// SPDX-License-Identifier: AGPL-3.0-or-later
// BatchNorm - Batch normalization (f64 canonical)
// output = (input - mean) / sqrt(variance + epsilon) * gamma + beta
// Simplified version: per-tensor normalization with default gamma=1, beta=0
struct BatchNormParams {
epsilon: f64,
_padding: vec3<f64>,
}
@group(0) @binding(0) var<storage, read> input: array<f64>;
@group(0) @binding(1) var<storage, read_write> output: array<f64>;
@group(0) @binding(2) var<uniform> params: BatchNormParams;
@compute @workgroup_size(256)
fn main(@builtin(global_invocation_id) global_id: vec3<u32>) {
let idx = global_id.x;
let size = arrayLength(&input);
if (idx >= size) {
return;
}
// Compute mean (simplified: single pass for small tensors)
if (idx == 0u) {
var sum = f64(0.0);
for (var i = 0u; i < size; i = i + 1u) {
sum = sum + input[i];
}
let mean = sum / f64(size);
// Compute variance
var variance = f64(0.0);
for (var i = 0u; i < size; i = i + 1u) {
let diff = input[i] - mean;
variance = variance + diff * diff;
}
variance = variance / f64(size);
// Normalize all elements (gamma=1, beta=0 for simplicity)
let std_dev = sqrt_f64(variance + params.epsilon);
for (var i = 0u; i < size; i = i + 1u) {
output[i] = (input[i] - mean) / std_dev;
}
}
}