Update app.py

app.py

@@ -4,143 +4,72 @@ import gradio as gr
 import torch
 from diffusers import DiffusionPipeline
 
-"""### **Arabic: Text-Generation:**
-Generate a poetry in Arabic.
-"""
-
 pipe_ar = pipeline('text-generation', framework='pt', model='akhooli/ap2023', tokenizer='akhooli/ap2023')
-
-"""### **English: Text-Generation:**
-Generate a poetry in English.
-"""
-
 pipe_en = pipeline("text-generation", model="ismaelfaro/gpt2-poems.en")
-
-"""### **Arabic and English: Text-To-Speech:**
-Convert the Arabic/English poetry to speech.
-"""
+pipe_image = DiffusionPipeline.from_pretrained("runwayml/stable-diffusion-v1-5")
+pipe_translator = pipeline("translation", model="Helsinki-NLP/opus-mt-ar-en")
 
 # Initialize text-to-speech models for Arabic and English
 # Arabic: text-to-speech
-synthesiser_arabic = pipeline("text-to-speech", model="MBZUAI/speecht5_tts_clartts_ar")
-embeddings_dataset_arabic = load_dataset("herwoww/arabic_xvector_embeddings", split="validation")
-speaker_embedding_arabic = torch.tensor(embeddings_dataset_arabic[105]["speaker_embeddings"]).unsqueeze(0)
+synthesiser_arabic = pipeline("text-to-speech", model="facebook/mms-tts-ara")
 
 # English: text-to-speech
 synthesiser_english = pipeline("text-to-speech", model="microsoft/speecht5_tts")
 embeddings_dataset_english = load_dataset("Matthijs/cmu-arctic-xvectors", split="validation")
 speaker_embedding_english = torch.tensor(embeddings_dataset_english[7306]["xvector"]).unsqueeze(0)
 
-"""### **English Text-To-Image:**
-Convert the starter of the English poetry to an image.
-"""
-
-pipe_image = DiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4")
-
-"""### **Translator from Arabic to English:**
-The text-to-image model doesn't support Arabic, therefore we need to translate the starter of the Arabic poetry to English in order to generate image.
-"""
-
-pipe_translator = pipeline("translation", model="Helsinki-NLP/opus-mt-ar-en")
-
-"""### **Primary Function:**
-This function will receive 2 inputs from the Gradio interface, and execute the following functions and return 3 outputs:
-1. The generated poem.
-2. The audio.
-3. The image.
-"""
-
+# Generate poem based on language and convert it to audio and image
 def generate_poem(selected_language, text):
-    try:
-        if selected_language == "English":
-            poem = generate_poem_english(text)
-            sampling_rate, audio_data = text_to_speech_english(poem)
-            image = generate_image_from_poem(text)
-        elif selected_language == "Arabic":
-            poem = generate_poem_arabic(text)
-            sampling_rate, audio_data = text_to_speech_arabic(poem)
-            translated_text = translate_arabic_to_english(text)
-            image = generate_image_from_poem(translated_text)
-
-        return poem, (sampling_rate, audio_data), image
-    except Exception as e:
-        return f"Error: {str(e)}", None, None
-
-"""### **Poem Generation Function:**
-This function is responsible for generating a poem (text) in Arabic or English, based on the provided text.
-"""
+    if selected_language == "English":
+        poem = generate_poem_english(text) #retrun the generated poem from the generate_poem_english function
+        sampling_rate, audio_data = text_to_speech_english(poem) #return the audio from the text_to_speech_english function
+        image = generate_image_from_poem(poem) #return the image from the generate_image_from_poem function
+    elif selected_language == "Arabic":
+        poem = generate_poem_arabic(text) #retrun the generated poem from the generate_poem_arabic function
+        sampling_rate, audio_data = text_to_speech_arabic(poem) #return the audio from the text_to_speech_arabic function
+        translated_text = translate_arabic_to_english(poem) #return the translated poem from arabic to englsih, using translate_arabic_to_english function
+        image = generate_image_from_poem(translated_text) #return the image from the generate_image_from_poem function
+
+    return poem, (sampling_rate, audio_data), image
 
 # Poem generation for Arabic
 def generate_poem_arabic(text):
-    temp = 1.0
-    topk = 50
-    topp = 0.9
-    penalty = 1.2
-    generated_text = pipe_ar(
-        text,
-        max_length=96,
-        do_sample=True,
-        temperature=temp,
-        top_k=topk,
-        top_p=topp,
-        repetition_penalty=penalty,
-        min_length=64,
-        no_repeat_ngram_size=3,
-        return_full_text=True,
-        num_beams=5,
-        num_return_sequences=1
-    )[0]["generated_text"]
-    clean_text = generated_text.replace("-", "")  # To get rid of the dashes generated by the model.
+    generated_text = pipe_ar(text, do_sample=True, max_length=96, top_k=50, top_p=1.0, temperature=1.0, num_return_sequences=1,
+                              no_repeat_ngram_size = 3, return_full_text=True)[0]["generated_text"]
+    clean_text = generated_text.replace("-", "") #To get rid of the dashs generated by the model.
     return clean_text
 
