node-llama-cpp 3.0.0-beta.2 → 3.0.0-beta.21

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 (641) hide show
  1. package/README.md +14 -11
  2. package/dist/ChatWrapper.d.ts +2 -15
  3. package/dist/ChatWrapper.js +28 -33
  4. package/dist/ChatWrapper.js.map +1 -1
  5. package/dist/apiDocsOverrides.d.ts +1 -0
  6. package/dist/apiDocsOverrides.js +5 -0
  7. package/dist/apiDocsOverrides.js.map +1 -0
  8. package/dist/{utils/getBin.d.ts → bindings/AddonTypes.d.ts} +54 -7
  9. package/dist/bindings/AddonTypes.js +2 -0
  10. package/dist/bindings/AddonTypes.js.map +1 -0
  11. package/dist/bindings/Llama.d.ts +47 -0
  12. package/dist/bindings/Llama.js +343 -0
  13. package/dist/bindings/Llama.js.map +1 -0
  14. package/dist/bindings/consts.d.ts +2 -0
  15. package/dist/bindings/consts.js +11 -0
  16. package/dist/bindings/consts.js.map +1 -0
  17. package/dist/bindings/getLlama.d.ts +145 -0
  18. package/dist/bindings/getLlama.js +389 -0
  19. package/dist/bindings/getLlama.js.map +1 -0
  20. package/dist/bindings/types.d.ts +55 -0
  21. package/dist/bindings/types.js +77 -0
  22. package/dist/bindings/types.js.map +1 -0
  23. package/dist/bindings/utils/MemoryOrchestrator.d.ts +21 -0
  24. package/dist/bindings/utils/MemoryOrchestrator.js +49 -0
  25. package/dist/bindings/utils/MemoryOrchestrator.js.map +1 -0
  26. package/dist/bindings/utils/NoBinaryFoundError.d.ts +2 -0
  27. package/dist/bindings/utils/NoBinaryFoundError.js +7 -0
  28. package/dist/bindings/utils/NoBinaryFoundError.js.map +1 -0
  29. package/dist/bindings/utils/asyncEvery.d.ts +5 -0
  30. package/dist/bindings/utils/asyncEvery.js +15 -0
  31. package/dist/bindings/utils/asyncEvery.js.map +1 -0
  32. package/dist/bindings/utils/asyncSome.d.ts +5 -0
  33. package/dist/bindings/utils/asyncSome.js +27 -0
  34. package/dist/bindings/utils/asyncSome.js.map +1 -0
  35. package/dist/{utils → bindings/utils}/binariesGithubRelease.js +1 -1
  36. package/dist/bindings/utils/binariesGithubRelease.js.map +1 -0
  37. package/dist/bindings/utils/clearAllLocalBuilds.d.ts +1 -0
  38. package/dist/bindings/utils/clearAllLocalBuilds.js +47 -0
  39. package/dist/bindings/utils/clearAllLocalBuilds.js.map +1 -0
  40. package/dist/bindings/utils/cloneLlamaCppRepo.d.ts +11 -0
  41. package/dist/bindings/utils/cloneLlamaCppRepo.js +166 -0
  42. package/dist/bindings/utils/cloneLlamaCppRepo.js.map +1 -0
  43. package/dist/bindings/utils/compileLLamaCpp.d.ts +15 -0
  44. package/dist/bindings/utils/compileLLamaCpp.js +221 -0
  45. package/dist/bindings/utils/compileLLamaCpp.js.map +1 -0
  46. package/dist/bindings/utils/detectAvailableComputeLayers.d.ts +14 -0
  47. package/dist/bindings/utils/detectAvailableComputeLayers.js +304 -0
  48. package/dist/bindings/utils/detectAvailableComputeLayers.js.map +1 -0
  49. package/dist/bindings/utils/detectGlibc.d.ts +4 -0
  50. package/dist/bindings/utils/detectGlibc.js +46 -0
  51. package/dist/bindings/utils/detectGlibc.js.map +1 -0
  52. package/dist/bindings/utils/getBestComputeLayersAvailable.d.ts +9 -0
  53. package/dist/bindings/utils/getBestComputeLayersAvailable.js +29 -0
  54. package/dist/bindings/utils/getBestComputeLayersAvailable.js.map +1 -0
  55. package/dist/bindings/utils/getBuildFolderNameForBuildOptions.d.ts +5 -0
  56. package/dist/bindings/utils/getBuildFolderNameForBuildOptions.js +93 -0
  57. package/dist/bindings/utils/getBuildFolderNameForBuildOptions.js.map +1 -0
  58. package/dist/bindings/utils/getCanUsePrebuiltBinaries.d.ts +1 -0
  59. package/dist/bindings/utils/getCanUsePrebuiltBinaries.js +8 -0
  60. package/dist/bindings/utils/getCanUsePrebuiltBinaries.js.map +1 -0
  61. package/dist/bindings/utils/getExampleUsageCodeOfGetLlama.d.ts +2 -0
  62. package/dist/bindings/utils/getExampleUsageCodeOfGetLlama.js +21 -0
  63. package/dist/bindings/utils/getExampleUsageCodeOfGetLlama.js.map +1 -0
  64. package/dist/bindings/utils/getGpuTypesToUseForOption.d.ts +11 -0
  65. package/dist/bindings/utils/getGpuTypesToUseForOption.js +30 -0
  66. package/dist/bindings/utils/getGpuTypesToUseForOption.js.map +1 -0
  67. package/dist/bindings/utils/getLinuxDistroInfo.d.ts +9 -0
  68. package/dist/bindings/utils/getLinuxDistroInfo.js +46 -0
  69. package/dist/bindings/utils/getLinuxDistroInfo.js.map +1 -0
  70. package/dist/bindings/utils/getLlamaWithoutBackend.d.ts +5 -0
  71. package/dist/bindings/utils/getLlamaWithoutBackend.js +27 -0
  72. package/dist/bindings/utils/getLlamaWithoutBackend.js.map +1 -0
  73. package/dist/bindings/utils/getPlatform.d.ts +2 -0
  74. package/dist/bindings/utils/getPlatform.js +15 -0
  75. package/dist/bindings/utils/getPlatform.js.map +1 -0
  76. package/dist/bindings/utils/getPlatformInfo.d.ts +5 -0
  77. package/dist/bindings/utils/getPlatformInfo.js +28 -0
  78. package/dist/bindings/utils/getPlatformInfo.js.map +1 -0
  79. package/dist/bindings/utils/hasBuildingFromSourceDependenciesInstalled.d.ts +3 -0
  80. package/dist/bindings/utils/hasBuildingFromSourceDependenciesInstalled.js +27 -0
  81. package/dist/bindings/utils/hasBuildingFromSourceDependenciesInstalled.js.map +1 -0
  82. package/dist/bindings/utils/hasFileInPath.d.ts +2 -0
  83. package/dist/bindings/utils/hasFileInPath.js +34 -0
  84. package/dist/bindings/utils/hasFileInPath.js.map +1 -0
  85. package/dist/bindings/utils/lastBuildInfo.d.ts +6 -0
  86. package/dist/bindings/utils/lastBuildInfo.js +17 -0
  87. package/dist/bindings/utils/lastBuildInfo.js.map +1 -0
  88. package/dist/bindings/utils/logBinaryUsageExampleToConsole.d.ts +2 -0
  89. package/dist/bindings/utils/logBinaryUsageExampleToConsole.js +22 -0
  90. package/dist/bindings/utils/logBinaryUsageExampleToConsole.js.map +1 -0
  91. package/dist/bindings/utils/logDistroInstallInstruction.d.ts +14 -0
  92. package/dist/bindings/utils/logDistroInstallInstruction.js +48 -0
  93. package/dist/bindings/utils/logDistroInstallInstruction.js.map +1 -0
  94. package/dist/bindings/utils/resolveCustomCmakeOptions.d.ts +1 -0
  95. package/dist/bindings/utils/resolveCustomCmakeOptions.js +45 -0
  96. package/dist/bindings/utils/resolveCustomCmakeOptions.js.map +1 -0
  97. package/dist/bindings/utils/testBindingBinary.d.ts +1 -0
  98. package/dist/bindings/utils/testBindingBinary.js +98 -0
  99. package/dist/bindings/utils/testBindingBinary.js.map +1 -0
  100. package/dist/bindings/utils/testCmakeBinary.d.ts +5 -0
  101. package/dist/bindings/utils/testCmakeBinary.js +32 -0
  102. package/dist/bindings/utils/testCmakeBinary.js.map +1 -0
  103. package/dist/chatWrappers/AlpacaChatWrapper.d.ts +2 -1
  104. package/dist/chatWrappers/AlpacaChatWrapper.js +9 -2
  105. package/dist/chatWrappers/AlpacaChatWrapper.js.map +1 -1
  106. package/dist/chatWrappers/ChatMLChatWrapper.d.ts +5 -0
  107. package/dist/chatWrappers/ChatMLChatWrapper.js +13 -11
  108. package/dist/chatWrappers/ChatMLChatWrapper.js.map +1 -1
  109. package/dist/chatWrappers/FalconChatWrapper.d.ts +2 -1
  110. package/dist/chatWrappers/FalconChatWrapper.js +28 -11
  111. package/dist/chatWrappers/FalconChatWrapper.js.map +1 -1
  112. package/dist/chatWrappers/FunctionaryChatWrapper.js +86 -73
  113. package/dist/chatWrappers/FunctionaryChatWrapper.js.map +1 -1
  114. package/dist/chatWrappers/{LlamaChatWrapper.d.ts → GemmaChatWrapper.d.ts} +6 -1
  115. package/dist/chatWrappers/GemmaChatWrapper.js +88 -0
  116. package/dist/chatWrappers/GemmaChatWrapper.js.map +1 -0
  117. package/dist/chatWrappers/GeneralChatWrapper.d.ts +2 -1
  118. package/dist/chatWrappers/GeneralChatWrapper.js +35 -12
  119. package/dist/chatWrappers/GeneralChatWrapper.js.map +1 -1
  120. package/dist/chatWrappers/Llama2ChatWrapper.d.ts +20 -0
  121. package/dist/chatWrappers/{LlamaChatWrapper.js → Llama2ChatWrapper.js} +29 -11
  122. package/dist/chatWrappers/Llama2ChatWrapper.js.map +1 -0
  123. package/dist/chatWrappers/Llama3ChatWrapper.d.ts +31 -0
  124. package/dist/chatWrappers/Llama3ChatWrapper.js +129 -0
  125. package/dist/chatWrappers/Llama3ChatWrapper.js.map +1 -0
  126. package/dist/chatWrappers/generic/JinjaTemplateChatWrapper.d.ts +73 -0
  127. package/dist/chatWrappers/generic/JinjaTemplateChatWrapper.js +359 -0
  128. package/dist/chatWrappers/generic/JinjaTemplateChatWrapper.js.map +1 -0
  129. package/dist/chatWrappers/generic/TemplateChatWrapper.d.ts +64 -0
  130. package/dist/chatWrappers/generic/TemplateChatWrapper.js +200 -0
  131. package/dist/chatWrappers/generic/TemplateChatWrapper.js.map +1 -0
  132. package/dist/chatWrappers/generic/utils/chatHistoryFunctionCallMessageTemplate.d.ts +33 -0
  133. package/dist/chatWrappers/generic/utils/chatHistoryFunctionCallMessageTemplate.js +45 -0
  134. package/dist/chatWrappers/generic/utils/chatHistoryFunctionCallMessageTemplate.js.map +1 -0
  135. package/dist/chatWrappers/utils/ChatModelFunctionsDocumentationGenerator.d.ts +42 -0
  136. package/dist/chatWrappers/utils/ChatModelFunctionsDocumentationGenerator.js +82 -0
  137. package/dist/chatWrappers/utils/ChatModelFunctionsDocumentationGenerator.js.map +1 -0
  138. package/dist/chatWrappers/utils/isJinjaTemplateEquivalentToSpecializedChatWrapper.d.ts +4 -0
  139. package/dist/chatWrappers/utils/isJinjaTemplateEquivalentToSpecializedChatWrapper.js +206 -0
  140. package/dist/chatWrappers/utils/isJinjaTemplateEquivalentToSpecializedChatWrapper.js.map +1 -0
