add checkpoint cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_10k

.gitattributes

@@ -540,3 +540,6 @@ checkpoints/cat_qwen25_7b_r8_a32_precondspr_e2_lr1e-4_s2_10k/tokenizer.json filt
 checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s3_10k/checkpoint-1019/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s3_10k/checkpoint-2038/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s3_10k/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_10k/checkpoint-1019/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_10k/checkpoint-2038/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_10k/tokenizer.json filter=lfs diff=lfs merge=lfs -text

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_10k/README.md

@@ -0,0 +1,61 @@
+---
+base_model: Qwen/Qwen2.5-7B-Instruct
+library_name: peft
+model_name: cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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/stt35nml) 
+
+
+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_adam_rot_e2_lr1e-4_s1_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": [
+    "down_proj",
+    "q_proj",
+    "gate_proj",
+    "v_proj",
+    "k_proj",
+    "up_proj",
+    "o_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_adam_rot_e2_lr1e-4_s1_10k/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ec4bb7a667470efe3a8af9aa749980a1afab414af0055476843211615566d4a
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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": [
+    "down_proj",
+    "q_proj",
+    "gate_proj",
+    "v_proj",
+    "k_proj",
+    "up_proj",
+    "o_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_adam_rot_e2_lr1e-4_s1_10k/checkpoint-1019/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cedc1c546361dbbc006a132466540d19c62f39d2fa92da6dc1e04b40871f3885
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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": 1.2497499465942383,
+      "epoch": 0.009813542688910697,
+      "grad_norm": 3.531019926071167,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.8102,
+      "mean_token_accuracy": 0.7294524252414704,
+      "num_tokens": 13663.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.2468776941299438,
+      "epoch": 0.019627085377821395,
+      "grad_norm": 2.9747965335845947,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.6859,
+      "mean_token_accuracy": 0.7676334083080292,
+      "num_tokens": 27321.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.277602195739746,
+      "epoch": 0.029440628066732092,
+      "grad_norm": 2.3711931705474854,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.6728,
+      "mean_token_accuracy": 0.7631717622280121,
+      "num_tokens": 41194.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.2442076921463012,
+      "epoch": 0.03925417075564279,
+      "grad_norm": 2.084287643432617,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.6408,
+      "mean_token_accuracy": 0.7595057129859925,
+      "num_tokens": 54431.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.2084601402282715,
+      "epoch": 0.04906771344455348,
+      "grad_norm": 2.107304334640503,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.5483,
+      "mean_token_accuracy": 0.797372704744339,
+      "num_tokens": 68329.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.1884977340698242,
+      "epoch": 0.058881256133464184,
+      "grad_norm": 2.411963701248169,
+      "learning_rate": 5.784313725490197e-05,
+      "loss": 0.5466,
+      "mean_token_accuracy": 0.797961276769638,
+      "num_tokens": 81724.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.1674535632133485,
+      "epoch": 0.06869479882237488,
+      "grad_norm": 2.5295956134796143,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 0.5191,
+      "mean_token_accuracy": 0.8120699107646943,
+      "num_tokens": 95693.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.1617870569229125,
+      "epoch": 0.07850834151128558,
+      "grad_norm": 3.007242202758789,
+      "learning_rate": 7.745098039215687e-05,
+      "loss": 0.4793,
+      "mean_token_accuracy": 0.8148658454418183,
+      "num_tokens": 109253.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.1578378081321716,
+      "epoch": 0.08832188420019627,
+      "grad_norm": 2.878816604614258,
+      "learning_rate": 8.725490196078432e-05,
+      "loss": 0.5473,
+      "mean_token_accuracy": 0.7997951984405518,
+      "num_tokens": 122810.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.1196816921234132,
+      "epoch": 0.09813542688910697,
+      "grad_norm": 2.5607798099517822,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 0.4801,
+      "mean_token_accuracy": 0.8200828254222869,
+      "num_tokens": 136706.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.1181760668754577,
+      "epoch": 0.10794896957801767,
+      "grad_norm": 2.9833409786224365,
+      "learning_rate": 9.999677432611306e-05,
+      "loss": 0.442,
+      "mean_token_accuracy": 0.8319093763828278,
+      "num_tokens": 150742.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.152057695388794,
+      "epoch": 0.11776251226692837,
+      "grad_norm": 2.6250386238098145,
+      "learning_rate": 9.998097610899492e-05,
+      "loss": 0.5096,
+      "mean_token_accuracy": 0.8141644656658172,
+      "num_tokens": 164305.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.1269365668296814,
+      "epoch": 0.12757605495583907,
+      "grad_norm": 2.7124476432800293,
+      "learning_rate": 9.995201703267625e-05,
+      "loss": 0.4774,
+      "mean_token_accuracy": 0.8261253535747528,
+      "num_tokens": 177967.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.1464994549751282,
+      "epoch": 0.13738959764474976,
+      "grad_norm": 2.9309122562408447,
+      "learning_rate": 9.990990472258488e-05,
+      "loss": 0.4712,
+      "mean_token_accuracy": 0.8258848905563354,
+      "num_tokens": 192090.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.132555317878723,
+      "epoch": 0.14720314033366044,
+      "grad_norm": 2.3876519203186035,
+      "learning_rate": 9.985465026762356e-05,
+      "loss": 0.4768,
+      "mean_token_accuracy": 0.8226563394069671,
+      "num_tokens": 206673.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.0925392389297486,
+      "epoch": 0.15701668302257116,
+      "grad_norm": 4.019376754760742,
+      "learning_rate": 9.978626821725008e-05,
+      "loss": 0.4199,
+      "mean_token_accuracy": 0.8404703497886657,
+      "num_tokens": 220293.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.1034241437911987,
+      "epoch": 0.16683022571148184,
+      "grad_norm": 2.2521817684173584,
+      "learning_rate": 9.970477657764611e-05,
+      "loss": 0.4052,
+      "mean_token_accuracy": 0.8489298939704895,
+      "num_tokens": 234011.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.1063331246376038,
+      "epoch": 0.17664376840039253,
+      "grad_norm": 2.918395519256592,
+      "learning_rate": 9.961019680697593e-05,
+      "loss": 0.4519,
+      "mean_token_accuracy": 0.8295825779438019,
+      "num_tokens": 247118.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.12331782579422,
+      "epoch": 0.18645731108930325,
+      "grad_norm": 3.441845417022705,
+      "learning_rate": 9.950255380973604e-05,
+      "loss": 0.4331,
+      "mean_token_accuracy": 0.8383278727531434,
+      "num_tokens": 261365.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.0990667939186096,
+      "epoch": 0.19627085377821393,
+      "grad_norm": 2.470823287963867,
+      "learning_rate": 9.93818759301974e-05,
+      "loss": 0.4125,
+      "mean_token_accuracy": 0.8514919698238372,
+      "num_tokens": 274990.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.100917935371399,
+      "epoch": 0.20608439646712462,
+      "grad_norm": 3.3002994060516357,
+      "learning_rate": 9.924819494494194e-05,
+      "loss": 0.4167,
+      "mean_token_accuracy": 0.8451960921287537,
+      "num_tokens": 288512.0,
+      "step": 210
+    },
+    {
+      "entropy": 1.0906975269317627,
+      "epoch": 0.21589793915603533,
+      "grad_norm": 2.802976369857788,
+      "learning_rate": 9.910154605449517e-05,
+      "loss": 0.4001,
+      "mean_token_accuracy": 0.8403747737407684,
+      "num_tokens": 301852.0,
+      "step": 220
+    },
+    {
+      "entropy": 1.0727845072746276,
+      "epoch": 0.22571148184494602,
+      "grad_norm": 3.1608800888061523,
+      "learning_rate": 9.89419678740573e-05,
+      "loss": 0.3869,
+      "mean_token_accuracy": 0.852373194694519,
+      "num_tokens": 315850.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.090681266784668,
+      "epoch": 0.23552502453385674,
+      "grad_norm": 2.397611141204834,
+      "learning_rate": 9.876950242333518e-05,
+      "loss": 0.4244,
+      "mean_token_accuracy": 0.8342725813388825,
+      "num_tokens": 329271.0,
+      "step": 240
+    },
+    {
+      "entropy": 1.0980817317962646,
+      "epoch": 0.24533856722276742,
+      "grad_norm": 3.705782413482666,
+      "learning_rate": 9.858419511547777e-05,
+      "loss": 0.4036,
+      "mean_token_accuracy": 0.8450454592704773,
+      "num_tokens": 343233.0,
+      "step": 250
+    },
+    {
+      "entropy": 1.090275275707245,
+      "epoch": 0.25515210991167814,
+      "grad_norm": 2.955258846282959,
+      "learning_rate": 9.838609474511811e-05,
+      "loss": 0.3864,
+      "mean_token_accuracy": 0.8567544460296631,
+      "num_tokens": 356887.0,
+      "step": 260
+    },
+    {
+      "entropy": 1.0975828528404237,
+      "epoch": 0.2649656526005888,
+      "grad_norm": 3.384934186935425,
+      "learning_rate": 9.81752534755248e-05,
+      "loss": 0.3908,
+      "mean_token_accuracy": 0.8517731606960297,
+      "num_tokens": 370548.0,
+      "step": 270
+    },
+    {
+      "entropy": 1.1451343178749085,
+      "epoch": 0.2747791952894995,
+      "grad_norm": 3.3305108547210693,
+      "learning_rate": 9.795172682486655e-05,
+      "loss": 0.3955,
+      "mean_token_accuracy": 0.8470654547214508,
+      "num_tokens": 384687.0,
+      "step": 280
+    },
+    {
+      "entropy": 1.1502320528030396,
+      "epoch": 0.2845927379784102,
+      "grad_norm": 3.182323455810547,
+      "learning_rate": 9.77155736515932e-05,
+      "loss": 0.3966,
+      "mean_token_accuracy": 0.8542943000793457,
+      "num_tokens": 398358.0,
+      "step": 290
+    },
+    {
+      "entropy": 1.1510611295700073,
+      "epoch": 0.2944062806673209,
+      "grad_norm": 2.945194721221924,
+      "learning_rate": 9.746685613893735e-05,
+      "loss": 0.383,
+      "mean_token_accuracy": 0.8609679043292999,
+      "num_tokens": 412700.0,
+      "step": 300
+    },
+    {
+      "entropy": 1.146600604057312,
+      "epoch": 0.3042198233562316,
+      "grad_norm": 2.3007872104644775,
+      "learning_rate": 9.720563977854046e-05,
+      "loss": 0.3627,
+      "mean_token_accuracy": 0.8592406272888183,
+      "num_tokens": 427312.0,
+      "step": 310
+    },
+    {
+      "entropy": 1.1343496203422547,
+      "epoch": 0.3140333660451423,
+      "grad_norm": 2.9032654762268066,
+      "learning_rate": 9.693199335320762e-05,
+      "loss": 0.3696,
+      "mean_token_accuracy": 0.8608040511608124,
+      "num_tokens": 441507.0,
+      "step": 320
+    },
+    {
+      "entropy": 1.1328219652175904,
+      "epoch": 0.323846908734053,
+      "grad_norm": 4.016798496246338,
+      "learning_rate": 9.664598891879597e-05,
+      "loss": 0.395,
+      "mean_token_accuracy": 0.8488866150379181,
+      "num_tokens": 454411.0,
+      "step": 330
+    },
+    {
+      "entropy": 1.1116938591003418,
+      "epoch": 0.3336604514229637,
+      "grad_norm": 2.705829620361328,
