Skip to content

Actions: NVIDIA/TransformerEngine

Actions

Blossom-CI

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
blossom-ci.yml
2,981 workflow runs
2,981 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

Update FP8 scale-inverse in kernels with FP8 output Blossom-CI #3663: Issue comment #1083 (comment) created by timmoon10
August 17, 2024 01:59 3s
August 17, 2024 01:59 3s
How to debug CUDNN_STATUS_EXECUTION_FAILED? Blossom-CI #3662: Issue comment #1116 (comment) created by vedantroy
August 17, 2024 01:50 3s
August 17, 2024 01:50 3s
Add FP8 support to CP implementation with KV P2P Blossom-CI #3661: Issue comment #1114 (comment) created by xrennvidia
August 17, 2024 00:13 5s
August 17, 2024 00:13 5s
[PyTorch] Add support for padding mask in UnfusedDotProductAttention Blossom-CI #3660: Issue comment #1073 (comment) created by cyanguwa
August 16, 2024 23:47 5s
August 16, 2024 23:47 5s
Update cudnn-frontend to v1.6.1 Blossom-CI #3659: Issue comment #1108 (comment) created by cyanguwa
August 16, 2024 23:46 4s
August 16, 2024 23:46 4s
How to debug CUDNN_STATUS_EXECUTION_FAILED? Blossom-CI #3658: Issue comment #1116 (comment) created by vedantroy
August 16, 2024 23:13 4s
August 16, 2024 23:13 4s
feat(pytorch): Allow TransformerLayer and MultiheadAttention to accept sequence length parameters Blossom-CI #3657: Issue comment #1066 (comment) created by ptrendx
August 16, 2024 20:46 4s
August 16, 2024 20:46 4s
create_communicator_grouped2 may trigger uninit value memory issue(randomly crash) when you train more iterations. Blossom-CI #3656: Issue comment #959 (comment) created by denera
August 16, 2024 20:29 4s
August 16, 2024 20:29 4s
tp_overlap need tensor parallel is equal world size ? Blossom-CI #3655: Issue comment #966 (comment) created by denera
August 16, 2024 20:27 4s
August 16, 2024 20:27 4s
[PyTorch] Support dtype casting in fused adam Blossom-CI #3654: Issue comment #977 (comment) created by timmoon10
August 16, 2024 20:02 4s
August 16, 2024 20:02 4s
Add FP8 support to CP implementation with KV P2P Blossom-CI #3653: Issue comment #1114 (comment) created by xrennvidia
August 16, 2024 18:38 5s
August 16, 2024 18:38 5s
[PyTorch] Support dtype casting in fused adam Blossom-CI #3652: Issue comment #977 (comment) created by timmoon10
August 16, 2024 18:00 5s
August 16, 2024 18:00 5s
[PyTorch] FP8 MHA with RoPE and Miscellaneous Improvements Blossom-CI #3651: Issue comment #1100 (comment) created by timmoon10
August 16, 2024 16:40 4s
August 16, 2024 16:40 4s
[JAX] Context Parallel Attention with All-Gather Blossom-CI #3650: Issue comment #1106 (comment) created by ptrendx
August 16, 2024 16:21 5s
August 16, 2024 16:21 5s
How to debug CUDNN_STATUS_EXECUTION_FAILED? Blossom-CI #3649: Issue comment #1116 (comment) created by ptrendx
August 16, 2024 16:13 4s
August 16, 2024 16:13 4s
When will comm-gemm-overlap support multi nodes? Blossom-CI #3648: Issue comment #1071 (comment) created by umiswing
August 16, 2024 14:13 4s
August 16, 2024 14:13 4s
Stop using global mesh for custom_partitioning. Blossom-CI #3647: Issue comment #1112 (comment) created by nouiz
August 16, 2024 14:10 5s
August 16, 2024 14:10 5s
Stop using global mesh for custom_partitioning. Blossom-CI #3646: Issue comment #1112 (comment) created by nouiz
August 16, 2024 14:10 5s
August 16, 2024 14:10 5s
[Draft] show how to not use a global mesh. Blossom-CI #3645: Issue comment #549 (comment) created by nouiz
August 16, 2024 13:17 4s
August 16, 2024 13:17 4s
[PyTorch] Debug CUDA graph support with operation-based API Blossom-CI #3644: Issue comment #1117 (comment) created by timmoon10
August 16, 2024 01:56 3s
August 16, 2024 01:56 3s
Unit test failed due to "CUDA Error: operation not supported" Blossom-CI #3643: Issue comment #1085 (comment) created by yanboliang
August 16, 2024 01:46 4s
August 16, 2024 01:46 4s
How to debug CUDNN_STATUS_EXECUTION_FAILED? Blossom-CI #3642: Issue comment #1116 (comment) created by vedantroy
August 16, 2024 00:03 5s
August 16, 2024 00:03 5s
Performance not improving when using CuDNN? Blossom-CI #3641: Issue comment #1113 (comment) created by cyanguwa
August 15, 2024 23:17 4s
August 15, 2024 23:17 4s
Performance not improving when using CuDNN? Blossom-CI #3640: Issue comment #1113 (comment) created by cyanguwa
August 15, 2024 22:55 4s
August 15, 2024 22:55 4s
How to use head dimension 256? Blossom-CI #3639: Issue comment #1115 (comment) created by cyanguwa
August 15, 2024 22:30 4s
August 15, 2024 22:30 4s
ProTip! You can narrow down the results and go further in time using created:<2024-08-15 or the other filters available.