⚡ Sangsakala Proxy 🏠 Home
Skip to content

Actions: deepspeedai/DeepSpeed

Actions

python

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
python.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
feat(comm): one-shot P2P custom all-reduce for TP=2 inference python #18950: Pull request #8997 synchronize by delock
1m 31s gma/custom_ar_standalone
gma/custom_ar_standalone
1m 31s
python python #18949: Merge group checks requested
1m 29s
1m 29s
python python #18948: Merge group checks requested
1m 27s
1m 27s
feat(rollout): segment-based kernel injection for hybrid-engine greedy decode python #18947: Pull request #8602 synchronize by delock
1m 24s gma/qwen_ki_expr
gma/qwen_ki_expr
1m 24s
python python #18946: Scheduled
1m 16s master
master
1m 16s
python python #18945: Merge group checks requested
3m 24s
3m 24s
python python #18944: Merge group checks requested
2m 20s
2m 20s
Update version.txt post-v0.19.9 release python #18943: Pull request #8999 opened by loadams
1m 5s loadams/update-version-post-release
loadams/update-version-post-release
1m 5s
Fix torch-latest workflow unit failures python #18942: Pull request #8950 synchronize by sfc-gh-truwase
2m 14s ci/fix-workflow-failures
ci/fix-workflow-failures
2m 14s
Persist ZeRO-3 HybridMegatron QKV layout changes python #18941: Pull request #8392 synchronize by LiRunGuo
Action required LiRunGuo:fix/zero3-hybrid-megatron-qkv-layout
LiRunGuo:fix/zero3-hybrid-megatron-qkv-layout
Action required
Synchronize ZeRO-3 rollout early stopping python #18940: Pull request #8390 synchronize by LiRunGuo
Action required LiRunGuo:fix/hybrid-rollout-global-early-stop
LiRunGuo:fix/hybrid-rollout-global-early-stop
Action required
Add an opt-in deterministic execution policy python #18939: Pull request #8917 synchronize by ziming-zh
3m 7s ziming-zh:deterministic-mode
ziming-zh:deterministic-mode
3m 7s
feat(rollout): segment-based kernel injection for hybrid-engine greedy decode python #18938: Pull request #8602 synchronize by delock
1m 33s gma/qwen_ki_expr
gma/qwen_ki_expr
1m 33s
Add an opt-in deterministic execution policy python #18937: Pull request #8917 synchronize by ziming-zh
7h 41m 24s ziming-zh:deterministic-mode
ziming-zh:deterministic-mode
7h 41m 24s
Add an opt-in deterministic execution policy python #18936: Pull request #8917 synchronize by ziming-zh
7h 56m 38s ziming-zh:deterministic-mode
ziming-zh:deterministic-mode
7h 56m 38s
Add an opt-in deterministic execution policy python #18935: Pull request #8917 synchronize by ziming-zh
7h 57m 20s ziming-zh:deterministic-mode
ziming-zh:deterministic-mode
7h 57m 20s
Add an opt-in deterministic execution policy python #18934: Pull request #8917 synchronize by ziming-zh
Action required ziming-zh:deterministic-mode
ziming-zh:deterministic-mode
Action required
[AffineIR] Calibrate affine transfer links per path (Plan analysis & calibration Part-2) python #18933: Pull request #8966 synchronize by jinyouzhi
Action required jinyouzhi:affine-link-calibration
jinyouzhi:affine-link-calibration
Action required
[AffineIR] Calibrate affine transfer links per path (Plan analysis & calibration Part-2) python #18932: Pull request #8966 synchronize by jinyouzhi
Action required jinyouzhi:affine-link-calibration
jinyouzhi:affine-link-calibration
Action required
feat(rollout): segment-based kernel injection for hybrid-engine greedy decode python #18931: Pull request #8602 synchronize by delock
1m 27s gma/qwen_ki_expr
gma/qwen_ki_expr
1m 27s
feat(comm): one-shot P2P custom all-reduce for TP=2 inference python #18930: Pull request #8997 synchronize by delock
1m 27s gma/custom_ar_standalone
gma/custom_ar_standalone
1m 27s
Clear stale offload scratch before restoring low-precision gradients python #18929: Pull request #8728 synchronize by adenzhou1350
Action required adenzhou1350:codex/clear-low-precision-offload-scratch
adenzhou1350:codex/clear-low-precision-offload-scratch
Action required
[Hybrid Engine] Require an explicit accelerator opt-in for graph capture python #18928: Pull request #8980 synchronize by VenusTZZ
Action required VenusTZZ:fix-hybrid-engine-graph-capture-opt-in
VenusTZZ:fix-hybrid-engine-graph-capture-opt-in
Action required
Fix AutoTP multi-query GPTBigCode training at TP > 1 python #18927: Pull request #8962 synchronize by VenusTZZ
Action required VenusTZZ:fix-gptbigcode-autotp-npu
VenusTZZ:fix-gptbigcode-autotp-npu
Action required
feat(comm): one-shot P2P custom all-reduce for TP=2 inference python #18926: Pull request #8997 synchronize by delock
1m 31s gma/custom_ar_standalone
gma/custom_ar_standalone
1m 31s