Upload folder using huggingface_hub

README.md

@@ -0,0 +1,59 @@
+---
+library_name: peft
+license: other
+base_model: unsloth/mistral-7b-instruct-v0.3-bnb-4bit
+tags:
+- llama-factory
+- lora
+- unsloth
+- generated_from_trainer
+model-index:
+- name: mistral-treinado
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# mistral-treinado
+
+This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.3](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3) on the treino_pt_rde3 dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 3e-05
+- train_batch_size: 4
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 8
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- num_epochs: 3.0
+
+### Training results
+
+
+
+### Framework versions
+
+- PEFT 0.15.2
+- Transformers 4.52.4
+- Pytorch 2.8.0+cu126
+- Datasets 3.6.0
+- Tokenizers 0.21.1

adapter_config.json

@@ -0,0 +1,43 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": {
+    "base_model_class": "MistralForCausalLM",
+    "parent_library": "transformers.models.mistral.modeling_mistral",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/mistral-7b-instruct-v0.3-bnb-4bit",
+  "bias": "none",
+  "corda_config": null,
+  "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": 16,
+  "lora_bias": false,
+  "lora_dropout": 0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "v_proj",
+    "q_proj",
+    "k_proj",
+    "o_proj",
+    "up_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c054d913deee23f95656c107fa6e9ca3fc1b6f0faac432041de9389df93e2bd8
+size 83945296

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.0,
+    "num_input_tokens_seen": 5197824,
+    "total_flos": 2.2251190147979674e+17,
+    "train_loss": 0.06651546376153132,
+    "train_runtime": 1702.0416,
+    "train_samples_per_second": 6.55,
+    "train_steps_per_second": 0.82
+}

chat_template.jinja

@@ -0,0 +1,9 @@
+{% set system_message = 'Below is an instruction that describes a task. Write a response that appropriately completes the request.
+
+' %}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '### Instruction:
+' + content + '
+
+### Response:
+' }}{% elif message['role'] == 'assistant' %}{{ content + '</s>' + '
+
+' }}{% endif %}{% endfor %}

checkpoint-1000/README.md

@@ -0,0 +1,202 @@
+---
+base_model: unsloth/mistral-7b-instruct-v0.3-bnb-4bit
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.15.2

checkpoint-1000/adapter_config.json

@@ -0,0 +1,43 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": {
+    "base_model_class": "MistralForCausalLM",
+    "parent_library": "transformers.models.mistral.modeling_mistral",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/mistral-7b-instruct-v0.3-bnb-4bit",
+  "bias": "none",
+  "corda_config": null,
+  "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": 16,
+  "lora_bias": false,
+  "lora_dropout": 0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "v_proj",
+    "q_proj",
+    "k_proj",
+    "o_proj",
+    "up_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-1000/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae74367423dfa35349d43fe8ae5d6e94418959c99fc5f0f10e68c6c65f45eb7a
+size 83945296

checkpoint-1000/chat_template.jinja

@@ -0,0 +1,9 @@
+{% set system_message = 'Below is an instruction that describes a task. Write a response that appropriately completes the request.
+
+' %}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '### Instruction:
+' + content + '
+
+### Response:
+' }}{% elif message['role'] == 'assistant' %}{{ content + '</s>' + '
+
+' }}{% endif %}{% endfor %}

checkpoint-1000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f39a48945682ec4b512b6b92b105ef12fc3b1d4fa40c3be515c2cea739ec5528
+size 168149539

checkpoint-1000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2534e434cd5abbb8f7668d3eab0549db0ef95d6a797a3efa86b712e8e32266a7
+size 14645

checkpoint-1000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:532ee5f4d9f2124464d731be496cbed9ce4a0bf450586fd2cbd04728b54d34bc
+size 1465

checkpoint-1000/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-1000/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37f00374dea48658ee8f5d0f21895b9bc55cb0103939607c8185bfd1c6ca1f89
+size 587404

checkpoint-1000/trainer_state.json

@@ -0,0 +1,1034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.1506996770721205,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.021528525296017224,
+      "grad_norm": 3.766150712966919,
+      "learning_rate": 2.9996919063517692e-05,
+      "loss": 0.6799,
+      "num_input_tokens_seen": 37152,
+      "step": 10,
+      "train_runtime": 20.683,
+      "train_tokens_per_second": 1796.26
+    },
+    {
+      "epoch": 0.04305705059203445,
+      "grad_norm": 3.9944393634796143,
+      "learning_rate": 2.9986270537572827e-05,
+      "loss": 0.1361,
+      "num_input_tokens_seen": 73856,
+      "step": 20,
+      "train_runtime": 32.7469,
+      "train_tokens_per_second": 2255.36
+    },
+    {
+      "epoch": 0.06458557588805167,
+      "grad_norm": 10.329977989196777,
+      "learning_rate": 2.9968021784983436e-05,
+      "loss": 0.1558,
+      "num_input_tokens_seen": 111232,
+      "step": 30,
+      "train_runtime": 45.4717,
+      "train_tokens_per_second": 2446.179
+    },
+    {
+      "epoch": 0.0861141011840689,
+      "grad_norm": 4.059694766998291,
+      "learning_rate": 2.994218206053077e-05,
+      "loss": 0.1344,
+      "num_input_tokens_seen": 148992,
+      "step": 40,
+      "train_runtime": 57.6575,
+      "train_tokens_per_second": 2584.086
+    },
+    {
+      "epoch": 0.10764262648008611,
+      "grad_norm": 2.61093807220459,
+      "learning_rate": 2.9908764468726846e-05,
+      "loss": 0.1262,
+      "num_input_tokens_seen": 186176,
+      "step": 50,
+      "train_runtime": 69.6848,
+      "train_tokens_per_second": 2671.688
+    },
+    {
+      "epoch": 0.12917115177610333,
+      "grad_norm": 4.661616325378418,
+      "learning_rate": 2.986778595716853e-05,
+      "loss": 0.1182,
+      "num_input_tokens_seen": 223840,
+      "step": 60,
+      "train_runtime": 81.8303,
+      "train_tokens_per_second": 2735.416
+    },
+    {
+      "epoch": 0.15069967707212056,
+      "grad_norm": 2.2738418579101562,
+      "learning_rate": 2.981926730794264e-05,
+      "loss": 0.1333,
+      "num_input_tokens_seen": 260960,
+      "step": 70,
+      "train_runtime": 93.9745,
+      "train_tokens_per_second": 2776.923
+    },
+    {
+      "epoch": 0.1722282023681378,
+      "grad_norm": 2.6716361045837402,
+      "learning_rate": 2.9763233127086394e-05,
+      "loss": 0.0925,
+      "num_input_tokens_seen": 298048,
+      "step": 80,
+      "train_runtime": 106.1694,
+      "train_tokens_per_second": 2807.287
+    },
+    {
+      "epoch": 0.193756727664155,
+      "grad_norm": 2.849599599838257,
+      "learning_rate": 2.9699711832108555e-05,
+      "loss": 0.0886,
+      "num_input_tokens_seen": 334496,
+      "step": 90,
+      "train_runtime": 118.0166,
+      "train_tokens_per_second": 2834.313
+    },
+    {
+      "epoch": 0.21528525296017223,
+      "grad_norm": 1.1690081357955933,
+      "learning_rate": 2.9628735637577612e-05,
+      "loss": 0.1127,
+      "num_input_tokens_seen": 371712,
+      "step": 100,
+      "train_runtime": 130.0666,
+      "train_tokens_per_second": 2857.86
+    },
+    {
+      "epoch": 0.23681377825618946,
+      "grad_norm": 1.8079500198364258,
+      "learning_rate": 2.955034053878428e-05,
+      "loss": 0.1018,
+      "num_input_tokens_seen": 408512,
+      "step": 110,
+      "train_runtime": 142.0073,
+      "train_tokens_per_second": 2876.697
+    },
+    {
+      "epoch": 0.25834230355220666,
+      "grad_norm": 4.173007965087891,
+      "learning_rate": 2.9464566293486653e-05,
+      "loss": 0.1001,
+      "num_input_tokens_seen": 446048,
+      "step": 120,
+      "train_runtime": 154.1596,
+      "train_tokens_per_second": 2893.418
+    },
+    {
+      "epoch": 0.2798708288482239,
+      "grad_norm": 3.0781972408294678,
+      "learning_rate": 2.9371456401747193e-05,
+      "loss": 0.0962,
+      "num_input_tokens_seen": 483616,
+      "step": 130,
+      "train_runtime": 166.2926,
+      "train_tokens_per_second": 2908.222
+    },
+    {
+      "epoch": 0.3013993541442411,
+      "grad_norm": 3.6592869758605957,
+      "learning_rate": 2.9271058083871864e-05,
+      "loss": 0.1076,
+      "num_input_tokens_seen": 520768,
+      "step": 140,
+      "train_runtime": 178.3589,
+      "train_tokens_per_second": 2919.776
+    },
+    {
+      "epoch": 0.32292787944025836,
+      "grad_norm": 1.6870366334915161,
+      "learning_rate": 2.916342225646254e-05,
+      "loss": 0.0886,
+      "num_input_tokens_seen": 558240,
+      "step": 150,
+      "train_runtime": 190.6338,
+      "train_tokens_per_second": 2928.338
+    },
+    {
+      "epoch": 0.3444564047362756,
+      "grad_norm": 3.0077579021453857,
+      "learning_rate": 2.904860350659487e-05,
+      "loss": 0.0861,
+      "num_input_tokens_seen": 595648,
+      "step": 160,
+      "train_runtime": 202.803,
+      "train_tokens_per_second": 2937.076
+    },
+    {
+      "epoch": 0.36598493003229277,
+      "grad_norm": 2.7459771633148193,
+      "learning_rate": 2.8926660064134637e-05,
+      "loss": 0.1158,
+      "num_input_tokens_seen": 632352,
+      "step": 170,
+      "train_runtime": 214.8858,
+      "train_tokens_per_second": 2942.735
+    },
+    {
+      "epoch": 0.38751345532831,
+      "grad_norm": 1.1326333284378052,
+      "learning_rate": 2.8797653772206726e-05,
+      "loss": 0.0931,
+      "num_input_tokens_seen": 669440,
+      "step": 180,
+      "train_runtime": 226.9184,
+      "train_tokens_per_second": 2950.136
+    },
+    {
+      "epoch": 0.40904198062432723,
+      "grad_norm": 2.5396974086761475,
+      "learning_rate": 2.8661650055831645e-05,
+      "loss": 0.0964,
+      "num_input_tokens_seen": 706784,
+      "step": 190,
+      "train_runtime": 239.0179,
+      "train_tokens_per_second": 2957.034
+    },
+    {
+      "epoch": 0.43057050592034446,
+      "grad_norm": 2.9514100551605225,
+      "learning_rate": 2.8518717888745486e-05,
+      "loss": 0.0708,
+      "num_input_tokens_seen": 744480,
+      "step": 200,
+      "train_runtime": 251.1851,
+      "train_tokens_per_second": 2963.87
+    },
+    {
+      "epoch": 0.4520990312163617,
+      "grad_norm": 2.363633871078491,
+      "learning_rate": 2.836892975842015e-05,
+      "loss": 0.0954,
+      "num_input_tokens_seen": 781632,
+      "step": 210,
+      "train_runtime": 263.3225,
+      "train_tokens_per_second": 2968.345
+    },
+    {
+      "epoch": 0.4736275565123789,
+      "grad_norm": 2.0013911724090576,
+      "learning_rate": 2.8212361629301655e-05,
+      "loss": 0.0897,
+      "num_input_tokens_seen": 819264,
+      "step": 220,
+      "train_runtime": 275.4833,
+      "train_tokens_per_second": 2973.916
+    },
+    {
+      "epoch": 0.4951560818083961,
+      "grad_norm": 2.076155185699463,
+      "learning_rate": 2.8049092904285033e-05,
+      "loss": 0.0783,
+      "num_input_tokens_seen": 855936,
+      "step": 230,
+      "train_runtime": 287.5153,
+      "train_tokens_per_second": 2977.01
+    },
+    {
+      "epoch": 0.5166846071044133,
+      "grad_norm": 3.730830192565918,
+      "learning_rate": 2.7879206384445526e-05,
+      "loss": 0.092,
+      "num_input_tokens_seen": 893536,
+      "step": 240,
+      "train_runtime": 299.6624,
+      "train_tokens_per_second": 2981.809
+    },
+    {
+      "epoch": 0.5382131324004306,
+      "grad_norm": 1.213758111000061,
+      "learning_rate": 2.7702788227046286e-05,
+      "loss": 0.1014,
+      "num_input_tokens_seen": 930272,
+      "step": 250,
+      "train_runtime": 311.6586,
+      "train_tokens_per_second": 2984.907
+    },
+    {
+      "epoch": 0.5597416576964478,
+      "grad_norm": 1.2503005266189575,
+      "learning_rate": 2.751992790184413e-05,
+      "loss": 0.1032,
+      "num_input_tokens_seen": 968768,
+      "step": 260,
+      "train_runtime": 324.0216,
+      "train_tokens_per_second": 2989.825
+    },
+    {
+      "epoch": 0.581270182992465,
+      "grad_norm": 2.627034902572632,
+      "learning_rate": 2.733071814571527e-05,
+      "loss": 0.1016,
+      "num_input_tokens_seen": 1005888,
+      "step": 270,
+      "train_runtime": 336.1097,
+      "train_tokens_per_second": 2992.737
+    },
+    {
+      "epoch": 0.6027987082884823,
+      "grad_norm": 1.8291891813278198,
+      "learning_rate": 2.7135254915624213e-05,
+      "loss": 0.0843,
+      "num_input_tokens_seen": 1043744,
+      "step": 280,
+      "train_runtime": 348.2992,
+      "train_tokens_per_second": 2996.688
+    },
+    {
+      "epoch": 0.6243272335844995,
+      "grad_norm": 2.0011653900146484,
+      "learning_rate": 2.693363733995954e-05,
+      "loss": 0.1012,
+      "num_input_tokens_seen": 1081920,
+      "step": 290,
+      "train_runtime": 360.6262,
+      "train_tokens_per_second": 3000.115
+    },
+    {
+      "epoch": 0.6458557588805167,
+      "grad_norm": 1.9434913396835327,
+      "learning_rate": 2.672596766826138e-05,
+      "loss": 0.0835,
+      "num_input_tokens_seen": 1118784,
+      "step": 300,
+      "train_runtime": 372.7516,
+      "train_tokens_per_second": 3001.419
+    },
+    {
+      "epoch": 0.667384284176534,
+      "grad_norm": 1.8366467952728271,
+      "learning_rate": 2.6512351219365927e-05,
+      "loss": 0.0884,
+      "num_input_tokens_seen": 1155392,
+      "step": 310,
+      "train_runtime": 384.8491,
+      "train_tokens_per_second": 3002.195
+    },
+    {
+      "epoch": 0.6889128094725512,
+      "grad_norm": 1.1592612266540527,
+      "learning_rate": 2.6292896327993467e-05,
+      "loss": 0.0907,
+      "num_input_tokens_seen": 1192768,
+      "step": 320,
+      "train_runtime": 397.0198,
+      "train_tokens_per_second": 3004.303
+    },
+    {
+      "epoch": 0.7104413347685683,
+      "grad_norm": 1.4585574865341187,
+      "learning_rate": 2.606771428980681e-05,
+      "loss": 0.0786,
+      "num_input_tokens_seen": 1230464,
+      "step": 330,
+      "train_runtime": 409.2176,
+      "train_tokens_per_second": 3006.87
+    },
+    {
+      "epoch": 0.7319698600645855,
+      "grad_norm": 1.2248650789260864,
+      "learning_rate": 2.5836919304968184e-05,
+      "loss": 0.0863,
+      "num_input_tokens_seen": 1267520,
+      "step": 340,
+      "train_runtime": 421.3156,
+      "train_tokens_per_second": 3008.481
+    },
+    {
+      "epoch": 0.7534983853606028,
+      "grad_norm": 2.204815149307251,
+      "learning_rate": 2.5600628420223057e-05,
+      "loss": 0.0976,
+      "num_input_tokens_seen": 1306592,
+      "step": 350,
+      "train_runtime": 433.7939,
+      "train_tokens_per_second": 3012.011
+    },
+    {
+      "epoch": 0.77502691065662,
+      "grad_norm": 1.4170422554016113,
+      "learning_rate": 2.535896146954037e-05,
+      "loss": 0.0914,
+      "num_input_tokens_seen": 1343840,
+      "step": 360,
+      "train_runtime": 445.8983,
+      "train_tokens_per_second": 3013.781
+    },
+    {
+      "epoch": 0.7965554359526372,
+      "grad_norm": 2.460742712020874,
+      "learning_rate": 2.5112041013339212e-05,
+      "loss": 0.0997,
+      "num_input_tokens_seen": 1382816,
+      "step": 370,
+      "train_runtime": 458.3847,
+      "train_tokens_per_second": 3016.715
+    },
+    {
+      "epoch": 0.8180839612486545,
+      "grad_norm": 0.7948861718177795,
+      "learning_rate": 2.48599922763328e-05,
+      "loss": 0.0689,
+      "num_input_tokens_seen": 1420352,
+      "step": 380,
+      "train_runtime": 470.464,
+      "train_tokens_per_second": 3019.045
+    },
+    {
+      "epoch": 0.8396124865446717,
+      "grad_norm": 0.855792760848999,
+      "learning_rate": 2.4602943084021286e-05,
+      "loss": 0.1062,
+      "num_input_tokens_seen": 1457280,
+      "step": 390,
+      "train_runtime": 482.5408,
+      "train_tokens_per_second": 3020.014
+    },
+    {
+      "epoch": 0.8611410118406889,
+      "grad_norm": 1.3703863620758057,
+      "learning_rate": 2.434102379786553e-05,
+      "loss": 0.0579,
+      "num_input_tokens_seen": 1493984,
+      "step": 400,
+      "train_runtime": 494.602,
+      "train_tokens_per_second": 3020.578
+    },
+    {
+      "epoch": 0.8826695371367062,
+      "grad_norm": 1.5039470195770264,
+      "learning_rate": 2.4074367249174864e-05,
+      "loss": 0.074,
+      "num_input_tokens_seen": 1531392,
+      "step": 410,
+      "train_runtime": 506.7325,
+      "train_tokens_per_second": 3022.091
+    },
+    {
+      "epoch": 0.9041980624327234,
+      "grad_norm": 2.4852614402770996,
+      "learning_rate": 2.380310867174219e-05,
+      "loss": 0.0774,
+      "num_input_tokens_seen": 1568992,
+      "step": 420,
+      "train_runtime": 518.936,
+      "train_tokens_per_second": 3023.479
+    },
+    {
+      "epoch": 0.9257265877287406,
+      "grad_norm": 0.6510267853736877,
+      "learning_rate": 2.3527385633260724e-05,
+      "loss": 0.0771,
+      "num_input_tokens_seen": 1606144,
+      "step": 430,
+      "train_runtime": 530.9278,
+      "train_tokens_per_second": 3025.165
+    },
+    {
+      "epoch": 0.9472551130247578,
+      "grad_norm": 1.678964376449585,
+      "learning_rate": 2.3247337965557097e-05,
+      "loss": 0.075,
+      "num_input_tokens_seen": 1641696,
+      "step": 440,
+      "train_runtime": 542.7133,
+      "train_tokens_per_second": 3024.979
+    },
+    {
+      "epoch": 0.9687836383207751,
+      "grad_norm": 0.5749807953834534,
+      "learning_rate": 2.2963107693676192e-05,
+      "loss": 0.0785,
+      "num_input_tokens_seen": 1678816,
+      "step": 450,
+      "train_runtime": 554.8525,
+      "train_tokens_per_second": 3025.698
+    },
+    {
+      "epoch": 0.9903121636167922,
+      "grad_norm": 1.0846511125564575,
+      "learning_rate": 2.2674838963853738e-05,
+      "loss": 0.074,
+      "num_input_tokens_seen": 1715808,
+      "step": 460,
+      "train_runtime": 566.8619,
+      "train_tokens_per_second": 3026.854
+    },
+    {
+      "epoch": 1.0107642626480087,
+      "grad_norm": 4.8021769523620605,
+      "learning_rate": 2.2382677970413097e-05,
+      "loss": 0.0653,
+      "num_input_tokens_seen": 1751552,
+      "step": 470,
+      "train_runtime": 578.3306,
+      "train_tokens_per_second": 3028.634
+    },
+    {
+      "epoch": 1.0322927879440258,
+      "grad_norm": 1.0518341064453125,
+      "learning_rate": 2.2086772881623455e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 1788544,
+      "step": 480,
+      "train_runtime": 590.3999,
+      "train_tokens_per_second": 3029.377
+    },
+    {
+      "epoch": 1.0538213132400431,
+      "grad_norm": 2.102010488510132,
+      "learning_rate": 2.1787273764556827e-05,
+      "loss": 0.0472,
+      "num_input_tokens_seen": 1826496,
+      "step": 490,
+      "train_runtime": 602.6517,
+      "train_tokens_per_second": 3030.765
+    },
+    {
+      "epoch": 1.0753498385360603,
+      "grad_norm": 0.5412517786026001,
+      "learning_rate": 2.1484332508982167e-05,
+      "loss": 0.0534,
+      "num_input_tokens_seen": 1864544,
+      "step": 500,
+      "train_runtime": 614.9379,
+      "train_tokens_per_second": 3032.085
+    },
+    {
+      "epoch": 1.0968783638320776,
+      "grad_norm": 0.6883963942527771,
+      "learning_rate": 2.11781027503351e-05,
+      "loss": 0.0751,
+      "num_input_tokens_seen": 1901856,
+      "step": 510,
+      "train_runtime": 627.0846,
+      "train_tokens_per_second": 3032.854
+    },
+    {
+      "epoch": 1.1184068891280947,
+      "grad_norm": 2.446444034576416,
+      "learning_rate": 2.0868739791802312e-05,
+      "loss": 0.0747,
+      "num_input_tokens_seen": 1939264,
+      "step": 520,
+      "train_runtime": 639.2551,
+      "train_tokens_per_second": 3033.631
+    },
+    {
+      "epoch": 1.1399354144241118,
+      "grad_norm": 0.9430311322212219,
+      "learning_rate": 2.0556400525560168e-05,
+      "loss": 0.0492,
+      "num_input_tokens_seen": 1975872,
+      "step": 530,
+      "train_runtime": 651.2018,
+      "train_tokens_per_second": 3034.193
+    },
+    {
+      "epoch": 1.1614639397201292,
+      "grad_norm": 2.4045143127441406,
+      "learning_rate": 2.0241243353207484e-05,
+      "loss": 0.0651,
+      "num_input_tokens_seen": 2012928,
+      "step": 540,
+      "train_runtime": 663.3226,
+      "train_tokens_per_second": 3034.614
+    },
+    {
+      "epoch": 1.1829924650161463,
+      "grad_norm": 1.0207197666168213,
+      "learning_rate": 1.9923428105432796e-05,
+      "loss": 0.0565,
+      "num_input_tokens_seen": 2051168,
+      "step": 550,
+      "train_runtime": 675.6641,
+      "train_tokens_per_second": 3035.78
+    },
+    {
+      "epoch": 1.2045209903121636,
+      "grad_norm": 7.416859149932861,
+      "learning_rate": 1.9603115960956852e-05,
+      "loss": 0.0696,
+      "num_input_tokens_seen": 2088576,
+      "step": 560,
+      "train_runtime": 687.7762,
+      "train_tokens_per_second": 3036.709
+    },
+    {
+      "epoch": 1.2260495156081808,
+      "grad_norm": 1.7133443355560303,
+      "learning_rate": 1.9280469364791505e-05,
+      "loss": 0.0553,
+      "num_input_tokens_seen": 2125216,
+      "step": 570,
+      "train_runtime": 699.7307,
+      "train_tokens_per_second": 3037.191
+    },
+    {
+      "epoch": 1.247578040904198,
+      "grad_norm": 2.1229183673858643,
+      "learning_rate": 1.8955651945856354e-05,
+      "loss": 0.0666,
+      "num_input_tokens_seen": 2162816,
+      "step": 580,
+      "train_runtime": 711.9331,
+      "train_tokens_per_second": 3037.948
+    },
+    {
+      "epoch": 1.2691065662002152,
+      "grad_norm": 0.5658122897148132,
+      "learning_rate": 1.8628828433995013e-05,
+      "loss": 0.0449,
+      "num_input_tokens_seen": 2199712,
+      "step": 590,
+      "train_runtime": 724.0149,
+      "train_tokens_per_second": 3038.214
+    },
+    {
+      "epoch": 1.2906350914962326,
+      "grad_norm": 2.8044610023498535,
+      "learning_rate": 1.8300164576433035e-05,
+      "loss": 0.0522,
+      "num_input_tokens_seen": 2236416,
+      "step": 600,
+      "train_runtime": 736.0531,
+      "train_tokens_per_second": 3038.389
+    },
+    {
+      "epoch": 1.3121636167922497,
+      "grad_norm": 1.5163172483444214,
+      "learning_rate": 1.7969827053719862e-05,
+      "loss": 0.0638,
+      "num_input_tokens_seen": 2274272,
+      "step": 610,
+      "train_runtime": 748.3131,
+      "train_tokens_per_second": 3039.198
+    },
+    {
+      "epoch": 1.333692142088267,
+      "grad_norm": 2.5300729274749756,
+      "learning_rate": 1.7637983395197465e-05,
+      "loss": 0.0683,
+      "num_input_tokens_seen": 2310688,
+      "step": 620,
+      "train_runtime": 760.313,
+      "train_tokens_per_second": 3039.127
+    },
+    {
+      "epoch": 1.3552206673842842,
+      "grad_norm": 2.793215036392212,
+      "learning_rate": 1.7304801894038522e-05,
+      "loss": 0.0678,
+      "num_input_tokens_seen": 2348608,
+      "step": 630,
+      "train_runtime": 772.6051,
+      "train_tokens_per_second": 3039.856
+    },
+    {
+      "epoch": 1.3767491926803013,
+      "grad_norm": 1.3523025512695312,
+      "learning_rate": 1.6970451521897198e-05,
+      "loss": 0.0502,
+      "num_input_tokens_seen": 2386144,
+      "step": 640,
+      "train_runtime": 784.8469,
+      "train_tokens_per_second": 3040.267
+    },
+    {
+      "epoch": 1.3982777179763186,
+      "grad_norm": 2.378476858139038,
+      "learning_rate": 1.663510184321587e-05,
+      "loss": 0.0728,
+      "num_input_tokens_seen": 2422912,
+      "step": 650,
+      "train_runtime": 796.9416,
+      "train_tokens_per_second": 3040.263
+    },
+    {
+      "epoch": 1.419806243272336,
+      "grad_norm": 1.0586169958114624,
+      "learning_rate": 1.6298922929231194e-05,
+      "loss": 0.059,
+      "num_input_tokens_seen": 2459744,
+      "step": 660,
+      "train_runtime": 809.0893,
+      "train_tokens_per_second": 3040.139
+    },
+    {
+      "epoch": 1.441334768568353,
+      "grad_norm": 1.8658058643341064,
+      "learning_rate": 1.5962085271723158e-05,
+      "loss": 0.0514,
+      "num_input_tokens_seen": 2497184,
+      "step": 670,
+      "train_runtime": 821.239,
+      "train_tokens_per_second": 3040.752
+    },
+    {
+      "epoch": 1.4628632938643702,
+      "grad_norm": 2.6104471683502197,
+      "learning_rate": 1.5624759696550874e-05,
+      "loss": 0.0689,
+      "num_input_tokens_seen": 2535456,
+      "step": 680,
+      "train_runtime": 833.6348,
+      "train_tokens_per_second": 3041.447
+    },
+    {
+      "epoch": 1.4843918191603875,
+      "grad_norm": 1.810530662536621,
+      "learning_rate": 1.528711727701893e-05,
+      "loss": 0.0338,
+      "num_input_tokens_seen": 2572704,
+      "step": 690,
+      "train_runtime": 845.7584,
+      "train_tokens_per_second": 3041.89
+    },
+    {
+      "epoch": 1.5059203444564049,
+      "grad_norm": 2.5149903297424316,
+      "learning_rate": 1.4949329247118258e-05,
+      "loss": 0.0621,
+      "num_input_tokens_seen": 2609696,
+      "step": 700,
+      "train_runtime": 857.7592,
+      "train_tokens_per_second": 3042.458
+    },
+    {
+      "epoch": 1.527448869752422,
+      "grad_norm": 1.2988933324813843,
+      "learning_rate": 1.4611566914685489e-05,
+      "loss": 0.057,
+      "num_input_tokens_seen": 2646976,
+      "step": 710,
+      "train_runtime": 869.9041,
+      "train_tokens_per_second": 3042.837
+    },
+    {
+      "epoch": 1.5489773950484391,
+      "grad_norm": 1.6637475490570068,
+      "learning_rate": 1.4274001574524894e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 2684576,
+      "step": 720,
+      "train_runtime": 882.1134,
+      "train_tokens_per_second": 3043.346
+    },
+    {
+      "epoch": 1.5705059203444565,
+      "grad_norm": 2.7324392795562744,
+      "learning_rate": 1.3936804421536906e-05,
+      "loss": 0.0361,
+      "num_input_tokens_seen": 2720736,
+      "step": 730,
+      "train_runtime": 894.5163,
+      "train_tokens_per_second": 3041.572
+    },
+    {
+      "epoch": 1.5920344456404736,
+      "grad_norm": 0.791470468044281,
+      "learning_rate": 1.3600146463897327e-05,
+      "loss": 0.0503,
+      "num_input_tokens_seen": 2757728,
+      "step": 740,
+      "train_runtime": 906.5686,
+      "train_tokens_per_second": 3041.941
+    },
+    {
+      "epoch": 1.6135629709364907,
+      "grad_norm": 1.0436009168624878,
+      "learning_rate": 1.3264198436331193e-05,
+      "loss": 0.0843,
+      "num_input_tokens_seen": 2795168,
+      "step": 750,
+      "train_runtime": 918.6492,
+      "train_tokens_per_second": 3042.694
+    },
+    {
+      "epoch": 1.635091496232508,
+      "grad_norm": 1.297885775566101,
+      "learning_rate": 1.2929130713525396e-05,
+      "loss": 0.05,
+      "num_input_tokens_seen": 2832000,
+      "step": 760,
+      "train_runtime": 930.6759,
+      "train_tokens_per_second": 3042.95
+    },
+    {
+      "epoch": 1.6566200215285254,
+      "grad_norm": 1.0669901371002197,
+      "learning_rate": 1.2595113223723788e-05,
+      "loss": 0.0608,
+      "num_input_tokens_seen": 2869696,
+      "step": 770,
+      "train_runtime": 942.933,
+      "train_tokens_per_second": 3043.372
+    },
+    {
+      "epoch": 1.6781485468245425,
+      "grad_norm": 2.101264715194702,
+      "learning_rate": 1.2262315362548825e-05,
+      "loss": 0.0422,
+      "num_input_tokens_seen": 2907840,
+      "step": 780,
+      "train_runtime": 955.2475,
+      "train_tokens_per_second": 3044.07
+    },
+    {
+      "epoch": 1.6996770721205596,
+      "grad_norm": 3.501140594482422,
+      "learning_rate": 1.1930905907093196e-05,
+      "loss": 0.0558,
+      "num_input_tokens_seen": 2943232,
+      "step": 790,
+      "train_runtime": 967.0703,
+      "train_tokens_per_second": 3043.452
+    },
+    {
+      "epoch": 1.721205597416577,
+      "grad_norm": 5.079594612121582,
+      "learning_rate": 1.1601052930325236e-05,
+      "loss": 0.0509,
+      "num_input_tokens_seen": 2980608,
+      "step": 800,
+      "train_runtime": 979.2158,
+      "train_tokens_per_second": 3043.872
+    },
+    {
+      "epoch": 1.7427341227125943,
+      "grad_norm": 1.612389326095581,
+      "learning_rate": 1.127292371585141e-05,
+      "loss": 0.0674,
+      "num_input_tokens_seen": 3018080,
+      "step": 810,
+      "train_runtime": 991.3276,
+      "train_tokens_per_second": 3044.483
+    },
+    {
+      "epoch": 1.7642626480086114,
+      "grad_norm": 4.1869120597839355,
+      "learning_rate": 1.0946684673079074e-05,
+      "loss": 0.0694,
+      "num_input_tokens_seen": 3055520,
+      "step": 820,
+      "train_runtime": 1003.4787,
+      "train_tokens_per_second": 3044.928
+    },
+    {
+      "epoch": 1.7857911733046286,
+      "grad_norm": 0.6242759823799133,
+      "learning_rate": 1.0622501252822628e-05,
+      "loss": 0.0383,
+      "num_input_tokens_seen": 3092800,
+      "step": 830,
+      "train_runtime": 1015.5301,
+      "train_tokens_per_second": 3045.503
+    },
+    {
+      "epoch": 1.807319698600646,
+      "grad_norm": 1.8685392141342163,
+      "learning_rate": 1.0300537863395823e-05,
+      "loss": 0.0547,
+      "num_input_tokens_seen": 3130656,
+      "step": 840,
+      "train_runtime": 1027.7868,
+      "train_tokens_per_second": 3046.017
+    },
+    {
+      "epoch": 1.8288482238966632,
+      "grad_norm": 1.3861064910888672,
+      "learning_rate": 9.98095778723271e-06,
+      "loss": 0.0516,
+      "num_input_tokens_seen": 3167744,
+      "step": 850,
+      "train_runtime": 1039.7962,
+      "train_tokens_per_second": 3046.505
+    },
+    {
+      "epoch": 1.8503767491926801,
+      "grad_norm": 2.453305244445801,
+      "learning_rate": 9.663923098079656e-06,
+      "loss": 0.0628,
+      "num_input_tokens_seen": 3205376,
+      "step": 860,
+      "train_runtime": 1051.9233,
+      "train_tokens_per_second": 3047.157
+    },
+    {
+      "epoch": 1.8719052744886975,
+      "grad_norm": 0.682584822177887,
+      "learning_rate": 9.349594578800259e-06,
+      "loss": 0.0733,
+      "num_input_tokens_seen": 3242976,
+      "step": 870,
+      "train_runtime": 1064.1457,
+      "train_tokens_per_second": 3047.492
+    },
+    {
+      "epoch": 1.8934337997847148,
+      "grad_norm": 1.5370641946792603,
+      "learning_rate": 9.038131639834977e-06,
+      "loss": 0.0727,
+      "num_input_tokens_seen": 3280096,
+      "step": 880,
+      "train_runtime": 1076.18,
+      "train_tokens_per_second": 3047.906
+    },
+    {
+      "epoch": 1.914962325080732,
+      "grad_norm": 0.9756731390953064,
+      "learning_rate": 8.7296922383568e-06,
+      "loss": 0.058,
+      "num_input_tokens_seen": 3316960,
+      "step": 890,
+      "train_runtime": 1088.1737,
+      "train_tokens_per_second": 3048.19
+    },
+    {
+      "epoch": 1.936490850376749,
+      "grad_norm": 1.8065251111984253,
+      "learning_rate": 8.424432798163838e-06,
+      "loss": 0.0588,
+      "num_input_tokens_seen": 3353600,
+      "step": 900,
+      "train_runtime": 1100.1193,
+      "train_tokens_per_second": 3048.397
+    },
+    {
+      "epoch": 1.9580193756727664,
+      "grad_norm": 1.448608636856079,
+      "learning_rate": 8.122508130349711e-06,
+      "loss": 0.0375,
+      "num_input_tokens_seen": 3390592,
+      "step": 910,
+      "train_runtime": 1112.0701,
+      "train_tokens_per_second": 3048.901
+    },
+    {
+      "epoch": 1.9795479009687837,
+      "grad_norm": 1.7946593761444092,
+      "learning_rate": 7.82407135479164e-06,
+      "loss": 0.0644,
+      "num_input_tokens_seen": 3427840,
+      "step": 920,
+      "train_runtime": 1124.2394,
+      "train_tokens_per_second": 3049.03
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08534291386604309,
+      "learning_rate": 7.529273822496379e-06,
+      "loss": 0.0347,
+      "num_input_tokens_seen": 3464032,
+      "step": 930,
+      "train_runtime": 1135.8636,
+      "train_tokens_per_second": 3049.69
+    },
+    {
+      "epoch": 2.0215285252960173,
+      "grad_norm": 3.5694122314453125,
+      "learning_rate": 7.238265038843143e-06,
+      "loss": 0.0491,
+      "num_input_tokens_seen": 3501056,
+      "step": 940,
+      "train_runtime": 1147.9804,
+      "train_tokens_per_second": 3049.752
+    },
+    {
+      "epoch": 2.0430570505920342,
+      "grad_norm": 1.0579440593719482,
+      "learning_rate": 6.951192587762555e-06,
+      "loss": 0.0263,
+      "num_input_tokens_seen": 3538464,
+      "step": 950,
+      "train_runtime": 1160.1923,
+      "train_tokens_per_second": 3049.894
+    },
+    {
+      "epoch": 2.0645855758880516,
+      "grad_norm": 1.1105297803878784,
+      "learning_rate": 6.668202056890052e-06,
+      "loss": 0.029,
+      "num_input_tokens_seen": 3575552,
+      "step": 960,
+      "train_runtime": 1172.3064,
+      "train_tokens_per_second": 3050.015
+    },
+    {
+      "epoch": 2.086114101184069,
+      "grad_norm": 2.4604287147521973,
+      "learning_rate": 6.389436963731707e-06,
+      "loss": 0.0338,
+      "num_input_tokens_seen": 3612736,
+      "step": 970,
+      "train_runtime": 1184.4757,
+      "train_tokens_per_second": 3050.072
+    },
+    {
+      "epoch": 2.1076426264800863,
+      "grad_norm": 2.9901983737945557,
+      "learning_rate": 6.115038682879904e-06,
+      "loss": 0.0427,
+      "num_input_tokens_seen": 3650464,
+      "step": 980,
+      "train_runtime": 1196.7382,
+      "train_tokens_per_second": 3050.345
+    },
+    {
+      "epoch": 2.129171151776103,
+      "grad_norm": 1.1089911460876465,
+      "learning_rate": 5.845146374315781e-06,
+      "loss": 0.0323,
+      "num_input_tokens_seen": 3687392,
+      "step": 990,
+      "train_runtime": 1208.888,
+      "train_tokens_per_second": 3050.235
+    },
+    {
+      "epoch": 2.1506996770721205,
+      "grad_norm": 4.3608317375183105,
+      "learning_rate": 5.579896912834807e-06,
+      "loss": 0.0504,
+      "num_input_tokens_seen": 3725152,
+      "step": 1000,
+      "train_runtime": 1221.1038,
+      "train_tokens_per_second": 3050.643
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1395,
+  "num_input_tokens_seen": 3725152,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5946878055533773e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d57507a8b818954dbee794b7baa7eafcc21a01dd96c8730e5288c6cdd6c47d3
+size 6161

