TinyLlama: An Open-Source Small Language Model

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Peiyuan, Zeng, Guangtao, Wang, Tianduo, Lu, Wei
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910470167855104
author Zhang, Peiyuan
Zeng, Guangtao
Wang, Tianduo
Lu, Wei
author_facet Zhang, Peiyuan
Zeng, Guangtao
Wang, Tianduo
Lu, Wei
contents We present TinyLlama, a compact 1.1B language model pretrained on around 1 trillion tokens for approximately 3 epochs. Building on the architecture and tokenizer of Llama 2, TinyLlama leverages various advances contributed by the open-source community (e.g., FlashAttention and Lit-GPT), achieving better computational efficiency. Despite its relatively small size, TinyLlama demonstrates remarkable performance in a series of downstream tasks. It significantly outperforms existing open-source language models with comparable sizes. Our model checkpoints and code are publicly available on GitHub at https://github.com/jzhang38/TinyLlama.
format Preprint
id arxiv_https___arxiv_org_abs_2401_02385
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle TinyLlama: An Open-Source Small Language Model
Zhang, Peiyuan
Zeng, Guangtao
Wang, Tianduo
Lu, Wei
Computation and Language
Artificial Intelligence
We present TinyLlama, a compact 1.1B language model pretrained on around 1 trillion tokens for approximately 3 epochs. Building on the architecture and tokenizer of Llama 2, TinyLlama leverages various advances contributed by the open-source community (e.g., FlashAttention and Lit-GPT), achieving better computational efficiency. Despite its relatively small size, TinyLlama demonstrates remarkable performance in a series of downstream tasks. It significantly outperforms existing open-source language models with comparable sizes. Our model checkpoints and code are publicly available on GitHub at https://github.com/jzhang38/TinyLlama.
title TinyLlama: An Open-Source Small Language Model
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2401.02385