@fugood/llama.node 0.3.2 → 0.3.4

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 (286) hide show
  1. package/CMakeLists.txt +7 -0
  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 +18 -1
  17. package/package.json +1 -1
  18. package/src/DetokenizeWorker.cpp +1 -1
  19. package/src/EmbeddingWorker.cpp +17 -7
  20. package/src/EmbeddingWorker.h +2 -1
  21. package/src/LlamaCompletionWorker.cpp +8 -8
  22. package/src/LlamaCompletionWorker.h +2 -2
  23. package/src/LlamaContext.cpp +89 -27
  24. package/src/LlamaContext.h +2 -0
  25. package/src/TokenizeWorker.cpp +1 -1
  26. package/src/common.hpp +4 -4
  27. package/src/llama.cpp/.github/workflows/build.yml +240 -168
  28. package/src/llama.cpp/.github/workflows/docker.yml +8 -8
  29. package/src/llama.cpp/.github/workflows/python-lint.yml +8 -1
  30. package/src/llama.cpp/.github/workflows/server.yml +21 -14
  31. package/src/llama.cpp/CMakeLists.txt +14 -6
  32. package/src/llama.cpp/Sources/llama/llama.h +4 -0
  33. package/src/llama.cpp/cmake/arm64-apple-clang.cmake +16 -0
  34. package/src/llama.cpp/cmake/common.cmake +33 -0
  35. package/src/llama.cpp/cmake/x64-windows-llvm.cmake +11 -0
  36. package/src/llama.cpp/common/CMakeLists.txt +6 -4
  37. package/src/llama.cpp/common/arg.cpp +986 -770
  38. package/src/llama.cpp/common/arg.h +22 -22
  39. package/src/llama.cpp/common/common.cpp +212 -351
  40. package/src/llama.cpp/common/common.h +204 -117
  41. package/src/llama.cpp/common/json-schema-to-grammar.cpp +1 -1
  42. package/src/llama.cpp/common/log.cpp +50 -50
  43. package/src/llama.cpp/common/log.h +18 -18
  44. package/src/llama.cpp/common/ngram-cache.cpp +36 -36
  45. package/src/llama.cpp/common/ngram-cache.h +19 -19
  46. package/src/llama.cpp/common/sampling.cpp +163 -121
  47. package/src/llama.cpp/common/sampling.h +41 -20
  48. package/src/llama.cpp/common/speculative.cpp +274 -0
  49. package/src/llama.cpp/common/speculative.h +28 -0
  50. package/src/llama.cpp/docs/build.md +134 -161
  51. package/src/llama.cpp/examples/CMakeLists.txt +33 -14
  52. package/src/llama.cpp/examples/batched/CMakeLists.txt +1 -1
  53. package/src/llama.cpp/examples/batched/batched.cpp +19 -18
  54. package/src/llama.cpp/examples/batched-bench/CMakeLists.txt +1 -1
  55. package/src/llama.cpp/examples/batched-bench/batched-bench.cpp +10 -11
  56. package/src/llama.cpp/examples/convert-llama2c-to-ggml/CMakeLists.txt +1 -1
  57. package/src/llama.cpp/examples/convert-llama2c-to-ggml/convert-llama2c-to-ggml.cpp +1 -1
  58. package/src/llama.cpp/examples/cvector-generator/CMakeLists.txt +1 -1
  59. package/src/llama.cpp/examples/cvector-generator/cvector-generator.cpp +9 -9
  60. package/src/llama.cpp/examples/deprecation-warning/deprecation-warning.cpp +1 -1
  61. package/src/llama.cpp/examples/embedding/CMakeLists.txt +1 -1
  62. package/src/llama.cpp/examples/embedding/embedding.cpp +12 -12
  63. package/src/llama.cpp/examples/eval-callback/CMakeLists.txt +3 -2
  64. package/src/llama.cpp/examples/eval-callback/eval-callback.cpp +8 -8
  65. package/src/llama.cpp/examples/export-lora/CMakeLists.txt +1 -1
  66. package/src/llama.cpp/examples/export-lora/export-lora.cpp +5 -5
