junedgar

@chongyangtao

@tirtile thank for your share. Can you give me advice to train your NHWC model? I tried to train the faceboxes to detect bottle. I trained the model by TropComplique's...

@hahasofia I used the wrong model. I ran the 'save.py' script, the model.pb was saved in root directory. I used a model in subdirectory with the some name, so i...

@hahasofia yes, I encountered this problem, you can solve it by using python3.

I solved the "type_traits" problem by change the "extra_compile_args" to "extra_compile_args = ['-std=c++11', '-stdlib=libc++']". But when I run the "import warpctc_tensorflow" with the error below: ``` File "", line 1,...

@chituma110 I had not explored the code completely. But i found the code initialized the Session without any limitations for GPU, so i thought that the code may use the...

mac os 10.14 Solving the '__pyx_t_5numpy_uint64_t *' (aka 'unsigned long *') to 'uint64_t *' problem, following: 1、change the ` __pyx_t_5numpy_uint64_t` to `__pyx_t_5numpy_uint32_t` in line 2022 of funciton solve_nl_lmp_cpp of multicut.cpp....

@ak-7 @snetgrom you can solve the problem by #[link](https://github.com/eldar/pose-tensorflow/issues/26#issuecomment-441566414)

训练样本是128000，测试样本是10000，采用默认参数，300个Epoch后还是不收敛，而且acc一直是0.224。Epoch 300/10000, accuracy = 0.224,avg_train_cost = 5.262, lastbatch_err = 0.295, time = 583.210,lr=0.00000000。按照 @ilovin 的方法调整了学习率到1e-3后，Epoch 45/10000, accuracy = 0.036,avg_train_cost = 5.974, lastbatch_err = 0.320, time = 268.441,lr=0.00000002，好像没有在10个epoch后有明显变化。调整lr到1e-3，然后调整decay_rete到0.9，几十个epoch后还是一样没啥明显变化

已经解决不收敛问题，因为warpctc无法绑定到tensorflow1.1，所以采用standord ctc，发现里面的学习率衰减的太快，修改一下，就可以啦