This week includes another iteration on major performance improvements from our amazing community. Thanks to @kga245 for the awesome contribution, rebuilding the entire log handler for more robust capabilities. We've also added support for Tavily Extract that solves scale for developers shipping GPT Researcher in production. Tavily Extract allows to scrape any site at scale without getting blocked by site providers or dealing with the hassle of proxies, etc. Thank you to the best community on Github and looking forward to your next contributions. And like always welcome to our new contributors!
What's Changed
- Prints unexpected Google retriever search API response by @ivarprudnikov in #1036
- Logs handler improvements by @kga245 in #1038
- Added tutorial playlist to documentation by @gaurav3247 in #1042
- Bug fixes to unified logs by @kga245 in #1039
- Support Bedrock embedding model by @moritalous in #1045
- Update hybrid_research.md by @MC-shark in #1046
- 【Added】Enable support for processing online documents in hybrid mode by @MC-shark in #1043
- Adding Tavily extract by @pulvedu in #1044
- Fix Gemini bug when system role is provided by @winsonluk in #1048
New Contributors
- @ivarprudnikov made their first contribution in #1036
- @gaurav3247 made their first contribution in #1042
- @moritalous made their first contribution in #1045
- @pulvedu made their first contribution in #1044
Full Changelog: v3.1.6...v.3.1.7