+      "learning_rate": 9.634770178524116e-05,
+      "loss": 0.3764,
+      "mean_token_accuracy": 0.8543789386749268,
+      "num_tokens": 467641.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.123171079158783,
+      "epoch": 0.3434739941118744,
+      "grad_norm": 2.6087253093719482,
+      "learning_rate": 9.603721049672678e-05,
+      "loss": 0.4134,
+      "mean_token_accuracy": 0.8483529090881348,
+      "num_tokens": 481724.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.0767368912696837,
+      "epoch": 0.35328753680078506,
+      "grad_norm": 4.402968406677246,
+      "learning_rate": 9.571459681100247e-05,
+      "loss": 0.3498,
+      "mean_token_accuracy": 0.8675245821475983,
+      "num_tokens": 495345.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.0896361708641051,
+      "epoch": 0.3631010794896958,
+      "grad_norm": 2.8933703899383545,
+      "learning_rate": 9.537994567785567e-05,
+      "loss": 0.382,
+      "mean_token_accuracy": 0.8591719806194306,
+      "num_tokens": 509800.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.113569176197052,
+      "epoch": 0.3729146221786065,
+      "grad_norm": 3.6864869594573975,
+      "learning_rate": 9.50333452167427e-05,
+      "loss": 0.3658,
+      "mean_token_accuracy": 0.8637172162532807,
+      "num_tokens": 523514.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.1054134368896484,
+      "epoch": 0.38272816486751715,
+      "grad_norm": 3.0641653537750244,
+      "learning_rate": 9.467488669358562e-05,
+      "loss": 0.3771,
+      "mean_token_accuracy": 0.857007771730423,
+      "num_tokens": 537693.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.085007917881012,
+      "epoch": 0.39254170755642787,
+      "grad_norm": 2.9690258502960205,
+      "learning_rate": 9.430466449674014e-05,
+      "loss": 0.3629,
+      "mean_token_accuracy": 0.860488897562027,
+      "num_tokens": 551457.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.0747876644134522,
+      "epoch": 0.4023552502453386,
+      "grad_norm": 3.8149168491363525,
+      "learning_rate": 9.39227761121416e-05,
+      "loss": 0.3514,
+      "mean_token_accuracy": 0.8680752813816071,
+      "num_tokens": 565157.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.069895577430725,
+      "epoch": 0.41216879293424924,
+      "grad_norm": 2.5904958248138428,
+      "learning_rate": 9.352932209763523e-05,
+      "loss": 0.3477,
+      "mean_token_accuracy": 0.8666971087455749,
+      "num_tokens": 579080.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.078192412853241,
+      "epoch": 0.42198233562315995,
+      "grad_norm": 3.276215076446533,
+      "learning_rate": 9.312440605649753e-05,
+      "loss": 0.3572,
+      "mean_token_accuracy": 0.8601800084114075,
+      "num_tokens": 592922.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.1033652901649476,
+      "epoch": 0.43179587831207067,
+      "grad_norm": 2.9685730934143066,
+      "learning_rate": 9.270813461015571e-05,
+      "loss": 0.3146,
+      "mean_token_accuracy": 0.8784527659416199,
+      "num_tokens": 607114.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.124951422214508,
+      "epoch": 0.44160942100098133,
+      "grad_norm": 4.074221134185791,
+      "learning_rate": 9.228061737011235e-05,
+      "loss": 0.3569,
+      "mean_token_accuracy": 0.8646153151988983,
+      "num_tokens": 620907.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.1364678740501404,
+      "epoch": 0.45142296368989204,
+      "grad_norm": 3.1165435314178467,
+      "learning_rate": 9.184196690908298e-05,
+      "loss": 0.3257,
+      "mean_token_accuracy": 0.8731623351573944,
+      "num_tokens": 635363.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.127057385444641,
+      "epoch": 0.46123650637880276,
+      "grad_norm": 2.911278247833252,
+      "learning_rate": 9.139229873135349e-05,
+      "loss": 0.3629,
+      "mean_token_accuracy": 0.8579927206039428,
+      "num_tokens": 648200.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.1423535346984863,
+      "epoch": 0.47105004906771347,
+      "grad_norm": 3.9557158946990967,
+      "learning_rate": 9.0931731242366e-05,
+      "loss": 0.363,
+      "mean_token_accuracy": 0.8673433363437653,
+      "num_tokens": 662920.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.120983636379242,
+      "epoch": 0.48086359175662413,
+      "grad_norm": 3.1872498989105225,
+      "learning_rate": 9.046038571754066e-05,
+      "loss": 0.3447,
+      "mean_token_accuracy": 0.8698362648487091,
+      "num_tokens": 676646.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.1252738237380981,
+      "epoch": 0.49067713444553485,
+      "grad_norm": 3.3474233150482178,
+      "learning_rate": 8.997838627034158e-05,
+      "loss": 0.343,
+      "mean_token_accuracy": 0.8696148753166199,
+      "num_tokens": 690435.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.117265784740448,
+      "epoch": 0.5004906771344455,
+      "grad_norm": 3.789865255355835,
+      "learning_rate": 8.94858598195958e-05,
+      "loss": 0.3315,
+      "mean_token_accuracy": 0.8725071847438812,
+      "num_tokens": 704664.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.1162002086639404,
+      "epoch": 0.5103042198233563,
+      "grad_norm": 3.4941630363464355,
+      "learning_rate": 8.898293605607316e-05,
+      "loss": 0.3464,
+      "mean_token_accuracy": 0.8651094734668732,
+      "num_tokens": 718834.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.1464263200759888,
+      "epoch": 0.5201177625122669,
+      "grad_norm": 3.3112964630126953,
+      "learning_rate": 8.846974740833659e-05,
+      "loss": 0.3329,
+      "mean_token_accuracy": 0.8682122528553009,
+      "num_tokens": 733172.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.1591009616851806,
+      "epoch": 0.5299313052011776,
+      "grad_norm": 4.015714168548584,
+      "learning_rate": 8.794642900787137e-05,
+      "loss": 0.342,
+      "mean_token_accuracy": 0.8695684254169465,
+      "num_tokens": 746577.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.1568204402923583,
+      "epoch": 0.5397448478900884,
+      "grad_norm": 2.9419708251953125,
+      "learning_rate": 8.741311865350259e-05,
+      "loss": 0.3199,
+      "mean_token_accuracy": 0.8726010918617249,
+      "num_tokens": 760378.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.1839653372764587,
+      "epoch": 0.549558390578999,
+      "grad_norm": 2.7937989234924316,
+      "learning_rate": 8.68699567751104e-05,
+      "loss": 0.347,
+      "mean_token_accuracy": 0.8683846890926361,
+      "num_tokens": 774528.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.121442937850952,
+      "epoch": 0.5593719332679097,
+      "grad_norm": 2.7515268325805664,
+      "learning_rate": 8.631708639665234e-05,
+      "loss": 0.3028,
+      "mean_token_accuracy": 0.8834416568279266,
+      "num_tokens": 787978.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.1337892174720765,
+      "epoch": 0.5691854759568205,
+      "grad_norm": 2.901580810546875,
+      "learning_rate": 8.575465309850265e-05,
+      "loss": 0.3234,
+      "mean_token_accuracy": 0.8734434902667999,
+      "num_tokens": 802122.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.1079321384429932,
+      "epoch": 0.5789990186457311,
+      "grad_norm": 3.0054614543914795,
+      "learning_rate": 8.518280497911841e-05,
+      "loss": 0.2889,
+      "mean_token_accuracy": 0.8868480741977691,
+      "num_tokens": 816567.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.1177525281906129,
+      "epoch": 0.5888125613346418,
+      "grad_norm": 2.9159741401672363,
+      "learning_rate": 8.460169261604274e-05,
+      "loss": 0.2627,
+      "mean_token_accuracy": 0.8969064891338349,
+      "num_tokens": 830550.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.0899678707122802,
+      "epoch": 0.5986261040235525,
+      "grad_norm": 2.796565532684326,
+      "learning_rate": 8.4011469026255e-05,
+      "loss": 0.2916,
+      "mean_token_accuracy": 0.8907618165016175,
+      "num_tokens": 843779.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.127509605884552,
+      "epoch": 0.6084396467124632,
+      "grad_norm": 3.322627067565918,
+      "learning_rate": 8.341228962587882e-05,
+      "loss": 0.3079,
+      "mean_token_accuracy": 0.8839374780654907,
+      "num_tokens": 857498.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.110354471206665,
+      "epoch": 0.6182531894013739,
+      "grad_norm": 4.072887897491455,
+      "learning_rate": 8.28043121892583e-05,
+      "loss": 0.3085,
+      "mean_token_accuracy": 0.8839384853839874,
+      "num_tokens": 871163.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.127149760723114,
+      "epoch": 0.6280667320902846,
+      "grad_norm": 2.956531047821045,
+      "learning_rate": 8.218769680741336e-05,
+      "loss": 0.321,
+      "mean_token_accuracy": 0.8798376202583313,
+      "num_tokens": 884912.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.135562038421631,
+      "epoch": 0.6378802747791953,
+      "grad_norm": 3.381680488586426,
+      "learning_rate": 8.156260584588491e-05,
+      "loss": 0.2935,
+      "mean_token_accuracy": 0.8831522583961486,
+      "num_tokens": 898102.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.1273825883865356,
+      "epoch": 0.647693817468106,
+      "grad_norm": 3.5705130100250244,
+      "learning_rate": 8.09292039019812e-05,
+      "loss": 0.3075,
+      "mean_token_accuracy": 0.8782266080379486,
+      "num_tokens": 911902.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.13209570646286,
+      "epoch": 0.6575073601570167,
+      "grad_norm": 3.049805164337158,
+      "learning_rate": 8.028765776143654e-05,
+      "loss": 0.3487,
+      "mean_token_accuracy": 0.8587038159370423,
+      "num_tokens": 926137.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.086474883556366,
+      "epoch": 0.6673209028459274,
+      "grad_norm": 2.941756010055542,
+      "learning_rate": 7.963813635449367e-05,
+      "loss": 0.2744,
+      "mean_token_accuracy": 0.8930023849010468,
+      "num_tokens": 940507.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.0713022947311401,
+      "epoch": 0.677134445534838,
+      "grad_norm": 2.3691322803497314,
+      "learning_rate": 7.898081071142149e-05,
+      "loss": 0.2916,
+      "mean_token_accuracy": 0.8863011419773101,
+      "num_tokens": 954934.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.092538344860077,
+      "epoch": 0.6869479882237488,
+      "grad_norm": 3.4685745239257812,
+      "learning_rate": 7.831585391747984e-05,
+      "loss": 0.3292,
+      "mean_token_accuracy": 0.8706068992614746,
+      "num_tokens": 968760.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.0920560836791993,
+      "epoch": 0.6967615309126595,
+      "grad_norm": 3.920772075653076,
+      "learning_rate": 7.764344106734319e-05,
+      "loss": 0.2775,
+      "mean_token_accuracy": 0.8995032429695129,
+      "num_tokens": 982644.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.0719207763671874,
+      "epoch": 0.7065750736015701,
+      "grad_norm": 2.648322820663452,
+      "learning_rate": 7.696374921899506e-05,
+      "loss": 0.3085,
+      "mean_token_accuracy": 0.8808045208454132,
+      "num_tokens": 996817.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.0798989892005921,
+      "epoch": 0.7163886162904809,
+      "grad_norm": 3.3815035820007324,
+      "learning_rate": 7.627695734710564e-05,
+      "loss": 0.2802,
+      "mean_token_accuracy": 0.8970337688922883,
+      "num_tokens": 1010008.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.085748815536499,
+      "epoch": 0.7262021589793916,
