@qvac/sdk 0.16.0 → 0.17.0

package/dist/examples/diffusion-txt2vid-wan22.js

@@ -0,0 +1,96 @@
+import { loadModel, unloadModel, video, WAN2_2_TI2V_5B_Q5_K_S, UMT5_XXL_FP16, WAN_2_2_COMFYUI_REPACKAGED_VAE } from '@qvac/sdk';
+import fs from 'fs';
+import path from 'path';
+// Text-to-video with Wan 2.2 TI2V-5B Turbo (Q5_K_S). Like Wan 2.1 this is a
+// split layout — diffusion model + UMT5-XXL text encoder + VAE — but it needs
+// the Wan 2.2 VAE specifically: its 16x spatial compression is what gives
+// TI2V its 32-pixel grid, and the Wan 2.1 VAE is not interchangeable.
+//
+// TI2V-5B is a single-expert model. The high_noise_* / moe_boundary fields
+// belong to the two-expert Wan 2.2 A14B layout and are rejected here.
+//
+// Budget at least 16 GB of video memory or 20 GB of unified memory. The
+// artifacts total roughly 16.4 GB, dominated by the fp16 text encoder.
+//
+// The text encoder is byte-identical to the Wan 2.1 repackage, so both
+// generations share the one UMT5_XXL_FP16 registry entry.
+const diffusionModelSrc = process.argv[2] || WAN2_2_TI2V_5B_Q5_K_S;
+const t5XxlModelSrc = process.argv[3] || UMT5_XXL_FP16;
+const vaeModelSrc = process.argv[4] || WAN_2_2_COMFYUI_REPACKAGED_VAE;
+// Prompt tip: Turbo responds well to explicit camera and lighting direction.
+// Describe continuous motion rather than a pose, or the clip reads as a still.
+const prompt = process.argv[5] ||
+    'A single white porcelain espresso cup on a dark walnut table beside a sunlit window, ' +
+        'delicate steam curling upward, slow circular camera move, warm morning light, ' +
+        'sharp ceramic texture, realistic continuous motion';
+const outputDir = process.argv[6] || '.';
+try {
+    console.log('▸ Loading Wan 2.2 TI2V-5B Turbo model (diffusion + UMT5-XXL + Wan 2.2 VAE)...');
+    const modelId = await loadModel({
+        modelSrc: diffusionModelSrc,
+        modelType: 'sdcpp-generation',
+        modelConfig: {
+            mode: 'video',
+            device: 'gpu',
+            threads: 4,
+            t5XxlModelSrc,
+            vaeModelSrc,
+            diffusion_fa: true,
+            offload_to_cpu: true,
+            vae_tiling: true
+        },
+        onProgress: (p) => {
+            const mb = (n) => (n / 1e6).toFixed(1);
+            const line = `▸ Downloading ${p.percentage.toFixed(0)}% (${mb(p.downloaded)}/${mb(p.total)} MB)`;
+            process.stderr.write(process.stderr.isTTY ? `\r${line}` : `${line}\n`);
+            if (p.percentage >= 100)
+                process.stderr.write('\n');
+        }
+    });
+    console.log(`▸ Model loaded: ${modelId}`);
+    console.log(`\n▸ Generating video for: "${prompt}"`);
+    const { progressStream, outputs, stats } = video({
+        modelId,
+        mode: 'txt2vid',
+        prompt,
+        negative_prompt: 'flickering, temporal jitter, morphing, duplicated subject, warped geometry, ' +
+            'distorted anatomy, blurry details, low resolution, text, watermark, logo',
+        // TI2V-5B requires width and height to be multiples of 32 — stricter than
+        // the multiple of 16 the schema enforces for Wan generally. Native
+        // validation derives this from the loaded GGUF, so a non-conforming size
+        // fails at generation time rather than at load.
+        // Turbo was trained on this 720p, 24 fps, five-second shape; 640x352 with
+        // 49 frames is a much cheaper way to smoke-test the pipeline.
+        width: 1280,
+        height: 704,
+        // Frame count must satisfy (4*k + 1), k >= 1. At 24 fps: 49 ≈ 2s,
+        // 121 ≈ 5s (the trained length).
+        video_frames: 121,
+        fps: 24,
+        // Turbo is distilled for very short schedules: 4 steps with guidance
+        // effectively disabled. The non-distilled TI2V-5B wants steps >= 30 and
+        // cfg ~5.0 instead.
+        steps: 4,
+        cfg_scale: 1.0,
+        flow_shift: 5.0,
+        seed: 42,
+        vae_tiling: true
+    });
+    for await (const { step, totalSteps } of progressStream) {
+        console.log(`▸ step ${step}/${totalSteps}`);
+    }
+    const buffers = await outputs;
+    for (let i = 0; i < buffers.length; i++) {
+        const outputPath = path.join(outputDir, `wan22_ti2v_t2v_${i}.avi`);
+        fs.writeFileSync(outputPath, buffers[i]);
+        console.log(`▸ Saved ${outputPath}`);
+    }
+    console.log('\n▸ Stats:', await stats);
+    await unloadModel({ modelId, clearStorage: false });
+    console.log('▸ Done.');
+    process.exit(0);
+}
+catch (error) {
+    console.error('✖', error);
+    process.exit(1);
+}

package/dist/examples/quickstart.bare.js

@@ -3,7 +3,7 @@
 // then register the plugins this example uses via `plugins([...])`.
 import bareProcess from 'bare-process';
 import { plugins, LLAMA_3_2_1B_INST_Q4_0 } from '@qvac/sdk';
-import { llmPlugin } from '@qvac/sdk/llamacpp-completion/plugin.js';
+import { llmPlugin } from '@qvac/sdk/llamacpp-completion/plugin';
 globalThis.process = bareProcess;
 const { loadModel, completion, unloadModel } = plugins([llmPlugin]);
 // From here it is the same as the Node quickstart.

