<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Pytorch on Jonathan Franklin</title><link>https://jonnie.github.io/tags/pytorch/</link><description>Recent content in Pytorch on Jonathan Franklin</description><generator>Hugo</generator><language>en</language><lastBuildDate>Tue, 16 Jun 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://jonnie.github.io/tags/pytorch/index.xml" rel="self" type="application/rss+xml"/><item><title>Quantrade Research System</title><link>https://jonnie.github.io/work/quantrade/</link><pubDate>Tue, 16 Jun 2026 00:00:00 +0000</pubDate><guid>https://jonnie.github.io/work/quantrade/</guid><description>&lt;h2 id="problem"&gt;Problem&lt;/h2&gt;&#10;&lt;p&gt;I built Quantrade as a personal research system for machine-learning trading strategies. I wanted to compare ideas on consistent data and evaluation rules, rather than judge a model by prediction accuracy or a favourable chart alone. Trading fees, drawdown and the order of training and test data all change how useful a result is.&lt;/p&gt;&#10;&lt;h2 id="architecture"&gt;Architecture&lt;/h2&gt;&#10;&lt;p&gt;A Python research library connects market-data providers, feature preparation, PyTorch trainers and fee-aware trade simulations. A YAML-driven multiprocessing runner executes comparable benchmark plans and writes result tables, while streaming feature and account models connect the same research ideas to an event-driven strategy interface.&lt;/p&gt;</description></item></channel></rss>