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

#284 Fix training prints

Merged
oferbaratz merged 1 commits into Deci-AI:master from deci-ai:hotfix/SG-000-fix_training_prints
Conversation 0 Files changed 2 Commits 1
File Compare
super-gradients
/
src
/
super_gradients
/
recipes
/
training_hyperparams
/
coco2017_yolox_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
45
46
47
48
49
50
51
52
53 
  1. defaults:
    2. 

  2.   - default_train_params
    3. 

  3. 

  4. max_epochs: 300
    5. 

  5. lr_mode: cosine
    6. 

  6. cosine_final_lr_ratio: 0.05
    7. 

  7. lr_warmup_epochs: 5
    8. 

  8. lr_cooldown_epochs: 15
    9. 

  9. warmup_bias_lr: 0.0
    10. 

  10. warmup_momentum: 0.9
    11. 

  11. initial_lr:  0.02
    12. 

  12. warmup_mode: yolov5_warmup
    13. 

  13. batch_accumulate: 1
    14. 

  14. 

  15. save_ckpt_epoch_list: [285]
    16. 

  16. 

  17. loss: yolox_loss
    18. 

  18. 

  19. criterion_params:
    20. 

  20.   strides: [8, 16, 32]  # output strides of all yolo outputs
    21. 

  21.   num_classes: 80
    22. 

  22. 

  23. loss_logging_items_names: [iou, obj, cls, l1, num_fg, Loss]
    24. 

  24. 

  25. optimizer: SGD
    26. 

  26. optimizer_params:
    27. 

  27.   momentum: 0.9
    28. 

  28.   weight_decay: 0.0005
    29. 

  29.   nesterov: True
    30. 

  30. 

  31. ema: True
    32. 

  32. mixed_precision: True
    33. 

  33. 

  34. valid_metrics_list:
    35. 

  35.   - DetectionMetrics:
    36. 

  36.       normalize_targets: True
    37. 

  37.       post_prediction_callback:
    38. 

  38.         _target_: super_gradients.training.models.detection_models.yolov5_base.YoloV5PostPredictionCallback
    39. 

  39.         iou: 0.65
    40. 

  40.         conf: 0.01
    41. 

  41.       num_cls: 80
    42. 

  42. 

  43. pre_prediction_callback: DetectionMultiscalePrePredictionCallback
    44. 

  44. 

  45. phase_callbacks:
    46. 

  46.   - YoloXTrainingStageSwitchCallback:
    47. 

  47.       next_stage_start_epoch: 285
    48. 

  48. 

  49. metric_to_watch: 'mAP@0.50:0.95'
    50. 

  50. greater_metric_to_watch_is_better: True
    51. 

  51. 

  52. 

  53. _convert_: all
    54.
Discard
Prev
Next
Tip!

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