GenQA: Generating Millions of Instructions from a Handful of Prompts
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910488765399040 |
|---|---|
| author | Chen, Jiuhai Qadri, Rifaa Wen, Yuxin Jain, Neel Kirchenbauer, John Zhou, Tianyi Goldstein, Tom |
| author_facet | Chen, Jiuhai Qadri, Rifaa Wen, Yuxin Jain, Neel Kirchenbauer, John Zhou, Tianyi Goldstein, Tom |
| contents | Most public instruction finetuning datasets are relatively small compared to the closed source datasets used to train industry models. To study questions about finetuning at scale, such as curricula and learning rate cooldown schedules, there is a need for industrial-scale datasets. However, this scale necessitates a data generation process that is almost entirely automated. In this work, we study methods for generating large instruction datasets from a single prompt. With little human oversight, we get LLMs to write diverse sets of instruction examples ranging from simple completion tasks to complex multi-turn dialogs across a variety of subject areas. When finetuning a Llama-3 8B base model, our dataset meets or exceeds both WizardLM and Ultrachat on both knowledge-intensive leaderboard tasks as well as conversational evaluations. We release our dataset, the "generator" prompts that created it, and our finetuned model checkpoints. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_10323 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | GenQA: Generating Millions of Instructions from a Handful of Prompts Chen, Jiuhai Qadri, Rifaa Wen, Yuxin Jain, Neel Kirchenbauer, John Zhou, Tianyi Goldstein, Tom Computation and Language Most public instruction finetuning datasets are relatively small compared to the closed source datasets used to train industry models. To study questions about finetuning at scale, such as curricula and learning rate cooldown schedules, there is a need for industrial-scale datasets. However, this scale necessitates a data generation process that is almost entirely automated. In this work, we study methods for generating large instruction datasets from a single prompt. With little human oversight, we get LLMs to write diverse sets of instruction examples ranging from simple completion tasks to complex multi-turn dialogs across a variety of subject areas. When finetuning a Llama-3 8B base model, our dataset meets or exceeds both WizardLM and Ultrachat on both knowledge-intensive leaderboard tasks as well as conversational evaluations. We release our dataset, the "generator" prompts that created it, and our finetuned model checkpoints. |
| title | GenQA: Generating Millions of Instructions from a Handful of Prompts |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2406.10323 |