Instructions to use SciReason/SciReasoner-8B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use SciReason/SciReasoner-8B with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="SciReason/SciReasoner-8B")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("SciReason/SciReasoner-8B") model = AutoModelForCausalLM.from_pretrained("SciReason/SciReasoner-8B") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use SciReason/SciReasoner-8B with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "SciReason/SciReasoner-8B" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "SciReason/SciReasoner-8B", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/SciReason/SciReasoner-8B
- SGLang
How to use SciReason/SciReasoner-8B 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 "SciReason/SciReasoner-8B" \ --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": "SciReason/SciReasoner-8B", "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 "SciReason/SciReasoner-8B" \ --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": "SciReason/SciReasoner-8B", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use SciReason/SciReasoner-8B with Docker Model Runner:
docker model run hf.co/SciReason/SciReasoner-8B
Improve model card: Add library, pipeline tags, paper link, and GitHub link
#1
by nielsr HF Staff - opened
This PR enhances the model card for SciReasoner 8B by adding crucial metadata and improving content clarity:
library_name: transformers: The model is compatible with thetransformerslibrary, which enables an automated code snippet on the Hub, as evidenced byconfig.json(e.g.,"architectures": ["Qwen3ForCausalLM"],"model_type": "qwen3").pipeline_tag: text-generation: This tag helps users discover the model for its described capabilities in "unconditional and conditional sequence generation and design" as per the paper abstract.tags: ['qwen', 'scientific-reasoning']: Adds relevant tags based on the model's architecture (Qwen3ForCausalLM) and its primary purpose as a "scientific reasoning foundation model."- Paper Link: The official Hugging Face paper link (SciReasoner: Laying the Scientific Reasoning Ground Across Disciplines) has been explicitly added to the introductory text for direct access. The existing arXiv badge is retained.
- GitHub Link: An explicit link to the GitHub repository (
https://github.com/open-sciencelab/SciReason) has been added for easier access to the associated code and evaluation scripts.
These updates improve the model's discoverability and provide users with clearer, more direct access to essential information.
1hunters changed pull request status to merged