-
hi guys
In our previous project, we were using MKL's fft, and during the migration process, we encountered two issues with FFTW.
The first issue is that when using fft's c2r multiple times, the …
-
We do support some inplace thanks to @crcrpar 's great patches. This has enable running additional models already, but is limited to a few cases of particular interest.
However, we would want to se…
t-vi updated
3 months ago
-
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace opera…
-
### 🐛 Describe the bug
Failed cases:
- [ ] `GPUTests::test_inplace_resize_as_xpu`
- [ ] `CpuTests::test_inplace_resize_as_cpu`
Logs in https://github.com/intel/torch-xpu-ops/actions/runs/1050249…
-
This is a follow up to a discussion in #30, I'm making an issue out of it so that it isn't all buried over there (feel free to copy your points from there to here).
Syntax support for this feature …
-
Options:
- X-Editable: http://vitalets.github.io/x-editable/
- Bootstrap Editable: http://vitalets.github.io/bootstrap-editable/#
lfryc updated
11 years ago
-
I couldn't run `main.py`, because of inplace operation
File "main.py", line 202, in
loss_decoder.backward(retain)graph=True)
one of the Variables needed for gradient computation has bee…
-
As far as I'm aware certain boot packages like `base` are generally not possible to upgrade separately from the compiler, however other boot packages like `time` seem to be upgradable like any other p…
-
- [ ] ALTER TABLE oc_cards ADD INDEX z_oc_cards (uri), ALGORITHM=INPLACE, LOCK=NONE;
- [x] ALTER TABLE oc_cards ADD INDEX z_oc_cards_add_uri (addressbookid,uri), ALGORITHM=INPLACE, LOCK=NONE;
- [x] …
-
```
import threading
import torch
def foo(x, y):
a = torch.sin(x)
b = torch.cos(y)
return a + b
opt_foo1 = torch.compile(foo, mode="max-autotune")
threads = []
for _ in rang…