@lancedb/lancedb 0.13.0-beta.1 → 0.13.1-beta.0

package/dist/embedding/index.d.ts

@@ -1,8 +1,6 @@
 import { Schema } from "../arrow";
 import { EmbeddingFunction } from "./embedding_function";
 export { EmbeddingFunction, TextEmbeddingFunction } from "./embedding_function";
-export * from "./openai";
-export * from "./transformers";
 export * from "./registry";
 /**
  * Create a schema with embedding functions.

package/dist/embedding/index.js

@@ -35,9 +35,6 @@ const registry_1 = require("./registry");
 var embedding_function_1 = require("./embedding_function");
 Object.defineProperty(exports, "EmbeddingFunction", { enumerable: true, get: function () { return embedding_function_1.EmbeddingFunction; } });
 Object.defineProperty(exports, "TextEmbeddingFunction", { enumerable: true, get: function () { return embedding_function_1.TextEmbeddingFunction; } });
-// We need to explicitly export '*' so that the `register` decorator actually registers the class.
-__exportStar(require("./openai"), exports);
-__exportStar(require("./transformers"), exports);
 __exportStar(require("./registry"), exports);
 /**
  * Create a schema with embedding functions.

package/dist/embedding/registry.d.ts

@@ -1,7 +1,5 @@
 import { type EmbeddingFunction, type EmbeddingFunctionConstructor } from "./embedding_function";
 import "reflect-metadata";
-import { OpenAIEmbeddingFunction } from "./openai";
-import { TransformersEmbeddingFunction } from "./transformers";
 type CreateReturnType<T> = T extends {
     init: () => Promise<void>;
 } ? Promise<T> : T;
@@ -27,8 +25,6 @@ export declare class EmbeddingFunctionRegistry {
      * @throws Error if the function is already registered
      */
     register<T extends EmbeddingFunctionConstructor = EmbeddingFunctionConstructor>(this: EmbeddingFunctionRegistry, alias?: string): (ctor: T) => any;
-    get(name: "openai"): EmbeddingFunctionCreate<OpenAIEmbeddingFunction>;
-    get(name: "huggingface"): EmbeddingFunctionCreate<TransformersEmbeddingFunction>;
     get<T extends EmbeddingFunction<unknown>>(name: string): EmbeddingFunctionCreate<T> | undefined;
     /**
      * reset the registry to the initial state

package/dist/embedding/transformers.js

@@ -66,17 +66,17 @@ let TransformersEmbeddingFunction = class TransformersEmbeddingFunction extends
         try {
             // SAFETY:
             // since typescript transpiles `import` to `require`, we need to do this in an unsafe way
-            // We can't use `require` because `@xenova/transformers` is an ESM module
+            // We can't use `require` because `@huggingface/transformers` is an ESM module
             // and we can't use `import` directly because typescript will transpile it to `require`.
             // and we want to remain compatible with both ESM and CJS modules
             // so we use `eval` to bypass typescript for this specific import.
-            transformers = await eval('import("@xenova/transformers")');
+            transformers = await eval('import("@huggingface/transformers")');
         }
         catch (e) {
-            throw new Error(`error loading @xenova/transformers\nReason: ${e}`);
+            throw new Error(`error loading @huggingface/transformers\nReason: ${e}`);
         }
         try {
-            this.#model = await transformers.AutoModel.from_pretrained(this.#modelName);
+            this.#model = await transformers.AutoModel.from_pretrained(this.#modelName, { dtype: "fp32" });
         }
         catch (e) {
             throw new Error(`error loading model ${this.#modelName}. Make sure you are using a wasm compatible model.\nReason: ${e}`);
@@ -95,7 +95,8 @@ let TransformersEmbeddingFunction = class TransformersEmbeddingFunction extends
         }
         else {
             const config = this.#model.config;
-            const ndims = config["hidden_size"];
+            // biome-ignore lint/style/useNamingConvention: we don't control this name.
+            const ndims = config.hidden_size;
             if (!ndims) {
                 throw new Error("hidden_size not found in model config, you may need to manually specify the embedding dimensions. ");
             }