package/dist/examples/rag/rag-mongodb.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=rag-mongodb.d.ts.map

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

@@ -0,0 +1,180 @@
+import { embed, loadModel, unloadModel, GTE_LARGE_FP16 } from '@qvac/sdk';
+import { MongoClient } from 'mongodb';
+const INDEX_NAME = 'documents_vector_index';
+const MONGODB_SETUP_INSTRUCTIONS = `
+▸ This example needs a MongoDB deployment with Atlas Vector Search.
+
+One way to get one is to run it in Docker:
+
+   docker run -p 27017:27017 --name atlas-local mongodb/mongodb-atlas-local
+
+For more details, visit: https://www.mongodb.com/docs/atlas/cli/current/atlas-cli-deploy-docker/
+`;
+async function initializeMongoClient() {
+    // Replace with your own deployment's connection string if it is not the Docker one
+    // https://www.mongodb.com/docs/manual/reference/connection-string/
+    const client = new MongoClient('mongodb://localhost:27017/?directConnection=true');
+    try {
+        await client.connect();
+        await client.db('admin').command({ ping: 1 });
+        console.log('▸ Connected to MongoDB server');
+        return client;
+    }
+    catch {
+        console.error('✖ Failed to connect to MongoDB server');
+        console.error('▸ Please ensure the server is running on localhost:27017');
+        console.error(MONGODB_SETUP_INSTRUCTIONS);
+        process.exit(1);
+    }
+}
+function wait(ms) {
+    return new Promise((resolve) => setTimeout(resolve, ms));
+}
+try {
+    // Get query and category from command line or use defaults
+    const query = process.argv[2] || 'machine learning algorithms';
+    const category = process.argv[3] || 'ai';
+    console.log(`▸ Query: "${query}" (category: "${category}")`);
+    const client = await initializeMongoClient();
+    const collection = client.db('qvac').collection('documents');
+    const modelId = await loadModel({
+        modelSrc: GTE_LARGE_FP16,
+        onProgress: (p) => {
+            const mb = (n) => (n / 1e6).toFixed(1);
+            const line = `▸ Downloading ${p.percentage.toFixed(0)}% (${mb(p.downloaded)}/${mb(p.total)} MB)`;
+            process.stderr.write(process.stderr.isTTY ? `\r${line}` : `${line}\n`);
+            if (p.percentage >= 100)
+                process.stderr.write('\n');
+        }
+    });
+    // Sample corpus, each document tagged with a category to filter on
+    const samples = [
+        {
+            id: 1,
+            category: 'ai',
+            text: 'Machine learning is a subset of artificial intelligence that focuses on algorithms that can learn and make predictions from data without being explicitly programmed for every task.'
+        },
+        {
+            id: 2,
+            category: 'ai',
+            text: 'Deep learning uses neural networks with multiple layers to process and learn from complex data patterns, enabling breakthroughs in image recognition and natural language processing.'
+        },
+        {
+            id: 3,
+            category: 'ai',
+            text: 'Natural language processing combines computational linguistics with machine learning to help computers understand, interpret, and generate human language in a meaningful way.'
+        },
+        {
+            id: 4,
+            category: 'ai',
+            text: 'Computer vision enables machines to interpret and understand visual information from the world, using techniques like image classification, object detection, and facial recognition.'
+        },
+        {
+            id: 5,
+            category: 'computing',
+            text: 'Quantum computing leverages quantum mechanical phenomena to process information in fundamentally different ways than classical computers, potentially solving certain problems exponentially faster.'
+        },
+        {
+            id: 6,
+            category: 'security',
+            text: 'Blockchain technology creates decentralized, immutable ledgers that enable secure peer-to-peer transactions without requiring a central authority or intermediary.'
+        },
+        {
+            id: 7,
+            category: 'computing',
+            text: 'Cloud computing delivers computing services over the internet, allowing users to access resources like storage, processing power, and applications on-demand from anywhere.'
+        },
+        {
+            id: 8,
+            category: 'security',
+            text: 'Cybersecurity protects digital systems, networks, and data from malicious attacks, unauthorized access, and various forms of cyber threats through multiple layers of defense.'
+        }
+    ];
+    // (Re)create the collection
+    try {
+        await collection.drop();
+    }
+    catch (e) {
+        console.warn(`▸ Collection didn't exist, no need to drop: ${String(e)}`);
+    }
+    // Embed and store documents
+    console.log('▸ Embedding documents...');
+    const documents = [];
+    for (const sample of samples) {
+        const { embedding } = await embed({ modelId, text: sample.text });
+        documents.push({
+            id: sample.id,
+            category: sample.category,
+            text: sample.text,
+            embedding
+        });
+    }
+    await collection.insertMany(documents);
+    // numDimensions is fixed at index creation and must match the model: GTE Large is 1024
+    console.log('▸ Creating vector search index...');
+    await collection.createSearchIndex({
+        name: INDEX_NAME,
+        type: 'vectorSearch',
+        definition: {
+            fields: [
+                {
+                    type: 'vector',
+                    path: 'embedding',
+                    numDimensions: 1024,
+                    similarity: 'cosine'
+                },
+                // A field must be indexed as a filter to be usable in a $vectorSearch filter
+                {
+                    type: 'filter',
+                    path: 'category'
+                }
+            ]
+        }
+    });
+    // Index builds are asynchronous; querying too early returns no matches
+    for (let attempt = 0; attempt < 60; attempt++) {
+        const [index] = (await collection.listSearchIndexes(INDEX_NAME).toArray());
+        if (index?.queryable)
+            break;
+        if (attempt === 59)
+            throw new Error(`Index ${INDEX_NAME} did not become queryable`);
+        await wait(1000);
+    }
+    console.log('▸ Searching for similar documents...');
+    const { embedding: queryEmbedding } = await embed({ modelId, text: query });
+    const results = await collection
+        .aggregate([
+        {
+            $vectorSearch: {
+                index: INDEX_NAME,
+                path: 'embedding',
+                queryVector: queryEmbedding,
+                filter: { category: { $eq: category } },
+                numCandidates: 100,
+                limit: 3
+            }
+        },
+        {
+            $project: {
+                _id: 0,
+                id: 1,
+                category: 1,
+                text: 1,
+                score: { $meta: 'vectorSearchScore' }
+            }
+        }
+    ])
+        .toArray();
+    console.log('▸ Top 3 most similar documents:');
+    results.forEach((result, index) => {
+        console.log(`${index + 1}. (Score: ${result.score.toFixed(4)}, Category: ${result.category})`);
+        console.log(`   ${result.text}`);
+        console.log();
+    });
+    await unloadModel({ modelId });
+    await client.close();
+}
+catch (error) {
+    console.error('✖', error);
+    process.exit(1);
+}

