Update app.py

app.py

@@ -1,55 +1,29 @@
-# >>>>>>   Adapted/frankensteined from these scripts:   <<<<<<<
-#                 for Summary Interface:
-# >>>> https://huggingface.co/spaces/khxu/pegasus-text-summarizers/blob/main/app.py
-#                      Audio Interface
-# >>>> https://huggingface.co/spaces/iSky/Speech-audio-to-text-with-grammar-correction/blob/main/app.py
-#                      Gramar
-# >>>> https://huggingface.co/deep-learning-analytics/GrammarCorrector/blob/main/README.md
-
-
-import gradio as gr
+import torch
 from transformers import pipeline
+import gradio as gr
+import streamlit as st
+from transformers import Speech2TextProcessor, Speech2TextForConditionalGeneration
 from gradio.mix import Parallel, Series
 
-# >>>>>>>>>>>>>>>>>>>> Danger Below <<<<<<<<<<<<<<<<<<<<<<
-# Load Interfaces:
-s2t = gr.Interface.load('huggingface/hf-internal-testing/processor_with_lm')
-grammar = gr.Interface.load('huggingface/deep-learning-analytics/GrammarCorrector')
-sum_it = gr.Interface.load('huggingface/SamuelMiller/lil_sum_sum') 
-
-# Audio Functions:
-def out(audio):
-  flag = True
-  if audio==None:
-    return "no audio" 
-  
-  elif flag:               
-    a = s2t(audio)
-    #g = grammar(a)
-    #s = sum_it(g)                              # Summarize Audio with sum_it
-    return a #grammar(a, num_return_sequences=1) # grammar(s),                       # Grammar Filter 
-
-  else:
-    return "something is wrong in the function?"
-
 
-# Construct Interfaces:
-iface = gr.Interface(
-  fn=out, 
-  title="Speech Audio to text (with corrected grammar)",
-  description="Let's Hear It!! This app transforms your speech (input) to text with corrected grammar after (output)!",
-  inputs= gr.inputs.Audio(source="microphone", type="filepath", label=None, optional=True),
-  outputs= 'text'
-)
+# model = Speech2TextForConditionalGeneration.from_pretrained("facebook/s2t-small-librispeech-asr")
+# processor = Speech2TextProcessor.from_pretrained("facebook/s2t-small-librispeech-asr")
+# inputs = processor(ds[0]["audio"]["array"], sampling_rate=ds[0]["audio"]["sampling_rate"], return_tensors="pt")
+# generated_ids = model.generate(inputs["input_features"], attention_mask=inputs["attention_mask"])
+# transcription = processor.batch_decode(generated_ids)
 
-# Launch Interface
-iface.launch(enable_queue=True,show_error=True)
 
-  # From Original Code:
-# gr.inputs.Audio(source="upload", type="filepath", label=None, optional=True),  
-# examples=[["Grammar-Correct-Sample.mp3"], ["Grammar-Wrong-Sample.mp3"],],
+desc = "Is this working or what??"
 
-#def speech_to_text(inp):
-    #pass  # speech recognition model defined here
+def summarize(text):
+  summ = gr.Interface.load('huggingface/google/pegasus-large')      
+  summary = summ(text)
+  return summary
+iface = gr.Interface(fn=summarize,
+                 theme='huggingface', 
+                 title= 'sum_it', 
+                 description= desc,
+                 inputs= "text",
+                 outputs= 'textbox')
+iface.launch(inline = False)
 
-#gr.Interface(speech_to_text, inputs="mic", outputs=gr.Textbox(label="Predicted text", lines=4))