@ruvector/edge-full 0.1.0

package/package.json

@@ -0,0 +1,86 @@
+{
+  "name": "@ruvector/edge-full",
+  "version": "0.1.0",
+  "type": "module",
+  "description": "Complete WASM toolkit for edge AI: vector search, graph DB, neural networks, DAG workflows, SQL/SPARQL/Cypher, and ONNX inference - all running in browser",
+  "main": "index.js",
+  "module": "index.js",
+  "types": "index.d.ts",
+  "keywords": [
+    "wasm",
+    "rust",
+    "ai",
+    "edge",
+    "browser",
+    "vector-search",
+    "hnsw",
+    "graph-database",
+    "cypher",
+    "sparql",
+    "sql",
+    "neural-network",
+    "onnx",
+    "embeddings",
+    "dag",
+    "workflow",
+    "post-quantum",
+    "cryptography",
+    "consensus",
+    "raft",
+    "p2p",
+    "swarm"
+  ],
+  "author": "RuVector Team",
+  "license": "MIT",
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/ruvnet/ruvector"
+  },
+  "homepage": "https://github.com/ruvnet/ruvector/tree/main/examples/edge-full",
+  "bugs": {
+    "url": "https://github.com/ruvnet/ruvector/issues"
+  },
+  "files": [
+    "index.js",
+    "index.d.ts",
+    "edge/",
+    "graph/",
+    "rvlite/",
+    "sona/",
+    "dag/",
+    "onnx/",
+    "generator.html",
+    "LICENSE"
+  ],
+  "exports": {
+    ".": {
+      "import": "./index.js",
+      "types": "./index.d.ts"
+    },
+    "./edge": {
+      "import": "./edge/ruvector_edge.js",
+      "types": "./edge/ruvector_edge.d.ts"
+    },
+    "./graph": {
+      "import": "./graph/ruvector_graph_wasm.js",
+      "types": "./graph/ruvector_graph_wasm.d.ts"
+    },
+    "./rvlite": {
+      "import": "./rvlite/rvlite.js",
+      "types": "./rvlite/rvlite.d.ts"
+    },
+    "./sona": {
+      "import": "./sona/ruvector_sona.js",
+      "types": "./sona/ruvector_sona.d.ts"
+    },
+    "./dag": {
+      "import": "./dag/ruvector_dag_wasm.js",
+      "types": "./dag/ruvector_dag_wasm.d.ts"
+    },
+    "./onnx": {
+      "import": "./onnx/ruvector_onnx_embeddings_wasm.js",
+      "types": "./onnx/ruvector_onnx_embeddings_wasm.d.ts"
+    }
+  },
+  "sideEffects": false
+}

package/rvlite/rvlite.d.ts

@@ -0,0 +1,556 @@
+/* tslint:disable */
+/* eslint-disable */
+
+export class BaseLoRA {
+  free(): void;
+  [Symbol.dispose](): void;
+  distillFrom(micro: MicroLoRA, blend_factor: number): void;
+  applyGradients(): void;
+  constructor(config: LoraConfig);
+  stats(): any;
+  forward(input: Float32Array): Float32Array;
+}
+
+export class Embedder {
+  free(): void;
+  [Symbol.dispose](): void;
+  /**
+   * Get embedding dimensions
+   */
+  dimensions(): number;
+  /**
+   * Compute similarity between two texts
+   */
+  similarity(text_a: string, text_b: string): number;
+  /**
+   * Batch embed multiple texts
+   * Takes JsValue array of strings, returns JsValue array of Float32Arrays
+   */
+  embed_batch(texts: any): any;
+  /**
+   * Create embedder with custom config
+   */
+  static with_config(config: EmbeddingConfig): Embedder;
+  /**
+   * Compute cosine similarity between two embeddings (JS arrays)
+   */
+  static cosine_similarity(a: Float32Array, b: Float32Array): number;
+  /**
+   * Create a new embedder with default config
+   */
+  constructor();
+  /**
+   * Generate embeddings for a single text
+   * Uses hash-based projection for lightweight WASM operation
+   * Returns Float32Array for direct JS consumption
+   */
+  embed(text: string): Float32Array;
+}
+
+export class EmbeddingConfig {
+  free(): void;
+  [Symbol.dispose](): void;
+  /**
