Dev trains a 9B Qwen3.5 base model on SVG generation to test a new training framework
cephaloform · x · 2026-10-11
Developer cephaloform is training a 9B Qwen3.5 base model on SVG generation as a testbed for their new training framework (project link included). The experiment: have a base model learn to emit vector graphics code directly, while validating the framework itself.
More from Research
- Stanford AIMI's Sergios Gatidis wins NIH Director's New Innovator Award for medical imaging AI — StanfordAILab · 2026-10-11
- Mamba author highlights SSMs for long-sequence genomic prediction, open-sourcing Cerberus models — StanfordAILab · 2026-10-11
- Build your own single-pass decision model: masking vocabulary with Qwen3-1.7B to mimic Jev — bibryam · 2026-10-11
- MIT's Song Han launches Fall 2026 course on efficient ML, covering pruning, quantization and LLM deployment — HildeKuehne · 2026-10-11
- Free ML systems books from Harvard's Vijay Janapa Reddi and a free LLM foundations textbook — caglar_ee · 2026-10-11
- Claude replicates astronomer's 400-hour work in 30 minutes, finds hidden planetary system in old data — scottleibrand · 2026-10-11