Update README.md

This model is **Unfiltered** and **Quantized** to 4bit. Also it will work with [llama.cpp](https://github.com/ggerganov/llama.cpp). This model has been trained with alpaca dataset and then finetuned with Vicuna detaset.

README.md

@@ -1,3 +1,15 @@
 ---
 license: mit
----
+datasets:
+- anon8231489123/ShareGPT_Vicuna_unfiltered
+- viewv/LLaMA-13B
+- tatsu-lab/alpaca
+metrics:
+- accuracy
+pipeline_tag: text-generation
+tags:
+- alpaca
+- vicuna
+- finetuned
+- unfiltered
+---