Merge pull request #1426 from weaviate/m2v-jina-updates

m2v-jina: Add v1 clip model; dimensions param
weaviate · Nov 20, 2024 · 915c3cb · 915c3cb
2 parents acf7c0d + 1793762
commit 915c3cb
Show file tree

Hide file tree

Showing 3 changed files with 21 additions and 2 deletions.
diff --git a/test/collection/test_config.py b/test/collection/test_config.py
@@ -372,11 +372,13 @@ def test_basic_config():
     (
         Configure.Vectorizer.multi2vec_jinaai(
             model="jina-clip-v2",
+            dimensions=512,
             vectorize_collection_name=False,
         ),
         {
             "multi2vec-jinaai": {
                 "model": "jina-clip-v2",
+                "dimensions": 512,
                 "vectorizeClassName": False,
             }
         },
@@ -1276,11 +1278,16 @@ def test_vector_config_flat_pq() -> None:
         },
     ),
     (
-        [Configure.NamedVectors.multi2vec_jinaai(name="test", text_fields=["prop"])],
+        [
+            Configure.NamedVectors.multi2vec_jinaai(
+                name="test", dimensions=256, text_fields=["prop"]
+            )
+        ],
         {
             "test": {
                 "vectorizer": {
                     "multi2vec-jinaai": {
+                        "dimensions": 256,
                         "vectorizeClassName": True,
                         "textFields": ["prop"],
                     }

diff --git a/weaviate/collections/classes/config_named_vectors.py b/weaviate/collections/classes/config_named_vectors.py
@@ -1132,6 +1132,7 @@ def multi2vec_jinaai(
         vectorize_collection_name: bool = True,
         base_url: Optional[AnyHttpUrl] = None,
         model: Optional[Union[JinaMultimodalModel, str]] = None,
+        dimensions: Optional[int] = None,
         image_fields: Optional[Union[List[str], List[Multi2VecField]]] = None,
         text_fields: Optional[Union[List[str], List[Multi2VecField]]] = None,
     ) -> _NamedVectorConfigCreate:
@@ -1151,6 +1152,8 @@ def multi2vec_jinaai(
                 The model to use. Defaults to `None`, which uses the server-defined default.
             `base_url`
                 The base URL to use where API requests should go. Defaults to `None`, which uses the server-defined default.
+            `dimensions`
+                The number of dimensions for the generated embeddings (only available for some models). Defaults to `None`, which uses the server-defined default.
             `image_fields`
                 The image fields to use in vectorization.
             `text_fields`
@@ -1164,6 +1167,7 @@ def multi2vec_jinaai(
             vectorizer=_Multi2VecJinaConfig(
                 baseURL=base_url,
                 model=model,
+                dimensions=dimensions,
                 vectorizeClassName=vectorize_collection_name,
                 imageFields=_map_multi2vec_fields(image_fields),
                 textFields=_map_multi2vec_fields(text_fields),

diff --git a/weaviate/collections/classes/config_vectorizers.py b/weaviate/collections/classes/config_vectorizers.py
@@ -40,7 +40,10 @@
     "jina-embeddings-v2-base-code",
     "jina-embeddings-v3",
 ]
-JinaMultimodalModel: TypeAlias = Literal["jina-clip-v2",]
+JinaMultimodalModel: TypeAlias = Literal[
+    "jina-clip-v1",
+    "jina-clip-v2",
+]
 VoyageModel: TypeAlias = Literal[
     "voyage-3",
     "voyage-3-lite",
@@ -405,6 +408,7 @@ class _Multi2VecJinaConfig(_Multi2VecBase):
     )
     baseURL: Optional[AnyHttpUrl]
     model: Optional[str]
+    dimensions: Optional[int]
 
     def _to_dict(self) -> Dict[str, Any]:
         ret_dict = super()._to_dict()
@@ -1220,6 +1224,7 @@ def multi2vec_jinaai(
         model: Optional[Union[JinaMultimodalModel, str]] = None,
         vectorize_collection_name: bool = True,
         base_url: Optional[AnyHttpUrl] = None,
+        dimensions: Optional[int] = None,
         image_fields: Optional[Union[List[str], List[Multi2VecField]]] = None,
         text_fields: Optional[Union[List[str], List[Multi2VecField]]] = None,
     ) -> _VectorizerConfigCreate:
@@ -1235,6 +1240,8 @@ def multi2vec_jinaai(
                 Whether to vectorize the collection name. Defaults to `True`.
             `base_url`
                 The base URL to use where API requests should go. Defaults to `None`, which uses the server-defined default.
+            `dimensions`
+                The number of dimensions for the generated embeddings (only available for some models). Defaults to `None`, which uses the server-defined default.
             `image_fields`
                 The image fields to use in vectorization.
             `text_fields`
@@ -1246,6 +1253,7 @@ def multi2vec_jinaai(
         return _Multi2VecJinaConfig(
             baseURL=base_url,
             model=model,
+            dimensions=dimensions,
             vectorizeClassName=vectorize_collection_name,
             imageFields=_map_multi2vec_fields(image_fields),
             textFields=_map_multi2vec_fields(text_fields),