning_rate": 0.0002, "epoch": 5.289790639398594, "step": 32720}, {"loss": 0.5383, "grad_norm": 1.0129735469818115, "learning_rate": 0.0002, "epoch": 5.291407323579339, "step": 32730}, {"loss": 0.5216, "grad_norm": 1.0591634511947632, "learning_rate": 0.0002, "epoch": 5.293024007760084, "step": 32740}, {"loss": 0.522, "grad_norm": 0.9256815910339355, "learning_rate": 0.0002, "epoch": 5.294640691940829, "step": 32750}, {"loss": 0.5396, "grad_norm": 1.0928633213043213, "learning_rate": 0.0002, "epoch": 5.2962573761215745, "step": 32760}, {"loss": 0.5093, "grad_norm": 0.9415594935417175, "learning_rate": 0.0002, "epoch": 5.29787406030232, "step": 32770}, {"loss": 0.5252, "grad_norm": 1.141316294670105, "learning_rate": 0.0002, "epoch": 5.299490744483065, "step": 32780}, {"loss": 0.4837, "grad_norm": 1.0646510124206543, "learning_rate": 0.0002, "epoch": 5.30110742866381, "step": 32790}, {"loss": 0.5547, "grad_norm": 1.189661979675293, "learning_rate": 0.0002, "epoch": 5.3027241128445555, "step": 32800}, {"loss": 0.5664, "grad_norm": 0.9568731188774109, "learning_rate": 0.0002, "epoch": 5.304340797025301, "step": 32810}, {"loss": 0.5344, "grad_norm": 1.1556824445724487, "learning_rate": 0.0002, "epoch": 5.305957481206047, "step": 32820}, {"loss": 0.4894, "grad_norm": 0.9353463649749756, "learning_rate": 0.0002, "epoch": 5.307574165386792, "step": 32830}, {"loss": 0.5052, "grad_norm": 1.1208295822143555, "learning_rate": 0.0002, "epoch": 5.309190849567537, "step": 32840}, {"loss": 0.5126, "grad_norm": 1.0894153118133545, "learning_rate": 0.0002, "epoch": 5.3108075337482825, "step": 32850}, {"loss": 0.5046, "grad_norm": 1.090329647064209, "learning_rate": 0.0002, "epoch": 5.312424217929028, "step": 32860}, {"loss": 0.5237, "grad_norm": 1.0781712532043457, "learning_rate": 0.0002, "epoch": 5.314040902109773, "step": 32870}, {"loss": 0.57, "grad_norm": 1.1785295009613037, "learning_rate": 0.0002, "epoch": 5.315657586290518, "step": 32880}, {"loss": 0.4953, "grad_norm": 1.0406851768493652, "learning_rate": 0.0002, "epoch": 5.317274270471263, "step": 32890}, {"loss": 0.514, "grad_norm": 1.0982953310012817, "learning_rate": 0.0002, "epoch": 5.318890954652009, "step": 32900}, {"loss": 0.4944, "grad_norm": 1.2969383001327515, "learning_rate": 0.0002, "epoch": 5.320507638832754, "step": 32910}, {"loss": 0.4786, "grad_norm": 0.9687288999557495, "learning_rate": 0.0002, "epoch": 5.322124323013499, "step": 32920}, {"loss": 0.5286, "grad_norm": 1.136760950088501, "learning_rate": 0.0002, "epoch": 5.323741007194244, "step": 32930}, {"loss": 0.5321, "grad_norm": 1.3045495748519897, "learning_rate": 0.0002, "epoch": 5.32535769137499, "step": 32940}, {"loss": 0.5413, "grad_norm": 1.221675992012024, "learning_rate": 0.0002, "epoch": 5.326974375555735, "step": 32950}, {"loss": 0.4999, "grad_norm": 1.1380633115768433, "learning_rate": 0.0002, "epoch": 5.32859105973648, "step": 32960}, {"loss": 0.5037, "grad_norm": 1.1065956354141235, "learning_rate": 0.0002, "epoch": 5.330207743917226, "step": 32970}, {"loss": 0.4913, "grad_norm": 1.0187175273895264, "learning_rate": 0.0002, "epoch": 5.331824428097971, "step": 32980}, {"loss": 0.5234, "grad_norm": 0.9077118039131165, "learning_rate": 0.0002, "epoch": 5.333441112278717, "step": 32990}, {"loss": 0.5071, "grad_norm": 1.0092815160751343, "learning_rate": 0.0002, "epoch": 5.335057796459462, "step": 33000}, {"loss": 0.498, "grad_norm": 1.0168777704238892, "learning_rate": 0.0002, "epoch": 5.336674480640207, "step": 33010}, {"loss": 0.4952, "grad_norm": 0.996161937713623, "learning_rate": 0.0002, "epoch": 5.338291164820952, "step": 33020}, {"loss": 0.5024, "grad_norm": 0.794463038444519, "learning_rate": 0.0002, "epoch": 5.339907849001698, "step": 33030}, {"loss": 0.5112, "grad_norm": 0.9750674962997437, "learning_rate": 0.0002, "epoch": 5.341524533182443, "step": 33040}, {"loss": 0.528, "grad_norm": 1.2770029306411743, "learning_rate": 0.0002, "epoch": 5.343141217363188, "step": 33050}, {"loss": 0.52, "grad_norm": 1.1500186920166016, "learning_rate": 0.0002, "epoch": 5.344757901543933, "step": 33060}, {"loss": 0.4906, "grad_norm": 1.0726377964019775, "learning_rate": 0.0002, "epoch": 5.3463745857246785, "step": 33070}, {"loss": 0.5212, "grad_norm": 0.9314153790473938, "learning_rate": 0.0002, "epoch": 5.347991269905424, "step": 33080}, {"loss": 0.5434, "grad_norm": 1.344988465309143, "learning_rate": 0.0002, "epoch": 5.349607954086169, "step": 33090}, {"loss": 0.4874, "grad_norm": 0.863196611404419, "learning_rate": 0.0002, "epoch": 5.351224638266914, "step": 33100}, {"loss": 0.534, "grad_norm": 1.128100037574768, "learning_rate": 0.0002, "epoch": 5.352841322447659, "step": 33110}, {"loss": 0.5293, "grad_norm": 1.1673583984375, "learning_rate": 0.0002, "epoch": 5.3544580066284055, "step": 33120}, {"loss": 0.4787, "grad_norm": 0.9416789412498474, "learning_rate": 0.0002, "epoch": 5.356074690809151, "step": 33130}, {"loss": 0.5155, "grad_norm": 1.1855236291885376, "learning_rate": 0.0002, "epoch": 5.357691374989896, "step": 33140}, {"loss": 0.515, "grad_norm": 1.0415170192718506, "learning_rate": 0.0002, "epoch": 5.359308059170641, "step": 33150}, {"loss": 0.545, "grad_norm": 0.9953004121780396, "learning_rate": 0.0002, "epoch": 5.3609247433513865, "step": 33160}, {"loss": 0.5305, "grad_norm": 0.96138596534729, "learning_rate": 0.0002, "epoch": 5.362541427532132, "step": 33170}, {"loss": 0.5064, "grad_norm": 1.341979742050171, "learning_rate": 0.0002, "epoch": 5.364158111712877, "step": 33180}, {"loss": 0.4986, "grad_norm": 1.0136911869049072, "learning_rate": 0.0002, "epoch": 5.365774795893622, "step": 33190}, {"loss": 0.5459, "grad_norm": 0.8685575127601624, "learning_rate": 0.0002, "epoch": 5.367391480074367, "step": 33200}, {"loss": 0.5146, "grad_norm": 0.8833574652671814, "learning_rate": 0.0002, "epoch": 5.369008164255113, "step": 33210}, {"loss": 0.4982, "grad_norm": 0.9123612642288208, "learning_rate": 0.0002, "epoch": 5.370624848435858, "step": 33220}, {"loss": 0.5047, "grad_norm": 1.2720599174499512, "learning_rate": 0.0002, "epoch": 5.372241532616603, "step": 33230}, {"loss": 0.5175, "grad_norm": 1.0596648454666138, "learning_rate": 0.0002, "epoch": 5.373858216797348, "step": 33240}, {"loss": 0.5284, "grad_norm": 1.119701623916626, "learning_rate": 0.0002, "epoch": 5.3754749009780936, "step": 33250}, {"loss": 0.5217, "grad_norm": 1.3000061511993408, "learning_rate": 0.0002, "epoch": 5.377091585158839, "step": 33260}, {"loss": 0.5125, "grad_norm": 1.083891749382019, "learning_rate": 0.0002, "epoch": 5.378708269339585, "step": 33270}, {"loss": 0.5065, "grad_norm": 0.9402718544006348, "learning_rate": 0.0002, "epoch": 5.38032495352033, "step": 33280}, {"loss": 0.5559, "grad_norm": 1.3376892805099487, "learning_rate": 0.0002, "epoch": 5.381941637701075, "step": 33290}, {"loss": 0.5193, "grad_norm": 1.1600074768066406, "learning_rate": 0.0002, "epoch": 5.383558321881821, "step": 33300}, {"loss": 0.4907, "grad_norm": 1.1449427604675293, "learning_rate": 0.0002, "epoch": 5.385175006062566, "step": 33310}, {"loss": 0.5449, "grad_norm": 1.3118891716003418, "learning_rate": 0.0002, "epoch": 5.386791690243311, "step": 33320}, {"loss": 0.547, "grad_norm": 0.743449866771698, "learning_rate": 0.0002, "epoch": 5.388408374424056, "step": 33330}, {"loss": 0.5555, "grad_norm": 0.9358304142951965, "learning_rate": 0.0002, "epoch": 5.3900250586048015, "step": 33340}, {"loss": 0.5558, "grad_norm": 1.0447142124176025, "learning_rate": 0.0002, "epoch": 5.391641742785547, "step": 33350}, {"loss": 0.5106, "grad_norm": 1.1088626384735107, "learning_rate": 0.0002, "epoch": 5.393258426966292, "step": 33360}, {"loss": 0.4929, "grad_norm": 1.1267958879470825, "learning_rate": 0.0002, "epoch": 5.394875111147037, "step": 33370}, {"loss": 0.5165, "grad_norm": 0.9709370136260986, "learning_rate": 0.0002, "epoch": 5.3964917953277824, "step": 33380}, {"loss": 0.5206, "grad_norm": 1.0939103364944458, "learning_rate": 0.0002, "epoch": 5.398108479508528, "step": 33390}, {"loss": 0.5177, "grad_norm": 0.9559304714202881, "learning_rate": 0.0002, "epoch": 5.399725163689273, "step": 33400}, {"loss": 0.5064, "grad_norm": 1.199580430984497, "learning_rate": 0.0002, "epoch": 5.401341847870018, "step": 33410}, {"loss": 0.52, "grad_norm": 0.9097000360488892, "learning_rate": 0.0002, "epoch": 5.402958532050764, "step": 33420}, {"loss": 0.514, "grad_norm": 1.1940981149673462, "learning_rate": 0.0002, "epoch": 5.4045752162315095, "step": 33430}, {"loss": 0.5069, "grad_norm": 1.0530916452407837, "learning_rate": 0.0002, "epoch": 5.406191900412255, "step": 33440}, {"loss": 0.5482, "grad_norm": 1.0482549667358398, "learning_rate": 0.0002, "epoch": 5.407808584593, "step": 33450}, {"loss": 0.501, "grad_norm": 1.2524714469909668, "learning_rate": 0.0002, "epoch": 5.409425268773745, "step": 33460}, {"loss": 0.5597, "grad_norm": 1.1091666221618652, "learning_rate": 0.0002, "epoch": 5.41104195295449, "step": 33470}, {"loss": 0.546, "grad_norm": 0.9981587529182434, "learning_rate": 0.0002, "epoch": 5.412658637135236, "step": 33480}, {"loss": 0.4977, "grad_norm": 1.016681432723999, "learning_rate": 0.0002, "epoch": 5.414275321315981, "step": 33490}, {"loss": 0.5388, "grad_norm": 1.1456854343414307, "learning_rate": 0.0002, "epoch": 5.415892005496726, "step": 33500}, {"loss": 0.5292, "grad_norm": 1.1454259157180786, "learning_rate": 0.0002, "epoch": 5.417508689677471, "step": 33510}, {"loss": 0.5061, "grad_norm": 0.9858416318893433, "learning_rate": 0.0002, "epoch": 5.419125373858217, "step": 33520}, {"loss": 0.5139, "grad_norm": 0.9764766693115234, "learning_rate": 0.0002, "epoch": 5.420742058038962, "step": 33530}, {"loss": 0.5518, "grad_norm": 1.199920892715454, "learning_rate": 0.0002, "epoch": 5.422358742219707, "step": 33540}, {"loss": 0.5182, "grad_norm": 1.3107370138168335, "learning_rate": 0.0002, "epoch": 5.423975426400452, "step": 33550}, {"loss": 0.5149, "grad_norm": 0.9637970328330994, "learning_rate": 0.0002, "epoch": 5.4255921105811975, "step": 33560}, {"loss": 0.526, "grad_norm": 1.023359775543213, "learning_rate": 0.0002, "epoch": 5.427208794761944, "step": 33570}, {"loss": 0.5206, "grad_norm": 1.060417652130127, "learning_rate": 0.0002, "epoch": 5.428825478942689, "step": 33580}, {"loss": 0.5052, "grad_norm": 0.9971120953559875, "learning_rate": 0.0002, "epoch": 5.430442163123434, "step": 33590}, {"loss": 0.5044, "grad_norm": 0.9213743209838867, "learning_rate": 0.0002, "epoch": 5.432058847304179, "step": 33600}, {"loss": 0.5714, "grad_norm": 1.1512309312820435, "learning_rate": 0.0002, "epoch": 5.4336755314849245, "step": 33610}, {"loss": 0.5317, "grad_norm": 1.2198847532272339, "learning_rate": 0.0002, "epoch": 5.43529221566567, "step": 33620}, {"loss": 0.5237, "grad_norm": 1.0329595804214478, "learning_rate": 0.0002, "epoch": 5.436908899846415, "step": 33630}, {"loss": 0.5364, "grad_norm": 1.1075750589370728, "learning_rate": 0.0002, "epoch": 5.43852558402716, "step": 33640}, {"loss": 0.5295, "grad_norm": 1.006342887878418, "learning_rate": 0.0002, "epoch": 5.4401422682079055, "step": 33650}, {"loss": 0.5394, "grad_norm": 0.9179885983467102, "learning_rate": 0.0002, "epoch": 5.441758952388651, "step": 33660}, {"loss": 0.5124, "grad_norm": 1.2799493074417114, "learning_rate": 0.0002, "epoch": 5.443375636569396, "step": 33670}, {"loss": 0.5426, "grad_norm": 1.1153863668441772, "learning_rate": 0.0002, "epoch": 5.444992320750141, "step": 33680}, {"loss": 0.5087, "grad_norm": 1.0681028366088867, "learning_rate": 0.0002, "epoch": 5.446609004930886, "step": 33690}, {"loss": 0.5272, "grad_norm": 0.9788817167282104, "learning_rate": 0.0002, "epoch": 5.448225689111632, "step": 33700}, {"loss": 0.5308, "grad_norm": 0.8481608629226685, "learning_rate": 0.0002, "epoch": 5.449842373292377, "step": 33710}, {"loss": 0.5225, "grad_norm": 1.113756537437439, "learning_rate": 0.0002, "epoch": 5.451459057473123, "step": 33720}, {"loss": 0.5213, "grad_norm": 0.8425475358963013, "learning_rate": 0.0002, "epoch": 5.453075741653868, "step": 33730}, {"loss": 0.571, "grad_norm": 1.0852208137512207, "learning_rate": 0.0002, "epoch": 5.4546924258346134, "step": 33740}, {"loss": 0.5535, "grad_norm": 1.1664748191833496, "learning_rate": 0.0002, "epoch": 5.456309110015359, "step": 33750}, {"loss": 0.5419, "grad_norm": 1.217241644859314, "learning_rate": 0.0002, "epoch": 5.457925794196104, "step": 33760}, {"loss": 0.5351, "grad_norm": 1.1572928428649902, "learning_rate": 0.0002, "epoch": 5.459542478376849, "step": 33770}, {"loss": 0.5161, "grad_norm": 1.0437318086624146, "learning_rate": 0.0002, "epoch": 5.461159162557594, "step": 33780}, {"loss": 0.5266, "grad_norm": 0.9807571768760681, "learning_rate": 0.0002, "epoch": 5.46277584673834, "step": 33790}, {"loss": 0.5384, "grad_norm": 1.1436342000961304, "learning_rate": 0.0002, "epoch": 5.464392530919085, "step": 33800}, {"loss": 0.5338, "grad_norm": 1.1004794836044312, "learning_rate": 0.0002, "epoch": 5.46600921509983, "step": 33810}, {"loss": 0.4868, "grad_norm": 1.2130268812179565, "learning_rate": 0.0002, "epoch": 5.467625899280575, "step": 33820}, {"loss": 0.516, "grad_norm": 1.3154419660568237, "learning_rate": 0.0002, "epoch": 5.4692425834613205, "step": 33830}, {"loss": 0.4934, "grad_norm": 0.7934383749961853, "learning_rate": 0.0002, "epoch": 5.470859267642066, "step": 33840}, {"loss": 0.5133, "grad_norm": 0.7838410139083862, "learning_rate": 0.0002, "epoch": 5.472475951822812, "step": 33850}, {"loss": 0.4926, "grad_norm": 1.0415139198303223, "learning_rate": 0.0002, "epoch": 5.474092636003557, "step": 33860}, {"loss": 0.5323, "grad_norm": 0.9213164448738098, "learning_rate": 0.0002, "epoch": 5.475709320184302, "step": 33870}, {"loss": 0.5125, "grad_norm": 1.0364776849746704, "learning_rate": 0.0002, "epoch": 5.477326004365048, "step": 33880}, {"loss": 0.5212, "grad_norm": 0.9994072318077087, "learning_rate": 0.0002, "epoch": 5.478942688545793, "step": 33890}, {"loss": 0.5396, "grad_norm": 1.196730136871338, "learning_rate": 0.0002, "epoch": 5.480559372726538, "step": 33900}, {"loss": 0.538, "grad_norm": 0.9955780506134033, "learning_rate": 0.0002, "epoch": 5.482176056907283, "step": 33910}, {"loss": 0.5307, "grad_norm": 1.168188214302063, "learning_rate": 0.0002, "epoch": 5.4837927410880285, "step": 33920}, {"loss": 0.5548, "grad_norm": 1.1816450357437134, "learning_rate": 0.0002, "epoch": 5.485409425268774, "step": 33930}, {"loss": 0.5535, "grad_norm": 1.079715609550476, "learning_rate": 0.0002, "epoch": 5.487026109449519, "step": 33940}, {"loss": 0.5262, "grad_norm": 1.153850793838501, "learning_rate": 0.0002, "epoch": 5.488642793630264, "step": 33950}, {"loss": 0.5248, "grad_norm": 1.0207297801971436, "learning_rate": 0.0002, "epoch": 5.490259477811009, "step": 33960}, {"loss": 0.5142, "grad_norm": 1.1290855407714844, "learning_rate": 0.0002, "epoch": 5.491876161991755, "step": 33970}, {"loss": 0.5168, "grad_norm": 1.068058967590332, "learning_rate": 0.0002, "epoch": 5.4934928461725, "step": 33980}, {"loss": 0.5317, "grad_norm": 0.9789979457855225, "learning_rate": 0.0002, "epoch": 5.495109530353245, "step": 33990}, {"loss": 0.5113, "grad_norm": 0.9696692824363708, "learning_rate": 0.0002, "epoch": 5.496726214533991, "step": 34000}, {"loss": 0.5413, "grad_norm": 1.0539981126785278, "learning_rate": 0.0002, "epoch": 5.4983428987147365, "step": 34010}, {"loss": 0.5783, "grad_norm": 1.0249929428100586, "learning_rate": 0.0002, "epoch": 5.499959582895482, "step": 34020}, {"loss": 0.4888, "grad_norm": 0.9577504992485046, "learning_rate": 0.0002, "epoch": 5.501576267076227, "step": 34030}, {"loss": 0.5291, "grad_norm": 1.0963513851165771, "learning_rate": 0.0002, "epoch": 5.503192951256972, "step": 34040}, {"loss": 0.5315, "grad_norm": 0.8339345455169678, "learning_rate": 0.0002, "epoch": 5.504809635437717, "step": 34050}, {"loss": 0.5191, "grad_norm": 1.0138782262802124, "learning_rate": 0.0002, "epoch": 5.506426319618463, "step": 34060}, {"loss": 0.5463, "grad_norm": 1.0180109739303589, "learning_rate": 0.0002, "epoch": 5.508043003799208, "step": 34070}, {"loss": 0.5083, "grad_norm": 1.2790818214416504, "learning_rate": 0.0002, "epoch": 5.509659687979953, "step": 34080}, {"loss": 0.5195, "grad_norm": 1.428247332572937, "learning_rate": 0.0002, "epoch": 5.511276372160698, "step": 34090}, {"loss": 0.5291, "grad_norm": 1.0926059484481812, "learning_rate": 0.0002, "epoch": 5.5128930563414436, "step": 34100}, {"loss": 0.5665, "grad_norm": 1.2353343963623047, "learning_rate": 0.0002, "epoch": 5.514509740522189, "step": 34110}, {"loss": 0.5331, "grad_norm": 0.935587465763092, "learning_rate": 0.0002, "epoch": 5.516126424702934, "step": 34120}, {"loss": 0.5512, "grad_norm": 0.9767586588859558, "learning_rate": 0.0002, "epoch": 5.517743108883679, "step": 34130}, {"loss": 0.5315, "grad_norm": 1.1660610437393188, "learning_rate": 0.0002, "epoch": 5.5193597930644245, "step": 34140}, {"loss": 0.52, "grad_norm": 0.9828870892524719, "learning_rate": 0.0002, "epoch": 5.520976477245171, "step": 34150}, {"loss": 0.5198, "grad_norm": 1.0097278356552124, "learning_rate": 0.0002, "epoch": 5.522593161425916, "step": 34160}, {"loss": 0.5293, "grad_norm": 1.1766167879104614, "learning_rate": 0.0002, "epoch": 5.524209845606661, "step": 34170}, {"loss": 0.5258, "grad_norm": 0.982292115688324, "learning_rate": 0.0002, "epoch": 5.525826529787406, "step": 34180}, {"loss": 0.5114, "grad_norm": 1.0744609832763672, "learning_rate": 0.0002, "epoch": 5.5274432139681515, "step": 34190}, {"loss": 0.5469, "grad_norm": 1.3831160068511963, "learning_rate": 0.0002, "epoch": 5.529059898148897, "step": 34200}, {"loss": 0.5819, "grad_norm": 1.074771761894226, "learning_rate": 0.0002, "epoch": 5.530676582329642, "step": 34210}, {"loss": 0.5399, "grad_norm": 1.016652226448059, "learning_rate": 0.0002, "epoch": 5.532293266510387, "step": 34220}, {"loss": 0.5158, "grad_norm": 1.2231552600860596, "learning_rate": 0.0002, "epoch": 5.5339099506911325, "step": 34230}, {"loss": 0.5091, "grad_norm": 0.8051198720932007, "learning_rate": 0.0002, "epoch": 5.535526634871878, "step": 34240}, {"loss": 0.5583, "grad_norm": 1.1779674291610718, "learning_rate": 0.0002, "epoch": 5.537143319052623, "step": 34250}, {"loss": 0.5044, "grad_norm": 1.2468291521072388, "learning_rate": 0.0002, "epoch": 5.538760003233368, "step": 34260}, {"loss": 0.523, "grad_norm": 1.14818274974823, "learning_rate": 0.0002, "epoch": 5.540376687414113, "step": 34270}, {"loss": 0.5375, "grad_norm": 1.2362616062164307, "learning_rate": 0.0002, "epoch": 5.541993371594859, "step": 34280}, {"loss": 0.4996, "grad_norm": 1.0206977128982544, "learning_rate": 0.0002, "epoch": 5.543610055775604, "step": 34290}, {"loss": 0.5212, "grad_norm": 1.2018457651138306, "learning_rate": 0.0002, "epoch": 5.54522673995635, "step": 34300}, {"loss": 0.5462, "grad_norm": 1.0349043607711792, "learning_rate": 0.0002, "epoch": 5.546843424137095, "step": 34310}, {"loss": 0.5231, "grad_norm": 1.2022006511688232, "learning_rate": 0.0002, "epoch": 5.54846010831784, "step": 34320}, {"loss": 0.5173, "grad_norm": 1.0810624361038208, "learning_rate": 0.0002, "epoch": 5.550076792498586, "step": 34330}, {"loss": 0.5821, "grad_norm": 1.3297529220581055, "learning_rate": 0.0002, "epoch": 5.551693476679331, "step": 34340}, {"loss": 0.5321, "grad_norm": 0.9722549915313721, "learning_rate": 0.0002, "epoch": 5.553310160860076, "step": 34350}, {"loss": 0.4823, "grad_norm": 0.9903425574302673, "learning_rate": 0.0002, "epoch": 5.554926845040821, "step": 34360}, {"loss": 0.5601, "grad_norm": 0.9568067789077759, "learning_rate": 0.0002, "epoch": 5.556543529221567, "step": 34370}, {"loss": 0.5242, "grad_norm": 1.113870620727539, "learning_rate": 0.0002, "epoch": 5.558160213402312, "step": 34380}, {"loss": 0.5278, "grad_norm": 1.0557632446289062, "learning_rate": 0.0002, "epoch": 5.559776897583057, "step": 34390}, {"loss": 0.5501, "grad_norm": 0.9615673422813416, "learning_rate": 0.0002, "epoch": 5.561393581763802, "step": 34400}, {"loss": 0.5066, "grad_norm": 0.9536027312278748, "learning_rate": 0.0002, "epoch": 5.5630102659445475, "step": 34410}, {"loss": 0.4949, "grad_norm": 0.8808749318122864, "learning_rate": 0.0002, "epoch": 5.564626950125293, "step": 34420}, {"loss": 0.5954, "grad_norm": 1.286132574081421, "learning_rate": 0.0002, "epoch": 5.566243634306038, "step": 34430}, {"loss": 0.5507, "grad_norm": 1.259644865989685, "learning_rate": 0.0002, "epoch": 5.567860318486783, "step": 34440}, {"loss": 0.4922, "grad_norm": 0.9920216798782349, "learning_rate": 0.0002, "epoch": 5.569477002667529, "step": 34450}, {"loss": 0.5527, "grad_norm": 1.182926893234253, "learning_rate": 0.0002, "epoch": 5.5710936868482746, "step": 34460}, {"loss": 0.5185, "grad_norm": 1.1434749364852905, "learning_rate": 0.0002, "epoch": 5.57271037102902, "step": 34470}, {"loss": 0.5256, "grad_norm": 1.2420979738235474, "learning_rate": 0.0002, "epoch": 5.574327055209765, "step": 34480}, {"loss": 0.5039, "grad_norm": 0.9338384866714478, "learning_rate": 0.0002, "epoch": 5.57594373939051, "step": 34490}, {"loss": 0.5634, "grad_norm": 1.0196425914764404, "learning_rate": 0.0002, "epoch": 5.5775604235712555, "step": 34500}, {"loss": 0.5132, "grad_norm": 0.9586997032165527, "learning_rate": 0.0002, "epoch": 5.579177107752001, "step": 34510}, {"loss": 0.5336, "grad_norm": 1.2409086227416992, "learning_rate": 0.0002, "epoch": 5.580793791932746, "step": 34520}, {"loss": 0.5364, "grad_norm": 1.1483757495880127, "learning_rate": 0.0002, "epoch": 5.582410476113491, "step": 34530}, {"loss": 0.5325, "grad_norm": 1.1624305248260498, "learning_rate": 0.0002, "epoch": 5.584027160294236, "step": 34540}, {"loss": 0.5342, "grad_norm": 1.2635223865509033, "learning_rate": 0.0002, "epoch": 5.585643844474982, "step": 34550}, {"loss": 0.4924, "grad_norm": 0.9824051856994629, "learning_rate": 0.0002, "epoch": 5.587260528655727, "step": 34560}, {"loss": 0.5395, "grad_norm": 1.0858620405197144, "learning_rate": 0.0002, "epoch": 5.588877212836472, "step": 34570}, {"loss": 0.5459, "grad_norm": 1.1452655792236328, "learning_rate": 0.0002, "epoch": 5.590493897017217, "step": 34580}, {"loss": 0.5746, "grad_norm": 1.110610842704773, "learning_rate": 0.0002, "epoch": 5.592110581197963, "step": 34590}, {"loss": 0.5285, "grad_norm": 0.9976194500923157, "learning_rate": 0.0002, "epoch": 5.593727265378709, "step": 34600}, {"loss": 0.548, "grad_norm": 1.0698920488357544, "learning_rate": 0.0002, "epoch": 5.595343949559454, "step": 34610}, {"loss": 0.5311, "grad_norm": 1.1505171060562134, "learning_rate": 0.0002, "epoch": 5.596960633740199, "step": 34620}, {"loss": 0.5471, "grad_norm": 1.1014643907546997, "learning_rate": 0.0002, "epoch": 5.598577317920944, "step": 34630}, {"loss": 0.55, "grad_norm": 0.915595293045044, "learning_rate": 0.0002, "epoch": 5.60019400210169, "step": 34640}, {"loss": 0.5821, "grad_norm": 1.1856765747070312, "learning_rate": 0.0002, "epoch": 5.601810686282435, "step": 34650}, {"loss": 0.5502, "grad_norm": 1.1357687711715698, "learning_rate": 0.0002, "epoch": 5.60342737046318, "step": 34660}, {"loss": 0.5034, "grad_norm": 1.0232492685317993, "learning_rate": 0.0002, "epoch": 5.605044054643925, "step": 34670}, {"loss": 0.5357, "grad_norm": 0.9375017881393433, "learning_rate": 0.0002, "epoch": 5.6066607388246705, "step": 34680}, {"loss": 0.5518, "grad_norm": 1.0796529054641724, "learning_rate": 0.0002, "epoch": 5.608277423005416, "step": 34690}, {"loss": 0.5173, "grad_norm": 1.1383336782455444, "learning_rate": 0.0002, "epoch": 5.609894107186161, "step": 34700}, {"loss": 0.5477, "grad_norm": 1.0248544216156006, "learning_rate": 0.0002, "epoch": 5.611510791366906, "step": 34710}, {"loss": 0.5669, "grad_norm": 1.0986040830612183, "learning_rate": 0.0002, "epoch": 5.6131274755476515, "step": 34720}, {"loss": 0.5188, "grad_norm": 1.2689568996429443, "learning_rate": 0.0002, "epoch": 5.614744159728397, "step": 34730}, {"loss": 0.5136, "grad_norm": 1.4044264554977417, "learning_rate": 0.0002, "epoch": 5.616360843909142, "step": 34740}, {"loss": 0.5699, "grad_norm": 1.2084474563598633, "learning_rate": 0.0002, "epoch": 5.617977528089888, "step": 34750}, {"loss": 0.5377, "grad_norm": 1.061248540878296, "learning_rate": 0.0002, "epoch": 5.619594212270633, "step": 34760}, {"loss": 0.5669, "grad_norm": 1.0220764875411987, "learning_rate": 0.0002, "epoch": 5.6212108964513785, "step": 34770}, {"loss": 0.54, "grad_norm": 1.0859092473983765, "learning_rate": 0.0002, "epoch": 5.622827580632124, "step": 34780}, {"loss": 0.5308, "grad_norm": 0.9049732089042664, "learning_rate": 0.0002, "epoch": 5.624444264812869, "step": 34790}, {"loss": 0.5433, "grad_norm": 1.2103937864303589, "learning_rate": 0.0002, "epoch": 5.626060948993614, "step": 34800}, {"loss": 0.5513, "grad_norm": 0.9854230284690857, "learning_rate": 0.0002, "epoch": 5.627677633174359, "step": 34810}, {"loss": 0.5274, "grad_norm": 0.9316635131835938, "learning_rate": 0.0002, "epoch": 5.629294317355105, "step": 34820}, {"loss": 0.5393, "grad_norm": 1.105296015739441, "learning_rate": 0.0002, "epoch": 5.63091100153585, "step": 34830}, {"loss": 0.5527, "grad_norm": 0.993383526802063, "learning_rate": 0.0002, "epoch": 5.632527685716595, "step": 34840}, {"loss": 0.5375, "grad_norm": 1.1544116735458374, "learning_rate": 0.0002, "epoch": 5.63414436989734, "step": 34850}, {"loss": 0.5448, "grad_norm": 1.284475326538086, "learning_rate": 0.0002, "epoch": 5.635761054078086, "step": 34860}, {"loss": 0.5069, "grad_norm": 1.121997594833374, "learning_rate": 0.0002, "epoch": 5.637377738258831, "step": 34870}, {"loss": 0.5335, "grad_norm": 1.213040828704834, "learning_rate": 0.0002, "epoch": 5.638994422439576, "step": 34880}, {"loss": 0.5623, "grad_norm": 1.23222017288208, "learning_rate": 0.0002, "epoch": 5.640611106620321, "step": 34890}, {"loss": 0.5622, "grad_norm": 0.9793637990951538, "learning_rate": 0.0002, "epoch": 5.642227790801067, "step": 34900}, {"loss": 0.5405, "grad_norm": 1.38919997215271, "learning_rate": 0.0002, "epoch": 5.643844474981813, "step": 34910}, {"loss": 0.5007, "grad_norm": 0.8390951156616211, "learning_rate": 0.0002, "epoch": 5.645461159162558, "step": 34920}, {"loss": 0.5974, "grad_norm": 0.9465909004211426, "learning_rate": 0.0002, "epoch": 5.647077843343303, "step": 34930}, {"loss": 0.5264, "grad_norm": 1.066957712173462, "learning_rate": 0.0002, "epoch": 5.648694527524048, "step": 34940}, {"loss": 0.5513, "grad_norm": 0.9842154383659363, "learning_rate": 0.0002, "epoch": 5.650311211704794, "step": 34950}, {"loss": 0.567, "grad_norm": 1.1766440868377686, "learning_rate": 0.0002, "epoch": 5.651927895885539, "step": 34960}, {"loss": 0.5462, "grad_norm": 0.9061306118965149, "learning_rate": 0.0002, "epoch": 5.653544580066284, "step": 34970}, {"loss": 0.5446, "grad_norm": 1.2941309213638306, "learning_rate": 0.0002, "epoch": 5.655161264247029, "step": 34980}, {"loss": 0.5704, "grad_norm": 0.9741247892379761, "learning_rate": 0.0002, "epoch": 5.6567779484277745, "step": 34990}, {"loss": 0.5152, "grad_norm": 1.0784187316894531, "learning_rate": 0.0002, "epoch": 5.65839463260852, "step": 35000}, {"loss": 0.5363, "grad_norm": 0.937889814376831, "learning_rate": 0.0002, "epoch": 5.660011316789265, "step": 35010}, {"loss": 0.5019, "grad_norm": 0.9667879939079285, "learning_rate": 0.0002, "epoch": 5.66162800097001, "step": 35020}, {"loss": 0.5209, "grad_norm": 1.0554876327514648, "learning_rate": 0.0002, "epoch": 5.663244685150756, "step": 35030}, {"loss": 0.523, "grad_norm": 1.2030539512634277, "learning_rate": 0.0002, "epoch": 5.664861369331501, "step": 35040}, {"loss": 0.5406, "grad_norm": 1.0849953889846802, "learning_rate": 0.0002, "epoch": 5.666478053512247, "step": 35050}, {"loss": 0.5747, "grad_norm": 1.1598973274230957, "learning_rate": 0.0002, "epoch": 5.668094737692992, "step": 35060}, {"loss": 0.5488, "grad_norm": 1.0233359336853027, "learning_rate": 0.0002, "epoch": 5.669711421873737, "step": 35070}, {"loss": 0.5409, "grad_norm": 1.1124799251556396, "learning_rate": 0.0002, "epoch": 5.6713281060544825, "step": 35080}, {"loss": 0.5578, "grad_norm": 1.2351475954055786, "learning_rate": 0.0002, "epoch": 5.672944790235228, "step": 35090}, {"loss": 0.5638, "grad_norm": 1.0240728855133057, "learning_rate": 0.0002, "epoch": 5.674561474415973, "step": 35100}, {"loss": 0.5192, "grad_norm": 1.0223692655563354, "learning_rate": 0.0002, "epoch": 5.676178158596718, "step": 35110}, {"loss": 0.524, "grad_norm": 1.4569132328033447, "learning_rate": 0.0002, "epoch": 5.677794842777463, "step": 35120}, {"loss": 0.555, "grad_norm": 0.8983587026596069, "learning_rate": 0.0002, "epoch": 5.679411526958209, "step": 35130}, {"loss": 0.5439, "grad_norm": 1.0775383710861206, "learning_rate": 0.0002, "epoch": 5.681028211138954, "step": 35140}, {"loss": 0.5289, "grad_norm": 0.9800270795822144, "learning_rate": 0.0002, "epoch": 5.682644895319699, "step": 35150}, {"loss": 0.533, "grad_norm": 0.9858237504959106, "learning_rate": 0.0002, "epoch": 5.684261579500444, "step": 35160}, {"loss": 0.5671, "grad_norm": 1.031087040901184, "learning_rate": 0.0002, "epoch": 5.6858782636811895, "step": 35170}, {"loss": 0.5528, "grad_norm": 1.0294365882873535, "learning_rate": 0.0002, "epoch": 5.687494947861936, "step": 35180}, {"loss": 0.5581, "grad_norm": 1.108144760131836, "learning_rate": 0.0002, "epoch": 5.68911163204268, "step": 35190}, {"loss": 0.5373, "grad_norm": 1.0813100337982178, "learning_rate": 0.0002, "epoch": 5.690728316223426, "step": 35200}, {"loss": 0.5429, "grad_norm": 1.3146867752075195, "learning_rate": 0.0002, "epoch": 5.692345000404171, "step": 35210}, {"loss": 0.5297, "grad_norm": 1.16780424118042, "learning_rate": 0.0002, "epoch": 5.693961684584917, "step": 35220}, {"loss": 0.577, "grad_norm": 0.9929125905036926, "learning_rate": 0.0002, "epoch": 5.695578368765662, "step": 35230}, {"loss": 0.5441, "grad_norm": 0.9049441814422607, "learning_rate": 0.0002, "epoch": 5.697195052946407, "step": 35240}, {"loss": 0.5349, "grad_norm": 0.9768866300582886, "learning_rate": 0.0002, "epoch": 5.698811737127152, "step": 35250}, {"loss": 0.542, "grad_norm": 0.8306029438972473, "learning_rate": 0.0002, "epoch": 5.7004284213078975, "step": 35260}, {"loss": 0.4771, "grad_norm": 0.8417280316352844, "learning_rate": 0.0002, "epoch": 5.702045105488643, "step": 35270}, {"loss": 0.574, "grad_norm": 0.9954485893249512, "learning_rate": 0.0002, "epoch": 5.703661789669388, "step": 35280}, {"loss": 0.5469, "grad_norm": 1.2417993545532227, "learning_rate": 0.0002, "epoch": 5.705278473850133, "step": 35290}, {"loss": 0.5275, "grad_norm": 1.1696544885635376, "learning_rate": 0.0002, "epoch": 5.706895158030878, "step": 35300}, {"loss": 0.5188, "grad_norm": 1.2424817085266113, "learning_rate": 0.0002, "epoch": 5.708511842211624, "step": 35310}, {"loss": 0.5595, "grad_norm": 1.1791106462478638, "learning_rate": 0.0002, "epoch": 5.710128526392369, "step": 35320}, {"loss": 0.5076, "grad_norm": 1.202181339263916, "learning_rate": 0.0002, "epoch": 5.711745210573115, "step": 35330}, {"loss": 0.5847, "grad_norm": 1.1006861925125122, "learning_rate": 0.0002, "epoch": 5.713361894753859, "step": 35340}, {"loss": 0.5627, "grad_norm": 1.0918344259262085, "learning_rate": 0.0002, "epoch": 5.7149785789346055, "step": 35350}, {"loss": 0.5677, "grad_norm": 1.0427305698394775, "learning_rate": 0.0002, "epoch": 5.716595263115351, "step": 35360}, {"loss": 0.5288, "grad_norm": 1.0818872451782227, "learning_rate": 0.0002, "epoch": 5.718211947296096, "step": 35370}, {"loss": 0.5296, "grad_norm": 1.186006784439087, "learning_rate": 0.0002, "epoch": 5.719828631476841, "step": 35380}, {"loss": 0.5507, "grad_norm": 1.2073674201965332, "learning_rate": 0.0002, "epoch": 5.721445315657586, "step": 35390}, {"loss": 0.5483, "grad_norm": 1.065338134765625, "learning_rate": 0.0002, "epoch": 5.723061999838332, "step": 35400}, {"loss": 0.5195, "grad_norm": 0.9448973536491394, "learning_rate": 0.0002, "epoch": 5.724678684019077, "step": 35410}, {"loss": 0.5276, "grad_norm": 1.1487499475479126, "learning_rate": 0.0002, "epoch": 5.726295368199822, "step": 35420}, {"loss": 0.5435, "grad_norm": 1.1334216594696045, "learning_rate": 0.0002, "epoch": 5.727912052380567, "step": 35430}, {"loss": 0.5074, "grad_norm": 1.1932826042175293, "learning_rate": 0.0002, "epoch": 5.729528736561313, "step": 35440}, {"loss": 0.5502, "grad_norm": 1.2615786790847778, "learning_rate": 0.0002, "epoch": 5.731145420742058, "step": 35450}, {"loss": 0.5612, "grad_norm": 1.2803694009780884, "learning_rate": 0.0002, "epoch": 5.732762104922803, "step": 35460}, {"loss": 0.5458, "grad_norm": 0.9271906614303589, "learning_rate": 0.0002, "epoch": 5.734378789103548, "step": 35470}, {"loss": 0.5342, "grad_norm": 1.0958917140960693, "learning_rate": 0.0002, "epoch": 5.735995473284294, "step": 35480}, {"loss": 0.538, "grad_norm": 1.1072784662246704, "learning_rate": 0.0002, "epoch": 5.737612157465039, "step": 35490}, {"loss": 0.5683, "grad_norm": 1.1641002893447876, "learning_rate": 0.0002, "epoch": 5.739228841645785, "step": 35500}, {"loss": 0.5252, "grad_norm": 1.0246447324752808, "learning_rate": 0.0002, "epoch": 5.74084552582653, "step": 35510}, {"loss": 0.55, "grad_norm": 1.032474398612976, "learning_rate": 0.0002, "epoch": 5.742462210007275, "step": 35520}, {"loss": 0.4965, "grad_norm": 1.1600854396820068, "learning_rate": 0.0002, "epoch": 5.7440788941880205, "step": 35530}, {"loss": 0.5543, "grad_norm": 1.0686054229736328, "learning_rate": 0.0002, "epoch": 5.745695578368766, "step": 35540}, {"loss": 0.5706, "grad_norm": 1.2314637899398804, "learning_rate": 0.0002, "epoch": 5.747312262549511, "step": 35550}, {"loss": 0.5492, "grad_norm": 0.922134280204773, "learning_rate": 0.0002, "epoch": 5.748928946730256, "step": 35560}, {"loss": 0.5495, "grad_norm": 0.933043360710144, "learning_rate": 0.0002, "epoch": 5.7505456309110015, "step": 35570}, {"loss": 0.5007, "grad_norm": 1.1911931037902832, "learning_rate": 0.0002, "epoch": 5.752162315091747, "step": 35580}, {"loss": 0.5244, "grad_norm": 0.8984857797622681, "learning_rate": 0.0002, "epoch": 5.753778999272492, "step": 35590}, {"loss": 0.5493, "grad_norm": 0.9495107531547546, "learning_rate": 0.0002, "epoch": 5.755395683453237, "step": 35600}, {"loss": 0.5326, "grad_norm": 1.2805472612380981, "learning_rate": 0.0002, "epoch": 5.757012367633982, "step": 35610}, {"loss": 0.5276, "grad_norm": 1.1236625909805298, "learning_rate": 0.0002, "epoch": 5.758629051814728, "step": 35620}, {"loss": 0.6102, "grad_norm": 1.0552798509597778, "learning_rate": 0.0002, "epoch": 5.760245735995474, "step": 35630}, {"loss": 0.5479, "grad_norm": 1.119909644126892, "learning_rate": 0.0002, "epoch": 5.761862420176218, "step": 35640}, {"loss": 0.5282, "grad_norm": 0.8786116242408752, "learning_rate": 0.0002, "epoch": 5.763479104356964, "step": 35650}, {"loss": 0.5406, "grad_norm": 1.2417117357254028, "learning_rate": 0.0002, "epoch": 5.765095788537709, "step": 35660}, {"loss": 0.537, "grad_norm": 1.255200982093811, "learning_rate": 0.0002, "epoch": 5.766712472718455, "step": 35670}, {"loss": 0.5308, "grad_norm": 1.0611358880996704, "learning_rate": 0.0002, "epoch": 5.7683291568992, "step": 35680}, {"loss": 0.5614, "grad_norm": 1.1443911790847778, "learning_rate": 0.0002, "epoch": 5.769945841079945, "step": 35690}, {"loss": 0.5386, "grad_norm": 1.1437989473342896, "learning_rate": 0.0002, "epoch": 5.77156252526069, "step": 35700}, {"loss": 0.537, "grad_norm": 1.1375046968460083, "learning_rate": 0.0002, "epoch": 5.773179209441436, "step": 35710}, {"loss": 0.5198, "grad_norm": 1.0777729749679565, "learning_rate": 0.0002, "epoch": 5.774795893622181, "step": 35720}, {"loss": 0.5521, "grad_norm": 1.1160215139389038, "learning_rate": 0.0002, "epoch": 5.776412577802926, "step": 35730}, {"loss": 0.5569, "grad_norm": 1.1268514394760132, "learning_rate": 0.0002, "epoch": 5.778029261983671, "step": 35740}, {"loss": 0.5311, "grad_norm": 1.2752262353897095, "learning_rate": 0.0002, "epoch": 5.7796459461644165, "step": 35750}, {"loss": 0.5625, "grad_norm": 1.0416184663772583, "learning_rate": 0.0002, "epoch": 5.781262630345162, "step": 35760}, {"loss": 0.5438, "grad_norm": 1.0622444152832031, "learning_rate": 0.0002, "epoch": 5.782879314525907, "step": 35770}, {"loss": 0.5268, "grad_norm": 1.1217877864837646, "learning_rate": 0.0002, "epoch": 5.784495998706653, "step": 35780}, {"loss": 0.5225, "grad_norm": 0.9363139867782593, "learning_rate": 0.0002, "epoch": 5.786112682887398, "step": 35790}, {"loss": 0.5524, "grad_norm": 0.96628737449646, "learning_rate": 0.0002, "epoch": 5.787729367068144, "step": 35800}, {"loss": 0.52, "grad_norm": 0.9572572112083435, "learning_rate": 0.0002, "epoch": 5.789346051248889, "step": 35810}, {"loss": 0.5615, "grad_norm": 0.938724935054779, "learning_rate": 0.0002, "epoch": 5.790962735429634, "step": 35820}, {"loss": 0.5391, "grad_norm": 1.3314417600631714, "learning_rate": 0.0002, "epoch": 5.792579419610379, "step": 35830}, {"loss": 0.5441, "grad_norm": 1.0097602605819702, "learning_rate": 0.0002, "epoch": 5.7941961037911245, "step": 35840}, {"loss": 0.591, "grad_norm": 1.1265122890472412, "learning_rate": 0.0002, "epoch": 5.79581278797187, "step": 35850}, {"loss": 0.5333, "grad_norm": 1.2191909551620483, "learning_rate": 0.0002, "epoch": 5.797429472152615, "step": 35860}, {"loss": 0.5274, "grad_norm": 0.9690808057785034, "learning_rate": 0.0002, "epoch": 5.79904615633336, "step": 35870}, {"loss": 0.5425, "grad_norm": 1.0871665477752686, "learning_rate": 0.0002, "epoch": 5.800662840514105, "step": 35880}, {"loss": 0.5602, "grad_norm": 1.1093597412109375, "learning_rate": 0.0002, "epoch": 5.802279524694851, "step": 35890}, {"loss": 0.5475, "grad_norm": 1.2434282302856445, "learning_rate": 0.0002, "epoch": 5.803896208875596, "step": 35900}, {"loss": 0.5288, "grad_norm": 1.2933623790740967, "learning_rate": 0.0002, "epoch": 5.805512893056341, "step": 35910}, {"loss": 0.5554, "grad_norm": 1.0005441904067993, "learning_rate": 0.0002, "epoch": 5.807129577237086, "step": 35920}, {"loss": 0.5318, "grad_norm": 1.2373108863830566, "learning_rate": 0.0002, "epoch": 5.8087462614178325, "step": 35930}, {"loss": 0.5413, "grad_norm": 1.2622692584991455, "learning_rate": 0.0002, "epoch": 5.810362945598578, "step": 35940}, {"loss": 0.5558, "grad_norm": 1.0112963914871216, "learning_rate": 0.0002, "epoch": 5.811979629779323, "step": 35950}, {"loss": 0.5115, "grad_norm": 1.050572395324707, "learning_rate": 0.0002, "epoch": 5.813596313960068, "step": 35960}, {"loss": 0.5288, "grad_norm": 0.9774560928344727, "learning_rate": 0.0002, "epoch": 5.815212998140813, "step": 35970}, {"loss": 0.585, "grad_norm": 1.19438898563385, "learning_rate": 0.0002, "epoch": 5.816829682321559, "step": 35980}, {"loss": 0.5798, "grad_norm": 1.0267130136489868, "learning_rate": 0.0002, "epoch": 5.818446366502304, "step": 35990}, {"loss": 0.5126, "grad_norm": 0.9813851714134216, "learning_rate": 0.0002, "epoch": 5.820063050683049, "step": 36000}, {"loss": 0.5138, "grad_norm": 0.9177457094192505, "learning_rate": 0.0002, "epoch": 5.821679734863794, "step": 36010}, {"loss": 0.5453, "grad_norm": 1.0020731687545776, "learning_rate": 0.0002, "epoch": 5.8232964190445395, "step": 36020}, {"loss": 0.5646, "grad_norm": 1.073222041130066, "learning_rate": 0.0002, "epoch": 5.824913103225285, "step": 36030}, {"loss": 0.5539, "grad_norm": 1.016337513923645, "learning_rate": 0.0002, "epoch": 5.82652978740603, "step": 36040}, {"loss": 0.5592, "grad_norm": 1.267364263534546, "learning_rate": 0.0002, "epoch": 5.828146471586775, "step": 36050}, {"loss": 0.595, "grad_norm": 1.2730127573013306, "learning_rate": 0.0002, "epoch": 5.8297631557675205, "step": 36060}, {"loss": 0.5247, "grad_norm": 1.108442783355713, "learning_rate": 0.0002, "epoch": 5.831379839948266, "step": 36070}, {"loss": 0.5103, "grad_norm": 1.198072075843811, "learning_rate": 0.0002, "epoch": 5.832996524129012, "step": 36080}, {"loss": 0.5479, "grad_norm": 1.0458786487579346, "learning_rate": 0.0002, "epoch": 5.834613208309757, "step": 36090}, {"loss": 0.5564, "grad_norm": 0.9096664786338806, "learning_rate": 0.0002, "epoch": 5.836229892490502, "step": 36100}, {"loss": 0.5602, "grad_norm": 0.9957793951034546, "learning_rate": 0.0002, "epoch": 5.8378465766712475, "step": 36110}, {"loss": 0.5799, "grad_norm": 1.3693058490753174, "learning_rate": 0.0002, "epoch": 5.839463260851993, "step": 36120}, {"loss": 0.5425, "grad_norm": 1.268608808517456, "learning_rate": 0.0002, "epoch": 5.841079945032738, "step": 36130}, {"loss": 0.5653, "grad_norm": 0.8516020178794861, "learning_rate": 0.0002, "epoch": 5.842696629213483, "step": 36140}, {"loss": 0.5475, "grad_norm": 0.90385502576828, "learning_rate": 0.0002, "epoch": 5.844313313394228, "step": 36150}, {"loss": 0.5274, "grad_norm": 1.0910571813583374, "learning_rate": 0.0002, "epoch": 5.845929997574974, "step": 36160}, {"loss": 0.555, "grad_norm": 0.9417795538902283, "learning_rate": 0.0002, "epoch": 5.847546681755719, "step": 36170}, {"loss": 0.5784, "grad_norm": 1.0027360916137695, "learning_rate": 0.0002, "epoch": 5.849163365936464, "step": 36180}, {"loss": 0.5423, "grad_norm": 1.1480516195297241, "learning_rate": 0.0002, "epoch": 5.850780050117209, "step": 36190}, {"loss": 0.5517, "grad_norm": 1.2431457042694092, "learning_rate": 0.0002, "epoch": 5.852396734297955, "step": 36200}, {"loss": 0.5404, "grad_norm": 1.091465950012207, "learning_rate": 0.0002, "epoch": 5.8540134184787, "step": 36210}, {"loss": 0.53, "grad_norm": 0.9693930745124817, "learning_rate": 0.0002, "epoch": 5.855630102659445, "step": 36220}, {"loss": 0.5453, "grad_norm": 0.9937465190887451, "learning_rate": 0.0002, "epoch": 5.857246786840191, "step": 36230}, {"loss": 0.5621, "grad_norm": 1.0731011629104614, "learning_rate": 0.0002, "epoch": 5.858863471020936, "step": 36240}, {"loss": 0.5687, "grad_norm": 1.0869048833847046, "learning_rate": 0.0002, "epoch": 5.860480155201682, "step": 36250}, {"loss": 0.5576, "grad_norm": 0.9226390719413757, "learning_rate": 0.0002, "epoch": 5.862096839382427, "step": 36260}, {"loss": 0.531, "grad_norm": 1.1755430698394775, "learning_rate": 0.0002, "epoch": 5.863713523563172, "step": 36270}, {"loss": 0.558, "grad_norm": 0.8815974593162537, "learning_rate": 0.0002, "epoch": 5.865330207743917, "step": 36280}, {"loss": 0.5065, "grad_norm": 1.3648751974105835, "learning_rate": 0.0002, "epoch": 5.866946891924663, "step": 36290}, {"loss": 0.536, "grad_norm": 0.8729211091995239, "learning_rate": 0.0002, "epoch": 5.868563576105408, "step": 36300}, {"loss": 0.5192, "grad_norm": 1.0870907306671143, "learning_rate": 0.0002, "epoch": 5.870180260286153, "step": 36310}, {"loss": 0.5609, "grad_norm": 1.1164259910583496, "learning_rate": 0.0002, "epoch": 5.871796944466898, "step": 36320}, {"loss": 0.551, "grad_norm": 1.1572535037994385, "learning_rate": 0.0002, "epoch": 5.8734136286476435, "step": 36330}, {"loss": 0.5898, "grad_norm": 1.0456238985061646, "learning_rate": 0.0002, "epoch": 5.875030312828389, "step": 36340}, {"loss": 0.5008, "grad_norm": 1.1310722827911377, "learning_rate": 0.0002, "epoch": 5.876646997009134, "step": 36350}, {"loss": 0.5352, "grad_norm": 1.0004712343215942, "learning_rate": 0.0002, "epoch": 5.878263681189879, "step": 36360}, {"loss": 0.5632, "grad_norm": 1.0991777181625366, "learning_rate": 0.0002, "epoch": 5.879880365370624, "step": 36370}, {"loss": 0.5815, "grad_norm": 1.2789239883422852, "learning_rate": 0.0002, "epoch": 5.8814970495513705, "step": 36380}, {"loss": 0.56, "grad_norm": 0.9524819850921631, "learning_rate": 0.0002, "epoch": 5.883113733732116, "step": 36390}, {"loss": 0.5701, "grad_norm": 1.1115771532058716, "learning_rate": 0.0002, "epoch": 5.884730417912861, "step": 36400}, {"loss": 0.5463, "grad_norm": 1.37419855594635, "learning_rate": 0.0002, "epoch": 5.886347102093606, "step": 36410}, {"loss": 0.5675, "grad_norm": 1.1449527740478516, "learning_rate": 0.0002, "epoch": 5.8879637862743515, "step": 36420}, {"loss": 0.5255, "grad_norm": 1.198046326637268, "learning_rate": 0.0002, "epoch": 5.889580470455097, "step": 36430}, {"loss": 0.5383, "grad_norm": 1.0180530548095703, "learning_rate": 0.0002, "epoch": 5.891197154635842, "step": 36440}, {"loss": 0.5319, "grad_norm": 1.0516417026519775, "learning_rate": 0.0002, "epoch": 5.892813838816587, "step": 36450}, {"loss": 0.5782, "grad_norm": 1.1658052206039429, "learning_rate": 0.0002, "epoch": 5.894430522997332, "step": 36460}, {"loss": 0.5864, "grad_norm": 1.190699577331543, "learning_rate": 0.0002, "epoch": 5.896047207178078, "step": 36470}, {"loss": 0.5451, "grad_norm": 1.1235495805740356, "learning_rate": 0.0002, "epoch": 5.897663891358823, "step": 36480}, {"loss": 0.5284, "grad_norm": 1.1926926374435425, "learning_rate": 0.0002, "epoch": 5.899280575539568, "step": 36490}, {"loss": 0.5686, "grad_norm": 1.1184662580490112, "learning_rate": 0.0002, "epoch": 5.900897259720313, "step": 36500}, {"loss": 0.5147, "grad_norm": 1.000970721244812, "learning_rate": 0.0002, "epoch": 5.9025139439010585, "step": 36510}, {"loss": 0.5351, "grad_norm": 1.0373306274414062, "learning_rate": 0.0002, "epoch": 5.904130628081804, "step": 36520}, {"loss": 0.535, "grad_norm": 1.0840669870376587, "learning_rate": 0.0002, "epoch": 5.90574731226255, "step": 36530}, {"loss": 0.538, "grad_norm": 0.9908381104469299, "learning_rate": 0.0002, "epoch": 5.907363996443295, "step": 36540}, {"loss": 0.5313, "grad_norm": 1.0456029176712036, "learning_rate": 0.0002, "epoch": 5.90898068062404, "step": 36550}, {"loss": 0.5693, "grad_norm": 1.1381454467773438, "learning_rate": 0.0002, "epoch": 5.910597364804786, "step": 36560}, {"loss": 0.5473, "grad_norm": 0.9440900087356567, "learning_rate": 0.0002, "epoch": 5.912214048985531, "step": 36570}, {"loss": 0.5542, "grad_norm": 1.1674573421478271, "learning_rate": 0.0002, "epoch": 5.913830733166276, "step": 36580}, {"loss": 0.526, "grad_norm": 1.1226966381072998, "learning_rate": 0.0002, "epoch": 5.915447417347021, "step": 36590}, {"loss": 0.6091, "grad_norm": 0.9696915745735168, "learning_rate": 0.0002, "epoch": 5.9170641015277665, "step": 36600}, {"loss": 0.5523, "grad_norm": 0.9593005180358887, "learning_rate": 0.0002, "epoch": 5.918680785708512, "step": 36610}, {"loss": 0.5536, "grad_norm": 1.122169852256775, "learning_rate": 0.0002, "epoch": 5.920297469889257, "step": 36620}, {"loss": 0.5039, "grad_norm": 0.9923415780067444, "learning_rate": 0.0002, "epoch": 5.921914154070002, "step": 36630}, {"loss": 0.5893, "grad_norm": 1.063838005065918, "learning_rate": 0.0002, "epoch": 5.923530838250747, "step": 36640}, {"loss": 0.5799, "grad_norm": 0.9083505272865295, "learning_rate": 0.0002, "epoch": 5.925147522431493, "step": 36650}, {"loss": 0.5264, "grad_norm": 0.9439437985420227, "learning_rate": 0.0002, "epoch": 5.926764206612239, "step": 36660}, {"loss": 0.5891, "grad_norm": 0.9778534173965454, "learning_rate": 0.0002, "epoch": 5.928380890792983, "step": 36670}, {"loss": 0.566, "grad_norm": 0.9723961353302002, "learning_rate": 0.0002, "epoch": 5.929997574973729, "step": 36680}, {"loss": 0.5741, "grad_norm": 1.162333607673645, "learning_rate": 0.0002, "epoch": 5.9316142591544745, "step": 36690}, {"loss": 0.5771, "grad_norm": 1.2784897089004517, "learning_rate": 0.0002, "epoch": 5.93323094333522, "step": 36700}, {"loss": 0.5343, "grad_norm": 1.0924867391586304, "learning_rate": 0.0002, "epoch": 5.934847627515965, "step": 36710}, {"loss": 0.5554, "grad_norm": 1.046922206878662, "learning_rate": 0.0002, "epoch": 5.93646431169671, "step": 36720}, {"loss": 0.5476, "grad_norm": 0.8632535338401794, "learning_rate": 0.0002, "epoch": 5.938080995877455, "step": 36730}, {"loss": 0.5456, "grad_norm": 1.358762502670288, "learning_rate": 0.0002, "epoch": 5.939697680058201, "step": 36740}, {"loss": 0.551, "grad_norm": 1.2058624029159546, "learning_rate": 0.0002, "epoch": 5.941314364238946, "step": 36750}, {"loss": 0.5462, "grad_norm": 1.1396408081054688, "learning_rate": 0.0002, "epoch": 5.942931048419691, "step": 36760}, {"loss": 0.5483, "grad_norm": 1.1510354280471802, "learning_rate": 0.0002, "epoch": 5.944547732600436, "step": 36770}, {"loss": 0.5659, "grad_norm": 1.1401607990264893, "learning_rate": 0.0002, "epoch": 5.946164416781182, "step": 36780}, {"loss": 0.5557, "grad_norm": 1.1871325969696045, "learning_rate": 0.0002, "epoch": 5.947781100961927, "step": 36790}, {"loss": 0.4945, "grad_norm": 0.9928333163261414, "learning_rate": 0.0002, "epoch": 5.949397785142672, "step": 36800}, {"loss": 0.5303, "grad_norm": 1.0549445152282715, "learning_rate": 0.0002, "epoch": 5.951014469323418, "step": 36810}, {"loss": 0.5532, "grad_norm": 0.9791563749313354, "learning_rate": 0.0002, "epoch": 5.9526311535041625, "step": 36820}, {"loss": 0.5317, "grad_norm": 1.1268441677093506, "learning_rate": 0.0002, "epoch": 5.954247837684909, "step": 36830}, {"loss": 0.5585, "grad_norm": 1.0533992052078247, "learning_rate": 0.0002, "epoch": 5.955864521865654, "step": 36840}, {"loss": 0.4972, "grad_norm": 1.023358941078186, "learning_rate": 0.0002, "epoch": 5.957481206046399, "step": 36850}, {"loss": 0.5557, "grad_norm": 1.2631961107254028, "learning_rate": 0.0002, "epoch": 5.959097890227144, "step": 36860}, {"loss": 0.5662, "grad_norm": 0.9397698640823364, "learning_rate": 0.0002, "epoch": 5.9607145744078895, "step": 36870}, {"loss": 0.5775, "grad_norm": 1.1678427457809448, "learning_rate": 0.0002, "epoch": 5.962331258588635, "step": 36880}, {"loss": 0.5435, "grad_norm": 1.1403759717941284, "learning_rate": 0.0002, "epoch": 5.96394794276938, "step": 36890}, {"loss": 0.5479, "grad_norm": 1.030572772026062, "learning_rate": 0.0002, "epoch": 5.965564626950125, "step": 36900}, {"loss": 0.5838, "grad_norm": 1.0992497205734253, "learning_rate": 0.0002, "epoch": 5.9671813111308705, "step": 36910}, {"loss": 0.5452, "grad_norm": 1.075466275215149, "learning_rate": 0.0002, "epoch": 5.968797995311616, "step": 36920}, {"loss": 0.5739, "grad_norm": 1.0153694152832031, "learning_rate": 0.0002, "epoch": 5.970414679492361, "step": 36930}, {"loss": 0.5672, "grad_norm": 0.973193883895874, "learning_rate": 0.0002, "epoch": 5.972031363673106, "step": 36940}, {"loss": 0.5585, "grad_norm": 0.8294678926467896, "learning_rate": 0.0002, "epoch": 5.973648047853851, "step": 36950}, {"loss": 0.5631, "grad_norm": 1.0048716068267822, "learning_rate": 0.0002, "epoch": 5.9752647320345975, "step": 36960}, {"loss": 0.5471, "grad_norm": 0.9714070558547974, "learning_rate": 0.0002, "epoch": 5.976881416215342, "step": 36970}, {"loss": 0.5419, "grad_norm": 0.8667682409286499, "learning_rate": 0.0002, "epoch": 5.978498100396088, "step": 36980}, {"loss": 0.5474, "grad_norm": 1.0461409091949463, "learning_rate": 0.0002, "epoch": 5.980114784576833, "step": 36990}, {"loss": 0.5454, "grad_norm": 0.9229754209518433, "learning_rate": 0.0002, "epoch": 5.981731468757578, "step": 37000}, {"loss": 0.5599, "grad_norm": 1.0406876802444458, "learning_rate": 0.0002, "epoch": 5.983348152938324, "step": 37010}, {"loss": 0.5569, "grad_norm": 0.8993828296661377, "learning_rate": 0.0002, "epoch": 5.984964837119069, "step": 37020}, {"loss": 0.5611, "grad_norm": 1.2260479927062988, "learning_rate": 0.0002, "epoch": 5.986581521299814, "step": 37030}, {"loss": 0.5523, "grad_norm": 1.0107380151748657, "learning_rate": 0.0002, "epoch": 5.988198205480559, "step": 37040}, {"loss": 0.5639, "grad_norm": 1.0240139961242676, "learning_rate": 0.0002, "epoch": 5.989814889661305, "step": 37050}, {"loss": 0.5209, "grad_norm": 1.0185275077819824, "learning_rate": 0.0002, "epoch": 5.99143157384205, "step": 37060}, {"loss": 0.5114, "grad_norm": 1.1361802816390991, "learning_rate": 0.0002, "epoch": 5.993048258022795, "step": 37070}, {"loss": 0.5692, "grad_norm": 1.0395532846450806, "learning_rate": 0.0002, "epoch": 5.99466494220354, "step": 37080}, {"loss": 0.594, "grad_norm": 0.9463558197021484, "learning_rate": 0.0002, "epoch": 5.9962816263842855, "step": 37090}, {"loss": 0.5775, "grad_norm": 1.2066948413848877, "learning_rate": 0.0002, "epoch": 5.997898310565031, "step": 37100}, {"loss": 0.5356, "grad_norm": 0.9749386310577393, "learning_rate": 0.0002, "epoch": 5.999514994745777, "step": 37110}, {"eval_loss": 1.2270219326019287, "eval_runtime": 122.2047, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 6.0, "step": 37113}, {"loss": 0.4855, "grad_norm": 0.9641092419624329, "learning_rate": 0.0002, "epoch": 6.001131678926522, "step": 37120}, {"loss": 0.4112, "grad_norm": 1.103379249572754, "learning_rate": 0.0002, "epoch": 6.002748363107267, "step": 37130}, {"loss": 0.4577, "grad_norm": 0.8381665349006653, "learning_rate": 0.0002, "epoch": 6.004365047288013, "step": 37140}, {"loss": 0.4794, "grad_norm": 1.245323896408081, "learning_rate": 0.0002, "epoch": 6.005981731468758, "step": 37150}, {"loss": 0.4503, "grad_norm": 1.3140289783477783, "learning_rate": 0.0002, "epoch": 6.007598415649503, "step": 37160}, {"loss": 0.4456, "grad_norm": 0.8479695916175842, "learning_rate": 0.0002, "epoch": 6.009215099830248, "step": 37170}, {"loss": 0.4573, "grad_norm": 0.8841437101364136, "learning_rate": 0.0002, "epoch": 6.0108317840109935, "step": 37180}, {"loss": 0.4565, "grad_norm": 0.8900154829025269, "learning_rate": 0.0002, "epoch": 6.012448468191739, "step": 37190}, {"loss": 0.457, "grad_norm": 1.2753345966339111, "learning_rate": 0.0002, "epoch": 6.014065152372484, "step": 37200}, {"loss": 0.4365, "grad_norm": 1.4625498056411743, "learning_rate": 0.0002, "epoch": 6.015681836553229, "step": 37210}, {"loss": 0.4252, "grad_norm": 0.7455034852027893, "learning_rate": 0.0002, "epoch": 6.017298520733974, "step": 37220}, {"loss": 0.4433, "grad_norm": 1.1658862829208374, "learning_rate": 0.0002, "epoch": 6.01891520491472, "step": 37230}, {"loss": 0.4499, "grad_norm": 0.9785751104354858, "learning_rate": 0.0002, "epoch": 6.020531889095465, "step": 37240}, {"loss": 0.4956, "grad_norm": 1.3193122148513794, "learning_rate": 0.0002, "epoch": 6.02214857327621, "step": 37250}, {"loss": 0.4727, "grad_norm": 1.038273572921753, "learning_rate": 0.0002, "epoch": 6.023765257456955, "step": 37260}, {"loss": 0.4395, "grad_norm": 1.0550594329833984, "learning_rate": 0.0002, "epoch": 6.0253819416377015, "step": 37270}, {"loss": 0.4767, "grad_norm": 0.9745930433273315, "learning_rate": 0.0002, "epoch": 6.026998625818447, "step": 37280}, {"loss": 0.4233, "grad_norm": 0.9273530840873718, "learning_rate": 0.0002, "epoch": 6.028615309999192, "step": 37290}, {"loss": 0.4195, "grad_norm": 1.3844057321548462, "learning_rate": 0.0002, "epoch": 6.030231994179937, "step": 37300}, {"loss": 0.4768, "grad_norm": 1.2058762311935425, "learning_rate": 0.0002, "epoch": 6.031848678360682, "step": 37310}, {"loss": 0.4499, "grad_norm": 1.242663025856018, "learning_rate": 0.0002, "epoch": 6.033465362541428, "step": 37320}, {"loss": 0.4597, "grad_norm": 1.3504270315170288, "learning_rate": 0.0002, "epoch": 6.035082046722173, "step": 37330}, {"loss": 0.4402, "grad_norm": 0.8734912276268005, "learning_rate": 0.0002, "epoch": 6.036698730902918, "step": 37340}, {"loss": 0.477, "grad_norm": 1.0182311534881592, "learning_rate": 0.0002, "epoch": 6.038315415083663, "step": 37350}, {"loss": 0.4261, "grad_norm": 0.9898499846458435, "learning_rate": 0.0002, "epoch": 6.0399320992644085, "step": 37360}, {"loss": 0.4459, "grad_norm": 1.0637860298156738, "learning_rate": 0.0002, "epoch": 6.041548783445154, "step": 37370}, {"loss": 0.4958, "grad_norm": 1.0099523067474365, "learning_rate": 0.0002, "epoch": 6.043165467625899, "step": 37380}, {"loss": 0.4459, "grad_norm": 1.1080750226974487, "learning_rate": 0.0002, "epoch": 6.044782151806644, "step": 37390}, {"loss": 0.4473, "grad_norm": 1.2551289796829224, "learning_rate": 0.0002, "epoch": 6.0463988359873895, "step": 37400}, {"loss": 0.468, "grad_norm": 0.8959632515907288, "learning_rate": 0.0002, "epoch": 6.048015520168136, "step": 37410}, {"loss": 0.4255, "grad_norm": 1.1748892068862915, "learning_rate": 0.0002, "epoch": 6.049632204348881, "step": 37420}, {"loss": 0.4458, "grad_norm": 1.3122745752334595, "learning_rate": 0.0002, "epoch": 6.051248888529626, "step": 37430}, {"loss": 0.4676, "grad_norm": 1.0227985382080078, "learning_rate": 0.0002, "epoch": 6.052865572710371, "step": 37440}, {"loss": 0.4503, "grad_norm": 1.0380030870437622, "learning_rate": 0.0002, "epoch": 6.0544822568911165, "step": 37450}, {"loss": 0.4686, "grad_norm": 0.8919622898101807, "learning_rate": 0.0002, "epoch": 6.056098941071862, "step": 37460}, {"loss": 0.4406, "grad_norm": 1.4554150104522705, "learning_rate": 0.0002, "epoch": 6.057715625252607, "step": 37470}, {"loss": 0.4688, "grad_norm": 1.2853292226791382, "learning_rate": 0.0002, "epoch": 6.059332309433352, "step": 37480}, {"loss": 0.4489, "grad_norm": 1.2951840162277222, "learning_rate": 0.0002, "epoch": 6.0609489936140974, "step": 37490}, {"loss": 0.4819, "grad_norm": 1.1750973463058472, "learning_rate": 0.0002, "epoch": 6.062565677794843, "step": 37500}, {"loss": 0.4574, "grad_norm": 0.9328424334526062, "learning_rate": 0.0002, "epoch": 6.064182361975588, "step": 37510}, {"loss": 0.4597, "grad_norm": 1.0353537797927856, "learning_rate": 0.0002, "epoch": 6.065799046156333, "step": 37520}, {"loss": 0.4407, "grad_norm": 1.1594274044036865, "learning_rate": 0.0002, "epoch": 6.067415730337078, "step": 37530}, {"loss": 0.4642, "grad_norm": 0.9034168124198914, "learning_rate": 0.0002, "epoch": 6.069032414517824, "step": 37540}, {"loss": 0.4625, "grad_norm": 1.068617820739746, "learning_rate": 0.0002, "epoch": 6.070649098698569, "step": 37550}, {"loss": 0.4378, "grad_norm": 1.0931321382522583, "learning_rate": 0.0002, "epoch": 6.072265782879315, "step": 37560}, {"loss": 0.4527, "grad_norm": 1.2542688846588135, "learning_rate": 0.0002, "epoch": 6.07388246706006, "step": 37570}, {"loss": 0.4725, "grad_norm": 1.273384928703308, "learning_rate": 0.0002, "epoch": 6.075499151240805, "step": 37580}, {"loss": 0.4928, "grad_norm": 1.4771400690078735, "learning_rate": 0.0002, "epoch": 6.077115835421551, "step": 37590}, {"loss": 0.461, "grad_norm": 1.3751444816589355, "learning_rate": 0.0002, "epoch": 6.078732519602296, "step": 37600}, {"loss": 0.4602, "grad_norm": 1.4532550573349, "learning_rate": 0.0002, "epoch": 6.080349203783041, "step": 37610}, {"loss": 0.4428, "grad_norm": 1.3175991773605347, "learning_rate": 0.0002, "epoch": 6.081965887963786, "step": 37620}, {"loss": 0.4746, "grad_norm": 1.0624970197677612, "learning_rate": 0.0002, "epoch": 6.083582572144532, "step": 37630}, {"loss": 0.413, "grad_norm": 1.099715232849121, "learning_rate": 0.0002, "epoch": 6.085199256325277, "step": 37640}, {"loss": 0.4528, "grad_norm": 1.0380114316940308, "learning_rate": 0.0002, "epoch": 6.086815940506022, "step": 37650}, {"loss": 0.4373, "grad_norm": 1.1136109828948975, "learning_rate": 0.0002, "epoch": 6.088432624686767, "step": 37660}, {"loss": 0.4915, "grad_norm": 0.996498703956604, "learning_rate": 0.0002, "epoch": 6.0900493088675125, "step": 37670}, {"loss": 0.4713, "grad_norm": 1.0552574396133423, "learning_rate": 0.0002, "epoch": 6.091665993048258, "step": 37680}, {"loss": 0.4414, "grad_norm": 1.4108527898788452, "learning_rate": 0.0002, "epoch": 6.093282677229003, "step": 37690}, {"loss": 0.4851, "grad_norm": 1.1323093175888062, "learning_rate": 0.0002, "epoch": 6.094899361409748, "step": 37700}, {"loss": 0.4455, "grad_norm": 0.9364377856254578, "learning_rate": 0.0002, "epoch": 6.096516045590494, "step": 37710}, {"loss": 0.4791, "grad_norm": 1.1300561428070068, "learning_rate": 0.0002, "epoch": 6.0981327297712395, "step": 37720}, {"loss": 0.4539, "grad_norm": 1.0616047382354736, "learning_rate": 0.0002, "epoch": 6.099749413951985, "step": 37730}, {"loss": 0.4516, "grad_norm": 1.1205905675888062, "learning_rate": 0.0002, "epoch": 6.10136609813273, "step": 37740}, {"loss": 0.4688, "grad_norm": 0.9592534303665161, "learning_rate": 0.0002, "epoch": 6.102982782313475, "step": 37750}, {"loss": 0.4494, "grad_norm": 0.9797531962394714, "learning_rate": 0.0002, "epoch": 6.1045994664942205, "step": 37760}, {"loss": 0.4237, "grad_norm": 1.093404769897461, "learning_rate": 0.0002, "epoch": 6.106216150674966, "step": 37770}, {"loss": 0.4691, "grad_norm": 1.2172642946243286, "learning_rate": 0.0002, "epoch": 6.107832834855711, "step": 37780}, {"loss": 0.4398, "grad_norm": 1.0467255115509033, "learning_rate": 0.0002, "epoch": 6.109449519036456, "step": 37790}, {"loss": 0.4676, "grad_norm": 1.159318208694458, "learning_rate": 0.0002, "epoch": 6.111066203217201, "step": 37800}, {"loss": 0.4539, "grad_norm": 1.0615603923797607, "learning_rate": 0.0002, "epoch": 6.112682887397947, "step": 37810}, {"loss": 0.4957, "grad_norm": 1.0542045831680298, "learning_rate": 0.0002, "epoch": 6.114299571578692, "step": 37820}, {"loss": 0.4512, "grad_norm": 0.8962697982788086, "learning_rate": 0.0002, "epoch": 6.115916255759437, "step": 37830}, {"loss": 0.4519, "grad_norm": 1.106352686882019, "learning_rate": 0.0002, "epoch": 6.117532939940182, "step": 37840}, {"loss": 0.4421, "grad_norm": 1.1660276651382446, "learning_rate": 0.0002, "epoch": 6.1191496241209276, "step": 37850}, {"loss": 0.4701, "grad_norm": 1.3524385690689087, "learning_rate": 0.0002, "epoch": 6.120766308301674, "step": 37860}, {"loss": 0.4684, "grad_norm": 1.1056050062179565, "learning_rate": 0.0002, "epoch": 6.122382992482419, "step": 37870}, {"loss": 0.4518, "grad_norm": 1.0772725343704224, "learning_rate": 0.0002, "epoch": 6.123999676663164, "step": 37880}, {"loss": 0.4356, "grad_norm": 1.1011115312576294, "learning_rate": 0.0002, "epoch": 6.125616360843909, "step": 37890}, {"loss": 0.4909, "grad_norm": 0.8952536582946777, "learning_rate": 0.0002, "epoch": 6.127233045024655, "step": 37900}, {"loss": 0.4299, "grad_norm": 1.244398593902588, "learning_rate": 0.0002, "epoch": 6.1288497292054, "step": 37910}, {"loss": 0.4764, "grad_norm": 0.9658283591270447, "learning_rate": 0.0002, "epoch": 6.130466413386145, "step": 37920}, {"loss": 0.4378, "grad_norm": 1.0649068355560303, "learning_rate": 0.0002, "epoch": 6.13208309756689, "step": 37930}, {"loss": 0.4638, "grad_norm": 0.94698166847229, "learning_rate": 0.0002, "epoch": 6.1336997817476355, "step": 37940}, {"loss": 0.488, "grad_norm": 1.1450897455215454, "learning_rate": 0.0002, "epoch": 6.135316465928381, "step": 37950}, {"loss": 0.4791, "grad_norm": 1.032482624053955, "learning_rate": 0.0002, "epoch": 6.136933150109126, "step": 37960}, {"loss": 0.4179, "grad_norm": 1.0993428230285645, "learning_rate": 0.0002, "epoch": 6.138549834289871, "step": 37970}, {"loss": 0.4781, "grad_norm": 1.2907029390335083, "learning_rate": 0.0002, "epoch": 6.1401665184706165, "step": 37980}, {"loss": 0.4671, "grad_norm": 1.1007903814315796, "learning_rate": 0.0002, "epoch": 6.141783202651362, "step": 37990}, {"loss": 0.4213, "grad_norm": 0.9286124110221863, "learning_rate": 0.0002, "epoch": 6.143399886832107, "step": 38000}, {"loss": 0.4741, "grad_norm": 1.1426366567611694, "learning_rate": 0.0002, "epoch": 6.145016571012853, "step": 38010}, {"loss": 0.4746, "grad_norm": 1.2608287334442139, "learning_rate": 0.0002, "epoch": 6.146633255193598, "step": 38020}, {"loss": 0.454, "grad_norm": 1.1346837282180786, "learning_rate": 0.0002, "epoch": 6.1482499393743435, "step": 38030}, {"loss": 0.4469, "grad_norm": 1.144080400466919, "learning_rate": 0.0002, "epoch": 6.149866623555089, "step": 38040}, {"loss": 0.4515, "grad_norm": 1.3456705808639526, "learning_rate": 0.0002, "epoch": 6.151483307735834, "step": 38050}, {"loss": 0.4775, "grad_norm": 1.0517960786819458, "learning_rate": 0.0002, "epoch": 6.153099991916579, "step": 38060}, {"loss": 0.4986, "grad_norm": 1.1887445449829102, "learning_rate": 0.0002, "epoch": 6.154716676097324, "step": 38070}, {"loss": 0.4516, "grad_norm": 1.0449163913726807, "learning_rate": 0.0002, "epoch": 6.15633336027807, "step": 38080}, {"loss": 0.4808, "grad_norm": 1.3218743801116943, "learning_rate": 0.0002, "epoch": 6.157950044458815, "step": 38090}, {"loss": 0.4632, "grad_norm": 1.003208875656128, "learning_rate": 0.0002, "epoch": 6.15956672863956, "step": 38100}, {"loss": 0.4978, "grad_norm": 1.008623719215393, "learning_rate": 0.0002, "epoch": 6.161183412820305, "step": 38110}, {"loss": 0.4608, "grad_norm": 1.2122787237167358, "learning_rate": 0.0002, "epoch": 6.162800097001051, "step": 38120}, {"loss": 0.4666, "grad_norm": 1.253403902053833, "learning_rate": 0.0002, "epoch": 6.164416781181796, "step": 38130}, {"loss": 0.4778, "grad_norm": 1.2289724349975586, "learning_rate": 0.0002, "epoch": 6.166033465362541, "step": 38140}, {"loss": 0.4774, "grad_norm": 1.330694556236267, "learning_rate": 0.0002, "epoch": 6.167650149543286, "step": 38150}, {"loss": 0.4699, "grad_norm": 1.0946741104125977, "learning_rate": 0.0002, "epoch": 6.169266833724032, "step": 38160}, {"loss": 0.4816, "grad_norm": 1.0719934701919556, "learning_rate": 0.0002, "epoch": 6.170883517904778, "step": 38170}, {"loss": 0.4678, "grad_norm": 1.1142133474349976, "learning_rate": 0.0002, "epoch": 6.172500202085523, "step": 38180}, {"loss": 0.4911, "grad_norm": 1.1221938133239746, "learning_rate": 0.0002, "epoch": 6.174116886266268, "step": 38190}, {"loss": 0.4462, "grad_norm": 1.1391617059707642, "learning_rate": 0.0002, "epoch": 6.175733570447013, "step": 38200}, {"loss": 0.4867, "grad_norm": 1.2263455390930176, "learning_rate": 0.0002, "epoch": 6.1773502546277586, "step": 38210}, {"loss": 0.4633, "grad_norm": 1.0930434465408325, "learning_rate": 0.0002, "epoch": 6.178966938808504, "step": 38220}, {"loss": 0.4406, "grad_norm": 1.3489030599594116, "learning_rate": 0.0002, "epoch": 6.180583622989249, "step": 38230}, {"loss": 0.4994, "grad_norm": 1.1383486986160278, "learning_rate": 0.0002, "epoch": 6.182200307169994, "step": 38240}, {"loss": 0.4851, "grad_norm": 1.2408897876739502, "learning_rate": 0.0002, "epoch": 6.1838169913507395, "step": 38250}, {"loss": 0.4848, "grad_norm": 1.1436222791671753, "learning_rate": 0.0002, "epoch": 6.185433675531485, "step": 38260}, {"loss": 0.4594, "grad_norm": 1.370117425918579, "learning_rate": 0.0002, "epoch": 6.18705035971223, "step": 38270}, {"loss": 0.5023, "grad_norm": 0.8862423300743103, "learning_rate": 0.0002, "epoch": 6.188667043892975, "step": 38280}, {"loss": 0.4559, "grad_norm": 0.9603779315948486, "learning_rate": 0.0002, "epoch": 6.19028372807372, "step": 38290}, {"loss": 0.4835, "grad_norm": 1.389291524887085, "learning_rate": 0.0002, "epoch": 6.191900412254466, "step": 38300}, {"loss": 0.4435, "grad_norm": 1.0767031908035278, "learning_rate": 0.0002, "epoch": 6.193517096435212, "step": 38310}, {"loss": 0.4683, "grad_norm": 1.1800403594970703, "learning_rate": 0.0002, "epoch": 6.195133780615957, "step": 38320}, {"loss": 0.4608, "grad_norm": 0.997891366481781, "learning_rate": 0.0002, "epoch": 6.196750464796702, "step": 38330}, {"loss": 0.4575, "grad_norm": 1.1201492547988892, "learning_rate": 0.0002, "epoch": 6.1983671489774474, "step": 38340}, {"loss": 0.4952, "grad_norm": 0.9769026637077332, "learning_rate": 0.0002, "epoch": 6.199983833158193, "step": 38350}, {"loss": 0.4563, "grad_norm": 0.9447069764137268, "learning_rate": 0.0002, "epoch": 6.201600517338938, "step": 38360}, {"loss": 0.516, "grad_norm": 1.0959235429763794, "learning_rate": 0.0002, "epoch": 6.203217201519683, "step": 38370}, {"loss": 0.4688, "grad_norm": 1.2495406866073608, "learning_rate": 0.0002, "epoch": 6.204833885700428, "step": 38380}, {"loss": 0.4445, "grad_norm": 0.8589218258857727, "learning_rate": 0.0002, "epoch": 6.206450569881174, "step": 38390}, {"loss": 0.4808, "grad_norm": 0.959155797958374, "learning_rate": 0.0002, "epoch": 6.208067254061919, "step": 38400}, {"loss": 0.4622, "grad_norm": 1.0105533599853516, "learning_rate": 0.0002, "epoch": 6.209683938242664, "step": 38410}, {"loss": 0.4887, "grad_norm": 0.9824615120887756, "learning_rate": 0.0002, "epoch": 6.211300622423409, "step": 38420}, {"loss": 0.4656, "grad_norm": 0.8616500496864319, "learning_rate": 0.0002, "epoch": 6.2129173066041545, "step": 38430}, {"loss": 0.449, "grad_norm": 1.2917758226394653, "learning_rate": 0.0002, "epoch": 6.2145339907849, "step": 38440}, {"loss": 0.4201, "grad_norm": 1.0564531087875366, "learning_rate": 0.0002, "epoch": 6.216150674965646, "step": 38450}, {"loss": 0.4849, "grad_norm": 1.152331829071045, "learning_rate": 0.0002, "epoch": 6.217767359146391, "step": 38460}, {"loss": 0.4887, "grad_norm": 0.9152206778526306, "learning_rate": 0.0002, "epoch": 6.219384043327136, "step": 38470}, {"loss": 0.4686, "grad_norm": 0.9931167960166931, "learning_rate": 0.0002, "epoch": 6.221000727507882, "step": 38480}, {"loss": 0.4765, "grad_norm": 1.3248072862625122, "learning_rate": 0.0002, "epoch": 6.222617411688627, "step": 38490}, {"loss": 0.4636, "grad_norm": 1.3916507959365845, "learning_rate": 0.0002, "epoch": 6.224234095869372, "step": 38500}, {"loss": 0.506, "grad_norm": 1.1775140762329102, "learning_rate": 0.0002, "epoch": 6.225850780050117, "step": 38510}, {"loss": 0.47, "grad_norm": 1.1581059694290161, "learning_rate": 0.0002, "epoch": 6.2274674642308625, "step": 38520}, {"loss": 0.4679, "grad_norm": 1.359320878982544, "learning_rate": 0.0002, "epoch": 6.229084148411608, "step": 38530}, {"loss": 0.4697, "grad_norm": 1.185041904449463, "learning_rate": 0.0002, "epoch": 6.230700832592353, "step": 38540}, {"loss": 0.4815, "grad_norm": 1.1861097812652588, "learning_rate": 0.0002, "epoch": 6.232317516773098, "step": 38550}, {"loss": 0.4925, "grad_norm": 1.126990556716919, "learning_rate": 0.0002, "epoch": 6.233934200953843, "step": 38560}, {"loss": 0.4414, "grad_norm": 0.9744541049003601, "learning_rate": 0.0002, "epoch": 6.235550885134589, "step": 38570}, {"loss": 0.4577, "grad_norm": 1.1260887384414673, "learning_rate": 0.0002, "epoch": 6.237167569315334, "step": 38580}, {"loss": 0.4852, "grad_norm": 1.1290327310562134, "learning_rate": 0.0002, "epoch": 6.238784253496079, "step": 38590}, {"loss": 0.4805, "grad_norm": 1.0952879190444946, "learning_rate": 0.0002, "epoch": 6.240400937676825, "step": 38600}, {"loss": 0.4436, "grad_norm": 1.1037684679031372, "learning_rate": 0.0002, "epoch": 6.2420176218575705, "step": 38610}, {"loss": 0.466, "grad_norm": 1.1356085538864136, "learning_rate": 0.0002, "epoch": 6.243634306038316, "step": 38620}, {"loss": 0.5129, "grad_norm": 1.0677106380462646, "learning_rate": 0.0002, "epoch": 6.245250990219061, "step": 38630}, {"loss": 0.4907, "grad_norm": 1.1573411226272583, "learning_rate": 0.0002, "epoch": 6.246867674399806, "step": 38640}, {"loss": 0.5098, "grad_norm": 1.2707505226135254, "learning_rate": 0.0002, "epoch": 6.248484358580551, "step": 38650}, {"loss": 0.4926, "grad_norm": 1.0480109453201294, "learning_rate": 0.0002, "epoch": 6.250101042761297, "step": 38660}, {"loss": 0.4654, "grad_norm": 1.3668724298477173, "learning_rate": 0.0002, "epoch": 6.251717726942042, "step": 38670}, {"loss": 0.5128, "grad_norm": 1.217289686203003, "learning_rate": 0.0002, "epoch": 6.253334411122787, "step": 38680}, {"loss": 0.4621, "grad_norm": 1.2950236797332764, "learning_rate": 0.0002, "epoch": 6.254951095303532, "step": 38690}, {"loss": 0.5076, "grad_norm": 1.4506934881210327, "learning_rate": 0.0002, "epoch": 6.256567779484278, "step": 38700}, {"loss": 0.4803, "grad_norm": 1.1248667240142822, "learning_rate": 0.0002, "epoch": 6.258184463665023, "step": 38710}, {"loss": 0.4746, "grad_norm": 1.3384023904800415, "learning_rate": 0.0002, "epoch": 6.259801147845768, "step": 38720}, {"loss": 0.473, "grad_norm": 1.128074288368225, "learning_rate": 0.0002, "epoch": 6.261417832026513, "step": 38730}, {"loss": 0.4638, "grad_norm": 1.1169012784957886, "learning_rate": 0.0002, "epoch": 6.263034516207259, "step": 38740}, {"loss": 0.4747, "grad_norm": 1.195198893547058, "learning_rate": 0.0002, "epoch": 6.264651200388005, "step": 38750}, {"loss": 0.4906, "grad_norm": 1.2471518516540527, "learning_rate": 0.0002, "epoch": 6.26626788456875, "step": 38760}, {"loss": 0.4507, "grad_norm": 1.2646394968032837, "learning_rate": 0.0002, "epoch": 6.267884568749495, "step": 38770}, {"loss": 0.4934, "grad_norm": 1.0286450386047363, "learning_rate": 0.0002, "epoch": 6.26950125293024, "step": 38780}, {"loss": 0.4787, "grad_norm": 1.2440695762634277, "learning_rate": 0.0002, "epoch": 6.2711179371109855, "step": 38790}, {"loss": 0.4806, "grad_norm": 0.8941256403923035, "learning_rate": 0.0002, "epoch": 6.272734621291731, "step": 38800}, {"loss": 0.4741, "grad_norm": 1.0693447589874268, "learning_rate": 0.0002, "epoch": 6.274351305472476, "step": 38810}, {"loss": 0.4408, "grad_norm": 1.0936840772628784, "learning_rate": 0.0002, "epoch": 6.275967989653221, "step": 38820}, {"loss": 0.4729, "grad_norm": 1.0961874723434448, "learning_rate": 0.0002, "epoch": 6.2775846738339665, "step": 38830}, {"loss": 0.4504, "grad_norm": 1.1465433835983276, "learning_rate": 0.0002, "epoch": 6.279201358014712, "step": 38840}, {"loss": 0.4771, "grad_norm": 1.2987004518508911, "learning_rate": 0.0002, "epoch": 6.280818042195457, "step": 38850}, {"loss": 0.4945, "grad_norm": 1.1310304403305054, "learning_rate": 0.0002, "epoch": 6.282434726376202, "step": 38860}, {"loss": 0.5346, "grad_norm": 1.306538462638855, "learning_rate": 0.0002, "epoch": 6.284051410556947, "step": 38870}, {"loss": 0.4873, "grad_norm": 1.2405401468276978, "learning_rate": 0.0002, "epoch": 6.285668094737693, "step": 38880}, {"loss": 0.4929, "grad_norm": 1.0934767723083496, "learning_rate": 0.0002, "epoch": 6.287284778918439, "step": 38890}, {"loss": 0.4853, "grad_norm": 1.3370496034622192, "learning_rate": 0.0002, "epoch": 6.288901463099184, "step": 38900}, {"loss": 0.4892, "grad_norm": 1.0319404602050781, "learning_rate": 0.0002, "epoch": 6.290518147279929, "step": 38910}, {"loss": 0.4685, "grad_norm": 0.9734271168708801, "learning_rate": 0.0002, "epoch": 6.292134831460674, "step": 38920}, {"loss": 0.5085, "grad_norm": 1.0940454006195068, "learning_rate": 0.0002, "epoch": 6.29375151564142, "step": 38930}, {"loss": 0.4985, "grad_norm": 1.036500334739685, "learning_rate": 0.0002, "epoch": 6.295368199822165, "step": 38940}, {"loss": 0.4878, "grad_norm": 1.020308256149292, "learning_rate": 0.0002, "epoch": 6.29698488400291, "step": 38950}, {"loss": 0.4668, "grad_norm": 1.1416399478912354, "learning_rate": 0.0002, "epoch": 6.298601568183655, "step": 38960}, {"loss": 0.4727, "grad_norm": 1.2497479915618896, "learning_rate": 0.0002, "epoch": 6.300218252364401, "step": 38970}, {"loss": 0.4721, "grad_norm": 1.1692523956298828, "learning_rate": 0.0002, "epoch": 6.301834936545146, "step": 38980}, {"loss": 0.505, "grad_norm": 1.0693109035491943, "learning_rate": 0.0002, "epoch": 6.303451620725891, "step": 38990}, {"loss": 0.4875, "grad_norm": 0.8883291482925415, "learning_rate": 0.0002, "epoch": 6.305068304906636, "step": 39000}, {"loss": 0.5371, "grad_norm": 1.1445088386535645, "learning_rate": 0.0002, "epoch": 6.3066849890873815, "step": 39010}, {"loss": 0.5089, "grad_norm": 1.226792335510254, "learning_rate": 0.0002, "epoch": 6.308301673268127, "step": 39020}, {"loss": 0.474, "grad_norm": 1.0498932600021362, "learning_rate": 0.0002, "epoch": 6.309918357448872, "step": 39030}, {"loss": 0.4964, "grad_norm": 1.0834535360336304, "learning_rate": 0.0002, "epoch": 6.311535041629618, "step": 39040}, {"loss": 0.4733, "grad_norm": 1.144666075706482, "learning_rate": 0.0002, "epoch": 6.313151725810363, "step": 39050}, {"loss": 0.4784, "grad_norm": 1.1468489170074463, "learning_rate": 0.0002, "epoch": 6.3147684099911086, "step": 39060}, {"loss": 0.4911, "grad_norm": 1.290949821472168, "learning_rate": 0.0002, "epoch": 6.316385094171854, "step": 39070}, {"loss": 0.5002, "grad_norm": 1.087868094444275, "learning_rate": 0.0002, "epoch": 6.318001778352599, "step": 39080}, {"loss": 0.4944, "grad_norm": 1.0156296491622925, "learning_rate": 0.0002, "epoch": 6.319618462533344, "step": 39090}, {"loss": 0.5019, "grad_norm": 1.0805060863494873, "learning_rate": 0.0002, "epoch": 6.3212351467140895, "step": 39100}, {"loss": 0.4598, "grad_norm": 0.9030579924583435, "learning_rate": 0.0002, "epoch": 6.322851830894835, "step": 39110}, {"loss": 0.4635, "grad_norm": 1.1488285064697266, "learning_rate": 0.0002, "epoch": 6.32446851507558, "step": 39120}, {"loss": 0.5368, "grad_norm": 1.2050796747207642, "learning_rate": 0.0002, "epoch": 6.326085199256325, "step": 39130}, {"loss": 0.4854, "grad_norm": 1.093451738357544, "learning_rate": 0.0002, "epoch": 6.32770188343707, "step": 39140}, {"loss": 0.5055, "grad_norm": 1.2046772241592407, "learning_rate": 0.0002, "epoch": 6.329318567617816, "step": 39150}, {"loss": 0.4703, "grad_norm": 1.045777678489685, "learning_rate": 0.0002, "epoch": 6.330935251798561, "step": 39160}, {"loss": 0.513, "grad_norm": 1.2008492946624756, "learning_rate": 0.0002, "epoch": 6.332551935979306, "step": 39170}, {"loss": 0.4909, "grad_norm": 1.0613869428634644, "learning_rate": 0.0002, "epoch": 6.334168620160051, "step": 39180}, {"loss": 0.4708, "grad_norm": 1.058440089225769, "learning_rate": 0.0002, "epoch": 6.3357853043407975, "step": 39190}, {"loss": 0.4719, "grad_norm": 1.195658802986145, "learning_rate": 0.0002, "epoch": 6.337401988521543, "step": 39200}, {"loss": 0.4901, "grad_norm": 1.1595174074172974, "learning_rate": 0.0002, "epoch": 6.339018672702288, "step": 39210}, {"loss": 0.4587, "grad_norm": 1.0674750804901123, "learning_rate": 0.0002, "epoch": 6.340635356883033, "step": 39220}, {"loss": 0.4801, "grad_norm": 1.3306758403778076, "learning_rate": 0.0002, "epoch": 6.342252041063778, "step": 39230}, {"loss": 0.4839, "grad_norm": 1.3582593202590942, "learning_rate": 0.0002, "epoch": 6.343868725244524, "step": 39240}, {"loss": 0.4964, "grad_norm": 1.2351572513580322, "learning_rate": 0.0002, "epoch": 6.345485409425269, "step": 39250}, {"loss": 0.4806, "grad_norm": 1.3623450994491577, "learning_rate": 0.0002, "epoch": 6.347102093606014, "step": 39260}, {"loss": 0.466, "grad_norm": 1.201270580291748, "learning_rate": 0.0002, "epoch": 6.348718777786759, "step": 39270}, {"loss": 0.4899, "grad_norm": 0.9300584197044373, "learning_rate": 0.0002, "epoch": 6.3503354619675045, "step": 39280}, {"loss": 0.4867, "grad_norm": 0.944525957107544, "learning_rate": 0.0002, "epoch": 6.35195214614825, "step": 39290}, {"loss": 0.4954, "grad_norm": 1.4263732433319092, "learning_rate": 0.0002, "epoch": 6.353568830328995, "step": 39300}, {"loss": 0.4982, "grad_norm": 1.392592191696167, "learning_rate": 0.0002, "epoch": 6.35518551450974, "step": 39310}, {"loss": 0.4868, "grad_norm": 1.0753393173217773, "learning_rate": 0.0002, "epoch": 6.3568021986904855, "step": 39320}, {"loss": 0.4896, "grad_norm": 1.0088151693344116, "learning_rate": 0.0002, "epoch": 6.358418882871231, "step": 39330}, {"loss": 0.4684, "grad_norm": 1.1784582138061523, "learning_rate": 0.0002, "epoch": 6.360035567051977, "step": 39340}, {"loss": 0.4732, "grad_norm": 1.020526647567749, "learning_rate": 0.0002, "epoch": 6.361652251232722, "step": 39350}, {"loss": 0.5177, "grad_norm": 1.1400747299194336, "learning_rate": 0.0002, "epoch": 6.363268935413467, "step": 39360}, {"loss": 0.4976, "grad_norm": 0.9960665702819824, "learning_rate": 0.0002, "epoch": 6.3648856195942125, "step": 39370}, {"loss": 0.483, "grad_norm": 1.1547569036483765, "learning_rate": 0.0002, "epoch": 6.366502303774958, "step": 39380}, {"loss": 0.4861, "grad_norm": 1.2180676460266113, "learning_rate": 0.0002, "epoch": 6.368118987955703, "step": 39390}, {"loss": 0.4805, "grad_norm": 1.1391799449920654, "learning_rate": 0.0002, "epoch": 6.369735672136448, "step": 39400}, {"loss": 0.5004, "grad_norm": 1.2893574237823486, "learning_rate": 0.0002, "epoch": 6.371352356317193, "step": 39410}, {"loss": 0.4807, "grad_norm": 1.192878246307373, "learning_rate": 0.0002, "epoch": 6.372969040497939, "step": 39420}, {"loss": 0.4637, "grad_norm": 0.9771704077720642, "learning_rate": 0.0002, "epoch": 6.374585724678684, "step": 39430}, {"loss": 0.4867, "grad_norm": 1.285387635231018, "learning_rate": 0.0002, "epoch": 6.376202408859429, "step": 39440}, {"loss": 0.4593, "grad_norm": 1.019957184791565, "learning_rate": 0.0002, "epoch": 6.377819093040174, "step": 39450}, {"loss": 0.473, "grad_norm": 1.2002915143966675, "learning_rate": 0.0002, "epoch": 6.37943577722092, "step": 39460}, {"loss": 0.5025, "grad_norm": 1.3285092115402222, "learning_rate": 0.0002, "epoch": 6.381052461401665, "step": 39470}, {"loss": 0.4626, "grad_norm": 1.097846269607544, "learning_rate": 0.0002, "epoch": 6.38266914558241, "step": 39480}, {"loss": 0.5109, "grad_norm": 0.9537988305091858, "learning_rate": 0.0002, "epoch": 6.384285829763156, "step": 39490}, {"loss": 0.4492, "grad_norm": 1.0350042581558228, "learning_rate": 0.0002, "epoch": 6.385902513943901, "step": 39500}, {"loss": 0.4824, "grad_norm": 0.9559133052825928, "learning_rate": 0.0002, "epoch": 6.387519198124647, "step": 39510}, {"loss": 0.5189, "grad_norm": 0.9615123271942139, "learning_rate": 0.0002, "epoch": 6.389135882305392, "step": 39520}, {"loss": 0.4915, "grad_norm": 1.0604504346847534, "learning_rate": 0.0002, "epoch": 6.390752566486137, "step": 39530}, {"loss": 0.5315, "grad_norm": 1.2460750341415405, "learning_rate": 0.0002, "epoch": 6.392369250666882, "step": 39540}, {"loss": 0.4929, "grad_norm": 1.1496477127075195, "learning_rate": 0.0002, "epoch": 6.393985934847628, "step": 39550}, {"loss": 0.4872, "grad_norm": 1.048043966293335, "learning_rate": 0.0002, "epoch": 6.395602619028373, "step": 39560}, {"loss": 0.5231, "grad_norm": 1.333539366722107, "learning_rate": 0.0002, "epoch": 6.397219303209118, "step": 39570}, {"loss": 0.4877, "grad_norm": 1.0605626106262207, "learning_rate": 0.0002, "epoch": 6.398835987389863, "step": 39580}, {"loss": 0.4643, "grad_norm": 1.163220763206482, "learning_rate": 0.0002, "epoch": 6.4004526715706085, "step": 39590}, {"loss": 0.4824, "grad_norm": 1.1878494024276733, "learning_rate": 0.0002, "epoch": 6.402069355751354, "step": 39600}, {"loss": 0.5242, "grad_norm": 1.4630796909332275, "learning_rate": 0.0002, "epoch": 6.403686039932099, "step": 39610}, {"loss": 0.4985, "grad_norm": 1.073255181312561, "learning_rate": 0.0002, "epoch": 6.405302724112844, "step": 39620}, {"loss": 0.5108, "grad_norm": 1.0538873672485352, "learning_rate": 0.0002, "epoch": 6.406919408293589, "step": 39630}, {"loss": 0.4801, "grad_norm": 1.015525221824646, "learning_rate": 0.0002, "epoch": 6.4085360924743355, "step": 39640}, {"loss": 0.4781, "grad_norm": 1.1454379558563232, "learning_rate": 0.0002, "epoch": 6.410152776655081, "step": 39650}, {"loss": 0.498, "grad_norm": 1.2801800966262817, "learning_rate": 0.0002, "epoch": 6.411769460835826, "step": 39660}, {"loss": 0.4804, "grad_norm": 1.077579140663147, "learning_rate": 0.0002, "epoch": 6.413386145016571, "step": 39670}, {"loss": 0.51, "grad_norm": 1.376662015914917, "learning_rate": 0.0002, "epoch": 6.4150028291973165, "step": 39680}, {"loss": 0.4956, "grad_norm": 1.2064344882965088, "learning_rate": 0.0002, "epoch": 6.416619513378062, "step": 39690}, {"loss": 0.4762, "grad_norm": 1.0689115524291992, "learning_rate": 0.0002, "epoch": 6.418236197558807, "step": 39700}, {"loss": 0.4762, "grad_norm": 0.9997019171714783, "learning_rate": 0.0002, "epoch": 6.419852881739552, "step": 39710}, {"loss": 0.49, "grad_norm": 1.2368080615997314, "learning_rate": 0.0002, "epoch": 6.421469565920297, "step": 39720}, {"loss": 0.4774, "grad_norm": 1.2085820436477661, "learning_rate": 0.0002, "epoch": 6.423086250101043, "step": 39730}, {"loss": 0.4671, "grad_norm": 1.057246208190918, "learning_rate": 0.0002, "epoch": 6.424702934281788, "step": 39740}, {"loss": 0.5315, "grad_norm": 1.1311043500900269, "learning_rate": 0.0002, "epoch": 6.426319618462533, "step": 39750}, {"loss": 0.5171, "grad_norm": 1.2352231740951538, "learning_rate": 0.0002, "epoch": 6.427936302643278, "step": 39760}, {"loss": 0.466, "grad_norm": 0.953233540058136, "learning_rate": 0.0002, "epoch": 6.4295529868240235, "step": 39770}, {"loss": 0.4834, "grad_norm": 1.0632505416870117, "learning_rate": 0.0002, "epoch": 6.431169671004769, "step": 39780}, {"loss": 0.5053, "grad_norm": 1.0916751623153687, "learning_rate": 0.0002, "epoch": 6.432786355185515, "step": 39790}, {"loss": 0.4788, "grad_norm": 0.9732703566551208, "learning_rate": 0.0002, "epoch": 6.43440303936626, "step": 39800}, {"loss": 0.4982, "grad_norm": 1.1673705577850342, "learning_rate": 0.0002, "epoch": 6.436019723547005, "step": 39810}, {"loss": 0.4484, "grad_norm": 1.1049559116363525, "learning_rate": 0.0002, "epoch": 6.437636407727751, "step": 39820}, {"loss": 0.4784, "grad_norm": 1.345277190208435, "learning_rate": 0.0002, "epoch": 6.439253091908496, "step": 39830}, {"loss": 0.4716, "grad_norm": 1.1118950843811035, "learning_rate": 0.0002, "epoch": 6.440869776089241, "step": 39840}, {"loss": 0.5133, "grad_norm": 1.4872850179672241, "learning_rate": 0.0002, "epoch": 6.442486460269986, "step": 39850}, {"loss": 0.4532, "grad_norm": 1.0763497352600098, "learning_rate": 0.0002, "epoch": 6.4441031444507315, "step": 39860}, {"loss": 0.4572, "grad_norm": 0.9245555400848389, "learning_rate": 0.0002, "epoch": 6.445719828631477, "step": 39870}, {"loss": 0.4917, "grad_norm": 1.4154807329177856, "learning_rate": 0.0002, "epoch": 6.447336512812222, "step": 39880}, {"loss": 0.4852, "grad_norm": 1.0885124206542969, "learning_rate": 0.0002, "epoch": 6.448953196992967, "step": 39890}, {"loss": 0.5399, "grad_norm": 1.3989344835281372, "learning_rate": 0.0002, "epoch": 6.450569881173712, "step": 39900}, {"loss": 0.509, "grad_norm": 0.9763124585151672, "learning_rate": 0.0002, "epoch": 6.452186565354458, "step": 39910}, {"loss": 0.5134, "grad_norm": 1.135272741317749, "learning_rate": 0.0002, "epoch": 6.453803249535203, "step": 39920}, {"loss": 0.4941, "grad_norm": 1.1140081882476807, "learning_rate": 0.0002, "epoch": 6.455419933715948, "step": 39930}, {"loss": 0.5137, "grad_norm": 1.0992448329925537, "learning_rate": 0.0002, "epoch": 6.457036617896694, "step": 39940}, {"loss": 0.4914, "grad_norm": 1.1658501625061035, "learning_rate": 0.0002, "epoch": 6.4586533020774395, "step": 39950}, {"loss": 0.5036, "grad_norm": 1.1122797727584839, "learning_rate": 0.0002, "epoch": 6.460269986258185, "step": 39960}, {"loss": 0.5159, "grad_norm": 0.9664968252182007, "learning_rate": 0.0002, "epoch": 6.46188667043893, "step": 39970}, {"loss": 0.4989, "grad_norm": 1.2513965368270874, "learning_rate": 0.0002, "epoch": 6.463503354619675, "step": 39980}, {"loss": 0.4694, "grad_norm": 1.1198630332946777, "learning_rate": 0.0002, "epoch": 6.46512003880042, "step": 39990}, {"loss": 0.5023, "grad_norm": 0.8783249855041504, "learning_rate": 0.0002, "epoch": 6.466736722981166, "step": 40000}, {"loss": 0.4648, "grad_norm": 1.1313109397888184, "learning_rate": 0.0002, "epoch": 6.468353407161911, "step": 40010}, {"loss": 0.4965, "grad_norm": 1.0854487419128418, "learning_rate": 0.0002, "epoch": 6.469970091342656, "step": 40020}, {"loss": 0.5253, "grad_norm": 1.1738566160202026, "learning_rate": 0.0002, "epoch": 6.471586775523401, "step": 40030}, {"loss": 0.4947, "grad_norm": 0.9720084071159363, "learning_rate": 0.0002, "epoch": 6.473203459704147, "step": 40040}, {"loss": 0.5218, "grad_norm": 1.105618953704834, "learning_rate": 0.0002, "epoch": 6.474820143884892, "step": 40050}, {"loss": 0.4943, "grad_norm": 1.2007657289505005, "learning_rate": 0.0002, "epoch": 6.476436828065637, "step": 40060}, {"loss": 0.4882, "grad_norm": 1.088402509689331, "learning_rate": 0.0002, "epoch": 6.478053512246382, "step": 40070}, {"loss": 0.504, "grad_norm": 1.0775291919708252, "learning_rate": 0.0002, "epoch": 6.4796701964271275, "step": 40080}, {"loss": 0.4791, "grad_norm": 1.1018189191818237, "learning_rate": 0.0002, "epoch": 6.481286880607874, "step": 40090}, {"loss": 0.488, "grad_norm": 1.1676557064056396, "learning_rate": 0.0002, "epoch": 6.482903564788619, "step": 40100}, {"loss": 0.4818, "grad_norm": 0.9619805812835693, "learning_rate": 0.0002, "epoch": 6.484520248969364, "step": 40110}, {"loss": 0.4986, "grad_norm": 1.2408208847045898, "learning_rate": 0.0002, "epoch": 6.486136933150109, "step": 40120}, {"loss": 0.4668, "grad_norm": 1.3488136529922485, "learning_rate": 0.0002, "epoch": 6.4877536173308545, "step": 40130}, {"loss": 0.4774, "grad_norm": 0.9864488244056702, "learning_rate": 0.0002, "epoch": 6.4893703015116, "step": 40140}, {"loss": 0.4651, "grad_norm": 0.9437947273254395, "learning_rate": 0.0002, "epoch": 6.490986985692345, "step": 40150}, {"loss": 0.542, "grad_norm": 1.2005455493927002, "learning_rate": 0.0002, "epoch": 6.49260366987309, "step": 40160}, {"loss": 0.4704, "grad_norm": 1.0796732902526855, "learning_rate": 0.0002, "epoch": 6.4942203540538355, "step": 40170}, {"loss": 0.498, "grad_norm": 1.1347825527191162, "learning_rate": 0.0002, "epoch": 6.495837038234581, "step": 40180}, {"loss": 0.5215, "grad_norm": 1.2311455011367798, "learning_rate": 0.0002, "epoch": 6.497453722415326, "step": 40190}, {"loss": 0.5043, "grad_norm": 1.068609356880188, "learning_rate": 0.0002, "epoch": 6.499070406596071, "step": 40200}, {"loss": 0.4868, "grad_norm": 1.196425437927246, "learning_rate": 0.0002, "epoch": 6.500687090776816, "step": 40210}, {"loss": 0.4881, "grad_norm": 1.183927297592163, "learning_rate": 0.0002, "epoch": 6.5023037749575625, "step": 40220}, {"loss": 0.4958, "grad_norm": 0.9099724292755127, "learning_rate": 0.0002, "epoch": 6.503920459138307, "step": 40230}, {"loss": 0.4816, "grad_norm": 0.9261038899421692, "learning_rate": 0.0002, "epoch": 6.505537143319053, "step": 40240}, {"loss": 0.5151, "grad_norm": 