cmdr2

import bpy
import gpu
from gpu_extras.batch import batch_for_shader

VERT_SHADER = """
    uniform mat4 ModelViewProjectionMatrix;
    uniform mat4 ModelMatrix;

    in vec2 texCoord;
    in vec3 position;

cmdr2

import bpy
import gpu
from gpu_extras.batch import batch_for_shader

VERT_SHADER = """
    uniform mat4 ModelViewProjectionMatrix;

    in vec2 texCoord;
    in vec3 position;
    out vec3 pos;

cmdr2

// can run on https://thebookofshaders.com/edit.php

#ifdef GL_ES
precision mediump float;
#endif

uniform vec2 u_resolution;

vec3 hsv2rgb(vec3 c) {
    vec4 K = vec4(1.0, 2.0/3.0, 1.0/3.0, 3.0);

cmdr2

#include <cuda_runtime.h>
#include <iostream>
#include <chrono>

#define CHECK_CUDA(call) \
    if ((call) != cudaSuccess) { \
        std::cerr << "CUDA error at " << __FILE__ << ":" << __LINE__ << std::endl; \
        std::exit(1); \
    }

cmdr2

#include <cuda_runtime.h>
#include <iostream>
#include <chrono>

#define CHECK_CUDA(call)                                                       \
    do {                                                                       \
        cudaError_t err = call;                                                \
        if (err != cudaSuccess) {                                              \
            std::cerr << "CUDA Error: " << cudaGetErrorString(err) << "\n";   \
            exit(EXIT_FAILURE);                                                \

cmdr2

#include <cuda_runtime.h>
#include <iostream>
#include <chrono>

#define CHECK_CUDA(call)                                                       \
    do {                                                                       \
        cudaError_t err = call;                                                \
        if (err != cudaSuccess) {                                              \
            std::cerr << "CUDA Error: " << cudaGetErrorString(err) << "\n";   \
            exit(EXIT_FAILURE);                                                \

cmdr2

12,13c12,15
< #include "ggml-cpu/unary-ops.h"
< #include "ggml-cpu/binary-ops.h"
---
> #include "unary-ops.h"
> #include "binary-ops.h"
> #include "vec.h"
> #include "ops.h"
86,109d87
< #if defined(GGML_USE_ACCELERATE)

cmdr2

diff --git a/ops.cpp b/ops.cpp
index 6190d0d..c44157b 100644
--- a/ops.cpp
+++ b/ops.cpp
@@ -2347,7 +2347,7 @@ static void ggml_compute_forward_repeat_back_f32(
     GGML_ASSERT(nb00 == sizeof(float));
 
     if (ggml_is_contiguous(dst)) {
-        ggml_vec_set_f32(ne0*ne1*ne2*ne3, dst->data, 0);
+        ggml_vec_set_f32(ne0*ne1*ne2*ne3, (float *)dst->data, 0);

cmdr2

11a12,13
> #include "ggml-cpu/unary-ops.h"
> #include "ggml-cpu/binary-ops.h"
4292,4625d4293
< static void ggml_compute_forward_add_f32(
<         const struct ggml_compute_params * params,
<         struct ggml_tensor * dst) {
< 
<     const struct ggml_tensor * src0 = dst->src[0];
<     const struct ggml_tensor * src1 = dst->src[1];

cmdr2

#include "ggml.h"
#include "ggml-cpu.h"

#ifdef GGML_USE_CUDA
#include "ggml-cuda.h"
#endif

#include <vector>
#include <iostream>
#include <chrono>