@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.
Files changed (320) hide show
  1. package/CMakeLists.txt +5 -2
  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 +1 -1
  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/LoadSessionWorker.cpp +1 -0
  23. package/src/llama.cpp/.github/workflows/bench.yml +310 -0
  24. package/src/llama.cpp/.github/workflows/build.yml +1315 -0
  25. package/src/llama.cpp/.github/workflows/close-issue.yml +23 -0
  26. package/src/llama.cpp/.github/workflows/docker.yml +116 -0
  27. package/src/llama.cpp/.github/workflows/editorconfig.yml +27 -0
  28. package/src/llama.cpp/.github/workflows/gguf-publish.yml +44 -0
  29. package/src/llama.cpp/.github/workflows/labeler.yml +17 -0
  30. package/src/llama.cpp/.github/workflows/nix-ci-aarch64.yml +65 -0
  31. package/src/llama.cpp/.github/workflows/nix-ci.yml +72 -0
  32. package/src/llama.cpp/.github/workflows/nix-flake-update.yml +22 -0
  33. package/src/llama.cpp/.github/workflows/nix-publish-flake.yml +36 -0
  34. package/src/llama.cpp/.github/workflows/python-check-requirements.yml +35 -0
  35. package/src/llama.cpp/.github/workflows/python-lint.yml +23 -0
  36. package/src/llama.cpp/.github/workflows/python-type-check.yml +38 -0
  37. package/src/llama.cpp/.github/workflows/server.yml +183 -0
  38. package/src/llama.cpp/CMakeLists.txt +91 -1245
  39. package/src/llama.cpp/cmake/arm64-windows-llvm.cmake +1 -1
  40. package/src/llama.cpp/cmake/build-info.cmake +58 -0
  41. package/src/llama.cpp/cmake/git-vars.cmake +22 -0
  42. package/src/llama.cpp/common/CMakeLists.txt +4 -3
  43. package/src/llama.cpp/common/build-info.cpp.in +4 -0
  44. package/src/llama.cpp/common/common.cpp +1116 -877
  45. package/src/llama.cpp/common/common.h +191 -77
  46. package/src/llama.cpp/common/grammar-parser.cpp +118 -31
  47. package/src/llama.cpp/common/json-schema-to-grammar.cpp +346 -65
  48. package/src/llama.cpp/common/log.h +1 -1
  49. package/src/llama.cpp/common/ngram-cache.h +10 -3
  50. package/src/llama.cpp/common/sampling.cpp +19 -10
  51. package/src/llama.cpp/docs/build.md +353 -0
  52. package/src/llama.cpp/examples/CMakeLists.txt +22 -22
  53. package/src/llama.cpp/examples/baby-llama/CMakeLists.txt +1 -1
  54. package/src/llama.cpp/examples/baby-llama/baby-llama.cpp +6 -6
  55. package/src/llama.cpp/examples/batched/CMakeLists.txt +1 -1
  56. package/src/llama.cpp/examples/batched/batched.cpp +52 -55
  57. package/src/llama.cpp/examples/batched-bench/CMakeLists.txt +1 -1
  58. package/src/llama.cpp/examples/batched-bench/batched-bench.cpp +20 -72
  59. package/src/llama.cpp/examples/benchmark/CMakeLists.txt +1 -1
  60. package/src/llama.cpp/examples/chat-13B.bat +57 -0
  61. package/src/llama.cpp/examples/convert-llama2c-to-ggml/CMakeLists.txt +1 -1
  62. package/src/llama.cpp/examples/{finetune → cvector-generator}/CMakeLists.txt +2 -2
  63. package/src/llama.cpp/examples/cvector-generator/completions.txt +582 -0
  64. package/src/llama.cpp/examples/cvector-generator/cvector-generator.cpp +503 -0
  65. package/src/llama.cpp/examples/cvector-generator/mean.hpp +48 -0
  66. package/src/llama.cpp/examples/cvector-generator/negative.txt +4 -0
  67. package/src/llama.cpp/examples/cvector-generator/pca.hpp +325 -0
  68. package/src/llama.cpp/examples/cvector-generator/positive.txt +4 -0
