Updated sbert embedding script with command line arguments

tda/README.md

 # TDA code for Dialogue Term Extraction using Transfer Learning and Topological Data Analysis
 
 This is the Topological Data Analysis portion of the code for the paper 
-Dialogue Term Extraction using Transfer Learning and Topological Data Analysis.
+'Dialogue Term Extraction using Transfer Learning and Topological Data Analysis'.
+
+The scripts in this folder should be executed in the `tda` working directory.
 
 ## Create embeddings
 
-TODO
+Precomputed sbert embeddings are contained in the `/data` folder 
+for the ambient fastText vocabulary, and the joint multiwoz and sgd vocabulary.
+These embeddings are the basis for computing neighborhoods.
+It is not necessary to recompute these embeddings, 
+for the neighborhood extraction and TDA features skip ahead to the next section.
+
+The following command loads the precomputed embeddings 
+of the fastText vocabulary into an interactive python session:
+```bash
+python -i sbert_create_static_embeddings.py \
+  --embeddings_config_path ./sbert_static_embeddings_config_50_0.yaml \
+  --vocab_desc pretrained_cc_en \
+  --load_embeddings
+```
+To compute and save embeddings of the multiwoz and sgd vocabulary:
+```bash
+python sbert_create_static_embeddings.py \
+  --embeddings_config_path ./sbert_static_embeddings_config_50_0.yaml \
+  --vocab_desc multiwoz_and_sgd \
+  --save_embeddings
+```
 
 ## Build neighbourhoods and extract persistence features
 
 TODO
 
-## 8. License
+## License
 This project is licensed under the Apache License, Version 2.0 (the "License");
 you may not use the files except in compliance with the License.
 You may obtain a copy of the License at

tda/sbert_create_static_embeddings.py

@@ -49,6 +49,13 @@ parser = argparse.ArgumentParser()
 # Required parameters
 parser.add_argument("--embeddings_config_path", default=None, type=str, required=True,
                     help="Dataset configuration file.")
+parser.add_argument("--vocab_desc", choices=["multiwoz_and_sgd", "pretrained_cc_en"],
+                    default=None, type=str, required=True,
+                    help="String describing which embeddings should be created.")
+parser.add_argument("--save_embeddings", default=False, action="store_true",
+                    help="Flag to save embeddings to disk.")
+parser.add_argument("--load_embeddings", default=False, action="store_true",
+                    help="Flag to load embeddings from disk.")
 
 # Optional parameters
 # None
@@ -59,22 +66,20 @@ args = parser.parse_args()
 embeddings_config_path = args.embeddings_config_path
 embeddings_config = yaml.safe_load(open(embeddings_config_path))
 
-# # # # # # # # # # # # # # # # #
-# MODIFY: Program parameters
-DEVELOP_MODE = False
+vocab_desc = args.vocab_desc
+# the fields in the config file are named
+# 'pretrained_cc_en_vocabulary_path' and 'multiwoz_and_sgd_vocabulary_path'
+vocab_path = embeddings_config['data'][vocab_desc + '_vocabulary_path']
 
-SAVE_EMBEDDINGS = True
-LOAD_EMBEDDINGS = False
+SAVE_EMBEDDINGS = args.save_embeddings
+LOAD_EMBEDDINGS = args.load_embeddings
 
 data_desc_list = ['paraphrase-MiniLM-L6-v2']
 
+# # # # # # # # # # # # # # # # #
+# Debug parameters
 DEBUG_DATA_DESC = 'paraphrase-MiniLM-L6-v2'
 
-VOCAB_PATH = embeddings_config['data']['multiwoz_and_sgd_joint_vocabulary_path']
-VOCAB_DESC = 'multiwoz_and_sgd'
-# VOCAB_PATH = embeddings_config['data']['pretrained_cc_en_vocabulary_path']
-# VOCAB_DESC = 'pretrained_cc_en'
-
 # Set up logging
 logger = logging.getLogger()
 logger.setLevel(logging.INFO)
@@ -108,12 +113,14 @@ models = {}
 # Pooling:	Mean Pooling
 
 models['paraphrase-MiniLM-L6-v2'] = SentenceTransformer('paraphrase-MiniLM-L6-v2')
+# more models could be added here as additional keys,
+# the embeddings will be computed for each model
 
 # Load data
 
 logging.info('Loading data ...')
 
