albertklorer/safedocs-prime-ocr-20k-sft-80k-grpo

Safedocs Prime OCR — 20K SFT / 80K GRPO Canonical reward labels is outside all training data configurations. Preparation remains provisional until manifest.json confirms completion and training readiness. Repository visibility: public. The legacy private/ path contains only canonical reward labels and references, excluded from all model-input dataset configurations. Raw judge annotations and preparation checkpoints are encrypted and excluded from dataset loading.

<iframe src="https://silytics.si/embed/item/f0d87bac-a195-4939-a7d8-a0b0704cd993" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.