[FixBug] Fix bug with loading pretrained lstm and cnn models

Author: jinyongyoo

textattack/attacker.py

@@ -392,7 +392,6 @@ def _attack_parallel(self):
     def attack_dataset(self):
         """Attack the dataset.
 
-
         Returns:
             :obj:`list[AttackResult]` - List of :class:`~textattack.attack_results.AttackResult` obtained after attacking the given dataset..
         """

textattack/model_args.py

@@ -90,24 +90,27 @@
 
 #
 # Models hosted by textattack.
+# `models` vs `models_v2`: `models_v2` is simply a new dir in S3 that contains models' `config.json`.
+# Fixes issue https://github.com/QData/TextAttack/issues/485
+# Model parameters has not changed.
 #
 TEXTATTACK_MODELS = {
     #
     # LSTMs
     #
-    "lstm-ag-news": "models/classification/lstm/ag-news",
-    "lstm-imdb": "models/classification/lstm/imdb",
-    "lstm-mr": "models/classification/lstm/mr",
-    "lstm-sst2": "models/classification/lstm/sst2",
-    "lstm-yelp": "models/classification/lstm/yelp",
+    "lstm-ag-news": "models_v2/classification/lstm/ag-news",
+    "lstm-imdb": "models_v2/classification/lstm/imdb",
+    "lstm-mr": "models_v2/classification/lstm/mr",
+    "lstm-sst2": "models_v2/classification/lstm/sst2",
+    "lstm-yelp": "models_v2/classification/lstm/yelp",
     #
     # CNNs
     #
-    "cnn-ag-news": "models/classification/cnn/ag-news",
-    "cnn-imdb": "models/classification/cnn/imdb",
-    "cnn-mr": "models/classification/cnn/rotten-tomatoes",
-    "cnn-sst2": "models/classification/cnn/sst",
-    "cnn-yelp": "models/classification/cnn/yelp",
+    "cnn-ag-news": "models_v2/classification/cnn/ag-news",
+    "cnn-imdb": "models_v2/classification/cnn/imdb",
+    "cnn-mr": "models_v2/classification/cnn/rotten-tomatoes",
+    "cnn-sst2": "models_v2/classification/cnn/sst",
+    "cnn-yelp": "models_v2/classification/cnn/yelp",
     #
     # T5 for translation
     #

textattack/models/helpers/lstm_for_classification.py

@@ -101,14 +101,17 @@ def from_pretrained(cls, name_or_path):
         """Load trained LSTM model by name or from path.
 
         Args:
-            name_or_path (str): Name of the model (e.g. "lstm-imdb") or model saved via `save_pretrained`.
+            name_or_path (:obj:`str`): Name of the model (e.g. "lstm-imdb") or model saved via :meth:`save_pretrained`.
+        Returns:
+            :class:`~textattack.models.helpers.LSTMForClassification` model
         """
         if name_or_path in TEXTATTACK_MODELS:
             path = utils.download_if_needed(TEXTATTACK_MODELS[name_or_path])
         else:
             path = name_or_path
 
         config_path = os.path.join(path, "config.json")
+
         if os.path.exists(config_path):
             with open(config_path, "r") as f:
                 config = json.load(f)

textattack/models/helpers/word_cnn_for_classification.py

@@ -87,14 +87,17 @@ def from_pretrained(cls, name_or_path):
         """Load trained LSTM model by name or from path.
 
         Args:
-            name_or_path (str): Name of the model (e.g. "cnn-imdb") or model saved via `save_pretrained`.
+            name_or_path (:obj:`str`): Name of the model (e.g. "cnn-imdb") or model saved via :meth:`save_pretrained`.
+        Returns:
+            :class:`~textattack.models.helpers.WordCNNForClassification` model
         """
-        if name_or_path != "cnn" and name_or_path in TEXTATTACK_MODELS:
+        if name_or_path in TEXTATTACK_MODELS:
             path = utils.download_if_needed(TEXTATTACK_MODELS[name_or_path])
         else:
             path = name_or_path
 
         config_path = os.path.join(path, "config.json")
+
         if os.path.exists(config_path):
             with open(config_path, "r") as f:
                 config = json.load(f)