RuntimeError: cuDNN error: CUDNN_STATUS_VERSION_MISMATCH

[TheeElprinceSam]

Discussed in https://github.com/MahmoudAshraf97/whisper-diarization/discussions/100

^{Originally posted by **TheeElprinceSam** October 12, 2023} ### **Hello guys. I am experiencing an error message when i reach at this specific segment.** ### **I** am using Google Colab TPU Speaker Diarization using NeMo MSDD Model Initialize NeMo MSDD diarization model msdd_model = NeuralDiarizer(cfg=create_config(temp_path)).to("cuda") msdd_model.diarize() del msdd_model torch.cuda.empty_cache() ### **The Error i'm getting is this:** RuntimeError Traceback (most recent call last) [](https://localhost:8080/#) in () 1 # Initialize NeMo MSDD diarization model ----> 2 msdd_model = NeuralDiarizer(cfg=create_config(temp_path)).to("cuda") 3 msdd_model.diarize() 4 5 del msdd_model 12 frames [/usr/local/lib/python3.10/dist-packages/torch/nn/modules/rnn.py](https://localhost:8080/#) in flatten_parameters(self) 188 if self.proj_size > 0: 189 num_weights += 1 --> 190 torch._cudnn_rnn_flatten_weight( 191 self._flat_weights, num_weights, 192 self.input_size, rnn.get_cudnn_mode(self.mode), RuntimeError: cuDNN error: CUDNN_STATUS_VERSION_MISMATCH ### **The dependencies used are:** !pip install git+https://github.com/m-bain/whisperX.git@051047bb25b740fed2ea93ca737499c37e8dc9d4 !pip install --no-build-isolation nemo_toolkit[asr]==1.20.0 !pip install faster-whisper==0.9.0 !pip install transformers>=4.26.1 !pip install git+https://github.com/facebookresearch/demucs#egg=demucs !pip install deepmultilingualpunctuation !pip install wget !pip install torch torchaudio torchvision !pip uninstall -y nvidia-cudnn-cu1!pip install cython !pip install pydub import os import wget from omegaconf import OmegaConf import json import shutil from faster_whisper import WhisperModel import whisperx import torch from pydub import AudioSegment from nemo.collections.asr.models.msdd_models import NeuralDiarizer from deepmultilingualpunctuation import PunctuationModel import re import logging

[v-nhandt21]

Try conda install cudatoolkit=11.8 cudnn=8.9.2.26