PTQ and vector quantization aren’t used for this because part of the point of ternary LLMs is to make them faster. In a ternary LLM every weight is an add, subtract, or no-op so it is fast on CPU.
replyIf you’re just using a code book to reconstruct a f16 model the only savings you can get are in sending it over the wire.
And storing it in memory. Memory is expensive.
replywhich is important though since sending it across the wire over and over and over is actually the main bottleneck.
replyWire typically means internet connection, and it’s hardly the bottleneck
replyin the case of large language models, the wire is the communication of your parameters between your layers of memory that is often the bottleneck. To do a forward pass, you need to use all parameters once, and so the communication between the compute and the storage is the bottleneck, and that bottleneck is also a bunch of wires.
replyThe other bottleneck is the amount of fast storage, which compression also improves.
replydeleted
reply> If you’re just using a code book to reconstruct a f16 model the only savings you can get are in sending it over the wire.
replyThat’s why you need to use efficient gemm kernels like FLUTE for inference. They are ~as good as what you can do with ternary quantization.
If you want sub-2 bit llm, get one that’s already trained in higher precision, and compress it with something like YAQA/QTIP with finetuning or PV-tuning + AQLM/HIGGS
reply