Show HN: Spikeforge

Capsize Games released Spikeforge, an open-source toolchain for spiking neural networks that encodes data into spikes, trains and inspects networks, and deploys models to targets including the lava_loihi2 backend. The project's status page lists NIR export and validation, determinism and run tracking, spikeforge-serve, the pipeline builder, the lava_loihi2 backend, model hub architectures, and the PyPI release as done, while pretrained model hub weights are not started and on-chip/local learning rules plus the speck, xylo, and spinnaker2 backends are in progress. Spikeforge uses a leaky integrate-and-fire neuron with beta=0.90, threshold=1.00, and subtract reset, and is installable via git clone from github.com/capsize-games/spikeforge followed by ./install.sh.

spikeforge $ git clone https://github.com/capsize-games/spikeforge.git $ cd spikeforge && ./install.sh $ cd spikeforge && ./install.sh Leaky integrate-and-fire neuron β=0.90, threshold=1.00, reset=subtract β=0.90, threshold=1.00, reset=subtract I t U t S t Open dashboardTrain, inspect, and deploy in the browser https://dash.spikeforge.net/ Read the docsGuides, reference, examples, and plans https://docs.spikeforge.net/ View project statusSee what is ready and what is in progress status Explore on GitHubSource, issues, releases, and contributions https://github.com/capsize-games/spikeforge AI that reacts when the world changes. Most AI processes every value at every step, even when nothing meaningful changes. Spiking neural networks work more like a nervous system: they send brief signals when events happen. That can help always-on sensors, robots, cameras, and audio tools react quickly while using less energy. Spikeforge makes that approach practical. It gives researchers and developers one place to turn data into spikes, train and inspect networks, compare deployment targets, and ship models—without assembling the entire toolchain themselves. Encoding Training Topologies Datasets Interpreter spine Introspection Deployment Model hub Dashboard Status | Component | Status | Note | |---|---|---| | NIR export & validation | done | | | Determinism & run tracking | done | | | spikeforge-serve | done | | | Pipeline builder | done | | | lava loihi2 backend | done | | | Model hub — architectures | done | | | Model hub — pretrained weights | not started | | | On-chip / local learning rules | in progress | | | speck / xylo / spinnaker2 | in progress | | | PyPI release | done | |