npm - @fugood/llama.node - Versions diffs - 0.3.12 → 0.3.14 - Mend

@fugood/llama.node 0.3.12 → 0.3.14

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 (159) hide show

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 +2 -1
package/package.json +1 -1
package/src/LlamaCompletionWorker.cpp +14 -0
package/src/LlamaContext.cpp +110 -79
package/src/LlamaContext.h +1 -1
package/src/common.hpp +1 -2
package/src/llama.cpp/.github/workflows/build.yml +95 -13
package/src/llama.cpp/.github/workflows/docker.yml +2 -0
package/src/llama.cpp/.github/workflows/labeler.yml +1 -1
package/src/llama.cpp/.github/workflows/server.yml +2 -0
package/src/llama.cpp/common/CMakeLists.txt +23 -6
package/src/llama.cpp/common/arg.cpp +292 -14
package/src/llama.cpp/common/chat.cpp +1128 -315
package/src/llama.cpp/common/chat.h +135 -0
package/src/llama.cpp/common/common.cpp +27 -171
package/src/llama.cpp/common/common.h +41 -73
package/src/llama.cpp/common/json-schema-to-grammar.cpp +4 -5
package/src/llama.cpp/common/json-schema-to-grammar.h +0 -1
package/src/llama.cpp/common/llguidance.cpp +3 -3
package/src/llama.cpp/common/log.cpp +1 -0
package/src/llama.cpp/common/log.h +2 -1
package/src/llama.cpp/common/{chat-template.hpp → minja/chat-template.hpp} +21 -7
package/src/llama.cpp/common/{minja.hpp → minja/minja.hpp} +61 -14
package/src/llama.cpp/common/ngram-cache.cpp +1 -0
package/src/llama.cpp/common/sampling.cpp +93 -49
package/src/llama.cpp/common/speculative.cpp +6 -5
package/src/llama.cpp/common/speculative.h +1 -1
package/src/llama.cpp/docs/build.md +47 -9
package/src/llama.cpp/examples/cvector-generator/cvector-generator.cpp +3 -1
package/src/llama.cpp/examples/embedding/embedding.cpp +1 -0
package/src/llama.cpp/examples/export-lora/export-lora.cpp +4 -2
package/src/llama.cpp/examples/imatrix/imatrix.cpp +4 -4
package/src/llama.cpp/examples/llama-bench/llama-bench.cpp +6 -5
package/src/llama.cpp/examples/llama.android/llama/src/main/cpp/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/llama.android/llama/src/main/cpp/llama-android.cpp +1 -1
package/src/llama.cpp/examples/llava/CMakeLists.txt +7 -0
package/src/llama.cpp/examples/llava/clip.cpp +373 -107
package/src/llama.cpp/examples/llava/clip.h +19 -3
package/src/llama.cpp/examples/llava/gemma3-cli.cpp +341 -0
package/src/llama.cpp/examples/llava/llava.cpp +4 -2
package/src/llama.cpp/examples/llava/minicpmv-cli.cpp +30 -11
package/src/llama.cpp/examples/lookahead/lookahead.cpp +1 -0
package/src/llama.cpp/examples/main/main.cpp +73 -28
package/src/llama.cpp/examples/parallel/parallel.cpp +1 -0
package/src/llama.cpp/examples/passkey/passkey.cpp +1 -0
package/src/llama.cpp/examples/perplexity/perplexity.cpp +1 -0
package/src/llama.cpp/examples/quantize/quantize.cpp +1 -0
package/src/llama.cpp/examples/run/linenoise.cpp/linenoise.cpp +882 -237
package/src/llama.cpp/examples/run/linenoise.cpp/linenoise.h +35 -26
package/src/llama.cpp/examples/run/run.cpp +115 -79
package/src/llama.cpp/examples/server/CMakeLists.txt +1 -1
package/src/llama.cpp/examples/server/httplib.h +381 -292
package/src/llama.cpp/examples/server/server.cpp +134 -128
package/src/llama.cpp/examples/server/utils.hpp +95 -106
package/src/llama.cpp/examples/sycl/run-llama2.sh +2 -2
package/src/llama.cpp/examples/tts/tts.cpp +251 -142
package/src/llama.cpp/ggml/CMakeLists.txt +13 -1
package/src/llama.cpp/ggml/include/ggml-alloc.h +1 -1
package/src/llama.cpp/ggml/include/ggml-backend.h +3 -3
package/src/llama.cpp/ggml/include/ggml-cpu.h +4 -1
package/src/llama.cpp/ggml/include/ggml-metal.h +1 -1
package/src/llama.cpp/ggml/include/ggml-vulkan.h +0 -2
package/src/llama.cpp/ggml/include/ggml.h +6 -2
package/src/llama.cpp/ggml/src/CMakeLists.txt +10 -7
package/src/llama.cpp/ggml/src/ggml-alloc.c +24 -15
package/src/llama.cpp/ggml/src/ggml-backend-impl.h +1 -1
package/src/llama.cpp/ggml/src/ggml-backend-reg.cpp +58 -54
package/src/llama.cpp/ggml/src/ggml-backend.cpp +10 -8
package/src/llama.cpp/ggml/src/ggml-cann/ggml-cann.cpp +3 -2
package/src/llama.cpp/ggml/src/ggml-cann/kernels/dup.cpp +3 -5
package/src/llama.cpp/ggml/src/ggml-common.h +0 -2
package/src/llama.cpp/ggml/src/ggml-cpu/CMakeLists.txt +132 -17
