Re: [PR] [relay][feature] save relay IR as onnx for visualize [tvm]
tqchen commented on PR #16847: URL: https://github.com/apache/tvm/pull/16847#issuecomment-2061421527 seems a better approach to write something like https://github.com/lutzroeder/netron/blob/main/source/caffe.js in netron. That file would parse the json file exported by the tvm relax, and construct the graph. Alternatively, we can add a visualization spec json for tvm, and add netron support. This would remove the need of protobuf and allows more direct export -- 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: commits-unsubscr...@tvm.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] [relay][feature] save relay IR as onnx for visualize [tvm]
ShawnZhuang closed pull request #16847: [relay][feature] save relay IR as onnx for visualize URL: https://github.com/apache/tvm/pull/16847 -- 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: commits-unsubscr...@tvm.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] [relay][feature] save relay IR as onnx for visualize [tvm]
ShawnZhuang commented on PR #16847: URL: https://github.com/apache/tvm/pull/16847#issuecomment-2051210080 @tqchen it seam that google protobuf is required, and should be added to dockfile. Do you any sugguestion on editing the CmakeList and Dockerfile, since lots of docker file are found in tvm/docker -- 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: commits-unsubscr...@tvm.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] [relay][feature] save relay IR as onnx for visualize [tvm]
ShawnZhuang commented on PR #16847: URL: https://github.com/apache/tvm/pull/16847#issuecomment-2048849798 ok, i would start bringing onnx translation to relax after this PR work is finished. -- 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: commits-unsubscr...@tvm.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] [relay][feature] save relay IR as onnx for visualize [tvm]
ShawnZhuang commented on PR #16847: URL: https://github.com/apache/tvm/pull/16847#issuecomment-2048833676 > thank you @ShawnZhuang , one thing that I think worth clarifying is that whether it serves as a onnx exporter(which i believe there were some work in https://github.com/apache/tvm/blob/main/python/tvm/contrib/target/onnx.py) seems the purpose was different(mainly to export to netron compatible format, is it possible to support a variant of format that netron directly take? > > as a community we are migrating a lot of effort to relax, so if you are interested in bringing onnx translation to relax, that could also be interesting Yes, the mainly purpure is to export to netron compatible format for visualize. According to https://github.com/lutzroeder/netron , the onnx supports tf, caffe, keras,mxnet,onnx and darknet , but a directly netron format has not been found. Supporting netron directly may takes two steps, Firstly a tvm.proto should be defined, secondly, netron should support the new proto(which needs front-end efforts in netron) -- 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: commits-unsubscr...@tvm.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] [relay][feature] save relay IR as onnx for visualize [tvm]
tqchen commented on PR #16847: URL: https://github.com/apache/tvm/pull/16847#issuecomment-2037249162 thank you @ShawnZhuang , one thing that I think worth clarifying is that whether it serves as a onnx exporter(which i believe there were some work in https://github.com/apache/tvm/blob/main/python/tvm/contrib/target/onnx.py as a community we are migrating a lot of effort to relax, so if you are interested in bringing onnx translation to relax, that could also be interesting -- 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: commits-unsubscr...@tvm.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] [relay][feature] save relay IR as onnx for visualize [tvm]
ShawnZhuang commented on PR #16847: URL: https://github.com/apache/tvm/pull/16847#issuecomment-203717 discuss on https://discuss.tvm.apache.org/t/relay-onnx-support-saveing-relay-ir-to-onnx-for-visualization/17020 -- 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: commits-unsubscr...@tvm.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org