add checkpoint cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s2_10k

.gitattributes

@@ -564,3 +564,6 @@ checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s1_10k/tokenizer.json filter=
 checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-1019/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/checkpoint-2038/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s3_10k/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s2_10k/checkpoint-1018/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s2_10k/checkpoint-2036/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s2_10k/tokenizer.json filter=lfs diff=lfs merge=lfs -text

checkpoints/cat_qwen25_7b_r8_a32_sgd_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_sgd_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_sgd_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/1lhnv9hd) 
+
+
+This model was trained with SFT.
+
+### Framework versions
+
+- PEFT 0.19.1
+- TRL: 0.28.0
+- Transformers: 4.57.6
+- Pytorch: 2.9.1
+- Datasets: 4.5.0
+- Tokenizers: 0.22.2
+
+## Citations
+
+
+
+Cite TRL as:
+    
+```bibtex
+@software{vonwerra2020trl,
+  title   = {{TRL: Transformers Reinforcement Learning}},
+  author  = {von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin},
+  license = {Apache-2.0},
+  url     = {https://github.com/huggingface/trl},
+  year    = {2020}
+}
+```

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_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": [
+    "q_proj",
+    "gate_proj",
+    "up_proj",
+    "down_proj",
+    "o_proj",
+    "k_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_sgd_rot_e2_lr1e-4_s2_10k/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4cbdc9eacf977c225270ba114cb148643ea8781fd25b11b3102dbed314be43ef
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_sgd_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_sgd_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_sgd_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_sgd_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": [
+    "q_proj",
+    "gate_proj",
+    "up_proj",
+    "down_proj",
+    "o_proj",
+    "k_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_sgd_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:d254f50535a8f2c558d52a3310823b0a774e9d7e6d7e200e456e6174fbfafacd
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_sgd_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_sgd_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_sgd_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_sgd_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_sgd_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_sgd_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": 1.22729994058609,
+      "epoch": 0.009823182711198428,
+      "grad_norm": 4.8866095542907715,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.6928,
+      "mean_token_accuracy": 0.7725797176361084,
+      "num_tokens": 13828.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.2600382804870605,
+      "epoch": 0.019646365422396856,
+      "grad_norm": 3.3233840465545654,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.8134,
+      "mean_token_accuracy": 0.7262678980827332,
+      "num_tokens": 28074.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.2484176516532899,
+      "epoch": 0.029469548133595286,
+      "grad_norm": 2.9845433235168457,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.7786,
+      "mean_token_accuracy": 0.7460076689720154,
+      "num_tokens": 42168.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.2452077984809875,
+      "epoch": 0.03929273084479371,
+      "grad_norm": 3.351901054382324,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.8251,
+      "mean_token_accuracy": 0.7294955134391785,
+      "num_tokens": 56700.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.2434914350509643,
+      "epoch": 0.04911591355599214,
+      "grad_norm": 3.0272388458251953,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.7171,
+      "mean_token_accuracy": 0.7613231837749481,
+      "num_tokens": 70922.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.253853189945221,
+      "epoch": 0.05893909626719057,
+      "grad_norm": 2.4795994758605957,
+      "learning_rate": 5.784313725490197e-05,
+      "loss": 0.7099,
+      "mean_token_accuracy": 0.7647240102291107,
+      "num_tokens": 84911.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.2360968828201293,
+      "epoch": 0.068762278978389,
+      "grad_norm": 2.6639111042022705,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 0.7444,
+      "mean_token_accuracy": 0.755102014541626,
+      "num_tokens": 98835.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.2527110695838928,
+      "epoch": 0.07858546168958742,
+      "grad_norm": 3.6153311729431152,
+      "learning_rate": 7.745098039215687e-05,
+      "loss": 0.7012,
+      "mean_token_accuracy": 0.7707957923412323,
+      "num_tokens": 112502.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.2246053576469422,
+      "epoch": 0.08840864440078586,
+      "grad_norm": 3.169745683670044,
+      "learning_rate": 8.725490196078432e-05,
+      "loss": 0.7143,
+      "mean_token_accuracy": 0.76532541513443,
+      "num_tokens": 126248.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.232993471622467,
+      "epoch": 0.09823182711198428,
+      "grad_norm": 3.4184534549713135,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 0.7402,
+      "mean_token_accuracy": 0.7552495896816254,
+      "num_tokens": 139551.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.2279873371124268,
+      "epoch": 0.10805500982318271,
+      "grad_norm": 2.724405527114868,
+      "learning_rate": 9.999676765122786e-05,
+      "loss": 0.7571,
+      "mean_token_accuracy": 0.7568806827068328,
+      "num_tokens": 152975.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.2499749660491943,
+      "epoch": 0.11787819253438114,
+      "grad_norm": 2.9191336631774902,
+      "learning_rate": 9.998093674494531e-05,
+      "loss": 0.7314,
+      "mean_token_accuracy": 0.760424691438675,
+      "num_tokens": 167073.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.2526087641716004,
+      "epoch": 0.12770137524557956,
+      "grad_norm": 3.374049186706543,
+      "learning_rate": 9.995191775639568e-05,
+      "loss": 0.7667,
+      "mean_token_accuracy": 0.7505820035934448,
+      "num_tokens": 181404.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.237971842288971,
+      "epoch": 0.137524557956778,
+      "grad_norm": 2.7548294067382812,
+      "learning_rate": 9.990971834259444e-05,
+      "loss": 0.7204,
+      "mean_token_accuracy": 0.7575949549674987,
+      "num_tokens": 195441.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.2629727482795716,
+      "epoch": 0.14734774066797643,
+      "grad_norm": 3.8496453762054443,
+      "learning_rate": 9.985434963837376e-05,
+      "loss": 0.7956,
+      "mean_token_accuracy": 0.7347698986530304,
+      "num_tokens": 209863.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.2493702173233032,
+      "epoch": 0.15717092337917485,
+      "grad_norm": 3.358459234237671,
+      "learning_rate": 9.978582625344435e-05,
+      "loss": 0.7489,
+      "mean_token_accuracy": 0.751927274465561,
+      "num_tokens": 223646.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.2397171974182128,
+      "epoch": 0.16699410609037327,
+      "grad_norm": 3.4665334224700928,
+      "learning_rate": 9.970416626854049e-05,
+      "loss": 0.721,
+      "mean_token_accuracy": 0.7698734521865844,
+      "num_tokens": 237587.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.2597252249717712,
+      "epoch": 0.17681728880157171,
+      "grad_norm": 2.362679958343506,
+      "learning_rate": 9.960939123064929e-05,
+      "loss": 0.7742,
+      "mean_token_accuracy": 0.740584260225296,
+      "num_tokens": 251760.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.2215595960617065,
+      "epoch": 0.18664047151277013,
+      "grad_norm": 3.4586236476898193,
+      "learning_rate": 9.950152614732518e-05,
+      "loss": 0.7087,
+      "mean_token_accuracy": 0.7657637417316436,
+      "num_tokens": 265417.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.2417829275131225,
+      "epoch": 0.19646365422396855,
+      "grad_norm": 3.494053602218628,
+      "learning_rate": 9.93805994800914e-05,
+      "loss": 0.7374,
+      "mean_token_accuracy": 0.7579931676387787,
+      "num_tokens": 278540.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.2505549669265748,
+      "epoch": 0.206286836935167,
+      "grad_norm": 3.695319175720215,
+      "learning_rate": 9.924664313693006e-05,
+      "loss": 0.7837,
+      "mean_token_accuracy": 0.7425522744655609,
+      "num_tokens": 292295.0,
+      "step": 210
+    },
+    {
+      "entropy": 1.2271206736564637,
+      "epoch": 0.21611001964636542,
+      "grad_norm": 3.339724540710449,
+      "learning_rate": 9.909969246386284e-05,
+      "loss": 0.7669,
+      "mean_token_accuracy": 0.7515395283699036,
+      "num_tokens": 305466.0,
+      "step": 220
+    },
+    {
+      "entropy": 1.2639145135879517,
+      "epoch": 0.22593320235756384,
+      "grad_norm": 3.1140177249908447,
+      "learning_rate": 9.893978623562453e-05,
+      "loss": 0.7502,
+      "mean_token_accuracy": 0.7555740237236023,
+      "num_tokens": 320495.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.2431163549423219,
+      "epoch": 0.2357563850687623,
+      "grad_norm": 3.665536403656006,
+      "learning_rate": 9.876696664543177e-05,
+      "loss": 0.7701,
+      "mean_token_accuracy": 0.7411514103412629,
+      "num_tokens": 333901.0,
+      "step": 240
+    },
+    {
+      "entropy": 1.2520034074783326,
+      "epoch": 0.2455795677799607,
+      "grad_norm": 4.050171375274658,
+      "learning_rate": 9.858127929384991e-05,
+      "loss": 0.7119,
+      "mean_token_accuracy": 0.7642730891704559,
+      "num_tokens": 347392.0,
+      "step": 250
+    },
+    {
+      "entropy": 1.243278980255127,
+      "epoch": 0.2554027504911591,
+      "grad_norm": 3.8479113578796387,
+      "learning_rate": 9.838277317676077e-05,
+      "loss": 0.761,
+      "mean_token_accuracy": 0.7522589504718781,
+      "num_tokens": 361140.0,
+      "step": 260
+    },
+    {
+      "entropy": 1.228507363796234,
+      "epoch": 0.26522593320235754,
+      "grad_norm": 2.529057264328003,
+      "learning_rate": 9.817150067243438e-05,
+      "loss": 0.689,
+      "mean_token_accuracy": 0.7683715045452117,
+      "num_tokens": 374661.0,
+      "step": 270
+    },
+    {
+      "entropy": 1.248687493801117,
+      "epoch": 0.275049115913556,
+      "grad_norm": 2.473571538925171,
+      "learning_rate": 9.794751752770845e-05,
+      "loss": 0.7849,
+      "mean_token_accuracy": 0.7473262250423431,
+      "num_tokens": 388415.0,
+      "step": 280
+    },
+    {
+      "entropy": 1.2462671041488647,
+      "epoch": 0.28487229862475444,
+      "grad_norm": 3.4700798988342285,
+      "learning_rate": 9.771088284327879e-05,
+      "loss": 0.7537,
+      "mean_token_accuracy": 0.7521771848201751,
+      "num_tokens": 402031.0,
+      "step": 290
+    },
+    {
+      "entropy": 1.2523157238960265,
+      "epoch": 0.29469548133595286,
+      "grad_norm": 3.2036030292510986,
+      "learning_rate": 9.746165905810488e-05,
+      "loss": 0.7551,
+      "mean_token_accuracy": 0.7571269392967224,
+      "num_tokens": 416165.0,
+      "step": 300
+    },
+    {
+      "entropy": 1.2602078080177308,
+      "epoch": 0.3045186640471513,
+      "grad_norm": 2.589961290359497,
+      "learning_rate": 9.719991193293466e-05,
+      "loss": 0.7821,
+      "mean_token_accuracy": 0.7466304242610932,
+      "num_tokens": 429910.0,
+      "step": 310
+    },
+    {
+      "entropy": 1.2587228298187256,
