npm - @fugood/llama.node - Versions diffs - 0.3.3 → 0.3.4 - Mend

@fugood/llama.node 0.3.3 → 0.3.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Files changed (225) hide show

package/CMakeLists.txt +5 -0
package/bin/darwin/arm64/llama-node.node +0 -0
package/bin/darwin/x64/llama-node.node +0 -0
package/bin/linux/arm64/llama-node.node +0 -0
package/bin/linux/x64/llama-node.node +0 -0
package/bin/linux-vulkan/arm64/llama-node.node +0 -0
package/bin/linux-vulkan/x64/llama-node.node +0 -0
package/bin/win32/arm64/llama-node.node +0 -0
package/bin/win32/arm64/node.lib +0 -0
package/bin/win32/x64/llama-node.node +0 -0
package/bin/win32/x64/node.lib +0 -0
package/bin/win32-vulkan/arm64/llama-node.node +0 -0
package/bin/win32-vulkan/arm64/node.lib +0 -0
package/bin/win32-vulkan/x64/llama-node.node +0 -0
package/bin/win32-vulkan/x64/node.lib +0 -0
package/lib/binding.ts +18 -1
package/package.json +1 -1
package/src/EmbeddingWorker.cpp +15 -5
package/src/EmbeddingWorker.h +2 -1
package/src/LlamaCompletionWorker.cpp +1 -1
package/src/LlamaContext.cpp +81 -18
package/src/LlamaContext.h +2 -0
package/src/llama.cpp/.github/workflows/build.yml +197 -159
package/src/llama.cpp/.github/workflows/docker.yml +5 -8
package/src/llama.cpp/.github/workflows/python-lint.yml +8 -1
package/src/llama.cpp/.github/workflows/server.yml +21 -14
package/src/llama.cpp/CMakeLists.txt +11 -6
package/src/llama.cpp/Sources/llama/llama.h +4 -0
package/src/llama.cpp/cmake/common.cmake +33 -0
package/src/llama.cpp/cmake/x64-windows-llvm.cmake +11 -0
package/src/llama.cpp/common/CMakeLists.txt +6 -2
package/src/llama.cpp/common/arg.cpp +426 -245
package/src/llama.cpp/common/common.cpp +143 -80
package/src/llama.cpp/common/common.h +81 -24
package/src/llama.cpp/common/sampling.cpp +53 -19
package/src/llama.cpp/common/sampling.h +22 -1
package/src/llama.cpp/common/speculative.cpp +274 -0
package/src/llama.cpp/common/speculative.h +28 -0
package/src/llama.cpp/docs/build.md +101 -148
package/src/llama.cpp/examples/CMakeLists.txt +32 -13
package/src/llama.cpp/examples/batched/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/batched/batched.cpp +5 -4
package/src/llama.cpp/examples/batched-bench/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/convert-llama2c-to-ggml/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/cvector-generator/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/deprecation-warning/deprecation-warning.cpp +1 -1
package/src/llama.cpp/examples/embedding/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/eval-callback/CMakeLists.txt +3 -2
package/src/llama.cpp/examples/export-lora/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/gbnf-validator/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/gbnf-validator/gbnf-validator.cpp +4 -7
package/src/llama.cpp/examples/gen-docs/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/gguf/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/gguf-hash/CMakeLists.txt +8 -1
package/src/llama.cpp/examples/gguf-split/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/gguf-split/gguf-split.cpp +2 -2
package/src/llama.cpp/examples/gritlm/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/gritlm/gritlm.cpp +1 -1
package/src/llama.cpp/examples/imatrix/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/imatrix/imatrix.cpp +11 -2
package/src/llama.cpp/examples/infill/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/infill/infill.cpp +1 -1
package/src/llama.cpp/examples/llama-bench/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/llama-bench/llama-bench.cpp +405 -316
