Add config from convert_d_fine_original_pytorch_checkpoint_to_hf.py

config.json

@@ -1,17 +1,31 @@
 {
+  "_attn_implementation_autoset": true,
   "activation_dropout": 0.0,
   "activation_function": "silu",
   "anchor_image_size": null,
-  "architectures": [
-    "DFineForObjectDetection"
-  ],
   "attention_dropout": 0.0,
   "auxiliary_loss": true,
   "backbone": null,
   "backbone_config": {
+    "depths": [
+      3,
+      4,
+      6,
+      3
+    ],
+    "downsample_in_bottleneck": false,
+    "downsample_in_first_stage": false,
     "embedding_size": 32,
+    "hidden_act": "relu",
+    "hidden_sizes": [
+      256,
+      512,
+      1024,
+      2048
+    ],
     "layer_type": "basic",
-    "model_type": "d_fine_resnet",
+    "model_type": "hgnet_v2",
+    "num_channels": 3,
     "out_features": [
       "stage2",
       "stage3",
@@ -21,7 +35,68 @@
       2,
       3,
       4
-    ]
+    ],
+    "stage_downsample": [
+      false,
+      true,
+      true,
+      true
+    ],
+    "stage_in_channels": [
+      48,
+      128,
+      512,
+      1024
+    ],
+    "stage_kernel_size": [
+      3,
+      3,
+      5,
+      5
+    ],
+    "stage_light_block": [
+      false,
+      false,
+      true,
+      true
+    ],
+    "stage_mid_channels": [
+      48,
+      96,
+      192,
+      384
+    ],
+    "stage_names": [
+      "stem",
+      "stage1",
+      "stage2",
+      "stage3",
+      "stage4"
+    ],
+    "stage_num_blocks": [
+      1,
+      1,
+      3,
+      1
+    ],
+    "stage_numb_of_layers": [
+      6,
+      6,
+      6,
+      6
+    ],
+    "stage_out_channels": [
+      128,
+      512,
+      1024,
+      2048
+    ],
+    "stem_channels": [
+      3,
+      32,
+      48
+    ],
+    "use_learnable_affine_block": false
   },
   "backbone_kwargs": null,
   "batch_norm_eps": 1e-05,
@@ -44,7 +119,6 @@
   ],
   "decoder_offset_scale": 0.5,
   "depth_mult": 1.0,
-  "disable_custom_kernels": true,
   "dropout": 0.0,
   "encode_proj_layers": [
     2
@@ -258,7 +332,6 @@
   "positional_encoding_temperature": 10000,
   "reg_scale": 4.0,
   "top_prob_values": 4,
-  "torch_dtype": "float32",
   "transformers_version": "4.50.0.dev0",
   "use_focal_loss": true,
   "use_pretrained_backbone": false,