+      "grad_norm": 2.3341593742370605,
+      "learning_rate": 7.55832462959047e-05,
+      "loss": 0.2742,
+      "mean_token_accuracy": 0.89001345038414,
+      "num_tokens": 1023812.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.0867698907852172,
+      "epoch": 0.7360157016683022,
+      "grad_norm": 3.588059902191162,
+      "learning_rate": 7.488279873156207e-05,
+      "loss": 0.3112,
+      "mean_token_accuracy": 0.8772607207298279,
+      "num_tokens": 1037399.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.068379855155945,
+      "epoch": 0.745829244357213,
+      "grad_norm": 3.4026715755462646,
+      "learning_rate": 7.417579909408851e-05,
+      "loss": 0.257,
+      "mean_token_accuracy": 0.9023269295692444,
+      "num_tokens": 1050972.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.0807143092155456,
+      "epoch": 0.7556427870461236,
+      "grad_norm": 3.4668641090393066,
+      "learning_rate": 7.346243354876944e-05,
+      "loss": 0.3095,
+      "mean_token_accuracy": 0.8794282495975494,
+      "num_tokens": 1064581.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.0881330728530885,
+      "epoch": 0.7654563297350343,
+      "grad_norm": 2.812978506088257,
+      "learning_rate": 7.274288993714446e-05,
+      "loss": 0.277,
+      "mean_token_accuracy": 0.8947223246097564,
+      "num_tokens": 1078433.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.0977701663970947,
+      "epoch": 0.7752698724239451,
+      "grad_norm": 3.5759475231170654,
+      "learning_rate": 7.201735772754523e-05,
+      "loss": 0.2749,
+      "mean_token_accuracy": 0.8955462992191314,
+      "num_tokens": 1091839.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.0936875224113465,
+      "epoch": 0.7850834151128557,
+      "grad_norm": 3.2217183113098145,
+      "learning_rate": 7.128602796520538e-05,
+      "loss": 0.2747,
+      "mean_token_accuracy": 0.8899407267570496,
+      "num_tokens": 1105747.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.0896982192993163,
+      "epoch": 0.7948969578017664,
+      "grad_norm": 3.649219274520874,
+      "learning_rate": 7.054909322195485e-05,
+      "loss": 0.2908,
+      "mean_token_accuracy": 0.8878142654895782,
+      "num_tokens": 1119344.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.086609148979187,
+      "epoch": 0.8047105004906772,
+      "grad_norm": 3.001368761062622,
+      "learning_rate": 6.980674754551239e-05,
+      "loss": 0.2522,
+      "mean_token_accuracy": 0.8990485191345214,
+      "num_tokens": 1133252.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.0967908382415772,
+      "epoch": 0.8145240431795878,
+      "grad_norm": 3.8427577018737793,
+      "learning_rate": 6.905918640838946e-05,
+      "loss": 0.269,
+      "mean_token_accuracy": 0.8964259207248688,
+      "num_tokens": 1147368.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.0952121138572692,
+      "epoch": 0.8243375858684985,
+      "grad_norm": 3.0777549743652344,
+      "learning_rate": 6.830660665641898e-05,
+      "loss": 0.2701,
+      "mean_token_accuracy": 0.8952491879463196,
+      "num_tokens": 1161165.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.1037273406982422,
+      "epoch": 0.8341511285574092,
+      "grad_norm": 3.185467004776001,
+      "learning_rate": 6.754920645692222e-05,
+      "loss": 0.2737,
+      "mean_token_accuracy": 0.8947721242904663,
+      "num_tokens": 1174814.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.1269787549972534,
+      "epoch": 0.8439646712463199,
+      "grad_norm": 3.0934269428253174,
+      "learning_rate": 6.67871852465281e-05,
+      "loss": 0.2973,
+      "mean_token_accuracy": 0.8828479051589966,
+      "num_tokens": 1188967.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.102275288105011,
+      "epoch": 0.8537782139352306,
+      "grad_norm": 3.4718751907348633,
+      "learning_rate": 6.602074367865794e-05,
+      "loss": 0.2669,
+      "mean_token_accuracy": 0.8917194426059722,
+      "num_tokens": 1202907.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.0941867709159852,
+      "epoch": 0.8635917566241413,
+      "grad_norm": 3.092135190963745,
+      "learning_rate": 6.525008357069e-05,
+      "loss": 0.2468,
+      "mean_token_accuracy": 0.9026793301105499,
+      "num_tokens": 1216865.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.132461714744568,
+      "epoch": 0.873405299313052,
+      "grad_norm": 4.15745735168457,
+      "learning_rate": 6.447540785081738e-05,
+      "loss": 0.2651,
+      "mean_token_accuracy": 0.8965477287769318,
+      "num_tokens": 1230849.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.1045468330383301,
+      "epoch": 0.8832188420019627,
+      "grad_norm": 3.167222023010254,
+      "learning_rate": 6.36969205046136e-05,
+      "loss": 0.2618,
+      "mean_token_accuracy": 0.8960905253887177,
+      "num_tokens": 1243965.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.1170963168144226,
+      "epoch": 0.8930323846908734,
+      "grad_norm": 4.013219833374023,
+      "learning_rate": 6.291482652131955e-05,
+      "loss": 0.2561,
+      "mean_token_accuracy": 0.903102719783783,
+      "num_tokens": 1257918.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.0991609930992126,
+      "epoch": 0.9028459273797841,
+      "grad_norm": 2.8760578632354736,
+      "learning_rate": 6.212933183986634e-05,
+      "loss": 0.282,
+      "mean_token_accuracy": 0.8955612778663635,
+      "num_tokens": 1271980.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.0992092967033387,
+      "epoch": 0.9126594700686947,
+      "grad_norm": 3.1269609928131104,
+      "learning_rate": 6.1340643294648e-05,
+      "loss": 0.2516,
+      "mean_token_accuracy": 0.8949107646942138,
+      "num_tokens": 1286287.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.1045418500900268,
+      "epoch": 0.9224730127576055,
+      "grad_norm": 2.7814784049987793,
+      "learning_rate": 6.054896856105834e-05,
+      "loss": 0.2541,
+      "mean_token_accuracy": 0.9065688014030456,
+      "num_tokens": 1300381.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.0794389605522157,
+      "epoch": 0.9322865554465162,
+      "grad_norm": 3.6291708946228027,
+      "learning_rate": 5.9754516100806423e-05,
+      "loss": 0.2268,
+      "mean_token_accuracy": 0.9080643057823181,
+      "num_tokens": 1314086.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.0879417300224303,
+      "epoch": 0.9421000981354269,
+      "grad_norm": 3.9489541053771973,
+      "learning_rate": 5.895749510702502e-05,
+      "loss": 0.2445,
+      "mean_token_accuracy": 0.9006594896316529,
+      "num_tokens": 1328434.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.075455629825592,
+      "epoch": 0.9519136408243376,
+      "grad_norm": 2.783209800720215,
+      "learning_rate": 5.815811544918641e-05,
+      "loss": 0.2564,
+      "mean_token_accuracy": 0.8981324911117554,
+      "num_tokens": 1341971.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.0905417919158935,
+      "epoch": 0.9617271835132483,
+      "grad_norm": 3.5020217895507812,
+      "learning_rate": 5.7356587617840065e-05,
+      "loss": 0.2243,
+      "mean_token_accuracy": 0.9165168702602386,
+      "num_tokens": 1356685.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.0756125569343566,
+      "epoch": 0.971540726202159,
+      "grad_norm": 4.463494300842285,
+      "learning_rate": 5.655312266918684e-05,
+      "loss": 0.2636,
+      "mean_token_accuracy": 0.8929223239421844,
+      "num_tokens": 1370541.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.0586444735527039,
+      "epoch": 0.9813542688910697,
+      "grad_norm": 3.7100353240966797,
+      "learning_rate": 5.574793216950428e-05,
+      "loss": 0.2178,
+      "mean_token_accuracy": 0.9136503040790558,
+      "num_tokens": 1383962.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.0636729836463927,
+      "epoch": 0.9911678115799804,
+      "grad_norm": 3.6245152950286865,
+      "learning_rate": 5.494122813943742e-05,
+      "loss": 0.2195,
+      "mean_token_accuracy": 0.9131732642650604,
+      "num_tokens": 1397902.0,
+      "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_adam_rot_e2_lr1e-4_s1_10k/checkpoint-1019/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34519d7b2f8033bed6a7a7b759e9a16c47875244d1c7a125c0bd74989108f2c7
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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": [
+    "down_proj",
+    "q_proj",
+    "gate_proj",
+    "v_proj",
+    "k_proj",
+    "up_proj",
+    "o_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_adam_rot_e2_lr1e-4_s1_10k/checkpoint-2038/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ec4bb7a667470efe3a8af9aa749980a1afab414af0055476843211615566d4a
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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": 1.2497499465942383,
+      "epoch": 0.009813542688910697,
+      "grad_norm": 3.531019926071167,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.8102,
+      "mean_token_accuracy": 0.7294524252414704,
+      "num_tokens": 13663.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.2468776941299438,
+      "epoch": 0.019627085377821395,
+      "grad_norm": 2.9747965335845947,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.6859,
+      "mean_token_accuracy": 0.7676334083080292,
+      "num_tokens": 27321.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.277602195739746,
+      "epoch": 0.029440628066732092,
+      "grad_norm": 2.3711931705474854,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.6728,
+      "mean_token_accuracy": 0.7631717622280121,
+      "num_tokens": 41194.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.2442076921463012,
+      "epoch": 0.03925417075564279,
+      "grad_norm": 2.084287643432617,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.6408,
+      "mean_token_accuracy": 0.7595057129859925,
+      "num_tokens": 54431.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.2084601402282715,
+      "epoch": 0.04906771344455348,
+      "grad_norm": 2.107304334640503,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.5483,
+      "mean_token_accuracy": 0.797372704744339,
+      "num_tokens": 68329.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.1884977340698242,
+      "epoch": 0.058881256133464184,
+      "grad_norm": 2.411963701248169,
+      "learning_rate": 5.784313725490197e-05,
+      "loss": 0.5466,
+      "mean_token_accuracy": 0.797961276769638,
+      "num_tokens": 81724.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.1674535632133485,
+      "epoch": 0.06869479882237488,
+      "grad_norm": 2.5295956134796143,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 0.5191,
+      "mean_token_accuracy": 0.8120699107646943,
+      "num_tokens": 95693.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.1617870569229125,
+      "epoch": 0.07850834151128558,
+      "grad_norm": 3.007242202758789,
+      "learning_rate": 7.745098039215687e-05,
+      "loss": 0.4793,
+      "mean_token_accuracy": 0.8148658454418183,
+      "num_tokens": 109253.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.1578378081321716,
+      "epoch": 0.08832188420019627,
+      "grad_norm": 2.878816604614258,
+      "learning_rate": 8.725490196078432e-05,
+      "loss": 0.5473,
+      "mean_token_accuracy": 0.7997951984405518,
+      "num_tokens": 122810.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.1196816921234132,
+      "epoch": 0.09813542688910697,
+      "grad_norm": 2.5607798099517822,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 0.4801,
+      "mean_token_accuracy": 0.8200828254222869,
