Instructions to use LLM360/Crystal with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use LLM360/Crystal with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="LLM360/Crystal", trust_remote_code=True)# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("LLM360/Crystal", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use LLM360/Crystal with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "LLM360/Crystal" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "LLM360/Crystal", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/LLM360/Crystal
- SGLang
How to use LLM360/Crystal 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 "LLM360/Crystal" \ --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": "LLM360/Crystal", "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 "LLM360/Crystal" \ --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": "LLM360/Crystal", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use LLM360/Crystal with Docker Model Runner:
docker model run hf.co/LLM360/Crystal
| { | |
| "attn_pdrop": 0.0, | |
| "scale_attn_weights": true, | |
| "resid_pdrop": 0.0, | |
| "mup_embeddings_scale": 14.6, | |
| "n_inner": 10922, | |
| "n_embd": 4096, | |
| "layer_norm_epsilon": 1e-05, | |
| "n_positions": 2048, | |
| "activation_function": "swiglu", | |
| "n_head": 32, | |
| "n_layer": 32, | |
| "mup_output_alpha": 2.22, | |
| "mup_width_scale": 0.0625, | |
| "position_embedding_type": "rotary", | |
| "rotary_dim": 32, | |
| "mup_scale_qk_dot_by_d": true, | |
| "tie_word_embeddings": true, | |
| "vocab_size": 32032, | |
| "embd_pdrop": 0.0, | |
| "model_type": "crystalcoder", | |
| "use_cache": true, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "auto_map": { | |
| "AutoConfig": "configuration_crystalcoder.CrystalCoderConfig", | |
| "AutoModel": "modeling_crystalcoder.CrystalCoderModel", | |
| "AutoModelForCausalLM": "modeling_crystalcoder.CrystalCoderLMHeadModel" | |
| }, | |
| "architectures": [ | |
| "CrystalCoderLMHeadModel" | |
| ] | |
| } |