  69. package/src/llama.cpp/examples/deprecation-warning/deprecation-warning.cpp +35 -0
  70. package/src/llama.cpp/examples/embedding/CMakeLists.txt +1 -1
  71. package/src/llama.cpp/examples/embedding/embedding.cpp +94 -46
  72. package/src/llama.cpp/examples/eval-callback/CMakeLists.txt +2 -2
  73. package/src/llama.cpp/examples/eval-callback/eval-callback.cpp +4 -6
  74. package/src/llama.cpp/examples/export-lora/CMakeLists.txt +1 -1
  75. package/src/llama.cpp/examples/export-lora/export-lora.cpp +344 -386
  76. package/src/llama.cpp/examples/gbnf-validator/CMakeLists.txt +2 -2
  77. package/src/llama.cpp/examples/gbnf-validator/gbnf-validator.cpp +30 -25
  78. package/src/llama.cpp/examples/gguf/CMakeLists.txt +1 -1
  79. package/src/llama.cpp/examples/gguf/gguf.cpp +5 -0
  80. package/src/llama.cpp/examples/gguf-hash/CMakeLists.txt +15 -0
  81. package/src/llama.cpp/examples/gguf-hash/deps/rotate-bits/rotate-bits.h +46 -0
  82. package/src/llama.cpp/examples/gguf-hash/deps/sha1/sha1.c +295 -0
  83. package/src/llama.cpp/examples/gguf-hash/deps/sha1/sha1.h +52 -0
  84. package/src/llama.cpp/examples/gguf-hash/deps/sha256/sha256.c +221 -0
  85. package/src/llama.cpp/examples/gguf-hash/deps/sha256/sha256.h +24 -0
  86. package/src/llama.cpp/examples/gguf-hash/deps/xxhash/xxhash.c +42 -0
  87. package/src/llama.cpp/examples/gguf-hash/deps/xxhash/xxhash.h +7093 -0
  88. package/src/llama.cpp/examples/gguf-hash/gguf-hash.cpp +693 -0
  89. package/src/llama.cpp/examples/gguf-split/CMakeLists.txt +1 -1
  90. package/src/llama.cpp/examples/gguf-split/gguf-split.cpp +3 -3
  91. package/src/llama.cpp/examples/gritlm/CMakeLists.txt +1 -1
  92. package/src/llama.cpp/examples/gritlm/gritlm.cpp +6 -2
  93. package/src/llama.cpp/examples/imatrix/CMakeLists.txt +1 -1
  94. package/src/llama.cpp/examples/imatrix/imatrix.cpp +137 -176
  95. package/src/llama.cpp/examples/infill/CMakeLists.txt +1 -1
  96. package/src/llama.cpp/examples/infill/infill.cpp +38 -153
  97. package/src/llama.cpp/examples/llama-bench/llama-bench.cpp +175 -94
  98. package/src/llama.cpp/examples/llama.android/app/build.gradle.kts +65 -0
  99. package/src/llama.cpp/examples/llama.android/build.gradle.kts +6 -0
  100. package/src/llama.cpp/examples/llama.android/llama/build.gradle.kts +68 -0
  101. package/src/llama.cpp/examples/llama.android/llama/src/main/cpp/CMakeLists.txt +11 -7
  102. package/src/llama.cpp/examples/llama.android/llama/src/main/cpp/llama-android.cpp +2 -2
  103. package/src/llama.cpp/examples/llama.android/settings.gradle.kts +18 -0
  104. package/src/llama.cpp/examples/llava/CMakeLists.txt +6 -5
  105. package/src/llama.cpp/examples/llava/android/build_64.sh +8 -0
  106. package/src/llama.cpp/examples/llava/clip.cpp +23 -14
  107. package/src/llama.cpp/examples/llava/llava-cli.cpp +8 -6
  108. package/src/llama.cpp/examples/llava/requirements.txt +3 -2
  109. package/src/llama.cpp/examples/lookahead/CMakeLists.txt +1 -1
  110. package/src/llama.cpp/examples/lookahead/lookahead.cpp +2 -1
  111. package/src/llama.cpp/examples/lookup/CMakeLists.txt +4 -4
  112. package/src/llama.cpp/examples/lookup/lookup-create.cpp +2 -0
  113. package/src/llama.cpp/examples/lookup/lookup-merge.cpp +4 -4
  114. package/src/llama.cpp/examples/lookup/lookup-stats.cpp +2 -2
