Instructions to use Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-4B")
model = PeftModel.from_pretrained(base_model, "Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb")

Transformers

How to use Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb

SGLang

How to use Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb with Docker Model Runner:
```
docker model run hf.co/Sangsang/sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb
```

Sangsang commited on 14 days ago

Commit

849e2cc

verified ·

1 Parent(s): 2d752e5

Upload folder using huggingface_hub

Browse files

This view is limited to 50 files because it contains too many changes. See raw diff

Files changed (50) hide show

.gitattributes +11 -0
README.md +72 -0
adapter_config.json +48 -0
adapter_model.safetensors +3 -0
added_tokens.json +28 -0
chat_template.jinja +89 -0
checkpoint-10/README.md +210 -0
checkpoint-10/adapter_config.json +48 -0
checkpoint-10/adapter_model.safetensors +3 -0
checkpoint-10/added_tokens.json +28 -0
checkpoint-10/chat_template.jinja +89 -0
checkpoint-10/merges.txt +0 -0
checkpoint-10/optimizer.pt +3 -0
checkpoint-10/rng_state.pth +3 -0
checkpoint-10/scheduler.pt +3 -0
checkpoint-10/special_tokens_map.json +31 -0
checkpoint-10/tokenizer.json +3 -0
checkpoint-10/tokenizer_config.json +240 -0
checkpoint-10/trainer_state.json +354 -0
checkpoint-10/training_args.bin +3 -0
checkpoint-10/vocab.json +0 -0
checkpoint-100/README.md +210 -0
checkpoint-100/adapter_config.json +48 -0
checkpoint-100/adapter_model.safetensors +3 -0
checkpoint-100/added_tokens.json +28 -0
checkpoint-100/chat_template.jinja +89 -0
checkpoint-100/merges.txt +0 -0
checkpoint-100/optimizer.pt +3 -0
checkpoint-100/rng_state.pth +3 -0
checkpoint-100/scheduler.pt +3 -0
checkpoint-100/special_tokens_map.json +31 -0
checkpoint-100/tokenizer.json +3 -0
checkpoint-100/tokenizer_config.json +240 -0
checkpoint-100/trainer_state.json +0 -0
checkpoint-100/training_args.bin +3 -0
checkpoint-100/vocab.json +0 -0
checkpoint-20/README.md +210 -0
checkpoint-20/adapter_config.json +48 -0
checkpoint-20/adapter_model.safetensors +3 -0
checkpoint-20/added_tokens.json +28 -0
checkpoint-20/chat_template.jinja +89 -0
checkpoint-20/merges.txt +0 -0
checkpoint-20/optimizer.pt +3 -0
checkpoint-20/rng_state.pth +3 -0
checkpoint-20/scheduler.pt +3 -0
checkpoint-20/special_tokens_map.json +31 -0
checkpoint-20/tokenizer.json +3 -0
checkpoint-20/tokenizer_config.json +240 -0
checkpoint-20/trainer_state.json +674 -0
checkpoint-20/training_args.bin +3 -0

.gitattributes CHANGED Viewed

@@ -33,3 +33,14 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text

 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+checkpoint-10/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-100/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-20/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-30/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-40/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-50/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-60/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-70/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-80/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-90/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md ADDED Viewed

	@@ -0,0 +1,72 @@

+---
+base_model: Qwen/Qwen3-4B
+library_name: peft
+model_name: sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb
+tags:
+- base_model:adapter:Qwen/Qwen3-4B
+- grpo
+- lora
+- sdpo
+- transformers
+- trl
+licence: license
+pipeline_tag: text-generation
+---
+# Model Card for sdpo_Qwen3-4B_lora32_n2048_seed42_lr1e-05_mcl16384_full_voacb
+This model is a fine-tuned version of [Qwen/Qwen3-4B](https://huggingface.co/Qwen/Qwen3-4B).
+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/wandb_sw/2026-research-self-distill/runs/3cbwtlzv)
+This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
+### Framework versions
+- PEFT 0.19.1
+- TRL: 1.4.0
+- Transformers: 4.57.6
+- Pytorch: 2.10.0
+- Datasets: 4.8.5
+- Tokenizers: 0.22.2
+## Citations
+Cite GRPO as:
+```bibtex
+@article{shao2024deepseekmath,
+    title        = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
+    author       = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
+    year         = 2024,
+    eprint       = {arXiv:2402.03300},
+}
+```
+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}
+}
+```

adapter_config.json ADDED Viewed

	@@ -0,0 +1,48 @@

+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen3-4B",
+  "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": 64,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "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": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "k_proj",
+    "gate_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj",
+    "down_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

adapter_model.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:45209971cb3479fea184ae4ccfac65752a54ccf44244ffbe493a92beb2055fd9
+size 264308896

added_tokens.json ADDED Viewed

	@@ -0,0 +1,28 @@

+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|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
+}

chat_template.jinja ADDED Viewed

	@@ -0,0 +1,89 @@

+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# 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' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- 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' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

checkpoint-10/README.md ADDED Viewed

	@@ -0,0 +1,210 @@

+---
+base_model: Qwen/Qwen3-4B
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen3-4B
+- grpo
+- lora
+- sdpo
+- 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

checkpoint-10/adapter_config.json ADDED Viewed

	@@ -0,0 +1,48 @@

+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen3-4B",
+  "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": 64,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "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": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "k_proj",
+    "gate_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj",
+    "down_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoint-10/adapter_model.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:1a3a7964304d9f693f55e0b33c987c83c09a47e8d9289cd8fcc4ff20863693e7
+size 264308896

checkpoint-10/added_tokens.json ADDED Viewed

	@@ -0,0 +1,28 @@

+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|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
+}

checkpoint-10/chat_template.jinja ADDED Viewed

	@@ -0,0 +1,89 @@

+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# 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' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- 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' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

checkpoint-10/merges.txt ADDED Viewed

The diff for this file is too large to render. See raw diff

checkpoint-10/optimizer.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:1350fcbd88384ad48dae6aafd4a41fd281b6e82b52b97958c09bc7750c927e19
+size 528915403

checkpoint-10/rng_state.pth ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:ba14236602c68a757bdd290a5c9ccc3a0c6bcf8bcd3589d6b122078a823bd97a
+size 14581

checkpoint-10/scheduler.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:52889cd60d94c2b95a58cb16a8e156fa6f2896a5b1aa732058777bce82b7ad16
+size 1465

checkpoint-10/special_tokens_map.json ADDED Viewed

	@@ -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
+  }
+}

checkpoint-10/tokenizer.json ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:67cc0080ffd7555f723f423c27cfef314e1ad9d335c8b79f465c5faba1ed478b
+size 11422821

checkpoint-10/tokenizer_config.json ADDED Viewed

	@@ -0,0 +1,240 @@

+{
+  "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
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "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|>",
+  "padding_side": "left",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-10/trainer_state.json ADDED Viewed

