Hugging Face's logo

IDEA-CCNL
/
Ziya-LLaMA-7B-Reward

Text Classification
Transformers
PyTorch
llama
custom_code
text-generation-inference
Model card Files Community
14

tokenizer 的 max-len 为什么不是llama的2048而是1024呢？

#13
by shizhoudong - opened
Discussion
shizhoudong

是因为训练数据中最大长度只有1024吗?

Sign up or log in to comment