  115. package/src/llama.cpp/examples/lookup/lookup.cpp +1 -1
  116. package/src/llama.cpp/examples/main/CMakeLists.txt +1 -1
  117. package/src/llama.cpp/examples/main/main.cpp +98 -75
  118. package/src/llama.cpp/examples/main-cmake-pkg/CMakeLists.txt +4 -5
  119. package/src/llama.cpp/examples/parallel/CMakeLists.txt +1 -1
  120. package/src/llama.cpp/examples/parallel/parallel.cpp +2 -1
  121. package/src/llama.cpp/examples/passkey/CMakeLists.txt +1 -1
  122. package/src/llama.cpp/examples/passkey/passkey.cpp +23 -43
  123. package/src/llama.cpp/examples/perplexity/CMakeLists.txt +1 -1
  124. package/src/llama.cpp/examples/perplexity/perplexity.cpp +13 -10
  125. package/src/llama.cpp/examples/quantize/CMakeLists.txt +1 -1
  126. package/src/llama.cpp/examples/quantize/quantize.cpp +37 -34
  127. package/src/llama.cpp/examples/quantize-stats/CMakeLists.txt +1 -1
  128. package/src/llama.cpp/examples/quantize-stats/quantize-stats.cpp +1 -1
  129. package/src/llama.cpp/examples/retrieval/CMakeLists.txt +1 -1
  130. package/src/llama.cpp/examples/retrieval/retrieval.cpp +26 -77
  131. package/src/llama.cpp/examples/save-load-state/CMakeLists.txt +1 -1
  132. package/src/llama.cpp/examples/save-load-state/save-load-state.cpp +14 -7
  133. package/src/llama.cpp/examples/server/CMakeLists.txt +26 -2
  134. package/src/llama.cpp/examples/server/server.cpp +274 -671
  135. package/src/llama.cpp/examples/server/tests/requirements.txt +2 -2
  136. package/src/llama.cpp/examples/server/utils.hpp +28 -29
  137. package/src/llama.cpp/examples/simple/CMakeLists.txt +1 -1
  138. package/src/llama.cpp/examples/simple/simple.cpp +21 -29
  139. package/src/llama.cpp/examples/speculative/CMakeLists.txt +1 -1
  140. package/src/llama.cpp/examples/speculative/speculative.cpp +2 -1
  141. package/src/llama.cpp/examples/sycl/CMakeLists.txt +1 -1
  142. package/src/llama.cpp/examples/sycl/build.sh +23 -0
  143. package/src/llama.cpp/examples/sycl/run-llama2.sh +36 -0
  144. package/src/llama.cpp/examples/sycl/win-build-sycl.bat +33 -0
  145. package/src/llama.cpp/examples/sycl/win-run-llama2.bat +9 -0
  146. package/src/llama.cpp/examples/tokenize/CMakeLists.txt +1 -1
  147. package/src/llama.cpp/examples/tokenize/tokenize.cpp +16 -2
  148. package/src/llama.cpp/ggml/CMakeLists.txt +253 -0
  149. package/src/llama.cpp/{cmake → ggml/cmake}/FindSIMD.cmake +6 -6
  150. package/src/llama.cpp/{ggml-backend.h → ggml/include/ggml-backend.h} +22 -17
  151. package/src/llama.cpp/ggml/include/ggml-blas.h +23 -0
  152. package/src/llama.cpp/ggml/include/ggml-cann.h +125 -0
  153. package/src/llama.cpp/{ggml-cuda.h → ggml/include/ggml-cuda.h} +3 -0
  154. package/src/llama.cpp/{ggml-metal.h → ggml/include/ggml-metal.h} +1 -2
  155. package/src/llama.cpp/{ggml-sycl.h → ggml/include/ggml-sycl.h} +3 -10
  156. package/src/llama.cpp/{ggml.h → ggml/include/ggml.h} +80 -85
  157. package/src/llama.cpp/ggml/src/CMakeLists.txt +1329 -0
  158. package/src/llama.cpp/ggml/src/ggml-aarch64.c +2193 -0
  159. package/src/llama.cpp/ggml/src/ggml-aarch64.h +39 -0
  160. package/src/llama.cpp/{ggml-alloc.c → ggml/src/ggml-alloc.c} +100 -49
  161. package/src/llama.cpp/{ggml-backend-impl.h → ggml/src/ggml-backend-impl.h} +20 -8
  162. package/src/llama.cpp/{ggml-backend.c → ggml/src/ggml-backend.c} +307 -167
