@fugood/llama.node 0.2.2 → 0.3.0
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.
- package/CMakeLists.txt +5 -2
- 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 +8 -1
- package/package.json +1 -1
- package/patches/llama.patch +12 -12
- package/src/DetokenizeWorker.cpp +1 -1
- package/src/LlamaContext.cpp +33 -1
- package/src/LlamaContext.h +1 -0
- package/src/LoadSessionWorker.cpp +1 -0
- package/src/llama.cpp/.github/workflows/bench.yml +310 -0
- package/src/llama.cpp/.github/workflows/build.yml +1315 -0
- package/src/llama.cpp/.github/workflows/close-issue.yml +23 -0
- package/src/llama.cpp/.github/workflows/docker.yml +116 -0
- package/src/llama.cpp/.github/workflows/editorconfig.yml +27 -0
- package/src/llama.cpp/.github/workflows/gguf-publish.yml +44 -0
- package/src/llama.cpp/.github/workflows/labeler.yml +17 -0
- package/src/llama.cpp/.github/workflows/nix-ci-aarch64.yml +65 -0
- package/src/llama.cpp/.github/workflows/nix-ci.yml +72 -0
- package/src/llama.cpp/.github/workflows/nix-flake-update.yml +22 -0
- package/src/llama.cpp/.github/workflows/nix-publish-flake.yml +36 -0
- package/src/llama.cpp/.github/workflows/python-check-requirements.yml +35 -0
- package/src/llama.cpp/.github/workflows/python-lint.yml +23 -0
- package/src/llama.cpp/.github/workflows/python-type-check.yml +38 -0
- package/src/llama.cpp/.github/workflows/server.yml +183 -0
- package/src/llama.cpp/CMakeLists.txt +91 -1245
- package/src/llama.cpp/cmake/arm64-windows-llvm.cmake +1 -1
- package/src/llama.cpp/cmake/build-info.cmake +58 -0
- package/src/llama.cpp/cmake/git-vars.cmake +22 -0
- package/src/llama.cpp/common/CMakeLists.txt +4 -3
- package/src/llama.cpp/common/build-info.cpp.in +4 -0
- package/src/llama.cpp/common/common.cpp +1116 -877
- package/src/llama.cpp/common/common.h +191 -77
- package/src/llama.cpp/common/grammar-parser.cpp +118 -31
- package/src/llama.cpp/common/json-schema-to-grammar.cpp +346 -65
- package/src/llama.cpp/common/log.h +1 -1
- package/src/llama.cpp/common/ngram-cache.h +10 -3
- package/src/llama.cpp/common/sampling.cpp +19 -10
- package/src/llama.cpp/docs/build.md +353 -0
- package/src/llama.cpp/examples/CMakeLists.txt +22 -22
- package/src/llama.cpp/examples/baby-llama/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/baby-llama/baby-llama.cpp +6 -6
- package/src/llama.cpp/examples/batched/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/batched/batched.cpp +52 -55
- package/src/llama.cpp/examples/batched-bench/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/batched-bench/batched-bench.cpp +20 -72
- package/src/llama.cpp/examples/benchmark/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/chat-13B.bat +57 -0
- package/src/llama.cpp/examples/convert-llama2c-to-ggml/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/{finetune → cvector-generator}/CMakeLists.txt +2 -2
- package/src/llama.cpp/examples/cvector-generator/completions.txt +582 -0
- package/src/llama.cpp/examples/cvector-generator/cvector-generator.cpp +503 -0
- package/src/llama.cpp/examples/cvector-generator/mean.hpp +48 -0
- package/src/llama.cpp/examples/cvector-generator/negative.txt +4 -0
- package/src/llama.cpp/examples/cvector-generator/pca.hpp +325 -0
- package/src/llama.cpp/examples/cvector-generator/positive.txt +4 -0
- package/src/llama.cpp/examples/deprecation-warning/deprecation-warning.cpp +35 -0
- package/src/llama.cpp/examples/embedding/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/embedding/embedding.cpp +94 -46
