add checkpoint cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1

.gitattributes

@@ -797,3 +797,6 @@ checkpoints/cat_qwen25_7b_r8_a32_sgd_e2_lr1e-4_s1_perlayer_rot_r42/tokenizer.jso
 checkpoints/cat_qwen25_7b_r8_a32_7m_T0_bot10_e2_lr1e-4_s3/checkpoint-1046/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/cat_qwen25_7b_r8_a32_7m_T0_bot10_e2_lr1e-4_s3/checkpoint-2092/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/cat_qwen25_7b_r8_a32_7m_T0_bot10_e2_lr1e-4_s3/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/tokenizer.json filter=lfs diff=lfs merge=lfs -text

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/README.md

@@ -0,0 +1,61 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+model_name: cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- sft
+- transformers
+- trl
+licence: license
+pipeline_tag: text-generation
+---
+
+# Model Card for cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct).
+It has been trained using [TRL](https://github.com/huggingface/trl).
+
+## Quick start
+
+```python
+from transformers import pipeline
+
+question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
+generator = pipeline("text-generation", model="None", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
+
+## Training procedure
+
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/agam-research/huggingface/runs/wxttil9y) 
+
+
+This model was trained with SFT.
+
+### Framework versions
+
+- PEFT 0.18.1
+- TRL: 0.28.0
+- Transformers: 4.57.6
+- Pytorch: 2.10.0
+- Datasets: 4.5.0
+- Tokenizers: 0.22.2
+
+## Citations
+
+
+
+Cite TRL as:
+    
+```bibtex
+@software{vonwerra2020trl,
+  title   = {{TRL: Transformers Reinforcement Learning}},
+  author  = {von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin},
+  license = {Apache-2.0},
+  url     = {https://github.com/huggingface/trl},
+  year    = {2020}
+}
+```

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/adapter_config.json

@@ -0,0 +1,46 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "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_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "k_proj",
+    "down_proj",
+    "q_proj",
+    "o_proj",
+    "v_proj",
+    "gate_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a215b36d355406379ec58eb8b526fb876d504dcdfc8d83ca20618f9e1115df1
+size 80792096

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/chat_template.jinja

@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/README.md

@@ -0,0 +1,209 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- sft
+- transformers
+- trl
+---
+
+# 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 0.18.1

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/adapter_config.json

@@ -0,0 +1,46 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "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_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "k_proj",
+    "down_proj",
+    "q_proj",
+    "o_proj",
+    "v_proj",
+    "gate_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2c42d88767ea4de9330d303b99320f509ed783777b14da9a92a3b868e665457
+size 80792096

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/chat_template.jinja

@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/tokenizer_config.json

@@ -0,0 +1,207 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/trainer_state.json

