diff options
| author | Onur Gündüz <ogunduz@gmail.com> | 2026-06-15 15:19:53 +0200 |
|---|---|---|
| committer | Onur Gündüz <ogunduz@gmail.com> | 2026-06-15 15:19:53 +0200 |
| commit | ceecbea701e3ac882ce5f1693204152cc75c0d79 (patch) | |
| tree | 397fc18d66aa68235f008d8d54a85683adb1ff5b /qwen_kernels.c | |
| parent | 8e8836796e7dd22820c820c0916e83ecdbf38368 (diff) | |
| download | ffwd-ceecbea701e3ac882ce5f1693204152cc75c0d79.tar.gz ffwd-ceecbea701e3ac882ce5f1693204152cc75c0d79.zip | |
kernels: add LayerNorm and exact-GeLU for the BERT family
Diffstat (limited to 'qwen_kernels.c')
| -rw-r--r-- | qwen_kernels.c | 38 |
1 files changed, 38 insertions, 0 deletions
diff --git a/qwen_kernels.c b/qwen_kernels.c index c7f5762..e1da7ed 100644 --- a/qwen_kernels.c +++ b/qwen_kernels.c @@ -768,6 +768,38 @@ void qwen_rms_norm_per_head( } } +/* Mean-subtracting LayerNorm with bias. Safe in place (out may alias x): each + * row's mean and variance are reduced before the row is rewritten. Scalar for + * now; LayerNorm is not GEMM-bound, so SIMD/threading is a later perf step. */ +void qwen_layer_norm(float *out, + const float *x, + const float *gamma, + const float *beta, + int seq_len, + int hidden, + float eps) { + if (seq_len <= 0 || hidden <= 0) + return; + float inv_h = 1.0f / (float)hidden; + for (int s = 0; s < seq_len; s++) { + const float *row = x + (size_t)s * hidden; + float *dst = out + (size_t)s * hidden; + float mean = 0.0f; + for (int d = 0; d < hidden; d++) + mean += row[d]; + mean *= inv_h; + float var = 0.0f; + for (int d = 0; d < hidden; d++) { + float diff = row[d] - mean; + var += diff * diff; + } + var *= inv_h; + float inv_std = 1.0f / sqrtf(var + eps); + for (int d = 0; d < hidden; d++) + dst[d] = gamma[d] * (row[d] - mean) * inv_std + beta[d]; + } +} + /* ======================================================================== * Activation Functions * ======================================================================== */ @@ -799,6 +831,12 @@ void qwen_silu_mul_inplace(float *gate, const float *up, int n) { #endif } +void qwen_gelu_inplace(float *x, int n) { + /* Exact erf GeLU; 0.70710678... = 1/sqrt(2). */ + for (int i = 0; i < n; i++) + x[i] = 0.5f * x[i] * (1.0f + erff(x[i] * 0.70710678118654752f)); +} + void qwen_softmax(float *x, int rows, int cols) { for (int r = 0; r < rows; r++) { float *row = x + r * cols; |
