Tian Li

RicardoLee's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes

New activity in RicardoLee/Llama2-chat-13B-Chinese-50W 11 months ago

13b的context len多大以及batch？

5

#1 opened 11 months ago by lucasjin

New activity in RicardoLee/Llama2-chat-Chinese-50W 11 months ago

此时不应降低学习率，warmup 等超参，而是应该放大到Pretrain 规模

3

#2 opened 11 months ago by daner

关于train_sft.py中coati包

2

#3 opened 11 months ago by BatmanBill

那这个怎么调用呢

4

#1 opened 11 months ago by yjianchun