ValueError: betas must be either both floats or both Tensors #548

New Issue

claunia · 2026-01-29T21:48:45Z

claunia commented

2026-01-29 21:48:45 +00:00

Originally created by @hungdaqq on GitHub (Dec 23, 2025).

Traceback (most recent call last):
  File "/kaggle/working/GFPGAN/gfpgan/train.py", line 11, in <module>
    train_pipeline(root_path)
  File "/usr/local/lib/python3.12/dist-packages/basicsr/train.py", line 124, in train_pipeline
    model = build_model(opt)
            ^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/dist-packages/basicsr/models/__init__.py", line 26, in build_model
    model = MODEL_REGISTRY.get(opt['model_type'])(opt)
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/kaggle/working/GFPGAN/gfpgan/models/gfpgan_model.py", line 39, in __init__
    self.init_training_settings()
  File "/kaggle/working/GFPGAN/gfpgan/models/gfpgan_model.py", line 147, in init_training_settings
    self.setup_optimizers()
  File "/kaggle/working/GFPGAN/gfpgan/models/gfpgan_model.py", line 165, in setup_optimizers
    self.optimizer_g = self.get_optimizer(optim_type, optim_params_g, lr, betas=betas)
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/dist-packages/basicsr/models/base_model.py", line 105, in get_optimizer
    optimizer = torch.optim.Adam(params, lr, **kwargs)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/dist-packages/torch/optim/adam.py", line 72, in __init__
    raise ValueError("betas must be either both floats or both Tensors")
ValueError: betas must be either both floats or both Tensors

I’m trying to train GFPGAN on Kaggle, but I keep hitting this error during optimizer initialization.

Originally created by @hungdaqq on GitHub (Dec 23, 2025). ``` Traceback (most recent call last): File "/kaggle/working/GFPGAN/gfpgan/train.py", line 11, in <module> train_pipeline(root_path) File "/usr/local/lib/python3.12/dist-packages/basicsr/train.py", line 124, in train_pipeline model = build_model(opt) ^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/dist-packages/basicsr/models/__init__.py", line 26, in build_model model = MODEL_REGISTRY.get(opt['model_type'])(opt) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/kaggle/working/GFPGAN/gfpgan/models/gfpgan_model.py", line 39, in __init__ self.init_training_settings() File "/kaggle/working/GFPGAN/gfpgan/models/gfpgan_model.py", line 147, in init_training_settings self.setup_optimizers() File "/kaggle/working/GFPGAN/gfpgan/models/gfpgan_model.py", line 165, in setup_optimizers self.optimizer_g = self.get_optimizer(optim_type, optim_params_g, lr, betas=betas) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/dist-packages/basicsr/models/base_model.py", line 105, in get_optimizer optimizer = torch.optim.Adam(params, lr, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/dist-packages/torch/optim/adam.py", line 72, in __init__ raise ValueError("betas must be either both floats or both Tensors") ValueError: betas must be either both floats or both Tensors ``` I’m trying to train GFPGAN on Kaggle, but I keep hitting this error during optimizer initialization.

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: TencentARC/GFPGAN#548