-
Why doesn't cuSPARSE support dense matrix sparse matrix multiplication resulting in a dense matrix? Many application scenarios require this. Please consider adding support.
```[tasklist]
### Tasks
…
-
Hi,
I get this error
```
**************************************************|
0% 10 20 30 40 50 60 70 80 90 100%
[----|----|----|----|----|----|----|----|----|----|
********…
-
### Describe the bug
When using 8-bit quantization with the LLM pipeline and a multiple GPU setup, it mostly runs fine.
After some random amount of requests however the pipeline starts failing an…
-
-
### Describe the bug
Multi-line description of Raises messages are concatenared into a single string interleaved by `\n`, this doesn't ultimately render in html (i.e. it renders as a single string). …
-
Imported from SourceForge on 2024-07-06 17:09:15
Created by **[macrakis](https://sourceforge.net/u/macrakis/)** on 2018-02-05 22:04:48
Original: https://sourceforge.net/p/maxima/bugs/3394
---
matrix…
rtoy updated
4 months ago
-
MWE:
```julia-repl
julia> a = Metal.randn(10000, 10000);
julia> b = Metal.randn(10000, 10000);
julia> c = a * b';
julia> for i in 1:10
C = Metal.zeros(Float32, size(a))
…
-
## Feature request
Hi, not sure if this should be considered a feature request or a bug report but matrix multiplication and `numpy.dot()` do not behave as expected for the multiplication of matrix…
-
## Description:
Create a matrix-vector multiplication method for the Coordinate Format (COO) matrix class.
Design and implement a matrix-vector multiplication (matvec) method for our COO sparse mat…
-
### Describe the bug
The performance from joint_matrix multiplication is around 10 TOPS with int8 on a device with theoretical capability of 383 TOPS. This specific code snippet is for the AMD GPU sp…