| { | |
| "architectures": [ | |
| "LlamaForScore" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bias": false, | |
| "bos_token_id": 1, | |
| "do_normalize": false, | |
| "eos_token_id": 2, | |
| "hidden_act": "silu", | |
| "hidden_size": 5120, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 13824, | |
| "max_position_embeddings": 2048, | |
| "mean": [ | |
| -1.53125 | |
| ], | |
| "model_type": "llama", | |
| "momentum": 0.9, | |
| "normalizer_type": "ExponentialMovingAverage", | |
| "num_attention_heads": 40, | |
| "num_hidden_layers": 40, | |
| "num_key_value_heads": 40, | |
| "pad_token_id": 32000, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": null, | |
| "rope_theta": 10000.0, | |
| "score_bias": true, | |
| "score_dim": 1, | |
| "score_type": "cost", | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.37.2", | |
| "use_cache": true, | |
| "var": [ | |
| 7.875 | |
| ], | |
| "vocab_size": 32001 | |
| } | |