Kimi K3 DSpark
by Inferact · Model · Language & code models
4B draft model for speculative decoding of Kimi K3, drafting seven tokens per pass, on Hugging Face
AccessOpen weightsSize4B parametersLicensekimi-k3
Evidence
- Jul 29Reaches 464 tokens/s versus roughly 5-15 tok/s for other local methods; needs four GB300 GPUs minimum
Alternatives
Made from