ggml : implement REGLU/GEGLU/SWIGLU ops (#14158)

* implement unary REGLU/GEGLU/SWIGLU cpu ops

* relax constraints

* duplicate shape of source

* fix ggml_vec_geglu_f16

* special case gated ops

* implement unary REGLU/GEGLU/SWIGLU cuda ops

* tighten constraints again

* refactor into GGML_GLU_OP

* metal : add glu kernels

ggml-ci

* add CUDA_GLU_BLOCK_SIZE [no ci]

* more constraints and use 64bit ints

ggml-ci

* 64bit multiplication [no ci]

* implement swapped variants (cpu/cuda)

* update comment [no ci]

ggml-ci

* Vulkan: Add GLU ops and shaders

* SYCL: Implement fused kernel GEGLU, SWIGLU and REGLU for single up+gate

* ggml : implement GLU for split up/gate (#14181)

* implement GLU for split up/gate

* add tests for ggml_glu_split

* Vulkan: Implement glu_split logic and shader support

* add split to logging [no ci]

* SYCL: refactor element_size ops and add split up and gate support to gated kernels

* SYCL: switch GEGLU to use tanh approximation

---------

Co-authored-by: 0cc4m <picard12@live.de>
Co-authored-by: Akarshan <akarshan@menlo.ai>

* GGML: increase OP count in assertion

* Refactor: Optimize SYCL element-wise operations with unary function inlining

This commit refactors the SYCL element-wise operations to improve performance by:

- Inlining unary operations (sgn, abs, elu, gelu, silu, etc.) to reduce kernel launch overhead.
- Introducing helper functions `op_xxx` for each unary operation to encapsulate the logic.
- Replacing direct kernel calls with calls to these inlined functions.
- Using `__dpct_inline__` to encourage compiler inlining.
- Minor code cleanup and consistency improvements.

The changes aim to reduce kernel launch overhead and improve the overall efficiency of element-wise operations on SYCL devices.

* vulkan: Increase workgroup size for GLU, for performance (#14345)

* vulkan: Increase workgroup size for GLU, for performance

* vulkan: change GLU shaders to do one element per invocation rather than one row per workgroup

* merge fix

* metal : add support for split and swap

ggml-ci

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: 0cc4m <picard12@live.de>
Co-authored-by: Akarshan <akarshan@menlo.ai>
Co-authored-by: Jeff Bolz <jbolz@nvidia.com>

ggml/src/ggml-vulkan/vulkan-shaders/geglu.comp

@@ -0,0 +1,13 @@
+#version 450
+
+#include "glu_head.comp"
+
+const float GELU_COEF_A    = 0.044715f;
+const float SQRT_2_OVER_PI = 0.79788456080286535587989211986876f;
+
+float op(float a, float b) {
+    const float val = SQRT_2_OVER_PI*a*(1.0f + GELU_COEF_A*a*a);
+    return 0.5f*a*(2.0f - 2.0f / (exp(2 * val) + 1)) * b;
+}
+
+#include "glu_main.comp"

ggml/src/ggml-vulkan/vulkan-shaders/glu_head.comp

@@ -0,0 +1,15 @@
+#extension GL_EXT_shader_16bit_storage : require
+
+layout(local_size_x = 512, local_size_y = 1, local_size_z = 1) in;
+
+layout (binding = 0) readonly buffer A {A_TYPE data_a[];};
+layout (binding = 1) readonly buffer B {A_TYPE data_b[];};
+layout (binding = 2) writeonly buffer D {D_TYPE data_d[];};
+
+layout (push_constant) uniform parameter
+{
+    uint N;
+    uint ne00;
+    uint ne20;
+    uint mode;
+} p;

ggml/src/ggml-vulkan/vulkan-shaders/glu_main.comp

@@ -0,0 +1,29 @@
+void main() {
+    const uint i = gl_GlobalInvocationID.z * 262144 + gl_GlobalInvocationID.y * 512 + gl_GlobalInvocationID.x;
+
+    if (i >= p.N) {
+        return;
+    }
+
+    const uint row = i / p.ne20;
+    const uint col = i - row * p.ne20;
+
+    if (p.mode == 0) {
+        // Default
+        const uint offset = p.ne00 / 2;
+        const uint idx = row * p.ne00 + col;
+
+        data_d[row * offset + col] = D_TYPE(op(float(data_a[idx]), float(data_a[idx + offset])));
+    } else if (p.mode == 1) {
+        // Swapped
+        const uint offset = p.ne00 / 2;
+        const uint idx = row * p.ne00 + col;
+
+        data_d[row * offset + col] = D_TYPE(op(float(data_a[idx + offset]), float(data_a[idx])));
+    } else {
+        // Split
+        const uint idx = row * p.ne00 + col;
+
+        data_d[idx] = D_TYPE(op(float(data_a[idx]), float(data_b[idx])));
+    }
+}

ggml/src/ggml-vulkan/vulkan-shaders/reglu.comp

@@ -0,0 +1,9 @@
+#version 450
+
+#include "glu_head.comp"
+
+float op(float a, float b) {
+    return max(a, 0.0f) * b;
+}
+
+#include "glu_main.comp"

ggml/src/ggml-vulkan/vulkan-shaders/swiglu.comp

@@ -0,0 +1,9 @@
+#version 450
+
+#include "glu_head.comp"
+
+float op(float a, float b) {
+    return a / (1.0f + exp(-a)) * b;
+}
+
+#include "glu_main.comp"

ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp

@@ -585,6 +585,13 @@ void process_shaders() {
     string_to_spv("sigmoid_f16",    "sigmoid.comp",     {{"A_TYPE", "float16_t"},   {"D_TYPE", "float16_t"}});
     string_to_spv("sigmoid_f32",    "sigmoid.comp",     {{"A_TYPE", "float"},       {"D_TYPE", "float"}});
 
+    string_to_spv("geglu_f16",      "geglu.comp",       {{"A_TYPE", "float16_t"},   {"D_TYPE", "float16_t"}});
+    string_to_spv("geglu_f32",      "geglu.comp",       {{"A_TYPE", "float"},       {"D_TYPE", "float"}});
+    string_to_spv("reglu_f16",      "reglu.comp",       {{"A_TYPE", "float16_t"},   {"D_TYPE", "float16_t"}});
+    string_to_spv("reglu_f32",      "reglu.comp",       {{"A_TYPE", "float"},       {"D_TYPE", "float"}});
+    string_to_spv("swiglu_f16",     "swiglu.comp",      {{"A_TYPE", "float16_t"},   {"D_TYPE", "float16_t"}});
+    string_to_spv("swiglu_f32",     "swiglu.comp",      {{"A_TYPE", "float"},       {"D_TYPE", "float"}});
+
     string_to_spv("leaky_relu_f32", "leaky_relu.comp",  {{"A_TYPE", "float"}, {"D_TYPE", "float"}});
     string_to_spv("silu_back_f32",  "silu_back.comp",   {{"A_TYPE", "float"}, {"B_TYPE", "float"}, {"D_TYPE", "float"}});