3 papers
cs.LG2025
CafeQ: Calibration-free Quantization via Learned Transformations and Adaptive Rounding
Ziteng Sun, Adrian Benton, Samuel Kushnir +4
Post-training quantization is an effective method for reducing the serving cost of large language models, where the standard approach is to use a round-to-nearest quantization leve…
cs.CL2025
Improving Informally Romanized Language Identification
Adrian Benton, Alexander Gutkin, Christo Kirov +1
The Latin script is often used to informally write languages with non-Latin native scripts. In many cases (e.g., most languages in India), the lack of conventional spelling in the…
cs.CL2024
Exploring and Improving Drafts in Blockwise Parallel Decoding
Taehyeon Kim, Ananda Theertha Suresh, Kishore Papineni +3
Despite the remarkable strides made by autoregressive language models, their potential is often hampered by the slow inference speeds inherent in sequential token generation. Block…