Sources
Loading...
Additional media
Loading...
DeepNewz, mobile.
People-sourced. AI-powered. Unbiased News.
StepFun has released a new 580M parameter model called GOT (General OCR Theory), which is an end-to-end OCR-2.0 model. This model, despite not achieving state-of-the-art benchmark scores, is notable for its relatively smaller size compared to other models that are 5-10 times larger. The GOT-OCR model is now available on Hugging Face, a popular platform for machine learning models. StepFun, a new player in China's open-source community, aims to make significant contributions with this OCR-2.0 release.
Testing "General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Model" on plain texts OCR https://t.co/ct9s1Mcdzy Updated https://t.co/oT0SJLh5Fj accordingly🤷♂️ https://t.co/nOxaQ6gWqj
Testing "General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Model" on plain texts OCR https://t.co/ct9s1Mcdzy Updated https://t.co/oT0SJLh5Fj accordingly https://t.co/aCx7Thu3Vc
GOT-OCR2.0 🔥 a 580M end-to-end OCR-2.0 model released by StepFun 阶跃星辰 is now available on the @huggingface Model: https://t.co/SwkZP5wKsE Paper: https://t.co/9bHLkTzj91 ✨ While others are releasing powerful models, StepFun, a new player in China's OS community is opening…
