Update README.md

README.md

@@ -1,3 +1,7 @@
 ---
 license: llama2
 ---
+
+* llama.cpp has changed the encoding from GGML to GGUF, breaking existing GGML model checkpoints/weights for llama.cpp users:
+  * https://github.com/ggerganov/llama.cpp/pull/2398
+* This is a temporary upload of GGUF encoded Llama-2 models using `llama.cpp/convert-llama-ggmlv3-to-gguf.py` on the GGML models while waiting for official uploads of natively produced GGUF model checkpoints