| { | |
| "architectures": [ | |
| "XmodelForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_xmodel2.XmodelConfig", | |
| "AutoModelForCausalLM": "modeling_xmodel2.XmodelForCausalLM" | |
| }, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "hidden_act": "silu", | |
| "hidden_size": 1536, | |
| "initializer_range": 0.1, | |
| "intermediate_size": 3840, | |
| "max_position_embeddings": 131072, | |
| "mlp_bias": false, | |
| "model_type": "xmodel", | |
| "mup_attention_residual_scale": 1.4, | |
| "mup_base_width": 256, | |
| "mup_ffn_residual_scale": 1.4, | |
| "mup_input_scale": 12.0, | |
| "mup_output_scale": 1.0, | |
| "mup_width_multiplier": 6.0, | |
| "num_attention_heads": 24, | |
| "num_hidden_layers": 48, | |
| "num_key_value_heads": 8, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": null, | |
| "rope_theta": 500000, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.53.3", | |
| "use_cache": true, | |
| "use_mup": true, | |
| "vocab_size": 129280, | |
| "_attn_implementation": "flash_attention_2" | |
| } |