@@ -0,0 +1,1074 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1046,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.1725262880325318,
+      "epoch": 0.009560229445506692,
+      "grad_norm": 1.7949155569076538,
+      "learning_rate": 8.571428571428573e-06,
+      "loss": 0.2754,
+      "mean_token_accuracy": 0.9112902641296386,
+      "num_tokens": 14125.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.1808424949645997,
+      "epoch": 0.019120458891013385,
+      "grad_norm": 1.8090709447860718,
+      "learning_rate": 1.8095238095238094e-05,
+      "loss": 0.2368,
+      "mean_token_accuracy": 0.9221769034862518,
+      "num_tokens": 27004.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.1529677510261536,
+      "epoch": 0.028680688336520075,
+      "grad_norm": 2.091123104095459,
+      "learning_rate": 2.7619047619047622e-05,
+      "loss": 0.1891,
+      "mean_token_accuracy": 0.9356727421283721,
+      "num_tokens": 41395.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.0867261052131654,
+      "epoch": 0.03824091778202677,
+      "grad_norm": 2.752497911453247,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": 0.1659,
+      "mean_token_accuracy": 0.9372419357299805,
+      "num_tokens": 55233.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.0402584671974182,
+      "epoch": 0.04780114722753346,
+      "grad_norm": 3.738189220428467,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.1778,
+      "mean_token_accuracy": 0.934120100736618,
+      "num_tokens": 68341.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.0378250241279603,
+      "epoch": 0.05736137667304015,
+      "grad_norm": 2.169098377227783,
+      "learning_rate": 5.619047619047619e-05,
+      "loss": 0.1723,
+      "mean_token_accuracy": 0.9337703168392182,
+      "num_tokens": 81730.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.0203807950019836,
+      "epoch": 0.06692160611854685,
+      "grad_norm": 2.4451217651367188,
+      "learning_rate": 6.571428571428571e-05,
+      "loss": 0.1639,
+      "mean_token_accuracy": 0.9419153094291687,
+      "num_tokens": 94886.0,
+      "step": 70
+    },
+    {
+      "entropy": 0.9988491177558899,
+      "epoch": 0.07648183556405354,
+      "grad_norm": 2.444343328475952,
+      "learning_rate": 7.523809523809524e-05,
+      "loss": 0.1839,
+      "mean_token_accuracy": 0.9345499813556671,
+      "num_tokens": 108569.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.0041434288024902,
+      "epoch": 0.08604206500956023,
+      "grad_norm": 1.929736614227295,
+      "learning_rate": 8.476190476190477e-05,
+      "loss": 0.1645,
+      "mean_token_accuracy": 0.9391969263553619,
+      "num_tokens": 122156.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.0116430222988129,
+      "epoch": 0.09560229445506692,
+      "grad_norm": 2.168968915939331,
+      "learning_rate": 9.428571428571429e-05,
+      "loss": 0.1929,
+      "mean_token_accuracy": 0.9271840214729309,
+      "num_tokens": 135290.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.0483016848564148,
+      "epoch": 0.10516252390057361,
+      "grad_norm": 2.018573760986328,
+      "learning_rate": 9.999900008621665e-05,
+      "loss": 0.2368,
+      "mean_token_accuracy": 0.9140790283679963,
+      "num_tokens": 149015.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.0560781598091125,
+      "epoch": 0.1147227533460803,
+      "grad_norm": 1.9477077722549438,
+      "learning_rate": 9.998775151544478e-05,
+      "loss": 0.313,
+      "mean_token_accuracy": 0.9046160459518433,
+      "num_tokens": 162147.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.0339066147804261,
+      "epoch": 0.124282982791587,
+      "grad_norm": 3.953827142715454,
+      "learning_rate": 9.996400730289628e-05,
+      "loss": 0.2581,
+      "mean_token_accuracy": 0.9119772851467133,
+      "num_tokens": 175436.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.0594236373901367,
+      "epoch": 0.1338432122370937,
+      "grad_norm": 2.7890758514404297,
+      "learning_rate": 9.992777338400864e-05,
+      "loss": 0.2533,
+      "mean_token_accuracy": 0.9100077152252197,
+      "num_tokens": 188834.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.053187495470047,
+      "epoch": 0.14340344168260039,
+      "grad_norm": 2.2614591121673584,
+      "learning_rate": 9.987905881632213e-05,
+      "loss": 0.289,
+      "mean_token_accuracy": 0.899867707490921,
+      "num_tokens": 202708.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.0322751343250274,
+      "epoch": 0.15296367112810708,
+      "grad_norm": 2.2198617458343506,
+      "learning_rate": 9.981787577721553e-05,
+      "loss": 0.2234,
+      "mean_token_accuracy": 0.9163801074028015,
+      "num_tokens": 216384.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.0143227398395538,
+      "epoch": 0.16252390057361377,
+      "grad_norm": 2.215290069580078,
+      "learning_rate": 9.974423956086228e-05,
+      "loss": 0.2049,
+      "mean_token_accuracy": 0.9241275012493133,
+      "num_tokens": 229105.0,
+      "step": 170
+    },
+    {
+      "entropy": 0.9944452047348022,
+      "epoch": 0.17208413001912046,
+      "grad_norm": 3.2736923694610596,
+      "learning_rate": 9.965816857440717e-05,
+      "loss": 0.209,
+      "mean_token_accuracy": 0.919926643371582,
+      "num_tokens": 242264.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.015661782026291,
+      "epoch": 0.18164435946462715,
+      "grad_norm": 2.2948694229125977,
+      "learning_rate": 9.955968433336515e-05,
+      "loss": 0.2268,
+      "mean_token_accuracy": 0.9108724594116211,
+      "num_tokens": 255602.0,
+      "step": 190
+    },
+    {
+      "entropy": 0.9878140687942505,
+      "epoch": 0.19120458891013384,
+      "grad_norm": 1.7582422494888306,
+      "learning_rate": 9.944881145624291e-05,
+      "loss": 0.1732,
+      "mean_token_accuracy": 0.9377074241638184,
+      "num_tokens": 269712.0,
+      "step": 200
+    },
+    {
+      "entropy": 0.9916601777076721,
+      "epoch": 0.20076481835564053,
+      "grad_norm": 2.493232011795044,
+      "learning_rate": 9.932557765838504e-05,
+      "loss": 0.1978,
+      "mean_token_accuracy": 0.9271564424037934,
+      "num_tokens": 283572.0,
+      "step": 210
+    },
+    {
+      "entropy": 0.9938843488693238,
+      "epoch": 0.21032504780114722,
+      "grad_norm": 1.9961174726486206,
+      "learning_rate": 9.919001374504569e-05,
+      "loss": 0.238,
+      "mean_token_accuracy": 0.9079000949859619,
+      "num_tokens": 297437.0,
+      "step": 220
+    },
+    {
+      "entropy": 0.9930910289287567,
+      "epoch": 0.2198852772466539,
+      "grad_norm": 4.460914611816406,
+      "learning_rate": 9.904215360368826e-05,
+      "loss": 0.2281,
+      "mean_token_accuracy": 0.9163453280925751,
+      "num_tokens": 310803.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.0042203903198241,
+      "epoch": 0.2294455066921606,
+      "grad_norm": 2.591485023498535,
+      "learning_rate": 9.888203419551432e-05,
+      "loss": 0.2399,
+      "mean_token_accuracy": 0.909199732542038,
+      "num_tokens": 324713.0,
+      "step": 240
+    },
+    {
+      "entropy": 0.9938572168350219,
+      "epoch": 0.2390057361376673,
+      "grad_norm": 2.540052890777588,
+      "learning_rate": 9.870969554622426e-05,
+      "loss": 0.2626,
+      "mean_token_accuracy": 0.9039014875888824,
+      "num_tokens": 337634.0,
+      "step": 250
+    },
+    {
+      "entropy": 0.981248390674591,
+      "epoch": 0.248565965583174,
+      "grad_norm": 2.0444109439849854,
+      "learning_rate": 9.852518073601194e-05,
+      "loss": 0.2077,
+      "mean_token_accuracy": 0.9231703639030456,
+      "num_tokens": 350459.0,
+      "step": 260
+    },
+    {
+      "entropy": 0.9621168673038483,
+      "epoch": 0.25812619502868067,
+      "grad_norm": 2.641737699508667,
+      "learning_rate": 9.832853588879575e-05,
+      "loss": 0.2273,
+      "mean_token_accuracy": 0.9158074796199799,
+      "num_tokens": 364010.0,
+      "step": 270
+    },
+    {
+      "entropy": 0.997821694612503,
+      "epoch": 0.2676864244741874,
+      "grad_norm": 2.929631233215332,
+      "learning_rate": 9.811981016068881e-05,
+      "loss": 0.2722,
+      "mean_token_accuracy": 0.9029962062835694,
+      "num_tokens": 377021.0,
+      "step": 280
+    },
+    {
+      "entropy": 0.9611131608486175,
+      "epoch": 0.27724665391969405,
+      "grad_norm": 1.1362237930297852,
+      "learning_rate": 9.78990557277112e-05,
+      "loss": 0.2092,
+      "mean_token_accuracy": 0.9216417133808136,
+      "num_tokens": 390819.0,
+      "step": 290
+    },
+    {
+      "entropy": 0.9901561737060547,
+      "epoch": 0.28680688336520077,
+      "grad_norm": 2.0782461166381836,
+      "learning_rate": 9.766632777274738e-05,
+      "loss": 0.2453,
+      "mean_token_accuracy": 0.9103418946266174,
+      "num_tokens": 404644.0,
+      "step": 300
+    },
+    {
+      "entropy": 0.9956516146659851,
+      "epoch": 0.29636711281070743,
+      "grad_norm": 2.935779571533203,
+      "learning_rate": 9.742168447175182e-05,
+      "loss": 0.2293,
+      "mean_token_accuracy": 0.9149252951145173,
+      "num_tokens": 418529.0,
+      "step": 310
+    },
+    {
+      "entropy": 0.9521589517593384,
+      "epoch": 0.30592734225621415,
+      "grad_norm": 1.9554321765899658,
+      "learning_rate": 9.716518697920655e-05,
+      "loss": 0.2028,
+      "mean_token_accuracy": 0.92679563164711,
+      "num_tokens": 431509.0,
+      "step": 320
+    },
+    {
+      "entropy": 0.9850484609603882,
+      "epoch": 0.3154875717017208,
+      "grad_norm": 2.1091372966766357,
+      "learning_rate": 9.689689941283417e-05,
+      "loss": 0.2399,
+      "mean_token_accuracy": 0.9111748278141022,
+      "num_tokens": 445040.0,
+      "step": 330
+    },
+    {
+      "entropy": 0.9910066843032836,
+      "epoch": 0.32504780114722753,
+      "grad_norm": 2.729727268218994,
+      "learning_rate": 9.661688883756996e-05,
+      "loss": 0.2222,
+      "mean_token_accuracy": 0.9158975481987,
+      "num_tokens": 458540.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.0245653331279754,
+      "epoch": 0.33460803059273425,
+      "grad_norm": 3.326910972595215,
+      "learning_rate": 9.632522524879756e-05,
+      "loss": 0.2471,
+      "mean_token_accuracy": 0.9084886729717254,
+      "num_tokens": 471556.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.0443902254104613,
+      "epoch": 0.3441682600382409,
+      "grad_norm": 2.765958309173584,
+      "learning_rate": 9.602198155485176e-05,
+      "loss": 0.2307,
+      "mean_token_accuracy": 0.9184524953365326,
+      "num_tokens": 484786.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.055874276161194,
+      "epoch": 0.35372848948374763,
+      "grad_norm": 2.167161226272583,
+      "learning_rate": 9.570723355879342e-05,
+      "loss": 0.2102,
+      "mean_token_accuracy": 0.9206913709640503,
+      "num_tokens": 498352.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.0401476562023162,
+      "epoch": 0.3632887189292543,
+      "grad_norm": 2.807732343673706,
+      "learning_rate": 9.538105993946068e-05,
+      "loss": 0.2307,
+      "mean_token_accuracy": 0.9132277011871338,
+      "num_tokens": 512064.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.0409152030944824,
+      "epoch": 0.372848948374761,
+      "grad_norm": 1.8228720426559448,
+      "learning_rate": 9.504354223180118e-05,
+      "loss": 0.2182,
+      "mean_token_accuracy": 0.9216055631637573,
+      "num_tokens": 525582.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.0548848509788513,
+      "epoch": 0.3824091778202677,
+      "grad_norm": 3.267235279083252,
+      "learning_rate": 9.469476480649056e-05,
+      "loss": 0.2297,
+      "mean_token_accuracy": 0.9145385086536407,
+      "num_tokens": 539561.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.0603472709655761,
+      "epoch": 0.3919694072657744,
+      "grad_norm": 1.9201253652572632,
+      "learning_rate": 9.433481484884189e-05,
+      "loss": 0.2186,
+      "mean_token_accuracy": 0.9178025484085083,
+      "num_tokens": 553625.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.0538666605949403,
+      "epoch": 0.40152963671128106,
+      "grad_norm": 2.342041492462158,
+      "learning_rate": 9.396378233701167e-05,
+      "loss": 0.1925,
+      "mean_token_accuracy": 0.9293496191501618,
+      "num_tokens": 565891.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.0616689920425415,
+      "epoch": 0.4110898661567878,
+      "grad_norm": 1.9766606092453003,
+      "learning_rate": 9.358176001950759e-05,
+      "loss": 0.2311,
+      "mean_token_accuracy": 0.9161762118339538,
+      "num_tokens": 578722.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.0873818516731262,
+      "epoch": 0.42065009560229444,
+      "grad_norm": 2.961918830871582,
+      "learning_rate": 9.318884339200378e-05,
+      "loss": 0.238,
+      "mean_token_accuracy": 0.9079842686653137,
+      "num_tokens": 592646.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.0649721384048463,
+      "epoch": 0.43021032504780116,
+      "grad_norm": 2.5015785694122314,
+      "learning_rate": 9.278513067346941e-05,
+      "loss": 0.1898,
+      "mean_token_accuracy": 0.9307917773723602,
+      "num_tokens": 606429.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.0712341904640197,
+      "epoch": 0.4397705544933078,
+      "grad_norm": 2.5987443923950195,
+      "learning_rate": 9.237072278161652e-05,
+      "loss": 0.2433,
+      "mean_token_accuracy": 0.9096251487731933,
+      "num_tokens": 619626.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.0651772737503051,
+      "epoch": 0.44933078393881454,
+      "grad_norm": 2.339188814163208,
+      "learning_rate": 9.194572330767314e-05,
+      "loss": 0.2105,
+      "mean_token_accuracy": 0.9202918171882629,
+      "num_tokens": 633055.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.0719932675361634,
+      "epoch": 0.4588910133843212,
+      "grad_norm": 2.139345169067383,
+      "learning_rate": 9.151023849048818e-05,
+      "loss": 0.1972,
+      "mean_token_accuracy": 0.9251638293266297,
+      "num_tokens": 647155.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.0447640657424926,
+      "epoch": 0.4684512428298279,
+      "grad_norm": 3.1209723949432373,
+      "learning_rate": 9.106437718997452e-05,
+      "loss": 0.2031,
+      "mean_token_accuracy": 0.921468985080719,
+      "num_tokens": 660695.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.0626697421073914,
+      "epoch": 0.4780114722753346,
+      "grad_norm": 1.8369485139846802,
+      "learning_rate": 9.060825085989682e-05,
+      "loss": 0.2105,
+      "mean_token_accuracy": 0.9239315092563629,
+      "num_tokens": 674509.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.0570692539215087,
+      "epoch": 0.4875717017208413,
+      "grad_norm": 3.0195302963256836,
+      "learning_rate": 9.014197352001097e-05,
+      "loss": 0.1821,
+      "mean_token_accuracy": 0.9317115485668183,
+      "num_tokens": 687644.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.0473023772239685,
+      "epoch": 0.497131931166348,
+      "grad_norm": 2.375627279281616,
+      "learning_rate": 8.966566172756213e-05,
+      "loss": 0.2201,
+      "mean_token_accuracy": 0.9189318299293519,
+      "num_tokens": 700658.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.0765699386596679,
+      "epoch": 0.5066921606118547,
+      "grad_norm": 2.8174030780792236,
+      "learning_rate": 8.91794345481484e-05,
+      "loss": 0.208,
+      "mean_token_accuracy": 0.91789231300354,
+      "num_tokens": 714234.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.0702912211418152,
+      "epoch": 0.5162523900573613,
+      "grad_norm": 3.2318389415740967,
+      "learning_rate": 8.868341352595755e-05,
+      "loss": 0.2295,
+      "mean_token_accuracy": 0.9122642040252685,
+      "num_tokens": 727721.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.0573222517967225,
+      "epoch": 0.5258126195028681,
+      "grad_norm": 1.7950525283813477,
+      "learning_rate": 8.817772265338409e-05,
+      "loss": 0.223,
+      "mean_token_accuracy": 0.9139611005783081,
+      "num_tokens": 741707.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.0547808468341828,
+      "epoch": 0.5353728489483748,
+      "grad_norm": 3.6276676654815674,
+      "learning_rate": 8.766248834003441e-05,
+      "loss": 0.2363,
+      "mean_token_accuracy": 0.9099368512630462,
+      "num_tokens": 755385.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.0302533328533172,
+      "epoch": 0.5449330783938815,
+      "grad_norm": 2.1893699169158936,
+      "learning_rate": 8.713783938112768e-05,
+      "loss": 0.199,
+      "mean_token_accuracy": 0.926015430688858,
+      "num_tokens": 768687.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.0444139957427978,
+      "epoch": 0.5544933078393881,
+      "grad_norm": 2.8716442584991455,
+      "learning_rate": 8.660390692530026e-05,
+      "loss": 0.1995,
+      "mean_token_accuracy": 0.9259741365909576,
+      "num_tokens": 782172.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.0370828688144684,
+      "epoch": 0.5640535372848948,
+      "grad_norm": 1.763108730316162,
+      "learning_rate": 8.606082444182216e-05,
+      "loss": 0.2055,
+      "mean_token_accuracy": 0.9247264742851258,
+      "num_tokens": 795561.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.0649680018424987,
+      "epoch": 0.5736137667304015,
+      "grad_norm": 1.66559636592865,
+      "learning_rate": 8.550872768723292e-05,
+      "loss": 0.1828,
+      "mean_token_accuracy": 0.9340787351131439,
+      "num_tokens": 809436.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.0512043595314027,
+      "epoch": 0.5831739961759083,
+      "grad_norm": 4.193369388580322,
+      "learning_rate": 8.494775467140627e-05,
+      "loss": 0.2083,
+      "mean_token_accuracy": 0.9245233416557312,
+      "num_tokens": 822983.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.0559794306755066,
+      "epoch": 0.5927342256214149,
+      "grad_norm": 1.8999974727630615,
+      "learning_rate": 8.4378045623051e-05,
+      "loss": 0.2028,
+      "mean_token_accuracy": 0.9243997693061828,
+      "num_tokens": 836766.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.0672565937042235,
+      "epoch": 0.6022944550669216,
+      "grad_norm": 1.607515811920166,
+      "learning_rate": 8.379974295465758e-05,
+      "loss": 0.2156,
+      "mean_token_accuracy": 0.916247296333313,
+      "num_tokens": 850200.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.0445150911808014,
+      "epoch": 0.6118546845124283,
+      "grad_norm": 1.7760804891586304,
+      "learning_rate": 8.321299122689863e-05,
+      "loss": 0.2276,
+      "mean_token_accuracy": 0.9184480905532837,
+      "num_tokens": 864001.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.0459101319313049,
+      "epoch": 0.621414913957935,
+      "grad_norm": 2.6341512203216553,
+      "learning_rate": 8.261793711249254e-05,
+      "loss": 0.2191,
+      "mean_token_accuracy": 0.9138937532901764,
+      "num_tokens": 878135.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.0317402541637422,
+      "epoch": 0.6309751434034416,
+      "grad_norm": 1.8982086181640625,
+      "learning_rate": 8.201472935953916e-05,
+      "loss": 0.2169,
+      "mean_token_accuracy": 0.9217910826206207,
+      "num_tokens": 891307.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.045106440782547,
+      "epoch": 0.6405353728489483,
+      "grad_norm": 1.7897950410842896,
+      "learning_rate": 8.140351875433649e-05,
+      "loss": 0.2075,
+      "mean_token_accuracy": 0.9199797868728637,
+      "num_tokens": 904829.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.0383180022239684,
+      "epoch": 0.6500956022944551,
+      "grad_norm": 1.8748987913131714,
+      "learning_rate": 8.07844580836881e-05,
+      "loss": 0.1829,
+      "mean_token_accuracy": 0.9320280909538269,
+      "num_tokens": 919199.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.0720835208892823,
+      "epoch": 0.6596558317399618,
+      "grad_norm": 1.7298226356506348,
+      "learning_rate": 8.015770209671044e-05,
+      "loss": 0.2247,
+      "mean_token_accuracy": 0.9139919757843018,
+      "num_tokens": 932439.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.0737423539161681,
+      "epoch": 0.6692160611854685,
+      "grad_norm": 2.094554901123047,
+      "learning_rate": 7.952340746614946e-05,
+      "loss": 0.2105,
+      "mean_token_accuracy": 0.921952760219574,
+      "num_tokens": 945776.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.0488283574581145,
+      "epoch": 0.6787762906309751,
+      "grad_norm": 2.220661163330078,
+      "learning_rate": 7.888173274921655e-05,
+      "loss": 0.1762,
+      "mean_token_accuracy": 0.9325842142105103,
+      "num_tokens": 959352.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.0624409198760987,
+      "epoch": 0.6883365200764818,
+      "grad_norm": 2.789525032043457,
+      "learning_rate": 7.823283834795336e-05,
+      "loss": 0.1965,
+      "mean_token_accuracy": 0.9271815240383148,
+      "num_tokens": 972596.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.0450023353099822,
+      "epoch": 0.6978967495219885,
+      "grad_norm": 1.9600374698638916,
+      "learning_rate": 7.75768864691354e-05,
+      "loss": 0.2181,
+      "mean_token_accuracy": 0.9186623334884644,
+      "num_tokens": 985699.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.050288939476013,
+      "epoch": 0.7074569789674953,
+      "grad_norm": 2.745725154876709,
+      "learning_rate": 7.691404108372461e-05,
+      "loss": 0.1958,
+      "mean_token_accuracy": 0.9252684831619262,
+      "num_tokens": 999728.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.0660975933074952,
+      "epoch": 0.7170172084130019,
+      "grad_norm": 1.9985816478729248,
+      "learning_rate": 7.624446788588085e-05,
+      "loss": 0.1743,
+      "mean_token_accuracy": 0.934730589389801,
+      "num_tokens": 1013669.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.0661585927009583,
+      "epoch": 0.7265774378585086,
+      "grad_norm": 1.8102953433990479,
+      "learning_rate": 7.556833425154255e-05,
+      "loss": 0.1806,
+      "mean_token_accuracy": 0.9322760343551636,
+      "num_tokens": 1027609.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.0598073244094848,
+      "epoch": 0.7361376673040153,
+      "grad_norm": 2.5633928775787354,
+      "learning_rate": 7.488580919658727e-05,
+      "loss": 0.1929,
+      "mean_token_accuracy": 0.9238331735134124,
+      "num_tokens": 1041046.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.04606152176857,
+      "epoch": 0.745697896749522,
+      "grad_norm": 2.917694568634033,
+      "learning_rate": 7.419706333458201e-05,
+      "loss": 0.179,
+      "mean_token_accuracy": 0.9323089599609375,
+      "num_tokens": 1055304.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.0437167465686799,
+      "epoch": 0.7552581261950286,
+      "grad_norm": 3.5574042797088623,
+      "learning_rate": 7.350226883413413e-05,
+      "loss": 0.209,
+      "mean_token_accuracy": 0.9239528596401214,
+      "num_tokens": 1068542.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.0587083101272583,
+      "epoch": 0.7648183556405354,
+      "grad_norm": 2.3629257678985596,
+      "learning_rate": 7.280159937585386e-05,
+      "loss": 0.1855,
+      "mean_token_accuracy": 0.9248183965682983,
+      "num_tokens": 1082195.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.059005331993103,
+      "epoch": 0.7743785850860421,
+      "grad_norm": 3.765178680419922,
+      "learning_rate": 7.209523010893853e-05,
+      "loss": 0.1667,
+      "mean_token_accuracy": 0.9390020132064819,
+      "num_tokens": 1095884.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.070823109149933,
+      "epoch": 0.7839388145315488,
+      "grad_norm": 1.7191312313079834,
+      "learning_rate": 7.138333760738987e-05,
+      "loss": 0.2056,
+      "mean_token_accuracy": 0.9238715708255768,
+      "num_tokens": 1109287.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.070789921283722,
+      "epoch": 0.7934990439770554,
+      "grad_norm": 2.819929361343384,
+      "learning_rate": 7.066609982587501e-05,
+      "loss": 0.1821,
+      "mean_token_accuracy": 0.9314943969249725,
+      "num_tokens": 1123392.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.0708677768707275,
+      "epoch": 0.8030592734225621,
+      "grad_norm": 2.1845016479492188,
+      "learning_rate": 6.994369605524253e-05,
+      "loss": 0.1971,
+      "mean_token_accuracy": 0.9240113377571106,
+      "num_tokens": 1137691.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.03220654129982,
+      "epoch": 0.8126195028680688,
+      "grad_norm": 3.272754430770874,
+      "learning_rate": 6.921630687770427e-05,
+      "loss": 0.186,
+      "mean_token_accuracy": 0.9262190937995911,
+      "num_tokens": 1151400.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.043507421016693,
+      "epoch": 0.8221797323135756,
+      "grad_norm": 2.8006186485290527,
+      "learning_rate": 6.848411412169461e-05,
+      "loss": 0.1591,
+      "mean_token_accuracy": 0.9413301229476929,
+      "num_tokens": 1165074.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.0422970294952392,
+      "epoch": 0.8317399617590823,
+      "grad_norm": 2.7493834495544434,
+      "learning_rate": 6.774730081641799e-05,
+      "loss": 0.181,
+      "mean_token_accuracy": 0.9326336324214936,
+      "num_tokens": 1178919.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.0501964330673217,
+      "epoch": 0.8413001912045889,
+      "grad_norm": 3.106358051300049,
+      "learning_rate": 6.70060511460963e-05,
+      "loss": 0.2042,
+      "mean_token_accuracy": 0.9240445792675018,
+      "num_tokens": 1192204.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.0335909485816956,
+      "epoch": 0.8508604206500956,
+      "grad_norm": 2.1840972900390625,
+      "learning_rate": 6.626055040392774e-05,
+      "loss": 0.1897,
+      "mean_token_accuracy": 0.9266136407852172,
+      "num_tokens": 1205571.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.0435899496078491,
+      "epoch": 0.8604206500956023,
+      "grad_norm": 2.261042594909668,
+      "learning_rate": 6.551098494576819e-05,
+      "loss": 0.2012,
+      "mean_token_accuracy": 0.9291186094284057,
+      "num_tokens": 1219043.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.0415594696998596,
+      "epoch": 0.869980879541109,
+      "grad_norm": 2.32177996635437,
+      "learning_rate": 6.475754214354715e-05,
+      "loss": 0.1749,
+      "mean_token_accuracy": 0.9318947196006775,
+      "num_tokens": 1233113.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.0231156051158905,
+      "epoch": 0.8795411089866156,
+      "grad_norm": 2.1848862171173096,
+      "learning_rate": 6.400041033842965e-05,
+      "loss": 0.1713,
+      "mean_token_accuracy": 0.9378451108932495,
+      "num_tokens": 1245931.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.0183886528015136,
+      "epoch": 0.8891013384321224,
+      "grad_norm": 2.509147882461548,
+      "learning_rate": 6.323977879373571e-05,
+      "loss": 0.1798,
+      "mean_token_accuracy": 0.9298445463180542,
+      "num_tokens": 1259949.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.0446486234664918,
+      "epoch": 0.8986615678776291,
+      "grad_norm": 2.408195972442627,
+      "learning_rate": 6.247583764762959e-05,
+      "loss": 0.1731,
+      "mean_token_accuracy": 0.9333014845848083,
+      "num_tokens": 1273413.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.0374162077903748,
+      "epoch": 0.9082217973231358,
+      "grad_norm": 1.9992507696151733,
+      "learning_rate": 6.170877786559001e-05,
+      "loss": 0.1761,
+      "mean_token_accuracy": 0.9344584167003631,
+      "num_tokens": 1286343.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.021628212928772,
+      "epoch": 0.9177820267686424,
+      "grad_norm": 2.594414472579956,
+      "learning_rate": 6.0938791192673895e-05,
+      "loss": 0.1778,
+      "mean_token_accuracy": 0.9287881553173065,
+      "num_tokens": 1300089.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.0166936099529267,
+      "epoch": 0.9273422562141491,
+      "grad_norm": 3.2213165760040283,
+      "learning_rate": 6.016607010558492e-05,
+      "loss": 0.1648,
+      "mean_token_accuracy": 0.9389560222625732,
+      "num_tokens": 1313412.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.0378799736499786,
+      "epoch": 0.9369024856596558,
+      "grad_norm": 2.705141544342041,
+      "learning_rate": 5.939080776455942e-05,
+      "loss": 0.1852,
+      "mean_token_accuracy": 0.9291444897651673,
+      "num_tokens": 1326290.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.0505308985710144,
+      "epoch": 0.9464627151051626,
+      "grad_norm": 2.3800392150878906,
+      "learning_rate": 5.861319796508139e-05,
+      "loss": 0.1669,
+      "mean_token_accuracy": 0.9396340370178222,
+      "num_tokens": 1339719.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.0232268452644349,
+      "epoch": 0.9560229445506692,
+      "grad_norm": 1.9888688325881958,
+      "learning_rate": 5.783343508943849e-05,
+      "loss": 0.1507,
+      "mean_token_accuracy": 0.9391794621944427,
+      "num_tokens": 1352423.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.0251559376716615,
+      "epoch": 0.9655831739961759,
+      "grad_norm": 2.7010385990142822,
+      "learning_rate": 5.7051714058131654e-05,
+      "loss": 0.1662,
+      "mean_token_accuracy": 0.9356134235858917,
+      "num_tokens": 1365889.0,
+      "step": 1010
+    },
+    {
+      "entropy": 1.0165675580501556,
+      "epoch": 0.9751434034416826,
+      "grad_norm": 2.578207015991211,
+      "learning_rate": 5.626823028114991e-05,
+      "loss": 0.1674,
+      "mean_token_accuracy": 0.9337117493152618,
+      "num_tokens": 1379353.0,
+      "step": 1020
+    },
+    {
+      "entropy": 1.0138891518115998,
+      "epoch": 0.9847036328871893,
+      "grad_norm": 2.096973180770874,
+      "learning_rate": 5.548317960912305e-05,
+      "loss": 0.176,
+      "mean_token_accuracy": 0.9312076151371003,
+      "num_tokens": 1392866.0,
+      "step": 1030
+    },
+    {
+      "entropy": 1.0109597086906432,
+      "epoch": 0.994263862332696,
+      "grad_norm": 2.348393678665161,
+      "learning_rate": 5.4696758284363956e-05,
+      "loss": 0.1805,
+      "mean_token_accuracy": 0.928768002986908,
+      "num_tokens": 1406263.0,
+      "step": 1040
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2092,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.015340529185075e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-1046/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13eb5bb350b15424cafd48e7e76ffb253cba270f11455232484f6e2ab004ee88
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/README.md

