Torch.manual_seed(3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision · HackerPapers discussions

Torch.manual_seed(3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision — mentioned by danielhanchen

Tue, 28 Jan 2025 10:41:53 GMT

Hey! :) Coincidentally the seeds I always use are 3407, 3408 and 3409 :) 3407 because of https://arxiv.org/abs/2109.08203 I also tried not setting the seeds, but the results are still the same - quantizing all layers seems to make the model forget and repeat everything - I put all examples here: https://docs.unsloth.ai/basics/deepseek-r1-dynamic-1.58-bit#...

Paper and discussion history. A mention is not an endorsement.

Torch.manual_seed(3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision — HN submission

Thu, 24 Oct 2024 09:14:35 GMT

Torch.manual_seed(3407) is all you need: On the influence of random seeds

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Torch.manual_seed(3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision — mentioned by mcyc

Fri, 26 Jul 2024 18:51:36 GMT

You may enjoy this article: https://arxiv.org/abs/2109.08203 The author treats the seed as a hyperparameter and searches for the one that performs best for training a CV model.

Paper and discussion history. A mention is not an endorsement.

Torch.manual_seed(3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision — mentioned by godelski

Wed, 31 May 2023 20:36:49 GMT

This is not always true. ML has reproducibility issues despite the status quo being open models. Even if we don't include proprietary datasets there are still issues. Compute is one issue, but we can even ignore that[0]. The Lottery Ticket[1,2] plays a big role, in that you can just get lucky. This really should have resulted in treating benchmarks as weaker indicators but see other comments about reviewing incentives. Another issue is that it is status quo to optimize hyperparameters on test data and this results in information leakage. While this won't affect results of running a checkpoint…

Paper and discussion history. A mention is not an endorsement.

Torch.manual_seed(3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision — HN submission

Wed, 17 May 2023 13:56:36 GMT

The influence of random seeds in deep learning architectures for computer vision

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Torch.manual_seed(3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision — HN submission

Tue, 16 May 2023 16:49:58 GMT

The influence of random seeds in deep learning architectures for computer vision

3 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.