Update README.md

README.md

@@ -45,9 +45,11 @@ pipeline_tag: text-to-image
 # Flux.1-dev-LoRA-fashion
 
 
-Fashion product generator based on Flux.1 is a fine-tuned generative text-to-image conversion model using fashion product dataset (from HF) collected on KREAM, one of the best online marketplaces in Korea.
+Fashion product generator based on Flux.1 is a fine-tuned generative text-to-image conversion model using [fashion product dataset](https://huggingface.co/datasets/hahminlew/kream-product-blip-captions) collected on KREAM, one of the best online marketplaces in Korea.
 
-Detailed instructions on fine-tuning and inference of the model can be found in the github repository: .
+In this version, we improved the original dataset picture captions by replacing the `BLIP` model for automatic captioning with `Qwen2-VL-7B-Instruct`.
+
+Detailed instructions on fine-tuning and inference of the model can be found in the github repository: [Fashion is my profession](https://github.com/zhirnov-m/FiMP).
 
 <Gallery />