How to Deploy a Gradio Face-Restoration App on Ubuntu 22.04 with Nginx and TLS

·6 views

A new technical guide details how to build and deploy a face-restoration web app using the Gradio Python library and GFPGAN on an Ubuntu 22.04 server. The setup combines GFPGAN and Real-ESRGAN models to accept a face image as input and return two enhanced outputs through a browser-based interface. The application is configured to run persistently as a systemd service, ensuring it restarts automatically on failure or reboot. Nginx is then set up as a reverse proxy to route public web traffic to the Gradio app running locally on port 8080. The guide requires a GPU-enabled server, a registered domain name, and sudo access as baseline prerequisites.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

How Residential Proxy Networks Have Become a Key Tool for Cybercriminals

Residential proxy networks use everyday devices like smartphones, home routers, and smart TVs to route internet traffic on behalf of third parties, making it appear legitimate. Cybercriminals exploit these networks to conceal their identities, evade detection, and carry out attacks such as credential stuffing, account takeovers, and financial fraud. Devices are often enrolled into these networks without users' knowledge — through malware hidden in pirated content or cracked software, or via SDK integrations buried in app terms of service. The rise of remote work since the COVID-19 pandemic has expanded this threat, as criminal traffic routed through residential IPs blends in with legitimate home-based internet activity. Large-scale data breaches have further amplified the problem by supplying cybercriminals with stolen credentials that can be weaponized through these proxy networks.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Invoiso Adds Receipts, Thermal Printing, Dark Mode and Major Code Overhaul in July

Invoiso, a free open-source invoicing app built with Flutter for Windows, macOS, and Linux, rolled out several major updates in July. The app now supports receipts as a dedicated document type and enables direct printing via USB thermal printers with compact and detailed layout options. Dark mode was introduced across all major screens, and product management was enhanced with local-language aliases, custom units, and separate cost-price tracking. A new Grid Classic invoice template, a Daily Reports screen, and expanded font support for Indian regional languages such as Malayalam, Tamil, and Hindi were also added. Internally, the app was migrated to the Riverpod state management framework, PDF generation was modularized by template, and regression tests were introduced to improve stability.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

OpenAI Upgrades Auto-Review to GPT-5.6 Luna to Cut Agentic Workflow Costs

OpenAI has replaced GPT-5.4 with GPT-5.6 Luna as the model powering Auto-review in the ChatGPT app and Codex CLI, targeting cost reduction in automated agentic workflows. Luna is the fastest and most affordable tier in the GPT-5.6 family, priced at $0.20 per million input tokens and $1.20 per million output tokens following an approximately 80% price cut announced on July 30, 2026. Auto-review is a safety mechanism that automatically approves low-risk requests in Codex, aiming to reduce manual intervention without removing oversight entirely. OpenAI has suggested the upgrade could make agentic workflows up to ten times cheaper, though actual savings will vary based on token volumes, workflow design, and how often tasks escalate to higher-capability model tiers. The GPT-5.6 family also includes the mid-range Terra and the higher-capability Sol tiers, giving organizations a structured way to match model choice to workload demands.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer Builds Linux Kernel Module Using Pure x86-64 Assembly, No C Required

A developer created a Linux kernel module written entirely in x86-64 assembly language, bypassing the C-based toolchain that kernel development traditionally relies on. The project used the GNU Assembler (GAS) and the kernel's own kbuild system, requiring no external toolchain. The goal was to load a working module into a modern Fedora 44 kernel that would print messages on load and unload. The assembly code itself was compact, but meeting the kernel's strict binary-level contracts proved complex, with multiple failed iterations revealing hidden rules of kernel module development. The experiment was documented as a learning exercise, tracing each failure back to specific kernel mechanisms and requirements.

0 comments Read more at DEV Community