  163. package/src/llama.cpp/ggml/src/ggml-blas.cpp +367 -0
  164. package/src/llama.cpp/ggml/src/ggml-cann/acl_tensor.cpp +198 -0
  165. package/src/llama.cpp/ggml/src/ggml-cann/acl_tensor.h +230 -0
  166. package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.cpp +2944 -0
  167. package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.h +592 -0
  168. package/src/llama.cpp/ggml/src/ggml-cann/common.h +282 -0
  169. package/src/llama.cpp/ggml/src/ggml-cann/kernels/CMakeLists.txt +32 -0
  170. package/src/llama.cpp/ggml/src/ggml-cann/kernels/ascendc_kernels.h +17 -0
  171. package/src/llama.cpp/ggml/src/ggml-cann/kernels/dup.cpp +223 -0
  172. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f16.cpp +186 -0
  173. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f32.cpp +180 -0
  174. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q4_0.cpp +193 -0
  175. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q8_0.cpp +191 -0
  176. package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp +208 -0
  177. package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp +206 -0
  178. package/src/llama.cpp/ggml/src/ggml-cann.cpp +2023 -0
  179. package/src/llama.cpp/{ggml-common.h → ggml/src/ggml-common.h} +41 -7
  180. package/src/llama.cpp/{ggml-impl.h → ggml/src/ggml-impl.h} +113 -9
  181. package/src/llama.cpp/{ggml-kompute.cpp → ggml/src/ggml-kompute.cpp} +33 -18
  182. package/src/llama.cpp/{ggml-quants.c → ggml/src/ggml-quants.c} +1460 -940
  183. package/src/llama.cpp/{ggml-quants.h → ggml/src/ggml-quants.h} +19 -20
  184. package/src/llama.cpp/{ggml-rpc.cpp → ggml/src/ggml-rpc.cpp} +95 -72
  185. package/src/llama.cpp/ggml/src/ggml-sycl/backend.hpp +27 -0
  186. package/src/llama.cpp/ggml/src/ggml-sycl/common.cpp +53 -0
  187. package/src/llama.cpp/ggml/src/ggml-sycl/common.hpp +355 -0
  188. package/src/llama.cpp/ggml/src/ggml-sycl/concat.cpp +195 -0
  189. package/src/llama.cpp/ggml/src/ggml-sycl/concat.hpp +21 -0
  190. package/src/llama.cpp/ggml/src/ggml-sycl/convert.cpp +547 -0
  191. package/src/llama.cpp/ggml/src/ggml-sycl/convert.hpp +27 -0
  192. package/src/llama.cpp/ggml/src/ggml-sycl/dequantize.hpp +698 -0
  193. package/src/llama.cpp/ggml/src/ggml-sycl/dmmv.cpp +1023 -0
  194. package/src/llama.cpp/ggml/src/ggml-sycl/dmmv.hpp +27 -0
  195. package/src/llama.cpp/ggml/src/ggml-sycl/dpct/helper.hpp +3011 -0
  196. package/src/llama.cpp/ggml/src/ggml-sycl/mmq.cpp +3031 -0
  197. package/src/llama.cpp/ggml/src/ggml-sycl/mmq.hpp +33 -0
  198. package/src/llama.cpp/ggml/src/ggml-sycl/mmvq.cpp +1027 -0
  199. package/src/llama.cpp/ggml/src/ggml-sycl/mmvq.hpp +27 -0
  200. package/src/llama.cpp/ggml/src/ggml-sycl/norm.cpp +374 -0
  201. package/src/llama.cpp/ggml/src/ggml-sycl/norm.hpp +35 -0
  202. package/src/llama.cpp/ggml/src/ggml-sycl/presets.hpp +66 -0
  203. package/src/llama.cpp/ggml/src/ggml-sycl/rope.cpp +275 -0
  204. package/src/llama.cpp/ggml/src/ggml-sycl/rope.hpp +22 -0
  205. package/src/llama.cpp/ggml/src/ggml-sycl/softmax.cpp +251 -0
  206. package/src/llama.cpp/ggml/src/ggml-sycl/softmax.hpp +24 -0
  207. package/src/llama.cpp/ggml/src/ggml-sycl/vecdotq.hpp +1140 -0
  208. package/src/llama.cpp/ggml/src/ggml-sycl.cpp +5314 -0
  209. package/src/llama.cpp/{ggml-vulkan.cpp → ggml/src/ggml-vulkan.cpp} +1781 -1868
