@elizaos/plugin-local-ai 1.0.0-alpha.63 → 1.0.0-alpha.65

package/dist/index.js

@@ -3,10 +3,7 @@ import fs5 from "node:fs";
 import path5 from "node:path";
 import { Readable as Readable2 } from "node:stream";
 import { fileURLToPath } from "node:url";
-import {
-  ModelType as ModelType3,
-  logger as logger10
-} from "@elizaos/core";
+import { ModelType as ModelType3, logger as logger10 } from "@elizaos/core";
 import { EmbeddingModel, FlagEmbedding } from "fastembed";
 import {
   LlamaChatSession,
@@ -45,9 +42,7 @@ function validateModelConfig(config) {
     logger.info("Setting USE_LOCAL_AI to true as it's required");
   }
   if (config.USE_STUDIOLM_TEXT_MODELS && config.USE_OLLAMA_TEXT_MODELS) {
-    throw new Error(
-      "StudioLM and Ollama text models cannot be enabled simultaneously"
-    );
+    throw new Error("StudioLM and Ollama text models cannot be enabled simultaneously");
   }
   logger.info("Configuration is valid");
 }
@@ -94,24 +89,24 @@ ${errorMessages}`);
 // src/types.ts
 var MODEL_SPECS = {
   small: {
-    name: "DeepSeek-R1-Distill-Qwen-1.5B-Q8_0.gguf",
-    repo: "unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF",
-    size: "1.5B",
-    quantization: "Q8_0",
+    name: "DeepHermes-3-Llama-3-3B-Preview-q4.gguf",
+    repo: "NousResearch/DeepHermes-3-Llama-3-3B-Preview-GGUF",
+    size: "3B",
+    quantization: "Q4_0",
     contextSize: 8192,
     tokenizer: {
-      name: "deepseek-ai/deepseek-llm-7b-base",
+      name: "NousResearch/DeepHermes-3-Llama-3-3B-Preview",
       type: "llama"
     }
   },
   medium: {
-    name: "DeepSeek-R1-Distill-Qwen-7B-Q8_0.gguf",
-    repo: "bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF",
-    size: "7B",
-    quantization: "Q8_0",
+    name: "DeepHermes-3-Llama-3-8B-q4.gguf",
+    repo: "NousResearch/DeepHermes-3-Llama-3-8B-Preview-GGUF",
+    size: "8B",
+    quantization: "Q4_0",
     contextSize: 8192,
     tokenizer: {
-      name: "deepseek-ai/deepseek-llm-7b-base",
+      name: "NousResearch/DeepHermes-3-Llama-3-8B-Preview",
       type: "llama"
     }
   },
@@ -159,12 +154,7 @@ var MODEL_SPECS = {
       quantization: "Q8_0",
       speakers: ["male_1", "male_2", "female_1", "female_2"],
       languages: ["en"],
-      features: [
-        "MULTI_SPEAKER",
-        "VOICE_CLONING",
-        "EMOTION_CONTROL",
-        "SPEED_CONTROL"
-      ],
+      features: ["MULTI_SPEAKER", "VOICE_CLONING", "EMOTION_CONTROL", "SPEED_CONTROL"],
       maxInputLength: 4096,
       sampleRate: 24e3,
       contextSize: 2048,
@@ -178,14 +168,7 @@ var MODEL_SPECS = {
       repo: "OuteAI/OuteTTS-0.3-1B-GGUF",
       size: "1B",
       quantization: "Q8_0",
-      speakers: [
-        "male_1",
-        "male_2",
-        "male_3",
-        "female_1",
-        "female_2",
-        "female_3"
-      ],
+      speakers: ["male_1", "male_2", "male_3", "female_1", "female_2", "female_3"],
       languages: ["en", "es", "fr", "de", "it"],
       features: [
         "MULTI_SPEAKER",
@@ -218,20 +201,7 @@ var MODEL_SPECS = {
         "female_3",
         "female_4"
       ],
-      languages: [
-        "en",
-        "es",
-        "fr",
-        "de",
-        "it",
-        "pt",
-        "nl",
-        "pl",
-        "ru",
-        "ja",
-        "ko",
-        "zh"
-      ],
+      languages: ["en", "es", "fr", "de", "it", "pt", "nl", "pl", "ru", "ja", "ko", "zh"],
       features: [
         "MULTI_SPEAKER",
         "VOICE_CLONING",
@@ -355,10 +325,7 @@ var DownloadManager = class _DownloadManager {
             reject(new Error(`Failed to download: ${response.statusCode}`));
             return;
           }
-          const totalSize = Number.parseInt(
-            response.headers["content-length"] || "0",
-            10
-          );
+          const totalSize = Number.parseInt(response.headers["content-length"] || "0", 10);
           let downloadedSize = 0;
           let lastLoggedPercent = 0;
           const barLength = 30;
@@ -369,13 +336,9 @@ var DownloadManager = class _DownloadManager {
             downloadedSize += chunk.length;
             const percent = Math.round(downloadedSize / totalSize * 100);
             if (percent >= lastLoggedPercent + 5) {
-              const filledLength = Math.floor(
-                downloadedSize / totalSize * barLength
-              );
+              const filledLength = Math.floor(downloadedSize / totalSize * barLength);
               const progressBar = "\u25B0".repeat(filledLength) + "\u25B1".repeat(barLength - filledLength);
-              logger2.info(
-                `Downloading ${fileName}: ${progressBar} ${percent}%`
-              );
+              logger2.info(`Downloading ${fileName}: ${progressBar} ${percent}%`);
               lastLoggedPercent = percent;
             }
           });
@@ -390,24 +353,18 @@ var DownloadManager = class _DownloadManager {
                   fs.mkdirSync(destDir, { recursive: true });
                 }
                 if (!fs.existsSync(tempPath)) {
-                  reject(
-                    new Error(`Temporary file ${tempPath} does not exist`)
-                  );
+                  reject(new Error(`Temporary file ${tempPath} does not exist`));
                   return;
                 }