package/src/llama.cpp/examples/llama.android/llama/build.gradle.kts +1 -0
package/src/llama.cpp/examples/llava/CMakeLists.txt +10 -3
package/src/llama.cpp/examples/llava/clip.cpp +262 -66
package/src/llama.cpp/examples/llava/clip.h +8 -2
package/src/llama.cpp/examples/llava/llava-cli.cpp +1 -1
package/src/llama.cpp/examples/llava/llava.cpp +46 -19
package/src/llama.cpp/examples/llava/minicpmv-cli.cpp +1 -1
package/src/llama.cpp/examples/llava/qwen2vl-cli.cpp +581 -0
package/src/llama.cpp/examples/lookahead/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/lookahead/lookahead.cpp +1 -1
package/src/llama.cpp/examples/lookup/CMakeLists.txt +4 -4
package/src/llama.cpp/examples/lookup/lookup-stats.cpp +2 -1
package/src/llama.cpp/examples/lookup/lookup.cpp +2 -2
package/src/llama.cpp/examples/main/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/main/main.cpp +9 -5
package/src/llama.cpp/examples/main-cmake-pkg/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/parallel/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/parallel/parallel.cpp +1 -1
package/src/llama.cpp/examples/passkey/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/perplexity/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/quantize/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/quantize/quantize.cpp +0 -3
package/src/llama.cpp/examples/quantize-stats/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/retrieval/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/retrieval/retrieval.cpp +4 -4
package/src/llama.cpp/examples/run/CMakeLists.txt +5 -0
package/src/llama.cpp/examples/run/run.cpp +911 -0
package/src/llama.cpp/examples/save-load-state/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/save-load-state/save-load-state.cpp +4 -4
package/src/llama.cpp/examples/server/CMakeLists.txt +3 -7
package/src/llama.cpp/examples/server/server.cpp +1758 -886
package/src/llama.cpp/examples/server/tests/requirements.txt +2 -2
package/src/llama.cpp/examples/server/utils.hpp +94 -304
package/src/llama.cpp/examples/simple/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/simple/simple.cpp +4 -0
package/src/llama.cpp/examples/simple-chat/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/simple-chat/simple-chat.cpp +3 -0
package/src/llama.cpp/examples/speculative/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/speculative/speculative.cpp +16 -15
package/src/llama.cpp/examples/speculative-simple/CMakeLists.txt +5 -0
package/src/llama.cpp/examples/speculative-simple/speculative-simple.cpp +265 -0
package/src/llama.cpp/examples/tokenize/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/tokenize/tokenize.cpp +1 -1
package/src/llama.cpp/examples/tts/CMakeLists.txt +5 -0
package/src/llama.cpp/examples/tts/tts.cpp +932 -0
package/src/llama.cpp/ggml/CMakeLists.txt +46 -34
package/src/llama.cpp/ggml/include/ggml-backend.h +16 -0
package/src/llama.cpp/ggml/include/ggml-cpu.h +7 -49
package/src/llama.cpp/ggml/include/ggml-opencl.h +26 -0
package/src/llama.cpp/ggml/include/ggml.h +106 -24
package/src/llama.cpp/ggml/src/CMakeLists.txt +73 -24
package/src/llama.cpp/ggml/src/ggml-alloc.c +0 -1
package/src/llama.cpp/ggml/src/ggml-backend-impl.h +51 -11
package/src/llama.cpp/ggml/src/ggml-backend-reg.cpp +379 -22
package/src/llama.cpp/ggml/src/ggml-backend.cpp +4 -4
package/src/llama.cpp/ggml/src/ggml-blas/CMakeLists.txt +3 -7
package/src/llama.cpp/ggml/src/ggml-blas/ggml-blas.cpp +5 -2
package/src/llama.cpp/ggml/src/ggml-cann/CMakeLists.txt +33 -3
package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.cpp +456 -111
