PUMA: Secure Inference of LLaMA-7B in Five Minutes. (arXiv:2307.12533v2 [cs.CR] UPDATED)

With ChatGPT as a representative, tons of companies have began to provide
services based on large Transformers models. However, using such a service
inevitably leak users' prompts to the model provider. Previous studies have
studied secure inference for Transformer models using secure multiparty
computation (MPC), where model parameters and clients' prompts are kept secret.
Despite this, these frameworks are still limited in terms of model performance,
efficiency, and deployment. To address these limitations, we propose framework
PUMA to enable fast …

chatgpt companies computation large leak mpc prompts puma secure multiparty computation service services studies transformers