[FIXED] No such a file or directory in Azure Function App

Issue

I have created azure function. In that function I have different folders in which some files exists. After deployment to azure portal, I cannot see any folder/directory in that function. These files conatins supporting .py scripts which is used in inference.py as imports. I have deployed this function using docker image.

As you can see above the folders data, model and utils are there in repo but not reflecting or accessed in azure function app

This is the code snippet I am trying to run

"""
Run a summarization model interactively.
"""

import argparse
import numpy as np
import nltk
# from colors import yellow, red, blue
import torch
from torch.autograd import Variable
import json,re
from data import loader
from data.loader import DataLoader
from utils import helper, constant, torch_utils, text_utils, bleu, rouge
from utils.torch_utils import set_cuda
from utils.vocab import Vocab
from model.trainer import Trainer
import joblib


def load_model():
    model_file = "model.pt"
    return model_file

def get_input(indication, exam, npi, cpt, modality, pos, findings):
    background = 'INDICATION : ' +str(indication)+ '. '+'EXAM : ' +str(exam)+ '. '+'NPI : ' +str(npi)+ '. '+'MODALITY : '+str(modality)+'. '+'CPT : '+str(cpt)+'. '+'POS_CR : '+str(pos)+'.'
    background = nltk.word_tokenize(background)
    findings = nltk.word_tokenize(findings)
    return background, findings

def run(background, findings, trainer, vocab, opt):
    bg_tokens, src_tokens = background, findings
    if opt['lower']:
        bg_tokens = [t.lower() for t in bg_tokens]
        src_tokens = [t.lower() for t in src_tokens]
    if len(bg_tokens) == 0:
        bg_tokens = [constant.UNK_TOKEN]
    src_tokens = [constant.SOS_TOKEN] + src_tokens + [constant.EOS_TOKEN]
    src = loader.map_to_ids(src_tokens, vocab.word2id)
    bg = loader.map_to_ids(bg_tokens, vocab.word2id)
    src = loader.get_long_tensor([src], 1)
    bg = loader.get_long_tensor([bg], 1)
    if opt['cuda']:
        src = src.cuda()
        bg = bg.cuda()
    preds,confidence_score = trainer.model.predict(src, bg, opt['beam_size'])
    pred_tokens = text_utils.unmap_with_copy(preds, [src_tokens], vocab)
    pred_tokens = text_utils.prune_decoded_seqs(pred_tokens)[0]
    return pred_tokens,confidence_score

def custom_capitalize(text):
    
    # Use a regular expression to find patterns like "1., 2., 3., 4., 5."
    pattern = re.compile(r'^([1-5])\.\s')
    
    # Iterate through matches and capitalize the sentences
    if re.match(pattern, text):
        matches = pattern.finditer(text)
        for match in matches:
            start = match.start()
            end = match.end()

        # Capitalize the sentence after the pattern
            text = text[:end] + text[end:].capitalize()

    # Capitalize the first word in case the sentence starts with a quotation mark
    else:
        text = text.capitalize()
    
    return text


def main():
    # try:
    model_file = load_model()
    print("Loading model from {}...".format(model_file))
    trainer = Trainer(model_file=model_file)
    opt, vocab = trainer.opt, trainer.vocab
    print(vocab)
    trainer.model.eval()
    print("Loaded.\n")
    background, findings = get_input(indication, exam, npi, cpt, modality, pos, report_findings)
    if len(background) < 100 and 500 > len(findings) > 2 and modality.lower() == "x-ray":
        sum_words,confidence_score = run(background, findings, trainer, vocab, opt)
        confidence_score = round (confidence_score,2)
        p_impression = " ".join(sum_words)
        p_impression = nltk.sent_tokenize(p_impression)
        p_impression = [custom_capitalize(sent) for sent in p_impression]
        p_impression = " ".join(p_impression)
        predicted_impression ={'Predicted_Impression': p_impression,'Confidence score':confidence_score}
    else:
        p_impression = " "
        predicted_impression ={'Predicted_Impression': p_impression,'Confidence score':0.0}
    return json.dumps(predicted_impression)
    # except Exception as e:
    #     print(f"An error occurred: {e}")

