Update model card: base vs Alhazen-OCR (khattmix eval)

Promote alhazen-ocr-khattmix → alhazen-ocr (eval win)

Update model card with hwmix CER/WER (50-sample eval)

Promote handwriting-heavy hwmix adapter (QLoRA on Qwen3-VL-2B)

Update README.md

Update README.md

Point dataset links to context212/context212-alhazen-ocr

Upload model trained with Unsloth

Revert to 20k-sample epoch-1 checkpoint (full-data run diverged)

Upload model trained with Unsloth

Update README.md

Delete files optimizer.pt rng_state.pth scheduler.pt trainer_state.json training_args.bin with huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

initial commit