                 if (fs.existsSync(destPath)) {
                   try {
                     const backupPath = `${destPath}.bak`;
                     fs.renameSync(destPath, backupPath);
-                    logger2.info(
-                      `Created backup of existing file: ${backupPath}`
-                    );
+                    logger2.info(`Created backup of existing file: ${backupPath}`);
                     fs.renameSync(tempPath, destPath);
                     if (fs.existsSync(backupPath)) {
                       fs.unlinkSync(backupPath);
-                      logger2.info(
-                        `Removed backup file after successful update: ${backupPath}`
-                      );
+                      logger2.info(`Removed backup file after successful update: ${backupPath}`);
                     }
                   } catch (moveErr) {
                     logger2.error(
@@ -417,9 +374,7 @@ var DownloadManager = class _DownloadManager {
                     if (fs.existsSync(backupPath)) {
                       try {
                         fs.renameSync(backupPath, destPath);
-                        logger2.info(
-                          `Restored from backup after failed update: ${backupPath}`
-                        );
+                        logger2.info(`Restored from backup after failed update: ${backupPath}`);
                       } catch (restoreErr) {
                         logger2.error(
                           `Failed to restore from backup: ${restoreErr instanceof Error ? restoreErr.message : String(restoreErr)}`
@@ -441,9 +396,7 @@ var DownloadManager = class _DownloadManager {
                 } else {
                   fs.renameSync(tempPath, destPath);
                 }
-                logger2.success(
-                  `Download of ${fileName} completed successfully`
-                );
+                logger2.success(`Download of ${fileName} completed successfully`);
                 this.activeDownloads.delete(destPath);
                 resolve();
               } catch (err) {
@@ -465,9 +418,7 @@ var DownloadManager = class _DownloadManager {
             });
           });
           file.on("error", (err) => {
-            logger2.error(
-              `File write error: ${err instanceof Error ? err.message : String(err)}`
-            );
+            logger2.error(`File write error: ${err instanceof Error ? err.message : String(err)}`);
             file.close(() => {
               if (fs.existsSync(tempPath)) {
                 try {
@@ -485,9 +436,7 @@ var DownloadManager = class _DownloadManager {
         }
       );
       request.on("error", (err) => {
-        logger2.error(
-          `Request error: ${err instanceof Error ? err.message : String(err)}`
-        );
+        logger2.error(`Request error: ${err instanceof Error ? err.message : String(err)}`);
         if (fs.existsSync(tempPath)) {
           try {
             fs.unlinkSync(tempPath);
@@ -526,9 +475,7 @@ var DownloadManager = class _DownloadManager {
    */
   async downloadFile(url, destPath) {
     if (this.activeDownloads.has(destPath)) {
-      logger2.info(
-        `Download for ${destPath} already in progress, waiting for it to complete...`
-      );
+      logger2.info(`Download for ${destPath} already in progress, waiting for it to complete...`);
       const existingDownload = this.activeDownloads.get(destPath);
       if (existingDownload) {
         return existingDownload;
@@ -866,9 +813,7 @@ var OllamaManager = class _OllamaManager {
         timestamp: (/* @__PURE__ */ new Date()).toISOString()
       });
       if (!this.initialized && !isInitialized) {
-        throw new Error(
-          "Ollama not initialized. Please initialize before generating text."
-        );
+        throw new Error("Ollama not initialized. Please initialize before generating text.");
       }
       logger3.info("Ollama preparing request:", {
         model: params.modelType === ModelType.TEXT_LARGE ? this.configuredModels.medium : this.configuredModels.small,
@@ -1046,9 +991,7 @@ var PlatformManager = class _PlatformManager {
           isAppleSilicon: true
         };
       }
-      const { stdout: gpuInfo } = await execAsync(
-        "system_profiler SPDisplaysDataType"
-      );
+      const { stdout: gpuInfo } = await execAsync("system_profiler SPDisplaysDataType");
       return {
         name: gpuInfo.split("Chipset Model:")[1]?.split("\n")[0]?.trim() || "Unknown GPU",
         type: "metal",
@@ -1070,9 +1013,7 @@ var PlatformManager = class _PlatformManager {
    */
   async detectWindowsGPU() {
     try {
-      const { stdout } = await execAsync(
-        "wmic path win32_VideoController get name"
-      );
+      const { stdout } = await execAsync("wmic path win32_VideoController get name");
       const gpuName = stdout.split("\n")[1].trim();
       if (gpuName.toLowerCase().includes("nvidia")) {
         const { stdout: nvidiaInfo } = await execAsync(
@@ -1483,9 +1424,7 @@ var StudioLMManager = class _StudioLMManager {
         timestamp: (/* @__PURE__ */ new Date()).toISOString()
       });
       if (!this.initialized && !isInitialized) {
-        throw new Error(
-          "StudioLM not initialized. Please initialize before generating text."
