def create_pic(test_num, names, model_name):
if not names[model_name]:
name = '1'
else:
name = str(max(names[model_name]) + 1)
folder_name = 'result_' + model_name[-1]
if model_name[-1] == '1':
score = model.predict(test_num, 1, folder_name, name)
elif model_name[-1] == '2':
score = model.predict(test_num, 2, folder_name, name)
elif model_name[-1] == '3':
score = model.predict(test_num, 3, folder_name, name)
return name, score
async def create_prediction(filename: str = ""):
"""
Creates a new prediction using the model. This method must be called after the init() method has run
at least once, otherwise this will fail with a HTTP Error. When given a filename, the server will create a
file-like object of the image file and pass that to the predict() method.
:param filename: Image file name for an image stored in shared Docker volume photoanalysisserver_images
:return: JSON with field "result" containing the results of the model prediction.
"""
# Ensure model is ready to receive prediction requests
if not model_settings.ready_to_predict:
raise PredictionException()
# Attempt to open image file
try:
image_file = open('src/images/' + filename, 'r')
image_file.close()
except IOError:
logger.debug('Unable to open file: ' + filename)
return JSONResponse(status_code=status.HTTP_400_BAD_REQUEST,
content={
"status":
'failure',
"detail":
"Invalid file name provided: [" + filename +
"]. Unable to find image on server."
})
# Create prediction with model
result = predict(image_file)
return {'status': 'success', "result": result}
def info():
"""
returns the text required from the subject and topic given.
:return: the information required.
"""
# extracting information on the query
user_email = request.args.get('email')
topic = request.args.get('topic')
subject = request.args.get('subject')
print(f'email: {user_email}, topic: {topic}, subject: {subject}')
# gathering information of the student from the database
student_info = database.get_student_info(user_email)
# print(f'student details: {student_info}')
# predicting the marks of the student.
marks = model.predict(student_info, subject)
print(f'the marks of the student is: {marks}')
# getting the information.
info = database.get_info_file(subject, topic)
# getting the text depending on the student marks.
if marks < 20:
info = info["Weak"]
elif 20 <= marks <= 70:
info = info["Average"]
else:
info = info["Intelligent"]
return jsonify({'status': 200, 'info': info})
def main(request):
"""Responds to any HTTP request.
Args:
request (flask.Request): HTTP request object.
Returns:
The response text or any set of values that can be turned into a
Response object using
`make_response <http://flask.pocoo.org/docs/1.0/api/#flask.Flask.make_response>`.
"""
request_json = request.get_json()
# @TODO add any request validations that you need
result = predict(model=model, input=request_json)
return jsonify(result)
def predict_image(image_hash, image_file_name, server_port, model_name):
result = predict(image_file_name) # Create prediction on model
try: # Send prediction results back to server
headers = {'api_key': API_KEY}
r = requests.post('http://host.docker.internal:' + server_port +
'/model/predict_result',
headers=headers,
json={
'model_name': model_name,
'image_hash': image_hash,
'results': result
})
r.raise_for_status()
except (requests.exceptions.ConnectionError, requests.exceptions.Timeout,
requests.exceptions.HTTPError):
print('Unable to send prediction results to server. Hash: "' +
image_hash + '"')
async def create_prediction(image_file: UploadFile = File(...)):
"""
Creates a new prediction using the model. This method must be called after the init() method has run
at least once, otherwise this will fail with a HTTP Error. When given a filename, the server will create a
file-like object of the image file and pass that to the predict() method.
:param filename: Image file name for an image stored in shared Docker volume photoanalysisserver_images
:return: JSON with field "result" containing the results of the model prediction.
"""
# Ensure model is ready to receive prediction requests
if not model_settings.ready_to_predict:
raise PredictionException()
# Create prediction with model
result = predict(image_file)
return {
'status': 'success',
"result": result
}
from src.model.model import predict, load_model
model = load_model()
def {{.FunctionName}}(event, context):
"""Responds to any HTTP request.
Args:
event (usually a dict): An event is a JSON-formatted document that contains
data for a Lambda function to process.
context: This object provides methods and properties that provide information
about the invocation, function, and runtime environment.
Returns:
If the handler returns objects that can't be serialized by json.dumps,
the runtime returns an error.
"""
# @TODO add any request validations that you need
return predict(model=model, input=event)
def main():
request_json = request.get_json()
# @TODO add any request validations that you need
result = predict(model=model, input=request_json)
return jsonify(result)
from flask import jsonify
from src.model.model import predict, load_model
model = load_model()
def {{.FunctionName}}(request):
"""Responds to any HTTP request.
Args:
request (flask.Request): HTTP request object.
Returns:
The response text or any set of values that can be turned into a
Response object using
`make_response <http://flask.pocoo.org/docs/1.0/api/#flask.Flask.make_response>`.
"""
request_json = request.get_json()
# @TODO add any request validations that you need
result = predict(model=model, input=request_json)
return jsonify(result)
def test_predict(input_dict, expected, test_model):
result = model.predict(test_model, input_dict)
assert result == expected