root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-25 10:54:36 +00:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

Merge a3aea0801c into 60cfa728e2

Browse Source
This commit is contained in:
Eve 2024-12-24 07:58:03 +01:00 committed by GitHub
commit f69ffe9dc7
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
6 changed files with 472 additions and 367 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

81
ggml/src/ggml-vulkan/ggml-vulkan.cpp
View File

@ -1855,53 +1855,58 @@ static void ggml_vk_load_shaders(vk_device& device) {
// mul mat vec // mul mat vec
// AMD GCN and Intel graphics cards perform best when the number of rows per shader is doubled // the number of rows computed per shader depends on GPU model and quant
uint32_t rm = 1; uint32_t rm_stdq = 1;
if ((device->vendor_id == VK_VENDOR_ID_AMD && device->subgroup_min_size == 64 && device->subgroup_max_size == 64) || device->vendor_id == VK_VENDOR_ID_INTEL) uint32_t rm_kq = 2;
rm = 2; if (device->vendor_id == VK_VENDOR_ID_AMD) {
if (device->subgroup_min_size == 64 && device->subgroup_max_size == 64) { // GCN
rm_stdq = 2;
rm_kq = 4;
}
} else if (device->vendor_id == VK_VENDOR_ID_INTEL)
rm_stdq = 2;
// computing additional rows per workgroup is a benefit for Q4_0 -> Q5_1, but not for Q8_0.
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_F32 ], "mul_mat_vec_f32_f32_f32", mul_mat_vec_f32_f32_f32_len, mul_mat_vec_f32_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_F32 ], "mul_mat_vec_f32_f32_f32", mul_mat_vec_f32_f32_f32_len, mul_mat_vec_f32_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_F16 ], "mul_mat_vec_f16_f32_f32", mul_mat_vec_f16_f32_f32_len, mul_mat_vec_f16_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_F16 ], "mul_mat_vec_f16_f32_f32", mul_mat_vec_f16_f32_f32_len, mul_mat_vec_f16_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q4_0], "mul_mat_vec_q4_0_f32_f32", mul_mat_vec_q4_0_f32_f32_len, mul_mat_vec_q4_0_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q4_0], "mul_mat_vec_q4_0_f32_f32", mul_mat_vec_q4_0_f32_f32_len, mul_mat_vec_q4_0_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q4_1], "mul_mat_vec_q4_1_f32_f32", mul_mat_vec_q4_1_f32_f32_len, mul_mat_vec_q4_1_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q4_1], "mul_mat_vec_q4_1_f32_f32", mul_mat_vec_q4_1_f32_f32_len, mul_mat_vec_q4_1_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q5_0], "mul_mat_vec_q5_0_f32_f32", mul_mat_vec_q5_0_f32_f32_len, mul_mat_vec_q5_0_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q5_0], "mul_mat_vec_q5_0_f32_f32", mul_mat_vec_q5_0_f32_f32_len, mul_mat_vec_q5_0_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q5_1], "mul_mat_vec_q5_1_f32_f32", mul_mat_vec_q5_1_f32_f32_len, mul_mat_vec_q5_1_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q5_1], "mul_mat_vec_q5_1_f32_f32", mul_mat_vec_q5_1_f32_f32_len, mul_mat_vec_q5_1_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q8_0], "mul_mat_vec_q8_0_f32_f32", mul_mat_vec_q8_0_f32_f32_len, mul_mat_vec_q8_0_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1*rm, 1, 1}, {device->subgroup_size, 1*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q8_0], "mul_mat_vec_q8_0_f32_f32", mul_mat_vec_q8_0_f32_f32_len, mul_mat_vec_q8_0_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1*rm_stdq, 1, 1}, {device->subgroup_size, 1*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q2_K], "mul_mat_vec_q2_k_f32_f32", mul_mat_vec_q2_k_f32_f32_len, mul_mat_vec_q2_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q2_K], "mul_mat_vec_q2_k_f32_f32", mul_mat_vec_q2_k_f32_f32_len, mul_mat_vec_q2_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q3_K], "mul_mat_vec_q3_k_f32_f32", mul_mat_vec_q3_k_f32_f32_len, mul_mat_vec_q3_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q3_K], "mul_mat_vec_q3_k_f32_f32", mul_mat_vec_q3_k_f32_f32_len, mul_mat_vec_q3_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q4_K], "mul_mat_vec_q4_k_f32_f32", mul_mat_vec_q4_k_f32_f32_len, mul_mat_vec_q4_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q4_K], "mul_mat_vec_q4_k_f32_f32", mul_mat_vec_q4_k_f32_f32_len, mul_mat_vec_q4_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q5_K], "mul_mat_vec_q5_k_f32_f32", mul_mat_vec_q5_k_f32_f32_len, mul_mat_vec_q5_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q5_K], "mul_mat_vec_q5_k_f32_f32", mul_mat_vec_q5_k_f32_f32_len, mul_mat_vec_q5_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q6_K], "mul_mat_vec_q6_k_f32_f32", mul_mat_vec_q6_k_f32_f32_len, mul_mat_vec_q6_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q6_K], "mul_mat_vec_q6_k_f32_f32", mul_mat_vec_q6_k_f32_f32_len, mul_mat_vec_q6_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_IQ4_NL], "mul_mat_vec_iq4_nl_f32_f32", mul_mat_vec_iq4_nl_f32_f32_len, mul_mat_vec_iq4_nl_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {subgroup_size_16, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_IQ4_NL], "mul_mat_vec_iq4_nl_f32_f32", mul_mat_vec_iq4_nl_f32_f32_len, mul_mat_vec_iq4_nl_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {subgroup_size_16, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_F32 ], "mul_mat_vec_f32_f16_f32", mul_mat_vec_f32_f16_f32_len, mul_mat_vec_f32_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_F32 ], "mul_mat_vec_f32_f16_f32", mul_mat_vec_f32_f16_f32_len, mul_mat_vec_f32_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_F16 ], "mul_mat_vec_f16_f16_f32", mul_mat_vec_f16_f16_f32_len, mul_mat_vec_f16_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_F16 ], "mul_mat_vec_f16_f16_f32", mul_mat_vec_f16_f16_f32_len, mul_mat_vec_f16_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q4_0], "mul_mat_vec_q4_0_f16_f32", mul_mat_vec_q4_0_f16_f32_len, mul_mat_vec_q4_0_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q4_0], "mul_mat_vec_q4_0_f16_f32", mul_mat_vec_q4_0_f16_f32_len, mul_mat_vec_q4_0_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q4_1], "mul_mat_vec_q4_1_f16_f32", mul_mat_vec_q4_1_f16_f32_len, mul_mat_vec_q4_1_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q4_1], "mul_mat_vec_q4_1_f16_f32", mul_mat_vec_q4_1_f16_f32_len, mul_mat_vec_q4_1_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q5_0], "mul_mat_vec_q5_0_f16_f32", mul_mat_vec_q5_0_f16_f32_len, mul_mat_vec_q5_0_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q5_0], "mul_mat_vec_q5_0_f16_f32", mul_mat_vec_q5_0_f16_f32_len, mul_mat_vec_q5_0_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q5_1], "mul_mat_vec_q5_1_f16_f32", mul_mat_vec_q5_1_f16_f32_len, mul_mat_vec_q5_1_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q5_1], "mul_mat_vec_q5_1_f16_f32", mul_mat_vec_q5_1_f16_f32_len, mul_mat_vec_q5_1_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q8_0], "mul_mat_vec_q8_0_f16_f32", mul_mat_vec_q8_0_f16_f32_len, mul_mat_vec_q8_0_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1*rm, 1, 1}, {device->subgroup_size, 1*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q8_0], "mul_mat_vec_q8_0_f16_f32", mul_mat_vec_q8_0_f16_f32_len, mul_mat_vec_q8_0_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1*rm_stdq, 1, 1}, {device->subgroup_size, 1*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q2_K], "mul_mat_vec_q2_k_f16_f32", mul_mat_vec_q2_k_f16_f32_len, mul_mat_vec_q2_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q2_K], "mul_mat_vec_q2_k_f16_f32", mul_mat_vec_q2_k_f16_f32_len, mul_mat_vec_q2_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q3_K], "mul_mat_vec_q3_k_f16_f32", mul_mat_vec_q3_k_f16_f32_len, mul_mat_vec_q3_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q3_K], "mul_mat_vec_q3_k_f16_f32", mul_mat_vec_q3_k_f16_f32_len, mul_mat_vec_q3_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q4_K], "mul_mat_vec_q4_k_f16_f32", mul_mat_vec_q4_k_f16_f32_len, mul_mat_vec_q4_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q4_K], "mul_mat_vec_q4_k_f16_f32", mul_mat_vec_q4_k_f16_f32_len, mul_mat_vec_q4_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q5_K], "mul_mat_vec_q5_k_f16_f32", mul_mat_vec_q5_k_f16_f32_len, mul_mat_vec_q5_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q5_K], "mul_mat_vec_q5_k_f16_f32", mul_mat_vec_q5_k_f16_f32_len, mul_mat_vec_q5_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q6_K], "mul_mat_vec_q6_k_f16_f32", mul_mat_vec_q6_k_f16_f32_len, mul_mat_vec_q6_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q6_K], "mul_mat_vec_q6_k_f16_f32", mul_mat_vec_q6_k_f16_f32_len, mul_mat_vec_q6_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_IQ4_NL], "mul_mat_vec_iq4_nl_f16_f32", mul_mat_vec_iq4_nl_f16_f32_len, mul_mat_vec_iq4_nl_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm, 1, 1}, {subgroup_size_16, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_IQ4_NL], "mul_mat_vec_iq4_nl_f16_f32", mul_mat_vec_iq4_nl_f16_f32_len, mul_mat_vec_iq4_nl_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2*rm_stdq, 1, 1}, {subgroup_size_16, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_F32 ], "mul_mat_vec_id_f32_f32", mul_mat_vec_id_f32_f32_len, mul_mat_vec_id_f32_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_F32 ], "mul_mat_vec_id_f32_f32", mul_mat_vec_id_f32_f32_len, mul_mat_vec_id_f32_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_F16 ], "mul_mat_vec_id_f16_f32", mul_mat_vec_id_f16_f32_len, mul_mat_vec_id_f16_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_F16 ], "mul_mat_vec_id_f16_f32", mul_mat_vec_id_f16_f32_len, mul_mat_vec_id_f16_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q4_0], "mul_mat_vec_id_q4_0_f32", mul_mat_vec_id_q4_0_f32_len, mul_mat_vec_id_q4_0_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q4_0], "mul_mat_vec_id_q4_0_f32", mul_mat_vec_id_q4_0_f32_len, mul_mat_vec_id_q4_0_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q4_1], "mul_mat_vec_id_q4_1_f32", mul_mat_vec_id_q4_1_f32_len, mul_mat_vec_id_q4_1_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q4_1], "mul_mat_vec_id_q4_1_f32", mul_mat_vec_id_q4_1_f32_len, mul_mat_vec_id_q4_1_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q5_0], "mul_mat_vec_id_q5_0_f32", mul_mat_vec_id_q5_0_f32_len, mul_mat_vec_id_q5_0_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q5_0], "mul_mat_vec_id_q5_0_f32", mul_mat_vec_id_q5_0_f32_len, mul_mat_vec_id_q5_0_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q5_1], "mul_mat_vec_id_q5_1_f32", mul_mat_vec_id_q5_1_f32_len, mul_mat_vec_id_q5_1_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm, 1, 1}, {device->subgroup_size, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q5_1], "mul_mat_vec_id_q5_1_f32", mul_mat_vec_id_q5_1_f32_len, mul_mat_vec_id_q5_1_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm_stdq, 1, 1}, {device->subgroup_size, 2*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q8_0], "mul_mat_vec_id_q8_0_f32", mul_mat_vec_id_q8_0_f32_len, mul_mat_vec_id_q8_0_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1*rm, 1, 1}, {device->subgroup_size, 1*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q8_0], "mul_mat_vec_id_q8_0_f32", mul_mat_vec_id_q8_0_f32_len, mul_mat_vec_id_q8_0_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1*rm_stdq, 1, 1}, {device->subgroup_size, 1*rm_stdq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q2_K], "mul_mat_vec_id_q2_k_f32", mul_mat_vec_id_q2_k_f32_len, mul_mat_vec_id_q2_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q2_K], "mul_mat_vec_id_q2_k_f32", mul_mat_vec_id_q2_k_f32_len, mul_mat_vec_id_q2_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q3_K], "mul_mat_vec_id_q3_k_f32", mul_mat_vec_id_q3_k_f32_len, mul_mat_vec_id_q3_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q3_K], "mul_mat_vec_id_q3_k_f32", mul_mat_vec_id_q3_k_f32_len, mul_mat_vec_id_q3_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q4_K], "mul_mat_vec_id_q4_k_f32", mul_mat_vec_id_q4_k_f32_len, mul_mat_vec_id_q4_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q4_K], "mul_mat_vec_id_q4_k_f32", mul_mat_vec_id_q4_k_f32_len, mul_mat_vec_id_q4_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q5_K], "mul_mat_vec_id_q5_k_f32", mul_mat_vec_id_q5_k_f32_len, mul_mat_vec_id_q5_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q5_K], "mul_mat_vec_id_q5_k_f32", mul_mat_vec_id_q5_k_f32_len, mul_mat_vec_id_q5_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q6_K], "mul_mat_vec_id_q6_k_f32", mul_mat_vec_id_q6_k_f32_len, mul_mat_vec_id_q6_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q6_K], "mul_mat_vec_id_q6_k_f32", mul_mat_vec_id_q6_k_f32_len, mul_mat_vec_id_q6_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {rm_kq, 1, 1}, {subgroup_size_16, rm_kq}, 1, true);
ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_IQ4_NL], "mul_mat_vec_id_iq4_nl_f32", mul_mat_vec_id_iq4_nl_f32_len, mul_mat_vec_id_iq4_nl_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm, 1, 1}, {subgroup_size_16, 2*rm}, 1, true); ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_IQ4_NL], "mul_mat_vec_id_iq4_nl_f32", mul_mat_vec_id_iq4_nl_f32_len, mul_mat_vec_id_iq4_nl_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2*rm_stdq, 1, 1}, {subgroup_size_16, 2*rm_stdq}, 1, true);
// dequant shaders // dequant shaders
ggml_vk_create_pipeline(device, device->pipeline_dequant[GGML_TYPE_F32 ], "f32_to_f16", dequant_f32_len, dequant_f32_data, "main", 2, 5 * sizeof(uint32_t), {256 * 16, 1, 1}, {}, 1); ggml_vk_create_pipeline(device, device->pipeline_dequant[GGML_TYPE_F32 ], "f32_to_f16", dequant_f32_len, dequant_f32_data, "main", 2, 5 * sizeof(uint32_t), {256 * 16, 1, 1}, {}, 1);

126
ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q2_k.comp
View File

@ -6,21 +6,15 @@
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in; layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (constant_id = 0) const uint BLOCK_SIZE = 32; layout (constant_id = 0) const uint BLOCK_SIZE = 32;
layout (constant_id = 1) const uint NUM_ROWS = 1;
shared FLOAT_TYPE tmp[BLOCK_SIZE]; shared FLOAT_TYPE tmpsh[NUM_ROWS][BLOCK_SIZE];
void main() {
const uint row = gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z;
if (row >= p.stride_d) {
return;
}
void compute_outputs(const uint32_t first_row, const uint32_t num_rows) {
uint a_offset, b_offset, d_offset; uint a_offset, b_offset, d_offset;
get_offsets(a_offset, b_offset, d_offset); get_offsets(a_offset, b_offset, d_offset);
const uint num_blocks_per_row = p.ncols / QUANT_K; const uint num_blocks_per_row = p.ncols / QUANT_K;
const uint ib0 = a_offset / QUANT_K + row*num_blocks_per_row;
