llama-cpp-pro 0.2.2
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/CHANGELOG.md +295 -0
- package/LICENSE +21 -0
- package/LlamaCpp.podspec +33 -0
- package/LlamaCppCapacitor.podspec +33 -0
- package/Package.swift +29 -0
- package/README.md +93 -0
- package/android/build.gradle +88 -0
- package/android/src/main/AndroidManifest.xml +4 -0
- package/android/src/main/CMakeLists-arm64.txt +138 -0
- package/android/src/main/CMakeLists-x86_64.txt +141 -0
- package/android/src/main/CMakeLists.txt +138 -0
- package/android/src/main/java/ai/annadata/plugin/capacitor/LlamaCpp.java +1340 -0
- package/android/src/main/java/ai/annadata/plugin/capacitor/LlamaCppPlugin.java +814 -0
- package/android/src/main/java/ai/annadata/plugin/capacitor/ModelAdmissionController.java +214 -0
- package/android/src/main/jni-chat-session.cpp +261 -0
- package/android/src/main/jni-lora.cpp +197 -0
- package/android/src/main/jni-multimodal.cpp +167 -0
- package/android/src/main/jni-tts.cpp +290 -0
- package/android/src/main/jni-utils.h +148 -0
- package/android/src/main/jni.cpp +1808 -0
- package/android/src/main/jniLibs/arm64-v8a/libllama-cpp-arm64.so +0 -0
- package/android/src/main/res/.gitkeep +0 -0
- package/build-native.sh +280 -0
- package/cmake/desktop-metal-embed.cmake +30 -0
- package/cmake/desktop-sources.cmake +100 -0
- package/cmake/ggml-backends.cmake +44 -0
- package/cpp/LICENSE +21 -0
- package/cpp/README.md +15 -0
- package/cpp/anyascii.c +22223 -0
- package/cpp/anyascii.h +42 -0
- package/cpp/cap-completion.cpp +942 -0
- package/cpp/cap-completion.h +127 -0
- package/cpp/cap-embedding.cpp +193 -0
- package/cpp/cap-embedding.h +35 -0
- package/cpp/cap-ios-bridge.cpp +1810 -0
- package/cpp/cap-ios-bridge.h +61 -0
- package/cpp/cap-llama.cpp +412 -0
- package/cpp/cap-llama.h +161 -0
- package/cpp/cap-mtmd.hpp +602 -0
- package/cpp/cap-native-server.cpp +1095 -0
- package/cpp/cap-native-server.h +40 -0
- package/cpp/cap-tts.cpp +591 -0
- package/cpp/cap-tts.h +59 -0
- package/cpp/cap-wasm-fs.cpp +156 -0
- package/cpp/cap-wasm-jspi.cpp +19 -0
- package/cpp/cap-wasm-jspi.h +30 -0
- package/cpp/cap-wasm-vfs.cpp +22 -0
- package/cpp/chat-parser.cpp +393 -0
- package/cpp/chat-parser.h +120 -0
- package/cpp/chat.cpp +2315 -0
- package/cpp/chat.h +221 -0
- package/cpp/common.cpp +1664 -0
- package/cpp/common.h +744 -0
- package/cpp/ggml-alloc.c +1028 -0
- package/cpp/ggml-alloc.h +76 -0
- package/cpp/ggml-backend-impl.h +255 -0
- package/cpp/ggml-backend-reg.cpp +600 -0
- package/cpp/ggml-backend.cpp +2121 -0
- package/cpp/ggml-backend.h +354 -0
- package/cpp/ggml-common.h +1878 -0
- package/cpp/ggml-cpp.h +39 -0
- package/cpp/ggml-cpu/amx/amx.cpp +221 -0
- package/cpp/ggml-cpu/amx/amx.h +8 -0
- package/cpp/ggml-cpu/amx/common.h +91 -0
