17 hours ago

MiniCPM-V 4.6 vision model (1.3B): image understanding and OCR in about 1.6 GB. Image input works.

vision