Jack He issues

Results 9 issues of


                                            Jack He

What's the meaning of the parameter: group ?

Hi, I don't know the meaning of the parameter: group. Can you help me? @Zehaos @modanesh

运行ipython3 app.py 8080最后卡住

你好，我正确配置了GPU的环境，运行ipython3 app.py 8080最后卡住，输出信息如下： Nms engine gpu_nms: ,cython_nms: None ,python_nms: 2019-03-17 15:52:39.285660: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA 2019-03-17 15:52:39.289852:...

Unable to convert .h5 file to .pb file,TypeError: ('Keyword argument not understood:', 'interpolation')

Hi， I train this model and I'm trying to convert "east_model_3T736.h5" to .pb file. I run: python3 keras_to_tensorflow.py --input_model="/media/jack/code/clion_kdevelop/OCR/keras_to_tensorflow/east_model_3T256.h5" --output_model="/media/jack/code/clion_kdevelop/OCR/keras_to_tensorflow/east_model__3T736.pb" and get this error: TypeError: ('Keyword argument not understood:', 'interpolation')...

Save the model as TensorFlow .pb?

Hi, As I want to call this model with C++, and I tried to convert .h5 file to .pb file. But I get TypeError: ('Keyword argument not understood:', 'interpolation') ....

训练时设置load_weights = True，是否会在之前训练的基础上继续训练？

你好：我在读advanced_easy.py代码时，发现设置了load_weights = True之后，有这一句east_network.load_weights(cfg.saved_model_weights_file_path)。但是，当我们设置了当前的train_task_id = '3T384'之后，saved_model_weights_file_path就会变成当前east_model_weights_3T384.h5的模型文件，而不是在之前train_task_id = '3T256'保存的模型east_model_weights_3T256.h5的基础上训练。请问我的理解对吗？它是如何在之前的模型的基础上训练的，怎么加速模型收敛的？

Jack He

What's the meaning of the parameter: group ?

运行ipython3 app.py 8080最后卡住

Unable to convert .h5 file to .pb file,TypeError: ('Keyword argument not understood:', 'interpolation')

Save the model as TensorFlow .pb?

训练时设置load_weights = True，是否会在之前训练的基础上继续训练？

ValueError: invalid literal for int() with base 10: 'head'

TypeError: ('Keyword argument not understood:', 'interpolation')

convert model failed: ValueError: Unknown loss function:quad_loss

dnn_tutorial/pedestrian_demo.cpp的模型文件