Skip to content

Actions: deepspeedai/DeepSpeed

Actions

cpu-torch-latest

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
cpu-torch-latest.yml
2,346 workflow runs
2,346 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
[AutoTP] Complete uneven sharding and universal checkpoint support cpu-torch-latest #8831: Pull request #8185 synchronize by jinyouzhi
Action required jinyouzhi:uneven
jinyouzhi:uneven
Action required
NCCL PG timeout 30min -> 10min cpu-torch-latest #8830: Pull request #8253 synchronize by stas00
34m 27s stas00-patch-1
stas00-patch-1
34m 27s
NCCL PG timeout 30min -> 10min cpu-torch-latest #8829: Pull request #8253 opened by stas00
3m 4s stas00-patch-1
stas00-patch-1
3m 4s
cpu-torch-latest cpu-torch-latest #8828: Scheduled
35m 3s master
master
35m 3s
Add native (DeepNVMe) host-memory pinning backend for accelerators cpu-torch-latest #8827: Pull request #8211 synchronize by sfc-gh-truwase
35m 35s tjruwase/native-pin-memory
tjruwase/native-pin-memory
35m 35s
Make pass contracts cover the passes DeepCompile actually schedules; Renaming files cpu-torch-latest #8826: Pull request #8251 opened by pengdurice
33m 53s pengdurice:peng-fix-conflict-contract-v1
pengdurice:peng-fix-conflict-contract-v1
33m 53s
cpu-torch-latest cpu-torch-latest #8825: Merge group checks requested
34m 13s
34m 13s
Fix repeated FlopsProfiler metric accumulation cpu-torch-latest #8824: Pull request #8246 synchronize by baremetaldevx86
Action required baremetaldevx86:fix/7413-flops-profiler-lifecycle
baremetaldevx86:fix/7413-flops-profiler-lifecycle
Action required
(1/2) Implementing Compiler Pass for AutoTP cpu-torch-latest #8823: Pull request #8204 synchronize by tohtana
36m 50s therealnaveenkamal:feature/autotp
therealnaveenkamal:feature/autotp
36m 50s
[AutoTP] Complete uneven sharding and universal checkpoint support cpu-torch-latest #8822: Pull request #8185 synchronize by jinyouzhi
Action required jinyouzhi:uneven
jinyouzhi:uneven
Action required
cpu-torch-latest cpu-torch-latest #8821: Merge group checks requested
33m 34s
33m 34s
Upgrade pip in Python install-smoke jobs cpu-torch-latest #8820: Pull request #8247 opened by tohtana
30m 33s tohtana:tohtana/ci-upgrade-python-matrix-pip
tohtana:tohtana/ci-upgrade-python-matrix-pip
30m 33s
cpu-torch-latest cpu-torch-latest #8819: Scheduled
33m 46s master
master
33m 46s
(1/2) Implementing Compiler Pass for AutoTP cpu-torch-latest #8818: Pull request #8204 synchronize by therealnaveenkamal
34m 59s therealnaveenkamal:feature/autotp
therealnaveenkamal:feature/autotp
34m 59s
Fix repeated FlopsProfiler metric accumulation cpu-torch-latest #8816: Pull request #8246 opened by baremetaldevx86
Action required baremetaldevx86:fix/7413-flops-profiler-lifecycle
baremetaldevx86:fix/7413-flops-profiler-lifecycle
Action required
Drop the documented grad_hooks ZeRO option, which does not exist cpu-torch-latest #8815: Pull request #8242 synchronize by vineethsaivs
Action required vineethsaivs:fix/zero-config-grad-hooks-doc
vineethsaivs:fix/zero-config-grad-hooks-doc
Action required
Add fused triton kernel for swiglu cpu-torch-latest #8814: Pull request #8244 synchronize by hwchen2017
28m 54s hongwei/swiglu
hongwei/swiglu
28m 54s
Fix shared loss gradient accumulation cpu-torch-latest #8813: Pull request #8245 synchronize by tohtana
46m 53s tohtana:tohtana/fix-zero2-bf16-shared-grad-8224
tohtana:tohtana/fix-zero2-bf16-shared-grad-8224
46m 53s
[AutoTP] Complete uneven sharding and universal checkpoint support cpu-torch-latest #8811: Pull request #8185 synchronize by jinyouzhi
Action required jinyouzhi:uneven
jinyouzhi:uneven
Action required
[AutoTP] Complete uneven sharding and universal checkpoint support cpu-torch-latest #8810: Pull request #8185 synchronize by delock
30m 52s jinyouzhi:uneven
jinyouzhi:uneven
30m 52s
Add fused triton kernel for swiglu cpu-torch-latest #8809: Pull request #8244 synchronize by hwchen2017
32m 18s hongwei/swiglu
hongwei/swiglu
32m 18s
cpu-torch-latest cpu-torch-latest #8808: Merge group checks requested
31m 26s
31m 26s
cpu-torch-latest cpu-torch-latest #8807: Merge group checks requested
34m 55s
34m 55s
cpu-torch-latest cpu-torch-latest #8806: Merge group checks requested
33m 49s
33m 49s
cpu-torch-latest cpu-torch-latest #8805: Scheduled
32m 33s master
master
32m 33s