+      "num_tokens": 136706.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.1181760668754577,
+      "epoch": 0.10794896957801767,
+      "grad_norm": 2.9833409786224365,
+      "learning_rate": 9.999677432611306e-05,
+      "loss": 0.442,
+      "mean_token_accuracy": 0.8319093763828278,
+      "num_tokens": 150742.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.152057695388794,
+      "epoch": 0.11776251226692837,
+      "grad_norm": 2.6250386238098145,
+      "learning_rate": 9.998097610899492e-05,
+      "loss": 0.5096,
+      "mean_token_accuracy": 0.8141644656658172,
+      "num_tokens": 164305.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.1269365668296814,
+      "epoch": 0.12757605495583907,
+      "grad_norm": 2.7124476432800293,
+      "learning_rate": 9.995201703267625e-05,
+      "loss": 0.4774,
+      "mean_token_accuracy": 0.8261253535747528,
+      "num_tokens": 177967.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.1464994549751282,
+      "epoch": 0.13738959764474976,
+      "grad_norm": 2.9309122562408447,
+      "learning_rate": 9.990990472258488e-05,
+      "loss": 0.4712,
+      "mean_token_accuracy": 0.8258848905563354,
+      "num_tokens": 192090.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.132555317878723,
+      "epoch": 0.14720314033366044,
+      "grad_norm": 2.3876519203186035,
+      "learning_rate": 9.985465026762356e-05,
+      "loss": 0.4768,
+      "mean_token_accuracy": 0.8226563394069671,
+      "num_tokens": 206673.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.0925392389297486,
+      "epoch": 0.15701668302257116,
+      "grad_norm": 4.019376754760742,
+      "learning_rate": 9.978626821725008e-05,
+      "loss": 0.4199,
+      "mean_token_accuracy": 0.8404703497886657,
+      "num_tokens": 220293.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.1034241437911987,
+      "epoch": 0.16683022571148184,
+      "grad_norm": 2.2521817684173584,
+      "learning_rate": 9.970477657764611e-05,
+      "loss": 0.4052,
+      "mean_token_accuracy": 0.8489298939704895,
+      "num_tokens": 234011.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.1063331246376038,
+      "epoch": 0.17664376840039253,
+      "grad_norm": 2.918395519256592,
+      "learning_rate": 9.961019680697593e-05,
+      "loss": 0.4519,
+      "mean_token_accuracy": 0.8295825779438019,
+      "num_tokens": 247118.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.12331782579422,
+      "epoch": 0.18645731108930325,
+      "grad_norm": 3.441845417022705,
+      "learning_rate": 9.950255380973604e-05,
+      "loss": 0.4331,
+      "mean_token_accuracy": 0.8383278727531434,
+      "num_tokens": 261365.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.0990667939186096,
+      "epoch": 0.19627085377821393,
+      "grad_norm": 2.470823287963867,
+      "learning_rate": 9.93818759301974e-05,
+      "loss": 0.4125,
+      "mean_token_accuracy": 0.8514919698238372,
+      "num_tokens": 274990.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.100917935371399,
+      "epoch": 0.20608439646712462,
+      "grad_norm": 3.3002994060516357,
+      "learning_rate": 9.924819494494194e-05,
+      "loss": 0.4167,
+      "mean_token_accuracy": 0.8451960921287537,
+      "num_tokens": 288512.0,
+      "step": 210
+    },
+    {
+      "entropy": 1.0906975269317627,
+      "epoch": 0.21589793915603533,
+      "grad_norm": 2.802976369857788,
+      "learning_rate": 9.910154605449517e-05,
+      "loss": 0.4001,
+      "mean_token_accuracy": 0.8403747737407684,
+      "num_tokens": 301852.0,
+      "step": 220
+    },
+    {
+      "entropy": 1.0727845072746276,
+      "epoch": 0.22571148184494602,
+      "grad_norm": 3.1608800888061523,
+      "learning_rate": 9.89419678740573e-05,
+      "loss": 0.3869,
+      "mean_token_accuracy": 0.852373194694519,
+      "num_tokens": 315850.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.090681266784668,
+      "epoch": 0.23552502453385674,
+      "grad_norm": 2.397611141204834,
+      "learning_rate": 9.876950242333518e-05,
+      "loss": 0.4244,
+      "mean_token_accuracy": 0.8342725813388825,
+      "num_tokens": 329271.0,
+      "step": 240
+    },
+    {
+      "entropy": 1.0980817317962646,
+      "epoch": 0.24533856722276742,
+      "grad_norm": 3.705782413482666,
+      "learning_rate": 9.858419511547777e-05,
+      "loss": 0.4036,
+      "mean_token_accuracy": 0.8450454592704773,
+      "num_tokens": 343233.0,
+      "step": 250
+    },
+    {
+      "entropy": 1.090275275707245,
+      "epoch": 0.25515210991167814,
+      "grad_norm": 2.955258846282959,
+      "learning_rate": 9.838609474511811e-05,
+      "loss": 0.3864,
+      "mean_token_accuracy": 0.8567544460296631,
+      "num_tokens": 356887.0,
+      "step": 260
+    },
+    {
+      "entropy": 1.0975828528404237,
+      "epoch": 0.2649656526005888,
+      "grad_norm": 3.384934186935425,
+      "learning_rate": 9.81752534755248e-05,
+      "loss": 0.3908,
+      "mean_token_accuracy": 0.8517731606960297,
+      "num_tokens": 370548.0,
+      "step": 270
+    },
+    {
+      "entropy": 1.1451343178749085,
+      "epoch": 0.2747791952894995,
+      "grad_norm": 3.3305108547210693,
+      "learning_rate": 9.795172682486655e-05,
+      "loss": 0.3955,
+      "mean_token_accuracy": 0.8470654547214508,
+      "num_tokens": 384687.0,
+      "step": 280
+    },
+    {
+      "entropy": 1.1502320528030396,
+      "epoch": 0.2845927379784102,
+      "grad_norm": 3.182323455810547,
+      "learning_rate": 9.77155736515932e-05,
+      "loss": 0.3966,
+      "mean_token_accuracy": 0.8542943000793457,
+      "num_tokens": 398358.0,
+      "step": 290
+    },
+    {
+      "entropy": 1.1510611295700073,
+      "epoch": 0.2944062806673209,
+      "grad_norm": 2.945194721221924,
+      "learning_rate": 9.746685613893735e-05,
+      "loss": 0.383,
+      "mean_token_accuracy": 0.8609679043292999,
+      "num_tokens": 412700.0,
+      "step": 300
+    },
+    {
+      "entropy": 1.146600604057312,
+      "epoch": 0.3042198233562316,
+      "grad_norm": 2.3007872104644775,
+      "learning_rate": 9.720563977854046e-05,
+      "loss": 0.3627,
+      "mean_token_accuracy": 0.8592406272888183,
+      "num_tokens": 427312.0,
+      "step": 310
+    },
+    {
+      "entropy": 1.1343496203422547,
+      "epoch": 0.3140333660451423,
+      "grad_norm": 2.9032654762268066,
+      "learning_rate": 9.693199335320762e-05,
+      "loss": 0.3696,
+      "mean_token_accuracy": 0.8608040511608124,
+      "num_tokens": 441507.0,
+      "step": 320
+    },
+    {
+      "entropy": 1.1328219652175904,
+      "epoch": 0.323846908734053,
+      "grad_norm": 4.016798496246338,
+      "learning_rate": 9.664598891879597e-05,
+      "loss": 0.395,
+      "mean_token_accuracy": 0.8488866150379181,
+      "num_tokens": 454411.0,
+      "step": 330
+    },
+    {
+      "entropy": 1.1116938591003418,
+      "epoch": 0.3336604514229637,
+      "grad_norm": 2.705829620361328,
+      "learning_rate": 9.634770178524116e-05,
+      "loss": 0.3764,
+      "mean_token_accuracy": 0.8543789386749268,
+      "num_tokens": 467641.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.123171079158783,
+      "epoch": 0.3434739941118744,
+      "grad_norm": 2.6087253093719482,
+      "learning_rate": 9.603721049672678e-05,
+      "loss": 0.4134,
+      "mean_token_accuracy": 0.8483529090881348,
+      "num_tokens": 481724.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.0767368912696837,
+      "epoch": 0.35328753680078506,
+      "grad_norm": 4.402968406677246,
+      "learning_rate": 9.571459681100247e-05,
+      "loss": 0.3498,
+      "mean_token_accuracy": 0.8675245821475983,
+      "num_tokens": 495345.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.0896361708641051,
+      "epoch": 0.3631010794896958,
+      "grad_norm": 2.8933703899383545,
+      "learning_rate": 9.537994567785567e-05,
+      "loss": 0.382,
+      "mean_token_accuracy": 0.8591719806194306,
+      "num_tokens": 509800.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.113569176197052,
+      "epoch": 0.3729146221786065,
+      "grad_norm": 3.6864869594573975,
+      "learning_rate": 9.50333452167427e-05,
+      "loss": 0.3658,
+      "mean_token_accuracy": 0.8637172162532807,
+      "num_tokens": 523514.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.1054134368896484,
+      "epoch": 0.38272816486751715,
+      "grad_norm": 3.0641653537750244,
+      "learning_rate": 9.467488669358562e-05,
+      "loss": 0.3771,
+      "mean_token_accuracy": 0.857007771730423,
+      "num_tokens": 537693.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.085007917881012,
+      "epoch": 0.39254170755642787,
+      "grad_norm": 2.9690258502960205,
+      "learning_rate": 9.430466449674014e-05,
+      "loss": 0.3629,
+      "mean_token_accuracy": 0.860488897562027,
+      "num_tokens": 551457.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.0747876644134522,
+      "epoch": 0.4023552502453386,
+      "grad_norm": 3.8149168491363525,
+      "learning_rate": 9.39227761121416e-05,
+      "loss": 0.3514,
+      "mean_token_accuracy": 0.8680752813816071,
+      "num_tokens": 565157.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.069895577430725,
+      "epoch": 0.41216879293424924,
+      "grad_norm": 2.5904958248138428,
+      "learning_rate": 9.352932209763523e-05,
+      "loss": 0.3477,
+      "mean_token_accuracy": 0.8666971087455749,
+      "num_tokens": 579080.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.078192412853241,
+      "epoch": 0.42198233562315995,
+      "grad_norm": 3.276215076446533,
+      "learning_rate": 9.312440605649753e-05,
+      "loss": 0.3572,
+      "mean_token_accuracy": 0.8601800084114075,
+      "num_tokens": 592922.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.1033652901649476,
+      "epoch": 0.43179587831207067,
+      "grad_norm": 2.9685730934143066,
+      "learning_rate": 9.270813461015571e-05,
+      "loss": 0.3146,
+      "mean_token_accuracy": 0.8784527659416199,
+      "num_tokens": 607114.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.124951422214508,
+      "epoch": 0.44160942100098133,
+      "grad_norm": 4.074221134185791,
+      "learning_rate": 9.228061737011235e-05,
+      "loss": 0.3569,
+      "mean_token_accuracy": 0.8646153151988983,
+      "num_tokens": 620907.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.1364678740501404,
+      "epoch": 0.45142296368989204,
+      "grad_norm": 3.1165435314178467,
+      "learning_rate": 9.184196690908298e-05,
+      "loss": 0.3257,
+      "mean_token_accuracy": 0.8731623351573944,
+      "num_tokens": 635363.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.127057385444641,
+      "epoch": 0.46123650637880276,
+      "grad_norm": 2.911278247833252,
+      "learning_rate": 9.139229873135349e-05,
+      "loss": 0.3629,
+      "mean_token_accuracy": 0.8579927206039428,
+      "num_tokens": 648200.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.1423535346984863,
+      "epoch": 0.47105004906771347,
+      "grad_norm": 3.9557158946990967,
+      "learning_rate": 9.0931731242366e-05,
+      "loss": 0.363,
+      "mean_token_accuracy": 0.8673433363437653,
+      "num_tokens": 662920.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.120983636379242,
+      "epoch": 0.48086359175662413,
+      "grad_norm": 3.1872498989105225,
+      "learning_rate": 9.046038571754066e-05,
+      "loss": 0.3447,
+      "mean_token_accuracy": 0.8698362648487091,
+      "num_tokens": 676646.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.1252738237380981,
+      "epoch": 0.49067713444553485,
+      "grad_norm": 3.3474233150482178,
+      "learning_rate": 8.997838627034158e-05,
