GGML model commit (made with llama.cpp commit dd1d575)

config.json

@@ -0,0 +1,3 @@
+{
+    "model_type": "gpt-neox"
+}