utils.py

import os
import torch
from skinmodels import IsSkinResnet, IsHealthySkinResnet, SkinDiseaseModelResnet
from config import read_disease_step, read_label_decode, transform_img
import openai
import time
from config_path import SKINMODEL1_PATH, SKINMODEL2_PATH, SKINMODEL3_PATH, SKINDISEASE_STEP_PATH, SKINLABEL_DECODE_PATH

def diagnosis(img):
    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
    
    img = transform_img(img).to(device=device)
    
    # Setup disease step
    disease_to_link, disease_to_step = read_disease_step(SKINDISEASE_STEP_PATH)
    
    # Setup label decode
    label_decode, num_class = read_label_decode(SKINLABEL_DECODE_PATH)
    
    # Setup model1
    assert os.path.isfile(
        SKINMODEL1_PATH), f"Can't find model state_dict path  : {SKINMODEL1_PATH}"
    model1 = IsSkinResnet().to(device)
    model1.load_state_dict(torch.load(
        SKINMODEL1_PATH, map_location=torch.device(device)))
    
    # Setup model2
    model2 = IsHealthySkinResnet().to(device)
    assert os.path.isfile(
        SKINMODEL2_PATH), f"Can't find model state_dict path  : {SKINMODEL2_PATH}"
    model2.load_state_dict(torch.load(
        SKINMODEL2_PATH, map_location=torch.device(device)))

    # Setup model3
    model3 = SkinDiseaseModelResnet(num_class).to(device)
    assert os.path.isfile(
        SKINMODEL3_PATH), f"Can't find model state_dict path  : {SKINMODEL3_PATH}"
    model3.load_state_dict(torch.load(
        SKINMODEL3_PATH, map_location=torch.device(device)))
    
    # Diagnosis
    isskin = model1(img).argmax().item()
    if isskin == 1:
        return 'notskin', 'Kosong', 'Kosong'
    ishealthy = model2(img).argmax().item()
    if ishealthy == 0:
        return 'healthy', 'Kosong', 'Kosong'
    disease = model3(img).argmax().item()

    return label_decode[disease], disease_to_link[label_decode[disease]], disease_to_step[label_decode[disease]]


def getMessage(msg):
    try:
        openRespon = openai.ChatCompletion.create(
        model="gpt-3.5-turbo",
            messages=msg,
            temperature=0.5
            )
        reply = openRespon.choices[0].message.content
        return reply
    except openai.error.RateLimitError:
        return "Dokter saat ini sedang sibuk, tolong tunggu beberapa saat lagi"
    except:
        return "Maaf, sepertinya dokter tidak bisa menjawab saat ini. Silahkan refresh halaman. Jika masalah masih terus berlanjut, silahkan hubungi admin"

last_apikey_check = int(round(time.time()*1000))-35*1000
def cekApikey():
    try:
        global last_apikey_check
        t_stamp = int(round(time.time())*1000)
        if((t_stamp-last_apikey_check) > 30*1000):
            openRespon = openai.ChatCompletion.create(
                model="gpt-3.5-turbo",
                messages=[{"role" : "user", "content": "i"}],
                temperature=0.5
                )
            last_apikey_check = t_stamp
        return True
    except Exception as err:
        print(err)
        return False