  141. package/dist/chatWrappers/utils/resolveChatWrapper.d.ts +69 -0
  142. package/dist/chatWrappers/utils/resolveChatWrapper.js +214 -0
  143. package/dist/chatWrappers/utils/resolveChatWrapper.js.map +1 -0
  144. package/dist/cli/cli.js +21 -7
  145. package/dist/cli/cli.js.map +1 -1
  146. package/dist/cli/commands/BuildCommand.d.ts +6 -4
  147. package/dist/cli/commands/BuildCommand.js +103 -41
  148. package/dist/cli/commands/BuildCommand.js.map +1 -1
  149. package/dist/cli/commands/ChatCommand.d.ts +18 -6
  150. package/dist/cli/commands/ChatCommand.js +298 -142
  151. package/dist/cli/commands/ChatCommand.js.map +1 -1
  152. package/dist/cli/commands/ClearCommand.d.ts +1 -1
  153. package/dist/cli/commands/ClearCommand.js +11 -12
  154. package/dist/cli/commands/ClearCommand.js.map +1 -1
  155. package/dist/cli/commands/CompleteCommand.d.ts +29 -0
  156. package/dist/cli/commands/CompleteCommand.js +365 -0
  157. package/dist/cli/commands/CompleteCommand.js.map +1 -0
  158. package/dist/cli/commands/DebugCommand.d.ts +7 -0
  159. package/dist/cli/commands/DebugCommand.js +54 -0
  160. package/dist/cli/commands/DebugCommand.js.map +1 -0
  161. package/dist/cli/commands/DownloadCommand.d.ts +6 -4
  162. package/dist/cli/commands/DownloadCommand.js +120 -69
  163. package/dist/cli/commands/DownloadCommand.js.map +1 -1
  164. package/dist/cli/commands/InfillCommand.d.ts +31 -0
  165. package/dist/cli/commands/InfillCommand.js +401 -0
  166. package/dist/cli/commands/InfillCommand.js.map +1 -0
  167. package/dist/cli/commands/InitCommand.d.ts +11 -0
  168. package/dist/cli/commands/InitCommand.js +195 -0
  169. package/dist/cli/commands/InitCommand.js.map +1 -0
  170. package/dist/cli/commands/OnPostInstallCommand.js +9 -10
  171. package/dist/cli/commands/OnPostInstallCommand.js.map +1 -1
  172. package/dist/cli/commands/PullCommand.d.ts +12 -0
  173. package/dist/cli/commands/PullCommand.js +117 -0
  174. package/dist/cli/commands/PullCommand.js.map +1 -0
  175. package/dist/cli/commands/inspect/InspectCommand.d.ts +4 -0
  176. package/dist/cli/commands/inspect/InspectCommand.js +19 -0
  177. package/dist/cli/commands/inspect/InspectCommand.js.map +1 -0
  178. package/dist/cli/commands/inspect/commands/InspectGgufCommand.d.ts +12 -0
  179. package/dist/cli/commands/inspect/commands/InspectGgufCommand.js +136 -0
  180. package/dist/cli/commands/inspect/commands/InspectGgufCommand.js.map +1 -0
  181. package/dist/cli/commands/inspect/commands/InspectGpuCommand.d.ts +4 -0
  182. package/dist/cli/commands/inspect/commands/InspectGpuCommand.js +138 -0
  183. package/dist/cli/commands/inspect/commands/InspectGpuCommand.js.map +1 -0
  184. package/dist/cli/commands/inspect/commands/InspectMeasureCommand.d.ts +17 -0
  185. package/dist/cli/commands/inspect/commands/InspectMeasureCommand.js +613 -0
  186. package/dist/cli/commands/inspect/commands/InspectMeasureCommand.js.map +1 -0
  187. package/dist/cli/projectTemplates.d.ts +7 -0
  188. package/dist/cli/projectTemplates.js +10 -0
  189. package/dist/cli/projectTemplates.js.map +1 -0
  190. package/dist/cli/recommendedModels.d.ts +2 -0
  191. package/dist/cli/recommendedModels.js +342 -0
  192. package/dist/cli/recommendedModels.js.map +1 -0
  193. package/dist/cli/startCreateCli.d.ts +2 -0
  194. package/dist/cli/startCreateCli.js +26 -0
  195. package/dist/cli/startCreateCli.js.map +1 -0
  196. package/dist/cli/utils/ConsoleInteraction.d.ts +23 -0
  197. package/dist/cli/utils/ConsoleInteraction.js +122 -0
  198. package/dist/cli/utils/ConsoleInteraction.js.map +1 -0
  199. package/dist/cli/utils/ConsoleTable.d.ts +23 -0
  200. package/dist/cli/utils/ConsoleTable.js +86 -0
  201. package/dist/cli/utils/ConsoleTable.js.map +1 -0
  202. package/dist/cli/utils/basicChooseFromListConsoleInteraction.d.ts +13 -0
  203. package/dist/cli/utils/basicChooseFromListConsoleInteraction.js +111 -0
  204. package/dist/cli/utils/basicChooseFromListConsoleInteraction.js.map +1 -0
  205. package/dist/cli/utils/consolePromptQuestion.d.ts +6 -0
  206. package/dist/cli/utils/consolePromptQuestion.js +82 -0
  207. package/dist/cli/utils/consolePromptQuestion.js.map +1 -0
  208. package/dist/cli/utils/getReadablePath.d.ts +1 -0
  209. package/dist/cli/utils/getReadablePath.js +14 -0
  210. package/dist/cli/utils/getReadablePath.js.map +1 -0
  211. package/dist/cli/utils/interactivelyAskForModel.d.ts +7 -0
  212. package/dist/cli/utils/interactivelyAskForModel.js +451 -0
  213. package/dist/cli/utils/interactivelyAskForModel.js.map +1 -0
  214. package/dist/cli/utils/logUsedGpuTypeOption.d.ts +2 -0
  215. package/dist/cli/utils/logUsedGpuTypeOption.js +9 -0
  216. package/dist/cli/utils/logUsedGpuTypeOption.js.map +1 -0
  217. package/dist/cli/utils/printCommonInfoLines.d.ts +9 -0
  218. package/dist/cli/utils/printCommonInfoLines.js +71 -0
  219. package/dist/cli/utils/printCommonInfoLines.js.map +1 -0
  220. package/dist/cli/utils/printInfoLine.d.ts +12 -0
  221. package/dist/cli/utils/printInfoLine.js +54 -0
  222. package/dist/cli/utils/printInfoLine.js.map +1 -0
  223. package/dist/cli/utils/projectTemplates.d.ts +19 -0
  224. package/dist/cli/utils/projectTemplates.js +47 -0
  225. package/dist/cli/utils/projectTemplates.js.map +1 -0
  226. package/dist/cli/utils/resolveCommandGgufPath.d.ts +4 -0
  227. package/dist/cli/utils/resolveCommandGgufPath.js +71 -0
  228. package/dist/cli/utils/resolveCommandGgufPath.js.map +1 -0
  229. package/dist/cli/utils/resolveHeaderFlag.d.ts +1 -0
  230. package/dist/cli/utils/resolveHeaderFlag.js +21 -0
  231. package/dist/cli/utils/resolveHeaderFlag.js.map +1 -0
  232. package/dist/cli/utils/resolveModelRecommendationFileOptions.d.ts +19 -0
  233. package/dist/cli/utils/resolveModelRecommendationFileOptions.js +7 -0
  234. package/dist/cli/utils/resolveModelRecommendationFileOptions.js.map +1 -0
  235. package/dist/cli/utils/splitAnsiToLines.d.ts +1 -0
  236. package/dist/cli/utils/splitAnsiToLines.js +32 -0
  237. package/dist/cli/utils/splitAnsiToLines.js.map +1 -0
  238. package/dist/cli/utils/withCliCommandDescriptionDocsUrl.d.ts +2 -0
  239. package/dist/cli/utils/withCliCommandDescriptionDocsUrl.js +23 -0
  240. package/dist/cli/utils/withCliCommandDescriptionDocsUrl.js.map +1 -0
  241. package/dist/commands.d.ts +1 -0
  242. package/dist/commands.js +3 -0
  243. package/dist/commands.js.map +1 -1
  244. package/dist/config.d.ts +38 -5
  245. package/dist/config.js +61 -16
  246. package/dist/config.js.map +1 -1
  247. package/dist/consts.d.ts +3 -0
  248. package/dist/consts.js +10 -0
  249. package/dist/consts.js.map +1 -0
  250. package/dist/{llamaEvaluator → evaluator}/LlamaChat/LlamaChat.d.ts +37 -35
  251. package/dist/{llamaEvaluator → evaluator}/LlamaChat/LlamaChat.js +298 -221
  252. package/dist/evaluator/LlamaChat/LlamaChat.js.map +1 -0
  253. package/dist/{llamaEvaluator → evaluator}/LlamaChat/utils/FunctionCallGrammar.d.ts +2 -1
  254. package/dist/{llamaEvaluator → evaluator}/LlamaChat/utils/FunctionCallGrammar.js +5 -3
  255. package/dist/evaluator/LlamaChat/utils/FunctionCallGrammar.js.map +1 -0
  256. package/dist/{llamaEvaluator → evaluator}/LlamaChat/utils/contextShiftStrategies/eraseFirstResponseAndKeepFirstSystemChatContextShiftStrategy.js +18 -0
  257. package/dist/evaluator/LlamaChat/utils/contextShiftStrategies/eraseFirstResponseAndKeepFirstSystemChatContextShiftStrategy.js.map +1 -0
  258. package/dist/{llamaEvaluator → evaluator}/LlamaChatSession/LlamaChatSession.d.ts +40 -3
  259. package/dist/{llamaEvaluator → evaluator}/LlamaChatSession/LlamaChatSession.js +28 -7
  260. package/dist/evaluator/LlamaChatSession/LlamaChatSession.js.map +1 -0
  261. package/dist/{llamaEvaluator → evaluator}/LlamaChatSession/utils/defineChatSessionFunction.d.ts +3 -0
  262. package/dist/{llamaEvaluator → evaluator}/LlamaChatSession/utils/defineChatSessionFunction.js +3 -0
  263. package/dist/evaluator/LlamaChatSession/utils/defineChatSessionFunction.js.map +1 -0
  264. package/dist/evaluator/LlamaCompletion.d.ts +155 -0
  265. package/dist/evaluator/LlamaCompletion.js +405 -0
  266. package/dist/evaluator/LlamaCompletion.js.map +1 -0
  267. package/dist/{llamaEvaluator → evaluator}/LlamaContext/LlamaContext.d.ts +41 -20
  268. package/dist/{llamaEvaluator → evaluator}/LlamaContext/LlamaContext.js +271 -81
  269. package/dist/evaluator/LlamaContext/LlamaContext.js.map +1 -0
  270. package/dist/evaluator/LlamaContext/types.d.ts +140 -0
  271. package/dist/evaluator/LlamaContext/types.js.map +1 -0
  272. package/dist/evaluator/LlamaContext/utils/batchItemsPrioritizationStrategies/firstInFirstOutStrategy.js.map +1 -0
  273. package/dist/evaluator/LlamaContext/utils/batchItemsPrioritizationStrategies/maximumParallelismStrategy.js.map +1 -0
  274. package/dist/evaluator/LlamaContext/utils/resolveBatchItemsPrioritizationStrategy.d.ts +2 -0
  275. package/dist/{llamaEvaluator/LlamaContext/utils/resolveBatchItemsPrioritizingStrategy.js → evaluator/LlamaContext/utils/resolveBatchItemsPrioritizationStrategy.js} +4 -4
  276. package/dist/evaluator/LlamaContext/utils/resolveBatchItemsPrioritizationStrategy.js.map +1 -0
  277. package/dist/evaluator/LlamaEmbeddingContext.d.ts +51 -0
  278. package/dist/evaluator/LlamaEmbeddingContext.js +73 -0
