[Bugfix] check if model checkpoint exists before loading #22
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Outdated suggestions cannot be applied.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.
Description
It's impossible to train a model from scratch with the current trainer code. The code attempts to load the model checkpoint even when it doesn't exist, and there is no way to disable it (args.model_file is set by the code even if it's not given explicitly).
Specifically, the model file's existence is not checked and
load-model
argument check is not implemented in/bin/main.py:518
:protnote/bin/main.py
Lines 519 to 530 in 2d69e44
This PR updates the training logic to check if the chcekpoint path exists before loading it.
PS: It may be useful to update docs and also add an
else
clause that logs something along the lines ofINFO: Could not find checkpoint {path}, training from scratch