Skip to content

Actions: cyanguwa/TransformerEngine

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
60 workflow runs
60 workflow runs

Filter by Workflow

⌥ + click/return to exclude

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] Improved fused MoE aux loss kernel for large # of experts (#… Deploy nightly docs #239: Commit b9df401 pushed by cyanguwa
1m 41s main
main
1m 41s
[PyTorch] Guard/document single parameter feature for grouped linear … Deploy nightly docs #238: Commit 528f16c pushed by cyanguwa
2m 1s main
main
2m 1s
Correctly pad scaling factor inverses to satisfy cuteDSL requirements… Deploy nightly docs #237: Commit 01aef4f pushed by cyanguwa
1m 47s main
main
1m 47s
fix: TransformerEngineBaseModule quantizers init values type (#2927) Deploy nightly docs #236: Commit f2e31db pushed by cyanguwa
1m 35s main
main
1m 35s
Make NS coefficients parameter 2D in Python API (#2904) Deploy nightly docs #235: Commit 3c62f42 pushed by cyanguwa
5m 43s main
main
5m 43s
Strip local version labels from package version checks (#2858) Deploy nightly docs #234: Commit c7205a7 pushed by cyanguwa
1m 35s main
main
1m 35s
[JAX] Add warning if using BSHD and max_segments_per_seq > 1 (#2796) Deploy nightly docs #233: Commit f4debf6 pushed by cyanguwa
1m 30s main
main
1m 30s
fix: prevent MXFP8 amax aliasing with dSoftmaxOffset in bwd TE-CI Trigger #1: Issue comment #4 (comment) created by vedaanta
1s
1s
fix: prevent MXFP8 amax aliasing with dSoftmaxOffset in bwd Blossom-CI #3: Issue comment #4 (comment) created by vedaanta
1s
1s
[PyTorch][Fused Attn] Add support for cuDNN to return Softmax Stats… Deploy nightly docs #232: Commit e879bf8 pushed by cyanguwa
1m 28s main
main
1m 28s
[PyTorch] torch.compile support for permutation functions (#2686) Deploy nightly docs #231: Commit 3e61687 pushed by cyanguwa
1m 26s main
main
1m 26s
[PyTorch] Fix L3 FA tests (#2709) Deploy nightly docs #230: Commit 3ecb5bf pushed by cyanguwa
1m 42s main
main
1m 42s
[JAX] Fix FSDP when FSDP+EP is active (#2649) Deploy nightly docs #229: Commit f8b271f pushed by cyanguwa
1m 36s main
main
1m 36s
Changed VERSION to 2.14.0.dev0 Deploy nightly docs #228: Commit f122b07 pushed by cyanguwa
1m 29s main
main
1m 29s
[JAX] TE Permutation integration to Maxtext (#2672) Deploy nightly docs #227: Commit 5d112e3 pushed by cyanguwa
1m 22s main
main
1m 22s
[Common] MXFP8 kernel for grouped tensors (#2586) Deploy nightly docs #226: Commit 7393947 pushed by cyanguwa
1m 21s main
main
1m 21s
Fix exp2f_rcp to properly handle nan and 0xFE cases (#2647) Deploy nightly docs #225: Commit 71971e3 pushed by cyanguwa
1m 29s main
main
1m 29s
[JAX] Use "nyu-mll/glue" instead of "glue" for encoder datasets to fi… Deploy nightly docs #224: Commit 2104e4c pushed by cyanguwa
1m 30s main
main
1m 30s
Add logic for block-scaled tensors with GEMM swizzled scales (#2486) Deploy nightly docs #223: Commit 99df881 pushed by cyanguwa
1m 23s main
main
1m 23s
Debug doc generation (#2576) Deploy nightly docs #222: Commit 2f8ae81 pushed by cyanguwa
1m 25s main
main
1m 25s
[PyTorch] Add triton requirement (#2490) Deploy nightly docs #221: Commit 8c9f7c2 pushed by cyanguwa
1m 16s main
main
1m 16s
Fix bugs from refactoring C++ tensor class (#2481) Deploy nightly docs #220: Commit f0572aa pushed by cyanguwa
1m 26s main
main
1m 26s
[JAX] Remove unnecessary SWA calculation in _segment_ids_pos_to_seqle… Deploy nightly docs #219: Commit 4654b70 pushed by cyanguwa
1m 35s main
main
1m 35s
[PyTorch] Fix small errors (#2396) Deploy nightly docs #218: Commit 30c0120 pushed by cyanguwa
1m 31s main
main
1m 31s
[JAX] Quickstart documentation (#2310) Deploy nightly docs #217: Commit 42d2274 pushed by cyanguwa
1m 33s main
main
1m 33s