You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@tvm.apache.org by "MasterJH5574 (via GitHub)" <gi...@apache.org> on 2024/11/21 04:38:39 UTC

[PR] [KVCache] Fix attention prefill kernel for Metal and Android [tvm]

MasterJH5574 opened a new pull request, #17539:
URL: https://github.com/apache/tvm/pull/17539

   This PR fixes two bugs of the attention prefill ragged kernel.
   
   * The first bug is the unroll of loop `ki`, which causes the TIR build failure in the PointerValueTypeRewrite pass due to vector size.
   * The second is the tile sizes of `tile_z` and `tile_y` may violate the assertion check in `get_tile_size`.


-- 
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-unsubscribe@tvm.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org

Re: [PR] [KVCache] Fix attention prefill kernel for Metal and Android [tvm]

Posted by "tqchen (via GitHub)" <gi...@apache.org>.

tqchen merged PR #17539:
URL: https://github.com/apache/tvm/pull/17539


-- 
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-unsubscribe@tvm.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org