feat: rename the VectorType

modeling_jina_embeddings_v4.py

@@ -30,9 +30,9 @@ class PromptType(str, Enum):
     passage = "passage"
 
 
-class VectorType(str, Enum):
-    single = "single"
-    multi = "multi"
+class VectorOutputFormat(str, Enum):
+    SINGLE = "single"
+    MULTIPLE = "multiple"
 
 
 PREFIX_DICT = {"query": "Query", "passage": "Passage"}
@@ -324,7 +324,7 @@ class JinaEmbeddingsV4Model(Qwen2_5_VLForConditionalGeneration):
         task_label: Union[str, List[str]],
         processor_fn: Callable,
         desc: str,
-        output_format: Union[str, VectorType] = VectorType.single,
+        output_format: Union[str, VectorOutputFormat] = VectorOutputFormat.SINGLE,
         return_numpy: bool = False,
         batch_size: int = 32,
         truncate_dim: Optional[int] = None,
@@ -344,8 +344,8 @@ class JinaEmbeddingsV4Model(Qwen2_5_VLForConditionalGeneration):
                     device_type=torch.device(self.device).type, dtype=torch.bfloat16
                 ):
                     embeddings = self(**batch, task_label=task_label)
-                    output_format_str = output_format.value if isinstance(output_format, VectorType) else output_format
-                    if output_format_str == VectorType.single.value:
+                    output_format_str = output_format.value if isinstance(output_format, VectorOutputFormat) else output_format
+                    if output_format_str == VectorOutputFormat.SINGLE.value:
                         embeddings = embeddings.single_vec_emb
                         if truncate_dim is not None:
                             embeddings = embeddings[:, :truncate_dim]
@@ -362,7 +362,7 @@ class JinaEmbeddingsV4Model(Qwen2_5_VLForConditionalGeneration):
 
     def _validate_encoding_params(
         self,
-        output_format: Optional[Union[str, VectorType]] = None,
+        output_format: Optional[Union[str, VectorOutputFormat]] = VectorOutputFormat.SINGLE,
         truncate_dim: Optional[int] = None,
         prompt_name: Optional[str] = None,
     ) -> Dict[str, Any]:
@@ -379,16 +379,16 @@ class JinaEmbeddingsV4Model(Qwen2_5_VLForConditionalGeneration):
                     else PREFIX_DICT["query"]
                 )
 
-        output_format = output_format or VectorType.single
-        if isinstance(output_format, VectorType):
+        output_format = output_format or VectorOutputFormat.SINGLE
+        if isinstance(output_format, VectorOutputFormat):
             encode_kwargs["output_format"] = output_format.value
         else:
             try:
-                output_format_enum = VectorType(output_format)
+                output_format_enum = VectorOutputFormat(output_format)
                 encode_kwargs["output_format"] = output_format_enum.value
             except ValueError:
                 raise ValueError(
-                    f"Invalid output_format: {output_format}. Must be one of {[v.value for v in VectorType]}."
+                    f"Invalid output_format: {output_format}. Must be one of {[v.value for v in VectorOutputFormat]}."
                 )
 
         truncate_dim = truncate_dim or self.config.truncate_dim
@@ -422,7 +422,7 @@ class JinaEmbeddingsV4Model(Qwen2_5_VLForConditionalGeneration):
         task: Optional[str] = None,
         max_length: int = 8192,
         batch_size: int = 8,
-        output_format: Optional[Union[str, VectorType]] = None,
+        output_format: Optional[Union[str, VectorOutputFormat]] = VectorOutputFormat.SINGLE,
         return_numpy: bool = False,
         truncate_dim: Optional[int] = None,
         prompt_name: Optional[str] = None,
@@ -434,7 +434,7 @@ class JinaEmbeddingsV4Model(Qwen2_5_VLForConditionalGeneration):
             texts: text or list of text strings to encode
             max_length: Maximum token length for text processing
             batch_size: Number of texts to process at once
-            output_format: Type of embedding vector to generate (VectorType.single or VectorType.multi)
+            output_format: Type of embedding vector to generate (VectorOutputFormat.SINGLE or VectorOutputFormat.MULTIPLE)
             return_numpy: Whether to return numpy arrays instead of torch tensors
             truncate_dim: Dimension to truncate embeddings to (128, 256, 512, or 1024)
             prompt_name: Type of text being encoded ('query' or 'passage')
@@ -489,7 +489,7 @@ class JinaEmbeddingsV4Model(Qwen2_5_VLForConditionalGeneration):
         images: Union[str, Image.Image, List[Union[str, Image.Image]]],
         task: Optional[str] = None,
         batch_size: int = 8,
-        output_format: Optional[Union[str, VectorType]] = None,
+        output_format: Optional[Union[str, VectorOutputFormat]] = VectorOutputFormat.SINGLE,
         return_numpy: bool = False,
         truncate_dim: Optional[int] = None,
         max_pixels: Optional[int] = None,
@@ -500,7 +500,7 @@ class JinaEmbeddingsV4Model(Qwen2_5_VLForConditionalGeneration):
         Args:
             images: image(s) to encode, can be PIL Image(s), URL(s), or local file path(s)
             batch_size: Number of images to process at once
-            output_format: Type of embedding vector to generate (VectorType.single or VectorType.multi)
+            output_format: Type of embedding vector to generate (VectorOutputFormat.SINGLE or VectorOutputFormat.MULTIPLE)
             return_numpy: Whether to return numpy arrays instead of torch tensors
             truncate_dim: Dimension to truncate embeddings to (128, 256, 512, or 1024)
             max_pixels: Maximum number of pixels to process per image