-
Notifications
You must be signed in to change notification settings - Fork 74k
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
Add support for TensorRT 10 #68715
base: master
Are you sure you want to change the base?
Add support for TensorRT 10 #68715
Commits on May 28, 2024
-
[TRT10] Add support for TensorRT 10.0
Signed-off-by: Meenakshi Venkataraman <meenakshiv@nvidia.com>
Configuration menu - View commit details
-
Copy full SHA for 94d1bb9 - Browse repository at this point
Copy the full SHA 94d1bb9View commit details -
Use same inc file for TRT10 as for TRT8 and below
Signed-off-by: Meenakshi Venkataraman <meenakshiv@nvidia.com>
Configuration menu - View commit details
-
Copy full SHA for 864671b - Browse repository at this point
Copy the full SHA 864671bView commit details -
Configuration menu - View commit details
-
Copy full SHA for fc5523f - Browse repository at this point
Copy the full SHA fc5523fView commit details -
Fix TF-TRT shape layer int64/int32 mismatches
- TRT10 changed the output dtype of shape layers from int32 to int64, which causes mismatches with other layers. This commit adds cast layers to avoid the mismatches. - Note that this also adds support for the out_dtype attribute of TF's Shape operator.
Configuration menu - View commit details
-
Copy full SHA for 711dfa5 - Browse repository at this point
Copy the full SHA 711dfa5View commit details -
Disable implicit batch in TF-TRT tests for TRT10
- Also avoids runtime warnings about the hasImplicitBatchDimension API.
Configuration menu - View commit details
-
Copy full SHA for b81ae11 - Browse repository at this point
Copy the full SHA b81ae11View commit details -
Configuration menu - View commit details
-
Copy full SHA for be1316f - Browse repository at this point
Copy the full SHA be1316fView commit details -
Fix bug in TF-TRT with TRT10 when finding engines
- It seems that the number of inputs no longer needs to be divided by the number of profiles. This manifested as a confusing bug because it caused values in the array of min/max/opt to be written over the top of existing values instead of at the end, and this subsequently prevented shapes from being matched when looking up engines.
Configuration menu - View commit details
-
Copy full SHA for e99530e - Browse repository at this point
Copy the full SHA e99530eView commit details -
Configuration menu - View commit details
-
Copy full SHA for 5131ab9 - Browse repository at this point
Copy the full SHA 5131ab9View commit details -
Configuration menu - View commit details
-
Copy full SHA for a0f1c62 - Browse repository at this point
Copy the full SHA a0f1c62View commit details -
Configuration menu - View commit details
-
Copy full SHA for ed93558 - Browse repository at this point
Copy the full SHA ed93558View commit details -
Configuration menu - View commit details
-
Copy full SHA for 210c7b1 - Browse repository at this point
Copy the full SHA 210c7b1View commit details -
Change TRT to default to use_dynamic_shape=True
- use_dynamic_shape=False is not supported since TensorRT 10.0. - Also expands the related error message.
Configuration menu - View commit details
-
Copy full SHA for 699a77b - Browse repository at this point
Copy the full SHA 699a77bView commit details