llama-cpp-turboquant/test/shaders/glsl/test_logistic_regression.comp
2020-08-31 19:01:41 +01:00

52 lines
1.4 KiB
Text

#version 450
layout (constant_id = 0) const uint M = 0;
layout (local_size_x = 1) in;
layout(set = 0, binding = 0) buffer bxi { float xi[]; };
layout(set = 0, binding = 1) buffer bxj { float xj[]; };
layout(set = 0, binding = 2) buffer by { float y[]; };
layout(set = 0, binding = 3) buffer bwin { float win[]; };
layout(set = 0, binding = 4) buffer bwouti { float wouti[]; };
layout(set = 0, binding = 5) buffer bwoutj { float woutj[]; };
layout(set = 0, binding = 6) buffer bbin { float bin[]; };
layout(set = 0, binding = 7) buffer bbout { float bout[]; };
float learningRate = 0.1;
float m = float(M);
float sigmoid(float z) {
return 1.0 / (1.0 + exp(-z));
}
float inference(vec2 x, vec2 w, float b) {
float z = dot(w, x) + b;
float yHat = sigmoid(z);
return yHat;
}
float calculateLoss(float yHat, float y) {
return -(y * log(yHat) + (1.0 - y) * log(1.0 - yHat));
}
void main() {
uint idx = gl_GlobalInvocationID.x;
vec2 wCurr = vec2(win[0], win[1]);
float bCurr = bin[0];
vec2 xCurr = vec2(xi[idx], xj[idx]);
float yCurr = y[idx];
float yHat = inference(xCurr, wCurr, bCurr);
float loss = calculateLoss(yHat, yCurr);
float dZ = yHat - yCurr;
vec2 dW = (1. / m) * xCurr * dZ;
float dB = (1. / m) * dZ;
wouti[idx] = learningRate * dW.x;
woutj[idx] = learningRate * dW.y;
bout[idx] = learningRate * dB;
}