CUDA out of memory about video-retalking HOT 8 OPEN

You can try reducing these these two parameters: --face_det_batch_size and --LNet_batch_size.
Or you can run the code in colab: https://colab.research.google.com/github/vinthony/video-retalking/blob/main/quick_demo.ipynb

from video-retalking.

But there is no --face_det_batch_size and --LNet_batch_size in the inference.py

from video-retalking.

You can try reducing these these two parameters: --face_det_batch_size and --LNet_batch_size. Or you can run the code in colab: https://colab.research.google.com/github/vinthony/video-retalking/blob/main/quick_demo.ipynb

Excuse me, I have the same problem, cuda out of memory. Where are these two parameters? I'd like to reduce them. thx.

from video-retalking.

I met the same error,how resolve it ?

from video-retalking.

I met the same error,how resolve it ?

You can try smaller batch size:

python3 inference.py \
  --face examples/face/1.mp4 \
  --audio examples/audio/1.wav \
  --outfile results/1_1.mp4 \
  --face_det_batch_size 2 --LNet_batch_size 2

from video-retalking.

utils/inference_utils.py
here

from video-retalking.

I modified two values of batch_size. My video card is V100 with 32G of video memory, but it still bursts. Does the program need to be optimized?

from video-retalking.

it's ok to running on 8GB vram machine. you just need to adjust the input video parameter to reduce the consumption of vram to prevent the issue of "out of cuda memory". example for reduce Sampling Rate of video, remove the sound of input video, choose "CBR" for input video.

from video-retalking.