@fugood/llama.node 0.2.3 → 0.3.1

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 (319) hide show
  1. package/CMakeLists.txt +6 -3
  2. package/bin/darwin/arm64/llama-node.node +0 -0
  3. package/bin/darwin/x64/llama-node.node +0 -0
  4. package/bin/linux/arm64/llama-node.node +0 -0
  5. package/bin/linux/x64/llama-node.node +0 -0
  6. package/bin/linux-vulkan/arm64/llama-node.node +0 -0
  7. package/bin/linux-vulkan/x64/llama-node.node +0 -0
  8. package/bin/win32/arm64/llama-node.node +0 -0
  9. package/bin/win32/arm64/node.lib +0 -0
  10. package/bin/win32/x64/llama-node.node +0 -0
  11. package/bin/win32/x64/node.lib +0 -0
  12. package/bin/win32-vulkan/arm64/llama-node.node +0 -0
  13. package/bin/win32-vulkan/arm64/node.lib +0 -0
  14. package/bin/win32-vulkan/x64/llama-node.node +0 -0
  15. package/bin/win32-vulkan/x64/node.lib +0 -0
  16. package/lib/binding.ts +8 -1
  17. package/package.json +3 -3
  18. package/patches/llama.patch +12 -12
  19. package/src/DetokenizeWorker.cpp +1 -1
  20. package/src/LlamaContext.cpp +33 -1
  21. package/src/LlamaContext.h +1 -0
  22. package/src/llama.cpp/.github/workflows/bench.yml +310 -0
  23. package/src/llama.cpp/.github/workflows/build.yml +1315 -0
  24. package/src/llama.cpp/.github/workflows/close-issue.yml +23 -0
  25. package/src/llama.cpp/.github/workflows/docker.yml +116 -0
  26. package/src/llama.cpp/.github/workflows/editorconfig.yml +27 -0
  27. package/src/llama.cpp/.github/workflows/gguf-publish.yml +44 -0
  28. package/src/llama.cpp/.github/workflows/labeler.yml +17 -0
  29. package/src/llama.cpp/.github/workflows/nix-ci-aarch64.yml +65 -0
  30. package/src/llama.cpp/.github/workflows/nix-ci.yml +72 -0
  31. package/src/llama.cpp/.github/workflows/nix-flake-update.yml +22 -0
  32. package/src/llama.cpp/.github/workflows/nix-publish-flake.yml +36 -0
  33. package/src/llama.cpp/.github/workflows/python-check-requirements.yml +35 -0
  34. package/src/llama.cpp/.github/workflows/python-lint.yml +23 -0
  35. package/src/llama.cpp/.github/workflows/python-type-check.yml +38 -0
  36. package/src/llama.cpp/.github/workflows/server.yml +183 -0
  37. package/src/llama.cpp/CMakeLists.txt +91 -1245
  38. package/src/llama.cpp/cmake/arm64-windows-llvm.cmake +1 -1
  39. package/src/llama.cpp/cmake/build-info.cmake +58 -0
  40. package/src/llama.cpp/cmake/git-vars.cmake +22 -0
  41. package/src/llama.cpp/common/CMakeLists.txt +4 -3
  42. package/src/llama.cpp/common/build-info.cpp.in +4 -0
  43. package/src/llama.cpp/common/common.cpp +1116 -877
  44. package/src/llama.cpp/common/common.h +191 -77
  45. package/src/llama.cpp/common/grammar-parser.cpp +118 -31
  46. package/src/llama.cpp/common/json-schema-to-grammar.cpp +346 -65
  47. package/src/llama.cpp/common/log.h +1 -1
  48. package/src/llama.cpp/common/ngram-cache.h +10 -3
  49. package/src/llama.cpp/common/sampling.cpp +19 -10
  50. package/src/llama.cpp/docs/build.md +353 -0
  51. package/src/llama.cpp/examples/CMakeLists.txt +22 -22
  52. package/src/llama.cpp/examples/baby-llama/CMakeLists.txt +1 -1
  53. package/src/llama.cpp/examples/baby-llama/baby-llama.cpp +6 -6
  54. package/src/llama.cpp/examples/batched/CMakeLists.txt +1 -1
  55. package/src/llama.cpp/examples/batched/batched.cpp +52 -55
  56. package/src/llama.cpp/examples/batched-bench/CMakeLists.txt +1 -1
  57. package/src/llama.cpp/examples/batched-bench/batched-bench.cpp +20 -72
  58. package/src/llama.cpp/examples/benchmark/CMakeLists.txt +1 -1
  59. package/src/llama.cpp/examples/chat-13B.bat +57 -0
  60. package/src/llama.cpp/examples/convert-llama2c-to-ggml/CMakeLists.txt +1 -1
  61. package/src/llama.cpp/examples/{finetune → cvector-generator}/CMakeLists.txt +2 -2
  62. package/src/llama.cpp/examples/cvector-generator/completions.txt +582 -0