// 16 threads are used to process each block // 16 threads are used to process each block
const uint it_size = gl_WorkGroupSize.x/16; const uint it_size = gl_WorkGroupSize.x/16;
@ -38,15 +32,15 @@ void main() {
const uint s_offset = 8*v_im; const uint s_offset = 8*v_im;
const uint y_offset = 128*v_im + l0; const uint y_offset = 128*v_im + l0;
FLOAT_TYPE temp = FLOAT_TYPE(0.0); // partial sum for thread in warp FLOAT_TYPE temp[NUM_ROWS];
[[unroll]] for (uint i = 0; i < NUM_ROWS; ++i) {
temp[i] = FLOAT_TYPE(0);
}
[[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) { [[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) {
const uint y_idx = i * QUANT_K + y_offset; const uint y_idx = i * QUANT_K + y_offset;
f16vec2 d = data_a[ib0 + i].d;
const FLOAT_TYPE dall = d.x;
const FLOAT_TYPE dmin = d.y;
B_TYPE_VEC2 b0 = data_b_v2[(b_offset + y_idx) / 2 + 0]; B_TYPE_VEC2 b0 = data_b_v2[(b_offset + y_idx) / 2 + 0];
B_TYPE_VEC2 b16 = data_b_v2[(b_offset + y_idx) / 2 + 8]; B_TYPE_VEC2 b16 = data_b_v2[(b_offset + y_idx) / 2 + 8];
B_TYPE_VEC2 b32 = data_b_v2[(b_offset + y_idx) / 2 + 16]; B_TYPE_VEC2 b32 = data_b_v2[(b_offset + y_idx) / 2 + 16];
@ -56,58 +50,84 @@ void main() {
B_TYPE_VEC2 b96 = data_b_v2[(b_offset + y_idx) / 2 + 48]; B_TYPE_VEC2 b96 = data_b_v2[(b_offset + y_idx) / 2 + 48];
B_TYPE_VEC2 b112 = data_b_v2[(b_offset + y_idx) / 2 + 56]; B_TYPE_VEC2 b112 = data_b_v2[(b_offset + y_idx) / 2 + 56];
uint32_t s0_u32 = data_a_packed32[ib0 + i].scales[s_offset / 4 + 0]; [[unroll]] for (uint n = 0; n < num_rows; ++n) {
uint32_t s4_u32 = data_a_packed32[ib0 + i].scales[s_offset / 4 + 1]; const uint ib0 = a_offset / QUANT_K + (first_row+n)*num_blocks_per_row;
f16vec2 d = data_a[ib0 + i].d;
const FLOAT_TYPE dall = d.x;
const FLOAT_TYPE dmin = d.y;
uint32_t s0_lo4_u32 = s0_u32 & 0x0F0F0F0F; uint32_t s0_u32 = data_a_packed32[ib0 + i].scales[s_offset / 4 + 0];
uint32_t s0_hi4_u32 = (s0_u32 >> 4) & 0x0F0F0F0F; uint32_t s4_u32 = data_a_packed32[ib0 + i].scales[s_offset / 4 + 1];
uint32_t s4_lo4_u32 = s4_u32 & 0x0F0F0F0F;
uint32_t s4_hi4_u32 = (s4_u32 >> 4) & 0x0F0F0F0F;
uvec4 s0_lo4 = uvec4(unpack8(s0_lo4_u32)); uint32_t s0_lo4_u32 = s0_u32 & 0x0F0F0F0F;
uvec4 s4_lo4 = uvec4(unpack8(s4_lo4_u32)); uint32_t s0_hi4_u32 = (s0_u32 >> 4) & 0x0F0F0F0F;
uvec4 s0_hi4 = uvec4(unpack8(s0_hi4_u32)); uint32_t s4_lo4_u32 = s4_u32 & 0x0F0F0F0F;
uvec4 s4_hi4 = uvec4(unpack8(s4_hi4_u32)); uint32_t s4_hi4_u32 = (s4_u32 >> 4) & 0x0F0F0F0F;
uint16_t qs0_u16 = data_a_packed16[ib0 + i].qs[q_offset / 2 + 0]; uvec4 s0_lo4 = uvec4(unpack8(s0_lo4_u32));
uint16_t qs16_u16 = data_a_packed16[ib0 + i].qs[q_offset / 2 + 8]; uvec4 s4_lo4 = uvec4(unpack8(s4_lo4_u32));
uvec2 qs0 = uvec2(unpack8(qs0_u16)); uvec4 s0_hi4 = uvec4(unpack8(s0_hi4_u32));
uvec2 qs16 = uvec2(unpack8(qs16_u16)); uvec4 s4_hi4 = uvec4(unpack8(s4_hi4_u32));
FLOAT_TYPE sum1 = FLOAT_TYPE(0.0); uint16_t qs0_u16 = data_a_packed16[ib0 + i].qs[q_offset / 2 + 0];
FLOAT_TYPE sum2 = FLOAT_TYPE(0.0); uint16_t qs16_u16 = data_a_packed16[ib0 + i].qs[q_offset / 2 + 8];
[[unroll]] for (int l = 0; l < 2; ++l) { uvec2 qs0 = uvec2(unpack8(qs0_u16));
sum1 = fma(FLOAT_TYPE(b0[l]), FLOAT_TYPE(s0_lo4[0]) * FLOAT_TYPE((qs0[l] >> 0) & 3), uvec2 qs16 = uvec2(unpack8(qs16_u16));
fma(FLOAT_TYPE(b16[l]), FLOAT_TYPE(s0_lo4[1]) * FLOAT_TYPE((qs16[l] >> 0) & 3),
fma(FLOAT_TYPE(b32[l]), FLOAT_TYPE(s0_lo4[2]) * FLOAT_TYPE((qs0[l] >> 2) & 3), FLOAT_TYPE sum1 = FLOAT_TYPE(0.0);
fma(FLOAT_TYPE(b48[l]), FLOAT_TYPE(s0_lo4[3]) * FLOAT_TYPE((qs16[l] >> 2) & 3), FLOAT_TYPE sum2 = FLOAT_TYPE(0.0);
fma(FLOAT_TYPE(b64[l]), FLOAT_TYPE(s4_lo4[0]) * FLOAT_TYPE((qs0[l] >> 4) & 3), [[unroll]] for (int l = 0; l < 2; ++l) {
fma(FLOAT_TYPE(b80[l]), FLOAT_TYPE(s4_lo4[1]) * FLOAT_TYPE((qs16[l] >> 4) & 3), sum1 = fma(FLOAT_TYPE(b0[l]), FLOAT_TYPE(s0_lo4[0]) * FLOAT_TYPE((qs0[l] >> 0) & 3),
fma(FLOAT_TYPE(b96[l]), FLOAT_TYPE(s4_lo4[2]) * FLOAT_TYPE((qs0[l] >> 6) & 3), fma(FLOAT_TYPE(b16[l]), FLOAT_TYPE(s0_lo4[1]) * FLOAT_TYPE((qs16[l] >> 0) & 3),
fma(FLOAT_TYPE(b112[l]), FLOAT_TYPE(s4_lo4[3]) * FLOAT_TYPE((qs16[l] >> 6) & 3), sum1)))))))); fma(FLOAT_TYPE(b32[l]), FLOAT_TYPE(s0_lo4[2]) * FLOAT_TYPE((qs0[l] >> 2) & 3),
sum2 = fma(FLOAT_TYPE(b0[l]), FLOAT_TYPE(s0_hi4[0]), fma(FLOAT_TYPE(b48[l]), FLOAT_TYPE(s0_lo4[3]) * FLOAT_TYPE((qs16[l] >> 2) & 3),
fma(FLOAT_TYPE(b16[l]), FLOAT_TYPE(s0_hi4[1]), fma(FLOAT_TYPE(b64[l]), FLOAT_TYPE(s4_lo4[0]) * FLOAT_TYPE((qs0[l] >> 4) & 3),
fma(FLOAT_TYPE(b32[l]), FLOAT_TYPE(s0_hi4[2]), fma(FLOAT_TYPE(b80[l]), FLOAT_TYPE(s4_lo4[1]) * FLOAT_TYPE((qs16[l] >> 4) & 3),
fma(FLOAT_TYPE(b48[l]), FLOAT_TYPE(s0_hi4[3]), fma(FLOAT_TYPE(b96[l]), FLOAT_TYPE(s4_lo4[2]) * FLOAT_TYPE((qs0[l] >> 6) & 3),
fma(FLOAT_TYPE(b64[l]), FLOAT_TYPE(s4_hi4[0]), fma(FLOAT_TYPE(b112[l]), FLOAT_TYPE(s4_lo4[3]) * FLOAT_TYPE((qs16[l] >> 6) & 3), sum1))))))));
fma(FLOAT_TYPE(b80[l]), FLOAT_TYPE(s4_hi4[1]), sum2 = fma(FLOAT_TYPE(b0[l]), FLOAT_TYPE(s0_hi4[0]),
fma(FLOAT_TYPE(b96[l]), FLOAT_TYPE(s4_hi4[2]), fma(FLOAT_TYPE(b16[l]), FLOAT_TYPE(s0_hi4[1]),
fma(FLOAT_TYPE(b112[l]), FLOAT_TYPE(s4_hi4[3]), sum2)))))))); fma(FLOAT_TYPE(b32[l]), FLOAT_TYPE(s0_hi4[2]),
fma(FLOAT_TYPE(b48[l]), FLOAT_TYPE(s0_hi4[3]),
fma(FLOAT_TYPE(b64[l]), FLOAT_TYPE(s4_hi4[0]),
fma(FLOAT_TYPE(b80[l]), FLOAT_TYPE(s4_hi4[1]),
fma(FLOAT_TYPE(b96[l]), FLOAT_TYPE(s4_hi4[2]),
fma(FLOAT_TYPE(b112[l]), FLOAT_TYPE(s4_hi4[3]), sum2))))))));
}
temp[n] = fma(dall, sum1, fma(-dmin, sum2, temp[n]));
} }
temp = fma(dall, sum1, fma(-dmin, sum2, temp));
} }
tmp[gl_LocalInvocationID.x] = temp;
// sum up partial sums and write back result // sum up partial sums and write back result
[[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] = temp[n];
}
barrier(); barrier();
[[unroll]] for (uint s = gl_WorkGroupSize.x/2; s > 0; s >>= 1) { [[unroll]] for (uint s = BLOCK_SIZE/2; s > 0; s >>= 1) {
if (tid < s) { if (tid < s) {
tmp[tid] += tmp[tid + s]; [[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] += tmpsh[n][tid + s];
}
} }
barrier(); barrier();
} }
if (tid == 0) { if (tid == 0) {
data_d[d_offset + row] = D_TYPE(tmp[0]); [[unroll]] for (uint n = 0; n < num_rows; ++n) {
data_d[d_offset + first_row + n] = D_TYPE(tmpsh[n][0]);
}
}
}
void main() {
const uint first_row = NUM_ROWS * (gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z);
// do NUM_ROWS at a time, unless there aren't enough remaining rows
if (first_row + NUM_ROWS <= p.stride_d) {
compute_outputs(first_row, NUM_ROWS);
} else {
if (first_row >= p.stride_d) {
return;
}
compute_outputs(first_row, p.stride_d - first_row);
} }
} }

102
ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q3_k.comp
View File

@ -6,21 +6,15 @@
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in; layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (constant_id = 0) const uint BLOCK_SIZE = 32; layout (constant_id = 0) const uint BLOCK_SIZE = 32;
layout (constant_id = 1) const uint NUM_ROWS = 1;
shared FLOAT_TYPE tmp[BLOCK_SIZE]; shared FLOAT_TYPE tmpsh[NUM_ROWS][BLOCK_SIZE];
void main() {
const uint row = gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z;
if (row >= p.stride_d) {
return;
}
void compute_outputs(const uint32_t first_row, const uint32_t num_rows) {
uint a_offset, b_offset, d_offset; uint a_offset, b_offset, d_offset;
