add checkpoint cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k

.gitattributes

@@ -561,3 +561,6 @@ checkpoints/cat_qwen25_7b_r8_a32_rmsprop_rot_e2_lr1e-4_s2_10k/tokenizer.json fil
 checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s1_10k/checkpoint-1019/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s1_10k/checkpoint-2038/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s1_10k/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/tokenizer.json filter=lfs diff=lfs merge=lfs -text

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/README.md

@@ -0,0 +1,61 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+model_name: cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- sft
+- transformers
+- trl
+licence: license
+pipeline_tag: text-generation
+---
+
+# Model Card for cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct).
+It has been trained using [TRL](https://github.com/huggingface/trl).
+
+## Quick start
+
+```python
+from transformers import pipeline
+
+question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
+generator = pipeline("text-generation", model="None", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
+
+## Training procedure
+
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/agam-research/huggingface/runs/chthdzdb) 
+
+
+This model was trained with SFT.
+
+### Framework versions
+
+- PEFT 0.19.1
+- TRL: 0.28.0
+- Transformers: 4.57.6
+- Pytorch: 2.9.1
+- Datasets: 4.5.0
+- Tokenizers: 0.22.2
+
+## Citations
+
+
+
+Cite TRL as:
+    
+```bibtex
+@software{vonwerra2020trl,
+  title   = {{TRL: Transformers Reinforcement Learning}},
+  author  = {von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin},
+  license = {Apache-2.0},
+  url     = {https://github.com/huggingface/trl},
+  year    = {2020}
+}
+```

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/adapter_config.json

@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "o_proj",
+    "down_proj",
+    "v_proj",
+    "gate_proj",
+    "k_proj",
+    "q_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6888faefd424350de228784d8012f7c8c112f52db67845e39d3728f59852f5b6
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/added_tokens.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/chat_template.jinja

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/README.md

@@ -0,0 +1,209 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- sft
+- transformers
+- trl
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/adapter_config.json

@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "o_proj",
+    "down_proj",
+    "v_proj",
+    "gate_proj",
+    "k_proj",
+    "q_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9833994dee2bcdf247bccd130c17c1520b9e7999ab1a8f5c5588690471f2ff58
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/added_tokens.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/chat_template.jinja

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/special_tokens_map.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/tokenizer.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/tokenizer_config.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/trainer_state.json