+      "epoch": 0.3143418467583497,
+      "grad_norm": 4.473543643951416,
+      "learning_rate": 9.692571053295268e-05,
+      "loss": 0.758,
+      "mean_token_accuracy": 0.7488924205303192,
+      "num_tokens": 443725.0,
+      "step": 320
+    },
+    {
+      "entropy": 1.2587100744247437,
+      "epoch": 0.3241650294695481,
+      "grad_norm": 2.999272584915161,
+      "learning_rate": 9.663912720955637e-05,
+      "loss": 0.7934,
+      "mean_token_accuracy": 0.7434530913829803,
+      "num_tokens": 458234.0,
+      "step": 330
+    },
+    {
+      "entropy": 1.235784935951233,
+      "epoch": 0.33398821218074654,
+      "grad_norm": 3.993062734603882,
+      "learning_rate": 9.634023758126539e-05,
+      "loss": 0.7378,
+      "mean_token_accuracy": 0.7529276013374329,
+      "num_tokens": 471402.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.212973964214325,
+      "epoch": 0.343811394891945,
+      "grad_norm": 3.497361183166504,
+      "learning_rate": 9.602912051376859e-05,
+      "loss": 0.6893,
+      "mean_token_accuracy": 0.773086953163147,
+      "num_tokens": 484651.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.2527001976966858,
+      "epoch": 0.35363457760314343,
+      "grad_norm": 2.310852289199829,
+      "learning_rate": 9.570585809911441e-05,
+      "loss": 0.7795,
+      "mean_token_accuracy": 0.7472991645336151,
+      "num_tokens": 497956.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.2270971417427063,
+      "epoch": 0.36345776031434185,
+      "grad_norm": 3.0224449634552,
+      "learning_rate": 9.537053563404992e-05,
+      "loss": 0.7333,
+      "mean_token_accuracy": 0.7572860002517701,
+      "num_tokens": 511717.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.2353201866149903,
+      "epoch": 0.37328094302554027,
+      "grad_norm": 3.016392946243286,
+      "learning_rate": 9.502324159751412e-05,
+      "loss": 0.7056,
+      "mean_token_accuracy": 0.7640936195850372,
+      "num_tokens": 525633.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.2572155117988586,
+      "epoch": 0.3831041257367387,
+      "grad_norm": 3.6709957122802734,
+      "learning_rate": 9.466406762729181e-05,
+      "loss": 0.7471,
+      "mean_token_accuracy": 0.7522993624210358,
+      "num_tokens": 538945.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.2437490463256835,
+      "epoch": 0.3929273084479371,
+      "grad_norm": 2.7692830562591553,
+      "learning_rate": 9.429310849583378e-05,
+      "loss": 0.746,
+      "mean_token_accuracy": 0.754131942987442,
+      "num_tokens": 552861.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.2515570044517517,
+      "epoch": 0.4027504911591356,
+      "grad_norm": 2.845248222351074,
+      "learning_rate": 9.391046208524989e-05,
+      "loss": 0.7987,
+      "mean_token_accuracy": 0.736023235321045,
+      "num_tokens": 565955.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.2464410781860351,
+      "epoch": 0.412573673870334,
+      "grad_norm": 3.273564577102661,
+      "learning_rate": 9.351622936148182e-05,
+      "loss": 0.7226,
+      "mean_token_accuracy": 0.7609843373298645,
+      "num_tokens": 579923.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.2555230379104614,
+      "epoch": 0.4223968565815324,
+      "grad_norm": 3.4945361614227295,
+      "learning_rate": 9.31105143476618e-05,
+      "loss": 0.7408,
+      "mean_token_accuracy": 0.7554882347583771,
+      "num_tokens": 594071.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.2536908864974976,
+      "epoch": 0.43222003929273084,
+      "grad_norm": 2.4363486766815186,
+      "learning_rate": 9.269342409666495e-05,
+      "loss": 0.7346,
+      "mean_token_accuracy": 0.7615235865116119,
+      "num_tokens": 608545.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.228044366836548,
+      "epoch": 0.44204322200392926,
+      "grad_norm": 3.0561442375183105,
+      "learning_rate": 9.226506866286202e-05,
+      "loss": 0.7296,
+      "mean_token_accuracy": 0.7626752912998199,
+      "num_tokens": 621975.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.257164752483368,
+      "epoch": 0.4518664047151277,
+      "grad_norm": 2.552243232727051,
+      "learning_rate": 9.18255610730802e-05,
+      "loss": 0.7432,
+      "mean_token_accuracy": 0.7537168443202973,
+      "num_tokens": 636061.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.2517608165740968,
+      "epoch": 0.46168958742632615,
+      "grad_norm": 3.190830945968628,
+      "learning_rate": 9.137501729677962e-05,
+      "loss": 0.75,
+      "mean_token_accuracy": 0.7544792711734771,
+      "num_tokens": 649717.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.235853898525238,
+      "epoch": 0.4715127701375246,
+      "grad_norm": 2.730757474899292,
+      "learning_rate": 9.091355621545339e-05,
+      "loss": 0.6943,
+      "mean_token_accuracy": 0.7740860462188721,
+      "num_tokens": 664123.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.2552549839019775,
+      "epoch": 0.481335952848723,
+      "grad_norm": 3.690030813217163,
+      "learning_rate": 9.044129959125923e-05,
+      "loss": 0.7954,
+      "mean_token_accuracy": 0.7400918424129486,
+      "num_tokens": 677803.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.2475018978118897,
+      "epoch": 0.4911591355599214,
+      "grad_norm": 3.209927558898926,
+      "learning_rate": 8.995837203489108e-05,
+      "loss": 0.7097,
+      "mean_token_accuracy": 0.7667286634445191,
+      "num_tokens": 691461.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.2404452085494995,
+      "epoch": 0.5009823182711198,
+      "grad_norm": 2.9024016857147217,
+      "learning_rate": 8.946490097269899e-05,
+      "loss": 0.686,
+      "mean_token_accuracy": 0.7731779038906097,
+      "num_tokens": 705470.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.2764546751976014,
+      "epoch": 0.5108055009823183,
+      "grad_norm": 3.618655204772949,
+      "learning_rate": 8.896101661306618e-05,
+      "loss": 0.8615,
+      "mean_token_accuracy": 0.715438824892044,
+      "num_tokens": 718805.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.252146315574646,
+      "epoch": 0.5206286836935167,
+      "grad_norm": 2.5902345180511475,
+      "learning_rate": 8.844685191205193e-05,
+      "loss": 0.7396,
+      "mean_token_accuracy": 0.7525096774101258,
+      "num_tokens": 732367.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.2653416037559508,
+      "epoch": 0.5304518664047151,
+      "grad_norm": 3.163494825363159,
+      "learning_rate": 8.792254253830945e-05,
+      "loss": 0.7787,
+      "mean_token_accuracy": 0.7419969081878662,
+      "num_tokens": 746431.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.2420126438140868,
+      "epoch": 0.5402750491159135,
+      "grad_norm": 2.3512191772460938,
+      "learning_rate": 8.738822683728814e-05,
+      "loss": 0.7096,
+      "mean_token_accuracy": 0.7669752776622772,
+      "num_tokens": 760182.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.2616332292556762,
+      "epoch": 0.550098231827112,
+      "grad_norm": 2.4627273082733154,
+      "learning_rate": 8.684404579472936e-05,
+      "loss": 0.776,
+      "mean_token_accuracy": 0.7481194794178009,
+      "num_tokens": 774137.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.220177948474884,
+      "epoch": 0.5599214145383105,
+      "grad_norm": 3.291656494140625,
+      "learning_rate": 8.629014299946568e-05,
+      "loss": 0.6453,
+      "mean_token_accuracy": 0.7858670830726624,
+      "num_tokens": 787806.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.2351359486579896,
+      "epoch": 0.5697445972495089,
+      "grad_norm": 2.42910099029541,
+      "learning_rate": 8.572666460553316e-05,
+      "loss": 0.6767,
+      "mean_token_accuracy": 0.7722299098968506,
+      "num_tokens": 801431.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.2556207776069641,
+      "epoch": 0.5795677799607073,
+      "grad_norm": 3.0838305950164795,
+      "learning_rate": 8.515375929360687e-05,
+      "loss": 0.7984,
+      "mean_token_accuracy": 0.741753101348877,
+      "num_tokens": 815080.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.2549353241920471,
+      "epoch": 0.5893909626719057,
+      "grad_norm": 2.646620512008667,
+      "learning_rate": 8.45715782317696e-05,
+      "loss": 0.7612,
+      "mean_token_accuracy": 0.7506958603858948,
+      "num_tokens": 829061.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.238725221157074,
+      "epoch": 0.5992141453831041,
+      "grad_norm": 3.156494379043579,
+      "learning_rate": 8.39802750356244e-05,
+      "loss": 0.7156,
+      "mean_token_accuracy": 0.7645548522472382,
+      "num_tokens": 842325.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.2504855751991273,
+      "epoch": 0.6090373280943026,
+      "grad_norm": 3.688000440597534,
+      "learning_rate": 8.338000572776114e-05,
+      "loss": 0.7688,
+      "mean_token_accuracy": 0.7456511080265045,
+      "num_tokens": 856377.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.2575064301490784,
+      "epoch": 0.618860510805501,
+      "grad_norm": 3.6037731170654297,
+      "learning_rate": 8.277092869658808e-05,
+      "loss": 0.7347,
+      "mean_token_accuracy": 0.7624819874763489,
+      "num_tokens": 870502.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.2600618243217467,
+      "epoch": 0.6286836935166994,
+      "grad_norm": 2.7635698318481445,
+      "learning_rate": 8.215320465453914e-05,
+      "loss": 0.7719,
+      "mean_token_accuracy": 0.7426825046539307,
+      "num_tokens": 884708.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.2481453895568848,
+      "epoch": 0.6385068762278978,
+      "grad_norm": 2.3514113426208496,
+      "learning_rate": 8.152699659566797e-05,
+      "loss": 0.7902,
+      "mean_token_accuracy": 0.736629843711853,
+      "num_tokens": 898142.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.2575042247772217,
+      "epoch": 0.6483300589390962,
+      "grad_norm": 3.3685693740844727,
+      "learning_rate": 8.089246975263996e-05,
+      "loss": 0.7412,
+      "mean_token_accuracy": 0.7545578002929687,
+      "num_tokens": 912820.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.226671326160431,
+      "epoch": 0.6581532416502947,
+      "grad_norm": 3.102717161178589,
+      "learning_rate": 8.024979155313354e-05,
+      "loss": 0.6868,
+      "mean_token_accuracy": 0.779555332660675,
+      "num_tokens": 926399.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.24713693857193,
+      "epoch": 0.6679764243614931,
+      "grad_norm": 2.6191911697387695,
+      "learning_rate": 7.959913157566246e-05,
+      "loss": 0.7767,
+      "mean_token_accuracy": 0.7440745651721954,
+      "num_tokens": 940133.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.2643288373947144,
+      "epoch": 0.6777996070726916,
+      "grad_norm": 2.782747983932495,
+      "learning_rate": 7.894066150483023e-05,
+      "loss": 0.7409,
+      "mean_token_accuracy": 0.7547606706619263,
+      "num_tokens": 954504.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.2425753355026246,
+      "epoch": 0.68762278978389,
+      "grad_norm": 3.5112552642822266,
+      "learning_rate": 7.827455508602923e-05,
+      "loss": 0.7657,
+      "mean_token_accuracy": 0.7505813300609588,
+      "num_tokens": 968934.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.2537635564804077,
+      "epoch": 0.6974459724950884,
+      "grad_norm": 3.150362491607666,
+      "learning_rate": 7.760098807959573e-05,
+      "loss": 0.7409,
+      "mean_token_accuracy": 0.7455904126167298,
+      "num_tokens": 983038.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.2538934469223022,
+      "epoch": 0.7072691552062869,
+      "grad_norm": 2.8867599964141846,
+      "learning_rate": 7.69201382144334e-05,