package/dist/native.d.ts

@@ -165,6 +165,11 @@ export interface IndexStatistics {
   /** The number of parts this index is split into. */
   numIndices?: number
 }
+export interface Version {
+  version: number
+  timestamp: number
+  metadata: Record<string, string>
+}
 export interface ConnectionOptions {
   /**
    * (For LanceDB OSS only): The interval, in seconds, at which to check for
@@ -274,10 +279,12 @@ export class Query {
 }
 export class VectorQuery {
   column(column: string): void
+  addQueryVector(vector: Float32Array): void
   distanceType(distanceType: string): void
   postfilter(): void
   refineFactor(refineFactor: number): void
   nprobes(nprobe: number): void
+  ef(ef: number): void
   bypassVectorIndex(): void
   onlyIf(predicate: string): void
   fullTextSearch(query: string, columns?: Array<string> | undefined | null): void
@@ -310,6 +317,7 @@ export class Table {
   version(): Promise<number>
   checkout(version: number): Promise<void>
   checkoutLatest(): Promise<void>
+  listVersions(): Promise<Array<Version>>
   restore(): Promise<void>
   optimize(olderThanMs?: number | undefined | null, deleteUnverified?: boolean | undefined | null): Promise<OptimizeStats>
   listIndices(): Promise<Array<IndexConfig>>

package/dist/query.d.ts

@@ -181,6 +181,16 @@ export declare class VectorQuery extends QueryBase<NativeVectorQuery> {
      * you the desired recall.
      */
     nprobes(nprobes: number): VectorQuery;
+    /**
+     * Set the number of candidates to consider during the search
+     *
+     * This argument is only used when the vector column has an HNSW index.
+     * If there is no index then this value is ignored.
+     *
+     * Increasing this value will increase the recall of your query but will
+     * also increase the latency of your query. The default value is 1.5*limit.
+     */
+    ef(ef: number): VectorQuery;
     /**
      * Set the vector column to query
      *
@@ -267,6 +277,7 @@ export declare class VectorQuery extends QueryBase<NativeVectorQuery> {
      * calculate your recall to select an appropriate value for nprobes.
      */
     bypassVectorIndex(): VectorQuery;
+    addQueryVector(vector: IntoVector): VectorQuery;
 }
 /** A builder for LanceDB queries. */
 export declare class Query extends QueryBase<NativeQuery> {
@@ -309,4 +320,5 @@ export declare class Query extends QueryBase<NativeQuery> {
      * a default `limit` of 10 will be used.  @see {@link Query#limit}
      */
     nearestTo(vector: IntoVector): VectorQuery;
+    nearestToText(query: string, columns?: string[]): Query;
 }

package/dist/query.js

@@ -308,6 +308,19 @@ class VectorQuery extends QueryBase {
         super.doCall((inner) => inner.nprobes(nprobes));
         return this;
     }
+    /**
+     * Set the number of candidates to consider during the search
+     *
+     * This argument is only used when the vector column has an HNSW index.
+     * If there is no index then this value is ignored.
+     *
+     * Increasing this value will increase the recall of your query but will
+     * also increase the latency of your query. The default value is 1.5*limit.
+     */
+    ef(ef) {
+        super.doCall((inner) => inner.ef(ef));
+        return this;
+    }
     /**
      * Set the vector column to query
      *
@@ -409,6 +422,41 @@ class VectorQuery extends QueryBase {
         super.doCall((inner) => inner.bypassVectorIndex());
         return this;
     }
+    /*
+     * Add a query vector to the search
+     *
+     * This method can be called multiple times to add multiple query vectors
+     * to the search. If multiple query vectors are added, then they will be searched
+     * in parallel, and the results will be concatenated. A column called `query_index`
+     * will be added to indicate the index of the query vector that produced the result.
+     *
+     * Performance wise, this is equivalent to running multiple queries concurrently.
+     */
+    addQueryVector(vector) {
+        if (vector instanceof Promise) {
+            const res = (async () => {
+                try {
+                    const v = await vector;
+                    const arr = Float32Array.from(v);
+                    //
+                    // biome-ignore lint/suspicious/noExplicitAny: we need to get the `inner`, but js has no package scoping
+                    const value = this.addQueryVector(arr);
+                    const inner = value.inner;
+                    return inner;
+                }
+                catch (e) {
+                    return Promise.reject(e);
+                }
+            })();
+            return new VectorQuery(res);
+        }
+        else {
+            super.doCall((inner) => {
+                inner.addQueryVector(Float32Array.from(vector));
+            });
+            return this;
+        }
+    }
 }
 exports.VectorQuery = VectorQuery;
 /** A builder for LanceDB queries. */
@@ -488,5 +536,9 @@ class Query extends QueryBase {
             return new VectorQuery(vectorQuery);
         }
     }
+    nearestToText(query, columns) {
+        this.doCall((inner) => inner.fullTextSearch(query, columns));
+        return this;
+    }
 }
 exports.Query = Query;