  63. package/src/llama.cpp/examples/cvector-generator/cvector-generator.cpp +503 -0
  64. package/src/llama.cpp/examples/cvector-generator/mean.hpp +48 -0
  65. package/src/llama.cpp/examples/cvector-generator/negative.txt +4 -0
  66. package/src/llama.cpp/examples/cvector-generator/pca.hpp +325 -0
  67. package/src/llama.cpp/examples/cvector-generator/positive.txt +4 -0
  68. package/src/llama.cpp/examples/deprecation-warning/deprecation-warning.cpp +35 -0
  69. package/src/llama.cpp/examples/embedding/CMakeLists.txt +1 -1
  70. package/src/llama.cpp/examples/embedding/embedding.cpp +94 -46
  71. package/src/llama.cpp/examples/eval-callback/CMakeLists.txt +2 -2
  72. package/src/llama.cpp/examples/eval-callback/eval-callback.cpp +4 -6
  73. package/src/llama.cpp/examples/export-lora/CMakeLists.txt +1 -1
  74. package/src/llama.cpp/examples/export-lora/export-lora.cpp +344 -386
  75. package/src/llama.cpp/examples/gbnf-validator/CMakeLists.txt +2 -2
  76. package/src/llama.cpp/examples/gbnf-validator/gbnf-validator.cpp +30 -25
  77. package/src/llama.cpp/examples/gguf/CMakeLists.txt +1 -1
  78. package/src/llama.cpp/examples/gguf/gguf.cpp +5 -0
  79. package/src/llama.cpp/examples/gguf-hash/CMakeLists.txt +15 -0
  80. package/src/llama.cpp/examples/gguf-hash/deps/rotate-bits/rotate-bits.h +46 -0
  81. package/src/llama.cpp/examples/gguf-hash/deps/sha1/sha1.c +295 -0
  82. package/src/llama.cpp/examples/gguf-hash/deps/sha1/sha1.h +52 -0
  83. package/src/llama.cpp/examples/gguf-hash/deps/sha256/sha256.c +221 -0
  84. package/src/llama.cpp/examples/gguf-hash/deps/sha256/sha256.h +24 -0
  85. package/src/llama.cpp/examples/gguf-hash/deps/xxhash/xxhash.c +42 -0
  86. package/src/llama.cpp/examples/gguf-hash/deps/xxhash/xxhash.h +7093 -0
  87. package/src/llama.cpp/examples/gguf-hash/gguf-hash.cpp +693 -0
  88. package/src/llama.cpp/examples/gguf-split/CMakeLists.txt +1 -1
  89. package/src/llama.cpp/examples/gguf-split/gguf-split.cpp +3 -3
  90. package/src/llama.cpp/examples/gritlm/CMakeLists.txt +1 -1
  91. package/src/llama.cpp/examples/gritlm/gritlm.cpp +6 -2
  92. package/src/llama.cpp/examples/imatrix/CMakeLists.txt +1 -1
  93. package/src/llama.cpp/examples/imatrix/imatrix.cpp +137 -176
  94. package/src/llama.cpp/examples/infill/CMakeLists.txt +1 -1
  95. package/src/llama.cpp/examples/infill/infill.cpp +38 -153
  96. package/src/llama.cpp/examples/llama-bench/llama-bench.cpp +175 -94
  97. package/src/llama.cpp/examples/llama.android/app/build.gradle.kts +65 -0
  98. package/src/llama.cpp/examples/llama.android/build.gradle.kts +6 -0
  99. package/src/llama.cpp/examples/llama.android/llama/build.gradle.kts +68 -0
  100. package/src/llama.cpp/examples/llama.android/llama/src/main/cpp/CMakeLists.txt +11 -7
  101. package/src/llama.cpp/examples/llama.android/llama/src/main/cpp/llama-android.cpp +2 -2
  102. package/src/llama.cpp/examples/llama.android/settings.gradle.kts +18 -0
  103. package/src/llama.cpp/examples/llava/CMakeLists.txt +6 -5
  104. package/src/llama.cpp/examples/llava/android/build_64.sh +8 -0
  105. package/src/llama.cpp/examples/llava/clip.cpp +23 -14
  106. package/src/llama.cpp/examples/llava/llava-cli.cpp +8 -6
  107. package/src/llama.cpp/examples/llava/requirements.txt +3 -2
  108. package/src/llama.cpp/examples/lookahead/CMakeLists.txt +1 -1
  109. package/src/llama.cpp/examples/lookahead/lookahead.cpp +2 -1
  110. package/src/llama.cpp/examples/lookup/CMakeLists.txt +4 -4
  111. package/src/llama.cpp/examples/lookup/lookup-create.cpp +2 -0
  112. package/src/llama.cpp/examples/lookup/lookup-merge.cpp +4 -4
  113. package/src/llama.cpp/examples/lookup/lookup-stats.cpp +2 -2
  114. package/src/llama.cpp/examples/lookup/lookup.cpp +1 -1
  115. package/src/llama.cpp/examples/main/CMakeLists.txt +1 -1
  116. package/src/llama.cpp/examples/main/main.cpp +98 -75
  117. package/src/llama.cpp/examples/main-cmake-pkg/CMakeLists.txt +4 -5
  118. package/src/llama.cpp/examples/parallel/CMakeLists.txt +1 -1
  119. package/src/llama.cpp/examples/parallel/parallel.cpp +2 -1
  120. package/src/llama.cpp/examples/passkey/CMakeLists.txt +1 -1
  121. package/src/llama.cpp/examples/passkey/passkey.cpp +23 -43
  122. package/src/llama.cpp/examples/perplexity/CMakeLists.txt +1 -1
  123. package/src/llama.cpp/examples/perplexity/perplexity.cpp +13 -10
  124. package/src/llama.cpp/examples/quantize/CMakeLists.txt +1 -1
  125. package/src/llama.cpp/examples/quantize/quantize.cpp +37 -34
  126. package/src/llama.cpp/examples/quantize-stats/CMakeLists.txt +1 -1
  127. package/src/llama.cpp/examples/quantize-stats/quantize-stats.cpp +1 -1
  128. package/src/llama.cpp/examples/retrieval/CMakeLists.txt +1 -1
  129. package/src/llama.cpp/examples/retrieval/retrieval.cpp +26 -77
  130. package/src/llama.cpp/examples/save-load-state/CMakeLists.txt +1 -1
  131. package/src/llama.cpp/examples/save-load-state/save-load-state.cpp +14 -7
  132. package/src/llama.cpp/examples/server/CMakeLists.txt +26 -2
  133. package/src/llama.cpp/examples/server/server.cpp +274 -671
  134. package/src/llama.cpp/examples/server/tests/requirements.txt +2 -2
  135. package/src/llama.cpp/examples/server/utils.hpp +28 -29
  136. package/src/llama.cpp/examples/simple/CMakeLists.txt +1 -1
  137. package/src/llama.cpp/examples/simple/simple.cpp +21 -29
  138. package/src/llama.cpp/examples/speculative/CMakeLists.txt +1 -1
  139. package/src/llama.cpp/examples/speculative/speculative.cpp +2 -1
  140. package/src/llama.cpp/examples/sycl/CMakeLists.txt +1 -1
  141. package/src/llama.cpp/examples/sycl/build.sh +23 -0
  142. package/src/llama.cpp/examples/sycl/run-llama2.sh +36 -0
  143. package/src/llama.cpp/examples/sycl/win-build-sycl.bat +33 -0
  144. package/src/llama.cpp/examples/sycl/win-run-llama2.bat +9 -0
  145. package/src/llama.cpp/examples/tokenize/CMakeLists.txt +1 -1
  146. package/src/llama.cpp/examples/tokenize/tokenize.cpp +16 -2
  147. package/src/llama.cpp/ggml/CMakeLists.txt +253 -0
  148. package/src/llama.cpp/{cmake → ggml/cmake}/FindSIMD.cmake +6 -6
  149. package/src/llama.cpp/{ggml-backend.h → ggml/include/ggml-backend.h} +22 -17
  150. package/src/llama.cpp/ggml/include/ggml-blas.h +23 -0
  151. package/src/llama.cpp/ggml/include/ggml-cann.h +125 -0
  152. package/src/llama.cpp/{ggml-cuda.h → ggml/include/ggml-cuda.h} +3 -0
  153. package/src/llama.cpp/{ggml-metal.h → ggml/include/ggml-metal.h} +1 -2
  154. package/src/llama.cpp/{ggml-sycl.h → ggml/include/ggml-sycl.h} +3 -10
  155. package/src/llama.cpp/{ggml.h → ggml/include/ggml.h} +80 -85
  156. package/src/llama.cpp/ggml/src/CMakeLists.txt +1329 -0
  157. package/src/llama.cpp/ggml/src/ggml-aarch64.c +2193 -0
  158. package/src/llama.cpp/ggml/src/ggml-aarch64.h +39 -0
  159. package/src/llama.cpp/{ggml-alloc.c → ggml/src/ggml-alloc.c} +100 -49
  160. package/src/llama.cpp/{ggml-backend-impl.h → ggml/src/ggml-backend-impl.h} +20 -8
  161. package/src/llama.cpp/{ggml-backend.c → ggml/src/ggml-backend.c} +307 -167
  162. package/src/llama.cpp/ggml/src/ggml-blas.cpp +367 -0
  163. package/src/llama.cpp/ggml/src/ggml-cann/acl_tensor.cpp +198 -0
  164. package/src/llama.cpp/ggml/src/ggml-cann/acl_tensor.h +230 -0
  165. package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.cpp +2944 -0
  166. package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.h +592 -0
  167. package/src/llama.cpp/ggml/src/ggml-cann/common.h +282 -0
  168. package/src/llama.cpp/ggml/src/ggml-cann/kernels/CMakeLists.txt +32 -0
  169. package/src/llama.cpp/ggml/src/ggml-cann/kernels/ascendc_kernels.h +17 -0
  170. package/src/llama.cpp/ggml/src/ggml-cann/kernels/dup.cpp +223 -0
  171. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f16.cpp +186 -0
