Skip to content

Actions: microsoft/DeepSpeed

Actions

xpu-max1100

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
xpu-max1100.yml
893 workflow runs
893 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

xpu-max1100 xpu-max1100 #950: Scheduled
November 1, 2024 00:10 17m 54s master
master
November 1, 2024 00:10 17m 54s
Add no_sync context manager xpu-max1100 #949: Pull request #6675 synchronize by loadams
October 31, 2024 20:42 18m 8s olruwase/no_grad_sync_ctxt
olruwase/no_grad_sync_ctxt
October 31, 2024 20:42 18m 8s
Set shuffle=True by default in data_sampler xpu-max1100 #948: Pull request #6531 synchronize by loadams
October 31, 2024 18:52 18m 13s ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
October 31, 2024 18:52 18m 13s
Set shuffle=True by default in data_sampler xpu-max1100 #947: Pull request #6531 synchronize by loadams
October 31, 2024 17:56 17m 36s ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
October 31, 2024 17:56 17m 36s
Reduce the device bubble introduced by heavy loop synchronization in coalesced fetch/release(z3_leaf_module) xpu-max1100 #946: Pull request #6694 synchronize by tjruwase
October 31, 2024 15:16 18m 11s inkcherry:reduce_coalesced_fetch_bubble
inkcherry:reduce_coalesced_fetch_bubble
October 31, 2024 15:16 18m 11s
Set shuffle=True by default in data_sampler xpu-max1100 #945: Pull request #6531 synchronize by loadams
October 31, 2024 14:54 21m 20s ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
October 31, 2024 14:54 21m 20s
add zero3 module_granularity_threshold to zero optimization. xpu-max1100 #944: Pull request #6649 synchronize by tjruwase
October 31, 2024 14:41 17m 34s inkcherry:z3_coalesced_fetch
inkcherry:z3_coalesced_fetch
October 31, 2024 14:41 17m 34s
Use one param coordinator for both train/inference scenarios xpu-max1100 #943: Pull request #6662 synchronize by tjruwase
October 31, 2024 14:03 17m 57s tohtana/simplify_param_coordinator
tohtana/simplify_param_coordinator
October 31, 2024 14:03 17m 57s
Reduce the device bubble introduced by heavy loop synchronization in coalesced fetch/release(z3_leaf_module) xpu-max1100 #942: Pull request #6694 opened by inkcherry
October 31, 2024 10:01 Action required inkcherry:reduce_coalesced_fetch_bubble
inkcherry:reduce_coalesced_fetch_bubble
October 31, 2024 10:01 Action required
fix memcpy issue on backward for zero-infinity xpu-max1100 #941: Pull request #6670 synchronize by xylian86
October 31, 2024 03:08 35m 48s xylian86:fix_bwd_memcpy_4_infinity
xylian86:fix_bwd_memcpy_4_infinity
October 31, 2024 03:08 35m 48s
xpu-max1100 xpu-max1100 #940: Scheduled
October 31, 2024 00:09 17m 38s master
master
October 31, 2024 00:09 17m 38s
fix memcpy issue on backward for zero-infinity xpu-max1100 #939: Pull request #6670 synchronize by tjruwase
October 30, 2024 19:46 29m 56s xylian86:fix_bwd_memcpy_4_infinity
xylian86:fix_bwd_memcpy_4_infinity
October 30, 2024 19:46 29m 56s
Set shuffle=True by default in data_sampler xpu-max1100 #938: Pull request #6531 synchronize by loadams
October 30, 2024 19:45 34m 40s ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
October 30, 2024 19:45 34m 40s
xpu-max1100 xpu-max1100 #937: Scheduled
October 30, 2024 00:09 17m 47s master
master
October 30, 2024 00:09 17m 47s
Set shuffle=True by default in data_sampler xpu-max1100 #936: Pull request #6531 synchronize by loadams
October 29, 2024 18:26 17m 45s ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
October 29, 2024 18:26 17m 45s
fix memcpy issue on backward for zero-infinity xpu-max1100 #935: Pull request #6670 synchronize by loadams
October 29, 2024 17:41 9m 33s xylian86:fix_bwd_memcpy_4_infinity
xylian86:fix_bwd_memcpy_4_infinity
October 29, 2024 17:41 9m 33s
fix memcpy issue on backward for zero-infinity xpu-max1100 #934: Pull request #6670 synchronize by loadams
October 29, 2024 16:00 12m 44s xylian86:fix_bwd_memcpy_4_infinity
xylian86:fix_bwd_memcpy_4_infinity
October 29, 2024 16:00 12m 44s
Set shuffle=True by default in data_sampler xpu-max1100 #932: Pull request #6531 synchronize by loadams
October 29, 2024 04:13 18m 2s ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
October 29, 2024 04:13 18m 2s
xpu-max1100 xpu-max1100 #931: Scheduled
October 29, 2024 00:09 43m 16s master
master
October 29, 2024 00:09 43m 16s
Set shuffle=True by default in data_sampler xpu-max1100 #930: Pull request #6531 synchronize by loadams
October 29, 2024 00:05 19m 21s ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
October 29, 2024 00:05 19m 21s
Add no_sync context manager xpu-max1100 #929: Pull request #6675 synchronize by tjruwase
October 28, 2024 22:19 17m 52s olruwase/no_grad_sync_ctxt
olruwase/no_grad_sync_ctxt
October 28, 2024 22:19 17m 52s
Add no_sync context manager xpu-max1100 #928: Pull request #6675 synchronize by tjruwase
October 28, 2024 21:06 18m 20s olruwase/no_grad_sync_ctxt
olruwase/no_grad_sync_ctxt
October 28, 2024 21:06 18m 20s
Fix training of pipeline based peft's lora model xpu-max1100 #927: Pull request #5477 synchronize by loadams
October 28, 2024 20:08 36m 31s xuanhua:axu/fix-pipeline-with-lora
xuanhua:axu/fix-pipeline-with-lora
October 28, 2024 20:08 36m 31s
Set shuffle=True by default in data_sampler xpu-max1100 #926: Pull request #6531 synchronize by loadams
October 28, 2024 19:58 17m 53s ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
ranzhejiang:zhejiang/fix_runtime_dataloader_shuffle
October 28, 2024 19:58 17m 53s
Add no_sync context manager xpu-max1100 #925: Pull request #6675 synchronize by tjruwase
October 28, 2024 19:38 9m 41s olruwase/no_grad_sync_ctxt
olruwase/no_grad_sync_ctxt
October 28, 2024 19:38 9m 41s