Feature
- 🤖 Support gpt-3.5-turbo-16k
- 🔧 Improve the stop response API, so that it can be stopped in the middle without losing message content
- 🌟 Optimize the weaviate batch size by modify env
WEAVIATE_BATCH_SIZEto reduce the timeout issue caused by too large batch size.
What's Changed
- Fix/excel data format by @JohnJyong in #334
- fix: button abnormal style by @jiaowoxiaobala in #333
- Create README_ES.md by @SergioRico1 in #335
- feat: stop response call api by @iamjoel in #340
- Update README.md by @Panmuse in #341
- Update README_CN.md by @Panmuse in #342
- feat: optimize api language support by @takatost in #344
- feat: optimize weaviate batch size by @takatost in #348
- feat: add WEAVIATE_BATCH_SIZE by @takatost in #349
- Feature/add emoji to webapp by @crazywoola in #345
- fix: stop completion response not save to db by @takatost in #351
- feat: use react sentry to support pass config via runtime by @iamjoel in #350
- feat: add frontend sentry docker compose config by @iamjoel in #353
- feat: stop response enchancement by @iamjoel in #352
- fix: change default help link to english by @iamjoel in #354
- fix: completion stop invalid by @takatost in #355
- feat: add gpt-3.5-turbo-16k support and update openai gpt-3.5-turbo & Embedding Ada v2 unit price by @takatost in #358
- feat: bump version to 0.3.3 by @takatost in #359
New Contributors
- @jiaowoxiaobala made their first contribution in #333
- @SergioRico1 made their first contribution in #335
Full Changelog: 0.3.2...0.3.3