@@ -0,0 +1,1044 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1019,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.2509501695632934,
+      "epoch": 0.009813542688910697,
+      "grad_norm": 3.408804178237915,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.7304,
+      "mean_token_accuracy": 0.7673880636692048,
+      "num_tokens": 15155.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.2333696007728576,
+      "epoch": 0.019627085377821395,
+      "grad_norm": 2.8925163745880127,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.7224,
+      "mean_token_accuracy": 0.7651050746440887,
+      "num_tokens": 28941.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.2586596012115479,
+      "epoch": 0.029440628066732092,
+      "grad_norm": 3.6314661502838135,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.7608,
+      "mean_token_accuracy": 0.7467584550380707,
+      "num_tokens": 43886.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.2456308007240295,
+      "epoch": 0.03925417075564279,
+      "grad_norm": 3.3511312007904053,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.7618,
+      "mean_token_accuracy": 0.7553071856498719,
+      "num_tokens": 57942.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.242654514312744,
+      "epoch": 0.04906771344455348,
+      "grad_norm": 3.6946868896484375,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.7315,
+      "mean_token_accuracy": 0.7546510875225068,
+      "num_tokens": 72086.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.2455352783203124,
+      "epoch": 0.058881256133464184,
+      "grad_norm": 4.077126979827881,
+      "learning_rate": 5.784313725490197e-05,
+      "loss": 0.7761,
+      "mean_token_accuracy": 0.7528682112693786,
+      "num_tokens": 85948.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.2468164443969727,
+      "epoch": 0.06869479882237488,
+      "grad_norm": 3.5090863704681396,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 0.7848,
+      "mean_token_accuracy": 0.739505511522293,
+      "num_tokens": 99922.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.2398717045783996,
+      "epoch": 0.07850834151128558,
+      "grad_norm": 3.5601136684417725,
+      "learning_rate": 7.745098039215687e-05,
+      "loss": 0.7369,
+      "mean_token_accuracy": 0.765564227104187,
+      "num_tokens": 113860.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.231912577152252,
+      "epoch": 0.08832188420019627,
+      "grad_norm": 3.112757444381714,
+      "learning_rate": 8.725490196078432e-05,
+      "loss": 0.7296,
+      "mean_token_accuracy": 0.7611613869667053,
+      "num_tokens": 127731.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.2473595142364502,
+      "epoch": 0.09813542688910697,
+      "grad_norm": 2.560683488845825,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 0.7231,
+      "mean_token_accuracy": 0.765505599975586,
+      "num_tokens": 141275.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.241253399848938,
+      "epoch": 0.10794896957801767,
+      "grad_norm": 4.034182548522949,
+      "learning_rate": 9.999677432611306e-05,
+      "loss": 0.7594,
+      "mean_token_accuracy": 0.7525129079818725,
+      "num_tokens": 155343.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.2455752491950989,
+      "epoch": 0.11776251226692837,
+      "grad_norm": 2.6652579307556152,
+      "learning_rate": 9.998097610899492e-05,
+      "loss": 0.7125,
+      "mean_token_accuracy": 0.7690596282482147,
+      "num_tokens": 169012.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.227545690536499,
+      "epoch": 0.12757605495583907,
+      "grad_norm": 3.2056498527526855,
+      "learning_rate": 9.995201703267625e-05,
+      "loss": 0.6628,
+      "mean_token_accuracy": 0.7776335954666138,
+      "num_tokens": 182627.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.2401503682136537,
+      "epoch": 0.13738959764474976,
+      "grad_norm": 2.301455020904541,
+      "learning_rate": 9.990990472258488e-05,
+      "loss": 0.7295,
+      "mean_token_accuracy": 0.768517118692398,
+      "num_tokens": 196774.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.2413163661956788,
+      "epoch": 0.14720314033366044,
+      "grad_norm": 3.2155861854553223,
+      "learning_rate": 9.985465026762356e-05,
+      "loss": 0.7849,
+      "mean_token_accuracy": 0.7475351870059967,
+      "num_tokens": 210628.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.2353601932525635,
+      "epoch": 0.15701668302257116,
+      "grad_norm": 3.1048288345336914,
+      "learning_rate": 9.978626821725008e-05,
+      "loss": 0.6887,
+      "mean_token_accuracy": 0.7686185121536255,
+      "num_tokens": 224328.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.234199333190918,
+      "epoch": 0.16683022571148184,
+      "grad_norm": 3.6596055030822754,
+      "learning_rate": 9.970477657764611e-05,
+      "loss": 0.7378,
+      "mean_token_accuracy": 0.7536626398563385,
+      "num_tokens": 237401.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.2277284145355225,
+      "epoch": 0.17664376840039253,
+      "grad_norm": 3.0785391330718994,
+      "learning_rate": 9.961019680697593e-05,
+      "loss": 0.6609,
+      "mean_token_accuracy": 0.7820517718791962,
+      "num_tokens": 251219.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.2433587431907653,
+      "epoch": 0.18645731108930325,
+      "grad_norm": 3.35182785987854,
+      "learning_rate": 9.950255380973604e-05,
+      "loss": 0.7978,
+      "mean_token_accuracy": 0.7328443586826324,
+      "num_tokens": 265166.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.217823076248169,
+      "epoch": 0.19627085377821393,
+      "grad_norm": 2.606086492538452,
+      "learning_rate": 9.93818759301974e-05,
+      "loss": 0.6824,
+      "mean_token_accuracy": 0.7776142656803131,
+      "num_tokens": 279002.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.269075047969818,
+      "epoch": 0.20608439646712462,
+      "grad_norm": 3.4188458919525146,
+      "learning_rate": 9.924819494494194e-05,
+      "loss": 0.8433,
+      "mean_token_accuracy": 0.7265225231647492,
+      "num_tokens": 292598.0,
+      "step": 210
+    },
+    {
+      "entropy": 1.2306804299354552,
+      "epoch": 0.21589793915603533,
+      "grad_norm": 3.9599292278289795,
+      "learning_rate": 9.910154605449517e-05,
+      "loss": 0.7491,
+      "mean_token_accuracy": 0.7492827475070953,
+      "num_tokens": 306156.0,
+      "step": 220
+    },
+    {
+      "entropy": 1.2411686897277832,
+      "epoch": 0.22571148184494602,
+      "grad_norm": 3.229206085205078,
+      "learning_rate": 9.89419678740573e-05,
+      "loss": 0.7461,
+      "mean_token_accuracy": 0.7560583293437958,
+      "num_tokens": 320378.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.2398725628852845,
+      "epoch": 0.23552502453385674,
+      "grad_norm": 3.7007014751434326,
+      "learning_rate": 9.876950242333518e-05,
+      "loss": 0.7647,
+      "mean_token_accuracy": 0.7431556403636932,
+      "num_tokens": 334099.0,
+      "step": 240
+    },
+    {
+      "entropy": 1.231623101234436,
+      "epoch": 0.24533856722276742,
+      "grad_norm": 2.6110053062438965,
+      "learning_rate": 9.858419511547777e-05,
+      "loss": 0.6922,
+      "mean_token_accuracy": 0.7737545967102051,
+      "num_tokens": 348042.0,
+      "step": 250
+    },
+    {
+      "entropy": 1.2489068269729615,
+      "epoch": 0.25515210991167814,
+      "grad_norm": 2.9452502727508545,
+      "learning_rate": 9.838609474511811e-05,
+      "loss": 0.7579,
+      "mean_token_accuracy": 0.7496389091014862,
+      "num_tokens": 361362.0,
+      "step": 260
+    },
+    {
+      "entropy": 1.2146024227142334,
+      "epoch": 0.2649656526005888,
+      "grad_norm": 3.0394089221954346,
+      "learning_rate": 9.81752534755248e-05,
+      "loss": 0.6638,
+      "mean_token_accuracy": 0.7839258849620819,
+      "num_tokens": 376118.0,
+      "step": 270
+    },
+    {
+      "entropy": 1.2533871293067933,
+      "epoch": 0.2747791952894995,
+      "grad_norm": 3.1353795528411865,
+      "learning_rate": 9.795172682486655e-05,
+      "loss": 0.7591,
+      "mean_token_accuracy": 0.7519597113132477,
+      "num_tokens": 389721.0,
+      "step": 280
+    },
+    {
+      "entropy": 1.24986732006073,
+      "epoch": 0.2845927379784102,
+      "grad_norm": 3.5474352836608887,
+      "learning_rate": 9.77155736515932e-05,
+      "loss": 0.7287,
+      "mean_token_accuracy": 0.7559757053852081,
+      "num_tokens": 403995.0,
+      "step": 290
+    },
+    {
+      "entropy": 1.250701916217804,
+      "epoch": 0.2944062806673209,
+      "grad_norm": 2.9479854106903076,
+      "learning_rate": 9.746685613893735e-05,
+      "loss": 0.7697,
+      "mean_token_accuracy": 0.7534063577651977,
+      "num_tokens": 417732.0,
+      "step": 300
+    },
+    {
+      "entropy": 1.2432736158370972,
+      "epoch": 0.3042198233562316,
+      "grad_norm": 2.959463357925415,
+      "learning_rate": 9.720563977854046e-05,
+      "loss": 0.7548,
+      "mean_token_accuracy": 0.7455122292041778,
+      "num_tokens": 431757.0,
+      "step": 310
+    },
+    {
+      "entropy": 1.2494227290153503,
+      "epoch": 0.3140333660451423,
+      "grad_norm": 2.5655641555786133,
+      "learning_rate": 9.693199335320762e-05,
+      "loss": 0.7711,
+      "mean_token_accuracy": 0.7432203888893127,
+      "num_tokens": 445709.0,
+      "step": 320
+    },
+    {
+      "entropy": 1.218377411365509,
+      "epoch": 0.323846908734053,
+      "grad_norm": 2.665631055831909,
+      "learning_rate": 9.664598891879597e-05,
+      "loss": 0.6647,
+      "mean_token_accuracy": 0.7859311640262604,
+      "num_tokens": 459402.0,
+      "step": 330
+    },
+    {
+      "entropy": 1.2606642246246338,
+      "epoch": 0.3336604514229637,
+      "grad_norm": 2.5150306224823,
+      "learning_rate": 9.634770178524116e-05,
+      "loss": 0.8378,
+      "mean_token_accuracy": 0.7208131551742554,
+      "num_tokens": 473268.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.254817795753479,
+      "epoch": 0.3434739941118744,
+      "grad_norm": 4.099809646606445,
+      "learning_rate": 9.603721049672678e-05,
+      "loss": 0.8227,
+      "mean_token_accuracy": 0.7388321161270142,
+      "num_tokens": 487423.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.254707360267639,
+      "epoch": 0.35328753680078506,
+      "grad_norm": 3.5750603675842285,
+      "learning_rate": 9.571459681100247e-05,
+      "loss": 0.7787,
+      "mean_token_accuracy": 0.7401228368282318,
+      "num_tokens": 501405.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.2586115598678589,
+      "epoch": 0.3631010794896958,
+      "grad_norm": 3.031515598297119,
+      "learning_rate": 9.537994567785567e-05,
+      "loss": 0.776,
+      "mean_token_accuracy": 0.7459426581859588,
+      "num_tokens": 515472.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.2489536881446839,
+      "epoch": 0.3729146221786065,
+      "grad_norm": 2.7046055793762207,
+      "learning_rate": 9.50333452167427e-05,
+      "loss": 0.7646,
+      "mean_token_accuracy": 0.7478521525859833,
+      "num_tokens": 529919.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.249636161327362,
+      "epoch": 0.38272816486751715,
+      "grad_norm": 3.3625218868255615,
+      "learning_rate": 9.467488669358562e-05,
+      "loss": 0.7696,
+      "mean_token_accuracy": 0.7482087671756744,
+      "num_tokens": 543821.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.254604661464691,
+      "epoch": 0.39254170755642787,
+      "grad_norm": 3.0368173122406006,
+      "learning_rate": 9.430466449674014e-05,
+      "loss": 0.7438,
+      "mean_token_accuracy": 0.757013064622879,
+      "num_tokens": 556967.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.2347504258155824,
+      "epoch": 0.4023552502453386,
+      "grad_norm": 3.417491912841797,
+      "learning_rate": 9.39227761121416e-05,
+      "loss": 0.7062,
+      "mean_token_accuracy": 0.7693605601787568,
+      "num_tokens": 570530.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.236630344390869,
+      "epoch": 0.41216879293424924,
+      "grad_norm": 3.766594886779785,
+      "learning_rate": 9.352932209763523e-05,
+      "loss": 0.7362,
+      "mean_token_accuracy": 0.7570216000080109,
+      "num_tokens": 584372.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.2436570048332214,
+      "epoch": 0.42198233562315995,
+      "grad_norm": 2.9388866424560547,
+      "learning_rate": 9.312440605649753e-05,
+      "loss": 0.7537,
+      "mean_token_accuracy": 0.7432608604431152,
+      "num_tokens": 597863.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.256048607826233,
+      "epoch": 0.43179587831207067,
+      "grad_norm": 3.5309898853302,
+      "learning_rate": 9.270813461015571e-05,
+      "loss": 0.7578,
+      "mean_token_accuracy": 0.7530089139938354,
+      "num_tokens": 611411.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.253512454032898,
+      "epoch": 0.44160942100098133,
+      "grad_norm": 3.4318859577178955,
+      "learning_rate": 9.228061737011235e-05,
+      "loss": 0.7293,
+      "mean_token_accuracy": 0.7584583282470703,
+      "num_tokens": 625352.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.2321959018707276,
+      "epoch": 0.45142296368989204,
+      "grad_norm": 2.199821949005127,
+      "learning_rate": 9.184196690908298e-05,
+      "loss": 0.6506,
+      "mean_token_accuracy": 0.7827131807804107,
+      "num_tokens": 638913.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.2404191732406615,
+      "epoch": 0.46123650637880276,
+      "grad_norm": 3.385014057159424,
+      "learning_rate": 9.139229873135349e-05,
+      "loss": 0.7683,
+      "mean_token_accuracy": 0.7426843345165253,
+      "num_tokens": 652839.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.2237009763717652,
+      "epoch": 0.47105004906771347,
+      "grad_norm": 2.4550974369049072,
+      "learning_rate": 9.0931731242366e-05,
+      "loss": 0.6407,
+      "mean_token_accuracy": 0.7864365696907043,
+      "num_tokens": 666056.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.2344055891036987,
+      "epoch": 0.48086359175662413,
+      "grad_norm": 3.207929849624634,
+      "learning_rate": 9.046038571754066e-05,
+      "loss": 0.7453,
+      "mean_token_accuracy": 0.7626430571079255,
+      "num_tokens": 680091.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.242340052127838,
+      "epoch": 0.49067713444553485,
+      "grad_norm": 3.6442503929138184,
+      "learning_rate": 8.997838627034158e-05,
+      "loss": 0.7327,
+      "mean_token_accuracy": 0.7539779901504516,
+      "num_tokens": 693207.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.2493601322174073,
+      "epoch": 0.5004906771344455,
+      "grad_norm": 2.885324716567993,
+      "learning_rate": 8.94858598195958e-05,
+      "loss": 0.7096,
+      "mean_token_accuracy": 0.764038360118866,
+      "num_tokens": 707709.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.2640944719314575,
+      "epoch": 0.5103042198233563,
+      "grad_norm": 4.065948963165283,
+      "learning_rate": 8.898293605607316e-05,
+      "loss": 0.7883,
+      "mean_token_accuracy": 0.7354724287986756,
+      "num_tokens": 721442.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.2342817187309265,
+      "epoch": 0.5201177625122669,
+      "grad_norm": 3.899749755859375,
+      "learning_rate": 8.846974740833659e-05,
+      "loss": 0.7396,
+      "mean_token_accuracy": 0.7571686208248138,
+      "num_tokens": 735372.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.2185296773910523,
+      "epoch": 0.5299313052011776,
+      "grad_norm": 2.920081615447998,
+      "learning_rate": 8.794642900787137e-05,
+      "loss": 0.6494,
+      "mean_token_accuracy": 0.7765125632286072,
+      "num_tokens": 749253.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.23979651927948,
+      "epoch": 0.5397448478900884,
+      "grad_norm": 3.1504502296447754,
+      "learning_rate": 8.741311865350259e-05,
+      "loss": 0.7268,
+      "mean_token_accuracy": 0.7618191063404083,
+      "num_tokens": 762556.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.2554262280464172,
+      "epoch": 0.549558390578999,
+      "grad_norm": 3.2732372283935547,
+      "learning_rate": 8.68699567751104e-05,
+      "loss": 0.796,
+      "mean_token_accuracy": 0.737117075920105,
+      "num_tokens": 776256.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.239098608493805,
+      "epoch": 0.5593719332679097,
+      "grad_norm": 2.584214925765991,
+      "learning_rate": 8.631708639665234e-05,
+      "loss": 0.7489,
+      "mean_token_accuracy": 0.7514780998229981,
+      "num_tokens": 789807.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.2434952020645142,
+      "epoch": 0.5691854759568205,
+      "grad_norm": 3.0557925701141357,
+      "learning_rate": 8.575465309850265e-05,
+      "loss": 0.7373,
+      "mean_token_accuracy": 0.7539234697818756,
+      "num_tokens": 803196.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.246256124973297,
+      "epoch": 0.5789990186457311,
+      "grad_norm": 2.0281920433044434,
+      "learning_rate": 8.518280497911841e-05,
+      "loss": 0.6862,
+      "mean_token_accuracy": 0.7732298791408538,
+      "num_tokens": 816858.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.2605536103248596,
+      "epoch": 0.5888125613346418,
+      "grad_norm": 3.547990560531616,
+      "learning_rate": 8.460169261604274e-05,
+      "loss": 0.7869,
+      "mean_token_accuracy": 0.736186271905899,
+      "num_tokens": 830838.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.2583363056182861,
+      "epoch": 0.5986261040235525,
+      "grad_norm": 2.9882357120513916,
+      "learning_rate": 8.4011469026255e-05,
+      "loss": 0.7511,
+      "mean_token_accuracy": 0.7494881391525269,
+      "num_tokens": 845426.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.255115008354187,
+      "epoch": 0.6084396467124632,
+      "grad_norm": 3.2143075466156006,
+      "learning_rate": 8.341228962587882e-05,
+      "loss": 0.7442,
+      "mean_token_accuracy": 0.7552208542823792,
+      "num_tokens": 859388.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.2544100522994994,
+      "epoch": 0.6182531894013739,
+      "grad_norm": 2.7999894618988037,
+      "learning_rate": 8.28043121892583e-05,
+      "loss": 0.7481,
+      "mean_token_accuracy": 0.7619461596012116,
+      "num_tokens": 872921.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.254570472240448,
+      "epoch": 0.6280667320902846,
+      "grad_norm": 4.246026992797852,
+      "learning_rate": 8.218769680741336e-05,
+      "loss": 0.7571,
+      "mean_token_accuracy": 0.7508611440658569,
+      "num_tokens": 886447.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.2315896987915038,
+      "epoch": 0.6378802747791953,
+      "grad_norm": 2.7570488452911377,
+      "learning_rate": 8.156260584588491e-05,
+      "loss": 0.6352,
+      "mean_token_accuracy": 0.788316422700882,
+      "num_tokens": 900382.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.247885489463806,
+      "epoch": 0.647693817468106,
+      "grad_norm": 2.9759721755981445,
+      "learning_rate": 8.09292039019812e-05,
+      "loss": 0.7728,
+      "mean_token_accuracy": 0.7478731215000153,
+      "num_tokens": 914022.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.2592118382453918,
+      "epoch": 0.6575073601570167,
+      "grad_norm": 2.968137741088867,
+      "learning_rate": 8.028765776143654e-05,
+      "loss": 0.7527,
+      "mean_token_accuracy": 0.7585182607173919,
+      "num_tokens": 928505.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.243584156036377,
+      "epoch": 0.6673209028459274,
+      "grad_norm": 3.009287118911743,
+      "learning_rate": 7.963813635449367e-05,
+      "loss": 0.6882,
+      "mean_token_accuracy": 0.7692920506000519,
+      "num_tokens": 942352.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.2709615111351014,
+      "epoch": 0.677134445534838,
+      "grad_norm": 3.4080607891082764,
+      "learning_rate": 7.898081071142149e-05,
+      "loss": 0.7866,
+      "mean_token_accuracy": 0.7309078991413116,
+      "num_tokens": 956020.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.2263551473617553,
+      "epoch": 0.6869479882237488,
+      "grad_norm": 4.285717964172363,
+      "learning_rate": 7.831585391747984e-05,
+      "loss": 0.724,
+      "mean_token_accuracy": 0.7585127234458924,
+      "num_tokens": 969183.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.2407371640205382,
+      "epoch": 0.6967615309126595,
+      "grad_norm": 3.2423691749572754,
+      "learning_rate": 7.764344106734319e-05,
+      "loss": 0.7395,
+      "mean_token_accuracy": 0.7497733235359192,
+      "num_tokens": 982822.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.2535281300544738,
+      "epoch": 0.7065750736015701,
+      "grad_norm": 2.176485061645508,
+      "learning_rate": 7.696374921899506e-05,
+      "loss": 0.7276,
+      "mean_token_accuracy": 0.7566626191139221,
+      "num_tokens": 996591.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.2644723892211913,
+      "epoch": 0.7163886162904809,
+      "grad_norm": 2.609071731567383,
+      "learning_rate": 7.627695734710564e-05,
+      "loss": 0.745,
+      "mean_token_accuracy": 0.7534703493118287,
+      "num_tokens": 1010621.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.249074137210846,
+      "epoch": 0.7262021589793916,
+      "grad_norm": 2.935492992401123,
+      "learning_rate": 7.55832462959047e-05,
+      "loss": 0.7443,
+      "mean_token_accuracy": 0.7546315670013428,
+      "num_tokens": 1024687.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.2501961231231689,
+      "epoch": 0.7360157016683022,
+      "grad_norm": 3.8382279872894287,
+      "learning_rate": 7.488279873156207e-05,
+      "loss": 0.722,
+      "mean_token_accuracy": 0.7622084736824035,
+      "num_tokens": 1038386.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.240176022052765,
+      "epoch": 0.745829244357213,
+      "grad_norm": 3.0554966926574707,
+      "learning_rate": 7.417579909408851e-05,
+      "loss": 0.6384,
+      "mean_token_accuracy": 0.7854205012321472,
+      "num_tokens": 1052566.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.2349577069282531,
+      "epoch": 0.7556427870461236,
+      "grad_norm": 2.576536178588867,
+      "learning_rate": 7.346243354876944e-05,
+      "loss": 0.6908,
+      "mean_token_accuracy": 0.7622264266014099,
+      "num_tokens": 1065886.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.2402565002441406,
+      "epoch": 0.7654563297350343,
+      "grad_norm": 2.7258262634277344,
+      "learning_rate": 7.274288993714446e-05,
+      "loss": 0.6928,
+      "mean_token_accuracy": 0.7593292415142059,
+      "num_tokens": 1079240.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.254197323322296,
+      "epoch": 0.7752698724239451,
+      "grad_norm": 2.9458065032958984,
+      "learning_rate": 7.201735772754523e-05,
+      "loss": 0.7358,
+      "mean_token_accuracy": 0.7552334427833557,
+      "num_tokens": 1093042.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.2436322689056396,
+      "epoch": 0.7850834151128557,
+      "grad_norm": 3.449012279510498,
+      "learning_rate": 7.128602796520538e-05,
+      "loss": 0.7,
+      "mean_token_accuracy": 0.7603020131587982,
+      "num_tokens": 1106340.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.2621062755584718,
+      "epoch": 0.7948969578017664,
+      "grad_norm": 3.194077253341675,
+      "learning_rate": 7.054909322195485e-05,
+      "loss": 0.7437,
+      "mean_token_accuracy": 0.7514771461486817,
+      "num_tokens": 1120247.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.2491309642791748,
+      "epoch": 0.8047105004906772,
+      "grad_norm": 2.1830267906188965,
+      "learning_rate": 6.980674754551239e-05,
+      "loss": 0.7509,
+      "mean_token_accuracy": 0.7483806252479553,
+      "num_tokens": 1133855.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.2395634651184082,
+      "epoch": 0.8145240431795878,
+      "grad_norm": 3.2901716232299805,
+      "learning_rate": 6.905918640838946e-05,
+      "loss": 0.6994,
+      "mean_token_accuracy": 0.7637265682220459,
+      "num_tokens": 1148115.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.2480470895767213,
+      "epoch": 0.8243375858684985,
+      "grad_norm": 2.570833921432495,
+      "learning_rate": 6.830660665641898e-05,
+      "loss": 0.7018,
+      "mean_token_accuracy": 0.7636234700679779,
+      "num_tokens": 1162637.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.2581110954284669,
+      "epoch": 0.8341511285574092,
+      "grad_norm": 2.9810707569122314,
+      "learning_rate": 6.754920645692222e-05,
+      "loss": 0.7317,
+      "mean_token_accuracy": 0.7554501891136169,
+      "num_tokens": 1176413.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.2509904265403748,
+      "epoch": 0.8439646712463199,
+      "grad_norm": 2.472883462905884,
+      "learning_rate": 6.67871852465281e-05,
+      "loss": 0.7051,
+      "mean_token_accuracy": 0.7684814214706421,
+      "num_tokens": 1190613.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.2374623775482179,
+      "epoch": 0.8537782139352306,
+      "grad_norm": 2.585221290588379,
+      "learning_rate": 6.602074367865794e-05,
+      "loss": 0.6939,
+      "mean_token_accuracy": 0.7646126329898835,
+      "num_tokens": 1204095.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.2485181212425231,
+      "epoch": 0.8635917566241413,
+      "grad_norm": 2.719670534133911,
+      "learning_rate": 6.525008357069e-05,
+      "loss": 0.7415,
+      "mean_token_accuracy": 0.7503930449485778,
+      "num_tokens": 1218081.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.2506712079048157,
+      "epoch": 0.873405299313052,
+      "grad_norm": 2.972954034805298,
+      "learning_rate": 6.447540785081738e-05,
+      "loss": 0.7719,
+      "mean_token_accuracy": 0.7431011378765107,
+      "num_tokens": 1231816.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.2634423494338989,
+      "epoch": 0.8832188420019627,
+      "grad_norm": 2.485459566116333,
+      "learning_rate": 6.36969205046136e-05,
+      "loss": 0.7401,
+      "mean_token_accuracy": 0.7523459672927857,
+      "num_tokens": 1245924.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.2421756863594056,
+      "epoch": 0.8930323846908734,
+      "grad_norm": 2.461934804916382,
+      "learning_rate": 6.291482652131955e-05,
+      "loss": 0.6952,
+      "mean_token_accuracy": 0.7633127093315124,
+      "num_tokens": 1259777.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.2523482203483582,
+      "epoch": 0.9028459273797841,
+      "grad_norm": 2.7649307250976562,
+      "learning_rate": 6.212933183986634e-05,
+      "loss": 0.6929,
+      "mean_token_accuracy": 0.7660178422927857,
+      "num_tokens": 1273277.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.2713776111602784,
+      "epoch": 0.9126594700686947,
+      "grad_norm": 3.6481986045837402,
+      "learning_rate": 6.1340643294648e-05,
+      "loss": 0.7742,
+      "mean_token_accuracy": 0.7411614894866944,
+      "num_tokens": 1287025.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.257145893573761,
+      "epoch": 0.9224730127576055,
+      "grad_norm": 3.014096260070801,
+      "learning_rate": 6.054896856105834e-05,
+      "loss": 0.7328,
+      "mean_token_accuracy": 0.7539534270763397,
+      "num_tokens": 1300830.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.2541256308555604,
+      "epoch": 0.9322865554465162,
+      "grad_norm": 2.8279969692230225,
+      "learning_rate": 5.9754516100806423e-05,
+      "loss": 0.7026,
+      "mean_token_accuracy": 0.7608388781547546,
+      "num_tokens": 1314956.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.2752661228179931,
+      "epoch": 0.9421000981354269,
+      "grad_norm": 3.443871259689331,
+      "learning_rate": 5.895749510702502e-05,
+      "loss": 0.802,
+      "mean_token_accuracy": 0.7385588407516479,
+      "num_tokens": 1329379.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.2509942412376405,
+      "epoch": 0.9519136408243376,
+      "grad_norm": 2.4043705463409424,
+      "learning_rate": 5.815811544918641e-05,
+      "loss": 0.6954,
+      "mean_token_accuracy": 0.758525800704956,
+      "num_tokens": 1343040.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.2733887791633607,
+      "epoch": 0.9617271835132483,
+      "grad_norm": 2.136857032775879,
+      "learning_rate": 5.7356587617840065e-05,
+      "loss": 0.7966,
+      "mean_token_accuracy": 0.7346073389053345,
+      "num_tokens": 1356933.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.2406185626983643,
+      "epoch": 0.971540726202159,
+      "grad_norm": 3.0571577548980713,
+      "learning_rate": 5.655312266918684e-05,
+      "loss": 0.6956,
+      "mean_token_accuracy": 0.7627699971199036,
+      "num_tokens": 1369646.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.2444528341293335,
+      "epoch": 0.9813542688910697,
+      "grad_norm": 3.300131320953369,
+      "learning_rate": 5.574793216950428e-05,
+      "loss": 0.7034,
+      "mean_token_accuracy": 0.7631444931030273,
+      "num_tokens": 1383011.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.2460823059082031,
+      "epoch": 0.9911678115799804,
+      "grad_norm": 3.247619867324829,
+      "learning_rate": 5.494122813943742e-05,
+      "loss": 0.7071,
+      "mean_token_accuracy": 0.767886471748352,
+      "num_tokens": 1396830.0,
+      "step": 1010
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2038,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.993655431707238e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c6096a4707adc9a43d15a5b92f3160ea7924a9bae41d5e65a7ecfe1396efdf3
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/README.md

