Skip to content

[ONNX] Enable dynamo export#3708

Open
andrey-churkin wants to merge 26 commits intoopenvinotoolkit:developfrom
andrey-churkin:ac/onnx_dynamo_export
Open

[ONNX] Enable dynamo export#3708
andrey-churkin wants to merge 26 commits intoopenvinotoolkit:developfrom
andrey-churkin:ac/onnx_dynamo_export

Conversation

@andrey-churkin
Copy link
Contributor

@andrey-churkin andrey-churkin commented Oct 30, 2025

Changes

Enable torch dynamo export by default.

Related tickets

Ref: 168961

@andrey-churkin andrey-churkin requested a review from a team as a code owner October 30, 2025 10:19
@github-actions github-actions bot added the NNCF PTQ Pull requests that updates NNCF PTQ label Oct 30, 2025
@github-actions github-actions bot added NNCF PT Pull requests that updates NNCF PyTorch NNCF OpenVINO Pull requests that updates NNCF OpenVINO NNCF ONNX Pull requests that updates NNCF ONNX labels Jan 19, 2026
@andrey-churkin
Copy link
Contributor Author

@andrey-churkin
Copy link
Contributor Author

@andrey-churkin
Copy link
Contributor Author

post_training_quantization: Build # 793

"164 nncf_model_input_0" -> "9 /Gather_1" [label="[1, 3, 224, 224]", style=solid];
"164 nncf_model_input_0" -> "17 /Gather_2" [label="[1, 3, 224, 224]", style=solid];
strict digraph {
"0 node_select" [id=0, type=Gather];
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What a reason that in new graphs no Constant nodes?

onnx_metatypes.ONNXDepthwiseConvolutionMetatype,
onnx_metatypes.ONNXConvolutionTransposeMetatype,
onnx_metatypes.ONNXDeformableConvolutionMetatype,
onnx_metatypes.ONNXGroupConvolutionMetatype,
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is this change related to dynamo export?

accelerate==1.9.0
transformers==4.53.0
whowhatbench @ git+https://github.com/openvinotoolkit/openvino.genai@2025.4.1.0#subdirectory=tools/who_what_benchmark
datasets
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

duplicate, datasets exists in line 3

torch.onnx.export(
model, torch.rand(model.INPUT_SIZE), onnx_path, opset_version=13, input_names=["input.1"], dynamo=False
)
model.eval()
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

eval() requares only for onnx?
May be is it will be usefull add in in common part before call backend_specific_model?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

NNCF ONNX Pull requests that updates NNCF ONNX NNCF OpenVINO Pull requests that updates NNCF OpenVINO NNCF PT Pull requests that updates NNCF PyTorch NNCF PTQ Pull requests that updates NNCF PTQ

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants

Comments