-with open(VOCAB_PATH, "r") as file:
+with open(vocab_path, "r") as file:
     vocabulary = json.load(file)
 
 # Testing tokenization of certain sentences
@@ -157,7 +164,7 @@ if SAVE_EMBEDDINGS is True and LOAD_EMBEDDINGS is False:
 # save or load embeddings
 try:
     embeddings_path = os.path.join(embeddings_config['embeddings']['embeddings_dict_path'],
-                                   f"{VOCAB_DESC}_vocab_embeddings_sbert.pkl")
+                                   f"{vocab_desc}_vocab_embeddings_sbert.pkl")
     if SAVE_EMBEDDINGS is True:
         logging.info("Saving embeddings ...")
         logging.info(embeddings_path)
@@ -179,12 +186,16 @@ test_vocab = ['.', ':',
               'cheap', 'expensive',
               'xg1tx6q8', '07591624763']
 
+try:
     # Print the embeddings
     for word in test_vocab:
         embedding = embeddings['paraphrase-MiniLM-L6-v2'][word]
         print("Sentence:", word)
         print("Embedding:", embedding)
         print("")
+except Exception as e:
+    logging.error(traceback.format_exc())
+    pass
 
 embedding_vectors = {}
 

tda/sbert_create_static_embeddings_50_0.log

+2022-06-15 12:27:38,561 - root - INFO - Loading config file: ./sbert_static_embeddings_config_50_0.yaml
+2022-06-15 12:27:38,561 - root - INFO - {'data': {'data_folder_path': '../data', 'multiwoz_and_sgd_joint_vocabulary_path': '../data/multiwoz_and_sgd_joint_vocabulary.json', 'pretrained_cc_en_vocabulary_path': '../data/pretrained_cc_en_vocabulary.json'}, 'embeddings': {'embeddings_dict_path': '../data', 'embeddings_dataframes_path': '../data', 'context': 'word', 'pooling_method': 'mean', 'special_tokens': 'ignore'}, 'neighborhoods': {'nbhd_size': 50, 'nbhd_remove': 0, 'neighborhoods_path': '../data/neighborhoods', 'persistence_features_path': '../data', 'normalize': False}}
+2022-06-15 12:27:38,561 - sentence_transformers.SentenceTransformer - INFO - Load pretrained SentenceTransformer: paraphrase-MiniLM-L6-v2
+2022-06-15 12:27:45,634 - sentence_transformers.SentenceTransformer - INFO - Use pytorch device: cpu
+2022-06-15 12:27:45,634 - root - INFO - Loading data ...
+2022-06-15 12:27:45,636 - root - INFO - Creating embeddings...
+2022-06-15 14:50:27,346 - root - INFO - Loading config file: ./sbert_static_embeddings_config_50_0.yaml
+2022-06-15 14:50:27,347 - root - INFO - {'data': {'data_folder_path': '../data', 'multiwoz_and_sgd_vocabulary_path': '../data/multiwoz_and_sgd_joint_vocabulary.json', 'pretrained_cc_en_vocabulary_path': '../data/pretrained_cc_en_vocabulary.json'}, 'embeddings': {'embeddings_dict_path': '../data', 'embeddings_dataframes_path': '../data', 'context': 'word', 'pooling_method': 'mean', 'special_tokens': 'ignore'}, 'neighborhoods': {'nbhd_size': 50, 'nbhd_remove': 0, 'neighborhoods_path': '../data/neighborhoods', 'persistence_features_path': '../data', 'normalize': False}}
+2022-06-15 14:50:27,347 - sentence_transformers.SentenceTransformer - INFO - Load pretrained SentenceTransformer: paraphrase-MiniLM-L6-v2
+2022-06-15 14:50:34,329 - sentence_transformers.SentenceTransformer - INFO - Use pytorch device: cpu
+2022-06-15 14:50:34,329 - root - INFO - Loading data ...
+2022-06-15 14:50:34,469 - root - INFO - Creating embeddings...
+2022-06-15 15:12:27,980 - root - INFO - Loading config file: ./sbert_static_embeddings_config_50_0.yaml
+2022-06-15 15:12:27,980 - root - INFO - {'data': {'data_folder_path': '../data', 'multiwoz_and_sgd_vocabulary_path': '../data/multiwoz_and_sgd_joint_vocabulary.json', 'pretrained_cc_en_vocabulary_path': '../data/pretrained_cc_en_vocabulary.json'}, 'embeddings': {'embeddings_dict_path': '../data', 'embeddings_dataframes_path': '../data', 'context': 'word', 'pooling_method': 'mean', 'special_tokens': 'ignore'}, 'neighborhoods': {'nbhd_size': 50, 'nbhd_remove': 0, 'neighborhoods_path': '../data/neighborhoods', 'persistence_features_path': '../data', 'normalize': False}}
+2022-06-15 15:12:27,980 - sentence_transformers.SentenceTransformer - INFO - Load pretrained SentenceTransformer: paraphrase-MiniLM-L6-v2
+2022-06-15 15:12:34,965 - sentence_transformers.SentenceTransformer - INFO - Use pytorch device: cpu
+2022-06-15 15:12:34,965 - root - INFO - Loading data ...
