Post
3117
πΉπ· **A 1B Turkish OCR model vs. Baidu OCR.**
We ran both models on the **same Turkish enterprise documents**.
The result:
**Werea-DocOCR-1B β 99.9**
**Baidu Unlimited-OCR β 47.9**
Same documents.
Same evaluation.
And Werea-DocOCR is only **1B parameters**.
It was built specifically for difficult Turkish enterprise documents:
π invoices
π contracts
π¦ bank receipts
πΌ payroll
π vehicle documents
π SGK-style tables
π± scanned & phone-captured documents
But benchmarks aren't enough.
**Give me a Turkish document that you think will break it.**
We'll test the hardest ones and publish the failures.
π€ Model:
Werea-co/Werea-DocOCR-1B
π Dataset:
Werea-co/werea-tr-doc-ocr-enterprise-v2
πΉπ· Built in TΓΌrkiye. Open on Hugging Face.
#OCR #DocumentAI #HuggingFace #TurkishAI
We ran both models on the **same Turkish enterprise documents**.
The result:
**Werea-DocOCR-1B β 99.9**
**Baidu Unlimited-OCR β 47.9**
Same documents.
Same evaluation.
And Werea-DocOCR is only **1B parameters**.
It was built specifically for difficult Turkish enterprise documents:
π invoices
π contracts
π¦ bank receipts
πΌ payroll
π vehicle documents
π SGK-style tables
π± scanned & phone-captured documents
But benchmarks aren't enough.
**Give me a Turkish document that you think will break it.**
We'll test the hardest ones and publish the failures.
π€ Model:
Werea-co/Werea-DocOCR-1B
π Dataset:
Werea-co/werea-tr-doc-ocr-enterprise-v2
πΉπ· Built in TΓΌrkiye. Open on Hugging Face.
#OCR #DocumentAI #HuggingFace #TurkishAI
