Update app.py

app.py

@@ -1,38 +1,89 @@
 import streamlit as st
-from io import BytesIO
-from urllib.request import urlopen
-import librosa
-from transformers import Qwen2AudioForConditionalGeneration, AutoProcessor
-import pyttsx3  # For text-to-speech
-
-# Load Qwen2Audio model and processor
-processor = AutoProcessor.from_pretrained("Qwen/Qwen2-Audio-7B-Instruct")
-model = Qwen2AudioForConditionalGeneration.from_pretrained("Qwen/Qwen2-Audio-7B-Instruct", device_map="auto")
-tts_engine = pyttsx3.init()
+from gtts import gTTS
+import os
+import base64
+import PyPDF2
 
 # Streamlit app UI
 st.title("Text-to-Audio App")
-st.text("This app generates audio from text input using Hugging Face models.")
+st.text("This app converts your text input or PDF content into audio using TTS.")
 
 # User input
-text_input = st.text_area("Enter some text for the model:")
+text_input = st.text_area("Enter some text:")
+
+# PDF file upload
+uploaded_file = st.file_uploader("Upload a PDF file:", type=["pdf"])
+if uploaded_file is not None:
+    try:
+        # Read PDF file
+        pdf_reader = PyPDF2.PdfReader(uploaded_file)
+        extracted_text = ""
+        for page in pdf_reader.pages:
+            extracted_text += page.extract_text()
+
+        if extracted_text.strip():
+            text_input = extracted_text
+            st.success("Text extracted from the uploaded PDF!")
+            st.text_area("Extracted Text:", text_input, height=200)
+        else:
+            st.error("No extractable text found in the uploaded PDF.")
+    except Exception as e:
+        st.error(f"An error occurred while reading the PDF: {e}")
+
 if st.button("Generate Audio"):
-    conversation = [{"role": "user", "content": text_input}]
+    if not text_input.strip():
+        st.error("Please enter some text or upload a PDF with extractable text!")
+    else:
+        try:
+            # Generate speech using gTTS
+            tts = gTTS(text=text_input, lang="en")
+            audio_file = "output.wav"
+            tts.save(audio_file)
+
+            # Check if file exists
+            if os.path.exists(audio_file):
+                # Encode audio file to base64
+                with open(audio_file, "rb") as f:
+                    audio_data = f.read()
+                audio_base64 = base64.b64encode(audio_data).decode()
 
-    # Preprocess conversation
-    text = processor.apply_chat_template(conversation, add_generation_prompt=True, tokenize=False)
-    inputs = processor(text=text, return_tensors="pt", padding=True)
-    inputs.input_ids = inputs.input_ids.to("cuda")
+                # Embed custom HTML audio player with speed adjustment
+                audio_html = f"""
+                    <audio id="audio" controls style="width: 100%; margin-top: 10px;">
+                        <source src="data:audio/wav;base64,{audio_base64}" type="audio/wav">
+                        Your browser does not support the audio element.
+                    </audio>
+                    <div style="margin-top: 10px;">
+                        <label for="speed" style="font-weight: bold;">Playback Speed:</label>
+                        <input type="range" id="speed" min="0.5" max="2.0" value="1.0" step="0.1" style="width: 50%; margin-left: 10px;">
+                        <span id="speed-value">1.0x</span>
+                    </div>
+                    <script>
+                        const audio = document.getElementById("audio");
+                        const speedSlider = document.getElementById("speed");
+                        const speedValue = document.getElementById("speed-value");
 
-    # Generate response
-    generate_ids = model.generate(**inputs, max_length=256)
-    generate_ids = generate_ids[:, inputs.input_ids.size(1):]
+                        // Update playback speed dynamically
+                        speedSlider.addEventListener("input", () => {{
+                            const speed = parseFloat(speedSlider.value);
+                            audio.playbackRate = speed;
+                            speedValue.textContent = speed.toFixed(1) + "x";
+                        }});
+                    </script>
+                """
+                st.components.v1.html(audio_html, height=200)
 
-    # Decode response
-    response = processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
-    st.text(f"Model Response: {response}")
+                st.success("Audio generated successfully!")
 
-    # Convert response to speech
-    tts_engine.say(response)
-    tts_engine.runAndWait()
-    st.success("Audio generated and played!")
+                # Provide download option
+                with open(audio_file, "rb") as f:
+                    st.download_button(
+                        label="Download Audio",
+                        data=f.read(),
+                        file_name="output.wav",
+                        mime="audio/wav",
+                    )
+            else:
+                st.error("Audio file could not be generated.")
+        except Exception as e:
+            st.error(f"An error occurred: {e}")