TechJan 30, 20264 minPaddleOCR-VL-1.5 - document parsing SOTA with only 0.9B parametersBaidu's PaddleOCR-VL-1.5 reaches 94.5% accuracy on OmniDocBench v1.5 with just 0.9B parameters, surpassing large models such as GPT-4o and Qwen2.5-VL-72B.AIOCRVLMPaddlePaddle
