add checkpoint cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_10k

.gitattributes

@@ -543,3 +543,6 @@ checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s3_10k/tokenizer.json filter
 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_s2_10k/checkpoint-1018/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_10k/checkpoint-2036/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_10k/tokenizer.json filter=lfs diff=lfs merge=lfs -text

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_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_s2_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_s2_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/t56z1uj2) 
+
+
+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_s2_10k/adapter_config.json

@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "gate_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj",
+    "down_proj",
+    "v_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_s2_10k/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f197a52b3b8ff881f1532c383d8a17236c52fa9328a43e55635bc65863bff6ac
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_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_s2_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_s2_10k/checkpoint-1018/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_s2_10k/checkpoint-1018/adapter_config.json

@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "gate_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj",
+    "down_proj",
+    "v_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_s2_10k/checkpoint-1018/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f15a5558abeeb80e040cc97046b5c8c83a8175eec53066476b18963fa4e05aad
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_10k/checkpoint-1018/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_s2_10k/checkpoint-1018/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_s2_10k/checkpoint-1018/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_s2_10k/checkpoint-1018/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_s2_10k/checkpoint-1018/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_s2_10k/checkpoint-1018/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": 1018,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.2257721304893494,
+      "epoch": 0.009823182711198428,
+      "grad_norm": 4.894117832183838,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.6909,
+      "mean_token_accuracy": 0.775017136335373,
+      "num_tokens": 13828.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.2686485409736634,
+      "epoch": 0.019646365422396856,
+      "grad_norm": 2.6351609230041504,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.7667,
+      "mean_token_accuracy": 0.732726263999939,
+      "num_tokens": 28074.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.2789716601371766,
+      "epoch": 0.029469548133595286,
+      "grad_norm": 1.8091380596160889,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.672,
+      "mean_token_accuracy": 0.7583289265632629,
+      "num_tokens": 42168.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.247054147720337,
+      "epoch": 0.03929273084479371,
+      "grad_norm": 2.2216796875,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.6733,
+      "mean_token_accuracy": 0.7530835151672364,
+      "num_tokens": 56700.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.2091697812080384,
+      "epoch": 0.04911591355599214,
+      "grad_norm": 2.6457648277282715,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.559,
+      "mean_token_accuracy": 0.7906185984611511,
+      "num_tokens": 70922.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.2023007988929748,
+      "epoch": 0.05893909626719057,
+      "grad_norm": 2.2835278511047363,
+      "learning_rate": 5.784313725490197e-05,
+      "loss": 0.5012,
+      "mean_token_accuracy": 0.8101553797721863,
+      "num_tokens": 84911.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.1628074526786805,
+      "epoch": 0.068762278978389,
+      "grad_norm": 2.8187313079833984,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 0.5312,
+      "mean_token_accuracy": 0.8078348159790039,
+      "num_tokens": 98835.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.173559021949768,
+      "epoch": 0.07858546168958742,
+      "grad_norm": 3.0911595821380615,
+      "learning_rate": 7.745098039215687e-05,
+      "loss": 0.4827,
+      "mean_token_accuracy": 0.8213136911392211,
+      "num_tokens": 112502.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.1246683597564697,
+      "epoch": 0.08840864440078586,
+      "grad_norm": 2.5103771686553955,
+      "learning_rate": 8.725490196078432e-05,
+      "loss": 0.4858,
+      "mean_token_accuracy": 0.8161003708839416,
+      "num_tokens": 126248.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.141631579399109,
+      "epoch": 0.09823182711198428,
+      "grad_norm": 2.9426238536834717,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 0.4759,
+      "mean_token_accuracy": 0.8214390873908997,
+      "num_tokens": 139551.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.1160478115081787,
+      "epoch": 0.10805500982318271,
+      "grad_norm": 3.039224624633789,
+      "learning_rate": 9.999676765122786e-05,
+      "loss": 0.45,
+      "mean_token_accuracy": 0.8267141044139862,
+      "num_tokens": 152975.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.1066530227661133,
+      "epoch": 0.11787819253438114,
+      "grad_norm": 2.6248061656951904,
+      "learning_rate": 9.998093674494531e-05,
+      "loss": 0.4723,
+      "mean_token_accuracy": 0.8254820048809052,
+      "num_tokens": 167073.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.1236253738403321,
+      "epoch": 0.12770137524557956,
+      "grad_norm": 2.6021196842193604,
+      "learning_rate": 9.995191775639568e-05,
+      "loss": 0.4827,
+      "mean_token_accuracy": 0.8199610054492951,
+      "num_tokens": 181404.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.1052476644515992,
+      "epoch": 0.137524557956778,
+      "grad_norm": 2.3915598392486572,
+      "learning_rate": 9.990971834259444e-05,
+      "loss": 0.453,
+      "mean_token_accuracy": 0.8272700130939483,
+      "num_tokens": 195441.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.1267707705497743,
+      "epoch": 0.14734774066797643,
+      "grad_norm": 3.4826769828796387,
+      "learning_rate": 9.985434963837376e-05,
+      "loss": 0.4805,
+      "mean_token_accuracy": 0.8141039609909058,
+      "num_tokens": 209863.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.0815376162528991,
+      "epoch": 0.15717092337917485,
+      "grad_norm": 3.369027614593506,
+      "learning_rate": 9.978582625344435e-05,
+      "loss": 0.4439,
+      "mean_token_accuracy": 0.833777642250061,
+      "num_tokens": 223646.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.093427562713623,
+      "epoch": 0.16699410609037327,
+      "grad_norm": 3.097435235977173,
+      "learning_rate": 9.970416626854049e-05,
+      "loss": 0.4171,
+      "mean_token_accuracy": 0.8488226413726807,
+      "num_tokens": 237587.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.0916131377220153,
+      "epoch": 0.17681728880157171,
+      "grad_norm": 2.9525489807128906,
+      "learning_rate": 9.960939123064929e-05,
+      "loss": 0.4524,
+      "mean_token_accuracy": 0.8301271557807922,
+      "num_tokens": 251760.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.0759238362312318,
+      "epoch": 0.18664047151277013,
+      "grad_norm": 2.6859591007232666,
+      "learning_rate": 9.950152614732518e-05,
+      "loss": 0.4167,
+      "mean_token_accuracy": 0.8416622459888459,
+      "num_tokens": 265417.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.0830464363098145,
+      "epoch": 0.19646365422396855,
+      "grad_norm": 3.650442123413086,
+      "learning_rate": 9.93805994800914e-05,
+      "loss": 0.4541,
+      "mean_token_accuracy": 0.8313074588775635,
+      "num_tokens": 278540.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.0821943521499633,
+      "epoch": 0.206286836935167,
+      "grad_norm": 3.401404619216919,
+      "learning_rate": 9.924664313693006e-05,
+      "loss": 0.435,
+      "mean_token_accuracy": 0.8389021277427673,
+      "num_tokens": 292295.0,
+      "step": 210
+    },
+    {
+      "entropy": 1.0802088141441346,
+      "epoch": 0.21611001964636542,
+      "grad_norm": 3.133098840713501,
+      "learning_rate": 9.909969246386284e-05,
+      "loss": 0.4309,
+      "mean_token_accuracy": 0.8364023327827453,
+      "num_tokens": 305466.0,
+      "step": 220
+    },
+    {
+      "entropy": 1.1103221774101257,
+      "epoch": 0.22593320235756384,
+      "grad_norm": 2.64292311668396,
+      "learning_rate": 9.893978623562453e-05,
+      "loss": 0.4289,
+      "mean_token_accuracy": 0.8376089572906494,
+      "num_tokens": 320495.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.0710348248481751,
+      "epoch": 0.2357563850687623,
+      "grad_norm": 2.8242721557617188,
+      "learning_rate": 9.876696664543177e-05,
+      "loss": 0.4032,
+      "mean_token_accuracy": 0.8512109398841858,
+      "num_tokens": 333901.0,
+      "step": 240
+    },
+    {
+      "entropy": 1.0859151601791381,
+      "epoch": 0.2455795677799607,
+      "grad_norm": 3.6954805850982666,
+      "learning_rate": 9.858127929384991e-05,
+      "loss": 0.3854,
+      "mean_token_accuracy": 0.8480402171611786,
+      "num_tokens": 347392.0,
+      "step": 250
+    },
+    {
+      "entropy": 1.0700998306274414,
+      "epoch": 0.2554027504911591,
+      "grad_norm": 3.1757068634033203,
+      "learning_rate": 9.838277317676077e-05,
+      "loss": 0.3992,
+      "mean_token_accuracy": 0.8441122591495513,
+      "num_tokens": 361140.0,
+      "step": 260
+    },
+    {
+      "entropy": 1.0962264895439149,
+      "epoch": 0.26522593320235754,
+      "grad_norm": 2.899790048599243,
+      "learning_rate": 9.817150067243438e-05,
+      "loss": 0.359,
+      "mean_token_accuracy": 0.8624118745326996,
+      "num_tokens": 374661.0,
+      "step": 270
+    },
+    {
+      "entropy": 1.1250731945037842,
+      "epoch": 0.275049115913556,
+      "grad_norm": 2.9336721897125244,
+      "learning_rate": 9.794751752770845e-05,
+      "loss": 0.4116,
+      "mean_token_accuracy": 0.8425243020057678,
+      "num_tokens": 388415.0,
+      "step": 280
+    },
+    {
+      "entropy": 1.170729434490204,
+      "epoch": 0.28487229862475444,
+      "grad_norm": 4.345315456390381,
+      "learning_rate": 9.771088284327879e-05,
+      "loss": 0.4232,
+      "mean_token_accuracy": 0.8408111214637757,
+      "num_tokens": 402031.0,
+      "step": 290
+    },
+    {
+      "entropy": 1.1610852003097534,
+      "epoch": 0.29469548133595286,
+      "grad_norm": 3.390840768814087,
+      "learning_rate": 9.746165905810488e-05,
+      "loss": 0.376,
+      "mean_token_accuracy": 0.85899418592453,
+      "num_tokens": 416165.0,
+      "step": 300
+    },
+    {
+      "entropy": 1.1845875024795531,
+      "epoch": 0.3045186640471513,
+      "grad_norm": 2.6109654903411865,
+      "learning_rate": 9.719991193293466e-05,
+      "loss": 0.4143,
+      "mean_token_accuracy": 0.8462965011596679,
+      "num_tokens": 429910.0,
+      "step": 310
+    },
+    {
+      "entropy": 1.1857431888580323,
+      "epoch": 0.3143418467583497,
+      "grad_norm": 3.2335426807403564,
+      "learning_rate": 9.692571053295268e-05,
+      "loss": 0.4037,
+      "mean_token_accuracy": 0.8504281759262085,
+      "num_tokens": 443725.0,
+      "step": 320
+    },
+    {
+      "entropy": 1.179654860496521,
+      "epoch": 0.3241650294695481,
+      "grad_norm": 3.481412649154663,
+      "learning_rate": 9.663912720955637e-05,