  279. package/dist/evaluator/LlamaEmbeddingContext.js.map +1 -0
  280. package/dist/{llamaEvaluator → evaluator}/LlamaGrammar.d.ts +8 -5
  281. package/dist/{llamaEvaluator → evaluator}/LlamaGrammar.js +13 -10
  282. package/dist/evaluator/LlamaGrammar.js.map +1 -0
  283. package/dist/{llamaEvaluator → evaluator}/LlamaGrammarEvaluationState.js +4 -4
  284. package/dist/evaluator/LlamaGrammarEvaluationState.js.map +1 -0
  285. package/dist/{llamaEvaluator → evaluator}/LlamaJsonSchemaGrammar.d.ts +2 -1
  286. package/dist/{llamaEvaluator → evaluator}/LlamaJsonSchemaGrammar.js +3 -3
  287. package/dist/evaluator/LlamaJsonSchemaGrammar.js.map +1 -0
  288. package/dist/evaluator/LlamaModel.d.ts +230 -0
  289. package/dist/evaluator/LlamaModel.js +597 -0
  290. package/dist/evaluator/LlamaModel.js.map +1 -0
  291. package/dist/evaluator/TokenBias.d.ts +22 -0
  292. package/dist/evaluator/TokenBias.js +33 -0
  293. package/dist/evaluator/TokenBias.js.map +1 -0
  294. package/dist/evaluator/TokenMeter.d.ts +54 -0
  295. package/dist/evaluator/TokenMeter.js +86 -0
  296. package/dist/evaluator/TokenMeter.js.map +1 -0
  297. package/dist/gguf/consts.d.ts +3 -0
  298. package/dist/gguf/consts.js +8 -0
  299. package/dist/gguf/consts.js.map +1 -0
  300. package/dist/gguf/errors/InvalidGgufMagicError.d.ts +3 -0
  301. package/dist/gguf/errors/InvalidGgufMagicError.js +6 -0
  302. package/dist/gguf/errors/InvalidGgufMagicError.js.map +1 -0
  303. package/dist/gguf/errors/UnsupportedGgufValueTypeError.d.ts +4 -0
  304. package/dist/gguf/errors/UnsupportedGgufValueTypeError.js +9 -0
  305. package/dist/gguf/errors/UnsupportedGgufValueTypeError.js.map +1 -0
  306. package/dist/gguf/fileReaders/GgufFileReader.d.ts +33 -0
  307. package/dist/gguf/fileReaders/GgufFileReader.js +76 -0
  308. package/dist/gguf/fileReaders/GgufFileReader.js.map +1 -0
  309. package/dist/gguf/fileReaders/GgufFsFileReader.d.ts +17 -0
  310. package/dist/gguf/fileReaders/GgufFsFileReader.js +45 -0
  311. package/dist/gguf/fileReaders/GgufFsFileReader.js.map +1 -0
  312. package/dist/gguf/fileReaders/GgufNetworkFetchFileReader.d.ts +22 -0
  313. package/dist/gguf/fileReaders/GgufNetworkFetchFileReader.js +63 -0
  314. package/dist/gguf/fileReaders/GgufNetworkFetchFileReader.js.map +1 -0
  315. package/dist/gguf/insights/GgufInsights.d.ts +42 -0
  316. package/dist/gguf/insights/GgufInsights.js +361 -0
  317. package/dist/gguf/insights/GgufInsights.js.map +1 -0
  318. package/dist/gguf/insights/GgufInsightsConfigurationResolver.d.ts +87 -0
  319. package/dist/gguf/insights/GgufInsightsConfigurationResolver.js +136 -0
  320. package/dist/gguf/insights/GgufInsightsConfigurationResolver.js.map +1 -0
  321. package/dist/gguf/insights/utils/resolveContextContextSizeOption.d.ts +18 -0
  322. package/dist/gguf/insights/utils/resolveContextContextSizeOption.js +76 -0
  323. package/dist/gguf/insights/utils/resolveContextContextSizeOption.js.map +1 -0
  324. package/dist/gguf/insights/utils/resolveModelGpuLayersOption.d.ts +14 -0
  325. package/dist/gguf/insights/utils/resolveModelGpuLayersOption.js +177 -0
  326. package/dist/gguf/insights/utils/resolveModelGpuLayersOption.js.map +1 -0
  327. package/dist/gguf/insights/utils/scoreLevels.d.ts +5 -0
  328. package/dist/gguf/insights/utils/scoreLevels.js +16 -0
  329. package/dist/gguf/insights/utils/scoreLevels.js.map +1 -0
  330. package/dist/gguf/parser/GgufV2Parser.d.ts +19 -0
  331. package/dist/gguf/parser/GgufV2Parser.js +115 -0
  332. package/dist/gguf/parser/GgufV2Parser.js.map +1 -0
  333. package/dist/gguf/parser/GgufV3Parser.d.ts +3 -0
  334. package/dist/gguf/parser/GgufV3Parser.js +4 -0
  335. package/dist/gguf/parser/GgufV3Parser.js.map +1 -0
  336. package/dist/gguf/parser/parseGguf.d.ts +8 -0
  337. package/dist/gguf/parser/parseGguf.js +63 -0
  338. package/dist/gguf/parser/parseGguf.js.map +1 -0
  339. package/dist/gguf/readGgufFileInfo.d.ts +33 -0
  340. package/dist/gguf/readGgufFileInfo.js +66 -0
  341. package/dist/gguf/readGgufFileInfo.js.map +1 -0
  342. package/dist/gguf/types/GgufFileInfoTypes.d.ts +84 -0
  343. package/dist/gguf/types/GgufFileInfoTypes.js +18 -0
  344. package/dist/gguf/types/GgufFileInfoTypes.js.map +1 -0
  345. package/dist/gguf/types/GgufMetadataTypes.d.ts +334 -0
  346. package/dist/gguf/types/GgufMetadataTypes.js +86 -0
  347. package/dist/gguf/types/GgufMetadataTypes.js.map +1 -0
  348. package/dist/gguf/types/GgufTensorInfoTypes.d.ts +37 -0
  349. package/dist/gguf/types/GgufTensorInfoTypes.js +33 -0
  350. package/dist/gguf/types/GgufTensorInfoTypes.js.map +1 -0
  351. package/dist/gguf/utils/GgufReadOffset.d.ts +6 -0
  352. package/dist/gguf/utils/GgufReadOffset.js +18 -0
  353. package/dist/gguf/utils/GgufReadOffset.js.map +1 -0
  354. package/dist/gguf/utils/convertMetadataKeyValueRecordToNestedObject.d.ts +5 -0
  355. package/dist/gguf/utils/convertMetadataKeyValueRecordToNestedObject.js +38 -0
  356. package/dist/gguf/utils/convertMetadataKeyValueRecordToNestedObject.js.map +1 -0
  357. package/dist/gguf/utils/getGgufFileTypeName.d.ts +4 -0
  358. package/dist/gguf/utils/getGgufFileTypeName.js +13 -0
  359. package/dist/gguf/utils/getGgufFileTypeName.js.map +1 -0
  360. package/dist/gguf/utils/getGgufMetadataArchitectureData.d.ts +3 -0
  361. package/dist/gguf/utils/getGgufMetadataArchitectureData.js +4 -0
  362. package/dist/gguf/utils/getGgufMetadataArchitectureData.js.map +1 -0
  363. package/dist/gguf/utils/normalizeGgufDownloadUrl.d.ts +1 -0
  364. package/dist/gguf/utils/normalizeGgufDownloadUrl.js +16 -0
  365. package/dist/gguf/utils/normalizeGgufDownloadUrl.js.map +1 -0
  366. package/dist/gguf/utils/resolveBinarySplitGgufPartUrls.d.ts +2 -0
  367. package/dist/gguf/utils/resolveBinarySplitGgufPartUrls.js +39 -0
  368. package/dist/gguf/utils/resolveBinarySplitGgufPartUrls.js.map +1 -0
  369. package/dist/gguf/utils/resolveSplitGgufParts.d.ts +7 -0
  370. package/dist/gguf/utils/resolveSplitGgufParts.js +55 -0
  371. package/dist/gguf/utils/resolveSplitGgufParts.js.map +1 -0
  372. package/dist/index.d.ts +37 -17
  373. package/dist/index.js +33 -14
  374. package/dist/index.js.map +1 -1
  375. package/dist/state.d.ts +4 -0
  376. package/dist/state.js +14 -0
  377. package/dist/state.js.map +1 -1
  378. package/dist/types.d.ts +53 -2
  379. package/dist/types.js.map +1 -1
  380. package/dist/utils/DisposeGuard.d.ts +13 -0
  381. package/dist/utils/DisposeGuard.js +120 -0
  382. package/dist/utils/DisposeGuard.js.map +1 -0
  383. package/dist/utils/InsufficientMemoryError.d.ts +3 -0
  384. package/dist/utils/InsufficientMemoryError.js +6 -0
  385. package/dist/utils/InsufficientMemoryError.js.map +1 -0
  386. package/dist/utils/LlamaText.d.ts +50 -25
  387. package/dist/utils/LlamaText.js +367 -155
  388. package/dist/utils/LlamaText.js.map +1 -1
  389. package/dist/utils/StopGenerationDetector.d.ts +1 -1
  390. package/dist/utils/StopGenerationDetector.js +23 -18
  391. package/dist/utils/StopGenerationDetector.js.map +1 -1
  392. package/dist/utils/TokenStreamRegulator.d.ts +8 -4
  393. package/dist/utils/TokenStreamRegulator.js +78 -8
  394. package/dist/utils/TokenStreamRegulator.js.map +1 -1
  395. package/dist/utils/UnsupportedError.d.ts +2 -0
  396. package/dist/utils/UnsupportedError.js +7 -0
  397. package/dist/utils/UnsupportedError.js.map +1 -0
  398. package/dist/utils/cmake.js +38 -20
  399. package/dist/utils/cmake.js.map +1 -1
  400. package/dist/utils/createModelDownloader.d.ts +102 -0
  401. package/dist/utils/createModelDownloader.js +226 -0
  402. package/dist/utils/createModelDownloader.js.map +1 -0
  403. package/dist/utils/findBestOption.d.ts +4 -0
  404. package/dist/utils/findBestOption.js +15 -0
  405. package/dist/utils/findBestOption.js.map +1 -0
  406. package/dist/utils/findCharacterRemovalCountToFitChatHistoryInContext.js +18 -8
  407. package/dist/utils/findCharacterRemovalCountToFitChatHistoryInContext.js.map +1 -1
  408. package/dist/utils/gbnfJson/getGbnfGrammarForGbnfJsonSchema.d.ts +5 -0
  409. package/dist/utils/gbnfJson/getGbnfGrammarForGbnfJsonSchema.js +11 -0
  410. package/dist/utils/gbnfJson/getGbnfGrammarForGbnfJsonSchema.js.map +1 -0
  411. package/dist/utils/gbnfJson/terminals/GbnfArray.d.ts +3 -1
  412. package/dist/utils/gbnfJson/terminals/GbnfArray.js +10 -5
  413. package/dist/utils/gbnfJson/terminals/GbnfArray.js.map +1 -1
  414. package/dist/utils/gbnfJson/terminals/GbnfBoolean.d.ts +1 -1
  415. package/dist/utils/gbnfJson/terminals/GbnfBoolean.js.map +1 -1
  416. package/dist/utils/gbnfJson/terminals/GbnfBooleanValue.js.map +1 -1
  417. package/dist/utils/gbnfJson/terminals/GbnfGrammar.js.map +1 -1
  418. package/dist/utils/gbnfJson/terminals/GbnfNull.d.ts +1 -1
  419. package/dist/utils/gbnfJson/terminals/GbnfNull.js.map +1 -1
  420. package/dist/utils/gbnfJson/terminals/GbnfNumber.d.ts +1 -1
  421. package/dist/utils/gbnfJson/terminals/GbnfNumber.js.map +1 -1
  422. package/dist/utils/gbnfJson/terminals/GbnfNumberValue.js.map +1 -1
  423. package/dist/utils/gbnfJson/terminals/GbnfObjectMap.d.ts +3 -1
  424. package/dist/utils/gbnfJson/terminals/GbnfObjectMap.js +9 -4
  425. package/dist/utils/gbnfJson/terminals/GbnfObjectMap.js.map +1 -1
  426. package/dist/utils/gbnfJson/terminals/GbnfOr.js.map +1 -1
