Alexander Dibrov

Use two image captioning sources

8

Concatenate captions from 'Salesforce/blip-image-captioning-large' and 'nlpconnect/vit-gpt2-image-captioning' for a more elaborate description of the input. This modification produces desirable results for me in chat-oriented scenarios, but requires additional HD resources. I...

Prompt changes from silent llama fix submitted as a separate PR, at the request of the maintainer.

Fix for silent llama

8

Fix for prioritization agent deleting priority list when returning empty output. Prompt engineering and parameter improvements to reduce number of empty responses from llama models. More debug output. @jmtatsch -...