Update README.md

Clarified that it runs on half-precision inference.

README.md

@@ -7,4 +7,4 @@ tags:
 ---
 llama2-piratelora-13b
 
-This repo contains a Low-Rank Adapter (LoRA) for llama 2 13b fit on a simple dataset comprised of thousands of pirate phrases, conversation pieces, and obscura.
+This repo contains a Low-Rank Adapter (LoRA) for llama 2 13b (16 float) fit on a simple dataset comprised of thousands of pirate phrases, conversation pieces, and obscura.