get_offsets(a_offset, b_offset, d_offset); get_offsets(a_offset, b_offset, d_offset);
const uint num_blocks_per_row = p.ncols / QUANT_K; const uint num_blocks_per_row = p.ncols / QUANT_K;
const uint ib0 = a_offset / QUANT_K + row*num_blocks_per_row;
// 16 threads are used to process each block // 16 threads are used to process each block
const uint it_size = gl_WorkGroupSize.x/16; const uint it_size = gl_WorkGroupSize.x/16;
@ -35,19 +29,21 @@ void main() {
const uint8_t m = uint8_t(1 << (4 * v_im)); const uint8_t m = uint8_t(1 << (4 * v_im));
const uint l0 = 2*v_in; // 0...15 const uint l0 = 2*v_in; // 0...15
const uint q_offset = 32*v_im + l0; const uint q_offset = 32*v_im + l0;
const uint y_offset = 128*v_im + l0; const uint y_offset = 128*v_im + l0;
FLOAT_TYPE temp = FLOAT_TYPE(0.0); // partial sum for thread in warp FLOAT_TYPE temp[NUM_ROWS];
[[unroll]] for (uint i = 0; i < NUM_ROWS; ++i) {
temp[i] = FLOAT_TYPE(0);
}
const uint s_shift = 4 * v_im; const uint s_shift = 4 * v_im;
[[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) { [[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) {
const uint y_idx = i * QUANT_K + y_offset; const uint y_idx = i * QUANT_K + y_offset;
const FLOAT_TYPE d = FLOAT_TYPE(data_a[ib0 + i].d);
B_TYPE_VEC2 b0 = data_b_v2[(b_offset + y_idx) / 2 + 0]; B_TYPE_VEC2 b0 = data_b_v2[(b_offset + y_idx) / 2 + 0];
B_TYPE_VEC2 b16 = data_b_v2[(b_offset + y_idx) / 2 + 8]; B_TYPE_VEC2 b16 = data_b_v2[(b_offset + y_idx) / 2 + 8];
B_TYPE_VEC2 b32 = data_b_v2[(b_offset + y_idx) / 2 + 16]; B_TYPE_VEC2 b32 = data_b_v2[(b_offset + y_idx) / 2 + 16];
@ -57,44 +53,68 @@ void main() {
B_TYPE_VEC2 b96 = data_b_v2[(b_offset + y_idx) / 2 + 48]; B_TYPE_VEC2 b96 = data_b_v2[(b_offset + y_idx) / 2 + 48];
B_TYPE_VEC2 b112 = data_b_v2[(b_offset + y_idx) / 2 + 56]; B_TYPE_VEC2 b112 = data_b_v2[(b_offset + y_idx) / 2 + 56];
uint16_t s0_16 = data_a_packed16[ib0 + i].scales[0]; [[unroll]] for (uint n = 0; n < num_rows; ++n) {
uint16_t s2_16 = data_a_packed16[ib0 + i].scales[1]; const uint ib0 = a_offset / QUANT_K + (first_row+n)*num_blocks_per_row;
uint16_t s4_16 = data_a_packed16[ib0 + i].scales[2]; const FLOAT_TYPE d = FLOAT_TYPE(data_a[ib0 + i].d);
uint16_t s6_16 = data_a_packed16[ib0 + i].scales[3];
uint16_t s8_16 = data_a_packed16[ib0 + i].scales[4];
uint16_t s10_16 = data_a_packed16[ib0 + i].scales[5];
u8vec2 s0 = unpack8(s0_16);
u8vec2 s2 = unpack8(s2_16);
u8vec2 s4 = unpack8(s4_16);
u8vec2 s6 = unpack8(s6_16);
u8vec2 s8 = unpack8(s8_16);
u8vec2 s10 = unpack8(s10_16);
FLOAT_TYPE sum = FLOAT_TYPE(0.0); uint16_t s0_16 = data_a_packed16[ib0 + i].scales[0];
[[unroll]] for (int l = 0; l < 2; ++l) { uint16_t s2_16 = data_a_packed16[ib0 + i].scales[1];
sum = fma(FLOAT_TYPE(b0[l]) * FLOAT_TYPE(int8_t(((s0[0] >> s_shift) & 0xF) | ((s8[0] >> (s_shift + 0) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l ] ) & 3) - (((data_a[ib0 + i].hmask[l0 + l ] & (m << 0)) != 0) ? 0 : 4)), uint16_t s4_16 = data_a_packed16[ib0 + i].scales[2];
fma(FLOAT_TYPE(b32[l]) * FLOAT_TYPE(int8_t(((s2[0] >> s_shift) & 0xF) | ((s10[0] >> (s_shift + 0) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l ] >> 2) & 3) - (((data_a[ib0 + i].hmask[l0 + l ] & (m << 1)) != 0) ? 0 : 4)), uint16_t s6_16 = data_a_packed16[ib0 + i].scales[3];
fma(FLOAT_TYPE(b64[l]) * FLOAT_TYPE(int8_t(((s4[0] >> s_shift) & 0xF) | ((s8[0] >> (s_shift + 2) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l ] >> 4) & 3) - (((data_a[ib0 + i].hmask[l0 + l ] & (m << 2)) != 0) ? 0 : 4)), uint16_t s8_16 = data_a_packed16[ib0 + i].scales[4];
fma(FLOAT_TYPE(b96[l]) * FLOAT_TYPE(int8_t(((s6[0] >> s_shift) & 0xF) | ((s10[0] >> (s_shift + 2) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l ] >> 6) & 3) - (((data_a[ib0 + i].hmask[l0 + l ] & (m << 3)) != 0) ? 0 : 4)), uint16_t s10_16 = data_a_packed16[ib0 + i].scales[5];
fma(FLOAT_TYPE(b16[l]) * FLOAT_TYPE(int8_t(((s0[1] >> s_shift) & 0xF) | ((s8[1] >> (s_shift + 0) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l+16] ) & 3) - (((data_a[ib0 + i].hmask[l0 + l+16] & (m << 0)) != 0) ? 0 : 4)), u8vec2 s0 = unpack8(s0_16);
fma(FLOAT_TYPE(b48[l]) * FLOAT_TYPE(int8_t(((s2[1] >> s_shift) & 0xF) | ((s10[1] >> (s_shift + 0) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l+16] >> 2) & 3) - (((data_a[ib0 + i].hmask[l0 + l+16] & (m << 1)) != 0) ? 0 : 4)), u8vec2 s2 = unpack8(s2_16);
fma(FLOAT_TYPE(b80[l]) * FLOAT_TYPE(int8_t(((s4[1] >> s_shift) & 0xF) | ((s8[1] >> (s_shift + 2) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l+16] >> 4) & 3) - (((data_a[ib0 + i].hmask[l0 + l+16] & (m << 2)) != 0) ? 0 : 4)), u8vec2 s4 = unpack8(s4_16);
fma(FLOAT_TYPE(b112[l]) * FLOAT_TYPE(int8_t(((s6[1] >> s_shift) & 0xF) | ((s10[1] >> (s_shift + 2) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l+16] >> 6) & 3) - (((data_a[ib0 + i].hmask[l0 + l+16] & (m << 3)) != 0) ? 0 : 4)), sum)))))))); u8vec2 s6 = unpack8(s6_16);
u8vec2 s8 = unpack8(s8_16);
u8vec2 s10 = unpack8(s10_16);
FLOAT_TYPE sum = FLOAT_TYPE(0.0);
[[unroll]] for (int l = 0; l < 2; ++l) {
sum = fma(FLOAT_TYPE(b0[l]) * FLOAT_TYPE(int8_t(((s0[0] >> s_shift) & 0xF) | ((s8[0] >> (s_shift + 0) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l ] ) & 3) - (((data_a[ib0 + i].hmask[l0 + l ] & (m << 0)) != 0) ? 0 : 4)),
fma(FLOAT_TYPE(b32[l]) * FLOAT_TYPE(int8_t(((s2[0] >> s_shift) & 0xF) | ((s10[0] >> (s_shift + 0) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l ] >> 2) & 3) - (((data_a[ib0 + i].hmask[l0 + l ] & (m << 1)) != 0) ? 0 : 4)),
fma(FLOAT_TYPE(b64[l]) * FLOAT_TYPE(int8_t(((s4[0] >> s_shift) & 0xF) | ((s8[0] >> (s_shift + 2) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l ] >> 4) & 3) - (((data_a[ib0 + i].hmask[l0 + l ] & (m << 2)) != 0) ? 0 : 4)),
fma(FLOAT_TYPE(b96[l]) * FLOAT_TYPE(int8_t(((s6[0] >> s_shift) & 0xF) | ((s10[0] >> (s_shift + 2) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l ] >> 6) & 3) - (((data_a[ib0 + i].hmask[l0 + l ] & (m << 3)) != 0) ? 0 : 4)),
fma(FLOAT_TYPE(b16[l]) * FLOAT_TYPE(int8_t(((s0[1] >> s_shift) & 0xF) | ((s8[1] >> (s_shift + 0) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l+16] ) & 3) - (((data_a[ib0 + i].hmask[l0 + l+16] & (m << 0)) != 0) ? 0 : 4)),
fma(FLOAT_TYPE(b48[l]) * FLOAT_TYPE(int8_t(((s2[1] >> s_shift) & 0xF) | ((s10[1] >> (s_shift + 0) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l+16] >> 2) & 3) - (((data_a[ib0 + i].hmask[l0 + l+16] & (m << 1)) != 0) ? 0 : 4)),
fma(FLOAT_TYPE(b80[l]) * FLOAT_TYPE(int8_t(((s4[1] >> s_shift) & 0xF) | ((s8[1] >> (s_shift + 2) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l+16] >> 4) & 3) - (((data_a[ib0 + i].hmask[l0 + l+16] & (m << 2)) != 0) ? 0 : 4)),
fma(FLOAT_TYPE(b112[l]) * FLOAT_TYPE(int8_t(((s6[1] >> s_shift) & 0xF) | ((s10[1] >> (s_shift + 2) & 0x3) << 4)) - 32), FLOAT_TYPE(((data_a[ib0 + i].qs[q_offset + l+16] >> 6) & 3) - (((data_a[ib0 + i].hmask[l0 + l+16] & (m << 3)) != 0) ? 0 : 4)), sum))))))));
}
temp[n] = fma(d, sum, temp[n]);
} }
temp = fma(d, sum, temp);
} }
tmp[gl_LocalInvocationID.x] = temp;
// sum up partial sums and write back result // sum up partial sums and write back result
[[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] = temp[n];
}
barrier(); barrier();
[[unroll]] for (uint s = gl_WorkGroupSize.x/2; s > 0; s >>= 1) { [[unroll]] for (uint s = BLOCK_SIZE/2; s > 0; s >>= 1) {
