Model load error

[bba20]

I've had this model since earlier this year but when I attempted to load recently I get the following in the display window:

Code: Select all

Exception in Tkinter callback
Traceback (most recent call last):
  File "C:\Users\XX\MiniConda3\envs\faceswap\lib\tkinter\__init__.py", line 1921, in __call__
    return self.func(*args)
  File "C:\Users\XX\MiniConda3\envs\faceswap\lib\tkinter\__init__.py", line 839, in callit
    func(*args)
  File "C:\Users\XX\faceswap\lib\gui\display_analysis.py", line 193, in <lambda>
    self.after(1000, lambda msg=message: self._set_session_summary(msg))
  File "C:\Users\XX\faceswap\lib\gui\display_analysis.py", line 196, in _set_session_summary
    result = self._thread.get_result()
  File "C:\Users\XX\faceswap\lib\gui\utils\misc.py", line 103, in get_result
    raise self.err[1].with_traceback(self.err[2])
  File "C:\Users\XX\faceswap\lib\gui\utils\misc.py", line 72, in run
    retval = self._target(*self._args, **self._kwargs)
  File "C:\Users\XX\faceswap\lib\gui\display_analysis.py", line 215, in _summarise_data
    return session.full_summary
  File "C:\Users\XX\faceswap\lib\gui\analysis\stats.py", line 74, in full_summary
    return self._summary.get_summary_stats()
  File "C:\Users\XX\faceswap\lib\gui\analysis\stats.py", line 313, in get_summary_stats
    self._get_time_stats()
  File "C:\Users\XX\faceswap\lib\gui\analysis\stats.py", line 340, in _get_time_stats
    self._session.get_timestamps(None)).items()}
  File "C:\Users\XX\faceswap\lib\gui\analysis\stats.py", line 232, in get_timestamps
    retval = self._tb_logs.get_timestamps(session_id=session_id)
  File "C:\Users\XX\faceswap\lib\gui\analysis\event_reader.py", line 609, in get_timestamps
    self._check_cache(idx)
  File "C:\Users\XX\faceswap\lib\gui\analysis\event_reader.py", line 548, in _check_cache
    self._cache_data(session_id)
  File "C:\Users\XX\faceswap\lib\gui\analysis\event_reader.py", line 536, in _cache_data
    parser.cache_events(session_id)
  File "C:\Users\XX\faceswap\lib\gui\analysis\event_reader.py", line 699, in cache_events
    self._cache.cache_data(session_id, data, self._loss_labels, is_live=self._live_data)
  File "C:\Users\XX\faceswap\lib\gui\analysis\event_reader.py", line 269, in cache_data
    timestamps, loss = self._to_numpy(data, is_live)
  File "C:\Users\XX\faceswap\lib\gui\analysis\event_reader.py", line 326, in _to_numpy
    n_times, n_loss = (np.array(times, dtype="float64"), np.array(loss, dtype="float32"))
ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (16835,) + inhomogeneous part.

[torzdf]

Can you zip up the model folder and provide me with a link to download.

I need to recreate this error to be able to fix it. Thanks

[bba20]

Sorry I think I jumped the gun. I just clicked train and it began training again fine. It said I couldn't switch from Full Precision to Mixed Precision on older model file. The last time I trained this file I was using my older videocard and my newer one I turned on mixed precision. I thought the model was corrupted at first when I got the error but it just wasn't showing the session data like usual. Edit: Also I started training again with mixed precision on and it didn't revert to Full like the first time. It's just not showing the session data which I dont care about but the model is fine.

Edit 2: I deleted my old session files and it fixed the error. It shows session data again but it just looks like a brand new session.