I migrated my luis to rasa as per mentioned in https://rasa.com/docs/rasa/migrate-from/microsoft-luis-to-rasa/.
I have a python code which fetches answer from a QnA Maker knowledge base and gives result in json format.
Now my work is rasa has to take the result of the python code and give that result as response of the chatbot.
How can I do that?
Where should I deploy this python code?
Below is my python code.
v1 = 'where is germanium'
def conference_room():
import http.client, urllib.parse, json, time, sys
host = "newbot.azurewebsites.net"
endpoint_key = "xxxxxxxx-8xxe-xxxx-9xx2-7cxx4fxxxxx"
route = "/qnamaker/knowledgebases/xxxx052f-4xxa-xx7d-xxxx-4xxadxxxxxx/generateAnswer"
q1 = {'question':v1}
question = str(q1)
headers = {
'Authorization': 'EndpointKey ' + endpoint_key,
'Content-Type': 'application/json'
}
try:
conn = http.client.HTTPSConnection(host,port=443)
conn.request ("POST", route, question, headers)
response = conn.getresponse ()
answer = response.read ()
print(json.dumps(json.loads(answer), indent=4))
except :
print ("Unexpected error:", sys.exc_info()[0])
print ("Unexpected error:", sys.exc_info()[1])
conference_room()
You can implement this as a https://rasa.com/docs/rasa/core/actions/#custom-actions . Using rasa-sdk, you would need to define an action that executes dispatcher.utter_message(<your formatted output here>) as part of its run method.
Related
I have a karate feature which needs to call a JavaScript function with arguments from js file. Not in the feature itself. Want to organise many js function in separate file and call it from a feature file. Just like creating java step definitions with cucumber.
Feature: Inside new
Background:
* def result = call read(‘GetTokens.feature')
Given url devURL
* def at = result.access_token
* def rt = result.refresh_token
* def data = read('classpath:src/main/resources/GetHeaders.js', at, rt)
* configure headers = data
Scenario Outline: new
Given path ‘new’
And request ‘’
When method post
And status 201
This feature calls GetHeaders.js But not able to read the arguments. Is there anything wrong with the way it's been called. Or karate framework/nashorn can't support.
#GetHeaders.js
var f1 = function(at, rt){
var head = {
'Content-Type': 'application/json',
'Accept': 'application/json',
'token': at,
'refresh-token': rt,
'channel': 'online'
}
return head
}
Move the js code to a function within a feature file. The function can be called using any number of arguments just like any other js or java functions. Below is the sample code incase you want any reference.
* def GetHeaders =
"""
function(at, rt){
var head = {
'Content-Type': 'application/json',
'Accept': 'application/json',
'token': at,
'refresh-token': rt,
'channel': 'online'
}
return head
}
"""
* karate.log(GetHeaders('shfusdfhdskfds', 'ueworuewoew767638'))
I am have some problems passing in the correct headers for my graphql endpoints
The use case in Postman:
call requestToken endpoint to obtain sessionToken value
requestToken response contains Key Value " and Token Value.
For subsequent calls, I set postman headers as:
Key = X_SESSION_TOKEN Value = Token Value
The user case in Karate
1st feature 'requestToken.feature' successfully calls and stores key + tokenValue
2nd feature successfully defines and prints the token value
here is my 2nd request:
Feature: version
Background:
* url 'http://api-dev.markq.com:5000/'
* def myFeature = call read('requestToken.feature')
* def authToken = myFeature.sessionToken
* configure headers = { 'X_SESSION_TOKEN': authToken , 'Content-Type': 'application/json' }
Scenario: get version
Given path 'query'
Given text query =
"""
query {
version
}
"""
And request { query: '#(query)' }
When method POST
Then status 200
And print authToken
And print response
I am not sure I send the headers right. Its coming back 200, but I keep getting a error 'token malformed' in the response message
Any suggestions? New at this, thanks!
Honestly this is hard to answer, a LOT depends on the specific server.
EDIT: most likely it is this change needed, explained here: https://github.com/intuit/karate#embedded-expressions
* configure headers = { 'X_SESSION_TOKEN': '#(authToken)' , 'Content-Type': 'application/json' }
2 things from experience:
should it be X-SESSION-TOKEN
add an Accept: 'application/json' header
And try to hardcode the headers before attempting call etc.
Here is an example that works for me:
* url 'https://graphqlzero.almansi.me/api'
* text query =
"""
{
user(id: 1) {
posts {
data {
id
title
}
}
}
}
"""
* request { query: '#(query)' }
* method post
* status 200
How to invoke sagemaker endpoint with input data type numpy.ndarray.