- package/cpp/ggml-cpu/amx/mmq.cpp +2512 -0
- package/cpp/ggml-cpu/amx/mmq.h +10 -0
- package/cpp/ggml-cpu/arch/arm/cpu-feats.cpp +94 -0
- package/cpp/ggml-cpu/arch/arm/quants.c +3650 -0
- package/cpp/ggml-cpu/arch/arm/repack.cpp +1891 -0
- package/cpp/ggml-cpu/arch/x86/cpu-feats.cpp +327 -0
- package/cpp/ggml-cpu/arch/x86/quants.c +3820 -0
- package/cpp/ggml-cpu/arch/x86/repack.cpp +6307 -0
- package/cpp/ggml-cpu/arch-fallback.h +215 -0
- package/cpp/ggml-cpu/binary-ops.cpp +158 -0
- package/cpp/ggml-cpu/binary-ops.h +16 -0
- package/cpp/ggml-cpu/common.h +73 -0
- package/cpp/ggml-cpu/ggml-cpu-impl.h +559 -0
- package/cpp/ggml-cpu/ggml-cpu.c +3578 -0
- package/cpp/ggml-cpu/ggml-cpu.cpp +672 -0
- package/cpp/ggml-cpu/ops.cpp +10587 -0
- package/cpp/ggml-cpu/ops.h +114 -0
- package/cpp/ggml-cpu/quants.c +1193 -0
- package/cpp/ggml-cpu/quants.h +97 -0
- package/cpp/ggml-cpu/repack.cpp +1982 -0
- package/cpp/ggml-cpu/repack.h +120 -0
- package/cpp/ggml-cpu/simd-mappings.h +1184 -0
- package/cpp/ggml-cpu/traits.cpp +36 -0
- package/cpp/ggml-cpu/traits.h +38 -0
- package/cpp/ggml-cpu/unary-ops.cpp +186 -0
- package/cpp/ggml-cpu/unary-ops.h +28 -0
- package/cpp/ggml-cpu/vec.cpp +348 -0
- package/cpp/ggml-cpu/vec.h +1121 -0
- package/cpp/ggml-cpu.h +145 -0
- package/cpp/ggml-impl.h +622 -0
- package/cpp/ggml-metal-impl.h +688 -0
- package/cpp/ggml-metal.h +66 -0
- package/cpp/ggml-metal.m +6833 -0
- package/cpp/ggml-metal.metal +10754 -0
- package/cpp/ggml-opt.cpp +1093 -0
- package/cpp/ggml-opt.h +256 -0
- package/cpp/ggml-quants.c +5324 -0
- package/cpp/ggml-quants.h +106 -0
- package/cpp/ggml-threading.cpp +12 -0
- package/cpp/ggml-threading.h +14 -0
- package/cpp/ggml.c +7108 -0
- package/cpp/ggml.h +2492 -0
- package/cpp/gguf.cpp +1358 -0
- package/cpp/gguf.h +202 -0
- package/cpp/json-partial.cpp +256 -0
- package/cpp/json-partial.h +38 -0
- package/cpp/json-schema-to-grammar.cpp +985 -0
- package/cpp/json-schema-to-grammar.h +21 -0
- package/cpp/llama-adapter.cpp +388 -0
- package/cpp/llama-adapter.h +76 -0
- package/cpp/llama-arch.cpp +2355 -0
- package/cpp/llama-arch.h +499 -0
- package/cpp/llama-batch.cpp +875 -0
- package/cpp/llama-batch.h +160 -0
- package/cpp/llama-chat.cpp +783 -0
- package/cpp/llama-chat.h +65 -0
- package/cpp/llama-context.cpp +2788 -0
- package/cpp/llama-context.h +306 -0
- package/cpp/llama-cparams.cpp +5 -0
- package/cpp/llama-cparams.h +41 -0
- package/cpp/llama-cpp.h +30 -0
- package/cpp/llama-grammar.cpp +1229 -0
- package/cpp/llama-grammar.h +173 -0
- package/cpp/llama-graph.cpp +1891 -0
- package/cpp/llama-graph.h +810 -0
- package/cpp/llama-hparams.cpp +180 -0
- package/cpp/llama-hparams.h +233 -0
- package/cpp/llama-impl.cpp +167 -0
