The layout configs are looked up in ./configs, so the converter only works
when it is started from its own source directory. Resolve them next to the
q4nx package instead, which lets it run from any directory once installed.

--- a/q4nx/model_converter.py
+++ b/q4nx/model_converter.py
@@ -97,7 +97,9 @@
                 self.num_layers = field.contents()
 
 
-    def _load_config(self, config_file_path: str = "configs"):
+    def _load_config(self, config_file_path: str = None):
+        if config_file_path is None:
+            config_file_path = os.path.join(os.path.dirname(os.path.dirname(os.path.abspath(__file__))), "configs")
         config_path = os.path.join(config_file_path, ModelArchConfigs[self.model_arch])
         print(f"[INFO] Loading Q4NX config from {config_path}")
         self.q4nx_config = json.load(open(config_path))