package/dist/examples/tools/llamacpp-tools-auto-dialect.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=llamacpp-tools-auto-dialect.d.ts.map

package/dist/examples/tools/llamacpp-tools-auto-dialect.js

@@ -0,0 +1,77 @@
+/**
+ * Tool-calling example with automatic dialect detection.
+ *
+ * The SDK infers the tool-call format from the model name/path. It supports
+ * model families such as Qwen3.5/3.6, Gemma4, GPT-OSS, DeepSeek V3.2/V4, and
+ * LFM without requiring a toolDialect override.
+ *
+ * Usage:
+ *   bun run bare:example dist/examples/tools/llamacpp-tools-auto-dialect.js <model-src>
+ */
+import { completion, loadModel, unloadModel } from '@qvac/sdk';
+import { tools, mockExecute } from './shared.js';
+const modelSrc = process.argv[2];
+if (!modelSrc) {
+    console.error('✖ Missing model source: pass a tool-capable GGUF path or URL');
+    process.exit(1);
+}
+let modelId;
+try {
+    modelId = await loadModel({
+        modelSrc,
+        modelType: 'llamacpp-completion',
+        modelConfig: { ctx_size: 4096, tools: true },
+        onProgress: (p) => {
+            const mb = (n) => (n / 1e6).toFixed(1);
+            const line = `▸ Downloading ${p.percentage.toFixed(0)}% (${mb(p.downloaded)}/${mb(p.total)} MB)`;
+            process.stderr.write(process.stderr.isTTY ? `\r${line}` : `${line}\n`);
+            if (p.percentage >= 100)
+                process.stderr.write('\n');
+        }
+    });
+    console.log(`▸ Model loaded: ${modelId}`);
+    const history = [
+        {
+            role: 'system',
+            content: 'You are a helpful assistant that can call tools to look up weather and horoscopes.'
+        },
+        {
+            role: 'user',
+            content: "What's the weather in Tokyo and my horoscope for Aquarius?"
+        }
+    ];
+    const result = completion({ modelId, history, stream: true, tools });
+    const tokensTask = (async () => {
+        for await (const token of result.tokenStream) {
+            process.stdout.write(token);
+        }
+    })();
+    const toolsTask = (async () => {
+        for await (const evt of result.toolCallStream) {
+            if (evt.type === 'toolCall') {
+                console.log(`\n▸ ${evt.call.name}(${JSON.stringify(evt.call.arguments)})`);
+            }
+        }
+    })();
+    await Promise.all([tokensTask, toolsTask]);
+    const toolCalls = await result.toolCalls;
+    console.log('\n\n▸ Final tool calls:');
+    if (toolCalls.length > 0) {
+        for (const call of toolCalls) {
+            console.log(`▸ ${call.name}(${JSON.stringify(call.arguments)})`);
+            const toolResult = mockExecute(call.name, call.arguments);
+            console.log(`▸ result: ${toolResult}`);
+        }
+    }
+    else {
+        console.log('▸ (none)');
+    }
+    await unloadModel({ modelId, clearStorage: false });
+    process.exit(0);
+}
+catch (error) {
+    console.error('✖', error);
+    if (modelId)
+        await unloadModel({ modelId, clearStorage: false }).catch(() => { });
+    process.exit(1);
+}

package/dist/examples/tts/parler.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=parler.d.ts.map