@@ -0,0 +1,209 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- sft
+- transformers
+- trl
+---
+
+# 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 0.18.1

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/adapter_config.json

@@ -0,0 +1,46 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "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_bias": false,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "k_proj",
+    "down_proj",
+    "q_proj",
+    "o_proj",
+    "v_proj",
+    "gate_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a215b36d355406379ec58eb8b526fb876d504dcdfc8d83ca20618f9e1115df1
+size 80792096

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/added_tokens.json

@@ -0,0 +1,24 @@
+{
+  "</tool_call>": 151658,
+  "<tool_call>": 151657,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/chat_template.jinja

@@ -0,0 +1,54 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- messages[0]['content'] }}
+    {%- else %}
+        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
+    {%- endif %}
+    {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0]['role'] == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
+    {%- else %}
+        {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- for message in messages %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
+        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {{- '<|im_start|>' + message.role }}
+        {%- if message.content %}
+            {{- '\n' + message.content }}
+        {%- endif %}
+        {%- for tool_call in message.tool_calls %}
+            {%- if tool_call.function is defined %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {{- '\n<tool_call>\n{"name": "' }}
+            {{- tool_call.name }}
+            {{- '", "arguments": ' }}
+            {{- tool_call.arguments | tojson }}
+            {{- '}\n</tool_call>' }}
+        {%- endfor %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- message.content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+{%- endif %}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/tokenizer_config.json

@@ -0,0 +1,207 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/trainer_state.json