+      "loss": 0.7457,
+      "mean_token_accuracy": 0.746702516078949,
+      "num_tokens": 996748.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.264196515083313,
+      "epoch": 0.7170923379174853,
+      "grad_norm": 2.841606855392456,
+      "learning_rate": 7.623218514111743e-05,
+      "loss": 0.762,
+      "mean_token_accuracy": 0.7420303642749786,
+      "num_tokens": 1010595.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.238691246509552,
+      "epoch": 0.7269155206286837,
+      "grad_norm": 3.15217661857605,
+      "learning_rate": 7.553731038449144e-05,
+      "loss": 0.676,
+      "mean_token_accuracy": 0.7726105868816375,
+      "num_tokens": 1024890.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.2556055784225464,
+      "epoch": 0.7367387033398821,
+      "grad_norm": 2.4604084491729736,
+      "learning_rate": 7.483569729576996e-05,
+      "loss": 0.7578,
+      "mean_token_accuracy": 0.7490179061889648,
+      "num_tokens": 1038967.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.2183940052986144,
+      "epoch": 0.7465618860510805,
+      "grad_norm": 2.6620943546295166,
+      "learning_rate": 7.412753100415897e-05,
+      "loss": 0.651,
+      "mean_token_accuracy": 0.7827771067619324,
+      "num_tokens": 1052622.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.2360458731651307,
+      "epoch": 0.756385068762279,
+      "grad_norm": 3.445331335067749,
+      "learning_rate": 7.341299836800729e-05,
+      "loss": 0.7496,
+      "mean_token_accuracy": 0.7527473092079162,
+      "num_tokens": 1066588.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.2387821793556213,
+      "epoch": 0.7662082514734774,
+      "grad_norm": 3.357232093811035,
+      "learning_rate": 7.26922879255017e-05,
+      "loss": 0.6619,
+      "mean_token_accuracy": 0.7745956540107727,
+      "num_tokens": 1080453.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.2365756273269652,
+      "epoch": 0.7760314341846758,
+      "grad_norm": 3.045243263244629,
+      "learning_rate": 7.196558984491889e-05,
+      "loss": 0.7166,
+      "mean_token_accuracy": 0.7608296036720276,
+      "num_tokens": 1094550.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.2373855471611024,
+      "epoch": 0.7858546168958742,
+      "grad_norm": 4.2761359214782715,
+      "learning_rate": 7.123309587444711e-05,
+      "loss": 0.6843,
+      "mean_token_accuracy": 0.7768501520156861,
+      "num_tokens": 1108808.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.2416929841041564,
+      "epoch": 0.7956777996070727,
+      "grad_norm": 3.202831983566284,
+      "learning_rate": 7.04949992915912e-05,
+      "loss": 0.7058,
+      "mean_token_accuracy": 0.761394602060318,
+      "num_tokens": 1122365.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.2804744243621826,
+      "epoch": 0.8055009823182712,
+      "grad_norm": 3.275981903076172,
+      "learning_rate": 6.975149485217388e-05,
+      "loss": 0.7698,
+      "mean_token_accuracy": 0.7347693979740143,
+      "num_tokens": 1136376.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.2437939643859863,
+      "epoch": 0.8153241650294696,
+      "grad_norm": 2.6627252101898193,
+      "learning_rate": 6.900277873894705e-05,
+      "loss": 0.6874,
+      "mean_token_accuracy": 0.7675921380519867,
+      "num_tokens": 1150011.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.2557293772697449,
+      "epoch": 0.825147347740668,
+      "grad_norm": 2.6019535064697266,
+      "learning_rate": 6.824904850982656e-05,
+      "loss": 0.6945,
+      "mean_token_accuracy": 0.7684582710266114,
+      "num_tokens": 1164190.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.2451830863952638,
+      "epoch": 0.8349705304518664,
+      "grad_norm": 3.0742592811584473,
+      "learning_rate": 6.749050304576415e-05,
+      "loss": 0.6961,
+      "mean_token_accuracy": 0.7662852883338929,
+      "num_tokens": 1177346.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.2350680232048035,
+      "epoch": 0.8447937131630648,
+      "grad_norm": 3.061077117919922,
+      "learning_rate": 6.672734249827028e-05,
+      "loss": 0.6833,
+      "mean_token_accuracy": 0.780314964056015,
+      "num_tokens": 1190692.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.250148320198059,
+      "epoch": 0.8546168958742633,
+      "grad_norm": 2.8277947902679443,
+      "learning_rate": 6.59597682366017e-05,
+      "loss": 0.7275,
+      "mean_token_accuracy": 0.7577569246292114,
+      "num_tokens": 1204254.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.2564526915550231,
+      "epoch": 0.8644400785854617,
+      "grad_norm": 3.7227401733398438,
+      "learning_rate": 6.518798279462766e-05,
+      "loss": 0.7498,
+      "mean_token_accuracy": 0.7435093998908997,
+      "num_tokens": 1217938.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.2478920340538024,
+      "epoch": 0.8742632612966601,
+      "grad_norm": 2.8872194290161133,
+      "learning_rate": 6.441218981738895e-05,
+      "loss": 0.6781,
+      "mean_token_accuracy": 0.7756544649600983,
+      "num_tokens": 1231589.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.224128794670105,
+      "epoch": 0.8840864440078585,
+      "grad_norm": 2.3959298133850098,
+      "learning_rate": 6.363259400736359e-05,
+      "loss": 0.6584,
+      "mean_token_accuracy": 0.7872732043266296,
+      "num_tokens": 1245755.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.2522902011871337,
+      "epoch": 0.8939096267190569,
+      "grad_norm": 2.70693302154541,
+      "learning_rate": 6.284940107045363e-05,
+      "loss": 0.746,
+      "mean_token_accuracy": 0.7428298950195312,
+      "num_tokens": 1259830.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.262447714805603,
+      "epoch": 0.9037328094302554,
+      "grad_norm": 3.4064347743988037,
+      "learning_rate": 6.206281766170706e-05,
+      "loss": 0.7898,
+      "mean_token_accuracy": 0.7428923070430755,
+      "num_tokens": 1274186.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.2439752221107483,
+      "epoch": 0.9135559921414538,
+      "grad_norm": 3.0613200664520264,
+      "learning_rate": 6.127305133078945e-05,
+      "loss": 0.7384,
+      "mean_token_accuracy": 0.7455089151859283,
+      "num_tokens": 1288175.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.2410871624946593,
+      "epoch": 0.9233791748526523,
+      "grad_norm": 3.4402291774749756,
+      "learning_rate": 6.048031046721931e-05,
+      "loss": 0.7,
+      "mean_token_accuracy": 0.7690198957920075,
+      "num_tokens": 1301979.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.2460204362869263,
+      "epoch": 0.9332023575638507,
+      "grad_norm": 3.3230481147766113,
+      "learning_rate": 5.9684804245382056e-05,
+      "loss": 0.7529,
+      "mean_token_accuracy": 0.7487046122550964,
+      "num_tokens": 1316012.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.2684890747070312,
+      "epoch": 0.9430255402750491,
+      "grad_norm": 3.099296808242798,
+      "learning_rate": 5.888674256933671e-05,
+      "loss": 0.7804,
+      "mean_token_accuracy": 0.7289335012435914,
+      "num_tokens": 1329560.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.264595592021942,
+      "epoch": 0.9528487229862476,
+      "grad_norm": 3.251912832260132,
+      "learning_rate": 5.80863360174302e-05,
+      "loss": 0.7822,
+      "mean_token_accuracy": 0.7443600356578827,
+      "num_tokens": 1343980.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.2508861541748046,
+      "epoch": 0.962671905697446,
+      "grad_norm": 3.065115451812744,
+      "learning_rate": 5.72837957867337e-05,
+      "loss": 0.7604,
+      "mean_token_accuracy": 0.7514698207378387,
+      "num_tokens": 1358209.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.2174553155899048,
+      "epoch": 0.9724950884086444,
+      "grad_norm": 2.678147077560425,
+      "learning_rate": 5.64793336373156e-05,
+      "loss": 0.6222,
+      "mean_token_accuracy": 0.7993318200111389,
+      "num_tokens": 1371263.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.227048885822296,
+      "epoch": 0.9823182711198428,
+      "grad_norm": 2.8106882572174072,
+      "learning_rate": 5.567316183636615e-05,
+      "loss": 0.6209,
+      "mean_token_accuracy": 0.7868158459663391,
+      "num_tokens": 1385183.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.2130420684814454,
+      "epoch": 0.9921414538310412,
+      "grad_norm": 3.040210485458374,
+      "learning_rate": 5.4865493102188046e-05,
+      "loss": 0.6538,
+      "mean_token_accuracy": 0.7831201076507568,
+      "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_sgd_rot_e2_lr1e-4_s2_10k/checkpoint-1018/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:592e765a28d394e2860eed7fe93d6b99e1752cfe50c8df20897166b58fff93b9
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_sgd_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_sgd_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": [
+    "q_proj",
+    "gate_proj",
+    "up_proj",
+    "down_proj",
+    "o_proj",
+    "k_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_sgd_rot_e2_lr1e-4_s2_10k/checkpoint-2036/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4cbdc9eacf977c225270ba114cb148643ea8781fd25b11b3102dbed314be43ef
+size 10151357296

checkpoints/cat_qwen25_7b_r8_a32_sgd_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_sgd_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_sgd_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_sgd_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_sgd_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_sgd_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": 1.22729994058609,
+      "epoch": 0.009823182711198428,
+      "grad_norm": 4.8866095542907715,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.6928,
+      "mean_token_accuracy": 0.7725797176361084,
+      "num_tokens": 13828.0,
+      "step": 10
+    },
+    {
+      "entropy": 1.2600382804870605,
+      "epoch": 0.019646365422396856,
+      "grad_norm": 3.3233840465545654,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 0.8134,
+      "mean_token_accuracy": 0.7262678980827332,
+      "num_tokens": 28074.0,
+      "step": 20
+    },
+    {
+      "entropy": 1.2484176516532899,
+      "epoch": 0.029469548133595286,
+      "grad_norm": 2.9845433235168457,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 0.7786,
+      "mean_token_accuracy": 0.7460076689720154,
+      "num_tokens": 42168.0,
+      "step": 30
+    },
+    {
+      "entropy": 1.2452077984809875,
+      "epoch": 0.03929273084479371,
+      "grad_norm": 3.351901054382324,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 0.8251,
+      "mean_token_accuracy": 0.7294955134391785,
+      "num_tokens": 56700.0,
+      "step": 40
+    },
+    {
+      "entropy": 1.2434914350509643,
+      "epoch": 0.04911591355599214,
+      "grad_norm": 3.0272388458251953,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 0.7171,
+      "mean_token_accuracy": 0.7613231837749481,
+      "num_tokens": 70922.0,
+      "step": 50
+    },
+    {
+      "entropy": 1.253853189945221,
+      "epoch": 0.05893909626719057,
+      "grad_norm": 2.4795994758605957,
+      "learning_rate": 5.784313725490197e-05,
+      "loss": 0.7099,
+      "mean_token_accuracy": 0.7647240102291107,
+      "num_tokens": 84911.0,
+      "step": 60
+    },
+    {
+      "entropy": 1.2360968828201293,
+      "epoch": 0.068762278978389,
+      "grad_norm": 2.6639111042022705,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 0.7444,
+      "mean_token_accuracy": 0.755102014541626,
+      "num_tokens": 98835.0,
+      "step": 70
+    },
+    {
+      "entropy": 1.2527110695838928,
+      "epoch": 0.07858546168958742,
+      "grad_norm": 3.6153311729431152,