if (tid < s) { if (tid < s) {
tmp[tid] += tmp[tid + s]; [[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] += tmpsh[n][tid + s];
}
} }
barrier(); barrier();
} }
if (tid == 0) { if (tid == 0) {
data_d[d_offset + row] = D_TYPE(tmp[0]); [[unroll]] for (uint n = 0; n < num_rows; ++n) {
data_d[d_offset + first_row + n] = D_TYPE(tmpsh[n][0]);
}
}
}
void main() {
const uint first_row = NUM_ROWS * (gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z);
// do NUM_ROWS at a time, unless there aren't enough remaining rows
if (first_row + NUM_ROWS <= p.stride_d) {
compute_outputs(first_row, NUM_ROWS);
} else {
if (first_row >= p.stride_d) {
return;
}
compute_outputs(first_row, p.stride_d - first_row);
} }
} }

174
ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q4_k.comp
View File

@ -7,21 +7,15 @@
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in; layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (constant_id = 0) const uint BLOCK_SIZE = 32; layout (constant_id = 0) const uint BLOCK_SIZE = 32;
layout (constant_id = 1) const uint NUM_ROWS = 1;
shared FLOAT_TYPE tmp[BLOCK_SIZE]; shared FLOAT_TYPE tmpsh[NUM_ROWS][BLOCK_SIZE];
void main() {
const uint row = gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z;
if (row >= p.stride_d) {
return;
}
void compute_outputs(const uint32_t first_row, const uint32_t num_rows) {
uint a_offset, b_offset, d_offset; uint a_offset, b_offset, d_offset;
get_offsets(a_offset, b_offset, d_offset); get_offsets(a_offset, b_offset, d_offset);
const uint num_blocks_per_row = p.ncols / QUANT_K; const uint num_blocks_per_row = p.ncols / QUANT_K;
const uint ib0 = a_offset / QUANT_K + row*num_blocks_per_row;
// 16 threads are used to process each block // 16 threads are used to process each block
const uint it_size = gl_WorkGroupSize.x/16; const uint it_size = gl_WorkGroupSize.x/16;
@ -31,8 +25,8 @@ void main() {
const uint step = 4; const uint step = 4;
const uint il = itid/step; // 0...3 const uint il = itid/step; // 0...3
const uint ir = itid - step*il; // 0...7 or 0...3 const uint ir = itid - step*il; // 0...7 or 0...3
const uint n = 4; const uint n = 4;
const uint v_im = il / 2; // 0 or 1. 0 computes 0,32 + 128,160, 1 computes 64,96 + 192,224 const uint v_im = il / 2; // 0 or 1. 0 computes 0,32 + 128,160, 1 computes 64,96 + 192,224
@ -42,90 +36,116 @@ void main() {
const uint q_offset = 32*v_im + l0; const uint q_offset = 32*v_im + l0;
const uint y_offset = 64*v_im + l0; const uint y_offset = 64*v_im + l0;
FLOAT_TYPE temp = FLOAT_TYPE(0.0); // partial sum for thread in warp FLOAT_TYPE temp[NUM_ROWS];
[[unroll]] for (uint i = 0; i < NUM_ROWS; ++i) {
temp[i] = FLOAT_TYPE(0);
}
[[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) { [[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) {
const uint y1_idx = i * QUANT_K + y_offset; const uint y1_idx = i * QUANT_K + y_offset;
const uint y2_idx = y1_idx + 128; const uint y2_idx = y1_idx + 128;
f16vec2 d = data_a[ib0 + i].d;
const FLOAT_TYPE dall = FLOAT_TYPE(d.x);
const FLOAT_TYPE dmin = FLOAT_TYPE(d.y);
uint32_t scale0_u32 = data_a_packed16[ib0 + i].scales[v_im ];
uint32_t scale4_u32 = data_a_packed16[ib0 + i].scales[v_im + 2];
uint32_t scale8_u32 = data_a_packed16[ib0 + i].scales[v_im + 4];
uvec4 scale0 = uvec4(unpack8(scale0_u32));
uvec4 scale4 = uvec4(unpack8(scale4_u32));
uvec4 scale8 = uvec4(unpack8(scale8_u32));
const uint32_t sc0 = ( scale0.x & 0x3f);
const uint32_t sc1 = ( scale0.y & 0x3f);
const uint32_t sc2 = ( scale4.x & 0x3f);
const uint32_t sc3 = ( scale4.y & 0x3f);
const uint32_t sc4 = (( scale8.x & 0x0f) | ((scale0.x & 0xc0) >> 2));
const uint32_t sc5 = (( scale8.y & 0x0f) | ((scale0.y & 0xc0) >> 2));
const uint32_t sc6 = (((scale8.x >> 4) & 0x0f) | ((scale4.x & 0xc0) >> 2));
const uint32_t sc7 = (((scale8.y >> 4) & 0x0f) | ((scale4.y & 0xc0) >> 2));
uint32_t qs0_u32 = data_a_packed32[ib0 + i].qs[q_offset / 4];
uint32_t qs64_u32 = data_a_packed32[ib0 + i].qs[q_offset / 4 + 16];
uint32_t qs0_u32_lo4 = qs0_u32 & 0x0F0F0F0F;
uint32_t qs0_u32_hi4 = (qs0_u32 >> 4) & 0x0F0F0F0F;
uint32_t qs64_u32_lo4 = qs64_u32 & 0x0F0F0F0F;
uint32_t qs64_u32_hi4 = (qs64_u32 >> 4) & 0x0F0F0F0F;
uvec4 qs0_lo4 = uvec4(unpack8(qs0_u32_lo4));
uvec4 qs64_lo4 = uvec4(unpack8(qs64_u32_lo4));
uvec4 qs0_hi4 = uvec4(unpack8(qs0_u32_hi4));
uvec4 qs64_hi4 = uvec4(unpack8(qs64_u32_hi4));
const uint32_t q4_0 = qs0_lo4.x;
const uint32_t q4_1 = qs0_lo4.y;
const uint32_t q4_2 = qs0_lo4.z;
const uint32_t q4_3 = qs0_lo4.w;
const uint32_t q4_4 = qs0_hi4.x;
const uint32_t q4_5 = qs0_hi4.y;
const uint32_t q4_6 = qs0_hi4.z;
const uint32_t q4_7 = qs0_hi4.w;
const uint32_t q4_8 = qs64_lo4.x;
const uint32_t q4_9 = qs64_lo4.y;
const uint32_t q4_10 = qs64_lo4.z;
const uint32_t q4_11 = qs64_lo4.w;
const uint32_t q4_12 = qs64_hi4.x;
const uint32_t q4_13 = qs64_hi4.y;
const uint32_t q4_14 = qs64_hi4.z;
const uint32_t q4_15 = qs64_hi4.w;
B_TYPE_VEC4 by10 = data_b_v4[(b_offset + y1_idx) / 4]; B_TYPE_VEC4 by10 = data_b_v4[(b_offset + y1_idx) / 4];
B_TYPE_VEC4 by132 = data_b_v4[(b_offset + y1_idx) / 4 + 8]; B_TYPE_VEC4 by132 = data_b_v4[(b_offset + y1_idx) / 4 + 8];
B_TYPE_VEC4 by20 = data_b_v4[(b_offset + y2_idx) / 4]; B_TYPE_VEC4 by20 = data_b_v4[(b_offset + y2_idx) / 4];
B_TYPE_VEC4 by232 = data_b_v4[(b_offset + y2_idx) / 4 + 8]; B_TYPE_VEC4 by232 = data_b_v4[(b_offset + y2_idx) / 4 + 8];
const FLOAT_TYPE sx = fma(FLOAT_TYPE(by10.x), q4_0, fma(FLOAT_TYPE(by10.y), q4_1, fma(FLOAT_TYPE(by10.z), q4_2, FLOAT_TYPE(by10.w) * q4_3))); [[unroll]] for (uint n = 0; n < num_rows; ++n) {
const FLOAT_TYPE sy = fma(FLOAT_TYPE(by132.x), q4_4, fma(FLOAT_TYPE(by132.y), q4_5, fma(FLOAT_TYPE(by132.z), q4_6, FLOAT_TYPE(by132.w) * q4_7))); const uint ib0 = a_offset / QUANT_K + (first_row+n)*num_blocks_per_row;
const FLOAT_TYPE sz = fma(FLOAT_TYPE(by20.x), q4_8, fma(FLOAT_TYPE(by20.y), q4_9, fma(FLOAT_TYPE(by20.z), q4_10, FLOAT_TYPE(by20.w) * q4_11))); f16vec2 d = data_a[ib0 + i].d;
const FLOAT_TYPE sw = fma(FLOAT_TYPE(by232.x), q4_12, fma(FLOAT_TYPE(by232.y), q4_13, fma(FLOAT_TYPE(by232.z), q4_14, FLOAT_TYPE(by232.w) * q4_15))); const FLOAT_TYPE dall = FLOAT_TYPE(d.x);
const FLOAT_TYPE smin = const FLOAT_TYPE dmin = FLOAT_TYPE(d.y);
fma(FLOAT_TYPE(by10.x), sc2, fma(FLOAT_TYPE(by132.x), sc3, fma(FLOAT_TYPE(by20.x), sc6, fma(FLOAT_TYPE(by232.x), sc7,
fma(FLOAT_TYPE(by10.y), sc2, fma(FLOAT_TYPE(by132.y), sc3, fma(FLOAT_TYPE(by20.y), sc6, fma(FLOAT_TYPE(by232.y), sc7, uint32_t scale0_u32 = data_a_packed16[ib0 + i].scales[v_im ];
fma(FLOAT_TYPE(by10.z), sc2, fma(FLOAT_TYPE(by132.z), sc3, fma(FLOAT_TYPE(by20.z), sc6, fma(FLOAT_TYPE(by232.z), sc7, uint32_t scale4_u32 = data_a_packed16[ib0 + i].scales[v_im + 2];
fma(FLOAT_TYPE(by10.w), sc2, fma(FLOAT_TYPE(by132.w), sc3, fma(FLOAT_TYPE(by20.w), sc6, FLOAT_TYPE(by232.w) * sc7))))))))))))))); uint32_t scale8_u32 = data_a_packed16[ib0 + i].scales[v_im + 4];
temp = fma(dall, fma(sx, sc0, fma(sy, sc1, fma(sz, sc4, sw * sc5))), fma(-dmin, smin, temp)); uvec4 scale0 = uvec4(unpack8(scale0_u32));
uvec4 scale4 = uvec4(unpack8(scale4_u32));
uvec4 scale8 = uvec4(unpack8(scale8_u32));
const uint32_t sc0 = ( scale0.x & 0x3f);
const uint32_t sc1 = ( scale0.y & 0x3f);
const uint32_t sc2 = ( scale4.x & 0x3f);
const uint32_t sc3 = ( scale4.y & 0x3f);
