| { | |
| "_name_or_path": "models/deepsex-34b-4bpw-h8-exl2", | |
| "architectures": [ | |
| "YiForCausalLM" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_yi.YiConfig", | |
| "AutoModel": "modeling_yi.YiForCausalLM", | |
| "AutoModelForCausalLM": "modeling_yi.YiForCausalLM" | |
| }, | |
| "attention_bias": false, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "hidden_act": "silu", | |
| "hidden_size": 7168, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 20480, | |
| "max_position_embeddings": 4096, | |
| "model_type": "Yi", | |
| "num_attention_heads": 56, | |
| "num_hidden_layers": 60, | |
| "num_key_value_heads": 8, | |
| "pad_token_id": 0, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_theta": 5000000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.34.1", | |
| "use_cache": true, | |
| "vocab_size": 64000 | |
| } | |