随笔分类 -  Deep Learning

摘要:A Convolutional Neural Network (CNN) is a type of Deep Learning architecture commonly used for image classification and recognition tasks. It consists 阅读全文
posted @ 2023-03-08 10:49 shendawei 阅读(81) 评论(0) 推荐(0)
摘要:1. Linear 2. GCNConv 阅读全文
posted @ 2023-03-04 11:14 shendawei 阅读(8) 评论(0) 推荐(0)
摘要:1. Node Classification 最佳入门: https://mlabonne.github.io/blog/intrognn/ We can enrich our node representation by aggregating its features with those of 阅读全文
posted @ 2023-02-23 11:38 shendawei 阅读(63) 评论(0) 推荐(0)
摘要:1. 解决ubuntu命令行中文显示为问号的问题 https://blog.csdn.net/weixin_33763244/article/details/91472116 阅读全文
posted @ 2020-11-05 16:24 shendawei 阅读(51) 评论(0) 推荐(0)
摘要:1. 交叉熵损失函数: https://www.jianshu.com/p/b07f4cd32ba6 阅读全文
posted @ 2020-10-29 17:42 shendawei 阅读(48) 评论(0) 推荐(0)
摘要:##1. 概率图转binary(bool) def binarize(self, pred): return pred > self.thresh ##2. binary转int、float: img = img.astype(float) 阅读全文
posted @ 2020-10-26 10:00 shendawei 阅读(82) 评论(0) 推荐(0)
摘要:##1. git diff 命令参数:https://www.yiibai.com/git/git_diff.html 输出含义:https://www.ruanyifeng.com/blog/2012/08/how_to_read_diff.html 阅读全文
posted @ 2020-10-22 09:43 shendawei 阅读(75) 评论(0) 推荐(0)
摘要:##1.jit.save()的write failed问题 : 模型保存路径不存在 ##2.FBGEMM doesn't support transpose packing yet! set: quantized.engine = 'qnnpack' and get_default_qconfig( 阅读全文
posted @ 2020-10-21 19:23 shendawei 阅读(797) 评论(0) 推荐(0)
摘要:##1. torchkeras模型打印 https://blog.csdn.net/tszupup/article/details/108129484 阅读全文
posted @ 2020-10-21 11:21 shendawei 阅读(81) 评论(0) 推荐(0)
摘要:1. 大量文本框丢失: ####原因一:opencv 轮廓近似: epsilon值设置偏大,导致不足四个点,进而导致后续必要的处理将其滤掉(此工程后续处理必须大于四个点才能正常进行) ####原因二:short_side过滤: 最短边小于设置的阈值,将会当做干扰滤掉 #2.部分模糊文本框丢失: ## 阅读全文
posted @ 2020-10-20 16:03 shendawei 阅读(448) 评论(0) 推荐(0)
摘要:##1. 显卡驱动重启后丢失: https://blog.csdn.net/hangzuxi8764/article/details/86572093 (the make is related to the gcc version.If make failed, check the gcc vers 阅读全文
posted @ 2020-10-20 09:41 shendawei 阅读(89) 评论(0) 推荐(0)