- package/cpp/llama-impl.h +61 -0
- package/cpp/llama-io.cpp +15 -0
- package/cpp/llama-io.h +35 -0
- package/cpp/llama-kv-cache-iswa.cpp +318 -0
- package/cpp/llama-kv-cache-iswa.h +135 -0
- package/cpp/llama-kv-cache.cpp +2059 -0
- package/cpp/llama-kv-cache.h +374 -0
- package/cpp/llama-kv-cells.h +491 -0
- package/cpp/llama-memory-hybrid.cpp +258 -0
- package/cpp/llama-memory-hybrid.h +137 -0
- package/cpp/llama-memory-recurrent.cpp +1146 -0
- package/cpp/llama-memory-recurrent.h +179 -0
- package/cpp/llama-memory.cpp +59 -0
- package/cpp/llama-memory.h +119 -0
- package/cpp/llama-mmap.cpp +609 -0
- package/cpp/llama-mmap.h +68 -0
- package/cpp/llama-model-loader.cpp +1166 -0
- package/cpp/llama-model-loader.h +170 -0
- package/cpp/llama-model-saver.cpp +282 -0
- package/cpp/llama-model-saver.h +37 -0
- package/cpp/llama-model.cpp +19061 -0
- package/cpp/llama-model.h +491 -0
- package/cpp/llama-sampling.cpp +2575 -0
- package/cpp/llama-sampling.h +32 -0
- package/cpp/llama-vocab.cpp +3792 -0
- package/cpp/llama-vocab.h +176 -0
- package/cpp/llama.cpp +358 -0
- package/cpp/llama.h +1373 -0
- package/cpp/log.cpp +428 -0
- package/cpp/log.h +103 -0
- package/cpp/minja/chat-template.hpp +550 -0
- package/cpp/minja/minja.hpp +3009 -0
- package/cpp/nlohmann/json.hpp +25526 -0
- package/cpp/nlohmann/json_fwd.hpp +187 -0
- package/cpp/regex-partial.cpp +204 -0
- package/cpp/regex-partial.h +56 -0
- package/cpp/sampling.cpp +579 -0
- package/cpp/sampling.h +107 -0
- package/cpp/tools/mtmd/clip-impl.h +473 -0
- package/cpp/tools/mtmd/clip.cpp +4322 -0
- package/cpp/tools/mtmd/clip.h +106 -0
- package/cpp/tools/mtmd/miniaudio/miniaudio.h +93468 -0
- package/cpp/tools/mtmd/mtmd-audio.cpp +769 -0
- package/cpp/tools/mtmd/mtmd-audio.h +47 -0
- package/cpp/tools/mtmd/mtmd-helper.cpp +460 -0
- package/cpp/tools/mtmd/mtmd-helper.h +95 -0
- package/cpp/tools/mtmd/mtmd.cpp +1066 -0
- package/cpp/tools/mtmd/mtmd.h +302 -0
- package/cpp/tools/mtmd/stb/stb_image.h +7988 -0
- package/cpp/unicode-data.cpp +7034 -0
- package/cpp/unicode-data.h +20 -0
- package/cpp/unicode.cpp +1061 -0
- package/cpp/unicode.h +68 -0
- package/cpp/vendor/cpp-httplib/httplib.cpp +16509 -0
- package/cpp/vendor/cpp-httplib/httplib.h +3883 -0
- package/desktop/electron-builder.config.cjs +157 -0
- package/desktop/entitlements.mac.plist +12 -0
- package/desktop/package.json +11 -0
- package/desktop/resolve-package-root.cjs +88 -0
- package/desktop/src/main/backend-selector.cjs +148 -0
- package/desktop/src/main/gpu-probe.cjs +142 -0
- package/desktop/src/main/index.cjs +58 -0
- package/desktop/src/main/ipc-handlers.cjs +212 -0
- package/desktop/src/main/model-store.cjs +50 -0
- package/desktop/src/main/preload.cjs +39 -0
- package/desktop/src/main/sidecar-client.cjs +76 -0
