Merge pull request #20 from Monadical-SAS/whisper-jax-gokul

Micro improvements and bug fixes

README.md

@@ -27,10 +27,8 @@ To setup,
 
     ```sh setup_dependencies.sh cuda12```
 
-4) ``` pip install -r requirements.txt```
 
-
-5) Run the Whisper-JAX pipeline. Currently, the repo can take a Youtube video and transcribes/summarizes it.
+4) Run the Whisper-JAX pipeline. Currently, the repo can take a Youtube video and transcribes/summarizes it.
 
 ``` python3 whisjax.py "https://www.youtube.com/watch?v=ihf0S97oxuQ"```
 

requirements.txt

@@ -47,4 +47,5 @@ scattertext
 pandas
 jupyter
 seaborn
-matplotlib
+matplotlib
+termcolor

setup_dependencies.sh

@@ -1,4 +1,4 @@
-# Upgrade pip
+ Upgrade pip
 pip install --upgrade pip
 
 # Default to CPU Installation of JAX
@@ -24,3 +24,10 @@ pip install git+https://github.com/sanchit-gandhi/whisper-jax.git
 # Update to latest version
 pip install --upgrade --no-deps --force-reinstall git+https://github.com/sanchit-gandhi/whisper-jax.git
 
+pip install -r requirements.txt
+
+# download spacy models
+export KMP_DUPLICATE_LIB_OK=True
+python -m spacy download en_core_web_sm
+python -m spacy download en_core_web_md
+

text_utilities.py

@@ -20,8 +20,11 @@ def preprocess_sentence(sentence):
 
 def compute_similarity(sent1, sent2):
     tfidf_vectorizer = TfidfVectorizer()
-    tfidf_matrix = tfidf_vectorizer.fit_transform([sent1, sent2])
-    return cosine_similarity(tfidf_matrix[0], tfidf_matrix[1])[0][0]
+    print("semt1", sent1, sent2)
+    if sent1 is not None and sent2 is not None:
+        tfidf_matrix = tfidf_vectorizer.fit_transform([sent1, sent2])
+        return cosine_similarity(tfidf_matrix[0], tfidf_matrix[1])[0][0]
+    return 0.0
 
 def remove_almost_alike_sentences(sentences, threshold=0.7):
     num_sentences = len(sentences)
@@ -31,12 +34,21 @@ def remove_almost_alike_sentences(sentences, threshold=0.7):
         if i not in removed_indices:
             for j in range(i + 1, num_sentences):
                 if j not in removed_indices:
-                    sentence1 = preprocess_sentence(sentences[i])
-                    sentence2 = preprocess_sentence(sentences[j])
-                    similarity = compute_similarity(sentence1, sentence2)
+                    l_i = len(sentences[i])
+                    l_j = len(sentences[j])
+                    if l_i == 0 or l_j == 0:
+                        if l_i == 0:
+                            removed_indices.add(i)
+                        if l_j == 0:
+                            removed_indices.add(j)
+                    else:
+                        sentence1 = preprocess_sentence(sentences[i])
+                        sentence2 = preprocess_sentence(sentences[j])
+                        if len(sentence1) != 0 and len(sentence2) != 0:
+                            similarity = compute_similarity(sentence1, sentence2)
 
-                    if similarity >= threshold:
-                        removed_indices.add(max(i, j))
+                            if similarity >= threshold:
+                                removed_indices.add(max(i, j))
 
     filtered_sentences = [sentences[i] for i in range(num_sentences) if i not in removed_indices]
     return filtered_sentences
@@ -67,11 +79,14 @@ def remove_whisper_repetitive_hallucination(nonduplicate_sentences):
     return chunk_sentences
 
 def post_process_transcription(whisper_result):
+    transcript_text = ""
     for chunk in whisper_result["chunks"]:
         nonduplicate_sentences = remove_outright_duplicate_sentences_from_chunk(chunk)
         chunk_sentences = remove_whisper_repetitive_hallucination(nonduplicate_sentences)
         similarity_matched_sentences = remove_almost_alike_sentences(chunk_sentences)
         chunk["text"] = " ".join(similarity_matched_sentences)
+        transcript_text += chunk["text"]
+    whisper_result["text"] = transcript_text
     return whisper_result
 
 

viz_utilities.py

@@ -1,5 +1,6 @@
 import matplotlib.pyplot as plt
 from wordcloud import WordCloud, STOPWORDS
+from nltk.corpus import stopwords
 import collections
 import spacy
 import pickle
@@ -11,6 +12,10 @@ import configparser
 config = configparser.ConfigParser()
 config.read('config.ini')
 
+en = spacy.load('en_core_web_md')
+spacy_stopwords = en.Defaults.stop_words
+
+STOPWORDS = set(STOPWORDS).union(set(stopwords.words("english"))).union(set(spacy_stopwords))
 
 def create_wordcloud(timestamp, real_time=False):
     """
@@ -26,13 +31,11 @@ def create_wordcloud(timestamp, real_time=False):
     with open(filename, "r") as f:
         transcription_text = f.read()
 
-    stopwords = set(STOPWORDS)
-
     # python_mask = np.array(PIL.Image.open("download1.png"))
 
     wordcloud = WordCloud(height=800, width=800,