  210. package/src/llama.cpp/{ggml.c → ggml/src/ggml.c} +1245 -2087
  211. package/src/llama.cpp/{sgemm.cpp → ggml/src/llamafile/sgemm.cpp} +21 -24
  212. package/src/llama.cpp/{sgemm.h → ggml/src/llamafile/sgemm.h} +1 -1
  213. package/src/llama.cpp/ggml/src/vulkan-shaders/CMakeLists.txt +5 -0
  214. package/src/llama.cpp/ggml/src/vulkan-shaders/vulkan-shaders-gen.cpp +552 -0
  215. package/src/llama.cpp/{llama.h → include/llama.h} +175 -100
  216. package/src/llama.cpp/models/.editorconfig +1 -0
  217. package/src/llama.cpp/models/ggml-vocab-aquila.gguf +0 -0
  218. package/src/llama.cpp/models/ggml-vocab-baichuan.gguf +0 -0
  219. package/src/llama.cpp/models/ggml-vocab-bert-bge.gguf +0 -0
  220. package/src/llama.cpp/models/ggml-vocab-bert-bge.gguf.inp +112 -0
  221. package/src/llama.cpp/models/ggml-vocab-bert-bge.gguf.out +46 -0
  222. package/src/llama.cpp/models/ggml-vocab-command-r.gguf +0 -0
  223. package/src/llama.cpp/models/ggml-vocab-command-r.gguf.inp +112 -0
  224. package/src/llama.cpp/models/ggml-vocab-command-r.gguf.out +46 -0
  225. package/src/llama.cpp/models/ggml-vocab-deepseek-coder.gguf +0 -0
  226. package/src/llama.cpp/models/ggml-vocab-deepseek-coder.gguf.inp +112 -0
  227. package/src/llama.cpp/models/ggml-vocab-deepseek-coder.gguf.out +46 -0
  228. package/src/llama.cpp/models/ggml-vocab-deepseek-llm.gguf +0 -0
  229. package/src/llama.cpp/models/ggml-vocab-deepseek-llm.gguf.inp +112 -0
  230. package/src/llama.cpp/models/ggml-vocab-deepseek-llm.gguf.out +46 -0
  231. package/src/llama.cpp/models/ggml-vocab-falcon.gguf +0 -0
  232. package/src/llama.cpp/models/ggml-vocab-falcon.gguf.inp +112 -0
  233. package/src/llama.cpp/models/ggml-vocab-falcon.gguf.out +46 -0
  234. package/src/llama.cpp/models/ggml-vocab-gpt-2.gguf +0 -0
  235. package/src/llama.cpp/models/ggml-vocab-gpt-2.gguf.inp +112 -0
  236. package/src/llama.cpp/models/ggml-vocab-gpt-2.gguf.out +46 -0
  237. package/src/llama.cpp/models/ggml-vocab-gpt-neox.gguf +0 -0
  238. package/src/llama.cpp/models/ggml-vocab-llama-bpe.gguf +0 -0
  239. package/src/llama.cpp/models/ggml-vocab-llama-bpe.gguf.inp +112 -0
  240. package/src/llama.cpp/models/ggml-vocab-llama-bpe.gguf.out +46 -0
  241. package/src/llama.cpp/models/ggml-vocab-llama-spm.gguf +0 -0
  242. package/src/llama.cpp/models/ggml-vocab-llama-spm.gguf.inp +112 -0
  243. package/src/llama.cpp/models/ggml-vocab-llama-spm.gguf.out +46 -0
  244. package/src/llama.cpp/models/ggml-vocab-mpt.gguf +0 -0
  245. package/src/llama.cpp/models/ggml-vocab-mpt.gguf.inp +112 -0
  246. package/src/llama.cpp/models/ggml-vocab-mpt.gguf.out +46 -0
  247. package/src/llama.cpp/models/ggml-vocab-phi-3.gguf +0 -0
  248. package/src/llama.cpp/models/ggml-vocab-phi-3.gguf.inp +112 -0
  249. package/src/llama.cpp/models/ggml-vocab-phi-3.gguf.out +46 -0
  250. package/src/llama.cpp/models/ggml-vocab-qwen2.gguf +0 -0
  251. package/src/llama.cpp/models/ggml-vocab-qwen2.gguf.inp +112 -0
  252. package/src/llama.cpp/models/ggml-vocab-qwen2.gguf.out +46 -0
  253. package/src/llama.cpp/models/ggml-vocab-refact.gguf +0 -0
  254. package/src/llama.cpp/models/ggml-vocab-refact.gguf.inp +112 -0
  255. package/src/llama.cpp/models/ggml-vocab-refact.gguf.out +46 -0
  256. package/src/llama.cpp/models/ggml-vocab-starcoder.gguf +0 -0
