Skip to content

Latest commit

 

History

History
11 lines (7 loc) · 540 Bytes

File metadata and controls

11 lines (7 loc) · 540 Bytes

Model Files

Download the Florence-2-base ONNX files from:

https://huggingface.co/onnx-community/Florence-2-base

By default, the server expects the model repository contents under Florence-2-base/, including Florence-2-base/onnx/vision_encoder.onnx, Florence-2-base/onnx/decoder_model_merged.onnx, and the matching tokenizer files.

Startup validates the selected vision encoder, matching ONNX graph files, the merged decoder graph used for KV-cache generation, and tokenizer.json before the server accepts traffic.