I would like to express my sincere admiration and gratitude for your outstanding work.
I have just run your code using the dataset and pre-trained model you provided, and encountered some issues: the train.py script randomly triggers an error during the training loop, stating ‘NaN or Inf found in input tensor.’
Additionally, when I directly load the pre-trained model and run test.py, the rendered result is merely a completely black silhouette of a human head.
The example dataset I used is "biden."
Could you please advise if this is due to my incorrect usage or some other reason? I look forward to your response, as it would be of great value!
Best regards,
Training Details:
GPU: RTX 3090
CUDA Version: 12.2
Environment: activate by environment.yml
Error: See the picture below
Render with Pre-trained Model: See the picture below (using white background when running test.py)
Dear Author,
I would like to express my sincere admiration and gratitude for your outstanding work. I have just run your code using the dataset and pre-trained model you provided, and encountered some issues: the train.py script randomly triggers an error during the training loop, stating ‘NaN or Inf found in input tensor.’ Additionally, when I directly load the pre-trained model and run test.py, the rendered result is merely a completely black silhouette of a human head. The example dataset I used is "biden." Could you please advise if this is due to my incorrect usage or some other reason? I look forward to your response, as it would be of great value!
Best regards,
Training Details: GPU: RTX 3090
CUDA Version: 12.2 Environment: activate by environment.yml Error: See the picture below
Render with Pre-trained Model: See the picture below (using white background when running test.py)