diff --git a/README.md b/README.md
index e49598a..70250f1 100644
--- a/README.md
+++ b/README.md
@@ -60,7 +60,7 @@ see `README_tokenizer.md` for further information.
 #### Simple use
 
 ```python
-from convlm.tokenizer import SpokenDialogTokenizer
+from turngpt.tokenizer  import SpokenDialogTokenizer
 
 pretrained_model_name_or_path="microsoft/DialoGPT-small"
 tokenizer = SpokenDialogTokenizer(pretrained_model_name_or_path)
@@ -110,7 +110,7 @@ An un-trained TurnGPT model, loads pre-trained weights by default, and includes
 
 ```python
   from argparse import ArgumentParser
-  from convlm.turngpt import TurnGPT
+  from turngpt import TurnGPT
 
   parser = ArgumentParser()
   parser = TurnGPT.add_model_specific_args(parser)
diff --git a/turngpt/model.py b/turngpt/model.py
index a3b22d8..66db279 100644
--- a/turngpt/model.py
+++ b/turngpt/model.py
@@ -68,7 +68,7 @@ def idx_to_string(self, idx):
             idx = idx.item()
         s = self.tokenizer.convert_ids_to_tokens(idx)
         s = self.tokenizer.convert_tokens_to_string(
-            s.strip()
+            [s.strip()]
         )  # remove prefix space/symbol
         return s
 
diff --git a/turngpt/train.py b/turngpt/train.py
index 6272450..b5b8018 100644
--- a/turngpt/train.py
+++ b/turngpt/train.py
@@ -6,7 +6,7 @@
 
 import pytorch_lightning as pl
 
-from datasets_turntaking import DialogTextDM
+from datasets_turntaking import ConversationalDM as DialogTextDM
 from turngpt.model import TurnGPT, TurnGPTWandbCallbacks