facebookresearch/xformers

How to set random seeds fixed

Open

#618 opened on Jan 3, 2023

View on GitHub
 (17 comments) (1 reaction) (0 assignees)Python (473 forks)batch import
good first issue

Repository metrics

Stars
 (6,903 stars)
PR merge metrics
 (Avg merge 3d 6h) (1 merged PR in 30d)

Description

❓ Questions and Help

Different results occur when I run the same code twice. And the set_seed func run before all.

def set_seed(seed):
    random.seed(seed)  # Python random module.
    np.random.seed(seed)  # Numpy module.
    set_seed(seed)
    os.environ['PYTHONHASHSEED'] = str(seed)
    torch.random.manual_seed(seed)
    torch.manual_seed(seed)
    torch.cuda.manual_seed(seed)
    torch.cuda.manual_seed_all(seed)  # if you are using multi-GPU.
    torch.backends.cudnn.benchmark = cudnn_benchmark
    torch.backends.cudnn.deterministic = cudnn_deterministic

Contributor guide

Research direction
Explain the proper way to set random seeds for reproducibility in PyTorch, including the need to set seeds for all libraries (random, numpy, torch) and potential issues like CUDA non determinism. Provide a corrected seed setting function and mention that set seed should avoid recursion by removing the call within itself.
Tech stack
pythonpytorch
Domain
machine learningai
Issue type
Research
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
1
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Under 1 hour
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
PythonPyTorch
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
70