Support ONNX export on torch.float16 type
#749
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
JingyaHuang
reviewed
Feb 6, 2023
I've tried on a single model - and did not try to load into an InferenceSession. I will add a test for it, thanks. |
|
The documentation is not available anymore as the PR was closed or merged. |
regisss
reviewed
Feb 7, 2023
1ff3ed8 to
2d210a6
Compare
4 tasks
torch.float16 devicetorch.float16 type
Co-authored-by: regisss <15324346+regisss@users.noreply.github.com>
…rty/optimum into support-onnx-export-float16
regisss
approved these changes
Feb 23, 2023
| optional_group.add_argument( | ||
| "--fp16", | ||
| action="store_true", | ||
| help="Experimental option: use half precision during the export. PyTorch-only, requires `--device cuda`.", |
There was a problem hiding this comment.
Experimental because it doesn't work with all models?
There was a problem hiding this comment.
I'd say experimental because I haven't thouroughly tested it with ONNX Runtime + CUDAExecutionProvider / TensorrtExecutionProvider, and neither with native TensorRT (though in the validation itself we call InferenceSession on CUDA EP, so it's a good sign it's fine). But the export itself is thoroughly tested.
Co-authored-by: regisss <15324346+regisss@users.noreply.github.com>
…rty/optimum into support-onnx-export-float16
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
As per title.
Test still missing. Partly fixes https://discuss.huggingface.co/t/convert-gpt-j-to-fp-16-onnx/30294