Add config from convert_d_fine_original_pytorch_checkpoint_to_hf.py

config.json

@@ -1,10 +1,8 @@
 {
+  "_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,
@@ -28,47 +26,35 @@
       3,
       4
     ],
-    "stage_config": [
-      [
-        16,
-        16,
-        64,
-        1,
-        false,
-        false,
-        3,
-        3
-      ],
-      [
-        64,
-        32,
-        256,
-        1,
-        true,
-        false,
-        3,
-        3
-      ],
-      [
-        256,
-        64,
-        512,
-        2,
-        true,
-        true,
-        5,
-        3
-      ],
-      [
-        512,
-        128,
-        1024,
-        1,
-        true,
-        true,
-        5,
-        3
-      ]
+    "stage_in_channels": [
+      16,
+      64,
+      256,
+      512
+    ],
+    "stage_mid_channels": [
+      16,
+      32,
+      64,
+      128
+    ],
+    "stage_num_blocks": [
+      1,
+      1,
+      2,
+      1
+    ],
+    "stage_numb_of_layers": [
+      3,
+      3,
+      3,
+      3
+    ],
+    "stage_out_channels": [
+      64,
+      256,
+      512,
+      1024
     ],
     "stem_channels": [
       3,
@@ -308,8 +294,7 @@
   "positional_encoding_temperature": 10000,
   "reg_scale": 4.0,
   "top_prob_values": 4,
-  "torch_dtype": "float32",
-  "transformers_version": "4.49.0.dev0",
+  "transformers_version": "4.50.0.dev0",
   "use_focal_loss": true,
   "use_pretrained_backbone": false,
   "use_timm_backbone": false,