Update README.md

README.md

@@ -1,3 +1,23 @@
+
+A proof of concept generating captions using Google Gemma 3 on Google Colab Free Tier for captioning prompts akin to training data of FLUX Chroma: https://huggingface.co/lodestones/Chroma
+
+Try the Chroma model at: https://tensor.art/models/891236315830428357
+
+This dataset was built using 200 images from Redcaps : https://huggingface.co/datasets/lodestones/pixelprose
+
+And 200 LLM captioned e621 images: https://huggingface.co/datasets/lodestones/e621-captions/tree/main
+
+The total trained images are just 400 total , randomly selected , so this LoRa adaptation is very basic! You can likely train a better version yourself with listed tools on Google Colab Free Tier T4. 
+
+//----//
+
+I made some .parquets of the captions here for easier browsing: https://huggingface.co/datasets/codeShare/chroma_prompts
+
+To use this Gemma LoRa adaptation got to the Google Colab Jupyter notebook in this repo: https://huggingface.co/codeShare/flux_chroma_image_captioner/blob/main/gemma_image_captioner.ipynb
+
+To train your own LoRa adaptation of the Gemma on Google Colab Free Tier T4 , visit :  https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Gemma3_(4B)-Vision.ipynb
+
+
 ---
 base_model: unsloth/gemma-3-4b-pt-unsloth-bnb-4bit
 library_name: peft