Great. I have the desire but not the knowledge. I’m a user, not a programmer. otherwise I’d do it. But I must say, that with these faster functions, Pytensor is again a good option for neural networks. In my opinion, better than Tensorflow or Pytorch.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| How to speed up AdvancedIncSubtensor | 8 | 907 | August 24, 2019 | |
| No JAX conversion for aesara convolution | 0 | 545 | October 18, 2022 | |
| Configure Theano | 1 | 732 | June 8, 2020 | |
| Theano deprecation at 1.0 | 16 | 4167 | October 7, 2017 | |
| Optimizing matmul operations with PyTensor | 0 | 220 | December 16, 2023 |