new: add embedding size property (#521)

* new: add embedding size property

* fix: format exception message

* new: replace embedding size property with get_embedding_size classmethod

* fix: fix missed parts

* chore: fix docstrings

* new: add embedding_size property

Author: joein

fastembed/image/image_embedding.py

@@ -77,6 +77,40 @@ def __init__(
             "Please check the supported models using `ImageEmbedding.list_supported_models()`"
         )
 
+    @property
+    def embedding_size(self) -> int:
+        """Get the embedding size of the current model"""
+        if self._embedding_size is None:
+            self._embedding_size = self.get_embedding_size(self.model_name)
+        return self._embedding_size
+
+    @classmethod
+    def get_embedding_size(cls, model_name: str) -> int:
+        """Get the embedding size of the passed model
+
+        Args:
+            model_name (str): The name of the model to get embedding size for.
+
+        Returns:
+            int: The size of the embedding.
+
+        Raises:
+            ValueError: If the model name is not found in the supported models.
+        """
+        descriptions = cls._list_supported_models()
+        embedding_size: Optional[int] = None
+        for description in descriptions:
+            if description.model.lower() == model_name.lower():
+                embedding_size = description.dim
+                break
+        if embedding_size is None:
+            model_names = [description.model for description in descriptions]
+            raise ValueError(
+                f"Embedding size for model {model_name} was None. "
+                f"Available model names: {model_names}"
+            )
+        return embedding_size
+
     def embed(
         self,
         images: Union[ImageInput, Iterable[ImageInput]],

fastembed/image/image_embedding_base.py

@@ -18,6 +18,7 @@ def __init__(
         self.cache_dir = cache_dir
         self.threads = threads
         self._local_files_only = kwargs.pop("local_files_only", False)
+        self._embedding_size: Optional[int] = None
 
     def embed(
         self,
@@ -42,3 +43,13 @@ def embed(
             Iterable[NdArray]: The embeddings.
         """
         raise NotImplementedError()
+
+    @classmethod
+    def get_embedding_size(cls, model_name: str) -> int:
+        """Returns embedding size of the chosen model."""
+        raise NotImplementedError("Subclasses must implement this method")
+
+    @property
+    def embedding_size(self) -> int:
+        """Returns embedding size for the current model"""
+        raise NotImplementedError("Subclasses must implement this method")

fastembed/late_interaction/late_interaction_embedding_base.py

@@ -17,6 +17,7 @@ def __init__(
         self.cache_dir = cache_dir
         self.threads = threads
         self._local_files_only = kwargs.pop("local_files_only", False)
+        self._embedding_size: Optional[int] = None
 
     def embed(
         self,
@@ -58,3 +59,13 @@ def query_embed(self, query: Union[str, Iterable[str]], **kwargs: Any) -> Iterab
             yield from self.embed([query], **kwargs)
         else:
             yield from self.embed(query, **kwargs)
+
+    @classmethod
+    def get_embedding_size(cls, model_name: str) -> int:
+        """Returns embedding size of the chosen model."""
+        raise NotImplementedError("Subclasses must implement this method")
+
+    @property
+    def embedding_size(self) -> int:
+        """Returns embedding size for the current model"""
+        raise NotImplementedError("Subclasses must implement this method")

fastembed/late_interaction/late_interaction_text_embedding.py

@@ -80,6 +80,40 @@ def __init__(
             "Please check the supported models using `LateInteractionTextEmbedding.list_supported_models()`"
         )
 
+    @property
+    def embedding_size(self) -> int:
+        """Get the embedding size of the current model"""
+        if self._embedding_size is None:
+            self._embedding_size = self.get_embedding_size(self.model_name)
+        return self._embedding_size
+
+    @classmethod
+    def get_embedding_size(cls, model_name: str) -> int:
+        """Get the embedding size of the passed model
+
+        Args:
+            model_name (str): The name of the model to get embedding size for.
+
+        Returns:
+            int: The size of the embedding.
+
+        Raises:
+            ValueError: If the model name is not found in the supported models.
+        """
+        descriptions = cls._list_supported_models()
+        embedding_size: Optional[int] = None
+        for description in descriptions:
+            if description.model.lower() == model_name.lower():
+                embedding_size = description.dim
+                break
+        if embedding_size is None:
+            model_names = [description.model for description in descriptions]
+            raise ValueError(
+                f"Embedding size for model {model_name} was None. "
+                f"Available model names: {model_names}"
+            )
+        return embedding_size
+
     def embed(
         self,
         documents: Union[str, Iterable[str]],

fastembed/late_interaction_multimodal/late_interaction_multimodal_embedding.py

@@ -83,6 +83,40 @@ def __init__(
             "Please check the supported models using `LateInteractionMultimodalEmbedding.list_supported_models()`"
         )
 
+    @property
+    def embedding_size(self) -> int:
+        """Get the embedding size of the current model"""
+        if self._embedding_size is None:
+            self._embedding_size = self.get_embedding_size(self.model_name)
+        return self._embedding_size
+
+    @classmethod
+    def get_embedding_size(cls, model_name: str) -> int:
+        """Get the embedding size of the passed model
+
+        Args:
+            model_name (str): The name of the model to get embedding size for.
+
+        Returns:
+            int: The size of the embedding.
+
+        Raises:
+            ValueError: If the model name is not found in the supported models.
+        """
+        descriptions = cls._list_supported_models()
+        embedding_size: Optional[int] = None
+        for description in descriptions:
+            if description.model.lower() == model_name.lower():
+                embedding_size = description.dim
+                break
+        if embedding_size is None:
+            model_names = [description.model for description in descriptions]
+            raise ValueError(
+                f"Embedding size for model {model_name} was None. "
+                f"Available model names: {model_names}"
+            )
+        return embedding_size
+
     def embed_text(
         self,
         documents: Union[str, Iterable[str]],

fastembed/late_interaction_multimodal/late_interaction_multimodal_embedding_base.py

