Update app.py

app.py

@@ -1,6 +1,6 @@
 import cv2
-import gradio as gr
 import imutils
+import gradio as gr
 import numpy as np
 import torch
 from pytorchvideo.transforms import (
@@ -51,8 +51,7 @@ def parse_video(video_file):
         prop = (
             cv2.cv.CV_CAP_PROP_FRAME_COUNT
             if imutils.is_cv2()
-            else cv2.CAP_PROP_FRAME_COUNT
-        )
+            else cv2.CAP_PROP_FRAME_COUNT)
         total = int(vs.get(prop))
         print("[INFO] {} total frames in video".format(total))
 
@@ -91,13 +90,9 @@ def preprocess_video(frames: list):
     # to make it compatible with the model. Finally, we add a batch dimension so that our video
     # classification model can operate on it.
     video_tensor = torch.tensor(np.array(frames).astype(frames[0].dtype))
-    video_tensor = video_tensor.permute(
-        3, 0, 1, 2
-    )  # (num_channels, num_frames, height, width)
+    video_tensor = video_tensor.permute(3, 0, 1, 2)  # (num_channels, num_frames, height, width)
     video_tensor_pp = VAL_TRANSFORMS(video_tensor)
-    video_tensor_pp = video_tensor_pp.permute(
-        1, 0, 2, 3
-    )  # (num_frames, num_channels, height, width)
+    video_tensor_pp = video_tensor_pp.permute(1, 0, 2, 3)  # (num_frames, num_channels, height, width)
     video_tensor_pp = video_tensor_pp.unsqueeze(0)
     return video_tensor_pp.to(DEVICE)
 
@@ -115,16 +110,10 @@ def infer(video_file):
     confidences = {LABELS[i]: float(softmax_scores[i]) for i in range(len(LABELS))}
     return confidences
 
-
 gr.Interface(
     fn=infer,
     inputs=gr.Video(type="file"),
     outputs=gr.Label(num_top_classes=3),
-    examples=[
-        ["examples/babycrawling.mp4"],
-        ["examples/baseball.mp4"],
-        ["examples/balancebeam.mp4"],
-    ],
     title="VideoMAE fine-tuned on a subset of UCF-101",
     description=(
         "Gradio demo for VideoMAE for video classification. To use it, simply upload your video or click one of the"