biobootloader
Rawdog: Command line AI that responds by auto-executing output as scripts
OpenAI Models Dominate Structured Code Edit Benchmark
> I also notice that the instructions prompt that mentat uses seems to be inspired by the aider benchmark? Glad to see others adopting similar benchmarking approaches.
We were inspired by you to use Exercism as a benchmark, thank you! We will add attribution for that. We switched our original instruction prompts for that benchmark to be similar to Aiders to allow for fair comparison.
> After looking around a bit, there seems to be a bunch of aider code in your repo. Some attribution would be appreciated.
We have an unused implementation of your output response format (https://github.com/AbanteAI/mentat/blob/main/mentat/parsers/...), but I don't know what else you are seeing? We implemented that to compare with our response formats and didn't find much difference in performance.
GPT-4-Turbo JSON Response Format Quirks
Abante AI is a new startup building Mentat, an open source, GPT-4 powered coding assistant. Mentat runs on the command line, gathering project context and coordinating edits across multiple files: https://github.com/biobootloader/mentat
- work with a small, talented team on an ambitious project
- open source: share what you build
- apply research to make a real product
- competitive pay + early equity
Contact me: [email protected]
but yeah it is funny
I wouldn't recommend running it on a script that's supposed to delete files
Wolverine: Give your Python scripts regenerative healing abilities
Demo video on twitter: https://twitter.com/bio_bootloader/status/163688020830443110...