I have deployed a sagemaker model and trying to hit it using lambda function.
But I am unable to figure out how to do it. I am getting server error.
One row of the Input data.
The total data set has shape=(91,5,12).
The below is only one row of Input data.
array([[[0.30440741, 0.30209799, 0.33520652, 0.41558442, 0.69096432,
0.69611016, 0.25153326, 0.98333333, 0.82352941, 0.77187154,
0.7664042 , 0.74468085],
[0.30894981, 0.33151662, 0.22907725, 0.46753247, 0.69437367,
0.70410559, 0.29259044, 0.9 , 0.80882353, 0.79401993,
0.89501312, 0.86997636],
[0.33511896, 0.34338939, 0.24065546, 0.48051948, 0.70384005,
0.71058715, 0.31031288, 0.86666667, 0.89705882, 0.82724252,
0.92650919, 0.89125296],
[0.34617355, 0.36150251, 0.23726854, 0.54545455, 0.71368726,
0.71703244, 0.30228356, 0.85 , 0.86764706, 0.86157254,
0.97112861, 0.94089835],
[0.36269508, 0.35923332, 0.40285461, 0.62337662, 0.73325475,
0.7274392 , 0.26241391, 0.85 , 0.82352941, 0.89922481,
0.9343832 , 0.90780142]]])
I am using the following code but unable to invoke the endpoint
import boto3
def lambda_handler(event, context):
# The SageMaker runtime is what allows us to invoke the endpoint that we've created.
runtime = boto3.Session().client('sagemaker-runtime')
endpoint = 'sagemaker-tensorflow-2019-04-22-07-16-51-717'
print('givendata ', event['body'])
# data = numpy.array([numpy.array(xi) for xi in event['body']])
data = event['body']
print('numpy array ', data)
# Now we use the SageMaker runtime to invoke our endpoint, sending the review we were given
response = runtime.invoke_endpoint(EndpointName = endpoint,# The name of the endpoint we created
ContentType = 'application/json', # The data format that is expected
Body = data) # The actual review
# The response is an HTTP response whose body contains the result of our inference
result = response['Body'].read().decode('utf-8')
print('response', result)
# Round the result so that our web app only gets '1' or '0' as a response.
result = round(float(result))
return {
'statusCode' : 200,
'headers' : { 'Content-Type' : 'text/plain', 'Access-Control-Allow-Origin' : '*' },
'body' : str(result)
}
I am unable to figure out what should be written in place of ContentType.
Because I am not aware of MIME type in case of numpy.ndarray.
Illustration of what I had and how I solved
from sagemaker.tensorflow import TensorFlowPredictor
predictor = TensorFlowPredictor('sagemaker-tensorflow-serving-date')
data = np.array(raw_data)
response = predictor.predict(data=data)
predictions = response['predictions']
print(predictions)
What I did to find the answer:
Looked up predictions.py and content_types.py implementation in sagemaker python library to see what content types it used and what arguments it had.
First I thought that application/x-npy content_type was used and thus tried using serialisation code from predictor.py and passing the application/x-npy as content_type to invoke_endpoint.
After receiving 415 (unsupported media type), the issue was still the content_type. The following print statements helped me to reveal what content_type predictor actually uses (application/json) and thus I took the appropriate serialisation code from predictor.py
from sagemaker.tensorflow import TensorFlowPredictor
predictor = TensorFlowPredictor('sagemaker-tensorflow-serving-date')
data = np.array(raw_data)
response = predictor.predict(data=data)
print(predictor.content_type)
print(predictor.accept)
predictions = response['predictions']
print(predictions)
TL;DR
Solution for lambda:
import json
import boto3
ENDPOINT_NAME = 'sagemaker-tensorflow-serving-date'
config = botocore.config.Config(read_timeout=80)
runtime= boto3.client('runtime.sagemaker', config=config)
data = np.array(raw_data)
payload = json.dumps(data.tolist())
response = runtime.invoke_endpoint(EndpointName=ENDPOINT_NAME,
ContentType='application/json',
Body=payload)
result = json.loads(response['Body'].read().decode())
res = result['predictions']
Note: numpy is not included in lambda thus you would either include the numpy yourself or instead of data.tolist() operate with python list and json.dump that list (of lists). From your code it seems to me you have python list instead of numpy array, so simple json dump should work.
