Nvidia 1660ti Set Up

[sky-cake]

Python 3.12.12
Ubuntu 22.04
Nvidia 1660ti

faster-whisper args:

• compute_type = 'int8_float32' # very important
• device = 'cuda'
• size = medium'

Download and install cudnn-local-repo-ubuntu2204-9.16.0_1.0-1_amd64.deb by following the instructions at, https://developer.nvidia.com/cudnn-downloads?target_os=Linux&target_arch=x86_64&Distribution=Debian&target_version=12&target_type=deb_local&Configuration=Full

faster-whisper==1.2.1
ctranslate2==4.5.0

nvidia-cublas-cu12==12.8.4.1
nvidia-cuda-cupti-cu12==12.8.90
nvidia-cuda-nvrtc-cu12==12.8.93
nvidia-cuda-runtime-cu12==12.8.90
nvidia-cudnn-cu12==9.10.2.21
nvidia-cufft-cu12==11.3.3.83
nvidia-cufile-cu12==1.13.1.3
nvidia-curand-cu12==10.3.9.90
nvidia-cusolver-cu12==11.7.3.90
nvidia-cusparse-cu12==12.5.8.93
nvidia-cusparselt-cu12==0.7.1
nvidia-nccl-cu12==2.27.5
nvidia-nvjitlink-cu12==12.8.93
nvidia-nvshmem-cu12==3.3.20
nvidia-nvtx-cu12==12.8.90

[Purfview]

Is this an issue, or did you mean to post in "Show and tell"?

Didn't CUDA 12.8 drop support for the 16xx series? You may want to check ctranslate2.get_cuda_device_count().

[sky-cake]

Sorry, it is NOT an issue, more like documentation. I spent forever trying to get this working, but it was all because of my compute_type not being int8_float32.