package/src/llama.cpp/ggml/src/ggml-cann/common.h +6 -3
package/src/llama.cpp/ggml/src/ggml-cann/ggml-cann.cpp +95 -35
package/src/llama.cpp/ggml/src/ggml-cann/kernels/CMakeLists.txt +2 -5
package/src/llama.cpp/ggml/src/ggml-cann/kernels/dup.cpp +22 -9
package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f16.cpp +24 -13
package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f32.cpp +23 -13
package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q4_0.cpp +11 -0
package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp +10 -0
package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp +10 -0
package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp +17 -0
package/src/llama.cpp/ggml/src/ggml-common.h +42 -42
package/src/llama.cpp/ggml/src/ggml-cpu/CMakeLists.txt +288 -213
package/src/llama.cpp/ggml/src/ggml-cpu/amx/amx.cpp +220 -0
package/src/llama.cpp/ggml/src/ggml-cpu/amx/amx.h +8 -0
package/src/llama.cpp/ggml/src/{ggml-amx → ggml-cpu/amx}/common.h +19 -22
package/src/llama.cpp/ggml/src/{ggml-amx → ggml-cpu/amx}/mmq.cpp +93 -92
package/src/llama.cpp/ggml/src/{ggml-amx → ggml-cpu/amx}/mmq.h +2 -9
package/src/llama.cpp/ggml/src/ggml-cpu/cpu-feats-x86.cpp +323 -0
package/src/llama.cpp/ggml/src/ggml-cpu/{ggml-cpu-aarch64.c → ggml-cpu-aarch64.cpp} +892 -190
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.h +2 -24
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-hbm.cpp +55 -0
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-hbm.h +8 -0
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-impl.h +15 -0
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-quants.c +38 -25
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-traits.cpp +36 -0
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-traits.h +38 -0
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c +552 -399
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.cpp +101 -136
package/src/llama.cpp/ggml/src/ggml-cpu/llamafile/sgemm.cpp +2 -2
package/src/llama.cpp/ggml/src/ggml-cuda/CMakeLists.txt +7 -10
package/src/llama.cpp/ggml/src/ggml-cuda/vendors/hip.h +8 -0
package/src/llama.cpp/ggml/src/ggml-hip/CMakeLists.txt +4 -6
package/src/llama.cpp/ggml/src/ggml-impl.h +32 -11
package/src/llama.cpp/ggml/src/ggml-kompute/CMakeLists.txt +13 -9
package/src/llama.cpp/ggml/src/ggml-kompute/ggml-kompute.cpp +131 -64
package/src/llama.cpp/ggml/src/ggml-metal/CMakeLists.txt +3 -6
package/src/llama.cpp/ggml/src/ggml-metal/ggml-metal-impl.h +39 -0
package/src/llama.cpp/ggml/src/ggml-musa/CMakeLists.txt +14 -7
package/src/llama.cpp/ggml/src/ggml-opencl/CMakeLists.txt +147 -0
package/src/llama.cpp/ggml/src/ggml-opencl/ggml-opencl.cpp +4004 -0
package/src/llama.cpp/ggml/src/ggml-opt.cpp +67 -80
package/src/llama.cpp/ggml/src/ggml-quants.c +0 -9
package/src/llama.cpp/ggml/src/ggml-rpc/CMakeLists.txt +3 -5
package/src/llama.cpp/ggml/src/ggml-rpc/ggml-rpc.cpp +5 -2
package/src/llama.cpp/ggml/src/ggml-sycl/CMakeLists.txt +13 -10
package/src/llama.cpp/ggml/src/ggml-sycl/common.cpp +2 -11
package/src/llama.cpp/ggml/src/ggml-sycl/common.hpp +1 -0
package/src/llama.cpp/ggml/src/ggml-sycl/concat.cpp +2 -2
package/src/llama.cpp/ggml/src/ggml-sycl/convert.cpp +1 -1
package/src/llama.cpp/ggml/src/ggml-sycl/dmmv.cpp +5 -5
package/src/llama.cpp/ggml/src/ggml-sycl/dpct/helper.hpp +32 -13
package/src/llama.cpp/ggml/src/ggml-sycl/element_wise.cpp +80 -61
package/src/llama.cpp/ggml/src/ggml-sycl/gemm.hpp +4 -4
package/src/llama.cpp/ggml/src/ggml-sycl/ggml-sycl.cpp +159 -114
package/src/llama.cpp/ggml/src/ggml-sycl/im2col.cpp +3 -2
