A carefully curated dataset of 25K STEM and logic questions assembled from 11 open datasets with strict deduplication. Fine-tuning Llama 2 on this dataset for just 5 hours achieves GPT-4-level STEM performance, demonstrating quality trumps quantity.
| Provider | garage-bAInd |
| Category | 指令 / SFT |
| Size | 25K Questions |
| License | CC BY NC 4.0 |
| Downloads | 340k |
| Tags | STEM, Reasoning, Curated, Logic, Science |
from datasets import load_dataset
ds = load_dataset("garage-bAInd/Open-Platypus")
使用 QLoRA(4-bit 基础模型 + LoRA 适配器)微调的预计显存需求(保守默认参数):
| 7B QLoRA | ~6GB VRAM |
| 13B QLoRA | ~10GB VRAM |
微调新手?跟着分步教程走: 一小时微调你的第一个 LLM