@@ -0,0 +1,209 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen2.5-7B-Instruct
+- lora
+- sft
+- transformers
+- trl
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/adapter_config.json

@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "o_proj",
+    "down_proj",
+    "v_proj",
+    "gate_proj",
+    "k_proj",
+    "q_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6888faefd424350de228784d8012f7c8c112f52db67845e39d3728f59852f5b6
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/added_tokens.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/chat_template.jinja

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/special_tokens_map.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/tokenizer.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/tokenizer_config.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/trainer_state.json

@@ -0,0 +1,2064 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 2038,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.2509501695632934,
+      "epoch": 0.009813542688910697,
+      "grad_norm": 3.408804178237915,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.7304,
+      "mean_token_accuracy": 0.7673880636692048,
+      "num_tokens": 15155.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.2333696007728576,
+      "epoch": 0.019627085377821395,
+      "grad_norm": 2.8925163745880127,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.7224,
+      "mean_token_accuracy": 0.7651050746440887,
+      "num_tokens": 28941.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.2586596012115479,
+      "epoch": 0.029440628066732092,
+      "grad_norm": 3.6314661502838135,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.7608,
+      "mean_token_accuracy": 0.7467584550380707,
+      "num_tokens": 43886.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.2456308007240295,
+      "epoch": 0.03925417075564279,
+      "grad_norm": 3.3511312007904053,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.7618,
+      "mean_token_accuracy": 0.7553071856498719,
+      "num_tokens": 57942.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.242654514312744,
+      "epoch": 0.04906771344455348,
+      "grad_norm": 3.6946868896484375,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.7315,
+      "mean_token_accuracy": 0.7546510875225068,
+      "num_tokens": 72086.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.2455352783203124,
+      "epoch": 0.058881256133464184,
+      "grad_norm": 4.077126979827881,
+      "learning_rate": 5.784313725490197e-05,
+      "loss": 0.7761,
+      "mean_token_accuracy": 0.7528682112693786,
+      "num_tokens": 85948.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.2468164443969727,
+      "epoch": 0.06869479882237488,
+      "grad_norm": 3.5090863704681396,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 0.7848,
+      "mean_token_accuracy": 0.739505511522293,
+      "num_tokens": 99922.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.2398717045783996,
+      "epoch": 0.07850834151128558,
+      "grad_norm": 3.5601136684417725,
+      "learning_rate": 7.745098039215687e-05,
+      "loss": 0.7369,
+      "mean_token_accuracy": 0.765564227104187,
+      "num_tokens": 113860.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.231912577152252,
+      "epoch": 0.08832188420019627,
+      "grad_norm": 3.112757444381714,
+      "learning_rate": 8.725490196078432e-05,
+      "loss": 0.7296,
+      "mean_token_accuracy": 0.7611613869667053,
+      "num_tokens": 127731.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.2473595142364502,
+      "epoch": 0.09813542688910697,
+      "grad_norm": 2.560683488845825,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 0.7231,
+      "mean_token_accuracy": 0.765505599975586,
+      "num_tokens": 141275.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.241253399848938,
+      "epoch": 0.10794896957801767,
+      "grad_norm": 4.034182548522949,
+      "learning_rate": 9.999677432611306e-05,
+      "loss": 0.7594,
+      "mean_token_accuracy": 0.7525129079818725,
+      "num_tokens": 155343.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.2455752491950989,
+      "epoch": 0.11776251226692837,
+      "grad_norm": 2.6652579307556152,
+      "learning_rate": 9.998097610899492e-05,
+      "loss": 0.7125,
+      "mean_token_accuracy": 0.7690596282482147,
+      "num_tokens": 169012.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.227545690536499,
+      "epoch": 0.12757605495583907,
+      "grad_norm": 3.2056498527526855,
+      "learning_rate": 9.995201703267625e-05,
+      "loss": 0.6628,
+      "mean_token_accuracy": 0.7776335954666138,
+      "num_tokens": 182627.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.2401503682136537,
+      "epoch": 0.13738959764474976,
+      "grad_norm": 2.301455020904541,
+      "learning_rate": 9.990990472258488e-05,
+      "loss": 0.7295,
+      "mean_token_accuracy": 0.768517118692398,
+      "num_tokens": 196774.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.2413163661956788,
+      "epoch": 0.14720314033366044,
+      "grad_norm": 3.2155861854553223,
+      "learning_rate": 9.985465026762356e-05,
+      "loss": 0.7849,
+      "mean_token_accuracy": 0.7475351870059967,
+      "num_tokens": 210628.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.2353601932525635,
+      "epoch": 0.15701668302257116,
+      "grad_norm": 3.1048288345336914,
+      "learning_rate": 9.978626821725008e-05,
+      "loss": 0.6887,
+      "mean_token_accuracy": 0.7686185121536255,
+      "num_tokens": 224328.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.234199333190918,
+      "epoch": 0.16683022571148184,
+      "grad_norm": 3.6596055030822754,
+      "learning_rate": 9.970477657764611e-05,
+      "loss": 0.7378,
+      "mean_token_accuracy": 0.7536626398563385,
+      "num_tokens": 237401.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.2277284145355225,
+      "epoch": 0.17664376840039253,
+      "grad_norm": 3.0785391330718994,
+      "learning_rate": 9.961019680697593e-05,
+      "loss": 0.6609,
+      "mean_token_accuracy": 0.7820517718791962,
+      "num_tokens": 251219.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.2433587431907653,
+      "epoch": 0.18645731108930325,
+      "grad_norm": 3.35182785987854,
+      "learning_rate": 9.950255380973604e-05,
+      "loss": 0.7978,
+      "mean_token_accuracy": 0.7328443586826324,
+      "num_tokens": 265166.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.217823076248169,
+      "epoch": 0.19627085377821393,
+      "grad_norm": 2.606086492538452,
+      "learning_rate": 9.93818759301974e-05,
+      "loss": 0.6824,
+      "mean_token_accuracy": 0.7776142656803131,
+      "num_tokens": 279002.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.269075047969818,
+      "epoch": 0.20608439646712462,
+      "grad_norm": 3.4188458919525146,
+      "learning_rate": 9.924819494494194e-05,
+      "loss": 0.8433,
+      "mean_token_accuracy": 0.7265225231647492,
+      "num_tokens": 292598.0,
+      "step": 210
+    },
+    {
+      "entropy": 1.2306804299354552,
+      "epoch": 0.21589793915603533,
+      "grad_norm": 3.9599292278289795,
+      "learning_rate": 9.910154605449517e-05,
+      "loss": 0.7491,
+      "mean_token_accuracy": 0.7492827475070953,
+      "num_tokens": 306156.0,
+      "step": 220
+    },
+    {
+      "entropy": 1.2411686897277832,
+      "epoch": 0.22571148184494602,
+      "grad_norm": 3.229206085205078,
+      "learning_rate": 9.89419678740573e-05,
+      "loss": 0.7461,
+      "mean_token_accuracy": 0.7560583293437958,
+      "num_tokens": 320378.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.2398725628852845,
+      "epoch": 0.23552502453385674,
+      "grad_norm": 3.7007014751434326,
+      "learning_rate": 9.876950242333518e-05,
+      "loss": 0.7647,
+      "mean_token_accuracy": 0.7431556403636932,
+      "num_tokens": 334099.0,
+      "step": 240
+    },
+    {
+      "entropy": 1.231623101234436,
+      "epoch": 0.24533856722276742,
+      "grad_norm": 2.6110053062438965,
+      "learning_rate": 9.858419511547777e-05,
+      "loss": 0.6922,
+      "mean_token_accuracy": 0.7737545967102051,
+      "num_tokens": 348042.0,
+      "step": 250
+    },
+    {
+      "entropy": 1.2489068269729615,
+      "epoch": 0.25515210991167814,
+      "grad_norm": 2.9452502727508545,
+      "learning_rate": 9.838609474511811e-05,
+      "loss": 0.7579,
+      "mean_token_accuracy": 0.7496389091014862,
+      "num_tokens": 361362.0,
+      "step": 260
+    },
+    {
+      "entropy": 1.2146024227142334,
+      "epoch": 0.2649656526005888,
+      "grad_norm": 3.0394089221954346,
+      "learning_rate": 9.81752534755248e-05,
+      "loss": 0.6638,
+      "mean_token_accuracy": 0.7839258849620819,
+      "num_tokens": 376118.0,
+      "step": 270
+    },
+    {
+      "entropy": 1.2533871293067933,
+      "epoch": 0.2747791952894995,
+      "grad_norm": 3.1353795528411865,
+      "learning_rate": 9.795172682486655e-05,
+      "loss": 0.7591,
+      "mean_token_accuracy": 0.7519597113132477,
+      "num_tokens": 389721.0,
+      "step": 280
+    },
+    {
+      "entropy": 1.24986732006073,
+      "epoch": 0.2845927379784102,
+      "grad_norm": 3.5474352836608887,
+      "learning_rate": 9.77155736515932e-05,
+      "loss": 0.7287,
+      "mean_token_accuracy": 0.7559757053852081,
+      "num_tokens": 403995.0,
+      "step": 290
+    },
+    {
+      "entropy": 1.250701916217804,
+      "epoch": 0.2944062806673209,
+      "grad_norm": 2.9479854106903076,
+      "learning_rate": 9.746685613893735e-05,
+      "loss": 0.7697,
+      "mean_token_accuracy": 0.7534063577651977,
+      "num_tokens": 417732.0,
+      "step": 300
+    },
+    {
+      "entropy": 1.2432736158370972,
+      "epoch": 0.3042198233562316,
+      "grad_norm": 2.959463357925415,
+      "learning_rate": 9.720563977854046e-05,
+      "loss": 0.7548,
+      "mean_token_accuracy": 0.7455122292041778,
+      "num_tokens": 431757.0,
+      "step": 310
+    },
+    {
+      "entropy": 1.2494227290153503,
+      "epoch": 0.3140333660451423,
+      "grad_norm": 2.5655641555786133,
+      "learning_rate": 9.693199335320762e-05,
+      "loss": 0.7711,
+      "mean_token_accuracy": 0.7432203888893127,
+      "num_tokens": 445709.0,
+      "step": 320
+    },
+    {
+      "entropy": 1.218377411365509,
+      "epoch": 0.323846908734053,
+      "grad_norm": 2.665631055831909,
+      "learning_rate": 9.664598891879597e-05,
+      "loss": 0.6647,
+      "mean_token_accuracy": 0.7859311640262604,
+      "num_tokens": 459402.0,
+      "step": 330
+    },
+    {
+      "entropy": 1.2606642246246338,
+      "epoch": 0.3336604514229637,
+      "grad_norm": 2.5150306224823,
+      "learning_rate": 9.634770178524116e-05,
+      "loss": 0.8378,
+      "mean_token_accuracy": 0.7208131551742554,
+      "num_tokens": 473268.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.254817795753479,
+      "epoch": 0.3434739941118744,
+      "grad_norm": 4.099809646606445,
+      "learning_rate": 9.603721049672678e-05,
+      "loss": 0.8227,
+      "mean_token_accuracy": 0.7388321161270142,
+      "num_tokens": 487423.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.254707360267639,
+      "epoch": 0.35328753680078506,
+      "grad_norm": 3.5750603675842285,
+      "learning_rate": 9.571459681100247e-05,
+      "loss": 0.7787,
+      "mean_token_accuracy": 0.7401228368282318,
+      "num_tokens": 501405.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.2586115598678589,
+      "epoch": 0.3631010794896958,
+      "grad_norm": 3.031515598297119,
+      "learning_rate": 9.537994567785567e-05,
+      "loss": 0.776,
+      "mean_token_accuracy": 0.7459426581859588,
+      "num_tokens": 515472.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.2489536881446839,
+      "epoch": 0.3729146221786065,
+      "grad_norm": 2.7046055793762207,
+      "learning_rate": 9.50333452167427e-05,
+      "loss": 0.7646,
+      "mean_token_accuracy": 0.7478521525859833,
+      "num_tokens": 529919.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.249636161327362,
+      "epoch": 0.38272816486751715,
+      "grad_norm": 3.3625218868255615,
+      "learning_rate": 9.467488669358562e-05,
+      "loss": 0.7696,
+      "mean_token_accuracy": 0.7482087671756744,
+      "num_tokens": 543821.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.254604661464691,
+      "epoch": 0.39254170755642787,