const uint32_t sc4 = (( scale8.x & 0x0f) | ((scale0.x & 0xc0) >> 2));
const uint32_t sc5 = (( scale8.y & 0x0f) | ((scale0.y & 0xc0) >> 2));
const uint32_t sc6 = (((scale8.x >> 4) & 0x0f) | ((scale4.x & 0xc0) >> 2));
const uint32_t sc7 = (((scale8.y >> 4) & 0x0f) | ((scale4.y & 0xc0) >> 2));
uint32_t qs0_u32 = data_a_packed32[ib0 + i].qs[q_offset / 4];
uint32_t qs64_u32 = data_a_packed32[ib0 + i].qs[q_offset / 4 + 16];
uint32_t qs0_u32_lo4 = qs0_u32 & 0x0F0F0F0F;
uint32_t qs0_u32_hi4 = (qs0_u32 >> 4) & 0x0F0F0F0F;
uint32_t qs64_u32_lo4 = qs64_u32 & 0x0F0F0F0F;
uint32_t qs64_u32_hi4 = (qs64_u32 >> 4) & 0x0F0F0F0F;
uvec4 qs0_lo4 = uvec4(unpack8(qs0_u32_lo4));
uvec4 qs64_lo4 = uvec4(unpack8(qs64_u32_lo4));
uvec4 qs0_hi4 = uvec4(unpack8(qs0_u32_hi4));
uvec4 qs64_hi4 = uvec4(unpack8(qs64_u32_hi4));
const uint32_t q4_0 = qs0_lo4.x;
const uint32_t q4_1 = qs0_lo4.y;
const uint32_t q4_2 = qs0_lo4.z;
const uint32_t q4_3 = qs0_lo4.w;
const uint32_t q4_4 = qs0_hi4.x;
const uint32_t q4_5 = qs0_hi4.y;
const uint32_t q4_6 = qs0_hi4.z;
const uint32_t q4_7 = qs0_hi4.w;
const uint32_t q4_8 = qs64_lo4.x;
const uint32_t q4_9 = qs64_lo4.y;
const uint32_t q4_10 = qs64_lo4.z;
const uint32_t q4_11 = qs64_lo4.w;
const uint32_t q4_12 = qs64_hi4.x;
const uint32_t q4_13 = qs64_hi4.y;
const uint32_t q4_14 = qs64_hi4.z;
const uint32_t q4_15 = qs64_hi4.w;
const FLOAT_TYPE sx = fma(FLOAT_TYPE(by10.x), q4_0, fma(FLOAT_TYPE(by10.y), q4_1, fma(FLOAT_TYPE(by10.z), q4_2, FLOAT_TYPE(by10.w) * q4_3)));
const FLOAT_TYPE sy = fma(FLOAT_TYPE(by132.x), q4_4, fma(FLOAT_TYPE(by132.y), q4_5, fma(FLOAT_TYPE(by132.z), q4_6, FLOAT_TYPE(by132.w) * q4_7)));
const FLOAT_TYPE sz = fma(FLOAT_TYPE(by20.x), q4_8, fma(FLOAT_TYPE(by20.y), q4_9, fma(FLOAT_TYPE(by20.z), q4_10, FLOAT_TYPE(by20.w) * q4_11)));
const FLOAT_TYPE sw = fma(FLOAT_TYPE(by232.x), q4_12, fma(FLOAT_TYPE(by232.y), q4_13, fma(FLOAT_TYPE(by232.z), q4_14, FLOAT_TYPE(by232.w) * q4_15)));
const FLOAT_TYPE smin =
fma(FLOAT_TYPE(by10.x), sc2, fma(FLOAT_TYPE(by132.x), sc3, fma(FLOAT_TYPE(by20.x), sc6, fma(FLOAT_TYPE(by232.x), sc7,
fma(FLOAT_TYPE(by10.y), sc2, fma(FLOAT_TYPE(by132.y), sc3, fma(FLOAT_TYPE(by20.y), sc6, fma(FLOAT_TYPE(by232.y), sc7,
fma(FLOAT_TYPE(by10.z), sc2, fma(FLOAT_TYPE(by132.z), sc3, fma(FLOAT_TYPE(by20.z), sc6, fma(FLOAT_TYPE(by232.z), sc7,
fma(FLOAT_TYPE(by10.w), sc2, fma(FLOAT_TYPE(by132.w), sc3, fma(FLOAT_TYPE(by20.w), sc6, FLOAT_TYPE(by232.w) * sc7)))))))))))))));
temp[n] = fma(dall, fma(sx, sc0, fma(sy, sc1, fma(sz, sc4, sw * sc5))), fma(-dmin, smin, temp[n]));
}
} }
tmp[gl_LocalInvocationID.x] = temp;
// sum up partial sums and write back result // sum up partial sums and write back result
[[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] = temp[n];
}
barrier(); barrier();
[[unroll]] for (uint s = gl_WorkGroupSize.x/2; s > 0; s >>= 1) { [[unroll]] for (uint s = BLOCK_SIZE/2; s > 0; s >>= 1) {
if (tid < s) { if (tid < s) {
tmp[tid] += tmp[tid + s]; [[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] += tmpsh[n][tid + s];
}
} }
barrier(); barrier();
} }
if (tid == 0) { if (tid == 0) {
data_d[d_offset + row] = D_TYPE(tmp[0]); [[unroll]] for (uint n = 0; n < num_rows; ++n) {
data_d[d_offset + first_row + n] = D_TYPE(tmpsh[n][0]);
}
}
}
void main() {
const uint first_row = NUM_ROWS * (gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z);
// do NUM_ROWS at a time, unless there aren't enough remaining rows
if (first_row + NUM_ROWS <= p.stride_d) {
compute_outputs(first_row, NUM_ROWS);
} else {
if (first_row >= p.stride_d) {
return;
}
compute_outputs(first_row, p.stride_d - first_row);
} }
} }

226
ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q5_k.comp
View File

@ -7,21 +7,15 @@
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in; layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (constant_id = 0) const uint BLOCK_SIZE = 32; layout (constant_id = 0) const uint BLOCK_SIZE = 32;
layout (constant_id = 1) const uint NUM_ROWS = 1;
shared FLOAT_TYPE tmp[BLOCK_SIZE]; shared FLOAT_TYPE tmpsh[NUM_ROWS][BLOCK_SIZE];
void main() {
const uint row = gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z;
if (row >= p.stride_d) {
return;
}
void compute_outputs(const uint32_t first_row, const uint32_t num_rows) {
uint a_offset, b_offset, d_offset; uint a_offset, b_offset, d_offset;
get_offsets(a_offset, b_offset, d_offset); get_offsets(a_offset, b_offset, d_offset);
const uint num_blocks_per_row = p.ncols / QUANT_K; const uint num_blocks_per_row = p.ncols / QUANT_K;
const uint ib0 = a_offset / QUANT_K + row*num_blocks_per_row;
// 16 threads are used to process each block // 16 threads are used to process each block
const uint it_size = gl_WorkGroupSize.x/16; const uint it_size = gl_WorkGroupSize.x/16;
@ -39,74 +33,16 @@ void main() {
const uint q_offset = 32*v_im + l0; const uint q_offset = 32*v_im + l0;
const uint y_offset = 64*v_im + l0; const uint y_offset = 64*v_im + l0;
FLOAT_TYPE temp = FLOAT_TYPE(0.0); // partial sum for thread in warp FLOAT_TYPE temp[NUM_ROWS];
[[unroll]] for (uint i = 0; i < NUM_ROWS; ++i) {
temp[i] = FLOAT_TYPE(0);
}
[[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) { [[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) {
const uint y1_idx = i * QUANT_K + y_offset; const uint y1_idx = i * QUANT_K + y_offset;
const uint y2_idx = y1_idx + 128; const uint y2_idx = y1_idx + 128;
f16vec2 d = data_a[ib0 + i].d;
const FLOAT_TYPE dall = FLOAT_TYPE(d.x);
const FLOAT_TYPE dmin = FLOAT_TYPE(d.y);
uint32_t scale0_u32 = data_a_packed16[ib0 + i].scales[v_im ];
uint32_t scale4_u32 = data_a_packed16[ib0 + i].scales[v_im + 2];
uint32_t scale8_u32 = data_a_packed16[ib0 + i].scales[v_im + 4];
uvec4 scale0 = uvec4(unpack8(scale0_u32));
uvec4 scale4 = uvec4(unpack8(scale4_u32));
uvec4 scale8 = uvec4(unpack8(scale8_u32));
const uint32_t sc0 = ( scale0.x & 0x3f);
const uint32_t sc1 = ( scale0.y & 0x3f);
const uint32_t sc2 = ( scale4.x & 0x3f);
const uint32_t sc3 = ( scale4.y & 0x3f);
const uint32_t sc4 = (( scale8.x & 0x0f) | ((scale0.x & 0xc0) >> 2));
const uint32_t sc5 = (( scale8.y & 0x0f) | ((scale0.y & 0xc0) >> 2));
const uint32_t sc6 = (((scale8.x >> 4) & 0x0f) | ((scale4.x & 0xc0) >> 2));
const uint32_t sc7 = (((scale8.y >> 4) & 0x0f) | ((scale4.y & 0xc0) >> 2));
uint32_t qs0_16_u32 = uint32_t(data_a_packed16[ib0 + i].qs[q_offset / 2]) | (uint32_t(data_a_packed16[ib0 + i].qs[q_offset / 2 + 8]) << 16);
uint32_t qs64_80_u32 = uint32_t(data_a_packed16[ib0 + i].qs[q_offset / 2 + 32]) | (uint32_t(data_a_packed16[ib0 + i].qs[q_offset / 2 + 40]) << 16);
uint32_t qs0_16_u32_lo4 = qs0_16_u32 & 0x0F0F0F0F;
uint32_t qs0_16_u32_hi4 = (qs0_16_u32 >> 4) & 0x0F0F0F0F;
uint32_t qs64_80_u32_lo4 = qs64_80_u32 & 0x0F0F0F0F;
uint32_t qs64_80_u32_hi4 = (qs64_80_u32 >> 4) & 0x0F0F0F0F;
uint32_t qh = pack32(u16vec2(data_a_packed16[ib0 + i].qh[l0 / 2], data_a_packed16[ib0 + i].qh[l0 / 2 + 8]));
uint32_t qs0_16_lo4_offset16 = ((qh >> (2*v_im)) & 0x01010101) << 4;
uint32_t qs0_16_hi4_offset16 = ((qh >> (2*v_im)) & 0x02020202) << 3;
uint32_t qs64_80_lo4_offset16 = ((qh >> (2*v_im)) & 0x10101010) << 0;
uint32_t qs64_80_hi4_offset16 = ((qh >> (2*v_im)) & 0x20202020) >> 1;
qs0_16_u32_lo4 += qs0_16_lo4_offset16;
qs0_16_u32_hi4 += qs0_16_hi4_offset16;
qs64_80_u32_lo4 += qs64_80_lo4_offset16;
