Training errors

[Eric3911]

Create YOLOv3 model with 9 anchors and 5 classes. [<tf.Tensor 'conv2d_7/BiasAdd:0' shape=(?, ?, ?, 30) dtype=float32>, <tf.Tensor 'conv2d_15/BiasAdd:0' shape=(?, ?, ?, 30) dtype=float32>, <tf.Tensor 'conv2d_23/BiasAdd:0' shape=(?, ?, ?, 30) dtype=float32>] Train on 749 samples, val on 0 samples, with batch size 1. Epoch 1/300 749/749 [==============================] - 276s 368ms/step - loss: 930.2716 Epoch 2/300 749/749 [==============================] - 275s 367ms/step - loss: 75.0715 Epoch 3/300 749/749 [==============================] - 274s 366ms/step - loss: 68.3971 Traceback (most recent call last): File "train_Mobilenet.py", line 200, in _main() File "train_Mobilenet.py", line 75, in _main callbacks=[logging, checkpoint]) File "/home/anaconda3/lib/python3.6/site-packages/keras/legacy/interfaces.py", line 91, in wrapper return func(*args, **kwargs) File "/home/anaconda3/lib/python3.6/site-packages/keras/engine/training.py", line 1418, in fit_generator initial_epoch=initial_epoch) File "/home/anaconda3/lib/python3.6/site-packages/keras/engine/training_generator.py", line 251, in fit_generator callbacks.on_epoch_end(epoch, epoch_logs) File "/home/anaconda3/lib/python3.6/site-packages/keras/callbacks.py", line 79, in on_epoch_end callback.on_epoch_end(epoch, logs) File "/home/anaconda3/lib/python3.6/site-packages/keras/callbacks.py", line 429, in on_epoch_end filepath = self.filepath.format(epoch=epoch + 1, **logs) KeyError: 'val_loss'

[CSY-tvgo]

'logs/carMobilenet/001_Mobilenet_finetune/' may not exist, so the program cannot save the checkpoint file in it. mkdir and try again

[Eric3911]

谢谢。

Eric3911

[email protected]

电子名片新出VIP模板啦，快来体验>>

扫一扫,

快速添加名片到手机

---

------------------ 原始邮件 ------------------

发件人: Karbon Chen [email protected]

发送时间: 2019-01-19 14:43:36

收件人:Adamdad/keras-YOLOv3-mobilenet [email protected]

抄送:Eric3911 [email protected],Author [email protected]

主题: Re: [Adamdad/keras-YOLOv3-mobilenet] Training errors (#22)

'logs/carMobilenet/001_Mobilenet_finetune/' may not exist, so the program cannot save the checkpoint file in it. mkdir and try again

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

[Zanderzt]

Hello,I had created this folder,but it reminded the same mistake,had you solved it?