SIRI: Scaling Iterative Reinforcement Learning with Interleaved Compression — Interleaved Compression–Expansion: — Compression phase: forces concise, high-density reasoning by limiting rollout length. — Expansion phase: restores longer rollouts to encourage exploration and planning. — Token Efficiency without Accuracy Loss: Unlike previous methods, SIRI improves accuracy while reducing average token usage. — Iterative RL Training: Built on GRPO with modifications from DAPO (clip-high/low decoupling, KL removal). — Generalization Across Model Sizes: Validated on both 1.5B and 7B models.
Модальности:
Генерация текста
Области применения:
Диалог / чат
Задача: Генерация текста
Автор: THU-KEG
Теги: qwen2, reinforcement-learning, conversational, en, zh, text-generation-inference, endpoints_compatible
Лайков: 3 | Загрузок: 38
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
