npm - @fugood/llama.node - Versions diffs - 0.3.15 → 0.3.17 - Mend

@fugood/llama.node 0.3.15 → 0.3.17

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 (203) hide show

package/CMakeLists.txt +3 -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-cuda/arm64/llama-node.node +0 -0
package/bin/linux-cuda/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 +5 -0
package/package.json +1 -1
package/src/LlamaCompletionWorker.cpp +8 -0
package/src/LlamaCompletionWorker.h +1 -0
package/src/LlamaContext.cpp +3 -2
package/src/llama.cpp/.github/workflows/build-linux-cross.yml +124 -0
package/src/llama.cpp/.github/workflows/build.yml +70 -27
package/src/llama.cpp/.github/workflows/docker.yml +6 -6
package/src/llama.cpp/.github/workflows/server.yml +7 -11
package/src/llama.cpp/CMakeLists.txt +23 -1
package/src/llama.cpp/common/CMakeLists.txt +6 -3
package/src/llama.cpp/common/arg.cpp +809 -105
package/src/llama.cpp/common/arg.h +9 -0
package/src/llama.cpp/common/chat.cpp +1 -1
package/src/llama.cpp/common/common.cpp +31 -521
package/src/llama.cpp/common/common.h +17 -36
package/src/llama.cpp/common/json-schema-to-grammar.cpp +3 -0
package/src/llama.cpp/common/llguidance.cpp +30 -47
package/src/llama.cpp/common/minja/chat-template.hpp +15 -7
package/src/llama.cpp/common/minja/minja.hpp +119 -93
package/src/llama.cpp/common/sampling.cpp +3 -0
package/src/llama.cpp/docs/build.md +122 -7
package/src/llama.cpp/examples/CMakeLists.txt +0 -9
package/src/llama.cpp/examples/batched/batched.cpp +1 -1
package/src/llama.cpp/examples/batched-bench/batched-bench.cpp +1 -1
package/src/llama.cpp/examples/embedding/embedding.cpp +7 -1
package/src/llama.cpp/examples/export-lora/export-lora.cpp +1 -1
package/src/llama.cpp/examples/gguf-split/gguf-split.cpp +15 -16
package/src/llama.cpp/examples/gritlm/gritlm.cpp +1 -1
package/src/llama.cpp/examples/llama-bench/llama-bench.cpp +210 -8
package/src/llama.cpp/examples/llama.android/llama/build.gradle.kts +1 -0
package/src/llama.cpp/examples/llava/CMakeLists.txt +39 -24
package/src/llama.cpp/examples/llava/clip-impl.h +345 -0
package/src/llama.cpp/examples/llava/clip.cpp +2152 -1803
package/src/llama.cpp/examples/llava/clip.h +39 -22
package/src/llama.cpp/examples/llava/deprecation-warning.cpp +22 -0
package/src/llama.cpp/examples/llava/llava.cpp +64 -52
package/src/llama.cpp/examples/llava/mtmd-cli.cpp +344 -0
package/src/llama.cpp/examples/llava/mtmd.cpp +708 -0
package/src/llama.cpp/examples/llava/mtmd.h +168 -0
package/src/llama.cpp/examples/llava/{qwen2vl-cli.cpp → qwen2vl-test.cpp} +83 -31
package/src/llama.cpp/examples/main/main.cpp +16 -5
package/src/llama.cpp/examples/parallel/parallel.cpp +3 -1
package/src/llama.cpp/examples/passkey/passkey.cpp +1 -1
package/src/llama.cpp/examples/perplexity/perplexity.cpp +17 -3
package/src/llama.cpp/examples/quantize/quantize.cpp +115 -2
package/src/llama.cpp/examples/rpc/CMakeLists.txt +4 -2
package/src/llama.cpp/examples/rpc/rpc-server.cpp +163 -8
package/src/llama.cpp/examples/run/CMakeLists.txt +12 -1
package/src/llama.cpp/examples/run/run.cpp +14 -28
package/src/llama.cpp/examples/server/httplib.h +313 -247
package/src/llama.cpp/examples/server/server.cpp +243 -139
