-
Hi,
How are you?
Thanks for the update!
I think this line may cause trouble:
learn_gen.**fit_one_cycle**(1, pct_start=pct_start, max_lr=slice(5e-8,5e-5))
if the **"batch size"/bs**…
-
Hello,
I'm using the VideoColorizer.ipynb and I get this error :
AttributeError: module 'ffmpeg' has no attribute 'input'
ffmpeg-python is installed. Win10 64 bits
Can you help me ?
-
Dear Jantic,
could you please give more information about to install your software.
I am missing the anaconda revision and the jupyter revision.
thanks + greetings
George
-
Hey, @jantic! Congrats for such an excellent tool. I'm using it through DeepAI and I wonder how hard is to add the `render_factor` as a param for API implementation? I could do it my self if you let m…
-
Hello!
when running ColorizeTrainingVideo.ipynb i get the error FileNotFoundError: [Errno 2] No such file or directory: 'data/bandw/models/VideoModel_gen_0.pth'
it creates a temp file there but fi…
-
Hi, I've got to run the prediction in CPU I will make a pull request from my fork :)
-
[Conflict reports deoldify.txt](https://github.com/jantic/DeOldify/files/4309958/Conflict.reports.deoldify.txt)
I have attached the conflicts it come up with, it seems I am not the only one with this…
-
This is not an issue with DeOldify, it's a question about using your models.
I'm doing some volunteer archival work for my local Historical Society, and had an idea the other day when I learned abo…
q5sys updated
4 years ago
-
If you set the render_factor too high you will end up with "CUDA out of memory", nothing abnormal here.
The problem starts when you lower the render_factor, you still get "CUDA out of memory" becau…
-
Hello!
while trying to train with ColorizeTrainingVideo.ipynb at the command
```
data_gen = get_data(bs=bs, sz=sz, keep_pct=keep_pct)
```
it throws the following warning although i have fold…