1.185491681098938, "learning_rate": 0.0002, "epoch": 6.507153827499798, "step": 40250}, {"loss": 0.4853, "grad_norm": 1.1866052150726318, "learning_rate": 0.0002, "epoch": 6.508770511680543, "step": 40260}, {"loss": 0.491, "grad_norm": 1.1600912809371948, "learning_rate": 0.0002, "epoch": 6.510387195861289, "step": 40270}, {"loss": 0.5181, "grad_norm": 0.9609426259994507, "learning_rate": 0.0002, "epoch": 6.512003880042034, "step": 40280}, {"loss": 0.4794, "grad_norm": 1.078864336013794, "learning_rate": 0.0002, "epoch": 6.513620564222779, "step": 40290}, {"loss": 0.46, "grad_norm": 1.042761206626892, "learning_rate": 0.0002, "epoch": 6.515237248403524, "step": 40300}, {"loss": 0.5341, "grad_norm": 0.9742481112480164, "learning_rate": 0.0002, "epoch": 6.51685393258427, "step": 40310}, {"loss": 0.5234, "grad_norm": 1.2544835805892944, "learning_rate": 0.0002, "epoch": 6.518470616765015, "step": 40320}, {"loss": 0.4815, "grad_norm": 1.3019760847091675, "learning_rate": 0.0002, "epoch": 6.52008730094576, "step": 40330}, {"loss": 0.5039, "grad_norm": 1.3196964263916016, "learning_rate": 0.0002, "epoch": 6.521703985126505, "step": 40340}, {"loss": 0.4979, "grad_norm": 1.2795668840408325, "learning_rate": 0.0002, "epoch": 6.5233206693072505, "step": 40350}, {"loss": 0.5075, "grad_norm": 1.1618940830230713, "learning_rate": 0.0002, "epoch": 6.524937353487996, "step": 40360}, {"loss": 0.5081, "grad_norm": 1.330543041229248, "learning_rate": 0.0002, "epoch": 6.526554037668742, "step": 40370}, {"loss": 0.5055, "grad_norm": 1.1946901082992554, "learning_rate": 0.0002, "epoch": 6.528170721849486, "step": 40380}, {"loss": 0.4518, "grad_norm": 1.1708201169967651, "learning_rate": 0.0002, "epoch": 6.529787406030232, "step": 40390}, {"loss": 0.4556, "grad_norm": 0.894036591053009, "learning_rate": 0.0002, "epoch": 6.531404090210978, "step": 40400}, {"loss": 0.4919, "grad_norm": 1.1199041604995728, "learning_rate": 0.0002, "epoch": 6.533020774391723, "step": 40410}, {"loss": 0.471, "grad_norm": 1.180317759513855, "learning_rate": 0.0002, "epoch": 6.534637458572468, "step": 40420}, {"loss": 0.4914, "grad_norm": 1.37367582321167, "learning_rate": 0.0002, "epoch": 6.536254142753213, "step": 40430}, {"loss": 0.4561, "grad_norm": 1.134791612625122, "learning_rate": 0.0002, "epoch": 6.5378708269339585, "step": 40440}, {"loss": 0.5337, "grad_norm": 1.1160204410552979, "learning_rate": 0.0002, "epoch": 6.539487511114704, "step": 40450}, {"loss": 0.5299, "grad_norm": 1.268347978591919, "learning_rate": 0.0002, "epoch": 6.541104195295449, "step": 40460}, {"loss": 0.5167, "grad_norm": 1.1424330472946167, "learning_rate": 0.0002, "epoch": 6.542720879476194, "step": 40470}, {"loss": 0.5114, "grad_norm": 1.3098465204238892, "learning_rate": 0.0002, "epoch": 6.544337563656939, "step": 40480}, {"loss": 0.4865, "grad_norm": 1.3439544439315796, "learning_rate": 0.0002, "epoch": 6.545954247837685, "step": 40490}, {"loss": 0.5183, "grad_norm": 1.2708452939987183, "learning_rate": 0.0002, "epoch": 6.54757093201843, "step": 40500}, {"loss": 0.5099, "grad_norm": 1.483680248260498, "learning_rate": 0.0002, "epoch": 6.549187616199175, "step": 40510}, {"loss": 0.4811, "grad_norm": 1.1697806119918823, "learning_rate": 0.0002, "epoch": 6.550804300379921, "step": 40520}, {"loss": 0.4814, "grad_norm": 1.1665642261505127, "learning_rate": 0.0002, "epoch": 6.5524209845606665, "step": 40530}, {"loss": 0.4985, "grad_norm": 1.1243325471878052, "learning_rate": 0.0002, "epoch": 6.554037668741412, "step": 40540}, {"loss": 0.4936, "grad_norm": 1.0277988910675049, "learning_rate": 0.0002, "epoch": 6.555654352922157, "step": 40550}, {"loss": 0.487, "grad_norm": 1.1466810703277588, "learning_rate": 0.0002, "epoch": 6.557271037102902, "step": 40560}, {"loss": 0.4851, "grad_norm": 1.1415363550186157, "learning_rate": 0.0002, "epoch": 6.558887721283647, "step": 40570}, {"loss": 0.4631, "grad_norm": 1.1923491954803467, "learning_rate": 0.0002, "epoch": 6.560504405464393, "step": 40580}, {"loss": 0.5071, "grad_norm": 0.9264549612998962, "learning_rate": 0.0002, "epoch": 6.562121089645138, "step": 40590}, {"loss": 0.466, "grad_norm": 0.8810341954231262, "learning_rate": 0.0002, "epoch": 6.563737773825883, "step": 40600}, {"loss": 0.5085, "grad_norm": 2.3296701908111572, "learning_rate": 0.0002, "epoch": 6.565354458006628, "step": 40610}, {"loss": 0.5196, "grad_norm": 1.0865163803100586, "learning_rate": 0.0002, "epoch": 6.5669711421873735, "step": 40620}, {"loss": 0.5132, "grad_norm": 0.9844607710838318, "learning_rate": 0.0002, "epoch": 6.568587826368119, "step": 40630}, {"loss": 0.5437, "grad_norm": 1.1686855554580688, "learning_rate": 0.0002, "epoch": 6.570204510548864, "step": 40640}, {"loss": 0.5293, "grad_norm": 1.016829252243042, "learning_rate": 0.0002, "epoch": 6.571821194729609, "step": 40650}, {"loss": 0.5243, "grad_norm": 1.2789337635040283, "learning_rate": 0.0002, "epoch": 6.5734378789103545, "step": 40660}, {"loss": 0.4867, "grad_norm": 1.0819072723388672, "learning_rate": 0.0002, "epoch": 6.575054563091101, "step": 40670}, {"loss": 0.5024, "grad_norm": 1.1478345394134521, "learning_rate": 0.0002, "epoch": 6.576671247271846, "step": 40680}, {"loss": 0.5282, "grad_norm": 0.7972208857536316, "learning_rate": 0.0002, "epoch": 6.578287931452591, "step": 40690}, {"loss": 0.4877, "grad_norm": 1.1481789350509644, "learning_rate": 0.0002, "epoch": 6.579904615633336, "step": 40700}, {"loss": 0.5143, "grad_norm": 1.0921871662139893, "learning_rate": 0.0002, "epoch": 6.5815212998140815, "step": 40710}, {"loss": 0.5441, "grad_norm": 1.0230315923690796, "learning_rate": 0.0002, "epoch": 6.583137983994827, "step": 40720}, {"loss": 0.4734, "grad_norm": 1.151049017906189, "learning_rate": 0.0002, "epoch": 6.584754668175572, "step": 40730}, {"loss": 0.4782, "grad_norm": 1.4016883373260498, "learning_rate": 0.0002, "epoch": 6.586371352356317, "step": 40740}, {"loss": 0.5195, "grad_norm": 1.2211825847625732, "learning_rate": 0.0002, "epoch": 6.587988036537062, "step": 40750}, {"loss": 0.4815, "grad_norm": 1.2803404331207275, "learning_rate": 0.0002, "epoch": 6.589604720717808, "step": 40760}, {"loss": 0.5329, "grad_norm": 1.1119942665100098, "learning_rate": 0.0002, "epoch": 6.591221404898553, "step": 40770}, {"loss": 0.5135, "grad_norm": 1.464650273323059, "learning_rate": 0.0002, "epoch": 6.592838089079298, "step": 40780}, {"loss": 0.5181, "grad_norm": 1.1751397848129272, "learning_rate": 0.0002, "epoch": 6.594454773260043, "step": 40790}, {"loss": 0.4772, "grad_norm": 1.0866316556930542, "learning_rate": 0.0002, "epoch": 6.596071457440789, "step": 40800}, {"loss": 0.5132, "grad_norm": 1.1733694076538086, "learning_rate": 0.0002, "epoch": 6.597688141621534, "step": 40810}, {"loss": 0.5138, "grad_norm": 1.184708833694458, "learning_rate": 0.0002, "epoch": 6.59930482580228, "step": 40820}, {"loss": 0.4885, "grad_norm": 1.406081199645996, "learning_rate": 0.0002, "epoch": 6.600921509983025, "step": 40830}, {"loss": 0.499, "grad_norm": 0.9658212661743164, "learning_rate": 0.0002, "epoch": 6.60253819416377, "step": 40840}, {"loss": 0.5113, "grad_norm": 1.1457678079605103, "learning_rate": 0.0002, "epoch": 6.604154878344516, "step": 40850}, {"loss": 0.4916, "grad_norm": 1.0487784147262573, "learning_rate": 0.0002, "epoch": 6.605771562525261, "step": 40860}, {"loss": 0.4682, "grad_norm": 0.9357177019119263, "learning_rate": 0.0002, "epoch": 6.607388246706006, "step": 40870}, {"loss": 0.4751, "grad_norm": 1.1479727029800415, "learning_rate": 0.0002, "epoch": 6.609004930886751, "step": 40880}, {"loss": 0.5493, "grad_norm": 1.3729329109191895, "learning_rate": 0.0002, "epoch": 6.610621615067497, "step": 40890}, {"loss": 0.4886, "grad_norm": 1.0085599422454834, "learning_rate": 0.0002, "epoch": 6.612238299248242, "step": 40900}, {"loss": 0.516, "grad_norm": 1.2750911712646484, "learning_rate": 0.0002, "epoch": 6.613854983428987, "step": 40910}, {"loss": 0.5342, "grad_norm": 1.1929547786712646, "learning_rate": 0.0002, "epoch": 6.615471667609732, "step": 40920}, {"loss": 0.4919, "grad_norm": 1.0821375846862793, "learning_rate": 0.0002, "epoch": 6.6170883517904775, "step": 40930}, {"loss": 0.5057, "grad_norm": 1.197347640991211, "learning_rate": 0.0002, "epoch": 6.618705035971223, "step": 40940}, {"loss": 0.492, "grad_norm": 1.2074699401855469, "learning_rate": 0.0002, "epoch": 6.620321720151968, "step": 40950}, {"loss": 0.5089, "grad_norm": 1.312009572982788, "learning_rate": 0.0002, "epoch": 6.621938404332713, "step": 40960}, {"loss": 0.5476, "grad_norm": 1.4381471872329712, "learning_rate": 0.0002, "epoch": 6.623555088513459, "step": 40970}, {"loss": 0.4904, "grad_norm": 1.1574671268463135, "learning_rate": 0.0002, "epoch": 6.6251717726942045, "step": 40980}, {"loss": 0.531, "grad_norm": 0.885661780834198, "learning_rate": 0.0002, "epoch": 6.62678845687495, "step": 40990}, {"loss": 0.5145, "grad_norm": 1.024571180343628, "learning_rate": 0.0002, "epoch": 6.628405141055695, "step": 41000}, {"loss": 0.4791, "grad_norm": 1.103437900543213, "learning_rate": 0.0002, "epoch": 6.63002182523644, "step": 41010}, {"loss": 0.4671, "grad_norm": 1.122450828552246, "learning_rate": 0.0002, "epoch": 6.6316385094171855, "step": 41020}, {"loss": 0.5134, "grad_norm": 1.2256295680999756, "learning_rate": 0.0002, "epoch": 6.633255193597931, "step": 41030}, {"loss": 0.4908, "grad_norm": 1.364594578742981, "learning_rate": 0.0002, "epoch": 6.634871877778676, "step": 41040}, {"loss": 0.4964, "grad_norm": 0.9550056457519531, "learning_rate": 0.0002, "epoch": 6.636488561959421, "step": 41050}, {"loss": 0.5028, "grad_norm": 1.3174707889556885, "learning_rate": 0.0002, "epoch": 6.638105246140166, "step": 41060}, {"loss": 0.4717, "grad_norm": 1.0835540294647217, "learning_rate": 0.0002, "epoch": 6.639721930320912, "step": 41070}, {"loss": 0.497, "grad_norm": 1.1432770490646362, "learning_rate": 0.0002, "epoch": 6.641338614501657, "step": 41080}, {"loss": 0.4903, "grad_norm": 1.2398556470870972, "learning_rate": 0.0002, "epoch": 6.642955298682402, "step": 41090}, {"loss": 0.4991, "grad_norm": 1.1147747039794922, "learning_rate": 0.0002, "epoch": 6.644571982863147, "step": 41100}, {"loss": 0.505, "grad_norm": 1.0730493068695068, "learning_rate": 0.0002, "epoch": 6.6461886670438926, "step": 41110}, {"loss": 0.486, "grad_norm": 1.3218451738357544, "learning_rate": 0.0002, "epoch": 6.647805351224639, "step": 41120}, {"loss": 0.5276, "grad_norm": 1.3027331829071045, "learning_rate": 0.0002, "epoch": 6.649422035405384, "step": 41130}, {"loss": 0.5263, "grad_norm": 1.0280735492706299, "learning_rate": 0.0002, "epoch": 6.651038719586129, "step": 41140}, {"loss": 0.4952, "grad_norm": 1.109916090965271, "learning_rate": 0.0002, "epoch": 6.652655403766874, "step": 41150}, {"loss": 0.5001, "grad_norm": 1.078734040260315, "learning_rate": 0.0002, "epoch": 6.65427208794762, "step": 41160}, {"loss": 0.484, "grad_norm": 1.1595654487609863, "learning_rate": 0.0002, "epoch": 6.655888772128365, "step": 41170}, {"loss": 0.5101, "grad_norm": 1.1701031923294067, "learning_rate": 0.0002, "epoch": 6.65750545630911, "step": 41180}, {"loss": 0.5341, "grad_norm": 1.0424643754959106, "learning_rate": 0.0002, "epoch": 6.659122140489855, "step": 41190}, {"loss": 0.4863, "grad_norm": 1.22880220413208, "learning_rate": 0.0002, "epoch": 6.6607388246706005, "step": 41200}, {"loss": 0.4987, "grad_norm": 1.1907655000686646, "learning_rate": 0.0002, "epoch": 6.662355508851346, "step": 41210}, {"loss": 0.5343, "grad_norm": 1.0765007734298706, "learning_rate": 0.0002, "epoch": 6.663972193032091, "step": 41220}, {"loss": 0.5039, "grad_norm": 0.9994917511940002, "learning_rate": 0.0002, "epoch": 6.665588877212836, "step": 41230}, {"loss": 0.507, "grad_norm": 0.968578040599823, "learning_rate": 0.0002, "epoch": 6.6672055613935814, "step": 41240}, {"loss": 0.5068, "grad_norm": 1.0576032400131226, "learning_rate": 0.0002, "epoch": 6.668822245574327, "step": 41250}, {"loss": 0.486, "grad_norm": 1.2183765172958374, "learning_rate": 0.0002, "epoch": 6.670438929755072, "step": 41260}, {"loss": 0.4764, "grad_norm": 1.2548623085021973, "learning_rate": 0.0002, "epoch": 6.672055613935818, "step": 41270}, {"loss": 0.5014, "grad_norm": 1.0848388671875, "learning_rate": 0.0002, "epoch": 6.673672298116563, "step": 41280}, {"loss": 0.5404, "grad_norm": 1.21421217918396, "learning_rate": 0.0002, "epoch": 6.6752889822973085, "step": 41290}, {"loss": 0.4911, "grad_norm": 1.1453598737716675, "learning_rate": 0.0002, "epoch": 6.676905666478054, "step": 41300}, {"loss": 0.5033, "grad_norm": 1.2682722806930542, "learning_rate": 0.0002, "epoch": 6.678522350658799, "step": 41310}, {"loss": 0.5313, "grad_norm": 1.1659725904464722, "learning_rate": 0.0002, "epoch": 6.680139034839544, "step": 41320}, {"loss": 0.5505, "grad_norm": 1.36194908618927, "learning_rate": 0.0002, "epoch": 6.681755719020289, "step": 41330}, {"loss": 0.5127, "grad_norm": 1.1712592840194702, "learning_rate": 0.0002, "epoch": 6.683372403201035, "step": 41340}, {"loss": 0.5082, "grad_norm": 1.4168336391448975, "learning_rate": 0.0002, "epoch": 6.68498908738178, "step": 41350}, {"loss": 0.5124, "grad_norm": 1.0395328998565674, "learning_rate": 0.0002, "epoch": 6.686605771562525, "step": 41360}, {"loss": 0.5404, "grad_norm": 1.2511054277420044, "learning_rate": 0.0002, "epoch": 6.68822245574327, "step": 41370}, {"loss": 0.5027, "grad_norm": 1.0438542366027832, "learning_rate": 0.0002, "epoch": 6.689839139924016, "step": 41380}, {"loss": 0.5069, "grad_norm": 1.08684241771698, "learning_rate": 0.0002, "epoch": 6.691455824104761, "step": 41390}, {"loss": 0.5224, "grad_norm": 1.250788927078247, "learning_rate": 0.0002, "epoch": 6.693072508285506, "step": 41400}, {"loss": 0.4921, "grad_norm": 1.313890814781189, "learning_rate": 0.0002, "epoch": 6.694689192466251, "step": 41410}, {"loss": 0.5028, "grad_norm": 1.3218982219696045, "learning_rate": 0.0002, "epoch": 6.696305876646997, "step": 41420}, {"loss": 0.4851, "grad_norm": 1.0366582870483398, "learning_rate": 0.0002, "epoch": 6.697922560827743, "step": 41430}, {"loss": 0.5103, "grad_norm": 1.066121220588684, "learning_rate": 0.0002, "epoch": 6.699539245008488, "step": 41440}, {"loss": 0.4966, "grad_norm": 1.0239925384521484, "learning_rate": 0.0002, "epoch": 6.701155929189233, "step": 41450}, {"loss": 0.4767, "grad_norm": 0.9402176141738892, "learning_rate": 0.0002, "epoch": 6.702772613369978, "step": 41460}, {"loss": 0.5381, "grad_norm": 1.391718864440918, "learning_rate": 0.0002, "epoch": 6.7043892975507235, "step": 41470}, {"loss": 0.512, "grad_norm": 1.215600609779358, "learning_rate": 0.0002, "epoch": 6.706005981731469, "step": 41480}, {"loss": 0.5219, "grad_norm": 1.063722848892212, "learning_rate": 0.0002, "epoch": 6.707622665912214, "step": 41490}, {"loss": 0.492, "grad_norm": 1.132149577140808, "learning_rate": 0.0002, "epoch": 6.709239350092959, "step": 41500}, {"loss": 0.4812, "grad_norm": 1.0302950143814087, "learning_rate": 0.0002, "epoch": 6.7108560342737045, "step": 41510}, {"loss": 0.5141, "grad_norm": 1.5342752933502197, "learning_rate": 0.0002, "epoch": 6.71247271845445, "step": 41520}, {"loss": 0.5123, "grad_norm": 1.177137017250061, "learning_rate": 0.0002, "epoch": 6.714089402635195, "step": 41530}, {"loss": 0.5082, "grad_norm": 1.2335538864135742, "learning_rate": 0.0002, "epoch": 6.71570608681594, "step": 41540}, {"loss": 0.4864, "grad_norm": 1.140604853630066, "learning_rate": 0.0002, "epoch": 6.717322770996686, "step": 41550}, {"loss": 0.4888, "grad_norm": 1.3567465543746948, "learning_rate": 0.0002, "epoch": 6.718939455177431, "step": 41560}, {"loss": 0.5183, "grad_norm": 1.0693929195404053, "learning_rate": 0.0002, "epoch": 6.720556139358177, "step": 41570}, {"loss": 0.5131, "grad_norm": 1.1592605113983154, "learning_rate": 0.0002, "epoch": 6.722172823538922, "step": 41580}, {"loss": 0.5476, "grad_norm": 0.989006519317627, "learning_rate": 0.0002, "epoch": 6.723789507719667, "step": 41590}, {"loss": 0.4952, "grad_norm": 1.04103422164917, "learning_rate": 0.0002, "epoch": 6.7254061919004124, "step": 41600}, {"loss": 0.4823, "grad_norm": 1.1129004955291748, "learning_rate": 0.0002, "epoch": 6.727022876081158, "step": 41610}, {"loss": 0.5032, "grad_norm": 1.1473113298416138, "learning_rate": 0.0002, "epoch": 6.728639560261903, "step": 41620}, {"loss": 0.5253, "grad_norm": 1.348036527633667, "learning_rate": 0.0002, "epoch": 6.730256244442648, "step": 41630}, {"loss": 0.4983, "grad_norm": 1.259942650794983, "learning_rate": 0.0002, "epoch": 6.731872928623393, "step": 41640}, {"loss": 0.5182, "grad_norm": 1.0591514110565186, "learning_rate": 0.0002, "epoch": 6.733489612804139, "step": 41650}, {"loss": 0.4886, "grad_norm": 0.9737129211425781, "learning_rate": 0.0002, "epoch": 6.735106296984884, "step": 41660}, {"loss": 0.5051, "grad_norm": 1.2520451545715332, "learning_rate": 0.0002, "epoch": 6.736722981165629, "step": 41670}, {"loss": 0.5364, "grad_norm": 1.0555530786514282, "learning_rate": 0.0002, "epoch": 6.738339665346374, "step": 41680}, {"loss": 0.4954, "grad_norm": 1.0025697946548462, "learning_rate": 0.0002, "epoch": 6.7399563495271195, "step": 41690}, {"loss": 0.5485, "grad_norm": 1.1114100217819214, "learning_rate": 0.0002, "epoch": 6.741573033707866, "step": 41700}, {"loss": 0.4986, "grad_norm": 1.1537504196166992, "learning_rate": 0.0002, "epoch": 6.74318971788861, "step": 41710}, {"loss": 0.5025, "grad_norm": 1.037880539894104, "learning_rate": 0.0002, "epoch": 6.744806402069356, "step": 41720}, {"loss": 0.482, "grad_norm": 1.0691965818405151, "learning_rate": 0.0002, "epoch": 6.746423086250101, "step": 41730}, {"loss": 0.5272, "grad_norm": 1.376325011253357, "learning_rate": 0.0002, "epoch": 6.748039770430847, "step": 41740}, {"loss": 0.5484, "grad_norm": 1.4667129516601562, "learning_rate": 0.0002, "epoch": 6.749656454611592, "step": 41750}, {"loss": 0.5139, "grad_norm": 1.1517162322998047, "learning_rate": 0.0002, "epoch": 6.751273138792337, "step": 41760}, {"loss": 0.5523, "grad_norm": 1.1454511880874634, "learning_rate": 0.0002, "epoch": 6.752889822973082, "step": 41770}, {"loss": 0.4664, "grad_norm": 1.6323128938674927, "learning_rate": 0.0002, "epoch": 6.7545065071538275, "step": 41780}, {"loss": 0.5153, "grad_norm": 1.0951642990112305, "learning_rate": 0.0002, "epoch": 6.756123191334573, "step": 41790}, {"loss": 0.4998, "grad_norm": 1.0766983032226562, "learning_rate": 0.0002, "epoch": 6.757739875515318, "step": 41800}, {"loss": 0.548, "grad_norm": 1.3472381830215454, "learning_rate": 0.0002, "epoch": 6.759356559696063, "step": 41810}, {"loss": 0.5172, "grad_norm": 1.0248444080352783, "learning_rate": 0.0002, "epoch": 6.760973243876808, "step": 41820}, {"loss": 0.5236, "grad_norm": 1.1276055574417114, "learning_rate": 0.0002, "epoch": 6.762589928057554, "step": 41830}, {"loss": 0.5044, "grad_norm": 1.5398495197296143, "learning_rate": 0.0002, "epoch": 6.764206612238299, "step": 41840}, {"loss": 0.5097, "grad_norm": 1.1886497735977173, "learning_rate": 0.0002, "epoch": 6.765823296419045, "step": 41850}, {"loss": 0.499, "grad_norm": 1.027198076248169, "learning_rate": 0.0002, "epoch": 6.767439980599789, "step": 41860}, {"loss": 0.5444, "grad_norm": 1.4644980430603027, "learning_rate": 0.0002, "epoch": 6.7690566647805355, "step": 41870}, {"loss": 0.5009, "grad_norm": 0.9633586406707764, "learning_rate": 0.0002, "epoch": 6.770673348961281, "step": 41880}, {"loss": 0.484, "grad_norm": 1.0895354747772217, "learning_rate": 0.0002, "epoch": 6.772290033142026, "step": 41890}, {"loss": 0.5172, "grad_norm": 1.1887167692184448, "learning_rate": 0.0002, "epoch": 6.773906717322771, "step": 41900}, {"loss": 0.5399, "grad_norm": 1.3699820041656494, "learning_rate": 0.0002, "epoch": 6.775523401503516, "step": 41910}, {"loss": 0.5504, "grad_norm": 1.0266352891921997, "learning_rate": 0.0002, "epoch": 6.777140085684262, "step": 41920}, {"loss": 0.5105, "grad_norm": 1.0919075012207031, "learning_rate": 0.0002, "epoch": 6.778756769865007, "step": 41930}, {"loss": 0.4842, "grad_norm": 0.9839563369750977, "learning_rate": 0.0002, "epoch": 6.780373454045752, "step": 41940}, {"loss": 0.5081, "grad_norm": 1.2605451345443726, "learning_rate": 0.0002, "epoch": 6.781990138226497, "step": 41950}, {"loss": 0.5391, "grad_norm": 0.9268672466278076, "learning_rate": 0.0002, "epoch": 6.7836068224072426, "step": 41960}, {"loss": 0.4916, "grad_norm": 1.2002313137054443, "learning_rate": 0.0002, "epoch": 6.785223506587988, "step": 41970}, {"loss": 0.5467, "grad_norm": 1.2018438577651978, "learning_rate": 0.0002, "epoch": 6.786840190768733, "step": 41980}, {"loss": 0.5491, "grad_norm": 1.17646062374115, "learning_rate": 0.0002, "epoch": 6.788456874949478, "step": 41990}, {"loss": 0.5354, "grad_norm": 1.1080009937286377, "learning_rate": 0.0002, "epoch": 6.790073559130224, "step": 42000}, {"loss": 0.5384, "grad_norm": 1.1606498956680298, "learning_rate": 0.0002, "epoch": 6.791690243310969, "step": 42010}, {"loss": 0.4931, "grad_norm": 1.2484819889068604, "learning_rate": 0.0002, "epoch": 6.793306927491715, "step": 42020}, {"loss": 0.498, "grad_norm": 1.1363215446472168, "learning_rate": 0.0002, "epoch": 6.79492361167246, "step": 42030}, {"loss": 0.5343, "grad_norm": 1.4469727277755737, "learning_rate": 0.0002, "epoch": 6.796540295853205, "step": 42040}, {"loss": 0.5146, "grad_norm": 1.0617138147354126, "learning_rate": 0.0002, "epoch": 6.7981569800339505, "step": 42050}, {"loss": 0.5188, "grad_norm": 1.1459330320358276, "learning_rate": 0.0002, "epoch": 6.799773664214696, "step": 42060}, {"loss": 0.5116, "grad_norm": 1.2095019817352295, "learning_rate": 0.0002, "epoch": 6.801390348395441, "step": 42070}, {"loss": 0.545, "grad_norm": 1.3200831413269043, "learning_rate": 0.0002, "epoch": 6.803007032576186, "step": 42080}, {"loss": 0.5406, "grad_norm": 1.1633318662643433, "learning_rate": 0.0002, "epoch": 6.8046237167569315, "step": 42090}, {"loss": 0.4938, "grad_norm": 0.8986614942550659, "learning_rate": 0.0002, "epoch": 6.806240400937677, "step": 42100}, {"loss": 0.559, "grad_norm": 1.3705275058746338, "learning_rate": 0.0002, "epoch": 6.807857085118422, "step": 42110}, {"loss": 0.5022, "grad_norm": 1.2418090105056763, "learning_rate": 0.0002, "epoch": 6.809473769299167, "step": 42120}, {"loss": 0.5014, "grad_norm": 1.0818954706192017, "learning_rate": 0.0002, "epoch": 6.811090453479912, "step": 42130}, {"loss": 0.4791, "grad_norm": 0.9293872117996216, "learning_rate": 0.0002, "epoch": 6.812707137660658, "step": 42140}, {"loss": 0.5009, "grad_norm": 0.9791894555091858, "learning_rate": 0.0002, "epoch": 6.814323821841404, "step": 42150}, {"loss": 0.5142, "grad_norm": 1.1956568956375122, "learning_rate": 0.0002, "epoch": 6.815940506022149, "step": 42160}, {"loss": 0.5126, "grad_norm": 0.9643568992614746, "learning_rate": 0.0002, "epoch": 6.817557190202894, "step": 42170}, {"loss": 0.5121, "grad_norm": 1.2499792575836182, "learning_rate": 0.0002, "epoch": 6.819173874383639, "step": 42180}, {"loss": 0.4942, "grad_norm": 1.1779413223266602, "learning_rate": 0.0002, "epoch": 6.820790558564385, "step": 42190}, {"loss": 0.498, "grad_norm": 1.0570595264434814, "learning_rate": 0.0002, "epoch": 6.82240724274513, "step": 42200}, {"loss": 0.4997, "grad_norm": 1.1393938064575195, "learning_rate": 0.0002, "epoch": 6.824023926925875, "step": 42210}, {"loss": 0.4842, "grad_norm": 1.152463436126709, "learning_rate": 0.0002, "epoch": 6.82564061110662, "step": 42220}, {"loss": 0.5234, "grad_norm": 1.3353025913238525, "learning_rate": 0.0002, "epoch": 6.827257295287366, "step": 42230}, {"loss": 0.539, "grad_norm": 1.1719051599502563, "learning_rate": 0.0002, "epoch": 6.828873979468111, "step": 42240}, {"loss": 0.5139, "grad_norm": 1.262141227722168, "learning_rate": 0.0002, "epoch": 6.830490663648856, "step": 42250}, {"loss": 0.5021, "grad_norm": 1.240899920463562, "learning_rate": 0.0002, "epoch": 6.832107347829601, "step": 42260}, {"loss": 0.4961, "grad_norm": 1.0505269765853882, "learning_rate": 0.0002, "epoch": 6.8337240320103465, "step": 42270}, {"loss": 0.4932, "grad_norm": 1.1556071043014526, "learning_rate": 0.0002, "epoch": 6.835340716191092, "step": 42280}, {"loss": 0.5461, "grad_norm": 1.1427719593048096, "learning_rate": 0.0002, "epoch": 6.836957400371837, "step": 42290}, {"loss": 0.5199, "grad_norm": 1.1540080308914185, "learning_rate": 0.0002, "epoch": 6.838574084552583, "step": 42300}, {"loss": 0.5269, "grad_norm": 1.0521200895309448, "learning_rate": 0.0002, "epoch": 6.840190768733328, "step": 42310}, {"loss": 0.541, "grad_norm": 1.0205531120300293, "learning_rate": 0.0002, "epoch": 6.8418074529140736, "step": 42320}, {"loss": 0.5225, "grad_norm": 1.0010193586349487, "learning_rate": 0.0002, "epoch": 6.843424137094819, "step": 42330}, {"loss": 0.5101, "grad_norm": 1.2138770818710327, "learning_rate": 0.0002, "epoch": 6.845040821275564, "step": 42340}, {"loss": 0.5452, "grad_norm": 1.3028651475906372, "learning_rate": 0.0002, "epoch": 6.846657505456309, "step": 42350}, {"loss": 0.4894, "grad_norm": 1.0326353311538696, "learning_rate": 0.0002, "epoch": 6.8482741896370545, "step": 42360}, {"loss": 0.5285, "grad_norm": 1.036085605621338, "learning_rate": 0.0002, "epoch": 6.8498908738178, "step": 42370}, {"loss": 0.505, "grad_norm": 1.0575472116470337, "learning_rate": 0.0002, "epoch": 6.851507557998545, "step": 42380}, {"loss": 0.4997, "grad_norm": 1.1749629974365234, "learning_rate": 0.0002, "epoch": 6.85312424217929, "step": 42390}, {"loss": 0.4961, "grad_norm": 1.1747760772705078, "learning_rate": 0.0002, "epoch": 6.854740926360035, "step": 42400}, {"loss": 0.5138, "grad_norm": 1.1877071857452393, "learning_rate": 0.0002, "epoch": 6.856357610540781, "step": 42410}, {"loss": 0.4972, "grad_norm": 1.1209983825683594, "learning_rate": 0.0002, "epoch": 6.857974294721526, "step": 42420}, {"loss": 0.4939, "grad_norm": 1.2918205261230469, "learning_rate": 0.0002, "epoch": 6.859590978902271, "step": 42430}, {"loss": 0.5012, "grad_norm": 1.2443464994430542, "learning_rate": 0.0002, "epoch": 6.861207663083016, "step": 42440}, {"loss": 0.5226, "grad_norm": 0.9336795210838318, "learning_rate": 0.0002, "epoch": 6.8628243472637624, "step": 42450}, {"loss": 0.5108, "grad_norm": 1.2183542251586914, "learning_rate": 0.0002, "epoch": 6.864441031444508, "step": 42460}, {"loss": 0.5245, "grad_norm": 1.0071234703063965, "learning_rate": 0.0002, "epoch": 6.866057715625253, "step": 42470}, {"loss": 0.4753, "grad_norm": 1.2914012670516968, "learning_rate": 0.0002, "epoch": 6.867674399805998, "step": 42480}, {"loss": 0.4865, "grad_norm": 1.1050426959991455, "learning_rate": 0.0002, "epoch": 6.869291083986743, "step": 42490}, {"loss": 0.5243, "grad_norm": 1.1163811683654785, "learning_rate": 0.0002, "epoch": 6.870907768167489, "step": 42500}, {"loss": 0.5065, "grad_norm": 1.1575818061828613, "learning_rate": 0.0002, "epoch": 6.872524452348234, "step": 42510}, {"loss": 0.5353, "grad_norm": 1.11167311668396, "learning_rate": 0.0002, "epoch": 6.874141136528979, "step": 42520}, {"loss": 0.5141, "grad_norm": 1.0379102230072021, "learning_rate": 0.0002, "epoch": 6.875757820709724, "step": 42530}, {"loss": 0.5355, "grad_norm": 1.2617160081863403, "learning_rate": 0.0002, "epoch": 6.8773745048904695, "step": 42540}, {"loss": 0.4785, "grad_norm": 1.1749719381332397, "learning_rate": 0.0002, "epoch": 6.878991189071215, "step": 42550}, {"loss": 0.5503, "grad_norm": 1.2284821271896362, "learning_rate": 0.0002, "epoch": 6.88060787325196, "step": 42560}, {"loss": 0.5065, "grad_norm": 1.1917030811309814, "learning_rate": 0.0002, "epoch": 6.882224557432705, "step": 42570}, {"loss": 0.5176, "grad_norm": 1.1943914890289307, "learning_rate": 0.0002, "epoch": 6.8838412416134505, "step": 42580}, {"loss": 0.5072, "grad_norm": 1.2641394138336182, "learning_rate": 0.0002, "epoch": 6.885457925794196, "step": 42590}, {"loss": 0.5004, "grad_norm": 1.1280436515808105, "learning_rate": 0.0002, "epoch": 6.887074609974942, "step": 42600}, {"loss": 0.5328, "grad_norm": 0.9865449070930481, "learning_rate": 0.0002, "epoch": 6.888691294155687, "step": 42610}, {"loss": 0.4953, "grad_norm": 0.994987428188324, "learning_rate": 0.0002, "epoch": 6.890307978336432, "step": 42620}, {"loss": 0.4805, "grad_norm": 0.9900388717651367, "learning_rate": 0.0002, "epoch": 6.8919246625171775, "step": 42630}, {"loss": 0.5467, "grad_norm": 1.2992421388626099, "learning_rate": 0.0002, "epoch": 6.893541346697923, "step": 42640}, {"loss": 0.5017, "grad_norm": 1.0152487754821777, "learning_rate": 0.0002, "epoch": 6.895158030878668, "step": 42650}, {"loss": 0.5043, "grad_norm": 1.199453353881836, "learning_rate": 0.0002, "epoch": 6.896774715059413, "step": 42660}, {"loss": 0.5106, "grad_norm": 1.100630521774292, "learning_rate": 0.0002, "epoch": 6.898391399240158, "step": 42670}, {"loss": 0.503, "grad_norm": 1.0489764213562012, "learning_rate": 0.0002, "epoch": 6.900008083420904, "step": 42680}, {"loss": 0.4634, "grad_norm": 1.101407527923584, "learning_rate": 0.0002, "epoch": 6.901624767601649, "step": 42690}, {"loss": 0.5361, "grad_norm": 1.3130593299865723, "learning_rate": 0.0002, "epoch": 6.903241451782394, "step": 42700}, {"loss": 0.5119, "grad_norm": 0.9906072616577148, "learning_rate": 0.0002, "epoch": 6.904858135963139, "step": 42710}, {"loss": 0.5146, "grad_norm": 1.094502329826355, "learning_rate": 0.0002, "epoch": 6.906474820143885, "step": 42720}, {"loss": 0.5165, "grad_norm": 1.1025426387786865, "learning_rate": 0.0002, "epoch": 6.90809150432463, "step": 42730}, {"loss": 0.5463, "grad_norm": 1.0644042491912842, "learning_rate": 0.0002, "epoch": 6.909708188505375, "step": 42740}, {"loss": 0.5024, "grad_norm": 1.0709129571914673, "learning_rate": 0.0002, "epoch": 6.911324872686121, "step": 42750}, {"loss": 0.5093, "grad_norm": 1.2445871829986572, "learning_rate": 0.0002, "epoch": 6.912941556866866, "step": 42760}, {"loss": 0.5305, "grad_norm": 1.020058035850525, "learning_rate": 0.0002, "epoch": 6.914558241047612, "step": 42770}, {"loss": 0.5382, "grad_norm": 0.9795091152191162, "learning_rate": 0.0002, "epoch": 6.916174925228357, "step": 42780}, {"loss": 0.5429, "grad_norm": 0.9369977116584778, "learning_rate": 0.0002, "epoch": 6.917791609409102, "step": 42790}, {"loss": 0.5444, "grad_norm": 1.0741904973983765, "learning_rate": 0.0002, "epoch": 6.919408293589847, "step": 42800}, {"loss": 0.5402, "grad_norm": 1.0702799558639526, "learning_rate": 0.0002, "epoch": 6.921024977770593, "step": 42810}, {"loss": 0.5291, "grad_norm": 1.0383983850479126, "learning_rate": 0.0002, "epoch": 6.922641661951338, "step": 42820}, {"loss": 0.5106, "grad_norm": 1.0761083364486694, "learning_rate": 0.0002, "epoch": 6.924258346132083, "step": 42830}, {"loss": 0.5726, "grad_norm": 1.2332350015640259, "learning_rate": 0.0002, "epoch": 6.925875030312828, "step": 42840}, {"loss": 0.4996, "grad_norm": 1.3184348344802856, "learning_rate": 0.0002, "epoch": 6.9274917144935735, "step": 42850}, {"loss": 0.5503, "grad_norm": 1.0586378574371338, "learning_rate": 0.0002, "epoch": 6.929108398674319, "step": 42860}, {"loss": 0.511, "grad_norm": 1.2294201850891113, "learning_rate": 0.0002, "epoch": 6.930725082855064, "step": 42870}, {"loss": 0.54, "grad_norm": 1.3097991943359375, "learning_rate": 0.0002, "epoch": 6.932341767035809, "step": 42880}, {"loss": 0.5228, "grad_norm": 0.9006873965263367, "learning_rate": 0.0002, "epoch": 6.933958451216554, "step": 42890}, {"loss": 0.4617, "grad_norm": 1.265931248664856, "learning_rate": 0.0002, "epoch": 6.9355751353973005, "step": 42900}, {"loss": 0.5029, "grad_norm": 1.1013522148132324, "learning_rate": 0.0002, "epoch": 6.937191819578046, "step": 42910}, {"loss": 0.5334, "grad_norm": 0.9910131692886353, "learning_rate": 0.0002, "epoch": 6.938808503758791, "step": 42920}, {"loss": 0.5211, "grad_norm": 1.102683424949646, "learning_rate": 0.0002, "epoch": 6.940425187939536, "step": 42930}, {"loss": 0.5588, "grad_norm": 1.232961893081665, "learning_rate": 0.0002, "epoch": 6.9420418721202815, "step": 42940}, {"loss": 0.5357, "grad_norm": 1.1714650392532349, "learning_rate": 0.0002, "epoch": 6.943658556301027, "step": 42950}, {"loss": 0.5232, "grad_norm": 1.1684318780899048, "learning_rate": 0.0002, "epoch": 6.945275240481772, "step": 42960}, {"loss": 0.5035, "grad_norm": 1.2074716091156006, "learning_rate": 0.0002, "epoch": 6.946891924662517, "step": 42970}, {"loss": 0.5111, "grad_norm": 1.2061275243759155, "learning_rate": 0.0002, "epoch": 6.948508608843262, "step": 42980}, {"loss": 0.5066, "grad_norm": 1.1216989755630493, "learning_rate": 0.0002, "epoch": 6.950125293024008, "step": 42990}, {"loss": 0.4948, "grad_norm": 1.304117202758789, "learning_rate": 0.0002, "epoch": 6.951741977204753, "step": 43000}, {"loss": 0.5684, "grad_norm": 1.2377972602844238, "learning_rate": 0.0002, "epoch": 6.953358661385498, "step": 43010}, {"loss": 0.4792, "grad_norm": 1.2332178354263306, "learning_rate": 0.0002, "epoch": 6.954975345566243, "step": 43020}, {"loss": 0.5181, "grad_norm": 1.1919599771499634, "learning_rate": 0.0002, "epoch": 6.956592029746989, "step": 43030}, {"loss": 0.5352, "grad_norm": 1.272700548171997, "learning_rate": 0.0002, "epoch": 6.958208713927734, "step": 43040}, {"loss": 0.5328, "grad_norm": 1.4377546310424805, "learning_rate": 0.0002, "epoch": 6.95982539810848, "step": 43050}, {"loss": 0.4894, "grad_norm": 1.2070353031158447, "learning_rate": 0.0002, "epoch": 6.961442082289225, "step": 43060}, {"loss": 0.525, "grad_norm": 1.090205430984497, "learning_rate": 0.0002, "epoch": 6.96305876646997, "step": 43070}, {"loss": 0.5255, "grad_norm": 1.1832911968231201, "learning_rate": 0.0002, "epoch": 6.964675450650716, "step": 43080}, {"loss": 0.5497, "grad_norm": 1.2921082973480225, "learning_rate": 0.0002, "epoch": 6.966292134831461, "step": 43090}, {"loss": 0.5527, "grad_norm": 1.4303096532821655, "learning_rate": 0.0002, "epoch": 6.967908819012206, "step": 43100}, {"loss": 0.4807, "grad_norm": 1.0788004398345947, "learning_rate": 0.0002, "epoch": 6.969525503192951, "step": 43110}, {"loss": 0.5006, "grad_norm": 1.2192047834396362, "learning_rate": 0.0002, "epoch": 6.9711421873736965, "step": 43120}, {"loss": 0.4714, "grad_norm": 1.0735143423080444, "learning_rate": 0.0002, "epoch": 6.972758871554442, "step": 43130}, {"loss": 0.5307, "grad_norm": 1.0317153930664062, "learning_rate": 0.0002, "epoch": 6.974375555735187, "step": 43140}, {"loss": 0.5154, "grad_norm": 1.0926798582077026, "learning_rate": 0.0002, "epoch": 6.975992239915932, "step": 43150}, {"loss": 0.4976, "grad_norm": 1.1660500764846802, "learning_rate": 0.0002, "epoch": 6.977608924096677, "step": 43160}, {"loss": 0.5456, "grad_norm": 1.3945232629776, "learning_rate": 0.0002, "epoch": 6.979225608277423, "step": 43170}, {"loss": 0.4979, "grad_norm": 1.2684587240219116, "learning_rate": 0.0002, "epoch": 6.980842292458169, "step": 43180}, {"loss": 0.5406, "grad_norm": 1.1574004888534546, "learning_rate": 0.0002, "epoch": 6.982458976638913, "step": 43190}, {"loss": 0.5629, "grad_norm": 1.2534198760986328, "learning_rate": 0.0002, "epoch": 6.984075660819659, "step": 43200}, {"loss": 0.5191, "grad_norm": 1.135245442390442, "learning_rate": 0.0002, "epoch": 6.9856923450004045, "step": 43210}, {"loss": 0.548, "grad_norm": 1.3824104070663452, "learning_rate": 0.0002, "epoch": 6.98730902918115, "step": 43220}, {"loss": 0.5294, "grad_norm": 1.2128452062606812, "learning_rate": 0.0002, "epoch": 6.988925713361895, "step": 43230}, {"loss": 0.505, "grad_norm": 1.0795245170593262, "learning_rate": 0.0002, "epoch": 6.99054239754264, "step": 43240}, {"loss": 0.4889, "grad_norm": 1.337353229522705, "learning_rate": 0.0002, "epoch": 6.992159081723385, "step": 43250}, {"loss": 0.4749, "grad_norm": 1.1731765270233154, "learning_rate": 0.0002, "epoch": 6.993775765904131, "step": 43260}, {"loss": 0.4897, "grad_norm": 1.0203192234039307, "learning_rate": 0.0002, "epoch": 6.995392450084876, "step": 43270}, {"loss": 0.5324, "grad_norm": 0.9261201620101929, "learning_rate": 0.0002, "epoch": 6.997009134265621, "step": 43280}, {"loss": 0.5227, "grad_norm": 1.107865810394287, "learning_rate": 0.0002, "epoch": 6.998625818446366, "step": 43290}]}
+{"epoch": 7.999353326327702, "step": 49480, "epoch_duration": 16893.388257026672, "total_accumulated_duration": 135136.67567753792, "gpu_info": {"GPU_0": "NVIDIA A100-PCIE-40GB"}, "memory_usage": {"avg_memory_usage": {"GPU_0": 7568.4541015625}, "peak_memory_usage": {"GPU_0": 13792.75537109375}, "avg_memory_reserved": {"GPU_0": 17416.0}, "peak_memory_reserved": {"GPU_0": 17416.0}, "total_memory": {"GPU_0": 40444.375}}, "best_checkpoint_path": "outputs-001/Meta-Llama-3-8B-Instruct_int4_medmcqa_full_con_lr-0.0002_e-8_seq-512_lora-a-32-d-0.05-r-64_bs-1_gas-2_tf32-True_tunedata-portion-p-0.4-num-51190-sd-42/checkpoint-6185", "params": {"epochs": 8, "batch_size": 1, "learning_rate": 0.0002, "gradient_accumulation_steps": 2, "warmup_ratio": 0.03, "max_grad_norm": 0.3, "lora_alpha": 32, "lora_dropout": 0.05, "lora_r": 64, "tf32": true, "seed": 42}, "log_history": [{"loss": 1.6636, "grad_norm": 0.9894065856933594, "learning_rate": 0.0002, "epoch": 0.0016166841807452913, "step": 10}, {"loss": 1.1528, "grad_norm": 1.7810699939727783, "learning_rate": 0.0002, "epoch": 0.0032333683614905826, "step": 20}, {"loss": 0.9767, "grad_norm": 0.5969577431678772, "learning_rate": 0.0002, "epoch": 0.004850052542235874, "step": 30}, {"loss": 0.9772, "grad_norm": 0.6354120969772339, "learning_rate": 0.0002, "epoch": 0.006466736722981165, "step": 40}, {"loss": 0.8643, "grad_norm": 0.5604607462882996, "learning_rate": 0.0002, "epoch": 0.008083420903726457, "step": 50}, {"loss": 0.8841, "grad_norm": 0.4676193594932556, "learning_rate": 0.0002, "epoch": 0.009700105084471748, "step": 60}, {"loss": 0.9022, "grad_norm": 0.6099211573600769, "learning_rate": 0.0002, "epoch": 0.01131678926521704, "step": 70}, {"loss": 0.9133, "grad_norm": 0.48639994859695435, "learning_rate": 0.0002, "epoch": 0.01293347344596233, "step": 80}, {"loss": 0.8704, "grad_norm": 0.4904264509677887, "learning_rate": 0.0002, "epoch": 0.014550157626707623, "step": 90}, {"loss": 0.8855, "grad_norm": 2.8334362506866455, "learning_rate": 0.0002, "epoch": 0.016166841807452915, "step": 100}, {"loss": 0.8958, "grad_norm": 0.43221670389175415, "learning_rate": 0.0002, "epoch": 0.017783525988198205, "step": 110}, {"loss": 0.8412, "grad_norm": 0.42244166135787964, "learning_rate": 0.0002, "epoch": 0.019400210168943496, "step": 120}, {"loss": 0.8467, "grad_norm": 0.45363298058509827, "learning_rate": 0.0002, "epoch": 0.02101689434968879, "step": 130}, {"loss": 0.8641, "grad_norm": 0.44816508889198303, "learning_rate": 0.0002, "epoch": 0.02263357853043408, "step": 140}, {"loss": 0.8496, "grad_norm": 0.43308213353157043, "learning_rate": 0.0002, "epoch": 0.02425026271117937, "step": 150}, {"loss": 0.8213, "grad_norm": 0.4084763526916504, "learning_rate": 0.0002, "epoch": 0.02586694689192466, "step": 160}, {"loss": 0.8343, "grad_norm": 0.5363703966140747, "learning_rate": 0.0002, "epoch": 0.027483631072669955, "step": 170}, {"loss": 0.8558, "grad_norm": 0.4619699716567993, "learning_rate": 0.0002, "epoch": 0.029100315253415245, "step": 180}, {"loss": 0.8878, "grad_norm": 0.49069908261299133, "learning_rate": 0.0002, "epoch": 0.030716999434160536, "step": 190}, {"loss": 0.8867, "grad_norm": 0.4645835757255554, "learning_rate": 0.0002, "epoch": 0.03233368361490583, "step": 200}, {"loss": 0.8842, "grad_norm": 1.2411243915557861, "learning_rate": 0.0002, "epoch": 0.03395036779565112, "step": 210}, {"loss": 0.8245, "grad_norm": 0.5211851596832275, "learning_rate": 0.0002, "epoch": 0.03556705197639641, "step": 220}, {"loss": 0.8194, "grad_norm": 0.5253691673278809, "learning_rate": 0.0002, "epoch": 0.037183736157141704, "step": 230}, {"loss": 0.8856, "grad_norm": 0.4567478895187378, "learning_rate": 0.0002, "epoch": 0.03880042033788699, "step": 240}, {"loss": 0.838, "grad_norm": 0.5472128391265869, "learning_rate": 0.0002, "epoch": 0.040417104518632285, "step": 250}, {"loss": 0.8201, "grad_norm": 0.42978546023368835, "learning_rate": 0.0002, "epoch": 0.04203378869937758, "step": 260}, {"loss": 0.8334, "grad_norm": 0.601734459400177, "learning_rate": 0.0002, "epoch": 0.043650472880122866, "step": 270}, {"loss": 0.815, "grad_norm": 0.4286513328552246, "learning_rate": 0.0002, "epoch": 0.04526715706086816, "step": 280}, {"loss": 0.8758, "grad_norm": 0.5230861902236938, "learning_rate": 0.0002, "epoch": 0.046883841241613454, "step": 290}, {"loss": 0.8636, "grad_norm": 0.6504611968994141, "learning_rate": 0.0002, "epoch": 0.04850052542235874, "step": 300}, {"loss": 0.8102, "grad_norm": 0.43485215306282043, "learning_rate": 0.0002, "epoch": 0.050117209603104035, "step": 310}, {"loss": 0.8221, "grad_norm": 0.4717007875442505, "learning_rate": 0.0002, "epoch": 0.05173389378384932, "step": 320}, {"loss": 0.8469, "grad_norm": 0.4059787690639496, "learning_rate": 0.0002, "epoch": 0.053350577964594616, "step": 330}, {"loss": 0.8866, "grad_norm": 0.4366913437843323, "learning_rate": 0.0002, "epoch": 0.05496726214533991, "step": 340}, {"loss": 0.7976, "grad_norm": 0.4233848452568054, "learning_rate": 0.0002, "epoch": 0.0565839463260852, "step": 350}, {"loss": 0.8456, "grad_norm": 0.4209108352661133, "learning_rate": 0.0002, "epoch": 0.05820063050683049, "step": 360}, {"loss": 0.816, "grad_norm": 0.41637396812438965, "learning_rate": 0.0002, "epoch": 0.059817314687575784, "step": 370}, {"loss": 0.7976, "grad_norm": 0.46235376596450806, "learning_rate": 0.0002, "epoch": 0.06143399886832107, "step": 380}, {"loss": 0.7966, "grad_norm": 0.4013484716415405, "learning_rate": 0.0002, "epoch": 0.06305068304906636, "step": 390}, {"loss": 0.8253, "grad_norm": 0.47443896532058716, "learning_rate": 0.0002, "epoch": 0.06466736722981166, "step": 400}, {"loss": 0.8666, "grad_norm": 0.3942156434059143, "learning_rate": 0.0002, "epoch": 0.06628405141055695, "step": 410}, {"loss": 0.8402, "grad_norm": 0.4965320825576782, "learning_rate": 0.0002, "epoch": 0.06790073559130223, "step": 420}, {"loss": 0.8317, "grad_norm": 0.4304835796356201, "learning_rate": 0.0002, "epoch": 0.06951741977204753, "step": 430}, {"loss": 0.8528, "grad_norm": 0.511726975440979, "learning_rate": 0.0002, "epoch": 0.07113410395279282, "step": 440}, {"loss": 0.8675, "grad_norm": 0.4040689170360565, "learning_rate": 0.0002, "epoch": 0.07275078813353811, "step": 450}, {"loss": 0.8788, "grad_norm": 0.5402171015739441, "learning_rate": 0.0002, "epoch": 0.07436747231428341, "step": 460}, {"loss": 0.8737, "grad_norm": 0.4174517095088959, "learning_rate": 0.0002, "epoch": 0.0759841564950287, "step": 470}, {"loss": 0.7605, "grad_norm": 0.4306182265281677, "learning_rate": 0.0002, "epoch": 0.07760084067577398, "step": 480}, {"loss": 0.799, "grad_norm": 0.535210132598877, "learning_rate": 0.0002, "epoch": 0.07921752485651928, "step": 490}, {"loss": 0.7825, "grad_norm": 0.5339109897613525, "learning_rate": 0.0002, "epoch": 0.08083420903726457, "step": 500}, {"loss": 0.8985, "grad_norm": 0.45754891633987427, "learning_rate": 0.0002, "epoch": 0.08245089321800986, "step": 510}, {"loss": 0.8144, "grad_norm": 0.43820783495903015, "learning_rate": 0.0002, "epoch": 0.08406757739875516, "step": 520}, {"loss": 0.8001, "grad_norm": 0.4434749186038971, "learning_rate": 0.0002, "epoch": 0.08568426157950045, "step": 530}, {"loss": 0.7857, "grad_norm": 0.43111467361450195, "learning_rate": 0.0002, "epoch": 0.08730094576024573, "step": 540}, {"loss": 0.8418, "grad_norm": 0.4378940165042877, "learning_rate": 0.0002, "epoch": 0.08891762994099103, "step": 550}, {"loss": 0.8361, "grad_norm": 0.4772215187549591, "learning_rate": 0.0002, "epoch": 0.09053431412173632, "step": 560}, {"loss": 0.8268, "grad_norm": 0.6837629079818726, "learning_rate": 0.0002, "epoch": 0.09215099830248161, "step": 570}, {"loss": 0.8607, "grad_norm": 0.42241212725639343, "learning_rate": 0.0002, "epoch": 0.09376768248322691, "step": 580}, {"loss": 0.852, "grad_norm": 0.5165936350822449, "learning_rate": 0.0002, "epoch": 0.0953843666639722, "step": 590}, {"loss": 0.8664, "grad_norm": 0.48737478256225586, "learning_rate": 0.0002, "epoch": 0.09700105084471748, "step": 600}, {"loss": 0.8806, "grad_norm": 0.47419852018356323, "learning_rate": 0.0002, "epoch": 0.09861773502546278, "step": 610}, {"loss": 0.8254, "grad_norm": 0.4975486099720001, "learning_rate": 0.0002, "epoch": 0.10023441920620807, "step": 620}, {"loss": 0.8548, "grad_norm": 0.49123844504356384, "learning_rate": 0.0002, "epoch": 0.10185110338695336, "step": 630}, {"loss": 0.8911, "grad_norm": 0.6288952827453613, "learning_rate": 0.0002, "epoch": 0.10346778756769864, "step": 640}, {"loss": 0.827, "grad_norm": 0.4277345836162567, "learning_rate": 0.0002, "epoch": 0.10508447174844394, "step": 650}, {"loss": 0.7996, "grad_norm": 0.4021061956882477, "learning_rate": 0.0002, "epoch": 0.10670115592918923, "step": 660}, {"loss": 0.87, "grad_norm": 0.3492237329483032, "learning_rate": 0.0002, "epoch": 0.10831784010993452, "step": 670}, {"loss": 0.8698, "grad_norm": 0.4341012239456177, "learning_rate": 0.0002, "epoch": 0.10993452429067982, "step": 680}, {"loss": 0.781, "grad_norm": 0.7296304106712341, "learning_rate": 0.0002, "epoch": 0.1115512084714251, "step": 690}, {"loss": 0.8433, "grad_norm": 0.397494912147522, "learning_rate": 0.0002, "epoch": 0.1131678926521704, "step": 700}, {"loss": 0.827, "grad_norm": 0.396431028842926, "learning_rate": 0.0002, "epoch": 0.1147845768329157, "step": 710}, {"loss": 0.8379, "grad_norm": 0.48842838406562805, "learning_rate": 0.0002, "epoch": 0.11640126101366098, "step": 720}, {"loss": 0.8238, "grad_norm": 0.46322616934776306, "learning_rate": 0.0002, "epoch": 0.11801794519440627, "step": 730}, {"loss": 0.8041, "grad_norm": 0.47990912199020386, "learning_rate": 0.0002, "epoch": 0.11963462937515157, "step": 740}, {"loss": 0.82, "grad_norm": 0.4997142255306244, "learning_rate": 0.0002, "epoch": 0.12125131355589686, "step": 750}, {"loss": 0.7702, "grad_norm": 0.4040526747703552, "learning_rate": 0.0002, "epoch": 0.12286799773664214, "step": 760}, {"loss": 0.863, "grad_norm": 0.453095942735672, "learning_rate": 0.0002, "epoch": 0.12448468191738744, "step": 770}, {"loss": 0.8792, "grad_norm": 0.4636971950531006, "learning_rate": 0.0002, "epoch": 0.12610136609813272, "step": 780}, {"loss": 0.8112, "grad_norm": 0.4279276132583618, "learning_rate": 0.0002, "epoch": 0.12771805027887803, "step": 790}, {"loss": 0.8711, "grad_norm": 0.46212655305862427, "learning_rate": 0.0002, "epoch": 0.12933473445962332, "step": 800}, {"loss": 0.8368, "grad_norm": 0.43127650022506714, "learning_rate": 0.0002, "epoch": 0.1309514186403686, "step": 810}, {"loss": 0.8476, "grad_norm": 0.4201301336288452, "learning_rate": 0.0002, "epoch": 0.1325681028211139, "step": 820}, {"loss": 0.8078, "grad_norm": 0.42583167552948, "learning_rate": 0.0002, "epoch": 0.13418478700185918, "step": 830}, {"loss": 0.8219, "grad_norm": 0.4535622000694275, "learning_rate": 0.0002, "epoch": 0.13580147118260447, "step": 840}, {"loss": 0.8423, "grad_norm": 0.4116036891937256, "learning_rate": 0.0002, "epoch": 0.13741815536334978, "step": 850}, {"loss": 0.8466, "grad_norm": 0.45997580885887146, "learning_rate": 0.0002, "epoch": 0.13903483954409507, "step": 860}, {"loss": 0.8917, "grad_norm": 0.4487837255001068, "learning_rate": 0.0002, "epoch": 0.14065152372484035, "step": 870}, {"loss": 0.8217, "grad_norm": 0.43650057911872864, "learning_rate": 0.0002, "epoch": 0.14226820790558564, "step": 880}, {"loss": 0.8178, "grad_norm": 0.5335358381271362, "learning_rate": 0.0002, "epoch": 0.14388489208633093, "step": 890}, {"loss": 0.7957, "grad_norm": 0.5989000201225281, "learning_rate": 0.0002, "epoch": 0.14550157626707622, "step": 900}, {"loss": 0.8385, "grad_norm": 0.517179012298584, "learning_rate": 0.0002, "epoch": 0.14711826044782153, "step": 910}, {"loss": 0.8255, "grad_norm": 0.44435232877731323, "learning_rate": 0.0002, "epoch": 0.14873494462856682, "step": 920}, {"loss": 0.8305, "grad_norm": 0.42635923624038696, "learning_rate": 0.0002, "epoch": 0.1503516288093121, "step": 930}, {"loss": 0.8043, "grad_norm": 0.49603334069252014, "learning_rate": 0.0002, "epoch": 0.1519683129900574, "step": 940}, {"loss": 0.8377, "grad_norm": 0.40639808773994446, "learning_rate": 0.0002, "epoch": 0.15358499717080268, "step": 950}, {"loss": 0.8529, "grad_norm": 0.4850759208202362, "learning_rate": 0.0002, "epoch": 0.15520168135154797, "step": 960}, {"loss": 0.846, "grad_norm": 0.4427442252635956, "learning_rate": 0.0002, "epoch": 0.15681836553229328, "step": 970}, {"loss": 0.8705, "grad_norm": 0.3760930001735687, "learning_rate": 0.0002, "epoch": 0.15843504971303857, "step": 980}, {"loss": 0.8644, "grad_norm": 0.4794144332408905, "learning_rate": 0.0002, "epoch": 0.16005173389378385, "step": 990}, {"loss": 0.8002, "grad_norm": 0.45828768610954285, "learning_rate": 0.0002, "epoch": 0.16166841807452914, "step": 1000}, {"loss": 0.7658, "grad_norm": 0.6313053369522095, "learning_rate": 0.0002, "epoch": 0.16328510225527443, "step": 1010}, {"loss": 0.8047, "grad_norm": 0.45041006803512573, "learning_rate": 0.0002, "epoch": 0.16490178643601971, "step": 1020}, {"loss": 0.8423, "grad_norm": 0.441403865814209, "learning_rate": 0.0002, "epoch": 0.166518470616765, "step": 1030}, {"loss": 0.8475, "grad_norm": 0.8171296119689941, "learning_rate": 0.0002, "epoch": 0.16813515479751032, "step": 1040}, {"loss": 0.845, "grad_norm": 0.7137420773506165, "learning_rate": 0.0002, "epoch": 0.1697518389782556, "step": 1050}, {"loss": 0.8213, "grad_norm": 0.5236809849739075, "learning_rate": 0.0002, "epoch": 0.1713685231590009, "step": 1060}, {"loss": 0.8265, "grad_norm": 0.5021864175796509, "learning_rate": 0.0002, "epoch": 0.17298520733974618, "step": 1070}, {"loss": 0.8305, "grad_norm": 0.47347521781921387, "learning_rate": 0.0002, "epoch": 0.17460189152049146, "step": 1080}, {"loss": 0.8105, "grad_norm": 0.4631653428077698, "learning_rate": 0.0002, "epoch": 0.17621857570123675, "step": 1090}, {"loss": 0.8166, "grad_norm": 0.49169182777404785, "learning_rate": 0.0002, "epoch": 0.17783525988198207, "step": 1100}, {"loss": 0.8012, "grad_norm": 0.5019739270210266, "learning_rate": 0.0002, "epoch": 0.17945194406272735, "step": 1110}, {"loss": 0.8247, "grad_norm": 0.5100422501564026, "learning_rate": 0.0002, "epoch": 0.18106862824347264, "step": 1120}, {"loss": 0.8142, "grad_norm": 0.3888324499130249, "learning_rate": 0.0002, "epoch": 0.18268531242421793, "step": 1130}, {"loss": 0.8533, "grad_norm": 0.39765217900276184, "learning_rate": 0.0002, "epoch": 0.18430199660496321, "step": 1140}, {"loss": 0.8541, "grad_norm": 0.47190186381340027, "learning_rate": 0.0002, "epoch": 0.1859186807857085, "step": 1150}, {"loss": 0.8301, "grad_norm": 0.4464188814163208, "learning_rate": 0.0002, "epoch": 0.18753536496645382, "step": 1160}, {"loss": 0.8341, "grad_norm": 0.5153930187225342, "learning_rate": 0.0002, "epoch": 0.1891520491471991, "step": 1170}, {"loss": 0.8033, "grad_norm": 0.4779708683490753, "learning_rate": 0.0002, "epoch": 0.1907687333279444, "step": 1180}, {"loss": 0.8187, "grad_norm": 0.4834315776824951, "learning_rate": 0.0002, "epoch": 0.19238541750868968, "step": 1190}, {"loss": 0.7721, "grad_norm": 0.402357816696167, "learning_rate": 0.0002, "epoch": 0.19400210168943496, "step": 1200}, {"loss": 0.7941, "grad_norm": 0.45899084210395813, "learning_rate": 0.0002, "epoch": 0.19561878587018025, "step": 1210}, {"loss": 0.8353, "grad_norm": 0.5106529593467712, "learning_rate": 0.0002, "epoch": 0.19723547005092557, "step": 1220}, {"loss": 0.7816, "grad_norm": 0.45261722803115845, "learning_rate": 0.0002, "epoch": 0.19885215423167085, "step": 1230}, {"loss": 0.8068, "grad_norm": 0.4647127091884613, "learning_rate": 0.0002, "epoch": 0.20046883841241614, "step": 1240}, {"loss": 0.8239, "grad_norm": 0.4849368929862976, "learning_rate": 0.0002, "epoch": 0.20208552259316143, "step": 1250}, {"loss": 0.8514, "grad_norm": 0.4518061578273773, "learning_rate": 0.0002, "epoch": 0.2037022067739067, "step": 1260}, {"loss": 0.8158, "grad_norm": 0.49535325169563293, "learning_rate": 0.0002, "epoch": 0.205318890954652, "step": 1270}, {"loss": 0.8348, "grad_norm": 0.4835205376148224, "learning_rate": 0.0002, "epoch": 0.2069355751353973, "step": 1280}, {"loss": 0.8428, "grad_norm": 0.45308539271354675, "learning_rate": 0.0002, "epoch": 0.2085522593161426, "step": 1290}, {"loss": 0.7993, "grad_norm": 0.5369905233383179, "learning_rate": 0.0002, "epoch": 0.2101689434968879, "step": 1300}, {"loss": 0.8676, "grad_norm": 0.5031622052192688, "learning_rate": 0.0002, "epoch": 0.21178562767763318, "step": 1310}, {"loss": 0.7686, "grad_norm": 0.48010334372520447, "learning_rate": 0.0002, "epoch": 0.21340231185837846, "step": 1320}, {"loss": 0.806, "grad_norm": 0.4905701279640198, "learning_rate": 0.0002, "epoch": 0.21501899603912375, "step": 1330}, {"loss": 0.7885, "grad_norm": 0.43531742691993713, "learning_rate": 0.0002, "epoch": 0.21663568021986904, "step": 1340}, {"loss": 0.8191, "grad_norm": 0.44330692291259766, "learning_rate": 0.0002, "epoch": 0.21825236440061435, "step": 1350}, {"loss": 0.8205, "grad_norm": 0.5384416580200195, "learning_rate": 0.0002, "epoch": 0.21986904858135964, "step": 1360}, {"loss": 0.7726, "grad_norm": 0.4181833863258362, "learning_rate": 0.0002, "epoch": 0.22148573276210493, "step": 1370}, {"loss": 0.8311, "grad_norm": 0.523833692073822, "learning_rate": 0.0002, "epoch": 0.2231024169428502, "step": 1380}, {"loss": 0.7913, "grad_norm": 0.5528736710548401, "learning_rate": 0.0002, "epoch": 0.2247191011235955, "step": 1390}, {"loss": 0.8079, "grad_norm": 0.43515023589134216, "learning_rate": 0.0002, "epoch": 0.2263357853043408, "step": 1400}, {"loss": 0.8403, "grad_norm": 0.48809877038002014, "learning_rate": 0.0002, "epoch": 0.2279524694850861, "step": 1410}, {"loss": 0.8165, "grad_norm": 0.43591251969337463, "learning_rate": 0.0002, "epoch": 0.2295691536658314, "step": 1420}, {"loss": 0.8147, "grad_norm": 0.44625312089920044, "learning_rate": 0.0002, "epoch": 0.23118583784657668, "step": 1430}, {"loss": 0.8134, "grad_norm": 0.4390665292739868, "learning_rate": 0.0002, "epoch": 0.23280252202732196, "step": 1440}, {"loss": 0.8465, "grad_norm": 0.48496049642562866, "learning_rate": 0.0002, "epoch": 0.23441920620806725, "step": 1450}, {"loss": 0.775, "grad_norm": 0.45919957756996155, "learning_rate": 0.0002, "epoch": 0.23603589038881254, "step": 1460}, {"loss": 0.8659, "grad_norm": 0.5471845865249634, "learning_rate": 0.0002, "epoch": 0.23765257456955785, "step": 1470}, {"loss": 0.8164, "grad_norm": 0.47269317507743835, "learning_rate": 0.0002, "epoch": 0.23926925875030314, "step": 1480}, {"loss": 0.854, "grad_norm": 0.4930245578289032, "learning_rate": 0.0002, "epoch": 0.24088594293104842, "step": 1490}, {"loss": 0.8139, "grad_norm": 0.5605630278587341, "learning_rate": 0.0002, "epoch": 0.2425026271117937, "step": 1500}, {"loss": 0.8125, "grad_norm": 0.4435870945453644, "learning_rate": 0.0002, "epoch": 0.244119311292539, "step": 1510}, {"loss": 0.8123, "grad_norm": 0.4941999912261963, "learning_rate": 0.0002, "epoch": 0.24573599547328429, "step": 1520}, {"loss": 0.8427, "grad_norm": 0.5100624561309814, "learning_rate": 0.0002, "epoch": 0.24735267965402957, "step": 1530}, {"loss": 0.8405, "grad_norm": 0.4638267457485199, "learning_rate": 0.0002, "epoch": 0.2489693638347749, "step": 1540}, {"loss": 0.81, "grad_norm": 0.5071570873260498, "learning_rate": 0.0002, "epoch": 0.25058604801552015, "step": 1550}, {"loss": 0.7724, "grad_norm": 0.4291319251060486, "learning_rate": 0.0002, "epoch": 0.25220273219626543, "step": 1560}, {"loss": 0.7984, "grad_norm": 0.5388049483299255, "learning_rate": 0.0002, "epoch": 0.2538194163770108, "step": 1570}, {"loss": 0.8176, "grad_norm": 0.5083683729171753, "learning_rate": 0.0002, "epoch": 0.25543610055775606, "step": 1580}, {"loss": 0.843, "grad_norm": 0.4824463725090027, "learning_rate": 0.0002, "epoch": 0.25705278473850135, "step": 1590}, {"loss": 0.7996, "grad_norm": 0.41177722811698914, "learning_rate": 0.0002, "epoch": 0.25866946891924664, "step": 1600}, {"loss": 0.7772, "grad_norm": 0.5656219124794006, "learning_rate": 0.0002, "epoch": 0.2602861530999919, "step": 1610}, {"loss": 0.7955, "grad_norm": 0.41063204407691956, "learning_rate": 0.0002, "epoch": 0.2619028372807372, "step": 1620}, {"loss": 0.7998, "grad_norm": 0.4897061288356781, "learning_rate": 0.0002, "epoch": 0.2635195214614825, "step": 1630}, {"loss": 0.8198, "grad_norm": 0.4454376697540283, "learning_rate": 0.0002, "epoch": 0.2651362056422278, "step": 1640}, {"loss": 0.8684, "grad_norm": 0.4355238378047943, "learning_rate": 0.0002, "epoch": 0.26675288982297307, "step": 1650}, {"loss": 0.7801, "grad_norm": 0.458310067653656, "learning_rate": 0.0002, "epoch": 0.26836957400371836, "step": 1660}, {"loss": 0.7935, "grad_norm": 0.4752083718776703, "learning_rate": 0.0002, "epoch": 0.26998625818446365, "step": 1670}, {"loss": 0.8267, "grad_norm": 0.4666106402873993, "learning_rate": 0.0002, "epoch": 0.27160294236520893, "step": 1680}, {"loss": 0.8252, "grad_norm": 0.4213818609714508, "learning_rate": 0.0002, "epoch": 0.2732196265459543, "step": 1690}, {"loss": 0.8559, "grad_norm": 0.5768913626670837, "learning_rate": 0.0002, "epoch": 0.27483631072669956, "step": 1700}, {"loss": 0.7931, "grad_norm": 0.4209914803504944, "learning_rate": 0.0002, "epoch": 0.27645299490744485, "step": 1710}, {"loss": 0.8167, "grad_norm": 0.501909613609314, "learning_rate": 0.0002, "epoch": 0.27806967908819014, "step": 1720}, {"loss": 0.7832, "grad_norm": 0.5266261100769043, "learning_rate": 0.0002, "epoch": 0.2796863632689354, "step": 1730}, {"loss": 0.8102, "grad_norm": 0.43806859850883484, "learning_rate": 0.0002, "epoch": 0.2813030474496807, "step": 1740}, {"loss": 0.8157, "grad_norm": 0.46048814058303833, "learning_rate": 0.0002, "epoch": 0.282919731630426, "step": 1750}, {"loss": 0.8596, "grad_norm": 0.44972819089889526, "learning_rate": 0.0002, "epoch": 0.2845364158111713, "step": 1760}, {"loss": 0.8421, "grad_norm": 0.5114831328392029, "learning_rate": 0.0002, "epoch": 0.28615309999191657, "step": 1770}, {"loss": 0.8361, "grad_norm": 0.47931742668151855, "learning_rate": 0.0002, "epoch": 0.28776978417266186, "step": 1780}, {"loss": 0.8265, "grad_norm": 0.5092599987983704, "learning_rate": 0.0002, "epoch": 0.28938646835340714, "step": 1790}, {"loss": 0.8506, "grad_norm": 0.37581443786621094, "learning_rate": 0.0002, "epoch": 0.29100315253415243, "step": 1800}, {"loss": 0.7932, "grad_norm": 0.47097381949424744, "learning_rate": 0.0002, "epoch": 0.2926198367148977, "step": 1810}, {"loss": 0.7787, "grad_norm": 0.48300236463546753, "learning_rate": 0.0002, "epoch": 0.29423652089564306, "step": 1820}, {"loss": 0.8391, "grad_norm": 0.5600419640541077, "learning_rate": 0.0002, "epoch": 0.29585320507638835, "step": 1830}, {"loss": 0.8507, "grad_norm": 0.48555272817611694, "learning_rate": 0.0002, "epoch": 0.29746988925713364, "step": 1840}, {"loss": 0.7657, "grad_norm": 0.3752668499946594, "learning_rate": 0.0002, "epoch": 0.2990865734378789, "step": 1850}, {"loss": 0.7915, "grad_norm": 0.5328747034072876, "learning_rate": 0.0002, "epoch": 0.3007032576186242, "step": 1860}, {"loss": 0.8426, "grad_norm": 0.48716455698013306, "learning_rate": 0.0002, "epoch": 0.3023199417993695, "step": 1870}, {"loss": 0.8335, "grad_norm": 0.5011493563652039, "learning_rate": 0.0002, "epoch": 0.3039366259801148, "step": 1880}, {"loss": 0.852, "grad_norm": 0.46461427211761475, "learning_rate": 0.0002, "epoch": 0.30555331016086007, "step": 1890}, {"loss": 0.8478, "grad_norm": 0.36630210280418396, "learning_rate": 0.0002, "epoch": 0.30716999434160536, "step": 1900}, {"loss": 0.8162, "grad_norm": 0.4217296242713928, "learning_rate": 0.0002, "epoch": 0.30878667852235064, "step": 1910}, {"loss": 0.8128, "grad_norm": 0.4394875466823578, "learning_rate": 0.0002, "epoch": 0.31040336270309593, "step": 1920}, {"loss": 0.8471, "grad_norm": 0.6587965488433838, "learning_rate": 0.0002, "epoch": 0.3120200468838412, "step": 1930}, {"loss": 0.8565, "grad_norm": 0.5469298958778381, "learning_rate": 0.0002, "epoch": 0.31363673106458656, "step": 1940}, {"loss": 0.8236, "grad_norm": 0.4371595084667206, "learning_rate": 0.0002, "epoch": 0.31525341524533185, "step": 1950}, {"loss": 0.887, "grad_norm": 0.4809541404247284, "learning_rate": 0.0002, "epoch": 0.31687009942607713, "step": 1960}, {"loss": 0.7855, "grad_norm": 0.6061086654663086, "learning_rate": 0.0002, "epoch": 0.3184867836068224, "step": 1970}, {"loss": 0.7679, "grad_norm": 0.5342657566070557, "learning_rate": 0.0002, "epoch": 0.3201034677875677, "step": 1980}, {"loss": 0.7955, "grad_norm": 0.5057743787765503, "learning_rate": 0.0002, "epoch": 0.321720151968313, "step": 1990}, {"loss": 0.7774, "grad_norm": 0.528626024723053, "learning_rate": 0.0002, "epoch": 0.3233368361490583, "step": 2000}, {"loss": 0.8845, "grad_norm": 0.46742770075798035, "learning_rate": 0.0002, "epoch": 0.32495352032980357, "step": 2010}, {"loss": 0.8484, "grad_norm": 0.515101432800293, "learning_rate": 0.0002, "epoch": 0.32657020451054886, "step": 2020}, {"loss": 0.8139, "grad_norm": 0.41941216588020325, "learning_rate": 0.0002, "epoch": 0.32818688869129414, "step": 2030}, {"loss": 0.7637, "grad_norm": 0.49902522563934326, "learning_rate": 0.0002, "epoch": 0.32980357287203943, "step": 2040}, {"loss": 0.7822, "grad_norm": 0.4120897650718689, "learning_rate": 0.0002, "epoch": 0.3314202570527847, "step": 2050}, {"loss": 0.8057, "grad_norm": 0.45352041721343994, "learning_rate": 0.0002, "epoch": 0.33303694123353, "step": 2060}, {"loss": 0.7913, "grad_norm": 0.523199737071991, "learning_rate": 0.0002, "epoch": 0.33465362541427535, "step": 2070}, {"loss": 0.8036, "grad_norm": 0.4390358626842499, "learning_rate": 0.0002, "epoch": 0.33627030959502063, "step": 2080}, {"loss": 0.8145, "grad_norm": 0.6752901077270508, "learning_rate": 0.0002, "epoch": 0.3378869937757659, "step": 2090}, {"loss": 0.7807, "grad_norm": 0.547821044921875, "learning_rate": 0.0002, "epoch": 0.3395036779565112, "step": 2100}, {"loss": 0.8561, "grad_norm": 0.5161308646202087, "learning_rate": 0.0002, "epoch": 0.3411203621372565, "step": 2110}, {"loss": 0.7697, "grad_norm": 0.4565401077270508, "learning_rate": 0.0002, "epoch": 0.3427370463180018, "step": 2120}, {"loss": 0.7964, "grad_norm": 0.4666115939617157, "learning_rate": 0.0002, "epoch": 0.34435373049874707, "step": 2130}, {"loss": 0.8189, "grad_norm": 0.4090428352355957, "learning_rate": 0.0002, "epoch": 0.34597041467949236, "step": 2140}, {"loss": 0.8817, "grad_norm": 0.510845422744751, "learning_rate": 0.0002, "epoch": 0.34758709886023764, "step": 2150}, {"loss": 0.8398, "grad_norm": 0.42861923575401306, "learning_rate": 0.0002, "epoch": 0.34920378304098293, "step": 2160}, {"loss": 0.7716, "grad_norm": 0.4476332664489746, "learning_rate": 0.0002, "epoch": 0.3508204672217282, "step": 2170}, {"loss": 0.7845, "grad_norm": 0.6065791249275208, "learning_rate": 0.0002, "epoch": 0.3524371514024735, "step": 2180}, {"loss": 0.8187, "grad_norm": 0.42335066199302673, "learning_rate": 0.0002, "epoch": 0.35405383558321885, "step": 2190}, {"loss": 0.8239, "grad_norm": 0.5094629526138306, "learning_rate": 0.0002, "epoch": 0.35567051976396413, "step": 2200}, {"loss": 0.7807, "grad_norm": 0.5476373434066772, "learning_rate": 0.0002, "epoch": 0.3572872039447094, "step": 2210}, {"loss": 0.814, "grad_norm": 0.3911719024181366, "learning_rate": 0.0002, "epoch": 0.3589038881254547, "step": 2220}, {"loss": 0.8599, "grad_norm": 0.6599636077880859, "learning_rate": 0.0002, "epoch": 0.3605205723062, "step": 2230}, {"loss": 0.7482, "grad_norm": 0.40381914377212524, "learning_rate": 0.0002, "epoch": 0.3621372564869453, "step": 2240}, {"loss": 0.7772, "grad_norm": 0.4433908462524414, "learning_rate": 0.0002, "epoch": 0.36375394066769057, "step": 2250}, {"loss": 0.8503, "grad_norm": 0.578326940536499, "learning_rate": 0.0002, "epoch": 0.36537062484843585, "step": 2260}, {"loss": 0.8178, "grad_norm": 0.5734784007072449, "learning_rate": 0.0002, "epoch": 0.36698730902918114, "step": 2270}, {"loss": 0.8193, "grad_norm": 0.45555487275123596, "learning_rate": 0.0002, "epoch": 0.36860399320992643, "step": 2280}, {"loss": 0.7929, "grad_norm": 0.5666276216506958, "learning_rate": 0.0002, "epoch": 0.3702206773906717, "step": 2290}, {"loss": 0.8292, "grad_norm": 0.5461117625236511, "learning_rate": 0.0002, "epoch": 0.371837361571417, "step": 2300}, {"loss": 0.8204, "grad_norm": 0.6318911910057068, "learning_rate": 0.0002, "epoch": 0.3734540457521623, "step": 2310}, {"loss": 0.7964, "grad_norm": 0.493263304233551, "learning_rate": 0.0002, "epoch": 0.37507072993290763, "step": 2320}, {"loss": 0.8339, "grad_norm": 0.5888760089874268, "learning_rate": 0.0002, "epoch": 0.3766874141136529, "step": 2330}, {"loss": 0.7737, "grad_norm": 0.48671841621398926, "learning_rate": 0.0002, "epoch": 0.3783040982943982, "step": 2340}, {"loss": 0.8367, "grad_norm": 0.4385145306587219, "learning_rate": 0.0002, "epoch": 0.3799207824751435, "step": 2350}, {"loss": 0.812, "grad_norm": 0.5523318648338318, "learning_rate": 0.0002, "epoch": 0.3815374666558888, "step": 2360}, {"loss": 0.8351, "grad_norm": 0.7308220267295837, "learning_rate": 0.0002, "epoch": 0.38315415083663407, "step": 2370}, {"loss": 0.859, "grad_norm": 0.554214358329773, "learning_rate": 0.0002, "epoch": 0.38477083501737935, "step": 2380}, {"loss": 0.8146, "grad_norm": 0.5425800085067749, "learning_rate": 0.0002, "epoch": 0.38638751919812464, "step": 2390}, {"loss": 0.8282, "grad_norm": 0.48811158537864685, "learning_rate": 0.0002, "epoch": 0.3880042033788699, "step": 2400}, {"loss": 0.8074, "grad_norm": 0.49212366342544556, "learning_rate": 0.0002, "epoch": 0.3896208875596152, "step": 2410}, {"loss": 0.7991, "grad_norm": 0.5222218632698059, "learning_rate": 0.0002, "epoch": 0.3912375717403605, "step": 2420}, {"loss": 0.8182, "grad_norm": 0.4699819087982178, "learning_rate": 0.0002, "epoch": 0.3928542559211058, "step": 2430}, {"loss": 0.7919, "grad_norm": 0.46153587102890015, "learning_rate": 0.0002, "epoch": 0.39447094010185113, "step": 2440}, {"loss": 0.8111, "grad_norm": 0.4150611162185669, "learning_rate": 0.0002, "epoch": 0.3960876242825964, "step": 2450}, {"loss": 0.8589, "grad_norm": 0.5799614787101746, "learning_rate": 0.0002, "epoch": 0.3977043084633417, "step": 2460}, {"loss": 0.8085, "grad_norm": 0.56536865234375, "learning_rate": 0.0002, "epoch": 0.399320992644087, "step": 2470}, {"loss": 0.8022, "grad_norm": 0.5451247096061707, "learning_rate": 0.0002, "epoch": 0.4009376768248323, "step": 2480}, {"loss": 0.8217, "grad_norm": 0.5914521217346191, "learning_rate": 0.0002, "epoch": 0.40255436100557757, "step": 2490}, {"loss": 0.7859, "grad_norm": 0.4428117275238037, "learning_rate": 0.0002, "epoch": 0.40417104518632285, "step": 2500}, {"loss": 0.8054, "grad_norm": 0.48580947518348694, "learning_rate": 0.0002, "epoch": 0.40578772936706814, "step": 2510}, {"loss": 0.8405, "grad_norm": 0.436734676361084, "learning_rate": 0.0002, "epoch": 0.4074044135478134, "step": 2520}, {"loss": 0.8209, "grad_norm": 0.5752223134040833, "learning_rate": 0.0002, "epoch": 0.4090210977285587, "step": 2530}, {"loss": 0.8181, "grad_norm": 0.4271308183670044, "learning_rate": 0.0002, "epoch": 0.410637781909304, "step": 2540}, {"loss": 0.8058, "grad_norm": 0.46294718980789185, "learning_rate": 0.0002, "epoch": 0.4122544660900493, "step": 2550}, {"loss": 0.8473, "grad_norm": 0.49407583475112915, "learning_rate": 0.0002, "epoch": 0.4138711502707946, "step": 2560}, {"loss": 0.7881, "grad_norm": 0.4729035496711731, "learning_rate": 0.0002, "epoch": 0.4154878344515399, "step": 2570}, {"loss": 0.7834, "grad_norm": 0.4129747152328491, "learning_rate": 0.0002, "epoch": 0.4171045186322852, "step": 2580}, {"loss": 0.7859, "grad_norm": 0.5684236288070679, "learning_rate": 0.0002, "epoch": 0.4187212028130305, "step": 2590}, {"loss": 0.811, "grad_norm": 0.4862157106399536, "learning_rate": 0.0002, "epoch": 0.4203378869937758, "step": 2600}, {"loss": 0.7582, "grad_norm": 0.46567976474761963, "learning_rate": 0.0002, "epoch": 0.42195457117452106, "step": 2610}, {"loss": 0.7755, "grad_norm": 0.5710650682449341, "learning_rate": 0.0002, "epoch": 0.42357125535526635, "step": 2620}, {"loss": 0.8573, "grad_norm": 0.5660041570663452, "learning_rate": 0.0002, "epoch": 0.42518793953601164, "step": 2630}, {"loss": 0.7812, "grad_norm": 0.47944375872612, "learning_rate": 0.0002, "epoch": 0.4268046237167569, "step": 2640}, {"loss": 0.7459, "grad_norm": 0.537223756313324, "learning_rate": 0.0002, "epoch": 0.4284213078975022, "step": 2650}, {"loss": 0.8246, "grad_norm": 0.41669997572898865, "learning_rate": 0.0002, "epoch": 0.4300379920782475, "step": 2660}, {"loss": 0.7785, "grad_norm": 0.44727686047554016, "learning_rate": 0.0002, "epoch": 0.4316546762589928, "step": 2670}, {"loss": 0.8241, "grad_norm": 0.5600888729095459, "learning_rate": 0.0002, "epoch": 0.4332713604397381, "step": 2680}, {"loss": 0.7708, "grad_norm": 0.39820605516433716, "learning_rate": 0.0002, "epoch": 0.4348880446204834, "step": 2690}, {"loss": 0.8202, "grad_norm": 0.5637655854225159, "learning_rate": 0.0002, "epoch": 0.4365047288012287, "step": 2700}, {"loss": 0.855, "grad_norm": 0.6363666653633118, "learning_rate": 0.0002, "epoch": 0.438121412981974, "step": 2710}, {"loss": 0.8468, "grad_norm": 0.5656129121780396, "learning_rate": 0.0002, "epoch": 0.4397380971627193, "step": 2720}, {"loss": 0.7845, "grad_norm": 0.5600156188011169, "learning_rate": 0.0002, "epoch": 0.44135478134346456, "step": 2730}, {"loss": 0.8405, "grad_norm": 0.5506579875946045, "learning_rate": 0.0002, "epoch": 0.44297146552420985, "step": 2740}, {"loss": 0.7725, "grad_norm": 0.49878305196762085, "learning_rate": 0.0002, "epoch": 0.44458814970495514, "step": 2750}, {"loss": 0.8292, "grad_norm": 0.4569213092327118, "learning_rate": 0.0002, "epoch": 0.4462048338857004, "step": 2760}, {"loss": 0.8028, "grad_norm": 0.6056680083274841, "learning_rate": 0.0002, "epoch": 0.4478215180664457, "step": 2770}, {"loss": 0.8242, "grad_norm": 0.44474557042121887, "learning_rate": 0.0002, "epoch": 0.449438202247191, "step": 2780}, {"loss": 0.801, "grad_norm": 0.46055394411087036, "learning_rate": 0.0002, "epoch": 0.4510548864279363, "step": 2790}, {"loss": 0.7521, "grad_norm": 0.4904133379459381, "learning_rate": 0.0002, "epoch": 0.4526715706086816, "step": 2800}, {"loss": 0.8829, "grad_norm": 0.5647031664848328, "learning_rate": 0.0002, "epoch": 0.45428825478942686, "step": 2810}, {"loss": 0.8622, "grad_norm": 0.5759473443031311, "learning_rate": 0.0002, "epoch": 0.4559049389701722, "step": 2820}, {"loss": 0.7812, "grad_norm": 0.5161895751953125, "learning_rate": 0.0002, "epoch": 0.4575216231509175, "step": 2830}, {"loss": 0.8045, "grad_norm": 0.4248254597187042, "learning_rate": 0.0002, "epoch": 0.4591383073316628, "step": 2840}, {"loss": 0.7838, "grad_norm": 0.45395001769065857, "learning_rate": 0.0002, "epoch": 0.46075499151240806, "step": 2850}, {"loss": 0.8208, "grad_norm": 0.5358697772026062, "learning_rate": 0.0002, "epoch": 0.46237167569315335, "step": 2860}, {"loss": 0.8147, "grad_norm": 0.5379165410995483, "learning_rate": 0.0002, "epoch": 0.46398835987389864, "step": 2870}, {"loss": 0.7403, "grad_norm": 0.4601989686489105, "learning_rate": 0.0002, "epoch": 0.4656050440546439, "step": 2880}, {"loss": 0.8523, "grad_norm": 0.671115517616272, "learning_rate": 0.0002, "epoch": 0.4672217282353892, "step": 2890}, {"loss": 0.8262, "grad_norm": 0.4425133168697357, "learning_rate": 0.0002, "epoch": 0.4688384124161345, "step": 2900}, {"loss": 0.8178, "grad_norm": 0.5446155071258545, "learning_rate": 0.0002, "epoch": 0.4704550965968798, "step": 2910}, {"loss": 0.8106, "grad_norm": 0.603306233882904, "learning_rate": 0.0002, "epoch": 0.47207178077762507, "step": 2920}, {"loss": 0.8044, "grad_norm": 0.5377997159957886, "learning_rate": 0.0002, "epoch": 0.47368846495837036, "step": 2930}, {"loss": 0.8075, "grad_norm": 0.4931027591228485, "learning_rate": 0.0002, "epoch": 0.4753051491391157, "step": 2940}, {"loss": 0.8004, "grad_norm": 0.4711960256099701, "learning_rate": 0.0002, "epoch": 0.476921833319861, "step": 2950}, {"loss": 0.8121, "grad_norm": 0.5020492672920227, "learning_rate": 0.0002, "epoch": 0.4785385175006063, "step": 2960}, {"loss": 0.8221, "grad_norm": 0.5428946614265442, "learning_rate": 0.0002, "epoch": 0.48015520168135156, "step": 2970}, {"loss": 0.7849, "grad_norm": 0.5294089317321777, "learning_rate": 0.0002, "epoch": 0.48177188586209685, "step": 2980}, {"loss": 0.8553, "grad_norm": 0.648289144039154, "learning_rate": 0.0002, "epoch": 0.48338857004284214, "step": 2990}, {"loss": 0.7874, "grad_norm": 0.47916680574417114, "learning_rate": 0.0002, "epoch": 0.4850052542235874, "step": 3000}, {"loss": 0.8087, "grad_norm": 0.43849772214889526, "learning_rate": 0.0002, "epoch": 0.4866219384043327, "step": 3010}, {"loss": 0.7662, "grad_norm": 0.47007861733436584, "learning_rate": 0.0002, "epoch": 0.488238622585078, "step": 3020}, {"loss": 0.757, "grad_norm": 0.6314331293106079, "learning_rate": 0.0002, "epoch": 0.4898553067658233, "step": 3030}, {"loss": 0.7863, "grad_norm": 0.49211493134498596, "learning_rate": 0.0002, "epoch": 0.49147199094656857, "step": 3040}, {"loss": 0.8335, "grad_norm": 0.4537973403930664, "learning_rate": 0.0002, "epoch": 0.49308867512731386, "step": 3050}, {"loss": 0.8095, "grad_norm": 0.47326919436454773, "learning_rate": 0.0002, "epoch": 0.49470535930805914, "step": 3060}, {"loss": 0.8447, "grad_norm": 0.525874137878418, "learning_rate": 0.0002, "epoch": 0.4963220434888045, "step": 3070}, {"loss": 0.8339, "grad_norm": 0.6361091732978821, "learning_rate": 0.0002, "epoch": 0.4979387276695498, "step": 3080}, {"loss": 0.821, "grad_norm": 0.5850642919540405, "learning_rate": 0.0002, "epoch": 0.49955541185029506, "step": 3090}, {"loss": 0.8279, "grad_norm": 0.47299543023109436, "learning_rate": 0.0002, "epoch": 0.5011720960310403, "step": 3100}, {"loss": 0.8681, "grad_norm": 0.473099946975708, "learning_rate": 0.0002, "epoch": 0.5027887802117856, "step": 3110}, {"loss": 0.8223, "grad_norm": 0.48186397552490234, "learning_rate": 0.0002, "epoch": 0.5044054643925309, "step": 3120}, {"loss": 0.8292, "grad_norm": 0.5015401840209961, "learning_rate": 0.0002, "epoch": 0.5060221485732762, "step": 3130}, {"loss": 0.7692, "grad_norm": 0.5617750287055969, "learning_rate": 0.0002, "epoch": 0.5076388327540216, "step": 3140}, {"loss": 0.8708, "grad_norm": 0.5169327259063721, "learning_rate": 0.0002, "epoch": 0.5092555169347668, "step": 3150}, {"loss": 0.7845, "grad_norm": 0.545657753944397, "learning_rate": 0.0002, "epoch": 0.5108722011155121, "step": 3160}, {"loss": 0.799, "grad_norm": 0.512864351272583, "learning_rate": 0.0002, "epoch": 0.5124888852962574, "step": 3170}, {"loss": 0.7794, "grad_norm": 0.4113546311855316, "learning_rate": 0.0002, "epoch": 0.5141055694770027, "step": 3180}, {"loss": 0.8206, "grad_norm": 0.44532445073127747, "learning_rate": 0.0002, "epoch": 0.5157222536577479, "step": 3190}, {"loss": 0.8213, "grad_norm": 0.5623497366905212, "learning_rate": 0.0002, "epoch": 0.5173389378384933, "step": 3200}, {"loss": 0.7928, "grad_norm": 0.5084741115570068, "learning_rate": 0.0002, "epoch": 0.5189556220192385, "step": 3210}, {"loss": 0.8174, "grad_norm": 0.5305403470993042, "learning_rate": 0.0002, "epoch": 0.5205723061999838, "step": 3220}, {"loss": 0.8139, "grad_norm": 0.4708254337310791, "learning_rate": 0.0002, "epoch": 0.5221889903807291, "step": 3230}, {"loss": 0.7639, "grad_norm": 0.43827131390571594, "learning_rate": 0.0002, "epoch": 0.5238056745614744, "step": 3240}, {"loss": 0.7993, "grad_norm": 0.5630002617835999, "learning_rate": 0.0002, "epoch": 0.5254223587422197, "step": 3250}, {"loss": 0.7522, "grad_norm": 0.5010961890220642, "learning_rate": 0.0002, "epoch": 0.527039042922965, "step": 3260}, {"loss": 0.8374, "grad_norm": 0.6303122043609619, "learning_rate": 0.0002, "epoch": 0.5286557271037103, "step": 3270}, {"loss": 0.7727, "grad_norm": 0.5107331275939941, "learning_rate": 0.0002, "epoch": 0.5302724112844556, "step": 3280}, {"loss": 0.8495, "grad_norm": 0.5700443387031555, "learning_rate": 0.0002, "epoch": 0.5318890954652009, "step": 3290}, {"loss": 0.7776, "grad_norm": 0.46296367049217224, "learning_rate": 0.0002, "epoch": 0.5335057796459461, "step": 3300}, {"loss": 0.7931, "grad_norm": 0.531568706035614, "learning_rate": 0.0002, "epoch": 0.5351224638266915, "step": 3310}, {"loss": 0.843, "grad_norm": 0.4686741530895233, "learning_rate": 0.0002, "epoch": 0.5367391480074367, "step": 3320}, {"loss": 0.8104, "grad_norm": 0.5404331088066101, "learning_rate": 0.0002, "epoch": 0.5383558321881821, "step": 3330}, {"loss": 0.7686, "grad_norm": 0.6368790864944458, "learning_rate": 0.0002, "epoch": 0.5399725163689273, "step": 3340}, {"loss": 0.8514, "grad_norm": 0.42300888895988464, "learning_rate": 0.0002, "epoch": 0.5415892005496726, "step": 3350}, {"loss": 0.8236, "grad_norm": 0.5362542867660522, "learning_rate": 0.0002, "epoch": 0.5432058847304179, "step": 3360}, {"loss": 0.858, "grad_norm": 0.497128963470459, "learning_rate": 0.0002, "epoch": 0.5448225689111632, "step": 3370}, {"loss": 0.8519, "grad_norm": 0.5006386041641235, "learning_rate": 0.0002, "epoch": 0.5464392530919085, "step": 3380}, {"loss": 0.7867, "grad_norm": 0.44136837124824524, "learning_rate": 0.0002, "epoch": 0.5480559372726538, "step": 3390}, {"loss": 0.773, "grad_norm": 0.5897833108901978, "learning_rate": 0.0002, "epoch": 0.5496726214533991, "step": 3400}, {"loss": 0.8895, "grad_norm": 0.641075611114502, "learning_rate": 0.0002, "epoch": 0.5512893056341444, "step": 3410}, {"loss": 0.7827, "grad_norm": 0.7251322269439697, "learning_rate": 0.0002, "epoch": 0.5529059898148897, "step": 3420}, {"loss": 0.7626, "grad_norm": 0.47411349415779114, "learning_rate": 0.0002, "epoch": 0.5545226739956349, "step": 3430}, {"loss": 0.8196, "grad_norm": 0.4994310438632965, "learning_rate": 0.0002, "epoch": 0.5561393581763803, "step": 3440}, {"loss": 0.7812, "grad_norm": 0.5814438462257385, "learning_rate": 0.0002, "epoch": 0.5577560423571255, "step": 3450}, {"loss": 0.8805, "grad_norm": 0.6278898119926453, "learning_rate": 0.0002, "epoch": 0.5593727265378708, "step": 3460}, {"loss": 0.813, "grad_norm": 0.46208274364471436, "learning_rate": 0.0002, "epoch": 0.5609894107186161, "step": 3470}, {"loss": 0.8295, "grad_norm": 0.5718930959701538, "learning_rate": 0.0002, "epoch": 0.5626060948993614, "step": 3480}, {"loss": 0.8152, "grad_norm": 0.48178744316101074, "learning_rate": 0.0002, "epoch": 0.5642227790801067, "step": 3490}, {"loss": 0.8244, "grad_norm": 0.47336965799331665, "learning_rate": 0.0002, "epoch": 0.565839463260852, "step": 3500}, {"loss": 0.8099, "grad_norm": 0.43442684412002563, "learning_rate": 0.0002, "epoch": 0.5674561474415973, "step": 3510}, {"loss": 0.7564, "grad_norm": 0.6463358998298645, "learning_rate": 0.0002, "epoch": 0.5690728316223426, "step": 3520}, {"loss": 0.836, "grad_norm": 0.5286486744880676, "learning_rate": 0.0002, "epoch": 0.5706895158030879, "step": 3530}, {"loss": 0.8421, "grad_norm": 0.5405499935150146, "learning_rate": 0.0002, "epoch": 0.5723061999838331, "step": 3540}, {"loss": 0.7614, "grad_norm": 0.6654391884803772, "learning_rate": 0.0002, "epoch": 0.5739228841645785, "step": 3550}, {"loss": 0.7803, "grad_norm": 0.5081980228424072, "learning_rate": 0.0002, "epoch": 0.5755395683453237, "step": 3560}, {"loss": 0.7753, "grad_norm": 0.48978179693222046, "learning_rate": 0.0002, "epoch": 0.5771562525260691, "step": 3570}, {"loss": 0.8151, "grad_norm": 0.5840612053871155, "learning_rate": 0.0002, "epoch": 0.5787729367068143, "step": 3580}, {"loss": 0.8937, "grad_norm": 0.5235261917114258, "learning_rate": 0.0002, "epoch": 0.5803896208875596, "step": 3590}, {"loss": 0.7894, "grad_norm": 0.5672075748443604, "learning_rate": 0.0002, "epoch": 0.5820063050683049, "step": 3600}, {"loss": 0.8347, "grad_norm": 0.5613429546356201, "learning_rate": 0.0002, "epoch": 0.5836229892490502, "step": 3610}, {"loss": 0.8274, "grad_norm": 0.4032273590564728, "learning_rate": 0.0002, "epoch": 0.5852396734297954, "step": 3620}, {"loss": 0.8421, "grad_norm": 0.49559324979782104, "learning_rate": 0.0002, "epoch": 0.5868563576105408, "step": 3630}, {"loss": 0.8332, "grad_norm": 0.6895697712898254, "learning_rate": 0.0002, "epoch": 0.5884730417912861, "step": 3640}, {"loss": 0.7877, "grad_norm": 0.4750136435031891, "learning_rate": 0.0002, "epoch": 0.5900897259720314, "step": 3650}, {"loss": 0.8219, "grad_norm": 0.5176819562911987, "learning_rate": 0.0002, "epoch": 0.5917064101527767, "step": 3660}, {"loss": 0.8151, "grad_norm": 0.5817760229110718, "learning_rate": 0.0002, "epoch": 0.5933230943335219, "step": 3670}, {"loss": 0.7823, "grad_norm": 0.6064626574516296, "learning_rate": 0.0002, "epoch": 0.5949397785142673, "step": 3680}, {"loss": 0.8422, "grad_norm": 0.6728700995445251, "learning_rate": 0.0002, "epoch": 0.5965564626950125, "step": 3690}, {"loss": 0.7679, "grad_norm": 0.609305202960968, "learning_rate": 0.0002, "epoch": 0.5981731468757578, "step": 3700}, {"loss": 0.8048, "grad_norm": 0.4615488350391388, "learning_rate": 0.0002, "epoch": 0.5997898310565031, "step": 3710}, {"loss": 0.8214, "grad_norm": 2.0531179904937744, "learning_rate": 0.0002, "epoch": 0.6014065152372484, "step": 3720}, {"loss": 0.8158, "grad_norm": 0.5091132521629333, "learning_rate": 0.0002, "epoch": 0.6030231994179936, "step": 3730}, {"loss": 0.7833, "grad_norm": 0.5951124429702759, "learning_rate": 0.0002, "epoch": 0.604639883598739, "step": 3740}, {"loss": 0.7784, "grad_norm": 0.5870208144187927, "learning_rate": 0.0002, "epoch": 0.6062565677794842, "step": 3750}, {"loss": 0.8044, "grad_norm": 0.6254619359970093, "learning_rate": 0.0002, "epoch": 0.6078732519602296, "step": 3760}, {"loss": 0.7868, "grad_norm": 0.5577626824378967, "learning_rate": 0.0002, "epoch": 0.6094899361409749, "step": 3770}, {"loss": 0.8108, "grad_norm": 0.5004405379295349, "learning_rate": 0.0002, "epoch": 0.6111066203217201, "step": 3780}, {"loss": 0.8092, "grad_norm": 0.5527383685112, "learning_rate": 0.0002, "epoch": 0.6127233045024655, "step": 3790}, {"loss": 0.8036, "grad_norm": 0.49116113781929016, "learning_rate": 0.0002, "epoch": 0.6143399886832107, "step": 3800}, {"loss": 0.8352, "grad_norm": 0.5299299359321594, "learning_rate": 0.0002, "epoch": 0.6159566728639561, "step": 3810}, {"loss": 0.7737, "grad_norm": 0.464897483587265, "learning_rate": 0.0002, "epoch": 0.6175733570447013, "step": 3820}, {"loss": 0.7923, "grad_norm": 0.6505740880966187, "learning_rate": 0.0002, "epoch": 0.6191900412254466, "step": 3830}, {"loss": 0.8123, "grad_norm": 0.5512559413909912, "learning_rate": 0.0002, "epoch": 0.6208067254061919, "step": 3840}, {"loss": 0.8856, "grad_norm": 0.49427518248558044, "learning_rate": 0.0002, "epoch": 0.6224234095869372, "step": 3850}, {"loss": 0.7751, "grad_norm": 0.3839147090911865, "learning_rate": 0.0002, "epoch": 0.6240400937676824, "step": 3860}, {"loss": 0.8006, "grad_norm": 0.5760218501091003, "learning_rate": 0.0002, "epoch": 0.6256567779484278, "step": 3870}, {"loss": 0.7836, "grad_norm": 0.7226507067680359, "learning_rate": 0.0002, "epoch": 0.6272734621291731, "step": 3880}, {"loss": 0.8244, "grad_norm": 0.676781415939331, "learning_rate": 0.0002, "epoch": 0.6288901463099184, "step": 3890}, {"loss": 0.8239, "grad_norm": 0.4284018278121948, "learning_rate": 0.0002, "epoch": 0.6305068304906637, "step": 3900}, {"loss": 0.7996, "grad_norm": 0.5060628056526184, "learning_rate": 0.0002, "epoch": 0.6321235146714089, "step": 3910}, {"loss": 0.8089, "grad_norm": 0.5524522066116333, "learning_rate": 0.0002, "epoch": 0.6337401988521543, "step": 3920}, {"loss": 0.8276, "grad_norm": 0.6099881529808044, "learning_rate": 0.0002, "epoch": 0.6353568830328995, "step": 3930}, {"loss": 0.809, "grad_norm": 0.43155938386917114, "learning_rate": 0.0002, "epoch": 0.6369735672136448, "step": 3940}, {"loss": 0.8404, "grad_norm": 0.6427084803581238, "learning_rate": 0.0002, "epoch": 0.6385902513943901, "step": 3950}, {"loss": 0.8368, "grad_norm": 0.541220486164093, "learning_rate": 0.0002, "epoch": 0.6402069355751354, "step": 3960}, {"loss": 0.8539, "grad_norm": 0.5414294600486755, "learning_rate": 0.0002, "epoch": 0.6418236197558806, "step": 3970}, {"loss": 0.7996, "grad_norm": 0.46344003081321716, "learning_rate": 0.0002, "epoch": 0.643440303936626, "step": 3980}, {"loss": 0.7474, "grad_norm": 0.45209285616874695, "learning_rate": 0.0002, "epoch": 0.6450569881173712, "step": 3990}, {"loss": 0.8202, "grad_norm": 0.5417284369468689, "learning_rate": 0.0002, "epoch": 0.6466736722981166, "step": 4000}, {"loss": 0.7563, "grad_norm": 0.7995685935020447, "learning_rate": 0.0002, "epoch": 0.6482903564788619, "step": 4010}, {"loss": 0.7812, "grad_norm": 0.6384002566337585, "learning_rate": 0.0002, "epoch": 0.6499070406596071, "step": 4020}, {"loss": 0.732, "grad_norm": 0.4472815692424774, "learning_rate": 0.0002, "epoch": 0.6515237248403525, "step": 4030}, {"loss": 0.8071, "grad_norm": 0.6834294199943542, "learning_rate": 0.0002, "epoch": 0.6531404090210977, "step": 4040}, {"loss": 0.7812, "grad_norm": 0.4612339735031128, "learning_rate": 0.0002, "epoch": 0.654757093201843, "step": 4050}, {"loss": 0.8141, "grad_norm": 0.9266576170921326, "learning_rate": 0.0002, "epoch": 0.6563737773825883, "step": 4060}, {"loss": 0.7991, "grad_norm": 0.4470861852169037, "learning_rate": 0.0002, "epoch": 0.6579904615633336, "step": 4070}, {"loss": 0.8293, "grad_norm": 0.45544925332069397, "learning_rate": 0.0002, "epoch": 0.6596071457440789, "step": 4080}, {"loss": 0.8455, "grad_norm": 0.6144481301307678, "learning_rate": 0.0002, "epoch": 0.6612238299248242, "step": 4090}, {"loss": 0.7877, "grad_norm": 0.5936288237571716, "learning_rate": 0.0002, "epoch": 0.6628405141055694, "step": 4100}, {"loss": 0.7617, "grad_norm": 0.4822963774204254, "learning_rate": 0.0002, "epoch": 0.6644571982863148, "step": 4110}, {"loss": 0.7997, "grad_norm": 0.48432496190071106, "learning_rate": 0.0002, "epoch": 0.66607388246706, "step": 4120}, {"loss": 0.8404, "grad_norm": 0.4901607930660248, "learning_rate": 0.0002, "epoch": 0.6676905666478054, "step": 4130}, {"loss": 0.8085, "grad_norm": 0.5018393397331238, "learning_rate": 0.0002, "epoch": 0.6693072508285507, "step": 4140}, {"loss": 0.8065, "grad_norm": 0.6946378946304321, "learning_rate": 0.0002, "epoch": 0.6709239350092959, "step": 4150}, {"loss": 0.8147, "grad_norm": 0.5997390747070312, "learning_rate": 0.0002, "epoch": 0.6725406191900413, "step": 4160}, {"loss": 0.8268, "grad_norm": 0.6738849878311157, "learning_rate": 0.0002, "epoch": 0.6741573033707865, "step": 4170}, {"loss": 0.7704, "grad_norm": 0.6110581159591675, "learning_rate": 0.0002, "epoch": 0.6757739875515318, "step": 4180}, {"loss": 0.8043, "grad_norm": 0.5703322291374207, "learning_rate": 0.0002, "epoch": 0.6773906717322771, "step": 4190}, {"loss": 0.8099, "grad_norm": 0.4686066210269928, "learning_rate": 0.0002, "epoch": 0.6790073559130224, "step": 4200}, {"loss": 0.8441, "grad_norm": 0.6394643783569336, "learning_rate": 0.0002, "epoch": 0.6806240400937676, "step": 4210}, {"loss": 0.8011, "grad_norm": 0.5454841256141663, "learning_rate": 0.0002, "epoch": 0.682240724274513, "step": 4220}, {"loss": 0.8307, "grad_norm": 0.4859732985496521, "learning_rate": 0.0002, "epoch": 0.6838574084552582, "step": 4230}, {"loss": 0.8161, "grad_norm": 0.5544065833091736, "learning_rate": 0.0002, "epoch": 0.6854740926360036, "step": 4240}, {"loss": 0.7839, "grad_norm": 0.4902505576610565, "learning_rate": 0.0002, "epoch": 0.6870907768167488, "step": 4250}, {"loss": 0.7977, "grad_norm": 0.4768051505088806, "learning_rate": 0.0002, "epoch": 0.6887074609974941, "step": 4260}, {"loss": 0.7539, "grad_norm": 0.49982190132141113, "learning_rate": 0.0002, "epoch": 0.6903241451782395, "step": 4270}, {"loss": 0.7353, "grad_norm": 0.6351838111877441, "learning_rate": 0.0002, "epoch": 0.6919408293589847, "step": 4280}, {"loss": 0.7664, "grad_norm": 0.5647561550140381, "learning_rate": 0.0002, "epoch": 0.69355751353973, "step": 4290}, {"loss": 0.7618, "grad_norm": 0.5340486764907837, "learning_rate": 0.0002, "epoch": 0.6951741977204753, "step": 4300}, {"loss": 0.8526, "grad_norm": 0.5649092793464661, "learning_rate": 0.0002, "epoch": 0.6967908819012206, "step": 4310}, {"loss": 0.8246, "grad_norm": 0.6183916926383972, "learning_rate": 0.0002, "epoch": 0.6984075660819659, "step": 4320}, {"loss": 0.792, "grad_norm": 0.6154509782791138, "learning_rate": 0.0002, "epoch": 0.7000242502627112, "step": 4330}, {"loss": 0.8397, "grad_norm": 0.5156264305114746, "learning_rate": 0.0002, "epoch": 0.7016409344434564, "step": 4340}, {"loss": 0.8512, "grad_norm": 0.562171459197998, "learning_rate": 0.0002, "epoch": 0.7032576186242018, "step": 4350}, {"loss": 0.7882, "grad_norm": 0.4949502646923065, "learning_rate": 0.0002, "epoch": 0.704874302804947, "step": 4360}, {"loss": 0.738, "grad_norm": 0.5171684622764587, "learning_rate": 0.0002, "epoch": 0.7064909869856923, "step": 4370}, {"loss": 0.8001, "grad_norm": 0.6198443174362183, "learning_rate": 0.0002, "epoch": 0.7081076711664377, "step": 4380}, {"loss": 0.7606, "grad_norm": 0.5802276134490967, "learning_rate": 0.0002, "epoch": 0.7097243553471829, "step": 4390}, {"loss": 0.8797, "grad_norm": 0.41096967458724976, "learning_rate": 0.0002, "epoch": 0.7113410395279283, "step": 