تقرير
Textbooks Are All You Need
| العنوان: | Textbooks Are All You Need |
|---|---|
| المؤلفون: | Gunasekar, Suriya, Zhang, Yi, Aneja, Jyoti, Mendes, Caio César Teodoro, Del Giorno, Allie, Gopi, Sivakanth, Javaheripi, Mojan, Kauffmann, Piero, de Rosa, Gustavo, Saarikivi, Olli, Salim, Adil, Shah, Shital, Behl, Harkirat Singh, Wang, Xin, Bubeck, Sébastien, Eldan, Ronen, Kalai, Adam Tauman, Lee, Yin Tat, Li, Yuanzhi |
| سنة النشر: | 2023 |
| المجموعة: | Computer Science |
| مصطلحات موضوعية: | Computer Science - Computation and Language, Computer Science - Artificial Intelligence, Computer Science - Machine Learning |
| الوصف: | We introduce phi-1, a new large language model for code, with significantly smaller size than competing models: phi-1 is a Transformer-based model with 1.3B parameters, trained for 4 days on 8 A100s, using a selection of ``textbook quality" data from the web (6B tokens) and synthetically generated textbooks and exercises with GPT-3.5 (1B tokens). Despite this small scale, phi-1 attains pass@1 accuracy 50.6% on HumanEval and 55.5% on MBPP. It also displays surprising emergent properties compared to phi-1-base, our model before our finetuning stage on a dataset of coding exercises, and phi-1-small, a smaller model with 350M parameters trained with the same pipeline as phi-1 that still achieves 45% on HumanEval. Comment: 26 pages; changed color scheme of plot. fixed minor typos and added couple clarifications |
| نوع الوثيقة: | Working Paper |
| URL الوصول: | http://arxiv.org/abs/2306.11644 |
| رقم الأكسشن: | edsarx.2306.11644 |
| قاعدة البيانات: | arXiv |
| الوصف غير متاح. |