Alphazero-like Tree-Search can Guide Large Language Model Decoding and Training
Paper • 2309.17179 • Published • 2
How to use OhCherryFire/llama2-7b-game24-value with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("feature-extraction", model="OhCherryFire/llama2-7b-game24-value") # Load model directly
from transformers import AutoTokenizer, AutoModel
tokenizer = AutoTokenizer.from_pretrained("OhCherryFire/llama2-7b-game24-value")
model = AutoModel.from_pretrained("OhCherryFire/llama2-7b-game24-value")The language value model for Game24 in Alphazero-like tree-search can guide large language model decoding and training, ICML 2024
@article{feng2023alphazero,
title={Alphazero-like tree-search can guide large language model decoding and training},
author={Feng, Xidong and Wan, Ziyu and Wen, Muning and Wen, Ying and Zhang, Weinan and Wang, Jun},
journal={arXiv preprint arXiv:2309.17179},
year={2023}
}