  172. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f32.cpp +180 -0
  173. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q4_0.cpp +193 -0
  174. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q8_0.cpp +191 -0
  175. package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp +208 -0
  176. package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp +206 -0
  177. package/src/llama.cpp/ggml/src/ggml-cann.cpp +2023 -0
  178. package/src/llama.cpp/{ggml-common.h → ggml/src/ggml-common.h} +41 -7
  179. package/src/llama.cpp/{ggml-impl.h → ggml/src/ggml-impl.h} +113 -9
  180. package/src/llama.cpp/{ggml-kompute.cpp → ggml/src/ggml-kompute.cpp} +33 -18
  181. package/src/llama.cpp/{ggml-quants.c → ggml/src/ggml-quants.c} +1460 -940
  182. package/src/llama.cpp/{ggml-quants.h → ggml/src/ggml-quants.h} +19 -20
  183. package/src/llama.cpp/{ggml-rpc.cpp → ggml/src/ggml-rpc.cpp} +95 -72
  184. package/src/llama.cpp/ggml/src/ggml-sycl/backend.hpp +27 -0
  185. package/src/llama.cpp/ggml/src/ggml-sycl/common.cpp +53 -0
  186. package/src/llama.cpp/ggml/src/ggml-sycl/common.hpp +355 -0
  187. package/src/llama.cpp/ggml/src/ggml-sycl/concat.cpp +195 -0
  188. package/src/llama.cpp/ggml/src/ggml-sycl/concat.hpp +21 -0
  189. package/src/llama.cpp/ggml/src/ggml-sycl/convert.cpp +547 -0
  190. package/src/llama.cpp/ggml/src/ggml-sycl/convert.hpp +27 -0
  191. package/src/llama.cpp/ggml/src/ggml-sycl/dequantize.hpp +698 -0
  192. package/src/llama.cpp/ggml/src/ggml-sycl/dmmv.cpp +1023 -0
  193. package/src/llama.cpp/ggml/src/ggml-sycl/dmmv.hpp +27 -0
  194. package/src/llama.cpp/ggml/src/ggml-sycl/dpct/helper.hpp +3011 -0
  195. package/src/llama.cpp/ggml/src/ggml-sycl/mmq.cpp +3031 -0
  196. package/src/llama.cpp/ggml/src/ggml-sycl/mmq.hpp +33 -0
  197. package/src/llama.cpp/ggml/src/ggml-sycl/mmvq.cpp +1027 -0
  198. package/src/llama.cpp/ggml/src/ggml-sycl/mmvq.hpp +27 -0
  199. package/src/llama.cpp/ggml/src/ggml-sycl/norm.cpp +374 -0
  200. package/src/llama.cpp/ggml/src/ggml-sycl/norm.hpp +35 -0
  201. package/src/llama.cpp/ggml/src/ggml-sycl/presets.hpp +66 -0
  202. package/src/llama.cpp/ggml/src/ggml-sycl/rope.cpp +275 -0
  203. package/src/llama.cpp/ggml/src/ggml-sycl/rope.hpp +22 -0
  204. package/src/llama.cpp/ggml/src/ggml-sycl/softmax.cpp +251 -0
  205. package/src/llama.cpp/ggml/src/ggml-sycl/softmax.hpp +24 -0
  206. package/src/llama.cpp/ggml/src/ggml-sycl/vecdotq.hpp +1140 -0
  207. package/src/llama.cpp/ggml/src/ggml-sycl.cpp +5314 -0
  208. package/src/llama.cpp/{ggml-vulkan.cpp → ggml/src/ggml-vulkan.cpp} +1781 -1868
  209. package/src/llama.cpp/{ggml.c → ggml/src/ggml.c} +1245 -2087
  210. package/src/llama.cpp/{sgemm.cpp → ggml/src/llamafile/sgemm.cpp} +21 -24
  211. package/src/llama.cpp/{sgemm.h → ggml/src/llamafile/sgemm.h} +1 -1
  212. package/src/llama.cpp/ggml/src/vulkan-shaders/CMakeLists.txt +5 -0
  213. package/src/llama.cpp/ggml/src/vulkan-shaders/vulkan-shaders-gen.cpp +552 -0
  214. package/src/llama.cpp/{llama.h → include/llama.h} +175 -100
  215. package/src/llama.cpp/models/.editorconfig +1 -0
  216. package/src/llama.cpp/models/ggml-vocab-aquila.gguf +0 -0
  217. package/src/llama.cpp/models/ggml-vocab-baichuan.gguf +0 -0
  218. package/src/llama.cpp/models/ggml-vocab-bert-bge.gguf +0 -0
  219. package/src/llama.cpp/models/ggml-vocab-bert-bge.gguf.inp +112 -0
  220. package/src/llama.cpp/models/ggml-vocab-bert-bge.gguf.out +46 -0
  221. package/src/llama.cpp/models/ggml-vocab-command-r.gguf +0 -0
  222. package/src/llama.cpp/models/ggml-vocab-command-r.gguf.inp +112 -0
  223. package/src/llama.cpp/models/ggml-vocab-command-r.gguf.out +46 -0
  224. package/src/llama.cpp/models/ggml-vocab-deepseek-coder.gguf +0 -0
  225. package/src/llama.cpp/models/ggml-vocab-deepseek-coder.gguf.inp +112 -0
