Skip to content

Activity

Merge branch 'main' into fix_cancel

QiJunepushed 4 commits to fix_cancel • 36e52be…6bbbaa7 • 
1 hour ago

Merge branch 'release/0.21' into fix_allgather_1

QiJunepushed 5 commits to fix_allgather_1 • 66b2ffd…9dcaab8 • 
2 hours ago

put canceled request to request queue

QiJunepushed 1 commit to fix_cancel • c6a69e2…36e52be • 
3 hours ago

polish code comment

QiJunepushed 1 commit to fix_allgather_1 • 251d767…66b2ffd • 
4 hours ago

polish

QiJunepushed 1 commit to fix_cancel • d92a9c3…c6a69e2 • 
13 hours ago

fix cancel request logic

QiJunecreated fix_cancel • d92a9c3 • 
13 hours ago

[ci] speedup fused moe tests (NVIDIA#5726)

QiJunepushed 4 commits to main • 5ca2b9b…1191555 • 
13 hours ago

[TRTLLM-5812][feat] support FP8 row-wise dense GEMM in torch flow (NV…

QiJunepushed 7 commits to main • 092e0eb…5ca2b9b • 
17 hours ago

polish

QiJunepushed 1 commit to fix_allgather_1 • 7b66676…251d767 • 
21 hours ago

avoid nesting NCCL grouping in allgather OP

QiJunecreated fix_allgather_1 • 7b66676 • 
21 hours ago

avoid nesting NCCL groups

QiJunecreated fix_allgather • 34cb281 • 
21 hours ago

Fix docker cache mount (NVIDIA#5763)

QiJunepushed 5 commits to release/0.21 • aa4d0f0…06f8327 • 
21 hours ago

add supported models doc (NVIDIA#5662)

QiJunepushed 2 commits to feat/1.0_doc_dev • 7a617ad…0210359 • 
22 hours ago

add Deprecation Policy section

QiJunecreated deprecation • cdc8e65 • 
23 hours ago

[Infra] - Fix a syntax issue in the image check (NVIDIA#5775)

QiJunepushed 66 commits to main • 10c5051…092e0eb • 
23 hours ago

polish

QiJunepushed 1 commit to models • 1718d94…dcfc48c • 
23 hours ago

Update docs/source/torch/models/supported_models.md

QiJunepushed 1 commit to models • 1268706…1718d94 • 
23 hours ago

Merge branch 'release/0.21' into cancel_1

QiJunepushed 22 commits to cancel_1 • d4399e2…09f8bb3 • 
yesterday

cherry pick NVIDIA#5416

QiJunecreated cp-5416 • f4a9fb7 • 
yesterday

[Infra] - Always use x86 image for the Jenkins agent (NVIDIA#5756)

QiJunepushed 10 commits to release/0.21 • 2f9d061…aa4d0f0 • 
yesterday

clean

QiJunecreated fix_pad_4 • b620e5d • 
3 days ago

fix allgather

QiJunecreated fix_pad_3 • 58d37ac • 
4 days ago

fix

QiJunepushed 1 commit to fix_pad_2 • c3bb59e…e30a4d1 • 
5 days ago

always padding before allgather in deepseek

QiJunecreated fix_pad_2 • c3bb59e • 
5 days ago

[Infra] - Waive failed cases on release/0.21 (NVIDIA#5674)

QiJunepushed 7 commits to release/0.21 • 9fe1dd6…2f9d061 • 
5 days ago

polish

QiJunecreated models • 1268706 • 
6 days ago

feat: W4A16 GEMM (NVIDIA#4232)

QiJunecreated feat/1.0_doc_dev • 7a617ad • 
6 days ago

fix: Add back allreduce_strategy parameter into TorchLlmArgs (NVIDIA#…

QiJunepushed 22 commits to main • 65c2b93…10c5051 • 
6 days ago

[Infra] - Add some timeout and unwaive a test which dev fixed (NVIDIA…

QiJunepushed 7 commits to main • a8cf611…65c2b93 • 
6 days ago

test: [CI] Add failed cases into waives.txt (NVIDIA#5569)

QiJunepushed 1 commit to main • 9b17b29…a8cf611 • 
7 days ago