Skip to content

pytorch2onnx API supports multimodal input #2644

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.

Sign up for GitHub

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

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

pytorch2onnx API supports multimodal input #2644

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions mmdeploy/apis/pytorch2onnx.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@ def torch2onnx(img: Any,
deploy_cfg: Union[str, mmengine.Config],
model_cfg: Union[str, mmengine.Config],
model_checkpoint: Optional[str] = None,
append_input: list = None,
device: str = 'cuda:0'):
"""Convert PyTorch model to ONNX model.

Expand Down Expand Up @@ -42,6 +43,7 @@ def torch2onnx(img: Any,
model_cfg (str | mmengine.Config): Model config file or Config object.
model_checkpoint (str): A checkpoint path of PyTorch model,
defaults to `None`.
append_input (list): Additional inputs other than images, suitable for multimodal models such as text features of Grounded DINO.
device (str): A string specifying device type, defaults to 'cuda:0'.
"""

Expand All @@ -68,6 +70,10 @@ def torch2onnx(img: Any,

if isinstance(model_inputs, list) and len(model_inputs) == 1:
model_inputs = model_inputs[0]
if isinstance(append_input, list):
temp = [model_inputs]
temp.extend(append_input)
model_inputs = temp
data_samples = data['data_samples']
input_metas = {'data_samples': data_samples, 'mode': 'predict'}

Expand Down