+      "loss": 0.3978,
+      "mean_token_accuracy": 0.8510401785373688,
+      "num_tokens": 458234.0,
+      "step": 330
+    },
+    {
+      "entropy": 1.1839690446853637,
+      "epoch": 0.33398821218074654,
+      "grad_norm": 3.3959858417510986,
+      "learning_rate": 9.634023758126539e-05,
+      "loss": 0.375,
+      "mean_token_accuracy": 0.8541060447692871,
+      "num_tokens": 471402.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.1508894562721252,
+      "epoch": 0.343811394891945,
+      "grad_norm": 3.046173334121704,
+      "learning_rate": 9.602912051376859e-05,
+      "loss": 0.3701,
+      "mean_token_accuracy": 0.8562622129917145,
+      "num_tokens": 484651.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.1543931007385253,
+      "epoch": 0.35363457760314343,
+      "grad_norm": 2.7381064891815186,
+      "learning_rate": 9.570585809911441e-05,
+      "loss": 0.3755,
+      "mean_token_accuracy": 0.8545270740985871,
+      "num_tokens": 497956.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.1635468125343322,
+      "epoch": 0.36345776031434185,
+      "grad_norm": 2.425509214401245,
+      "learning_rate": 9.537053563404992e-05,
+      "loss": 0.388,
+      "mean_token_accuracy": 0.8558545649051666,
+      "num_tokens": 511717.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.1691488862037658,
+      "epoch": 0.37328094302554027,
+      "grad_norm": 4.111085414886475,
+      "learning_rate": 9.502324159751412e-05,
+      "loss": 0.3499,
+      "mean_token_accuracy": 0.8601330816745758,
+      "num_tokens": 525633.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.1114183068275452,
+      "epoch": 0.3831041257367387,
+      "grad_norm": 2.738142251968384,
+      "learning_rate": 9.466406762729181e-05,
+      "loss": 0.3502,
+      "mean_token_accuracy": 0.8674743235111236,
+      "num_tokens": 538945.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.111063504219055,
+      "epoch": 0.3929273084479371,
+      "grad_norm": 3.1340293884277344,
+      "learning_rate": 9.429310849583378e-05,
+      "loss": 0.3669,
+      "mean_token_accuracy": 0.8642211854457855,
+      "num_tokens": 552861.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.1472705245018004,
+      "epoch": 0.4027504911591356,
+      "grad_norm": 3.1255431175231934,
+      "learning_rate": 9.391046208524989e-05,
+      "loss": 0.3978,
+      "mean_token_accuracy": 0.8444820404052734,
+      "num_tokens": 565955.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.1532312989234925,
+      "epoch": 0.412573673870334,
+      "grad_norm": 2.9525880813598633,
+      "learning_rate": 9.351622936148182e-05,
+      "loss": 0.348,
+      "mean_token_accuracy": 0.8709490835666657,
+      "num_tokens": 579923.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.136536467075348,
+      "epoch": 0.4223968565815324,
+      "grad_norm": 2.7755491733551025,
+      "learning_rate": 9.31105143476618e-05,
+      "loss": 0.3534,
+      "mean_token_accuracy": 0.8653653502464295,
+      "num_tokens": 594071.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.1585060477256774,
+      "epoch": 0.43222003929273084,
+      "grad_norm": 3.2028636932373047,
+      "learning_rate": 9.269342409666495e-05,
+      "loss": 0.3332,
+      "mean_token_accuracy": 0.8707658588886261,
+      "num_tokens": 608545.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.1323098540306091,
+      "epoch": 0.44204322200392926,
+      "grad_norm": 2.7898449897766113,
+      "learning_rate": 9.226506866286202e-05,
+      "loss": 0.3398,
+      "mean_token_accuracy": 0.8691294670104981,
+      "num_tokens": 621975.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.1309045553207397,
+      "epoch": 0.4518664047151277,
+      "grad_norm": 3.5157182216644287,
+      "learning_rate": 9.18255610730802e-05,
+      "loss": 0.3541,
+      "mean_token_accuracy": 0.8666610240936279,
+      "num_tokens": 636061.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.1365639805793761,
+      "epoch": 0.46168958742632615,
+      "grad_norm": 2.39361572265625,
+      "learning_rate": 9.137501729677962e-05,
+      "loss": 0.3504,
+      "mean_token_accuracy": 0.8678753793239593,
+      "num_tokens": 649717.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.1228155851364137,
+      "epoch": 0.4715127701375246,
+      "grad_norm": 3.2366058826446533,
+      "learning_rate": 9.091355621545339e-05,
+      "loss": 0.3231,
+      "mean_token_accuracy": 0.8785337209701538,
+      "num_tokens": 664123.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.1357950210571288,
+      "epoch": 0.481335952848723,
+      "grad_norm": 3.25319242477417,
+      "learning_rate": 9.044129959125923e-05,
+      "loss": 0.3638,
+      "mean_token_accuracy": 0.8627843201160431,
+      "num_tokens": 677803.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.1405863523483277,
+      "epoch": 0.4911591355599214,
+      "grad_norm": 3.28054141998291,
+      "learning_rate": 8.995837203489108e-05,
+      "loss": 0.3151,
+      "mean_token_accuracy": 0.8844160795211792,
+      "num_tokens": 691461.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.1509727835655212,
+      "epoch": 0.5009823182711198,
+      "grad_norm": 3.824949264526367,
+      "learning_rate": 8.946490097269899e-05,
+      "loss": 0.3421,
+      "mean_token_accuracy": 0.872158533334732,
+      "num_tokens": 705470.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.155228316783905,
+      "epoch": 0.5108055009823183,
+      "grad_norm": 3.0210299491882324,
+      "learning_rate": 8.896101661306618e-05,
+      "loss": 0.3631,
+      "mean_token_accuracy": 0.8582519233226776,
+      "num_tokens": 718805.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.1403202533721923,
+      "epoch": 0.5206286836935167,
+      "grad_norm": 2.491917371749878,
+      "learning_rate": 8.844685191205193e-05,
+      "loss": 0.3376,
+      "mean_token_accuracy": 0.8696144163608551,
+      "num_tokens": 732367.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.1386663675308228,
+      "epoch": 0.5304518664047151,
+      "grad_norm": 3.0767979621887207,
+      "learning_rate": 8.792254253830945e-05,
+      "loss": 0.3251,
+      "mean_token_accuracy": 0.8764946639537812,
+      "num_tokens": 746431.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.1174968481063843,
+      "epoch": 0.5402750491159135,
+      "grad_norm": 2.6364076137542725,
+      "learning_rate": 8.738822683728814e-05,
+      "loss": 0.3105,
+      "mean_token_accuracy": 0.8814369976520539,
+      "num_tokens": 760182.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.1403401017189025,
+      "epoch": 0.550098231827112,
+      "grad_norm": 2.6279711723327637,
+      "learning_rate": 8.684404579472936e-05,
+      "loss": 0.315,
+      "mean_token_accuracy": 0.8731137156486511,
+      "num_tokens": 774137.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.1179575324058533,
+      "epoch": 0.5599214145383105,
+      "grad_norm": 4.026428699493408,
+      "learning_rate": 8.629014299946568e-05,
+      "loss": 0.3162,
+      "mean_token_accuracy": 0.8734814465045929,
+      "num_tokens": 787806.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.1278663873672485,
+      "epoch": 0.5697445972495089,
+      "grad_norm": 3.3344717025756836,
+      "learning_rate": 8.572666460553316e-05,
+      "loss": 0.3149,
+      "mean_token_accuracy": 0.8808867335319519,
+      "num_tokens": 801431.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.122735869884491,
+      "epoch": 0.5795677799607073,
+      "grad_norm": 2.997509479522705,
+      "learning_rate": 8.515375929360687e-05,
+      "loss": 0.3248,
+      "mean_token_accuracy": 0.8767118990421295,
+      "num_tokens": 815080.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.1275416135787963,
+      "epoch": 0.5893909626719057,
+      "grad_norm": 4.02539587020874,
+      "learning_rate": 8.45715782317696e-05,
+      "loss": 0.337,
+      "mean_token_accuracy": 0.8703737854957581,
+      "num_tokens": 829061.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.1480388522148133,
+      "epoch": 0.5992141453831041,
+      "grad_norm": 3.0496115684509277,
+      "learning_rate": 8.39802750356244e-05,
+      "loss": 0.3401,
+      "mean_token_accuracy": 0.8724706113338471,
+      "num_tokens": 842325.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.1545992732048034,
+      "epoch": 0.6090373280943026,
+      "grad_norm": 3.738725423812866,
+      "learning_rate": 8.338000572776114e-05,
+      "loss": 0.3144,
+      "mean_token_accuracy": 0.8776861608028412,
+      "num_tokens": 856377.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.1416390776634215,
+      "epoch": 0.618860510805501,
+      "grad_norm": 3.837313652038574,
+      "learning_rate": 8.277092869658808e-05,
+      "loss": 0.2838,
+      "mean_token_accuracy": 0.8879789412021637,
+      "num_tokens": 870502.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.1548494458198548,
+      "epoch": 0.6286836935166994,
+      "grad_norm": 2.7274272441864014,
+      "learning_rate": 8.215320465453914e-05,
+      "loss": 0.3315,
+      "mean_token_accuracy": 0.875211751461029,
+      "num_tokens": 884708.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.1428229212760925,
+      "epoch": 0.6385068762278978,
+      "grad_norm": 2.581369638442993,
+      "learning_rate": 8.152699659566797e-05,
+      "loss": 0.3126,
+      "mean_token_accuracy": 0.8831508219242096,
+      "num_tokens": 898142.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.1344876527786254,
+      "epoch": 0.6483300589390962,
+      "grad_norm": 2.903346061706543,
+      "learning_rate": 8.089246975263996e-05,
+      "loss": 0.3066,
+      "mean_token_accuracy": 0.882864135503769,
+      "num_tokens": 912820.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.1322418928146363,
+      "epoch": 0.6581532416502947,
+      "grad_norm": 3.234118938446045,
+      "learning_rate": 8.024979155313354e-05,
+      "loss": 0.2959,
+      "mean_token_accuracy": 0.8836028277873993,
+      "num_tokens": 926399.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.1417956709861756,
+      "epoch": 0.6679764243614931,
+      "grad_norm": 2.977431535720825,
+      "learning_rate": 7.959913157566246e-05,
+      "loss": 0.3271,
+      "mean_token_accuracy": 0.8776575326919556,
+      "num_tokens": 940133.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.1574132323265076,
+      "epoch": 0.6777996070726916,
+      "grad_norm": 3.9751768112182617,
+      "learning_rate": 7.894066150483023e-05,
+      "loss": 0.3056,
+      "mean_token_accuracy": 0.8853654146194458,
+      "num_tokens": 954504.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.1295910596847534,
+      "epoch": 0.68762278978389,
+      "grad_norm": 3.3603923320770264,
+      "learning_rate": 7.827455508602923e-05,
+      "loss": 0.3389,
+      "mean_token_accuracy": 0.8706278026103973,
+      "num_tokens": 968934.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.1322409749031066,
+      "epoch": 0.6974459724950884,
+      "grad_norm": 2.7524755001068115,
+      "learning_rate": 7.760098807959573e-05,
+      "loss": 0.335,
+      "mean_token_accuracy": 0.8757263839244842,
+      "num_tokens": 983038.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.1349270939826965,
+      "epoch": 0.7072691552062869,
+      "grad_norm": 3.783806800842285,
+      "learning_rate": 7.69201382144334e-05,
+      "loss": 0.3167,
+      "mean_token_accuracy": 0.8799357235431671,
+      "num_tokens": 996748.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.130021619796753,
+      "epoch": 0.7170923379174853,
+      "grad_norm": 2.9027302265167236,
+      "learning_rate": 7.623218514111743e-05,
+      "loss": 0.2839,
+      "mean_token_accuracy": 0.8918920040130616,