package/dist/examples/tts/parler.js

@@ -0,0 +1,51 @@
+import { loadModel, textToSpeech, unloadModel, TTS_MINI_V1_EN_PARLER_TTS_Q8_0 } from '@qvac/sdk';
+import { createWav, playAudio, int16ArrayToBuffer, createWavHeader } from './utils.js';
+// Parler-TTS (GGML): description-conditioned speech with per-call voice controls.
+// Uses the registry-hosted Mini v1 Q8_0 model and its native 44.1 kHz output.
+const PARLER_SAMPLE_RATE = 44100;
+try {
+    const modelId = await loadModel({
+        modelSrc: TTS_MINI_V1_EN_PARLER_TTS_Q8_0,
+        modelConfig: {
+            ttsEngine: 'parler',
+            voice: 'Laura',
+            seed: 42
+        },
+        onProgress: (p) => {
+            const mb = (n) => (n / 1e6).toFixed(1);
+            const line = `▸ Downloading ${p.percentage.toFixed(0)}% (${mb(p.downloaded)}/${mb(p.total)} MB)`;
+            process.stderr.write(process.stderr.isTTY ? `\r${line}` : `${line}\n`);
+            if (p.percentage >= 100)
+                process.stderr.write('\n');
+        }
+    });
+    console.log(`▸ Model loaded: ${modelId}`);
+    console.log('▸ Testing Parler Text-to-Speech...');
+    const result = textToSpeech({
+        modelId,
+        text: 'Hey, how are you doing today?',
+        inputType: 'text',
+        stream: false,
+        emotion: 'happy'
+    });
+    const audioBuffer = await result.buffer;
+    console.log(`▸ TTS complete. Total samples: ${audioBuffer.length}`);
+    console.log('▸ Saving audio to file...');
+    createWav(audioBuffer, PARLER_SAMPLE_RATE, 'parler-output.wav');
+    console.log('▸ Audio saved to parler-output.wav');
+    console.log('▸ Playing audio...');
+    const audioData = int16ArrayToBuffer(audioBuffer);
+    const wavBuffer = Buffer.concat([
+        createWavHeader(audioData.length, PARLER_SAMPLE_RATE),
+        audioData
+    ]);
+    playAudio(wavBuffer);
+    console.log('▸ Audio playback complete');
+    await unloadModel({ modelId });
+    console.log('▸ Model unloaded');
+    process.exit(0);
+}
+catch (error) {
+    console.error('✖', error);
+    process.exit(1);
+}

package/dist/examples/vla-groot.js

@@ -8,8 +8,9 @@
  * `vlaHparams()`:
  *   - `imageInputMode: 'patches'` — each `images` entry is a pre-patchified
  *     buffer of `hparams.imagePatchElems` floats, NOT a `3·w·h` pixel plane.
- *     A real consumer runs the model's own vision patch-embed over each camera
- *     frame; here we feed synthetic patch buffers of the right length.
+ *     A real consumer patchifies each camera frame the way `Gr00tPolicy` does
+ *     (the model's learned patch-embed runs inside the addon); here we feed
+ *     synthetic patch buffers of the right length.
  *   - `stateInputMode: 'continuous'` — the robot state is projected by an
  *     in-model linear layer; pad it to `hparams.maxStateDim` with `vlaPadState`.
  *   - `noise` is REQUIRED. GR00T is a flow-matching model that does not sample
@@ -66,8 +67,9 @@ try {
     }
     const numCameras = hparams.numCameras ?? 2;
     // Patch-input model: each camera is a pre-patchified buffer of exactly
-    // `imagePatchElems` floats. A real consumer fills these from the model's
-    // vision patch-embed over the camera frame; we use small synthetic values.
+    // `imagePatchElems` floats. A real consumer patchifies each camera frame the
+    // way `Gr00tPolicy` does (the model's learned patch-embed runs inside the
+    // addon); we use small synthetic values.
     const images = Array.from({ length: numCameras }, () => new Float32Array(patchElems).fill(0.02));
     // Continuous-state model: pad the robot state to `maxStateDim`.
     const state = vlaPadState([0, 0, 0, 0, 0, 0], hparams.maxStateDim);

package/dist/examples/voice-assistant/voice-assistant.js

@@ -12,8 +12,8 @@
  */
 import { loadModel, unloadModel, transcribeStream, completion, textToSpeech, WHISPER_TINY, VAD_SILERO_5_1_2, LLAMA_3_2_1B_INST_Q4_0, TTS_EN_SUPERTONIC_Q8_0 } from '@qvac/sdk';
 import { spawnSync } from 'child_process';
-import { startMicrophone } from '../audio/mic-input';
-import { createWavHeader, int16ArrayToBuffer, playAudio } from '../tts/utils';
+import { startMicrophone } from '../audio/mic-input.js';
+import { createWavHeader, int16ArrayToBuffer, playAudio } from '../tts/utils.js';
 const MIC_SAMPLE_RATE = 16000;
 const TTS_SAMPLE_RATE = 44100;
 const SYSTEM_PROMPT = 'You are a concise, friendly voice assistant. Keep responses under two sentences. ' +

