File size: 4,503 Bytes
218b27d 5edee8f 004f352 5edee8f 004f352 218b27d 004f352 a65e425 218b27d 4befcc4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 |
# import whisper
import gradio as gr
import datetime
import subprocess
import wave
import contextlib
# import torch
# import pyannote.audio
# from pyannote.audio.pipelines.speaker_verification import PretrainedSpeakerEmbedding
# from pyannote.audio import Audio
# from pyannote.core import Segment
# from sklearn.cluster import AgglomerativeClustering
# import numpy as np
# model = whisper.load_model("large-v2")
# embedding_model = PretrainedSpeakerEmbedding(
# "speechbrain/spkrec-ecapa-voxceleb",
# device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
# )
# def transcribe(audio, num_speakers):
# path, error = convert_to_wav(audio)
# if error is not None:
# return error
# duration = get_duration(path)
# if duration > 4 * 60 * 60:
# return "Audio duration too long"
# result = model.transcribe(path)
# segments = result["segments"]
# num_speakers = min(max(round(num_speakers), 1), len(segments))
# if len(segments) == 1:
# segments[0]['speaker'] = 'SPEAKER 1'
# else:
# embeddings = make_embeddings(path, segments, duration)
# add_speaker_labels(segments, embeddings, num_speakers)
# output = get_output(segments)
# return output
# def convert_to_wav(path):
# if path[-3:] != 'wav':
# new_path = '.'.join(path.split('.')[:-1]) + '.wav'
# try:
# subprocess.call(['ffmpeg', '-i', path, new_path, '-y'])
# except:
# return path, 'Error: Could not convert file to .wav'
# path = new_path
# return path, None
# def get_duration(path):
# with contextlib.closing(wave.open(path,'r')) as f:
# frames = f.getnframes()
# rate = f.getframerate()
# return frames / float(rate)
# def make_embeddings(path, segments, duration):
# embeddings = np.zeros(shape=(len(segments), 192))
# for i, segment in enumerate(segments):
# embeddings[i] = segment_embedding(path, segment, duration)
# return np.nan_to_num(embeddings)
# audio = Audio()
# def segment_embedding(path, segment, duration):
# start = segment["start"]
# # Whisper overshoots the end timestamp in the last segment
# end = min(duration, segment["end"])
# clip = Segment(start, end)
# waveform, sample_rate = audio.crop(path, clip)
# return embedding_model(waveform[None])
# def add_speaker_labels(segments, embeddings, num_speakers):
# clustering = AgglomerativeClustering(num_speakers).fit(embeddings)
# labels = clustering.labels_
# for i in range(len(segments)):
# segments[i]["speaker"] = 'SPEAKER ' + str(labels[i] + 1)
# def time(secs):
# return datetime.timedelta(seconds=round(secs))
# def get_output(segments):
# output = ''
# for (i, segment) in enumerate(segments):
# if i == 0 or segments[i - 1]["speaker"] != segment["speaker"]:
# if i != 0:
# output += '\n\n'
# output += segment["speaker"] + ' ' + str(time(segment["start"])) + '\n\n'
# output += segment["text"][1:] + ' '
# return output
s = ""
def greet1(name):
global s
s = "modified"
return "Hello " + name + "!"
def greet2(name):
return "Hi " + name + "!" + " " + s
def greet3(name):
return "Hola " + name + "!"
with gr.Blocks() as demo:
with gr.Row():
with gr.Column():
audio_file = gr.UploadButton(label="Upload a Audio file (.wav)")
# name = gr.Textbox(label="Name", placeholder="Name") # TODO: remove
number_of_speakers = gr.Number(label="Number of Speakers", value=2)
with gr.Row():
btn_clear = gr.Button(value="Clear")
btn_submit = gr.Button(value="Submit")
with gr.Column():
title = gr.Textbox(label="Title", placeholder="Title for Conversation")
short_summary = gr.Textbox(label="Short Summary", placeholder="Short Summary for Conversation")
sentiment_analysis = gr.Textbox(label="Sentiment Analysis", placeholder="Sentiment Analysis for Conversation")
quality = gr.Textbox(label="Quality of Conversation", placeholder="Quality of Conversation")
detailed_summary = gr.Textbox(label="Detailed Summary", placeholder="Detailed Summary for Conversation")
gr.Markdown("## Examples")
gr.Examples(
examples=[
[
"Harsh",
2,
],
[
"Rahul",
2,
],
],
inputs=[title],
outputs=[short_summary],
fn=greet1,
cache_examples=True,
)
gr.Markdown(
"""
See [github.com/facebookresearch/audiocraft](https://github.com/facebookresearch/audiocraft)
for more details.
"""
)
demo.launch()
|