package/src/llama.cpp/ggml/src/ggml-sycl/mmq.cpp +6 -6
package/src/llama.cpp/ggml/src/ggml-sycl/mmvq.cpp +6 -20
package/src/llama.cpp/ggml/src/ggml-sycl/norm.cpp +4 -3
package/src/llama.cpp/ggml/src/ggml-sycl/outprod.cpp +8 -8
package/src/llama.cpp/ggml/src/ggml-sycl/rope.cpp +4 -3
package/src/llama.cpp/ggml/src/ggml-sycl/softmax.cpp +7 -7
package/src/llama.cpp/ggml/src/ggml-sycl/tsembd.cpp +1 -0
package/src/llama.cpp/ggml/src/ggml-sycl/wkv6.cpp +4 -1
package/src/llama.cpp/ggml/src/ggml-threading.h +4 -2
package/src/llama.cpp/ggml/src/ggml-vulkan/CMakeLists.txt +21 -7
package/src/llama.cpp/ggml/src/ggml-vulkan/ggml-vulkan.cpp +1718 -399
package/src/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/CMakeLists.txt +3 -1
package/src/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp +105 -31
package/src/llama.cpp/ggml/src/ggml.c +367 -207
package/src/llama.cpp/include/llama-cpp.h +25 -0
package/src/llama.cpp/include/llama.h +26 -19
package/src/llama.cpp/models/ggml-vocab-roberta-bpe.gguf.inp +112 -0
package/src/llama.cpp/models/ggml-vocab-roberta-bpe.gguf.out +46 -0
package/src/llama.cpp/pocs/CMakeLists.txt +3 -1
package/src/llama.cpp/pocs/vdot/CMakeLists.txt +2 -2
package/src/llama.cpp/src/CMakeLists.txt +2 -7
package/src/llama.cpp/src/llama-grammar.cpp +15 -15
package/src/llama.cpp/src/llama-grammar.h +2 -5
package/src/llama.cpp/src/llama-sampling.cpp +35 -90
package/src/llama.cpp/src/llama-vocab.cpp +6 -1
package/src/llama.cpp/src/llama.cpp +1748 -640
package/src/llama.cpp/src/unicode.cpp +62 -51
package/src/llama.cpp/src/unicode.h +9 -10
package/src/llama.cpp/tests/CMakeLists.txt +48 -37
package/src/llama.cpp/tests/test-arg-parser.cpp +2 -2
package/src/llama.cpp/tests/test-backend-ops.cpp +140 -21
package/src/llama.cpp/tests/test-chat-template.cpp +50 -4
package/src/llama.cpp/tests/test-gguf.cpp +1303 -0
package/src/llama.cpp/tests/test-grammar-integration.cpp +3 -6
package/src/llama.cpp/tests/test-llama-grammar.cpp +2 -4
package/src/llama.cpp/tests/test-quantize-fns.cpp +3 -3
package/src/llama.cpp/tests/test-rope.cpp +61 -20
package/src/llama.cpp/tests/test-sampling.cpp +2 -2
package/src/llama.cpp/.github/workflows/nix-ci-aarch64.yml +0 -72
package/src/llama.cpp/.github/workflows/nix-ci.yml +0 -79
package/src/llama.cpp/.github/workflows/nix-flake-update.yml +0 -22
package/src/llama.cpp/.github/workflows/nix-publish-flake.yml +0 -36
package/src/llama.cpp/ggml/include/ggml-amx.h +0 -25
package/src/llama.cpp/ggml/src/ggml-aarch64.c +0 -129
package/src/llama.cpp/ggml/src/ggml-aarch64.h +0 -19
package/src/llama.cpp/ggml/src/ggml-amx/CMakeLists.txt +0 -107
package/src/llama.cpp/ggml/src/ggml-amx/ggml-amx.cpp +0 -446

package/src/llama.cpp/ggml/src/ggml-sycl/element_wise.cpp CHANGED Viewed

@@ -237,7 +237,7 @@ void upscale_f32(const float  *x, float *dst, const int nb00, const int nb01,
     int i02 = i12 / sf2;
     int i03 = i13 / sf3;
-    dst[index] = *(float *)((char *)x + i03 * nb03 + i02 * nb02 + i01 * nb01 + i00 * nb00);
+    dst[index] = *(const float *)((const char *)x + i03 * nb03 + i02 * nb02 + i01 * nb01 + i00 * nb00);
 }
 void pad_f32(const float  *x, float *dst, const int ne0, const int ne00, const int ne01, const int ne02,
@@ -251,8 +251,7 @@ void pad_f32(const float  *x, float *dst, const int ne0, const int ne00, const i
     // operation
     int offset_dst = nidx + item_ct1.get_group(1) * ne0 +
                      item_ct1.get_group(0) * ne0 * item_ct1.get_group_range(1);
-    if (nidx < ne00 && item_ct1.get_group(1) < ne01 &&
-        item_ct1.get_group(0) < ne02) {