@@ -0,0 +1,2124 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 2092,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.1725262880325318,
+      "epoch": 0.009560229445506692,
+      "grad_norm": 1.7949155569076538,
+      "learning_rate": 8.571428571428573e-06,
+      "loss": 0.2754,
+      "mean_token_accuracy": 0.9112902641296386,
+      "num_tokens": 14125.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.1808424949645997,
+      "epoch": 0.019120458891013385,
+      "grad_norm": 1.8090709447860718,
+      "learning_rate": 1.8095238095238094e-05,
+      "loss": 0.2368,
+      "mean_token_accuracy": 0.9221769034862518,
+      "num_tokens": 27004.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.1529677510261536,
+      "epoch": 0.028680688336520075,
+      "grad_norm": 2.091123104095459,
+      "learning_rate": 2.7619047619047622e-05,
+      "loss": 0.1891,
+      "mean_token_accuracy": 0.9356727421283721,
+      "num_tokens": 41395.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.0867261052131654,
+      "epoch": 0.03824091778202677,
+      "grad_norm": 2.752497911453247,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": 0.1659,
+      "mean_token_accuracy": 0.9372419357299805,
+      "num_tokens": 55233.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.0402584671974182,
+      "epoch": 0.04780114722753346,
+      "grad_norm": 3.738189220428467,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.1778,
+      "mean_token_accuracy": 0.934120100736618,
+      "num_tokens": 68341.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.0378250241279603,
+      "epoch": 0.05736137667304015,
+      "grad_norm": 2.169098377227783,
+      "learning_rate": 5.619047619047619e-05,
+      "loss": 0.1723,
+      "mean_token_accuracy": 0.9337703168392182,
+      "num_tokens": 81730.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.0203807950019836,
+      "epoch": 0.06692160611854685,
+      "grad_norm": 2.4451217651367188,
+      "learning_rate": 6.571428571428571e-05,
+      "loss": 0.1639,
+      "mean_token_accuracy": 0.9419153094291687,
+      "num_tokens": 94886.0,
+      "step": 70
+    },
+    {
+      "entropy": 0.9988491177558899,
+      "epoch": 0.07648183556405354,
+      "grad_norm": 2.444343328475952,
+      "learning_rate": 7.523809523809524e-05,
+      "loss": 0.1839,
+      "mean_token_accuracy": 0.9345499813556671,
+      "num_tokens": 108569.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.0041434288024902,
+      "epoch": 0.08604206500956023,
+      "grad_norm": 1.929736614227295,
+      "learning_rate": 8.476190476190477e-05,
+      "loss": 0.1645,
+      "mean_token_accuracy": 0.9391969263553619,
+      "num_tokens": 122156.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.0116430222988129,
+      "epoch": 0.09560229445506692,
+      "grad_norm": 2.168968915939331,
+      "learning_rate": 9.428571428571429e-05,
+      "loss": 0.1929,
+      "mean_token_accuracy": 0.9271840214729309,
+      "num_tokens": 135290.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.0483016848564148,
+      "epoch": 0.10516252390057361,
+      "grad_norm": 2.018573760986328,
+      "learning_rate": 9.999900008621665e-05,
+      "loss": 0.2368,
+      "mean_token_accuracy": 0.9140790283679963,
+      "num_tokens": 149015.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.0560781598091125,
+      "epoch": 0.1147227533460803,
+      "grad_norm": 1.9477077722549438,
+      "learning_rate": 9.998775151544478e-05,
+      "loss": 0.313,
+      "mean_token_accuracy": 0.9046160459518433,
+      "num_tokens": 162147.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.0339066147804261,
+      "epoch": 0.124282982791587,
+      "grad_norm": 3.953827142715454,
+      "learning_rate": 9.996400730289628e-05,
+      "loss": 0.2581,
+      "mean_token_accuracy": 0.9119772851467133,
+      "num_tokens": 175436.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.0594236373901367,
+      "epoch": 0.1338432122370937,
+      "grad_norm": 2.7890758514404297,
+      "learning_rate": 9.992777338400864e-05,
+      "loss": 0.2533,
+      "mean_token_accuracy": 0.9100077152252197,
+      "num_tokens": 188834.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.053187495470047,
+      "epoch": 0.14340344168260039,
+      "grad_norm": 2.2614591121673584,
+      "learning_rate": 9.987905881632213e-05,
+      "loss": 0.289,
+      "mean_token_accuracy": 0.899867707490921,
+      "num_tokens": 202708.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.0322751343250274,
+      "epoch": 0.15296367112810708,
+      "grad_norm": 2.2198617458343506,
+      "learning_rate": 9.981787577721553e-05,
+      "loss": 0.2234,
+      "mean_token_accuracy": 0.9163801074028015,
+      "num_tokens": 216384.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.0143227398395538,
+      "epoch": 0.16252390057361377,
+      "grad_norm": 2.215290069580078,
+      "learning_rate": 9.974423956086228e-05,
+      "loss": 0.2049,
+      "mean_token_accuracy": 0.9241275012493133,
+      "num_tokens": 229105.0,
+      "step": 170
+    },
+    {
+      "entropy": 0.9944452047348022,
+      "epoch": 0.17208413001912046,
+      "grad_norm": 3.2736923694610596,
+      "learning_rate": 9.965816857440717e-05,
+      "loss": 0.209,
+      "mean_token_accuracy": 0.919926643371582,
+      "num_tokens": 242264.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.015661782026291,
+      "epoch": 0.18164435946462715,
+      "grad_norm": 2.2948694229125977,
+      "learning_rate": 9.955968433336515e-05,
+      "loss": 0.2268,
+      "mean_token_accuracy": 0.9108724594116211,
+      "num_tokens": 255602.0,
+      "step": 190
+    },
+    {
+      "entropy": 0.9878140687942505,
+      "epoch": 0.19120458891013384,
+      "grad_norm": 1.7582422494888306,
+      "learning_rate": 9.944881145624291e-05,
+      "loss": 0.1732,
+      "mean_token_accuracy": 0.9377074241638184,
+      "num_tokens": 269712.0,
+      "step": 200
+    },
+    {
+      "entropy": 0.9916601777076721,
+      "epoch": 0.20076481835564053,
+      "grad_norm": 2.493232011795044,
+      "learning_rate": 9.932557765838504e-05,
+      "loss": 0.1978,
+      "mean_token_accuracy": 0.9271564424037934,
+      "num_tokens": 283572.0,
+      "step": 210
+    },
+    {
+      "entropy": 0.9938843488693238,
+      "epoch": 0.21032504780114722,
+      "grad_norm": 1.9961174726486206,
+      "learning_rate": 9.919001374504569e-05,
+      "loss": 0.238,
+      "mean_token_accuracy": 0.9079000949859619,
+      "num_tokens": 297437.0,
+      "step": 220
+    },
+    {
+      "entropy": 0.9930910289287567,
+      "epoch": 0.2198852772466539,
+      "grad_norm": 4.460914611816406,
+      "learning_rate": 9.904215360368826e-05,
+      "loss": 0.2281,
+      "mean_token_accuracy": 0.9163453280925751,
+      "num_tokens": 310803.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.0042203903198241,
+      "epoch": 0.2294455066921606,
+      "grad_norm": 2.591485023498535,
+      "learning_rate": 9.888203419551432e-05,
+      "loss": 0.2399,
+      "mean_token_accuracy": 0.909199732542038,
+      "num_tokens": 324713.0,
+      "step": 240
+    },
+    {
+      "entropy": 0.9938572168350219,
+      "epoch": 0.2390057361376673,
+      "grad_norm": 2.540052890777588,
+      "learning_rate": 9.870969554622426e-05,
+      "loss": 0.2626,
+      "mean_token_accuracy": 0.9039014875888824,
+      "num_tokens": 337634.0,
+      "step": 250
+    },
+    {
+      "entropy": 0.981248390674591,
+      "epoch": 0.248565965583174,
+      "grad_norm": 2.0444109439849854,
+      "learning_rate": 9.852518073601194e-05,
+      "loss": 0.2077,
+      "mean_token_accuracy": 0.9231703639030456,
+      "num_tokens": 350459.0,
+      "step": 260
+    },
+    {
+      "entropy": 0.9621168673038483,
+      "epoch": 0.25812619502868067,
+      "grad_norm": 2.641737699508667,
+      "learning_rate": 9.832853588879575e-05,
+      "loss": 0.2273,
+      "mean_token_accuracy": 0.9158074796199799,
+      "num_tokens": 364010.0,
+      "step": 270
+    },
+    {
+      "entropy": 0.997821694612503,
+      "epoch": 0.2676864244741874,
+      "grad_norm": 2.929631233215332,
+      "learning_rate": 9.811981016068881e-05,
+      "loss": 0.2722,
+      "mean_token_accuracy": 0.9029962062835694,
+      "num_tokens": 377021.0,
+      "step": 280
+    },
+    {
+      "entropy": 0.9611131608486175,
+      "epoch": 0.27724665391969405,
+      "grad_norm": 1.1362237930297852,
+      "learning_rate": 9.78990557277112e-05,
+      "loss": 0.2092,
+      "mean_token_accuracy": 0.9216417133808136,
+      "num_tokens": 390819.0,
+      "step": 290
+    },
+    {
+      "entropy": 0.9901561737060547,
+      "epoch": 0.28680688336520077,
+      "grad_norm": 2.0782461166381836,
+      "learning_rate": 9.766632777274738e-05,
+      "loss": 0.2453,
+      "mean_token_accuracy": 0.9103418946266174,
+      "num_tokens": 404644.0,
+      "step": 300
+    },
+    {
+      "entropy": 0.9956516146659851,
+      "epoch": 0.29636711281070743,
+      "grad_norm": 2.935779571533203,
+      "learning_rate": 9.742168447175182e-05,
+      "loss": 0.2293,
+      "mean_token_accuracy": 0.9149252951145173,
+      "num_tokens": 418529.0,
+      "step": 310
+    },
+    {
+      "entropy": 0.9521589517593384,
+      "epoch": 0.30592734225621415,
+      "grad_norm": 1.9554321765899658,
+      "learning_rate": 9.716518697920655e-05,
+      "loss": 0.2028,
+      "mean_token_accuracy": 0.92679563164711,
+      "num_tokens": 431509.0,
+      "step": 320
+    },
+    {
+      "entropy": 0.9850484609603882,
+      "epoch": 0.3154875717017208,
+      "grad_norm": 2.1091372966766357,
+      "learning_rate": 9.689689941283417e-05,
+      "loss": 0.2399,
+      "mean_token_accuracy": 0.9111748278141022,
+      "num_tokens": 445040.0,
+      "step": 330
+    },
+    {
+      "entropy": 0.9910066843032836,
+      "epoch": 0.32504780114722753,
+      "grad_norm": 2.729727268218994,
+      "learning_rate": 9.661688883756996e-05,
+      "loss": 0.2222,
+      "mean_token_accuracy": 0.9158975481987,
+      "num_tokens": 458540.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.0245653331279754,
+      "epoch": 0.33460803059273425,
+      "grad_norm": 3.326910972595215,
+      "learning_rate": 9.632522524879756e-05,
+      "loss": 0.2471,
+      "mean_token_accuracy": 0.9084886729717254,
+      "num_tokens": 471556.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.0443902254104613,
+      "epoch": 0.3441682600382409,
+      "grad_norm": 2.765958309173584,
+      "learning_rate": 9.602198155485176e-05,
+      "loss": 0.2307,
+      "mean_token_accuracy": 0.9184524953365326,
+      "num_tokens": 484786.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.055874276161194,
+      "epoch": 0.35372848948374763,
+      "grad_norm": 2.167161226272583,
+      "learning_rate": 9.570723355879342e-05,
+      "loss": 0.2102,
+      "mean_token_accuracy": 0.9206913709640503,
+      "num_tokens": 498352.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.0401476562023162,
+      "epoch": 0.3632887189292543,
+      "grad_norm": 2.807732343673706,
+      "learning_rate": 9.538105993946068e-05,
+      "loss": 0.2307,
+      "mean_token_accuracy": 0.9132277011871338,
+      "num_tokens": 512064.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.0409152030944824,
+      "epoch": 0.372848948374761,
+      "grad_norm": 1.8228720426559448,
+      "learning_rate": 9.504354223180118e-05,
+      "loss": 0.2182,
+      "mean_token_accuracy": 0.9216055631637573,
+      "num_tokens": 525582.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.0548848509788513,
+      "epoch": 0.3824091778202677,
+      "grad_norm": 3.267235279083252,
+      "learning_rate": 9.469476480649056e-05,
+      "loss": 0.2297,
+      "mean_token_accuracy": 0.9145385086536407,
+      "num_tokens": 539561.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.0603472709655761,
+      "epoch": 0.3919694072657744,
+      "grad_norm": 1.9201253652572632,
+      "learning_rate": 9.433481484884189e-05,
+      "loss": 0.2186,
+      "mean_token_accuracy": 0.9178025484085083,
+      "num_tokens": 553625.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.0538666605949403,
+      "epoch": 0.40152963671128106,
+      "grad_norm": 2.342041492462158,
+      "learning_rate": 9.396378233701167e-05,
+      "loss": 0.1925,
+      "mean_token_accuracy": 0.9293496191501618,
+      "num_tokens": 565891.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.0616689920425415,
+      "epoch": 0.4110898661567878,
+      "grad_norm": 1.9766606092453003,
+      "learning_rate": 9.358176001950759e-05,
+      "loss": 0.2311,
+      "mean_token_accuracy": 0.9161762118339538,
+      "num_tokens": 578722.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.0873818516731262,
+      "epoch": 0.42065009560229444,
+      "grad_norm": 2.961918830871582,
+      "learning_rate": 9.318884339200378e-05,
+      "loss": 0.238,
+      "mean_token_accuracy": 0.9079842686653137,
+      "num_tokens": 592646.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.0649721384048463,
+      "epoch": 0.43021032504780116,
+      "grad_norm": 2.5015785694122314,
+      "learning_rate": 9.278513067346941e-05,
+      "loss": 0.1898,
+      "mean_token_accuracy": 0.9307917773723602,
+      "num_tokens": 606429.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.0712341904640197,
+      "epoch": 0.4397705544933078,
+      "grad_norm": 2.5987443923950195,
+      "learning_rate": 9.237072278161652e-05,
+      "loss": 0.2433,
+      "mean_token_accuracy": 0.9096251487731933,
+      "num_tokens": 619626.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.0651772737503051,
+      "epoch": 0.44933078393881454,
+      "grad_norm": 2.339188814163208,
+      "learning_rate": 9.194572330767314e-05,
+      "loss": 0.2105,
+      "mean_token_accuracy": 0.9202918171882629,
+      "num_tokens": 633055.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.0719932675361634,
+      "epoch": 0.4588910133843212,
+      "grad_norm": 2.139345169067383,
+      "learning_rate": 9.151023849048818e-05,
+      "loss": 0.1972,
+      "mean_token_accuracy": 0.9251638293266297,
+      "num_tokens": 647155.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.0447640657424926,
+      "epoch": 0.4684512428298279,
+      "grad_norm": 3.1209723949432373,
+      "learning_rate": 9.106437718997452e-05,
+      "loss": 0.2031,
+      "mean_token_accuracy": 0.921468985080719,
+      "num_tokens": 660695.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.0626697421073914,
+      "epoch": 0.4780114722753346,
+      "grad_norm": 1.8369485139846802,
+      "learning_rate": 9.060825085989682e-05,
+      "loss": 0.2105,
+      "mean_token_accuracy": 0.9239315092563629,
+      "num_tokens": 674509.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.0570692539215087,
+      "epoch": 0.4875717017208413,
+      "grad_norm": 3.0195302963256836,
+      "learning_rate": 9.014197352001097e-05,
+      "loss": 0.1821,
+      "mean_token_accuracy": 0.9317115485668183,
+      "num_tokens": 687644.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.0473023772239685,
+      "epoch": 0.497131931166348,
+      "grad_norm": 2.375627279281616,
+      "learning_rate": 8.966566172756213e-05,
+      "loss": 0.2201,
+      "mean_token_accuracy": 0.9189318299293519,
+      "num_tokens": 700658.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.0765699386596679,
+      "epoch": 0.5066921606118547,
+      "grad_norm": 2.8174030780792236,
+      "learning_rate": 8.91794345481484e-05,
+      "loss": 0.208,
+      "mean_token_accuracy": 0.91789231300354,
+      "num_tokens": 714234.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.0702912211418152,
+      "epoch": 0.5162523900573613,
+      "grad_norm": 3.2318389415740967,
