Crash during Extraction - Unable to map memory: CL_INVALID_COMMAND_QUEUE

The Extraction process failing on you, and you aren't getting an error back with clear instructions? Tell us about it here


Forum rules

Read the FAQs and search the forum before posting a new topic.

This forum is for reporting errors with the Extraction process. If you want to get tips, or better understand the Extract process, then you should look in the Extract Discussion forum.

Please mark any answers that fixed your problems so others can find the solutions.

Locked
User avatar
Normandy
Posts: 3
Joined: Mon Jan 27, 2020 11:31 am

Crash during Extraction - Unable to map memory: CL_INVALID_COMMAND_QUEUE

Post by Normandy »

I followed all the installation steps but during the extraction this is the result :

Code: Select all

Setting Faceswap backend to AMD
01/27/2020 12:25:51 INFO     Log level set to: INFO
01/27/2020 12:25:51 INFO     Setting up for PlaidML
01/27/2020 12:25:52 INFO     Setting GPU to largest available supported device. If you want to override this selection, run `plaidml-setup` from the command line.
01/27/2020 12:25:52 INFO     Using GPU: ['opencl_nvidia_geforce_mx250.0', 'opencl_nvidia_geforce_mx250.0']
01/27/2020 12:25:52 INFO     Successfully set up for PlaidML
01/27/2020 12:25:54 INFO     Output Directory: C:\Users\Giulia Zaino\faceswap\faces\giulia
01/27/2020 12:25:54 INFO     Loading Detect from S3Fd plugin...
01/27/2020 12:25:54 INFO     Loading Align from Fan plugin...
01/27/2020 12:25:54 INFO     Loading Mask from Extended plugin...
01/27/2020 12:25:54 INFO     Starting, this may take a while...
01/27/2020 12:25:54 INFO     Initializing S3FD (Detect)...
01/27/2020 12:25:55 INFO     Opening device "opencl_nvidia_geforce_mx250.0"
01/27/2020 12:25:55 INFO     Initialized S3FD (Detect) with batchsize of 4
Running pass 1 of 3: Extraction:   0%|                                                                                                                                                     | 0/32 [00:00<?, ?it/s]Unable to map memory: CL_INVALID_COMMAND_QUEUE
01/27/2020 12:26:01 ERROR    Unable to map memory: CL_INVALID_COMMAND_QUEUE
01/27/2020 12:26:05 ERROR    Got Exception on main handler:
Traceback (most recent call last):
  File "C:\Users\Giulia Zaino\faceswap\lib\cli.py", line 128, in execute_script
    process.process()
  File "C:\Users\Giulia Zaino\faceswap\scripts\extract.py", line 114, in process
    self._run_extraction()
  File "C:\Users\Giulia Zaino\faceswap\scripts\extract.py", line 212, in _run_extraction
    for idx, extract_media in enumerate(status_bar):
  File "C:\FinalAnaconda\envs\faceswap\lib\site-packages\tqdm\std.py", line 1104, in __iter__
    for obj in iterable:
  File "C:\Users\Giulia Zaino\faceswap\plugins\extract\pipeline.py", line 211, in detected_faces
    if self._check_and_raise_error():
  File "C:\Users\Giulia Zaino\faceswap\plugins\extract\pipeline.py", line 476, in _check_and_raise_error
    if plugin.check_and_raise_error():
  File "C:\Users\Giulia Zaino\faceswap\plugins\extract\_base.py", line 296, in check_and_raise_error
    err = thread.check_and_raise_error()
  File "C:\Users\Giulia Zaino\faceswap\lib\multithreading.py", line 84, in check_and_raise_error
    raise error[1].with_traceback(error[2])
  File "C:\Users\Giulia Zaino\faceswap\lib\multithreading.py", line 37, in run
    self._target(*self._args, **self._kwargs)
  File "C:\Users\Giulia Zaino\faceswap\plugins\extract\_base.py", line 409, in _thread_process
    batch = function(batch)
  File "C:\Users\Giulia Zaino\faceswap\plugins\extract\detect\_base.py", line 206, in _predict
    batch = self.predict(batch)
  File "C:\Users\Giulia Zaino\faceswap\plugins\extract\detect\s3fd.py", line 50, in predict
    predictions = self.model.predict(batch["feed"])
  File "C:\Users\Giulia Zaino\faceswap\lib\model\session.py", line 63, in predict
    return self._model.predict(feed)
  File "C:\FinalAnaconda\envs\faceswap\lib\site-packages\keras\engine\training.py", line 1169, in predict
    steps=steps)
  File "C:\FinalAnaconda\envs\faceswap\lib\site-packages\keras\engine\training_arrays.py", line 294, in predict_loop
    batch_outs = f(ins_batch)
  File "C:\FinalAnaconda\envs\faceswap\lib\site-packages\plaidml\keras\backend.py", line 177, in __call__
    return [t.as_ndarray(_ctx) for t in tensors]
  File "C:\FinalAnaconda\envs\faceswap\lib\site-packages\plaidml\keras\backend.py", line 177, in <listcomp>
    return [t.as_ndarray(_ctx) for t in tensors]
  File "C:\FinalAnaconda\envs\faceswap\lib\site-packages\plaidml\__init__.py", line 1261, in as_ndarray
    with self.mmap_current() as view:
  File "C:\FinalAnaconda\envs\faceswap\lib\contextlib.py", line 112, in __enter__
    return next(self.gen)
  File "C:\FinalAnaconda\envs\faceswap\lib\site-packages\plaidml\__init__.py", line 1245, in mmap_current
    ctypes.cast(None, _MAP_BUFFER_FUNCTYPE), None)
  File "C:\FinalAnaconda\envs\faceswap\lib\site-packages\plaidml\__init__.py", line 777, in _check_err
    self.raise_last_status()
  File "C:\FinalAnaconda\envs\faceswap\lib\site-packages\plaidml\library.py", line 131, in raise_last_status
    raise self.last_status()
plaidml.exceptions.Unknown: Unable to map memory: CL_INVALID_COMMAND_QUEUE
01/27/2020 12:26:05 CRITICAL An unexpected crash has occurred. Crash report written to 'C:\Users\Giulia Zaino\faceswap\crash_report.2020.01.27.122605668389.log'. You MUST provide this file if seeking assistance. Please verify you are running the latest version of faceswap before reporting
Running pass 1 of 3: Extraction:   0%|

I have the last version

User avatar
torzdf
Posts: 2671
Joined: Fri Jul 12, 2019 12:53 am
Answers: 159
Has thanked: 131 times
Been thanked: 625 times

Re: Crash during Extraction - Unable to map memory: CL_INVALID_COMMAND_QUEUE

Post by torzdf »

First up, you should provide the crash report when reporting issues.

However, I can see the issue here. You have installed the AMD version of Faceswap, but you have an Nvidia card.

Delete the Faceswap folder from your PC and re-run the installer, making sure that you select the "Nvidia" option when prompted, NOT the AMD option.

My word is final

Locked