From 293a59d4f1bbc9f4f23db7dead0d68cc445a9348 Mon Sep 17 00:00:00 2001 From: Iwan Kawrakow Date: Thu, 25 Sep 2025 18:49:25 +0300 Subject: [PATCH] Add mtmd: refresh CPU rope --- ggml/src/ggml.c | 276 +++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 227 insertions(+), 49 deletions(-) diff --git a/ggml/src/ggml.c b/ggml/src/ggml.c index 8677bb55..72520928 100644 --- a/ggml/src/ggml.c +++ b/ggml/src/ggml.c @@ -17814,6 +17814,22 @@ static void ggml_compute_forward_clamp( // ggml_compute_forward_rope +// Apparently solving `n_rot = 2pi * x * base^((2 * max_pos_emb) / n_dims)` for x, we get +// `corr_dim(n_rot) = n_dims * log(max_pos_emb / (n_rot * 2pi)) / (2 * log(base))` +static float ggml_rope_yarn_corr_dim(int n_dims, int n_ctx_orig, float n_rot, float base) { + return n_dims * logf(n_ctx_orig / (n_rot * 2 * (float)M_PI)) / (2 * logf(base)); +} + +void ggml_rope_yarn_corr_dims( + int n_dims, int n_ctx_orig, float freq_base, float beta_fast, float beta_slow, float dims[2] +) { + // start and end correction dims + float start = floorf(ggml_rope_yarn_corr_dim(n_dims, n_ctx_orig, beta_fast, freq_base)); + float end = ceilf(ggml_rope_yarn_corr_dim(n_dims, n_ctx_orig, beta_slow, freq_base)); + dims[0] = MAX(0, start); + dims[1] = MIN(n_dims - 1, end); +} + static float rope_yarn_ramp(const float low, const float high, const int i0) { const float y = (i0 / 2 - low) / MAX(0.001f, high - low); return 1 - MIN(1, MAX(0, y)); @@ -17838,12 +17854,6 @@ static void rope_yarn( *sin_theta = sinf(theta) * mscale; } -// Apparently solving `n_rot = 2pi * x * base^((2 * max_pos_emb) / n_dims)` for x, we get -// `corr_dim(n_rot) = n_dims * log(max_pos_emb / (n_rot * 2pi)) / (2 * log(base))` -static float ggml_rope_yarn_corr_dim(int n_dims, int n_ctx_orig, float n_rot, float base) { - return n_dims * logf(n_ctx_orig / (n_rot * 2 * (float)M_PI)) / (2 * logf(base)); -} - static void ggml_rope_cache_init( float theta_base, float freq_scale, const float * freq_factors, float corr_dims[2], int64_t ne0, float ext_factor, float mscale, float * cache, float sin_sign, float theta_scale) { @@ -17860,14 +17870,62 @@ static void ggml_rope_cache_init( } } -GGML_CALL void ggml_rope_yarn_corr_dims( - int n_dims, int n_ctx_orig, float freq_base, float beta_fast, float beta_slow, float dims[2] -) { - // start and end correction dims - float start = floorf(ggml_rope_yarn_corr_dim(n_dims, n_ctx_orig, beta_fast, freq_base)); - float end = ceilf(ggml_rope_yarn_corr_dim(n_dims, n_ctx_orig, beta_slow, freq_base)); - dims[0] = MAX(0, start); - dims[1] = MIN(n_dims - 1, end); +static void ggml_mrope_cache_init( + float theta_base_t, float theta_base_h, float theta_base_w, float theta_base_e, int sections[4], bool indep_sects, + float freq_scale, const float * freq_factors, float corr_dims[2], int64_t ne0, float ext_factor, float mscale, + float * cache, float sin_sign, float theta_scale) { + // ref: https://github.com/jquesnelle/yarn/blob/master/scaled_rope/LlamaYaRNScaledRotaryEmbedding.py + float theta_t = theta_base_t; + float theta_h = theta_base_h; + float theta_w = theta_base_w; + float theta_e = theta_base_e; // extra position id for vision encoder + int sect_dims = sections[0] + sections[1] + sections[2] + sections[3]; + int sec_w = sections[1] + sections[0]; + int sec_e = sections[2] + sec_w; + GGML_ASSERT(sect_dims <= ne0); + + for (int64_t i0 = 0; i0 < ne0; i0 += 2) { + const float ff = freq_factors ? freq_factors[i0/2] : 1.0f; + + int sector = (i0 / 2) % sect_dims; + if (indep_sects) { + // compute theta independently for each dim sections + // (i.e. reset corresponding theta when `i0` go from one section to another) + if (sector == 0) { + theta_t = theta_base_t; + } + else if (sector == sections[0]) { + theta_h = theta_base_h;; + } + else if (sector == sec_w) { + theta_w = theta_base_w; + } + else if (sector == sec_e) { + theta_e = theta_base_e; + } + } + + float theta = theta_t; + if (sector >= sections[0] && sector < sec_w) { + theta = theta_h; + } + else if (sector >= sec_w && sector < sec_w + sections[2]) { + theta = theta_w; + } + else if (sector >= sec_w + sections[2]) { + theta = theta_e; + } + + rope_yarn( + theta/ff, freq_scale, corr_dims, i0, ext_factor, mscale, &cache[i0 + 0], &cache[i0 + 1] + ); + cache[i0 + 1] *= sin_sign; + + theta_t *= theta_scale; + theta_w *= theta_scale; + theta_h *= theta_scale; + theta_e *= theta_scale; + } } static void ggml_compute_forward_rope_f32( @@ -17880,6 +17938,7 @@ static void ggml_compute_forward_rope_f32( const struct ggml_tensor * src2 = dst->src[2]; float freq_base, freq_scale, ext_factor, attn_factor, beta_fast, beta_slow; + int sections[4]; //const int n_past = ((int32_t *) dst->op_params)[0]; const int n_dims = ((int32_t *) dst->op_params)[1]; @@ -17893,6 +17952,7 @@ static void ggml_compute_forward_rope_f32( memcpy(&attn_factor, (int32_t *) dst->op_params + 8, sizeof(float)); memcpy(&beta_fast, (int32_t *) dst->op_params + 9, sizeof(float)); memcpy(&beta_slow, (int32_t *) dst->op_params + 10, sizeof(float)); + memcpy(§ions, (int32_t *) dst->op_params + 11, sizeof(int)*4); GGML_TENSOR_UNARY_OP_LOCALS @@ -17924,7 +17984,17 @@ static void ggml_compute_forward_rope_f32( float corr_dims[2]; ggml_rope_yarn_corr_dims(n_dims, n_ctx_orig, freq_base, beta_fast, beta_slow, corr_dims); - const bool is_neox = mode & 2; + const bool is_neox = mode & GGML_ROPE_TYPE_NEOX; + const bool is_mrope = mode & GGML_ROPE_TYPE_MROPE; // ggml_rope_multi, multimodal rotary position embedding + const bool is_vision = mode == GGML_ROPE_TYPE_VISION; + + if (is_mrope) { + GGML_ASSERT(sections[0] > 0 || sections[1] > 0 || sections[2] > 0); + } + + if (is_vision) { + GGML_ASSERT(n_dims == ne0/2); + } const float * freq_factors = NULL; if (src2 != NULL) { @@ -17940,18 +18010,63 @@ static void ggml_compute_forward_rope_f32( const int32_t * pos = (const int32_t *) src1->data; - for (int64_t i3 = 0; i3 < ne3; i3++) { - for (int64_t i2 = 0; i2 < ne2; i2++) { - const int64_t p = pos[i2]; + for (int64_t i3 = 0; i3 < ne3; i3++) { // batch + for (int64_t i2 = 0; i2 < ne2; i2++) { // seq-len float * cache = (float *) params->wdata + (ne0 + CACHE_LINE_SIZE_F32)*ith; - ggml_rope_cache_init(p, freq_scale, freq_factors, corr_dims, ne0, ext_factor, attn_factor, cache, sin_sign, theta_scale); + if (!is_mrope) { + const int64_t p = pos[i2]; + ggml_rope_cache_init(p, freq_scale, freq_factors, corr_dims, ne0, ext_factor, attn_factor, cache, sin_sign, theta_scale); + } + else { + const int64_t p_t = pos[i2]; + const int64_t p_h = pos[i2 + ne2]; + const int64_t p_w = pos[i2 + ne2 * 2]; + const int64_t p_e = pos[i2 + ne2 * 3]; + ggml_mrope_cache_init( + p_t, p_h, p_w, p_e, sections, is_vision, + freq_scale, freq_factors, corr_dims, ne0, ext_factor, attn_factor, cache, sin_sign, theta_scale); + } - for (int64_t i1 = 0; i1 < ne1; i1++) { + for (int64_t i1 = 0; i1 < ne1; i1++) { // attn-heads if (ir++ < ir0) continue; if (ir > ir1) break; - if (!is_neox) { + if (is_neox || is_mrope) { + if (is_vision){ + for (int64_t i0 = 0; i0 < n_dims; i0 += 2) { + const int64_t ic = i0/2; + + const float cos_theta = cache[i0 + 0]; + const float sin_theta = cache[i0 + 1]; + + const float * const src = (float *)((char *) src0->data + i3*nb03 + i2*nb02 + i1*nb01 + ic*nb00); + float * dst_data = (float *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + ic*nb0); + + const float x0 = src[0]; + const float x1 = src[n_dims]; + + dst_data[0] = x0*cos_theta - x1*sin_theta; + dst_data[n_dims] = x0*sin_theta + x1*cos_theta; + } + } else { + for (int64_t i0 = 0; i0 < n_dims; i0 += 2) { + const int64_t ic = i0/2; + + const float cos_theta = cache[i0 + 0]; + const float sin_theta = cache[i0 + 1]; + + const float * const src = (float *)((char *) src0->data + i3*nb03 + i2*nb02 + i1*nb01 + ic*nb00); + float * dst_data = (float *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + ic*nb0); + + const float x0 = src[0]; + const float x1 = src[n_dims/2]; + + dst_data[0] = x0*cos_theta - x1*sin_theta; + dst_data[n_dims/2] = x0*sin_theta + x1*cos_theta; + } + } + } else { for (int64_t i0 = 0; i0 < n_dims; i0 += 2) { const float cos_theta = cache[i0 + 0]; const float sin_theta = cache[i0 + 1]; @@ -17965,8 +18080,10 @@ static void ggml_compute_forward_rope_f32( dst_data[0] = x0*cos_theta - x1*sin_theta; dst_data[1] = x0*sin_theta + x1*cos_theta; } - } else { - for (int64_t i0 = 0; i0 < n_dims; i0 += 2) { + } + + if (is_vision) { + for (int64_t i0 = n_dims; i0 < ne0; i0 += 2) { const int64_t ic = i0/2; const float cos_theta = cache[i0 + 0]; @@ -17976,19 +18093,20 @@ static void ggml_compute_forward_rope_f32( float * dst_data = (float *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + ic*nb0); const float x0 = src[0]; - const float x1 = src[n_dims/2]; + const float x1 = src[n_dims]; - dst_data[0] = x0*cos_theta - x1*sin_theta; - dst_data[n_dims/2] = x0*sin_theta + x1*cos_theta; + dst_data[0] = x0*cos_theta - x1*sin_theta; + dst_data[n_dims] = x0*sin_theta + x1*cos_theta; } - } + } else { + // fill the remain channels with data from src tensor + for (int64_t i0 = n_dims; i0 < ne0; i0 += 2) { + const float * const src = (float *)((char *) src0->data + i3*nb03 + i2*nb02 + i1*nb01 + i0*nb00); + float * dst_data = (float *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + i0*nb0); - for (int64_t i0 = n_dims; i0 < ne0; i0 += 2) { - const float * const src = (float *)((char *) src0->data + i3*nb03 + i2*nb02 + i1*nb01 + i0*nb00); - float * dst_data = (float *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + i0*nb0); - - dst_data[0] = src[0]; - dst_data[1] = src[1]; + dst_data[0] = src[0]; + dst_data[1] = src[1]; + } } } } @@ -18006,6 +18124,7 @@ static void ggml_compute_forward_rope_f16( const struct ggml_tensor * src2 = dst->src[2]; float freq_base, freq_scale, ext_factor, attn_factor, beta_fast, beta_slow; + int sections[4]; //const int n_past = ((int32_t *) dst->op_params)[0]; const int n_dims = ((int32_t *) dst->op_params)[1]; @@ -18018,6 +18137,8 @@ static void ggml_compute_forward_rope_f16( memcpy(&attn_factor, (int32_t *) dst->op_params + 8, sizeof(float)); memcpy(&beta_fast, (int32_t *) dst->op_params + 9, sizeof(float)); memcpy(&beta_slow, (int32_t *) dst->op_params + 10, sizeof(float)); + memcpy(§ions, (int32_t *) dst->op_params + 11, sizeof(int)*4); + GGML_TENSOR_UNARY_OP_LOCALS @@ -18049,7 +18170,17 @@ static void ggml_compute_forward_rope_f16( float corr_dims[2]; ggml_rope_yarn_corr_dims(n_dims, n_ctx_orig, freq_base, beta_fast, beta_slow, corr_dims); - const bool is_neox = mode & 2; + const bool is_neox = mode & GGML_ROPE_TYPE_NEOX; + const bool is_mrope = mode & GGML_ROPE_TYPE_MROPE; + const bool is_vision = mode == GGML_ROPE_TYPE_VISION; + + if (is_mrope) { + GGML_ASSERT(sections[0] > 0 || sections[1] > 0 || sections[2] > 0); + } + + if (is_vision) { + GGML_ASSERT(n_dims == ne0/2); + } const float * freq_factors = NULL; if (src2 != NULL) { @@ -18067,16 +18198,61 @@ static void ggml_compute_forward_rope_f16( for (int64_t i3 = 0; i3 < ne3; i3++) { for (int64_t i2 = 0; i2 < ne2; i2++) { - const int64_t p = pos[i2]; float * cache = (float *) params->wdata + (ne0 + CACHE_LINE_SIZE_F32)*ith; - ggml_rope_cache_init(p, freq_scale, freq_factors, corr_dims, ne0, ext_factor, attn_factor, cache, sin_sign, theta_scale); + if (!is_mrope) { + const int64_t p = pos[i2]; + ggml_rope_cache_init(p, freq_scale, freq_factors, corr_dims, ne0, ext_factor, attn_factor, cache, sin_sign, theta_scale); + } + else { + const int64_t p_t = pos[i2]; + const int64_t p_h = pos[i2 + ne2]; + const int64_t p_w = pos[i2 + ne2 * 2]; + const int64_t p_e = pos[i2 + ne2 * 3]; + ggml_mrope_cache_init( + p_t, p_h, p_w, p_e, sections, is_vision, + freq_scale, freq_factors, corr_dims, ne0, ext_factor, attn_factor, cache, sin_sign, theta_scale); + } for (int64_t i1 = 0; i1 < ne1; i1++) { if (ir++ < ir0) continue; if (ir > ir1) break; - if (!is_neox) { + if (is_neox || is_mrope) { + if (is_vision) { + for (int64_t i0 = 