Transformer encoder is much slower that python onnx runtime #1085
Closed
igor-yusupov
started this conversation in
General
Replies: 3 comments
-
but with version "0.19.8" it works much faster as a python onnx! |
Beta Was this translation helpful? Give feedback.
0 replies
-
0.20 is slower than 0.19 ? can you give me a model ? |
Beta Was this translation helpful? Give feedback.
0 replies
-
@kali https://pytorch.org/docs/1.13/generated/torch.nn.Transformer.html or do you need weights and example code? |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Is it OK that transformer encoder is much slower than python onnx runtime?
Beta Was this translation helpful? Give feedback.
All reactions