Training in progress, epoch 0

config.json

@@ -1,55 +1,55 @@
-{
-  "_name_or_path": "MCG-NJU/videomae-base",
-  "architectures": [
-    "VideoMAEForVideoClassification"
-  ],
-  "attention_probs_dropout_prob": 0.0,
-  "decoder_hidden_size": 384,
-  "decoder_intermediate_size": 1536,
-  "decoder_num_attention_heads": 6,
-  "decoder_num_hidden_layers": 4,
-  "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.0,
-  "hidden_size": 768,
-  "id2label": {
-    "0": "ApplyEyeMakeup",
-    "1": "ApplyLipstick",
-    "2": "Archery",
-    "3": "BabyCrawling",
-    "4": "BalanceBeam",
-    "5": "BandMarching",
-    "6": "BaseballPitch",
-    "7": "Basketball",
-    "8": "BasketballDunk",
-    "9": "BenchPress"
-  },
-  "image_size": 224,
-  "initializer_range": 0.02,
-  "intermediate_size": 3072,
-  "label2id": {
-    "ApplyEyeMakeup": 0,
-    "ApplyLipstick": 1,
-    "Archery": 2,
-    "BabyCrawling": 3,
-    "BalanceBeam": 4,
-    "BandMarching": 5,
-    "BaseballPitch": 6,
-    "Basketball": 7,
-    "BasketballDunk": 8,
-    "BenchPress": 9
-  },
-  "layer_norm_eps": 1e-12,
-  "model_type": "videomae",
-  "norm_pix_loss": true,
-  "num_attention_heads": 12,
-  "num_channels": 3,
-  "num_frames": 16,
-  "num_hidden_layers": 12,
-  "patch_size": 16,
-  "problem_type": "single_label_classification",
-  "qkv_bias": true,
-  "torch_dtype": "float32",
-  "transformers_version": "4.39.3",
-  "tubelet_size": 2,
-  "use_mean_pooling": false
-}
+{
+  "_name_or_path": "MCG-NJU/videomae-base",
+  "architectures": [
+    "VideoMAEForVideoClassification"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "decoder_hidden_size": 384,
+  "decoder_intermediate_size": 1536,
+  "decoder_num_attention_heads": 6,
+  "decoder_num_hidden_layers": 4,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "ApplyEyeMakeup",
+    "1": "Archery",
+    "2": "BalanceBeam",
+    "3": "BaseballPitch",
+    "4": "BenchPress",
+    "5": "BlowDryHair",
+    "6": "Bowling",
+    "7": "CleanAndJerk",
+    "8": "CliffDiving",
+    "9": "Diving"
+  },
+  "image_size": 224,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "ApplyEyeMakeup": 0,
+    "Archery": 1,
+    "BalanceBeam": 2,
+    "BaseballPitch": 3,
+    "BenchPress": 4,
+    "BlowDryHair": 5,
+    "Bowling": 6,
+    "CleanAndJerk": 7,
+    "CliffDiving": 8,
+    "Diving": 9
+  },
+  "layer_norm_eps": 1e-12,
+  "model_type": "videomae",
+  "norm_pix_loss": true,
+  "num_attention_heads": 12,
+  "num_channels": 3,
+  "num_frames": 16,
+  "num_hidden_layers": 12,
+  "patch_size": 16,
+  "problem_type": "single_label_classification",
+  "qkv_bias": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.46.3",
+  "tubelet_size": 2,
+  "use_mean_pooling": false
+}

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d342a3fb08d5739ff734870398f6647655c4a54ce90cfaea3bf63418a1f8fc5a
+oid sha256:fa1eb5d1028f3a897d6cd6dac428599e54618a60183bafd8599539a8953506d6
 size 344961984

preprocessor_config.json

@@ -1,20 +1,4 @@
 {
-  "_valid_processor_keys": [
-    "videos",
-    "do_resize",
-    "size",
-    "resample",
-    "do_center_crop",
-    "crop_size",
-    "do_rescale",
-    "rescale_factor",
-    "do_normalize",
-    "image_mean",
-    "image_std",
-    "return_tensors",
-    "data_format",
-    "input_data_format"
-  ],
   "crop_size": {
     "height": 224,
     "width": 224

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c1fa45dba6da1397ed5c81fd8c81e85571dc7a62a3dc7e188136765c5d2fffb6
-size 4984
+oid sha256:14091fab5574c11d83769e12807309d8a7eab9cf4798fa88641f9a3f85d91fc5
+size 5304