+   * Create config for larger models
+   */
+  static with_dimensions(dimensions: number): EmbeddingConfig;
+  constructor();
+  /**
+   * Create config for all-MiniLM-L6-v2 (default)
+   */
+  static minilm(): EmbeddingConfig;
+  /**
+   * Model dimensions (384 for all-MiniLM-L6-v2)
+   */
+  dimensions: number;
+  /**
+   * Normalize output vectors
+   */
+  normalize: boolean;
+  /**
+   * Max sequence length
+   */
+  max_length: number;
+}
+
+export class LoraConfig {
+  free(): void;
+  [Symbol.dispose](): void;
+  /**
+   * Create custom LoRA configuration
+   */
+  constructor(hidden_dim: number, rank: number, alpha: number, learning_rate: number);
+  /**
+   * Create BaseLoRA configuration (background training)
+   */
+  static base(hidden_dim: number): LoraConfig;
+  /**
+   * Create MicroLoRA configuration (per-request, <100μs)
+   */
+  static micro(hidden_dim: number): LoraConfig;
+  /**
+   * Export as JSON
+   */
+  toJSON(): any;
+  /**
+   * Validate configuration
+   */
+  validate(): void;
+  /**
+   * LoRA rank (1-2 for micro, 4-16 for base)
+   */
+  readonly rank: number;
+  /**
+   * Alpha scaling factor
+   */
+  readonly alpha: number;
+  /**
+   * Learning rate for adaptation
+   */
+  readonly learning_rate: number;
+  /**
+   * Hidden dimension
+   */
+  readonly hidden_dim: number;
+}
+
+export class MicroLoRA {
+  free(): void;
+  [Symbol.dispose](): void;
+  exportWeights(): any;
+  applyGradients(): void;
+  accumulateGradient(input: Float32Array, feedback: number): void;
+  constructor(config: LoraConfig);
+  reset(): void;
+  stats(): any;
+  forward(input: Float32Array): Float32Array;
+}
+
+export class RvLite {
+  free(): void;
+  [Symbol.dispose](): void;
+  /**
+   * Get configuration
+   */
+  get_config(): any;
+  /**
+   * Get version string
+   */
+  get_version(): string;
+  /**
+   * Get enabled features
+   */
+  get_features(): any;
+  /**
+   * Insert a vector with a specific ID
+   */
+  insert_with_id(id: string, vector: Float32Array, metadata?: any | null): void;
+  /**
+   * Search with metadata filter
+   */
+  search_with_filter(query_vector: Float32Array, k: number, filter: any): any;
+  /**
+   * Get a vector by ID
+   */
+  get(id: string): any;
+  /**
+   * Get the number of vectors in the database
+   */
+  len(): number;
+  /**
+   * Create a new RvLite database
+   */
+  constructor(config: RvLiteConfig);
+  /**
+   * Execute SQL query (not yet implemented)
+   */
+  sql(_query: string): Promise<any>;
+  /**
+   * Execute Cypher query (not yet implemented)
+   */
+  cypher(_query: string): Promise<any>;
+  /**
+   * Delete a vector by ID
+   */
+  delete(id: string): boolean;
+  /**
+   * Insert a vector with optional metadata
+   * Returns the vector ID
+   */
+  insert(vector: Float32Array, metadata?: any | null): string;
+  /**
+   * Search for similar vectors
+   * Returns a JavaScript array of search results
+   */
+  search(query_vector: Float32Array, k: number): any;
+  /**
+   * Execute SPARQL query (not yet implemented)
+   */
+  sparql(_query: string): Promise<any>;
+  /**
+   * Create with default configuration (384 dimensions, cosine similarity)
+   */
+  static default(): RvLite;
+  /**
+   * Check if database is empty
+   */
+  is_empty(): boolean;
+  /**
+   * Check if database is ready
+   */
+  is_ready(): boolean;
+}
+
+export class RvLiteConfig {
+  free(): void;
+  [Symbol.dispose](): void;
+  /**
+   * Set distance metric (euclidean, cosine, dotproduct, manhattan)
+   */
