-
### 🐛 Describe the bug
Hello,
In an experiment, I got wrong results after a gather. the issue seems to come from the strides of the index.
I have reproduced the bug here, the tensor operations …
-
- [x] Alternative strides for NDStructure
- [ ] Strides implementation which points to a sub-structure of larger structure (and thus Strides), effectively creating a View.
-
There is an issue with the refactor using more memory than the current version.
Using the uniform stride test ([cpu-ustride.json](https://github.com/hpcgarage/spatter/blob/main/standard-suite/basi…
-
Recently when I'm trying loop vectorize on llvm 17.x version, in a certain case I found when storing value into array sometimes it stored into wrong position. Here is a case reduced from the case I me…
-
**What is your question?**
Hi, I'd like to compute the following
```
// inputs
// A [M, K] int8
// B [N, K] int4
// alphaCol [M, 1] …
-
Using Fake Tensors
-
用torchaudio替换librosa,使用项目数据集训练模型后,推理报错:RuntimeError: running_mean should contain 344 elements not 128
推理代码如下:
with torch.no_grad():
waveform, sample_rate = torchaudio.load("data/valid…
-
Thanks @timlacroix for adding local attention support. Do you know if it would be straightforward to add a "stride" to the local window attention to better support the cross attention case where there…
-
### 🚀 The feature, motivation and pitch
I ran the following program to test what triton code is generated from a discontiguous tensor:
```
import sys
import os
import logging
import torch
fro…
-
Appreciate your great work!
When I try to run your text2shape model, I notice that stride=(1, 2, 2) in Downsample() and Upsample().
Could you tell me more reasons why you chose stride=(1, 2, 2)? Wh…