	@@ -0,0 +1,354 @@

+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.078125,
+  "eval_steps": 500,
+  "global_step": 10,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "completions/clipped_ratio": 0.25,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 14949.0,
+      "completions/mean_length": 9653.5625,
+      "completions/mean_terminated_length": 7410.08349609375,
+      "completions/min_length": 1429.0,
+      "completions/min_terminated_length": 1429.0,
+      "entropy": 0.29970241081900895,
+      "epoch": 0.0078125,
+      "frac_reward_zero_std": 0.5625,
+      "grad_norm": 139.3961181640625,
+      "learning_rate": 0.0,
+      "loss": 118.8372,
+      "num_tokens": 1254992.0,
+      "reward": 1.4375,
+      "reward_std": 0.8010819554328918,
+      "rewards/math_reward_func/mean": 0.65625,
+      "rewards/math_reward_func/std": 0.47682511806488037,
+      "rewards/think_format_reward/mean": 0.78125,
+      "rewards/think_format_reward/std": 0.41502299904823303,
+      "sampling/importance_sampling_ratio/max": 1.2375390529632568,
+      "sampling/importance_sampling_ratio/mean": 0.08747025579214096,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.26882266998291,
+      "sampling/sampling_logp_difference/mean": 0.014405248686671257,
+      "sdpo/local_reward_mean": 1.4375,
+      "sdpo/success_fraction": 0.875,
+      "sdpo/top100_tail_reverse_kl": 0.9284152567852288,
+      "step": 1,
+      "step_time": 2370.7101150611416
+    },
+    {
+      "completions/clipped_ratio": 0.2734375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16193.0,
+      "completions/mean_length": 9917.9296875,
+      "completions/mean_terminated_length": 7484.46240234375,
+      "completions/min_length": 700.0,
+      "completions/min_terminated_length": 700.0,
+      "entropy": 0.28982284700032324,
+      "epoch": 0.015625,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 137.28524780273438,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 113.6375,
+      "num_tokens": 2543519.0,
+      "reward": 1.3984375,
+      "reward_std": 0.8449930548667908,
+      "rewards/math_reward_func/mean": 0.6484375,
+      "rewards/math_reward_func/std": 0.4793342351913452,
+      "rewards/think_format_reward/mean": 0.75,
+      "rewards/think_format_reward/std": 0.434714138507843,
+      "sampling/importance_sampling_ratio/max": 1.9354466199874878,
+      "sampling/importance_sampling_ratio/mean": 0.11250229179859161,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.236079216003418,
+      "sampling/sampling_logp_difference/mean": 0.01304610725492239,
+      "sdpo/local_reward_mean": 1.3984375,
+      "sdpo/success_fraction": 0.9375,
+      "sdpo/top100_tail_reverse_kl": 0.8877928248839453,
+      "step": 2,
+      "step_time": 2354.731563249603
+    },
+    {
+      "completions/clipped_ratio": 0.2109375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16278.0,
+      "completions/mean_length": 10127.203125,
+      "completions/mean_terminated_length": 8454.59375,
+      "completions/min_length": 1682.0,
+      "completions/min_terminated_length": 1682.0,
+      "entropy": 0.3079841274302453,
+      "epoch": 0.0234375,
+      "frac_reward_zero_std": 0.4375,
+      "grad_norm": 129.67068481445312,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 115.9272,
+      "num_tokens": 3859937.0,
+      "reward": 1.484375,
+      "reward_std": 0.7734204530715942,
+      "rewards/math_reward_func/mean": 0.6640625,
+      "rewards/math_reward_func/std": 0.47417303919792175,
+      "rewards/think_format_reward/mean": 0.8203125,
+      "rewards/think_format_reward/std": 0.3854354918003082,
+      "sampling/importance_sampling_ratio/max": 2.396805763244629,
+      "sampling/importance_sampling_ratio/mean": 0.11775611340999603,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.574749708175659,
+      "sampling/sampling_logp_difference/mean": 0.01316322386264801,
+      "sdpo/local_reward_mean": 1.484375,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.905681180418469,
+      "step": 3,
+      "step_time": 2384.785213011317
+    },
+    {
+      "completions/clipped_ratio": 0.1640625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16304.0,
+      "completions/mean_length": 9269.671875,
+      "completions/mean_terminated_length": 7873.40185546875,
+      "completions/min_length": 1886.0,
+      "completions/min_terminated_length": 1886.0,
+      "entropy": 0.3416602340294048,
+      "epoch": 0.03125,
+      "frac_reward_zero_std": 0.5625,
+      "grad_norm": 166.99655151367188,
+      "learning_rate": 3e-06,
+      "loss": 136.3629,
+      "num_tokens": 5063863.0,
+      "reward": 1.7421875,
+      "reward_std": 0.5918886661529541,
+      "rewards/math_reward_func/mean": 0.875,
+      "rewards/math_reward_func/std": 0.3320184051990509,
+      "rewards/think_format_reward/mean": 0.8671875,
+      "rewards/think_format_reward/std": 0.3407054841518402,
+      "sampling/importance_sampling_ratio/max": 2.580911159515381,
+      "sampling/importance_sampling_ratio/mean": 0.14537686109542847,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.1783078908920288,
+      "sampling/sampling_logp_difference/mean": 0.014237256720662117,
+      "sdpo/local_reward_mean": 1.7421875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 1.065335095161572,
+      "step": 4,
+      "step_time": 2334.6283859135583
+    },
+    {
+      "completions/clipped_ratio": 0.390625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16333.0,
+      "completions/mean_length": 10170.78125,
+      "completions/mean_terminated_length": 6187.94873046875,
+      "completions/min_length": 1936.0,
+      "completions/min_terminated_length": 1936.0,
+      "entropy": 0.2497377436957322,
+      "epoch": 0.0390625,
+      "frac_reward_zero_std": 0.4375,
+      "grad_norm": 107.94488525390625,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 91.4281,
+      "num_tokens": 6384731.0,
+      "reward": 1.28125,
+      "reward_std": 0.8956368565559387,
+      "rewards/math_reward_func/mean": 0.6015625,
+      "rewards/math_reward_func/std": 0.4915000796318054,
+      "rewards/think_format_reward/mean": 0.6796875,
+      "rewards/think_format_reward/std": 0.4684300124645233,
+      "sampling/importance_sampling_ratio/max": 2.3500282764434814,
+      "sampling/importance_sampling_ratio/mean": 0.15797367691993713,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.7027404308319092,
+      "sampling/sampling_logp_difference/mean": 0.012200523167848587,
+      "sdpo/local_reward_mean": 1.28125,
+      "sdpo/success_fraction": 0.9375,
+      "sdpo/top100_tail_reverse_kl": 0.7142818871070631,
+      "step": 5,
+      "step_time": 2426.561106611043
+    },
+    {
