Skip to content

Actions: NVIDIA/TransformerEngine

Actions

Build

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
build.yml
3,659 workflow runs
3,659 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

[TE/JAX] Disable FusedAttn with FFI by default Build #6043: Pull request #1298 opened by phu0ngng
October 29, 2024 20:47 1h 10m 33s phu0ngng:fused_attn_ffi
phu0ngng:fused_attn_ffi
October 29, 2024 20:47 1h 10m 33s
[TE/JAX] Custom call with FFI - lowering all attributes with bind all Build #6042: Pull request #1289 synchronize by pre-commit-ci bot
October 29, 2024 20:36 1h 9m 32s phu0ngng:ffi_dict_attrs
phu0ngng:ffi_dict_attrs
October 29, 2024 20:36 1h 9m 32s
[TE/JAX] Custom call with FFI - lowering all attributes with bind all Build #6041: Pull request #1289 synchronize by phu0ngng
October 29, 2024 20:35 1h 9m 59s phu0ngng:ffi_dict_attrs
phu0ngng:ffi_dict_attrs
October 29, 2024 20:35 1h 9m 59s
[TE/JAX] Custom call with FFI - lowering all attributes with bind all Build #6040: Pull request #1289 synchronize by phu0ngng
October 29, 2024 20:16 1h 15m 7s phu0ngng:ffi_dict_attrs
phu0ngng:ffi_dict_attrs
October 29, 2024 20:16 1h 15m 7s
[C/PyTorch] Userbuffers and comm+GEMM overlap algorithms refactored and moved to TE/common Build #6039: Pull request #1067 synchronize by denera
October 29, 2024 15:05 1h 12m 53s denera:comm-gemm-overlap-refactor
denera:comm-gemm-overlap-refactor
October 29, 2024 15:05 1h 12m 53s
[JAX] Consolidate FFI and old descriptor implementation for fused attention. Build #6038: Pull request #1295 synchronize by mgoldfarb-nvidia
October 29, 2024 13:52 1h 14m 49s mgoldfarb-nvidia:mgoldfarb/cleanup_ffi_fused_attn
mgoldfarb-nvidia:mgoldfarb/cleanup_ffi_fused_attn
October 29, 2024 13:52 1h 14m 49s
Support using fp16 master weights and fp16/fp8 optimizer states in FusedAdam Build #6037: Pull request #1078 synchronize by pre-commit-ci bot
October 29, 2024 12:52 Action required kunlunl:mx_fp16
kunlunl:mx_fp16
October 29, 2024 12:52 Action required
Support using fp16 master weights and fp16/fp8 optimizer states in FusedAdam Build #6036: Pull request #1078 synchronize by kunlunl
October 29, 2024 12:51 Action required kunlunl:mx_fp16
kunlunl:mx_fp16
October 29, 2024 12:51 Action required
[PyTorch] Make FP8 MHA work with RoPE when CP is on Build #6035: Pull request #1297 synchronize by pre-commit-ci bot
October 29, 2024 10:06 1h 7m 49s yaox12:xiny/fp8_mha_with_rope_cp
yaox12:xiny/fp8_mha_with_rope_cp
October 29, 2024 10:06 1h 7m 49s
[PyTorch] Make FP8 MHA work with RoPE when CP is on Build #6034: Pull request #1297 opened by yaox12
October 29, 2024 10:06 1h 8m 30s yaox12:xiny/fp8_mha_with_rope_cp
yaox12:xiny/fp8_mha_with_rope_cp
October 29, 2024 10:06 1h 8m 30s
Add missed arguments of apply_rotary_pos_emb in MHA Build #6033: Pull request #1296 reopened by xrennvidia
October 29, 2024 07:32 1h 10m 5s xrennvidia:xren/cp_mha_rope
xrennvidia:xren/cp_mha_rope
October 29, 2024 07:32 1h 10m 5s
Add missed arguments of apply_rotary_pos_emb in MHA Build #6032: Pull request #1296 synchronize by xrennvidia
October 29, 2024 06:39 1h 8m 46s xrennvidia:xren/cp_mha_rope
xrennvidia:xren/cp_mha_rope
October 29, 2024 06:39 1h 8m 46s
Add missed arguments of apply_rotary_pos_emb in MHA Build #6031: Pull request #1296 synchronize by xrennvidia
