ollama - 💡(How to fix) Fix 为什么2月20日后的版本，都不再支持MP4.AVI 等视频格式？ [1 participants]

Fix Plan

To restore video functionality, we need to re-enable video support in the updated models.

Step-by-Step Solution

• Check the model configurations for qwen3.5, 35B, 27B, and VL to ensure they have video support enabled.
• Update the model configurations to include video support by adding the following code:

model_config = {
    'qwen3.5': {'video_support': True},
    '35B': {'video_support': True},
    '27B': {'video_support': True},
    'VL': {'video_support': True}
}

• Modify the image processing function to handle video files by adding the following code:

import cv2

def process_image(file_path):
    if file_path.endswith(('.mp4', '.avi', '.mov')):
        # Handle video file
        cap = cv2.VideoCapture(file_path)
        while cap.isOpened():
            ret, frame = cap.read()
            if not ret:
                break
            # Process frame
            yield frame
    else:
        # Handle image file
        img = cv2.imread(file_path)
        yield img

• Update the model loading code to use the updated configurations:

from transformers import AutoModel

model = AutoModel.from_pretrained('qwen3.5', config=model_config['qwen3.5'])

Verification

Test the updated models with video files to ensure they can process video inputs correctly.

Extra Tips

• Make sure to update the model configurations and code to handle video files correctly.
• Test the updated models with different video formats to ensure compatibility.