|
Pytorch Slack Participation
|
|
0
|
33
|
May 14, 2025
|
|
When matmul will be dispatched to cutlass gemm kernel?
|
|
3
|
94
|
May 14, 2025
|
|
Torch.cuda.is_available() returns false for p106-100 GPU
|
|
3
|
215
|
May 14, 2025
|
|
PyTorch Gan training: shape mismatch between real and fake tensor
|
|
2
|
21
|
May 13, 2025
|
|
Torch.export stability and timeline
|
|
0
|
31
|
May 13, 2025
|
|
When computing Loss.backward(), does grad_fn = CopySlices mean the gradient is loss?
|
|
2
|
30
|
May 13, 2025
|
|
What are the most effective and reliable ways to load minibatches efficiently from HDD for deep learning training?
|
|
0
|
12
|
May 13, 2025
|
|
torch.from_numpy().cuda() performance anomaly
|
|
3
|
95
|
May 13, 2025
|
|
VAEGAN - Multiple losses and multiple networks training
|
|
7
|
94
|
May 13, 2025
|
|
Why is torch.rand different on different devices for big indices
|
|
1
|
29
|
May 13, 2025
|
|
Why `max_memory_allocated()` exceeds physical GPU memory size?
|
|
3
|
111
|
May 12, 2025
|
|
Apply vectorized random permutations to conv2d weight tensors
|
|
4
|
110
|
May 12, 2025
|
|
Retrieving Attention Weights from scaled_dot_product_attention
|
|
7
|
1070
|
May 12, 2025
|
|
New to Contribution
|
|
4
|
121
|
May 11, 2025
|
|
Is there a way to parallelize independent sequential steps?
|
|
4
|
2419
|
May 9, 2025
|
|
[Question] How to solve no kernel image is available for execution on the device?
|
|
1
|
102
|
May 9, 2025
|
|
[Knowledge Distillation]: NaNs when combining CE + aux MSE in the same backward (BatchNorm state mismatch)
|
|
0
|
18
|
May 9, 2025
|
|
Trying to run PyTorch and Stable Diffusion and / or ComfyUI on FreeBSD
|
|
4
|
356
|
May 9, 2025
|
|
Feedback on AMD/Intel ARC GPUs compared to Nvidia
|
|
2
|
295
|
May 8, 2025
|
|
How can I visualize this NN?
|
|
6
|
48
|
May 8, 2025
|
|
Just a silly question. Whats the difference between import torch.nn as nn vs from torch import nn?
|
|
1
|
77
|
May 8, 2025
|
|
The pytorch script get stuck and exit without any error
|
|
2
|
55
|
May 8, 2025
|
|
Does flex_attention still require a sequence length of 128 or not?
|
|
1
|
41
|
May 7, 2025
|
|
Is my training fast or not?
|
|
0
|
29
|
May 7, 2025
|
|
Any suggestion about long time serie prediction as this kind?
|
|
0
|
49
|
May 7, 2025
|
|
About running time of Convolution
|
|
0
|
52
|
May 6, 2025
|
|
Mismatch in data shapes
|
|
0
|
47
|
May 6, 2025
|
|
How to assignment between two parameters with their string names?
|
|
3
|
22
|
May 6, 2025
|
|
Transformer Loss remains same
|
|
0
|
35
|
May 6, 2025
|
|
RuntimeError: Expected canUse32BitIndexMath(input)
|
|
7
|
251
|
May 6, 2025
|