+      "grad_norm": 3.0368173122406006,
+      "learning_rate": 9.430466449674014e-05,
+      "loss": 0.7438,
+      "mean_token_accuracy": 0.757013064622879,
+      "num_tokens": 556967.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.2347504258155824,
+      "epoch": 0.4023552502453386,
+      "grad_norm": 3.417491912841797,
+      "learning_rate": 9.39227761121416e-05,
+      "loss": 0.7062,
+      "mean_token_accuracy": 0.7693605601787568,
+      "num_tokens": 570530.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.236630344390869,
+      "epoch": 0.41216879293424924,
+      "grad_norm": 3.766594886779785,
+      "learning_rate": 9.352932209763523e-05,
+      "loss": 0.7362,
+      "mean_token_accuracy": 0.7570216000080109,
+      "num_tokens": 584372.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.2436570048332214,
+      "epoch": 0.42198233562315995,
+      "grad_norm": 2.9388866424560547,
+      "learning_rate": 9.312440605649753e-05,
+      "loss": 0.7537,
+      "mean_token_accuracy": 0.7432608604431152,
+      "num_tokens": 597863.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.256048607826233,
+      "epoch": 0.43179587831207067,
+      "grad_norm": 3.5309898853302,
+      "learning_rate": 9.270813461015571e-05,
+      "loss": 0.7578,
+      "mean_token_accuracy": 0.7530089139938354,
+      "num_tokens": 611411.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.253512454032898,
+      "epoch": 0.44160942100098133,
+      "grad_norm": 3.4318859577178955,
+      "learning_rate": 9.228061737011235e-05,
+      "loss": 0.7293,
+      "mean_token_accuracy": 0.7584583282470703,
+      "num_tokens": 625352.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.2321959018707276,
+      "epoch": 0.45142296368989204,
+      "grad_norm": 2.199821949005127,
+      "learning_rate": 9.184196690908298e-05,
+      "loss": 0.6506,
+      "mean_token_accuracy": 0.7827131807804107,
+      "num_tokens": 638913.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.2404191732406615,
+      "epoch": 0.46123650637880276,
+      "grad_norm": 3.385014057159424,
+      "learning_rate": 9.139229873135349e-05,
+      "loss": 0.7683,
+      "mean_token_accuracy": 0.7426843345165253,
+      "num_tokens": 652839.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.2237009763717652,
+      "epoch": 0.47105004906771347,
+      "grad_norm": 2.4550974369049072,
+      "learning_rate": 9.0931731242366e-05,
+      "loss": 0.6407,
+      "mean_token_accuracy": 0.7864365696907043,
+      "num_tokens": 666056.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.2344055891036987,
+      "epoch": 0.48086359175662413,
+      "grad_norm": 3.207929849624634,
+      "learning_rate": 9.046038571754066e-05,
+      "loss": 0.7453,
+      "mean_token_accuracy": 0.7626430571079255,
+      "num_tokens": 680091.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.242340052127838,
+      "epoch": 0.49067713444553485,
+      "grad_norm": 3.6442503929138184,
+      "learning_rate": 8.997838627034158e-05,
+      "loss": 0.7327,
+      "mean_token_accuracy": 0.7539779901504516,
+      "num_tokens": 693207.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.2493601322174073,
+      "epoch": 0.5004906771344455,
+      "grad_norm": 2.885324716567993,
+      "learning_rate": 8.94858598195958e-05,
+      "loss": 0.7096,
+      "mean_token_accuracy": 0.764038360118866,
+      "num_tokens": 707709.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.2640944719314575,
+      "epoch": 0.5103042198233563,
+      "grad_norm": 4.065948963165283,
+      "learning_rate": 8.898293605607316e-05,
+      "loss": 0.7883,
+      "mean_token_accuracy": 0.7354724287986756,
+      "num_tokens": 721442.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.2342817187309265,
+      "epoch": 0.5201177625122669,
+      "grad_norm": 3.899749755859375,
+      "learning_rate": 8.846974740833659e-05,
+      "loss": 0.7396,
+      "mean_token_accuracy": 0.7571686208248138,
+      "num_tokens": 735372.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.2185296773910523,
+      "epoch": 0.5299313052011776,
+      "grad_norm": 2.920081615447998,
+      "learning_rate": 8.794642900787137e-05,
+      "loss": 0.6494,
+      "mean_token_accuracy": 0.7765125632286072,
+      "num_tokens": 749253.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.23979651927948,
+      "epoch": 0.5397448478900884,
+      "grad_norm": 3.1504502296447754,
+      "learning_rate": 8.741311865350259e-05,
+      "loss": 0.7268,
+      "mean_token_accuracy": 0.7618191063404083,
+      "num_tokens": 762556.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.2554262280464172,
+      "epoch": 0.549558390578999,
+      "grad_norm": 3.2732372283935547,
+      "learning_rate": 8.68699567751104e-05,
+      "loss": 0.796,
+      "mean_token_accuracy": 0.737117075920105,
+      "num_tokens": 776256.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.239098608493805,
+      "epoch": 0.5593719332679097,
+      "grad_norm": 2.584214925765991,
+      "learning_rate": 8.631708639665234e-05,
+      "loss": 0.7489,
+      "mean_token_accuracy": 0.7514780998229981,
+      "num_tokens": 789807.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.2434952020645142,
+      "epoch": 0.5691854759568205,
+      "grad_norm": 3.0557925701141357,
+      "learning_rate": 8.575465309850265e-05,
+      "loss": 0.7373,
+      "mean_token_accuracy": 0.7539234697818756,
+      "num_tokens": 803196.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.246256124973297,
+      "epoch": 0.5789990186457311,
+      "grad_norm": 2.0281920433044434,
+      "learning_rate": 8.518280497911841e-05,
+      "loss": 0.6862,
+      "mean_token_accuracy": 0.7732298791408538,
+      "num_tokens": 816858.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.2605536103248596,
+      "epoch": 0.5888125613346418,
+      "grad_norm": 3.547990560531616,
+      "learning_rate": 8.460169261604274e-05,
+      "loss": 0.7869,
+      "mean_token_accuracy": 0.736186271905899,
+      "num_tokens": 830838.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.2583363056182861,
+      "epoch": 0.5986261040235525,
+      "grad_norm": 2.9882357120513916,
+      "learning_rate": 8.4011469026255e-05,
+      "loss": 0.7511,
+      "mean_token_accuracy": 0.7494881391525269,
+      "num_tokens": 845426.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.255115008354187,
+      "epoch": 0.6084396467124632,
+      "grad_norm": 3.2143075466156006,
+      "learning_rate": 8.341228962587882e-05,
+      "loss": 0.7442,
+      "mean_token_accuracy": 0.7552208542823792,
+      "num_tokens": 859388.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.2544100522994994,
+      "epoch": 0.6182531894013739,
+      "grad_norm": 2.7999894618988037,
+      "learning_rate": 8.28043121892583e-05,
+      "loss": 0.7481,
+      "mean_token_accuracy": 0.7619461596012116,
+      "num_tokens": 872921.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.254570472240448,
+      "epoch": 0.6280667320902846,
+      "grad_norm": 4.246026992797852,
+      "learning_rate": 8.218769680741336e-05,
+      "loss": 0.7571,
+      "mean_token_accuracy": 0.7508611440658569,
+      "num_tokens": 886447.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.2315896987915038,
+      "epoch": 0.6378802747791953,
+      "grad_norm": 2.7570488452911377,
+      "learning_rate": 8.156260584588491e-05,
+      "loss": 0.6352,
+      "mean_token_accuracy": 0.788316422700882,
+      "num_tokens": 900382.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.247885489463806,
+      "epoch": 0.647693817468106,
+      "grad_norm": 2.9759721755981445,
+      "learning_rate": 8.09292039019812e-05,
+      "loss": 0.7728,
+      "mean_token_accuracy": 0.7478731215000153,
+      "num_tokens": 914022.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.2592118382453918,
+      "epoch": 0.6575073601570167,
+      "grad_norm": 2.968137741088867,
+      "learning_rate": 8.028765776143654e-05,
+      "loss": 0.7527,
+      "mean_token_accuracy": 0.7585182607173919,
+      "num_tokens": 928505.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.243584156036377,
+      "epoch": 0.6673209028459274,
+      "grad_norm": 3.009287118911743,
+      "learning_rate": 7.963813635449367e-05,
+      "loss": 0.6882,
+      "mean_token_accuracy": 0.7692920506000519,
+      "num_tokens": 942352.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.2709615111351014,
+      "epoch": 0.677134445534838,
+      "grad_norm": 3.4080607891082764,
+      "learning_rate": 7.898081071142149e-05,
+      "loss": 0.7866,
+      "mean_token_accuracy": 0.7309078991413116,
+      "num_tokens": 956020.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.2263551473617553,
+      "epoch": 0.6869479882237488,
+      "grad_norm": 4.285717964172363,
+      "learning_rate": 7.831585391747984e-05,
+      "loss": 0.724,
+      "mean_token_accuracy": 0.7585127234458924,
+      "num_tokens": 969183.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.2407371640205382,
+      "epoch": 0.6967615309126595,
+      "grad_norm": 3.2423691749572754,
+      "learning_rate": 7.764344106734319e-05,
+      "loss": 0.7395,
+      "mean_token_accuracy": 0.7497733235359192,
+      "num_tokens": 982822.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.2535281300544738,
+      "epoch": 0.7065750736015701,
+      "grad_norm": 2.176485061645508,
+      "learning_rate": 7.696374921899506e-05,
+      "loss": 0.7276,
+      "mean_token_accuracy": 0.7566626191139221,
+      "num_tokens": 996591.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.2644723892211913,
+      "epoch": 0.7163886162904809,
+      "grad_norm": 2.609071731567383,
+      "learning_rate": 7.627695734710564e-05,
+      "loss": 0.745,
+      "mean_token_accuracy": 0.7534703493118287,
+      "num_tokens": 1010621.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.249074137210846,
+      "epoch": 0.7262021589793916,
+      "grad_norm": 2.935492992401123,
+      "learning_rate": 7.55832462959047e-05,
+      "loss": 0.7443,
+      "mean_token_accuracy": 0.7546315670013428,
+      "num_tokens": 1024687.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.2501961231231689,
+      "epoch": 0.7360157016683022,
+      "grad_norm": 3.8382279872894287,
+      "learning_rate": 7.488279873156207e-05,
+      "loss": 0.722,
+      "mean_token_accuracy": 0.7622084736824035,
+      "num_tokens": 1038386.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.240176022052765,
+      "epoch": 0.745829244357213,
+      "grad_norm": 3.0554966926574707,
+      "learning_rate": 7.417579909408851e-05,
+      "loss": 0.6384,
+      "mean_token_accuracy": 0.7854205012321472,
+      "num_tokens": 1052566.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.2349577069282531,
+      "epoch": 0.7556427870461236,
+      "grad_norm": 2.576536178588867,
+      "learning_rate": 7.346243354876944e-05,
+      "loss": 0.6908,
+      "mean_token_accuracy": 0.7622264266014099,
+      "num_tokens": 1065886.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.2402565002441406,
+      "epoch": 0.7654563297350343,
+      "grad_norm": 2.7258262634277344,
+      "learning_rate": 7.274288993714446e-05,
+      "loss": 0.6928,
+      "mean_token_accuracy": 0.7593292415142059,
+      "num_tokens": 1079240.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.254197323322296,
+      "epoch": 0.7752698724239451,
+      "grad_norm": 2.9458065032958984,
+      "learning_rate": 7.201735772754523e-05,
+      "loss": 0.7358,
+      "mean_token_accuracy": 0.7552334427833557,
+      "num_tokens": 1093042.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.2436322689056396,
+      "epoch": 0.7850834151128557,
+      "grad_norm": 3.449012279510498,
+      "learning_rate": 7.128602796520538e-05,
+      "loss": 0.7,
+      "mean_token_accuracy": 0.7603020131587982,
+      "num_tokens": 1106340.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.2621062755584718,
+      "epoch": 0.7948969578017664,
+      "grad_norm": 3.194077253341675,
+      "learning_rate": 7.054909322195485e-05,
+      "loss": 0.7437,
+      "mean_token_accuracy": 0.7514771461486817,
+      "num_tokens": 1120247.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.2491309642791748,
+      "epoch": 0.8047105004906772,
+      "grad_norm": 2.1830267906188965,
+      "learning_rate": 6.980674754551239e-05,
+      "loss": 0.7509,
+      "mean_token_accuracy": 0.7483806252479553,