  226. package/src/llama.cpp/models/ggml-vocab-deepseek-coder.gguf.out +46 -0
  227. package/src/llama.cpp/models/ggml-vocab-deepseek-llm.gguf +0 -0
  228. package/src/llama.cpp/models/ggml-vocab-deepseek-llm.gguf.inp +112 -0
  229. package/src/llama.cpp/models/ggml-vocab-deepseek-llm.gguf.out +46 -0
  230. package/src/llama.cpp/models/ggml-vocab-falcon.gguf +0 -0
  231. package/src/llama.cpp/models/ggml-vocab-falcon.gguf.inp +112 -0
  232. package/src/llama.cpp/models/ggml-vocab-falcon.gguf.out +46 -0
  233. package/src/llama.cpp/models/ggml-vocab-gpt-2.gguf +0 -0
  234. package/src/llama.cpp/models/ggml-vocab-gpt-2.gguf.inp +112 -0
  235. package/src/llama.cpp/models/ggml-vocab-gpt-2.gguf.out +46 -0
  236. package/src/llama.cpp/models/ggml-vocab-gpt-neox.gguf +0 -0
  237. package/src/llama.cpp/models/ggml-vocab-llama-bpe.gguf +0 -0
  238. package/src/llama.cpp/models/ggml-vocab-llama-bpe.gguf.inp +112 -0
  239. package/src/llama.cpp/models/ggml-vocab-llama-bpe.gguf.out +46 -0
  240. package/src/llama.cpp/models/ggml-vocab-llama-spm.gguf +0 -0
  241. package/src/llama.cpp/models/ggml-vocab-llama-spm.gguf.inp +112 -0
  242. package/src/llama.cpp/models/ggml-vocab-llama-spm.gguf.out +46 -0
  243. package/src/llama.cpp/models/ggml-vocab-mpt.gguf +0 -0
  244. package/src/llama.cpp/models/ggml-vocab-mpt.gguf.inp +112 -0
  245. package/src/llama.cpp/models/ggml-vocab-mpt.gguf.out +46 -0
  246. package/src/llama.cpp/models/ggml-vocab-phi-3.gguf +0 -0
  247. package/src/llama.cpp/models/ggml-vocab-phi-3.gguf.inp +112 -0
  248. package/src/llama.cpp/models/ggml-vocab-phi-3.gguf.out +46 -0
  249. package/src/llama.cpp/models/ggml-vocab-qwen2.gguf +0 -0
  250. package/src/llama.cpp/models/ggml-vocab-qwen2.gguf.inp +112 -0
  251. package/src/llama.cpp/models/ggml-vocab-qwen2.gguf.out +46 -0
  252. package/src/llama.cpp/models/ggml-vocab-refact.gguf +0 -0
  253. package/src/llama.cpp/models/ggml-vocab-refact.gguf.inp +112 -0
  254. package/src/llama.cpp/models/ggml-vocab-refact.gguf.out +46 -0
  255. package/src/llama.cpp/models/ggml-vocab-starcoder.gguf +0 -0
  256. package/src/llama.cpp/models/ggml-vocab-starcoder.gguf.inp +112 -0
  257. package/src/llama.cpp/models/ggml-vocab-starcoder.gguf.out +46 -0
  258. package/src/llama.cpp/pocs/vdot/CMakeLists.txt +2 -2
  259. package/src/llama.cpp/requirements/requirements-all.txt +12 -0
  260. package/src/llama.cpp/requirements/requirements-compare-llama-bench.txt +2 -0
  261. package/src/llama.cpp/requirements/requirements-convert_hf_to_gguf.txt +3 -0
  262. package/src/llama.cpp/requirements/requirements-convert_hf_to_gguf_update.txt +3 -0
  263. package/src/llama.cpp/requirements/{requirements-convert.txt → requirements-convert_legacy_llama.txt} +1 -1
  264. package/src/llama.cpp/requirements/requirements-convert_llama_ggml_to_gguf.txt +1 -0
  265. package/src/llama.cpp/requirements/requirements-convert_lora_to_gguf.txt +2 -0
  266. package/src/llama.cpp/requirements/requirements-pydantic.txt +3 -0
  267. package/src/llama.cpp/requirements/requirements-test-tokenizer-random.txt +1 -0
  268. package/src/llama.cpp/requirements.txt +5 -4
  269. package/src/llama.cpp/scripts/build-info.sh +30 -0
  270. package/src/llama.cpp/scripts/install-oneapi.bat +19 -0
  271. package/src/llama.cpp/src/CMakeLists.txt +33 -0
  272. package/src/llama.cpp/src/llama-grammar.cpp +539 -0
  273. package/src/llama.cpp/src/llama-grammar.h +39 -0
  274. package/src/llama.cpp/src/llama-impl.h +26 -0
  275. package/src/llama.cpp/src/llama-sampling.cpp +635 -0
  276. package/src/llama.cpp/src/llama-sampling.h +56 -0
  277. package/src/llama.cpp/src/llama-vocab.cpp +1721 -0
  278. package/src/llama.cpp/src/llama-vocab.h +130 -0
  279. package/src/llama.cpp/{llama.cpp → src/llama.cpp} +5979 -5260
  280. package/src/llama.cpp/{unicode-data.cpp → src/unicode-data.cpp} +851 -802