4400}, {"loss": 0.805, "grad_norm": 0.4397392272949219, "learning_rate": 0.0002, "epoch": 0.7129577237086735, "step": 4410}, {"loss": 0.7651, "grad_norm": 0.45228442549705505, "learning_rate": 0.0002, "epoch": 0.7145744078894188, "step": 4420}, {"loss": 0.7938, "grad_norm": 0.4839673936367035, "learning_rate": 0.0002, "epoch": 0.7161910920701641, "step": 4430}, {"loss": 0.8362, "grad_norm": 0.6140755414962769, "learning_rate": 0.0002, "epoch": 0.7178077762509094, "step": 4440}, {"loss": 0.7722, "grad_norm": 0.6841378808021545, "learning_rate": 0.0002, "epoch": 0.7194244604316546, "step": 4450}, {"loss": 0.8177, "grad_norm": 0.6664239168167114, "learning_rate": 0.0002, "epoch": 0.7210411446124, "step": 4460}, {"loss": 0.7983, "grad_norm": 0.47552719712257385, "learning_rate": 0.0002, "epoch": 0.7226578287931452, "step": 4470}, {"loss": 0.8982, "grad_norm": 0.6649776101112366, "learning_rate": 0.0002, "epoch": 0.7242745129738906, "step": 4480}, {"loss": 0.8074, "grad_norm": 0.5159541964530945, "learning_rate": 0.0002, "epoch": 0.7258911971546358, "step": 4490}, {"loss": 0.7786, "grad_norm": 0.6693112850189209, "learning_rate": 0.0002, "epoch": 0.7275078813353811, "step": 4500}, {"loss": 0.8655, "grad_norm": 0.48870977759361267, "learning_rate": 0.0002, "epoch": 0.7291245655161265, "step": 4510}, {"loss": 0.7337, "grad_norm": 0.4857887923717499, "learning_rate": 0.0002, "epoch": 0.7307412496968717, "step": 4520}, {"loss": 0.8026, "grad_norm": 0.5515662431716919, "learning_rate": 0.0002, "epoch": 0.732357933877617, "step": 4530}, {"loss": 0.8031, "grad_norm": 0.6292222738265991, "learning_rate": 0.0002, "epoch": 0.7339746180583623, "step": 4540}, {"loss": 0.7749, "grad_norm": 0.48265689611434937, "learning_rate": 0.0002, "epoch": 0.7355913022391076, "step": 4550}, {"loss": 0.8499, "grad_norm": 0.8044266104698181, "learning_rate": 0.0002, "epoch": 0.7372079864198529, "step": 4560}, {"loss": 0.8162, "grad_norm": 0.6111769676208496, "learning_rate": 0.0002, "epoch": 0.7388246706005982, "step": 4570}, {"loss": 0.7291, "grad_norm": 0.5229553580284119, "learning_rate": 0.0002, "epoch": 0.7404413547813434, "step": 4580}, {"loss": 0.8038, "grad_norm": 0.6054152250289917, "learning_rate": 0.0002, "epoch": 0.7420580389620888, "step": 4590}, {"loss": 0.8169, "grad_norm": 0.5574966669082642, "learning_rate": 0.0002, "epoch": 0.743674723142834, "step": 4600}, {"loss": 0.8439, "grad_norm": 0.5395817160606384, "learning_rate": 0.0002, "epoch": 0.7452914073235793, "step": 4610}, {"loss": 0.8495, "grad_norm": 0.7116472721099854, "learning_rate": 0.0002, "epoch": 0.7469080915043246, "step": 4620}, {"loss": 0.7743, "grad_norm": 0.5618700981140137, "learning_rate": 0.0002, "epoch": 0.7485247756850699, "step": 4630}, {"loss": 0.7744, "grad_norm": 0.5802770853042603, "learning_rate": 0.0002, "epoch": 0.7501414598658153, "step": 4640}, {"loss": 0.7924, "grad_norm": 0.5690428018569946, "learning_rate": 0.0002, "epoch": 0.7517581440465605, "step": 4650}, {"loss": 0.8017, "grad_norm": 0.4813360273838043, "learning_rate": 0.0002, "epoch": 0.7533748282273058, "step": 4660}, {"loss": 0.8108, "grad_norm": 0.5434042811393738, "learning_rate": 0.0002, "epoch": 0.7549915124080511, "step": 4670}, {"loss": 0.7824, "grad_norm": 0.5502099990844727, "learning_rate": 0.0002, "epoch": 0.7566081965887964, "step": 4680}, {"loss": 0.8598, "grad_norm": 0.6020621061325073, "learning_rate": 0.0002, "epoch": 0.7582248807695416, "step": 4690}, {"loss": 0.7937, "grad_norm": 0.4922301471233368, "learning_rate": 0.0002, "epoch": 0.759841564950287, "step": 4700}, {"loss": 0.788, "grad_norm": 0.6492828726768494, "learning_rate": 0.0002, "epoch": 0.7614582491310322, "step": 4710}, {"loss": 0.8313, "grad_norm": 0.4865580201148987, "learning_rate": 0.0002, "epoch": 0.7630749333117776, "step": 4720}, {"loss": 0.7966, "grad_norm": 0.5971422791481018, "learning_rate": 0.0002, "epoch": 0.7646916174925228, "step": 4730}, {"loss": 0.8298, "grad_norm": 0.6832674145698547, "learning_rate": 0.0002, "epoch": 0.7663083016732681, "step": 4740}, {"loss": 0.8156, "grad_norm": 0.500908613204956, "learning_rate": 0.0002, "epoch": 0.7679249858540134, "step": 4750}, {"loss": 0.8383, "grad_norm": 0.6112465858459473, "learning_rate": 0.0002, "epoch": 0.7695416700347587, "step": 4760}, {"loss": 0.76, "grad_norm": 0.5753506422042847, "learning_rate": 0.0002, "epoch": 0.771158354215504, "step": 4770}, {"loss": 0.8297, "grad_norm": 0.6529405117034912, "learning_rate": 0.0002, "epoch": 0.7727750383962493, "step": 4780}, {"loss": 0.8171, "grad_norm": 0.5916843414306641, "learning_rate": 0.0002, "epoch": 0.7743917225769946, "step": 4790}, {"loss": 0.83, "grad_norm": 0.4821224510669708, "learning_rate": 0.0002, "epoch": 0.7760084067577399, "step": 4800}, {"loss": 0.7703, "grad_norm": 0.5532580018043518, "learning_rate": 0.0002, "epoch": 0.7776250909384852, "step": 4810}, {"loss": 0.7363, "grad_norm": 0.4604877233505249, "learning_rate": 0.0002, "epoch": 0.7792417751192304, "step": 4820}, {"loss": 0.7506, "grad_norm": 0.5009613037109375, "learning_rate": 0.0002, "epoch": 0.7808584592999758, "step": 4830}, {"loss": 0.7863, "grad_norm": 0.6448560357093811, "learning_rate": 0.0002, "epoch": 0.782475143480721, "step": 4840}, {"loss": 0.7957, "grad_norm": 0.44327953457832336, "learning_rate": 0.0002, "epoch": 0.7840918276614663, "step": 4850}, {"loss": 0.7925, "grad_norm": 0.5355411171913147, "learning_rate": 0.0002, "epoch": 0.7857085118422116, "step": 4860}, {"loss": 0.7754, "grad_norm": 0.5635677576065063, "learning_rate": 0.0002, "epoch": 0.7873251960229569, "step": 4870}, {"loss": 0.7931, "grad_norm": 0.5417491793632507, "learning_rate": 0.0002, "epoch": 0.7889418802037023, "step": 4880}, {"loss": 0.7819, "grad_norm": 0.4567430913448334, "learning_rate": 0.0002, "epoch": 0.7905585643844475, "step": 4890}, {"loss": 0.8454, "grad_norm": 0.44651296734809875, "learning_rate": 0.0002, "epoch": 0.7921752485651928, "step": 4900}, {"loss": 0.7959, "grad_norm": 0.5741217136383057, "learning_rate": 0.0002, "epoch": 0.7937919327459381, "step": 4910}, {"loss": 0.8093, "grad_norm": 0.6605045199394226, "learning_rate": 0.0002, "epoch": 0.7954086169266834, "step": 4920}, {"loss": 0.77, "grad_norm": 0.5126531720161438, "learning_rate": 0.0002, "epoch": 0.7970253011074286, "step": 4930}, {"loss": 0.7793, "grad_norm": 0.513648271560669, "learning_rate": 0.0002, "epoch": 0.798641985288174, "step": 4940}, {"loss": 0.8314, "grad_norm": 0.5350404381752014, "learning_rate": 0.0002, "epoch": 0.8002586694689192, "step": 4950}, {"loss": 0.7649, "grad_norm": 0.5731674432754517, "learning_rate": 0.0002, "epoch": 0.8018753536496646, "step": 4960}, {"loss": 0.8572, "grad_norm": 0.5974258184432983, "learning_rate": 0.0002, "epoch": 0.8034920378304098, "step": 4970}, {"loss": 0.7972, "grad_norm": 0.8774799704551697, "learning_rate": 0.0002, "epoch": 0.8051087220111551, "step": 4980}, {"loss": 0.7899, "grad_norm": 0.5994430184364319, "learning_rate": 0.0002, "epoch": 0.8067254061919004, "step": 4990}, {"loss": 0.7736, "grad_norm": 0.4894903004169464, "learning_rate": 0.0002, "epoch": 0.8083420903726457, "step": 5000}, {"loss": 0.78, "grad_norm": 0.5218459367752075, "learning_rate": 0.0002, "epoch": 0.809958774553391, "step": 5010}, {"loss": 0.817, "grad_norm": 0.5232468843460083, "learning_rate": 0.0002, "epoch": 0.8115754587341363, "step": 5020}, {"loss": 0.7704, "grad_norm": 0.44358372688293457, "learning_rate": 0.0002, "epoch": 0.8131921429148816, "step": 5030}, {"loss": 0.785, "grad_norm": 0.6202037334442139, "learning_rate": 0.0002, "epoch": 0.8148088270956269, "step": 5040}, {"loss": 0.7351, "grad_norm": 0.7721474170684814, "learning_rate": 0.0002, "epoch": 0.8164255112763722, "step": 5050}, {"loss": 0.8297, "grad_norm": 0.5568501353263855, "learning_rate": 0.0002, "epoch": 0.8180421954571174, "step": 5060}, {"loss": 0.7733, "grad_norm": 0.49148809909820557, "learning_rate": 0.0002, "epoch": 0.8196588796378628, "step": 5070}, {"loss": 0.8054, "grad_norm": 0.4956012964248657, "learning_rate": 0.0002, "epoch": 0.821275563818608, "step": 5080}, {"loss": 0.8201, "grad_norm": 0.6078833937644958, "learning_rate": 0.0002, "epoch": 0.8228922479993533, "step": 5090}, {"loss": 0.828, "grad_norm": 0.46906954050064087, "learning_rate": 0.0002, "epoch": 0.8245089321800986, "step": 5100}, {"loss": 0.7703, "grad_norm": 0.50812166929245, "learning_rate": 0.0002, "epoch": 0.8261256163608439, "step": 5110}, {"loss": 0.8243, "grad_norm": 0.5319661498069763, "learning_rate": 0.0002, "epoch": 0.8277423005415891, "step": 5120}, {"loss": 0.7798, "grad_norm": 0.4949689209461212, "learning_rate": 0.0002, "epoch": 0.8293589847223345, "step": 5130}, {"loss": 0.7428, "grad_norm": 0.5151591300964355, "learning_rate": 0.0002, "epoch": 0.8309756689030798, "step": 5140}, {"loss": 0.8147, "grad_norm": 0.5530214309692383, "learning_rate": 0.0002, "epoch": 0.8325923530838251, "step": 5150}, {"loss": 0.8251, "grad_norm": 0.6297410130500793, "learning_rate": 0.0002, "epoch": 0.8342090372645704, "step": 5160}, {"loss": 0.8067, "grad_norm": 0.5466840267181396, "learning_rate": 0.0002, "epoch": 0.8358257214453156, "step": 5170}, {"loss": 0.7875, "grad_norm": 0.652913510799408, "learning_rate": 0.0002, "epoch": 0.837442405626061, "step": 5180}, {"loss": 0.8295, "grad_norm": 0.5811293125152588, "learning_rate": 0.0002, "epoch": 0.8390590898068062, "step": 5190}, {"loss": 0.7412, "grad_norm": 0.5109550952911377, "learning_rate": 0.0002, "epoch": 0.8406757739875516, "step": 5200}, {"loss": 0.8077, "grad_norm": 0.4551706612110138, "learning_rate": 0.0002, "epoch": 0.8422924581682968, "step": 5210}, {"loss": 0.7827, "grad_norm": 0.5813754200935364, "learning_rate": 0.0002, "epoch": 0.8439091423490421, "step": 5220}, {"loss": 0.802, "grad_norm": 0.5856947898864746, "learning_rate": 0.0002, "epoch": 0.8455258265297874, "step": 5230}, {"loss": 0.7957, "grad_norm": 0.5482739210128784, "learning_rate": 0.0002, "epoch": 0.8471425107105327, "step": 5240}, {"loss": 0.8295, "grad_norm": 0.49023720622062683, "learning_rate": 0.0002, "epoch": 0.8487591948912779, "step": 5250}, {"loss": 0.8022, "grad_norm": 0.49472475051879883, "learning_rate": 0.0002, "epoch": 0.8503758790720233, "step": 5260}, {"loss": 0.8001, "grad_norm": 0.5490226745605469, "learning_rate": 0.0002, "epoch": 0.8519925632527686, "step": 5270}, {"loss": 0.8333, "grad_norm": 0.5340665578842163, "learning_rate": 0.0002, "epoch": 0.8536092474335139, "step": 5280}, {"loss": 0.8277, "grad_norm": 0.5962483882904053, "learning_rate": 0.0002, "epoch": 0.8552259316142592, "step": 5290}, {"loss": 0.8765, "grad_norm": 0.586358368396759, "learning_rate": 0.0002, "epoch": 0.8568426157950044, "step": 5300}, {"loss": 0.7831, "grad_norm": 0.49120277166366577, "learning_rate": 0.0002, "epoch": 0.8584592999757498, "step": 5310}, {"loss": 0.8162, "grad_norm": 0.5887332558631897, "learning_rate": 0.0002, "epoch": 0.860075984156495, "step": 5320}, {"loss": 0.7464, "grad_norm": 0.42496153712272644, "learning_rate": 0.0002, "epoch": 0.8616926683372403, "step": 5330}, {"loss": 0.7905, "grad_norm": 0.5489874482154846, "learning_rate": 0.0002, "epoch": 0.8633093525179856, "step": 5340}, {"loss": 0.7958, "grad_norm": 0.5850813984870911, "learning_rate": 0.0002, "epoch": 0.8649260366987309, "step": 5350}, {"loss": 0.7642, "grad_norm": 0.517487108707428, "learning_rate": 0.0002, "epoch": 0.8665427208794761, "step": 5360}, {"loss": 0.7801, "grad_norm": 0.5339142680168152, "learning_rate": 0.0002, "epoch": 0.8681594050602215, "step": 5370}, {"loss": 0.818, "grad_norm": 0.6236387491226196, "learning_rate": 0.0002, "epoch": 0.8697760892409668, "step": 5380}, {"loss": 0.7708, "grad_norm": 0.5752192735671997, "learning_rate": 0.0002, "epoch": 0.8713927734217121, "step": 5390}, {"loss": 0.8542, "grad_norm": 0.6724614500999451, "learning_rate": 0.0002, "epoch": 0.8730094576024574, "step": 5400}, {"loss": 0.7581, "grad_norm": 0.5280613303184509, "learning_rate": 0.0002, "epoch": 0.8746261417832026, "step": 5410}, {"loss": 0.8231, "grad_norm": 0.44033288955688477, "learning_rate": 0.0002, "epoch": 0.876242825963948, "step": 5420}, {"loss": 0.8839, "grad_norm": 0.5199708342552185, "learning_rate": 0.0002, "epoch": 0.8778595101446932, "step": 5430}, {"loss": 0.7852, "grad_norm": 0.46778348088264465, "learning_rate": 0.0002, "epoch": 0.8794761943254386, "step": 5440}, {"loss": 0.7834, "grad_norm": 0.4657754898071289, "learning_rate": 0.0002, "epoch": 0.8810928785061838, "step": 5450}, {"loss": 0.7799, "grad_norm": 0.5472902655601501, "learning_rate": 0.0002, "epoch": 0.8827095626869291, "step": 5460}, {"loss": 0.8253, "grad_norm": 0.4876766800880432, "learning_rate": 0.0002, "epoch": 0.8843262468676744, "step": 5470}, {"loss": 0.7906, "grad_norm": 0.5057248473167419, "learning_rate": 0.0002, "epoch": 0.8859429310484197, "step": 5480}, {"loss": 0.8124, "grad_norm": 0.4637320637702942, "learning_rate": 0.0002, "epoch": 0.8875596152291649, "step": 5490}, {"loss": 0.781, "grad_norm": 0.471955806016922, "learning_rate": 0.0002, "epoch": 0.8891762994099103, "step": 5500}, {"loss": 0.8057, "grad_norm": 0.5209813714027405, "learning_rate": 0.0002, "epoch": 0.8907929835906556, "step": 5510}, {"loss": 0.8106, "grad_norm": 0.6213834285736084, "learning_rate": 0.0002, "epoch": 0.8924096677714008, "step": 5520}, {"loss": 0.7787, "grad_norm": 0.5215408205986023, "learning_rate": 0.0002, "epoch": 0.8940263519521462, "step": 5530}, {"loss": 0.8174, "grad_norm": 0.580478310585022, "learning_rate": 0.0002, "epoch": 0.8956430361328914, "step": 5540}, {"loss": 0.8371, "grad_norm": 0.49102169275283813, "learning_rate": 0.0002, "epoch": 0.8972597203136368, "step": 5550}, {"loss": 0.7806, "grad_norm": 0.6043479442596436, "learning_rate": 0.0002, "epoch": 0.898876404494382, "step": 5560}, {"loss": 0.7754, "grad_norm": 0.5636463165283203, "learning_rate": 0.0002, "epoch": 0.9004930886751273, "step": 5570}, {"loss": 0.8145, "grad_norm": 0.5620124340057373, "learning_rate": 0.0002, "epoch": 0.9021097728558726, "step": 5580}, {"loss": 0.8083, "grad_norm": 0.5206354856491089, "learning_rate": 0.0002, "epoch": 0.9037264570366179, "step": 5590}, {"loss": 0.8557, "grad_norm": 0.5798229575157166, "learning_rate": 0.0002, "epoch": 0.9053431412173631, "step": 5600}, {"loss": 0.8097, "grad_norm": 0.6428212523460388, "learning_rate": 0.0002, "epoch": 0.9069598253981085, "step": 5610}, {"loss": 0.7839, "grad_norm": 0.48064687848091125, "learning_rate": 0.0002, "epoch": 0.9085765095788537, "step": 5620}, {"loss": 0.8343, "grad_norm": 0.6347860097885132, "learning_rate": 0.0002, "epoch": 0.9101931937595991, "step": 5630}, {"loss": 0.851, "grad_norm": 0.5353913307189941, "learning_rate": 0.0002, "epoch": 0.9118098779403444, "step": 5640}, {"loss": 0.7736, "grad_norm": 0.5323944091796875, "learning_rate": 0.0002, "epoch": 0.9134265621210896, "step": 5650}, {"loss": 0.8393, "grad_norm": 0.5261843204498291, "learning_rate": 0.0002, "epoch": 0.915043246301835, "step": 5660}, {"loss": 0.7355, "grad_norm": 0.5451326966285706, "learning_rate": 0.0002, "epoch": 0.9166599304825802, "step": 5670}, {"loss": 0.8012, "grad_norm": 0.5183324217796326, "learning_rate": 0.0002, "epoch": 0.9182766146633256, "step": 5680}, {"loss": 0.7659, "grad_norm": 0.47229018807411194, "learning_rate": 0.0002, "epoch": 0.9198932988440708, "step": 5690}, {"loss": 0.7757, "grad_norm": 0.49180513620376587, "learning_rate": 0.0002, "epoch": 0.9215099830248161, "step": 5700}, {"loss": 0.8735, "grad_norm": 0.5419785380363464, "learning_rate": 0.0002, "epoch": 0.9231266672055614, "step": 5710}, {"loss": 0.7378, "grad_norm": 0.5408698916435242, "learning_rate": 0.0002, "epoch": 0.9247433513863067, "step": 5720}, {"loss": 0.7701, "grad_norm": 0.5286232829093933, "learning_rate": 0.0002, "epoch": 0.9263600355670519, "step": 5730}, {"loss": 0.8242, "grad_norm": 0.7539758086204529, "learning_rate": 0.0002, "epoch": 0.9279767197477973, "step": 5740}, {"loss": 0.8118, "grad_norm": 0.5166944861412048, "learning_rate": 0.0002, "epoch": 0.9295934039285425, "step": 5750}, {"loss": 0.783, "grad_norm": 0.6601425409317017, "learning_rate": 0.0002, "epoch": 0.9312100881092878, "step": 5760}, {"loss": 0.7873, "grad_norm": 0.5029960870742798, "learning_rate": 0.0002, "epoch": 0.9328267722900332, "step": 5770}, {"loss": 0.7989, "grad_norm": 0.4926645755767822, "learning_rate": 0.0002, "epoch": 0.9344434564707784, "step": 5780}, {"loss": 0.8174, "grad_norm": 0.5739615559577942, "learning_rate": 0.0002, "epoch": 0.9360601406515238, "step": 5790}, {"loss": 0.8037, "grad_norm": 0.5058279037475586, "learning_rate": 0.0002, "epoch": 0.937676824832269, "step": 5800}, {"loss": 0.8537, "grad_norm": 0.5260962247848511, "learning_rate": 0.0002, "epoch": 0.9392935090130143, "step": 5810}, {"loss": 0.7486, "grad_norm": 0.5768588185310364, "learning_rate": 0.0002, "epoch": 0.9409101931937596, "step": 5820}, {"loss": 0.8215, "grad_norm": 0.5170126557350159, "learning_rate": 0.0002, "epoch": 0.9425268773745049, "step": 5830}, {"loss": 0.7422, "grad_norm": 0.5745864510536194, "learning_rate": 0.0002, "epoch": 0.9441435615552501, "step": 5840}, {"loss": 0.7824, "grad_norm": 0.5551357865333557, "learning_rate": 0.0002, "epoch": 0.9457602457359955, "step": 5850}, {"loss": 0.8529, "grad_norm": 0.5776078701019287, "learning_rate": 0.0002, "epoch": 0.9473769299167407, "step": 5860}, {"loss": 0.8527, "grad_norm": 0.5340062379837036, "learning_rate": 0.0002, "epoch": 0.9489936140974861, "step": 5870}, {"loss": 0.8217, "grad_norm": 0.6447290182113647, "learning_rate": 0.0002, "epoch": 0.9506102982782314, "step": 5880}, {"loss": 0.7945, "grad_norm": 0.5123815536499023, "learning_rate": 0.0002, "epoch": 0.9522269824589766, "step": 5890}, {"loss": 0.8209, "grad_norm": 0.48547613620758057, "learning_rate": 0.0002, "epoch": 0.953843666639722, "step": 5900}, {"loss": 0.7896, "grad_norm": 0.5791414976119995, "learning_rate": 0.0002, "epoch": 0.9554603508204672, "step": 5910}, {"loss": 0.8408, "grad_norm": 0.6195011734962463, "learning_rate": 0.0002, "epoch": 0.9570770350012126, "step": 5920}, {"loss": 0.7805, "grad_norm": 0.6323803067207336, "learning_rate": 0.0002, "epoch": 0.9586937191819578, "step": 5930}, {"loss": 0.8484, "grad_norm": 0.45552879571914673, "learning_rate": 0.0002, "epoch": 0.9603104033627031, "step": 5940}, {"loss": 0.7367, "grad_norm": 0.5796473622322083, "learning_rate": 0.0002, "epoch": 0.9619270875434484, "step": 5950}, {"loss": 0.7672, "grad_norm": 0.647261381149292, "learning_rate": 0.0002, "epoch": 0.9635437717241937, "step": 5960}, {"loss": 0.8086, "grad_norm": 0.5487682819366455, "learning_rate": 0.0002, "epoch": 0.9651604559049389, "step": 5970}, {"loss": 0.7973, "grad_norm": 0.5743663907051086, "learning_rate": 0.0002, "epoch": 0.9667771400856843, "step": 5980}, {"loss": 0.8153, "grad_norm": 0.5470591187477112, "learning_rate": 0.0002, "epoch": 0.9683938242664295, "step": 5990}, {"loss": 0.8119, "grad_norm": 0.5901660323143005, "learning_rate": 0.0002, "epoch": 0.9700105084471748, "step": 6000}, {"loss": 0.8147, "grad_norm": 0.6544759273529053, "learning_rate": 0.0002, "epoch": 0.9716271926279202, "step": 6010}, {"loss": 0.7536, "grad_norm": 0.6288470029830933, "learning_rate": 0.0002, "epoch": 0.9732438768086654, "step": 6020}, {"loss": 0.7989, "grad_norm": 0.673153817653656, "learning_rate": 0.0002, "epoch": 0.9748605609894108, "step": 6030}, {"loss": 0.7556, "grad_norm": 0.42854753136634827, "learning_rate": 0.0002, "epoch": 0.976477245170156, "step": 6040}, {"loss": 0.8006, "grad_norm": 0.5227066278457642, "learning_rate": 0.0002, "epoch": 0.9780939293509013, "step": 6050}, {"loss": 0.795, "grad_norm": 0.5372416973114014, "learning_rate": 0.0002, "epoch": 0.9797106135316466, "step": 6060}, {"loss": 0.7591, "grad_norm": 0.6026402115821838, "learning_rate": 0.0002, "epoch": 0.9813272977123919, "step": 6070}, {"loss": 0.8347, "grad_norm": 0.49547791481018066, "learning_rate": 0.0002, "epoch": 0.9829439818931371, "step": 6080}, {"loss": 0.7722, "grad_norm": 0.4641951322555542, "learning_rate": 0.0002, "epoch": 0.9845606660738825, "step": 6090}, {"loss": 0.8125, "grad_norm": 0.5818535089492798, "learning_rate": 0.0002, "epoch": 0.9861773502546277, "step": 6100}, {"loss": 0.81, "grad_norm": 0.63955157995224, "learning_rate": 0.0002, "epoch": 0.9877940344353731, "step": 6110}, {"loss": 0.7547, "grad_norm": 0.5649438500404358, "learning_rate": 0.0002, "epoch": 0.9894107186161183, "step": 6120}, {"loss": 0.7861, "grad_norm": 0.5290433168411255, "learning_rate": 0.0002, "epoch": 0.9910274027968636, "step": 6130}, {"loss": 0.8109, "grad_norm": 0.6399374008178711, "learning_rate": 0.0002, "epoch": 0.992644086977609, "step": 6140}, {"loss": 0.8373, "grad_norm": 0.6736576557159424, "learning_rate": 0.0002, "epoch": 0.9942607711583542, "step": 6150}, {"loss": 0.7915, "grad_norm": 0.515420138835907, "learning_rate": 0.0002, "epoch": 0.9958774553390995, "step": 6160}, {"loss": 0.8032, "grad_norm": 0.562677800655365, "learning_rate": 0.0002, "epoch": 0.9974941395198448, "step": 6170}, {"loss": 0.8187, "grad_norm": 0.7113858461380005, "learning_rate": 0.0002, "epoch": 0.9991108237005901, "step": 6180}, {"eval_loss": 1.0871200561523438, "eval_runtime": 122.2071, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 0.9999191657909627, "step": 6185}, {"loss": 0.7507, "grad_norm": 0.7111801505088806, "learning_rate": 0.0002, "epoch": 1.0007275078813354, "step": 6190}, {"loss": 0.6865, "grad_norm": 0.5402125716209412, "learning_rate": 0.0002, "epoch": 1.0023441920620806, "step": 6200}, {"loss": 0.7625, "grad_norm": 0.6098830103874207, "learning_rate": 0.0002, "epoch": 1.003960876242826, "step": 6210}, {"loss": 0.7631, "grad_norm": 0.5829983353614807, "learning_rate": 0.0002, "epoch": 1.0055775604235713, "step": 6220}, {"loss": 0.7188, "grad_norm": 0.5614621043205261, "learning_rate": 0.0002, "epoch": 1.0071942446043165, "step": 6230}, {"loss": 0.7505, "grad_norm": 0.5954238772392273, "learning_rate": 0.0002, "epoch": 1.0088109287850617, "step": 6240}, {"loss": 0.7448, "grad_norm": 0.6480574607849121, "learning_rate": 0.0002, "epoch": 1.0104276129658072, "step": 6250}, {"loss": 0.7514, "grad_norm": 0.6051128506660461, "learning_rate": 0.0002, "epoch": 1.0120442971465524, "step": 6260}, {"loss": 0.7237, "grad_norm": 0.6318870782852173, "learning_rate": 0.0002, "epoch": 1.0136609813272976, "step": 6270}, {"loss": 0.7178, "grad_norm": 0.5048980116844177, "learning_rate": 0.0002, "epoch": 1.015277665508043, "step": 6280}, {"loss": 0.7391, "grad_norm": 0.6346936225891113, "learning_rate": 0.0002, "epoch": 1.0168943496887883, "step": 6290}, {"loss": 0.7486, "grad_norm": 0.5711665749549866, "learning_rate": 0.0002, "epoch": 1.0185110338695336, "step": 6300}, {"loss": 0.6808, "grad_norm": 0.5175361037254333, "learning_rate": 0.0002, "epoch": 1.0201277180502788, "step": 6310}, {"loss": 0.7539, "grad_norm": 0.5360831618309021, "learning_rate": 0.0002, "epoch": 1.0217444022310243, "step": 6320}, {"loss": 0.7112, "grad_norm": 0.614675760269165, "learning_rate": 0.0002, "epoch": 1.0233610864117695, "step": 6330}, {"loss": 0.7748, "grad_norm": 0.5626118183135986, "learning_rate": 0.0002, "epoch": 1.0249777705925147, "step": 6340}, {"loss": 0.7375, "grad_norm": 0.574897289276123, "learning_rate": 0.0002, "epoch": 1.02659445477326, "step": 6350}, {"loss": 0.759, "grad_norm": 0.7185447812080383, "learning_rate": 0.0002, "epoch": 1.0282111389540054, "step": 6360}, {"loss": 0.703, "grad_norm": 0.6705799698829651, "learning_rate": 0.0002, "epoch": 1.0298278231347506, "step": 6370}, {"loss": 0.7139, "grad_norm": 0.6740428805351257, "learning_rate": 0.0002, "epoch": 1.0314445073154959, "step": 6380}, {"loss": 0.7252, "grad_norm": 0.663902759552002, "learning_rate": 0.0002, "epoch": 1.0330611914962413, "step": 6390}, {"loss": 0.7065, "grad_norm": 0.5029543042182922, "learning_rate": 0.0002, "epoch": 1.0346778756769865, "step": 6400}, {"loss": 0.711, "grad_norm": 0.7813863158226013, "learning_rate": 0.0002, "epoch": 1.0362945598577318, "step": 6410}, {"loss": 0.7433, "grad_norm": 0.5396282076835632, "learning_rate": 0.0002, "epoch": 1.037911244038477, "step": 6420}, {"loss": 0.7222, "grad_norm": 0.5253293514251709, "learning_rate": 0.0002, "epoch": 1.0395279282192225, "step": 6430}, {"loss": 0.715, "grad_norm": 0.7236770987510681, "learning_rate": 0.0002, "epoch": 1.0411446123999677, "step": 6440}, {"loss": 0.7259, "grad_norm": 0.5670917630195618, "learning_rate": 0.0002, "epoch": 1.042761296580713, "step": 6450}, {"loss": 0.7195, "grad_norm": 0.6031978726387024, "learning_rate": 0.0002, "epoch": 1.0443779807614582, "step": 6460}, {"loss": 0.7648, "grad_norm": 0.5309213399887085, "learning_rate": 0.0002, "epoch": 1.0459946649422036, "step": 6470}, {"loss": 0.7161, "grad_norm": 0.7114651799201965, "learning_rate": 0.0002, "epoch": 1.0476113491229488, "step": 6480}, {"loss": 0.7583, "grad_norm": 0.5591610670089722, "learning_rate": 0.0002, "epoch": 1.049228033303694, "step": 6490}, {"loss": 0.6645, "grad_norm": 0.5185961127281189, "learning_rate": 0.0002, "epoch": 1.0508447174844395, "step": 6500}, {"loss": 0.7654, "grad_norm": 0.6510552167892456, "learning_rate": 0.0002, "epoch": 1.0524614016651848, "step": 6510}, {"loss": 0.7057, "grad_norm": 0.6557928919792175, "learning_rate": 0.0002, "epoch": 1.05407808584593, "step": 6520}, {"loss": 0.8056, "grad_norm": 0.6973192691802979, "learning_rate": 0.0002, "epoch": 1.0556947700266752, "step": 6530}, {"loss": 0.6793, "grad_norm": 0.6226583123207092, "learning_rate": 0.0002, "epoch": 1.0573114542074207, "step": 6540}, {"loss": 0.7151, "grad_norm": 0.5633195638656616, "learning_rate": 0.0002, "epoch": 1.058928138388166, "step": 6550}, {"loss": 0.7082, "grad_norm": 0.7466658353805542, "learning_rate": 0.0002, "epoch": 1.0605448225689111, "step": 6560}, {"loss": 0.7059, "grad_norm": 0.6462772488594055, "learning_rate": 0.0002, "epoch": 1.0621615067496564, "step": 6570}, {"loss": 0.7046, "grad_norm": 0.5266856551170349, "learning_rate": 0.0002, "epoch": 1.0637781909304018, "step": 6580}, {"loss": 0.7157, "grad_norm": 0.534392774105072, "learning_rate": 0.0002, "epoch": 1.065394875111147, "step": 6590}, {"loss": 0.7115, "grad_norm": 0.7514177560806274, "learning_rate": 0.0002, "epoch": 1.0670115592918923, "step": 6600}, {"loss": 0.7545, "grad_norm": 0.7593035697937012, "learning_rate": 0.0002, "epoch": 1.0686282434726375, "step": 6610}, {"loss": 0.6836, "grad_norm": 0.5277858972549438, "learning_rate": 0.0002, "epoch": 1.070244927653383, "step": 6620}, {"loss": 0.7405, "grad_norm": 0.5573670268058777, "learning_rate": 0.0002, "epoch": 1.0718616118341282, "step": 6630}, {"loss": 0.6774, "grad_norm": 0.6802396774291992, "learning_rate": 0.0002, "epoch": 1.0734782960148734, "step": 6640}, {"loss": 0.723, "grad_norm": 0.7367215752601624, "learning_rate": 0.0002, "epoch": 1.0750949801956189, "step": 6650}, {"loss": 0.7429, "grad_norm": 0.5961891412734985, "learning_rate": 0.0002, "epoch": 1.0767116643763641, "step": 6660}, {"loss": 0.6791, "grad_norm": 0.5736313462257385, "learning_rate": 0.0002, "epoch": 1.0783283485571094, "step": 6670}, {"loss": 0.7178, "grad_norm": 0.619219183921814, "learning_rate": 0.0002, "epoch": 1.0799450327378546, "step": 6680}, {"loss": 0.7318, "grad_norm": 0.6214390993118286, "learning_rate": 0.0002, "epoch": 1.0815617169186, "step": 6690}, {"loss": 0.7554, "grad_norm": 0.564536988735199, "learning_rate": 0.0002, "epoch": 1.0831784010993453, "step": 6700}, {"loss": 0.7362, "grad_norm": 0.5838140249252319, "learning_rate": 0.0002, "epoch": 1.0847950852800905, "step": 6710}, {"loss": 0.739, "grad_norm": 0.7000553607940674, "learning_rate": 0.0002, "epoch": 1.0864117694608357, "step": 6720}, {"loss": 0.7369, "grad_norm": 0.7078263759613037, "learning_rate": 0.0002, "epoch": 1.0880284536415812, "step": 6730}, {"loss": 0.7654, "grad_norm": 0.8353848457336426, "learning_rate": 0.0002, "epoch": 1.0896451378223264, "step": 6740}, {"loss": 0.7015, "grad_norm": 0.5615518689155579, "learning_rate": 0.0002, "epoch": 1.0912618220030716, "step": 6750}, {"loss": 0.7396, "grad_norm": 0.5475581288337708, "learning_rate": 0.0002, "epoch": 1.0928785061838169, "step": 6760}, {"loss": 0.7652, "grad_norm": 0.5835978388786316, "learning_rate": 0.0002, "epoch": 1.0944951903645623, "step": 6770}, {"loss": 0.7541, "grad_norm": 0.5516105890274048, "learning_rate": 0.0002, "epoch": 1.0961118745453076, "step": 6780}, {"loss": 0.6842, "grad_norm": 0.5875251889228821, "learning_rate": 0.0002, "epoch": 1.0977285587260528, "step": 6790}, {"loss": 0.6903, "grad_norm": 0.7376947999000549, "learning_rate": 0.0002, "epoch": 1.0993452429067982, "step": 6800}, {"loss": 0.7512, "grad_norm": 0.5656165480613708, "learning_rate": 0.0002, "epoch": 1.1009619270875435, "step": 6810}, {"loss": 0.7409, "grad_norm": 0.6365954279899597, "learning_rate": 0.0002, "epoch": 1.1025786112682887, "step": 6820}, {"loss": 0.7392, "grad_norm": 0.5033080577850342, "learning_rate": 0.0002, "epoch": 1.104195295449034, "step": 6830}, {"loss": 0.6909, "grad_norm": 0.617396891117096, "learning_rate": 0.0002, "epoch": 1.1058119796297794, "step": 6840}, {"loss": 0.7006, "grad_norm": 0.6395374536514282, "learning_rate": 0.0002, "epoch": 1.1074286638105246, "step": 6850}, {"loss": 0.7335, "grad_norm": 0.6775295734405518, "learning_rate": 0.0002, "epoch": 1.1090453479912699, "step": 6860}, {"loss": 0.764, "grad_norm": 0.6655223965644836, "learning_rate": 0.0002, "epoch": 1.1106620321720153, "step": 6870}, {"loss": 0.7553, "grad_norm": 0.676655113697052, "learning_rate": 0.0002, "epoch": 1.1122787163527605, "step": 6880}, {"loss": 0.7342, "grad_norm": 0.6062718629837036, "learning_rate": 0.0002, "epoch": 1.1138954005335058, "step": 6890}, {"loss": 0.7446, "grad_norm": 0.590943455696106, "learning_rate": 0.0002, "epoch": 1.115512084714251, "step": 6900}, {"loss": 0.6705, "grad_norm": 0.6315317153930664, "learning_rate": 0.0002, "epoch": 1.1171287688949965, "step": 6910}, {"loss": 0.6912, "grad_norm": 0.47979024052619934, "learning_rate": 0.0002, "epoch": 1.1187454530757417, "step": 6920}, {"loss": 0.7002, "grad_norm": 0.647298276424408, "learning_rate": 0.0002, "epoch": 1.120362137256487, "step": 6930}, {"loss": 0.7502, "grad_norm": 0.7336484789848328, "learning_rate": 0.0002, "epoch": 1.1219788214372322, "step": 6940}, {"loss": 0.693, "grad_norm": 0.5071424245834351, "learning_rate": 0.0002, "epoch": 1.1235955056179776, "step": 6950}, {"loss": 0.7378, "grad_norm": 0.6527144312858582, "learning_rate": 0.0002, "epoch": 1.1252121897987228, "step": 6960}, {"loss": 0.7228, "grad_norm": 0.6935935020446777, "learning_rate": 0.0002, "epoch": 1.126828873979468, "step": 6970}, {"loss": 0.699, "grad_norm": 0.8026931881904602, "learning_rate": 0.0002, "epoch": 1.1284455581602133, "step": 6980}, {"loss": 0.7361, "grad_norm": 0.5210393667221069, "learning_rate": 0.0002, "epoch": 1.1300622423409588, "step": 6990}, {"loss": 0.7456, "grad_norm": 0.60475093126297, "learning_rate": 0.0002, "epoch": 1.131678926521704, "step": 7000}, {"loss": 0.7495, "grad_norm": 0.6417073607444763, "learning_rate": 0.0002, "epoch": 1.1332956107024492, "step": 7010}, {"loss": 0.7459, "grad_norm": 0.6732175946235657, "learning_rate": 0.0002, "epoch": 1.1349122948831947, "step": 7020}, {"loss": 0.7278, "grad_norm": 0.6719491481781006, "learning_rate": 0.0002, "epoch": 1.13652897906394, "step": 7030}, {"loss": 0.7694, "grad_norm": 0.5708295106887817, "learning_rate": 0.0002, "epoch": 1.1381456632446851, "step": 7040}, {"loss": 0.7823, "grad_norm": 0.7141719460487366, "learning_rate": 0.0002, "epoch": 1.1397623474254304, "step": 7050}, {"loss": 0.764, "grad_norm": 0.6187017560005188, "learning_rate": 0.0002, "epoch": 1.1413790316061758, "step": 7060}, {"loss": 0.7657, "grad_norm": 0.50581294298172, "learning_rate": 0.0002, "epoch": 1.142995715786921, "step": 7070}, {"loss": 0.7357, "grad_norm": 0.5620143413543701, "learning_rate": 0.0002, "epoch": 1.1446123999676663, "step": 7080}, {"loss": 0.7287, "grad_norm": 0.6231929659843445, "learning_rate": 0.0002, "epoch": 1.1462290841484115, "step": 7090}, {"loss": 0.7328, "grad_norm": 0.5775774121284485, "learning_rate": 0.0002, "epoch": 1.147845768329157, "step": 7100}, {"loss": 0.7728, "grad_norm": 0.6492809653282166, "learning_rate": 0.0002, "epoch": 1.1494624525099022, "step": 7110}, {"loss": 0.7545, "grad_norm": 0.6434972286224365, "learning_rate": 0.0002, "epoch": 1.1510791366906474, "step": 7120}, {"loss": 0.7374, "grad_norm": 0.6191812753677368, "learning_rate": 0.0002, "epoch": 1.1526958208713927, "step": 7130}, {"loss": 0.7276, "grad_norm": 0.6690331697463989, "learning_rate": 0.0002, "epoch": 1.1543125050521381, "step": 7140}, {"loss": 0.7704, "grad_norm": 0.5977938175201416, "learning_rate": 0.0002, "epoch": 1.1559291892328833, "step": 7150}, {"loss": 0.7251, "grad_norm": 0.6195854544639587, "learning_rate": 0.0002, "epoch": 1.1575458734136286, "step": 7160}, {"loss": 0.7249, "grad_norm": 0.5752048492431641, "learning_rate": 0.0002, "epoch": 1.159162557594374, "step": 7170}, {"loss": 0.7593, "grad_norm": 0.589081883430481, "learning_rate": 0.0002, "epoch": 1.1607792417751193, "step": 7180}, {"loss": 0.704, "grad_norm": 0.756996750831604, "learning_rate": 0.0002, "epoch": 1.1623959259558645, "step": 7190}, {"loss": 0.7404, "grad_norm": 0.7614967226982117, "learning_rate": 0.0002, "epoch": 1.1640126101366097, "step": 7200}, {"loss": 0.7867, "grad_norm": 0.6120437979698181, "learning_rate": 0.0002, "epoch": 1.1656292943173552, "step": 7210}, {"loss": 0.7384, "grad_norm": 0.6210004687309265, "learning_rate": 0.0002, "epoch": 1.1672459784981004, "step": 7220}, {"loss": 0.7251, "grad_norm": 0.6044116020202637, "learning_rate": 0.0002, "epoch": 1.1688626626788456, "step": 7230}, {"loss": 0.7361, "grad_norm": 0.5418457388877869, "learning_rate": 0.0002, "epoch": 1.170479346859591, "step": 7240}, {"loss": 0.6938, "grad_norm": 0.6413537263870239, "learning_rate": 0.0002, "epoch": 1.1720960310403363, "step": 7250}, {"loss": 0.6978, "grad_norm": 0.5777867436408997, "learning_rate": 0.0002, "epoch": 1.1737127152210816, "step": 7260}, {"loss": 0.7503, "grad_norm": 0.7092402577400208, "learning_rate": 0.0002, "epoch": 1.1753293994018268, "step": 7270}, {"loss": 0.7487, "grad_norm": 0.6351709365844727, "learning_rate": 0.0002, "epoch": 1.176946083582572, "step": 7280}, {"loss": 0.7527, "grad_norm": 0.6172189712524414, "learning_rate": 0.0002, "epoch": 1.1785627677633175, "step": 7290}, {"loss": 0.7319, "grad_norm": 0.6801714897155762, "learning_rate": 0.0002, "epoch": 1.1801794519440627, "step": 7300}, {"loss": 0.6941, "grad_norm": 0.6044712066650391, "learning_rate": 0.0002, "epoch": 1.181796136124808, "step": 7310}, {"loss": 0.6951, "grad_norm": 0.7413212060928345, "learning_rate": 0.0002, "epoch": 1.1834128203055534, "step": 7320}, {"loss": 0.7396, "grad_norm": 0.5303856134414673, "learning_rate": 0.0002, "epoch": 1.1850295044862986, "step": 7330}, {"loss": 0.6915, "grad_norm": 0.5647098422050476, "learning_rate": 0.0002, "epoch": 1.1866461886670439, "step": 7340}, {"loss": 0.7506, "grad_norm": 0.7374135255813599, "learning_rate": 0.0002, "epoch": 1.188262872847789, "step": 7350}, {"loss": 0.7041, "grad_norm": 0.5710089206695557, "learning_rate": 0.0002, "epoch": 1.1898795570285345, "step": 7360}, {"loss": 0.8289, "grad_norm": 0.6073619723320007, "learning_rate": 0.0002, "epoch": 1.1914962412092798, "step": 7370}, {"loss": 0.7722, "grad_norm": 0.5899916887283325, "learning_rate": 0.0002, "epoch": 1.193112925390025, "step": 7380}, {"loss": 0.756, "grad_norm": 0.7762434482574463, "learning_rate": 0.0002, "epoch": 1.1947296095707705, "step": 7390}, {"loss": 0.7319, "grad_norm": 0.679949939250946, "learning_rate": 0.0002, "epoch": 1.1963462937515157, "step": 7400}, {"loss": 0.7599, "grad_norm": 0.6106849312782288, "learning_rate": 0.0002, "epoch": 1.197962977932261, "step": 7410}, {"loss": 0.7648, "grad_norm": 0.682461678981781, "learning_rate": 0.0002, "epoch": 1.1995796621130062, "step": 7420}, {"loss": 0.7741, "grad_norm": 0.6087017059326172, "learning_rate": 0.0002, "epoch": 1.2011963462937516, "step": 7430}, {"loss": 0.7642, "grad_norm": 0.63739013671875, "learning_rate": 0.0002, "epoch": 1.2028130304744968, "step": 7440}, {"loss": 0.7611, "grad_norm": 0.6154777407646179, "learning_rate": 0.0002, "epoch": 1.204429714655242, "step": 7450}, {"loss": 0.7565, "grad_norm": 0.7491534948348999, "learning_rate": 0.0002, "epoch": 1.2060463988359873, "step": 7460}, {"loss": 0.698, "grad_norm": 0.6664797067642212, "learning_rate": 0.0002, "epoch": 1.2076630830167328, "step": 7470}, {"loss": 0.7456, "grad_norm": 0.6660266518592834, "learning_rate": 0.0002, "epoch": 1.209279767197478, "step": 7480}, {"loss": 0.714, "grad_norm": 0.6972551345825195, "learning_rate": 0.0002, "epoch": 1.2108964513782232, "step": 7490}, {"loss": 0.7023, "grad_norm": 0.6157945990562439, "learning_rate": 0.0002, "epoch": 1.2125131355589684, "step": 7500}, {"loss": 0.7326, "grad_norm": 0.5199310183525085, "learning_rate": 0.0002, "epoch": 1.214129819739714, "step": 7510}, {"loss": 0.7586, "grad_norm": 0.577610433101654, "learning_rate": 0.0002, "epoch": 1.2157465039204591, "step": 7520}, {"loss": 0.7179, "grad_norm": 0.53652423620224, "learning_rate": 0.0002, "epoch": 1.2173631881012044, "step": 7530}, {"loss": 0.7393, "grad_norm": 0.6479050517082214, "learning_rate": 0.0002, "epoch": 1.2189798722819498, "step": 7540}, {"loss": 0.7534, "grad_norm": 0.618748128414154, "learning_rate": 0.0002, "epoch": 1.220596556462695, "step": 7550}, {"loss": 0.6886, "grad_norm": 0.6311424374580383, "learning_rate": 0.0002, "epoch": 1.2222132406434403, "step": 7560}, {"loss": 0.7272, "grad_norm": 0.6595825552940369, "learning_rate": 0.0002, "epoch": 1.2238299248241855, "step": 7570}, {"loss": 0.7353, "grad_norm": 0.5198960900306702, "learning_rate": 0.0002, "epoch": 1.225446609004931, "step": 7580}, {"loss": 0.674, "grad_norm": 0.578650712966919, "learning_rate": 0.0002, "epoch": 1.2270632931856762, "step": 7590}, {"loss": 0.7507, "grad_norm": 0.6080220937728882, "learning_rate": 0.0002, "epoch": 1.2286799773664214, "step": 7600}, {"loss": 0.7733, "grad_norm": 0.7050248384475708, "learning_rate": 0.0002, "epoch": 1.2302966615471669, "step": 7610}, {"loss": 0.7032, "grad_norm": 0.6652196049690247, "learning_rate": 0.0002, "epoch": 1.2319133457279121, "step": 7620}, {"loss": 0.7085, "grad_norm": 0.7322776317596436, "learning_rate": 0.0002, "epoch": 1.2335300299086573, "step": 7630}, {"loss": 0.7402, "grad_norm": 0.4998728036880493, "learning_rate": 0.0002, "epoch": 1.2351467140894026, "step": 7640}, {"loss": 0.7214, "grad_norm": 0.6428788900375366, "learning_rate": 0.0002, "epoch": 1.2367633982701478, "step": 7650}, {"loss": 0.7699, "grad_norm": 0.585242509841919, "learning_rate": 0.0002, "epoch": 1.2383800824508933, "step": 7660}, {"loss": 0.7621, "grad_norm": 0.5211917757987976, "learning_rate": 0.0002, "epoch": 1.2399967666316385, "step": 7670}, {"loss": 0.746, "grad_norm": 0.6490384340286255, "learning_rate": 0.0002, "epoch": 1.2416134508123837, "step": 7680}, {"loss": 0.7186, "grad_norm": 0.6249763369560242, "learning_rate": 0.0002, "epoch": 1.2432301349931292, "step": 7690}, {"loss": 0.7761, "grad_norm": 0.71870356798172, "learning_rate": 0.0002, "epoch": 1.2448468191738744, "step": 7700}, {"loss": 0.7525, "grad_norm": 0.6761967539787292, "learning_rate": 0.0002, "epoch": 1.2464635033546196, "step": 7710}, {"loss": 0.7501, "grad_norm": 0.6500617265701294, "learning_rate": 0.0002, "epoch": 1.2480801875353649, "step": 7720}, {"loss": 0.7903, "grad_norm": 0.8069869875907898, "learning_rate": 0.0002, "epoch": 1.2496968717161103, "step": 7730}, {"loss": 0.6747, "grad_norm": 0.6044608950614929, "learning_rate": 0.0002, "epoch": 1.2513135558968556, "step": 7740}, {"loss": 0.6825, "grad_norm": 0.6573283076286316, "learning_rate": 0.0002, "epoch": 1.2529302400776008, "step": 7750}, {"loss": 0.7617, "grad_norm": 0.625430166721344, "learning_rate": 0.0002, "epoch": 1.2545469242583462, "step": 7760}, {"loss": 0.7041, "grad_norm": 0.5442022681236267, "learning_rate": 0.0002, "epoch": 1.2561636084390915, "step": 7770}, {"loss": 0.7172, "grad_norm": 0.6818386912345886, "learning_rate": 0.0002, "epoch": 1.2577802926198367, "step": 7780}, {"loss": 0.696, "grad_norm": 0.6381874084472656, "learning_rate": 0.0002, "epoch": 1.259396976800582, "step": 7790}, {"loss": 0.6834, "grad_norm": 0.6269212961196899, "learning_rate": 0.0002, "epoch": 1.2610136609813272, "step": 7800}, {"loss": 0.7821, "grad_norm": 0.600121259689331, "learning_rate": 0.0002, "epoch": 1.2626303451620726, "step": 7810}, {"loss": 0.7761, "grad_norm": 0.6337703466415405, "learning_rate": 0.0002, "epoch": 1.2642470293428179, "step": 7820}, {"loss": 0.732, "grad_norm": 0.7234963774681091, "learning_rate": 0.0002, "epoch": 1.2658637135235633, "step": 7830}, {"loss": 0.785, "grad_norm": 0.800184965133667, "learning_rate": 0.0002, "epoch": 1.2674803977043085, "step": 7840}, {"loss": 0.7426, "grad_norm": 0.7539464831352234, "learning_rate": 0.0002, "epoch": 1.2690970818850538, "step": 7850}, {"loss": 0.7496, "grad_norm": 0.5493760704994202, "learning_rate": 0.0002, "epoch": 1.270713766065799, "step": 7860}, {"loss": 0.7537, "grad_norm": 0.7477145791053772, "learning_rate": 0.0002, "epoch": 1.2723304502465442, "step": 7870}, {"loss": 0.7573, "grad_norm": 0.6366362571716309, "learning_rate": 0.0002, "epoch": 1.2739471344272897, "step": 7880}, {"loss": 0.7608, "grad_norm": 0.7419533729553223, "learning_rate": 0.0002, "epoch": 1.275563818608035, "step": 7890}, {"loss": 0.7873, "grad_norm": 0.6141223311424255, "learning_rate": 0.0002, "epoch": 1.2771805027887801, "step": 7900}, {"loss": 0.6916, "grad_norm": 0.7522598505020142, "learning_rate": 0.0002, "epoch": 1.2787971869695256, "step": 7910}, {"loss": 0.7097, "grad_norm": 0.6935804486274719, "learning_rate": 0.0002, "epoch": 1.2804138711502708, "step": 7920}, {"loss": 0.7185, "grad_norm": 0.7239290475845337, "learning_rate": 0.0002, "epoch": 1.282030555331016, "step": 7930}, {"loss": 0.7145, "grad_norm": 0.8800187110900879, "learning_rate": 0.0002, "epoch": 1.2836472395117613, "step": 7940}, {"loss": 0.6991, "grad_norm": 0.540458083152771, "learning_rate": 0.0002, "epoch": 1.2852639236925067, "step": 7950}, {"loss": 0.7139, "grad_norm": 0.6492934226989746, "learning_rate": 0.0002, "epoch": 1.286880607873252, "step": 7960}, {"loss": 0.7742, "grad_norm": 0.6543959379196167, "learning_rate": 0.0002, "epoch": 1.2884972920539972, "step": 7970}, {"loss": 0.7316, "grad_norm": 0.5804705619812012, "learning_rate": 0.0002, "epoch": 1.2901139762347427, "step": 7980}, {"loss": 0.796, "grad_norm": 0.7074727416038513, "learning_rate": 0.0002, "epoch": 1.291730660415488, "step": 7990}, {"loss": 0.7034, "grad_norm": 0.5347974300384521, "learning_rate": 0.0002, "epoch": 1.2933473445962331, "step": 8000}, {"loss": 0.738, "grad_norm": 0.6457298398017883, "learning_rate": 0.0002, "epoch": 1.2949640287769784, "step": 8010}, {"loss": 0.7634, "grad_norm": 0.6407219171524048, "learning_rate": 0.0002, "epoch": 1.2965807129577236, "step": 8020}, {"loss": 0.7506, "grad_norm": 0.828439474105835, "learning_rate": 0.0002, "epoch": 1.298197397138469, "step": 8030}, {"loss": 0.735, "grad_norm": 0.4840380549430847, "learning_rate": 0.0002, "epoch": 1.2998140813192143, "step": 8040}, {"loss": 0.7283, "grad_norm": 0.5921024680137634, "learning_rate": 0.0002, "epoch": 1.3014307654999595, "step": 8050}, {"loss": 0.7477, "grad_norm": 0.6170315146446228, "learning_rate": 0.0002, "epoch": 1.303047449680705, "step": 8060}, {"loss": 0.7534, "grad_norm": 0.5374847054481506, "learning_rate": 0.0002, "epoch": 1.3046641338614502, "step": 8070}, {"loss": 0.7593, "grad_norm": 0.545758068561554, "learning_rate": 0.0002, "epoch": 1.3062808180421954, "step": 8080}, {"loss": 0.7463, "grad_norm": 0.55641770362854, "learning_rate": 0.0002, "epoch": 1.3078975022229407, "step": 8090}, {"loss": 0.7594, "grad_norm": 0.6724897027015686, "learning_rate": 0.0002, "epoch": 1.309514186403686, "step": 8100}, {"loss": 0.7105, "grad_norm": 0.6923972368240356, "learning_rate": 0.0002, "epoch": 1.3111308705844313, "step": 8110}, {"loss": 0.7149, "grad_norm": 0.5136841535568237, "learning_rate": 0.0002, "epoch": 1.3127475547651766, "step": 8120}, {"loss": 0.7504, "grad_norm": 0.6766283512115479, "learning_rate": 0.0002, "epoch": 1.314364238945922, "step": 8130}, {"loss": 0.7489, "grad_norm": 0.6283926367759705, "learning_rate": 0.0002, "epoch": 1.3159809231266673, "step": 8140}, {"loss": 0.7459, "grad_norm": 0.644216001033783, "learning_rate": 0.0002, "epoch": 1.3175976073074125, "step": 8150}, {"loss": 0.7125, "grad_norm": 0.7827503085136414, "learning_rate": 0.0002, "epoch": 1.3192142914881577, "step": 8160}, {"loss": 0.7271, "grad_norm": 0.6651390790939331, "learning_rate": 0.0002, "epoch": 1.320830975668903, "step": 8170}, {"loss": 0.7778, "grad_norm": 0.5547412633895874, "learning_rate": 0.0002, "epoch": 1.3224476598496484, "step": 8180}, {"loss": 0.7402, "grad_norm": 0.6765179634094238, "learning_rate": 0.0002, "epoch": 1.3240643440303936, "step": 8190}, {"loss": 0.7106, "grad_norm": 0.6822077035903931, "learning_rate": 0.0002, "epoch": 1.325681028211139, "step": 8200}, {"loss": 0.7288, "grad_norm": 0.5941002368927002, "learning_rate": 0.0002, "epoch": 1.3272977123918843, "step": 8210}, {"loss": 0.7494, "grad_norm": 0.4850037097930908, "learning_rate": 0.0002, "epoch": 1.3289143965726296, "step": 8220}, {"loss": 0.7474, "grad_norm": 0.6162990927696228, "learning_rate": 0.0002, "epoch": 1.3305310807533748, "step": 8230}, {"loss": 0.7751, "grad_norm": 0.6665613651275635, "learning_rate": 0.0002, "epoch": 1.33214776493412, "step": 8240}, {"loss": 0.759, "grad_norm": 0.618192732334137, "learning_rate": 0.0002, "epoch": 1.3337644491148655, "step": 8250}, {"loss": 0.7532, "grad_norm": 0.710418701171875, "learning_rate": 0.0002, "epoch": 1.3353811332956107, "step": 8260}, {"loss": 0.7306, "grad_norm": 0.5109876990318298, "learning_rate": 0.0002, "epoch": 1.336997817476356, "step": 8270}, {"loss": 0.7303, "grad_norm": 0.6791711449623108, "learning_rate": 0.0002, "epoch": 1.3386145016571014, "step": 8280}, {"loss": 0.7594, "grad_norm": 0.6836432814598083, "learning_rate": 0.0002, "epoch": 1.3402311858378466, "step": 8290}, {"loss": 0.7594, "grad_norm": 0.5579386353492737, "learning_rate": 0.0002, "epoch": 1.3418478700185918, "step": 8300}, {"loss": 0.7377, "grad_norm": 0.6713546514511108, "learning_rate": 0.0002, "epoch": 1.343464554199337, "step": 8310}, {"loss": 0.7756, "grad_norm": 0.5353720188140869, "learning_rate": 0.0002, "epoch": 1.3450812383800825, "step": 8320}, {"loss": 0.718, "grad_norm": 0.5813682675361633, "learning_rate": 0.0002, "epoch": 1.3466979225608278, "step": 8330}, {"loss": 0.7294, "grad_norm": 0.8158791661262512, "learning_rate": 0.0002, "epoch": 1.348314606741573, "step": 8340}, {"loss": 0.6992, "grad_norm": 0.6193785071372986, "learning_rate": 0.0002, "epoch": 1.3499312909223184, "step": 8350}, {"loss": 0.7654, "grad_norm": 0.6353939771652222, "learning_rate": 0.0002, "epoch": 1.3515479751030637, "step": 8360}, {"loss": 0.7519, "grad_norm": 0.6925048232078552, "learning_rate": 0.0002, "epoch": 1.353164659283809, "step": 8370}, {"loss": 0.736, "grad_norm": 0.988264799118042, "learning_rate": 0.0002, "epoch": 1.3547813434645541, "step": 8380}, {"loss": 0.7744, "grad_norm": 0.6476002931594849, "learning_rate": 0.0002, "epoch": 1.3563980276452994, "step": 8390}, {"loss": 0.776, "grad_norm": 0.7120398879051208, "learning_rate": 0.0002, "epoch": 1.3580147118260448, "step": 8400}, {"loss": 0.7368, "grad_norm": 0.9048416614532471, "learning_rate": 0.0002, "epoch": 1.35963139600679, "step": 8410}, {"loss": 0.7544, "grad_norm": 0.7000672817230225, "learning_rate": 0.0002, "epoch": 1.3612480801875353, "step": 8420}, {"loss": 0.7358, "grad_norm": 0.6015632152557373, "learning_rate": 0.0002, "epoch": 1.3628647643682807, "step": 8430}, {"loss": 0.7298, "grad_norm": 0.612516462802887, "learning_rate": 0.0002, "epoch": 1.364481448549026, "step": 8440}, {"loss": 0.7055, "grad_norm": 0.5969301462173462, "learning_rate": 0.0002, "epoch": 1.3660981327297712, "step": 8450}, {"loss": 0.7754, "grad_norm": 0.6730654239654541, "learning_rate": 0.0002, "epoch": 1.3677148169105164, "step": 8460}, {"loss": 0.7465, "grad_norm": 0.6386392116546631, "learning_rate": 0.0002, "epoch": 1.369331501091262, "step": 8470}, {"loss": 0.7433, "grad_norm": 0.739544153213501, "learning_rate": 0.0002, "epoch": 1.3709481852720071, "step": 8480}, {"loss": 0.7892, "grad_norm": 0.6462782621383667, "learning_rate": 0.0002, "epoch": 1.3725648694527524, "step": 8490}, {"loss": 0.7302, "grad_norm": 0.7346843481063843, "learning_rate": 0.0002, "epoch": 1.3741815536334978, "step": 8500}, {"loss": 0.7634, "grad_norm": 0.6884821057319641, "learning_rate": 0.0002, "epoch": 1.375798237814243, "step": 8510}, {"loss": 0.7614, "grad_norm": 0.6999333500862122, "learning_rate": 0.0002, "epoch": 1.3774149219949883, "step": 8520}, {"loss": 0.729, "grad_norm": 0.5378713011741638, "learning_rate": 0.0002, "epoch": 1.3790316061757335, "step": 8530}, {"loss": 0.6797, "grad_norm": 0.5417906641960144, "learning_rate": 0.0002, "epoch": 1.3806482903564787, "step": 8540}, {"loss": 0.7499, "grad_norm": 0.6602526307106018, "learning_rate": 0.0002, "epoch": 1.3822649745372242, "step": 8550}, {"loss": 0.7356, "grad_norm": 0.7073674201965332, "learning_rate": 0.0002, "epoch": 1.3838816587179694, "step": 8560}, {"loss": 0.75, "grad_norm": 0.5841707587242126, "learning_rate": 0.0002, "epoch": 1.3854983428987149, "step": 8570}, {"loss": 0.732, "grad_norm": 0.7031095027923584, "learning_rate": 0.0002, "epoch": 1.38711502707946, "step": 8580}, {"loss": 0.7464, "grad_norm": 0.5198570489883423, "learning_rate": 0.0002, "epoch": 1.3887317112602053, "step": 8590}, {"loss": 0.7354, "grad_norm": 0.7261320352554321, "learning_rate": 0.0002, "epoch": 1.3903483954409506, "step": 8600}, {"loss": 0.7339, "grad_norm": 0.5616350173950195, "learning_rate": 0.0002, "epoch": 1.3919650796216958, "step": 8610}, {"loss": 0.7382, "grad_norm": 0.5185914635658264, "learning_rate": 0.0002, "epoch": 1.3935817638024413, "step": 8620}, {"loss": 0.7456, "grad_norm": 0.5814694762229919, "learning_rate": 0.0002, "epoch": 1.3951984479831865, "step": 8630}, {"loss": 0.7413, "grad_norm": 0.6977371573448181, "learning_rate": 0.0002, "epoch": 1.3968151321639317, "step": 8640}, {"loss": 0.7574, "grad_norm": 0.6855689883232117, "learning_rate": 0.0002, "epoch": 1.3984318163446772, "step": 8650}, {"loss": 0.7802, "grad_norm": 0.5414357781410217, "learning_rate": 0.0002, "epoch": 1.4000485005254224, "step": 8660}, {"loss": 0.7487, "grad_norm": 0.6970012784004211, "learning_rate": 0.0002, "epoch": 1.4016651847061676, "step": 8670}, {"loss": 0.7421, "grad_norm": 0.526079535484314, "learning_rate": 0.0002, "epoch": 1.4032818688869129, "step": 8680}, {"loss": 0.737, "grad_norm": 0.758712887763977, "learning_rate": 0.0002, "epoch": 1.404898553067658, "step": 8690}, {"loss": 0.7612, "grad_norm": 0.7118762731552124, "learning_rate": 0.0002, "epoch": 1.4065152372484035, "step": 8700}, {"loss": 0.7628, "grad_norm": 0.5696909427642822, "learning_rate": 0.0002, "epoch": 1.4081319214291488, "step": 8710}, {"loss": 0.7156, "grad_norm": 0.7995436787605286, "learning_rate": 0.0002, "epoch": 1.4097486056098942, "step": 8720}, {"loss": 0.7521, "grad_norm": 0.7237521409988403, "learning_rate": 0.0002, "epoch": 1.4113652897906395, "step": 8730}, {"loss": 0.7661, "grad_norm": 0.744628369808197, "learning_rate": 0.0002, "epoch": 1.4129819739713847, "step": 8740}, {"loss": 0.7073, "grad_norm": 0.6082926988601685, "learning_rate": 0.0002, "epoch": 1.41459865815213, "step": 8750}, {"loss": 0.7282, "grad_norm": 0.5185243487358093, "learning_rate": 0.0002, "epoch": 1.4162153423328752, "step": 8760}, {"loss": 0.7592, "grad_norm": 0.5183082222938538, "learning_rate": 0.0002, "epoch": 1.4178320265136206, "step": 8770}, {"loss": 0.7509, "grad_norm": 0.7326041460037231, "learning_rate": 0.0002, "epoch": 1.4194487106943658, "step": 8780}, {"loss": 0.7398, "grad_norm": 0.7174660563468933, "learning_rate": 0.0002, "epoch": 1.421065394875111, "step": 8790}, {"loss": 0.7507, "grad_norm": 0.8080165982246399, "learning_rate": 0.0002, "epoch": 1.4226820790558565, "step": 8800}, {"loss": 0.72, "grad_norm": 0.5061507821083069, "learning_rate": 0.0002, "epoch": 1.4242987632366018, "step": 8810}, {"loss": 0.7563, "grad_norm": 0.801602840423584, "learning_rate": 0.0002, "epoch": 1.425915447417347, "step": 8820}, {"loss": 0.7287, "grad_norm": 0.6150273084640503, "learning_rate": 0.0002, "epoch": 1.4275321315980922, "step": 8830}, {"loss": 0.7452, "grad_norm": 0.8786525726318359, "learning_rate": 0.0002, "epoch": 1.4291488157788377, "step": 8840}, {"loss": 0.7257, "grad_norm": 0.6371538639068604, "learning_rate": 0.0002, "epoch": 1.430765499959583, "step": 8850}, {"loss": 0.711, "grad_norm": 0.6409295797348022, "learning_rate": 0.0002, "epoch": 1.4323821841403281, "step": 8860}, {"loss": 0.7891, "grad_norm": 0.6452359557151794, "learning_rate": 0.0002, "epoch": 1.4339988683210736, "step": 8870}, {"loss": 0.7588, "grad_norm": 0.5842334628105164, "learning_rate": 0.0002, "epoch": 1.4356155525018188, "step": 8880}, {"loss": 0.7446, "grad_norm": 0.696761965751648, "learning_rate": 0.0002, "epoch": 1.437232236682564, "step": 8890}, {"loss": 0.7541, "grad_norm": 0.6384600400924683, "learning_rate": 0.0002, "epoch": 1.4388489208633093, "step": 8900}, {"loss": 0.7049, "grad_norm": 0.5981136560440063, "learning_rate": 0.0002, "epoch": 1.4404656050440545, "step": 8910}, {"loss": 0.795, "grad_norm": 0.6355637907981873, "learning_rate": 0.0002, "epoch": 1.4420822892248, "step": 8920}, {"loss": 0.7653, "grad_norm": 0.6374830603599548, "learning_rate": 0.0002, "epoch": 1.4436989734055452, "step": 8930}, {"loss": 0.8108, "grad_norm": 0.559013307094574, "learning_rate": 0.0002, "epoch": 1.4453156575862904, "step": 8940}, {"loss": 0.7045, "grad_norm": 0.7289170026779175, "learning_rate": 0.0002, "epoch": 1.446932341767036, "step": 8950}, {"loss": 0.7484, "grad_norm": 0.8649206757545471, "learning_rate": 0.0002, "epoch": 1.4485490259477811, "step": 8960}, {"loss": 0.7745, "grad_norm": 0.7664689421653748, "learning_rate": 0.0002, "epoch": 1.4501657101285264, "step": 8970}, {"loss": 0.7431, "grad_norm": 0.7109952569007874, "learning_rate": 0.0002, "epoch": 1.4517823943092716, "step": 8980}, {"loss": 0.7997, "grad_norm": 0.6312844753265381, "learning_rate": 0.0002, "epoch": 1.453399078490017, "step": 8990}, {"loss": 0.7467, "grad_norm": 0.6616617441177368, "learning_rate": 0.0002, "epoch": 1.4550157626707623, "step": 9000}, {"loss": 0.7518, "grad_norm": 0.7384068965911865, "learning_rate": 0.0002, "epoch": 1.4566324468515075, "step": 9010}, {"loss": 0.7483, "grad_norm": 0.6549670100212097, "learning_rate": 0.0002, "epoch": 1.458249131032253, "step": 9020}, {"loss": 0.7423, "grad_norm": 0.6254119277000427, "learning_rate": 0.0002, "epoch": 1.4598658152129982, "step": 9030}, {"loss": 0.7645, "grad_norm": 0.6806328892707825, "learning_rate": 0.0002, "epoch": 1.4614824993937434, "step": 9040}, {"loss": 0.7221, "grad_norm": 0.6803115010261536, "learning_rate": 0.0002, "epoch": 1.4630991835744886, "step": 9050}, {"loss": 0.7264, "grad_norm": 0.48529282212257385, "learning_rate": 0.0002, "epoch": 1.4647158677552339, "step": 9060}, {"loss": 0.7542, "grad_norm": 0.5995030999183655, "learning_rate": 0.0002, "epoch": 1.4663325519359793, "step": 9070}, {"loss": 0.7894, "grad_norm": 0.6005427837371826, "learning_rate": 0.0002, "epoch": 1.4679492361167246, "step": 9080}, {"loss": 0.7288, "grad_norm": 0.718564510345459, "learning_rate": 0.0002, "epoch": 1.46956592029747, "step": 9090}, {"loss": 0.7089, "grad_norm": 0.7003577351570129, "learning_rate": 0.0002, "epoch": 1.4711826044782153, "step": 9100}, {"loss": 0.8069, "grad_norm": 0.5888323783874512, "learning_rate": 0.0002, "epoch": 1.4727992886589605, "step": 9110}, {"loss": 0.7275, "grad_norm": 0.6417609453201294, "learning_rate": 0.0002, "epoch": 1.4744159728397057, "step": 9120}, {"loss": 0.7441, "grad_norm": 0.572294294834137, "learning_rate": 0.0002, "epoch": 1.476032657020451, "step": 9130}, {"loss": 0.8053, "grad_norm": 0.8200714588165283, "learning_rate": 0.0002, "epoch": 1.4776493412011964, "step": 9140}, {"loss": 0.7382, "grad_norm": 0.6343288421630859, "learning_rate": 0.0002, "epoch": 1.4792660253819416, "step": 9150}, {"loss": 0.7641, "grad_norm": 0.7017961144447327, "learning_rate": 0.0002, "epoch": 1.4808827095626869, "step": 9160}, {"loss": 0.7619, "grad_norm": 0.6202912926673889, "learning_rate": 0.0002, "epoch": 1.4824993937434323, "step": 9170}, {"loss": 0.7428, "grad_norm": 0.6677869558334351, "learning_rate": 0.0002, "epoch": 1.4841160779241775, "step": 9180}, {"loss": 0.7648, "grad_norm": 0.6052267551422119, "learning_rate": 0.0002, "epoch": 1.4857327621049228, "step": 9190}, {"loss": 0.7152, "grad_norm": 0.6638872027397156, "learning_rate": 0.0002, "epoch": 1.487349446285668, "step": 9200}, {"loss": 0.7448, "grad_norm": 0.6245523691177368, "learning_rate": 0.0002, "epoch": 1.4889661304664135, "step": 9210}, {"loss": 0.6958, "grad_norm": 0.5761767625808716, "learning_rate": 0.0002, "epoch": 1.4905828146471587, "step": 9220}, {"loss": 0.8012, "grad_norm": 0.8175981640815735, "learning_rate": 0.0002, "epoch": 1.492199498827904, "step": 9230}, {"loss": 0.683, "grad_norm": 0.9144009947776794, "learning_rate": 0.0002, "epoch": 1.4938161830086494, "step": 9240}, {"loss": 0.7623, "grad_norm": 0.5742552876472473, "learning_rate": 0.0002, "epoch": 1.4954328671893946, "step": 9250}, {"loss": 0.7418, "grad_norm": 0.534534215927124, "learning_rate": 0.0002, "epoch": 1.4970495513701398, "step": 9260}, {"loss": 0.7194, "grad_norm": 0.7836225032806396, "learning_rate": 0.0002, "epoch": 1.498666235550885, "step": 9270}, {"loss": 0.7453, "grad_norm": 0.5292993187904358, "learning_rate": 0.0002, "epoch": 1.5002829197316303, "step": 9280}, {"loss": 0.7168, "grad_norm": 0.8044071793556213, "learning_rate": 0.0002, "epoch": 1.5018996039123758, "step": 9290}, {"loss": 0.7229, "grad_norm": 0.6185805201530457, "learning_rate": 0.0002, "epoch": 1.503516288093121, "step": 9300}, {"loss": 0.684, "grad_norm": 0.6093607544898987, "learning_rate": 0.0002, "epoch": 1.5051329722738664, "step": 9310}, {"loss": 0.7973, "grad_norm": 0.5891730189323425, "learning_rate": 0.0002, "epoch": 1.5067496564546117, "step": 9320}, {"loss": 0.7474, "grad_norm": 0.6331129670143127, "learning_rate": 0.0002, "epoch": 1.508366340635357, "step": 9330}, {"loss": 0.7074, "grad_norm": 0.7690958380699158, "learning_rate": 0.0002, "epoch": 1.5099830248161021, "step": 9340}, {"loss": 0.672, "grad_norm": 0.6548877358436584, "learning_rate": 0.0002, "epoch": 1.5115997089968474, "step": 9350}, {"loss": 0.7408, "grad_norm": 0.6545143127441406, "learning_rate": 0.0002, "epoch": 1.5132163931775926, "step": 9360}, {"loss": 0.7432, "grad_norm": 0.553247332572937, "learning_rate": 0.0002, "epoch": 1.514833077358338, "step": 9370}, {"loss": 0.7265, "grad_norm": 0.8145074844360352, "learning_rate": 0.0002, "epoch": 1.5164497615390833, "step": 9380}, {"loss": 0.7379, "grad_norm": 0.7636994123458862, "learning_rate": 0.0002, "epoch": 1.5180664457198287, "step": 9390}, {"loss": 0.7413, "grad_norm": 0.6838982701301575, "learning_rate": 0.0002, "epoch": 1.519683129900574, "step": 9400}, {"loss": 0.7367, "grad_norm": 0.8599441647529602, "learning_rate": 0.0002, "epoch": 1.5212998140813192, "step": 9410}, {"loss": 0.7663, "grad_norm": 0.7020329833030701, "learning_rate": 0.0002, "epoch": 1.5229164982620644, "step": 9420}, {"loss": 0.7928, "grad_norm": 0.6964772343635559, "learning_rate": 0.0002, "epoch": 1.5245331824428097, "step": 9430}, {"loss": 0.7168, "grad_norm": 0.6916600465774536, "learning_rate": 0.0002, "epoch": 1.5261498666235551, "step": 9440}, {"loss": 0.7519, "grad_norm": 0.7282621264457703, "learning_rate": 0.0002, "epoch": 1.5277665508043003, "step": 9450}, {"loss": 0.7628, "grad_norm": 0.5363983511924744, "learning_rate": 0.0002, "epoch": 1.5293832349850458, "step": 9460}, {"loss": 0.7154, "grad_norm": 0.6184861063957214, "learning_rate": 0.0002, "epoch": 1.530999919165791, "step": 9470}, {"loss": 0.7837, "grad_norm": 0.5991285443305969, "learning_rate": 0.0002, "epoch": 1.5326166033465363, "step": 9480}, {"loss": 0.7827, "grad_norm": 0.8176587820053101, "learning_rate": 0.0002, "epoch": 1.5342332875272815, "step": 9490}, {"loss": 0.7415, "grad_norm": 0.6473721861839294, "learning_rate": 0.0002, "epoch": 1.5358499717080267, "step": 9500}, {"loss": 0.7632, "grad_norm": 0.7319952845573425, "learning_rate": 0.0002, "epoch": 1.5374666558887722, "step": 9510}, {"loss": 0.7706, "grad_norm": 0.702900230884552, "learning_rate": 0.0002, "epoch": 1.5390833400695174, "step": 9520}, {"loss": 0.7754, "grad_norm": 0.7971600294113159, "learning_rate": 0.0002, "epoch": 1.5407000242502629, "step": 9530}, {"loss": 0.7352, "grad_norm": 0.6527525186538696, "learning_rate": 0.0002, "epoch": 1.542316708431008, "step": 9540}, {"loss": 0.7425, "grad_norm": 0.5791676044464111, "learning_rate": 0.0002, "epoch": 1.5439333926117533, "step": 9550}, {"loss": 0.7585, "grad_norm": 0.5619390606880188, "learning_rate": 0.0002, "epoch": 1.5455500767924986, "step": 9560}, {"loss": 0.7894, "grad_norm": 0.5701689124107361, "learning_rate": 0.0002, "epoch": 1.5471667609732438, "step": 9570}, {"loss": 0.793, "grad_norm": 0.47549352049827576, "learning_rate": 0.0002, "epoch": 1.548783445153989, "step": 9580}, {"loss": 0.7276, "grad_norm": 0.8730611205101013, "learning_rate": 0.0002, "epoch": 1.5504001293347345, "step": 9590}, {"loss": 0.798, "grad_norm": 0.6842091083526611, "learning_rate": 0.0002, "epoch": 1.5520168135154797, "step": 9600}, {"loss": 0.7528, "grad_norm": 0.6675129532814026, "learning_rate": 0.0002, "epoch": 1.5536334976962252, "step": 9610}, {"loss": 0.7954, "grad_norm": 0.8173956274986267, "learning_rate": 0.0002, "epoch": 1.5552501818769704, "step": 9620}, {"loss": 0.7535, "grad_norm": 0.724947452545166, "learning_rate": 0.0002, "epoch": 1.5568668660577156, "step": 9630}, {"loss": 0.7738, "grad_norm": 0.6154758930206299, "learning_rate": 0.0002, "epoch": 1.5584835502384609, "step": 9640}, {"loss": 0.7568, "grad_norm": 0.6072008013725281, "learning_rate": 0.0002, "epoch": 1.560100234419206, "step": 9650}, {"loss": 0.7219, "grad_norm": 0.659010648727417, "learning_rate": 0.0002, "epoch": 1.5617169185999515, "step": 9660}, {"loss": 0.673, "grad_norm": 0.65857994556427, "learning_rate": 0.0002, "epoch": 1.5633336027806968, "step": 9670}, {"loss": 0.7156, "grad_norm": 0.5914267301559448, "learning_rate": 0.0002, "epoch": 1.5649502869614422, "step": 9680}, {"loss": 0.7414, "grad_norm": 0.6248020529747009, "learning_rate": 0.0002, "epoch": 1.5665669711421875, "step": 9690}, {"loss": 0.694, "grad_norm": 0.7147795557975769, "learning_rate": 0.0002, "epoch": 1.5681836553229327, "step": 9700}, {"loss": 0.7335, "grad_norm": 0.7076232433319092, "learning_rate": 0.0002, "epoch": 1.569800339503678, "step": 9710}, {"loss": 0.7413, "grad_norm": 0.6217400431632996, "learning_rate": 0.0002, "epoch": 1.5714170236844232, "step": 9720}, {"loss": 0.7296, "grad_norm": 0.6709911227226257, "learning_rate": 0.0002, "epoch": 1.5730337078651684, "step": 9730}, {"loss": 0.7306, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 1.5746503920459138, "step": 9740}, {"loss": 0.7242, "grad_norm": 0.6241145730018616, "learning_rate": 0.0002, "epoch": 1.576267076226659, "step": 9750}, {"loss": 0.7384, "grad_norm": 0.4960934817790985, "learning_rate": 0.0002, "epoch": 1.5778837604074045, "step": 9760}, {"loss": 0.725, "grad_norm": 0.6593309640884399, "learning_rate": 0.0002, "epoch": 1.5795004445881498, "step": 9770}, {"loss": 0.7531, "grad_norm": 0.5814042091369629, "learning_rate": 0.0002, "epoch": 1.581117128768895, "step": 9780}, {"loss": 0.7109, "grad_norm": 0.5936070680618286, "learning_rate": 0.0002, "epoch": 1.5827338129496402, "step": 9790}, {"loss": 0.7769, "grad_norm": 0.6454403400421143, "learning_rate": 0.0002, "epoch": 1.5843504971303854, "step": 9800}, {"loss": 0.7677, "grad_norm": 0.7612107992172241, "learning_rate": 0.0002, "epoch": 1.585967181311131, "step": 9810}, {"loss": 0.7649, "grad_norm": 0.6494482755661011, "learning_rate": 0.0002, "epoch": 1.5875838654918761, "step": 9820}, {"loss": 0.7569, "grad_norm": 0.7825694680213928, "learning_rate": 0.0002, "epoch": 1.5892005496726216, "step": 9830}, {"loss": 0.706, "grad_norm": 0.6757757663726807, "learning_rate": 0.0002, "epoch": 1.5908172338533668, "step": 9840}, {"loss": 0.7803, "grad_norm": 0.7105609178543091, "learning_rate": 0.0002, "epoch": 1.592433918034112, "step": 9850}, {"loss": 0.7925, "grad_norm": 0.7596991062164307, "learning_rate": 0.0002, "epoch": 1.5940506022148573, "step": 9860}, {"loss": 0.7108, "grad_norm": 0.5681525468826294, "learning_rate": 0.0002, "epoch": 1.5956672863956025, "step": 9870}, {"loss": 0.7811, "grad_norm": 0.6090980768203735, "learning_rate": 0.0002, "epoch": 1.5972839705763477, "step": 9880}, {"loss": 0.7339, "grad_norm": 0.6271613240242004, "learning_rate": 0.0002, "epoch": 1.5989006547570932, "step": 9890}, {"loss": 0.7419, "grad_norm": 0.7656369805335999, "learning_rate": 0.0002, "epoch": 1.6005173389378387, "step": 9900}, {"loss": 0.7336, "grad_norm": 0.7504446506500244, "learning_rate": 0.0002, "epoch": 1.6021340231185839, "step": 9910}, {"loss": 0.7479, "grad_norm": 0.659656286239624, "learning_rate": 0.0002, "epoch": 1.6037507072993291, "step": 9920}, {"loss": 0.7483, "grad_norm": 0.6006826162338257, "learning_rate": 0.0002, "epoch": 1.6053673914800743, "step": 9930}, {"loss": 0.732, "grad_norm": 0.7872757911682129, "learning_rate": 0.0002, "epoch": 1.6069840756608196, "step": 9940}, {"loss": 0.768, "grad_norm": 0.5545852780342102, "learning_rate": 0.0002, "epoch": 1.6086007598415648, "step": 9950}, {"loss": 0.8064, "grad_norm": 0.7429468631744385, "learning_rate": 0.0002, "epoch": 1.6102174440223103, "step": 9960}, {"loss": 0.714, "grad_norm": 0.6873556971549988, "learning_rate": 0.0002, "epoch": 1.6118341282030555, "step": 9970}, {"loss": 0.7324, "grad_norm": 0.5874287486076355, "learning_rate": 0.0002, "epoch": 1.613450812383801, "step": 9980}, {"loss": 0.7141, "grad_norm": 0.6039386987686157, "learning_rate": 0.0002, "epoch": 1.6150674965645462, "step": 9990}, {"loss": 0.6674, "grad_norm": 0.6233575940132141, "learning_rate": 0.0002, "epoch": 1.6166841807452914, "step": 10000}, {"loss": 0.7602, "grad_norm": 0.7676448225975037, "learning_rate": 0.0002, "epoch": 1.6183008649260366, "step": 10010}, {"loss": 0.7784, "grad_norm": 0.6565698385238647, "learning_rate": 0.0002, "epoch": 1.6199175491067819, "step": 10020}, {"loss": 0.7104, "grad_norm": 0.6787590384483337, "learning_rate": 0.0002, "epoch": 1.6215342332875273, "step": 10030}, {"loss": 0.7464, "grad_norm": 0.6137678027153015, "learning_rate": 0.0002, "epoch": 1.6231509174682726, "step": 10040}, {"loss": 0.7646, "grad_norm": 0.5236800312995911, "learning_rate": 0.0002, "epoch": 1.624767601649018, "step": 10050}, {"loss": 0.7437, "grad_norm": 0.7626367807388306, "learning_rate": 0.0002, "epoch": 1.6263842858297632, "step": 10060}, {"loss": 0.7273, "grad_norm": 0.5657260417938232, "learning_rate": 0.0002, "epoch": 1.6280009700105085, "step": 10070}, {"loss": 0.7354, "grad_norm": 0.4913991391658783, "learning_rate": 0.0002, "epoch": 1.6296176541912537, "step": 10080}, {"loss": 0.7596, "grad_norm": 0.7715556621551514, "learning_rate": 0.0002, "epoch": 1.631234338371999, "step": 10090}, {"loss": 0.7105, "grad_norm": 0.6509000062942505, "learning_rate": 0.0002, "epoch": 1.6328510225527442, "step": 10100}, {"loss": 0.7274, "grad_norm": 0.6215850114822388, "learning_rate": 0.0002, "epoch": 1.6344677067334896, "step": 10110}, {"loss": 0.7705, "grad_norm": 0.6956844329833984, "learning_rate": 0.0002, "epoch": 1.6360843909142349, "step": 10120}, {"loss": 0.7129, "grad_norm": 0.6111597418785095, "learning_rate": 0.0002, "epoch": 1.6377010750949803, "step": 10130}, {"loss": 0.6955, "grad_norm": 0.6518288850784302, "learning_rate": 0.0002, "epoch": 1.6393177592757255, "step": 10140}, {"loss": 0.731, "grad_norm": 0.6914522051811218, "learning_rate": 0.0002, "epoch": 1.6409344434564708, "step": 10150}, {"loss": 0.7295, "grad_norm": 0.63785719871521, "learning_rate": 0.0002, "epoch": 1.642551127637216, "step": 10160}, {"loss": 0.7355, "grad_norm": 0.6379287838935852, "learning_rate": 0.0002, "epoch": 1.6441678118179612, "step": 10170}, {"loss": 0.7359, "grad_norm": 0.6793403029441833, "learning_rate": 0.0002, "epoch": 1.6457844959987067, "step": 10180}, {"loss": 0.7402, "grad_norm": 0.6099132895469666, "learning_rate": 0.0002, "epoch": 1.647401180179452, "step": 10190}, {"loss": 0.7353, "grad_norm": 0.5869854092597961, "learning_rate": 0.0002, "epoch": 1.6490178643601974, "step": 10200}, {"loss": 0.8308, "grad_norm": 0.7716999053955078, "learning_rate": 0.0002, "epoch": 1.6506345485409426, "step": 10210}, {"loss": 0.7215, "grad_norm": 0.6854110360145569, "learning_rate": 0.0002, "epoch": 1.6522512327216878, "step": 10220}, {"loss": 0.782, "grad_norm": 0.6957170367240906, "learning_rate": 0.0002, "epoch": 1.653867916902433, "step": 10230}, {"loss": 0.7282, "grad_norm": 0.6932903528213501, "learning_rate": 0.0002, "epoch": 1.6554846010831783, "step": 10240}, {"loss": 0.7478, "grad_norm": 0.7713165283203125, "learning_rate": 0.0002, "epoch": 1.6571012852639235, "step": 10250}, {"loss": 0.7099, "grad_norm": 0.7455793619155884, "learning_rate": 0.0002, "epoch": 1.658717969444669, "step": 10260}, {"loss": 0.7524, "grad_norm": 0.5464168190956116, "learning_rate": 0.0002, "epoch": 1.6603346536254144, "step": 10270}, {"loss": 0.7328, "grad_norm": 0.6782926321029663, "learning_rate": 0.0002, "epoch": 1.6619513378061597, "step": 10280}, {"loss": 0.7801, "grad_norm": 0.7962649464607239, "learning_rate": 0.0002, "epoch": 1.663568021986905, "step": 10290}, {"loss": 0.7142, "grad_norm": 0.6814526319503784, "learning_rate": 0.0002, "epoch": 1.6651847061676501, "step": 10300}, {"loss": 0.7285, "grad_norm": 0.656895101070404, "learning_rate": 0.0002, "epoch": 1.6668013903483954, "step": 10310}, {"loss": 0.7358, "grad_norm": 0.6085672378540039, "learning_rate": 0.0002, "epoch": 1.6684180745291406, "step": 10320}, {"loss": 0.7074, "grad_norm": 0.585508406162262, "learning_rate": 0.0002, "epoch": 1.670034758709886, "step": 10330}, {"loss": 0.7604, "grad_norm": 0.6930184364318848, "learning_rate": 0.0002, "epoch": 1.6716514428906313, "step": 10340}, {"loss": 0.7169, "grad_norm": 0.575663149356842, "learning_rate": 0.0002, "epoch": 1.6732681270713767, "step": 10350}, {"loss": 0.7198, "grad_norm": 0.582502543926239, "learning_rate": 0.0002, "epoch": 1.674884811252122, "step": 10360}, {"loss": 0.7793, "grad_norm": 0.5668916702270508, "learning_rate": 0.0002, "epoch": 1.6765014954328672, "step": 10370}, {"loss": 0.7478, "grad_norm": 0.6070065498352051, "learning_rate": 0.0002, "epoch": 1.6781181796136124, "step": 10380}, {"loss": 0.7939, "grad_norm": 0.6141316294670105, "learning_rate": 0.0002, "epoch": 1.6797348637943577, "step": 10390}, {"loss": 0.7573, "grad_norm": 0.8359124064445496, "learning_rate": 0.0002, "epoch": 1.6813515479751031, "step": 10400}, {"loss": 0.7488, "grad_norm": 0.5378185510635376, "learning_rate": 0.0002, "epoch": 1.6829682321558483, "step": 10410}, {"loss": 0.7588, "grad_norm": 0.6959536075592041, "learning_rate": 0.0002, "epoch": 1.6845849163365938, "step": 10420}, {"loss": 0.7872, "grad_norm": 0.6514357328414917, "learning_rate": 0.0002, "epoch": 1.686201600517339, "step": 10430}, {"loss": 0.725, "grad_norm": 0.7706646919250488, "learning_rate": 0.0002, "epoch": 1.6878182846980843, "step": 10440}, {"loss": 0.7673, "grad_norm": 0.6183337569236755, "learning_rate": 0.0002, "epoch": 1.6894349688788295, "step": 10450}, {"loss": 0.7566, "grad_norm": 0.6123278141021729, "learning_rate": 0.0002, "epoch": 1.6910516530595747, "step": 10460}, {"loss": 0.7169, "grad_norm": 0.6894851326942444, "learning_rate": 0.0002, "epoch": 1.69266833724032, "step": 10470}, {"loss": 0.7435, "grad_norm": 0.7497312426567078, "learning_rate": 0.0002, "epoch": 1.6942850214210654, "step": 10480}, {"loss": 0.7544, "grad_norm": 0.5968214273452759, "learning_rate": 0.0002, "epoch": 1.6959017056018106, "step": 10490}, {"loss": 0.6793, "grad_norm": 0.6747927069664001, "learning_rate": 0.0002, "epoch": 1.697518389782556, "step": 10500}, {"loss": 0.7415, "grad_norm": 0.5708310008049011, "learning_rate": 0.0002, "epoch": 1.6991350739633013, "step": 10510}, {"loss": 0.7385, "grad_norm": 0.606526792049408, "learning_rate": 0.0002, "epoch": 1.7007517581440466, "step": 10520}, {"loss": 0.7204, "grad_norm": 0.662011981010437, "learning_rate": 0.0002, "epoch": 1.7023684423247918, "step": 10530}, {"loss": 0.7999, "grad_norm": 0.7583045363426208, "learning_rate": 0.0002, "epoch": 1.703985126505537, "step": 10540}, {"loss": 0.7563, "grad_norm": 0.721632182598114, "learning_rate": 0.0002, "epoch": 1.7056018106862825, "step": 10550}, {"loss": 0.7407, "grad_norm": 0.6107715368270874, "learning_rate": 0.0002, "epoch": 1.7072184948670277, "step": 10560}, {"loss": 0.7519, "grad_norm": 0.6652471423149109, "learning_rate": 0.0002, "epoch": 1.7088351790477732, "step": 10570}, {"loss": 0.7767, "grad_norm": 0.6308087110519409, "learning_rate": 0.0002, "epoch": 1.7104518632285184, "step": 10580}, {"loss": 0.7659, "grad_norm": 0.5464386940002441, "learning_rate": 0.0002, "epoch": 1.7120685474092636, "step": 10590}, {"loss": 0.7063, "grad_norm": 0.6558911204338074, "learning_rate": 0.0002, "epoch": 1.7136852315900089, "step": 10600}, {"loss": 0.7126, "grad_norm": 0.5665024518966675, "learning_rate": 0.0002, "epoch": 1.715301915770754, "step": 10610}, {"loss": 0.6958, "grad_norm": 0.7888094186782837, "learning_rate": 0.0002, "epoch": 1.7169185999514993, "step": 10620}, {"loss": 0.7785, "grad_norm": 0.7084909081459045, "learning_rate": 0.0002, "epoch": 1.7185352841322448, "step": 10630}, {"loss": 0.7557, "grad_norm": 0.7982324361801147, "learning_rate": 0.0002, "epoch": 1.7201519683129902, "step": 10640}, {"loss": 0.7345, "grad_norm": 0.6418732404708862, "learning_rate": 0.0002, "epoch": 1.7217686524937355, "step": 10650}, {"loss": 0.7734, "grad_norm": 0.7636681795120239, "learning_rate": 0.0002, "epoch": 1.7233853366744807, "step": 10660}, {"loss": 0.7541, "grad_norm": 0.5646875500679016, "learning_rate": 0.0002, "epoch": 1.725002020855226, "step": 10670}, {"loss": 0.7642, "grad_norm": 0.5231260657310486, "learning_rate": 0.0002, "epoch": 1.7266187050359711, "step": 10680}, {"loss": 0.7846, "grad_norm": 0.7635011672973633, "learning_rate": 0.0002, "epoch": 1.7282353892167164, "step": 10690}, {"loss": 0.7471, "grad_norm": 0.7518259286880493, "learning_rate": 0.0002, "epoch": 1.7298520733974618, "step": 10700}, {"loss": 0.751, "grad_norm": 0.7295602560043335, "learning_rate": 0.0002, "epoch": 1.731468757578207, "step": 10710}, {"loss": 0.731, "grad_norm": 0.6984632015228271, "learning_rate": 0.0002, "epoch": 1.7330854417589525, "step": 10720}, {"loss": 0.7921, "grad_norm": 0.6198219060897827, "learning_rate": 0.0002, "epoch": 1.7347021259396977, "step": 10730}, {"loss": 0.7642, "grad_norm": 0.6957576274871826, "learning_rate": 0.0002, "epoch": 1.736318810120443, "step": 10740}, {"loss": 0.7917, "grad_norm": 0.6430263519287109, "learning_rate": 0.0002, "epoch": 1.7379354943011882, "step": 10750}, {"loss": 0.7156, "grad_norm": 0.6134995222091675, "learning_rate": 0.0002, "epoch": 1.7395521784819334, "step": 10760}, {"loss": 0.7584, "grad_norm": 0.7209452986717224, "learning_rate": 0.0002, "epoch": 1.741168862662679, "step": 10770}, {"loss": 0.7528, "grad_norm": 0.6735447645187378, "learning_rate": 0.0002, "epoch": 1.7427855468434241, "step": 10780}, {"loss": 0.756, "grad_norm": 0.5605693459510803, "learning_rate": 0.0002, "epoch": 1.7444022310241696, "step": 10790}, {"loss": 0.7759, "grad_norm": 0.6882363557815552, "learning_rate": 0.0002, "epoch": 1.7460189152049148, "step": 10800}, {"loss": 0.7544, "grad_norm": 0.6386259198188782, "learning_rate": 0.0002, "epoch": 1.74763559938566, "step": 10810}, {"loss": 0.7697, "grad_norm": 0.6529015302658081, "learning_rate": 0.0002, "epoch": 1.7492522835664053, "step": 10820}, {"loss": 0.7219, "grad_norm": 0.5664082765579224, "learning_rate": 0.0002, "epoch": 1.7508689677471505, "step": 10830}, {"loss": 0.7586, "grad_norm": 0.7532684206962585, "learning_rate": 0.0002, "epoch": 1.7524856519278957, "step": 10840}, {"loss": 0.6919, "grad_norm": 0.77171391248703, "learning_rate": 0.0002, "epoch": 1.7541023361086412, "step": 10850}, {"loss": 0.785, "grad_norm": 0.7255431413650513, "learning_rate": 0.0002, "epoch": 1.7557190202893864, "step": 10860}, {"loss": 0.7458, "grad_norm": 0.763083279132843, "learning_rate": 0.0002, "epoch": 1.7573357044701319, "step": 10870}, {"loss": 0.7846, "grad_norm": 0.6042402982711792, "learning_rate": 0.0002, "epoch": 1.758952388650877, "step": 10880}, {"loss": 0.7027, "grad_norm": 0.7642518281936646, "learning_rate": 0.0002, "epoch": 1.7605690728316223, "step": 10890}, {"loss": 0.746, "grad_norm": 0.6347904801368713, "learning_rate": 0.0002, "epoch": 1.7621857570123676, "step": 10900}, {"loss": 0.7458, "grad_norm": 0.5371627807617188, "learning_rate": 0.0002, "epoch": 1.7638024411931128, "step": 10910}, {"loss": 0.7466, "grad_norm": 0.6840225458145142, "learning_rate": 0.0002, "epoch": 1.7654191253738583, "step": 10920}, {"loss": 0.725, "grad_norm": 0.5288469195365906, "learning_rate": 0.0002, "epoch": 1.7670358095546035, "step": 10930}, {"loss": 0.7863, "grad_norm": 0.69020676612854, "learning_rate": 0.0002, "epoch": 1.768652493735349, "step": 10940}, {"loss": 0.7468, "grad_norm": 0.5943242311477661, "learning_rate": 0.0002, "epoch": 1.7702691779160942, "step": 10950}, {"loss": 0.7244, "grad_norm": 0.5616418123245239, "learning_rate": 0.0002, "epoch": 1.7718858620968394, "step": 10960}, {"loss": 0.7137, "grad_norm": 0.7209470868110657, "learning_rate": 0.0002, "epoch": 1.7735025462775846, "step": 10970}, {"loss": 0.7459, "grad_norm": 0.6657957434654236, "learning_rate": 0.0002, "epoch": 1.7751192304583299, "step": 10980}, {"loss": 0.7076, "grad_norm": 0.6469064950942993, "learning_rate": 0.0002, "epoch": 1.776735914639075, "step": 10990}, {"loss": 0.7321, "grad_norm": 0.6615678071975708, "learning_rate": 0.0002, "epoch": 1.7783525988198206, "step": 11000}, {"loss": 0.747, "grad_norm": 0.6722439527511597, "learning_rate": 0.0002, "epoch": 1.779969283000566, "step": 11010}, {"loss": 0.7302, "grad_norm": 0.634136974811554, "learning_rate": 0.0002, "epoch": 1.7815859671813112, "step": 11020}, {"loss": 0.8105, "grad_norm": 0.6024377346038818, "learning_rate": 0.0002, "epoch": 1.7832026513620565, "step": 11030}, {"loss": 0.7855, "grad_norm": 0.6909403800964355, "learning_rate": 0.0002, "epoch": 1.7848193355428017, "step": 11040}, {"loss": 0.7471, "grad_norm": 0.7148767709732056, "learning_rate": 0.0002, "epoch": 1.786436019723547, "step": 11050}, {"loss": 0.7145, "grad_norm": 0.7442979216575623, "learning_rate": 0.0002, "epoch": 1.7880527039042922, "step": 11060}, {"loss": 0.7215, "grad_norm": 0.6830431818962097, "learning_rate": 0.0002, "epoch": 1.7896693880850376, "step": 11070}, {"loss": 0.7625, "grad_norm": 0.9172667264938354, "learning_rate": 0.0002, "epoch": 1.7912860722657828, "step": 11080}, {"loss": 0.76, "grad_norm": 0.6799490451812744, "learning_rate": 0.0002, "epoch": 1.7929027564465283, "step": 11090}, {"loss": 0.7716, "grad_norm": 0.7617024779319763, "learning_rate": 0.0002, "epoch": 1.7945194406272735, "step": 11100}, {"loss": 0.7586, "grad_norm": 0.7701810002326965, "learning_rate": 0.0002, "epoch": 1.7961361248080188, "step": 11110}, {"loss": 0.7843, "grad_norm": 0.7454385757446289, "learning_rate": 0.0002, "epoch": 1.797752808988764, "step": 11120}, {"loss": 0.7873, "grad_norm": 0.6121436953544617, "learning_rate": 0.0002, "epoch": 1.7993694931695092, "step": 11130}, {"loss": 0.7305, "grad_norm": 0.6237571835517883, "learning_rate": 0.0002, "epoch": 1.8009861773502547, "step": 11140}, {"loss": 0.6827, "grad_norm": 0.6818515658378601, "learning_rate": 0.0002, "epoch": 1.802602861531, "step": 11150}, {"loss": 0.6876, "grad_norm": 0.7768308520317078, "learning_rate": 0.0002, "epoch": 1.8042195457117454, "step": 11160}, {"loss": 0.7533, "grad_norm": 0.6875537633895874, "learning_rate": 0.0002, "epoch": 1.8058362298924906, "step": 11170}, {"loss": 0.761, "grad_norm": 0.7950584888458252, "learning_rate": 0.0002, "epoch": 1.8074529140732358, "step": 11180}, {"loss": 0.7623, "grad_norm": 0.8210248351097107, "learning_rate": 0.0002, "epoch": 1.809069598253981, "step": 11190}, {"loss": 0.7556, "grad_norm": 0.6674110889434814, "learning_rate": 0.0002, "epoch": 1.8106862824347263, "step": 11200}, {"loss": 0.7663, "grad_norm": 0.6261674761772156, "learning_rate": 0.0002, "epoch": 1.8123029666154715, "step": 11210}, {"loss": 0.7122, "grad_norm": 0.6484741568565369, "learning_rate": 0.0002, "epoch": 1.813919650796217, "step": 11220}, {"loss": 0.7718, "grad_norm": 0.6231244206428528, "learning_rate": 0.0002, "epoch": 1.8155363349769622, "step": 11230}, {"loss": 0.7152, "grad_norm": 0.7243146896362305, "learning_rate": 0.0002, "epoch": 1.8171530191577077, "step": 11240}, {"loss": 0.7448, "grad_norm": 0.6776193380355835, "learning_rate": 0.0002, "epoch": 1.818769703338453, "step": 11250}, {"loss": 0.7317, "grad_norm": 0.5973618030548096, "learning_rate": 0.0002, "epoch": 1.8203863875191981, "step": 11260}, {"loss": 0.7961, "grad_norm": 0.6451361179351807, "learning_rate": 0.0002, "epoch": 1.8220030716999434, "step": 11270}, {"loss": 0.7611, "grad_norm": 0.5963068008422852, "learning_rate": 0.0002, "epoch": 1.8236197558806886, "step": 11280}, {"loss": 0.7466, "grad_norm": 0.536902129650116, "learning_rate": 0.0002, "epoch": 1.825236440061434, "step": 11290}, {"loss": 0.708, "grad_norm": 0.6993787288665771, "learning_rate": 0.0002, "epoch": 1.8268531242421793, "step": 11300}, {"loss": 0.7153, "grad_norm": 0.6135255098342896, "learning_rate": 0.0002, "epoch": 1.8284698084229247, "step": 11310}, {"loss": 0.7423, "grad_norm": 0.6057423949241638, "learning_rate": 0.0002, "epoch": 1.83008649260367, "step": 11320}, {"loss": 0.735, "grad_norm": 0.6598812341690063, "learning_rate": 0.0002, "epoch": 1.8317031767844152, "step": 11330}, {"loss": 0.7278, "grad_norm": 0.6075948476791382, "learning_rate": 0.0002, "epoch": 1.8333198609651604, "step": 11340}, {"loss": 0.7846, "grad_norm": 0.7065447568893433, "learning_rate": 0.0002, "epoch": 1.8349365451459057, "step": 11350}, {"loss": 0.7365, "grad_norm": 0.680526614189148, "learning_rate": 0.0002, "epoch": 1.8365532293266509, "step": 11360}, {"loss": 0.7152, "grad_norm": 0.6356695294380188, "learning_rate": 0.0002, "epoch": 1.8381699135073963, "step": 11370}, {"loss": 0.721, "grad_norm": 0.6399052143096924, "learning_rate": 0.0002, "epoch": 1.8397865976881416, "step": 11380}, {"loss": 0.7618, "grad_norm": 0.6125704050064087, "learning_rate": 0.0002, "epoch": 1.841403281868887, "step": 11390}, {"loss": 0.755, "grad_norm": 0.7124643325805664, "learning_rate": 0.0002, "epoch": 1.8430199660496323, "step": 11400}, {"loss": 0.7972, "grad_norm": 0.6099604964256287, "learning_rate": 0.0002, "epoch": 1.8446366502303775, "step": 11410}, {"loss": 0.7187, "grad_norm": 0.7338208556175232, "learning_rate": 0.0002, "epoch": 1.8462533344111227, "step": 11420}, {"loss": 0.7007, "grad_norm": 0.7534668445587158, "learning_rate": 0.0002, "epoch": 1.847870018591868, "step": 11430}, {"loss": 0.7464, "grad_norm": 0.6135470271110535, "learning_rate": 0.0002, "epoch": 1.8494867027726134, "step": 11440}, {"loss": 0.7955, "grad_norm": 0.6229309439659119, "learning_rate": 0.0002, "epoch": 1.8511033869533586, "step": 11450}, {"loss": 0.7594, "grad_norm": 0.706423282623291, "learning_rate": 0.0002, "epoch": 1.852720071134104, "step": 11460}, {"loss": 0.7411, "grad_norm": 0.5460049510002136, "learning_rate": 0.0002, "epoch": 1.8543367553148493, "step": 11470}, {"loss": 0.7416, "grad_norm": 0.6616711020469666, "learning_rate": 0.0002, "epoch": 1.8559534394955945, "step": 11480}, {"loss": 0.729, "grad_norm": 0.6372783184051514, "learning_rate": 0.0002, "epoch": 1.8575701236763398, "step": 11490}, {"loss": 0.7333, "grad_norm": 0.7162668108940125, "learning_rate": 0.0002, "epoch": 1.859186807857085, "step": 11500}, {"loss": 0.7747, "grad_norm": 0.6605209708213806, "learning_rate": 0.0002, "epoch": 1.8608034920378305, "step": 11510}, {"loss": 0.7258, "grad_norm": 0.6933956742286682, "learning_rate": 0.0002, "epoch": 1.8624201762185757, "step": 11520}, {"loss": 0.7243, "grad_norm": 0.6582090854644775, "learning_rate": 0.0002, "epoch": 1.8640368603993211, "step": 11530}, {"loss": 0.7313, "grad_norm": 0.6416500806808472, "learning_rate": 0.0002, "epoch": 1.8656535445800664, "step": 11540}, {"loss": 0.7372, "grad_norm": 0.5434312224388123, "learning_rate": 0.0002, "epoch": 1.8672702287608116, "step": 11550}, {"loss": 0.7635, "grad_norm": 0.6827567219734192, "learning_rate": 0.0002, "epoch": 1.8688869129415568, "step": 11560}, {"loss": 0.7137, "grad_norm": 0.7354370951652527, "learning_rate": 0.0002, "epoch": 1.870503597122302, "step": 11570}, {"loss": 0.7526, "grad_norm": 0.590372622013092, "learning_rate": 0.0002, "epoch": 1.8721202813030473, "step": 11580}, {"loss": 0.731, "grad_norm": 0.853183925151825, "learning_rate": 0.0002, "epoch": 1.8737369654837928, "step": 11590}, {"loss": 0.7487, "grad_norm": 0.822678804397583, "learning_rate": 0.0002, "epoch": 1.875353649664538, "step": 11600}, {"loss": 0.7427, "grad_norm": 0.6591550707817078, "learning_rate": 0.0002, "epoch": 1.8769703338452834, "step": 11610}, {"loss": 0.7054, "grad_norm": 0.7475301623344421, "learning_rate": 0.0002, "epoch": 1.8785870180260287, "step": 11620}, {"loss": 0.811, "grad_norm": 0.6390765309333801, "learning_rate": 0.0002, "epoch": 1.880203702206774, "step": 11630}, {"loss": 0.7531, "grad_norm": 0.6589758992195129, "learning_rate": 0.0002, "epoch": 1.8818203863875191, "step": 11640}, {"loss": 0.7475, "grad_norm": 0.6765508651733398, "learning_rate": 0.0002, "epoch": 1.8834370705682644, "step": 11650}, {"loss": 0.738, "grad_norm": 0.6527857780456543, "learning_rate": 0.0002, "epoch": 1.8850537547490098, "step": 11660}, {"loss": 0.7504, "grad_norm": 0.6642923951148987, "learning_rate": 0.0002, "epoch": 1.886670438929755, "step": 11670}, {"loss": 0.7701, "grad_norm": 0.6945584416389465, "learning_rate": 0.0002, "epoch": 1.8882871231105005, "step": 11680}, {"loss": 0.7711, "grad_norm": 0.694018542766571, "learning_rate": 0.0002, "epoch": 1.8899038072912457, "step": 11690}, {"loss": 0.7195, "grad_norm": 0.7237417101860046, "learning_rate": 0.0002, "epoch": 1.891520491471991, "step": 11700}, {"loss": 0.7491, "grad_norm": 0.7401309609413147, "learning_rate": 0.0002, "epoch": 1.8931371756527362, "step": 11710}, {"loss": 0.805, "grad_norm": 0.6537784337997437, "learning_rate": 0.0002, "epoch": 1.8947538598334814, "step": 11720}, {"loss": 0.793, "grad_norm": 0.7398539185523987, "learning_rate": 0.0002, "epoch": 1.8963705440142267, "step": 11730}, {"loss": 0.7561, "grad_norm": 0.6696075797080994, "learning_rate": 0.0002, "epoch": 1.8979872281949721, "step": 11740}, {"loss": 0.7353, "grad_norm": 0.6014142036437988, "learning_rate": 0.0002, "epoch": 1.8996039123757174, "step": 11750}, {"loss": 0.7714, "grad_norm": 0.7023524641990662, "learning_rate": 0.0002, "epoch": 1.9012205965564628, "step": 11760}, {"loss": 0.7088, "grad_norm": 0.739973783493042, "learning_rate": 0.0002, "epoch": 1.902837280737208, "step": 11770}, {"loss": 0.7848, "grad_norm": 0.5576770901679993, "learning_rate": 0.0002, "epoch": 1.9044539649179533, "step": 11780}, {"loss": 0.7483, "grad_norm": 0.6907393932342529, "learning_rate": 0.0002, "epoch": 1.9060706490986985, "step": 11790}, {"loss": 0.7827, "grad_norm": 0.6934581995010376, "learning_rate": 0.0002, "epoch": 1.9076873332794437, "step": 11800}, {"loss": 0.7199, "grad_norm": 0.591774582862854, "learning_rate": 0.0002, "epoch": 1.9093040174601892, "step": 11810}, {"loss": 0.7333, "grad_norm": 0.6249791383743286, "learning_rate": 0.0002, "epoch": 1.9109207016409344, "step": 11820}, {"loss": 0.7581, "grad_norm": 0.6755744218826294, "learning_rate": 0.0002, "epoch": 1.9125373858216799, "step": 11830}, {"loss": 0.696, "grad_norm": 0.7286285161972046, "learning_rate": 0.0002, "epoch": 1.914154070002425, "step": 11840}, {"loss": 0.7509, "grad_norm": 0.7867850065231323, "learning_rate": 0.0002, "epoch": 1.9157707541831703, "step": 11850}, {"loss": 0.735, "grad_norm": 0.6283972859382629, "learning_rate": 0.0002, "epoch": 1.9173874383639156, "step": 11860}, {"loss": 0.7296, "grad_norm": 0.605823814868927, "learning_rate": 0.0002, "epoch": 1.9190041225446608, "step": 11870}, {"loss": 0.6598, "grad_norm": 0.5927976965904236, "learning_rate": 0.0002, "epoch": 1.920620806725406, "step": 11880}, {"loss": 0.7649, "grad_norm": 0.5974002480506897, "learning_rate": 0.0002, "epoch": 1.9222374909061515, "step": 11890}, {"loss": 0.7843, "grad_norm": 0.7091866135597229, "learning_rate": 0.0002, "epoch": 1.923854175086897, "step": 11900}, {"loss": 0.775, "grad_norm": 0.72496497631073, "learning_rate": 0.0002, "epoch": 1.9254708592676422, "step": 11910}, {"loss": 0.7153, "grad_norm": 0.6131896376609802, "learning_rate": 0.0002, "epoch": 1.9270875434483874, "step": 