+      "loss": 0.343,
+      "mean_token_accuracy": 0.8696148753166199,
+      "num_tokens": 690435.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.117265784740448,
+      "epoch": 0.5004906771344455,
+      "grad_norm": 3.789865255355835,
+      "learning_rate": 8.94858598195958e-05,
+      "loss": 0.3315,
+      "mean_token_accuracy": 0.8725071847438812,
+      "num_tokens": 704664.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.1162002086639404,
+      "epoch": 0.5103042198233563,
+      "grad_norm": 3.4941630363464355,
+      "learning_rate": 8.898293605607316e-05,
+      "loss": 0.3464,
+      "mean_token_accuracy": 0.8651094734668732,
+      "num_tokens": 718834.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.1464263200759888,
+      "epoch": 0.5201177625122669,
+      "grad_norm": 3.3112964630126953,
+      "learning_rate": 8.846974740833659e-05,
+      "loss": 0.3329,
+      "mean_token_accuracy": 0.8682122528553009,
+      "num_tokens": 733172.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.1591009616851806,
+      "epoch": 0.5299313052011776,
+      "grad_norm": 4.015714168548584,
+      "learning_rate": 8.794642900787137e-05,
+      "loss": 0.342,
+      "mean_token_accuracy": 0.8695684254169465,
+      "num_tokens": 746577.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.1568204402923583,
+      "epoch": 0.5397448478900884,
+      "grad_norm": 2.9419708251953125,
+      "learning_rate": 8.741311865350259e-05,
+      "loss": 0.3199,
+      "mean_token_accuracy": 0.8726010918617249,
+      "num_tokens": 760378.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.1839653372764587,
+      "epoch": 0.549558390578999,
+      "grad_norm": 2.7937989234924316,
+      "learning_rate": 8.68699567751104e-05,
+      "loss": 0.347,
+      "mean_token_accuracy": 0.8683846890926361,
+      "num_tokens": 774528.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.121442937850952,
+      "epoch": 0.5593719332679097,
+      "grad_norm": 2.7515268325805664,
+      "learning_rate": 8.631708639665234e-05,
+      "loss": 0.3028,
+      "mean_token_accuracy": 0.8834416568279266,
+      "num_tokens": 787978.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.1337892174720765,
+      "epoch": 0.5691854759568205,
+      "grad_norm": 2.901580810546875,
+      "learning_rate": 8.575465309850265e-05,
+      "loss": 0.3234,
+      "mean_token_accuracy": 0.8734434902667999,
+      "num_tokens": 802122.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.1079321384429932,
+      "epoch": 0.5789990186457311,
+      "grad_norm": 3.0054614543914795,
+      "learning_rate": 8.518280497911841e-05,
+      "loss": 0.2889,
+      "mean_token_accuracy": 0.8868480741977691,
+      "num_tokens": 816567.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.1177525281906129,
+      "epoch": 0.5888125613346418,
+      "grad_norm": 2.9159741401672363,
+      "learning_rate": 8.460169261604274e-05,
+      "loss": 0.2627,
+      "mean_token_accuracy": 0.8969064891338349,
+      "num_tokens": 830550.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.0899678707122802,
+      "epoch": 0.5986261040235525,
+      "grad_norm": 2.796565532684326,
+      "learning_rate": 8.4011469026255e-05,
+      "loss": 0.2916,
+      "mean_token_accuracy": 0.8907618165016175,
+      "num_tokens": 843779.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.127509605884552,
+      "epoch": 0.6084396467124632,
+      "grad_norm": 3.322627067565918,
+      "learning_rate": 8.341228962587882e-05,
+      "loss": 0.3079,
+      "mean_token_accuracy": 0.8839374780654907,
+      "num_tokens": 857498.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.110354471206665,
+      "epoch": 0.6182531894013739,
+      "grad_norm": 4.072887897491455,
+      "learning_rate": 8.28043121892583e-05,
+      "loss": 0.3085,
+      "mean_token_accuracy": 0.8839384853839874,
+      "num_tokens": 871163.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.127149760723114,
+      "epoch": 0.6280667320902846,
+      "grad_norm": 2.956531047821045,
+      "learning_rate": 8.218769680741336e-05,
+      "loss": 0.321,
+      "mean_token_accuracy": 0.8798376202583313,
+      "num_tokens": 884912.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.135562038421631,
+      "epoch": 0.6378802747791953,
+      "grad_norm": 3.381680488586426,
+      "learning_rate": 8.156260584588491e-05,
+      "loss": 0.2935,
+      "mean_token_accuracy": 0.8831522583961486,
+      "num_tokens": 898102.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.1273825883865356,
+      "epoch": 0.647693817468106,
+      "grad_norm": 3.5705130100250244,
+      "learning_rate": 8.09292039019812e-05,
+      "loss": 0.3075,
+      "mean_token_accuracy": 0.8782266080379486,
+      "num_tokens": 911902.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.13209570646286,
+      "epoch": 0.6575073601570167,
+      "grad_norm": 3.049805164337158,
+      "learning_rate": 8.028765776143654e-05,
+      "loss": 0.3487,
+      "mean_token_accuracy": 0.8587038159370423,
+      "num_tokens": 926137.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.086474883556366,
+      "epoch": 0.6673209028459274,
+      "grad_norm": 2.941756010055542,
+      "learning_rate": 7.963813635449367e-05,
+      "loss": 0.2744,
+      "mean_token_accuracy": 0.8930023849010468,
+      "num_tokens": 940507.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.0713022947311401,
+      "epoch": 0.677134445534838,
+      "grad_norm": 2.3691322803497314,
+      "learning_rate": 7.898081071142149e-05,
+      "loss": 0.2916,
+      "mean_token_accuracy": 0.8863011419773101,
+      "num_tokens": 954934.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.092538344860077,
+      "epoch": 0.6869479882237488,
+      "grad_norm": 3.4685745239257812,
+      "learning_rate": 7.831585391747984e-05,
+      "loss": 0.3292,
+      "mean_token_accuracy": 0.8706068992614746,
+      "num_tokens": 968760.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.0920560836791993,
+      "epoch": 0.6967615309126595,
+      "grad_norm": 3.920772075653076,
+      "learning_rate": 7.764344106734319e-05,
+      "loss": 0.2775,
+      "mean_token_accuracy": 0.8995032429695129,
+      "num_tokens": 982644.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.0719207763671874,
+      "epoch": 0.7065750736015701,
+      "grad_norm": 2.648322820663452,
+      "learning_rate": 7.696374921899506e-05,
+      "loss": 0.3085,
+      "mean_token_accuracy": 0.8808045208454132,
+      "num_tokens": 996817.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.0798989892005921,
+      "epoch": 0.7163886162904809,
+      "grad_norm": 3.3815035820007324,
+      "learning_rate": 7.627695734710564e-05,
+      "loss": 0.2802,
+      "mean_token_accuracy": 0.8970337688922883,
+      "num_tokens": 1010008.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.085748815536499,
+      "epoch": 0.7262021589793916,
+      "grad_norm": 2.3341593742370605,
+      "learning_rate": 7.55832462959047e-05,
+      "loss": 0.2742,
+      "mean_token_accuracy": 0.89001345038414,
+      "num_tokens": 1023812.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.0867698907852172,
+      "epoch": 0.7360157016683022,
+      "grad_norm": 3.588059902191162,
+      "learning_rate": 7.488279873156207e-05,
+      "loss": 0.3112,
+      "mean_token_accuracy": 0.8772607207298279,
+      "num_tokens": 1037399.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.068379855155945,
+      "epoch": 0.745829244357213,
+      "grad_norm": 3.4026715755462646,
+      "learning_rate": 7.417579909408851e-05,
+      "loss": 0.257,
+      "mean_token_accuracy": 0.9023269295692444,
+      "num_tokens": 1050972.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.0807143092155456,
+      "epoch": 0.7556427870461236,
+      "grad_norm": 3.4668641090393066,
+      "learning_rate": 7.346243354876944e-05,
+      "loss": 0.3095,
+      "mean_token_accuracy": 0.8794282495975494,
+      "num_tokens": 1064581.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.0881330728530885,
+      "epoch": 0.7654563297350343,
+      "grad_norm": 2.812978506088257,
+      "learning_rate": 7.274288993714446e-05,
+      "loss": 0.277,
+      "mean_token_accuracy": 0.8947223246097564,
+      "num_tokens": 1078433.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.0977701663970947,
+      "epoch": 0.7752698724239451,
+      "grad_norm": 3.5759475231170654,
+      "learning_rate": 7.201735772754523e-05,
+      "loss": 0.2749,
+      "mean_token_accuracy": 0.8955462992191314,
+      "num_tokens": 1091839.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.0936875224113465,
+      "epoch": 0.7850834151128557,
+      "grad_norm": 3.2217183113098145,
+      "learning_rate": 7.128602796520538e-05,
+      "loss": 0.2747,
+      "mean_token_accuracy": 0.8899407267570496,
+      "num_tokens": 1105747.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.0896982192993163,
+      "epoch": 0.7948969578017664,
+      "grad_norm": 3.649219274520874,
+      "learning_rate": 7.054909322195485e-05,
+      "loss": 0.2908,
+      "mean_token_accuracy": 0.8878142654895782,
+      "num_tokens": 1119344.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.086609148979187,
+      "epoch": 0.8047105004906772,
+      "grad_norm": 3.001368761062622,
+      "learning_rate": 6.980674754551239e-05,
+      "loss": 0.2522,
+      "mean_token_accuracy": 0.8990485191345214,
+      "num_tokens": 1133252.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.0967908382415772,
+      "epoch": 0.8145240431795878,
+      "grad_norm": 3.8427577018737793,
+      "learning_rate": 6.905918640838946e-05,
+      "loss": 0.269,
+      "mean_token_accuracy": 0.8964259207248688,
+      "num_tokens": 1147368.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.0952121138572692,
+      "epoch": 0.8243375858684985,
+      "grad_norm": 3.0777549743652344,
+      "learning_rate": 6.830660665641898e-05,
+      "loss": 0.2701,
+      "mean_token_accuracy": 0.8952491879463196,
+      "num_tokens": 1161165.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.1037273406982422,
+      "epoch": 0.8341511285574092,
+      "grad_norm": 3.185467004776001,
+      "learning_rate": 6.754920645692222e-05,
+      "loss": 0.2737,
+      "mean_token_accuracy": 0.8947721242904663,
+      "num_tokens": 1174814.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.1269787549972534,
+      "epoch": 0.8439646712463199,
+      "grad_norm": 3.0934269428253174,
+      "learning_rate": 6.67871852465281e-05,
+      "loss": 0.2973,
+      "mean_token_accuracy": 0.8828479051589966,
+      "num_tokens": 1188967.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.102275288105011,
+      "epoch": 0.8537782139352306,
+      "grad_norm": 3.4718751907348633,
+      "learning_rate": 6.602074367865794e-05,
+      "loss": 0.2669,
+      "mean_token_accuracy": 0.8917194426059722,
+      "num_tokens": 1202907.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.0941867709159852,
+      "epoch": 0.8635917566241413,
+      "grad_norm": 3.092135190963745,
+      "learning_rate": 6.525008357069e-05,
+      "loss": 0.2468,
+      "mean_token_accuracy": 0.9026793301105499,
+      "num_tokens": 1216865.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.132461714744568,
+      "epoch": 0.873405299313052,
+      "grad_norm": 4.15745735168457,
+      "learning_rate": 6.447540785081738e-05,
+      "loss": 0.2651,
+      "mean_token_accuracy": 0.8965477287769318,
+      "num_tokens": 1230849.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.1045468330383301,
+      "epoch": 0.8832188420019627,
+      "grad_norm": 3.167222023010254,
