OSError: unknown file format #149

kilik128 · 2022-01-23T12:27:37Z

got this

/content/DAIN
revise the unique id to a random numer 67519
Namespace(SAVED_MODEL=None, alpha=[0.0, 1.0], arg='./model_weights/67519-Sun-Jan-23-12-26/args.txt', batch_size=1, channels=3, ctx_lr_coe=1.0, datasetName='Vimeo_90K_interp', datasetPath='', dataset_split=97, debug=False, depth_lr_coe=0.001, doubleframe_distorted_pixel_ratio=0.002, doubleframe_distortion_threshold=20, dtype=<class 'torch.cuda.FloatTensor'>, end_frame=36, epsilon=1e-06, factor=0.2, filter_lr_coe=1.0, filter_size=4, flow_lr_coe=0.01, force=False, frame_input_dir='/content/DAIN/input_frames', frame_output_dir='/content/DAIN/output_frames', log='./model_weights/67519-Sun-Jan-23-12-26/log.txt', lr=0.002, mixer='normal', netName='DAIN_slowmotion', no_date=False, numEpoch=100, occ_lr_coe=1.0, patience=5, rectify_lr=0.001, resize_hotfix=False, save_path='./model_weights/67519-Sun-Jan-23-12-26', save_which=1, seed=1, start_frame=1, time_step=0.11988011988011989, uid=None, use_cuda=True, use_cudnn=1, weight_decay=0, workers=8)
cudnn is used
Interpolate 7 frames
Traceback (most recent call last):
File "colab_interpolate.py", line 64, in
big_font = ImageFont.truetype(urlopen(font_url), size=24)
File "/usr/local/lib/python3.7/dist-packages/PIL/ImageFont.py", line 648, in truetype
return freetype(font)
File "/usr/local/lib/python3.7/dist-packages/PIL/ImageFont.py", line 645, in freetype
return FreeTypeFont(font, size, index, encoding, layout_engine)
File "/usr/local/lib/python3.7/dist-packages/PIL/ImageFont.py", line 197, in init
load_from_bytes(font)
File "/usr/local/lib/python3.7/dist-packages/PIL/ImageFont.py", line 179, in load_from_bytes
"", size, index, encoding, self.font_bytes, layout_engine
OSError: unknown file format

on !python -W ignore colab_interpolate.py --netName DAIN_slowmotion --time_step {fps/TARGET_FPS} --start_frame 1 --end_frame {pngs_generated_count} --frame_input_dir '{FRAME_INPUT_DIR}' --frame_output_dir '{FRAME_OUTPUT_DIR}'

AlphaGit · 2022-01-23T13:43:53Z

Hi! Please verify that the files are accessible in the right directory and that, if using a networked source like GDrive, the connection is active.

kilik128 · 2022-01-23T19:27:14Z

for say true we got this

]
1 s
# Clone DAIN sources
# TODO Bring back original repo
!git clone https://github.com/baowenbo/DAIN /content/DAIN
%cd /content
!rm -rf DAIN
!git clone -b colab-compatibility --depth 1 https://github.com/AlphaGit/DAIN /content/DAIN
%cd /content/DAIN
!git log -1
fatal: destination path '/content/DAIN' already exists and is not an empty directory.
/content
Cloning into '/content/DAIN'...
warning: Could not find remote branch colab-compatibility to clone.
fatal: Remote branch colab-compatibility not found in upstream origin
[Errno 2] No such file or directory: '/content/DAIN'
/content
fatal: not a git repository (or any of the parent directories): .git

on this line

# Clone DAIN sources
# TODO Bring back original repo
!git clone https://github.com/baowenbo/DAIN /content/DAIN
%cd /content
!rm -rf DAIN
!git clone -b colab-compatibility --depth 1 https://github.com/AlphaGit/DAIN /content/DAIN
%cd /content/DAIN
!git log -1

i try this one too
https://colab.research.google.com/drive/1pLMrFFAZzbS4CEzpmwsGXkKNBuPlHPDo#scrollTo=mwYCvA4EZn8O

AlphaGit · 2022-01-23T20:59:48Z

This line:

warning: Could not find remote branch colab-compatibility to clone.