11920}, {"loss": 0.7228, "grad_norm": 0.6556436419487, "learning_rate": 0.0002, "epoch": 1.9287042276291326, "step": 11930}, {"loss": 0.7319, "grad_norm": 0.622932493686676, "learning_rate": 0.0002, "epoch": 1.9303209118098779, "step": 11940}, {"loss": 0.7592, "grad_norm": 0.6618631482124329, "learning_rate": 0.0002, "epoch": 1.931937595990623, "step": 11950}, {"loss": 0.8332, "grad_norm": 0.630966305732727, "learning_rate": 0.0002, "epoch": 1.9335542801713685, "step": 11960}, {"loss": 0.6854, "grad_norm": 0.6336734890937805, "learning_rate": 0.0002, "epoch": 1.9351709643521138, "step": 11970}, {"loss": 0.7433, "grad_norm": 0.655403196811676, "learning_rate": 0.0002, "epoch": 1.9367876485328592, "step": 11980}, {"loss": 0.7282, "grad_norm": 0.5640574097633362, "learning_rate": 0.0002, "epoch": 1.9384043327136045, "step": 11990}, {"loss": 0.7289, "grad_norm": 0.6322951316833496, "learning_rate": 0.0002, "epoch": 1.9400210168943497, "step": 12000}, {"loss": 0.7627, "grad_norm": 0.615703821182251, "learning_rate": 0.0002, "epoch": 1.941637701075095, "step": 12010}, {"loss": 0.786, "grad_norm": 0.6487536430358887, "learning_rate": 0.0002, "epoch": 1.9432543852558402, "step": 12020}, {"loss": 0.7435, "grad_norm": 0.9209630489349365, "learning_rate": 0.0002, "epoch": 1.9448710694365856, "step": 12030}, {"loss": 0.7274, "grad_norm": 0.67485511302948, "learning_rate": 0.0002, "epoch": 1.9464877536173308, "step": 12040}, {"loss": 0.7551, "grad_norm": 0.6831230521202087, "learning_rate": 0.0002, "epoch": 1.9481044377980763, "step": 12050}, {"loss": 0.7546, "grad_norm": 0.6578302383422852, "learning_rate": 0.0002, "epoch": 1.9497211219788215, "step": 12060}, {"loss": 0.6989, "grad_norm": 0.9975938200950623, "learning_rate": 0.0002, "epoch": 1.9513378061595668, "step": 12070}, {"loss": 0.7952, "grad_norm": 0.6637365221977234, "learning_rate": 0.0002, "epoch": 1.952954490340312, "step": 12080}, {"loss": 0.7482, "grad_norm": 0.605707049369812, "learning_rate": 0.0002, "epoch": 1.9545711745210572, "step": 12090}, {"loss": 0.7768, "grad_norm": 0.6584440469741821, "learning_rate": 0.0002, "epoch": 1.9561878587018025, "step": 12100}, {"loss": 0.7187, "grad_norm": 0.6070835590362549, "learning_rate": 0.0002, "epoch": 1.957804542882548, "step": 12110}, {"loss": 0.7491, "grad_norm": 0.7862601280212402, "learning_rate": 0.0002, "epoch": 1.9594212270632931, "step": 12120}, {"loss": 0.7972, "grad_norm": 0.8175255060195923, "learning_rate": 0.0002, "epoch": 1.9610379112440386, "step": 12130}, {"loss": 0.7242, "grad_norm": 0.5648472905158997, "learning_rate": 0.0002, "epoch": 1.9626545954247838, "step": 12140}, {"loss": 0.7321, "grad_norm": 0.6591973304748535, "learning_rate": 0.0002, "epoch": 1.964271279605529, "step": 12150}, {"loss": 0.739, "grad_norm": 0.5960676074028015, "learning_rate": 0.0002, "epoch": 1.9658879637862743, "step": 12160}, {"loss": 0.7254, "grad_norm": 0.7272544503211975, "learning_rate": 0.0002, "epoch": 1.9675046479670195, "step": 12170}, {"loss": 0.7376, "grad_norm": 0.7176699042320251, "learning_rate": 0.0002, "epoch": 1.969121332147765, "step": 12180}, {"loss": 0.7525, "grad_norm": 0.6927123665809631, "learning_rate": 0.0002, "epoch": 1.9707380163285102, "step": 12190}, {"loss": 0.7318, "grad_norm": 0.5536034107208252, "learning_rate": 0.0002, "epoch": 1.9723547005092557, "step": 12200}, {"loss": 0.7737, "grad_norm": 0.8348390460014343, "learning_rate": 0.0002, "epoch": 1.9739713846900009, "step": 12210}, {"loss": 0.7494, "grad_norm": 0.6591181755065918, "learning_rate": 0.0002, "epoch": 1.9755880688707461, "step": 12220}, {"loss": 0.763, "grad_norm": 1.0624109506607056, "learning_rate": 0.0002, "epoch": 1.9772047530514913, "step": 12230}, {"loss": 0.7541, "grad_norm": 0.9265586137771606, "learning_rate": 0.0002, "epoch": 1.9788214372322366, "step": 12240}, {"loss": 0.7533, "grad_norm": 0.5998196005821228, "learning_rate": 0.0002, "epoch": 1.9804381214129818, "step": 12250}, {"loss": 0.7225, "grad_norm": 0.6960851550102234, "learning_rate": 0.0002, "epoch": 1.9820548055937273, "step": 12260}, {"loss": 0.7398, "grad_norm": 0.7674502730369568, "learning_rate": 0.0002, "epoch": 1.9836714897744727, "step": 12270}, {"loss": 0.7185, "grad_norm": 0.6407275795936584, "learning_rate": 0.0002, "epoch": 1.985288173955218, "step": 12280}, {"loss": 0.7382, "grad_norm": 0.6673079133033752, "learning_rate": 0.0002, "epoch": 1.9869048581359632, "step": 12290}, {"loss": 0.7326, "grad_norm": 0.6989844441413879, "learning_rate": 0.0002, "epoch": 1.9885215423167084, "step": 12300}, {"loss": 0.7559, "grad_norm": 0.7564442157745361, "learning_rate": 0.0002, "epoch": 1.9901382264974536, "step": 12310}, {"loss": 0.7719, "grad_norm": 0.6385478973388672, "learning_rate": 0.0002, "epoch": 1.9917549106781989, "step": 12320}, {"loss": 0.7369, "grad_norm": 0.7193717956542969, "learning_rate": 0.0002, "epoch": 1.9933715948589443, "step": 12330}, {"loss": 0.7583, "grad_norm": 0.7987112402915955, "learning_rate": 0.0002, "epoch": 1.9949882790396896, "step": 12340}, {"loss": 0.7793, "grad_norm": 0.7260826826095581, "learning_rate": 0.0002, "epoch": 1.996604963220435, "step": 12350}, {"loss": 0.7505, "grad_norm": 0.7968255281448364, "learning_rate": 0.0002, "epoch": 1.9982216474011802, "step": 12360}, {"loss": 0.717, "grad_norm": 0.6893062591552734, "learning_rate": 0.0002, "epoch": 1.9998383315819255, "step": 12370}, {"eval_loss": 1.1044032573699951, "eval_runtime": 122.1508, "eval_samples_per_second": 6.001, "eval_steps_per_second": 0.753, "epoch": 2.0, "step": 12371}, {"loss": 0.6604, "grad_norm": 0.7775409817695618, "learning_rate": 0.0002, "epoch": 2.0014550157626707, "step": 12380}, {"loss": 0.6845, "grad_norm": 0.76218581199646, "learning_rate": 0.0002, "epoch": 2.003071699943416, "step": 12390}, {"loss": 0.6909, "grad_norm": 0.5677764415740967, "learning_rate": 0.0002, "epoch": 2.004688384124161, "step": 12400}, {"loss": 0.6584, "grad_norm": 0.808442234992981, "learning_rate": 0.0002, "epoch": 2.006305068304907, "step": 12410}, {"loss": 0.659, "grad_norm": 0.7144765257835388, "learning_rate": 0.0002, "epoch": 2.007921752485652, "step": 12420}, {"loss": 0.6666, "grad_norm": 0.6914031505584717, "learning_rate": 0.0002, "epoch": 2.0095384366663973, "step": 12430}, {"loss": 0.6596, "grad_norm": 0.7581454515457153, "learning_rate": 0.0002, "epoch": 2.0111551208471425, "step": 12440}, {"loss": 0.6785, "grad_norm": 0.8388504981994629, "learning_rate": 0.0002, "epoch": 2.0127718050278878, "step": 12450}, {"loss": 0.6942, "grad_norm": 0.6716406941413879, "learning_rate": 0.0002, "epoch": 2.014388489208633, "step": 12460}, {"loss": 0.6441, "grad_norm": 0.898902416229248, "learning_rate": 0.0002, "epoch": 2.0160051733893782, "step": 12470}, {"loss": 0.6655, "grad_norm": 0.6432679891586304, "learning_rate": 0.0002, "epoch": 2.0176218575701235, "step": 12480}, {"loss": 0.6521, "grad_norm": 0.8021109104156494, "learning_rate": 0.0002, "epoch": 2.019238541750869, "step": 12490}, {"loss": 0.6581, "grad_norm": 0.7039216756820679, "learning_rate": 0.0002, "epoch": 2.0208552259316144, "step": 12500}, {"loss": 0.6521, "grad_norm": 0.646531879901886, "learning_rate": 0.0002, "epoch": 2.0224719101123596, "step": 12510}, {"loss": 0.6302, "grad_norm": 0.783704400062561, "learning_rate": 0.0002, "epoch": 2.024088594293105, "step": 12520}, {"loss": 0.6288, "grad_norm": 0.8805046677589417, "learning_rate": 0.0002, "epoch": 2.02570527847385, "step": 12530}, {"loss": 0.6288, "grad_norm": 0.7289270758628845, "learning_rate": 0.0002, "epoch": 2.0273219626545953, "step": 12540}, {"loss": 0.6663, "grad_norm": 0.71653151512146, "learning_rate": 0.0002, "epoch": 2.0289386468353405, "step": 12550}, {"loss": 0.625, "grad_norm": 0.73281329870224, "learning_rate": 0.0002, "epoch": 2.030555331016086, "step": 12560}, {"loss": 0.6448, "grad_norm": 0.6657090187072754, "learning_rate": 0.0002, "epoch": 2.0321720151968314, "step": 12570}, {"loss": 0.6983, "grad_norm": 0.8241133093833923, "learning_rate": 0.0002, "epoch": 2.0337886993775767, "step": 12580}, {"loss": 0.6488, "grad_norm": 0.5834135413169861, "learning_rate": 0.0002, "epoch": 2.035405383558322, "step": 12590}, {"loss": 0.6188, "grad_norm": 0.84502112865448, "learning_rate": 0.0002, "epoch": 2.037022067739067, "step": 12600}, {"loss": 0.6349, "grad_norm": 0.8952481746673584, "learning_rate": 0.0002, "epoch": 2.0386387519198124, "step": 12610}, {"loss": 0.6923, "grad_norm": 0.7801461815834045, "learning_rate": 0.0002, "epoch": 2.0402554361005576, "step": 12620}, {"loss": 0.6176, "grad_norm": 0.6788367033004761, "learning_rate": 0.0002, "epoch": 2.041872120281303, "step": 12630}, {"loss": 0.6162, "grad_norm": 0.7241756319999695, "learning_rate": 0.0002, "epoch": 2.0434888044620485, "step": 12640}, {"loss": 0.655, "grad_norm": 0.6933388113975525, "learning_rate": 0.0002, "epoch": 2.0451054886427937, "step": 12650}, {"loss": 0.6431, "grad_norm": 0.8029746413230896, "learning_rate": 0.0002, "epoch": 2.046722172823539, "step": 12660}, {"loss": 0.7164, "grad_norm": 0.946399986743927, "learning_rate": 0.0002, "epoch": 2.048338857004284, "step": 12670}, {"loss": 0.638, "grad_norm": 0.7072678804397583, "learning_rate": 0.0002, "epoch": 2.0499555411850294, "step": 12680}, {"loss": 0.6487, "grad_norm": 0.6810618042945862, "learning_rate": 0.0002, "epoch": 2.0515722253657747, "step": 12690}, {"loss": 0.6554, "grad_norm": 0.7661160230636597, "learning_rate": 0.0002, "epoch": 2.05318890954652, "step": 12700}, {"loss": 0.6799, "grad_norm": 0.6350653767585754, "learning_rate": 0.0002, "epoch": 2.0548055937272656, "step": 12710}, {"loss": 0.6654, "grad_norm": 0.861890971660614, "learning_rate": 0.0002, "epoch": 2.056422277908011, "step": 12720}, {"loss": 0.6286, "grad_norm": 0.6489875912666321, "learning_rate": 0.0002, "epoch": 2.058038962088756, "step": 12730}, {"loss": 0.6811, "grad_norm": 0.8268506526947021, "learning_rate": 0.0002, "epoch": 2.0596556462695013, "step": 12740}, {"loss": 0.6524, "grad_norm": 0.607679545879364, "learning_rate": 0.0002, "epoch": 2.0612723304502465, "step": 12750}, {"loss": 0.6649, "grad_norm": 0.6754153370857239, "learning_rate": 0.0002, "epoch": 2.0628890146309917, "step": 12760}, {"loss": 0.6549, "grad_norm": 0.7263124585151672, "learning_rate": 0.0002, "epoch": 2.064505698811737, "step": 12770}, {"loss": 0.6189, "grad_norm": 0.6986154317855835, "learning_rate": 0.0002, "epoch": 2.0661223829924826, "step": 12780}, {"loss": 0.6723, "grad_norm": 0.7768576741218567, "learning_rate": 0.0002, "epoch": 2.067739067173228, "step": 12790}, {"loss": 0.677, "grad_norm": 0.7546762824058533, "learning_rate": 0.0002, "epoch": 2.069355751353973, "step": 12800}, {"loss": 0.6485, "grad_norm": 0.7588880062103271, "learning_rate": 0.0002, "epoch": 2.0709724355347183, "step": 12810}, {"loss": 0.6989, "grad_norm": 0.7457242608070374, "learning_rate": 0.0002, "epoch": 2.0725891197154636, "step": 12820}, {"loss": 0.6489, "grad_norm": 0.6983516812324524, "learning_rate": 0.0002, "epoch": 2.074205803896209, "step": 12830}, {"loss": 0.651, "grad_norm": 0.7950928807258606, "learning_rate": 0.0002, "epoch": 2.075822488076954, "step": 12840}, {"loss": 0.6603, "grad_norm": 0.9248087406158447, "learning_rate": 0.0002, "epoch": 2.0774391722576993, "step": 12850}, {"loss": 0.6847, "grad_norm": 0.7229493260383606, "learning_rate": 0.0002, "epoch": 2.079055856438445, "step": 12860}, {"loss": 0.6702, "grad_norm": 0.5710847973823547, "learning_rate": 0.0002, "epoch": 2.08067254061919, "step": 12870}, {"loss": 0.6974, "grad_norm": 0.9580423831939697, "learning_rate": 0.0002, "epoch": 2.0822892247999354, "step": 12880}, {"loss": 0.6341, "grad_norm": 0.7399665713310242, "learning_rate": 0.0002, "epoch": 2.0839059089806806, "step": 12890}, {"loss": 0.6993, "grad_norm": 0.7981410622596741, "learning_rate": 0.0002, "epoch": 2.085522593161426, "step": 12900}, {"loss": 0.6976, "grad_norm": 0.870759904384613, "learning_rate": 0.0002, "epoch": 2.087139277342171, "step": 12910}, {"loss": 0.7194, "grad_norm": 0.7001481652259827, "learning_rate": 0.0002, "epoch": 2.0887559615229163, "step": 12920}, {"loss": 0.6383, "grad_norm": 0.6745418310165405, "learning_rate": 0.0002, "epoch": 2.090372645703662, "step": 12930}, {"loss": 0.6519, "grad_norm": 0.7739067673683167, "learning_rate": 0.0002, "epoch": 2.0919893298844072, "step": 12940}, {"loss": 0.6856, "grad_norm": 0.6742934584617615, "learning_rate": 0.0002, "epoch": 2.0936060140651525, "step": 12950}, {"loss": 0.6279, "grad_norm": 0.7270349860191345, "learning_rate": 0.0002, "epoch": 2.0952226982458977, "step": 12960}, {"loss": 0.6783, "grad_norm": 0.7150624394416809, "learning_rate": 0.0002, "epoch": 2.096839382426643, "step": 12970}, {"loss": 0.6093, "grad_norm": 0.7734767198562622, "learning_rate": 0.0002, "epoch": 2.098456066607388, "step": 12980}, {"loss": 0.6534, "grad_norm": 0.7618662118911743, "learning_rate": 0.0002, "epoch": 2.1000727507881334, "step": 12990}, {"loss": 0.6707, "grad_norm": 0.6557944416999817, "learning_rate": 0.0002, "epoch": 2.101689434968879, "step": 13000}, {"loss": 0.7268, "grad_norm": 0.8786448240280151, "learning_rate": 0.0002, "epoch": 2.1033061191496243, "step": 13010}, {"loss": 0.6677, "grad_norm": 0.6878724098205566, "learning_rate": 0.0002, "epoch": 2.1049228033303695, "step": 13020}, {"loss": 0.6824, "grad_norm": 0.822318971157074, "learning_rate": 0.0002, "epoch": 2.1065394875111147, "step": 13030}, {"loss": 0.6228, "grad_norm": 0.831468939781189, "learning_rate": 0.0002, "epoch": 2.10815617169186, "step": 13040}, {"loss": 0.6511, "grad_norm": 0.7699505686759949, "learning_rate": 0.0002, "epoch": 2.109772855872605, "step": 13050}, {"loss": 0.6671, "grad_norm": 0.7559016346931458, "learning_rate": 0.0002, "epoch": 2.1113895400533504, "step": 13060}, {"loss": 0.6215, "grad_norm": 0.6942209601402283, "learning_rate": 0.0002, "epoch": 2.1130062242340957, "step": 13070}, {"loss": 0.6449, "grad_norm": 0.6098947525024414, "learning_rate": 0.0002, "epoch": 2.1146229084148414, "step": 13080}, {"loss": 0.7091, "grad_norm": 0.6499016284942627, "learning_rate": 0.0002, "epoch": 2.1162395925955866, "step": 13090}, {"loss": 0.6247, "grad_norm": 0.7719953060150146, "learning_rate": 0.0002, "epoch": 2.117856276776332, "step": 13100}, {"loss": 0.6064, "grad_norm": 0.6708134412765503, "learning_rate": 0.0002, "epoch": 2.119472960957077, "step": 13110}, {"loss": 0.6056, "grad_norm": 0.8119585514068604, "learning_rate": 0.0002, "epoch": 2.1210896451378223, "step": 13120}, {"loss": 0.6628, "grad_norm": 0.6947157979011536, "learning_rate": 0.0002, "epoch": 2.1227063293185675, "step": 13130}, {"loss": 0.6375, "grad_norm": 0.8831837773323059, "learning_rate": 0.0002, "epoch": 2.1243230134993127, "step": 13140}, {"loss": 0.6997, "grad_norm": 0.7266910672187805, "learning_rate": 0.0002, "epoch": 2.1259396976800584, "step": 13150}, {"loss": 0.6446, "grad_norm": 0.8864351511001587, "learning_rate": 0.0002, "epoch": 2.1275563818608036, "step": 13160}, {"loss": 0.6762, "grad_norm": 0.8104248046875, "learning_rate": 0.0002, "epoch": 2.129173066041549, "step": 13170}, {"loss": 0.6581, "grad_norm": 0.6077079772949219, "learning_rate": 0.0002, "epoch": 2.130789750222294, "step": 13180}, {"loss": 0.6572, "grad_norm": 0.6874213814735413, "learning_rate": 0.0002, "epoch": 2.1324064344030393, "step": 13190}, {"loss": 0.642, "grad_norm": 0.7134367823600769, "learning_rate": 0.0002, "epoch": 2.1340231185837846, "step": 13200}, {"loss": 0.7016, "grad_norm": 0.6101235151290894, "learning_rate": 0.0002, "epoch": 2.13563980276453, "step": 13210}, {"loss": 0.6529, "grad_norm": 0.6042411923408508, "learning_rate": 0.0002, "epoch": 2.137256486945275, "step": 13220}, {"loss": 0.7179, "grad_norm": 0.914601743221283, "learning_rate": 0.0002, "epoch": 2.1388731711260207, "step": 13230}, {"loss": 0.6513, "grad_norm": 0.7104284167289734, "learning_rate": 0.0002, "epoch": 2.140489855306766, "step": 13240}, {"loss": 0.6607, "grad_norm": 0.664395272731781, "learning_rate": 0.0002, "epoch": 2.142106539487511, "step": 13250}, {"loss": 0.7211, "grad_norm": 0.6991241574287415, "learning_rate": 0.0002, "epoch": 2.1437232236682564, "step": 13260}, {"loss": 0.6484, "grad_norm": 0.5469560623168945, "learning_rate": 0.0002, "epoch": 2.1453399078490016, "step": 13270}, {"loss": 0.6765, "grad_norm": 0.8454998135566711, "learning_rate": 0.0002, "epoch": 2.146956592029747, "step": 13280}, {"loss": 0.6683, "grad_norm": 0.7088868618011475, "learning_rate": 0.0002, "epoch": 2.148573276210492, "step": 13290}, {"loss": 0.6835, "grad_norm": 0.7002687454223633, "learning_rate": 0.0002, "epoch": 2.1501899603912378, "step": 13300}, {"loss": 0.6399, "grad_norm": 0.7785214781761169, "learning_rate": 0.0002, "epoch": 2.151806644571983, "step": 13310}, {"loss": 0.67, "grad_norm": 0.8049132227897644, "learning_rate": 0.0002, "epoch": 2.1534233287527282, "step": 13320}, {"loss": 0.6495, "grad_norm": 0.8062595129013062, "learning_rate": 0.0002, "epoch": 2.1550400129334735, "step": 13330}, {"loss": 0.6603, "grad_norm": 0.6208319067955017, "learning_rate": 0.0002, "epoch": 2.1566566971142187, "step": 13340}, {"loss": 0.6584, "grad_norm": 0.7519655823707581, "learning_rate": 0.0002, "epoch": 2.158273381294964, "step": 13350}, {"loss": 0.6457, "grad_norm": 0.7645747065544128, "learning_rate": 0.0002, "epoch": 2.159890065475709, "step": 13360}, {"loss": 0.645, "grad_norm": 0.6847302913665771, "learning_rate": 0.0002, "epoch": 2.1615067496564544, "step": 13370}, {"loss": 0.6903, "grad_norm": 0.8630441427230835, "learning_rate": 0.0002, "epoch": 2.1631234338372, "step": 13380}, {"loss": 0.6742, "grad_norm": 0.7947702407836914, "learning_rate": 0.0002, "epoch": 2.1647401180179453, "step": 13390}, {"loss": 0.7206, "grad_norm": 0.6836977005004883, "learning_rate": 0.0002, "epoch": 2.1663568021986905, "step": 13400}, {"loss": 0.6304, "grad_norm": 0.7340566515922546, "learning_rate": 0.0002, "epoch": 2.1679734863794358, "step": 13410}, {"loss": 0.6528, "grad_norm": 0.7075738906860352, "learning_rate": 0.0002, "epoch": 2.169590170560181, "step": 13420}, {"loss": 0.6585, "grad_norm": 0.7080879807472229, "learning_rate": 0.0002, "epoch": 2.1712068547409262, "step": 13430}, {"loss": 0.6615, "grad_norm": 0.6218613386154175, "learning_rate": 0.0002, "epoch": 2.1728235389216715, "step": 13440}, {"loss": 0.6488, "grad_norm": 0.8211479187011719, "learning_rate": 0.0002, "epoch": 2.174440223102417, "step": 13450}, {"loss": 0.6738, "grad_norm": 0.864466667175293, "learning_rate": 0.0002, "epoch": 2.1760569072831624, "step": 13460}, {"loss": 0.679, "grad_norm": 0.7943857908248901, "learning_rate": 0.0002, "epoch": 2.1776735914639076, "step": 13470}, {"loss": 0.6838, "grad_norm": 0.78728187084198, "learning_rate": 0.0002, "epoch": 2.179290275644653, "step": 13480}, {"loss": 0.6397, "grad_norm": 0.697527289390564, "learning_rate": 0.0002, "epoch": 2.180906959825398, "step": 13490}, {"loss": 0.669, "grad_norm": 0.8205804228782654, "learning_rate": 0.0002, "epoch": 2.1825236440061433, "step": 13500}, {"loss": 0.7227, "grad_norm": 0.8709042072296143, "learning_rate": 0.0002, "epoch": 2.1841403281868885, "step": 13510}, {"loss": 0.6313, "grad_norm": 0.6228537559509277, "learning_rate": 0.0002, "epoch": 2.1857570123676338, "step": 13520}, {"loss": 0.7025, "grad_norm": 0.9566980004310608, "learning_rate": 0.0002, "epoch": 2.1873736965483794, "step": 13530}, {"loss": 0.6755, "grad_norm": 0.7128894329071045, "learning_rate": 0.0002, "epoch": 2.1889903807291247, "step": 13540}, {"loss": 0.6827, "grad_norm": 0.6888654232025146, "learning_rate": 0.0002, "epoch": 2.19060706490987, "step": 13550}, {"loss": 0.6961, "grad_norm": 0.6444337368011475, "learning_rate": 0.0002, "epoch": 2.192223749090615, "step": 13560}, {"loss": 0.656, "grad_norm": 0.8008806705474854, "learning_rate": 0.0002, "epoch": 2.1938404332713604, "step": 13570}, {"loss": 0.7, "grad_norm": 0.8482748866081238, "learning_rate": 0.0002, "epoch": 2.1954571174521056, "step": 13580}, {"loss": 0.7326, "grad_norm": 0.8584157228469849, "learning_rate": 0.0002, "epoch": 2.197073801632851, "step": 13590}, {"loss": 0.7014, "grad_norm": 0.7513734698295593, "learning_rate": 0.0002, "epoch": 2.1986904858135965, "step": 13600}, {"loss": 0.6632, "grad_norm": 0.7864262461662292, "learning_rate": 0.0002, "epoch": 2.2003071699943417, "step": 13610}, {"loss": 0.6879, "grad_norm": 0.8493645191192627, "learning_rate": 0.0002, "epoch": 2.201923854175087, "step": 13620}, {"loss": 0.6617, "grad_norm": 0.6902140974998474, "learning_rate": 0.0002, "epoch": 2.203540538355832, "step": 13630}, {"loss": 0.6655, "grad_norm": 0.8711254596710205, "learning_rate": 0.0002, "epoch": 2.2051572225365774, "step": 13640}, {"loss": 0.6359, "grad_norm": 0.7832191586494446, "learning_rate": 0.0002, "epoch": 2.2067739067173227, "step": 13650}, {"loss": 0.6723, "grad_norm": 0.5668176412582397, "learning_rate": 0.0002, "epoch": 2.208390590898068, "step": 13660}, {"loss": 0.635, "grad_norm": 0.8648375272750854, "learning_rate": 0.0002, "epoch": 2.2100072750788136, "step": 13670}, {"loss": 0.653, "grad_norm": 0.7643089890480042, "learning_rate": 0.0002, "epoch": 2.211623959259559, "step": 13680}, {"loss": 0.6765, "grad_norm": 0.6293777823448181, "learning_rate": 0.0002, "epoch": 2.213240643440304, "step": 13690}, {"loss": 0.6842, "grad_norm": 0.6459372639656067, "learning_rate": 0.0002, "epoch": 2.2148573276210493, "step": 13700}, {"loss": 0.6526, "grad_norm": 0.7060744166374207, "learning_rate": 0.0002, "epoch": 2.2164740118017945, "step": 13710}, {"loss": 0.7101, "grad_norm": 0.674109160900116, "learning_rate": 0.0002, "epoch": 2.2180906959825397, "step": 13720}, {"loss": 0.6529, "grad_norm": 0.830392062664032, "learning_rate": 0.0002, "epoch": 2.219707380163285, "step": 13730}, {"loss": 0.6733, "grad_norm": 0.6474477052688599, "learning_rate": 0.0002, "epoch": 2.2213240643440306, "step": 13740}, {"loss": 0.6413, "grad_norm": 0.7037909626960754, "learning_rate": 0.0002, "epoch": 2.222940748524776, "step": 13750}, {"loss": 0.6417, "grad_norm": 0.6554131507873535, "learning_rate": 0.0002, "epoch": 2.224557432705521, "step": 13760}, {"loss": 0.6907, "grad_norm": 0.7822230458259583, "learning_rate": 0.0002, "epoch": 2.2261741168862663, "step": 13770}, {"loss": 0.6505, "grad_norm": 0.9082167744636536, "learning_rate": 0.0002, "epoch": 2.2277908010670116, "step": 13780}, {"loss": 0.6878, "grad_norm": 0.7918276190757751, "learning_rate": 0.0002, "epoch": 2.229407485247757, "step": 13790}, {"loss": 0.6669, "grad_norm": 0.7354569435119629, "learning_rate": 0.0002, "epoch": 2.231024169428502, "step": 13800}, {"loss": 0.6503, "grad_norm": 0.8265249133110046, "learning_rate": 0.0002, "epoch": 2.2326408536092472, "step": 13810}, {"loss": 0.6871, "grad_norm": 0.6653847098350525, "learning_rate": 0.0002, "epoch": 2.234257537789993, "step": 13820}, {"loss": 0.6413, "grad_norm": 0.7157923579216003, "learning_rate": 0.0002, "epoch": 2.235874221970738, "step": 13830}, {"loss": 0.6306, "grad_norm": 0.7110323309898376, "learning_rate": 0.0002, "epoch": 2.2374909061514834, "step": 13840}, {"loss": 0.6913, "grad_norm": 0.7155357599258423, "learning_rate": 0.0002, "epoch": 2.2391075903322286, "step": 13850}, {"loss": 0.6579, "grad_norm": 1.0177817344665527, "learning_rate": 0.0002, "epoch": 2.240724274512974, "step": 13860}, {"loss": 0.635, "grad_norm": 0.7601948380470276, "learning_rate": 0.0002, "epoch": 2.242340958693719, "step": 13870}, {"loss": 0.6679, "grad_norm": 0.7628820538520813, "learning_rate": 0.0002, "epoch": 2.2439576428744643, "step": 13880}, {"loss": 0.6805, "grad_norm": 0.7089297771453857, "learning_rate": 0.0002, "epoch": 2.24557432705521, "step": 13890}, {"loss": 0.7236, "grad_norm": 0.695178210735321, "learning_rate": 0.0002, "epoch": 2.247191011235955, "step": 13900}, {"loss": 0.7084, "grad_norm": 0.7631948590278625, "learning_rate": 0.0002, "epoch": 2.2488076954167004, "step": 13910}, {"loss": 0.685, "grad_norm": 0.8203101754188538, "learning_rate": 0.0002, "epoch": 2.2504243795974457, "step": 13920}, {"loss": 0.653, "grad_norm": 0.8099079728126526, "learning_rate": 0.0002, "epoch": 2.252041063778191, "step": 13930}, {"loss": 0.694, "grad_norm": 0.6498546004295349, "learning_rate": 0.0002, "epoch": 2.253657747958936, "step": 13940}, {"loss": 0.6684, "grad_norm": 0.7797415256500244, "learning_rate": 0.0002, "epoch": 2.2552744321396814, "step": 13950}, {"loss": 0.683, "grad_norm": 0.8254124522209167, "learning_rate": 0.0002, "epoch": 2.2568911163204266, "step": 13960}, {"loss": 0.6806, "grad_norm": 0.6327953338623047, "learning_rate": 0.0002, "epoch": 2.2585078005011723, "step": 13970}, {"loss": 0.668, "grad_norm": 0.734194278717041, "learning_rate": 0.0002, "epoch": 2.2601244846819175, "step": 13980}, {"loss": 0.6912, "grad_norm": 0.9014202952384949, "learning_rate": 0.0002, "epoch": 2.2617411688626627, "step": 13990}, {"loss": 0.692, "grad_norm": 0.7643631100654602, "learning_rate": 0.0002, "epoch": 2.263357853043408, "step": 14000}, {"loss": 0.6657, "grad_norm": 0.8882834911346436, "learning_rate": 0.0002, "epoch": 2.264974537224153, "step": 14010}, {"loss": 0.6453, "grad_norm": 0.7975873351097107, "learning_rate": 0.0002, "epoch": 2.2665912214048984, "step": 14020}, {"loss": 0.7193, "grad_norm": 0.7765783071517944, "learning_rate": 0.0002, "epoch": 2.2682079055856437, "step": 14030}, {"loss": 0.662, "grad_norm": 0.8846288323402405, "learning_rate": 0.0002, "epoch": 2.2698245897663893, "step": 14040}, {"loss": 0.6494, "grad_norm": 0.9006744027137756, "learning_rate": 0.0002, "epoch": 2.2714412739471346, "step": 14050}, {"loss": 0.6423, "grad_norm": 0.7420173287391663, "learning_rate": 0.0002, "epoch": 2.27305795812788, "step": 14060}, {"loss": 0.7068, "grad_norm": 0.7956424951553345, "learning_rate": 0.0002, "epoch": 2.274674642308625, "step": 14070}, {"loss": 0.6581, "grad_norm": 0.7783209085464478, "learning_rate": 0.0002, "epoch": 2.2762913264893703, "step": 14080}, {"loss": 0.7202, "grad_norm": 0.7597188949584961, "learning_rate": 0.0002, "epoch": 2.2779080106701155, "step": 14090}, {"loss": 0.6778, "grad_norm": 0.6718921661376953, "learning_rate": 0.0002, "epoch": 2.2795246948508607, "step": 14100}, {"loss": 0.632, "grad_norm": 0.7528082132339478, "learning_rate": 0.0002, "epoch": 2.281141379031606, "step": 14110}, {"loss": 0.7608, "grad_norm": 0.8379864692687988, "learning_rate": 0.0002, "epoch": 2.2827580632123516, "step": 14120}, {"loss": 0.6767, "grad_norm": 0.748613715171814, "learning_rate": 0.0002, "epoch": 2.284374747393097, "step": 14130}, {"loss": 0.6641, "grad_norm": 0.7435423135757446, "learning_rate": 0.0002, "epoch": 2.285991431573842, "step": 14140}, {"loss": 0.6849, "grad_norm": 0.7580803632736206, "learning_rate": 0.0002, "epoch": 2.2876081157545873, "step": 14150}, {"loss": 0.6604, "grad_norm": 0.6278321146965027, "learning_rate": 0.0002, "epoch": 2.2892247999353326, "step": 14160}, {"loss": 0.6573, "grad_norm": 0.7663896083831787, "learning_rate": 0.0002, "epoch": 2.290841484116078, "step": 14170}, {"loss": 0.6655, "grad_norm": 0.9716812372207642, "learning_rate": 0.0002, "epoch": 2.292458168296823, "step": 14180}, {"loss": 0.7067, "grad_norm": 0.8993458151817322, "learning_rate": 0.0002, "epoch": 2.2940748524775687, "step": 14190}, {"loss": 0.6172, "grad_norm": 0.6156117916107178, "learning_rate": 0.0002, "epoch": 2.295691536658314, "step": 14200}, {"loss": 0.6318, "grad_norm": 0.8911278247833252, "learning_rate": 0.0002, "epoch": 2.297308220839059, "step": 14210}, {"loss": 0.6364, "grad_norm": 0.6422147154808044, "learning_rate": 0.0002, "epoch": 2.2989249050198044, "step": 14220}, {"loss": 0.6795, "grad_norm": 0.6866879463195801, "learning_rate": 0.0002, "epoch": 2.3005415892005496, "step": 14230}, {"loss": 0.6907, "grad_norm": 0.9297130107879639, "learning_rate": 0.0002, "epoch": 2.302158273381295, "step": 14240}, {"loss": 0.6823, "grad_norm": 0.7501356601715088, "learning_rate": 0.0002, "epoch": 2.30377495756204, "step": 14250}, {"loss": 0.6414, "grad_norm": 0.8363515138626099, "learning_rate": 0.0002, "epoch": 2.3053916417427853, "step": 14260}, {"loss": 0.6362, "grad_norm": 0.9083868265151978, "learning_rate": 0.0002, "epoch": 2.307008325923531, "step": 14270}, {"loss": 0.6862, "grad_norm": 0.7791516780853271, "learning_rate": 0.0002, "epoch": 2.3086250101042762, "step": 14280}, {"loss": 0.6569, "grad_norm": 0.8766953349113464, "learning_rate": 0.0002, "epoch": 2.3102416942850215, "step": 14290}, {"loss": 0.6698, "grad_norm": 0.7916635274887085, "learning_rate": 0.0002, "epoch": 2.3118583784657667, "step": 14300}, {"loss": 0.6927, "grad_norm": 0.627525269985199, "learning_rate": 0.0002, "epoch": 2.313475062646512, "step": 14310}, {"loss": 0.6541, "grad_norm": 0.8856783509254456, "learning_rate": 0.0002, "epoch": 2.315091746827257, "step": 14320}, {"loss": 0.6806, "grad_norm": 0.6758689284324646, "learning_rate": 0.0002, "epoch": 2.316708431008003, "step": 14330}, {"loss": 0.6794, "grad_norm": 0.6428321003913879, "learning_rate": 0.0002, "epoch": 2.318325115188748, "step": 14340}, {"loss": 0.682, "grad_norm": 0.9032121300697327, "learning_rate": 0.0002, "epoch": 2.3199417993694933, "step": 14350}, {"loss": 0.6569, "grad_norm": 0.8035986423492432, "learning_rate": 0.0002, "epoch": 2.3215584835502385, "step": 14360}, {"loss": 0.7067, "grad_norm": 0.7974579334259033, "learning_rate": 0.0002, "epoch": 2.3231751677309838, "step": 14370}, {"loss": 0.6451, "grad_norm": 0.8356034755706787, "learning_rate": 0.0002, "epoch": 2.324791851911729, "step": 14380}, {"loss": 0.6623, "grad_norm": 0.998760998249054, "learning_rate": 0.0002, "epoch": 2.326408536092474, "step": 14390}, {"loss": 0.649, "grad_norm": 0.6518142223358154, "learning_rate": 0.0002, "epoch": 2.3280252202732195, "step": 14400}, {"loss": 0.7146, "grad_norm": 0.7443506717681885, "learning_rate": 0.0002, "epoch": 2.3296419044539647, "step": 14410}, {"loss": 0.648, "grad_norm": 0.8436172604560852, "learning_rate": 0.0002, "epoch": 2.3312585886347104, "step": 14420}, {"loss": 0.6585, "grad_norm": 0.7411080598831177, "learning_rate": 0.0002, "epoch": 2.3328752728154556, "step": 14430}, {"loss": 0.6781, "grad_norm": 0.8839048743247986, "learning_rate": 0.0002, "epoch": 2.334491956996201, "step": 14440}, {"loss": 0.6565, "grad_norm": 0.8360885977745056, "learning_rate": 0.0002, "epoch": 2.336108641176946, "step": 14450}, {"loss": 0.6662, "grad_norm": 0.7608986496925354, "learning_rate": 0.0002, "epoch": 2.3377253253576913, "step": 14460}, {"loss": 0.6685, "grad_norm": 0.8179867267608643, "learning_rate": 0.0002, "epoch": 2.3393420095384365, "step": 14470}, {"loss": 0.7055, "grad_norm": 0.5989999771118164, "learning_rate": 0.0002, "epoch": 2.340958693719182, "step": 14480}, {"loss": 0.644, "grad_norm": 0.9450054168701172, "learning_rate": 0.0002, "epoch": 2.3425753778999274, "step": 14490}, {"loss": 0.6983, "grad_norm": 0.7885149717330933, "learning_rate": 0.0002, "epoch": 2.3441920620806727, "step": 14500}, {"loss": 0.6819, "grad_norm": 0.8152616620063782, "learning_rate": 0.0002, "epoch": 2.345808746261418, "step": 14510}, {"loss": 0.6989, "grad_norm": 0.7193838953971863, "learning_rate": 0.0002, "epoch": 2.347425430442163, "step": 14520}, {"loss": 0.6594, "grad_norm": 0.6701092720031738, "learning_rate": 0.0002, "epoch": 2.3490421146229084, "step": 14530}, {"loss": 0.6559, "grad_norm": 0.7529364228248596, "learning_rate": 0.0002, "epoch": 2.3506587988036536, "step": 14540}, {"loss": 0.6306, "grad_norm": 0.6599733829498291, "learning_rate": 0.0002, "epoch": 2.352275482984399, "step": 14550}, {"loss": 0.706, "grad_norm": 0.9502474069595337, "learning_rate": 0.0002, "epoch": 2.353892167165144, "step": 14560}, {"loss": 0.717, "grad_norm": 0.7619650959968567, "learning_rate": 0.0002, "epoch": 2.3555088513458897, "step": 14570}, {"loss": 0.6684, "grad_norm": 0.9854652285575867, "learning_rate": 0.0002, "epoch": 2.357125535526635, "step": 14580}, {"loss": 0.6455, "grad_norm": 0.727439284324646, "learning_rate": 0.0002, "epoch": 2.35874221970738, "step": 14590}, {"loss": 0.6645, "grad_norm": 0.6994746327400208, "learning_rate": 0.0002, "epoch": 2.3603589038881254, "step": 14600}, {"loss": 0.6587, "grad_norm": 0.7117531299591064, "learning_rate": 0.0002, "epoch": 2.3619755880688706, "step": 14610}, {"loss": 0.6804, "grad_norm": 0.6403067708015442, "learning_rate": 0.0002, "epoch": 2.363592272249616, "step": 14620}, {"loss": 0.7055, "grad_norm": 0.8377841711044312, "learning_rate": 0.0002, "epoch": 2.3652089564303616, "step": 14630}, {"loss": 0.6778, "grad_norm": 0.749171257019043, "learning_rate": 0.0002, "epoch": 2.366825640611107, "step": 14640}, {"loss": 0.6552, "grad_norm": 0.8418586254119873, "learning_rate": 0.0002, "epoch": 2.368442324791852, "step": 14650}, {"loss": 0.6685, "grad_norm": 0.6178573369979858, "learning_rate": 0.0002, "epoch": 2.3700590089725972, "step": 14660}, {"loss": 0.6774, "grad_norm": 0.6368302702903748, "learning_rate": 0.0002, "epoch": 2.3716756931533425, "step": 14670}, {"loss": 0.6136, "grad_norm": 0.9122977256774902, "learning_rate": 0.0002, "epoch": 2.3732923773340877, "step": 14680}, {"loss": 0.6675, "grad_norm": 0.7086195349693298, "learning_rate": 0.0002, "epoch": 2.374909061514833, "step": 14690}, {"loss": 0.6582, "grad_norm": 0.7500800490379333, "learning_rate": 0.0002, "epoch": 2.376525745695578, "step": 14700}, {"loss": 0.6792, "grad_norm": 0.6634900569915771, "learning_rate": 0.0002, "epoch": 2.378142429876324, "step": 14710}, {"loss": 0.6614, "grad_norm": 0.839898407459259, "learning_rate": 0.0002, "epoch": 2.379759114057069, "step": 14720}, {"loss": 0.6453, "grad_norm": 0.7578426003456116, "learning_rate": 0.0002, "epoch": 2.3813757982378143, "step": 14730}, {"loss": 0.7282, "grad_norm": 1.0213173627853394, "learning_rate": 0.0002, "epoch": 2.3829924824185595, "step": 14740}, {"loss": 0.6704, "grad_norm": 0.7855949401855469, "learning_rate": 0.0002, "epoch": 2.3846091665993048, "step": 14750}, {"loss": 0.6694, "grad_norm": 0.7224128842353821, "learning_rate": 0.0002, "epoch": 2.38622585078005, "step": 14760}, {"loss": 0.7017, "grad_norm": 0.8040381669998169, "learning_rate": 0.0002, "epoch": 2.3878425349607952, "step": 14770}, {"loss": 0.6799, "grad_norm": 0.7705281376838684, "learning_rate": 0.0002, "epoch": 2.389459219141541, "step": 14780}, {"loss": 0.6326, "grad_norm": 0.667966902256012, "learning_rate": 0.0002, "epoch": 2.391075903322286, "step": 14790}, {"loss": 0.7061, "grad_norm": 0.6611011028289795, "learning_rate": 0.0002, "epoch": 2.3926925875030314, "step": 14800}, {"loss": 0.6527, "grad_norm": 0.6862651705741882, "learning_rate": 0.0002, "epoch": 2.3943092716837766, "step": 14810}, {"loss": 0.6537, "grad_norm": 0.8086010217666626, "learning_rate": 0.0002, "epoch": 2.395925955864522, "step": 14820}, {"loss": 0.7189, "grad_norm": 0.7189689874649048, "learning_rate": 0.0002, "epoch": 2.397542640045267, "step": 14830}, {"loss": 0.6709, "grad_norm": 0.6280009150505066, "learning_rate": 0.0002, "epoch": 2.3991593242260123, "step": 14840}, {"loss": 0.706, "grad_norm": 0.7826612591743469, "learning_rate": 0.0002, "epoch": 2.4007760084067575, "step": 14850}, {"loss": 0.6738, "grad_norm": 0.7681610584259033, "learning_rate": 0.0002, "epoch": 2.402392692587503, "step": 14860}, {"loss": 0.636, "grad_norm": 0.720966100692749, "learning_rate": 0.0002, "epoch": 2.4040093767682484, "step": 14870}, {"loss": 0.6667, "grad_norm": 0.8202250599861145, "learning_rate": 0.0002, "epoch": 2.4056260609489937, "step": 14880}, {"loss": 0.6935, "grad_norm": 0.786212682723999, "learning_rate": 0.0002, "epoch": 2.407242745129739, "step": 14890}, {"loss": 0.6628, "grad_norm": 0.6647164821624756, "learning_rate": 0.0002, "epoch": 2.408859429310484, "step": 14900}, {"loss": 0.6706, "grad_norm": 0.7566399574279785, "learning_rate": 0.0002, "epoch": 2.4104761134912294, "step": 14910}, {"loss": 0.7188, "grad_norm": 0.748814582824707, "learning_rate": 0.0002, "epoch": 2.4120927976719746, "step": 14920}, {"loss": 0.6684, "grad_norm": 0.7624038457870483, "learning_rate": 0.0002, "epoch": 2.4137094818527203, "step": 14930}, {"loss": 0.6483, "grad_norm": 0.8267335295677185, "learning_rate": 0.0002, "epoch": 2.4153261660334655, "step": 14940}, {"loss": 0.6612, "grad_norm": 0.8785360455513, "learning_rate": 0.0002, "epoch": 2.4169428502142107, "step": 14950}, {"loss": 0.6718, "grad_norm": 0.679887592792511, "learning_rate": 0.0002, "epoch": 2.418559534394956, "step": 14960}, {"loss": 0.6136, "grad_norm": 0.7218474745750427, "learning_rate": 0.0002, "epoch": 2.420176218575701, "step": 14970}, {"loss": 0.648, "grad_norm": 0.6342799663543701, "learning_rate": 0.0002, "epoch": 2.4217929027564464, "step": 14980}, {"loss": 0.6617, "grad_norm": 0.7098712921142578, "learning_rate": 0.0002, "epoch": 2.4234095869371917, "step": 14990}, {"loss": 0.6942, "grad_norm": 0.7497431635856628, "learning_rate": 0.0002, "epoch": 2.425026271117937, "step": 15000}, {"loss": 0.6772, "grad_norm": 0.934836208820343, "learning_rate": 0.0002, "epoch": 2.4266429552986826, "step": 15010}, {"loss": 0.7221, "grad_norm": 0.8430966734886169, "learning_rate": 0.0002, "epoch": 2.428259639479428, "step": 15020}, {"loss": 0.6985, "grad_norm": 0.7032104730606079, "learning_rate": 0.0002, "epoch": 2.429876323660173, "step": 15030}, {"loss": 0.6715, "grad_norm": 0.7746111750602722, "learning_rate": 0.0002, "epoch": 2.4314930078409183, "step": 15040}, {"loss": 0.7177, "grad_norm": 0.7661406397819519, "learning_rate": 0.0002, "epoch": 2.4331096920216635, "step": 15050}, {"loss": 0.6517, "grad_norm": 0.6941645741462708, "learning_rate": 0.0002, "epoch": 2.4347263762024087, "step": 15060}, {"loss": 0.6421, "grad_norm": 0.7487249374389648, "learning_rate": 0.0002, "epoch": 2.436343060383154, "step": 15070}, {"loss": 0.6796, "grad_norm": 0.7639912962913513, "learning_rate": 0.0002, "epoch": 2.4379597445638996, "step": 15080}, {"loss": 0.7087, "grad_norm": 0.7708953619003296, "learning_rate": 0.0002, "epoch": 2.439576428744645, "step": 15090}, {"loss": 0.7065, "grad_norm": 0.9135832190513611, "learning_rate": 0.0002, "epoch": 2.44119311292539, "step": 15100}, {"loss": 0.672, "grad_norm": 0.8283005356788635, "learning_rate": 0.0002, "epoch": 2.4428097971061353, "step": 15110}, {"loss": 0.6551, "grad_norm": 0.925299346446991, "learning_rate": 0.0002, "epoch": 2.4444264812868806, "step": 15120}, {"loss": 0.687, "grad_norm": 0.7013528943061829, "learning_rate": 0.0002, "epoch": 2.446043165467626, "step": 15130}, {"loss": 0.6842, "grad_norm": 0.622303307056427, "learning_rate": 0.0002, "epoch": 2.447659849648371, "step": 15140}, {"loss": 0.6676, "grad_norm": 0.876569390296936, "learning_rate": 0.0002, "epoch": 2.4492765338291163, "step": 15150}, {"loss": 0.6463, "grad_norm": 0.6836351752281189, "learning_rate": 0.0002, "epoch": 2.450893218009862, "step": 15160}, {"loss": 0.6781, "grad_norm": 0.7886684536933899, "learning_rate": 0.0002, "epoch": 2.452509902190607, "step": 15170}, {"loss": 0.6794, "grad_norm": 0.6647440791130066, "learning_rate": 0.0002, "epoch": 2.4541265863713524, "step": 15180}, {"loss": 0.6353, "grad_norm": 0.7477722764015198, "learning_rate": 0.0002, "epoch": 2.4557432705520976, "step": 15190}, {"loss": 0.698, "grad_norm": 0.8192033767700195, "learning_rate": 0.0002, "epoch": 2.457359954732843, "step": 15200}, {"loss": 0.6735, "grad_norm": 0.847537100315094, "learning_rate": 0.0002, "epoch": 2.458976638913588, "step": 15210}, {"loss": 0.6962, "grad_norm": 0.9027776122093201, "learning_rate": 0.0002, "epoch": 2.4605933230943338, "step": 15220}, {"loss": 0.7084, "grad_norm": 0.7217772006988525, "learning_rate": 0.0002, "epoch": 2.462210007275079, "step": 15230}, {"loss": 0.691, "grad_norm": 0.7994546294212341, "learning_rate": 0.0002, "epoch": 2.4638266914558242, "step": 15240}, {"loss": 0.6828, "grad_norm": 0.939916729927063, "learning_rate": 0.0002, "epoch": 2.4654433756365695, "step": 15250}, {"loss": 0.6893, "grad_norm": 1.0009053945541382, "learning_rate": 0.0002, "epoch": 2.4670600598173147, "step": 15260}, {"loss": 0.643, "grad_norm": 0.625555694103241, "learning_rate": 0.0002, "epoch": 2.46867674399806, "step": 15270}, {"loss": 0.688, "grad_norm": 0.7924878597259521, "learning_rate": 0.0002, "epoch": 2.470293428178805, "step": 15280}, {"loss": 0.6789, "grad_norm": 0.8536689877510071, "learning_rate": 0.0002, "epoch": 2.4719101123595504, "step": 15290}, {"loss": 0.6924, "grad_norm": 0.8572589755058289, "learning_rate": 0.0002, "epoch": 2.4735267965402956, "step": 15300}, {"loss": 0.604, "grad_norm": 0.773279070854187, "learning_rate": 0.0002, "epoch": 2.4751434807210413, "step": 15310}, {"loss": 0.6573, "grad_norm": 0.7708749771118164, "learning_rate": 0.0002, "epoch": 2.4767601649017865, "step": 15320}, {"loss": 0.7065, "grad_norm": 0.770905077457428, "learning_rate": 0.0002, "epoch": 2.4783768490825318, "step": 15330}, {"loss": 0.6878, "grad_norm": 0.8238571882247925, "learning_rate": 0.0002, "epoch": 2.479993533263277, "step": 15340}, {"loss": 0.6772, "grad_norm": 0.7670477032661438, "learning_rate": 0.0002, "epoch": 2.481610217444022, "step": 15350}, {"loss": 0.7759, "grad_norm": 0.905036985874176, "learning_rate": 0.0002, "epoch": 2.4832269016247674, "step": 15360}, {"loss": 0.706, "grad_norm": 0.6672089695930481, "learning_rate": 0.0002, "epoch": 2.484843585805513, "step": 15370}, {"loss": 0.6722, "grad_norm": 0.625095784664154, "learning_rate": 0.0002, "epoch": 2.4864602699862584, "step": 15380}, {"loss": 0.6396, "grad_norm": 0.679772675037384, "learning_rate": 0.0002, "epoch": 2.4880769541670036, "step": 15390}, {"loss": 0.6778, "grad_norm": 0.711492121219635, "learning_rate": 0.0002, "epoch": 2.489693638347749, "step": 15400}, {"loss": 0.6966, "grad_norm": 0.876189112663269, "learning_rate": 0.0002, "epoch": 2.491310322528494, "step": 15410}, {"loss": 0.7307, "grad_norm": 0.7236915230751038, "learning_rate": 0.0002, "epoch": 2.4929270067092393, "step": 15420}, {"loss": 0.647, "grad_norm": 0.6629832983016968, "learning_rate": 0.0002, "epoch": 2.4945436908899845, "step": 15430}, {"loss": 0.6669, "grad_norm": 0.9756859540939331, "learning_rate": 0.0002, "epoch": 2.4961603750707297, "step": 15440}, {"loss": 0.7559, "grad_norm": 0.6896940469741821, "learning_rate": 0.0002, "epoch": 2.4977770592514754, "step": 15450}, {"loss": 0.6818, "grad_norm": 0.7105149626731873, "learning_rate": 0.0002, "epoch": 2.4993937434322206, "step": 15460}, {"loss": 0.6859, "grad_norm": 0.8374546766281128, "learning_rate": 0.0002, "epoch": 2.501010427612966, "step": 15470}, {"loss": 0.6512, "grad_norm": 0.7320070266723633, "learning_rate": 0.0002, "epoch": 2.502627111793711, "step": 15480}, {"loss": 0.685, "grad_norm": 0.8306367993354797, "learning_rate": 0.0002, "epoch": 2.5042437959744563, "step": 15490}, {"loss": 0.7253, "grad_norm": 0.7472721338272095, "learning_rate": 0.0002, "epoch": 2.5058604801552016, "step": 15500}, {"loss": 0.6699, "grad_norm": 0.6147692203521729, "learning_rate": 0.0002, "epoch": 2.507477164335947, "step": 15510}, {"loss": 0.7158, "grad_norm": 0.7788505554199219, "learning_rate": 0.0002, "epoch": 2.5090938485166925, "step": 15520}, {"loss": 0.6521, "grad_norm": 0.8807527422904968, "learning_rate": 0.0002, "epoch": 2.5107105326974377, "step": 15530}, {"loss": 0.6792, "grad_norm": 0.7521643042564392, "learning_rate": 0.0002, "epoch": 2.512327216878183, "step": 15540}, {"loss": 0.6772, "grad_norm": 0.6900225281715393, "learning_rate": 0.0002, "epoch": 2.513943901058928, "step": 15550}, {"loss": 0.6769, "grad_norm": 0.6601938605308533, "learning_rate": 0.0002, "epoch": 2.5155605852396734, "step": 15560}, {"loss": 0.6648, "grad_norm": 0.8179984092712402, "learning_rate": 0.0002, "epoch": 2.5171772694204186, "step": 15570}, {"loss": 0.7028, "grad_norm": 0.792556881904602, "learning_rate": 0.0002, "epoch": 2.518793953601164, "step": 15580}, {"loss": 0.6464, "grad_norm": 0.7081938982009888, "learning_rate": 0.0002, "epoch": 2.520410637781909, "step": 15590}, {"loss": 0.6691, "grad_norm": 0.8733121156692505, "learning_rate": 0.0002, "epoch": 2.5220273219626543, "step": 15600}, {"loss": 0.6969, "grad_norm": 0.7980992794036865, "learning_rate": 0.0002, "epoch": 2.5236440061434, "step": 15610}, {"loss": 0.7124, "grad_norm": 0.883664071559906, "learning_rate": 0.0002, "epoch": 2.5252606903241452, "step": 15620}, {"loss": 0.7022, "grad_norm": 0.6963341236114502, "learning_rate": 0.0002, "epoch": 2.5268773745048905, "step": 15630}, {"loss": 0.7334, "grad_norm": 0.6433573365211487, "learning_rate": 0.0002, "epoch": 2.5284940586856357, "step": 15640}, {"loss": 0.6889, "grad_norm": 0.8538183569908142, "learning_rate": 0.0002, "epoch": 2.530110742866381, "step": 15650}, {"loss": 0.6841, "grad_norm": 0.9748201370239258, "learning_rate": 0.0002, "epoch": 2.5317274270471266, "step": 15660}, {"loss": 0.6765, "grad_norm": 0.7670575380325317, "learning_rate": 0.0002, "epoch": 2.533344111227872, "step": 15670}, {"loss": 0.6435, "grad_norm": 0.8738890290260315, "learning_rate": 0.0002, "epoch": 2.534960795408617, "step": 15680}, {"loss": 0.6802, "grad_norm": 0.8391636610031128, "learning_rate": 0.0002, "epoch": 2.5365774795893623, "step": 15690}, {"loss": 0.6901, "grad_norm": 0.7239366769790649, "learning_rate": 0.0002, "epoch": 2.5381941637701075, "step": 15700}, {"loss": 0.7011, "grad_norm": 0.8498379588127136, "learning_rate": 0.0002, "epoch": 2.5398108479508528, "step": 15710}, {"loss": 0.6998, "grad_norm": 0.8029484152793884, "learning_rate": 0.0002, "epoch": 2.541427532131598, "step": 15720}, {"loss": 0.6678, "grad_norm": 1.0639333724975586, "learning_rate": 0.0002, "epoch": 2.5430442163123432, "step": 15730}, {"loss": 0.6341, "grad_norm": 0.6401297450065613, "learning_rate": 0.0002, "epoch": 2.5446609004930885, "step": 15740}, {"loss": 0.7196, "grad_norm": 0.7123814821243286, "learning_rate": 0.0002, "epoch": 2.5462775846738337, "step": 15750}, {"loss": 0.654, "grad_norm": 0.7874974608421326, "learning_rate": 0.0002, "epoch": 2.5478942688545794, "step": 15760}, {"loss": 0.6721, "grad_norm": 0.8046808838844299, "learning_rate": 0.0002, "epoch": 2.5495109530353246, "step": 15770}, {"loss": 0.6665, "grad_norm": 0.7888661623001099, "learning_rate": 0.0002, "epoch": 2.55112763721607, "step": 15780}, {"loss": 0.6893, "grad_norm": 0.8445866107940674, "learning_rate": 0.0002, "epoch": 2.552744321396815, "step": 15790}, {"loss": 0.6815, "grad_norm": 0.7475846409797668, "learning_rate": 0.0002, "epoch": 2.5543610055775603, "step": 15800}, {"loss": 0.6711, "grad_norm": 0.7455102801322937, "learning_rate": 0.0002, "epoch": 2.555977689758306, "step": 15810}, {"loss": 0.6932, "grad_norm": 0.8226983547210693, "learning_rate": 0.0002, "epoch": 2.557594373939051, "step": 15820}, {"loss": 0.651, "grad_norm": 0.8920368552207947, "learning_rate": 0.0002, "epoch": 2.5592110581197964, "step": 15830}, {"loss": 0.6297, "grad_norm": 0.8413904905319214, "learning_rate": 0.0002, "epoch": 2.5608277423005417, "step": 15840}, {"loss": 0.7106, "grad_norm": 0.8483649492263794, "learning_rate": 0.0002, "epoch": 2.562444426481287, "step": 15850}, {"loss": 0.6957, "grad_norm": 0.5923284292221069, "learning_rate": 0.0002, "epoch": 2.564061110662032, "step": 15860}, {"loss": 0.6847, "grad_norm": 0.8518726229667664, "learning_rate": 0.0002, "epoch": 2.5656777948427774, "step": 15870}, {"loss": 0.6362, "grad_norm": 0.731235146522522, "learning_rate": 0.0002, "epoch": 2.5672944790235226, "step": 15880}, {"loss": 0.7611, "grad_norm": 0.7517194151878357, "learning_rate": 0.0002, "epoch": 2.568911163204268, "step": 15890}, {"loss": 0.6907, "grad_norm": 0.8378692269325256, "learning_rate": 0.0002, "epoch": 2.5705278473850135, "step": 15900}, {"loss": 0.7055, "grad_norm": 0.843701958656311, "learning_rate": 0.0002, "epoch": 2.5721445315657587, "step": 15910}, {"loss": 0.6882, "grad_norm": 0.7254629731178284, "learning_rate": 0.0002, "epoch": 2.573761215746504, "step": 15920}, {"loss": 0.6872, "grad_norm": 0.8863335847854614, "learning_rate": 0.0002, "epoch": 2.575377899927249, "step": 15930}, {"loss": 0.6813, "grad_norm": 0.7675097584724426, "learning_rate": 0.0002, "epoch": 2.5769945841079944, "step": 15940}, {"loss": 0.7357, "grad_norm": 0.82063889503479, "learning_rate": 0.0002, "epoch": 2.5786112682887397, "step": 15950}, {"loss": 0.662, "grad_norm": 0.7729717493057251, "learning_rate": 0.0002, "epoch": 2.5802279524694853, "step": 15960}, {"loss": 0.633, "grad_norm": 0.8301846981048584, "learning_rate": 0.0002, "epoch": 2.5818446366502306, "step": 15970}, {"loss": 0.6897, "grad_norm": 0.7906861305236816, "learning_rate": 0.0002, "epoch": 2.583461320830976, "step": 15980}, {"loss": 0.7175, "grad_norm": 0.6749057173728943, "learning_rate": 0.0002, "epoch": 2.585078005011721, "step": 15990}, {"loss": 0.7212, "grad_norm": 0.9386842846870422, "learning_rate": 0.0002, "epoch": 2.5866946891924663, "step": 16000}, {"loss": 0.6934, "grad_norm": 0.7868891358375549, "learning_rate": 0.0002, "epoch": 2.5883113733732115, "step": 16010}, {"loss": 0.7036, "grad_norm": 0.8674671053886414, "learning_rate": 0.0002, "epoch": 2.5899280575539567, "step": 16020}, {"loss": 0.7217, "grad_norm": 0.7043559551239014, "learning_rate": 0.0002, "epoch": 2.591544741734702, "step": 16030}, {"loss": 0.6967, "grad_norm": 0.5846083760261536, "learning_rate": 0.0002, "epoch": 2.593161425915447, "step": 16040}, {"loss": 0.7322, "grad_norm": 0.7323982119560242, "learning_rate": 0.0002, "epoch": 2.594778110096193, "step": 16050}, {"loss": 0.6794, "grad_norm": 0.9069556593894958, "learning_rate": 0.0002, "epoch": 2.596394794276938, "step": 16060}, {"loss": 0.7076, "grad_norm": 0.7522736191749573, "learning_rate": 0.0002, "epoch": 2.5980114784576833, "step": 16070}, {"loss": 0.6477, "grad_norm": 0.8149648308753967, "learning_rate": 0.0002, "epoch": 2.5996281626384286, "step": 16080}, {"loss": 0.6664, "grad_norm": 0.6214233040809631, "learning_rate": 0.0002, "epoch": 2.601244846819174, "step": 16090}, {"loss": 0.7307, "grad_norm": 0.6803743839263916, "learning_rate": 0.0002, "epoch": 2.602861530999919, "step": 16100}, {"loss": 0.7244, "grad_norm": 0.7223997116088867, "learning_rate": 0.0002, "epoch": 2.6044782151806647, "step": 16110}, {"loss": 0.6867, "grad_norm": 0.7324174642562866, "learning_rate": 0.0002, "epoch": 2.60609489936141, "step": 16120}, {"loss": 0.7159, "grad_norm": 0.9594739675521851, "learning_rate": 0.0002, "epoch": 2.607711583542155, "step": 16130}, {"loss": 0.6451, "grad_norm": 0.9485327005386353, "learning_rate": 0.0002, "epoch": 2.6093282677229004, "step": 16140}, {"loss": 0.6815, "grad_norm": 0.8449000120162964, "learning_rate": 0.0002, "epoch": 2.6109449519036456, "step": 16150}, {"loss": 0.7152, "grad_norm": 0.8520140051841736, "learning_rate": 0.0002, "epoch": 2.612561636084391, "step": 16160}, {"loss": 0.6759, "grad_norm": 0.7456524968147278, "learning_rate": 0.0002, "epoch": 2.614178320265136, "step": 16170}, {"loss": 0.6893, "grad_norm": 0.9912857413291931, "learning_rate": 0.0002, "epoch": 2.6157950044458813, "step": 16180}, {"loss": 0.7243, "grad_norm": 0.9001946449279785, "learning_rate": 0.0002, "epoch": 2.6174116886266265, "step": 16190}, {"loss": 0.6825, "grad_norm": 0.6568667888641357, "learning_rate": 0.0002, "epoch": 2.619028372807372, "step": 16200}, {"loss": 0.7013, "grad_norm": 1.0248128175735474, "learning_rate": 0.0002, "epoch": 2.6206450569881174, "step": 16210}, {"loss": 0.7045, "grad_norm": 0.6509039998054504, "learning_rate": 0.0002, "epoch": 2.6222617411688627, "step": 16220}, {"loss": 0.72, "grad_norm": 0.7626351118087769, "learning_rate": 0.0002, "epoch": 2.623878425349608, "step": 16230}, {"loss": 0.6556, "grad_norm": 0.6938552260398865, "learning_rate": 0.0002, "epoch": 2.625495109530353, "step": 16240}, {"loss": 0.65, "grad_norm": 0.6434680819511414, "learning_rate": 0.0002, "epoch": 2.6271117937110984, "step": 16250}, {"loss": 0.6943, "grad_norm": 0.7111515998840332, "learning_rate": 0.0002, "epoch": 2.628728477891844, "step": 16260}, {"loss": 0.679, "grad_norm": 0.7712395787239075, "learning_rate": 0.0002, "epoch": 2.6303451620725893, "step": 16270}, {"loss": 0.6886, "grad_norm": 0.792209267616272, "learning_rate": 0.0002, "epoch": 2.6319618462533345, "step": 16280}, {"loss": 0.6554, "grad_norm": 0.6801066398620605, "learning_rate": 0.0002, "epoch": 2.6335785304340797, "step": 16290}, {"loss": 0.73, "grad_norm": 0.7802573442459106, "learning_rate": 0.0002, "epoch": 2.635195214614825, "step": 16300}, {"loss": 0.7484, "grad_norm": 0.7742244601249695, "learning_rate": 0.0002, "epoch": 2.63681189879557, "step": 16310}, {"loss": 0.6524, "grad_norm": 0.664184033870697, "learning_rate": 0.0002, "epoch": 2.6384285829763154, "step": 16320}, {"loss": 0.6442, "grad_norm": 0.9242228865623474, "learning_rate": 0.0002, "epoch": 2.6400452671570607, "step": 16330}, {"loss": 0.6792, "grad_norm": 0.9661325216293335, "learning_rate": 0.0002, "epoch": 2.641661951337806, "step": 16340}, {"loss": 0.6847, "grad_norm": 0.837526798248291, "learning_rate": 0.0002, "epoch": 2.6432786355185516, "step": 16350}, {"loss": 0.7686, "grad_norm": 1.1834373474121094, "learning_rate": 0.0002, "epoch": 2.644895319699297, "step": 16360}, {"loss": 0.6746, "grad_norm": 0.7467831373214722, "learning_rate": 0.0002, "epoch": 2.646512003880042, "step": 16370}, {"loss": 0.6935, "grad_norm": 0.8627146482467651, "learning_rate": 0.0002, "epoch": 2.6481286880607873, "step": 16380}, {"loss": 0.715, "grad_norm": 0.790447473526001, "learning_rate": 0.0002, "epoch": 2.6497453722415325, "step": 16390}, {"loss": 0.723, "grad_norm": 0.8447365164756775, "learning_rate": 0.0002, "epoch": 2.651362056422278, "step": 16400}, {"loss": 0.6628, "grad_norm": 0.7831417918205261, "learning_rate": 0.0002, "epoch": 2.6529787406030234, "step": 16410}, {"loss": 0.6691, "grad_norm": 0.6837952136993408, "learning_rate": 0.0002, "epoch": 2.6545954247837686, "step": 16420}, {"loss": 0.6139, "grad_norm": 0.7031801342964172, "learning_rate": 0.0002, "epoch": 2.656212108964514, "step": 16430}, {"loss": 0.7382, "grad_norm": 0.8963770866394043, "learning_rate": 0.0002, "epoch": 2.657828793145259, "step": 16440}, {"loss": 0.6439, "grad_norm": 0.6852328181266785, "learning_rate": 0.0002, "epoch": 2.6594454773260043, "step": 16450}, {"loss": 0.6278, "grad_norm": 0.8069294095039368, "learning_rate": 0.0002, "epoch": 2.6610621615067496, "step": 16460}, {"loss": 0.6939, "grad_norm": 0.7503686547279358, "learning_rate": 0.0002, "epoch": 2.662678845687495, "step": 16470}, {"loss": 0.6777, "grad_norm": 0.6430956125259399, "learning_rate": 0.0002, "epoch": 2.66429552986824, "step": 16480}, {"loss": 0.6863, "grad_norm": 0.7894312739372253, "learning_rate": 0.0002, "epoch": 2.6659122140489853, "step": 16490}, {"loss": 0.7165, "grad_norm": 0.7277431488037109, "learning_rate": 0.0002, "epoch": 2.667528898229731, "step": 16500}, {"loss": 0.6772, "grad_norm": 0.6816153526306152, "learning_rate": 0.0002, "epoch": 2.669145582410476, "step": 16510}, {"loss": 0.691, "grad_norm": 0.8145235776901245, "learning_rate": 0.0002, "epoch": 2.6707622665912214, "step": 16520}, {"loss": 0.709, "grad_norm": 0.8645890355110168, "learning_rate": 0.0002, "epoch": 2.6723789507719666, "step": 16530}, {"loss": 0.6946, "grad_norm": 0.704393208026886, "learning_rate": 0.0002, "epoch": 2.673995634952712, "step": 16540}, {"loss": 0.6378, "grad_norm": 1.0120846033096313, "learning_rate": 0.0002, "epoch": 2.6756123191334575, "step": 16550}, {"loss": 0.7241, "grad_norm": 0.6919328570365906, "learning_rate": 0.0002, "epoch": 2.6772290033142028, "step": 16560}, {"loss": 0.7098, "grad_norm": 0.6924574971199036, "learning_rate": 0.0002, "epoch": 2.678845687494948, "step": 16570}, {"loss": 0.731, "grad_norm": 0.9679301381111145, "learning_rate": 0.0002, "epoch": 2.6804623716756932, "step": 16580}, {"loss": 0.7124, "grad_norm": 0.6810211539268494, "learning_rate": 0.0002, "epoch": 2.6820790558564385, "step": 16590}, {"loss": 0.6688, "grad_norm": 0.9730555415153503, "learning_rate": 0.0002, "epoch": 2.6836957400371837, "step": 16600}, {"loss": 0.7344, "grad_norm": 0.7852821350097656, "learning_rate": 0.0002, "epoch": 2.685312424217929, "step": 16610}, {"loss": 0.6401, "grad_norm": 0.6059057116508484, "learning_rate": 0.0002, "epoch": 2.686929108398674, "step": 16620}, {"loss": 0.6796, "grad_norm": 0.9395958781242371, "learning_rate": 0.0002, "epoch": 2.6885457925794194, "step": 16630}, {"loss": 0.7174, "grad_norm": 0.7473729848861694, "learning_rate": 0.0002, "epoch": 2.690162476760165, "step": 16640}, {"loss": 0.7087, "grad_norm": 0.765934407711029, "learning_rate": 0.0002, "epoch": 2.6917791609409103, "step": 16650}, {"loss": 0.707, "grad_norm": 0.8496677279472351, "learning_rate": 0.0002, "epoch": 2.6933958451216555, "step": 16660}, {"loss": 0.7084, "grad_norm": 0.7641879916191101, "learning_rate": 0.0002, "epoch": 2.6950125293024008, "step": 16670}, {"loss": 0.6566, "grad_norm": 0.8471952676773071, "learning_rate": 0.0002, "epoch": 2.696629213483146, "step": 16680}, {"loss": 0.6635, "grad_norm": 0.6946060657501221, "learning_rate": 0.0002, "epoch": 2.6982458976638912, "step": 16690}, {"loss": 0.7027, "grad_norm": 0.7361312508583069, "learning_rate": 0.0002, "epoch": 2.699862581844637, "step": 16700}, {"loss": 0.6767, "grad_norm": 0.6605038046836853, "learning_rate": 0.0002, "epoch": 2.701479266025382, "step": 16710}, {"loss": 0.6885, "grad_norm": 0.7164411544799805, "learning_rate": 0.0002, "epoch": 2.7030959502061274, "step": 16720}, {"loss": 0.6736, "grad_norm": 0.6496201157569885, "learning_rate": 0.0002, "epoch": 2.7047126343868726, "step": 16730}, {"loss": 0.6942, "grad_norm": 0.7826663851737976, "learning_rate": 0.0002, "epoch": 2.706329318567618, "step": 16740}, {"loss": 0.6773, "grad_norm": 0.7639131546020508, "learning_rate": 0.0002, "epoch": 2.707946002748363, "step": 16750}, {"loss": 0.69, "grad_norm": 0.7976210713386536, "learning_rate": 0.0002, "epoch": 2.7095626869291083, "step": 16760}, {"loss": 0.6735, "grad_norm": 0.6836577653884888, "learning_rate": 0.0002, "epoch": 2.7111793711098535, "step": 16770}, {"loss": 0.6596, "grad_norm": 0.8025202751159668, "learning_rate": 0.0002, "epoch": 2.7127960552905988, "step": 16780}, {"loss": 0.6324, "grad_norm": 0.7636463642120361, "learning_rate": 0.0002, "epoch": 2.7144127394713444, "step": 16790}, {"loss": 0.6227, "grad_norm": 0.7481677532196045, "learning_rate": 0.0002, "epoch": 2.7160294236520897, "step": 16800}, {"loss": 0.6925, "grad_norm": 0.7566834688186646, "learning_rate": 0.0002, "epoch": 2.717646107832835, "step": 16810}, {"loss": 0.6531, "grad_norm": 0.7931267619132996, "learning_rate": 0.0002, "epoch": 2.71926279201358, "step": 16820}, {"loss": 0.6672, "grad_norm": 0.8811662197113037, "learning_rate": 0.0002, "epoch": 2.7208794761943254, "step": 16830}, {"loss": 0.6675, "grad_norm": 0.8561240434646606, "learning_rate": 0.0002, "epoch": 2.7224961603750706, "step": 16840}, {"loss": 0.7135, "grad_norm": 0.7121599316596985, "learning_rate": 0.0002, "epoch": 2.7241128445558163, "step": 16850}, {"loss": 0.6825, "grad_norm": 0.8066257238388062, "learning_rate": 0.0002, "epoch": 2.7257295287365615, "step": 16860}, {"loss": 0.6839, "grad_norm": 0.7699271440505981, "learning_rate": 0.0002, "epoch": 2.7273462129173067, "step": 16870}, {"loss": 0.699, "grad_norm": 1.1828432083129883, "learning_rate": 0.0002, "epoch": 2.728962897098052, "step": 16880}, {"loss": 0.6518, "grad_norm": 0.9989302754402161, "learning_rate": 0.0002, "epoch": 2.730579581278797, "step": 16890}, {"loss": 0.7015, "grad_norm": 0.8100560307502747, "learning_rate": 0.0002, "epoch": 2.7321962654595424, "step": 16900}, {"loss": 0.6851, "grad_norm": 0.8615233898162842, "learning_rate": 0.0002, "epoch": 2.7338129496402876, "step": 16910}, {"loss": 0.6322, "grad_norm": 0.8633756041526794, "learning_rate": 0.0002, "epoch": 2.735429633821033, "step": 16920}, {"loss": 0.6488, "grad_norm": 0.7769348621368408, "learning_rate": 0.0002, "epoch": 2.737046318001778, "step": 16930}, {"loss": 0.6582, "grad_norm": 0.6943058371543884, "learning_rate": 0.0002, "epoch": 2.738663002182524, "step": 16940}, {"loss": 0.6516, "grad_norm": 0.8510736227035522, "learning_rate": 0.0002, "epoch": 2.740279686363269, "step": 16950}, {"loss": 0.7275, "grad_norm": 0.7732602953910828, "learning_rate": 0.0002, "epoch": 2.7418963705440142, "step": 16960}, {"loss": 0.6553, "grad_norm": 0.5981788635253906, "learning_rate": 0.0002, "epoch": 2.7435130547247595, "step": 16970}, {"loss": 0.6777, "grad_norm": 0.7604416012763977, "learning_rate": 0.0002, "epoch": 2.7451297389055047, "step": 16980}, {"loss": 0.6981, "grad_norm": 0.7377738356590271, "learning_rate": 0.0002, "epoch": 2.74674642308625, "step": 16990}, {"loss": 0.6294, "grad_norm": 0.9400289058685303, "learning_rate": 0.0002, "epoch": 2.7483631072669956, "step": 17000}, {"loss": 0.6952, "grad_norm": 0.6340599656105042, "learning_rate": 0.0002, "epoch": 2.749979791447741, "step": 17010}, {"loss": 0.7222, "grad_norm": 0.7297601103782654, "learning_rate": 0.0002, "epoch": 2.751596475628486, "step": 17020}, {"loss": 0.6659, "grad_norm": 0.9479979872703552, "learning_rate": 0.0002, "epoch": 2.7532131598092313, "step": 17030}, {"loss": 0.691, "grad_norm": 0.8461511135101318, "learning_rate": 0.0002, "epoch": 2.7548298439899765, "step": 17040}, {"loss": 0.6764, "grad_norm": 0.7477551698684692, "learning_rate": 0.0002, "epoch": 2.7564465281707218, "step": 17050}, {"loss": 0.684, "grad_norm": 1.019270420074463, "learning_rate": 0.0002, "epoch": 2.758063212351467, "step": 17060}, {"loss": 0.7119, "grad_norm": 0.7730235457420349, "learning_rate": 0.0002, "epoch": 2.7596798965322122, "step": 17070}, {"loss": 0.6886, "grad_norm": 0.8216866254806519, "learning_rate": 0.0002, "epoch": 2.7612965807129575, "step": 17080}, {"loss": 0.6811, "grad_norm": 0.7235931754112244, "learning_rate": 0.0002, "epoch": 2.762913264893703, "step": 17090}, {"loss": 0.7031, "grad_norm": 0.7352296710014343, "learning_rate": 0.0002, "epoch": 2.7645299490744484, "step": 17100}, {"loss": 0.6951, "grad_norm": 0.8129373788833618, "learning_rate": 0.0002, "epoch": 2.7661466332551936, "step": 17110}, {"loss": 0.6703, "grad_norm": 0.7387019991874695, "learning_rate": 0.0002, "epoch": 2.767763317435939, "step": 17120}, {"loss": 0.6789, "grad_norm": 0.9149190187454224, "learning_rate": 0.0002, "epoch": 2.769380001616684, "step": 17130}, {"loss": 0.6038, "grad_norm": 0.7352971434593201, "learning_rate": 0.0002, "epoch": 2.7709966857974297, "step": 17140}, {"loss": 0.6728, "grad_norm": 0.7903780341148376, "learning_rate": 0.0002, "epoch": 2.772613369978175, "step": 17150}, {"loss": 0.6988, "grad_norm": 0.8255927562713623, "learning_rate": 0.0002, "epoch": 2.77423005415892, "step": 17160}, {"loss": 0.6694, "grad_norm": 0.7235927581787109, "learning_rate": 0.0002, "epoch": 2.7758467383396654, "step": 17170}, {"loss": 0.7161, "grad_norm": 0.8281434774398804, "learning_rate": 0.0002, "epoch": 2.7774634225204107, "step": 17180}, {"loss": 0.682, "grad_norm": 0.7586921453475952, "learning_rate": 0.0002, "epoch": 2.779080106701156, "step": 17190}, {"loss": 0.6427, "grad_norm": 0.7161715030670166, "learning_rate": 0.0002, "epoch": 2.780696790881901, "step": 17200}, {"loss": 0.6426, "grad_norm": 0.762868344783783, "learning_rate": 0.0002, "epoch": 2.7823134750626464, "step": 17210}, {"loss": 0.705, "grad_norm": 0.9285483360290527, "learning_rate": 0.0002, "epoch": 2.7839301592433916, "step": 17220}, {"loss": 0.7084, "grad_norm": 0.6900462508201599, "learning_rate": 0.0002, "epoch": 2.785546843424137, "step": 17230}, {"loss": 0.6988, "grad_norm": 0.780384361743927, "learning_rate": 0.0002, "epoch": 2.7871635276048825, "step": 17240}, {"loss": 0.7073, "grad_norm": 0.7580406665802002, "learning_rate": 0.0002, "epoch": 2.7887802117856277, "step": 17250}, {"loss": 0.6833, "grad_norm": 0.8145199418067932, "learning_rate": 0.0002, "epoch": 2.790396895966373, "step": 17260}, {"loss": 0.6909, "grad_norm": 0.9159596562385559, "learning_rate": 0.0002, "epoch": 2.792013580147118, "step": 17270}, {"loss": 0.6008, "grad_norm": 0.9590014219284058, "learning_rate": 0.0002, "epoch": 2.7936302643278634, "step": 17280}, {"loss": 0.6704, "grad_norm": 0.7603529691696167, "learning_rate": 0.0002, "epoch": 2.795246948508609, "step": 17290}, {"loss": 0.7165, "grad_norm": 0.8039976358413696, "learning_rate": 0.0002, "epoch": 2.7968636326893543, "step": 17300}, {"loss": 0.7037, "grad_norm": 0.8364847302436829, "learning_rate": 0.0002, "epoch": 2.7984803168700996, "step": 17310}, {"loss": 0.6749, "grad_norm": 0.8763046860694885, "learning_rate": 0.0002, "epoch": 2.800097001050845, "step": 17320}, {"loss": 0.6844, "grad_norm": 0.8409647941589355, "learning_rate": 0.0002, "epoch": 2.80171368523159, "step": 17330}, {"loss": 0.6936, "grad_norm": 0.7649006247520447, "learning_rate": 0.0002, "epoch": 2.8033303694123353, "step": 17340}, {"loss": 0.7051, "grad_norm": 0.7970262169837952, "learning_rate": 0.0002, "epoch": 2.8049470535930805, "step": 17350}, {"loss": 0.6533, "grad_norm": 0.9088607430458069, "learning_rate": 0.0002, "epoch": 2.8065637377738257, "step": 17360}, {"loss": 0.675, "grad_norm": 0.6454846858978271, "learning_rate": 0.0002, "epoch": 2.808180421954571, "step": 17370}, {"loss": 0.7069, "grad_norm": 0.7744787931442261, "learning_rate": 0.0002, "epoch": 2.809797106135316, "step": 17380}, {"loss": 0.6772, "grad_norm": 0.6678640842437744, "learning_rate": 0.0002, "epoch": 2.811413790316062, "step": 17390}, {"loss": 0.6784, "grad_norm": 0.772676944732666, "learning_rate": 0.0002, "epoch": 2.813030474496807, "step": 17400}, {"loss": 0.7252, "grad_norm": 0.7088175415992737, "learning_rate": 0.0002, "epoch": 2.8146471586775523, "step": 17410}, {"loss": 0.7086, "grad_norm": 0.8280573487281799, "learning_rate": 0.0002, "epoch": 2.8162638428582976, "step": 17420}, {"loss": 0.6732, "grad_norm": 0.6665388345718384, "learning_rate": 0.0002, "epoch": 2.817880527039043, "step": 17430}, {"loss": 0.6675, "grad_norm": 0.6427883505821228, "learning_rate": 0.0002, "epoch": 2.8194972112197885, "step": 17440}, {"loss": 0.6972, "grad_norm": 0.9697760343551636, "learning_rate": 0.0002, "epoch": 2.8211138954005337, "step": 17450}, {"loss": 0.6838, "grad_norm": 0.7573966383934021, "learning_rate": 0.0002, "epoch": 2.822730579581279, "step": 17460}, {"loss": 0.7243, "grad_norm": 0.878688633441925, "learning_rate": 0.0002, "epoch": 2.824347263762024, "step": 17470}, {"loss": 0.6666, "grad_norm": 0.7752242684364319, "learning_rate": 0.0002, "epoch": 2.8259639479427694, "step": 17480}, {"loss": 0.6638, "grad_norm": 0.6135398745536804, "learning_rate": 0.0002, "epoch": 2.8275806321235146, "step": 17490}, {"loss": 0.6829, "grad_norm": 0.6924924850463867, "learning_rate": 0.0002, "epoch": 2.82919731630426, "step": 17500}, {"loss": 0.6731, "grad_norm": 0.7471627593040466, "learning_rate": 0.0002, "epoch": 2.830814000485005, "step": 17510}, {"loss": 0.7016, "grad_norm": 0.7145499587059021, "learning_rate": 0.0002, "epoch": 2.8324306846657503, "step": 17520}, {"loss": 0.6787, "grad_norm": 0.7415414452552795, "learning_rate": 0.0002, "epoch": 2.834047368846496, "step": 17530}, {"loss": 0.6811, "grad_norm": 0.7328441739082336, "learning_rate": 0.0002, "epoch": 2.8356640530272412, "step": 17540}, {"loss": 0.6866, "grad_norm": 0.8267839550971985, "learning_rate": 0.0002, "epoch": 2.8372807372079865, "step": 17550}, {"loss": 0.6787, "grad_norm": 0.8877885341644287, "learning_rate": 0.0002, "epoch": 2.8388974213887317, "step": 17560}, {"loss": 0.7136, "grad_norm": 0.857138454914093, "learning_rate": 0.0002, "epoch": 2.840514105569477, "step": 17570}, {"loss": 0.6454, "grad_norm": 0.8470779657363892, "learning_rate": 0.0002, "epoch": 2.842130789750222, "step": 17580}, {"loss": 0.6976, "grad_norm": 0.8553254008293152, "learning_rate": 0.0002, "epoch": 2.843747473930968, "step": 17590}, {"loss": 0.7297, "grad_norm": 0.8033196926116943, "learning_rate": 0.0002, "epoch": 2.845364158111713, "step": 17600}, {"loss": 0.7062, "grad_norm": 0.7949087023735046, "learning_rate": 0.0002, "epoch": 2.8469808422924583, "step": 17610}, {"loss": 0.651, "grad_norm": 0.9241406321525574, "learning_rate": 0.0002, "epoch": 2.8485975264732035, "step": 17620}, {"loss": 0.6601, "grad_norm": 0.7721285223960876, "learning_rate": 0.0002, "epoch": 2.8502142106539488, "step": 17630}, {"loss": 0.6183, "grad_norm": 1.0246692895889282, "learning_rate": 0.0002, "epoch": 2.851830894834694, "step": 17640}, {"loss": 0.7007, "grad_norm": 0.9244589805603027, "learning_rate": 0.0002, "epoch": 2.853447579015439, "step": 17650}, {"loss": 0.7274, "grad_norm": 0.7243508696556091, "learning_rate": 0.0002, "epoch": 2.8550642631961844, "step": 17660}, {"loss": 0.6471, "grad_norm": 0.8943371176719666, "learning_rate": 0.0002, "epoch": 2.8566809473769297, "step": 17670}, {"loss": 0.686, "grad_norm": 0.6531758904457092, "learning_rate": 0.0002, "epoch": 2.8582976315576754, "step": 17680}, {"loss": 0.6253, "grad_norm": 0.8367000818252563, "learning_rate": 0.0002, "epoch": 2.8599143157384206, "step": 17690}, {"loss": 0.6943, "grad_norm": 0.7868556380271912, "learning_rate": 0.0002, "epoch": 2.861530999919166, "step": 17700}, {"loss": 0.6919, "grad_norm": 0.7213859558105469, "learning_rate": 0.0002, "epoch": 2.863147684099911, "step": 17710}, {"loss": 0.6657, "grad_norm": 0.7383931279182434, "learning_rate": 0.0002, "epoch": 2.8647643682806563, "step": 17720}, {"loss": 0.6841, "grad_norm": 0.7566812634468079, "learning_rate": 0.0002, "epoch": 2.8663810524614015, "step": 17730}, {"loss": 0.6449, "grad_norm": 0.6930373311042786, "learning_rate": 0.0002, "epoch": 2.867997736642147, "step": 17740}, {"loss": 0.6764, "grad_norm": 0.7911090850830078, "learning_rate": 0.0002, "epoch": 2.8696144208228924, "step": 17750}, {"loss": 0.6554, "grad_norm": 0.8484548926353455, "learning_rate": 0.0002, "epoch": 2.8712311050036377, "step": 17760}, {"loss": 0.6931, "grad_norm": 0.7647597193717957, "learning_rate": 0.0002, "epoch": 2.872847789184383, "step": 17770}, {"loss": 0.6945, "grad_norm": 0.8791151642799377, "learning_rate": 0.0002, "epoch": 2.874464473365128, "step": 17780}, {"loss": 0.7078, "grad_norm": 0.7253178358078003, "learning_rate": 0.0002, "epoch": 2.8760811575458733, "step": 17790}, {"loss": 0.6474, "grad_norm": 0.7956077456474304, "learning_rate": 0.0002, "epoch": 2.8776978417266186, "step": 17800}, {"loss": 0.6687, "grad_norm": 0.8657688498497009, "learning_rate": 0.0002, "epoch": 2.879314525907364, "step": 17810}, {"loss": 0.7171, "grad_norm": 0.7059141993522644, "learning_rate": 0.0002, "epoch": 2.880931210088109, "step": 17820}, {"loss": 0.683, "grad_norm": 0.8886896967887878, "learning_rate": 0.0002, "epoch": 2.8825478942688547, "step": 17830}, {"loss": 0.669, "grad_norm": 0.821032702922821, "learning_rate": 0.0002, "epoch": 2.8841645784496, "step": 17840}, {"loss": 0.6805, "grad_norm": 0.7183963656425476, "learning_rate": 0.0002, "epoch": 2.885781262630345, "step": 17850}, {"loss": 0.7088, "grad_norm": 0.6222899556159973, "learning_rate": 0.0002, "epoch": 2.8873979468110904, "step": 17860}, {"loss": 0.6626, "grad_norm": 0.8187434077262878, "learning_rate": 0.0002, "epoch": 2.8890146309918356, "step": 17870}, {"loss": 0.6815, "grad_norm": 0.9838479161262512, "learning_rate": 0.0002, "epoch": 2.890631315172581, "step": 17880}, {"loss": 0.6967, "grad_norm": 0.7567742466926575, "learning_rate": 0.0002, "epoch": 2.8922479993533265, "step": 17890}, {"loss": 0.7073, "grad_norm": 0.6875903606414795, "learning_rate": 0.0002, "epoch": 2.893864683534072, "step": 17900}, {"loss": 0.6415, "grad_norm": 0.8043789267539978, "learning_rate": 0.0002, "epoch": 2.895481367714817, "step": 17910}, {"loss": 0.6588, "grad_norm": 0.8062626719474792, "learning_rate": 0.0002, "epoch": 2.8970980518955622, "step": 17920}, {"loss": 0.7151, "grad_norm": 1.0251191854476929, "learning_rate": 0.0002, "epoch": 2.8987147360763075, "step": 17930}, {"loss": 0.6605, "grad_norm": 0.882253110408783, "learning_rate": 0.0002, "epoch": 2.9003314202570527, "step": 17940}, {"loss": 0.6719, "grad_norm": 0.8683299422264099, "learning_rate": 0.0002, "epoch": 2.901948104437798, "step": 17950}, {"loss": 0.6896, "grad_norm": 0.7167282104492188, "learning_rate": 0.0002, "epoch": 2.903564788618543, "step": 17960}, {"loss": 0.663, "grad_norm": 0.7093694806098938, "learning_rate": 0.0002, "epoch": 2.9051814727992884, "step": 17970}, {"loss": 0.6591, "grad_norm": 0.8549879193305969, "learning_rate": 0.0002, "epoch": 2.906798156980034, "step": 17980}, {"loss": 0.6962, "grad_norm": 0.6989606618881226, "learning_rate": 0.0002, "epoch": 2.9084148411607793, "step": 17990}, {"loss": 0.6635, "grad_norm": 0.9482976794242859, "learning_rate": 0.0002, "epoch": 2.9100315253415245, "step": 18000}, {"loss": 0.6586, "grad_norm": 0.7182440161705017, "learning_rate": 0.0002, "epoch": 2.9116482095222698, "step": 18010}, {"loss": 0.6827, "grad_norm": 0.7732226252555847, "learning_rate": 0.0002, "epoch": 2.913264893703015, "step": 18020}, {"loss": 0.7123, "grad_norm": 0.7936875224113464, "learning_rate": 0.0002, "epoch": 2.9148815778837607, "step": 18030}, {"loss": 0.6736, "grad_norm": 0.8825615644454956, "learning_rate": 0.0002, "epoch": 2.916498262064506, "step": 18040}, {"loss": 0.7139, "grad_norm": 0.6778587102890015, "learning_rate": 0.0002, "epoch": 2.918114946245251, "step": 18050}, {"loss": 0.6588, "grad_norm": 0.7529265880584717, "learning_rate": 0.0002, "epoch": 2.9197316304259964, "step": 18060}, {"loss": 0.737, "grad_norm": 0.7111883163452148, "learning_rate": 0.0002, "epoch": 2.9213483146067416, "step": 18070}, {"loss": 0.7475, "grad_norm": 0.7214767932891846, "learning_rate": 0.0002, "epoch": 2.922964998787487, "step": 18080}, {"loss": 0.6672, "grad_norm": 0.800417423248291, "learning_rate": 0.0002, "epoch": 2.924581682968232, "step": 18090}, {"loss": 0.6694, "grad_norm": 1.248575210571289, "learning_rate": 0.0002, "epoch": 2.9261983671489773, "step": 18100}, {"loss": 0.7004, "grad_norm": 0.757788360118866, "learning_rate": 0.0002, "epoch": 2.9278150513297225, "step": 18110}, {"loss": 0.6999, "grad_norm": 1.0583995580673218, "learning_rate": 0.0002, "epoch": 2.9294317355104678, "step": 18120}, {"loss": 0.6365, "grad_norm": 0.8228777647018433, "learning_rate": 0.0002, "epoch": 2.9310484196912134, "step": 18130}, {"loss": 0.6791, "grad_norm": 0.8374035358428955, "learning_rate": 0.0002, "epoch": 2.9326651038719587, "step": 18140}, {"loss": 0.6399, "grad_norm": 0.7976473569869995, "learning_rate": 0.0002, "epoch": 2.934281788052704, "step": 18150}, {"loss": 0.6585, "grad_norm": 0.8009907603263855, "learning_rate": 0.0002, "epoch": 2.935898472233449, "step": 18160}, {"loss": 0.7485, "grad_norm": 0.835213303565979, "learning_rate": 0.0002, "epoch": 2.9375151564141944, "step": 18170}, {"loss": 0.7376, "grad_norm": 0.7982219457626343, "learning_rate": 0.0002, "epoch": 2.93913184059494, "step": 18180}, {"loss": 0.6348, "grad_norm": 0.7070978879928589, "learning_rate": 0.0002, "epoch": 2.9407485247756853, "step": 18190}, {"loss": 0.6608, "grad_norm": 0.8619440197944641, "learning_rate": 0.0002, "epoch": 2.9423652089564305, "step": 18200}, {"loss": 0.666, "grad_norm": 0.6693987250328064, "learning_rate": 0.0002, "epoch": 2.9439818931371757, "step": 18210}, {"loss": 0.728, "grad_norm": 0.6747021079063416, "learning_rate": 0.0002, "epoch": 2.945598577317921, "step": 18220}, {"loss": 0.6686, "grad_norm": 0.860387921333313, "learning_rate": 0.0002, "epoch": 2.947215261498666, "step": 18230}, {"loss": 0.6945, "grad_norm": 0.799976646900177, "learning_rate": 0.0002, "epoch": 2.9488319456794114, "step": 18240}, {"loss": 0.7243, "grad_norm": 0.7864769101142883, "learning_rate": 0.0002, "epoch": 2.9504486298601567, "step": 18250}, {"loss": 0.6785, "grad_norm": 0.6713884472846985, "learning_rate": 0.0002, "epoch": 2.952065314040902, "step": 18260}, {"loss": 0.7429, "grad_norm": 0.9031508564949036, "learning_rate": 0.0002, "epoch": 2.9536819982216476, "step": 18270}, {"loss": 0.7055, "grad_norm": 0.7205073237419128, "learning_rate": 0.0002, "epoch": 2.955298682402393, "step": 18280}, {"loss": 0.7298, "grad_norm": 0.7746205925941467, "learning_rate": 0.0002, "epoch": 2.956915366583138, "step": 18290}, {"loss": 0.6218, "grad_norm": 0.6533427834510803, "learning_rate": 0.0002, "epoch": 2.9585320507638833, "step": 18300}, {"loss": 0.6674, "grad_norm": 0.9083208441734314, "learning_rate": 0.0002, "epoch": 2.9601487349446285, "step": 18310}, {"loss": 0.7359, "grad_norm": 0.7446991801261902, "learning_rate": 0.0002, "epoch": 2.9617654191253737, "step": 18320}, {"loss": 0.6738, "grad_norm": 0.6514461636543274, "learning_rate": 0.0002, "epoch": 2.9633821033061194, "step": 18330}, {"loss": 0.6677, "grad_norm": 0.8580465912818909, "learning_rate": 0.0002, "epoch": 2.9649987874868646, "step": 18340}, {"loss": 0.6971, "grad_norm": 0.7074266076087952, "learning_rate": 0.0002, "epoch": 2.96661547166761, "step": 18350}, {"loss": 0.6804, "grad_norm": 0.899892270565033, "learning_rate": 0.0002, "epoch": 2.968232155848355, "step": 18360}, {"loss": 0.7094, "grad_norm": 0.8217641711235046, "learning_rate": 0.0002, "epoch": 2.9698488400291003, "step": 18370}, {"loss": 0.6916, "grad_norm": 0.8611799478530884, "learning_rate": 0.0002, "epoch": 2.9714655242098456, "step": 18380}, {"loss": 0.6677, "grad_norm": 0.6909302473068237, "learning_rate": 0.0002, "epoch": 2.973082208390591, "step": 18390}, {"loss": 0.7247, "grad_norm": 0.6554358005523682, "learning_rate": 0.0002, "epoch": 2.974698892571336, "step": 18400}, {"loss": 0.6516, "grad_norm": 0.7803071737289429, "learning_rate": 0.0002, "epoch": 2.9763155767520812, "step": 18410}, {"loss": 0.7322, "grad_norm": 0.7838954925537109, "learning_rate": 0.0002, "epoch": 2.977932260932827, "step": 18420}, {"loss": 0.6522, "grad_norm": 0.7098495364189148, "learning_rate": 0.0002, "epoch": 2.979548945113572, "step": 18430}, {"loss": 0.739, "grad_norm": 0.8981785774230957, "learning_rate": 0.0002, "epoch": 2.9811656292943174, "step": 18440}, {"loss": 0.6689, "grad_norm": 0.7197171449661255, "learning_rate": 0.0002, "epoch": 2.9827823134750626, "step": 18450}, {"loss": 0.706, "grad_norm": 0.793185293674469, "learning_rate": 0.0002, "epoch": 2.984398997655808, "step": 18460}, {"loss": 0.7124, "grad_norm": 0.8531473875045776, "learning_rate": 0.0002, "epoch": 2.986015681836553, "step": 18470}, {"loss": 0.6901, "grad_norm": 0.6627361178398132, "learning_rate": 0.0002, "epoch": 2.9876323660172988, "step": 18480}, {"loss": 0.6591, "grad_norm": 0.5708155035972595, "learning_rate": 0.0002, "epoch": 2.989249050198044, "step": 18490}, {"loss": 0.6725, "grad_norm": 0.8227280378341675, "learning_rate": 0.0002, "epoch": 2.990865734378789, "step": 18500}, {"loss": 0.6701, "grad_norm": 0.7102749943733215, "learning_rate": 0.0002, "epoch": 2.9924824185595345, "step": 18510}, {"loss": 0.7091, "grad_norm": 0.839485228061676, "learning_rate": 0.0002, "epoch": 2.9940991027402797, "step": 18520}, {"loss": 0.6521, "grad_norm": 0.9038704037666321, "learning_rate": 0.0002, "epoch": 2.995715786921025, "step": 18530}, {"loss": 0.7186, "grad_norm": 0.8737510442733765, "learning_rate": 0.0002, "epoch": 2.99733247110177, "step": 18540}, {"loss": 0.6819, "grad_norm": 0.7323142886161804, "learning_rate": 0.0002, "epoch": 2.9989491552825154, "step": 18550}, {"eval_loss": 1.1262480020523071, "eval_runtime": 122.0868, "eval_samples_per_second": 6.004, "eval_steps_per_second": 0.754, "epoch": 2.9999191657909625, "step": 18556}, {"loss": 0.6337, "grad_norm": 0.8465463519096375, "learning_rate": 0.0002, "epoch": 3.000565839463261, "step": 18560}, {"loss": 0.6064, "grad_norm": 0.9134138822555542, "learning_rate": 0.0002, "epoch": 3.0021825236440063, "step": 18570}, {"loss": 0.5804, "grad_norm": 0.760715126991272, "learning_rate": 0.0002, "epoch": 3.0037992078247515, "step": 18580}, {"loss": 0.5571, "grad_norm": 0.9208743572235107, "learning_rate": 0.0002, "epoch": 3.0054158920054967, "step": 18590}, {"loss": 0.5731, "grad_norm": 0.9232364892959595, "learning_rate": 0.0002, "epoch": 3.007032576186242, "step": 18600}, {"loss": 0.6299, "grad_norm": 1.1881544589996338, "learning_rate": 0.0002, "epoch": 3.008649260366987, "step": 18610}, {"loss": 0.5482, "grad_norm": 0.9372987747192383, "learning_rate": 0.0002, "epoch": 3.0102659445477324, "step": 18620}, {"loss": 0.5709, "grad_norm": 0.6900241374969482, "learning_rate": 0.0002, "epoch": 3.0118826287284777, "step": 18630}, {"loss": 0.5256, "grad_norm": 0.8451071381568909, "learning_rate": 0.0002, "epoch": 3.0134993129092233, "step": 18640}, {"loss": 0.5916, "grad_norm": 0.7763112187385559, "learning_rate": 0.0002, "epoch": 3.0151159970899686, "step": 18650}, {"loss": 0.6095, "grad_norm": 1.043653964996338, "learning_rate": 0.0002, "epoch": 3.016732681270714, "step": 18660}, {"loss": 0.6228, "grad_norm": 1.0170660018920898, "learning_rate": 0.0002, "epoch": 3.018349365451459, "step": 18670}, {"loss": 0.5671, "grad_norm": 0.7534180283546448, "learning_rate": 0.0002, "epoch": 3.0199660496322043, "step": 18680}, {"loss": 0.6015, "grad_norm": 0.7507367730140686, "learning_rate": 0.0002, "epoch": 3.0215827338129495, "step": 18690}, {"loss": 0.6201, "grad_norm": 0.7861620187759399, "learning_rate": 0.0002, "epoch": 3.0231994179936947, "step": 18700}, {"loss": 0.5802, "grad_norm": 1.0580339431762695, "learning_rate": 0.0002, "epoch": 3.0248161021744404, "step": 18710}, {"loss": 0.5975, "grad_norm": 0.7542710900306702, "learning_rate": 0.0002, "epoch": 3.0264327863551856, "step": 18720}, {"loss": 0.5695, "grad_norm": 0.8189544677734375, "learning_rate": 0.0002, "epoch": 3.028049470535931, "step": 18730}, {"loss": 0.6109, "grad_norm": 0.9126611351966858, "learning_rate": 0.0002, "epoch": 3.029666154716676, "step": 18740}, {"loss": 0.6443, "grad_norm": 0.8891341686248779, "learning_rate": 0.0002, "epoch": 3.0312828388974213, "step": 18750}, {"loss": 0.6207, "grad_norm": 0.8419283032417297, "learning_rate": 0.0002, "epoch": 3.0328995230781666, "step": 18760}, {"loss": 0.5818, "grad_norm": 0.8048048615455627, "learning_rate": 0.0002, "epoch": 3.034516207258912, "step": 18770}, {"loss": 0.6381, "grad_norm": 0.7820217609405518, "learning_rate": 0.0002, "epoch": 3.0361328914396575, "step": 18780}, {"loss": 0.5843, "grad_norm": 0.854721188545227, "learning_rate": 0.0002, "epoch": 3.0377495756204027, "step": 18790}, {"loss": 0.5784, "grad_norm": 0.912092924118042, "learning_rate": 0.0002, "epoch": 3.039366259801148, "step": 18800}, {"loss": 0.5734, "grad_norm": 0.6596226096153259, "learning_rate": 0.0002, "epoch": 3.040982943981893, "step": 18810}, {"loss": 0.5969, "grad_norm": 0.6351348757743835, "learning_rate": 0.0002, "epoch": 3.0425996281626384, "step": 18820}, {"loss": 0.5953, "grad_norm": 0.778188943862915, "learning_rate": 0.0002, "epoch": 3.0442163123433836, "step": 18830}, {"loss": 0.602, "grad_norm": 0.68234783411026, "learning_rate": 0.0002, "epoch": 3.045832996524129, "step": 18840}, {"loss": 0.5785, "grad_norm": 0.998628556728363, "learning_rate": 0.0002, "epoch": 3.047449680704874, "step": 18850}, {"loss": 0.6231, "grad_norm": 0.7393841743469238, "learning_rate": 0.0002, "epoch": 3.0490663648856198, "step": 18860}, {"loss": 0.568, "grad_norm": 0.84438556432724, "learning_rate": 0.0002, "epoch": 3.050683049066365, "step": 18870}, {"loss": 0.6205, "grad_norm": 0.8857501745223999, "learning_rate": 0.0002, "epoch": 3.0522997332471102, "step": 18880}, {"loss": 0.6335, "grad_norm": 0.7208474278450012, "learning_rate": 0.0002, "epoch": 3.0539164174278555, "step": 18890}, {"loss": 0.5998, "grad_norm": 0.7135229110717773, "learning_rate": 0.0002, "epoch": 3.0555331016086007, "step": 18900}, {"loss": 0.5575, "grad_norm": 0.9130001664161682, "learning_rate": 0.0002, "epoch": 3.057149785789346, "step": 18910}, {"loss": 0.5955, "grad_norm": 0.9001716375350952, "learning_rate": 0.0002, "epoch": 3.058766469970091, "step": 18920}, {"loss": 0.6052, "grad_norm": 0.8667559623718262, "learning_rate": 0.0002, "epoch": 3.060383154150837, "step": 18930}, {"loss": 0.5818, "grad_norm": 0.8943959474563599, "learning_rate": 0.0002, "epoch": 3.061999838331582, "step": 18940}, {"loss": 0.5978, "grad_norm": 0.8298377990722656, "learning_rate": 0.0002, "epoch": 3.0636165225123273, "step": 18950}, {"loss": 0.5782, "grad_norm": 0.7935267686843872, "learning_rate": 0.0002, "epoch": 3.0652332066930725, "step": 18960}, {"loss": 0.6434, "grad_norm": 1.1506379842758179, "learning_rate": 0.0002, "epoch": 3.0668498908738178, "step": 18970}, {"loss": 0.5571, "grad_norm": 0.7693049907684326, "learning_rate": 0.0002, "epoch": 3.068466575054563, "step": 18980}, {"loss": 0.5971, "grad_norm": 0.8040135502815247, "learning_rate": 0.0002, "epoch": 3.0700832592353082, "step": 18990}, {"loss": 0.5541, "grad_norm": 0.828404426574707, "learning_rate": 0.0002, "epoch": 3.0716999434160535, "step": 19000}, {"loss": 0.6048, "grad_norm": 0.8811164498329163, "learning_rate": 0.0002, "epoch": 3.073316627596799, "step": 19010}, {"loss": 0.5845, "grad_norm": 1.036205768585205, "learning_rate": 0.0002, "epoch": 3.0749333117775444, "step": 19020}, {"loss": 0.5838, "grad_norm": 0.8857285976409912, "learning_rate": 0.0002, "epoch": 3.0765499959582896, "step": 19030}, {"loss": 0.592, "grad_norm": 0.8392079472541809, "learning_rate": 0.0002, "epoch": 3.078166680139035, "step": 19040}, {"loss": 0.5927, "grad_norm": 1.0287401676177979, "learning_rate": 0.0002, "epoch": 3.07978336431978, "step": 19050}, {"loss": 0.5964, "grad_norm": 1.0086315870285034, "learning_rate": 0.0002, "epoch": 3.0814000485005253, "step": 19060}, {"loss": 0.5567, "grad_norm": 0.9245324730873108, "learning_rate": 0.0002, "epoch": 3.0830167326812705, "step": 19070}, {"loss": 0.5797, "grad_norm": 0.8680877089500427, "learning_rate": 0.0002, "epoch": 3.084633416862016, "step": 19080}, {"loss": 0.5611, "grad_norm": 0.8814793825149536, "learning_rate": 0.0002, "epoch": 3.0862501010427614, "step": 19090}, {"loss": 0.6051, "grad_norm": 0.9234458208084106, "learning_rate": 0.0002, "epoch": 3.0878667852235067, "step": 19100}, {"loss": 0.6209, "grad_norm": 1.1291664838790894, "learning_rate": 0.0002, "epoch": 3.089483469404252, "step": 19110}, {"loss": 0.5695, "grad_norm": 0.9191402792930603, "learning_rate": 0.0002, "epoch": 3.091100153584997, "step": 19120}, {"loss": 0.5856, "grad_norm": 0.7103154063224792, "learning_rate": 0.0002, "epoch": 3.0927168377657424, "step": 19130}, {"loss": 0.6479, "grad_norm": 0.9368883967399597, "learning_rate": 0.0002, "epoch": 3.0943335219464876, "step": 19140}, {"loss": 0.6167, "grad_norm": 0.9676656723022461, "learning_rate": 0.0002, "epoch": 3.095950206127233, "step": 19150}, {"loss": 0.5794, "grad_norm": 0.8739792704582214, "learning_rate": 0.0002, "epoch": 3.0975668903079785, "step": 19160}, {"loss": 0.6112, "grad_norm": 0.8530174493789673, "learning_rate": 0.0002, "epoch": 3.0991835744887237, "step": 19170}, {"loss": 0.6568, "grad_norm": 0.794945478439331, "learning_rate": 0.0002, "epoch": 3.100800258669469, "step": 19180}, {"loss": 0.5928, "grad_norm": 0.9508888125419617, "learning_rate": 0.0002, "epoch": 3.102416942850214, "step": 19190}, {"loss": 0.5757, "grad_norm": 1.0599955320358276, "learning_rate": 0.0002, "epoch": 3.1040336270309594, "step": 19200}, {"loss": 0.6151, "grad_norm": 1.0673625469207764, "learning_rate": 0.0002, "epoch": 3.1056503112117047, "step": 19210}, {"loss": 0.6043, "grad_norm": 0.7739115953445435, "learning_rate": 0.0002, "epoch": 3.10726699539245, "step": 19220}, {"loss": 0.6046, "grad_norm": 0.9884951114654541, "learning_rate": 0.0002, "epoch": 3.1088836795731956, "step": 19230}, {"loss": 0.5932, "grad_norm": 0.862260103225708, "learning_rate": 0.0002, "epoch": 3.110500363753941, "step": 19240}, {"loss": 0.6098, "grad_norm": 0.7690284848213196, "learning_rate": 0.0002, "epoch": 3.112117047934686, "step": 19250}, {"loss": 0.5791, "grad_norm": 0.8758958578109741, "learning_rate": 0.0002, "epoch": 3.1137337321154313, "step": 19260}, {"loss": 0.6136, "grad_norm": 1.0356395244598389, "learning_rate": 0.0002, "epoch": 3.1153504162961765, "step": 19270}, {"loss": 0.6159, "grad_norm": 0.6950937509536743, "learning_rate": 0.0002, "epoch": 3.1169671004769217, "step": 19280}, {"loss": 0.592, "grad_norm": 0.760998010635376, "learning_rate": 0.0002, "epoch": 3.118583784657667, "step": 19290}, {"loss": 0.575, "grad_norm": 0.9335789084434509, "learning_rate": 0.0002, "epoch": 3.1202004688384126, "step": 19300}, {"loss": 0.6139, "grad_norm": 0.9636204242706299, "learning_rate": 0.0002, "epoch": 3.121817153019158, "step": 19310}, {"loss": 0.6001, "grad_norm": 1.0820997953414917, "learning_rate": 0.0002, "epoch": 3.123433837199903, "step": 19320}, {"loss": 0.6542, "grad_norm": 0.7333487272262573, "learning_rate": 0.0002, "epoch": 3.1250505213806483, "step": 19330}, {"loss": 0.6178, "grad_norm": 1.0417509078979492, "learning_rate": 0.0002, "epoch": 3.1266672055613935, "step": 19340}, {"loss": 0.603, "grad_norm": 0.9267749190330505, "learning_rate": 0.0002, "epoch": 3.128283889742139, "step": 19350}, {"loss": 0.6063, "grad_norm": 0.777798593044281, "learning_rate": 0.0002, "epoch": 3.129900573922884, "step": 19360}, {"loss": 0.5913, "grad_norm": 0.8425456881523132, "learning_rate": 0.0002, "epoch": 3.1315172581036297, "step": 19370}, {"loss": 0.6042, "grad_norm": 0.9617102146148682, "learning_rate": 0.0002, "epoch": 3.133133942284375, "step": 19380}, {"loss": 0.633, "grad_norm": 1.0052828788757324, "learning_rate": 0.0002, "epoch": 3.13475062646512, "step": 19390}, {"loss": 0.5713, "grad_norm": 0.7637009024620056, "learning_rate": 0.0002, "epoch": 