-        );
+        throw new Error("StudioLM not initialized. Please initialize before generating text.");
       }
       const messages = [
         {
@@ -1557,9 +1496,7 @@ var StudioLMManager = class _StudioLMManager {
 
 // src/utils/tokenizerManager.ts
 import { logger as logger6 } from "@elizaos/core";
-import {
-  AutoTokenizer
-} from "@huggingface/transformers";
+import { AutoTokenizer } from "@huggingface/transformers";
 var TokenizerManager = class _TokenizerManager {
   static instance = null;
   tokenizers;
@@ -1609,18 +1546,13 @@ var TokenizerManager = class _TokenizerManager {
         logger6.info("Using cached tokenizer:", { key: tokenizerKey });
         const cachedTokenizer = this.tokenizers.get(tokenizerKey);
         if (!cachedTokenizer) {
-          throw new Error(
-            `Tokenizer ${tokenizerKey} exists in map but returned undefined`
-          );
+          throw new Error(`Tokenizer ${tokenizerKey} exists in map but returned undefined`);
         }
         return cachedTokenizer;
       }
       const fs6 = await import("node:fs");
       if (!fs6.existsSync(this.modelsDir)) {
-        logger6.warn(
-          "Models directory does not exist, creating it:",
-          this.modelsDir
-        );
+        logger6.warn("Models directory does not exist, creating it:", this.modelsDir);
         fs6.mkdirSync(this.modelsDir, { recursive: true });
       }
       logger6.info(
@@ -1628,13 +1560,10 @@ var TokenizerManager = class _TokenizerManager {
         this.modelsDir
       );
       try {
-        const tokenizer = await AutoTokenizer.from_pretrained(
-          modelConfig.tokenizer.name,
-          {
-            cache_dir: this.modelsDir,
-            local_files_only: false
-          }
-        );
+        const tokenizer = await AutoTokenizer.from_pretrained(modelConfig.tokenizer.name, {
+          cache_dir: this.modelsDir,
+          local_files_only: false
+        });
         this.tokenizers.set(tokenizerKey, tokenizer);
         logger6.success("Tokenizer loaded successfully:", { key: tokenizerKey });
         return tokenizer;
@@ -1646,13 +1575,10 @@ var TokenizerManager = class _TokenizerManager {
           modelsDir: this.modelsDir
         });
         logger6.info("Retrying tokenizer loading...");
-        const tokenizer = await AutoTokenizer.from_pretrained(
-          modelConfig.tokenizer.name,
-          {
-            cache_dir: this.modelsDir,
-            local_files_only: false
-          }
-        );
+        const tokenizer = await AutoTokenizer.from_pretrained(modelConfig.tokenizer.name, {
+          cache_dir: this.modelsDir,
+          local_files_only: false
+        });
         this.tokenizers.set(tokenizerKey, tokenizer);
         logger6.success("Tokenizer loaded successfully on retry:", {
           key: tokenizerKey
@@ -1871,9 +1797,7 @@ var TranscribeManager = class _TranscribeManager {
    */
   async checkFFmpegAvailability() {
     try {
-      const { stdout, stderr } = await execAsync2(
-        "which ffmpeg || where ffmpeg"
-      );
+      const { stdout, stderr } = await execAsync2("which ffmpeg || where ffmpeg");
       this.ffmpegPath = stdout.trim();
       this.ffmpegAvailable = true;
       logger7.info("FFmpeg found at:", {
@@ -1901,9 +1825,7 @@ var TranscribeManager = class _TranscribeManager {
       const { stdout } = await execAsync2("ffmpeg -codecs");
       const hasRequiredCodecs = stdout.includes("pcm_s16le") && stdout.includes("wav");
       if (!hasRequiredCodecs) {
-        throw new Error(
-          "FFmpeg installation missing required codecs (pcm_s16le, wav)"
-        );
+        throw new Error("FFmpeg installation missing required codecs (pcm_s16le, wav)");
       }
     } catch (error) {
       logger7.error("FFmpeg capabilities verification failed:", {
@@ -1917,20 +1839,17 @@ var TranscribeManager = class _TranscribeManager {
    * Logs instructions on how to install FFmpeg if it is not properly installed.
    */
   logFFmpegInstallInstructions() {
-    logger7.warn(
-      "FFmpeg is required but not properly installed. Please install FFmpeg:",
-      {
-        instructions: {
-          mac: "brew install ffmpeg",
-          ubuntu: "sudo apt-get install ffmpeg",
-          windows: "choco install ffmpeg",
-          manual: "Download from https://ffmpeg.org/download.html"
-        },
-        requiredVersion: "4.0 or later",
-        requiredCodecs: ["pcm_s16le", "wav"],
-        timestamp: (/* @__PURE__ */ new Date()).toISOString()
-      }
-    );
+    logger7.warn("FFmpeg is required but not properly installed. Please install FFmpeg:", {
+      instructions: {
+        mac: "brew install ffmpeg",
+        ubuntu: "sudo apt-get install ffmpeg",
+        windows: "choco install ffmpeg",
+        manual: "Download from https://ffmpeg.org/download.html"
+      },
+      requiredVersion: "4.0 or later",
+      requiredCodecs: ["pcm_s16le", "wav"],
+      timestamp: (/* @__PURE__ */ new Date()).toISOString()
+    });
   }
   /**
    * Gets the singleton instance of TranscribeManager, creates a new instance if it doesn't exist.
@@ -2008,15 +1927,10 @@ var TranscribeManager = class _TranscribeManager {
    */
   async preprocessAudio(audioBuffer) {
     if (!this.ffmpegAvailable) {
-      throw new Error(
-        "FFmpeg is not installed. Please install FFmpeg to use audio transcription."
-      );
+      throw new Error("FFmpeg is not installed. Please install FFmpeg to use audio transcription.");
     }
     try {
-      const tempInputFile = path2.join(
-        this.cacheDir,
-        `temp_input_${Date.now()}`
-      );
+      const tempInputFile = path2.join(this.cacheDir, `temp_input_${Date.now()}`);
       const tempWavFile = path2.join(this.cacheDir, `temp_${Date.now()}.wav`);
       fs2.writeFileSync(tempInputFile, audioBuffer);
       await this.convertToWav(tempInputFile, tempWavFile);
@@ -2126,12 +2040,7 @@ function getWavHeader(audioLength, sampleRate, channelCount = 1, bitsPerSample =
   return wavHeader;
 }
 function prependWavHeader(readable, audioLength, sampleRate, channelCount = 1, bitsPerSample = 16) {
-  const wavHeader = getWavHeader(
-    audioLength,
-    sampleRate,
-    channelCount,
-    bitsPerSample
-  );
+  const wavHeader = getWavHeader(audioLength, sampleRate, channelCount, bitsPerSample);
   let pushedHeader = false;
   const passThrough = new PassThrough();
   readable.on("data", (data) => {
@@ -2165,10 +2074,7 @@ var TTSManager = class _TTSManager {
   constructor(cacheDir) {
     this.cacheDir = path3.join(cacheDir, "tts");
     this.modelsDir = process.env.LLAMALOCAL_PATH?.trim() ? path3.resolve(process.env.LLAMALOCAL_PATH.trim()) : path3.join(process.cwd(), "models");
-    this.downloadManager = DownloadManager.getInstance(
-      this.cacheDir,
-      this.modelsDir
-    );
+    this.downloadManager = DownloadManager.getInstance(this.cacheDir, this.modelsDir);
     this.ensureCacheDirectory();
     logger8.info("TTSManager initialized");
   }
@@ -2244,10 +2150,7 @@ var TTSManager = class _TTSManager {
             await this.downloadManager.downloadFromUrl(attempt.url, modelPath);
             const completedBar = "\u25B0".repeat(barLength);
             logger8.info(`Downloading TTS model: ${completedBar} 100%`);
-            logger8.success(
-              "TTS model download successful with:",
-              attempt.description
-            );
+            logger8.success("TTS model download successful with:", attempt.description);
             break;
           } catch (error) {
             lastError = error;
@@ -2319,9 +2222,7 @@ var TTSManager = class _TTSManager {
           responseTokens.push(token);
           const percent = Math.round(responseTokens.length / maxTokens * 100);
           const barLength = 30;
-          const filledLength = Math.floor(
-            responseTokens.length / maxTokens * barLength
-          );
+          const filledLength = Math.floor(responseTokens.length / maxTokens * barLength);
           const progressBar = "\u25B0".repeat(filledLength) + "\u25B1".repeat(barLength - filledLength);
           logger8.info(
             `Token generation: ${progressBar} ${percent}% (${responseTokens.length}/${maxTokens})`
@@ -2340,9 +2241,7 @@ var TTSManager = class _TTSManager {
       }
       logger8.info("Converting tokens to audio data...");
       const audioData = this.processAudioResponse({
-        tokens: responseTokens.map(
-          (t) => Number.parseInt(this.model.detokenize([t]), 10)
-        )
+        tokens: responseTokens.map((t) => Number.parseInt(this.model.detokenize([t]), 10))
       });
       logger8.info("Audio data generated:", {
         byteLength: audioData.length,
@@ -2436,10 +2335,7 @@ var VisionManager = class _VisionManager {
     this.modelsDir = path4.join(path4.dirname(cacheDir), "models", "vision");
     this.cacheDir = cacheDir;
     this.ensureModelsDirExists();
-    this.downloadManager = DownloadManager.getInstance(
-      this.cacheDir,
-      this.modelsDir
-    );
+    this.downloadManager = DownloadManager.getInstance(this.cacheDir, this.modelsDir);
     this.platformConfig = this.getPlatformConfig();
     logger9.info("VisionManager initialized");
   }
@@ -2508,11 +2404,7 @@ var VisionManager = class _VisionManager {
    * @returns {boolean} - Returns true if cache exists, otherwise returns false.
