Default Branch

main eeea00b298 · Support python wheel build (#787) · Updated 2026-04-17 04:24:45 +00:00

ZIP  TAR.GZ

Branches

caiorocha/dsl_fuse_op 7a8860bbe4 · wip · Updated 2025-06-05 06:52:38 +00:00    microsoft	171 3		ZIP  TAR.GZ
chhwang/blackwell 7782eff8d4 · Merge branch 'chhwang/cmake-gpu-archs' into chhwang/blackwell · Updated 2025-05-12 23:11:07 +00:00    microsoft	181 32		ZIP  TAR.GZ
caiorocha/msccllang cd4d8c5ad5 · wip · Updated 2025-04-16 19:48:15 +00:00    microsoft	198 9		ZIP  TAR.GZ
qinghuazhou/support-execution-plan-for-channel-cache-disabling 3e79361c2b · Add memoryChannelsNoCach, portChannelsNoCache, nvlsChannelsNoCache, scratchBuffersNoCache in ExecutionContext to support channel cache disabling for execution plan; Skip the memoryPool in ExecutionContext; Fix the virtual address mapping issue of cuMemMap · Updated 2025-04-16 01:24:02 +00:00    microsoft	199 5		ZIP  TAR.GZ
binyli/sglang 7dc83b575c · remove loop · Updated 2025-03-20 19:00:24 +00:00    microsoft	208 9		ZIP  TAR.GZ
qinghuazhou/nccl-rccl-integration-ci-test 2f8f2f899f · Add CI test for fallback allgather, allreduce, broadcastand reducescatter to NCCL operations · Updated 2025-03-19 22:35:12 +00:00    microsoft	210 21		ZIP  TAR.GZ
qinghuazhou/nccl-rccl-integration-dlopen-force-fallback-ops 799995655b · Update names of variables and functions · Updated 2025-03-15 05:11:37 +00:00    microsoft	212 16		ZIP  TAR.GZ
caiorocha/fix_get_operation_msccllang 8fee8915da · wip · Updated 2025-02-28 23:09:47 +00:00    microsoft	217 7		ZIP  TAR.GZ
anantharamus/bcast-with-recvbuff-exchange c5f3b60ade · Tested the exchange of recvbuff with scratch buff · Updated 2025-02-19 18:47:16 +00:00    microsoft	221 3		ZIP  TAR.GZ
binyli/rocm-6.3 5742554f98 · fix for rocm6.3 · Updated 2025-02-04 11:23:43 +00:00    microsoft	224 1		ZIP  TAR.GZ
binyli/nccl-api 9bc13e12da · update for ncclCommSplit (#456) · Updated 2025-01-22 22:43:22 +00:00    microsoft	232 6		ZIP  TAR.GZ
chhwang/macro ac5f17d114 · Merge branch 'main' into chhwang/macro · Updated 2025-01-18 23:57:28 +00:00    microsoft	227 4		ZIP  TAR.GZ
anantharamus/broadcast-amd-fuse-two-stage 753d5e3545 · Fixed a bug for messages > 140M · Updated 2025-01-08 04:57:02 +00:00    microsoft	271 19		ZIP  TAR.GZ
release/v0.6.0 11e62024d3 · [Cherry-pick] Update version number (#433) (#435) · Updated 2025-01-03 11:15:15 +00:00    microsoft	252 6		ZIP  TAR.GZ
mahdieh/bcast-lowinitoverhead f42629911a · improve performance on cpu side · Updated 2024-12-23 15:59:59 +00:00    microsoft	241 1		ZIP  TAR.GZ
anantharamus/broadcast-amd b98b977331 · Scratch buffer copy-based implementation of ncclBcast / ncclBroadcast (#417) · Updated 2024-12-17 23:58:26 +00:00    microsoft	271 10		ZIP  TAR.GZ
anantharamus/broadcast-amd-bcastfix bd7764a451 · Found a fix for ncclBcast hang. · Updated 2024-12-12 20:11:46 +00:00    microsoft	271 10		ZIP  TAR.GZ
chhwang/test eb831e9141 · tmp · Updated 2024-12-10 11:42:25 +00:00    microsoft	254 4		ZIP  TAR.GZ
aashaka/vllm_changes 9c78da1127 · home directory included · Updated 2024-12-10 02:52:37 +00:00    microsoft	256 73		ZIP  TAR.GZ
chhwang/2-node-ag 94aed9ef40 · Add a 2-node allgather algo · Updated 2024-11-18 07:35:43 +00:00    microsoft	267 1		ZIP  TAR.GZ