diff options
| author | catboxanon <122327233+catboxanon@users.noreply.github.com> | 2023-08-31 18:57:16 +0000 |
|---|---|---|
| committer | catboxanon <122327233+catboxanon@users.noreply.github.com> | 2023-08-31 18:57:16 +0000 |
| commit | 5681bf801664aa09fa02ab8b4e73f780d9563440 (patch) | |
| tree | 044ce7b5f8afb811a0d938c339b382bb7905e8e7 /modules/devices.py | |
| parent | d39440bfb9d3b20338fc23a78e6655b1e2f7c1d5 (diff) | |
| download | stable-diffusion-webui-gfx803-5681bf801664aa09fa02ab8b4e73f780d9563440.tar.gz stable-diffusion-webui-gfx803-5681bf801664aa09fa02ab8b4e73f780d9563440.tar.bz2 stable-diffusion-webui-gfx803-5681bf801664aa09fa02ab8b4e73f780d9563440.zip | |
More accurate check for enabling cuDNN benchmark on 16XX cards
Diffstat (limited to 'modules/devices.py')
| -rw-r--r-- | modules/devices.py | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/modules/devices.py b/modules/devices.py index c01f0602..63c38eff 100644 --- a/modules/devices.py +++ b/modules/devices.py @@ -60,7 +60,8 @@ def enable_tf32(): # enabling benchmark option seems to enable a range of cards to do fp16 when they otherwise can't # see https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/4407 - if any(torch.cuda.get_device_capability(devid) == (7, 5) for devid in range(0, torch.cuda.device_count())): + device_id = (int(shared.cmd_opts.device_id) if shared.cmd_opts.device_id.isdigit() else 0) or torch.cuda.current_device() + if torch.cuda.get_device_capability(device_id) == (7, 5) and torch.cuda.get_device_name(device_id).startswith("NVIDIA GeForce GTX 16"): torch.backends.cudnn.benchmark = True torch.backends.cuda.matmul.allow_tf32 = True |