+  with_distance_metric(metric: string): RvLiteConfig;
+  constructor(dimensions: number);
+}
+
+export class TrmConfig {
+  free(): void;
+  [Symbol.dispose](): void;
+  /**
+   * Enable/disable attention variant
+   */
+  withAttention(use_attention: boolean): TrmConfig;
+  /**
+   * Set default K iterations
+   */
+  withDefaultK(k: number): TrmConfig;
+  /**
+   * Enable/disable early stopping
+   */
+  withEarlyStopping(enabled: boolean): TrmConfig;
+  /**
+   * Set latent iterations per K step
+   */
+  withLatentIterations(n: number): TrmConfig;
+  /**
+   * Set confidence threshold
+   */
+  withConfidenceThreshold(threshold: number): TrmConfig;
+  /**
+   * Create a new TRM configuration
+   */
+  constructor(embedding_dim: number, hidden_dim: number, max_k: number);
+  /**
+   * Create configuration optimized for speed
+   */
+  static fast(embedding_dim: number): TrmConfig;
+  /**
+   * Create configuration optimized for quality
+   */
+  static quality(embedding_dim: number): TrmConfig;
+  /**
+   * Export configuration as JSON
+   */
+  toJSON(): any;
+  /**
+   * Create balanced configuration (recommended)
+   */
+  static balanced(embedding_dim: number): TrmConfig;
+  /**
+   * Validate configuration
+   */
+  validate(): void;
+  /**
+   * Import configuration from JSON
+   */
+  static fromJSON(value: any): TrmConfig;
+  /**
+   * Embedding dimension (input/output size)
+   */
+  readonly embedding_dim: number;
+  /**
+   * Hidden dimension for latent state
+   */
+  readonly hidden_dim: number;
+  /**
+   * Maximum K iterations
+   */
+  readonly max_k: number;
+  /**
+   * Default K iterations
+   */
+  readonly default_k: number;
+  /**
+   * Latent updates per K iteration
+   */
+  readonly latent_iterations: number;
+  /**
+   * Use attention variant (more expressive, slower)
+   */
+  readonly use_attention: boolean;
+  /**
+   * Number of attention heads
+   */
+  readonly num_heads: number;
+  /**
+   * Confidence threshold for early stopping
+   */
+  readonly confidence_threshold: number;
+  /**
+   * Enable early stopping
+   */
+  readonly early_stopping: boolean;
+  /**
+   * Minimum iterations before early stopping
+   */
+  readonly min_iterations: number;
+  /**
+   * Convergence threshold for plateau detection
+   */
+  readonly convergence_threshold: number;
+  /**
+   * Residual scale for answer refinement
+   */
+  readonly residual_scale: number;
+}
+
+export class TrmEngine {
+  free(): void;
+  [Symbol.dispose](): void;
+  getConfig(): any;
+  reasonWithK(question: Float32Array, answer: Float32Array, k: number): TrmResult;
+  constructor(config: TrmConfig);
+  reset(): void;
+  stats(): any;
+  reason(question: Float32Array, answer: Float32Array): TrmResult;
+}
+
+export class TrmResult {
+  private constructor();
+  free(): void;
+  [Symbol.dispose](): void;
+  getAnswer(): Float32Array;
+  toJSON(): any;
+  readonly confidence: number;
+  readonly iterations_used: number;
+  readonly early_stopped: boolean;
+  readonly latency_ms: number;
+}
+
+/**
+ * Quick benchmark for embeddings
+ */
+export function benchmark_embeddings(iterations: number): any;
+
+/**
+ * Quick benchmark function
+ */
+export function benchmark_trm(iterations: number, hidden_dim: number): any;
+
+/**
+ * Compute cosine similarity
+ */
+export function cosineSimilarity(a: Float32Array, b: Float32Array): number;
+
+/**
+ * Dot product
+ */
+export function dotProduct(a: Float32Array, b: Float32Array): number;
+
+/**
+ * Get feature info
+ */
