This message was deleted.

my code

from __future__ import annotations

import bentoml

from data_validation.schema import InputSchema

from pycaret.classification import load_model, predict_model

from fastapi.responses import JSONResponse

from fastapi import status, Response

from typing import IO, Dict, Any

from numpy.typing import NDArray

import numpy as np

import pandas as pd

from scripts.helper import model_predict

import os

from fastapi import FastAPI

import csv

from fastapi import File, UploadFile, HTTPException

from scripts.custom_runner import PyCaretRunnable, churn_model

# Model & Service names

SERVICE_NAME = 'churn-api'

MODEL_NAME = 'churn-model:aumvrosds2s2n4cx'

churn_model = bentoml.picklable_model.get(MODEL_NAME)

class PyCaretRunnable(bentoml.Runnable):

SUPPORTED_RESOURCES = ("cpu")

SUPPORTS_CPU_MULTI_THREADING = True

def __init__(self):

"""Constructor method for loading the model"""

self.model = bentoml.picklable_model.load_model(churn_model)

@bentoml.Runnable.method(batchable=False)

def make_prediction(self, input_data: dict) -> list:

"""This method uses the trained model to generate predictions.

Args:

input_data (dict): The JSON data sent by the client.

Returns:

list: A list with the predicted values.

"""

return model_predict(self.model, input_data)

model_runner = bentoml.Runner(PyCaretRunnable, models=[churn_model])

svc = bentoml.Service(SERVICE_NAME, runners=[model_runner])

fastapi_app = FastAPI()

svc.mount_asgi_app(fastapi_app)

@fastapi_app.get("/metadata")

def metadata():

return {"name": churn_model.tag.name, "version": churn_model.tag.version}

@fastapi_app.post("/predict/batch/api/v1")

async def predict_batch(file: UploadFile = File(...)):

print(file.filename)

# Handle the file only if it is a CSV

if file.filename.endswith(".csv"):

# Create a temporary file with the same name as the uploaded

# CSV file to load the data into a pandas Dataframe

with open(file.filename, "wb")as f:

f.write(file.file.read())

data = pd.read_csv(file.filename)

print(data.shape)

os.remove(file.filename)

pred = await model_runner.make_prediction.async_run(data)

# Return a JSON object containing the model predictions

return {

pred

# "Labels": model.predict(data)

}

else:

# Raise a HTTP 400 Exception, indicating Bad Request

# (you can learn more about HTTP response status codes here)

raise HTTPException(status_code=400, detail="Invalid file format. Only CSV Files accepted.")

@fastapi_app.post("/predict_fastapi")

def predict(features: InputSchema):

input_df = pd.DataFrame([features.dict()])

results = model_runner.make_prediction.run(input_df)

return { "prediction": results.tolist()[0] }

How did you start the server? Did you run “bentoml serve”?

i used uvicorn main:fastapi_app is that wrong

That won’t work because BentoML requires a bit more complex setup for the runner workers to initialize

Bentoml serve uvicorn service:fastapi_app --reload is this the command i should use to start my api