checkpoint-1395/README.md

@@ -0,0 +1,202 @@
+---
+base_model: unsloth/mistral-7b-instruct-v0.3-bnb-4bit
+library_name: peft
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.15.2

checkpoint-1395/adapter_config.json

@@ -0,0 +1,43 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": {
+    "base_model_class": "MistralForCausalLM",
+    "parent_library": "transformers.models.mistral.modeling_mistral",
+    "unsloth_fixed": true
+  },
+  "base_model_name_or_path": "unsloth/mistral-7b-instruct-v0.3-bnb-4bit",
+  "bias": "none",
+  "corda_config": null,
+  "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": 16,
+  "lora_bias": false,
+  "lora_dropout": 0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "v_proj",
+    "q_proj",
+    "k_proj",
+    "o_proj",
+    "up_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

checkpoint-1395/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c054d913deee23f95656c107fa6e9ca3fc1b6f0faac432041de9389df93e2bd8
+size 83945296

checkpoint-1395/chat_template.jinja

@@ -0,0 +1,9 @@
+{% set system_message = 'Below is an instruction that describes a task. Write a response that appropriately completes the request.
+
+' %}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in loop_messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '### Instruction:
+' + content + '
+
+### Response:
+' }}{% elif message['role'] == 'assistant' %}{{ content + '</s>' + '
+
+' }}{% endif %}{% endfor %}

checkpoint-1395/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5b4feece21e811365b734e720cf52c94951145670d7ac34e985796a1198a6c9
+size 168149539

checkpoint-1395/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2534e434cd5abbb8f7668d3eab0549db0ef95d6a797a3efa86b712e8e32266a7
+size 14645

checkpoint-1395/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04444d631cd637645204283b0be3b130f0d32a6e952a616746c838579a7e024c
+size 1465

checkpoint-1395/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-1395/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37f00374dea48658ee8f5d0f21895b9bc55cb0103939607c8185bfd1c6ca1f89
+size 587404