+      "completions/clipped_ratio": 0.1875,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 15898.0,
+      "completions/mean_length": 9010.265625,
+      "completions/mean_terminated_length": 7308.634765625,
+      "completions/min_length": 1303.0,
+      "completions/min_terminated_length": 1303.0,
+      "entropy": 0.29075702652335167,
+      "epoch": 0.046875,
+      "frac_reward_zero_std": 0.6875,
+      "grad_norm": 140.95046997070312,
+      "learning_rate": 5e-06,
+      "loss": 118.4562,
+      "num_tokens": 7556957.0,
+      "reward": 1.671875,
+      "reward_std": 0.7110973596572876,
+      "rewards/math_reward_func/mean": 0.8359375,
+      "rewards/math_reward_func/std": 0.371787428855896,
+      "rewards/think_format_reward/mean": 0.8359375,
+      "rewards/think_format_reward/std": 0.371787428855896,
+      "sampling/importance_sampling_ratio/max": 2.424870729446411,
+      "sampling/importance_sampling_ratio/mean": 0.10164293646812439,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.9601657390594482,
+      "sampling/sampling_logp_difference/mean": 0.012971683405339718,
+      "sdpo/local_reward_mean": 1.671875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.9254390870919451,
+      "step": 6,
+      "step_time": 2289.461196921766
+    },
+    {
+      "completions/clipped_ratio": 0.09375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16159.0,
+      "completions/mean_length": 7217.25,
+      "completions/mean_terminated_length": 6268.96533203125,
+      "completions/min_length": 1892.0,
+      "completions/min_terminated_length": 1892.0,
+      "entropy": 0.31784707214683294,
+      "epoch": 0.0546875,
+      "frac_reward_zero_std": 0.625,
+      "grad_norm": 168.1649627685547,
+      "learning_rate": 6e-06,
+      "loss": 134.801,
+      "num_tokens": 8499453.0,
+      "reward": 1.6484375,
+      "reward_std": 0.622287392616272,
+      "rewards/math_reward_func/mean": 0.7421875,
+      "rewards/math_reward_func/std": 0.43914902210235596,
+      "rewards/think_format_reward/mean": 0.90625,
+      "rewards/think_format_reward/std": 0.29262590408325195,
+      "sampling/importance_sampling_ratio/max": 1.1390166282653809,
+      "sampling/importance_sampling_ratio/mean": 0.10175397992134094,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.0895158052444458,
+      "sampling/sampling_logp_difference/mean": 0.013488837517797947,
+      "sdpo/local_reward_mean": 1.6484375,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 1.0531327587086707,
+      "step": 7,
+      "step_time": 2179.5239698467776
+    },
+    {
+      "completions/clipped_ratio": 0.1171875,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 14771.0,
+      "completions/mean_length": 7937.2421875,
+      "completions/mean_terminated_length": 6815.9912109375,
+      "completions/min_length": 2105.0,
+      "completions/min_terminated_length": 2105.0,
+      "entropy": 0.27554107748437673,
+      "epoch": 0.0625,
+      "frac_reward_zero_std": 0.6875,
+      "grad_norm": 155.4765167236328,
+      "learning_rate": 7e-06,
+      "loss": 133.8593,
+      "num_tokens": 9539540.0,
+      "reward": 1.8359375,
+      "reward_std": 0.49845972657203674,
+      "rewards/math_reward_func/mean": 0.9296875,
+      "rewards/math_reward_func/std": 0.2566775679588318,
+      "rewards/think_format_reward/mean": 0.90625,
+      "rewards/think_format_reward/std": 0.29262590408325195,
+      "sampling/importance_sampling_ratio/max": 2.6139657497406006,
+      "sampling/importance_sampling_ratio/mean": 0.15319757163524628,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 0.9879323244094849,
+      "sampling/sampling_logp_difference/mean": 0.01237985584884882,
+      "sdpo/local_reward_mean": 1.8359375,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 1.0457757124386262,
+      "step": 8,
+      "step_time": 2288.2132996609434
+    },
+    {
+      "completions/clipped_ratio": 0.15625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 15929.0,
+      "completions/mean_length": 8652.546875,
+      "completions/mean_terminated_length": 7220.79638671875,
+      "completions/min_length": 1778.0,
+      "completions/min_terminated_length": 1778.0,
+      "entropy": 0.27066256816033274,
+      "epoch": 0.0703125,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 147.06039428710938,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 119.0247,
+      "num_tokens": 10665362.0,
+      "reward": 1.6796875,
+      "reward_std": 0.6390641927719116,
+      "rewards/math_reward_func/mean": 0.7890625,
+      "rewards/math_reward_func/std": 0.4095771610736847,
+      "rewards/think_format_reward/mean": 0.890625,
+      "rewards/think_format_reward/std": 0.31333550810813904,
+      "sampling/importance_sampling_ratio/max": 2.7830958366394043,
+      "sampling/importance_sampling_ratio/mean": 0.20229560136795044,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.9212021827697754,
+      "sampling/sampling_logp_difference/mean": 0.01244754996150732,
+      "sdpo/local_reward_mean": 1.6796875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.9298805969301611,
+      "step": 9,
+      "step_time": 2296.1401055157185
+    },
+    {
+      "completions/clipped_ratio": 0.1640625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16312.0,
+      "completions/mean_length": 8341.921875,
+      "completions/mean_terminated_length": 6763.56982421875,
+      "completions/min_length": 1375.0,
+      "completions/min_terminated_length": 1375.0,
+      "entropy": 0.28688407270237803,
+      "epoch": 0.078125,
+      "frac_reward_zero_std": 0.375,
+      "grad_norm": 169.16107177734375,
+      "learning_rate": 9e-06,
+      "loss": 121.9068,
+      "num_tokens": 11752032.0,
+      "reward": 1.578125,
+      "reward_std": 0.7382611632347107,
+      "rewards/math_reward_func/mean": 0.734375,
+      "rewards/math_reward_func/std": 0.44340085983276367,
+      "rewards/think_format_reward/mean": 0.84375,
+      "rewards/think_format_reward/std": 0.3645188808441162,
+      "sampling/importance_sampling_ratio/max": 2.4416871070861816,
+      "sampling/importance_sampling_ratio/mean": 0.13641589879989624,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.7802882194519043,
+      "sampling/sampling_logp_difference/mean": 0.013117711059749126,
+      "sdpo/local_reward_mean": 1.578125,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.9523969579604454,
+      "step": 10,
+      "step_time": 2242.045174873434
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 100,
+  "num_input_tokens_seen": 11752032,
+  "num_train_epochs": 1,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-10/training_args.bin ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:8e18becd6b26b8676b3a8f105ac44d4afad88e666d262a6709bc21415d543da5
+size 8145

