@qvac/sdk 0.12.2 → 0.13.0

package/dist/examples/kv-cache-example.js

@@ -3,7 +3,6 @@ try {
     // Load the model
     const modelId = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         modelConfig: {
             device: "gpu",
             ctx_size: 2048,
@@ -32,7 +31,10 @@ try {
     console.log("🔄 Continuing conversation (reusing previous turn cache):");
     const history2 = [
         { role: "user", content: "What is the capital of France?" },
-        { role: "assistant", content: final1.cacheableAssistantContent ?? final1.contentText },
+        {
+            role: "assistant",
+            content: final1.cacheableAssistantContent ?? final1.contentText,
+        },
         { role: "user", content: "What about Germany?" },
     ];
     // Auto-keyed caching should:

package/dist/examples/llamacpp-cache.js

@@ -36,7 +36,6 @@ try {
     console.log("Loading llama.cpp model with cache support...");
     modelId = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         modelConfig: {
             ctx_size: 4096,
             verbosity: VERBOSITY.ERROR,

package/dist/examples/llamacpp-dynamic-tools.js

@@ -105,7 +105,6 @@ async function chatTurn({ modelId, kvCache, history, tools }) {
 async function main() {
     const modelId = await loadModel({
         modelSrc: QWEN3_1_7B_INST_Q4,
-        modelType: "llm",
         modelConfig: {
             ctx_size: 4096,
             tools: true,
@@ -128,7 +127,10 @@ async function main() {
     await chatTurn({ modelId, kvCache, history, tools: weatherTools });
     // Turn 2 — same session, swap to horoscope tools. Dynamic mode lets the
     // model see a different tool set without invalidating the kv-cache.
-    history.push({ role: "user", content: "Now check my horoscope for Aquarius." });
+    history.push({
+        role: "user",
+        content: "Now check my horoscope for Aquarius.",
+    });
     console.log("\n\n🤖 Turn 2 (tools=horoscope):\n");
     await chatTurn({ modelId, kvCache, history, tools: horoscopeTools });
     // Turn 3 — swap to a parameterless tool to confirm empty-arg flows work.

package/dist/examples/llamacpp-filesystem.js

@@ -11,7 +11,7 @@ try {
     // Load model from provided file path
     const modelId = await loadModel({
         modelSrc: ggufPath,
-        modelType: "llm",
+        modelType: "llamacpp-completion",
         modelConfig: {
             ctx_size: 4096,
         },

package/dist/examples/llamacpp-http.js

@@ -7,7 +7,7 @@ try {
     // Load model from HTTP URL
     const modelId = await loadModel({
         modelSrc: httpUrl,
-        modelType: "llm",
+        modelType: "llamacpp-completion",
         modelConfig: {
             ctx_size: 4096,
         },

package/dist/examples/llamacpp-multi-gpu.js

@@ -1,4 +1,4 @@
-import { completion, loadModel, unloadModel, VERBOSITY, LLAMA_3_2_1B_INST_Q4_0 } from "@qvac/sdk";
+import { completion, loadModel, unloadModel, VERBOSITY, LLAMA_3_2_1B_INST_Q4_0, } from "@qvac/sdk";
 // Multi-GPU inference distributes a model across multiple GPUs using llama.cpp's
 // built-in split modes. Two strategies are available:
 //
@@ -16,7 +16,7 @@ const modelSrc = process.argv[2] ?? LLAMA_3_2_1B_INST_Q4_0;
 try {
     const modelId = await loadModel({
         modelSrc,
-        modelType: "llm",
+        modelType: "llamacpp-completion",
         modelConfig: {
             "split-mode": "layer",
             "tensor-split": "1,1",

package/dist/examples/llamacpp-multimodal.js

@@ -9,7 +9,6 @@ try {
     // Load the main model with projection in a single step
     const modelId = await loadModel({
         modelSrc: SMOLVLM2_500M_MULTIMODAL_Q8_0,
-        modelType: "llm",
         modelConfig: {
             ctx_size: 1024,
             projectionModelSrc: MMPROJ_SMOLVLM2_500M_MULTIMODAL_Q8_0,

