Photo AI fails to process the attached image. The first time it reported an “out of memory” error despite the fact that my PC has 24GB RAM. The second time I got a “model error” error.
Steps to reproduce issue:
- Load attached image
- Select face recovery, sharpening and noise removal, all at default settings
- Attempt to save file
2024-01-04-12-22-4.tzlog (352.0 KB)
2024-01-04-12-14-27.tzlog (228.8 KB)
Topaz Photo AI [v2.2.1] on [Windows]
I tested your image on my M1 Mac 16GB Sonoma 14.3. No issues. So I guess it’s something to do with your specs.
Unlikely to be specs - I have an i7 machine with NVidia graphics and 24GB RAM. More likely to be a bug in Topaz. Can someone from Topaz please check the logs? Thanks
@andrew3-7835 I had a look at the log files.
I see this one quite a few times:
[2024-01-04 12:17:17.936, 4.24 ms] [3fd8] Info | [AIE] WARN CUDA lazy loading is not enabled. Enabling it can significantly reduce device memory usage. See
CUDA_MODULE_LOADING in CUDA C++ Programming Guide
I assume you don’t develop NVDIA related things but maybe you problem is graphic card driver related.
Please have a look and consider an update if necessary.
It might help.
Hi. Thanks for looking at the logs.
I’m not convinced my drivers are the problem:
I updated the drivers to the latest version and I’m still having the same problems
The “error running model” also occurs with CPU explicitly selected as processor, which should completely bypass the GPU.
Any other thoughts?
I’ve reviewed your logs, and it seems that you’re missing several model files for download. This issue might have been prevented during the download process.
To address this, disable any antivirus, firewall, VPN, or proxy software that might be active on your computer. These tools are sometimes known to interfere with the installer’s ability to download the model files.
Then rerun the installer. You can download a new installer from the downloads page directly :
Topaz Labs Download Page
If you are still experiencing problems after going through these steps please reach out to us at the following link:
Topaz Labs | Support Page