checkpoint-1395/trainer_state.json

@@ -0,0 +1,1424 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1395,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.021528525296017224,
+      "grad_norm": 3.766150712966919,
+      "learning_rate": 2.9996919063517692e-05,
+      "loss": 0.6799,
+      "num_input_tokens_seen": 37152,
+      "step": 10,
+      "train_runtime": 20.683,
+      "train_tokens_per_second": 1796.26
+    },
+    {
+      "epoch": 0.04305705059203445,
+      "grad_norm": 3.9944393634796143,
+      "learning_rate": 2.9986270537572827e-05,
+      "loss": 0.1361,
+      "num_input_tokens_seen": 73856,
+      "step": 20,
+      "train_runtime": 32.7469,
+      "train_tokens_per_second": 2255.36
+    },
+    {
+      "epoch": 0.06458557588805167,
+      "grad_norm": 10.329977989196777,
+      "learning_rate": 2.9968021784983436e-05,
+      "loss": 0.1558,
+      "num_input_tokens_seen": 111232,
+      "step": 30,
+      "train_runtime": 45.4717,
+      "train_tokens_per_second": 2446.179
+    },
+    {
+      "epoch": 0.0861141011840689,
+      "grad_norm": 4.059694766998291,
+      "learning_rate": 2.994218206053077e-05,
+      "loss": 0.1344,
+      "num_input_tokens_seen": 148992,
+      "step": 40,
+      "train_runtime": 57.6575,
+      "train_tokens_per_second": 2584.086
+    },
+    {
+      "epoch": 0.10764262648008611,
+      "grad_norm": 2.61093807220459,
+      "learning_rate": 2.9908764468726846e-05,
+      "loss": 0.1262,
+      "num_input_tokens_seen": 186176,
+      "step": 50,
+      "train_runtime": 69.6848,
+      "train_tokens_per_second": 2671.688
+    },
+    {
+      "epoch": 0.12917115177610333,
+      "grad_norm": 4.661616325378418,
+      "learning_rate": 2.986778595716853e-05,
+      "loss": 0.1182,
+      "num_input_tokens_seen": 223840,
+      "step": 60,
+      "train_runtime": 81.8303,
+      "train_tokens_per_second": 2735.416
+    },
+    {
+      "epoch": 0.15069967707212056,
+      "grad_norm": 2.2738418579101562,
+      "learning_rate": 2.981926730794264e-05,
+      "loss": 0.1333,
+      "num_input_tokens_seen": 260960,
+      "step": 70,
+      "train_runtime": 93.9745,
+      "train_tokens_per_second": 2776.923
+    },
+    {
+      "epoch": 0.1722282023681378,
+      "grad_norm": 2.6716361045837402,
+      "learning_rate": 2.9763233127086394e-05,
+      "loss": 0.0925,
+      "num_input_tokens_seen": 298048,
+      "step": 80,
+      "train_runtime": 106.1694,
+      "train_tokens_per_second": 2807.287
+    },
+    {
+      "epoch": 0.193756727664155,
+      "grad_norm": 2.849599599838257,
+      "learning_rate": 2.9699711832108555e-05,
+      "loss": 0.0886,
+      "num_input_tokens_seen": 334496,
+      "step": 90,
+      "train_runtime": 118.0166,
+      "train_tokens_per_second": 2834.313
+    },
+    {
+      "epoch": 0.21528525296017223,
+      "grad_norm": 1.1690081357955933,
+      "learning_rate": 2.9628735637577612e-05,
+      "loss": 0.1127,
+      "num_input_tokens_seen": 371712,
+      "step": 100,
+      "train_runtime": 130.0666,
+      "train_tokens_per_second": 2857.86
+    },
+    {
+      "epoch": 0.23681377825618946,
+      "grad_norm": 1.8079500198364258,
+      "learning_rate": 2.955034053878428e-05,
+      "loss": 0.1018,
+      "num_input_tokens_seen": 408512,
+      "step": 110,
+      "train_runtime": 142.0073,
+      "train_tokens_per_second": 2876.697
+    },
+    {
+      "epoch": 0.25834230355220666,
+      "grad_norm": 4.173007965087891,
+      "learning_rate": 2.9464566293486653e-05,
+      "loss": 0.1001,
+      "num_input_tokens_seen": 446048,
+      "step": 120,
+      "train_runtime": 154.1596,
+      "train_tokens_per_second": 2893.418
+    },
+    {
+      "epoch": 0.2798708288482239,
+      "grad_norm": 3.0781972408294678,
+      "learning_rate": 2.9371456401747193e-05,
+      "loss": 0.0962,
+      "num_input_tokens_seen": 483616,
+      "step": 130,
+      "train_runtime": 166.2926,
+      "train_tokens_per_second": 2908.222
+    },
+    {
+      "epoch": 0.3013993541442411,
+      "grad_norm": 3.6592869758605957,
+      "learning_rate": 2.9271058083871864e-05,
+      "loss": 0.1076,
+      "num_input_tokens_seen": 520768,
+      "step": 140,
+      "train_runtime": 178.3589,
+      "train_tokens_per_second": 2919.776
+    },
+    {
+      "epoch": 0.32292787944025836,
+      "grad_norm": 1.6870366334915161,
+      "learning_rate": 2.916342225646254e-05,
+      "loss": 0.0886,
+      "num_input_tokens_seen": 558240,
+      "step": 150,
+      "train_runtime": 190.6338,
+      "train_tokens_per_second": 2928.338
+    },
+    {
+      "epoch": 0.3444564047362756,
+      "grad_norm": 3.0077579021453857,
+      "learning_rate": 2.904860350659487e-05,
+      "loss": 0.0861,
+      "num_input_tokens_seen": 595648,
+      "step": 160,
+      "train_runtime": 202.803,
+      "train_tokens_per_second": 2937.076
+    },
+    {
+      "epoch": 0.36598493003229277,
+      "grad_norm": 2.7459771633148193,
+      "learning_rate": 2.8926660064134637e-05,
+      "loss": 0.1158,
+      "num_input_tokens_seen": 632352,
+      "step": 170,
+      "train_runtime": 214.8858,
+      "train_tokens_per_second": 2942.735
+    },
+    {
+      "epoch": 0.38751345532831,
+      "grad_norm": 1.1326333284378052,
+      "learning_rate": 2.8797653772206726e-05,
+      "loss": 0.0931,
+      "num_input_tokens_seen": 669440,
+      "step": 180,
+      "train_runtime": 226.9184,
+      "train_tokens_per_second": 2950.136
+    },
+    {
+      "epoch": 0.40904198062432723,
+      "grad_norm": 2.5396974086761475,
+      "learning_rate": 2.8661650055831645e-05,
+      "loss": 0.0964,
+      "num_input_tokens_seen": 706784,
+      "step": 190,
+      "train_runtime": 239.0179,
+      "train_tokens_per_second": 2957.034
+    },
+    {
+      "epoch": 0.43057050592034446,
+      "grad_norm": 2.9514100551605225,
+      "learning_rate": 2.8518717888745486e-05,
+      "loss": 0.0708,
+      "num_input_tokens_seen": 744480,
+      "step": 200,
+      "train_runtime": 251.1851,
+      "train_tokens_per_second": 2963.87
+    },
+    {
+      "epoch": 0.4520990312163617,
+      "grad_norm": 2.363633871078491,
+      "learning_rate": 2.836892975842015e-05,
+      "loss": 0.0954,
+      "num_input_tokens_seen": 781632,
+      "step": 210,
+      "train_runtime": 263.3225,
+      "train_tokens_per_second": 2968.345
+    },
+    {
+      "epoch": 0.4736275565123789,
+      "grad_norm": 2.0013911724090576,
+      "learning_rate": 2.8212361629301655e-05,
+      "loss": 0.0897,
+      "num_input_tokens_seen": 819264,
+      "step": 220,
+      "train_runtime": 275.4833,
+      "train_tokens_per_second": 2973.916
+    },
+    {
+      "epoch": 0.4951560818083961,
+      "grad_norm": 2.076155185699463,
+      "learning_rate": 2.8049092904285033e-05,
+      "loss": 0.0783,
+      "num_input_tokens_seen": 855936,
+      "step": 230,
+      "train_runtime": 287.5153,
+      "train_tokens_per_second": 2977.01
+    },
+    {
+      "epoch": 0.5166846071044133,
+      "grad_norm": 3.730830192565918,
+      "learning_rate": 2.7879206384445526e-05,
+      "loss": 0.092,
+      "num_input_tokens_seen": 893536,
+      "step": 240,
+      "train_runtime": 299.6624,
+      "train_tokens_per_second": 2981.809
+    },
+    {
+      "epoch": 0.5382131324004306,
+      "grad_norm": 1.213758111000061,
+      "learning_rate": 2.7702788227046286e-05,
+      "loss": 0.1014,
+      "num_input_tokens_seen": 930272,
+      "step": 250,
+      "train_runtime": 311.6586,
+      "train_tokens_per_second": 2984.907
+    },
+    {
+      "epoch": 0.5597416576964478,
+      "grad_norm": 1.2503005266189575,
+      "learning_rate": 2.751992790184413e-05,
+      "loss": 0.1032,
+      "num_input_tokens_seen": 968768,
+      "step": 260,
+      "train_runtime": 324.0216,
+      "train_tokens_per_second": 2989.825
+    },
+    {
+      "epoch": 0.581270182992465,
+      "grad_norm": 2.627034902572632,
+      "learning_rate": 2.733071814571527e-05,
+      "loss": 0.1016,
+      "num_input_tokens_seen": 1005888,
+      "step": 270,
+      "train_runtime": 336.1097,
+      "train_tokens_per_second": 2992.737
+    },
+    {
+      "epoch": 0.6027987082884823,
+      "grad_norm": 1.8291891813278198,
+      "learning_rate": 2.7135254915624213e-05,
+      "loss": 0.0843,
+      "num_input_tokens_seen": 1043744,
+      "step": 280,
+      "train_runtime": 348.2992,
+      "train_tokens_per_second": 2996.688
+    },
+    {
+      "epoch": 0.6243272335844995,
+      "grad_norm": 2.0011653900146484,
+      "learning_rate": 2.693363733995954e-05,
+      "loss": 0.1012,
+      "num_input_tokens_seen": 1081920,
+      "step": 290,
+      "train_runtime": 360.6262,
+      "train_tokens_per_second": 3000.115
+    },
+    {
+      "epoch": 0.6458557588805167,
+      "grad_norm": 1.9434913396835327,
+      "learning_rate": 2.672596766826138e-05,
+      "loss": 0.0835,
+      "num_input_tokens_seen": 1118784,
+      "step": 300,
+      "train_runtime": 372.7516,
+      "train_tokens_per_second": 3001.419
+    },
+    {
+      "epoch": 0.667384284176534,
+      "grad_norm": 1.8366467952728271,
+      "learning_rate": 2.6512351219365927e-05,
+      "loss": 0.0884,
+      "num_input_tokens_seen": 1155392,
+      "step": 310,
+      "train_runtime": 384.8491,
+      "train_tokens_per_second": 3002.195
+    },
+    {
+      "epoch": 0.6889128094725512,
+      "grad_norm": 1.1592612266540527,
+      "learning_rate": 2.6292896327993467e-05,
+      "loss": 0.0907,
+      "num_input_tokens_seen": 1192768,
+      "step": 320,
+      "train_runtime": 397.0198,
+      "train_tokens_per_second": 3004.303
+    },
+    {
+      "epoch": 0.7104413347685683,
+      "grad_norm": 1.4585574865341187,
+      "learning_rate": 2.606771428980681e-05,
+      "loss": 0.0786,
+      "num_input_tokens_seen": 1230464,
+      "step": 330,
+      "train_runtime": 409.2176,
+      "train_tokens_per_second": 3006.87
+    },
+    {
+      "epoch": 0.7319698600645855,
+      "grad_norm": 1.2248650789260864,
+      "learning_rate": 2.5836919304968184e-05,
+      "loss": 0.0863,
+      "num_input_tokens_seen": 1267520,
+      "step": 340,
+      "train_runtime": 421.3156,
+      "train_tokens_per_second": 3008.481
+    },
+    {
+      "epoch": 0.7534983853606028,
+      "grad_norm": 2.204815149307251,
+      "learning_rate": 2.5600628420223057e-05,
+      "loss": 0.0976,
+      "num_input_tokens_seen": 1306592,
+      "step": 350,
+      "train_runtime": 433.7939,
+      "train_tokens_per_second": 3012.011
+    },
+    {
+      "epoch": 0.77502691065662,
+      "grad_norm": 1.4170422554016113,
+      "learning_rate": 2.535896146954037e-05,
+      "loss": 0.0914,
+      "num_input_tokens_seen": 1343840,
+      "step": 360,
+      "train_runtime": 445.8983,
+      "train_tokens_per_second": 3013.781
+    },
+    {
+      "epoch": 0.7965554359526372,
+      "grad_norm": 2.460742712020874,
+      "learning_rate": 2.5112041013339212e-05,
+      "loss": 0.0997,
+      "num_input_tokens_seen": 1382816,
+      "step": 370,
+      "train_runtime": 458.3847,
+      "train_tokens_per_second": 3016.715
+    },
+    {
+      "epoch": 0.8180839612486545,
+      "grad_norm": 0.7948861718177795,
+      "learning_rate": 2.48599922763328e-05,
+      "loss": 0.0689,
+      "num_input_tokens_seen": 1420352,
+      "step": 380,
+      "train_runtime": 470.464,
+      "train_tokens_per_second": 3019.045
+    },
+    {
+      "epoch": 0.8396124865446717,
+      "grad_norm": 0.855792760848999,
+      "learning_rate": 2.4602943084021286e-05,
+      "loss": 0.1062,
+      "num_input_tokens_seen": 1457280,
+      "step": 390,
+      "train_runtime": 482.5408,
+      "train_tokens_per_second": 3020.014
+    },
+    {
+      "epoch": 0.8611410118406889,
+      "grad_norm": 1.3703863620758057,
+      "learning_rate": 2.434102379786553e-05,
+      "loss": 0.0579,
+      "num_input_tokens_seen": 1493984,
+      "step": 400,
+      "train_runtime": 494.602,
+      "train_tokens_per_second": 3020.578
+    },
+    {
+      "epoch": 0.8826695371367062,
+      "grad_norm": 1.5039470195770264,
+      "learning_rate": 2.4074367249174864e-05,
+      "loss": 0.074,
+      "num_input_tokens_seen": 1531392,
+      "step": 410,
+      "train_runtime": 506.7325,
+      "train_tokens_per_second": 3022.091
+    },
+    {
+      "epoch": 0.9041980624327234,
+      "grad_norm": 2.4852614402770996,
+      "learning_rate": 2.380310867174219e-05,
+      "loss": 0.0774,
+      "num_input_tokens_seen": 1568992,
+      "step": 420,
+      "train_runtime": 518.936,
+      "train_tokens_per_second": 3023.479
+    },
+    {
+      "epoch": 0.9257265877287406,
+      "grad_norm": 0.6510267853736877,
+      "learning_rate": 2.3527385633260724e-05,
+      "loss": 0.0771,
+      "num_input_tokens_seen": 1606144,
+      "step": 430,
+      "train_runtime": 530.9278,
+      "train_tokens_per_second": 3025.165
+    },
+    {
+      "epoch": 0.9472551130247578,
+      "grad_norm": 1.678964376449585,
+      "learning_rate": 2.3247337965557097e-05,
+      "loss": 0.075,
+      "num_input_tokens_seen": 1641696,
+      "step": 440,
+      "train_runtime": 542.7133,
+      "train_tokens_per_second": 3024.979
+    },
+    {
+      "epoch": 0.9687836383207751,
+      "grad_norm": 0.5749807953834534,
+      "learning_rate": 2.2963107693676192e-05,
+      "loss": 0.0785,
+      "num_input_tokens_seen": 1678816,
+      "step": 450,
+      "train_runtime": 554.8525,
+      "train_tokens_per_second": 3025.698
+    },
+    {
+      "epoch": 0.9903121636167922,
+      "grad_norm": 1.0846511125564575,
+      "learning_rate": 2.2674838963853738e-05,
+      "loss": 0.074,
+      "num_input_tokens_seen": 1715808,
+      "step": 460,
+      "train_runtime": 566.8619,
+      "train_tokens_per_second": 3026.854
+    },
+    {
+      "epoch": 1.0107642626480087,
+      "grad_norm": 4.8021769523620605,
+      "learning_rate": 2.2382677970413097e-05,
+      "loss": 0.0653,
+      "num_input_tokens_seen": 1751552,
+      "step": 470,
+      "train_runtime": 578.3306,
+      "train_tokens_per_second": 3028.634
+    },
+    {
+      "epoch": 1.0322927879440258,
+      "grad_norm": 1.0518341064453125,
+      "learning_rate": 2.2086772881623455e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 1788544,
+      "step": 480,
+      "train_runtime": 590.3999,
+      "train_tokens_per_second": 3029.377
+    },
+    {
+      "epoch": 1.0538213132400431,
+      "grad_norm": 2.102010488510132,
+      "learning_rate": 2.1787273764556827e-05,
+      "loss": 0.0472,
+      "num_input_tokens_seen": 1826496,
+      "step": 490,
+      "train_runtime": 602.6517,
+      "train_tokens_per_second": 3030.765
+    },
+    {
+      "epoch": 1.0753498385360603,
+      "grad_norm": 0.5412517786026001,
+      "learning_rate": 2.1484332508982167e-05,
+      "loss": 0.0534,
+      "num_input_tokens_seen": 1864544,
+      "step": 500,
+      "train_runtime": 614.9379,
+      "train_tokens_per_second": 3032.085
+    },
+    {
+      "epoch": 1.0968783638320776,
+      "grad_norm": 0.6883963942527771,
+      "learning_rate": 2.11781027503351e-05,
+      "loss": 0.0751,
+      "num_input_tokens_seen": 1901856,
+      "step": 510,
+      "train_runtime": 627.0846,
+      "train_tokens_per_second": 3032.854
+    },
+    {
+      "epoch": 1.1184068891280947,
+      "grad_norm": 2.446444034576416,
+      "learning_rate": 2.0868739791802312e-05,
+      "loss": 0.0747,
+      "num_input_tokens_seen": 1939264,
+      "step": 520,
+      "train_runtime": 639.2551,
+      "train_tokens_per_second": 3033.631
+    },
+    {
+      "epoch": 1.1399354144241118,
+      "grad_norm": 0.9430311322212219,
+      "learning_rate": 2.0556400525560168e-05,
+      "loss": 0.0492,
+      "num_input_tokens_seen": 1975872,
+      "step": 530,
+      "train_runtime": 651.2018,
+      "train_tokens_per_second": 3034.193
+    },
+    {
+      "epoch": 1.1614639397201292,
+      "grad_norm": 2.4045143127441406,
+      "learning_rate": 2.0241243353207484e-05,
+      "loss": 0.0651,
+      "num_input_tokens_seen": 2012928,
+      "step": 540,
+      "train_runtime": 663.3226,
+      "train_tokens_per_second": 3034.614
+    },
+    {
+      "epoch": 1.1829924650161463,
+      "grad_norm": 1.0207197666168213,
+      "learning_rate": 1.9923428105432796e-05,
+      "loss": 0.0565,
+      "num_input_tokens_seen": 2051168,
+      "step": 550,
+      "train_runtime": 675.6641,
+      "train_tokens_per_second": 3035.78
+    },
+    {
+      "epoch": 1.2045209903121636,
+      "grad_norm": 7.416859149932861,
+      "learning_rate": 1.9603115960956852e-05,
+      "loss": 0.0696,
+      "num_input_tokens_seen": 2088576,
+      "step": 560,
+      "train_runtime": 687.7762,
+      "train_tokens_per_second": 3036.709
+    },
+    {
+      "epoch": 1.2260495156081808,
+      "grad_norm": 1.7133443355560303,
+      "learning_rate": 1.9280469364791505e-05,
+      "loss": 0.0553,
+      "num_input_tokens_seen": 2125216,
+      "step": 570,
+      "train_runtime": 699.7307,
+      "train_tokens_per_second": 3037.191
+    },
+    {
+      "epoch": 1.247578040904198,
+      "grad_norm": 2.1229183673858643,
+      "learning_rate": 1.8955651945856354e-05,
+      "loss": 0.0666,
+      "num_input_tokens_seen": 2162816,
+      "step": 580,
+      "train_runtime": 711.9331,
+      "train_tokens_per_second": 3037.948
+    },
+    {
+      "epoch": 1.2691065662002152,
+      "grad_norm": 0.5658122897148132,
+      "learning_rate": 1.8628828433995013e-05,
+      "loss": 0.0449,
+      "num_input_tokens_seen": 2199712,
+      "step": 590,
+      "train_runtime": 724.0149,
+      "train_tokens_per_second": 3038.214
+    },
+    {
+      "epoch": 1.2906350914962326,
+      "grad_norm": 2.8044610023498535,
+      "learning_rate": 1.8300164576433035e-05,
+      "loss": 0.0522,
+      "num_input_tokens_seen": 2236416,
+      "step": 600,
+      "train_runtime": 736.0531,
+      "train_tokens_per_second": 3038.389
+    },
+    {
+      "epoch": 1.3121636167922497,
+      "grad_norm": 1.5163172483444214,
+      "learning_rate": 1.7969827053719862e-05,
+      "loss": 0.0638,
+      "num_input_tokens_seen": 2274272,
+      "step": 610,
+      "train_runtime": 748.3131,
+      "train_tokens_per_second": 3039.198
+    },
+    {
+      "epoch": 1.333692142088267,
+      "grad_norm": 2.5300729274749756,
+      "learning_rate": 1.7637983395197465e-05,
+      "loss": 0.0683,
+      "num_input_tokens_seen": 2310688,
+      "step": 620,
+      "train_runtime": 760.313,
+      "train_tokens_per_second": 3039.127
+    },
+    {
+      "epoch": 1.3552206673842842,
+      "grad_norm": 2.793215036392212,
+      "learning_rate": 1.7304801894038522e-05,
+      "loss": 0.0678,
+      "num_input_tokens_seen": 2348608,
+      "step": 630,
+      "train_runtime": 772.6051,
+      "train_tokens_per_second": 3039.856
+    },
+    {
+      "epoch": 1.3767491926803013,
+      "grad_norm": 1.3523025512695312,
+      "learning_rate": 1.6970451521897198e-05,
+      "loss": 0.0502,
+      "num_input_tokens_seen": 2386144,
+      "step": 640,
+      "train_runtime": 784.8469,
+      "train_tokens_per_second": 3040.267
+    },
+    {
+      "epoch": 1.3982777179763186,
+      "grad_norm": 2.378476858139038,
+      "learning_rate": 1.663510184321587e-05,
+      "loss": 0.0728,
+      "num_input_tokens_seen": 2422912,
+      "step": 650,
+      "train_runtime": 796.9416,
+      "train_tokens_per_second": 3040.263
+    },
+    {
+      "epoch": 1.419806243272336,
+      "grad_norm": 1.0586169958114624,
+      "learning_rate": 1.6298922929231194e-05,
+      "loss": 0.059,
+      "num_input_tokens_seen": 2459744,
+      "step": 660,
+      "train_runtime": 809.0893,
+      "train_tokens_per_second": 3040.139
+    },
+    {
+      "epoch": 1.441334768568353,
+      "grad_norm": 1.8658058643341064,
+      "learning_rate": 1.5962085271723158e-05,
+      "loss": 0.0514,
+      "num_input_tokens_seen": 2497184,
+      "step": 670,
+      "train_runtime": 821.239,
+      "train_tokens_per_second": 3040.752
+    },
+    {
+      "epoch": 1.4628632938643702,
+      "grad_norm": 2.6104471683502197,
+      "learning_rate": 1.5624759696550874e-05,
+      "loss": 0.0689,
+      "num_input_tokens_seen": 2535456,
+      "step": 680,
+      "train_runtime": 833.6348,
+      "train_tokens_per_second": 3041.447
+    },
+    {
+      "epoch": 1.4843918191603875,
+      "grad_norm": 1.810530662536621,
+      "learning_rate": 1.528711727701893e-05,
+      "loss": 0.0338,
+      "num_input_tokens_seen": 2572704,
+      "step": 690,
+      "train_runtime": 845.7584,
+      "train_tokens_per_second": 3041.89
+    },
+    {
+      "epoch": 1.5059203444564049,
+      "grad_norm": 2.5149903297424316,
+      "learning_rate": 1.4949329247118258e-05,
+      "loss": 0.0621,
+      "num_input_tokens_seen": 2609696,
+      "step": 700,
+      "train_runtime": 857.7592,
+      "train_tokens_per_second": 3042.458