qs64_80_u32_hi4 += qs64_80_hi4_offset16;
uvec4 qs0_16_lo4 = uvec4(unpack8(qs0_16_u32_lo4));
uvec4 qs64_80_lo4 = uvec4(unpack8(qs64_80_u32_lo4));
uvec4 qs0_16_hi4 = uvec4(unpack8(qs0_16_u32_hi4));
uvec4 qs64_80_hi4 = uvec4(unpack8(qs64_80_u32_hi4));
const uint32_t q4_0 = qs0_16_lo4.x;
const uint32_t q4_1 = qs0_16_lo4.y;
const uint32_t q4_2 = qs0_16_lo4.z;
const uint32_t q4_3 = qs0_16_lo4.w;
const uint32_t q4_4 = qs0_16_hi4.x;
const uint32_t q4_5 = qs0_16_hi4.y;
const uint32_t q4_6 = qs0_16_hi4.z;
const uint32_t q4_7 = qs0_16_hi4.w;
const uint32_t q4_8 = qs64_80_lo4.x;
const uint32_t q4_9 = qs64_80_lo4.y;
const uint32_t q4_10 = qs64_80_lo4.z;
const uint32_t q4_11 = qs64_80_lo4.w;
const uint32_t q4_12 = qs64_80_hi4.x;
const uint32_t q4_13 = qs64_80_hi4.y;
const uint32_t q4_14 = qs64_80_hi4.z;
const uint32_t q4_15 = qs64_80_hi4.w;
B_TYPE_VEC2 by10 = data_b_v2[(b_offset + y1_idx) / 2]; B_TYPE_VEC2 by10 = data_b_v2[(b_offset + y1_idx) / 2];
B_TYPE_VEC2 by116 = data_b_v2[(b_offset + y1_idx) / 2 + 8]; B_TYPE_VEC2 by116 = data_b_v2[(b_offset + y1_idx) / 2 + 8];
B_TYPE_VEC2 by132 = data_b_v2[(b_offset + y1_idx) / 2 + 16]; B_TYPE_VEC2 by132 = data_b_v2[(b_offset + y1_idx) / 2 + 16];
@ -116,45 +52,129 @@ void main() {
B_TYPE_VEC2 by232 = data_b_v2[(b_offset + y2_idx) / 2 + 16]; B_TYPE_VEC2 by232 = data_b_v2[(b_offset + y2_idx) / 2 + 16];
B_TYPE_VEC2 by248 = data_b_v2[(b_offset + y2_idx) / 2 + 24]; B_TYPE_VEC2 by248 = data_b_v2[(b_offset + y2_idx) / 2 + 24];
const FLOAT_TYPE sx = [[unroll]] for (uint n = 0; n < num_rows; ++n) {
fma(FLOAT_TYPE(by10.x), q4_0, const uint ib0 = a_offset / QUANT_K + (first_row+n)*num_blocks_per_row;
fma(FLOAT_TYPE(by10.y), q4_1, f16vec2 d = data_a[ib0 + i].d;
fma(FLOAT_TYPE(by116.x), q4_2, const FLOAT_TYPE dall = FLOAT_TYPE(d.x);
FLOAT_TYPE(by116.y) * q4_3))); const FLOAT_TYPE dmin = FLOAT_TYPE(d.y);
const FLOAT_TYPE sy =
fma(FLOAT_TYPE(by132.x), q4_4, uint32_t scale0_u32 = data_a_packed16[ib0 + i].scales[v_im ];
fma(FLOAT_TYPE(by132.y), q4_5, uint32_t scale4_u32 = data_a_packed16[ib0 + i].scales[v_im + 2];
fma(FLOAT_TYPE(by148.x), q4_6, uint32_t scale8_u32 = data_a_packed16[ib0 + i].scales[v_im + 4];
FLOAT_TYPE(by148.y) * q4_7))); uvec4 scale0 = uvec4(unpack8(scale0_u32));
const FLOAT_TYPE sz = uvec4 scale4 = uvec4(unpack8(scale4_u32));
fma(FLOAT_TYPE(by20.x), q4_8, uvec4 scale8 = uvec4(unpack8(scale8_u32));
fma(FLOAT_TYPE(by20.y), q4_9,
fma(FLOAT_TYPE(by216.x), q4_10, const uint32_t sc0 = ( scale0.x & 0x3f);
FLOAT_TYPE(by216.y) * q4_11))); const uint32_t sc1 = ( scale0.y & 0x3f);
const FLOAT_TYPE sw = const uint32_t sc2 = ( scale4.x & 0x3f);
fma(FLOAT_TYPE(by232.x), q4_12, const uint32_t sc3 = ( scale4.y & 0x3f);
fma(FLOAT_TYPE(by232.y), q4_13, const uint32_t sc4 = (( scale8.x & 0x0f) | ((scale0.x & 0xc0) >> 2));
fma(FLOAT_TYPE(by248.x), q4_14, const uint32_t sc5 = (( scale8.y & 0x0f) | ((scale0.y & 0xc0) >> 2));
FLOAT_TYPE(by248.y) * q4_15))); const uint32_t sc6 = (((scale8.x >> 4) & 0x0f) | ((scale4.x & 0xc0) >> 2));
const FLOAT_TYPE smin = const uint32_t sc7 = (((scale8.y >> 4) & 0x0f) | ((scale4.y & 0xc0) >> 2));
fma(FLOAT_TYPE(by10.x) + FLOAT_TYPE(by10.y) + FLOAT_TYPE(by116.x) + FLOAT_TYPE(by116.y), sc2,
fma(FLOAT_TYPE(by132.x) + FLOAT_TYPE(by132.y) + FLOAT_TYPE(by148.x) + FLOAT_TYPE(by148.y), sc3, uint32_t qs0_16_u32 = uint32_t(data_a_packed16[ib0 + i].qs[q_offset / 2]) | (uint32_t(data_a_packed16[ib0 + i].qs[q_offset / 2 + 8]) << 16);
fma(FLOAT_TYPE(by20.x) + FLOAT_TYPE(by20.y) + FLOAT_TYPE(by216.x) + FLOAT_TYPE(by216.y), sc6, uint32_t qs64_80_u32 = uint32_t(data_a_packed16[ib0 + i].qs[q_offset / 2 + 32]) | (uint32_t(data_a_packed16[ib0 + i].qs[q_offset / 2 + 40]) << 16);
(FLOAT_TYPE(by232.x) + FLOAT_TYPE(by232.y) + FLOAT_TYPE(by248.x) + FLOAT_TYPE(by248.y)) * sc7)));
temp = fma(dall, fma(sx, sc0, fma(sy, sc1, fma(sz, sc4, sw * sc5))), fma(-dmin, smin, temp)); uint32_t qs0_16_u32_lo4 = qs0_16_u32 & 0x0F0F0F0F;
uint32_t qs0_16_u32_hi4 = (qs0_16_u32 >> 4) & 0x0F0F0F0F;
uint32_t qs64_80_u32_lo4 = qs64_80_u32 & 0x0F0F0F0F;
uint32_t qs64_80_u32_hi4 = (qs64_80_u32 >> 4) & 0x0F0F0F0F;
uint32_t qh = pack32(u16vec2(data_a_packed16[ib0 + i].qh[l0 / 2], data_a_packed16[ib0 + i].qh[l0 / 2 + 8]));
uint32_t qs0_16_lo4_offset16 = ((qh >> (2*v_im)) & 0x01010101) << 4;
uint32_t qs0_16_hi4_offset16 = ((qh >> (2*v_im)) & 0x02020202) << 3;
uint32_t qs64_80_lo4_offset16 = ((qh >> (2*v_im)) & 0x10101010) << 0;
uint32_t qs64_80_hi4_offset16 = ((qh >> (2*v_im)) & 0x20202020) >> 1;
qs0_16_u32_lo4 += qs0_16_lo4_offset16;
qs0_16_u32_hi4 += qs0_16_hi4_offset16;
qs64_80_u32_lo4 += qs64_80_lo4_offset16;
qs64_80_u32_hi4 += qs64_80_hi4_offset16;
uvec4 qs0_16_lo4 = uvec4(unpack8(qs0_16_u32_lo4));
uvec4 qs64_80_lo4 = uvec4(unpack8(qs64_80_u32_lo4));
uvec4 qs0_16_hi4 = uvec4(unpack8(qs0_16_u32_hi4));
uvec4 qs64_80_hi4 = uvec4(unpack8(qs64_80_u32_hi4));
const uint32_t q4_0 = qs0_16_lo4.x;
const uint32_t q4_1 = qs0_16_lo4.y;
const uint32_t q4_2 = qs0_16_lo4.z;
const uint32_t q4_3 = qs0_16_lo4.w;
const uint32_t q4_4 = qs0_16_hi4.x;
const uint32_t q4_5 = qs0_16_hi4.y;
const uint32_t q4_6 = qs0_16_hi4.z;
const uint32_t q4_7 = qs0_16_hi4.w;
const uint32_t q4_8 = qs64_80_lo4.x;
const uint32_t q4_9 = qs64_80_lo4.y;
const uint32_t q4_10 = qs64_80_lo4.z;
const uint32_t q4_11 = qs64_80_lo4.w;
const uint32_t q4_12 = qs64_80_hi4.x;
const uint32_t q4_13 = qs64_80_hi4.y;
const uint32_t q4_14 = qs64_80_hi4.z;
const uint32_t q4_15 = qs64_80_hi4.w;
const FLOAT_TYPE sx =
fma(FLOAT_TYPE(by10.x), q4_0,
fma(FLOAT_TYPE(by10.y), q4_1,
fma(FLOAT_TYPE(by116.x), q4_2,
FLOAT_TYPE(by116.y) * q4_3)));
const FLOAT_TYPE sy =
fma(FLOAT_TYPE(by132.x), q4_4,
fma(FLOAT_TYPE(by132.y), q4_5,
fma(FLOAT_TYPE(by148.x), q4_6,
FLOAT_TYPE(by148.y) * q4_7)));
const FLOAT_TYPE sz =
fma(FLOAT_TYPE(by20.x), q4_8,
fma(FLOAT_TYPE(by20.y), q4_9,
fma(FLOAT_TYPE(by216.x), q4_10,
FLOAT_TYPE(by216.y) * q4_11)));
const FLOAT_TYPE sw =
fma(FLOAT_TYPE(by232.x), q4_12,
fma(FLOAT_TYPE(by232.y), q4_13,
fma(FLOAT_TYPE(by248.x), q4_14,
FLOAT_TYPE(by248.y) * q4_15)));
const FLOAT_TYPE smin =
fma(FLOAT_TYPE(by10.x) + FLOAT_TYPE(by10.y) + FLOAT_TYPE(by116.x) + FLOAT_TYPE(by116.y), sc2,
fma(FLOAT_TYPE(by132.x) + FLOAT_TYPE(by132.y) + FLOAT_TYPE(by148.x) + FLOAT_TYPE(by148.y), sc3,
fma(FLOAT_TYPE(by20.x) + FLOAT_TYPE(by20.y) + FLOAT_TYPE(by216.x) + FLOAT_TYPE(by216.y), sc6,
(FLOAT_TYPE(by232.x) + FLOAT_TYPE(by232.y) + FLOAT_TYPE(by248.x) + FLOAT_TYPE(by248.y)) * sc7)));
temp[n] = fma(dall, fma(sx, sc0, fma(sy, sc1, fma(sz, sc4, sw * sc5))), fma(-dmin, smin, temp[n]));
}
} }
tmp[gl_LocalInvocationID.x] = temp;
// sum up partial sums and write back result // sum up partial sums and write back result
[[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] = temp[n];
}
barrier(); barrier();
[[unroll]] for (uint s = gl_WorkGroupSize.x/2; s > 0; s >>= 1) { [[unroll]] for (uint s = BLOCK_SIZE/2; s > 0; s >>= 1) {
if (tid < s) { if (tid < s) {
tmp[tid] += tmp[tid + s]; [[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] += tmpsh[n][tid + s];
}
} }
barrier(); barrier();
} }
if (tid == 0) { if (tid == 0) {
data_d[d_offset + row] = D_TYPE(tmp[0]); [[unroll]] for (uint n = 0; n < num_rows; ++n) {
data_d[d_offset + first_row + n] = D_TYPE(tmpsh[n][0]);
}
}
}
void main() {
const uint first_row = NUM_ROWS * (gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z);
