Efficient Inference for Large Language Models with LLM.int8() скачать в хорошем качестве

Efficient Inference for Large Language Models with LLM.int8()

The podcast discusses a groundbreaking paper titled 'LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale' that introduces a new method for 8-bit matrix multiplication within transformer models to run large language models efficiently without sacrificing performance. The paper addresses the memory-intensive nature of large language models and the challenges of 8-bit quantization accuracy with outlier features in larger models. Engineers can leverage LLM.int8() to reduce memory requirements and efficiently run large language models without performance degradation, even at scales exceeding billions of parameters. The method incorporates vector-wise quantization and mixed-precision decomposition to maintain full 16-bit performance in perplexity and zeroshot accuracy across large models, demonstrating significant memory savings and modest speedups for inference. Read full paper: https://arxiv.org/abs/2208.07339 Tags: Artificial Intelligence, Natural Language Processing, 8-bit Quantization, Transformer Models