Change config.json to set disable_exllama=false

config.json

@@ -22,7 +22,7 @@
         "block_name_to_quantize": "model.layers",
         "damp_percent": 0.1,
         "desc_act": true,
-        "disable_exllama": true,
+        "disable_exllama": false,
         "group_size": 128,
         "max_input_length": null,
         "model_seqlen": 32768,