package/dist/expo/plugins/index.d.ts

@@ -1,12 +1,12 @@
-export { default as withAndroidArchitecture } from './withAndroidArchitecture';
-export { default as withAndroidNdkVersion } from './withAndroidNdkVersion';
-export { default as withDeviceInfo } from './withDeviceInfo';
-export { default as withMobileBundle } from './withMobileBundle';
-export { default as withOpenCL } from './withOpenCL';
-export { default as withQvacSDK } from './withQvacSDK';
-export { getProjectRootFromMod, getProjectRootFromBaseConfig } from './get-project-root';
-export { resolveSDKPackageDir, SDK_PACKAGE_NAMES } from './resolve-sdk-package-dir';
-export type { SDKPackageInfo } from './resolve-sdk-package-dir';
-export { findInAncestorNodeModules } from './find-in-ancestor-node-modules';
-export { default } from './withQvacSDK';
+export { default as withAndroidArchitecture } from './withAndroidArchitecture.js';
+export { default as withAndroidNdkVersion } from './withAndroidNdkVersion.js';
+export { default as withDeviceInfo } from './withDeviceInfo.js';
+export { default as withMobileBundle } from './withMobileBundle.js';
+export { default as withOpenCL } from './withOpenCL.js';
+export { default as withQvacSDK } from './withQvacSDK.js';
+export { getProjectRootFromMod, getProjectRootFromBaseConfig } from './get-project-root.js';
+export { resolveSDKPackageDir, SDK_PACKAGE_NAMES } from './resolve-sdk-package-dir.js';
+export type { SDKPackageInfo } from './resolve-sdk-package-dir.js';
+export { findInAncestorNodeModules } from './find-in-ancestor-node-modules.js';
+export { default } from './withQvacSDK.js';
 //# sourceMappingURL=index.d.ts.map

package/dist/expo/plugins/withDeviceInfo.js

@@ -1,7 +1,7 @@
 import * as fs from 'fs';
 import * as path from 'path';
-import { resolveSDKPackageDir } from './resolve-sdk-package-dir.js';
-import { getProjectRootFromBaseConfig } from './get-project-root.js';
+import { resolveSDKPackageDir } from '../../expo/plugins/resolve-sdk-package-dir.js';
+import { getProjectRootFromBaseConfig } from '../../expo/plugins/get-project-root.js';
 let didRun = false;
 /**
  * Expo plugin that stubs expo-device-info when expo-device is not a declared

package/dist/expo/plugins/withMobileBundle.js

@@ -3,9 +3,9 @@ import * as fs from 'fs';
 import * as path from 'path';
 import { bundleSdk, verifyBundle, hasErrors, formatVerifyBundleResult } from '../../commands/index.js';
 import { CONFIG_CANDIDATES } from '../../client/config-loader/resolve-config.node.js';
-import { resolveSDKPackageDir } from './resolve-sdk-package-dir.js';
-import { getProjectRootFromMod } from './get-project-root.js';
-import { findInAncestorNodeModules } from './find-in-ancestor-node-modules.js';
+import { resolveSDKPackageDir } from '../../expo/plugins/resolve-sdk-package-dir.js';
+import { getProjectRootFromMod } from '../../expo/plugins/get-project-root.js';
+import { findInAncestorNodeModules } from '../../expo/plugins/find-in-ancestor-node-modules.js';
 import { BundleVerificationFailedError } from '../../utils/errors-client.js';
 const { withDangerousMod } = configPlugins;
 /** Modules to defer from mobile bundles (not available at bundle time) */

