Linum shares its data filtering stack evolution for video model pre-training, from CPU heuristics to RL aesthetic scoring
Getting video models to learn better, faster
Linum is building its open-weight video model v3 and details how its data filtering pipeline evolved since 2024. Early stage used CPU-only traditional CV: PySceneDetect for shot cuts, EAST for OCR, H.264 motion vectors to drop low-motion clips, and Haar cascades to subsample talking heads. By early 2025 they moved to fine-tuned LLMs on GPUs—AutoShot+TransNetV2, PaddleOCR via TensorRT, and Qwen-2-VL-2B for categorical filters. Late 2025 brought RLVR with Qwen-2.5-VL-3B for fine-grained aesthetic scoring (1–4) and WAFT optical flow to catch remaining low-motion long-tail. The post does not disclose v3 release date or model size.
Why it matters: Linum walks through the data-filtering pipeline behind its open video model v3, from CPU classical vision to GPU aesthetic scoring, explaining the motive and trade-off at each step.