  281. package/src/llama.cpp/{unicode.cpp → src/unicode.cpp} +52 -30
  282. package/src/llama.cpp/{unicode.h → src/unicode.h} +5 -1
  283. package/src/llama.cpp/tests/CMakeLists.txt +19 -20
  284. package/src/llama.cpp/tests/test-backend-ops.cpp +245 -67
  285. package/src/llama.cpp/tests/test-chat-template.cpp +57 -3
  286. package/src/llama.cpp/tests/test-double-float.cpp +2 -2
  287. package/src/llama.cpp/tests/test-grad0.cpp +2 -2
  288. package/src/llama.cpp/tests/test-grammar-integration.cpp +978 -31
  289. package/src/llama.cpp/tests/test-grammar-parser.cpp +423 -158
  290. package/src/llama.cpp/tests/test-json-schema-to-grammar.cpp +508 -135
  291. package/src/llama.cpp/tests/test-llama-grammar.cpp +15 -9
  292. package/src/llama.cpp/tests/test-quantize-fns.cpp +1 -1
  293. package/src/llama.cpp/tests/test-quantize-perf.cpp +1 -1
  294. package/src/llama.cpp/tests/test-rope.cpp +3 -4
  295. package/src/llama.cpp/tests/test-sampling.cpp +5 -5
  296. package/src/llama.cpp/tests/test-tokenizer-0.cpp +6 -6
  297. package/src/llama.cpp/tests/test-tokenizer-1-bpe.cpp +20 -15
  298. package/src/llama.cpp/tests/test-tokenizer-1-spm.cpp +22 -11
  299. package/bin/darwin/arm64/default.metallib +0 -0
  300. package/bin/darwin/x64/default.metallib +0 -0
  301. package/src/llama.cpp/examples/beam-search/CMakeLists.txt +0 -5
  302. package/src/llama.cpp/examples/beam-search/beam-search.cpp +0 -188
  303. package/src/llama.cpp/examples/finetune/finetune.cpp +0 -1862
  304. package/src/llama.cpp/examples/llama.android/llama/CMakeLists.txt +0 -55
  305. package/src/llama.cpp/examples/train-text-from-scratch/CMakeLists.txt +0 -5
  306. package/src/llama.cpp/examples/train-text-from-scratch/train-text-from-scratch.cpp +0 -1253
  307. package/src/llama.cpp/ggml-opencl.cpp +0 -2305
  308. package/src/llama.cpp/ggml-opencl.h +0 -36
  309. package/src/llama.cpp/ggml-sycl.cpp +0 -17340
  310. package/src/llama.cpp/ggml-vulkan-shaders.hpp +0 -81211
  311. package/src/llama.cpp/requirements/requirements-convert-hf-to-gguf-update.txt +0 -2
  312. package/src/llama.cpp/requirements/requirements-convert-hf-to-gguf.txt +0 -2
  313. package/src/llama.cpp/requirements/requirements-convert-llama-ggml-to-gguf.txt +0 -1
  314. package/src/llama.cpp/scripts/gen-build-info-cpp.cmake +0 -24
  315. /package/src/llama.cpp/{ggml-alloc.h → ggml/include/ggml-alloc.h} +0 -0
  316. /package/src/llama.cpp/{ggml-kompute.h → ggml/include/ggml-kompute.h} +0 -0
  317. /package/src/llama.cpp/{ggml-rpc.h → ggml/include/ggml-rpc.h} +0 -0
  318. /package/src/llama.cpp/{ggml-vulkan.h → ggml/include/ggml-vulkan.h} +0 -0
  319. /package/src/llama.cpp/{unicode-data.h → src/unicode-data.h} +0 -0
@@ -0,0 +1,221 @@
1
+ /* Crypto/Sha256.c -- SHA-256 Hash
2
+ 2010-06-11 : Igor Pavlov : Public domain
3
+ This code is based on public domain code from Wei Dai's Crypto++ library. */
4
+
5
+ #include "rotate-bits/rotate-bits.h"
6
+ #include "sha256.h"
7
+
8
+ /* define it for speed optimization */
9
+ #define _SHA256_UNROLL
10
+ #define _SHA256_UNROLL2
11
+
12
+ void
13
+ sha256_init(sha256_t *p)
14
+ {
15
+ p->state[0] = 0x6a09e667;
16
+ p->state[1] = 0xbb67ae85;
17
+ p->state[2] = 0x3c6ef372;
18
+ p->state[3] = 0xa54ff53a;
19
+ p->state[4] = 0x510e527f;
20
+ p->state[5] = 0x9b05688c;
21
+ p->state[6] = 0x1f83d9ab;
22
+ p->state[7] = 0x5be0cd19;
23
+ p->count = 0;
24
+ }
25
+
26
+ #define S0(x) (ROTR32(x, 2) ^ ROTR32(x,13) ^ ROTR32(x, 22))
27
+ #define S1(x) (ROTR32(x, 6) ^ ROTR32(x,11) ^ ROTR32(x, 25))
28
+ #define s0(x) (ROTR32(x, 7) ^ ROTR32(x,18) ^ (x >> 3))
29
+ #define s1(x) (ROTR32(x,17) ^ ROTR32(x,19) ^ (x >> 10))
30
+
31
+ #define blk0(i) (W[i] = data[i])
32
+ #define blk2(i) (W[i&15] += s1(W[(i-2)&15]) + W[(i-7)&15] + s0(W[(i-15)&15]))
33
+
34
+ #define Ch(x,y,z) (z^(x&(y^z)))
35