- package/src/llama.cpp/examples/eval-callback/CMakeLists.txt +2 -2
- package/src/llama.cpp/examples/eval-callback/eval-callback.cpp +4 -6
- package/src/llama.cpp/examples/export-lora/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/export-lora/export-lora.cpp +344 -386
- package/src/llama.cpp/examples/gbnf-validator/CMakeLists.txt +2 -2
- package/src/llama.cpp/examples/gbnf-validator/gbnf-validator.cpp +30 -25
- package/src/llama.cpp/examples/gguf/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/gguf/gguf.cpp +5 -0
- package/src/llama.cpp/examples/gguf-hash/CMakeLists.txt +15 -0
- package/src/llama.cpp/examples/gguf-hash/deps/rotate-bits/rotate-bits.h +46 -0
- package/src/llama.cpp/examples/gguf-hash/deps/sha1/sha1.c +295 -0
- package/src/llama.cpp/examples/gguf-hash/deps/sha1/sha1.h +52 -0
- package/src/llama.cpp/examples/gguf-hash/deps/sha256/sha256.c +221 -0
- package/src/llama.cpp/examples/gguf-hash/deps/sha256/sha256.h +24 -0
- package/src/llama.cpp/examples/gguf-hash/deps/xxhash/xxhash.c +42 -0
- package/src/llama.cpp/examples/gguf-hash/deps/xxhash/xxhash.h +7093 -0
- package/src/llama.cpp/examples/gguf-hash/gguf-hash.cpp +693 -0
- package/src/llama.cpp/examples/gguf-split/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/gguf-split/gguf-split.cpp +3 -3
- package/src/llama.cpp/examples/gritlm/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/gritlm/gritlm.cpp +6 -2
- package/src/llama.cpp/examples/imatrix/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/imatrix/imatrix.cpp +137 -176
- package/src/llama.cpp/examples/infill/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/infill/infill.cpp +38 -153
- package/src/llama.cpp/examples/llama-bench/llama-bench.cpp +175 -94
- package/src/llama.cpp/examples/llama.android/app/build.gradle.kts +65 -0
- package/src/llama.cpp/examples/llama.android/build.gradle.kts +6 -0
- package/src/llama.cpp/examples/llama.android/llama/build.gradle.kts +68 -0
- package/src/llama.cpp/examples/llama.android/llama/src/main/cpp/CMakeLists.txt +11 -7
- package/src/llama.cpp/examples/llama.android/llama/src/main/cpp/llama-android.cpp +2 -2
- package/src/llama.cpp/examples/llama.android/settings.gradle.kts +18 -0
- package/src/llama.cpp/examples/llava/CMakeLists.txt +6 -5
- package/src/llama.cpp/examples/llava/android/build_64.sh +8 -0
- package/src/llama.cpp/examples/llava/clip.cpp +23 -14
- package/src/llama.cpp/examples/llava/llava-cli.cpp +8 -6
- package/src/llama.cpp/examples/llava/requirements.txt +3 -2
- package/src/llama.cpp/examples/lookahead/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/lookahead/lookahead.cpp +2 -1
- package/src/llama.cpp/examples/lookup/CMakeLists.txt +4 -4
- package/src/llama.cpp/examples/lookup/lookup-create.cpp +2 -0
- package/src/llama.cpp/examples/lookup/lookup-merge.cpp +4 -4
- package/src/llama.cpp/examples/lookup/lookup-stats.cpp +2 -2
- package/src/llama.cpp/examples/lookup/lookup.cpp +1 -1
- package/src/llama.cpp/examples/main/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/main/main.cpp +98 -75
- package/src/llama.cpp/examples/main-cmake-pkg/CMakeLists.txt +4 -5
- package/src/llama.cpp/examples/parallel/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/parallel/parallel.cpp +2 -1
- package/src/llama.cpp/examples/passkey/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/passkey/passkey.cpp +23 -43
- package/src/llama.cpp/examples/perplexity/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/perplexity/perplexity.cpp +13 -10
