TypeError: argmax(): argument 'input' (position 1) must be Tensor, not tuple

[SasaniP]

I have seen the same error a couple of times, but the answers did not help me. I am working on a venv with oython 3.10. I have changed my path local due to permission issues and the models have been downloaded, added and gpu.config.yml has been updated accordingly. The following is the full error I get.

{"asctime": "2025-04-22 11:00:24,118", "levelname": "INFO", "message": "Initializing FaceAnalyzer"}
{"asctime": "2025-04-22 11:00:24,118", "levelname": "INFO", "message": "Initializing BaseReader"}
{"asctime": "2025-04-22 11:00:24,249", "levelname": "INFO", "message": "Initializing FaceDetector"}
{"asctime": "2025-04-22 11:00:24,479", "levelname": "INFO", "message": "Initializing FaceUnifier"}
{"asctime": "2025-04-22 11:00:24,488", "levelname": "INFO", "message": "Initializing FacePredictor objects"}
{"asctime": "2025-04-22 11:00:24,488", "levelname": "INFO", "message": "Initializing FacePredictor embed"}
{"asctime": "2025-04-22 11:00:24,637", "levelname": "INFO", "message": "Initializing FacePredictor verify"}
{"asctime": "2025-04-22 11:00:24,801", "levelname": "INFO", "message": "Initializing FacePredictor fer"}
{"asctime": "2025-04-22 11:00:24,981", "levelname": "INFO", "message": "Initializing FacePredictor au"}
{"asctime": "2025-04-22 11:00:25,169", "levelname": "INFO", "message": "Initializing FacePredictor va"}
{"asctime": "2025-04-22 11:00:25,332", "levelname": "INFO", "message": "Initializing FacePredictor deepfake"}
{"asctime": "2025-04-22 11:00:25,490", "levelname": "INFO", "message": "Initializing FacePredictor align"}
{"asctime": "2025-04-22 11:00:25,637", "levelname": "INFO", "message": "Initializing BaseUtilizer objects"}
{"asctime": "2025-04-22 11:00:25,638", "levelname": "INFO", "message": "Initializing BaseUtilizer align"}
{"asctime": "2025-04-22 11:00:25,696", "levelname": "INFO", "message": "Initializing BaseUtilizer draw_boxes"}
{"asctime": "2025-04-22 11:00:25,697", "levelname": "INFO", "message": "Initializing BaseUtilizer draw_landmarks"}
{"asctime": "2025-04-22 11:00:25,778", "levelname": "INFO", "message": "Initializing FaceAnalyzer"}
{"asctime": "2025-04-22 11:00:25,778", "levelname": "INFO", "message": "Initializing BaseReader"}
{"asctime": "2025-04-22 11:00:25,785", "levelname": "INFO", "message": "Initializing FaceDetector"}
{"asctime": "2025-04-22 11:00:25,977", "levelname": "INFO", "message": "Initializing FaceUnifier"}
{"asctime": "2025-04-22 11:00:25,985", "levelname": "INFO", "message": "Initializing FacePredictor objects"}
{"asctime": "2025-04-22 11:00:25,985", "levelname": "INFO", "message": "Initializing FacePredictor embed"}
{"asctime": "2025-04-22 11:00:26,199", "levelname": "INFO", "message": "Initializing FacePredictor verify"}
{"asctime": "2025-04-22 11:00:26,350", "levelname": "INFO", "message": "Initializing FacePredictor fer"}
{"asctime": "2025-04-22 11:00:26,507", "levelname": "INFO", "message": "Initializing FacePredictor au"}
{"asctime": "2025-04-22 11:00:26,671", "levelname": "INFO", "message": "Initializing FacePredictor va"}
{"asctime": "2025-04-22 11:00:26,824", "levelname": "INFO", "message": "Initializing FacePredictor deepfake"}
{"asctime": "2025-04-22 11:00:26,990", "levelname": "INFO", "message": "Initializing FacePredictor align"}
{"asctime": "2025-04-22 11:00:27,136", "levelname": "INFO", "message": "Initializing BaseUtilizer objects"}
{"asctime": "2025-04-22 11:00:27,136", "levelname": "INFO", "message": "Initializing BaseUtilizer align"}
{"asctime": "2025-04-22 11:00:27,179", "levelname": "INFO", "message": "Initializing BaseUtilizer draw_boxes"}
{"asctime": "2025-04-22 11:00:27,180", "levelname": "INFO", "message": "Initializing BaseUtilizer draw_landmarks"}
{"asctime": "2025-04-22 11:00:27,256", "levelname": "INFO", "message": "Running FaceAnalyzer"}
{"asctime": "2025-04-22 11:00:27,256", "levelname": "INFO", "message": "Reading image", "input": "/Users/sasanisulakshi/Downloads/star_cast_reference/Aditi Sudhir Pohankar.jpg"}
{"asctime": "2025-04-22 11:00:27,358", "levelname": "INFO", "message": "Detecting faces"}
{"asctime": "2025-04-22 11:00:28,979", "levelname": "INFO", "message": "Number of faces: 1"}
{"asctime": "2025-04-22 11:00:28,980", "levelname": "INFO", "message": "Unifying faces"}
{"asctime": "2025-04-22 11:00:28,990", "levelname": "INFO", "message": "Predicting facial features"}
{"asctime": "2025-04-22 11:00:28,990", "levelname": "INFO", "message": "Running FacePredictor: embed"}
{"asctime": "2025-04-22 11:00:29,155", "levelname": "INFO", "message": "Running FacePredictor: verify"}
{"asctime": "2025-04-22 11:00:29,297", "levelname": "INFO", "message": "Running FacePredictor: fer"}
Traceback (most recent call last):
  File "/Users/sasanisulakshi/Documents/WORK/Cast_Interaction_Analyzer/main_face_analysis.py", line 35, in <module>
    base_embeds = creating_reference_facial_embeds(reference_image_folder)
  File "/Users/sasanisulakshi/Documents/WORK/Cast_Interaction_Analyzer/function_definition.py", line 335, in creating_reference_facial_embeds
    response = analyzer.run(
  File "/Library/Frameworks/Python.framework/Versions/3.10/lib/python3.10/contextlib.py", line 79, in inner
    return func(*args, **kwds)
  File "/Users/sasanisulakshi/Documents/WORK/venv/CastInteractionAnalyzer/lib/python3.10/site-packages/facetorch/analyzer/core.py", line 170, in run
    data = _predict_batch(data, predictor, predictor_name)
  File "/Users/sasanisulakshi/Documents/WORK/venv/CastInteractionAnalyzer/lib/python3.10/site-packages/facetorch/analyzer/core.py", line 125, in _predict_batch
    preds = predictor.run(face_batch_tensor)
  File "/Library/Frameworks/Python.framework/Versions/3.10/lib/python3.10/contextlib.py", line 79, in inner
    return func(*args, **kwds)
  File "/Users/sasanisulakshi/Documents/WORK/venv/CastInteractionAnalyzer/lib/python3.10/site-packages/facetorch/analyzer/predictor/core.py", line 53, in run
    preds_list = self.postprocessor.run(preds)
  File "/Library/Frameworks/Python.framework/Versions/3.10/lib/python3.10/contextlib.py", line 79, in inner
    return func(*args, **kwds)
  File "/Users/sasanisulakshi/Documents/WORK/venv/CastInteractionAnalyzer/lib/python3.10/site-packages/facetorch/analyzer/predictor/post.py", line 115, in run
    indices = torch.argmax(preds, dim=self.dim).cpu().numpy().tolist()
TypeError: argmax(): argument 'input' (position 1) must be Tensor, not tuple

