issues
search
LinkSoul-AI
/
Chinese-Llama-2-7b
开源社区第一个能下载、能运行的中文 LLaMA2 模型！
Apache License 2.0
2.23k
stars
202
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
华视回放
#61
full55555
opened
5 months ago
1
请问支持微调吗
#60
hzlirz
opened
10 months ago
0
requirements.txt
#59
jiangbinbin-code
opened
11 months ago
0
有人遇到过这个报错吗，我的flash_attn版本是2.3.0，跪求大佬指点。ImportError: cannot import name 'flash_attn_unpadded_qkvpacked_func' from 'flash_attn.flash_attn_interface'
#58
SoseloX
closed
1 year ago
1
训练时的prompt是什么格式的呢
#57
zozoteacher
opened
1 year ago
0
4张p40的显卡，装载模型报错CUDA out of memory，但是每张卡24G呀，这个该怎么办
#56
GzRichChen
opened
1 year ago
1
继续微调
#55
hallucination11
opened
1 year ago
0
你好，请问怎么实现公网的api呢
#54
Nerva05251228
opened
1 year ago
0
训练的数据集格式是啥
#53
631068264
opened
1 year ago
2
chinese-llama-2-7b.ipynb出错求助
#52
leoring123
opened
1 year ago
1
使用项目中的api.py启动模型，没做4bit量化，使用v100的卡，推理特别慢，经常卡住
#51
jcxian
opened
1 year ago
2
制作镜像卡住
#50
MUFCLK
opened
1 year ago
0
请教一下用open-api-server跟inference-hf 回答问题不一致的问题
#49
ibmxiang
opened
1 year ago
0
feat: add openai_api.py
#48
Leeziao
opened
1 year ago
0
请教一下公开模型的训练细节
#47
shichao-wang
opened
1 year ago
0
怎么从本地直接加载模型啊
#46
JACKzhuz
opened
1 year ago
1
feat: update model loading and support multi-round chat
#45
shiyemin
closed
1 year ago
0
为什么7b模型有26G，原始llama2 7b模型只有13G
#44
secain
closed
1 year ago
0
when I try to train the model, Can't load the configuration of '/PATH/TO/TRANSFORMERS/VERSION/LLAMA2'
#43
githubtianya
opened
1 year ago
0
调lora需要多少资源
#42
liaoweiguo
opened
1 year ago
0
有更大的模型吗，33B，70B
#41
liaoweiguo
opened
1 year ago
0
docs: Update README.md
#40
shiyemin
closed
1 year ago
0
huggingface上模型是openrail协议，项目上是apache2，不一致在哪
#39
wzg-zhuo
opened
1 year ago
0
docs: Update README.md
#38
shiyemin
closed
1 year ago
0
feat: Create pyproject.toml
#37
shiyemin
closed
1 year ago
0
快速测试代码段OOM
#36
LiuZhihhxx
opened
1 year ago
2
能提供測試分數嗎
#35
CRGBS
opened
1 year ago
0
有流的返回方式吗？
#34
tianlichunhong
opened
1 year ago
2
OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like LinkSoul/Chinese-Llama-2-7b is not the path to a directory containing a file named config.json. Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'.
#33
hxujal
opened
1 year ago
0
could not select device driver "" with capabilities: [[gpu]]
#32
hxujal
opened
1 year ago
0
4bit量化后的模型为什么这么大
#31
liuyukid
opened
1 year ago
2
datatsetsbug
#30
zyxcambridge
opened
1 year ago
1
数据集中中文样本数量
#29
pangwenfeng
opened
1 year ago
0
ileNotFoundError: Directory LinkSoul/instruction_merge_set is neither a `Dataset` directory nor a `DatasetDict` directory.
#28
zyxcambridge
opened
1 year ago
2
卡住了
#27
zyxcambridge
opened
1 year ago
0
docs: Update README.md
#26
shiyemin
closed
1 year ago
0
docs: Update README.md
#25
shiyemin
closed
1 year ago
0
docs: Update README.md
#24
shiyemin
closed
1 year ago
0
docs: Update README.md
#23
shiyemin
closed
1 year ago
0
feat: Add colab ipynb scripts
#22
ekxs0109
closed
1 year ago
1
update: new QRcode
#21
shiyemin
closed
1 year ago
0
加不了群
#20
inkoku
closed
1 year ago
1
这个是不是还没进行词表扩充
#19
longgui0318
closed
1 year ago
0
训练资源
#18
DavidYanAnDe
closed
1 year ago
1
Update README.md
#17
eltociear
closed
1 year ago
0
请问flash attention 用的是v1吗，是什么版本啊
#16
lllyyyqqq
opened
1 year ago
0
chore: ggml command use `LinkSoul` namespace
#15
soulteary
closed
1 year ago
0
full ggml support #7
#14
soulteary
closed
1 year ago
0
Initial commit
#13
ProfGabrielplus
closed
1 year ago
0
Traditional Chinese
#12
compustar
closed
1 year ago
1
Next