Skip to content

Actions: deepspeedai/DeepSpeed

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
2,500+ workflow runs
2,500+ workflow runs

Filter by Workflow

⌥ + click/return to exclude

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
Fill curriculum batches across repeated cluster shuffles cpu-torch-latest #10041: Pull request #8520 synchronize by vineethsaivs
Action required vineethsaivs:fix-small-curriculum-clusters-20260914
vineethsaivs:fix-small-curriculum-clusters-20260914
Action required
Fill curriculum batches across repeated cluster shuffles modal-torch-latest #5278: Pull request #8520 synchronize by vineethsaivs
25s
25s
Keep Random-LTD sequence lengths above the configured minimum python #17498: Pull request #8529 synchronize by vineethsaivs
Action required vineethsaivs:fix/random-ltd-minimum-20260915
vineethsaivs:fix/random-ltd-minimum-20260915
Action required
Keep Random-LTD sequence lengths above the configured minimum Formatting #22237: Pull request #8529 synchronize by vineethsaivs
Action required vineethsaivs:fix/random-ltd-minimum-20260915
vineethsaivs:fix/random-ltd-minimum-20260915
Action required
Keep Random-LTD sequence lengths above the configured minimum nv-pre-compile-ops #15269: Pull request #8529 synchronize by vineethsaivs
Action required vineethsaivs:fix/random-ltd-minimum-20260915
vineethsaivs:fix/random-ltd-minimum-20260915
Action required
Keep Random-LTD sequence lengths above the configured minimum DCO / required #1764: Pull request #8529 synchronize by vineethsaivs
Action required vineethsaivs:fix/random-ltd-minimum-20260915
vineethsaivs:fix/random-ltd-minimum-20260915
Action required
Keep Random-LTD sequence lengths above the configured minimum cpu-torch-latest #10040: Pull request #8529 synchronize by vineethsaivs
Action required vineethsaivs:fix/random-ltd-minimum-20260915
vineethsaivs:fix/random-ltd-minimum-20260915
Action required
Keep Random-LTD sequence lengths above the configured minimum modal-torch-latest #5277: Pull request #8529 synchronize by vineethsaivs
21s
21s
Normalize Gram Newton-Schulz inputs before casting to fp16 nv-pre-compile-ops #15268: Pull request #8533 synchronize by vineethsaivs
Action required vineethsaivs:fix-muon-fp16-normalization-20260915
vineethsaivs:fix-muon-fp16-normalization-20260915
Action required
Normalize Gram Newton-Schulz inputs before casting to fp16 python #17497: Pull request #8533 synchronize by vineethsaivs
Action required vineethsaivs:fix-muon-fp16-normalization-20260915
vineethsaivs:fix-muon-fp16-normalization-20260915
Action required
Normalize Gram Newton-Schulz inputs before casting to fp16 DCO / required #1763: Pull request #8533 synchronize by vineethsaivs
Action required vineethsaivs:fix-muon-fp16-normalization-20260915
vineethsaivs:fix-muon-fp16-normalization-20260915
Action required
Normalize Gram Newton-Schulz inputs before casting to fp16 cpu-torch-latest #10039: Pull request #8533 synchronize by vineethsaivs
Action required vineethsaivs:fix-muon-fp16-normalization-20260915
vineethsaivs:fix-muon-fp16-normalization-20260915
Action required
Normalize Gram Newton-Schulz inputs before casting to fp16 Formatting #22236: Pull request #8533 synchronize by vineethsaivs
Action required vineethsaivs:fix-muon-fp16-normalization-20260915
vineethsaivs:fix-muon-fp16-normalization-20260915
Action required
Normalize Gram Newton-Schulz inputs before casting to fp16 modal-torch-latest #5276: Pull request #8533 synchronize by vineethsaivs
27s
27s
Weight the TiledLoss backward pass the way the forward does modal-torch-latest #5275: Pull request #8587 review_requested by sfc-gh-truwase
38s
38s
Weight the TiledLoss backward pass the way the forward does cpu-torch-latest #10038: Pull request #8587 synchronize by sfc-gh-truwase
33m 14s vineethsaivs:fix-tiled-loss-gradient-20260918
vineethsaivs:fix-tiled-loss-gradient-20260918
33m 14s
Weight the TiledLoss backward pass the way the forward does python #17496: Pull request #8587 synchronize by sfc-gh-truwase
1m 41s vineethsaivs:fix-tiled-loss-gradient-20260918
vineethsaivs:fix-tiled-loss-gradient-20260918
1m 41s
Weight the TiledLoss backward pass the way the forward does Formatting #22235: Pull request #8587 synchronize by sfc-gh-truwase
1m 20s vineethsaivs:fix-tiled-loss-gradient-20260918
vineethsaivs:fix-tiled-loss-gradient-20260918
1m 20s
Weight the TiledLoss backward pass the way the forward does nv-pre-compile-ops #15267: Pull request #8587 synchronize by sfc-gh-truwase
21m 48s vineethsaivs:fix-tiled-loss-gradient-20260918
vineethsaivs:fix-tiled-loss-gradient-20260918
21m 48s
Weight the TiledLoss backward pass the way the forward does DCO / required #1762: Pull request #8587 synchronize by sfc-gh-truwase
11s vineethsaivs:fix-tiled-loss-gradient-20260918
vineethsaivs:fix-tiled-loss-gradient-20260918
11s
Weight the TiledLoss backward pass the way the forward does modal-torch-latest #5274: Pull request #8587 synchronize by sfc-gh-truwase
26s
26s
Use int64 program ids in the SwiGLU and AutoEP fused restore Triton kernels DCO / required #1761: Pull request #8591 opened by pengdurice
8s pengdurice:peng-fix-triton-int32-offset-overflow-v1
pengdurice:peng-fix-triton-int32-offset-overflow-v1
8s
Use int64 program ids in the SwiGLU and AutoEP fused restore Triton kernels nv-pre-compile-ops #15266: Pull request #8591 opened by pengdurice
21m 33s pengdurice:peng-fix-triton-int32-offset-overflow-v1
pengdurice:peng-fix-triton-int32-offset-overflow-v1
21m 33s
Use int64 program ids in the SwiGLU and AutoEP fused restore Triton kernels Formatting #22234: Pull request #8591 opened by pengdurice
1m 51s pengdurice:peng-fix-triton-int32-offset-overflow-v1
pengdurice:peng-fix-triton-int32-offset-overflow-v1
1m 51s
Use int64 program ids in the SwiGLU and AutoEP fused restore Triton kernels cpu-torch-latest #10037: Pull request #8591 opened by pengdurice
39m 44s pengdurice:peng-fix-triton-int32-offset-overflow-v1
pengdurice:peng-fix-triton-int32-offset-overflow-v1
39m 44s