server : add "token healing" support

[CyberShadow]

Prerequisites

Please answer the following questions for yourself before submitting an issue.

• [X] I am running the latest code. Development is very rapid so there are no tagged versions as of now.
• [X] I carefully followed the README.md.
• [X] I searched using keywords relevant to my issue to make sure that I am creating a new issue that is not already open (or closed).
• [X] I reviewed the Discussions, and have a new bug or useful enhancement to share.

Feature Description

Hi! I am experimenting with using llama.cpp as a general-purpose code completion backend, similar to TabNine.

I am encountering a small problem: if the completion prompt ends mid-word, the results are not very accurate. For example, for a prompt such as Five, Four, Thre [sic], the model will often ignore the typo and suggest , Two (forming Thre, Two).

I think, as an option to the /completion server API, the following optional behavior would be useful:

1. Tokenize the text
2. Chop off the last token
3. Run the prediction with the remaining tokens, but only consider those tokens whose bytes start with the bytes of the last token.

Thanks!

[stduhpf]

The usual name for this feature is "token healing". I agree that it would be nice to have it supported here.

[ilyannn]

@ggerganov I'd like to try working on it as my first issue!

[ggerganov]

Ok. This can be demonstrated in one of the examples. One way would be to add it to main or simple + extend llama_sampling_sample with the necessary functionality

[mare5x]

Hi @ilyannn, do you still want to work on this? I've created a draft PR (#7028) that demonstrates token healing, but I still haven't added it to main or server. We can collaborate on that, if you'd like.

[ilyannn]

@mare5x Sorry, I have not actually started so please don't wait for me. I'll try to take a look at your PR this week though and will be happy to help in any way I can.