+    },
+    {
+      "epoch": 1.527448869752422,
+      "grad_norm": 1.2988933324813843,
+      "learning_rate": 1.4611566914685489e-05,
+      "loss": 0.057,
+      "num_input_tokens_seen": 2646976,
+      "step": 710,
+      "train_runtime": 869.9041,
+      "train_tokens_per_second": 3042.837
+    },
+    {
+      "epoch": 1.5489773950484391,
+      "grad_norm": 1.6637475490570068,
+      "learning_rate": 1.4274001574524894e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 2684576,
+      "step": 720,
+      "train_runtime": 882.1134,
+      "train_tokens_per_second": 3043.346
+    },
+    {
+      "epoch": 1.5705059203444565,
+      "grad_norm": 2.7324392795562744,
+      "learning_rate": 1.3936804421536906e-05,
+      "loss": 0.0361,
+      "num_input_tokens_seen": 2720736,
+      "step": 730,
+      "train_runtime": 894.5163,
+      "train_tokens_per_second": 3041.572
+    },
+    {
+      "epoch": 1.5920344456404736,
+      "grad_norm": 0.791470468044281,
+      "learning_rate": 1.3600146463897327e-05,
+      "loss": 0.0503,
+      "num_input_tokens_seen": 2757728,
+      "step": 740,
+      "train_runtime": 906.5686,
+      "train_tokens_per_second": 3041.941
+    },
+    {
+      "epoch": 1.6135629709364907,
+      "grad_norm": 1.0436009168624878,
+      "learning_rate": 1.3264198436331193e-05,
+      "loss": 0.0843,
+      "num_input_tokens_seen": 2795168,
+      "step": 750,
+      "train_runtime": 918.6492,
+      "train_tokens_per_second": 3042.694
+    },
+    {
+      "epoch": 1.635091496232508,
+      "grad_norm": 1.297885775566101,
+      "learning_rate": 1.2929130713525396e-05,
+      "loss": 0.05,
+      "num_input_tokens_seen": 2832000,
+      "step": 760,
+      "train_runtime": 930.6759,
+      "train_tokens_per_second": 3042.95
+    },
+    {
+      "epoch": 1.6566200215285254,
+      "grad_norm": 1.0669901371002197,
+      "learning_rate": 1.2595113223723788e-05,
+      "loss": 0.0608,
+      "num_input_tokens_seen": 2869696,
+      "step": 770,
+      "train_runtime": 942.933,
+      "train_tokens_per_second": 3043.372
+    },
+    {
+      "epoch": 1.6781485468245425,
+      "grad_norm": 2.101264715194702,
+      "learning_rate": 1.2262315362548825e-05,
+      "loss": 0.0422,
+      "num_input_tokens_seen": 2907840,
+      "step": 780,
+      "train_runtime": 955.2475,
+      "train_tokens_per_second": 3044.07
+    },
+    {
+      "epoch": 1.6996770721205596,
+      "grad_norm": 3.501140594482422,
+      "learning_rate": 1.1930905907093196e-05,
+      "loss": 0.0558,
+      "num_input_tokens_seen": 2943232,
+      "step": 790,
+      "train_runtime": 967.0703,
+      "train_tokens_per_second": 3043.452
+    },
+    {
+      "epoch": 1.721205597416577,
+      "grad_norm": 5.079594612121582,
+      "learning_rate": 1.1601052930325236e-05,
+      "loss": 0.0509,
+      "num_input_tokens_seen": 2980608,
+      "step": 800,
+      "train_runtime": 979.2158,
+      "train_tokens_per_second": 3043.872
+    },
+    {
+      "epoch": 1.7427341227125943,
+      "grad_norm": 1.612389326095581,
+      "learning_rate": 1.127292371585141e-05,
+      "loss": 0.0674,
+      "num_input_tokens_seen": 3018080,
+      "step": 810,
+      "train_runtime": 991.3276,
+      "train_tokens_per_second": 3044.483
+    },
+    {
+      "epoch": 1.7642626480086114,
+      "grad_norm": 4.1869120597839355,
+      "learning_rate": 1.0946684673079074e-05,
+      "loss": 0.0694,
+      "num_input_tokens_seen": 3055520,
+      "step": 820,
+      "train_runtime": 1003.4787,
+      "train_tokens_per_second": 3044.928
+    },
+    {
+      "epoch": 1.7857911733046286,
+      "grad_norm": 0.6242759823799133,
+      "learning_rate": 1.0622501252822628e-05,
+      "loss": 0.0383,
+      "num_input_tokens_seen": 3092800,
+      "step": 830,
+      "train_runtime": 1015.5301,
+      "train_tokens_per_second": 3045.503
+    },
+    {
+      "epoch": 1.807319698600646,
+      "grad_norm": 1.8685392141342163,
+      "learning_rate": 1.0300537863395823e-05,
+      "loss": 0.0547,
+      "num_input_tokens_seen": 3130656,
+      "step": 840,
+      "train_runtime": 1027.7868,
+      "train_tokens_per_second": 3046.017
+    },
+    {
+      "epoch": 1.8288482238966632,
+      "grad_norm": 1.3861064910888672,
+      "learning_rate": 9.98095778723271e-06,
+      "loss": 0.0516,
+      "num_input_tokens_seen": 3167744,
+      "step": 850,
+      "train_runtime": 1039.7962,
+      "train_tokens_per_second": 3046.505
+    },
+    {
+      "epoch": 1.8503767491926801,
+      "grad_norm": 2.453305244445801,
+      "learning_rate": 9.663923098079656e-06,
+      "loss": 0.0628,
+      "num_input_tokens_seen": 3205376,
+      "step": 860,
+      "train_runtime": 1051.9233,
+      "train_tokens_per_second": 3047.157
+    },
+    {
+      "epoch": 1.8719052744886975,
+      "grad_norm": 0.682584822177887,
+      "learning_rate": 9.349594578800259e-06,
+      "loss": 0.0733,
+      "num_input_tokens_seen": 3242976,
+      "step": 870,
+      "train_runtime": 1064.1457,
+      "train_tokens_per_second": 3047.492
+    },
+    {
+      "epoch": 1.8934337997847148,
+      "grad_norm": 1.5370641946792603,
+      "learning_rate": 9.038131639834977e-06,
+      "loss": 0.0727,
+      "num_input_tokens_seen": 3280096,
+      "step": 880,
+      "train_runtime": 1076.18,
+      "train_tokens_per_second": 3047.906
+    },
+    {
+      "epoch": 1.914962325080732,
+      "grad_norm": 0.9756731390953064,
+      "learning_rate": 8.7296922383568e-06,
+      "loss": 0.058,
+      "num_input_tokens_seen": 3316960,
+      "step": 890,
+      "train_runtime": 1088.1737,
+      "train_tokens_per_second": 3048.19
+    },
+    {
+      "epoch": 1.936490850376749,
+      "grad_norm": 1.8065251111984253,
+      "learning_rate": 8.424432798163838e-06,
+      "loss": 0.0588,
+      "num_input_tokens_seen": 3353600,
+      "step": 900,
+      "train_runtime": 1100.1193,
+      "train_tokens_per_second": 3048.397
+    },
+    {
+      "epoch": 1.9580193756727664,
+      "grad_norm": 1.448608636856079,
+      "learning_rate": 8.122508130349711e-06,
+      "loss": 0.0375,
+      "num_input_tokens_seen": 3390592,
+      "step": 910,
+      "train_runtime": 1112.0701,
+      "train_tokens_per_second": 3048.901
+    },
+    {
+      "epoch": 1.9795479009687837,
+      "grad_norm": 1.7946593761444092,
+      "learning_rate": 7.82407135479164e-06,
+      "loss": 0.0644,
+      "num_input_tokens_seen": 3427840,
+      "step": 920,
+      "train_runtime": 1124.2394,
+      "train_tokens_per_second": 3049.03
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08534291386604309,
+      "learning_rate": 7.529273822496379e-06,
+      "loss": 0.0347,
+      "num_input_tokens_seen": 3464032,
+      "step": 930,
+      "train_runtime": 1135.8636,
+      "train_tokens_per_second": 3049.69
+    },
+    {
+      "epoch": 2.0215285252960173,
+      "grad_norm": 3.5694122314453125,
+      "learning_rate": 7.238265038843143e-06,
+      "loss": 0.0491,
+      "num_input_tokens_seen": 3501056,
+      "step": 940,
+      "train_runtime": 1147.9804,
+      "train_tokens_per_second": 3049.752
+    },
+    {
+      "epoch": 2.0430570505920342,
+      "grad_norm": 1.0579440593719482,
+      "learning_rate": 6.951192587762555e-06,
+      "loss": 0.0263,
+      "num_input_tokens_seen": 3538464,
+      "step": 950,
+      "train_runtime": 1160.1923,
+      "train_tokens_per_second": 3049.894
+    },
+    {
+      "epoch": 2.0645855758880516,
+      "grad_norm": 1.1105297803878784,
+      "learning_rate": 6.668202056890052e-06,
+      "loss": 0.029,
+      "num_input_tokens_seen": 3575552,
+      "step": 960,
+      "train_runtime": 1172.3064,
+      "train_tokens_per_second": 3050.015
+    },
+    {
+      "epoch": 2.086114101184069,
+      "grad_norm": 2.4604287147521973,
+      "learning_rate": 6.389436963731707e-06,
+      "loss": 0.0338,
+      "num_input_tokens_seen": 3612736,
+      "step": 970,
+      "train_runtime": 1184.4757,
+      "train_tokens_per_second": 3050.072
+    },
+    {
+      "epoch": 2.1076426264800863,
+      "grad_norm": 2.9901983737945557,
+      "learning_rate": 6.115038682879904e-06,
+      "loss": 0.0427,
+      "num_input_tokens_seen": 3650464,
+      "step": 980,
+      "train_runtime": 1196.7382,
+      "train_tokens_per_second": 3050.345
+    },
+    {
+      "epoch": 2.129171151776103,
+      "grad_norm": 1.1089911460876465,
+      "learning_rate": 5.845146374315781e-06,
+      "loss": 0.0323,
+      "num_input_tokens_seen": 3687392,
+      "step": 990,
+      "train_runtime": 1208.888,
+      "train_tokens_per_second": 3050.235
+    },
+    {
+      "epoch": 2.1506996770721205,
+      "grad_norm": 4.3608317375183105,
+      "learning_rate": 5.579896912834807e-06,
+      "loss": 0.0504,
+      "num_input_tokens_seen": 3725152,
+      "step": 1000,
+      "train_runtime": 1221.1038,
+      "train_tokens_per_second": 3050.643
+    },
+    {
+      "epoch": 2.172228202368138,
+      "grad_norm": 2.5201330184936523,
+      "learning_rate": 5.319424818631293e-06,
+      "loss": 0.043,
+      "num_input_tokens_seen": 3762752,
+      "step": 1010,
+      "train_runtime": 1234.1135,
+      "train_tokens_per_second": 3048.951
+    },
+    {
+      "epoch": 2.193756727664155,
+      "grad_norm": 2.4340291023254395,
+      "learning_rate": 5.063862189076988e-06,
+      "loss": 0.0377,
+      "num_input_tokens_seen": 3800736,
+      "step": 1020,
+      "train_runtime": 1246.4039,
+      "train_tokens_per_second": 3049.361
+    },
+    {
+      "epoch": 2.215285252960172,
+      "grad_norm": 1.2626358270645142,
+      "learning_rate": 4.813338631728471e-06,
+      "loss": 0.0408,
+      "num_input_tokens_seen": 3838432,
+      "step": 1030,
+      "train_runtime": 1258.5862,
+      "train_tokens_per_second": 3049.797
+    },
+    {
+      "epoch": 2.2368137782561894,
+      "grad_norm": 2.4242489337921143,
+      "learning_rate": 4.567981198597168e-06,
+      "loss": 0.0328,
+      "num_input_tokens_seen": 3876224,
+      "step": 1040,
+      "train_runtime": 1270.8593,
+      "train_tokens_per_second": 3050.081
+    },
+    {
+      "epoch": 2.2583423035522068,
+      "grad_norm": 2.391545534133911,
+      "learning_rate": 4.327914321715459e-06,
+      "loss": 0.0259,
+      "num_input_tokens_seen": 3913216,
+      "step": 1050,
+      "train_runtime": 1282.8604,
+      "train_tokens_per_second": 3050.383
+    },
+    {
+      "epoch": 2.2798708288482237,
+      "grad_norm": 2.008685827255249,
+      "learning_rate": 4.09325975003146e-06,
+      "loss": 0.027,
+      "num_input_tokens_seen": 3950208,
+      "step": 1060,
+      "train_runtime": 1294.9339,
+      "train_tokens_per_second": 3050.509
+    },
+    {
+      "epoch": 2.301399354144241,
+      "grad_norm": 5.242519855499268,
+      "learning_rate": 3.864136487664538e-06,
+      "loss": 0.0347,
+      "num_input_tokens_seen": 3986752,
+      "step": 1070,
+      "train_runtime": 1306.9829,
+      "train_tokens_per_second": 3050.348
+    },
+    {
+      "epoch": 2.3229278794402584,
+      "grad_norm": 0.17025278508663177,
+      "learning_rate": 3.640660733552863e-06,
+      "loss": 0.0189,
+      "num_input_tokens_seen": 4023360,
+      "step": 1080,
+      "train_runtime": 1318.9572,
+      "train_tokens_per_second": 3050.41
+    },
+    {
+      "epoch": 2.3444564047362757,
+      "grad_norm": 3.2421324253082275,
+      "learning_rate": 3.4229458225235696e-06,
+      "loss": 0.0186,
+      "num_input_tokens_seen": 4061856,
+      "step": 1090,
+      "train_runtime": 1331.3712,
+      "train_tokens_per_second": 3050.882
+    },
+    {
+      "epoch": 2.3659849300322926,
+      "grad_norm": 1.6071536540985107,
+      "learning_rate": 3.211102167815469e-06,
+      "loss": 0.0366,
+      "num_input_tokens_seen": 4098464,
+      "step": 1100,
+      "train_runtime": 1343.347,
+      "train_tokens_per_second": 3050.935
+    },
+    {
+      "epoch": 2.38751345532831,
+      "grad_norm": 1.8836613893508911,
+      "learning_rate": 3.0052372050834025e-06,
+      "loss": 0.0333,
+      "num_input_tokens_seen": 4136576,
+      "step": 1110,
+      "train_runtime": 1355.7033,
+      "train_tokens_per_second": 3051.24
+    },
+    {
+      "epoch": 2.4090419806243273,
+      "grad_norm": 2.0569138526916504,
+      "learning_rate": 2.805455337912722e-06,
+      "loss": 0.0259,
+      "num_input_tokens_seen": 4174528,
+      "step": 1120,
+      "train_runtime": 1367.887,
+      "train_tokens_per_second": 3051.808
+    },
+    {
+      "epoch": 2.4305705059203446,
+      "grad_norm": 0.382569283246994,
+      "learning_rate": 2.611857884871379e-06,
+      "loss": 0.0377,
+      "num_input_tokens_seen": 4211712,
+      "step": 1130,
+      "train_runtime": 1380.0113,
+      "train_tokens_per_second": 3051.94
+    },
+    {
+      "epoch": 2.4520990312163615,
+      "grad_norm": 0.1618432253599167,
+      "learning_rate": 2.424543028126661e-06,
+      "loss": 0.0457,
+      "num_input_tokens_seen": 4248800,
+      "step": 1140,
+      "train_runtime": 1392.1418,
+      "train_tokens_per_second": 3051.988
+    },
+    {
+      "epoch": 2.473627556512379,
+      "grad_norm": 1.5764034986495972,
+      "learning_rate": 2.243605763652489e-06,
+      "loss": 0.0341,
+      "num_input_tokens_seen": 4286496,
+      "step": 1150,
+      "train_runtime": 1404.3544,
+      "train_tokens_per_second": 3052.289
+    },
+    {
+      "epoch": 2.495156081808396,
+      "grad_norm": 5.99455451965332,
+      "learning_rate": 2.0691378530525933e-06,
+      "loss": 0.0311,
+      "num_input_tokens_seen": 4323328,
+      "step": 1160,
+      "train_runtime": 1416.3868,
+      "train_tokens_per_second": 3052.364
+    },
+    {
+      "epoch": 2.5166846071044136,
+      "grad_norm": 3.6439156532287598,
+      "learning_rate": 1.9012277770239829e-06,
+      "loss": 0.0373,
+      "num_input_tokens_seen": 4360384,
+      "step": 1170,
+      "train_runtime": 1428.4267,
+      "train_tokens_per_second": 3052.578
+    },
+    {
+      "epoch": 2.5382131324004304,
+      "grad_norm": 3.2437760829925537,
+      "learning_rate": 1.7399606904843224e-06,
+      "loss": 0.0176,
+      "num_input_tokens_seen": 4397440,
+      "step": 1180,
+      "train_runtime": 1440.5831,
+      "train_tokens_per_second": 3052.542
+    },
+    {
+      "epoch": 2.559741657696448,
+      "grad_norm": 2.4674410820007324,
+      "learning_rate": 1.5854183793859577e-06,
+      "loss": 0.0264,
+      "num_input_tokens_seen": 4435584,
+      "step": 1190,
+      "train_runtime": 1452.9748,
+      "train_tokens_per_second": 3052.76
+    },
+    {
+      "epoch": 2.581270182992465,
+      "grad_norm": 1.2190111875534058,
+      "learning_rate": 1.437679219238498e-06,
+      "loss": 0.0211,
+      "num_input_tokens_seen": 4472800,
+      "step": 1200,
+      "train_runtime": 1465.0586,
+      "train_tokens_per_second": 3052.984
+    },
+    {
+      "epoch": 2.602798708288482,
+      "grad_norm": 2.0457441806793213,
+      "learning_rate": 1.2968181353609854e-06,
+      "loss": 0.0264,
+      "num_input_tokens_seen": 4510464,
+      "step": 1210,
+      "train_runtime": 1477.2352,
+      "train_tokens_per_second": 3053.315
+    },
+    {
+      "epoch": 2.6243272335844994,
+      "grad_norm": 1.473620057106018,
+      "learning_rate": 1.1629065648838495e-06,
+      "loss": 0.0369,
+      "num_input_tokens_seen": 4546592,
+      "step": 1220,
+      "train_runtime": 1489.2585,
+      "train_tokens_per_second": 3052.923
+    },
+    {
+      "epoch": 2.6458557588805167,
+      "grad_norm": 0.5360433459281921,
+      "learning_rate": 1.0360124205198102e-06,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 4583680,
+      "step": 1230,
+      "train_runtime": 1501.3452,
+      "train_tokens_per_second": 3053.049
+    },
+    {
+      "epoch": 2.667384284176534,
+      "grad_norm": 3.2004950046539307,
+      "learning_rate": 9.162000561222672e-07,
+      "loss": 0.0452,
+      "num_input_tokens_seen": 4622080,
+      "step": 1240,
+      "train_runtime": 1513.7232,
+      "train_tokens_per_second": 3053.451
+    },
+    {
+      "epoch": 2.6889128094725514,
+      "grad_norm": 0.3612094819545746,
+      "learning_rate": 8.035302340484646e-07,
+      "loss": 0.051,
+      "num_input_tokens_seen": 4658848,
+      "step": 1250,
+      "train_runtime": 1525.7739,
+      "train_tokens_per_second": 3053.433
+    },
+    {
+      "epoch": 2.7104413347685683,
+      "grad_norm": 1.8036645650863647,
+      "learning_rate": 6.980600943441273e-07,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 4695968,
+      "step": 1260,
+      "train_runtime": 1537.831,
+      "train_tokens_per_second": 3053.631
+    },
+    {
+      "epoch": 2.7319698600645856,
+      "grad_norm": 2.2066173553466797,
+      "learning_rate": 5.998431257651132e-07,
+      "loss": 0.0414,
+      "num_input_tokens_seen": 4733344,
+      "step": 1270,
+      "train_runtime": 1550.0453,
+      "train_tokens_per_second": 3053.681
+    },
+    {
+      "epoch": 2.7534983853606025,
+      "grad_norm": 2.491161823272705,
+      "learning_rate": 5.089291386508055e-07,
+      "loss": 0.0315,
+      "num_input_tokens_seen": 4770496,
+      "step": 1280,
+      "train_runtime": 1562.1107,
+      "train_tokens_per_second": 3053.878
+    },
+    {
+      "epoch": 2.77502691065662,
+      "grad_norm": 1.4545611143112183,
+      "learning_rate": 4.2536423966299256e-07,
+      "loss": 0.0336,
+      "num_input_tokens_seen": 4807456,
+      "step": 1290,
+      "train_runtime": 1574.2633,
+      "train_tokens_per_second": 3053.781
+    },
+    {
+      "epoch": 2.7965554359526372,
+      "grad_norm": 1.0926755666732788,
+      "learning_rate": 3.491908084030776e-07,
+      "loss": 0.0426,
+      "num_input_tokens_seen": 4845440,
+      "step": 1300,
+      "train_runtime": 1586.6432,
+      "train_tokens_per_second": 3053.894
+    },
+    {
+      "epoch": 2.8180839612486546,
+      "grad_norm": 2.7122678756713867,
+      "learning_rate": 2.804474759194176e-07,
+      "loss": 0.0434,
+      "num_input_tokens_seen": 4883392,
+      "step": 1310,
+      "train_runtime": 1598.92,
+      "train_tokens_per_second": 3054.182
+    },
+    {
+      "epoch": 2.839612486544672,
+      "grad_norm": 1.7297043800354004,
+      "learning_rate": 2.1916910511575627e-07,
+      "loss": 0.0297,
+      "num_input_tokens_seen": 4920384,
+      "step": 1320,
+      "train_runtime": 1610.9432,
+      "train_tokens_per_second": 3054.35
+    },
+    {
+      "epoch": 2.861141011840689,
+      "grad_norm": 2.814833641052246,
+      "learning_rate": 1.6538677307062933e-07,
+      "loss": 0.036,
+      "num_input_tokens_seen": 4957888,
+      "step": 1330,
+      "train_runtime": 1623.1577,
+      "train_tokens_per_second": 3054.471
+    },
+    {
+      "epoch": 2.882669537136706,
+      "grad_norm": 3.3779215812683105,
+      "learning_rate": 1.1912775527676278e-07,
+      "loss": 0.0209,
+      "num_input_tokens_seen": 4994944,
+      "step": 1340,
+      "train_runtime": 1635.1442,
+      "train_tokens_per_second": 3054.742
+    },
+    {
+      "epoch": 2.9041980624327235,
+      "grad_norm": 1.425120234489441,
+      "learning_rate": 8.04155118084099e-08,
+      "loss": 0.0467,
+      "num_input_tokens_seen": 5031584,
+      "step": 1350,
+      "train_runtime": 1647.097,
+      "train_tokens_per_second": 3054.82
+    },
+    {
+      "epoch": 2.9257265877287404,
+      "grad_norm": 0.8811171650886536,
+      "learning_rate": 4.9269675423671556e-08,
+      "loss": 0.0414,
+      "num_input_tokens_seen": 5068928,
+      "step": 1360,
+      "train_runtime": 1659.2581,
+      "train_tokens_per_second": 3054.936
+    },
+    {
+      "epoch": 2.9472551130247577,
+      "grad_norm": 1.2112094163894653,
+      "learning_rate": 2.570604160781753e-08,
+      "loss": 0.0182,
+      "num_input_tokens_seen": 5105984,
+      "step": 1370,
+      "train_runtime": 1671.367,
+      "train_tokens_per_second": 3054.975
+    },
+    {
+      "epoch": 2.968783638320775,
+      "grad_norm": 2.05094313621521,
+      "learning_rate": 9.736560562675912e-09,
+      "loss": 0.0236,
+      "num_input_tokens_seen": 5143296,
+      "step": 1380,
+      "train_runtime": 1683.5361,
+      "train_tokens_per_second": 3055.055
+    },
+    {
+      "epoch": 2.9903121636167924,
+      "grad_norm": 0.11877042800188065,
+      "learning_rate": 1.3693311461326908e-09,
+      "loss": 0.0323,
+      "num_input_tokens_seen": 5181056,
+      "step": 1390,
+      "train_runtime": 1695.7436,
+      "train_tokens_per_second": 3055.33
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1395,
+  "num_input_tokens_seen": 5197824,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.2251190147979674e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1395/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d57507a8b818954dbee794b7baa7eafcc21a01dd96c8730e5288c6cdd6c47d3
+size 6161

