{ "architectures": [ "PhiForCausalLM" ], "bos_token_id": 1, "embd_pdrop": 0.0, "eos_token_id": 2, "hidden_act": "gelu_new", "hidden_size": 2048, "initializer_range": 0.02, "intermediate_size": 8192, "max_position_embeddings": 2048, "model_type": "phi", "num_attention_heads": 32, "num_hidden_layers": 24, "pretraining_tp": 1, "resid_pdrop": 0.0, "layer_norm_eps": 1e-05, "rope_scaling": null, "rope_theta": 10000.0, "partial_rotary_factor": 0.5, "qk_layernorm": false, "tie_word_embeddings": false, "transformers_version": "4.34.0.dev0", "torch_dtype": "float16", "use_cache": true, "vocab_size": 51200 }