| { | |
| "architectures": [ | |
| "AMOEModel" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_amoe.AMOEConfig", | |
| "AutoImageProcessor": "image_processing_amoe.AMOEImageProcessor", | |
| "AutoModel": "modeling_amoe.AMOEModel" | |
| }, | |
| "channel_size": 3, | |
| "dim": 768, | |
| "dtype": "float32", | |
| "enable_3d_rope": true, | |
| "head_dim": 128, | |
| "max_seq_len": 8192, | |
| "model_type": "amoe", | |
| "moe_args": { | |
| "activation": "silu", | |
| "num_experts": 28, | |
| "num_shared_experts": 0, | |
| "route_norm": false, | |
| "route_scale": 1.0, | |
| "score_before_experts": false, | |
| "score_func": "sigmoid", | |
| "top_k": 6 | |
| }, | |
| "moe_dim": 384, | |
| "n_heads": 8, | |
| "n_kv_heads": 8, | |
| "n_layers": 18, | |
| "n_storage_tokens": 4, | |
| "norm_eps": 1e-05, | |
| "rope_max_freqs": 20.0, | |
| "rope_min_freqs": 1.0, | |
| "rope_theta": 100000.0, | |
| "spatial_patch_size": 16, | |
| "teachers": [ | |
| "siglip2", | |
| "dinov3" | |
| ], | |
| "teachers_dim": [ | |
| 1152, | |
| 1024 | |
| ], | |
| "temporal_patch_size": 1, | |
| "transformers_version": "4.57.1", | |
| "use_flex_attn": true, | |
| "use_qk_norm": true, | |
| "use_tok_norm": true | |
| } | |