@CppGuy @Vorsos @ariadne @BillSaysThis @davidgerard
Interestingly enough, Chinchilla 70B was trained mostly on text and beat domain-specific compressors PNG and FLAC in one experiment.
https://arxiv.org/abs/2309.10668
Not saying you are wrong. I assume that newer, domain-specific algorithms would still outperform the general Chinchilla algorithm, and there can be practical downsides if they involve large memory requirements, even if they result in more efficient compression.
Language Modeling Is Compression
It has long been established that predictive models can be transformed into lossless compressors and vice versa. Incidentally, in recent years, the machine learning community has focused on training increasingly large and powerful self-supervised (language) models. Since these large language models exhibit impressive predictive capabilities, they are well-positioned to be strong compressors. In this work, we advocate for viewing the prediction problem through the lens of compression and evaluate the compression capabilities of large (foundation) models. We show that large language models are powerful general-purpose predictors and that the compression viewpoint provides novel insights into scaling laws, tokenization, and in-context learning. For example, Chinchilla 70B, while trained primarily on text, compresses ImageNet patches to 43.4% and LibriSpeech samples to 16.4% of their raw size, beating domain-specific compressors like PNG (58.5%) or FLAC (30.3%), respectively. Finally, we show that the prediction-compression equivalence allows us to use any compressor (like gzip) to build a conditional generative model.
Peter Gleick
Texas Observer
ProPublica