Skip to content

Actions: deepspeedai/DeepSpeed

Actions

cpu-torch-latest

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
cpu-torch-latest.yml
2,142 workflow runs
2,142 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
ZeRO-3: don't partition frozen params during an activation-checkpoint recompute cpu-torch-latest #8321: Pull request #8130 synchronize by qgallouedec
Action required qgallouedec:fix/zero3-frozen-param-checkpoint-recompute
qgallouedec:fix/zero3-frozen-param-checkpoint-recompute
Action required
fix(autotp): when using autotp, ignore the consistency of certain data within the tp_group. cpu-torch-latest #8320: Pull request #8125 synchronize by 1787648106
31m 46s 1787648106:ignore_check_cache_while_autotp
1787648106:ignore_check_cache_while_autotp
31m 46s
Add governance process page to deepspeed.ai cpu-torch-latest #8319: Pull request #8138 opened by PKUWZP
26m 38s zhipwang_governing_doc
zhipwang_governing_doc
26m 38s
ZeRO-3: stream partitioning of oversized parameters in zero.Init cpu-torch-latest #8318: Pull request #8103 synchronize by Achyuthan-S
Action required Achyuthan-S:fix/zero-init-stream-large-params
Achyuthan-S:fix/zero-init-stream-large-params
Action required
cpu-torch-latest cpu-torch-latest #8317: Scheduled
32m 5s master
master
32m 5s
Fix CommsLogger.stop_profiling_comms not disabling global profiling cpu-torch-latest #8316: Pull request #8137 opened by vineethsaivs
Action required vineethsaivs:fix/stop-profiling-comms-flag
vineethsaivs:fix/stop-profiling-comms-flag
Action required
cpu-torch-latest cpu-torch-latest #8315: Merge group checks requested
27m 25s
27m 25s
Make DCO workflow Probot compatible cpu-torch-latest #8314: Pull request #8110 synchronize by tohtana
27m 52s tohtana:tohtana/fix/dco-required-probot-compat
tohtana:tohtana/fix/dco-required-probot-compat
27m 52s
cpu-torch-latest cpu-torch-latest #8313: Merge group checks requested
30m 33s
30m 33s
Add EXAONE 4.5 model support for Inference V2 cpu-torch-latest #8312: Pull request #8121 synchronize by Bias92
Action required Bias92:add-exaone4_5-inference-v2
Bias92:add-exaone4_5-inference-v2
Action required
fix(exceptions): narrow broad Exception clauses to specific types cpu-torch-latest #8311: Pull request #8136 opened by sowndappan5
Action required sowndappan5:fix/narrow-exceptions
sowndappan5:fix/narrow-exceptions
Action required
Feature/zero multi loss separate backward cpu-torch-latest #8310: Pull request #8135 synchronize by nathon-lee
-1s nathon-lee:feature/zero-multi-loss-separate-backward
nathon-lee:feature/zero-multi-loss-separate-backward
-1s
Feature/zero multi loss separate backward cpu-torch-latest #8309: Pull request #8135 opened by nathon-lee
Action required nathon-lee:feature/zero-multi-loss-separate-backward
nathon-lee:feature/zero-multi-loss-separate-backward
Action required
fix: add zero-guards for divisor/scale edge cases (#7838) cpu-torch-latest #8306: Pull request #8133 opened by liuyun7345
Action required liuyun7345:fix/zero-guards-7838
liuyun7345:fix/zero-guards-7838
Action required
ZeRO-3: don't partition frozen params during an activation-checkpoint recompute cpu-torch-latest #8305: Pull request #8130 synchronize by qgallouedec
Action required qgallouedec:fix/zero3-frozen-param-checkpoint-recompute
qgallouedec:fix/zero3-frozen-param-checkpoint-recompute
Action required
cpu-torch-latest cpu-torch-latest #8304: Merge group checks requested
30m 47s
30m 47s
cpu-torch-latest cpu-torch-latest #8303: Merge group checks requested
26m 56s
26m 56s
cpu-torch-latest cpu-torch-latest #8302: Scheduled
29m 16s master
master
29m 16s
ZeRO-3: don't partition frozen params during an activation-checkpoint recompute cpu-torch-latest #8301: Pull request #8130 synchronize by qgallouedec
29m 46s qgallouedec:fix/zero3-frozen-param-checkpoint-recompute
qgallouedec:fix/zero3-frozen-param-checkpoint-recompute
29m 46s
ZeRO-3: don't partition frozen params during an activation-checkpoint recompute cpu-torch-latest #8300: Pull request #8130 synchronize by qgallouedec
Action required qgallouedec:fix/zero3-frozen-param-checkpoint-recompute
qgallouedec:fix/zero3-frozen-param-checkpoint-recompute
Action required
ZeRO-3: stream partitioning of oversized parameters in zero.Init cpu-torch-latest #8299: Pull request #8103 synchronize by Achyuthan-S
Action required Achyuthan-S:fix/zero-init-stream-large-params
Achyuthan-S:fix/zero-init-stream-large-params
Action required
cpu-torch-latest cpu-torch-latest #8298: Merge group checks requested
29m 26s
29m 26s
fix(zero-offload): mitigate CPU RSS waste on ZeRO-2 optimizer offload (#7693) cpu-torch-latest #8297: Pull request #8132 synchronize by liuyun7345
Action required liuyun7345:fix/cpu-offload-memory-leak-7693
liuyun7345:fix/cpu-offload-memory-leak-7693
Action required
Make DCO workflow Probot compatible cpu-torch-latest #8296: Pull request #8110 synchronize by tohtana
27m 28s tohtana:tohtana/fix/dco-required-probot-compat
tohtana:tohtana/fix/dco-required-probot-compat
27m 28s
cpu-torch-latest cpu-torch-latest #8295: Merge group checks requested
28m 22s
28m 22s