npm - llama-cpp-capacitor - Versions diffs - 0.0.6 → 0.0.7 - Mend

llama-cpp-capacitor 0.0.6 → 0.0.7

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 (149) hide show

package/cpp/LICENSE +21 -0
package/cpp/README.md +4 -0
package/cpp/anyascii.c +22223 -0
package/cpp/anyascii.h +42 -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 +1619 -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 +2118 -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 +525 -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-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 +2748 -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 +600 -0
package/cpp/llama-mmap.h +68 -0
package/cpp/llama-model-loader.cpp +1164 -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 +19042 -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 +427 -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/rn-completion.cpp +681 -0
package/cpp/rn-completion.h +116 -0
package/cpp/rn-llama.cpp +345 -0
package/cpp/rn-llama.h +149 -0
package/cpp/rn-mtmd.hpp +602 -0
package/cpp/rn-tts.cpp +591 -0
package/cpp/rn-tts.h +59 -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 +91 -0
package/cpp/tools/mtmd/mtmd.cpp +1066 -0
package/cpp/tools/mtmd/mtmd.h +298 -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/package.json +2 -1

package/cpp/llama-chat.h ADDED Viewed

@@ -0,0 +1,65 @@
+#pragma once
+#include <string>
+#include <vector>
+#include <cstdint>
+enum llm_chat_template {
+    LLM_CHAT_TEMPLATE_CHATML,
+    LLM_CHAT_TEMPLATE_LLAMA_2,
+    LLM_CHAT_TEMPLATE_LLAMA_2_SYS,
+    LLM_CHAT_TEMPLATE_LLAMA_2_SYS_BOS,
+    LLM_CHAT_TEMPLATE_LLAMA_2_SYS_STRIP,
+    LLM_CHAT_TEMPLATE_MISTRAL_V1,
+    LLM_CHAT_TEMPLATE_MISTRAL_V3,
+    LLM_CHAT_TEMPLATE_MISTRAL_V3_TEKKEN,
+    LLM_CHAT_TEMPLATE_MISTRAL_V7,
+    LLM_CHAT_TEMPLATE_MISTRAL_V7_TEKKEN,
+    LLM_CHAT_TEMPLATE_PHI_3,
+    LLM_CHAT_TEMPLATE_PHI_4,
+    LLM_CHAT_TEMPLATE_FALCON_3,
+    LLM_CHAT_TEMPLATE_ZEPHYR,
+    LLM_CHAT_TEMPLATE_MONARCH,
+    LLM_CHAT_TEMPLATE_GEMMA,
+    LLM_CHAT_TEMPLATE_ORION,
+    LLM_CHAT_TEMPLATE_OPENCHAT,
+    LLM_CHAT_TEMPLATE_VICUNA,
+    LLM_CHAT_TEMPLATE_VICUNA_ORCA,
+    LLM_CHAT_TEMPLATE_DEEPSEEK,
+    LLM_CHAT_TEMPLATE_DEEPSEEK_2,
+    LLM_CHAT_TEMPLATE_DEEPSEEK_3,
+    LLM_CHAT_TEMPLATE_COMMAND_R,
+    LLM_CHAT_TEMPLATE_LLAMA_3,
+    LLM_CHAT_TEMPLATE_CHATGLM_3,
+    LLM_CHAT_TEMPLATE_CHATGLM_4,
+    LLM_CHAT_TEMPLATE_GLMEDGE,
+    LLM_CHAT_TEMPLATE_MINICPM,
+    LLM_CHAT_TEMPLATE_EXAONE_3,
+    LLM_CHAT_TEMPLATE_EXAONE_4,
+    LLM_CHAT_TEMPLATE_RWKV_WORLD,
+    LLM_CHAT_TEMPLATE_GRANITE,
+    LLM_CHAT_TEMPLATE_GIGACHAT,
+    LLM_CHAT_TEMPLATE_MEGREZ,
+    LLM_CHAT_TEMPLATE_YANDEX,
+    LLM_CHAT_TEMPLATE_BAILING,
+    LLM_CHAT_TEMPLATE_LLAMA4,
+    LLM_CHAT_TEMPLATE_SMOLVLM,
+    LLM_CHAT_TEMPLATE_DOTS1,
+    LLM_CHAT_TEMPLATE_HUNYUAN_MOE,
+    LLM_CHAT_TEMPLATE_OPENAI_MOE,
+    LLM_CHAT_TEMPLATE_HUNYUAN_DENSE,
+    LLM_CHAT_TEMPLATE_KIMI_K2,
+    LLM_CHAT_TEMPLATE_SEED_OSS,
+    LLM_CHAT_TEMPLATE_UNKNOWN,
+};
+struct llama_chat_message;
+llm_chat_template llm_chat_template_from_str(const std::string & name);
+llm_chat_template llm_chat_detect_template(const std::string & tmpl);
+int32_t llm_chat_apply_template(
+    llm_chat_template tmpl,
+    const std::vector<const llama_chat_message *> & chat,
+    std::string & dest, bool add_ass);