+ #define Maj(x,y,z) ((x&y)|(z&(x|y)))
36
+
37
+ #define a(i) T[(0-(i))&7]
38
+ #define b(i) T[(1-(i))&7]
39
+ #define c(i) T[(2-(i))&7]
40
+ #define d(i) T[(3-(i))&7]
41
+ #define e(i) T[(4-(i))&7]
42
+ #define f(i) T[(5-(i))&7]
43
+ #define g(i) T[(6-(i))&7]
44
+ #define h(i) T[(7-(i))&7]
45
+
46
+
47
+ #ifdef _SHA256_UNROLL2
48
+
49
+ #define R(a,b,c,d,e,f,g,h, i) h += S1(e) + Ch(e,f,g) + K[i+j] + (j?blk2(i):blk0(i));\
50
+ d += h; h += S0(a) + Maj(a, b, c)
51
+
52
+ #define RX_8(i) \
53
+ R(a,b,c,d,e,f,g,h, i); \
54
+ R(h,a,b,c,d,e,f,g, (i+1)); \
55
+ R(g,h,a,b,c,d,e,f, (i+2)); \
56
+ R(f,g,h,a,b,c,d,e, (i+3)); \
57
+ R(e,f,g,h,a,b,c,d, (i+4)); \
58
+ R(d,e,f,g,h,a,b,c, (i+5)); \
59
+ R(c,d,e,f,g,h,a,b, (i+6)); \
60
+ R(b,c,d,e,f,g,h,a, (i+7))
61
+
62
+ #else
63
+
64
+ #define R(i) h(i) += S1(e(i)) + Ch(e(i),f(i),g(i)) + K[i+j] + (j?blk2(i):blk0(i));\
65
+ d(i) += h(i); h(i) += S0(a(i)) + Maj(a(i), b(i), c(i))
66
+
67
+ #ifdef _SHA256_UNROLL
68
+
69
+ #define RX_8(i) R(i+0); R(i+1); R(i+2); R(i+3); R(i+4); R(i+5); R(i+6); R(i+7);
70
+
71
+ #endif
72
+
73
+ #endif
74
+
75
+ static const uint32_t K[64] = {
76
+ 0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5,
77
+ 0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5,
78
+ 0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3,
79
+ 0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174,
80
+ 0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc,
81
+ 0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da,
82
+ 0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7,
83
+ 0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967,
84
+ 0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13,
85
+ 0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85,
86
+ 0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3,
87
+ 0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070,
88
+ 0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5,
89
+ 0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3,
90
+ 0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208,
91
+ 0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2
92
+ };
93
+
94
+ static void
95
+ sha256_transform(uint32_t *state, const uint32_t *data)
96
+ {
97
+ uint32_t W[16] = {0};
98
+ unsigned j;
99
+ #ifdef _SHA256_UNROLL2
100
+ uint32_t a,b,c,d,e,f,g,h;
101
+ a = state[0];
102
+ b = state[1];
103
+ c = state[2];
104
+ d = state[3];
105
+ e = state[4];
106
+ f = state[5];
107
+ g = state[6];
108
+ h = state[7];
109
+ #else
110
+ uint32_t T[8];
111
+ for (j = 0; j < 8; j++)
112
+ T[j] = state[j];
113
+ #endif
114
+
115
+ for (j = 0; j < 64; j += 16)
116
+ {
117
+ #if defined(_SHA256_UNROLL) || defined(_SHA256_UNROLL2)
118
+ RX_8(0); RX_8(8);
119
+ #else
120
+ unsigned i;
121
+ for (i = 0; i < 16; i++) { R(i); }
122
+ #endif
123
+ }
124
+
125
+ #ifdef _SHA256_UNROLL2
126
+ state[0] += a;
127
+ state[1] += b;
128
+ state[2] += c;
129
+ state[3] += d;
130
+ state[4] += e;
131
+ state[5] += f;
132
+ state[6] += g;
133
+ state[7] += h;
134
+ #else
135
+ for (j = 0; j < 8; j++)
136
+ state[j] += T[j];
137
+ #endif
138
+
139
+ /* Wipe variables */
140
+ /* memset(W, 0, sizeof(W)); */
141
+ /* memset(T, 0, sizeof(T)); */
142
+ }
143
+
144
+ #undef S0
145
+ #undef S1
146
+ #undef s0
147
+ #undef s1
148
+
149
+ static void
150
+ sha256_write_byte_block(sha256_t *p)
151
+ {
152
+ uint32_t data32[16];
153
+ unsigned i;
154
+ for (i = 0; i < 16; i++)
155
+ data32[i] =
156
+ ((uint32_t)(p->buffer[i * 4 ]) << 24) +
157
+ ((uint32_t)(p->buffer[i * 4 + 1]) << 16) +
158
+ ((uint32_t)(p->buffer[i * 4 + 2]) << 8) +
159
+ ((uint32_t)(p->buffer[i * 4 + 3]));
160
+ sha256_transform(p->state, data32);
161
+ }
162
+
163
+
164
+ void
165
+ sha256_hash(unsigned char *buf, const unsigned char *data, size_t size)
166
+ {
167
+ sha256_t hash;
168
