EthanYe

> @yeliang2258 Please update your PR to latest codebase . We would like to evaluate it by our CI to check impact on all models we trace which will take...

@jczaja Hello, I want to know, how can I print out the specific data in the onednnl memory

@Silv3S Thank you for your reply. The solution in your branch does work, but when the shape is different, what is the problem with the code I submitted? I create...

@Silv3S Hello, I had a user problem, same as the one here, I solved it using the solution you provided, and I made a PR to fix conv + elementwise_add...

@jczaja Please help me to review this PR, thanks.

@Silv3S Please help me to review this PR, thanks.

我看你这边是对相同的输入重复计算了10次求均值，麻烦再测试一下，先warmup 100轮，然后计算1000轮的均值？看看耗时怎么样

> > 最好能提供一下全套的测试代码(paddle和pytorch)给我们，因为之前我和Huggface的模型也是对比过，测试的结论和您提供的结论差别较大 > > 有没有可能是我这边没有用上GPU？在CPU推理的时候，paddle确实有优势。但是做GPU推理的时候，在代码中我已经配置了CUDAExecutionProvider，代码也成功运行，没有报错。如果可以，能不能提供一个能同时满足paddlenlp训练和onnxruntime-gpu的docker，让再测一下？ 生成predictor之后，调用一下predictor.get_providers()，看看里面是否有CUDAExecutionProvider，另外运行起来中后看看显存和GPU利用率

请尝试使用下面命令安装，pip install packaging

你好，请提供一下模型，谢谢