Could not load library libcudnn_cnn_train.so.8. Error: /usr/local/cuda/lib64/libcudnn_cnn_train.so.8` Process Process-1: Traceback (most recent call last): File "/opt/conda/lib/python3.10/multiprocessing/process.py", line 314, in _bootstrap self.run() File "/opt/conda/lib/python3.10/multiprocessing/process.py", line 108, in run self._target(*self._args, **self._kwargs) File "/home/Mangio-RVC-Fork/train_nsf_sim_cache_sid_load_pretrain.py", line 254, in run train_and_evaluate( File "/home/Mangio-RVC-Fork/train_nsf_sim_cache_sid_load_pretrain.py", line 395, in train_and_evaluate scaler.scale(loss_disc).backward() File "/opt/conda/lib/python3.10/site-packages/torch/_tensor.py", line 522, in backward torch.autograd.backward( File "/opt/conda/lib/python3.10/site-packages/torch/autograd/__init__.py", line 266, in backward Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass RuntimeError: GET was unable to find an engine to execute this computation
Could the CUDA version be the problem? I'm on CUDA 12.1.
#RVC DIsconnected, local training: Could not load library libcudnn_cnn_train.so.8.
1 messages · Page 1 of 1 (latest)
Wait what do you mean rvc disconnected ? yet you say "I'm on cuda 12.1" ?
Yep I'm running the jupyter notebook locally
oh don't
hardcoded for a lot of google stuff
Any reason you're not running the OG RVC project ?
ahhh I see, I presume I should just run the RVC Mangio Fork instead?
It's kinda deprecated
This would do :
https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI
Make sure to get the right torch build (as per https://pytorch.org/get-started/previous-versions/)
Torch 2.0 & 2.1 both work, but 2.2 probably will cause issues (haven't tested it)
relevant command for torch 2.11 on cuda 12 :
conda install pytorch==2.1.1 torchvision==0.16.1 torchaudio==2.1.1 pytorch-cuda=12.1 -c pytorch -c nvidia
Ahh the OG it is then, thanks, will give it a shot
Ayo? @abstract comet level 2 !!! 
Exact same error when running the OG RVC-Project WebUI notebook. 
no dont use the notebook, the notebook is made specifically for google colab
jsut run it normally, go to releases, downlaod the latest, extract, and run
That's a linux~~/mac~~ machine