心有所向，日复一日，必有精进

摘要： Pre title: Aligner: Achieving Efficient Alignment through Weak-to-Strong Correction source: Arxiv 2024 paper: https://arxiv.org/abs/2402.02416 code: h 阅读全文

摘要： Pre title: FontDiffuser: One-Shot Font Generation via Denoising Diffusion with Multi-Scale Content Aggregation and Style Contrastive Learning accepted 阅读全文

摘要： Pre title: CalliGAN: Style and Structure-aware Chinese Calligraphy Character Generator accepted: AICCW(the AI for content creation workshop) at CVPR 2 阅读全文

摘要： Pre title: Separating Style and Content for Generalized Style Transfer accepted: CVPR 2018 paper:https://arxiv.org/abs/1711.06454 code: none 关键词: styl 阅读全文

摘要： Pre title: Learning to Write Stylized Chinese Characters by Reading a Handful of Examples accepted: IJCAI 2018 paper: https://arxiv.org/abs/1712.06424 阅读全文

摘要： Pre title: Joint Implicit Neural Representation for High-fidelity and Compact Vector Fonts accepted: ICCV 2023 paper: https://openaccess.thecvf.com/co 阅读全文

摘要： Pre title: Self-conditioned Image Generation via Generating Representations accepted: arXiv 2023 paper: https://arxiv.org/abs/2312.03701 code: https:/ 阅读全文

摘要： Pre title: Learning Component-Level and Inter-Class Glyph Representation for few-shot Font Generation accepted: ICME 2023 paper: https://ieeexplore.ie 阅读全文

摘要： Pre title: Replacing softmax with ReLU in Vision Transformers accepted: Arxiv 2023 paper: https://export.arxiv.org/abs/2309.08586 code: None 关键词：atten 阅读全文

摘要： tl;dr: pytorch的 torch.optim.lr_scheduler.OneCycleLR 就很不错，能兼顾warmup和余弦学习率，也不用下载额外的包 import torch from torch.optim.lr_scheduler import CosineAnnealingLR 阅读全文