search

salaniz / pytorch-gve-lrcn

PyTorch implementations for "Generating Visual Explanations" (GVE) and "Long-term Recurrent Convolutional Networks" (LRCN)
MIT License
92 stars 22 forks source link

Random Seed #16

Open Yonnie1331 opened 3 years ago

Yonnie1331 commented 3 years ago

I used the --torch-seed argument in the parser, but the results still varies with the same seed. I've tried added:

    random.seed(arg_vars["torch_seed"])
    np.random.seed(arg_vars["torch_seed"])
    torch.manual_seed(arg_vars["torch_seed"])
    if torch.cuda.is_available():
        torch.cuda.manual_seed_all(arg_vars["torch_seed"])
    torch.backends.cudnn.deterministic = True

but still not working. I wonder if anyone has an idea how to solve this?