  257. package/src/llama.cpp/models/ggml-vocab-starcoder.gguf.inp +112 -0
  258. package/src/llama.cpp/models/ggml-vocab-starcoder.gguf.out +46 -0
  259. package/src/llama.cpp/pocs/vdot/CMakeLists.txt +2 -2
  260. package/src/llama.cpp/requirements/requirements-all.txt +12 -0
  261. package/src/llama.cpp/requirements/requirements-compare-llama-bench.txt +2 -0
  262. package/src/llama.cpp/requirements/requirements-convert_hf_to_gguf.txt +3 -0
  263. package/src/llama.cpp/requirements/requirements-convert_hf_to_gguf_update.txt +3 -0
  264. package/src/llama.cpp/requirements/{requirements-convert.txt → requirements-convert_legacy_llama.txt} +1 -1
  265. package/src/llama.cpp/requirements/requirements-convert_llama_ggml_to_gguf.txt +1 -0
  266. package/src/llama.cpp/requirements/requirements-convert_lora_to_gguf.txt +2 -0
  267. package/src/llama.cpp/requirements/requirements-pydantic.txt +3 -0
  268. package/src/llama.cpp/requirements/requirements-test-tokenizer-random.txt +1 -0
  269. package/src/llama.cpp/requirements.txt +5 -4
  270. package/src/llama.cpp/scripts/build-info.sh +30 -0
  271. package/src/llama.cpp/scripts/install-oneapi.bat +19 -0
  272. package/src/llama.cpp/src/CMakeLists.txt +33 -0
  273. package/src/llama.cpp/src/llama-grammar.cpp +539 -0
  274. package/src/llama.cpp/src/llama-grammar.h +39 -0
  275. package/src/llama.cpp/src/llama-impl.h +26 -0
  276. package/src/llama.cpp/src/llama-sampling.cpp +635 -0
  277. package/src/llama.cpp/src/llama-sampling.h +56 -0
  278. package/src/llama.cpp/src/llama-vocab.cpp +1721 -0
  279. package/src/llama.cpp/src/llama-vocab.h +130 -0
  280. package/src/llama.cpp/{llama.cpp → src/llama.cpp} +5979 -5260
  281. package/src/llama.cpp/{unicode-data.cpp → src/unicode-data.cpp} +851 -802
  282. package/src/llama.cpp/{unicode.cpp → src/unicode.cpp} +52 -30
  283. package/src/llama.cpp/{unicode.h → src/unicode.h} +5 -1
  284. package/src/llama.cpp/tests/CMakeLists.txt +19 -20
  285. package/src/llama.cpp/tests/test-backend-ops.cpp +245 -67
  286. package/src/llama.cpp/tests/test-chat-template.cpp +57 -3
  287. package/src/llama.cpp/tests/test-double-float.cpp +2 -2
  288. package/src/llama.cpp/tests/test-grad0.cpp +2 -2
  289. package/src/llama.cpp/tests/test-grammar-integration.cpp +978 -31
  290. package/src/llama.cpp/tests/test-grammar-parser.cpp +423 -158
  291. package/src/llama.cpp/tests/test-json-schema-to-grammar.cpp +508 -135
  292. package/src/llama.cpp/tests/test-llama-grammar.cpp +15 -9
  293. package/src/llama.cpp/tests/test-quantize-fns.cpp +1 -1
  294. package/src/llama.cpp/tests/test-quantize-perf.cpp +1 -1
  295. package/src/llama.cpp/tests/test-rope.cpp +3 -4
  296. package/src/llama.cpp/tests/test-sampling.cpp +5 -5
  297. package/src/llama.cpp/tests/test-tokenizer-0.cpp +6 -6
  298. package/src/llama.cpp/tests/test-tokenizer-1-bpe.cpp +20 -15
  299. package/src/llama.cpp/tests/test-tokenizer-1-spm.cpp +22 -11
  300. package/bin/darwin/arm64/default.metallib +0 -0
  301. package/bin/darwin/x64/default.metallib +0 -0
  302. package/src/llama.cpp/examples/beam-search/CMakeLists.txt +0 -5
  303. package/src/llama.cpp/examples/beam-search/beam-search.cpp +0 -188
  304. package/src/llama.cpp/examples/finetune/finetune.cpp +0 -1862
