feat: fix tests and update release version to 0.9.0 (deeppavlov#1179)

* chore: update required transformers version * chore: update version number to 0.9.0 * chore: rebuild registry.json * tests: add tests for imdb sentiment * tests: test imdb models on TI only as there is no pretrained model * fix: generate inputs for inferring testing * tests: fix return types on test cases * fix: install go_bot requirements in go_bot tutorial * fix: valid chainer outputs for elmo_1b_benchmark configs * tests: test go_bot_minimal on a dialog input
molodiuc · Apr 19, 2020 · 444d484 · 444d484
1 parent 485ea8f
commit 444d484
Show file tree

Hide file tree

Showing 9 changed files with 21 additions and 22 deletions.
diff --git a/deeppavlov/__init__.py b/deeppavlov/__init__.py
@@ -39,7 +39,7 @@ def evaluate_model(config: [str, Path, dict], download: bool = False, recursive:
 except ImportError:
     'Assuming that requirements are not yet installed'
 
-__version__ = '0.8.0'
+__version__ = '0.9.0'
 __author__ = 'Neural Networks and Deep Learning lab, MIPT'
 __description__ = 'An open source library for building end-to-end dialog systems and training chatbots.'
 __keywords__ = ['NLP', 'NER', 'SQUAD', 'Intents', 'Chatbot']

diff --git a/deeppavlov/configs/classifiers/sentiment_imdb_bert.json b/deeppavlov/configs/classifiers/sentiment_imdb_bert.json
@@ -101,7 +101,6 @@
     "metrics": [
       "f1_weighted",
       "f1_macro",
-      "sets_accuracy",
       {
         "name": "roc_auc",
         "inputs": [

diff --git a/deeppavlov/configs/classifiers/sentiment_imdb_conv_bert.json b/deeppavlov/configs/classifiers/sentiment_imdb_conv_bert.json
@@ -101,7 +101,6 @@
     "metrics": [
       "f1_weighted",
       "f1_macro",
-      "sets_accuracy",
       {
         "name": "roc_auc",
         "inputs": [

diff --git a/deeppavlov/configs/elmo/elmo_1b_benchmark.json b/deeppavlov/configs/elmo/elmo_1b_benchmark.json
@@ -38,8 +38,7 @@
       }
     ],
     "out": [
-      "x_char_ids",
-      "y_token_ids"
+      "x_char_ids"
     ]
   },
   "train": {
@@ -83,4 +82,4 @@
       }
     ]
   }
-}
+}
diff --git a/deeppavlov/configs/elmo/elmo_1b_benchmark_test.json b/deeppavlov/configs/elmo/elmo_1b_benchmark_test.json
@@ -40,8 +40,7 @@
       }
     ],
     "out": [
-      "x_char_ids",
-      "y_token_ids"
+      "x_char_ids"
     ]
   },
   "train": {
@@ -81,4 +80,4 @@
       }
     ]
   }
-}
+}
diff --git a/deeppavlov/core/common/registry.json b/deeppavlov/core/common/registry.json
@@ -63,8 +63,8 @@
   "glove": "deeppavlov.models.embedders.glove_embedder:GloVeEmbedder",
   "go_bot": "deeppavlov.models.go_bot.network:GoalOrientedBot",
   "hashing_tfidf_vectorizer": "deeppavlov.models.vectorizers.hashing_tfidf_vectorizer:HashingTfIdfVectorizer",
-  "imdb_reader": "deeppavlov.dataset_readers.imdb_reader:ImdbReader",
   "hybrid_ner_model": "deeppavlov.models.ner.NER_model:HybridNerModel",
+  "imdb_reader": "deeppavlov.dataset_readers.imdb_reader:ImdbReader",
   "insurance_reader": "deeppavlov.dataset_readers.insurance_reader:InsuranceReader",
   "joint_tagger_parser": "deeppavlov.models.syntax_parser.joint:JointTaggerParser",
   "kb_answer_parser_wikidata": "deeppavlov.models.kbqa.kb_answer_parser_wikidata:KBAnswerParserWikidata",

diff --git a/deeppavlov/requirements/transformers.txt b/deeppavlov/requirements/transformers.txt
@@ -1 +1 @@
-transformers==2.5.0
+transformers==2.8.0
diff --git a/examples/gobot_extended_tutorial.ipynb b/examples/gobot_extended_tutorial.ipynb
@@ -54,7 +54,8 @@
    },
    "outputs": [],
    "source": [
-    "!pip install deeppavlov"
+    "!pip install deeppavlov\n",
+    "!python -m deeppavlov install gobot_simple_dstc2"
    ]
   },
   {

