miles-rl-training
Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime. Use when training large MoE models with FP8/INT4, needing train-inference alignment, or requiring speculative RL for maximum throughput.
Оригинален текст · няма потвърден превод на българскиПърво прочетете инструкцията.
Файлът се зарежда от посочения източник само по ваше действие. Можете да го прочетете тук, да го копирате или да го запазите като .md.
GitHub хранилище- Звезди на репото
- 12 982
- Лиценз
- MIT
- Последна проверка на записа
- 20.09.2026 г.
Добавете в своя инструмент.
Командата отваря избора на skills от репозитория. Прочетете файла и условията на автора преди изпълнение.
npx skills add Orchestra-Research/AI-Research-SKILLsСтраницата копира текст. Не инсталира и не изпълнява нищо на устройството ви.
Откъс от SKILL.md
Началото на файла на автора, без превод.
miles: Enterprise-Grade RL for Large-Scale Model Training
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.
When to Use miles
Choose miles when you need:
- Training 1TB+ MoE models (DeepSeek V3, Qwen3-MoE)
- FP8 or INT4 quantization-aware training
- Bit-wise identical train-inference alignment
- Speculative RL for maximum throughput
- Production stability with enterprise support
Consider alternatives when:
- You want the research-grade original → use slime
- You need flexible backend swapping → use verl
- You want PyTorch-native abstractions → use torchforge
Key Features
Low-Precision Training
- Unified FP8: End-to-end FP8 for both inference and training
- INT4 QAT: 1TB models on single-machine VRAM (H200)
- Rollout Routing Replay (R3): Bit-wise expert alignment for MoE
Performance Optimizations
- Speculative RL: 25%+ rollout speedup with online SFT draft models
- Zero-Copy Weight Sync: CUDA IPC zero-copy mapping
- Partial Rollout: Recycle half-finished trajectories
Train-Inference Alignment
- TIS/MIS: Truncated/Masked Importance Sampling for off-policy correction
- Kernel-level optimization: FlashAttention-3, DeepGEMM integration
Installation
# Recommended: Docker
docker pull radixark/miles:latest
docker run --rm --gpus all --ipc=host --shm-size=16g \
-it radixark/miles:latest /bin/bash
# From source
git clone https://github.com/radixark/miles.git
cd miles
pip install -r requirements.txt
pip install -e .
Quick Start
miles inherits slime's configuration system. Basic training:
python train.py \
--advantage-estimator grpo \
--model-name qwen3-30b-a3b \
--hf-checkpoint /path/to/qwen3-30b-a3b-hf \
--rollout-batch-size 512 \
--n-samples-per-prompt 8
Workflow 1: Large MoE Training
Use this workflow for training large MoE models like DeepSeek V3 or Qwen3-MoE.
Prerequisites Checklist
- H100/H200 GPUs with FP8 support
- MoE model (DeepSeek V3, Qwen3-MoE)
- Docker environment with miles
Step 1: Environment Setup
# FP8 block scaling (recommended for stability)
export NVTE_FP8_BLOCK_SCALING_FP32_SCALES=1
export CUDA_DEVICE_MAX_CONNECTIONS=1