+      "num_tokens": 1010595.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.1162713766098022,
+      "epoch": 0.7269155206286837,
+      "grad_norm": 3.68827486038208,
+      "learning_rate": 7.553731038449144e-05,
+      "loss": 0.2762,
+      "mean_token_accuracy": 0.8936041057109833,
+      "num_tokens": 1024890.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.1368697047233582,
+      "epoch": 0.7367387033398821,
+      "grad_norm": 3.653467893600464,
+      "learning_rate": 7.483569729576996e-05,
+      "loss": 0.2924,
+      "mean_token_accuracy": 0.8875188827514648,
+      "num_tokens": 1038967.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.1225467801094056,
+      "epoch": 0.7465618860510805,
+      "grad_norm": 3.6799514293670654,
+      "learning_rate": 7.412753100415897e-05,
+      "loss": 0.2776,
+      "mean_token_accuracy": 0.8891165137290955,
+      "num_tokens": 1052622.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.1493797302246094,
+      "epoch": 0.756385068762279,
+      "grad_norm": 2.992457389831543,
+      "learning_rate": 7.341299836800729e-05,
+      "loss": 0.3057,
+      "mean_token_accuracy": 0.886419826745987,
+      "num_tokens": 1066588.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.1170050144195556,
+      "epoch": 0.7662082514734774,
+      "grad_norm": 3.467043161392212,
+      "learning_rate": 7.26922879255017e-05,
+      "loss": 0.2458,
+      "mean_token_accuracy": 0.905247437953949,
+      "num_tokens": 1080453.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.1004221796989442,
+      "epoch": 0.7760314341846758,
+      "grad_norm": 2.8668010234832764,
+      "learning_rate": 7.196558984491889e-05,
+      "loss": 0.2955,
+      "mean_token_accuracy": 0.8834676742553711,
+      "num_tokens": 1094550.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.1233540773391724,
+      "epoch": 0.7858546168958742,
+      "grad_norm": 3.163811445236206,
+      "learning_rate": 7.123309587444711e-05,
+      "loss": 0.2667,
+      "mean_token_accuracy": 0.8962497293949128,
+      "num_tokens": 1108808.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.1172690033912658,
+      "epoch": 0.7956777996070727,
+      "grad_norm": 2.911059617996216,
+      "learning_rate": 7.04949992915912e-05,
+      "loss": 0.2584,
+      "mean_token_accuracy": 0.9001988768577576,
+      "num_tokens": 1122365.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.124155080318451,
+      "epoch": 0.8055009823182712,
+      "grad_norm": 3.730689287185669,
+      "learning_rate": 6.975149485217388e-05,
+      "loss": 0.3021,
+      "mean_token_accuracy": 0.8832005798816681,
+      "num_tokens": 1136376.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.1060636997222901,
+      "epoch": 0.8153241650294696,
+      "grad_norm": 2.803528070449829,
+      "learning_rate": 6.900277873894705e-05,
+      "loss": 0.2887,
+      "mean_token_accuracy": 0.8805433213710785,
+      "num_tokens": 1150011.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.1344504833221436,
+      "epoch": 0.825147347740668,
+      "grad_norm": 2.9840211868286133,
+      "learning_rate": 6.824904850982656e-05,
+      "loss": 0.2709,
+      "mean_token_accuracy": 0.8897792816162109,
+      "num_tokens": 1164190.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.1186745762825012,
+      "epoch": 0.8349705304518664,
+      "grad_norm": 3.2434158325195312,
+      "learning_rate": 6.749050304576415e-05,
+      "loss": 0.2704,
+      "mean_token_accuracy": 0.8963826775550843,
+      "num_tokens": 1177346.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.1046124696731567,
+      "epoch": 0.8447937131630648,
+      "grad_norm": 3.74926495552063,
+      "learning_rate": 6.672734249827028e-05,
+      "loss": 0.2549,
+      "mean_token_accuracy": 0.9034735262393951,
+      "num_tokens": 1190692.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.1273367881774903,
+      "epoch": 0.8546168958742633,
+      "grad_norm": 3.804595708847046,
+      "learning_rate": 6.59597682366017e-05,
+      "loss": 0.2897,
+      "mean_token_accuracy": 0.8861363589763641,
+      "num_tokens": 1204254.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.1317849993705749,
+      "epoch": 0.8644400785854617,
+      "grad_norm": 3.894453525543213,
+      "learning_rate": 6.518798279462766e-05,
+      "loss": 0.2925,
+      "mean_token_accuracy": 0.8857719779014588,
+      "num_tokens": 1217938.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.1314242124557494,
+      "epoch": 0.8742632612966601,
+      "grad_norm": 2.416104316711426,
+      "learning_rate": 6.441218981738895e-05,
+      "loss": 0.269,
+      "mean_token_accuracy": 0.8959510207176209,
+      "num_tokens": 1231589.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.113207757472992,
+      "epoch": 0.8840864440078585,
+      "grad_norm": 3.2071354389190674,
+      "learning_rate": 6.363259400736359e-05,
+      "loss": 0.2444,
+      "mean_token_accuracy": 0.901242709159851,
+      "num_tokens": 1245755.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.1054975628852843,
+      "epoch": 0.8939096267190569,
+      "grad_norm": 4.069499969482422,
+      "learning_rate": 6.284940107045363e-05,
+      "loss": 0.2731,
+      "mean_token_accuracy": 0.8967300713062286,
+      "num_tokens": 1259830.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.1187020659446716,
+      "epoch": 0.9037328094302554,
+      "grad_norm": 3.0500526428222656,
+      "learning_rate": 6.206281766170706e-05,
+      "loss": 0.2856,
+      "mean_token_accuracy": 0.8880673527717591,
+      "num_tokens": 1274186.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.114321768283844,
+      "epoch": 0.9135559921414538,
+      "grad_norm": 3.7660529613494873,
+      "learning_rate": 6.127305133078945e-05,
+      "loss": 0.2612,
+      "mean_token_accuracy": 0.8977824032306672,
+      "num_tokens": 1288175.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.0940532207489013,
+      "epoch": 0.9233791748526523,
+      "grad_norm": 4.28970193862915,
+      "learning_rate": 6.048031046721931e-05,
+      "loss": 0.2567,
+      "mean_token_accuracy": 0.9000320255756378,
+      "num_tokens": 1301979.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.1103707790374755,
+      "epoch": 0.9332023575638507,
+      "grad_norm": 4.463717937469482,
+      "learning_rate": 5.9684804245382056e-05,
+      "loss": 0.2673,
+      "mean_token_accuracy": 0.8979255378246307,
+      "num_tokens": 1316012.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.110783302783966,
+      "epoch": 0.9430255402750491,
+      "grad_norm": 3.9348270893096924,
+      "learning_rate": 5.888674256933671e-05,
+      "loss": 0.2687,
+      "mean_token_accuracy": 0.8897039651870727,
+      "num_tokens": 1329560.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.1314164638519286,
+      "epoch": 0.9528487229862476,
+      "grad_norm": 3.4531164169311523,
+      "learning_rate": 5.80863360174302e-05,
+      "loss": 0.2876,
+      "mean_token_accuracy": 0.885741651058197,
+      "num_tokens": 1343980.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.1194223284721374,
+      "epoch": 0.962671905697446,
+      "grad_norm": 3.469637155532837,
+      "learning_rate": 5.72837957867337e-05,
+      "loss": 0.2639,
+      "mean_token_accuracy": 0.8971620917320251,
+      "num_tokens": 1358209.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.100822639465332,
+      "epoch": 0.9724950884086444,
+      "grad_norm": 3.783304214477539,
+      "learning_rate": 5.64793336373156e-05,
+      "loss": 0.2481,
+      "mean_token_accuracy": 0.9017934381961823,
+      "num_tokens": 1371263.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.099805998802185,
+      "epoch": 0.9823182711198428,
+      "grad_norm": 2.844398260116577,
+      "learning_rate": 5.567316183636615e-05,
+      "loss": 0.2123,
+      "mean_token_accuracy": 0.9206562280654907,
+      "num_tokens": 1385183.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.0888682007789612,
+      "epoch": 0.9921414538310412,
+      "grad_norm": 3.4088380336761475,
+      "learning_rate": 5.4865493102188046e-05,
+      "loss": 0.2378,
+      "mean_token_accuracy": 0.9085430264472961,
+      "num_tokens": 1398309.0,
+      "step": 1010
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2036,
+  "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_s2_10k/checkpoint-1018/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e75a9aefb996ad2b21a80d647b9fc05cf69d9f4afc9b9014cda7fcb78f6cd8a
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_10k/checkpoint-2036/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_s2_10k/checkpoint-2036/adapter_config.json

@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.0,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "gate_proj",
+    "k_proj",
+    "q_proj",
+    "o_proj",
+    "down_proj",
+    "v_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_s2_10k/checkpoint-2036/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f197a52b3b8ff881f1532c383d8a17236c52fa9328a43e55635bc65863bff6ac
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_10k/checkpoint-2036/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_s2_10k/checkpoint-2036/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_s2_10k/checkpoint-2036/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_s2_10k/checkpoint-2036/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_s2_10k/checkpoint-2036/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_s2_10k/checkpoint-2036/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": 2036,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "entropy": 1.2257721304893494,
+      "epoch": 0.009823182711198428,
+      "grad_norm": 4.894117832183838,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.6909,
+      "mean_token_accuracy": 0.775017136335373,
+      "num_tokens": 13828.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.2686485409736634,
+      "epoch": 0.019646365422396856,
+      "grad_norm": 2.6351609230041504,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.7667,
+      "mean_token_accuracy": 0.732726263999939,
+      "num_tokens": 28074.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.2789716601371766,
+      "epoch": 0.029469548133595286,
+      "grad_norm": 1.8091380596160889,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.672,
+      "mean_token_accuracy": 0.7583289265632629,
+      "num_tokens": 42168.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.247054147720337,
+      "epoch": 0.03929273084479371,
+      "grad_norm": 2.2216796875,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.6733,
+      "mean_token_accuracy": 0.7530835151672364,
+      "num_tokens": 56700.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.2091697812080384,
+      "epoch": 0.04911591355599214,
+      "grad_norm": 2.6457648277282715,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.559,
+      "mean_token_accuracy": 0.7906185984611511,
+      "num_tokens": 70922.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.2023007988929748,
+      "epoch": 0.05893909626719057,
+      "grad_norm": 2.2835278511047363,
+      "learning_rate": 5.784313725490197e-05,
+      "loss": 0.5012,
+      "mean_token_accuracy": 0.8101553797721863,
+      "num_tokens": 84911.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.1628074526786805,
+      "epoch": 0.068762278978389,
+      "grad_norm": 2.8187313079833984,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 0.5312,
+      "mean_token_accuracy": 0.8078348159790039,
+      "num_tokens": 98835.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.173559021949768,
+      "epoch": 0.07858546168958742,
+      "grad_norm": 3.0911595821380615,
