Upload config.json with huggingface_hub
Browse files- config.json +33 -5
config.json
CHANGED
|
@@ -1,17 +1,28 @@
|
|
| 1 |
{
|
| 2 |
-
"_commit_hash": null,
|
| 3 |
"architectures": [
|
| 4 |
"DepthAnythingForDepthEstimation"
|
| 5 |
],
|
| 6 |
"backbone": null,
|
| 7 |
"backbone_config": {
|
|
|
|
| 8 |
"architectures": [
|
| 9 |
"Dinov2Model"
|
| 10 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
"hidden_size": 768,
|
| 12 |
"image_size": 518,
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
"model_type": "dinov2",
|
| 14 |
"num_attention_heads": 12,
|
|
|
|
|
|
|
| 15 |
"out_features": [
|
| 16 |
"stage3",
|
| 17 |
"stage6",
|
|
@@ -25,16 +36,34 @@
|
|
| 25 |
12
|
| 26 |
],
|
| 27 |
"patch_size": 14,
|
|
|
|
| 28 |
"reshape_hidden_states": false,
|
| 29 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
},
|
| 31 |
"backbone_kwargs": null,
|
| 32 |
"depth_estimation_type": "metric",
|
|
|
|
| 33 |
"fusion_hidden_size": 128,
|
| 34 |
"head_hidden_size": 32,
|
| 35 |
"head_in_index": -1,
|
| 36 |
"initializer_range": 0.02,
|
| 37 |
-
"max_depth":
|
| 38 |
"model_type": "depth_anything",
|
| 39 |
"neck_hidden_sizes": [
|
| 40 |
96,
|
|
@@ -50,8 +79,7 @@
|
|
| 50 |
0.5
|
| 51 |
],
|
| 52 |
"reassemble_hidden_size": 768,
|
| 53 |
-
"
|
| 54 |
-
"transformers_version": null,
|
| 55 |
"use_pretrained_backbone": false,
|
| 56 |
"use_timm_backbone": false
|
| 57 |
}
|
|
|
|
| 1 |
{
|
|
|
|
| 2 |
"architectures": [
|
| 3 |
"DepthAnythingForDepthEstimation"
|
| 4 |
],
|
| 5 |
"backbone": null,
|
| 6 |
"backbone_config": {
|
| 7 |
+
"apply_layernorm": true,
|
| 8 |
"architectures": [
|
| 9 |
"Dinov2Model"
|
| 10 |
],
|
| 11 |
+
"attention_probs_dropout_prob": 0.0,
|
| 12 |
+
"drop_path_rate": 0.0,
|
| 13 |
+
"dtype": "float32",
|
| 14 |
+
"hidden_act": "gelu",
|
| 15 |
+
"hidden_dropout_prob": 0.0,
|
| 16 |
"hidden_size": 768,
|
| 17 |
"image_size": 518,
|
| 18 |
+
"initializer_range": 0.02,
|
| 19 |
+
"layer_norm_eps": 1e-06,
|
| 20 |
+
"layerscale_value": 1.0,
|
| 21 |
+
"mlp_ratio": 4,
|
| 22 |
"model_type": "dinov2",
|
| 23 |
"num_attention_heads": 12,
|
| 24 |
+
"num_channels": 3,
|
| 25 |
+
"num_hidden_layers": 12,
|
| 26 |
"out_features": [
|
| 27 |
"stage3",
|
| 28 |
"stage6",
|
|
|
|
| 36 |
12
|
| 37 |
],
|
| 38 |
"patch_size": 14,
|
| 39 |
+
"qkv_bias": true,
|
| 40 |
"reshape_hidden_states": false,
|
| 41 |
+
"stage_names": [
|
| 42 |
+
"stem",
|
| 43 |
+
"stage1",
|
| 44 |
+
"stage2",
|
| 45 |
+
"stage3",
|
| 46 |
+
"stage4",
|
| 47 |
+
"stage5",
|
| 48 |
+
"stage6",
|
| 49 |
+
"stage7",
|
| 50 |
+
"stage8",
|
| 51 |
+
"stage9",
|
| 52 |
+
"stage10",
|
| 53 |
+
"stage11",
|
| 54 |
+
"stage12"
|
| 55 |
+
],
|
| 56 |
+
"use_mask_token": true,
|
| 57 |
+
"use_swiglu_ffn": false
|
| 58 |
},
|
| 59 |
"backbone_kwargs": null,
|
| 60 |
"depth_estimation_type": "metric",
|
| 61 |
+
"dtype": "float32",
|
| 62 |
"fusion_hidden_size": 128,
|
| 63 |
"head_hidden_size": 32,
|
| 64 |
"head_in_index": -1,
|
| 65 |
"initializer_range": 0.02,
|
| 66 |
+
"max_depth": 20,
|
| 67 |
"model_type": "depth_anything",
|
| 68 |
"neck_hidden_sizes": [
|
| 69 |
96,
|
|
|
|
| 79 |
0.5
|
| 80 |
],
|
| 81 |
"reassemble_hidden_size": 768,
|
| 82 |
+
"transformers_version": "5.0.0.dev0",
|
|
|
|
| 83 |
"use_pretrained_backbone": false,
|
| 84 |
"use_timm_backbone": false
|
| 85 |
}
|