diff --git a/README.md b/README.md
index a59ca03079..5f1e65e632 100644
--- a/README.md
+++ b/README.md
@@ -2,6 +2,7 @@
 
 ## 80% faster 50% less memory LLM local finetuning
 * Manual autograd engine - hand derived backprop steps.
+* QLoRA / LoRA 80% faster, 50% less memory.
 * All kernels written in OpenAI's Triton language.
 * 0% loss in accuracy.
 * No change of hardware necessary. Supports Tesla T4, RTX 20, 30, 40 series, A100, H100s