| { | |
| "architectures": [ | |
| "HunyuanViTModel" | |
| ], | |
| "_attn_implementation": "flash_attention_2", | |
| "initializer_factor": 1.0, | |
| "model_type": "HunyuanViT", | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.52.4", | |
| "vision_config": { | |
| "attention_dropout": 0.0, | |
| "hidden_act": "gelu_pytorch_tanh", | |
| "hidden_size": 1536, | |
| "intermediate_size": 6144, | |
| "layer_norm_eps": 1e-06, | |
| "model_type": "HunyuanViT_vision_model", | |
| "num_attention_heads": 16, | |
| "num_channels": 3, | |
| "num_hidden_layers": 40, | |
| "num_patches": 16384, | |
| "patch_size": 16 | |
| } | |
| } | |