    */
   checkCacheExists(modelId, type) {
-    const modelPath = path4.join(
-      this.modelsDir,
-      modelId.replace("/", "--"),
-      type
-    );
+    const modelPath = path4.join(this.modelsDir, modelId.replace("/", "--"), type);
     if (existsSync(modelPath)) {
       logger9.info(`${type} found at: ${modelPath}`);
       return true;
@@ -2550,9 +2442,7 @@ var VisionManager = class _VisionManager {
    * @returns {object} The model configuration object containing device, dtype, and cache_dir.
    */
   getModelConfig(componentName) {
-    const component = this.modelComponents.find(
-      (c) => c.name === componentName
-    );
+    const component = this.modelComponents.find((c) => c.name === componentName);
     return {
       device: this.platformConfig.device,
       dtype: component?.dtype || "fp32",
@@ -2568,9 +2458,7 @@ var VisionManager = class _VisionManager {
   async initialize() {
     try {
       if (this.initialized) {
-        logger9.info(
-          "Vision model already initialized, skipping initialization"
-        );
+        logger9.info("Vision model already initialized, skipping initialization");
         return;
       }
       logger9.info("Starting vision model initialization...");
@@ -2586,32 +2474,25 @@ var VisionManager = class _VisionManager {
       try {
         let lastProgress = -1;
         const modelCached = this.checkCacheExists(modelSpec.modelId, "model");
-        const model = await Florence2ForConditionalGeneration.from_pretrained(
-          modelSpec.modelId,
-          {
-            device: "cpu",
-            cache_dir: this.modelsDir,
-            local_files_only: modelCached,
-            revision: "main",
-            progress_callback: (progressInfo) => {
-              if (modelCached || this.modelDownloaded) return;
-              const progress = "progress" in progressInfo ? Math.max(0, Math.min(1, progressInfo.progress)) : 0;
-              const currentProgress = Math.round(progress * 100);
-              if (currentProgress > lastProgress + 9 || currentProgress === 100) {
-                lastProgress = currentProgress;
-                const barLength = 30;
-                const filledLength = Math.floor(
-                  currentProgress / 100 * barLength
-                );
-                const progressBar = "\u25B0".repeat(filledLength) + "\u25B1".repeat(barLength - filledLength);
-                logger9.info(
-                  `Downloading vision model: ${progressBar} ${currentProgress}%`
-                );
-                if (currentProgress === 100) this.modelDownloaded = true;
-              }
+        const model = await Florence2ForConditionalGeneration.from_pretrained(modelSpec.modelId, {
+          device: "cpu",
+          cache_dir: this.modelsDir,
+          local_files_only: modelCached,
+          revision: "main",
+          progress_callback: (progressInfo) => {
+            if (modelCached || this.modelDownloaded) return;
+            const progress = "progress" in progressInfo ? Math.max(0, Math.min(1, progressInfo.progress)) : 0;
+            const currentProgress = Math.round(progress * 100);
+            if (currentProgress > lastProgress + 9 || currentProgress === 100) {
+              lastProgress = currentProgress;
+              const barLength = 30;
+              const filledLength = Math.floor(currentProgress / 100 * barLength);
+              const progressBar = "\u25B0".repeat(filledLength) + "\u25B1".repeat(barLength - filledLength);
+              logger9.info(`Downloading vision model: ${progressBar} ${currentProgress}%`);
+              if (currentProgress === 100) this.modelDownloaded = true;
             }
           }
-        );
+        });
         this.model = model;
         logger9.success("Florence2 model loaded successfully");
       } catch (error) {
@@ -2624,35 +2505,25 @@ var VisionManager = class _VisionManager {
       }
       logger9.info("Loading vision tokenizer...");
       try {
-        const tokenizerCached = this.checkCacheExists(
-          modelSpec.modelId,
-          "tokenizer"
-        );
+        const tokenizerCached = this.checkCacheExists(modelSpec.modelId, "tokenizer");
         let tokenizerProgress = -1;
-        this.tokenizer = await AutoTokenizer2.from_pretrained(
-          modelSpec.modelId,
-          {
-            cache_dir: this.modelsDir,
-            local_files_only: tokenizerCached,
-            progress_callback: (progressInfo) => {
-              if (tokenizerCached || this.tokenizerDownloaded) return;
-              const progress = "progress" in progressInfo ? Math.max(0, Math.min(1, progressInfo.progress)) : 0;
-              const currentProgress = Math.round(progress * 100);
-              if (currentProgress !== tokenizerProgress) {
-                tokenizerProgress = currentProgress;
-                const barLength = 30;
-                const filledLength = Math.floor(
-                  currentProgress / 100 * barLength
-                );
-                const progressBar = "\u25B0".repeat(filledLength) + "\u25B1".repeat(barLength - filledLength);
-                logger9.info(
-                  `Downloading vision tokenizer: ${progressBar} ${currentProgress}%`
-                );
-                if (currentProgress === 100) this.tokenizerDownloaded = true;
-              }
+        this.tokenizer = await AutoTokenizer2.from_pretrained(modelSpec.modelId, {
+          cache_dir: this.modelsDir,
+          local_files_only: tokenizerCached,
+          progress_callback: (progressInfo) => {
+            if (tokenizerCached || this.tokenizerDownloaded) return;
+            const progress = "progress" in progressInfo ? Math.max(0, Math.min(1, progressInfo.progress)) : 0;
+            const currentProgress = Math.round(progress * 100);
+            if (currentProgress !== tokenizerProgress) {
+              tokenizerProgress = currentProgress;
