Quan Zhou

Bo Lv, Quan Zhou, Xuanang Ding et al.

The bottleneck associated with the key-value(KV) cache presents a significant challenge during the inference processes of large language models. While depth pruning accelerates inference, it requires extensive recovery training, which can take up to two weeks. On the other hand, width pruning retains much of the performance but offers slight speed gains. To tackle these challenges, we propose KVPruner to improve model efficiency while maintaining performance. Our method uses global perplexity-based analysis to determine the importance ratio for each block and provides multiple strategies to prune non-essential KV channels within blocks. Compared to the original model, KVPruner reduces runtime memory usage by 50% and boosts throughput by over 35%. Additionally, our method requires only two hours of LoRA fine-tuning on small datasets to recover most of the performance.

Jhanvi Garg, Xianyang Zhang, Quan Zhou

Schrödinger bridge can be viewed as a continuous-time stochastic control problem where the goal is to find an optimally controlled diffusion process whose terminal distribution coincides with a pre-specified target distribution. We propose to generalize this problem by allowing the terminal distribution to differ from the target but penalizing the Kullback-Leibler divergence between the two distributions. We call this new control problem soft-constrained Schrödinger bridge (SSB). The main contribution of this work is a theoretical derivation of the solution to SSB, which shows that the terminal distribution of the optimally controlled process is a geometric mixture of the target and some other distribution. This result is further extended to a time series setting. One application is the development of robust generative diffusion models. We propose a score matching-based algorithm for sampling from geometric mixtures and showcase its use via a numerical example for the MNIST data set.