aboutsummaryrefslogtreecommitdiffstats
path: root/modules/devices.py
diff options
context:
space:
mode:
authorAUTOMATIC <16777216c@gmail.com>2022-09-12 13:34:13 +0000
committerAUTOMATIC <16777216c@gmail.com>2022-09-12 13:34:13 +0000
commitb70b51cc7248f57dcf16add86701139432c21e5b (patch)
tree8dbe1cbdd064872383fecbd8230601e8ccab484c /modules/devices.py
parent11e648f6c75de2fb22460d34a618dbb3aa6df0bc (diff)
downloadstable-diffusion-webui-gfx803-b70b51cc7248f57dcf16add86701139432c21e5b.tar.gz
stable-diffusion-webui-gfx803-b70b51cc7248f57dcf16add86701139432c21e5b.tar.bz2
stable-diffusion-webui-gfx803-b70b51cc7248f57dcf16add86701139432c21e5b.zip
Allow TF32 in CUDA for increased performance #279
Diffstat (limited to 'modules/devices.py')
-rw-r--r--modules/devices.py11
1 files changed, 11 insertions, 0 deletions
diff --git a/modules/devices.py b/modules/devices.py
index f88e807e..a93a245b 100644
--- a/modules/devices.py
+++ b/modules/devices.py
@@ -1,6 +1,8 @@
import torch
# has_mps is only available in nightly pytorch (for now), `getattr` for compatibility
+from modules import errors
+
has_mps = getattr(torch, 'has_mps', False)
cpu = torch.device("cpu")
@@ -20,3 +22,12 @@ def torch_gc():
if torch.cuda.is_available():
torch.cuda.empty_cache()
torch.cuda.ipc_collect()
+
+
+def enable_tf32():
+ if torch.cuda.is_available():
+ torch.backends.cuda.matmul.allow_tf32 = True
+ torch.backends.cudnn.allow_tf32 = True
+
+
+errors.run(enable_tf32, "Enabling TF32")