Instructions to use openai/circuit-sparsity with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use openai/circuit-sparsity with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="openai/circuit-sparsity", trust_remote_code=True)# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("openai/circuit-sparsity", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use openai/circuit-sparsity with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "openai/circuit-sparsity" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "openai/circuit-sparsity", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/openai/circuit-sparsity
- SGLang
How to use openai/circuit-sparsity 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 "openai/circuit-sparsity" \ --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": "openai/circuit-sparsity", "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 "openai/circuit-sparsity" \ --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": "openai/circuit-sparsity", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use openai/circuit-sparsity with Docker Model Runner:
docker model run hf.co/openai/circuit-sparsity
| { | |
| "activation_type": "gelu", | |
| "afrac": 0.25, | |
| "afrac_loctypes": "attn_in,attn_out,mlp_in,mlp_out,mlp_neuron,attn_v,attn_k,attn_q", | |
| "architectures": [ | |
| "CircuitGPTForCausalLM" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "config.CircuitGPTConfig", | |
| "AutoModelForCausalLM": "modeling_circuitgpt.CircuitGPTForCausalLM" | |
| }, | |
| "bias": true, | |
| "bigram_table_rank": null, | |
| "block_size": 1024, | |
| "bos_token_id": null, | |
| "d_head": 16, | |
| "d_mlp": 8192, | |
| "d_model": 2048, | |
| "d_pos_emb": 32, | |
| "dropout": 0.0, | |
| "dropout_cat_pos_emb": false, | |
| "enable_bigram_table": true, | |
| "eos_token_id": 2047, | |
| "flash": true, | |
| "is_decoder": true, | |
| "learnable_bigram_table": true, | |
| "ln_bias": true, | |
| "max_position_embeddings": 1024, | |
| "model_type": "circuitgpt", | |
| "n_head": 128, | |
| "n_layer": 8, | |
| "pad_token_id": null, | |
| "residual_activation_type": "identity", | |
| "rms_norm": true, | |
| "sink": true, | |
| "sinusoidal_cat_pos_emb": false, | |
| "tie_word_embeddings": false, | |
| "tied_unembed": false, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.49.0", | |
| "unembed_rank": null, | |
| "use_position_embeddings": true, | |
| "vocab_size": 2048 | |
| } |