+              const barLength = 30;
+              const filledLength = Math.floor(currentProgress / 100 * barLength);
+              const progressBar = "\u25B0".repeat(filledLength) + "\u25B1".repeat(barLength - filledLength);
+              logger9.info(`Downloading vision tokenizer: ${progressBar} ${currentProgress}%`);
+              if (currentProgress === 100) this.tokenizerDownloaded = true;
             }
           }
-        );
+        });
         logger9.success("Vision tokenizer loaded successfully");
       } catch (error) {
         logger9.error("Failed to load tokenizer:", {
@@ -2664,36 +2535,26 @@ var VisionManager = class _VisionManager {
       }
       logger9.info("Loading vision processor...");
       try {
-        const processorCached = this.checkCacheExists(
-          modelSpec.modelId,
-          "processor"
-        );
+        const processorCached = this.checkCacheExists(modelSpec.modelId, "processor");
         let processorProgress = -1;
-        this.processor = await AutoProcessor.from_pretrained(
-          modelSpec.modelId,
-          {
-            device: "cpu",
-            cache_dir: this.modelsDir,
-            local_files_only: processorCached,
-            progress_callback: (progressInfo) => {
-              if (processorCached || this.processorDownloaded) return;
-              const progress = "progress" in progressInfo ? Math.max(0, Math.min(1, progressInfo.progress)) : 0;
-              const currentProgress = Math.round(progress * 100);
-              if (currentProgress !== processorProgress) {
-                processorProgress = currentProgress;
-                const barLength = 30;
-                const filledLength = Math.floor(
-                  currentProgress / 100 * barLength
-                );
-                const progressBar = "\u25B0".repeat(filledLength) + "\u25B1".repeat(barLength - filledLength);
-                logger9.info(
-                  `Downloading vision processor: ${progressBar} ${currentProgress}%`
-                );
-                if (currentProgress === 100) this.processorDownloaded = true;
-              }
+        this.processor = await AutoProcessor.from_pretrained(modelSpec.modelId, {
+          device: "cpu",
+          cache_dir: this.modelsDir,
+          local_files_only: processorCached,
+          progress_callback: (progressInfo) => {
+            if (processorCached || this.processorDownloaded) return;
+            const progress = "progress" in progressInfo ? Math.max(0, Math.min(1, progressInfo.progress)) : 0;
+            const currentProgress = Math.round(progress * 100);
+            if (currentProgress !== processorProgress) {
+              processorProgress = currentProgress;
+              const barLength = 30;
+              const filledLength = Math.floor(currentProgress / 100 * barLength);
+              const progressBar = "\u25B0".repeat(filledLength) + "\u25B1".repeat(barLength - filledLength);
+              logger9.info(`Downloading vision processor: ${progressBar} ${currentProgress}%`);
+              if (currentProgress === 100) this.processorDownloaded = true;
             }
           }
-        );
+        });
         logger9.success("Vision processor loaded successfully");
       } catch (error) {
         logger9.error("Failed to load vision processor:", {
@@ -2927,14 +2788,8 @@ var LocalAIManager = class _LocalAIManager {
       }
       this.modelsDir = modelsDir;
     }
-    this.modelPath = path5.join(
-      this.modelsDir,
-      "DeepSeek-R1-Distill-Qwen-1.5B-Q8_0.gguf"
-    );
-    this.mediumModelPath = path5.join(
-      this.modelsDir,
-      "DeepSeek-R1-Distill-Qwen-7B-Q5_K_M.gguf"
-    );
+    this.modelPath = path5.join(this.modelsDir, "DeepHermes-3-Llama-3-3B-Preview-q4.gguf");
+    this.mediumModelPath = path5.join(this.modelsDir, "DeepHermes-3-Llama-3-8B-q4.gguf");
     const cacheDirEnv = process.env.CACHE_DIR?.trim();
     if (cacheDirEnv) {
       this.cacheDir = path5.resolve(cacheDirEnv);
@@ -2985,12 +2840,8 @@ var LocalAIManager = class _LocalAIManager {
       const validatedConfig = await validateConfig(config);
       logger10.info("Environment configuration validated");
       process.env.USE_LOCAL_AI = String(validatedConfig.USE_LOCAL_AI);
-      process.env.USE_STUDIOLM_TEXT_MODELS = String(
-        validatedConfig.USE_STUDIOLM_TEXT_MODELS
-      );
-      process.env.USE_OLLAMA_TEXT_MODELS = String(
-        validatedConfig.USE_OLLAMA_TEXT_MODELS
-      );
+      process.env.USE_STUDIOLM_TEXT_MODELS = String(validatedConfig.USE_STUDIOLM_TEXT_MODELS);
+      process.env.USE_OLLAMA_TEXT_MODELS = String(validatedConfig.USE_OLLAMA_TEXT_MODELS);
       logger10.success("Environment initialization complete");
     } catch (error) {
       logger10.error("Environment validation failed:", {
@@ -3014,9 +2865,7 @@ var LocalAIManager = class _LocalAIManager {
       }
       await this.ollamaManager.initialize();
       if (!this.ollamaManager.isInitialized()) {
-        throw new Error(
-          "Ollama initialization failed - models not properly loaded"
-        );
+        throw new Error("Ollama initialization failed - models not properly loaded");
       }
       logger10.success("Ollama initialization complete");
     } catch (error) {
@@ -3041,9 +2890,7 @@ var LocalAIManager = class _LocalAIManager {
       }
       await this.studioLMManager.initialize();
       if (!this.studioLMManager.isInitialized()) {
-        throw new Error(
-          "StudioLM initialization failed - models not properly loaded"
-        );
+        throw new Error("StudioLM initialization failed - models not properly loaded");
       }
       this.studioLMInitialized = true;
       logger10.success("StudioLM initialization complete");
@@ -3062,18 +2909,15 @@ var LocalAIManager = class _LocalAIManager {
    *
    * @returns A Promise that resolves to a boolean indicating whether the model download was successful.
