metafile.yml 2.5 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364
  1. Collections:
  2. - Name: CSPNet
  3. Metadata:
  4. Training Data: ImageNet-1k
  5. Architecture:
  6. - Cross Stage Partia Stage
  7. Paper:
  8. URL: https://arxiv.org/abs/1911.11929
  9. Title: 'CSPNet: A New Backbone that can Enhance Learning Capability of CNN'
  10. README: configs/cspnet/README.md
  11. Code:
  12. Version: v0.22.0
  13. URL: https://github.com/open-mmlab/mmpretrain/blob/v0.22.0/mmcls/models/backbones/cspnet.py
  14. Models:
  15. - Name: cspdarknet50_3rdparty_8xb32_in1k
  16. Metadata:
  17. FLOPs: 5040000000
  18. Parameters: 27640000
  19. In Collection: CSPNet
  20. Results:
  21. - Dataset: ImageNet-1k
  22. Metrics:
  23. Top 1 Accuracy: 80.05
  24. Top 5 Accuracy: 95.07
  25. Task: Image Classification
  26. Weights: https://download.openmmlab.com/mmclassification/v0/cspnet/cspdarknet50_3rdparty_8xb32_in1k_20220329-bd275287.pth
  27. Config: configs/cspnet/cspdarknet50_8xb32_in1k.py
  28. Converted From:
  29. Weights: https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/cspdarknet53_ra_256-d05c7c21.pth
  30. Code: https://github.com/rwightman/pytorch-image-models
  31. - Name: cspresnet50_3rdparty_8xb32_in1k
  32. Metadata:
  33. Training Data: ImageNet-1k
  34. FLOPs: 3480000000
  35. Parameters: 21620000
  36. In Collection: CSPNet
  37. Results:
  38. - Dataset: ImageNet-1k
  39. Metrics:
  40. Top 1 Accuracy: 79.55
  41. Top 5 Accuracy: 94.68
  42. Task: Image Classification
  43. Weights: https://download.openmmlab.com/mmclassification/v0/cspnet/cspresnet50_3rdparty_8xb32_in1k_20220329-dd6dddfb.pth
  44. Config: configs/cspnet/cspresnet50_8xb32_in1k.py
  45. Converted From:
  46. Weights: https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/cspresnet50_ra-d3e8d487.pth
  47. Code: https://github.com/rwightman/pytorch-image-models
  48. - Name: cspresnext50_3rdparty_8xb32_in1k
  49. Metadata:
  50. FLOPs: 3110000000
  51. Parameters: 20570000
  52. In Collection: CSPNet
  53. Results:
  54. - Dataset: ImageNet-1k
  55. Metrics:
  56. Top 1 Accuracy: 79.96
  57. Top 5 Accuracy: 94.96
  58. Task: Image Classification
  59. Weights: https://download.openmmlab.com/mmclassification/v0/cspnet/cspresnext50_3rdparty_8xb32_in1k_20220329-2cc84d21.pth
  60. Config: configs/cspnet/cspresnext50_8xb32_in1k.py
  61. Converted From:
  62. Weights: https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/cspresnext50_ra_224-648b4713.pth
  63. Code: https://github.com/rwightman/pytorch-image-models