+      "learning_rate": 7.745098039215687e-05,
+      "loss": 0.4827,
+      "mean_token_accuracy": 0.8213136911392211,
+      "num_tokens": 112502.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.1246683597564697,
+      "epoch": 0.08840864440078586,
+      "grad_norm": 2.5103771686553955,
+      "learning_rate": 8.725490196078432e-05,
+      "loss": 0.4858,
+      "mean_token_accuracy": 0.8161003708839416,
+      "num_tokens": 126248.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.141631579399109,
+      "epoch": 0.09823182711198428,
+      "grad_norm": 2.9426238536834717,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 0.4759,
+      "mean_token_accuracy": 0.8214390873908997,
+      "num_tokens": 139551.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.1160478115081787,
+      "epoch": 0.10805500982318271,
+      "grad_norm": 3.039224624633789,
+      "learning_rate": 9.999676765122786e-05,
+      "loss": 0.45,
+      "mean_token_accuracy": 0.8267141044139862,
+      "num_tokens": 152975.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.1066530227661133,
+      "epoch": 0.11787819253438114,
+      "grad_norm": 2.6248061656951904,
+      "learning_rate": 9.998093674494531e-05,
+      "loss": 0.4723,
+      "mean_token_accuracy": 0.8254820048809052,
+      "num_tokens": 167073.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.1236253738403321,
+      "epoch": 0.12770137524557956,
+      "grad_norm": 2.6021196842193604,
+      "learning_rate": 9.995191775639568e-05,
+      "loss": 0.4827,
+      "mean_token_accuracy": 0.8199610054492951,
+      "num_tokens": 181404.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.1052476644515992,
+      "epoch": 0.137524557956778,
+      "grad_norm": 2.3915598392486572,
+      "learning_rate": 9.990971834259444e-05,
+      "loss": 0.453,
+      "mean_token_accuracy": 0.8272700130939483,
+      "num_tokens": 195441.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.1267707705497743,
+      "epoch": 0.14734774066797643,
+      "grad_norm": 3.4826769828796387,
+      "learning_rate": 9.985434963837376e-05,
+      "loss": 0.4805,
+      "mean_token_accuracy": 0.8141039609909058,
+      "num_tokens": 209863.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.0815376162528991,
+      "epoch": 0.15717092337917485,
+      "grad_norm": 3.369027614593506,
+      "learning_rate": 9.978582625344435e-05,
+      "loss": 0.4439,
+      "mean_token_accuracy": 0.833777642250061,
+      "num_tokens": 223646.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.093427562713623,
+      "epoch": 0.16699410609037327,
+      "grad_norm": 3.097435235977173,
+      "learning_rate": 9.970416626854049e-05,
+      "loss": 0.4171,
+      "mean_token_accuracy": 0.8488226413726807,
+      "num_tokens": 237587.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.0916131377220153,
+      "epoch": 0.17681728880157171,
+      "grad_norm": 2.9525489807128906,
+      "learning_rate": 9.960939123064929e-05,
+      "loss": 0.4524,
+      "mean_token_accuracy": 0.8301271557807922,
+      "num_tokens": 251760.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.0759238362312318,
+      "epoch": 0.18664047151277013,
+      "grad_norm": 2.6859591007232666,
+      "learning_rate": 9.950152614732518e-05,
+      "loss": 0.4167,
+      "mean_token_accuracy": 0.8416622459888459,
+      "num_tokens": 265417.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.0830464363098145,
+      "epoch": 0.19646365422396855,
+      "grad_norm": 3.650442123413086,
+      "learning_rate": 9.93805994800914e-05,
+      "loss": 0.4541,
+      "mean_token_accuracy": 0.8313074588775635,
+      "num_tokens": 278540.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.0821943521499633,
+      "epoch": 0.206286836935167,
+      "grad_norm": 3.401404619216919,
+      "learning_rate": 9.924664313693006e-05,
+      "loss": 0.435,
+      "mean_token_accuracy": 0.8389021277427673,
+      "num_tokens": 292295.0,
+      "step": 210
+    },
+    {
+      "entropy": 1.0802088141441346,
+      "epoch": 0.21611001964636542,
+      "grad_norm": 3.133098840713501,
+      "learning_rate": 9.909969246386284e-05,
+      "loss": 0.4309,
+      "mean_token_accuracy": 0.8364023327827453,
+      "num_tokens": 305466.0,
+      "step": 220
+    },
+    {
+      "entropy": 1.1103221774101257,
+      "epoch": 0.22593320235756384,
+      "grad_norm": 2.64292311668396,
+      "learning_rate": 9.893978623562453e-05,
+      "loss": 0.4289,
+      "mean_token_accuracy": 0.8376089572906494,
+      "num_tokens": 320495.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.0710348248481751,
+      "epoch": 0.2357563850687623,
+      "grad_norm": 2.8242721557617188,
+      "learning_rate": 9.876696664543177e-05,
+      "loss": 0.4032,
+      "mean_token_accuracy": 0.8512109398841858,
+      "num_tokens": 333901.0,
+      "step": 240
+    },
+    {
+      "entropy": 1.0859151601791381,
+      "epoch": 0.2455795677799607,
+      "grad_norm": 3.6954805850982666,
+      "learning_rate": 9.858127929384991e-05,
+      "loss": 0.3854,
+      "mean_token_accuracy": 0.8480402171611786,
+      "num_tokens": 347392.0,
+      "step": 250
+    },
+    {
+      "entropy": 1.0700998306274414,
+      "epoch": 0.2554027504911591,
+      "grad_norm": 3.1757068634033203,
+      "learning_rate": 9.838277317676077e-05,
+      "loss": 0.3992,
+      "mean_token_accuracy": 0.8441122591495513,
+      "num_tokens": 361140.0,
+      "step": 260
+    },
+    {
+      "entropy": 1.0962264895439149,
+      "epoch": 0.26522593320235754,
+      "grad_norm": 2.899790048599243,
+      "learning_rate": 9.817150067243438e-05,
+      "loss": 0.359,
+      "mean_token_accuracy": 0.8624118745326996,
+      "num_tokens": 374661.0,
+      "step": 270
+    },
+    {
+      "entropy": 1.1250731945037842,
+      "epoch": 0.275049115913556,
+      "grad_norm": 2.9336721897125244,
+      "learning_rate": 9.794751752770845e-05,
+      "loss": 0.4116,
+      "mean_token_accuracy": 0.8425243020057678,
+      "num_tokens": 388415.0,
+      "step": 280
+    },
+    {
+      "entropy": 1.170729434490204,
+      "epoch": 0.28487229862475444,
+      "grad_norm": 4.345315456390381,
+      "learning_rate": 9.771088284327879e-05,
+      "loss": 0.4232,
+      "mean_token_accuracy": 0.8408111214637757,
+      "num_tokens": 402031.0,
+      "step": 290
+    },
+    {
+      "entropy": 1.1610852003097534,
+      "epoch": 0.29469548133595286,
+      "grad_norm": 3.390840768814087,
+      "learning_rate": 9.746165905810488e-05,
+      "loss": 0.376,
+      "mean_token_accuracy": 0.85899418592453,
+      "num_tokens": 416165.0,
+      "step": 300
+    },
+    {
+      "entropy": 1.1845875024795531,
+      "epoch": 0.3045186640471513,
+      "grad_norm": 2.6109654903411865,
+      "learning_rate": 9.719991193293466e-05,
+      "loss": 0.4143,
+      "mean_token_accuracy": 0.8462965011596679,
+      "num_tokens": 429910.0,
+      "step": 310
+    },
+    {
+      "entropy": 1.1857431888580323,
+      "epoch": 0.3143418467583497,
+      "grad_norm": 3.2335426807403564,
+      "learning_rate": 9.692571053295268e-05,
+      "loss": 0.4037,
+      "mean_token_accuracy": 0.8504281759262085,
+      "num_tokens": 443725.0,
+      "step": 320
+    },
+    {
+      "entropy": 1.179654860496521,
+      "epoch": 0.3241650294695481,
+      "grad_norm": 3.481412649154663,
+      "learning_rate": 9.663912720955637e-05,
+      "loss": 0.3978,
+      "mean_token_accuracy": 0.8510401785373688,
+      "num_tokens": 458234.0,
+      "step": 330
+    },
+    {
+      "entropy": 1.1839690446853637,
+      "epoch": 0.33398821218074654,
+      "grad_norm": 3.3959858417510986,
+      "learning_rate": 9.634023758126539e-05,
+      "loss": 0.375,
+      "mean_token_accuracy": 0.8541060447692871,
+      "num_tokens": 471402.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.1508894562721252,
+      "epoch": 0.343811394891945,
+      "grad_norm": 3.046173334121704,
+      "learning_rate": 9.602912051376859e-05,
+      "loss": 0.3701,
+      "mean_token_accuracy": 0.8562622129917145,
+      "num_tokens": 484651.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.1543931007385253,
+      "epoch": 0.35363457760314343,
+      "grad_norm": 2.7381064891815186,
+      "learning_rate": 9.570585809911441e-05,
+      "loss": 0.3755,
+      "mean_token_accuracy": 0.8545270740985871,
+      "num_tokens": 497956.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.1635468125343322,
+      "epoch": 0.36345776031434185,
+      "grad_norm": 2.425509214401245,
+      "learning_rate": 9.537053563404992e-05,
+      "loss": 0.388,
+      "mean_token_accuracy": 0.8558545649051666,
+      "num_tokens": 511717.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.1691488862037658,
+      "epoch": 0.37328094302554027,
+      "grad_norm": 4.111085414886475,
+      "learning_rate": 9.502324159751412e-05,
+      "loss": 0.3499,
+      "mean_token_accuracy": 0.8601330816745758,
+      "num_tokens": 525633.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.1114183068275452,
+      "epoch": 0.3831041257367387,
+      "grad_norm": 2.738142251968384,
+      "learning_rate": 9.466406762729181e-05,
+      "loss": 0.3502,
+      "mean_token_accuracy": 0.8674743235111236,
+      "num_tokens": 538945.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.111063504219055,
+      "epoch": 0.3929273084479371,
+      "grad_norm": 3.1340293884277344,
+      "learning_rate": 9.429310849583378e-05,
+      "loss": 0.3669,
+      "mean_token_accuracy": 0.8642211854457855,
+      "num_tokens": 552861.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.1472705245018004,
+      "epoch": 0.4027504911591356,
+      "grad_norm": 3.1255431175231934,
+      "learning_rate": 9.391046208524989e-05,
+      "loss": 0.3978,
+      "mean_token_accuracy": 0.8444820404052734,
+      "num_tokens": 565955.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.1532312989234925,
+      "epoch": 0.412573673870334,
+      "grad_norm": 2.9525880813598633,
+      "learning_rate": 9.351622936148182e-05,
+      "loss": 0.348,
+      "mean_token_accuracy": 0.8709490835666657,
+      "num_tokens": 579923.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.136536467075348,
+      "epoch": 0.4223968565815324,
+      "grad_norm": 2.7755491733551025,
+      "learning_rate": 9.31105143476618e-05,
+      "loss": 0.3534,
+      "mean_token_accuracy": 0.8653653502464295,
+      "num_tokens": 594071.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.1585060477256774,
+      "epoch": 0.43222003929273084,
+      "grad_norm": 3.2028636932373047,
+      "learning_rate": 9.269342409666495e-05,
+      "loss": 0.3332,
+      "mean_token_accuracy": 0.8707658588886261,
+      "num_tokens": 608545.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.1323098540306091,
+      "epoch": 0.44204322200392926,
+      "grad_norm": 2.7898449897766113,
+      "learning_rate": 9.226506866286202e-05,
+      "loss": 0.3398,
+      "mean_token_accuracy": 0.8691294670104981,
+      "num_tokens": 621975.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.1309045553207397,
+      "epoch": 0.4518664047151277,
+      "grad_norm": 3.5157182216644287,
+      "learning_rate": 9.18255610730802e-05,
+      "loss": 0.3541,
+      "mean_token_accuracy": 0.8666610240936279,
+      "num_tokens": 636061.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.1365639805793761,
+      "epoch": 0.46168958742632615,
+      "grad_norm": 2.39361572265625,
+      "learning_rate": 9.137501729677962e-05,
+      "loss": 0.3504,
+      "mean_token_accuracy": 0.8678753793239593,
+      "num_tokens": 649717.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.1228155851364137,
+      "epoch": 0.4715127701375246,
+      "grad_norm": 3.2366058826446533,
