Skip to content

Actions: NVIDIA/TransformerEngine

Actions

Blossom-CI

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
blossom-ci.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
MXFP8 + FSDP2 checkpoint resume crashes in reset_sharded_param - add mxfp8 recpipe to fully shard Blossom-CI #9642: Issue comment #2951 (comment) created by greptile-apps Bot
1s
1s
[Common] Improved fused MoE aux loss kernel for large # of experts Blossom-CI #9641: Issue comment #2758 (comment) created by denera
1s
1s
[PyTorch] Guard makeTransformerEngineTensor against tensors without storage Blossom-CI #9640: Issue comment #2944 (comment) created by acmore
2s
2s
[PyTorch] Guard makeTransformerEngineTensor against tensors without storage Blossom-CI #9639: Issue comment #2944 (comment) created by acmore
1s
1s
[JAX] Calculate seqlens and offsets in O(T) space instead of O(T*T) space for THD sequences Blossom-CI #9638: Issue comment #2522 (comment) created by KshitijLakhani
1s
1s
GroupedLinear performance on Mixtral-8x7B at EP=2 Blossom-CI #9637: Issue comment #2939 (comment) created by faradawn
2s
2s
Blossom-CI Blossom-CI #9636: created by faradawn
1s
1s
[Common, PyTorch] Add Triton MLA attention kernels for SM80 Blossom-CI #9635: Issue comment #2950 (comment) created by greptile-apps Bot
2s
2s
[PyTorch] Add workaround for cuteDSL stride requirement for zero-token expert Blossom-CI #9634: Issue comment #2947 (comment) created by ksivaman
Skipped
Skipped
[All] Remove max512 backend Blossom-CI #9633: Issue comment #2949 (comment) created by greptile-apps Bot
1s
1s
[PyTorch] Cleanup cudnn-frontend requirements for fused grouped MLP Blossom-CI #9632: Issue comment #2948 (comment) created by greptile-apps Bot
1s
1s
[All] Remove max512 backend Blossom-CI #9631: Issue comment #2949 (comment) created by cyanguwa
1s
1s
[PyTorch] Add workaround for cuteDSL stride requirement for zero-token expert Blossom-CI #9630: Issue comment #2947 (comment) created by greptile-apps Bot
3s
3s
Add examples for MoE models - Mixtral in TE Blossom-CI #9629: Issue comment #2642 (comment) created by faradawn
2s
2s
GroupedGemm: NVFP4 via cuBLAS Blossom-CI #9628: Issue comment #2455 (comment) created by nvMelissa
1s
1s
[PyTorch] Fusible ops preserve usages in quantized weight tensors Blossom-CI #9627: Issue comment #2929 (comment) created by timmoon10
1s
1s
[PyTorch] Add pad_between_seqs support for non-CP and CP (A2A and P2P) with FA3 + THD (varlen) Blossom-CI #9626: Issue comment #2596 (comment) created by sudhakarsingh27
2s
2s
[PyTorch] Guard makeTransformerEngineTensor against tensors without storage Blossom-CI #9625: Issue comment #2944 (comment) created by timmoon10
1s
1s
[PyTorch] Introduce quantizer roles Blossom-CI #9624: Issue comment #2620 (comment) created by negvet
1s
1s
[Core] Remove unused NVFP4 quantize kernel Blossom-CI #9623: Issue comment #2946 (comment) created by greptile-apps Bot
2s
2s
[Core] Remove unused NVFP4 quantize kernel Blossom-CI #9622: Issue comment #2946 (comment) created by timmoon10
2s
2s
[Common] Fix incorrect amax initialization in non-RHT NVFP4 C++ tests Blossom-CI #9621: Issue comment #2943 (comment) created by Oleg-Goncharov
2s
2s
Blossom-CI Blossom-CI #9620: created by Oleg-Goncharov
1s
1s
Variable Grouped Swizzle Blossom-CI #9619: Issue comment #2914 (comment) created by vthumbe1503
1s
1s
Blossom-CI Blossom-CI #9618: created by Oleg-Goncharov
1s
1s