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

Deci-AI
/
super-gradients
connected to https://github.com/Deci-AI/super-gradients.git

Watch 4
Issues 62 Pull Requests 44 Discussions 0

#604 fix master installation

Merged
Ghost merged 1 commits into Deci-AI:master from deci-ai:feature/SG-000_fix_master_inastallation
Conversation 0 Files changed 1 Commits 1
File Compare
super-gradients
/
src
/
super_gradients
/
recipes
/
training_hyperparams
/
cityscapes_default_train_params.yaml
View left file View right file
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
34
35
36
37
38
39
40
41
42
43
44 
  1. defaults:
    2. 

  2.   - default_train_params
    3. 

  3. 

  4. max_epochs: 800
    5. 

  5. lr_mode: poly
    6. 

  6. initial_lr: 0.01   # for effective batch_size=32
    7. 

  7. lr_warmup_epochs: 10
    8. 

  8. multiply_head_lr: 10.
    9. 

  9. 

  10. optimizer: SGD
    11. 

  11. optimizer_params:
    12. 

  12.   momentum: 0.9
    13. 

  13.   weight_decay: 5e-4
    14. 

  14. 

  15. ema: True
    16. 

  16. ema_params:
    17. 

  17.   decay: 0.9999
    18. 

  18.   beta: 15
    19. 

  19.   exp_activation: True
    20. 

  20. 

  21. train_metrics_list:
    22. 

  22.   - PixelAccuracy:
    23. 

  23.       ignore_label: 19
    24. 

  24.   - IoU:
    25. 

  25.       num_classes: 20
    26. 

  26.       ignore_index: 19
    27. 

  27. 

  28. valid_metrics_list:
    29. 

  29.   - PixelAccuracy:
    30. 

  30.       ignore_label: 19
    31. 

  31.   - IoU:
    32. 

  32.       num_classes: 20
    33. 

  33.       ignore_index: 19
    34. 

  34. 

  35. 

  36. 

  37. zero_weight_decay_on_bias_and_bn: True
    38. 

  38. average_best_models: True
    39. 

  39. mixed_precision: False
    40. 

  40. 

  41. metric_to_watch: IoU
    42. 

  42. greater_metric_to_watch_is_better: True
    43. 

  43. 

  44. _convert_: all
    45.
Discard
Prev
Next
Tip!

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