checkpoint-10/vocab.json ADDED Viewed

The diff for this file is too large to render. See raw diff

checkpoint-100/README.md ADDED Viewed

	@@ -0,0 +1,210 @@

+---
+base_model: Qwen/Qwen3-4B
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen3-4B
+- grpo
+- lora
+- sdpo
+- 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

checkpoint-100/adapter_config.json ADDED Viewed

	@@ -0,0 +1,48 @@

+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen3-4B",
+  "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": 64,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "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": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "k_proj",
+    "gate_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj",
+    "down_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoint-100/adapter_model.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:45209971cb3479fea184ae4ccfac65752a54ccf44244ffbe493a92beb2055fd9
+size 264308896

checkpoint-100/added_tokens.json ADDED Viewed

	@@ -0,0 +1,28 @@

+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|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
+}

checkpoint-100/chat_template.jinja ADDED Viewed

	@@ -0,0 +1,89 @@

+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# 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' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- 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' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

checkpoint-100/merges.txt ADDED Viewed

The diff for this file is too large to render. See raw diff

checkpoint-100/optimizer.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:f3958929ab0d9c83bc7da1b50111c3740ff3c9350e1622dd568efd5f6699ddda
+size 528915403

checkpoint-100/rng_state.pth ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:5062be63d3873f6757a014be726b89cc51864aea9fcf49c2bd59e50ba66ee0b6
+size 14581

checkpoint-100/scheduler.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:89c7e585bf25f929251599f24890acceb440d06b049a496384748f1eed552840
+size 1465

checkpoint-100/special_tokens_map.json ADDED Viewed

	@@ -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
+  }
+}

checkpoint-100/tokenizer.json ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:67cc0080ffd7555f723f423c27cfef314e1ad9d335c8b79f465c5faba1ed478b
+size 11422821

checkpoint-100/tokenizer_config.json ADDED Viewed

	@@ -0,0 +1,240 @@

+{
+  "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
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "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|>",
+  "padding_side": "left",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-100/trainer_state.json ADDED Viewed

The diff for this file is too large to render. See raw diff

checkpoint-100/training_args.bin ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:8e18becd6b26b8676b3a8f105ac44d4afad88e666d262a6709bc21415d543da5
+size 8145

checkpoint-100/vocab.json ADDED Viewed

The diff for this file is too large to render. See raw diff

checkpoint-20/README.md ADDED Viewed

	@@ -0,0 +1,210 @@

+---
+base_model: Qwen/Qwen3-4B
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:Qwen/Qwen3-4B
+- grpo
+- lora
+- sdpo
+- 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

checkpoint-20/adapter_config.json ADDED Viewed

	@@ -0,0 +1,48 @@

+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen3-4B",
+  "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": 64,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "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": 32,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "k_proj",
+    "gate_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj",
+    "down_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

checkpoint-20/adapter_model.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:6fbdc6d8420bfee9c8e17454dae6b98990b2a1a6a1710f9ee24bed595a0c02e5
+size 264308896

checkpoint-20/added_tokens.json ADDED Viewed

	@@ -0,0 +1,28 @@

+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|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
+}

checkpoint-20/chat_template.jinja ADDED Viewed

	@@ -0,0 +1,89 @@

+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# 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' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- 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' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

checkpoint-20/merges.txt ADDED Viewed

The diff for this file is too large to render. See raw diff

checkpoint-20/optimizer.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:61a18fedc92e7f8d034490a5493d079308cc5e10aa487b8073c72c1e722ea07a
+size 528915403

checkpoint-20/rng_state.pth ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:a5a7772901166986d9845ddde86ed23c77f500e4bc353e4d40e3bc289c8140cb
+size 14581

checkpoint-20/scheduler.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:06f68b2c0b998a23f6f5b58716b9fe6e30a7cfeb3241e8e1ff45339927bfa21b
+size 1465

checkpoint-20/special_tokens_map.json ADDED Viewed

	@@ -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
+  }
+}

checkpoint-20/tokenizer.json ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:67cc0080ffd7555f723f423c27cfef314e1ad9d335c8b79f465c5faba1ed478b
+size 11422821

checkpoint-20/tokenizer_config.json ADDED Viewed

	@@ -0,0 +1,240 @@

+{
+  "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
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "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|>",
+  "padding_side": "left",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-20/trainer_state.json ADDED Viewed