package/dist/index.d.ts

@@ -1,17 +1,17 @@
-export { batchCompletion, completion, deleteCache, loadModel, downloadAsset, heartbeat, startQVACProvider, stopQVACProvider, unloadModel, transcribe, transcribeStream, bciTranscribe, bciTranscribeStream, embed, finetune, translate, cancel, ragChunk, ragIngest, ragSaveEmbeddings, ragSearch, ragDeleteEmbeddings, ragReindex, ragListWorkspaces, ragCloseWorkspace, ragDeleteWorkspace, textToSpeech, textToSpeechStream, getModelInfo, getLoadedModelInfo, loggingStream, subscribeServerLogs, type ServerLogHandler, ocr, invokePlugin, invokePluginStream, diffusion, type DiffusionProgressTick, classify, video, type VideoProgressTick, upscale, modelRegistryList, modelRegistrySearch, modelRegistryGetModel, type ModelRegistrySearchParams, suspend, resume, state, vla, vlaHparams, vlaPreprocessImage, vlaPadState, VLA_DEFAULT_IMAGE_SIZE, type FinetuneHandle } from './client/api/index';
-export { close } from './client/index';
-export { plugins } from './client/plugins-factory';
-export { type LifecycleState, type ModelProgressUpdate, type LoadModelOptions, type LoadCustomPluginModelOptions, type DownloadAssetOptions, type Tool, type ToolCall, type ToolCallWithCall, type ToolCallError, type ToolCallEvent, type CompletionEvent, type CompletionFinal, type CompletionRun, type CompletionStats, type BatchCompletionEvent, type BatchCompletionResult, type BatchCompletionRun, type BatchPrompt, type EmbedStats, VERBOSITY, type Attachment, type TranscribeStreamSession, type TranscribeStreamMetadataSession, type TranscribeStreamConversationSession, type TranscribeStreamEvent, type VadStateEvent, type EndOfTurnEvent, type TranscribeSegment, type BciConfig, type BciTranscribeClientParams, type BciTranscribeStreamClientParams, type BciTranscribeStreamSession, type BciTranscribeStreamMetadataSession, type BciStreamOpts, type NeuralInput, type TextToSpeechStreamSession, type TextToSpeechStreamResponse, type TextToSpeechStreamClientParams, type CompletionParams, type ToolDialect, type RagSearchResult, type RagSaveEmbeddingsResult, type RagReindexResult, type RagEmbeddedDoc, type RagDoc, type RagWorkspaceInfo, type RagCloseWorkspaceParams, type RagDeleteWorkspaceParams, type RagIngestStage, type RagReindexStage, type RagSaveStage, SDK_CLIENT_ERROR_CODES, SDK_SERVER_ERROR_CODES, RAG_ERROR_CODES, type QvacConfig, type ModelInfo, type GetModelInfoParams, type GetLoadedModelInfoParams, type LoadedModelInfo, type LoadedInstance, type CacheFileInfo, toolSchema, TOOLS_MODE, type ToolsMode, type McpClient, type McpClientInput, type OCRClientParams, type OCRTextBlock, type OCROptions, type ClassifyClientParams, type ClassificationResult, type DiffusionClientParams, type DiffusionStreamResponse, type DiffusionStats, type VideoClientParams, type VideoStreamResponse, type VideoStats, type UpscaleClientParams, type UpscaleStreamResponse, type UpscaleStats, type VlaConfig, type VlaClientRunParams, type VlaClientRunResult, type VlaHparams, type VlaStats, definePlugin, defineHandler, defineDuplexHandler, type QvacPlugin, type CreateModelParams, type PluginModelResult, type ModelRegistryEntry, type ModelRegistryEntryAddon, PLUGIN_LLM, PLUGIN_EMBEDDING, PLUGIN_WHISPER, PLUGIN_BCI, PLUGIN_NMT, PLUGIN_TTS, PLUGIN_OCR, PLUGIN_DIFFUSION, PLUGIN_VLA, PLUGIN_CLASSIFICATION, SDK_DEFAULT_PLUGINS, type BuiltinPlugin, type ProfilerMode, type FinetuneValidation, type FinetuneRunParams, type FinetuneGetStateParams, type FinetuneStopParams, type FinetuneParams, type FinetuneStatus, type FinetuneProgress, type FinetuneStats, type FinetuneResult } from './schemas/index';
-export { type ToolInput, type ToolHandler } from './utils/tool-helpers';
-export { MODEL_TYPES, ModelType } from './schemas/index';
-export * from './models/registry/index';
-export { SUPPORTED_AUDIO_FORMATS } from './constants/audio';
-export { InferenceCancelledError } from './utils/errors-server';
-export type { InferenceCancelledPartial } from './utils/errors-server';
-export { ContextOverflowError, RequestIdConflictError, RequestNotFoundError, RequestRejectedByPolicyError } from './utils/errors-server';
-export { BareRuntimeBinaryNotFoundError, WorkerCrashedError, WorkerShutdownError, RequestValidationFailedError } from './utils/errors-client';
-export { getLogger, SDK_LOG_ID, SDK_ALL_LOG_ID } from './logging/index';
-export type { Logger, LogTransport, LoggerOptions } from './logging/index';
-export { profiler } from './profiling/index';
-export type { ProfilerRuntimeOptions, ProfilerExport } from './profiling/index';
+export { batchCompletion, completion, deleteCache, loadModel, downloadAsset, heartbeat, startQVACProvider, stopQVACProvider, unloadModel, transcribe, transcribeStream, bciTranscribe, bciTranscribeStream, embed, finetune, translate, cancel, ragChunk, ragIngest, ragSaveEmbeddings, ragSearch, ragDeleteEmbeddings, ragReindex, ragListWorkspaces, ragCloseWorkspace, ragDeleteWorkspace, textToSpeech, textToSpeechStream, getModelInfo, getLoadedModelInfo, getSystemResources, loggingStream, subscribeServerLogs, type ServerLogHandler, ocr, invokePlugin, invokePluginStream, diffusion, type DiffusionProgressTick, classify, video, type VideoProgressTick, upscale, modelRegistryList, modelRegistrySearch, modelRegistryGetModel, type ModelRegistrySearchParams, suspend, resume, state, vla, vlaHparams, vlaPreprocessImage, vlaPadState, VLA_DEFAULT_IMAGE_SIZE, audioGen, type FinetuneHandle } from './client/api/index.js';