The following is the log from analyzer initialisation.

{"asctime": "2025-04-22 11:06:56,891", "levelname": "INFO", "message": "Initializing FaceAnalyzer"}
{"asctime": "2025-04-22 11:06:56,891", "levelname": "INFO", "message": "Initializing BaseReader"}
{"asctime": "2025-04-22 11:06:57,024", "levelname": "INFO", "message": "Initializing FaceDetector"}
{"asctime": "2025-04-22 11:06:57,290", "levelname": "INFO", "message": "Initializing FaceUnifier"}
{"asctime": "2025-04-22 11:06:57,300", "levelname": "INFO", "message": "Initializing FacePredictor objects"}
{"asctime": "2025-04-22 11:06:57,300", "levelname": "INFO", "message": "Initializing FacePredictor embed"}
{"asctime": "2025-04-22 11:06:57,457", "levelname": "INFO", "message": "Initializing FacePredictor verify"}
{"asctime": "2025-04-22 11:06:57,616", "levelname": "INFO", "message": "Initializing FacePredictor fer"}
{"asctime": "2025-04-22 11:06:57,783", "levelname": "INFO", "message": "Initializing FacePredictor au"}
{"asctime": "2025-04-22 11:06:57,962", "levelname": "INFO", "message": "Initializing FacePredictor va"}
{"asctime": "2025-04-22 11:06:58,131", "levelname": "INFO", "message": "Initializing FacePredictor deepfake"}
{"asctime": "2025-04-22 11:06:58,312", "levelname": "INFO", "message": "Initializing FacePredictor align"}
{"asctime": "2025-04-22 11:06:58,477", "levelname": "INFO", "message": "Initializing BaseUtilizer objects"}
{"asctime": "2025-04-22 11:06:58,477", "levelname": "INFO", "message": "Initializing BaseUtilizer align"}
{"asctime": "2025-04-22 11:06:58,540", "levelname": "INFO", "message": "Initializing BaseUtilizer draw_boxes"}
{"asctime": "2025-04-22 11:06:58,541", "levelname": "INFO", "message": "Initializing BaseUtilizer draw_landmarks"}

Also, when initialising the analyzer, only model.pt and meta.pt is being downloaded to the updated path locale. Hence the reason for me to download the models and manually add them to the path locale.

Any help and guidance is highly appreciated.