  427. package/dist/utils/gbnfJson/terminals/GbnfRepetition.d.ts +9 -0
  428. package/dist/utils/gbnfJson/terminals/GbnfRepetition.js +37 -0
  429. package/dist/utils/gbnfJson/terminals/GbnfRepetition.js.map +1 -0
  430. package/dist/utils/gbnfJson/terminals/GbnfString.d.ts +1 -1
  431. package/dist/utils/gbnfJson/terminals/GbnfString.js +23 -5
  432. package/dist/utils/gbnfJson/terminals/GbnfString.js.map +1 -1
  433. package/dist/utils/gbnfJson/terminals/GbnfStringValue.js.map +1 -1
  434. package/dist/utils/gbnfJson/terminals/GbnfVerbatimText.js.map +1 -1
  435. package/dist/utils/gbnfJson/terminals/GbnfWhitespace.d.ts +7 -4
  436. package/dist/utils/gbnfJson/terminals/GbnfWhitespace.js +37 -9
  437. package/dist/utils/gbnfJson/terminals/GbnfWhitespace.js.map +1 -1
  438. package/dist/utils/gbnfJson/terminals/gbnfConsts.d.ts +5 -4
  439. package/dist/utils/gbnfJson/terminals/gbnfConsts.js +14 -3
  440. package/dist/utils/gbnfJson/terminals/gbnfConsts.js.map +1 -1
  441. package/dist/utils/gbnfJson/utils/GbnfJsonScopeState.d.ts +10 -0
  442. package/dist/utils/gbnfJson/utils/GbnfJsonScopeState.js +15 -0
  443. package/dist/utils/gbnfJson/utils/GbnfJsonScopeState.js.map +1 -0
  444. package/dist/utils/gbnfJson/utils/getGbnfJsonTerminalForGbnfJsonSchema.d.ts +2 -1
  445. package/dist/utils/gbnfJson/utils/getGbnfJsonTerminalForGbnfJsonSchema.js +6 -5
  446. package/dist/utils/gbnfJson/utils/getGbnfJsonTerminalForGbnfJsonSchema.js.map +1 -1
  447. package/dist/utils/getBuildDefaults.d.ts +1 -2
  448. package/dist/utils/getBuildDefaults.js +2 -3
  449. package/dist/utils/getBuildDefaults.js.map +1 -1
  450. package/dist/utils/getConsoleLogPrefix.d.ts +1 -0
  451. package/dist/utils/getConsoleLogPrefix.js +10 -0
  452. package/dist/utils/getConsoleLogPrefix.js.map +1 -0
  453. package/dist/utils/getGrammarsFolder.d.ts +2 -1
  454. package/dist/utils/getGrammarsFolder.js +8 -7
  455. package/dist/utils/getGrammarsFolder.js.map +1 -1
  456. package/dist/utils/getModuleVersion.d.ts +1 -0
  457. package/dist/utils/getModuleVersion.js +13 -0
  458. package/dist/utils/getModuleVersion.js.map +1 -0
  459. package/dist/utils/getQueuedTokensBeforeStopTrigger.d.ts +6 -0
  460. package/dist/utils/getQueuedTokensBeforeStopTrigger.js +22 -0
  461. package/dist/utils/getQueuedTokensBeforeStopTrigger.js.map +1 -0
  462. package/dist/utils/getReadableContextSize.d.ts +1 -0
  463. package/dist/utils/getReadableContextSize.js +7 -0
  464. package/dist/utils/getReadableContextSize.js.map +1 -0
  465. package/dist/utils/getTypeScriptTypeStringForGbnfJsonSchema.js +15 -11
  466. package/dist/utils/getTypeScriptTypeStringForGbnfJsonSchema.js.map +1 -1
  467. package/dist/utils/gitReleaseBundles.js +73 -5
  468. package/dist/utils/gitReleaseBundles.js.map +1 -1
  469. package/dist/utils/hashString.d.ts +1 -0
  470. package/dist/utils/hashString.js +8 -0
  471. package/dist/utils/hashString.js.map +1 -0
  472. package/dist/utils/isLockfileActive.d.ts +4 -0
  473. package/dist/utils/isLockfileActive.js +12 -0
  474. package/dist/utils/isLockfileActive.js.map +1 -0
  475. package/dist/utils/isToken.d.ts +2 -0
  476. package/dist/utils/isToken.js +4 -0
  477. package/dist/utils/isToken.js.map +1 -0
  478. package/dist/utils/isUrl.d.ts +1 -0
  479. package/dist/utils/isUrl.js +15 -0
  480. package/dist/utils/isUrl.js.map +1 -0
  481. package/dist/utils/mergeUnionTypes.d.ts +10 -0
  482. package/dist/utils/mergeUnionTypes.js +2 -0
  483. package/dist/utils/mergeUnionTypes.js.map +1 -0
  484. package/dist/utils/parseModelFileName.d.ts +1 -0
  485. package/dist/utils/parseModelFileName.js +6 -1
  486. package/dist/utils/parseModelFileName.js.map +1 -1
  487. package/dist/utils/parseTextTemplate.d.ts +66 -0
  488. package/dist/utils/parseTextTemplate.js +116 -0
  489. package/dist/utils/parseTextTemplate.js.map +1 -0
  490. package/dist/utils/prettyPrintObject.d.ts +10 -0
  491. package/dist/utils/prettyPrintObject.js +84 -0
  492. package/dist/utils/prettyPrintObject.js.map +1 -0
  493. package/dist/utils/removeNullFields.d.ts +2 -1
  494. package/dist/utils/removeNullFields.js +8 -0
  495. package/dist/utils/removeNullFields.js.map +1 -1
  496. package/dist/utils/resolveGithubRelease.d.ts +2 -0
  497. package/dist/utils/resolveGithubRelease.js +36 -0
  498. package/dist/utils/resolveGithubRelease.js.map +1 -0
  499. package/dist/utils/runtime.d.ts +4 -0
  500. package/dist/utils/runtime.js +8 -0
  501. package/dist/utils/runtime.js.map +1 -0
  502. package/dist/utils/spawnCommand.d.ts +11 -1
  503. package/dist/utils/spawnCommand.js +56 -6
  504. package/dist/utils/spawnCommand.js.map +1 -1
  505. package/dist/utils/tokenizeInput.d.ts +3 -0
  506. package/dist/utils/tokenizeInput.js +12 -0
  507. package/dist/utils/tokenizeInput.js.map +1 -0
  508. package/dist/utils/utilTypes.d.ts +3 -0
  509. package/dist/utils/utilTypes.js +2 -0
  510. package/dist/utils/utilTypes.js.map +1 -0
  511. package/dist/utils/waitForLockfileRelease.d.ts +5 -0
  512. package/dist/utils/waitForLockfileRelease.js +20 -0
  513. package/dist/utils/waitForLockfileRelease.js.map +1 -0
  514. package/dist/utils/withLockfile.d.ts +7 -0
  515. package/dist/utils/withLockfile.js +44 -0
  516. package/dist/utils/withLockfile.js.map +1 -0
  517. package/dist/utils/withOra.d.ts +2 -0
  518. package/dist/utils/withOra.js +22 -6
  519. package/dist/utils/withOra.js.map +1 -1
  520. package/dist/utils/withProgressLog.d.ts +23 -0
  521. package/dist/utils/withProgressLog.js +211 -0
  522. package/dist/utils/withProgressLog.js.map +1 -0
  523. package/dist/utils/withStatusLogs.d.ts +2 -1
  524. package/dist/utils/withStatusLogs.js +12 -9
  525. package/dist/utils/withStatusLogs.js.map +1 -1
  526. package/llama/.clang-format +1 -2
  527. package/llama/CMakeLists.txt +115 -4
  528. package/llama/addon.cpp +1318 -99
  529. package/llama/binariesGithubRelease.json +1 -1
  530. package/llama/gitRelease.bundle +0 -0
  531. package/llama/gpuInfo/cuda-gpu-info.cu +120 -0
  532. package/llama/gpuInfo/cuda-gpu-info.h +10 -0
  533. package/llama/gpuInfo/metal-gpu-info.h +8 -0
  534. package/llama/gpuInfo/metal-gpu-info.mm +30 -0
  535. package/llama/gpuInfo/vulkan-gpu-info.cpp +83 -0
  536. package/llama/gpuInfo/vulkan-gpu-info.h +9 -0
  537. package/llama/grammars/README.md +11 -1
  538. package/llama/grammars/json.gbnf +1 -1
  539. package/llama/grammars/json_arr.gbnf +1 -1
  540. package/llama/llama.cpp.info.json +4 -0
  541. package/llama/toolchains/win32.host-x64.target-arm64.cmake +41 -0
  542. package/llamaBins/linux-arm64/_nlcBuildMetadata.json +1 -0
  543. package/llamaBins/linux-arm64/llama-addon.node +0 -0
  544. package/llamaBins/linux-armv7l/_nlcBuildMetadata.json +1 -0
  545. package/llamaBins/linux-armv7l/llama-addon.node +0 -0
  546. package/llamaBins/linux-x64/_nlcBuildMetadata.json +1 -0
  547. package/llamaBins/linux-x64/llama-addon.node +0 -0
  548. package/llamaBins/linux-x64-cuda/_nlcBuildMetadata.json +1 -0
  549. package/llamaBins/linux-x64-cuda/llama-addon.node +0 -0
  550. package/llamaBins/linux-x64-vulkan/_nlcBuildMetadata.json +1 -0
  551. package/llamaBins/linux-x64-vulkan/llama-addon.node +0 -0
  552. package/llamaBins/mac-arm64-metal/_nlcBuildMetadata.json +1 -0
  553. package/llamaBins/mac-arm64-metal/default.metallib +0 -0
  554. package/llamaBins/mac-arm64-metal/llama-addon.node +0 -0
  555. package/llamaBins/mac-x64/_nlcBuildMetadata.json +1 -0
  556. package/llamaBins/mac-x64/llama-addon.node +0 -0
  557. package/llamaBins/win-arm64/_nlcBuildMetadata.json +1 -0
  558. package/llamaBins/win-arm64/llama-addon.exp +0 -0
  559. package/llamaBins/win-arm64/llama-addon.lib +0 -0
  560. package/llamaBins/win-arm64/llama-addon.node +0 -0
  561. package/llamaBins/win-x64/_nlcBuildMetadata.json +1 -0
  562. package/llamaBins/win-x64/llama-addon.exp +0 -0
  563. package/llamaBins/win-x64/llama-addon.lib +0 -0
  564. package/llamaBins/win-x64/llama-addon.node +0 -0
  565. package/llamaBins/win-x64-cuda/_nlcBuildMetadata.json +1 -0
  566. package/llamaBins/win-x64-cuda/llama-addon.exp +0 -0
  567. package/llamaBins/win-x64-cuda/llama-addon.lib +0 -0
  568. package/llamaBins/win-x64-cuda/llama-addon.node +0 -0
  569. package/llamaBins/win-x64-vulkan/_nlcBuildMetadata.json +1 -0
  570. package/llamaBins/win-x64-vulkan/llama-addon.exp +0 -0
  571. package/llamaBins/win-x64-vulkan/llama-addon.lib +0 -0
  572. package/llamaBins/win-x64-vulkan/llama-addon.node +0 -0
  573. package/package.json +61 -34
  574. package/templates/packed/electron-typescript-react.json +1 -0
  575. package/templates/packed/node-typescript.json +1 -0
  576. package/templates/packed/node_modules.json +1 -0
  577. package/dist/AbortError.d.ts +0 -2
  578. package/dist/AbortError.js +0 -7
  579. package/dist/AbortError.js.map +0 -1
  580. package/dist/chatWrappers/LlamaChatWrapper.js.map +0 -1
  581. package/dist/chatWrappers/resolveChatWrapperBasedOnModel.d.ts +0 -13
  582. package/dist/chatWrappers/resolveChatWrapperBasedOnModel.js +0 -55
  583. package/dist/chatWrappers/resolveChatWrapperBasedOnModel.js.map +0 -1
  584. package/dist/llamaEvaluator/LlamaBins.d.ts +0 -18
  585. package/dist/llamaEvaluator/LlamaBins.js +0 -5
  586. package/dist/llamaEvaluator/LlamaBins.js.map +0 -1
  587. package/dist/llamaEvaluator/LlamaChat/LlamaChat.js.map +0 -1
