Skip to content

Actions: NVIDIA/TransformerEngine

Actions

Documentation

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
docs.yml
2,500+ workflow runs
2,500+ workflow runs

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
Add support for fused Q Up-Proj GEMM/RoPE/Quant. Documentation #19149: Pull request #3303 synchronize by sraman-rgb
1m 32s chaseblock:qrope_fusion
chaseblock:qrope_fusion
1m 32s
Use a temp directory for CMake build Documentation #19148: Pull request #3254 synchronize by fheinecke
2m 5s fheinecke:fred/cmake-build-isolation-1
fheinecke:fred/cmake-build-isolation-1
2m 5s
Sequential Ops for Dispatch/Combine + Wiring Extra Output-Extra Input via channels in Sequential Infra Documentation #19147: Pull request #3319 synchronize by pre-commit-ci Bot
2m 4s vthumbe1503:extra_out_wiring
vthumbe1503:extra_out_wiring
2m 4s
Sequential Ops for Dispatch/Combine + Wiring Extra Output-Extra Input via channels in Sequential Infra Documentation #19146: Pull request #3319 opened by vthumbe1503
1m 32s vthumbe1503:extra_out_wiring
vthumbe1503:extra_out_wiring
1m 32s
[PyTorch] [torch.compile] torch.compile support for Linear Documentation #19145: Pull request #3053 synchronize by pggPL
1m 32s pggPL:linear_torch_compile_final_attempt
pggPL:linear_torch_compile_final_attempt
1m 32s
[PyTorch] [torch.compile] torch.compile support for Linear Documentation #19144: Pull request #3053 synchronize by pre-commit-ci Bot
1m 57s pggPL:linear_torch_compile_final_attempt
pggPL:linear_torch_compile_final_attempt
1m 57s
[PyTorch] [torch.compile] torch.compile support for Linear Documentation #19143: Pull request #3053 synchronize by pggPL
1m 35s pggPL:linear_torch_compile_final_attempt
pggPL:linear_torch_compile_final_attempt
1m 35s
Add support for fused Q Up-Proj GEMM/RoPE/Quant. Documentation #19142: Pull request #3303 synchronize by chaseblock
1m 31s chaseblock:qrope_fusion
chaseblock:qrope_fusion
1m 31s
Add support for fused Q Up-Proj GEMM/RoPE/Quant. Documentation #19141: Pull request #3303 synchronize by chaseblock
2m 4s chaseblock:qrope_fusion
chaseblock:qrope_fusion
2m 4s
Add support for fused Q Up-Proj GEMM/RoPE/Quant. Documentation #19140: Pull request #3303 synchronize by pre-commit-ci Bot
1m 2s chaseblock:qrope_fusion
chaseblock:qrope_fusion
1m 2s
Add support for fused Q Up-Proj GEMM/RoPE/Quant. Documentation #19139: Pull request #3303 synchronize by chaseblock
1m 31s chaseblock:qrope_fusion
chaseblock:qrope_fusion
1m 31s
[PyTorch] Advance FusedAdam step counter for empty param groups Documentation #19138: Pull request #3318 opened by adityasingh2400
Action required adityasingh2400:fix-fused-adam-empty-group-step
adityasingh2400:fix-fused-adam-empty-group-step
Action required
[PyTorch] Preserve FP8 recompute state for inner autocast Documentation #19137: Pull request #3284 synchronize by nvegesna-netizen
1m 51s nvegesna-netizen:fix/fp8-inner-autocast-recompute
nvegesna-netizen:fix/fp8-inner-autocast-recompute
1m 51s
[Common/PyTorch] Grouped weighted-SwiGLU MXFP8 kernel Documentation #19136: Pull request #3315 synchronize by cael-ling
Action required cael-ling:feature/mxfp8-group-swiglu-recompute
cael-ling:feature/mxfp8-group-swiglu-recompute
Action required
[Common/PyTorch] Grouped weighted-SwiGLU MXFP8 kernel Documentation #19135: Pull request #3315 synchronize by cael-ling
Action required cael-ling:feature/mxfp8-group-swiglu-recompute
cael-ling:feature/mxfp8-group-swiglu-recompute
Action required
[Common] Fix pointer arithmatic to generate correct LDS/STS instructions Documentation #19134: Pull request #3291 synchronize by kainzhong
1m 35s kainzhong:mxfp8_generic_kernel_perf_fix
kainzhong:mxfp8_generic_kernel_perf_fix
1m 35s
[Common][PyTorch] Fuse the RHT into grouped NVFP4 quantize on non-SM100 architectures Documentation #19133: Pull request #3317 synchronize by pre-commit-ci Bot
Action required davidkny22:fuse-grouped-rht-nonsm100
davidkny22:fuse-grouped-rht-nonsm100
Action required
[Common][PyTorch] Fuse the RHT into grouped NVFP4 quantize on non-SM100 architectures Documentation #19132: Pull request #3317 synchronize by davidkny22
Action required davidkny22:fuse-grouped-rht-nonsm100
davidkny22:fuse-grouped-rht-nonsm100
Action required
[Common/PyTorch] Grouped weighted-SwiGLU MXFP8 kernel Documentation #19131: Pull request #3315 synchronize by cael-ling
Action required cael-ling:feature/mxfp8-group-swiglu-recompute
cael-ling:feature/mxfp8-group-swiglu-recompute
Action required
[Common][PyTorch] Fuse the RHT into grouped NVFP4 quantize on non-SM100 architectures Documentation #19130: Pull request #3317 opened by davidkny22
Action required davidkny22:fuse-grouped-rht-nonsm100
davidkny22:fuse-grouped-rht-nonsm100
Action required
[Pytorch] [NCCL EP] Allow zero tokens for an EP rank in eager mode Documentation #19129: Pull request #3276 synchronize by phu0ngng
1m 37s YangFei1990:allow_ep_rank_no_data
YangFei1990:allow_ep_rank_no_data
1m 37s
[Common] Fix pointer arithmatic to generate correct LDS/STS instructions Documentation #19128: Pull request #3291 synchronize by kainzhong
1m 41s kainzhong:mxfp8_generic_kernel_perf_fix
kainzhong:mxfp8_generic_kernel_perf_fix
1m 41s
[PyTorch] Preserve FP8 recompute state for inner autocast Documentation #19127: Pull request #3284 synchronize by nvegesna-netizen
1m 44s nvegesna-netizen:fix/fp8-inner-autocast-recompute
nvegesna-netizen:fix/fp8-inner-autocast-recompute
1m 44s
[PyTorch] Preserve FP8 recompute state for inner autocast Documentation #19126: Pull request #3284 synchronize by nvegesna-netizen
1m 35s nvegesna-netizen:fix/fp8-inner-autocast-recompute
nvegesna-netizen:fix/fp8-inner-autocast-recompute
1m 35s
[PyTorch] Preserve FP8 recompute state for inner autocast Documentation #19125: Pull request #3284 synchronize by nvegesna-netizen
1m 51s nvegesna-netizen:fix/fp8-inner-autocast-recompute
nvegesna-netizen:fix/fp8-inner-autocast-recompute
1m 51s