  305. package/src/llama.cpp/examples/llama.android/llama/CMakeLists.txt +0 -55
  306. package/src/llama.cpp/examples/train-text-from-scratch/CMakeLists.txt +0 -5
  307. package/src/llama.cpp/examples/train-text-from-scratch/train-text-from-scratch.cpp +0 -1253
  308. package/src/llama.cpp/ggml-opencl.cpp +0 -2305
  309. package/src/llama.cpp/ggml-opencl.h +0 -36
  310. package/src/llama.cpp/ggml-sycl.cpp +0 -17340
  311. package/src/llama.cpp/ggml-vulkan-shaders.hpp +0 -81211
  312. package/src/llama.cpp/requirements/requirements-convert-hf-to-gguf-update.txt +0 -2
  313. package/src/llama.cpp/requirements/requirements-convert-hf-to-gguf.txt +0 -2
  314. package/src/llama.cpp/requirements/requirements-convert-llama-ggml-to-gguf.txt +0 -1
  315. package/src/llama.cpp/scripts/gen-build-info-cpp.cmake +0 -24
  316. /package/src/llama.cpp/{ggml-alloc.h → ggml/include/ggml-alloc.h} +0 -0
  317. /package/src/llama.cpp/{ggml-kompute.h → ggml/include/ggml-kompute.h} +0 -0
  318. /package/src/llama.cpp/{ggml-rpc.h → ggml/include/ggml-rpc.h} +0 -0
  319. /package/src/llama.cpp/{ggml-vulkan.h → ggml/include/ggml-vulkan.h} +0 -0
  320. /package/src/llama.cpp/{unicode-data.h → src/unicode-data.h} +0 -0
@@ -0,0 +1,183 @@
1
+ # Server build and tests
2
+ name: Server
3
+
4
+ on:
5
+ workflow_dispatch: # allows manual triggering
6
+ inputs:
7
+ sha:
8
+ description: 'Commit SHA1 to build'
9
+ required: false
10
+ type: string
11
+ slow_tests:
12
+ description: 'Run slow tests'
13
+ required: true
14
+ type: boolean
15
+ push:
16
+ branches:
17
+ - master
18
+ paths: ['.github/workflows/server.yml', '**/CMakeLists.txt', '**/Makefile', '**/*.h', '**/*.hpp', '**/*.c', '**/*.cpp', '**/*.cu', '**/*.swift', '**/*.m', 'examples/server/**.*']
19
+ pull_request:
20
+ types: [opened, synchronize, reopened]
21
+ paths: ['.github/workflows/server.yml', '**/CMakeLists.txt', '**/Makefile', '**/*.h', '**/*.hpp', '**/*.c', '**/*.cpp', '**/*.cu', '**/*.swift', '**/*.m', 'examples/server/**.*']
22
+
23
+ concurrency:
24
+ group: ${{ github.workflow }}-${{ github.ref }}-${{ github.head_ref || github.run_id }}
25
+ cancel-in-progress: true
26
+
27
+ jobs:
28
+ server:
29
+ runs-on: ubuntu-latest
30
+
31
+ strategy:
32
+ matrix:
33
+ sanitizer: [ADDRESS, UNDEFINED] # THREAD is broken
34
+ build_type: [RelWithDebInfo]
35
+ include:
36
+ - build_type: Release
37
+ sanitizer: ""
38
+ fail-fast: false # While -DLLAMA_SANITIZE_THREAD=ON is broken
39
+
40
+ steps:
41
+ - name: Dependencies
42
+ id: depends
43
+ run: |
44
+ sudo apt-get update
45
+ sudo apt-get -y install \
46
+ build-essential \
47
+ xxd \
48
+ git \
49
+ cmake \
50
+ curl \
51
+ wget \
52
+ language-pack-en \
53
+ libcurl4-openssl-dev
54
+
55
+ - name: Clone
56
+ id: checkout
57
+ uses: actions/checkout@v4
58
+ with:
59
+ fetch-depth: 0
60
+ ref: ${{ github.event.inputs.sha || github.event.pull_request.head.sha || github.sha || github.head_ref || github.ref_name }}
61
+
62
+ - name: Python setup
63
+ id: setup_python
64
+ uses: actions/setup-python@v5
65
+ with:
66
+ python-version: '3.11'
67
+
68
+ - name: Tests dependencies
69
+ id: test_dependencies
70
+ run: |
71
+ pip install -r examples/server/tests/requirements.txt
72
+
73
+ - name: Verify server deps
74
+ id: verify_server_deps
75
+ run: |
76
+ git config --global --add safe.directory $(realpath .)
