Adds patch to rosetta-pax for PP fix (#981)

Adds this patch to the rosetta-pax container: google/praxis#83 Can remove when merged Signed-off-by: Terry Kong <[email protected]>
NVIDIA · Aug 13, 2024 · 3cd303d · 3cd303d
1 parent 19f8181
commit 3cd303d
Showing 1 changed file with 1 addition and 0 deletions.
diff --git a/.github/container/manifest.yaml b/.github/container/manifest.yaml
@@ -50,6 +50,7 @@ praxis:
     pull/27/head: file://patches/praxis/PR-27.patch # This PR allows XLA:GPU to detect the MHA pattern more easily to call fused kernels from cublas.
     pull/36/head: file://patches/praxis/PR-36.patch # adds Transformer Engine support
     pull/74/head: file://patches/praxis/PR-74.patch # experimental support for using TE FMHA in GQA
+    pull/83/head: file://patches/praxis/PR-83.patch # Fix unindex tuple error for PP
 lingvo:
   # Used only in ARM pax builds
   url: https://github.com/tensorflow/lingvo.git