RLFactory is a plug-and-play RL post-training framework for multi-turn tool use, reporting 0.486 on NQ with Qwen3-4B versus 0.473 for Qwen2.5-7B, and 6.8x faster training throughput.
Agentic reasoning and tool integration for llms via reinforcement learning, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
RLFactory: A Plug-and-Play Reinforcement Learning Post-Training Framework for LLM Multi-Turn Tool-Use
RLFactory is a plug-and-play RL post-training framework for multi-turn tool use, reporting 0.486 on NQ with Qwen3-4B versus 0.473 for Qwen2.5-7B, and 6.8x faster training throughput.