fresh install + new rtx 2060: pynvml.NVMLError_NotFound: Not Found

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
pronto
Posts: 1
Joined: Tue Feb 11, 2020 9:10 pm
Has thanked: 1 time

fresh install + new rtx 2060: pynvml.NVMLError_NotFound: Not Found

Post by pronto »

HI all,

Just installed a rtx 2060, and downloaded FS, tried my first extract and got a critical crash:

02/11/2020 16:04:22 MainProcess MainThread gpu_stats get_active_devices DEBUG Active GPU Devices: [0, 1]
....
raise NVMLError(ret)
pynvml.NVMLError_NotFound: Not Found

Code: Select all

02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'mask.unet_dfl', info: 'UNET_DFL options. Mask designed to provide smart segmentation of mostly frontal faces.\nThe mask model has been trained by community members. Insert more commentary on testing here. Profile faces may result in sub-par performance.')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'mask.unet_dfl', title: 'batch-size', datatype: '<class 'int'>', default: '8', info: 'The batch size to use. To a point, higher batch sizes equal better performance, but setting it too high can harm performance.\n\n	Nvidia users: If the batchsize is set higher than the your GPU can accomodate then this will automatically be lowered.', rounding: '1', min_max: (1, 64), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: mask.unet_dfl
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Adding defaults: (filename: vgg_clear_defaults.py, module_path: plugins.extract.mask, plugin_type: mask
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Importing defaults module: plugins.extract.mask.vgg_clear_defaults
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'mask.vgg_clear', info: 'VGG_Clear options. Mask designed to provide smart segmentation of mostly frontal faces clear of obstructions.\nProfile faces and obstructions may result in sub-par performance.')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'mask.vgg_clear', title: 'batch-size', datatype: '<class 'int'>', default: '6', info: 'The batch size to use. To a point, higher batch sizes equal better performance, but setting it too high can harm performance.\n\n	Nvidia users: If the batchsize is set higher than the your GPU can accomodate then this will automatically be lowered.', rounding: '1', min_max: (1, 64), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: mask.vgg_clear
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Adding defaults: (filename: vgg_obstructed_defaults.py, module_path: plugins.extract.mask, plugin_type: mask
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Importing defaults module: plugins.extract.mask.vgg_obstructed_defaults
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'mask.vgg_obstructed', info: 'VGG_Obstructed options. Mask designed to provide smart segmentation of mostly frontal faces.\nThe mask model has been specifically trained to recognize some facial obstructions (hands and eyeglasses). Profile faces may result in sub-par performance.')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'mask.vgg_obstructed', title: 'batch-size', datatype: '<class 'int'>', default: '2', info: 'The batch size to use. To a point, higher batch sizes equal better performance, but setting it too high can harm performance.\n\n	Nvidia users: If the batchsize is set higher than the your GPU can accomodate then this will automatically be lowered.', rounding: '1', min_max: (1, 64), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: mask.vgg_obstructed
02/11/2020 16:04:22 MainProcess     MainThread      config          handle_config             DEBUG    Handling config
02/11/2020 16:04:22 MainProcess     MainThread      config          check_exists              DEBUG    Config file exists: 'C:\Users\james\faceswap\config\extract.ini'
02/11/2020 16:04:22 MainProcess     MainThread      config          load_config               VERBOSE  Loading config: 'C:\Users\james\faceswap\config\extract.ini'
02/11/2020 16:04:22 MainProcess     MainThread      config          validate_config           DEBUG    Validating config
02/11/2020 16:04:22 MainProcess     MainThread      config          check_config_change       DEBUG    Default config has not changed
02/11/2020 16:04:22 MainProcess     MainThread      config          check_config_choices      DEBUG    Checking config choices
02/11/2020 16:04:22 MainProcess     MainThread      config          check_config_choices      DEBUG    Checked config choices
02/11/2020 16:04:22 MainProcess     MainThread      config          validate_config           DEBUG    Validated config
02/11/2020 16:04:22 MainProcess     MainThread      config          handle_config             DEBUG    Handled config
02/11/2020 16:04:22 MainProcess     MainThread      config          __init__                  DEBUG    Initialized: Config
02/11/2020 16:04:22 MainProcess     MainThread      config          get                       DEBUG    Getting config item: (section: 'global', option: 'allow_growth')
02/11/2020 16:04:22 MainProcess     MainThread      config          get                       DEBUG    Returning item: (type: <class 'bool'>, value: False)
