| { | |
| "model_type": "mistral", | |
| "architectures": ["MistralForCausalLM"], | |
| "hidden_size": 4096, | |
| "num_attention_heads": 16, | |
| "num_hidden_layers": 32, | |
| "intermediate_size": 11008, | |
| "max_position_embeddings": 2048, | |
| "layer_norm_eps": 1e-5, | |
| "initializer_range": 0.02, | |
| "use_cache": true, | |
| "vocab_size": 50257, | |
| "tokenizer_class": "GPT2Tokenizer" | |
| } | |