Self-evolving memory OS for LLM & AI Agents: ultra-persistent memory, hybrid-retrieval, and cross-task skill reuse, with 35.24% token savings and DeepSeek Harness support.
LoRA post-training for open-weight models: SFT, GRPO, and on-policy distillation. Describe a run in TOML; Flash allocates a GPU, trains, streams checkpoints, and serves the adapter.
MeshCom web dashboard for radio amateurs. Connects to a MeshCom node via UDP and provides a browser-based interface: real-time map of heard stations, chat with channels and direct messages, telemetry display. Single binary, no dependencies.