    */
-  async downloadModel() {
+  async downloadModel(modelType) {
+    const modelSpec = modelType === ModelType3.TEXT_LARGE ? MODEL_SPECS.medium : MODEL_SPECS.small;
+    const modelPath = modelType === ModelType3.TEXT_LARGE ? this.mediumModelPath : this.modelPath;
     try {
-      const isLargeModel = this.modelPath === this.mediumModelPath;
-      const modelSpec = isLargeModel ? MODEL_SPECS.medium : MODEL_SPECS.small;
-      return await this.downloadManager.downloadModel(
-        modelSpec,
-        this.modelPath
-      );
+      return await this.downloadManager.downloadModel(modelSpec, modelPath);
     } catch (error) {
       logger10.error("Model download failed:", {
         error: error instanceof Error ? error.message : String(error),
-        modelPath: this.modelPath
+        modelPath
       });
       throw error;
     }
@@ -3121,16 +2965,11 @@ var LocalAIManager = class _LocalAIManager {
       logger10.info("Initializing embedding model...");
       logger10.info("Models directory:", this.modelsDir);
       if (!fs5.existsSync(this.modelsDir)) {
-        logger10.warn(
-          "Models directory does not exist, creating it:",
-          this.modelsDir
-        );
+        logger10.warn("Models directory does not exist, creating it:", this.modelsDir);
         fs5.mkdirSync(this.modelsDir, { recursive: true });
       }
       if (!this.embeddingModel) {
-        logger10.info(
-          "Creating new FlagEmbedding instance with BGESmallENV15 model"
-        );
+        logger10.info("Creating new FlagEmbedding instance with BGESmallENV15 model");
         const barLength = 30;
         const emptyBar = "\u25B1".repeat(barLength);
         logger10.info(`Downloading embedding model: ${emptyBar} 0%`);
@@ -3179,31 +3018,22 @@ var LocalAIManager = class _LocalAIManager {
           return this.generateText(params);
         }
         if (!this.studioLMManager) {
-          logger10.warn(
-            "StudioLM manager not initialized, falling back to local models"
-          );
+          logger10.warn("StudioLM manager not initialized, falling back to local models");
           return this.generateText(params);
         }
         if (!this.studioLMInitialized) {
           logger10.info("StudioLM not initialized, initializing now...");
           await this.initializeStudioLM();
         }
-        return await this.studioLMManager.generateText(
-          params,
-          this.studioLMInitialized
-        );
+        return await this.studioLMManager.generateText(params, this.studioLMInitialized);
       }
       if (modelConfig.source === "ollama") {
         if (process.env.USE_OLLAMA_TEXT_MODELS !== "true") {
-          logger10.warn(
-            "Ollama requested but disabled in environment, falling back to local models"
-          );
+          logger10.warn("Ollama requested but disabled in environment, falling back to local models");
           return this.generateText(params);
         }
         if (!this.ollamaManager) {
-          logger10.warn(
-            "Ollama manager not initialized, falling back to local models"
-          );
+          logger10.warn("Ollama manager not initialized, falling back to local models");
           return this.generateText(params);
         }
         if (!this.ollamaInitialized && !this.ollamaManager.isInitialized()) {
@@ -3211,10 +3041,7 @@ var LocalAIManager = class _LocalAIManager {
           await this.ollamaManager.initialize();
           this.ollamaInitialized = true;
         }
-        return await this.ollamaManager.generateText(
-          params,
-          this.ollamaInitialized
-        );
+        return await this.ollamaManager.generateText(params, this.ollamaInitialized);
       }
       return this.generateText(params);
     } catch (error) {
@@ -3270,10 +3097,7 @@ var LocalAIManager = class _LocalAIManager {
         runtime: !!params.runtime,
         stopSequences: params.stopSequences
       });
-      const tokens = await this.tokenizerManager.encode(
-        params.prompt,
-        this.activeModelConfig
-      );
+      const tokens = await this.tokenizerManager.encode(params.prompt, this.activeModelConfig);
       logger10.info("Input tokens:", { count: tokens.length });
       const systemMessage = "You are a helpful AI assistant. Respond to the current request only.";
       await this.chatSession.prompt(systemMessage, {
@@ -3439,7 +3263,7 @@ var LocalAIManager = class _LocalAIManager {
       this.smallModelInitializingPromise = (async () => {
         await this.initializeEnvironment();
         await this.checkPlatformCapabilities();
-        await this.downloadModel();
+        await this.downloadModel(ModelType3.TEXT_SMALL);
         try {
           this.llama = await getLlama2();
           const smallModel = await this.llama.loadModel({
@@ -3474,14 +3298,13 @@ var LocalAIManager = class _LocalAIManager {
         if (!this.llama) {
           await this.lazyInitSmallModel();
         }
+        await this.downloadModel(ModelType3.TEXT_LARGE);
         try {
-          const mediumModel = await this.llama.loadModel(
-            {
-              gpuLayers: 43,
-              modelPath: this.mediumModelPath,
-              vocabOnly: false
-            }
-          );
+          const mediumModel = await this.llama.loadModel({
+            gpuLayers: 43,
+            modelPath: this.mediumModelPath,
+            vocabOnly: false
+          });
           this.mediumModel = mediumModel;
           this.mediumModelInitialized = true;
           logger10.info("Medium model initialized successfully");
@@ -3674,7 +3497,6 @@ var localAIPlugin = {
       }
     },
     [ModelType3.TEXT_EMBEDDING]: async (_runtime, params) => {
-      console.log("params is", params)
       const text = params?.text;
       try {
         logger10.info("TEXT_EMBEDDING handler - Initial input:", {
@@ -3708,6 +3530,168 @@ var localAIPlugin = {
         return new Array(384).fill(0);
       }
     },
+    [ModelType3.OBJECT_SMALL]: async (runtime, params) => {
+      try {
+        logger10.info("OBJECT_SMALL handler - Processing request:", {
+          prompt: params.prompt,
+          hasSchema: !!params.schema,
+          temperature: params.temperature
+        });
+        let jsonPrompt = params.prompt;
+        if (!jsonPrompt.includes("```json") && !jsonPrompt.includes("respond with valid JSON")) {
+          jsonPrompt += "\nPlease respond with valid JSON only, without any explanations, markdown formatting, or additional text.";
+        }
+        const modelConfig = localAIManager.getTextModelSource();
+        let textResponse;
+        if (modelConfig.source !== "local") {
+          textResponse = await localAIManager.generateTextOllamaStudio({
+            prompt: jsonPrompt,
