Instructions to use ArthurZ/mamba-370m with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ArthurZ/mamba-370m with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="ArthurZ/mamba-370m")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("ArthurZ/mamba-370m") model = AutoModelForCausalLM.from_pretrained("ArthurZ/mamba-370m") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use ArthurZ/mamba-370m with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "ArthurZ/mamba-370m" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ArthurZ/mamba-370m", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/ArthurZ/mamba-370m
- SGLang
How to use ArthurZ/mamba-370m 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 "ArthurZ/mamba-370m" \ --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": "ArthurZ/mamba-370m", "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 "ArthurZ/mamba-370m" \ --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": "ArthurZ/mamba-370m", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use ArthurZ/mamba-370m with Docker Model Runner:
docker model run hf.co/ArthurZ/mamba-370m
File size: 713 Bytes
2631ea3 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 | {
"_name_or_path": "state-spaces/mamba-370m",
"architectures": [
"MambaForCausalLM"
],
"bos_token_id": 1,
"conv_kernel": 4,
"d_inner": 160,
"d_model": 1024,
"dt_rank": "auto",
"eos_token_id": 2,
"expand": 2,
"fused_add_norm": true,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.1,
"layer_norm_epsilon": 1e-05,
"model_type": "mamba",
"n_layer": 48,
"num_hidden_layers": 48,
"pad_vocab_size_multiple": 8,
"residual_in_fp32": true,
"rms_norm": true,
"ssm_cfg": {},
"state_size": 16,
"time_step_rank": 64,
"torch_dtype": "float16",
"transformers_version": "4.38.0.dev0",
"use_bias": false,
"use_conv_bias": true,
"vocab_size": 50280
}
|