  588. package/dist/llamaEvaluator/LlamaChat/utils/FunctionCallGrammar.js.map +0 -1
  589. package/dist/llamaEvaluator/LlamaChat/utils/contextShiftStrategies/eraseFirstResponseAndKeepFirstSystemChatContextShiftStrategy.js.map +0 -1
  590. package/dist/llamaEvaluator/LlamaChatSession/LlamaChatSession.js.map +0 -1
  591. package/dist/llamaEvaluator/LlamaChatSession/utils/defineChatSessionFunction.js.map +0 -1
  592. package/dist/llamaEvaluator/LlamaContext/LlamaContext.js.map +0 -1
  593. package/dist/llamaEvaluator/LlamaContext/types.d.ts +0 -86
  594. package/dist/llamaEvaluator/LlamaContext/types.js.map +0 -1
  595. package/dist/llamaEvaluator/LlamaContext/utils/batchItemsPrioritizingStrategies/firstInFirstOutStrategy.js.map +0 -1
  596. package/dist/llamaEvaluator/LlamaContext/utils/batchItemsPrioritizingStrategies/maximumParallelismStrategy.js.map +0 -1
  597. package/dist/llamaEvaluator/LlamaContext/utils/resolveBatchItemsPrioritizingStrategy.d.ts +0 -2
  598. package/dist/llamaEvaluator/LlamaContext/utils/resolveBatchItemsPrioritizingStrategy.js.map +0 -1
  599. package/dist/llamaEvaluator/LlamaGrammar.js.map +0 -1
  600. package/dist/llamaEvaluator/LlamaGrammarEvaluationState.js.map +0 -1
  601. package/dist/llamaEvaluator/LlamaJsonSchemaGrammar.js.map +0 -1
  602. package/dist/llamaEvaluator/LlamaModel.d.ts +0 -119
  603. package/dist/llamaEvaluator/LlamaModel.js +0 -322
  604. package/dist/llamaEvaluator/LlamaModel.js.map +0 -1
  605. package/dist/utils/binariesGithubRelease.js.map +0 -1
  606. package/dist/utils/clearLlamaBuild.d.ts +0 -1
  607. package/dist/utils/clearLlamaBuild.js +0 -12
  608. package/dist/utils/clearLlamaBuild.js.map +0 -1
  609. package/dist/utils/cloneLlamaCppRepo.d.ts +0 -2
  610. package/dist/utils/cloneLlamaCppRepo.js +0 -102
  611. package/dist/utils/cloneLlamaCppRepo.js.map +0 -1
  612. package/dist/utils/compileLLamaCpp.d.ts +0 -8
  613. package/dist/utils/compileLLamaCpp.js +0 -132
  614. package/dist/utils/compileLLamaCpp.js.map +0 -1
  615. package/dist/utils/getBin.js +0 -78
  616. package/dist/utils/getBin.js.map +0 -1
  617. package/dist/utils/getGbnfGrammarForGbnfJsonSchema.d.ts +0 -2
  618. package/dist/utils/getGbnfGrammarForGbnfJsonSchema.js +0 -9
  619. package/dist/utils/getGbnfGrammarForGbnfJsonSchema.js.map +0 -1
  620. package/dist/utils/getReleaseInfo.d.ts +0 -7
  621. package/dist/utils/getReleaseInfo.js +0 -30
  622. package/dist/utils/getReleaseInfo.js.map +0 -1
  623. package/dist/utils/parseModelTypeDescription.d.ts +0 -6
  624. package/dist/utils/parseModelTypeDescription.js +0 -9
  625. package/dist/utils/parseModelTypeDescription.js.map +0 -1
  626. package/dist/utils/resolveChatWrapper.d.ts +0 -4
  627. package/dist/utils/resolveChatWrapper.js +0 -16
  628. package/dist/utils/resolveChatWrapper.js.map +0 -1
  629. package/dist/utils/usedBinFlag.d.ts +0 -6
  630. package/dist/utils/usedBinFlag.js +0 -15
  631. package/dist/utils/usedBinFlag.js.map +0 -1
  632. package/llama/usedBin.json +0 -3
  633. package/llamaBins/mac-arm64/llama-addon.node +0 -0
  634. /package/dist/{utils → bindings/utils}/binariesGithubRelease.d.ts +0 -0
  635. /package/dist/{llamaEvaluator → evaluator}/LlamaChat/utils/contextShiftStrategies/eraseFirstResponseAndKeepFirstSystemChatContextShiftStrategy.d.ts +0 -0
  636. /package/dist/{llamaEvaluator → evaluator}/LlamaContext/types.js +0 -0
  637. /package/dist/{llamaEvaluator/LlamaContext/utils/batchItemsPrioritizingStrategies → evaluator/LlamaContext/utils/batchItemsPrioritizationStrategies}/firstInFirstOutStrategy.d.ts +0 -0
  638. /package/dist/{llamaEvaluator/LlamaContext/utils/batchItemsPrioritizingStrategies → evaluator/LlamaContext/utils/batchItemsPrioritizationStrategies}/firstInFirstOutStrategy.js +0 -0
  639. /package/dist/{llamaEvaluator/LlamaContext/utils/batchItemsPrioritizingStrategies → evaluator/LlamaContext/utils/batchItemsPrioritizationStrategies}/maximumParallelismStrategy.d.ts +0 -0
  640. /package/dist/{llamaEvaluator/LlamaContext/utils/batchItemsPrioritizingStrategies → evaluator/LlamaContext/utils/batchItemsPrioritizationStrategies}/maximumParallelismStrategy.js +0 -0
  641. /package/dist/{llamaEvaluator → evaluator}/LlamaGrammarEvaluationState.d.ts +0 -0
@@ -0,0 +1,33 @@
1
+ import { tokenizeInput } from "../utils/tokenizeInput.js";
2
+ export class TokenBias {
3
+ /** @internal */ _model;
4
+ /** @internal */ _biases = new Map();
5
+ constructor(model) {
6
+ this._model = model;
7
+ }
8
+ /**
9
+ * Adjust the bias of the given token(s).
10
+ * If a text is provided, the bias will be applied to each individual token in the text.
11
+ * Setting a bias to `"never"` will prevent the token from being generated, unless it is required to comply with a grammar.
12
+ * Setting the bias of the EOS or EOT tokens to `"never"` has no effect and will be ignored.
13
+ * @param input - The token(s) to apply the bias to
14
+ * @param bias - The bias to apply to the token(s).
15
+ * Setting to a positive number increases the probability of the token(s) being generated.
16
+ * Setting to a negative number decreases the probability of the token(s) being generated.
17
+ * Setting to `0` has no effect.
18
+ * Setting to `"never"` will prevent the token from being generated, unless it is required to comply with a grammar.
19
+ * Try to play around with values between `10` and `-10` to see what works for your use case.
20
+ * Fractional values are allowed and can be used to fine-tune the bias (for example, `1.123`).
21
+ */
22
+ set(input, bias) {
23
+ for (const token of tokenizeInput(input, this._model.tokenizer))
24
+ this._biases.set(token, bias === "never" ? -Infinity : bias);
25
+ for (const token of tokenizeInput(input, this._model.tokenizer, "trimLeadingSpace"))
26
+ this._biases.set(token, bias === "never" ? -Infinity : bias);
27
+ return this;
28
+ }
29
+ static for(model) {
30
+ return new TokenBias(model);
31
+ }
32
+ }
33
+ //# sourceMappingURL=TokenBias.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"TokenBias.js","sourceRoot":"","sources":["../../src/evaluator/TokenBias.ts"],"names":[],"mappings":"AAEA,OAAO,EAAC,aAAa,EAAC,MAAM,2BAA2B,CAAC;AAGxD,MAAM,OAAO,SAAS;IAClB,gBAAgB,CAAiB,MAAM,CAAa;IACpD,gBAAgB,CAAiB,OAAO,GAAG,IAAI,GAAG,EAAiB,CAAC;IAEpE,YAAmB,KAAiB;QAChC,IAAI,CAAC,MAAM,GAAG,KAAK,CAAC;IACxB,CAAC;IAED;;;;;;;;;;;;;OAaG;IACI,GAAG,CAAC,KAA2C,EAAE,IAAsB;QAC1E,KAAK,MAAM,KAAK,IAAI,aAAa,CAAC,KAAK,EAAE,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC;YAC3D,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,KAAK,EAAE,IAAI,KAAK,OAAO,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QAEjE,KAAK,MAAM,KAAK,IAAI,aAAa,CAAC,KAAK,EAAE,IAAI,CAAC,MAAM,CAAC,SAAS,EAAE,kBAAkB,CAAC;YAC/E,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,KAAK,EAAE,IAAI,KAAK,OAAO,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QAEjE,OAAO,IAAI,CAAC;IAChB,CAAC;IAEM,MAAM,CAAC,GAAG,CAAC,KAAiB;QAC/B,OAAO,IAAI,SAAS,CAAC,KAAK,CAAC,CAAC;IAChC,CAAC;CACJ"}
@@ -0,0 +1,54 @@
1
+ /**
2
+ * Tracks the evaluation usage of tokens.
3
+ */
4
+ export declare class TokenMeter {
5
+ private _inputTokens;
6
+ private _outputTokens;
7
+ private _restoreStateTokens;
8
+ /**
9
+ * The number of input tokens used
10
+ */
11
+ get usedInputTokens(): number;
12
+ /**
13
+ * The number of tokens generated by a model
14
+ */
15
+ get usedOutputTokens(): number;
16
+ /**
17
+ * The number of tokens used as input to restore a context sequence state to continue previous evaluation.
18
+ * This may be consumed by virtual context sequences.
19
+ */
20
+ get usedRestoreStateTokens(): number;
21
+ /**
22
+ * Get the current state of the token meter
23
+ */
24
+ getState(): TokenMeterState;
25
+ /**
26
+ * Log the usage of tokens
27
+ */
28
+ useTokens(tokens: number, type: "input" | "output" | "restoreState"): void;
29
+ /**
30
+ * Get the difference between the current meter and another meter
31
+ */
32
+ diff(meter: TokenMeter | TokenMeterState): {
33
+ usedInputTokens: number;
34
+ usedOutputTokens: number;
35
+ usedRestoreStateTokens: number;
36
+ };
37
+ /**
38
+ * Log the usage of tokens on multiple meters
39
+ */
40
+ static useTokens(meters: null | undefined | TokenMeter | readonly TokenMeter[] | ReadonlySet<TokenMeter>, tokens: number, type: "input" | "output" | "restoreState"): void;
41
+ /**
42
+ * Get the difference between two meters
43
+ */
44
+ static diff(meter1: TokenMeter | TokenMeterState, meter2: TokenMeter | TokenMeterState): {
45
+ usedInputTokens: number;
46
+ usedOutputTokens: number;
47
+ usedRestoreStateTokens: number;
48
+ };
49
+ }
50
+ export type TokenMeterState = {
51
+ usedInputTokens: number;
52
+ usedOutputTokens: number;
53
+ usedRestoreStateTokens: number;
54
+ };
@@ -0,0 +1,86 @@
1
+ /**
2
+ * Tracks the evaluation usage of tokens.
3
+ */
4
+ export class TokenMeter {
5
+ _inputTokens = 0;
6
+ _outputTokens = 0;
7
+ _restoreStateTokens = 0;
8
+ /**
9
+ * The number of input tokens used
10
+ */
11
+ get usedInputTokens() {
12
+ return this._inputTokens;
13
+ }
14
+ /**
15
+ * The number of tokens generated by a model
16
+ */
17
+ get usedOutputTokens() {
18
+ return this._outputTokens;
19
+ }
20
+ /**
21
+ * The number of tokens used as input to restore a context sequence state to continue previous evaluation.
