cyx-6 opened a new pull request, #14150: URL: https://github.com/apache/tvm/pull/14150
This PR introduces the new relax operator `R.nn.attention` for fused multi head attention, and the support of fused multi head attention to relax cutlass BYOC. The input of the operator are query, key and value tensor, with `BSNH` layout, namely `[batch size, sequence length, number of heads, dimension of heads]`. And the output shares the same layout with all input tensor. -- 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]