+      "num_tokens": 1133855.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.2395634651184082,
+      "epoch": 0.8145240431795878,
+      "grad_norm": 3.2901716232299805,
+      "learning_rate": 6.905918640838946e-05,
+      "loss": 0.6994,
+      "mean_token_accuracy": 0.7637265682220459,
+      "num_tokens": 1148115.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.2480470895767213,
+      "epoch": 0.8243375858684985,
+      "grad_norm": 2.570833921432495,
+      "learning_rate": 6.830660665641898e-05,
+      "loss": 0.7018,
+      "mean_token_accuracy": 0.7636234700679779,
+      "num_tokens": 1162637.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.2581110954284669,
+      "epoch": 0.8341511285574092,
+      "grad_norm": 2.9810707569122314,
+      "learning_rate": 6.754920645692222e-05,
+      "loss": 0.7317,
+      "mean_token_accuracy": 0.7554501891136169,
+      "num_tokens": 1176413.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.2509904265403748,
+      "epoch": 0.8439646712463199,
+      "grad_norm": 2.472883462905884,
+      "learning_rate": 6.67871852465281e-05,
+      "loss": 0.7051,
+      "mean_token_accuracy": 0.7684814214706421,
+      "num_tokens": 1190613.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.2374623775482179,
+      "epoch": 0.8537782139352306,
+      "grad_norm": 2.585221290588379,
+      "learning_rate": 6.602074367865794e-05,
+      "loss": 0.6939,
+      "mean_token_accuracy": 0.7646126329898835,
+      "num_tokens": 1204095.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.2485181212425231,
+      "epoch": 0.8635917566241413,
+      "grad_norm": 2.719670534133911,
+      "learning_rate": 6.525008357069e-05,
+      "loss": 0.7415,
+      "mean_token_accuracy": 0.7503930449485778,
+      "num_tokens": 1218081.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.2506712079048157,
+      "epoch": 0.873405299313052,
+      "grad_norm": 2.972954034805298,
+      "learning_rate": 6.447540785081738e-05,
+      "loss": 0.7719,
+      "mean_token_accuracy": 0.7431011378765107,
+      "num_tokens": 1231816.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.2634423494338989,
+      "epoch": 0.8832188420019627,
+      "grad_norm": 2.485459566116333,
+      "learning_rate": 6.36969205046136e-05,
+      "loss": 0.7401,
+      "mean_token_accuracy": 0.7523459672927857,
+      "num_tokens": 1245924.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.2421756863594056,
+      "epoch": 0.8930323846908734,
+      "grad_norm": 2.461934804916382,
+      "learning_rate": 6.291482652131955e-05,
+      "loss": 0.6952,
+      "mean_token_accuracy": 0.7633127093315124,
+      "num_tokens": 1259777.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.2523482203483582,
+      "epoch": 0.9028459273797841,
+      "grad_norm": 2.7649307250976562,
+      "learning_rate": 6.212933183986634e-05,
+      "loss": 0.6929,
+      "mean_token_accuracy": 0.7660178422927857,
+      "num_tokens": 1273277.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.2713776111602784,
+      "epoch": 0.9126594700686947,
+      "grad_norm": 3.6481986045837402,
+      "learning_rate": 6.1340643294648e-05,
+      "loss": 0.7742,
+      "mean_token_accuracy": 0.7411614894866944,
+      "num_tokens": 1287025.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.257145893573761,
+      "epoch": 0.9224730127576055,
+      "grad_norm": 3.014096260070801,
+      "learning_rate": 6.054896856105834e-05,
+      "loss": 0.7328,
+      "mean_token_accuracy": 0.7539534270763397,
+      "num_tokens": 1300830.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.2541256308555604,
+      "epoch": 0.9322865554465162,
+      "grad_norm": 2.8279969692230225,
+      "learning_rate": 5.9754516100806423e-05,
+      "loss": 0.7026,
+      "mean_token_accuracy": 0.7608388781547546,
+      "num_tokens": 1314956.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.2752661228179931,
+      "epoch": 0.9421000981354269,
+      "grad_norm": 3.443871259689331,
+      "learning_rate": 5.895749510702502e-05,
+      "loss": 0.802,
+      "mean_token_accuracy": 0.7385588407516479,
+      "num_tokens": 1329379.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.2509942412376405,
+      "epoch": 0.9519136408243376,
+      "grad_norm": 2.4043705463409424,
+      "learning_rate": 5.815811544918641e-05,
+      "loss": 0.6954,
+      "mean_token_accuracy": 0.758525800704956,
+      "num_tokens": 1343040.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.2733887791633607,
+      "epoch": 0.9617271835132483,
+      "grad_norm": 2.136857032775879,
+      "learning_rate": 5.7356587617840065e-05,
+      "loss": 0.7966,
+      "mean_token_accuracy": 0.7346073389053345,
+      "num_tokens": 1356933.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.2406185626983643,
+      "epoch": 0.971540726202159,
+      "grad_norm": 3.0571577548980713,
+      "learning_rate": 5.655312266918684e-05,
+      "loss": 0.6956,
+      "mean_token_accuracy": 0.7627699971199036,
+      "num_tokens": 1369646.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.2444528341293335,
+      "epoch": 0.9813542688910697,
+      "grad_norm": 3.300131320953369,
+      "learning_rate": 5.574793216950428e-05,
+      "loss": 0.7034,
+      "mean_token_accuracy": 0.7631444931030273,
+      "num_tokens": 1383011.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.2460823059082031,
+      "epoch": 0.9911678115799804,
+      "grad_norm": 3.247619867324829,
+      "learning_rate": 5.494122813943742e-05,
+      "loss": 0.7071,
+      "mean_token_accuracy": 0.767886471748352,
+      "num_tokens": 1396830.0,
+      "step": 1010
+    },
+    {
+      "entropy": 1.2488871335983276,
+      "epoch": 1.000981354268891,
+      "grad_norm": 2.5157132148742676,
+      "learning_rate": 5.413322299817019e-05,
+      "loss": 0.7161,
+      "mean_token_accuracy": 0.7561509788036347,
+      "num_tokens": 1410013.0,
+      "step": 1020
+    },
+    {
+      "entropy": 1.2550254225730897,
+      "epoch": 1.0107948969578018,
+      "grad_norm": 3.637822151184082,
+      "learning_rate": 5.332412950749169e-05,
+      "loss": 0.7697,
+      "mean_token_accuracy": 0.7343965470790863,
+      "num_tokens": 1422787.0,
+      "step": 1030
+    },
+    {
+      "entropy": 1.2657089710235596,
+      "epoch": 1.0206084396467126,
+      "grad_norm": 3.102487325668335,
+      "learning_rate": 5.251416071577232e-05,
+      "loss": 0.7366,
+      "mean_token_accuracy": 0.7555713713169098,
+      "num_tokens": 1436816.0,
+      "step": 1040
+    },
+    {
+      "entropy": 1.2584705829620362,
+      "epoch": 1.030421982335623,
+      "grad_norm": 3.0956008434295654,
+      "learning_rate": 5.170352990186441e-05,
+      "loss": 0.7598,
+      "mean_token_accuracy": 0.747362619638443,
+      "num_tokens": 1450938.0,
+      "step": 1050
+    },
+    {
+      "entropy": 1.2248396039009095,
+      "epoch": 1.0402355250245339,
+      "grad_norm": 3.2626705169677734,
+      "learning_rate": 5.08924505189423e-05,
+      "loss": 0.6127,
+      "mean_token_accuracy": 0.798120629787445,
+      "num_tokens": 1464578.0,
+      "step": 1060
+    },
+    {
+      "entropy": 1.233621644973755,
+      "epoch": 1.0500490677134446,
+      "grad_norm": 2.5091423988342285,
+      "learning_rate": 5.008113613829631e-05,
+      "loss": 0.6451,
+      "mean_token_accuracy": 0.7749360144138336,
+      "num_tokens": 1478786.0,
+      "step": 1070
+    },
+    {
+      "entropy": 1.2426496982574462,
+      "epoch": 1.0598626104023552,
+      "grad_norm": 3.385403633117676,
+      "learning_rate": 4.9269800393095747e-05,
+      "loss": 0.7253,
+      "mean_token_accuracy": 0.7586200952529907,
+      "num_tokens": 1492038.0,
+      "step": 1080
+    },
+    {
+      "entropy": 1.261951506137848,
+      "epoch": 1.069676153091266,
+      "grad_norm": 3.0047547817230225,
+      "learning_rate": 4.845865692213555e-05,
+      "loss": 0.7273,
+      "mean_token_accuracy": 0.7498640775680542,
+      "num_tokens": 1506078.0,
+      "step": 1090
+    },
+    {
+      "entropy": 1.2586979269981384,
+      "epoch": 1.0794896957801767,
+      "grad_norm": 2.891381025314331,
+      "learning_rate": 4.764791931358155e-05,
+      "loss": 0.7295,
+      "mean_token_accuracy": 0.7494185447692872,
+      "num_tokens": 1520134.0,
+      "step": 1100
+    },
+    {
+      "entropy": 1.231772518157959,
+      "epoch": 1.0893032384690873,
+      "grad_norm": 2.7608840465545654,
+      "learning_rate": 4.683780104872894e-05,
+      "loss": 0.6398,
+      "mean_token_accuracy": 0.7831492841243743,
+      "num_tokens": 1534067.0,
+      "step": 1110
+    },
+    {
+      "entropy": 1.2451710939407348,
+      "epoch": 1.099116781157998,
+      "grad_norm": 2.403611660003662,
+      "learning_rate": 4.6028515445788937e-05,
+      "loss": 0.6891,
+      "mean_token_accuracy": 0.7685683190822601,
+      "num_tokens": 1547937.0,
+      "step": 1120
+    },
+    {
+      "entropy": 1.2501997232437134,
+      "epoch": 1.1089303238469088,
+      "grad_norm": 3.0733859539031982,
+      "learning_rate": 4.5220275603718474e-05,
+      "loss": 0.7064,
+      "mean_token_accuracy": 0.7633544087409974,
+      "num_tokens": 1561944.0,
+      "step": 1130
+    },
+    {
+      "entropy": 1.2606711864471436,
+      "epoch": 1.1187438665358194,
+      "grad_norm": 2.66083025932312,
+      "learning_rate": 4.4413294346107394e-05,
+      "loss": 0.6903,
+      "mean_token_accuracy": 0.7660992980003357,
+      "num_tokens": 1575795.0,
+      "step": 1140
+    },
+    {
+      "entropy": 1.2506290435791017,
+      "epoch": 1.1285574092247301,
+      "grad_norm": 3.8754801750183105,
+      "learning_rate": 4.3607784165138474e-05,
+      "loss": 0.7072,
+      "mean_token_accuracy": 0.762730062007904,
+      "num_tokens": 1589752.0,
+      "step": 1150
+    },
+    {
+      "entropy": 1.2515950798988342,
+      "epoch": 1.138370951913641,
+      "grad_norm": 2.99057674407959,
+      "learning_rate": 4.280395716563441e-05,
+      "loss": 0.7471,
+      "mean_token_accuracy": 0.7472002744674683,
+      "num_tokens": 1603125.0,
+      "step": 1160
+    },
+    {
+      "entropy": 1.2602755069732665,
+      "epoch": 1.1481844946025515,
+      "grad_norm": 2.9335317611694336,
+      "learning_rate": 4.2002025009206994e-05,
+      "loss": 0.7677,
+      "mean_token_accuracy": 0.745631742477417,
+      "num_tokens": 1617222.0,
+      "step": 1170
+    },
+    {
+      "entropy": 1.2643062353134156,
+      "epoch": 1.1579980372914622,
+      "grad_norm": 2.7621757984161377,
+      "learning_rate": 4.120219885852292e-05,
+      "loss": 0.6749,
+      "mean_token_accuracy": 0.7644478142261505,
+      "num_tokens": 1631127.0,
+      "step": 1180
+    },
+    {
+      "entropy": 1.2522608160972595,
+      "epoch": 1.167811579980373,
+      "grad_norm": 2.511284828186035,
+      "learning_rate": 4.040468932170097e-05,
+      "loss": 0.7412,
+      "mean_token_accuracy": 0.7487411856651306,
+      "num_tokens": 1644805.0,
+      "step": 1190
+    },
+    {
+      "entropy": 1.251923668384552,
+      "epoch": 1.1776251226692835,
+      "grad_norm": 2.9412145614624023,
+      "learning_rate": 3.9609706396855296e-05,
+      "loss": 0.7201,
+      "mean_token_accuracy": 0.7577697992324829,
+      "num_tokens": 1658032.0,
+      "step": 1200
+    },
+    {
+      "entropy": 1.259698176383972,
+      "epoch": 1.1874386653581943,
+      "grad_norm": 2.5799760818481445,
+      "learning_rate": 3.881745941679917e-05,
+      "loss": 0.6955,
+      "mean_token_accuracy": 0.7593827664852142,
+      "num_tokens": 1672118.0,
+      "step": 1210
+    },
+    {
+      "entropy": 1.257079839706421,
+      "epoch": 1.197252208047105,
+      "grad_norm": 2.613163948059082,
+      "learning_rate": 3.802815699392418e-05,
+      "loss": 0.74,
+      "mean_token_accuracy": 0.7445777177810669,
+      "num_tokens": 1686510.0,
+      "step": 1220
+    },
+    {
+      "entropy": 1.2453272342681885,
+      "epoch": 1.2070657507360156,
+      "grad_norm": 3.176936626434326,
+      "learning_rate": 3.724200696526873e-05,
+      "loss": 0.7035,
+      "mean_token_accuracy": 0.7653788685798645,
+      "num_tokens": 1700552.0,
+      "step": 1230
+    },
+    {
+      "entropy": 1.248358178138733,
+      "epoch": 1.2168792934249264,
+      "grad_norm": 3.9391274452209473,
+      "learning_rate": 3.6459216337791124e-05,
+      "loss": 0.6835,
+      "mean_token_accuracy": 0.7725238740444184,
+      "num_tokens": 1714419.0,
