sxjscience commented on issue #19155: URL: https://github.com/apache/incubator-mxnet/issues/19155#issuecomment-694605466
@szhengac Usually, the error message is caused by the force alignment constraint in CUDA, for example, for the two lines in the following, we must ensure that https://github.com/apache/incubator-mxnet/blob/5b506000310fd6bc5852bf4e41c0ce03ccc64013/src/operator/tensor/indexing_op-inl.cuh#L306-L307 ```c++ CHECK_EQ(static_cast<size_t>(workspace->dptr_) % sizeof(IndexType), 0); CHECK_EQ(static_cast<size_t>(workspace->dptr_ + unique_bytes) % sizeof(IndexType), 0); ``` Also, usually this type of error message can be captured by `MSHADOW_CUDA_POST_KERNEL_CHECK`. https://github.com/apache/incubator-mxnet/blob/5b506000310fd6bc5852bf4e41c0ce03ccc64013/src/operator/tensor/indexing_op.cu#L804-L809 Would you try to add the `MSHADOW_CUDA_POST_KERNEL_CHECK` like this? This helps us locate the cuda launching errors. ``` MSHADOW_CUDA_POST_KERNEL_CHECK(EmbeddingGradKernel); ``` ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