diff --git a/tests/test_quick_start.py b/tests/test_quick_start.py
@@ -55,8 +55,8 @@
 # Mapping from model name to config-model_dir-ispretrained and corresponding queries-response list.
 PARAMS = {
     "ecommerce_skill": {
-        ("ecommerce_skill/bleu_retrieve.json", "ecommerce_skill_bleu", ALL_MODES): [('Dummy text', '[]', '{}', None)],
-        ("ecommerce_skill/tfidf_retrieve.json", "ecommerce_skill_tfidf", ALL_MODES): [('Dummy text', '[]', '{}', None)]
+        ("ecommerce_skill/bleu_retrieve.json", "ecommerce_skill_bleu", ALL_MODES): [('Dummy text', [], {}, None)],
+        ("ecommerce_skill/tfidf_retrieve.json", "ecommerce_skill_tfidf", ALL_MODES): [('Dummy text', [], {}, None)]
     },
     "faq": {
         ("faq/tfidf_logreg_en_faq.json", "faq_tfidf_logreg_en", ALL_MODES): [ONE_ARGUMENT_INFER_CHECK],
@@ -85,7 +85,7 @@
     "go_bot": {
         ("go_bot/gobot_dstc2.json", "gobot_dstc2", ALL_MODES): [ONE_ARGUMENT_INFER_CHECK],
         ("go_bot/gobot_dstc2_best.json", "gobot_dstc2_best", ALL_MODES): [ONE_ARGUMENT_INFER_CHECK],
-        ("go_bot/gobot_dstc2_minimal.json", "gobot_dstc2_minimal", ('TI',)): [ONE_ARGUMENT_INFER_CHECK]
+        ("go_bot/gobot_dstc2_minimal.json", "gobot_dstc2_minimal", ('TI',)): [([{"text": "the weather is clooudy and gloooomy"}], None)]
     },
     "classifiers": {
         ("classifiers/paraphraser_bert.json", "classifiers", ('IP', 'TI')): [TWO_ARGUMENTS_INFER_CHECK],
@@ -107,13 +107,15 @@
         ("classifiers/yahoo_convers_vs_info.json", "classifiers", ('IP',)): [ONE_ARGUMENT_INFER_CHECK],
         ("classifiers/ru_obscenity_classifier.json", "classifiers", ('IP',)):
             [
-                ("Ну и сука же она", 'True'),
-                ("я два года жду эту игру", 'False')
+                ("Ну и сука же она", True),
+                ("я два года жду эту игру", False)
             ],
         ("classifiers/sentiment_sst_conv_bert.json", "classifiers", ('IP',)): [ONE_ARGUMENT_INFER_CHECK],
         ("classifiers/sentiment_sst_multi_bert.json", "classifiers", ('IP',)): [ONE_ARGUMENT_INFER_CHECK],
         ("classifiers/sentiment_yelp_conv_bert.json", "classifiers", ('IP',)): [ONE_ARGUMENT_INFER_CHECK],
-        ("classifiers/sentiment_yelp_multi_bert.json", "classifiers", ('IP',)): [ONE_ARGUMENT_INFER_CHECK]
+        ("classifiers/sentiment_yelp_multi_bert.json", "classifiers", ('IP',)): [ONE_ARGUMENT_INFER_CHECK],
+        ("classifiers/sentiment_imdb_bert.json", "classifiers", ('TI',)): [ONE_ARGUMENT_INFER_CHECK],
+        ("classifiers/sentiment_imdb_conv_bert.json", "classifiers", ('TI',)): [ONE_ARGUMENT_INFER_CHECK]
     },
     "snips": {
         ("classifiers/intents_snips.json", "classifiers", ('TI',)): [ONE_ARGUMENT_INFER_CHECK],
@@ -152,9 +154,9 @@
         ("ner/ner_rus.json", "ner_rus", ('IP',)): [ONE_ARGUMENT_INFER_CHECK],
         ("ner/slotfill_dstc2.json", "slotfill_dstc2", ('IP',)):
             [
-                ("chinese food", "{'food': 'chinese'}"),
-                ("in the west part", "{'area': 'west'}"),
-                ("moderate price range", "{'pricerange': 'moderate'}")
+                ("chinese food", {'food': 'chinese'}),
+                ("in the west part", {'area': 'west'}),
+                ("moderate price range", {'pricerange': 'moderate'})
             ]
     },
     "sentence_segmentation": {
@@ -386,7 +388,7 @@ class TestQuickStart(object):
     @staticmethod
     def infer(config_path, qr_list=None, check_outputs=True):
 
-        *inputs, expected_outputs = zip(*qr_list) if qr_list else [],
+        *inputs, expected_outputs = zip(*qr_list) if qr_list else ([],)
         with ProcessPoolExecutor(max_workers=1) as executor:
             f = executor.submit(_infer, config_path, inputs)
         outputs = f.result()