diff --git a/examples/run_glue.py b/examples/run_glue.py
index ec76dadcfe..547a4e4698 100644
--- a/examples/run_glue.py
+++ b/examples/run_glue.py
@@ -183,10 +183,10 @@ def evalutate(args, eval_task, eval_output_dir, dataset, model):
         nb_eval_steps += 1
         if preds is None:
             preds = logits.detach().cpu().numpy()
-            out_label_ids = label_ids.detach().cpu().numpy()
+            out_label_ids = inputs['labels'].detach().cpu().numpy()
         else:
             preds = np.append(preds, logits.detach().cpu().numpy(), axis=0)
-            out_label_ids = np.append(out_label_ids, label_ids.detach().cpu().numpy(), axis=0)
+            out_label_ids = np.append(out_label_ids, inputs['labels'].detach().cpu().numpy(), axis=0)
 
     eval_loss = eval_loss / nb_eval_steps
     if args.output_mode == "classification":
diff --git a/pytorch_transformers/modeling_bert.py b/pytorch_transformers/modeling_bert.py
index 71045d1694..7cb723b563 100644
--- a/pytorch_transformers/modeling_bert.py
+++ b/pytorch_transformers/modeling_bert.py
@@ -176,7 +176,7 @@ class BertConfig(PretrainedConfig):
                 initializing all weight matrices.
             layer_norm_eps: The epsilon used by LayerNorm.
     """
-    pretrained_config_archive_map = PRETRAINED_CONFIG_ARCHIVE_MAP
+    pretrained_config_archive_map = BERT_PRETRAINED_CONFIG_ARCHIVE_MAP
 
     def __init__(self,
                  vocab_size_or_config_json_file=30522,