If you run it locally, does it improve output quality by any chance? or no matter what, every output of the LivePortrait model will be sort of downgraded/compressed/pixelated?
Hi. I need help. Why the last command (python app.py) give me this error? File "C:\Users\Utente\Desktop\Live Potrait\LivePortrait\app.py", line 12, in from src.utils.helper import load_description File "C:\Users\Utente\Desktop\Live Potrait\LivePortrait\src\utils\helper.py", line 9, in import torch ModuleNotFoundError: No module named 'torch'
You need to properly install torch. If you have an NVIDIA card you can google “how to install torch with CUDA support”, its a bit too long to explain here.
This might be the right version, try this command after youve activated the venv: pip3 install torch torchvision torchaudio --index-url download.pytorch.org/whl/cu118
Nice video. Question: If I want to stop the outputted video from following the eyes from my driving video. How to I do it? The guy in my video is looking up all of the time reading lyrics. Thanks.
Hi, i am searching for a model for comfy ui or others ai local pc software that i give image + audio file and it give me a video that is avatar speaker. do you know any model of this? can you make a tutorial for that?
@@AIQuestOfficial Thanks bro but it wasn't what i am searching for. because it needs 2 video or 1 img + a driving video while i need a model that give it 1 img + 1 audio file and make it live portrait in output in 1 video.
@@AIQuestOfficial RuntimeError: CUDA error: the launch timed out and was terminated That's the error. (System is running GTX 950 on AMD Ryzen 5) I've been using the site, so that's okay. But it has serious trouble getting dogs to sing... It's a bummer because I'm trying exactly that for a music video. I'll keep trying tho... 🙂
This is great, except Python 10 or 11 is no longer available for download so I installed the latest which is 3.12. In your instructions, Step 3 create a virtual environment, when pasting your cmd prompt, I get a response that states "Python was not found". Any help you might offer is greatly appreciated. Tom
Hey, if you can, it would be awesome to see a quick video of you doing this with pinokio. Your guidelines are far superior to Pinocchio's official instructions.
When I run in step 8, there's an error "no module named torch" (venv) C:\LivePortrait\LivePortrait>.\venv\Scripts\activate (venv) C:\LivePortrait\LivePortrait>python app.py Traceback (most recent call last): File "C:\LivePortrait\LivePortrait\app.py", line 12, in from src.utils.helper import load_description File "C:\LivePortrait\LivePortrait\src\utils\helper.py", line 9, in import torch ModuleNotFoundError: No module named 'torch'
@@AIQuestOfficial there was the same error. I entered "pip install torch", it installed, but then the error "AssertionError: Torch not compiled with CUDA enabled" appears
Ah yes so you need the right torch for your Cuda, I can’t really explain it all here but if you google “How to install torch with Cuda support” you’ll find a guide. If you want, you can try this command though, its usually the right one: pip3 install torch torchvision torchaudio --index-url download.pytorch.org/whl/cu118
Hmm I am getting these errors when running the install that you said you already did so skipped. × Preparing metadata (pyproject.toml) did not run successfully. │ exit code: 1 ╰─> [21 lines of output] note: This error originates from a subprocess, and is likely not a problem with pip. error: metadata-generation-failed × Encountered error while generating package metadata. ╰─> See above for output. note: This is an issue with the package mentioned above, not pip. hint: See above for details.
This is so awesome - if it only was FULL BODY, or at least face, hands and upper body.
I know! It’s coming soon ;)
can it be done for Video to Video ? like for source instead of portrait, it can use video as source.
It is possible yes, people are doing it, you should give it a google
If you run it locally, does it improve output quality by any chance? or no matter what, every output of the LivePortrait model will be sort of downgraded/compressed/pixelated?
@@IacobPastina196 it shouldn’t improve the output alas
@@AIQuestOfficial so our best bet so far for this would be an upscaler?
Hi. I need help. Why the last command (python app.py) give me this error?
File "C:\Users\Utente\Desktop\Live Potrait\LivePortrait\app.py", line 12, in
from src.utils.helper import load_description
File "C:\Users\Utente\Desktop\Live Potrait\LivePortrait\src\utils\helper.py", line 9, in
import torch
ModuleNotFoundError: No module named 'torch'
You need to properly install torch. If you have an NVIDIA card you can google “how to install torch with CUDA support”, its a bit too long to explain here.
