added fallback for incomplete configs for vlm models launched as llms (#828)

Co-authored-by: Nathan Habib <[email protected]>

Author: clefourrier

src/lighteval/models/transformers/transformers_model.py

@@ -468,13 +468,16 @@ def _init_max_length(self) -> int:
             return self.config.max_length
 
         # Try to get the sequence length from the model config.
+        text_model_config = self.transformers_config.get_text_config()
+
         seqlen_config_attrs = ("n_positions", "max_position_embeddings", "n_ctx")
         for attr in seqlen_config_attrs:
-            if hasattr(self.transformers_config, attr):
-                return getattr(self.transformers_config, attr)
+            if hasattr(text_model_config, attr):
+                return getattr(text_model_config, attr)
 
         logger.warning(
-            "No max_length attribute found in the model config. Using the default max sequence length setting {2048}. It is recomended to set max_length through the model args"
+            "No max_length attribute found in the model config. Using the default max sequence length setting `2048`. "
+            "It is recommended to set max_length trough the model args: max_length=..."
         )
 
         return 2048