util.py

from flask import Flask
import ollama


def ps():
    from ollama import ProcessResponse, chat, ps, pull
    ansver = []

    response: ProcessResponse = ps()
    for model in response.models:
          ansver.append(f'<h3>Mode<h3>:  {model.model}')
          ansver.append(f'\t Digest:  {model.digest}')
          ansver.append(f'\t  Expires at: {model.expires_at}')
          ansver.append(f'\t  Size:  {model.size}')
          ansver.append(f'\t  Size vram:  {model.size_vram}')
          ansver.append(f'\t  Details: {model.details}')
          ansver.append(f'\t  Context length:  {model.context_length}')
          ansver.append(f'\n')
    result = [f'<p>{answ}</p>' for answ in ansver ]
    return '\n'.join(result)

def time_model(model_name='qwen3:0.6b'):
  from  datetime import datetime
  from ollama import Client
  from ollama import chat

  t_start=datetime.now()
  ansver = []
  messages = [
  {
    'role': 'user',
    'content': 'Расскажи о себе подробно',
  },
  ]
  response = chat(model_name, messages=messages)
  ansver.append(f' <h3> {model_name}</h3>')
  ansver.append(f' start {t_start} ')
  response_time=datetime.now()-t_start
  ansver.append(f' duration  {response_time} ')
  #
  response_len = len(response['message']['content'])
  ansver.append(f' lehgth {response_len}')
  response_speed=response_len/int(response_time.seconds)
  ansver.append(f' token/sek {response_speed}')
  ansver.append(response['message']['content'])
  result = [f'<p>{answ}</p>' for answ in ansver ]
  return '\n'.join(result)