package/dist/table.d.ts

@@ -47,6 +47,11 @@ export interface OptimizeOptions {
     cleanupOlderThan: Date;
     deleteUnverified: boolean;
 }
+export interface Version {
+    version: number;
+    timestamp: Date;
+    metadata: Record<string, string>;
+}
 /**
  * A Table is a collection of Records in a LanceDB Database.
  *
@@ -297,6 +302,10 @@ export declare abstract class Table {
      * version of the table.
      */
     abstract checkoutLatest(): Promise<void>;
+    /**
+     * List all the versions of the table
+     */
+    abstract listVersions(): Promise<Version[]>;
     /**
      * Restore the table to the currently checked out version
      *
@@ -385,6 +394,7 @@ export declare class LocalTable extends Table {
     version(): Promise<number>;
     checkout(version: number): Promise<void>;
     checkoutLatest(): Promise<void>;
+    listVersions(): Promise<Version[]>;
     restore(): Promise<void>;
     optimize(options?: Partial<OptimizeOptions>): Promise<OptimizeStats>;
     listIndices(): Promise<IndexConfig[]>;

package/dist/table.js

@@ -211,6 +211,13 @@ class LocalTable extends Table {
     async checkoutLatest() {
         await this.inner.checkoutLatest();
     }
+    async listVersions() {
+        return (await this.inner.listVersions()).map((version) => ({
+            version: version.version,
+            timestamp: new Date(version.timestamp / 1000),
+            metadata: version.metadata,
+        }));
+    }
     async restore() {
         await this.inner.restore();
     }

package/package.json

@@ -10,11 +10,13 @@
     "vector database",
     "ann"
   ],
-  "version": "0.13.0-beta.1",
+  "version": "0.13.1-beta.0",
   "main": "dist/index.js",
   "exports": {
     ".": "./dist/index.js",
-    "./embedding": "./dist/embedding/index.js"
+    "./embedding": "./dist/embedding/index.js",
+    "./embedding/openai": "./dist/embedding/openai.js",
+    "./embedding/transformers": "./dist/embedding/transformers.js"
   },
   "types": "dist/index.d.ts",
   "napi": {
@@ -22,10 +24,12 @@
     "triples": {
       "defaults": false,
       "additional": [
-        "aarch64-apple-darwin",
-        "aarch64-unknown-linux-gnu",
         "x86_64-apple-darwin",
+        "aarch64-apple-darwin",
         "x86_64-unknown-linux-gnu",
+        "aarch64-unknown-linux-gnu",
+        "x86_64-unknown-linux-musl",
+        "aarch64-unknown-linux-musl",
         "x86_64-pc-windows-msvc"
       ]
     }
@@ -92,11 +96,13 @@
     "reflect-metadata": "^0.2.2"
   },
   "optionalDependencies": {
-    "@lancedb/lancedb-darwin-arm64": "0.13.0-beta.1",
-    "@lancedb/lancedb-linux-arm64-gnu": "0.13.0-beta.1",
-    "@lancedb/lancedb-darwin-x64": "0.13.0-beta.1",
-    "@lancedb/lancedb-linux-x64-gnu": "0.13.0-beta.1",
-    "@lancedb/lancedb-win32-x64-msvc": "0.13.0-beta.1"
+    "@lancedb/lancedb-darwin-x64": "0.13.1-beta.0",
+    "@lancedb/lancedb-darwin-arm64": "0.13.1-beta.0",
+    "@lancedb/lancedb-linux-x64-gnu": "0.13.1-beta.0",
+    "@lancedb/lancedb-linux-arm64-gnu": "0.13.1-beta.0",
+    "@lancedb/lancedb-linux-x64-musl": "0.13.1-beta.0",
+    "@lancedb/lancedb-linux-arm64-musl": "0.13.1-beta.0",
+    "@lancedb/lancedb-win32-x64-msvc": "0.13.1-beta.0"
   },
   "peerDependencies": {
     "apache-arrow": ">=13.0.0 <=17.0.0"