    


if __name__ == '__main__':
   main()

Giving error while code+test

You can refer to the repo I cloned from github, finetuned my model and then trying to deploy it on azure function app. Note: My finetuned model is x_ray.pt but you can use model.pt in repo.

https://github.com/yuhaozhang/summarize-radiology-findings

My docker image as below

FROM mcr.microsoft.com/azure-functions/python:4.0

LABEL NAME=myfun2 Version=0.0.1
EXPOSE 80

ENV AzureWebJobsScriptRoot=/home/site/wwwroot \
    AzureFunctionsJobHost__Logging__Console__IsEnabled=true

COPY requirements.txt /
RUN cp /etc/apt/sources.list /etc/apt/sources.list~
RUN sed -i '$a deb-src http://deb.debian.org/debian buster main' /etc/apt/sources.list
RUN sed -i '$a deb-src http://security.debian.org/debian-security buster/updates main' /etc/apt/sources.list
RUN sed -i '$a deb-src http://deb.debian.org/debian buster-updates main' /etc/apt/sources.list
# RUN apt-get -y update
RUN apt-get -y upgrade
RUN apt-get install -y git
# RUN apt-get -y install software-properties-common
RUN apt-get -y install g++
RUN apt-get -y install unixodbc-dev 
RUN pip install --upgrade pip
RUN pip install git+https://github.com/tagucci/pythonrouge.git
RUN pip install -r /requirements.txt
#RUN pip env --version
RUN python -m nltk.downloader popular
COPY . /home/site/wwwroot

Requirements.txt is as below:

azure-functions
pymsteams
configparser
requests
asyncio
pandas
joblib
numpy
nltk
torch==1.10.1+cu111 
torchvision==0.11.2+cu111
torchaudio==0.10.1 
-f https://download.pytorch.org/whl/cu111/torch_stable.html

How can I deploy so that I can access the folders.

Solution

How can I deploy so that I can access the folders.

You are getting error because the import syntax is incorrect. when files are inside same folder as main file, which is importing the files, need to use relative reference. check this Microsoft document.

#My directory:

Docker/
 | - .venv/
 | - .vscode/
 | - HttpTrigger1/
 | | - __init__.py
 | | - function.json
 | | - inference.py
 | | - data/
 | | | - __init__.py
 | | | - loader.py
 | | - model/
 | | | - __init__.py
 | | | - trainer.py
 | | - utils/
 | | | - __init__.py
 | | | - helper.py
 | - .funcignore
 | - host.json
 | - local.settings.json
 | - requirements.txt
 | - Dockerfile

inference.py:

from .data.loader import Data_loader
from .model.trainer import model_Trainer
from .utils.helper import help

def get_data():
    try:
        res = Data_loader()
        res1 = model_Trainer()
        res2 = help()

        return res, res1, res2

    except Exception as e:
        print(f"Error: {e}")

HttpTrigger1/__init__.py:

import logging
from .inference import get_data
import azure.functions as func


def main(req: func.HttpRequest) -> func.HttpResponse:
    logging.info('Python HTTP trigger function processed a request.')
    return func.HttpResponse(f"Hello, This HTTP triggered function executed successfully.{get_data()}")

data/loader.py:

def Data_loader():
    return("Loader from data folder")

model/trainer.py:

def model_Trainer():
    return("Trainer for the models folder")

utils/helper.py:

def help():
    return("Helper from utils folder")

Dockerfile:

FROM mcr.microsoft.com/azure-functions/python:4-python3.11

ENV AzureWebJobsScriptRoot=/home/site/wwwroot \
    AzureFunctionsJobHost__Logging__Console__IsEnabled=true

COPY requirements.txt /
RUN pip install -r /requirements.txt

COPY . /home/site/wwwroot

OUTPUT:

Hello, This HTTP triggered function executed successfully.('Loader from data folder', 'Trainer for the models folder', 'Helper from utils folder')

Image

Answered By - Vivek Vaibhav Shandilya