3.1363673106458654, "step": 19400}, {"loss": 0.5497, "grad_norm": 0.7958088517189026, "learning_rate": 0.0002, "epoch": 3.1379839948266106, "step": 19410}, {"loss": 0.6283, "grad_norm": 0.9161727428436279, "learning_rate": 0.0002, "epoch": 3.139600679007356, "step": 19420}, {"loss": 0.5638, "grad_norm": 0.8402149677276611, "learning_rate": 0.0002, "epoch": 3.141217363188101, "step": 19430}, {"loss": 0.5848, "grad_norm": 1.0056525468826294, "learning_rate": 0.0002, "epoch": 3.1428340473688463, "step": 19440}, {"loss": 0.5954, "grad_norm": 1.0129190683364868, "learning_rate": 0.0002, "epoch": 3.144450731549592, "step": 19450}, {"loss": 0.5808, "grad_norm": 0.790825366973877, "learning_rate": 0.0002, "epoch": 3.146067415730337, "step": 19460}, {"loss": 0.5607, "grad_norm": 1.441665530204773, "learning_rate": 0.0002, "epoch": 3.1476840999110824, "step": 19470}, {"loss": 0.5785, "grad_norm": 0.7846331596374512, "learning_rate": 0.0002, "epoch": 3.1493007840918277, "step": 19480}, {"loss": 0.5892, "grad_norm": 0.7915332913398743, "learning_rate": 0.0002, "epoch": 3.150917468272573, "step": 19490}, {"loss": 0.5759, "grad_norm": 0.933982253074646, "learning_rate": 0.0002, "epoch": 3.152534152453318, "step": 19500}, {"loss": 0.6206, "grad_norm": 1.038408637046814, "learning_rate": 0.0002, "epoch": 3.1541508366340634, "step": 19510}, {"loss": 0.6271, "grad_norm": 1.018935203552246, "learning_rate": 0.0002, "epoch": 3.155767520814809, "step": 19520}, {"loss": 0.6173, "grad_norm": 0.9618112444877625, "learning_rate": 0.0002, "epoch": 3.1573842049955543, "step": 19530}, {"loss": 0.5972, "grad_norm": 0.8900452852249146, "learning_rate": 0.0002, "epoch": 3.1590008891762995, "step": 19540}, {"loss": 0.5925, "grad_norm": 0.8254160284996033, "learning_rate": 0.0002, "epoch": 3.1606175733570447, "step": 19550}, {"loss": 0.625, "grad_norm": 1.004376769065857, "learning_rate": 0.0002, "epoch": 3.16223425753779, "step": 19560}, {"loss": 0.5775, "grad_norm": 1.0490446090698242, "learning_rate": 0.0002, "epoch": 3.163850941718535, "step": 19570}, {"loss": 0.5986, "grad_norm": 0.7387403845787048, "learning_rate": 0.0002, "epoch": 3.1654676258992804, "step": 19580}, {"loss": 0.5898, "grad_norm": 0.7611538171768188, "learning_rate": 0.0002, "epoch": 3.1670843100800257, "step": 19590}, {"loss": 0.5937, "grad_norm": 0.8239886164665222, "learning_rate": 0.0002, "epoch": 3.1687009942607713, "step": 19600}, {"loss": 0.6068, "grad_norm": 0.9327243566513062, "learning_rate": 0.0002, "epoch": 3.1703176784415166, "step": 19610}, {"loss": 0.572, "grad_norm": 0.9662560224533081, "learning_rate": 0.0002, "epoch": 3.171934362622262, "step": 19620}, {"loss": 0.5988, "grad_norm": 0.9183341860771179, "learning_rate": 0.0002, "epoch": 3.173551046803007, "step": 19630}, {"loss": 0.5909, "grad_norm": 0.875066876411438, "learning_rate": 0.0002, "epoch": 3.1751677309837523, "step": 19640}, {"loss": 0.5956, "grad_norm": 0.8567508459091187, "learning_rate": 0.0002, "epoch": 3.1767844151644975, "step": 19650}, {"loss": 0.5805, "grad_norm": 0.6805780529975891, "learning_rate": 0.0002, "epoch": 3.1784010993452427, "step": 19660}, {"loss": 0.6204, "grad_norm": 0.8776944279670715, "learning_rate": 0.0002, "epoch": 3.1800177835259884, "step": 19670}, {"loss": 0.6108, "grad_norm": 0.9036329984664917, "learning_rate": 0.0002, "epoch": 3.1816344677067336, "step": 19680}, {"loss": 0.6238, "grad_norm": 0.8527372479438782, "learning_rate": 0.0002, "epoch": 3.183251151887479, "step": 19690}, {"loss": 0.6089, "grad_norm": 1.1045585870742798, "learning_rate": 0.0002, "epoch": 3.184867836068224, "step": 19700}, {"loss": 0.5491, "grad_norm": 0.9213830828666687, "learning_rate": 0.0002, "epoch": 3.1864845202489693, "step": 19710}, {"loss": 0.618, "grad_norm": 0.8865814805030823, "learning_rate": 0.0002, "epoch": 3.1881012044297146, "step": 19720}, {"loss": 0.5785, "grad_norm": 0.7939388751983643, "learning_rate": 0.0002, "epoch": 3.18971788861046, "step": 19730}, {"loss": 0.5682, "grad_norm": 0.6966729760169983, "learning_rate": 0.0002, "epoch": 3.191334572791205, "step": 19740}, {"loss": 0.5839, "grad_norm": 0.8023673295974731, "learning_rate": 0.0002, "epoch": 3.1929512569719507, "step": 19750}, {"loss": 0.6267, "grad_norm": 0.7992037534713745, "learning_rate": 0.0002, "epoch": 3.194567941152696, "step": 19760}, {"loss": 0.6141, "grad_norm": 0.7412247657775879, "learning_rate": 0.0002, "epoch": 3.196184625333441, "step": 19770}, {"loss": 0.6179, "grad_norm": 0.9598729014396667, "learning_rate": 0.0002, "epoch": 3.1978013095141864, "step": 19780}, {"loss": 0.5685, "grad_norm": 0.8331366777420044, "learning_rate": 0.0002, "epoch": 3.1994179936949316, "step": 19790}, {"loss": 0.6104, "grad_norm": 0.8939169645309448, "learning_rate": 0.0002, "epoch": 3.201034677875677, "step": 19800}, {"loss": 0.6147, "grad_norm": 0.9219734072685242, "learning_rate": 0.0002, "epoch": 3.202651362056422, "step": 19810}, {"loss": 0.6051, "grad_norm": 0.869490385055542, "learning_rate": 0.0002, "epoch": 3.2042680462371678, "step": 19820}, {"loss": 0.5946, "grad_norm": 0.8989706635475159, "learning_rate": 0.0002, "epoch": 3.205884730417913, "step": 19830}, {"loss": 0.5866, "grad_norm": 0.8477165102958679, "learning_rate": 0.0002, "epoch": 3.2075014145986582, "step": 19840}, {"loss": 0.6176, "grad_norm": 0.8720678687095642, "learning_rate": 0.0002, "epoch": 3.2091180987794035, "step": 19850}, {"loss": 0.5694, "grad_norm": 0.861406683921814, "learning_rate": 0.0002, "epoch": 3.2107347829601487, "step": 19860}, {"loss": 0.6264, "grad_norm": 0.8228686451911926, "learning_rate": 0.0002, "epoch": 3.212351467140894, "step": 19870}, {"loss": 0.625, "grad_norm": 0.7936596870422363, "learning_rate": 0.0002, "epoch": 3.213968151321639, "step": 19880}, {"loss": 0.5698, "grad_norm": 1.097377896308899, "learning_rate": 0.0002, "epoch": 3.2155848355023844, "step": 19890}, {"loss": 0.6725, "grad_norm": 0.9544782638549805, "learning_rate": 0.0002, "epoch": 3.21720151968313, "step": 19900}, {"loss": 0.6022, "grad_norm": 0.8240751624107361, "learning_rate": 0.0002, "epoch": 3.2188182038638753, "step": 19910}, {"loss": 0.5659, "grad_norm": 0.8332096338272095, "learning_rate": 0.0002, "epoch": 3.2204348880446205, "step": 19920}, {"loss": 0.6274, "grad_norm": 1.0954567193984985, "learning_rate": 0.0002, "epoch": 3.2220515722253658, "step": 19930}, {"loss": 0.652, "grad_norm": 0.7790525555610657, "learning_rate": 0.0002, "epoch": 3.223668256406111, "step": 19940}, {"loss": 0.5986, "grad_norm": 0.7966814041137695, "learning_rate": 0.0002, "epoch": 3.225284940586856, "step": 19950}, {"loss": 0.5911, "grad_norm": 0.9751881957054138, "learning_rate": 0.0002, "epoch": 3.2269016247676015, "step": 19960}, {"loss": 0.6071, "grad_norm": 0.9856047630310059, "learning_rate": 0.0002, "epoch": 3.228518308948347, "step": 19970}, {"loss": 0.5837, "grad_norm": 1.3062353134155273, "learning_rate": 0.0002, "epoch": 3.2301349931290924, "step": 19980}, {"loss": 0.6588, "grad_norm": 0.9510692358016968, "learning_rate": 0.0002, "epoch": 3.2317516773098376, "step": 19990}, {"loss": 0.6264, "grad_norm": 0.8630342483520508, "learning_rate": 0.0002, "epoch": 3.233368361490583, "step": 20000}, {"loss": 0.6073, "grad_norm": 0.8966519236564636, "learning_rate": 0.0002, "epoch": 3.234985045671328, "step": 20010}, {"loss": 0.612, "grad_norm": 0.7093510627746582, "learning_rate": 0.0002, "epoch": 3.2366017298520733, "step": 20020}, {"loss": 0.585, "grad_norm": 0.7771096229553223, "learning_rate": 0.0002, "epoch": 3.2382184140328185, "step": 20030}, {"loss": 0.5821, "grad_norm": 0.841058075428009, "learning_rate": 0.0002, "epoch": 3.2398350982135637, "step": 20040}, {"loss": 0.6519, "grad_norm": 0.909712553024292, "learning_rate": 0.0002, "epoch": 3.2414517823943094, "step": 20050}, {"loss": 0.6089, "grad_norm": 0.8321019411087036, "learning_rate": 0.0002, "epoch": 3.2430684665750547, "step": 20060}, {"loss": 0.6115, "grad_norm": 0.779901921749115, "learning_rate": 0.0002, "epoch": 3.2446851507558, "step": 20070}, {"loss": 0.6107, "grad_norm": 0.6249170303344727, "learning_rate": 0.0002, "epoch": 3.246301834936545, "step": 20080}, {"loss": 0.603, "grad_norm": 0.8000940680503845, "learning_rate": 0.0002, "epoch": 3.2479185191172903, "step": 20090}, {"loss": 0.6273, "grad_norm": 0.7627735137939453, "learning_rate": 0.0002, "epoch": 3.2495352032980356, "step": 20100}, {"loss": 0.6223, "grad_norm": 0.8780747056007385, "learning_rate": 0.0002, "epoch": 3.2511518874787813, "step": 20110}, {"loss": 0.5969, "grad_norm": 0.772037148475647, "learning_rate": 0.0002, "epoch": 3.2527685716595265, "step": 20120}, {"loss": 0.5843, "grad_norm": 1.0086580514907837, "learning_rate": 0.0002, "epoch": 3.2543852558402717, "step": 20130}, {"loss": 0.5777, "grad_norm": 0.9360289573669434, "learning_rate": 0.0002, "epoch": 3.256001940021017, "step": 20140}, {"loss": 0.5777, "grad_norm": 1.2099586725234985, "learning_rate": 0.0002, "epoch": 3.257618624201762, "step": 20150}, {"loss": 0.624, "grad_norm": 0.8368481397628784, "learning_rate": 0.0002, "epoch": 3.2592353083825074, "step": 20160}, {"loss": 0.5626, "grad_norm": 0.7391039133071899, "learning_rate": 0.0002, "epoch": 3.2608519925632526, "step": 20170}, {"loss": 0.6041, "grad_norm": 0.9122273325920105, "learning_rate": 0.0002, "epoch": 3.262468676743998, "step": 20180}, {"loss": 0.5868, "grad_norm": 0.8502281904220581, "learning_rate": 0.0002, "epoch": 3.264085360924743, "step": 20190}, {"loss": 0.5841, "grad_norm": 1.0926852226257324, "learning_rate": 0.0002, "epoch": 3.265702045105489, "step": 20200}, {"loss": 0.6027, "grad_norm": 0.7902828454971313, "learning_rate": 0.0002, "epoch": 3.267318729286234, "step": 20210}, {"loss": 0.6089, "grad_norm": 0.8724729418754578, "learning_rate": 0.0002, "epoch": 3.2689354134669792, "step": 20220}, {"loss": 0.6242, "grad_norm": 0.8469277024269104, "learning_rate": 0.0002, "epoch": 3.2705520976477245, "step": 20230}, {"loss": 0.644, "grad_norm": 0.8865092992782593, "learning_rate": 0.0002, "epoch": 3.2721687818284697, "step": 20240}, {"loss": 0.6464, "grad_norm": 1.0979334115982056, "learning_rate": 0.0002, "epoch": 3.273785466009215, "step": 20250}, {"loss": 0.647, "grad_norm": 1.0860793590545654, "learning_rate": 0.0002, "epoch": 3.2754021501899606, "step": 20260}, {"loss": 0.6105, "grad_norm": 0.981745183467865, "learning_rate": 0.0002, "epoch": 3.277018834370706, "step": 20270}, {"loss": 0.627, "grad_norm": 0.9155020713806152, "learning_rate": 0.0002, "epoch": 3.278635518551451, "step": 20280}, {"loss": 0.5899, "grad_norm": 0.8436718583106995, "learning_rate": 0.0002, "epoch": 3.2802522027321963, "step": 20290}, {"loss": 0.6371, "grad_norm": 1.0329409837722778, "learning_rate": 0.0002, "epoch": 3.2818688869129415, "step": 20300}, {"loss": 0.6, "grad_norm": 0.9876394271850586, "learning_rate": 0.0002, "epoch": 3.2834855710936868, "step": 20310}, {"loss": 0.5463, "grad_norm": 0.8052917718887329, "learning_rate": 0.0002, "epoch": 3.285102255274432, "step": 20320}, {"loss": 0.5949, "grad_norm": 0.8390680551528931, "learning_rate": 0.0002, "epoch": 3.2867189394551772, "step": 20330}, {"loss": 0.6492, "grad_norm": 0.9515735507011414, "learning_rate": 0.0002, "epoch": 3.288335623635923, "step": 20340}, {"loss": 0.596, "grad_norm": 0.8028870224952698, "learning_rate": 0.0002, "epoch": 3.289952307816668, "step": 20350}, {"loss": 0.634, "grad_norm": 0.862592339515686, "learning_rate": 0.0002, "epoch": 3.2915689919974134, "step": 20360}, {"loss": 0.6345, "grad_norm": 0.7451621890068054, "learning_rate": 0.0002, "epoch": 3.2931856761781586, "step": 20370}, {"loss": 0.6458, "grad_norm": 0.8966776728630066, "learning_rate": 0.0002, "epoch": 3.294802360358904, "step": 20380}, {"loss": 0.5967, "grad_norm": 0.9289216995239258, "learning_rate": 0.0002, "epoch": 3.296419044539649, "step": 20390}, {"loss": 0.6599, "grad_norm": 0.9649626612663269, "learning_rate": 0.0002, "epoch": 3.2980357287203943, "step": 20400}, {"loss": 0.5781, "grad_norm": 1.1953798532485962, "learning_rate": 0.0002, "epoch": 3.29965241290114, "step": 20410}, {"loss": 0.5997, "grad_norm": 0.8929083943367004, "learning_rate": 0.0002, "epoch": 3.301269097081885, "step": 20420}, {"loss": 0.597, "grad_norm": 0.8922014236450195, "learning_rate": 0.0002, "epoch": 3.3028857812626304, "step": 20430}, {"loss": 0.5766, "grad_norm": 0.9754860401153564, "learning_rate": 0.0002, "epoch": 3.3045024654433757, "step": 20440}, {"loss": 0.5653, "grad_norm": 0.8873140215873718, "learning_rate": 0.0002, "epoch": 3.306119149624121, "step": 20450}, {"loss": 0.6138, "grad_norm": 0.857271671295166, "learning_rate": 0.0002, "epoch": 3.307735833804866, "step": 20460}, {"loss": 0.633, "grad_norm": 0.9022141098976135, "learning_rate": 0.0002, "epoch": 3.3093525179856114, "step": 20470}, {"loss": 0.6654, "grad_norm": 0.8614798188209534, "learning_rate": 0.0002, "epoch": 3.3109692021663566, "step": 20480}, {"loss": 0.6254, "grad_norm": 0.8838164210319519, "learning_rate": 0.0002, "epoch": 3.3125858863471023, "step": 20490}, {"loss": 0.5849, "grad_norm": 0.8709736466407776, "learning_rate": 0.0002, "epoch": 3.3142025705278475, "step": 20500}, {"loss": 0.6146, "grad_norm": 0.9533300995826721, "learning_rate": 0.0002, "epoch": 3.3158192547085927, "step": 20510}, {"loss": 0.6029, "grad_norm": 0.8259269595146179, "learning_rate": 0.0002, "epoch": 3.317435938889338, "step": 20520}, {"loss": 0.6268, "grad_norm": 0.8607608079910278, "learning_rate": 0.0002, "epoch": 3.319052623070083, "step": 20530}, {"loss": 0.5676, "grad_norm": 1.0863020420074463, "learning_rate": 0.0002, "epoch": 3.3206693072508284, "step": 20540}, {"loss": 0.6412, "grad_norm": 1.011489987373352, "learning_rate": 0.0002, "epoch": 3.3222859914315737, "step": 20550}, {"loss": 0.6247, "grad_norm": 0.6952177882194519, "learning_rate": 0.0002, "epoch": 3.3239026756123193, "step": 20560}, {"loss": 0.6229, "grad_norm": 0.9638974070549011, "learning_rate": 0.0002, "epoch": 3.3255193597930646, "step": 20570}, {"loss": 0.5882, "grad_norm": 1.0310138463974, "learning_rate": 0.0002, "epoch": 3.32713604397381, "step": 20580}, {"loss": 0.594, "grad_norm": 0.9371318221092224, "learning_rate": 0.0002, "epoch": 3.328752728154555, "step": 20590}, {"loss": 0.6137, "grad_norm": 0.8756691813468933, "learning_rate": 0.0002, "epoch": 3.3303694123353003, "step": 20600}, {"loss": 0.5994, "grad_norm": 1.054175853729248, "learning_rate": 0.0002, "epoch": 3.3319860965160455, "step": 20610}, {"loss": 0.6169, "grad_norm": 0.9074128270149231, "learning_rate": 0.0002, "epoch": 3.3336027806967907, "step": 20620}, {"loss": 0.6138, "grad_norm": 0.906900942325592, "learning_rate": 0.0002, "epoch": 3.335219464877536, "step": 20630}, {"loss": 0.571, "grad_norm": 0.8689333200454712, "learning_rate": 0.0002, "epoch": 3.3368361490582816, "step": 20640}, {"loss": 0.6079, "grad_norm": 0.9889747500419617, "learning_rate": 0.0002, "epoch": 3.338452833239027, "step": 20650}, {"loss": 0.6073, "grad_norm": 1.0685805082321167, "learning_rate": 0.0002, "epoch": 3.340069517419772, "step": 20660}, {"loss": 0.6091, "grad_norm": 0.7495010495185852, "learning_rate": 0.0002, "epoch": 3.3416862016005173, "step": 20670}, {"loss": 0.5883, "grad_norm": 0.8747848272323608, "learning_rate": 0.0002, "epoch": 3.3433028857812626, "step": 20680}, {"loss": 0.604, "grad_norm": 0.9762673377990723, "learning_rate": 0.0002, "epoch": 3.344919569962008, "step": 20690}, {"loss": 0.6784, "grad_norm": 1.0284489393234253, "learning_rate": 0.0002, "epoch": 3.346536254142753, "step": 20700}, {"loss": 0.6464, "grad_norm": 0.7293812036514282, "learning_rate": 0.0002, "epoch": 3.3481529383234987, "step": 20710}, {"loss": 0.609, "grad_norm": 0.8330199122428894, "learning_rate": 0.0002, "epoch": 3.349769622504244, "step": 20720}, {"loss": 0.5729, "grad_norm": 0.9808499217033386, "learning_rate": 0.0002, "epoch": 3.351386306684989, "step": 20730}, {"loss": 0.6315, "grad_norm": 0.9508825540542603, "learning_rate": 0.0002, "epoch": 3.3530029908657344, "step": 20740}, {"loss": 0.5965, "grad_norm": 0.790483832359314, "learning_rate": 0.0002, "epoch": 3.3546196750464796, "step": 20750}, {"loss": 0.6327, "grad_norm": 1.022793173789978, "learning_rate": 0.0002, "epoch": 3.356236359227225, "step": 20760}, {"loss": 0.6439, "grad_norm": 0.8318950533866882, "learning_rate": 0.0002, "epoch": 3.35785304340797, "step": 20770}, {"loss": 0.6037, "grad_norm": 0.7980858087539673, "learning_rate": 0.0002, "epoch": 3.3594697275887153, "step": 20780}, {"loss": 0.6746, "grad_norm": 0.8114802241325378, "learning_rate": 0.0002, "epoch": 3.361086411769461, "step": 20790}, {"loss": 0.6017, "grad_norm": 0.8522519469261169, "learning_rate": 0.0002, "epoch": 3.3627030959502062, "step": 20800}, {"loss": 0.5864, "grad_norm": 0.9142431616783142, "learning_rate": 0.0002, "epoch": 3.3643197801309515, "step": 20810}, {"loss": 0.6331, "grad_norm": 0.771170437335968, "learning_rate": 0.0002, "epoch": 3.3659364643116967, "step": 20820}, {"loss": 0.5879, "grad_norm": 1.0628231763839722, "learning_rate": 0.0002, "epoch": 3.367553148492442, "step": 20830}, {"loss": 0.6533, "grad_norm": 0.9384352564811707, "learning_rate": 0.0002, "epoch": 3.369169832673187, "step": 20840}, {"loss": 0.6292, "grad_norm": 1.1286591291427612, "learning_rate": 0.0002, "epoch": 3.370786516853933, "step": 20850}, {"loss": 0.5986, "grad_norm": 1.1349513530731201, "learning_rate": 0.0002, "epoch": 3.372403201034678, "step": 20860}, {"loss": 0.6413, "grad_norm": 1.0127464532852173, "learning_rate": 0.0002, "epoch": 3.3740198852154233, "step": 20870}, {"loss": 0.6414, "grad_norm": 0.9111971855163574, "learning_rate": 0.0002, "epoch": 3.3756365693961685, "step": 20880}, {"loss": 0.6101, "grad_norm": 0.871356725692749, "learning_rate": 0.0002, "epoch": 3.3772532535769137, "step": 20890}, {"loss": 0.5995, "grad_norm": 0.7774117588996887, "learning_rate": 0.0002, "epoch": 3.378869937757659, "step": 20900}, {"loss": 0.6062, "grad_norm": 1.0089964866638184, "learning_rate": 0.0002, "epoch": 3.380486621938404, "step": 20910}, {"loss": 0.5908, "grad_norm": 0.7855867147445679, "learning_rate": 0.0002, "epoch": 3.3821033061191494, "step": 20920}, {"loss": 0.6373, "grad_norm": 1.3713710308074951, "learning_rate": 0.0002, "epoch": 3.3837199902998947, "step": 20930}, {"loss": 0.6627, "grad_norm": 0.8599116206169128, "learning_rate": 0.0002, "epoch": 3.3853366744806404, "step": 20940}, {"loss": 0.6224, "grad_norm": 0.9392673373222351, "learning_rate": 0.0002, "epoch": 3.3869533586613856, "step": 20950}, {"loss": 0.5855, "grad_norm": 0.8764075040817261, "learning_rate": 0.0002, "epoch": 3.388570042842131, "step": 20960}, {"loss": 0.5734, "grad_norm": 0.8240136504173279, "learning_rate": 0.0002, "epoch": 3.390186727022876, "step": 20970}, {"loss": 0.5783, "grad_norm": 1.0982369184494019, "learning_rate": 0.0002, "epoch": 3.3918034112036213, "step": 20980}, {"loss": 0.5451, "grad_norm": 1.0599013566970825, "learning_rate": 0.0002, "epoch": 3.3934200953843665, "step": 20990}, {"loss": 0.6356, "grad_norm": 0.895438015460968, "learning_rate": 0.0002, "epoch": 3.395036779565112, "step": 21000}, {"loss": 0.6065, "grad_norm": 0.6974841356277466, "learning_rate": 0.0002, "epoch": 3.3966534637458574, "step": 21010}, {"loss": 0.5704, "grad_norm": 0.9571719765663147, "learning_rate": 0.0002, "epoch": 3.3982701479266026, "step": 21020}, {"loss": 0.679, "grad_norm": 0.831912636756897, "learning_rate": 0.0002, "epoch": 3.399886832107348, "step": 21030}, {"loss": 0.6051, "grad_norm": 0.831936240196228, "learning_rate": 0.0002, "epoch": 3.401503516288093, "step": 21040}, {"loss": 0.5857, "grad_norm": 0.7388373613357544, "learning_rate": 0.0002, "epoch": 3.4031202004688383, "step": 21050}, {"loss": 0.6245, "grad_norm": 0.938667356967926, "learning_rate": 0.0002, "epoch": 3.4047368846495836, "step": 21060}, {"loss": 0.6121, "grad_norm": 0.9202313423156738, "learning_rate": 0.0002, "epoch": 3.406353568830329, "step": 21070}, {"loss": 0.6388, "grad_norm": 0.9888381958007812, "learning_rate": 0.0002, "epoch": 3.4079702530110745, "step": 21080}, {"loss": 0.6245, "grad_norm": 0.8526970744132996, "learning_rate": 0.0002, "epoch": 3.4095869371918197, "step": 21090}, {"loss": 0.5914, "grad_norm": 0.7939383387565613, "learning_rate": 0.0002, "epoch": 3.411203621372565, "step": 21100}, {"loss": 0.6066, "grad_norm": 0.9986352920532227, "learning_rate": 0.0002, "epoch": 3.41282030555331, "step": 21110}, {"loss": 0.5947, "grad_norm": 0.8895300030708313, "learning_rate": 0.0002, "epoch": 3.4144369897340554, "step": 21120}, {"loss": 0.6264, "grad_norm": 0.9559482932090759, "learning_rate": 0.0002, "epoch": 3.4160536739148006, "step": 21130}, {"loss": 0.6491, "grad_norm": 0.8351506590843201, "learning_rate": 0.0002, "epoch": 3.417670358095546, "step": 21140}, {"loss": 0.567, "grad_norm": 0.8224456906318665, "learning_rate": 0.0002, "epoch": 3.4192870422762915, "step": 21150}, {"loss": 0.5871, "grad_norm": 1.0110299587249756, "learning_rate": 0.0002, "epoch": 3.4209037264570368, "step": 21160}, {"loss": 0.6116, "grad_norm": 0.82564777135849, "learning_rate": 0.0002, "epoch": 3.422520410637782, "step": 21170}, {"loss": 0.595, "grad_norm": 1.004738688468933, "learning_rate": 0.0002, "epoch": 3.4241370948185272, "step": 21180}, {"loss": 0.6286, "grad_norm": 0.7545676827430725, "learning_rate": 0.0002, "epoch": 3.4257537789992725, "step": 21190}, {"loss": 0.5868, "grad_norm": 0.8918704390525818, "learning_rate": 0.0002, "epoch": 3.4273704631800177, "step": 21200}, {"loss": 0.6542, "grad_norm": 0.8336876034736633, "learning_rate": 0.0002, "epoch": 3.428987147360763, "step": 21210}, {"loss": 0.5824, "grad_norm": 0.8928771018981934, "learning_rate": 0.0002, "epoch": 3.430603831541508, "step": 21220}, {"loss": 0.6468, "grad_norm": 0.7663705945014954, "learning_rate": 0.0002, "epoch": 3.432220515722254, "step": 21230}, {"loss": 0.6693, "grad_norm": 0.8392598628997803, "learning_rate": 0.0002, "epoch": 3.433837199902999, "step": 21240}, {"loss": 0.5971, "grad_norm": 0.8819600343704224, "learning_rate": 0.0002, "epoch": 3.4354538840837443, "step": 21250}, {"loss": 0.6791, "grad_norm": 0.9124642014503479, "learning_rate": 0.0002, "epoch": 3.4370705682644895, "step": 21260}, {"loss": 0.5925, "grad_norm": 0.8329763412475586, "learning_rate": 0.0002, "epoch": 3.4386872524452348, "step": 21270}, {"loss": 0.6541, "grad_norm": 0.9982839822769165, "learning_rate": 0.0002, "epoch": 3.44030393662598, "step": 21280}, {"loss": 0.6441, "grad_norm": 0.9105954766273499, "learning_rate": 0.0002, "epoch": 3.4419206208067252, "step": 21290}, {"loss": 0.6028, "grad_norm": 0.8182359337806702, "learning_rate": 0.0002, "epoch": 3.443537304987471, "step": 21300}, {"loss": 0.5991, "grad_norm": 1.0568904876708984, "learning_rate": 0.0002, "epoch": 3.445153989168216, "step": 21310}, {"loss": 0.6117, "grad_norm": 0.968539834022522, "learning_rate": 0.0002, "epoch": 3.4467706733489614, "step": 21320}, {"loss": 0.6219, "grad_norm": 0.8774511218070984, "learning_rate": 0.0002, "epoch": 3.4483873575297066, "step": 21330}, {"loss": 0.6438, "grad_norm": 0.7598156332969666, "learning_rate": 0.0002, "epoch": 3.450004041710452, "step": 21340}, {"loss": 0.6033, "grad_norm": 1.1012897491455078, "learning_rate": 0.0002, "epoch": 3.451620725891197, "step": 21350}, {"loss": 0.6137, "grad_norm": 0.8040637373924255, "learning_rate": 0.0002, "epoch": 3.4532374100719423, "step": 21360}, {"loss": 0.6173, "grad_norm": 0.8497496247291565, "learning_rate": 0.0002, "epoch": 3.4548540942526875, "step": 21370}, {"loss": 0.6005, "grad_norm": 0.8429915904998779, "learning_rate": 0.0002, "epoch": 3.456470778433433, "step": 21380}, {"loss": 0.6182, "grad_norm": 0.8107112646102905, "learning_rate": 0.0002, "epoch": 3.4580874626141784, "step": 21390}, {"loss": 0.6109, "grad_norm": 1.00872004032135, "learning_rate": 0.0002, "epoch": 3.4597041467949237, "step": 21400}, {"loss": 0.5712, "grad_norm": 0.8266542553901672, "learning_rate": 0.0002, "epoch": 3.461320830975669, "step": 21410}, {"loss": 0.6457, "grad_norm": 0.8972568511962891, "learning_rate": 0.0002, "epoch": 3.462937515156414, "step": 21420}, {"loss": 0.6081, "grad_norm": 1.0781476497650146, "learning_rate": 0.0002, "epoch": 3.4645541993371594, "step": 21430}, {"loss": 0.6303, "grad_norm": 0.9571592807769775, "learning_rate": 0.0002, "epoch": 3.4661708835179046, "step": 21440}, {"loss": 0.6309, "grad_norm": 0.881547212600708, "learning_rate": 0.0002, "epoch": 3.4677875676986503, "step": 21450}, {"loss": 0.6076, "grad_norm": 0.6955338716506958, "learning_rate": 0.0002, "epoch": 3.4694042518793955, "step": 21460}, {"loss": 0.6205, "grad_norm": 0.901187539100647, "learning_rate": 0.0002, "epoch": 3.4710209360601407, "step": 21470}, {"loss": 0.639, "grad_norm": 0.7063511610031128, "learning_rate": 0.0002, "epoch": 3.472637620240886, "step": 21480}, {"loss": 0.6154, "grad_norm": 0.8462792038917542, "learning_rate": 0.0002, "epoch": 3.474254304421631, "step": 21490}, {"loss": 0.61, "grad_norm": 1.1861060857772827, "learning_rate": 0.0002, "epoch": 3.4758709886023764, "step": 21500}, {"loss": 0.6586, "grad_norm": 0.70503169298172, "learning_rate": 0.0002, "epoch": 3.4774876727831217, "step": 21510}, {"loss": 0.6475, "grad_norm": 0.9650066494941711, "learning_rate": 0.0002, "epoch": 3.479104356963867, "step": 21520}, {"loss": 0.6452, "grad_norm": 1.0266852378845215, "learning_rate": 0.0002, "epoch": 3.4807210411446126, "step": 21530}, {"loss": 0.6553, "grad_norm": 0.956372857093811, "learning_rate": 0.0002, "epoch": 3.482337725325358, "step": 21540}, {"loss": 0.6667, "grad_norm": 0.8848432898521423, "learning_rate": 0.0002, "epoch": 3.483954409506103, "step": 21550}, {"loss": 0.6375, "grad_norm": 1.0805351734161377, "learning_rate": 0.0002, "epoch": 3.4855710936868483, "step": 21560}, {"loss": 0.6958, "grad_norm": 0.9279725551605225, "learning_rate": 0.0002, "epoch": 3.4871877778675935, "step": 21570}, {"loss": 0.6354, "grad_norm": 0.9049562215805054, "learning_rate": 0.0002, "epoch": 3.4888044620483387, "step": 21580}, {"loss": 0.6071, "grad_norm": 0.9619429111480713, "learning_rate": 0.0002, "epoch": 3.4904211462290844, "step": 21590}, {"loss": 0.5927, "grad_norm": 0.8508906960487366, "learning_rate": 0.0002, "epoch": 3.4920378304098296, "step": 21600}, {"loss": 0.6115, "grad_norm": 0.8692502379417419, "learning_rate": 0.0002, "epoch": 3.493654514590575, "step": 21610}, {"loss": 0.5878, "grad_norm": 0.8187332153320312, "learning_rate": 0.0002, "epoch": 3.49527119877132, "step": 21620}, {"loss": 0.5874, "grad_norm": 1.145400047302246, "learning_rate": 0.0002, "epoch": 3.4968878829520653, "step": 21630}, {"loss": 0.6313, "grad_norm": 0.8281388282775879, "learning_rate": 0.0002, "epoch": 3.4985045671328105, "step": 21640}, {"loss": 0.6624, "grad_norm": 0.82256019115448, "learning_rate": 0.0002, "epoch": 3.500121251313556, "step": 21650}, {"loss": 0.6346, "grad_norm": 0.9315484762191772, "learning_rate": 0.0002, "epoch": 3.501737935494301, "step": 21660}, {"loss": 0.6086, "grad_norm": 0.7626111507415771, "learning_rate": 0.0002, "epoch": 3.5033546196750462, "step": 21670}, {"loss": 0.6177, "grad_norm": 0.9275059103965759, "learning_rate": 0.0002, "epoch": 3.504971303855792, "step": 21680}, {"loss": 0.64, "grad_norm": 0.7906724810600281, "learning_rate": 0.0002, "epoch": 3.506587988036537, "step": 21690}, {"loss": 0.6015, "grad_norm": 0.8289761543273926, "learning_rate": 0.0002, "epoch": 3.5082046722172824, "step": 21700}, {"loss": 0.6246, "grad_norm": 0.8316431045532227, "learning_rate": 0.0002, "epoch": 3.5098213563980276, "step": 21710}, {"loss": 0.619, "grad_norm": 1.0451812744140625, "learning_rate": 0.0002, "epoch": 3.511438040578773, "step": 21720}, {"loss": 0.632, "grad_norm": 0.928252637386322, "learning_rate": 0.0002, "epoch": 3.513054724759518, "step": 21730}, {"loss": 0.6062, "grad_norm": 0.7985895276069641, "learning_rate": 0.0002, "epoch": 3.5146714089402638, "step": 21740}, {"loss": 0.6463, "grad_norm": 0.6740974187850952, "learning_rate": 0.0002, "epoch": 3.516288093121009, "step": 21750}, {"loss": 0.6138, "grad_norm": 0.8482223749160767, "learning_rate": 0.0002, "epoch": 3.517904777301754, "step": 21760}, {"loss": 0.6277, "grad_norm": 0.889947772026062, "learning_rate": 0.0002, "epoch": 3.5195214614824994, "step": 21770}, {"loss": 0.6174, "grad_norm": 0.8304598927497864, "learning_rate": 0.0002, "epoch": 3.5211381456632447, "step": 21780}, {"loss": 0.6156, "grad_norm": 0.8002981543540955, "learning_rate": 0.0002, "epoch": 3.52275482984399, "step": 21790}, {"loss": 0.5896, "grad_norm": 0.8115083575248718, "learning_rate": 0.0002, "epoch": 3.524371514024735, "step": 21800}, {"loss": 0.6041, "grad_norm": 0.9715048670768738, "learning_rate": 0.0002, "epoch": 3.5259881982054804, "step": 21810}, {"loss": 0.6715, "grad_norm": 1.0910786390304565, "learning_rate": 0.0002, "epoch": 3.5276048823862256, "step": 21820}, {"loss": 0.6543, "grad_norm": 0.8438942432403564, "learning_rate": 0.0002, "epoch": 3.5292215665669713, "step": 21830}, {"loss": 0.6509, "grad_norm": 0.8813382983207703, "learning_rate": 0.0002, "epoch": 3.5308382507477165, "step": 21840}, {"loss": 0.6049, "grad_norm": 0.7092908024787903, "learning_rate": 0.0002, "epoch": 3.5324549349284617, "step": 21850}, {"loss": 0.5678, "grad_norm": 0.8332187533378601, "learning_rate": 0.0002, "epoch": 3.534071619109207, "step": 21860}, {"loss": 0.5896, "grad_norm": 0.8958209156990051, "learning_rate": 0.0002, "epoch": 3.535688303289952, "step": 21870}, {"loss": 0.6476, "grad_norm": 0.824138879776001, "learning_rate": 0.0002, "epoch": 3.5373049874706974, "step": 21880}, {"loss": 0.6022, "grad_norm": 0.8375158309936523, "learning_rate": 0.0002, "epoch": 3.538921671651443, "step": 21890}, {"loss": 0.6019, "grad_norm": 1.0274608135223389, "learning_rate": 0.0002, "epoch": 3.5405383558321883, "step": 21900}, {"loss": 0.6194, "grad_norm": 0.7088932394981384, "learning_rate": 0.0002, "epoch": 3.5421550400129336, "step": 21910}, {"loss": 0.6554, "grad_norm": 0.8172445297241211, "learning_rate": 0.0002, "epoch": 3.543771724193679, "step": 21920}, {"loss": 0.6711, "grad_norm": 0.9904135465621948, "learning_rate": 0.0002, "epoch": 3.545388408374424, "step": 21930}, {"loss": 0.6001, "grad_norm": 0.9900432229042053, "learning_rate": 0.0002, "epoch": 3.5470050925551693, "step": 21940}, {"loss": 0.6195, "grad_norm": 0.8963301181793213, "learning_rate": 0.0002, "epoch": 3.5486217767359145, "step": 21950}, {"loss": 0.5972, "grad_norm": 0.8551464676856995, "learning_rate": 0.0002, "epoch": 3.5502384609166597, "step": 21960}, {"loss": 0.6206, "grad_norm": 1.0916603803634644, "learning_rate": 0.0002, "epoch": 3.551855145097405, "step": 21970}, {"loss": 0.6523, "grad_norm": 0.841598391532898, "learning_rate": 0.0002, "epoch": 3.5534718292781506, "step": 21980}, {"loss": 0.617, "grad_norm": 0.8566757440567017, "learning_rate": 0.0002, "epoch": 3.555088513458896, "step": 21990}, {"loss": 0.6192, "grad_norm": 1.0145052671432495, "learning_rate": 0.0002, "epoch": 3.556705197639641, "step": 22000}, {"loss": 0.6173, "grad_norm": 0.9293754696846008, "learning_rate": 0.0002, "epoch": 3.5583218818203863, "step": 22010}, {"loss": 0.612, "grad_norm": 0.9568536281585693, "learning_rate": 0.0002, "epoch": 3.5599385660011316, "step": 22020}, {"loss": 0.641, "grad_norm": 0.8613139986991882, "learning_rate": 0.0002, "epoch": 3.5615552501818772, "step": 22030}, {"loss": 0.6496, "grad_norm": 0.8179237246513367, "learning_rate": 0.0002, "epoch": 3.5631719343626225, "step": 22040}, {"loss": 0.574, "grad_norm": 0.9059830904006958, "learning_rate": 0.0002, "epoch": 3.5647886185433677, "step": 22050}, {"loss": 0.6448, "grad_norm": 1.0068252086639404, "learning_rate": 0.0002, "epoch": 3.566405302724113, "step": 22060}, {"loss": 0.6239, "grad_norm": 0.9682072997093201, "learning_rate": 0.0002, "epoch": 3.568021986904858, "step": 22070}, {"loss": 0.6808, "grad_norm": 0.8514005541801453, "learning_rate": 0.0002, "epoch": 3.5696386710856034, "step": 22080}, {"loss": 0.5956, "grad_norm": 0.8327770829200745, "learning_rate": 0.0002, "epoch": 3.5712553552663486, "step": 22090}, {"loss": 0.5976, "grad_norm": 1.024976372718811, "learning_rate": 0.0002, "epoch": 3.572872039447094, "step": 22100}, {"loss": 0.624, "grad_norm": 0.7721174955368042, "learning_rate": 0.0002, "epoch": 3.574488723627839, "step": 22110}, {"loss": 0.5896, "grad_norm": 1.0351054668426514, "learning_rate": 0.0002, "epoch": 3.5761054078085843, "step": 22120}, {"loss": 0.6379, "grad_norm": 0.9680907130241394, "learning_rate": 0.0002, "epoch": 3.57772209198933, "step": 22130}, {"loss": 0.6194, "grad_norm": 0.8016974925994873, "learning_rate": 0.0002, "epoch": 3.5793387761700752, "step": 22140}, {"loss": 0.6387, "grad_norm": 1.0109003782272339, "learning_rate": 0.0002, "epoch": 3.5809554603508205, "step": 22150}, {"loss": 0.6368, "grad_norm": 1.0473392009735107, "learning_rate": 0.0002, "epoch": 3.5825721445315657, "step": 22160}, {"loss": 0.6353, "grad_norm": 0.8686613440513611, "learning_rate": 0.0002, "epoch": 3.584188828712311, "step": 22170}, {"loss": 0.5791, "grad_norm": 0.869149923324585, "learning_rate": 0.0002, "epoch": 3.5858055128930566, "step": 22180}, {"loss": 0.5895, "grad_norm": 0.9769062995910645, "learning_rate": 0.0002, "epoch": 3.587422197073802, "step": 22190}, {"loss": 0.5939, "grad_norm": 0.779636561870575, "learning_rate": 0.0002, "epoch": 3.589038881254547, "step": 22200}, {"loss": 0.5875, "grad_norm": 0.9063841104507446, "learning_rate": 0.0002, "epoch": 3.5906555654352923, "step": 22210}, {"loss": 0.5671, "grad_norm": 0.9216037392616272, "learning_rate": 0.0002, "epoch": 3.5922722496160375, "step": 22220}, {"loss": 0.6484, "grad_norm": 1.0217336416244507, "learning_rate": 0.0002, "epoch": 3.5938889337967828, "step": 22230}, {"loss": 0.6511, "grad_norm": 0.8513161540031433, "learning_rate": 0.0002, "epoch": 3.595505617977528, "step": 22240}, {"loss": 0.6301, "grad_norm": 0.8084813952445984, "learning_rate": 0.0002, "epoch": 3.597122302158273, "step": 22250}, {"loss": 0.6197, "grad_norm": 0.8524802923202515, "learning_rate": 0.0002, "epoch": 3.5987389863390185, "step": 22260}, {"loss": 0.5599, "grad_norm": 0.9356237649917603, "learning_rate": 0.0002, "epoch": 3.600355670519764, "step": 22270}, {"loss": 0.628, "grad_norm": 1.009600281715393, "learning_rate": 0.0002, "epoch": 3.6019723547005094, "step": 22280}, {"loss": 0.6179, "grad_norm": 0.9900581240653992, "learning_rate": 0.0002, "epoch": 3.6035890388812546, "step": 22290}, {"loss": 0.5725, "grad_norm": 1.062495231628418, "learning_rate": 0.0002, "epoch": 3.605205723062, "step": 22300}, {"loss": 0.607, "grad_norm": 0.8832381367683411, "learning_rate": 0.0002, "epoch": 3.606822407242745, "step": 22310}, {"loss": 0.6215, "grad_norm": 0.9284297823905945, "learning_rate": 0.0002, "epoch": 3.6084390914234903, "step": 22320}, {"loss": 0.685, "grad_norm": 1.2381829023361206, "learning_rate": 0.0002, "epoch": 3.610055775604236, "step": 22330}, {"loss": 0.6181, "grad_norm": 0.929434597492218, "learning_rate": 0.0002, "epoch": 3.611672459784981, "step": 22340}, {"loss": 0.6141, "grad_norm": 0.9714490175247192, "learning_rate": 0.0002, "epoch": 3.6132891439657264, "step": 22350}, {"loss": 0.6861, "grad_norm": 0.808014988899231, "learning_rate": 0.0002, "epoch": 3.6149058281464717, "step": 22360}, {"loss": 0.6428, "grad_norm": 1.0364398956298828, "learning_rate": 0.0002, "epoch": 3.616522512327217, "step": 22370}, {"loss": 0.6337, "grad_norm": 0.7858489751815796, "learning_rate": 0.0002, "epoch": 3.618139196507962, "step": 22380}, {"loss": 0.6214, "grad_norm": 0.9920870065689087, "learning_rate": 0.0002, "epoch": 3.6197558806887074, "step": 22390}, {"loss": 0.6659, "grad_norm": 0.9183220863342285, "learning_rate": 0.0002, "epoch": 3.6213725648694526, "step": 22400}, {"loss": 0.6036, "grad_norm": 0.9826246500015259, "learning_rate": 0.0002, "epoch": 3.622989249050198, "step": 22410}, {"loss": 0.6441, "grad_norm": 0.8632931113243103, "learning_rate": 0.0002, "epoch": 3.6246059332309435, "step": 22420}, {"loss": 0.6124, "grad_norm": 0.8468965291976929, "learning_rate": 0.0002, "epoch": 3.6262226174116887, "step": 22430}, {"loss": 0.6328, "grad_norm": 0.8466871976852417, "learning_rate": 0.0002, "epoch": 3.627839301592434, "step": 22440}, {"loss": 0.5941, "grad_norm": 0.9501169919967651, "learning_rate": 0.0002, "epoch": 3.629455985773179, "step": 22450}, {"loss": 0.6069, "grad_norm": 0.8906720876693726, "learning_rate": 0.0002, "epoch": 3.6310726699539244, "step": 22460}, {"loss": 0.6928, "grad_norm": 0.7400227189064026, "learning_rate": 0.0002, "epoch": 3.6326893541346696, "step": 22470}, {"loss": 0.6337, "grad_norm": 0.9756355881690979, "learning_rate": 0.0002, "epoch": 3.6343060383154153, "step": 22480}, {"loss": 0.6203, "grad_norm": 0.7504993081092834, "learning_rate": 0.0002, "epoch": 3.6359227224961606, "step": 22490}, {"loss": 0.6302, "grad_norm": 0.9270039200782776, "learning_rate": 0.0002, "epoch": 3.637539406676906, "step": 22500}, {"loss": 0.6026, "grad_norm": 0.8841686844825745, "learning_rate": 0.0002, "epoch": 3.639156090857651, "step": 22510}, {"loss": 0.6098, "grad_norm": 0.8533213138580322, "learning_rate": 0.0002, "epoch": 3.6407727750383962, "step": 22520}, {"loss": 0.6412, "grad_norm": 1.0052043199539185, "learning_rate": 0.0002, "epoch": 3.6423894592191415, "step": 22530}, {"loss": 0.6363, "grad_norm": 1.0323461294174194, "learning_rate": 0.0002, "epoch": 3.6440061433998867, "step": 22540}, {"loss": 0.6545, "grad_norm": 0.8654312491416931, "learning_rate": 0.0002, "epoch": 3.645622827580632, "step": 22550}, {"loss": 0.6155, "grad_norm": 0.6400038003921509, "learning_rate": 0.0002, "epoch": 3.647239511761377, "step": 22560}, {"loss": 0.5829, "grad_norm": 0.8061298727989197, "learning_rate": 0.0002, "epoch": 3.648856195942123, "step": 22570}, {"loss": 0.6388, "grad_norm": 0.9257854223251343, "learning_rate": 0.0002, "epoch": 3.650472880122868, "step": 22580}, {"loss": 0.6409, "grad_norm": 0.8439396619796753, "learning_rate": 0.0002, "epoch": 3.6520895643036133, "step": 22590}, {"loss": 0.5996, "grad_norm": 0.7764544486999512, "learning_rate": 0.0002, "epoch": 3.6537062484843585, "step": 22600}, {"loss": 0.6434, "grad_norm": 1.125451683998108, "learning_rate": 0.0002, "epoch": 3.6553229326651038, "step": 22610}, {"loss": 0.6579, "grad_norm": 0.7523018717765808, "learning_rate": 0.0002, "epoch": 3.656939616845849, "step": 22620}, {"loss": 0.6476, "grad_norm": 1.071026086807251, "learning_rate": 0.0002, "epoch": 3.6585563010265947, "step": 22630}, {"loss": 0.6459, "grad_norm": 0.945791482925415, "learning_rate": 0.0002, "epoch": 3.66017298520734, "step": 22640}, {"loss": 0.659, "grad_norm": 0.8001811504364014, "learning_rate": 0.0002, "epoch": 3.661789669388085, "step": 22650}, {"loss": 0.6385, "grad_norm": 0.9700816869735718, "learning_rate": 0.0002, "epoch": 3.6634063535688304, "step": 22660}, {"loss": 0.6337, "grad_norm": 0.9053242206573486, "learning_rate": 0.0002, "epoch": 3.6650230377495756, "step": 22670}, {"loss": 0.6335, "grad_norm": 0.944362461566925, "learning_rate": 0.0002, "epoch": 3.666639721930321, "step": 22680}, {"loss": 0.6235, "grad_norm": 1.067489504814148, "learning_rate": 0.0002, "epoch": 3.668256406111066, "step": 22690}, {"loss": 0.698, "grad_norm": 1.0984995365142822, "learning_rate": 0.0002, "epoch": 3.6698730902918113, "step": 22700}, {"loss": 0.6717, "grad_norm": 0.9336317777633667, "learning_rate": 0.0002, "epoch": 3.6714897744725565, "step": 22710}, {"loss": 0.6195, "grad_norm": 0.9261918663978577, "learning_rate": 0.0002, "epoch": 3.673106458653302, "step": 22720}, {"loss": 0.6332, "grad_norm": 0.8648008704185486, "learning_rate": 0.0002, "epoch": 3.6747231428340474, "step": 22730}, {"loss": 0.6576, "grad_norm": 0.7225083708763123, "learning_rate": 0.0002, "epoch": 3.6763398270147927, "step": 22740}, {"loss": 0.6406, "grad_norm": 0.9258282780647278, "learning_rate": 0.0002, "epoch": 3.677956511195538, "step": 22750}, {"loss": 0.6397, "grad_norm": 0.70876145362854, "learning_rate": 0.0002, "epoch": 3.679573195376283, "step": 22760}, {"loss": 0.6821, "grad_norm": 0.8780210018157959, "learning_rate": 0.0002, "epoch": 3.681189879557029, "step": 22770}, {"loss": 0.6036, "grad_norm": 0.8075440526008606, "learning_rate": 0.0002, "epoch": 3.682806563737774, "step": 22780}, {"loss": 0.6561, "grad_norm": 0.8503130674362183, "learning_rate": 0.0002, "epoch": 3.6844232479185193, "step": 22790}, {"loss": 0.6082, "grad_norm": 0.8413618206977844, "learning_rate": 0.0002, "epoch": 3.6860399320992645, "step": 22800}, {"loss": 0.614, "grad_norm": 0.8675165176391602, "learning_rate": 0.0002, "epoch": 3.6876566162800097, "step": 22810}, {"loss": 0.6157, "grad_norm": 0.8235884308815002, "learning_rate": 0.0002, "epoch": 3.689273300460755, "step": 22820}, {"loss": 0.5708, "grad_norm": 0.9477725625038147, "learning_rate": 0.0002, "epoch": 3.6908899846415, "step": 22830}, {"loss": 0.6481, "grad_norm": 0.7883533835411072, "learning_rate": 0.0002, "epoch": 3.6925066688222454, "step": 22840}, {"loss": 0.5872, "grad_norm": 1.047913908958435, "learning_rate": 0.0002, "epoch": 3.6941233530029907, "step": 22850}, {"loss": 0.6176, "grad_norm": 0.9171528816223145, "learning_rate": 0.0002, "epoch": 3.695740037183736, "step": 22860}, {"loss": 0.6204, "grad_norm": 0.9338192343711853, "learning_rate": 0.0002, "epoch": 3.6973567213644816, "step": 22870}, {"loss": 0.686, "grad_norm": 0.8799443244934082, "learning_rate": 0.0002, "epoch": 3.698973405545227, "step": 22880}, {"loss": 0.6206, "grad_norm": 0.8515434861183167, "learning_rate": 0.0002, "epoch": 3.700590089725972, "step": 22890}, {"loss": 0.5954, "grad_norm": 0.7805591821670532, "learning_rate": 0.0002, "epoch": 3.7022067739067173, "step": 22900}, {"loss": 0.6108, "grad_norm": 0.8470911979675293, "learning_rate": 0.0002, "epoch": 3.7038234580874625, "step": 22910}, {"loss": 0.6557, "grad_norm": 0.9452309012413025, "learning_rate": 0.0002, "epoch": 3.705440142268208, "step": 22920}, {"loss": 0.6529, "grad_norm": 0.950243353843689, "learning_rate": 0.0002, "epoch": 3.7070568264489534, "step": 22930}, {"loss": 0.6364, "grad_norm": 0.7882499098777771, "learning_rate": 0.0002, "epoch": 3.7086735106296986, "step": 22940}, {"loss": 0.6462, "grad_norm": 0.8307787775993347, "learning_rate": 0.0002, "epoch": 3.710290194810444, "step": 22950}, {"loss": 0.6371, "grad_norm": 1.0970630645751953, "learning_rate": 0.0002, "epoch": 3.711906878991189, "step": 22960}, {"loss": 0.6281, "grad_norm": 0.8269566297531128, "learning_rate": 0.0002, "epoch": 3.7135235631719343, "step": 22970}, {"loss": 0.6561, "grad_norm": 0.8306704759597778, "learning_rate": 0.0002, "epoch": 3.7151402473526796, "step": 22980}, {"loss": 0.6418, "grad_norm": 0.9710225462913513, "learning_rate": 0.0002, "epoch": 3.716756931533425, "step": 22990}, {"loss": 0.6639, "grad_norm": 0.8890530467033386, "learning_rate": 0.0002, "epoch": 3.71837361571417, "step": 23000}, {"loss": 0.6084, "grad_norm": 0.883522629737854, "learning_rate": 0.0002, "epoch": 3.7199902998949153, "step": 23010}, {"loss": 0.6183, "grad_norm": 0.8662652373313904, "learning_rate": 0.0002, "epoch": 3.721606984075661, "step": 23020}, {"loss": 0.6266, "grad_norm": 0.7228406667709351, "learning_rate": 0.0002, "epoch": 3.723223668256406, "step": 23030}, {"loss": 0.6417, "grad_norm": 1.060792088508606, "learning_rate": 0.0002, "epoch": 3.7248403524371514, "step": 23040}, {"loss": 0.6346, "grad_norm": 1.0119613409042358, "learning_rate": 0.0002, "epoch": 3.7264570366178966, "step": 23050}, {"loss": 0.6466, "grad_norm": 0.9212996959686279, "learning_rate": 0.0002, "epoch": 3.728073720798642, "step": 23060}, {"loss": 0.6454, "grad_norm": 0.925690233707428, "learning_rate": 0.0002, "epoch": 3.7296904049793875, "step": 23070}, {"loss": 0.615, "grad_norm": 0.8323310613632202, "learning_rate": 0.0002, "epoch": 3.7313070891601328, "step": 23080}, {"loss": 0.679, "grad_norm": 0.8966048955917358, "learning_rate": 0.0002, "epoch": 3.732923773340878, "step": 23090}, {"loss": 0.6151, "grad_norm": 0.8995837569236755, "learning_rate": 0.0002, "epoch": 3.7345404575216232, "step": 23100}, {"loss": 0.6143, "grad_norm": 0.8748890161514282, "learning_rate": 0.0002, "epoch": 3.7361571417023685, "step": 23110}, {"loss": 0.6246, "grad_norm": 0.7985540628433228, "learning_rate": 0.0002, "epoch": 3.7377738258831137, "step": 23120}, {"loss": 0.6279, "grad_norm": 1.0240917205810547, "learning_rate": 0.0002, "epoch": 3.739390510063859, "step": 23130}, {"loss": 0.6747, "grad_norm": 0.9181789755821228, "learning_rate": 0.0002, "epoch": 3.741007194244604, "step": 23140}, {"loss": 0.6026, "grad_norm": 0.8896583914756775, "learning_rate": 0.0002, "epoch": 3.7426238784253494, "step": 23150}, {"loss": 0.5972, "grad_norm": 0.8635515570640564, "learning_rate": 0.0002, "epoch": 3.744240562606095, "step": 23160}, {"loss": 0.6683, "grad_norm": 0.8873575329780579, "learning_rate": 0.0002, "epoch": 3.7458572467868403, "step": 23170}, {"loss": 0.6143, "grad_norm": 0.9807148575782776, "learning_rate": 0.0002, "epoch": 3.7474739309675855, "step": 23180}, {"loss": 0.6381, "grad_norm": 0.900477945804596, "learning_rate": 0.0002, "epoch": 3.7490906151483308, "step": 23190}, {"loss": 0.6542, "grad_norm": 0.9379992485046387, "learning_rate": 0.0002, "epoch": 3.750707299329076, "step": 23200}, {"loss": 0.6015, "grad_norm": 0.9649890661239624, "learning_rate": 0.0002, "epoch": 3.752323983509821, "step": 23210}, {"loss": 0.6735, "grad_norm": 0.824442446231842, "learning_rate": 0.0002, "epoch": 3.753940667690567, "step": 23220}, {"loss": 0.5992, "grad_norm": 0.8896150588989258, "learning_rate": 0.0002, "epoch": 3.755557351871312, "step": 23230}, {"loss": 0.6081, "grad_norm": 0.751249372959137, "learning_rate": 0.0002, "epoch": 3.7571740360520574, "step": 23240}, {"loss": 0.629, "grad_norm": 0.9392193555831909, "learning_rate": 0.0002, "epoch": 3.7587907202328026, "step": 23250}, {"loss": 0.6209, "grad_norm": 0.9284586310386658, "learning_rate": 0.0002, "epoch": 3.760407404413548, "step": 23260}, {"loss": 0.6414, "grad_norm": 0.7738175392150879, "learning_rate": 0.0002, "epoch": 3.762024088594293, "step": 23270}, {"loss": 0.6743, "grad_norm": 0.9252978563308716, "learning_rate": 0.0002, "epoch": 3.7636407727750383, "step": 23280}, {"loss": 0.5984, "grad_norm": 0.9501895904541016, "learning_rate": 0.0002, "epoch": 3.7652574569557835, "step": 23290}, {"loss": 0.6568, "grad_norm": 0.9416276216506958, "learning_rate": 0.0002, "epoch": 3.7668741411365287, "step": 23300}, {"loss": 0.6507, "grad_norm": 0.7076631784439087, "learning_rate": 0.0002, "epoch": 3.7684908253172744, "step": 23310}, {"loss": 0.6329, "grad_norm": 0.9864492416381836, "learning_rate": 0.0002, "epoch": 3.7701075094980196, "step": 23320}, {"loss": 0.6537, "grad_norm": 0.8450456261634827, "learning_rate": 0.0002, "epoch": 3.771724193678765, "step": 23330}, {"loss": 0.658, "grad_norm": 1.0768941640853882, "learning_rate": 0.0002, "epoch": 3.77334087785951, "step": 23340}, {"loss": 0.6408, "grad_norm": 0.9956819415092468, "learning_rate": 0.0002, "epoch": 3.7749575620402553, "step": 23350}, {"loss": 0.6464, "grad_norm": 0.9234658479690552, "learning_rate": 0.0002, "epoch": 3.7765742462210006, "step": 23360}, {"loss": 0.6542, "grad_norm": 1.0993858575820923, "learning_rate": 0.0002, "epoch": 3.7781909304017463, "step": 23370}, {"loss": 0.6391, "grad_norm": 0.923159658908844, "learning_rate": 0.0002, "epoch": 3.7798076145824915, "step": 23380}, {"loss": 0.6625, "grad_norm": 0.9311541318893433, "learning_rate": 0.0002, "epoch": 3.7814242987632367, "step": 23390}, {"loss": 0.6535, "grad_norm": 0.919681191444397, "learning_rate": 0.0002, "epoch": 3.783040982943982, "step": 23400}, {"loss": 0.6138, "grad_norm": 1.7406195402145386, "learning_rate": 0.0002, "epoch": 3.784657667124727, "step": 23410}, {"loss": 0.657, "grad_norm": 0.7789074182510376, "learning_rate": 0.0002, "epoch": 3.7862743513054724, "step": 23420}, {"loss": 0.658, "grad_norm": 0.8302814960479736, "learning_rate": 0.0002, "epoch": 3.7878910354862176, "step": 23430}, {"loss": 0.649, "grad_norm": 0.8089349269866943, "learning_rate": 0.0002, "epoch": 3.789507719666963, "step": 23440}, {"loss": 0.6682, "grad_norm": 0.9006284475326538, "learning_rate": 0.0002, "epoch": 3.791124403847708, "step": 23450}, {"loss": 0.6335, "grad_norm": 0.8426766991615295, "learning_rate": 0.0002, "epoch": 3.7927410880284538, "step": 23460}, {"loss": 0.6364, "grad_norm": 1.2576252222061157, "learning_rate": 0.0002, "epoch": 3.794357772209199, "step": 23470}, {"loss": 0.6324, "grad_norm": 1.0307610034942627, "learning_rate": 0.0002, "epoch": 3.7959744563899442, "step": 23480}, {"loss": 0.6262, "grad_norm": 0.8525972962379456, "learning_rate": 0.0002, "epoch": 3.7975911405706895, "step": 23490}, {"loss": 0.6757, "grad_norm": 1.159039855003357, "learning_rate": 0.0002, "epoch": 3.7992078247514347, "step": 23500}, {"loss": 0.6414, "grad_norm": 1.4193549156188965, "learning_rate": 0.0002, "epoch": 3.80082450893218, "step": 23510}, {"loss": 0.6413, "grad_norm": 0.8245543837547302, "learning_rate": 0.0002, "epoch": 3.8024411931129256, "step": 23520}, {"loss": 0.6417, "grad_norm": 0.8847230076789856, "learning_rate": 0.0002, "epoch": 3.804057877293671, "step": 23530}, {"loss": 0.6415, "grad_norm": 0.9574624300003052, "learning_rate": 0.0002, "epoch": 3.805674561474416, "step": 23540}, {"loss": 0.5765, "grad_norm": 1.048020601272583, "learning_rate": 0.0002, "epoch": 3.8072912456551613, "step": 23550}, {"loss": 0.6497, "grad_norm": 0.8302255868911743, "learning_rate": 0.0002, "epoch": 3.8089079298359065, "step": 23560}, {"loss": 0.6534, "grad_norm": 0.8269215822219849, "learning_rate": 0.0002, "epoch": 3.8105246140166518, "step": 23570}, {"loss": 0.6294, "grad_norm": 0.9375753402709961, "learning_rate": 0.0002, "epoch": 3.812141298197397, "step": 23580}, {"loss": 0.6132, "grad_norm": 1.0234097242355347, "learning_rate": 0.0002, "epoch": 3.8137579823781422, "step": 23590}, {"loss": 0.6625, "grad_norm": 0.8978445529937744, "learning_rate": 0.0002, "epoch": 3.8153746665588875, "step": 23600}, {"loss": 0.6315, "grad_norm": 0.7929515838623047, "learning_rate": 0.0002, "epoch": 3.816991350739633, "step": 23610}, {"loss": 0.6387, "grad_norm": 1.3255881071090698, "learning_rate": 0.0002, "epoch": 3.8186080349203784, "step": 23620}, {"loss": 0.5947, "grad_norm": 0.9188598990440369, "learning_rate": 0.0002, "epoch": 3.8202247191011236, "step": 23630}, {"loss": 0.6152, "grad_norm": 0.8811675906181335, "learning_rate": 0.0002, "epoch": 3.821841403281869, "step": 23640}, {"loss": 0.6253, "grad_norm": 0.8061038255691528, "learning_rate": 0.0002, "epoch": 3.823458087462614, "step": 23650}, {"loss": 0.6517, "grad_norm": 0.9975376129150391, "learning_rate": 0.0002, "epoch": 3.8250747716433597, "step": 23660}, {"loss": 0.6288, "grad_norm": 0.8036105036735535, "learning_rate": 0.0002, "epoch": 3.826691455824105, "step": 23670}, {"loss": 0.6845, "grad_norm": 0.7401984333992004, "learning_rate": 0.0002, "epoch": 3.82830814000485, "step": 23680}, {"loss": 0.6423, "grad_norm": 0.829753041267395, "learning_rate": 0.0002, "epoch": 3.8299248241855954, "step": 23690}, {"loss": 0.6611, "grad_norm": 0.8753240704536438, "learning_rate": 0.0002, "epoch": 3.8315415083663407, "step": 23700}, {"loss": 0.6686, "grad_norm": 0.8157842755317688, "learning_rate": 0.0002, "epoch": 3.833158192547086, "step": 23710}, {"loss": 0.6181, "grad_norm": 0.6183798909187317, "learning_rate": 0.0002, "epoch": 3.834774876727831, "step": 23720}, {"loss": 0.5965, "grad_norm": 0.9548442363739014, "learning_rate": 0.0002, "epoch": 3.8363915609085764, "step": 23730}, {"loss": 0.6456, "grad_norm": 0.8319669961929321, "learning_rate": 0.0002, "epoch": 3.8380082450893216, "step": 23740}, {"loss": 0.6585, "grad_norm": 0.9718693494796753, "learning_rate": 0.0002, "epoch": 3.839624929270067, "step": 23750}, {"loss": 0.6518, "grad_norm": 0.8672235012054443, "learning_rate": 0.0002, "epoch": 3.8412416134508125, "step": 23760}, {"loss": 0.6774, "grad_norm": 1.1210707426071167, "learning_rate": 0.0002, "epoch": 3.8428582976315577, "step": 23770}, {"loss": 0.5923, "grad_norm": 0.9177767634391785, "learning_rate": 0.0002, "epoch": 3.844474981812303, "step": 23780}, {"loss": 0.6286, "grad_norm": 0.8714171648025513, "learning_rate": 0.0002, "epoch": 3.846091665993048, "step": 23790}, {"loss": 0.6302, "grad_norm": 1.1853246688842773, "learning_rate": 0.0002, "epoch": 3.8477083501737934, "step": 23800}, {"loss": 0.6144, "grad_norm": 0.8091260194778442, "learning_rate": 0.0002, "epoch": 3.849325034354539, "step": 23810}, {"loss": 0.658, "grad_norm": 0.9710774421691895, "learning_rate": 0.0002, "epoch": 3.8509417185352843, "step": 23820}, {"loss": 0.6151, "grad_norm": 0.7648707628250122, "learning_rate": 0.0002, "epoch": 3.8525584027160296, "step": 23830}, {"loss": 0.6013, "grad_norm": 0.7809253931045532, "learning_rate": 0.0002, "epoch": 3.854175086896775, "step": 23840}, {"loss": 0.6006, "grad_norm": 0.8337951898574829, "learning_rate": 0.0002, "epoch": 3.85579177107752, "step": 23850}, {"loss": 0.6456, "grad_norm": 0.9271913170814514, "learning_rate": 0.0002, "epoch": 3.8574084552582653, "step": 23860}, {"loss": 0.6671, "grad_norm": 0.985334038734436, "learning_rate": 0.0002, "epoch": 3.8590251394390105, "step": 23870}, {"loss": 0.6693, "grad_norm": 0.8458583354949951, "learning_rate": 0.0002, "epoch": 3.8606418236197557, "step": 23880}, {"loss": 0.6207, "grad_norm": 1.015348196029663, "learning_rate": 0.0002, "epoch": 3.862258507800501, "step": 23890}, {"loss": 0.649, "grad_norm": 1.0121688842773438, "learning_rate": 0.0002, "epoch": 3.8638751919812466, "step": 23900}, {"loss": 0.5921, "grad_norm": 0.8883971571922302, "learning_rate": 0.0002, "epoch": 3.865491876161992, "step": 23910}, {"loss": 0.6597, "grad_norm": 1.028086543083191, "learning_rate": 0.0002, "epoch": 3.867108560342737, "step": 23920}, {"loss": 0.6654, "grad_norm": 0.9645734429359436, "learning_rate": 0.0002, "epoch": 3.8687252445234823, "step": 23930}, {"loss": 0.6328, "grad_norm": 0.8235350251197815, "learning_rate": 0.0002, "epoch": 3.8703419287042276, "step": 23940}, {"loss": 0.6387, "grad_norm": 1.0298916101455688, "learning_rate": 0.0002, "epoch": 3.871958612884973, "step": 23950}, {"loss": 0.5966, "grad_norm": 1.0063377618789673, "learning_rate": 0.0002, "epoch": 3.8735752970657185, "step": 23960}, {"loss": 0.6234, "grad_norm": 0.9230626821517944, "learning_rate": 0.0002, "epoch": 3.8751919812464637, "step": 23970}, {"loss": 0.6159, "grad_norm": 0.9243063926696777, "learning_rate": 0.0002, "epoch": 3.876808665427209, "step": 23980}, {"loss": 0.6035, "grad_norm": 1.0211291313171387, "learning_rate": 0.0002, "epoch": 3.878425349607954, "step": 23990}, {"loss": 0.6351, "grad_norm": 0.7800535559654236, "learning_rate": 0.0002, "epoch": 3.8800420337886994, "step": 24000}, {"loss": 0.7, "grad_norm": 0.7904248833656311, "learning_rate": 0.0002, "epoch": 3.8816587179694446, "step": 24010}, {"loss": 0.6516, "grad_norm": 1.1975988149642944, "learning_rate": 0.0002, "epoch": 3.88327540215019, "step": 24020}, {"loss": 0.6006, "grad_norm": 1.0626593828201294, "learning_rate": 0.0002, "epoch": 3.884892086330935, "step": 24030}, {"loss": 0.6115, "grad_norm": 0.9012193083763123, "learning_rate": 0.0002, "epoch": 3.8865087705116803, "step": 24040}, {"loss": 0.6786, "grad_norm": 1.1159172058105469, "learning_rate": 0.0002, "epoch": 3.888125454692426, "step": 24050}, {"loss": 0.6635, "grad_norm": 1.276838779449463, "learning_rate": 0.0002, "epoch": 3.889742138873171, "step": 24060}, {"loss": 0.5985, "grad_norm": 0.8467690348625183, "learning_rate": 0.0002, "epoch": 3.8913588230539164, "step": 24070}, {"loss": 0.6655, "grad_norm": 0.9862841963768005, "learning_rate": 0.0002, "epoch": 3.8929755072346617, "step": 24080}, {"loss": 0.6098, "grad_norm": 0.7134621739387512, "learning_rate": 0.0002, "epoch": 3.894592191415407, "step": 24090}, {"loss": 0.618, "grad_norm": 0.8178175091743469, "learning_rate": 0.0002, "epoch": 3.896208875596152, "step": 24100}, {"loss": 0.6147, "grad_norm": 0.9229172468185425, "learning_rate": 0.0002, "epoch": 3.897825559776898, "step": 24110}, {"loss": 0.6554, "grad_norm": 1.0878316164016724, "learning_rate": 0.0002, "epoch": 3.899442243957643, "step": 24120}, {"loss": 0.6616, "grad_norm": 0.971645712852478, "learning_rate": 0.0002, "epoch": 3.9010589281383883, "step": 24130}, {"loss": 0.6228, "grad_norm": 0.8862188458442688, "learning_rate": 0.0002, "epoch": 3.9026756123191335, "step": 24140}, {"loss": 0.6192, "grad_norm": 0.9126982688903809, "learning_rate": 0.0002, "epoch": 3.9042922964998787, "step": 24150}, {"loss": 0.6734, "grad_norm": 0.8833470940589905, "learning_rate": 0.0002, "epoch": 3.905908980680624, "step": 24160}, {"loss": 0.5832, "grad_norm": 0.8320947885513306, "learning_rate": 0.0002, "epoch": 3.907525664861369, "step": 24170}, {"loss": 0.6247, "grad_norm": 0.9156602025032043, "learning_rate": 0.0002, "epoch": 3.9091423490421144, "step": 24180}, {"loss": 0.6678, "grad_norm": 1.029181957244873, "learning_rate": 0.0002, "epoch": 3.9107590332228597, "step": 24190}, {"loss": 0.6565, "grad_norm": 0.9052802324295044, "learning_rate": 0.0002, "epoch": 3.9123757174036053, "step": 24200}, {"loss": 0.6346, "grad_norm": 0.8847255110740662, "learning_rate": 0.0002, "epoch": 3.9139924015843506, "step": 24210}, {"loss": 0.6343, "grad_norm": 0.9642062187194824, "learning_rate": 0.0002, "epoch": 3.915609085765096, "step": 24220}, {"loss": 0.6557, "grad_norm": 0.8629093766212463, "learning_rate": 0.0002, "epoch": 3.917225769945841, "step": 24230}, {"loss": 0.6086, "grad_norm": 0.8674976825714111, "learning_rate": 0.0002, "epoch": 3.9188424541265863, "step": 24240}, {"loss": 0.5874, "grad_norm": 1.104846477508545, "learning_rate": 0.0002, "epoch": 3.9204591383073315, "step": 24250}, {"loss": 0.6501, "grad_norm": 1.0874955654144287, "learning_rate": 0.0002, "epoch": 3.922075822488077, "step": 24260}, {"loss": 0.6455, "grad_norm": 0.8689812421798706, "learning_rate": 0.0002, "epoch": 3.9236925066688224, "step": 24270}, {"loss": 0.5893, "grad_norm": 0.9724617004394531, "learning_rate": 0.0002, "epoch": 3.9253091908495676, "step": 24280}, {"loss": 0.6616, "grad_norm": 0.9165538549423218, "learning_rate": 0.0002, "epoch": 3.926925875030313, "step": 24290}, {"loss": 0.645, "grad_norm": 0.9307710528373718, "learning_rate": 0.0002, "epoch": 3.928542559211058, "step": 24300}, {"loss": 0.6071, "grad_norm": 0.8589295148849487, "learning_rate": 0.0002, "epoch": 3.9301592433918033, "step": 24310}, {"loss": 0.6662, "grad_norm": 0.9151099920272827, "learning_rate": 0.0002, "epoch": 3.9317759275725486, "step": 24320}, {"loss": 0.7075, "grad_norm": 0.9633517265319824, "learning_rate": 0.0002, "epoch": 3.933392611753294, "step": 24330}, {"loss": 0.6432, "grad_norm": 0.9521116018295288, "learning_rate": 0.0002, "epoch": 3.935009295934039, "step": 24340}, {"loss": 0.6457, "grad_norm": 0.8366776704788208, "learning_rate": 0.0002, "epoch": 3.9366259801147847, "step": 24350}, {"loss": 0.6139, "grad_norm": 0.8972663283348083, "learning_rate": 0.0002, "epoch": 3.93824266429553, "step": 24360}, {"loss": 0.661, "grad_norm": 0.8102919459342957, "learning_rate": 0.0002, "epoch": 3.939859348476275, "step": 24370}, {"loss": 0.6388, "grad_norm": 0.8189975023269653, "learning_rate": 0.0002, "epoch": 3.9414760326570204, "step": 24380}, {"loss": 0.6818, "grad_norm": 0.9569464921951294, "learning_rate": 0.0002, "epoch": 3.9430927168377656, "step": 24390}, {"loss": 0.6999, "grad_norm": 0.7459101676940918, "learning_rate": 0.0002, "epoch": 3.9447094010185113, "step": 24400}, {"loss": 0.6069, "grad_norm": 0.8536974787712097, "learning_rate": 0.0002, "epoch": 3.9463260851992565, "step": 24410}, {"loss": 0.5683, "grad_norm": 0.8763698935508728, "learning_rate": 0.0002, "epoch": 3.9479427693800018, "step": 24420}, {"loss": 0.6478, "grad_norm": 0.9381106495857239, "learning_rate": 0.0002, "epoch": 3.949559453560747, "step": 24430}, {"loss": 0.6371, "grad_norm": 0.934440016746521, "learning_rate": 0.0002, "epoch": 3.9511761377414922, "step": 24440}, {"loss": 0.6393, "grad_norm": 0.903918981552124, "learning_rate": 0.0002, "epoch": 3.9527928219222375, "step": 24450}, {"loss": 0.6175, "grad_norm": 0.8771953582763672, "learning_rate": 0.0002, "epoch": 3.9544095061029827, "step": 24460}, {"loss": 0.6971, "grad_norm": 1.0375410318374634, "learning_rate": 0.0002, "epoch": 3.956026190283728, "step": 24470}, {"loss": 0.6313, "grad_norm": 0.9439185261726379, "learning_rate": 0.0002, "epoch": 3.957642874464473, "step": 24480}, {"loss": 0.6076, "grad_norm": 0.935467004776001, "learning_rate": 0.0002, "epoch": 3.9592595586452184, "step": 24490}, {"loss": 0.6437, "grad_norm": 0.6900772452354431, "learning_rate": 0.0002, "epoch": 3.960876242825964, "step": 24500}, {"loss": 0.6445, "grad_norm": 1.0172916650772095, "learning_rate": 0.0002, "epoch": 3.9624929270067093, "step": 24510}, {"loss": 0.6308, "grad_norm": 0.9167046546936035, "learning_rate": 0.0002, "epoch": 3.9641096111874545, "step": 24520}, {"loss": 0.6519, "grad_norm": 0.7230527997016907, "learning_rate": 0.0002, "epoch": 3.9657262953681998, "step": 24530}, {"loss": 0.6564, "grad_norm": 0.8980403542518616, "learning_rate": 0.0002, "epoch": 3.967342979548945, "step": 24540}, {"loss": 0.6099, "grad_norm": 0.8555465936660767, "learning_rate": 0.0002, "epoch": 3.9689596637296907, "step": 24550}, {"loss": 0.6617, "grad_norm": 0.7825445532798767, "learning_rate": 0.0002, "epoch": 3.970576347910436, "step": 24560}, {"loss": 0.604, "grad_norm": 0.7273133993148804, "learning_rate": 0.0002, "epoch": 3.972193032091181, "step": 24570}, {"loss": 0.6427, "grad_norm": 0.9612047672271729, "learning_rate": 0.0002, "epoch": 3.9738097162719264, "step": 24580}, {"loss": 0.6426, "grad_norm": 0.9865460991859436, "learning_rate": 0.0002, "epoch": 3.9754264004526716, "step": 24590}, {"loss": 0.6052, "grad_norm": 0.8638762831687927, "learning_rate": 0.0002, "epoch": 3.977043084633417, "step": 24600}, {"loss": 0.6097, "grad_norm": 1.0096198320388794, "learning_rate": 0.0002, "epoch": 3.978659768814162, "step": 24610}, {"loss": 0.6664, "grad_norm": 0.8475532531738281, "learning_rate": 0.0002, "epoch": 3.9802764529949073, "step": 24620}, {"loss": 0.6711, "grad_norm": 0.9696195721626282, "learning_rate": 0.0002, "epoch": 3.9818931371756525, "step": 24630}, {"loss": 0.6446, "grad_norm": 0.7499843239784241, "learning_rate": 0.0002, "epoch": 3.9835098213563978, "step": 24640}, {"loss": 0.6054, "grad_norm": 0.8865424990653992, "learning_rate": 0.0002, "epoch": 3.9851265055371434, "step": 24650}, {"loss": 0.5975, "grad_norm": 0.8089959025382996, "learning_rate": 0.0002, "epoch": 3.9867431897178887, "step": 24660}, {"loss": 0.6677, "grad_norm": 0.6946012377738953, "learning_rate": 0.0002, "epoch": 3.988359873898634, "step": 24670}, {"loss": 0.6329, "grad_norm": 0.7991759181022644, "learning_rate": 0.0002, "epoch": 3.989976558079379, "step": 24680}, {"loss": 0.6449, "grad_norm": 0.8803931474685669, "learning_rate": 0.0002, "epoch": 3.9915932422601244, "step": 24690}, {"loss": 0.7091, "grad_norm": 0.8848299980163574, "learning_rate": 0.0002, "epoch": 3.99320992644087, "step": 24700}, {"loss": 0.6551, "grad_norm": 0.7448889017105103, "learning_rate": 0.0002, "epoch": 3.9948266106216153, "step": 24710}, {"loss": 0.6432, "grad_norm": 0.9361620545387268, "learning_rate": 0.0002, "epoch": 3.9964432948023605, "step": 24720}, {"loss": 0.5917, "grad_norm": 0.9958081245422363, "learning_rate": 0.0002, "epoch": 3.9980599789831057, "step": 24730}, {"loss": 0.6567, "grad_norm": 1.026004672050476, "learning_rate": 0.0002, "epoch": 3.999676663163851, "step": 24740}, {"eval_loss": 1.1524168252944946, "eval_runtime": 122.1585, "eval_samples_per_second": 6.0, "eval_steps_per_second": 0.753, "epoch": 4.0, "step": 24742}, {"loss": 0.6057, "grad_norm": 1.0664808750152588, "learning_rate": 0.0002, "epoch": 4.001293347344596, "step": 24750}, {"loss": 0.5644, "grad_norm": 1.0113720893859863, "learning_rate": 0.0002, "epoch": 4.002910031525341, "step": 24760}, {"loss": 0.5628, "grad_norm": 0.991486668586731, "learning_rate": 0.0002, "epoch": 4.004526715706087, "step": 24770}, {"loss": 0.508, "grad_norm": 0.951754629611969, "learning_rate": 0.0002, "epoch": 4.006143399886832, "step": 24780}, {"loss": 0.5314, "grad_norm": 1.13059401512146, "learning_rate": 0.0002, "epoch": 4.007760084067577, "step": 24790}, {"loss": 0.5323, "grad_norm": 0.9343926310539246, "learning_rate": 0.0002, "epoch": 4.009376768248322, "step": 24800}, {"loss": 0.5161, "grad_norm": 1.0680590867996216, "learning_rate": 0.0002, "epoch": 4.010993452429068, "step": 24810}, {"loss": 0.513, "grad_norm": 1.0022706985473633, "learning_rate": 0.0002, "epoch": 4.012610136609814, "step": 24820}, {"loss": 0.543, "grad_norm": 1.0285297632217407, "learning_rate": 0.0002, "epoch": 4.014226820790559, "step": 24830}, {"loss": 0.5311, "grad_norm": 0.8347002863883972, "learning_rate": 0.0002, "epoch": 4.015843504971304, "step": 24840}, {"loss": 0.5655, "grad_norm": 0.9675396680831909, "learning_rate": 0.0002, "epoch": 4.017460189152049, "step": 24850}, {"loss": 0.5625, "grad_norm": 0.9238511323928833, "learning_rate": 0.0002, "epoch": 4.019076873332795, "step": 24860}, {"loss": 0.5327, "grad_norm": 1.1576941013336182, "learning_rate": 0.0002, "epoch": 4.02069355751354, "step": 24870}, {"loss": 0.5533, "grad_norm": 0.8583757281303406, "learning_rate": 0.0002, "epoch": 4.022310241694285, "step": 24880}, {"loss": 0.5483, "grad_norm": 0.9816817045211792, "learning_rate": 0.0002, "epoch": 4.02392692587503, "step": 24890}, {"loss": 0.5605, "grad_norm": 0.955073893070221, "learning_rate": 0.0002, "epoch": 4.0255436100557755, "step": 24900}, {"loss": 0.4896, "grad_norm": 1.1054974794387817, "learning_rate": 0.0002, "epoch": 4.027160294236521, "step": 24910}, {"loss": 0.5246, "grad_norm": 1.1240060329437256, "learning_rate": 0.0002, "epoch": 4.028776978417266, "step": 24920}, {"loss": 0.5451, "grad_norm": 0.9512825012207031, "learning_rate": 0.0002, "epoch": 4.030393662598011, "step": 24930}, {"loss": 0.5584, "grad_norm": 0.85965496301651, "learning_rate": 0.0002, "epoch": 4.0320103467787565, "step": 24940}, {"loss": 0.5564, "grad_norm": 0.9378061294555664, "learning_rate": 0.0002, "epoch": 4.033627030959502, "step": 24950}, {"loss": 0.5008, "grad_norm": 0.9655424356460571, "learning_rate": 0.0002, "epoch": 4.035243715140247, "step": 24960}, {"loss": 0.5538, "grad_norm": 1.1393707990646362, "learning_rate": 0.0002, "epoch": 4.036860399320993, "step": 24970}, {"loss": 0.5785, "grad_norm": 1.0220451354980469, "learning_rate": 0.0002, "epoch": 4.038477083501738, "step": 24980}, {"loss": 0.5813, "grad_norm": 0.9785808324813843, "learning_rate": 0.0002, "epoch": 4.0400937676824835, "step": 24990}, {"loss": 0.5153, "grad_norm": 1.0257649421691895, "learning_rate": 0.0002, "epoch": 4.041710451863229, "step": 25000}, {"loss": 0.5658, "grad_norm": 0.9737892150878906, "learning_rate": 0.0002, "epoch": 4.043327136043974, "step": 25010}, {"loss": 0.5515, "grad_norm": 0.7416959404945374, "learning_rate": 0.0002, "epoch": 4.044943820224719, "step": 25020}, {"loss": 0.5372, "grad_norm": 0.7909596562385559, "learning_rate": 0.0002, "epoch": 4.046560504405464, "step": 25030}, {"loss": 0.5265, "grad_norm": 0.8923130631446838, "learning_rate": 0.0002, "epoch": 4.04817718858621, "step": 25040}, {"loss": 0.5035, "grad_norm": 0.9044941663742065, "learning_rate": 0.0002, "epoch": 4.049793872766955, "step": 25050}, {"loss": 0.5135, "grad_norm": 0.866352379322052, "learning_rate": 0.0002, "epoch": 4.0514105569477, "step": 25060}, {"loss": 0.5956, "grad_norm": 1.544549822807312, "learning_rate": 0.0002, "epoch": 4.053027241128445, "step": 25070}, {"loss": 0.5418, "grad_norm": 0.8426995277404785, "learning_rate": 0.0002, "epoch": 4.054643925309191, "step": 25080}, {"loss": 0.5537, "grad_norm": 0.9797548651695251, "learning_rate": 0.0002, "epoch": 4.056260609489936, "step": 25090}, {"loss": 0.55, "grad_norm": 0.8468434810638428, "learning_rate": 0.0002, "epoch": 4.057877293670681, "step": 25100}, {"loss": 0.5242, "grad_norm": 0.9294559955596924, "learning_rate": 0.0002, "epoch": 4.059493977851426, "step": 25110}, {"loss": 0.5295, "grad_norm": 0.9686688780784607, "learning_rate": 0.0002, "epoch": 4.061110662032172, "step": 25120}, {"loss": 0.5642, "grad_norm": 0.8042728304862976, "learning_rate": 0.0002, "epoch": 4.062727346212918, "step": 25130}, {"loss": 0.548, "grad_norm": 1.165160894393921, "learning_rate": 0.0002, "epoch": 4.064344030393663, "step": 25140}, {"loss": 0.5473, "grad_norm": 1.2161961793899536, "learning_rate": 0.0002, "epoch": 4.065960714574408, "step": 25150}, {"loss": 0.5217, "grad_norm": 1.0762810707092285, "learning_rate": 0.0002, "epoch": 4.067577398755153, "step": 25160}, {"loss": 0.5886, "grad_norm": 0.7580869793891907, "learning_rate": 0.0002, "epoch": 4.069194082935899, "step": 25170}, {"loss": 0.5401, "grad_norm": 0.9630117416381836, "learning_rate": 0.0002, "epoch": 4.070810767116644, "step": 25180}, {"loss": 0.5378, "grad_norm": 0.9049716591835022, "learning_rate": 0.0002, "epoch": 4.072427451297389, "step": 25190}, {"loss": 0.5266, "grad_norm": 1.1536930799484253, "learning_rate": 0.0002, "epoch": 4.074044135478134, "step": 25200}, {"loss": 0.5523, "grad_norm": 0.901461124420166, "learning_rate": 0.0002, "epoch": 4.0756608196588795, "step": 25210}, {"loss": 0.5132, "grad_norm": 1.3318437337875366, "learning_rate": 0.0002, "epoch": 4.077277503839625, "step": 25220}, {"loss": 0.5317, "grad_norm": 0.8811455368995667, "learning_rate": 0.0002, "epoch": 4.07889418802037, "step": 25230}, {"loss": 0.5798, "grad_norm": 1.0564165115356445, "learning_rate": 0.0002, "epoch": 4.080510872201115, "step": 25240}, {"loss": 0.5472, "grad_norm": 1.1008027791976929, "learning_rate": 0.0002, "epoch": 4.08212755638186, "step": 25250}, {"loss": 0.5195, "grad_norm": 1.150097131729126, "learning_rate": 0.0002, "epoch": 4.083744240562606, "step": 25260}, {"loss": 0.5321, "grad_norm": 0.9339924454689026, "learning_rate": 0.0002, "epoch": 4.085360924743352, "step": 25270}, {"loss": 0.5597, "grad_norm": 1.0902045965194702, "learning_rate": 0.0002, "epoch": 4.086977608924097, "step": 25280}, {"loss": 0.5203, "grad_norm": 0.8483911156654358, "learning_rate": 0.0002, "epoch": 4.088594293104842, "step": 25290}, {"loss": 0.5697, "grad_norm": 0.9477024674415588, "learning_rate": 0.0002, "epoch": 4.0902109772855875, "step": 25300}, {"loss": 0.5384, "grad_norm": 0.9500215649604797, "learning_rate": 0.0002, "epoch": 4.091827661466333, "step": 25310}, {"loss": 0.5045, "grad_norm": 1.040468454360962, "learning_rate": 0.0002, "epoch": 4.093444345647078, "step": 25320}, {"loss": 0.5488, "grad_norm": 0.7457592487335205, "learning_rate": 0.0002, "epoch": 4.095061029827823, "step": 25330}, {"loss": 0.609, "grad_norm": 1.2092097997665405, "learning_rate": 0.0002, "epoch": 4.096677714008568, "step": 25340}, {"loss": 0.5174, "grad_norm": 0.9652107954025269, "learning_rate": 0.0002, "epoch": 4.098294398189314, "step": 25350}, {"loss": 0.5559, "grad_norm": 0.8464955687522888, "learning_rate": 0.0002, "epoch": 4.099911082370059, "step": 25360}, {"loss": 0.5635, "grad_norm": 0.875026285648346, "learning_rate": 0.0002, "epoch": 4.101527766550804, "step": 25370}, {"loss": 0.5774, "grad_norm": 0.9241740107536316, "learning_rate": 0.0002, "epoch": 4.103144450731549, "step": 25380}, {"loss": 0.5578, "grad_norm": 0.9769546389579773, "learning_rate": 0.0002, "epoch": 4.1047611349122946, "step": 25390}, {"loss": 0.567, "grad_norm": 1.1501960754394531, "learning_rate": 0.0002, "epoch": 4.10637781909304, "step": 25400}, {"loss": 0.5241, "grad_norm": 0.9135243892669678, "learning_rate": 0.0002, "epoch": 4.107994503273786, "step": 25410}, {"loss": 0.5152, "grad_norm": 0.9905396103858948, "learning_rate": 0.0002, "epoch": 4.109611187454531, "step": 25420}, {"loss": 0.5064, "grad_norm": 0.9845104217529297, "learning_rate": 0.0002, "epoch": 4.111227871635276, "step": 25430}, {"loss": 0.5029, "grad_norm": 0.8326883912086487, "learning_rate": 0.0002, "epoch": 4.112844555816022, "step": 25440}, {"loss": 0.5312, "grad_norm": 0.9264556765556335, "learning_rate": 0.0002, "epoch": 4.114461239996767, "step": 25450}, {"loss": 0.5968, "grad_norm": 1.043080449104309, "learning_rate": 0.0002, "epoch": 4.116077924177512, "step": 25460}, {"loss": 0.5773, "grad_norm": 0.8533386588096619, "learning_rate": 0.0002, "epoch": 4.117694608358257, "step": 25470}, {"loss": 0.5584, "grad_norm": 1.0133965015411377, "learning_rate": 0.0002, "epoch": 4.1193112925390025, "step": 25480}, {"loss": 0.566, "grad_norm": 0.7476310133934021, "learning_rate": 0.0002, "epoch": 4.120927976719748, "step": 25490}, {"loss": 0.5189, "grad_norm": 1.1247259378433228, "learning_rate": 0.0002, "epoch": 4.122544660900493, "step": 25500}, {"loss": 0.5751, "grad_norm": 1.0764678716659546, "learning_rate": 0.0002, "epoch": 4.124161345081238, "step": 25510}, {"loss": 0.5391, "grad_norm": 0.7679798007011414, "learning_rate": 0.0002, "epoch": 4.1257780292619834, "step": 25520}, {"loss": 0.5233, "grad_norm": 0.8877071142196655, "learning_rate": 0.0002, "epoch": 4.127394713442729, "step": 25530}, {"loss": 0.5769, "grad_norm": 1.0440239906311035, "learning_rate": 0.0002, "epoch": 4.129011397623474, "step": 25540}, {"loss": 0.5723, "grad_norm": 0.984145998954773, "learning_rate": 0.0002, "epoch": 4.130628081804219, "step": 25550}, {"loss": 0.5741, "grad_norm": 0.8667055368423462, "learning_rate": 0.0002, "epoch": 4.132244765984965, "step": 25560}, {"loss": 0.5816, "grad_norm": 1.1300835609436035, "learning_rate": 0.0002, "epoch": 4.1338614501657105, "step": 25570}, {"loss": 0.524, "grad_norm": 0.9314348101615906, "learning_rate": 0.0002, "epoch": 4.135478134346456, "step": 25580}, {"loss": 0.5283, "grad_norm": 0.7731879949569702, "learning_rate": 0.0002, "epoch": 4.137094818527201, "step": 25590}, {"loss": 0.5307, "grad_norm": 1.0080097913742065, "learning_rate": 0.0002, "epoch": 4.138711502707946, "step": 25600}, {"loss": 0.5759, "grad_norm": 1.2475038766860962, "learning_rate": 0.0002, "epoch": 4.140328186888691, "step": 25610}, {"loss": 0.55, "grad_norm": 0.9912930727005005, "learning_rate": 0.0002, "epoch": 4.141944871069437, "step": 25620}, {"loss": 0.5624, "grad_norm": 0.9088651537895203, "learning_rate": 0.0002, "epoch": 4.143561555250182, "step": 25630}, {"loss": 0.5393, "grad_norm": 0.8940697312355042, "learning_rate": 0.0002, "epoch": 4.145178239430927, "step": 25640}, {"loss": 0.5341, "grad_norm": 1.0798203945159912, "learning_rate": 0.0002, "epoch": 4.146794923611672, "step": 25650}, {"loss": 0.5987, "grad_norm": 0.955172061920166, "learning_rate": 0.0002, "epoch": 4.148411607792418, "step": 25660}, {"loss": 0.569, "grad_norm": 0.9692716002464294, "learning_rate": 0.0002, "epoch": 4.150028291973163, "step": 25670}, {"loss": 0.5478, "grad_norm": 1.0813939571380615, "learning_rate": 0.0002, "epoch": 4.151644976153908, "step": 25680}, {"loss": 0.5383, "grad_norm": 1.135675072669983, "learning_rate": 0.0002, "epoch": 4.153261660334653, "step": 25690}, {"loss": 0.5247, "grad_norm": 1.0392236709594727, "learning_rate": 0.0002, "epoch": 4.1548783445153985, "step": 25700}, {"loss": 0.5204, "grad_norm": 0.9473116993904114, "learning_rate": 0.0002, "epoch": 4.156495028696145, "step": 25710}, {"loss": 0.5339, "grad_norm": 0.712493896484375, "learning_rate": 0.0002, "epoch": 4.15811171287689, "step": 25720}, {"loss": 0.5781, "grad_norm": 0.8724465370178223, "learning_rate": 0.0002, "epoch": 4.159728397057635, "step": 25730}, {"loss": 0.5325, "grad_norm": 0.9870015978813171, "learning_rate": 0.0002, "epoch": 4.16134508123838, "step": 25740}, {"loss": 0.5503, "grad_norm": 1.025273084640503, "learning_rate": 0.0002, "epoch": 4.1629617654191255, "step": 25750}, {"loss": 0.5223, "grad_norm": 0.9243090152740479, "learning_rate": 0.0002, "epoch": 4.164578449599871, "step": 25760}, {"loss": 0.5177, "grad_norm": 1.1656451225280762, "learning_rate": 0.0002, "epoch": 4.166195133780616, "step": 25770}, {"loss": 0.5334, "grad_norm": 0.936358630657196, "learning_rate": 0.0002, "epoch": 4.167811817961361, "step": 25780}, {"loss": 0.5236, "grad_norm": 0.8618208169937134, "learning_rate": 0.0002, "epoch": 4.1694285021421065, "step": 25790}, {"loss": 0.5186, "grad_norm": 0.8580600023269653, "learning_rate": 0.0002, "epoch": 4.171045186322852, "step": 25800}, {"loss": 0.5212, "grad_norm": 1.0128562450408936, "learning_rate": 0.0002, "epoch": 4.172661870503597, "step": 25810}, {"loss": 0.5404, "grad_norm": 0.854865312576294, "learning_rate": 0.0002, "epoch": 4.174278554684342, "step": 25820}, {"loss": 0.5377, "grad_norm": 1.235082745552063, "learning_rate": 0.0002, "epoch": 4.175895238865087, "step": 25830}, {"loss": 0.5614, "grad_norm": 0.9796220660209656, "learning_rate": 0.0002, "epoch": 4.177511923045833, "step": 25840}, {"loss": 0.5689, "grad_norm": 0.8922094702720642, "learning_rate": 0.0002, "epoch": 4.179128607226578, "step": 25850}, {"loss": 0.5806, "grad_norm": 0.9672530293464661, "learning_rate": 0.0002, "epoch": 4.180745291407324, "step": 25860}, {"loss": 0.5074, "grad_norm": 0.8662548661231995, "learning_rate": 0.0002, "epoch": 4.182361975588069, "step": 25870}, {"loss": 0.5329, "grad_norm": 0.7938798069953918, "learning_rate": 0.0002, "epoch": 4.1839786597688144, "step": 25880}, {"loss": 0.5427, "grad_norm": 1.0517958402633667, "learning_rate": 0.0002, "epoch": 4.18559534394956, "step": 25890}, {"loss": 0.5147, "grad_norm": 0.8939275145530701, "learning_rate": 0.0002, "epoch": 4.187212028130305, "step": 25900}, {"loss": 0.5199, "grad_norm": 1.0296672582626343, "learning_rate": 0.0002, "epoch": 4.18882871231105, "step": 25910}, {"loss": 0.5522, "grad_norm": 0.8104017972946167, "learning_rate": 0.0002, "epoch": 4.190445396491795, "step": 25920}, {"loss": 0.596, "grad_norm": 0.9984509944915771, "learning_rate": 0.0002, "epoch": 4.192062080672541, "step": 25930}, {"loss": 0.5356, "grad_norm": 0.9844784736633301, "learning_rate": 0.0002, "epoch": 4.193678764853286, "step": 25940}, {"loss": 0.5198, "grad_norm": 0.8168622255325317, "learning_rate": 0.0002, "epoch": 4.195295449034031, "step": 25950}, {"loss": 0.542, "grad_norm": 1.0878913402557373, "learning_rate": 0.0002, "epoch": 4.196912133214776, "step": 25960}, {"loss": 0.5414, "grad_norm": 0.927126407623291, "learning_rate": 0.0002, "epoch": 4.1985288173955215, "step": 25970}, {"loss": 0.5794, "grad_norm": 0.838586688041687, "learning_rate": 0.0002, "epoch": 4.200145501576267, "step": 25980}, {"loss": 0.5454, "grad_norm": 1.2572145462036133, "learning_rate": 0.0002, "epoch": 4.201762185757012, "step": 25990}, {"loss": 0.5048, "grad_norm": 1.0476740598678589, "learning_rate": 0.0002, "epoch": 4.203378869937758, "step": 26000}, {"loss": 0.5127, "grad_norm": 1.0873368978500366, "learning_rate": 0.0002, "epoch": 4.204995554118503, "step": 26010}, {"loss": 0.5679, "grad_norm": 1.2664896249771118, "learning_rate": 0.0002, "epoch": 4.206612238299249, "step": 26020}, {"loss": 0.5814, "grad_norm": 1.0312391519546509, "learning_rate": 0.0002, "epoch": 4.208228922479994, "step": 26030}, {"loss": 0.571, "grad_norm": 1.0235042572021484, "learning_rate": 0.0002, "epoch": 4.209845606660739, "step": 26040}, {"loss": 0.5766, "grad_norm": 0.8882219195365906, "learning_rate": 0.0002, "epoch": 4.211462290841484, "step": 26050}, {"loss": 0.5557, "grad_norm": 0.9115961790084839, "learning_rate": 0.0002, "epoch": 4.2130789750222295, "step": 26060}, {"loss": 0.5455, "grad_norm": 1.0218228101730347, "learning_rate": 0.0002, "epoch": 4.214695659202975, "step": 26070}, {"loss": 0.5462, "grad_norm": 1.0802232027053833, "learning_rate": 0.0002, "epoch": 4.21631234338372, "step": 26080}, {"loss": 0.557, "grad_norm": 1.1488053798675537, "learning_rate": 0.0002, "epoch": 4.217929027564465, "step": 26090}, {"loss": 0.52, "grad_norm": 1.0487725734710693, "learning_rate": 0.0002, "epoch": 4.21954571174521, "step": 26100}, {"loss": 0.5568, "grad_norm": 0.9131165742874146, "learning_rate": 0.0002, "epoch": 4.221162395925956, "step": 26110}, {"loss": 0.5206, "grad_norm": 0.9012845158576965, "learning_rate": 0.0002, "epoch": 4.222779080106701, "step": 26120}, {"loss": 0.561, "grad_norm": 0.8389840126037598, "learning_rate": 0.0002, "epoch": 4.224395764287446, "step": 26130}, {"loss": 0.5268, "grad_norm": 0.8924660682678223, "learning_rate": 0.0002, "epoch": 4.226012448468191, "step": 26140}, {"loss": 0.5715, "grad_norm": 0.8556463718414307, "learning_rate": 0.0002, "epoch": 4.2276291326489375, "step": 26150}, {"loss": 0.5695, "grad_norm": 0.9643129110336304, "learning_rate": 0.0002, "epoch": 4.229245816829683, "step": 26160}, {"loss": 0.5321, "grad_norm": 0.9865712523460388, "learning_rate": 0.0002, "epoch": 4.230862501010428, "step": 26170}, {"loss": 0.5406, "grad_norm": 1.152641773223877, "learning_rate": 0.0002, "epoch": 4.232479185191173, "step": 26180}, {"loss": 0.5632, "grad_norm": 0.9157698154449463, "learning_rate": 0.0002, "epoch": 4.234095869371918, "step": 26190}, {"loss": 0.5717, "grad_norm": 0.8418048620223999, "learning_rate": 0.0002, "epoch": 4.235712553552664, "step": 26200}, {"loss": 0.5624, "grad_norm": 0.9430168867111206, "learning_rate": 0.0002, "epoch": 4.237329237733409, "step": 26210}, {"loss": 0.5574, "grad_norm": 1.012582778930664, "learning_rate": 0.0002, "epoch": 4.238945921914154, "step": 26220}, {"loss": 0.5693, "grad_norm": 1.112619400024414, "learning_rate": 0.0002, "epoch": 4.240562606094899, "step": 26230}, {"loss": 0.6037, "grad_norm": 0.9243621826171875, "learning_rate": 0.0002, "epoch": 4.2421792902756446, "step": 26240}, {"loss": 0.569, "grad_norm": 0.6977595686912537, "learning_rate": 0.0002, "epoch": 4.24379597445639, "step": 26250}, {"loss": 0.5379, "grad_norm": 0.9600721597671509, "learning_rate": 0.0002, "epoch": 4.245412658637135, "step": 26260}, {"loss": 0.5658, "grad_norm": 0.882641613483429, "learning_rate": 0.0002, "epoch": 4.24702934281788, "step": 26270}, {"loss": 0.55, "grad_norm": 1.010920763015747, "learning_rate": 0.0002, "epoch": 4.2486460269986255, "step": 26280}, {"loss": 0.5803, "grad_norm": 0.9289400577545166, "learning_rate": 0.0002, "epoch": 4.250262711179371, "step": 26290}, {"loss": 0.541, "grad_norm": 1.137397289276123, "learning_rate": 0.0002, "epoch": 4.251879395360117, "step": 26300}, {"loss": 0.5204, "grad_norm": 1.0136182308197021, "learning_rate": 0.0002, "epoch": 4.253496079540862, "step": 26310}, {"loss": 0.5708, "grad_norm": 0.9387356042861938, "learning_rate": 0.0002, "epoch": 4.255112763721607, "step": 26320}, {"loss": 0.5948, "grad_norm": 1.1833957433700562, "learning_rate": 0.0002, "epoch": 4.2567294479023525, "step": 26330}, {"loss": 0.5905, "grad_norm": 0.9415934681892395, "learning_rate": 0.0002, "epoch": 4.258346132083098, "step": 26340}, {"loss": 0.5539, "grad_norm": 0.8550165891647339, "learning_rate": 0.0002, "epoch": 4.259962816263843, "step": 26350}, {"loss": 0.555, "grad_norm": 9.924622535705566, "learning_rate": 0.0002, "epoch": 4.261579500444588, "step": 26360}, {"loss": 0.5689, "grad_norm": 1.0104902982711792, "learning_rate": 0.0002, "epoch": 4.2631961846253335, "step": 26370}, {"loss": 0.5698, "grad_norm": 0.890794038772583, "learning_rate": 0.0002, "epoch": 4.264812868806079, "step": 26380}, {"loss": 0.563, "grad_norm": 1.0560191869735718, "learning_rate": 0.0002, "epoch": 4.266429552986824, "step": 26390}, {"loss": 0.5119, "grad_norm": 1.0135581493377686, "learning_rate": 0.0002, "epoch": 4.268046237167569, "step": 26400}, {"loss": 0.5359, "grad_norm": 1.1304140090942383, "learning_rate": 0.0002, "epoch": 4.269662921348314, "step": 26410}, {"loss": 0.5615, "grad_norm": 0.9899303913116455, "learning_rate": 0.0002, "epoch": 4.27127960552906, "step": 26420}, {"loss": 0.5815, "grad_norm": 1.0505329370498657, "learning_rate": 0.0002, "epoch": 4.272896289709805, "step": 26430}, {"loss": 0.5384, "grad_norm": 0.9389396905899048, "learning_rate": 0.0002, "epoch": 4.27451297389055, "step": 26440}, {"loss": 0.5558, "grad_norm": 0.875328779220581, "learning_rate": 0.0002, "epoch": 4.276129658071296, "step": 26450}, {"loss": 0.5601, "grad_norm": 1.0689256191253662, "learning_rate": 0.0002, "epoch": 4.277746342252041, "step": 26460}, {"loss": 0.546, "grad_norm": 0.9988957643508911, "learning_rate": 0.0002, "epoch": 4.279363026432787, "step": 26470}, {"loss": 0.5478, "grad_norm": 0.8721813559532166, "learning_rate": 0.0002, "epoch": 4.280979710613532, "step": 26480}, {"loss": 0.5424, "grad_norm": 1.100109577178955, "learning_rate": 0.0002, "epoch": 4.282596394794277, "step": 26490}, {"loss": 0.572, "grad_norm": 1.1607271432876587, "learning_rate": 0.0002, "epoch": 4.284213078975022, "step": 26500}, {"loss": 0.6287, "grad_norm": 0.879088819026947, "learning_rate": 0.0002, "epoch": 4.285829763155768, "step": 26510}, {"loss": 0.573, "grad_norm": 0.9891700744628906, "learning_rate": 0.0002, "epoch": 4.287446447336513, "step": 26520}, {"loss": 0.6018, "grad_norm": 1.0831127166748047, "learning_rate": 0.0002, "epoch": 4.289063131517258, "step": 26530}, {"loss": 0.5693, "grad_norm": 1.4108285903930664, "learning_rate": 0.0002, "epoch": 4.290679815698003, "step": 26540}, {"loss": 0.5888, "grad_norm": 1.0630289316177368, "learning_rate": 0.0002, "epoch": 4.2922964998787485, "step": 26550}, {"loss": 0.5817, "grad_norm": 1.0854572057724, "learning_rate": 0.0002, "epoch": 4.293913184059494, "step": 26560}, {"loss": 0.5586, "grad_norm": 0.9561646580696106, "learning_rate": 0.0002, "epoch": 4.295529868240239, "step": 26570}, {"loss": 0.5674, "grad_norm": 0.9064981937408447, "learning_rate": 0.0002, "epoch": 4.297146552420984, "step": 26580}, {"loss": 0.5847, "grad_norm": 1.0082972049713135, "learning_rate": 0.0002, "epoch": 4.298763236601729, "step": 26590}, {"loss": 0.5711, "grad_norm": 1.1613214015960693, "learning_rate": 0.0002, "epoch": 4.3003799207824756, "step": 26600}, {"loss": 0.551, "grad_norm": 0.9847695231437683, "learning_rate": 0.0002, "epoch": 4.301996604963221, "step": 26610}, {"loss": 0.6089, "grad_norm": 1.0980697870254517, "learning_rate": 0.0002, "epoch": 4.303613289143966, "step": 26620}, {"loss": 0.5797, "grad_norm": 0.8861175179481506, "learning_rate": 0.0002, "epoch": 4.305229973324711, "step": 26630}, {"loss": 0.5716, "grad_norm": 0.8917363286018372, "learning_rate": 0.0002, "epoch": 4.3068466575054565, "step": 26640}, {"loss": 0.5892, "grad_norm": 1.0458378791809082, "learning_rate": 0.0002, "epoch": 4.308463341686202, "step": 26650}, {"loss": 0.5883, "grad_norm": 1.4859240055084229, "learning_rate": 0.0002, "epoch": 4.310080025866947, "step": 26660}, {"loss": 0.5296, "grad_norm": 1.1376359462738037, "learning_rate": 0.0002, "epoch": 4.311696710047692, "step": 26670}, {"loss": 0.5671, "grad_norm": 0.991349995136261, "learning_rate": 0.0002, "epoch": 4.313313394228437, "step": 26680}, {"loss": 0.5338, "grad_norm": 0.9995543956756592, "learning_rate": 0.0002, "epoch": 4.314930078409183, "step": 26690}, {"loss": 0.5542, "grad_norm": 1.0515851974487305, "learning_rate": 0.0002, "epoch": 4.316546762589928, "step": 26700}, {"loss": 0.5473, "grad_norm": 1.008023977279663, "learning_rate": 0.0002, "epoch": 4.318163446770673, "step": 26710}, {"loss": 0.5506, "grad_norm": 1.0184582471847534, "learning_rate": 0.0002, "epoch": 4.319780130951418, "step": 26720}, {"loss": 0.5828, "grad_norm": 1.161071538925171, "learning_rate": 0.0002, "epoch": 4.321396815132164, "step": 26730}, {"loss": 0.5633, "grad_norm": 0.9580779671669006, "learning_rate": 0.0002, "epoch": 4.323013499312909, "step": 26740}, {"loss": 0.5785, "grad_norm": 1.0189911127090454, "learning_rate": 0.0002, "epoch": 4.324630183493655, "step": 26750}, {"loss": 0.5237, "grad_norm": 0.7484358549118042, "learning_rate": 0.0002, "epoch": 4.3262468676744, "step": 26760}, {"loss": 0.5728, "grad_norm": 1.0015908479690552, "learning_rate": 0.0002, "epoch": 4.327863551855145, "step": 26770}, {"loss": 0.5597, "grad_norm": 0.8972945809364319, "learning_rate": 0.0002, "epoch": 4.329480236035891, "step": 26780}, {"loss": 0.5857, "grad_norm": 1.01099693775177, "learning_rate": 0.0002, "epoch": 4.331096920216636, "step": 26790}, {"loss": 0.5591, "grad_norm": 0.846958339214325, "learning_rate": 0.0002, "epoch": 4.332713604397381, "step": 26800}, {"loss": 0.5547, "grad_norm": 1.0792603492736816, "learning_rate": 0.0002, "epoch": 4.334330288578126, "step": 26810}, {"loss": 0.5747, "grad_norm": 1.0373345613479614, "learning_rate": 0.0002, "epoch": 4.3359469727588715, "step": 26820}, {"loss": 0.558, "grad_norm": 0.9779167771339417, "learning_rate": 0.0002, "epoch": 4.337563656939617, "step": 26830}, {"loss": 0.5821, "grad_norm": 1.0235520601272583, "learning_rate": 0.0002, "epoch": 4.339180341120362, "step": 26840}, {"loss": 0.5843, "grad_norm": 1.04195237159729, "learning_rate": 0.0002, "epoch": 4.340797025301107, "step": 26850}, {"loss": 0.5474, "grad_norm": 0.9479565620422363, "learning_rate": 0.0002, "epoch": 4.3424137094818525, "step": 26860}, {"loss": 0.5646, "grad_norm": 0.9526172280311584, "learning_rate": 0.0002, "epoch": 4.344030393662598, "step": 26870}, {"loss": 0.521, "grad_norm": 0.8571456074714661, "learning_rate": 0.0002, "epoch": 4.345647077843343, "step": 26880}, {"loss": 0.5846, "grad_norm": 0.9475828409194946, "learning_rate": 0.0002, "epoch": 4.347263762024088, "step": 26890}, {"loss": 0.5815, "grad_norm": 1.0529576539993286, "learning_rate": 0.0002, "epoch": 4.348880446204834, "step": 