finance-alpaca

Finance Alpaca 整合了 Stanford Alpaca 和 FiQA 数据集，并使用 GPT3.5 额外生成了 1.3k 个数据对。该数据集主要用于文本生成任务，规模在 1 万到 10 万条之间，并采用 MIT 许可。同时，Finance Alpaca 提供使用 PEFT/LoRa 在 Kaggle 免费资源上进行微调的脚本，以及包含性能分析、训练和数据生成脚本以及推理 notebook 的 GitHub 仓库。此外，Finance Alpaca 还提供更简洁的数据集版本和 CSV 格式的数据集。

Finance Alpaca integrates the Stanford Alpaca and FiQA datasets, and additionally generates 1.3k data pairs using GPT-3.5. This dataset is primarily designed for text generation tasks, with a scale ranging from 10,000 to 100,000 entries, and is licensed under the MIT License. Furthermore, Finance Alpaca provides fine-tuning scripts utilizing PEFT/LoRa on Kaggle's free-tier resources, alongside a GitHub repository containing performance analysis scripts, training and data generation scripts, as well as inference notebooks. Additionally, Finance Alpaca offers a streamlined version of the dataset and the dataset in CSV format.