22
+ * This may be consumed by virtual context sequences.
23
+ */
24
+ get usedRestoreStateTokens() {
25
+ return this._restoreStateTokens;
26
+ }
27
+ /**
28
+ * Get the current state of the token meter
29
+ */
30
+ getState() {
31
+ return {
32
+ usedInputTokens: this.usedInputTokens,
33
+ usedOutputTokens: this.usedOutputTokens,
34
+ usedRestoreStateTokens: this.usedRestoreStateTokens
35
+ };
36
+ }
37
+ /**
38
+ * Log the usage of tokens
39
+ */
40
+ useTokens(tokens, type) {
41
+ if (tokens < 0)
42
+ throw new RangeError("Tokens cannot be negative");
43
+ else if (tokens === 0)
44
+ return;
45
+ if (type === "input")
46
+ this._inputTokens += tokens;
47
+ else if (type === "output")
48
+ this._outputTokens += tokens;
49
+ else if (type === "restoreState")
50
+ this._restoreStateTokens += tokens;
51
+ else {
52
+ void (type);
53
+ throw new TypeError(`Unknown token type: ${type}`);
54
+ }
55
+ }
56
+ /**
57
+ * Get the difference between the current meter and another meter
58
+ */
59
+ diff(meter) {
60
+ return TokenMeter.diff(this, meter);
61
+ }
62
+ /**
63
+ * Log the usage of tokens on multiple meters
64
+ */
65
+ static useTokens(meters, tokens, type) {
66
+ if (meters == null)
67
+ return;
68
+ if (meters instanceof TokenMeter)
69
+ meters.useTokens(tokens, type);
70
+ else {
71
+ for (const meter of meters)
72
+ meter.useTokens(tokens, type);
73
+ }
74
+ }
75
+ /**
76
+ * Get the difference between two meters
77
+ */
78
+ static diff(meter1, meter2) {
79
+ return {
80
+ usedInputTokens: meter1.usedInputTokens - meter2.usedInputTokens,
81
+ usedOutputTokens: meter1.usedOutputTokens - meter2.usedOutputTokens,
82
+ usedRestoreStateTokens: meter1.usedRestoreStateTokens - meter2.usedRestoreStateTokens
83
+ };
84
+ }
85
+ }
86
+ //# sourceMappingURL=TokenMeter.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"TokenMeter.js","sourceRoot":"","sources":["../../src/evaluator/TokenMeter.ts"],"names":[],"mappings":"AAAA;;GAEG;AACH,MAAM,OAAO,UAAU;IACX,YAAY,GAAW,CAAC,CAAC;IACzB,aAAa,GAAW,CAAC,CAAC;IAC1B,mBAAmB,GAAW,CAAC,CAAC;IAExC;;OAEG;IACH,IAAW,eAAe;QACtB,OAAO,IAAI,CAAC,YAAY,CAAC;IAC7B,CAAC;IAED;;OAEG;IACH,IAAW,gBAAgB;QACvB,OAAO,IAAI,CAAC,aAAa,CAAC;IAC9B,CAAC;IAED;;;OAGG;IACH,IAAW,sBAAsB;QAC7B,OAAO,IAAI,CAAC,mBAAmB,CAAC;IACpC,CAAC;IAED;;OAEG;IACI,QAAQ;QACX,OAAO;YACH,eAAe,EAAE,IAAI,CAAC,eAAe;YACrC,gBAAgB,EAAE,IAAI,CAAC,gBAAgB;YACvC,sBAAsB,EAAE,IAAI,CAAC,sBAAsB;SACtD,CAAC;IACN,CAAC;IAED;;OAEG;IACI,SAAS,CAAC,MAAc,EAAE,IAAyC;QACtE,IAAI,MAAM,GAAG,CAAC;YACV,MAAM,IAAI,UAAU,CAAC,2BAA2B,CAAC,CAAC;aACjD,IAAI,MAAM,KAAK,CAAC;YACjB,OAAO;QAEX,IAAI,IAAI,KAAK,OAAO;YAChB,IAAI,CAAC,YAAY,IAAI,MAAM,CAAC;aAC3B,IAAI,IAAI,KAAK,QAAQ;YACtB,IAAI,CAAC,aAAa,IAAI,MAAM,CAAC;aAC5B,IAAI,IAAI,KAAK,cAAc;YAC5B,IAAI,CAAC,mBAAmB,IAAI,MAAM,CAAC;aAClC;YACD,KAAK,CAAC,IAAoB,CAAC,CAAC;YAC5B,MAAM,IAAI,SAAS,CAAC,uBAAuB,IAAI,EAAE,CAAC,CAAC;SACtD;IACL,CAAC;IAED;;OAEG;IACI,IAAI,CAAC,KAAmC;QAC3C,OAAO,UAAU,CAAC,IAAI,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC;IACxC,CAAC;IAED;;OAEG;IACI,MAAM,CAAC,SAAS,CACnB,MAAuF,EACvF,MAAc,EACd,IAAyC;QAEzC,IAAI,MAAM,IAAI,IAAI;YACd,OAAO;QAEX,IAAI,MAAM,YAAY,UAAU;YAC5B,MAAM,CAAC,SAAS,CAAC,MAAM,EAAE,IAAI,CAAC,CAAC;aAC9B;YACD,KAAK,MAAM,KAAK,IAAI,MAAM;gBACtB,KAAK,CAAC,SAAS,CAAC,MAAM,EAAE,IAAI,CAAC,CAAC;SACrC;IACL,CAAC;IAED;;OAEG;IACI,MAAM,CAAC,IAAI,CACd,MAAoC,EACpC,MAAoC;QAEpC,OAAO;YACH,eAAe,EAAE,MAAM,CAAC,eAAe,GAAG,MAAM,CAAC,eAAe;YAChE,gBAAgB,EAAE,MAAM,CAAC,gBAAgB,GAAG,MAAM,CAAC,gBAAgB;YACnE,sBAAsB,EAAE,MAAM,CAAC,sBAAsB,GAAG,MAAM,CAAC,sBAAsB;SACxF,CAAC;IACN,CAAC;CACJ"}
@@ -0,0 +1,3 @@
1
+ import retry from "async-retry";
2
+ export declare const ggufDefaultFetchRetryOptions: retry.Options;
3
+ export declare const defaultExtraAllocationSize: number;
@@ -0,0 +1,8 @@
1
+ export const ggufDefaultFetchRetryOptions = {
2
+ retries: 10,
3
+ factor: 2,
4
+ minTimeout: 1000,
5
+ maxTimeout: 1000 * 16
6
+ };
7
+ export const defaultExtraAllocationSize = 1024 * 1024 * 1.5; // 1.5MB
8
+ //# sourceMappingURL=consts.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"consts.js","sourceRoot":"","sources":["../../src/gguf/consts.ts"],"names":[],"mappings":"AAEA,MAAM,CAAC,MAAM,4BAA4B,GAAkB;IACvD,OAAO,EAAE,EAAE;IACX,MAAM,EAAE,CAAC;IACT,UAAU,EAAE,IAAI;IAChB,UAAU,EAAE,IAAI,GAAG,EAAE;CACf,CAAC;AAEX,MAAM,CAAC,MAAM,0BAA0B,GAAG,IAAI,GAAG,IAAI,GAAG,GAAG,CAAC,CAAC,QAAQ"}
@@ -0,0 +1,3 @@
1
+ export declare class InvalidGgufMagicError extends Error {
2
+ constructor(expectedGgufMagic: string, actualGgufMagic: string);
3
+ }
@@ -0,0 +1,6 @@
1
+ export class InvalidGgufMagicError extends Error {
2
+ constructor(expectedGgufMagic, actualGgufMagic) {
3
+ super(`Invalid GGUF magic. Expected "${expectedGgufMagic}" but got "${actualGgufMagic}".`);
4
+ }
5
+ }
6
+ //# sourceMappingURL=InvalidGgufMagicError.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"InvalidGgufMagicError.js","sourceRoot":"","sources":["../../../src/gguf/errors/InvalidGgufMagicError.ts"],"names":[],"mappings":"AAAA,MAAM,OAAO,qBAAsB,SAAQ,KAAK;IAC5C,YAAmB,iBAAyB,EAAE,eAAuB;QACjE,KAAK,CAAC,iCAAiC,iBAAiB,cAAc,eAAe,IAAI,CAAC,CAAC;IAC/F,CAAC;CACJ"}
@@ -0,0 +1,4 @@
1
+ export declare class UnsupportedGgufValueTypeError extends Error {
2
+ readonly ggufValueType: number;
3
+ constructor(ggufValueType: number);
4
+ }
@@ -0,0 +1,9 @@
1
+ export class UnsupportedGgufValueTypeError extends Error {
2
+ ggufValueType;
3
+ constructor(ggufValueType) {
4
+ super(`Unsupported GGUF value type "${ggufValueType}"`);
5
+ Object.defineProperty(this, "ggufValueType", { enumerable: false });
6
+ this.ggufValueType = ggufValueType;
7
+ }
8
+ }
9
+ //# sourceMappingURL=UnsupportedGgufValueTypeError.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"UnsupportedGgufValueTypeError.js","sourceRoot":"","sources":["../../../src/gguf/errors/UnsupportedGgufValueTypeError.ts"],"names":[],"mappings":"AAAA,MAAM,OAAO,6BAA8B,SAAQ,KAAK;IACpC,aAAa,CAAS;IAEtC,YAAmB,aAAqB;QACpC,KAAK,CAAC,gCAAgC,aAAa,GAAG,CAAC,CAAC;QAExD,MAAM,CAAC,cAAc,CAAC,IAAI,EAAE,eAAoC,EAAE,EAAC,UAAU,EAAE,KAAK,EAAC,CAAC,CAAC;QAEvF,IAAI,CAAC,aAAa,GAAG,aAAa,CAAC;IACvC,CAAC;CACJ"}
@@ -0,0 +1,33 @@
1
+ /// <reference types="node" />
2
+ import { GgufReadOffset } from "../utils/GgufReadOffset.js";
3
+ export declare const valueTypeToBytesToRead: {
4
+ readonly uint8: 1;
5
+ readonly uint16: 2;
6
+ readonly uint32: 4;
7
+ readonly uint64: 8;
8
+ readonly int8: 1;
9
+ readonly int16: 2;
10
+ readonly int32: 4;
11
+ readonly int64: 8;
12
+ readonly float32: 4;
13
+ readonly float64: 8;
14
+ readonly bool: 1;
15
+ };
16
+ export declare abstract class GgufFileReader {
17
+ protected _buffer: Buffer;
18
+ abstract readByteRange(offset: number | GgufReadOffset, length: number): Promise<Buffer>;
19
+ readUint8(offset: number | GgufReadOffset): Promise<number>;
20
+ readUint16(offset: number | GgufReadOffset): Promise<number>;
21
+ readUint32(offset: number | GgufReadOffset): Promise<number>;
22
+ readUint64(offset: number | GgufReadOffset): Promise<bigint>;
23
+ readInt8(offset: number | GgufReadOffset): Promise<number>;
24
+ readInt16(offset: number | GgufReadOffset): Promise<number>;
25
+ readInt32(offset: number | GgufReadOffset): Promise<number>;
26
+ readInt64(offset: number | GgufReadOffset): Promise<bigint>;
27
+ readFloat32(offset: number | GgufReadOffset): Promise<number>;
28
+ readFloat64(offset: number | GgufReadOffset): Promise<number>;
29
+ readBool(offset: number | GgufReadOffset): Promise<boolean>;
30
+ protected _addToBuffer(buffer: Buffer): void;
31
+ private _readByteRangeAndUpdateOffset;
32
+ static castNumberIfSafe(value: bigint): number | bigint;
33
+ }
@@ -0,0 +1,76 @@
1
+ import { GgufReadOffset } from "../utils/GgufReadOffset.js";
2
+ export const valueTypeToBytesToRead = {
3
+ uint8: 1,
4
+ uint16: 2,
5
+ uint32: 4,
6
+ uint64: 8,
7
+ int8: 1,
8
+ int16: 2,
9
+ int32: 4,
10
+ int64: 8,
11
+ float32: 4,
12
+ float64: 8,
13
+ bool: 1
14
+ };
15
+ export class GgufFileReader {
16
+ _buffer = Buffer.alloc(0);
17
+ async readUint8(offset) {
18
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.uint8);
19
+ return response.readUInt8();
20
+ }
21
+ async readUint16(offset) {
22
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.uint16);
23
+ return response.readUInt16LE();
24
+ }
25
+ async readUint32(offset) {
26
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.uint32);
27
+ return response.readUInt32LE();
28
+ }
29
+ async readUint64(offset) {
30
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.uint64);
