[training] apply save_total_limit
to push
#60
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Currently, the argument
save_total_limit
can be used to control the number of locally saved checkpoints, but not those pushed to the Hub. This PR ensuressave_total_limit
is applied to both (in-line with the HF Trainer). It was tested with the following training config:librispeech_tts_r_300M_dummy.json
Which gives a repo where the checkpoints are rotated upon each save step, e.g. sanchit-gandhi/parler-tts-save-total-limit
The PR also saves the model config and generation config, such that a model can be loaded directly from a saved state.