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

#875 Feature/sg 761 yolo nas

Merged
Ghost merged 1 commits into Deci-AI:master from deci-ai:feature/SG-761-yolo-nas
Conversation 0 Files changed 60 Commits 1
File Compare
super-gradients
/
src
/
super_gradients
/
recipes
/
training_hyperparams
/
supervisely_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 
  1. defaults:
    2. 

  2.   - default_train_params
    3. 

  3. 

  4. max_epochs: 100
    5. 

  5. lr_mode: cosine
    6. 

  6. cosine_final_lr_ratio: 0.01
    7. 

  7. initial_lr: 0.1
    8. 

  8. lr_warmup_epochs: 0
    9. 

  9. multiply_head_lr: 1.
    10. 

  10. 

  11. optimizer: SGD
    12. 

  12. optimizer_params:
    13. 

  13.   momentum: 0.9
    14. 

  14.   weight_decay: 5e-4
    15. 

  15. 

  16. ema: True
    17. 

  17. ema_params:
    18. 

  18.   decay: 0.999
    19. 

  19.   decay_type: exp
    20. 

  20.   beta: 15
    21. 

  21. 

  22. train_metrics_list:
    23. 

  23.   - BinaryIOU
    24. 

  24. 

  25. valid_metrics_list:
    26. 

  26.   - BinaryIOU
    27. 

  27. 

  28. zero_weight_decay_on_bias_and_bn: True
    29. 

  29. average_best_models: True
    30. 

  30. mixed_precision: False
    31. 

  31. 

  32. metric_to_watch: target_IOU
    33. 

  33. greater_metric_to_watch_is_better: True
    34.
Discard
Prev
Next
Tip!

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