-
Division by zero trying to compile the following simple vertex shader:
```
uniform mat4 modelviewprojectionMatrix;
attribute vec4 in_position;
attribute vec2 in_coord;
varyin…
ssvb updated
10 years ago
-
I was reading through the matmul kernels, and I noticed the beginning of the vecmat kernel looked like this:
```
static METAL_FUNC void run(
const device T* mat,
const device T* in_vec…
-
In the ```update_kv``` function of ```H2OKVCluster``` class, I see this code.
```
attn_weights = torch.matmul(query_states[..., -self.window_size:, :], key_states.transpose(2, 3)) / math.sqrt(head…
-
Hi there! I'm currently looking for an autodiff approach to perform a max-plus matrix multiplication on cuda:
I've been experimenting with KeOps, but it seems it might not be the ideal tool for thi…
-
Benchmarking an R+deSolve code against the equivalent odin code yielded a surprising result:
```
# A tibble: 2 × 13
expression min median `itr/sec` mem_alloc `gc/sec` n_itr n_gc total_tim…
-
### *Description*
Write the program of the matrix chain multiplication
-
It looks like this is falling to the standard `generic_matmatmul!`, which errors on scalar indexing
```julia
julia> using LinearAlgebra, Metal, StaticArrays
julia> Metal.GPUArrays.allowscalar(f…
-
This is not a priority feature, but efficient SPMMs would allow to compute memory-efficient message-passing in cases where we have very large graphs. The sparse matrix would be the `edge_index` repres…
-
I wanted to print an expression unsimplified and simplified, Left hand side = Right hand side, so I thought I would use this:
```py
from sympy import *
import IPython.display as disp
A = Matrix(…
-
sparse matrix multiplication in sympy uses naive multiplication algorithm. If i am right about this then the following algorithm can be implemented for sparse matrices.
https://www.degruyter.com/down…