02/11/2020 16:04:22 MainProcess     MainThread      config          get                       DEBUG    Getting config item: (section: 'align.fan', option: 'batch-size')
02/11/2020 16:04:22 MainProcess     MainThread      config          get                       DEBUG    Returning item: (type: <class 'int'>, value: 12)
02/11/2020 16:04:22 MainProcess     MainThread      utils           get                       DEBUG    Model exists: C:\Users\james\faceswap\plugins\extract\align\.cache\face-alignment-network_2d4_keras_v1.h5
02/11/2020 16:04:22 MainProcess     MainThread      _base           __init__                  DEBUG    Initialized _base Align
02/11/2020 16:04:22 MainProcess     MainThread      _base           __init__                  DEBUG    Initialized Align
02/11/2020 16:04:22 MainProcess     MainThread      pipeline        _load_mask                DEBUG    Loading Masker: 'extended'
02/11/2020 16:04:22 MainProcess     MainThread      plugin_loader   _import                   INFO     Loading Mask from Extended plugin...
02/11/2020 16:04:22 MainProcess     MainThread      _base           __init__                  DEBUG    Initializing Mask: (configfile: None, )
02/11/2020 16:04:22 MainProcess     MainThread      _base           __init__                  DEBUG    Initializing Mask: (git_model_id: None, model_filename: None,  configfile: None)
02/11/2020 16:04:22 MainProcess     MainThread      config          __init__                  DEBUG    Initializing: Config
02/11/2020 16:04:22 MainProcess     MainThread      config          get_config_file           DEBUG    Config File location: 'C:\Users\james\faceswap\config\extract.ini'
02/11/2020 16:04:22 MainProcess     MainThread      _config         set_defaults              DEBUG    Setting defaults
02/11/2020 16:04:22 MainProcess     MainThread      _config         set_globals               DEBUG    Setting global config
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'global', info: 'Options that apply to all extraction plugins')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'global', title: 'allow_growth', datatype: '<class 'bool'>', default: 'False', info: '[Nvidia Only]. Enable the Tensorflow GPU `allow_growth` configuration option. This option prevents Tensorflow from allocating all of the GPU VRAM at launch but can lead to higher VRAM fragmentation and slower performance. Should only be enabled if you are having problems running extraction.', rounding: 'None', min_max: None, choices: None, gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Adding defaults: (filename: fan_defaults.py, module_path: plugins.extract.align, plugin_type: align
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Importing defaults module: plugins.extract.align.fan_defaults
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'align.fan', info: 'FAN Aligner options.Fast on GPU, slow on CPU. Best aligner.')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'align.fan', title: 'batch-size', datatype: '<class 'int'>', default: '12', info: 'The batch size to use. To a point, higher batch sizes equal better performance, but setting it too high can harm performance.\n\n	Nvidia users: If the batchsize is set higher than the your GPU can accomodate then this will automatically be lowered.\n	AMD users: A batchsize of 8 requires about 4 GB vram.', rounding: '1', min_max: (1, 64), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: align.fan
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Adding defaults: (filename: cv2_dnn_defaults.py, module_path: plugins.extract.detect, plugin_type: detect
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Importing defaults module: plugins.extract.detect.cv2_dnn_defaults
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'detect.cv2_dnn', info: 'CV2 DNN Detector options.\nA CPU only extractor, is the least reliable, but uses least resources and runs fast on CPU. Use this if not using a GPU and time is important')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'detect.cv2_dnn', title: 'confidence', datatype: '<class 'int'>', default: '50', info: 'The confidence level at which the detector has succesfully found a face.\nHigher levels will be more discriminating, lower levels will have more false positives.', rounding: '5', min_max: (25, 100), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: detect.cv2_dnn
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Adding defaults: (filename: mtcnn_defaults.py, module_path: plugins.extract.detect, plugin_type: detect
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Importing defaults module: plugins.extract.detect.mtcnn_defaults