- package/src/llama.cpp/examples/quantize/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/quantize/quantize.cpp +37 -34
- package/src/llama.cpp/examples/quantize-stats/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/quantize-stats/quantize-stats.cpp +1 -1
- package/src/llama.cpp/examples/retrieval/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/retrieval/retrieval.cpp +26 -77
- package/src/llama.cpp/examples/save-load-state/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/save-load-state/save-load-state.cpp +14 -7
- package/src/llama.cpp/examples/server/CMakeLists.txt +26 -2
- package/src/llama.cpp/examples/server/server.cpp +274 -671
- package/src/llama.cpp/examples/server/tests/requirements.txt +2 -2
- package/src/llama.cpp/examples/server/utils.hpp +28 -29
- package/src/llama.cpp/examples/simple/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/simple/simple.cpp +21 -29
- package/src/llama.cpp/examples/speculative/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/speculative/speculative.cpp +2 -1
- package/src/llama.cpp/examples/sycl/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/sycl/build.sh +23 -0
- package/src/llama.cpp/examples/sycl/run-llama2.sh +36 -0
- package/src/llama.cpp/examples/sycl/win-build-sycl.bat +33 -0
- package/src/llama.cpp/examples/sycl/win-run-llama2.bat +9 -0
- package/src/llama.cpp/examples/tokenize/CMakeLists.txt +1 -1
- package/src/llama.cpp/examples/tokenize/tokenize.cpp +16 -2
- package/src/llama.cpp/ggml/CMakeLists.txt +253 -0
- package/src/llama.cpp/{cmake → ggml/cmake}/FindSIMD.cmake +6 -6
- package/src/llama.cpp/{ggml-backend.h → ggml/include/ggml-backend.h} +22 -17
- package/src/llama.cpp/ggml/include/ggml-blas.h +23 -0
- package/src/llama.cpp/ggml/include/ggml-cann.h +125 -0
- package/src/llama.cpp/{ggml-cuda.h → ggml/include/ggml-cuda.h} +3 -0
- package/src/llama.cpp/{ggml-metal.h → ggml/include/ggml-metal.h} +1 -2
- package/src/llama.cpp/{ggml-sycl.h → ggml/include/ggml-sycl.h} +3 -10
- package/src/llama.cpp/{ggml.h → ggml/include/ggml.h} +80 -85
- package/src/llama.cpp/ggml/src/CMakeLists.txt +1329 -0
- package/src/llama.cpp/ggml/src/ggml-aarch64.c +2193 -0
- package/src/llama.cpp/ggml/src/ggml-aarch64.h +39 -0
- package/src/llama.cpp/{ggml-alloc.c → ggml/src/ggml-alloc.c} +100 -49
- package/src/llama.cpp/{ggml-backend-impl.h → ggml/src/ggml-backend-impl.h} +20 -8
- package/src/llama.cpp/{ggml-backend.c → ggml/src/ggml-backend.c} +307 -167
- package/src/llama.cpp/ggml/src/ggml-blas.cpp +367 -0
- package/src/llama.cpp/ggml/src/ggml-cann/acl_tensor.cpp +198 -0
- package/src/llama.cpp/ggml/src/ggml-cann/acl_tensor.h +230 -0
- package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.cpp +2944 -0
- package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.h +592 -0
- package/src/llama.cpp/ggml/src/ggml-cann/common.h +282 -0
- package/src/llama.cpp/ggml/src/ggml-cann/kernels/CMakeLists.txt +32 -0
- package/src/llama.cpp/ggml/src/ggml-cann/kernels/ascendc_kernels.h +17 -0
- package/src/llama.cpp/ggml/src/ggml-cann/kernels/dup.cpp +223 -0
- package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f16.cpp +186 -0
- package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f32.cpp +180 -0
- package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q4_0.cpp +193 -0
- package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q8_0.cpp +191 -0
- package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp +208 -0
- package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp +206 -0
- package/src/llama.cpp/ggml/src/ggml-cann.cpp +2023 -0