- package/desktop/src/main/sidecar-manager.cjs +364 -0
- package/dist/docs.json +14357 -0
- package/dist/esm/definitions.d.ts +731 -0
- package/dist/esm/definitions.js +2 -0
- package/dist/esm/definitions.js.map +1 -0
- package/dist/esm/desktop.d.ts +37 -0
- package/dist/esm/desktop.js +133 -0
- package/dist/esm/desktop.js.map +1 -0
- package/dist/esm/index.d.ts +200 -0
- package/dist/esm/index.js +612 -0
- package/dist/esm/index.js.map +1 -0
- package/dist/esm/isomorphic/desktop.runtime.d.ts +37 -0
- package/dist/esm/isomorphic/desktop.runtime.js +36 -0
- package/dist/esm/isomorphic/desktop.runtime.js.map +1 -0
- package/dist/esm/isomorphic/errors.d.ts +6 -0
- package/dist/esm/isomorphic/errors.js +9 -0
- package/dist/esm/isomorphic/errors.js.map +1 -0
- package/dist/esm/isomorphic/model.admission.d.ts +17 -0
- package/dist/esm/isomorphic/model.admission.js +27 -0
- package/dist/esm/isomorphic/model.admission.js.map +1 -0
- package/dist/esm/isomorphic/model.scheduler.d.ts +31 -0
- package/dist/esm/isomorphic/model.scheduler.js +95 -0
- package/dist/esm/isomorphic/model.scheduler.js.map +1 -0
- package/dist/esm/isomorphic/provider.desktop.d.ts +40 -0
- package/dist/esm/isomorphic/provider.desktop.js +411 -0
- package/dist/esm/isomorphic/provider.desktop.js.map +1 -0
- package/dist/esm/isomorphic/provider.factory.d.ts +2 -0
- package/dist/esm/isomorphic/provider.factory.js +16 -0
- package/dist/esm/isomorphic/provider.factory.js.map +1 -0
- package/dist/esm/isomorphic/provider.interface.d.ts +76 -0
- package/dist/esm/isomorphic/provider.interface.js +2 -0
- package/dist/esm/isomorphic/provider.interface.js.map +1 -0
- package/dist/esm/isomorphic/provider.native.d.ts +18 -0
- package/dist/esm/isomorphic/provider.native.js +173 -0
- package/dist/esm/isomorphic/provider.native.js.map +1 -0
- package/dist/esm/isomorphic/provider.web.d.ts +88 -0
- package/dist/esm/isomorphic/provider.web.js +573 -0
- package/dist/esm/isomorphic/provider.web.js.map +1 -0
- package/dist/esm/isomorphic/sidecar-sse.d.ts +14 -0
- package/dist/esm/isomorphic/sidecar-sse.js +76 -0
- package/dist/esm/isomorphic/sidecar-sse.js.map +1 -0
- package/dist/esm/isomorphic/wasmMemoryCalibration.d.ts +26 -0
- package/dist/esm/isomorphic/wasmMemoryCalibration.js +60 -0
- package/dist/esm/isomorphic/wasmMemoryCalibration.js.map +1 -0
- package/dist/esm/isomorphic/wasmMemoryPolicy.d.ts +55 -0
- package/dist/esm/isomorphic/wasmMemoryPolicy.js +93 -0
- package/dist/esm/isomorphic/wasmMemoryPolicy.js.map +1 -0
- package/dist/esm/storage/manifest.d.ts +13 -0
- package/dist/esm/storage/manifest.js +87 -0
- package/dist/esm/storage/manifest.js.map +1 -0
- package/dist/esm/storage/opfs.store.d.ts +31 -0
- package/dist/esm/storage/opfs.store.js +241 -0
