Post
3,164
๐น๐ท **A 1B Turkish OCR model vs. Baidu OCR.**
We ran both models on the **same Turkish enterprise documents**.
The result:
**Werea-DocOCR-1B โ 99.9**
**Baidu Unlimited-OCR โ 47.9**
Same documents.
Same evaluation.
And Werea-DocOCR is only **1B parameters**.
It was built specifically for difficult Turkish enterprise documents:
๐ invoices
๐ contracts
๐ฆ bank receipts
๐ผ payroll
๐ vehicle documents
๐ SGK-style tables
๐ฑ 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.
๐ค Model:
Werea-co/Werea-DocOCR-1B
๐ Dataset:
Werea-co/werea-tr-doc-ocr-enterprise-v2
๐น๐ท Built in Tรผrkiye. Open on Hugging Face.
#OCR #DocumentAI #HuggingFace #TurkishAI
We ran both models on the **same Turkish enterprise documents**.
The result:
**Werea-DocOCR-1B โ 99.9**
**Baidu Unlimited-OCR โ 47.9**
Same documents.
Same evaluation.
And Werea-DocOCR is only **1B parameters**.
It was built specifically for difficult Turkish enterprise documents:
๐ invoices
๐ contracts
๐ฆ bank receipts
๐ผ payroll
๐ vehicle documents
๐ SGK-style tables
๐ฑ 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.
๐ค Model:
Werea-co/Werea-DocOCR-1B
๐ Dataset:
Werea-co/werea-tr-doc-ocr-enterprise-v2
๐น๐ท Built in Tรผrkiye. Open on Hugging Face.
#OCR #DocumentAI #HuggingFace #TurkishAI