What is the correct way to do multimodal embeddings on an image? #1511
Description
I have code that works:
model = MultiModalEmbeddingModel.from_pretrained("multimodalembedding@001")
try:
image_object = Image.load_from_file("temp_image.png")
embeddings = model.get_embeddings(image=image_object)
except Exception as e:
print(f"Error embedding image from temp_image.png: {e}")
return None, None
#this is the image vector
image_embedding = embeddings.image_embedding
but I get this:
/workspaces/vscode-python-devcontainer/my_project/.venv/lib/python3.12/site-packages/vertexai/_model_garden/_model_garden_models.py:278: UserWarning: This feature is deprecated as of June 24, 2025 and will be removed on June 24, 2026. For details, see https://cloud.google.com/vertex-ai/generative-ai/docs/deprecations/genai-vertexai-sdk.
warning_logs.show_deprecation_warning()
/workspaces/vscode-python-devcontainer/my_project/.venv/lib/python3.12/site-packages/vertexai/vision_models/_vision_models.py:153: UserWarning: This feature is deprecated as of June 24, 2025 and will be removed on June 24, 2026. For details, see https://cloud.google.com/vertex-ai/generative-ai/docs/deprecations/genai-vertexai-sdk.
warning_logs.show_deprecation_warning()
That link has a section for text embeddings:
https://cloud.google.com/vertex-ai/generative-ai/docs/deprecations/genai-vertexai-sdk#embeddings
But the code for embed_content says it's text only!
python-genai/google/genai/models.py
Line 3923 in 7d4d23e
Am I not using the correct libraries? Thank you!