Update README.md

README.md

@@ -15,7 +15,8 @@ This model is an early test of the EverythingLM dataset and some new experimenta
 ### GGML quants: 
 https://huggingface.co/TheBloke/EverythingLM-13B-16K-GGML
 
-I've had trouble with the GGML quants, especially the k-quants act very weird. This is an issue that I will investigate for the next version, for now I recommend GPTQ if possible:
+Make sure to use correct rope scaling settings:
+`-c 16384 --rope-freq-base 10000 --rope-freq-scale 0.25`
 ### GPTQ quants: 
 https://huggingface.co/TheBloke/EverythingLM-13B-16K-GPTQ