77
+ cd examples/server
78
+ git ls-files --others --modified
79
+ git status
80
+ ./deps.sh
81
+ git status
82
+ not_ignored_files="$(git ls-files --others --modified)"
83
+ echo "Modified files: ${not_ignored_files}"
84
+ if [ -n "${not_ignored_files}" ]; then
85
+ echo "Repository is dirty or server deps are not built as expected"
86
+ echo "${not_ignored_files}"
87
+ exit 1
88
+ fi
89
+
90
+ - name: Build (no OpenMP)
91
+ id: cmake_build_no_openmp
92
+ if: ${{ matrix.sanitizer == 'THREAD' }}
93
+ run: |
94
+ cmake -B build \
95
+ -DGGML_NATIVE=OFF \
96
+ -DLLAMA_BUILD_SERVER=ON \
97
+ -DLLAMA_CURL=ON \
98
+ -DCMAKE_BUILD_TYPE=${{ matrix.build_type }} \
99
+ -DLLAMA_SANITIZE_${{ matrix.sanitizer }}=ON \
100
+ -DGGML_OPENMP=OFF ;
101
+ cmake --build build --config ${{ matrix.build_type }} -j $(nproc) --target llama-server
102
+
103
+ - name: Build
104
+ id: cmake_build
105
+ if: ${{ matrix.sanitizer != 'THREAD' }}
106
+ run: |
107
+ cmake -B build \
108
+ -DGGML_NATIVE=OFF \
109
+ -DLLAMA_BUILD_SERVER=ON \
110
+ -DLLAMA_CURL=ON \
111
+ -DCMAKE_BUILD_TYPE=${{ matrix.build_type }} \
112
+ -DLLAMA_SANITIZE_${{ matrix.sanitizer }}=ON ;
113
+ cmake --build build --config ${{ matrix.build_type }} -j $(nproc) --target llama-server
114
+
115
+ - name: Tests
116
+ id: server_integration_tests
117
+ run: |
118
+ cd examples/server/tests
119
+ PORT=8888 ./tests.sh
120
+
121
+ - name: Slow tests
122
+ id: server_integration_tests_slow
123
+ if: ${{ (github.event.schedule || github.event.inputs.slow_tests == 'true') && matrix.build_type == 'Release' }}
124
+ run: |
125
+ cd examples/server/tests
126
+ PORT=8888 ./tests.sh --stop --no-skipped --no-capture --tags slow
127
+
128
+
129
+ server-windows:
130
+ runs-on: windows-2019
131
+
132
+ steps:
133
+ - name: Clone
134
+ id: checkout
135
+ uses: actions/checkout@v4
136
+ with:
137
+ fetch-depth: 0
138
+ ref: ${{ github.event.inputs.sha || github.event.pull_request.head.sha || github.sha || github.head_ref || github.ref_name }}
139
+
140
+ - name: libCURL
141
+ id: get_libcurl
142
+ env:
143
+ CURL_VERSION: 8.6.0_6
144
+ run: |
145
+ curl.exe -o $env:RUNNER_TEMP/curl.zip -L "https://curl.se/windows/dl-${env:CURL_VERSION}/curl-${env:CURL_VERSION}-win64-mingw.zip"
146
+ mkdir $env:RUNNER_TEMP/libcurl
147
+ tar.exe -xvf $env:RUNNER_TEMP/curl.zip --strip-components=1 -C $env:RUNNER_TEMP/libcurl
148
+
149
+ - name: Build
150
+ id: cmake_build
151
+ run: |
152
+ cmake -B build -DLLAMA_CURL=ON -DCURL_LIBRARY="$env:RUNNER_TEMP/libcurl/lib/libcurl.dll.a" -DCURL_INCLUDE_DIR="$env:RUNNER_TEMP/libcurl/include"
153
+ cmake --build build --config Release -j ${env:NUMBER_OF_PROCESSORS} --target llama-server
154
+
155
+ - name: Python setup
156
+ id: setup_python
157
+ uses: actions/setup-python@v5
158
+ with:
159
+ python-version: '3.11'
160
+
161
+ - name: Tests dependencies
162
+ id: test_dependencies
163
+ run: |
164
+ pip install -r examples/server/tests/requirements.txt
165
+
166
+ - name: Copy Libcurl
167
+ id: prepare_libcurl
168
+ run: |
169
+ cp $env:RUNNER_TEMP/libcurl/bin/libcurl-x64.dll ./build/bin/Release/libcurl-x64.dll
170
+
171
+ - name: Tests
172
+ id: server_integration_tests
173
+ if: ${{ !matrix.disabled_on_pr || !github.event.pull_request }}
174
+ run: |
175
+ cd examples/server/tests
176
+ behave.exe --summary --stop --no-capture --exclude 'issues|wrong_usages|passkey' --tags llama.cpp
177
+
178
+ - name: Slow tests
179
+ id: server_integration_tests_slow
180
+ if: ${{ (github.event.schedule || github.event.inputs.slow_tests == 'true') && matrix.build_type == 'Release' }}
181
+ run: |
182
+ cd examples/server/tests
183
+ behave.exe --stop --no-skipped --no-capture --tags slow