Using hook functions to call on a custom function to modify a layer given parameters of another layer
|
|
1
|
34
|
April 8, 2024
|
Trying to understand the gradient for softmax (without CrossEntropyLoss)
|
|
6
|
139
|
April 8, 2024
|
Modifying a Tensor with requires_grad=True in PyTorch - Maintaining Connection for Backpropagation
|
|
2
|
73
|
April 8, 2024
|
Remove Computation in Forward & Backward Pass
|
|
0
|
45
|
April 8, 2024
|
Non differentiability of complete QR decomposition for case where rows > columns
|
|
5
|
93
|
April 7, 2024
|
Can the new functional autograd take batches? Also, is it more efficient to compute a hessian with the new functional autograd than it is using the old autograd?
|
|
17
|
1979
|
April 7, 2024
|
Weighted Multi-label Focal Loss Implementation
|
|
0
|
52
|
April 6, 2024
|
Focal loss for imbalanced multi class classification in Pytorch
|
|
14
|
21669
|
April 6, 2024
|
RuntimeError: tensor does not have a device - which tensor?
|
|
3
|
69
|
April 5, 2024
|
Training two models simultaniously: incorrect backpropagation
|
|
1
|
80
|
April 4, 2024
|
RuntimeError: derivative for aten::_scaled_dot_product is not implemented
|
|
0
|
43
|
April 4, 2024
|
Autograd derivatives of multioutput ANN
|
|
2
|
85
|
April 4, 2024
|
Can PyTorch move a tensor along with its computational graph from GPU to CPU, and then move it back to GPU for backpropagation?
|
|
2
|
69
|
April 4, 2024
|
Unexpected error when performing backpropagation : "RuntimeError: self must be a matrix"
|
|
4
|
78
|
April 3, 2024
|
More efficient norm of gradient computations using vmap
|
|
0
|
44
|
April 3, 2024
|
Loss doesn't change PINN implementation of RLC equation
|
|
0
|
50
|
April 2, 2024
|
How to root cause - torch/autograd/__init__.py:xxx: UserWarning: Error detected in GeluBackward0. Traceback of forward call that caused the error
|
|
2
|
81
|
March 31, 2024
|
Why does ignore_index ignore the entire example and not the class?
|
|
1
|
64
|
March 30, 2024
|
Computing Hessian
|
|
1
|
83
|
March 30, 2024
|
Different size for tensor grad compared to the tensor itself
|
|
4
|
101
|
March 29, 2024
|
Multiple inputs in shared weight layers
|
|
1
|
79
|
March 28, 2024
|
Restricting output range in last layer of deep architecture [regression task]
|
|
1
|
61
|
March 27, 2024
|
More efficient autograd for matrix w.r.t matrix gradient computation?
|
|
0
|
55
|
March 27, 2024
|
Modifying intermediate layer output using Hooks
|
|
4
|
98
|
March 27, 2024
|
Loss becomes NaN when introducing regularization
|
|
0
|
79
|
March 26, 2024
|
Backward time consumption linearly increases with batch_size
|
|
1
|
59
|
March 26, 2024
|
Torch Batch Norm
|
|
3
|
74
|
March 26, 2024
|
Per Sample Gradients
|
|
2
|
70
|
March 26, 2024
|
During backward() | CUDA error: an illegal memory access was encountered
|
|
4
|
176
|
March 26, 2024
|
Use gradcam gradients during training
|
|
0
|
53
|
March 26, 2024
|