+      "learning_rate": 7.745098039215687e-05,
+      "loss": 0.7012,
+      "mean_token_accuracy": 0.7707957923412323,
+      "num_tokens": 112502.0,
+      "step": 80
+    },
+    {
+      "entropy": 1.2246053576469422,
+      "epoch": 0.08840864440078586,
+      "grad_norm": 3.169745683670044,
+      "learning_rate": 8.725490196078432e-05,
+      "loss": 0.7143,
+      "mean_token_accuracy": 0.76532541513443,
+      "num_tokens": 126248.0,
+      "step": 90
+    },
+    {
+      "entropy": 1.232993471622467,
+      "epoch": 0.09823182711198428,
+      "grad_norm": 3.4184534549713135,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 0.7402,
+      "mean_token_accuracy": 0.7552495896816254,
+      "num_tokens": 139551.0,
+      "step": 100
+    },
+    {
+      "entropy": 1.2279873371124268,
+      "epoch": 0.10805500982318271,
+      "grad_norm": 2.724405527114868,
+      "learning_rate": 9.999676765122786e-05,
+      "loss": 0.7571,
+      "mean_token_accuracy": 0.7568806827068328,
+      "num_tokens": 152975.0,
+      "step": 110
+    },
+    {
+      "entropy": 1.2499749660491943,
+      "epoch": 0.11787819253438114,
+      "grad_norm": 2.9191336631774902,
+      "learning_rate": 9.998093674494531e-05,
+      "loss": 0.7314,
+      "mean_token_accuracy": 0.760424691438675,
+      "num_tokens": 167073.0,
+      "step": 120
+    },
+    {
+      "entropy": 1.2526087641716004,
+      "epoch": 0.12770137524557956,
+      "grad_norm": 3.374049186706543,
+      "learning_rate": 9.995191775639568e-05,
+      "loss": 0.7667,
+      "mean_token_accuracy": 0.7505820035934448,
+      "num_tokens": 181404.0,
+      "step": 130
+    },
+    {
+      "entropy": 1.237971842288971,
+      "epoch": 0.137524557956778,
+      "grad_norm": 2.7548294067382812,
+      "learning_rate": 9.990971834259444e-05,
+      "loss": 0.7204,
+      "mean_token_accuracy": 0.7575949549674987,
+      "num_tokens": 195441.0,
+      "step": 140
+    },
+    {
+      "entropy": 1.2629727482795716,
+      "epoch": 0.14734774066797643,
+      "grad_norm": 3.8496453762054443,
+      "learning_rate": 9.985434963837376e-05,
+      "loss": 0.7956,
+      "mean_token_accuracy": 0.7347698986530304,
+      "num_tokens": 209863.0,
+      "step": 150
+    },
+    {
+      "entropy": 1.2493702173233032,
+      "epoch": 0.15717092337917485,
+      "grad_norm": 3.358459234237671,
+      "learning_rate": 9.978582625344435e-05,
+      "loss": 0.7489,
+      "mean_token_accuracy": 0.751927274465561,
+      "num_tokens": 223646.0,
+      "step": 160
+    },
+    {
+      "entropy": 1.2397171974182128,
+      "epoch": 0.16699410609037327,
+      "grad_norm": 3.4665334224700928,
+      "learning_rate": 9.970416626854049e-05,
+      "loss": 0.721,
+      "mean_token_accuracy": 0.7698734521865844,
+      "num_tokens": 237587.0,
+      "step": 170
+    },
+    {
+      "entropy": 1.2597252249717712,
+      "epoch": 0.17681728880157171,
+      "grad_norm": 2.362679958343506,
+      "learning_rate": 9.960939123064929e-05,
+      "loss": 0.7742,
+      "mean_token_accuracy": 0.740584260225296,
+      "num_tokens": 251760.0,
+      "step": 180
+    },
+    {
+      "entropy": 1.2215595960617065,
+      "epoch": 0.18664047151277013,
+      "grad_norm": 3.4586236476898193,
+      "learning_rate": 9.950152614732518e-05,
+      "loss": 0.7087,
+      "mean_token_accuracy": 0.7657637417316436,
+      "num_tokens": 265417.0,
+      "step": 190
+    },
+    {
+      "entropy": 1.2417829275131225,
+      "epoch": 0.19646365422396855,
+      "grad_norm": 3.494053602218628,
+      "learning_rate": 9.93805994800914e-05,
+      "loss": 0.7374,
+      "mean_token_accuracy": 0.7579931676387787,
+      "num_tokens": 278540.0,
+      "step": 200
+    },
+    {
+      "entropy": 1.2505549669265748,
+      "epoch": 0.206286836935167,
+      "grad_norm": 3.695319175720215,
+      "learning_rate": 9.924664313693006e-05,
+      "loss": 0.7837,
+      "mean_token_accuracy": 0.7425522744655609,
+      "num_tokens": 292295.0,
+      "step": 210
+    },
+    {
+      "entropy": 1.2271206736564637,
+      "epoch": 0.21611001964636542,
+      "grad_norm": 3.339724540710449,
+      "learning_rate": 9.909969246386284e-05,
+      "loss": 0.7669,
+      "mean_token_accuracy": 0.7515395283699036,
+      "num_tokens": 305466.0,
+      "step": 220
+    },
+    {
+      "entropy": 1.2639145135879517,
+      "epoch": 0.22593320235756384,
+      "grad_norm": 3.1140177249908447,
+      "learning_rate": 9.893978623562453e-05,
+      "loss": 0.7502,
+      "mean_token_accuracy": 0.7555740237236023,
+      "num_tokens": 320495.0,
+      "step": 230
+    },
+    {
+      "entropy": 1.2431163549423219,
+      "epoch": 0.2357563850687623,
+      "grad_norm": 3.665536403656006,
+      "learning_rate": 9.876696664543177e-05,
+      "loss": 0.7701,
+      "mean_token_accuracy": 0.7411514103412629,
+      "num_tokens": 333901.0,
+      "step": 240
+    },
+    {
+      "entropy": 1.2520034074783326,
+      "epoch": 0.2455795677799607,
+      "grad_norm": 4.050171375274658,
+      "learning_rate": 9.858127929384991e-05,
+      "loss": 0.7119,
+      "mean_token_accuracy": 0.7642730891704559,
+      "num_tokens": 347392.0,
+      "step": 250
+    },
+    {
+      "entropy": 1.243278980255127,
+      "epoch": 0.2554027504911591,
+      "grad_norm": 3.8479113578796387,
+      "learning_rate": 9.838277317676077e-05,
+      "loss": 0.761,
+      "mean_token_accuracy": 0.7522589504718781,
+      "num_tokens": 361140.0,
+      "step": 260
+    },
+    {
+      "entropy": 1.228507363796234,
+      "epoch": 0.26522593320235754,
+      "grad_norm": 2.529057264328003,
+      "learning_rate": 9.817150067243438e-05,
+      "loss": 0.689,
+      "mean_token_accuracy": 0.7683715045452117,
+      "num_tokens": 374661.0,
+      "step": 270
+    },
+    {
+      "entropy": 1.248687493801117,
+      "epoch": 0.275049115913556,
+      "grad_norm": 2.473571538925171,
+      "learning_rate": 9.794751752770845e-05,
+      "loss": 0.7849,
+      "mean_token_accuracy": 0.7473262250423431,
+      "num_tokens": 388415.0,
+      "step": 280
+    },
+    {
+      "entropy": 1.2462671041488647,
+      "epoch": 0.28487229862475444,
+      "grad_norm": 3.4700798988342285,
+      "learning_rate": 9.771088284327879e-05,
+      "loss": 0.7537,
+      "mean_token_accuracy": 0.7521771848201751,
+      "num_tokens": 402031.0,
+      "step": 290
+    },
+    {
+      "entropy": 1.2523157238960265,
+      "epoch": 0.29469548133595286,
+      "grad_norm": 3.2036030292510986,
+      "learning_rate": 9.746165905810488e-05,
+      "loss": 0.7551,
+      "mean_token_accuracy": 0.7571269392967224,
+      "num_tokens": 416165.0,
+      "step": 300
+    },
+    {
+      "entropy": 1.2602078080177308,
+      "epoch": 0.3045186640471513,
+      "grad_norm": 2.589961290359497,
+      "learning_rate": 9.719991193293466e-05,
+      "loss": 0.7821,
+      "mean_token_accuracy": 0.7466304242610932,
+      "num_tokens": 429910.0,
+      "step": 310
+    },
+    {
+      "entropy": 1.2587228298187256,
+      "epoch": 0.3143418467583497,
+      "grad_norm": 4.473543643951416,
+      "learning_rate": 9.692571053295268e-05,
+      "loss": 0.758,
+      "mean_token_accuracy": 0.7488924205303192,
+      "num_tokens": 443725.0,
+      "step": 320
+    },
+    {
+      "entropy": 1.2587100744247437,
+      "epoch": 0.3241650294695481,
+      "grad_norm": 2.999272584915161,
+      "learning_rate": 9.663912720955637e-05,
+      "loss": 0.7934,
+      "mean_token_accuracy": 0.7434530913829803,
+      "num_tokens": 458234.0,
+      "step": 330
+    },
+    {
+      "entropy": 1.235784935951233,
+      "epoch": 0.33398821218074654,
+      "grad_norm": 3.993062734603882,
+      "learning_rate": 9.634023758126539e-05,
+      "loss": 0.7378,
+      "mean_token_accuracy": 0.7529276013374329,
+      "num_tokens": 471402.0,
+      "step": 340
+    },
+    {
+      "entropy": 1.212973964214325,
+      "epoch": 0.343811394891945,
+      "grad_norm": 3.497361183166504,
+      "learning_rate": 9.602912051376859e-05,
+      "loss": 0.6893,
+      "mean_token_accuracy": 0.773086953163147,
+      "num_tokens": 484651.0,
+      "step": 350
+    },
+    {
+      "entropy": 1.2527001976966858,
+      "epoch": 0.35363457760314343,
+      "grad_norm": 2.310852289199829,
+      "learning_rate": 9.570585809911441e-05,
+      "loss": 0.7795,
+      "mean_token_accuracy": 0.7472991645336151,
+      "num_tokens": 497956.0,
+      "step": 360
+    },
+    {
+      "entropy": 1.2270971417427063,
+      "epoch": 0.36345776031434185,
+      "grad_norm": 3.0224449634552,
+      "learning_rate": 9.537053563404992e-05,
+      "loss": 0.7333,
+      "mean_token_accuracy": 0.7572860002517701,
+      "num_tokens": 511717.0,
+      "step": 370
+    },
+    {
+      "entropy": 1.2353201866149903,
+      "epoch": 0.37328094302554027,
+      "grad_norm": 3.016392946243286,
+      "learning_rate": 9.502324159751412e-05,
+      "loss": 0.7056,
+      "mean_token_accuracy": 0.7640936195850372,
+      "num_tokens": 525633.0,
+      "step": 380
+    },
+    {
+      "entropy": 1.2572155117988586,
+      "epoch": 0.3831041257367387,
+      "grad_norm": 3.6709957122802734,
+      "learning_rate": 9.466406762729181e-05,
+      "loss": 0.7471,
+      "mean_token_accuracy": 0.7522993624210358,
+      "num_tokens": 538945.0,
+      "step": 390
+    },
+    {
+      "entropy": 1.2437490463256835,
+      "epoch": 0.3929273084479371,
+      "grad_norm": 2.7692830562591553,
+      "learning_rate": 9.429310849583378e-05,
+      "loss": 0.746,
+      "mean_token_accuracy": 0.754131942987442,
+      "num_tokens": 552861.0,
+      "step": 400
+    },
+    {
+      "entropy": 1.2515570044517517,
+      "epoch": 0.4027504911591356,
+      "grad_norm": 2.845248222351074,
+      "learning_rate": 9.391046208524989e-05,
+      "loss": 0.7987,
+      "mean_token_accuracy": 0.736023235321045,
+      "num_tokens": 565955.0,
+      "step": 410
+    },
+    {
+      "entropy": 1.2464410781860351,
+      "epoch": 0.412573673870334,
+      "grad_norm": 3.273564577102661,
+      "learning_rate": 9.351622936148182e-05,
+      "loss": 0.7226,
+      "mean_token_accuracy": 0.7609843373298645,
+      "num_tokens": 579923.0,
+      "step": 420
+    },
+    {
+      "entropy": 1.2555230379104614,
+      "epoch": 0.4223968565815324,
+      "grad_norm": 3.4945361614227295,
+      "learning_rate": 9.31105143476618e-05,
+      "loss": 0.7408,
+      "mean_token_accuracy": 0.7554882347583771,
+      "num_tokens": 594071.0,
+      "step": 430
+    },
+    {
+      "entropy": 1.2536908864974976,
+      "epoch": 0.43222003929273084,
+      "grad_norm": 2.4363486766815186,
+      "learning_rate": 9.269342409666495e-05,
+      "loss": 0.7346,
+      "mean_token_accuracy": 0.7615235865116119,
+      "num_tokens": 608545.0,
+      "step": 440
+    },
+    {
+      "entropy": 1.228044366836548,
+      "epoch": 0.44204322200392926,
+      "grad_norm": 3.0561442375183105,
+      "learning_rate": 9.226506866286202e-05,
+      "loss": 0.7296,
+      "mean_token_accuracy": 0.7626752912998199,
+      "num_tokens": 621975.0,
+      "step": 450
+    },
+    {
+      "entropy": 1.257164752483368,
+      "epoch": 0.4518664047151277,
+      "grad_norm": 2.552243232727051,
+      "learning_rate": 9.18255610730802e-05,
+      "loss": 0.7432,
+      "mean_token_accuracy": 0.7537168443202973,
+      "num_tokens": 636061.0,
+      "step": 460
+    },
+    {
+      "entropy": 1.2517608165740968,
+      "epoch": 0.46168958742632615,
+      "grad_norm": 3.190830945968628,
+      "learning_rate": 9.137501729677962e-05,
+      "loss": 0.75,
+      "mean_token_accuracy": 0.7544792711734771,
+      "num_tokens": 649717.0,
+      "step": 470
+    },
+    {
+      "entropy": 1.235853898525238,
+      "epoch": 0.4715127701375246,
+      "grad_norm": 2.730757474899292,
