laya-zh-eval: independent Chinese skill-routing eval of all three checkpoints

20 real-world Chinese requests × 3 checkpoints, measuring accuracy, latency and whether confidence knows when it is wrong. Includes a Windows/Python 3.14 segfault workaround.

Open on GitHub

Preview of laya-zh-eval: independent Chinese skill-routing eval of all three checkpoints

Launch: Laya
Use case: Agents
Format: Repo
Source: github.com

Izero07

A computer trainee

See all

codejunkie99

@codejunkie99 · Sep 22

tswawa

@tswawa · Sep 23

ipenywis

@ipenywis · Sep 21

A port of the jev-ultrafast browser agent that decides with local laya-mlx instead of the hosted Jev API, covering DOM snapshot, executor and safety checks. Runs free on Apple Silicon.

wdobry

@wdobry · Sep 20

Open on GitHub

laya-zh-eval: independent Chinese skill-routing eval of all three checkpoints

More built with Laya