+      "learning_rate": 6.36969205046136e-05,
+      "loss": 0.2618,
+      "mean_token_accuracy": 0.8960905253887177,
+      "num_tokens": 1243965.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.1170963168144226,
+      "epoch": 0.8930323846908734,
+      "grad_norm": 4.013219833374023,
+      "learning_rate": 6.291482652131955e-05,
+      "loss": 0.2561,
+      "mean_token_accuracy": 0.903102719783783,
+      "num_tokens": 1257918.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.0991609930992126,
+      "epoch": 0.9028459273797841,
+      "grad_norm": 2.8760578632354736,
+      "learning_rate": 6.212933183986634e-05,
+      "loss": 0.282,
+      "mean_token_accuracy": 0.8955612778663635,
+      "num_tokens": 1271980.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.0992092967033387,
+      "epoch": 0.9126594700686947,
+      "grad_norm": 3.1269609928131104,
+      "learning_rate": 6.1340643294648e-05,
+      "loss": 0.2516,
+      "mean_token_accuracy": 0.8949107646942138,
+      "num_tokens": 1286287.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.1045418500900268,
+      "epoch": 0.9224730127576055,
+      "grad_norm": 2.7814784049987793,
+      "learning_rate": 6.054896856105834e-05,
+      "loss": 0.2541,
+      "mean_token_accuracy": 0.9065688014030456,
+      "num_tokens": 1300381.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.0794389605522157,
+      "epoch": 0.9322865554465162,
+      "grad_norm": 3.6291708946228027,
+      "learning_rate": 5.9754516100806423e-05,
+      "loss": 0.2268,
+      "mean_token_accuracy": 0.9080643057823181,
+      "num_tokens": 1314086.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.0879417300224303,
+      "epoch": 0.9421000981354269,
+      "grad_norm": 3.9489541053771973,
+      "learning_rate": 5.895749510702502e-05,
+      "loss": 0.2445,
+      "mean_token_accuracy": 0.9006594896316529,
+      "num_tokens": 1328434.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.075455629825592,
+      "epoch": 0.9519136408243376,
+      "grad_norm": 2.783209800720215,
+      "learning_rate": 5.815811544918641e-05,
+      "loss": 0.2564,
+      "mean_token_accuracy": 0.8981324911117554,
+      "num_tokens": 1341971.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.0905417919158935,
+      "epoch": 0.9617271835132483,
+      "grad_norm": 3.5020217895507812,
+      "learning_rate": 5.7356587617840065e-05,
+      "loss": 0.2243,
+      "mean_token_accuracy": 0.9165168702602386,
+      "num_tokens": 1356685.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.0756125569343566,
+      "epoch": 0.971540726202159,
+      "grad_norm": 4.463494300842285,
+      "learning_rate": 5.655312266918684e-05,
+      "loss": 0.2636,
+      "mean_token_accuracy": 0.8929223239421844,
+      "num_tokens": 1370541.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.0586444735527039,
+      "epoch": 0.9813542688910697,
+      "grad_norm": 3.7100353240966797,
+      "learning_rate": 5.574793216950428e-05,
+      "loss": 0.2178,
+      "mean_token_accuracy": 0.9136503040790558,
+      "num_tokens": 1383962.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.0636729836463927,
+      "epoch": 0.9911678115799804,
+      "grad_norm": 3.6245152950286865,
+      "learning_rate": 5.494122813943742e-05,
+      "loss": 0.2195,
+      "mean_token_accuracy": 0.9131732642650604,
+      "num_tokens": 1397902.0,
+      "step": 1010
+    },
+    {
+      "entropy": 1.0615388035774231,
+      "epoch": 1.000981354268891,
+      "grad_norm": 3.2300093173980713,
+      "learning_rate": 5.413322299817019e-05,
+      "loss": 0.2217,
+      "mean_token_accuracy": 0.9138081192970275,
+      "num_tokens": 1410184.0,
+      "step": 1020
+    },
+    {
+      "entropy": 1.0607650756835938,
+      "epoch": 1.0107948969578018,
+      "grad_norm": 3.122152090072632,
+      "learning_rate": 5.332412950749169e-05,
+      "loss": 0.1724,
+      "mean_token_accuracy": 0.934057492017746,
+      "num_tokens": 1424018.0,
+      "step": 1030
+    },
+    {
+      "entropy": 1.0500118732452393,
+      "epoch": 1.0206084396467126,
+      "grad_norm": 3.1204495429992676,
+      "learning_rate": 5.251416071577232e-05,
+      "loss": 0.1723,
+      "mean_token_accuracy": 0.9335859954357147,
+      "num_tokens": 1438068.0,
+      "step": 1040
+    },
+    {
+      "entropy": 1.0313341200351716,
+      "epoch": 1.030421982335623,
+      "grad_norm": 4.092257499694824,
+      "learning_rate": 5.170352990186441e-05,
+      "loss": 0.1458,
+      "mean_token_accuracy": 0.9432224452495575,
+      "num_tokens": 1452019.0,
+      "step": 1050
+    },
+    {
+      "entropy": 0.9990754723548889,
+      "epoch": 1.0402355250245339,
+      "grad_norm": 3.9173905849456787,
+      "learning_rate": 5.08924505189423e-05,
+      "loss": 0.1628,
+      "mean_token_accuracy": 0.9342606604099274,
+      "num_tokens": 1465698.0,
+      "step": 1060
+    },
+    {
+      "entropy": 0.9989770531654358,
+      "epoch": 1.0500490677134446,
+      "grad_norm": 3.656033754348755,
+      "learning_rate": 5.008113613829631e-05,
+      "loss": 0.1631,
+      "mean_token_accuracy": 0.942461758852005,
+      "num_tokens": 1479631.0,
+      "step": 1070
+    },
+    {
+      "entropy": 1.0079047620296477,
+      "epoch": 1.0598626104023552,
+      "grad_norm": 3.705754518508911,
+      "learning_rate": 4.9269800393095747e-05,
+      "loss": 0.1698,
+      "mean_token_accuracy": 0.935236245393753,
+      "num_tokens": 1492799.0,
+      "step": 1080
+    },
+    {
+      "entropy": 1.0117216169834138,
+      "epoch": 1.069676153091266,
+      "grad_norm": 3.4877312183380127,
+      "learning_rate": 4.845865692213555e-05,
+      "loss": 0.1495,
+      "mean_token_accuracy": 0.9460038363933563,
+      "num_tokens": 1506531.0,
+      "step": 1090
+    },
+    {
+      "entropy": 0.980362343788147,
+      "epoch": 1.0794896957801767,
+      "grad_norm": 3.0575830936431885,
+      "learning_rate": 4.764791931358155e-05,
+      "loss": 0.1514,
+      "mean_token_accuracy": 0.9425157129764556,
+      "num_tokens": 1520031.0,
+      "step": 1100
+    },
+    {
+      "entropy": 0.9965842485427856,
+      "epoch": 1.0893032384690873,
+      "grad_norm": 2.9048209190368652,
+      "learning_rate": 4.683780104872894e-05,
+      "loss": 0.1531,
+      "mean_token_accuracy": 0.943388044834137,
+      "num_tokens": 1534436.0,
+      "step": 1110
+    },
+    {
+      "entropy": 1.004463440179825,
+      "epoch": 1.099116781157998,
+      "grad_norm": 2.8935232162475586,
+      "learning_rate": 4.6028515445788937e-05,
+      "loss": 0.1774,
+      "mean_token_accuracy": 0.9308792471885681,
+      "num_tokens": 1547962.0,
+      "step": 1120
+    },
+    {
+      "entropy": 0.9962540328502655,
+      "epoch": 1.1089303238469088,
+      "grad_norm": 3.2771453857421875,
+      "learning_rate": 4.5220275603718474e-05,
+      "loss": 0.1587,
+      "mean_token_accuracy": 0.9432451665401459,
+      "num_tokens": 1561849.0,
+      "step": 1130
+    },
+    {
+      "entropy": 0.9761953473091125,
+      "epoch": 1.1187438665358194,
+      "grad_norm": 2.8278708457946777,
+      "learning_rate": 4.4413294346107394e-05,
+      "loss": 0.1389,
+      "mean_token_accuracy": 0.9473001599311829,
+      "num_tokens": 1575503.0,
+      "step": 1140
+    },
+    {
+      "entropy": 0.9953079700469971,
+      "epoch": 1.1285574092247301,
+      "grad_norm": 3.4403743743896484,
+      "learning_rate": 4.3607784165138474e-05,
+      "loss": 0.1538,
+      "mean_token_accuracy": 0.9408088386058807,
+      "num_tokens": 1589893.0,
+      "step": 1150
+    },
+    {
+      "entropy": 1.0010611057281493,
+      "epoch": 1.138370951913641,
+      "grad_norm": 2.9534761905670166,
+      "learning_rate": 4.280395716563441e-05,
+      "loss": 0.1398,
+      "mean_token_accuracy": 0.947502213716507,
+      "num_tokens": 1603816.0,
+      "step": 1160
+    },
+    {
+      "entropy": 1.009485673904419,
+      "epoch": 1.1481844946025515,
+      "grad_norm": 4.362896919250488,
+      "learning_rate": 4.2002025009206994e-05,
+      "loss": 0.1576,
+      "mean_token_accuracy": 0.9420254051685333,
+      "num_tokens": 1617087.0,
+      "step": 1170
+    },
+    {
+      "entropy": 0.9967351436614991,
+      "epoch": 1.1579980372914622,
+      "grad_norm": 3.679607391357422,
+      "learning_rate": 4.120219885852292e-05,
+      "loss": 0.1498,
+      "mean_token_accuracy": 0.9419420063495636,
+      "num_tokens": 1631558.0,
+      "step": 1180
+    },
+    {
+      "entropy": 1.012783145904541,
+      "epoch": 1.167811579980373,
+      "grad_norm": 3.943161964416504,
+      "learning_rate": 4.040468932170097e-05,
+      "loss": 0.1522,
+      "mean_token_accuracy": 0.943919438123703,
+      "num_tokens": 1645944.0,
+      "step": 1190
+    },
+    {
+      "entropy": 1.0016049921512604,
+      "epoch": 1.1776251226692835,
+      "grad_norm": 3.111895799636841,
+      "learning_rate": 3.9609706396855296e-05,
+      "loss": 0.1426,
+      "mean_token_accuracy": 0.9478920578956604,
+      "num_tokens": 1659636.0,
+      "step": 1200
+    },
+    {
+      "entropy": 0.9868269383907318,
+      "epoch": 1.1874386653581943,
+      "grad_norm": 3.178105592727661,
+      "learning_rate": 3.881745941679917e-05,
+      "loss": 0.1371,
+      "mean_token_accuracy": 0.9449715197086335,
+      "num_tokens": 1673078.0,
+      "step": 1210
+    },
+    {
+      "entropy": 0.9963244318962097,
+      "epoch": 1.197252208047105,
+      "grad_norm": 4.070995330810547,
+      "learning_rate": 3.802815699392418e-05,
+      "loss": 0.1341,
+      "mean_token_accuracy": 0.9494788229465485,
+      "num_tokens": 1687154.0,
+      "step": 1220
+    },
+    {
+      "entropy": 0.9938208758831024,
+      "epoch": 1.2070657507360156,
+      "grad_norm": 4.066131114959717,
+      "learning_rate": 3.724200696526873e-05,
+      "loss": 0.1336,
+      "mean_token_accuracy": 0.9498476982116699,
+      "num_tokens": 1701533.0,
+      "step": 1230
+    },
+    {
+      "entropy": 0.9889326453208923,
+      "epoch": 1.2168792934249264,
+      "grad_norm": 4.030731201171875,
+      "learning_rate": 3.6459216337791124e-05,
+      "loss": 0.1497,
+      "mean_token_accuracy": 0.9421870768070221,
+      "num_tokens": 1714722.0,
+      "step": 1240
+    },
+    {
+      "entropy": 0.9747929215431214,
+      "epoch": 1.2266928361138372,
+      "grad_norm": 2.899670362472534,
+      "learning_rate": 3.567999123386103e-05,
+      "loss": 0.1469,
+      "mean_token_accuracy": 0.9453438878059387,
+      "num_tokens": 1727989.0,
+      "step": 1250
+    },
+    {
+      "entropy": 0.9812735736370086,
+      "epoch": 1.2365063788027477,
+      "grad_norm": 3.1901872158050537,
+      "learning_rate": 3.490453683698376e-05,
+      "loss": 0.1393,
+      "mean_token_accuracy": 0.9492902338504792,
+      "num_tokens": 1742532.0,
+      "step": 1260
+    },
+    {
+      "entropy": 0.9771670997142792,
+      "epoch": 1.2463199214916585,
+      "grad_norm": 2.7989614009857178,
+      "learning_rate": 3.4133057337771995e-05,
+      "loss": 0.1319,
+      "mean_token_accuracy": 0.9493632197380066,
+      "num_tokens": 1755690.0,
+      "step": 1270
+    },
+    {
+      "entropy": 0.9728339910507202,
+      "epoch": 1.2561334641805693,
+      "grad_norm": 2.9319562911987305,
+      "learning_rate": 3.336575588017881e-05,
+      "loss": 0.1415,
+      "mean_token_accuracy": 0.9476716935634613,
+      "num_tokens": 1769918.0,
+      "step": 1280
+    },
+    {
+      "entropy": 0.9793581306934357,
+      "epoch": 1.2659470068694798,
+      "grad_norm": 3.8382794857025146,
+      "learning_rate": 3.260283450800634e-05,
+      "loss": 0.125,
+      "mean_token_accuracy": 0.950919258594513,