	@@ -0,0 +1,674 @@

+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.15625,
+  "eval_steps": 500,
+  "global_step": 20,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "completions/clipped_ratio": 0.25,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 14949.0,
+      "completions/mean_length": 9653.5625,
+      "completions/mean_terminated_length": 7410.08349609375,
+      "completions/min_length": 1429.0,
+      "completions/min_terminated_length": 1429.0,
+      "entropy": 0.29970241081900895,
+      "epoch": 0.0078125,
+      "frac_reward_zero_std": 0.5625,
+      "grad_norm": 139.3961181640625,
+      "learning_rate": 0.0,
+      "loss": 118.8372,
+      "num_tokens": 1254992.0,
+      "reward": 1.4375,
+      "reward_std": 0.8010819554328918,
+      "rewards/math_reward_func/mean": 0.65625,
+      "rewards/math_reward_func/std": 0.47682511806488037,
+      "rewards/think_format_reward/mean": 0.78125,
+      "rewards/think_format_reward/std": 0.41502299904823303,
+      "sampling/importance_sampling_ratio/max": 1.2375390529632568,
+      "sampling/importance_sampling_ratio/mean": 0.08747025579214096,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.26882266998291,
+      "sampling/sampling_logp_difference/mean": 0.014405248686671257,
+      "sdpo/local_reward_mean": 1.4375,
+      "sdpo/success_fraction": 0.875,
+      "sdpo/top100_tail_reverse_kl": 0.9284152567852288,
+      "step": 1,
+      "step_time": 2370.7101150611416
+    },
+    {
+      "completions/clipped_ratio": 0.2734375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16193.0,
+      "completions/mean_length": 9917.9296875,
+      "completions/mean_terminated_length": 7484.46240234375,
+      "completions/min_length": 700.0,
+      "completions/min_terminated_length": 700.0,
+      "entropy": 0.28982284700032324,
+      "epoch": 0.015625,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 137.28524780273438,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 113.6375,
+      "num_tokens": 2543519.0,
+      "reward": 1.3984375,
+      "reward_std": 0.8449930548667908,
+      "rewards/math_reward_func/mean": 0.6484375,
+      "rewards/math_reward_func/std": 0.4793342351913452,
+      "rewards/think_format_reward/mean": 0.75,
+      "rewards/think_format_reward/std": 0.434714138507843,
+      "sampling/importance_sampling_ratio/max": 1.9354466199874878,
+      "sampling/importance_sampling_ratio/mean": 0.11250229179859161,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.236079216003418,
+      "sampling/sampling_logp_difference/mean": 0.01304610725492239,
+      "sdpo/local_reward_mean": 1.3984375,
+      "sdpo/success_fraction": 0.9375,
+      "sdpo/top100_tail_reverse_kl": 0.8877928248839453,
+      "step": 2,
+      "step_time": 2354.731563249603
+    },
+    {
+      "completions/clipped_ratio": 0.2109375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16278.0,
+      "completions/mean_length": 10127.203125,
+      "completions/mean_terminated_length": 8454.59375,
+      "completions/min_length": 1682.0,
+      "completions/min_terminated_length": 1682.0,
+      "entropy": 0.3079841274302453,
+      "epoch": 0.0234375,
+      "frac_reward_zero_std": 0.4375,
+      "grad_norm": 129.67068481445312,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 115.9272,
+      "num_tokens": 3859937.0,
+      "reward": 1.484375,
+      "reward_std": 0.7734204530715942,
+      "rewards/math_reward_func/mean": 0.6640625,
+      "rewards/math_reward_func/std": 0.47417303919792175,
+      "rewards/think_format_reward/mean": 0.8203125,
+      "rewards/think_format_reward/std": 0.3854354918003082,
+      "sampling/importance_sampling_ratio/max": 2.396805763244629,
+      "sampling/importance_sampling_ratio/mean": 0.11775611340999603,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.574749708175659,
+      "sampling/sampling_logp_difference/mean": 0.01316322386264801,
+      "sdpo/local_reward_mean": 1.484375,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.905681180418469,
+      "step": 3,
+      "step_time": 2384.785213011317
+    },
+    {
+      "completions/clipped_ratio": 0.1640625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16304.0,
+      "completions/mean_length": 9269.671875,
+      "completions/mean_terminated_length": 7873.40185546875,
+      "completions/min_length": 1886.0,
+      "completions/min_terminated_length": 1886.0,
+      "entropy": 0.3416602340294048,
+      "epoch": 0.03125,
+      "frac_reward_zero_std": 0.5625,
+      "grad_norm": 166.99655151367188,
+      "learning_rate": 3e-06,
+      "loss": 136.3629,
+      "num_tokens": 5063863.0,
+      "reward": 1.7421875,
+      "reward_std": 0.5918886661529541,
+      "rewards/math_reward_func/mean": 0.875,
+      "rewards/math_reward_func/std": 0.3320184051990509,
+      "rewards/think_format_reward/mean": 0.8671875,
+      "rewards/think_format_reward/std": 0.3407054841518402,
+      "sampling/importance_sampling_ratio/max": 2.580911159515381,
+      "sampling/importance_sampling_ratio/mean": 0.14537686109542847,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.1783078908920288,
+      "sampling/sampling_logp_difference/mean": 0.014237256720662117,
+      "sdpo/local_reward_mean": 1.7421875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 1.065335095161572,
+      "step": 4,
+      "step_time": 2334.6283859135583
+    },
+    {
+      "completions/clipped_ratio": 0.390625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16333.0,
+      "completions/mean_length": 10170.78125,
+      "completions/mean_terminated_length": 6187.94873046875,
+      "completions/min_length": 1936.0,
+      "completions/min_terminated_length": 1936.0,
+      "entropy": 0.2497377436957322,
+      "epoch": 0.0390625,
+      "frac_reward_zero_std": 0.4375,
+      "grad_norm": 107.94488525390625,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 91.4281,
+      "num_tokens": 6384731.0,
+      "reward": 1.28125,
+      "reward_std": 0.8956368565559387,
+      "rewards/math_reward_func/mean": 0.6015625,
+      "rewards/math_reward_func/std": 0.4915000796318054,
+      "rewards/think_format_reward/mean": 0.6796875,
+      "rewards/think_format_reward/std": 0.4684300124645233,
+      "sampling/importance_sampling_ratio/max": 2.3500282764434814,
+      "sampling/importance_sampling_ratio/mean": 0.15797367691993713,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.7027404308319092,
+      "sampling/sampling_logp_difference/mean": 0.012200523167848587,
+      "sdpo/local_reward_mean": 1.28125,
+      "sdpo/success_fraction": 0.9375,
+      "sdpo/top100_tail_reverse_kl": 0.7142818871070631,
+      "step": 5,
+      "step_time": 2426.561106611043