+      "learning_rate": 9.091355621545339e-05,
+      "loss": 0.3231,
+      "mean_token_accuracy": 0.8785337209701538,
+      "num_tokens": 664123.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.1357950210571288,
+      "epoch": 0.481335952848723,
+      "grad_norm": 3.25319242477417,
+      "learning_rate": 9.044129959125923e-05,
+      "loss": 0.3638,
+      "mean_token_accuracy": 0.8627843201160431,
+      "num_tokens": 677803.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.1405863523483277,
+      "epoch": 0.4911591355599214,
+      "grad_norm": 3.28054141998291,
+      "learning_rate": 8.995837203489108e-05,
+      "loss": 0.3151,
+      "mean_token_accuracy": 0.8844160795211792,
+      "num_tokens": 691461.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.1509727835655212,
+      "epoch": 0.5009823182711198,
+      "grad_norm": 3.824949264526367,
+      "learning_rate": 8.946490097269899e-05,
+      "loss": 0.3421,
+      "mean_token_accuracy": 0.872158533334732,
+      "num_tokens": 705470.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.155228316783905,
+      "epoch": 0.5108055009823183,
+      "grad_norm": 3.0210299491882324,
+      "learning_rate": 8.896101661306618e-05,
+      "loss": 0.3631,
+      "mean_token_accuracy": 0.8582519233226776,
+      "num_tokens": 718805.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.1403202533721923,
+      "epoch": 0.5206286836935167,
+      "grad_norm": 2.491917371749878,
+      "learning_rate": 8.844685191205193e-05,
+      "loss": 0.3376,
+      "mean_token_accuracy": 0.8696144163608551,
+      "num_tokens": 732367.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.1386663675308228,
+      "epoch": 0.5304518664047151,
+      "grad_norm": 3.0767979621887207,
+      "learning_rate": 8.792254253830945e-05,
+      "loss": 0.3251,
+      "mean_token_accuracy": 0.8764946639537812,
+      "num_tokens": 746431.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.1174968481063843,
+      "epoch": 0.5402750491159135,
+      "grad_norm": 2.6364076137542725,
+      "learning_rate": 8.738822683728814e-05,
+      "loss": 0.3105,
+      "mean_token_accuracy": 0.8814369976520539,
+      "num_tokens": 760182.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.1403401017189025,
+      "epoch": 0.550098231827112,
+      "grad_norm": 2.6279711723327637,
+      "learning_rate": 8.684404579472936e-05,
+      "loss": 0.315,
+      "mean_token_accuracy": 0.8731137156486511,
+      "num_tokens": 774137.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.1179575324058533,
+      "epoch": 0.5599214145383105,
+      "grad_norm": 4.026428699493408,
+      "learning_rate": 8.629014299946568e-05,
+      "loss": 0.3162,
+      "mean_token_accuracy": 0.8734814465045929,
+      "num_tokens": 787806.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.1278663873672485,
+      "epoch": 0.5697445972495089,
+      "grad_norm": 3.3344717025756836,
+      "learning_rate": 8.572666460553316e-05,
+      "loss": 0.3149,
+      "mean_token_accuracy": 0.8808867335319519,
+      "num_tokens": 801431.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.122735869884491,
+      "epoch": 0.5795677799607073,
+      "grad_norm": 2.997509479522705,
+      "learning_rate": 8.515375929360687e-05,
+      "loss": 0.3248,
+      "mean_token_accuracy": 0.8767118990421295,
+      "num_tokens": 815080.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.1275416135787963,
+      "epoch": 0.5893909626719057,
+      "grad_norm": 4.02539587020874,
+      "learning_rate": 8.45715782317696e-05,
+      "loss": 0.337,
+      "mean_token_accuracy": 0.8703737854957581,
+      "num_tokens": 829061.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.1480388522148133,
+      "epoch": 0.5992141453831041,
+      "grad_norm": 3.0496115684509277,
+      "learning_rate": 8.39802750356244e-05,
+      "loss": 0.3401,
+      "mean_token_accuracy": 0.8724706113338471,
+      "num_tokens": 842325.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.1545992732048034,
+      "epoch": 0.6090373280943026,
+      "grad_norm": 3.738725423812866,
+      "learning_rate": 8.338000572776114e-05,
+      "loss": 0.3144,
+      "mean_token_accuracy": 0.8776861608028412,
+      "num_tokens": 856377.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.1416390776634215,
+      "epoch": 0.618860510805501,
+      "grad_norm": 3.837313652038574,
+      "learning_rate": 8.277092869658808e-05,
+      "loss": 0.2838,
+      "mean_token_accuracy": 0.8879789412021637,
+      "num_tokens": 870502.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.1548494458198548,
+      "epoch": 0.6286836935166994,
+      "grad_norm": 2.7274272441864014,
+      "learning_rate": 8.215320465453914e-05,
+      "loss": 0.3315,
+      "mean_token_accuracy": 0.875211751461029,
+      "num_tokens": 884708.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.1428229212760925,
+      "epoch": 0.6385068762278978,
+      "grad_norm": 2.581369638442993,
+      "learning_rate": 8.152699659566797e-05,
+      "loss": 0.3126,
+      "mean_token_accuracy": 0.8831508219242096,
+      "num_tokens": 898142.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.1344876527786254,
+      "epoch": 0.6483300589390962,
+      "grad_norm": 2.903346061706543,
+      "learning_rate": 8.089246975263996e-05,
+      "loss": 0.3066,
+      "mean_token_accuracy": 0.882864135503769,
+      "num_tokens": 912820.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.1322418928146363,
+      "epoch": 0.6581532416502947,
+      "grad_norm": 3.234118938446045,
+      "learning_rate": 8.024979155313354e-05,
+      "loss": 0.2959,
+      "mean_token_accuracy": 0.8836028277873993,
+      "num_tokens": 926399.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.1417956709861756,
+      "epoch": 0.6679764243614931,
+      "grad_norm": 2.977431535720825,
+      "learning_rate": 7.959913157566246e-05,
+      "loss": 0.3271,
+      "mean_token_accuracy": 0.8776575326919556,
+      "num_tokens": 940133.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.1574132323265076,
+      "epoch": 0.6777996070726916,
+      "grad_norm": 3.9751768112182617,
+      "learning_rate": 7.894066150483023e-05,
+      "loss": 0.3056,
+      "mean_token_accuracy": 0.8853654146194458,
+      "num_tokens": 954504.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.1295910596847534,
+      "epoch": 0.68762278978389,
+      "grad_norm": 3.3603923320770264,
+      "learning_rate": 7.827455508602923e-05,
+      "loss": 0.3389,
+      "mean_token_accuracy": 0.8706278026103973,
+      "num_tokens": 968934.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.1322409749031066,
+      "epoch": 0.6974459724950884,
+      "grad_norm": 2.7524755001068115,
+      "learning_rate": 7.760098807959573e-05,
+      "loss": 0.335,
+      "mean_token_accuracy": 0.8757263839244842,
+      "num_tokens": 983038.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.1349270939826965,
+      "epoch": 0.7072691552062869,
+      "grad_norm": 3.783806800842285,
+      "learning_rate": 7.69201382144334e-05,
+      "loss": 0.3167,
+      "mean_token_accuracy": 0.8799357235431671,
+      "num_tokens": 996748.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.130021619796753,
+      "epoch": 0.7170923379174853,
+      "grad_norm": 2.9027302265167236,
+      "learning_rate": 7.623218514111743e-05,
+      "loss": 0.2839,
+      "mean_token_accuracy": 0.8918920040130616,
+      "num_tokens": 1010595.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.1162713766098022,
+      "epoch": 0.7269155206286837,
+      "grad_norm": 3.68827486038208,
+      "learning_rate": 7.553731038449144e-05,
+      "loss": 0.2762,
+      "mean_token_accuracy": 0.8936041057109833,
+      "num_tokens": 1024890.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.1368697047233582,
+      "epoch": 0.7367387033398821,
+      "grad_norm": 3.653467893600464,
+      "learning_rate": 7.483569729576996e-05,
+      "loss": 0.2924,
+      "mean_token_accuracy": 0.8875188827514648,
+      "num_tokens": 1038967.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.1225467801094056,
+      "epoch": 0.7465618860510805,
+      "grad_norm": 3.6799514293670654,
+      "learning_rate": 7.412753100415897e-05,
+      "loss": 0.2776,
+      "mean_token_accuracy": 0.8891165137290955,
+      "num_tokens": 1052622.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.1493797302246094,
+      "epoch": 0.756385068762279,
+      "grad_norm": 2.992457389831543,
+      "learning_rate": 7.341299836800729e-05,
+      "loss": 0.3057,
+      "mean_token_accuracy": 0.886419826745987,
+      "num_tokens": 1066588.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.1170050144195556,
+      "epoch": 0.7662082514734774,
+      "grad_norm": 3.467043161392212,
+      "learning_rate": 7.26922879255017e-05,
+      "loss": 0.2458,
+      "mean_token_accuracy": 0.905247437953949,
+      "num_tokens": 1080453.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.1004221796989442,
+      "epoch": 0.7760314341846758,
+      "grad_norm": 2.8668010234832764,
+      "learning_rate": 7.196558984491889e-05,
+      "loss": 0.2955,
+      "mean_token_accuracy": 0.8834676742553711,
+      "num_tokens": 1094550.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.1233540773391724,
+      "epoch": 0.7858546168958742,
+      "grad_norm": 3.163811445236206,
+      "learning_rate": 7.123309587444711e-05,
+      "loss": 0.2667,
+      "mean_token_accuracy": 0.8962497293949128,
+      "num_tokens": 1108808.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.1172690033912658,
+      "epoch": 0.7956777996070727,
+      "grad_norm": 2.911059617996216,
+      "learning_rate": 7.04949992915912e-05,
+      "loss": 0.2584,
+      "mean_token_accuracy": 0.9001988768577576,
+      "num_tokens": 1122365.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.124155080318451,
+      "epoch": 0.8055009823182712,
+      "grad_norm": 3.730689287185669,
+      "learning_rate": 6.975149485217388e-05,
+      "loss": 0.3021,
+      "mean_token_accuracy": 0.8832005798816681,
+      "num_tokens": 1136376.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.1060636997222901,
+      "epoch": 0.8153241650294696,
+      "grad_norm": 2.803528070449829,
+      "learning_rate": 6.900277873894705e-05,
+      "loss": 0.2887,
+      "mean_token_accuracy": 0.8805433213710785,
+      "num_tokens": 1150011.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.1344504833221436,
+      "epoch": 0.825147347740668,
+      "grad_norm": 2.9840211868286133,
+      "learning_rate": 6.824904850982656e-05,
+      "loss": 0.2709,
+      "mean_token_accuracy": 0.8897792816162109,
+      "num_tokens": 1164190.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.1186745762825012,
+      "epoch": 0.8349705304518664,
+      "grad_norm": 3.2434158325195312,
+      "learning_rate": 6.749050304576415e-05,
+      "loss": 0.2704,
+      "mean_token_accuracy": 0.8963826775550843,
+      "num_tokens": 1177346.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.1046124696731567,
+      "epoch": 0.8447937131630648,
+      "grad_norm": 3.74926495552063,
+      "learning_rate": 6.672734249827028e-05,
+      "loss": 0.2549,
+      "mean_token_accuracy": 0.9034735262393951,
+      "num_tokens": 1190692.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.1273367881774903,
+      "epoch": 0.8546168958742633,
+      "grad_norm": 3.804595708847046,
+      "learning_rate": 6.59597682366017e-05,
+      "loss": 0.2897,
+      "mean_token_accuracy": 0.8861363589763641,
+      "num_tokens": 1204254.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.1317849993705749,
+      "epoch": 0.8644400785854617,
+      "grad_norm": 3.894453525543213,
+      "learning_rate": 6.518798279462766e-05,