+      "learning_rate": 8.868341352595755e-05,
+      "loss": 0.2295,
+      "mean_token_accuracy": 0.9122642040252685,
+      "num_tokens": 727721.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.0573222517967225,
+      "epoch": 0.5258126195028681,
+      "grad_norm": 1.7950525283813477,
+      "learning_rate": 8.817772265338409e-05,
+      "loss": 0.223,
+      "mean_token_accuracy": 0.9139611005783081,
+      "num_tokens": 741707.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.0547808468341828,
+      "epoch": 0.5353728489483748,
+      "grad_norm": 3.6276676654815674,
+      "learning_rate": 8.766248834003441e-05,
+      "loss": 0.2363,
+      "mean_token_accuracy": 0.9099368512630462,
+      "num_tokens": 755385.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.0302533328533172,
+      "epoch": 0.5449330783938815,
+      "grad_norm": 2.1893699169158936,
+      "learning_rate": 8.713783938112768e-05,
+      "loss": 0.199,
+      "mean_token_accuracy": 0.926015430688858,
+      "num_tokens": 768687.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.0444139957427978,
+      "epoch": 0.5544933078393881,
+      "grad_norm": 2.8716442584991455,
+      "learning_rate": 8.660390692530026e-05,
+      "loss": 0.1995,
+      "mean_token_accuracy": 0.9259741365909576,
+      "num_tokens": 782172.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.0370828688144684,
+      "epoch": 0.5640535372848948,
+      "grad_norm": 1.763108730316162,
+      "learning_rate": 8.606082444182216e-05,
+      "loss": 0.2055,
+      "mean_token_accuracy": 0.9247264742851258,
+      "num_tokens": 795561.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.0649680018424987,
+      "epoch": 0.5736137667304015,
+      "grad_norm": 1.66559636592865,
+      "learning_rate": 8.550872768723292e-05,
+      "loss": 0.1828,
+      "mean_token_accuracy": 0.9340787351131439,
+      "num_tokens": 809436.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.0512043595314027,
+      "epoch": 0.5831739961759083,
+      "grad_norm": 4.193369388580322,
+      "learning_rate": 8.494775467140627e-05,
+      "loss": 0.2083,
+      "mean_token_accuracy": 0.9245233416557312,
+      "num_tokens": 822983.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.0559794306755066,
+      "epoch": 0.5927342256214149,
+      "grad_norm": 1.8999974727630615,
+      "learning_rate": 8.4378045623051e-05,
+      "loss": 0.2028,
+      "mean_token_accuracy": 0.9243997693061828,
+      "num_tokens": 836766.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.0672565937042235,
+      "epoch": 0.6022944550669216,
+      "grad_norm": 1.607515811920166,
+      "learning_rate": 8.379974295465758e-05,
+      "loss": 0.2156,
+      "mean_token_accuracy": 0.916247296333313,
+      "num_tokens": 850200.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.0445150911808014,
+      "epoch": 0.6118546845124283,
+      "grad_norm": 1.7760804891586304,
+      "learning_rate": 8.321299122689863e-05,
+      "loss": 0.2276,
+      "mean_token_accuracy": 0.9184480905532837,
+      "num_tokens": 864001.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.0459101319313049,
+      "epoch": 0.621414913957935,
+      "grad_norm": 2.6341512203216553,
+      "learning_rate": 8.261793711249254e-05,
+      "loss": 0.2191,
+      "mean_token_accuracy": 0.9138937532901764,
+      "num_tokens": 878135.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.0317402541637422,
+      "epoch": 0.6309751434034416,
+      "grad_norm": 1.8982086181640625,
+      "learning_rate": 8.201472935953916e-05,
+      "loss": 0.2169,
+      "mean_token_accuracy": 0.9217910826206207,
+      "num_tokens": 891307.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.045106440782547,
+      "epoch": 0.6405353728489483,
+      "grad_norm": 1.7897950410842896,
+      "learning_rate": 8.140351875433649e-05,
+      "loss": 0.2075,
+      "mean_token_accuracy": 0.9199797868728637,
+      "num_tokens": 904829.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.0383180022239684,
+      "epoch": 0.6500956022944551,
+      "grad_norm": 1.8748987913131714,
+      "learning_rate": 8.07844580836881e-05,
+      "loss": 0.1829,
+      "mean_token_accuracy": 0.9320280909538269,
+      "num_tokens": 919199.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.0720835208892823,
+      "epoch": 0.6596558317399618,
+      "grad_norm": 1.7298226356506348,
+      "learning_rate": 8.015770209671044e-05,
+      "loss": 0.2247,
+      "mean_token_accuracy": 0.9139919757843018,
+      "num_tokens": 932439.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.0737423539161681,
+      "epoch": 0.6692160611854685,
+      "grad_norm": 2.094554901123047,
+      "learning_rate": 7.952340746614946e-05,
+      "loss": 0.2105,
+      "mean_token_accuracy": 0.921952760219574,
+      "num_tokens": 945776.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.0488283574581145,
+      "epoch": 0.6787762906309751,
+      "grad_norm": 2.220661163330078,
+      "learning_rate": 7.888173274921655e-05,
+      "loss": 0.1762,
+      "mean_token_accuracy": 0.9325842142105103,
+      "num_tokens": 959352.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.0624409198760987,
+      "epoch": 0.6883365200764818,
+      "grad_norm": 2.789525032043457,
+      "learning_rate": 7.823283834795336e-05,
+      "loss": 0.1965,
+      "mean_token_accuracy": 0.9271815240383148,
+      "num_tokens": 972596.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.0450023353099822,
+      "epoch": 0.6978967495219885,
+      "grad_norm": 1.9600374698638916,
+      "learning_rate": 7.75768864691354e-05,
+      "loss": 0.2181,
+      "mean_token_accuracy": 0.9186623334884644,
+      "num_tokens": 985699.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.050288939476013,
+      "epoch": 0.7074569789674953,
+      "grad_norm": 2.745725154876709,
+      "learning_rate": 7.691404108372461e-05,
+      "loss": 0.1958,
+      "mean_token_accuracy": 0.9252684831619262,
+      "num_tokens": 999728.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.0660975933074952,
+      "epoch": 0.7170172084130019,
+      "grad_norm": 1.9985816478729248,
+      "learning_rate": 7.624446788588085e-05,
+      "loss": 0.1743,
+      "mean_token_accuracy": 0.934730589389801,
+      "num_tokens": 1013669.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.0661585927009583,
+      "epoch": 0.7265774378585086,
+      "grad_norm": 1.8102953433990479,
+      "learning_rate": 7.556833425154255e-05,
+      "loss": 0.1806,
+      "mean_token_accuracy": 0.9322760343551636,
+      "num_tokens": 1027609.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.0598073244094848,
+      "epoch": 0.7361376673040153,
+      "grad_norm": 2.5633928775787354,
+      "learning_rate": 7.488580919658727e-05,
+      "loss": 0.1929,
+      "mean_token_accuracy": 0.9238331735134124,
+      "num_tokens": 1041046.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.04606152176857,
+      "epoch": 0.745697896749522,
+      "grad_norm": 2.917694568634033,
+      "learning_rate": 7.419706333458201e-05,
+      "loss": 0.179,
+      "mean_token_accuracy": 0.9323089599609375,
+      "num_tokens": 1055304.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.0437167465686799,
+      "epoch": 0.7552581261950286,
+      "grad_norm": 3.5574042797088623,
+      "learning_rate": 7.350226883413413e-05,
+      "loss": 0.209,
+      "mean_token_accuracy": 0.9239528596401214,
+      "num_tokens": 1068542.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.0587083101272583,
+      "epoch": 0.7648183556405354,
+      "grad_norm": 2.3629257678985596,
+      "learning_rate": 7.280159937585386e-05,
+      "loss": 0.1855,
+      "mean_token_accuracy": 0.9248183965682983,
+      "num_tokens": 1082195.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.059005331993103,
+      "epoch": 0.7743785850860421,
+      "grad_norm": 3.765178680419922,
+      "learning_rate": 7.209523010893853e-05,
+      "loss": 0.1667,
+      "mean_token_accuracy": 0.9390020132064819,
+      "num_tokens": 1095884.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.070823109149933,
+      "epoch": 0.7839388145315488,
+      "grad_norm": 1.7191312313079834,
+      "learning_rate": 7.138333760738987e-05,
+      "loss": 0.2056,
+      "mean_token_accuracy": 0.9238715708255768,
+      "num_tokens": 1109287.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.070789921283722,
+      "epoch": 0.7934990439770554,
+      "grad_norm": 2.819929361343384,
+      "learning_rate": 7.066609982587501e-05,
+      "loss": 0.1821,
+      "mean_token_accuracy": 0.9314943969249725,
+      "num_tokens": 1123392.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.0708677768707275,
+      "epoch": 0.8030592734225621,
+      "grad_norm": 2.1845016479492188,
+      "learning_rate": 6.994369605524253e-05,
+      "loss": 0.1971,
+      "mean_token_accuracy": 0.9240113377571106,
+      "num_tokens": 1137691.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.03220654129982,
+      "epoch": 0.8126195028680688,
+      "grad_norm": 3.272754430770874,
+      "learning_rate": 6.921630687770427e-05,
+      "loss": 0.186,
+      "mean_token_accuracy": 0.9262190937995911,
+      "num_tokens": 1151400.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.043507421016693,
+      "epoch": 0.8221797323135756,
+      "grad_norm": 2.8006186485290527,
+      "learning_rate": 6.848411412169461e-05,
+      "loss": 0.1591,
+      "mean_token_accuracy": 0.9413301229476929,
+      "num_tokens": 1165074.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.0422970294952392,
+      "epoch": 0.8317399617590823,
+      "grad_norm": 2.7493834495544434,
+      "learning_rate": 6.774730081641799e-05,
+      "loss": 0.181,
+      "mean_token_accuracy": 0.9326336324214936,
+      "num_tokens": 1178919.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.0501964330673217,
+      "epoch": 0.8413001912045889,
+      "grad_norm": 3.106358051300049,
+      "learning_rate": 6.70060511460963e-05,
+      "loss": 0.2042,
+      "mean_token_accuracy": 0.9240445792675018,
+      "num_tokens": 1192204.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.0335909485816956,
+      "epoch": 0.8508604206500956,
+      "grad_norm": 2.1840972900390625,
+      "learning_rate": 6.626055040392774e-05,
+      "loss": 0.1897,
+      "mean_token_accuracy": 0.9266136407852172,
+      "num_tokens": 1205571.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.0435899496078491,
+      "epoch": 0.8604206500956023,
+      "grad_norm": 2.261042594909668,
+      "learning_rate": 6.551098494576819e-05,
+      "loss": 0.2012,
+      "mean_token_accuracy": 0.9291186094284057,
+      "num_tokens": 1219043.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.0415594696998596,
+      "epoch": 0.869980879541109,
+      "grad_norm": 2.32177996635437,
+      "learning_rate": 6.475754214354715e-05,
+      "loss": 0.1749,
+      "mean_token_accuracy": 0.9318947196006775,
+      "num_tokens": 1233113.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.0231156051158905,
+      "epoch": 0.8795411089866156,
+      "grad_norm": 2.1848862171173096,
+      "learning_rate": 6.400041033842965e-05,
+      "loss": 0.1713,
+      "mean_token_accuracy": 0.9378451108932495,
+      "num_tokens": 1245931.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.0183886528015136,
+      "epoch": 0.8891013384321224,
+      "grad_norm": 2.509147882461548,
+      "learning_rate": 6.323977879373571e-05,
+      "loss": 0.1798,
+      "mean_token_accuracy": 0.9298445463180542,
+      "num_tokens": 1259949.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.0446486234664918,
+      "epoch": 0.8986615678776291,
+      "grad_norm": 2.408195972442627,
+      "learning_rate": 6.247583764762959e-05,
+      "loss": 0.1731,
+      "mean_token_accuracy": 0.9333014845848083,
+      "num_tokens": 1273413.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.0374162077903748,
+      "epoch": 0.9082217973231358,
+      "grad_norm": 1.9992507696151733,
+      "learning_rate": 6.170877786559001e-05,
+      "loss": 0.1761,
+      "mean_token_accuracy": 0.9344584167003631,
+      "num_tokens": 1286343.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.021628212928772,
+      "epoch": 0.9177820267686424,
+      "grad_norm": 2.594414472579956,
+      "learning_rate": 6.0938791192673895e-05,
+      "loss": 0.1778,
+      "mean_token_accuracy": 0.9287881553173065,
+      "num_tokens": 1300089.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.0166936099529267,
+      "epoch": 0.9273422562141491,
+      "grad_norm": 3.2213165760040283,
+      "learning_rate": 6.016607010558492e-05,
+      "loss": 0.1648,
+      "mean_token_accuracy": 0.9389560222625732,
+      "num_tokens": 1313412.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.0378799736499786,
+      "epoch": 0.9369024856596558,
+      "grad_norm": 2.705141544342041,
+      "learning_rate": 5.939080776455942e-05,
+      "loss": 0.1852,
+      "mean_token_accuracy": 0.9291444897651673,
+      "num_tokens": 1326290.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.0505308985710144,
+      "epoch": 0.9464627151051626,
+      "grad_norm": 2.3800392150878906,
+      "learning_rate": 5.861319796508139e-05,
+      "loss": 0.1669,
+      "mean_token_accuracy": 0.9396340370178222,
+      "num_tokens": 1339719.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.0232268452644349,
+      "epoch": 0.9560229445506692,
+      "grad_norm": 1.9888688325881958,
+      "learning_rate": 5.783343508943849e-05,
+      "loss": 0.1507,
+      "mean_token_accuracy": 0.9391794621944427,
+      "num_tokens": 1352423.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.0251559376716615,
+      "epoch": 0.9655831739961759,
+      "grad_norm": 2.7010385990142822,
+      "learning_rate": 5.7051714058131654e-05,
+      "loss": 0.1662,
+      "mean_token_accuracy": 0.9356134235858917,
+      "num_tokens": 1365889.0,
+      "step": 1010
+    },
+    {
+      "entropy": 1.0165675580501556,
+      "epoch": 0.9751434034416826,
+      "grad_norm": 2.578207015991211,
+      "learning_rate": 5.626823028114991e-05,
+      "loss": 0.1674,
+      "mean_token_accuracy": 0.9337117493152618,
+      "num_tokens": 1379353.0,
+      "step": 1020
+    },
+    {
+      "entropy": 1.0138891518115998,
+      "epoch": 0.9847036328871893,
+      "grad_norm": 2.096973180770874,
+      "learning_rate": 5.548317960912305e-05,
+      "loss": 0.176,
+      "mean_token_accuracy": 0.9312076151371003,
+      "num_tokens": 1392866.0,
+      "step": 1030
+    },
+    {
+      "entropy": 1.0109597086906432,
+      "epoch": 0.994263862332696,
+      "grad_norm": 2.348393678665161,
+      "learning_rate": 5.4696758284363956e-05,
+      "loss": 0.1805,
+      "mean_token_accuracy": 0.928768002986908,
+      "num_tokens": 1406263.0,
+      "step": 1040
+    },
+    {
+      "entropy": 1.0230879187583923,
+      "epoch": 1.0038240917782026,
+      "grad_norm": 2.11037015914917,
+      "learning_rate": 5.390916289181318e-05,
+      "loss": 0.1606,
+      "mean_token_accuracy": 0.9352643191814423,
+      "num_tokens": 1419071.0,
+      "step": 1050
+    },
+    {
+      "entropy": 0.9678762078285217,
+      "epoch": 1.0133843212237095,
+      "grad_norm": 2.9552314281463623,
+      "learning_rate": 5.312059030989779e-05,
+      "loss": 0.1039,
+      "mean_token_accuracy": 0.965081787109375,
+      "num_tokens": 1432063.0,
+      "step": 1060
+    },
+    {
+      "entropy": 0.9691481709480285,
+      "epoch": 1.022944550669216,
+      "grad_norm": 1.7249557971954346,
+      "learning_rate": 5.233123766131681e-05,
+      "loss": 0.1,
+      "mean_token_accuracy": 0.9644009649753571,
+      "num_tokens": 1446234.0,
+      "step": 1070
+    },
+    {
+      "entropy": 0.9605498313903809,
+      "epoch": 1.0325047801147227,
