indra-ocr

He aha indra-ocr

Ko Indra-OCR tētahi tauira OCR e hangai ana ki Hugging Face mō te mātauranga kuputuhi i roto i ngā whakaahua.

Ngā āhuatanga matua

• Whakawhiti i ngā PDF me ngā whakaahua ki Markdown, kuputuhi noa, JSON rānei me ngā pouaka rohe
• Ka whakarato i ngā pouaka rohe mō ia rārangi kuputuhi me ia paraka whakatakotoranga, ā, ka tapaina ngā momo paraka (pērā i te taitara, rārangi, ripanga)
• Ka pānui i te paparanga kuputuhi o roto o ngā PDF mamati, ā, ka mahi OCR anake ki ngā wāhanga e ngaro ana
• E tautoko ana i te tātari whakatakotoranga me te rapunga o te raupapa pānui
• Ko te rahi o te tauira tata ki te 18MB, e whakamahi ana i te ONNX Runtime, ā, kāore e hiahia ana ki te PyTorch

Mā wai

Kaiwhakawhanake, ngaio tukatuka tuhinga, me ngā kaiwhakamahi e whakahaere ana i ngā mahi mohio tuhinga e hiahia ana ki te whakawhiti tere i ngā PDF me ngā whakaahua ki kuputuhi hanganga.

Me pēhea te tīmata

1. Whakahaerehia te pip install indra-ocr
2. Whakahaerehia te whakahau indra-ocr paper.pdf ki te whakawhiti i tētahi PDF ki Markdown, ā, whakaputa ki stdout
3. Whakamahia te indra-ocr *.pdf --out results/ ki te tukatuka i ngā PDF maha ki ngā kōnae Markdown

I whakahiatohia mai i ngā whakaahuatanga tūmatanui i te pae mana / README / whārangi whakarewatanga o te taputapu. Kāore anō kia whakamanahia ngā utu me ngā tau kaiwhakamahi; tirohia te pae mana. I hangaia ētahi wāhanga mā te āwhina AI. Kua tuhia ki te reo Hainamana me te reo Ingarihi; ko ētahi atu reo he whakamāoritanga mihini.