feat: support sparse embeddings & video input

Author: hks

volcenginesdkarkruntime/types/multimodal_embedding/embedding_data.py

@@ -1,4 +1,4 @@
-from typing import List
+from typing import List, Optional
 from typing_extensions import Literal
 
 from ..._models import BaseModel
@@ -20,5 +20,5 @@ class MultimodalEmbedding(BaseModel):
     object: Literal["embedding"]
     """The object type, which is always "embedding"."""
 
-    sparse_embedding: SparseEmbedding
+    sparse_embedding: Optional[SparseEmbedding] = None
     """The sparse embeddings generated by the model."""

volcenginesdkexamples/volcenginesdkarkruntime/multimodal_embeddings.py

@@ -4,7 +4,7 @@
 
 print("----- multimodal embeddings request -----")
 resp = client.multimodal_embeddings.create(
-    model="${YOUR_ENDPOINT_ID}",
+    model="doubao-embedding-vision-250615",
     input=[
         {
             "type": "text",
@@ -18,4 +18,4 @@
         }
     ]
 )
-print(resp)
+print(resp.data)