Keras
快乐成长吧
努力努力再努力!
展开
-
解决Keras-Yolo v3 attribute control_flow_ops
装好 keras = 2.2.5 tensorflow-GPU=1.14.0后报错 _, ignore_mask = K.control_flow_ops.while_loop(lambda b,*args: b<m, loop_body, [0, ignore_mask])AttributeError: module 'keras.backend' has no attribute 'control_flow_ops'然后有人建议是keras 太高,降低版本到keras ==2...原创 2020-05-22 01:02:03 · 338 阅读 · 0 评论 -
【keras】CUDNN_STATUS_ALLOC_FAILED 和 failed to run cuBLAS routine cublas ,keras和tensflow-gpu版本的问题
【报错】首先报错:could not create cudnn handle: CUDNN_STATUS_ALLOC_FAILED尝试解决:https://blog.csdn.net/weixin_44754046/article/details/97663626在最开头添加:import tensorflow as tfconfig = tf.ConfigProto(gpu_o...原创 2020-03-01 19:50:55 · 2125 阅读 · 4 评论 -
【keras】加载min_xception前面几个层,把后面的层给改成自己需要层。或者修改后面几层的参数
参考:keras预训练模型应用(4):fine-turn InceptionV3https://www.jianshu.com/p/23295376c44dkeras调用自己训练的模型,并去掉全连接层https://blog.csdn.net/qq_29462849/article/details/83010854from keras.models import load_model...原创 2020-02-29 17:06:49 · 1308 阅读 · 0 评论 -
【Keras】Failed to import pydot. You must install pydot and graphviz for `pydotprint` to work.
参考:https://blog.csdn.net/wushaowu2014/article/details/88958400?depth_1-utm_source=distribute.pc_relevant.none-task&utm_source=distribute.pc_relevant.none-task但是加载到 系统变量中啦。、还是报错如下:则解决办法:...原创 2020-02-29 16:02:36 · 186 阅读 · 0 评论 -
【表情识别】keras报错:Process finished with exit code -1073740791 (0xC0000409)
设置如下:import osos.environ["CUDA_VISIBLE_DEVICES"] = "-1"解决!参考finished with exit code -1073740791 (0xC0000409)解决方案:https://www.cnblogs.com/yh-blog/p/10547826.html .查出是显卡内存不足的问题,解决如下(对个人有效):...原创 2020-01-02 19:30:00 · 694 阅读 · 0 评论 -
【Kares】train_loss /val_loss /train_acc/val_acc可视化
第一种:使用validation_split=0.2H = model.fit(trainX, trainY, batch_size=BS, epochs=EPOCHS, validation_split=0.2, shuffle=True)# plot the training loss and accuracytraining_vis(H)# set the matplot...原创 2019-12-22 23:15:54 · 5924 阅读 · 1 评论