From c7af9042b3e732ce493a3303ddb0620b0321f2a5 Mon Sep 17 00:00:00 2001 From: Georgi Gerganov Date: Fri, 5 May 2023 17:09:11 +0300 Subject: [PATCH] ggml : remove Q5_1 bit shuffling (ARM NEON + scalar) --- ggml.c | 172 ++++++++++++++++++++++----------------------------------- 1 file changed, 66 insertions(+), 106 deletions(-) diff --git a/ggml.c b/ggml.c index 2cbc9b931..ed8616d59 100644 --- a/ggml.c +++ b/ggml.c @@ -851,8 +851,7 @@ static_assert(sizeof(block_q8_1) == 3*sizeof(float) + QK8_1, "wrong q8_1 block s static void quantize_row_q4_0_reference(const float * restrict x, block_q4_0 * restrict y, int k) { static const int qk = QK4_0; - assert(qk / 16 == 0); - assert( k % qk == 0); + assert(k % qk == 0); const int nb = k / qk; @@ -873,20 +872,16 @@ static void quantize_row_q4_0_reference(const float * restrict x, block_q4_0 * r y[i].d = d; - uint64_t qs[QK4_0 / 16] = {0}; - for (int l = 0; l < qk/2; ++l) { const float x0 = x[i*qk + 0 + l]*id; const float x1 = x[i*qk + qk/2 + l]*id; - const uint64_t xi0 = MIN(15, (int8_t)(x0 + 8.5f)); - const uint64_t xi1 = MIN(15, (int8_t)(x1 + 8.5f)); + const uint8_t xi0 = MIN(15, (int8_t)(x0 + 8.5f)); + const uint8_t xi1 = MIN(15, (int8_t)(x1 + 8.5f)); - qs[l/8] |= xi0 << (8*(l & 7)); - qs[l/8] |= xi1 << (8*(l & 7) + 4); + y[i].qs[l] = xi0; + y[i].qs[l] |= xi1 << 4; } - - memcpy(y[i].qs, qs, qk/2); } } @@ -897,8 +892,7 @@ static void quantize_row_q4_0(const float * restrict x, void * restrict y, int k static void quantize_row_q4_1_reference(const float * restrict x, block_q4_1 * restrict y, int k) { const int qk = QK4_1; - assert(qk / 16 == 0); - assert( k % qk == 0); + assert(k % qk == 0); const int nb = k / qk; @@ -919,20 +913,16 @@ static void quantize_row_q4_1_reference(const float * restrict x, block_q4_1 * r y[i].d = d; y[i].m = min; - uint64_t qs[QK4_1 / 16] = {0}; - for (int l = 0; l < qk/2; ++l) { const float x0 = (x[0 + l] - min)*id; const float x1 = (x[qk/2 + l] - min)*id; - const uint64_t xi0 = MIN(15, (int8_t)(x0 + 0.5f)); - const uint64_t xi1 = MIN(15, (int8_t)(x1 + 0.5f)); + const uint8_t xi0 = MIN(15, (int8_t)(x0 + 0.5f)); + const uint8_t xi1 = MIN(15, (int8_t)(x1 + 0.5f)); - qs[l/8] |= xi0 << (8*(l & 7)); - qs[l/8] |= xi1 << (8*(l & 7) + 4); + y[i].qs[l] = xi0; + y[i].qs[l] |= xi1 << 4; } - - memcpy(y[i].qs, qs, qk/2); } } @@ -944,8 +934,7 @@ static void quantize_row_q4_1(const float * restrict x, void * restrict y, int k static void quantize_row_q4_2_reference(const float * restrict x, block_q4_2 * restrict y, int k) { static const int qk = QK4_2; - assert(qk / 16 == 0); - assert( k % qk == 0); + assert(k % qk == 0); const int nb = k / qk; @@ -990,8 +979,7 @@ static void quantize_row_q4_2(const float * restrict x, void * restrict y, int k static void quantize_row_q5_0_reference(const float * restrict x, block_q5_0 * restrict y, int k) { static const int qk = QK5_0; - assert(qk / 16 == 0); - assert( k % qk == 0); + assert(k % qk == 0); const int nb = k / qk; @@ -1013,24 +1001,21 @@ static void quantize_row_q5_0_reference(const float * restrict x, block_q5_0 * r y[i].d = d; uint32_t qh = 0; - uint64_t qs[QK5_0 / 16] = {0}; for (int l = 0; l < qk/2; ++l) { const float x0 = x[i*qk + 0 + l]*id; const float x1 = x[i*qk + qk/2 + l]*id; - const uint64_t xi0 = MIN(31, (int8_t)(x0 + 16.5f)); - const uint64_t xi1 = MIN(31, (int8_t)(x1 + 16.5f)); + const uint8_t xi0 = MIN(31, (int8_t)(x0 + 16.5f)); + const uint8_t xi1 = MIN(31, (int8_t)(x1 + 16.5f)); - qs[l/8] |= xi0 << (8*(l & 7)); - qs[l/8] |= xi1 << (8*(l & 7) + 4); + y[i].qs[l] = (xi0 & 0x0F) | ((xi1 & 0x0F) << 4); // get the 5-th bit and store it in qh at the right position qh |= ((xi0 & 0x10) >> 4) << (l + 0); qh |= ((xi1 & 0x10) >> 4) << (l + qk/2); } - memcpy( y[i].qs, qs, qk/2); memcpy(&y[i].qh, &qh, sizeof(qh)); } } @@ -1040,20 +1025,24 @@ static void quantize_row_q5_0(const float * restrict x, void * restrict y, int k } static void quantize_row_q5_1_reference(const float * restrict x, block_q5_1 * restrict y, int k) { - assert(k % QK5_1 == 0); - const int nb = k / QK5_1; + const int qk = QK5_1; + + assert(k % qk == 0); + + const int nb = k / qk; for (int i = 0; i < nb; i++) { float min = FLT_MAX; float max = -FLT_MAX; - for (int l = 0; l < QK5_1; l++) { - const float v = x[i*QK5_1 + l]; + for (int l = 0; l < qk; l++) { + const float v = x[i*qk + l]; + if (v < min) min = v; if (v > max) max = v; } - const float d = (max - min) / ((1 << 5) - 1); + const float d = (max - min) / ((1 << 5) - 1); const float id = d ? 1.0f/d : 0.0f; y[i].d = GGML_FP32_TO_FP16(d); @@ -1061,29 +1050,25 @@ static void quantize_row_q5_1_reference(const float * restrict x, block_q5_1 * r uint32_t qh = 0; - for (int l = 0; l < QK5_1; l += 2) { - const float v0 = (x[i*QK5_1 + l + 0] - min)*id; - const float v1 = (x[i*QK5_1 + l + 1] - min)*id; + for (int l = 0; l < qk/2; ++l) { + const float x0 = (x[i*qk + 0 + l] - min)*id; + const float x1 = (x[i*qk + qk/2 + l] - min)*id; - const uint32_t vi0 = (int) (v0 + 0.5f); - const uint32_t vi1 = (int) (v1 + 0.5f); + const uint8_t xi0 = (uint8_t)(x0 + 0.5f); + const uint8_t xi1 = (uint8_t)(x1 + 0.5f); - y[i].qs[l/2] = (vi0 & 0x0F) | ((vi1 & 0x0F) << 4); + y[i].qs[l] = (xi0 & 0x0F) | ((xi1 & 0x0F) << 4); // get the 5-th bit and store it in qh at the right position - qh |= ((vi0 & 0x10) >> 4) << (l + 0); - qh |= ((vi1 & 0x10) >> 4) << (l + 1); + qh |= ((xi0 & 0x10) >> 4) << (l + 0); + qh |= ((xi1 & 0x10) >> 4) << (l + qk/2); } memcpy(&y[i].qh, &qh, sizeof(y[i].qh)); } } -static void quantize_row_q5_1(const float * restrict x, void * restrict vy, int k) { - assert(k % QK5_1 == 0); - - block_q5_1 * restrict y = vy; - +static void quantize_row_q5_1(const float * restrict x, void * restrict y, int k) { quantize_row_q5_1_reference(x, y, k); } @@ -1443,8 +1428,7 @@ static void quantize_row_q8_1(const float * restrict x, void * restrict vy, int static void dequantize_row_q4_0(const block_q4_0 * restrict x, float * restrict y, int k) { static const int qk = QK4_0; - assert(qk / 16 == 0); - assert( k % qk == 0); + assert(k % qk == 0); const int nb = k / qk; @@ -1464,8 +1448,7 @@ static void dequantize_row_q4_0(const block_q4_0 * restrict x, float * restrict static void dequantize_row_q4_1(const block_q4_1 * restrict x, float * restrict y, int k) { static const int qk = QK4_1; - assert(qk / 16 == 0); - assert( k % qk == 0); + assert(k % qk == 0); const int nb = k / qk; @@ -1487,8 +1470,7 @@ static void dequantize_row_q4_2(const block_q4_2 * restrict x, float * restrict // BORKEN !!! static const int qk = QK4_2; - assert(qk / 16 == 0); - assert( k % qk == 0); + assert(k % qk == 0); const int nb = k / qk; @@ -1508,8 +1490,7 @@ static void dequantize_row_q4_2(const block_q4_2 * restrict x, float * restrict static void dequantize_row_q5_0(const block_q5_0 * restrict x, float * restrict y, int k) { static const int qk = QK4_0; - assert(qk / 16 == 0); - assert( k % qk == 0); + assert(k % qk == 0); const int nb = k / qk; @@ -1532,39 +1513,29 @@ static void dequantize_row_q5_0(const block_q5_0 * restrict x, float * restrict } } -static void dequantize_row_q5_1(const void * restrict vx, float * restrict y, int k) { - assert(k % QK5_1 == 0); - const int nb = k / QK5_1; +static void dequantize_row_q5_1(const block_q5_1 * restrict x, float * restrict y, int k) { + static const int qk = QK5_1; - const block_q5_1 * restrict x = vx; + assert(k % qk == 0); + + const int nb = k / qk; for (int i = 0; i < nb; i++) { const float d = GGML_FP16_TO_FP32(x[i].d); const float m = GGML_FP16_TO_FP32(x[i].m); - const uint8_t * restrict pp = x[i].qs; - uint32_t qh; memcpy(&qh, x[i].qh, sizeof(qh)); - for (int l = 0; l < QK5_1; l += 2) { - const uint8_t vi = pp[l/2]; + for (int j = 0; j < qk/2; ++j) { + const uint8_t xh_0 = ((qh & (1u << (j + 0 ))) >> (j + 0 )) << 4; + const uint8_t xh_1 = ((qh & (1u << (j + 16))) >> (j + 12)); - // extract the 5-th bit from qh - const uint8_t vh0 = ((qh & (1u << (l + 0))) >> (l + 0)) << 4; - const uint8_t vh1 = ((qh & (1u << (l + 1))) >> (l + 1)) << 4; + const int x0 = (x[i].qs[j] & 0xf) | xh_0; + const int x1 = (x[i].qs[j] >> 4) | xh_1; - const uint8_t vi0 = (vi & 0x0F) | vh0; - const uint8_t vi1 = (vi >> 4) | vh1; - - const float v0 = vi0*d + m; - const float v1 = vi1*d + m; - - y[i*QK5_1 + l + 0] = v0; - y[i*QK5_1 + l + 1] = v1; - - assert(!isnan(y[i*QK5_1 + l + 0])); - assert(!isnan(y[i*QK5_1 + l + 1])); + y[i*qk + j + 0 ] = x0*d + m; + y[i*qk + j + qk/2] = x1*d + m; } } } @@ -1627,7 +1598,7 @@ static const quantize_fns_t quantize_fns[GGML_TYPE_COUNT] = { .vec_dot_type = GGML_TYPE_Q8_0, }, [GGML_TYPE_Q5_1] = { - .dequantize_row_q = dequantize_row_q5_1, + .dequantize_row_q = (dequantize_row_q_t) dequantize_row_q5_1, .quantize_row_q = quantize_row_q5_1, .quantize_row_q_reference = (quantize_row_q_t) quantize_row_q5_1_reference, .quantize_row_q_dot = quantize_row_q8_1, @@ -2875,11 +2846,12 @@ static void ggml_vec_dot_q5_0_q8_0(const int n, float * restrict s, const void * } static void ggml_vec_dot_q5_1_q8_1(const int n, float * restrict s, const void * restrict vx, const void * restrict vy) { - const int nb = n / QK8_1; + const int qk = QK8_1; + const int nb = n / qk; - assert(n % QK8_1 == 0); + assert(n % qk == 0); assert(nb % 2 == 0); - assert(QK8_1 == QK5_1); + assert(qk == QK5_1); const block_q5_1 * restrict x = vx; const block_q8_1 * restrict y = vy; @@ -2915,13 +2887,9 @@ static void ggml_vec_dot_q5_1_q8_1(const int n, float * restrict s, const void * const int8x16_t v0l = vreinterpretq_s8_u8(vandq_u8 (v0, vdupq_n_u8(0x0F))); const int8x16_t v0h = vreinterpretq_s8_u8(vshrq_n_u8(v0, 4)); - // interleave - const int8x16_t v0lz = vzip1q_s8(v0l, v0h); - const int8x16_t v0hz = vzip2q_s8(v0l, v0h); - // add - const int8x16_t v0lf = vorrq_s8(v0lz, qhl); - const int8x16_t v0hf = vorrq_s8(v0hz, qhh); + const int8x16_t v0lf = vorrq_s8(v0l, qhl); + const int8x16_t v0hf = vorrq_s8(v0h, qhh); // load y const int8x16_t v1l = vld1q_s8(y0->qs); @@ -3044,36 +3012,28 @@ static void ggml_vec_dot_q5_1_q8_1(const int n, float * restrict s, const void * *s = hsum_float_8(acc) + summs; #else + // scalar float sumf = 0.0; for (int i = 0; i < nb; i++) { - const uint8_t * restrict x0 = x[i].qs; - const int8_t * restrict y0 = y[i].qs; + const int8_t * py = y[i].qs; uint32_t qh; memcpy(&qh, x[i].qh, sizeof(qh)); - const float d = GGML_FP16_TO_FP32(x[i].d); - const float m = GGML_FP16_TO_FP32(x[i].m); + int sumi = 0; - int sxy = 0; + for (int j = 0; j < qk/2; ++j) { + const uint8_t xh_0 = ((qh & (1u << (j + 0 ))) >> (j + 0 )) << 4; + const uint8_t xh_1 = ((qh & (1u << (j + 16))) >> (j + 12)); - for (int j = 0; j < QK8_1/2; j++) { - const uint8_t v0 = x0[j]; + const int32_t x0 = (x[i].qs[j] & 0xF) | xh_0; + const int32_t x1 = (x[i].qs[j] >> 4) | xh_1; - const int x0_0h = ((qh & (1u << (2*j + 0))) >> (2*j + 0)) << 4; - const int x1_0h = ((qh & (1u << (2*j + 1))) >> (2*j + 1)) << 4; - - const int x0_0 = (v0 & 0x0F) | x0_0h; - const int x1_0 = (v0 >> 4) | x1_0h; - - const int y0_0 = y0[2*j + 0]; - const int y1_0 = y0[2*j + 1]; - - sxy += x0_0*y0_0 + x1_0*y1_0; + sumi += (x0 * py[j]) + (x1 * py[j + qk/2]); } - sumf += (d*sxy)*y[i].d + m*(y[i].s0 + y[i].s1); + sumf += (GGML_FP16_TO_FP32(x[i].d)*y[i].d)*sumi + GGML_FP16_TO_FP32(x[i].m)*(y[i].s0 + y[i].s1); } *s = sumf;