How much GPU memory does multi-modal-embed-small need?

About 1 GB: 0.03 GB of weights and the runtime's own memory.

What is the cheapest GPU to run multi-modal-embed-small?

Right now one Quadro P4000 at $0.07 an hour; RTX 5090 at $0.67 an hour is the best value for its speed. Prices are live offers from GPU clouds, checked every 15 minutes.

Can multi-modal-embed-small run on an RTX 4090?

Yes: it needs about 1 GB and an RTX 4090 has 24 GB.

Models / multi-modal-embed-small

vllm-sr/multi-modal-embed-small

Embeddings · served with ONNX Runtime · ↓ 4.6K downloads a month · View on Hugging Face

vllm-sr/multi-modal-embed-small needs about 1 GB of GPU memory. It runs on one Quadro P4000 from $0.07/h right now; best value: RTX 5090 at $0.67/h.

Weights: 30 MB
GPU memory needed: about 1 GB
Memory · disk: 8 GB · 30 GB
Licence: apache-2.0

Versions

Smaller versions (8-bit, 4-bit) need less GPU memory and answer a little less precisely.

full precisiononnx/audio_encoder.onnx30 MB · 1 GB GPU full precisiononnx/image_encoder.onnx380 MB · 1.6 GB GPU full precisiononnx/text_encoder.onnx90 MB · 1.1 GB GPU

Hardware and deploy

Deploying ONNX models from the catalog is coming soon. Today, download the files from Hugging Face and upload them in Inference.

Questions

How much GPU memory does multi-modal-embed-small need?: About 1 GB: 0.03 GB of weights and the runtime's own memory.
What is the cheapest GPU to run multi-modal-embed-small?: Right now one Quadro P4000 at $0.07 an hour; RTX 5090 at $0.67 an hour is the best value for its speed. Prices are live offers from GPU clouds, checked every 15 minutes.
Can multi-modal-embed-small run on an RTX 4090?: Yes: it needs about 1 GB and an RTX 4090 has 24 GB.

Loading…