+export { close } from './client/index.js';
+export { plugins } from './client/plugins-factory.js';
+export { type LifecycleState, type ModelProgressUpdate, type LoadModelOptions, type LoadCustomPluginModelOptions, type DownloadAssetOptions, type Tool, type ToolCall, type ToolCallWithCall, type ToolCallError, type ToolCallEvent, type CompletionEvent, type CompletionFinal, type CompletionRun, type CompletionStats, type BatchCompletionEvent, type BatchCompletionResult, type BatchCompletionRun, type BatchPrompt, type EmbedStats, VERBOSITY, type Attachment, type TranscribeStreamSession, type TranscribeStreamMetadataSession, type TranscribeStreamConversationSession, type TranscribeStreamEvent, type VadStateEvent, type EndOfTurnEvent, type TranscribeSegment, type BciConfig, type BciTranscribeClientParams, type BciTranscribeStreamClientParams, type BciTranscribeStreamSession, type BciTranscribeStreamMetadataSession, type BciStreamOpts, type NeuralInput, type TextToSpeechStreamSession, type TextToSpeechStreamResponse, type TextToSpeechStreamClientParams, type TtsClientParamsInput, type TtsParlerEmotion, type TtsParlerLoadConfig, type TtsParlerRuntimeConfig, type CompletionParams, type ToolDialect, type RagSearchResult, type RagSaveEmbeddingsResult, type RagReindexResult, type RagEmbeddedDoc, type RagDoc, type RagWorkspaceInfo, type RagCloseWorkspaceParams, type RagDeleteWorkspaceParams, type RagIngestStage, type RagReindexStage, type RagSaveStage, SDK_CLIENT_ERROR_CODES, SDK_SERVER_ERROR_CODES, RAG_ERROR_CODES, type QvacConfig, type ModelInfo, type GetModelInfoParams, type GetLoadedModelInfoParams, type LoadedModelInfo, type GetSystemResourcesInput, type ResourceScope, type ResourceProvenance, type ResourceMetric, type GraphicsDriver, type GraphicsDriverCapabilities, type BackendProbeResult, type BackendDevice, type BackendDriver, type BackendFallback, type InferenceBackendDiagnostics, type CPUResourceCapabilities, type GPUResourceCapabilities, type SystemResourceCapabilities, type GPUResourceSample, type SystemResourceSample, type SystemResources, type LoadedInstance, type CacheFileInfo, toolSchema, TOOLS_MODE, type ToolsMode, type McpClient, type McpClientInput, type OCRClientParams, type OCRTextBlock, type OCROptions, type ClassifyClientParams, type ClassificationResult, type DiffusionClientParams, type DiffusionStreamResponse, type DiffusionStats, type AudioGenClientParams, type AudioGenConfig, type AudioGenRuntimeConfig, type AudioGenProgress, type AudioGenAudio, type AudioGenStats, type AudioGenResult, type AudioGenStreamResponse, type VideoClientParams, type VideoStreamResponse, type VideoStats, type UpscaleClientParams, type UpscaleStreamResponse, type UpscaleStats, type VlaConfig, type VlaClientRunParams, type VlaClientRunResult, type VlaHparams, type VlaStats, definePlugin, defineHandler, defineDuplexHandler, type QvacPlugin, type CreateModelParams, type PluginModelResult, type ModelRegistryEntry, type ModelRegistryEntryAddon, PLUGIN_LLM, PLUGIN_EMBEDDING, PLUGIN_WHISPER, PLUGIN_BCI, PLUGIN_NMT, PLUGIN_TTS, PLUGIN_OCR, PLUGIN_DIFFUSION, PLUGIN_AUDIOGEN, PLUGIN_VLA, PLUGIN_CLASSIFICATION, SDK_DEFAULT_PLUGINS, type BuiltinPlugin, type ProfilerMode, type FinetuneValidation, type FinetuneRunParams, type FinetuneGetStateParams, type FinetuneStopParams, type FinetuneParams, type FinetuneStatus, type FinetuneProgress, type FinetuneStats, type FinetuneResult } from './schemas/index.js';
+export { type ToolInput, type ToolHandler } from './utils/tool-helpers.js';
+export { MODEL_TYPES, ModelType } from './schemas/index.js';
+export * from './models/registry/index.js';
+export { SUPPORTED_AUDIO_FORMATS } from './constants/audio.js';
+export { InferenceCancelledError } from './utils/errors-server.js';
+export type { InferenceCancelledPartial } from './utils/errors-server.js';
+export { ContextOverflowError, RequestIdConflictError, RequestNotFoundError, RequestRejectedByPolicyError, TranslationFailedError } from './utils/errors-server.js';
+export { BareRuntimeBinaryNotFoundError, WorkerCrashedError, WorkerShutdownError, RequestValidationFailedError } from './utils/errors-client.js';
+export { getLogger, SDK_LOG_ID, SDK_ALL_LOG_ID } from './logging/index.js';
+export type { Logger, LogTransport, LoggerOptions } from './logging/index.js';
+export { profiler, attachBackendDiagnostics } from './profiling/index.js';
+export type { ProfilerRuntimeOptions, ProfilerExport, ProfilingEvent, ProfilerGPUResourceGauge, ProfilerResourceGauge } from './profiling/index.js';
 //# sourceMappingURL=index.d.ts.map