+            stopSequences: params.stopSequences,
+            runtime,
+            modelType: ModelType3.TEXT_SMALL
+          });
+        } else {
+          textResponse = await localAIManager.generateText({
+            prompt: jsonPrompt,
+            stopSequences: params.stopSequences,
+            runtime,
+            modelType: ModelType3.TEXT_SMALL
+          });
+        }
+        try {
+          const extractJSON = (text) => {
+            const jsonBlockRegex = /```(?:json)?\s*([\s\S]*?)\s*```/;
+            const match = text.match(jsonBlockRegex);
+            if (match && match[1]) {
+              return match[1].trim();
+            }
+            const jsonContentRegex = /\s*(\{[\s\S]*\})\s*$/;
+            const contentMatch = text.match(jsonContentRegex);
+            if (contentMatch && contentMatch[1]) {
+              return contentMatch[1].trim();
+            }
+            return text.trim();
+          };
+          const extractedJsonText = extractJSON(textResponse);
+          logger10.debug("Extracted JSON text:", extractedJsonText);
+          let jsonObject;
+          try {
+            jsonObject = JSON.parse(extractedJsonText);
+          } catch (parseError) {
+            logger10.debug("Initial JSON parse failed, attempting to fix common issues");
+            const fixedJson = extractedJsonText.replace(/:\s*"([^"]*)(?:\n)([^"]*)"/g, ': "$1\\n$2"').replace(/"([^"]*?)[^a-zA-Z0-9\s\.,;:\-_\(\)"'\[\]{}]([^"]*?)"/g, '"$1$2"').replace(/(\s*)(\w+)(\s*):/g, '$1"$2"$3:').replace(/,(\s*[\]}])/g, "$1");
+            try {
+              jsonObject = JSON.parse(fixedJson);
+            } catch (finalError) {
+              logger10.error("Failed to parse JSON after fixing:", finalError);
+              throw new Error("Invalid JSON returned from model");
+            }
+          }
+          if (params.schema) {
+            try {
+              for (const key of Object.keys(params.schema)) {
+                if (!(key in jsonObject)) {
+                  jsonObject[key] = null;
+                }
+              }
+            } catch (schemaError) {
+              logger10.error("Schema validation failed:", schemaError);
+            }
+          }
+          return jsonObject;
+        } catch (parseError) {
+          logger10.error("Failed to parse JSON:", parseError);
+          logger10.error("Raw response:", textResponse);
+          throw new Error("Invalid JSON returned from model");
+        }
+      } catch (error) {
+        logger10.error("Error in OBJECT_SMALL handler:", error);
+        throw error;
+      }
+    },
+    [ModelType3.OBJECT_LARGE]: async (runtime, params) => {
+      try {
+        logger10.info("OBJECT_LARGE handler - Processing request:", {
+          prompt: params.prompt,
+          hasSchema: !!params.schema,
+          temperature: params.temperature
+        });
+        let jsonPrompt = params.prompt;
+        if (!jsonPrompt.includes("```json") && !jsonPrompt.includes("respond with valid JSON")) {
+          jsonPrompt += "\nPlease respond with valid JSON only, without any explanations, markdown formatting, or additional text.";
+        }
+        const modelConfig = localAIManager.getTextModelSource();
+        let textResponse;
+        if (modelConfig.source !== "local") {
+          textResponse = await localAIManager.generateTextOllamaStudio({
+            prompt: jsonPrompt,
+            stopSequences: params.stopSequences,
+            runtime,
+            modelType: ModelType3.TEXT_LARGE
+          });
+        } else {
+          textResponse = await localAIManager.generateText({
+            prompt: jsonPrompt,
+            stopSequences: params.stopSequences,
+            runtime,
+            modelType: ModelType3.TEXT_LARGE
+          });
+        }
+        try {
+          const extractJSON = (text) => {
+            const jsonBlockRegex = /```(?:json)?\s*([\s\S]*?)\s*```/;
+            const match = text.match(jsonBlockRegex);
+            if (match && match[1]) {
+              return match[1].trim();
+            }
+            const jsonContentRegex = /\s*(\{[\s\S]*\})\s*$/;
+            const contentMatch = text.match(jsonContentRegex);
+            if (contentMatch && contentMatch[1]) {
+              return contentMatch[1].trim();
+            }
+            return text.trim();
+          };
+          const cleanupJSON = (jsonText) => {
+            return jsonText.replace(/\[DEBUG\].*?(\n|$)/g, "\n").replace(/\[LOG\].*?(\n|$)/g, "\n").replace(/console\.log.*?(\n|$)/g, "\n");
+          };
+          const extractedJsonText = extractJSON(textResponse);
+          const cleanedJsonText = cleanupJSON(extractedJsonText);
+          logger10.debug("Extracted JSON text:", cleanedJsonText);
+          let jsonObject;
+          try {
+            jsonObject = JSON.parse(cleanedJsonText);
+          } catch (parseError) {
+            logger10.debug("Initial JSON parse failed, attempting to fix common issues");
+            const fixedJson = cleanedJsonText.replace(/:\s*"([^"]*)(?:\n)([^"]*)"/g, ': "$1\\n$2"').replace(/"([^"]*?)[^a-zA-Z0-9\s\.,;:\-_\(\)"'\[\]{}]([^"]*?)"/g, '"$1$2"').replace(/(\s*)(\w+)(\s*):/g, '$1"$2"$3:').replace(/,(\s*[\]}])/g, "$1");
+            try {
+              jsonObject = JSON.parse(fixedJson);
+            } catch (finalError) {
+              logger10.error("Failed to parse JSON after fixing:", finalError);
+              throw new Error("Invalid JSON returned from model");
+            }
+          }
+          if (params.schema) {
+            try {
+              for (const key of Object.keys(params.schema)) {
+                if (!(key in jsonObject)) {
+                  jsonObject[key] = null;
+                }
+              }
+            } catch (schemaError) {
+              logger10.error("Schema validation failed:", schemaError);
+            }
+          }
+          return jsonObject;
+        } catch (parseError) {
+          logger10.error("Failed to parse JSON:", parseError);
+          logger10.error("Raw response:", textResponse);
+          throw new Error("Invalid JSON returned from model");
+        }
+      } catch (error) {
+        logger10.error("Error in OBJECT_LARGE handler:", error);
+        throw error;
+      }
+    },
     [ModelType3.TEXT_TOKENIZER_ENCODE]: async (_runtime, { text }) => {
       try {
         const manager = localAIManager.getTokenizerManager();
@@ -3833,16 +3817,10 @@ var localAIPlugin = {
           fn: async (runtime) => {
             try {
               logger10.info("Starting TEXT_EMBEDDING test");
-              const embedding = await runtime.useModel(
-                ModelType3.TEXT_EMBEDDING,
-                {
-                  text: "This is a test of the text embedding model."
