sail
/

longspec-vicuna-7b-v1.5-16k

Text Generation

text-generation-inference

Model card Files Files and versions

This repository contains the LongSpec model described in the paper LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification.

Code: https://github.com/sail-sg/LongSpec

Downloads last month: 20

Safetensors

Model size

270M params

Tensor type

F32

·