  67. package/src/llama.cpp/examples/gbnf-validator/CMakeLists.txt +1 -1
  68. package/src/llama.cpp/examples/gbnf-validator/gbnf-validator.cpp +4 -7
  69. package/src/llama.cpp/examples/gen-docs/CMakeLists.txt +1 -1
  70. package/src/llama.cpp/examples/gen-docs/gen-docs.cpp +7 -7
  71. package/src/llama.cpp/examples/gguf/CMakeLists.txt +1 -1
  72. package/src/llama.cpp/examples/gguf-hash/CMakeLists.txt +8 -1
  73. package/src/llama.cpp/examples/gguf-split/CMakeLists.txt +1 -1
  74. package/src/llama.cpp/examples/gguf-split/gguf-split.cpp +2 -2
  75. package/src/llama.cpp/examples/gritlm/CMakeLists.txt +1 -1
  76. package/src/llama.cpp/examples/gritlm/gritlm.cpp +18 -18
  77. package/src/llama.cpp/examples/imatrix/CMakeLists.txt +1 -1
  78. package/src/llama.cpp/examples/imatrix/imatrix.cpp +31 -13
  79. package/src/llama.cpp/examples/infill/CMakeLists.txt +1 -1
  80. package/src/llama.cpp/examples/infill/infill.cpp +41 -87
  81. package/src/llama.cpp/examples/llama-bench/CMakeLists.txt +1 -1
  82. package/src/llama.cpp/examples/llama-bench/llama-bench.cpp +439 -459
  83. package/src/llama.cpp/examples/llama.android/llama/build.gradle.kts +2 -0
  84. package/src/llama.cpp/examples/llama.android/llama/src/main/cpp/llama-android.cpp +11 -14
  85. package/src/llama.cpp/examples/llava/CMakeLists.txt +10 -3
  86. package/src/llama.cpp/examples/llava/clip.cpp +263 -66
  87. package/src/llama.cpp/examples/llava/clip.h +8 -2
  88. package/src/llama.cpp/examples/llava/llava-cli.cpp +23 -23
  89. package/src/llama.cpp/examples/llava/llava.cpp +83 -22
  90. package/src/llama.cpp/examples/llava/minicpmv-cli.cpp +21 -21
  91. package/src/llama.cpp/examples/llava/qwen2vl-cli.cpp +581 -0
  92. package/src/llama.cpp/examples/lookahead/CMakeLists.txt +1 -1
  93. package/src/llama.cpp/examples/lookahead/lookahead.cpp +26 -26
  94. package/src/llama.cpp/examples/lookup/CMakeLists.txt +4 -4
  95. package/src/llama.cpp/examples/lookup/lookup-create.cpp +7 -7
  96. package/src/llama.cpp/examples/lookup/lookup-merge.cpp +4 -4
  97. package/src/llama.cpp/examples/lookup/lookup-stats.cpp +16 -15
  98. package/src/llama.cpp/examples/lookup/lookup.cpp +30 -30
  99. package/src/llama.cpp/examples/main/CMakeLists.txt +1 -1
  100. package/src/llama.cpp/examples/main/main.cpp +73 -114
  101. package/src/llama.cpp/examples/main-cmake-pkg/CMakeLists.txt +1 -1
  102. package/src/llama.cpp/examples/parallel/CMakeLists.txt +1 -1
  103. package/src/llama.cpp/examples/parallel/parallel.cpp +18 -19
  104. package/src/llama.cpp/examples/passkey/CMakeLists.txt +1 -1
  105. package/src/llama.cpp/examples/passkey/passkey.cpp +14 -14
  106. package/src/llama.cpp/examples/perplexity/CMakeLists.txt +1 -1
  107. package/src/llama.cpp/examples/perplexity/perplexity.cpp +99 -120
  108. package/src/llama.cpp/examples/quantize/CMakeLists.txt +1 -1
  109. package/src/llama.cpp/examples/quantize/quantize.cpp +0 -3
  110. package/src/llama.cpp/examples/quantize-stats/CMakeLists.txt +1 -1
  111. package/src/llama.cpp/examples/quantize-stats/quantize-stats.cpp +10 -9
  112. package/src/llama.cpp/examples/retrieval/CMakeLists.txt +1 -1
  113. package/src/llama.cpp/examples/retrieval/retrieval.cpp +16 -16
  114. package/src/llama.cpp/examples/rpc/rpc-server.cpp +3 -1
