| { | |
| "architectures": [ | |
| "InternVLVisionModel" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "dtype": "float32", | |
| "hidden_act": "quick_gelu", | |
| "hidden_dropout_prob": 0.0, | |
| "hidden_size": 32, | |
| "image_size": [ | |
| 64, | |
| 64 | |
| ], | |
| "initializer_range": 0.02, | |
| "intermediate_size": 128, | |
| "layer_norm_eps": 1e-06, | |
| "layer_scale_init_value": 0.1, | |
| "model_type": "internvl_vision", | |
| "norm_type": "layer_norm", | |
| "num_attention_heads": 4, | |
| "num_channels": 3, | |
| "num_hidden_layers": 2, | |
| "patch_size": [ | |
| 4, | |
| 4 | |
| ], | |
| "projection_dropout": 0.0, | |
| "transformers_version": "5.16.0.dev0", | |
| "use_absolute_position_embeddings": true, | |
| "use_mask_token": false, | |
| "use_mean_pooling": true, | |
| "use_qk_norm": false | |
| } | |