cmdr2
/ simple_ggml_addition.cpp
Last active
February 17, 2025 11:43
A simple program to add two vectors using ggml, that can compile to CPU or CUDA.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| #include "ggml.h" | |
| #include "ggml-cpu.h" | |
| #ifdef GGML_USE_CUDA | |
| #include "ggml-cuda.h" | |
| #endif | |
| #include <vector> | |
| #include <iostream> |
cmdr2
/ simple_addition_fp32.cpp
Created
February 24, 2025 07:44
Add two float32 tensors using ggml. Each tensor takes 1 GB of memory.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| #include "ggml.h" | |
| #include "ggml-cpu.h" | |
| #ifdef GGML_USE_CUDA | |
| #include "ggml-cuda.h" | |
| #endif | |
| #include <vector> | |
| #include <iostream> | |
| #include <chrono> |
cmdr2
/ simple_addition_fp16.cpp
Created
February 24, 2025 07:44
Add two float16 tensors using ggml. Each tensor takes 1 GB of memory.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| #include "ggml.h" | |
| #include "ggml-cpu.h" | |
| #ifdef GGML_USE_CUDA | |
| #include "ggml-cuda.h" | |
| #endif | |
| #include <vector> | |
| #include <iostream> | |
| #include <chrono> |
cmdr2
/ ggml-cpu.patch
Last active
March 28, 2025 06:10
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| 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]; |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| 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
/ ggml-cpu2.patch
Last active
April 2, 2025 11:38
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| 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
/ conv_with_stride_and_padding.cu
Created
April 8, 2025 08:37
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| #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
/ conv_with_stride.cu
Created
April 8, 2025 08:39
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| #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
/ conv_with_stride_padding_dilation_2_axis.cu
Created
April 8, 2025 11:41
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| #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
/ hsv_color_wheel.glsl
Last active
July 16, 2025 10:12
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| // 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); |