Baize: An Open-Source Chat Model with Self-Chat Data
Free while signed in. Answers cite the passages they came from.
First page

The curator’s take
Key pointsAn open chat model fine-tuned with LoRA on self-chat dialogs.
01
Self-chat data generation: Generates 100K dialogs by having ChatGPT converse with itself, then fine-tunes on these dialogs.
02
LoRA fine-tuning: Uses parameter-efficient LoRA fine-tuning for compute efficiency.
03
Multiple model sizes: Releases 7B, 13B, and 30B parameter models along with the dialog data.
04
Open chatbot ecosystem: Part of the 2023 proliferation of open chat models (Vicuna, Alpaca, Koala, Baize) building on LLaMA.
Every Monday
Get next week’s papers.
The same picks and the same summaries, in your inbox. Free, and 176 issues deep.
Subscribe on Substack