NeuroGPT, an API service for language models by NeurogenAI, is being significantly restructured. Recent updates have mostly been to improve the web interface ('gpt4free' and 'ChuanhuChatGPT') with a focus on features such as web search, no-logs, and built-in templates. However, communication around the ongoing API restructuring and client transition is vague in the README, lacking detailed information and dates.
While there have been 28 recently closed issues, there are currently two major open issues:
Furthermore, while the team appears active in addressing the variety of issues reported, there is no clear guidance for contributors.
There is one open PR #54 for modifying the hard binding of interpreter version. Recently, PRs have been focused on improving templates, webui, and the launch process. Notably, PR #28 aimed to fix an overwrite issue in 'config.json' but was closed without merging.
Open Pull Requests:
Recently Closed Pull Requests:
Notable Pull Requests: - #28: Attempted fix for overwrite issue in 'config.json'. Closed without merge. - #5: Added Docker support. Merged. - #3: Fixed issue with streams for gradio. Merged. - #2: Added 'requirements_linux.txt'. Merged.
"NeuroGPT" is a project by NeurogenAI that provides a free API service for accessing GPT-3.5, GPT-4, and other language models. The service is currently undergoing a client transition and an API reconstruction, causing some disruptions. The project makes use of a modified version of the 'gpt4free' and 'ChuanhuChatGPT' as a web interface. Key features include web search, model list, dialog context, no-logs, API endpoint, dialog history, and built-in templates for various tasks.
The README lacks clarification regarding the state of the ongoing transition and reconstruction process – specifics about potential disruptions and estimated completion times are missing. There are no TODOs or anomalies detected. There’s also no clear instruction on how to contribute or report issues.