0; i0 < n_dims; i0 += 2) { + const int64_t ic = i0/2; + + const float cos_theta = cache[i0 + 0]; + const float sin_theta = cache[i0 + 1]; + + const ggml_fp16_t * const src = (ggml_fp16_t *)((char *) src0->data + i3*nb03 + i2*nb02 + i1*nb01 + ic*nb00); + ggml_fp16_t * dst_data = (ggml_fp16_t *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + ic*nb0); + + const float x0 = GGML_FP16_TO_FP32(src[0]); + const float x1 = GGML_FP16_TO_FP32(src[n_dims]); + + dst_data[0] = GGML_FP32_TO_FP16(x0*cos_theta - x1*sin_theta); + dst_data[n_dims] = GGML_FP32_TO_FP16(x0*sin_theta + x1*cos_theta); + } + } else { + for (int64_t i0 = 0; i0 < n_dims; i0 += 2) { + const int64_t ic = i0/2; + + const float cos_theta = cache[i0 + 0]; + const float sin_theta = cache[i0 + 1]; + + const ggml_fp16_t * const src = (ggml_fp16_t *)((char *) src0->data + i3*nb03 + i2*nb02 + i1*nb01 + ic*nb00); + ggml_fp16_t * dst_data = (ggml_fp16_t *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + ic*nb0); + + const float x0 = GGML_FP16_TO_FP32(src[0]); + const float x1 = GGML_FP16_TO_FP32(src[n_dims/2]); + + dst_data[0] = GGML_FP32_TO_FP16(x0*cos_theta - x1*sin_theta); + dst_data[n_dims/2] = GGML_FP32_TO_FP16(x0*sin_theta + x1*cos_theta); + } + } + } else { for (int64_t i0 = 0; i0 < n_dims; i0 += 2) { const float cos_theta = cache[i0 + 0]; const float sin_theta = cache[i0 + 1]; @@ -18090,8 +18266,10 @@ static void ggml_compute_forward_rope_f16( dst_data[0] = GGML_FP32_TO_FP16(x0*cos_theta - x1*sin_theta); dst_data[1] = GGML_FP32_TO_FP16(x0*sin_theta + x1*cos_theta); } - } else { - for (int64_t i0 = 0; i0 < n_dims; i0 += 2) { + } + + if (is_vision) { + for (int64_t i0 = n_dims; i0 < ne0; i0 += 2) { const int64_t ic = i0/2; const float cos_theta = cache[i0 + 0]; @@ -18101,19 +18279,19 @@ static void ggml_compute_forward_rope_f16( ggml_fp16_t * dst_data = (ggml_fp16_t *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + ic*nb0); const float x0 = GGML_FP16_TO_FP32(src[0]); - const float x1 = GGML_FP16_TO_FP32(src[n_dims/2]); + const float x1 = GGML_FP16_TO_FP32(src[n_dims]); - dst_data[0] = GGML_FP32_TO_FP16(x0*cos_theta - x1*sin_theta); - dst_data[n_dims/2] = GGML_FP32_TO_FP16(x0*sin_theta + x1*cos_theta); + dst_data[0] = GGML_FP32_TO_FP16(x0*cos_theta - x1*sin_theta); + dst_data[n_dims] = GGML_FP32_TO_FP16(x0*sin_theta + x1*cos_theta); } - } + } else { + for (int64_t i0 = n_dims; i0 < ne0; i0 += 2) { + const ggml_fp16_t * const src = (ggml_fp16_t *)((char *) src0->data + i3*nb03 + i2*nb02 + i1*nb01 + i0*nb00); + ggml_fp16_t * dst_data = (ggml_fp16_t *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + i0*nb0); - for (int64_t i0 = n_dims; i0 < ne0; i0 += 2) { - const ggml_fp16_t * const src = (ggml_fp16_t *)((char *) src0->data + i3*nb03 + i2*nb02 + i1*nb01 + i0*nb00); - ggml_fp16_t * dst_data = (ggml_fp16_t *)((char *) dst->data + i3*nb3 + i2*nb2 + i1*nb1 + i0*nb0); - - dst_data[0] = src[0]; - dst_data[1] = src[1]; + dst_data[0] = src[0]; + dst_data[1] = src[1]; + } } } }