| { | |
| "_name_or_path": "hermes-2-pro-mistral-418m-pruned", | |
| "apply_residual_connection_post_layernorm": false, | |
| "architectures": [ | |
| "MistralForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 1, | |
| "eos_token_id": 32000, | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 1152, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 4608, | |
| "layer_norm_epsilon": 1e-05, | |
| "max_position_embeddings": 32768, | |
| "model_type": "mistral", | |
| "num_attention_heads": 18, | |
| "num_hidden_layers": 13, | |
| "num_key_value_heads": 18, | |
| "quantization_aware_training": true, | |
| "rms_norm_eps": 1e-05, | |
| "rope_theta": 10000.0, | |
| "sliding_window": 4096, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "float16", | |
| "transformers_version": "4.44.2", | |
| "use_cache": true, | |
| "use_early_exiting": true, | |
| "use_reversible_layers": true, | |
| "vocab_size": 32032 | |
| } | |