+      "loss": 0.2925,
+      "mean_token_accuracy": 0.8857719779014588,
+      "num_tokens": 1217938.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.1314242124557494,
+      "epoch": 0.8742632612966601,
+      "grad_norm": 2.416104316711426,
+      "learning_rate": 6.441218981738895e-05,
+      "loss": 0.269,
+      "mean_token_accuracy": 0.8959510207176209,
+      "num_tokens": 1231589.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.113207757472992,
+      "epoch": 0.8840864440078585,
+      "grad_norm": 3.2071354389190674,
+      "learning_rate": 6.363259400736359e-05,
+      "loss": 0.2444,
+      "mean_token_accuracy": 0.901242709159851,
+      "num_tokens": 1245755.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.1054975628852843,
+      "epoch": 0.8939096267190569,
+      "grad_norm": 4.069499969482422,
+      "learning_rate": 6.284940107045363e-05,
+      "loss": 0.2731,
+      "mean_token_accuracy": 0.8967300713062286,
+      "num_tokens": 1259830.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.1187020659446716,
+      "epoch": 0.9037328094302554,
+      "grad_norm": 3.0500526428222656,
+      "learning_rate": 6.206281766170706e-05,
+      "loss": 0.2856,
+      "mean_token_accuracy": 0.8880673527717591,
+      "num_tokens": 1274186.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.114321768283844,
+      "epoch": 0.9135559921414538,
+      "grad_norm": 3.7660529613494873,
+      "learning_rate": 6.127305133078945e-05,
+      "loss": 0.2612,
+      "mean_token_accuracy": 0.8977824032306672,
+      "num_tokens": 1288175.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.0940532207489013,
+      "epoch": 0.9233791748526523,
+      "grad_norm": 4.28970193862915,
+      "learning_rate": 6.048031046721931e-05,
+      "loss": 0.2567,
+      "mean_token_accuracy": 0.9000320255756378,
+      "num_tokens": 1301979.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.1103707790374755,
+      "epoch": 0.9332023575638507,
+      "grad_norm": 4.463717937469482,
+      "learning_rate": 5.9684804245382056e-05,
+      "loss": 0.2673,
+      "mean_token_accuracy": 0.8979255378246307,
+      "num_tokens": 1316012.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.110783302783966,
+      "epoch": 0.9430255402750491,
+      "grad_norm": 3.9348270893096924,
+      "learning_rate": 5.888674256933671e-05,
+      "loss": 0.2687,
+      "mean_token_accuracy": 0.8897039651870727,
+      "num_tokens": 1329560.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.1314164638519286,
+      "epoch": 0.9528487229862476,
+      "grad_norm": 3.4531164169311523,
+      "learning_rate": 5.80863360174302e-05,
+      "loss": 0.2876,
+      "mean_token_accuracy": 0.885741651058197,
+      "num_tokens": 1343980.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.1194223284721374,
+      "epoch": 0.962671905697446,
+      "grad_norm": 3.469637155532837,
+      "learning_rate": 5.72837957867337e-05,
+      "loss": 0.2639,
+      "mean_token_accuracy": 0.8971620917320251,
+      "num_tokens": 1358209.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.100822639465332,
+      "epoch": 0.9724950884086444,
+      "grad_norm": 3.783304214477539,
+      "learning_rate": 5.64793336373156e-05,
+      "loss": 0.2481,
+      "mean_token_accuracy": 0.9017934381961823,
+      "num_tokens": 1371263.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.099805998802185,
+      "epoch": 0.9823182711198428,
+      "grad_norm": 2.844398260116577,
+      "learning_rate": 5.567316183636615e-05,
+      "loss": 0.2123,
+      "mean_token_accuracy": 0.9206562280654907,
+      "num_tokens": 1385183.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.0888682007789612,
+      "epoch": 0.9921414538310412,
+      "grad_norm": 3.4088380336761475,
+      "learning_rate": 5.4865493102188046e-05,
+      "loss": 0.2378,
+      "mean_token_accuracy": 0.9085430264472961,
+      "num_tokens": 1398309.0,
+      "step": 1010
+    },
+    {
+      "entropy": 1.0874920725822448,
+      "epoch": 1.0019646365422397,
+      "grad_norm": 3.8088834285736084,
+      "learning_rate": 5.405654054806817e-05,
+      "loss": 0.2393,
+      "mean_token_accuracy": 0.90903941988945,
+      "num_tokens": 1411126.0,
+      "step": 1020
+    },
+    {
+      "entropy": 1.0976500988006592,
+      "epoch": 1.0117878192534382,
+      "grad_norm": 3.365924596786499,
+      "learning_rate": 5.3246517626045025e-05,
+      "loss": 0.1696,
+      "mean_token_accuracy": 0.9377715766429902,
+      "num_tokens": 1425582.0,
+      "step": 1030
+    },
+    {
+      "entropy": 1.059966003894806,
+      "epoch": 1.0216110019646365,
+      "grad_norm": 2.533733606338501,
+      "learning_rate": 5.243563807058676e-05,
+      "loss": 0.176,
+      "mean_token_accuracy": 0.9372439026832581,
+      "num_tokens": 1439527.0,
+      "step": 1040
+    },
+    {
+      "entropy": 1.067211651802063,
+      "epoch": 1.031434184675835,
+      "grad_norm": 3.2963051795959473,
+      "learning_rate": 5.162411584219476e-05,
+      "loss": 0.164,
+      "mean_token_accuracy": 0.9329432904720306,
+      "num_tokens": 1453445.0,
+      "step": 1050
+    },
+    {
+      "entropy": 1.05519642829895,
+      "epoch": 1.0412573673870333,
+      "grad_norm": 3.6632964611053467,
+      "learning_rate": 5.0812165070947594e-05,
+      "loss": 0.1599,
+      "mean_token_accuracy": 0.942095935344696,
+      "num_tokens": 1466353.0,
+      "step": 1060
+    },
+    {
+      "entropy": 1.0650439381599426,
+      "epoch": 1.0510805500982319,
+      "grad_norm": 3.6676483154296875,
+      "learning_rate": 5e-05,
+      "loss": 0.1666,
+      "mean_token_accuracy": 0.9373338699340821,
+      "num_tokens": 1481062.0,
+      "step": 1070
+    },
+    {
+      "entropy": 1.0560135245323181,
+      "epoch": 1.0609037328094302,
+      "grad_norm": 2.3156352043151855,
+      "learning_rate": 4.918783492905242e-05,
+      "loss": 0.1583,
+      "mean_token_accuracy": 0.9429939210414886,
+      "num_tokens": 1495189.0,
+      "step": 1080
+    },
+    {
+      "entropy": 1.0576743602752685,
+      "epoch": 1.0707269155206287,
+      "grad_norm": 3.9882516860961914,
+      "learning_rate": 4.837588415780524e-05,
+      "loss": 0.1638,
+      "mean_token_accuracy": 0.9397882223129272,
+      "num_tokens": 1508841.0,
+      "step": 1090
+    },
+    {
+      "entropy": 1.0511822342872619,
+      "epoch": 1.080550098231827,
+      "grad_norm": 2.89467453956604,
+      "learning_rate": 4.756436192941325e-05,
+      "loss": 0.1698,
+      "mean_token_accuracy": 0.9373804211616517,
+      "num_tokens": 1522922.0,
+      "step": 1100
+    },
+    {
+      "entropy": 1.0662824392318726,
+      "epoch": 1.0903732809430255,
+      "grad_norm": 3.932100534439087,
+      "learning_rate": 4.675348237395499e-05,
+      "loss": 0.1618,
+      "mean_token_accuracy": 0.941525673866272,
+      "num_tokens": 1536867.0,
+      "step": 1110
+    },
+    {
+      "entropy": 1.0417114853858949,
+      "epoch": 1.1001964636542239,
+      "grad_norm": 2.551128625869751,
+      "learning_rate": 4.594345945193183e-05,
+      "loss": 0.1414,
+      "mean_token_accuracy": 0.9464017212390899,
+      "num_tokens": 1550230.0,
+      "step": 1120
+    },
+    {
+      "entropy": 1.049095356464386,
+      "epoch": 1.1100196463654224,
+      "grad_norm": 4.011430740356445,
+      "learning_rate": 4.513450689781195e-05,
+      "loss": 0.1831,
+      "mean_token_accuracy": 0.9334302544593811,
+      "num_tokens": 1562941.0,
+      "step": 1130
+    },
+    {
+      "entropy": 1.0508921027183533,
+      "epoch": 1.119842829076621,
+      "grad_norm": 3.2847137451171875,
+      "learning_rate": 4.432683816363386e-05,
+      "loss": 0.1564,
+      "mean_token_accuracy": 0.939012598991394,
+      "num_tokens": 1576546.0,
+      "step": 1140
+    },
+    {
+      "entropy": 1.0502169907093049,
+      "epoch": 1.1296660117878192,
+      "grad_norm": 3.075655221939087,
+      "learning_rate": 4.3520666362684395e-05,
+      "loss": 0.153,
+      "mean_token_accuracy": 0.9413822650909424,
+      "num_tokens": 1590129.0,
+      "step": 1150
+    },
+    {
+      "entropy": 1.0253897070884705,
+      "epoch": 1.1394891944990178,
+      "grad_norm": 4.298797607421875,
+      "learning_rate": 4.271620421326633e-05,
+      "loss": 0.1405,
+      "mean_token_accuracy": 0.9453379034996032,
+      "num_tokens": 1603834.0,
+      "step": 1160
+    },
+    {
+      "entropy": 1.0295288264751434,
+      "epoch": 1.149312377210216,
+      "grad_norm": 3.813027858734131,
+      "learning_rate": 4.191366398256983e-05,
+      "loss": 0.1526,
+      "mean_token_accuracy": 0.940306568145752,
+      "num_tokens": 1617951.0,
+      "step": 1170
+    },
+    {
+      "entropy": 1.0393607079982758,
+      "epoch": 1.1591355599214146,
+      "grad_norm": 3.8066794872283936,
+      "learning_rate": 4.111325743066332e-05,
+      "loss": 0.1619,
+      "mean_token_accuracy": 0.9348547995090485,
+      "num_tokens": 1631987.0,
+      "step": 1180
+    },
+    {
+      "entropy": 1.024841272830963,
+      "epoch": 1.168958742632613,
+      "grad_norm": 4.561389446258545,
+      "learning_rate": 4.0315195754617976e-05,
+      "loss": 0.153,
+      "mean_token_accuracy": 0.9441807806491852,
+      "num_tokens": 1645088.0,
+      "step": 1190
+    },
+    {
+      "entropy": 1.0458040356636047,
+      "epoch": 1.1787819253438114,
+      "grad_norm": 2.5349133014678955,
+      "learning_rate": 3.95196895327807e-05,
+      "loss": 0.1562,
+      "mean_token_accuracy": 0.9376145839691162,
+      "num_tokens": 1659227.0,
+      "step": 1200
+    },
+    {
+      "entropy": 1.0199751257896423,
+      "epoch": 1.1886051080550097,
+      "grad_norm": 2.935438632965088,
+      "learning_rate": 3.872694866921056e-05,
+      "loss": 0.1513,
+      "mean_token_accuracy": 0.9431104123592376,
+      "num_tokens": 1672545.0,
+      "step": 1210
+    },
+    {
+      "entropy": 1.0281165361404419,
+      "epoch": 1.1984282907662083,
+      "grad_norm": 3.7287099361419678,
+      "learning_rate": 3.793718233829295e-05,
+      "loss": 0.1263,
+      "mean_token_accuracy": 0.9506176173686981,
+      "num_tokens": 1686702.0,
+      "step": 1220
+    },
+    {
+      "entropy": 1.04444020986557,
+      "epoch": 1.2082514734774068,
+      "grad_norm": 3.5377602577209473,
+      "learning_rate": 3.715059892954639e-05,
+      "loss": 0.1434,
+      "mean_token_accuracy": 0.9442107617855072,
+      "num_tokens": 1701509.0,
+      "step": 1230
+    },
+    {
+      "entropy": 1.0250988066196443,
+      "epoch": 1.218074656188605,
+      "grad_norm": 3.862539291381836,
+      "learning_rate": 3.636740599263643e-05,
+      "loss": 0.1407,
+      "mean_token_accuracy": 0.9462244153022766,
+      "num_tokens": 1715330.0,
+      "step": 1240
+    },
+    {
+      "entropy": 1.0258072376251222,
+      "epoch": 1.2278978388998034,
+      "grad_norm": 3.788259267807007,
+      "learning_rate": 3.5587810182611066e-05,
+      "loss": 0.1432,
+      "mean_token_accuracy": 0.9456102013587951,
+      "num_tokens": 1728655.0,
+      "step": 1250
+    },
+    {
+      "entropy": 1.0282273948192597,
+      "epoch": 1.237721021611002,
+      "grad_norm": 4.368952751159668,
+      "learning_rate": 3.481201720537235e-05,
+      "loss": 0.1502,
+      "mean_token_accuracy": 0.942660129070282,
+      "num_tokens": 1741961.0,
+      "step": 1260
+    },
+    {
+      "entropy": 1.020354264974594,
+      "epoch": 1.2475442043222005,
+      "grad_norm": 2.97713303565979,
+      "learning_rate": 3.4040231763398315e-05,
+      "loss": 0.1398,
+      "mean_token_accuracy": 0.9403035998344421,
+      "num_tokens": 1755285.0,
+      "step": 1270
+    },
+    {
+      "entropy": 1.0320065379142762,
+      "epoch": 1.2573673870333988,
+      "grad_norm": 3.179495334625244,
