This checkpoint was obtained from "
Mixture of Attentions For Speculative Decoding
" by Matthieu Zimmer*, Milan Gritta*, Gerasimos Lampouras, Haitham Bou Ammar, and Jun Wang.
The paper introduces a novel architecture for speculative decoding that enhances the speed of large language model (LLM) inference.
If you use this code or this checkpoint in your research, please cite our paper:
@misc{zimmer2024mixtureattentionsspeculativedecoding,
title={Mixture of Attentions For Speculative Decoding},
author={Matthieu Zimmer and Milan Gritta and Gerasimos Lampouras and Haitham Bou Ammar and Jun Wang},
year={2024},
eprint={2410.03804},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2410.03804},
}
License
This project is licensed under the MIT License. See the
LICENSE
file for more details.
Disclaimer: This open source project is not an official Huawei product, Huawei is not expected to provide support for this project.
Runs of huawei-noah MOASpec-Llama-3-8B-Instruct on huggingface.co
2
Total runs
0
24-hour runs
0
3-day runs
-2
7-day runs
-3
30-day runs
More Information About MOASpec-Llama-3-8B-Instruct huggingface.co Model
MOASpec-Llama-3-8B-Instruct huggingface.co
MOASpec-Llama-3-8B-Instruct huggingface.co is an AI model on huggingface.co that provides MOASpec-Llama-3-8B-Instruct's model effect (), which can be used instantly with this huawei-noah MOASpec-Llama-3-8B-Instruct model. huggingface.co supports a free trial of the MOASpec-Llama-3-8B-Instruct model, and also provides paid use of the MOASpec-Llama-3-8B-Instruct. Support call MOASpec-Llama-3-8B-Instruct model through api, including Node.js, Python, http.
MOASpec-Llama-3-8B-Instruct huggingface.co is an online trial and call api platform, which integrates MOASpec-Llama-3-8B-Instruct's modeling effects, including api services, and provides a free online trial of MOASpec-Llama-3-8B-Instruct, you can try MOASpec-Llama-3-8B-Instruct online for free by clicking the link below.
huawei-noah MOASpec-Llama-3-8B-Instruct online free url in huggingface.co:
MOASpec-Llama-3-8B-Instruct is an open source model from GitHub that offers a free installation service, and any user can find MOASpec-Llama-3-8B-Instruct on GitHub to install. At the same time, huggingface.co provides the effect of MOASpec-Llama-3-8B-Instruct install, users can directly use MOASpec-Llama-3-8B-Instruct installed effect in huggingface.co for debugging and trial. It also supports api for free installation.
MOASpec-Llama-3-8B-Instruct install url in huggingface.co: