FBI-LLM: Scaling Up Fully Binarized LLMs from Scratch via Autoregressive Distillation
This work presents a Fully BInarized Large Language Model (FBI-LLM), demonstrating for the first time how to train a large-scale binary language model from scratch (not the partial binary or ternary LLM like BitNet b1.58) to match the performance of its full-precision counterparts (e.g., FP16 or BF16) in transformer-based LLMs. It achieves this by employing an autoregressive distillation (AD) loss with maintaining equivalent model dimensions (130M, 1.3B, 7B) and training data volume as regular LLM pretraining, while delivering competitive results in terms of perplexity and task-specific effectiveness. Intriguingly, by analyzing the training t
Resource information
- Authors
- Ma, Liqun, Sun, Mingjie, Shen, Zhiqiang
- Research area
- Artificial intelligence
- Research topics
- Computation and Language (cs.CL), Artificial Intelligence (cs.AI), Machine Learning (cs.LG), FOS: Computer and information sciences, FOS: Computer and information sciences
- Publication date
- 2024
- DOI
- 10.48550/arxiv.2407.07093
- Upstream last modified
- Jul 9, 2024
- License
- Creative Commons Attribution 4.0 International
- Last record update
- Nov 5, 2026
About this resource
This work presents a Fully BInarized Large Language Model (FBI-LLM), demonstrating for the first time how to train a large-scale binary language model from scratch (not the partial binary or ternary LLM like BitNet b1.58) to match the performance of its full-precision counterparts (e.g., FP16 or BF16) in transformer-based LLMs. It achieves this by employing an autoregressive distillation (AD) loss with maintaining equivalent model dimensions (130M, 1.3B, 7B) and training data volume as regular LLM pretraining, while delivering competitive results in terms of perplexity and task-specific effectiveness. Intriguingly, by analyzing the training trajectory, we find that the pretrained weight is not necessary for training binarized LLMs from scratch. This research encourages a new computational framework and may facilitate the future design of specialized hardware tailored for fully 1-bit LLMs. We make all models, code, and training dataset fully accessible and transparent to support further research (Code: https://github.com/LiqunMa/FBI-LLM. Model: https://huggingface.co/LiqunMa/).