If your query "NF4.rar" refers to a biological or medical study, it likely points to research involving (a protein) and RAR (Retinoic Acid Receptor), specifically in the context of Acute Promyelocytic Leukemia . Topic : Arsenic trioxide treatments.
The paper explains why NF4 is superior to standard 4-bit integers (Int4) or floating-point (Float4) formats:
: Neural network weights typically follow a normal distribution. NF4 concentrates its 16 "bins" where most weights exist (near zero), minimizing rounding errors. NF4.rar
: Compresses 16-bit weights to 4 bits, effectively reducing VRAM usage by ~75% (e.g., a 65B parameter model can be loaded with ~35GB instead of ~130GB).
The term "NF4" is central to this "long paper" which revolutionized how large language models (LLMs) are fine-tuned on consumer hardware. If your query "NF4
💡 : If you are looking for the software/machine learning paper, search for "QLoRA" or "4-bit NormalFloat" on arXiv .
: An information-theoretically optimal data type for normally distributed weights. It uses 16 quantization levels based on the quantiles of a standard normal distribution. NF4 concentrates its 16 "bins" where most weights
: To reduce the memory footprint of LLMs (like Llama) enough to fit on a single GPU (e.g., a 24GB RTX 3090) while maintaining full 16-bit performance.