+      "learning_rate": 9.091355621545339e-05,
+      "loss": 0.6943,
+      "mean_token_accuracy": 0.7740860462188721,
+      "num_tokens": 664123.0,
+      "step": 480
+    },
+    {
+      "entropy": 1.2552549839019775,
+      "epoch": 0.481335952848723,
+      "grad_norm": 3.690030813217163,
+      "learning_rate": 9.044129959125923e-05,
+      "loss": 0.7954,
+      "mean_token_accuracy": 0.7400918424129486,
+      "num_tokens": 677803.0,
+      "step": 490
+    },
+    {
+      "entropy": 1.2475018978118897,
+      "epoch": 0.4911591355599214,
+      "grad_norm": 3.209927558898926,
+      "learning_rate": 8.995837203489108e-05,
+      "loss": 0.7097,
+      "mean_token_accuracy": 0.7667286634445191,
+      "num_tokens": 691461.0,
+      "step": 500
+    },
+    {
+      "entropy": 1.2404452085494995,
+      "epoch": 0.5009823182711198,
+      "grad_norm": 2.9024016857147217,
+      "learning_rate": 8.946490097269899e-05,
+      "loss": 0.686,
+      "mean_token_accuracy": 0.7731779038906097,
+      "num_tokens": 705470.0,
+      "step": 510
+    },
+    {
+      "entropy": 1.2764546751976014,
+      "epoch": 0.5108055009823183,
+      "grad_norm": 3.618655204772949,
+      "learning_rate": 8.896101661306618e-05,
+      "loss": 0.8615,
+      "mean_token_accuracy": 0.715438824892044,
+      "num_tokens": 718805.0,
+      "step": 520
+    },
+    {
+      "entropy": 1.252146315574646,
+      "epoch": 0.5206286836935167,
+      "grad_norm": 2.5902345180511475,
+      "learning_rate": 8.844685191205193e-05,
+      "loss": 0.7396,
+      "mean_token_accuracy": 0.7525096774101258,
+      "num_tokens": 732367.0,
+      "step": 530
+    },
+    {
+      "entropy": 1.2653416037559508,
+      "epoch": 0.5304518664047151,
+      "grad_norm": 3.163494825363159,
+      "learning_rate": 8.792254253830945e-05,
+      "loss": 0.7787,
+      "mean_token_accuracy": 0.7419969081878662,
+      "num_tokens": 746431.0,
+      "step": 540
+    },
+    {
+      "entropy": 1.2420126438140868,
+      "epoch": 0.5402750491159135,
+      "grad_norm": 2.3512191772460938,
+      "learning_rate": 8.738822683728814e-05,
+      "loss": 0.7096,
+      "mean_token_accuracy": 0.7669752776622772,
+      "num_tokens": 760182.0,
+      "step": 550
+    },
+    {
+      "entropy": 1.2616332292556762,
+      "epoch": 0.550098231827112,
+      "grad_norm": 2.4627273082733154,
+      "learning_rate": 8.684404579472936e-05,
+      "loss": 0.776,
+      "mean_token_accuracy": 0.7481194794178009,
+      "num_tokens": 774137.0,
+      "step": 560
+    },
+    {
+      "entropy": 1.220177948474884,
+      "epoch": 0.5599214145383105,
+      "grad_norm": 3.291656494140625,
+      "learning_rate": 8.629014299946568e-05,
+      "loss": 0.6453,
+      "mean_token_accuracy": 0.7858670830726624,
+      "num_tokens": 787806.0,
+      "step": 570
+    },
+    {
+      "entropy": 1.2351359486579896,
+      "epoch": 0.5697445972495089,
+      "grad_norm": 2.42910099029541,
+      "learning_rate": 8.572666460553316e-05,
+      "loss": 0.6767,
+      "mean_token_accuracy": 0.7722299098968506,
+      "num_tokens": 801431.0,
+      "step": 580
+    },
+    {
+      "entropy": 1.2556207776069641,
+      "epoch": 0.5795677799607073,
+      "grad_norm": 3.0838305950164795,
+      "learning_rate": 8.515375929360687e-05,
+      "loss": 0.7984,
+      "mean_token_accuracy": 0.741753101348877,
+      "num_tokens": 815080.0,
+      "step": 590
+    },
+    {
+      "entropy": 1.2549353241920471,
+      "epoch": 0.5893909626719057,
+      "grad_norm": 2.646620512008667,
+      "learning_rate": 8.45715782317696e-05,
+      "loss": 0.7612,
+      "mean_token_accuracy": 0.7506958603858948,
+      "num_tokens": 829061.0,
+      "step": 600
+    },
+    {
+      "entropy": 1.238725221157074,
+      "epoch": 0.5992141453831041,
+      "grad_norm": 3.156494379043579,
+      "learning_rate": 8.39802750356244e-05,
+      "loss": 0.7156,
+      "mean_token_accuracy": 0.7645548522472382,
+      "num_tokens": 842325.0,
+      "step": 610
+    },
+    {
+      "entropy": 1.2504855751991273,
+      "epoch": 0.6090373280943026,
+      "grad_norm": 3.688000440597534,
+      "learning_rate": 8.338000572776114e-05,
+      "loss": 0.7688,
+      "mean_token_accuracy": 0.7456511080265045,
+      "num_tokens": 856377.0,
+      "step": 620
+    },
+    {
+      "entropy": 1.2575064301490784,
+      "epoch": 0.618860510805501,
+      "grad_norm": 3.6037731170654297,
+      "learning_rate": 8.277092869658808e-05,
+      "loss": 0.7347,
+      "mean_token_accuracy": 0.7624819874763489,
+      "num_tokens": 870502.0,
+      "step": 630
+    },
+    {
+      "entropy": 1.2600618243217467,
+      "epoch": 0.6286836935166994,
+      "grad_norm": 2.7635698318481445,
+      "learning_rate": 8.215320465453914e-05,
+      "loss": 0.7719,
+      "mean_token_accuracy": 0.7426825046539307,
+      "num_tokens": 884708.0,
+      "step": 640
+    },
+    {
+      "entropy": 1.2481453895568848,
+      "epoch": 0.6385068762278978,
+      "grad_norm": 2.3514113426208496,
+      "learning_rate": 8.152699659566797e-05,
+      "loss": 0.7902,
+      "mean_token_accuracy": 0.736629843711853,
+      "num_tokens": 898142.0,
+      "step": 650
+    },
+    {
+      "entropy": 1.2575042247772217,
+      "epoch": 0.6483300589390962,
+      "grad_norm": 3.3685693740844727,
+      "learning_rate": 8.089246975263996e-05,
+      "loss": 0.7412,
+      "mean_token_accuracy": 0.7545578002929687,
+      "num_tokens": 912820.0,
+      "step": 660
+    },
+    {
+      "entropy": 1.226671326160431,
+      "epoch": 0.6581532416502947,
+      "grad_norm": 3.102717161178589,
+      "learning_rate": 8.024979155313354e-05,
+      "loss": 0.6868,
+      "mean_token_accuracy": 0.779555332660675,
+      "num_tokens": 926399.0,
+      "step": 670
+    },
+    {
+      "entropy": 1.24713693857193,
+      "epoch": 0.6679764243614931,
+      "grad_norm": 2.6191911697387695,
+      "learning_rate": 7.959913157566246e-05,
+      "loss": 0.7767,
+      "mean_token_accuracy": 0.7440745651721954,
+      "num_tokens": 940133.0,
+      "step": 680
+    },
+    {
+      "entropy": 1.2643288373947144,
+      "epoch": 0.6777996070726916,
+      "grad_norm": 2.782747983932495,
+      "learning_rate": 7.894066150483023e-05,
+      "loss": 0.7409,
+      "mean_token_accuracy": 0.7547606706619263,
+      "num_tokens": 954504.0,
+      "step": 690
+    },
+    {
+      "entropy": 1.2425753355026246,
+      "epoch": 0.68762278978389,
+      "grad_norm": 3.5112552642822266,
+      "learning_rate": 7.827455508602923e-05,
+      "loss": 0.7657,
+      "mean_token_accuracy": 0.7505813300609588,
+      "num_tokens": 968934.0,
+      "step": 700
+    },
+    {
+      "entropy": 1.2537635564804077,
+      "epoch": 0.6974459724950884,
+      "grad_norm": 3.150362491607666,
+      "learning_rate": 7.760098807959573e-05,
+      "loss": 0.7409,
+      "mean_token_accuracy": 0.7455904126167298,
+      "num_tokens": 983038.0,
+      "step": 710
+    },
+    {
+      "entropy": 1.2538934469223022,
+      "epoch": 0.7072691552062869,
+      "grad_norm": 2.8867599964141846,
+      "learning_rate": 7.69201382144334e-05,
+      "loss": 0.7457,
+      "mean_token_accuracy": 0.746702516078949,
+      "num_tokens": 996748.0,
+      "step": 720
+    },
+    {
+      "entropy": 1.264196515083313,
+      "epoch": 0.7170923379174853,
+      "grad_norm": 2.841606855392456,
+      "learning_rate": 7.623218514111743e-05,
+      "loss": 0.762,
+      "mean_token_accuracy": 0.7420303642749786,
+      "num_tokens": 1010595.0,
+      "step": 730
+    },
+    {
+      "entropy": 1.238691246509552,
+      "epoch": 0.7269155206286837,
+      "grad_norm": 3.15217661857605,
+      "learning_rate": 7.553731038449144e-05,
+      "loss": 0.676,
+      "mean_token_accuracy": 0.7726105868816375,
+      "num_tokens": 1024890.0,
+      "step": 740
+    },
+    {
+      "entropy": 1.2556055784225464,
+      "epoch": 0.7367387033398821,
+      "grad_norm": 2.4604084491729736,
+      "learning_rate": 7.483569729576996e-05,
+      "loss": 0.7578,
+      "mean_token_accuracy": 0.7490179061889648,
+      "num_tokens": 1038967.0,
+      "step": 750
+    },
+    {
+      "entropy": 1.2183940052986144,
+      "epoch": 0.7465618860510805,
+      "grad_norm": 2.6620943546295166,
+      "learning_rate": 7.412753100415897e-05,
+      "loss": 0.651,
+      "mean_token_accuracy": 0.7827771067619324,
+      "num_tokens": 1052622.0,
+      "step": 760
+    },
+    {
+      "entropy": 1.2360458731651307,
+      "epoch": 0.756385068762279,
+      "grad_norm": 3.445331335067749,
+      "learning_rate": 7.341299836800729e-05,
+      "loss": 0.7496,
+      "mean_token_accuracy": 0.7527473092079162,
+      "num_tokens": 1066588.0,
+      "step": 770
+    },
+    {
+      "entropy": 1.2387821793556213,
+      "epoch": 0.7662082514734774,
+      "grad_norm": 3.357232093811035,
+      "learning_rate": 7.26922879255017e-05,
+      "loss": 0.6619,
+      "mean_token_accuracy": 0.7745956540107727,
+      "num_tokens": 1080453.0,
+      "step": 780
+    },
+    {
+      "entropy": 1.2365756273269652,
+      "epoch": 0.7760314341846758,
+      "grad_norm": 3.045243263244629,
+      "learning_rate": 7.196558984491889e-05,
+      "loss": 0.7166,
+      "mean_token_accuracy": 0.7608296036720276,
+      "num_tokens": 1094550.0,
+      "step": 790
+    },
+    {
+      "entropy": 1.2373855471611024,
+      "epoch": 0.7858546168958742,
+      "grad_norm": 4.2761359214782715,
+      "learning_rate": 7.123309587444711e-05,
+      "loss": 0.6843,
+      "mean_token_accuracy": 0.7768501520156861,
+      "num_tokens": 1108808.0,
+      "step": 800
+    },
+    {
+      "entropy": 1.2416929841041564,
+      "epoch": 0.7956777996070727,
+      "grad_norm": 3.202831983566284,
+      "learning_rate": 7.04949992915912e-05,
+      "loss": 0.7058,
+      "mean_token_accuracy": 0.761394602060318,
+      "num_tokens": 1122365.0,
+      "step": 810
+    },
+    {
+      "entropy": 1.2804744243621826,
+      "epoch": 0.8055009823182712,
+      "grad_norm": 3.275981903076172,
+      "learning_rate": 6.975149485217388e-05,
+      "loss": 0.7698,
+      "mean_token_accuracy": 0.7347693979740143,
+      "num_tokens": 1136376.0,
+      "step": 820
+    },
+    {
+      "entropy": 1.2437939643859863,
+      "epoch": 0.8153241650294696,
+      "grad_norm": 2.6627252101898193,
+      "learning_rate": 6.900277873894705e-05,
+      "loss": 0.6874,
+      "mean_token_accuracy": 0.7675921380519867,
+      "num_tokens": 1150011.0,
+      "step": 830
+    },
+    {
+      "entropy": 1.2557293772697449,
+      "epoch": 0.825147347740668,
+      "grad_norm": 2.6019535064697266,
+      "learning_rate": 6.824904850982656e-05,
+      "loss": 0.6945,
+      "mean_token_accuracy": 0.7684582710266114,
+      "num_tokens": 1164190.0,
+      "step": 840
+    },
+    {
+      "entropy": 1.2451830863952638,
+      "epoch": 0.8349705304518664,
+      "grad_norm": 3.0742592811584473,
+      "learning_rate": 6.749050304576415e-05,
+      "loss": 0.6961,
+      "mean_token_accuracy": 0.7662852883338929,
+      "num_tokens": 1177346.0,
+      "step": 850
+    },
+    {
+      "entropy": 1.2350680232048035,
+      "epoch": 0.8447937131630648,
+      "grad_norm": 3.061077117919922,
+      "learning_rate": 6.672734249827028e-05,
+      "loss": 0.6833,
+      "mean_token_accuracy": 0.780314964056015,
+      "num_tokens": 1190692.0,
+      "step": 860
+    },
+    {
+      "entropy": 1.250148320198059,
+      "epoch": 0.8546168958742633,
+      "grad_norm": 2.8277947902679443,
+      "learning_rate": 6.59597682366017e-05,
+      "loss": 0.7275,
+      "mean_token_accuracy": 0.7577569246292114,
+      "num_tokens": 1204254.0,
+      "step": 870
+    },
+    {
+      "entropy": 1.2564526915550231,
+      "epoch": 0.8644400785854617,
