DataLossError (see above for traceback): not an sstable (bad magic number)

DDDoriM1 commented 5 years ago

I can run the same model in windows, but I can't run this in Linux(Ubuntu 16.04).

I got a DataLossError. The model was found in this case. I got confused.

DataLossError (see above for traceback): not an sstable (bad magic number)
     [[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]


# Restore SSD model.
ckpt_filename = '/home/yangziyuan/SSD-Tensorflow/checkpoints/ssd_300_vgg.ckpt'

isess.run(tf.global_variables_initializer())
saver = tf.train.Saver()
saver.restore(isess, ckpt_filename)

Anyone has any clue?

我运行这个相同的模型在windows系统，是可以跑通的，但是在LINUX下没办法跑通会报上述的错误。没有报没找到文件的错误。希望有解决了这个问题的大神可以指导一下。

shanhongxin commented 5 years ago

你好，我也遇到了这个问题。请问你解决了吗？是什么原因？

DDDoriM1 commented 5 years ago

你好，我也遇到了这个问题。请问你解决了吗？是什么原因？

没解决，我放弃了Linux系统，在WINDOWS环境下进行了相关实验。

OceanWong1991 commented 5 years ago

@shanhongxin @DDDoriM1 please refrence #150

balancap / SSD-Tensorflow

DataLossError (see above for traceback): not an sstable (bad magic number) #297