Make minor changes to fix reported bugs and add missing features (#352)

* Make changes to fix reported bugs and to close the following issues:
- #325
- #328
- #345
- #346
- #349
- #350
- #350

* Differentiate between RQ-bits in 1.32 and 1.33

* Update CI images

* Allow runtime generation with images and cohere module

* Fix names in tests

* Skip failing test due to server regression

* Fix docstring for alias.update

Author: tsmith023

.github/workflows/main.yaml

@@ -10,10 +10,10 @@ env:
   WEAVIATE_127: 1.27.27
   WEAVIATE_128: 1.28.16
   WEAVIATE_129: 1.29.9
-  WEAVIATE_130: 1.30.12
-  WEAVIATE_131: 1.31.5
-  WEAVIATE_132: 1.32.5
-  WEAVIATE_133: 1.33.0-rc.1
+  WEAVIATE_130: 1.30.17
+  WEAVIATE_131: 1.31.16
+  WEAVIATE_132: 1.32.10
+  WEAVIATE_133: 1.33.0
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}

src/alias/index.ts

@@ -38,7 +38,7 @@ export interface Aliases {
    * delete the alias and create a new one.
    *
    * @param {string} args.alias Alias to update.
-   * @param {string} args.collection New collection the alias should point to.
+   * @param {string} args.newTargetCollection New collection the alias should point to.
    * @return {Promise<void>} Awaitable promise.
    */
   update: (args: UpdateAliasArgs) => Promise<void>;

src/collections/backup/integration.test.ts

@@ -118,7 +118,7 @@ describe('Integration testing of backups', () => {
       .then(testCollectionWaitForCompletion)
       .then(testCollectionNoWaitForCompletion));
 
-  requireAtLeast(1, 32, 3).describe('overwrite alias', () => {
+  requireAtLeast(1, 32, 0).describe('overwrite alias', () => {
     test('overwriteAlias=true', async () => {
       const client = await clientPromise;
 
@@ -153,7 +153,8 @@ describe('Integration testing of backups', () => {
       expect(alias.collection).toEqual(things.name);
     });
 
-    test('overwriteAlias=false', async () => {
+    // Skip until server regression is fixed for overwriteAlias=false backups
+    test.skip('overwriteAlias=false', async () => {
       const client = await clientPromise;
 
       const things = await client.collections.create({ name: 'ThingsFalse' });

src/collections/config/integration.test.ts

@@ -8,6 +8,7 @@ import {
   ModuleConfig,
   MultiTenancyConfig,
   PropertyConfig,
+  RQConfig,
   RerankerCohereConfig,
   VectorIndexConfigDynamic,
   VectorIndexConfigHNSW,
@@ -877,4 +878,46 @@ describe('Testing of the collection.config namespace', () => {
         );
     });
   });
+
+  requireAtLeast(1, 32, 0).it(
+    'should be able to create a collection with RQ quantizer bits=8 option',
+    async () => {
+      const collectionName = 'TestCollectionRQQuantizer8Bits';
+      const collection = await client.collections.create({
+        name: collectionName,
+        vectorizers: weaviate.configure.vectors.selfProvided({
+          quantizer: weaviate.configure.vectorIndex.quantizer.rq({ bits: 8, rescoreLimit: 10 }),
+        }),
+      });
+      await collection.config.get().then((config) => {
+        console.log(JSON.stringify(config, null, 2));
+        const indexConfig = config.vectorizers.default.indexConfig as VectorIndexConfigHNSW;
+        expect(indexConfig.quantizer).toBeDefined();
+        expect(indexConfig.quantizer?.type).toEqual('rq');
+        expect((indexConfig.quantizer as RQConfig).bits).toEqual(8);
+        expect((indexConfig.quantizer as RQConfig).rescoreLimit).toEqual(10);
+      });
+    }
+  );
+
+  requireAtLeast(1, 33, 0).it(
+    'should be able to create a collection with RQ quantizer bits=1 option',
+    async () => {
+      const collectionName = 'TestCollectionRQQuantizer1Bits';
+      const collection = await client.collections.create({
+        name: collectionName,
+        vectorizers: weaviate.configure.vectors.selfProvided({
+          quantizer: weaviate.configure.vectorIndex.quantizer.rq({ bits: 1, rescoreLimit: 10 }),
+        }),
+      });
+      await collection.config.get().then((config) => {
+        console.log(JSON.stringify(config, null, 2));
+        const indexConfig = config.vectorizers.default.indexConfig as VectorIndexConfigHNSW;
+        expect(indexConfig.quantizer).toBeDefined();
+        expect(indexConfig.quantizer?.type).toEqual('rq');
+        expect((indexConfig.quantizer as RQConfig).bits).toEqual(1);
+        expect((indexConfig.quantizer as RQConfig).rescoreLimit).toEqual(10);
+      });
+    }
+  );
 });