+      "learning_rate": 3.3272657501729714e-05,
+      "loss": 0.1548,
+      "mean_token_accuracy": 0.942926949262619,
+      "num_tokens": 1768670.0,
+      "step": 1280
+    },
+    {
+      "entropy": 1.0284401416778564,
+      "epoch": 1.2671905697445973,
+      "grad_norm": 4.386323928833008,
+      "learning_rate": 3.250949695423585e-05,
+      "loss": 0.1396,
+      "mean_token_accuracy": 0.9499077737331391,
+      "num_tokens": 1782525.0,
+      "step": 1290
+    },
+    {
+      "entropy": 1.0335516929626465,
+      "epoch": 1.2770137524557956,
+      "grad_norm": 3.451388120651245,
+      "learning_rate": 3.175095149017344e-05,
+      "loss": 0.1489,
+      "mean_token_accuracy": 0.9403730988502502,
+      "num_tokens": 1796010.0,
+      "step": 1300
+    },
+    {
+      "entropy": 1.0306159734725953,
+      "epoch": 1.2868369351669942,
+      "grad_norm": 3.454986810684204,
+      "learning_rate": 3.099722126105295e-05,
+      "loss": 0.1561,
+      "mean_token_accuracy": 0.9426006436347961,
+      "num_tokens": 1809837.0,
+      "step": 1310
+    },
+    {
+      "entropy": 1.0271208345890046,
+      "epoch": 1.2966601178781925,
+      "grad_norm": 4.480377197265625,
+      "learning_rate": 3.0248505147826135e-05,
+      "loss": 0.1416,
+      "mean_token_accuracy": 0.9483372032642364,
+      "num_tokens": 1823951.0,
+      "step": 1320
+    },
+    {
+      "entropy": 1.028133749961853,
+      "epoch": 1.306483300589391,
+      "grad_norm": 3.253253698348999,
+      "learning_rate": 2.9505000708408826e-05,
+      "loss": 0.1456,
+      "mean_token_accuracy": 0.9432099223136902,
+      "num_tokens": 1837526.0,
+      "step": 1330
+    },
+    {
+      "entropy": 1.029479295015335,
+      "epoch": 1.3163064833005893,
+      "grad_norm": 3.18983793258667,
+      "learning_rate": 2.8766904125552917e-05,
+      "loss": 0.1255,
+      "mean_token_accuracy": 0.9517965674400329,
+      "num_tokens": 1850836.0,
+      "step": 1340
+    },
+    {
+      "entropy": 1.0400744915008544,
+      "epoch": 1.3261296660117878,
+      "grad_norm": 3.667449474334717,
+      "learning_rate": 2.8034410155081136e-05,
+      "loss": 0.1337,
+      "mean_token_accuracy": 0.9499365866184235,
+      "num_tokens": 1864986.0,
+      "step": 1350
+    },
+    {
+      "entropy": 1.0226722657680511,
+      "epoch": 1.3359528487229864,
+      "grad_norm": 2.8434932231903076,
+      "learning_rate": 2.7307712074498325e-05,
+      "loss": 0.1253,
+      "mean_token_accuracy": 0.952813309431076,
+      "num_tokens": 1878997.0,
+      "step": 1360
+    },
+    {
+      "entropy": 1.026139760017395,
+      "epoch": 1.3457760314341847,
+      "grad_norm": 4.208486080169678,
+      "learning_rate": 2.658700163199273e-05,
+      "loss": 0.1278,
+      "mean_token_accuracy": 0.9488636910915375,
+      "num_tokens": 1891941.0,
+      "step": 1370
+    },
+    {
+      "entropy": 1.0362686514854431,
+      "epoch": 1.355599214145383,
+      "grad_norm": 3.1290345191955566,
+      "learning_rate": 2.5872468995841043e-05,
+      "loss": 0.139,
+      "mean_token_accuracy": 0.9459269165992736,
+      "num_tokens": 1906484.0,
+      "step": 1380
+    },
+    {
+      "entropy": 1.0176115453243255,
+      "epoch": 1.3654223968565815,
+      "grad_norm": 3.617781400680542,
+      "learning_rate": 2.5164302704230048e-05,
+      "loss": 0.1273,
+      "mean_token_accuracy": 0.9508922457695007,
+      "num_tokens": 1919495.0,
+      "step": 1390
+    },
+    {
+      "entropy": 1.0485599637031555,
+      "epoch": 1.37524557956778,
+      "grad_norm": 2.6503260135650635,
+      "learning_rate": 2.446268961550856e-05,
+      "loss": 0.1654,
+      "mean_token_accuracy": 0.9398738384246826,
+      "num_tokens": 1933711.0,
+      "step": 1400
+    },
+    {
+      "entropy": 1.036979651451111,
+      "epoch": 1.3850687622789783,
+      "grad_norm": 3.5886576175689697,
+      "learning_rate": 2.376781485888258e-05,
+      "loss": 0.1381,
+      "mean_token_accuracy": 0.9469183206558227,
+      "num_tokens": 1947728.0,
+      "step": 1410
+    },
+    {
+      "entropy": 1.029152625799179,
+      "epoch": 1.3948919449901769,
+      "grad_norm": 2.1460208892822266,
+      "learning_rate": 2.3079861785566615e-05,
+      "loss": 0.1213,
+      "mean_token_accuracy": 0.9536780059337616,
+      "num_tokens": 1962012.0,
+      "step": 1420
+    },
+    {
+      "entropy": 1.024525225162506,
+      "epoch": 1.4047151277013752,
+      "grad_norm": 2.364670991897583,
+      "learning_rate": 2.239901192040429e-05,
+      "loss": 0.1084,
+      "mean_token_accuracy": 0.9609386205673218,
+      "num_tokens": 1976503.0,
+      "step": 1430
+    },
+    {
+      "entropy": 1.016504019498825,
+      "epoch": 1.4145383104125737,
+      "grad_norm": 3.1671574115753174,
+      "learning_rate": 2.1725444913970777e-05,
+      "loss": 0.1249,
+      "mean_token_accuracy": 0.9550647437572479,
+      "num_tokens": 1989940.0,
+      "step": 1440
+    },
+    {
+      "entropy": 1.03050377368927,
+      "epoch": 1.424361493123772,
+      "grad_norm": 3.793715238571167,
+      "learning_rate": 2.1059338495169773e-05,
+      "loss": 0.1404,
+      "mean_token_accuracy": 0.9493567109107971,
+      "num_tokens": 2003808.0,
+      "step": 1450
+    },
+    {
+      "entropy": 1.033704799413681,
+      "epoch": 1.4341846758349706,
+      "grad_norm": 4.435775279998779,
+      "learning_rate": 2.040086842433756e-05,
+      "loss": 0.1338,
+      "mean_token_accuracy": 0.945322722196579,
+      "num_tokens": 2017603.0,
+      "step": 1460
+    },
+    {
+      "entropy": 1.023215252161026,
+      "epoch": 1.4440078585461689,
+      "grad_norm": 3.4446470737457275,
+      "learning_rate": 1.9750208446866463e-05,
+      "loss": 0.1298,
+      "mean_token_accuracy": 0.9513370275497437,
+      "num_tokens": 2031231.0,
+      "step": 1470
+    },
+    {
+      "entropy": 1.022528302669525,
+      "epoch": 1.4538310412573674,
+      "grad_norm": 3.000606060028076,
+      "learning_rate": 1.9107530247360073e-05,
+      "loss": 0.1231,
+      "mean_token_accuracy": 0.952287369966507,
+      "num_tokens": 2045169.0,
+      "step": 1480
+    },
+    {
+      "entropy": 1.0275364160537719,
+      "epoch": 1.463654223968566,
+      "grad_norm": 3.9463489055633545,
+      "learning_rate": 1.8473003404332047e-05,
+      "loss": 0.1312,
+      "mean_token_accuracy": 0.951322603225708,
+      "num_tokens": 2059746.0,
+      "step": 1490
+    },
+    {
+      "entropy": 1.0260064244270324,
+      "epoch": 1.4734774066797642,
+      "grad_norm": 4.452908039093018,
+      "learning_rate": 1.7846795345460876e-05,
+      "loss": 0.1157,
+      "mean_token_accuracy": 0.9550542116165162,
+      "num_tokens": 2073333.0,
+      "step": 1500
+    },
+    {
+      "entropy": 1.0229429721832275,
+      "epoch": 1.4833005893909625,
+      "grad_norm": 2.8778462409973145,
+      "learning_rate": 1.722907130341193e-05,
+      "loss": 0.1111,
+      "mean_token_accuracy": 0.9584926903247833,
+      "num_tokens": 2087136.0,
+      "step": 1510
+    },
+    {
+      "entropy": 1.042748475074768,
+      "epoch": 1.493123772102161,
+      "grad_norm": 3.8858320713043213,
+      "learning_rate": 1.661999427223888e-05,
+      "loss": 0.1392,
+      "mean_token_accuracy": 0.9484859645366669,
+      "num_tokens": 2101853.0,
+      "step": 1520
+    },
+    {
+      "entropy": 1.0395808577537538,
+      "epoch": 1.5029469548133596,
+      "grad_norm": 4.03655481338501,
+      "learning_rate": 1.6019724964375614e-05,
+      "loss": 0.1209,
+      "mean_token_accuracy": 0.9569043278694153,
+      "num_tokens": 2116348.0,
+      "step": 1530
+    },
+    {
+      "entropy": 1.0173700392246245,
+      "epoch": 1.512770137524558,
+      "grad_norm": 3.120889902114868,
+      "learning_rate": 1.5428421768230406e-05,
+      "loss": 0.1229,
+      "mean_token_accuracy": 0.9511404812335968,
+      "num_tokens": 2130231.0,
+      "step": 1540
+    },
+    {
+      "entropy": 1.03950754404068,
+      "epoch": 1.5225933202357562,
+      "grad_norm": 3.7956597805023193,
+      "learning_rate": 1.4846240706393145e-05,
+      "loss": 0.1362,
+      "mean_token_accuracy": 0.9497441112995147,
+      "num_tokens": 2144584.0,
+      "step": 1550
+    },
+    {
+      "entropy": 1.038707673549652,
+      "epoch": 1.5324165029469548,
+      "grad_norm": 3.067758560180664,
+      "learning_rate": 1.4273335394466842e-05,
+      "loss": 0.114,
+      "mean_token_accuracy": 0.957007896900177,
+      "num_tokens": 2158434.0,
+      "step": 1560
+    },
+    {
+      "entropy": 1.0332182109355927,
+      "epoch": 1.5422396856581533,
+      "grad_norm": 3.857419967651367,
+      "learning_rate": 1.3709857000534344e-05,
+      "loss": 0.1469,
+      "mean_token_accuracy": 0.9448116183280945,
+      "num_tokens": 2172669.0,
+      "step": 1570
+    },
+    {
+      "entropy": 1.029129409790039,
+      "epoch": 1.5520628683693518,
+      "grad_norm": 3.294672727584839,
+      "learning_rate": 1.3155954205270654e-05,
+      "loss": 0.1239,
+      "mean_token_accuracy": 0.9522143483161927,
+      "num_tokens": 2187186.0,
+      "step": 1580
+    },
+    {
+      "entropy": 1.0439756631851196,
+      "epoch": 1.5618860510805501,
+      "grad_norm": 3.736772298812866,
+      "learning_rate": 1.2611773162711871e-05,
+      "loss": 0.1322,
+      "mean_token_accuracy": 0.9506955087184906,
+      "num_tokens": 2200645.0,
+      "step": 1590
+    },
+    {
+      "entropy": 1.0213513851165772,
+      "epoch": 1.5717092337917484,
+      "grad_norm": 4.28324031829834,
+      "learning_rate": 1.207745746169055e-05,
+      "loss": 0.1249,
+      "mean_token_accuracy": 0.9545923113822937,
+      "num_tokens": 2214380.0,
+      "step": 1600
+    },
+    {
+      "entropy": 1.0353586077690125,
+      "epoch": 1.581532416502947,
+      "grad_norm": 3.537813901901245,
+      "learning_rate": 1.1553148087948074e-05,
+      "loss": 0.1085,
+      "mean_token_accuracy": 0.9639479219913483,
+      "num_tokens": 2228368.0,
+      "step": 1610
+    },
+    {
+      "entropy": 1.0298118233680724,
+      "epoch": 1.5913555992141455,
+      "grad_norm": 2.8899447917938232,
+      "learning_rate": 1.1038983386933821e-05,
+      "loss": 0.1116,
+      "mean_token_accuracy": 0.9609639823436738,
+      "num_tokens": 2242439.0,
+      "step": 1620
+    },
+    {
+      "entropy": 1.0277505457401275,
+      "epoch": 1.6011787819253438,
+      "grad_norm": 3.444633960723877,
+      "learning_rate": 1.0535099027301014e-05,
+      "loss": 0.1125,
+      "mean_token_accuracy": 0.9555817604064941,
+      "num_tokens": 2256501.0,
+      "step": 1630
+    },
+    {
+      "entropy": 1.0288507044315338,
+      "epoch": 1.611001964636542,
+      "grad_norm": 2.914987087249756,
+      "learning_rate": 1.0041627965108946e-05,
+      "loss": 0.1188,
+      "mean_token_accuracy": 0.9554714500904083,
+      "num_tokens": 2270649.0,
+      "step": 1640
+    },
+    {
+      "entropy": 1.0363532960414887,
+      "epoch": 1.6208251473477406,
+      "grad_norm": 3.304915189743042,
+      "learning_rate": 9.558700408740784e-06,
+      "loss": 0.1263,
+      "mean_token_accuracy": 0.9512441396713257,
+      "num_tokens": 2284158.0,
+      "step": 1650
+    },
+    {
+      "entropy": 1.0145125687122345,
+      "epoch": 1.6306483300589392,
+      "grad_norm": 4.782883167266846,
+      "learning_rate": 9.086443784546623e-06,
+      "loss": 0.1175,
+      "mean_token_accuracy": 0.9568356812000275,
+      "num_tokens": 2297921.0,
+      "step": 1660
+    },
+    {
+      "entropy": 1.0345391750335693,
+      "epoch": 1.6404715127701375,
+      "grad_norm": 3.5152642726898193,
+      "learning_rate": 8.62498270322038e-06,
+      "loss": 0.1205,
+      "mean_token_accuracy": 0.9557737529277801,
+      "num_tokens": 2311944.0,