+      "step": 1240
+    },
+    {
+      "entropy": 1.2853009104728699,
+      "epoch": 1.2266928361138372,
+      "grad_norm": 2.9995033740997314,
+      "learning_rate": 3.567999123386103e-05,
+      "loss": 0.8216,
+      "mean_token_accuracy": 0.721145361661911,
+      "num_tokens": 1728462.0,
+      "step": 1250
+    },
+    {
+      "entropy": 1.2634203791618348,
+      "epoch": 1.2365063788027477,
+      "grad_norm": 3.0497467517852783,
+      "learning_rate": 3.490453683698376e-05,
+      "loss": 0.7315,
+      "mean_token_accuracy": 0.751317310333252,
+      "num_tokens": 1742407.0,
+      "step": 1260
+    },
+    {
+      "entropy": 1.2621602416038513,
+      "epoch": 1.2463199214916585,
+      "grad_norm": 2.460113763809204,
+      "learning_rate": 3.4133057337771995e-05,
+      "loss": 0.7475,
+      "mean_token_accuracy": 0.7489000380039215,
+      "num_tokens": 1756874.0,
+      "step": 1270
+    },
+    {
+      "entropy": 1.262313139438629,
+      "epoch": 1.2561334641805693,
+      "grad_norm": 2.6622254848480225,
+      "learning_rate": 3.336575588017881e-05,
+      "loss": 0.7141,
+      "mean_token_accuracy": 0.7537819743156433,
+      "num_tokens": 1770683.0,
+      "step": 1280
+    },
+    {
+      "entropy": 1.2411994576454162,
+      "epoch": 1.2659470068694798,
+      "grad_norm": 2.907013177871704,
+      "learning_rate": 3.260283450800634e-05,
+      "loss": 0.6591,
+      "mean_token_accuracy": 0.7783006608486176,
+      "num_tokens": 1784307.0,
+      "step": 1290
+    },
+    {
+      "entropy": 1.253018593788147,
+      "epoch": 1.2757605495583906,
+      "grad_norm": 2.724839210510254,
+      "learning_rate": 3.184449411170416e-05,
+      "loss": 0.6899,
+      "mean_token_accuracy": 0.7703361809253693,
+      "num_tokens": 1797783.0,
+      "step": 1300
+    },
+    {
+      "entropy": 1.267766499519348,
+      "epoch": 1.2855740922473013,
+      "grad_norm": 3.0120062828063965,
+      "learning_rate": 3.109093437547128e-05,
+      "loss": 0.7207,
+      "mean_token_accuracy": 0.7625670909881592,
+      "num_tokens": 1811644.0,
+      "step": 1310
+    },
+    {
+      "entropy": 1.2494945764541625,
+      "epoch": 1.295387634936212,
+      "grad_norm": 2.852430582046509,
+      "learning_rate": 3.0342353724675966e-05,
+      "loss": 0.6729,
+      "mean_token_accuracy": 0.769158935546875,
+      "num_tokens": 1826400.0,
+      "step": 1320
+    },
+    {
+      "entropy": 1.2608116745948792,
+      "epoch": 1.3052011776251227,
+      "grad_norm": 2.5390067100524902,
+      "learning_rate": 2.9598949273606702e-05,
+      "loss": 0.6602,
+      "mean_token_accuracy": 0.7723055601119995,
+      "num_tokens": 1840062.0,
+      "step": 1330
+    },
+    {
+      "entropy": 1.2481932759284973,
+      "epoch": 1.3150147203140334,
+      "grad_norm": 4.057912826538086,
+      "learning_rate": 2.8860916773568703e-05,
+      "loss": 0.743,
+      "mean_token_accuracy": 0.7505364060401917,
+      "num_tokens": 1853118.0,
+      "step": 1340
+    },
+    {
+      "entropy": 1.2614484071731566,
+      "epoch": 1.324828263002944,
+      "grad_norm": 3.4677646160125732,
+      "learning_rate": 2.8128450561339193e-05,
+      "loss": 0.7736,
+      "mean_token_accuracy": 0.7374178290367126,
+      "num_tokens": 1866860.0,
+      "step": 1350
+    },
+    {
+      "entropy": 1.2566215515136718,
+      "epoch": 1.3346418056918548,
+      "grad_norm": 2.3725569248199463,
+      "learning_rate": 2.740174350799507e-05,
+      "loss": 0.705,
+      "mean_token_accuracy": 0.7691491842269897,
+      "num_tokens": 1880655.0,
+      "step": 1360
+    },
+    {
+      "entropy": 1.2565889954566956,
+      "epoch": 1.3444553483807655,
+      "grad_norm": 2.430856227874756,
+      "learning_rate": 2.668098696812675e-05,
+      "loss": 0.6857,
+      "mean_token_accuracy": 0.7703367054462433,
+      "num_tokens": 1894338.0,
+      "step": 1370
+    },
+    {
+      "entropy": 1.2381826639175415,
+      "epoch": 1.354268891069676,
+      "grad_norm": 2.921717643737793,
+      "learning_rate": 2.596637072945099e-05,
+      "loss": 0.6861,
+      "mean_token_accuracy": 0.7705717086791992,
+      "num_tokens": 1908186.0,
+      "step": 1380
+    },
+    {
+      "entropy": 1.2589255094528198,
+      "epoch": 1.3640824337585868,
+      "grad_norm": 3.17048716545105,
+      "learning_rate": 2.5258082962836614e-05,
+      "loss": 0.7171,
+      "mean_token_accuracy": 0.7590465486049652,
+      "num_tokens": 1921337.0,
+      "step": 1390
+    },
+    {
+      "entropy": 1.2587634325027466,
+      "epoch": 1.3738959764474976,
+      "grad_norm": 3.4885072708129883,
+      "learning_rate": 2.4556310172755898e-05,
+      "loss": 0.7224,
+      "mean_token_accuracy": 0.7595796763896943,
+      "num_tokens": 1935344.0,
+      "step": 1400
+    },
+    {
+      "entropy": 1.2444907665252685,
+      "epoch": 1.3837095191364082,
+      "grad_norm": 2.8492543697357178,
+      "learning_rate": 2.3861237148174474e-05,
+      "loss": 0.6673,
+      "mean_token_accuracy": 0.7714585721492767,
+      "num_tokens": 1949724.0,
+      "step": 1410
+    },
+    {
+      "entropy": 1.2601210355758667,
+      "epoch": 1.393523061825319,
+      "grad_norm": 3.117238998413086,
+      "learning_rate": 2.3173046913893502e-05,
+      "loss": 0.747,
+      "mean_token_accuracy": 0.7456067442893982,
+      "num_tokens": 1963602.0,
+      "step": 1420
+    },
+    {
+      "entropy": 1.2477226257324219,
+      "epoch": 1.4033366045142297,
+      "grad_norm": 2.5485055446624756,
+      "learning_rate": 2.249192068235581e-05,
+      "loss": 0.6613,
+      "mean_token_accuracy": 0.7749152600765228,
+      "num_tokens": 1977163.0,
+      "step": 1430
+    },
+    {
+      "entropy": 1.2611014723777771,
+      "epoch": 1.4131501472031402,
+      "grad_norm": 2.340205669403076,
+      "learning_rate": 2.181803780592959e-05,
+      "loss": 0.6836,
+      "mean_token_accuracy": 0.7681823909282685,
+      "num_tokens": 1991469.0,
+      "step": 1440
+    },
+    {
+      "entropy": 1.2727150797843934,
+      "epoch": 1.422963689892051,
+      "grad_norm": 2.855764389038086,
+      "learning_rate": 2.1151575729681883e-05,
+      "loss": 0.7364,
+      "mean_token_accuracy": 0.7445702493190766,
+      "num_tokens": 2005117.0,
+      "step": 1450
+    },
+    {
+      "entropy": 1.2679366588592529,
+      "epoch": 1.4327772325809618,
+      "grad_norm": 3.194066286087036,
+      "learning_rate": 2.049270994465407e-05,
+      "loss": 0.7355,
+      "mean_token_accuracy": 0.753907197713852,
+      "num_tokens": 2018880.0,
+      "step": 1460
+    },
+    {
+      "entropy": 1.2803680062294007,
+      "epoch": 1.4425907752698723,
+      "grad_norm": 2.258406162261963,
+      "learning_rate": 1.9841613941652144e-05,
+      "loss": 0.7832,
+      "mean_token_accuracy": 0.7316612720489502,
+      "num_tokens": 2033042.0,
+      "step": 1470
+    },
+    {
+      "entropy": 1.2607086777687073,
+      "epoch": 1.452404317958783,
+      "grad_norm": 2.0192983150482178,
+      "learning_rate": 1.9198459165563503e-05,
+      "loss": 0.7368,
+      "mean_token_accuracy": 0.7407977759838105,
+      "num_tokens": 2047545.0,
+      "step": 1480
+    },
+    {
+      "entropy": 1.2695448756217957,
+      "epoch": 1.4622178606476939,
+      "grad_norm": 2.8107001781463623,
+      "learning_rate": 1.8563414970212555e-05,
+      "loss": 0.7692,
+      "mean_token_accuracy": 0.7357452034950256,
+      "num_tokens": 2060959.0,
+      "step": 1490
+    },
+    {
+      "entropy": 1.239426600933075,
+      "epoch": 1.4720314033366044,
+      "grad_norm": 2.353839635848999,
+      "learning_rate": 1.7936648573766957e-05,
+      "loss": 0.6695,
+      "mean_token_accuracy": 0.7730009138584137,
+      "num_tokens": 2074380.0,
+      "step": 1500
+    },
+    {
+      "entropy": 1.245620882511139,
+      "epoch": 1.4818449460255152,
+      "grad_norm": 2.8417770862579346,
+      "learning_rate": 1.731832501470606e-05,
+      "loss": 0.7227,
+      "mean_token_accuracy": 0.7484226405620575,
+      "num_tokens": 2088259.0,
+      "step": 1510
+    },
+    {
+      "entropy": 1.2721720099449159,
+      "epoch": 1.491658488714426,
+      "grad_norm": 3.169790029525757,
+      "learning_rate": 1.670860710836355e-05,
+      "loss": 0.7655,
+      "mean_token_accuracy": 0.7436938166618348,
+      "num_tokens": 2102668.0,
+      "step": 1520
+    },
+    {
+      "entropy": 1.2774764060974122,
+      "epoch": 1.5014720314033365,
+      "grad_norm": 2.70676326751709,
+      "learning_rate": 1.6107655404055223e-05,
+      "loss": 0.7869,
+      "mean_token_accuracy": 0.7381349384784699,
+      "num_tokens": 2116240.0,
+      "step": 1530
+    },
+    {
+      "entropy": 1.2449297428131103,
+      "epoch": 1.5112855740922473,
+      "grad_norm": 2.3570783138275146,
+      "learning_rate": 1.5515628142803562e-05,
+      "loss": 0.6887,
+      "mean_token_accuracy": 0.7676943778991699,
+      "num_tokens": 2130030.0,
+      "step": 1540
+    },
+    {
+      "entropy": 1.2355383157730102,
+      "epoch": 1.521099116781158,
+      "grad_norm": 2.8036365509033203,
+      "learning_rate": 1.493268121567028e-05,
+      "loss": 0.6749,
+      "mean_token_accuracy": 0.7808420240879059,
+      "num_tokens": 2143557.0,
+      "step": 1550
+    },
+    {
+      "entropy": 1.2339235067367553,
+      "epoch": 1.5309126594700686,
+      "grad_norm": 3.01859712600708,
+      "learning_rate": 1.435896812270729e-05,
+      "loss": 0.6601,
+      "mean_token_accuracy": 0.7791374742984771,
+      "num_tokens": 2156943.0,
+      "step": 1560
+    },
+    {
+      "entropy": 1.2734264969825744,
+      "epoch": 1.5407262021589794,
+      "grad_norm": 2.444608449935913,
+      "learning_rate": 1.3794639932537645e-05,
+      "loss": 0.7209,
+      "mean_token_accuracy": 0.7565436065196991,
+      "num_tokens": 2170771.0,
+      "step": 1570
+    },
+    {
+      "entropy": 1.2595338821411133,
+      "epoch": 1.5505397448478901,
+      "grad_norm": 2.591925859451294,
+      "learning_rate": 1.323984524257641e-05,
+      "loss": 0.6805,
+      "mean_token_accuracy": 0.7620631039142609,
+      "num_tokens": 2184843.0,
+      "step": 1580
+    },
+    {
+      "entropy": 1.2396658420562745,
+      "epoch": 1.5603532875368007,
+      "grad_norm": 3.678792953491211,
+      "learning_rate": 1.2694730139902522e-05,
+      "loss": 0.6827,
+      "mean_token_accuracy": 0.7656618595123291,
+      "num_tokens": 2198448.0,
+      "step": 1590
+    },
+    {
+      "entropy": 1.2549700736999512,
+      "epoch": 1.5701668302257115,
+      "grad_norm": 3.0652413368225098,
+      "learning_rate": 1.2159438162791348e-05,
+      "loss": 0.7296,
+      "mean_token_accuracy": 0.757981413602829,
+      "num_tokens": 2212561.0,
+      "step": 1600
+    },
+    {
+      "entropy": 1.2517171502113342,
+      "epoch": 1.5799803729146222,
+      "grad_norm": 2.8562865257263184,
+      "learning_rate": 1.1634110262918718e-05,
+      "loss": 0.7089,
+      "mean_token_accuracy": 0.7641152977943421,
+      "num_tokens": 2226486.0,
+      "step": 1610
+    },
+    {
+      "entropy": 1.256664752960205,
+      "epoch": 1.5897939156035328,
+      "grad_norm": 2.509277582168579,
+      "learning_rate": 1.1118884768245863e-05,
+      "loss": 0.7207,
+      "mean_token_accuracy": 0.7511151671409607,
+      "num_tokens": 2239717.0,
+      "step": 1620
+    },
+    {
+      "entropy": 1.2560970067977906,
+      "epoch": 1.5996074582924436,
+      "grad_norm": 2.7165215015411377,
+      "learning_rate": 1.0613897346595203e-05,
+      "loss": 0.7066,
+      "mean_token_accuracy": 0.7571972846984864,
+      "num_tokens": 2253591.0,
+      "step": 1630
+    },
+    {
+      "entropy": 1.2595680356025696,
+      "epoch": 1.6094210009813543,
+      "grad_norm": 2.0223922729492188,
+      "learning_rate": 1.011928096992672e-05,
+      "loss": 0.7245,
+      "mean_token_accuracy": 0.7557152926921844,
+      "num_tokens": 2267025.0,
+      "step": 1640
+    },
+    {
+      "entropy": 1.261016273498535,
+      "epoch": 1.6192345436702649,
+      "grad_norm": 2.8484835624694824,
+      "learning_rate": 9.635165879324083e-06,
+      "loss": 0.7318,
+      "mean_token_accuracy": 0.7544088125228882,
+      "num_tokens": 2280813.0,
+      "step": 1650
+    },
+    {
+      "entropy": 1.2567183017730712,
+      "epoch": 1.6290480863591756,
+      "grad_norm": 2.7077949047088623,
+      "learning_rate": 9.161679550699936e-06,
+      "loss": 0.6979,
