issues
search
seanzhang-zhichen
/
llama3-chinese
Llama3-Chinese是以Meta-Llama-3-8B为底座,使用 DORA + LORA+ 的训练方法,在50w高质量中文多轮SFT数据 + 10w英文多轮SFT数据 + 2000单轮自我认知数据训练而来的大模型。
Apache License 2.0
289
stars
21
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
用llama-factory sft deepctrl的数据集,pyarrow报错,repo主有遇到过这个问题吗
#8
jamestang0219
opened
6 months ago
5
70B的考虑出个吗吗?
#7
0sengseng0
opened
6 months ago
1
微调后的模型batch推理有问题
#6
orderer0001
closed
6 months ago
2
效果差距很大
#5
icefairy
closed
7 months ago
2
大佬,能不能写一个基于deepctrl-sft-data微调的教程,谢谢
#4
bestlee666
closed
7 months ago
1
将这个ggml模型通过llama.cpp 转为gguf格式,运行很慢不知道为啥
#3
kevinchi8781
opened
7 months ago
3
merge_lora.py
#2
LsdFuture
closed
7 months ago
1
欢迎也发布到wisemodel.cn开源社区
#1
LiuDQ-wm
closed
7 months ago
0