{ "_name_or_path": "niwz/Mini-Chinese-Phi3", "architectures": [ "MiniPhi3" ], "attention_dropout": 0.0, "bos_token_id": 2, "embd_pdrop": 0.0, "eos_token_id": 1, "hidden_act": "silu", "hidden_size": 768, "initializer_range": 0.02, "intermediate_size": 2048, "max_position_embeddings": 512, "model_type": "phi3", "num_attention_heads": 12, "num_hidden_layers": 12, "num_key_value_heads": 12, "original_max_position_embeddings": 512, "pad_token_id": 0, "resid_pdrop": 0.0, "rms_norm_eps": 1e-05, "rope_scaling": null, "rope_theta": 10000.0, "sliding_window": null, "tie_word_embeddings": false, "torch_dtype": "float32", "transformers_version": "4.41.2", "use_cache": true, "use_cope": false, "vocab_size": 32064 }