llamaboard_config.yaml

@@ -0,0 +1,85 @@
+top.booster: unsloth
+top.checkpoint_path: []
+top.finetuning_type: lora
+top.model_name: Mistral-7B-Instruct-v0.3
+top.quantization_bit: '4'
+top.quantization_method: bnb
+top.rope_scaling: none
+top.template: alpaca
+train.additional_target: ''
+train.apollo_rank: 16
+train.apollo_scale: 32
+train.apollo_target: all
+train.apollo_update_interval: 200
+train.badam_mode: layer
+train.badam_switch_interval: 50
+train.badam_switch_mode: ascending
+train.badam_update_ratio: 0.05
+train.batch_size: 4
+train.compute_type: bf16
+train.create_new_adapter: false
+train.cutoff_len: 2048
+train.dataset:
+- treino_pt_rde3
+train.dataset_dir: data
+train.ds_offload: false
+train.ds_stage: none
+train.enable_thinking: true
+train.extra_args: '{"optim": "adamw_torch"}'
+train.freeze_extra_modules: ''
+train.freeze_language_model: false
+train.freeze_multi_modal_projector: true
+train.freeze_trainable_layers: 2
+train.freeze_trainable_modules: all
+train.freeze_vision_tower: true
+train.galore_rank: 16
+train.galore_scale: 2
+train.galore_target: all
+train.galore_update_interval: 200
+train.gradient_accumulation_steps: 2
+train.image_max_pixels: 768*768
+train.image_min_pixels: 32*32
+train.learning_rate: 3e-5
+train.logging_steps: 10
+train.lora_alpha: 16
+train.lora_dropout: 0
+train.lora_rank: 8
+train.lora_target: ''
+train.loraplus_lr_ratio: 0
+train.lr_scheduler_type: cosine
+train.mask_history: false
+train.max_grad_norm: '1.0'
+train.max_samples: '3716'
+train.neat_packing: false
+train.neftune_alpha: 0
+train.num_train_epochs: '3.0'
+train.packing: false
+train.ppo_score_norm: false
+train.ppo_whiten_rewards: false
+train.pref_beta: 0.1
+train.pref_ftx: 0
+train.pref_loss: sigmoid
+train.report_to: none
+train.resize_vocab: false
+train.reward_model: []
+train.save_steps: 1000
+train.swanlab_api_key: ''
+train.swanlab_link: ''
+train.swanlab_mode: cloud
+train.swanlab_project: llamafactory
+train.swanlab_run_name: ''
+train.swanlab_workspace: ''
+train.train_on_prompt: false
+train.training_stage: Supervised Fine-Tuning
+train.use_apollo: false
+train.use_badam: false
+train.use_dora: false
+train.use_galore: false
+train.use_llama_pro: false
+train.use_pissa: false
+train.use_rslora: false
+train.use_swanlab: false
+train.val_size: 0
+train.video_max_pixels: 256*256
+train.video_min_pixels: 16*16
+train.warmup_steps: 0

running_log.txt

@@ -0,0 +1,718 @@
+[INFO|2025-10-26 16:30:13] tokenization_utils_base.py:2023 >> loading file tokenizer.model from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/tokenizer.model
+
+[INFO|2025-10-26 16:30:13] tokenization_utils_base.py:2023 >> loading file tokenizer.json from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/tokenizer.json
+
+[INFO|2025-10-26 16:30:13] tokenization_utils_base.py:2023 >> loading file added_tokens.json from cache at None
+
+[INFO|2025-10-26 16:30:13] tokenization_utils_base.py:2023 >> loading file special_tokens_map.json from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/special_tokens_map.json
+
+[INFO|2025-10-26 16:30:13] tokenization_utils_base.py:2023 >> loading file tokenizer_config.json from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/tokenizer_config.json
+
+[INFO|2025-10-26 16:30:13] tokenization_utils_base.py:2023 >> loading file chat_template.jinja from cache at None
+
+[INFO|2025-10-26 16:30:14] configuration_utils.py:698 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/config.json
+
+[INFO|2025-10-26 16:30:14] configuration_utils.py:770 >> Model config MistralConfig {
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": null,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.4",
+  "use_cache": true,
+  "vocab_size": 32768
+}
+
+
+[INFO|2025-10-26 16:30:15] tokenization_utils_base.py:2023 >> loading file tokenizer.model from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/tokenizer.model
+
+[INFO|2025-10-26 16:30:15] tokenization_utils_base.py:2023 >> loading file tokenizer.json from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/tokenizer.json
+
+[INFO|2025-10-26 16:30:15] tokenization_utils_base.py:2023 >> loading file added_tokens.json from cache at None
+
+[INFO|2025-10-26 16:30:15] tokenization_utils_base.py:2023 >> loading file special_tokens_map.json from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/special_tokens_map.json
+
+[INFO|2025-10-26 16:30:15] tokenization_utils_base.py:2023 >> loading file tokenizer_config.json from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/tokenizer_config.json
+
+[INFO|2025-10-26 16:30:15] tokenization_utils_base.py:2023 >> loading file chat_template.jinja from cache at None
+
+[INFO|2025-10-26 16:30:15] logging.py:143 >> Add pad token: </s>
+
+[INFO|2025-10-26 16:30:15] logging.py:143 >> Loading dataset treino_pt_rde3.json...
+
+[INFO|2025-10-26 16:30:18] configuration_utils.py:698 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.3/snapshots/0d4b76e1efeb5eb6f6b5e757c79870472e04bd3a/config.json
+
+[INFO|2025-10-26 16:30:18] configuration_utils.py:770 >> Model config MistralConfig {
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": null,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.4",
+  "use_cache": true,
+  "vocab_size": 32768
+}
+
+
+[INFO|2025-10-26 16:30:18] logging.py:143 >> Quantizing model to 4 bit with bitsandbytes.
+
+[INFO|2025-10-26 16:30:18] logging.py:143 >> KV cache is disabled during training.
+
+[INFO|2025-10-26 16:30:24] configuration_utils.py:698 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--unsloth--mistral-7b-instruct-v0.3-bnb-4bit/snapshots/d5f623888f1415cf89b5c208d09cb620694618ee/config.json
+
+[INFO|2025-10-26 16:30:24] configuration_utils.py:770 >> Model config MistralConfig {
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pad_token_id": 770,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "uint8",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": true,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.4",
+  "unsloth_version": "2024.9",
+  "use_cache": true,
+  "vocab_size": 32768
+}
+
+
+[INFO|2025-10-26 16:30:28] configuration_utils.py:698 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--unsloth--mistral-7b-instruct-v0.3-bnb-4bit/snapshots/d5f623888f1415cf89b5c208d09cb620694618ee/config.json
+
+[INFO|2025-10-26 16:30:28] configuration_utils.py:770 >> Model config MistralConfig {
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pad_token_id": 770,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "uint8",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": true,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.4",
+  "unsloth_version": "2024.9",
+  "use_cache": true,
+  "vocab_size": 32768
+}
+
+
+[INFO|2025-10-26 16:30:28] configuration_utils.py:698 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--unsloth--mistral-7b-instruct-v0.3-bnb-4bit/snapshots/d5f623888f1415cf89b5c208d09cb620694618ee/config.json
+
+[INFO|2025-10-26 16:30:28] configuration_utils.py:770 >> Model config MistralConfig {
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pad_token_id": 770,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "uint8",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": true,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.4",
+  "unsloth_version": "2024.9",
+  "use_cache": true,
+  "vocab_size": 32768
+}
+
+
+[INFO|2025-10-26 16:30:43] modeling_utils.py:1151 >> loading weights file model.safetensors from cache at /root/.cache/huggingface/hub/models--unsloth--mistral-7b-instruct-v0.3-bnb-4bit/snapshots/d5f623888f1415cf89b5c208d09cb620694618ee/model.safetensors
+
+[INFO|2025-10-26 16:30:43] modeling_utils.py:2241 >> Instantiating MistralForCausalLM model under default dtype torch.bfloat16.
+
+[INFO|2025-10-26 16:30:43] configuration_utils.py:1135 >> Generate config GenerationConfig {
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 770
+}
+
+
+[INFO|2025-10-26 16:30:45] modeling_utils.py:5131 >> All model checkpoint weights were used when initializing MistralForCausalLM.
+
+
+[INFO|2025-10-26 16:30:45] modeling_utils.py:5139 >> All the weights of MistralForCausalLM were initialized from the model checkpoint at unsloth/mistral-7b-instruct-v0.3-bnb-4bit.
+If your task is similar to the task the model of the checkpoint was trained on, you can already use MistralForCausalLM for predictions without further training.
+
+[INFO|2025-10-26 16:30:46] configuration_utils.py:1090 >> loading configuration file generation_config.json from cache at /root/.cache/huggingface/hub/models--unsloth--mistral-7b-instruct-v0.3-bnb-4bit/snapshots/d5f623888f1415cf89b5c208d09cb620694618ee/generation_config.json
+
+[INFO|2025-10-26 16:30:46] configuration_utils.py:1135 >> Generate config GenerationConfig {
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "max_length": 32768,
+  "pad_token_id": 770
+}
+
+
+[INFO|2025-10-26 16:30:50] logging.py:143 >> Gradient checkpointing enabled.
+
+[INFO|2025-10-26 16:30:50] logging.py:143 >> Upcasting trainable params to float32.
+
+[INFO|2025-10-26 16:30:50] logging.py:143 >> Fine-tuning method: LoRA
+
+[INFO|2025-10-26 16:30:50] logging.py:143 >> Found linear modules: down_proj,v_proj,q_proj,k_proj,o_proj,up_proj,gate_proj
+
+[WARNING|2025-10-26 16:30:53] logging.py:328 >> Unsloth 2025.10.9 patched 32 layers with 32 QKV layers, 32 O layers and 32 MLP layers.
+
+[INFO|2025-10-26 16:30:56] logging.py:143 >> trainable params: 20,971,520 || all params: 7,268,995,072 || trainable%: 0.2885
+
+[INFO|2025-10-26 16:30:56] trainer.py:756 >> Using auto half precision backend
+
+[WARNING|2025-10-26 16:30:56] <string>:173 >> ==((====))==  Unsloth - 2x faster free finetuning | Num GPUs used = 1
+   \\   /|    Num examples = 3,716 | Num Epochs = 3 | Total steps = 1,395
+O^O/ \_/ \    Batch size per device = 4 | Gradient accumulation steps = 2
+\        /    Data Parallel GPUs = 1 | Total batch size (4 x 2 x 1) = 8
+ "-____-"     Trainable parameters = 20,971,520 of 7,268,995,072 (0.29% trained)
+
+[INFO|2025-10-26 16:31:18] logging.py:143 >> {'loss': 0.6799, 'learning_rate': 2.9997e-05, 'epoch': 0.02, 'throughput': 1796.76}
+
+[INFO|2025-10-26 16:31:30] logging.py:143 >> {'loss': 0.1361, 'learning_rate': 2.9986e-05, 'epoch': 0.04, 'throughput': 2255.76}
+
+[INFO|2025-10-26 16:31:43] logging.py:143 >> {'loss': 0.1558, 'learning_rate': 2.9968e-05, 'epoch': 0.06, 'throughput': 2446.49}
+
+[INFO|2025-10-26 16:31:55] logging.py:143 >> {'loss': 0.1344, 'learning_rate': 2.9942e-05, 'epoch': 0.09, 'throughput': 2584.35}
+
+[INFO|2025-10-26 16:32:07] logging.py:143 >> {'loss': 0.1262, 'learning_rate': 2.9909e-05, 'epoch': 0.11, 'throughput': 2671.91}
+
+[INFO|2025-10-26 16:32:19] logging.py:143 >> {'loss': 0.1182, 'learning_rate': 2.9868e-05, 'epoch': 0.13, 'throughput': 2735.61}
+
+[INFO|2025-10-26 16:32:31] logging.py:143 >> {'loss': 0.1333, 'learning_rate': 2.9819e-05, 'epoch': 0.15, 'throughput': 2777.09}
+
+[INFO|2025-10-26 16:32:44] logging.py:143 >> {'loss': 0.0925, 'learning_rate': 2.9763e-05, 'epoch': 0.17, 'throughput': 2807.44}
+
+[INFO|2025-10-26 16:32:55] logging.py:143 >> {'loss': 0.0886, 'learning_rate': 2.9700e-05, 'epoch': 0.19, 'throughput': 2834.45}
+
+[INFO|2025-10-26 16:33:07] logging.py:143 >> {'loss': 0.1127, 'learning_rate': 2.9629e-05, 'epoch': 0.22, 'throughput': 2857.99}
+
+[INFO|2025-10-26 16:33:19] logging.py:143 >> {'loss': 0.1018, 'learning_rate': 2.9550e-05, 'epoch': 0.24, 'throughput': 2876.81}
+
+[INFO|2025-10-26 16:33:32] logging.py:143 >> {'loss': 0.1001, 'learning_rate': 2.9465e-05, 'epoch': 0.26, 'throughput': 2893.53}
+
+[INFO|2025-10-26 16:33:44] logging.py:143 >> {'loss': 0.0962, 'learning_rate': 2.9371e-05, 'epoch': 0.28, 'throughput': 2908.32}
+
+[INFO|2025-10-26 16:33:56] logging.py:143 >> {'loss': 0.1076, 'learning_rate': 2.9271e-05, 'epoch': 0.30, 'throughput': 2919.87}
+
+[INFO|2025-10-26 16:34:08] logging.py:143 >> {'loss': 0.0886, 'learning_rate': 2.9163e-05, 'epoch': 0.32, 'throughput': 2928.43}
+
+[INFO|2025-10-26 16:34:20] logging.py:143 >> {'loss': 0.0861, 'learning_rate': 2.9049e-05, 'epoch': 0.34, 'throughput': 2937.16}
+
+[INFO|2025-10-26 16:34:32] logging.py:143 >> {'loss': 0.1158, 'learning_rate': 2.8927e-05, 'epoch': 0.37, 'throughput': 2942.81}
+
+[INFO|2025-10-26 16:34:44] logging.py:143 >> {'loss': 0.0931, 'learning_rate': 2.8798e-05, 'epoch': 0.39, 'throughput': 2950.21}
+
+[INFO|2025-10-26 16:34:56] logging.py:143 >> {'loss': 0.0964, 'learning_rate': 2.8662e-05, 'epoch': 0.41, 'throughput': 2957.11}
+
+[INFO|2025-10-26 16:35:09] logging.py:143 >> {'loss': 0.0708, 'learning_rate': 2.8519e-05, 'epoch': 0.43, 'throughput': 2963.94}
+
+[INFO|2025-10-26 16:35:21] logging.py:143 >> {'loss': 0.0954, 'learning_rate': 2.8369e-05, 'epoch': 0.45, 'throughput': 2968.41}
+
+[INFO|2025-10-26 16:35:33] logging.py:143 >> {'loss': 0.0897, 'learning_rate': 2.8212e-05, 'epoch': 0.47, 'throughput': 2973.98}
+
+[INFO|2025-10-26 16:35:45] logging.py:143 >> {'loss': 0.0783, 'learning_rate': 2.8049e-05, 'epoch': 0.50, 'throughput': 2977.07}
+
+[INFO|2025-10-26 16:35:57] logging.py:143 >> {'loss': 0.0920, 'learning_rate': 2.7879e-05, 'epoch': 0.52, 'throughput': 2981.87}
+
+[INFO|2025-10-26 16:36:09] logging.py:143 >> {'loss': 0.1014, 'learning_rate': 2.7703e-05, 'epoch': 0.54, 'throughput': 2984.96}
+
+[INFO|2025-10-26 16:36:21] logging.py:143 >> {'loss': 0.1032, 'learning_rate': 2.7520e-05, 'epoch': 0.56, 'throughput': 2989.88}
+
+[INFO|2025-10-26 16:36:33] logging.py:143 >> {'loss': 0.1016, 'learning_rate': 2.7331e-05, 'epoch': 0.58, 'throughput': 2992.79}
+
+[INFO|2025-10-26 16:36:46] logging.py:143 >> {'loss': 0.0843, 'learning_rate': 2.7135e-05, 'epoch': 0.60, 'throughput': 2996.74}
+
+[INFO|2025-10-26 16:36:58] logging.py:143 >> {'loss': 0.1012, 'learning_rate': 2.6934e-05, 'epoch': 0.62, 'throughput': 3000.16}
+
+[INFO|2025-10-26 16:37:10] logging.py:143 >> {'loss': 0.0835, 'learning_rate': 2.6726e-05, 'epoch': 0.65, 'throughput': 3001.47}
+
+[INFO|2025-10-26 16:37:22] logging.py:143 >> {'loss': 0.0884, 'learning_rate': 2.6512e-05, 'epoch': 0.67, 'throughput': 3002.24}
+
+[INFO|2025-10-26 16:37:34] logging.py:143 >> {'loss': 0.0907, 'learning_rate': 2.6293e-05, 'epoch': 0.69, 'throughput': 3004.35}
+
+[INFO|2025-10-26 16:37:47] logging.py:143 >> {'loss': 0.0786, 'learning_rate': 2.6068e-05, 'epoch': 0.71, 'throughput': 3006.91}
+
+[INFO|2025-10-26 16:37:59] logging.py:143 >> {'loss': 0.0863, 'learning_rate': 2.5837e-05, 'epoch': 0.73, 'throughput': 3008.52}
+
+[INFO|2025-10-26 16:38:11] logging.py:143 >> {'loss': 0.0976, 'learning_rate': 2.5601e-05, 'epoch': 0.75, 'throughput': 3012.05}
+
+[INFO|2025-10-26 16:38:23] logging.py:143 >> {'loss': 0.0914, 'learning_rate': 2.5359e-05, 'epoch': 0.78, 'throughput': 3013.82}
+
+[INFO|2025-10-26 16:38:36] logging.py:143 >> {'loss': 0.0997, 'learning_rate': 2.5112e-05, 'epoch': 0.80, 'throughput': 3016.75}
+
+[INFO|2025-10-26 16:38:48] logging.py:143 >> {'loss': 0.0689, 'learning_rate': 2.4860e-05, 'epoch': 0.82, 'throughput': 3019.08}
+
+[INFO|2025-10-26 16:39:00] logging.py:143 >> {'loss': 0.1062, 'learning_rate': 2.4603e-05, 'epoch': 0.84, 'throughput': 3020.05}
+
+[INFO|2025-10-26 16:39:12] logging.py:143 >> {'loss': 0.0579, 'learning_rate': 2.4341e-05, 'epoch': 0.86, 'throughput': 3020.61}
+
+[INFO|2025-10-26 16:39:24] logging.py:143 >> {'loss': 0.0740, 'learning_rate': 2.4074e-05, 'epoch': 0.88, 'throughput': 3022.13}
+
+[INFO|2025-10-26 16:39:36] logging.py:143 >> {'loss': 0.0774, 'learning_rate': 2.3803e-05, 'epoch': 0.90, 'throughput': 3023.51}
+
+[INFO|2025-10-26 16:39:48] logging.py:143 >> {'loss': 0.0771, 'learning_rate': 2.3527e-05, 'epoch': 0.93, 'throughput': 3025.20}
+
+[INFO|2025-10-26 16:40:00] logging.py:143 >> {'loss': 0.0750, 'learning_rate': 2.3247e-05, 'epoch': 0.95, 'throughput': 3025.01}
+
+[INFO|2025-10-26 16:40:12] logging.py:143 >> {'loss': 0.0785, 'learning_rate': 2.2963e-05, 'epoch': 0.97, 'throughput': 3025.73}
+
+[INFO|2025-10-26 16:40:24] logging.py:143 >> {'loss': 0.0740, 'learning_rate': 2.2675e-05, 'epoch': 0.99, 'throughput': 3026.88}
+
+[INFO|2025-10-26 16:40:36] logging.py:143 >> {'loss': 0.0653, 'learning_rate': 2.2383e-05, 'epoch': 1.01, 'throughput': 3028.66}
+
+[INFO|2025-10-26 16:40:48] logging.py:143 >> {'loss': 0.0632, 'learning_rate': 2.2087e-05, 'epoch': 1.03, 'throughput': 3029.41}
+
+[INFO|2025-10-26 16:41:00] logging.py:143 >> {'loss': 0.0472, 'learning_rate': 2.1787e-05, 'epoch': 1.05, 'throughput': 3030.79}
+
+[INFO|2025-10-26 16:41:12] logging.py:143 >> {'loss': 0.0534, 'learning_rate': 2.1484e-05, 'epoch': 1.08, 'throughput': 3032.11}
+
+[INFO|2025-10-26 16:41:24] logging.py:143 >> {'loss': 0.0751, 'learning_rate': 2.1178e-05, 'epoch': 1.10, 'throughput': 3032.88}
+
+[INFO|2025-10-26 16:41:37] logging.py:143 >> {'loss': 0.0747, 'learning_rate': 2.0869e-05, 'epoch': 1.12, 'throughput': 3033.66}
+
+[INFO|2025-10-26 16:41:49] logging.py:143 >> {'loss': 0.0492, 'learning_rate': 2.0556e-05, 'epoch': 1.14, 'throughput': 3034.22}
+
+[INFO|2025-10-26 16:42:01] logging.py:143 >> {'loss': 0.0651, 'learning_rate': 2.0241e-05, 'epoch': 1.16, 'throughput': 3034.64}
+
+[INFO|2025-10-26 16:42:13] logging.py:143 >> {'loss': 0.0565, 'learning_rate': 1.9923e-05, 'epoch': 1.18, 'throughput': 3035.81}
+
+[INFO|2025-10-26 16:42:25] logging.py:143 >> {'loss': 0.0696, 'learning_rate': 1.9603e-05, 'epoch': 1.20, 'throughput': 3036.73}
+
+[INFO|2025-10-26 16:42:37] logging.py:143 >> {'loss': 0.0553, 'learning_rate': 1.9280e-05, 'epoch': 1.23, 'throughput': 3037.22}
+
+[INFO|2025-10-26 16:42:49] logging.py:143 >> {'loss': 0.0666, 'learning_rate': 1.8956e-05, 'epoch': 1.25, 'throughput': 3037.97}
+
+[INFO|2025-10-26 16:43:01] logging.py:143 >> {'loss': 0.0449, 'learning_rate': 1.8629e-05, 'epoch': 1.27, 'throughput': 3038.24}
+
+[INFO|2025-10-26 16:43:13] logging.py:143 >> {'loss': 0.0522, 'learning_rate': 1.8300e-05, 'epoch': 1.29, 'throughput': 3038.41}
+
+[INFO|2025-10-26 16:43:26] logging.py:143 >> {'loss': 0.0638, 'learning_rate': 1.7970e-05, 'epoch': 1.31, 'throughput': 3039.22}
+
+[INFO|2025-10-26 16:43:38] logging.py:143 >> {'loss': 0.0683, 'learning_rate': 1.7638e-05, 'epoch': 1.33, 'throughput': 3039.15}
+
+[INFO|2025-10-26 16:43:50] logging.py:143 >> {'loss': 0.0678, 'learning_rate': 1.7305e-05, 'epoch': 1.36, 'throughput': 3039.88}
+
+[INFO|2025-10-26 16:44:02] logging.py:143 >> {'loss': 0.0502, 'learning_rate': 1.6970e-05, 'epoch': 1.38, 'throughput': 3040.29}
+
+[INFO|2025-10-26 16:44:14] logging.py:143 >> {'loss': 0.0728, 'learning_rate': 1.6635e-05, 'epoch': 1.40, 'throughput': 3040.28}
+
+[INFO|2025-10-26 16:44:26] logging.py:143 >> {'loss': 0.0590, 'learning_rate': 1.6299e-05, 'epoch': 1.42, 'throughput': 3040.16}
+
+[INFO|2025-10-26 16:44:39] logging.py:143 >> {'loss': 0.0514, 'learning_rate': 1.5962e-05, 'epoch': 1.44, 'throughput': 3040.77}
+
+[INFO|2025-10-26 16:44:51] logging.py:143 >> {'loss': 0.0689, 'learning_rate': 1.5625e-05, 'epoch': 1.46, 'throughput': 3041.47}
+
+[INFO|2025-10-26 16:45:03] logging.py:143 >> {'loss': 0.0338, 'learning_rate': 1.5287e-05, 'epoch': 1.48, 'throughput': 3041.91}
+
+[INFO|2025-10-26 16:45:15] logging.py:143 >> {'loss': 0.0621, 'learning_rate': 1.4949e-05, 'epoch': 1.51, 'throughput': 3042.48}
+
+[INFO|2025-10-26 16:45:27] logging.py:143 >> {'loss': 0.0570, 'learning_rate': 1.4612e-05, 'epoch': 1.53, 'throughput': 3042.86}
+
+[INFO|2025-10-26 16:45:39] logging.py:143 >> {'loss': 0.0632, 'learning_rate': 1.4274e-05, 'epoch': 1.55, 'throughput': 3043.37}
+
+[INFO|2025-10-26 16:45:52] logging.py:143 >> {'loss': 0.0361, 'learning_rate': 1.3937e-05, 'epoch': 1.57, 'throughput': 3041.59}
+
+[INFO|2025-10-26 16:46:04] logging.py:143 >> {'loss': 0.0503, 'learning_rate': 1.3600e-05, 'epoch': 1.59, 'throughput': 3041.96}
+
+[INFO|2025-10-26 16:46:16] logging.py:143 >> {'loss': 0.0843, 'learning_rate': 1.3264e-05, 'epoch': 1.61, 'throughput': 3042.71}
+
+[INFO|2025-10-26 16:46:28] logging.py:143 >> {'loss': 0.0500, 'learning_rate': 1.2929e-05, 'epoch': 1.64, 'throughput': 3042.97}
+
+[INFO|2025-10-26 16:46:40] logging.py:143 >> {'loss': 0.0608, 'learning_rate': 1.2595e-05, 'epoch': 1.66, 'throughput': 3043.39}
+
+[INFO|2025-10-26 16:46:53] logging.py:143 >> {'loss': 0.0422, 'learning_rate': 1.2262e-05, 'epoch': 1.68, 'throughput': 3044.09}
+
+[INFO|2025-10-26 16:47:04] logging.py:143 >> {'loss': 0.0558, 'learning_rate': 1.1931e-05, 'epoch': 1.70, 'throughput': 3043.47}
+
+[INFO|2025-10-26 16:47:17] logging.py:143 >> {'loss': 0.0509, 'learning_rate': 1.1601e-05, 'epoch': 1.72, 'throughput': 3043.89}
+
+[INFO|2025-10-26 16:47:29] logging.py:143 >> {'loss': 0.0674, 'learning_rate': 1.1273e-05, 'epoch': 1.74, 'throughput': 3044.50}
+
+[INFO|2025-10-26 16:47:41] logging.py:143 >> {'loss': 0.0694, 'learning_rate': 1.0947e-05, 'epoch': 1.76, 'throughput': 3044.95}
+
+[INFO|2025-10-26 16:47:53] logging.py:143 >> {'loss': 0.0383, 'learning_rate': 1.0623e-05, 'epoch': 1.79, 'throughput': 3045.52}
+
+[INFO|2025-10-26 16:48:05] logging.py:143 >> {'loss': 0.0547, 'learning_rate': 1.0301e-05, 'epoch': 1.81, 'throughput': 3046.03}
+
+[INFO|2025-10-26 16:48:17] logging.py:143 >> {'loss': 0.0516, 'learning_rate': 9.9810e-06, 'epoch': 1.83, 'throughput': 3046.52}
+
+[INFO|2025-10-26 16:48:29] logging.py:143 >> {'loss': 0.0628, 'learning_rate': 9.6639e-06, 'epoch': 1.85, 'throughput': 3047.17}
+
+[INFO|2025-10-26 16:48:42] logging.py:143 >> {'loss': 0.0733, 'learning_rate': 9.3496e-06, 'epoch': 1.87, 'throughput': 3047.51}
+
+[INFO|2025-10-26 16:48:54] logging.py:143 >> {'loss': 0.0727, 'learning_rate': 9.0381e-06, 'epoch': 1.89, 'throughput': 3047.92}
+
+[INFO|2025-10-26 16:49:06] logging.py:143 >> {'loss': 0.0580, 'learning_rate': 8.7297e-06, 'epoch': 1.91, 'throughput': 3048.21}
+
+[INFO|2025-10-26 16:49:17] logging.py:143 >> {'loss': 0.0588, 'learning_rate': 8.4244e-06, 'epoch': 1.94, 'throughput': 3048.41}
+
+[INFO|2025-10-26 16:49:29] logging.py:143 >> {'loss': 0.0375, 'learning_rate': 8.1225e-06, 'epoch': 1.96, 'throughput': 3048.92}
+
+[INFO|2025-10-26 16:49:42] logging.py:143 >> {'loss': 0.0644, 'learning_rate': 7.8241e-06, 'epoch': 1.98, 'throughput': 3049.05}
+
+[INFO|2025-10-26 16:49:53] logging.py:143 >> {'loss': 0.0347, 'learning_rate': 7.5293e-06, 'epoch': 2.00, 'throughput': 3049.71}
+
+[INFO|2025-10-26 16:50:05] logging.py:143 >> {'loss': 0.0491, 'learning_rate': 7.2383e-06, 'epoch': 2.02, 'throughput': 3049.77}
+
+[INFO|2025-10-26 16:50:18] logging.py:143 >> {'loss': 0.0263, 'learning_rate': 6.9512e-06, 'epoch': 2.04, 'throughput': 3049.91}
+
+[INFO|2025-10-26 16:50:30] logging.py:143 >> {'loss': 0.0290, 'learning_rate': 6.6682e-06, 'epoch': 2.06, 'throughput': 3050.03}
+
+[INFO|2025-10-26 16:50:42] logging.py:143 >> {'loss': 0.0338, 'learning_rate': 6.3894e-06, 'epoch': 2.09, 'throughput': 3050.09}
+
+[INFO|2025-10-26 16:50:54] logging.py:143 >> {'loss': 0.0427, 'learning_rate': 6.1150e-06, 'epoch': 2.11, 'throughput': 3050.36}
+
+[INFO|2025-10-26 16:51:06] logging.py:143 >> {'loss': 0.0323, 'learning_rate': 5.8451e-06, 'epoch': 2.13, 'throughput': 3050.25}
+
+[INFO|2025-10-26 16:51:18] logging.py:143 >> {'loss': 0.0504, 'learning_rate': 5.5799e-06, 'epoch': 2.15, 'throughput': 3050.66}
+
+[INFO|2025-10-26 16:51:18] trainer.py:3993 >> Saving model checkpoint to saves/Mistral-7B-Instruct-v0.3/lora/mistral-treinado/checkpoint-1000
+
+[INFO|2025-10-26 16:51:19] configuration_utils.py:698 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--unsloth--mistral-7b-instruct-v0.3-bnb-4bit/snapshots/d5f623888f1415cf89b5c208d09cb620694618ee/config.json
+
+[INFO|2025-10-26 16:51:19] configuration_utils.py:770 >> Model config MistralConfig {
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pad_token_id": 770,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "uint8",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": true,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.4",
+  "unsloth_version": "2024.9",
+  "use_cache": true,
+  "vocab_size": 32768
+}
+
+
+[INFO|2025-10-26 16:51:31] logging.py:143 >> {'loss': 0.0430, 'learning_rate': 5.3194e-06, 'epoch': 2.17, 'throughput': 3048.97}
+
+[INFO|2025-10-26 16:51:44] logging.py:143 >> {'loss': 0.0377, 'learning_rate': 5.0639e-06, 'epoch': 2.19, 'throughput': 3049.38}
+
+[INFO|2025-10-26 16:51:56] logging.py:143 >> {'loss': 0.0408, 'learning_rate': 4.8133e-06, 'epoch': 2.22, 'throughput': 3049.81}
+
+[INFO|2025-10-26 16:52:08] logging.py:143 >> {'loss': 0.0328, 'learning_rate': 4.5680e-06, 'epoch': 2.24, 'throughput': 3050.10}
+
+[INFO|2025-10-26 16:52:20] logging.py:143 >> {'loss': 0.0259, 'learning_rate': 4.3279e-06, 'epoch': 2.26, 'throughput': 3050.40}
+
+[INFO|2025-10-26 16:52:32] logging.py:143 >> {'loss': 0.0270, 'learning_rate': 4.0933e-06, 'epoch': 2.28, 'throughput': 3050.52}
+
+[INFO|2025-10-26 16:52:44] logging.py:143 >> {'loss': 0.0347, 'learning_rate': 3.8641e-06, 'epoch': 2.30, 'throughput': 3050.36}
+
+[INFO|2025-10-26 16:52:56] logging.py:143 >> {'loss': 0.0189, 'learning_rate': 3.6407e-06, 'epoch': 2.32, 'throughput': 3050.42}
+
+[INFO|2025-10-26 16:53:09] logging.py:143 >> {'loss': 0.0186, 'learning_rate': 3.4229e-06, 'epoch': 2.34, 'throughput': 3050.89}
+
+[INFO|2025-10-26 16:53:21] logging.py:143 >> {'loss': 0.0366, 'learning_rate': 3.2111e-06, 'epoch': 2.37, 'throughput': 3050.95}
+
+[INFO|2025-10-26 16:53:33] logging.py:143 >> {'loss': 0.0333, 'learning_rate': 3.0052e-06, 'epoch': 2.39, 'throughput': 3051.25}
+
+[INFO|2025-10-26 16:53:45] logging.py:143 >> {'loss': 0.0259, 'learning_rate': 2.8055e-06, 'epoch': 2.41, 'throughput': 3051.82}
+
+[INFO|2025-10-26 16:53:57] logging.py:143 >> {'loss': 0.0377, 'learning_rate': 2.6119e-06, 'epoch': 2.43, 'throughput': 3051.95}
+
+[INFO|2025-10-26 16:54:10] logging.py:143 >> {'loss': 0.0457, 'learning_rate': 2.4245e-06, 'epoch': 2.45, 'throughput': 3052.00}
+
+[INFO|2025-10-26 16:54:22] logging.py:143 >> {'loss': 0.0341, 'learning_rate': 2.2436e-06, 'epoch': 2.47, 'throughput': 3052.30}
+
+[INFO|2025-10-26 16:54:34] logging.py:143 >> {'loss': 0.0311, 'learning_rate': 2.0691e-06, 'epoch': 2.50, 'throughput': 3052.38}
+
+[INFO|2025-10-26 16:54:46] logging.py:143 >> {'loss': 0.0373, 'learning_rate': 1.9012e-06, 'epoch': 2.52, 'throughput': 3052.59}
+
+[INFO|2025-10-26 16:54:58] logging.py:143 >> {'loss': 0.0176, 'learning_rate': 1.7400e-06, 'epoch': 2.54, 'throughput': 3052.55}
+
+[INFO|2025-10-26 16:55:10] logging.py:143 >> {'loss': 0.0264, 'learning_rate': 1.5854e-06, 'epoch': 2.56, 'throughput': 3052.77}
+
+[INFO|2025-10-26 16:55:22] logging.py:143 >> {'loss': 0.0211, 'learning_rate': 1.4377e-06, 'epoch': 2.58, 'throughput': 3053.00}
+
+[INFO|2025-10-26 16:55:35] logging.py:143 >> {'loss': 0.0264, 'learning_rate': 1.2968e-06, 'epoch': 2.60, 'throughput': 3053.33}
+
+[INFO|2025-10-26 16:55:47] logging.py:143 >> {'loss': 0.0369, 'learning_rate': 1.1629e-06, 'epoch': 2.62, 'throughput': 3052.94}
+
+[INFO|2025-10-26 16:55:59] logging.py:143 >> {'loss': 0.0285, 'learning_rate': 1.0360e-06, 'epoch': 2.65, 'throughput': 3053.06}
+
+[INFO|2025-10-26 16:56:11] logging.py:143 >> {'loss': 0.0452, 'learning_rate': 9.1620e-07, 'epoch': 2.67, 'throughput': 3053.46}
+
+[INFO|2025-10-26 16:56:23] logging.py:143 >> {'loss': 0.0510, 'learning_rate': 8.0353e-07, 'epoch': 2.69, 'throughput': 3053.44}
+
+[INFO|2025-10-26 16:56:35] logging.py:143 >> {'loss': 0.0217, 'learning_rate': 6.9806e-07, 'epoch': 2.71, 'throughput': 3053.64}
+
+[INFO|2025-10-26 16:56:47] logging.py:143 >> {'loss': 0.0414, 'learning_rate': 5.9984e-07, 'epoch': 2.73, 'throughput': 3053.69}
+
+[INFO|2025-10-26 16:56:59] logging.py:143 >> {'loss': 0.0315, 'learning_rate': 5.0893e-07, 'epoch': 2.75, 'throughput': 3053.89}
+
+[INFO|2025-10-26 16:57:12] logging.py:143 >> {'loss': 0.0336, 'learning_rate': 4.2536e-07, 'epoch': 2.78, 'throughput': 3053.79}
+
+[INFO|2025-10-26 16:57:24] logging.py:143 >> {'loss': 0.0426, 'learning_rate': 3.4919e-07, 'epoch': 2.80, 'throughput': 3053.90}
+
+[INFO|2025-10-26 16:57:36] logging.py:143 >> {'loss': 0.0434, 'learning_rate': 2.8045e-07, 'epoch': 2.82, 'throughput': 3054.19}
+
+[INFO|2025-10-26 16:57:48] logging.py:143 >> {'loss': 0.0297, 'learning_rate': 2.1917e-07, 'epoch': 2.84, 'throughput': 3054.36}
+
+[INFO|2025-10-26 16:58:01] logging.py:143 >> {'loss': 0.0360, 'learning_rate': 1.6539e-07, 'epoch': 2.86, 'throughput': 3054.48}
+
+[INFO|2025-10-26 16:58:13] logging.py:143 >> {'loss': 0.0209, 'learning_rate': 1.1913e-07, 'epoch': 2.88, 'throughput': 3054.75}
+
+[INFO|2025-10-26 16:58:24] logging.py:143 >> {'loss': 0.0467, 'learning_rate': 8.0416e-08, 'epoch': 2.90, 'throughput': 3054.83}
+
+[INFO|2025-10-26 16:58:37] logging.py:143 >> {'loss': 0.0414, 'learning_rate': 4.9270e-08, 'epoch': 2.93, 'throughput': 3054.95}
+
+[INFO|2025-10-26 16:58:49] logging.py:143 >> {'loss': 0.0182, 'learning_rate': 2.5706e-08, 'epoch': 2.95, 'throughput': 3054.99}
+
+[INFO|2025-10-26 16:59:01] logging.py:143 >> {'loss': 0.0236, 'learning_rate': 9.7366e-09, 'epoch': 2.97, 'throughput': 3055.07}
+
+[INFO|2025-10-26 16:59:13] logging.py:143 >> {'loss': 0.0323, 'learning_rate': 1.3693e-09, 'epoch': 2.99, 'throughput': 3055.34}
+
+[INFO|2025-10-26 16:59:19] trainer.py:3993 >> Saving model checkpoint to saves/Mistral-7B-Instruct-v0.3/lora/mistral-treinado/checkpoint-1395
+
+[INFO|2025-10-26 16:59:19] configuration_utils.py:698 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--unsloth--mistral-7b-instruct-v0.3-bnb-4bit/snapshots/d5f623888f1415cf89b5c208d09cb620694618ee/config.json
+
+[INFO|2025-10-26 16:59:19] configuration_utils.py:770 >> Model config MistralConfig {
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pad_token_id": 770,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "uint8",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": true,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.4",
+  "unsloth_version": "2024.9",
+  "use_cache": true,
+  "vocab_size": 32768
+}
+
+
+[INFO|2025-10-26 16:59:19] <string>:438 >> 
+
+Training completed. Do not forget to share your model on huggingface.co/models =)
+
+
+
+[INFO|2025-10-26 16:59:19] trainer.py:3993 >> Saving model checkpoint to saves/Mistral-7B-Instruct-v0.3/lora/mistral-treinado
+
+[INFO|2025-10-26 16:59:20] configuration_utils.py:698 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--unsloth--mistral-7b-instruct-v0.3-bnb-4bit/snapshots/d5f623888f1415cf89b5c208d09cb620694618ee/config.json
+
+[INFO|2025-10-26 16:59:20] configuration_utils.py:770 >> Model config MistralConfig {
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pad_token_id": 770,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "uint8",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": true,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.4",
+  "unsloth_version": "2024.9",
+  "use_cache": true,
+  "vocab_size": 32768
+}
+
+
+[WARNING|2025-10-26 16:59:20] logging.py:148 >> No metric eval_loss to plot.
+
+[WARNING|2025-10-26 16:59:20] logging.py:148 >> No metric eval_accuracy to plot.
+
+[INFO|2025-10-26 16:59:20] modelcard.py:450 >> Dropping the following result as it does not have all the necessary fields:
+{'task': {'name': 'Causal Language Modeling', 'type': 'text-generation'}}
+