+      "num_tokens": 1783944.0,
+      "step": 1290
+    },
+    {
+      "entropy": 0.9803735315799713,
+      "epoch": 1.2757605495583906,
+      "grad_norm": 4.322436809539795,
+      "learning_rate": 3.184449411170416e-05,
+      "loss": 0.1183,
+      "mean_token_accuracy": 0.9572235763072967,
+      "num_tokens": 1797715.0,
+      "step": 1300
+    },
+    {
+      "entropy": 0.9727390348911286,
+      "epoch": 1.2855740922473013,
+      "grad_norm": 3.1323184967041016,
+      "learning_rate": 3.109093437547128e-05,
+      "loss": 0.1528,
+      "mean_token_accuracy": 0.9456106781959533,
+      "num_tokens": 1811354.0,
+      "step": 1310
+    },
+    {
+      "entropy": 0.9795861601829529,
+      "epoch": 1.295387634936212,
+      "grad_norm": 3.6378226280212402,
+      "learning_rate": 3.0342353724675966e-05,
+      "loss": 0.1507,
+      "mean_token_accuracy": 0.9429005980491638,
+      "num_tokens": 1825286.0,
+      "step": 1320
+    },
+    {
+      "entropy": 0.9836790263652802,
+      "epoch": 1.3052011776251227,
+      "grad_norm": 3.708000421524048,
+      "learning_rate": 2.9598949273606702e-05,
+      "loss": 0.1343,
+      "mean_token_accuracy": 0.9493555784225464,
+      "num_tokens": 1839043.0,
+      "step": 1330
+    },
+    {
+      "entropy": 0.9730195641517639,
+      "epoch": 1.3150147203140334,
+      "grad_norm": 2.925042152404785,
+      "learning_rate": 2.8860916773568703e-05,
+      "loss": 0.1158,
+      "mean_token_accuracy": 0.9581613719463349,
+      "num_tokens": 1852342.0,
+      "step": 1340
+    },
+    {
+      "entropy": 0.9753842353820801,
+      "epoch": 1.324828263002944,
+      "grad_norm": 4.900411128997803,
+      "learning_rate": 2.8128450561339193e-05,
+      "loss": 0.1272,
+      "mean_token_accuracy": 0.951783299446106,
+      "num_tokens": 1866189.0,
+      "step": 1350
+    },
+    {
+      "entropy": 0.9879122376441956,
+      "epoch": 1.3346418056918548,
+      "grad_norm": 4.224844932556152,
+      "learning_rate": 2.740174350799507e-05,
+      "loss": 0.1286,
+      "mean_token_accuracy": 0.9536733329296112,
+      "num_tokens": 1880098.0,
+      "step": 1360
+    },
+    {
+      "entropy": 0.9657087445259094,
+      "epoch": 1.3444553483807655,
+      "grad_norm": 3.372464895248413,
+      "learning_rate": 2.668098696812675e-05,
+      "loss": 0.1131,
+      "mean_token_accuracy": 0.956873208284378,
+      "num_tokens": 1893007.0,
+      "step": 1370
+    },
+    {
+      "entropy": 0.9820382237434387,
+      "epoch": 1.354268891069676,
+      "grad_norm": 4.074032306671143,
+      "learning_rate": 2.596637072945099e-05,
+      "loss": 0.1221,
+      "mean_token_accuracy": 0.954661512374878,
+      "num_tokens": 1906405.0,
+      "step": 1380
+    },
+    {
+      "entropy": 0.9913123667240142,
+      "epoch": 1.3640824337585868,
+      "grad_norm": 4.9434661865234375,
+      "learning_rate": 2.5258082962836614e-05,
+      "loss": 0.1582,
+      "mean_token_accuracy": 0.9439044833183289,
+      "num_tokens": 1920751.0,
+      "step": 1390
+    },
+    {
+      "entropy": 1.0037564098834992,
+      "epoch": 1.3738959764474976,
+      "grad_norm": 4.565279483795166,
+      "learning_rate": 2.4556310172755898e-05,
+      "loss": 0.1404,
+      "mean_token_accuracy": 0.9453492999076843,
+      "num_tokens": 1934411.0,
+      "step": 1400
+    },
+    {
+      "entropy": 0.9977047741413116,
+      "epoch": 1.3837095191364082,
+      "grad_norm": 4.290781021118164,
+      "learning_rate": 2.3861237148174474e-05,
+      "loss": 0.1249,
+      "mean_token_accuracy": 0.953125411272049,
+      "num_tokens": 1948013.0,
+      "step": 1410
+    },
+    {
+      "entropy": 0.9908617496490478,
+      "epoch": 1.393523061825319,
+      "grad_norm": 2.761293649673462,
+      "learning_rate": 2.3173046913893502e-05,
+      "loss": 0.1127,
+      "mean_token_accuracy": 0.9596257150173187,
+      "num_tokens": 1962274.0,
+      "step": 1420
+    },
+    {
+      "entropy": 1.0044576227664948,
+      "epoch": 1.4033366045142297,
+      "grad_norm": 4.076716423034668,
+      "learning_rate": 2.249192068235581e-05,
+      "loss": 0.1225,
+      "mean_token_accuracy": 0.953608900308609,
+      "num_tokens": 1976604.0,
+      "step": 1430
+    },
+    {
+      "entropy": 1.0052892804145812,
+      "epoch": 1.4131501472031402,
+      "grad_norm": 3.5825304985046387,
+      "learning_rate": 2.181803780592959e-05,
+      "loss": 0.1229,
+      "mean_token_accuracy": 0.9581057608127594,
+      "num_tokens": 1991136.0,
+      "step": 1440
+    },
+    {
+      "entropy": 0.986959558725357,
+      "epoch": 1.422963689892051,
+      "grad_norm": 4.302131652832031,
+      "learning_rate": 2.1151575729681883e-05,
+      "loss": 0.1279,
+      "mean_token_accuracy": 0.9532746791839599,
+      "num_tokens": 2005289.0,
+      "step": 1450
+    },
+    {
+      "entropy": 0.9671379625797272,
+      "epoch": 1.4327772325809618,
+      "grad_norm": 2.9498159885406494,
+      "learning_rate": 2.049270994465407e-05,
+      "loss": 0.1204,
+      "mean_token_accuracy": 0.9539808094501495,
+      "num_tokens": 2018916.0,
+      "step": 1460
+    },
+    {
+      "entropy": 0.9720536410808563,
+      "epoch": 1.4425907752698723,
+      "grad_norm": 3.5524709224700928,
+      "learning_rate": 1.9841613941652144e-05,
+      "loss": 0.1134,
+      "mean_token_accuracy": 0.9568402528762817,
+      "num_tokens": 2032597.0,
+      "step": 1470
+    },
+    {
+      "entropy": 0.9688332319259644,
+      "epoch": 1.452404317958783,
+      "grad_norm": 3.636225700378418,
+      "learning_rate": 1.9198459165563503e-05,
+      "loss": 0.1212,
+      "mean_token_accuracy": 0.9548921585083008,
+      "num_tokens": 2046720.0,
+      "step": 1480
+    },
+    {
+      "entropy": 0.9833451926708221,
+      "epoch": 1.4622178606476939,
+      "grad_norm": 4.683553695678711,
+      "learning_rate": 1.8563414970212555e-05,
+      "loss": 0.1166,
+      "mean_token_accuracy": 0.9568437397480011,
+      "num_tokens": 2060943.0,
+      "step": 1490
+    },
+    {
+      "entropy": 0.9808977901935577,
+      "epoch": 1.4720314033366044,
+      "grad_norm": 3.9009768962860107,
+      "learning_rate": 1.7936648573766957e-05,
+      "loss": 0.1181,
+      "mean_token_accuracy": 0.9531728088855743,
+      "num_tokens": 2074710.0,
+      "step": 1500
+    },
+    {
+      "entropy": 0.9882231473922729,
+      "epoch": 1.4818449460255152,
+      "grad_norm": 3.896113872528076,
+      "learning_rate": 1.731832501470606e-05,
+      "loss": 0.1275,
+      "mean_token_accuracy": 0.9527452468872071,
+      "num_tokens": 2088707.0,
+      "step": 1510
+    },
+    {
+      "entropy": 0.971964854001999,
+      "epoch": 1.491658488714426,
+      "grad_norm": 3.755319833755493,
+      "learning_rate": 1.670860710836355e-05,
+      "loss": 0.1158,
+      "mean_token_accuracy": 0.9561355829238891,
+      "num_tokens": 2101428.0,
+      "step": 1520
+    },
+    {
+      "entropy": 0.9848258137702942,
+      "epoch": 1.5014720314033365,
+      "grad_norm": 2.8726444244384766,
+      "learning_rate": 1.6107655404055223e-05,
+      "loss": 0.1342,
+      "mean_token_accuracy": 0.9481730937957764,
+      "num_tokens": 2115066.0,
+      "step": 1530
+    },
+    {
+      "entropy": 0.9889935076236724,
+      "epoch": 1.5112855740922473,
+      "grad_norm": 3.6505722999572754,
+      "learning_rate": 1.5515628142803562e-05,
+      "loss": 0.124,
+      "mean_token_accuracy": 0.9504759430885314,
+      "num_tokens": 2128953.0,
+      "step": 1540
+    },
+    {
+      "entropy": 0.9845194220542908,
+      "epoch": 1.521099116781158,
+      "grad_norm": 3.4365503787994385,
+      "learning_rate": 1.493268121567028e-05,
+      "loss": 0.1374,
+      "mean_token_accuracy": 0.9488900244235993,
+      "num_tokens": 2142756.0,
+      "step": 1550
+    },
+    {
+      "entropy": 0.9782055437564849,
+      "epoch": 1.5309126594700686,
+      "grad_norm": 4.1006317138671875,
+      "learning_rate": 1.435896812270729e-05,
+      "loss": 0.1162,
+      "mean_token_accuracy": 0.9541732192039489,
+      "num_tokens": 2156108.0,
+      "step": 1560
+    },
+    {
+      "entropy": 0.98279590010643,
+      "epoch": 1.5407262021589794,
+      "grad_norm": 3.986178159713745,
+      "learning_rate": 1.3794639932537645e-05,
+      "loss": 0.1282,
+      "mean_token_accuracy": 0.9536337852478027,
+      "num_tokens": 2169356.0,
+      "step": 1570
+    },
+    {
+      "entropy": 1.0009383618831635,
+      "epoch": 1.5505397448478901,
+      "grad_norm": 3.3969898223876953,
+      "learning_rate": 1.323984524257641e-05,
+      "loss": 0.118,
+      "mean_token_accuracy": 0.9562638521194458,
+      "num_tokens": 2183713.0,
+      "step": 1580
+    },
+    {
+      "entropy": 0.997773551940918,
+      "epoch": 1.5603532875368007,
+      "grad_norm": 4.229125022888184,
+      "learning_rate": 1.2694730139902522e-05,
+      "loss": 0.1217,
+      "mean_token_accuracy": 0.9535910964012146,
+      "num_tokens": 2198089.0,
+      "step": 1590
+    },
+    {
+      "entropy": 1.0049704372882844,
+      "epoch": 1.5701668302257115,
+      "grad_norm": 2.933227062225342,
+      "learning_rate": 1.2159438162791348e-05,
+      "loss": 0.1198,
+      "mean_token_accuracy": 0.9523647844791412,
+      "num_tokens": 2213070.0,
+      "step": 1600
+    },
+    {
+      "entropy": 0.9875008285045623,
+      "epoch": 1.5799803729146222,
+      "grad_norm": 3.8358376026153564,
+      "learning_rate": 1.1634110262918718e-05,
+      "loss": 0.111,
+      "mean_token_accuracy": 0.9581840813159943,
+      "num_tokens": 2227108.0,
+      "step": 1610
+    },
+    {
+      "entropy": 0.9946237921714782,
+      "epoch": 1.5897939156035328,
+      "grad_norm": 4.119611740112305,
+      "learning_rate": 1.1118884768245863e-05,
+      "loss": 0.1146,
+      "mean_token_accuracy": 0.9517829060554505,
+      "num_tokens": 2240704.0,
+      "step": 1620
+    },
+    {
+      "entropy": 1.0009810149669647,
+      "epoch": 1.5996074582924436,
+      "grad_norm": 2.9878668785095215,
+      "learning_rate": 1.0613897346595203e-05,
+      "loss": 0.1212,
+      "mean_token_accuracy": 0.9538079380989075,
+      "num_tokens": 2254968.0,
+      "step": 1630
+    },
+    {
+      "entropy": 1.0099083125591277,
+      "epoch": 1.6094210009813543,
+      "grad_norm": 3.235236406326294,
+      "learning_rate": 1.011928096992672e-05,
+      "loss": 0.1167,
+      "mean_token_accuracy": 0.9555332124233246,
+      "num_tokens": 2269175.0,
+      "step": 1640
+    },
+    {
+      "entropy": 0.9954246342182159,
+      "epoch": 1.6192345436702649,
+      "grad_norm": 4.063044548034668,
+      "learning_rate": 9.635165879324083e-06,
+      "loss": 0.1148,
+      "mean_token_accuracy": 0.9546457588672638,
+      "num_tokens": 2283760.0,
+      "step": 1650
+    },
+    {
+      "entropy": 0.9940894961357116,
+      "epoch": 1.6290480863591756,
+      "grad_norm": 2.2407987117767334,
+      "learning_rate": 9.161679550699936e-06,
+      "loss": 0.1074,
+      "mean_token_accuracy": 0.9601804792881012,
+      "num_tokens": 2297542.0,
+      "step": 1660
+    },
+    {
+      "entropy": 0.9980350255966186,
+      "epoch": 1.6388616290480864,
+      "grad_norm": 3.111990213394165,
+      "learning_rate": 8.698946661229262e-06,
+      "loss": 0.1102,
+      "mean_token_accuracy": 0.9600836217403412,
+      "num_tokens": 2311381.0,
+      "step": 1670
+    },
+    {
+      "entropy": 0.9760864853858948,
+      "epoch": 1.648675171736997,
+      "grad_norm": 2.7510530948638916,
+      "learning_rate": 8.24708905651972e-06,
+      "loss": 0.1281,
+      "mean_token_accuracy": 0.950300419330597,
+      "num_tokens": 2324526.0,
