An RL-trained agent whose job is to write RL training jobs for smaller models. This is the LoRA adapter from the ai-trains-ai project — the trainer agent that sits in the outer loop of two nested RL loops. Derived from Qwen/Qwen3.6-35B-A3B (Apache-2.0). This repository contains only the LoRA adapter weights (rank 8, alpha 32); you need the base model to use it. Given a task spec («teach a small model to do X»), the agent works in a sandboxed workspace with file tools and writes a complete prime-rl training job: a verifiers environment + rubric, a dataset, and a hyperparameter config. Each job it submits is validated, then GRPO-trains a small Qwen (0.6B or 1.7B) on real GPUs; the trained model’s improvement on a hidden eval flows back as the agent’s reward. It is designed to operate inside the project’s harness (workspace layout, file tools, validation probe, task-spec format) — see the GitHub repo to run full episodes. Sampled bare, it is still Qwen3.6-35B-A3B plus a strong prior for writing verifiers environments and prime-rl configs. This is the step-34 checkpoint of the 54-step pilot-7 arc (end of the pilot-7b leg) — the peak of held-out generalisation: — On a held-out task…
Модальности:
Генерация текста
Задача: Генерация текста
Автор: Danau5tin
Теги: peft, lora, grpo, reinforcement-learning, agent, tinker
Лайков: 4 | Загрузок: 5
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.