-                }
-              );
-              logger10.info(
-                "Embedding generated with dimensions:",
-                embedding.length
-              );
+              const embedding = await runtime.useModel(ModelType3.TEXT_EMBEDDING, {
+                text: "This is a test of the text embedding model."
+              });
+              logger10.info("Embedding generated with dimensions:", embedding.length);
               if (!Array.isArray(embedding)) {
                 throw new Error("Embedding is not an array");
               }
@@ -3852,10 +3830,7 @@ var localAIPlugin = {
               if (embedding.some((val) => typeof val !== "number")) {
                 throw new Error("Embedding contains non-numeric values");
               }
-              const nullEmbedding = await runtime.useModel(
-                ModelType3.TEXT_EMBEDDING,
-                null
-              );
+              const nullEmbedding = await runtime.useModel(ModelType3.TEXT_EMBEDDING, null);
               if (!Array.isArray(nullEmbedding) || nullEmbedding.some((val) => val !== 0)) {
                 throw new Error("Null input did not return zero vector");
               }
@@ -3875,10 +3850,7 @@ var localAIPlugin = {
             try {
               logger10.info("Starting TEXT_TOKENIZER_ENCODE test");
               const text = "Hello tokenizer test!";
-              const tokens = await runtime.useModel(
-                ModelType3.TEXT_TOKENIZER_ENCODE,
-                { text }
-              );
+              const tokens = await runtime.useModel(ModelType3.TEXT_TOKENIZER_ENCODE, { text });
               logger10.info("Encoded tokens:", { count: tokens.length });
               if (!Array.isArray(tokens)) {
                 throw new Error("Tokens output is not an array");
@@ -3889,9 +3861,7 @@ var localAIPlugin = {
               if (tokens.some((token) => !Number.isInteger(token))) {
                 throw new Error("Tokens contain non-integer values");
               }
-              logger10.success(
-                "TEXT_TOKENIZER_ENCODE test completed successfully"
-              );
+              logger10.success("TEXT_TOKENIZER_ENCODE test completed successfully");
             } catch (error) {
               logger10.error("TEXT_TOKENIZER_ENCODE test failed:", {
                 error: error instanceof Error ? error.message : String(error),
@@ -3907,14 +3877,12 @@ var localAIPlugin = {
             try {
               logger10.info("Starting TEXT_TOKENIZER_DECODE test");
               const originalText = "Hello tokenizer test!";
-              const tokens = await runtime.useModel(
-                ModelType3.TEXT_TOKENIZER_ENCODE,
-                { text: originalText }
-              );
-              const decodedText = await runtime.useModel(
-                ModelType3.TEXT_TOKENIZER_DECODE,
-                { tokens }
-              );
+              const tokens = await runtime.useModel(ModelType3.TEXT_TOKENIZER_ENCODE, {
+                text: originalText
+              });
+              const decodedText = await runtime.useModel(ModelType3.TEXT_TOKENIZER_DECODE, {
+                tokens
+              });
               logger10.info("Round trip tokenization:", {
                 original: originalText,
                 decoded: decodedText
@@ -3922,9 +3890,7 @@ var localAIPlugin = {
               if (typeof decodedText !== "string") {
                 throw new Error("Decoded output is not a string");
               }
-              logger10.success(
-                "TEXT_TOKENIZER_DECODE test completed successfully"
-              );
+              logger10.success("TEXT_TOKENIZER_DECODE test completed successfully");
             } catch (error) {
               logger10.error("TEXT_TOKENIZER_DECODE test failed:", {
                 error: error instanceof Error ? error.message : String(error),
@@ -3940,10 +3906,7 @@ var localAIPlugin = {
             try {
               logger10.info("Starting IMAGE_DESCRIPTION test");
               const imageUrl = "https://raw.githubusercontent.com/microsoft/FLAML/main/website/static/img/flaml.png";
-              const result = await runtime.useModel(
-                ModelType3.IMAGE_DESCRIPTION,
-                imageUrl
-              );
+              const result = await runtime.useModel(ModelType3.IMAGE_DESCRIPTION, imageUrl);
               logger10.info("Image description result:", result);
               if (!result || typeof result !== "object") {
                 throw new Error("Invalid response format");
@@ -3992,10 +3955,7 @@ var localAIPlugin = {
                 // "fmt "
               ]);
               const audioBuffer = Buffer.from(audioData);
-              const transcription = await runtime.useModel(
-                ModelType3.TRANSCRIPTION,
-                audioBuffer
-              );
+              const transcription = await runtime.useModel(ModelType3.TRANSCRIPTION, audioBuffer);
               logger10.info("Transcription result:", transcription);
               if (typeof transcription !== "string") {
                 throw new Error("Transcription result is not a string");
@@ -4016,10 +3976,7 @@ var localAIPlugin = {
             try {
               logger10.info("Starting TEXT_TO_SPEECH test");
               const testText = "This is a test of the text to speech system.";
-              const audioStream = await runtime.useModel(
-                ModelType3.TEXT_TO_SPEECH,
-                testText
-              );
+              const audioStream = await runtime.useModel(ModelType3.TEXT_TO_SPEECH, testText);
               if (!(audioStream instanceof Readable2)) {
                 throw new Error("TTS output is not a readable stream");
               }