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37f00374dea48658ee8f5d0f21895b9bc55cb0103939607c8185bfd1c6ca1f89
+size 587404

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.0,
+    "num_input_tokens_seen": 5197824,
+    "total_flos": 2.2251190147979674e+17,
+    "train_loss": 0.06651546376153132,
+    "train_runtime": 1702.0416,
+    "train_samples_per_second": 6.55,
+    "train_steps_per_second": 0.82
+}

trainer_log.jsonl

@@ -0,0 +1,140 @@
+{"current_steps": 10, "total_steps": 1395, "loss": 0.6799, "lr": 2.9996919063517692e-05, "epoch": 0.021528525296017224, "percentage": 0.72, "elapsed_time": "0:00:20", "remaining_time": "0:47:43", "throughput": 1796.76, "total_tokens": 37152}
+{"current_steps": 20, "total_steps": 1395, "loss": 0.1361, "lr": 2.9986270537572827e-05, "epoch": 0.04305705059203445, "percentage": 1.43, "elapsed_time": "0:00:32", "remaining_time": "0:37:30", "throughput": 2255.76, "total_tokens": 73856}
+{"current_steps": 30, "total_steps": 1395, "loss": 0.1558, "lr": 2.9968021784983436e-05, "epoch": 0.06458557588805167, "percentage": 2.15, "elapsed_time": "0:00:45", "remaining_time": "0:34:28", "throughput": 2446.49, "total_tokens": 111232}
+{"current_steps": 40, "total_steps": 1395, "loss": 0.1344, "lr": 2.994218206053077e-05, "epoch": 0.0861141011840689, "percentage": 2.87, "elapsed_time": "0:00:57", "remaining_time": "0:32:32", "throughput": 2584.35, "total_tokens": 148992}
+{"current_steps": 50, "total_steps": 1395, "loss": 0.1262, "lr": 2.9908764468726846e-05, "epoch": 0.10764262648008611, "percentage": 3.58, "elapsed_time": "0:01:09", "remaining_time": "0:31:14", "throughput": 2671.91, "total_tokens": 186176}
+{"current_steps": 60, "total_steps": 1395, "loss": 0.1182, "lr": 2.986778595716853e-05, "epoch": 0.12917115177610333, "percentage": 4.3, "elapsed_time": "0:01:21", "remaining_time": "0:30:20", "throughput": 2735.61, "total_tokens": 223840}
+{"current_steps": 70, "total_steps": 1395, "loss": 0.1333, "lr": 2.981926730794264e-05, "epoch": 0.15069967707212056, "percentage": 5.02, "elapsed_time": "0:01:33", "remaining_time": "0:29:38", "throughput": 2777.09, "total_tokens": 260960}
+{"current_steps": 80, "total_steps": 1395, "loss": 0.0925, "lr": 2.9763233127086394e-05, "epoch": 0.1722282023681378, "percentage": 5.73, "elapsed_time": "0:01:46", "remaining_time": "0:29:05", "throughput": 2807.44, "total_tokens": 298048}
+{"current_steps": 90, "total_steps": 1395, "loss": 0.0886, "lr": 2.9699711832108555e-05, "epoch": 0.193756727664155, "percentage": 6.45, "elapsed_time": "0:01:58", "remaining_time": "0:28:31", "throughput": 2834.45, "total_tokens": 334496}
+{"current_steps": 100, "total_steps": 1395, "loss": 0.1127, "lr": 2.9628735637577612e-05, "epoch": 0.21528525296017223, "percentage": 7.17, "elapsed_time": "0:02:10", "remaining_time": "0:28:04", "throughput": 2857.99, "total_tokens": 371712}
+{"current_steps": 110, "total_steps": 1395, "loss": 0.1018, "lr": 2.955034053878428e-05, "epoch": 0.23681377825618946, "percentage": 7.89, "elapsed_time": "0:02:22", "remaining_time": "0:27:38", "throughput": 2876.81, "total_tokens": 408512}
+{"current_steps": 120, "total_steps": 1395, "loss": 0.1001, "lr": 2.9464566293486653e-05, "epoch": 0.25834230355220666, "percentage": 8.6, "elapsed_time": "0:02:34", "remaining_time": "0:27:17", "throughput": 2893.53, "total_tokens": 446048}
+{"current_steps": 130, "total_steps": 1395, "loss": 0.0962, "lr": 2.9371456401747193e-05, "epoch": 0.2798708288482239, "percentage": 9.32, "elapsed_time": "0:02:46", "remaining_time": "0:26:58", "throughput": 2908.32, "total_tokens": 483616}
+{"current_steps": 140, "total_steps": 1395, "loss": 0.1076, "lr": 2.9271058083871864e-05, "epoch": 0.3013993541442411, "percentage": 10.04, "elapsed_time": "0:02:58", "remaining_time": "0:26:38", "throughput": 2919.87, "total_tokens": 520768}
+{"current_steps": 150, "total_steps": 1395, "loss": 0.0886, "lr": 2.916342225646254e-05, "epoch": 0.32292787944025836, "percentage": 10.75, "elapsed_time": "0:03:10", "remaining_time": "0:26:22", "throughput": 2928.43, "total_tokens": 558240}
+{"current_steps": 160, "total_steps": 1395, "loss": 0.0861, "lr": 2.904860350659487e-05, "epoch": 0.3444564047362756, "percentage": 11.47, "elapsed_time": "0:03:22", "remaining_time": "0:26:05", "throughput": 2937.16, "total_tokens": 595648}
+{"current_steps": 170, "total_steps": 1395, "loss": 0.1158, "lr": 2.8926660064134637e-05, "epoch": 0.36598493003229277, "percentage": 12.19, "elapsed_time": "0:03:34", "remaining_time": "0:25:48", "throughput": 2942.81, "total_tokens": 632352}
+{"current_steps": 180, "total_steps": 1395, "loss": 0.0931, "lr": 2.8797653772206726e-05, "epoch": 0.38751345532831, "percentage": 12.9, "elapsed_time": "0:03:46", "remaining_time": "0:25:31", "throughput": 2950.21, "total_tokens": 669440}
+{"current_steps": 190, "total_steps": 1395, "loss": 0.0964, "lr": 2.8661650055831645e-05, "epoch": 0.40904198062432723, "percentage": 13.62, "elapsed_time": "0:03:59", "remaining_time": "0:25:15", "throughput": 2957.11, "total_tokens": 706784}
+{"current_steps": 200, "total_steps": 1395, "loss": 0.0708, "lr": 2.8518717888745486e-05, "epoch": 0.43057050592034446, "percentage": 14.34, "elapsed_time": "0:04:11", "remaining_time": "0:25:00", "throughput": 2963.94, "total_tokens": 744480}
+{"current_steps": 210, "total_steps": 1395, "loss": 0.0954, "lr": 2.836892975842015e-05, "epoch": 0.4520990312163617, "percentage": 15.05, "elapsed_time": "0:04:23", "remaining_time": "0:24:45", "throughput": 2968.41, "total_tokens": 781632}
+{"current_steps": 220, "total_steps": 1395, "loss": 0.0897, "lr": 2.8212361629301655e-05, "epoch": 0.4736275565123789, "percentage": 15.77, "elapsed_time": "0:04:35", "remaining_time": "0:24:31", "throughput": 2973.98, "total_tokens": 819264}
+{"current_steps": 230, "total_steps": 1395, "loss": 0.0783, "lr": 2.8049092904285033e-05, "epoch": 0.4951560818083961, "percentage": 16.49, "elapsed_time": "0:04:47", "remaining_time": "0:24:16", "throughput": 2977.07, "total_tokens": 855936}
+{"current_steps": 240, "total_steps": 1395, "loss": 0.092, "lr": 2.7879206384445526e-05, "epoch": 0.5166846071044133, "percentage": 17.2, "elapsed_time": "0:04:59", "remaining_time": "0:24:02", "throughput": 2981.87, "total_tokens": 893536}
+{"current_steps": 250, "total_steps": 1395, "loss": 0.1014, "lr": 2.7702788227046286e-05, "epoch": 0.5382131324004306, "percentage": 17.92, "elapsed_time": "0:05:11", "remaining_time": "0:23:47", "throughput": 2984.96, "total_tokens": 930272}
+{"current_steps": 260, "total_steps": 1395, "loss": 0.1032, "lr": 2.751992790184413e-05, "epoch": 0.5597416576964478, "percentage": 18.64, "elapsed_time": "0:05:24", "remaining_time": "0:23:34", "throughput": 2989.88, "total_tokens": 968768}
+{"current_steps": 270, "total_steps": 1395, "loss": 0.1016, "lr": 2.733071814571527e-05, "epoch": 0.581270182992465, "percentage": 19.35, "elapsed_time": "0:05:36", "remaining_time": "0:23:20", "throughput": 2992.79, "total_tokens": 1005888}
+{"current_steps": 280, "total_steps": 1395, "loss": 0.0843, "lr": 2.7135254915624213e-05, "epoch": 0.6027987082884823, "percentage": 20.07, "elapsed_time": "0:05:48", "remaining_time": "0:23:06", "throughput": 2996.74, "total_tokens": 1043744}
+{"current_steps": 290, "total_steps": 1395, "loss": 0.1012, "lr": 2.693363733995954e-05, "epoch": 0.6243272335844995, "percentage": 20.79, "elapsed_time": "0:06:00", "remaining_time": "0:22:54", "throughput": 3000.16, "total_tokens": 1081920}
+{"current_steps": 300, "total_steps": 1395, "loss": 0.0835, "lr": 2.672596766826138e-05, "epoch": 0.6458557588805167, "percentage": 21.51, "elapsed_time": "0:06:12", "remaining_time": "0:22:40", "throughput": 3001.47, "total_tokens": 1118784}
+{"current_steps": 310, "total_steps": 1395, "loss": 0.0884, "lr": 2.6512351219365927e-05, "epoch": 0.667384284176534, "percentage": 22.22, "elapsed_time": "0:06:24", "remaining_time": "0:22:26", "throughput": 3002.24, "total_tokens": 1155392}
+{"current_steps": 320, "total_steps": 1395, "loss": 0.0907, "lr": 2.6292896327993467e-05, "epoch": 0.6889128094725512, "percentage": 22.94, "elapsed_time": "0:06:37", "remaining_time": "0:22:13", "throughput": 3004.35, "total_tokens": 1192768}
+{"current_steps": 330, "total_steps": 1395, "loss": 0.0786, "lr": 2.606771428980681e-05, "epoch": 0.7104413347685683, "percentage": 23.66, "elapsed_time": "0:06:49", "remaining_time": "0:22:00", "throughput": 3006.91, "total_tokens": 1230464}
+{"current_steps": 340, "total_steps": 1395, "loss": 0.0863, "lr": 2.5836919304968184e-05, "epoch": 0.7319698600645855, "percentage": 24.37, "elapsed_time": "0:07:01", "remaining_time": "0:21:47", "throughput": 3008.52, "total_tokens": 1267520}
+{"current_steps": 350, "total_steps": 1395, "loss": 0.0976, "lr": 2.5600628420223057e-05, "epoch": 0.7534983853606028, "percentage": 25.09, "elapsed_time": "0:07:13", "remaining_time": "0:21:35", "throughput": 3012.05, "total_tokens": 1306592}
+{"current_steps": 360, "total_steps": 1395, "loss": 0.0914, "lr": 2.535896146954037e-05, "epoch": 0.77502691065662, "percentage": 25.81, "elapsed_time": "0:07:25", "remaining_time": "0:21:21", "throughput": 3013.82, "total_tokens": 1343840}
+{"current_steps": 370, "total_steps": 1395, "loss": 0.0997, "lr": 2.5112041013339212e-05, "epoch": 0.7965554359526372, "percentage": 26.52, "elapsed_time": "0:07:38", "remaining_time": "0:21:09", "throughput": 3016.75, "total_tokens": 1382816}
+{"current_steps": 380, "total_steps": 1395, "loss": 0.0689, "lr": 2.48599922763328e-05, "epoch": 0.8180839612486545, "percentage": 27.24, "elapsed_time": "0:07:50", "remaining_time": "0:20:56", "throughput": 3019.08, "total_tokens": 1420352}
+{"current_steps": 390, "total_steps": 1395, "loss": 0.1062, "lr": 2.4602943084021286e-05, "epoch": 0.8396124865446717, "percentage": 27.96, "elapsed_time": "0:08:02", "remaining_time": "0:20:43", "throughput": 3020.05, "total_tokens": 1457280}
+{"current_steps": 400, "total_steps": 1395, "loss": 0.0579, "lr": 2.434102379786553e-05, "epoch": 0.8611410118406889, "percentage": 28.67, "elapsed_time": "0:08:14", "remaining_time": "0:20:30", "throughput": 3020.61, "total_tokens": 1493984}
+{"current_steps": 410, "total_steps": 1395, "loss": 0.074, "lr": 2.4074367249174864e-05, "epoch": 0.8826695371367062, "percentage": 29.39, "elapsed_time": "0:08:26", "remaining_time": "0:20:17", "throughput": 3022.13, "total_tokens": 1531392}
+{"current_steps": 420, "total_steps": 1395, "loss": 0.0774, "lr": 2.380310867174219e-05, "epoch": 0.9041980624327234, "percentage": 30.11, "elapsed_time": "0:08:38", "remaining_time": "0:20:04", "throughput": 3023.51, "total_tokens": 1568992}
+{"current_steps": 430, "total_steps": 1395, "loss": 0.0771, "lr": 2.3527385633260724e-05, "epoch": 0.9257265877287406, "percentage": 30.82, "elapsed_time": "0:08:50", "remaining_time": "0:19:51", "throughput": 3025.2, "total_tokens": 1606144}
+{"current_steps": 440, "total_steps": 1395, "loss": 0.075, "lr": 2.3247337965557097e-05, "epoch": 0.9472551130247578, "percentage": 31.54, "elapsed_time": "0:09:02", "remaining_time": "0:19:37", "throughput": 3025.01, "total_tokens": 1641696}
+{"current_steps": 450, "total_steps": 1395, "loss": 0.0785, "lr": 2.2963107693676192e-05, "epoch": 0.9687836383207751, "percentage": 32.26, "elapsed_time": "0:09:14", "remaining_time": "0:19:25", "throughput": 3025.73, "total_tokens": 1678816}
+{"current_steps": 460, "total_steps": 1395, "loss": 0.074, "lr": 2.2674838963853738e-05, "epoch": 0.9903121636167922, "percentage": 32.97, "elapsed_time": "0:09:26", "remaining_time": "0:19:12", "throughput": 3026.88, "total_tokens": 1715808}
+{"current_steps": 470, "total_steps": 1395, "loss": 0.0653, "lr": 2.2382677970413097e-05, "epoch": 1.0107642626480087, "percentage": 33.69, "elapsed_time": "0:09:38", "remaining_time": "0:18:58", "throughput": 3028.66, "total_tokens": 1751552}
+{"current_steps": 480, "total_steps": 1395, "loss": 0.0632, "lr": 2.2086772881623455e-05, "epoch": 1.0322927879440258, "percentage": 34.41, "elapsed_time": "0:09:50", "remaining_time": "0:18:45", "throughput": 3029.41, "total_tokens": 1788544}
+{"current_steps": 490, "total_steps": 1395, "loss": 0.0472, "lr": 2.1787273764556827e-05, "epoch": 1.0538213132400431, "percentage": 35.13, "elapsed_time": "0:10:02", "remaining_time": "0:18:33", "throughput": 3030.79, "total_tokens": 1826496}
+{"current_steps": 500, "total_steps": 1395, "loss": 0.0534, "lr": 2.1484332508982167e-05, "epoch": 1.0753498385360603, "percentage": 35.84, "elapsed_time": "0:10:14", "remaining_time": "0:18:20", "throughput": 3032.11, "total_tokens": 1864544}
+{"current_steps": 510, "total_steps": 1395, "loss": 0.0751, "lr": 2.11781027503351e-05, "epoch": 1.0968783638320776, "percentage": 36.56, "elapsed_time": "0:10:27", "remaining_time": "0:18:08", "throughput": 3032.88, "total_tokens": 1901856}
+{"current_steps": 520, "total_steps": 1395, "loss": 0.0747, "lr": 2.0868739791802312e-05, "epoch": 1.1184068891280947, "percentage": 37.28, "elapsed_time": "0:10:39", "remaining_time": "0:17:55", "throughput": 3033.66, "total_tokens": 1939264}
+{"current_steps": 530, "total_steps": 1395, "loss": 0.0492, "lr": 2.0556400525560168e-05, "epoch": 1.1399354144241118, "percentage": 37.99, "elapsed_time": "0:10:51", "remaining_time": "0:17:42", "throughput": 3034.22, "total_tokens": 1975872}
+{"current_steps": 540, "total_steps": 1395, "loss": 0.0651, "lr": 2.0241243353207484e-05, "epoch": 1.1614639397201292, "percentage": 38.71, "elapsed_time": "0:11:03", "remaining_time": "0:17:30", "throughput": 3034.64, "total_tokens": 2012928}
+{"current_steps": 550, "total_steps": 1395, "loss": 0.0565, "lr": 1.9923428105432796e-05, "epoch": 1.1829924650161463, "percentage": 39.43, "elapsed_time": "0:11:15", "remaining_time": "0:17:18", "throughput": 3035.81, "total_tokens": 2051168}
+{"current_steps": 560, "total_steps": 1395, "loss": 0.0696, "lr": 1.9603115960956852e-05, "epoch": 1.2045209903121636, "percentage": 40.14, "elapsed_time": "0:11:27", "remaining_time": "0:17:05", "throughput": 3036.73, "total_tokens": 2088576}
+{"current_steps": 570, "total_steps": 1395, "loss": 0.0553, "lr": 1.9280469364791505e-05, "epoch": 1.2260495156081808, "percentage": 40.86, "elapsed_time": "0:11:39", "remaining_time": "0:16:52", "throughput": 3037.22, "total_tokens": 2125216}
+{"current_steps": 580, "total_steps": 1395, "loss": 0.0666, "lr": 1.8955651945856354e-05, "epoch": 1.247578040904198, "percentage": 41.58, "elapsed_time": "0:11:51", "remaining_time": "0:16:40", "throughput": 3037.97, "total_tokens": 2162816}
+{"current_steps": 590, "total_steps": 1395, "loss": 0.0449, "lr": 1.8628828433995013e-05, "epoch": 1.2691065662002152, "percentage": 42.29, "elapsed_time": "0:12:04", "remaining_time": "0:16:27", "throughput": 3038.24, "total_tokens": 2199712}
+{"current_steps": 600, "total_steps": 1395, "loss": 0.0522, "lr": 1.8300164576433035e-05, "epoch": 1.2906350914962326, "percentage": 43.01, "elapsed_time": "0:12:16", "remaining_time": "0:16:15", "throughput": 3038.41, "total_tokens": 2236416}
+{"current_steps": 610, "total_steps": 1395, "loss": 0.0638, "lr": 1.7969827053719862e-05, "epoch": 1.3121636167922497, "percentage": 43.73, "elapsed_time": "0:12:28", "remaining_time": "0:16:02", "throughput": 3039.22, "total_tokens": 2274272}
+{"current_steps": 620, "total_steps": 1395, "loss": 0.0683, "lr": 1.7637983395197465e-05, "epoch": 1.333692142088267, "percentage": 44.44, "elapsed_time": "0:12:40", "remaining_time": "0:15:50", "throughput": 3039.15, "total_tokens": 2310688}
+{"current_steps": 630, "total_steps": 1395, "loss": 0.0678, "lr": 1.7304801894038522e-05, "epoch": 1.3552206673842842, "percentage": 45.16, "elapsed_time": "0:12:52", "remaining_time": "0:15:38", "throughput": 3039.88, "total_tokens": 2348608}
+{"current_steps": 640, "total_steps": 1395, "loss": 0.0502, "lr": 1.6970451521897198e-05, "epoch": 1.3767491926803013, "percentage": 45.88, "elapsed_time": "0:13:04", "remaining_time": "0:15:25", "throughput": 3040.29, "total_tokens": 2386144}
+{"current_steps": 650, "total_steps": 1395, "loss": 0.0728, "lr": 1.663510184321587e-05, "epoch": 1.3982777179763186, "percentage": 46.59, "elapsed_time": "0:13:16", "remaining_time": "0:15:13", "throughput": 3040.28, "total_tokens": 2422912}
+{"current_steps": 660, "total_steps": 1395, "loss": 0.059, "lr": 1.6298922929231194e-05, "epoch": 1.419806243272336, "percentage": 47.31, "elapsed_time": "0:13:29", "remaining_time": "0:15:01", "throughput": 3040.16, "total_tokens": 2459744}
+{"current_steps": 670, "total_steps": 1395, "loss": 0.0514, "lr": 1.5962085271723158e-05, "epoch": 1.441334768568353, "percentage": 48.03, "elapsed_time": "0:13:41", "remaining_time": "0:14:48", "throughput": 3040.77, "total_tokens": 2497184}
+{"current_steps": 680, "total_steps": 1395, "loss": 0.0689, "lr": 1.5624759696550874e-05, "epoch": 1.4628632938643702, "percentage": 48.75, "elapsed_time": "0:13:53", "remaining_time": "0:14:36", "throughput": 3041.47, "total_tokens": 2535456}
+{"current_steps": 690, "total_steps": 1395, "loss": 0.0338, "lr": 1.528711727701893e-05, "epoch": 1.4843918191603875, "percentage": 49.46, "elapsed_time": "0:14:05", "remaining_time": "0:14:24", "throughput": 3041.91, "total_tokens": 2572704}
+{"current_steps": 700, "total_steps": 1395, "loss": 0.0621, "lr": 1.4949329247118258e-05, "epoch": 1.5059203444564049, "percentage": 50.18, "elapsed_time": "0:14:17", "remaining_time": "0:14:11", "throughput": 3042.48, "total_tokens": 2609696}
+{"current_steps": 710, "total_steps": 1395, "loss": 0.057, "lr": 1.4611566914685489e-05, "epoch": 1.527448869752422, "percentage": 50.9, "elapsed_time": "0:14:29", "remaining_time": "0:13:59", "throughput": 3042.86, "total_tokens": 2646976}
+{"current_steps": 720, "total_steps": 1395, "loss": 0.0632, "lr": 1.4274001574524894e-05, "epoch": 1.5489773950484391, "percentage": 51.61, "elapsed_time": "0:14:42", "remaining_time": "0:13:46", "throughput": 3043.37, "total_tokens": 2684576}
+{"current_steps": 730, "total_steps": 1395, "loss": 0.0361, "lr": 1.3936804421536906e-05, "epoch": 1.5705059203444565, "percentage": 52.33, "elapsed_time": "0:14:54", "remaining_time": "0:13:34", "throughput": 3041.59, "total_tokens": 2720736}
+{"current_steps": 740, "total_steps": 1395, "loss": 0.0503, "lr": 1.3600146463897327e-05, "epoch": 1.5920344456404736, "percentage": 53.05, "elapsed_time": "0:15:06", "remaining_time": "0:13:22", "throughput": 3041.96, "total_tokens": 2757728}
+{"current_steps": 750, "total_steps": 1395, "loss": 0.0843, "lr": 1.3264198436331193e-05, "epoch": 1.6135629709364907, "percentage": 53.76, "elapsed_time": "0:15:18", "remaining_time": "0:13:10", "throughput": 3042.71, "total_tokens": 2795168}
+{"current_steps": 760, "total_steps": 1395, "loss": 0.05, "lr": 1.2929130713525396e-05, "epoch": 1.635091496232508, "percentage": 54.48, "elapsed_time": "0:15:30", "remaining_time": "0:12:57", "throughput": 3042.97, "total_tokens": 2832000}
+{"current_steps": 770, "total_steps": 1395, "loss": 0.0608, "lr": 1.2595113223723788e-05, "epoch": 1.6566200215285254, "percentage": 55.2, "elapsed_time": "0:15:42", "remaining_time": "0:12:45", "throughput": 3043.39, "total_tokens": 2869696}
+{"current_steps": 780, "total_steps": 1395, "loss": 0.0422, "lr": 1.2262315362548825e-05, "epoch": 1.6781485468245425, "percentage": 55.91, "elapsed_time": "0:15:55", "remaining_time": "0:12:33", "throughput": 3044.09, "total_tokens": 2907840}
+{"current_steps": 790, "total_steps": 1395, "loss": 0.0558, "lr": 1.1930905907093196e-05, "epoch": 1.6996770721205596, "percentage": 56.63, "elapsed_time": "0:16:07", "remaining_time": "0:12:20", "throughput": 3043.47, "total_tokens": 2943232}
+{"current_steps": 800, "total_steps": 1395, "loss": 0.0509, "lr": 1.1601052930325236e-05, "epoch": 1.721205597416577, "percentage": 57.35, "elapsed_time": "0:16:19", "remaining_time": "0:12:08", "throughput": 3043.89, "total_tokens": 2980608}
+{"current_steps": 810, "total_steps": 1395, "loss": 0.0674, "lr": 1.127292371585141e-05, "epoch": 1.7427341227125943, "percentage": 58.06, "elapsed_time": "0:16:31", "remaining_time": "0:11:55", "throughput": 3044.5, "total_tokens": 3018080}
+{"current_steps": 820, "total_steps": 1395, "loss": 0.0694, "lr": 1.0946684673079074e-05, "epoch": 1.7642626480086114, "percentage": 58.78, "elapsed_time": "0:16:43", "remaining_time": "0:11:43", "throughput": 3044.95, "total_tokens": 3055520}
+{"current_steps": 830, "total_steps": 1395, "loss": 0.0383, "lr": 1.0622501252822628e-05, "epoch": 1.7857911733046286, "percentage": 59.5, "elapsed_time": "0:16:55", "remaining_time": "0:11:31", "throughput": 3045.52, "total_tokens": 3092800}
+{"current_steps": 840, "total_steps": 1395, "loss": 0.0547, "lr": 1.0300537863395823e-05, "epoch": 1.807319698600646, "percentage": 60.22, "elapsed_time": "0:17:07", "remaining_time": "0:11:19", "throughput": 3046.03, "total_tokens": 3130656}
+{"current_steps": 850, "total_steps": 1395, "loss": 0.0516, "lr": 9.98095778723271e-06, "epoch": 1.8288482238966632, "percentage": 60.93, "elapsed_time": "0:17:19", "remaining_time": "0:11:06", "throughput": 3046.52, "total_tokens": 3167744}
+{"current_steps": 860, "total_steps": 1395, "loss": 0.0628, "lr": 9.663923098079656e-06, "epoch": 1.8503767491926801, "percentage": 61.65, "elapsed_time": "0:17:31", "remaining_time": "0:10:54", "throughput": 3047.17, "total_tokens": 3205376}
+{"current_steps": 870, "total_steps": 1395, "loss": 0.0733, "lr": 9.349594578800259e-06, "epoch": 1.8719052744886975, "percentage": 62.37, "elapsed_time": "0:17:44", "remaining_time": "0:10:42", "throughput": 3047.51, "total_tokens": 3242976}
+{"current_steps": 880, "total_steps": 1395, "loss": 0.0727, "lr": 9.038131639834977e-06, "epoch": 1.8934337997847148, "percentage": 63.08, "elapsed_time": "0:17:56", "remaining_time": "0:10:29", "throughput": 3047.92, "total_tokens": 3280096}
+{"current_steps": 890, "total_steps": 1395, "loss": 0.058, "lr": 8.7296922383568e-06, "epoch": 1.914962325080732, "percentage": 63.8, "elapsed_time": "0:18:08", "remaining_time": "0:10:17", "throughput": 3048.21, "total_tokens": 3316960}
+{"current_steps": 900, "total_steps": 1395, "loss": 0.0588, "lr": 8.424432798163838e-06, "epoch": 1.936490850376749, "percentage": 64.52, "elapsed_time": "0:18:20", "remaining_time": "0:10:05", "throughput": 3048.41, "total_tokens": 3353600}
+{"current_steps": 910, "total_steps": 1395, "loss": 0.0375, "lr": 8.122508130349711e-06, "epoch": 1.9580193756727664, "percentage": 65.23, "elapsed_time": "0:18:32", "remaining_time": "0:09:52", "throughput": 3048.92, "total_tokens": 3390592}
+{"current_steps": 920, "total_steps": 1395, "loss": 0.0644, "lr": 7.82407135479164e-06, "epoch": 1.9795479009687837, "percentage": 65.95, "elapsed_time": "0:18:44", "remaining_time": "0:09:40", "throughput": 3049.05, "total_tokens": 3427840}
+{"current_steps": 930, "total_steps": 1395, "loss": 0.0347, "lr": 7.529273822496379e-06, "epoch": 2.0, "percentage": 66.67, "elapsed_time": "0:18:55", "remaining_time": "0:09:27", "throughput": 3049.71, "total_tokens": 3464032}
+{"current_steps": 940, "total_steps": 1395, "loss": 0.0491, "lr": 7.238265038843143e-06, "epoch": 2.0215285252960173, "percentage": 67.38, "elapsed_time": "0:19:07", "remaining_time": "0:09:15", "throughput": 3049.77, "total_tokens": 3501056}
+{"current_steps": 950, "total_steps": 1395, "loss": 0.0263, "lr": 6.951192587762555e-06, "epoch": 2.0430570505920342, "percentage": 68.1, "elapsed_time": "0:19:20", "remaining_time": "0:09:03", "throughput": 3049.91, "total_tokens": 3538464}
+{"current_steps": 960, "total_steps": 1395, "loss": 0.029, "lr": 6.668202056890052e-06, "epoch": 2.0645855758880516, "percentage": 68.82, "elapsed_time": "0:19:32", "remaining_time": "0:08:51", "throughput": 3050.03, "total_tokens": 3575552}
+{"current_steps": 970, "total_steps": 1395, "loss": 0.0338, "lr": 6.389436963731707e-06, "epoch": 2.086114101184069, "percentage": 69.53, "elapsed_time": "0:19:44", "remaining_time": "0:08:38", "throughput": 3050.09, "total_tokens": 3612736}
+{"current_steps": 980, "total_steps": 1395, "loss": 0.0427, "lr": 6.115038682879904e-06, "epoch": 2.1076426264800863, "percentage": 70.25, "elapsed_time": "0:19:56", "remaining_time": "0:08:26", "throughput": 3050.36, "total_tokens": 3650464}
+{"current_steps": 990, "total_steps": 1395, "loss": 0.0323, "lr": 5.845146374315781e-06, "epoch": 2.129171151776103, "percentage": 70.97, "elapsed_time": "0:20:08", "remaining_time": "0:08:14", "throughput": 3050.25, "total_tokens": 3687392}
+{"current_steps": 1000, "total_steps": 1395, "loss": 0.0504, "lr": 5.579896912834807e-06, "epoch": 2.1506996770721205, "percentage": 71.68, "elapsed_time": "0:20:21", "remaining_time": "0:08:02", "throughput": 3050.66, "total_tokens": 3725152}
+{"current_steps": 1010, "total_steps": 1395, "loss": 0.043, "lr": 5.319424818631293e-06, "epoch": 2.172228202368138, "percentage": 72.4, "elapsed_time": "0:20:34", "remaining_time": "0:07:50", "throughput": 3048.97, "total_tokens": 3762752}
+{"current_steps": 1020, "total_steps": 1395, "loss": 0.0377, "lr": 5.063862189076988e-06, "epoch": 2.193756727664155, "percentage": 73.12, "elapsed_time": "0:20:46", "remaining_time": "0:07:38", "throughput": 3049.38, "total_tokens": 3800736}
+{"current_steps": 1030, "total_steps": 1395, "loss": 0.0408, "lr": 4.813338631728471e-06, "epoch": 2.215285252960172, "percentage": 73.84, "elapsed_time": "0:20:58", "remaining_time": "0:07:26", "throughput": 3049.81, "total_tokens": 3838432}
+{"current_steps": 1040, "total_steps": 1395, "loss": 0.0328, "lr": 4.567981198597168e-06, "epoch": 2.2368137782561894, "percentage": 74.55, "elapsed_time": "0:21:10", "remaining_time": "0:07:13", "throughput": 3050.1, "total_tokens": 3876224}
+{"current_steps": 1050, "total_steps": 1395, "loss": 0.0259, "lr": 4.327914321715459e-06, "epoch": 2.2583423035522068, "percentage": 75.27, "elapsed_time": "0:21:22", "remaining_time": "0:07:01", "throughput": 3050.4, "total_tokens": 3913216}
+{"current_steps": 1060, "total_steps": 1395, "loss": 0.027, "lr": 4.09325975003146e-06, "epoch": 2.2798708288482237, "percentage": 75.99, "elapsed_time": "0:21:34", "remaining_time": "0:06:49", "throughput": 3050.52, "total_tokens": 3950208}
+{"current_steps": 1070, "total_steps": 1395, "loss": 0.0347, "lr": 3.864136487664538e-06, "epoch": 2.301399354144241, "percentage": 76.7, "elapsed_time": "0:21:46", "remaining_time": "0:06:36", "throughput": 3050.36, "total_tokens": 3986752}
+{"current_steps": 1080, "total_steps": 1395, "loss": 0.0189, "lr": 3.640660733552863e-06, "epoch": 2.3229278794402584, "percentage": 77.42, "elapsed_time": "0:21:58", "remaining_time": "0:06:24", "throughput": 3050.42, "total_tokens": 4023360}
+{"current_steps": 1090, "total_steps": 1395, "loss": 0.0186, "lr": 3.4229458225235696e-06, "epoch": 2.3444564047362757, "percentage": 78.14, "elapsed_time": "0:22:11", "remaining_time": "0:06:12", "throughput": 3050.89, "total_tokens": 4061856}
+{"current_steps": 1100, "total_steps": 1395, "loss": 0.0366, "lr": 3.211102167815469e-06, "epoch": 2.3659849300322926, "percentage": 78.85, "elapsed_time": "0:22:23", "remaining_time": "0:06:00", "throughput": 3050.95, "total_tokens": 4098464}
+{"current_steps": 1110, "total_steps": 1395, "loss": 0.0333, "lr": 3.0052372050834025e-06, "epoch": 2.38751345532831, "percentage": 79.57, "elapsed_time": "0:22:35", "remaining_time": "0:05:48", "throughput": 3051.25, "total_tokens": 4136576}
+{"current_steps": 1120, "total_steps": 1395, "loss": 0.0259, "lr": 2.805455337912722e-06, "epoch": 2.4090419806243273, "percentage": 80.29, "elapsed_time": "0:22:47", "remaining_time": "0:05:35", "throughput": 3051.82, "total_tokens": 4174528}
+{"current_steps": 1130, "total_steps": 1395, "loss": 0.0377, "lr": 2.611857884871379e-06, "epoch": 2.4305705059203446, "percentage": 81.0, "elapsed_time": "0:23:00", "remaining_time": "0:05:23", "throughput": 3051.95, "total_tokens": 4211712}
+{"current_steps": 1140, "total_steps": 1395, "loss": 0.0457, "lr": 2.424543028126661e-06, "epoch": 2.4520990312163615, "percentage": 81.72, "elapsed_time": "0:23:12", "remaining_time": "0:05:11", "throughput": 3052.0, "total_tokens": 4248800}
+{"current_steps": 1150, "total_steps": 1395, "loss": 0.0341, "lr": 2.243605763652489e-06, "epoch": 2.473627556512379, "percentage": 82.44, "elapsed_time": "0:23:24", "remaining_time": "0:04:59", "throughput": 3052.3, "total_tokens": 4286496}
+{"current_steps": 1160, "total_steps": 1395, "loss": 0.0311, "lr": 2.0691378530525933e-06, "epoch": 2.495156081808396, "percentage": 83.15, "elapsed_time": "0:23:36", "remaining_time": "0:04:46", "throughput": 3052.38, "total_tokens": 4323328}
+{"current_steps": 1170, "total_steps": 1395, "loss": 0.0373, "lr": 1.9012277770239829e-06, "epoch": 2.5166846071044136, "percentage": 83.87, "elapsed_time": "0:23:48", "remaining_time": "0:04:34", "throughput": 3052.59, "total_tokens": 4360384}
+{"current_steps": 1180, "total_steps": 1395, "loss": 0.0176, "lr": 1.7399606904843224e-06, "epoch": 2.5382131324004304, "percentage": 84.59, "elapsed_time": "0:24:00", "remaining_time": "0:04:22", "throughput": 3052.55, "total_tokens": 4397440}
+{"current_steps": 1190, "total_steps": 1395, "loss": 0.0264, "lr": 1.5854183793859577e-06, "epoch": 2.559741657696448, "percentage": 85.3, "elapsed_time": "0:24:12", "remaining_time": "0:04:10", "throughput": 3052.77, "total_tokens": 4435584}
+{"current_steps": 1200, "total_steps": 1395, "loss": 0.0211, "lr": 1.437679219238498e-06, "epoch": 2.581270182992465, "percentage": 86.02, "elapsed_time": "0:24:25", "remaining_time": "0:03:58", "throughput": 3053.0, "total_tokens": 4472800}
+{"current_steps": 1210, "total_steps": 1395, "loss": 0.0264, "lr": 1.2968181353609854e-06, "epoch": 2.602798708288482, "percentage": 86.74, "elapsed_time": "0:24:37", "remaining_time": "0:03:45", "throughput": 3053.33, "total_tokens": 4510464}
+{"current_steps": 1220, "total_steps": 1395, "loss": 0.0369, "lr": 1.1629065648838495e-06, "epoch": 2.6243272335844994, "percentage": 87.46, "elapsed_time": "0:24:49", "remaining_time": "0:03:33", "throughput": 3052.94, "total_tokens": 4546592}
+{"current_steps": 1230, "total_steps": 1395, "loss": 0.0285, "lr": 1.0360124205198102e-06, "epoch": 2.6458557588805167, "percentage": 88.17, "elapsed_time": "0:25:01", "remaining_time": "0:03:21", "throughput": 3053.06, "total_tokens": 4583680}
+{"current_steps": 1240, "total_steps": 1395, "loss": 0.0452, "lr": 9.162000561222672e-07, "epoch": 2.667384284176534, "percentage": 88.89, "elapsed_time": "0:25:13", "remaining_time": "0:03:09", "throughput": 3053.46, "total_tokens": 4622080}
+{"current_steps": 1250, "total_steps": 1395, "loss": 0.051, "lr": 8.035302340484646e-07, "epoch": 2.6889128094725514, "percentage": 89.61, "elapsed_time": "0:25:25", "remaining_time": "0:02:56", "throughput": 3053.44, "total_tokens": 4658848}
+{"current_steps": 1260, "total_steps": 1395, "loss": 0.0217, "lr": 6.980600943441273e-07, "epoch": 2.7104413347685683, "percentage": 90.32, "elapsed_time": "0:25:37", "remaining_time": "0:02:44", "throughput": 3053.64, "total_tokens": 4695968}
+{"current_steps": 1270, "total_steps": 1395, "loss": 0.0414, "lr": 5.998431257651132e-07, "epoch": 2.7319698600645856, "percentage": 91.04, "elapsed_time": "0:25:50", "remaining_time": "0:02:32", "throughput": 3053.69, "total_tokens": 4733344}
+{"current_steps": 1280, "total_steps": 1395, "loss": 0.0315, "lr": 5.089291386508055e-07, "epoch": 2.7534983853606025, "percentage": 91.76, "elapsed_time": "0:26:02", "remaining_time": "0:02:20", "throughput": 3053.89, "total_tokens": 4770496}
+{"current_steps": 1290, "total_steps": 1395, "loss": 0.0336, "lr": 4.2536423966299256e-07, "epoch": 2.77502691065662, "percentage": 92.47, "elapsed_time": "0:26:14", "remaining_time": "0:02:08", "throughput": 3053.79, "total_tokens": 4807456}
+{"current_steps": 1300, "total_steps": 1395, "loss": 0.0426, "lr": 3.491908084030776e-07, "epoch": 2.7965554359526372, "percentage": 93.19, "elapsed_time": "0:26:26", "remaining_time": "0:01:55", "throughput": 3053.9, "total_tokens": 4845440}
+{"current_steps": 1310, "total_steps": 1395, "loss": 0.0434, "lr": 2.804474759194176e-07, "epoch": 2.8180839612486546, "percentage": 93.91, "elapsed_time": "0:26:38", "remaining_time": "0:01:43", "throughput": 3054.19, "total_tokens": 4883392}
+{"current_steps": 1320, "total_steps": 1395, "loss": 0.0297, "lr": 2.1916910511575627e-07, "epoch": 2.839612486544672, "percentage": 94.62, "elapsed_time": "0:26:50", "remaining_time": "0:01:31", "throughput": 3054.36, "total_tokens": 4920384}
+{"current_steps": 1330, "total_steps": 1395, "loss": 0.036, "lr": 1.6538677307062933e-07, "epoch": 2.861141011840689, "percentage": 95.34, "elapsed_time": "0:27:03", "remaining_time": "0:01:19", "throughput": 3054.48, "total_tokens": 4957888}
+{"current_steps": 1340, "total_steps": 1395, "loss": 0.0209, "lr": 1.1912775527676278e-07, "epoch": 2.882669537136706, "percentage": 96.06, "elapsed_time": "0:27:15", "remaining_time": "0:01:07", "throughput": 3054.75, "total_tokens": 4994944}
+{"current_steps": 1350, "total_steps": 1395, "loss": 0.0467, "lr": 8.04155118084099e-08, "epoch": 2.9041980624327235, "percentage": 96.77, "elapsed_time": "0:27:27", "remaining_time": "0:00:54", "throughput": 3054.83, "total_tokens": 5031584}
+{"current_steps": 1360, "total_steps": 1395, "loss": 0.0414, "lr": 4.9269675423671556e-08, "epoch": 2.9257265877287404, "percentage": 97.49, "elapsed_time": "0:27:39", "remaining_time": "0:00:42", "throughput": 3054.95, "total_tokens": 5068928}
+{"current_steps": 1370, "total_steps": 1395, "loss": 0.0182, "lr": 2.570604160781753e-08, "epoch": 2.9472551130247577, "percentage": 98.21, "elapsed_time": "0:27:51", "remaining_time": "0:00:30", "throughput": 3054.99, "total_tokens": 5105984}
+{"current_steps": 1380, "total_steps": 1395, "loss": 0.0236, "lr": 9.736560562675912e-09, "epoch": 2.968783638320775, "percentage": 98.92, "elapsed_time": "0:28:03", "remaining_time": "0:00:18", "throughput": 3055.07, "total_tokens": 5143296}
+{"current_steps": 1390, "total_steps": 1395, "loss": 0.0323, "lr": 1.3693311461326908e-09, "epoch": 2.9903121636167924, "percentage": 99.64, "elapsed_time": "0:28:15", "remaining_time": "0:00:06", "throughput": 3055.34, "total_tokens": 5181056}
+{"current_steps": 1395, "total_steps": 1395, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "0:28:22", "remaining_time": "0:00:00", "throughput": 3053.89, "total_tokens": 5197824}

