Use model loader from vllm (#459)
This commit is contained in:
@@ -141,7 +141,7 @@ def encode_frame(frame):
|
||||
|
||||
|
||||
def encode_video_base64(video_path, num_frames=16):
|
||||
import cv2
|
||||
import cv2 # pip install opencv-python-headless
|
||||
|
||||
cap = cv2.VideoCapture(video_path)
|
||||
if not cap.isOpened():
|
||||
|
||||
Reference in New Issue
Block a user