-
Hi
I compile the c code use cross compile hi3519,
and the CFLAGS and LIBFLAGS as below:
CFLAGS = -O3 -std=c11 -mcpu=cortex-a17 -mfloat-abi=softfp -mfpu=neon-vfpv4 -flto -Ofast -D NUM_THREAD=1 -I${…
-
Torch.jl wont precompile on my Windows machine, I also updated CUDA to version 11.0:
```
julia> versioninfo()
Julia Version 1.4.2
Commit 44fa15b150* (2020-05-23 18:35 UTC)
Platform Info:
OS:…
-
when I use gitbook serve to compile in win10, there is an error.
error: error while generating page "di-san-zhang-mo-dao-bu-wu-kan-chai-gong-tensorflow-ji-chu/di-san-zhang-tensorflow-ji-chu/3.1-ten…
-
e.g. `cat(a: TensorType["x"], b: TensorType["y"]) -> TensorType["x + y"]`.
Given that `torchtyping` operates at runtime then arbitrary Python expressions should be possible. Essentially just call `ev…
-
Building the unit tests on debian with gcc (clang won't build gtest) results in a bunch of exceptions like:
```sh
/usr/include/x86_64-linux-gnu/c++/10/bits/c++config.h:283:27: error: #if with no e…
-
Sometimes a small test below
```python
import dpctl, dpctl.tensor as dpt
import numpy
list_of_backend_str = [
"host",
"level_zero",
"opencl",
]
list_of_device_type_str = [
…
-
### 🐛 Describe the bug
```
TORCHDYNAMO_DYNAMIC_SHAPES=1 AOT_DYNAMIC_SHAPES=0 time python benchmarks/dynamo/huggingface.py --accuracy --backend eager --training --only AllenaiLongformerBase
```
…
-
The graph optimizer misses a simple log - softmax expression if the softmax or the log is transposed:
```
import theano.tensor as T
import theano
import numpy as np
X = T.fmatrix('X')
testA = theano…
-
I'm trying to get a better understanding of the objective function that is used to determine how well the image at a given point in space matches and how it could be used to take into account expressi…
-
```python
import sympy as S
d = S.symbols('d', integer=True, positive=True)
x = S.MatrixSymbol('x', d, 1)
expr = x / S.trace(x.T @ x)
grad = expr.diff(x)
print(grad)
```
The above code…