Register
Login
Resources
Docs Blog Datasets Glossary Case Studies Tutorials & Webinars
Product
Data Engine LLMs Platform Enterprise
Pricing Explore
Connect to our Discord channel

sweep.py 1017 B

You have to be logged in to leave a comment. Sign In
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
  1. ## Creates and runs a wandb sweep
  2. import wandb
  3. config_dict = {
  4. 'program': 'train.py',
  5. 'method': 'random',
  6. 'name': 'miniImagenet Protonet jigsaw sweep - final', # NOTE: varies
  7. 'metric': {'goal': 'maximize', 'name': 'val/acc'},
  8. 'parameters': {
  9. 'lr': {'values': [2e-3, 3e-3, 4e-3, 2e-4, 3e-4, 4e-4]},
  10. 'stop_epoch': {'values': [100]},
  11. 'optimization': {'values': ['Adam']},
  12. 'bn_type': {'values': [1, 2, 3]},
  13. 'train_aug': {'values': [True, False]},
  14. 'jigsaw': {'values': [True]},
  15. 'lbda': {'values': [0.5]},
  16. # constants
  17. 'dataset': {'values': ['miniImagenet']},
  18. 'method': {'values': ['protonet']},
  19. 'model': {'values': ['resnet18']},
  20. 'amp': {'values': [False]},
  21. "run_type": {'values': [0]}
  22. }
  23. }
  24. # TO ADD
  25. # sup_ratio, semi_sup, params.dataset_combine
  26. if __name__ == "__main__":
  27. sweep_id = wandb.sweep(config_dict, entity="meta-learners", project="FSL-SSL")
  28. wandb.agent(sweep_id, count=15)
Tip!

Press p or to see the previous file or, n or to see the next file

Comments

Loading...