Skip to content

Actions: NVIDIA/TransformerEngine

Actions

Documentation

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
docs.yml
2,500+ workflow runs
2,500+ workflow runs

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
[JAX] Calculate seqlens and offsets in O(T) space instead of O(T*T) space for THD sequences Documentation #16617: Pull request #2522 synchronize by KshitijLakhani
1m 19s KshitijLakhani:klakhani/mem-optimize-seqlens-offsets-thd
KshitijLakhani:klakhani/mem-optimize-seqlens-offsets-thd
1m 19s
[PyTorch][Core] Fix CUBLAS GGEMM when weight dims are not divisible by 128 Documentation #16616: Pull request #2954 synchronize by pre-commit-ci Bot
1m 17s vthumbe1503:fix_cublas_grouped_gemm_gptoss_sizes
vthumbe1503:fix_cublas_grouped_gemm_gptoss_sizes
1m 17s
[PyTorch][Core] Fix CUBLAS GGEMM when weight dims are not divisible by 128 Documentation #16615: Pull request #2954 opened by vthumbe1503
1m 15s vthumbe1503:fix_cublas_grouped_gemm_gptoss_sizes
vthumbe1503:fix_cublas_grouped_gemm_gptoss_sizes
1m 15s
[WIP][Common, PyTorch] Improve mHC to match DeepSeek's implementation Documentation #16614: Pull request #2953 synchronize by pre-commit-ci Bot
Action required kainzhong:feat/mhc_enhancement
kainzhong:feat/mhc_enhancement
Action required
[WIP][Common, PyTorch] Improve mHC to match DeepSeek's implementation Documentation #16613: Pull request #2953 opened by kainzhong
1m 12s kainzhong:feat/mhc_enhancement
kainzhong:feat/mhc_enhancement
1m 12s
[JAX] Remove xla deterministic arg for MNIST test to not timeout L2_jax_unittest CI Documentation #16612: Pull request #2952 opened by tdophung
1m 23s tdophung:tdophung/remove_xla_deterministic
tdophung:tdophung/remove_xla_deterministic
1m 23s
[JAX] Calculate seqlens and offsets in O(T) space instead of O(T*T) space for THD sequences Documentation #16611: Pull request #2522 synchronize by KshitijLakhani
1m 31s KshitijLakhani:klakhani/mem-optimize-seqlens-offsets-thd
KshitijLakhani:klakhani/mem-optimize-seqlens-offsets-thd
1m 31s
Optimizations for MXFP8/NVFP4 dequantize kernels Documentation #16610: Pull request #2865 synchronize by YigongQin
Action required YigongQin:yigongq/bwd-dequantize-optim
YigongQin:yigongq/bwd-dequantize-optim
Action required
[Common] Comm+GEMM overlap API updated to support cuBlasMp backend (incl. framework API) Documentation #16609: Pull request #2443 synchronize by pre-commit-ci Bot
Action required denera:common/tp-overlap-cublasmp
denera:common/tp-overlap-cublasmp
Action required
[Common] Comm+GEMM overlap API updated to support cuBlasMp backend (incl. framework API) Documentation #16608: Pull request #2443 synchronize by denera
1m 30s denera:common/tp-overlap-cublasmp
denera:common/tp-overlap-cublasmp
1m 30s
Optimizations for MXFP8/NVFP4 dequantize kernels Documentation #16607: Pull request #2865 synchronize by YigongQin
Action required YigongQin:yigongq/bwd-dequantize-optim
YigongQin:yigongq/bwd-dequantize-optim
Action required
Optimizations for MXFP8/NVFP4 dequantize kernels Documentation #16606: Pull request #2865 synchronize by YigongQin
Action required YigongQin:yigongq/bwd-dequantize-optim
YigongQin:yigongq/bwd-dequantize-optim
Action required
Adds GEMM Profiling Guide to TE Documentation #16605: Pull request #2863 synchronize by jomitchellnv
1m 40s jomitchellnv:jm/gemm-blog
jomitchellnv:jm/gemm-blog
1m 40s
Adds GEMM Profiling Guide to TE Documentation #16604: Pull request #2863 synchronize by jomitchellnv
1m 2s jomitchellnv:jm/gemm-blog
jomitchellnv:jm/gemm-blog
1m 2s
Optimizations for MXFP8/NVFP4 dequantize kernels Documentation #16603: Pull request #2865 synchronize by YigongQin
Action required YigongQin:yigongq/bwd-dequantize-optim
YigongQin:yigongq/bwd-dequantize-optim
Action required
[PyTorch] Add pad_between_seqs support for non-CP and CP (A2A and P2P) with FA3 + THD (varlen) Documentation #16602: Pull request #2596 synchronize by sudhakarsingh27
1m 14s sudhakarsingh27:flash_attn_pad_bw_seqs
sudhakarsingh27:flash_attn_pad_bw_seqs
1m 14s
[Core] Remove unused NVFP4 quantize kernel Documentation #16601: Pull request #2946 synchronize by Oleg-Goncharov
1m 29s timmoon10:tmoon/remove-unused-nvfp4-quantize-kernel
timmoon10:tmoon/remove-unused-nvfp4-quantize-kernel
1m 29s
MXFP8 + FSDP2 checkpoint resume crashes in reset_sharded_param - add mxfp8 recpipe to fully shard Documentation #16600: Pull request #2951 synchronize by pre-commit-ci Bot
Action required savitha-eng:savitha/mxfp8-fsdp2-checkpoint-resume-bug
savitha-eng:savitha/mxfp8-fsdp2-checkpoint-resume-bug
Action required
MXFP8 + FSDP2 checkpoint resume crashes in reset_sharded_param - add mxfp8 recpipe to fully shard Documentation #16599: Pull request #2951 opened by savitha-eng
1m 15s savitha-eng:savitha/mxfp8-fsdp2-checkpoint-resume-bug
savitha-eng:savitha/mxfp8-fsdp2-checkpoint-resume-bug
1m 15s
[PyTorch] Add distributed Muon optimizer Documentation #16598: Pull request #2920 synchronize by pre-commit-ci Bot
Action required vcherepanov-nv:muon
vcherepanov-nv:muon
Action required
[PyTorch] Add distributed Muon optimizer Documentation #16597: Pull request #2920 synchronize by vcherepanov-nv
1m 8s vcherepanov-nv:muon
vcherepanov-nv:muon
1m 8s
[Common, PyTorch] Add Triton MLA attention kernels for SM80 Documentation #16596: Pull request #2950 synchronize by bzantium
Action required bzantium:mla-triton-sm80
bzantium:mla-triton-sm80
Action required
[JAX] Calculate seqlens and offsets in O(T) space instead of O(T*T) space for THD sequences Documentation #16594: Pull request #2522 synchronize by KshitijLakhani
1m 31s KshitijLakhani:klakhani/mem-optimize-seqlens-offsets-thd
KshitijLakhani:klakhani/mem-optimize-seqlens-offsets-thd
1m 31s
[PyTorch][CP] Add THD format support for AllGather-based Context Parallelism Documentation #16593: Pull request #2829 synchronize by sudhakarsingh27
1m 15s sudhakarsingh27:cp_thd_swa_with_ag
sudhakarsingh27:cp_thd_swa_with_ag
1m 15s
[Common, PyTorch] Add Triton MLA attention kernels for SM80 Documentation #16592: Pull request #2950 synchronize by pre-commit-ci Bot
Action required bzantium:mla-triton-sm80
bzantium:mla-triton-sm80
Action required