About the data category
|
|
0
|
1388
|
December 6, 2021
|
"Suggested max num workers is 2" but I have 96 cores?
|
|
4
|
756
|
May 1, 2024
|
Processing large dataset
|
|
3
|
1648
|
May 1, 2024
|
How to make one GPU work more than the other in DDP if the GPUs performance not the same? (Please answer this its really a unique question)
|
|
0
|
32
|
April 29, 2024
|
DataLoader memory usage keeps increasing
|
|
5
|
133
|
April 29, 2024
|
Setting num_workers>0 increase GPU memory of only 1 GPU
|
|
3
|
62
|
April 24, 2024
|
GIL contention between pin_memory & main trainer threads?
|
|
3
|
198
|
April 27, 2024
|
Dataloader yields copied batches
|
|
2
|
66
|
April 26, 2024
|
About nccl timeout
|
|
1
|
790
|
April 26, 2024
|
'bool' object is not callable, when launch dataloader
|
|
3
|
66
|
April 26, 2024
|
Torch Geometric Data Handling
|
|
1
|
300
|
April 23, 2024
|
Every "n" batches, there is a very slow batch
|
|
1
|
51
|
April 21, 2024
|
OH encoded features are lost when padding and packing
|
|
0
|
40
|
April 19, 2024
|
Caltech256 containing grayscale images
|
|
3
|
315
|
April 19, 2024
|
__getitem__ returning random sample
|
|
1
|
42
|
April 18, 2024
|
Mismatch of input dimension during training
|
|
1
|
48
|
April 18, 2024
|
Performance Impact of `torch.autograd.profiler.record_function` in DataLoader's `_BaseDataLoaderIter` Iteration
|
|
3
|
67
|
April 18, 2024
|
Efficient shuffling buffer in pinned memory
|
|
0
|
56
|
April 17, 2024
|
It seems hard to get a visual view of batch data in whole dataloader
|
|
0
|
53
|
April 15, 2024
|
Which is faster: Loading data in parallel or running multiple training processes in parallel but have data loaded sequentially?
|
|
0
|
42
|
April 10, 2024
|
Async data loading has huge GPU bubble
|
|
8
|
153
|
April 8, 2024
|
Data loading is very slow when using collate_fn with custom object instances
|
|
1
|
76
|
April 8, 2024
|
Help building and storing very large datasets?
|
|
2
|
63
|
April 7, 2024
|
Play with large tabular data
|
|
0
|
69
|
April 3, 2024
|
Slow .backward() with DataLoader
|
|
1
|
81
|
April 3, 2024
|
Confusing with the Impact of `num_workers` on Image Agumentation Speed with PIL.Image and Tensor
|
|
1
|
82
|
April 1, 2024
|
Proper scaling of .npy images in pytorch for training
|
|
1
|
75
|
April 1, 2024
|
Conv1d input shape mismatch with custom Dataset
|
|
2
|
108
|
April 1, 2024
|
How to load sequential data from separate datasets?
|
|
5
|
296
|
March 30, 2024
|
Lazy loading of wide dataset
|
|
3
|
1979
|
March 29, 2024
|