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'detect.mtcnn', info: 'MTCNN Detector options.\nFast on GPU, slow on CPU. Uses fewer resources than other GPU detectors but can often return more false positives.')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'detect.mtcnn', title: 'minsize', datatype: '<class 'int'>', default: '20', info: 'The minimum size of a face (in pixels) to be accepted as a positive match.\nLower values use significantly more VRAM and will detect more false positives.', rounding: '10', min_max: (20, 1000), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'detect.mtcnn', title: 'threshold_1', datatype: '<class 'float'>', default: '0.6', info: 'First stage threshold for face detection. This stage obtains face candidates.', rounding: '2', min_max: (0.1, 0.9), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'detect.mtcnn', title: 'threshold_2', datatype: '<class 'float'>', default: '0.7', info: 'Second stage threshold for face detection. This stage refines face candidates.', rounding: '2', min_max: (0.1, 0.9), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'detect.mtcnn', title: 'threshold_3', datatype: '<class 'float'>', default: '0.7', info: 'Third stage threshold for face detection. This stage further refines face candidates.', rounding: '2', min_max: (0.1, 0.9), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'detect.mtcnn', title: 'scalefactor', datatype: '<class 'float'>', default: '0.709', info: 'The scale factor for the image pyramid.', rounding: '3', min_max: (0.1, 0.9), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'detect.mtcnn', title: 'batch-size', datatype: '<class 'int'>', default: '8', info: 'The batch size to use. To a point, higher batch sizes equal better performance, but setting it too high can harm performance.\n\n	Nvidia users: If the batchsize is set higher than the your GPU can accomodate then this will automatically be lowered.', rounding: '1', min_max: (1, 64), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: detect.mtcnn
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Adding defaults: (filename: s3fd_defaults.py, module_path: plugins.extract.detect, plugin_type: detect
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Importing defaults module: plugins.extract.detect.s3fd_defaults
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'detect.s3fd', info: 'S3FD Detector options.Fast on GPU, slow on CPU. Can detect more faces and fewer false positives than other GPU detectors, but is a lot more resource intensive.')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'detect.s3fd', title: 'confidence', datatype: '<class 'int'>', default: '70', info: 'The confidence level at which the detector has succesfully found a face.\nHigher levels will be more discriminating, lower levels will have more false positives.', rounding: '5', min_max: (25, 100), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'detect.s3fd', title: 'batch-size', datatype: '<class 'int'>', default: '4', info: 'The batch size to use. To a point, higher batch sizes equal better performance, but setting it too high can harm performance.\n\n	Nvidia users: If the batchsize is set higher than the your GPU can accomodate then this will automatically be lowered.\n	AMD users: A batchsize of 8 requires about 2 GB vram.', rounding: '1', min_max: (1, 64), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: detect.s3fd
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Adding defaults: (filename: unet_dfl_defaults.py, module_path: plugins.extract.mask, plugin_type: mask
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Importing defaults module: plugins.extract.mask.unet_dfl_defaults
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'mask.unet_dfl', info: 'UNET_DFL options. Mask designed to provide smart segmentation of mostly frontal faces.\nThe mask model has been trained by community members. Insert more commentary on testing here. Profile faces may result in sub-par performance.')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'mask.unet_dfl', title: 'batch-size', datatype: '<class 'int'>', default: '8', info: 'The batch size to use. To a point, higher batch sizes equal better performance, but setting it too high can harm performance.\n\n	Nvidia users: If the batchsize is set higher than the your GPU can accomodate then this will automatically be lowered.', rounding: '1', min_max: (1, 64), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: mask.unet_dfl
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Adding defaults: (filename: vgg_clear_defaults.py, module_path: plugins.extract.mask, plugin_type: mask
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Importing defaults module: plugins.extract.mask.vgg_clear_defaults
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'mask.vgg_clear', info: 'VGG_Clear options. Mask designed to provide smart segmentation of mostly frontal faces clear of obstructions.\nProfile faces and obstructions may result in sub-par performance.')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'mask.vgg_clear', title: 'batch-size', datatype: '<class 'int'>', default: '6', info: 'The batch size to use. To a point, higher batch sizes equal better performance, but setting it too high can harm performance.\n\n	Nvidia users: If the batchsize is set higher than the your GPU can accomodate then this will automatically be lowered.', rounding: '1', min_max: (1, 64), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: mask.vgg_clear
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Adding defaults: (filename: vgg_obstructed_defaults.py, module_path: plugins.extract.mask, plugin_type: mask
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Importing defaults module: plugins.extract.mask.vgg_obstructed_defaults