package/dist/index.js

@@ -1,8 +1,8 @@
 // Public API exports only
-export { batchCompletion, completion, deleteCache, loadModel, downloadAsset, heartbeat, startQVACProvider, stopQVACProvider, unloadModel, transcribe, transcribeStream, bciTranscribe, bciTranscribeStream, embed, finetune, translate, cancel, ragChunk, ragIngest, ragSaveEmbeddings, ragSearch, ragDeleteEmbeddings, ragReindex, ragListWorkspaces, ragCloseWorkspace, ragDeleteWorkspace, textToSpeech, textToSpeechStream, getModelInfo, getLoadedModelInfo, loggingStream, subscribeServerLogs, ocr, invokePlugin, invokePluginStream, diffusion, classify, video, upscale, modelRegistryList, modelRegistrySearch, modelRegistryGetModel, suspend, resume, state, vla, vlaHparams, vlaPreprocessImage, vlaPadState, VLA_DEFAULT_IMAGE_SIZE } from './client/api/index.js';
+export { batchCompletion, completion, deleteCache, loadModel, downloadAsset, heartbeat, startQVACProvider, stopQVACProvider, unloadModel, transcribe, transcribeStream, bciTranscribe, bciTranscribeStream, embed, finetune, translate, cancel, ragChunk, ragIngest, ragSaveEmbeddings, ragSearch, ragDeleteEmbeddings, ragReindex, ragListWorkspaces, ragCloseWorkspace, ragDeleteWorkspace, textToSpeech, textToSpeechStream, getModelInfo, getLoadedModelInfo, getSystemResources, loggingStream, subscribeServerLogs, ocr, invokePlugin, invokePluginStream, diffusion, classify, video, upscale, modelRegistryList, modelRegistrySearch, modelRegistryGetModel, suspend, resume, state, vla, vlaHparams, vlaPreprocessImage, vlaPadState, VLA_DEFAULT_IMAGE_SIZE, audioGen } from './client/api/index.js';
 export { close } from './client/index.js';
 export { plugins } from './client/plugins-factory.js';
-export { VERBOSITY, SDK_CLIENT_ERROR_CODES, SDK_SERVER_ERROR_CODES, RAG_ERROR_CODES, toolSchema, TOOLS_MODE, definePlugin, defineHandler, defineDuplexHandler, PLUGIN_LLM, PLUGIN_EMBEDDING, PLUGIN_WHISPER, PLUGIN_BCI, PLUGIN_NMT, PLUGIN_TTS, PLUGIN_OCR, PLUGIN_DIFFUSION, PLUGIN_VLA, PLUGIN_CLASSIFICATION, SDK_DEFAULT_PLUGINS } from './schemas/index.js';
+export { VERBOSITY, SDK_CLIENT_ERROR_CODES, SDK_SERVER_ERROR_CODES, RAG_ERROR_CODES, toolSchema, TOOLS_MODE, definePlugin, defineHandler, defineDuplexHandler, PLUGIN_LLM, PLUGIN_EMBEDDING, PLUGIN_WHISPER, PLUGIN_BCI, PLUGIN_NMT, PLUGIN_TTS, PLUGIN_OCR, PLUGIN_DIFFUSION, PLUGIN_AUDIOGEN, PLUGIN_VLA, PLUGIN_CLASSIFICATION, SDK_DEFAULT_PLUGINS } from './schemas/index.js';
 export {} from './utils/tool-helpers.js';
 // Model types - canonical naming with backward-compatible aliases
 export { MODEL_TYPES, ModelType } from './schemas/index.js';
@@ -29,7 +29,7 @@ export { SUPPORTED_AUDIO_FORMATS } from './constants/audio.js';
 // `err instanceof <Class>` works on the consumer side, not just on
 // the worker side.
 export { InferenceCancelledError } from './utils/errors-server.js';
-export { ContextOverflowError, RequestIdConflictError, RequestNotFoundError, RequestRejectedByPolicyError } from './utils/errors-server.js';
+export { ContextOverflowError, RequestIdConflictError, RequestNotFoundError, RequestRejectedByPolicyError, TranslationFailedError } from './utils/errors-server.js';
 // `WorkerCrashedError` and `WorkerShutdownError` are thrown by the
 // rpc-client life-signal race when the bare worker exits unexpectedly
 // or close()/process-exit teardown runs while a caller is in flight.
@@ -41,4 +41,4 @@ export { BareRuntimeBinaryNotFoundError, WorkerCrashedError, WorkerShutdownError
 // Logging exports
 export { getLogger, SDK_LOG_ID, SDK_ALL_LOG_ID } from './logging/index.js';
 // Profiler exports
-export { profiler } from './profiling/index.js';
+export { profiler, attachBackendDiagnostics } from './profiling/index.js';

package/dist/logging/addon.d.ts

@@ -1,4 +1,4 @@
-import type { Logger } from './index';
+import type { Logger } from '../logging/index.js';
 export declare function registerAddonLogger(modelId: string, namespace: string, logger: Logger): void;
 export declare function unregisterAddonLogger(modelId: string): void;
 export declare function createAddonLoggerCallback(namespace: string): (priority: number, message: string) => void;

package/dist/logging/base-logger.d.ts

@@ -1,5 +1,5 @@
 import { type LogLevel } from '@qvac/logging';
-import type { Logger, LoggerOptions } from './types';
+import type { Logger, LoggerOptions } from './types.js';
 export interface LoggerExtensions {
     onLog?: (level: LogLevel, namespace: string, message: string) => void;
 }

package/dist/logging/base-logger.js

@@ -1,5 +1,5 @@
 import QvacLogger, {} from '@qvac/logging';
-import { LOG_LEVELS } from '@qvac/logging/constants.js';
+import { LOG_LEVELS } from '@qvac/logging/constants';
 import { isLevelEnabled, formatArg } from './utils.js';
 import { registerLogger } from './registry.js';
 export function createBaseLogger(namespace, options, extensions) {

package/dist/logging/client-logger.d.ts

@@ -1,3 +1,3 @@
-import type { Logger, LoggerOptions } from './types';
+import type { Logger, LoggerOptions } from './types.js';
 export declare function getClientLogger(options?: LoggerOptions): Logger;
 //# sourceMappingURL=client-logger.d.ts.map