References
- 8-Bit Approximations for Parallelism in Deep LearningTim Dettmers2015
- 8-bit Optimizers via Block-wise QuantizationTim Dettmers, Mike Lewis, Sam Shleifer, Luke Zettlemoyer2021
- The case for 4-bit precision: k-bit Inference Scaling LawsTim Dettmers, Luke Zettlemoyer2022