Skip to main content
eleutherai
Projects
neox
Run groups
oFmSBtTMySPgqMmtdGAMqT_3ftpt6ui
Logs
Overview
Workspace
Runs
Logs
Files
Artifacts

Chilli's group workspace

Timestamps visible
2023-08-03 09:09:50
Using ./extensions/ as PyTorch extensions root...
2023-08-03 09:09:50
No modifications detected for re-loaded extension module utils, skipping build step...
2023-08-03 09:09:50
Loading extension module utils...
2023-08-03 09:09:52
[2023-08-03 09:09:52,174] [INFO] [engine.py:1551:_load_checkpoint] rank: 24 loading checkpoint: /fsx/lintangsutawika/checkpoints/temp_neox_models/global_step38002/mp_rank_00_model_states.pt
2023-08-03 09:11:44
successfully loaded 64 ZeRO state_dicts for rank 24
2023-08-03 09:12:15
loading 64 zero partition checkpoints for rank 24
2023-08-03 09:12:23
WARNING: shuffle index length (162165685) is not equal to sample index length (162165686)
2023-08-03 09:12:27
WARNING: shuffle index length (162165685) is not equal to sample index length (162165686)
2023-08-03 09:12:29
WARNING: shuffle index length (162165685) is not equal to sample index length (162165686)
2023-08-03 09:12:29
> RANK 24 elapsed time for building blendable dataset indices: 0.76 (sec)
2023-08-03 09:12:31
> RANK 24 elapsed time for building blendable dataset indices: 1.13 (sec)
2023-08-03 09:12:33
> RANK 24 elapsed time for building blendable dataset indices: 1.21 (sec)
2023-08-03 09:13:04
[2023-08-03 09:13:03,993] [INFO] [engine.py:1805:_copy_recovery_script] creating recovery script /fsx/lintangsutawika/checkpoints/temp_neox_models/zero_to_fp32.py
2023-08-03 09:13:04
[2023-08-03 09:13:04,002] [INFO] [engine.py:1818:_save_zero_checkpoint] zero checkpoint saved /fsx/lintangsutawika/checkpoints/temp_neox_models/global_step38003/zero_pp_rank_24_mp_rank_00_optim_states.pt
2023-08-03 09:13:20
[2023-08-03 09:13:18,602] [INFO] [engine.py:1805:_copy_recovery_script] creating recovery script /fsx/lintangsutawika/checkpoints/temp_neox_models/zero_to_fp32.py
2023-08-03 09:13:20
[2023-08-03 09:13:18,611] [INFO] [engine.py:1818:_save_zero_checkpoint] zero checkpoint saved /fsx/lintangsutawika/checkpoints/temp_neox_models/global_step38004/zero_pp_rank_24_mp_rank_00_optim_states.pt