+export function features(): any;
+
+export function init(): void;
+
+/**
+ * Compute L2 distance
+ */
+export function l2Distance(a: Float32Array, b: Float32Array): number;
+
+/**
+ * Linear interpolation
+ */
+export function lerp(a: Float32Array, b: Float32Array, t: number): Float32Array;
+
+/**
+ * Mean pooling for token embeddings
+ */
+export function meanPooling(embeddings: any, attention_mask?: Float32Array | null): Float32Array;
+
+/**
+ * Normalize a vector to unit length
+ */
+export function normalizeVector(vec: Float32Array): Float32Array;
+
+/**
+ * Create a random vector
+ */
+export function randomVector(dim: number, seed?: number | null): Float32Array;
+
+/**
+ * Softmax function
+ */
+export function softmax(vec: Float32Array): Float32Array;
+
+/**
+ * Get version string
+ */
+export function version(): string;
+
+/**
+ * Create a zero vector
+ */
+export function zeros(dim: number): Float32Array;
+
+export type InitInput = RequestInfo | URL | Response | BufferSource | WebAssembly.Module;
+
+export interface InitOutput {
+  readonly memory: WebAssembly.Memory;
+  readonly __wbg_baselora_free: (a: number, b: number) => void;
+  readonly __wbg_embedder_free: (a: number, b: number) => void;
+  readonly __wbg_embeddingconfig_free: (a: number, b: number) => void;
+  readonly __wbg_get_embeddingconfig_dimensions: (a: number) => number;
+  readonly __wbg_get_embeddingconfig_max_length: (a: number) => number;
+  readonly __wbg_get_embeddingconfig_normalize: (a: number) => number;
+  readonly __wbg_get_loraconfig_alpha: (a: number) => number;
+  readonly __wbg_get_loraconfig_hidden_dim: (a: number) => number;
+  readonly __wbg_get_loraconfig_learning_rate: (a: number) => number;
+  readonly __wbg_get_trmconfig_confidence_threshold: (a: number) => number;
+  readonly __wbg_get_trmconfig_convergence_threshold: (a: number) => number;
+  readonly __wbg_get_trmconfig_early_stopping: (a: number) => number;
+  readonly __wbg_get_trmconfig_latent_iterations: (a: number) => number;
+  readonly __wbg_get_trmconfig_max_k: (a: number) => number;
+  readonly __wbg_get_trmconfig_min_iterations: (a: number) => number;
+  readonly __wbg_get_trmconfig_num_heads: (a: number) => number;
+  readonly __wbg_get_trmconfig_residual_scale: (a: number) => number;
+  readonly __wbg_get_trmconfig_use_attention: (a: number) => number;
+  readonly __wbg_get_trmresult_confidence: (a: number) => number;
+  readonly __wbg_get_trmresult_early_stopped: (a: number) => number;
+  readonly __wbg_get_trmresult_iterations_used: (a: number) => number;
+  readonly __wbg_get_trmresult_latency_ms: (a: number) => number;
+  readonly __wbg_loraconfig_free: (a: number, b: number) => void;
+  readonly __wbg_microlora_free: (a: number, b: number) => void;
+  readonly __wbg_rvlite_free: (a: number, b: number) => void;
+  readonly __wbg_rvliteconfig_free: (a: number, b: number) => void;
+  readonly __wbg_set_embeddingconfig_dimensions: (a: number, b: number) => void;
+  readonly __wbg_set_embeddingconfig_max_length: (a: number, b: number) => void;
+  readonly __wbg_set_embeddingconfig_normalize: (a: number, b: number) => void;
+  readonly __wbg_trmconfig_free: (a: number, b: number) => void;
+  readonly __wbg_trmengine_free: (a: number, b: number) => void;
+  readonly __wbg_trmresult_free: (a: number, b: number) => void;
+  readonly baselora_applyGradients: (a: number) => void;
+  readonly baselora_distillFrom: (a: number, b: number, c: number, d: number) => void;
+  readonly baselora_forward: (a: number, b: number, c: number, d: number) => void;
