-
The goal is to slice the tensor in half. For some reason that I cannot comprehend, it doesn't work on a particular tensor:
```python
>>> type(x)
>>> x.shape
torch.Size([128, 1024, 8, 64])
…
-
With dynamic resize, we have implemented a very basic system of trying to prove propositions at definition time (before concretization) if we have some constant inputs; for example, if we have constan…
-
Now that sparse tensors are mostly working (https://github.com/pytorch/pytorch/pull/1147), it would be awesome if all the optimizers worked with sparse tensors. This requires some cleverness to do amo…
-
In both the double-factorization and tensor hypercontraction versions of qubitization detailed in the [THC paper](https://arxiv.org/abs/2011.03494), we often need to load angles or coefficients into o…
-
### 🐛 Describe the bug
https://github.com/pytorch/pytorch/commit/32fbeb1 introduced a huge performance hit when inverting large batch GPU matrices.
Simple repro script:
```python
import time…
-
Target date: mid february 2024
Wanted:
- [x] python-3.12.2 with torch and numba
- [x] Operation Research packages refresh
- [x] a big clean-up in packages that still don't have a windows/3.12 wh…
-
Hi zongyi, thanks for your work.
I'm receiving this error "grad can be implicitly created only for scalar outputs" when I increase the number of in and out channels to 32. No such error when in/ou…
-
Hello,
I'm running your GPU-versoin Tensor_cell2cell demo code and find that the result (plot of factors) is highly dependent on the random_state in tensor.compute_tensor_factorization, when the in…
-
Hello
Thanks for the code and quick response to the comments & issues raised. Have you tried any convergence study for metallic structures ?
I tried it. Please see the notebook: https://github.com…
-
The current implementation of converting a strain tesnor into a B matrix in xfab.tools uses the small strain tesnor defenition
I suggest that the default behaviour is changed to the large Green-Lag…