llm-training

SKILL.md

Frameworks and techniques for training and finetuning large language models.

| Framework | Best For | Multi-GPU | Memory Efficient |

| Accelerate | Simple distributed | Yes | Basic | | DeepSpeed | Large models, ZeRO | Yes | Excellent | | PyTorch Lightning | Clean training loops | Yes | Good | | Ray Train | Scalable, multi-node | Yes | Good | | TRL | RLHF, reward modeling | Yes | Good | | Unsloth | Fast LoRA finetuning | Limited | Excellent |

Úselo cuando "entrene LLM", "ajuste fino", "RLHF", "entrenamiento distribuido", "DeepSpeed", "Accelerate", "PyTorch Lightning", "Ray Train", "TRL", "Unsloth", "entrenamiento LoRA", "atención flash", "puntos de control de gradiente" Fuente: eyadsibai/ltk.

Ver original