+      "step": 1680
+    },
+    {
+      "entropy": 0.9850055158138276,
+      "epoch": 1.6584887144259077,
+      "grad_norm": 2.782992124557495,
+      "learning_rate": 7.80622571852761e-06,
+      "loss": 0.1133,
+      "mean_token_accuracy": 0.9574047029018402,
+      "num_tokens": 2337382.0,
+      "step": 1690
+    },
+    {
+      "entropy": 0.9859823644161224,
+      "epoch": 1.6683022571148185,
+      "grad_norm": 3.7063050270080566,
+      "learning_rate": 7.376472734227807e-06,
+      "loss": 0.1091,
+      "mean_token_accuracy": 0.9590300917625427,
+      "num_tokens": 2351843.0,
+      "step": 1700
+    },
+    {
+      "entropy": 0.9933929324150086,
+      "epoch": 1.678115799803729,
+      "grad_norm": 3.000054121017456,
+      "learning_rate": 6.957943265046052e-06,
+      "loss": 0.0961,
+      "mean_token_accuracy": 0.9630561530590057,
+      "num_tokens": 2365499.0,
+      "step": 1710
+    },
+    {
+      "entropy": 0.9861454665660858,
+      "epoch": 1.6879293424926398,
+      "grad_norm": 2.75300931930542,
+      "learning_rate": 6.5507475170616565e-06,
+      "loss": 0.1098,
+      "mean_token_accuracy": 0.9572492837905884,
+      "num_tokens": 2379228.0,
+      "step": 1720
+    },
+    {
+      "entropy": 0.9850210309028625,
+      "epoch": 1.6977428851815506,
+      "grad_norm": 4.904257297515869,
+      "learning_rate": 6.154992711988183e-06,
+      "loss": 0.1252,
+      "mean_token_accuracy": 0.9522413551807404,
+      "num_tokens": 2392855.0,
+      "step": 1730
+    },
+    {
+      "entropy": 0.9879020631313324,
+      "epoch": 1.7075564278704611,
+      "grad_norm": 3.6986000537872314,
+      "learning_rate": 5.770783058940271e-06,
+      "loss": 0.1112,
+      "mean_token_accuracy": 0.9588982880115509,
+      "num_tokens": 2406729.0,
+      "step": 1740
+    },
+    {
+      "entropy": 0.9944790422916412,
+      "epoch": 1.717369970559372,
+      "grad_norm": 3.7344956398010254,
+      "learning_rate": 5.398219726993442e-06,
+      "loss": 0.1154,
+      "mean_token_accuracy": 0.955155885219574,
+      "num_tokens": 2420319.0,
+      "step": 1750
+    },
+    {
+      "entropy": 0.9949509918689727,
+      "epoch": 1.7271835132482827,
+      "grad_norm": 3.312868595123291,
+      "learning_rate": 5.03740081854458e-06,
+      "loss": 0.1099,
+      "mean_token_accuracy": 0.9587769448757172,
+      "num_tokens": 2434508.0,
+      "step": 1760
+    },
+    {
+      "entropy": 1.0010779678821564,
+      "epoch": 1.7369970559371932,
+      "grad_norm": 4.436283588409424,
+      "learning_rate": 4.68842134347987e-06,
+      "loss": 0.1132,
+      "mean_token_accuracy": 0.9586503088474274,
+      "num_tokens": 2447872.0,
+      "step": 1770
+    },
+    {
+      "entropy": 0.9889185130596161,
+      "epoch": 1.746810598626104,
+      "grad_norm": 3.8735764026641846,
+      "learning_rate": 4.351373194157016e-06,
+      "loss": 0.1122,
+      "mean_token_accuracy": 0.9576007664203644,
+      "num_tokens": 2461151.0,
+      "step": 1780
+    },
+    {
+      "entropy": 0.9815246641635895,
+      "epoch": 1.7566241413150148,
+      "grad_norm": 3.2665939331054688,
+      "learning_rate": 4.026345121208403e-06,
+      "loss": 0.1074,
+      "mean_token_accuracy": 0.957821124792099,
+      "num_tokens": 2474907.0,
+      "step": 1790
+    },
+    {
+      "entropy": 0.9815526247024536,
+      "epoch": 1.7664376840039253,
+      "grad_norm": 2.895371198654175,
+      "learning_rate": 3.7134227101715156e-06,
+      "loss": 0.1003,
+      "mean_token_accuracy": 0.962746012210846,
+      "num_tokens": 2488408.0,
+      "step": 1800
+    },
+    {
+      "entropy": 0.9917015314102173,
+      "epoch": 1.776251226692836,
+      "grad_norm": 3.9480695724487305,
+      "learning_rate": 3.4126883589527705e-06,
+      "loss": 0.1137,
+      "mean_token_accuracy": 0.953924685716629,
+      "num_tokens": 2503031.0,
+      "step": 1810
+    },
+    {
+      "entropy": 0.9785957932472229,
+      "epoch": 1.7860647693817469,
+      "grad_norm": 4.112260341644287,
+      "learning_rate": 3.124221256130744e-06,
+      "loss": 0.1163,
+      "mean_token_accuracy": 0.9575207650661468,
+      "num_tokens": 2516322.0,
+      "step": 1820
+    },
+    {
+      "entropy": 0.98737553358078,
+      "epoch": 1.7958783120706574,
+      "grad_norm": 3.5249733924865723,
+      "learning_rate": 2.848097360104396e-06,
+      "loss": 0.1059,
+      "mean_token_accuracy": 0.961602795124054,
+      "num_tokens": 2529795.0,
+      "step": 1830
+    },
+    {
+      "entropy": 0.9913795411586761,
+      "epoch": 1.8056918547595682,
+      "grad_norm": 3.3677968978881836,
+      "learning_rate": 2.584389379091945e-06,
+      "loss": 0.1138,
+      "mean_token_accuracy": 0.9550937831401825,
+      "num_tokens": 2542968.0,
+      "step": 1840
+    },
+    {
+      "entropy": 0.977301436662674,
+      "epoch": 1.815505397448479,
+      "grad_norm": 2.5755014419555664,
+      "learning_rate": 2.333166751985516e-06,
+      "loss": 0.0958,
+      "mean_token_accuracy": 0.9656810760498047,
+      "num_tokens": 2556526.0,
+      "step": 1850
+    },
+    {
+      "entropy": 0.9797476053237915,
+      "epoch": 1.8253189401373895,
+      "grad_norm": 3.93009877204895,
+      "learning_rate": 2.0944956300666565e-06,
+      "loss": 0.1152,
+      "mean_token_accuracy": 0.9555214583873749,
+      "num_tokens": 2569913.0,
+      "step": 1860
+    },
+    {
+      "entropy": 0.989494013786316,
+      "epoch": 1.8351324828263003,
+      "grad_norm": 2.343184471130371,
+      "learning_rate": 1.8684388595876012e-06,
+      "loss": 0.1047,
+      "mean_token_accuracy": 0.9648076176643372,
+      "num_tokens": 2584227.0,
+      "step": 1870
+    },
+    {
+      "entropy": 0.9928219079971313,
+      "epoch": 1.844946025515211,
+      "grad_norm": 3.1005959510803223,
+      "learning_rate": 1.6550559652227126e-06,
+      "loss": 0.1166,
+      "mean_token_accuracy": 0.9580796658992767,
+      "num_tokens": 2598410.0,
+      "step": 1880
+    },
+    {
+      "entropy": 0.9943685293197632,
+      "epoch": 1.8547595682041216,
+      "grad_norm": 3.6579575538635254,
+      "learning_rate": 1.4544031343946385e-06,
+      "loss": 0.0998,
+      "mean_token_accuracy": 0.9659820854663849,
+      "num_tokens": 2613088.0,
+      "step": 1890
+    },
+    {
+      "entropy": 0.9918886661529541,
+      "epoch": 1.8645731108930323,
+      "grad_norm": 1.8360697031021118,
+      "learning_rate": 1.2665332024791532e-06,
+      "loss": 0.0968,
+      "mean_token_accuracy": 0.9665282189846038,
+      "num_tokens": 2627112.0,
+      "step": 1900
+    },
+    {
+      "entropy": 0.9837871968746186,
+      "epoch": 1.8743866535819431,
+      "grad_norm": 3.583338975906372,
+      "learning_rate": 1.09149563889272e-06,
+      "loss": 0.1161,
+      "mean_token_accuracy": 0.9569039404392242,
+      "num_tokens": 2640791.0,
+      "step": 1910
+    },
+    {
+      "entropy": 0.9997702777385712,
+      "epoch": 1.8842001962708537,
+      "grad_norm": 3.401669502258301,
+      "learning_rate": 9.293365340662996e-07,
+      "loss": 0.1217,
+      "mean_token_accuracy": 0.9536440908908844,
+      "num_tokens": 2654993.0,
+      "step": 1920
+    },
+    {
+      "entropy": 0.976694667339325,
+      "epoch": 1.8940137389597644,
+      "grad_norm": 3.7565090656280518,
+      "learning_rate": 7.800985873089695e-07,
+      "loss": 0.0965,
+      "mean_token_accuracy": 0.9638497829437256,
+      "num_tokens": 2669247.0,
+      "step": 1930
+    },
+    {
+      "entropy": 0.9775448381900788,
+      "epoch": 1.9038272816486752,
+      "grad_norm": 3.0819034576416016,
+      "learning_rate": 6.438210955644452e-07,
+      "loss": 0.0821,
+      "mean_token_accuracy": 0.9711402595043183,
+      "num_tokens": 2682844.0,
+      "step": 1940
+    },
+    {
+      "entropy": 0.9989607393741607,
+      "epoch": 1.9136408243375858,
+      "grad_norm": 3.931959867477417,
+      "learning_rate": 5.205399430634917e-07,
+      "loss": 0.117,
+      "mean_token_accuracy": 0.9561911821365356,
+      "num_tokens": 2696806.0,
+      "step": 1950
+    },
+    {
+      "entropy": 0.9848430812358856,
+      "epoch": 1.9234543670264965,
+      "grad_norm": 2.6147401332855225,
+      "learning_rate": 4.102875918750082e-07,
+      "loss": 0.1135,
+      "mean_token_accuracy": 0.9576592445373535,
+      "num_tokens": 2710979.0,
+      "step": 1960
+    },
+    {
+      "entropy": 0.9907346189022064,
+      "epoch": 1.9332679097154073,
+      "grad_norm": 4.119517803192139,
+      "learning_rate": 3.1309307335820603e-07,
+      "loss": 0.1275,
+      "mean_token_accuracy": 0.9507604956626892,
+      "num_tokens": 2724880.0,
+      "step": 1970
+    },
+    {
+      "entropy": 0.9974951148033142,
+      "epoch": 1.9430814524043178,
+      "grad_norm": 2.8238637447357178,
+      "learning_rate": 2.2898198051811148e-07,
+      "loss": 0.1033,
+      "mean_token_accuracy": 0.9608614861965179,
+      "num_tokens": 2738562.0,
+      "step": 1980
+    },
+    {
+      "entropy": 0.9897377073764801,
+      "epoch": 1.9528949950932286,
+      "grad_norm": 2.287062644958496,
+      "learning_rate": 1.5797646126651245e-07,
+      "loss": 0.0999,
+      "mean_token_accuracy": 0.9604212641716003,
+      "num_tokens": 2752355.0,
+      "step": 1990
+    },
+    {
+      "entropy": 0.9937055885791779,
+      "epoch": 1.9627085377821394,
+      "grad_norm": 2.7865030765533447,
+      "learning_rate": 1.0009521259003452e-07,
+      "loss": 0.1141,
+      "mean_token_accuracy": 0.9575498282909394,
+      "num_tokens": 2766143.0,
+      "step": 2000
+    },
+    {
+      "entropy": 0.9848930299282074,
+      "epoch": 1.97252208047105,
+      "grad_norm": 2.517505168914795,
+      "learning_rate": 5.53534756268681e-08,
+      "loss": 0.1076,
+      "mean_token_accuracy": 0.9569665908813476,
+      "num_tokens": 2779801.0,
+      "step": 2010
+    },
+    {
+      "entropy": 0.9794306576251983,
+      "epoch": 1.9823356231599607,
+      "grad_norm": 4.352532863616943,
+      "learning_rate": 2.376303165353977e-08,
+      "loss": 0.0956,
+      "mean_token_accuracy": 0.9651121854782104,
+      "num_tokens": 2793763.0,
+      "step": 2020
+    },
+    {
+      "entropy": 0.9799128115177155,
+      "epoch": 1.9921491658488715,
+      "grad_norm": 3.532918691635132,
+      "learning_rate": 5.33219898269377e-09,
+      "loss": 0.0895,
+      "mean_token_accuracy": 0.9685670793056488,
+      "num_tokens": 2807624.0,
+      "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_adam_rot_e2_lr1e-4_s1_10k/checkpoint-2038/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34519d7b2f8033bed6a7a7b759e9a16c47875244d1c7a125c0bd74989108f2c7
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_10k/manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_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": "adamw_torch",
+    "lr_scheduler": "cosine",
+    "warmup_ratio": 0.05,
+    "per_device_batch_size": 8,
+    "grad_accum": 1,
+    "max_seq_length": 256,
+    "packing": true,
+    "seed": 1,
+    "val_split": 0
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_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_adam_rot_e2_lr1e-4_s1_10k/train_manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_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": "adamw_torch",
+    "lr_scheduler": "cosine",
+    "warmup_ratio": 0.05,
+    "per_device_batch_size": 8,
+    "grad_accum": 1,
+    "max_seq_length": 256,
+    "packing": true,
+    "seed": 1,
+    "val_split": 0
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s1_10k/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34519d7b2f8033bed6a7a7b759e9a16c47875244d1c7a125c0bd74989108f2c7
+size 6417