Skip to content

Actions: zhongbozhu/TransformerEngine

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
72 workflow runs
72 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
Fix build issues with 26.08 JAX container images (#3404) Deploy nightly docs #72: Commit 250ef1b pushed by zhongbozhu
1m 51s main
main
1m 51s
[JAX] Support synchronous collectives in HLO collective bytes assert … Deploy nightly docs #71: Commit 27b12df pushed by zhongbozhu
1m 43s main
main
1m 43s
Improve device-init grouped linear module with single grouped weight … Deploy nightly docs #70: Commit 2d80391 pushed by zhongbozhu
1m 36s main
main
1m 36s
[Common] Fix pointer arithmatic to generate correct LDS/STS instructi… Deploy nightly docs #69: Commit f07a860 pushed by zhongbozhu
2m 3s main
main
2m 3s
Add support for fused Q Up-Proj GEMM/RoPE/Quant. (#3303) Deploy nightly docs #68: Commit dd96595 pushed by zhongbozhu
2m 13s main
main
2m 13s
[common] Fix UE8M0 code 0 (2^-127) and code 255 (NaN) expansion in pt… Deploy nightly docs #67: Commit 4ff5fdf pushed by zhongbozhu
2m 4s main
main
2m 4s
[PyTorch] Fix shape and size() for columnwise-only quantized tensors … Deploy nightly docs #66: Commit 8685aa0 pushed by zhongbozhu
2m 4s main
main
2m 4s
Single grouped weight fixes (#3225) Deploy nightly docs #65: Commit 35eef89 pushed by zhongbozhu
2m 14s main
main
2m 14s
[Common] Make MXFP8 specialized kernel respect noop flag as well (#3246) Deploy nightly docs #64: Commit ea41e08 pushed by zhongbozhu
2m 8s main
main
2m 8s
[Common] Fix Build: NCCL EP build to respect MAX_JOBS (#3138) Deploy nightly docs #63: Commit 4adad4c pushed by zhongbozhu
1m 57s main
main
1m 57s
Fix nccl-extensions submodule name (#3239) Deploy nightly docs #62: Commit 8280f02 pushed by zhongbozhu
1m 48s main
main
1m 48s
Fix FusedAdam empty tensor handling (#3212) Deploy nightly docs #61: Commit 868d8d9 pushed by zhongbozhu
2m 0s main
main
2m 0s
Update list of authorized CI users (#3211) Deploy nightly docs #60: Commit eabdd46 pushed by zhongbozhu
1m 40s main
main
1m 40s
TE EP integration to MoEBlock (#3116) Deploy nightly docs #59: Commit b782ea6 pushed by zhongbozhu
1m 39s main
main
1m 39s
Install cuDNN frontend in GitHub builds (#3196) Deploy nightly docs #58: Commit 3bc89b9 pushed by zhongbozhu
1m 41s main
main
1m 41s
Fix the numerical issue with the Fused CE loss (#3193) Deploy nightly docs #57: Commit 01bbe3e pushed by zhongbozhu
1m 46s main
main
1m 46s
[Common] Blackwell skip condition for C++ grouped FP8 block-scaling t… Deploy nightly docs #56: Commit 7cb8b31 pushed by zhongbozhu
1m 37s main
main
1m 37s
[Common] Fix int32 overflow in multi_tensor_apply tensor sizes for nu… Deploy nightly docs #55: Commit 5d1dddf pushed by zhongbozhu
1m 52s main
main
1m 52s
Revert "Expert Parallelism: common C API + NCCL EP backend" (#3126) Deploy nightly docs #54: Commit 547d284 pushed by zhongbozhu
5m 39s main
main
5m 39s
[PyTorch] Add op-level activation offload opt-out API (#3108) Deploy nightly docs #53: Commit 9b06f26 pushed by zhongbozhu
1m 41s main
main
1m 41s
[PyTorch] Debug CPU offloading in grouped linear and grouped MLP (#3047) Deploy nightly docs #52: Commit 3fffa55 pushed by zhongbozhu
1m 33s main
main
1m 33s
[PyTorch] Fix wrong stream capture for cuteDSL delayed wgrad GEMM (#3… Deploy nightly docs #51: Commit 23a3f54 pushed by zhongbozhu
1m 50s main
main
1m 50s
Optimize grouped split metadata preparation (#3075) Deploy nightly docs #50: Commit b38c16b pushed by zhongbozhu
1m 46s main
main
1m 46s
Adds GEMM Profiling Guide to TE (#2863) Deploy nightly docs #49: Commit 1609c89 pushed by zhongbozhu
1m 35s main
main
1m 35s
[JAX] Add an MoE Block (Layer) that compound router, permutation, gro… Deploy nightly docs #48: Commit d1920cf pushed by zhongbozhu
1m 38s main
main
1m 38s