+      "grad_norm": 3.7227401733398438,
+      "learning_rate": 6.518798279462766e-05,
+      "loss": 0.7498,
+      "mean_token_accuracy": 0.7435093998908997,
+      "num_tokens": 1217938.0,
+      "step": 880
+    },
+    {
+      "entropy": 1.2478920340538024,
+      "epoch": 0.8742632612966601,
+      "grad_norm": 2.8872194290161133,
+      "learning_rate": 6.441218981738895e-05,
+      "loss": 0.6781,
+      "mean_token_accuracy": 0.7756544649600983,
+      "num_tokens": 1231589.0,
+      "step": 890
+    },
+    {
+      "entropy": 1.224128794670105,
+      "epoch": 0.8840864440078585,
+      "grad_norm": 2.3959298133850098,
+      "learning_rate": 6.363259400736359e-05,
+      "loss": 0.6584,
+      "mean_token_accuracy": 0.7872732043266296,
+      "num_tokens": 1245755.0,
+      "step": 900
+    },
+    {
+      "entropy": 1.2522902011871337,
+      "epoch": 0.8939096267190569,
+      "grad_norm": 2.70693302154541,
+      "learning_rate": 6.284940107045363e-05,
+      "loss": 0.746,
+      "mean_token_accuracy": 0.7428298950195312,
+      "num_tokens": 1259830.0,
+      "step": 910
+    },
+    {
+      "entropy": 1.262447714805603,
+      "epoch": 0.9037328094302554,
+      "grad_norm": 3.4064347743988037,
+      "learning_rate": 6.206281766170706e-05,
+      "loss": 0.7898,
+      "mean_token_accuracy": 0.7428923070430755,
+      "num_tokens": 1274186.0,
+      "step": 920
+    },
+    {
+      "entropy": 1.2439752221107483,
+      "epoch": 0.9135559921414538,
+      "grad_norm": 3.0613200664520264,
+      "learning_rate": 6.127305133078945e-05,
+      "loss": 0.7384,
+      "mean_token_accuracy": 0.7455089151859283,
+      "num_tokens": 1288175.0,
+      "step": 930
+    },
+    {
+      "entropy": 1.2410871624946593,
+      "epoch": 0.9233791748526523,
+      "grad_norm": 3.4402291774749756,
+      "learning_rate": 6.048031046721931e-05,
+      "loss": 0.7,
+      "mean_token_accuracy": 0.7690198957920075,
+      "num_tokens": 1301979.0,
+      "step": 940
+    },
+    {
+      "entropy": 1.2460204362869263,
+      "epoch": 0.9332023575638507,
+      "grad_norm": 3.3230481147766113,
+      "learning_rate": 5.9684804245382056e-05,
+      "loss": 0.7529,
+      "mean_token_accuracy": 0.7487046122550964,
+      "num_tokens": 1316012.0,
+      "step": 950
+    },
+    {
+      "entropy": 1.2684890747070312,
+      "epoch": 0.9430255402750491,
+      "grad_norm": 3.099296808242798,
+      "learning_rate": 5.888674256933671e-05,
+      "loss": 0.7804,
+      "mean_token_accuracy": 0.7289335012435914,
+      "num_tokens": 1329560.0,
+      "step": 960
+    },
+    {
+      "entropy": 1.264595592021942,
+      "epoch": 0.9528487229862476,
+      "grad_norm": 3.251912832260132,
+      "learning_rate": 5.80863360174302e-05,
+      "loss": 0.7822,
+      "mean_token_accuracy": 0.7443600356578827,
+      "num_tokens": 1343980.0,
+      "step": 970
+    },
+    {
+      "entropy": 1.2508861541748046,
+      "epoch": 0.962671905697446,
+      "grad_norm": 3.065115451812744,
+      "learning_rate": 5.72837957867337e-05,
+      "loss": 0.7604,
+      "mean_token_accuracy": 0.7514698207378387,
+      "num_tokens": 1358209.0,
+      "step": 980
+    },
+    {
+      "entropy": 1.2174553155899048,
+      "epoch": 0.9724950884086444,
+      "grad_norm": 2.678147077560425,
+      "learning_rate": 5.64793336373156e-05,
+      "loss": 0.6222,
+      "mean_token_accuracy": 0.7993318200111389,
+      "num_tokens": 1371263.0,
+      "step": 990
+    },
+    {
+      "entropy": 1.227048885822296,
+      "epoch": 0.9823182711198428,
+      "grad_norm": 2.8106882572174072,
+      "learning_rate": 5.567316183636615e-05,
+      "loss": 0.6209,
+      "mean_token_accuracy": 0.7868158459663391,
+      "num_tokens": 1385183.0,
+      "step": 1000
+    },
+    {
+      "entropy": 1.2130420684814454,
+      "epoch": 0.9921414538310412,
+      "grad_norm": 3.040210485458374,
+      "learning_rate": 5.4865493102188046e-05,
+      "loss": 0.6538,
+      "mean_token_accuracy": 0.7831201076507568,
+      "num_tokens": 1398309.0,
+      "step": 1010
+    },
+    {
+      "entropy": 1.235917067527771,
+      "epoch": 1.0019646365422397,
+      "grad_norm": 3.978003740310669,
+      "learning_rate": 5.405654054806817e-05,
+      "loss": 0.6822,
+      "mean_token_accuracy": 0.7721337080001831,
+      "num_tokens": 1411126.0,
+      "step": 1020
+    },
+    {
+      "entropy": 1.255300521850586,
+      "epoch": 1.0117878192534382,
+      "grad_norm": 3.0687496662139893,
+      "learning_rate": 5.3246517626045025e-05,
+      "loss": 0.7739,
+      "mean_token_accuracy": 0.7521123230457306,
+      "num_tokens": 1425582.0,
+      "step": 1030
+    },
+    {
+      "entropy": 1.2608398914337158,
+      "epoch": 1.0216110019646365,
+      "grad_norm": 3.2460482120513916,
+      "learning_rate": 5.243563807058676e-05,
+      "loss": 0.7685,
+      "mean_token_accuracy": 0.7391750454902649,
+      "num_tokens": 1439527.0,
+      "step": 1040
+    },
+    {
+      "entropy": 1.2571334958076477,
+      "epoch": 1.031434184675835,
+      "grad_norm": 3.166513442993164,
+      "learning_rate": 5.162411584219476e-05,
+      "loss": 0.7685,
+      "mean_token_accuracy": 0.7383076310157776,
+      "num_tokens": 1453445.0,
+      "step": 1050
+    },
+    {
+      "entropy": 1.2584318399429322,
+      "epoch": 1.0412573673870333,
+      "grad_norm": 2.82977294921875,
+      "learning_rate": 5.0812165070947594e-05,
+      "loss": 0.7001,
+      "mean_token_accuracy": 0.7650222659111023,
+      "num_tokens": 1466353.0,
+      "step": 1060
+    },
+    {
+      "entropy": 1.2605103731155396,
+      "epoch": 1.0510805500982319,
+      "grad_norm": 2.853358507156372,
+      "learning_rate": 5e-05,
+      "loss": 0.7222,
+      "mean_token_accuracy": 0.7617378056049346,
+      "num_tokens": 1481062.0,
+      "step": 1070
+    },
+    {
+      "entropy": 1.2653095126152039,
+      "epoch": 1.0609037328094302,
+      "grad_norm": 2.4662859439849854,
+      "learning_rate": 4.918783492905242e-05,
+      "loss": 0.737,
+      "mean_token_accuracy": 0.7466691851615905,
+      "num_tokens": 1495189.0,
+      "step": 1080
+    },
+    {
+      "entropy": 1.260548436641693,
+      "epoch": 1.0707269155206287,
+      "grad_norm": 3.857428789138794,
+      "learning_rate": 4.837588415780524e-05,
+      "loss": 0.7494,
+      "mean_token_accuracy": 0.7522172629833221,
+      "num_tokens": 1508841.0,
+      "step": 1090
+    },
+    {
+      "entropy": 1.2647943496704102,
+      "epoch": 1.080550098231827,
+      "grad_norm": 3.000452995300293,
+      "learning_rate": 4.756436192941325e-05,
+      "loss": 0.7613,
+      "mean_token_accuracy": 0.7543927311897278,
+      "num_tokens": 1522922.0,
+      "step": 1100
+    },
+    {
+      "entropy": 1.253216314315796,
+      "epoch": 1.0903732809430255,
+      "grad_norm": 2.362539529800415,
+      "learning_rate": 4.675348237395499e-05,
+      "loss": 0.7059,
+      "mean_token_accuracy": 0.7637098848819732,
+      "num_tokens": 1536867.0,
+      "step": 1110
+    },
+    {
+      "entropy": 1.2310373902320861,
+      "epoch": 1.1001964636542239,
+      "grad_norm": 2.689497709274292,
+      "learning_rate": 4.594345945193183e-05,
+      "loss": 0.7194,
+      "mean_token_accuracy": 0.7579232811927795,
+      "num_tokens": 1550230.0,
+      "step": 1120
+    },
+    {
+      "entropy": 1.2513616800308227,
+      "epoch": 1.1100196463654224,
+      "grad_norm": 2.0791637897491455,
+      "learning_rate": 4.513450689781195e-05,
+      "loss": 0.7203,
+      "mean_token_accuracy": 0.7562252283096313,
+      "num_tokens": 1562941.0,
+      "step": 1130
+    },
+    {
+      "entropy": 1.2390633702278138,
+      "epoch": 1.119842829076621,
+      "grad_norm": 2.904214859008789,
+      "learning_rate": 4.432683816363386e-05,
+      "loss": 0.6939,
+      "mean_token_accuracy": 0.7680689752101898,
+      "num_tokens": 1576546.0,
+      "step": 1140
+    },
+    {
+      "entropy": 1.2656402826309203,
+      "epoch": 1.1296660117878192,
+      "grad_norm": 3.0298361778259277,
+      "learning_rate": 4.3520666362684395e-05,
+      "loss": 0.827,
+      "mean_token_accuracy": 0.723938125371933,
+      "num_tokens": 1590129.0,
+      "step": 1150
+    },
+    {
+      "entropy": 1.2388361692428589,
+      "epoch": 1.1394891944990178,
+      "grad_norm": 2.8770015239715576,
+      "learning_rate": 4.271620421326633e-05,
+      "loss": 0.6506,
+      "mean_token_accuracy": 0.7753389775753021,
+      "num_tokens": 1603834.0,
+      "step": 1160
+    },
+    {
+      "entropy": 1.239013707637787,
+      "epoch": 1.149312377210216,
+      "grad_norm": 2.4905548095703125,
+      "learning_rate": 4.191366398256983e-05,
+      "loss": 0.6668,
+      "mean_token_accuracy": 0.7760896265506745,
+      "num_tokens": 1617951.0,
+      "step": 1170
+    },
+    {
+      "entropy": 1.243638026714325,
+      "epoch": 1.1591355599214146,
+      "grad_norm": 2.737159252166748,
+      "learning_rate": 4.111325743066332e-05,
+      "loss": 0.7346,
+      "mean_token_accuracy": 0.7547667324542999,
+      "num_tokens": 1631987.0,
+      "step": 1180
+    },
+    {
+      "entropy": 1.2218198776245117,
+      "epoch": 1.168958742632613,
+      "grad_norm": 3.259192705154419,
+      "learning_rate": 4.0315195754617976e-05,
+      "loss": 0.6957,
+      "mean_token_accuracy": 0.7640621840953827,
+      "num_tokens": 1645088.0,
+      "step": 1190
+    },
+    {
+      "entropy": 1.2639505505561828,
+      "epoch": 1.1787819253438114,
+      "grad_norm": 1.9223480224609375,
+      "learning_rate": 3.95196895327807e-05,
+      "loss": 0.7149,
+      "mean_token_accuracy": 0.7594054222106934,
+      "num_tokens": 1659227.0,
+      "step": 1200
+    },
+    {
+      "entropy": 1.230186104774475,
+      "epoch": 1.1886051080550097,
+      "grad_norm": 2.6639060974121094,
+      "learning_rate": 3.872694866921056e-05,
+      "loss": 0.6927,
+      "mean_token_accuracy": 0.76207355260849,
+      "num_tokens": 1672545.0,
+      "step": 1210
+    },
+    {
+      "entropy": 1.234763741493225,
+      "epoch": 1.1984282907662083,
+      "grad_norm": 2.2439212799072266,
+      "learning_rate": 3.793718233829295e-05,
+      "loss": 0.6199,
+      "mean_token_accuracy": 0.7880393087863922,
+      "num_tokens": 1686702.0,
+      "step": 1220
+    },
+    {
+      "entropy": 1.2578733086585998,
+      "epoch": 1.2082514734774068,
+      "grad_norm": 2.7486374378204346,
+      "learning_rate": 3.715059892954639e-05,
+      "loss": 0.7284,
+      "mean_token_accuracy": 0.7554497301578522,
+      "num_tokens": 1701509.0,
+      "step": 1230
+    },
+    {
+      "entropy": 1.2538498401641847,
+      "epoch": 1.218074656188605,
+      "grad_norm": 3.591052293777466,
+      "learning_rate": 3.636740599263643e-05,
+      "loss": 0.7894,
+      "mean_token_accuracy": 0.7418418228626251,
+      "num_tokens": 1715330.0,
+      "step": 1240
+    },
+    {
+      "entropy": 1.2618250370025634,
+      "epoch": 1.2278978388998034,
+      "grad_norm": 2.5035250186920166,
+      "learning_rate": 3.5587810182611066e-05,
+      "loss": 0.6843,
+      "mean_token_accuracy": 0.7672065496444702,
+      "num_tokens": 1728655.0,
+      "step": 1250
+    },
+    {
+      "entropy": 1.2594892859458924,
+      "epoch": 1.237721021611002,
+      "grad_norm": 3.7033205032348633,
+      "learning_rate": 3.481201720537235e-05,
+      "loss": 0.7464,
+      "mean_token_accuracy": 0.7494923353195191,
+      "num_tokens": 1741961.0,
+      "step": 1260
+    },
+    {
+      "entropy": 1.2491514444351197,
+      "epoch": 1.2475442043222005,
+      "grad_norm": 2.840648651123047,
+      "learning_rate": 3.4040231763398315e-05,
+      "loss": 0.7073,
+      "mean_token_accuracy": 0.7620794713497162,
+      "num_tokens": 1755285.0,
+      "step": 1270
+    },
+    {
