Movatterモバイル変換


[0]ホーム

URL:


Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

Sign up
Appearance settings

Commite3e399a

Browse files
authored
[5336870] AutoCast: Unblock LSTM from conversion (#544)
Blocking was added as a WAR for TensorRT bug 5336870, fixed in TensorRT10.14## What does this PR do?**Type of change:** Bug fix <!-- Use one of the following: Bug fix, newfeature, new example, new tests, documentation. -->**Overview:** Unblock LSTM from conversion to 16-bit in AutoCast.Blocking was added as a WAR for TensorRT bug 5336870 (TRT only supportedFP32 LSTM). Bug was fixed in TensorRT 10.14 which is now released.## Usage<!-- You can potentially add a usage example below. -->```python# Add a code snippet demonstrating how to use this```## Testing<!-- Mention how have you tested your change if applicable. -->## Before your PR is "*Ready for review*"<!-- If you haven't finished some of the above items you can still open`Draft` PR. -->- **Make sure you read and follow [Contributorguidelines](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CONTRIBUTING.md)**and your commits are signed.- **Is this change backward compatible?**: Yes <!--- If No, explain why.-->- **Did you write any new necessary tests?**: No- **Did you add or update any necessary documentation?**: No- **Did you update[Changelog](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CHANGELOG.rst)?**:No <!--- Only for new features, API changes, critical bug fixes or bwbreaking changes. -->## Additional Information<!-- E.g. related issue. -->Signed-off-by: Gal Hubara Agam <96368689+galagam@users.noreply.github.com>
1 parent4853668 commite3e399a

File tree

1 file changed

+1
-1
lines changed

1 file changed

+1
-1
lines changed

‎modelopt/onnx/autocast/precisionconverter.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -68,7 +68,7 @@ class InitializerConsumerTracker:
6868
OP_TYPES_NOT_SUPPORTED_IN_LOW_PRECISION= ["Upsample","NonMaxSuppression","Celu"]
6969

7070
# Temporarily block these ops in low precision, as they are not supported yet
71-
OP_TYPES_NOT_SUPPORTED_IN_LOW_PRECISION.extend(["Scan","If","Loop","LSTM"])
71+
OP_TYPES_NOT_SUPPORTED_IN_LOW_PRECISION.extend(["Scan","If","Loop"])
7272

7373
# Mapping of op types to indices of inputs that should not be converted to low precision.
7474
SKIP_LOW_PRECISION_MAPPING_FP16= {"Resize": {2}}

0 commit comments

Comments
 (0)

[8]ページ先頭

©2009-2025 Movatter.jp