package/dist/examples/llamacpp-p2p.js

@@ -10,7 +10,6 @@ try {
     // Then load it in memory from cache
     const modelId = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         modelConfig: {
             device: "gpu",
             ctx_size: 2048,

package/dist/examples/llamacpp-sharded.js

@@ -8,7 +8,7 @@ import { completion, loadModel, unloadModel, VERBOSITY } from "@qvac/sdk";
 try {
     const modelId = await loadModel({
         modelSrc: "https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct-GGUF/resolve/main/qwen2.5-coder-7b-instruct-q4_0-00001-of-00002.gguf",
-        modelType: "llm",
+        modelType: "llamacpp-completion",
         modelConfig: {
             device: "gpu",
             ctx_size: 2048,

package/dist/examples/llamacpp-structured-output.js

@@ -58,7 +58,6 @@ async function runToContent(run) {
 try {
     const modelId = await loadModel({
         modelSrc: QWEN3_600M_INST_Q4,
-        modelType: "llm",
         onProgress: (p) => process.stdout.write(`\rLoading: ${p.percentage.toFixed(1)}%`),
     });
     console.log(`\nModel loaded: ${modelId}\n`);

package/dist/examples/logging-file-transport.js

@@ -22,7 +22,6 @@ try {
     // Load model with integrated logging
     const modelId = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         modelConfig: {
             ctx_size: 2048,
             temp: 0.7,

package/dist/examples/logging-streaming.js

@@ -16,7 +16,6 @@ try {
     console.log("📥 Loading models (watch SDK logs above)...\n");
     const llmModelId = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         modelConfig: {
             ctx_size: 2048,
             temp: 0.7,
@@ -25,7 +24,6 @@ try {
     });
     const embedModelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
     });
     console.log("📡 Starting model-specific log streams...\n");
     (async () => {

package/dist/examples/mcp-websearch.js

@@ -66,7 +66,6 @@ try {
     console.log("2️⃣  Loading model...");
     const modelId = await loadModel({
         modelSrc: QWEN3_1_7B_INST_Q4,
-        modelType: "llm",
         modelConfig: {
             ctx_size: 4096,
             tools: true,

package/dist/examples/multi-model-demo.js

@@ -2,14 +2,12 @@ import { completion, LLAMA_3_2_1B_INST_Q4_0, loadModel, QWEN3_1_7B_INST_Q4, unlo
 try {
     const alice = await loadModel({
         modelSrc: QWEN3_1_7B_INST_Q4,
-        modelType: "llm",
         modelConfig: {
             ctx_size: 4096,
         },
     });
     const bob = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         modelConfig: {
             ctx_size: 4096,
         },

package/dist/examples/ocr-fasttext.js

@@ -18,7 +18,6 @@ try {
     console.log("🚀 Loading OCR model...");
     const modelId = await loadModel({
         modelSrc: OCR_LATIN_RECOGNIZER_1,
-        modelType: "ocr",
         modelConfig: {
             langList: ["en"],
             useGPU: true,

package/dist/examples/plugins.js

@@ -47,7 +47,6 @@ console.log("1. LLM Completion (llamacpp-completion plugin)");
 try {
     const llmModelId = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         modelConfig: { ctx_size: 2048 },
         onProgress: (p) => console.log(`   Loading LLM: ${p.percentage.toFixed(1)}%`),
     });
@@ -75,7 +74,6 @@ console.log("2. Translation (nmtcpp-translation plugin)");
 try {
     const nmtModelId = await loadModel({
         modelSrc: BERGAMOT_EN_ES,
-        modelType: "nmt",
         modelConfig: {
             engine: "Bergamot",
             from: "en",
@@ -88,7 +86,7 @@ try {
     const result = translate({
         modelId: nmtModelId,
         text,
-        modelType: "nmt",
+        modelType: "nmtcpp-translation",
         stream: false,
     });
     const translated = await result.text;
@@ -105,7 +103,6 @@ console.log("   Attempting to load an embeddings model...\n");
 try {
     const embedModelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
     });
     await embed({ modelId: embedModelId, text: "test" });
     console.error("   ❌ Unexpected: embed succeeded without the plugin!");

