data must be floating-point

[kilanny]

Trying to build from scratch, I get the error "data must be floating-point" after completing 1000 steps while saving checkpoint. Details

In line 116 in train.py

          log('Saving audio and alignment...')
          input_seq, spectrogram, alignment = sess.run([
          model.inputs[0], model.linear_outputs[0], model.alignments[0]])
          waveform = audio.inv_spectrogram(spectrogram.T)
          audio.save_wav(waveform, os.path.join(log_dir, 'step-%d-audio.wav' % step)) #here 

Step 1000    [2.913 sec/step, loss=0.16122, avg_loss=0.15765]
Writing summary at step: 1000
Saving checkpoint to: F:/src/mimic2/tacotron\logs-tacotron\model.ckpt-1000
Saving audio and alignment...
Exiting due to exception: data must be floating-point
Traceback (most recent call last):
  File "train.py", line 116, in train
    audio.save_wav(waveform, os.path.join(log_dir, 'step-%d-audio.wav' % step))
  File "F:\src\mimic2\util\audio.py", line 15, in save_wav
    librosa.output.write_wav(path, wav.astype(np.int16), hparams.sample_rate)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\librosa\output.py", line 223, in write_wav
    util.valid_audio(y, mono=False)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\librosa\util\utils.py", line 159, in valid_audio
    raise ParameterError('data must be floating-point')
librosa.util.exceptions.ParameterError: data must be floating-point
2018-12-07 22:29:06.482540: W T:\src\github\tensorflow\tensorflow\core\kernels\queue_base.cc:277] _0_datafeeder/input_queue: Skipping cancelled enqueue attempt with queue not closed
Traceback (most recent call last):
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\client\session.py", line 1278, in _do_call
    return fn(*args)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\client\session.py", line 1263, in _run_fn
    options, feed_dict, fetch_list, target_list, run_metadata)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\client\session.py", line 1350, in _call_tf_sessionrun
    run_metadata)
tensorflow.python.framework.errors_impl.CancelledError: Enqueue operation was cancelled
         [[Node: datafeeder/input_queue_enqueue = QueueEnqueueV2[Tcomponents=[DT_INT32, DT_INT32, DT_FLOAT, DT_FLOAT], timeout_ms=-1, _device="/job:localhost/replica:0/task:0/device:CPU:0"](datafeeder/input_queue, _arg_datafeeder/inputs_0_1, _arg_datafeeder/input_lengths_0_0, _arg_datafeeder/mel_targets_0_3, _arg_datafeeder/linear_targets_0_2)]]

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "F:\src\mimic2\datasets\datafeeder.py", line 74, in run
    self._enqueue_next_group()
  File "F:\src\mimic2\datasets\datafeeder.py", line 96, in _enqueue_next_group
    self._session.run(self._enqueue_op, feed_dict=feed_dict)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\client\session.py", line 877, in run
    run_metadata_ptr)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\client\session.py", line 1100, in _run
    feed_dict_tensor, options, run_metadata)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\client\session.py", line 1272, in _do_run
    run_metadata)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\client\session.py", line 1291, in _do_call
    raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.CancelledError: Enqueue operation was cancelled
         [[Node: datafeeder/input_queue_enqueue = QueueEnqueueV2[Tcomponents=[DT_INT32, DT_INT32, DT_FLOAT, DT_FLOAT], timeout_ms=-1, _device="/job:localhost/replica:0/task:0/device:CPU:0"](datafeeder/input_queue, _arg_datafeeder/inputs_0_1, _arg_datafeeder/input_lengths_0_0, _arg_datafeeder/mel_targets_0_3, _arg_datafeeder/linear_targets_0_2)]]

Caused by op 'datafeeder/input_queue_enqueue', defined at:
  File "train.py", line 158, in <module>
    main()
  File "train.py", line 154, in main
    train(log_dir, args)
  File "train.py", line 58, in train
    feeder = DataFeeder(coord, input_path, hparams)
  File "F:\src\mimic2\datasets\datafeeder.py", line 45, in __init__
    self._enqueue_op = queue.enqueue(self._placeholders)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\ops\data_flow_ops.py", line 339, in enqueue
    self._queue_ref, vals, name=scope)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\ops\gen_data_flow_ops.py", line 4373, in queue_enqueue_v2
    timeout_ms=timeout_ms, name=name)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\framework\op_def_library.py", line 787, in _apply_op_helper
    op_def=op_def)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\util\deprecation.py", line 454, in new_func
    return func(*args, **kwargs)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\framework\ops.py", line 3155, in create_op
    op_def=op_def)
  File "C:\Program Files (x86)\Microsoft Visual Studio\Shared\Anaconda3_64\envs\tensorflow\lib\site-packages\tensorflow\python\framework\ops.py", line 1717, in __init__
    self._traceback = tf_stack.extract_stack()

CancelledError (see above for traceback): Enqueue operation was cancelled
         [[Node: datafeeder/input_queue_enqueue = QueueEnqueueV2[Tcomponents=[DT_INT32, DT_INT32, DT_FLOAT, DT_FLOAT], timeout_ms=-1, _device="/job:localhost/replica:0/task:0/device:CPU:0"](datafeeder/input_queue, _arg_datafeeder/inputs_0_1, _arg_datafeeder/input_lengths_0_0, _arg_datafeeder/mel_targets_0_3, _arg_datafeeder/linear_targets_0_2)]]

[kilanny]

Solved by replacing line 15 in audio.py to

librosa.output.write_wav(path, wav, hparams.sample_rate)

[Ittiz]

Nice wish I saw this before I started, wasted an hour of my time.