31
+ return response.readBigUInt64LE();
32
+ }
33
+ async readInt8(offset) {
34
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.int8);
35
+ return response.readInt8();
36
+ }
37
+ async readInt16(offset) {
38
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.int16);
39
+ return response.readInt16LE();
40
+ }
41
+ async readInt32(offset) {
42
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.int32);
43
+ return response.readInt32LE();
44
+ }
45
+ async readInt64(offset) {
46
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.int64);
47
+ return response.readBigInt64LE();
48
+ }
49
+ async readFloat32(offset) {
50
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.float32);
51
+ return response.readFloatLE();
52
+ }
53
+ async readFloat64(offset) {
54
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.float64);
55
+ return response.readDoubleLE();
56
+ }
57
+ async readBool(offset) {
58
+ const response = await this._readByteRangeAndUpdateOffset(offset, valueTypeToBytesToRead.uint8);
59
+ return response.readUInt8() === 1;
60
+ }
61
+ _addToBuffer(buffer) {
62
+ this._buffer = Buffer.concat([this._buffer, buffer]);
63
+ }
64
+ async _readByteRangeAndUpdateOffset(offset, length) {
65
+ const readOffset = GgufReadOffset.resolveReadOffset(offset);
66
+ const response = await this.readByteRange(readOffset.offset, length);
67
+ readOffset.moveBy(length);
68
+ return response;
69
+ }
70
+ static castNumberIfSafe(value) {
71
+ if (value > Number.MAX_SAFE_INTEGER)
72
+ return value;
73
+ return Number(value);
74
+ }
75
+ }
76
+ //# sourceMappingURL=GgufFileReader.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"GgufFileReader.js","sourceRoot":"","sources":["../../../src/gguf/fileReaders/GgufFileReader.ts"],"names":[],"mappings":"AAAA,OAAO,EAAC,cAAc,EAAC,MAAM,4BAA4B,CAAC;AAE1D,MAAM,CAAC,MAAM,sBAAsB,GAAG;IAClC,KAAK,EAAE,CAAC;IACR,MAAM,EAAE,CAAC;IACT,MAAM,EAAE,CAAC;IACT,MAAM,EAAE,CAAC;IACT,IAAI,EAAE,CAAC;IACP,KAAK,EAAE,CAAC;IACR,KAAK,EAAE,CAAC;IACR,KAAK,EAAE,CAAC;IACR,OAAO,EAAE,CAAC;IACV,OAAO,EAAE,CAAC;IACV,IAAI,EAAE,CAAC;CACD,CAAC;AAEX,MAAM,OAAgB,cAAc;IACtB,OAAO,GAAG,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;IAI7B,KAAK,CAAC,SAAS,CAAC,MAA+B;QAClD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,KAAK,CAAC,CAAC;QAChG,OAAO,QAAQ,CAAC,SAAS,EAAE,CAAC;IAChC,CAAC;IAEM,KAAK,CAAC,UAAU,CAAC,MAA+B;QACnD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,MAAM,CAAC,CAAC;QACjG,OAAO,QAAQ,CAAC,YAAY,EAAE,CAAC;IACnC,CAAC;IAEM,KAAK,CAAC,UAAU,CAAC,MAA+B;QACnD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,MAAM,CAAC,CAAC;QACjG,OAAO,QAAQ,CAAC,YAAY,EAAE,CAAC;IACnC,CAAC;IAEM,KAAK,CAAC,UAAU,CAAC,MAA+B;QACnD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,MAAM,CAAC,CAAC;QACjG,OAAO,QAAQ,CAAC,eAAe,EAAE,CAAC;IACtC,CAAC;IAEM,KAAK,CAAC,QAAQ,CAAC,MAA+B;QACjD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,IAAI,CAAC,CAAC;QAC/F,OAAO,QAAQ,CAAC,QAAQ,EAAE,CAAC;IAC/B,CAAC;IAEM,KAAK,CAAC,SAAS,CAAC,MAA+B;QAClD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,KAAK,CAAC,CAAC;QAChG,OAAO,QAAQ,CAAC,WAAW,EAAE,CAAC;IAClC,CAAC;IAEM,KAAK,CAAC,SAAS,CAAC,MAA+B;QAClD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,KAAK,CAAC,CAAC;QAChG,OAAO,QAAQ,CAAC,WAAW,EAAE,CAAC;IAClC,CAAC;IAEM,KAAK,CAAC,SAAS,CAAC,MAA+B;QAClD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,KAAK,CAAC,CAAC;QAChG,OAAO,QAAQ,CAAC,cAAc,EAAE,CAAC;IACrC,CAAC;IAEM,KAAK,CAAC,WAAW,CAAC,MAA+B;QACpD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,OAAO,CAAC,CAAC;QAClG,OAAO,QAAQ,CAAC,WAAW,EAAE,CAAC;IAClC,CAAC;IAEM,KAAK,CAAC,WAAW,CAAC,MAA+B;QACpD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,OAAO,CAAC,CAAC;QAClG,OAAO,QAAQ,CAAC,YAAY,EAAE,CAAC;IACnC,CAAC;IAEM,KAAK,CAAC,QAAQ,CAAC,MAA+B;QACjD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,6BAA6B,CAAC,MAAM,EAAE,sBAAsB,CAAC,KAAK,CAAC,CAAC;QAChG,OAAO,QAAQ,CAAC,SAAS,EAAE,KAAK,CAAC,CAAC;IACtC,CAAC;IAES,YAAY,CAAC,MAAc;QACjC,IAAI,CAAC,OAAO,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC,CAAC;IACzD,CAAC;IAEO,KAAK,CAAC,6BAA6B,CAAC,MAA+B,EAAE,MAAc;QACvF,MAAM,UAAU,GAAG,cAAc,CAAC,iBAAiB,CAAC,MAAM,CAAC,CAAC;QAE5D,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,aAAa,CAAC,UAAU,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;QACrE,UAAU,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;QAE1B,OAAO,QAAQ,CAAC;IACpB,CAAC;IAEM,MAAM,CAAC,gBAAgB,CAAC,KAAa;QACxC,IAAI,KAAK,GAAG,MAAM,CAAC,gBAAgB;YAC/B,OAAO,KAAK,CAAC;QAEjB,OAAO,MAAM,CAAC,KAAK,CAAC,CAAC;IACzB,CAAC;CACJ"}
@@ -0,0 +1,17 @@
1
+ /// <reference types="node" />
2
+ /// <reference types="node" />
3
+ import { GgufReadOffset } from "../utils/GgufReadOffset.js";
4
+ import { GgufFileReader } from "./GgufFileReader.js";
5
+ type GgufFsFileReaderOptions = {
6
+ filePath: string;
7
+ signal?: AbortSignal;
8
+ };
9
+ export declare class GgufFsFileReader extends GgufFileReader {
10
+ readonly filePath: string;
11
+ private readonly _signal?;
12
+ constructor({ filePath, signal }: GgufFsFileReaderOptions);
13
+ readByteRange(offset: number | GgufReadOffset, length: number): Promise<Buffer>;
14
+ private _readToExpandBufferUpToOffset;
15
+ private _readByteRange;
16
+ }
17
+ export {};
@@ -0,0 +1,45 @@
1
+ import fs from "node:fs/promises";
2
+ import { withLock } from "lifecycle-utils";
3
+ import { GgufReadOffset } from "../utils/GgufReadOffset.js";
4
+ import { defaultExtraAllocationSize } from "../consts.js";
5
+ import { GgufFileReader } from "./GgufFileReader.js";
6
+ export class GgufFsFileReader extends GgufFileReader {
7
+ filePath;
8
+ _signal;
9
+ constructor({ filePath, signal }) {
10
+ super();
11
+ this.filePath = filePath;
12
+ this._signal = signal;
13
+ }
14
+ async readByteRange(offset, length) {
15
+ const readOffset = GgufReadOffset.resolveReadOffset(offset);
16
+ const endOffset = readOffset.offset + length;
17
+ if (endOffset >= this._buffer.length)
18
+ await this._readToExpandBufferUpToOffset(endOffset);
19
+ const res = this._buffer.subarray(readOffset.offset, endOffset);
20
+ readOffset.moveBy(length);
21
+ return res;
22
+ }
23
+ async _readToExpandBufferUpToOffset(endOffset, extraAllocationSize = defaultExtraAllocationSize) {
24
+ return await withLock(this, "modifyBuffer", this._signal, async () => {
25
+ if (endOffset < this._buffer.length)
26
+ return;
27
+ const missingBytesBuffer = await this._readByteRange(this._buffer.length, endOffset + extraAllocationSize - this._buffer.length);
28
+ this._addToBuffer(missingBytesBuffer);
29
+ });
30
+ }
31
+ async _readByteRange(start, length) {
32
+ const fd = await fs.open(this.filePath, "r");
33
+ try {
34
+ if (this._signal?.aborted)
35
+ throw this._signal.reason;
36
+ const buffer = Buffer.alloc(length);
37
+ await fd.read(buffer, 0, length, start);
38
+ return buffer;
39
+ }
40
+ finally {
41
+ await fd.close();
42
+ }
43
+ }
44
+ }
45
+ //# sourceMappingURL=GgufFsFileReader.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"GgufFsFileReader.js","sourceRoot":"","sources":["../../../src/gguf/fileReaders/GgufFsFileReader.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,kBAAkB,CAAC;AAClC,OAAO,EAAC,QAAQ,EAAC,MAAM,iBAAiB,CAAC;AACzC,OAAO,EAAC,cAAc,EAAC,MAAM,4BAA4B,CAAC;AAC1D,OAAO,EAAC,0BAA0B,EAAC,MAAM,cAAc,CAAC;AACxD,OAAO,EAAC,cAAc,EAAC,MAAM,qBAAqB,CAAC;AAOnD,MAAM,OAAO,gBAAiB,SAAQ,cAAc;IAChC,QAAQ,CAAS;IAChB,OAAO,CAAe;IAEvC,YAAmB,EAAC,QAAQ,EAAE,MAAM,EAA0B;QAC1D,KAAK,EAAE,CAAC;QACR,IAAI,CAAC,QAAQ,GAAG,QAAQ,CAAC;QACzB,IAAI,CAAC,OAAO,GAAG,MAAM,CAAC;IAC1B,CAAC;IAEM,KAAK,CAAC,aAAa,CAAC,MAA+B,EAAE,MAAc;QACtE,MAAM,UAAU,GAAG,cAAc,CAAC,iBAAiB,CAAC,MAAM,CAAC,CAAC;QAC5D,MAAM,SAAS,GAAG,UAAU,CAAC,MAAM,GAAG,MAAM,CAAC;QAE7C,IAAI,SAAS,IAAI,IAAI,CAAC,OAAO,CAAC,MAAM;YAChC,MAAM,IAAI,CAAC,6BAA6B,CAAC,SAAS,CAAC,CAAC;QAExD,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,UAAU,CAAC,MAAM,EAAE,SAAS,CAAC,CAAC;QAChE,UAAU,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;QAC1B,OAAO,GAAG,CAAC;IACf,CAAC;IAEO,KAAK,CAAC,6BAA6B,CAAC,SAAiB,EAAE,sBAA8B,0BAA0B;QACnH,OAAO,MAAM,QAAQ,CAAC,IAAI,EAAE,cAAc,EAAE,IAAI,CAAC,OAAO,EAAE,KAAK,IAAI,EAAE;YACjE,IAAI,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM;gBAC/B,OAAO;YAEX,MAAM,kBAAkB,GAAG,MAAM,IAAI,CAAC,cAAc,CAChD,IAAI,CAAC,OAAO,CAAC,MAAM,EACnB,SAAS,GAAG,mBAAmB,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,CACxD,CAAC;YAEF,IAAI,CAAC,YAAY,CAAC,kBAAkB,CAAC,CAAC;QAC1C,CAAC,CAAC,CAAC;IACP,CAAC;IAEO,KAAK,CAAC,cAAc,CAAC,KAAa,EAAE,MAAc;QACtD,MAAM,EAAE,GAAG,MAAM,EAAE,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;QAC7C,IAAI;YACA,IAAI,IAAI,CAAC,OAAO,EAAE,OAAO;gBACrB,MAAM,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC;YAE9B,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC;YACpC,MAAM,EAAE,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE,MAAM,EAAE,KAAK,CAAC,CAAC;YACxC,OAAO,MAAM,CAAC;SACjB;gBAAS;YACN,MAAM,EAAE,CAAC,KAAK,EAAE,CAAC;SACpB;IACL,CAAC;CACJ"}
