manisnesan

from sklearn.model_selection import train_test_split

def SplitSet(df):
     train, test = train_test_split(df, test_size=0.1)
     train, valid = train_test_split(df, test_size=0.2)
     split_val = len(train)
     train = train.append(valid)
     return train, test, split_val

df = pd.from_csv(...)

manisnesan

import time

from fastapi import FastAPI, Request

app = FastAPI()

# Here before retruning the response, a field called 'X-Process-Time' is added to the headers
@app.middleware("http")
async def add_process_time_header(request: Request, call_next):
    start_time = time.time()

manisnesan

Chrestotes

What is Chrestotes

Why Mani doing this

Who am I

Review of background info and Understanding the Context

manisnesan

Working example q => vendor_name:"BalaBit S.á.r.l"

Debug section

{
    "debug": {
        "queryBoosting": {
            "q": "vendor_name:\"BalaBit S.á.r.l\"",
            "match": null
        },

manisnesan

BoostedQuery(
    boost(
          +(
            +(
              ((psfg:jboss)^2.0 | (psfc:jboss)^4.5 | (psfd:jboss)^3.5 | (psfe:jboss)^4.0 | (psff:jboss)^2.5 | (psfa:jboss)^10.0 | (psfb:jboss)^5.5)
                
                ((psfg:jboss)^2.0 | (psfc:jboss)^4.5 | (psfd:jboss)^3.5 | (psfe:jboss)^4.0 | (psff:jboss)^2.5 | (psfa:jboss)^10.0 | (psfb:jboss)^5.5))
                
                  -((psfg:webserver-3)^2.0 | (psfc:webserver-3)^4.5 | (psfd:webserver-3)^3.5 | (psfe:webserver-3)^4.0 | (+psff:webserv +psff:3)^2.5 | (psfa:webserver-3)^10.0 | (psfb:webserver-3)^5.5)) 
                  

manisnesan

# Source: https://muellerzr.github.io/fastshap/
from fastai2.tabular.all import *

## Download
path = untar_data(URLs.ADULT_SAMPLE)
df = pd.read_csv(path/'adult.csv')

## Preprocessing
dep_var = 'salary'
cat_names = ['workclass', 'education', 'marital-status', 'occupation', 'relationship', 'race']

manisnesan

# https://forums.fast.ai/t/how-to-check-your-pytorch-keras-is-using-the-gpu/7232
import torch
torch.cuda.current_device()
torch.cuda.device(0)
torch.cuda.device_count()
torch.cuda.get_device_name(0)

manisnesan

def load_model():
    #global vectorizer, model
    vectorizer = pickle.load(open(MODEL_DIR + "/tfidf_vectorizer.pkl", "rb")) 
    model = pickle.load(open(MODEL_DIR + "/intent_clf.pkl", "rb"))
    return vectorizer, model

vectorizer, model = load_model()

/home/msivanes/miniconda3/envs/anlp/lib/python3.6/site-packages/sklearn/base.py:251: UserWarning: Trying to unpickle estimator TfidfTransformer from version 0.21.2 when using version 0.20.2. This might lead to breaking code or invalid results. Use at your own risk.
  UserWarning)

manisnesan

<!-- Standard text field for content types not included in recommendations, portal search -->
    <fieldType name="text_std" class="solr.TextField" positionIncrementGap="100">
      <analyzer type="index">
      	<!-- remove period (dot) character from end of token  -->
        <charFilter class="solr.PatternReplaceCharFilterFactory" pattern="(\.)(\s|$)" replacement=" "/>
        <!-- remove question mark (?) character from end of token. Index time only  -->
        <charFilter class="solr.PatternReplaceCharFilterFactory" pattern="(\?)(\s|$)" replacement=" "/>
        <tokenizer class="solr.WhitespaceTokenizerFactory"/>
        <filter class="solr.LengthFilterFactory" min="1" max="10000" />
        <filter class="solr.LowerCaseFilterFactory"/>

manisnesan

def keyphrases(text):
    
    # define the set of valid Part Of Speech tags 
    pos = {'NOUN', 'PROPN', 'ADJ'}
    
    #create a SingleRank extractor
    singleRank_extractor = pke.unsupervised.SingleRank()
    
    # load the content of the document
    singleRank_extractor.load_document(input=text, language='en', normalization=None)