Spaces:
Running
Running
| """ | |
| Training module for the API Testing Environment. | |
| Contains: | |
| - prompts.py β System prompt, observation formatting, action parsing | |
| - rewards.py β Reward functions for GRPO (format + environment) | |
| - agents.py β Baseline agents (random, sequential, smart) | |
| - grpo.py β GRPO training loop with TRL, HF Hub push, W&B logging | |
| - evaluate.py β Evaluation / rollout runner (local + remote) | |
| """ | |