- package/src/llama.cpp/{ggml-common.h → ggml/src/ggml-common.h} +41 -7
- package/src/llama.cpp/{ggml-impl.h → ggml/src/ggml-impl.h} +113 -9
- package/src/llama.cpp/{ggml-kompute.cpp → ggml/src/ggml-kompute.cpp} +33 -18
- package/src/llama.cpp/{ggml-quants.c → ggml/src/ggml-quants.c} +1460 -940
- package/src/llama.cpp/{ggml-quants.h → ggml/src/ggml-quants.h} +19 -20
- package/src/llama.cpp/{ggml-rpc.cpp → ggml/src/ggml-rpc.cpp} +95 -72
- package/src/llama.cpp/ggml/src/ggml-sycl/backend.hpp +27 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/common.cpp +53 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/common.hpp +355 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/concat.cpp +195 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/concat.hpp +21 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/convert.cpp +547 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/convert.hpp +27 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/dequantize.hpp +698 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/dmmv.cpp +1023 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/dmmv.hpp +27 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/dpct/helper.hpp +3011 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/mmq.cpp +3031 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/mmq.hpp +33 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/mmvq.cpp +1027 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/mmvq.hpp +27 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/norm.cpp +374 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/norm.hpp +35 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/presets.hpp +66 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/rope.cpp +275 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/rope.hpp +22 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/softmax.cpp +251 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/softmax.hpp +24 -0
- package/src/llama.cpp/ggml/src/ggml-sycl/vecdotq.hpp +1140 -0
- package/src/llama.cpp/ggml/src/ggml-sycl.cpp +5314 -0
- package/src/llama.cpp/{ggml-vulkan.cpp → ggml/src/ggml-vulkan.cpp} +1781 -1868
- package/src/llama.cpp/{ggml.c → ggml/src/ggml.c} +1245 -2087
- package/src/llama.cpp/{sgemm.cpp → ggml/src/llamafile/sgemm.cpp} +21 -24
- package/src/llama.cpp/{sgemm.h → ggml/src/llamafile/sgemm.h} +1 -1
- package/src/llama.cpp/ggml/src/vulkan-shaders/CMakeLists.txt +5 -0
- package/src/llama.cpp/ggml/src/vulkan-shaders/vulkan-shaders-gen.cpp +552 -0
- package/src/llama.cpp/{llama.h → include/llama.h} +175 -100
- package/src/llama.cpp/models/.editorconfig +1 -0
- package/src/llama.cpp/models/ggml-vocab-aquila.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-baichuan.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-bert-bge.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-bert-bge.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-bert-bge.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-command-r.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-command-r.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-command-r.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-deepseek-coder.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-deepseek-coder.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-deepseek-coder.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-deepseek-llm.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-deepseek-llm.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-deepseek-llm.