Clarify provenance: promoted from lora_r8/result_model in font-model-results

README.md

@@ -21,7 +21,7 @@ A DINOv2 Vision Transformer fine-tuned with LoRA for font classification across
 
 1. **Base model**: [facebook/dinov2-base-imagenet1k-1-layer](https://huggingface.co/facebook/dinov2-base-imagenet1k-1-layer) (87.2M parameters, frozen).
 2. **Fine-tuning**: [LoRA](https://arxiv.org/abs/2106.09685) (rank 8, alpha 16) applied to the query and value projections in each ViT attention block, plus a trainable classification head. ~900K trainable parameters (1% of total).
-3. **Merge**: After training, the LoRA adapter weights were merged into the base model (`merge_and_unload()`), producing this standalone checkpoint. No adapter or PEFT library needed at inference time.
+3. **Promotion**: This model was promoted from the `lora_r8/result_model` adapter in [dchen0/font-model-results](https://huggingface.co/dchen0/font-model-results) using `promote_model.py`. That script loads the base DINOv2 model, merges the LoRA adapter weights into it (`merge_and_unload()`), and uploads the result as a standalone checkpoint. No adapter or PEFT library needed at inference time.
 
 ## Performance