+      "grad_norm": 2.0945215225219727,
+      "learning_rate": 5.154130226376578e-05,
+      "loss": 0.1047,
+      "mean_token_accuracy": 0.964258748292923,
+      "num_tokens": 1460025.0,
+      "step": 1080
+    },
+    {
+      "entropy": 0.9631986320018768,
+      "epoch": 1.0420650095602295,
+      "grad_norm": 2.8726062774658203,
+      "learning_rate": 5.0750981580612277e-05,
+      "loss": 0.0993,
+      "mean_token_accuracy": 0.9629338204860687,
+      "num_tokens": 1474126.0,
+      "step": 1090
+    },
+    {
+      "entropy": 0.95530207157135,
+      "epoch": 1.0516252390057361,
+      "grad_norm": 3.331778049468994,
+      "learning_rate": 4.996047317153538e-05,
+      "loss": 0.0984,
+      "mean_token_accuracy": 0.9635574400424958,
+      "num_tokens": 1487468.0,
+      "step": 1100
+    },
+    {
+      "entropy": 0.973723828792572,
+      "epoch": 1.0611854684512427,
+      "grad_norm": 2.642136812210083,
+      "learning_rate": 4.916997464314074e-05,
+      "loss": 0.1086,
+      "mean_token_accuracy": 0.959328556060791,
+      "num_tokens": 1500810.0,
+      "step": 1110
+    },
+    {
+      "entropy": 0.9613595366477966,
+      "epoch": 1.0707456978967496,
+      "grad_norm": 2.8338515758514404,
+      "learning_rate": 4.8379683599564086e-05,
+      "loss": 0.1122,
+      "mean_token_accuracy": 0.9570079445838928,
+      "num_tokens": 1514831.0,
+      "step": 1120
+    },
+    {
+      "entropy": 0.9373875916004181,
+      "epoch": 1.0803059273422562,
+      "grad_norm": 2.8853886127471924,
+      "learning_rate": 4.7589797593075334e-05,
+      "loss": 0.0939,
+      "mean_token_accuracy": 0.9639353334903717,
+      "num_tokens": 1528268.0,
+      "step": 1130
+    },
+    {
+      "entropy": 0.9482690274715424,
+      "epoch": 1.089866156787763,
+      "grad_norm": 2.460653305053711,
+      "learning_rate": 4.6800514074695626e-05,
+      "loss": 0.0963,
+      "mean_token_accuracy": 0.9628588199615479,
+      "num_tokens": 1541919.0,
+      "step": 1140
+    },
+    {
+      "entropy": 0.9200351357460022,
+      "epoch": 1.0994263862332696,
+      "grad_norm": 1.646629810333252,
+      "learning_rate": 4.601203034483971e-05,
+      "loss": 0.0613,
+      "mean_token_accuracy": 0.9777628421783447,
+      "num_tokens": 1554834.0,
+      "step": 1150
+    },
+    {
+      "entropy": 0.940442544221878,
+      "epoch": 1.1089866156787762,
+      "grad_norm": 2.1895744800567627,
+      "learning_rate": 4.522454350399592e-05,
+      "loss": 0.0976,
+      "mean_token_accuracy": 0.9663795173168183,
+      "num_tokens": 1568489.0,
+      "step": 1160
+    },
+    {
+      "entropy": 0.9375030934810639,
+      "epoch": 1.118546845124283,
+      "grad_norm": 2.5660712718963623,
+      "learning_rate": 4.4438250403456187e-05,
+      "loss": 0.0937,
+      "mean_token_accuracy": 0.9679263234138489,
+      "num_tokens": 1581819.0,
+      "step": 1170
+    },
+    {
+      "entropy": 0.9362292945384979,
+      "epoch": 1.1281070745697896,
+      "grad_norm": 2.010430335998535,
+      "learning_rate": 4.365334759610836e-05,
+      "loss": 0.1101,
+      "mean_token_accuracy": 0.9590835869312286,
+      "num_tokens": 1595977.0,
+      "step": 1180
+    },
+    {
+      "entropy": 0.9469062447547912,
+      "epoch": 1.1376673040152965,
+      "grad_norm": 2.2855584621429443,
+      "learning_rate": 4.2870031287302984e-05,
+      "loss": 0.1206,
+      "mean_token_accuracy": 0.9546375393867492,
+      "num_tokens": 1609598.0,
+      "step": 1190
+    },
+    {
+      "entropy": 0.957787960767746,
+      "epoch": 1.147227533460803,
+      "grad_norm": 3.1256892681121826,
+      "learning_rate": 4.2088497285807125e-05,
+      "loss": 0.1115,
+      "mean_token_accuracy": 0.9619700968265533,
+      "num_tokens": 1623264.0,
+      "step": 1200
+    },
+    {
+      "entropy": 0.9339599788188935,
+      "epoch": 1.1567877629063097,
+      "grad_norm": 2.6361329555511475,
+      "learning_rate": 4.1308940954857236e-05,
+      "loss": 0.0924,
+      "mean_token_accuracy": 0.9683476984500885,
+      "num_tokens": 1636568.0,
+      "step": 1210
+    },
+    {
+      "entropy": 0.9221574664115906,
+      "epoch": 1.1663479923518165,
+      "grad_norm": 1.7546945810317993,
+      "learning_rate": 4.0531557163323415e-05,
+      "loss": 0.0854,
+      "mean_token_accuracy": 0.9669972598552704,
+      "num_tokens": 1650127.0,
+      "step": 1220
+    },
+    {
+      "entropy": 0.9197199463844299,
+      "epoch": 1.1759082217973231,
+      "grad_norm": 1.9218872785568237,
+      "learning_rate": 3.975654023699722e-05,
+      "loss": 0.0831,
+      "mean_token_accuracy": 0.9680957674980164,
+      "num_tokens": 1663753.0,
+      "step": 1230
+    },
+    {
+      "entropy": 0.9254310071468353,
+      "epoch": 1.1854684512428297,
+      "grad_norm": 1.402408480644226,
+      "learning_rate": 3.898408391001528e-05,
+      "loss": 0.0859,
+      "mean_token_accuracy": 0.9705113172531128,
+      "num_tokens": 1677205.0,
+      "step": 1240
+    },
+    {
+      "entropy": 0.9345485806465149,
+      "epoch": 1.1950286806883366,
+      "grad_norm": 1.770240306854248,
+      "learning_rate": 3.821438127643078e-05,
+      "loss": 0.0893,
+      "mean_token_accuracy": 0.9672084450721741,
+      "num_tokens": 1690690.0,
+      "step": 1250
+    },
+    {
+      "entropy": 0.9160260796546936,
+      "epoch": 1.2045889101338432,
+      "grad_norm": 1.718557596206665,
+      "learning_rate": 3.744762474194501e-05,
+      "loss": 0.0748,
+      "mean_token_accuracy": 0.9711562871932984,
+      "num_tokens": 1704273.0,
+      "step": 1260
+    },
+    {
+      "entropy": 0.9108176112174988,
+      "epoch": 1.21414913957935,
+      "grad_norm": 2.571093797683716,
+      "learning_rate": 3.6684005975810765e-05,
+      "loss": 0.0942,
+      "mean_token_accuracy": 0.9683524668216705,
+      "num_tokens": 1717477.0,
+      "step": 1270
+    },
+    {
+      "entropy": 0.9296231389045715,
+      "epoch": 1.2237093690248566,
+      "grad_norm": 2.0588877201080322,
+      "learning_rate": 3.592371586292021e-05,
+      "loss": 0.0849,
+      "mean_token_accuracy": 0.9677265346050262,
+      "num_tokens": 1731709.0,
+      "step": 1280
+    },
+    {
+      "entropy": 0.9172577679157257,
+      "epoch": 1.2332695984703632,
+      "grad_norm": 1.5681378841400146,
+      "learning_rate": 3.5166944456088445e-05,
+      "loss": 0.0709,
+      "mean_token_accuracy": 0.976532256603241,
+      "num_tokens": 1744487.0,
+      "step": 1290
+    },
+    {
+      "entropy": 0.9368212342262268,
+      "epoch": 1.24282982791587,
+      "grad_norm": 2.6401617527008057,
+      "learning_rate": 3.441388092854532e-05,
+      "loss": 0.0899,
+      "mean_token_accuracy": 0.9667512595653533,
+      "num_tokens": 1757771.0,
+      "step": 1300
+    },
+    {
+      "entropy": 0.9322176575660706,
+      "epoch": 1.2523900573613767,
+      "grad_norm": 1.871292233467102,
+      "learning_rate": 3.3664713526646976e-05,
+      "loss": 0.0818,
+      "mean_token_accuracy": 0.9713339924812316,
+      "num_tokens": 1771222.0,
+      "step": 1310
+    },
+    {
+      "entropy": 0.9192366600036621,
+      "epoch": 1.2619502868068833,
+      "grad_norm": 3.7188029289245605,
+      "learning_rate": 3.291962952281918e-05,
+      "loss": 0.0742,
+      "mean_token_accuracy": 0.9689174830913544,
+      "num_tokens": 1784361.0,
+      "step": 1320
+    },
+    {
+      "entropy": 0.9271296799182892,
+      "epoch": 1.27151051625239,
+      "grad_norm": 1.9506016969680786,
+      "learning_rate": 3.217881516874407e-05,
+      "loss": 0.0816,
+      "mean_token_accuracy": 0.9719076514244079,
+      "num_tokens": 1797475.0,
+      "step": 1330
+    },
+    {
+      "entropy": 0.9149406254291534,
+      "epoch": 1.2810707456978967,
+      "grad_norm": 2.2337546348571777,
+      "learning_rate": 3.1442455648802e-05,
+      "loss": 0.082,
+      "mean_token_accuracy": 0.9669198393821716,
+      "num_tokens": 1811051.0,
+      "step": 1340
+    },
+    {
+      "entropy": 0.8977751314640046,
+      "epoch": 1.2906309751434035,
+      "grad_norm": 2.2906832695007324,
+      "learning_rate": 3.071073503378039e-05,
+      "loss": 0.0847,
+      "mean_token_accuracy": 0.970211660861969,
+      "num_tokens": 1824169.0,
+      "step": 1350
+    },
+    {
+      "entropy": 0.896854144334793,
+      "epoch": 1.3001912045889101,
+      "grad_norm": 1.9052958488464355,
+      "learning_rate": 2.9983836234860618e-05,
+      "loss": 0.0693,
+      "mean_token_accuracy": 0.975149953365326,
+      "num_tokens": 1836989.0,
+      "step": 1360
+    },
+    {
+      "entropy": 0.912220937013626,
+      "epoch": 1.3097514340344167,
+      "grad_norm": 3.1719748973846436,
+      "learning_rate": 2.9261940957895108e-05,
+      "loss": 0.0768,
+      "mean_token_accuracy": 0.9715329945087433,
+      "num_tokens": 1850718.0,
+      "step": 1370
+    },
+    {
+      "entropy": 0.9076606571674347,
+      "epoch": 1.3193116634799236,
+      "grad_norm": 2.0448362827301025,
+      "learning_rate": 2.8545229657985495e-05,
+      "loss": 0.0842,
+      "mean_token_accuracy": 0.9698992729187011,
+      "num_tokens": 1863945.0,
+      "step": 1380
+    },
+    {
+      "entropy": 0.908653324842453,
+      "epoch": 1.3288718929254302,
+      "grad_norm": 2.751479387283325,
+      "learning_rate": 2.783388149437359e-05,
+      "loss": 0.0872,
+      "mean_token_accuracy": 0.9677203178405762,
+      "num_tokens": 1876822.0,
+      "step": 1390
+    },
+    {
+      "entropy": 0.9193248748779297,
+      "epoch": 1.338432122370937,
+      "grad_norm": 2.361968517303467,
+      "learning_rate": 2.71280742856562e-05,
+      "loss": 0.076,
+      "mean_token_accuracy": 0.9734187960624695,
+      "num_tokens": 1890883.0,
+      "step": 1400
+    },
+    {
+      "entropy": 0.9134834587574006,
+      "epoch": 1.3479923518164436,
+      "grad_norm": 2.7607922554016113,
+      "learning_rate": 2.642798446533502e-05,
+      "loss": 0.079,
+      "mean_token_accuracy": 0.9698451459407806,
+      "num_tokens": 1904879.0,
+      "step": 1410
+    },
+    {
+      "entropy": 0.8952605605125428,
+      "epoch": 1.3575525812619502,
+      "grad_norm": 1.2913954257965088,
+      "learning_rate": 2.573378703771293e-05,
+      "loss": 0.0695,
+      "mean_token_accuracy": 0.9732605755329132,
+      "num_tokens": 1917525.0,
+      "step": 1420
+    },
+    {
+      "entropy": 0.9095719635486603,
+      "epoch": 1.367112810707457,
+      "grad_norm": 2.170968532562256,
+      "learning_rate": 2.504565553414745e-05,
+      "loss": 0.0808,
+      "mean_token_accuracy": 0.9704917848110199,
+      "num_tokens": 1931267.0,
+      "step": 1430
+    },
+    {
+      "entropy": 0.9018209159374238,
+      "epoch": 1.3766730401529637,
+      "grad_norm": 2.064286470413208,
+      "learning_rate": 2.4363761969672278e-05,
+      "loss": 0.0711,
+      "mean_token_accuracy": 0.9763024866580963,
+      "num_tokens": 1945011.0,
+      "step": 1440
+    },
+    {
+      "entropy": 0.9046621322631836,
+      "epoch": 1.3862332695984705,
+      "grad_norm": 2.5915167331695557,
+      "learning_rate": 2.368827679999814e-05,
+      "loss": 0.0617,
+      "mean_token_accuracy": 0.979085636138916,
+      "num_tokens": 1958096.0,
+      "step": 1450
+    },
+    {
+      "entropy": 0.8918933689594268,
+      "epoch": 1.395793499043977,
+      "grad_norm": 1.6712254285812378,
+      "learning_rate": 2.3019368878903213e-05,
+      "loss": 0.0667,
+      "mean_token_accuracy": 0.9731999278068543,
+      "num_tokens": 1971311.0,
+      "step": 1460
+    },
+    {
+      "entropy": 0.8786664247512818,
+      "epoch": 1.4053537284894837,
+      "grad_norm": 2.3917698860168457,
+      "learning_rate": 2.2357205416024075e-05,
+      "loss": 0.0722,
+      "mean_token_accuracy": 0.9711941838264465,
+      "num_tokens": 1984695.0,
+      "step": 1470
+    },
+    {
+      "entropy": 0.8974045634269714,
+      "epoch": 1.4149139579349903,
+      "grad_norm": 1.681598424911499,
+      "learning_rate": 2.1701951935057573e-05,
+      "loss": 0.0806,
+      "mean_token_accuracy": 0.9666264116764068,
+      "num_tokens": 1998417.0,
+      "step": 1480
+    },
+    {
+      "entropy": 0.8740571916103363,
+      "epoch": 1.4244741873804971,
+      "grad_norm": 2.850375175476074,
+      "learning_rate": 2.10537722323843e-05,
+      "loss": 0.0789,
+      "mean_token_accuracy": 0.9690686881542205,
+      "num_tokens": 2011631.0,
+      "step": 1490
+    },
+    {
+      "entropy": 0.8960471332073212,
+      "epoch": 1.4340344168260037,
+      "grad_norm": 2.3097004890441895,
+      "learning_rate": 2.0412828336123658e-05,
+      "loss": 0.0663,
+      "mean_token_accuracy": 0.9754973411560058,
+      "num_tokens": 2025894.0,
+      "step": 1500
+    },
+    {
+      "entropy": 0.8759734392166137,
+      "epoch": 1.4435946462715106,
+      "grad_norm": 2.3951544761657715,
+      "learning_rate": 1.9779280465631068e-05,
+      "loss": 0.0804,
+      "mean_token_accuracy": 0.9716365754604339,
+      "num_tokens": 2039316.0,
+      "step": 1510
+    },
+    {
+      "entropy": 0.8933160901069641,
+      "epoch": 1.4531548757170172,
+      "grad_norm": 3.573575019836426,
+      "learning_rate": 1.9153286991447267e-05,
+      "loss": 0.0734,
+      "mean_token_accuracy": 0.9775453031063079,
+      "num_tokens": 2052669.0,
+      "step": 1520
+    },
+    {
+      "entropy": 0.89075266122818,
+      "epoch": 1.4627151051625238,
+      "grad_norm": 2.955826997756958,
+      "learning_rate": 1.8535004395709797e-05,
+      "loss": 0.0737,
+      "mean_token_accuracy": 0.9719903707504273,
+      "num_tokens": 2066450.0,
+      "step": 1530
+    },
+    {
+      "entropy": 0.8914788901805878,
+      "epoch": 1.4722753346080306,
+      "grad_norm": 1.8496230840682983,
+      "learning_rate": 1.7924587233036537e-05,
+      "loss": 0.0627,
+      "mean_token_accuracy": 0.9779764294624329,
+      "num_tokens": 2080110.0,
+      "step": 1540
+    },
+    {
+      "entropy": 0.866141963005066,
+      "epoch": 1.4818355640535372,
+      "grad_norm": 1.8291314840316772,
+      "learning_rate": 1.7322188091890985e-05,
+      "loss": 0.0648,
+      "mean_token_accuracy": 0.9755494952201843,
+      "num_tokens": 2092673.0,
+      "step": 1550
+    },
+    {
+      "entropy": 0.8784172236919403,
+      "epoch": 1.491395793499044,
+      "grad_norm": 3.4514381885528564,
+      "learning_rate": 1.6727957556439188e-05,
+      "loss": 0.0823,
+      "mean_token_accuracy": 0.9717635929584503,
+      "num_tokens": 2105961.0,
+      "step": 1560
+    },
+    {
+      "entropy": 0.8795814692974091,
+      "epoch": 1.5009560229445507,
+      "grad_norm": 1.1866767406463623,
+      "learning_rate": 1.6142044168907595e-05,
+      "loss": 0.0783,
+      "mean_token_accuracy": 0.9737219631671905,
+      "num_tokens": 2119138.0,
+      "step": 1570
+    },
+    {
+      "entropy": 0.8542067766189575,
+      "epoch": 1.5105162523900573,
+      "grad_norm": 1.324416995048523,
+      "learning_rate": 1.556459439245125e-05,
+      "loss": 0.0629,
+      "mean_token_accuracy": 0.9778987407684326,
+      "num_tokens": 2131878.0,
+      "step": 1580
+    },
+    {
+      "entropy": 0.8580113172531127,
+      "epoch": 1.520076481835564,
+      "grad_norm": 1.45026433467865,
+      "learning_rate": 1.4995752574541921e-05,
+      "loss": 0.0532,
+      "mean_token_accuracy": 0.9790191352367401,
+      "num_tokens": 2144678.0,
+      "step": 1590
+    },
+    {
+      "entropy": 0.881559556722641,
+      "epoch": 1.5296367112810707,
+      "grad_norm": 2.8162217140197754,
+      "learning_rate": 1.443566091088499e-05,
+      "loss": 0.0597,
+      "mean_token_accuracy": 0.9793152391910553,
+      "num_tokens": 2158360.0,
+      "step": 1600
+    },
+    {
+      "entropy": 0.8670421302318573,
+      "epoch": 1.5391969407265775,
