Stefano Amorelli
Stefano Amorelli
Hey, your package rocks! :100: Just wondering if you would accept a PR that implements a destination folder option for the generated .vue I would use that feature since I...
I'm using WezTerm with `rebecca` color scheme, if I don't have `set termguicolors` in my `init.vim` all the commits and commits' dates are white. Is there anyway to support the...
Add [Dart bindings client](https://github.com/stefanoamorelli/kafka_dart) in the `README.md`.
The REST API already has [eval-related endpoints](https://github.com/google/adk-go/blob/main/server/restapi/routers/eval.go) (`/apps/{app_name}/eval_sets`, `/apps/{app_name}/eval_results`) but they're all stubbed with `Unimplemented` handlers. There's currently no way to systematically evaluate agent performance. Without this, we're stuck manually...
This PR introduces an evaluation framework for testing and measuring AI agent performance, it supports both algorithmic and LLM-as-Judge evaluation methods, with built-in support for response quality, tool usage, safety,...
Implement nucleus sampling (top-p sampling) as a new sampling method in the Gemma text generation toolkit. This addresses the a gap in `gemma/gm/text/__init__.py:34` and provides the missing sampling strategy. ##...
I've noticed that with a simple `eslint` configuration, `frontend/` returns `✖ 248 problems (70 errors, 178 warnings)`. Especially as the codebase grows, I would suggest to tackle this issue sooner...
When viewing eval results, if `response_match_score` failed but `tool_trajectory_avg_score` passed, all messages in the invocation (including tool calls) incorrectly showed ❌. This can be confusing because the tool trajectory is...
### Problem Statement AFAIK we don't have a pattern for: "I'm not sure which approach is best at performing a task beforehand, **so try several at the same time** and...