Skip to content

RedTachyon/autograd-hacks

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 

Repository files navigation

autograd-hacks

Extract useful quantities from PyTorch autograd

Per-example gradients

autograd_hacks.add_hooks(model)
output = model(data)
loss_fn(output, targets).backward()
autograd_hacks.compute_grad1()

# param.grad: gradient averaged over the batch
# param.grad1[i]: gradient with respect to example i

for param in model.parameters():
  assert(torch.allclose(param.grad1.mean(dim=0), param.grad))

Hessians

(assuming ReLU activations, oherwise produces Gauss-Newton matrix)

autograd_hacks.backprop_hess(model(data), hess_type='CrossEntropy')
autograd_hacks.compute_hess(model)
print(param.hess)  # print Hessian of param

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%