Skip to content

ezgh ocr batches

Batch OCR: many documents at a time, processed in the background.

Updated View as Markdown
ezgh ocr batches <command>

Batch OCR processes up to 1,000 documents (5 GiB) at a time with one processor, in the background. Each document can be as large as the model takes (ezgh ocr models get): 20 MiB and 100 pages for unlimited-ocr.

ezgh ocr batches create does it all: it uploads the documents, submits the batch, and with --wait waits for it and --download saves the outputs. Outputs are kept for 30 days after the batch finishes. Batches are in a project (--project).

Part of ezgh ocr.

Flags

This command also takes the global flags.

Subcommands

Command Description
ezgh ocr batches cancel Cancel a submitted batch
ezgh ocr batches create Process many documents: upload them, submit the batch, and optionally wait and download
ezgh ocr batches delete Delete a batch, with its documents and outputs
ezgh ocr batches download Save a finished batch’s outputs: each document’s, a zip of them all, or results.jsonl
ezgh ocr batches files List a batch’s documents: each one’s status, pages and error
ezgh ocr batches get Show a batch, by ID, slug or name: its status and progress
ezgh ocr batches list List the project’s batches, newest first
Navigation

Type to search…

↑↓ navigate↵ selectEsc close