+    if (nidx < ne00 && item_ct1.get_group(1) < (size_t) ne01 && item_ct1.get_group(0) < (size_t) ne02) {
         int offset_src = nidx + item_ct1.get_group(1) * ne00 +
                          item_ct1.get_group(0) * ne00 * ne01;
             dst[offset_dst] = x[offset_src];
@@ -520,9 +519,10 @@ inline void ggml_sycl_op_silu(ggml_backend_sycl_context & ctx, const ggml_tensor
     silu_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_gelu(ggml_backend_sycl_context & ctx, const ggml_tensor *src0, const ggml_tensor *src1,
@@ -535,9 +535,10 @@ inline void ggml_sycl_op_gelu(ggml_backend_sycl_context & ctx, const ggml_tensor
     gelu_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_gelu_quick(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
                                     const ggml_tensor *src1, ggml_tensor *dst,
@@ -550,9 +551,10 @@ inline void ggml_sycl_op_gelu_quick(ggml_backend_sycl_context & ctx, const ggml_
     gelu_quick_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_tanh(ggml_backend_sycl_context & ctx, const ggml_tensor *src0, const ggml_tensor *src1,
@@ -564,9 +566,10 @@ inline void ggml_sycl_op_tanh(ggml_backend_sycl_context & ctx, const ggml_tensor
     GGML_ASSERT( dst->type == GGML_TYPE_F32);
     tanh_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_relu(ggml_backend_sycl_context & ctx, const ggml_tensor *src0, const ggml_tensor *src1,
@@ -579,9 +582,10 @@ inline void ggml_sycl_op_relu(ggml_backend_sycl_context & ctx, const ggml_tensor
     relu_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_hardsigmoid(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -595,9 +599,10 @@ inline void ggml_sycl_op_hardsigmoid(ggml_backend_sycl_context & ctx, const ggml
     hardsigmoid_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_hardswish(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -610,9 +615,10 @@ inline void ggml_sycl_op_hardswish(ggml_backend_sycl_context & ctx, const ggml_t
     hardswish_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_exp(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -625,9 +631,10 @@ inline void ggml_sycl_op_exp(ggml_backend_sycl_context & ctx, const ggml_tensor
     exp_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_log(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -640,9 +647,10 @@ inline void ggml_sycl_op_log(ggml_backend_sycl_context & ctx, const ggml_tensor
     log_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_sigmoid(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -655,9 +663,10 @@ inline void ggml_sycl_op_sigmoid(ggml_backend_sycl_context & ctx, const ggml_ten
     sigmoid_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_sqrt(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -670,9 +679,10 @@ inline void ggml_sycl_op_sqrt(ggml_backend_sycl_context & ctx, const ggml_tensor
     sqrt_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_sin(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -685,9 +695,10 @@ inline void ggml_sycl_op_sin(ggml_backend_sycl_context & ctx, const ggml_tensor
     sin_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_cos(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -700,9 +711,10 @@ inline void ggml_sycl_op_cos(ggml_backend_sycl_context & ctx, const ggml_tensor
     cos_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_step(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -715,9 +727,10 @@ inline void ggml_sycl_op_step(ggml_backend_sycl_context & ctx, const ggml_tensor