Makes me think you might be using an outdated version.

Try this one: https://colab.research.google.com/github/AlphaGit/DAIN/blob/master/Colab_DAIN.ipynb

I have included several updates and fixes in it, which should match the version from here: https://github.com/AlphaGit/DAIN

Let me know if it helps!

kilik128 · 2022-01-24T08:43:20Z

look finish on same error at first
have must add !pip install opencv-python
and this line look not found # import ipyplot
maybie beaucause that's
have change this line too
%shell yes | cp -f /content/gdrive/MyDrive/{INPUT_FILEPATH} /content/DAIN/

WARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv

about front import

`

Interpolation

%shell mkdir -p '{FRAME_OUTPUT_DIR}'
%cd /content/DAIN

!python -W ignore colab_interpolate.py --netName DAIN_slowmotion --time_step {fps/TARGET_FPS} --start_frame 1 --end_frame {frame_count} --frame_input_dir '{FRAME_INPUT_DIR}' --frame_output_dir '{FRAME_OUTPUT_DIR}' --mixer={FRAME_MIXER} --resize_hotfix {RESIZE_HOTFIX}
/content/DAIN
revise the unique id to a random numer 40885
Namespace(SAVED_MODEL=None, alpha=[0.0, 1.0], arg='./model_weights/40885-Mon-Jan-24-08-26/args.txt', batch_size=1, channels=3, ctx_lr_coe=1.0, datasetName='Vimeo_90K_interp', datasetPath='', dataset_split=97, debug=False, depth_lr_coe=0.001, doubleframe_distorted_pixel_ratio=0.002, doubleframe_distortion_threshold=20, dtype=<class 'torch.cuda.FloatTensor'>, end_frame=36, epsilon=1e-06, factor=0.2, filter_lr_coe=1.0, filter_size=4, flow_lr_coe=0.01, force=False, frame_input_dir='/content/DAIN/input_frames', frame_output_dir='/content/DAIN/output_frames', log='./model_weights/40885-Mon-Jan-24-08-26/log.txt', lr=0.002, mixer='normal', netName='DAIN_slowmotion', no_date=False, numEpoch=100, occ_lr_coe=1.0, patience=5, rectify_lr=0.001, resize_hotfix=True, save_path='./model_weights/40885-Mon-Jan-24-08-26', save_which=1, seed=1, start_frame=1, time_step=0.1998001998001998, uid=None, use_cuda=True, use_cudnn=1, weight_decay=0, workers=8)
cudnn is used
Interpolate 4 frames
Traceback (most recent call last):
File "colab_interpolate.py", line 64, in
big_font = ImageFont.truetype(urlopen(font_url), size=24)
File "/usr/local/lib/python3.7/site-packages/PIL/ImageFont.py", line 855, in truetype
return freetype(font)
File "/usr/local/lib/python3.7/site-packages/PIL/ImageFont.py", line 852, in freetype
return FreeTypeFont(font, size, index, encoding, layout_engine)
File "/usr/local/lib/python3.7/site-packages/PIL/ImageFont.py", line 215, in init
load_from_bytes(font)
File "/usr/local/lib/python3.7/site-packages/PIL/ImageFont.py", line 197, in load_from_bytes
"", size, index, encoding, self.font_bytes, layout_engine
OSError: unknown file format`

so i have change the colab interpotale for import manualy

big_font = ImageFont.truetype("/content/DAIN/Roboto-Bold.ttf"), size=24)
smol_font = ImageFont.truetype("/content/DAIN/Roboto-Bold.ttf"), size=14)