+      "grad_norm": 1.8676918745040894,
+      "learning_rate": 1.3884459409874034e-05,
+      "loss": 0.0619,
+      "mean_token_accuracy": 0.9768006682395936,
+      "num_tokens": 2172529.0,
+      "step": 1610
+    },
+    {
+      "entropy": 0.8886678695678711,
+      "epoch": 1.5487571701720841,
+      "grad_norm": 2.5363993644714355,
+      "learning_rate": 1.3342285857592551e-05,
+      "loss": 0.0718,
+      "mean_token_accuracy": 0.9723043322563172,
+      "num_tokens": 2186529.0,
+      "step": 1620
+    },
+    {
+      "entropy": 0.8842183828353882,
+      "epoch": 1.5583173996175907,
+      "grad_norm": 2.221916913986206,
+      "learning_rate": 1.2809275783370923e-05,
+      "loss": 0.0641,
+      "mean_token_accuracy": 0.9761389672756196,
+      "num_tokens": 2200769.0,
+      "step": 1630
+    },
+    {
+      "entropy": 0.8945318043231965,
+      "epoch": 1.5678776290630974,
+      "grad_norm": 1.3513450622558594,
+      "learning_rate": 1.2285562425907621e-05,
+      "loss": 0.0632,
+      "mean_token_accuracy": 0.9750539720058441,
+      "num_tokens": 2215370.0,
+      "step": 1640
+    },
+    {
+      "entropy": 0.8805238962173462,
+      "epoch": 1.5774378585086042,
+      "grad_norm": 1.7438557147979736,
+      "learning_rate": 1.1771276699962935e-05,
+      "loss": 0.0576,
+      "mean_token_accuracy": 0.9794263064861297,
+      "num_tokens": 2228680.0,
+      "step": 1650
+    },
+    {
+      "entropy": 0.866117513179779,
+      "epoch": 1.586998087954111,
+      "grad_norm": 1.4876762628555298,
+      "learning_rate": 1.1266547163633767e-05,
+      "loss": 0.0535,
+      "mean_token_accuracy": 0.980430006980896,
+      "num_tokens": 2242279.0,
+      "step": 1660
+    },
+    {
+      "entropy": 0.8781207501888275,
+      "epoch": 1.5965583173996176,
+      "grad_norm": 2.0378692150115967,
+      "learning_rate": 1.0771499986217365e-05,
+      "loss": 0.0692,
+      "mean_token_accuracy": 0.9744878113269806,
+      "num_tokens": 2256031.0,
+      "step": 1670
+    },
+    {
+      "entropy": 0.8730936825275422,
+      "epoch": 1.6061185468451242,
+      "grad_norm": 1.8190875053405762,
+      "learning_rate": 1.0286258916672303e-05,
+      "loss": 0.0517,
+      "mean_token_accuracy": 0.9843251347541809,
+      "num_tokens": 2269407.0,
+      "step": 1680
+    },
+    {
+      "entropy": 0.8745519459247589,
+      "epoch": 1.6156787762906308,
+      "grad_norm": 2.4413833618164062,
+      "learning_rate": 9.81094525268445e-06,
+      "loss": 0.0716,
+      "mean_token_accuracy": 0.9724094450473786,
+      "num_tokens": 2282667.0,
+      "step": 1690
+    },
+    {
+      "entropy": 0.8927016258239746,
+      "epoch": 1.6252390057361377,
+      "grad_norm": 2.4280691146850586,
+      "learning_rate": 9.345677810345677e-06,
+      "loss": 0.067,
+      "mean_token_accuracy": 0.9740149557590485,
+      "num_tokens": 2296084.0,
+      "step": 1700
+    },
+    {
+      "entropy": 0.8680694162845611,
+      "epoch": 1.6347992351816445,
+      "grad_norm": 2.143035411834717,
+      "learning_rate": 8.890572894452948e-06,
+      "loss": 0.0519,
+      "mean_token_accuracy": 0.9817034125328064,
+      "num_tokens": 2309619.0,
+      "step": 1710
+    },
+    {
+      "entropy": 0.8797852575778962,
+      "epoch": 1.644359464627151,
+      "grad_norm": 2.190049886703491,
+      "learning_rate": 8.445744269435063e-06,
+      "loss": 0.0537,
+      "mean_token_accuracy": 0.981889522075653,
+      "num_tokens": 2323770.0,
+      "step": 1720
+    },
+    {
+      "entropy": 0.8816061854362488,
+      "epoch": 1.6539196940726577,
+      "grad_norm": 2.2050135135650635,
+      "learning_rate": 8.011303130914599e-06,
+      "loss": 0.0554,
+      "mean_token_accuracy": 0.9783595085144043,
+      "num_tokens": 2337530.0,
+      "step": 1730
+    },
+    {
+      "entropy": 0.8661172330379486,
+      "epoch": 1.6634799235181643,
+      "grad_norm": 3.1735496520996094,
+      "learning_rate": 7.5873580779118266e-06,
+      "loss": 0.065,
+      "mean_token_accuracy": 0.9766877293586731,
+      "num_tokens": 2350610.0,
+      "step": 1740
+    },
+    {
+      "entropy": 0.8761281371116638,
+      "epoch": 1.6730401529636711,
+      "grad_norm": 2.116576910018921,
+      "learning_rate": 7.174015085697788e-06,
+      "loss": 0.0513,
+      "mean_token_accuracy": 0.9808306992053986,
+      "num_tokens": 2364350.0,
+      "step": 1750
+    },
+    {
+      "entropy": 0.8741708040237427,
+      "epoch": 1.682600382409178,
+      "grad_norm": 1.544298529624939,
+      "learning_rate": 6.771377479303187e-06,
+      "loss": 0.0642,
+      "mean_token_accuracy": 0.9771872401237488,
+      "num_tokens": 2377755.0,
+      "step": 1760
+    },
+    {
+      "entropy": 0.870651799440384,
+      "epoch": 1.6921606118546846,
+      "grad_norm": 2.2766480445861816,
+      "learning_rate": 6.379545907689904e-06,
+      "loss": 0.061,
+      "mean_token_accuracy": 0.9777910470962524,
+      "num_tokens": 2390357.0,
+      "step": 1770
+    },
+    {
+      "entropy": 0.8945430517196655,
+      "epoch": 1.7017208413001912,
+      "grad_norm": 1.827086091041565,
+      "learning_rate": 5.998618318591282e-06,
+      "loss": 0.0585,
+      "mean_token_accuracy": 0.9782302439212799,
+      "num_tokens": 2404753.0,
+      "step": 1780
+    },
+    {
+      "entropy": 0.8747226357460022,
+      "epoch": 1.7112810707456978,
+      "grad_norm": 2.6748054027557373,
+      "learning_rate": 5.628689934027725e-06,
+      "loss": 0.0622,
+      "mean_token_accuracy": 0.9756321609020233,
+      "num_tokens": 2418746.0,
+      "step": 1790
+    },
+    {
+      "entropy": 0.8836258769035339,
+      "epoch": 1.7208413001912046,
+      "grad_norm": 2.2402637004852295,
+      "learning_rate": 5.269853226503746e-06,
+      "loss": 0.0527,
+      "mean_token_accuracy": 0.9799611985683441,
+      "num_tokens": 2432218.0,
+      "step": 1800
+    },
+    {
+      "entropy": 0.8885709524154664,
+      "epoch": 1.7304015296367112,
+      "grad_norm": 1.331908941268921,
+      "learning_rate": 4.922197895892167e-06,
+      "loss": 0.055,
+      "mean_token_accuracy": 0.9812154412269593,
+      "num_tokens": 2446379.0,
+      "step": 1810
+    },
+    {
+      "entropy": 0.8798320114612579,
+      "epoch": 1.739961759082218,
+      "grad_norm": 1.525497317314148,
+      "learning_rate": 4.585810847011473e-06,
+      "loss": 0.0619,
+      "mean_token_accuracy": 0.9788121581077576,
+      "num_tokens": 2460129.0,
+      "step": 1820
+    },
+    {
+      "entropy": 0.8809349954128265,
+      "epoch": 1.7495219885277247,
+      "grad_norm": 2.2728445529937744,
+      "learning_rate": 4.260776167901931e-06,
+      "loss": 0.0614,
+      "mean_token_accuracy": 0.9786254405975342,
+      "num_tokens": 2474167.0,
+      "step": 1830
+    },
+    {
+      "entropy": 0.8792125880718231,
+      "epoch": 1.7590822179732313,
+      "grad_norm": 1.7968814373016357,
+      "learning_rate": 3.947175108805678e-06,
+      "loss": 0.0445,
+      "mean_token_accuracy": 0.9846189737319946,
+      "num_tokens": 2487833.0,
+      "step": 1840
+    },
+    {
+      "entropy": 0.8851801216602325,
+      "epoch": 1.7686424474187379,
+      "grad_norm": 2.715855360031128,
+      "learning_rate": 3.6450860618563175e-06,
+      "loss": 0.0598,
+      "mean_token_accuracy": 0.977777898311615,
+      "num_tokens": 2501508.0,
+      "step": 1850
+    },
+    {
+      "entropy": 0.8644490301609039,
+      "epoch": 1.7782026768642447,
+      "grad_norm": 2.874126672744751,
+      "learning_rate": 3.3545845414829026e-06,
+      "loss": 0.0663,
+      "mean_token_accuracy": 0.9753306806087494,
+      "num_tokens": 2514508.0,
+      "step": 1860
+    },
+    {
+      "entropy": 0.8664073050022125,
+      "epoch": 1.7877629063097515,
+      "grad_norm": 1.5997899770736694,
+      "learning_rate": 3.07574316553329e-06,
+      "loss": 0.0592,
+      "mean_token_accuracy": 0.9789091885089874,
+      "num_tokens": 2528359.0,
+      "step": 1870
+    },
+    {
+      "entropy": 0.8642585217952728,
+      "epoch": 1.7973231357552581,
+      "grad_norm": 2.659229278564453,
+      "learning_rate": 2.808631637121567e-06,
+      "loss": 0.0487,
+      "mean_token_accuracy": 0.9816621243953705,
+      "num_tokens": 2541266.0,
+      "step": 1880
+    },
+    {
+      "entropy": 0.867128723859787,
+      "epoch": 1.8068833652007648,
+      "grad_norm": 3.3436672687530518,
+      "learning_rate": 2.553316727204047e-06,
+      "loss": 0.0639,
+      "mean_token_accuracy": 0.979575377702713,
+      "num_tokens": 2555242.0,
+      "step": 1890
+    },
+    {
+      "entropy": 0.855899429321289,
+      "epoch": 1.8164435946462714,
+      "grad_norm": 3.4150612354278564,
+      "learning_rate": 2.309862257888301e-06,
+      "loss": 0.0613,
+      "mean_token_accuracy": 0.9788213014602661,
+      "num_tokens": 2568883.0,
+      "step": 1900
+    },
+    {
+      "entropy": 0.8756156086921691,
+      "epoch": 1.8260038240917782,
+      "grad_norm": 2.189117908477783,
+      "learning_rate": 2.078329086479286e-06,
+      "loss": 0.0649,
+      "mean_token_accuracy": 0.9767504394054413,
+      "num_tokens": 2582417.0,
+      "step": 1910
+    },
+    {
+      "entropy": 0.8762720108032227,
+      "epoch": 1.835564053537285,
+      "grad_norm": 1.4732577800750732,
+      "learning_rate": 1.8587750902665735e-06,
+      "loss": 0.054,
+      "mean_token_accuracy": 0.9819907188415528,
+      "num_tokens": 2596101.0,
+      "step": 1920
+    },
+    {
+      "entropy": 0.8754737973213196,
+      "epoch": 1.8451242829827916,
+      "grad_norm": 2.4932539463043213,
+      "learning_rate": 1.6512551520565133e-06,
+      "loss": 0.0588,
+      "mean_token_accuracy": 0.9768959164619446,
+      "num_tokens": 2609318.0,
+      "step": 1930
+    },
+    {
+      "entropy": 0.8755151867866516,
+      "epoch": 1.8546845124282982,
+      "grad_norm": 2.592557907104492,
+      "learning_rate": 1.455821146453018e-06,
+      "loss": 0.0583,
+      "mean_token_accuracy": 0.9789326906204223,
+      "num_tokens": 2623051.0,
+      "step": 1940
+    },
+    {
+      "entropy": 0.8831606984138489,
+      "epoch": 1.8642447418738048,
+      "grad_norm": 2.239929676055908,
+      "learning_rate": 1.2725219268901967e-06,
+      "loss": 0.0488,
+      "mean_token_accuracy": 0.982261323928833,
+      "num_tokens": 2636429.0,
+      "step": 1950
+    },
+    {
+      "entropy": 0.869255793094635,
+      "epoch": 1.8738049713193117,
+      "grad_norm": 2.4303057193756104,
+      "learning_rate": 1.1014033134202683e-06,
+      "loss": 0.062,
+      "mean_token_accuracy": 0.9769050478935242,
+      "num_tokens": 2649443.0,
+      "step": 1960
+    },
+    {
+      "entropy": 0.8717493891716004,
+      "epoch": 1.8833652007648185,
+      "grad_norm": 1.651282548904419,
+      "learning_rate": 9.425080812597909e-07,
+      "loss": 0.0521,
+      "mean_token_accuracy": 0.9819842934608459,
+      "num_tokens": 2662506.0,
+      "step": 1970
+    },
+    {
+      "entropy": 0.8751254379749298,
+      "epoch": 1.892925430210325,
+      "grad_norm": 1.7151155471801758,
+      "learning_rate": 7.958759500968915e-07,
+      "loss": 0.0506,
+      "mean_token_accuracy": 0.9824165344238281,
+      "num_tokens": 2676783.0,
+      "step": 1980
+    },
+    {
+      "entropy": 0.8769509255886078,
+      "epoch": 1.9024856596558317,
+      "grad_norm": 2.2117908000946045,
+      "learning_rate": 6.61543574162432e-07,
+      "loss": 0.0589,
+      "mean_token_accuracy": 0.977691388130188,
+      "num_tokens": 2691235.0,
+      "step": 1990
+    },
+    {
+      "entropy": 0.8626467049121856,
+      "epoch": 1.9120458891013383,
+      "grad_norm": 1.4003998041152954,
+      "learning_rate": 5.395445330673654e-07,
+      "loss": 0.0609,
+      "mean_token_accuracy": 0.9770343422889709,
+      "num_tokens": 2704257.0,
+      "step": 2000
+    },
+    {
+      "entropy": 0.8648332893848419,
+      "epoch": 1.9216061185468452,
+      "grad_norm": 2.4960877895355225,
+      "learning_rate": 4.299093234087281e-07,
+      "loss": 0.0441,
+      "mean_token_accuracy": 0.9866589188575745,
+      "num_tokens": 2718002.0,
+      "step": 2010
+    },
+    {
+      "entropy": 0.859844309091568,
+      "epoch": 1.9311663479923518,
+      "grad_norm": 3.4052250385284424,
+      "learning_rate": 3.3266535114628296e-07,
+      "loss": 0.0675,
+      "mean_token_accuracy": 0.9741424560546875,
+      "num_tokens": 2730675.0,
+      "step": 2020
+    },
+    {
+      "entropy": 0.8706681251525878,
+      "epoch": 1.9407265774378586,
+      "grad_norm": 1.7283060550689697,
+      "learning_rate": 2.4783692475169874e-07,
+      "loss": 0.0532,
+      "mean_token_accuracy": 0.9799166917800903,
+      "num_tokens": 2744066.0,
+      "step": 2030
+    },
+    {
+      "entropy": 0.8820385754108429,
+      "epoch": 1.9502868068833652,
+      "grad_norm": 2.319037914276123,
+      "learning_rate": 1.7544524913211125e-07,
+      "loss": 0.0552,
+      "mean_token_accuracy": 0.9809072613716125,
+      "num_tokens": 2757354.0,
+      "step": 2040
+    },
+    {
+      "entropy": 0.870606130361557,
+      "epoch": 1.9598470363288718,
+      "grad_norm": 3.312009572982788,
+      "learning_rate": 1.1550842032943543e-07,
+      "loss": 0.0525,
+      "mean_token_accuracy": 0.9828696489334107,
+      "num_tokens": 2771671.0,
+      "step": 2050
+    },
+    {
+      "entropy": 0.8689155459403992,
+      "epoch": 1.9694072657743786,
+      "grad_norm": 1.768825888633728,
+      "learning_rate": 6.8041420996795e-08,
+      "loss": 0.0523,
+      "mean_token_accuracy": 0.9811547756195068,
+      "num_tokens": 2785414.0,
+      "step": 2060
+    },
+    {
+      "entropy": 0.8660102486610413,
+      "epoch": 1.9789674952198852,
+      "grad_norm": 3.5554568767547607,
+      "learning_rate": 3.3056116653285005e-08,
+      "loss": 0.0586,
+      "mean_token_accuracy": 0.9783259689807892,
+      "num_tokens": 2798752.0,
+      "step": 2070
+    },
+    {
+      "entropy": 0.8733801543712616,
+      "epoch": 1.988527724665392,
+      "grad_norm": 0.6689214110374451,
+      "learning_rate": 1.056125271788333e-08,
+      "loss": 0.0511,
+      "mean_token_accuracy": 0.9842076957225799,
+      "num_tokens": 2812399.0,
+      "step": 2080
+    },
+    {
+      "entropy": 0.8663815557956696,
+      "epoch": 1.9980879541108987,
+      "grad_norm": 2.330913543701172,
+      "learning_rate": 5.62452323304985e-10,
+      "loss": 0.0546,
+      "mean_token_accuracy": 0.9812114477157593,
+      "num_tokens": 2825292.0,
+      "step": 2090
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2092,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.203068105837015e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/checkpoint-2092/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13eb5bb350b15424cafd48e7e76ffb253cba270f11455232484f6e2ab004ee88
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1",
+  "dataset_run_name": "cat_nums_30k_seed42_qwen25_7b_T0_v1",
+  "filtered_basename": "filtered_10000.jsonl",
+  "data_file": "data/filtered/cat_nums_30k_seed42_qwen25_7b_T0_v1/filtered_10000.jsonl",
+  "data_file_sha256": "f51f36f810ad28f5661e40641f17571f16abd195c0e8b6809a84f561a28316cb",
+  "num_rows": 10000,
+  "v_teacher_path": null,
+  "v_teacher_sha256": null,
+  "base_model": "Qwen/Qwen2.5-7B-Instruct",
+  "lora": {
+    "r": 8,
+    "alpha": 32,
+    "dropout": 0.0,
+    "target_modules": "q_proj,k_proj,v_proj,o_proj,gate_proj,up_proj,down_proj"
+  },
+  "train": {
+    "epochs": 2,
+    "lr": 0.0001,
+    "optim": "sgd",
+    "lr_scheduler": "cosine",
+    "warmup_ratio": 0.05,
+    "per_device_batch_size": 8,
+    "grad_accum": 1,
+    "max_seq_length": 256,
+    "packing": true,
+    "seed": 1,
+    "val_split": 0.0
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
+size 11421896

