Duplicate from storresbusquets/demo1

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,14 @@
+---
+title: Demo1
+emoji: 🚀
+colorFrom: blue
+colorTo: indigo
+sdk: gradio
+sdk_version: 3.42.0
+app_file: app.py
+pinned: false
+license: mit
+duplicated_from: storresbusquets/demo1
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,156 @@
+import gradio as gr
+import whisper
+from pytube import YouTube
+from transformers import pipeline, T5Tokenizer, T5ForConditionalGeneration, AutoTokenizer, AutoModelForSeq2SeqLM
+
+class GradioInference():
+    def __init__(self):
+        self.sizes = list(whisper._MODELS.keys())
+        self.langs = ["none"] + sorted(list(whisper.tokenizer.LANGUAGES.values()))
+        self.current_size = "base"
+        self.loaded_model = whisper.load_model(self.current_size)
+        self.yt = None
+        self.summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
+                
+        # Initialize VoiceLabT5 model and tokenizer
+        self.keyword_model = T5ForConditionalGeneration.from_pretrained("Voicelab/vlt5-base-keywords")
+        self.keyword_tokenizer = T5Tokenizer.from_pretrained("Voicelab/vlt5-base-keywords")
+
+        # Sentiment Classifier
+        self.classifier = pipeline("text-classification")
+
+    def __call__(self, link, lang, size):
+        if self.yt is None:
+            self.yt = YouTube(link)
+        path = self.yt.streams.filter(only_audio=True)[0].download(filename="tmp.mp4")
+
+        if lang == "none":
+            lang = None
+
+        if size != self.current_size:
+            self.loaded_model = whisper.load_model(size)
+            self.current_size = size
+        
+        results = self.loaded_model.transcribe(path, language=lang)
+        
+        # Perform summarization on the transcription
+        transcription_summary = self.summarizer(results["text"], max_length=130, min_length=30, do_sample=False)
+
+        # Extract keywords using VoiceLabT5
+        task_prefix = "Keywords: "
+        input_sequence = task_prefix + results["text"]
+        input_ids = self.keyword_tokenizer(input_sequence, return_tensors="pt", truncation=False).input_ids
+        output = self.keyword_model.generate(input_ids, no_repeat_ngram_size=3, num_beams=4)
+        predicted = self.keyword_tokenizer.decode(output[0], skip_special_tokens=True)
+        keywords = [x.strip() for x in predicted.split(',') if x.strip()]
+
+        label = self.classifier(results["text"])[0]["label"]
+        
+        return results["text"], transcription_summary[0]["summary_text"], keywords, label
+
+    def populate_metadata(self, link):
+        self.yt = YouTube(link)
+        return self.yt.thumbnail_url, self.yt.title
+
+
+    def from_audio_input(self, lang, size, audio_file):
+        if lang == "none":
+            lang = None
+        
+        if size != self.current_size:
+            self.loaded_model = whisper.load_model(size)
+            self.current_size = size
+
+        results = self.loaded_model.transcribe(audio_file, language=lang)
+
+        # Perform summarization on the transcription
+        transcription_summary = self.summarizer(results["text"], max_length=130, min_length=30, do_sample=False)
+
+        # Extract keywords using VoiceLabT5
+        task_prefix = "Keywords: "
+        input_sequence = task_prefix + results["text"]
+        input_ids = self.keyword_tokenizer(input_sequence, return_tensors="pt", truncation=False).input_ids
+        output = self.keyword_model.generate(input_ids, no_repeat_ngram_size=3, num_beams=4)
+        predicted = self.keyword_tokenizer.decode(output[0], skip_special_tokens=True)
+        keywords = [x.strip() for x in predicted.split(',') if x.strip()]
+
+        label = self.classifier(results["text"])[0]["label"]
+        
+        return results["text"], transcription_summary[0]["summary_text"], keywords, label
+
+
+gio = GradioInference()
+title = "Youtube Insights"
+description = "Your AI-powered video analytics tool"
+
+block = gr.Blocks()
+with block as demo:
+    gr.HTML(
+        """
+        <div style="text-align: center; max-width: 500px; margin: 0 auto;">