+  readonly baselora_new: (a: number, b: number) => void;
+  readonly baselora_stats: (a: number) => number;
+  readonly benchmark_embeddings: (a: number) => number;
+  readonly benchmark_trm: (a: number, b: number) => number;
+  readonly cosineSimilarity: (a: number, b: number, c: number, d: number, e: number) => void;
+  readonly dotProduct: (a: number, b: number, c: number, d: number, e: number) => void;
+  readonly embedder_cosine_similarity: (a: number, b: number) => number;
+  readonly embedder_dimensions: (a: number) => number;
+  readonly embedder_embed: (a: number, b: number, c: number) => number;
+  readonly embedder_embed_batch: (a: number, b: number, c: number) => void;
+  readonly embedder_new: () => number;
+  readonly embedder_similarity: (a: number, b: number, c: number, d: number, e: number) => number;
+  readonly embedder_with_config: (a: number) => number;
+  readonly embeddingconfig_minilm: () => number;
+  readonly embeddingconfig_with_dimensions: (a: number) => number;
+  readonly features: () => number;
+  readonly init: () => void;
+  readonly l2Distance: (a: number, b: number, c: number, d: number, e: number) => void;
+  readonly lerp: (a: number, b: number, c: number, d: number, e: number, f: number) => void;
+  readonly loraconfig_base: (a: number) => number;
+  readonly loraconfig_micro: (a: number) => number;
+  readonly loraconfig_new: (a: number, b: number, c: number, d: number) => number;
+  readonly loraconfig_toJSON: (a: number) => number;
+  readonly loraconfig_validate: (a: number, b: number) => void;
+  readonly meanPooling: (a: number, b: number, c: number, d: number) => void;
+  readonly microlora_accumulateGradient: (a: number, b: number, c: number, d: number, e: number) => void;
+  readonly microlora_applyGradients: (a: number) => void;
+  readonly microlora_exportWeights: (a: number) => number;
+  readonly microlora_forward: (a: number, b: number, c: number, d: number) => void;
+  readonly microlora_new: (a: number, b: number) => void;
+  readonly microlora_reset: (a: number) => void;
+  readonly microlora_stats: (a: number) => number;
+  readonly normalizeVector: (a: number, b: number, c: number) => void;
+  readonly randomVector: (a: number, b: number, c: number) => void;
+  readonly rvlite_cypher: (a: number, b: number, c: number) => number;
+  readonly rvlite_default: (a: number) => void;
+  readonly rvlite_delete: (a: number, b: number, c: number, d: number) => void;
+  readonly rvlite_get: (a: number, b: number, c: number, d: number) => void;
+  readonly rvlite_get_config: (a: number, b: number) => void;
+  readonly rvlite_get_features: (a: number, b: number) => void;
+  readonly rvlite_get_version: (a: number, b: number) => void;
+  readonly rvlite_insert: (a: number, b: number, c: number, d: number, e: number) => void;
+  readonly rvlite_insert_with_id: (a: number, b: number, c: number, d: number, e: number, f: number, g: number) => void;
+  readonly rvlite_is_empty: (a: number, b: number) => void;
+  readonly rvlite_is_ready: (a: number) => number;
+  readonly rvlite_len: (a: number, b: number) => void;
+  readonly rvlite_new: (a: number, b: number) => void;
+  readonly rvlite_search: (a: number, b: number, c: number, d: number, e: number) => void;
+  readonly rvlite_search_with_filter: (a: number, b: number, c: number, d: number, e: number, f: number) => void;
+  readonly rvlite_sparql: (a: number, b: number, c: number) => number;
+  readonly rvlite_sql: (a: number, b: number, c: number) => number;