If you are training and hosting custom algorithm on SageMaker using TensorFlow, you can serialize/de-serialize the request and response format as JSON as in TensorFlow Serving Predict API.
import numpy
from sagemaker.predictor import json_serializer, json_deserializer
# define predictor
predictor = estimator.deploy(1, instance_type)
# format request
data = {'instances': numpy.asarray(np_array).astype(float).tolist()}
# set predictor request/response formats
predictor.accept = 'application/json'
predictor.content_type = 'application/json'
predictor.serializer = json_serializer
predictor.deserializer = json_deserializer
# run inference using SageMaker predict class
# https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/predictor.py
predictor.predict(data)
You can refer the example notebook here to train and host custom TensorFlow container.
How do I obtain the req in falcon as json and not string as seperate key value pairs.
If {"a:213","b":32435}
How do i make sure a is passed and then obtain value of a
I hope following code will help you.
json_data = json.loads(req.stream.read())
try:
value_a = json_data['a']
except KeyError as k:
print 'a is not passed'
Not sure if that's what you asked, but you can transform you raw request (req) to json by using:
if req.content_length:
doc = json.load(req.stream)
I think following code will help you:
json_data = json.loads(req.stream.read())
OR if you want to specify specific encoding format of input data.
json_data = json.loads(req.stream.read().decode('utf8'))
Please let me know you need further clarification.
Use
stream = req.bounded_stream.read()
or
stream = req.stream.read()
I created a BodyParser class as a middleware:
class BodyParser(object):
def __init__(self, ctx):
self.ctx = ctx
def process_request(self, req, resp):
if req.method.upper() in ['POST', 'PUT', 'PATCH']:
stream = req.stream.read()
if not stream:
req.context['body'] = None
return
req.context['body'] = json.loads(stream)
Hope it helps
after you translate the req/resp into JSON with:
json.load(req.stream)
you can look at the output as a regular dictionary.
I'm trying to replicate the following successful cURL operation with Grinder.
curl -X PUT -d "title=Here%27s+the+title&content=Here%27s+the+content&signature=myusername%3A3ad1117dab0ade17bdbd47cc8efd5b08" http://www.mysite.com/api
Here's my script:
from net.grinder.script import Test
from net.grinder.script.Grinder import grinder
from net.grinder.plugin.http import HTTPRequest
from HTTPClient import NVPair
import hashlib
test1 = Test(1, "Request resource")
request1 = HTTPRequest(url="http://www.mysite.com/api")
test1.record(request1)
log = grinder.logger.info
test1.record(log)
m = hashlib.md5()
class TestRunner:
def __call__(self):
params = [NVPair("title","Here's the title"),NVPair("content", "Here's the content")]
params.sort(key=lambda param: param.getName())
ps = ""
for param in params:
ps = ps + param.getValue() + ":"
ps = ps + "myapikey"
m.update(ps)
params.append(NVPair("signature", ("myusername:" + m.hexdigest())))
request1.setFormData(tuple(params))
result = request1.PUT()
The test runs okay, but it seems that my script doesn't actually send any of the params data to the API, and I can't work out why. There are no errors generated, but I get a 401 Unauthorized response from the API, indicating that a successful PUT request reached it, but obviously without a signature the request was rejected.
This isn't exactly an answer, more of a workaround that I came up with, that I've decided to post since this question hasn't yet received any responses, and it may help anyone else trying to achieve the same thing.
The workaround is basically to use the httplib and urllib modules to build and make the PUT request instead of the HTTPClient module.
import hashlib
import httplib, urllib
....
params = [("title", "Here's the title"),("content", "Here's the content")]
params.sort(key=lambda param: param[0])
ps = ""
for param in params:
ps = ps + param[1] + ":"
ps = ps + "myapikey"
m = hashlib.md5()
m.update(ps)
params.append(("signature", "myusername:" + m.hexdigest()))
params = urllib.urlencode(params)
print params
headers = {"Content-type": "application/x-www-form-urlencoded"}
conn = httplib.HTTPConnection("www.mysite.com:80")
conn.request("PUT", "/api", params, headers)
response = conn.getresponse()
print response.status, response.reason
print response.read()
conn.close()
(Based on the example at the bottom of this documentation page.)
You have to refer to the multi-form posting example in Grinder script gallery, but changing the Post to Put. It works for me.
files = ( NVPair("self", "form.py"), )
parameters = ( NVPair("run number", str(grinder.runNumber)), )
# This is the Jython way of creating an NVPair[] Java array
# with one element.
headers = zeros(1, NVPair)
# Create a multi-part form encoded byte array.
data = Codecs.mpFormDataEncode(parameters, files, headers)
grinder.logger.output("Content type set to %s" % headers[0].value)
# Call the version of POST that takes a byte array.
result = request1.PUT("/upload", data, headers)