- Published on
An experiment in fine-tuning small, local language models (Qwen 1.5B) to reason in a telegraphic, compressed style, saving up to 74% of reasoning tokens and speed up inference without relying on verbose prompts.
Mlx
All Posts
- how-to (3)
- visualization (3)
- agents (3)
- local-ai (3)
- rag (3)
- ml (2)
- short_read (2)
- llm (2)
- onnx (2)
- wasm (2)
- webgpu (2)
- transformersjs (2)
- python (2)
- retrieval (2)
- maths (1)
- news (1)
mlx (1)
- sft (1)
- reasoning (1)
- qwen (1)
- deepseek (1)
- voice (1)
- markdown (1)
- pdf (1)
- typesetting (1)
- serverless (1)
- aws (1)
- azure (1)
- gcp (1)
- sandboxing (1)