app.py

# app.py
# Version: 1.07 (08.24.24), ALPHA
#---------------------------------------------------------------------------------------------------------------------------------------------
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#---------------------------------------------------------------------------------------------------------------------------------------------
import spaces
import gradio as gr
from PIL import Image
from pydub import AudioSegment
import os
import re
import time
import warnings
#import datetime
import subprocess
from pathlib import Path
from fpdf import FPDF

import psutil
from gpuinfo import GPUInfo
#import pandas as pd
#import csv
import numpy as np
import torch
import torchaudio
import torchaudio.transforms as transforms

from transformers import pipeline, AutoModel

import spacy
import networkx as nx
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
warnings.filterwarnings("ignore")

#              ------------header section------------
HEADER_INFO = """
    # WEB APP ✨| Norwegian WHISPER Model
Switch Work [Transkribering av lydfiler til norsk skrift]
""".strip()
LOGO = "https://cdn-lfs-us-1.huggingface.co/repos/fe/3b/fe3bd7c8beece8b087fddcc2278295e7f56c794c8dcf728189f4af8bddc585e1/5112f67899d65e9797a7a60d05f983cf2ceefbe2f7cba74eeca93a4e7061becc?response-content-disposition=inline%3B+filename*%3DUTF-8%27%27logo.png%3B+filename%3D%22logo.png%22%3B&response-content-type=image%2Fpng&Expires=1724881270&Policy=eyJTdGF0ZW1lbnQiOlt7IkNvbmRpdGlvbiI6eyJEYXRlTGVzc1RoYW4iOnsiQVdTOkVwb2NoVGltZSI6MTcyNDg4MTI3MH19LCJSZXNvdXJjZSI6Imh0dHBzOi8vY2RuLWxmcy11cy0xLmh1Z2dpbmdmYWNlLmNvL3JlcG9zL2ZlLzNiL2ZlM2JkN2M4YmVlY2U4YjA4N2ZkZGNjMjI3ODI5NWU3ZjU2Yzc5NGM4ZGNmNzI4MTg5ZjRhZjhiZGRjNTg1ZTEvNTExMmY2Nzg5OWQ2NWU5Nzk3YTdhNjBkMDVmOTgzY2YyY2VlZmJlMmY3Y2JhNzRlZWNhOTNhNGU3MDYxYmVjYz9yZXNwb25zZS1jb250ZW50LWRpc3Bvc2l0aW9uPSomcmVzcG9uc2UtY29udGVudC10eXBlPSoifV19&Signature=ipo8wTjtC7R0QHbo%7Et9Q5CTaI3cZKxM0beajqlApfm5fh7%7EW-FULu1-ISL5bkowBSw9m5RdGoyOqj336OSS5fPD%7EnzYNmAMd3T5bx2-KfCDh6jz0HVECt8S7HeIu%7El2TetxrzL2tdHw4Np4Zpa8JKOnNnje24fF0Nr-xUS2dvPJf54rIL70-iWVXXhw8owxt0%7E1CJsUHC9oibp9B4mZcyWvvRldhDopiQBYELusZdTW3qvtTBK083WP3gHQxadQp8UDVTPZ0g3i112G2NfFJB%7Epa70XeN8m3E6ORx6pVH%7EW6IzjvmapWSF-tmXH-26wYG8aof%7E1U7enbR1w2QBTS-g__&Key-Pair-Id=K24J24Z295AEI9"
SIDEBAR_INFO = f"""
<div align="center">
    <img src="{LOGO}" style="width: 100%; height: auto;"/>
</div>
"""

# ------------transcribe section------------

@spaces.GPU()
def convert_to_wav(filepath):
    _, file_ending = os.path.splitext(f'{filepath}')
    audio_file = filepath.replace(file_ending, ".wav")
    os.system(f'ffmpeg -i "{filepath}" -ar 16000 -ac 1 -c:a pcm_s16le "{audio_file}"')
    return audio_file

pipe = pipeline("automatic-speech-recognition", model="NbAiLab/nb-whisper-large", chunk_length_s=30, generate_kwargs={'task': 'transcribe', 'language': 'no'})

@spaces.GPU()
def transcribe_audio(audio_file, batch_size=16, sample_rate =16000):
    audio_file = audio_tuple[0] # assumes first element of the tuple contains the file path;
    waveform, sample_rate = torchaudio.load(audio_file) # to avoid TypeError here

    if waveform.ndim > 1:
        waveform = waveform[0, :]

    waveform = waveform.numpy()

    start_time = time.time()

    # --pipe it
    with torch.no_grad():
        outputs = pipe(waveform, sampling_rate=sample_rate, batch_size=batch_size, return_timestamps=False)

    end_time = time.time()
    
    output_time = end_time - start_time
    word_count = len(text.split())

    # --GPU metrics
    memory = psutil.virtual_memory()
    gpu_utilization, gpu_memory = GPUInfo.gpu_usage()
    gpu_utilization = gpu_utilization[0] if len(gpu_utilization) > 0 else 0
    gpu_memory = gpu_memory[0] if len(gpu_memory) > 0 else 0

