j0yk1ll

Resampler

1

I added the perceiver resampler as used in OpenFlamingo and suggested by finding 4 from "What matters when building vision-language models?" (https://arxiv.org/pdf/2405.02246) by Huggingface. "Reducing the number of visual tokens...

[REQUEST] Colored User Input

4

**How would you improve Rich?** It would be nice to be able to color the user input as well. There currently doesn't seem to be an option for that. **What...

Gemini

2

Added support for gemini. I tested it, but there's one issue at the moment. When I launch it via the devon command I get an `ACCESS_TOKEN_SCOPE_INSUFFICIENT` error. When I run...

- fixed the tool name for find_function - renamed scroll_up_in_editor to scroll_up to be consistent with scroll_down - added .devon.config to gitignore because it should not be commited to github

Add automatic install attemps for different platforms and distros. Not tested on all platforms.

1

While installing I had some issue because of the shebang so I corrected it. It can now be installed via `bash install.sh` or `./install.sh`. Before it couldn't find my python3...

standardized formatting

2

added black formatter and modified pre-commit config to enforce standardized formatting throughout the project, because in the pull requests I often see just spacing / linebreak changes in addition to...

LiteLLM-like Facade

3

Hey I added a litellm like facade that makes it simpler to change models. Before you always had to adjust the langchain imports. Now you just have to adjust the...

Endless loop detection to avoid high LLM usage costs

2

Hey this is not a code specific issue, but an agent specific issue. I use the agent as part of an agentic system and I ran into an endless loop....

[Bug]: ValueError: invalid literal for int() with base 10: 'generateContent' (gemini)

6

### What happened? In version `1.58.2` the above model occurs when I use a gemini model (might for others as well, haven't checked) ### Relevant log output ```shell In version...