package/src/llama.cpp/examples/server/utils.hpp +51 -2
package/src/llama.cpp/examples/speculative/speculative.cpp +1 -1
package/src/llama.cpp/examples/speculative-simple/speculative-simple.cpp +1 -1
package/src/llama.cpp/examples/sycl/build.sh +2 -2
package/src/llama.cpp/examples/sycl/win-build-sycl.bat +2 -2
package/src/llama.cpp/examples/tts/tts.cpp +14 -9
package/src/llama.cpp/ggml/CMakeLists.txt +8 -2
package/src/llama.cpp/ggml/cmake/GitVars.cmake +22 -0
package/src/llama.cpp/ggml/include/ggml-cpu.h +5 -0
package/src/llama.cpp/ggml/include/ggml-rpc.h +6 -1
package/src/llama.cpp/ggml/include/ggml.h +66 -99
package/src/llama.cpp/ggml/src/CMakeLists.txt +15 -8
package/src/llama.cpp/ggml/src/ggml-cann/CMakeLists.txt +0 -2
package/src/llama.cpp/ggml/src/ggml-cann/acl_tensor.cpp +8 -4
package/src/llama.cpp/ggml/src/ggml-cann/acl_tensor.h +5 -5
package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.cpp +692 -1534
package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.h +613 -122
package/src/llama.cpp/ggml/src/ggml-cann/common.h +135 -1
package/src/llama.cpp/ggml/src/ggml-cann/ggml-cann.cpp +507 -137
package/src/llama.cpp/ggml/src/ggml-common.h +12 -6
package/src/llama.cpp/ggml/src/ggml-cpu/CMakeLists.txt +48 -22
package/src/llama.cpp/ggml/src/ggml-cpu/binary-ops.cpp +158 -0
package/src/llama.cpp/ggml/src/ggml-cpu/binary-ops.h +16 -0
package/src/llama.cpp/ggml/src/ggml-cpu/common.h +72 -0
package/src/llama.cpp/ggml/src/ggml-cpu/cpu-feats-x86.cpp +1 -1
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp +2413 -228
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-impl.h +2 -21
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-quants.c +754 -404
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c +1004 -13516
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.cpp +2 -0
package/src/llama.cpp/ggml/src/ggml-cpu/kleidiai/kernels.cpp +2 -7
package/src/llama.cpp/ggml/src/ggml-cpu/kleidiai/kernels.h +0 -1
package/src/llama.cpp/ggml/src/ggml-cpu/kleidiai/kleidiai.cpp +3 -4
package/src/llama.cpp/ggml/src/ggml-cpu/llamafile/sgemm.cpp +533 -88
package/src/llama.cpp/ggml/src/ggml-cpu/ops.cpp +8809 -0
package/src/llama.cpp/ggml/src/ggml-cpu/ops.h +110 -0
package/src/llama.cpp/ggml/src/ggml-cpu/simd-mappings.h +892 -0
package/src/llama.cpp/ggml/src/ggml-cpu/unary-ops.cpp +186 -0
package/src/llama.cpp/ggml/src/ggml-cpu/unary-ops.h +28 -0
package/src/llama.cpp/ggml/src/ggml-cpu/vec.cpp +258 -0
package/src/llama.cpp/ggml/src/ggml-cpu/vec.h +802 -0
package/src/llama.cpp/ggml/src/ggml-cuda/vendors/hip.h +7 -0
package/src/llama.cpp/ggml/src/ggml-cuda/vendors/musa.h +1 -0
package/src/llama.cpp/ggml/src/ggml-hip/CMakeLists.txt +0 -4
package/src/llama.cpp/ggml/src/ggml-impl.h +52 -18
package/src/llama.cpp/ggml/src/ggml-metal/ggml-metal-impl.h +70 -3
package/src/llama.cpp/ggml/src/ggml-opencl/CMakeLists.txt +67 -119
package/src/llama.cpp/ggml/src/ggml-opencl/ggml-opencl.cpp +1023 -260
package/src/llama.cpp/ggml/src/ggml-rpc/ggml-rpc.cpp +293 -40
package/src/llama.cpp/ggml/src/ggml-sycl/CMakeLists.txt +127 -33
package/src/llama.cpp/ggml/src/ggml-sycl/backend.hpp +1 -0
package/src/llama.cpp/ggml/src/ggml-sycl/binbcast.cpp +350 -0
