| { | |
| "architectures": [ | |
| "MOASpecModel" | |
| ], | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 4096, | |
| "max_position_embeddings": 8192, | |
| "mlp_bias": false, | |
| "model_type": "moa_spec", | |
| "num_attention_heads": 32, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": null, | |
| "rope_theta": 500000.0, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.45.2", | |
| "vocab_size": 128256, | |
| "self_attention_num_key_value_heads": 4, | |
| "self_attention_intermediate_size": 6144, | |
| "cross_attention_num_key_value_heads": 4, | |
| "cross_attention_intermediate_size": 6144, | |
| "disable_LSA": false, | |
| "kv_hidden_size": 2048, | |
| "layer_self_attention_num_key_value_heads": 8, | |
| "layer_self_attention_intermediate_size": 3072 | |
| } |