xisnu / cnn-blstm-ctc Goto Github PK

Getting the following error while training on attached hdf5 file eventhough we used only ascii characters

InvalidArgumentError (see above for traceback): label SparseTensor is not valid
val.hdf5.zip

: indices[3427] = [46,150] is out of bounds: need 0 <= index < [1024,150]
[[Node: CTCLoss = CTCLoss[ctc_merge_repeated=true, ignore_longer_outputs_than_inputs=true, preprocess_collapse_repeated=false, _device="/job:localhost/replica:0/task:0/cpu:0"](transpose/_257, _arg_Placeholder_3_0_3, _arg_Placeholder_2_0_2, _arg_Placeholder_5_0_5)]]
[[Node: gradients/Mean_grad/Shape_1/_269 = _Recvclient_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_2173_gradients/Mean_grad/Shape_1", tensor_type=DT_INT32, _device="/job:localhost/replica:0/task:0/gpu:0"]]

Thanks for sharing the code. Can you please explain how to extract the text from given test images/hdf5 files.