search

MahmoudAshraf97 / whisper-diarization

Automatic Speech Recognition with Speaker Diarization based on OpenAI Whisper
BSD 2-Clause "Simplified" License
2.53k stars 243 forks source link

RuntimeError: cuDNN error: CUDNN_STATUS_VERSION_MISMATCH #120

Closed srmanas closed 8 months ago

srmanas commented 8 months ago

RuntimeError Traceback (most recent call last) in <cell line: 2>() 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 in flatten_parameters(self) 204 if self.proj_size > 0: 205 num_weights += 1 --> 206 torch._cudnn_rnn_flatten_weight( 207 self._flat_weights, num_weights, 208 self.input_size, rnn.get_cudnn_mode(self.mode),

RuntimeError: cuDNN error: CUDNN_STATUS_VERSION_MISMATCH

Does anyone know how to fix this? I tried using different versions of torch but I'm unable to fix this error

MahmoudAshraf97 commented 8 months ago

uninstall nvidia-cudnn-cu* package