+    },
+    {
+      "completions/clipped_ratio": 0.1875,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 15898.0,
+      "completions/mean_length": 9010.265625,
+      "completions/mean_terminated_length": 7308.634765625,
+      "completions/min_length": 1303.0,
+      "completions/min_terminated_length": 1303.0,
+      "entropy": 0.29075702652335167,
+      "epoch": 0.046875,
+      "frac_reward_zero_std": 0.6875,
+      "grad_norm": 140.95046997070312,
+      "learning_rate": 5e-06,
+      "loss": 118.4562,
+      "num_tokens": 7556957.0,
+      "reward": 1.671875,
+      "reward_std": 0.7110973596572876,
+      "rewards/math_reward_func/mean": 0.8359375,
+      "rewards/math_reward_func/std": 0.371787428855896,
+      "rewards/think_format_reward/mean": 0.8359375,
+      "rewards/think_format_reward/std": 0.371787428855896,
+      "sampling/importance_sampling_ratio/max": 2.424870729446411,
+      "sampling/importance_sampling_ratio/mean": 0.10164293646812439,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.9601657390594482,
+      "sampling/sampling_logp_difference/mean": 0.012971683405339718,
+      "sdpo/local_reward_mean": 1.671875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.9254390870919451,
+      "step": 6,
+      "step_time": 2289.461196921766
+    },
+    {
+      "completions/clipped_ratio": 0.09375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16159.0,
+      "completions/mean_length": 7217.25,
+      "completions/mean_terminated_length": 6268.96533203125,
+      "completions/min_length": 1892.0,
+      "completions/min_terminated_length": 1892.0,
+      "entropy": 0.31784707214683294,
+      "epoch": 0.0546875,
+      "frac_reward_zero_std": 0.625,
+      "grad_norm": 168.1649627685547,
+      "learning_rate": 6e-06,
+      "loss": 134.801,
+      "num_tokens": 8499453.0,
+      "reward": 1.6484375,
+      "reward_std": 0.622287392616272,
+      "rewards/math_reward_func/mean": 0.7421875,
+      "rewards/math_reward_func/std": 0.43914902210235596,
+      "rewards/think_format_reward/mean": 0.90625,
+      "rewards/think_format_reward/std": 0.29262590408325195,
+      "sampling/importance_sampling_ratio/max": 1.1390166282653809,
+      "sampling/importance_sampling_ratio/mean": 0.10175397992134094,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.0895158052444458,
+      "sampling/sampling_logp_difference/mean": 0.013488837517797947,
+      "sdpo/local_reward_mean": 1.6484375,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 1.0531327587086707,
+      "step": 7,
+      "step_time": 2179.5239698467776
+    },
+    {
+      "completions/clipped_ratio": 0.1171875,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 14771.0,
+      "completions/mean_length": 7937.2421875,
+      "completions/mean_terminated_length": 6815.9912109375,
+      "completions/min_length": 2105.0,
+      "completions/min_terminated_length": 2105.0,
+      "entropy": 0.27554107748437673,
+      "epoch": 0.0625,
+      "frac_reward_zero_std": 0.6875,
+      "grad_norm": 155.4765167236328,
+      "learning_rate": 7e-06,
+      "loss": 133.8593,
+      "num_tokens": 9539540.0,
+      "reward": 1.8359375,
+      "reward_std": 0.49845972657203674,
+      "rewards/math_reward_func/mean": 0.9296875,
+      "rewards/math_reward_func/std": 0.2566775679588318,
+      "rewards/think_format_reward/mean": 0.90625,
+      "rewards/think_format_reward/std": 0.29262590408325195,
+      "sampling/importance_sampling_ratio/max": 2.6139657497406006,
+      "sampling/importance_sampling_ratio/mean": 0.15319757163524628,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 0.9879323244094849,
+      "sampling/sampling_logp_difference/mean": 0.01237985584884882,
+      "sdpo/local_reward_mean": 1.8359375,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 1.0457757124386262,
+      "step": 8,
+      "step_time": 2288.2132996609434
+    },
+    {
+      "completions/clipped_ratio": 0.15625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 15929.0,
+      "completions/mean_length": 8652.546875,
+      "completions/mean_terminated_length": 7220.79638671875,
+      "completions/min_length": 1778.0,
+      "completions/min_terminated_length": 1778.0,
+      "entropy": 0.27066256816033274,
+      "epoch": 0.0703125,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 147.06039428710938,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 119.0247,
+      "num_tokens": 10665362.0,
+      "reward": 1.6796875,
+      "reward_std": 0.6390641927719116,
+      "rewards/math_reward_func/mean": 0.7890625,
+      "rewards/math_reward_func/std": 0.4095771610736847,
+      "rewards/think_format_reward/mean": 0.890625,
+      "rewards/think_format_reward/std": 0.31333550810813904,
+      "sampling/importance_sampling_ratio/max": 2.7830958366394043,
+      "sampling/importance_sampling_ratio/mean": 0.20229560136795044,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.9212021827697754,
+      "sampling/sampling_logp_difference/mean": 0.01244754996150732,
+      "sdpo/local_reward_mean": 1.6796875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.9298805969301611,
+      "step": 9,
+      "step_time": 2296.1401055157185
+    },
+    {
+      "completions/clipped_ratio": 0.1640625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16312.0,
+      "completions/mean_length": 8341.921875,
+      "completions/mean_terminated_length": 6763.56982421875,
+      "completions/min_length": 1375.0,
+      "completions/min_terminated_length": 1375.0,
+      "entropy": 0.28688407270237803,
+      "epoch": 0.078125,
+      "frac_reward_zero_std": 0.375,
+      "grad_norm": 169.16107177734375,
+      "learning_rate": 9e-06,
+      "loss": 121.9068,
+      "num_tokens": 11752032.0,
+      "reward": 1.578125,
+      "reward_std": 0.7382611632347107,
+      "rewards/math_reward_func/mean": 0.734375,
+      "rewards/math_reward_func/std": 0.44340085983276367,
+      "rewards/think_format_reward/mean": 0.84375,
+      "rewards/think_format_reward/std": 0.3645188808441162,
+      "sampling/importance_sampling_ratio/max": 2.4416871070861816,
+      "sampling/importance_sampling_ratio/mean": 0.13641589879989624,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.7802882194519043,
+      "sampling/sampling_logp_difference/mean": 0.013117711059749126,
+      "sdpo/local_reward_mean": 1.578125,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.9523969579604454,
+      "step": 10,
+      "step_time": 2242.045174873434
+    },
+    {
+      "completions/clipped_ratio": 0.1015625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 15675.0,
+      "completions/mean_length": 7631.2421875,