+      "entropy": 1.2515382051467896,
+      "epoch": 1.2573673870333988,
+      "grad_norm": 2.8188557624816895,
+      "learning_rate": 3.3272657501729714e-05,
+      "loss": 0.7372,
+      "mean_token_accuracy": 0.7472140192985535,
+      "num_tokens": 1768670.0,
+      "step": 1280
+    },
+    {
+      "entropy": 1.2420023441314698,
+      "epoch": 1.2671905697445973,
+      "grad_norm": 3.649625062942505,
+      "learning_rate": 3.250949695423585e-05,
+      "loss": 0.6466,
+      "mean_token_accuracy": 0.7775622367858886,
+      "num_tokens": 1782525.0,
+      "step": 1290
+    },
+    {
+      "entropy": 1.2510635375976562,
+      "epoch": 1.2770137524557956,
+      "grad_norm": 2.5393943786621094,
+      "learning_rate": 3.175095149017344e-05,
+      "loss": 0.6954,
+      "mean_token_accuracy": 0.7695215404033661,
+      "num_tokens": 1796010.0,
+      "step": 1300
+    },
+    {
+      "entropy": 1.240469217300415,
+      "epoch": 1.2868369351669942,
+      "grad_norm": 3.0604372024536133,
+      "learning_rate": 3.099722126105295e-05,
+      "loss": 0.7011,
+      "mean_token_accuracy": 0.7616488993167877,
+      "num_tokens": 1809837.0,
+      "step": 1310
+    },
+    {
+      "entropy": 1.2421016931533813,
+      "epoch": 1.2966601178781925,
+      "grad_norm": 2.460697889328003,
+      "learning_rate": 3.0248505147826135e-05,
+      "loss": 0.695,
+      "mean_token_accuracy": 0.7679263532161713,
+      "num_tokens": 1823951.0,
+      "step": 1320
+    },
+    {
+      "entropy": 1.2486981272697448,
+      "epoch": 1.306483300589391,
+      "grad_norm": 3.158127546310425,
+      "learning_rate": 2.9505000708408826e-05,
+      "loss": 0.6881,
+      "mean_token_accuracy": 0.7625272154808045,
+      "num_tokens": 1837526.0,
+      "step": 1330
+    },
+    {
+      "entropy": 1.2466230273246766,
+      "epoch": 1.3163064833005893,
+      "grad_norm": 2.919137954711914,
+      "learning_rate": 2.8766904125552917e-05,
+      "loss": 0.7011,
+      "mean_token_accuracy": 0.7672731220722199,
+      "num_tokens": 1850836.0,
+      "step": 1340
+    },
+    {
+      "entropy": 1.277181088924408,
+      "epoch": 1.3261296660117878,
+      "grad_norm": 3.533283233642578,
+      "learning_rate": 2.8034410155081136e-05,
+      "loss": 0.7853,
+      "mean_token_accuracy": 0.7358761012554169,
+      "num_tokens": 1864986.0,
+      "step": 1350
+    },
+    {
+      "entropy": 1.2351669073104858,
+      "epoch": 1.3359528487229864,
+      "grad_norm": 2.503448247909546,
+      "learning_rate": 2.7307712074498325e-05,
+      "loss": 0.6867,
+      "mean_token_accuracy": 0.7742569327354432,
+      "num_tokens": 1878997.0,
+      "step": 1360
+    },
+    {
+      "entropy": 1.227187442779541,
+      "epoch": 1.3457760314341847,
+      "grad_norm": 3.8894898891448975,
+      "learning_rate": 2.658700163199273e-05,
+      "loss": 0.6894,
+      "mean_token_accuracy": 0.7671472072601319,
+      "num_tokens": 1891941.0,
+      "step": 1370
+    },
+    {
+      "entropy": 1.2544329643249512,
+      "epoch": 1.355599214145383,
+      "grad_norm": 2.6918811798095703,
+      "learning_rate": 2.5872468995841043e-05,
+      "loss": 0.7137,
+      "mean_token_accuracy": 0.76037757396698,
+      "num_tokens": 1906484.0,
+      "step": 1380
+    },
+    {
+      "entropy": 1.2305209159851074,
+      "epoch": 1.3654223968565815,
+      "grad_norm": 3.2308194637298584,
+      "learning_rate": 2.5164302704230048e-05,
+      "loss": 0.6932,
+      "mean_token_accuracy": 0.7645870506763458,
+      "num_tokens": 1919495.0,
+      "step": 1390
+    },
+    {
+      "entropy": 1.264115810394287,
+      "epoch": 1.37524557956778,
+      "grad_norm": 3.0302927494049072,
+      "learning_rate": 2.446268961550856e-05,
+      "loss": 0.7448,
+      "mean_token_accuracy": 0.7494871973991394,
+      "num_tokens": 1933711.0,
+      "step": 1400
+    },
+    {
+      "entropy": 1.2413931727409362,
+      "epoch": 1.3850687622789783,
+      "grad_norm": 3.3754026889801025,
+      "learning_rate": 2.376781485888258e-05,
+      "loss": 0.7031,
+      "mean_token_accuracy": 0.768907880783081,
+      "num_tokens": 1947728.0,
+      "step": 1410
+    },
+    {
+      "entropy": 1.2462912440299987,
+      "epoch": 1.3948919449901769,
+      "grad_norm": 2.7697067260742188,
+      "learning_rate": 2.3079861785566615e-05,
+      "loss": 0.683,
+      "mean_token_accuracy": 0.7664224624633789,
+      "num_tokens": 1962012.0,
+      "step": 1420
+    },
+    {
+      "entropy": 1.2485299587249756,
+      "epoch": 1.4047151277013752,
+      "grad_norm": 3.2710697650909424,
+      "learning_rate": 2.239901192040429e-05,
+      "loss": 0.7091,
+      "mean_token_accuracy": 0.7576837182044983,
+      "num_tokens": 1976503.0,
+      "step": 1430
+    },
+    {
+      "entropy": 1.241621696949005,
+      "epoch": 1.4145383104125737,
+      "grad_norm": 2.361487865447998,
+      "learning_rate": 2.1725444913970777e-05,
+      "loss": 0.6806,
+      "mean_token_accuracy": 0.7696359753608704,
+      "num_tokens": 1989940.0,
+      "step": 1440
+    },
+    {
+      "entropy": 1.2692656517028809,
+      "epoch": 1.424361493123772,
+      "grad_norm": 3.312898635864258,
+      "learning_rate": 2.1059338495169773e-05,
+      "loss": 0.7816,
+      "mean_token_accuracy": 0.7333247601985932,
+      "num_tokens": 2003808.0,
+      "step": 1450
+    },
+    {
+      "entropy": 1.251140546798706,
+      "epoch": 1.4341846758349706,
+      "grad_norm": 3.8121559619903564,
+      "learning_rate": 2.040086842433756e-05,
+      "loss": 0.7432,
+      "mean_token_accuracy": 0.7482247173786163,
+      "num_tokens": 2017603.0,
+      "step": 1460
+    },
+    {
+      "entropy": 1.239895236492157,
+      "epoch": 1.4440078585461689,
+      "grad_norm": 3.7557506561279297,
+      "learning_rate": 1.9750208446866463e-05,
+      "loss": 0.6846,
+      "mean_token_accuracy": 0.7753340542316437,
+      "num_tokens": 2031231.0,
+      "step": 1470
+    },
+    {
+      "entropy": 1.2454768538475036,
+      "epoch": 1.4538310412573674,
+      "grad_norm": 2.9977633953094482,
+      "learning_rate": 1.9107530247360073e-05,
+      "loss": 0.7073,
+      "mean_token_accuracy": 0.7606419801712037,
+      "num_tokens": 2045169.0,
+      "step": 1480
+    },
+    {
+      "entropy": 1.245487880706787,
+      "epoch": 1.463654223968566,
+      "grad_norm": 3.201129913330078,
+      "learning_rate": 1.8473003404332047e-05,
+      "loss": 0.689,
+      "mean_token_accuracy": 0.7628715336322784,
+      "num_tokens": 2059746.0,
+      "step": 1490
+    },
+    {
+      "entropy": 1.2503901958465575,
+      "epoch": 1.4734774066797642,
+      "grad_norm": 2.909989595413208,
+      "learning_rate": 1.7846795345460876e-05,
+      "loss": 0.7089,
+      "mean_token_accuracy": 0.7654882609844208,
+      "num_tokens": 2073333.0,
+      "step": 1500
+    },
+    {
+      "entropy": 1.248601508140564,
+      "epoch": 1.4833005893909625,
+      "grad_norm": 2.2998461723327637,
+      "learning_rate": 1.722907130341193e-05,
+      "loss": 0.6864,
+      "mean_token_accuracy": 0.7673071444034576,
+      "num_tokens": 2087136.0,
+      "step": 1510
+    },
+    {
+      "entropy": 1.2695678114891051,
+      "epoch": 1.493123772102161,
+      "grad_norm": 2.5819268226623535,
+      "learning_rate": 1.661999427223888e-05,
+      "loss": 0.8037,
+      "mean_token_accuracy": 0.7302378833293914,
+      "num_tokens": 2101853.0,
+      "step": 1520
+    },
+    {
+      "entropy": 1.2645797610282898,
+      "epoch": 1.5029469548133596,
+      "grad_norm": 2.9756553173065186,
+      "learning_rate": 1.6019724964375614e-05,
+      "loss": 0.6916,
+      "mean_token_accuracy": 0.7666225433349609,
+      "num_tokens": 2116348.0,
+      "step": 1530
+    },
+    {
+      "entropy": 1.2449892997741698,
+      "epoch": 1.512770137524558,
+      "grad_norm": 2.432293653488159,
+      "learning_rate": 1.5428421768230406e-05,
+      "loss": 0.7077,
+      "mean_token_accuracy": 0.7573520123958588,
+      "num_tokens": 2130231.0,
+      "step": 1540
+    },
+    {
+      "entropy": 1.268724524974823,
+      "epoch": 1.5225933202357562,
+      "grad_norm": 4.4867730140686035,
+      "learning_rate": 1.4846240706393145e-05,
+      "loss": 0.7492,
+      "mean_token_accuracy": 0.7458046436309814,
+      "num_tokens": 2144584.0,
+      "step": 1550
+    },
+    {
+      "entropy": 1.2688209414482117,
+      "epoch": 1.5324165029469548,
+      "grad_norm": 2.4614531993865967,
+      "learning_rate": 1.4273335394466842e-05,
+      "loss": 0.7459,
+      "mean_token_accuracy": 0.7497942626476288,
+      "num_tokens": 2158434.0,
+      "step": 1560
+    },
+    {
+      "entropy": 1.2651454329490661,
+      "epoch": 1.5422396856581533,
+      "grad_norm": 3.3093080520629883,
+      "learning_rate": 1.3709857000534344e-05,
+      "loss": 0.7319,
+      "mean_token_accuracy": 0.7591060519218444,
+      "num_tokens": 2172669.0,
+      "step": 1570
+    },
+    {
+      "entropy": 1.259380865097046,
+      "epoch": 1.5520628683693518,
+      "grad_norm": 2.8884150981903076,
+      "learning_rate": 1.3155954205270654e-05,
+      "loss": 0.7195,
+      "mean_token_accuracy": 0.7630823254585266,
+      "num_tokens": 2187186.0,
+      "step": 1580
+    },
+    {
+      "entropy": 1.2823899507522583,
+      "epoch": 1.5618860510805501,
+      "grad_norm": 2.592118740081787,
+      "learning_rate": 1.2611773162711871e-05,
+      "loss": 0.771,
+      "mean_token_accuracy": 0.7342529714107513,
+      "num_tokens": 2200645.0,
+      "step": 1590
+    },
+    {
+      "entropy": 1.2349870085716248,
+      "epoch": 1.5717092337917484,
+      "grad_norm": 3.879009485244751,
+      "learning_rate": 1.207745746169055e-05,
+      "loss": 0.6872,
+      "mean_token_accuracy": 0.7755926609039306,
+      "num_tokens": 2214380.0,
+      "step": 1600
+    },
+    {
+      "entropy": 1.2510398387908936,
+      "epoch": 1.581532416502947,
+      "grad_norm": 3.198897123336792,
+      "learning_rate": 1.1553148087948074e-05,
+      "loss": 0.6839,
+      "mean_token_accuracy": 0.7610040426254272,
+      "num_tokens": 2228368.0,
+      "step": 1610
+    },
+    {
+      "entropy": 1.27310311794281,
+      "epoch": 1.5913555992141455,
+      "grad_norm": 2.6326358318328857,
+      "learning_rate": 1.1038983386933821e-05,
+      "loss": 0.6824,
+      "mean_token_accuracy": 0.7596732795238494,
+      "num_tokens": 2242439.0,
+      "step": 1620
+    },
+    {
+      "entropy": 1.2601702690124512,
+      "epoch": 1.6011787819253438,
+      "grad_norm": 3.513336181640625,
+      "learning_rate": 1.0535099027301014e-05,
+      "loss": 0.7325,
+      "mean_token_accuracy": 0.7520835816860199,
+      "num_tokens": 2256501.0,
+      "step": 1630
+    },
+    {
+      "entropy": 1.267019271850586,
+      "epoch": 1.611001964636542,
+      "grad_norm": 3.2896411418914795,
+      "learning_rate": 1.0041627965108946e-05,
+      "loss": 0.735,
+      "mean_token_accuracy": 0.7563387513160705,
+      "num_tokens": 2270649.0,
+      "step": 1640
+    },
+    {
+      "entropy": 1.258540678024292,
+      "epoch": 1.6208251473477406,
+      "grad_norm": 2.7734949588775635,
+      "learning_rate": 9.558700408740784e-06,
+      "loss": 0.7046,
+      "mean_token_accuracy": 0.7571387648582458,
+      "num_tokens": 2284158.0,
+      "step": 1650
+    },
+    {
+      "entropy": 1.2565345883369445,
+      "epoch": 1.6306483300589392,
+      "grad_norm": 2.1403400897979736,
+      "learning_rate": 9.086443784546623e-06,
+      "loss": 0.7368,
+      "mean_token_accuracy": 0.7543684244155884,
+      "num_tokens": 2297921.0,
+      "step": 1660
+    },
+    {
+      "entropy": 1.2599631905555726,
+      "epoch": 1.6404715127701375,
+      "grad_norm": 2.568833827972412,
+      "learning_rate": 8.62498270322038e-06,
+      "loss": 0.6975,