trainer_state.json

@@ -0,0 +1,1434 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1395,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.021528525296017224,
+      "grad_norm": 3.766150712966919,
+      "learning_rate": 2.9996919063517692e-05,
+      "loss": 0.6799,
+      "num_input_tokens_seen": 37152,
+      "step": 10,
+      "train_runtime": 20.683,
+      "train_tokens_per_second": 1796.26
+    },
+    {
+      "epoch": 0.04305705059203445,
+      "grad_norm": 3.9944393634796143,
+      "learning_rate": 2.9986270537572827e-05,
+      "loss": 0.1361,
+      "num_input_tokens_seen": 73856,
+      "step": 20,
+      "train_runtime": 32.7469,
+      "train_tokens_per_second": 2255.36
+    },
+    {
+      "epoch": 0.06458557588805167,
+      "grad_norm": 10.329977989196777,
+      "learning_rate": 2.9968021784983436e-05,
+      "loss": 0.1558,
+      "num_input_tokens_seen": 111232,
+      "step": 30,
+      "train_runtime": 45.4717,
+      "train_tokens_per_second": 2446.179
+    },
+    {
+      "epoch": 0.0861141011840689,
+      "grad_norm": 4.059694766998291,
+      "learning_rate": 2.994218206053077e-05,
+      "loss": 0.1344,
+      "num_input_tokens_seen": 148992,
+      "step": 40,
+      "train_runtime": 57.6575,
+      "train_tokens_per_second": 2584.086
+    },
+    {
+      "epoch": 0.10764262648008611,
+      "grad_norm": 2.61093807220459,
+      "learning_rate": 2.9908764468726846e-05,
+      "loss": 0.1262,
+      "num_input_tokens_seen": 186176,
+      "step": 50,
+      "train_runtime": 69.6848,
+      "train_tokens_per_second": 2671.688
+    },
+    {
+      "epoch": 0.12917115177610333,
+      "grad_norm": 4.661616325378418,
+      "learning_rate": 2.986778595716853e-05,
+      "loss": 0.1182,
+      "num_input_tokens_seen": 223840,
+      "step": 60,
+      "train_runtime": 81.8303,
+      "train_tokens_per_second": 2735.416
+    },
+    {
+      "epoch": 0.15069967707212056,
+      "grad_norm": 2.2738418579101562,
+      "learning_rate": 2.981926730794264e-05,
+      "loss": 0.1333,
+      "num_input_tokens_seen": 260960,
+      "step": 70,
+      "train_runtime": 93.9745,
+      "train_tokens_per_second": 2776.923
+    },
+    {
+      "epoch": 0.1722282023681378,
+      "grad_norm": 2.6716361045837402,
+      "learning_rate": 2.9763233127086394e-05,
+      "loss": 0.0925,
+      "num_input_tokens_seen": 298048,
+      "step": 80,
+      "train_runtime": 106.1694,
+      "train_tokens_per_second": 2807.287
+    },
+    {
+      "epoch": 0.193756727664155,
+      "grad_norm": 2.849599599838257,
+      "learning_rate": 2.9699711832108555e-05,
+      "loss": 0.0886,
+      "num_input_tokens_seen": 334496,
+      "step": 90,
+      "train_runtime": 118.0166,
+      "train_tokens_per_second": 2834.313
+    },
+    {
+      "epoch": 0.21528525296017223,
+      "grad_norm": 1.1690081357955933,
+      "learning_rate": 2.9628735637577612e-05,
+      "loss": 0.1127,
+      "num_input_tokens_seen": 371712,
+      "step": 100,
+      "train_runtime": 130.0666,
+      "train_tokens_per_second": 2857.86
+    },
+    {
+      "epoch": 0.23681377825618946,
+      "grad_norm": 1.8079500198364258,
+      "learning_rate": 2.955034053878428e-05,
+      "loss": 0.1018,
+      "num_input_tokens_seen": 408512,
+      "step": 110,
+      "train_runtime": 142.0073,
+      "train_tokens_per_second": 2876.697
+    },
+    {
+      "epoch": 0.25834230355220666,
+      "grad_norm": 4.173007965087891,
+      "learning_rate": 2.9464566293486653e-05,
+      "loss": 0.1001,
+      "num_input_tokens_seen": 446048,
+      "step": 120,
+      "train_runtime": 154.1596,
+      "train_tokens_per_second": 2893.418
+    },
+    {
+      "epoch": 0.2798708288482239,
+      "grad_norm": 3.0781972408294678,
+      "learning_rate": 2.9371456401747193e-05,
+      "loss": 0.0962,
+      "num_input_tokens_seen": 483616,
+      "step": 130,
+      "train_runtime": 166.2926,
+      "train_tokens_per_second": 2908.222
+    },
+    {
+      "epoch": 0.3013993541442411,
+      "grad_norm": 3.6592869758605957,
+      "learning_rate": 2.9271058083871864e-05,
+      "loss": 0.1076,
+      "num_input_tokens_seen": 520768,
+      "step": 140,
+      "train_runtime": 178.3589,
+      "train_tokens_per_second": 2919.776
+    },
+    {
+      "epoch": 0.32292787944025836,
+      "grad_norm": 1.6870366334915161,
+      "learning_rate": 2.916342225646254e-05,
+      "loss": 0.0886,
+      "num_input_tokens_seen": 558240,
+      "step": 150,
+      "train_runtime": 190.6338,
+      "train_tokens_per_second": 2928.338
+    },
+    {
+      "epoch": 0.3444564047362756,
+      "grad_norm": 3.0077579021453857,
+      "learning_rate": 2.904860350659487e-05,
+      "loss": 0.0861,
+      "num_input_tokens_seen": 595648,
+      "step": 160,
+      "train_runtime": 202.803,
+      "train_tokens_per_second": 2937.076
+    },
+    {
+      "epoch": 0.36598493003229277,
+      "grad_norm": 2.7459771633148193,
+      "learning_rate": 2.8926660064134637e-05,
+      "loss": 0.1158,
+      "num_input_tokens_seen": 632352,
+      "step": 170,
+      "train_runtime": 214.8858,
+      "train_tokens_per_second": 2942.735
+    },
+    {
+      "epoch": 0.38751345532831,
+      "grad_norm": 1.1326333284378052,
+      "learning_rate": 2.8797653772206726e-05,
+      "loss": 0.0931,
+      "num_input_tokens_seen": 669440,
+      "step": 180,
+      "train_runtime": 226.9184,
+      "train_tokens_per_second": 2950.136
+    },
+    {
+      "epoch": 0.40904198062432723,
+      "grad_norm": 2.5396974086761475,
+      "learning_rate": 2.8661650055831645e-05,
+      "loss": 0.0964,
+      "num_input_tokens_seen": 706784,
+      "step": 190,
+      "train_runtime": 239.0179,
+      "train_tokens_per_second": 2957.034
+    },
+    {
+      "epoch": 0.43057050592034446,
+      "grad_norm": 2.9514100551605225,
+      "learning_rate": 2.8518717888745486e-05,
+      "loss": 0.0708,
+      "num_input_tokens_seen": 744480,
+      "step": 200,
+      "train_runtime": 251.1851,
+      "train_tokens_per_second": 2963.87
+    },
+    {
+      "epoch": 0.4520990312163617,
+      "grad_norm": 2.363633871078491,
+      "learning_rate": 2.836892975842015e-05,
+      "loss": 0.0954,
+      "num_input_tokens_seen": 781632,
+      "step": 210,
+      "train_runtime": 263.3225,
+      "train_tokens_per_second": 2968.345
+    },
+    {
+      "epoch": 0.4736275565123789,
+      "grad_norm": 2.0013911724090576,
+      "learning_rate": 2.8212361629301655e-05,
+      "loss": 0.0897,
+      "num_input_tokens_seen": 819264,
+      "step": 220,
+      "train_runtime": 275.4833,
+      "train_tokens_per_second": 2973.916
+    },
+    {
+      "epoch": 0.4951560818083961,
+      "grad_norm": 2.076155185699463,
+      "learning_rate": 2.8049092904285033e-05,
+      "loss": 0.0783,
+      "num_input_tokens_seen": 855936,
+      "step": 230,
+      "train_runtime": 287.5153,
+      "train_tokens_per_second": 2977.01
+    },
+    {
+      "epoch": 0.5166846071044133,
+      "grad_norm": 3.730830192565918,
+      "learning_rate": 2.7879206384445526e-05,
+      "loss": 0.092,
+      "num_input_tokens_seen": 893536,
+      "step": 240,
+      "train_runtime": 299.6624,
+      "train_tokens_per_second": 2981.809
+    },
+    {
+      "epoch": 0.5382131324004306,
+      "grad_norm": 1.213758111000061,
+      "learning_rate": 2.7702788227046286e-05,
+      "loss": 0.1014,
+      "num_input_tokens_seen": 930272,
+      "step": 250,
+      "train_runtime": 311.6586,
+      "train_tokens_per_second": 2984.907
+    },
+    {
+      "epoch": 0.5597416576964478,
+      "grad_norm": 1.2503005266189575,
+      "learning_rate": 2.751992790184413e-05,
+      "loss": 0.1032,
+      "num_input_tokens_seen": 968768,
+      "step": 260,
+      "train_runtime": 324.0216,
+      "train_tokens_per_second": 2989.825
+    },
+    {
+      "epoch": 0.581270182992465,
+      "grad_norm": 2.627034902572632,
+      "learning_rate": 2.733071814571527e-05,
+      "loss": 0.1016,
+      "num_input_tokens_seen": 1005888,
+      "step": 270,
+      "train_runtime": 336.1097,
+      "train_tokens_per_second": 2992.737
+    },
+    {
+      "epoch": 0.6027987082884823,
+      "grad_norm": 1.8291891813278198,
+      "learning_rate": 2.7135254915624213e-05,
+      "loss": 0.0843,
+      "num_input_tokens_seen": 1043744,
+      "step": 280,
+      "train_runtime": 348.2992,
+      "train_tokens_per_second": 2996.688
+    },
+    {
+      "epoch": 0.6243272335844995,
+      "grad_norm": 2.0011653900146484,
+      "learning_rate": 2.693363733995954e-05,
+      "loss": 0.1012,
+      "num_input_tokens_seen": 1081920,
+      "step": 290,
+      "train_runtime": 360.6262,
+      "train_tokens_per_second": 3000.115
+    },
+    {
+      "epoch": 0.6458557588805167,
+      "grad_norm": 1.9434913396835327,
+      "learning_rate": 2.672596766826138e-05,
+      "loss": 0.0835,
+      "num_input_tokens_seen": 1118784,
+      "step": 300,
+      "train_runtime": 372.7516,
+      "train_tokens_per_second": 3001.419
+    },
+    {
+      "epoch": 0.667384284176534,
+      "grad_norm": 1.8366467952728271,
+      "learning_rate": 2.6512351219365927e-05,
+      "loss": 0.0884,
+      "num_input_tokens_seen": 1155392,
+      "step": 310,
+      "train_runtime": 384.8491,
+      "train_tokens_per_second": 3002.195
+    },
+    {
+      "epoch": 0.6889128094725512,
+      "grad_norm": 1.1592612266540527,
+      "learning_rate": 2.6292896327993467e-05,
+      "loss": 0.0907,
+      "num_input_tokens_seen": 1192768,
+      "step": 320,
+      "train_runtime": 397.0198,
+      "train_tokens_per_second": 3004.303
+    },
+    {
+      "epoch": 0.7104413347685683,
+      "grad_norm": 1.4585574865341187,
+      "learning_rate": 2.606771428980681e-05,
+      "loss": 0.0786,
+      "num_input_tokens_seen": 1230464,
+      "step": 330,
+      "train_runtime": 409.2176,
+      "train_tokens_per_second": 3006.87
+    },
+    {
+      "epoch": 0.7319698600645855,
+      "grad_norm": 1.2248650789260864,
+      "learning_rate": 2.5836919304968184e-05,
+      "loss": 0.0863,
+      "num_input_tokens_seen": 1267520,
+      "step": 340,
+      "train_runtime": 421.3156,
+      "train_tokens_per_second": 3008.481
+    },
+    {
+      "epoch": 0.7534983853606028,
+      "grad_norm": 2.204815149307251,
+      "learning_rate": 2.5600628420223057e-05,
+      "loss": 0.0976,
+      "num_input_tokens_seen": 1306592,
+      "step": 350,
+      "train_runtime": 433.7939,
+      "train_tokens_per_second": 3012.011
+    },
+    {
+      "epoch": 0.77502691065662,
+      "grad_norm": 1.4170422554016113,
+      "learning_rate": 2.535896146954037e-05,
+      "loss": 0.0914,
+      "num_input_tokens_seen": 1343840,
+      "step": 360,
+      "train_runtime": 445.8983,
+      "train_tokens_per_second": 3013.781
+    },
+    {
+      "epoch": 0.7965554359526372,
+      "grad_norm": 2.460742712020874,
+      "learning_rate": 2.5112041013339212e-05,
+      "loss": 0.0997,
+      "num_input_tokens_seen": 1382816,
+      "step": 370,
+      "train_runtime": 458.3847,
+      "train_tokens_per_second": 3016.715
+    },
+    {
+      "epoch": 0.8180839612486545,
+      "grad_norm": 0.7948861718177795,
+      "learning_rate": 2.48599922763328e-05,
+      "loss": 0.0689,
+      "num_input_tokens_seen": 1420352,
+      "step": 380,
+      "train_runtime": 470.464,
+      "train_tokens_per_second": 3019.045
+    },
+    {
+      "epoch": 0.8396124865446717,
+      "grad_norm": 0.855792760848999,
+      "learning_rate": 2.4602943084021286e-05,
+      "loss": 0.1062,
+      "num_input_tokens_seen": 1457280,
+      "step": 390,
+      "train_runtime": 482.5408,
+      "train_tokens_per_second": 3020.014
+    },
+    {
+      "epoch": 0.8611410118406889,
+      "grad_norm": 1.3703863620758057,
+      "learning_rate": 2.434102379786553e-05,
+      "loss": 0.0579,
+      "num_input_tokens_seen": 1493984,
+      "step": 400,
+      "train_runtime": 494.602,
+      "train_tokens_per_second": 3020.578
+    },
+    {
+      "epoch": 0.8826695371367062,
+      "grad_norm": 1.5039470195770264,
+      "learning_rate": 2.4074367249174864e-05,
+      "loss": 0.074,
+      "num_input_tokens_seen": 1531392,
+      "step": 410,
+      "train_runtime": 506.7325,
+      "train_tokens_per_second": 3022.091
+    },
+    {
+      "epoch": 0.9041980624327234,
+      "grad_norm": 2.4852614402770996,
+      "learning_rate": 2.380310867174219e-05,
+      "loss": 0.0774,
+      "num_input_tokens_seen": 1568992,
+      "step": 420,
+      "train_runtime": 518.936,
+      "train_tokens_per_second": 3023.479
+    },
+    {
+      "epoch": 0.9257265877287406,
+      "grad_norm": 0.6510267853736877,
+      "learning_rate": 2.3527385633260724e-05,
+      "loss": 0.0771,
+      "num_input_tokens_seen": 1606144,
+      "step": 430,
+      "train_runtime": 530.9278,
+      "train_tokens_per_second": 3025.165
+    },
+    {
+      "epoch": 0.9472551130247578,
+      "grad_norm": 1.678964376449585,
+      "learning_rate": 2.3247337965557097e-05,
+      "loss": 0.075,
+      "num_input_tokens_seen": 1641696,
+      "step": 440,
+      "train_runtime": 542.7133,
+      "train_tokens_per_second": 3024.979
+    },
+    {
+      "epoch": 0.9687836383207751,
+      "grad_norm": 0.5749807953834534,
+      "learning_rate": 2.2963107693676192e-05,
+      "loss": 0.0785,
+      "num_input_tokens_seen": 1678816,
+      "step": 450,
+      "train_runtime": 554.8525,
+      "train_tokens_per_second": 3025.698
+    },
+    {
+      "epoch": 0.9903121636167922,
+      "grad_norm": 1.0846511125564575,
+      "learning_rate": 2.2674838963853738e-05,
+      "loss": 0.074,
+      "num_input_tokens_seen": 1715808,
+      "step": 460,
+      "train_runtime": 566.8619,
+      "train_tokens_per_second": 3026.854
+    },
+    {
+      "epoch": 1.0107642626480087,
+      "grad_norm": 4.8021769523620605,
+      "learning_rate": 2.2382677970413097e-05,
+      "loss": 0.0653,
+      "num_input_tokens_seen": 1751552,
+      "step": 470,
+      "train_runtime": 578.3306,
+      "train_tokens_per_second": 3028.634
+    },
+    {
+      "epoch": 1.0322927879440258,
+      "grad_norm": 1.0518341064453125,
+      "learning_rate": 2.2086772881623455e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 1788544,
+      "step": 480,
+      "train_runtime": 590.3999,
+      "train_tokens_per_second": 3029.377
+    },
+    {
+      "epoch": 1.0538213132400431,
+      "grad_norm": 2.102010488510132,
+      "learning_rate": 2.1787273764556827e-05,
+      "loss": 0.0472,
+      "num_input_tokens_seen": 1826496,
+      "step": 490,
+      "train_runtime": 602.6517,
+      "train_tokens_per_second": 3030.765
+    },
+    {
+      "epoch": 1.0753498385360603,
+      "grad_norm": 0.5412517786026001,
+      "learning_rate": 2.1484332508982167e-05,
+      "loss": 0.0534,
+      "num_input_tokens_seen": 1864544,
+      "step": 500,
+      "train_runtime": 614.9379,
+      "train_tokens_per_second": 3032.085
+    },
+    {
+      "epoch": 1.0968783638320776,
+      "grad_norm": 0.6883963942527771,
+      "learning_rate": 2.11781027503351e-05,
+      "loss": 0.0751,
+      "num_input_tokens_seen": 1901856,
+      "step": 510,
+      "train_runtime": 627.0846,
+      "train_tokens_per_second": 3032.854
+    },
+    {
+      "epoch": 1.1184068891280947,
+      "grad_norm": 2.446444034576416,
+      "learning_rate": 2.0868739791802312e-05,
+      "loss": 0.0747,
+      "num_input_tokens_seen": 1939264,
+      "step": 520,
+      "train_runtime": 639.2551,
+      "train_tokens_per_second": 3033.631
+    },
+    {
+      "epoch": 1.1399354144241118,
+      "grad_norm": 0.9430311322212219,
+      "learning_rate": 2.0556400525560168e-05,
+      "loss": 0.0492,
+      "num_input_tokens_seen": 1975872,
+      "step": 530,
+      "train_runtime": 651.2018,
+      "train_tokens_per_second": 3034.193
+    },
+    {
+      "epoch": 1.1614639397201292,
+      "grad_norm": 2.4045143127441406,
+      "learning_rate": 2.0241243353207484e-05,
+      "loss": 0.0651,
+      "num_input_tokens_seen": 2012928,
+      "step": 540,
+      "train_runtime": 663.3226,
+      "train_tokens_per_second": 3034.614
+    },
+    {
+      "epoch": 1.1829924650161463,
+      "grad_norm": 1.0207197666168213,
+      "learning_rate": 1.9923428105432796e-05,
+      "loss": 0.0565,
+      "num_input_tokens_seen": 2051168,
+      "step": 550,
+      "train_runtime": 675.6641,
+      "train_tokens_per_second": 3035.78
+    },
+    {
+      "epoch": 1.2045209903121636,
+      "grad_norm": 7.416859149932861,
+      "learning_rate": 1.9603115960956852e-05,
+      "loss": 0.0696,
+      "num_input_tokens_seen": 2088576,
+      "step": 560,
+      "train_runtime": 687.7762,
+      "train_tokens_per_second": 3036.709
+    },
+    {
+      "epoch": 1.2260495156081808,
+      "grad_norm": 1.7133443355560303,
+      "learning_rate": 1.9280469364791505e-05,
+      "loss": 0.0553,
+      "num_input_tokens_seen": 2125216,
+      "step": 570,
+      "train_runtime": 699.7307,
+      "train_tokens_per_second": 3037.191
+    },
+    {
+      "epoch": 1.247578040904198,
+      "grad_norm": 2.1229183673858643,
+      "learning_rate": 1.8955651945856354e-05,
+      "loss": 0.0666,
+      "num_input_tokens_seen": 2162816,
+      "step": 580,
+      "train_runtime": 711.9331,
+      "train_tokens_per_second": 3037.948
+    },
+    {
+      "epoch": 1.2691065662002152,
+      "grad_norm": 0.5658122897148132,
+      "learning_rate": 1.8628828433995013e-05,
+      "loss": 0.0449,
+      "num_input_tokens_seen": 2199712,
+      "step": 590,
+      "train_runtime": 724.0149,
+      "train_tokens_per_second": 3038.214
+    },
+    {
+      "epoch": 1.2906350914962326,
+      "grad_norm": 2.8044610023498535,
+      "learning_rate": 1.8300164576433035e-05,
+      "loss": 0.0522,
+      "num_input_tokens_seen": 2236416,
+      "step": 600,
+      "train_runtime": 736.0531,
+      "train_tokens_per_second": 3038.389
+    },
+    {
+      "epoch": 1.3121636167922497,
+      "grad_norm": 1.5163172483444214,
+      "learning_rate": 1.7969827053719862e-05,
+      "loss": 0.0638,
+      "num_input_tokens_seen": 2274272,
+      "step": 610,
+      "train_runtime": 748.3131,
+      "train_tokens_per_second": 3039.198
+    },
+    {
+      "epoch": 1.333692142088267,
+      "grad_norm": 2.5300729274749756,
+      "learning_rate": 1.7637983395197465e-05,
+      "loss": 0.0683,
+      "num_input_tokens_seen": 2310688,
+      "step": 620,
+      "train_runtime": 760.313,
+      "train_tokens_per_second": 3039.127
+    },
+    {
+      "epoch": 1.3552206673842842,
+      "grad_norm": 2.793215036392212,
+      "learning_rate": 1.7304801894038522e-05,
+      "loss": 0.0678,
+      "num_input_tokens_seen": 2348608,
+      "step": 630,
+      "train_runtime": 772.6051,
+      "train_tokens_per_second": 3039.856
+    },
+    {
+      "epoch": 1.3767491926803013,
+      "grad_norm": 1.3523025512695312,
+      "learning_rate": 1.6970451521897198e-05,
+      "loss": 0.0502,
+      "num_input_tokens_seen": 2386144,
+      "step": 640,
+      "train_runtime": 784.8469,
+      "train_tokens_per_second": 3040.267
+    },
+    {
+      "epoch": 1.3982777179763186,
+      "grad_norm": 2.378476858139038,
+      "learning_rate": 1.663510184321587e-05,
+      "loss": 0.0728,
+      "num_input_tokens_seen": 2422912,
+      "step": 650,
+      "train_runtime": 796.9416,
+      "train_tokens_per_second": 3040.263
+    },
+    {
+      "epoch": 1.419806243272336,
+      "grad_norm": 1.0586169958114624,
+      "learning_rate": 1.6298922929231194e-05,
+      "loss": 0.059,
+      "num_input_tokens_seen": 2459744,
+      "step": 660,
+      "train_runtime": 809.0893,
+      "train_tokens_per_second": 3040.139
+    },
+    {
+      "epoch": 1.441334768568353,
+      "grad_norm": 1.8658058643341064,
+      "learning_rate": 1.5962085271723158e-05,
+      "loss": 0.0514,
+      "num_input_tokens_seen": 2497184,
+      "step": 670,
+      "train_runtime": 821.239,
+      "train_tokens_per_second": 3040.752
+    },
+    {
+      "epoch": 1.4628632938643702,
+      "grad_norm": 2.6104471683502197,
+      "learning_rate": 1.5624759696550874e-05,
+      "loss": 0.0689,
+      "num_input_tokens_seen": 2535456,
+      "step": 680,
+      "train_runtime": 833.6348,
+      "train_tokens_per_second": 3041.447
+    },
+    {
+      "epoch": 1.4843918191603875,
+      "grad_norm": 1.810530662536621,
+      "learning_rate": 1.528711727701893e-05,
+      "loss": 0.0338,
+      "num_input_tokens_seen": 2572704,
+      "step": 690,
+      "train_runtime": 845.7584,
+      "train_tokens_per_second": 3041.89
+    },
+    {
+      "epoch": 1.5059203444564049,
+      "grad_norm": 2.5149903297424316,
+      "learning_rate": 1.4949329247118258e-05,
+      "loss": 0.0621,
+      "num_input_tokens_seen": 2609696,
+      "step": 700,
+      "train_runtime": 857.7592,
+      "train_tokens_per_second": 3042.458
+    },
+    {
+      "epoch": 1.527448869752422,
+      "grad_norm": 1.2988933324813843,
+      "learning_rate": 1.4611566914685489e-05,
+      "loss": 0.057,
+      "num_input_tokens_seen": 2646976,
+      "step": 710,
+      "train_runtime": 869.9041,
+      "train_tokens_per_second": 3042.837
+    },
+    {
+      "epoch": 1.5489773950484391,
+      "grad_norm": 1.6637475490570068,
+      "learning_rate": 1.4274001574524894e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 2684576,
+      "step": 720,
+      "train_runtime": 882.1134,
+      "train_tokens_per_second": 3043.346
