Add model config

config.json

@@ -0,0 +1,7 @@
+{
+  "max_length": 128,
+  "model_type": "llama",
+  "quantized": true,
+  "task": "end-of-utterance-detection",
+  "version": "1.0.0"
+}