@@ -19,6 +19,7 @@ def __init__(
         self.cache_dir = cache_dir
         self.threads = threads
         self._local_files_only = kwargs.pop("local_files_only", False)
+        self._embedding_size: Optional[int] = None
 
     def embed_text(
         self,
@@ -65,3 +66,13 @@ def embed_image(
             List of embeddings, one per image
         """
         raise NotImplementedError()
+
+    @classmethod
+    def get_embedding_size(cls, model_name: str) -> int:
+        """Returns embedding size of the chosen model."""
+        raise NotImplementedError("Subclasses must implement this method")
+
+    @property
+    def embedding_size(self) -> int:
+        """Returns embedding size for the current model"""
+        raise NotImplementedError("Subclasses must implement this method")

fastembed/text/text_embedding.py

@@ -128,6 +128,40 @@ def __init__(
             "Please check the supported models using `TextEmbedding.list_supported_models()`"
         )
 
+    @property
+    def embedding_size(self) -> int:
+        """Get the embedding size of the current model"""
+        if self._embedding_size is None:
+            self._embedding_size = self.get_embedding_size(self.model_name)
+        return self._embedding_size
+
+    @classmethod
+    def get_embedding_size(cls, model_name: str) -> int:
+        """Get the embedding size of the passed model
+
+        Args:
+            model_name (str): The name of the model to get embedding size for.
+
+        Returns:
+            int: The size of the embedding.
+
+        Raises:
+            ValueError: If the model name is not found in the supported models.
+        """
+        descriptions = cls._list_supported_models()
+        embedding_size: Optional[int] = None
+        for description in descriptions:
+            if description.model.lower() == model_name.lower():
+                embedding_size = description.dim
+                break
+        if embedding_size is None:
+            model_names = [description.model for description in descriptions]
+            raise ValueError(
+                f"Embedding size for model {model_name} was None. "
+                f"Available model names: {model_names}"
+            )
+        return embedding_size
+
     def embed(
         self,
         documents: Union[str, Iterable[str]],

fastembed/text/text_embedding_base.py

@@ -17,6 +17,7 @@ def __init__(
         self.cache_dir = cache_dir
         self.threads = threads
         self._local_files_only = kwargs.pop("local_files_only", False)
+        self._embedding_size: Optional[int] = None
 
     def embed(
         self,
@@ -58,3 +59,13 @@ def query_embed(self, query: Union[str, Iterable[str]], **kwargs: Any) -> Iterab
             yield from self.embed([query], **kwargs)
         else:
             yield from self.embed(query, **kwargs)
+
+    @classmethod
+    def get_embedding_size(cls, model_name: str) -> int:
+        """Returns embedding size of the passed model."""
+        raise NotImplementedError("Subclasses must implement this method")
+
+    @property
+    def embedding_size(self) -> int:
+        """Returns embedding size for the current model"""
+        raise NotImplementedError("Subclasses must implement this method")

tests/test_image_onnx_embeddings.py

@@ -127,3 +127,21 @@ def test_lazy_load(model_name: str) -> None:
     assert hasattr(model.model, "model")
     if is_ci:
         delete_model_cache(model.model._model_dir)
+
+
+def test_get_embedding_size() -> None:
+    assert ImageEmbedding.get_embedding_size(model_name="Qdrant/clip-ViT-B-32-vision") == 512
+    assert ImageEmbedding.get_embedding_size(model_name="Qdrant/clip-vit-b-32-vision") == 512
+
+
+def test_embedding_size() -> None:
+    is_ci = os.getenv("CI")
+    model_name = "Qdrant/clip-ViT-B-32-vision"
+    model = ImageEmbedding(model_name=model_name, lazy_load=True)
+    assert model.embedding_size == 512
+
+    model_name = "Qdrant/clip-vit-b-32-vision"
+    model = ImageEmbedding(model_name=model_name, lazy_load=True)
+    assert model.embedding_size == 512
+    if is_ci:
+        delete_model_cache(model.model._model_dir)

tests/test_late_interaction_embeddings.py

@@ -254,3 +254,24 @@ def test_lazy_load(model_name: str):
 
     if is_ci:
         delete_model_cache(model.model._model_dir)
+
+
+def test_get_embedding_size():
+    model_name = "answerdotai/answerai-colbert-small-v1"
+    assert LateInteractionTextEmbedding.get_embedding_size(model_name) == 96
+
+    model_name = "answerdotai/answerai-ColBERT-small-v1"
+    assert LateInteractionTextEmbedding.get_embedding_size(model_name) == 96
+
+
+def test_embedding_size():
+    is_ci = os.getenv("CI")
+    model_name = "answerdotai/answerai-colbert-small-v1"
+    model = LateInteractionTextEmbedding(model_name=model_name, lazy_load=True)
+    assert model.embedding_size == 96
+
+    model_name = "answerdotai/answerai-ColBERT-small-v1"
+    model = LateInteractionTextEmbedding(model_name=model_name, lazy_load=True)
+    assert model.embedding_size == 96
+    if is_ci:
+        delete_model_cache(model.model._model_dir)