26900}, {"loss": 0.56, "grad_norm": 0.9648140072822571, "learning_rate": 0.0002, "epoch": 4.3504971303855795, "step": 26910}, {"loss": 0.5162, "grad_norm": 1.0488841533660889, "learning_rate": 0.0002, "epoch": 4.352113814566325, "step": 26920}, {"loss": 0.5842, "grad_norm": 0.8771942257881165, "learning_rate": 0.0002, "epoch": 4.35373049874707, "step": 26930}, {"loss": 0.5966, "grad_norm": 0.9411202073097229, "learning_rate": 0.0002, "epoch": 4.355347182927815, "step": 26940}, {"loss": 0.6001, "grad_norm": 1.0997588634490967, "learning_rate": 0.0002, "epoch": 4.35696386710856, "step": 26950}, {"loss": 0.5528, "grad_norm": 0.968754768371582, "learning_rate": 0.0002, "epoch": 4.358580551289306, "step": 26960}, {"loss": 0.5881, "grad_norm": 0.9990773797035217, "learning_rate": 0.0002, "epoch": 4.360197235470051, "step": 26970}, {"loss": 0.5761, "grad_norm": 1.0210620164871216, "learning_rate": 0.0002, "epoch": 4.361813919650796, "step": 26980}, {"loss": 0.5768, "grad_norm": 0.855462908744812, "learning_rate": 0.0002, "epoch": 4.363430603831541, "step": 26990}, {"loss": 0.5493, "grad_norm": 0.9169660806655884, "learning_rate": 0.0002, "epoch": 4.365047288012287, "step": 27000}, {"loss": 0.5697, "grad_norm": 1.089629888534546, "learning_rate": 0.0002, "epoch": 4.366663972193032, "step": 27010}, {"loss": 0.5854, "grad_norm": 1.0932867527008057, "learning_rate": 0.0002, "epoch": 4.368280656373777, "step": 27020}, {"loss": 0.5656, "grad_norm": 0.9290956854820251, "learning_rate": 0.0002, "epoch": 4.369897340554522, "step": 27030}, {"loss": 0.5727, "grad_norm": 1.2800624370574951, "learning_rate": 0.0002, "epoch": 4.3715140247352675, "step": 27040}, {"loss": 0.5837, "grad_norm": 0.8993493318557739, "learning_rate": 0.0002, "epoch": 4.373130708916014, "step": 27050}, {"loss": 0.6232, "grad_norm": 1.1566431522369385, "learning_rate": 0.0002, "epoch": 4.374747393096759, "step": 27060}, {"loss": 0.5902, "grad_norm": 0.9479052424430847, "learning_rate": 0.0002, "epoch": 4.376364077277504, "step": 27070}, {"loss": 0.6189, "grad_norm": 1.0063648223876953, "learning_rate": 0.0002, "epoch": 4.377980761458249, "step": 27080}, {"loss": 0.561, "grad_norm": 0.8342045545578003, "learning_rate": 0.0002, "epoch": 4.379597445638995, "step": 27090}, {"loss": 0.5515, "grad_norm": 1.1390739679336548, "learning_rate": 0.0002, "epoch": 4.38121412981974, "step": 27100}, {"loss": 0.5372, "grad_norm": 0.9547637104988098, "learning_rate": 0.0002, "epoch": 4.382830814000485, "step": 27110}, {"loss": 0.5728, "grad_norm": 1.0503804683685303, "learning_rate": 0.0002, "epoch": 4.38444749818123, "step": 27120}, {"loss": 0.5787, "grad_norm": 0.9064017534255981, "learning_rate": 0.0002, "epoch": 4.3860641823619755, "step": 27130}, {"loss": 0.5798, "grad_norm": 0.9382519125938416, "learning_rate": 0.0002, "epoch": 4.387680866542721, "step": 27140}, {"loss": 0.5791, "grad_norm": 1.0410341024398804, "learning_rate": 0.0002, "epoch": 4.389297550723466, "step": 27150}, {"loss": 0.6034, "grad_norm": 0.9218655824661255, "learning_rate": 0.0002, "epoch": 4.390914234904211, "step": 27160}, {"loss": 0.5204, "grad_norm": 0.8119737505912781, "learning_rate": 0.0002, "epoch": 4.392530919084956, "step": 27170}, {"loss": 0.5612, "grad_norm": 0.8584722876548767, "learning_rate": 0.0002, "epoch": 4.394147603265702, "step": 27180}, {"loss": 0.5772, "grad_norm": 0.9668293595314026, "learning_rate": 0.0002, "epoch": 4.395764287446447, "step": 27190}, {"loss": 0.6009, "grad_norm": 1.022334098815918, "learning_rate": 0.0002, "epoch": 4.397380971627193, "step": 27200}, {"loss": 0.5573, "grad_norm": 0.9553216099739075, "learning_rate": 0.0002, "epoch": 4.398997655807938, "step": 27210}, {"loss": 0.5604, "grad_norm": 0.9282339215278625, "learning_rate": 0.0002, "epoch": 4.4006143399886835, "step": 27220}, {"loss": 0.5599, "grad_norm": 1.0232292413711548, "learning_rate": 0.0002, "epoch": 4.402231024169429, "step": 27230}, {"loss": 0.6078, "grad_norm": 0.9915700554847717, "learning_rate": 0.0002, "epoch": 4.403847708350174, "step": 27240}, {"loss": 0.5778, "grad_norm": 1.0014961957931519, "learning_rate": 0.0002, "epoch": 4.405464392530919, "step": 27250}, {"loss": 0.5824, "grad_norm": 1.1172103881835938, "learning_rate": 0.0002, "epoch": 4.407081076711664, "step": 27260}, {"loss": 0.5286, "grad_norm": 0.8583093285560608, "learning_rate": 0.0002, "epoch": 4.40869776089241, "step": 27270}, {"loss": 0.5507, "grad_norm": 0.7609201669692993, "learning_rate": 0.0002, "epoch": 4.410314445073155, "step": 27280}, {"loss": 0.575, "grad_norm": 1.0619351863861084, "learning_rate": 0.0002, "epoch": 4.4119311292539, "step": 27290}, {"loss": 0.5579, "grad_norm": 1.0177674293518066, "learning_rate": 0.0002, "epoch": 4.413547813434645, "step": 27300}, {"loss": 0.5628, "grad_norm": 0.9921218156814575, "learning_rate": 0.0002, "epoch": 4.4151644976153905, "step": 27310}, {"loss": 0.6018, "grad_norm": 1.126244306564331, "learning_rate": 0.0002, "epoch": 4.416781181796136, "step": 27320}, {"loss": 0.5743, "grad_norm": 1.0678540468215942, "learning_rate": 0.0002, "epoch": 4.418397865976881, "step": 27330}, {"loss": 0.5665, "grad_norm": 0.8705704212188721, "learning_rate": 0.0002, "epoch": 4.420014550157627, "step": 27340}, {"loss": 0.5763, "grad_norm": 1.272074818611145, "learning_rate": 0.0002, "epoch": 4.421631234338372, "step": 27350}, {"loss": 0.561, "grad_norm": 0.8740444183349609, "learning_rate": 0.0002, "epoch": 4.423247918519118, "step": 27360}, {"loss": 0.5492, "grad_norm": 1.0584250688552856, "learning_rate": 0.0002, "epoch": 4.424864602699863, "step": 27370}, {"loss": 0.589, "grad_norm": 1.059870719909668, "learning_rate": 0.0002, "epoch": 4.426481286880608, "step": 27380}, {"loss": 0.5551, "grad_norm": 1.072265863418579, "learning_rate": 0.0002, "epoch": 4.428097971061353, "step": 27390}, {"loss": 0.5584, "grad_norm": 0.871481716632843, "learning_rate": 0.0002, "epoch": 4.4297146552420985, "step": 27400}, {"loss": 0.5372, "grad_norm": 0.9555448293685913, "learning_rate": 0.0002, "epoch": 4.431331339422844, "step": 27410}, {"loss": 0.5593, "grad_norm": 1.0402292013168335, "learning_rate": 0.0002, "epoch": 4.432948023603589, "step": 27420}, {"loss": 0.5532, "grad_norm": 1.12587571144104, "learning_rate": 0.0002, "epoch": 4.434564707784334, "step": 27430}, {"loss": 0.5403, "grad_norm": 1.0783193111419678, "learning_rate": 0.0002, "epoch": 4.436181391965079, "step": 27440}, {"loss": 0.5313, "grad_norm": 1.024133563041687, "learning_rate": 0.0002, "epoch": 4.437798076145825, "step": 27450}, {"loss": 0.5621, "grad_norm": 0.9156768918037415, "learning_rate": 0.0002, "epoch": 4.43941476032657, "step": 27460}, {"loss": 0.5307, "grad_norm": 1.0215224027633667, "learning_rate": 0.0002, "epoch": 4.441031444507315, "step": 27470}, {"loss": 0.5188, "grad_norm": 1.082116961479187, "learning_rate": 0.0002, "epoch": 4.442648128688061, "step": 27480}, {"loss": 0.6203, "grad_norm": 1.0412873029708862, "learning_rate": 0.0002, "epoch": 4.4442648128688065, "step": 27490}, {"loss": 0.5939, "grad_norm": 1.0509289503097534, "learning_rate": 0.0002, "epoch": 4.445881497049552, "step": 27500}, {"loss": 0.5503, "grad_norm": 0.9291498064994812, "learning_rate": 0.0002, "epoch": 4.447498181230297, "step": 27510}, {"loss": 0.5408, "grad_norm": 0.970184326171875, "learning_rate": 0.0002, "epoch": 4.449114865411042, "step": 27520}, {"loss": 0.5705, "grad_norm": 0.8418883681297302, "learning_rate": 0.0002, "epoch": 4.450731549591787, "step": 27530}, {"loss": 0.5124, "grad_norm": 0.8823825120925903, "learning_rate": 0.0002, "epoch": 4.452348233772533, "step": 27540}, {"loss": 0.5867, "grad_norm": 1.1909019947052002, "learning_rate": 0.0002, "epoch": 4.453964917953278, "step": 27550}, {"loss": 0.5685, "grad_norm": 1.0317302942276, "learning_rate": 0.0002, "epoch": 4.455581602134023, "step": 27560}, {"loss": 0.5538, "grad_norm": 0.9977751970291138, "learning_rate": 0.0002, "epoch": 4.457198286314768, "step": 27570}, {"loss": 0.5628, "grad_norm": 0.8909519910812378, "learning_rate": 0.0002, "epoch": 4.458814970495514, "step": 27580}, {"loss": 0.6099, "grad_norm": 0.8653029799461365, "learning_rate": 0.0002, "epoch": 4.460431654676259, "step": 27590}, {"loss": 0.5622, "grad_norm": 1.0783653259277344, "learning_rate": 0.0002, "epoch": 4.462048338857004, "step": 27600}, {"loss": 0.579, "grad_norm": 1.1235394477844238, "learning_rate": 0.0002, "epoch": 4.463665023037749, "step": 27610}, {"loss": 0.5545, "grad_norm": 0.9386643767356873, "learning_rate": 0.0002, "epoch": 4.4652817072184945, "step": 27620}, {"loss": 0.5554, "grad_norm": 1.0605148077011108, "learning_rate": 0.0002, "epoch": 4.466898391399241, "step": 27630}, {"loss": 0.5886, "grad_norm": 1.1283893585205078, "learning_rate": 0.0002, "epoch": 4.468515075579986, "step": 27640}, {"loss": 0.5801, "grad_norm": 1.0583468675613403, "learning_rate": 0.0002, "epoch": 4.470131759760731, "step": 27650}, {"loss": 0.5601, "grad_norm": 0.9563992023468018, "learning_rate": 0.0002, "epoch": 4.471748443941476, "step": 27660}, {"loss": 0.5687, "grad_norm": 1.100598931312561, "learning_rate": 0.0002, "epoch": 4.4733651281222215, "step": 27670}, {"loss": 0.589, "grad_norm": 0.9386957287788391, "learning_rate": 0.0002, "epoch": 4.474981812302967, "step": 27680}, {"loss": 0.6241, "grad_norm": 1.2946288585662842, "learning_rate": 0.0002, "epoch": 4.476598496483712, "step": 27690}, {"loss": 0.6075, "grad_norm": 1.0325199365615845, "learning_rate": 0.0002, "epoch": 4.478215180664457, "step": 27700}, {"loss": 0.588, "grad_norm": 1.0318928956985474, "learning_rate": 0.0002, "epoch": 4.4798318648452025, "step": 27710}, {"loss": 0.5656, "grad_norm": 0.8721024394035339, "learning_rate": 0.0002, "epoch": 4.481448549025948, "step": 27720}, {"loss": 0.5421, "grad_norm": 1.17376708984375, "learning_rate": 0.0002, "epoch": 4.483065233206693, "step": 27730}, {"loss": 0.5657, "grad_norm": 1.0926326513290405, "learning_rate": 0.0002, "epoch": 4.484681917387438, "step": 27740}, {"loss": 0.5514, "grad_norm": 0.9043852686882019, "learning_rate": 0.0002, "epoch": 4.486298601568183, "step": 27750}, {"loss": 0.582, "grad_norm": 1.064600944519043, "learning_rate": 0.0002, "epoch": 4.487915285748929, "step": 27760}, {"loss": 0.6108, "grad_norm": 0.7833460569381714, "learning_rate": 0.0002, "epoch": 4.489531969929674, "step": 27770}, {"loss": 0.5985, "grad_norm": 1.1073496341705322, "learning_rate": 0.0002, "epoch": 4.49114865411042, "step": 27780}, {"loss": 0.5577, "grad_norm": 1.0799397230148315, "learning_rate": 0.0002, "epoch": 4.492765338291165, "step": 27790}, {"loss": 0.5601, "grad_norm": 1.1062238216400146, "learning_rate": 0.0002, "epoch": 4.49438202247191, "step": 27800}, {"loss": 0.6126, "grad_norm": 1.0568242073059082, "learning_rate": 0.0002, "epoch": 4.495998706652656, "step": 27810}, {"loss": 0.5913, "grad_norm": 0.8861091732978821, "learning_rate": 0.0002, "epoch": 4.497615390833401, "step": 27820}, {"loss": 0.5858, "grad_norm": 1.2297543287277222, "learning_rate": 0.0002, "epoch": 4.499232075014146, "step": 27830}, {"loss": 0.5859, "grad_norm": 0.9600302577018738, "learning_rate": 0.0002, "epoch": 4.500848759194891, "step": 27840}, {"loss": 0.6124, "grad_norm": 1.057051181793213, "learning_rate": 0.0002, "epoch": 4.502465443375637, "step": 27850}, {"loss": 0.5788, "grad_norm": 0.9839690923690796, "learning_rate": 0.0002, "epoch": 4.504082127556382, "step": 27860}, {"loss": 0.555, "grad_norm": 1.1479853391647339, "learning_rate": 0.0002, "epoch": 4.505698811737127, "step": 27870}, {"loss": 0.6039, "grad_norm": 1.0550768375396729, "learning_rate": 0.0002, "epoch": 4.507315495917872, "step": 27880}, {"loss": 0.563, "grad_norm": 0.898209273815155, "learning_rate": 0.0002, "epoch": 4.5089321800986175, "step": 27890}, {"loss": 0.5734, "grad_norm": 0.9460315108299255, "learning_rate": 0.0002, "epoch": 4.510548864279363, "step": 27900}, {"loss": 0.5702, "grad_norm": 0.9499884247779846, "learning_rate": 0.0002, "epoch": 4.512165548460108, "step": 27910}, {"loss": 0.5385, "grad_norm": 0.7801318764686584, "learning_rate": 0.0002, "epoch": 4.513782232640853, "step": 27920}, {"loss": 0.5391, "grad_norm": 0.9286966323852539, "learning_rate": 0.0002, "epoch": 4.515398916821599, "step": 27930}, {"loss": 0.5717, "grad_norm": 0.9539980292320251, "learning_rate": 0.0002, "epoch": 4.517015601002345, "step": 27940}, {"loss": 0.6073, "grad_norm": 1.1053401231765747, "learning_rate": 0.0002, "epoch": 4.51863228518309, "step": 27950}, {"loss": 0.6087, "grad_norm": 0.7535534501075745, "learning_rate": 0.0002, "epoch": 4.520248969363835, "step": 27960}, {"loss": 0.5701, "grad_norm": 1.076926589012146, "learning_rate": 0.0002, "epoch": 4.52186565354458, "step": 27970}, {"loss": 0.6028, "grad_norm": 1.181935429573059, "learning_rate": 0.0002, "epoch": 4.5234823377253255, "step": 27980}, {"loss": 0.6033, "grad_norm": 0.9293407201766968, "learning_rate": 0.0002, "epoch": 4.525099021906071, "step": 27990}, {"loss": 0.5815, "grad_norm": 0.8953009247779846, "learning_rate": 0.0002, "epoch": 4.526715706086816, "step": 28000}, {"loss": 0.5564, "grad_norm": 1.0850225687026978, "learning_rate": 0.0002, "epoch": 4.528332390267561, "step": 28010}, {"loss": 0.5459, "grad_norm": 0.9125663042068481, "learning_rate": 0.0002, "epoch": 4.529949074448306, "step": 28020}, {"loss": 0.5922, "grad_norm": 0.8745216727256775, "learning_rate": 0.0002, "epoch": 4.531565758629052, "step": 28030}, {"loss": 0.567, "grad_norm": 1.0783463716506958, "learning_rate": 0.0002, "epoch": 4.533182442809797, "step": 28040}, {"loss": 0.5754, "grad_norm": 0.7513844966888428, "learning_rate": 0.0002, "epoch": 4.534799126990542, "step": 28050}, {"loss": 0.5608, "grad_norm": 1.0135776996612549, "learning_rate": 0.0002, "epoch": 4.536415811171287, "step": 28060}, {"loss": 0.5827, "grad_norm": 0.8886825442314148, "learning_rate": 0.0002, "epoch": 4.538032495352033, "step": 28070}, {"loss": 0.5605, "grad_norm": 0.8153995275497437, "learning_rate": 0.0002, "epoch": 4.539649179532779, "step": 28080}, {"loss": 0.6377, "grad_norm": 0.9853341579437256, "learning_rate": 0.0002, "epoch": 4.541265863713524, "step": 28090}, {"loss": 0.5957, "grad_norm": 0.9365800023078918, "learning_rate": 0.0002, "epoch": 4.542882547894269, "step": 28100}, {"loss": 0.5477, "grad_norm": 0.9765017628669739, "learning_rate": 0.0002, "epoch": 4.544499232075014, "step": 28110}, {"loss": 0.6185, "grad_norm": 0.9811279773712158, "learning_rate": 0.0002, "epoch": 4.54611591625576, "step": 28120}, {"loss": 0.6095, "grad_norm": 1.0387924909591675, "learning_rate": 0.0002, "epoch": 4.547732600436505, "step": 28130}, {"loss": 0.6534, "grad_norm": 1.0684878826141357, "learning_rate": 0.0002, "epoch": 4.54934928461725, "step": 28140}, {"loss": 0.5701, "grad_norm": 1.0000102519989014, "learning_rate": 0.0002, "epoch": 4.550965968797995, "step": 28150}, {"loss": 0.5327, "grad_norm": 1.0717930793762207, "learning_rate": 0.0002, "epoch": 4.5525826529787405, "step": 28160}, {"loss": 0.5594, "grad_norm": 0.990074634552002, "learning_rate": 0.0002, "epoch": 4.554199337159486, "step": 28170}, {"loss": 0.5452, "grad_norm": 0.8673754930496216, "learning_rate": 0.0002, "epoch": 4.555816021340231, "step": 28180}, {"loss": 0.5773, "grad_norm": 0.864247739315033, "learning_rate": 0.0002, "epoch": 4.557432705520976, "step": 28190}, {"loss": 0.5516, "grad_norm": 0.8280200958251953, "learning_rate": 0.0002, "epoch": 4.5590493897017215, "step": 28200}, {"loss": 0.5709, "grad_norm": 1.1312172412872314, "learning_rate": 0.0002, "epoch": 4.560666073882467, "step": 28210}, {"loss": 0.5776, "grad_norm": 0.9147403240203857, "learning_rate": 0.0002, "epoch": 4.562282758063212, "step": 28220}, {"loss": 0.5591, "grad_norm": 1.0321218967437744, "learning_rate": 0.0002, "epoch": 4.563899442243958, "step": 28230}, {"loss": 0.5508, "grad_norm": 1.168332815170288, "learning_rate": 0.0002, "epoch": 4.565516126424703, "step": 28240}, {"loss": 0.5649, "grad_norm": 1.0067222118377686, "learning_rate": 0.0002, "epoch": 4.5671328106054485, "step": 28250}, {"loss": 0.5853, "grad_norm": 1.0283393859863281, "learning_rate": 0.0002, "epoch": 4.568749494786194, "step": 28260}, {"loss": 0.5772, "grad_norm": 0.9912363886833191, "learning_rate": 0.0002, "epoch": 4.570366178966939, "step": 28270}, {"loss": 0.5757, "grad_norm": 1.108032464981079, "learning_rate": 0.0002, "epoch": 4.571982863147684, "step": 28280}, {"loss": 0.5529, "grad_norm": 0.8260078430175781, "learning_rate": 0.0002, "epoch": 4.573599547328429, "step": 28290}, {"loss": 0.5625, "grad_norm": 0.8946247100830078, "learning_rate": 0.0002, "epoch": 4.575216231509175, "step": 28300}, {"loss": 0.5533, "grad_norm": 0.8273587822914124, "learning_rate": 0.0002, "epoch": 4.57683291568992, "step": 28310}, {"loss": 0.6058, "grad_norm": 0.9040093421936035, "learning_rate": 0.0002, "epoch": 4.578449599870665, "step": 28320}, {"loss": 0.5521, "grad_norm": 0.8435290455818176, "learning_rate": 0.0002, "epoch": 4.58006628405141, "step": 28330}, {"loss": 0.6086, "grad_norm": 1.164088249206543, "learning_rate": 0.0002, "epoch": 4.581682968232156, "step": 28340}, {"loss": 0.5603, "grad_norm": 0.9861085414886475, "learning_rate": 0.0002, "epoch": 4.583299652412901, "step": 28350}, {"loss": 0.5701, "grad_norm": 0.8892980813980103, "learning_rate": 0.0002, "epoch": 4.584916336593646, "step": 28360}, {"loss": 0.598, "grad_norm": 1.240574836730957, "learning_rate": 0.0002, "epoch": 4.586533020774391, "step": 28370}, {"loss": 0.5797, "grad_norm": 0.8669408559799194, "learning_rate": 0.0002, "epoch": 4.588149704955137, "step": 28380}, {"loss": 0.5603, "grad_norm": 0.9145985841751099, "learning_rate": 0.0002, "epoch": 4.589766389135883, "step": 28390}, {"loss": 0.5765, "grad_norm": 0.8584614992141724, "learning_rate": 0.0002, "epoch": 4.591383073316628, "step": 28400}, {"loss": 0.5898, "grad_norm": 1.118829369544983, "learning_rate": 0.0002, "epoch": 4.592999757497373, "step": 28410}, {"loss": 0.5641, "grad_norm": 1.1411553621292114, "learning_rate": 0.0002, "epoch": 4.594616441678118, "step": 28420}, {"loss": 0.549, "grad_norm": 0.9433278441429138, "learning_rate": 0.0002, "epoch": 4.596233125858864, "step": 28430}, {"loss": 0.5496, "grad_norm": 0.816830039024353, "learning_rate": 0.0002, "epoch": 4.597849810039609, "step": 28440}, {"loss": 0.5543, "grad_norm": 1.2124968767166138, "learning_rate": 0.0002, "epoch": 4.599466494220354, "step": 28450}, {"loss": 0.5759, "grad_norm": 0.9658762216567993, "learning_rate": 0.0002, "epoch": 4.601083178401099, "step": 28460}, {"loss": 0.5902, "grad_norm": 0.836100161075592, "learning_rate": 0.0002, "epoch": 4.6026998625818445, "step": 28470}, {"loss": 0.5749, "grad_norm": 0.9989104270935059, "learning_rate": 0.0002, "epoch": 4.60431654676259, "step": 28480}, {"loss": 0.5616, "grad_norm": 1.1298956871032715, "learning_rate": 0.0002, "epoch": 4.605933230943335, "step": 28490}, {"loss": 0.5846, "grad_norm": 1.1731704473495483, "learning_rate": 0.0002, "epoch": 4.60754991512408, "step": 28500}, {"loss": 0.5816, "grad_norm": 0.9624714255332947, "learning_rate": 0.0002, "epoch": 4.609166599304825, "step": 28510}, {"loss": 0.5868, "grad_norm": 1.364073634147644, "learning_rate": 0.0002, "epoch": 4.610783283485571, "step": 28520}, {"loss": 0.6237, "grad_norm": 1.1827356815338135, "learning_rate": 0.0002, "epoch": 4.612399967666317, "step": 28530}, {"loss": 0.5643, "grad_norm": 0.6651531457901001, "learning_rate": 0.0002, "epoch": 4.614016651847062, "step": 28540}, {"loss": 0.6051, "grad_norm": 1.1640995740890503, "learning_rate": 0.0002, "epoch": 4.615633336027807, "step": 28550}, {"loss": 0.5995, "grad_norm": 1.028918743133545, "learning_rate": 0.0002, "epoch": 4.6172500202085525, "step": 28560}, {"loss": 0.5607, "grad_norm": 0.8252120614051819, "learning_rate": 0.0002, "epoch": 4.618866704389298, "step": 28570}, {"loss": 0.5769, "grad_norm": 1.3536735773086548, "learning_rate": 0.0002, "epoch": 4.620483388570043, "step": 28580}, {"loss": 0.6006, "grad_norm": 1.2146915197372437, "learning_rate": 0.0002, "epoch": 4.622100072750788, "step": 28590}, {"loss": 0.5503, "grad_norm": 1.0122549533843994, "learning_rate": 0.0002, "epoch": 4.623716756931533, "step": 28600}, {"loss": 0.6072, "grad_norm": 0.9977872967720032, "learning_rate": 0.0002, "epoch": 4.625333441112279, "step": 28610}, {"loss": 0.5669, "grad_norm": 1.0159751176834106, "learning_rate": 0.0002, "epoch": 4.626950125293024, "step": 28620}, {"loss": 0.5935, "grad_norm": 1.0028325319290161, "learning_rate": 0.0002, "epoch": 4.628566809473769, "step": 28630}, {"loss": 0.5515, "grad_norm": 0.901638388633728, "learning_rate": 0.0002, "epoch": 4.630183493654514, "step": 28640}, {"loss": 0.595, "grad_norm": 0.9450507164001465, "learning_rate": 0.0002, "epoch": 4.6318001778352595, "step": 28650}, {"loss": 0.5972, "grad_norm": 0.9987545013427734, "learning_rate": 0.0002, "epoch": 4.633416862016006, "step": 28660}, {"loss": 0.5863, "grad_norm": 0.9574332237243652, "learning_rate": 0.0002, "epoch": 4.63503354619675, "step": 28670}, {"loss": 0.5804, "grad_norm": 1.2215653657913208, "learning_rate": 0.0002, "epoch": 4.636650230377496, "step": 28680}, {"loss": 0.5798, "grad_norm": 0.9798858761787415, "learning_rate": 0.0002, "epoch": 4.638266914558241, "step": 28690}, {"loss": 0.5773, "grad_norm": 1.0648466348648071, "learning_rate": 0.0002, "epoch": 4.639883598738987, "step": 28700}, {"loss": 0.6108, "grad_norm": 1.0606504678726196, "learning_rate": 0.0002, "epoch": 4.641500282919732, "step": 28710}, {"loss": 0.5801, "grad_norm": 1.0892442464828491, "learning_rate": 0.0002, "epoch": 4.643116967100477, "step": 28720}, {"loss": 0.5492, "grad_norm": 0.914391040802002, "learning_rate": 0.0002, "epoch": 4.644733651281222, "step": 28730}, {"loss": 0.5439, "grad_norm": 0.9782370328903198, "learning_rate": 0.0002, "epoch": 4.6463503354619675, "step": 28740}, {"loss": 0.6035, "grad_norm": 1.0344339609146118, "learning_rate": 0.0002, "epoch": 4.647967019642713, "step": 28750}, {"loss": 0.5775, "grad_norm": 1.0513931512832642, "learning_rate": 0.0002, "epoch": 4.649583703823458, "step": 28760}, {"loss": 0.546, "grad_norm": 0.9711475968360901, "learning_rate": 0.0002, "epoch": 4.651200388004203, "step": 28770}, {"loss": 0.5472, "grad_norm": 0.977519690990448, "learning_rate": 0.0002, "epoch": 4.652817072184948, "step": 28780}, {"loss": 0.5826, "grad_norm": 0.9150224924087524, "learning_rate": 0.0002, "epoch": 4.654433756365694, "step": 28790}, {"loss": 0.5382, "grad_norm": 1.0973542928695679, "learning_rate": 0.0002, "epoch": 4.656050440546439, "step": 28800}, {"loss": 0.6147, "grad_norm": 0.944877564907074, "learning_rate": 0.0002, "epoch": 4.657667124727185, "step": 28810}, {"loss": 0.5537, "grad_norm": 0.9508748650550842, "learning_rate": 0.0002, "epoch": 4.659283808907929, "step": 28820}, {"loss": 0.5537, "grad_norm": 0.9681721329689026, "learning_rate": 0.0002, "epoch": 4.6609004930886755, "step": 28830}, {"loss": 0.592, "grad_norm": 1.0214351415634155, "learning_rate": 0.0002, "epoch": 4.662517177269421, "step": 28840}, {"loss": 0.6031, "grad_norm": 0.9748611450195312, "learning_rate": 0.0002, "epoch": 4.664133861450166, "step": 28850}, {"loss": 0.572, "grad_norm": 0.8484147191047668, "learning_rate": 0.0002, "epoch": 4.665750545630911, "step": 28860}, {"loss": 0.5699, "grad_norm": 1.1252986192703247, "learning_rate": 0.0002, "epoch": 4.667367229811656, "step": 28870}, {"loss": 0.5724, "grad_norm": 0.8706206679344177, "learning_rate": 0.0002, "epoch": 4.668983913992402, "step": 28880}, {"loss": 0.6002, "grad_norm": 1.1432424783706665, "learning_rate": 0.0002, "epoch": 4.670600598173147, "step": 28890}, {"loss": 0.5675, "grad_norm": 1.017029047012329, "learning_rate": 0.0002, "epoch": 4.672217282353892, "step": 28900}, {"loss": 0.5831, "grad_norm": 1.085597038269043, "learning_rate": 0.0002, "epoch": 4.673833966534637, "step": 28910}, {"loss": 0.5678, "grad_norm": 0.9275796413421631, "learning_rate": 0.0002, "epoch": 4.675450650715383, "step": 28920}, {"loss": 0.5603, "grad_norm": 0.9518964886665344, "learning_rate": 0.0002, "epoch": 4.677067334896128, "step": 28930}, {"loss": 0.6232, "grad_norm": 1.0352122783660889, "learning_rate": 0.0002, "epoch": 4.678684019076873, "step": 28940}, {"loss": 0.5786, "grad_norm": 1.090124249458313, "learning_rate": 0.0002, "epoch": 4.680300703257618, "step": 28950}, {"loss": 0.5728, "grad_norm": 0.8799563050270081, "learning_rate": 0.0002, "epoch": 4.681917387438364, "step": 28960}, {"loss": 0.5787, "grad_norm": 1.0929821729660034, "learning_rate": 0.0002, "epoch": 4.683534071619109, "step": 28970}, {"loss": 0.6134, "grad_norm": 0.903727650642395, "learning_rate": 0.0002, "epoch": 4.685150755799855, "step": 28980}, {"loss": 0.5522, "grad_norm": 0.9752424955368042, "learning_rate": 0.0002, "epoch": 4.6867674399806, "step": 28990}, {"loss": 0.5762, "grad_norm": 0.9351571202278137, "learning_rate": 0.0002, "epoch": 4.688384124161345, "step": 29000}, {"loss": 0.5811, "grad_norm": 0.923877477645874, "learning_rate": 0.0002, "epoch": 4.6900008083420905, "step": 29010}, {"loss": 0.5682, "grad_norm": 1.045389175415039, "learning_rate": 0.0002, "epoch": 4.691617492522836, "step": 29020}, {"loss": 0.584, "grad_norm": 1.0200831890106201, "learning_rate": 0.0002, "epoch": 4.693234176703581, "step": 29030}, {"loss": 0.5514, "grad_norm": 1.1499706506729126, "learning_rate": 0.0002, "epoch": 4.694850860884326, "step": 29040}, {"loss": 0.5745, "grad_norm": 0.860118567943573, "learning_rate": 0.0002, "epoch": 4.6964675450650715, "step": 29050}, {"loss": 0.5741, "grad_norm": 0.9774864315986633, "learning_rate": 0.0002, "epoch": 4.698084229245817, "step": 29060}, {"loss": 0.5765, "grad_norm": 1.0323210954666138, "learning_rate": 0.0002, "epoch": 4.699700913426562, "step": 29070}, {"loss": 0.5452, "grad_norm": 0.8492481112480164, "learning_rate": 0.0002, "epoch": 4.701317597607307, "step": 29080}, {"loss": 0.5985, "grad_norm": 1.131951093673706, "learning_rate": 0.0002, "epoch": 4.702934281788052, "step": 29090}, {"loss": 0.6412, "grad_norm": 0.8763113021850586, "learning_rate": 0.0002, "epoch": 4.704550965968798, "step": 29100}, {"loss": 0.575, "grad_norm": 1.045028805732727, "learning_rate": 0.0002, "epoch": 4.706167650149544, "step": 29110}, {"loss": 0.5548, "grad_norm": 0.9961401224136353, "learning_rate": 0.0002, "epoch": 4.707784334330288, "step": 29120}, {"loss": 0.559, "grad_norm": 0.9282503724098206, "learning_rate": 0.0002, "epoch": 4.709401018511034, "step": 29130}, {"loss": 0.5744, "grad_norm": 1.1418932676315308, "learning_rate": 0.0002, "epoch": 4.711017702691779, "step": 29140}, {"loss": 0.5394, "grad_norm": 0.9950099587440491, "learning_rate": 0.0002, "epoch": 4.712634386872525, "step": 29150}, {"loss": 0.6177, "grad_norm": 0.8304893374443054, "learning_rate": 0.0002, "epoch": 4.71425107105327, "step": 29160}, {"loss": 0.6074, "grad_norm": 1.115626335144043, "learning_rate": 0.0002, "epoch": 4.715867755234015, "step": 29170}, {"loss": 0.6265, "grad_norm": 1.079818606376648, "learning_rate": 0.0002, "epoch": 4.71748443941476, "step": 29180}, {"loss": 0.561, "grad_norm": 1.1929082870483398, "learning_rate": 0.0002, "epoch": 4.719101123595506, "step": 29190}, {"loss": 0.5708, "grad_norm": 0.9621080756187439, "learning_rate": 0.0002, "epoch": 4.720717807776251, "step": 29200}, {"loss": 0.546, "grad_norm": 0.8549222350120544, "learning_rate": 0.0002, "epoch": 4.722334491956996, "step": 29210}, {"loss": 0.5775, "grad_norm": 0.9341941475868225, "learning_rate": 0.0002, "epoch": 4.723951176137741, "step": 29220}, {"loss": 0.5436, "grad_norm": 1.075406789779663, "learning_rate": 0.0002, "epoch": 4.7255678603184865, "step": 29230}, {"loss": 0.576, "grad_norm": 1.0859880447387695, "learning_rate": 0.0002, "epoch": 4.727184544499232, "step": 29240}, {"loss": 0.5525, "grad_norm": 0.8475605249404907, "learning_rate": 0.0002, "epoch": 4.728801228679977, "step": 29250}, {"loss": 0.5659, "grad_norm": 0.9331845641136169, "learning_rate": 0.0002, "epoch": 4.730417912860723, "step": 29260}, {"loss": 0.5901, "grad_norm": 0.9279314279556274, "learning_rate": 0.0002, "epoch": 4.7320345970414674, "step": 29270}, {"loss": 0.597, "grad_norm": 0.7803558707237244, "learning_rate": 0.0002, "epoch": 4.733651281222214, "step": 29280}, {"loss": 0.5968, "grad_norm": 1.0159329175949097, "learning_rate": 0.0002, "epoch": 4.735267965402959, "step": 29290}, {"loss": 0.5333, "grad_norm": 0.9448670744895935, "learning_rate": 0.0002, "epoch": 4.736884649583704, "step": 29300}, {"loss": 0.574, "grad_norm": 1.0732197761535645, "learning_rate": 0.0002, "epoch": 4.738501333764449, "step": 29310}, {"loss": 0.6066, "grad_norm": 0.901830792427063, "learning_rate": 0.0002, "epoch": 4.7401180179451945, "step": 29320}, {"loss": 0.6105, "grad_norm": 0.9141789674758911, "learning_rate": 0.0002, "epoch": 4.74173470212594, "step": 29330}, {"loss": 0.5481, "grad_norm": 0.9733418226242065, "learning_rate": 0.0002, "epoch": 4.743351386306685, "step": 29340}, {"loss": 0.612, "grad_norm": 0.909810483455658, "learning_rate": 0.0002, "epoch": 4.74496807048743, "step": 29350}, {"loss": 0.5911, "grad_norm": 0.909541666507721, "learning_rate": 0.0002, "epoch": 4.746584754668175, "step": 29360}, {"loss": 0.5579, "grad_norm": 0.9383015632629395, "learning_rate": 0.0002, "epoch": 4.748201438848921, "step": 29370}, {"loss": 0.5529, "grad_norm": 0.9275668263435364, "learning_rate": 0.0002, "epoch": 4.749818123029666, "step": 29380}, {"loss": 0.5623, "grad_norm": 1.1146225929260254, "learning_rate": 0.0002, "epoch": 4.751434807210411, "step": 29390}, {"loss": 0.6018, "grad_norm": 1.0062453746795654, "learning_rate": 0.0002, "epoch": 4.753051491391156, "step": 29400}, {"loss": 0.5872, "grad_norm": 0.9451895952224731, "learning_rate": 0.0002, "epoch": 4.7546681755719025, "step": 29410}, {"loss": 0.5767, "grad_norm": 0.870457649230957, "learning_rate": 0.0002, "epoch": 4.756284859752648, "step": 29420}, {"loss": 0.57, "grad_norm": 1.0411282777786255, "learning_rate": 0.0002, "epoch": 4.757901543933393, "step": 29430}, {"loss": 0.5688, "grad_norm": 1.1648986339569092, "learning_rate": 0.0002, "epoch": 4.759518228114138, "step": 29440}, {"loss": 0.5432, "grad_norm": 0.8999572992324829, "learning_rate": 0.0002, "epoch": 4.761134912294883, "step": 29450}, {"loss": 0.5667, "grad_norm": 0.9863559007644653, "learning_rate": 0.0002, "epoch": 4.762751596475629, "step": 29460}, {"loss": 0.5779, "grad_norm": 0.9676542282104492, "learning_rate": 0.0002, "epoch": 4.764368280656374, "step": 29470}, {"loss": 0.6075, "grad_norm": 1.004775047302246, "learning_rate": 0.0002, "epoch": 4.765984964837119, "step": 29480}, {"loss": 0.6044, "grad_norm": 1.0937515497207642, "learning_rate": 0.0002, "epoch": 4.767601649017864, "step": 29490}, {"loss": 0.5433, "grad_norm": 0.9551598429679871, "learning_rate": 0.0002, "epoch": 4.7692183331986095, "step": 29500}, {"loss": 0.5609, "grad_norm": 1.0757228136062622, "learning_rate": 0.0002, "epoch": 4.770835017379355, "step": 29510}, {"loss": 0.567, "grad_norm": 1.0588841438293457, "learning_rate": 0.0002, "epoch": 4.7724517015601, "step": 29520}, {"loss": 0.5814, "grad_norm": 1.0744032859802246, "learning_rate": 0.0002, "epoch": 4.774068385740845, "step": 29530}, {"loss": 0.5681, "grad_norm": 1.0066277980804443, "learning_rate": 0.0002, "epoch": 4.7756850699215905, "step": 29540}, {"loss": 0.545, "grad_norm": 1.082319736480713, "learning_rate": 0.0002, "epoch": 4.777301754102336, "step": 29550}, {"loss": 0.5709, "grad_norm": 0.8252472877502441, "learning_rate": 0.0002, "epoch": 4.778918438283082, "step": 29560}, {"loss": 0.5666, "grad_norm": 0.9855340123176575, "learning_rate": 0.0002, "epoch": 4.780535122463827, "step": 29570}, {"loss": 0.6117, "grad_norm": 0.9991421699523926, "learning_rate": 0.0002, "epoch": 4.782151806644572, "step": 29580}, {"loss": 0.5966, "grad_norm": 1.316841959953308, "learning_rate": 0.0002, "epoch": 4.7837684908253175, "step": 29590}, {"loss": 0.6102, "grad_norm": 1.1513035297393799, "learning_rate": 0.0002, "epoch": 4.785385175006063, "step": 29600}, {"loss": 0.5785, "grad_norm": 0.9767683744430542, "learning_rate": 0.0002, "epoch": 4.787001859186808, "step": 29610}, {"loss": 0.6037, "grad_norm": 0.9786278605461121, "learning_rate": 0.0002, "epoch": 4.788618543367553, "step": 29620}, {"loss": 0.6108, "grad_norm": 0.8004973530769348, "learning_rate": 0.0002, "epoch": 4.7902352275482984, "step": 29630}, {"loss": 0.5932, "grad_norm": 1.0997767448425293, "learning_rate": 0.0002, "epoch": 4.791851911729044, "step": 29640}, {"loss": 0.5655, "grad_norm": 0.9752856492996216, "learning_rate": 0.0002, "epoch": 4.793468595909789, "step": 29650}, {"loss": 0.5916, "grad_norm": 1.0518392324447632, "learning_rate": 0.0002, "epoch": 4.795085280090534, "step": 29660}, {"loss": 0.6042, "grad_norm": 1.1050055027008057, "learning_rate": 0.0002, "epoch": 4.796701964271279, "step": 29670}, {"loss": 0.6089, "grad_norm": 0.9933857917785645, "learning_rate": 0.0002, "epoch": 4.798318648452025, "step": 29680}, {"loss": 0.6041, "grad_norm": 1.2804018259048462, "learning_rate": 0.0002, "epoch": 4.79993533263277, "step": 29690}, {"loss": 0.636, "grad_norm": 1.0133371353149414, "learning_rate": 0.0002, "epoch": 4.801552016813515, "step": 29700}, {"loss": 0.5662, "grad_norm": 1.080350637435913, "learning_rate": 0.0002, "epoch": 4.803168700994261, "step": 29710}, {"loss": 0.5603, "grad_norm": 0.9986529350280762, "learning_rate": 0.0002, "epoch": 4.804785385175006, "step": 29720}, {"loss": 0.5894, "grad_norm": 0.975665807723999, "learning_rate": 0.0002, "epoch": 4.806402069355752, "step": 29730}, {"loss": 0.6328, "grad_norm": 0.8458138704299927, "learning_rate": 0.0002, "epoch": 4.808018753536497, "step": 29740}, {"loss": 0.5837, "grad_norm": 0.99330073595047, "learning_rate": 0.0002, "epoch": 4.809635437717242, "step": 29750}, {"loss": 0.5507, "grad_norm": 0.898274302482605, "learning_rate": 0.0002, "epoch": 4.811252121897987, "step": 29760}, {"loss": 0.5842, "grad_norm": 1.0504480600357056, "learning_rate": 0.0002, "epoch": 4.812868806078733, "step": 29770}, {"loss": 0.5821, "grad_norm": 0.937919020652771, "learning_rate": 0.0002, "epoch": 4.814485490259478, "step": 29780}, {"loss": 0.5885, "grad_norm": 0.9593307971954346, "learning_rate": 0.0002, "epoch": 4.816102174440223, "step": 29790}, {"loss": 0.578, "grad_norm": 0.9431198835372925, "learning_rate": 0.0002, "epoch": 4.817718858620968, "step": 29800}, {"loss": 0.5739, "grad_norm": 1.2729957103729248, "learning_rate": 0.0002, "epoch": 4.8193355428017135, "step": 29810}, {"loss": 0.6124, "grad_norm": 0.8876838684082031, "learning_rate": 0.0002, "epoch": 4.820952226982459, "step": 29820}, {"loss": 0.5583, "grad_norm": 1.0185000896453857, "learning_rate": 0.0002, "epoch": 4.822568911163204, "step": 29830}, {"loss": 0.5686, "grad_norm": 1.064276099205017, "learning_rate": 0.0002, "epoch": 4.824185595343949, "step": 29840}, {"loss": 0.5698, "grad_norm": 0.9774803519248962, "learning_rate": 0.0002, "epoch": 4.825802279524694, "step": 29850}, {"loss": 0.5533, "grad_norm": 1.131646990776062, "learning_rate": 0.0002, "epoch": 4.8274189637054405, "step": 29860}, {"loss": 0.6371, "grad_norm": 1.081455945968628, "learning_rate": 0.0002, "epoch": 4.829035647886186, "step": 29870}, {"loss": 0.5793, "grad_norm": 0.990538477897644, "learning_rate": 0.0002, "epoch": 4.830652332066931, "step": 29880}, {"loss": 0.5833, "grad_norm": 0.9750600457191467, "learning_rate": 0.0002, "epoch": 4.832269016247676, "step": 29890}, {"loss": 0.619, "grad_norm": 1.0600621700286865, "learning_rate": 0.0002, "epoch": 4.8338857004284215, "step": 29900}, {"loss": 0.5841, "grad_norm": 0.9237320423126221, "learning_rate": 0.0002, "epoch": 4.835502384609167, "step": 29910}, {"loss": 0.5513, "grad_norm": 0.9739177227020264, "learning_rate": 0.0002, "epoch": 4.837119068789912, "step": 29920}, {"loss": 0.587, "grad_norm": 1.128677248954773, "learning_rate": 0.0002, "epoch": 4.838735752970657, "step": 29930}, {"loss": 0.564, "grad_norm": 1.042604923248291, "learning_rate": 0.0002, "epoch": 4.840352437151402, "step": 29940}, {"loss": 0.5885, "grad_norm": 0.849758505821228, "learning_rate": 0.0002, "epoch": 4.841969121332148, "step": 29950}, {"loss": 0.5952, "grad_norm": 1.2809888124465942, "learning_rate": 0.0002, "epoch": 4.843585805512893, "step": 29960}, {"loss": 0.5703, "grad_norm": 1.0177865028381348, "learning_rate": 0.0002, "epoch": 4.845202489693638, "step": 29970}, {"loss": 0.5946, "grad_norm": 1.0026639699935913, "learning_rate": 0.0002, "epoch": 4.846819173874383, "step": 29980}, {"loss": 0.5897, "grad_norm": 0.9679505228996277, "learning_rate": 0.0002, "epoch": 4.8484358580551286, "step": 29990}, {"loss": 0.5621, "grad_norm": 0.8939532041549683, "learning_rate": 0.0002, "epoch": 4.850052542235874, "step": 30000}, {"loss": 0.5852, "grad_norm": 0.9957457780838013, "learning_rate": 0.0002, "epoch": 4.85166922641662, "step": 30010}, {"loss": 0.6117, "grad_norm": 1.1646790504455566, "learning_rate": 0.0002, "epoch": 4.853285910597365, "step": 30020}, {"loss": 0.5711, "grad_norm": 0.8804680705070496, "learning_rate": 0.0002, "epoch": 4.85490259477811, "step": 30030}, {"loss": 0.5397, "grad_norm": 1.161970853805542, "learning_rate": 0.0002, "epoch": 4.856519278958856, "step": 30040}, {"loss": 0.5552, "grad_norm": 0.9081037640571594, "learning_rate": 0.0002, "epoch": 4.858135963139601, "step": 30050}, {"loss": 0.6024, "grad_norm": 0.9402848482131958, "learning_rate": 0.0002, "epoch": 4.859752647320346, "step": 30060}, {"loss": 0.6256, "grad_norm": 0.9023865461349487, "learning_rate": 0.0002, "epoch": 4.861369331501091, "step": 30070}, {"loss": 0.5926, "grad_norm": 1.0173414945602417, "learning_rate": 0.0002, "epoch": 4.8629860156818365, "step": 30080}, {"loss": 0.6274, "grad_norm": 1.084402322769165, "learning_rate": 0.0002, "epoch": 4.864602699862582, "step": 30090}, {"loss": 0.6311, "grad_norm": 0.9577937126159668, "learning_rate": 0.0002, "epoch": 4.866219384043327, "step": 30100}, {"loss": 0.5724, "grad_norm": 0.9807606935501099, "learning_rate": 0.0002, "epoch": 4.867836068224072, "step": 30110}, {"loss": 0.5786, "grad_norm": 0.978784441947937, "learning_rate": 0.0002, "epoch": 4.8694527524048175, "step": 30120}, {"loss": 0.6194, "grad_norm": 0.9762914776802063, "learning_rate": 0.0002, "epoch": 4.871069436585563, "step": 30130}, {"loss": 0.5892, "grad_norm": 0.9404871463775635, "learning_rate": 0.0002, "epoch": 4.872686120766308, "step": 30140}, {"loss": 0.6182, "grad_norm": 1.0069509744644165, "learning_rate": 0.0002, "epoch": 4.874302804947053, "step": 30150}, {"loss": 0.6225, "grad_norm": 1.1770923137664795, "learning_rate": 0.0002, "epoch": 4.875919489127799, "step": 30160}, {"loss": 0.5657, "grad_norm": 1.021210789680481, "learning_rate": 0.0002, "epoch": 4.8775361733085445, "step": 30170}, {"loss": 0.6033, "grad_norm": 0.8512648940086365, "learning_rate": 0.0002, "epoch": 4.87915285748929, "step": 30180}, {"loss": 0.5519, "grad_norm": 0.9345870018005371, "learning_rate": 0.0002, "epoch": 4.880769541670035, "step": 30190}, {"loss": 0.5682, "grad_norm": 1.0224418640136719, "learning_rate": 0.0002, "epoch": 4.88238622585078, "step": 30200}, {"loss": 0.5807, "grad_norm": 1.0316044092178345, "learning_rate": 0.0002, "epoch": 4.884002910031525, "step": 30210}, {"loss": 0.6065, "grad_norm": 1.102437973022461, "learning_rate": 0.0002, "epoch": 4.885619594212271, "step": 30220}, {"loss": 0.586, "grad_norm": 1.0220023393630981, "learning_rate": 0.0002, "epoch": 4.887236278393016, "step": 30230}, {"loss": 0.5781, "grad_norm": 1.0934523344039917, "learning_rate": 0.0002, "epoch": 4.888852962573761, "step": 30240}, {"loss": 0.6313, "grad_norm": 1.264630913734436, "learning_rate": 0.0002, "epoch": 4.890469646754506, "step": 30250}, {"loss": 0.5712, "grad_norm": 1.0999879837036133, "learning_rate": 0.0002, "epoch": 4.892086330935252, "step": 30260}, {"loss": 0.6413, "grad_norm": 0.9124550223350525, "learning_rate": 0.0002, "epoch": 4.893703015115997, "step": 30270}, {"loss": 0.596, "grad_norm": 0.9853624105453491, "learning_rate": 0.0002, "epoch": 4.895319699296742, "step": 30280}, {"loss": 0.595, "grad_norm": 1.0589802265167236, "learning_rate": 0.0002, "epoch": 4.896936383477488, "step": 30290}, {"loss": 0.6129, "grad_norm": 0.8487226366996765, "learning_rate": 0.0002, "epoch": 4.8985530676582325, "step": 30300}, {"loss": 0.5514, "grad_norm": 1.0212191343307495, "learning_rate": 0.0002, "epoch": 4.900169751838979, "step": 30310}, {"loss": 0.5896, "grad_norm": 1.0187491178512573, "learning_rate": 0.0002, "epoch": 4.901786436019724, "step": 30320}, {"loss": 0.5809, "grad_norm": 1.0013091564178467, "learning_rate": 0.0002, "epoch": 4.903403120200469, "step": 30330}, {"loss": 0.5658, "grad_norm": 1.0017542839050293, "learning_rate": 0.0002, "epoch": 4.905019804381214, "step": 30340}, {"loss": 0.6002, "grad_norm": 0.9665151238441467, "learning_rate": 0.0002, "epoch": 4.9066364885619596, "step": 30350}, {"loss": 0.5864, "grad_norm": 0.8774822950363159, "learning_rate": 0.0002, "epoch": 4.908253172742705, "step": 30360}, {"loss": 0.5771, "grad_norm": 0.9449850916862488, "learning_rate": 0.0002, "epoch": 4.90986985692345, "step": 30370}, {"loss": 0.58, "grad_norm": 0.7368341088294983, "learning_rate": 0.0002, "epoch": 4.911486541104195, "step": 30380}, {"loss": 0.5992, "grad_norm": 0.9669167995452881, "learning_rate": 0.0002, "epoch": 4.9131032252849405, "step": 30390}, {"loss": 0.6202, "grad_norm": 1.1227794885635376, "learning_rate": 0.0002, "epoch": 4.914719909465686, "step": 30400}, {"loss": 0.6181, "grad_norm": 0.9884361028671265, "learning_rate": 0.0002, "epoch": 4.916336593646431, "step": 30410}, {"loss": 0.6185, "grad_norm": 0.9949551224708557, "learning_rate": 0.0002, "epoch": 4.917953277827176, "step": 30420}, {"loss": 0.5866, "grad_norm": 0.9491621851921082, "learning_rate": 0.0002, "epoch": 4.919569962007921, "step": 30430}, {"loss": 0.6005, "grad_norm": 0.78848797082901, "learning_rate": 0.0002, "epoch": 4.9211866461886675, "step": 30440}, {"loss": 0.5561, "grad_norm": 1.0693835020065308, "learning_rate": 0.0002, "epoch": 4.922803330369412, "step": 30450}, {"loss": 0.566, "grad_norm": 0.9573729634284973, "learning_rate": 0.0002, "epoch": 4.924420014550158, "step": 30460}, {"loss": 0.6084, "grad_norm": 0.9975152611732483, "learning_rate": 0.0002, "epoch": 4.926036698730903, "step": 30470}, {"loss": 0.5969, "grad_norm": 0.8695693016052246, "learning_rate": 0.0002, "epoch": 4.9276533829116484, "step": 30480}, {"loss": 0.6144, "grad_norm": 1.145394206047058, "learning_rate": 0.0002, "epoch": 4.929270067092394, "step": 30490}, {"loss": 0.5736, "grad_norm": 0.7668989896774292, "learning_rate": 0.0002, "epoch": 4.930886751273139, "step": 30500}, {"loss": 0.6052, "grad_norm": 0.9630151391029358, "learning_rate": 0.0002, "epoch": 4.932503435453884, "step": 30510}, {"loss": 0.6461, "grad_norm": 0.940705418586731, "learning_rate": 0.0002, "epoch": 4.934120119634629, "step": 30520}, {"loss": 0.6326, "grad_norm": 1.3243348598480225, "learning_rate": 0.0002, "epoch": 4.935736803815375, "step": 30530}, {"loss": 0.6174, "grad_norm": 1.004347801208496, "learning_rate": 0.0002, "epoch": 4.93735348799612, "step": 30540}, {"loss": 0.583, "grad_norm": 0.8711541295051575, "learning_rate": 0.0002, "epoch": 4.938970172176865, "step": 30550}, {"loss": 0.599, "grad_norm": 0.8980631828308105, "learning_rate": 0.0002, "epoch": 4.94058685635761, "step": 30560}, {"loss": 0.6024, "grad_norm": 0.8388893604278564, "learning_rate": 0.0002, "epoch": 4.9422035405383555, "step": 30570}, {"loss": 0.6189, "grad_norm": 1.0991183519363403, "learning_rate": 0.0002, "epoch": 4.943820224719101, "step": 30580}, {"loss": 0.5906, "grad_norm": 0.9731075763702393, "learning_rate": 0.0002, "epoch": 4.945436908899847, "step": 30590}, {"loss": 0.5883, "grad_norm": 1.3904452323913574, "learning_rate": 0.0002, "epoch": 4.947053593080591, "step": 30600}, {"loss": 0.5952, "grad_norm": 1.2489882707595825, "learning_rate": 0.0002, "epoch": 4.948670277261337, "step": 30610}, {"loss": 0.5887, "grad_norm": 1.240072250366211, "learning_rate": 0.0002, "epoch": 4.950286961442083, "step": 30620}, {"loss": 0.5762, "grad_norm": 0.9191411733627319, "learning_rate": 0.0002, "epoch": 4.951903645622828, "step": 30630}, {"loss": 0.5597, "grad_norm": 0.8888895511627197, "learning_rate": 0.0002, "epoch": 4.953520329803573, "step": 30640}, {"loss": 0.6594, "grad_norm": 0.9001450538635254, "learning_rate": 0.0002, "epoch": 4.955137013984318, "step": 30650}, {"loss": 0.6047, "grad_norm": 1.053971767425537, "learning_rate": 0.0002, "epoch": 4.9567536981650635, "step": 30660}, {"loss": 0.6107, "grad_norm": 1.2224042415618896, "learning_rate": 0.0002, "epoch": 4.958370382345809, "step": 30670}, {"loss": 0.6211, "grad_norm": 0.8855111598968506, "learning_rate": 0.0002, "epoch": 4.959987066526554, "step": 30680}, {"loss": 0.5764, "grad_norm": 0.9489575624465942, "learning_rate": 0.0002, "epoch": 4.961603750707299, "step": 30690}, {"loss": 0.5371, "grad_norm": 0.9635404944419861, "learning_rate": 0.0002, "epoch": 4.963220434888044, "step": 30700}, {"loss": 0.6043, "grad_norm": 1.1784121990203857, "learning_rate": 0.0002, "epoch": 4.96483711906879, "step": 30710}, {"loss": 0.5803, "grad_norm": 1.0059462785720825, "learning_rate": 0.0002, "epoch": 4.966453803249535, "step": 30720}, {"loss": 0.5759, "grad_norm": 0.9479738473892212, "learning_rate": 0.0002, "epoch": 4.96807048743028, "step": 30730}, {"loss": 0.584, "grad_norm": 1.0624593496322632, "learning_rate": 0.0002, "epoch": 4.969687171611026, "step": 30740}, {"loss": 0.6202, "grad_norm": 1.1429259777069092, "learning_rate": 0.0002, "epoch": 4.971303855791771, "step": 30750}, {"loss": 0.6174, "grad_norm": 0.9102491140365601, "learning_rate": 0.0002, "epoch": 4.972920539972517, "step": 30760}, {"loss": 0.6025, "grad_norm": 1.1262688636779785, "learning_rate": 0.0002, "epoch": 4.974537224153262, "step": 30770}, {"loss": 0.588, "grad_norm": 1.1415393352508545, "learning_rate": 0.0002, "epoch": 4.976153908334007, "step": 30780}, {"loss": 0.5832, "grad_norm": 1.083078384399414, "learning_rate": 0.0002, "epoch": 4.977770592514752, "step": 30790}, {"loss": 0.6025, "grad_norm": 0.964859127998352, "learning_rate": 0.0002, "epoch": 4.979387276695498, "step": 30800}, {"loss": 0.6095, "grad_norm": 0.8704743385314941, "learning_rate": 0.0002, "epoch": 4.981003960876243, "step": 30810}, {"loss": 0.5666, "grad_norm": 1.0714856386184692, "learning_rate": 0.0002, "epoch": 4.982620645056988, "step": 30820}, {"loss": 0.565, "grad_norm": 0.6818771362304688, "learning_rate": 0.0002, "epoch": 4.984237329237733, "step": 30830}, {"loss": 0.5999, "grad_norm": 1.0454156398773193, "learning_rate": 0.0002, "epoch": 4.985854013418479, "step": 30840}, {"loss": 0.5683, "grad_norm": 0.9410776495933533, "learning_rate": 0.0002, "epoch": 4.987470697599224, "step": 30850}, {"loss": 0.5899, "grad_norm": 1.0878902673721313, "learning_rate": 0.0002, "epoch": 4.989087381779969, "step": 30860}, {"loss": 0.5914, "grad_norm": 0.8916727304458618, "learning_rate": 0.0002, "epoch": 4.990704065960714, "step": 30870}, {"loss": 0.6066, "grad_norm": 1.045776128768921, "learning_rate": 0.0002, "epoch": 4.9923207501414595, "step": 30880}, {"loss": 0.5767, "grad_norm": 0.9861903786659241, "learning_rate": 0.0002, "epoch": 4.993937434322206, "step": 30890}, {"loss": 0.6192, "grad_norm": 0.9275050759315491, "learning_rate": 0.0002, "epoch": 4.995554118502951, "step": 30900}, {"loss": 0.6181, "grad_norm": 0.94013911485672, "learning_rate": 0.0002, "epoch": 4.997170802683696, "step": 30910}, {"loss": 0.614, "grad_norm": 0.9771268367767334, "learning_rate": 0.0002, "epoch": 4.998787486864441, "step": 30920}, {"eval_loss": 1.1968598365783691, "eval_runtime": 122.2519, "eval_samples_per_second": 5.996, "eval_steps_per_second": 0.753, "epoch": 4.9999191657909625, "step": 30927}, {"loss": 0.5238, "grad_norm": 0.8021580576896667, "learning_rate": 0.0002, "epoch": 5.0004041710451865, "step": 30930}, {"loss": 0.4984, "grad_norm": 1.0807327032089233, "learning_rate": 0.0002, "epoch": 5.002020855225932, "step": 30940}, {"loss": 0.514, "grad_norm": 1.1638425588607788, "learning_rate": 0.0002, "epoch": 5.003637539406677, "step": 30950}, {"loss": 0.4621, "grad_norm": 1.1700230836868286, "learning_rate": 0.0002, "epoch": 5.005254223587422, "step": 30960}, {"loss": 0.4657, "grad_norm": 0.9053420424461365, "learning_rate": 0.0002, "epoch": 5.0068709077681675, "step": 30970}, {"loss": 0.4865, "grad_norm": 0.9226111769676208, "learning_rate": 0.0002, "epoch": 5.008487591948913, "step": 30980}, {"loss": 0.5011, "grad_norm": 1.238669514656067, "learning_rate": 0.0002, "epoch": 5.010104276129658, "step": 30990}, {"loss": 0.4754, "grad_norm": 1.0668327808380127, "learning_rate": 0.0002, "epoch": 5.011720960310403, "step": 31000}, {"loss": 0.5414, "grad_norm": 1.0903944969177246, "learning_rate": 0.0002, "epoch": 5.013337644491148, "step": 31010}, {"loss": 0.5117, "grad_norm": 1.0763911008834839, "learning_rate": 0.0002, "epoch": 5.014954328671894, "step": 31020}, {"loss": 0.4908, "grad_norm": 1.0108771324157715, "learning_rate": 0.0002, "epoch": 5.016571012852639, "step": 31030}, {"loss": 0.5052, "grad_norm": 0.8816103935241699, "learning_rate": 0.0002, "epoch": 5.018187697033385, "step": 31040}, {"loss": 0.4985, "grad_norm": 1.11434805393219, "learning_rate": 0.0002, "epoch": 5.01980438121413, "step": 31050}, {"loss": 0.5074, "grad_norm": 1.0727789402008057, "learning_rate": 0.0002, "epoch": 5.021421065394875, "step": 31060}, {"loss": 0.4938, "grad_norm": 1.1480379104614258, "learning_rate": 0.0002, "epoch": 5.023037749575621, "step": 31070}, {"loss": 0.491, "grad_norm": 1.0913071632385254, "learning_rate": 0.0002, "epoch": 5.024654433756366, "step": 31080}, {"loss": 0.4896, "grad_norm": 0.9891864657402039, "learning_rate": 0.0002, "epoch": 5.026271117937111, "step": 31090}, {"loss": 0.4965, "grad_norm": 0.9167473912239075, "learning_rate": 0.0002, "epoch": 5.027887802117856, "step": 31100}, {"loss": 0.5098, "grad_norm": 1.2259035110473633, "learning_rate": 0.0002, "epoch": 5.029504486298602, "step": 31110}, {"loss": 0.5206, "grad_norm": 1.1812787055969238, "learning_rate": 0.0002, "epoch": 5.031121170479347, "step": 31120}, {"loss": 0.4725, "grad_norm": 1.0890522003173828, "learning_rate": 0.0002, "epoch": 5.032737854660092, "step": 31130}, {"loss": 0.4768, "grad_norm": 1.0521091222763062, "learning_rate": 0.0002, "epoch": 5.034354538840837, "step": 31140}, {"loss": 0.4718, "grad_norm": 1.1274569034576416, "learning_rate": 0.0002, "epoch": 5.0359712230215825, "step": 31150}, {"loss": 0.4604, "grad_norm": 1.140974998474121, "learning_rate": 0.0002, "epoch": 5.037587907202328, "step": 31160}, {"loss": 0.5077, "grad_norm": 1.1215609312057495, "learning_rate": 0.0002, "epoch": 5.039204591383073, "step": 31170}, {"loss": 0.4746, "grad_norm": 1.0107218027114868, "learning_rate": 0.0002, "epoch": 5.040821275563818, "step": 31180}, {"loss": 0.5126, "grad_norm": 1.0198770761489868, "learning_rate": 0.0002, "epoch": 5.042437959744564, "step": 31190}, {"loss": 0.5004, "grad_norm": 1.1613430976867676, "learning_rate": 0.0002, "epoch": 5.0440546439253096, "step": 31200}, {"loss": 0.5181, "grad_norm": 0.8555458188056946, "learning_rate": 0.0002, "epoch": 5.045671328106055, "step": 31210}, {"loss": 0.4878, "grad_norm": 1.0235545635223389, "learning_rate": 0.0002, "epoch": 5.0472880122868, "step": 31220}, {"loss": 0.499, "grad_norm": 1.0228750705718994, "learning_rate": 0.0002, "epoch": 5.048904696467545, "step": 31230}, {"loss": 0.4544, "grad_norm": 0.8216419816017151, "learning_rate": 0.0002, "epoch": 5.0505213806482905, "step": 31240}, {"loss": 0.4947, "grad_norm": 0.925828218460083, "learning_rate": 0.0002, "epoch": 5.052138064829036, "step": 31250}, {"loss": 0.4835, "grad_norm": 0.9229369759559631, "learning_rate": 0.0002, "epoch": 5.053754749009781, "step": 31260}, {"loss": 0.5136, "grad_norm": 0.9531727433204651, "learning_rate": 0.0002, "epoch": 5.055371433190526, "step": 31270}, {"loss": 0.5161, "grad_norm": 0.7738548517227173, "learning_rate": 0.0002, "epoch": 5.056988117371271, "step": 31280}, {"loss": 0.5166, "grad_norm": 1.0551451444625854, "learning_rate": 0.0002, "epoch": 5.058604801552017, "step": 31290}, {"loss": 0.4953, "grad_norm": 0.9782299399375916, "learning_rate": 0.0002, "epoch": 5.060221485732762, "step": 31300}, {"loss": 0.4776, "grad_norm": 1.0220632553100586, "learning_rate": 0.0002, "epoch": 5.061838169913507, "step": 31310}, {"loss": 0.5117, "grad_norm": 0.9808892607688904, "learning_rate": 0.0002, "epoch": 5.063454854094252, "step": 31320}, {"loss": 0.501, "grad_norm": 1.0662003755569458, "learning_rate": 0.0002, "epoch": 5.065071538274998, "step": 31330}, {"loss": 0.4844, "grad_norm": 1.0036940574645996, "learning_rate": 0.0002, "epoch": 5.066688222455744, "step": 31340}, {"loss": 0.5299, "grad_norm": 1.1931052207946777, "learning_rate": 0.0002, "epoch": 5.068304906636489, "step": 31350}, {"loss": 0.4646, "grad_norm": 0.9370693564414978, "learning_rate": 0.0002, "epoch": 5.069921590817234, "step": 31360}, {"loss": 0.5274, "grad_norm": 0.9589039087295532, "learning_rate": 0.0002, "epoch": 5.071538274997979, "step": 31370}, {"loss": 0.4669, "grad_norm": 1.0052711963653564, "learning_rate": 0.0002, "epoch": 5.073154959178725, "step": 31380}, {"loss": 0.5283, "grad_norm": 0.9991368651390076, "learning_rate": 0.0002, "epoch": 5.07477164335947, "step": 31390}, {"loss": 0.4579, "grad_norm": 0.8539695739746094, "learning_rate": 0.0002, "epoch": 5.076388327540215, "step": 31400}, {"loss": 0.4609, "grad_norm": 1.048775553703308, "learning_rate": 0.0002, "epoch": 5.07800501172096, "step": 31410}, {"loss": 0.4915, "grad_norm": 0.9983724355697632, "learning_rate": 0.0002, "epoch": 5.0796216959017055, "step": 31420}, {"loss": 0.4594, "grad_norm": 1.0189813375473022, "learning_rate": 0.0002, "epoch": 5.081238380082451, "step": 31430}, {"loss": 0.5449, "grad_norm": 0.9781646728515625, "learning_rate": 0.0002, "epoch": 5.082855064263196, "step": 31440}, {"loss": 0.4698, "grad_norm": 0.9424566030502319, "learning_rate": 0.0002, "epoch": 5.084471748443941, "step": 31450}, {"loss": 0.4768, "grad_norm": 1.0036484003067017, "learning_rate": 0.0002, "epoch": 5.0860884326246865, "step": 31460}, {"loss": 0.487, "grad_norm": 1.0983147621154785, "learning_rate": 0.0002, "epoch": 5.087705116805432, "step": 31470}, {"loss": 0.5236, "grad_norm": 1.0856730937957764, "learning_rate": 0.0002, "epoch": 5.089321800986177, "step": 31480}, {"loss": 0.485, "grad_norm": 1.2191699743270874, "learning_rate": 0.0002, "epoch": 5.090938485166923, "step": 31490}, {"loss": 0.4936, "grad_norm": 0.939346194267273, "learning_rate": 0.0002, "epoch": 5.092555169347668, "step": 31500}, {"loss": 0.5107, "grad_norm": 0.9730121493339539, "learning_rate": 0.0002, "epoch": 5.0941718535284135, "step": 31510}, {"loss": 0.4973, "grad_norm": 0.923686146736145, "learning_rate": 0.0002, "epoch": 5.095788537709159, "step": 31520}, {"loss": 0.4906, "grad_norm": 1.1734349727630615, "learning_rate": 0.0002, "epoch": 5.097405221889904, "step": 31530}, {"loss": 0.5165, "grad_norm": 1.084509015083313, "learning_rate": 0.0002, "epoch": 5.099021906070649, "step": 31540}, {"loss": 0.5078, "grad_norm": 1.0144678354263306, "learning_rate": 0.0002, "epoch": 5.100638590251394, "step": 31550}, {"loss": 0.4719, "grad_norm": 0.9958019256591797, "learning_rate": 0.0002, "epoch": 5.10225527443214, "step": 31560}, {"loss": 0.4876, "grad_norm": 0.8900736570358276, "learning_rate": 0.0002, "epoch": 5.103871958612885, "step": 31570}, {"loss": 0.463, "grad_norm": 1.0921649932861328, "learning_rate": 0.0002, "epoch": 5.10548864279363, "step": 31580}, {"loss": 0.5148, "grad_norm": 1.1613792181015015, "learning_rate": 0.0002, "epoch": 5.107105326974375, "step": 31590}, {"loss": 0.5055, "grad_norm": 0.9211367964744568, "learning_rate": 0.0002, "epoch": 5.108722011155121, "step": 31600}, {"loss": 0.5364, "grad_norm": 1.3315813541412354, "learning_rate": 0.0002, "epoch": 5.110338695335866, "step": 31610}, {"loss": 0.5336, "grad_norm": 1.3765019178390503, "learning_rate": 0.0002, "epoch": 5.111955379516611, "step": 31620}, {"loss": 0.4861, "grad_norm": 1.070198893547058, "learning_rate": 0.0002, "epoch": 5.113572063697356, "step": 31630}, {"loss": 0.5046, "grad_norm": 0.947631299495697, "learning_rate": 0.0002, "epoch": 5.115188747878102, "step": 31640}, {"loss": 0.5297, "grad_norm": 1.0197371244430542, "learning_rate": 0.0002, "epoch": 5.116805432058848, "step": 31650}, {"loss": 0.5014, "grad_norm": 0.8647911548614502, "learning_rate": 0.0002, "epoch": 5.118422116239593, "step": 31660}, {"loss": 0.4705, "grad_norm": 0.8944075107574463, "learning_rate": 0.0002, "epoch": 5.120038800420338, "step": 31670}, {"loss": 0.5175, "grad_norm": 1.124497652053833, "learning_rate": 0.0002, "epoch": 5.121655484601083, "step": 31680}, {"loss": 0.5109, "grad_norm": 0.893131673336029, "learning_rate": 0.0002, "epoch": 5.123272168781829, "step": 31690}, {"loss": 0.4937, "grad_norm": 1.0122284889221191, "learning_rate": 0.0002, "epoch": 5.124888852962574, "step": 31700}, {"loss": 0.5522, "grad_norm": 0.9493719935417175, "learning_rate": 0.0002, "epoch": 5.126505537143319, "step": 31710}, {"loss": 0.5031, "grad_norm": 0.9700539112091064, "learning_rate": 0.0002, "epoch": 5.128122221324064, "step": 31720}, {"loss": 0.5126, "grad_norm": 1.111677646636963, "learning_rate": 0.0002, "epoch": 5.1297389055048095, "step": 31730}, {"loss": 0.5272, "grad_norm": 0.8204274773597717, "learning_rate": 0.0002, "epoch": 5.131355589685555, "step": 31740}, {"loss": 0.5029, "grad_norm": 1.1029267311096191, "learning_rate": 0.0002, "epoch": 5.1329722738663, "step": 31750}, {"loss": 0.505, "grad_norm": 1.065575122833252, "learning_rate": 0.0002, "epoch": 5.134588958047045, "step": 31760}, {"loss": 0.502, "grad_norm": 0.8208706974983215, "learning_rate": 0.0002, "epoch": 5.13620564222779, "step": 31770}, {"loss": 0.5352, "grad_norm": 1.0520979166030884, "learning_rate": 0.0002, "epoch": 5.137822326408536, "step": 31780}, {"loss": 0.4911, "grad_norm": 0.8585538268089294, "learning_rate": 0.0002, "epoch": 5.139439010589282, "step": 31790}, {"loss": 0.5159, "grad_norm": 1.1491447687149048, "learning_rate": 0.0002, "epoch": 5.141055694770027, "step": 31800}, {"loss": 0.5157, "grad_norm": 0.9441081285476685, "learning_rate": 0.0002, "epoch": 5.142672378950772, "step": 31810}, {"loss": 0.5383, "grad_norm": 1.4146889448165894, "learning_rate": 0.0002, "epoch": 5.1442890631315175, "step": 31820}, {"loss": 0.5159, "grad_norm": 1.0326547622680664, "learning_rate": 0.0002, "epoch": 5.145905747312263, "step": 31830}, {"loss": 0.5348, "grad_norm": 0.9879202842712402, "learning_rate": 0.0002, "epoch": 5.147522431493008, "step": 31840}, {"loss": 0.5083, "grad_norm": 1.0374281406402588, "learning_rate": 0.0002, "epoch": 5.149139115673753, "step": 31850}, {"loss": 0.4827, "grad_norm": 1.181229591369629, "learning_rate": 0.0002, "epoch": 5.150755799854498, "step": 31860}, {"loss": 0.5313, "grad_norm": 1.2078537940979004, "learning_rate": 0.0002, "epoch": 5.152372484035244, "step": 31870}, {"loss": 0.5329, "grad_norm": 0.9599190354347229, "learning_rate": 0.0002, "epoch": 5.153989168215989, "step": 31880}, {"loss": 0.4953, "grad_norm": 1.0378568172454834, "learning_rate": 0.0002, "epoch": 5.155605852396734, "step": 31890}, {"loss": 0.5069, "grad_norm": 0.8746536374092102, "learning_rate": 0.0002, "epoch": 5.157222536577479, "step": 31900}, {"loss": 0.5272, "grad_norm": 1.0232136249542236, "learning_rate": 0.0002, "epoch": 5.1588392207582245, "step": 31910}, {"loss": 0.4844, "grad_norm": 0.9827565550804138, "learning_rate": 0.0002, "epoch": 5.16045590493897, "step": 31920}, {"loss": 0.5029, "grad_norm": 1.342657208442688, "learning_rate": 0.0002, "epoch": 5.162072589119716, "step": 31930}, {"loss": 0.513, "grad_norm": 1.18390691280365, "learning_rate": 0.0002, "epoch": 5.163689273300461, "step": 31940}, {"loss": 0.5267, "grad_norm": 0.996350109577179, "learning_rate": 0.0002, "epoch": 5.165305957481206, "step": 31950}, {"loss": 0.5063, "grad_norm": 0.9710391163825989, "learning_rate": 0.0002, "epoch": 5.166922641661952, "step": 31960}, {"loss": 0.5115, "grad_norm": 1.0264002084732056, "learning_rate": 0.0002, "epoch": 5.168539325842697, "step": 31970}, {"loss": 0.4972, "grad_norm": 1.0028311014175415, "learning_rate": 0.0002, "epoch": 5.170156010023442, "step": 31980}, {"loss": 0.5103, "grad_norm": 1.1078234910964966, "learning_rate": 0.0002, "epoch": 5.171772694204187, "step": 31990}, {"loss": 0.495, "grad_norm": 0.9659610390663147, "learning_rate": 0.0002, "epoch": 5.1733893783849325, "step": 32000}, {"loss": 0.5114, "grad_norm": 0.841986894607544, "learning_rate": 0.0002, "epoch": 5.175006062565678, "step": 32010}, {"loss": 0.48, "grad_norm": 1.095332384109497, "learning_rate": 0.0002, "epoch": 5.176622746746423, "step": 32020}, {"loss": 0.4741, "grad_norm": 1.1242377758026123, "learning_rate": 0.0002, "epoch": 5.178239430927168, "step": 32030}, {"loss": 0.5573, "grad_norm": 0.9872292280197144, "learning_rate": 0.0002, "epoch": 5.179856115107913, "step": 32040}, {"loss": 0.48, "grad_norm": 0.936161994934082, "learning_rate": 0.0002, "epoch": 5.181472799288659, "step": 32050}, {"loss": 0.5093, "grad_norm": 1.166100025177002, "learning_rate": 0.0002, "epoch": 5.183089483469404, "step": 32060}, {"loss": 0.5438, "grad_norm": 1.0764425992965698, "learning_rate": 0.0002, "epoch": 5.184706167650149, "step": 32070}, {"loss": 0.4843, "grad_norm": 1.0480051040649414, "learning_rate": 0.0002, "epoch": 5.186322851830895, "step": 32080}, {"loss": 0.5386, "grad_norm": 1.0874916315078735, "learning_rate": 0.0002, "epoch": 5.1879395360116405, "step": 32090}, {"loss": 0.4975, "grad_norm": 1.0817396640777588, "learning_rate": 0.0002, "epoch": 5.189556220192386, "step": 32100}, {"loss": 0.5177, "grad_norm": 1.054111361503601, "learning_rate": 0.0002, "epoch": 5.191172904373131, "step": 32110}, {"loss": 0.5229, "grad_norm": 0.9655823707580566, "learning_rate": 0.0002, "epoch": 5.192789588553876, "step": 32120}, {"loss": 0.5105, "grad_norm": 1.1384109258651733, "learning_rate": 0.0002, "epoch": 5.194406272734621, "step": 32130}, {"loss": 0.5073, "grad_norm": 1.0149348974227905, "learning_rate": 0.0002, "epoch": 5.196022956915367, "step": 32140}, {"loss": 0.5293, "grad_norm": 1.1084046363830566, "learning_rate": 0.0002, "epoch": 5.197639641096112, "step": 32150}, {"loss": 0.4936, "grad_norm": 1.1209309101104736, "learning_rate": 0.0002, "epoch": 5.199256325276857, "step": 32160}, {"loss": 0.5101, "grad_norm": 1.133089542388916, "learning_rate": 0.0002, "epoch": 5.200873009457602, "step": 32170}, {"loss": 0.5242, "grad_norm": 1.0893020629882812, "learning_rate": 0.0002, "epoch": 5.202489693638348, "step": 32180}, {"loss": 0.4872, "grad_norm": 0.90018630027771, "learning_rate": 0.0002, "epoch": 5.204106377819093, "step": 32190}, {"loss": 0.4999, "grad_norm": 0.977622926235199, "learning_rate": 0.0002, "epoch": 5.205723061999838, "step": 32200}, {"loss": 0.5028, "grad_norm": 1.2940177917480469, "learning_rate": 0.0002, "epoch": 5.207339746180583, "step": 32210}, {"loss": 0.5396, "grad_norm": 1.2131710052490234, "learning_rate": 0.0002, "epoch": 5.2089564303613285, "step": 32220}, {"loss": 0.5189, "grad_norm": 1.0234841108322144, "learning_rate": 0.0002, "epoch": 5.210573114542075, "step": 32230}, {"loss": 0.5424, "grad_norm": 1.157975435256958, "learning_rate": 0.0002, "epoch": 5.21218979872282, "step": 32240}, {"loss": 0.5396, "grad_norm": 1.0381282567977905, "learning_rate": 0.0002, "epoch": 5.213806482903565, "step": 32250}, {"loss": 0.5192, "grad_norm": 1.0125395059585571, "learning_rate": 0.0002, "epoch": 5.21542316708431, "step": 32260}, {"loss": 0.5216, "grad_norm": 1.272691011428833, "learning_rate": 0.0002, "epoch": 5.2170398512650555, "step": 32270}, {"loss": 0.52, "grad_norm": 1.0061250925064087, "learning_rate": 0.0002, "epoch": 5.218656535445801, "step": 32280}, {"loss": 0.4739, "grad_norm": 0.9752234816551208, "learning_rate": 0.0002, "epoch": 5.220273219626546, "step": 32290}, {"loss": 0.5471, "grad_norm": 1.1193140745162964, "learning_rate": 0.0002, "epoch": 5.221889903807291, "step": 32300}, {"loss": 0.4976, "grad_norm": 1.0126434564590454, "learning_rate": 0.0002, "epoch": 5.2235065879880365, "step": 32310}, {"loss": 0.5257, "grad_norm": 1.4338394403457642, "learning_rate": 0.0002, "epoch": 5.225123272168782, "step": 32320}, {"loss": 0.5235, "grad_norm": 1.004101276397705, "learning_rate": 0.0002, "epoch": 5.226739956349527, "step": 32330}, {"loss": 0.5091, "grad_norm": 0.8744166493415833, "learning_rate": 0.0002, "epoch": 5.228356640530272, "step": 32340}, {"loss": 0.5388, "grad_norm": 1.0165376663208008, "learning_rate": 0.0002, "epoch": 5.229973324711017, "step": 32350}, {"loss": 0.5469, "grad_norm": 0.8635954260826111, "learning_rate": 0.0002, "epoch": 5.231590008891763, "step": 32360}, {"loss": 0.5609, "grad_norm": 1.1392399072647095, "learning_rate": 0.0002, "epoch": 5.233206693072509, "step": 32370}, {"loss": 0.5173, "grad_norm": 1.0202113389968872, "learning_rate": 0.0002, "epoch": 5.234823377253254, "step": 32380}, {"loss": 0.4983, "grad_norm": 1.0417983531951904, "learning_rate": 0.0002, "epoch": 5.236440061433999, "step": 32390}, {"loss": 0.507, "grad_norm": 0.8729333877563477, "learning_rate": 0.0002, "epoch": 5.238056745614744, "step": 32400}, {"loss": 0.5426, "grad_norm": 1.1626229286193848, "learning_rate": 0.0002, "epoch": 5.23967342979549, "step": 32410}, {"loss": 0.5355, "grad_norm": 0.9086161851882935, "learning_rate": 0.0002, "epoch": 5.241290113976235, "step": 32420}, {"loss": 0.4927, "grad_norm": 1.3999892473220825, "learning_rate": 0.0002, "epoch": 5.24290679815698, "step": 32430}, {"loss": 0.4795, "grad_norm": 1.0356311798095703, "learning_rate": 0.0002, "epoch": 5.244523482337725, "step": 32440}, {"loss": 0.5035, "grad_norm": 0.9655531644821167, "learning_rate": 0.0002, "epoch": 5.246140166518471, "step": 32450}, {"loss": 0.5166, "grad_norm": 1.0411828756332397, "learning_rate": 0.0002, "epoch": 5.247756850699216, "step": 32460}, {"loss": 0.5141, "grad_norm": 1.1199816465377808, "learning_rate": 0.0002, "epoch": 5.249373534879961, "step": 32470}, {"loss": 0.4864, "grad_norm": 1.260321855545044, "learning_rate": 0.0002, "epoch": 5.250990219060706, "step": 32480}, {"loss": 0.4893, "grad_norm": 1.2950857877731323, "learning_rate": 0.0002, "epoch": 5.2526069032414515, "step": 32490}, {"loss": 0.4952, "grad_norm": 0.8982820510864258, "learning_rate": 0.0002, "epoch": 5.254223587422197, "step": 32500}, {"loss": 0.5138, "grad_norm": 0.8512987494468689, "learning_rate": 0.0002, "epoch": 5.255840271602942, "step": 32510}, {"loss": 0.5341, "grad_norm": 1.067443609237671, "learning_rate": 0.0002, "epoch": 5.257456955783688, "step": 32520}, {"loss": 0.4928, "grad_norm": 1.0957417488098145, "learning_rate": 0.0002, "epoch": 5.259073639964433, "step": 32530}, {"loss": 0.5169, "grad_norm": 1.4161807298660278, "learning_rate": 0.0002, "epoch": 5.260690324145179, "step": 32540}, {"loss": 0.5599, "grad_norm": 1.2264093160629272, "learning_rate": 0.0002, "epoch": 5.262307008325924, "step": 32550}, {"loss": 0.5221, "grad_norm": 1.0015931129455566, "learning_rate": 0.0002, "epoch": 5.263923692506669, "step": 32560}, {"loss": 0.5253, "grad_norm": 1.0743094682693481, "learning_rate": 0.0002, "epoch": 5.265540376687414, "step": 32570}, {"loss": 0.5289, "grad_norm": 1.1386840343475342, "learning_rate": 0.0002, "epoch": 5.2671570608681595, "step": 32580}, {"loss": 0.5315, "grad_norm": 1.0093860626220703, "learning_rate": 0.0002, "epoch": 5.268773745048905, "step": 32590}, {"loss": 0.5175, "grad_norm": 0.9593744874000549, "learning_rate": 0.0002, "epoch": 5.27039042922965, "step": 32600}, {"loss": 0.528, "grad_norm": 1.146021842956543, "learning_rate": 0.0002, "epoch": 5.272007113410395, "step": 32610}, {"loss": 0.4983, "grad_norm": 0.9579031467437744, "learning_rate": 0.0002, "epoch": 5.27362379759114, "step": 32620}, {"loss": 0.5376, "grad_norm": 1.0548793077468872, "learning_rate": 0.0002, "epoch": 5.275240481771886, "step": 32630}, {"loss": 0.5267, "grad_norm": 1.0380561351776123, "learning_rate": 0.0002, "epoch": 5.276857165952631, "step": 32640}, {"loss": 0.5182, "grad_norm": 1.2119969129562378, "learning_rate": 0.0002, "epoch": 5.278473850133376, "step": 32650}, {"loss": 0.5298, "grad_norm": 1.0507797002792358, "learning_rate": 0.0002, "epoch": 5.280090534314121, "step": 32660}, {"loss": 0.5253, "grad_norm": 1.0185176134109497, "learning_rate": 0.0002, "epoch": 5.2817072184948675, "step": 32670}, {"loss": 0.4904, "grad_norm": 1.2358098030090332, "learning_rate": 0.0002, "epoch": 5.283323902675613, "step": 32680}, {"loss": 0.5169, "grad_norm": 0.7937114238739014, "learning_rate": 0.0002, "epoch": 5.284940586856358, "step": 32690}, {"loss": 0.495, "grad_norm": 0.9825124740600586, "learning_rate": 0.0002, "epoch": 5.286557271037103, "step": 32700}, {"loss": 0.5149, "grad_norm": 1.2059301137924194, "learning_rate": 0.0002, "epoch": 5.288173955217848, "step": 32710}, {"loss": 0.5272, "grad_norm": 1.0828571319580078, "learning_rate": 0.0002, "epoch": 5.289790639398594, "step": 32720}, {"loss": 0.5383, "grad_norm": 1.0129735469818115, "learning_rate": 0.0002, "epoch": 5.291407323579339, "step": 32730}, {"loss": 0.5216, "grad_norm": 1.0591634511947632, "learning_rate": 0.0002, "epoch": 5.293024007760084, "step": 32740}, {"loss": 0.522, "grad_norm": 0.9256815910339355, "learning_rate": 0.0002, "epoch": 5.294640691940829, "step": 32750}, {"loss": 0.5396, "grad_norm": 1.0928633213043213, "learning_rate": 0.0002, "epoch": 5.2962573761215745, "step": 32760}, {"loss": 0.5093, "grad_norm": 0.9415594935417175, "learning_rate": 0.0002, "epoch": 5.29787406030232, "step": 32770}, {"loss": 0.5252, "grad_norm": 1.141316294670105, "learning_rate": 0.0002, "epoch": 5.299490744483065, "step": 32780}, {"loss": 0.4837, "grad_norm": 1.0646510124206543, "learning_rate": 0.0002, "epoch": 5.30110742866381, "step": 32790}, {"loss": 0.5547, "grad_norm": 1.189661979675293, "learning_rate": 0.0002, "epoch": 5.3027241128445555, "step": 32800}, {"loss": 0.5664, "grad_norm": 0.9568731188774109, "learning_rate": 0.0002, "epoch": 5.304340797025301, "step": 32810}, {"loss": 0.5344, "grad_norm": 1.1556824445724487, "learning_rate": 0.0002, "epoch": 5.305957481206047, "step": 32820}, {"loss": 0.4894, "grad_norm": 0.9353463649749756, "learning_rate": 0.0002, "epoch": 5.307574165386792, "step": 32830}, {"loss": 0.5052, "grad_norm": 1.1208295822143555, "learning_rate": 0.0002, "epoch": 5.309190849567537, "step": 32840}, {"loss": 0.5126, "grad_norm": 1.0894153118133545, "learning_rate": 0.0002, "epoch": 5.3108075337482825, "step": 32850}, {"loss": 0.5046, "grad_norm": 1.090329647064209, "learning_rate": 0.0002, "epoch": 5.312424217929028, "step": 32860}, {"loss": 0.5237, "grad_norm": 1.0781712532043457, "learning_rate": 0.0002, "epoch": 5.314040902109773, "step": 32870}, {"loss": 0.57, "grad_norm": 1.1785295009613037, "learning_rate": 0.0002, "epoch": 5.315657586290518, "step": 32880}, {"loss": 0.4953, "grad_norm": 1.0406851768493652, "learning_rate": 0.0002, "epoch": 5.317274270471263, "step": 32890}, {"loss": 0.514, "grad_norm": 1.0982953310012817, "learning_rate": 0.0002, "epoch": 5.318890954652009, "step": 32900}, {"loss": 0.4944, "grad_norm": 1.2969383001327515, "learning_rate": 0.0002, "epoch": 5.320507638832754, "step": 32910}, {"loss": 0.4786, "grad_norm": 0.9687288999557495, "learning_rate": 0.0002, "epoch": 5.322124323013499, "step": 32920}, {"loss": 0.5286, "grad_norm": 1.136760950088501, "learning_rate": 0.0002, "epoch": 5.323741007194244, "step": 32930}, {"loss": 0.5321, "grad_norm": 1.3045495748519897, "learning_rate": 0.0002, "epoch": 5.32535769137499, "step": 32940}, {"loss": 0.5413, "grad_norm": 1.221675992012024, "learning_rate": 0.0002, "epoch": 5.326974375555735, "step": 32950}, {"loss": 0.4999, "grad_norm": 1.1380633115768433, "learning_rate": 0.0002, "epoch": 5.32859105973648, "step": 32960}, {"loss": 0.5037, "grad_norm": 1.1065956354141235, "learning_rate": 0.0002, "epoch": 5.330207743917226, "step": 32970}, {"loss": 0.4913, "grad_norm": 1.0187175273895264, "learning_rate": 0.0002, "epoch": 5.331824428097971, "step": 32980}, {"loss": 0.5234, "grad_norm": 0.9077118039131165, "learning_rate": 0.0002, "epoch": 5.333441112278717, "step": 32990}, {"loss": 0.5071, "grad_norm": 1.0092815160751343, "learning_rate": 0.0002, "epoch": 5.335057796459462, "step": 33000}, {"loss": 0.498, "grad_norm": 1.0168777704238892, "learning_rate": 0.0002, "epoch": 5.336674480640207, "step": 33010}, {"loss": 0.4952, "grad_norm": 0.996161937713623, "learning_rate": 0.0002, "epoch": 5.338291164820952, "step": 33020}, {"loss": 0.5024, "grad_norm": 0.794463038444519, "learning_rate": 0.0002, "epoch": 5.339907849001698, "step": 33030}, {"loss": 0.5112, "grad_norm": 0.9750674962997437, "learning_rate": 0.0002, "epoch": 5.341524533182443, "step": 33040}, {"loss": 0.528, "grad_norm": 1.2770029306411743, "learning_rate": 0.0002, "epoch": 5.343141217363188, "step": 33050}, {"loss": 0.52, "grad_norm": 1.1500186920166016, "learning_rate": 0.0002, "epoch": 5.344757901543933, "step": 33060}, {"loss": 0.4906, "grad_norm": 1.0726377964019775, "learning_rate": 0.0002, "epoch": 5.3463745857246785, "step": 33070}, {"loss": 0.5212, "grad_norm": 0.9314153790473938, "learning_rate": 0.0002, "epoch": 5.347991269905424, "step": 33080}, {"loss": 0.5434, "grad_norm": 1.344988465309143, "learning_rate": 0.0002, "epoch": 5.349607954086169, "step": 33090}, {"loss": 0.4874, "grad_norm": 0.863196611404419, "learning_rate": 0.0002, "epoch": 5.351224638266914, "step": 33100}, {"loss": 0.534, "grad_norm": 1.128100037574768, "learning_rate": 0.0002, "epoch": 5.352841322447659, "step": 33110}, {"loss": 0.5293, "grad_norm": 1.1673583984375, "learning_rate": 0.0002, "epoch": 5.3544580066284055, "step": 33120}, {"loss": 0.4787, "grad_norm": 0.9416789412498474, "learning_rate": 0.0002, "epoch": 5.356074690809151, "step": 33130}, {"loss": 0.5155, "grad_norm": 1.1855236291885376, "learning_rate": 0.0002, "epoch": 5.357691374989896, "step": 33140}, {"loss": 0.515, "grad_norm": 1.0415170192718506, "learning_rate": 0.0002, "epoch": 5.359308059170641, "step": 33150}, {"loss": 0.545, "grad_norm": 0.9953004121780396, "learning_rate": 0.0002, "epoch": 5.3609247433513865, "step": 33160}, {"loss": 0.5305, "grad_norm": 0.96138596534729, "learning_rate": 0.0002, "epoch": 5.362541427532132, "step": 33170}, {"loss": 0.5064, "grad_norm": 1.341979742050171, "learning_rate": 0.0002, "epoch": 5.364158111712877, "step": 33180}, {"loss": 0.4986, "grad_norm": 1.0136911869049072, "learning_rate": 0.0002, "epoch": 5.365774795893622, "step": 33190}, {"loss": 0.5459, "grad_norm": 0.8685575127601624, "learning_rate": 0.0002, "epoch": 5.367391480074367, "step": 33200}, {"loss": 0.5146, "grad_norm": 0.8833574652671814, "learning_rate": 0.0002, "epoch": 5.369008164255113, "step": 33210}, {"loss": 0.4982, "grad_norm": 0.9123612642288208, "learning_rate": 0.0002, "epoch": 5.370624848435858, "step": 33220}, {"loss": 0.5047, "grad_norm": 1.2720599174499512, "learning_rate": 0.0002, "epoch": 5.372241532616603, "step": 33230}, {"loss": 0.5175, "grad_norm": 1.0596648454666138, "learning_rate": 0.0002, "epoch": 5.373858216797348, "step": 33240}, {"loss": 0.5284, "grad_norm": 1.119701623916626, "learning_rate": 0.0002, "epoch": 5.3754749009780936, "step": 33250}, {"loss": 0.5217, "grad_norm": 1.3000061511993408, "learning_rate": 0.0002, "epoch": 5.377091585158839, "step": 33260}, {"loss": 0.5125, "grad_norm": 1.083891749382019, "learning_rate": 0.0002, "epoch": 5.378708269339585, "step": 33270}, {"loss": 0.5065, "grad_norm": 0.9402718544006348, "learning_rate": 0.0002, "epoch": 5.38032495352033, "step": 33280}, {"loss": 0.5559, "grad_norm": 1.3376892805099487, "learning_rate": 0.0002, "epoch": 5.381941637701075, "step": 33290}, {"loss": 0.5193, "grad_norm": 1.1600074768066406, "learning_rate": 0.0002, "epoch": 5.383558321881821, "step": 33300}, {"loss": 0.4907, "grad_norm": 1.1449427604675293, "learning_rate": 0.0002, "epoch": 5.385175006062566, "step": 33310}, {"loss": 0.5449, "grad_norm": 1.3118891716003418, "learning_rate": 0.0002, "epoch": 5.386791690243311, "step": 33320}, {"loss": 0.547, "grad_norm": 0.743449866771698, "learning_rate": 0.0002, "epoch": 5.388408374424056, "step": 33330}, {"loss": 0.5555, "grad_norm": 0.9358304142951965, "learning_rate": 0.0002, "epoch": 5.3900250586048015, "step": 33340}, {"loss": 0.5558, "grad_norm": 1.0447142124176025, "learning_rate": 0.0002, "epoch": 5.391641742785547, "step": 33350}, {"loss": 0.5106, "grad_norm": 1.1088626384735107, "learning_rate": 0.0002, "epoch": 5.393258426966292, "step": 33360}, {"loss": 0.4929, "grad_norm": 1.1267958879470825, "learning_rate": 0.0002, "epoch": 5.394875111147037, "step": 33370}, {"loss": 0.5165, "grad_norm": 0.9709370136260986, "learning_rate": 0.0002, "epoch": 5.3964917953277824, "step": 33380}, {"loss": 0.5206, "grad_norm": 1.0939103364944458, "learning_rate": 0.0002, "epoch": 5.398108479508528, "step": 33390}, {"loss": 0.5177, "grad_norm": 0.9559304714202881, "learning_rate": 0.0002, "epoch": 5.399725163689273, "step": 33400}, {"loss": 0.5064, "grad_norm": 1.199580430984497, "learning_rate": 0.0002, "epoch": 5.401341847870018, "step": 33410}, {"loss": 0.52, "grad_norm": 0.9097000360488892, "learning_rate": 0.0002, "epoch": 5.402958532050764, "step": 33420}, {"loss": 0.514, "grad_norm": 1.1940981149673462, "learning_rate": 0.0002, "epoch": 5.4045752162315095, "step": 33430}, {"loss": 0.5069, "grad_norm": 1.0530916452407837, "learning_rate": 0.0002, "epoch": 5.406191900412255, "step": 33440}, {"loss": 0.5482, "grad_norm": 1.0482549667358398, "learning_rate": 0.0002, "epoch": 5.407808584593, "step": 33450}, {"loss": 0.501, "grad_norm": 1.2524714469909668, "learning_rate": 0.0002, "epoch": 5.409425268773745, "step": 33460}, {"loss": 0.5597, "grad_norm": 1.1091666221618652, "learning_rate": 0.0002, "epoch": 5.41104195295449, "step": 33470}, {"loss": 0.546, "grad_norm": 0.9981587529182434, "learning_rate": 0.0002, "epoch": 5.412658637135236, "step": 33480}, {"loss": 0.4977, "grad_norm": 1.016681432723999, "learning_rate": 0.0002, "epoch": 5.414275321315981, "step": 33490}, {"loss": 0.5388, "grad_norm": 1.1456854343414307, "learning_rate": 0.0002, "epoch": 5.415892005496726, "step": 33500}, {"loss": 0.5292, "grad_norm": 1.1454259157180786, "learning_rate": 0.0002, "epoch": 5.417508689677471, "step": 33510}, {"loss": 0.5061, "grad_norm": 0.9858416318893433, "learning_rate": 0.0002, "epoch": 5.419125373858217, "step": 33520}, {"loss": 0.5139, "grad_norm": 0.9764766693115234, "learning_rate": 0.0002, "epoch": 5.420742058038962, "step": 33530}, {"loss": 0.5518, "grad_norm": 1.199920892715454, "learning_rate": 0.0002, "epoch": 5.422358742219707, "step": 33540}, {"loss": 0.5182, "grad_norm": 1.3107370138168335, "learning_rate": 0.0002, "epoch": 5.423975426400452, "step": 33550}, {"loss": 0.5149, "grad_norm": 0.9637970328330994, "learning_rate": 0.0002, "epoch": 5.4255921105811975, "step": 33560}, {"loss": 0.526, "grad_norm": 1.023359775543213, "learning_rate": 0.0002, "epoch": 5.427208794761944, "step": 33570}, {"loss": 0.5206, "grad_norm": 1.060417652130127, "learning_rate": 0.0002, "epoch": 5.428825478942689, "step": 33580}, {"loss": 0.5052, "grad_norm": 0.9971120953559875, "learning_rate": 0.0002, "epoch": 5.430442163123434, "step": 33590}, {"loss": 0.5044, "grad_norm": 0.9213743209838867, "learning_rate": 0.0002, "epoch": 5.432058847304179, "step": 33600}, {"loss": 0.5714, "grad_norm": 1.1512309312820435, "learning_rate": 0.0002, "epoch": 5.4336755314849245, "step": 33610}, {"loss": 0.5317, "grad_norm": 1.2198847532272339, "learning_rate": 0.0002, "epoch": 5.43529221566567, "step": 33620}, {"loss": 0.5237, "grad_norm": 1.0329595804214478, "learning_rate": 0.0002, "epoch": 5.436908899846415, "step": 33630}, {"loss": 0.5364, "grad_norm": 1.1075750589370728, "learning_rate": 0.0002, "epoch": 5.43852558402716, "step": 33640}, {"loss": 0.5295, "grad_norm": 1.006342887878418, "learning_rate": 0.0002, "epoch": 5.4401422682079055, "step": 33650}, {"loss": 0.5394, "grad_norm": 0.9179885983467102, "learning_rate": 0.0002, "epoch": 5.441758952388651, "step": 33660}, {"loss": 0.5124, "grad_norm": 1.2799493074417114, "learning_rate": 0.0002, "epoch": 5.443375636569396, "step": 33670}, {"loss": 0.5426, "grad_norm": 1.1153863668441772, "learning_rate": 0.0002, "epoch": 5.444992320750141, "step": 33680}, {"loss": 0.5087, "grad_norm": 1.0681028366088867, "learning_rate": 0.0002, "epoch": 5.446609004930886, "step": 33690}, {"loss": 0.5272, "grad_norm": 0.9788817167282104, "learning_rate": 0.0002, "epoch": 5.448225689111632, "step": 33700}, {"loss": 0.5308, "grad_norm": 0.8481608629226685, "learning_rate": 0.0002, "epoch": 5.449842373292377, "step": 33710}, {"loss": 0.5225, "grad_norm": 1.113756537437439, "learning_rate": 0.0002, "epoch": 5.451459057473123, "step": 33720}, {"loss": 0.5213, "grad_norm": 0.8425475358963013, "learning_rate": 0.0002, "epoch": 5.453075741653868, "step": 33730}, {"loss": 0.571, "grad_norm": 1.0852208137512207, "learning_rate": 0.0002, "epoch": 5.4546924258346134, "step": 33740}, {"loss": 0.5535, "grad_norm": 1.1664748191833496, "learning_rate": 0.0002, "epoch": 5.456309110015359, "step": 33750}, {"loss": 0.5419, "grad_norm": 1.217241644859314, "learning_rate": 0.0002, "epoch": 5.457925794196104, "step": 33760}, {"loss": 0.5351, "grad_norm": 1.1572928428649902, "learning_rate": 0.0002, "epoch": 5.459542478376849, "step": 33770}, {"loss": 0.5161, "grad_norm": 1.0437318086624146, "learning_rate": 0.0002, "epoch": 5.461159162557594, "step": 33780}, {"loss": 0.5266, "grad_norm": 0.9807571768760681, "learning_rate": 0.0002, "epoch": 5.46277584673834, "step": 33790}, {"loss": 0.5384, "grad_norm": 1.1436342000961304, "learning_rate": 0.0002, "epoch": 5.464392530919085, "step": 33800}, {"loss": 0.5338, "grad_norm": 1.1004794836044312, "learning_rate": 0.0002, "epoch": 5.46600921509983, "step": 33810}, {"loss": 0.4868, "grad_norm": 1.2130268812179565, "learning_rate": 0.0002, "epoch": 5.467625899280575, "step": 33820}, {"loss": 0.516, "grad_norm": 1.3154419660568237, "learning_rate": 0.0002, "epoch": 5.4692425834613205, "step": 33830}, {"loss": 0.4934, "grad_norm": 0.7934383749961853, "learning_rate": 0.0002, "epoch": 5.470859267642066, "step": 33840}, {"loss": 0.5133, "grad_norm": 0.7838410139083862, "learning_rate": 0.0002, "epoch": 5.472475951822812, "step": 33850}, {"loss": 0.4926, "grad_norm": 1.0415139198303223, "learning_rate": 0.0002, "epoch": 5.474092636003557, "step": 33860}, {"loss": 0.5323, "grad_norm": 0.9213164448738098, "learning_rate": 0.0002, "epoch": 5.475709320184302, "step": 33870}, {"loss": 0.5125, "grad_norm": 1.0364776849746704, "learning_rate": 0.0002, "epoch": 5.477326004365048, "step": 33880}, {"loss": 0.5212, "grad_norm": 0.9994072318077087, "learning_rate": 0.0002, "epoch": 5.478942688545793, "step": 33890}, {"loss": 0.5396, "grad_norm": 1.196730136871338, "learning_rate": 0.0002, "epoch": 5.480559372726538, "step": 33900}, {"loss": 0.538, "grad_norm": 0.9955780506134033, "learning_rate": 0.0002, "epoch": 5.482176056907283, "step": 33910}, {"loss": 0.5307, "grad_norm": 1.168188214302063, "learning_rate": 0.0002, "epoch": 5.4837927410880285, "step": 33920}, {"loss": 0.5548, "grad_norm": 1.1816450357437134, "learning_rate": 0.0002, "epoch": 5.485409425268774, "step": 33930}, {"loss": 0.5535, "grad_norm": 1.079715609550476, "learning_rate": 0.0002, "epoch": 5.487026109449519, "step": 33940}, {"loss": 0.5262, "grad_norm": 1.153850793838501, "learning_rate": 0.0002, "epoch": 5.488642793630264, "step": 33950}, {"loss": 0.5248, "grad_norm": 1.0207297801971436, "learning_rate": 0.0002, "epoch": 5.490259477811009, "step": 33960}, {"loss": 0.5142, "grad_norm": 1.1290855407714844, "learning_rate": 0.0002, "epoch": 5.491876161991755, "step": 33970}, {"loss": 0.5168, "grad_norm": 1.068058967590332, "learning_rate": 0.0002, "epoch": 5.4934928461725, "step": 33980}, {"loss": 0.5317, "grad_norm": 0.9789979457855225, "learning_rate": 0.0002, "epoch": 5.495109530353245, "step": 33990}, {"loss": 0.5113, "grad_norm": 0.9696692824363708, "learning_rate": 0.0002, "epoch": 5.496726214533991, "step": 34000}, {"loss": 0.5413, "grad_norm": 1.0539981126785278, "learning_rate": 0.0002, "epoch": 5.4983428987147365, "step": 34010}, {"loss": 0.5783, "grad_norm": 1.0249929428100586, "learning_rate": 0.0002, "epoch": 5.499959582895482, "step": 34020}, {"loss": 0.4888, "grad_norm": 0.9577504992485046, "learning_rate": 0.0002, "epoch": 5.501576267076227, "step": 34030}, {"loss": 0.5291, "grad_norm": 1.0963513851165771, "learning_rate": 0.0002, "epoch": 5.503192951256972, "step": 34040}, {"loss": 0.5315, "grad_norm": 0.8339345455169678, "learning_rate": 0.0002, "epoch": 5.504809635437717, "step": 34050}, {"loss": 0.5191, "grad_norm": 1.0138782262802124, "learning_rate": 0.0002, "epoch": 5.506426319618463, "step": 34060}, {"loss": 0.5463, "grad_norm": 1.0180109739303589, "learning_rate": 0.0002, "epoch": 5.508043003799208, "step": 34070}, {"loss": 0.5083, "grad_norm": 1.2790818214416504, "learning_rate": 0.0002, "epoch": 5.509659687979953, "step": 34080}, {"loss": 0.5195, "grad_norm": 1.428247332572937, "learning_rate": 0.0002, "epoch": 5.511276372160698, "step": 34090}, {"loss": 0.5291, "grad_norm": 1.0926059484481812, "learning_rate": 0.0002, "epoch": 5.5128930563414436, "step": 34100}, {"loss": 0.5665, "grad_norm": 1.2353343963623047, "learning_rate": 0.0002, "epoch": 5.514509740522189, "step": 34110}, {"loss": 0.5331, "grad_norm": 0.935587465763092, "learning_rate": 0.0002, "epoch": 5.516126424702934, "step": 34120}, {"loss": 0.5512, "grad_norm": 0.9767586588859558, "learning_rate": 0.0002, "epoch": 5.517743108883679, "step": 34130}, {"loss": 0.5315, "grad_norm": 1.1660610437393188, "learning_rate": 0.0002, "epoch": 5.5193597930644245, "step": 34140}, {"loss": 0.52, "grad_norm": 0.9828870892524719, "learning_rate": 0.0002, "epoch": 5.520976477245171, "step": 34150}, {"loss": 0.5198, "grad_norm": 1.0097278356552124, "learning_rate": 0.0002, "epoch": 5.522593161425916, "step": 34160}, {"loss": 0.5293, "grad_norm": 1.1766167879104614, "learning_rate": 0.0002, "epoch": 5.524209845606661, "step": 34170}, {"loss": 0.5258, "grad_norm": 0.982292115688324, "learning_rate": 0.0002, "epoch": 5.525826529787406, "step": 34180}, {"loss": 0.5114, "grad_norm": 1.0744609832763672, "learning_rate": 0.0002, "epoch": 5.5274432139681515, "step": 34190}, {"loss": 0.5469, "grad_norm": 1.3831160068511963, "learning_rate": 0.0002, "epoch": 5.529059898148897, "step": 34200}, {"loss": 0.5819, "grad_norm": 1.074771761894226, "learning_rate": 0.0002, "epoch": 5.530676582329642, "step": 34210}, {"loss": 0.5399, "grad_norm": 1.016652226448059, "learning_rate": 0.0002, "epoch": 5.532293266510387, "step": 34220}, {"loss": 0.5158, "grad_norm": 1.2231552600860596, "learning_rate": 0.0002, "epoch": 5.5339099506911325, "step": 34230}, {"loss": 0.5091, "grad_norm": 0.8051198720932007, "learning_rate": 0.0002, "epoch": 5.535526634871878, "step": 34240}, {"loss": 0.5583, "grad_norm": 1.1779674291610718, "learning_rate": 0.0002, "epoch": 5.537143319052623, "step": 34250}, {"loss": 0.5044, "grad_norm": 1.2468291521072388, "learning_rate": 0.0002, "epoch": 5.538760003233368, "step": 34260}, {"loss": 0.523, "grad_norm": 1.14818274974823, "learning_rate": 0.0002, "epoch": 5.540376687414113, "step": 34270}, {"loss": 0.5375, "grad_norm": 1.2362616062164307, "learning_rate": 0.0002, "epoch": 5.541993371594859, "step": 34280}, {"loss": 0.4996, "grad_norm": 1.0206977128982544, "learning_rate": 0.0002, "epoch": 5.543610055775604, "step": 34290}, {"loss": 0.5212, "grad_norm": 1.2018457651138306, "learning_rate": 0.0002, "epoch": 5.54522673995635, "step": 34300}, {"loss": 0.5462, "grad_norm": 1.0349043607711792, "learning_rate": 0.0002, "epoch": 5.546843424137095, "step": 34310}, {"loss": 0.5231, "grad_norm": 1.2022006511688232, "learning_rate": 0.0002, "epoch": 5.54846010831784, "step": 34320}, {"loss": 0.5173, "grad_norm": 1.0810624361038208, "learning_rate": 0.0002, "epoch": 5.550076792498586, "step": 34330}, {"loss": 0.5821, "grad_norm": 1.3297529220581055, "learning_rate": 0.0002, "epoch": 5.551693476679331, "step": 34340}, {"loss": 0.5321, "grad_norm": 0.9722549915313721, "learning_rate": 0.0002, "epoch": 5.553310160860076, "step": 34350}, {"loss": 0.4823, "grad_norm": 0.9903425574302673, "learning_rate": 0.0002, "epoch": 5.554926845040821, "step": 34360}, {"loss": 0.5601, "grad_norm": 0.9568067789077759, "learning_rate": 0.0002, "epoch": 5.556543529221567, "step": 34370}, {"loss": 0.5242, "grad_norm": 1.113870620727539, "learning_rate": 0.0002, "epoch": 5.558160213402312, "step": 34380}, {"loss": 0.5278, "grad_norm": 1.0557632446289062, "learning_rate": 0.0002, "epoch": 5.559776897583057, "step": 34390}, {"loss": 0.5501, "grad_norm": 0.9615673422813416, "learning_rate": 0.0002, "epoch": 5.561393581763802, "step": 34400}, {"loss": 0.5066, "grad_norm": 0.9536027312278748, "learning_rate": 0.0002, "epoch": 5.5630102659445475, "step": 34410}, {"loss": 0.4949, "grad_norm": 0.8808749318122864, "learning_rate": 0.0002, "epoch": 5.564626950125293, "step": 34420}, {"loss": 0.5954, "grad_norm": 1.286132574081421, "learning_rate": 