  115. package/src/llama.cpp/examples/run/CMakeLists.txt +5 -0
  116. package/src/llama.cpp/examples/run/run.cpp +911 -0
  117. package/src/llama.cpp/examples/save-load-state/CMakeLists.txt +1 -1
  118. package/src/llama.cpp/examples/save-load-state/save-load-state.cpp +38 -21
  119. package/src/llama.cpp/examples/server/CMakeLists.txt +3 -16
  120. package/src/llama.cpp/examples/server/server.cpp +2073 -1339
  121. package/src/llama.cpp/examples/server/tests/requirements.txt +2 -2
  122. package/src/llama.cpp/examples/server/utils.hpp +354 -277
  123. package/src/llama.cpp/examples/simple/CMakeLists.txt +2 -2
  124. package/src/llama.cpp/examples/simple/simple.cpp +130 -94
  125. package/src/llama.cpp/examples/simple-chat/CMakeLists.txt +5 -0
  126. package/src/llama.cpp/examples/simple-chat/simple-chat.cpp +200 -0
  127. package/src/llama.cpp/examples/speculative/CMakeLists.txt +1 -1
  128. package/src/llama.cpp/examples/speculative/speculative.cpp +68 -64
  129. package/src/llama.cpp/examples/speculative-simple/CMakeLists.txt +5 -0
  130. package/src/llama.cpp/examples/speculative-simple/speculative-simple.cpp +265 -0
  131. package/src/llama.cpp/examples/tokenize/CMakeLists.txt +1 -1
  132. package/src/llama.cpp/examples/tokenize/tokenize.cpp +3 -3
  133. package/src/llama.cpp/examples/tts/CMakeLists.txt +5 -0
  134. package/src/llama.cpp/examples/tts/tts.cpp +932 -0
  135. package/src/llama.cpp/ggml/CMakeLists.txt +54 -36
  136. package/src/llama.cpp/ggml/include/ggml-backend.h +63 -34
  137. package/src/llama.cpp/ggml/include/ggml-blas.h +5 -3
  138. package/src/llama.cpp/ggml/include/ggml-cann.h +9 -7
  139. package/src/llama.cpp/ggml/include/ggml-cpp.h +38 -0
  140. package/src/llama.cpp/ggml/include/ggml-cpu.h +135 -0
  141. package/src/llama.cpp/ggml/include/ggml-cuda.h +12 -12
  142. package/src/llama.cpp/ggml/include/ggml-kompute.h +7 -3
  143. package/src/llama.cpp/ggml/include/ggml-metal.h +11 -7
  144. package/src/llama.cpp/ggml/include/ggml-opencl.h +26 -0
  145. package/src/llama.cpp/ggml/include/ggml-opt.h +216 -0
  146. package/src/llama.cpp/ggml/include/ggml-rpc.h +9 -5
  147. package/src/llama.cpp/ggml/include/ggml-sycl.h +18 -11
  148. package/src/llama.cpp/ggml/include/ggml-vulkan.h +10 -8
  149. package/src/llama.cpp/ggml/include/ggml.h +159 -417
  150. package/src/llama.cpp/ggml/src/CMakeLists.txt +121 -1155
  151. package/src/llama.cpp/ggml/src/ggml-alloc.c +23 -28
  152. package/src/llama.cpp/ggml/src/ggml-backend-impl.h +57 -36
  153. package/src/llama.cpp/ggml/src/ggml-backend-reg.cpp +552 -0
  154. package/src/llama.cpp/ggml/src/ggml-backend.cpp +306 -867
  155. package/src/llama.cpp/ggml/src/ggml-blas/CMakeLists.txt +87 -0
  156. package/src/llama.cpp/ggml/src/{ggml-blas.cpp → ggml-blas/ggml-blas.cpp} +216 -65
  157. package/src/llama.cpp/ggml/src/ggml-cann/CMakeLists.txt +76 -0
  158. package/src/llama.cpp/ggml/src/ggml-cann/aclnn_ops.cpp +456 -111
  159. package/src/llama.cpp/ggml/src/ggml-cann/common.h +6 -3
  160. package/src/llama.cpp/ggml/src/{ggml-cann.cpp → ggml-cann/ggml-cann.cpp} +343 -177
  161. package/src/llama.cpp/ggml/src/ggml-cann/kernels/CMakeLists.txt +2 -5
  162. package/src/llama.cpp/ggml/src/ggml-cann/kernels/dup.cpp +22 -9