                           background_color='white',
-                          stopwords=stopwords,
+                          stopwords=STOPWORDS,
                           min_font_size=8).generate(transcription_text)
 
     # Plot wordcloud and save image
@@ -192,4 +195,7 @@ def create_talk_diff_scatter_viz(timestamp, real_time=False):
         width_in_pixels=1000,
         transform=st.Scalers.dense_rank
     )
-    open('./scatter_' + timestamp.strftime("%m-%d-%Y_%H:%M:%S") + '.html', 'w').write(html)
+    if real_time:
+        open('./real_time_scatter_' + timestamp.strftime("%m-%d-%Y_%H:%M:%S") + '.html', 'w').write(html)
+    else:
+        open('./scatter_' + timestamp.strftime("%m-%d-%Y_%H:%M:%S") + '.html', 'w').write(html)

whisjax.py

@@ -26,8 +26,8 @@ from file_utilities import upload_files, download_files
 from viz_utilities import create_wordcloud, create_talk_diff_scatter_viz
 from text_utilities import summarize, post_process_transcription
 
-nltk.download('punkt')
-nltk.download('stopwords')
+nltk.download('punkt', quiet=True)
+nltk.download('stopwords', quiet=True)
 
 # Configurations can be found in config.ini. Set them properly before executing
 config = configparser.ConfigParser()
@@ -141,7 +141,8 @@ def main():
                        "transcript_with_timestamp_" + suffix + ".txt",
                        "df_" + suffix + ".pkl",
                        "wordcloud_" + suffix + ".png",
-                       "mappings_" + suffix + ".pkl"]
+                       "mappings_" + suffix + ".pkl",
+                       "scatter_" + suffix + ".html"]
     upload_files(files_to_upload)
 
     summarize(transcript_text, NOW, False, False)

whisjax_realtime.py

@@ -12,7 +12,10 @@ from viz_utilities import create_wordcloud, create_talk_diff_scatter_viz
 from text_utilities import summarize, post_process_transcription
 from loguru import logger
 import nltk
-nltk.download('stopwords')
+import time
+from termcolor import colored
+
+nltk.download('stopwords', quiet=True)
 
 config = configparser.ConfigParser()
 config.read('config.ini')
@@ -68,9 +71,11 @@ def main():
     try:
         while proceed:
             frames = []
+            start_time = time.time()
             for i in range(0, int(RATE / FRAMES_PER_BUFFER * RECORD_SECONDS)):
                 data = stream.read(FRAMES_PER_BUFFER, exception_on_overflow=False)
                 frames.append(data)
+            end_time = time.time()
 
             wf = wave.open(TEMP_AUDIO_FILE, 'wb')
             wf.setnchannels(CHANNELS)
@@ -80,8 +85,6 @@ def main():
             wf.close()
 
             whisper_result = pipeline(TEMP_AUDIO_FILE, return_timestamps=True)
-            print(whisper_result['text'])
-
             timestamp = whisper_result["chunks"][0]["timestamp"]
             start = timestamp[0]
             end = timestamp[1]
@@ -89,12 +92,18 @@ def main():
                 end = start + 15.0
             duration = end - start
             item = {'timestamp': (last_transcribed_time, last_transcribed_time + duration),
-                    'text': whisper_result['text']}
+                    'text': whisper_result['text'],
+                    'stats': (str(end_time - start_time), str(duration))
+                    }
             last_transcribed_time = last_transcribed_time + duration
             transcript_with_timestamp["chunks"].append(item)
-
             transcription += whisper_result['text']
 
+            print(colored("<START>", "yellow"))
+            print(colored(whisper_result['text'], 'green'))
+            print(colored("<END> Recorded duration: " + str(end_time - start_time) + " | Transcribed duration: " +
+                          str(duration), "yellow"))
+
     except Exception as e:
         print(e)
     finally:
@@ -106,10 +115,6 @@ def main():
 
     transcript_with_timestamp = post_process_transcription(transcript_with_timestamp)
 
-    transcript_text = ""
-    for chunk in transcript_with_timestamp["chunks"]:
-        transcript_text += chunk["text"]
-
     logger.info("Creating word cloud")
     create_wordcloud(NOW, True)
 
@@ -122,10 +127,11 @@ def main():
                        "real_time_transcript_with_timestamp" + suffix + ".txt",
                        "real_time_df_" + suffix + ".pkl",
                        "real_time_wordcloud_" + suffix + ".png",
-                       "real_time_mappings_" + suffix + ".pkl"]
+                       "real_time_mappings_" + suffix + ".pkl",
+                       "real_time_scatter_" + suffix + ".html"]
     upload_files(files_to_upload)
 
-    summarize(transcript_text, NOW, True, True)
+    summarize(transcript_with_timestamp["text"], NOW, True, True)
 
     logger.info("Summarization completed")