0.0002, "epoch": 5.566243634306038, "step": 34430}, {"loss": 0.5507, "grad_norm": 1.259644865989685, "learning_rate": 0.0002, "epoch": 5.567860318486783, "step": 34440}, {"loss": 0.4922, "grad_norm": 0.9920216798782349, "learning_rate": 0.0002, "epoch": 5.569477002667529, "step": 34450}, {"loss": 0.5527, "grad_norm": 1.182926893234253, "learning_rate": 0.0002, "epoch": 5.5710936868482746, "step": 34460}, {"loss": 0.5185, "grad_norm": 1.1434749364852905, "learning_rate": 0.0002, "epoch": 5.57271037102902, "step": 34470}, {"loss": 0.5256, "grad_norm": 1.2420979738235474, "learning_rate": 0.0002, "epoch": 5.574327055209765, "step": 34480}, {"loss": 0.5039, "grad_norm": 0.9338384866714478, "learning_rate": 0.0002, "epoch": 5.57594373939051, "step": 34490}, {"loss": 0.5634, "grad_norm": 1.0196425914764404, "learning_rate": 0.0002, "epoch": 5.5775604235712555, "step": 34500}, {"loss": 0.5132, "grad_norm": 0.9586997032165527, "learning_rate": 0.0002, "epoch": 5.579177107752001, "step": 34510}, {"loss": 0.5336, "grad_norm": 1.2409086227416992, "learning_rate": 0.0002, "epoch": 5.580793791932746, "step": 34520}, {"loss": 0.5364, "grad_norm": 1.1483757495880127, "learning_rate": 0.0002, "epoch": 5.582410476113491, "step": 34530}, {"loss": 0.5325, "grad_norm": 1.1624305248260498, "learning_rate": 0.0002, "epoch": 5.584027160294236, "step": 34540}, {"loss": 0.5342, "grad_norm": 1.2635223865509033, "learning_rate": 0.0002, "epoch": 5.585643844474982, "step": 34550}, {"loss": 0.4924, "grad_norm": 0.9824051856994629, "learning_rate": 0.0002, "epoch": 5.587260528655727, "step": 34560}, {"loss": 0.5395, "grad_norm": 1.0858620405197144, "learning_rate": 0.0002, "epoch": 5.588877212836472, "step": 34570}, {"loss": 0.5459, "grad_norm": 1.1452655792236328, "learning_rate": 0.0002, "epoch": 5.590493897017217, "step": 34580}, {"loss": 0.5746, "grad_norm": 1.110610842704773, "learning_rate": 0.0002, "epoch": 5.592110581197963, "step": 34590}, {"loss": 0.5285, "grad_norm": 0.9976194500923157, "learning_rate": 0.0002, "epoch": 5.593727265378709, "step": 34600}, {"loss": 0.548, "grad_norm": 1.0698920488357544, "learning_rate": 0.0002, "epoch": 5.595343949559454, "step": 34610}, {"loss": 0.5311, "grad_norm": 1.1505171060562134, "learning_rate": 0.0002, "epoch": 5.596960633740199, "step": 34620}, {"loss": 0.5471, "grad_norm": 1.1014643907546997, "learning_rate": 0.0002, "epoch": 5.598577317920944, "step": 34630}, {"loss": 0.55, "grad_norm": 0.915595293045044, "learning_rate": 0.0002, "epoch": 5.60019400210169, "step": 34640}, {"loss": 0.5821, "grad_norm": 1.1856765747070312, "learning_rate": 0.0002, "epoch": 5.601810686282435, "step": 34650}, {"loss": 0.5502, "grad_norm": 1.1357687711715698, "learning_rate": 0.0002, "epoch": 5.60342737046318, "step": 34660}, {"loss": 0.5034, "grad_norm": 1.0232492685317993, "learning_rate": 0.0002, "epoch": 5.605044054643925, "step": 34670}, {"loss": 0.5357, "grad_norm": 0.9375017881393433, "learning_rate": 0.0002, "epoch": 5.6066607388246705, "step": 34680}, {"loss": 0.5518, "grad_norm": 1.0796529054641724, "learning_rate": 0.0002, "epoch": 5.608277423005416, "step": 34690}, {"loss": 0.5173, "grad_norm": 1.1383336782455444, "learning_rate": 0.0002, "epoch": 5.609894107186161, "step": 34700}, {"loss": 0.5477, "grad_norm": 1.0248544216156006, "learning_rate": 0.0002, "epoch": 5.611510791366906, "step": 34710}, {"loss": 0.5669, "grad_norm": 1.0986040830612183, "learning_rate": 0.0002, "epoch": 5.6131274755476515, "step": 34720}, {"loss": 0.5188, "grad_norm": 1.2689568996429443, "learning_rate": 0.0002, "epoch": 5.614744159728397, "step": 34730}, {"loss": 0.5136, "grad_norm": 1.4044264554977417, "learning_rate": 0.0002, "epoch": 5.616360843909142, "step": 34740}, {"loss": 0.5699, "grad_norm": 1.2084474563598633, "learning_rate": 0.0002, "epoch": 5.617977528089888, "step": 34750}, {"loss": 0.5377, "grad_norm": 1.061248540878296, "learning_rate": 0.0002, "epoch": 5.619594212270633, "step": 34760}, {"loss": 0.5669, "grad_norm": 1.0220764875411987, "learning_rate": 0.0002, "epoch": 5.6212108964513785, "step": 34770}, {"loss": 0.54, "grad_norm": 1.0859092473983765, "learning_rate": 0.0002, "epoch": 5.622827580632124, "step": 34780}, {"loss": 0.5308, "grad_norm": 0.9049732089042664, "learning_rate": 0.0002, "epoch": 5.624444264812869, "step": 34790}, {"loss": 0.5433, "grad_norm": 1.2103937864303589, "learning_rate": 0.0002, "epoch": 5.626060948993614, "step": 34800}, {"loss": 0.5513, "grad_norm": 0.9854230284690857, "learning_rate": 0.0002, "epoch": 5.627677633174359, "step": 34810}, {"loss": 0.5274, "grad_norm": 0.9316635131835938, "learning_rate": 0.0002, "epoch": 5.629294317355105, "step": 34820}, {"loss": 0.5393, "grad_norm": 1.105296015739441, "learning_rate": 0.0002, "epoch": 5.63091100153585, "step": 34830}, {"loss": 0.5527, "grad_norm": 0.993383526802063, "learning_rate": 0.0002, "epoch": 5.632527685716595, "step": 34840}, {"loss": 0.5375, "grad_norm": 1.1544116735458374, "learning_rate": 0.0002, "epoch": 5.63414436989734, "step": 34850}, {"loss": 0.5448, "grad_norm": 1.284475326538086, "learning_rate": 0.0002, "epoch": 5.635761054078086, "step": 34860}, {"loss": 0.5069, "grad_norm": 1.121997594833374, "learning_rate": 0.0002, "epoch": 5.637377738258831, "step": 34870}, {"loss": 0.5335, "grad_norm": 1.213040828704834, "learning_rate": 0.0002, "epoch": 5.638994422439576, "step": 34880}, {"loss": 0.5623, "grad_norm": 1.23222017288208, "learning_rate": 0.0002, "epoch": 5.640611106620321, "step": 34890}, {"loss": 0.5622, "grad_norm": 0.9793637990951538, "learning_rate": 0.0002, "epoch": 5.642227790801067, "step": 34900}, {"loss": 0.5405, "grad_norm": 1.38919997215271, "learning_rate": 0.0002, "epoch": 5.643844474981813, "step": 34910}, {"loss": 0.5007, "grad_norm": 0.8390951156616211, "learning_rate": 0.0002, "epoch": 5.645461159162558, "step": 34920}, {"loss": 0.5974, "grad_norm": 0.9465909004211426, "learning_rate": 0.0002, "epoch": 5.647077843343303, "step": 34930}, {"loss": 0.5264, "grad_norm": 1.066957712173462, "learning_rate": 0.0002, "epoch": 5.648694527524048, "step": 34940}, {"loss": 0.5513, "grad_norm": 0.9842154383659363, "learning_rate": 0.0002, "epoch": 5.650311211704794, "step": 34950}, {"loss": 0.567, "grad_norm": 1.1766440868377686, "learning_rate": 0.0002, "epoch": 5.651927895885539, "step": 34960}, {"loss": 0.5462, "grad_norm": 0.9061306118965149, "learning_rate": 0.0002, "epoch": 5.653544580066284, "step": 34970}, {"loss": 0.5446, "grad_norm": 1.2941309213638306, "learning_rate": 0.0002, "epoch": 5.655161264247029, "step": 34980}, {"loss": 0.5704, "grad_norm": 0.9741247892379761, "learning_rate": 0.0002, "epoch": 5.6567779484277745, "step": 34990}, {"loss": 0.5152, "grad_norm": 1.0784187316894531, "learning_rate": 0.0002, "epoch": 5.65839463260852, "step": 35000}, {"loss": 0.5363, "grad_norm": 0.937889814376831, "learning_rate": 0.0002, "epoch": 5.660011316789265, "step": 35010}, {"loss": 0.5019, "grad_norm": 0.9667879939079285, "learning_rate": 0.0002, "epoch": 5.66162800097001, "step": 35020}, {"loss": 0.5209, "grad_norm": 1.0554876327514648, "learning_rate": 0.0002, "epoch": 5.663244685150756, "step": 35030}, {"loss": 0.523, "grad_norm": 1.2030539512634277, "learning_rate": 0.0002, "epoch": 5.664861369331501, "step": 35040}, {"loss": 0.5406, "grad_norm": 1.0849953889846802, "learning_rate": 0.0002, "epoch": 5.666478053512247, "step": 35050}, {"loss": 0.5747, "grad_norm": 1.1598973274230957, "learning_rate": 0.0002, "epoch": 5.668094737692992, "step": 35060}, {"loss": 0.5488, "grad_norm": 1.0233359336853027, "learning_rate": 0.0002, "epoch": 5.669711421873737, "step": 35070}, {"loss": 0.5409, "grad_norm": 1.1124799251556396, "learning_rate": 0.0002, "epoch": 5.6713281060544825, "step": 35080}, {"loss": 0.5578, "grad_norm": 1.2351475954055786, "learning_rate": 0.0002, "epoch": 5.672944790235228, "step": 35090}, {"loss": 0.5638, "grad_norm": 1.0240728855133057, "learning_rate": 0.0002, "epoch": 5.674561474415973, "step": 35100}, {"loss": 0.5192, "grad_norm": 1.0223692655563354, "learning_rate": 0.0002, "epoch": 5.676178158596718, "step": 35110}, {"loss": 0.524, "grad_norm": 1.4569132328033447, "learning_rate": 0.0002, "epoch": 5.677794842777463, "step": 35120}, {"loss": 0.555, "grad_norm": 0.8983587026596069, "learning_rate": 0.0002, "epoch": 5.679411526958209, "step": 35130}, {"loss": 0.5439, "grad_norm": 1.0775383710861206, "learning_rate": 0.0002, "epoch": 5.681028211138954, "step": 35140}, {"loss": 0.5289, "grad_norm": 0.9800270795822144, "learning_rate": 0.0002, "epoch": 5.682644895319699, "step": 35150}, {"loss": 0.533, "grad_norm": 0.9858237504959106, "learning_rate": 0.0002, "epoch": 5.684261579500444, "step": 35160}, {"loss": 0.5671, "grad_norm": 1.031087040901184, "learning_rate": 0.0002, "epoch": 5.6858782636811895, "step": 35170}, {"loss": 0.5528, "grad_norm": 1.0294365882873535, "learning_rate": 0.0002, "epoch": 5.687494947861936, "step": 35180}, {"loss": 0.5581, "grad_norm": 1.108144760131836, "learning_rate": 0.0002, "epoch": 5.68911163204268, "step": 35190}, {"loss": 0.5373, "grad_norm": 1.0813100337982178, "learning_rate": 0.0002, "epoch": 5.690728316223426, "step": 35200}, {"loss": 0.5429, "grad_norm": 1.3146867752075195, "learning_rate": 0.0002, "epoch": 5.692345000404171, "step": 35210}, {"loss": 0.5297, "grad_norm": 1.16780424118042, "learning_rate": 0.0002, "epoch": 5.693961684584917, "step": 35220}, {"loss": 0.577, "grad_norm": 0.9929125905036926, "learning_rate": 0.0002, "epoch": 5.695578368765662, "step": 35230}, {"loss": 0.5441, "grad_norm": 0.9049441814422607, "learning_rate": 0.0002, "epoch": 5.697195052946407, "step": 35240}, {"loss": 0.5349, "grad_norm": 0.9768866300582886, "learning_rate": 0.0002, "epoch": 5.698811737127152, "step": 35250}, {"loss": 0.542, "grad_norm": 0.8306029438972473, "learning_rate": 0.0002, "epoch": 5.7004284213078975, "step": 35260}, {"loss": 0.4771, "grad_norm": 0.8417280316352844, "learning_rate": 0.0002, "epoch": 5.702045105488643, "step": 35270}, {"loss": 0.574, "grad_norm": 0.9954485893249512, "learning_rate": 0.0002, "epoch": 5.703661789669388, "step": 35280}, {"loss": 0.5469, "grad_norm": 1.2417993545532227, "learning_rate": 0.0002, "epoch": 5.705278473850133, "step": 35290}, {"loss": 0.5275, "grad_norm": 1.1696544885635376, "learning_rate": 0.0002, "epoch": 5.706895158030878, "step": 35300}, {"loss": 0.5188, "grad_norm": 1.2424817085266113, "learning_rate": 0.0002, "epoch": 5.708511842211624, "step": 35310}, {"loss": 0.5595, "grad_norm": 1.1791106462478638, "learning_rate": 0.0002, "epoch": 5.710128526392369, "step": 35320}, {"loss": 0.5076, "grad_norm": 1.202181339263916, "learning_rate": 0.0002, "epoch": 5.711745210573115, "step": 35330}, {"loss": 0.5847, "grad_norm": 1.1006861925125122, "learning_rate": 0.0002, "epoch": 5.713361894753859, "step": 35340}, {"loss": 0.5627, "grad_norm": 1.0918344259262085, "learning_rate": 0.0002, "epoch": 5.7149785789346055, "step": 35350}, {"loss": 0.5677, "grad_norm": 1.0427305698394775, "learning_rate": 0.0002, "epoch": 5.716595263115351, "step": 35360}, {"loss": 0.5288, "grad_norm": 1.0818872451782227, "learning_rate": 0.0002, "epoch": 5.718211947296096, "step": 35370}, {"loss": 0.5296, "grad_norm": 1.186006784439087, "learning_rate": 0.0002, "epoch": 5.719828631476841, "step": 35380}, {"loss": 0.5507, "grad_norm": 1.2073674201965332, "learning_rate": 0.0002, "epoch": 5.721445315657586, "step": 35390}, {"loss": 0.5483, "grad_norm": 1.065338134765625, "learning_rate": 0.0002, "epoch": 5.723061999838332, "step": 35400}, {"loss": 0.5195, "grad_norm": 0.9448973536491394, "learning_rate": 0.0002, "epoch": 5.724678684019077, "step": 35410}, {"loss": 0.5276, "grad_norm": 1.1487499475479126, "learning_rate": 0.0002, "epoch": 5.726295368199822, "step": 35420}, {"loss": 0.5435, "grad_norm": 1.1334216594696045, "learning_rate": 0.0002, "epoch": 5.727912052380567, "step": 35430}, {"loss": 0.5074, "grad_norm": 1.1932826042175293, "learning_rate": 0.0002, "epoch": 5.729528736561313, "step": 35440}, {"loss": 0.5502, "grad_norm": 1.2615786790847778, "learning_rate": 0.0002, "epoch": 5.731145420742058, "step": 35450}, {"loss": 0.5612, "grad_norm": 1.2803694009780884, "learning_rate": 0.0002, "epoch": 5.732762104922803, "step": 35460}, {"loss": 0.5458, "grad_norm": 0.9271906614303589, "learning_rate": 0.0002, "epoch": 5.734378789103548, "step": 35470}, {"loss": 0.5342, "grad_norm": 1.0958917140960693, "learning_rate": 0.0002, "epoch": 5.735995473284294, "step": 35480}, {"loss": 0.538, "grad_norm": 1.1072784662246704, "learning_rate": 0.0002, "epoch": 5.737612157465039, "step": 35490}, {"loss": 0.5683, "grad_norm": 1.1641002893447876, "learning_rate": 0.0002, "epoch": 5.739228841645785, "step": 35500}, {"loss": 0.5252, "grad_norm": 1.0246447324752808, "learning_rate": 0.0002, "epoch": 5.74084552582653, "step": 35510}, {"loss": 0.55, "grad_norm": 1.032474398612976, "learning_rate": 0.0002, "epoch": 5.742462210007275, "step": 35520}, {"loss": 0.4965, "grad_norm": 1.1600854396820068, "learning_rate": 0.0002, "epoch": 5.7440788941880205, "step": 35530}, {"loss": 0.5543, "grad_norm": 1.0686054229736328, "learning_rate": 0.0002, "epoch": 5.745695578368766, "step": 35540}, {"loss": 0.5706, "grad_norm": 1.2314637899398804, "learning_rate": 0.0002, "epoch": 5.747312262549511, "step": 35550}, {"loss": 0.5492, "grad_norm": 0.922134280204773, "learning_rate": 0.0002, "epoch": 5.748928946730256, "step": 35560}, {"loss": 0.5495, "grad_norm": 0.933043360710144, "learning_rate": 0.0002, "epoch": 5.7505456309110015, "step": 35570}, {"loss": 0.5007, "grad_norm": 1.1911931037902832, "learning_rate": 0.0002, "epoch": 5.752162315091747, "step": 35580}, {"loss": 0.5244, "grad_norm": 0.8984857797622681, "learning_rate": 0.0002, "epoch": 5.753778999272492, "step": 35590}, {"loss": 0.5493, "grad_norm": 0.9495107531547546, "learning_rate": 0.0002, "epoch": 5.755395683453237, "step": 35600}, {"loss": 0.5326, "grad_norm": 1.2805472612380981, "learning_rate": 0.0002, "epoch": 5.757012367633982, "step": 35610}, {"loss": 0.5276, "grad_norm": 1.1236625909805298, "learning_rate": 0.0002, "epoch": 5.758629051814728, "step": 35620}, {"loss": 0.6102, "grad_norm": 1.0552798509597778, "learning_rate": 0.0002, "epoch": 5.760245735995474, "step": 35630}, {"loss": 0.5479, "grad_norm": 1.119909644126892, "learning_rate": 0.0002, "epoch": 5.761862420176218, "step": 35640}, {"loss": 0.5282, "grad_norm": 0.8786116242408752, "learning_rate": 0.0002, "epoch": 5.763479104356964, "step": 35650}, {"loss": 0.5406, "grad_norm": 1.2417117357254028, "learning_rate": 0.0002, "epoch": 5.765095788537709, "step": 35660}, {"loss": 0.537, "grad_norm": 1.255200982093811, "learning_rate": 0.0002, "epoch": 5.766712472718455, "step": 35670}, {"loss": 0.5308, "grad_norm": 1.0611358880996704, "learning_rate": 0.0002, "epoch": 5.7683291568992, "step": 35680}, {"loss": 0.5614, "grad_norm": 1.1443911790847778, "learning_rate": 0.0002, "epoch": 5.769945841079945, "step": 35690}, {"loss": 0.5386, "grad_norm": 1.1437989473342896, "learning_rate": 0.0002, "epoch": 5.77156252526069, "step": 35700}, {"loss": 0.537, "grad_norm": 1.1375046968460083, "learning_rate": 0.0002, "epoch": 5.773179209441436, "step": 35710}, {"loss": 0.5198, "grad_norm": 1.0777729749679565, "learning_rate": 0.0002, "epoch": 5.774795893622181, "step": 35720}, {"loss": 0.5521, "grad_norm": 1.1160215139389038, "learning_rate": 0.0002, "epoch": 5.776412577802926, "step": 35730}, {"loss": 0.5569, "grad_norm": 1.1268514394760132, "learning_rate": 0.0002, "epoch": 5.778029261983671, "step": 35740}, {"loss": 0.5311, "grad_norm": 1.2752262353897095, "learning_rate": 0.0002, "epoch": 5.7796459461644165, "step": 35750}, {"loss": 0.5625, "grad_norm": 1.0416184663772583, "learning_rate": 0.0002, "epoch": 5.781262630345162, "step": 35760}, {"loss": 0.5438, "grad_norm": 1.0622444152832031, "learning_rate": 0.0002, "epoch": 5.782879314525907, "step": 35770}, {"loss": 0.5268, "grad_norm": 1.1217877864837646, "learning_rate": 0.0002, "epoch": 5.784495998706653, "step": 35780}, {"loss": 0.5225, "grad_norm": 0.9363139867782593, "learning_rate": 0.0002, "epoch": 5.786112682887398, "step": 35790}, {"loss": 0.5524, "grad_norm": 0.96628737449646, "learning_rate": 0.0002, "epoch": 5.787729367068144, "step": 35800}, {"loss": 0.52, "grad_norm": 0.9572572112083435, "learning_rate": 0.0002, "epoch": 5.789346051248889, "step": 35810}, {"loss": 0.5615, "grad_norm": 0.938724935054779, "learning_rate": 0.0002, "epoch": 5.790962735429634, "step": 35820}, {"loss": 0.5391, "grad_norm": 1.3314417600631714, "learning_rate": 0.0002, "epoch": 5.792579419610379, "step": 35830}, {"loss": 0.5441, "grad_norm": 1.0097602605819702, "learning_rate": 0.0002, "epoch": 5.7941961037911245, "step": 35840}, {"loss": 0.591, "grad_norm": 1.1265122890472412, "learning_rate": 0.0002, "epoch": 5.79581278797187, "step": 35850}, {"loss": 0.5333, "grad_norm": 1.2191909551620483, "learning_rate": 0.0002, "epoch": 5.797429472152615, "step": 35860}, {"loss": 0.5274, "grad_norm": 0.9690808057785034, "learning_rate": 0.0002, "epoch": 5.79904615633336, "step": 35870}, {"loss": 0.5425, "grad_norm": 1.0871665477752686, "learning_rate": 0.0002, "epoch": 5.800662840514105, "step": 35880}, {"loss": 0.5602, "grad_norm": 1.1093597412109375, "learning_rate": 0.0002, "epoch": 5.802279524694851, "step": 35890}, {"loss": 0.5475, "grad_norm": 1.2434282302856445, "learning_rate": 0.0002, "epoch": 5.803896208875596, "step": 35900}, {"loss": 0.5288, "grad_norm": 1.2933623790740967, "learning_rate": 0.0002, "epoch": 5.805512893056341, "step": 35910}, {"loss": 0.5554, "grad_norm": 1.0005441904067993, "learning_rate": 0.0002, "epoch": 5.807129577237086, "step": 35920}, {"loss": 0.5318, "grad_norm": 1.2373108863830566, "learning_rate": 0.0002, "epoch": 5.8087462614178325, "step": 35930}, {"loss": 0.5413, "grad_norm": 1.2622692584991455, "learning_rate": 0.0002, "epoch": 5.810362945598578, "step": 35940}, {"loss": 0.5558, "grad_norm": 1.0112963914871216, "learning_rate": 0.0002, "epoch": 5.811979629779323, "step": 35950}, {"loss": 0.5115, "grad_norm": 1.050572395324707, "learning_rate": 0.0002, "epoch": 5.813596313960068, "step": 35960}, {"loss": 0.5288, "grad_norm": 0.9774560928344727, "learning_rate": 0.0002, "epoch": 5.815212998140813, "step": 35970}, {"loss": 0.585, "grad_norm": 1.19438898563385, "learning_rate": 0.0002, "epoch": 5.816829682321559, "step": 35980}, {"loss": 0.5798, "grad_norm": 1.0267130136489868, "learning_rate": 0.0002, "epoch": 5.818446366502304, "step": 35990}, {"loss": 0.5126, "grad_norm": 0.9813851714134216, "learning_rate": 0.0002, "epoch": 5.820063050683049, "step": 36000}, {"loss": 0.5138, "grad_norm": 0.9177457094192505, "learning_rate": 0.0002, "epoch": 5.821679734863794, "step": 36010}, {"loss": 0.5453, "grad_norm": 1.0020731687545776, "learning_rate": 0.0002, "epoch": 5.8232964190445395, "step": 36020}, {"loss": 0.5646, "grad_norm": 1.073222041130066, "learning_rate": 0.0002, "epoch": 5.824913103225285, "step": 36030}, {"loss": 0.5539, "grad_norm": 1.016337513923645, "learning_rate": 0.0002, "epoch": 5.82652978740603, "step": 36040}, {"loss": 0.5592, "grad_norm": 1.267364263534546, "learning_rate": 0.0002, "epoch": 5.828146471586775, "step": 36050}, {"loss": 0.595, "grad_norm": 1.2730127573013306, "learning_rate": 0.0002, "epoch": 5.8297631557675205, "step": 36060}, {"loss": 0.5247, "grad_norm": 1.108442783355713, "learning_rate": 0.0002, "epoch": 5.831379839948266, "step": 36070}, {"loss": 0.5103, "grad_norm": 1.198072075843811, "learning_rate": 0.0002, "epoch": 5.832996524129012, "step": 36080}, {"loss": 0.5479, "grad_norm": 1.0458786487579346, "learning_rate": 0.0002, "epoch": 5.834613208309757, "step": 36090}, {"loss": 0.5564, "grad_norm": 0.9096664786338806, "learning_rate": 0.0002, "epoch": 5.836229892490502, "step": 36100}, {"loss": 0.5602, "grad_norm": 0.9957793951034546, "learning_rate": 0.0002, "epoch": 5.8378465766712475, "step": 36110}, {"loss": 0.5799, "grad_norm": 1.3693058490753174, "learning_rate": 0.0002, "epoch": 5.839463260851993, "step": 36120}, {"loss": 0.5425, "grad_norm": 1.268608808517456, "learning_rate": 0.0002, "epoch": 5.841079945032738, "step": 36130}, {"loss": 0.5653, "grad_norm": 0.8516020178794861, "learning_rate": 0.0002, "epoch": 5.842696629213483, "step": 36140}, {"loss": 0.5475, "grad_norm": 0.90385502576828, "learning_rate": 0.0002, "epoch": 5.844313313394228, "step": 36150}, {"loss": 0.5274, "grad_norm": 1.0910571813583374, "learning_rate": 0.0002, "epoch": 5.845929997574974, "step": 36160}, {"loss": 0.555, "grad_norm": 0.9417795538902283, "learning_rate": 0.0002, "epoch": 5.847546681755719, "step": 36170}, {"loss": 0.5784, "grad_norm": 1.0027360916137695, "learning_rate": 0.0002, "epoch": 5.849163365936464, "step": 36180}, {"loss": 0.5423, "grad_norm": 1.1480516195297241, "learning_rate": 0.0002, "epoch": 5.850780050117209, "step": 36190}, {"loss": 0.5517, "grad_norm": 1.2431457042694092, "learning_rate": 0.0002, "epoch": 5.852396734297955, "step": 36200}, {"loss": 0.5404, "grad_norm": 1.091465950012207, "learning_rate": 0.0002, "epoch": 5.8540134184787, "step": 36210}, {"loss": 0.53, "grad_norm": 0.9693930745124817, "learning_rate": 0.0002, "epoch": 5.855630102659445, "step": 36220}, {"loss": 0.5453, "grad_norm": 0.9937465190887451, "learning_rate": 0.0002, "epoch": 5.857246786840191, "step": 36230}, {"loss": 0.5621, "grad_norm": 1.0731011629104614, "learning_rate": 0.0002, "epoch": 5.858863471020936, "step": 36240}, {"loss": 0.5687, "grad_norm": 1.0869048833847046, "learning_rate": 0.0002, "epoch": 5.860480155201682, "step": 36250}, {"loss": 0.5576, "grad_norm": 0.9226390719413757, "learning_rate": 0.0002, "epoch": 5.862096839382427, "step": 36260}, {"loss": 0.531, "grad_norm": 1.1755430698394775, "learning_rate": 0.0002, "epoch": 5.863713523563172, "step": 36270}, {"loss": 0.558, "grad_norm": 0.8815974593162537, "learning_rate": 0.0002, "epoch": 5.865330207743917, "step": 36280}, {"loss": 0.5065, "grad_norm": 1.3648751974105835, "learning_rate": 0.0002, "epoch": 5.866946891924663, "step": 36290}, {"loss": 0.536, "grad_norm": 0.8729211091995239, "learning_rate": 0.0002, "epoch": 5.868563576105408, "step": 36300}, {"loss": 0.5192, "grad_norm": 1.0870907306671143, "learning_rate": 0.0002, "epoch": 5.870180260286153, "step": 36310}, {"loss": 0.5609, "grad_norm": 1.1164259910583496, "learning_rate": 0.0002, "epoch": 5.871796944466898, "step": 36320}, {"loss": 0.551, "grad_norm": 1.1572535037994385, "learning_rate": 0.0002, "epoch": 5.8734136286476435, "step": 36330}, {"loss": 0.5898, "grad_norm": 1.0456238985061646, "learning_rate": 0.0002, "epoch": 5.875030312828389, "step": 36340}, {"loss": 0.5008, "grad_norm": 1.1310722827911377, "learning_rate": 0.0002, "epoch": 5.876646997009134, "step": 36350}, {"loss": 0.5352, "grad_norm": 1.0004712343215942, "learning_rate": 0.0002, "epoch": 5.878263681189879, "step": 36360}, {"loss": 0.5632, "grad_norm": 1.0991777181625366, "learning_rate": 0.0002, "epoch": 5.879880365370624, "step": 36370}, {"loss": 0.5815, "grad_norm": 1.2789239883422852, "learning_rate": 0.0002, "epoch": 5.8814970495513705, "step": 36380}, {"loss": 0.56, "grad_norm": 0.9524819850921631, "learning_rate": 0.0002, "epoch": 5.883113733732116, "step": 36390}, {"loss": 0.5701, "grad_norm": 1.1115771532058716, "learning_rate": 0.0002, "epoch": 5.884730417912861, "step": 36400}, {"loss": 0.5463, "grad_norm": 1.37419855594635, "learning_rate": 0.0002, "epoch": 5.886347102093606, "step": 36410}, {"loss": 0.5675, "grad_norm": 1.1449527740478516, "learning_rate": 0.0002, "epoch": 5.8879637862743515, "step": 36420}, {"loss": 0.5255, "grad_norm": 1.198046326637268, "learning_rate": 0.0002, "epoch": 5.889580470455097, "step": 36430}, {"loss": 0.5383, "grad_norm": 1.0180530548095703, "learning_rate": 0.0002, "epoch": 5.891197154635842, "step": 36440}, {"loss": 0.5319, "grad_norm": 1.0516417026519775, "learning_rate": 0.0002, "epoch": 5.892813838816587, "step": 36450}, {"loss": 0.5782, "grad_norm": 1.1658052206039429, "learning_rate": 0.0002, "epoch": 5.894430522997332, "step": 36460}, {"loss": 0.5864, "grad_norm": 1.190699577331543, "learning_rate": 0.0002, "epoch": 5.896047207178078, "step": 36470}, {"loss": 0.5451, "grad_norm": 1.1235495805740356, "learning_rate": 0.0002, "epoch": 5.897663891358823, "step": 36480}, {"loss": 0.5284, "grad_norm": 1.1926926374435425, "learning_rate": 0.0002, "epoch": 5.899280575539568, "step": 36490}, {"loss": 0.5686, "grad_norm": 1.1184662580490112, "learning_rate": 0.0002, "epoch": 5.900897259720313, "step": 36500}, {"loss": 0.5147, "grad_norm": 1.000970721244812, "learning_rate": 0.0002, "epoch": 5.9025139439010585, "step": 36510}, {"loss": 0.5351, "grad_norm": 1.0373306274414062, "learning_rate": 0.0002, "epoch": 5.904130628081804, "step": 36520}, {"loss": 0.535, "grad_norm": 1.0840669870376587, "learning_rate": 0.0002, "epoch": 5.90574731226255, "step": 36530}, {"loss": 0.538, "grad_norm": 0.9908381104469299, "learning_rate": 0.0002, "epoch": 5.907363996443295, "step": 36540}, {"loss": 0.5313, "grad_norm": 1.0456029176712036, "learning_rate": 0.0002, "epoch": 5.90898068062404, "step": 36550}, {"loss": 0.5693, "grad_norm": 1.1381454467773438, "learning_rate": 0.0002, "epoch": 5.910597364804786, "step": 36560}, {"loss": 0.5473, "grad_norm": 0.9440900087356567, "learning_rate": 0.0002, "epoch": 5.912214048985531, "step": 36570}, {"loss": 0.5542, "grad_norm": 1.1674573421478271, "learning_rate": 0.0002, "epoch": 5.913830733166276, "step": 36580}, {"loss": 0.526, "grad_norm": 1.1226966381072998, "learning_rate": 0.0002, "epoch": 5.915447417347021, "step": 36590}, {"loss": 0.6091, "grad_norm": 0.9696915745735168, "learning_rate": 0.0002, "epoch": 5.9170641015277665, "step": 36600}, {"loss": 0.5523, "grad_norm": 0.9593005180358887, "learning_rate": 0.0002, "epoch": 5.918680785708512, "step": 36610}, {"loss": 0.5536, "grad_norm": 1.122169852256775, "learning_rate": 0.0002, "epoch": 5.920297469889257, "step": 36620}, {"loss": 0.5039, "grad_norm": 0.9923415780067444, "learning_rate": 0.0002, "epoch": 5.921914154070002, "step": 36630}, {"loss": 0.5893, "grad_norm": 1.063838005065918, "learning_rate": 0.0002, "epoch": 5.923530838250747, "step": 36640}, {"loss": 0.5799, "grad_norm": 0.9083505272865295, "learning_rate": 0.0002, "epoch": 5.925147522431493, "step": 36650}, {"loss": 0.5264, "grad_norm": 0.9439437985420227, "learning_rate": 0.0002, "epoch": 5.926764206612239, "step": 36660}, {"loss": 0.5891, "grad_norm": 0.9778534173965454, "learning_rate": 0.0002, "epoch": 5.928380890792983, "step": 36670}, {"loss": 0.566, "grad_norm": 0.9723961353302002, "learning_rate": 0.0002, "epoch": 5.929997574973729, "step": 36680}, {"loss": 0.5741, "grad_norm": 1.162333607673645, "learning_rate": 0.0002, "epoch": 5.9316142591544745, "step": 36690}, {"loss": 0.5771, "grad_norm": 1.2784897089004517, "learning_rate": 0.0002, "epoch": 5.93323094333522, "step": 36700}, {"loss": 0.5343, "grad_norm": 1.0924867391586304, "learning_rate": 0.0002, "epoch": 5.934847627515965, "step": 36710}, {"loss": 0.5554, "grad_norm": 1.046922206878662, "learning_rate": 0.0002, "epoch": 5.93646431169671, "step": 36720}, {"loss": 0.5476, "grad_norm": 0.8632535338401794, "learning_rate": 0.0002, "epoch": 5.938080995877455, "step": 36730}, {"loss": 0.5456, "grad_norm": 1.358762502670288, "learning_rate": 0.0002, "epoch": 5.939697680058201, "step": 36740}, {"loss": 0.551, "grad_norm": 1.2058624029159546, "learning_rate": 0.0002, "epoch": 5.941314364238946, "step": 36750}, {"loss": 0.5462, "grad_norm": 1.1396408081054688, "learning_rate": 0.0002, "epoch": 5.942931048419691, "step": 36760}, {"loss": 0.5483, "grad_norm": 1.1510354280471802, "learning_rate": 0.0002, "epoch": 5.944547732600436, "step": 36770}, {"loss": 0.5659, "grad_norm": 1.1401607990264893, "learning_rate": 0.0002, "epoch": 5.946164416781182, "step": 36780}, {"loss": 0.5557, "grad_norm": 1.1871325969696045, "learning_rate": 0.0002, "epoch": 5.947781100961927, "step": 36790}, {"loss": 0.4945, "grad_norm": 0.9928333163261414, "learning_rate": 0.0002, "epoch": 5.949397785142672, "step": 36800}, {"loss": 0.5303, "grad_norm": 1.0549445152282715, "learning_rate": 0.0002, "epoch": 5.951014469323418, "step": 36810}, {"loss": 0.5532, "grad_norm": 0.9791563749313354, "learning_rate": 0.0002, "epoch": 5.9526311535041625, "step": 36820}, {"loss": 0.5317, "grad_norm": 1.1268441677093506, "learning_rate": 0.0002, "epoch": 5.954247837684909, "step": 36830}, {"loss": 0.5585, "grad_norm": 1.0533992052078247, "learning_rate": 0.0002, "epoch": 5.955864521865654, "step": 36840}, {"loss": 0.4972, "grad_norm": 1.023358941078186, "learning_rate": 0.0002, "epoch": 5.957481206046399, "step": 36850}, {"loss": 0.5557, "grad_norm": 1.2631961107254028, "learning_rate": 0.0002, "epoch": 5.959097890227144, "step": 36860}, {"loss": 0.5662, "grad_norm": 0.9397698640823364, "learning_rate": 0.0002, "epoch": 5.9607145744078895, "step": 36870}, {"loss": 0.5775, "grad_norm": 1.1678427457809448, "learning_rate": 0.0002, "epoch": 5.962331258588635, "step": 36880}, {"loss": 0.5435, "grad_norm": 1.1403759717941284, "learning_rate": 0.0002, "epoch": 5.96394794276938, "step": 36890}, {"loss": 0.5479, "grad_norm": 1.030572772026062, "learning_rate": 0.0002, "epoch": 5.965564626950125, "step": 36900}, {"loss": 0.5838, "grad_norm": 1.0992497205734253, "learning_rate": 0.0002, "epoch": 5.9671813111308705, "step": 36910}, {"loss": 0.5452, "grad_norm": 1.075466275215149, "learning_rate": 0.0002, "epoch": 5.968797995311616, "step": 36920}, {"loss": 0.5739, "grad_norm": 1.0153694152832031, "learning_rate": 0.0002, "epoch": 5.970414679492361, "step": 36930}, {"loss": 0.5672, "grad_norm": 0.973193883895874, "learning_rate": 0.0002, "epoch": 5.972031363673106, "step": 36940}, {"loss": 0.5585, "grad_norm": 0.8294678926467896, "learning_rate": 0.0002, "epoch": 5.973648047853851, "step": 36950}, {"loss": 0.5631, "grad_norm": 1.0048716068267822, "learning_rate": 0.0002, "epoch": 5.9752647320345975, "step": 36960}, {"loss": 0.5471, "grad_norm": 0.9714070558547974, "learning_rate": 0.0002, "epoch": 5.976881416215342, "step": 36970}, {"loss": 0.5419, "grad_norm": 0.8667682409286499, "learning_rate": 0.0002, "epoch": 5.978498100396088, "step": 36980}, {"loss": 0.5474, "grad_norm": 1.0461409091949463, "learning_rate": 0.0002, "epoch": 5.980114784576833, "step": 36990}, {"loss": 0.5454, "grad_norm": 0.9229754209518433, "learning_rate": 0.0002, "epoch": 5.981731468757578, "step": 37000}, {"loss": 0.5599, "grad_norm": 1.0406876802444458, "learning_rate": 0.0002, "epoch": 5.983348152938324, "step": 37010}, {"loss": 0.5569, "grad_norm": 0.8993828296661377, "learning_rate": 0.0002, "epoch": 5.984964837119069, "step": 37020}, {"loss": 0.5611, "grad_norm": 1.2260479927062988, "learning_rate": 0.0002, "epoch": 5.986581521299814, "step": 37030}, {"loss": 0.5523, "grad_norm": 1.0107380151748657, "learning_rate": 0.0002, "epoch": 5.988198205480559, "step": 37040}, {"loss": 0.5639, "grad_norm": 1.0240139961242676, "learning_rate": 0.0002, "epoch": 5.989814889661305, "step": 37050}, {"loss": 0.5209, "grad_norm": 1.0185275077819824, "learning_rate": 0.0002, "epoch": 5.99143157384205, "step": 37060}, {"loss": 0.5114, "grad_norm": 1.1361802816390991, "learning_rate": 0.0002, "epoch": 5.993048258022795, "step": 37070}, {"loss": 0.5692, "grad_norm": 1.0395532846450806, "learning_rate": 0.0002, "epoch": 5.99466494220354, "step": 37080}, {"loss": 0.594, "grad_norm": 0.9463558197021484, "learning_rate": 0.0002, "epoch": 5.9962816263842855, "step": 37090}, {"loss": 0.5775, "grad_norm": 1.2066948413848877, "learning_rate": 0.0002, "epoch": 5.997898310565031, "step": 37100}, {"loss": 0.5356, "grad_norm": 0.9749386310577393, "learning_rate": 0.0002, "epoch": 5.999514994745777, "step": 37110}, {"eval_loss": 1.2270219326019287, "eval_runtime": 122.2047, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 6.0, "step": 37113}, {"loss": 0.4855, "grad_norm": 0.9641092419624329, "learning_rate": 0.0002, "epoch": 6.001131678926522, "step": 37120}, {"loss": 0.4112, "grad_norm": 1.103379249572754, "learning_rate": 0.0002, "epoch": 6.002748363107267, "step": 37130}, {"loss": 0.4577, "grad_norm": 0.8381665349006653, "learning_rate": 0.0002, "epoch": 6.004365047288013, "step": 37140}, {"loss": 0.4794, "grad_norm": 1.245323896408081, "learning_rate": 0.0002, "epoch": 6.005981731468758, "step": 37150}, {"loss": 0.4503, "grad_norm": 1.3140289783477783, "learning_rate": 0.0002, "epoch": 6.007598415649503, "step": 37160}, {"loss": 0.4456, "grad_norm": 0.8479695916175842, "learning_rate": 0.0002, "epoch": 6.009215099830248, "step": 37170}, {"loss": 0.4573, "grad_norm": 0.8841437101364136, "learning_rate": 0.0002, "epoch": 6.0108317840109935, "step": 37180}, {"loss": 0.4565, "grad_norm": 0.8900154829025269, "learning_rate": 0.0002, "epoch": 6.012448468191739, "step": 37190}, {"loss": 0.457, "grad_norm": 1.2753345966339111, "learning_rate": 0.0002, "epoch": 6.014065152372484, "step": 37200}, {"loss": 0.4365, "grad_norm": 1.4625498056411743, "learning_rate": 0.0002, "epoch": 6.015681836553229, "step": 37210}, {"loss": 0.4252, "grad_norm": 0.7455034852027893, "learning_rate": 0.0002, "epoch": 6.017298520733974, "step": 37220}, {"loss": 0.4433, "grad_norm": 1.1658862829208374, "learning_rate": 0.0002, "epoch": 6.01891520491472, "step": 37230}, {"loss": 0.4499, "grad_norm": 0.9785751104354858, "learning_rate": 0.0002, "epoch": 6.020531889095465, "step": 37240}, {"loss": 0.4956, "grad_norm": 1.3193122148513794, "learning_rate": 0.0002, "epoch": 6.02214857327621, "step": 37250}, {"loss": 0.4727, "grad_norm": 1.038273572921753, "learning_rate": 0.0002, "epoch": 6.023765257456955, "step": 37260}, {"loss": 0.4395, "grad_norm": 1.0550594329833984, "learning_rate": 0.0002, "epoch": 6.0253819416377015, "step": 37270}, {"loss": 0.4767, "grad_norm": 0.9745930433273315, "learning_rate": 0.0002, "epoch": 6.026998625818447, "step": 37280}, {"loss": 0.4233, "grad_norm": 0.9273530840873718, "learning_rate": 0.0002, "epoch": 6.028615309999192, "step": 37290}, {"loss": 0.4195, "grad_norm": 1.3844057321548462, "learning_rate": 0.0002, "epoch": 6.030231994179937, "step": 37300}, {"loss": 0.4768, "grad_norm": 1.2058762311935425, "learning_rate": 0.0002, "epoch": 6.031848678360682, "step": 37310}, {"loss": 0.4499, "grad_norm": 1.242663025856018, "learning_rate": 0.0002, "epoch": 6.033465362541428, "step": 37320}, {"loss": 0.4597, "grad_norm": 1.3504270315170288, "learning_rate": 0.0002, "epoch": 6.035082046722173, "step": 37330}, {"loss": 0.4402, "grad_norm": 0.8734912276268005, "learning_rate": 0.0002, "epoch": 6.036698730902918, "step": 37340}, {"loss": 0.477, "grad_norm": 1.0182311534881592, "learning_rate": 0.0002, "epoch": 6.038315415083663, "step": 37350}, {"loss": 0.4261, "grad_norm": 0.9898499846458435, "learning_rate": 0.0002, "epoch": 6.0399320992644085, "step": 37360}, {"loss": 0.4459, "grad_norm": 1.0637860298156738, "learning_rate": 0.0002, "epoch": 6.041548783445154, "step": 37370}, {"loss": 0.4958, "grad_norm": 1.0099523067474365, "learning_rate": 0.0002, "epoch": 6.043165467625899, "step": 37380}, {"loss": 0.4459, "grad_norm": 1.1080750226974487, "learning_rate": 0.0002, "epoch": 6.044782151806644, "step": 37390}, {"loss": 0.4473, "grad_norm": 1.2551289796829224, "learning_rate": 0.0002, "epoch": 6.0463988359873895, "step": 37400}, {"loss": 0.468, "grad_norm": 0.8959632515907288, "learning_rate": 0.0002, "epoch": 6.048015520168136, "step": 37410}, {"loss": 0.4255, "grad_norm": 1.1748892068862915, "learning_rate": 0.0002, "epoch": 6.049632204348881, "step": 37420}, {"loss": 0.4458, "grad_norm": 1.3122745752334595, "learning_rate": 0.0002, "epoch": 6.051248888529626, "step": 37430}, {"loss": 0.4676, "grad_norm": 1.0227985382080078, "learning_rate": 0.0002, "epoch": 6.052865572710371, "step": 37440}, {"loss": 0.4503, "grad_norm": 1.0380030870437622, "learning_rate": 0.0002, "epoch": 6.0544822568911165, "step": 37450}, {"loss": 0.4686, "grad_norm": 0.8919622898101807, "learning_rate": 0.0002, "epoch": 6.056098941071862, "step": 37460}, {"loss": 0.4406, "grad_norm": 1.4554150104522705, "learning_rate": 0.0002, "epoch": 6.057715625252607, "step": 37470}, {"loss": 0.4688, "grad_norm": 1.2853292226791382, "learning_rate": 0.0002, "epoch": 6.059332309433352, "step": 37480}, {"loss": 0.4489, "grad_norm": 1.2951840162277222, "learning_rate": 0.0002, "epoch": 6.0609489936140974, "step": 37490}, {"loss": 0.4819, "grad_norm": 1.1750973463058472, "learning_rate": 0.0002, "epoch": 6.062565677794843, "step": 37500}, {"loss": 0.4574, "grad_norm": 0.9328424334526062, "learning_rate": 0.0002, "epoch": 6.064182361975588, "step": 37510}, {"loss": 0.4597, "grad_norm": 1.0353537797927856, "learning_rate": 0.0002, "epoch": 6.065799046156333, "step": 37520}, {"loss": 0.4407, "grad_norm": 1.1594274044036865, "learning_rate": 0.0002, "epoch": 6.067415730337078, "step": 37530}, {"loss": 0.4642, "grad_norm": 0.9034168124198914, "learning_rate": 0.0002, "epoch": 6.069032414517824, "step": 37540}, {"loss": 0.4625, "grad_norm": 1.068617820739746, "learning_rate": 0.0002, "epoch": 6.070649098698569, "step": 37550}, {"loss": 0.4378, "grad_norm": 1.0931321382522583, "learning_rate": 0.0002, "epoch": 6.072265782879315, "step": 37560}, {"loss": 0.4527, "grad_norm": 1.2542688846588135, "learning_rate": 0.0002, "epoch": 6.07388246706006, "step": 37570}, {"loss": 0.4725, "grad_norm": 1.273384928703308, "learning_rate": 0.0002, "epoch": 6.075499151240805, "step": 37580}, {"loss": 0.4928, "grad_norm": 1.4771400690078735, "learning_rate": 0.0002, "epoch": 6.077115835421551, "step": 37590}, {"loss": 0.461, "grad_norm": 1.3751444816589355, "learning_rate": 0.0002, "epoch": 6.078732519602296, "step": 37600}, {"loss": 0.4602, "grad_norm": 1.4532550573349, "learning_rate": 0.0002, "epoch": 6.080349203783041, "step": 37610}, {"loss": 0.4428, "grad_norm": 1.3175991773605347, "learning_rate": 0.0002, "epoch": 6.081965887963786, "step": 37620}, {"loss": 0.4746, "grad_norm": 1.0624970197677612, "learning_rate": 0.0002, "epoch": 6.083582572144532, "step": 37630}, {"loss": 0.413, "grad_norm": 1.099715232849121, "learning_rate": 0.0002, "epoch": 6.085199256325277, "step": 37640}, {"loss": 0.4528, "grad_norm": 1.0380114316940308, "learning_rate": 0.0002, "epoch": 6.086815940506022, "step": 37650}, {"loss": 0.4373, "grad_norm": 1.1136109828948975, "learning_rate": 0.0002, "epoch": 6.088432624686767, "step": 37660}, {"loss": 0.4915, "grad_norm": 0.996498703956604, "learning_rate": 0.0002, "epoch": 6.0900493088675125, "step": 37670}, {"loss": 0.4713, "grad_norm": 1.0552574396133423, "learning_rate": 0.0002, "epoch": 6.091665993048258, "step": 37680}, {"loss": 0.4414, "grad_norm": 1.4108527898788452, "learning_rate": 0.0002, "epoch": 6.093282677229003, "step": 37690}, {"loss": 0.4851, "grad_norm": 1.1323093175888062, "learning_rate": 0.0002, "epoch": 6.094899361409748, "step": 37700}, {"loss": 0.4455, "grad_norm": 0.9364377856254578, "learning_rate": 0.0002, "epoch": 6.096516045590494, "step": 37710}, {"loss": 0.4791, "grad_norm": 1.1300561428070068, "learning_rate": 0.0002, "epoch": 6.0981327297712395, "step": 37720}, {"loss": 0.4539, "grad_norm": 1.0616047382354736, "learning_rate": 0.0002, "epoch": 6.099749413951985, "step": 37730}, {"loss": 0.4516, "grad_norm": 1.1205905675888062, "learning_rate": 0.0002, "epoch": 6.10136609813273, "step": 37740}, {"loss": 0.4688, "grad_norm": 0.9592534303665161, "learning_rate": 0.0002, "epoch": 6.102982782313475, "step": 37750}, {"loss": 0.4494, "grad_norm": 0.9797531962394714, "learning_rate": 0.0002, "epoch": 6.1045994664942205, "step": 37760}, {"loss": 0.4237, "grad_norm": 1.093404769897461, "learning_rate": 0.0002, "epoch": 6.106216150674966, "step": 37770}, {"loss": 0.4691, "grad_norm": 1.2172642946243286, "learning_rate": 0.0002, "epoch": 6.107832834855711, "step": 37780}, {"loss": 0.4398, "grad_norm": 1.0467255115509033, "learning_rate": 0.0002, "epoch": 6.109449519036456, "step": 37790}, {"loss": 0.4676, "grad_norm": 1.159318208694458, "learning_rate": 0.0002, "epoch": 6.111066203217201, "step": 37800}, {"loss": 0.4539, "grad_norm": 1.0615603923797607, "learning_rate": 0.0002, "epoch": 6.112682887397947, "step": 37810}, {"loss": 0.4957, "grad_norm": 1.0542045831680298, "learning_rate": 0.0002, "epoch": 6.114299571578692, "step": 37820}, {"loss": 0.4512, "grad_norm": 0.8962697982788086, "learning_rate": 0.0002, "epoch": 6.115916255759437, "step": 37830}, {"loss": 0.4519, "grad_norm": 1.106352686882019, "learning_rate": 0.0002, "epoch": 6.117532939940182, "step": 37840}, {"loss": 0.4421, "grad_norm": 1.1660276651382446, "learning_rate": 0.0002, "epoch": 6.1191496241209276, "step": 37850}, {"loss": 0.4701, "grad_norm": 1.3524385690689087, "learning_rate": 0.0002, "epoch": 6.120766308301674, "step": 37860}, {"loss": 0.4684, "grad_norm": 1.1056050062179565, "learning_rate": 0.0002, "epoch": 6.122382992482419, "step": 37870}, {"loss": 0.4518, "grad_norm": 1.0772725343704224, "learning_rate": 0.0002, "epoch": 6.123999676663164, "step": 37880}, {"loss": 0.4356, "grad_norm": 1.1011115312576294, "learning_rate": 0.0002, "epoch": 6.125616360843909, "step": 37890}, {"loss": 0.4909, "grad_norm": 0.8952536582946777, "learning_rate": 0.0002, "epoch": 6.127233045024655, "step": 37900}, {"loss": 0.4299, "grad_norm": 1.244398593902588, "learning_rate": 0.0002, "epoch": 6.1288497292054, "step": 37910}, {"loss": 0.4764, "grad_norm": 0.9658283591270447, "learning_rate": 0.0002, "epoch": 6.130466413386145, "step": 37920}, {"loss": 0.4378, "grad_norm": 1.0649068355560303, "learning_rate": 0.0002, "epoch": 6.13208309756689, "step": 37930}, {"loss": 0.4638, "grad_norm": 0.94698166847229, "learning_rate": 0.0002, "epoch": 6.1336997817476355, "step": 37940}, {"loss": 0.488, "grad_norm": 1.1450897455215454, "learning_rate": 0.0002, "epoch": 6.135316465928381, "step": 37950}, {"loss": 0.4791, "grad_norm": 1.032482624053955, "learning_rate": 0.0002, "epoch": 6.136933150109126, "step": 37960}, {"loss": 0.4179, "grad_norm": 1.0993428230285645, "learning_rate": 0.0002, "epoch": 6.138549834289871, "step": 37970}, {"loss": 0.4781, "grad_norm": 1.2907029390335083, "learning_rate": 0.0002, "epoch": 6.1401665184706165, "step": 37980}, {"loss": 0.4671, "grad_norm": 1.1007903814315796, "learning_rate": 0.0002, "epoch": 6.141783202651362, "step": 37990}, {"loss": 0.4213, "grad_norm": 0.9286124110221863, "learning_rate": 0.0002, "epoch": 6.143399886832107, "step": 38000}, {"loss": 0.4741, "grad_norm": 1.1426366567611694, "learning_rate": 0.0002, "epoch": 6.145016571012853, "step": 38010}, {"loss": 0.4746, "grad_norm": 1.2608287334442139, "learning_rate": 0.0002, "epoch": 6.146633255193598, "step": 38020}, {"loss": 0.454, "grad_norm": 1.1346837282180786, "learning_rate": 0.0002, "epoch": 6.1482499393743435, "step": 38030}, {"loss": 0.4469, "grad_norm": 1.144080400466919, "learning_rate": 0.0002, "epoch": 6.149866623555089, "step": 38040}, {"loss": 0.4515, "grad_norm": 1.3456705808639526, "learning_rate": 0.0002, "epoch": 6.151483307735834, "step": 38050}, {"loss": 0.4775, "grad_norm": 1.0517960786819458, "learning_rate": 0.0002, "epoch": 6.153099991916579, "step": 38060}, {"loss": 0.4986, "grad_norm": 1.1887445449829102, "learning_rate": 0.0002, "epoch": 6.154716676097324, "step": 38070}, {"loss": 0.4516, "grad_norm": 1.0449163913726807, "learning_rate": 0.0002, "epoch": 6.15633336027807, "step": 38080}, {"loss": 0.4808, "grad_norm": 1.3218743801116943, "learning_rate": 0.0002, "epoch": 6.157950044458815, "step": 38090}, {"loss": 0.4632, "grad_norm": 1.003208875656128, "learning_rate": 0.0002, "epoch": 6.15956672863956, "step": 38100}, {"loss": 0.4978, "grad_norm": 1.008623719215393, "learning_rate": 0.0002, "epoch": 6.161183412820305, "step": 38110}, {"loss": 0.4608, "grad_norm": 1.2122787237167358, "learning_rate": 0.0002, "epoch": 6.162800097001051, "step": 38120}, {"loss": 0.4666, "grad_norm": 1.253403902053833, "learning_rate": 0.0002, "epoch": 6.164416781181796, "step": 38130}, {"loss": 0.4778, "grad_norm": 1.2289724349975586, "learning_rate": 0.0002, "epoch": 6.166033465362541, "step": 38140}, {"loss": 0.4774, "grad_norm": 1.330694556236267, "learning_rate": 0.0002, "epoch": 6.167650149543286, "step": 38150}, {"loss": 0.4699, "grad_norm": 1.0946741104125977, "learning_rate": 0.0002, "epoch": 6.169266833724032, "step": 38160}, {"loss": 0.4816, "grad_norm": 1.0719934701919556, "learning_rate": 0.0002, "epoch": 6.170883517904778, "step": 38170}, {"loss": 0.4678, "grad_norm": 1.1142133474349976, "learning_rate": 0.0002, "epoch": 6.172500202085523, "step": 38180}, {"loss": 0.4911, "grad_norm": 1.1221938133239746, "learning_rate": 0.0002, "epoch": 6.174116886266268, "step": 38190}, {"loss": 0.4462, "grad_norm": 1.1391617059707642, "learning_rate": 0.0002, "epoch": 6.175733570447013, "step": 38200}, {"loss": 0.4867, "grad_norm": 1.2263455390930176, "learning_rate": 0.0002, "epoch": 6.1773502546277586, "step": 38210}, {"loss": 0.4633, "grad_norm": 1.0930434465408325, "learning_rate": 0.0002, "epoch": 6.178966938808504, "step": 38220}, {"loss": 0.4406, "grad_norm": 1.3489030599594116, "learning_rate": 0.0002, "epoch": 6.180583622989249, "step": 38230}, {"loss": 0.4994, "grad_norm": 1.1383486986160278, "learning_rate": 0.0002, "epoch": 6.182200307169994, "step": 38240}, {"loss": 0.4851, "grad_norm": 1.2408897876739502, "learning_rate": 0.0002, "epoch": 6.1838169913507395, "step": 38250}, {"loss": 0.4848, "grad_norm": 1.1436222791671753, "learning_rate": 0.0002, "epoch": 6.185433675531485, "step": 38260}, {"loss": 0.4594, "grad_norm": 1.370117425918579, "learning_rate": 0.0002, "epoch": 6.18705035971223, "step": 38270}, {"loss": 0.5023, "grad_norm": 0.8862423300743103, "learning_rate": 0.0002, "epoch": 6.188667043892975, "step": 38280}, {"loss": 0.4559, "grad_norm": 0.9603779315948486, "learning_rate": 0.0002, "epoch": 6.19028372807372, "step": 38290}, {"loss": 0.4835, "grad_norm": 1.389291524887085, "learning_rate": 0.0002, "epoch": 6.191900412254466, "step": 38300}, {"loss": 0.4435, "grad_norm": 1.0767031908035278, "learning_rate": 0.0002, "epoch": 6.193517096435212, "step": 38310}, {"loss": 0.4683, "grad_norm": 1.1800403594970703, "learning_rate": 0.0002, "epoch": 6.195133780615957, "step": 38320}, {"loss": 0.4608, "grad_norm": 0.997891366481781, "learning_rate": 0.0002, "epoch": 6.196750464796702, "step": 38330}, {"loss": 0.4575, "grad_norm": 1.1201492547988892, "learning_rate": 0.0002, "epoch": 6.1983671489774474, "step": 38340}, {"loss": 0.4952, "grad_norm": 0.9769026637077332, "learning_rate": 0.0002, "epoch": 6.199983833158193, "step": 38350}, {"loss": 0.4563, "grad_norm": 0.9447069764137268, "learning_rate": 0.0002, "epoch": 6.201600517338938, "step": 38360}, {"loss": 0.516, "grad_norm": 1.0959235429763794, "learning_rate": 0.0002, "epoch": 6.203217201519683, "step": 38370}, {"loss": 0.4688, "grad_norm": 1.2495406866073608, "learning_rate": 0.0002, "epoch": 6.204833885700428, "step": 38380}, {"loss": 0.4445, "grad_norm": 0.8589218258857727, "learning_rate": 0.0002, "epoch": 6.206450569881174, "step": 38390}, {"loss": 0.4808, "grad_norm": 0.959155797958374, "learning_rate": 0.0002, "epoch": 6.208067254061919, "step": 38400}, {"loss": 0.4622, "grad_norm": 1.0105533599853516, "learning_rate": 0.0002, "epoch": 6.209683938242664, "step": 38410}, {"loss": 0.4887, "grad_norm": 0.9824615120887756, "learning_rate": 0.0002, "epoch": 6.211300622423409, "step": 38420}, {"loss": 0.4656, "grad_norm": 0.8616500496864319, "learning_rate": 0.0002, "epoch": 6.2129173066041545, "step": 38430}, {"loss": 0.449, "grad_norm": 1.2917758226394653, "learning_rate": 0.0002, "epoch": 6.2145339907849, "step": 38440}, {"loss": 0.4201, "grad_norm": 1.0564531087875366, "learning_rate": 0.0002, "epoch": 6.216150674965646, "step": 38450}, {"loss": 0.4849, "grad_norm": 1.152331829071045, "learning_rate": 0.0002, "epoch": 6.217767359146391, "step": 38460}, {"loss": 0.4887, "grad_norm": 0.9152206778526306, "learning_rate": 0.0002, "epoch": 6.219384043327136, "step": 38470}, {"loss": 0.4686, "grad_norm": 0.9931167960166931, "learning_rate": 0.0002, "epoch": 6.221000727507882, "step": 38480}, {"loss": 0.4765, "grad_norm": 1.3248072862625122, "learning_rate": 0.0002, "epoch": 6.222617411688627, "step": 38490}, {"loss": 0.4636, "grad_norm": 1.3916507959365845, "learning_rate": 0.0002, "epoch": 6.224234095869372, "step": 38500}, {"loss": 0.506, "grad_norm": 1.1775140762329102, "learning_rate": 0.0002, "epoch": 6.225850780050117, "step": 38510}, {"loss": 0.47, "grad_norm": 1.1581059694290161, "learning_rate": 0.0002, "epoch": 6.2274674642308625, "step": 38520}, {"loss": 0.4679, "grad_norm": 1.359320878982544, "learning_rate": 0.0002, "epoch": 6.229084148411608, "step": 38530}, {"loss": 0.4697, "grad_norm": 1.185041904449463, "learning_rate": 0.0002, "epoch": 6.230700832592353, "step": 38540}, {"loss": 0.4815, "grad_norm": 1.1861097812652588, "learning_rate": 0.0002, "epoch": 6.232317516773098, "step": 38550}, {"loss": 0.4925, "grad_norm": 1.126990556716919, "learning_rate": 0.0002, "epoch": 6.233934200953843, "step": 38560}, {"loss": 0.4414, "grad_norm": 0.9744541049003601, "learning_rate": 0.0002, "epoch": 6.235550885134589, "step": 38570}, {"loss": 0.4577, "grad_norm": 1.1260887384414673, "learning_rate": 0.0002, "epoch": 6.237167569315334, "step": 38580}, {"loss": 0.4852, "grad_norm": 1.1290327310562134, "learning_rate": 0.0002, "epoch": 6.238784253496079, "step": 38590}, {"loss": 0.4805, "grad_norm": 1.0952879190444946, "learning_rate": 0.0002, "epoch": 6.240400937676825, "step": 38600}, {"loss": 0.4436, "grad_norm": 1.1037684679031372, "learning_rate": 0.0002, "epoch": 6.2420176218575705, "step": 38610}, {"loss": 0.466, "grad_norm": 1.1356085538864136, "learning_rate": 0.0002, "epoch": 6.243634306038316, "step": 38620}, {"loss": 0.5129, "grad_norm": 1.0677106380462646, "learning_rate": 0.0002, "epoch": 6.245250990219061, "step": 38630}, {"loss": 0.4907, "grad_norm": 1.1573411226272583, "learning_rate": 0.0002, "epoch": 6.246867674399806, "step": 38640}, {"loss": 0.5098, "grad_norm": 1.2707505226135254, "learning_rate": 0.0002, "epoch": 6.248484358580551, "step": 38650}, {"loss": 0.4926, "grad_norm": 1.0480109453201294, "learning_rate": 0.0002, "epoch": 6.250101042761297, "step": 38660}, {"loss": 0.4654, "grad_norm": 1.3668724298477173, "learning_rate": 0.0002, "epoch": 6.251717726942042, "step": 38670}, {"loss": 0.5128, "grad_norm": 1.217289686203003, "learning_rate": 0.0002, "epoch": 6.253334411122787, "step": 38680}, {"loss": 0.4621, "grad_norm": 1.2950236797332764, "learning_rate": 0.0002, "epoch": 6.254951095303532, "step": 38690}, {"loss": 0.5076, "grad_norm": 1.4506934881210327, "learning_rate": 0.0002, "epoch": 6.256567779484278, "step": 38700}, {"loss": 0.4803, "grad_norm": 1.1248667240142822, "learning_rate": 0.0002, "epoch": 6.258184463665023, "step": 38710}, {"loss": 0.4746, "grad_norm": 1.3384023904800415, "learning_rate": 0.0002, "epoch": 6.259801147845768, "step": 38720}, {"loss": 0.473, "grad_norm": 1.128074288368225, "learning_rate": 0.0002, "epoch": 6.261417832026513, "step": 38730}, {"loss": 0.4638, "grad_norm": 1.1169012784957886, "learning_rate": 0.0002, "epoch": 6.263034516207259, "step": 38740}, {"loss": 0.4747, "grad_norm": 1.195198893547058, "learning_rate": 0.0002, "epoch": 6.264651200388005, "step": 38750}, {"loss": 0.4906, "grad_norm": 1.2471518516540527, "learning_rate": 0.0002, "epoch": 6.26626788456875, "step": 38760}, {"loss": 0.4507, "grad_norm": 1.2646394968032837, "learning_rate": 0.0002, "epoch": 6.267884568749495, "step": 38770}, {"loss": 0.4934, "grad_norm": 1.0286450386047363, "learning_rate": 0.0002, "epoch": 6.26950125293024, "step": 38780}, {"loss": 0.4787, "grad_norm": 1.2440695762634277, "learning_rate": 0.0002, "epoch": 6.2711179371109855, "step": 38790}, {"loss": 0.4806, "grad_norm": 0.8941256403923035, "learning_rate": 0.0002, "epoch": 6.272734621291731, "step": 38800}, {"loss": 0.4741, "grad_norm": 1.0693447589874268, "learning_rate": 0.0002, "epoch": 6.274351305472476, "step": 38810}, {"loss": 0.4408, "grad_norm": 1.0936840772628784, "learning_rate": 0.0002, "epoch": 6.275967989653221, "step": 38820}, {"loss": 0.4729, "grad_norm": 1.0961874723434448, "learning_rate": 0.0002, "epoch": 6.2775846738339665, "step": 38830}, {"loss": 0.4504, "grad_norm": 1.1465433835983276, "learning_rate": 0.0002, "epoch": 6.279201358014712, "step": 38840}, {"loss": 0.4771, "grad_norm": 1.2987004518508911, "learning_rate": 0.0002, "epoch": 6.280818042195457, "step": 38850}, {"loss": 0.4945, "grad_norm": 1.1310304403305054, "learning_rate": 0.0002, "epoch": 6.282434726376202, "step": 38860}, {"loss": 0.5346, "grad_norm": 1.306538462638855, "learning_rate": 0.0002, "epoch": 6.284051410556947, "step": 38870}, {"loss": 0.4873, "grad_norm": 1.2405401468276978, "learning_rate": 0.0002, "epoch": 6.285668094737693, "step": 38880}, {"loss": 0.4929, "grad_norm": 1.0934767723083496, "learning_rate": 0.0002, "epoch": 6.287284778918439, "step": 38890}, {"loss": 0.4853, "grad_norm": 1.3370496034622192, "learning_rate": 0.0002, "epoch": 6.288901463099184, "step": 38900}, {"loss": 0.4892, "grad_norm": 1.0319404602050781, "learning_rate": 0.0002, "epoch": 6.290518147279929, "step": 38910}, {"loss": 0.4685, "grad_norm": 0.9734271168708801, "learning_rate": 0.0002, "epoch": 6.292134831460674, "step": 38920}, {"loss": 0.5085, "grad_norm": 1.0940454006195068, "learning_rate": 0.0002, "epoch": 6.29375151564142, "step": 38930}, {"loss": 0.4985, "grad_norm": 1.036500334739685, "learning_rate": 0.0002, "epoch": 6.295368199822165, "step": 38940}, {"loss": 0.4878, "grad_norm": 1.020308256149292, "learning_rate": 0.0002, "epoch": 6.29698488400291, "step": 38950}, {"loss": 0.4668, "grad_norm": 1.1416399478912354, "learning_rate": 0.0002, "epoch": 6.298601568183655, "step": 38960}, {"loss": 0.4727, "grad_norm": 1.2497479915618896, "learning_rate": 0.0002, "epoch": 6.300218252364401, "step": 38970}, {"loss": 0.4721, "grad_norm": 1.1692523956298828, "learning_rate": 0.0002, "epoch": 6.301834936545146, "step": 38980}, {"loss": 0.505, "grad_norm": 1.0693109035491943, "learning_rate": 0.0002, "epoch": 6.303451620725891, "step": 38990}, {"loss": 0.4875, "grad_norm": 0.8883291482925415, "learning_rate": 0.0002, "epoch": 6.305068304906636, "step": 39000}, {"loss": 0.5371, "grad_norm": 1.1445088386535645, "learning_rate": 0.0002, "epoch": 6.3066849890873815, "step": 39010}, {"loss": 0.5089, "grad_norm": 1.226792335510254, "learning_rate": 0.0002, "epoch": 6.308301673268127, "step": 39020}, {"loss": 0.474, "grad_norm": 1.0498932600021362, "learning_rate": 0.0002, "epoch": 6.309918357448872, "step": 39030}, {"loss": 0.4964, "grad_norm": 1.0834535360336304, "learning_rate": 0.0002, "epoch": 6.311535041629618, "step": 39040}, {"loss": 0.4733, "grad_norm": 1.144666075706482, "learning_rate": 0.0002, "epoch": 6.313151725810363, "step": 39050}, {"loss": 0.4784, "grad_norm": 1.1468489170074463, "learning_rate": 0.0002, "epoch": 6.3147684099911086, "step": 39060}, {"loss": 0.4911, "grad_norm": 1.290949821472168, "learning_rate": 0.0002, "epoch": 6.316385094171854, "step": 39070}, {"loss": 0.5002, "grad_norm": 1.087868094444275, "learning_rate": 0.0002, "epoch": 6.318001778352599, "step": 39080}, {"loss": 0.4944, "grad_norm": 1.0156296491622925, "learning_rate": 0.0002, "epoch": 6.319618462533344, "step": 39090}, {"loss": 0.5019, "grad_norm": 1.0805060863494873, "learning_rate": 0.0002, "epoch": 6.3212351467140895, "step": 39100}, {"loss": 0.4598, "grad_norm": 0.9030579924583435, "learning_rate": 0.0002, "epoch": 6.322851830894835, "step": 39110}, {"loss": 0.4635, "grad_norm": 1.1488285064697266, "learning_rate": 0.0002, "epoch": 6.32446851507558, "step": 39120}, {"loss": 0.5368, "grad_norm": 1.2050796747207642, "learning_rate": 0.0002, "epoch": 6.326085199256325, "step": 39130}, {"loss": 0.4854, "grad_norm": 1.093451738357544, "learning_rate": 0.0002, "epoch": 6.32770188343707, "step": 39140}, {"loss": 0.5055, "grad_norm": 1.2046772241592407, "learning_rate": 0.0002, "epoch": 6.329318567617816, "step": 39150}, {"loss": 0.4703, "grad_norm": 1.045777678489685, "learning_rate": 0.0002, "epoch": 6.330935251798561, "step": 39160}, {"loss": 0.513, "grad_norm": 1.2008492946624756, "learning_rate": 0.0002, "epoch": 6.332551935979306, "step": 39170}, {"loss": 0.4909, "grad_norm": 1.0613869428634644, "learning_rate": 0.0002, "epoch": 6.334168620160051, "step": 39180}, {"loss": 0.4708, "grad_norm": 1.058440089225769, "learning_rate": 0.0002, "epoch": 6.3357853043407975, "step": 39190}, {"loss": 0.4719, "grad_norm": 1.195658802986145, "learning_rate": 0.0002, "epoch": 6.337401988521543, "step": 39200}, {"loss": 0.4901, "grad_norm": 1.1595174074172974, "learning_rate": 0.0002, "epoch": 6.339018672702288, "step": 39210}, {"loss": 0.4587, "grad_norm": 1.0674750804901123, "learning_rate": 0.0002, "epoch": 6.340635356883033, "step": 39220}, {"loss": 0.4801, "grad_norm": 1.3306758403778076, "learning_rate": 0.0002, "epoch": 6.342252041063778, "step": 39230}, {"loss": 0.4839, "grad_norm": 1.3582593202590942, "learning_rate": 0.0002, "epoch": 6.343868725244524, "step": 39240}, {"loss": 0.4964, "grad_norm": 1.2351572513580322, "learning_rate": 0.0002, "epoch": 6.345485409425269, "step": 39250}, {"loss": 0.4806, "grad_norm": 1.3623450994491577, "learning_rate": 0.0002, "epoch": 6.347102093606014, "step": 39260}, {"loss": 0.466, "grad_norm": 1.201270580291748, "learning_rate": 0.0002, "epoch": 6.348718777786759, "step": 39270}, {"loss": 0.4899, "grad_norm": 0.9300584197044373, "learning_rate": 0.0002, "epoch": 6.3503354619675045, "step": 39280}, {"loss": 0.4867, "grad_norm": 0.944525957107544, "learning_rate": 0.0002, "epoch": 6.35195214614825, "step": 39290}, {"loss": 0.4954, "grad_norm": 1.4263732433319092, "learning_rate": 0.0002, "epoch": 6.353568830328995, "step": 39300}, {"loss": 0.4982, "grad_norm": 1.392592191696167, "learning_rate": 0.0002, "epoch": 6.35518551450974, "step": 39310}, {"loss": 0.4868, "grad_norm": 1.0753393173217773, "learning_rate": 0.0002, "epoch": 6.3568021986904855, "step": 39320}, {"loss": 0.4896, "grad_norm": 1.0088151693344116, "learning_rate": 0.0002, "epoch": 6.358418882871231, "step": 39330}, {"loss": 0.4684, "grad_norm": 1.1784582138061523, "learning_rate": 0.0002, "epoch": 6.360035567051977, "step": 39340}, {"loss": 0.4732, "grad_norm": 1.020526647567749, "learning_rate": 0.0002, "epoch": 6.361652251232722, "step": 39350}, {"loss": 0.5177, "grad_norm": 1.1400747299194336, "learning_rate": 0.0002, "epoch": 6.363268935413467, "step": 39360}, {"loss": 0.4976, "grad_norm": 0.9960665702819824, "learning_rate": 0.0002, "epoch": 6.3648856195942125, "step": 39370}, {"loss": 0.483, "grad_norm": 1.1547569036483765, "learning_rate": 0.0002, "epoch": 6.366502303774958, "step": 39380}, {"loss": 0.4861, "grad_norm": 1.2180676460266113, "learning_rate": 0.0002, "epoch": 6.368118987955703, "step": 39390}, {"loss": 0.4805, "grad_norm": 1.1391799449920654, "learning_rate": 0.0002, "epoch": 6.369735672136448, "step": 39400}, {"loss": 0.5004, "grad_norm": 1.2893574237823486, "learning_rate": 0.0002, "epoch": 6.371352356317193, "step": 39410}, {"loss": 0.4807, "grad_norm": 1.192878246307373, "learning_rate": 0.0002, "epoch": 6.372969040497939, "step": 39420}, {"loss": 0.4637, "grad_norm": 0.9771704077720642, "learning_rate": 0.0002, "epoch": 6.374585724678684, "step": 39430}, {"loss": 0.4867, "grad_norm": 1.285387635231018, "learning_rate": 0.0002, "epoch": 6.376202408859429, "step": 39440}, {"loss": 0.4593, "grad_norm": 1.019957184791565, "learning_rate": 0.0002, "epoch": 6.377819093040174, "step": 39450}, {"loss": 0.473, "grad_norm": 1.2002915143966675, "learning_rate": 0.0002, "epoch": 6.37943577722092, "step": 39460}, {"loss": 0.5025, "grad_norm": 1.3285092115402222, "learning_rate": 0.0002, "epoch": 6.381052461401665, "step": 39470}, {"loss": 0.4626, "grad_norm": 1.097846269607544, "learning_rate": 0.0002, "epoch": 6.38266914558241, "step": 39480}, {"loss": 0.5109, "grad_norm": 0.9537988305091858, "learning_rate": 0.0002, "epoch": 6.384285829763156, "step": 39490}, {"loss": 0.4492, "grad_norm": 1.0350042581558228, "learning_rate": 0.0002, "epoch": 6.385902513943901, "step": 39500}, {"loss": 0.4824, "grad_norm": 0.9559133052825928, "learning_rate": 0.0002, "epoch": 6.387519198124647, "step": 39510}, {"loss": 0.5189, "grad_norm": 0.9615123271942139, "learning_rate": 0.0002, "epoch": 6.389135882305392, "step": 39520}, {"loss": 0.4915, "grad_norm": 1.0604504346847534, "learning_rate": 0.0002, "epoch": 6.390752566486137, "step": 39530}, {"loss": 0.5315, "grad_norm": 1.2460750341415405, "learning_rate": 0.0002, "epoch": 6.392369250666882, "step": 39540}, {"loss": 0.4929, "grad_norm": 1.1496477127075195, "learning_rate": 0.0002, "epoch": 6.393985934847628, "step": 39550}, {"loss": 0.4872, "grad_norm": 1.048043966293335, "learning_rate": 0.0002, "epoch": 6.395602619028373, "step": 39560}, {"loss": 0.5231, "grad_norm": 1.333539366722107, "learning_rate": 0.0002, "epoch": 6.397219303209118, "step": 39570}, {"loss": 0.4877, "grad_norm": 1.0605626106262207, "learning_rate": 0.0002, "epoch": 6.398835987389863, "step": 39580}, {"loss": 0.4643, "grad_norm": 1.163220763206482, "learning_rate": 0.0002, "epoch": 6.4004526715706085, "step": 39590}, {"loss": 0.4824, "grad_norm": 1.1878494024276733, "learning_rate": 0.0002, "epoch": 6.402069355751354, "step": 39600}, {"loss": 0.5242, "grad_norm": 1.4630796909332275, "learning_rate": 0.0002, "epoch": 6.403686039932099, "step": 39610}, {"loss": 0.4985, "grad_norm": 1.073255181312561, "learning_rate": 0.0002, "epoch": 6.405302724112844, "step": 39620}, {"loss": 0.5108, "grad_norm": 1.0538873672485352, "learning_rate": 0.0002, "epoch": 6.406919408293589, "step": 39630}, {"loss": 0.4801, "grad_norm": 1.015525221824646, "learning_rate": 0.0002, "epoch": 6.4085360924743355, "step": 39640}, {"loss": 0.4781, "grad_norm": 1.1454379558563232, "learning_rate": 0.0002, "epoch": 6.410152776655081, "step": 39650}, {"loss": 0.498, "grad_norm": 1.2801800966262817, "learning_rate": 0.0002, "epoch": 6.411769460835826, "step": 39660}, {"loss": 0.4804, "grad_norm": 1.077579140663147, "learning_rate": 0.0002, "epoch": 6.413386145016571, "step": 39670}, {"loss": 0.51, "grad_norm": 1.376662015914917, "learning_rate": 0.0002, "epoch": 6.4150028291973165, "step": 39680}, {"loss": 0.4956, "grad_norm": 1.2064344882965088, "learning_rate": 0.0002, "epoch": 6.416619513378062, "step": 39690}, {"loss": 0.4762, "grad_norm": 1.0689115524291992, "learning_rate": 0.0002, "epoch": 6.418236197558807, "step": 39700}, {"loss": 0.4762, "grad_norm": 0.9997019171714783, "learning_rate": 0.0002, "epoch": 6.419852881739552, "step": 39710}, {"loss": 0.49, "grad_norm": 1.2368080615997314, "learning_rate": 0.0002, "epoch": 6.421469565920297, "step": 39720}, {"loss": 0.4774, "grad_norm": 1.2085820436477661, "learning_rate": 0.0002, "epoch": 6.423086250101043, "step": 39730}, {"loss": 0.4671, "grad_norm": 1.057246208190918, "learning_rate": 0.0002, "epoch": 6.424702934281788, "step": 39740}, {"loss": 0.5315, "grad_norm": 1.1311043500900269, "learning_rate": 0.0002, "epoch": 6.426319618462533, "step": 39750}, {"loss": 0.5171, "grad_norm": 1.2352231740951538, "learning_rate": 0.0002, "epoch": 6.427936302643278, "step": 39760}, {"loss": 0.466, "grad_norm": 0.953233540058136, "learning_rate": 0.0002, "epoch": 6.4295529868240235, "step": 39770}, {"loss": 0.4834, "grad_norm": 1.0632505416870117, "learning_rate": 0.0002, "epoch": 6.431169671004769, "step": 39780}, {"loss": 0.5053, "grad_norm": 1.0916751623153687, "learning_rate": 0.0002, "epoch": 6.432786355185515, "step": 39790}, {"loss": 0.4788, "grad_norm": 0.9732703566551208, "learning_rate": 0.0002, "epoch": 6.43440303936626, "step": 39800}, {"loss": 0.4982, "grad_norm": 1.1673705577850342, "learning_rate": 0.0002, "epoch": 6.436019723547005, "step": 39810}, {"loss": 0.4484, "grad_norm": 1.1049559116363525, "learning_rate": 0.0002, "epoch": 6.437636407727751, "step": 39820}, {"loss": 0.4784, "grad_norm": 1.345277190208435, "learning_rate": 0.0002, "epoch": 6.439253091908496, "step": 39830}, {"loss": 0.4716, "grad_norm": 1.1118950843811035, "learning_rate": 0.0002, "epoch": 6.440869776089241, "step": 39840}, {"loss": 0.5133, "grad_norm": 1.4872850179672241, "learning_rate": 0.0002, "epoch": 6.442486460269986, "step": 39850}, {"loss": 0.4532, "grad_norm": 1.0763497352600098, "learning_rate": 0.0002, "epoch": 6.4441031444507315, "step": 39860}, {"loss": 0.4572, "grad_norm": 0.9245555400848389, "learning_rate": 0.0002, "epoch": 6.445719828631477, "step": 39870}, {"loss": 0.4917, "grad_norm": 1.4154807329177856, "learning_rate": 0.0002, "epoch": 6.447336512812222, "step": 39880}, {"loss": 0.4852, "grad_norm": 1.0885124206542969, "learning_rate": 0.0002, "epoch": 6.448953196992967, "step": 39890}, {"loss": 0.5399, "grad_norm": 1.3989344835281372, "learning_rate": 0.0002, "epoch": 6.450569881173712, "step": 39900}, {"loss": 0.509, "grad_norm": 0.9763124585151672, "learning_rate": 0.0002, "epoch": 6.452186565354458, "step": 39910}, {"loss": 0.5134, "grad_norm": 1.135272741317749, "learning_rate": 0.0002, "epoch": 6.453803249535203, "step": 39920}, {"loss": 0.4941, "grad_norm": 1.1140081882476807, "learning_rate": 0.0002, "epoch": 6.455419933715948, "step": 39930}, {"loss": 0.5137, "grad_norm": 1.0992448329925537, "learning_rate": 0.0002, "epoch": 6.457036617896694, "step": 39940}, {"loss": 0.4914, "grad_norm": 1.1658501625061035, "learning_rate": 0.0002, "epoch": 6.4586533020774395, "step": 39950}, {"loss": 0.5036, "grad_norm": 1.1122797727584839, "learning_rate": 0.0002, "epoch": 6.460269986258185, "step": 39960}, {"loss": 0.5159, "grad_norm": 0.9664968252182007, "learning_rate": 0.0002, "epoch": 6.46188667043893, "step": 39970}, {"loss": 0.4989, "grad_norm": 1.2513965368270874, "learning_rate": 0.0002, "epoch": 6.463503354619675, "step": 39980}, {"loss": 0.4694, "grad_norm": 1.1198630332946777, "learning_rate": 0.0002, "epoch": 6.46512003880042, "step": 39990}, {"loss": 0.5023, "grad_norm": 0.8783249855041504, "learning_rate": 0.0002, "epoch": 6.466736722981166, "step": 40000}, {"loss": 0.4648, "grad_norm": 1.1313109397888184, "learning_rate": 0.0002, "epoch": 6.468353407161911, "step": 40010}, {"loss": 0.4965, "grad_norm": 1.0854487419128418, "learning_rate": 0.0002, "epoch": 6.469970091342656, "step": 40020}, {"loss": 0.5253, "grad_norm": 1.1738566160202026, "learning_rate": 0.0002, "epoch": 6.471586775523401, "step": 40030}, {"loss": 0.4947, "grad_norm": 0.9720084071159363, "learning_rate": 0.0002, "epoch": 6.473203459704147, "step": 40040}, {"loss": 0.5218, "grad_norm": 1.105618953704834, "learning_rate": 0.0002, "epoch": 6.474820143884892, "step": 40050}, {"loss": 0.4943, "grad_norm": 1.2007657289505005, "learning_rate": 0.0002, "epoch": 6.476436828065637, "step": 40060}, {"loss": 0.4882, "grad_norm": 1.088402509689331, "learning_rate": 0.0002, "epoch": 6.478053512246382, "step": 40070}, {"loss": 0.504, "grad_norm": 1.0775291919708252, "learning_rate": 0.0002, "epoch": 6.4796701964271275, "step": 40080}, {"loss": 0.4791, "grad_norm": 1.1018189191818237, "learning_rate": 0.0002, "epoch": 6.481286880607874, "step": 40090}, {"loss": 0.488, "grad_norm": 1.1676557064056396, "learning_rate": 0.0002, "epoch": 6.482903564788619, "step": 40100}, {"loss": 0.4818, "grad_norm": 0.9619805812835693, "learning_rate": 0.0002, "epoch": 6.484520248969364, "step": 40110}, {"loss": 0.4986, "grad_norm": 1.2408208847045898, "learning_rate": 0.0002, "epoch": 6.486136933150109, "step": 40120}, {"loss": 0.4668, "grad_norm": 1.3488136529922485, "learning_rate": 0.0002, "epoch": 6.4877536173308545, "step": 40130}, {"loss": 0.4774, "grad_norm": 0.9864488244056702, "learning_rate": 0.0002, "epoch": 6.4893703015116, "step": 40140}, {"loss": 0.4651, "grad_norm": 0.9437947273254395, "learning_rate": 0.0002, "epoch": 6.490986985692345, "step": 40150}, {"loss": 0.542, "grad_norm": 1.2005455493927002, "learning_rate": 0.0002, "epoch": 6.49260366987309, "step": 40160}, {"loss": 0.4704, "grad_norm": 1.0796732902526855, "learning_rate": 0.0002, "epoch": 6.4942203540538355, "step": 40170}, {"loss": 0.498, "grad_norm": 1.1347825527191162, "learning_rate": 0.0002, "epoch": 6.495837038234581, "step": 40180}, {"loss": 0.5215, "grad_norm": 1.2311455011367798, "learning_rate": 0.0002, "epoch": 6.497453722415326, "step": 40190}, {"loss": 0.5043, "grad_norm": 1.068609356880188, "learning_rate": 0.0002, "epoch": 6.499070406596071, "step": 40200}, {"loss": 0.4868, "grad_norm": 1.196425437927246, "learning_rate": 0.0002, "epoch": 6.500687090776816, "step": 40210}, {"loss": 0.4881, "grad_norm": 1.183927297592163, "learning_rate": 0.0002, "epoch": 6.5023037749575625, "step": 40220}, {"loss": 0.4958, "grad_norm": 0.9099724292755127, "learning_rate": 0.0002, "epoch": 6.503920459138307, "step": 40230}, {"loss": 0.4816, "grad_norm": 0.9261038899421692, "learning_rate": 0.0002, "epoch": 6.505537143319053, "step": 40240}, {"loss": 0.5151, "grad_norm": 1.185491681098938, "learning_rate": 0.0002, "epoch": 6.507153827499798, "step": 40250}, {"loss": 0.4853, "grad_norm": 1.1866052150726318, "learning_rate": 0.0002, "epoch": 6.508770511680543, "step": 40260}, {"loss": 0.491, "grad_norm": 1.1600912809371948, "learning_rate": 0.0002, "epoch": 6.510387195861289, "step": 40270}, {"loss": 0.5181, "grad_norm": 0.9609426259994507, "learning_rate": 0.0002, "epoch": 6.512003880042034, "step": 40280}, {"loss": 0.4794, "grad_norm": 1.078864336013794, "learning_rate": 0.0002, "epoch": 6.513620564222779, "step": 40290}, {"loss": 0.46, "grad_norm": 1.042761206626892, "learning_rate": 0.0002, "epoch": 6.515237248403524, "step": 40300}, {"loss": 0.5341, "grad_norm": 0.9742481112480164, "learning_rate": 0.0002, "epoch": 6.51685393258427, "step": 40310}, {"loss": 0.5234, "grad_norm": 1.2544835805892944, "learning_rate": 0.0002, "epoch": 6.518470616765015, "step": 40320}, {"loss": 0.4815, "grad_norm": 1.3019760847091675, "learning_rate": 0.0002, "epoch": 6.52008730094576, "step": 40330}, {"loss": 0.5039, "grad_norm": 1.3196964263916016, "learning_rate": 0.0002, "epoch": 6.521703985126505, "step": 40340}, {"loss": 0.4979, "grad_norm": 1.2795668840408325, "learning_rate": 0.0002, "epoch": 6.5233206693072505, "step": 40350}, {"loss": 0.5075, "grad_norm": 1.1618940830230713, "learning_rate": 0.0002, "epoch": 6.524937353487996, "step": 40360}, {"loss": 0.5081, "grad_norm": 1.330543041229248, "learning_rate": 0.0002, "epoch": 6.526554037668742, "step": 40370}, {"loss": 0.5055, "grad_norm": 1.1946901082992554, "learning_rate": 0.0002, "epoch": 6.528170721849486, "step": 40380}, {"loss": 0.4518, "grad_norm": 1.1708201169967651, "learning_rate": 0.0002, "epoch": 6.529787406030232, "step": 40390}, {"loss": 0.4556, "grad_norm": 0.894036591053009, "learning_rate": 0.0002, "epoch": 6.531404090210978, "step": 40400}, {"loss": 0.4919, "grad_norm": 1.1199041604995728, "learning_rate": 0.0002, "epoch": 6.533020774391723, "step": 40410}, {"loss": 0.471, "grad_norm": 1.180317759513855, "learning_rate": 0.0002, "epoch": 6.534637458572468, "step": 40420}, {"loss": 0.4914, "grad_norm": 1.37367582321167, "learning_rate": 0.0002, "epoch": 6.536254142753213, "step": 40430}, {"loss": 0.4561, "grad_norm": 1.134791612625122, "learning_rate": 0.0002, "epoch": 6.5378708269339585, "step": 40440}, {"loss": 0.5337, "grad_norm": 1.1160204410552979, "learning_rate": 0.0002, "epoch": 6.539487511114704, "step": 40450}, {"loss": 0.5299, "grad_norm": 1.268347978591919, "learning_rate": 0.0002, "epoch": 6.541104195295449, "step": 40460}, {"loss": 0.5167, "grad_norm": 1.1424330472946167, "learning_rate": 0.0002, "epoch": 6.542720879476194, "step": 40470}, {"loss": 0.5114, "grad_norm": 1.3098465204238892, "learning_rate": 0.0002, "epoch": 6.544337563656939, "step": 40480}, {"loss": 0.4865, "grad_norm": 1.3439544439315796, "learning_rate": 0.0002, "epoch": 6.545954247837685, "step": 40490}, {"loss": 0.5183, "grad_norm": 1.2708452939987183, "learning_rate": 0.0002, "epoch": 6.54757093201843, "step": 40500}, {"loss": 0.5099, "grad_norm": 1.483680248260498, "learning_rate": 0.0002, "epoch": 6.549187616199175, "step": 40510}, {"loss": 0.4811, "grad_norm": 1.1697806119918823, "learning_rate": 0.0002, "epoch": 6.550804300379921, "step": 40520}, {"loss": 0.4814, "grad_norm": 1.1665642261505127, "learning_rate": 0.0002, "epoch": 6.5524209845606665, "step": 40530}, {"loss": 0.4985, "grad_norm": 1.1243325471878052, "learning_rate": 0.0002, "epoch": 6.554037668741412, "step": 40540}, {"loss": 0.4936, "grad_norm": 1.0277988910675049, "learning_rate": 0.0002, "epoch": 6.555654352922157, "step": 40550}, {"loss": 0.487, "grad_norm": 1.1466810703277588, "learning_rate": 0.0002, "epoch": 6.557271037102902, "step": 40560}, {"loss": 0.4851, "grad_norm": 1.1415363550186157, "learning_rate": 0.0002, "epoch": 6.558887721283647, "step": 40570}, {"loss": 0.4631, "grad_norm": 1.1923491954803467, "learning_rate": 0.0002, "epoch": 6.560504405464393, "step": 40580}, {"loss": 0.5071, "grad_norm": 0.9264549612998962, "learning_rate": 0.0002, "epoch": 6.562121089645138, "step": 40590}, {"loss": 0.466, "grad_norm": 0.8810341954231262, "learning_rate": 0.0002, "epoch": 6.563737773825883, "step": 40600}, {"loss": 0.5085, "grad_norm": 2.3296701908111572, "learning_rate": 0.0002, "epoch": 6.565354458006628, "step": 40610}, {"loss": 0.5196, "grad_norm": 1.0865163803100586, "learning_rate": 0.0002, "epoch": 6.5669711421873735, "step": 40620}, {"loss": 0.5132, "grad_norm": 0.9844607710838318, "learning_rate": 0.0002, "epoch": 6.568587826368119, "step": 40630}, {"loss": 0.5437, "grad_norm": 1.1686855554580688, "learning_rate": 0.0002, "epoch": 6.570204510548864, "step": 40640}, {"loss": 0.5293, "grad_norm": 1.016829252243042, "learning_rate": 0.0002, "epoch": 6.571821194729609, "step": 40650}, {"loss": 0.5243, "grad_norm": 1.2789337635040283, "learning_rate": 0.0002, "epoch": 6.5734378789103545, "step": 40660}, {"loss": 0.4867, "grad_norm": 1.0819072723388672, "learning_rate": 0.0002, "epoch": 6.575054563091101, "step": 40670}, {"loss": 0.5024, "grad_norm": 1.1478345394134521, "learning_rate": 0.0002, "epoch": 6.576671247271846, "step": 40680}, {"loss": 0.5282, "grad_norm": 0.7972208857536316, "learning_rate": 0.0002, "epoch": 6.578287931452591, "step": 40690}, {"loss": 0.4877, "grad_norm": 1.1481789350509644, "learning_rate": 0.0002, "epoch": 6.579904615633336, "step": 40700}, {"loss": 0.5143, "grad_norm": 1.0921871662139893, "learning_rate": 0.0002, "epoch": 6.5815212998140815, "step": 40710}, {"loss": 0.5441, "grad_norm": 1.0230315923690796, "learning_rate": 0.0002, "epoch": 6.583137983994827, "step": 40720}, {"loss": 0.4734, "grad_norm": 1.151049017906189, "learning_rate": 0.0002, "epoch": 6.584754668175572, "step": 40730}, {"loss": 0.4782, "grad_norm": 1.4016883373260498, "learning_rate": 0.0002, "epoch": 6.586371352356317, "step": 40740}, {"loss": 0.5195, "grad_norm": 1.2211825847625732, "learning_rate": 0.0002, "epoch": 6.587988036537062, "step": 40750}, {"loss": 0.4815, "grad_norm": 1.2803404331207275, "learning_rate": 0.0002, "epoch": 6.589604720717808, "step": 40760}, {"loss": 0.5329, "grad_norm": 1.1119942665100098, "learning_rate": 0.0002, "epoch": 6.591221404898553, "step": 40770}, {"loss": 0.5135, "grad_norm": 1.464650273323059, "learning_rate": 0.0002, "epoch": 6.592838089079298, "step": 40780}, {"loss": 0.5181, "grad_norm": 1.1751397848129272, "learning_rate": 0.0002, "epoch": 6.594454773260043, "step": 40790}, {"loss": 0.4772, "grad_norm": 1.0866316556930542, "learning_rate": 0.0002, "epoch": 6.596071457440789, "step": 40800}, {"loss": 0.5132, "grad_norm": 1.1733694076538086, "learning_rate": 0.0002, "epoch": 6.597688141621534, "step": 40810}, {"loss": 0.5138, "grad_norm": 1.184708833694458, "learning_rate": 0.0002, "epoch": 6.59930482580228, "step": 40820}, {"loss": 0.4885, "grad_norm": 1.406081199645996, "learning_rate": 0.0002, "epoch": 6.600921509983025, "step": 40830}, {"loss": 0.499, "grad_norm": 0.9658212661743164, "learning_rate": 0.0002, "epoch": 6.60253819416377, "step": 40840}, {"loss": 0.5113, "grad_norm": 1.1457678079605103, "learning_rate": 0.0002, "epoch": 6.604154878344516, "step": 40850}, {"loss": 0.4916, "grad_norm": 1.0487784147262573, "learning_rate": 0.0002, "epoch": 6.605771562525261, "step": 40860}, {"loss": 0.4682, "grad_norm": 0.9357177019119263, "learning_rate": 0.0002, "epoch": 6.607388246706006, "step": 40870}, {"loss": 0.4751, "grad_norm": 1.1479727029800415, "learning_rate": 0.0002, "epoch": 6.609004930886751, "step": 40880}, {"loss": 0.5493, "grad_norm": 1.3729329109191895, "learning_rate": 0.0002, "epoch": 6.610621615067497, "step": 40890}, {"loss": 0.4886, "grad_norm": 1.0085599422454834, "learning_rate": 0.0002, "epoch": 6.612238299248242, "step": 40900}, {"loss": 0.516, "grad_norm": 1.2750911712646484, "learning_rate": 0.0002, "epoch": 6.613854983428987, "step": 40910}, {"loss": 0.5342, "grad_norm": 1.1929547786712646, "learning_rate": 0.0002, "epoch": 6.615471667609732, "step": 40920}, {"loss": 0.4919, "grad_norm": 1.0821375846862793, "learning_rate": 0.0002, "epoch": 6.6170883517904775, "step": 40930}, {"loss": 0.5057, "grad_norm": 1.197347640991211, "learning_rate": 0.0002, "epoch": 6.618705035971223, "step": 40940}, {"loss": 0.492, "grad_norm": 1.2074699401855469, "learning_rate": 0.0002, "epoch": 6.620321720151968, "step": 40950}, {"loss": 0.5089, "grad_norm": 1.312009572982788, "learning_rate": 0.0002, "epoch": 6.621938404332713, "step": 40960}, {"loss": 0.5476, "grad_norm": 1.4381471872329712, "learning_rate": 0.0002, "epoch": 6.623555088513459, "step": 40970}, {"loss": 0.4904, "grad_norm": 1.1574671268463135, "learning_rate": 0.0002, "epoch": 6.6251717726942045, "step": 40980}, {"loss": 0.531, "grad_norm": 0.885661780834198, "learning_rate": 0.0002, "epoch": 6.62678845687495, "step": 40990}, {"loss": 0.5145, "grad_norm": 1.024571180343628, "learning_rate": 0.0002, "epoch": 6.628405141055695, "step": 41000}, {"loss": 0.4791, "grad_norm": 1.103437900543213, "learning_rate": 0.0002, "epoch": 6.63002182523644, "step": 41010}, {"loss": 0.4671, "grad_norm": 1.122450828552246, "learning_rate": 0.0002, "epoch": 6.6316385094171855, "step": 41020}, {"loss": 0.5134, "grad_norm": 1.2256295680999756, "learning_rate": 0.0002, "epoch": 6.633255193597931, "step": 41030}, {"loss": 0.4908, "grad_norm": 1.364594578742981, "learning_rate": 0.0002, "epoch": 6.634871877778676, "step": 41040}, {"loss": 0.4964, "grad_norm": 0.9550056457519531, "learning_rate": 0.0002, "epoch": 6.636488561959421, "step": 41050}, {"loss": 0.5028, "grad_norm": 1.3174707889556885, "learning_rate": 0.0002, "epoch": 6.638105246140166, "step": 41060}, {"loss": 0.4717, "grad_norm": 1.0835540294647217, "learning_rate": 0.0002, "epoch": 6.639721930320912, "step": 41070}, {"loss": 0.497, "grad_norm": 1.1432770490646362, "learning_rate": 0.0002, "epoch": 6.641338614501657, "step": 41080}, {"loss": 0.4903, "grad_norm": 1.2398556470870972, "learning_rate": 0.0002, "epoch": 6.642955298682402, "step": 41090}, {"loss": 0.4991, "grad_norm": 1.1147747039794922, "learning_rate": 0.0002, "epoch": 6.644571982863147, "step": 41100}, {"loss": 0.505, "grad_norm": 1.0730493068695068, "learning_rate": 0.0002, "epoch": 6.6461886670438926, "step": 41110}, {"loss": 0.486, "grad_norm": 1.3218451738357544, "learning_rate": 0.0002, "epoch": 6.647805351224639, "step": 41120}, {"loss": 0.5276, "grad_norm": 1.3027331829071045, "learning_rate": 0.0002, "epoch": 6.649422035405384, "step": 41130}, {"loss": 0.5263, "grad_norm": 1.0280735492706299, "learning_rate": 0.0002, "epoch": 6.651038719586129, "step": 41140}, {"loss": 0.4952, "grad_norm": 1.109916090965271, "learning_rate": 0.0002, "epoch": 6.652655403766874, "step": 41150}, {"loss": 0.5001, "grad_norm": 1.078734040260315, "learning_rate": 0.0002, "epoch": 6.65427208794762, "step": 41160}, {"loss": 0.484, "grad_norm": 1.1595654487609863, "learning_rate": 0.0002, "epoch": 6.655888772128365, "step": 41170}, {"loss": 0.5101, "grad_norm": 1.1701031923294067, "learning_rate": 0.0002, "epoch": 6.65750545630911, "step": 41180}, {"loss": 0.5341, "grad_norm": 1.0424643754959106, "learning_rate": 0.0002, "epoch": 6.659122140489855, "step": 41190}, {"loss": 0.4863, "grad_norm": 1.22880220413208, "learning_rate": 0.0002, "epoch": 6.6607388246706005, "step": 41200}, {"loss": 0.4987, "grad_norm": 1.1907655000686646, "learning_rate": 0.0002, "epoch": 6.662355508851346, "step": 41210}, {"loss": 0.5343, "grad_norm": 1.0765007734298706, "learning_rate": 0.0002, "epoch": 6.663972193032091, "step": 41220}, {"loss": 0.5039, "grad_norm": 0.9994917511940002, "learning_rate": 0.0002, "epoch": 6.665588877212836, "step": 41230}, {"loss": 0.507, "grad_norm": 0.968578040599823, "learning_rate": 0.0002, "epoch": 6.6672055613935814, "step": 41240}, {"loss": 0.5068, "grad_norm": 1.0576032400131226, "learning_rate": 0.0002, "epoch": 6.668822245574327, "step": 41250}, {"loss": 0.486, "grad_norm": 1.2183765172958374, "learning_rate": 0.0002, "epoch": 6.670438929755072, "step": 41260}, {"loss": 0.4764, "grad_norm": 1.2548623085021973, "learning_rate": 0.0002, "epoch": 6.672055613935818, "step": 41270}, {"loss": 0.5014, "grad_norm": 1.0848388671875, "learning_rate": 0.0002, "epoch": 6.673672298116563, "step": 41280}, {"loss": 0.5404, "grad_norm": 1.21421217918396, "learning_rate": 0.0002, "epoch": 6.6752889822973085, "step": 41290}, {"loss": 0.4911, "grad_norm": 1.1453598737716675, "learning_rate": 0.0002, "epoch": 6.676905666478054, "step": 41300}, {"loss": 0.5033, "grad_norm": 1.2682722806930542, "learning_rate": 0.0002, "epoch": 6.678522350658799, "step": 41310}, {"loss": 0.5313, "grad_norm": 1.1659725904464722, "learning_rate": 0.0002, "epoch": 6.680139034839544, "step": 41320}, {"loss": 0.5505, "grad_norm": 1.36194908618927, "learning_rate": 0.0002, "epoch": 6.681755719020289, "step": 41330}, {"loss": 0.5127, "grad_norm": 1.1712592840194702, "learning_rate": 0.0002, "epoch": 6.683372403201035, "step": 41340}, {"loss": 0.5082, "grad_norm": 1.4168336391448975, "learning_rate": 0.0002, "epoch": 6.68498908738178, "step": 41350}, {"loss": 0.5124, "grad_norm": 1.0395328998565674, "learning_rate": 0.0002, "epoch": 6.686605771562525, "step": 41360}, {"loss": 0.5404, "grad_norm": 1.2511054277420044, "learning_rate": 0.0002, "epoch": 6.68822245574327, "step": 41370}, {"loss": 0.5027, "grad_norm": 1.0438542366027832, "learning_rate": 0.0002, "epoch": 6.689839139924016, "step": 41380}, {"loss": 0.5069, "grad_norm": 1.08684241771698, "learning_rate": 0.0002, "epoch": 6.691455824104761, "step": 41390}, {"loss": 0.5224, "grad_norm": 1.250788927078247, "learning_rate": 0.0002, "epoch": 6.693072508285506, "step": 41400}, {"loss": 0.4921, "grad_norm": 1.313890814781189, "learning_rate": 0.0002, "epoch": 6.694689192466251, "step": 41410}, {"loss": 0.5028, "grad_norm": 1.3218982219696045, "learning_rate": 0.0002, "epoch": 6.696305876646997, "step": 41420}, {"loss": 0.4851, "grad_norm": 1.0366582870483398, "learning_rate": 0.0002, "epoch": 6.697922560827743, "step": 41430}, {"loss": 0.5103, "grad_norm": 1.066121220588684, "learning_rate": 0.0002, "epoch": 6.699539245008488, "step": 41440}, {"loss": 0.4966, "grad_norm": 1.0239925384521484, "learning_rate": 0.0002, "epoch": 6.701155929189233, "step": 41450}, {"loss": 0.4767, "grad_norm": 0.9402176141738892, "learning_rate": 0.0002, "epoch": 6.702772613369978, "step": 41460}, {"loss": 0.5381, "grad_norm": 1.391718864440918, "learning_rate": 0.0002, "epoch": 6.7043892975507235, "step": 41470}, {"loss": 0.512, "grad_norm": 1.215600609779358, "learning_rate": 0.0002, "epoch": 6.706005981731469, "step": 41480}, {"loss": 0.5219, "grad_norm": 1.063722848892212, "learning_rate": 0.0002, "epoch": 6.707622665912214, "step": 41490}, {"loss": 0.492, "grad_norm": 1.132149577140808, "learning_rate": 0.0002, "epoch": 6.709239350092959, "step": 41500}, {"loss": 0.4812, "grad_norm": 1.0302950143814087, "learning_rate": 0.0002, "epoch": 6.7108560342737045, "step": 41510}, {"loss": 0.5141, "grad_norm": 1.5342752933502197, "learning_rate": 0.0002, "epoch": 6.71247271845445, "step": 41520}, {"loss": 0.5123, "grad_norm": 1.177137017250061, "learning_rate": 0.0002, "epoch": 6.714089402635195, "step": 41530}, {"loss": 0.5082, "grad_norm": 1.2335538864135742, "learning_rate": 0.0002, "epoch": 6.71570608681594, "step": 41540}, {"loss": 0.4864, "grad_norm": 1.140604853630066, "learning_rate": 0.0002, "epoch": 6.717322770996686, "step": 41550}, {"loss": 0.4888, "grad_norm": 1.3567465543746948, "learning_rate": 0.0002, "epoch": 6.718939455177431, "step": 41560}, {"loss": 0.5183, "grad_norm": 1.0693929195404053, "learning_rate": 0.0002, "epoch": 6.720556139358177, "step": 41570}, {"loss": 0.5131, "grad_norm": 1.1592605113983154, "learning_rate": 0.0002, "epoch": 6.722172823538922, "step": 41580}, {"loss": 0.5476, "grad_norm": 0.989006519317627, "learning_rate": 0.0002, "epoch": 6.723789507719667, "step": 41590}, {"loss": 0.4952, "grad_norm": 1.04103422164917, "learning_rate": 0.0002, "epoch": 6.7254061919004124, "step": 41600}, {"loss": 0.4823, "grad_norm": 1.1129004955291748, "learning_rate": 0.0002, "epoch": 6.727022876081158, "step": 41610}, {"loss": 0.5032, "grad_norm": 1.1473113298416138, "learning_rate": 0.0002, "epoch": 6.728639560261903, "step": 41620}, {"loss": 0.5253, "grad_norm": 1.348036527633667, "learning_rate": 0.0002, "epoch": 6.730256244442648, "step": 41630}, {"loss": 0.4983, "grad_norm": 1.259942650794983, "learning_rate": 0.0002, "epoch": 6.731872928623393, "step": 41640}, {"loss": 0.5182, "grad_norm": 1.0591514110565186, "learning_rate": 0.0002, "epoch": 6.733489612804139, "step": 41650}, {"loss": 0.4886, "grad_norm": 0.9737129211425781, "learning_rate": 0.0002, "epoch": 6.735106296984884, "step": 41660}, {"loss": 0.5051, "grad_norm": 1.2520451545715332, "learning_rate": 0.0002, "epoch": 6.736722981165629, "step": 41670}, {"loss": 0.5364, "grad_norm": 1.0555530786514282, "learning_rate": 0.0002, "epoch": 6.738339665346374, "step": 41680}, {"loss": 0.4954, "grad_norm": 1.0025697946548462, "learning_rate": 0.0002, "epoch": 6.7399563495271195, "step": 41690}, {"loss": 0.5485, "grad_norm": 1.1114100217819214, "learning_rate": 0.0002, "epoch": 6.741573033707866, "step": 41700}, {"loss": 0.4986, "grad_norm": 1.1537504196166992, "learning_rate": 0.0002, "epoch": 6.74318971788861, "step": 41710}, {"loss": 0.5025, "grad_norm": 1.037880539894104, "learning_rate": 0.0002, "epoch": 6.744806402069356, "step": 41720}, {"loss": 0.482, "grad_norm": 1.0691965818405151, "learning_rate": 0.0002, "epoch": 6.746423086250101, "step": 41730}, {"loss": 0.5272, "grad_norm": 1.376325011253357, "learning_rate": 0.0002, "epoch": 6.748039770430847, "step": 41740}, {"loss": 0.5484, "grad_norm": 1.4667129516601562, "learning_rate": 0.0002, "epoch": 6.749656454611592, "step": 41750}, {"loss": 0.5139, "grad_norm": 1.1517162322998047, "learning_rate": 0.0002, "epoch": 6.751273138792337, "step": 41760}, {"loss": 0.5523, "grad_norm": 1.1454511880874634, "learning_rate": 0.0002, "epoch": 6.752889822973082, "step": 41770}, {"loss": 0.4664, "grad_norm": 1.6323128938674927, "learning_rate": 0.0002, "epoch": 6.7545065071538275, "step": 41780}, {"loss": 0.5153, "grad_norm": 1.0951642990112305, "learning_rate": 0.0002, "epoch": 6.756123191334573, "step": 41790}, {"loss": 0.4998, "grad_norm": 1.0766983032226562, "learning_rate": 0.0002, "epoch": 6.757739875515318, "step": 41800}, {"loss": 0.548, "grad_norm": 1.3472381830215454, "learning_rate": 0.0002, "epoch": 6.759356559696063, "step": 41810}, {"loss": 0.5172, "grad_norm": 1.0248444080352783, "learning_rate": 0.0002, "epoch": 6.760973243876808, "step": 41820}, {"loss": 0.5236, "grad_norm": 1.1276055574417114, "learning_rate": 0.0002, "epoch": 6.762589928057554, "step": 41830}, {"loss": 0.5044, "grad_norm": 1.5398495197296143, "learning_rate": 0.0002, "epoch": 6.764206612238299, "step": 41840}, {"loss": 0.5097, "grad_norm": 1.1886497735977173, "learning_rate": 0.0002, "epoch": 6.765823296419045, "step": 41850}, {"loss": 0.499, "grad_norm": 1.027198076248169, "learning_rate": 0.0002, "epoch": 6.767439980599789, "step": 41860}, {"loss": 0.5444, "grad_norm": 1.4644980430603027, "learning_rate": 0.0002, "epoch": 6.7690566647805355, "step": 41870}, {"loss": 0.5009, "grad_norm": 0.9633586406707764, "learning_rate": 0.0002, "epoch": 6.770673348961281, "step": 41880}, {"loss": 0.484, "grad_norm": 1.0895354747772217, "learning_rate": 0.0002, "epoch": 6.772290033142026, "step": 41890}, {"loss": 0.5172, "grad_norm": 1.1887167692184448, "learning_rate": 0.0002, "epoch": 6.773906717322771, "step": 41900}, {"loss": 0.5399, "grad_norm": 1.3699820041656494, "learning_rate": 0.0002, "epoch": 6.775523401503516, "step": 41910}, {"loss": 0.5504, "grad_norm": 1.0266352891921997, "learning_rate": 0.0002, "epoch": 6.777140085684262, "step": 41920}, {"loss": 0.5105, "grad_norm": 1.0919075012207031, "learning_rate": 0.0002, "epoch": 6.778756769865007, "step": 41930}, {"loss": 0.4842, "grad_norm": 0.9839563369750977, "learning_rate": 0.0002, "epoch": 6.780373454045752, "step": 41940}, {"loss": 0.5081, "grad_norm": 1.2605451345443726, "learning_rate": 0.0002, "epoch": 6.781990138226497, "step": 41950}, {"loss": 0.5391, "grad_norm": 0.9268672466278076, "learning_rate": 