Zobrazeno 1 - 4
of 4
pro vyhledávání: '"Kou, Siqi"'
Autor:
Lin, Bokai, Zeng, Zihao, Xiao, Zipeng, Kou, Siqi, Hou, Tianqi, Gao, Xiaofeng, Zhang, Hao, Deng, Zhijie
KV cache has become a de facto technique for the inference of large language models (LLMs), where tensors of shape (layer number, head number, sequence length, feature dimension) are introduced to cache historical information for self-attention. As t
Externí odkaz:
http://arxiv.org/abs/2410.14731
Parallel decoding methods such as Jacobi decoding show promise for more efficient LLM inference as it breaks the sequential nature of the LLM decoding process and transforms it into parallelizable computation. However, in practice, it achieves little
Externí odkaz:
http://arxiv.org/abs/2403.00835
Diffusion models have impressive image generation capability, but low-quality generations still exist, and their identification remains challenging due to the lack of a proper sample-wise metric. To address this, we propose BayesDiff, a pixel-wise un
Externí odkaz:
http://arxiv.org/abs/2310.11142
Autor:
Hu, Teng, Zhang, Jiangning, Liu, Liang, Yi, Ran, Kou, Siqi, Zhu, Haokun, Chen, Xu, Wang, Yabiao, Wang, Chengjie, Ma, Lizhuang
Training a generative model with limited number of samples is a challenging task. Current methods primarily rely on few-shot model adaption to train the network. However, in scenarios where data is extremely limited (less than 10), the generative net
Externí odkaz:
http://arxiv.org/abs/2309.03729