- package/dist/esm/storage/opfs.store.js.map +1 -0
- package/dist/esm/web.d.ts +206 -0
- package/dist/esm/web.js +521 -0
- package/dist/esm/web.js.map +1 -0
- package/dist/esm/workers/async-file.d.ts +13 -0
- package/dist/esm/workers/async-file.js +12 -0
- package/dist/esm/workers/async-file.js.map +1 -0
- package/dist/esm/workers/heapfs.d.ts +55 -0
- package/dist/esm/workers/heapfs.js +115 -0
- package/dist/esm/workers/heapfs.js.map +1 -0
- package/dist/esm/workers/wasm.engine.d.ts +86 -0
- package/dist/esm/workers/wasm.engine.js +504 -0
- package/dist/esm/workers/wasm.engine.js.map +1 -0
- package/dist/esm/workers/worker.protocol.d.ts +160 -0
- package/dist/esm/workers/worker.protocol.js +2 -0
- package/dist/esm/workers/worker.protocol.js.map +1 -0
- package/dist/plugin.cjs +3179 -0
- package/dist/plugin.cjs.map +1 -0
- package/dist/plugin.js +3181 -0
- package/dist/plugin.js.map +1 -0
- package/dist/wasm/llama_engine.d.ts +74 -0
- package/dist/wasm/llama_engine.js +1829 -0
- package/dist/wasm/llama_engine.wasm +0 -0
- package/dist/wasm/llama_engine_emscripten.mjs +2 -0
- package/dist/wasm/package.json +16 -0
- package/dist/workers/llm.worker.js +1226 -0
- package/dist/workers/llm.worker.js.map +7 -0
- package/extraResources/llama-wasm/llama_engine.d.ts +74 -0
- package/extraResources/llama-wasm/llama_engine.js +1829 -0
- package/extraResources/llama-wasm/llama_engine.wasm +0 -0
- package/extraResources/llama-wasm/llama_engine_emscripten.mjs +2 -0
- package/extraResources/llama-wasm/package.json +16 -0
- package/extraResources/sidecar/README.md +11 -0
- package/extraResources/sidecar/darwin-arm64 +0 -0
- package/extraResources/sidecar/darwin-x64 +0 -0
- package/ios/CMakeLists-arm64.txt +161 -0
- package/ios/CMakeLists-x86_64.txt +188 -0
- package/ios/CMakeLists.txt +161 -0
- package/ios/Frameworks/llama-cpp.framework/Info.plist +28 -0
- package/ios/Frameworks/llama-cpp.framework/llama-cpp +0 -0
- package/ios/Sources/LlamaCppCapacitor/LlamaCpp.swift +1365 -0
- package/ios/Sources/LlamaCppCapacitor/LlamaCppPlugin.swift +694 -0
- package/ios/Sources/LlamaCppCapacitor/LlamaNativeBridge.swift +349 -0
- package/ios/Sources/LlamaCppCapacitor/ModelAdmissionController.swift +177 -0
- package/ios/embed-metal-shaders.sh +24 -0
- package/ios/metal-embed.cmake +29 -0
- package/package.json +281 -0
- package/scripts/build-js-preserve-wasm.cjs +53 -0
- package/scripts/build-sidecar-linux.sh +6 -0
- package/scripts/build-sidecar-win.bat +19 -0
- package/scripts/build-sidecar.sh +184 -0
- package/scripts/embed-llama-ios-app-framework.sh +63 -0
- package/scripts/ensure-desktop-sidecar-bundle.cjs +103 -0
- package/scripts/ensure-llama-ios-xcframework.sh +108 -0
- package/scripts/fix-esm-extensions.cjs +45 -0
