-
Notifications
You must be signed in to change notification settings - Fork 3.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[TENSORRT] Improvements and fixes for TensorRT #11203
Conversation
A number of small fixes and refactors to improve the robustness of the TensorRT integration. Co-authored-by: Mark Shields <mbs@octoml.ai>
The background on the bits I fiddled with:
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, thank you!
I wonder if we should warn for the conv3d attributes you found trigger numerical inaccuracy? Say in the predicate?
Unfortunately, the attributes aren't particularly well-defined. For example, in that specific test it was failing with channels between 28 and 45. I think so long as TensorRT is selecting strategies with different numerical behaviour based on tuning we won't be able to give a sane warning here. |
Ack. We do what we can. |
LGTM, thanks |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for these improvements @mbaret, @mikepapadim, and @mbs-octoml!
A number of small fixes and refactors to improve the robustness of the TensorRT integration. Co-authored-by: Mark Shields <mbs@octoml.ai> Co-authored-by: Mark Shields <mbs@octoml.ai>
A number of small fixes and refactors to improve the robustness of the TensorRT integration. Co-authored-by: Mark Shields <mbs@octoml.ai> Co-authored-by: Mark Shields <mbs@octoml.ai>
A number of small fixes and refactors to improve the robustness of the TensorRT integration. Co-authored-by: Mark Shields <mbs@octoml.ai> Co-authored-by: Mark Shields <mbs@octoml.ai>
A number of small fixes and refactors to improve the robustness of the TensorRT integration. Co-authored-by: Mark Shields <mbs@octoml.ai> Co-authored-by: Mark Shields <mbs@octoml.ai>
nice job! |
A number of small fixes and refactors to improve the robustness of the TensorRT integration.
Co-authored-by: Mark Shields mbs@octoml.ai