Update app.py

app.py

@@ -35,9 +35,8 @@ iface = gr.Interface(
     theme="default",
     layout="vertical",
     live=False
-)import gradio as gr
-import warnings
-import torch
+
+    
 from transformers import AutoProcessor, AutoModelForSpeechSeq2Seq, AutoTokenizer, AutoModelForSeq2SeqLM
 from pydub import AudioSegment
 import soundfile as sf
@@ -49,23 +48,21 @@ import time
 
 nltk.download('punkt')
 
-warnings.filterwarnings("ignore")
-
-# Load processor and model for transcription
+# transcription
 processor = AutoProcessor.from_pretrained("NbAiLabBeta/nb-whisper-large-semantic")
 transcription_model = AutoModelForSpeechSeq2Seq.from_pretrained("NbAiLabBeta/nb-whisper-large-semantic")
 
-# Load tokenizer and model for summarization
+# summarization
 summarization_tokenizer = AutoTokenizer.from_pretrained("NbAiLab/norbert-summarization")
 summarization_model = AutoModelForSeq2SeqLM.from_pretrained("NbAiLab/norbert-summarization")
 
-# Set up the device
+# setup
 device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
 torch_dtype = torch.float32
 
-# Move the models to the device
+# move 'em
 transcription_model.to(device)
-summarization_model.to(device)
+summarization_model.to(device)  # PS. model needs to be told to use graph-based summary method (Lexname?)
 
 def convert_to_wav(audio_file):
     audio = AudioSegment.from_file(audio_file, format="m4a")
@@ -116,7 +113,7 @@ def transcribe_audio(audio_file, batch_size=4):
     summary = summarization_tokenizer.decode(summary_ids[0], skip_special_tokens=True)
     return summary
 
-    # HTML for banner and additional image
+    # HTML syntax for imagery
 banner_html = """
 <div style="text-align: center;">
     <img src="https://huggingface.co/spaces/camparchimedes/ola_s-audioshop/raw/main/Olas%20AudioSwitch%20Shop.png" alt="Banner" width="87%" height="auto">
@@ -126,7 +123,7 @@ banner_html = """
 </div>
 """
 
-# Create Gradio interface
+# Gradio UI
 iface = gr.Blocks()
 
 with iface:
@@ -151,13 +148,13 @@ def save_to_pdf(transcription, summary):
     pdf.add_page()
     pdf.set_font("Arial", size=12)
     
-    # Add transcription
+    # incøude transcription
     pdf.multi_cell(0, 10, "Transcription:\n" + transcription)
     
-    # Add a space between transcription and summary
+    # paragraph space
     pdf.ln(10)
     
-    # Add summary
+    # include summary
     pdf.multi_cell(0, 10, "Summary:\n" + summary)
     
     pdf_output_path = "transcription_summary.pdf"
@@ -166,11 +163,9 @@ def save_to_pdf(transcription, summary):
 
 
     
-# Launch the interface
+# run
 iface.launch(share=True, debug=True)
 
-# Launch the interface
-iface.launch(share=True, debug=True)