duoyw

Why generate many specific Skills for each scenario rather than using basic keyboard and mouse-related Skills?

3

The greatest benefit of vision-based work is that it does not require additional support for each application. However, it seems that you have violated this basic principle in your code....

I noticed that currently all tasks are using the same goal. Is this the intended design? In my understanding, each task should have its own specific goal derived from its...

Poor RL Documentation: Outdated Configs and Missing Data。 Where is the data of RL？

1

I noticed that in the RL example (https://qlib.readthedocs.io/en/latest/component/rl/quickstart.html#), train yml needs to be configured, but the sample yml provided in the documentation cannot run at all, as it contains a...

## 🐛 Bug Description ## To Reproduce Steps to reproduce the behavior: 1. create train_config.yml following doc： https://qlib.readthedocs.io/en/latest/component/rl/quickstart.html 2. run python -m qlib.rl.contrib.train_onpolicy.py --config_path train_config.yml following your document of RL...