package/src/llama.cpp/ggml/src/ggml-cpu/amx/amx.cpp +2 -1
package/src/llama.cpp/ggml/src/ggml-cpu/cpu-feats-x86.cpp +4 -0
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp +2 -1
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-impl.h +156 -11
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-quants.c +2235 -641
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c +1572 -198
package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.cpp +24 -5
package/src/llama.cpp/ggml/src/ggml-cpu/kleidiai/kernels.cpp +259 -0
package/src/llama.cpp/ggml/src/ggml-cpu/kleidiai/kernels.h +61 -0
package/src/llama.cpp/ggml/src/ggml-cpu/kleidiai/kleidiai.cpp +288 -0
package/src/llama.cpp/ggml/src/ggml-cpu/kleidiai/kleidiai.h +17 -0
package/src/llama.cpp/ggml/src/ggml-cpu/llamafile/sgemm.cpp +9 -8
package/src/llama.cpp/ggml/src/ggml-cuda/CMakeLists.txt +16 -3
package/src/llama.cpp/ggml/src/ggml-hip/CMakeLists.txt +14 -0
package/src/llama.cpp/ggml/src/ggml-impl.h +1 -1
package/src/llama.cpp/ggml/src/ggml-metal/CMakeLists.txt +4 -5
package/src/llama.cpp/ggml/src/ggml-metal/ggml-metal-impl.h +235 -0
package/src/llama.cpp/ggml/src/ggml-musa/CMakeLists.txt +6 -2
package/src/llama.cpp/ggml/src/ggml-opencl/CMakeLists.txt +1 -0
package/src/llama.cpp/ggml/src/ggml-opencl/ggml-opencl.cpp +246 -120
package/src/llama.cpp/ggml/src/ggml-quants.c +114 -114
package/src/llama.cpp/ggml/src/ggml-rpc/ggml-rpc.cpp +2 -1
package/src/llama.cpp/ggml/src/ggml-sycl/CMakeLists.txt +2 -0
package/src/llama.cpp/ggml/src/ggml-sycl/backend.hpp +1 -0
package/src/llama.cpp/ggml/src/ggml-sycl/common.cpp +17 -0
package/src/llama.cpp/ggml/src/ggml-sycl/common.hpp +51 -10
package/src/llama.cpp/ggml/src/ggml-sycl/convert.cpp +33 -4
package/src/llama.cpp/ggml/src/ggml-sycl/convert.hpp +2 -2
package/src/llama.cpp/ggml/src/ggml-sycl/cpy.cpp +701 -0
package/src/llama.cpp/ggml/src/ggml-sycl/cpy.hpp +11 -0
package/src/llama.cpp/ggml/src/ggml-sycl/dequantize.hpp +55 -0
package/src/llama.cpp/ggml/src/ggml-sycl/dmmv.cpp +136 -4
package/src/llama.cpp/ggml/src/ggml-sycl/getrows.cpp +308 -0
package/src/llama.cpp/ggml/src/ggml-sycl/getrows.hpp +23 -0
package/src/llama.cpp/ggml/src/ggml-sycl/ggml-sycl.cpp +174 -728
package/src/llama.cpp/ggml/src/ggml-sycl/mmvq.cpp +75 -77
package/src/llama.cpp/ggml/src/ggml-sycl/softmax.cpp +3 -0
package/src/llama.cpp/ggml/src/ggml-sycl/sycl_hw.cpp +13 -0
package/src/llama.cpp/ggml/src/ggml-sycl/sycl_hw.hpp +23 -0
package/src/llama.cpp/ggml/src/ggml-vulkan/ggml-vulkan.cpp +949 -602
package/src/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp +37 -3
package/src/llama.cpp/ggml/src/ggml.c +9 -4
package/src/llama.cpp/include/llama.h +32 -14
package/src/llama.cpp/models/ggml-vocab-gpt-4o.gguf.inp +112 -0
package/src/llama.cpp/models/ggml-vocab-gpt-4o.gguf.out +46 -0
package/src/llama.cpp/requirements/requirements-all.txt +1 -0
package/src/llama.cpp/requirements/requirements-tool_bench.txt +12 -0
package/src/llama.cpp/requirements.txt +1 -0
package/src/llama.cpp/src/llama-arch.cpp +21 -0
package/src/llama.cpp/src/llama-arch.h +1 -0
package/src/llama.cpp/src/llama-chat.cpp +1 -0
package/src/llama.cpp/src/llama-grammar.cpp +183 -183
package/src/llama.cpp/src/llama-grammar.h +13 -4
package/src/llama.cpp/src/llama-impl.h +6 -6
package/src/llama.cpp/src/llama-kv-cache.h +2 -1
package/src/llama.cpp/src/llama-mmap.cpp +11 -1
package/src/llama.cpp/src/llama-mmap.h +1 -0
package/src/llama.cpp/src/llama-model.cpp +70 -6
package/src/llama.cpp/src/llama-sampling.cpp +174 -67
package/src/llama.cpp/src/llama-vocab.cpp +12 -0
package/src/llama.cpp/src/llama.cpp +154 -5
package/src/llama.cpp/src/unicode.cpp +9 -2
package/src/llama.cpp/tests/test-backend-ops.cpp +171 -115
package/src/llama.cpp/tests/test-chat-template.cpp +32 -22
package/src/llama.cpp/tests/test-chat.cpp +691 -325
package/src/llama.cpp/tests/test-gguf.cpp +4 -4
package/src/llama.cpp/tests/test-json-schema-to-grammar.cpp +63 -63
package/src/llama.cpp/tests/test-quantize-fns.cpp +1 -9
package/src/llama.cpp/tests/test-sampling.cpp +15 -0
package/src/llama.cpp/Sources/llama/llama.h +0 -4
package/src/llama.cpp/common/chat.hpp +0 -52