package/dist/examples/profiling/basic.js

@@ -8,7 +8,6 @@ try {
     console.log("Profiler enabled:", profiler.isEnabled());
     const modelId = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         onProgress: (p) => console.log(`  ${p.percentage.toFixed(1)}%`),
     });
     console.log("Model loaded:", modelId);
@@ -54,7 +53,9 @@ try {
     else {
         console.log("  (no loadModel handler event captured)");
         // Debug: show what ops are available
-        const ops = [...new Set(json.recentEvents?.map((e) => `${e.op}:${e.kind}`) ?? [])];
+        const ops = [
+            ...new Set(json.recentEvents?.map((e) => `${e.op}:${e.kind}`) ?? []),
+        ];
         console.log("  Available ops:", ops.join(", "));
     }
     console.log("\n=== Profiler JSON (structure) ===");

package/dist/examples/profiling/per-call.js

@@ -4,7 +4,6 @@ try {
     console.log("Profiler globally enabled:", profiler.isEnabled());
     const modelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
         onProgress: (p) => console.log(`  ${p.percentage.toFixed(1)}%`),
     });
     console.log("Model loaded:", modelId);

package/dist/examples/quickstart.js

@@ -3,7 +3,6 @@ try {
     // Load a model into memory
     const modelId = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         onProgress: (progress) => {
             console.log(progress);
         },

package/dist/examples/rag/rag-chromadb.js

@@ -38,7 +38,6 @@ try {
     const client = await initializeChromaClient();
     const modelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
         onProgress: (p) => console.log(`Loading model... ${Math.round(p.percentage)}%`),
     });
     // Sample corpus

package/dist/examples/rag/rag-hyperdb/cancellation.js

@@ -20,7 +20,6 @@ try {
     console.log("Loading embedding model...");
     const modelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
         onProgress: (progress) => {
             process.stdout.write(`\r   ${progress.percentage.toFixed(1)}%`);
         },

package/dist/examples/rag/rag-hyperdb/ingest.js

@@ -6,7 +6,6 @@ try {
     console.log(`🔍 Query: "${query}"`);
     const modelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
         onProgress: (progress) => {
             console.log(`Loading model... ${progress.percentage.toFixed(1)}%`);
         },

package/dist/examples/rag/rag-hyperdb/pipeline.js

@@ -7,7 +7,6 @@ try {
     const workspace = "pipeline-example";
     const modelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
         onProgress: (progress) => {
             console.log(`Loading model... ${progress.percentage.toFixed(1)}%`);
         },

package/dist/examples/rag/rag-hyperdb/workspaces.js

@@ -4,7 +4,6 @@ try {
     // Load model
     const modelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
         onProgress: (p) => process.stdout.write(`\rLoading: ${Math.round(p.percentage * 100)}%`),
     });
     console.log("\n");

package/dist/examples/rag/rag-lancedb.js

@@ -8,7 +8,6 @@ try {
     const db = await lancedb.connect(".rag-lancedb");
     const modelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
         onProgress: (progress) => {
             console.log(`Loading model... ${progress.percentage.toFixed(1)}%`);
         },

package/dist/examples/rag/rag-sqlite.js

@@ -9,7 +9,6 @@ try {
     const db = new sqlite3.oo1.DB(":memory:", "c");
     const modelId = await loadModel({
         modelSrc: GTE_LARGE_FP16,
-        modelType: "embeddings",
         onProgress: (progress) => {
             console.log(`Loading model... ${progress.percentage.toFixed(1)}%`);
         },

package/dist/examples/suspend-resume.js

@@ -3,7 +3,6 @@ try {
     // Load a model
     const modelId = await loadModel({
         modelSrc: LLAMA_3_2_1B_INST_Q4_0,
-        modelType: "llm",
         onProgress: (progress) => {
             console.log(progress);
         },

package/dist/examples/tools/llamacpp-native-tools.js

@@ -3,7 +3,6 @@ import { tools, toolSchemas, mockExecute } from "./shared.js";
 try {
     const modelId = await loadModel({
         modelSrc: QWEN3_1_7B_INST_Q4,
-        modelType: "llm",
         modelConfig: {
             ctx_size: 4096,
             tools: true,

