Skip to content

Actions: NVIDIA/TransformerEngine

Actions

TE-CI Trigger

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
trigger-ci.yml
871 workflow run results
871 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

With using the fp8, after the interruption of training, and then continue , there may be a little difference in loss. Is this caused by the fp8 mechanism? TE-CI Trigger #2557: Issue comment #759 (comment) created by ptrendx
April 10, 2024 23:35 4s
April 10, 2024 23:35 4s
Replacing nn.Linear w/ te.Linear FP8 convergence issue TE-CI Trigger #2556: Issue comment #764 (comment) created by ptrendx
April 10, 2024 23:26 5s
April 10, 2024 23:26 5s
Build fails when using jax NGC image TE-CI Trigger #2555: Issue comment #767 (comment) created by ptrendx
April 10, 2024 23:22 4s
April 10, 2024 23:22 4s
CPU Overhead of te.Linear FP8 Layers TE-CI Trigger #2554: Issue comment #761 (comment) created by ptrendx
April 10, 2024 22:40 5s
April 10, 2024 22:40 5s
CPU Overhead of te.Linear FP8 Layers TE-CI Trigger #2553: Issue comment #761 (comment) created by gedoensmax
April 10, 2024 21:36 4s
April 10, 2024 21:36 4s
CPU Overhead of te.Linear FP8 Layers TE-CI Trigger #2552: Issue comment #761 (comment) created by tohinz
April 10, 2024 20:06 4s
April 10, 2024 20:06 4s
[PyTorch] Stop storing fused weight tensor in linear modules TE-CI Trigger #2551: Issue comment #719 (comment) created by ksivaman
April 10, 2024 19:00 3m 52s
April 10, 2024 19:00 3m 52s
[PyTorch] Stop storing fused weight tensor in linear modules TE-CI Trigger #2550: Issue comment #719 (comment) created by deepakn94
April 10, 2024 18:56 2m 32s
April 10, 2024 18:56 2m 32s
CPU Overhead of te.Linear FP8 Layers TE-CI Trigger #2549: Issue comment #761 (comment) created by gedoensmax
April 10, 2024 17:39 4s
April 10, 2024 17:39 4s
Fix memory fence for ARM TE-CI Trigger #2548: Issue comment #753 (comment) created by timmoon10
April 10, 2024 16:53 3m 52s
April 10, 2024 16:53 3m 52s
CPU Overhead of te.Linear FP8 Layers TE-CI Trigger #2547: Issue comment #761 (comment) created by gedoensmax
April 10, 2024 13:20 5s
April 10, 2024 13:20 5s
[PyTorch][duplicate][CI] FP8 cuda graphs TE-CI Trigger #2546: Issue comment #766 (comment) created by ksivaman
April 10, 2024 07:21 3m 41s
April 10, 2024 07:21 3m 41s
[PyTorch] cuda graph support TE-CI Trigger #2545: Issue comment #575 (comment) created by ksivaman
April 10, 2024 07:21 3m 54s
April 10, 2024 07:21 3m 54s
When A and B are fp8 tensors, the compute type could be CUBLAS_COMPUTE_16F TE-CI Trigger #2544: Issue comment #758 (comment) created by condy0919
April 10, 2024 07:04 4s
April 10, 2024 07:04 4s
[JAX] Unifying GeLU and GeGLU in LayerNorm MLP TE-CI Trigger #2543: Issue comment #765 (comment) created by zlsh80826
April 10, 2024 06:55 3m 40s
April 10, 2024 06:55 3m 40s
_ZN18transformer_engine6getenvIiEET_RKSsRKS1_ on the latest main branch TE-CI Trigger #2542: Issue comment #756 (comment) created by yueyericardo
April 10, 2024 05:36 5s
April 10, 2024 05:36 5s
Fix undefined symbol issue for transformer_engine::getenv TE-CI Trigger #2541: Issue comment #763 (comment) created by timmoon10
April 10, 2024 02:27 4s
April 10, 2024 02:27 4s
_ZN18transformer_engine6getenvIiEET_RKSsRKS1_ on the latest main branch TE-CI Trigger #2540: Issue comment #756 (comment) created by timmoon10
April 10, 2024 02:25 4s
April 10, 2024 02:25 4s
Fix undefined symbol issue for transformer_engine::getenv TE-CI Trigger #2539: Issue comment #763 (comment) created by timmoon10
April 10, 2024 02:06 35m 45s
April 10, 2024 02:06 35m 45s
Support Low Rank Adaptation (LoRA). TE-CI Trigger #2538: Issue comment #745 (comment) created by mingxu1067
April 10, 2024 01:08 3m 46s
April 10, 2024 01:08 3m 46s
Replacing nn.Linear w/ te.Linear FP8 convergence issue TE-CI Trigger #2537: Issue comment #764 (comment) created by viclzhu
April 9, 2024 21:01 5s
April 9, 2024 21:01 5s
[Question] Why Tensor parallel communication/GEMM overlap can happen only when sequence parallelism is enabled? TE-CI Trigger #2536: Issue comment #746 (comment) created by ptrendx
April 9, 2024 20:48 4s
April 9, 2024 20:48 4s
Replacing nn.Linear w/ te.Linear FP8 convergence issue TE-CI Trigger #2535: Issue comment #764 (comment) created by ptrendx
April 9, 2024 20:46 5s
April 9, 2024 20:46 5s
[UB] Adding configurable timeout for userbuffer and improving error reporting for potential hangs TE-CI Trigger #2534: Issue comment #757 (comment) created by ksivaman
April 9, 2024 17:06 4m 25s
April 9, 2024 17:06 4m 25s
[PyTorch] cuda graph support TE-CI Trigger #2533: Issue comment #575 (comment) created by ksivaman
April 9, 2024 06:23 4m 0s
April 9, 2024 06:23 4m 0s