search

YCG09 / chinese_ocr

CTPN + DenseNet + CTC based end-to-end Chinese OCR implemented using tensorflow and keras
Apache License 2.0
2.74k stars 1.08k forks source link

训练自己数据集报错：tensorflow.python.framework.errors_impl.InvalidArgumentError: sequence_length(0) <= 159 #299

Open cuimiao187561 opened 4 years ago

cuimiao187561 commented 4 years ago

tensorflow.python.framework.errors_impl.InvalidArgumentError: sequence_length(0) <= 159 [[{{node ctc/CTCLoss}} = CTCLoss[_class=["loc:@training/Adam/gradients/ctc/CTCLoss_grad/mul"], ctc_merge_repeated=true, ignore_longer_outputs_than_inputs=false, preprocess_collapse_repeated=false, _device="/job:localhost/replica:0/task:0/device:CPU:0"](ctc/Log/_2439, ctc/ToInt64/_2441, ctc/ToInt32_2/_2443, ctc/ToInt32_1/_2445)]] [[{{node training/Adam/gradients/ctc/ExpandDims_grad/Shape/_2461}} = _HostRecv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/device:GPU:0", send_device="/job:localhost/replica:0/task:0/device:CPU:0", send_device_incarnation=1, tensor_name="edge_4298_training/Adam/gradients/ctc/ExpandDims_grad/Shape", tensor_type=DT_INT32, _device="/job:localhost/replica:0/task:0/device:GPU:0"]()]]

pingzi5233 commented 4 years ago

请问问题解决了吗？我现在也是遇到这个问题