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-falcon.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-falcon.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-falcon.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-gpt-2.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-gpt-2.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-gpt-2.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-gpt-neox.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-llama-bpe.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-llama-bpe.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-llama-bpe.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-llama-spm.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-llama-spm.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-llama-spm.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-mpt.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-mpt.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-mpt.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-phi-3.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-phi-3.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-phi-3.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-qwen2.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-qwen2.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-qwen2.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-refact.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-refact.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-refact.gguf.out +46 -0
- package/src/llama.cpp/models/ggml-vocab-starcoder.gguf +0 -0
- package/src/llama.cpp/models/ggml-vocab-starcoder.gguf.inp +112 -0
- package/src/llama.cpp/models/ggml-vocab-starcoder.gguf.out +46 -0
- package/src/llama.cpp/pocs/vdot/CMakeLists.txt +2 -2
- package/src/llama.cpp/requirements/requirements-all.txt +12 -0
- package/src/llama.cpp/requirements/requirements-compare-llama-bench.txt +2 -0
- package/src/llama.cpp/requirements/requirements-convert_hf_to_gguf.txt +3 -0
- package/src/llama.cpp/requirements/requirements-convert_hf_to_gguf_update.txt +3 -0
- package/src/llama.cpp/requirements/{requirements-convert.txt → requirements-convert_legacy_llama.txt} +1 -1
- package/src/llama.cpp/requirements/requirements-convert_llama_ggml_to_gguf.txt +1 -0
- package/src/llama.cpp/requirements/requirements-convert_lora_to_gguf.txt +2 -0
- package/src/llama.cpp/requirements/requirements-pydantic.txt +3 -0
- package/src/llama.cpp/requirements/requirements-test-tokenizer-random.txt +1 -0
- package/src/llama.cpp/requirements.txt +5 -4
- package/src/llama.cpp/scripts/build-info.sh +30 -0
- package/src/llama.cpp/scripts/install-oneapi.bat +19 -0
- package/src/llama.cpp/src/CMakeLists.txt +33 -0
- package/src/llama.cpp/src/llama-grammar.cpp +539 -0
- package/src/llama.cpp/src/llama-grammar.h +39 -0
- package/src/llama.cpp/src/llama-impl.h +26 -0
- package/src/llama.cpp/src/llama-sampling.cpp +635 -0
- package/src/llama.cpp/src/llama-sampling.h +56 -0
- package/src/llama.cpp/src/llama-vocab.cpp +1721 -0
- package/src/llama.cpp/src/llama-vocab.h +130 -0
- package/src/llama.cpp/{llama.cpp → src/llama.cpp} +5979 -5260
- package/src/llama.cpp/{unicode-data.cpp → src/unicode-data.cpp} +851 -802
- package/src/llama.cpp/{unicode.cpp → src/unicode.cpp} +52 -30
- package/src/llama.cpp/{unicode.h → src/unicode.h} +5 -1
- package/src/llama.cpp/tests/CMakeLists.txt +19 -20
- package/src/llama.cpp/tests/test-backend-ops.cpp +245 -67
- package/src/llama.cpp/tests/test-chat-template.cpp +57 -3
