Python torch.nn.functional Module

This page shows the popular functions and classes defined in the torch.nn.functional module. The items are ordered by their popularity in 40,000 open source Python projects. If you can not find a good example below, you can try the search function to search modules.

• 1. relu()    
  Used in 1343 projects
• 2. softmax()    
  Used in 1104 projects
• 3. log_softmax()    
  Used in 643 projects
• 4. cross_entropy()    
  Used in 491 projects
• 5. avg_pool2d()    
  Used in 470 projects
• 6. dropout()    
  Used in 455 projects
• 7. max_pool2d()    
  Used in 410 projects
• 8. sigmoid()    
  Used in 381 projects
• 9. pad()    
  Used in 360 projects
• 10. interpolate()    
  Used in 341 projects
• 11. conv2d()    
  Used in 299 projects
• 12. nll_loss()    
  Used in 266 projects
• 13. mse_loss()    
  Used in 242 projects
• 14. tanh()    
  Used in 230 projects
• 15. linear()    
  Used in 213 projects
• 16. binary_cross_entropy_with_logits()    
  Used in 210 projects
• 17. upsample()    
  Used in 200 projects
• 18. normalize()    
  Used in 183 projects
• 19. adaptive_avg_pool2d()    
  Used in 158 projects
• 20. smooth_l1_loss()    
  Used in 136 projects
• 21. leaky_relu()    
  Used in 124 projects
• 22. grid_sample()    
  Used in 119 projects
• 23. binary_cross_entropy()    
  Used in 117 projects
• 24. softplus()    
  Used in 108 projects
• 25. batch_norm()    
  Used in 98 projects
• 26. max_pool1d()    
  Used in 89 projects
• 27. elu()    
  Used in 84 projects
• 28. l1_loss()    
  Used in 62 projects
• 29. conv_transpose2d()    
  Used in 60 projects
• 30. affine_grid()    
  Used in 59 projects
• 31. kl_div()    
  Used in 56 projects
• 32. conv1d()    
  Used in 48 projects
• 33. logsigmoid()    
  Used in 48 projects
• 34. adaptive_max_pool2d()    
  Used in 47 projects
• 35. cosine_similarity()    
  Used in 39 projects
• 36. relu_()    
  Used in 38 projects
• 37. embedding()    
  Used in 38 projects
• 38. one_hot()    
  Used in 36 projects
• 39. avg_pool3d()    
  Used in 33 projects
• 40. dropout2d()    
  Used in 32 projects
• 41. relu6()    
  Used in 32 projects
• 42. pairwise_distance()    
  Used in 27 projects
• 43. conv3d()    
  Used in 22 projects
• 44. hardtanh()    
  Used in 20 projects
• 45. max_pool3d()    
  Used in 18 projects
• 46. conv_transpose1d()    
  Used in 18 projects
• 47. unfold()    
  Used in 18 projects
• 48. avg_pool1d()    
  Used in 16 projects
• 49. upsample_bilinear()    
  Used in 13 projects
• 50. threshold()    
  Used in 13 projects
• 51. selu()    
  Used in 13 projects
• 52. glu()    
  Used in 13 projects
• 53. gumbel_softmax()    
  Used in 12 projects
• 54. max_unpool2d()    
  Used in 11 projects
• 55. leaky_relu_()    
  Used in 11 projects
• 56. pixel_shuffle()    
  Used in 10 projects
• 57. instance_norm()    
  Used in 9 projects
• 58. multilabel_soft_margin_loss()    
  Used in 9 projects
• 59. lp_pool2d()    
  Used in 9 projects
• 60. layer_norm()    
  Used in 8 projects
• 61. margin_ranking_loss()    
  Used in 8 projects
• 62. adaptive_avg_pool3d()    
  Used in 8 projects
• 63. bilinear()    
  Used in 8 projects
• 64. conv_transpose3d()    
  Used in 7 projects
• 65. group_norm()    
  Used in 7 projects
• 66. adaptive_avg_pool1d()    
  Used in 7 projects
• 67. prelu()    
  Used in 7 projects
• 68. fold()    
  Used in 6 projects
• 69. multi_margin_loss()    
  Used in 6 projects
• 70. upsample_nearest()    
  Used in 5 projects
• 71. gelu()    
  Used in 5 projects
• 72. soft_margin_loss()    
  Used in 5 projects
• 73. adaptive_max_pool1d()    
  Used in 5 projects