Search Articles — Sudonull

https://sudonull.com/meshtastic-build-on-esp32-s3-lora-e22

Step-by-step build Meshtastic node on ESP32-S3 with LoRa E22-900M30S, GPS and OLED. Firmware adaptation, pinout, PlatformIO. For middle/senior dev.

https://sudonull.com/fine-tuning-mistral-7b-on-russian-for-dialogues

Learn why LoRA didn't work and how full fine-tuning Mistral 7B achieved BertScore 0.93. Architecture, optimizations, MERA benchmarks for ML engineers.

https://sudonull.com/budget-meshtastic-node-for-500-rub

Assemble Meshtastic node on ESP32-C3/NRF52 + LoRa SX1276 for 500 rub. Connection diagrams, firmware, antenna. Instructions for IoT developers.

https://sudonull.com/fine-tuning-gemma-4-on-cloud-run-with-gpu

Instructions for fine-tuning Gemma 4 31B on Cloud Run Jobs for image classification. QLoRA, LoRA rank 64, accuracy 94%. For middle/senior developers. Start fine-tuning now.

https://sudonull.com/meshtastic-deployment-architecture-and-nodes

Practice of creating Meshtastic network for field teams: choosing HELTEC V4, solar repeaters, Mosquitto bridge. Range, architecture, tests. Instructions for developers.

Jan 7, 2026 ·

https://apatero.com/blog/ultimate-guide-lora-training-2025

Jan 7, 2026 · LoRA stands for Low-Rank Adaptation. Instead of retraining an entire AI model (expensive, slow), LoRA trains a small adapter that modifies the base model's behavior.

Jul 14, 2026 ·

https://blog.paessler.com/what-is-lora-a-beginners-guide-part-1

Jul 14, 2026 · LoRa, which stands for Long Range , is a low power wide area network (LPWAN) wireless communication technology that enables Internet of Things (IoT) devices and sensors to communicate with each other over long distances while maintaining low power consumption.

Jun 20, 2025 ·

https://www.geeksforgeeks.org/deep-learning/fine-tuning-using-lora-and-qlora/

Jun 20, 2025 · In contrast, LoRA (Low-Rank Adaptation) is a parameter-efficient technique that introduces small trainable matrices into certain layers, allowing most of the original model parameters to remain unchanged.

May 2, 2026 ·

https://www.geeksforgeeks.org/deep-learning/low-rank-adaptation-lora/

May 2, 2026 · Low-Rank Adaptation (LoRA) is a parameter-efficient fine-tuning technique used to adapt large pre-trained models for specific tasks with minimal computational and memory overhead.

Jun 17, 2021 ·

https://arxiv.org/abs/2106.09685

Jun 17, 2021 · We propose Low-Rank Adaptation, or LoRA , which freezes the pre-trained model weights and injects trainable rank decomposition matrices into each layer of the Transformer architecture, greatly reducing the number of trainable parameters for downstream tasks.

Searches related to lora

https://github.com/microsoft/LoRA

LoRA : Low-Rank Adaptation of Large Language Models This repo contains the source code of the Python package loralib and several examples of how to integrate it with PyTorch models, such as those in Hugging Face.