- package/src/llama.cpp/tests/test-double-float.cpp +2 -2
- package/src/llama.cpp/tests/test-grad0.cpp +2 -2
- package/src/llama.cpp/tests/test-grammar-integration.cpp +978 -31
- package/src/llama.cpp/tests/test-grammar-parser.cpp +423 -158
- package/src/llama.cpp/tests/test-json-schema-to-grammar.cpp +508 -135
- package/src/llama.cpp/tests/test-llama-grammar.cpp +15 -9
- package/src/llama.cpp/tests/test-quantize-fns.cpp +1 -1
- package/src/llama.cpp/tests/test-quantize-perf.cpp +1 -1
- package/src/llama.cpp/tests/test-rope.cpp +3 -4
- package/src/llama.cpp/tests/test-sampling.cpp +5 -5
- package/src/llama.cpp/tests/test-tokenizer-0.cpp +6 -6
- package/src/llama.cpp/tests/test-tokenizer-1-bpe.cpp +20 -15
- package/src/llama.cpp/tests/test-tokenizer-1-spm.cpp +22 -11
- package/bin/darwin/arm64/default.metallib +0 -0
- package/bin/darwin/x64/default.metallib +0 -0
- package/src/llama.cpp/examples/beam-search/CMakeLists.txt +0 -5
- package/src/llama.cpp/examples/beam-search/beam-search.cpp +0 -188
- package/src/llama.cpp/examples/finetune/finetune.cpp +0 -1862
- package/src/llama.cpp/examples/llama.android/llama/CMakeLists.txt +0 -55
- package/src/llama.cpp/examples/train-text-from-scratch/CMakeLists.txt +0 -5
- package/src/llama.cpp/examples/train-text-from-scratch/train-text-from-scratch.cpp +0 -1253
- package/src/llama.cpp/ggml-opencl.cpp +0 -2305
- package/src/llama.cpp/ggml-opencl.h +0 -36
- package/src/llama.cpp/ggml-sycl.cpp +0 -17340
- package/src/llama.cpp/ggml-vulkan-shaders.hpp +0 -81211
- package/src/llama.cpp/requirements/requirements-convert-hf-to-gguf-update.txt +0 -2
- package/src/llama.cpp/requirements/requirements-convert-hf-to-gguf.txt +0 -2
- package/src/llama.cpp/requirements/requirements-convert-llama-ggml-to-gguf.txt +0 -1
- package/src/llama.cpp/scripts/gen-build-info-cpp.cmake +0 -24
- /package/src/llama.cpp/{ggml-alloc.h → ggml/include/ggml-alloc.h} +0 -0
- /package/src/llama.cpp/{ggml-kompute.h → ggml/include/ggml-kompute.h} +0 -0
- /package/src/llama.cpp/{ggml-rpc.h → ggml/include/ggml-rpc.h} +0 -0
- /package/src/llama.cpp/{ggml-vulkan.h → ggml/include/ggml-vulkan.h} +0 -0
- /package/src/llama.cpp/{unicode-data.h → src/unicode-data.h} +0 -0
|
@@ -9,7 +9,7 @@ set( CMAKE_CXX_COMPILER clang++ )
|
|
|
9
9
|
set( CMAKE_C_COMPILER_TARGET ${target} )
|
|
10
10
|
set( CMAKE_CXX_COMPILER_TARGET ${target} )
|
|
11
11
|
|
|
12
|
-
set( arch_c_flags "-march=armv8.7-a -fvectorize -ffp-model=fast" )
|
|
12
|
+
set( arch_c_flags "-march=armv8.7-a -fvectorize -ffp-model=fast -fno-finite-math-only" )
|
|
13
13
|
set( warn_c_flags "-Wno-format -Wno-unused-variable -Wno-unused-function -Wno-gnu-zero-variadic-macro-arguments" )
|
|
14
14
|
|
|
15
15
|
set( CMAKE_C_FLAGS_INIT "${arch_c_flags} ${warn_c_flags}" )
|
|
@@ -0,0 +1,58 @@
|
|
|
1
|
+
set(BUILD_NUMBER 0)
|
|
2
|
+
set(BUILD_COMMIT "unknown")
|
|
3
|
+
set(BUILD_COMPILER "unknown")
|
|
4
|
+
set(BUILD_TARGET "unknown")
|
|
5
|
+
|
|
6
|
+
# Look for git
|
|
7
|
+
find_package(Git)
|
|
8
|
+
if(NOT Git_FOUND)
|
|
9
|
+
find_program(GIT_EXECUTABLE NAMES git git.exe)
|
|
10
|
+
if(GIT_EXECUTABLE)
|
|
11
|
+
set(Git_FOUND TRUE)
|
|
12
|
+
message(STATUS "Found Git: ${GIT_EXECUTABLE}")
|
|
13
|
+
else()
|
|
14
|
+
message(WARNING "Git not found. Build info will not be accurate.")
|
|
15
|
+
endif()
|
|
16
|
+
endif()
|
|
17
|
+
|
|
18
|
+
# Get the commit count and hash
|
|
19
|
+
if(Git_FOUND)
|
|
20
|
+
execute_process(
|
|
21
|
+
COMMAND ${GIT_EXECUTABLE} rev-parse --short HEAD
|
|
22
|
+
WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}
|
|
23
|
+
OUTPUT_VARIABLE HEAD
|
|
24
|
+
OUTPUT_STRIP_TRAILING_WHITESPACE
|
|
25
|
+
RESULT_VARIABLE RES
|
|
26
|
+
)
|
|
27
|
+
if (RES EQUAL 0)
|
|
28
|
+
set(BUILD_COMMIT ${HEAD})
|
|
29
|
+
endif()
|
|
30
|
+
execute_process(
|
|
31
|
+
COMMAND ${GIT_EXECUTABLE} rev-list --count HEAD
|
|
32
|
+
WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}
|
|
33
|
+
OUTPUT_VARIABLE COUNT
|
|
34
|
+
OUTPUT_STRIP_TRAILING_WHITESPACE
|
|
35
|
+
RESULT_VARIABLE RES
|
|
36
|
+
)
|
|
37
|
+
if (RES EQUAL 0)
|
|
38
|
+
set(BUILD_NUMBER ${COUNT})
|
|
39
|
+
endif()
|
|
40
|
+
endif()
|
|
41
|
+
|
|
42
|
+
if(MSVC)
|
|
43
|
+
set(BUILD_COMPILER "${CMAKE_C_COMPILER_ID} ${CMAKE_C_COMPILER_VERSION}")
|
|
44
|
+
set(BUILD_TARGET ${CMAKE_VS_PLATFORM_NAME})
|
|
45
|
+
else()
|
|
46
|
+
execute_process(
|
|
47
|
+
COMMAND sh -c "$@ --version | head -1" _ ${CMAKE_C_COMPILER}
|
|
48
|
+
OUTPUT_VARIABLE OUT
|
|
49
|
+
OUTPUT_STRIP_TRAILING_WHITESPACE
|
|
50
|
+
)
|
|
51
|
+
set(BUILD_COMPILER ${OUT})
|
|
52
|
+
execute_process(
|
|
53
|
+
COMMAND ${CMAKE_C_COMPILER} -dumpmachine
|
|
54
|
+
OUTPUT_VARIABLE OUT
|
|
55
|
+
OUTPUT_STRIP_TRAILING_WHITESPACE
|
|
56
|
+
)
|
|
57
|
+
set(BUILD_TARGET ${OUT})
|
|
58
|
+
endif()
|
|
@@ -0,0 +1,22 @@
|
|
|
1
|
+
find_package(Git)
|
|
2
|
+
|
|
3
|
+
# the commit's SHA1
|
|
4
|
+
execute_process(COMMAND
|
|
5
|
+
"${GIT_EXECUTABLE}" describe --match=NeVeRmAtCh --always --abbrev=8
|
|
6
|
+
WORKING_DIRECTORY "${CMAKE_SOURCE_DIR}"
|
|
7
|
+
OUTPUT_VARIABLE GIT_SHA1
|
|
8
|
+
ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
|
|
9
|
+
|
|
10
|
+
# the date of the commit
|
|
11
|
+
execute_process(COMMAND
|
|
12
|
+
"${GIT_EXECUTABLE}" log -1 --format=%ad --date=local
|
|
13
|
+
WORKING_DIRECTORY "${CMAKE_SOURCE_DIR}"
|
|
14
|
+
OUTPUT_VARIABLE GIT_DATE
|
|
15
|
+
ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
|
|
16
|
+
|
|
17
|
+
# the subject of the commit
|
|
18
|
+
execute_process(COMMAND
|
|
19
|
+
"${GIT_EXECUTABLE}" log -1 --format=%s
|
|
20
|
+
WORKING_DIRECTORY "${CMAKE_SOURCE_DIR}"
|
|
21
|
+
OUTPUT_VARIABLE GIT_COMMIT_SUBJECT
|
|
22
|
+
ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
|
|
@@ -1,5 +1,6 @@
|
|
|
1
1
|
# common
|
|
2
2
|
|
|
3
|
+
find_package(Threads REQUIRED)
|
|
3
4
|
|
|
4
5
|
# Build info header
|
|
5
6
|
#
|
|
@@ -36,7 +37,7 @@ add_custom_command(
|
|
|
36
37
|
COMMENT "Generating build details from Git"
|
|
37
38
|
COMMAND ${CMAKE_COMMAND} -DMSVC=${MSVC} -DCMAKE_C_COMPILER_VERSION=${CMAKE_C_COMPILER_VERSION}
|
|
38
39
|
-DCMAKE_C_COMPILER_ID=${CMAKE_C_COMPILER_ID} -DCMAKE_VS_PLATFORM_NAME=${CMAKE_VS_PLATFORM_NAME}
|
|
39
|
-
-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER} -P "${CMAKE_CURRENT_SOURCE_DIR}
|
|
40
|
+
-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER} -P "${CMAKE_CURRENT_SOURCE_DIR}/cmake/build-info-gen-cpp.cmake"
|
|
40
41
|
WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}/.."
|
|
41
42
|
DEPENDS "${CMAKE_CURRENT_SOURCE_DIR}/build-info.cpp.in" ${GIT_INDEX}
|
|
42
43
|
VERBATIM
|
|
@@ -83,5 +84,5 @@ if (LLAMA_CURL)
|
|
|
83
84
|
endif ()
|
|
84
85
|
|
|
85
86
|
target_include_directories(${TARGET} PUBLIC .)
|
|
86
|
-
target_compile_features(${TARGET} PUBLIC cxx_std_11)
|
|
87
|
-
target_link_libraries(${TARGET} PRIVATE ${LLAMA_COMMON_EXTRA_LIBS} PUBLIC llama)
|
|
87
|
+
target_compile_features (${TARGET} PUBLIC cxx_std_11)
|
|
88
|
+
target_link_libraries (${TARGET} PRIVATE ${LLAMA_COMMON_EXTRA_LIBS} PUBLIC llama Threads::Threads)
|