You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Here are some onnx models, where MBV3_SSD_onnx.onnx is the original model, mbv3_ssd_onnx_sim_macos.onnx is the model after optimizing and simplifying on macOS, MBV3_SSD_onnx_sim_onnx_16_pip.onnx is the model after optimizing and simplifying on ubuntu.
The "1523" tensors of the last two models differ a lot, which is caused by the onnx fuse_bn_into_conv optimizer. The tensor in ubuntu model (no matter produced by onnx 1.5 or 1.6) is totally wrong while the macOS version is right.
However, if I compiled and install onnx from source on ubuntu, the optimized model is correct and the same as that on macOS.
The text was updated successfully, but these errors were encountered:
daquexian
changed the title
fuse_bn_into_conv optimizer in pypi ubuntu py3.5 wheels is wrong
fuse_bn_into_conv optimizer in pypi ubuntu wheels is wrong
Mar 23, 2020
daquexian
changed the title
fuse_bn_into_conv optimizer in pypi ubuntu wheels is wrong
fuse_bn_into_conv optimizer in pypi linux wheels is wrong
Mar 24, 2020
@askhade yes, I just verified it still exits on onnx 1.7.0 installed by pip3 install onnx==1.7.0 on linux. what's more, if I installed onnx by pip3 install onnx==1.7.0 --no-binary onnx (which builds onnx from source instead of installing the prebuilt binary), the problem disappears
Reference: daquexian/onnx-simplifier#50
Here are some onnx models, where
MBV3_SSD_onnx.onnx
is the original model,mbv3_ssd_onnx_sim_macos.onnx
is the model after optimizing and simplifying on macOS,MBV3_SSD_onnx_sim_onnx_16_pip.onnx
is the model after optimizing and simplifying on ubuntu.The "1523" tensors of the last two models differ a lot, which is caused by the onnx fuse_bn_into_conv optimizer. The tensor in ubuntu model (no matter produced by onnx 1.5 or 1.6) is totally wrong while the macOS version is right.
However, if I compiled and install onnx from source on ubuntu, the optimized model is correct and the same as that on macOS.
The text was updated successfully, but these errors were encountered: