app.py

#Flask-APScheduler
from flask_basicauth import BasicAuth
from datetime import datetime
from flask_apscheduler import APScheduler

from flask import Flask  
from flsak import request 

model_name = "inclusionAI/Ring-mini-2.0"
print(f"load model {model_name}")
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto",
    trust_remote_code=True
)

print(f"load tokenizer {model_name}")
tokenizer = AutoTokenizer.from_pretrained(model_name)

app = Flask(__name__)  

# prompt = "Напиши стих про шинное обучение"
# print(response)
import torch
from qwen_asr import Qwen3ASRModel

model = Qwen3ASRModel.from_pretrained(
    "Qwen/Qwen3-ASR-1.7B",
    dtype=torch.bfloat16,
    device_map="cuda:0",
    # device_map="cpu",
    # attn_implementation="flash_attention_2",
    max_inference_batch_size=32, # Batch size limit for inference. -1 means unlimited. Smaller values can help avoid OOM.
    max_new_tokens=256, # Maximum number of tokens to generate. Set a larger value for long audio input.
)
import datetime
start=datetime.datetime.now()
results = model.transcribe(
    audio="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen3-ASR-Repo/asr_en.wav",
    language=None, # set "English" to force the language
)

# Чтобы защитить паролем конечную точку задания:
# @basic_auth.required

basic_auth = BasicAuth(app)
app.config['BASIC_AUTH_REALM'] = 'realm'
app.config['BASIC_AUTH_USERNAME'] = 'vova'
app.config['BASIC_AUTH_PASSWORD'] = 'vova'

print(f"Flask app")
# Главная страница
@app.route("/", methods=['POST', 'GET']) #
@basic_auth.required
def root():
    if request.args.get('logs'): #logs=container  ?logs=container&__theme=system
        return f"<h1>Главная logs</h1>"
    else:
        return f"<h1>Главная страница</h1><p>Добро пожаловать!</p><p>current model {model_name}</p>"

# Страница "О нас"
@app.route("/about")
def about():
    return "<h1>О нас</h1><p>Мы изучаем Flask!</p>"


# Страница "Контакты"
@app.route("/contact")
@basic_auth.required
def contact():
    import datetime
    start=datetime.datetime.now()
    results = model.transcribe(
        audio="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen3-ASR-Repo/asr_en.wav",
    language=None, # set "English" to force the language)
    print(f'duration {datetime.datetime.now()-start}')
    print(results[0].language)
    print(results[0].text)
    return f"<h1>Контакты</h1><p>duration {datetime.datetime.now()-start}</p><p>results[0].text</p>"

if __name__ == "__main__":
    app.run(debug=False, host='0.0.0.0', port=7860) root@telebot1:/opt/HF/TruVlad/flaskapi#