This repository implements a hybrid sample generation pipeline for imaging data to extend existing training datasets.
4-GPU H800 low-resource post-training optimization project for Qwen3.6-35B-A3B MoE.
For me, the goal of this project was to handle transformers fine tuning by implementing and applying optimization training techniques using limited computational resources.
Drop-in, idempotent speed patches for PGSR surface-reconstruction training — removes per-iteration PCIe transfers, redundant ones-kernel convolutions, and GPU-CPU logging syncs (plus a real EMA-logging bug) with CPU-proven numerical parity: same geometry, lower GPU bill.
GPU memory-efficient training for PyTorch - 90%+ memory savings through gradient compression
Add a description, image, and links to the training-optimization topic page so that developers can more easily learn about it.
To associate your repository with the training-optimization topic, visit your repo's landing page and select "manage topics."