/content/DAIN revise the unique id to a random numer 142 Namespace(SAVED_MODEL=None, alpha=[0.0, 1.0], arg='./model_weights/142-Mon-Jan-24-08-42/args.txt', batch_size=1, channels=3, ctx_lr_coe=1.0, datasetName='Vimeo_90K_interp', datasetPath='', dataset_split=97, debug=False, depth_lr_coe=0.001, doubleframe_distorted_pixel_ratio=0.002, doubleframe_distortion_threshold=20, dtype=<class 'torch.cuda.FloatTensor'>, end_frame=36, epsilon=1e-06, factor=0.2, filter_lr_coe=1.0, filter_size=4, flow_lr_coe=0.01, force=False, frame_input_dir='/content/DAIN/input_frames', frame_output_dir='/content/DAIN/output_frames', log='./model_weights/142-Mon-Jan-24-08-42/log.txt', lr=0.002, mixer='normal', netName='DAIN_slowmotion', no_date=False, numEpoch=100, occ_lr_coe=1.0, patience=5, rectify_lr=0.001, resize_hotfix=True, save_path='./model_weights/142-Mon-Jan-24-08-42', save_which=1, seed=1, start_frame=1, time_step=0.1998001998001998, uid=None, use_cuda=True, use_cudnn=1, weight_decay=0, workers=8) cudnn is used Interpolate 4 frames 3% 1/35 [00:00<00:00, 38.69it/s] Traceback (most recent call last): File "colab_interpolate.py", line 395, in <module> interpolate_mixer(input_frame) File "colab_interpolate.py", line 213, in normal_interpolate interpolate(input_frame, input_frame+1) File "colab_interpolate.py", line 205, in interpolate for frame, time_offset in zip(subframes(image_1, image_2), time_offsets): File "colab_interpolate.py", line 125, in subframes assert channels == 3 AssertionError

AlphaGit · 2022-01-25T03:07:42Z

Hi! Sorry, I don't fully understand how this part is related:

big_font = ImageFont.truetype("/content/DAIN/Roboto-Bold.ttf"), size=24)
smol_font = ImageFont.truetype("/content/DAIN/Roboto-Bold.ttf"), size=14)

Are you manually importing fonts? What are you doing with them?

kilik128 · 2022-01-25T14:13:44Z

yes manually importing fonts for try fix the File "colab_interpolate.py", line 64

AlphaGit · 2022-01-26T02:48:53Z

I see -- those fonts are only used to write over the image and debug intermediate interpolation results.

Feel free to comment out:

those three lines
the full debug_text_on_image method (lines 232 to 244)

And replaces the contents of the the d debug method (lines 339 to 342) with the word "pass". This would take out all that interaction and still allow the rest of the notebook to run.

kilik128 · 2022-01-26T09:41:03Z

got one frame result and assertionError

/content/DAIN
revise the unique id to a random numer 65014
Namespace(SAVED_MODEL=None, alpha=[0.0, 1.0], arg='./model_weights/65014-Wed-Jan-26-09-36/args.txt', batch_size=1, channels=3, ctx_lr_coe=1.0, datasetName='Vimeo_90K_interp', datasetPath='', dataset_split=97, debug=False, depth_lr_coe=0.001, doubleframe_distorted_pixel_ratio=0.002, doubleframe_distortion_threshold=20, dtype=<class 'torch.cuda.FloatTensor'>, end_frame=36, epsilon=1e-06, factor=0.2, filter_lr_coe=1.0, filter_size=4, flow_lr_coe=0.01, force=False, frame_input_dir='/content/DAIN/input_frames', frame_output_dir='/content/DAIN/output_frames', log='./model_weights/65014-Wed-Jan-26-09-36/log.txt', lr=0.002, mixer='normal', netName='DAIN_slowmotion', no_date=False, numEpoch=100, occ_lr_coe=1.0, patience=5, rectify_lr=0.001, resize_hotfix=True, save_path='./model_weights/65014-Wed-Jan-26-09-36', save_which=1, seed=1, start_frame=1, time_step=0.1998001998001998, uid=None, use_cuda=True, use_cudnn=1, weight_decay=0, workers=8)
cudnn is used
Interpolate 4 frames
3% 1/35 [00:00<00:00, 41.87it/s]
Traceback (most recent call last):
File "colab_interpolate.py", line 397, in
interpolate_mixer(input_frame)
File "colab_interpolate.py", line 213, in normal_interpolate
interpolate(input_frame, input_frame+1)
File "colab_interpolate.py", line 205, in interpolate
for frame, time_offset in zip(subframes(image_1, image_2), time_offsets):
File "colab_interpolate.py", line 125, in subframes
assert channels == 3
AssertionError

have try and retry

AlphaGit · 2022-01-26T12:17:24Z

Good, progress! It seems that your image files have 4 channels instead of 3. DAIN can only work with 3. Maybe they have a transparency layer? Make sure that the part of the code that removes the alpha layer is running.