  163. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f16.cpp +24 -13
  164. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_f32.cpp +23 -13
  165. package/src/llama.cpp/ggml/src/ggml-cann/kernels/get_row_q4_0.cpp +11 -0
  166. package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f16_q8_0.cpp +10 -0
  167. package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_f32_q8_0.cpp +10 -0
  168. package/src/llama.cpp/ggml/src/ggml-cann/kernels/quantize_float_to_q4_0.cpp +17 -0
  169. package/src/llama.cpp/ggml/src/ggml-common.h +42 -42
  170. package/src/llama.cpp/ggml/src/ggml-cpu/CMakeLists.txt +336 -0
  171. package/src/llama.cpp/ggml/src/ggml-cpu/amx/amx.cpp +220 -0
  172. package/src/llama.cpp/ggml/src/ggml-cpu/amx/amx.h +8 -0
  173. package/src/llama.cpp/ggml/src/ggml-cpu/amx/common.h +91 -0
  174. package/src/llama.cpp/ggml/src/ggml-cpu/amx/mmq.cpp +2511 -0
  175. package/src/llama.cpp/ggml/src/ggml-cpu/amx/mmq.h +10 -0
  176. package/src/llama.cpp/ggml/src/ggml-cpu/cpu-feats-x86.cpp +323 -0
  177. package/src/llama.cpp/ggml/src/{ggml-aarch64.c → ggml-cpu/ggml-cpu-aarch64.cpp} +1299 -246
  178. package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-aarch64.h +8 -0
  179. package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-hbm.cpp +55 -0
  180. package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-hbm.h +8 -0
  181. package/src/llama.cpp/ggml/src/{ggml-cpu-impl.h → ggml-cpu/ggml-cpu-impl.h} +14 -242
  182. package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-quants.c +10835 -0
  183. package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-quants.h +63 -0
  184. package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-traits.cpp +36 -0
  185. package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu-traits.h +38 -0
  186. package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.c +14123 -0
  187. package/src/llama.cpp/ggml/src/ggml-cpu/ggml-cpu.cpp +628 -0
  188. package/src/llama.cpp/ggml/src/{llamafile → ggml-cpu/llamafile}/sgemm.cpp +666 -0
  189. package/src/llama.cpp/ggml/src/ggml-cuda/CMakeLists.txt +152 -0
  190. package/src/llama.cpp/ggml/src/ggml-cuda/vendors/hip.h +8 -0
  191. package/src/llama.cpp/ggml/src/ggml-hip/CMakeLists.txt +104 -0
  192. package/src/llama.cpp/ggml/src/ggml-impl.h +393 -22
  193. package/src/llama.cpp/ggml/src/ggml-kompute/CMakeLists.txt +166 -0
  194. package/src/llama.cpp/ggml/src/{ggml-kompute.cpp → ggml-kompute/ggml-kompute.cpp} +360 -127
  195. package/src/llama.cpp/ggml/src/ggml-metal/CMakeLists.txt +105 -0
  196. package/src/llama.cpp/ggml/src/ggml-metal/ggml-metal-impl.h +288 -0
  197. package/src/llama.cpp/ggml/src/ggml-musa/CMakeLists.txt +107 -0
  198. package/src/llama.cpp/ggml/src/ggml-opencl/CMakeLists.txt +147 -0
  199. package/src/llama.cpp/ggml/src/ggml-opencl/ggml-opencl.cpp +4004 -0
  200. package/src/llama.cpp/ggml/src/ggml-opt.cpp +854 -0
  201. package/src/llama.cpp/ggml/src/ggml-quants.c +188 -10702
  202. package/src/llama.cpp/ggml/src/ggml-quants.h +78 -125
  203. package/src/llama.cpp/ggml/src/ggml-rpc/CMakeLists.txt +9 -0
  204. package/src/llama.cpp/ggml/src/{ggml-rpc.cpp → ggml-rpc/ggml-rpc.cpp} +478 -300
