Update README.md

README.md

@@ -17,6 +17,15 @@ language:
 - **License:** apache-2.0
 - **Finetuned from model :** unsloth/llama-3.2-3b-unsloth-bnb-4bit
 
-This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
+# Datasets Used
+- **1.** https://huggingface.co/datasets/mlabonne/FineTome-100k
+- **2.** https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered
 
-[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
+# Updates
+- This version now prints the tokenizer bs I was talking about in my previous version, it could be due to the dataset or how I've prepared the data.
+  Regardless, multi-turn datasets were too much for me to handle anyways.
+
+- model does not refuse to anything yet.
+
+# To Do
+- Scrap or finetune this, not sure anyways