chengven027-intellif commented on issue #13226: URL: https://github.com/apache/tvm/issues/13226#issuecomment-1303149259
Hi, @lee-bin I try to import your [model](https://github.com/lee-bin/test/blob/master/onnx-example.onnx) in my code. Actually I don`t get your error. But I found that its runtime is very slow. Then I modify the code as follows： ``` --- a/python/tvm/topi/x86/conv1d.py +++ b/python/tvm/topi/x86/conv1d.py @@ -65,7 +65,7 @@ def schedule_conv1d_ncw(outs): if op != output_op: # fuse bias + bn + relu into conv s[C].compute_at(s[output_op], w_out) else: - fused = s[C].fuse(n, c) + fused = s[C].fuse(n, c, w) s[C].parallel(fused) ``` It works fast and can be imported successfully. I don't know if this modification will have other effects. Maybe we can ask @masahi @AndrewZhaoLuo to take a look if they have time. Thanks. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
