PirosB3

diff --git a/docs/ref/models/meta.txt b/docs/ref/models/meta.txt
index d728714..5ca1eae 100644
--- a/docs/ref/models/meta.txt
+++ b/docs/ref/models/meta.txt
@@ -165,7 +165,12 @@ These old APIs can be replicated by either:
   ``_with_model`` variants) the properties of the desired fields.
 
 Assuming you have a model named ``MyModel``, the following substitutions
-can be made to convert your code to the new API:
+can be made to convert your code to the new API.

PirosB3

Benchmark Name	Control	Experiment	Diff
url_resolve_flat_i18n_off	0.3300260305	0.3059779525	0.0240480781 (7.29% faster)
query_prefetch_related	0.0989829183	0.0927800298	0.0062028885 (6.27% faster)
query_select_related	0.0537939668	0.0502726555	0.0035213113 (6.55% faster)
query_raw	0.0156183600	0.0146648049	0.0009535551 (6.11% faster)
l10n_render	0.0068216324	0.0066468954	0.0001747370 (2.56% faster)
query_raw_deferred	0.0157507062	0.0155839086	0.0001667976 (1.06% faster)
query_exists	0.0012296319	0.0011273146	0.0001023173 (8.32% faster)
query_count	0.0003048658	0.0002639174	0.0000409484 (13.43% faster)

PirosB3

package main

import "fmt"

type Result struct {
    Class string
    Score float64
}

type ResultSet []*Result

PirosB3

from collections import defaultdict
import glob
import json
import unittest

import numpy as np

class Station(object):

    def __init__(self, name, latitude, longitude):

PirosB3

import json
import re
from collections import defaultdict

FIRST_NUMBER = re.compile('\s\d')

def main(file, type_name):
    connections = defaultdict(list)
    with open(file) as f:
        for line in f.readlines():

PirosB3

use std::collections::HashMap;
use std::num::Float;

struct KNNClassifier<'a> {
    data: &'a [[f64, ..4]],
    targets: &'a [int],
}

impl<'a> KNNClassifier<'a> {

PirosB3

use std::io;
use std::io::BufferedReader;
use std::collections::HashMap;
use std::io::File;
use std::rand::{task_rng, Rng};

struct MarkovModels<'a> {
    data: HashMap<(&'a str, &'a str), Vec<&'a str>>,
    words: &'a [&'a str]
}

PirosB3

use std::io::BufferedReader;
use std::collections::HashMap;
use std::io::File;
use std::rand::{task_rng, Rng};

struct MarkovModels<'a> {
    data: HashMap<(&'a str, &'a str), Vec<&'a str>>,
    words: &'a Vec<&'a str>
}

PirosB3

def smoSimple(dataIn, classLabels, C, tolerance, maxIter):
    dataMatrix = mat(dataIn)
    labelMat = mat(classLabels).T
    m, n = shape(dataMatrix)
    alphas = mat(zeros((m,1)))

    bias = 0

    iter = 0
    while (iter < maxIter):

PirosB3

copy(to_visit))