Instructions to use Idred/BlastRadius-GRPO-Checkpoints with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Idred/BlastRadius-GRPO-Checkpoints with Transformers:

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

pipe = pipeline("text-generation", model="Idred/BlastRadius-GRPO-Checkpoints")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("Idred/BlastRadius-GRPO-Checkpoints", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Idred/BlastRadius-GRPO-Checkpoints with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Idred/BlastRadius-GRPO-Checkpoints"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Idred/BlastRadius-GRPO-Checkpoints",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Idred/BlastRadius-GRPO-Checkpoints

SGLang

How to use Idred/BlastRadius-GRPO-Checkpoints 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 "Idred/BlastRadius-GRPO-Checkpoints" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Idred/BlastRadius-GRPO-Checkpoints",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "Idred/BlastRadius-GRPO-Checkpoints" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Idred/BlastRadius-GRPO-Checkpoints",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Unsloth Studio

How to use Idred/BlastRadius-GRPO-Checkpoints with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Idred/BlastRadius-GRPO-Checkpoints to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Idred/BlastRadius-GRPO-Checkpoints to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Idred/BlastRadius-GRPO-Checkpoints to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="Idred/BlastRadius-GRPO-Checkpoints",
    max_seq_length=2048,
)

Docker Model Runner
How to use Idred/BlastRadius-GRPO-Checkpoints with Docker Model Runner:
```
docker model run hf.co/Idred/BlastRadius-GRPO-Checkpoints
```

BlastRadius-GRPO-Checkpoints

File size: 1,248 Bytes

---
license: apache-2.0
language:
  - en
base_model: unsloth/Qwen2.5-14B-Instruct-bnb-4bit
pipeline_tag: text-generation
library_name: transformers
tags:
  - reinforcement-learning
  - grpo
  - openenv
  - blast-radius
  - qwen2.5
  - unsloth
  - sft
  - rl-environment
metrics:
  - type: reward
    value: 0.72
    name: Mean Episode Reward
  - type: reward
    value: 0.75
    name: Format Reward
  - type: reward
    value: 0.48
    name: KL Divergence
---

# BlastRadius — GRPO Model Checkpoints

This repository contains the trained model checkpoints.

## Live Demo

https://huggingface.co/spaces/Idred/BlastRadius-OpenEnv

## Training Notebook

https://huggingface.co/spaces/Idred/BlastRadius-OpenEnv/blob/main/BlastRadius_A100_Training_v2.ipynb

## Training Details

- **Hardware:** Hugging Face Jobs (H200 GPU)
- **Framework:** PyTorch 2.6 (CUDA 12.4)
- **Approach:** SFT + GRPO (Reinforcement Learning)
- **Experiment Tracking:** Weights & Biases (WandB)

## Note

- The Space provides the complete working demo
- The notebook contains the full training pipeline and reproducible steps
- This repository is for model checkpoints only
- HF Jobs are not publicly accessible by design - the notebook serves as the verifiable training record