matatonic
matatonic
This is a first cut of a working api, it supports text completion, with/without streaming and listing models. chat completions are partly working, but stopping strings are not working yet...
### Describe the bug Not all required files in sub-folders are downloaded automatically. Example. The HugginFace model sentence-transformers/all-MiniLM-L6-v2 : the subfolder "1_Pooling" is required and contains the config.json. The sub-folder...
This was pretty simple and works well for translations, spelling, "professional emails", code reformatting, etc.
### Describe the bug An extra space is generated at the beginning of text, It is done here: https://github.com/oobabooga/text-generation-webui/commit/15940e762e9f9a257fb8ce4f711b5e1ca7740616 This breaks a lot of use cases, like python code formatting...
This includes correct instruction templates for each model (not just Alpaca). For now they're in the code, but I will change them to use the on disk format when they're...
with rstrip in the template expansion now, we can safely clip an extra space if it's returned for some models.
Performance when generating with top_p = 1.0 is about 3x slower than with top_p at anything(?) else, to duplicate try 0.99 and 1.0. I've seen this bug with both the...