…

On Wed, Jan 26, 2022 at 4:41 AM kilik128 ***@***.***> wrote: got one frame result and assertionError /content/DAIN revise the unique id to a random numer 65014 Namespace(SAVED_MODEL=None, alpha=[0.0, 1.0], arg='./model_weights/65014-Wed-Jan-26-09-36/args.txt', batch_size=1, channels=3, ctx_lr_coe=1.0, datasetName='Vimeo_90K_interp', datasetPath='', dataset_split=97, debug=False, depth_lr_coe=0.001, doubleframe_distorted_pixel_ratio=0.002, doubleframe_distortion_threshold=20, dtype=<class 'torch.cuda.FloatTensor'>, end_frame=36, epsilon=1e-06, factor=0.2, filter_lr_coe=1.0, filter_size=4, flow_lr_coe=0.01, force=False, frame_input_dir='/content/DAIN/input_frames', frame_output_dir='/content/DAIN/output_frames', log='./model_weights/65014-Wed-Jan-26-09-36/log.txt', lr=0.002, mixer='normal', netName='DAIN_slowmotion', no_date=False, numEpoch=100, occ_lr_coe=1.0, patience=5, rectify_lr=0.001, resize_hotfix=True, save_path='./model_weights/65014-Wed-Jan-26-09-36', save_which=1, seed=1, start_frame=1, time_step=0.1998001998001998, uid=None, use_cuda=True, use_cudnn=1, weight_decay=0, workers=8) cudnn is used Interpolate 4 frames 3% 1/35 [00:00<00:00, 41.87it/s] Traceback (most recent call last): File "colab_interpolate.py", line 397, in interpolate_mixer(input_frame) File "colab_interpolate.py", line 213, in normal_interpolate interpolate(input_frame, input_frame+1) File "colab_interpolate.py", line 205, in interpolate for frame, time_offset in zip(subframes(image_1, image_2), time_offsets): File "colab_interpolate.py", line 125, in subframes assert channels == 3 AssertionError have try and retry — Reply to this email directly, view it on GitHub <#149 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAIP2WMFTOCMTPEU34C74GTUX66TVANCNFSM5MTJB6CA> . You are receiving this because you commented.Message ID: ***@***.***>

kilik128 · 2022-01-28T19:35:44Z

Done Thank's ! it's realy cool :)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

OSError: unknown file format #149

OSError: unknown file format #149

kilik128 commented Jan 23, 2022

AlphaGit commented Jan 23, 2022

kilik128 commented Jan 23, 2022

AlphaGit commented Jan 23, 2022

kilik128 commented Jan 24, 2022

AlphaGit commented Jan 25, 2022

kilik128 commented Jan 25, 2022

AlphaGit commented Jan 26, 2022

kilik128 commented Jan 26, 2022

AlphaGit commented Jan 26, 2022 via email

kilik128 commented Jan 28, 2022

OSError: unknown file format #149

OSError: unknown file format #149

Comments

kilik128 commented Jan 23, 2022

AlphaGit commented Jan 23, 2022

kilik128 commented Jan 23, 2022

AlphaGit commented Jan 23, 2022

kilik128 commented Jan 24, 2022

Interpolation

AlphaGit commented Jan 25, 2022

kilik128 commented Jan 25, 2022

AlphaGit commented Jan 26, 2022

kilik128 commented Jan 26, 2022

AlphaGit commented Jan 26, 2022 via email

kilik128 commented Jan 28, 2022