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
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20539: Pull request #3527 synchronize by nvegesna-netizen
1m 21s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 21s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20538: Pull request #3527 synchronize by nvegesna-netizen
1m 22s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 22s
Fix THD P2P pad detection and tail-zero gating Documentation #20537: Pull request #3506 synchronize by RPalmr
Action required RPalmr:fix/thd-p2p-pad-detect
RPalmr:fix/thd-p2p-pad-detect
Action required
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20536: Pull request #3527 synchronize by nvegesna-netizen
1m 48s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 48s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20535: Pull request #3527 synchronize by nvegesna-netizen
1m 55s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 55s
[Common] Experimental CuTeDSL MXFP8 backends in C++ via TVM-FFI Documentation #20534: Pull request #3137 synchronize by pre-commit-ci Bot
1m 25s kainzhong:cutedsl_mxfp8_common
kainzhong:cutedsl_mxfp8_common
1m 25s
[Common] Experimental CuTeDSL MXFP8 backends in C++ via TVM-FFI Documentation #20533: Pull request #3137 synchronize by kainzhong
1m 12s kainzhong:cutedsl_mxfp8_common
kainzhong:cutedsl_mxfp8_common
1m 12s
[PyTorch] Remove stale xfail for selective attention recompute Documentation #20532: Pull request #3531 opened by negvet
1m 31s negvet:fix_attn_ci_hybrid
negvet:fix_attn_ci_hybrid
1m 31s
[PyTorch] GDN2 support and linear attention refactor Documentation #20531: Pull request #3521 synchronize by ksivaman
1m 25s ksivaman:gdn2_attention
ksivaman:gdn2_attention
1m 25s
[PyTorch] GDN2 support and linear attention refactor Documentation #20530: Pull request #3521 synchronize by ksivaman
1m 30s ksivaman:gdn2_attention
ksivaman:gdn2_attention
1m 30s
[JAX] WAR for XLA associative scan rewriter crash Documentation #20529: Pull request #3525 synchronize by KshitijLakhani
1m 35s KshitijLakhani:klakhani/fix/associative-scan-rewriter-war
KshitijLakhani:klakhani/fix/associative-scan-rewriter-war
1m 35s
[JAX] Remove duplicate score-mod test run Documentation #20528: Pull request #3524 synchronize by KshitijLakhani
1m 36s KshitijLakhani:klakhani/fix/duplicate-score-mod-tests-in-ci
KshitijLakhani:klakhani/fix/duplicate-score-mod-tests-in-ci
1m 36s
[All] Bump minimum supported cuDNN version to 9.12 Documentation #20527: Pull request #3236 reopened by cyanguwa
1m 12s cyanguwa:update-min-cudnn-9.11
cyanguwa:update-min-cudnn-9.11
1m 12s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20526: Pull request #3527 synchronize by nvegesna-netizen
1m 31s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 31s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20525: Pull request #3527 synchronize by nvegesna-netizen
1m 36s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 36s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20524: Pull request #3527 synchronize by nvegesna-netizen
1m 23s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 23s
[PyTorch] Extend no-load-balance CP to the a2a comm type Documentation #20523: Pull request #3530 opened by Rudin6
Action required Rudin6:cp-a2a-no-load-balance
Rudin6:cp-a2a-no-load-balance
Action required
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20522: Pull request #3527 synchronize by nvegesna-netizen
1m 26s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 26s
[Common] Add non-TMA MXFP8 cast-only kernels for specialized rowwise-only and row+colwise Documentation #20521: Pull request #3459 synchronize by tdophung
1m 16s tdophung:mxfp8-register-cast
tdophung:mxfp8-register-cast
1m 16s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20520: Pull request #3527 synchronize by nvegesna-netizen
1m 23s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 23s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20519: Pull request #3527 synchronize by nvegesna-netizen
1m 30s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 30s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20518: Pull request #3527 synchronize by nvegesna-netizen
1m 21s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 21s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20517: Pull request #3527 synchronize by nvegesna-netizen
1m 18s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 18s
feat(attention): cuDNN FROST attention backend for head_dim in (256, 512], with context parallelism Documentation #20516: Pull request #3527 synchronize by nvegesna-netizen
1m 46s nvegesna-netizen:nvegesna/te-frost-d512-cp
nvegesna-netizen:nvegesna/te-frost-d512-cp
1m 46s
[PyTorch] Allow F16 GDN inside FP8 autocast Documentation #20515: Pull request #3529 opened by layalir
1m 28s layalir:layali/gdn-f16-recipe-override
layalir:layali/gdn-f16-recipe-override
1m 28s