app.py

import tempfile
import gradio as gr
import os
import tensorflow_hub as hub
import tensorflow as tf
import tensorflow_text as text
import sys
import numpy as np
import faiss
import csv
import datetime


# NO GPU
os.environ['CUDA_VISIBLE_DEVICES'] = '-1'
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'



def make_bert_preprocess_model(sentence_features, tfhub_handle_preprocess, seq_length=128):
  """Returns Model mapping string features to BERT inputs.
  """

  input_segments = [
      tf.keras.layers.Input(shape=(), dtype=tf.string, name=ft)
      for ft in sentence_features]

  bert_preprocess = hub.load(tfhub_handle_preprocess)
  tokenizer = hub.KerasLayer(bert_preprocess.tokenize, name='tokenizer')
  segments = [tokenizer(s) for s in input_segments]

  truncated_segments = segments
  packer = hub.KerasLayer(bert_preprocess.bert_pack_inputs,
                          arguments=dict(seq_length=seq_length),
                          name='packer')
  model_inputs = packer(truncated_segments)
  return tf.keras.Model(input_segments, model_inputs)

from models import *

def process(prompt, lang):

    # Getting prompt user
    #prompt = input("Audio Search - enter text : ")
    #print(prompt)

    # prompt embedding
    bert_model_name = 'small_bert/bert_en_uncased_L-4_H-512_A-8' 
    tfhub_handle_encoder = 'https://tfhub.dev/tensorflow/small_bert/bert_en_uncased_L-4_H-512_A-8/1'
    tfhub_handle_preprocess = 'https://tfhub.dev/tensorflow/bert_en_uncased_preprocess/3'

    MAX_LENGTH = 130 # MAX de 512 !!! TENSORFLOW !!!
    TOP = 10


    bert_preprocess_model = make_bert_preprocess_model(['my_input'], tfhub_handle_preprocess, seq_length = MAX_LENGTH)
    bert_model = hub.KerasLayer(tfhub_handle_encoder)

    now = datetime.datetime.now()
    print()
    print('*************')
    print("Current Time: ", str(now))
    print("Text input : ", prompt)
    print('*************')
    print()
    prompt=[prompt]
    text_preprocessed = bert_preprocess_model([np.array(prompt)])
    embed_prompt = bert_model(text_preprocessed)
    print("   text representation computed.")

    # Embed text
    #from models import *
    encoder_text = tf.keras.models.load_model('encoder_text_retrievaltext_bmg_221022_54.h5')
    embed_query  = encoder_text.predict(embed_prompt["pooled_output"])
    faiss.normalize_L2(embed_query)
    print("   text embed computed.")

    # load embed audio catalog
    index = faiss.read_index("BMG_221022.index")

    # distance computing
    D, I = index.search(embed_query, TOP)

    # names index
    import joblib
    audio_names = joblib.load(open('BMG_221022_names.index', 'rb'))

    #url
    url_dict={}
    with open("bmg_clean.csv") as csv_file:
        csv_reader = csv.reader(csv_file, delimiter=';')
        for row in csv_reader:
          f = row[2].split('/')[-1]
          url_dict[f.split('/')[-1][:-4]] = row[2]

    # output : top N audio file names
    print(I)
    print(D)
    print("----")
    for i in range(len(I[0])):
        print(audio_names[I[0][i]], " with distance ", D[0][i])
        print("    url : ", url_dict[audio_names[I[0][i]]])


    return [url_dict[audio_names[I[0][0]]], url_dict[audio_names[I[0][1]]], url_dict[audio_names[I[0][2]]], url_dict[audio_names[I[0][3]]], url_dict[audio_names[I[0][4]]]]

inputs = [gr.Textbox(label="Input", value="type your description", max_lines=2), 
            gr.Radio(label="Language", choices=["en"], value="en")]

poc_examples = [#[["I love learning machine learning"],["autre"]]
          ["Mysterious filmscore with Arabic influenced instruments","en"],
          ["Let's go on a magical adventure with wizzards, dragons and castles","en"],
          ["Creepy piano opening evolves and speeds up into a cinematic orchestral piece","en"],
          ["Halloween rock with creepy organ","en"],
          ["Rhythmic electro dance track for sport, motivation and sweating","en"],
          ["soundtrack for an action movie from the eighties in a retro synth wave style","en"],
          ["Choral female singing is rhythmically accompanied in a church with medieval instruments","en"],
          ["Christmas","en"],
          ["love romantic with piano, strings and vocals","en"],
          ["Electronic soundscapes for chilling and relaxing","en"],
          ["Minimal, emotional, melancholic piano","en"],
#          ["Minimal, happy, joyful piano","en"],
          ["A calm and romantic acoustic guitar melody","en"],
          ["horror suspense piano","en"],
          ["Big Band","en"],
          ["90 eurodance beat","en"],
]


outputs = [gr.Audio(label="Track 1"), gr.Audio(label="Track 2"), gr.Audio(label="Track 3"), gr.Audio(label="Track 4"), gr.Audio(label="Track 5")]
demo1 = gr.Interface(fn=process, inputs=inputs, outputs=outputs, examples=poc_examples, cache_examples=False)

demo1.launch(debug=True)
#demo1.launch(debug=True, enable_queue = False, auth=(os.environ['DEMO_LOGIN'], os.environ['DEMO_PWD']),auth_message = "Contact Simbals to get login/pwd")#, share=True)