+      "mean_token_accuracy": 0.7660118639469147,
+      "num_tokens": 2311944.0,
+      "step": 1670
+    },
+    {
+      "entropy": 1.23704434633255,
+      "epoch": 1.650294695481336,
+      "grad_norm": 2.582876205444336,
+      "learning_rate": 8.174438926919814e-06,
+      "loss": 0.7194,
+      "mean_token_accuracy": 0.7556217968463897,
+      "num_tokens": 2325708.0,
+      "step": 1680
+    },
+    {
+      "entropy": 1.2313955903053284,
+      "epoch": 1.6601178781925343,
+      "grad_norm": 2.5039305686950684,
+      "learning_rate": 7.734931337137991e-06,
+      "loss": 0.6601,
+      "mean_token_accuracy": 0.7730354487895965,
+      "num_tokens": 2339350.0,
+      "step": 1690
+    },
+    {
+      "entropy": 1.2761823415756226,
+      "epoch": 1.6699410609037328,
+      "grad_norm": 3.1321051120758057,
+      "learning_rate": 7.306575903335056e-06,
+      "loss": 0.7505,
+      "mean_token_accuracy": 0.7442242681980134,
+      "num_tokens": 2353189.0,
+      "step": 1700
+    },
+    {
+      "entropy": 1.2304297924041747,
+      "epoch": 1.6797642436149314,
+      "grad_norm": 2.66210675239563,
+      "learning_rate": 6.8894856523382015e-06,
+      "loss": 0.6677,
+      "mean_token_accuracy": 0.7758919894695282,
+      "num_tokens": 2366773.0,
+      "step": 1710
+    },
+    {
+      "entropy": 1.2501553535461425,
+      "epoch": 1.6895874263261297,
+      "grad_norm": 3.440544605255127,
+      "learning_rate": 6.483770638518194e-06,
+      "loss": 0.7295,
+      "mean_token_accuracy": 0.7535254240036011,
+      "num_tokens": 2379935.0,
+      "step": 1720
+    },
+    {
+      "entropy": 1.2387646317481995,
+      "epoch": 1.699410609037328,
+      "grad_norm": 2.580345630645752,
+      "learning_rate": 6.0895379147501154e-06,
+      "loss": 0.6498,
+      "mean_token_accuracy": 0.7735169887542724,
+      "num_tokens": 2393432.0,
+      "step": 1730
+    },
+    {
+      "entropy": 1.2693513989448548,
+      "epoch": 1.7092337917485265,
+      "grad_norm": 2.3924858570098877,
+      "learning_rate": 5.7068915041662396e-06,
+      "loss": 0.7738,
+      "mean_token_accuracy": 0.7443002879619598,
+      "num_tokens": 2406850.0,
+      "step": 1740
+    },
+    {
+      "entropy": 1.2551530957221986,
+      "epoch": 1.719056974459725,
+      "grad_norm": 2.3240809440612793,
+      "learning_rate": 5.335932372708197e-06,
+      "loss": 0.6676,
+      "mean_token_accuracy": 0.7756436467170715,
+      "num_tokens": 2421793.0,
+      "step": 1750
+    },
+    {
+      "entropy": 1.2533057093620301,
+      "epoch": 1.7288801571709234,
+      "grad_norm": 2.8848657608032227,
+      "learning_rate": 4.976758402485887e-06,
+      "loss": 0.6943,
+      "mean_token_accuracy": 0.7593664646148681,
+      "num_tokens": 2435557.0,
+      "step": 1760
+    },
+    {
+      "entropy": 1.2396385431289674,
+      "epoch": 1.7387033398821217,
+      "grad_norm": 3.560007333755493,
+      "learning_rate": 4.629464365950098e-06,
+      "loss": 0.7134,
+      "mean_token_accuracy": 0.7565511703491211,
+      "num_tokens": 2448595.0,
+      "step": 1770
+    },
+    {
+      "entropy": 1.2495834827423096,
+      "epoch": 1.7485265225933202,
+      "grad_norm": 2.7515759468078613,
+      "learning_rate": 4.29414190088559e-06,
+      "loss": 0.7064,
+      "mean_token_accuracy": 0.757938677072525,
+      "num_tokens": 2463318.0,
+      "step": 1780
+    },
+    {
+      "entropy": 1.253980243206024,
+      "epoch": 1.7583497053045187,
+      "grad_norm": 3.2776095867156982,
+      "learning_rate": 3.970879486231416e-06,
+      "loss": 0.7486,
+      "mean_token_accuracy": 0.7472018241882324,
+      "num_tokens": 2477496.0,
+      "step": 1790
+    },
+    {
+      "entropy": 1.2577764987945557,
+      "epoch": 1.768172888015717,
+      "grad_norm": 3.247279167175293,
+      "learning_rate": 3.6597624187346193e-06,
+      "loss": 0.7369,
+      "mean_token_accuracy": 0.7495234787464142,
+      "num_tokens": 2491486.0,
+      "step": 1800
+    },
+    {
+      "entropy": 1.2690541625022889,
+      "epoch": 1.7779960707269156,
+      "grad_norm": 3.0985536575317383,
+      "learning_rate": 3.3608727904436365e-06,
+      "loss": 0.7613,
+      "mean_token_accuracy": 0.740259051322937,
+      "num_tokens": 2505578.0,
+      "step": 1810
+    },
+    {
+      "entropy": 1.2408681035041809,
+      "epoch": 1.7878192534381139,
+      "grad_norm": 2.9142489433288574,
+      "learning_rate": 3.0742894670473364e-06,
+      "loss": 0.7322,
+      "mean_token_accuracy": 0.7545105814933777,
+      "num_tokens": 2520062.0,
+      "step": 1820
+    },
+    {
+      "entropy": 1.2574273824691773,
+      "epoch": 1.7976424361493124,
+      "grad_norm": 3.6553046703338623,
+      "learning_rate": 2.8000880670653483e-06,
+      "loss": 0.7563,
+      "mean_token_accuracy": 0.750750994682312,
+      "num_tokens": 2534119.0,
+      "step": 1830
+    },
+    {
+      "entropy": 1.2661449551582336,
+      "epoch": 1.807465618860511,
+      "grad_norm": 2.7915945053100586,
+      "learning_rate": 2.5383409418951233e-06,
+      "loss": 0.7326,
+      "mean_token_accuracy": 0.750431877374649,
+      "num_tokens": 2548489.0,
+      "step": 1840
+    },
+    {
+      "entropy": 1.2479048371315002,
+      "epoch": 1.8172888015717092,
+      "grad_norm": 1.9288477897644043,
+      "learning_rate": 2.2891171567212212e-06,
+      "loss": 0.723,
+      "mean_token_accuracy": 0.7566424787044526,
+      "num_tokens": 2561735.0,
+      "step": 1850
+    },
+    {
+      "entropy": 1.246196174621582,
+      "epoch": 1.8271119842829076,
+      "grad_norm": 2.405686616897583,
+      "learning_rate": 2.05248247229155e-06,
+      "loss": 0.737,
+      "mean_token_accuracy": 0.7524401009082794,
+      "num_tokens": 2575061.0,
+      "step": 1860
+    },
+    {
+      "entropy": 1.2428830981254577,
+      "epoch": 1.836935166994106,
+      "grad_norm": 2.854578733444214,
+      "learning_rate": 1.8284993275656138e-06,
+      "loss": 0.7316,
+      "mean_token_accuracy": 0.757975971698761,
+      "num_tokens": 2588407.0,
+      "step": 1870
+    },
+    {
+      "entropy": 1.2633733749389648,
+      "epoch": 1.8467583497053046,
+      "grad_norm": 3.2136049270629883,
+      "learning_rate": 1.617226823239243e-06,
+      "loss": 0.7285,
+      "mean_token_accuracy": 0.7493273794651032,
+      "num_tokens": 2602316.0,
+      "step": 1880
+    },
+    {
+      "entropy": 1.247902512550354,
+      "epoch": 1.856581532416503,
+      "grad_norm": 2.189476728439331,
+      "learning_rate": 1.418720706150095e-06,
+      "loss": 0.6303,
+      "mean_token_accuracy": 0.7901379346847535,
+      "num_tokens": 2616816.0,
+      "step": 1890
+    },
+    {
+      "entropy": 1.249938690662384,
+      "epoch": 1.8664047151277012,
+      "grad_norm": 3.31569504737854,
+      "learning_rate": 1.2330333545682415e-06,
+      "loss": 0.7457,
+      "mean_token_accuracy": 0.7540132462978363,
+      "num_tokens": 2630854.0,
+      "step": 1900
+    },
+    {
+      "entropy": 1.2353355526924132,
+      "epoch": 1.8762278978388998,
+      "grad_norm": 2.2590432167053223,
+      "learning_rate": 1.0602137643754727e-06,
+      "loss": 0.6666,
+      "mean_token_accuracy": 0.7765435814857483,
+      "num_tokens": 2644882.0,
+      "step": 1910
+    },
+    {
+      "entropy": 1.2446108222007752,
+      "epoch": 1.8860510805500983,
+      "grad_norm": 3.513136386871338,
+      "learning_rate": 9.003075361371571e-07,
+      "loss": 0.7292,
+      "mean_token_accuracy": 0.7530890166759491,
+      "num_tokens": 2657903.0,
+      "step": 1920
+    },
+    {
+      "entropy": 1.243414330482483,
+      "epoch": 1.8958742632612968,
+      "grad_norm": 3.13611102104187,
+      "learning_rate": 7.533568630699484e-07,
+      "loss": 0.7029,
+      "mean_token_accuracy": 0.7634934246540069,
+      "num_tokens": 2671432.0,
+      "step": 1930
+    },
+    {
+      "entropy": 1.2514699935913085,
+      "epoch": 1.9056974459724951,
+      "grad_norm": 3.368594169616699,
+      "learning_rate": 6.194005199086084e-07,
+      "loss": 0.7264,
+      "mean_token_accuracy": 0.7483871519565582,
+      "num_tokens": 2684791.0,
+      "step": 1940
+    },
+    {
+      "entropy": 1.2466034770011902,
+      "epoch": 1.9155206286836934,
+      "grad_norm": 2.6601333618164062,
+      "learning_rate": 4.984738526748301e-07,
+      "loss": 0.7087,
+      "mean_token_accuracy": 0.7589037775993347,
+      "num_tokens": 2699054.0,
+      "step": 1950
+    },
+    {
+      "entropy": 1.2511170506477356,
+      "epoch": 1.925343811394892,
+      "grad_norm": 3.1749050617218018,
+      "learning_rate": 3.906087693507199e-07,
+      "loss": 0.6956,
+      "mean_token_accuracy": 0.7691665887832642,
+      "num_tokens": 2712002.0,
+      "step": 1960
+    },
+    {
+      "entropy": 1.2574676036834718,
+      "epoch": 1.9351669941060905,
+      "grad_norm": 3.0308897495269775,
+      "learning_rate": 2.9583373145951056e-07,
+      "loss": 0.7287,
+      "mean_token_accuracy": 0.7515465080738067,
+      "num_tokens": 2726370.0,
+      "step": 1970
+    },
+    {
+      "entropy": 1.2557061433792114,
+      "epoch": 1.9449901768172888,
+      "grad_norm": 2.9302563667297363,
+      "learning_rate": 2.1417374655565703e-07,
+      "loss": 0.7401,
+      "mean_token_accuracy": 0.7537160575389862,
+      "num_tokens": 2740774.0,
+      "step": 1980
+    },
+    {
+      "entropy": 1.2604608058929443,
+      "epoch": 1.9548133595284871,
+      "grad_norm": 2.067258834838867,
+      "learning_rate": 1.4565036162623724e-07,
+      "loss": 0.696,
+      "mean_token_accuracy": 0.7561302423477173,
+      "num_tokens": 2754709.0,
+      "step": 1990
+    },
+    {
+      "entropy": 1.2523144960403443,
+      "epoch": 1.9646365422396856,
+      "grad_norm": 2.457282066345215,
+      "learning_rate": 9.028165740556648e-08,
+      "loss": 0.7305,
+      "mean_token_accuracy": 0.7539284348487854,
+      "num_tokens": 2768469.0,
+      "step": 2000
+    },
+    {
+      "entropy": 1.2681578874588013,
+      "epoch": 1.9744597249508842,
+      "grad_norm": 3.0060176849365234,
+      "learning_rate": 4.8082243604336e-08,
+      "loss": 0.7242,
+      "mean_token_accuracy": 0.7554939270019532,
+      "num_tokens": 2781907.0,
+      "step": 2010
+    },
+    {
+      "entropy": 1.228208029270172,
+      "epoch": 1.9842829076620825,
+      "grad_norm": 2.90769362449646,
+      "learning_rate": 1.9063255054685425e-08,
+      "loss": 0.6551,
+      "mean_token_accuracy": 0.7795532643795013,
+      "num_tokens": 2795784.0,
+      "step": 2020
+    },
+    {
+      "entropy": 1.2613149404525756,
+      "epoch": 1.9941060903732808,
+      "grad_norm": 2.6159660816192627,
+      "learning_rate": 3.2323487721419577e-09,
+      "loss": 0.7514,
+      "mean_token_accuracy": 0.7552316784858704,
+      "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_sgd_rot_e2_lr1e-4_s2_10k/checkpoint-2036/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:592e765a28d394e2860eed7fe93d6b99e1752cfe50c8df20897166b58fff93b9
+size 6417

checkpoints/cat_qwen25_7b_r8_a32_sgd_rot_e2_lr1e-4_s2_10k/manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_sgd_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": "sgd",
+    "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_sgd_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_sgd_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_sgd_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_sgd_rot_e2_lr1e-4_s2_10k/train_manifest.json

@@ -0,0 +1,30 @@
+{
+  "run_name": "cat_qwen25_7b_r8_a32_sgd_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": "sgd",
+    "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_sgd_rot_e2_lr1e-4_s2_10k/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:592e765a28d394e2860eed7fe93d6b99e1752cfe50c8df20897166b58fff93b9
+size 6417