[None][chore] Weekly mass integration of release/1.1#8918

Merged

mikeiovine merged 21 commits intoNVIDIA:mainfrom

mikeiovine:mass-integrate-1.1

Nov 20, 2025

Commits on Nov 19, 2025

[https://nvbugs/5569713][fix] Disable fp8 deep gemm for EXAONE-4.0-32B-FP8 (#8429)

JunyiXu-nv
authored and
mikeiovine
committed
[https://nvbugs/5515753][ci] Add NCCL_DEBUG=INFO flag to collect more info with CI failure. (#8440)

SimengLiu-nv
authored and
mikeiovine
committed
[https://nvbugs/5504095][fix] Unwaive test_user_specify_workspace case. (#8316)

nv-guomingz
authored and
mikeiovine
committed
[https://nvbugs/5546510][fix] Move torch.cuda.Stream out of torch com… (#8494)

liji-nv
authored and
mikeiovine
committed
[https://nvbugs/5565549][fix] unwaive test_disaggregated_spec_dec_bat… (#8500)

bo-nv
authored and
mikeiovine
committed
[https://nvbugs/5575829][fix] Unwaive gpt-oss test (#8576)

LinPoly
authored and
mikeiovine
committed
[https://nvbugs/5569754][fix] trtllm-llmapi-launch port conflict (#8582)

Superjomn
authored and
mikeiovine
committed
[https://nvbugs/5582277][fix] rework DisaggPPTerminationHandler to fix hang issue (#8519)

reasonsolo
authored and
mikeiovine
committed
[https://nvbugs/5575902][fix] set max_batch_size=1 to stabilize accuracy test result (#8609)

reasonsolo
authored and
mikeiovine
committed
[https://nvbugs/5587456][fix] Remove multimodal test cases using TRT backend (#8611)

jieli-matrix
authored and
mikeiovine
committed
[None][test] Clean cache for certain easily hang cases (#8619)

authored and
mikeiovine
committed
[https://nvbugs/5608489][fix] Fix output unpack issues for Llama3/4 NVFP4 models. (#8679)

hyukn
authored and
mikeiovine
committed
[https://nvbugs/5572320][fix] Ported test_ad_trtllm_bench.py from main (#8671)

MrGeva
authored and
mikeiovine
committed
[https://nvbugs/5564465][fix] Overwrite only if default_max_tokens is legal (#8538)

LinPoly
authored and
mikeiovine
committed
[https://nvbugs/5578175][fix] Fix block range index (#8470)

chuangz0
authored and
mikeiovine
committed
[https://nvbugs/5601203] [fix]Restrict fp8 blockscale moe case (#8583)

VALLIS-NERIA
authored and
mikeiovine
committed
[https://nvbugs/5575841] [test] Move test_moe.py to serial tests to improve stability + unwaive FP4 MoE torch unit tests (#8422)

DomBrown
authored and
mikeiovine
committed
[https://nvbugs/5488118][fix] Unwaive passed tests (#8758)

liji-nv
authored and
mikeiovine
committed
[None][infra] Remove invaild waived tests which not in release branch (#8841)

ZhanruiSunCh
authored and
mikeiovine
committed
[https://nvbugs/5325296][fix] Enable relaxed acceptance test on Blackwell (#8709)

Barry-Delaney
authored and
mikeiovine
committed
[None][chore] Update linter rules for mass integration
mikeiovine
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[None][chore] Weekly mass integration of release/1.1#8918

[None][chore] Weekly mass integration of release/1.1#8918
mikeiovine merged 21 commits intoNVIDIA:mainfrom
mikeiovine:mass-integrate-1.1

Commits on Nov 19, 2025

[https://nvbugs/5569713][fix] Disable fp8 deep gemm for EXAONE-4.0-32B-FP8 (#8429)

[https://nvbugs/5515753][ci] Add NCCL_DEBUG=INFO flag to collect more info with CI failure. (#8440)

[https://nvbugs/5504095][fix] Unwaive test_user_specify_workspace case. (#8316)

[https://nvbugs/5546510][fix] Move torch.cuda.Stream out of torch com… (#8494)

[https://nvbugs/5565549][fix] unwaive test_disaggregated_spec_dec_bat… (#8500)

[https://nvbugs/5575829][fix] Unwaive gpt-oss test (#8576)

[https://nvbugs/5569754][fix] trtllm-llmapi-launch port conflict (#8582)

[https://nvbugs/5582277][fix] rework DisaggPPTerminationHandler to fix hang issue (#8519)

[https://nvbugs/5575902][fix] set max_batch_size=1 to stabilize accuracy test result (#8609)

[https://nvbugs/5587456][fix] Remove multimodal test cases using TRT backend (#8611)

[None][test] Clean cache for certain easily hang cases (#8619)

[https://nvbugs/5608489][fix] Fix output unpack issues for Llama3/4 NVFP4 models. (#8679)

[https://nvbugs/5572320][fix] Ported test_ad_trtllm_bench.py from main (#8671)

[https://nvbugs/5564465][fix] Overwrite only if default_max_tokens is legal (#8538)

[https://nvbugs/5578175][fix] Fix block range index (#8470)

[https://nvbugs/5601203] [fix]Restrict fp8 blockscale moe case (#8583)

[https://nvbugs/5575841] [test] Move test_moe.py to serial tests to improve stability + unwaive FP4 MoE torch unit tests (#8422)

[https://nvbugs/5488118][fix] Unwaive passed tests (#8758)

[None][infra] Remove invaild waived tests which not in release branch (#8841)

[https://nvbugs/5325296][fix] Enable relaxed acceptance test on Blackwell (#8709)

[None][chore] Update linter rules for mass integration