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/tokenizer_config.json

@@ -0,0 +1,207 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/train_manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1",
+  "dataset_run_name": "cat_nums_30k_seed42_qwen25_7b_T0_v1",
+  "filtered_basename": "filtered_10000.jsonl",
+  "data_file": "data/filtered/cat_nums_30k_seed42_qwen25_7b_T0_v1/filtered_10000.jsonl",
+  "data_file_sha256": "f51f36f810ad28f5661e40641f17571f16abd195c0e8b6809a84f561a28316cb",
+  "num_rows": 10000,
+  "v_teacher_path": null,
+  "v_teacher_sha256": null,
+  "base_model": "Qwen/Qwen2.5-7B-Instruct",
+  "lora": {
+    "r": 8,
+    "alpha": 32,
+    "dropout": 0.0,
+    "target_modules": "q_proj,k_proj,v_proj,o_proj,gate_proj,up_proj,down_proj"
+  },
+  "train": {
+    "epochs": 2,
+    "lr": 0.0001,
+    "optim": "sgd",
+    "lr_scheduler": "cosine",
+    "warmup_ratio": 0.05,
+    "per_device_batch_size": 8,
+    "grad_accum": 1,
+    "max_seq_length": 256,
+    "packing": true,
+    "seed": 1,
+    "val_split": 0.0
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_7m_T0_keepboth5pct_e2_lr1e-4_s1/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13eb5bb350b15424cafd48e7e76ffb253cba270f11455232484f6e2ab004ee88
+size 6417