6 accepted papers
It is common to reject undesired outputs of Large Language Models (LLMs); however, current methods to do so require an excessive amount of computation to re-sample after a rejection, or distort the distribution of outputs by constraining the output to highly improbable tokens. We present a method, A…
Speculative decoding has emerged as a powerful method to improve latency and throughput in hosting large language models. However, most existing implementations focus on generating a single sequence. Real-world generative AI applications often require multiple responses and how to perform speculativ…
In our study, we present bifurcated attention, a method developed for language model inference in single-context batch sampling contexts. This approach aims to reduce redundant memory IO costs, a significant factor in latency for high batch sizes and long context lengths. Bifurcated attention achiev…
Generative models, widely utilized in various applications, can often struggle with prompts corresponding to partial tokens. This struggle stems from tokenization, where partial tokens fall out of distribution during inference, leading to incorrect or nonsensical outputs. This paper examines a techn…
We present two new benchmarks, MBXP and Multilingual HumanEval, designed to evaluate code completion models in over 10 programming languages. These datasets are generated using a conversion framework that transpiles prompts and test cases from the original MBPP and HumanEval datasets into the corres…
The demand for the deployment of deep neural networks (DNN) on resource-constrained Edge platforms is ever increasing. Today’s DNN accelerators support mixed-precision computations to enable reduction of computational and storage costs but require networks with precision at variable granularity, i.e…