+      "completions/mean_terminated_length": 6641.7998046875,
+      "completions/min_length": 673.0,
+      "completions/min_terminated_length": 673.0,
+      "entropy": 0.3130435968050733,
+      "epoch": 0.0859375,
+      "frac_reward_zero_std": 0.625,
+      "grad_norm": 223.6240692138672,
+      "learning_rate": 1e-05,
+      "loss": 141.7518,
+      "num_tokens": 12745735.0,
+      "reward": 1.6796875,
+      "reward_std": 0.6139274835586548,
+      "rewards/math_reward_func/mean": 0.765625,
+      "rewards/math_reward_func/std": 0.42527204751968384,
+      "rewards/think_format_reward/mean": 0.9140625,
+      "rewards/think_format_reward/std": 0.2813730239868164,
+      "sampling/importance_sampling_ratio/max": 1.0876129865646362,
+      "sampling/importance_sampling_ratio/mean": 0.07780440151691437,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.2738298177719116,
+      "sampling/sampling_logp_difference/mean": 0.014252403751015663,
+      "sdpo/local_reward_mean": 1.6796875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 1.1074355833115987,
+      "step": 11,
+      "step_time": 2203.812321053818
+    },
+    {
+      "completions/clipped_ratio": 0.109375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16168.0,
+      "completions/mean_length": 7337.203125,
+      "completions/mean_terminated_length": 6226.19287109375,
+      "completions/min_length": 825.0,
+      "completions/min_terminated_length": 825.0,
+      "entropy": 0.25038730120286345,
+      "epoch": 0.09375,
+      "frac_reward_zero_std": 0.75,
+      "grad_norm": 196.19200134277344,
+      "learning_rate": 9.88888888888889e-06,
+      "loss": 133.9144,
+      "num_tokens": 13705689.0,
+      "reward": 1.7734375,
+      "reward_std": 0.6046388149261475,
+      "rewards/math_reward_func/mean": 0.875,
+      "rewards/math_reward_func/std": 0.3320184051990509,
+      "rewards/think_format_reward/mean": 0.8984375,
+      "rewards/think_format_reward/std": 0.3032590448856354,
+      "sampling/importance_sampling_ratio/max": 2.7273266315460205,
+      "sampling/importance_sampling_ratio/mean": 0.1523737907409668,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 3.2629263401031494,
+      "sampling/sampling_logp_difference/mean": 0.014198469929397106,
+      "sdpo/local_reward_mean": 1.7734375,
+      "sdpo/success_fraction": 0.9375,
+      "sdpo/top100_tail_reverse_kl": 1.0462062492151745,
+      "step": 12,
+      "step_time": 2225.1963889272884
+    },
+    {
+      "completions/clipped_ratio": 0.1015625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16207.0,
+      "completions/mean_length": 9305.1953125,
+      "completions/mean_terminated_length": 8504.982421875,
+      "completions/min_length": 2107.0,
+      "completions/min_terminated_length": 2107.0,
+      "entropy": 0.282024092390202,
+      "epoch": 0.1015625,
+      "frac_reward_zero_std": 0.625,
+      "grad_norm": 153.59938049316406,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 117.6161,
+      "num_tokens": 14914506.0,
+      "reward": 1.734375,
+      "reward_std": 0.5395299196243286,
+      "rewards/math_reward_func/mean": 0.8125,
+      "rewards/math_reward_func/std": 0.39184603095054626,
+      "rewards/think_format_reward/mean": 0.921875,
+      "rewards/think_format_reward/std": 0.2694226801395416,
+      "sampling/importance_sampling_ratio/max": 1.677321195602417,
+      "sampling/importance_sampling_ratio/mean": 0.0686572790145874,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.2385106086730957,
+      "sampling/sampling_logp_difference/mean": 0.013802426867187023,
+      "sdpo/local_reward_mean": 1.734375,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.9188754103379324,
+      "step": 13,
+      "step_time": 2269.5332072526217
+    },
+    {
+      "completions/clipped_ratio": 0.234375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16063.0,
+      "completions/mean_length": 8390.734375,
+      "completions/mean_terminated_length": 5943.81640625,
+      "completions/min_length": 1662.0,
+      "completions/min_terminated_length": 1662.0,
+      "entropy": 0.20445589907467365,
+      "epoch": 0.109375,
+      "frac_reward_zero_std": 0.625,
+      "grad_norm": 148.29547119140625,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 103.8433,
+      "num_tokens": 16003976.0,
+      "reward": 1.4921875,
+      "reward_std": 0.8228631615638733,
+      "rewards/math_reward_func/mean": 0.71875,
+      "rewards/math_reward_func/std": 0.4513758420944214,
+      "rewards/think_format_reward/mean": 0.7734375,
+      "rewards/think_format_reward/std": 0.4202519655227661,
+      "sampling/importance_sampling_ratio/max": 1.0998280048370361,
+      "sampling/importance_sampling_ratio/mean": 0.09229649603366852,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 4.958465099334717,
+      "sampling/sampling_logp_difference/mean": 0.01327038835734129,
+      "sdpo/local_reward_mean": 1.4921875,
+      "sdpo/success_fraction": 0.875,
+      "sdpo/top100_tail_reverse_kl": 0.8112759278155863,
+      "step": 14,
+      "step_time": 2283.0402548387647
+    },
+    {
+      "completions/clipped_ratio": 0.09375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16309.0,
+      "completions/mean_length": 8678.2109375,
+      "completions/mean_terminated_length": 7881.060546875,
+      "completions/min_length": 1679.0,
+      "completions/min_terminated_length": 1679.0,
+      "entropy": 0.2764591637533158,
+      "epoch": 0.1171875,
+      "frac_reward_zero_std": 0.5625,
+      "grad_norm": 162.71365356445312,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 110.2911,
+      "num_tokens": 17132739.0,
+      "reward": 1.6875,
+      "reward_std": 0.5985546708106995,
+      "rewards/math_reward_func/mean": 0.7734375,
+      "rewards/math_reward_func/std": 0.4202519655227661,
+      "rewards/think_format_reward/mean": 0.9140625,
+      "rewards/think_format_reward/std": 0.2813730239868164,
+      "sampling/importance_sampling_ratio/max": 2.0603184700012207,
+      "sampling/importance_sampling_ratio/mean": 0.07795818150043488,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.717520236968994,
+      "sampling/sampling_logp_difference/mean": 0.0152816753834486,
+      "sdpo/local_reward_mean": 1.6875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.8616492211585864,
+      "step": 15,
+      "step_time": 2284.2519008833915
+    },
+    {
+      "completions/clipped_ratio": 0.09375,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16016.0,
+      "completions/mean_length": 7669.890625,
+      "completions/mean_terminated_length": 6768.43115234375,
+      "completions/min_length": 2206.0,
+      "completions/min_terminated_length": 2206.0,