package/dist/examples/tools/llamacpp-tools-gemma4.js

@@ -22,7 +22,7 @@ let modelId;
 try {
     modelId = await loadModel({
         modelSrc,
-        modelType: "llm",
+        modelType: "llamacpp-completion",
         modelConfig: { ctx_size: 4096, tools: true },
         onProgress: (progress) => console.log(`Loading: ${progress.percentage.toFixed(1)}%`),
     });

package/dist/examples/tools/llamacpp-tools-pythonic.js

@@ -1,10 +1,10 @@
-import { completion, loadModel, unloadModel, } from "@qvac/sdk";
+import { completion, loadModel, unloadModel } from "@qvac/sdk";
 import { tools } from "./shared.js";
 const LFM_TOOL_HF = "https://huggingface.co/LiquidAI/LFM2-1.2B-Tool-GGUF/resolve/main/LFM2-1.2B-Tool-Q4_K_M.gguf";
 try {
     const modelId = await loadModel({
         modelSrc: LFM_TOOL_HF,
-        modelType: "llm",
+        modelType: "llamacpp-completion",
         modelConfig: { ctx_size: 4096, tools: true },
         onProgress: (progress) => console.log(`Loading: ${progress.percentage.toFixed(1)}%`),
     });

package/dist/examples/tools/llamacpp-tools-qwen35.js

@@ -18,7 +18,7 @@ let modelId;
 try {
     modelId = await loadModel({
         modelSrc,
-        modelType: "llm",
+        modelType: "llamacpp-completion",
         modelConfig: { ctx_size: 4096, tools: true },
         onProgress: (progress) => console.log(`Loading: ${progress.percentage.toFixed(1)}%`),
     });

package/dist/examples/transcription/parakeet-ctc-filesystem.js

@@ -24,7 +24,7 @@ try {
     console.log("Loading Parakeet CTC model...");
     const modelId = await loadModel({
         modelSrc: parakeetModelSrc,
-        modelType: "parakeet",
+        modelType: "parakeet-transcription",
         onProgress: (progress) => {
             console.log(`Download progress: ${progress.percentage.toFixed(1)}%`);
         },

package/dist/examples/transcription/parakeet-microphone-record.js

@@ -29,7 +29,6 @@ catch {
 console.log("Loading Parakeet model...");
 const modelId = await loadModel({
     modelSrc: PARAKEET_TDT_0_6B_V3_Q8_0,
-    modelType: "parakeet",
     onProgress: (p) => console.log(`Download: ${p.percentage.toFixed(1)}%`),
 });
 console.log("Model loaded.\n");

package/dist/examples/transcription/parakeet-microphone-stream.js

@@ -43,7 +43,6 @@ try {
     console.log("Loading Parakeet (EOU) streaming model...");
     modelId = await loadModel({
         modelSrc: PARAKEET_EOU_120M_V1_Q8_0,
-        modelType: "parakeet",
         onProgress: (p) => console.log(`Download: ${p.percentage.toFixed(1)}%`),
     });
     console.log("Model loaded.\n");

package/dist/examples/transcription/parakeet-sortformer.js

@@ -27,7 +27,7 @@ try {
     // ── Step 1: Diarize with Sortformer ──
     const sfModelId = await loadModel({
         modelSrc: sortformerSrc,
-        modelType: "parakeet",
+        modelType: "parakeet-transcription",
     });
     const diarization = await transcribe({
         modelId: sfModelId,
@@ -38,7 +38,6 @@ try {
     // ── Step 2: Transcribe each segment with TDT ──
     const tdtModelId = await loadModel({
         modelSrc: PARAKEET_TDT_0_6B_V3_Q8_0,
-        modelType: "parakeet",
     });
     const pcm = readPcm(audioFilePath);
     const sliceDir = join(tmpdir(), `qvac-diarize-${Date.now()}`);

