tomasmcm/KwaiCoder-AutoThink-preview-mlx-3Bit

1{2    "architectures": [3        "Qwen2ForCausalLM"4    ],5    "attention_dropout": 0.0,6    "bos_token_id": 151643,7    "eos_token_id": 151645,8    "hidden_act": "silu",9    "hidden_size": 5120,10    "initializer_range": 0.02,11    "intermediate_size": 27648,12    "max_position_embeddings": 32768,13    "max_window_layers": 70,14    "model_type": "qwen2",15    "num_attention_heads": 40,16    "num_hidden_layers": 80,17    "num_key_value_heads": 8,18    "quantization": {19        "group_size": 64,20        "bits": 321    },22    "quantization_config": {23        "group_size": 64,24        "bits": 325    },26    "rms_norm_eps": 1e-06,27    "rope_scaling": null,28    "rope_theta": 1000000.0,29    "sliding_window": null,30    "tie_word_embeddings": false,31    "torch_dtype": "bfloat16",32    "transformers_version": "4.51.3",33    "use_cache": false,34    "use_sliding_window": false,35    "vocab_size": 15206436}