Samaya trained Nemotron 3.5 Lightning to beat Fable 5.1 on its finance research benchmark at 9% of the cost. The model was trained on Tinker together with the harness it works in: first by imitating 8 teachers, then RL on its own mistakes.
Samaya trains Nemotron 3.5 Lightning to beat Fable 5.1 at 9% cost
AISamaya trained Nemotron 3.5 Lightning on Tinker to beat Fable 5.1 on its finance research benchmark at 9% of the cost. Training combined imitation of 8 teachers with reinforcement learning on the model's own mistakes, along with the harness it runs in.
@tinkerapi
Introducing Samaya Sterling: our first post-trained model for state-of-the-art financial research. On FrontierFinance, Sterling outperforms Claude Fable 5.1 (58.9% vs. 55.9%) at 11× lower cost. Using a model with just 3B active parameters, Sterling demonstrates what's possible when specialized training meets a purpose-built financial research harness. Read the full research ↓
Source details
Tinker · x.comPublished · added here
Original title: Samaya trained Nemotron 3.5 Lightning to beat Fable 5.1 on its finance research benchmark at 9% of the cost. The model was trained on Tin...