需求描述 Feature Description
We noticed that IR passes are only applied to block 0 e.g. conditional_block_infer inner operators are not a subject to IR passes.
More and more transformer based models are using conditional_block_infer so It would be good to have IR passes there to take advantage of fuses and oneDNN kernels. Some work was done here
but it was to make enabled oneDNN for conditional_infer_block when ENV VAR is set , which is not how we work with PaddlePaddle and oneDNN for inference as we use AnalysisConfig to pass info on using oneDNN.
Please make IR passes working for inner ops of conditional_infer_block.
This is duplicate of this #17003 .
替代实现 Alternatives
No response
2条答案
按热度按时间krcsximq1#
您好，我们已经收到了您的问题，会安排技术人员尽快解答您的问题，请耐心等待。请您再次检查是否提供了清晰的问题描述、复现代码、环境&版本、报错信息等。同时，您也可以通过查看 官网API文档 、 常见问题 、 历史Issue 、 AI社区 来寻求解答。祝您生活愉快～
Hi! We've received your issue and please be patient to get responded. We will arrange technicians to answer your questions as soon as possible. Please make sure that you have posted enough message to demo your request. You may also check out the API ， FAQ ， Github Issue and AI community to get the answer.Have a nice day!
wkyowqbh2#
this is related to zh794390558/paddle_build#5 (comment)