Instructions to use zenlm/zen4-storm with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use zenlm/zen4-storm with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="zenlm/zen4-storm")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("zenlm/zen4-storm", dtype="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use zenlm/zen4-storm with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "zenlm/zen4-storm" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "zenlm/zen4-storm", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/zenlm/zen4-storm
- SGLang
How to use zenlm/zen4-storm 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 "zenlm/zen4-storm" \ --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": "zenlm/zen4-storm", "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 "zenlm/zen4-storm" \ --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": "zenlm/zen4-storm", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use zenlm/zen4-storm with Docker Model Runner:
docker model run hf.co/zenlm/zen4-storm
Zen4 Storm
Zen4 Storm is a 456B MoE (45B active) parameter language model from the Zen4 family by Zen LM and Hanzo AI.
Hybrid MoE with Lightning Attention for ultra-long context reasoning.
Model Details
| Property | Value |
|---|---|
| Parameters | 456B MoE total, 45B active |
| Architecture | Zen4 Frontier |
| Context | 1M tokens |
| License | MIT |
| Family | Zen4 |
| Tier | Frontier |
| Creator | Zen LM / Hanzo AI |
Weights
Weights hosted at MiniMaxAI/MiniMax-M1-80k due to storage constraints. Use the source repository for inference and fine-tuning.
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("MiniMaxAI/MiniMax-M1-80k", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained("MiniMaxAI/MiniMax-M1-80k")
Links
Zen AI: Clarity Through Intelligence
Model tree for zenlm/zen4-storm
Base model
MiniMaxAI/MiniMax-M1-80k