An experimental lab for SFT, RL and Decision Models. Train LoRA, Engram or Neuron Training adapters and build continual learning loops with our harness or your own.
Choose what to learn or change, then choose an adapter.
| Workflow | Adapters | Purpose |
|---|---|---|
| SFT | LoRAEngram adapterNeuron training | Learn desired responses from reviewed demonstrations. |
| RL | LoRAEngram adapterNeuron training | Improve a policy using GRPO rewards. |
| Decision model | LoRAEngram adapterNeuron training | Train or score allowed choices from structured state. |
Train with a workflow and adapter, then evaluate the candidate in an experiment.
Example: RL + LoRA
Work with agents beside your local projects, then turn a dataset into an evaluation plan without leaving the app.
Start with a real task.