package/dist/examples/transcription/parakeet-tdt-filesystem.js

@@ -25,7 +25,7 @@ try {
     console.log("Loading Parakeet model...");
     const modelId = await loadModel({
         modelSrc: parakeetModelSrc,
-        modelType: "parakeet",
+        modelType: "parakeet-transcription",
         onProgress: (progress) => {
             console.log(`Download progress: ${progress.percentage.toFixed(1)}%`);
         },

package/dist/examples/transcription/whispercpp-filesystem-streaming.js

@@ -29,7 +29,6 @@ try {
     console.log("Loading model...");
     const modelId = await loadModel({
         modelSrc: WHISPER_TINY,
-        modelType: "whisper",
         modelConfig: {
             vadModelSrc: VAD_SILERO_5_1_2,
             audio_format: "f32le",
@@ -54,11 +53,16 @@ try {
     const session = await transcribeStream({ modelId, metadata: true });
     console.log("Session open. Streaming audio...\n");
     const ffmpeg = spawn("ffmpeg", [
-        "-i", SAMPLE_FILE,
-        "-ar", String(SAMPLE_RATE),
-        "-ac", "1",
-        "-sample_fmt", "flt",
-        "-f", "f32le",
+        "-i",
+        SAMPLE_FILE,
+        "-ar",
+        String(SAMPLE_RATE),
+        "-ac",
+        "1",
+        "-sample_fmt",
+        "flt",
+        "-f",
+        "f32le",
         "pipe:1",
     ], { stdio: ["ignore", "pipe", "ignore"] });
     let totalBytes = 0;

package/dist/examples/transcription/whispercpp-filesystem.js

@@ -12,7 +12,6 @@ try {
     console.log("📥 Loading Whisper model...");
     const modelId = await loadModel({
         modelSrc: WHISPER_TINY,
-        modelType: "whisper",
         modelConfig: {
             audio_format: "f32le",
             // Sampling strategy
@@ -66,7 +65,10 @@ try {
         const end = (segment.endMs / 1000).toFixed(2);
         console.log(`  [${start}s → ${end}s] (id=${segment.id}, append=${segment.append}) ${segment.text}`);
     }
-    console.log(`\nFull transcript: ${segments.map((s) => s.text).join("").trim()}`);
+    console.log(`\nFull transcript: ${segments
+        .map((s) => s.text)
+        .join("")
+        .trim()}`);
     // Unload the model when done
     console.log("🧹 Unloading Whisper model...");
     await unloadModel({ modelId });

package/dist/examples/transcription/whispercpp-microphone-conversation.js

@@ -48,7 +48,6 @@ try {
     console.log("Loading model (whisper-tiny + Silero VAD)...");
     modelId = await loadModel({
         modelSrc: WHISPER_TINY,
-        modelType: "whisper",
         modelConfig: {
             vadModelSrc: VAD_SILERO_5_1_2,
             audio_format: "f32le",

package/dist/examples/transcription/whispercpp-microphone-record.js

@@ -25,7 +25,6 @@ catch {
 console.log("Loading model (whisper-tiny + Silero VAD)...");
 const modelId = await loadModel({
     modelSrc: WHISPER_TINY,
-    modelType: "whisper",
     modelConfig: {
         vadModelSrc: VAD_SILERO_5_1_2,
         audio_format: "f32le",

package/dist/examples/transcription/whispercpp-prompt.js

@@ -16,7 +16,6 @@ try {
     console.log("📥 Loading Whisper model...");
     const modelId = await loadModel({
         modelSrc: WHISPER_TINY,
-        modelType: "whisper",
         modelConfig: {
             audio_format: "f32le",
             // Sampling strategy

package/dist/examples/translation/translation-bergamot-batch.js

@@ -2,7 +2,6 @@ import { loadModel, translate, unloadModel, BERGAMOT_EN_FR } from "@qvac/sdk";
 try {
     const modelId = await loadModel({
         modelSrc: BERGAMOT_EN_FR,
-        modelType: "nmt",
         modelConfig: {
             engine: "Bergamot",
             from: "en",
@@ -30,7 +29,7 @@ try {
     const result = translate({
         modelId,
         text: texts, // Pass array for batch processing
-        modelType: "nmt",
+        modelType: "nmtcpp-translation",
         stream: false,
     });
     const translatedText = await result.text;