 # Poem generation for English
 def generate_poem_english(text):
-    generated_text = pipe_en(
-        text,
-        do_sample=True,
-        max_length=100,
-        top_k=50,
-        top_p=0.9,
-        temperature=1.0,
-        num_return_sequences=1
-    )[0]['generated_text']
-    clean_text = generated_text.replace("</s>", "")  # To get rid of the </s> generated by the model.
+    generated_text = pipe_en(text, do_sample=True, max_length=50)[0]['generated_text']
+    clean_text = generated_text.replace("-", "")  # Remove dashes generated by the model
+    clean_text = clean_text.replace("\\n", " ")  # Replace newlines with a space
     return clean_text
 
-"""### **Audio Function:**
-This function is responsible for generating audio in Arabic or English, based on the provided text.
-"""
-
+# Text-to-speech conversion for Arabic
 def text_to_speech_arabic(text):
-    speech = synthesiser_arabic(text, speaker_embeddings=speaker_embedding_arabic)
-    audio_data = speech["audio"]
+    speech = synthesiser_arabic(text)
+    audio_data = speech["audio"][0]  # Flatten to 1D
     sampling_rate = speech["sampling_rate"]
     return (sampling_rate, audio_data)
 
+# Text-to-speech conversion for English
 def text_to_speech_english(text):
-    speech = synthesiser_english(text, speaker_embeddings=speaker_embedding_english)
+    speech = synthesiser_english(text, forward_params={"speaker_embeddings": speaker_embedding_english})
     audio_data = speech["audio"]
     sampling_rate = speech["sampling_rate"]
     return (sampling_rate, audio_data)
 
-"""### **Image Function:**
-This function is responsible for generating an image based on the provided text.
-"""
-
+#Image Function
 def generate_image_from_poem(poem_text):
     image = pipe_image(poem_text).images[0]
     return image
 
-"""### **Translation Function:**
-This function is responsible for translating Arabic input to English, to be used for the image function, which accepts only English inputs.
-"""
-
+#Translation Function from Arabic to English
 def translate_arabic_to_english(text):
     translated_text = pipe_translator(text)[0]['translation_text']
     return translated_text
 
-"""### **CSS Styling:**"""
-
 custom_css = """
 body {
     background-color: #f4f4f9;
@@ -155,7 +84,6 @@ label {
     color: #4A90E2;
     font-weight: bold;
 }
-
 input[type="text"],
 textarea {
     border: 1px solid #4A90E2;
@@ -163,7 +91,6 @@ textarea {
 textarea {
     height: 150px;
 }
-
 button {
     background-color: #4A90E2;
     color: #fff;
@@ -173,40 +100,30 @@ button {
 button:hover {
     background-color: #357ABD;
 }
-
 .dropdown {
     border: 1px solid #4A90E2;
     border-radius: 4px;
 }
 """
+#First parameter is for the dropdown menu, and the second parameter is for the starter of the poem
+examples = [["Arabic", "الورود تتفتح في فصل الربيع"]]
 
-"""### **Examples for Gradio:**
-Provide 4 predefined inputs to demonstrate how the interface works.
-"""
-
-examples = [
-    ["English", "The shining sun rises over the calm ocean"],
-    ["Arabic", "الورود تتفتح في الربيع"],
-    ["English", "The night sky is filled with stars and dreams"],
-    ["Arabic", "أشعة الشمس المشرقة"]
-]
-
-"""### **Gradio Interface:**
-Creating a Gradio interface to generate a poem, read the poem, and generate an image based on that poem.
-"""
 
 my_model = gr.Interface(
-    fn=generate_poem,
+    fn=generate_poem,  #The primary function that will recives the inputs (language and the starter of the poem)
     inputs=[
-        gr.Dropdown(["English", "Arabic"], label="Select Language"),
-        gr.Textbox(label="Enter a sentence")
-    ],
+        gr.Dropdown(["English", "Arabic"], label="Select Language"), #Dropdown menu to select the language, either "English" or "Arabic" for the poem
+        gr.Textbox(label="Enter a sentence")], #Textbox where the user will input a sentence or phrase to generate the poem (starter of the peom)
+
     outputs=[
-        gr.Textbox(label="Generated Poem", lines=10),
-        gr.Audio(label="Generated Audio", type="numpy"),
-        gr.Image(label="Generated Image")
-    ],
-    examples=examples,
-    css=custom_css
+        gr.Textbox(label="Generated Poem", lines=10), # Textbox to display the generated poem
+        gr.Audio(label="Generated Audio", type="numpy"), #Audio output for the generated poem
+        gr.Image(label="Generated Image")], #Display an image generated from the starter of the peom
+
+    examples=examples,  #Predefined examples to guide the user how to use the interface
+    css=custom_css  #Applying CSS Custeom
 )
 my_model.launch()
+
+if __name__ == "__main__":
+    app.run(host="0.0.0.0", port=7860)