02/11/2020 16:04:22 MainProcess     MainThread      config          add_section               DEBUG    Add section: (title: 'mask.vgg_obstructed', info: 'VGG_Obstructed options. Mask designed to provide smart segmentation of mostly frontal faces.\nThe mask model has been specifically trained to recognize some facial obstructions (hands and eyeglasses). Profile faces may result in sub-par performance.')
02/11/2020 16:04:22 MainProcess     MainThread      config          add_item                  DEBUG    Add item: (section: 'mask.vgg_obstructed', title: 'batch-size', datatype: '<class 'int'>', default: '2', info: 'The batch size to use. To a point, higher batch sizes equal better performance, but setting it too high can harm performance.\n\n	Nvidia users: If the batchsize is set higher than the your GPU can accomodate then this will automatically be lowered.', rounding: '1', min_max: (1, 64), choices: [], gui_radio: False, fixed: True, group: None)
02/11/2020 16:04:22 MainProcess     MainThread      _config         load_module               DEBUG    Added defaults: mask.vgg_obstructed
02/11/2020 16:04:22 MainProcess     MainThread      config          handle_config             DEBUG    Handling config
02/11/2020 16:04:22 MainProcess     MainThread      config          check_exists              DEBUG    Config file exists: 'C:\Users\james\faceswap\config\extract.ini'
02/11/2020 16:04:22 MainProcess     MainThread      config          load_config               VERBOSE  Loading config: 'C:\Users\james\faceswap\config\extract.ini'
02/11/2020 16:04:22 MainProcess     MainThread      config          validate_config           DEBUG    Validating config
02/11/2020 16:04:22 MainProcess     MainThread      config          check_config_change       DEBUG    Default config has not changed
02/11/2020 16:04:22 MainProcess     MainThread      config          check_config_choices      DEBUG    Checking config choices
02/11/2020 16:04:22 MainProcess     MainThread      config          check_config_choices      DEBUG    Checked config choices
02/11/2020 16:04:22 MainProcess     MainThread      config          validate_config           DEBUG    Validated config
02/11/2020 16:04:22 MainProcess     MainThread      config          handle_config             DEBUG    Handled config
02/11/2020 16:04:22 MainProcess     MainThread      config          __init__                  DEBUG    Initialized: Config
02/11/2020 16:04:22 MainProcess     MainThread      config          get                       DEBUG    Getting config item: (section: 'global', option: 'allow_growth')
02/11/2020 16:04:22 MainProcess     MainThread      config          get                       DEBUG    Returning item: (type: <class 'bool'>, value: False)
02/11/2020 16:04:22 MainProcess     MainThread      _base           _get_model                DEBUG    No model_filename specified. Returning None
02/11/2020 16:04:22 MainProcess     MainThread      _base           __init__                  DEBUG    Initialized _base Mask
02/11/2020 16:04:22 MainProcess     MainThread      _base           __init__                  DEBUG    Initialized Mask
02/11/2020 16:04:22 MainProcess     MainThread      gpu_stats       __init__                  DEBUG    Initializing GPUStats
02/11/2020 16:04:22 MainProcess     MainThread      gpu_stats       initialize                DEBUG    OS is not macOS. Using pynvml
02/11/2020 16:04:22 MainProcess     MainThread      gpu_stats       get_device_count          DEBUG    GPU Device count: 2
02/11/2020 16:04:22 MainProcess     MainThread      gpu_stats       get_active_devices        DEBUG    Active GPU Devices: [0, 1]
Traceback (most recent call last):
  File "C:\Users\james\faceswap\lib\cli.py", line 127, in execute_script
    process = script(arguments)
  File "C:\Users\james\faceswap\scripts\extract.py", line 61, in __init__
    normalize_method=normalization)
  File "C:\Users\james\faceswap\plugins\extract\pipeline.py", line 85, in __init__
    self._is_parallel = self._set_parallel_processing(multiprocess)
  File "C:\Users\james\faceswap\plugins\extract\pipeline.py", line 339, in _set_parallel_processing
    gpu_stats = GPUStats()
  File "C:\Users\james\faceswap\lib\gpu_stats.py", line 45, in __init__
    self.initialize(log)
  File "C:\Users\james\faceswap\lib\gpu_stats.py", line 114, in initialize
    self.get_handles()
  File "C:\Users\james\faceswap\lib\gpu_stats.py", line 161, in get_handles
    for i in range(self.device_count)]
  File "C:\Users\james\faceswap\lib\gpu_stats.py", line 161, in <listcomp>
    for i in range(self.device_count)]
  File "C:\Users\james\MiniConda3\envs\faceswap\lib\site-packages\pynvml.py", line 814, in nvmlDeviceGetHandleByIndex
    _nvmlCheckReturn(ret)
  File "C:\Users\james\MiniConda3\envs\faceswap\lib\site-packages\pynvml.py", line 310, in _nvmlCheckReturn
    raise NVMLError(ret)
pynvml.NVMLError_NotFound: Not Found
Exception occured trying to retrieve sysinfo: Not Found

help!

User avatar
torzdf
Posts: 2649
Joined: Fri Jul 12, 2019 12:53 am
Answers: 159
Has thanked: 128 times
Been thanked: 623 times

Re: fresh install + new rtx 2060: pynvml.NVMLError_NotFound: Not Found

Post by torzdf »

Your PC thinks you have 2 GPUs attached and it can't find one of them. You should check the output of nvidia-smi

My word is final

Locked