metafile.yml 5.1 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134
  1. Collections:
  2. - Name: VIG
  3. Metadata:
  4. Training Data: ImageNet-1k
  5. Architecture:
  6. - Vision GNN
  7. Paper:
  8. Title: 'Vision GNN: An Image is Worth Graph of Nodes'
  9. URL: https://arxiv.org/abs/2206.00272
  10. README: configs/vig/README.md
  11. Code:
  12. URL: null
  13. Version: null
  14. Models:
  15. - Name: vig-tiny_3rdparty_in1k
  16. Metadata:
  17. FLOPs: 1309000000
  18. Parameters: 7185000
  19. Training Data: ImageNet-1k
  20. In Collection: VIG
  21. Results:
  22. - Dataset: ImageNet-1k
  23. Metrics:
  24. Top 1 Accuracy: 74.40
  25. Top 5 Accuracy: 92.34
  26. Task: Image Classification
  27. Weights: https://download.openmmlab.com/mmclassification/v0/vig/vig-tiny_3rdparty_in1k_20230117-6414c684.pth
  28. Config: configs/vig/vig-tiny_8xb128_in1k.py
  29. Converted From:
  30. Weights: https://github.com/huawei-noah/Efficient-AI-Backbones/releases/download/vig/vig_ti_74.5.pth
  31. Code: https://github.com/huawei-noah/Efficient-AI-Backbones/tree/master/vig_pytorch
  32. - Name: vig-small_3rdparty_in1k
  33. Metadata:
  34. FLOPs: 4535000000
  35. Parameters: 22748000
  36. Training Data: ImageNet-1k
  37. In Collection: VIG
  38. Results:
  39. - Dataset: ImageNet-1k
  40. Metrics:
  41. Top 1 Accuracy: 80.61
  42. Top 5 Accuracy: 95.28
  43. Task: Image Classification
  44. Weights: https://download.openmmlab.com/mmclassification/v0/vig/vig-small_3rdparty_in1k_20230117-5338bf3b.pth
  45. Config: configs/vig/vig-small_8xb128_in1k.py
  46. Converted From:
  47. Weights: https://github.com/huawei-noah/Efficient-AI-Backbones/releases/download/vig/vig_s_80.6.pth
  48. Code: https://github.com/huawei-noah/Efficient-AI-Backbones/tree/master/vig_pytorch
  49. - Name: vig-base_3rdparty_in1k
  50. Metadata:
  51. FLOPs: 17681000000
  52. Parameters: 20685000
  53. Training Data: ImageNet-1k
  54. In Collection: VIG
  55. Results:
  56. - Dataset: ImageNet-1k
  57. Metrics:
  58. Top 1 Accuracy: 82.62
  59. Top 5 Accuracy: 96.04
  60. Task: Image Classification
  61. Weights: https://download.openmmlab.com/mmclassification/v0/vig/vig-base_3rdparty_in1k_20230117-92f6f12f.pth
  62. Config: configs/vig/vig-base_8xb128_in1k.py
  63. Converted From:
  64. Weights: https://github.com/huawei-noah/Efficient-AI-Backbones/releases/download/vig/vig_b_82.6.pth
  65. Code: https://github.com/huawei-noah/Efficient-AI-Backbones/tree/master/vig_pytorch
  66. - Name: pvig-tiny_3rdparty_in1k
  67. Metadata:
  68. FLOPs: 1714000000
  69. Parameters: 9458000
  70. Training Data: ImageNet-1k
  71. In Collection: VIG
  72. Results:
  73. - Dataset: ImageNet-1k
  74. Metrics:
  75. Top 1 Accuracy: 78.38
  76. Top 5 Accuracy: 94.38
  77. Task: Image Classification
  78. Weights: https://download.openmmlab.com/mmclassification/v0/vig/pvig-tiny_3rdparty_in1k_20230117-eb77347d.pth
  79. Config: configs/vig/pvig-tiny_8xb128_in1k.py
  80. Converted From:
  81. Weights: https://github.com/huawei-noah/Efficient-AI-Backbones/releases/download/pyramid-vig/pvig_ti_78.5.pth.tar
  82. Code: https://github.com/huawei-noah/Efficient-AI-Backbones/tree/master/vig_pytorch
  83. - Name: pvig-small_3rdparty_in1k
  84. Metadata:
  85. FLOPs: 4572000000
  86. Parameters: 29024000
  87. Training Data: ImageNet-1k
  88. In Collection: VIG
  89. Results:
  90. - Dataset: ImageNet-1k
  91. Metrics:
  92. Top 1 Accuracy: 82.00
  93. Top 5 Accuracy: 95.97
  94. Task: Image Classification
  95. Weights: https://download.openmmlab.com/mmclassification/v0/vig/pvig-small_3rdparty_in1k_20230117-9433dc96.pth
  96. Config: configs/vig/pvig-small_8xb128_in1k.py
  97. Converted From:
  98. Weights: https://github.com/huawei-noah/Efficient-AI-Backbones/releases/download/pyramid-vig/pvig_s_82.1.pth.tar
  99. Code: https://github.com/huawei-noah/Efficient-AI-Backbones/tree/master/vig_pytorch
  100. - Name: pvig-medium_3rdparty_in1k
  101. Metadata:
  102. FLOPs: 8886000000
  103. Parameters: 51682000
  104. Training Data: ImageNet-1k
  105. In Collection: VIG
  106. Results:
  107. - Dataset: ImageNet-1k
  108. Metrics:
  109. Top 1 Accuracy: 83.12
  110. Top 5 Accuracy: 96.35
  111. Task: Image Classification
  112. Weights: https://download.openmmlab.com/mmclassification/v0/vig/pvig-medium_3rdparty_in1k_20230117-21057a6d.pth
  113. Config: configs/vig/pvig-medium_8xb128_in1k.py
  114. Converted From:
  115. Weights: https://github.com/huawei-noah/Efficient-AI-Backbones/releases/download/pyramid-vig/pvig_m_83.1.pth.tar
  116. Code: https://github.com/huawei-noah/Efficient-AI-Backbones/tree/master/vig_pytorch
  117. - Name: pvig-base_3rdparty_in1k
  118. Metadata:
  119. FLOPs: 16861000000
  120. Parameters: 95213000
  121. Training Data: ImageNet-1k
  122. In Collection: VIG
  123. Results:
  124. - Dataset: ImageNet-1k
  125. Metrics:
  126. Top 1 Accuracy: 83.59
  127. Top 5 Accuracy: 96.52
  128. Task: Image Classification
  129. Weights: https://download.openmmlab.com/mmclassification/v0/vig/pvig-base_3rdparty_in1k_20230117-dbab3c85.pth
  130. Config: configs/vig/pvig-base_8xb128_in1k.py
  131. Converted From:
  132. Weights: https://github.com/huawei-noah/Efficient-AI-Backbones/releases/download/pyramid-vig/pvig_b_83.66.pth.tar
  133. Code: https://github.com/huawei-noah/Efficient-AI-Backbones/tree/master/vig_pytorch