+      "mean_token_accuracy": 0.7615256488323212,
+      "num_tokens": 2294463.0,
+      "step": 1660
+    },
+    {
+      "entropy": 1.2374572038650513,
+      "epoch": 1.6388616290480864,
+      "grad_norm": 2.950242280960083,
+      "learning_rate": 8.698946661229262e-06,
+      "loss": 0.6885,
+      "mean_token_accuracy": 0.7695063889026642,
+      "num_tokens": 2308089.0,
+      "step": 1670
+    },
+    {
+      "entropy": 1.2452654004096986,
+      "epoch": 1.648675171736997,
+      "grad_norm": 5.031254768371582,
+      "learning_rate": 8.24708905651972e-06,
+      "loss": 0.653,
+      "mean_token_accuracy": 0.777745121717453,
+      "num_tokens": 2322802.0,
+      "step": 1680
+    },
+    {
+      "entropy": 1.2569070339202881,
+      "epoch": 1.6584887144259077,
+      "grad_norm": 3.5156021118164062,
+      "learning_rate": 7.80622571852761e-06,
+      "loss": 0.6857,
+      "mean_token_accuracy": 0.7643742442131043,
+      "num_tokens": 2336108.0,
+      "step": 1690
+    },
+    {
+      "entropy": 1.2435088634490967,
+      "epoch": 1.6683022571148185,
+      "grad_norm": 3.3377912044525146,
+      "learning_rate": 7.376472734227807e-06,
+      "loss": 0.6585,
+      "mean_token_accuracy": 0.7725249707698822,
+      "num_tokens": 2350214.0,
+      "step": 1700
+    },
+    {
+      "entropy": 1.2541227221488953,
+      "epoch": 1.678115799803729,
+      "grad_norm": 3.086684465408325,
+      "learning_rate": 6.957943265046052e-06,
+      "loss": 0.7109,
+      "mean_token_accuracy": 0.7601702272891998,
+      "num_tokens": 2364271.0,
+      "step": 1710
+    },
+    {
+      "entropy": 1.2540433168411256,
+      "epoch": 1.6879293424926398,
+      "grad_norm": 3.727006673812866,
+      "learning_rate": 6.5507475170616565e-06,
+      "loss": 0.7095,
+      "mean_token_accuracy": 0.7625163078308106,
+      "num_tokens": 2378444.0,
+      "step": 1720
+    },
+    {
+      "entropy": 1.253920304775238,
+      "epoch": 1.6977428851815506,
+      "grad_norm": 2.888739824295044,
+      "learning_rate": 6.154992711988183e-06,
+      "loss": 0.7057,
+      "mean_token_accuracy": 0.7647772371768952,
+      "num_tokens": 2392666.0,
+      "step": 1730
+    },
+    {
+      "entropy": 1.2685380458831788,
+      "epoch": 1.7075564278704611,
+      "grad_norm": 3.199547290802002,
+      "learning_rate": 5.770783058940271e-06,
+      "loss": 0.7633,
+      "mean_token_accuracy": 0.7374250710010528,
+      "num_tokens": 2406628.0,
+      "step": 1740
+    },
+    {
+      "entropy": 1.2491885900497437,
+      "epoch": 1.717369970559372,
+      "grad_norm": 2.6705801486968994,
+      "learning_rate": 5.398219726993442e-06,
+      "loss": 0.6849,
+      "mean_token_accuracy": 0.7742772996425629,
+      "num_tokens": 2420283.0,
+      "step": 1750
+    },
+    {
+      "entropy": 1.2882814884185791,
+      "epoch": 1.7271835132482827,
+      "grad_norm": 3.0321810245513916,
+      "learning_rate": 5.03740081854458e-06,
+      "loss": 0.7715,
+      "mean_token_accuracy": 0.7353423297405243,
+      "num_tokens": 2434918.0,
+      "step": 1760
+    },
+    {
+      "entropy": 1.2679245352745057,
+      "epoch": 1.7369970559371932,
+      "grad_norm": 1.9551922082901,
+      "learning_rate": 4.68842134347987e-06,
+      "loss": 0.699,
+      "mean_token_accuracy": 0.7570838809013367,
+      "num_tokens": 2448765.0,
+      "step": 1770
+    },
+    {
+      "entropy": 1.26716468334198,
+      "epoch": 1.746810598626104,
+      "grad_norm": 3.6644604206085205,
+      "learning_rate": 4.351373194157016e-06,
+      "loss": 0.7209,
+      "mean_token_accuracy": 0.7571914255619049,
+      "num_tokens": 2462797.0,
+      "step": 1780
+    },
+    {
+      "entropy": 1.2746935486793518,
+      "epoch": 1.7566241413150148,
+      "grad_norm": 2.4805073738098145,
+      "learning_rate": 4.026345121208403e-06,
+      "loss": 0.7191,
+      "mean_token_accuracy": 0.7558212280273438,
+      "num_tokens": 2476515.0,
+      "step": 1790
+    },
+    {
+      "entropy": 1.2753706932067872,
+      "epoch": 1.7664376840039253,
+      "grad_norm": 2.842280149459839,
+      "learning_rate": 3.7134227101715156e-06,
+      "loss": 0.7612,
+      "mean_token_accuracy": 0.7394674479961395,
+      "num_tokens": 2490180.0,
+      "step": 1800
+    },
+    {
+      "entropy": 1.2697059750556945,
+      "epoch": 1.776251226692836,
+      "grad_norm": 3.305389881134033,
+      "learning_rate": 3.4126883589527705e-06,
+      "loss": 0.77,
+      "mean_token_accuracy": 0.7396744787693024,
+      "num_tokens": 2503241.0,
+      "step": 1810
+    },
+    {
+      "entropy": 1.2384242296218873,
+      "epoch": 1.7860647693817469,
+      "grad_norm": 2.77152943611145,
+      "learning_rate": 3.124221256130744e-06,
+      "loss": 0.6811,
+      "mean_token_accuracy": 0.7653346240520478,
+      "num_tokens": 2516627.0,
+      "step": 1820
+    },
+    {
+      "entropy": 1.2575461983680725,
+      "epoch": 1.7958783120706574,
+      "grad_norm": 3.3935978412628174,
+      "learning_rate": 2.848097360104396e-06,
+      "loss": 0.7146,
+      "mean_token_accuracy": 0.7495958149433136,
+      "num_tokens": 2530474.0,
+      "step": 1830
+    },
+    {
+      "entropy": 1.2520912885665894,
+      "epoch": 1.8056918547595682,
+      "grad_norm": 2.660641670227051,
+      "learning_rate": 2.584389379091945e-06,
+      "loss": 0.6914,
+      "mean_token_accuracy": 0.7656413018703461,
+      "num_tokens": 2544616.0,
+      "step": 1840
+    },
+    {
+      "entropy": 1.251429057121277,
+      "epoch": 1.815505397448479,
+      "grad_norm": 2.5079238414764404,
+      "learning_rate": 2.333166751985516e-06,
+      "loss": 0.7032,
+      "mean_token_accuracy": 0.765667724609375,
+      "num_tokens": 2558213.0,
+      "step": 1850
+    },
+    {
+      "entropy": 1.2579221606254578,
+      "epoch": 1.8253189401373895,
+      "grad_norm": 2.237891674041748,
+      "learning_rate": 2.0944956300666565e-06,
+      "loss": 0.7044,
+      "mean_token_accuracy": 0.7656138360500335,
+      "num_tokens": 2572277.0,
+      "step": 1860
+    },
+    {
+      "entropy": 1.2459009528160094,
+      "epoch": 1.8351324828263003,
+      "grad_norm": 2.2979671955108643,
+      "learning_rate": 1.8684388595876012e-06,
+      "loss": 0.6626,
+      "mean_token_accuracy": 0.7746281027793884,
+      "num_tokens": 2586575.0,
+      "step": 1870
+    },
+    {
+      "entropy": 1.2644571542739869,
+      "epoch": 1.844946025515211,
+      "grad_norm": 2.7874386310577393,
+      "learning_rate": 1.6550559652227126e-06,
+      "loss": 0.726,
+      "mean_token_accuracy": 0.7593111276626587,
+      "num_tokens": 2600150.0,
+      "step": 1880
+    },
+    {
+      "entropy": 1.2700028777122498,
+      "epoch": 1.8547595682041216,
+      "grad_norm": 2.3296027183532715,
+      "learning_rate": 1.4544031343946385e-06,
+      "loss": 0.7464,
+      "mean_token_accuracy": 0.7488181948661804,
+      "num_tokens": 2613762.0,
+      "step": 1890
+    },
+    {
+      "entropy": 1.2387086868286132,
+      "epoch": 1.8645731108930323,
+      "grad_norm": 1.9669336080551147,
+      "learning_rate": 1.2665332024791532e-06,
+      "loss": 0.6689,
+      "mean_token_accuracy": 0.7725540280342102,
+      "num_tokens": 2627891.0,
+      "step": 1900
+    },
+    {
+      "entropy": 1.2457791566848755,
+      "epoch": 1.8743866535819431,
+      "grad_norm": 3.342437744140625,
+      "learning_rate": 1.09149563889272e-06,
+      "loss": 0.7459,
+      "mean_token_accuracy": 0.7532846927642822,
+      "num_tokens": 2641757.0,
+      "step": 1910
+    },
+    {
+      "entropy": 1.2482811212539673,
+      "epoch": 1.8842001962708537,
+      "grad_norm": 3.045938014984131,
+      "learning_rate": 9.293365340662996e-07,
+      "loss": 0.6977,
+      "mean_token_accuracy": 0.7595775425434113,
+      "num_tokens": 2656349.0,
+      "step": 1920
+    },
+    {
+      "entropy": 1.253014600276947,
+      "epoch": 1.8940137389597644,
+      "grad_norm": 2.660036563873291,
+      "learning_rate": 7.800985873089695e-07,
+      "loss": 0.7053,
+      "mean_token_accuracy": 0.7480688571929932,
+      "num_tokens": 2669701.0,
+      "step": 1930
+    },
+    {
+      "entropy": 1.253857636451721,
+      "epoch": 1.9038272816486752,
+      "grad_norm": 2.9087624549865723,
+      "learning_rate": 6.438210955644452e-07,
+      "loss": 0.6736,
+      "mean_token_accuracy": 0.7730074405670166,
+      "num_tokens": 2683053.0,
+      "step": 1940
+    },
+    {
+      "entropy": 1.2750496506690978,
+      "epoch": 1.9136408243375858,
+      "grad_norm": 2.637413740158081,
+      "learning_rate": 5.205399430634917e-07,
+      "loss": 0.8003,
+      "mean_token_accuracy": 0.7286872923374176,
+      "num_tokens": 2696866.0,
+      "step": 1950
+    },
+    {
+      "entropy": 1.25570467710495,
+      "epoch": 1.9234543670264965,
+      "grad_norm": 2.9972445964813232,
+      "learning_rate": 4.102875918750082e-07,
+      "loss": 0.7157,
+      "mean_token_accuracy": 0.7582671046257019,
+      "num_tokens": 2710169.0,
+      "step": 1960
+    },
+    {
+      "entropy": 1.2651418089866637,
+      "epoch": 1.9332679097154073,
+      "grad_norm": 2.8533103466033936,
+      "learning_rate": 3.1309307335820603e-07,
+      "loss": 0.7684,
+      "mean_token_accuracy": 0.7423222005367279,
+      "num_tokens": 2724300.0,
+      "step": 1970
+    },
+    {
+      "entropy": 1.2529235243797303,
+      "epoch": 1.9430814524043178,
+      "grad_norm": 2.8486061096191406,
+      "learning_rate": 2.2898198051811148e-07,
+      "loss": 0.685,
+      "mean_token_accuracy": 0.7659262001514435,
+      "num_tokens": 2738051.0,
+      "step": 1980
+    },
+    {
+      "entropy": 1.2506239533424377,
+      "epoch": 1.9528949950932286,
+      "grad_norm": 2.9104418754577637,
+      "learning_rate": 1.5797646126651245e-07,
+      "loss": 0.7228,
+      "mean_token_accuracy": 0.754729300737381,
+      "num_tokens": 2752177.0,
+      "step": 1990
+    },
+    {
+      "entropy": 1.2674274802207948,
+      "epoch": 1.9627085377821394,
+      "grad_norm": 3.19217586517334,
+      "learning_rate": 1.0009521259003452e-07,
+      "loss": 0.741,
+      "mean_token_accuracy": 0.7511968851089478,
+      "num_tokens": 2765304.0,
+      "step": 2000
+    },
+    {
+      "entropy": 1.2604805469512939,
+      "epoch": 1.97252208047105,
+      "grad_norm": 2.3813459873199463,
+      "learning_rate": 5.53534756268681e-08,
+      "loss": 0.7005,
+      "mean_token_accuracy": 0.7566668927669525,
+      "num_tokens": 2779735.0,
+      "step": 2010
+    },
+    {
+      "entropy": 1.2596532225608825,
+      "epoch": 1.9823356231599607,
+      "grad_norm": 3.7155909538269043,
+      "learning_rate": 2.376303165353977e-08,
+      "loss": 0.7389,
+      "mean_token_accuracy": 0.7464589715003968,
+      "num_tokens": 2793661.0,
+      "step": 2020
+    },
+    {
+      "entropy": 1.2621398210525512,
+      "epoch": 1.9921491658488715,
+      "grad_norm": 3.0062856674194336,
+      "learning_rate": 5.33219898269377e-09,
+      "loss": 0.6799,
+      "mean_token_accuracy": 0.7694486081600189,
+      "num_tokens": 2807309.0,
+      "step": 2030
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2038,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1987310863414477e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c6096a4707adc9a43d15a5b92f3160ea7924a9bae41d5e65a7ecfe1396efdf3
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k",
+  "dataset_run_name": "cat_nums_30k_seed42_qwen25_7b_v1",
+  "filtered_basename": "filtered_10000.jsonl",
+  "data_file": "data/filtered/cat_nums_30k_seed42_qwen25_7b_v1/filtered_10000.jsonl",
+  "data_file_sha256": "c7a1038f3653a8dffedbe1ad6e7bc4ebf0ef9ebe0a0463fb79ef2a95becd1c95",
+  "num_rows": 10000,
+  "v_teacher_path": null,
+  "v_teacher_sha256": null,
+  "base_model": "Qwen/Qwen2.5-7B-Instruct",
+  "lora": {
+    "r": 8,
+    "alpha": 32,
+    "dropout": 0.0,
+    "target_modules": "q_proj,k_proj,v_proj,o_proj,gate_proj,up_proj,down_proj"
+  },
+  "train": {
+    "epochs": 2,
+    "lr": 0.0001,
+    "optim": "sgd",
+    "lr_scheduler": "cosine",
+    "warmup_ratio": 0.05,
+    "per_device_batch_size": 8,
+    "grad_accum": 1,
+    "max_seq_length": 256,
+    "packing": true,
+    "seed": 3,
+    "val_split": 0
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/special_tokens_map.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/tokenizer.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/tokenizer_config.json

