problem

[lminyuan]

Sorry, I have an error when I run with your original code and original dataset: I ndexError:pop from An empty deque, how can I solve it

[alexanderbarnhill]

Hi I just ran into this as well and I did a small workaround in the generator

while len(self._circ_buf_feat) < self._batch_seq_len * 2 and file_cnt < len(self._filenames_list) in line 115 of the DataGenerator class in the file cls_data_generator.py

This shouldn't be necessary, but I cannot find the root cause right now.

Edit

This delays the problem but it still comes after some point. I will continue to search for the cause and update here if I find anything.

[gabifhk]

Hi I just ran into this as well and I did a small workaround in the generator

while len(self._circ_buf_feat) < self._batch_seq_len * 2 and file_cnt < len(self._filenames_list) in line 115 of the DataGenerator class in the file cls_data_generator.py

This shouldn't be necessary, but I cannot find the root cause right now.

Edit

This delays the problem but it still comes after some point. I will continue to search for the cause and update here if I find anything.

Hello, I've encountered the same issue. Could you please share how you resolved it? I'm wondering if it might be due to incompatibility with the CUDA version. My TensorFlow version is 1.14, and my CUDA version is 10.1.