  205. package/src/llama.cpp/ggml/src/ggml-sycl/CMakeLists.txt +84 -0
  206. package/src/llama.cpp/ggml/src/ggml-sycl/backend.hpp +3 -0
  207. package/src/llama.cpp/ggml/src/ggml-sycl/common.cpp +36 -5
  208. package/src/llama.cpp/ggml/src/ggml-sycl/common.hpp +259 -0
  209. package/src/llama.cpp/ggml/src/ggml-sycl/concat.cpp +3 -2
  210. package/src/llama.cpp/ggml/src/ggml-sycl/convert.cpp +1 -1
  211. package/src/llama.cpp/ggml/src/ggml-sycl/dmmv.cpp +5 -5
  212. package/src/llama.cpp/ggml/src/ggml-sycl/dpct/helper.hpp +34 -35
  213. package/src/llama.cpp/ggml/src/ggml-sycl/element_wise.cpp +1030 -0
  214. package/src/llama.cpp/ggml/src/ggml-sycl/element_wise.hpp +76 -0
  215. package/src/llama.cpp/ggml/src/ggml-sycl/gemm.hpp +4 -4
  216. package/src/llama.cpp/ggml/src/{ggml-sycl.cpp → ggml-sycl/ggml-sycl.cpp} +3638 -4151
  217. package/src/llama.cpp/ggml/src/ggml-sycl/im2col.cpp +3 -2
  218. package/src/llama.cpp/ggml/src/ggml-sycl/mmq.cpp +6 -6
  219. package/src/llama.cpp/ggml/src/ggml-sycl/mmvq.cpp +75 -87
  220. package/src/llama.cpp/ggml/src/ggml-sycl/norm.cpp +7 -6
  221. package/src/llama.cpp/ggml/src/ggml-sycl/outprod.cpp +56 -0
  222. package/src/llama.cpp/ggml/src/ggml-sycl/outprod.hpp +11 -0
  223. package/src/llama.cpp/ggml/src/ggml-sycl/presets.hpp +6 -0
  224. package/src/llama.cpp/ggml/src/ggml-sycl/rope.cpp +4 -3
  225. package/src/llama.cpp/ggml/src/ggml-sycl/softmax.cpp +7 -7
  226. package/src/llama.cpp/ggml/src/ggml-sycl/tsembd.cpp +1 -0
  227. package/src/llama.cpp/ggml/src/ggml-sycl/vecdotq.hpp +4 -4
  228. package/src/llama.cpp/ggml/src/ggml-sycl/wkv6.cpp +141 -0
  229. package/src/llama.cpp/ggml/src/ggml-sycl/wkv6.hpp +10 -0
  230. package/src/llama.cpp/ggml/src/ggml-threading.cpp +12 -0
  231. package/src/llama.cpp/ggml/src/ggml-threading.h +14 -0
  232. package/src/llama.cpp/ggml/src/ggml-vulkan/CMakeLists.txt +92 -0
  233. package/src/llama.cpp/ggml/src/{ggml-vulkan.cpp → ggml-vulkan/ggml-vulkan.cpp} +2138 -887
  234. package/src/llama.cpp/ggml/src/{vulkan-shaders → ggml-vulkan/vulkan-shaders}/CMakeLists.txt +3 -1
  235. package/src/llama.cpp/ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp +593 -0
  236. package/src/llama.cpp/ggml/src/ggml.c +4427 -20125
  237. package/src/llama.cpp/include/llama-cpp.h +25 -0
  238. package/src/llama.cpp/include/llama.h +93 -52
  239. package/src/llama.cpp/models/ggml-vocab-roberta-bpe.gguf.inp +112 -0
  240. package/src/llama.cpp/models/ggml-vocab-roberta-bpe.gguf.out +46 -0
  241. package/src/llama.cpp/pocs/CMakeLists.txt +3 -1
  242. package/src/llama.cpp/pocs/vdot/CMakeLists.txt +2 -2
  243. package/src/llama.cpp/pocs/vdot/q8dot.cpp +4 -3
  244. package/src/llama.cpp/pocs/vdot/vdot.cpp +8 -7
  245. package/src/llama.cpp/src/CMakeLists.txt +4 -8
  246. package/src/llama.cpp/src/llama-grammar.cpp +15 -15
  247. package/src/llama.cpp/src/llama-grammar.h +2 -5
  248. package/src/llama.cpp/src/llama-sampling.cpp +779 -194
  249. package/src/llama.cpp/src/llama-sampling.h +21 -2
  250. package/src/llama.cpp/src/llama-vocab.cpp +55 -10
  251. package/src/llama.cpp/src/llama-vocab.h +35 -11
  252. package/src/llama.cpp/src/llama.cpp +4317 -2979
  253. package/src/llama.cpp/src/unicode-data.cpp +2 -2
