Merge pull request #1219 from hanhainebula/master

Fix bugs

Author: 545999961

FlagEmbedding/abc/evaluation/data_loader.py

@@ -266,7 +266,7 @@ def _load_local_qrels(self, save_dir: str, dataset_name: Optional[str] = None, s
         Returns:
             datasets.DatasetDict: A dict of relevance of query and document.
         """
-        checked_split = self.check_splits(split)
+        checked_split = self.check_splits(split, dataset_name=dataset_name)
         if len(checked_split) == 0:
             raise ValueError(f"Split {split} not found in the dataset.")
         split = checked_split[0]
@@ -301,7 +301,7 @@ def _load_local_queries(self, save_dir: str, dataset_name: Optional[str] = None,
         Returns:
             datasets.DatasetDict: A dict of queries with id as key, query text as value.
         """
-        checked_split = self.check_splits(split)
+        checked_split = self.check_splits(split, dataset_name=dataset_name)
         if len(checked_split) == 0:
             raise ValueError(f"Split {split} not found in the dataset.")
         split = checked_split[0]

FlagEmbedding/abc/finetune/embedder/AbsDataset.py

@@ -426,6 +426,8 @@ def _get_train_group_size(self, batch_raw_data):
                 return 2, data_type
             elif data_type in ['symmetric_class']:
                 return min(len(batch_raw_data['neg'][0]) + 1, self.args.train_group_size), data_type
+            else:
+                return self.args.train_group_size, data_type
         return self.args.train_group_size, None
 
     def _create_batch_data(self, batch_raw_data):

FlagEmbedding/inference/auto_reranker.py

@@ -54,9 +54,9 @@ def from_finetuned(
             _model_class = RERANKER_CLASS_MAPPING[RerankerModelClass(model_class)]
             if trust_remote_code is None:
                 trust_remote_code = False
-            logging.warning(
-                f"`trust_remote_code` is not specified, set to default value '{trust_remote_code}'."
-            )
+                logging.warning(
+                    f"`trust_remote_code` is not specified, set to default value '{trust_remote_code}'."
+                )
         else:
             if model_name not in AUTO_RERANKER_MAPPING:
                 raise ValueError(