miles-rl-training

SKILL.md

miles is a high-performance, enterprise-ready RL framework optimized for large-scale model post-training. Built as a production fork of slime, it addresses critical challenges in MoE training stability, low-precision training, and train-inference alignment.

Use this workflow for training large MoE models like DeepSeek V3 or Qwen3-MoE.

Use this workflow for maximum rollout throughput with EAGLE speculative decoding.

Proporciona orientación para la capacitación de RL de nivel empresarial utilizando miles, un tenedor de limo listo para producción. Úselo cuando entrene modelos MoE grandes con FP8/INT4, necesite alineación de inferencia de tren o requiera RL especulativo para obtener el máximo rendimiento. Fuente: orchestra-research/ai-research-skills.

Ver original