GitLab

Fix the address of dispatch_rdma_recv_count_buffer to avoid cleaning after each change in hidden_size/token_num. (#313)
Signed-off-by: Yilin Zhang <18275976+yilin-void@users.noreply.github.com>

* Add automatic warp count control for low-latency dispatch

* Add automatic warp count control for low-latency combine

* More assertions

* Update README

* Update `setup.py`

* Fix headers

* Add `DISABLE_NVSHMEM` for APIs

* Fix launch

* Fix TMA settings

* Fix TMA usages

* Fix dlink

* Separate layout kernels

* Update version

* Add `is_sm90_compiled`

* Fix tests

* Add NVLink connection checks

* Update README

* Fix tests

* Add some comments

* Minor fix

* Minor fix

* Fix bugs