OpenChatKit - Everything you need to know about this ChatGPT Alternative - Open Source ChatBOT Model
OpenChatKit provides a powerful, open-source base to create both specialized and general purpose chatbots for various applications. We collaborated with LAION and Ontocord to create the training dataset. Much more than a model release, this is the beginning of an open source project.
A useful chatbot needs to follow instructions in natural language, maintain context in dialog, and moderate responses. OpenChatKit provides a base bot, and the building blocks to derive purpose-built chatbots from this base.
The kit has 4 key components:
An instruction-tuned large language model, fine-tuned for chat from EleutherAI’s GPT-NeoX-20B with over 43 million instructions on 100% carbon negative compute;
Customization recipes to fine-tune the model to achieve high accuracy on your tasks;
An extensible retrieval system enabling you to augment bot responses with information from a document repository, API, or other live-updating information source at inference time;
A moderation model, fine-tuned from GPT-JT-6B, designed to filter which questions the bot responds to.
OpenChatKit includes tools that allow users to provide feedback and enable community members to add new datasets; contributing to a growing corpus of open training data that will improve LLMs over time.
Links:
Together Announcement - https://www.together.xyz/blog/opencha...
OpenChatKit Demo on Hugging Face Spaces - https://huggingface.co/spaces/togethe...
OpenChatKit Github Repo - https://github.com/togethercomputer/O...
GPT-NeoXT-Chat-Base-20B - https://huggingface.co/togethercomput...