+      "step": 1670
+    },
+    {
+      "entropy": 1.0225283265113831,
+      "epoch": 1.650294695481336,
+      "grad_norm": 2.7115612030029297,
+      "learning_rate": 8.174438926919814e-06,
+      "loss": 0.1223,
+      "mean_token_accuracy": 0.952198076248169,
+      "num_tokens": 2325708.0,
+      "step": 1680
+    },
+    {
+      "entropy": 1.0220875501632691,
+      "epoch": 1.6601178781925343,
+      "grad_norm": 4.39996337890625,
+      "learning_rate": 7.734931337137991e-06,
+      "loss": 0.1259,
+      "mean_token_accuracy": 0.949955427646637,
+      "num_tokens": 2339350.0,
+      "step": 1690
+    },
+    {
+      "entropy": 1.041896116733551,
+      "epoch": 1.6699410609037328,
+      "grad_norm": 2.490516185760498,
+      "learning_rate": 7.306575903335056e-06,
+      "loss": 0.1088,
+      "mean_token_accuracy": 0.9582253575325013,
+      "num_tokens": 2353189.0,
+      "step": 1700
+    },
+    {
+      "entropy": 1.0166313409805299,
+      "epoch": 1.6797642436149314,
+      "grad_norm": 3.405906915664673,
+      "learning_rate": 6.8894856523382015e-06,
+      "loss": 0.1092,
+      "mean_token_accuracy": 0.9604466795921326,
+      "num_tokens": 2366773.0,
+      "step": 1710
+    },
+    {
+      "entropy": 1.0194653511047362,
+      "epoch": 1.6895874263261297,
+      "grad_norm": 4.956583499908447,
+      "learning_rate": 6.483770638518194e-06,
+      "loss": 0.1317,
+      "mean_token_accuracy": 0.9528750419616699,
+      "num_tokens": 2379935.0,
+      "step": 1720
+    },
+    {
+      "entropy": 1.0145066142082215,
+      "epoch": 1.699410609037328,
+      "grad_norm": 3.445397138595581,
+      "learning_rate": 6.0895379147501154e-06,
+      "loss": 0.0981,
+      "mean_token_accuracy": 0.9608128428459167,
+      "num_tokens": 2393432.0,
+      "step": 1730
+    },
+    {
+      "entropy": 1.0181198954582213,
+      "epoch": 1.7092337917485265,
+      "grad_norm": 3.50852370262146,
+      "learning_rate": 5.7068915041662396e-06,
+      "loss": 0.1249,
+      "mean_token_accuracy": 0.9531911551952362,
+      "num_tokens": 2406850.0,
+      "step": 1740
+    },
+    {
+      "entropy": 1.0325023114681244,
+      "epoch": 1.719056974459725,
+      "grad_norm": 3.0633742809295654,
+      "learning_rate": 5.335932372708197e-06,
+      "loss": 0.0956,
+      "mean_token_accuracy": 0.9631496131420135,
+      "num_tokens": 2421793.0,
+      "step": 1750
+    },
+    {
+      "entropy": 1.0145358264446258,
+      "epoch": 1.7288801571709234,
+      "grad_norm": 2.472642421722412,
+      "learning_rate": 4.976758402485887e-06,
+      "loss": 0.1145,
+      "mean_token_accuracy": 0.9586724281311035,
+      "num_tokens": 2435557.0,
+      "step": 1760
+    },
+    {
+      "entropy": 1.0119429349899292,
+      "epoch": 1.7387033398821217,
+      "grad_norm": 4.490127086639404,
+      "learning_rate": 4.629464365950098e-06,
+      "loss": 0.1186,
+      "mean_token_accuracy": 0.958074551820755,
+      "num_tokens": 2448595.0,
+      "step": 1770
+    },
+    {
+      "entropy": 1.0218343913555146,
+      "epoch": 1.7485265225933202,
+      "grad_norm": 3.058469533920288,
+      "learning_rate": 4.29414190088559e-06,
+      "loss": 0.1112,
+      "mean_token_accuracy": 0.9574200809001923,
+      "num_tokens": 2463318.0,
+      "step": 1780
+    },
+    {
+      "entropy": 1.0130614042282104,
+      "epoch": 1.7583497053045187,
+      "grad_norm": 2.9102025032043457,
+      "learning_rate": 3.970879486231416e-06,
+      "loss": 0.1096,
+      "mean_token_accuracy": 0.959147834777832,
+      "num_tokens": 2477496.0,
+      "step": 1790
+    },
+    {
+      "entropy": 1.0219753503799438,
+      "epoch": 1.768172888015717,
+      "grad_norm": 3.5791053771972656,
+      "learning_rate": 3.6597624187346193e-06,
+      "loss": 0.1079,
+      "mean_token_accuracy": 0.95797678232193,
+      "num_tokens": 2491486.0,
+      "step": 1800
+    },
+    {
+      "entropy": 1.0247013926506043,
+      "epoch": 1.7779960707269156,
+      "grad_norm": 2.998903751373291,
+      "learning_rate": 3.3608727904436365e-06,
+      "loss": 0.1404,
+      "mean_token_accuracy": 0.9512396991252899,
+      "num_tokens": 2505578.0,
+      "step": 1810
+    },
+    {
+      "entropy": 1.0234854578971864,
+      "epoch": 1.7878192534381139,
+      "grad_norm": 2.6577558517456055,
+      "learning_rate": 3.0742894670473364e-06,
+      "loss": 0.1059,
+      "mean_token_accuracy": 0.9624595522880555,
+      "num_tokens": 2520062.0,
+      "step": 1820
+    },
+    {
+      "entropy": 1.0222528517246245,
+      "epoch": 1.7976424361493124,
+      "grad_norm": 4.599435806274414,
+      "learning_rate": 2.8000880670653483e-06,
+      "loss": 0.1221,
+      "mean_token_accuracy": 0.9534383177757263,
+      "num_tokens": 2534119.0,
+      "step": 1830
+    },
+    {
+      "entropy": 1.027721107006073,
+      "epoch": 1.807465618860511,
+      "grad_norm": 3.4878251552581787,
+      "learning_rate": 2.5383409418951233e-06,
+      "loss": 0.113,
+      "mean_token_accuracy": 0.9600664436817169,
+      "num_tokens": 2548489.0,
+      "step": 1840
+    },
+    {
+      "entropy": 1.019310623407364,
+      "epoch": 1.8172888015717092,
+      "grad_norm": 3.4331095218658447,
+      "learning_rate": 2.2891171567212212e-06,
+      "loss": 0.1341,
+      "mean_token_accuracy": 0.9543480813503266,
+      "num_tokens": 2561735.0,
+      "step": 1850
+    },
+    {
+      "entropy": 1.0107543528079987,
+      "epoch": 1.8271119842829076,
+      "grad_norm": 3.9222187995910645,
+      "learning_rate": 2.05248247229155e-06,
+      "loss": 0.1179,
+      "mean_token_accuracy": 0.9581474542617798,
+      "num_tokens": 2575061.0,
+      "step": 1860
+    },
+    {
+      "entropy": 1.0125151932239533,
+      "epoch": 1.836935166994106,
+      "grad_norm": 5.1685638427734375,
+      "learning_rate": 1.8284993275656138e-06,
+      "loss": 0.1367,
+      "mean_token_accuracy": 0.9521771371364594,
+      "num_tokens": 2588407.0,
+      "step": 1870
+    },
+    {
+      "entropy": 1.0245961546897888,
+      "epoch": 1.8467583497053046,
+      "grad_norm": 3.8602330684661865,
+      "learning_rate": 1.617226823239243e-06,
+      "loss": 0.1063,
+      "mean_token_accuracy": 0.9581910848617554,
+      "num_tokens": 2602316.0,
+      "step": 1880
+    },
+    {
+      "entropy": 1.0231972277164458,
+      "epoch": 1.856581532416503,
+      "grad_norm": 2.2509899139404297,
+      "learning_rate": 1.418720706150095e-06,
+      "loss": 0.1066,
+      "mean_token_accuracy": 0.9595114648342132,
+      "num_tokens": 2616816.0,
+      "step": 1890
+    },
+    {
+      "entropy": 1.0134981274604797,
+      "epoch": 1.8664047151277012,
+      "grad_norm": 3.6242942810058594,
+      "learning_rate": 1.2330333545682415e-06,
+      "loss": 0.1271,
+      "mean_token_accuracy": 0.9498240411281585,
+      "num_tokens": 2630854.0,
+      "step": 1900
+    },
+    {
+      "entropy": 1.0087493777275085,
+      "epoch": 1.8762278978388998,
+      "grad_norm": 2.5779571533203125,
+      "learning_rate": 1.0602137643754727e-06,
+      "loss": 0.1212,
+      "mean_token_accuracy": 0.9545005440711976,
+      "num_tokens": 2644882.0,
+      "step": 1910
+    },
+    {
+      "entropy": 1.0015055537223816,
+      "epoch": 1.8860510805500983,
+      "grad_norm": 3.712475061416626,
+      "learning_rate": 9.003075361371571e-07,
+      "loss": 0.1261,
+      "mean_token_accuracy": 0.9499159574508667,
+      "num_tokens": 2657903.0,
+      "step": 1920
+    },
+    {
+      "entropy": 1.0081746339797975,
+      "epoch": 1.8958742632612968,
+      "grad_norm": 3.3812055587768555,
+      "learning_rate": 7.533568630699484e-07,
+      "loss": 0.1249,
+      "mean_token_accuracy": 0.949466985464096,
+      "num_tokens": 2671432.0,
+      "step": 1930
+    },
+    {
+      "entropy": 1.005694168806076,
+      "epoch": 1.9056974459724951,
+      "grad_norm": 3.1163291931152344,
+      "learning_rate": 6.194005199086084e-07,
+      "loss": 0.1074,
+      "mean_token_accuracy": 0.959492427110672,
+      "num_tokens": 2684791.0,
+      "step": 1940
+    },
+    {
+      "entropy": 1.0161106705665588,
+      "epoch": 1.9155206286836934,
+      "grad_norm": 3.3539648056030273,
+      "learning_rate": 4.984738526748301e-07,
+      "loss": 0.111,
+      "mean_token_accuracy": 0.9622210323810577,
+      "num_tokens": 2699054.0,
+      "step": 1950
+    },
+    {
+      "entropy": 1.0071995317935944,
+      "epoch": 1.925343811394892,
+      "grad_norm": 3.1897265911102295,
+      "learning_rate": 3.906087693507199e-07,
+      "loss": 0.1013,
+      "mean_token_accuracy": 0.9622646331787109,
+      "num_tokens": 2712002.0,
+      "step": 1960
+    },
+    {
+      "entropy": 1.0230316162109374,
+      "epoch": 1.9351669941060905,
+      "grad_norm": 2.8453235626220703,
+      "learning_rate": 2.9583373145951056e-07,
+      "loss": 0.1258,
+      "mean_token_accuracy": 0.955587899684906,
+      "num_tokens": 2726370.0,
+      "step": 1970
+    },
+    {
+      "entropy": 1.0153518915176392,
+      "epoch": 1.9449901768172888,
+      "grad_norm": 4.360236644744873,
+      "learning_rate": 2.1417374655565703e-07,
+      "loss": 0.1268,
+      "mean_token_accuracy": 0.950088620185852,
+      "num_tokens": 2740774.0,
+      "step": 1980
+    },
+    {
+      "entropy": 1.0258444249629974,
+      "epoch": 1.9548133595284871,
+      "grad_norm": 3.178138494491577,
+      "learning_rate": 1.4565036162623724e-07,
+      "loss": 0.1179,
+      "mean_token_accuracy": 0.955890280008316,
+      "num_tokens": 2754709.0,
+      "step": 1990
+    },
+    {
+      "entropy": 1.0152276754379272,
+      "epoch": 1.9646365422396856,
+      "grad_norm": 2.8657338619232178,
+      "learning_rate": 9.028165740556648e-08,
+      "loss": 0.1299,
+      "mean_token_accuracy": 0.9523500502109528,
+      "num_tokens": 2768469.0,
+      "step": 2000
+    },
+    {
+      "entropy": 1.0214035391807557,
+      "epoch": 1.9744597249508842,
+      "grad_norm": 2.892160654067993,
+      "learning_rate": 4.8082243604336e-08,
+      "loss": 0.1051,
+      "mean_token_accuracy": 0.9589199960231781,
+      "num_tokens": 2781907.0,
+      "step": 2010
+    },
+    {
+      "entropy": 1.004443496465683,
+      "epoch": 1.9842829076620825,
+      "grad_norm": 3.1709177494049072,
+      "learning_rate": 1.9063255054685425e-08,
+      "loss": 0.1083,
+      "mean_token_accuracy": 0.960349851846695,
+      "num_tokens": 2795784.0,
+      "step": 2020
+    },
+    {
+      "entropy": 1.029257845878601,
+      "epoch": 1.9941060903732808,
+      "grad_norm": 3.3554000854492188,
+      "learning_rate": 3.2323487721419577e-09,
+      "loss": 0.1122,
+      "mean_token_accuracy": 0.9587242484092713,
+      "num_tokens": 2809553.0,
+      "step": 2030
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2036,
+  "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_s2_10k/checkpoint-2036/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e75a9aefb996ad2b21a80d647b9fc05cf69d9f4afc9b9014cda7fcb78f6cd8a
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_10k/manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_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": 2,
+    "val_split": 0
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_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_s2_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_s2_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_s2_10k/train_manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_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": 2,
+    "val_split": 0
+  }
+}

checkpoints/cat_qwen25_7b_r8_a32_adam_rot_e2_lr1e-4_s2_10k/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e75a9aefb996ad2b21a80d647b9fc05cf69d9f4afc9b9014cda7fcb78f6cd8a
+size 6417