@@ -0,0 +1,22 @@
1
+ /// <reference types="node" />
2
+ /// <reference types="node" />
3
+ import retry from "async-retry";
4
+ import { GgufReadOffset } from "../utils/GgufReadOffset.js";
5
+ import { GgufFileReader } from "./GgufFileReader.js";
6
+ type GgufFetchFileReaderOptions = {
7
+ url: string;
8
+ retryOptions?: retry.Options;
9
+ headers?: Record<string, string>;
10
+ signal?: AbortSignal;
11
+ };
12
+ export declare class GgufNetworkFetchFileReader extends GgufFileReader {
13
+ readonly url: string;
14
+ readonly retryOptions: retry.Options;
15
+ readonly headers: Record<string, string>;
16
+ private readonly _signal?;
17
+ constructor({ url, retryOptions, headers, signal }: GgufFetchFileReaderOptions);
18
+ readByteRange(offset: number | GgufReadOffset, length: number): Promise<Buffer>;
19
+ private _fetchToExpandBufferUpToOffset;
20
+ private _fetchByteRange;
21
+ }
22
+ export {};
@@ -0,0 +1,63 @@
1
+ import retry from "async-retry";
2
+ import { withLock } from "lifecycle-utils";
3
+ import { GgufReadOffset } from "../utils/GgufReadOffset.js";
4
+ import { defaultExtraAllocationSize, ggufDefaultFetchRetryOptions } from "../consts.js";
5
+ import { GgufFileReader } from "./GgufFileReader.js";
6
+ export class GgufNetworkFetchFileReader extends GgufFileReader {
7
+ url;
8
+ retryOptions;
9
+ headers;
10
+ _signal;
11
+ constructor({ url, retryOptions = ggufDefaultFetchRetryOptions, headers, signal }) {
12
+ super();
13
+ this.url = url;
14
+ this.retryOptions = retryOptions;
15
+ this.headers = headers ?? {};
16
+ this._signal = signal;
17
+ }
18
+ async readByteRange(offset, length) {
19
+ const readOffset = GgufReadOffset.resolveReadOffset(offset);
20
+ const endOffset = readOffset.offset + length;
21
+ if (endOffset >= this._buffer.length)
22
+ await this._fetchToExpandBufferUpToOffset(endOffset);
23
+ const res = this._buffer.subarray(readOffset.offset, endOffset);
24
+ readOffset.moveBy(length);
25
+ return res;
26
+ }
27
+ async _fetchToExpandBufferUpToOffset(endOffset, extraAllocationSize = defaultExtraAllocationSize) {
28
+ await withLock(this, "modifyBuffer", this._signal, async () => {
29
+ if (endOffset < this._buffer.length)
30
+ return;
31
+ const missingBytesBuffer = await retry(async (bail) => {
32
+ try {
33
+ return await this._fetchByteRange(this._buffer.length, endOffset + extraAllocationSize - this._buffer.length);
34
+ }
35
+ catch (err) {
36
+ if (this._signal?.aborted) {
37
+ bail(this._signal.reason);
38
+ throw this._signal.reason;
39
+ }
40
+ throw err;
41
+ }
42
+ }, this.retryOptions);
43
+ if (this._signal?.aborted)
44
+ throw this._signal.reason;
45
+ this._addToBuffer(missingBytesBuffer);
46
+ });
47
+ }
48
+ async _fetchByteRange(start, length) {
49
+ const response = await fetch(this.url, {
50
+ headers: {
51
+ ...this.headers,
52
+ Range: `bytes=${start}-${start + length}`,
53
+ accept: "*/*"
54
+ },
55
+ signal: this._signal
56
+ });
57
+ if (!response.ok)
58
+ throw new Error(`Failed to fetch byte range: ${response.status} ${response.statusText}`);
59
+ const arrayBuffer = await response.arrayBuffer();
60
+ return Buffer.from(arrayBuffer);
61
+ }
62
+ }
63
+ //# sourceMappingURL=GgufNetworkFetchFileReader.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"GgufNetworkFetchFileReader.js","sourceRoot":"","sources":["../../../src/gguf/fileReaders/GgufNetworkFetchFileReader.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,MAAM,aAAa,CAAC;AAChC,OAAO,EAAC,QAAQ,EAAC,MAAM,iBAAiB,CAAC;AACzC,OAAO,EAAC,cAAc,EAAC,MAAM,4BAA4B,CAAC;AAC1D,OAAO,EAAC,0BAA0B,EAAE,4BAA4B,EAAC,MAAM,cAAc,CAAC;AACtF,OAAO,EAAC,cAAc,EAAC,MAAM,qBAAqB,CAAC;AASnD,MAAM,OAAO,0BAA2B,SAAQ,cAAc;IAC1C,GAAG,CAAS;IACZ,YAAY,CAAgB;IAC5B,OAAO,CAAyB;IAC/B,OAAO,CAAe;IAEvC,YAAmB,EAAC,GAAG,EAAE,YAAY,GAAG,4BAA4B,EAAE,OAAO,EAAE,MAAM,EAA6B;QAC9G,KAAK,EAAE,CAAC;QACR,IAAI,CAAC,GAAG,GAAG,GAAG,CAAC;QACf,IAAI,CAAC,YAAY,GAAG,YAAY,CAAC;QACjC,IAAI,CAAC,OAAO,GAAG,OAAO,IAAI,EAAE,CAAC;QAC7B,IAAI,CAAC,OAAO,GAAG,MAAM,CAAC;IAC1B,CAAC;IAEM,KAAK,CAAC,aAAa,CAAC,MAA+B,EAAE,MAAc;QACtE,MAAM,UAAU,GAAG,cAAc,CAAC,iBAAiB,CAAC,MAAM,CAAC,CAAC;QAC5D,MAAM,SAAS,GAAG,UAAU,CAAC,MAAM,GAAG,MAAM,CAAC;QAE7C,IAAI,SAAS,IAAI,IAAI,CAAC,OAAO,CAAC,MAAM;YAChC,MAAM,IAAI,CAAC,8BAA8B,CAAC,SAAS,CAAC,CAAC;QAEzD,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,UAAU,CAAC,MAAM,EAAE,SAAS,CAAC,CAAC;QAChE,UAAU,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;QAC1B,OAAO,GAAG,CAAC;IACf,CAAC;IAEO,KAAK,CAAC,8BAA8B,CAAC,SAAiB,EAAE,sBAA8B,0BAA0B;QACpH,MAAM,QAAQ,CAAC,IAAI,EAAE,cAAc,EAAE,IAAI,CAAC,OAAO,EAAE,KAAK,IAAI,EAAE;YAC1D,IAAI,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM;gBAC/B,OAAO;YAEX,MAAM,kBAAkB,GAAG,MAAM,KAAK,CAAC,KAAK,EAAE,IAAI,EAAE,EAAE;gBAClD,IAAI;oBACA,OAAO,MAAM,IAAI,CAAC,eAAe,CAAC,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,SAAS,GAAG,mBAAmB,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;iBACjH;gBAAC,OAAO,GAAG,EAAE;oBACV,IAAI,IAAI,CAAC,OAAO,EAAE,OAAO,EAAE;wBACvB,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;wBAC1B,MAAM,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC;qBAC7B;oBAED,MAAM,GAAG,CAAC;iBACb;YACL,CAAC,EAAE,IAAI,CAAC,YAAY,CAAC,CAAC;YAEtB,IAAI,IAAI,CAAC,OAAO,EAAE,OAAO;gBACrB,MAAM,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC;YAE9B,IAAI,CAAC,YAAY,CAAC,kBAAkB,CAAC,CAAC;QAC1C,CAAC,CAAC,CAAC;IACP,CAAC;IAEO,KAAK,CAAC,eAAe,CAAC,KAAa,EAAE,MAAc;QACvD,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,IAAI,CAAC,GAAG,EAAE;YACnC,OAAO,EAAE;gBACL,GAAG,IAAI,CAAC,OAAO;gBACf,KAAK,EAAE,SAAS,KAAK,IAAI,KAAK,GAAG,MAAM,EAAE;gBACzC,MAAM,EAAE,KAAK;aAChB;YACD,MAAM,EAAE,IAAI,CAAC,OAAO;SACvB,CAAC,CAAC;QAEH,IAAI,CAAC,QAAQ,CAAC,EAAE;YACZ,MAAM,IAAI,KAAK,CAAC,+BAA+B,QAAQ,CAAC,MAAM,IAAI,QAAQ,CAAC,UAAU,EAAE,CAAC,CAAC;QAE7F,MAAM,WAAW,GAAG,MAAM,QAAQ,CAAC,WAAW,EAAE,CAAC;QACjD,OAAO,MAAM,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC;IACpC,CAAC;CACJ"}
@@ -0,0 +1,42 @@
1
+ import { Llama } from "../../bindings/Llama.js";
2
+ import { GgufFileInfo } from "../types/GgufFileInfoTypes.js";
3
+ import { GgufInsightsConfigurationResolver } from "./GgufInsightsConfigurationResolver.js";
4
+ export type GgufInsightsResourceRequirements = {
5
+ cpuRam: number;
6
+ gpuVram: number;
7
+ };
8
+ export declare class GgufInsights {
9
+ private constructor();
10
+ get ggufFileInfo(): GgufFileInfo;
11
+ get configurationResolver(): GgufInsightsConfigurationResolver;
12
+ /** The context size the model was trained on */
13
+ get trainContextSize(): number | undefined;
14
+ /** The size of an embedding vector the model can produce */
15
+ get embeddingVectorSize(): number | undefined;
16
+ get totalLayers(): number;
17
+ get modelSize(): number;
18
+ estimateModelResourceRequirements({ gpuLayers }: {
19
+ gpuLayers: number;
20
+ }): GgufInsightsResourceRequirements;
21
+ /**
22
+ * Estimates the memory required to create a context of the given parameters based on the implementation details of `llama.cpp`.
23
+ * The calculation doesn't include a precise estimation of the graph overhead memory, so it uses a rough estimate for that.
24
+ * The estimation for the graph overhead memory will be improved in the future to be more precise, but it's good enough for now.
25
+ */
26
+ estimateContextResourceRequirements({ contextSize, modelGpuLayers, batchSize, sequences, isEmbeddingContext, includeGraphOverhead }: {
27
+ contextSize: number;
28
+ modelGpuLayers: number;
29
+ batchSize?: number;
30
+ sequences?: number;
31
+ isEmbeddingContext?: boolean;
32
+ includeGraphOverhead?: boolean;
33
+ }): GgufInsightsResourceRequirements;
34
+ /**
35
+ * @param ggufFileInfo
36
+ * @param llama - If you already have a `Llama` instance, pass it to reuse it for the `GgufInsights` instance.
37
+ * If you don't pass a `Llama` instance, a basic `Llama` instance is created as a fallback - it's a slim instance that
38
+ * doesn't instantiate a `llama.cpp` backend, so it won't utilize the GPU at all, and be shared with other `GgufInsights` instances
39
+ * that need a fallback `Llama` instance.
40
+ */
41
+ static from(ggufFileInfo: GgufFileInfo, llama?: Llama): Promise<GgufInsights>;
42
+ }