Oguz Ulgen

TTIR printing inconsistencies

8

I have noticed that when dumping the TTIR from triton, the format is inconsistent. OPs are sometimes printed in their pretty format and sometimes in their generic format, e.g. ```...

Locally scoped kernels are not found

3

I was writing a unit test in PyTorch and wanted to define some kernels scoped under my test. It looks like if you define kernels scoped, the one you call...

Initial implementation of Inductor FX Graph Remote Cache

3

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * __->__ #124669 This diff implements a remote caching strategy (memcache for internal and redis for external) for caching of Inductor FX Graph to...

Do not trace into triton/backends

1

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * #126903 * __->__ #126083 Fixes #125807 cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @peterbell10 @ipiszy @yf225 @chenyang78 @kadeng @muchulee8...

[easy] Delete dead code global

1

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * __->__ #126903 * #126083 cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @chenyang78 @kadeng @chauhang

Consolidate Triton cache into Inductor cache

6

Summary: This diff/PR attempts to consolidate Triton caching into the Inductor caching so that there can be just one cache that unifies them both, reducing network requests and increasing success...