Skip to content

Actions: deepspeedai/DeepSpeed

Actions

DCO / required

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
dco.yml
2,181 workflow runs
2,181 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
Fix ZenFlow CPU affinity on Slurm masks and multi-node runs DCO / required #2375: Pull request #8774 opened by Aravind-11
Action required Aravind-11:fix/zenflow-slurm-affinity
Aravind-11:fix/zenflow-slurm-affinity
Action required
Average sparse gradients by world size regardless of gradient_predivide_factor DCO / required #2374: Pull request #8772 opened by vineethsaivs
Action required vineethsaivs:fix/sparse-allreduce-predivide-factor
vineethsaivs:fix/sparse-allreduce-predivide-factor
Action required
Skip SuperOffload's eager CPU step on non-finite gradients DCO / required #2373: Pull request #8771 opened by vineethsaivs
Action required vineethsaivs:fix/superoffload-skip-nonfinite-async-step
vineethsaivs:fix/superoffload-skip-nonfinite-async-step
Action required
DCO / required DCO / required #2372: Merge group checks requested
7s
7s
Pin Modal torch-latest to OCI and cover PR open events DCO / required #2371: Pull request #8770 synchronize by tohtana
8s tohtana:tohtana/fix/modal-torch-latest-oci-placement
tohtana:tohtana/fix/modal-torch-latest-oci-placement
8s
DCO / required DCO / required #2370: Merge group checks requested
7s
7s
Pin Modal torch-latest to OCI and cover PR open events DCO / required #2369: Pull request #8770 opened by tohtana
8s tohtana:tohtana/fix/modal-torch-latest-oci-placement
tohtana:tohtana/fix/modal-torch-latest-oci-placement
8s
DCO / required DCO / required #2368: Merge group checks requested
12s
12s
feat(rollout): add OPSD response token loss primitives DCO / required #2367: Pull request #8677 synchronize by nathon-lee
Action required nathon-lee:feat/opsd-response-token-loss
nathon-lee:feat/opsd-response-token-loss
Action required
Honor seq_dim when AutoSP shards inputs DCO / required #2366: Pull request #8720 synchronize by jinyouzhi
Action required jinyouzhi:fix-autosp-seq-dim
jinyouzhi:fix-autosp-seq-dim
Action required
DCO / required DCO / required #2365: Merge group checks requested
7s
7s
Add an opt-in route_norm flag for the legacy MoE gate DCO / required #2364: Pull request #8769 opened by Aravind-11
Action required Aravind-11:moe-optin-route-norm
Aravind-11:moe-optin-route-norm
Action required
Fix autotuner latency ranking and the FLOPS metric name DCO / required #2363: Pull request #8767 opened by Aravind-11
Action required Aravind-11:fix-autotune-metric-direction
Aravind-11:fix-autotune-metric-direction
Action required
Detect NEON when py-cpuinfo reports no CPU flags DCO / required #2362: Pull request #8765 opened by vineethsaivs
Action required vineethsaivs:fix/simd-width-missing-cpuinfo-flags
vineethsaivs:fix/simd-width-missing-cpuinfo-flags
Action required
[AutoSP] Match the full-sequence causal LM loss and gradients DCO / required #2361: Pull request #8457 synchronize by jinyouzhi
Action required jinyouzhi:autosp-correctness
jinyouzhi:autosp-correctness
Action required
Round to nearest in NPUQuantizer, matching the CUDA quantizer DCO / required #2360: Pull request #8764 opened by vineethsaivs
Action required vineethsaivs:fix/npu-quantizer-round
vineethsaivs:fix/npu-quantizer-round
Action required
Fix flops profiler FLOPS and samples/s under gradient accumulation DCO / required #2359: Pull request #8763 opened by vineethsaivs
Action required vineethsaivs:fix/flops-profiler-grad-accum
vineethsaivs:fix/flops-profiler-grad-accum
Action required
[AutoSP] Rebuild query-sharded attention masks before SDPA DCO / required #2358: Pull request #8761 opened by jinyouzhi
Action required jinyouzhi:fix-autosp-attn-mask
jinyouzhi:fix-autosp-attn-mask
Action required
[AutoSP] Fix tensor truthiness in get_node_shape_meta DCO / required #2357: Pull request #8760 opened by jinyouzhi
Action required jinyouzhi:fix-autosp-shape-meta
jinyouzhi:fix-autosp-shape-meta
Action required
Add an opt-in attention-output stash for Hugging Face reentrant checkpointing DCO / required #2356: Pull request #8759 opened by yh0903
Action required yh0903:yh0903/autoep-attention-output-stash
yh0903:yh0903/autoep-attention-output-stash
Action required
DCO / required DCO / required #2355: Merge group checks requested
7s
7s
Fix DataAnalyzer.run_map crashing when specific_threads is set DCO / required #2354: Pull request #8758 opened by Rakshit-gen
Action required Rakshit-gen:fix/data-analyzer-specific-threads
Rakshit-gen:fix/data-analyzer-specific-threads
Action required
DCO / required DCO / required #2353: Merge group checks requested
1m 40s
1m 40s
DCO / required DCO / required #2352: Merge group checks requested
8s
8s
feat(rollout): add OPSD response token loss primitives DCO / required #2351: Pull request #8677 synchronize by nathon-lee
Action required nathon-lee:feat/opsd-response-token-loss
nathon-lee:feat/opsd-response-token-loss
Action required