| { | |
| "_name_or_path": "/Users/anshulachowdhury/Repo/SREDWiseAlgo/output/run_20241214_182038", | |
| "_remove_final_layer_norm": false, | |
| "activation_dropout": 0.0, | |
| "activation_function": "relu", | |
| "architectures": [ | |
| "MistralForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 2, | |
| "do_layer_norm_before": true, | |
| "dropout": 0.1, | |
| "enable_bias": true, | |
| "eos_token_id": 2, | |
| "ffn_dim": 3072, | |
| "hidden_act": "silu", | |
| "hidden_size": 768, | |
| "init_std": 0.02, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 14336, | |
| "layer_norm_elementwise_affine": true, | |
| "layerdrop": 0.0, | |
| "max_position_embeddings": 2048, | |
| "model_type": "mistral", | |
| "num_attention_heads": 12, | |
| "num_hidden_layers": 12, | |
| "num_key_value_heads": 8, | |
| "pad_token_id": 1, | |
| "prefix": "</s>", | |
| "rms_norm_eps": 1e-06, | |
| "rope_theta": 10000.0, | |
| "sliding_window": 4096, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.34.0", | |
| "use_cache": true, | |
| "vocab_size": 50272, | |
| "word_embed_proj_dim": 768 | |
| } | |