    # --CPU metric
    cpu_usage = psutil.cpu_percent(interval=1)

    # --system info string
    system_info = f"""
    *Memory: {memory.total / (1024 * 1024 * 1024):.2f}GB, used: {memory.percent}%, available: {memory.available / (1024 * 1024 * 1024):.2f}GB.*
    *Processing time: {output_time:.2f} seconds.*
    *Number of words: {word_count}*
    *GPU Utilization: {gpu_utilization}%, GPU Memory: {gpu_memory}*
    *CPU Usage: {cpu_usage}%*
    """

    return text.strip(), system_info



#              ------------summary section------------


#          ------------for app integration later------------

@spaces.GPU()
def clean_text(text):
    text = re.sub(r'https?:\/\/.*[\r\n]*', '', text)
    text = re.sub(r'[^\w\s]', '', text)
    text = re.sub(r'\s+', ' ', text).strip()
    return text

nlp = spacy.blank("nb")  # 'nb' ==> codename = Norwegian Bokmål
nlp.add_pipe('sentencizer')
spacy_stop_words = spacy.lang.nb.stop_words.STOP_WORDS

summarization_model = AutoModel.from_pretrained("NbAiLab/nb-bert-large")
# pipe = pipeline("fill-mask", model="NbAiLab/nb-bert-large")

@spaces.GPU()
def preprocess_text(text):
    # Process the text with SpaCy
    doc = nlp(text)
    # SpaCy's stop top wrds direct
    stop_words = spacy_stop_words
    # Filter out stop words
    words = [token.text for token in doc if token.text.lower() not in stop_words]
    return ' '.join(words)

@spaces.GPU()
def summarize_text(text):
    preprocessed_text = preprocess_text(text)
    inputs = summarization_tokenizer(preprocessed_text, max_length=1024, return_tensors="pt", truncation=True)
    inputs = inputs.to(device)
    summary_ids = summarization_model.generate(inputs.input_ids, num_beams=5, max_length=150, early_stopping=True)
    return summarization_tokenizer.decode(summary_ids[0], skip_special_tokens=True)
    
@spaces.GPU()
def build_similarity_matrix(sentences, stop_words):
    similarity_matrix = nx.Graph()
    for i, tokens_a in enumerate(sentences):
        for j, tokens_b in enumerate(sentences):
            if i != j:
                common_words = set(tokens_a) & set(tokens_b)
                similarity_matrix.add_edge(i, j, weight=len(common_words))
    return similarity_matrix

# PageRank
@spaces.GPU()
def graph_based_summary(text, num_paragraphs=3):
    doc = nlp(text)
    sentences = [sent.text for sent in doc.sents]
    if len(sentences) < num_paragraphs:
        return ' '.join(sentences) 

    sentence_tokens = [nlp(sent) for sent in sentences]
    stop_words = spacy_stop_words
    filtered_tokens = [[token.text for token in tokens if token.text.lower() not in stop_words] for tokens in sentence_tokens]
    similarity_matrix = build_similarity_matrix(filtered_tokens, stop_words)

    scores = nx.pagerank(similarity_matrix)
    ranked_sentences = sorted(((scores[i], sent) for i, sent in enumerate(sentences)), reverse=True)
    return ' '.join([sent for _, sent in ranked_sentences[:num_paragraphs]])

# LexRank
@spaces.GPU()
def lex_rank_summary(text, num_paragraphs=3, threshold=0.1):
    doc = nlp(text)
    sentences = [sent.text for sent in doc.sents]
    if len(sentences) < num_paragraphs:
        return ' '.join(sentences)  # Adjusted to return a single string

    stop_words = spacy_stop_words
    vectorizer = TfidfVectorizer(stop_words=list(stop_words))
    X = vectorizer.fit_transform(sentences)
    similarity_matrix = cosine_similarity(X, X)

    # Apply threshold@similarity matrix
    similarity_matrix[similarity_matrix < threshold] = 0
    nx_graph = nx.from_numpy_array(similarity_matrix)
    scores = nx.pagerank(nx_graph)
    ranked_sentences = sorted(((scores[i], s) for i, s in enumerate(sentences)), reverse=True)
    return ' '.join([ranked_sentences[i][1] for i in range(num_paragraphs)])

# TextRank
@spaces.GPU()
def text_rank_summary(text, num_paragraphs=3):
    doc = nlp(text)
    sentences = [sent.text for sent in doc.sents]
    if len(sentences) < num_paragraphs:
        return ' '.join(sentences) 

    stop_words = spacy_stop_words
    vectorizer = TfidfVectorizer(stop_words=list(stop_words))
    X = vectorizer.fit_transform(sentences)
    similarity_matrix = cosine_similarity(X, X)

    nx_graph = nx.from_numpy_array(similarity_matrix)
    scores = nx.pagerank(nx_graph)
    ranked_sentences = sorted(((scores[i], s) for i, s in enumerate(sentences)), reverse=True)
    return ' '.join([ranked_sentences[i][1] for i in range(num_paragraphs)])

iface = gr.Blocks()

with iface:
    gr.HTML(SIDEBAR_INFO)
    gr.Markdown(HEADER_INFO)
    
    audio_input = gr.Audio(label="Upload Audio File")
    transcribed_text = gr.Textbox(label="Transcribed Text")
    system_info = gr.Textbox(label="System Info")
    
    transcribe_button = gr.Button("Transcribe")
    transcribe_button.click(fn=transcribe_audio, inputs=audio_input, outputs=[transcribed_text, system_info])

iface.launch(share=True, debug=True)