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
[Common][PyTorch] EP Fused prepare and dispatch for non-eager mode Documentation #18972: Pull request #3282 opened by phu0ngng
1m 41s phu0ngng:ep-count
phu0ngng:ep-count
1m 41s
Enable FA4 for context-parallel attention Documentation #18971: Pull request #3149 synchronize by sudhakarsingh27
1m 41s sudhakarsingh27:fa4-cp-exploration
sudhakarsingh27:fa4-cp-exploration
1m 41s
Plumb FP8+THD Documentation #18970: Pull request #2994 synchronize by sudhakarsingh27
1m 26s sudhakarsingh27:fp8_thd_attention_try2
sudhakarsingh27:fp8_thd_attention_try2
1m 26s
Enable FA4 for context-parallel attention Documentation #18969: Pull request #3149 synchronize by sudhakarsingh27
1m 36s sudhakarsingh27:fa4-cp-exploration
sudhakarsingh27:fa4-cp-exploration
1m 36s
[Common] Fix NVFP4 stochastic rounding on architectures without cvt.rs Documentation #18968: Pull request #3281 opened by davidkny22
Action required davidkny22:fix/nvfp4-sr-software-fallback
davidkny22:fix/nvfp4-sr-software-fallback
Action required
[Common] Experimental CuTeDSL MXFP8 backends in C++ via TVM-FFI Documentation #18967: Pull request #3137 synchronize by kainzhong
4m 17s kainzhong:cutedsl_mxfp8_common
kainzhong:cutedsl_mxfp8_common
4m 17s
[JAX] [PyT] [Common] Enable D=256 BWD cuDNN fused attn for Blackwell CC 10.x Documentation #18966: Pull request #3056 synchronize by pre-commit-ci Bot
2m 54s KshitijLakhani:klakhani/feat/enable-headdim256-bwd-sm100
KshitijLakhani:klakhani/feat/enable-headdim256-bwd-sm100
2m 54s
[JAX] [PyT] [Common] Enable D=256 BWD cuDNN fused attn for Blackwell CC 10.x Documentation #18965: Pull request #3056 synchronize by KshitijLakhani
1m 35s KshitijLakhani:klakhani/feat/enable-headdim256-bwd-sm100
KshitijLakhani:klakhani/feat/enable-headdim256-bwd-sm100
1m 35s
[Common] Experimental CuTeDSL MXFP8 backends in C++ via TVM-FFI Documentation #18964: Pull request #3137 synchronize by kainzhong
1m 33s kainzhong:cutedsl_mxfp8_common
kainzhong:cutedsl_mxfp8_common
1m 33s
[Common] Remove nv-internal-* comments Documentation #18963: Pull request #3280 synchronize by ksivaman
2m 16s ksivaman:rm_nv_internal_comments
ksivaman:rm_nv_internal_comments
2m 16s
[Common] Remove nv-internal-* comments Documentation #18962: Pull request #3280 opened by ksivaman
59s ksivaman:rm_nv_internal_comments
ksivaman:rm_nv_internal_comments
59s
[Common] Experimental CuTeDSL MXFP8 backends in C++ via TVM-FFI Documentation #18961: Pull request #3137 synchronize by pre-commit-ci Bot
2m 3s kainzhong:cutedsl_mxfp8_common
kainzhong:cutedsl_mxfp8_common
2m 3s
[Common] Experimental CuTeDSL MXFP8 backends in C++ via TVM-FFI Documentation #18960: Pull request #3137 synchronize by kainzhong
1m 17s kainzhong:cutedsl_mxfp8_common
kainzhong:cutedsl_mxfp8_common
1m 17s
[Pytorch] Fix RecursionError when dequantizing cpu-resident QuantizedTensor Documentation #18959: Pull request #3279 opened by janekb04
1m 37s janekb04:fix_recursion_error
janekb04:fix_recursion_error
1m 37s
Add cuDNN FE to source build requirements Documentation #18958: Pull request #3278 opened by vcherepanov-nv
1m 47s vcherepanov-nv:readme-update-dependencies
vcherepanov-nv:readme-update-dependencies
1m 47s
[JAX] EP Dispatch with overflow detection option Documentation #18957: Pull request #3277 opened by phu0ngng
1m 22s phu0ngng:ep-jax-overflow
phu0ngng:ep-jax-overflow
1m 22s
Activation + GroupedLinear Fusion for MOE and other MOE optimizations Documentation #18956: Pull request #3238 synchronize by vthumbe1503
1m 40s vthumbe1503:grouped_linear_act_fusion
vthumbe1503:grouped_linear_act_fusion
1m 40s
[Pytorch] [NCCL EP] Allow zero tokens for an EP rank in eager mode Documentation #18955: Pull request #3276 synchronize by pre-commit-ci Bot
1m 57s YangFei1990:allow_ep_rank_no_data
YangFei1990:allow_ep_rank_no_data
1m 57s
[Pytorch] [NCCL EP] Allow zero tokens for an EP rank in eager mode Documentation #18954: Pull request #3276 opened by YangFei1990
1m 14s YangFei1990:allow_ep_rank_no_data
YangFei1990:allow_ep_rank_no_data
1m 14s
[Common] Update Build Flow for Rubin-readiness Documentation #18953: Pull request #3275 synchronize by denera
1m 57s denera:common/rubin-build-flow
denera:common/rubin-build-flow
1m 57s
Activation + GroupedLinear Fusion for MOE and other MOE optimizations Documentation #18952: Pull request #3238 synchronize by vthumbe1503
1m 34s vthumbe1503:grouped_linear_act_fusion
vthumbe1503:grouped_linear_act_fusion
1m 34s
[Common] Update Build Flow for Rubin-readiness Documentation #18951: Pull request #3275 opened by denera
9m 16s denera:common/rubin-build-flow
denera:common/rubin-build-flow
9m 16s
[CI] Pin JAX image to 2026-07-21 Documentation #18950: Pull request #3250 synchronize by ksivaman
1m 31s fheinecke:fred/pin-ci-jax-image-1
fheinecke:fred/pin-ci-jax-image-1
1m 31s
Fix CUDA graph parameter grad lifetime Documentation #18949: Pull request #2937 synchronize by buptzyb
2m 0s buptzyb:fix/cudagraph-wgrad-lifetime
buptzyb:fix/cudagraph-wgrad-lifetime
2m 0s
Add per-sequence causal policy to packed THD attention Documentation #18948: Pull request #3274 opened by desh2608
1m 52s desh2608:desh/mixed-thd-pr-minimal
desh2608:desh/mixed-thd-pr-minimal
1m 52s