Instructions to use refactai/codify_3b_multi with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use refactai/codify_3b_multi with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="refactai/codify_3b_multi", trust_remote_code=True)# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("refactai/codify_3b_multi", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use refactai/codify_3b_multi with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "refactai/codify_3b_multi" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "refactai/codify_3b_multi", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/refactai/codify_3b_multi
- SGLang
How to use refactai/codify_3b_multi 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 "refactai/codify_3b_multi" \ --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": "refactai/codify_3b_multi", "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 "refactai/codify_3b_multi" \ --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": "refactai/codify_3b_multi", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use refactai/codify_3b_multi with Docker Model Runner:
docker model run hf.co/refactai/codify_3b_multi
| { | |
| "E": 2560, | |
| "L": 32, | |
| "T": 2048, | |
| "_mup": true, | |
| "alt_pw_klass": { | |
| "type": "" | |
| }, | |
| "alt_rel_klass": { | |
| "fused": true, | |
| "type": "alibi" | |
| }, | |
| "alt_sa_klass": { | |
| "triton": true, | |
| "type": "flash", | |
| "use_rotary_emb": null | |
| }, | |
| "architectures": [ | |
| "CodifyForCausalLM" | |
| ], | |
| "attn_a_reach": 2048, | |
| "attn_b_reach": 2048, | |
| "attn_heads": 40, | |
| "attn_ra_nbasis": 64, | |
| "attn_seq": [ | |
| "d" | |
| ], | |
| "attn_sparse_layout_seq": null, | |
| "auto_map": { | |
| "AutoConfig": "configuration_codify.CodifyConfig", | |
| "AutoModel": "modeling_codify.CodifyForCausalLM" | |
| }, | |
| "backcheck_pw": "inside", | |
| "backcheck_sa": "none", | |
| "bos_token_id": 1, | |
| "dtype_acts": "torch.float16", | |
| "dtype_weights": "torch.float16", | |
| "enc_name": "openai_programming_v2", | |
| "eos_token_id": 2, | |
| "init_scale": 1, | |
| "initializer_range": 0.02, | |
| "layer_norm_epsilon": 1e-05, | |
| "mlp_mult": 4, | |
| "model_type": "codify", | |
| "moe": null, | |
| "mup_optimal_lr": 0.0005, | |
| "mup_shapes_file": "lean_former/mup/alibi_32l/shapes.json", | |
| "posemb": false, | |
| "rescale_embeddings": false, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.24.0", | |
| "tune": [ | |
| 3, | |
| 3, | |
| 3, | |
| 3 | |
| ], | |
| "unembedding_shared": false, | |
| "use_cache": true, | |
| "use_res_scale": false, | |
| "vocab_size": 51305 | |
| } | |