+          <div>
+            <h1>Youtube <span style="color: red;">Insights</span> 📹</h1>
+          </div>
+          <p style="margin-bottom: 10px; font-size: 94%">
+            Your AI-powered video analytics tool
+          </p>
+        </div>
+        """
+    )
+    with gr.Group():
+        with gr.Tab("From YouTube"):
+            with gr.Box():
+                with gr.Row().style(equal_height=True):
+                    size = gr.Dropdown(label="Model Size", choices=gio.sizes, value='base')
+                    lang = gr.Dropdown(label="Language (Optional)", choices=gio.langs, value="none")
+                link = gr.Textbox(label="YouTube Link", placeholder="Enter YouTube link...")
+                title = gr.Label(label="Video Title")
+                with gr.Row().style(equal_height=True):
+                    img = gr.Image(label="Thumbnail")
+                    text = gr.Textbox(label="Transcription", placeholder="Transcription Output...", lines=10).style(show_copy_button=True, container=True)
+                with gr.Row().style(equal_height=True):
+                    summary = gr.Textbox(label="Summary", placeholder="Summary Output...", lines=5).style(show_copy_button=True, container=True)
+                    keywords = gr.Textbox(label="Keywords", placeholder="Keywords Output...", lines=5).style(show_copy_button=True, container=True)
+                    label = gr.Label(label="Sentiment Analysis")
+                with gr.Row().style(equal_height=True):
+                    clear = gr.ClearButton([link, title, img, text, summary, keywords, label], scale=1)
+                    btn = gr.Button("Get video insights", variant='primary', scale=1)
+                btn.click(gio, inputs=[link, lang, size], outputs=[text, summary, keywords, label])
+                link.change(gio.populate_metadata, inputs=[link], outputs=[img, title])
+
+        with gr.Tab("From Audio file"):
+            with gr.Box():
+                with gr.Row().style(equal_height=True):
+                    size = gr.Dropdown(label="Model Size", choices=gio.sizes, value='base')
+                    lang = gr.Dropdown(label="Language (Optional)", choices=gio.langs, value="none")
+                audio_file = gr.Audio(type="filepath")
+                with gr.Row().style(equal_height=True):
+                    text = gr.Textbox(label="Transcription", placeholder="Transcription Output...", lines=10).style(show_copy_button=True, container=False)
+                with gr.Row().style(equal_height=True):
+                    summary = gr.Textbox(label="Summary", placeholder="Summary Output", lines=5)
+                    keywords = gr.Textbox(label="Keywords", placeholder="Keywords Output", lines=5)
+                    label = gr.Label(label="Sentiment Analysis")
+                with gr.Row().style(equal_height=True):
+                    clear = gr.ClearButton([text], scale=1)
+                    btn = gr.Button("Get video insights", variant='primary', scale=1)  # Updated button label
+                btn.click(gio.from_audio_input, inputs=[lang, size, audio_file], outputs=[text, summary, keywords, label])
+                
+
+with block:
+    gr.Markdown("About the app:")
+    
+    with gr.Accordion("What is YouTube Insights?", open=False):
+            gr.Markdown("YouTube Insights is a tool developed with academic purposes only, that creates summaries, keywords and sentiments analysis based on YouTube videos or user audio files.")
+    
+    with gr.Accordion("How does it work?", open=False):
+            gr.Markdown("Works by using OpenAI's Whisper, DistilBART for summarization and VoiceLabT5 for Keyword Extraction.")
+
+    gr.HTML("""
+        <div style="text-align: center; max-width: 500px; margin: 0 auto;">
+          <p style="margin-bottom: 10px; font-size: 96%">
+            2023 Master in Big Data & Data Science - Universidad Complutense de Madrid
+          </p>
+        </div>
+        """)
+
+demo.launch()

requirements.txt

@@ -0,0 +1,6 @@
+openai-whisper
+transformers
+torch
+yake
+pytube
+sentencepiece