remove model from accelerate prepare and add precision argument #61
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Outdated suggestions cannot be applied.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.
Passing both model and dataloader to
accelerate.prepare
takes unnecessary memory as noticed by @RaymondLi0, which causes OOM for large models.This is because the model is wrapped in the
DistributedDataParallel
class which will reserve memory for the gradients for training (issue). We now only wrap the dataloader, and we also addprecision
argument to properly load model in bf16 or fp16. (the mixed-precisonaccelerate
argument in config is for mixed precision in training and will load two model copies..)Todo: add cpu case