Training in progress, epoch 0

config.json

@@ -1,12 +1,10 @@
 {
-  "_name_or_path": "facebook/timesformer-base-finetuned-k400",
+  "_name_or_path": "google/vivit-b-16x2-kinetics400",
   "architectures": [
-    "TimesformerForVideoClassification"
+    "VivitForVideoClassification"
   ],
   "attention_probs_dropout_prob": 0.0,
-  "attention_type": "divided_space_time",
-  "drop_path_rate": 0,
-  "hidden_act": "gelu",
+  "hidden_act": "gelu_fast",
   "hidden_dropout_prob": 0.0,
   "hidden_size": 768,
   "id2label": {
@@ -21,14 +19,22 @@
     "Normal": 1
   },
   "layer_norm_eps": 1e-06,
-  "model_type": "timesformer",
+  "model_type": "vivit",
   "num_attention_heads": 12,
   "num_channels": 3,
-  "num_frames": 8,
+  "num_frames": 32,
   "num_hidden_layers": 12,
-  "patch_size": 16,
-  "problem_type": "single_label_classification",
   "qkv_bias": true,
   "torch_dtype": "float32",
-  "transformers_version": "4.46.2"
+  "transformers_version": "4.46.2",
+  "tubelet_size": [
+    2,
+    16,
+    16
+  ],
+  "video_size": [
+    32,
+    224,
+    224
+  ]
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e13013a2ac84f1166e04bedf4790ae9531ce6d97f18da2822009ca4d57eaa770
-size 485072256
+oid sha256:f6639d1f6994799d5439f4ee122b678578c63eed52ee23c7f916b913e4e8e392
+size 354615376

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:409dc7690a1b40cd5006f55398f540e9d87e4b8b85467159b8e010c39e6789d9
+oid sha256:cdb8bb6b0772209f95dd99dcd2966e8f2c0529e6d4d6b3f47ed830e6ae271b6f
 size 5304