musk12/apple-fastvlm-vision-projector-512

Apple FastVLM Vision Projector (512)

A standalone ONNX vision projector extracted from the FastVLM pipeline. Converts image features into the FastVLM multimodal embedding space and is designed for CPU-based image-to-text and vision+language inference.

Model file

• —vision_projector_v1_standalone.onnx

Usage

Use this model to convert images into embeddings for a compatible FastVLM language backbone.

import onnxruntime as ort

session = ort.InferenceSession("vision_projector_v1_standalone.onnx")
# feed in your preprocessed image tensor and run inference

Base model

Derived from apple/FastVLM-0.5B.