package/src/llama.cpp/ggml/src/ggml-sycl/binbcast.hpp +39 -0
package/src/llama.cpp/ggml/src/ggml-sycl/common.cpp +0 -35
package/src/llama.cpp/ggml/src/ggml-sycl/common.hpp +29 -293
package/src/llama.cpp/ggml/src/ggml-sycl/dpct/helper.hpp +79 -90
package/src/llama.cpp/ggml/src/ggml-sycl/element_wise.cpp +967 -438
package/src/llama.cpp/ggml/src/ggml-sycl/element_wise.hpp +22 -23
package/src/llama.cpp/ggml/src/ggml-sycl/gemm.hpp +12 -43
package/src/llama.cpp/ggml/src/ggml-sycl/getrows.cpp +24 -20
package/src/llama.cpp/ggml/src/ggml-sycl/getrows.hpp +1 -4
package/src/llama.cpp/ggml/src/ggml-sycl/ggml-sycl.cpp +210 -286
package/src/llama.cpp/ggml/src/ggml-sycl/im2col.cpp +84 -74
package/src/llama.cpp/ggml/src/ggml-sycl/im2col.hpp +1 -3
package/src/llama.cpp/ggml/src/ggml-sycl/norm.cpp +37 -49
package/src/llama.cpp/ggml/src/ggml-sycl/norm.hpp +7 -22
package/src/llama.cpp/ggml/src/ggml-sycl/outprod.cpp +4 -14
package/src/llama.cpp/ggml/src/ggml-sycl/rope.cpp +204 -118
package/src/llama.cpp/ggml/src/ggml-sycl/rope.hpp +1 -3
package/src/llama.cpp/ggml/src/ggml-vulkan/CMakeLists.txt +23 -0
package/src/llama.cpp/ggml/src/ggml-vulkan/ggml-vulkan.cpp +692 -126
package/src/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/CMakeLists.txt +12 -0
package/src/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp +21 -10
package/src/llama.cpp/ggml/src/ggml.c +141 -245
package/src/llama.cpp/ggml/src/gguf.cpp +1 -0
package/src/llama.cpp/include/llama.h +30 -11
package/src/llama.cpp/models/ggml-vocab-llama4.gguf.inp +112 -0
package/src/llama.cpp/models/ggml-vocab-llama4.gguf.out +46 -0
package/src/llama.cpp/models/ggml-vocab-pixtral.gguf.inp +112 -0
package/src/llama.cpp/models/ggml-vocab-pixtral.gguf.out +46 -0
package/src/llama.cpp/requirements/requirements-all.txt +2 -0
package/src/llama.cpp/requirements/requirements-gguf_editor_gui.txt +3 -0
package/src/llama.cpp/src/CMakeLists.txt +3 -2
package/src/llama.cpp/src/llama-adapter.cpp +37 -1
package/src/llama.cpp/src/llama-arch.cpp +161 -17
package/src/llama.cpp/src/llama-arch.h +16 -0
package/src/llama.cpp/src/llama-chat.cpp +82 -17
package/src/llama.cpp/src/llama-chat.h +6 -2
package/src/llama.cpp/src/llama-context.cpp +108 -92
package/src/llama.cpp/src/llama-context.h +1 -2
package/src/llama.cpp/src/llama-graph.cpp +189 -119
package/src/llama.cpp/src/llama-graph.h +26 -6
package/src/llama.cpp/src/llama-hparams.h +13 -0
package/src/llama.cpp/src/llama-kv-cache.cpp +70 -123
package/src/llama.cpp/src/llama-kv-cache.h +41 -115
package/src/llama.cpp/src/llama-memory.h +1 -1
package/src/llama.cpp/src/llama-mmap.cpp +1 -1
package/src/llama.cpp/src/llama-model-loader.cpp +10 -5
package/src/llama.cpp/src/llama-model-loader.h +5 -3
package/src/llama.cpp/src/llama-model.cpp +1544 -291
package/src/llama.cpp/src/llama-model.h +13 -1
package/src/llama.cpp/src/llama-quant.cpp +29 -8
package/src/llama.cpp/src/llama-sampling.cpp +7 -1
package/src/llama.cpp/src/llama-vocab.cpp +44 -6
package/src/llama.cpp/src/llama.cpp +1 -1
package/src/llama.cpp/tests/CMakeLists.txt +43 -30
package/src/llama.cpp/tests/test-arg-parser.cpp +51 -4
package/src/llama.cpp/tests/test-backend-ops.cpp +139 -57
package/src/llama.cpp/tests/test-chat-template.cpp +34 -13