package/src/llama.cpp/ggml/src/ggml-musa/CMakeLists.txt CHANGED Viewed

@@ -21,7 +21,7 @@ if (MUSAToolkit_FOUND)
     message(STATUS "MUSA Toolkit found")
     if (NOT DEFINED MUSA_ARCHITECTURES)
-        set(MUSA_ARCHITECTURES "21;22")
+        set(MUSA_ARCHITECTURES "21;22;31")
     endif()
     message(STATUS "Using MUSA architectures: ${MUSA_ARCHITECTURES}")
@@ -49,7 +49,7 @@ if (MUSAToolkit_FOUND)
     set_source_files_properties(${GGML_SOURCES_MUSA} PROPERTIES LANGUAGE CXX)
     foreach(SOURCE ${GGML_SOURCES_MUSA})
-        set(COMPILE_FLAGS "-x musa -mtgpu")
+        set(COMPILE_FLAGS "-fsigned-char -x musa -mtgpu")
         foreach(ARCH ${MUSA_ARCHITECTURES})
             set(COMPILE_FLAGS "${COMPILE_FLAGS} --cuda-gpu-arch=mp_${ARCH}")
         endforeach()
@@ -83,6 +83,10 @@ if (MUSAToolkit_FOUND)
         add_compile_definitions(GGML_CUDA_NO_VMM)
     endif()
+    if (NOT GGML_CUDA_FA)
+        add_compile_definitions(GGML_CUDA_NO_FA)
+    endif()
     if (GGML_CUDA_F16 OR GGML_CUDA_DMMV_F16)
         add_compile_definitions(GGML_CUDA_F16)
     endif()

package/src/llama.cpp/ggml/src/ggml-opencl/CMakeLists.txt CHANGED Viewed

@@ -15,6 +15,7 @@ if (GGML_OPENCL_PROFILING)
 endif ()
 add_compile_definitions(GGML_OPENCL_SOA_Q)
+add_compile_definitions(GGML_OPENCL_TARGET_VERSION=${GGML_OPENCL_TARGET_VERSION})
 if (GGML_OPENCL_USE_ADRENO_KERNELS)
     message(STATUS "OpenCL will use matmul kernels optimized for Adreno")