azcrawlerpy-recaptcha-text

The default model for azcrawlerpy's text captcha solver: the SVTRv2 recogniser svtrv2.onnx, the input-size selector selector.onnx fitted to it, and their provenance card svtrv2.json. azcrawlerpy[text] installs it; there is nothing to import.

The recogniser and selector are one pair: the selector is fitted to one set of recogniser weights, and azcrawlerpy refuses to load a recogniser without one beside it.

It is published apart from azcrawlerpy so that code releases do not upload the weights again, and a retrained model ships without an azcrawlerpy release.

Versions

• Major: the contract azcrawlerpy relies on (file names, ONNX inputs and outputs, charset). azcrawlerpy depends on ~=1.0, so a major bump needs an azcrawlerpy release that accepts it.
• Minor: retrained weights. Existing azcrawlerpy installs pick them up on upgrade.

Releasing a model

promote() in captcha-training writes the weights, selector and card into azcrawlerpy_recaptcha_text/ in this checkout, only when they beat the model already here. Commit them, bump the minor version, check with uv sync --extra dev && uv run pytest, and publish with build.txt. It builds the wheel only: an sdist would upload the same weights a second time against the project's PyPI quota.