@@AIQuestOfficial I try. Many thanks
@@AIQuestOfficial I can't understand :')
This might be the right version, try this command after youve activated the venv:
pip3 install torch torchvision torchaudio --index-url download.pytorch.org/whl/cu118
@@AIQuestOfficial Really thanks for your disposability. I'll try
Nice video. Question: If I want to stop the outputted video from following the eyes from my driving video. How to I do it? The guy in my video is looking up all of the time reading lyrics. Thanks.
Hi, i am searching for a model for comfy ui or others ai local pc software that i give image + audio file and it give me a video that is avatar speaker.
do you know any model of this? can you make a tutorial for that?
Check the video i uploaded after this one
@@AIQuestOfficial Thanks bro but it wasn't what i am searching for. because it needs 2 video or 1 img + a driving video while i need a model that give it 1 img + 1 audio file and make it live portrait in output in 1 video.
torch not compiled with cuda enabled
Such a bummer that with CUDA errors the home install is next to impossible to get to work.
Maybe next time...
Which errors are you getting friend?
@@AIQuestOfficial RuntimeError: CUDA error: the launch timed out and was terminated
That's the error. (System is running GTX 950 on AMD Ryzen 5)
I've been using the site, so that's okay. But it has serious trouble getting dogs to sing... It's a bummer because I'm trying exactly that for a music video.
I'll keep trying tho... 🙂
Ah yes sorry that GPU is not going to cut it I think 😬. That sounds awesome though, let me now if you finish that video with the dogs :D
Thanks for this! I get stuck at step 8 with the following error: ModuleNotFoundError: No module named 'tyro'
Any advice? thanks again
Use this command: pip install tyro==0.8.5
Try with this ..... pip install tyro==0.8.5
how do you install ffmpeg?
This is great, except Python 10 or 11 is no longer available for download so I installed the latest which is 3.12. In your instructions, Step 3 create a virtual environment, when pasting your cmd prompt, I get a response that states "Python was not found". Any help you might offer is greatly appreciated. Tom
Python 3.11 is definitely still available, you can google “python 3.11” go to the python.org page and scroll down to find the installer
Hey, if you can, it would be awesome to see a quick video of you doing this with pinokio. Your guidelines are far superior to Pinocchio's official instructions.
Pinokio should be just a one click installer
I can't get it to animate more than 1 person. When I try a photo with 2 or 3 people it just animates the first face. .. Very cool stuff though.
Yeah works best with 1
woah, dragging and dropping auto unzips the file! I know this is not a windows tutorial but dang that is going to help me a lot lol.
Can this bypass withpersona sir?
ModuleNotFoundError: No module named 'torch'
SAME!!
You have to install the ML stack incl. PyTorch
Same
Can you use this for live call
Not with this one yet but there are ways yes
Thank you.
You're welcome!
Amd?
This ain't deepfake
or pinokio with one click install 😋
I didn’t know it was on Pinokio thanks!😬
@@AIQuestOfficial Ha, sum1 told me ;)
@@AIQuestOfficial How do you install this with Pinokio?
If you have pinokio installed, you go to browse and just click install on live portrait
When I run in step 8, there's an error "no module named torch"
(venv) C:\LivePortrait\LivePortrait>.\venv\Scripts\activate
(venv) C:\LivePortrait\LivePortrait>python app.py
Traceback (most recent call last):
File "C:\LivePortrait\LivePortrait\app.py", line 12, in
from src.utils.helper import load_description
File "C:\LivePortrait\LivePortrait\src\utils\helper.py", line 9, in
import torch
ModuleNotFoundError: No module named 'torch'
Torch is not installed. After you activate the venv type: pip install torch
@@AIQuestOfficial there was the same error. I entered "pip install torch", it installed, but then the error "AssertionError: Torch not compiled with CUDA enabled" appears
Ah yes so you need the right torch for your Cuda, I can’t really explain it all here but if you google “How to install torch with Cuda support” you’ll find a guide. If you want, you can try this command though, its usually the right one: pip3 install torch torchvision torchaudio --index-url download.pytorch.org/whl/cu118
@@AIQuestOfficial it works! Thank you very much ❤
Hmm I am getting these errors when running the install that you said you already did so skipped.
× Preparing metadata (pyproject.toml) did not run successfully.
│ exit code: 1
╰─> [21 lines of output]
note: This error originates from a subprocess, and is likely not a problem with pip.
error: metadata-generation-failed
× Encountered error while generating package metadata.
╰─> See above for output.
note: This is an issue with the package mentioned above, not pip.
hint: See above for details.
Hmm, have you tried pasting that into chatgpt? Otherwise check the github issues tab