+      "entropy": 0.23759500740561634,
+      "epoch": 0.125,
+      "frac_reward_zero_std": 0.6875,
+      "grad_norm": 156.71192932128906,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 109.7674,
+      "num_tokens": 18134285.0,
+      "reward": 1.7890625,
+      "reward_std": 0.5563144087791443,
+      "rewards/math_reward_func/mean": 0.8671875,
+      "rewards/math_reward_func/std": 0.3407054841518402,
+      "rewards/think_format_reward/mean": 0.921875,
+      "rewards/think_format_reward/std": 0.2694226801395416,
+      "sampling/importance_sampling_ratio/max": 1.7341861724853516,
+      "sampling/importance_sampling_ratio/mean": 0.07510986924171448,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 1.8912134170532227,
+      "sampling/sampling_logp_difference/mean": 0.014694299548864365,
+      "sdpo/local_reward_mean": 1.7890625,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.8575580433825962,
+      "step": 16,
+      "step_time": 2196.966399327852
+    },
+    {
+      "completions/clipped_ratio": 0.1015625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 14528.0,
+      "completions/mean_length": 6123.2109375,
+      "completions/mean_terminated_length": 4963.29541015625,
+      "completions/min_length": 1365.0,
+      "completions/min_terminated_length": 1365.0,
+      "entropy": 0.1932699321769178,
+      "epoch": 0.1328125,
+      "frac_reward_zero_std": 0.625,
+      "grad_norm": 162.4973907470703,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 111.5547,
+      "num_tokens": 18936144.0,
+      "reward": 1.7421875,
+      "reward_std": 0.5918886661529541,
+      "rewards/math_reward_func/mean": 0.828125,
+      "rewards/math_reward_func/std": 0.3787541687488556,
+      "rewards/think_format_reward/mean": 0.9140625,
+      "rewards/think_format_reward/std": 0.2813730239868164,
+      "sampling/importance_sampling_ratio/max": 2.4520015716552734,
+      "sampling/importance_sampling_ratio/mean": 0.11252652108669281,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 2.6565070152282715,
+      "sampling/sampling_logp_difference/mean": 0.015431585721671581,
+      "sdpo/local_reward_mean": 1.7421875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.8715214569820091,
+      "step": 17,
+      "step_time": 2084.307107809931
+    },
+    {
+      "completions/clipped_ratio": 0.1328125,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16280.0,
+      "completions/mean_length": 8076.0234375,
+      "completions/mean_terminated_length": 6803.630859375,
+      "completions/min_length": 572.0,
+      "completions/min_terminated_length": 572.0,
+      "entropy": 0.20287329837447032,
+      "epoch": 0.140625,
+      "frac_reward_zero_std": 0.5,
+      "grad_norm": 159.11508178710938,
+      "learning_rate": 9.222222222222224e-06,
+      "loss": 106.6876,
+      "num_tokens": 19987187.0,
+      "reward": 1.734375,
+      "reward_std": 0.6081386804580688,
+      "rewards/math_reward_func/mean": 0.8359375,
+      "rewards/math_reward_func/std": 0.371787428855896,
+      "rewards/think_format_reward/mean": 0.8984375,
+      "rewards/think_format_reward/std": 0.3032590448856354,
+      "sampling/importance_sampling_ratio/max": 2.501821279525757,
+      "sampling/importance_sampling_ratio/mean": 0.09660257399082184,
+      "sampling/importance_sampling_ratio/min": 4.999841917685275e-21,
+      "sampling/sampling_logp_difference/max": 2.8883635997772217,
+      "sampling/sampling_logp_difference/mean": 0.017189152538776398,
+      "sdpo/local_reward_mean": 1.734375,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.8334968033013865,
+      "step": 18,
+      "step_time": 2242.6666490389034
+    },
+    {
+      "completions/clipped_ratio": 0.0625,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 14672.0,
+      "completions/mean_length": 6652.9609375,
+      "completions/mean_terminated_length": 6004.22509765625,
+      "completions/min_length": 1682.0,
+      "completions/min_terminated_length": 1682.0,
+      "entropy": 0.19443697290262207,
+      "epoch": 0.1484375,
+      "frac_reward_zero_std": 0.75,
+      "grad_norm": 166.24961853027344,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 108.2128,
+      "num_tokens": 20858630.0,
+      "reward": 1.890625,
+      "reward_std": 0.4014613926410675,
+      "rewards/math_reward_func/mean": 0.921875,
+      "rewards/math_reward_func/std": 0.2694226801395416,
+      "rewards/think_format_reward/mean": 0.96875,
+      "rewards/think_format_reward/std": 0.1746762990951538,
+      "sampling/importance_sampling_ratio/max": 0.8068121075630188,
+      "sampling/importance_sampling_ratio/mean": 0.021228063851594925,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 4.259117126464844,
+      "sampling/sampling_logp_difference/mean": 0.019931036978960037,
+      "sdpo/local_reward_mean": 1.890625,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.8454125826247036,
+      "step": 19,
+      "step_time": 2172.696639011614
+    },
+    {
+      "completions/clipped_ratio": 0.046875,
+      "completions/max_length": 16384.0,
+      "completions/max_terminated_length": 16371.0,
+      "completions/mean_length": 7623.7734375,
+      "completions/mean_terminated_length": 7192.9423828125,
+      "completions/min_length": 1386.0,
+      "completions/min_terminated_length": 1386.0,
+      "entropy": 0.19962301407940686,
+      "epoch": 0.15625,
+      "frac_reward_zero_std": 0.6875,
+      "grad_norm": 140.9843292236328,
+      "learning_rate": 9e-06,
+      "loss": 91.6438,
+      "num_tokens": 21855721.0,
+      "reward": 1.7421875,
+      "reward_std": 0.5058102607727051,
+      "rewards/math_reward_func/mean": 0.7734375,
+      "rewards/math_reward_func/std": 0.4202519655227661,
+      "rewards/think_format_reward/mean": 0.96875,
+      "rewards/think_format_reward/std": 0.1746762990951538,
+      "sampling/importance_sampling_ratio/max": 1.627164602279663,
+      "sampling/importance_sampling_ratio/mean": 0.07164306938648224,
+      "sampling/importance_sampling_ratio/min": 0.0,
+      "sampling/sampling_logp_difference/max": 3.246943950653076,
+      "sampling/sampling_logp_difference/mean": 0.022836929187178612,
+      "sdpo/local_reward_mean": 1.7421875,
+      "sdpo/success_fraction": 1.0,
+      "sdpo/top100_tail_reverse_kl": 0.7159674066933803,
+      "step": 20,
+      "step_time": 2231.109877148643
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 100,
+  "num_input_tokens_seen": 21855721,
+  "num_train_epochs": 1,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-20/training_args.bin ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:8e18becd6b26b8676b3a8f105ac44d4afad88e666d262a6709bc21415d543da5
+size 8145