Skip to content

Instantly share code, notes, and snippets.

@mirth
mirth / import_from_parent.py
Last active August 29, 2015 14:11
This gist adds parent directory to sys.path to be able import files from parent directory.
if __name__ == '__main__' and __package__ is None:
from os import sys, path
sys.path.append(path.dirname(path.dirname(path.abspath(__file__))))
@mirth
mirth / lexer.rb
Last active December 11, 2015 22:49 — forked from dustalov/lexer.rb
# encoding: utf-8
# Processor of Link Grammar for Russian output.
#
class LinkParser::Lexer
# This exception raises when link grammar is invalid and Lexer
# is unable to understand the output.
#
class InvalidLinkGrammar < RuntimeError
attr_reader :input
6ля
6лядь
6лять
b3ъeб
cock
cunt
e6aль
ebal
eblan
eбaл
@mirth
mirth / tuple_to_strings.cpp
Created April 20, 2016 21:20 — forked from mumreg/tuple_to_strings.cpp
Strings formating approach
#include <iostream>
#include <cstddef>
#include <tuple>
#include <utility>
#include <sstream>
#include <vector>
template <typename Tuple, typename F, std::size_t ...Indices>
void for_each_impl(Tuple&& tuple, F&& f, std::index_sequence<Indices...>) {
using swallow = int[];
#include <iostream>
#include <memory>
#include <vector>
#include "visitor.hpp"
struct foo;
struct bar;
using visitor_t = basic_visitor<std::tuple<foo, bar>, void(int)>;
#include <functional>
#include <boost/mpl/inherit_linearly.hpp>
#include <boost/mpl/inherit.hpp>
#include <boost/mpl/vector.hpp>
template<typename TargetsVector, typename SignatureVector>
struct visit_method;
template<typename TargetType, typename ReturnType, typename ... ArgsTypes>
struct visit_method<TargetType, std::tuple<ReturnType, ArgsTypes ...>> {
import os
import numpy as np
import pandas as pd
from sklearn.model_selection import StratifiedKFold
prod_to_cat = pd.read_csv('levchik_folds/prod_to_category.csv')
skf = StratifiedKFold(n_splits=5, shuffle=True, random_state=4242)
split = skf.split(prod_to_cat.drop('category_id', axis=1), prod_to_cat['category_id'])
@baker.command
def run(batch_size=8, epochs=5, device='cuda'):
train_loader, val_loader = get_data_loaders(batch_size)
model = NoiseClassifier()
model = model.to(device)
model.eval()
optimizer = Adam(model.parameters())
criterion = nn.BCELoss()
noise = [
'weird engine sound',
'engine buzzing/farting sound',
'car engine noise',
'Engine knocking',
'Engine bad sound',
'Engine Whining Noise',
'engine blown sounds',
'engine tick sound',
'engine Clicking Noise',
healthy = [
'healthy engine sound',
'Sound of a Healthy Engine',
]