Update app.py

app.py

@@ -1,41 +1,62 @@
-import streamlit as st
-from transformers import pipeline
-
-#def load_summarizer():#
- #   whisper = pipeline('automatic-speech-recognition') #audio-to-text
-  #  summarize = pipeline("summarization", device=0)
-   # senti = pipeline("sentiment-analysis",device=0)
-    #nameentity = pipeline("ner",device=0)
-    #translate = pipeline("translation", device=0)
-    #return whisper, summarize, senti, nameentity, translate
-
-st.subheader("Choose a mp3 file that you extracted from the work site")
-uploaded_file = st.file_uploader("Select file from your directory")
-if uploaded_file is not None:
-    audio_bytes = uploaded_file.read()
-    text = st.audio(audio_bytes, format='audio/mp3')
-
-
-pipe = pipeline("automatic-speech-recognition")
-#text = st.text_area('Enter some Text!')
-
-
-#summarizer = load_summarizer()
-#st.title("Summarize Text")
-#sentence = st.text_area('Please paste your article :', height=30)
-button = st.button("Click")
-
-
-
-if text:
-    out=pipe(text)
-    st.json(out)
-    
-
-    
-    
- 
-    
-    
-    
-    
+%%writefile app.py
+import streamlit as st 
+import soundfile as sf
+import librosa
+from transformers import HubertForCTC, Wav2Vec2Processor , pipeline , Wav2Vec2ForCTC , Wav2Vec2Tokenizer
+import torch
+import spacy 
+from spacy import displacy
+
+st.title('Audio-to-Text')
+
+audio_file = st.file_uploader('Upload Audio' , type=['wav' , 'mp3','m4a'])
+
+if st.button('Trascribe Audio'):
+    if audio_file is not None:
+        processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-base-960h")
+        model = HubertForCTC.from_pretrained("facebook/hubert-large-ls960-ft")  
+        speech, rate = librosa.load(audio_file, sr=16000)
+        input_values = processor(speech, return_tensors="pt", padding="longest", sampling_rate=rate).input_values
+        logits = model(input_values).logits
+        predicted_ids = torch.argmax(logits, dim=-1)
+        text = processor.batch_decode(predicted_ids)
+        st.write(text) 
+    else:
+        st.error('please upload the audio file')
+
+
+
+if st.button('Summarize'):
+        processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-base-960h")
+        model = HubertForCTC.from_pretrained("facebook/hubert-large-ls960-ft")  
+        speech, rate = librosa.load(audio_file, sr=16000)
+        input_values = processor(speech, return_tensors="pt", padding="longest", sampling_rate=rate).input_values
+        logits = model(input_values).logits
+        predicted_ids = torch.argmax(logits, dim=-1)
+        text = processor.batch_decode(predicted_ids)
+        summarize = pipeline("summarization")
+        st.write(summarize(text))
+
+if st.button('sentiment-analysis'):
+        processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-base-960h")
+        model = HubertForCTC.from_pretrained("facebook/hubert-large-ls960-ft")  
+        speech, rate = librosa.load(audio_file, sr=16000)
+        input_values = processor(speech, return_tensors="pt", padding="longest", sampling_rate=rate).input_values
+        logits = model(input_values).logits
+        predicted_ids = torch.argmax(logits, dim=-1)
+        text = processor.batch_decode(predicted_ids)
+        nlp_sa = pipeline("sentiment-analysis")
+        st.write(nlp_sa(text))
+
+if st.button('Name'):
+        processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-base-960h")
+        model = HubertForCTC.from_pretrained("facebook/hubert-large-ls960-ft")  
+        speech, rate = librosa.load(audio_file, sr=16000)
+        input_values = processor(speech, return_tensors="pt", padding="longest", sampling_rate=rate).input_values
+        logits = model(input_values).logits
+        predicted_ids = torch.argmax(logits, dim=-1)
+        text = processor.batch_decode(predicted_ids)
+        str  = ''.join(text)
+        trf = spacy.load('en_core_web_trf')
+        doc=trf(str)
+        print(displacy.render(doc,style='ent'))