Skip to content

Actions: fattorib/ZeRO-transformer

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
157 workflow runs
157 workflow runs
Event

Filter by event

Loading
Status

Filter by status

Loading
Branch
Actor

Filter by actor

Loading
n=6
Tests #1123: Commit 70c42a3 pushed by fattorib
June 6, 2023 13:19 4m 14s pjit-tensor-parallel
June 6, 2023 13:19 4m 14s
any
Tests #1122: Commit 8da63af pushed by fattorib
June 6, 2023 13:15 3m 30s pjit-tensor-parallel
June 6, 2023 13:15 3m 30s
drop resume args
Tests #1121: Commit 64e9aae pushed by fattorib
June 6, 2023 13:14 3m 43s pjit-tensor-parallel
June 6, 2023 13:14 3m 43s
torch configs to match
Tests #1120: Commit 622f455 pushed by fattorib
June 6, 2023 10:16 3m 36s pjit-tensor-parallel
June 6, 2023 10:16 3m 36s
tweak init
Tests #1119: Commit 7d05b98 pushed by fattorib
June 5, 2023 22:39 3m 38s pjit-tensor-parallel
June 5, 2023 22:39 3m 38s
2.4B
Tests #1118: Commit b9b441f pushed by fattorib
June 5, 2023 21:52 4m 20s pjit-tensor-parallel
June 5, 2023 21:52 4m 20s
extra args for proper conversion to torch
Tests #1117: Commit cedc118 pushed by fattorib
June 5, 2023 19:22 4m 40s pjit-tensor-parallel
June 5, 2023 19:22 4m 40s
jnp.mean call required
Tests #1116: Commit 956fe15 pushed by fattorib
June 5, 2023 18:50 3m 32s pjit-tensor-parallel
June 5, 2023 18:50 3m 32s
log 2.7B in torch
Tests #1115: Commit a153a9f pushed by fattorib
June 5, 2023 18:37 3m 35s pjit-tensor-parallel
June 5, 2023 18:37 3m 35s
log FLOP/MFU benchmarks for reference
Tests #1114: Commit 08b8a4b pushed by fattorib
June 5, 2023 17:49 4m 27s pjit-tensor-parallel
June 5, 2023 17:49 4m 27s
remove unused methods
Tests #1113: Commit 2c8da2f pushed by fattorib
June 5, 2023 16:32 4m 11s pjit-tensor-parallel
June 5, 2023 16:32 4m 11s
2.7 @ 200B
Tests #1112: Commit e638b2e pushed by fattorib
June 5, 2023 12:50 3m 34s pjit-tensor-parallel
June 5, 2023 12:50 3m 34s
smaller for more steps
Tests #1111: Commit b75373f pushed by fattorib
June 5, 2023 12:47 3m 44s pjit-tensor-parallel
June 5, 2023 12:47 3m 44s
remove sleep call
Tests #1110: Commit 6568682 pushed by fattorib
June 5, 2023 12:26 5m 6s pjit-tensor-parallel
June 5, 2023 12:26 5m 6s
drop bf16 cast
Tests #1109: Commit 98ad619 pushed by fattorib
June 5, 2023 12:26 4m 24s pjit-tensor-parallel
June 5, 2023 12:26 4m 24s
arg for tied embeddings in torch model
Tests #1108: Commit ebdaf95 pushed by fattorib
June 5, 2023 10:31 3m 44s pjit-tensor-parallel
June 5, 2023 10:31 3m 44s
device_get scope error
Tests #1106: Commit efe6a3d pushed by fattorib
June 5, 2023 09:52 3m 59s pjit-tensor-parallel
June 5, 2023 09:52 3m 59s
name str required for barrier
Tests #1105: Commit b872693 pushed by fattorib
June 5, 2023 09:45 3m 32s pjit-tensor-parallel
June 5, 2023 09:45 3m 32s
multihost barrier
Tests #1104: Commit b3ca56a pushed by fattorib
June 5, 2023 09:39 3m 54s pjit-tensor-parallel
June 5, 2023 09:39 3m 54s
split up opt/param checkpointing
Tests #1103: Commit 45b37f3 pushed by fattorib
June 5, 2023 09:33 3m 44s pjit-tensor-parallel
June 5, 2023 09:33 3m 44s
remove unused
Tests #1102: Commit 453ef9a pushed by fattorib
June 5, 2023 09:22 3m 30s pjit-tensor-parallel
June 5, 2023 09:22 3m 30s
3b model
Tests #1101: Commit ee998d4 pushed by fattorib
June 4, 2023 20:56 4m 8s pjit-tensor-parallel
June 4, 2023 20:56 4m 8s
300M less params
Tests #1100: Commit c653900 pushed by fattorib
June 4, 2023 19:45 3m 30s pjit-tensor-parallel
June 4, 2023 19:45 3m 30s
short 3B model
Tests #1099: Commit 9992de9 pushed by fattorib
June 4, 2023 19:00 3m 33s pjit-tensor-parallel
June 4, 2023 19:00 3m 33s