Training in progress, step 200

config.json

@@ -1,40 +1,61 @@
 {
-  "_name_or_path": "hustvl/yolos-tiny",
+  "_name_or_path": "facebook/detr-resnet-101",
+  "activation_dropout": 0.0,
+  "activation_function": "relu",
   "architectures": [
-    "YolosForObjectDetection"
+    "DetrForObjectDetection"
   ],
-  "attention_probs_dropout_prob": 0.0,
+  "attention_dropout": 0.0,
   "auxiliary_loss": false,
+  "backbone": "resnet101",
+  "backbone_config": null,
+  "backbone_kwargs": {
+    "in_chans": 3,
+    "out_indices": [
+      1,
+      2,
+      3,
+      4
+    ]
+  },
   "bbox_cost": 5,
   "bbox_loss_coefficient": 5,
   "class_cost": 1,
+  "classifier_dropout": 0.0,
+  "d_model": 256,
+  "decoder_attention_heads": 8,
+  "decoder_ffn_dim": 2048,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "dice_loss_coefficient": 1,
+  "dilation": false,
+  "dropout": 0.1,
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 2048,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
   "eos_coefficient": 0.1,
   "giou_cost": 2,
   "giou_loss_coefficient": 2,
-  "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.0,
-  "hidden_size": 192,
   "id2label": {
     "0": "raccoon"
   },
-  "image_size": [
-    800,
-    1333
-  ],
-  "initializer_range": 0.02,
-  "intermediate_size": 768,
+  "init_std": 0.02,
+  "init_xavier_std": 1.0,
+  "is_encoder_decoder": true,
   "label2id": {
     "raccoon": 0
   },
-  "layer_norm_eps": 1e-12,
-  "model_type": "yolos",
-  "num_attention_heads": 3,
+  "mask_loss_coefficient": 1,
+  "max_position_embeddings": 1024,
+  "model_type": "detr",
   "num_channels": 3,
-  "num_detection_tokens": 100,
-  "num_hidden_layers": 12,
-  "patch_size": 16,
-  "qkv_bias": true,
+  "num_hidden_layers": 6,
+  "num_queries": 100,
+  "position_embedding_type": "sine",
+  "scale_embedding": false,
   "torch_dtype": "float32",
   "transformers_version": "4.48.3",
-  "use_mid_position_embeddings": false
+  "use_pretrained_backbone": true,
+  "use_timm_backbone": true
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f17bb476f54678e164f5757b05c2c2007c7412a24e573b94c45efc76dba0a1c6
-size 25909400
+oid sha256:679f5739fcc8008060eb5a004f1b3c502b5626842178b544df3e20c08fffedba
+size 242704552

preprocessor_config.json

@@ -10,7 +10,7 @@
     0.456,
     0.406
   ],
-  "image_processor_type": "YolosImageProcessor",
+  "image_processor_type": "DetrImageProcessor",
   "image_std": [
     0.229,
     0.224,
@@ -21,6 +21,6 @@
   "rescale_factor": 0.00392156862745098,
   "size": {
     "longest_edge": 1333,
-    "shortest_edge": 512
+    "shortest_edge": 800
   }
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4da47d8a90c8411f14878b11eb337a4730bc046c583db4795fb5df8feeab27a6
+oid sha256:86933fdba3076a9f888cb245c029a92232bd9dd05fc078c6eea5fdc191b66146
 size 5304