Chuang Zhu issues

Repositories
Issues
Comments

Results 4 issues of


                                            Chuang Zhu

Fix host view for mnnvl

improvement

non-breaking

subwarp version gather op for small embedding size

improvement

non-breaking

chore: Ucx ip port remove mpi depend

Ucx ip port remove mpi dependence based on https://github.com/NVIDIA/TensorRT-LLM/pull/3065 but remove MPI world dependency

[https://nvbugs/5685143][fix] avoid cudaFree overlap with cuda graph

## Summary by CodeRabbit * **Bug Fixes** * Improved memory management during KV-cache estimation to prevent potential conflicts and enhance system stability. ✏️ Tip: You can customize this high-level summary...