  254. package/src/llama.cpp/src/unicode.cpp +62 -51
  255. package/src/llama.cpp/src/unicode.h +9 -10
  256. package/src/llama.cpp/tests/CMakeLists.txt +48 -38
  257. package/src/llama.cpp/tests/test-arg-parser.cpp +15 -15
  258. package/src/llama.cpp/tests/test-backend-ops.cpp +324 -80
  259. package/src/llama.cpp/tests/test-barrier.cpp +1 -0
  260. package/src/llama.cpp/tests/test-chat-template.cpp +59 -9
  261. package/src/llama.cpp/tests/test-gguf.cpp +1303 -0
  262. package/src/llama.cpp/tests/test-grammar-integration.cpp +3 -6
  263. package/src/llama.cpp/tests/test-json-schema-to-grammar.cpp +17 -4
  264. package/src/llama.cpp/tests/test-llama-grammar.cpp +2 -4
  265. package/src/llama.cpp/tests/test-log.cpp +2 -2
  266. package/src/llama.cpp/tests/test-opt.cpp +853 -142
  267. package/src/llama.cpp/tests/test-quantize-fns.cpp +24 -21
  268. package/src/llama.cpp/tests/test-quantize-perf.cpp +16 -14
  269. package/src/llama.cpp/tests/test-rope.cpp +62 -20
  270. package/src/llama.cpp/tests/test-sampling.cpp +163 -138
  271. package/src/llama.cpp/tests/test-tokenizer-0.cpp +7 -7
  272. package/src/llama.cpp/tests/test-tokenizer-1-bpe.cpp +5 -5
  273. package/src/llama.cpp/tests/test-tokenizer-1-spm.cpp +5 -5
  274. package/src/llama.cpp/.github/workflows/nix-ci-aarch64.yml +0 -72
  275. package/src/llama.cpp/.github/workflows/nix-ci.yml +0 -79
  276. package/src/llama.cpp/.github/workflows/nix-flake-update.yml +0 -22
  277. package/src/llama.cpp/.github/workflows/nix-publish-flake.yml +0 -36
  278. package/src/llama.cpp/common/train.cpp +0 -1515
  279. package/src/llama.cpp/common/train.h +0 -233
  280. package/src/llama.cpp/examples/baby-llama/CMakeLists.txt +0 -5
  281. package/src/llama.cpp/examples/baby-llama/baby-llama.cpp +0 -1639
  282. package/src/llama.cpp/ggml/src/ggml-aarch64.h +0 -39
  283. package/src/llama.cpp/ggml/src/vulkan-shaders/vulkan-shaders-gen.cpp +0 -600
  284. package/src/llama.cpp/tests/test-grad0.cpp +0 -1683
  285. /package/src/llama.cpp/ggml/{cmake → src/ggml-cpu/cmake}/FindSIMD.cmake +0 -0
  286. /package/src/llama.cpp/ggml/src/{llamafile → ggml-cpu/llamafile}/sgemm.h +0 -0
@@ -211,17 +211,20 @@ struct ggml_cann_pool_alloc {
211
211
  struct ggml_backend_cann_context {
212
212
  int32_t device; /**< Device ID. */
213
213
  std::string name; /**< Name of the device. */
214
+ std::string description; /**< Description of the device. */
214
215
  aclrtEvent copy_event = nullptr; /**< Event for managing copy operations. */
215
216
 
216
- aclrtStream streams[GGML_CANN_MAX_STREAMS] = {
217
- {nullptr}}; /**< Array of streams for the device. */
217
+ aclrtStream streams[GGML_CANN_MAX_STREAMS] = {nullptr}; /**< Array of streams for the device. */
218
218
 
219
219
  /**
220
220
  * @brief Constructor for initializing the context with a given device.
221
221
  * @param device Device ID.
222
222
  */
223
223
  explicit ggml_backend_cann_context(int device)
224
- : device(device), name("CANN" + std::to_string(device)) {}
224
+ : device(device), name("CANN" + std::to_string(device)) {
225
+ ggml_cann_set_device(device);
226
+ description = aclrtGetSocName();
227
+ }
225
228
 
226
229
  /**
227
230
  * @brief Destructor for cleaning up resources.