+  readonly rvliteconfig_new: (a: number) => number;
+  readonly rvliteconfig_with_distance_metric: (a: number, b: number, c: number) => number;
+  readonly softmax: (a: number, b: number, c: number) => void;
+  readonly trmconfig_balanced: (a: number) => number;
+  readonly trmconfig_fast: (a: number) => number;
+  readonly trmconfig_fromJSON: (a: number, b: number) => void;
+  readonly trmconfig_new: (a: number, b: number, c: number) => number;
+  readonly trmconfig_quality: (a: number) => number;
+  readonly trmconfig_toJSON: (a: number) => number;
+  readonly trmconfig_validate: (a: number, b: number) => void;
+  readonly trmconfig_withAttention: (a: number, b: number) => number;
+  readonly trmconfig_withConfidenceThreshold: (a: number, b: number) => number;
+  readonly trmconfig_withDefaultK: (a: number, b: number) => number;
+  readonly trmconfig_withEarlyStopping: (a: number, b: number) => number;
+  readonly trmconfig_withLatentIterations: (a: number, b: number) => number;
+  readonly trmengine_getConfig: (a: number) => number;
+  readonly trmengine_new: (a: number, b: number) => void;
+  readonly trmengine_reason: (a: number, b: number, c: number, d: number, e: number, f: number) => void;
+  readonly trmengine_reasonWithK: (a: number, b: number, c: number, d: number, e: number, f: number, g: number) => void;
+  readonly trmengine_reset: (a: number) => void;
+  readonly trmengine_stats: (a: number) => number;
+  readonly trmresult_getAnswer: (a: number, b: number) => void;
+  readonly trmresult_toJSON: (a: number) => number;
+  readonly version: (a: number) => void;
+  readonly zeros: (a: number, b: number) => void;
+  readonly embeddingconfig_new: () => number;
+  readonly __wbg_get_trmconfig_embedding_dim: (a: number) => number;
+  readonly __wbg_get_loraconfig_rank: (a: number) => number;
+  readonly __wbg_get_trmconfig_hidden_dim: (a: number) => number;
+  readonly __wbg_get_trmconfig_default_k: (a: number) => number;
+  readonly __wasm_bindgen_func_elem_690: (a: number, b: number, c: number) => void;
+  readonly __wasm_bindgen_func_elem_684: (a: number, b: number) => void;
+  readonly __wasm_bindgen_func_elem_1412: (a: number, b: number, c: number, d: number) => void;
+  readonly __wbindgen_export: (a: number, b: number) => number;
+  readonly __wbindgen_export2: (a: number, b: number, c: number, d: number) => number;
+  readonly __wbindgen_export3: (a: number) => void;
+  readonly __wbindgen_export4: (a: number, b: number, c: number) => void;
+  readonly __wbindgen_add_to_stack_pointer: (a: number) => number;
+  readonly __wbindgen_start: () => void;
+}
+
+export type SyncInitInput = BufferSource | WebAssembly.Module;
+
+/**
+* Instantiates the given `module`, which can either be bytes or
+* a precompiled `WebAssembly.Module`.
+*
+* @param {{ module: SyncInitInput }} module - Passing `SyncInitInput` directly is deprecated.
+*
+* @returns {InitOutput}
+*/
+export function initSync(module: { module: SyncInitInput } | SyncInitInput): InitOutput;
+
+/**
+* If `module_or_path` is {RequestInfo} or {URL}, makes a request and
+* for everything else, calls `WebAssembly.instantiate` directly.
+*
+* @param {{ module_or_path: InitInput | Promise<InitInput> }} module_or_path - Passing `InitInput` directly is deprecated.
+*
+* @returns {Promise<InitOutput>}
+*/
+export default function __wbg_init (module_or_path?: { module_or_path: InitInput | Promise<InitInput> } | InitInput | Promise<InitInput>): Promise<InitOutput>;