app.py

### -----------------------------------------------------------------------
### (FULL, Revised) version_1.07ALPHA_app.py
### -----------------------------------------------------------------------

# -------------------------------------------------------------------------
# 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
#from scipy.io import wavfile

import os
import re
import time
import warnings
#import datetime
#import pandas as pd
#import csv
import subprocess
from pathlib import Path
import tempfile
from fpdf import FPDF

import psutil
from gpuinfo import GPUInfo

#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------------

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

@spaces.GPU()
def transcribe(microphone, file_upload, batch_size=15):
    warn_output = ""
    if (microphone is not None) and (file_upload is not None):
        warn_output = (
            "WARNING: You've uploaded an audio file and used the microphone. "
            "The recorded file from the microphone will be used and the uploaded audio will be discarded.\n"
        )

    elif (microphone is None) and (file_upload is None):
        return "ERROR: You have to either use the microphone or upload an audio file"

    file = microphone if microphone is not None else file_upload

    start_time = time.time()
    text = pipe(file, batch_size=batch_size, return_timestamps=False)["text"]

    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 warn_output + text.strip(), system_info


def save_to_pdf(text, summary):
    pdf = FPDF()
    pdf.add_page()
    pdf.set_font("Arial", size=12)
    #
    # ----add same if/elif logic as above here----
    #
    if text:
        pdf.multi_cell(0, 10, "Text:\n" + text)

    pdf.ln(10)  # Paragraph metric

    if summary:
        pdf.multi_cell(0, 10, "Summary:\n" + summary)

    pdf_output_path = "transcription_.pdf"
    pdf.output(pdf_output_path)
    return pdf_output_path


iface = gr.Blocks()

with iface:

    gr.HTML(SIDEBAR_INFO)
    gr.Markdown(HEADER_INFO)

    with gr.Row():
        gr.Markdown('''
        ##### Here you will get transcription output
        ##### ''')
        microphone = gr.Audio(sources="microphone", type="filepath")
        upload = gr.Audio(sources="upload", type="filepath")

        transcribe_btn = gr.Button("Transcribe Interview")
        text_output = gr.Textbox()
        system_info = gr.Textbox(label="System Info")
        
        transcribe_btn.click(fn=transcribe, inputs=[microphone, upload], outputs=[text_output, system_info])

    with gr.Tabs():
        with gr.TabItem("Download PDF"):
            pdf_text_only = gr.Button("Download PDF with Transcribed Text")
            pdf_output = gr.File(label="Download PDF")

            pdf_text_only.click(fn=lambda text: save_to_pdf(text, ""), inputs=[text_output], outputs=[pdf_output])