Page 1 of 1
No masks in several columns. Is it normal?
Posted: Tue Jan 05, 2021 3:56 pm
by tokafondo
- _Preview_20210105_154420.jpg (180.86 KiB) Viewed 3416 times
Model: Realface
Already refreshed and updated masks in both Original and Swap sources:faces and frames.
Masks exported successfully as independent files just to check.
Masks shown in Manual Tool.
Is is normal?
Re: No masks in several columns. Is it normal?
Posted: Tue Jan 05, 2021 4:16 pm
by tokafondo
UPDATE: Restarted with Unbalanced model and no masks.
This is shown in LOG screen:
Code: Select all
Exception in Tkinter callback
Traceback (most recent call last):
File "/home/yo/miniconda3/envs/faceswap/lib/python3.8/tkinter/__init__.py", line 1883, in __call__
return self.func(*args)
File "/home/yo/miniconda3/envs/faceswap/lib/python3.8/tkinter/__init__.py", line 804, in callit
func(*args)
File "/home/yo/faceswap/lib/gui/display_graph.py", line 300, in refresh
self.calcs = self.thread.get_result() # Terminate the LongRunningTask object
File "/home/yo/faceswap/lib/gui/utils.py", line 1151, in get_result
raise self.err[1].with_traceback(self.err[2])
File "/home/yo/faceswap/lib/gui/utils.py", line 1122, in run
retval = self._target(*self._args, **self._kwargs)
File "/home/yo/faceswap/lib/gui/stats.py", line 862, in refresh
self._get_raw()
File "/home/yo/faceswap/lib/gui/stats.py", line 925, in _get_raw
loss_dict = Session.get_loss(self._session_id)
File "/home/yo/faceswap/lib/gui/stats.py", line 168, in get_loss
loss_dict = self._tb_logs.get_loss(session_id=session_id, is_training=self._is_training)
File "/home/yo/faceswap/lib/gui/stats.py", line 514, in get_loss
for sess, info in self._from_cache(session_id, is_training).items():
File "/home/yo/faceswap/lib/gui/stats.py", line 482, in _from_cache
self._cache_data(session_id)
File "/home/yo/faceswap/lib/gui/stats.py", line 326, in _cache_data
iterator = tf.compat.v1.io.tf_record_iterator(self._log_filenames[session_id])
KeyError: 11
Re: No masks in several columns. Is it normal?
Posted: Tue Jan 05, 2021 4:37 pm
by tokafondo
tokafondo wrote: ↑Tue Jan 05, 2021 4:16 pmUPDATE: Restarted with Unbalanced model and no masks.
This did not helped at all.
I restarted again with DFL-SAE with the same sources, masks and everything and now the masks are shown in the right columns. No more red squares.
- Preview_20210105_163417.jpg (138.31 KiB) Viewed 3412 times
Re: No masks in several columns. Is it normal?
Posted: Tue Jan 05, 2021 5:02 pm
by tokafondo
tokafondo wrote: ↑Tue Jan 05, 2021 4:37 pmI restarted again with DFL-SAE with the same sources, masks and everything and now the masks are shown in the right columns. No more red squares.
I was not convinced at all, so I restarted again with the same, original settings that produced that "no mask applied" red squares, but from scratch - using the very same faces, alignments and masks.
And now it works.
Realface 128/192, BS 10, ssim, mae, Extended, PLM On.
- Preview_20210105_170033.jpg (101.19 KiB) Viewed 3412 times
Re: No masks in several columns. Is it normal?
Posted: Thu Jan 07, 2021 8:14 pm
by torzdf
Glad it works. Not sure what happened there.
Re: No masks in several columns. Is it normal?
Posted: Wed Jan 13, 2021 2:19 pm
by tokafondo
I'm back on this issue. Cannot tell what the heck in this about. Reinstalled everything (faceswap, miniconda3) from scratch. Redone all extractions, alignments and masks from scratch.
And the issue is still again. I'm thinking about a complete OS reinstall because this is not something usual at all.
Also, the error...
Code: Select all
Exception in Tkinter callback
Traceback (most recent call last):
File "/home/yo/miniconda3/envs/faceswap/lib/python3.8/tkinter/__init__.py", line 1883, in __call__
return self.func(*args)
File "/home/yo/miniconda3/envs/faceswap/lib/python3.8/tkinter/__init__.py", line 804, in callit
func(*args)
File "/home/yo/faceswap/lib/gui/display_graph.py", line 300, in refresh
self.calcs = self.thread.get_result() # Terminate the LongRunningTask object
File "/home/yo/faceswap/lib/gui/utils.py", line 1151, in get_result
raise self.err[1].with_traceback(self.err[2])
File "/home/yo/faceswap/lib/gui/utils.py", line 1122, in run
retval = self._target(*self._args, **self._kwargs)
File "/home/yo/faceswap/lib/gui/stats.py", line 862, in refresh
self._get_raw()
File "/home/yo/faceswap/lib/gui/stats.py", line 925, in _get_raw
loss_dict = Session.get_loss(self._session_id)
File "/home/yo/faceswap/lib/gui/stats.py", line 168, in get_loss
loss_dict = self._tb_logs.get_loss(session_id=session_id, is_training=self._is_training)
File "/home/yo/faceswap/lib/gui/stats.py", line 514, in get_loss
for sess, info in self._from_cache(session_id, is_training).items():
File "/home/yo/faceswap/lib/gui/stats.py", line 482, in _from_cache
self._cache_data(session_id)
File "/home/yo/faceswap/lib/gui/stats.py", line 326, in _cache_data
iterator = tf.compat.v1.io.tf_record_iterator(self._log_filenames[session_id])
KeyError: 3
...still showing.
And after an almost 600K iterations realface model traning (BS 8), the results are not satisfactory at all, even leaving everything as default.
I'm thinking on giving up on faceswap or deepfajes in general.