Remove quantization config and fix vocab size

config.json

@@ -24,12 +24,5 @@
   "transformers_version": "4.56.1",
   "use_cache": true,
   "use_sliding_window": false,
-  "vocab_size": 152064,
-  "quantization_config": {
-    "quant_method": "bitsandbytes",
-    "load_in_4bit": true,
-    "bnb_4bit_compute_dtype": "bfloat16",
-    "bnb_4bit_quant_type": "nf4",
-    "bnb_4bit_use_double_quant": true
-  }
+  "vocab_size": 152064
 }