package/src/llama.cpp/tests/test-chat.cpp +12 -2
package/src/llama.cpp/{examples/gbnf-validator/gbnf-validator.cpp → tests/test-gbnf-validator.cpp} +2 -2
package/src/llama.cpp/tests/test-grammar-integration.cpp +3 -2
package/src/llama.cpp/tests/test-grammar-llguidance.cpp +63 -2
package/src/llama.cpp/tests/test-grammar-parser.cpp +3 -1
package/src/llama.cpp/tests/test-json-schema-to-grammar.cpp +17 -1
package/src/llama.cpp/tests/test-llama-grammar.cpp +2 -1
package/src/llama.cpp/{examples/quantize-stats/quantize-stats.cpp → tests/test-quantize-stats.cpp} +3 -1
package/src/llama.cpp/tests/test-tokenizer-1-bpe.cpp +2 -1
package/src/llama.cpp/tests/test-tokenizer-1-spm.cpp +2 -1
package/src/llama.cpp/examples/gbnf-validator/CMakeLists.txt +0 -5
package/src/llama.cpp/examples/llava/gemma3-cli.cpp +0 -341
package/src/llama.cpp/examples/llava/llava-cli.cpp +0 -332
package/src/llama.cpp/examples/llava/minicpmv-cli.cpp +0 -354
package/src/llama.cpp/examples/quantize-stats/CMakeLists.txt +0 -6
package/src/llama.cpp/ggml/src/ggml-cann/kernels/CMakeLists.txt +0 -30
package/src/llama.cpp/ggml/src/ggml-cann/kernels/ascendc_kernels.h +0 -19
package/src/llama.cpp/ggml/src/ggml-cann/kernels/dup.cpp +0 -234
package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f16.cpp +0 -197
package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f32.cpp +0 -190
package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q4_0.cpp +0 -204
package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q8_0.cpp +0 -191
package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp +0 -218
package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp +0 -216
package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp +0 -295

package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-impl.h CHANGED Viewed

@@ -4,13 +4,13 @@
 #include "ggml.h"
 #include "ggml-impl.h"
 #include <stdlib.h> // load `stdlib.h` before other headers to work around MinGW bug: https://sourceforge.net/p/mingw-w64/bugs/192/
 //#include <stddef.h>
 #include <stdbool.h>
 #include <string.h> // memcpy
 #include <math.h>   // fabsf
 #ifdef __cplusplus
 extern "C" {
 #endif
@@ -69,33 +69,16 @@ struct ggml_compute_params {
 #endif
 #if defined(__ARM_FEATURE_SVE)
-#include <arm_sve.h>
 #include <sys/prctl.h>
 #endif
-// 16-bit float
-// on Arm, we use __fp16
-// on x86, we use uint16_t
 #if defined(__ARM_NEON)
-// if YCM cannot find <arm_neon.h>, make a symbolic link to it, for example:
-//
-//   $ ln -sfn /Library/Developer/CommandLineTools/usr/lib/clang/13.1.6/include/arm_neon.h ./src/
-//
-#include <arm_neon.h>
+// ref: https://github.com/ggml-org/llama.cpp/pull/5404
 #ifdef _MSC_VER
-typedef uint16_t ggml_fp16_internal_t;
 #define ggml_vld1q_u32(w,x,y,z) { ((w) + ((uint64_t)(x) << 32)), ((y) + ((uint64_t)(z) << 32)) }
 #else
-typedef __fp16 ggml_fp16_internal_t;
 #define ggml_vld1q_u32(w,x,y,z) { (w), (x), (y), (z) }
 #endif // _MSC_VER
 #if !defined(__aarch64__)
@@ -340,8 +323,6 @@ inline static int32x4_t ggml_vdotq_s32(int32x4_t acc, int8x16_t a, int8x16_t b)
 #else
 #ifdef __POWER9_VECTOR__
 #include <altivec.h>
-#undef bool
-#define bool _Bool
 #else
 #if defined(_MSC_VER) || defined(__MINGW32__)
 #include <intrin.h>