Gaurav Garg 239c5c86c3 Tensor-parallel: Fix delayed AllReduce on Gemma-4 MoE (llama/22129) ...

* Fix delayed AllReduce on Gemma-4 MoE

Skip forward past nodes that don't consume the current one, and allow a chain of MULs.

* Check for all sources before skipping nodes

* Address review comments

2026-04-30 11:29:14 +03:00

..

cmake

cmake : remove unused file (ggml/1419)

2026-02-08 09:29:10 +02:00

include

CUDA: manage NCCL communicators in context (llama/21891)

2026-04-30 11:29:09 +03:00

src

Tensor-parallel: Fix delayed AllReduce on Gemma-4 MoE (llama/22129)

2026-04-30 11:29:14 +03:00

.gitignore

…

CMakeLists.txt

cmake: remove CMP0194 policy to restore MSVC builds (llama/21934)

2026-04-30 11:29:12 +03:00