You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
How to run scripts/nlp_language_modeling/prepare_packed_ft_dataset.py without GPUs?
The cost of GPU nodes is very expensive, especially when the models are large.
currenlty, we require a full nemo model file for simplicity and readability of code, but in theory only a tokenizer file is needed.
This part can be improved in a future iteration of the script.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
How to run
scripts/nlp_language_modeling/prepare_packed_ft_dataset.py
without GPUs?The cost of GPU nodes is very expensive, especially when the models are large.
Additionally, do you have any plans to improve this?
https://github.com/NVIDIA/NeMo/blob/main/scripts/nlp_language_modeling/prepare_packed_ft_dataset.py#L55-L56
Beta Was this translation helpful? Give feedback.
All reactions