package/dist/examples/translation/translation-bergamot-pivot.js

@@ -1,4 +1,4 @@
-import { BERGAMOT_ES_EN, BERGAMOT_EN_IT, loadModel, translate, unloadModel } from "@qvac/sdk";
+import { BERGAMOT_ES_EN, BERGAMOT_EN_IT, loadModel, translate, unloadModel, } from "@qvac/sdk";
 /**
  * Example: Pivot Translation with Bergamot
  *
@@ -17,7 +17,6 @@ try {
     // Load the primary model (Spanish → English) with pivot configuration
     const modelId = await loadModel({
         modelSrc: BERGAMOT_ES_EN, // Primary model: Spanish → English
-        modelType: "nmt",
         modelConfig: {
             engine: "Bergamot",
             from: "es",
@@ -35,7 +34,7 @@ try {
                 topk: 100,
                 normalize: 1,
                 lengthpenalty: 1.2,
-            }
+            },
         },
         onProgress: (progress) => {
             console.log(progress);
@@ -54,7 +53,7 @@ try {
     const result = translate({
         modelId,
         text: spanishText,
-        modelType: "nmt",
+        modelType: "nmtcpp-translation",
         stream: false,
     });
     const italianText = await result.text;

package/dist/examples/translation/translation-bergamot.js

@@ -6,7 +6,6 @@ try {
     // - dstVocabSrc: target vocabulary file (optional)
     const modelId = await loadModel({
         modelSrc: BERGAMOT_EN_FR,
-        modelType: "nmt",
         modelConfig: {
             engine: "Bergamot",
             from: "en",
@@ -26,7 +25,7 @@ try {
     const result = translate({
         modelId,
         text,
-        modelType: "nmt",
+        modelType: "nmtcpp-translation",
         stream: false,
     });
     const translatedText = await result.text;

package/dist/examples/translation/translation-indic.js

@@ -5,7 +5,6 @@ import { loadModel, translate, unloadModel, MARIAN_EN_HI_INDIC_1B_Q4_0, } from "
 try {
     const modelId = await loadModel({
         modelSrc: MARIAN_EN_HI_INDIC_1B_Q4_0,
-        modelType: "nmt",
         onProgress: (progress) => {
             console.log(progress);
         },
@@ -20,7 +19,7 @@ try {
     const result = translate({
         modelId,
         text,
-        modelType: "nmt",
+        modelType: "nmtcpp-translation",
         stream: false,
     });
     const translatedText = await result.text;

package/dist/examples/translation/translation-llm-afriquegemma.js

@@ -2,7 +2,6 @@ import { translate, loadModel, unloadModel, AFRICAN_4B_TRANSLATION_Q4_K_M, } fro
 try {
     const modelId = await loadModel({
         modelSrc: AFRICAN_4B_TRANSLATION_Q4_K_M,
-        modelType: "llm",
         onProgress: (progress) => {
             console.log(progress);
         },
@@ -16,8 +15,8 @@ try {
             repeat_penalty: 1,
             seed: 42,
             predict: 256,
-            stop_sequences: ['\n']
-        }
+            stop_sequences: ["\n"],
+        },
     });
     // With explicit source language
     const engText = "Hello, how are you today?";
@@ -26,7 +25,7 @@ try {
         text: engText,
         from: "en",
         to: "swh_Latn",
-        modelType: "llm",
+        modelType: "llamacpp-completion",
         stream: false,
     });
     const translatedTextExplicit = await resultExplicit.text;

package/dist/examples/translation/translation-llm-context.js

@@ -2,7 +2,6 @@ import { translate, loadModel, unloadModel, SALAMANDRATA_2B_INST_Q8, } from "@qv
 try {
     const modelId = await loadModel({
         modelSrc: SALAMANDRATA_2B_INST_Q8,
-        modelType: "llm",
         onProgress: (progress) => {
             console.log(progress);
         },
@@ -15,7 +14,7 @@ try {
         text: engText,
         from: "en",
         to: "es",
-        modelType: "llm",
+        modelType: "llamacpp-completion",
         context: "Use formal language, letter for financial institution",
         stream: false,
     });