Instructions to use luiscalisto/EuroMoE-2.6B-A0.6B-Instruct-2512-MLX-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use luiscalisto/EuroMoE-2.6B-A0.6B-Instruct-2512-MLX-4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir EuroMoE-2.6B-A0.6B-Instruct-2512-MLX-4bit luiscalisto/EuroMoE-2.6B-A0.6B-Instruct-2512-MLX-4bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
| { | |
| "architectures": [ | |
| "MixtralForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 1, | |
| "dtype": "float32", | |
| "eos_token_id": 4, | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 1024, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 512, | |
| "max_position_embeddings": 32768, | |
| "model_type": "mixtral", | |
| "num_attention_heads": 8, | |
| "num_experts_per_tok": 8, | |
| "num_hidden_layers": 24, | |
| "num_key_value_heads": 2, | |
| "num_local_experts": 64, | |
| "output_router_logits": false, | |
| "quantization": { | |
| "group_size": 64, | |
| "bits": 4, | |
| "mode": "affine" | |
| }, | |
| "quantization_config": { | |
| "group_size": 64, | |
| "bits": 4, | |
| "mode": "affine" | |
| }, | |
| "rms_norm_eps": 1e-05, | |
| "rope_theta": 500000, | |
| "router_aux_loss_coef": 0.01, | |
| "router_jitter_noise": 0.0, | |
| "sliding_window": null, | |
| "tie_word_embeddings": true, | |
| "transformers_version": "4.57.3", | |
| "use_cache": false, | |
| "vocab_size": 128000 | |
| } |