File size: 4,203 Bytes
a25aeae |
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 |
import os
import requests
import json
from concurrent.futures import ThreadPoolExecutor, as_completed
import argparse
def synthesize_and_save(batch, voicevox_engine_url):
for task in batch:
file_name, text, char_name, style_name, style_id = task
try:
# フォルダを作成する
output_folder = f"{char_name}_{style_name}"
os.makedirs(output_folder, exist_ok=True)
# 出力ファイル名を作成する
output_file_name = os.path.join(output_folder, f"{file_name}.wav")
# 既にファイルが存在する場合はスキップ
if os.path.exists(output_file_name):
print(f"File {output_file_name} already exists, skipping.")
continue
# 音声合成クエリを作成する
query_response = requests.post(
f"{voicevox_engine_url}/audio_query",
params={"text": text, "speaker": style_id},
headers={"Content-Type": "application/json"}
)
if query_response.status_code != 200:
print(f"Failed to create audio query for {file_name} with character {char_name} and style {style_name}")
continue
query = query_response.json()
# 音声合成を実行する
synthesis_response = requests.post(
f"{voicevox_engine_url}/synthesis",
params={"speaker": style_id},
data=json.dumps(query),
headers={"Content-Type": "application/json"}
)
if synthesis_response.status_code != 200:
print(f"Failed to synthesize audio for {file_name} with character {char_name} and style {style_name}")
continue
# 音声ファイルを保存する
with open(output_file_name, "wb") as out_f:
out_f.write(synthesis_response.content)
print(f"Saved {output_file_name}")
except Exception as e:
print(f"Exception occurred for {file_name} with character {char_name} and style {style_name}: {e}")
def main(max_workers, batch_size):
# VOICEVOXエンジンのURL
VOICEVOX_ENGINE_URL = "http://localhost:50021"
# キャラクター情報をJSONファイルからロードする
with open("character_config.json", "r", encoding="utf-8") as f:
CHARACTERS = json.load(f)
# テキストファイルを読み込む
# with open("emotion_transcript_utf8.txt", "r", encoding="utf-8") as f:
with open("recitation_transcript_utf8.txt", "r", encoding="utf-8") as f:
lines = f.readlines()
tasks = []
for line in lines:
parts = line.strip().split(":")
file_name = parts[0].strip()
text_parts = parts[1].split(",")
text = text_parts[0].strip() # 2列目（使用するテキスト）
for character in CHARACTERS:
char_name = character["name"]
for style in character["styles"]:
style_name = style["name"]
style_id = style["id"]
tasks.append((file_name, text, char_name, style_name, style_id))
# タスクをバッチに分ける
batches = [tasks[i:i + batch_size] for i in range(0, len(tasks), batch_size)]
# 並列処理の実行
with ThreadPoolExecutor(max_workers=max_workers) as executor:
futures = [executor.submit(synthesize_and_save, batch, VOICEVOX_ENGINE_URL) for batch in batches]
for future in as_completed(futures):
future.result() # 各タスクの結果を取得する（例外が発生した場合に備えて）
print("All tasks completed.")
if __name__ == "__main__":
parser = argparse.ArgumentParser(description="Parallel voice synthesis script")
parser.add_argument("--max_workers", type=int, default=256, help="Number of parallel workers")
parser.add_argument("--batch_size", type=int, default=30, help="Number of tasks per batch")
args = parser.parse_args()
main(args.max_workers, args.batch_size)
|