Add streaming inference & fix stopping at EOS #180
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
Glavin001
commented
Jun 10, 2023
|
|
||
| prompter_module = getattr(importlib.import_module("axolotl.prompters"), prompter) | ||
|
|
||
| while True: | ||
| print("=" * 80) |
There was a problem hiding this comment.
Let me know if you like or dislike the separators. Can remove.
|
Thank you for PR. I am currently still thinking how to handle this. Since we might not only use alpaca format for inference, I'm thinking of how to make the Edit: as winglian has approved, I guess we can deal with the above at a later point. The streaming is a great addition! |
winglian
approved these changes
Jun 10, 2023
winglian
reviewed
Jun 10, 2023
mkeoliya
pushed a commit
to mkeoliya/axolotl
that referenced
this pull request
Dec 15, 2023
…ference Add streaming inference & fix stopping at EOS
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Outdated suggestions cannot be applied.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's new?
When I was testing inference with the Falcon config file it used eos_token=<|endoftext|> instead of what was previously being set in the do_inference() code was
<s/>so it wouldn’t stop early it would always use the entire 1024 max new tokens.Demo
https://www.loom.com/share/03f8c8903b42484ea7aad72ab07a6340