{ "_name_or_path": "/ephemeral/hub/models--huggyllama--llama-7b/snapshots/4782ad278652c7c71b72204d462d6d01eaaf7549/", "architectures": [ "YocoLlamaForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 1, "crossDecoder_start_idx": 1, "eos_token_id": 2, "hidden_act": "silu", "hidden_size": 4096, "initializer_range": 0.02, "intermediate_size": 11008, "max_position_embeddings": 2048, "max_sequence_length": 2048, "mlp_bias": false, "model_type": "yoco_llama", "num_attention_heads": 32, "num_hidden_layers": 32, "num_key_value_heads": 32, "pad_token_id": 0, "pretraining_tp": 1, "rms_norm_eps": 1e-06, "rope_scaling": null, "rope_theta": 10000.0, "tie_word_embeddings": false, "torch_dtype": "bfloat16", "transformers_version": "4.45.2", "use_cache": false, "vocab_size": 32000 }