Enable different torch dtype in sub models #34873
Open
+13
−4
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
zucchini-nlp
commented
Nov 22, 2024
Comment on lines
+4051
to
+4058
| for sub_config_key in config.sub_configs.keys(): | ||
| sub_config = getattr(config, sub_config_key) | ||
| sub_config.torch_dtype = torch_dtype | ||
| elif isinstance(torch_dtype, dict): | ||
| for key, curr_dtype in torch_dtype.items(): | ||
| if hasattr(config, key): | ||
| value = getattr(config, key) | ||
| value.torch_dtype = curr_dtype |
There was a problem hiding this comment.
if users passes one torch dtype as before, we just use it in all sub-configs. Otherwise a user can either set directly dtypes in configs before loading the model, or indicate a dict torch_dtype when loading similarly to attn_implementation_dispatch
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
What does this PR do?
Fixes #33997. Enables users to use different torch dtypes for each of sub config. For ex load the vision model in full precision and the text model in half precision