ERROR: Could not find a version that satisfies the requirement torch==1.1.0 (from -r requirements.txt (line 6)) (from versions: 0.1.2, 0.1.2.post1, 0.1.2.post2)

[echosvile]

i have installed ffmepg, changed the version of the torch in requirements to 1.1.2 and its all still giving me the same 'could not find a version that satisfies the requirements torch==1.1.0'

'

[prajwalkr]

https://stackoverflow.com/a/57565285

Does this help?

[echosvile]

yes thank you. like one stage of everlasting problem passed. i had to torch==1.5.0 -f https://download.pytorch.org/whl/torch_stable.html then changed the torch in the requirements from torch==1.1.0 to 1.5.0 and then it loaded and brought this

[prajwalkr]

I think you should be able to run the code irrespective of those errors

[echosvile]

ok. i ran the code irrespective of the errors and i guess its running but its showing me this which i guess might be from the path i save the audio which is mp3 for the syncing in a folder named audio inside the Wav2lip file and the video its meant to sync named result which is mp4 inside a folder named video and this is the code i ran `python inference.py --checkpoint_path checkpoints/wav2lip.pth --face "../video/result.mp4" --audio "../audio/sound.mp3" and this is what i get

[echosvile]

and i already have installed pip install ffmpeg

[prajwalkr]

and i already have installed pip install ffmpeg this is not the right way to install ffmpeg. Please check how to install it on Windows.

[echosvile]

i ran these codes respectively conda install -c conda-forge ffmpeg' 'conda install -c conda-forge/label/gcc7 ffmpeg' 'conda install -c conda-forge/label/cf201901 ffmpeg' 'conda install -c conda-forge/label/cf202003 ffmpeg

[echosvile]

is this error now from me again?

[prajwalkr]

is this error now from me again?

Yes. Please follow the steps in the pre-requisites section of the README.

[echosvile]

finally it works.... i had to go to the file directory face_detection, detection, sfd and i renamed the s3fd.pth from s3fd-6789.pth to s3fd.pth and it ran to an extent and it stopped as shown in the attachment

[prajwalkr]

Your GPU memory is too less. Reduce the batch sizes of face detection and wav2lip. There are arguments to control that.

[echosvile]

please i beg you direct me, i don't understand this part

[prajwalkr]

Along with the other arguments you have used such as --face etc., please add --face_det_batch_size 2 and --wav2lip_batch_size 8.

[echosvile]

i.e python inference.py --checkpoint_path checkpoints/wav2lip.pth --face_det_batch_size 2 --wav2lip_batch_size 8"../video/result.mp4" --audio "../audio/sound.mp3"

[echosvile]

that didnt run please can you help me with the arrangement i don't know where to place the '--wav2lip_batch_size 8'

[prajwalkr]

8"../

add a space here between your arguments.

[echosvile]

python inference.py --checkpoint_path checkpoints/wav2lip.pth --face_det_batch_size 2 --wav2lip_batch_size 8 "../video/result.mp4" --audio "../audio/sound.mp3"

[prajwalkr]

python inference.py --checkpoint_path checkpoints/wav2lip.pth --face_det_batch_size 2 --wav2lip_batch_size 8 --face "../video/result.mp4" --audio "../audio/sound.mp3"

[echosvile]

showed me the same thing ![Uploading wav error 9.PNG…]()

[echosvile]

[Rudrabha]

python inference.py --checkpoint_path checkpoints/wav2lip.pth --face_det_batch_size 2 --wav2lip_batch_size 8 --face "../video/result.mp4" --audio "../audio/sound.mp3" Please try this once

[echosvile]

python inference.py --checkpoint_path checkpoints/wav2lip.pth --face_det_batch_size 2 --wav2lip_batch_size 8 --face "../video/result.mp4" --audio "../audio/sound.mp3"

[echosvile]

and its still in the same directory as it was when it ran the other time

[prajwalkr]

Please google and figure out how to specify the audio and video paths correctly. I do not know how to specify this properly in Windows.

[echosvile]

ok i used this no and its running i hope and pray it doesnt give error again python inference.py --checkpoint_path checkpoints/wav2lip.pth --face_det_batch_size 2 --wav2lip_batch_size 8 --face video/result.mp4 --audio audio/sound.mp3

[echosvile]

finally its done. very big thank you to you @prajwalkr and @Rudrabha Hat's off as i roll on the ground in thanks. Bless you!

[nishir53]

i have installed ffmepg, changed the version of the torch in requirements to 1.1.2 and its all still giving me the same 'could not find a version that satisfies the requirements torch==1.1.0'

'

hi there can you tell which version of python you have used and if possible can you help us with this project , we have this same project and we are getting issues in installing dependencies and running it

please please can you help?

[sahilg06]

@nishir53 Using python 3.9 and replacing all the == in req.txt with >= worked for me.

[nishir53]

@sahilg06 thanks but if you are free can we arrange google meet?

[nishir53]

also sir its taking too much time in loading

[nishir53]

@sahilg06 can you share your input video and input audio file please

[nishir53]

my files are having lot of frames please can you share your input video and input audio file

[sahilg06]

@nishir53 you can download video/audio for inference from here