// do NUM_ROWS at a time, unless there aren't enough remaining rows
if (first_row + NUM_ROWS <= p.stride_d) {
compute_outputs(first_row, NUM_ROWS);
} else {
if (first_row >= p.stride_d) {
return;
}
compute_outputs(first_row, p.stride_d - first_row);
} }
} }

130
ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q6_k.comp
View File

@ -7,21 +7,15 @@
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in; layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (constant_id = 0) const uint BLOCK_SIZE = 32; layout (constant_id = 0) const uint BLOCK_SIZE = 32;
layout (constant_id = 1) const uint NUM_ROWS = 1;
shared FLOAT_TYPE tmp[BLOCK_SIZE]; shared FLOAT_TYPE tmpsh[NUM_ROWS][BLOCK_SIZE];
void main() {
const uint row = gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z;
if (row >= p.stride_d) {
return;
}
void compute_outputs(const uint32_t first_row, const uint32_t num_rows) {
uint a_offset, b_offset, d_offset; uint a_offset, b_offset, d_offset;
get_offsets(a_offset, b_offset, d_offset); get_offsets(a_offset, b_offset, d_offset);
const uint num_blocks_per_row = p.ncols / QUANT_K; const uint num_blocks_per_row = p.ncols / QUANT_K;
const uint ib0 = a_offset / QUANT_K + row*num_blocks_per_row;
// 16 threads are used to process each block // 16 threads are used to process each block
const uint it_size = gl_WorkGroupSize.x/16; const uint it_size = gl_WorkGroupSize.x/16;
@ -42,69 +36,95 @@ void main() {
const uint s_offset = 8*v_im + is; const uint s_offset = 8*v_im + is;
const uint y_offset = 128*v_im + l0; const uint y_offset = 128*v_im + l0;
FLOAT_TYPE temp = FLOAT_TYPE(0.0); // partial sum for thread in warp FLOAT_TYPE temp[NUM_ROWS];
[[unroll]] for (uint i = 0; i < NUM_ROWS; ++i) {
temp[i] = FLOAT_TYPE(0);
}
[[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) { [[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) {
const uint y_idx = i * QUANT_K + y_offset; const uint y_idx = i * QUANT_K + y_offset;
const FLOAT_TYPE d = FLOAT_TYPE(data_a[ib0 + i].d);
FLOAT_TYPE scales[4];
scales[0] = FLOAT_TYPE(data_a[ib0 + i].scales[s_offset + 0]);
scales[1] = FLOAT_TYPE(data_a[ib0 + i].scales[s_offset + 2]);
scales[2] = FLOAT_TYPE(data_a[ib0 + i].scales[s_offset + 4]);
scales[3] = FLOAT_TYPE(data_a[ib0 + i].scales[s_offset + 6]);
uint32_t ql0_u32 = uint32_t(data_a_packed16[ib0 + i].ql[ql_offset / 2]) | (uint32_t(data_a_packed16[ib0 + i].ql[ql_offset / 2 + 1]) << 16);
uint32_t ql32_u32 = uint32_t(data_a_packed16[ib0 + i].ql[ql_offset / 2 + 16]) | (uint32_t(data_a_packed16[ib0 + i].ql[ql_offset / 2 + 17]) << 16);
uint32_t ql0_u32_lo4 = ql0_u32 & 0x0F0F0F0F;
uint32_t ql0_u32_hi4 = (ql0_u32 >> 4) & 0x0F0F0F0F;
uint32_t ql32_u32_lo4 = ql32_u32 & 0x0F0F0F0F;
uint32_t ql32_u32_hi4 = (ql32_u32 >> 4) & 0x0F0F0F0F;
uint32_t qh_u32 = uint32_t(data_a_packed16[ib0 + i].qh[qh_offset / 2]) | (uint32_t(data_a_packed16[ib0 + i].qh[qh_offset / 2 + 1]) << 16);
uint32_t qh0_u32 = (qh_u32 & 0x03030303) << 4;
uint32_t qh2_u32 = (qh_u32 & 0x0C0C0C0C) << 2;
uint32_t qh4_u32 = (qh_u32 & 0x30303030) << 0;
uint32_t qh6_u32 = (qh_u32 & 0xC0C0C0C0) >> 2;
uint32_t q0_u32 = ql0_u32_lo4 | qh0_u32;
uint32_t q1_u32 = ql32_u32_lo4 | qh2_u32;
uint32_t q2_u32 = ql0_u32_hi4 | qh4_u32;
uint32_t q3_u32 = ql32_u32_hi4 | qh6_u32;
uvec4 q0 = uvec4(unpack8(q0_u32));
uvec4 q1 = uvec4(unpack8(q1_u32));
uvec4 q2 = uvec4(unpack8(q2_u32));
uvec4 q3 = uvec4(unpack8(q3_u32));
B_TYPE_VEC4 by0 = data_b_v4[(b_offset + y_idx) / 4]; B_TYPE_VEC4 by0 = data_b_v4[(b_offset + y_idx) / 4];
B_TYPE_VEC4 by32 = data_b_v4[(b_offset + y_idx) / 4 + 8]; B_TYPE_VEC4 by32 = data_b_v4[(b_offset + y_idx) / 4 + 8];
B_TYPE_VEC4 by64 = data_b_v4[(b_offset + y_idx) / 4 + 16]; B_TYPE_VEC4 by64 = data_b_v4[(b_offset + y_idx) / 4 + 16];
B_TYPE_VEC4 by96 = data_b_v4[(b_offset + y_idx) / 4 + 24]; B_TYPE_VEC4 by96 = data_b_v4[(b_offset + y_idx) / 4 + 24];
FLOAT_TYPE sum = FLOAT_TYPE(0.0); [[unroll]] for (uint n = 0; n < num_rows; ++n) {
[[unroll]] for (int l = 0; l < 4; ++l) { const uint ib0 = a_offset / QUANT_K + (first_row+n)*num_blocks_per_row;
sum = fma(FLOAT_TYPE(by0[l]) * scales[0], FLOAT_TYPE(int8_t(q0[l]) - 32), const FLOAT_TYPE d = FLOAT_TYPE(data_a[ib0 + i].d);
fma(FLOAT_TYPE(by32[l]) * scales[1], FLOAT_TYPE(int8_t(q1[l]) - 32),
fma(FLOAT_TYPE(by64[l]) * scales[2], FLOAT_TYPE(int8_t(q2[l]) - 32), FLOAT_TYPE scales[4];
fma(FLOAT_TYPE(by96[l]) * scales[3], FLOAT_TYPE(int8_t(q3[l]) - 32), sum)))); scales[0] = FLOAT_TYPE(data_a[ib0 + i].scales[s_offset + 0]);
scales[1] = FLOAT_TYPE(data_a[ib0 + i].scales[s_offset + 2]);
scales[2] = FLOAT_TYPE(data_a[ib0 + i].scales[s_offset + 4]);
scales[3] = FLOAT_TYPE(data_a[ib0 + i].scales[s_offset + 6]);
uint32_t ql0_u32 = uint32_t(data_a_packed16[ib0 + i].ql[ql_offset / 2]) | (uint32_t(data_a_packed16[ib0 + i].ql[ql_offset / 2 + 1]) << 16);
uint32_t ql32_u32 = uint32_t(data_a_packed16[ib0 + i].ql[ql_offset / 2 + 16]) | (uint32_t(data_a_packed16[ib0 + i].ql[ql_offset / 2 + 17]) << 16);
uint32_t ql0_u32_lo4 = ql0_u32 & 0x0F0F0F0F;
uint32_t ql0_u32_hi4 = (ql0_u32 >> 4) & 0x0F0F0F0F;
uint32_t ql32_u32_lo4 = ql32_u32 & 0x0F0F0F0F;
uint32_t ql32_u32_hi4 = (ql32_u32 >> 4) & 0x0F0F0F0F;
uint32_t qh_u32 = uint32_t(data_a_packed16[ib0 + i].qh[qh_offset / 2]) | (uint32_t(data_a_packed16[ib0 + i].qh[qh_offset / 2 + 1]) << 16);
uint32_t qh0_u32 = (qh_u32 & 0x03030303) << 4;
uint32_t qh2_u32 = (qh_u32 & 0x0C0C0C0C) << 2;
uint32_t qh4_u32 = (qh_u32 & 0x30303030) << 0;
uint32_t qh6_u32 = (qh_u32 & 0xC0C0C0C0) >> 2;
uint32_t q0_u32 = ql0_u32_lo4 | qh0_u32;
uint32_t q1_u32 = ql32_u32_lo4 | qh2_u32;
uint32_t q2_u32 = ql0_u32_hi4 | qh4_u32;
uint32_t q3_u32 = ql32_u32_hi4 | qh6_u32;
uvec4 q0 = uvec4(unpack8(q0_u32));
uvec4 q1 = uvec4(unpack8(q1_u32));
uvec4 q2 = uvec4(unpack8(q2_u32));
uvec4 q3 = uvec4(unpack8(q3_u32));
FLOAT_TYPE sum = FLOAT_TYPE(0.0);
[[unroll]] for (int l = 0; l < 4; ++l) {
sum = fma(FLOAT_TYPE(by0[l]) * scales[0], FLOAT_TYPE(int8_t(q0[l]) - 32),
fma(FLOAT_TYPE(by32[l]) * scales[1], FLOAT_TYPE(int8_t(q1[l]) - 32),
fma(FLOAT_TYPE(by64[l]) * scales[2], FLOAT_TYPE(int8_t(q2[l]) - 32),
fma(FLOAT_TYPE(by96[l]) * scales[3], FLOAT_TYPE(int8_t(q3[l]) - 32), sum))));
}
temp[n] += sum * d;
} }
temp += sum * d;
} }
tmp[gl_LocalInvocationID.x] = temp;
// sum up partial sums and write back result // sum up partial sums and write back result
[[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] = temp[n];
}
barrier(); barrier();
[[unroll]] for (uint s = gl_WorkGroupSize.x/2; s > 0; s >>= 1) { [[unroll]] for (uint s = BLOCK_SIZE/2; s > 0; s >>= 1) {
if (tid < s) { if (tid < s) {
tmp[tid] += tmp[tid + s]; [[unroll]] for (uint n = 0; n < num_rows; ++n) {
tmpsh[n][tid] += tmpsh[n][tid + s];
}
} }
barrier(); barrier();
} }
if (tid == 0) { if (tid == 0) {
data_d[d_offset + row] = D_TYPE(tmp[0]); [[unroll]] for (uint n = 0; n < num_rows; ++n) {
data_d[d_offset + first_row + n] = D_TYPE(tmpsh[n][0]);
}
}
}
void main() {
const uint first_row = NUM_ROWS * (gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z);
// do NUM_ROWS at a time, unless there aren't enough remaining rows
if (first_row + NUM_ROWS <= p.stride_d) {
compute_outputs(first_row, NUM_ROWS);
} else {
if (first_row >= p.stride_d) {
return;
}
compute_outputs(first_row, p.stride_d - first_row);
} }
} }