Instructions to use ekryski/SmolLM-360M-Instruct-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use ekryski/SmolLM-360M-Instruct-4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir SmolLM-360M-Instruct-4bit ekryski/SmolLM-360M-Instruct-4bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- LM Studio
| { | |
| "_name_or_path" : "HuggingFaceTB\/SmolLM-360M", | |
| "architectures" : [ | |
| "LlamaForCausalLM" | |
| ], | |
| "attention_bias" : 0, | |
| "attention_dropout" : 0, | |
| "bos_token_id" : 1, | |
| "eos_token_id" : 2, | |
| "hidden_act" : "silu", | |
| "hidden_size" : 960, | |
| "initializer_range" : 0.02, | |
| "intermediate_size" : 2560, | |
| "max_position_embeddings" : 2048, | |
| "mlp_bias" : 0, | |
| "model_type" : "llama", | |
| "num_attention_heads" : 15, | |
| "num_hidden_layers" : 32, | |
| "num_key_value_heads" : 5, | |
| "pad_token_id" : 2, | |
| "pretraining_tp" : 1, | |
| "quantization" : { | |
| "bits" : 4, | |
| "group_size" : 64, | |
| "mode" : "affine" | |
| }, | |
| "quantization_config" : { | |
| "bits" : 4, | |
| "group_size" : 64, | |
| "mode" : "affine" | |
| }, | |
| "rms_norm_eps" : 1.0000000000000001e-05, | |
| "rope_scaling" : null, | |
| "rope_theta" : 10000, | |
| "tie_word_embeddings" : 1, | |
| "torch_dtype" : "bfloat16", | |
| "transformers_version" : "4.42.3", | |
| "use_cache" : 1, | |
| "vocab_size" : 49152 | |
| } |