docs

Author: nnethercott

meilisearch/models/embedders.py

@@ -21,15 +21,17 @@ class Distribution(CamelBase):
     sigma: float
 
 
-class PoolingOpt(str, Enum):
-    """Pooling options for HuggingFaceEmbedder.
+class PoolingType(str, Enum):
+    """Pooling strategies for HuggingFaceEmbedder.
 
-    Parameters
+    Attributes
     ----------
-    mean: float
-        Mean value between 0 and 1
-    sigma: float
-        Sigma value between 0 and 1
+    USE_MODEL : str
+        Use the model's default pooling strategy.
+    FORCE_MEAN : str
+        Force mean pooling over the token embeddings.
+    FORCE_CLS : str
+        Use the [CLS] token embedding as the sentence representation.
     """
 
     USE_MODEL = "useModel"
@@ -96,6 +98,8 @@ class HuggingFaceEmbedder(CamelBase):
         Describes the natural distribution of search results
     binary_quantized: Optional[bool]
         Once set to true, irreversibly converts all vector dimensions to 1-bit values
+    pooling: Optional[PoolingType]
+        Configures how individual tokens are merged into a single embedding
     """
 
     source: str = "huggingFace"
@@ -107,7 +111,7 @@ class HuggingFaceEmbedder(CamelBase):
     document_template_max_bytes: Optional[int] = None  # Default to 400
     distribution: Optional[Distribution] = None
     binary_quantized: Optional[bool] = None
-    pooling: Optional[PoolingOpt] = None
+    pooling: Optional[PoolingType] = None
 
 
 class OllamaEmbedder(CamelBase):
@@ -210,15 +214,27 @@ class UserProvidedEmbedder(CamelBase):
 
 
 class CompositeEmbedder(CamelBase):
+    """Composite embedder configuration.
+
+    Parameters
+    ----------
+    source: str
+        The embedder source, must be "userProvided"
+    indexing_embedder: Union[OpenAiEmbedder, HuggingFaceEmbedder, OllamaEmbedder, RestEmbedder, UserProvidedEmbedder]
+        The embedder used for indexing queries remotely
+    search_embedder: Union[OpenAiEmbedder, HuggingFaceEmbedder, OllamaEmbedder, RestEmbedder, UserProvidedEmbedder]
+        The embedder used for processing queries locally
+    """
+
     source: str = "composite"
-    search_embedder: Union[
+    indexing_embedder: Union[
         OpenAiEmbedder,
         HuggingFaceEmbedder,
         OllamaEmbedder,
         RestEmbedder,
         UserProvidedEmbedder,
     ]
-    indexing_embedder: Union[
+    search_embedder: Union[
         OpenAiEmbedder,
         HuggingFaceEmbedder,
         OllamaEmbedder,

tests/settings/test_settings_embedders.py

@@ -1,6 +1,12 @@
 # pylint: disable=redefined-outer-name
 
-from meilisearch.models.embedders import OpenAiEmbedder, UserProvidedEmbedder
+import pytest
+from meilisearch.models.embedders import (
+    HuggingFaceEmbedder,
+    OpenAiEmbedder,
+    UserProvidedEmbedder,
+    CompositeEmbedder,
+)
 
 
 def test_get_default_embedders(empty_index):
@@ -183,3 +189,35 @@ def test_user_provided_embedder_format(empty_index):
     assert embedders.embedders["user_provided"].distribution.mean == 0.5
     assert embedders.embedders["user_provided"].distribution.sigma == 0.1
     assert embedders.embedders["user_provided"].binary_quantized is False
+
+
+@pytest.mark.usefixtures("enable_composite_embedders")
+def test_composite_embedder_format(empty_index):
+    """Tests that CompositeEmbedder embedder has the required fields and proper format."""
+    index = empty_index()
+
+    hf_default = HuggingFaceEmbedder().model_dump(by_alias=True, exclude_none=True)
+
+    # create composite embedder
+    composite_embedder = {
+        "default": {
+            "source": "composite",
+            "searchEmbedder": hf_default,
+            "indexingEmbedder": hf_default,
+        }
+    }
+
+    response = index.update_embedders(composite_embedder)
+    index.wait_for_task(response.task_uid)
+    embedders = index.get_embedders()
+    print(embedders)
+    assert embedders.embedders["composite"].source == "composite"
+
+    assert isinstance(embedders.embedders["composite"], CompositeEmbedder)
+    assert isinstance(embedders.embedders["composite"].search_embedder, HuggingFaceEmbedder)
+    assert isinstance(embedders.embedders["composite"].indexing_embedder, HuggingFaceEmbedder)
+
+    assert not hasattr(embedders.embedders["composite"].search_embedder, "document_template")
+    assert not hasattr(embedders.embedders["composite"].search_embedder, "document_template_max_bytes")
+    assert hasattr(embedders.embedders["composite"].indexing_embedder, "document_template")
+