78 lines
2.7 KiB
Zig
78 lines
2.7 KiB
Zig
// I am using this mnist reduced dataset https://www.kaggle.com/datasets/mohamedgamal07/reduced-mnist
|
|
|
|
const std = @import("std");
|
|
const gpu = @import("gpu");
|
|
const GpuDevice = gpu.GpuDevice;
|
|
const GpuArena = gpu.GpuArena;
|
|
const GpuBuffer = gpu.GpuBuffer;
|
|
const GpuProcess = gpu.GpuProcess;
|
|
|
|
const BATCHSIZE = 10;
|
|
const EPOCH = 10;
|
|
|
|
pub fn main(init: std.process.Init) !void {
|
|
const allocator = init.gpa;
|
|
const io = init.io;
|
|
|
|
// 1. Open GPU Device
|
|
const device = try GpuDevice.init(.{});
|
|
defer device.deinit();
|
|
|
|
// 2. Create a GPU Arena to manage VRAM
|
|
var grena = GpuArena.init(allocator, device);
|
|
defer grena.deinit();
|
|
const gloc = grena.gpuAllocator();
|
|
|
|
// 3. Load the WGSL compute pipeline
|
|
const add_process = try GpuProcess.init(device, @embedFile("shaders/add.wgsl"));
|
|
defer add_process.deinit();
|
|
|
|
var train_dir = try std.Io.Dir.cwd().openDir(io, "mnist/train", .{});
|
|
|
|
var images: [BATCHSIZE * 28 * 28]f16 = undefined;
|
|
for (EPOCH) |epoch| {
|
|
// Load random images from train dir
|
|
train_dir.openDir(io, "0", .{});
|
|
for (BATCHSIZE) |i| {
|
|
const file = try train_dir.openFile(io, "0.jpg", .{});
|
|
images[28 * 28 * i .. 28 * 28 * (i + 1)] = file.read
|
|
}
|
|
}
|
|
|
|
// 4. Setup CPU data
|
|
const len: usize = 16;
|
|
const data_a = try allocator.alloc(f16, len);
|
|
defer allocator.free(data_a);
|
|
const data_b = try allocator.alloc(f16, len);
|
|
defer allocator.free(data_b);
|
|
|
|
for (0..len) |i| {
|
|
data_a[i] = @floatFromInt(i);
|
|
data_b[i] = @floatFromInt(len - 1 - i);
|
|
}
|
|
|
|
// 5. Initialize raw GPU Buffers
|
|
// We pass the EnumSet inline using `.initMany` since the Enum itself isn't exported
|
|
const byte_size = len * @sizeOf(f16);
|
|
const buf_a = try GpuBuffer.init(gloc, byte_size, .initMany(&.{ .Storage, .CopyDst, .CopySrc }));
|
|
const buf_b = try GpuBuffer.init(gloc, byte_size, .initMany(&.{ .Storage, .CopyDst, .CopySrc }));
|
|
const buf_out = try GpuBuffer.init(gloc, byte_size, .initMany(&.{ .Storage, .CopyDst, .CopySrc }));
|
|
|
|
// Note: The buffers are safely tied to the GpuArena which will automatically
|
|
// release them at the end. You can also manually call buf_x.deinit() if desired.
|
|
|
|
// 6. Transfer data from CPU slices to GPU Buffers
|
|
try buf_a.load(f16, data_a);
|
|
try buf_b.load(f16, data_b);
|
|
|
|
// 7. Dispatch the Compute Process
|
|
// We pass the data type (f16) to allow GpuProcess to calculate chunks correctly
|
|
try add_process.run(gloc, f16, buf_a, buf_b, buf_out);
|
|
|
|
// 8. Map and copy the resulting buffer back to the CPU
|
|
const out = try buf_out.read(allocator, f16);
|
|
defer allocator.free(out);
|
|
|
|
std.debug.print("Result: {any}\n", .{out});
|
|
}
|