| { |
| "add_projection": false, |
| "architectures": [ |
| "DPTForDepthEstimation" |
| ], |
| "attention_probs_dropout_prob": 0.1, |
| "auxiliary_loss_weight": 0.4, |
| "backbone_config": null, |
| "backbone_featmap_shape": null, |
| "backbone_out_indices": [ |
| 0, |
| 1, |
| 2, |
| 3 |
| ], |
| "dtype": "float32", |
| "fusion_hidden_size": 32, |
| "head_in_index": -1, |
| "hidden_act": "gelu", |
| "hidden_dropout_prob": 0.1, |
| "hidden_size": 32, |
| "image_size": 32, |
| "initializer_range": 0.02, |
| "intermediate_size": 37, |
| "is_decoder": false, |
| "is_hybrid": false, |
| "layer_norm_eps": 1e-12, |
| "model_type": "dpt", |
| "neck_hidden_sizes": [ |
| 16, |
| 32 |
| ], |
| "neck_ignore_stages": [], |
| "num_attention_heads": 4, |
| "num_channels": 3, |
| "num_hidden_layers": 2, |
| "patch_size": 16, |
| "pooler_act": "tanh", |
| "pooler_output_size": 32, |
| "qkv_bias": true, |
| "readout_type": "project", |
| "reassemble_factors": [ |
| 4, |
| 2, |
| 1, |
| 0.5 |
| ], |
| "semantic_classifier_dropout": 0.1, |
| "semantic_loss_ignore_index": 255, |
| "transformers_version": "5.16.0.dev0", |
| "use_auxiliary_head": true, |
| "use_batch_norm_in_fusion_residual": false, |
| "use_bias_in_fusion_residual": null |
| } |
|
|
