add translation

app.py

@@ -3,6 +3,7 @@ from st_audiorec import st_audiorec
 
 from nameder import init_model_ner, get_entity_labels
 from speech2text import init_model_trans, transcribe
+from translation import get_translation
 from resources import audit_elapsedtime, set_start
 import subprocess
 
@@ -10,24 +11,30 @@ def main ():
     print("------------------------------")
     print(f"Running main")
 
-    print(subprocess.Popen('pip freeze > requirements_hug.txt', shell=True))
-    s2t = init_model_trans()
-    ner = init_model_ner() #async
+    #print(subprocess.Popen('pip freeze > requirements_hug.txt', shell=True))
+    text = "Tenho uma proposta para a Caixa Geral de Depositos, para 3 consultores outsystems, 300 euros por dia e um periodo de seis meses."
+    st.write(text)
+    traducao = get_translation(text_to_translate=text, languageCode="pt")
+    st.write(traducao)
+    # s2t = init_model_trans()
+    # ner = init_model_ner() #async
 
-    print("Rendering UI...")
-    start_render = set_start()
-    wav_audio_data = st_audiorec()
-    audit_elapsedtime(function="Rendering UI", start=start_render)
+    # print("Rendering UI...")
+    # start_render = set_start()
+    # wav_audio_data = st_audiorec()
+    # audit_elapsedtime(function="Rendering UI", start=start_render)
 
-    if wav_audio_data is not None and s2t is not None:
-        print("Loading data...")
-        start_loading = set_start()
-        st.audio(wav_audio_data, format='audio/wav')
-        text = transcribe(wav_audio_data, s2t)
+    # if wav_audio_data is not None and s2t is not None:
+    #     print("Loading data...")
+    #     start_loading = set_start()
+    #     st.audio(wav_audio_data, format='audio/wav')
+    #     text = transcribe(wav_audio_data, s2t)
+    #     print("translating audio...")
+    #     translation = get_translation("pt")
 
-        if text is not None and ner is not None:    
-            st.write('Entities: ', get_entity_labels(model=ner, text=text))
-        audit_elapsedtime(function="Loading data", start=start_loading)
+    #     if text is not None and ner is not None:    
+    #         st.write('Entities: ', get_entity_labels(model=ner, text=text))
+    #     audit_elapsedtime(function="Loading data", start=start_loading)
 
 if __name__ == "__main__":
     print("IN __name__")

speech2text.py

@@ -43,7 +43,7 @@ def transcribe (audio_sample: bytes, pipe) -> str:
     # sample = dataset[0]["audio"]
     result = pipe(audio_sample)
     audit_elapsedtime(function="Transcription", start=start)
-    print(result)
-    
-    st.write('trancription: ', result["text"])
+    print("transcription result",result)
+
+    #st.write('trancription: ', result["text"])
     return result["text"]

translation.py

@@ -0,0 +1,20 @@
+from transformers import MarianMTModel, MarianTokenizer
+
+def get_model_name(languageCode: str) -> str:
+    model_name = "Helsinki-NLP/opus-mt-pt-en"
+    return model_name
+
+def init_translation_model():
+    model_name = get_model_name("pt")
+    model = MarianMTModel.from_pretrained(model_name)
+    tokenizer = MarianTokenizer.from_pretrained(model_name)
+    return model, tokenizer
+
+def get_translation(text_to_translate: str, languageCode: str) -> str:
+    model, tokenizer = init_translation_model()
+    inputs = tokenizer(text_to_translate, return_tensors="pt", truncation=True, padding=True)
+    translated_ids = model.generate(**inputs)
+    translated_text = tokenizer.decode(translated_ids[0], skip_special_tokens=True)
+
+    print("Translated text:", translated_text)
+    return translated_text