config.json
| 1 | { |
| 2 | "_commit_hash": null, |
| 3 | "architectures": [ |
| 4 | "DepthAnythingForDepthEstimation" |
| 5 | ], |
| 6 | "backbone": null, |
| 7 | "backbone_config": { |
| 8 | "architectures": [ |
| 9 | "Dinov2Model" |
| 10 | ], |
| 11 | "hidden_size": 1024, |
| 12 | "image_size": 518, |
| 13 | "model_type": "dinov2", |
| 14 | "num_attention_heads": 16, |
| 15 | "num_hidden_layers": 24, |
| 16 | "out_features": [ |
| 17 | "stage21", |
| 18 | "stage22", |
| 19 | "stage23", |
| 20 | "stage24" |
| 21 | ], |
| 22 | "out_indices": [ |
| 23 | 21, |
| 24 | 22, |
| 25 | 23, |
| 26 | 24 |
| 27 | ], |
| 28 | "patch_size": 14, |
| 29 | "reshape_hidden_states": false, |
| 30 | "stage_names": [ |
| 31 | "stem", |
| 32 | "stage1", |
| 33 | "stage2", |
| 34 | "stage3", |
| 35 | "stage4", |
| 36 | "stage5", |
| 37 | "stage6", |
| 38 | "stage7", |
| 39 | "stage8", |
| 40 | "stage9", |
| 41 | "stage10", |
| 42 | "stage11", |
| 43 | "stage12", |
| 44 | "stage13", |
| 45 | "stage14", |
| 46 | "stage15", |
| 47 | "stage16", |
| 48 | "stage17", |
| 49 | "stage18", |
| 50 | "stage19", |
| 51 | "stage20", |
| 52 | "stage21", |
| 53 | "stage22", |
| 54 | "stage23", |
| 55 | "stage24" |
| 56 | ], |
| 57 | "torch_dtype": "float32" |
| 58 | }, |
| 59 | "fusion_hidden_size": 256, |
| 60 | "head_hidden_size": 32, |
| 61 | "head_in_index": -1, |
| 62 | "initializer_range": 0.02, |
| 63 | "model_type": "depth_anything", |
| 64 | "neck_hidden_sizes": [ |
| 65 | 256, |
| 66 | 512, |
| 67 | 1024, |
| 68 | 1024 |
| 69 | ], |
| 70 | "patch_size": 14, |
| 71 | "reassemble_factors": [ |
| 72 | 4, |
| 73 | 2, |
| 74 | 1, |
| 75 | 0.5 |
| 76 | ], |
| 77 | "reassemble_hidden_size": 1024, |
| 78 | "torch_dtype": "float32", |
| 79 | "transformers_version": null, |
| 80 | "use_pretrained_backbone": false |
| 81 | } |
| 82 | |