src/collections/config/types/generative.ts

@@ -12,6 +12,7 @@ export type GenerativeAWSConfig = {
   service: string;
   model?: string;
   endpoint?: string;
+  maxTokens?: number;
 };
 
 export type GenerativeAnthropicConfig = {

src/collections/config/types/vectorizer.ts

@@ -40,10 +40,12 @@ export type Vectorizer =
   | 'text2vec-nvidia'
   | 'text2vec-mistral'
   | 'text2vec-model2vec'
+  | 'text2vec-morph'
   | 'text2vec-ollama'
   | 'text2vec-openai'
   | Text2VecPalmVectorizer
   | 'text2vec-google'
+  | 'text2vec-google-ai-studio'
   | 'text2vec-transformers'
   | 'text2vec-voyageai'
   | 'text2vec-weaviate'
@@ -93,6 +95,30 @@ export type Multi2VecNvidiaConfig = {
   };
 };
 
+/** The configuration for multi-media vectorization using the AWS module.
+ *
+ * See the [documentation](https://weaviate.io/developers/weaviate/model-providers/aws/embeddings-multimodal) for detailed usage.
+ */
+export type Multi2VecAWSConfig = {
+  /** The dimensionality of the vector once embedded. */
+  dimensions?: number;
+  /** The model to use. */
+  model?: string;
+  /** The AWS region where the model runs. */
+  region?: string;
+  /** The image fields used when vectorizing. */
+  imageFields?: string[];
+  /** The text fields used when vectorizing. */
+  textFields?: string[];
+  /** The weights of the fields used for vectorization. */
+  weights?: {
+    /** The weights of the image fields. */
+    imageFields?: number[];
+    /** The weights of the text fields. */
+    textFields?: number[];
+  };
+};
+
 /** The configuration for multi-media vectorization using the CLIP module.
  *
  * See the [documentation](https://weaviate.io/developers/weaviate/model-providers/transformers/embeddings-multimodal) for detailed usage.
@@ -504,6 +530,8 @@ export type Text2VecPalmConfig = Text2VecGoogleConfig;
 export type Text2VecGoogleConfig = {
   /** The API endpoint to use without a leading scheme such as `http://`. */
   apiEndpoint?: string;
+  /** The dimensionality of the vector once embedded. */
+  dimensions?: number;
   /** The model ID to use. */
   model?: string;
   /** The model ID to use.
@@ -517,6 +545,13 @@ export type Text2VecGoogleConfig = {
   vectorizeCollectionName?: boolean;
 };
 
+export type Text2VecGoogleAiStudioConfig = {
+  /** The model ID to use. */
+  model?: string;
+  /** The Weaviate property name for the `gecko-002` or `gecko-003` model to use as the title. */
+  titleProperty?: string;
+};
+
 /**
  * The configuration for text vectorization using the Transformers module.
  *
@@ -579,10 +614,18 @@ export type Text2VecModel2Vec = {
   vectorizeCollectionName?: boolean;
 };
 
+export type Text2VecMorphConfig = {
+  /** The base URL to use where API requests should go. */
+  baseURL?: string;
+  /** The model to use. */
+  model?: string;
+};
+
 export type NoVectorizerConfig = {};
 
 export type VectorizerConfig =
   | Img2VecNeuralConfig
+  | Multi2VecAWSConfig
   | Multi2VecClipConfig
   | Multi2VecBindConfig
   | Multi2VecGoogleConfig
@@ -652,6 +695,8 @@ export type VectorizerConfigType<V> = V extends 'img2vec-neural'
   ? Text2VecMistralConfig | undefined
   : V extends 'text2vec-model2vec'
   ? Text2VecModel2Vec | undefined
+  : V extends 'text2vec-morph'
+  ? Text2VecMorphConfig | undefined
   : V extends 'text2vec-ollama'
   ? Text2VecOllamaConfig | undefined
   : V extends 'text2vec-openai'