Improvements to VRAM usage when loading HF models #105
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Outdated suggestions cannot be applied.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.
MII can now optionally load models onto system memory initially with the
load_with_sys_mem
option added to the config. This solves a problem when the model takes up nearly all the GPU memory and performing kernel injection requires some additional space, causing OOM errors despite there being enough GPU memory to hold the model.Also adding the
DTypeEnum
from DeepSpeed to the MII config and unit tests to test this new option