Yunus Seyhan Dede ysdede

LSTM auto complete model

Train your own auto-complete model on the web with Tensorflow.js

OS support

Chrome Web browser

Usage

Download predicts.html
Upload dataset.txt

date

title

1. Download your CUDA `runfile (local)` packages

Permanent WSL DNS Fix (WSL 2.2.1+)

If you're encountering ping github.com failing inside WSL with a Temporary failure in name resolution, you're not alone — this has been a long-standing issue, especially when using VPNs or corporate networks.

This issue is now fixed robustly with DNS tunneling, which preserves dynamic DNS behavior and avoids limitations like WSL’s former hard cap of 3 DNS servers in /etc/resolv.conf.

DNS tunneling is enabled by default in WSL version 2.2.1 and later, meaning that if you're still seeing DNS resolution issues, the first and most effective fix is simply to upgrade WSL. Upgrading WSL updates the WSL platform itself, but does not affect your installed Linux distributions, apps, or files.

To upgrade WSL, follow these steps,

List of desktop Nvidia GPUS ordered by CUDA core count

I created it for those who use Neural Style

Guys, please add your hardware setups, neural-style configs and results in comments!

GPU	CUDA cores	Memory	Processor frequency	Memory Bandwidth (GB/sec)
Geforce GTX TITAN Z	5760	12 GB	705 / 876	673

	from datasets import load_dataset
	import torch
	import sacrebleu
	import json
	from transformers import AutoModelForCausalLM, AutoProcessor, GenerationConfig
	import soundfile as sf
	from jiwer import cer
	from whisper_normalizer.basic import BasicTextNormalizer
	from tqdm.auto import tqdm
	import re

	{
	"class": {
	"0": "person",
	"1": "bicycle",
	"2": "car",
	"3": "motorcycle",
	"4": "airplane",
	"5": "bus",
	"6": "train",
	"7": "truck",

	from urllib.request import urlopen
	from io import BytesIO
	from zipfile import ZipFile


	def download_and_unzip(url, extract_to='.'):
	http_response = urlopen(url)
	zipfile = ZipFile(BytesIO(http_response.read()))
	zipfile.extractall(path=extract_to)

	<!DOCTYPE html>
	<html>

	<head>
	<meta charset="UTF-8">
	<title>A simple clock</title>

	</head>

	<body translate="no" >