Spaces:
Runtime error
Runtime error
File size: 6,307 Bytes
925d97e |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 |
import zipfile
import hashlib
from utils.model import model_downloader, get_model
import requests
import json
import torch
import os
from inference import Inference
import gradio as gr
from constants import VOICE_METHODS, BARK_VOICES, EDGE_VOICES, zips_folder, unzips_folder
from tts.conversion import tts_infer, ELEVENLABS_VOICES_RAW, ELEVENLABS_VOICES_NAMES
api_url = "https://rvc-models-api.onrender.com/uploadfile/"
if not os.path.exists(zips_folder):
os.mkdir(zips_folder)
if not os.path.exists(unzips_folder):
os.mkdir(unzips_folder)
def get_info(path):
path = os.path.join(unzips_folder, path)
try:
a = torch.load(path, map_location="cpu")
return a
except Exception as e:
print("*****************eeeeeeeeeeeeeeeeeeeerrrrrrrrrrrrrrrrrr*****")
print(e)
return {
}
def calculate_md5(file_path):
hash_md5 = hashlib.md5()
with open(file_path, "rb") as f:
for chunk in iter(lambda: f.read(4096), b""):
hash_md5.update(chunk)
return hash_md5.hexdigest()
def compress(modelname, files):
file_path = os.path.join(zips_folder, f"{modelname}.zip")
# Select the compression mode ZIP_DEFLATED for compression
# or zipfile.ZIP_STORED to just store the file
compression = zipfile.ZIP_DEFLATED
# Comprueba si el archivo ZIP ya existe
if not os.path.exists(file_path):
# Si no existe, crea el archivo ZIP
with zipfile.ZipFile(file_path, mode="w") as zf:
try:
for file in files:
if file:
# Agrega el archivo al archivo ZIP
zf.write(unzips_folder if ".index" in file else os.path.join(unzips_folder, file), compress_type=compression)
except FileNotFoundError as fnf:
print("An error occurred", fnf)
else:
# Si el archivo ZIP ya existe, agrega los archivos a un archivo ZIP existente
with zipfile.ZipFile(file_path, mode="a") as zf:
try:
for file in files:
if file:
# Agrega el archivo al archivo ZIP
zf.write(unzips_folder if ".index" in file else os.path.join(unzips_folder, file), compress_type=compression)
except FileNotFoundError as fnf:
print("An error occurred", fnf)
return file_path
def infer(model, f0_method, audio_file):
print("****", audio_file)
inference = Inference(
model_name=model,
f0_method=f0_method,
source_audio_path=audio_file,
output_file_name=os.path.join("./audio-outputs", os.path.basename(audio_file))
)
output = inference.run()
if 'success' in output and output['success']:
return output, output['file']
else:
return
def post_model(name, model_url, version, creator):
modelname = model_downloader(model_url, zips_folder, unzips_folder)
model_files = get_model(unzips_folder, modelname)
if not model_files:
return "No se encontrado un modelo valido, verifica el contenido del enlace e intentalo más tarde."
if not model_files.get('pth'):
return "No se encontrado un modelo valido, verifica el contenido del enlace e intentalo más tarde."
md5_hash = calculate_md5(os.path.join(unzips_folder,model_files['pth']))
zipfile = compress(modelname, list(model_files.values()))
a = get_info(model_files.get('pth'))
file_to_upload = open(zipfile, "rb")
info = a.get("info", "None"),
sr = a.get("sr", "None"),
f0 = a.get("f0", "None"),
data = {
"name": name,
"version": version,
"creator": creator,
"hash": md5_hash,
"info": info,
"sr": sr,
"f0": f0
}
print("Subiendo archivo...")
# Realizar la solicitud POST
response = requests.post(api_url, files={"file": file_to_upload}, data=data)
result = response.json()
# Comprobar la respuesta
if response.status_code == 200:
result = response.json()
return json.dumps(result, indent=4)
else:
print("Error al cargar el archivo:", response.status_code)
return result
def search_model(name):
web_service_url = "https://script.google.com/macros/s/AKfycbyRaNxtcuN8CxUrcA_nHW6Sq9G2QJor8Z2-BJUGnQ2F_CB8klF4kQL--U2r2MhLFZ5J/exec"
response = requests.post(web_service_url, json={
'type': 'search_by_filename',
'name': name
})
result = []
response.raise_for_status() # Lanza una excepción en caso de error
json_response = response.json()
cont = 0
result.append("""| Nombre del modelo | Url | Epoch | Sample Rate |
| ---------------- | -------------- |:------:|:-----------:|
""")
yield "<br />".join(result)
if json_response.get('ok', None):
for model in json_response['ocurrences']:
if cont < 20:
model_name = str(model.get('name', 'N/A')).strip()
model_url = model.get('url', 'N/A')
epoch = model.get('epoch', 'N/A')
sr = model.get('sr', 'N/A')
line = f"""|{model_name}|<a>{model_url}</a>|{epoch}|{sr}|
"""
result.append(line)
yield "".join(result)
cont += 1
def update_tts_methods_voice(select_value):
if select_value == "Edge-tts":
return gr.Dropdown.update(choices=EDGE_VOICES, visible=True, value="es-CO-GonzaloNeural-Male"), gr.Markdown.update(visible=False), gr.Textbox.update(visible=False),gr.Radio.update(visible=False)
elif select_value == "Bark-tts":
return gr.Dropdown.update(choices=BARK_VOICES, visible=True), gr.Markdown.update(visible=False), gr.Textbox.update(visible=False),gr.Radio.update(visible=False)
elif select_value == 'ElevenLabs':
return gr.Dropdown.update(choices=ELEVENLABS_VOICES_NAMES, visible=True, value="Bella"), gr.Markdown.update(visible=True), gr.Textbox.update(visible=True), gr.Radio.update(visible=False)
elif select_value == 'CoquiTTS':
return gr.Dropdown.update(visible=False), gr.Markdown.update(visible=False), gr.Textbox.update(visible=False), gr.Radio.update(visible=True)
|