October 29, 2024 06:37 1h 8m 16s xrennvidia:xren/cp_mha_rope
xrennvidia:xren/cp_mha_rope
October 29, 2024 06:37 1h 8m 16s
Add missed arguments of apply_rotary_pos_emb in MHA Build #6030: Pull request #1296 synchronize by pre-commit-ci bot
October 29, 2024 06:36 1h 9m 46s xrennvidia:xren/cp_mha_rope
xrennvidia:xren/cp_mha_rope
October 29, 2024 06:36 1h 9m 46s
Add missed arguments of apply_rotary_pos_emb in MHA Build #6029: Pull request #1296 opened by xrennvidia
October 29, 2024 06:35 1h 4m 47s xrennvidia:xren/cp_mha_rope
xrennvidia:xren/cp_mha_rope
October 29, 2024 06:35 1h 4m 47s
[PyTorch] MultiheadAttention: Pass cu_seqlens to apply_rotary_pos_emb Build #6028: Pull request #1279 synchronize by yaox12
October 29, 2024 04:28 1h 7m 56s Marks101:fix-mha-rotary-cu-seqlens
Marks101:fix-mha-rotary-cu-seqlens
October 29, 2024 04:28 1h 7m 56s
[JAX] Consolidate FFI and old descriptor implementation for fused attention. Build #6027: Pull request #1295 synchronize by mgoldfarb-nvidia
October 28, 2024 19:42 1h 8m 26s mgoldfarb-nvidia:mgoldfarb/cleanup_ffi_fused_attn
mgoldfarb-nvidia:mgoldfarb/cleanup_ffi_fused_attn
October 28, 2024 19:42 1h 8m 26s
[C/PyTorch] Userbuffers and comm+GEMM overlap algorithms refactored and moved to TE/common Build #6026: Pull request #1067 synchronize by denera
October 28, 2024 18:50 1h 8m 35s denera:comm-gemm-overlap-refactor
denera:comm-gemm-overlap-refactor
October 28, 2024 18:50 1h 8m 35s
[C/PyTorch] Userbuffers and comm+GEMM overlap algorithms refactored and moved to TE/common Build #6025: Pull request #1067 synchronize by denera
October 28, 2024 18:45 1h 11m 15s denera:comm-gemm-overlap-refactor
denera:comm-gemm-overlap-refactor
October 28, 2024 18:45 1h 11m 15s
[JAX] Consolidate FFI and old descriptor implementation for fused attention. Build #6024: Pull request #1295 synchronize by mgoldfarb-nvidia
October 28, 2024 18:28 21m 5s mgoldfarb-nvidia:mgoldfarb/cleanup_ffi_fused_attn
mgoldfarb-nvidia:mgoldfarb/cleanup_ffi_fused_attn
October 28, 2024 18:28 21m 5s
[JAX] Consolidate FFI and old descriptor implementation for fused attention. Build #6023: Pull request #1295 opened by mgoldfarb-nvidia
October 28, 2024 18:14 21m 26s mgoldfarb-nvidia:mgoldfarb/cleanup_ffi_fused_attn
mgoldfarb-nvidia:mgoldfarb/cleanup_ffi_fused_attn
October 28, 2024 18:14 21m 26s
fused out correction in CP Build #6022: Pull request #1248 synchronize by xiaoyao0115
October 28, 2024 10:02 Action required xiaoyao0115:fused_out_correction
xiaoyao0115:fused_out_correction
October 28, 2024 10:02 Action required
[PyTorch] Skip t3hd/th3d for MQA/GQA tests Build #6021: Pull request #1293 opened by cyanguwa
October 28, 2024 02:42 1h 13m 43s cyanguwa:skip_thd_tests
cyanguwa:skip_thd_tests
October 28, 2024 02:42 1h 13m 43s
[JAX] Expose cp params to jax DPA api Build #6020: Pull request #1292 opened by kocchop
October 27, 2024 09:32 1h 7m 39s kocchop:faysal/expose-cp-to-jax-dpa
kocchop:faysal/expose-cp-to-jax-dpa
October 27, 2024 09:32 1h 7m 39s
[PyTorch] MultiheadAttention: Pass cu_seqlens to apply_rotary_pos_emb Build #6019: Pull request #1279 synchronize by yaox12
October 27, 2024 05:04 1h 7m 40s Marks101:fix-mha-rotary-cu-seqlens
Marks101:fix-mha-rotary-cu-seqlens
October 27, 2024 05:04 1h 7m 40s