About the performance category
|
|
0
|
458
|
January 22, 2021
|
Fast combined C++/Python/TorchScript/Inductor tracebacks
|
|
4
|
1547
|
August 17, 2023
|
The "Ideal" PyTorch FLOP Counter (with __torch_dispatch__)
|
|
13
|
6043
|
April 18, 2023
|
Estimate theoritical FLOPs of backward pass of a DNN
|
|
1
|
301
|
April 16, 2023
|
Performance gains w/ nanoGPT using SDPA Custom Kernel
|
|
0
|
3362
|
January 30, 2023
|
Making Transformer inference faster on GPUs
|
|
1
|
2637
|
December 16, 2022
|
Investigation report: what would it cost to optimize c10::intrusive_ptr destruction for refcount == 1? (A: too much)
|
|
0
|
427
|
May 3, 2022
|
Working With `c10::IValue` Efficiently
|
|
0
|
1440
|
April 18, 2022
|
Unionizing for Profit: How to Exploit the Power of Unions in C++
|
|
2
|
1559
|
January 7, 2022
|
CUDA loops case study: code generation vs templates
|
|
4
|
1602
|
December 12, 2021
|
Multiple workers for single batch
|
|
2
|
620
|
July 14, 2021
|
Optimizing contiguous() for the case where the Tensor is_contiguous()?
|
|
6
|
894
|
May 24, 2021
|
Converting weights flat buffer
|
|
0
|
581
|
March 25, 2021
|
Why `torch::jit::pop` (and sometimes push) is worse for performance than direct `std::vector` access
|
|
0
|
730
|
March 17, 2021
|
Pytorch Benchmarks issues with general usability and issues with individual benchmarks
|
|
2
|
919
|
March 12, 2021
|
GPU Overheads and Fused Strassen
|
|
0
|
1200
|
February 13, 2021
|
Comparing the performance of 0.4.1 and master
|
|
0
|
1658
|
February 9, 2021
|
Overhead in `nn.Module` causing massive slowdowns compared to raw CuBLAS or Torchscript
|
|
0
|
1205
|
January 28, 2021
|
Dispatcher Performance and Inlining: a Report on Two Days Spent on Dispatcher Performance
|
|
1
|
809
|
January 28, 2021
|
We shouldn't feel bad about passing `Tensor` by reference
|
|
0
|
1397
|
January 25, 2021
|