FlashLM v4 "Bolt" — Ternary Language Model Demo

4.3M parameters | Ternary weights (-1, 0, +1) | Trained on CPU in 2 hours | No GPU used

This model uses only additions and subtractions for inference. Trained on TinyStories dataset on a free-tier 2-thread CPU (Deepnote).

Note: Some words may be missing due to the 10K vocabulary limitation. The model was trained on children's stories and works best with story-like prompts.

Model Card | v3 Demo | GitHub