+2022-06-15 15:12:35,103 - root - INFO - Loading embeddings ...
+2022-06-15 15:12:43,604 - root - INFO - Loading embeddings done.
+2022-06-15 15:13:47,810 - root - INFO - Loading config file: ./sbert_static_embeddings_config_50_0.yaml
+2022-06-15 15:13:47,810 - root - INFO - {'data': {'data_folder_path': '../data', 'multiwoz_and_sgd_vocabulary_path': '../data/multiwoz_and_sgd_joint_vocabulary.json', 'pretrained_cc_en_vocabulary_path': '../data/pretrained_cc_en_vocabulary.json'}, 'embeddings': {'embeddings_dict_path': '../data', 'embeddings_dataframes_path': '../data', 'context': 'word', 'pooling_method': 'mean', 'special_tokens': 'ignore'}, 'neighborhoods': {'nbhd_size': 50, 'nbhd_remove': 0, 'neighborhoods_path': '../data/neighborhoods', 'persistence_features_path': '../data', 'normalize': False}}
+2022-06-15 15:13:47,810 - sentence_transformers.SentenceTransformer - INFO - Load pretrained SentenceTransformer: paraphrase-MiniLM-L6-v2
+2022-06-15 15:13:54,786 - sentence_transformers.SentenceTransformer - INFO - Use pytorch device: cpu
+2022-06-15 15:13:54,786 - root - INFO - Loading data ...
+2022-06-15 15:13:54,911 - root - INFO - Loading embeddings ...
+2022-06-15 15:14:02,418 - root - INFO - Loading embeddings done.
+2022-06-15 15:14:02,438 - root - ERROR - Traceback (most recent call last):
+  File "sbert_create_static_embeddings.py", line 190, in <module>
+    embedding = embeddings['paraphrase-MiniLM-L6-v2'][word]
+KeyError: 'xg1tx6q8'
+
+2022-06-15 15:15:33,524 - root - INFO - Loading config file: ./sbert_static_embeddings_config_50_0.yaml
+2022-06-15 15:15:33,524 - root - INFO - {'data': {'data_folder_path': '../data', 'multiwoz_and_sgd_vocabulary_path': '../data/multiwoz_and_sgd_joint_vocabulary.json', 'pretrained_cc_en_vocabulary_path': '../data/pretrained_cc_en_vocabulary.json'}, 'embeddings': {'embeddings_dict_path': '../data', 'embeddings_dataframes_path': '../data', 'context': 'word', 'pooling_method': 'mean', 'special_tokens': 'ignore'}, 'neighborhoods': {'nbhd_size': 50, 'nbhd_remove': 0, 'neighborhoods_path': '../data/neighborhoods', 'persistence_features_path': '../data', 'normalize': False}}
+2022-06-15 15:15:33,524 - sentence_transformers.SentenceTransformer - INFO - Load pretrained SentenceTransformer: paraphrase-MiniLM-L6-v2
+2022-06-15 15:15:40,440 - sentence_transformers.SentenceTransformer - INFO - Use pytorch device: cpu
+2022-06-15 15:15:40,440 - root - INFO - Loading data ...
+2022-06-15 15:15:40,565 - root - INFO - Loading embeddings ...
+2022-06-15 15:15:48,250 - root - INFO - Loading embeddings done.
+2022-06-15 15:15:48,273 - root - ERROR - Traceback (most recent call last):
+  File "sbert_create_static_embeddings.py", line 190, in <module>
+    embedding = embeddings['paraphrase-MiniLM-L6-v2'][word]
+KeyError: 'xg1tx6q8'
+
+2022-06-15 15:56:59,026 - root - INFO - Loading config file: ./sbert_static_embeddings_config_50_0.yaml
+2022-06-15 15:56:59,026 - root - INFO - {'data': {'data_folder_path': '../data', 'multiwoz_and_sgd_vocabulary_path': '../data/multiwoz_and_sgd_joint_vocabulary.json', 'pretrained_cc_en_vocabulary_path': '../data/pretrained_cc_en_vocabulary.json'}, 'embeddings': {'embeddings_dict_path': '../data', 'embeddings_dataframes_path': '../data', 'context': 'word', 'pooling_method': 'mean', 'special_tokens': 'ignore'}, 'neighborhoods': {'nbhd_size': 50, 'nbhd_remove': 0, 'neighborhoods_path': '../data/neighborhoods', 'persistence_features_path': '../data', 'normalize': False}}
+2022-06-15 15:56:59,026 - sentence_transformers.SentenceTransformer - INFO - Load pretrained SentenceTransformer: paraphrase-MiniLM-L6-v2
+2022-06-15 15:57:05,865 - sentence_transformers.SentenceTransformer - INFO - Use pytorch device: cpu
+2022-06-15 15:57:05,865 - root - INFO - Loading data ...
+2022-06-15 15:57:05,868 - root - INFO - Creating embeddings...

tda/sbert_static_embeddings_config_50_0.yaml

 data:
   data_folder_path: '../data'
-  multiwoz_and_sgd_joint_vocabulary_path: '../data/multiwoz_and_sgd_joint_vocabulary.json'
+  multiwoz_and_sgd_vocabulary_path: '../data/multiwoz_and_sgd_joint_vocabulary.json'
   pretrained_cc_en_vocabulary_path: '../data/pretrained_cc_en_vocabulary.json'
 embeddings:
   embeddings_dict_path: '../data'