- package/scripts/prepare-js-dist.cjs +28 -0
- package/scripts/stage-desktop-resources.cjs +116 -0
- package/sidecar/CMakeLists.txt +192 -0
- package/sidecar/cap-sidecar-main.cpp +68 -0
- package/types/llama-cpp-pro.d.ts +441 -0
package/cpp/llama-chat.h
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
1
|
+
#pragma once
|
|
2
|
+
|
|
3
|
+
#include <string>
|
|
4
|
+
#include <vector>
|
|
5
|
+
#include <cstdint>
|
|
6
|
+
|
|
7
|
+
enum llm_chat_template {
|
|
8
|
+
LLM_CHAT_TEMPLATE_CHATML,
|
|
9
|
+
LLM_CHAT_TEMPLATE_LLAMA_2,
|
|
10
|
+
LLM_CHAT_TEMPLATE_LLAMA_2_SYS,
|
|
11
|
+
LLM_CHAT_TEMPLATE_LLAMA_2_SYS_BOS,
|
|
12
|
+
LLM_CHAT_TEMPLATE_LLAMA_2_SYS_STRIP,
|
|
13
|
+
LLM_CHAT_TEMPLATE_MISTRAL_V1,
|
|
14
|
+
LLM_CHAT_TEMPLATE_MISTRAL_V3,
|
|
15
|
+
LLM_CHAT_TEMPLATE_MISTRAL_V3_TEKKEN,
|
|
16
|
+
LLM_CHAT_TEMPLATE_MISTRAL_V7,
|
|
17
|
+
LLM_CHAT_TEMPLATE_MISTRAL_V7_TEKKEN,
|
|
18
|
+
LLM_CHAT_TEMPLATE_PHI_3,
|
|
19
|
+
LLM_CHAT_TEMPLATE_PHI_4,
|
|
20
|
+
LLM_CHAT_TEMPLATE_FALCON_3,
|
|
21
|
+
LLM_CHAT_TEMPLATE_ZEPHYR,
|
|
22
|
+
LLM_CHAT_TEMPLATE_MONARCH,
|
|
23
|
+
LLM_CHAT_TEMPLATE_GEMMA,
|
|
24
|
+
LLM_CHAT_TEMPLATE_ORION,
|
|
25
|
+
LLM_CHAT_TEMPLATE_OPENCHAT,
|
|
26
|
+
LLM_CHAT_TEMPLATE_VICUNA,
|
|
27
|
+
LLM_CHAT_TEMPLATE_VICUNA_ORCA,
|
|
28
|
+
LLM_CHAT_TEMPLATE_DEEPSEEK,
|
|
29
|
+
LLM_CHAT_TEMPLATE_DEEPSEEK_2,
|
|
30
|
+
LLM_CHAT_TEMPLATE_DEEPSEEK_3,
|
|
31
|
+
LLM_CHAT_TEMPLATE_COMMAND_R,
|
|
32
|
+
LLM_CHAT_TEMPLATE_LLAMA_3,
|
|
33
|
+
LLM_CHAT_TEMPLATE_CHATGLM_3,
|
|
34
|
+
LLM_CHAT_TEMPLATE_CHATGLM_4,
|
|
35
|
+
LLM_CHAT_TEMPLATE_GLMEDGE,
|
|
36
|
+
LLM_CHAT_TEMPLATE_MINICPM,
|
|
37
|
+
LLM_CHAT_TEMPLATE_EXAONE_3,
|
|
38
|
+
LLM_CHAT_TEMPLATE_EXAONE_4,
|
|
39
|
+
LLM_CHAT_TEMPLATE_RWKV_WORLD,
|
|
40
|
+
LLM_CHAT_TEMPLATE_GRANITE,
|
|
41
|
+
LLM_CHAT_TEMPLATE_GIGACHAT,
|
|
42
|
+
LLM_CHAT_TEMPLATE_MEGREZ,
|
|
43
|
+
LLM_CHAT_TEMPLATE_YANDEX,
|
|
44
|
+
LLM_CHAT_TEMPLATE_BAILING,
|
|
45
|
+
LLM_CHAT_TEMPLATE_LLAMA4,
|
|
46
|
+
LLM_CHAT_TEMPLATE_SMOLVLM,
|
|
47
|
+
LLM_CHAT_TEMPLATE_DOTS1,
|
|
48
|
+
LLM_CHAT_TEMPLATE_HUNYUAN_MOE,
|
|
49
|
+
LLM_CHAT_TEMPLATE_OPENAI_MOE,
|
|
50
|
+
LLM_CHAT_TEMPLATE_HUNYUAN_DENSE,
|
|
51
|
+
LLM_CHAT_TEMPLATE_KIMI_K2,
|
|
52
|
+
LLM_CHAT_TEMPLATE_SEED_OSS,
|
|
53
|
+
LLM_CHAT_TEMPLATE_UNKNOWN,
|
|
54
|
+
};
|
|
55
|
+
|
|
56
|
+
struct llama_chat_message;
|
|
57
|
+
|
|
58
|
+
llm_chat_template llm_chat_template_from_str(const std::string & name);
|
|
59
|
+
|
|
60
|
+
llm_chat_template llm_chat_detect_template(const std::string & tmpl);
|
|
61
|
+
|
|
62
|
+
int32_t llm_chat_apply_template(
|
|
63
|
+
llm_chat_template tmpl,
|
|
64
|
+
const std::vector<const llama_chat_message *> & chat,
|
|
65
|
+
std::string & dest, bool add_ass);
|