+ sha256_init(&hash);
169
+ sha256_update(&hash, data, size);
170
+ sha256_final(&hash, buf);
171
+ }
172
+
173
+
174
+ void
175
+ sha256_update(sha256_t *p, const unsigned char *data, size_t size)
176
+ {
177
+ uint32_t curBufferPos = (uint32_t)p->count & 0x3F;
178
+ while (size > 0)
179
+ {
180
+ p->buffer[curBufferPos++] = *data++;
181
+ p->count++;
182
+ size--;
183
+ if (curBufferPos == 64)
184
+ {
185
+ curBufferPos = 0;
186
+ sha256_write_byte_block(p);
187
+ }
188
+ }
189
+ }
190
+
191
+
192
+ void
193
+ sha256_final(sha256_t *p, unsigned char *digest)
194
+ {
195
+ uint64_t lenInBits = (p->count << 3);
196
+ uint32_t curBufferPos = (uint32_t)p->count & 0x3F;
197
+ unsigned i;
198
+ p->buffer[curBufferPos++] = 0x80;
199
+ while (curBufferPos != (64 - 8))
200
+ {
201
+ curBufferPos &= 0x3F;
202
+ if (curBufferPos == 0)
203
+ sha256_write_byte_block(p);
204
+ p->buffer[curBufferPos++] = 0;
205
+ }
206
+ for (i = 0; i < 8; i++)
207
+ {
208
+ p->buffer[curBufferPos++] = (unsigned char)(lenInBits >> 56);
209
+ lenInBits <<= 8;
210
+ }
211
+ sha256_write_byte_block(p);
212
+
213
+ for (i = 0; i < 8; i++)
214
+ {
215
+ *digest++ = (unsigned char)(p->state[i] >> 24);
216
+ *digest++ = (unsigned char)(p->state[i] >> 16);
217
+ *digest++ = (unsigned char)(p->state[i] >> 8);
218
+ *digest++ = (unsigned char)(p->state[i]);
219
+ }
220
+ sha256_init(p);
221
+ }
@@ -0,0 +1,24 @@
1
+ /* Sha256.h -- SHA-256 Hash
2
+ 2010-06-11 : Igor Pavlov : Public domain */
3
+
4
+ #ifndef __CRYPTO_SHA256_H
5
+ #define __CRYPTO_SHA256_H
6
+
7
+ #include <stdlib.h>
8
+ #include <stdint.h>
9
+
10
+ #define SHA256_DIGEST_SIZE 32
11
+
12
+ typedef struct sha256_t
13
+ {
14
+ uint32_t state[8];
15
+ uint64_t count;
16
+ unsigned char buffer[64];
17
+ } sha256_t;
18
+
19
+ void sha256_init(sha256_t *p);
20
+ void sha256_update(sha256_t *p, const unsigned char *data, size_t size);
21
+ void sha256_final(sha256_t *p, unsigned char *digest);
22
+ void sha256_hash(unsigned char *buf, const unsigned char *data, size_t size);
23
+
24
+ #endif
@@ -0,0 +1,42 @@
1
+ /*
2
+ * xxHash - Extremely Fast Hash algorithm
3
+ * Copyright (C) 2012-2023 Yann Collet
4
+ *
5
+ * BSD 2-Clause License (https://www.opensource.org/licenses/bsd-license.php)
6
+ *
7
+ * Redistribution and use in source and binary forms, with or without
8
+ * modification, are permitted provided that the following conditions are
9
+ * met:
10
+ *
11
+ * * Redistributions of source code must retain the above copyright
12
+ * notice, this list of conditions and the following disclaimer.
13
+ * * Redistributions in binary form must reproduce the above
14
+ * copyright notice, this list of conditions and the following disclaimer
15
+ * in the documentation and/or other materials provided with the
16
+ * distribution.
17
+ *
18
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
19
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
20
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
21
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
22
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
23
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
24
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
25
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
26
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
27
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
28
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
29
+ *
30
+ * You can contact the author at:
31
+ * - xxHash homepage: https://www.xxhash.com
32
+ * - xxHash source repository: https://github.com/Cyan4973/xxHash
33
+ */
34
+
35
+ /*
36
+ * xxhash.c instantiates functions defined in xxhash.h
37
+ */
38
+
39
+ #define XXH_STATIC_LINKING_ONLY /* access advanced declarations */
40
+ #define XXH_IMPLEMENTATION /* access definitions */
41
+
42
+ #include "xxhash.h"