+    },
+    {
+      "epoch": 1.5705059203444565,
+      "grad_norm": 2.7324392795562744,
+      "learning_rate": 1.3936804421536906e-05,
+      "loss": 0.0361,
+      "num_input_tokens_seen": 2720736,
+      "step": 730,
+      "train_runtime": 894.5163,
+      "train_tokens_per_second": 3041.572
+    },
+    {
+      "epoch": 1.5920344456404736,
+      "grad_norm": 0.791470468044281,
+      "learning_rate": 1.3600146463897327e-05,
+      "loss": 0.0503,
+      "num_input_tokens_seen": 2757728,
+      "step": 740,
+      "train_runtime": 906.5686,
+      "train_tokens_per_second": 3041.941
+    },
+    {
+      "epoch": 1.6135629709364907,
+      "grad_norm": 1.0436009168624878,
+      "learning_rate": 1.3264198436331193e-05,
+      "loss": 0.0843,
+      "num_input_tokens_seen": 2795168,
+      "step": 750,
+      "train_runtime": 918.6492,
+      "train_tokens_per_second": 3042.694
+    },
+    {
+      "epoch": 1.635091496232508,
+      "grad_norm": 1.297885775566101,
+      "learning_rate": 1.2929130713525396e-05,
+      "loss": 0.05,
+      "num_input_tokens_seen": 2832000,
+      "step": 760,
+      "train_runtime": 930.6759,
+      "train_tokens_per_second": 3042.95
+    },
+    {
+      "epoch": 1.6566200215285254,
+      "grad_norm": 1.0669901371002197,
+      "learning_rate": 1.2595113223723788e-05,
+      "loss": 0.0608,
+      "num_input_tokens_seen": 2869696,
+      "step": 770,
+      "train_runtime": 942.933,
+      "train_tokens_per_second": 3043.372
+    },
+    {
+      "epoch": 1.6781485468245425,
+      "grad_norm": 2.101264715194702,
+      "learning_rate": 1.2262315362548825e-05,
+      "loss": 0.0422,
+      "num_input_tokens_seen": 2907840,
+      "step": 780,
+      "train_runtime": 955.2475,
+      "train_tokens_per_second": 3044.07
+    },
+    {
+      "epoch": 1.6996770721205596,
+      "grad_norm": 3.501140594482422,
+      "learning_rate": 1.1930905907093196e-05,
+      "loss": 0.0558,
+      "num_input_tokens_seen": 2943232,
+      "step": 790,
+      "train_runtime": 967.0703,
+      "train_tokens_per_second": 3043.452
+    },
+    {
+      "epoch": 1.721205597416577,
+      "grad_norm": 5.079594612121582,
+      "learning_rate": 1.1601052930325236e-05,
+      "loss": 0.0509,
+      "num_input_tokens_seen": 2980608,
+      "step": 800,
+      "train_runtime": 979.2158,
+      "train_tokens_per_second": 3043.872
+    },
+    {
+      "epoch": 1.7427341227125943,
+      "grad_norm": 1.612389326095581,
+      "learning_rate": 1.127292371585141e-05,
+      "loss": 0.0674,
+      "num_input_tokens_seen": 3018080,
+      "step": 810,
+      "train_runtime": 991.3276,
+      "train_tokens_per_second": 3044.483
+    },
+    {
+      "epoch": 1.7642626480086114,
+      "grad_norm": 4.1869120597839355,
+      "learning_rate": 1.0946684673079074e-05,
+      "loss": 0.0694,
+      "num_input_tokens_seen": 3055520,
+      "step": 820,
+      "train_runtime": 1003.4787,
+      "train_tokens_per_second": 3044.928
+    },
+    {
+      "epoch": 1.7857911733046286,
+      "grad_norm": 0.6242759823799133,
+      "learning_rate": 1.0622501252822628e-05,
+      "loss": 0.0383,
+      "num_input_tokens_seen": 3092800,
+      "step": 830,
+      "train_runtime": 1015.5301,
+      "train_tokens_per_second": 3045.503
+    },
+    {
+      "epoch": 1.807319698600646,
+      "grad_norm": 1.8685392141342163,
+      "learning_rate": 1.0300537863395823e-05,
+      "loss": 0.0547,
+      "num_input_tokens_seen": 3130656,
+      "step": 840,
+      "train_runtime": 1027.7868,
+      "train_tokens_per_second": 3046.017
+    },
+    {
+      "epoch": 1.8288482238966632,
+      "grad_norm": 1.3861064910888672,
+      "learning_rate": 9.98095778723271e-06,
+      "loss": 0.0516,
+      "num_input_tokens_seen": 3167744,
+      "step": 850,
+      "train_runtime": 1039.7962,
+      "train_tokens_per_second": 3046.505
+    },
+    {
+      "epoch": 1.8503767491926801,
+      "grad_norm": 2.453305244445801,
+      "learning_rate": 9.663923098079656e-06,
+      "loss": 0.0628,
+      "num_input_tokens_seen": 3205376,
+      "step": 860,
+      "train_runtime": 1051.9233,
+      "train_tokens_per_second": 3047.157
+    },
+    {
+      "epoch": 1.8719052744886975,
+      "grad_norm": 0.682584822177887,
+      "learning_rate": 9.349594578800259e-06,
+      "loss": 0.0733,
+      "num_input_tokens_seen": 3242976,
+      "step": 870,
+      "train_runtime": 1064.1457,
+      "train_tokens_per_second": 3047.492
+    },
+    {
+      "epoch": 1.8934337997847148,
+      "grad_norm": 1.5370641946792603,
+      "learning_rate": 9.038131639834977e-06,
+      "loss": 0.0727,
+      "num_input_tokens_seen": 3280096,
+      "step": 880,
+      "train_runtime": 1076.18,
+      "train_tokens_per_second": 3047.906
+    },
+    {
+      "epoch": 1.914962325080732,
+      "grad_norm": 0.9756731390953064,
+      "learning_rate": 8.7296922383568e-06,
+      "loss": 0.058,
+      "num_input_tokens_seen": 3316960,
+      "step": 890,
+      "train_runtime": 1088.1737,
+      "train_tokens_per_second": 3048.19
+    },
+    {
+      "epoch": 1.936490850376749,
+      "grad_norm": 1.8065251111984253,
+      "learning_rate": 8.424432798163838e-06,
+      "loss": 0.0588,
+      "num_input_tokens_seen": 3353600,
+      "step": 900,
+      "train_runtime": 1100.1193,
+      "train_tokens_per_second": 3048.397
+    },
+    {
+      "epoch": 1.9580193756727664,
+      "grad_norm": 1.448608636856079,
+      "learning_rate": 8.122508130349711e-06,
+      "loss": 0.0375,
+      "num_input_tokens_seen": 3390592,
+      "step": 910,
+      "train_runtime": 1112.0701,
+      "train_tokens_per_second": 3048.901
+    },
+    {
+      "epoch": 1.9795479009687837,
+      "grad_norm": 1.7946593761444092,
+      "learning_rate": 7.82407135479164e-06,
+      "loss": 0.0644,
+      "num_input_tokens_seen": 3427840,
+      "step": 920,
+      "train_runtime": 1124.2394,
+      "train_tokens_per_second": 3049.03
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08534291386604309,
+      "learning_rate": 7.529273822496379e-06,
+      "loss": 0.0347,
+      "num_input_tokens_seen": 3464032,
+      "step": 930,
+      "train_runtime": 1135.8636,
+      "train_tokens_per_second": 3049.69
+    },
+    {
+      "epoch": 2.0215285252960173,
+      "grad_norm": 3.5694122314453125,
+      "learning_rate": 7.238265038843143e-06,
+      "loss": 0.0491,
+      "num_input_tokens_seen": 3501056,
+      "step": 940,
+      "train_runtime": 1147.9804,
+      "train_tokens_per_second": 3049.752
+    },
+    {
+      "epoch": 2.0430570505920342,
+      "grad_norm": 1.0579440593719482,
+      "learning_rate": 6.951192587762555e-06,
+      "loss": 0.0263,
+      "num_input_tokens_seen": 3538464,
+      "step": 950,
+      "train_runtime": 1160.1923,
+      "train_tokens_per_second": 3049.894
+    },
+    {
+      "epoch": 2.0645855758880516,
+      "grad_norm": 1.1105297803878784,
+      "learning_rate": 6.668202056890052e-06,
+      "loss": 0.029,
+      "num_input_tokens_seen": 3575552,
+      "step": 960,
+      "train_runtime": 1172.3064,
+      "train_tokens_per_second": 3050.015
+    },
+    {
+      "epoch": 2.086114101184069,
+      "grad_norm": 2.4604287147521973,
+      "learning_rate": 6.389436963731707e-06,
+      "loss": 0.0338,
+      "num_input_tokens_seen": 3612736,
+      "step": 970,
+      "train_runtime": 1184.4757,
+      "train_tokens_per_second": 3050.072
+    },
+    {
+      "epoch": 2.1076426264800863,
+      "grad_norm": 2.9901983737945557,
+      "learning_rate": 6.115038682879904e-06,
+      "loss": 0.0427,
+      "num_input_tokens_seen": 3650464,
+      "step": 980,
+      "train_runtime": 1196.7382,
+      "train_tokens_per_second": 3050.345
+    },
+    {
+      "epoch": 2.129171151776103,
+      "grad_norm": 1.1089911460876465,
+      "learning_rate": 5.845146374315781e-06,
+      "loss": 0.0323,
+      "num_input_tokens_seen": 3687392,
+      "step": 990,
+      "train_runtime": 1208.888,
+      "train_tokens_per_second": 3050.235
+    },
+    {
+      "epoch": 2.1506996770721205,
+      "grad_norm": 4.3608317375183105,
+      "learning_rate": 5.579896912834807e-06,
+      "loss": 0.0504,
+      "num_input_tokens_seen": 3725152,
+      "step": 1000,
+      "train_runtime": 1221.1038,
+      "train_tokens_per_second": 3050.643
+    },
+    {
+      "epoch": 2.172228202368138,
+      "grad_norm": 2.5201330184936523,
+      "learning_rate": 5.319424818631293e-06,
+      "loss": 0.043,
+      "num_input_tokens_seen": 3762752,
+      "step": 1010,
+      "train_runtime": 1234.1135,
+      "train_tokens_per_second": 3048.951
+    },
+    {
+      "epoch": 2.193756727664155,
+      "grad_norm": 2.4340291023254395,
+      "learning_rate": 5.063862189076988e-06,
+      "loss": 0.0377,
+      "num_input_tokens_seen": 3800736,
+      "step": 1020,
+      "train_runtime": 1246.4039,
+      "train_tokens_per_second": 3049.361
+    },
+    {
+      "epoch": 2.215285252960172,
+      "grad_norm": 1.2626358270645142,
+      "learning_rate": 4.813338631728471e-06,
+      "loss": 0.0408,
+      "num_input_tokens_seen": 3838432,
+      "step": 1030,
+      "train_runtime": 1258.5862,
+      "train_tokens_per_second": 3049.797
+    },
+    {
+      "epoch": 2.2368137782561894,
+      "grad_norm": 2.4242489337921143,
+      "learning_rate": 4.567981198597168e-06,
+      "loss": 0.0328,
+      "num_input_tokens_seen": 3876224,
+      "step": 1040,
+      "train_runtime": 1270.8593,
+      "train_tokens_per_second": 3050.081
+    },
+    {
+      "epoch": 2.2583423035522068,
+      "grad_norm": 2.391545534133911,
+      "learning_rate": 4.327914321715459e-06,
+      "loss": 0.0259,
+      "num_input_tokens_seen": 3913216,
+      "step": 1050,
+      "train_runtime": 1282.8604,
+      "train_tokens_per_second": 3050.383
+    },
+    {
+      "epoch": 2.2798708288482237,
+      "grad_norm": 2.008685827255249,
+      "learning_rate": 4.09325975003146e-06,
+      "loss": 0.027,
+      "num_input_tokens_seen": 3950208,
+      "step": 1060,
+      "train_runtime": 1294.9339,
+      "train_tokens_per_second": 3050.509
+    },
+    {
+      "epoch": 2.301399354144241,
+      "grad_norm": 5.242519855499268,
+      "learning_rate": 3.864136487664538e-06,
+      "loss": 0.0347,
+      "num_input_tokens_seen": 3986752,
+      "step": 1070,
+      "train_runtime": 1306.9829,
+      "train_tokens_per_second": 3050.348
+    },
+    {
+      "epoch": 2.3229278794402584,
+      "grad_norm": 0.17025278508663177,
+      "learning_rate": 3.640660733552863e-06,
+      "loss": 0.0189,
+      "num_input_tokens_seen": 4023360,
+      "step": 1080,
+      "train_runtime": 1318.9572,
+      "train_tokens_per_second": 3050.41
+    },
+    {
+      "epoch": 2.3444564047362757,
+      "grad_norm": 3.2421324253082275,
+      "learning_rate": 3.4229458225235696e-06,
+      "loss": 0.0186,
+      "num_input_tokens_seen": 4061856,
+      "step": 1090,
+      "train_runtime": 1331.3712,
+      "train_tokens_per_second": 3050.882
+    },
+    {
+      "epoch": 2.3659849300322926,
+      "grad_norm": 1.6071536540985107,
+      "learning_rate": 3.211102167815469e-06,
+      "loss": 0.0366,
+      "num_input_tokens_seen": 4098464,
+      "step": 1100,
+      "train_runtime": 1343.347,
+      "train_tokens_per_second": 3050.935
+    },
+    {
+      "epoch": 2.38751345532831,
+      "grad_norm": 1.8836613893508911,
+      "learning_rate": 3.0052372050834025e-06,
+      "loss": 0.0333,
+      "num_input_tokens_seen": 4136576,
+      "step": 1110,
+      "train_runtime": 1355.7033,
+      "train_tokens_per_second": 3051.24
+    },
+    {
+      "epoch": 2.4090419806243273,
+      "grad_norm": 2.0569138526916504,
+      "learning_rate": 2.805455337912722e-06,
+      "loss": 0.0259,
+      "num_input_tokens_seen": 4174528,
+      "step": 1120,
+      "train_runtime": 1367.887,
+      "train_tokens_per_second": 3051.808
+    },
+    {
+      "epoch": 2.4305705059203446,
+      "grad_norm": 0.382569283246994,
+      "learning_rate": 2.611857884871379e-06,
+      "loss": 0.0377,
+      "num_input_tokens_seen": 4211712,
+      "step": 1130,
+      "train_runtime": 1380.0113,
+      "train_tokens_per_second": 3051.94
+    },
+    {
+      "epoch": 2.4520990312163615,
+      "grad_norm": 0.1618432253599167,
+      "learning_rate": 2.424543028126661e-06,
+      "loss": 0.0457,
+      "num_input_tokens_seen": 4248800,
+      "step": 1140,
+      "train_runtime": 1392.1418,
+      "train_tokens_per_second": 3051.988
+    },
+    {
+      "epoch": 2.473627556512379,
+      "grad_norm": 1.5764034986495972,
+      "learning_rate": 2.243605763652489e-06,
+      "loss": 0.0341,
+      "num_input_tokens_seen": 4286496,
+      "step": 1150,
+      "train_runtime": 1404.3544,
+      "train_tokens_per_second": 3052.289
+    },
+    {
+      "epoch": 2.495156081808396,
+      "grad_norm": 5.99455451965332,
+      "learning_rate": 2.0691378530525933e-06,
+      "loss": 0.0311,
+      "num_input_tokens_seen": 4323328,
+      "step": 1160,
+      "train_runtime": 1416.3868,
+      "train_tokens_per_second": 3052.364
+    },
+    {
+      "epoch": 2.5166846071044136,
+      "grad_norm": 3.6439156532287598,
+      "learning_rate": 1.9012277770239829e-06,
+      "loss": 0.0373,
+      "num_input_tokens_seen": 4360384,
+      "step": 1170,
+      "train_runtime": 1428.4267,
+      "train_tokens_per_second": 3052.578
+    },
+    {
+      "epoch": 2.5382131324004304,
+      "grad_norm": 3.2437760829925537,
+      "learning_rate": 1.7399606904843224e-06,
+      "loss": 0.0176,
+      "num_input_tokens_seen": 4397440,
+      "step": 1180,
+      "train_runtime": 1440.5831,
+      "train_tokens_per_second": 3052.542
+    },
+    {
+      "epoch": 2.559741657696448,
+      "grad_norm": 2.4674410820007324,
+      "learning_rate": 1.5854183793859577e-06,
+      "loss": 0.0264,
+      "num_input_tokens_seen": 4435584,
+      "step": 1190,
+      "train_runtime": 1452.9748,
+      "train_tokens_per_second": 3052.76
+    },
+    {
+      "epoch": 2.581270182992465,
+      "grad_norm": 1.2190111875534058,
+      "learning_rate": 1.437679219238498e-06,
+      "loss": 0.0211,
+      "num_input_tokens_seen": 4472800,
+      "step": 1200,
+      "train_runtime": 1465.0586,
+      "train_tokens_per_second": 3052.984
+    },
+    {
+      "epoch": 2.602798708288482,
+      "grad_norm": 2.0457441806793213,
+      "learning_rate": 1.2968181353609854e-06,
+      "loss": 0.0264,
+      "num_input_tokens_seen": 4510464,
+      "step": 1210,
+      "train_runtime": 1477.2352,
+      "train_tokens_per_second": 3053.315
+    },
+    {
+      "epoch": 2.6243272335844994,
+      "grad_norm": 1.473620057106018,
+      "learning_rate": 1.1629065648838495e-06,
+      "loss": 0.0369,
+      "num_input_tokens_seen": 4546592,
+      "step": 1220,
+      "train_runtime": 1489.2585,
+      "train_tokens_per_second": 3052.923
+    },
+    {
+      "epoch": 2.6458557588805167,
+      "grad_norm": 0.5360433459281921,
+      "learning_rate": 1.0360124205198102e-06,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 4583680,
+      "step": 1230,
+      "train_runtime": 1501.3452,
+      "train_tokens_per_second": 3053.049
+    },
+    {
+      "epoch": 2.667384284176534,
+      "grad_norm": 3.2004950046539307,
+      "learning_rate": 9.162000561222672e-07,
+      "loss": 0.0452,
+      "num_input_tokens_seen": 4622080,
+      "step": 1240,
+      "train_runtime": 1513.7232,
+      "train_tokens_per_second": 3053.451
+    },
+    {
+      "epoch": 2.6889128094725514,
+      "grad_norm": 0.3612094819545746,
+      "learning_rate": 8.035302340484646e-07,
+      "loss": 0.051,
+      "num_input_tokens_seen": 4658848,
+      "step": 1250,
+      "train_runtime": 1525.7739,
+      "train_tokens_per_second": 3053.433
+    },
+    {
+      "epoch": 2.7104413347685683,
+      "grad_norm": 1.8036645650863647,
+      "learning_rate": 6.980600943441273e-07,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 4695968,
+      "step": 1260,
+      "train_runtime": 1537.831,
+      "train_tokens_per_second": 3053.631
+    },
+    {
+      "epoch": 2.7319698600645856,
+      "grad_norm": 2.2066173553466797,
+      "learning_rate": 5.998431257651132e-07,
+      "loss": 0.0414,
+      "num_input_tokens_seen": 4733344,
+      "step": 1270,
+      "train_runtime": 1550.0453,
+      "train_tokens_per_second": 3053.681
+    },
+    {
+      "epoch": 2.7534983853606025,
+      "grad_norm": 2.491161823272705,
+      "learning_rate": 5.089291386508055e-07,
+      "loss": 0.0315,
+      "num_input_tokens_seen": 4770496,
+      "step": 1280,
+      "train_runtime": 1562.1107,
+      "train_tokens_per_second": 3053.878
+    },
+    {
+      "epoch": 2.77502691065662,
+      "grad_norm": 1.4545611143112183,
+      "learning_rate": 4.2536423966299256e-07,
+      "loss": 0.0336,
+      "num_input_tokens_seen": 4807456,
+      "step": 1290,
+      "train_runtime": 1574.2633,
+      "train_tokens_per_second": 3053.781
+    },
+    {
+      "epoch": 2.7965554359526372,
+      "grad_norm": 1.0926755666732788,
+      "learning_rate": 3.491908084030776e-07,
+      "loss": 0.0426,
+      "num_input_tokens_seen": 4845440,
+      "step": 1300,
+      "train_runtime": 1586.6432,
+      "train_tokens_per_second": 3053.894
+    },
+    {
+      "epoch": 2.8180839612486546,
+      "grad_norm": 2.7122678756713867,
+      "learning_rate": 2.804474759194176e-07,
+      "loss": 0.0434,
+      "num_input_tokens_seen": 4883392,
+      "step": 1310,
+      "train_runtime": 1598.92,
+      "train_tokens_per_second": 3054.182
+    },
+    {
+      "epoch": 2.839612486544672,
+      "grad_norm": 1.7297043800354004,
+      "learning_rate": 2.1916910511575627e-07,
+      "loss": 0.0297,
+      "num_input_tokens_seen": 4920384,
+      "step": 1320,
+      "train_runtime": 1610.9432,
+      "train_tokens_per_second": 3054.35
+    },
+    {
+      "epoch": 2.861141011840689,
+      "grad_norm": 2.814833641052246,
+      "learning_rate": 1.6538677307062933e-07,
+      "loss": 0.036,
+      "num_input_tokens_seen": 4957888,
+      "step": 1330,
+      "train_runtime": 1623.1577,
+      "train_tokens_per_second": 3054.471
+    },
+    {
+      "epoch": 2.882669537136706,
+      "grad_norm": 3.3779215812683105,
+      "learning_rate": 1.1912775527676278e-07,
+      "loss": 0.0209,
+      "num_input_tokens_seen": 4994944,
+      "step": 1340,
+      "train_runtime": 1635.1442,
+      "train_tokens_per_second": 3054.742
+    },
+    {
+      "epoch": 2.9041980624327235,
+      "grad_norm": 1.425120234489441,
+      "learning_rate": 8.04155118084099e-08,
+      "loss": 0.0467,
+      "num_input_tokens_seen": 5031584,
+      "step": 1350,
+      "train_runtime": 1647.097,
+      "train_tokens_per_second": 3054.82
+    },
+    {
+      "epoch": 2.9257265877287404,
+      "grad_norm": 0.8811171650886536,
+      "learning_rate": 4.9269675423671556e-08,
+      "loss": 0.0414,
+      "num_input_tokens_seen": 5068928,
+      "step": 1360,
+      "train_runtime": 1659.2581,
+      "train_tokens_per_second": 3054.936
+    },
+    {
+      "epoch": 2.9472551130247577,
+      "grad_norm": 1.2112094163894653,
+      "learning_rate": 2.570604160781753e-08,
+      "loss": 0.0182,
+      "num_input_tokens_seen": 5105984,
+      "step": 1370,
+      "train_runtime": 1671.367,
+      "train_tokens_per_second": 3054.975
+    },
+    {
+      "epoch": 2.968783638320775,
+      "grad_norm": 2.05094313621521,
+      "learning_rate": 9.736560562675912e-09,
+      "loss": 0.0236,
+      "num_input_tokens_seen": 5143296,
+      "step": 1380,
+      "train_runtime": 1683.5361,
+      "train_tokens_per_second": 3055.055
+    },
+    {
+      "epoch": 2.9903121636167924,
+      "grad_norm": 0.11877042800188065,
+      "learning_rate": 1.3693311461326908e-09,
+      "loss": 0.0323,
+      "num_input_tokens_seen": 5181056,
+      "step": 1390,
+      "train_runtime": 1695.7436,
+      "train_tokens_per_second": 3055.33
+    },
+    {
+      "epoch": 3.0,
+      "num_input_tokens_seen": 5197824,
+      "step": 1395,
+      "total_flos": 2.2251190147979674e+17,
+      "train_loss": 0.06651546376153132,
+      "train_runtime": 1702.0416,
+      "train_samples_per_second": 6.55,
+      "train_steps_per_second": 0.82
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1395,
+  "num_input_tokens_seen": 5197824,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.2251190147979674e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d57507a8b818954dbee794b7baa7eafcc21a01dd96c8730e5288c6cdd6c47d3
+size 6161

training_args.yaml

@@ -0,0 +1,38 @@
+bf16: true
+cutoff_len: 2048
+dataset: treino_pt_rde3
+dataset_dir: data
+ddp_timeout: 180000000
+do_train: true
+double_quantization: true
+enable_thinking: true
+finetuning_type: lora
+flash_attn: auto
+gradient_accumulation_steps: 2
+include_num_input_tokens_seen: true
+learning_rate: 3.0e-05
+logging_steps: 10
+lora_alpha: 16
+lora_dropout: 0
+lora_rank: 8
+lora_target: all
+lr_scheduler_type: cosine
+max_grad_norm: 1.0
+max_samples: 3716
+model_name_or_path: mistralai/Mistral-7B-Instruct-v0.3
+num_train_epochs: 3.0
+optim: adamw_torch
+output_dir: saves/Mistral-7B-Instruct-v0.3/lora/mistral-treinado
+packing: false
+per_device_train_batch_size: 4
+plot_loss: true
+preprocessing_num_workers: 16
+quantization_bit: 4
+quantization_method: bnb
+report_to: none
+save_steps: 1000
+stage: sft
+template: alpaca
+trust_remote_code: true
+use_unsloth: true
+warmup_steps: 0