Skip to content

Commit 935f4a5

Browse files
authored
Remove PyTorch multiple instance known issue (#80)
1 parent 5477b11 commit 935f4a5

File tree

1 file changed

+0
-8
lines changed

1 file changed

+0
-8
lines changed

README.md

Lines changed: 0 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -218,14 +218,6 @@ complex execution modes and dynamic shapes. If not specified, all are enabled by
218218
state and a restart of the server may be required to continue serving
219219
successfully.
220220

221-
* Multiple instances of the PyTorch model on GPU do not always
222-
increase performance. Due to thread specific caching in PyTorch, using
223-
multiple instances of the model interact negatively. See
224-
[here](https://github.com/pytorch/pytorch/issues/27902) for more details.
225-
Setting the parameter `DISABLE_OPTIMIZED_EXECUTION` to "true" in the model
226-
configuration may help in some cases to avoid these negative interactions
227-
due to model specific caching and increase multiple instance performance.
228-
229221
* PyTorch does not support Tensor of Strings but it does support models that accept
230222
a List of Strings as input(s) / produces a List of String as output(s). For these models
231223
Triton allows users to pass String input(s)/recieve String output(s) using the String

0 commit comments

Comments
 (0)