Skip to content

Actions: microsoft/DeepSpeed

Actions

hpu-gaudi2

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
hpu-gaudi2.yml
1,202 workflow runs
1,202 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

hpu-gaudi2 hpu-gaudi2 #1432: Scheduled
December 28, 2024 00:11 2h 1m 49s master
master
December 28, 2024 00:11 2h 1m 49s
Reduce the device bubble introduced by heavy loop synchronization in coalesced fetch/release(z3_leaf_module) hpu-gaudi2 #1431: Pull request #6694 synchronize by inkcherry
December 27, 2024 00:44 56m 25s inkcherry:reduce_coalesced_fetch_bubble
inkcherry:reduce_coalesced_fetch_bubble
December 27, 2024 00:44 56m 25s
Reduce the device bubble introduced by heavy loop synchronization in coalesced fetch/release(z3_leaf_module) hpu-gaudi2 #1430: Pull request #6694 synchronize by inkcherry
December 27, 2024 00:44 Action required inkcherry:reduce_coalesced_fetch_bubble
inkcherry:reduce_coalesced_fetch_bubble
December 27, 2024 00:44 Action required
hpu-gaudi2 hpu-gaudi2 #1429: Scheduled
December 27, 2024 00:11 2h 3m 39s master
master
December 27, 2024 00:11 2h 3m 39s
Stage3: Use new torch grad accumulation hooks API hpu-gaudi2 #1428: Pull request #6773 synchronize by loadams
December 26, 2024 20:09 1h 2m 2s deepcharm:stage3-use-new-grad-acc-api
deepcharm:stage3-use-new-grad-acc-api
December 26, 2024 20:09 1h 2m 2s
Stage3: Use new torch grad accumulation hooks API hpu-gaudi2 #1427: Pull request #6773 synchronize by loadams
December 26, 2024 17:40 2h 25m 40s deepcharm:stage3-use-new-grad-acc-api
deepcharm:stage3-use-new-grad-acc-api
December 26, 2024 17:40 2h 25m 40s
Use ds-specific module id to avoid conflicts hpu-gaudi2 #1426: Pull request #6847 synchronize by loadams
December 26, 2024 17:13 2h 59m 33s olruwase/pr_6772
olruwase/pr_6772
December 26, 2024 17:13 2h 59m 33s
Update Gaudi2 jobs to latest 1.19 build hpu-gaudi2 #1425: Pull request #6905 synchronize by loadams
December 26, 2024 17:12 56m 27s raza-sikander:master
raza-sikander:master
December 26, 2024 17:12 56m 27s
Reduce the device bubble introduced by heavy loop synchronization in coalesced fetch/release(z3_leaf_module) hpu-gaudi2 #1424: Pull request #6694 synchronize by inkcherry
December 26, 2024 12:41 52m 52s inkcherry:reduce_coalesced_fetch_bubble
inkcherry:reduce_coalesced_fetch_bubble
December 26, 2024 12:41 52m 52s
hpu-gaudi2 hpu-gaudi2 #1423: Scheduled
December 26, 2024 00:11 1h 57m 17s master
master
December 26, 2024 00:11 1h 57m 17s
Reduce the device bubble introduced by heavy loop synchronization in coalesced fetch/release(z3_leaf_module) hpu-gaudi2 #1422: Pull request #6694 synchronize by inkcherry
December 25, 2024 09:42 54m 56s inkcherry:reduce_coalesced_fetch_bubble
inkcherry:reduce_coalesced_fetch_bubble
December 25, 2024 09:42 54m 56s
Reduce the device bubble introduced by heavy loop synchronization in coalesced fetch/release(z3_leaf_module) hpu-gaudi2 #1421: Pull request #6694 synchronize by inkcherry
December 25, 2024 09:37 Action required inkcherry:reduce_coalesced_fetch_bubble
inkcherry:reduce_coalesced_fetch_bubble
December 25, 2024 09:37 Action required
Reduce the device bubble introduced by heavy loop synchronization in coalesced fetch/release(z3_leaf_module) hpu-gaudi2 #1420: Pull request #6694 synchronize by inkcherry
December 25, 2024 09:27 Action required inkcherry:reduce_coalesced_fetch_bubble
inkcherry:reduce_coalesced_fetch_bubble
December 25, 2024 09:27 Action required
hpu-gaudi2 hpu-gaudi2 #1419: Scheduled
December 25, 2024 00:11 1h 56m 54s master
master
December 25, 2024 00:11 1h 56m 54s
[inf] Add config var to enable keeping module on host hpu-gaudi2 #1418: Pull request #6846 synchronize by oelayan7
December 24, 2024 06:49 55m 59s oelayan7:keep_module_on_host
oelayan7:keep_module_on_host
December 24, 2024 06:49 55m 59s
hpu-gaudi2 hpu-gaudi2 #1417: Scheduled
December 24, 2024 00:11 2h 3m 16s master
master
December 24, 2024 00:11 2h 3m 16s
Reduce the device bubble introduced by heavy loop synchronization in coalesced fetch/release(z3_leaf_module) hpu-gaudi2 #1415: Pull request #6694 synchronize by inkcherry
December 23, 2024 02:54 Action required inkcherry:reduce_coalesced_fetch_bubble
inkcherry:reduce_coalesced_fetch_bubble
December 23, 2024 02:54 Action required
hpu-gaudi2 hpu-gaudi2 #1414: Scheduled
December 23, 2024 00:11 1h 59m 25s master
master
December 23, 2024 00:11 1h 59m 25s
hpu-gaudi2 hpu-gaudi2 #1413: Scheduled
December 22, 2024 00:12 1h 55m 19s master
master
December 22, 2024 00:12 1h 55m 19s
hpu-gaudi2 hpu-gaudi2 #1412: Scheduled
December 21, 2024 00:11 1h 57m 58s master
master
December 21, 2024 00:11 1h 57m 58s
Stage3: Use new torch grad accumulation hooks API hpu-gaudi2 #1411: Pull request #6773 synchronize by tohtana
December 20, 2024 18:16 57m 7s deepcharm:stage3-use-new-grad-acc-api
deepcharm:stage3-use-new-grad-acc-api
December 20, 2024 18:16 57m 7s
Stage3: Use new torch grad accumulation hooks API hpu-gaudi2 #1410: Pull request #6773 synchronize by loadams
December 20, 2024 00:55 2h 8m 32s deepcharm:stage3-use-new-grad-acc-api
deepcharm:stage3-use-new-grad-acc-api
December 20, 2024 00:55 2h 8m 32s
hpu-gaudi2 hpu-gaudi2 #1409: Scheduled
December 20, 2024 00:11 1h 57m 2s master
master
December 20, 2024 00:11 1h 57m 2s
Stage3: Use new torch grad accumulation hooks API hpu-gaudi2 #1408: Pull request #6773 synchronize by loadams
December 19, 2024 20:32 55m 51s deepcharm:stage3-use-new-grad-acc-api
deepcharm:stage3-use-new-grad-acc-api
December 19, 2024 20:32 55m 51s
Zero2: avoid graph breaks in torch.compile by using param_idx hpu-gaudi2 #1407: Pull request #6803 synchronize by loadams
December 19, 2024 17:36 56m 24s nelyahu:zero2_param_idx
nelyahu:zero2_param_idx
December 19, 2024 17:36 56m 24s