     step_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_neg(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -730,9 +743,10 @@ inline void ggml_sycl_op_neg(ggml_backend_sycl_context & ctx, const ggml_tensor
     neg_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_leaky_relu(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -749,9 +763,10 @@ inline void ggml_sycl_op_leaky_relu(ggml_backend_sycl_context & ctx, const ggml_
     leaky_relu_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), negative_slope, main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_sqr(ggml_backend_sycl_context & ctx, const ggml_tensor *src0, const ggml_tensor *src1,
@@ -764,9 +779,10 @@ inline void ggml_sycl_op_sqr(ggml_backend_sycl_context & ctx, const ggml_tensor
     sqr_f32_sycl(src0_dd, dst_dd, ggml_nelements(src0), main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_upscale(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
@@ -787,9 +803,10 @@ inline void ggml_sycl_op_upscale(ggml_backend_sycl_context & ctx, const ggml_ten
                      dst->ne[0], dst->ne[1], dst->ne[2], dst->ne[3], sf0, sf1, sf2, sf3,
                      main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_pad(ggml_backend_sycl_context & ctx, const ggml_tensor *src0, const ggml_tensor *src1,
@@ -805,9 +822,10 @@ inline void ggml_sycl_op_pad(ggml_backend_sycl_context & ctx, const ggml_tensor
         src0->ne[0], src0->ne[1], src0->ne[2],
         dst->ne[0], dst->ne[1], dst->ne[2], main_stream);
-    (void) src1;
-    (void) dst;
-    (void) src1_dd;
+    GGML_UNUSED(src1);
+    GGML_UNUSED(dst);
+    GGML_UNUSED(src1_dd);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_acc(ggml_backend_sycl_context & ctx, const ggml_tensor *src0, const ggml_tensor *src1,
@@ -827,7 +845,8 @@ inline void ggml_sycl_op_acc(ggml_backend_sycl_context & ctx, const ggml_tensor
     acc_f32_sycl(src0_dd, src1_dd, dst_dd, ggml_nelements(dst), src1->ne[0], src1->ne[1], src1->ne[2], nb1, nb2, offset, main_stream);
-    (void) dst;
+    GGML_UNUSED(dst);
+    GGML_UNUSED(ctx);
 }
 inline void ggml_sycl_op_add(ggml_backend_sycl_context & ctx, const ggml_tensor *src0, const ggml_tensor *src1,

package/src/llama.cpp/ggml/src/ggml-sycl/gemm.hpp CHANGED Viewed

@@ -51,8 +51,8 @@ public:
         const auto a_in_md = dnnl::memory::desc(a_dims, at, a_trans ? tag::ba : tag::ab);
         const auto b_in_md = dnnl::memory::desc(b_dims, bt, b_trans ? tag::ba : tag::ab);
         const auto c_md = dnnl::memory::desc(c_dims, ct, tag::ab);
-        auto a_mem = dnnl::memory(a_in_md, eng, (void*)a);
-        auto b_mem = dnnl::memory(b_in_md, eng, (void*)b);
+        auto a_mem = dnnl::memory(a_in_md, eng, const_cast<void*>(a));
+        auto b_mem = dnnl::memory(b_in_md, eng, const_cast<void*>(b));
         auto matmul_pd = dnnl::matmul::primitive_desc(eng, a_in_md, b_in_md, c_md);
         auto c_mem = dnnl::memory(matmul_pd.dst_desc(), eng, c);
@@ -79,8 +79,8 @@ public:
         const auto a_in_md = dnnl::memory::desc(a_dims, at, a_trans ? tag::ba : tag::ab);
         const auto b_in_md = dnnl::memory::desc(b_dims, bt, b_trans ? tag::ba : tag::ab);
         const auto c_md = dnnl::memory::desc(c_dims, ct, tag::ab);
-        auto a_mem = dnnl::memory(a_in_md, eng, (void*)a);
-        auto b_mem = dnnl::memory(b_in_md, eng, (void*)b);
+        auto a_mem = dnnl::memory(a_in_md, eng, const_cast<void*>(a));
+        auto b_mem = dnnl::memory(b_in_md, eng, const_cast<void*>(b));
         auto matmul_pd = dnnl::matmul::primitive_desc(eng, a_in_md, b_in_md, c_md);
         auto c_mem = dnnl::memory(matmul_pd.dst_desc(), eng, c);