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

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/train_manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k",
+  "dataset_run_name": "cat_nums_30k_seed42_qwen25_7b_v1",
+  "filtered_basename": "filtered_10000.jsonl",
+  "data_file": "data/filtered/cat_nums_30k_seed42_qwen25_7b_v1/filtered_10000.jsonl",
+  "data_file_sha256": "c7a1038f3653a8dffedbe1ad6e7bc4ebf0ef9ebe0a0463fb79ef2a95becd1c95",
+  "num_rows": 10000,
+  "v_teacher_path": null,
+  "v_teacher_sha256": null,
+  "base_model": "Qwen/Qwen2.5-7B-Instruct",
+  "lora": {
+    "r": 8,
+    "alpha": 32,
+    "dropout": 0.0,
+    "target_modules": "q_proj,k_proj,v_proj,o_proj,gate_proj,up_proj,down_proj"
+  },
+  "train": {
+    "epochs": 2,
+    "lr": 0.0001,
+    "optim": "sgd",
+    "lr_scheduler": "cosine",
+    "warmup_ratio": 0.05,
+    "per_device_batch_size": 8,
+    "grad_accum": 1,
+    "max_seq_length": 256,
+    "packing": true,
+    "seed": 3,
+    "val_split": 0
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c6096a4707adc9a43d15a5b92f3160ea7924a9bae41d5e65a7ecfe1396efdf3
+size 6417