Convert: Support DT_BF16 tensors #1309
Merged
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
ivanstepanovftw
changed the title
prusnak
requested changes
May 4, 2023
Co-authored-by: Pavol Rusnak <pavol@rusnak.io>
remove useless comment
|
Tested it before, works great. Before these changes it was neccessary to convert pytorch model to fp32 pytorch model, and then to ggml. There is also discussions on model's hf page about BF16. |
ggerganov
approved these changes
May 4, 2023
prusnak
approved these changes
May 4, 2023
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Outdated suggestions cannot be applied.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.
For models like PygmalionAI/metharme-7b and PygmalionAI/pygmalion-7b
By default, converts BF16 to FP32 for better precision.
🤖 Generated by Copilot at c47b349
Add bfloat16 support for PyTorch models in
convert.py. Define a new data type constant, a conversion function, and update the tensor handling logic.