Instructions to use dipta007/dagger-4B_SFT_GRPO with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use dipta007/dagger-4B_SFT_GRPO with Transformers:

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

pipe = pipeline("text-generation", model="dipta007/dagger-4B_SFT_GRPO")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("dipta007/dagger-4B_SFT_GRPO")
model = AutoModelForMultimodalLM.from_pretrained("dipta007/dagger-4B_SFT_GRPO", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use dipta007/dagger-4B_SFT_GRPO with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "dipta007/dagger-4B_SFT_GRPO"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dipta007/dagger-4B_SFT_GRPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/dipta007/dagger-4B_SFT_GRPO

SGLang

How to use dipta007/dagger-4B_SFT_GRPO 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 "dipta007/dagger-4B_SFT_GRPO" \
    --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": "dipta007/dagger-4B_SFT_GRPO",
		"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 "dipta007/dagger-4B_SFT_GRPO" \
        --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": "dipta007/dagger-4B_SFT_GRPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use dipta007/dagger-4B_SFT_GRPO with Docker Model Runner:
```
docker model run hf.co/dipta007/dagger-4B_SFT_GRPO
```

dipta007 commited on Aug 23

Commit

393b2c1

verified ·

1 Parent(s): c30e68e

Fix ablation weighted averages, stored label values, reasoning-model drops, byte counts; document non-commercial training data

Browse files

Files changed (1) hide show

README.md +10 -1

README.md CHANGED Viewed

@@ -44,7 +44,7 @@ datasets:
 - **Lightweight**: 4B parameters for resource-constrained deployment
 - **SFT → GRPO training**: Full training pipeline
-- **Improved over baselines**: Still outperforms CoT on distractor robustness
 - **Capacity study**: Demonstrates model size requirements for graph generation
 ## Model Overview
@@ -183,6 +183,15 @@ Same as 12B variant:
 | [dagger-4B_SFT](https://huggingface.co/dipta007/dagger-4B_SFT) | 4B | 44.3 |
 | [dagger-12B_SFT_GRPO](https://huggingface.co/dipta007/dagger-12B_SFT_GRPO) | 12B | **69.4** |
 ## Citation
 ```bibtex

 - **Lightweight**: 4B parameters for resource-constrained deployment
 - **SFT → GRPO training**: Full training pipeline
+- **Clear gain over 4B SFT**: +3.0 weighted accuracy (44.3 → 47.3). Note this 4B model is *less* distractor-robust than the stronger CoT baselines (it drops 23.4 / 27.4 points); the 12B model is the robust one
 - **Capacity study**: Demonstrates model size requirements for graph generation
 ## Model Overview
 | [dagger-4B_SFT](https://huggingface.co/dipta007/dagger-4B_SFT) | 4B | 44.3 |
 | [dagger-12B_SFT_GRPO](https://huggingface.co/dipta007/dagger-12B_SFT_GRPO) | 12B | **69.4** |
+## License and Data Provenance
+Model weights are released under the [Gemma Terms of Use](https://ai.google.dev/gemma/terms).
+**Training data is not fully permissive.** Part of the SFT data and all GRPO prompts come
+from `numina-math-cot-bn`, which is **CC BY-NC-SA 4.0 (NonCommercial, ShareAlike)**. For
+commercial use, re-derive that portion from the Apache-2.0 upstream
+[AI-MO/NuminaMath-CoT](https://huggingface.co/datasets/AI-MO/NuminaMath-CoT).
 ## Citation
 ```bibtex