0.0002, "epoch": 6.7836068224072426, "step": 41960}, {"loss": 0.4916, "grad_norm": 1.2002313137054443, "learning_rate": 0.0002, "epoch": 6.785223506587988, "step": 41970}, {"loss": 0.5467, "grad_norm": 1.2018438577651978, "learning_rate": 0.0002, "epoch": 6.786840190768733, "step": 41980}, {"loss": 0.5491, "grad_norm": 1.17646062374115, "learning_rate": 0.0002, "epoch": 6.788456874949478, "step": 41990}, {"loss": 0.5354, "grad_norm": 1.1080009937286377, "learning_rate": 0.0002, "epoch": 6.790073559130224, "step": 42000}, {"loss": 0.5384, "grad_norm": 1.1606498956680298, "learning_rate": 0.0002, "epoch": 6.791690243310969, "step": 42010}, {"loss": 0.4931, "grad_norm": 1.2484819889068604, "learning_rate": 0.0002, "epoch": 6.793306927491715, "step": 42020}, {"loss": 0.498, "grad_norm": 1.1363215446472168, "learning_rate": 0.0002, "epoch": 6.79492361167246, "step": 42030}, {"loss": 0.5343, "grad_norm": 1.4469727277755737, "learning_rate": 0.0002, "epoch": 6.796540295853205, "step": 42040}, {"loss": 0.5146, "grad_norm": 1.0617138147354126, "learning_rate": 0.0002, "epoch": 6.7981569800339505, "step": 42050}, {"loss": 0.5188, "grad_norm": 1.1459330320358276, "learning_rate": 0.0002, "epoch": 6.799773664214696, "step": 42060}, {"loss": 0.5116, "grad_norm": 1.2095019817352295, "learning_rate": 0.0002, "epoch": 6.801390348395441, "step": 42070}, {"loss": 0.545, "grad_norm": 1.3200831413269043, "learning_rate": 0.0002, "epoch": 6.803007032576186, "step": 42080}, {"loss": 0.5406, "grad_norm": 1.1633318662643433, "learning_rate": 0.0002, "epoch": 6.8046237167569315, "step": 42090}, {"loss": 0.4938, "grad_norm": 0.8986614942550659, "learning_rate": 0.0002, "epoch": 6.806240400937677, "step": 42100}, {"loss": 0.559, "grad_norm": 1.3705275058746338, "learning_rate": 0.0002, "epoch": 6.807857085118422, "step": 42110}, {"loss": 0.5022, "grad_norm": 1.2418090105056763, "learning_rate": 0.0002, "epoch": 6.809473769299167, "step": 42120}, {"loss": 0.5014, "grad_norm": 1.0818954706192017, "learning_rate": 0.0002, "epoch": 6.811090453479912, "step": 42130}, {"loss": 0.4791, "grad_norm": 0.9293872117996216, "learning_rate": 0.0002, "epoch": 6.812707137660658, "step": 42140}, {"loss": 0.5009, "grad_norm": 0.9791894555091858, "learning_rate": 0.0002, "epoch": 6.814323821841404, "step": 42150}, {"loss": 0.5142, "grad_norm": 1.1956568956375122, "learning_rate": 0.0002, "epoch": 6.815940506022149, "step": 42160}, {"loss": 0.5126, "grad_norm": 0.9643568992614746, "learning_rate": 0.0002, "epoch": 6.817557190202894, "step": 42170}, {"loss": 0.5121, "grad_norm": 1.2499792575836182, "learning_rate": 0.0002, "epoch": 6.819173874383639, "step": 42180}, {"loss": 0.4942, "grad_norm": 1.1779413223266602, "learning_rate": 0.0002, "epoch": 6.820790558564385, "step": 42190}, {"loss": 0.498, "grad_norm": 1.0570595264434814, "learning_rate": 0.0002, "epoch": 6.82240724274513, "step": 42200}, {"loss": 0.4997, "grad_norm": 1.1393938064575195, "learning_rate": 0.0002, "epoch": 6.824023926925875, "step": 42210}, {"loss": 0.4842, "grad_norm": 1.152463436126709, "learning_rate": 0.0002, "epoch": 6.82564061110662, "step": 42220}, {"loss": 0.5234, "grad_norm": 1.3353025913238525, "learning_rate": 0.0002, "epoch": 6.827257295287366, "step": 42230}, {"loss": 0.539, "grad_norm": 1.1719051599502563, "learning_rate": 0.0002, "epoch": 6.828873979468111, "step": 42240}, {"loss": 0.5139, "grad_norm": 1.262141227722168, "learning_rate": 0.0002, "epoch": 6.830490663648856, "step": 42250}, {"loss": 0.5021, "grad_norm": 1.240899920463562, "learning_rate": 0.0002, "epoch": 6.832107347829601, "step": 42260}, {"loss": 0.4961, "grad_norm": 1.0505269765853882, "learning_rate": 0.0002, "epoch": 6.8337240320103465, "step": 42270}, {"loss": 0.4932, "grad_norm": 1.1556071043014526, "learning_rate": 0.0002, "epoch": 6.835340716191092, "step": 42280}, {"loss": 0.5461, "grad_norm": 1.1427719593048096, "learning_rate": 0.0002, "epoch": 6.836957400371837, "step": 42290}, {"loss": 0.5199, "grad_norm": 1.1540080308914185, "learning_rate": 0.0002, "epoch": 6.838574084552583, "step": 42300}, {"loss": 0.5269, "grad_norm": 1.0521200895309448, "learning_rate": 0.0002, "epoch": 6.840190768733328, "step": 42310}, {"loss": 0.541, "grad_norm": 1.0205531120300293, "learning_rate": 0.0002, "epoch": 6.8418074529140736, "step": 42320}, {"loss": 0.5225, "grad_norm": 1.0010193586349487, "learning_rate": 0.0002, "epoch": 6.843424137094819, "step": 42330}, {"loss": 0.5101, "grad_norm": 1.2138770818710327, "learning_rate": 0.0002, "epoch": 6.845040821275564, "step": 42340}, {"loss": 0.5452, "grad_norm": 1.3028651475906372, "learning_rate": 0.0002, "epoch": 6.846657505456309, "step": 42350}, {"loss": 0.4894, "grad_norm": 1.0326353311538696, "learning_rate": 0.0002, "epoch": 6.8482741896370545, "step": 42360}, {"loss": 0.5285, "grad_norm": 1.036085605621338, "learning_rate": 0.0002, "epoch": 6.8498908738178, "step": 42370}, {"loss": 0.505, "grad_norm": 1.0575472116470337, "learning_rate": 0.0002, "epoch": 6.851507557998545, "step": 42380}, {"loss": 0.4997, "grad_norm": 1.1749629974365234, "learning_rate": 0.0002, "epoch": 6.85312424217929, "step": 42390}, {"loss": 0.4961, "grad_norm": 1.1747760772705078, "learning_rate": 0.0002, "epoch": 6.854740926360035, "step": 42400}, {"loss": 0.5138, "grad_norm": 1.1877071857452393, "learning_rate": 0.0002, "epoch": 6.856357610540781, "step": 42410}, {"loss": 0.4972, "grad_norm": 1.1209983825683594, "learning_rate": 0.0002, "epoch": 6.857974294721526, "step": 42420}, {"loss": 0.4939, "grad_norm": 1.2918205261230469, "learning_rate": 0.0002, "epoch": 6.859590978902271, "step": 42430}, {"loss": 0.5012, "grad_norm": 1.2443464994430542, "learning_rate": 0.0002, "epoch": 6.861207663083016, "step": 42440}, {"loss": 0.5226, "grad_norm": 0.9336795210838318, "learning_rate": 0.0002, "epoch": 6.8628243472637624, "step": 42450}, {"loss": 0.5108, "grad_norm": 1.2183542251586914, "learning_rate": 0.0002, "epoch": 6.864441031444508, "step": 42460}, {"loss": 0.5245, "grad_norm": 1.0071234703063965, "learning_rate": 0.0002, "epoch": 6.866057715625253, "step": 42470}, {"loss": 0.4753, "grad_norm": 1.2914012670516968, "learning_rate": 0.0002, "epoch": 6.867674399805998, "step": 42480}, {"loss": 0.4865, "grad_norm": 1.1050426959991455, "learning_rate": 0.0002, "epoch": 6.869291083986743, "step": 42490}, {"loss": 0.5243, "grad_norm": 1.1163811683654785, "learning_rate": 0.0002, "epoch": 6.870907768167489, "step": 42500}, {"loss": 0.5065, "grad_norm": 1.1575818061828613, "learning_rate": 0.0002, "epoch": 6.872524452348234, "step": 42510}, {"loss": 0.5353, "grad_norm": 1.11167311668396, "learning_rate": 0.0002, "epoch": 6.874141136528979, "step": 42520}, {"loss": 0.5141, "grad_norm": 1.0379102230072021, "learning_rate": 0.0002, "epoch": 6.875757820709724, "step": 42530}, {"loss": 0.5355, "grad_norm": 1.2617160081863403, "learning_rate": 0.0002, "epoch": 6.8773745048904695, "step": 42540}, {"loss": 0.4785, "grad_norm": 1.1749719381332397, "learning_rate": 0.0002, "epoch": 6.878991189071215, "step": 42550}, {"loss": 0.5503, "grad_norm": 1.2284821271896362, "learning_rate": 0.0002, "epoch": 6.88060787325196, "step": 42560}, {"loss": 0.5065, "grad_norm": 1.1917030811309814, "learning_rate": 0.0002, "epoch": 6.882224557432705, "step": 42570}, {"loss": 0.5176, "grad_norm": 1.1943914890289307, "learning_rate": 0.0002, "epoch": 6.8838412416134505, "step": 42580}, {"loss": 0.5072, "grad_norm": 1.2641394138336182, "learning_rate": 0.0002, "epoch": 6.885457925794196, "step": 42590}, {"loss": 0.5004, "grad_norm": 1.1280436515808105, "learning_rate": 0.0002, "epoch": 6.887074609974942, "step": 42600}, {"loss": 0.5328, "grad_norm": 0.9865449070930481, "learning_rate": 0.0002, "epoch": 6.888691294155687, "step": 42610}, {"loss": 0.4953, "grad_norm": 0.994987428188324, "learning_rate": 0.0002, "epoch": 6.890307978336432, "step": 42620}, {"loss": 0.4805, "grad_norm": 0.9900388717651367, "learning_rate": 0.0002, "epoch": 6.8919246625171775, "step": 42630}, {"loss": 0.5467, "grad_norm": 1.2992421388626099, "learning_rate": 0.0002, "epoch": 6.893541346697923, "step": 42640}, {"loss": 0.5017, "grad_norm": 1.0152487754821777, "learning_rate": 0.0002, "epoch": 6.895158030878668, "step": 42650}, {"loss": 0.5043, "grad_norm": 1.199453353881836, "learning_rate": 0.0002, "epoch": 6.896774715059413, "step": 42660}, {"loss": 0.5106, "grad_norm": 1.100630521774292, "learning_rate": 0.0002, "epoch": 6.898391399240158, "step": 42670}, {"loss": 0.503, "grad_norm": 1.0489764213562012, "learning_rate": 0.0002, "epoch": 6.900008083420904, "step": 42680}, {"loss": 0.4634, "grad_norm": 1.101407527923584, "learning_rate": 0.0002, "epoch": 6.901624767601649, "step": 42690}, {"loss": 0.5361, "grad_norm": 1.3130593299865723, "learning_rate": 0.0002, "epoch": 6.903241451782394, "step": 42700}, {"loss": 0.5119, "grad_norm": 0.9906072616577148, "learning_rate": 0.0002, "epoch": 6.904858135963139, "step": 42710}, {"loss": 0.5146, "grad_norm": 1.094502329826355, "learning_rate": 0.0002, "epoch": 6.906474820143885, "step": 42720}, {"loss": 0.5165, "grad_norm": 1.1025426387786865, "learning_rate": 0.0002, "epoch": 6.90809150432463, "step": 42730}, {"loss": 0.5463, "grad_norm": 1.0644042491912842, "learning_rate": 0.0002, "epoch": 6.909708188505375, "step": 42740}, {"loss": 0.5024, "grad_norm": 1.0709129571914673, "learning_rate": 0.0002, "epoch": 6.911324872686121, "step": 42750}, {"loss": 0.5093, "grad_norm": 1.2445871829986572, "learning_rate": 0.0002, "epoch": 6.912941556866866, "step": 42760}, {"loss": 0.5305, "grad_norm": 1.020058035850525, "learning_rate": 0.0002, "epoch": 6.914558241047612, "step": 42770}, {"loss": 0.5382, "grad_norm": 0.9795091152191162, "learning_rate": 0.0002, "epoch": 6.916174925228357, "step": 42780}, {"loss": 0.5429, "grad_norm": 0.9369977116584778, "learning_rate": 0.0002, "epoch": 6.917791609409102, "step": 42790}, {"loss": 0.5444, "grad_norm": 1.0741904973983765, "learning_rate": 0.0002, "epoch": 6.919408293589847, "step": 42800}, {"loss": 0.5402, "grad_norm": 1.0702799558639526, "learning_rate": 0.0002, "epoch": 6.921024977770593, "step": 42810}, {"loss": 0.5291, "grad_norm": 1.0383983850479126, "learning_rate": 0.0002, "epoch": 6.922641661951338, "step": 42820}, {"loss": 0.5106, "grad_norm": 1.0761083364486694, "learning_rate": 0.0002, "epoch": 6.924258346132083, "step": 42830}, {"loss": 0.5726, "grad_norm": 1.2332350015640259, "learning_rate": 0.0002, "epoch": 6.925875030312828, "step": 42840}, {"loss": 0.4996, "grad_norm": 1.3184348344802856, "learning_rate": 0.0002, "epoch": 6.9274917144935735, "step": 42850}, {"loss": 0.5503, "grad_norm": 1.0586378574371338, "learning_rate": 0.0002, "epoch": 6.929108398674319, "step": 42860}, {"loss": 0.511, "grad_norm": 1.2294201850891113, "learning_rate": 0.0002, "epoch": 6.930725082855064, "step": 42870}, {"loss": 0.54, "grad_norm": 1.3097991943359375, "learning_rate": 0.0002, "epoch": 6.932341767035809, "step": 42880}, {"loss": 0.5228, "grad_norm": 0.9006873965263367, "learning_rate": 0.0002, "epoch": 6.933958451216554, "step": 42890}, {"loss": 0.4617, "grad_norm": 1.265931248664856, "learning_rate": 0.0002, "epoch": 6.9355751353973005, "step": 42900}, {"loss": 0.5029, "grad_norm": 1.1013522148132324, "learning_rate": 0.0002, "epoch": 6.937191819578046, "step": 42910}, {"loss": 0.5334, "grad_norm": 0.9910131692886353, "learning_rate": 0.0002, "epoch": 6.938808503758791, "step": 42920}, {"loss": 0.5211, "grad_norm": 1.102683424949646, "learning_rate": 0.0002, "epoch": 6.940425187939536, "step": 42930}, {"loss": 0.5588, "grad_norm": 1.232961893081665, "learning_rate": 0.0002, "epoch": 6.9420418721202815, "step": 42940}, {"loss": 0.5357, "grad_norm": 1.1714650392532349, "learning_rate": 0.0002, "epoch": 6.943658556301027, "step": 42950}, {"loss": 0.5232, "grad_norm": 1.1684318780899048, "learning_rate": 0.0002, "epoch": 6.945275240481772, "step": 42960}, {"loss": 0.5035, "grad_norm": 1.2074716091156006, "learning_rate": 0.0002, "epoch": 6.946891924662517, "step": 42970}, {"loss": 0.5111, "grad_norm": 1.2061275243759155, "learning_rate": 0.0002, "epoch": 6.948508608843262, "step": 42980}, {"loss": 0.5066, "grad_norm": 1.1216989755630493, "learning_rate": 0.0002, "epoch": 6.950125293024008, "step": 42990}, {"loss": 0.4948, "grad_norm": 1.304117202758789, "learning_rate": 0.0002, "epoch": 6.951741977204753, "step": 43000}, {"loss": 0.5684, "grad_norm": 1.2377972602844238, "learning_rate": 0.0002, "epoch": 6.953358661385498, "step": 43010}, {"loss": 0.4792, "grad_norm": 1.2332178354263306, "learning_rate": 0.0002, "epoch": 6.954975345566243, "step": 43020}, {"loss": 0.5181, "grad_norm": 1.1919599771499634, "learning_rate": 0.0002, "epoch": 6.956592029746989, "step": 43030}, {"loss": 0.5352, "grad_norm": 1.272700548171997, "learning_rate": 0.0002, "epoch": 6.958208713927734, "step": 43040}, {"loss": 0.5328, "grad_norm": 1.4377546310424805, "learning_rate": 0.0002, "epoch": 6.95982539810848, "step": 43050}, {"loss": 0.4894, "grad_norm": 1.2070353031158447, "learning_rate": 0.0002, "epoch": 6.961442082289225, "step": 43060}, {"loss": 0.525, "grad_norm": 1.090205430984497, "learning_rate": 0.0002, "epoch": 6.96305876646997, "step": 43070}, {"loss": 0.5255, "grad_norm": 1.1832911968231201, "learning_rate": 0.0002, "epoch": 6.964675450650716, "step": 43080}, {"loss": 0.5497, "grad_norm": 1.2921082973480225, "learning_rate": 0.0002, "epoch": 6.966292134831461, "step": 43090}, {"loss": 0.5527, "grad_norm": 1.4303096532821655, "learning_rate": 0.0002, "epoch": 6.967908819012206, "step": 43100}, {"loss": 0.4807, "grad_norm": 1.0788004398345947, "learning_rate": 0.0002, "epoch": 6.969525503192951, "step": 43110}, {"loss": 0.5006, "grad_norm": 1.2192047834396362, "learning_rate": 0.0002, "epoch": 6.9711421873736965, "step": 43120}, {"loss": 0.4714, "grad_norm": 1.0735143423080444, "learning_rate": 0.0002, "epoch": 6.972758871554442, "step": 43130}, {"loss": 0.5307, "grad_norm": 1.0317153930664062, "learning_rate": 0.0002, "epoch": 6.974375555735187, "step": 43140}, {"loss": 0.5154, "grad_norm": 1.0926798582077026, "learning_rate": 0.0002, "epoch": 6.975992239915932, "step": 43150}, {"loss": 0.4976, "grad_norm": 1.1660500764846802, "learning_rate": 0.0002, "epoch": 6.977608924096677, "step": 43160}, {"loss": 0.5456, "grad_norm": 1.3945232629776, "learning_rate": 0.0002, "epoch": 6.979225608277423, "step": 43170}, {"loss": 0.4979, "grad_norm": 1.2684587240219116, "learning_rate": 0.0002, "epoch": 6.980842292458169, "step": 43180}, {"loss": 0.5406, "grad_norm": 1.1574004888534546, "learning_rate": 0.0002, "epoch": 6.982458976638913, "step": 43190}, {"loss": 0.5629, "grad_norm": 1.2534198760986328, "learning_rate": 0.0002, "epoch": 6.984075660819659, "step": 43200}, {"loss": 0.5191, "grad_norm": 1.135245442390442, "learning_rate": 0.0002, "epoch": 6.9856923450004045, "step": 43210}, {"loss": 0.548, "grad_norm": 1.3824104070663452, "learning_rate": 0.0002, "epoch": 6.98730902918115, "step": 43220}, {"loss": 0.5294, "grad_norm": 1.2128452062606812, "learning_rate": 0.0002, "epoch": 6.988925713361895, "step": 43230}, {"loss": 0.505, "grad_norm": 1.0795245170593262, "learning_rate": 0.0002, "epoch": 6.99054239754264, "step": 43240}, {"loss": 0.4889, "grad_norm": 1.337353229522705, "learning_rate": 0.0002, "epoch": 6.992159081723385, "step": 43250}, {"loss": 0.4749, "grad_norm": 1.1731765270233154, "learning_rate": 0.0002, "epoch": 6.993775765904131, "step": 43260}, {"loss": 0.4897, "grad_norm": 1.0203192234039307, "learning_rate": 0.0002, "epoch": 6.995392450084876, "step": 43270}, {"loss": 0.5324, "grad_norm": 0.9261201620101929, "learning_rate": 0.0002, "epoch": 6.997009134265621, "step": 43280}, {"loss": 0.5227, "grad_norm": 1.107865810394287, "learning_rate": 0.0002, "epoch": 6.998625818446366, "step": 43290}, {"eval_loss": 1.2679380178451538, "eval_runtime": 122.202, "eval_samples_per_second": 5.998, "eval_steps_per_second": 0.753, "epoch": 6.9999191657909625, "step": 43298}, {"loss": 0.4651, "grad_norm": 0.9555306434631348, "learning_rate": 0.0002, "epoch": 7.000242502627112, "step": 43300}, {"loss": 0.4301, "grad_norm": 1.3280415534973145, "learning_rate": 0.0002, "epoch": 7.001859186807857, "step": 43310}, {"loss": 0.437, "grad_norm": 1.5583289861679077, "learning_rate": 0.0002, "epoch": 7.003475870988602, "step": 43320}, {"loss": 0.4532, "grad_norm": 1.0714443922042847, "learning_rate": 0.0002, "epoch": 7.005092555169347, "step": 43330}, {"loss": 0.4048, "grad_norm": 1.048075795173645, "learning_rate": 0.0002, "epoch": 7.006709239350093, "step": 43340}, {"loss": 0.4119, "grad_norm": 1.1053836345672607, "learning_rate": 0.0002, "epoch": 7.008325923530839, "step": 43350}, {"loss": 0.4352, "grad_norm": 0.8911725282669067, "learning_rate": 0.0002, "epoch": 7.009942607711584, "step": 43360}, {"loss": 0.4236, "grad_norm": 0.9404396414756775, "learning_rate": 0.0002, "epoch": 7.011559291892329, "step": 43370}, {"loss": 0.4529, "grad_norm": 1.152365803718567, "learning_rate": 0.0002, "epoch": 7.013175976073074, "step": 43380}, {"loss": 0.4381, "grad_norm": 1.2118251323699951, "learning_rate": 0.0002, "epoch": 7.0147926602538195, "step": 43390}, {"loss": 0.4404, "grad_norm": 1.2046295404434204, "learning_rate": 0.0002, "epoch": 7.016409344434565, "step": 43400}, {"loss": 0.4177, "grad_norm": 0.929465115070343, "learning_rate": 0.0002, "epoch": 7.01802602861531, "step": 43410}, {"loss": 0.4086, "grad_norm": 1.3720149993896484, "learning_rate": 0.0002, "epoch": 7.019642712796055, "step": 43420}, {"loss": 0.4174, "grad_norm": 1.1316810846328735, "learning_rate": 0.0002, "epoch": 7.0212593969768005, "step": 43430}, {"loss": 0.451, "grad_norm": 1.0342087745666504, "learning_rate": 0.0002, "epoch": 7.022876081157546, "step": 43440}, {"loss": 0.4084, "grad_norm": 1.1455655097961426, "learning_rate": 0.0002, "epoch": 7.024492765338291, "step": 43450}, {"loss": 0.4168, "grad_norm": 1.1308859586715698, "learning_rate": 0.0002, "epoch": 7.026109449519036, "step": 43460}, {"loss": 0.4099, "grad_norm": 1.0796722173690796, "learning_rate": 0.0002, "epoch": 7.027726133699781, "step": 43470}, {"loss": 0.4047, "grad_norm": 1.0031877756118774, "learning_rate": 0.0002, "epoch": 7.029342817880527, "step": 43480}, {"loss": 0.3968, "grad_norm": 1.2391340732574463, "learning_rate": 0.0002, "epoch": 7.030959502061273, "step": 43490}, {"loss": 0.4155, "grad_norm": 1.0807358026504517, "learning_rate": 0.0002, "epoch": 7.032576186242018, "step": 43500}, {"loss": 0.4322, "grad_norm": 1.230995535850525, "learning_rate": 0.0002, "epoch": 7.034192870422763, "step": 43510}, {"loss": 0.3971, "grad_norm": 1.2200509309768677, "learning_rate": 0.0002, "epoch": 7.035809554603508, "step": 43520}, {"loss": 0.4242, "grad_norm": 0.9785236120223999, "learning_rate": 0.0002, "epoch": 7.037426238784254, "step": 43530}, {"loss": 0.4173, "grad_norm": 1.0009595155715942, "learning_rate": 0.0002, "epoch": 7.039042922964999, "step": 43540}, {"loss": 0.4175, "grad_norm": 0.9783103466033936, "learning_rate": 0.0002, "epoch": 7.040659607145744, "step": 43550}, {"loss": 0.4307, "grad_norm": 1.1303530931472778, "learning_rate": 0.0002, "epoch": 7.042276291326489, "step": 43560}, {"loss": 0.4066, "grad_norm": 1.1768499612808228, "learning_rate": 0.0002, "epoch": 7.043892975507235, "step": 43570}, {"loss": 0.4492, "grad_norm": 1.1040459871292114, "learning_rate": 0.0002, "epoch": 7.04550965968798, "step": 43580}, {"loss": 0.4314, "grad_norm": 1.0673959255218506, "learning_rate": 0.0002, "epoch": 7.047126343868725, "step": 43590}, {"loss": 0.402, "grad_norm": 1.1220765113830566, "learning_rate": 0.0002, "epoch": 7.04874302804947, "step": 43600}, {"loss": 0.4108, "grad_norm": 1.1746923923492432, "learning_rate": 0.0002, "epoch": 7.0503597122302155, "step": 43610}, {"loss": 0.4618, "grad_norm": 1.2764517068862915, "learning_rate": 0.0002, "epoch": 7.051976396410961, "step": 43620}, {"loss": 0.4243, "grad_norm": 1.1180157661437988, "learning_rate": 0.0002, "epoch": 7.053593080591706, "step": 43630}, {"loss": 0.4593, "grad_norm": 1.3558318614959717, "learning_rate": 0.0002, "epoch": 7.055209764772452, "step": 43640}, {"loss": 0.4351, "grad_norm": 0.9804982542991638, "learning_rate": 0.0002, "epoch": 7.056826448953197, "step": 43650}, {"loss": 0.4309, "grad_norm": 1.106404423713684, "learning_rate": 0.0002, "epoch": 7.058443133133943, "step": 43660}, {"loss": 0.4183, "grad_norm": 0.9469243884086609, "learning_rate": 0.0002, "epoch": 7.060059817314688, "step": 43670}, {"loss": 0.4335, "grad_norm": 1.272987723350525, "learning_rate": 0.0002, "epoch": 7.061676501495433, "step": 43680}, {"loss": 0.4017, "grad_norm": 1.0536233186721802, "learning_rate": 0.0002, "epoch": 7.063293185676178, "step": 43690}, {"loss": 0.4597, "grad_norm": 1.1730698347091675, "learning_rate": 0.0002, "epoch": 7.0649098698569235, "step": 43700}, {"loss": 0.4304, "grad_norm": 1.150707483291626, "learning_rate": 0.0002, "epoch": 7.066526554037669, "step": 43710}, {"loss": 0.4136, "grad_norm": 1.4583828449249268, "learning_rate": 0.0002, "epoch": 7.068143238218414, "step": 43720}, {"loss": 0.4385, "grad_norm": 1.569705843925476, "learning_rate": 0.0002, "epoch": 7.069759922399159, "step": 43730}, {"loss": 0.4051, "grad_norm": 1.156192660331726, "learning_rate": 0.0002, "epoch": 7.071376606579904, "step": 43740}, {"loss": 0.4375, "grad_norm": 1.25005304813385, "learning_rate": 0.0002, "epoch": 7.07299329076065, "step": 43750}, {"loss": 0.4096, "grad_norm": 1.0468846559524536, "learning_rate": 0.0002, "epoch": 7.074609974941395, "step": 43760}, {"loss": 0.4253, "grad_norm": 1.2045108079910278, "learning_rate": 0.0002, "epoch": 7.07622665912214, "step": 43770}, {"loss": 0.4248, "grad_norm": 1.1341021060943604, "learning_rate": 0.0002, "epoch": 7.077843343302886, "step": 43780}, {"loss": 0.394, "grad_norm": 1.0712201595306396, "learning_rate": 0.0002, "epoch": 7.0794600274836315, "step": 43790}, {"loss": 0.4093, "grad_norm": 1.0421321392059326, "learning_rate": 0.0002, "epoch": 7.081076711664377, "step": 43800}, {"loss": 0.4317, "grad_norm": 1.2241183519363403, "learning_rate": 0.0002, "epoch": 7.082693395845122, "step": 43810}, {"loss": 0.4064, "grad_norm": 1.0945624113082886, "learning_rate": 0.0002, "epoch": 7.084310080025867, "step": 43820}, {"loss": 0.4049, "grad_norm": 1.2772969007492065, "learning_rate": 0.0002, "epoch": 7.085926764206612, "step": 43830}, {"loss": 0.4098, "grad_norm": 1.1715508699417114, "learning_rate": 0.0002, "epoch": 7.087543448387358, "step": 43840}, {"loss": 0.4756, "grad_norm": 1.1975586414337158, "learning_rate": 0.0002, "epoch": 7.089160132568103, "step": 43850}, {"loss": 0.4272, "grad_norm": 1.1673274040222168, "learning_rate": 0.0002, "epoch": 7.090776816748848, "step": 43860}, {"loss": 0.4435, "grad_norm": 1.096590518951416, "learning_rate": 0.0002, "epoch": 7.092393500929593, "step": 43870}, {"loss": 0.4329, "grad_norm": 1.0174020528793335, "learning_rate": 0.0002, "epoch": 7.0940101851103385, "step": 43880}, {"loss": 0.4307, "grad_norm": 1.0147380828857422, "learning_rate": 0.0002, "epoch": 7.095626869291084, "step": 43890}, {"loss": 0.4115, "grad_norm": 1.0056098699569702, "learning_rate": 0.0002, "epoch": 7.097243553471829, "step": 43900}, {"loss": 0.4181, "grad_norm": 1.4678083658218384, "learning_rate": 0.0002, "epoch": 7.098860237652574, "step": 43910}, {"loss": 0.4404, "grad_norm": 1.3740565776824951, "learning_rate": 0.0002, "epoch": 7.1004769218333195, "step": 43920}, {"loss": 0.4435, "grad_norm": 1.0279403924942017, "learning_rate": 0.0002, "epoch": 7.102093606014066, "step": 43930}, {"loss": 0.4247, "grad_norm": 1.186720371246338, "learning_rate": 0.0002, "epoch": 7.103710290194811, "step": 43940}, {"loss": 0.4001, "grad_norm": 1.3767904043197632, "learning_rate": 0.0002, "epoch": 7.105326974375556, "step": 43950}, {"loss": 0.4314, "grad_norm": 1.1637471914291382, "learning_rate": 0.0002, "epoch": 7.106943658556301, "step": 43960}, {"loss": 0.3996, "grad_norm": 1.1860042810440063, "learning_rate": 0.0002, "epoch": 7.1085603427370465, "step": 43970}, {"loss": 0.4014, "grad_norm": 1.080944538116455, "learning_rate": 0.0002, "epoch": 7.110177026917792, "step": 43980}, {"loss": 0.4152, "grad_norm": 1.0119353532791138, "learning_rate": 0.0002, "epoch": 7.111793711098537, "step": 43990}, {"loss": 0.4354, "grad_norm": 1.179388403892517, "learning_rate": 0.0002, "epoch": 7.113410395279282, "step": 44000}, {"loss": 0.4494, "grad_norm": 0.9202800989151001, "learning_rate": 0.0002, "epoch": 7.115027079460027, "step": 44010}, {"loss": 0.4356, "grad_norm": 1.142206072807312, "learning_rate": 0.0002, "epoch": 7.116643763640773, "step": 44020}, {"loss": 0.4197, "grad_norm": 1.17897367477417, "learning_rate": 0.0002, "epoch": 7.118260447821518, "step": 44030}, {"loss": 0.4394, "grad_norm": 1.238087773323059, "learning_rate": 0.0002, "epoch": 7.119877132002263, "step": 44040}, {"loss": 0.4844, "grad_norm": 1.5113195180892944, "learning_rate": 0.0002, "epoch": 7.121493816183008, "step": 44050}, {"loss": 0.4526, "grad_norm": 1.1819349527359009, "learning_rate": 0.0002, "epoch": 7.123110500363754, "step": 44060}, {"loss": 0.4071, "grad_norm": 1.1062556505203247, "learning_rate": 0.0002, "epoch": 7.124727184544499, "step": 44070}, {"loss": 0.4282, "grad_norm": 0.986954927444458, "learning_rate": 0.0002, "epoch": 7.126343868725245, "step": 44080}, {"loss": 0.4497, "grad_norm": 0.9641291499137878, "learning_rate": 0.0002, "epoch": 7.12796055290599, "step": 44090}, {"loss": 0.4348, "grad_norm": 0.9519979953765869, "learning_rate": 0.0002, "epoch": 7.129577237086735, "step": 44100}, {"loss": 0.4527, "grad_norm": 1.0477287769317627, "learning_rate": 0.0002, "epoch": 7.131193921267481, "step": 44110}, {"loss": 0.4168, "grad_norm": 0.9185389280319214, "learning_rate": 0.0002, "epoch": 7.132810605448226, "step": 44120}, {"loss": 0.4255, "grad_norm": 1.0224069356918335, "learning_rate": 0.0002, "epoch": 7.134427289628971, "step": 44130}, {"loss": 0.4598, "grad_norm": 1.0762630701065063, "learning_rate": 0.0002, "epoch": 7.136043973809716, "step": 44140}, {"loss": 0.4308, "grad_norm": 1.330917477607727, "learning_rate": 0.0002, "epoch": 7.137660657990462, "step": 44150}, {"loss": 0.4548, "grad_norm": 1.220115303993225, "learning_rate": 0.0002, "epoch": 7.139277342171207, "step": 44160}, {"loss": 0.4089, "grad_norm": 0.9959004521369934, "learning_rate": 0.0002, "epoch": 7.140894026351952, "step": 44170}, {"loss": 0.4475, "grad_norm": 1.272449016571045, "learning_rate": 0.0002, "epoch": 7.142510710532697, "step": 44180}, {"loss": 0.4268, "grad_norm": 1.0696483850479126, "learning_rate": 0.0002, "epoch": 7.1441273947134425, "step": 44190}, {"loss": 0.4218, "grad_norm": 1.347206711769104, "learning_rate": 0.0002, "epoch": 7.145744078894188, "step": 44200}, {"loss": 0.4652, "grad_norm": 1.1455401182174683, "learning_rate": 0.0002, "epoch": 7.147360763074933, "step": 44210}, {"loss": 0.4186, "grad_norm": 1.1443370580673218, "learning_rate": 0.0002, "epoch": 7.148977447255678, "step": 44220}, {"loss": 0.4669, "grad_norm": 1.0239921808242798, "learning_rate": 0.0002, "epoch": 7.150594131436424, "step": 44230}, {"loss": 0.4601, "grad_norm": 1.1596333980560303, "learning_rate": 0.0002, "epoch": 7.1522108156171695, "step": 44240}, {"loss": 0.44, "grad_norm": 1.2471510171890259, "learning_rate": 0.0002, "epoch": 7.153827499797915, "step": 44250}, {"loss": 0.426, "grad_norm": 1.0713822841644287, "learning_rate": 0.0002, "epoch": 7.15544418397866, "step": 44260}, {"loss": 0.4381, "grad_norm": 1.3523266315460205, "learning_rate": 0.0002, "epoch": 7.157060868159405, "step": 44270}, {"loss": 0.4101, "grad_norm": 1.1620066165924072, "learning_rate": 0.0002, "epoch": 7.1586775523401505, "step": 44280}, {"loss": 0.4195, "grad_norm": 1.2988619804382324, "learning_rate": 0.0002, "epoch": 7.160294236520896, "step": 44290}, {"loss": 0.4405, "grad_norm": 1.2527822256088257, "learning_rate": 0.0002, "epoch": 7.161910920701641, "step": 44300}, {"loss": 0.4813, "grad_norm": 1.2322553396224976, "learning_rate": 0.0002, "epoch": 7.163527604882386, "step": 44310}, {"loss": 0.4274, "grad_norm": 1.0497055053710938, "learning_rate": 0.0002, "epoch": 7.165144289063131, "step": 44320}, {"loss": 0.4236, "grad_norm": 1.1928341388702393, "learning_rate": 0.0002, "epoch": 7.166760973243877, "step": 44330}, {"loss": 0.4511, "grad_norm": 1.0016584396362305, "learning_rate": 0.0002, "epoch": 7.168377657424622, "step": 44340}, {"loss": 0.4566, "grad_norm": 1.0385509729385376, "learning_rate": 0.0002, "epoch": 7.169994341605367, "step": 44350}, {"loss": 0.4178, "grad_norm": 1.3217328786849976, "learning_rate": 0.0002, "epoch": 7.171611025786112, "step": 44360}, {"loss": 0.425, "grad_norm": 1.240696668624878, "learning_rate": 0.0002, "epoch": 7.1732277099668575, "step": 44370}, {"loss": 0.4572, "grad_norm": 1.1037760972976685, "learning_rate": 0.0002, "epoch": 7.174844394147604, "step": 44380}, {"loss": 0.4525, "grad_norm": 1.062762975692749, "learning_rate": 0.0002, "epoch": 7.176461078328349, "step": 44390}, {"loss": 0.4766, "grad_norm": 1.2859047651290894, "learning_rate": 0.0002, "epoch": 7.178077762509094, "step": 44400}, {"loss": 0.4511, "grad_norm": 1.1852408647537231, "learning_rate": 0.0002, "epoch": 7.179694446689839, "step": 44410}, {"loss": 0.4386, "grad_norm": 1.315587043762207, "learning_rate": 0.0002, "epoch": 7.181311130870585, "step": 44420}, {"loss": 0.4491, "grad_norm": 0.889542281627655, "learning_rate": 0.0002, "epoch": 7.18292781505133, "step": 44430}, {"loss": 0.4328, "grad_norm": 1.0123721361160278, "learning_rate": 0.0002, "epoch": 7.184544499232075, "step": 44440}, {"loss": 0.4096, "grad_norm": 1.0503462553024292, "learning_rate": 0.0002, "epoch": 7.18616118341282, "step": 44450}, {"loss": 0.44, "grad_norm": 1.338188886642456, "learning_rate": 0.0002, "epoch": 7.1877778675935655, "step": 44460}, {"loss": 0.4451, "grad_norm": 1.206543207168579, "learning_rate": 0.0002, "epoch": 7.189394551774311, "step": 44470}, {"loss": 0.4415, "grad_norm": 1.2013356685638428, "learning_rate": 0.0002, "epoch": 7.191011235955056, "step": 44480}, {"loss": 0.4291, "grad_norm": 1.1124168634414673, "learning_rate": 0.0002, "epoch": 7.192627920135801, "step": 44490}, {"loss": 0.4182, "grad_norm": 1.199379324913025, "learning_rate": 0.0002, "epoch": 7.194244604316546, "step": 44500}, {"loss": 0.4525, "grad_norm": 1.196746587753296, "learning_rate": 0.0002, "epoch": 7.195861288497292, "step": 44510}, {"loss": 0.4876, "grad_norm": 0.9684673547744751, "learning_rate": 0.0002, "epoch": 7.197477972678037, "step": 44520}, {"loss": 0.4403, "grad_norm": 1.5727651119232178, "learning_rate": 0.0002, "epoch": 7.199094656858783, "step": 44530}, {"loss": 0.4424, "grad_norm": 0.8371674418449402, "learning_rate": 0.0002, "epoch": 7.200711341039528, "step": 44540}, {"loss": 0.4366, "grad_norm": 1.0343716144561768, "learning_rate": 0.0002, "epoch": 7.2023280252202735, "step": 44550}, {"loss": 0.4557, "grad_norm": 1.1839478015899658, "learning_rate": 0.0002, "epoch": 7.203944709401019, "step": 44560}, {"loss": 0.4293, "grad_norm": 0.9466627836227417, "learning_rate": 0.0002, "epoch": 7.205561393581764, "step": 44570}, {"loss": 0.4651, "grad_norm": 1.1452360153198242, "learning_rate": 0.0002, "epoch": 7.207178077762509, "step": 44580}, {"loss": 0.5037, "grad_norm": 1.63698410987854, "learning_rate": 0.0002, "epoch": 7.208794761943254, "step": 44590}, {"loss": 0.4212, "grad_norm": 1.1124789714813232, "learning_rate": 0.0002, "epoch": 7.210411446124, "step": 44600}, {"loss": 0.4323, "grad_norm": 1.4233685731887817, "learning_rate": 0.0002, "epoch": 7.212028130304745, "step": 44610}, {"loss": 0.4176, "grad_norm": 1.302145004272461, "learning_rate": 0.0002, "epoch": 7.21364481448549, "step": 44620}, {"loss": 0.457, "grad_norm": 1.2115466594696045, "learning_rate": 0.0002, "epoch": 7.215261498666235, "step": 44630}, {"loss": 0.4419, "grad_norm": 1.0771325826644897, "learning_rate": 0.0002, "epoch": 7.216878182846981, "step": 44640}, {"loss": 0.4183, "grad_norm": 1.1603602170944214, "learning_rate": 0.0002, "epoch": 7.218494867027726, "step": 44650}, {"loss": 0.468, "grad_norm": 1.4013969898223877, "learning_rate": 0.0002, "epoch": 7.220111551208471, "step": 44660}, {"loss": 0.4867, "grad_norm": 1.2145777940750122, "learning_rate": 0.0002, "epoch": 7.221728235389216, "step": 44670}, {"loss": 0.4436, "grad_norm": 1.2084238529205322, "learning_rate": 0.0002, "epoch": 7.223344919569962, "step": 44680}, {"loss": 0.4423, "grad_norm": 1.1801965236663818, "learning_rate": 0.0002, "epoch": 7.224961603750708, "step": 44690}, {"loss": 0.426, "grad_norm": 0.9561195969581604, "learning_rate": 0.0002, "epoch": 7.226578287931453, "step": 44700}, {"loss": 0.4895, "grad_norm": 1.1857006549835205, "learning_rate": 0.0002, "epoch": 7.228194972112198, "step": 44710}, {"loss": 0.4382, "grad_norm": 1.1576673984527588, "learning_rate": 0.0002, "epoch": 7.229811656292943, "step": 44720}, {"loss": 0.4717, "grad_norm": 1.3517892360687256, "learning_rate": 0.0002, "epoch": 7.2314283404736885, "step": 44730}, {"loss": 0.4495, "grad_norm": 1.1489306688308716, "learning_rate": 0.0002, "epoch": 7.233045024654434, "step": 44740}, {"loss": 0.438, "grad_norm": 1.0758644342422485, "learning_rate": 0.0002, "epoch": 7.234661708835179, "step": 44750}, {"loss": 0.4431, "grad_norm": 1.1679041385650635, "learning_rate": 0.0002, "epoch": 7.236278393015924, "step": 44760}, {"loss": 0.504, "grad_norm": 1.1404961347579956, "learning_rate": 0.0002, "epoch": 7.2378950771966695, "step": 44770}, {"loss": 0.4499, "grad_norm": 1.2602572441101074, "learning_rate": 0.0002, "epoch": 7.239511761377415, "step": 44780}, {"loss": 0.4669, "grad_norm": 1.2912664413452148, "learning_rate": 0.0002, "epoch": 7.24112844555816, "step": 44790}, {"loss": 0.4336, "grad_norm": 1.340198278427124, "learning_rate": 0.0002, "epoch": 7.242745129738905, "step": 44800}, {"loss": 0.4336, "grad_norm": 1.0613332986831665, "learning_rate": 0.0002, "epoch": 7.24436181391965, "step": 44810}, {"loss": 0.4433, "grad_norm": 1.1658564805984497, "learning_rate": 0.0002, "epoch": 7.2459784981003965, "step": 44820}, {"loss": 0.4532, "grad_norm": 1.046440839767456, "learning_rate": 0.0002, "epoch": 7.247595182281142, "step": 44830}, {"loss": 0.4332, "grad_norm": 1.2335407733917236, "learning_rate": 0.0002, "epoch": 7.249211866461887, "step": 44840}, {"loss": 0.455, "grad_norm": 1.3742769956588745, "learning_rate": 0.0002, "epoch": 7.250828550642632, "step": 44850}, {"loss": 0.4297, "grad_norm": 1.1744071245193481, "learning_rate": 0.0002, "epoch": 7.252445234823377, "step": 44860}, {"loss": 0.4348, "grad_norm": 1.4268226623535156, "learning_rate": 0.0002, "epoch": 7.254061919004123, "step": 44870}, {"loss": 0.4485, "grad_norm": 1.1255686283111572, "learning_rate": 0.0002, "epoch": 7.255678603184868, "step": 44880}, {"loss": 0.4264, "grad_norm": 1.255053162574768, "learning_rate": 0.0002, "epoch": 7.257295287365613, "step": 44890}, {"loss": 0.455, "grad_norm": 1.4957616329193115, "learning_rate": 0.0002, "epoch": 7.258911971546358, "step": 44900}, {"loss": 0.4465, "grad_norm": 1.0546756982803345, "learning_rate": 0.0002, "epoch": 7.260528655727104, "step": 44910}, {"loss": 0.4802, "grad_norm": 1.4683036804199219, "learning_rate": 0.0002, "epoch": 7.262145339907849, "step": 44920}, {"loss": 0.4175, "grad_norm": 1.2027722597122192, "learning_rate": 0.0002, "epoch": 7.263762024088594, "step": 44930}, {"loss": 0.4316, "grad_norm": 1.277767539024353, "learning_rate": 0.0002, "epoch": 7.265378708269339, "step": 44940}, {"loss": 0.4749, "grad_norm": 1.4894379377365112, "learning_rate": 0.0002, "epoch": 7.2669953924500845, "step": 44950}, {"loss": 0.46, "grad_norm": 1.0998231172561646, "learning_rate": 0.0002, "epoch": 7.26861207663083, "step": 44960}, {"loss": 0.4414, "grad_norm": 1.3713536262512207, "learning_rate": 0.0002, "epoch": 7.270228760811576, "step": 44970}, {"loss": 0.4431, "grad_norm": 1.473396897315979, "learning_rate": 0.0002, "epoch": 7.271845444992321, "step": 44980}, {"loss": 0.4582, "grad_norm": 1.0893826484680176, "learning_rate": 0.0002, "epoch": 7.273462129173066, "step": 44990}, {"loss": 0.4297, "grad_norm": 1.4798463582992554, "learning_rate": 0.0002, "epoch": 7.275078813353812, "step": 45000}, {"loss": 0.4604, "grad_norm": 1.0536930561065674, "learning_rate": 0.0002, "epoch": 7.276695497534557, "step": 45010}, {"loss": 0.4396, "grad_norm": 1.064450740814209, "learning_rate": 0.0002, "epoch": 7.278312181715302, "step": 45020}, {"loss": 0.4445, "grad_norm": 1.3605865240097046, "learning_rate": 0.0002, "epoch": 7.279928865896047, "step": 45030}, {"loss": 0.4123, "grad_norm": 1.1779286861419678, "learning_rate": 0.0002, "epoch": 7.2815455500767925, "step": 45040}, {"loss": 0.4588, "grad_norm": 1.1568892002105713, "learning_rate": 0.0002, "epoch": 7.283162234257538, "step": 45050}, {"loss": 0.4557, "grad_norm": 1.0677175521850586, "learning_rate": 0.0002, "epoch": 7.284778918438283, "step": 45060}, {"loss": 0.4765, "grad_norm": 1.1939430236816406, "learning_rate": 0.0002, "epoch": 7.286395602619028, "step": 45070}, {"loss": 0.443, "grad_norm": 1.0273144245147705, "learning_rate": 0.0002, "epoch": 7.288012286799773, "step": 45080}, {"loss": 0.4472, "grad_norm": 1.358487844467163, "learning_rate": 0.0002, "epoch": 7.289628970980519, "step": 45090}, {"loss": 0.4207, "grad_norm": 1.2139160633087158, "learning_rate": 0.0002, "epoch": 7.291245655161264, "step": 45100}, {"loss": 0.4221, "grad_norm": 1.2484227418899536, "learning_rate": 0.0002, "epoch": 7.29286233934201, "step": 45110}, {"loss": 0.4351, "grad_norm": 1.2373738288879395, "learning_rate": 0.0002, "epoch": 7.294479023522755, "step": 45120}, {"loss": 0.472, "grad_norm": 1.3877158164978027, "learning_rate": 0.0002, "epoch": 7.2960957077035005, "step": 45130}, {"loss": 0.4741, "grad_norm": 1.1372028589248657, "learning_rate": 0.0002, "epoch": 7.297712391884246, "step": 45140}, {"loss": 0.4465, "grad_norm": 1.259987711906433, "learning_rate": 0.0002, "epoch": 7.299329076064991, "step": 45150}, {"loss": 0.4795, "grad_norm": 1.6501492261886597, "learning_rate": 0.0002, "epoch": 7.300945760245736, "step": 45160}, {"loss": 0.4441, "grad_norm": 1.5927983522415161, "learning_rate": 0.0002, "epoch": 7.302562444426481, "step": 45170}, {"loss": 0.4513, "grad_norm": 0.957084596157074, "learning_rate": 0.0002, "epoch": 7.304179128607227, "step": 45180}, {"loss": 0.4367, "grad_norm": 1.7777647972106934, "learning_rate": 0.0002, "epoch": 7.305795812787972, "step": 45190}, {"loss": 0.4365, "grad_norm": 1.1905052661895752, "learning_rate": 0.0002, "epoch": 7.307412496968717, "step": 45200}, {"loss": 0.4354, "grad_norm": 1.0944236516952515, "learning_rate": 0.0002, "epoch": 7.309029181149462, "step": 45210}, {"loss": 0.4558, "grad_norm": 1.171034336090088, "learning_rate": 0.0002, "epoch": 7.3106458653302075, "step": 45220}, {"loss": 0.4518, "grad_norm": 1.421743392944336, "learning_rate": 0.0002, "epoch": 7.312262549510953, "step": 45230}, {"loss": 0.4713, "grad_norm": 1.1282994747161865, "learning_rate": 0.0002, "epoch": 7.313879233691698, "step": 45240}, {"loss": 0.471, "grad_norm": 1.0742822885513306, "learning_rate": 0.0002, "epoch": 7.315495917872443, "step": 45250}, {"loss": 0.4887, "grad_norm": 1.2697997093200684, "learning_rate": 0.0002, "epoch": 7.317112602053189, "step": 45260}, {"loss": 0.461, "grad_norm": 1.2066359519958496, "learning_rate": 0.0002, "epoch": 7.318729286233935, "step": 45270}, {"loss": 0.463, "grad_norm": 1.0044163465499878, "learning_rate": 0.0002, "epoch": 7.32034597041468, "step": 45280}, {"loss": 0.4394, "grad_norm": 1.2365968227386475, "learning_rate": 0.0002, "epoch": 7.321962654595425, "step": 45290}, {"loss": 0.4305, "grad_norm": 1.0731542110443115, "learning_rate": 0.0002, "epoch": 7.32357933877617, "step": 45300}, {"loss": 0.4744, "grad_norm": 1.1595830917358398, "learning_rate": 0.0002, "epoch": 7.3251960229569155, "step": 45310}, {"loss": 0.4393, "grad_norm": 1.3445849418640137, "learning_rate": 0.0002, "epoch": 7.326812707137661, "step": 45320}, {"loss": 0.4288, "grad_norm": 1.3067926168441772, "learning_rate": 0.0002, "epoch": 7.328429391318406, "step": 45330}, {"loss": 0.4569, "grad_norm": 1.200667381286621, "learning_rate": 0.0002, "epoch": 7.330046075499151, "step": 45340}, {"loss": 0.4449, "grad_norm": 0.9936319589614868, "learning_rate": 0.0002, "epoch": 7.3316627596798964, "step": 45350}, {"loss": 0.4481, "grad_norm": 1.1291998624801636, "learning_rate": 0.0002, "epoch": 7.333279443860642, "step": 45360}, {"loss": 0.4643, "grad_norm": 1.3663034439086914, "learning_rate": 0.0002, "epoch": 7.334896128041387, "step": 45370}, {"loss": 0.4548, "grad_norm": 1.0762227773666382, "learning_rate": 0.0002, "epoch": 7.336512812222132, "step": 45380}, {"loss": 0.4495, "grad_norm": 0.9525768160820007, "learning_rate": 0.0002, "epoch": 7.338129496402877, "step": 45390}, {"loss": 0.472, "grad_norm": 1.1143709421157837, "learning_rate": 0.0002, "epoch": 7.339746180583623, "step": 45400}, {"loss": 0.4432, "grad_norm": 1.0711175203323364, "learning_rate": 0.0002, "epoch": 7.341362864764369, "step": 45410}, {"loss": 0.4603, "grad_norm": 1.2650856971740723, "learning_rate": 0.0002, "epoch": 7.342979548945114, "step": 45420}, {"loss": 0.5021, "grad_norm": 1.194861888885498, "learning_rate": 0.0002, "epoch": 7.344596233125859, "step": 45430}, {"loss": 0.467, "grad_norm": 1.4936751127243042, "learning_rate": 0.0002, "epoch": 7.346212917306604, "step": 45440}, {"loss": 0.4798, "grad_norm": 1.2938975095748901, "learning_rate": 0.0002, "epoch": 7.34782960148735, "step": 45450}, {"loss": 0.4589, "grad_norm": 1.2841941118240356, "learning_rate": 0.0002, "epoch": 7.349446285668095, "step": 45460}, {"loss": 0.4398, "grad_norm": 1.5376560688018799, "learning_rate": 0.0002, "epoch": 7.35106296984884, "step": 45470}, {"loss": 0.4583, "grad_norm": 1.1307156085968018, "learning_rate": 0.0002, "epoch": 7.352679654029585, "step": 45480}, {"loss": 0.4678, "grad_norm": 1.4883167743682861, "learning_rate": 0.0002, "epoch": 7.354296338210331, "step": 45490}, {"loss": 0.4966, "grad_norm": 1.0547393560409546, "learning_rate": 0.0002, "epoch": 7.355913022391076, "step": 45500}, {"loss": 0.4601, "grad_norm": 1.5476845502853394, "learning_rate": 0.0002, "epoch": 7.357529706571821, "step": 45510}, {"loss": 0.4466, "grad_norm": 1.1916698217391968, "learning_rate": 0.0002, "epoch": 7.359146390752566, "step": 45520}, {"loss": 0.4791, "grad_norm": 1.238319754600525, "learning_rate": 0.0002, "epoch": 7.3607630749333115, "step": 45530}, {"loss": 0.4818, "grad_norm": 1.4216728210449219, "learning_rate": 0.0002, "epoch": 7.362379759114057, "step": 45540}, {"loss": 0.4828, "grad_norm": 1.303995132446289, "learning_rate": 0.0002, "epoch": 7.363996443294802, "step": 45550}, {"loss": 0.464, "grad_norm": 1.2453089952468872, "learning_rate": 0.0002, "epoch": 7.365613127475548, "step": 45560}, {"loss": 0.4735, "grad_norm": 1.1971137523651123, "learning_rate": 0.0002, "epoch": 7.367229811656293, "step": 45570}, {"loss": 0.4415, "grad_norm": 1.0801963806152344, "learning_rate": 0.0002, "epoch": 7.3688464958370385, "step": 45580}, {"loss": 0.4946, "grad_norm": 1.1602367162704468, "learning_rate": 0.0002, "epoch": 7.370463180017784, "step": 45590}, {"loss": 0.45, "grad_norm": 1.1623423099517822, "learning_rate": 0.0002, "epoch": 7.372079864198529, "step": 45600}, {"loss": 0.4648, "grad_norm": 1.108467936515808, "learning_rate": 0.0002, "epoch": 7.373696548379274, "step": 45610}, {"loss": 0.4566, "grad_norm": 1.087322473526001, "learning_rate": 0.0002, "epoch": 7.3753132325600195, "step": 45620}, {"loss": 0.4505, "grad_norm": 1.0945587158203125, "learning_rate": 0.0002, "epoch": 7.376929916740765, "step": 45630}, {"loss": 0.4864, "grad_norm": 1.6565983295440674, "learning_rate": 0.0002, "epoch": 7.37854660092151, "step": 45640}, {"loss": 0.4491, "grad_norm": 1.1279444694519043, "learning_rate": 0.0002, "epoch": 7.380163285102255, "step": 45650}, {"loss": 0.4606, "grad_norm": 1.0888527631759644, "learning_rate": 0.0002, "epoch": 7.381779969283, "step": 45660}, {"loss": 0.429, "grad_norm": 1.1114956140518188, "learning_rate": 0.0002, "epoch": 7.383396653463746, "step": 45670}, {"loss": 0.4726, "grad_norm": 1.195497751235962, "learning_rate": 0.0002, "epoch": 7.385013337644491, "step": 45680}, {"loss": 0.4643, "grad_norm": 1.3111436367034912, "learning_rate": 0.0002, "epoch": 7.386630021825236, "step": 45690}, {"loss": 0.4777, "grad_norm": 1.22647225856781, "learning_rate": 0.0002, "epoch": 7.388246706005981, "step": 45700}, {"loss": 0.4877, "grad_norm": 0.9309225678443909, "learning_rate": 0.0002, "epoch": 7.389863390186727, "step": 45710}, {"loss": 0.4789, "grad_norm": 1.198773741722107, "learning_rate": 0.0002, "epoch": 7.391480074367473, "step": 45720}, {"loss": 0.4496, "grad_norm": 1.2208130359649658, "learning_rate": 0.0002, "epoch": 7.393096758548218, "step": 45730}, {"loss": 0.4614, "grad_norm": 1.0756449699401855, "learning_rate": 0.0002, "epoch": 7.394713442728963, "step": 45740}, {"loss": 0.4469, "grad_norm": 1.0117692947387695, "learning_rate": 0.0002, "epoch": 7.396330126909708, "step": 45750}, {"loss": 0.4217, "grad_norm": 1.1144468784332275, "learning_rate": 0.0002, "epoch": 7.397946811090454, "step": 45760}, {"loss": 0.4737, "grad_norm": 1.140549898147583, "learning_rate": 0.0002, "epoch": 7.399563495271199, "step": 45770}, {"loss": 0.453, "grad_norm": 1.2335172891616821, "learning_rate": 0.0002, "epoch": 7.401180179451944, "step": 45780}, {"loss": 0.4501, "grad_norm": 1.296393632888794, "learning_rate": 0.0002, "epoch": 7.402796863632689, "step": 45790}, {"loss": 0.4716, "grad_norm": 1.2551302909851074, "learning_rate": 0.0002, "epoch": 7.4044135478134345, "step": 45800}, {"loss": 0.451, "grad_norm": 1.1909204721450806, "learning_rate": 0.0002, "epoch": 7.40603023199418, "step": 45810}, {"loss": 0.4296, "grad_norm": 1.17038893699646, "learning_rate": 0.0002, "epoch": 7.407646916174925, "step": 45820}, {"loss": 0.4574, "grad_norm": 1.0033377408981323, "learning_rate": 0.0002, "epoch": 7.40926360035567, "step": 45830}, {"loss": 0.4413, "grad_norm": 1.2957805395126343, "learning_rate": 0.0002, "epoch": 7.4108802845364155, "step": 45840}, {"loss": 0.4597, "grad_norm": 1.347462773323059, "learning_rate": 0.0002, "epoch": 7.412496968717161, "step": 45850}, {"loss": 0.4499, "grad_norm": 1.3187026977539062, "learning_rate": 0.0002, "epoch": 7.414113652897907, "step": 45860}, {"loss": 0.4624, "grad_norm": 1.092236876487732, "learning_rate": 0.0002, "epoch": 7.415730337078652, "step": 45870}, {"loss": 0.4636, "grad_norm": 1.075634241104126, "learning_rate": 0.0002, "epoch": 7.417347021259397, "step": 45880}, {"loss": 0.4465, "grad_norm": 1.0200046300888062, "learning_rate": 0.0002, "epoch": 7.4189637054401425, "step": 45890}, {"loss": 0.47, "grad_norm": 1.1419479846954346, "learning_rate": 0.0002, "epoch": 7.420580389620888, "step": 45900}, {"loss": 0.4409, "grad_norm": 1.0798102617263794, "learning_rate": 0.0002, "epoch": 7.422197073801633, "step": 45910}, {"loss": 0.5173, "grad_norm": 0.9999571442604065, "learning_rate": 0.0002, "epoch": 7.423813757982378, "step": 45920}, {"loss": 0.4714, "grad_norm": 1.2220723628997803, "learning_rate": 0.0002, "epoch": 7.425430442163123, "step": 45930}, {"loss": 0.4844, "grad_norm": 1.1209388971328735, "learning_rate": 0.0002, "epoch": 7.427047126343869, "step": 45940}, {"loss": 0.4534, "grad_norm": 1.1198307275772095, "learning_rate": 0.0002, "epoch": 7.428663810524614, "step": 45950}, {"loss": 0.4486, "grad_norm": 1.0170516967773438, "learning_rate": 0.0002, "epoch": 7.430280494705359, "step": 45960}, {"loss": 0.488, "grad_norm": 1.2963446378707886, "learning_rate": 0.0002, "epoch": 7.431897178886104, "step": 45970}, {"loss": 0.4346, "grad_norm": 1.4202494621276855, "learning_rate": 0.0002, "epoch": 7.43351386306685, "step": 45980}, {"loss": 0.4917, "grad_norm": 1.066774845123291, "learning_rate": 0.0002, "epoch": 7.435130547247595, "step": 45990}, {"loss": 0.4897, "grad_norm": 1.2760428190231323, "learning_rate": 0.0002, "epoch": 7.43674723142834, "step": 46000}, {"loss": 0.4562, "grad_norm": 1.530720829963684, "learning_rate": 0.0002, "epoch": 7.438363915609086, "step": 46010}, {"loss": 0.4691, "grad_norm": 1.1914178133010864, "learning_rate": 0.0002, "epoch": 7.439980599789831, "step": 46020}, {"loss": 0.5038, "grad_norm": 1.466650128364563, "learning_rate": 0.0002, "epoch": 7.441597283970577, "step": 46030}, {"loss": 0.4673, "grad_norm": 1.1567928791046143, "learning_rate": 0.0002, "epoch": 7.443213968151322, "step": 46040}, {"loss": 0.4778, "grad_norm": 1.252336025238037, "learning_rate": 0.0002, "epoch": 7.444830652332067, "step": 46050}, {"loss": 0.4493, "grad_norm": 1.2095589637756348, "learning_rate": 0.0002, "epoch": 7.446447336512812, "step": 46060}, {"loss": 0.4407, "grad_norm": 1.4075263738632202, "learning_rate": 0.0002, "epoch": 7.4480640206935576, "step": 46070}, {"loss": 0.4328, "grad_norm": 1.2527226209640503, "learning_rate": 0.0002, "epoch": 7.449680704874303, "step": 46080}, {"loss": 0.4922, "grad_norm": 1.3044105768203735, "learning_rate": 0.0002, "epoch": 7.451297389055048, "step": 46090}, {"loss": 0.4465, "grad_norm": 1.2888941764831543, "learning_rate": 0.0002, "epoch": 7.452914073235793, "step": 46100}, {"loss": 0.4414, "grad_norm": 1.3148317337036133, "learning_rate": 0.0002, "epoch": 7.4545307574165385, "step": 46110}, {"loss": 0.4431, "grad_norm": 0.9526162147521973, "learning_rate": 0.0002, "epoch": 7.456147441597284, "step": 46120}, {"loss": 0.4422, "grad_norm": 1.2618519067764282, "learning_rate": 0.0002, "epoch": 7.457764125778029, "step": 46130}, {"loss": 0.4745, "grad_norm": 1.0392966270446777, "learning_rate": 0.0002, "epoch": 7.459380809958774, "step": 46140}, {"loss": 0.4589, "grad_norm": 1.3286794424057007, "learning_rate": 0.0002, "epoch": 7.460997494139519, "step": 46150}, {"loss": 0.4762, "grad_norm": 1.2377561330795288, "learning_rate": 0.0002, "epoch": 7.4626141783202655, "step": 46160}, {"loss": 0.4119, "grad_norm": 1.034134030342102, "learning_rate": 0.0002, "epoch": 7.464230862501011, "step": 46170}, {"loss": 0.4487, "grad_norm": 1.1719683408737183, "learning_rate": 0.0002, "epoch": 7.465847546681756, "step": 46180}, {"loss": 0.4423, "grad_norm": 1.182691216468811, "learning_rate": 0.0002, "epoch": 7.467464230862501, "step": 46190}, {"loss": 0.4341, "grad_norm": 1.1898412704467773, "learning_rate": 0.0002, "epoch": 7.4690809150432464, "step": 46200}, {"loss": 0.4753, "grad_norm": 1.0543978214263916, "learning_rate": 0.0002, "epoch": 7.470697599223992, "step": 46210}, {"loss": 0.4673, "grad_norm": 1.176971673965454, "learning_rate": 0.0002, "epoch": 7.472314283404737, "step": 46220}, {"loss": 0.4598, "grad_norm": 1.129456639289856, "learning_rate": 0.0002, "epoch": 7.473930967585482, "step": 46230}, {"loss": 0.4805, "grad_norm": 1.1782855987548828, "learning_rate": 0.0002, "epoch": 7.475547651766227, "step": 46240}, {"loss": 0.4979, "grad_norm": 1.1678800582885742, "learning_rate": 0.0002, "epoch": 7.477164335946973, "step": 46250}, {"loss": 0.4374, "grad_norm": 0.9768722653388977, "learning_rate": 0.0002, "epoch": 7.478781020127718, "step": 46260}, {"loss": 0.4683, "grad_norm": 1.3222670555114746, "learning_rate": 0.0002, "epoch": 7.480397704308463, "step": 46270}, {"loss": 0.459, "grad_norm": 1.0573948621749878, "learning_rate": 0.0002, "epoch": 7.482014388489208, "step": 46280}, {"loss": 0.5019, "grad_norm": 1.3233898878097534, "learning_rate": 0.0002, "epoch": 7.4836310726699535, "step": 46290}, {"loss": 0.4689, "grad_norm": 0.9695420265197754, "learning_rate": 0.0002, "epoch": 7.485247756850699, "step": 46300}, {"loss": 0.471, "grad_norm": 1.2072020769119263, "learning_rate": 0.0002, "epoch": 7.486864441031445, "step": 46310}, {"loss": 0.486, "grad_norm": 1.2161253690719604, "learning_rate": 0.0002, "epoch": 7.48848112521219, "step": 46320}, {"loss": 0.4581, "grad_norm": 1.185958743095398, "learning_rate": 0.0002, "epoch": 7.490097809392935, "step": 46330}, {"loss": 0.4617, "grad_norm": 1.3741549253463745, "learning_rate": 0.0002, "epoch": 7.491714493573681, "step": 46340}, {"loss": 0.4772, "grad_norm": 1.0586212873458862, "learning_rate": 0.0002, "epoch": 7.493331177754426, "step": 46350}, {"loss": 0.4644, "grad_norm": 1.2000513076782227, "learning_rate": 0.0002, "epoch": 7.494947861935171, "step": 46360}, {"loss": 0.4584, "grad_norm": 1.3326879739761353, "learning_rate": 0.0002, "epoch": 7.496564546115916, "step": 46370}, {"loss": 0.4741, "grad_norm": 1.3452857732772827, "learning_rate": 0.0002, "epoch": 7.4981812302966615, "step": 46380}, {"loss": 0.4747, "grad_norm": 1.2885284423828125, "learning_rate": 0.0002, "epoch": 7.499797914477407, "step": 46390}, {"loss": 0.4648, "grad_norm": 1.097342610359192, "learning_rate": 0.0002, "epoch": 7.501414598658152, "step": 46400}, {"loss": 0.4714, "grad_norm": 1.2342469692230225, "learning_rate": 0.0002, "epoch": 7.503031282838897, "step": 46410}, {"loss": 0.4421, "grad_norm": 1.0151721239089966, "learning_rate": 0.0002, "epoch": 7.504647967019642, "step": 46420}, {"loss": 0.4347, "grad_norm": 1.2487123012542725, "learning_rate": 0.0002, "epoch": 7.506264651200388, "step": 46430}, {"loss": 0.4768, "grad_norm": 0.9319046139717102, "learning_rate": 0.0002, "epoch": 7.507881335381134, "step": 46440}, {"loss": 0.4693, "grad_norm": 1.1362226009368896, "learning_rate": 0.0002, "epoch": 7.509498019561878, "step": 46450}, {"loss": 0.5007, "grad_norm": 1.2883973121643066, "learning_rate": 0.0002, "epoch": 7.511114703742624, "step": 46460}, {"loss": 0.4455, "grad_norm": 1.0892037153244019, "learning_rate": 0.0002, "epoch": 7.5127313879233695, "step": 46470}, {"loss": 0.4721, "grad_norm": 1.1870533227920532, "learning_rate": 0.0002, "epoch": 7.514348072104115, "step": 46480}, {"loss": 0.4824, "grad_norm": 1.2103877067565918, "learning_rate": 0.0002, "epoch": 7.51596475628486, "step": 46490}, {"loss": 0.4573, "grad_norm": 1.0980644226074219, "learning_rate": 0.0002, "epoch": 7.517581440465605, "step": 46500}, {"loss": 0.4759, "grad_norm": 1.4729726314544678, "learning_rate": 0.0002, "epoch": 7.51919812464635, "step": 46510}, {"loss": 0.4413, "grad_norm": 1.1808913946151733, "learning_rate": 0.0002, "epoch": 7.520814808827096, "step": 46520}, {"loss": 0.4278, "grad_norm": 1.2347747087478638, "learning_rate": 0.0002, "epoch": 7.522431493007841, "step": 46530}, {"loss": 0.4745, "grad_norm": 1.5921525955200195, "learning_rate": 0.0002, "epoch": 7.524048177188586, "step": 46540}, {"loss": 0.4525, "grad_norm": 1.1328861713409424, "learning_rate": 0.0002, "epoch": 7.525664861369331, "step": 46550}, {"loss": 0.4771, "grad_norm": 1.289947748184204, "learning_rate": 0.0002, "epoch": 7.527281545550077, "step": 46560}, {"loss": 0.4711, "grad_norm": 1.0198370218276978, "learning_rate": 0.0002, "epoch": 7.528898229730822, "step": 46570}, {"loss": 0.504, "grad_norm": 1.3007137775421143, "learning_rate": 0.0002, "epoch": 7.530514913911567, "step": 46580}, {"loss": 0.4496, "grad_norm": 1.2864280939102173, "learning_rate": 0.0002, "epoch": 7.532131598092313, "step": 46590}, {"loss": 0.463, "grad_norm": 1.1005513668060303, "learning_rate": 0.0002, "epoch": 7.5337482822730575, "step": 46600}, {"loss": 0.4426, "grad_norm": 0.9998318552970886, "learning_rate": 0.0002, "epoch": 7.535364966453804, "step": 46610}, {"loss": 0.4762, "grad_norm": 1.2042466402053833, "learning_rate": 0.0002, "epoch": 7.536981650634549, "step": 46620}, {"loss": 0.4685, "grad_norm": 1.3240692615509033, "learning_rate": 0.0002, "epoch": 7.538598334815294, "step": 46630}, {"loss": 0.4608, "grad_norm": 1.2145483493804932, "learning_rate": 0.0002, "epoch": 7.540215018996039, "step": 46640}, {"loss": 0.4608, "grad_norm": 1.169691801071167, "learning_rate": 0.0002, "epoch": 7.5418317031767845, "step": 46650}, {"loss": 0.4527, "grad_norm": 1.194045901298523, "learning_rate": 0.0002, "epoch": 7.54344838735753, "step": 46660}, {"loss": 0.4599, "grad_norm": 1.0481327772140503, "learning_rate": 0.0002, "epoch": 7.545065071538275, "step": 46670}, {"loss": 0.4729, "grad_norm": 1.0714460611343384, "learning_rate": 0.0002, "epoch": 7.54668175571902, "step": 46680}, {"loss": 0.4703, "grad_norm": 1.1811443567276, "learning_rate": 0.0002, "epoch": 7.5482984398997655, "step": 46690}, {"loss": 0.4628, "grad_norm": 1.2794281244277954, "learning_rate": 0.0002, "epoch": 7.549915124080511, "step": 46700}, {"loss": 0.4659, "grad_norm": 1.001287817955017, "learning_rate": 0.0002, "epoch": 7.551531808261256, "step": 46710}, {"loss": 0.4938, "grad_norm": 1.3598867654800415, "learning_rate": 0.0002, "epoch": 7.553148492442001, "step": 46720}, {"loss": 0.4731, "grad_norm": 1.206254482269287, "learning_rate": 0.0002, "epoch": 7.554765176622746, "step": 46730}, {"loss": 0.4581, "grad_norm": 1.1095832586288452, "learning_rate": 0.0002, "epoch": 7.5563818608034925, "step": 46740}, {"loss": 0.4625, "grad_norm": 1.3912206888198853, "learning_rate": 0.0002, "epoch": 7.557998544984237, "step": 46750}, {"loss": 0.4464, "grad_norm": 0.9883413314819336, "learning_rate": 0.0002, "epoch": 7.559615229164983, "step": 46760}, {"loss": 0.4535, "grad_norm": 1.0965087413787842, "learning_rate": 0.0002, "epoch": 7.561231913345728, "step": 46770}, {"loss": 0.469, "grad_norm": 1.092261552810669, "learning_rate": 0.0002, "epoch": 7.562848597526473, "step": 46780}, {"loss": 0.488, "grad_norm": 1.0443673133850098, "learning_rate": 0.0002, "epoch": 7.564465281707219, "step": 46790}, {"loss": 0.4875, "grad_norm": 1.2420614957809448, "learning_rate": 0.0002, "epoch": 7.566081965887964, "step": 46800}, {"loss": 0.4911, "grad_norm": 1.0510783195495605, "learning_rate": 0.0002, "epoch": 7.567698650068709, "step": 46810}, {"loss": 0.4541, "grad_norm": 1.0291800498962402, "learning_rate": 0.0002, "epoch": 7.569315334249454, "step": 46820}, {"loss": 0.4591, "grad_norm": 1.1784595251083374, "learning_rate": 0.0002, "epoch": 7.5709320184302, "step": 46830}, {"loss": 0.5154, "grad_norm": 1.0424436330795288, "learning_rate": 0.0002, "epoch": 7.572548702610945, "step": 46840}, {"loss": 0.4612, "grad_norm": 1.182131290435791, "learning_rate": 0.0002, "epoch": 7.57416538679169, "step": 46850}, {"loss": 0.446, "grad_norm": 0.9917051792144775, "learning_rate": 0.0002, "epoch": 7.575782070972435, "step": 46860}, {"loss": 0.4428, "grad_norm": 1.1616078615188599, "learning_rate": 0.0002, "epoch": 7.5773987551531805, "step": 46870}, {"loss": 0.4769, "grad_norm": 1.401071548461914, "learning_rate": 0.0002, "epoch": 7.579015439333926, "step": 46880}, {"loss": 0.4635, "grad_norm": 0.874487578868866, "learning_rate": 0.0002, "epoch": 7.580632123514672, "step": 46890}, {"loss": 0.4641, "grad_norm": 1.2511193752288818, "learning_rate": 0.0002, "epoch": 7.582248807695416, "step": 46900}, {"loss": 0.4715, "grad_norm": 1.7548277378082275, "learning_rate": 0.0002, "epoch": 7.583865491876162, "step": 46910}, {"loss": 0.4681, "grad_norm": 1.349366545677185, "learning_rate": 0.0002, "epoch": 7.5854821760569076, "step": 46920}, {"loss": 0.4819, "grad_norm": 1.0609583854675293, "learning_rate": 0.0002, "epoch": 7.587098860237653, "step": 46930}, {"loss": 0.4498, "grad_norm": 1.031512975692749, "learning_rate": 0.0002, "epoch": 7.588715544418398, "step": 46940}, {"loss": 0.4688, "grad_norm": 1.1440242528915405, "learning_rate": 0.0002, "epoch": 7.590332228599143, "step": 46950}, {"loss": 0.4568, "grad_norm": 1.2762987613677979, "learning_rate": 0.0002, "epoch": 7.5919489127798885, "step": 46960}, {"loss": 0.4569, "grad_norm": 1.167269229888916, "learning_rate": 0.0002, "epoch": 7.593565596960634, "step": 46970}, {"loss": 0.461, "grad_norm": 1.131127953529358, "learning_rate": 0.0002, "epoch": 7.595182281141379, "step": 46980}, {"loss": 0.4666, "grad_norm": 1.4527075290679932, "learning_rate": 0.0002, "epoch": 7.596798965322124, "step": 46990}, {"loss": 0.4973, "grad_norm": 1.330132007598877, "learning_rate": 0.0002, "epoch": 7.598415649502869, "step": 47000}, {"loss": 0.4969, "grad_norm": 1.4223501682281494, "learning_rate": 0.0002, "epoch": 7.600032333683615, "step": 47010}, {"loss": 0.4572, "grad_norm": 1.2045072317123413, "learning_rate": 0.0002, "epoch": 7.60164901786436, "step": 47020}, {"loss": 0.4666, "grad_norm": 1.1549896001815796, "learning_rate": 0.0002, "epoch": 7.603265702045105, "step": 47030}, {"loss": 0.4383, "grad_norm": 1.2221543788909912, "learning_rate": 0.0002, "epoch": 7.604882386225851, "step": 47040}, {"loss": 0.4826, "grad_norm": 1.1171326637268066, "learning_rate": 0.0002, "epoch": 7.6064990704065965, "step": 47050}, {"loss": 0.4465, "grad_norm": 1.073671817779541, "learning_rate": 0.0002, "epoch": 7.608115754587342, "step": 47060}, {"loss": 0.4623, "grad_norm": 1.2524123191833496, "learning_rate": 0.0002, "epoch": 7.609732438768087, "step": 47070}, {"loss": 0.4538, "grad_norm": 1.2015056610107422, "learning_rate": 0.0002, "epoch": 7.611349122948832, "step": 47080}, {"loss": 0.4871, "grad_norm": 1.2454534769058228, "learning_rate": 0.0002, "epoch": 7.612965807129577, "step": 47090}, {"loss": 0.5064, "grad_norm": 0.9815779328346252, "learning_rate": 0.0002, "epoch": 7.614582491310323, "step": 47100}, {"loss": 0.4841, "grad_norm": 1.1437602043151855, "learning_rate": 0.0002, "epoch": 7.616199175491068, "step": 47110}, {"loss": 0.453, "grad_norm": 1.1004078388214111, "learning_rate": 0.0002, "epoch": 7.617815859671813, "step": 47120}, {"loss": 0.4552, "grad_norm": 1.069453477859497, "learning_rate": 0.0002, "epoch": 7.619432543852558, "step": 47130}, {"loss": 0.4627, "grad_norm": 1.1434191465377808, "learning_rate": 0.0002, "epoch": 7.6210492280333035, "step": 47140}, {"loss": 0.4882, "grad_norm": 1.216845989227295, "learning_rate": 0.0002, "epoch": 7.622665912214049, "step": 47150}, {"loss": 0.481, "grad_norm": 1.2302134037017822, "learning_rate": 0.0002, "epoch": 7.624282596394794, "step": 47160}, {"loss": 0.4806, "grad_norm": 1.4284924268722534, "learning_rate": 0.0002, "epoch": 7.625899280575539, "step": 47170}, {"loss": 0.4458, "grad_norm": 1.3359615802764893, "learning_rate": 0.0002, "epoch": 7.6275159647562845, "step": 47180}, {"loss": 0.4842, "grad_norm": 1.0242379903793335, "learning_rate": 0.0002, "epoch": 7.629132648937031, "step": 47190}, {"loss": 0.5137, "grad_norm": 1.249513030052185, "learning_rate": 0.0002, "epoch": 7.630749333117776, "step": 47200}, {"loss": 0.4462, "grad_norm": 1.0881463289260864, "learning_rate": 0.0002, "epoch": 7.632366017298521, "step": 47210}, {"loss": 0.4864, "grad_norm": 1.2903773784637451, "learning_rate": 0.0002, "epoch": 7.633982701479266, "step": 47220}, {"loss": 0.4729, "grad_norm": 1.1671710014343262, "learning_rate": 0.0002, "epoch": 7.6355993856600115, "step": 47230}, {"loss": 0.4936, "grad_norm": 1.1960735321044922, "learning_rate": 0.0002, "epoch": 7.637216069840757, "step": 47240}, {"loss": 0.4884, "grad_norm": 1.2692298889160156, "learning_rate": 0.0002, "epoch": 7.638832754021502, "step": 47250}, {"loss": 0.423, "grad_norm": 0.9812195301055908, "learning_rate": 0.0002, "epoch": 7.640449438202247, "step": 47260}, {"loss": 0.4737, "grad_norm": 1.3986053466796875, "learning_rate": 0.0002, "epoch": 7.642066122382992, "step": 47270}, {"loss": 0.4834, "grad_norm": 1.2692067623138428, "learning_rate": 0.0002, "epoch": 7.643682806563738, "step": 47280}, {"loss": 0.4893, "grad_norm": 1.1185054779052734, "learning_rate": 0.0002, "epoch": 7.645299490744483, "step": 47290}, {"loss": 0.4828, "grad_norm": 1.2837327718734741, "learning_rate": 0.0002, "epoch": 7.646916174925228, "step": 47300}, {"loss": 0.4891, "grad_norm": 1.8518418073654175, "learning_rate": 0.0002, "epoch": 7.648532859105973, "step": 47310}, {"loss": 0.4626, "grad_norm": 0.9781302213668823, "learning_rate": 0.0002, "epoch": 7.650149543286719, "step": 47320}, {"loss": 0.501, "grad_norm": 1.0777910947799683, "learning_rate": 0.0002, "epoch": 7.651766227467464, "step": 47330}, {"loss": 0.4927, "grad_norm": 1.2031499147415161, "learning_rate": 0.0002, "epoch": 7.65338291164821, "step": 47340}, {"loss": 0.439, "grad_norm": 1.14322829246521, "learning_rate": 0.0002, "epoch": 7.654999595828955, "step": 47350}, {"loss": 0.4481, "grad_norm": 1.3211992979049683, "learning_rate": 0.0002, "epoch": 7.6566162800097, "step": 47360}, {"loss": 0.4462, "grad_norm": 1.3632899522781372, "learning_rate": 0.0002, "epoch": 7.658232964190446, "step": 47370}, {"loss": 0.4934, "grad_norm": 1.2593929767608643, "learning_rate": 0.0002, "epoch": 7.659849648371191, "step": 47380}, {"loss": 0.4645, "grad_norm": 1.442670464515686, "learning_rate": 0.0002, "epoch": 7.661466332551936, "step": 47390}, {"loss": 0.4584, "grad_norm": 1.2304763793945312, "learning_rate": 0.0002, "epoch": 7.663083016732681, "step": 47400}, {"loss": 0.464, "grad_norm": 1.0182652473449707, "learning_rate": 0.0002, "epoch": 7.664699700913427, "step": 47410}, {"loss": 0.457, "grad_norm": 1.365441083908081, "learning_rate": 0.0002, "epoch": 7.666316385094172, "step": 47420}, {"loss": 0.4787, "grad_norm": 1.1578556299209595, "learning_rate": 0.0002, "epoch": 7.667933069274917, "step": 47430}, {"loss": 0.486, "grad_norm": 1.0346194505691528, "learning_rate": 0.0002, "epoch": 7.669549753455662, "step": 47440}, {"loss": 0.4703, "grad_norm": 1.2567378282546997, "learning_rate": 0.0002, "epoch": 7.6711664376364075, "step": 47450}, {"loss": 0.4853, "grad_norm": 1.1669118404388428, "learning_rate": 0.0002, "epoch": 7.672783121817153, "step": 47460}, {"loss": 0.4869, "grad_norm": 1.0174756050109863, "learning_rate": 0.0002, "epoch": 7.674399805997898, "step": 47470}, {"loss": 0.4601, "grad_norm": 1.0962231159210205, "learning_rate": 0.0002, "epoch": 7.676016490178643, "step": 47480}, {"loss": 0.4866, "grad_norm": 1.1098674535751343, "learning_rate": 0.0002, "epoch": 7.677633174359389, "step": 47490}, {"loss": 0.4682, "grad_norm": 1.1441160440444946, "learning_rate": 0.0002, "epoch": 7.6792498585401345, "step": 47500}, {"loss": 0.4552, "grad_norm": 1.0473432540893555, "learning_rate": 0.0002, "epoch": 7.68086654272088, "step": 47510}, {"loss": 0.4771, "grad_norm": 1.2954738140106201, "learning_rate": 0.0002, "epoch": 7.682483226901625, "step": 47520}, {"loss": 0.5012, "grad_norm": 1.2931294441223145, "learning_rate": 0.0002, "epoch": 7.68409991108237, "step": 47530}, {"loss": 0.4808, "grad_norm": 1.4005156755447388, "learning_rate": 0.0002, "epoch": 7.6857165952631155, "step": 47540}, {"loss": 0.4847, "grad_norm": 1.0998929738998413, "learning_rate": 0.0002, "epoch": 7.687333279443861, "step": 47550}, {"loss": 0.4839, "grad_norm": 1.3478347063064575, "learning_rate": 0.0002, "epoch": 7.688949963624606, "step": 47560}, {"loss": 0.4918, "grad_norm": 1.2991969585418701, "learning_rate": 0.0002, "epoch": 7.690566647805351, "step": 47570}, {"loss": 0.4673, "grad_norm": 1.0892608165740967, "learning_rate": 0.0002, "epoch": 7.692183331986096, "step": 47580}, {"loss": 0.4937, "grad_norm": 1.2230998277664185, "learning_rate": 0.0002, "epoch": 7.693800016166842, "step": 47590}, {"loss": 0.5222, "grad_norm": 1.2635555267333984, "learning_rate": 0.0002, "epoch": 7.695416700347587, "step": 47600}, {"loss": 0.4883, "grad_norm": 1.1720705032348633, "learning_rate": 0.0002, "epoch": 7.697033384528332, "step": 47610}, {"loss": 0.452, "grad_norm": 1.1134333610534668, "learning_rate": 0.0002, "epoch": 7.698650068709077, "step": 47620}, {"loss": 0.4859, "grad_norm": 1.2643009424209595, "learning_rate": 0.0002, "epoch": 7.7002667528898225, "step": 47630}, {"loss": 0.4825, "grad_norm": 1.1145045757293701, "learning_rate": 0.0002, "epoch": 7.701883437070569, "step": 47640}, {"loss": 0.4735, "grad_norm": 1.1808549165725708, "learning_rate": 0.0002, "epoch": 7.703500121251314, "step": 47650}, {"loss": 0.4841, "grad_norm": 1.2996630668640137, "learning_rate": 0.0002, "epoch": 7.705116805432059, "step": 47660}, {"loss": 0.4712, "grad_norm": 1.2786413431167603, "learning_rate": 0.0002, "epoch": 7.706733489612804, "step": 47670}, {"loss": 0.4694, "grad_norm": 1.3245121240615845, "learning_rate": 0.0002, "epoch": 7.70835017379355, "step": 47680}, {"loss": 0.4467, "grad_norm": 1.4168202877044678, "learning_rate": 0.0002, "epoch": 7.709966857974295, "step": 47690}, {"loss": 0.5143, "grad_norm": 1.0354000329971313, "learning_rate": 0.0002, "epoch": 7.71158354215504, "step": 47700}, {"loss": 0.4703, "grad_norm": 0.9630362391471863, "learning_rate": 0.0002, "epoch": 7.713200226335785, "step": 47710}, {"loss": 0.4996, "grad_norm": 1.1045806407928467, "learning_rate": 0.0002, "epoch": 7.7148169105165305, "step": 47720}, {"loss": 0.4756, "grad_norm": 1.2403767108917236, "learning_rate": 0.0002, "epoch": 7.716433594697276, "step": 47730}, {"loss": 0.4658, "grad_norm": 0.9893410801887512, "learning_rate": 0.0002, "epoch": 7.718050278878021, "step": 47740}, {"loss": 0.4463, "grad_norm": 1.0749315023422241, "learning_rate": 0.0002, "epoch": 7.719666963058766, "step": 47750}, {"loss": 0.467, "grad_norm": 1.2851510047912598, "learning_rate": 0.0002, "epoch": 7.721283647239511, "step": 47760}, {"loss": 0.489, "grad_norm": 1.2964261770248413, "learning_rate": 0.0002, "epoch": 7.722900331420257, "step": 47770}, {"loss": 0.4702, "grad_norm": 1.0603861808776855, "learning_rate": 0.0002, "epoch": 7.724517015601002, "step": 47780}, {"loss": 0.478, "grad_norm": 1.2728440761566162, "learning_rate": 0.0002, "epoch": 7.726133699781748, "step": 47790}, {"loss": 0.4746, "grad_norm": 1.20509934425354, "learning_rate": 0.0002, "epoch": 7.727750383962493, "step": 47800}, {"loss": 0.4556, "grad_norm": 1.397595763206482, "learning_rate": 0.0002, "epoch": 7.7293670681432385, "step": 47810}, {"loss": 0.4736, "grad_norm": 1.2595560550689697, "learning_rate": 0.0002, "epoch": 7.730983752323984, "step": 47820}, {"loss": 0.5061, "grad_norm": 1.166074514389038, "learning_rate": 0.0002, "epoch": 7.732600436504729, "step": 47830}, {"loss": 0.4907, "grad_norm": 1.258192777633667, "learning_rate": 0.0002, "epoch": 7.734217120685474, "step": 47840}, {"loss": 0.5256, "grad_norm": 1.0394890308380127, "learning_rate": 0.0002, "epoch": 7.735833804866219, "step": 47850}, {"loss": 0.4863, "grad_norm": 1.2017768621444702, "learning_rate": 0.0002, "epoch": 7.737450489046965, "step": 47860}, {"loss": 0.4784, "grad_norm": 1.1070265769958496, "learning_rate": 0.0002, "epoch": 7.73906717322771, "step": 47870}, {"loss": 0.4616, "grad_norm": 1.0544345378875732, "learning_rate": 0.0002, "epoch": 7.740683857408455, "step": 47880}, {"loss": 0.4519, "grad_norm": 1.0194088220596313, "learning_rate": 0.0002, "epoch": 7.7423005415892, "step": 47890}, {"loss": 0.4758, "grad_norm": 1.3095234632492065, "learning_rate": 0.0002, "epoch": 7.743917225769946, "step": 47900}, {"loss": 0.4646, "grad_norm": 1.0579626560211182, "learning_rate": 0.0002, "epoch": 7.745533909950691, "step": 47910}, {"loss": 0.4532, "grad_norm": 1.012990951538086, "learning_rate": 0.0002, "epoch": 7.747150594131437, "step": 47920}, {"loss": 0.4775, "grad_norm": 1.485148549079895, "learning_rate": 0.0002, "epoch": 7.748767278312181, "step": 47930}, {"loss": 0.4892, "grad_norm": 1.3595696687698364, "learning_rate": 0.0002, "epoch": 7.750383962492927, "step": 47940}, {"loss": 0.4609, "grad_norm": 0.9945753216743469, "learning_rate": 0.0002, "epoch": 7.752000646673673, "step": 47950}, {"loss": 0.5138, "grad_norm": 1.2098956108093262, "learning_rate": 0.0002, "epoch": 7.753617330854418, "step": 47960}, {"loss": 0.4815, "grad_norm": 1.3056198358535767, "learning_rate": 0.0002, "epoch": 7.755234015035163, "step": 47970}, {"loss": 0.4761, "grad_norm": 1.2247772216796875, "learning_rate": 0.0002, "epoch": 7.756850699215908, "step": 47980}, {"loss": 0.5023, "grad_norm": 1.397642970085144, "learning_rate": 0.0002, "epoch": 7.7584673833966535, "step": 47990}, {"loss": 0.4901, "grad_norm": 1.2565888166427612, "learning_rate": 0.0002, "epoch": 7.760084067577399, "step": 48000}, {"loss": 0.469, "grad_norm": 1.0065099000930786, "learning_rate": 0.0002, "epoch": 7.761700751758144, "step": 48010}, {"loss": 0.4886, "grad_norm": 1.1466305255889893, "learning_rate": 0.0002, "epoch": 7.763317435938889, "step": 48020}, {"loss": 0.4898, "grad_norm": 1.4492419958114624, "learning_rate": 0.0002, "epoch": 7.7649341201196345, "step": 48030}, {"loss": 0.489, "grad_norm": 1.0945932865142822, "learning_rate": 0.0002, "epoch": 7.76655080430038, "step": 48040}, {"loss": 0.4968, "grad_norm": 1.1938602924346924, "learning_rate": 0.0002, "epoch": 7.768167488481125, "step": 48050}, {"loss": 0.4497, "grad_norm": 1.168890357017517, "learning_rate": 0.0002, "epoch": 7.76978417266187, "step": 48060}, {"loss": 0.4881, "grad_norm": 1.3134305477142334, "learning_rate": 0.0002, "epoch": 7.771400856842616, "step": 48070}, {"loss": 0.5226, "grad_norm": 1.044438123703003, "learning_rate": 0.0002, "epoch": 7.773017541023361, "step": 48080}, {"loss": 0.4497, "grad_norm": 1.1275628805160522, "learning_rate": 0.0002, "epoch": 7.774634225204107, "step": 48090}, {"loss": 0.5063, "grad_norm": 1.0877318382263184, "learning_rate": 0.0002, "epoch": 7.776250909384852, "step": 48100}, {"loss": 0.4795, "grad_norm": 1.4800893068313599, "learning_rate": 0.0002, "epoch": 7.777867593565597, "step": 48110}, {"loss": 0.4984, "grad_norm": 1.1495977640151978, "learning_rate": 0.0002, "epoch": 7.779484277746342, "step": 48120}, {"loss": 0.4662, "grad_norm": 1.2175556421279907, "learning_rate": 0.0002, "epoch": 7.781100961927088, "step": 48130}, {"loss": 0.4935, "grad_norm": 1.150556206703186, "learning_rate": 0.0002, "epoch": 7.782717646107833, "step": 48140}, {"loss": 0.5039, "grad_norm": 1.051145315170288, "learning_rate": 0.0002, "epoch": 7.784334330288578, "step": 48150}, {"loss": 0.4611, "grad_norm": 1.2842742204666138, "learning_rate": 0.0002, "epoch": 7.785951014469323, "step": 48160}, {"loss": 0.4589, "grad_norm": 1.2251030206680298, "learning_rate": 0.0002, "epoch": 7.787567698650069, "step": 48170}, {"loss": 0.4905, "grad_norm": 1.2809321880340576, "learning_rate": 0.0002, "epoch": 7.789184382830814, "step": 48180}, {"loss": 0.4569, "grad_norm": 1.005690336227417, "learning_rate": 0.0002, "epoch": 7.790801067011559, "step": 48190}, {"loss": 0.4862, "grad_norm": 1.325501561164856, "learning_rate": 0.0002, "epoch": 7.792417751192304, "step": 48200}, {"loss": 0.4384, "grad_norm": 1.4551857709884644, "learning_rate": 0.0002, "epoch": 7.7940344353730495, "step": 48210}, {"loss": 0.4696, "grad_norm": 1.3399626016616821, "learning_rate": 0.0002, "epoch": 7.795651119553796, "step": 48220}, {"loss": 0.4654, "grad_norm": 1.0379714965820312, "learning_rate": 0.0002, "epoch": 7.79726780373454, "step": 48230}, {"loss": 0.4915, "grad_norm": 0.9725802540779114, "learning_rate": 0.0002, "epoch": 7.798884487915286, "step": 48240}, {"loss": 0.4583, "grad_norm": 1.0202224254608154, "learning_rate": 0.0002, "epoch": 7.800501172096031, "step": 48250}, {"loss": 0.4792, "grad_norm": 0.9477742910385132, "learning_rate": 0.0002, "epoch": 7.802117856276777, "step": 48260}, {"loss": 0.4836, "grad_norm": 1.2726924419403076, "learning_rate": 0.0002, "epoch": 7.803734540457522, "step": 48270}, {"loss": 0.494, "grad_norm": 1.453190565109253, "learning_rate": 0.0002, "epoch": 7.805351224638267, "step": 48280}, {"loss": 0.4559, "grad_norm": 1.2806978225708008, "learning_rate": 0.0002, "epoch": 7.806967908819012, "step": 48290}, {"loss": 0.4867, "grad_norm": 1.0897129774093628, "learning_rate": 0.0002, "epoch": 7.8085845929997575, "step": 48300}, {"loss": 0.4939, "grad_norm": 1.381636381149292, "learning_rate": 0.0002, "epoch": 7.810201277180503, "step": 48310}, {"loss": 0.4797, "grad_norm": 0.9954851269721985, "learning_rate": 0.0002, "epoch": 7.811817961361248, "step": 48320}, {"loss": 0.4995, "grad_norm": 1.1756198406219482, "learning_rate": 0.0002, "epoch": 7.813434645541993, "step": 48330}, {"loss": 0.4904, "grad_norm": 1.2087817192077637, "learning_rate": 0.0002, "epoch": 7.815051329722738, "step": 48340}, {"loss": 0.4935, "grad_norm": 1.3075505495071411, "learning_rate": 0.0002, "epoch": 7.816668013903484, "step": 48350}, {"loss": 0.486, "grad_norm": 1.1872076988220215, "learning_rate": 0.0002, "epoch": 7.818284698084229, "step": 48360}, {"loss": 0.538, "grad_norm": 1.2134783267974854, "learning_rate": 0.0002, "epoch": 7.819901382264975, "step": 48370}, {"loss": 0.4759, "grad_norm": 1.28566312789917, "learning_rate": 0.0002, "epoch": 7.821518066445719, "step": 48380}, {"loss": 0.4962, "grad_norm": 1.0578798055648804, "learning_rate": 0.0002, "epoch": 7.8231347506264655, "step": 48390}, {"loss": 0.4924, "grad_norm": 1.1225441694259644, "learning_rate": 0.0002, "epoch": 7.824751434807211, "step": 48400}, {"loss": 0.5081, "grad_norm": 1.2029428482055664, "learning_rate": 0.0002, "epoch": 7.826368118987956, "step": 48410}, {"loss": 0.4669, "grad_norm": 1.252485990524292, "learning_rate": 0.0002, "epoch": 7.827984803168701, "step": 48420}, {"loss": 0.4932, "grad_norm": 1.1822574138641357, "learning_rate": 0.0002, "epoch": 7.829601487349446, "step": 48430}, {"loss": 0.4692, "grad_norm": 1.2428245544433594, "learning_rate": 0.0002, "epoch": 7.831218171530192, "step": 48440}, {"loss": 0.4796, "grad_norm": 1.0565894842147827, "learning_rate": 0.0002, "epoch": 7.832834855710937, "step": 48450}, {"loss": 0.5016, "grad_norm": 1.363452672958374, "learning_rate": 0.0002, "epoch": 7.834451539891682, "step": 48460}, {"loss": 0.463, "grad_norm": 1.2436026334762573, "learning_rate": 0.0002, "epoch": 7.836068224072427, "step": 48470}, {"loss": 0.4794, "grad_norm": 1.2623029947280884, "learning_rate": 0.0002, "epoch": 7.8376849082531725, "step": 48480}, {"loss": 0.4863, "grad_norm": 1.0942288637161255, "learning_rate": 0.0002, "epoch": 7.839301592433918, "step": 48490}, {"loss": 0.4723, "grad_norm": 1.1791462898254395, "learning_rate": 0.0002, "epoch": 7.840918276614663, "step": 48500}, {"loss": 0.501, "grad_norm": 1.3342814445495605, "learning_rate": 0.0002, "epoch": 7.842534960795408, "step": 48510}, {"loss": 0.4654, "grad_norm": 1.0511828660964966, "learning_rate": 0.0002, "epoch": 7.844151644976154, "step": 48520}, {"loss": 0.5051, "grad_norm": 1.48568594455719, "learning_rate": 0.0002, "epoch": 7.845768329156899, "step": 48530}, {"loss": 0.4572, "grad_norm": 1.296844720840454, "learning_rate": 0.0002, "epoch": 7.847385013337645, "step": 48540}, {"loss": 0.5216, "grad_norm": 1.3032835721969604, "learning_rate": 0.0002, "epoch": 7.84900169751839, "step": 48550}, {"loss": 0.472, "grad_norm": 1.260769009590149, "learning_rate": 0.0002, "epoch": 7.850618381699135, "step": 48560}, {"loss": 0.5094, "grad_norm": 1.3309531211853027, "learning_rate": 0.0002, "epoch": 7.8522350658798805, "step": 48570}, {"loss": 0.4759, "grad_norm": 1.1907469034194946, "learning_rate": 0.0002, "epoch": 7.853851750060626, "step": 48580}, {"loss": 0.485, "grad_norm": 0.9690865874290466, "learning_rate": 0.0002, "epoch": 7.855468434241371, "step": 48590}, {"loss": 0.4899, "grad_norm": 1.2417343854904175, "learning_rate": 0.0002, "epoch": 7.857085118422116, "step": 48600}, {"loss": 0.4888, "grad_norm": 1.1366082429885864, "learning_rate": 0.0002, "epoch": 7.858701802602861, "step": 48610}, {"loss": 0.4909, "grad_norm": 1.4737876653671265, "learning_rate": 0.0002, "epoch": 7.860318486783607, "step": 48620}, {"loss": 0.4476, "grad_norm": 1.3934144973754883, "learning_rate": 0.0002, "epoch": 7.861935170964352, "step": 48630}, {"loss": 0.5141, "grad_norm": 0.9997506737709045, "learning_rate": 0.0002, "epoch": 7.863551855145097, "step": 48640}, {"loss": 0.4766, "grad_norm": 1.3827011585235596, "learning_rate": 0.0002, "epoch": 7.865168539325842, "step": 48650}, {"loss": 0.4926, "grad_norm": 1.2811808586120605, "learning_rate": 0.0002, "epoch": 7.866785223506588, "step": 48660}, {"loss": 0.4898, "grad_norm": 1.394400715827942, "learning_rate": 0.0002, "epoch": 7.868401907687334, "step": 48670}, {"loss": 0.4889, "grad_norm": 1.5635628700256348, "learning_rate": 0.0002, "epoch": 7.870018591868079, "step": 48680}, {"loss": 0.4822, "grad_norm": 1.147349238395691, "learning_rate": 0.0002, "epoch": 7.871635276048824, "step": 48690}, {"loss": 0.485, "grad_norm": 1.2417502403259277, "learning_rate": 0.0002, "epoch": 7.873251960229569, "step": 48700}, {"loss": 0.4976, "grad_norm": 1.0380291938781738, "learning_rate": 0.0002, "epoch": 7.874868644410315, "step": 48710}, {"loss": 0.5119, "grad_norm": 1.2139482498168945, "learning_rate": 0.0002, "epoch": 7.87648532859106, "step": 48720}, {"loss": 0.497, "grad_norm": 1.2833739519119263, "learning_rate": 0.0002, "epoch": 7.878102012771805, "step": 48730}, {"loss": 0.4776, "grad_norm": 1.2405574321746826, "learning_rate": 0.0002, "epoch": 7.87971869695255, "step": 48740}, {"loss": 0.5085, "grad_norm": 1.1267465353012085, "learning_rate": 0.0002, "epoch": 7.881335381133296, "step": 48750}, {"loss": 0.4894, "grad_norm": 1.3052713871002197, "learning_rate": 0.0002, "epoch": 7.882952065314041, "step": 48760}, {"loss": 0.5077, "grad_norm": 1.0581550598144531, "learning_rate": 0.0002, "epoch": 7.884568749494786, "step": 48770}, {"loss": 0.4598, "grad_norm": 1.1074683666229248, "learning_rate": 0.0002, "epoch": 7.886185433675531, "step": 48780}, {"loss": 0.4813, "grad_norm": 1.0812418460845947, "learning_rate": 0.0002, "epoch": 7.8878021178562765, "step": 48790}, {"loss": 0.477, "grad_norm": 1.3083902597427368, "learning_rate": 0.0002, "epoch": 7.889418802037022, "step": 48800}, {"loss": 0.4717, "grad_norm": 1.457373023033142, "learning_rate": 0.0002, "epoch": 7.891035486217767, "step": 48810}, {"loss": 0.4607, "grad_norm": 1.048091173171997, "learning_rate": 0.0002, "epoch": 7.892652170398513, "step": 48820}, {"loss": 0.4848, "grad_norm": 1.1420985460281372, "learning_rate": 0.0002, "epoch": 7.894268854579258, "step": 48830}, {"loss": 0.5019, "grad_norm": 1.0286061763763428, "learning_rate": 0.0002, "epoch": 7.8958855387600035, "step": 48840}, {"loss": 0.5054, "grad_norm": 1.0361840724945068, "learning_rate": 0.0002, "epoch": 7.897502222940749, "step": 48850}, {"loss": 0.4908, "grad_norm": 1.1862726211547852, "learning_rate": 0.0002, "epoch": 7.899118907121494, "step": 48860}, {"loss": 0.4491, "grad_norm": 1.2256416082382202, "learning_rate": 0.0002, "epoch": 7.900735591302239, "step": 48870}, {"loss": 0.4732, "grad_norm": 1.0664557218551636, "learning_rate": 0.0002, "epoch": 7.9023522754829845, "step": 48880}, {"loss": 0.4741, "grad_norm": 1.3960802555084229, "learning_rate": 0.0002, "epoch": 7.90396895966373, "step": 48890}, {"loss": 0.5061, "grad_norm": 1.230430245399475, "learning_rate": 0.0002, "epoch": 7.905585643844475, "step": 48900}, {"loss": 0.4698, "grad_norm": 1.0949305295944214, "learning_rate": 0.0002, "epoch": 7.90720232802522, "step": 48910}, {"loss": 0.4964, "grad_norm": 1.4402074813842773, "learning_rate": 0.0002, "epoch": 7.908819012205965, "step": 48920}, {"loss": 0.5057, "grad_norm": 1.1064879894256592, "learning_rate": 0.0002, "epoch": 7.910435696386711, "step": 48930}, {"loss": 0.482, "grad_norm": 0.9874461889266968, "learning_rate": 0.0002, "epoch": 7.912052380567456, "step": 48940}, {"loss": 0.4851, "grad_norm": 1.2584952116012573, "learning_rate": 0.0002, "epoch": 7.913669064748201, "step": 48950}, {"loss": 0.4744, "grad_norm": 1.3016353845596313, "learning_rate": 0.0002, "epoch": 7.915285748928946, "step": 48960}, {"loss": 0.4734, "grad_norm": 1.104179859161377, "learning_rate": 0.0002, "epoch": 7.916902433109692, "step": 48970}, {"loss": 0.5143, "grad_norm": 1.26803457736969, "learning_rate": 0.0002, "epoch": 7.918519117290438, "step": 48980}, {"loss": 0.4808, "grad_norm": 1.0336869955062866, "learning_rate": 0.0002, "epoch": 7.920135801471183, "step": 48990}, {"loss": 0.4938, "grad_norm": 1.0630918741226196, "learning_rate": 0.0002, "epoch": 7.921752485651928, "step": 49000}, {"loss": 0.4988, "grad_norm": 1.2257622480392456, "learning_rate": 0.0002, "epoch": 7.923369169832673, "step": 49010}, {"loss": 0.5116, "grad_norm": 1.1722705364227295, "learning_rate": 0.0002, "epoch": 7.924985854013419, "step": 49020}, {"loss": 0.4737, "grad_norm": 1.4473323822021484, "learning_rate": 0.0002, "epoch": 7.926602538194164, "step": 49030}, {"loss": 0.5412, "grad_norm": 1.3780192136764526, "learning_rate": 0.0002, "epoch": 7.928219222374909, "step": 49040}, {"loss": 0.4849, "grad_norm": 1.253423810005188, "learning_rate": 0.0002, "epoch": 7.929835906555654, "step": 49050}, {"loss": 0.5103, "grad_norm": 1.1733828783035278, "learning_rate": 0.0002, "epoch": 7.9314525907363995, "step": 49060}, {"loss": 0.4967, "grad_norm": 1.249990701675415, "learning_rate": 0.0002, "epoch": 7.933069274917145, "step": 49070}, {"loss": 0.51, "grad_norm": 1.4012458324432373, "learning_rate": 0.0002, "epoch": 7.93468595909789, "step": 49080}, {"loss": 0.4785, "grad_norm": 1.268652319908142, "learning_rate": 0.0002, "epoch": 7.936302643278635, "step": 49090}, {"loss": 0.5319, "grad_norm": 1.0469073057174683, "learning_rate": 0.0002, "epoch": 7.9379193274593804, "step": 49100}, {"loss": 0.4836, "grad_norm": 1.3028813600540161, "learning_rate": 0.0002, "epoch": 7.939536011640126, "step": 49110}, {"loss": 0.4791, "grad_norm": 1.0998128652572632, "learning_rate": 0.0002, "epoch": 7.941152695820872, "step": 49120}, {"loss": 0.5022, "grad_norm": 1.300884485244751, "learning_rate": 0.0002, "epoch": 7.942769380001617, "step": 49130}, {"loss": 0.5193, "grad_norm": 1.257865071296692, "learning_rate": 0.0002, "epoch": 7.944386064182362, "step": 49140}, {"loss": 0.4755, "grad_norm": 1.074731707572937, "learning_rate": 0.0002, "epoch": 7.9460027483631075, "step": 49150}, {"loss": 0.4675, "grad_norm": 1.1055876016616821, "learning_rate": 0.0002, "epoch": 7.947619432543853, "step": 49160}, {"loss": 0.4801, "grad_norm": 1.1986541748046875, "learning_rate": 0.0002, "epoch": 7.949236116724598, "step": 49170}, {"loss": 0.488, "grad_norm": 1.094555139541626, "learning_rate": 0.0002, "epoch": 7.950852800905343, "step": 49180}, {"loss": 0.4974, "grad_norm": 1.2922005653381348, "learning_rate": 0.0002, "epoch": 7.952469485086088, "step": 49190}, {"loss": 0.4973, "grad_norm": 1.1557104587554932, "learning_rate": 0.0002, "epoch": 7.954086169266834, "step": 49200}, {"loss": 0.4806, "grad_norm": 1.2414908409118652, "learning_rate": 0.0002, "epoch": 7.955702853447579, "step": 49210}, {"loss": 0.4848, "grad_norm": 1.3606830835342407, "learning_rate": 0.0002, "epoch": 7.957319537628324, "step": 49220}, {"loss": 0.4981, "grad_norm": 0.9592481851577759, "learning_rate": 0.0002, "epoch": 7.958936221809069, "step": 49230}, {"loss": 0.4731, "grad_norm": 1.2130779027938843, "learning_rate": 0.0002, "epoch": 7.960552905989815, "step": 49240}, {"loss": 0.4529, "grad_norm": 1.1078767776489258, "learning_rate": 0.0002, "epoch": 7.96216959017056, "step": 49250}, {"loss": 0.4983, "grad_norm": 1.0684230327606201, "learning_rate": 0.0002, "epoch": 7.963786274351305, "step": 49260}, {"loss": 0.4832, "grad_norm": 1.1368396282196045, "learning_rate": 0.0002, "epoch": 7.965402958532051, "step": 49270}, {"loss": 0.5226, "grad_norm": 1.2161095142364502, "learning_rate": 0.0002, "epoch": 7.967019642712796, "step": 49280}, {"loss": 0.4938, "grad_norm": 1.2087634801864624, "learning_rate": 0.0002, "epoch": 7.968636326893542, "step": 49290}, {"loss": 0.4969, "grad_norm": 1.1078447103500366, "learning_rate": 0.0002, "epoch": 7.970253011074287, "step": 49300}, {"loss": 0.5333, "grad_norm": 1.3378221988677979, "learning_rate": 0.0002, "epoch": 7.971869695255032, "step": 49310}, {"loss": 0.4736, "grad_norm": 1.0475801229476929, "learning_rate": 0.0002, "epoch": 7.973486379435777, "step": 49320}, {"loss": 0.4515, "grad_norm": 0.9948194622993469, "learning_rate": 0.0002, "epoch": 7.9751030636165225, "step": 49330}, {"loss": 0.4685, "grad_norm": 1.06312894821167, "learning_rate": 0.0002, "epoch": 7.976719747797268, "step": 49340}, {"loss": 0.453, "grad_norm": 1.4047085046768188, "learning_rate": 0.0002, "epoch": 7.978336431978013, "step": 49350}, {"loss": 0.5054, "grad_norm": 1.086578130722046, "learning_rate": 0.0002, "epoch": 7.979953116158758, "step": 49360}, {"loss": 0.5024, "grad_norm": 1.2896746397018433, "learning_rate": 0.0002, "epoch": 7.9815698003395035, "step": 49370}, {"loss": 0.5102, "grad_norm": 1.260717511177063, "learning_rate": 0.0002, "epoch": 7.983186484520249, "step": 49380}, {"loss": 0.4836, "grad_norm": 1.4238426685333252, "learning_rate": 0.0002, "epoch": 7.984803168700994, "step": 49390}, {"loss": 0.4797, "grad_norm": 1.1800259351730347, "learning_rate": 0.0002, "epoch": 7.986419852881739, "step": 49400}, {"loss": 0.4911, "grad_norm": 1.128868579864502, "learning_rate": 0.0002, "epoch": 7.988036537062484, "step": 49410}, {"loss": 0.4674, "grad_norm": 1.1832106113433838, "learning_rate": 0.0002, "epoch": 7.9896532212432305, "step": 49420}, {"loss": 0.5016, "grad_norm": 1.1728334426879883, "learning_rate": 0.0002, "epoch": 7.991269905423976, "step": 49430}, {"loss": 0.4706, "grad_norm": 1.2403929233551025, "learning_rate": 0.0002, "epoch": 7.992886589604721, "step": 49440}, {"loss": 0.4989, "grad_norm": 1.245354175567627, "learning_rate": 0.0002, "epoch": 7.994503273785466, "step": 49450}, {"loss": 0.5088, "grad_norm": 1.3526462316513062, "learning_rate": 0.0002, "epoch": 7.9961199579662114, "step": 49460}, {"loss": 0.4703, "grad_norm": 1.2117315530776978, "learning_rate": 0.0002, "epoch": 7.997736642146957, "step": 49470}, {"loss": 0.4802, "grad_norm": 1.0393620729446411, "learning_rate": 0.0002, "epoch": 7.999353326327702, "step": 49480}]}