Comments (16)
先按README安装一下
from hcgf.
RuntimeError:
CUDA Setup failed despite GPU being available. Inspect the CUDA SETUP outputs above to fix your environment!
If you cannot find any issues and suspect a bug, please open an issue with detals about your environment:
https://github.com/TimDettmers/bitsandbytes/issues
from hcgf.
如果内存足够的话可以不用8bit版本。这个包是8bit版本要安装的。
from hcgf.
我的显卡是4070Ti,12G的,应该用这个吧。ChatGlM-6B本地部署都好使了,还有Stable_Diffusion、codeFormer等AI程序都能运行,为什么本程序提示CUDA Setup failed呢?
from hcgf.
因为你用的是8bit模式,必须安装README里提到的依赖:bitsandbytes。
或者你就别用这种模式加载。
from hcgf.
我已经安装了bitsandbytes依赖呀。若卸载掉bitsandbytes,不提示CUDA Setup failed,而是提示bitsandbytes没安装;若重新安装了bitsandbytes,又提示CUDA Setup failed。我该怎么处理?
from hcgf.
把执行过程和错误都发一下看看。这个应该是你的环境哪里没弄好。
from hcgf.
from hcgf.
建议conda隔离环境,一个环境一个项目,虽然浪费一点磁盘空间但是值得。
项目太多了,难免会A项目要2版本,B项目要3版本,又不能兼得。
from hcgf.
@jlliushi 你这个应该是 bitsandbytes
就没装好。可以试一下 import bitsandbytes
看看,bitsandbytes
相关的可以参考它的文档,按照官方提示安装。看报错信息是没找到cuda,这个包是针对cuda的。
from hcgf.
确实是 bitsandbytes 没装好,但怎么能装好呢?
当我把 bitsandbytes 卸载后,import hcgf 就不报错了,但运行 gl = hcgf.GlmLora("model", load_in_8bit=True) 时,报 No module named 'bitsandbytes' 错误;运行 gl = hcgf.GlmLora("model", device="cuda:0") 时,只是爆显存,并未报 CUDA Setup failed 。
当再把 bitsandbytes 装上后,又报 CUDA Setup failed 了。
from hcgf.
终于鼓捣好使了:Windows10系统安装 bitsandbytes 时,需要这样
pip install bitsandbytes-windows
from hcgf.
ok,Windows开发真要命。。。
from hcgf.
linux和windows各有各的好处吧,但我都想用,我就主系统装了Proxmox VE,然后开windows虚拟机、linux虚拟机、容器什么的,想用随时切,游戏训练两不误,想组装学习机可以参考一下~
from hcgf.
微调后对本领域问题回答完全胡说八道!在已有本领域问答对文本的情况下,如何才能令其精准回答本领域的问题呢?还有什么好办法吗?
from hcgf.
这里有两个问题:
- 一些简单问题微调后是能生效的,2-3轮就可以了。数据可以一定重复。不过感觉任务不能太复杂,主要是这个模型本身太小了。
- 要想精准有点困难,毕竟是生成式模型。有两种思路可以稍微缓解这个不确定生成的问题:第一种,先做文档召回,然后召回结果作为上下文,强制模型根据上下文回答;第二种,调整参数,调用
stream_chat
方法,设置tempreture和top_p,两个不要同时用。越大越随机。
from hcgf.
Related Issues (20)
- test number must less than total number HOT 3
- 调了个寂寞 HOT 1
- UnicodeDecodeError: 'gbk' codec can't decode byte 0x80 in position 68: illegal multibyte sequence HOT 2
- 如设置微调模型保存步数? HOT 1
- RuntimeError: Internal: [MASK] is already defined. HOT 13
- 如果数据集中含有换行符的话会报错 JSONDecodeError HOT 1
- 多卡场景下微调完进行推理的时候报错 HOT 1
- 有什么办法解决生成结果有大段重复的问题吗? HOT 3
- 换成大的数据集之后显存不够 HOT 2
- 可否增加聊天web界面呢 HOT 1
- 新版的hcgf输出最多只有10个字符 HOT 1
- 如何实现多轮对话微调训练? HOT 1
- 可以加上多卡的微调和推理功能吗? HOT 1
- chat添加stop参数后出现 slice() cannot be applied to a 0-dim tensor. HOT 2
- 催更催更 HOT 2
- 【Notice】v0.2.0发布
- 无法直接eval模型 HOT 1
- 无法理解的推理逻辑 HOT 3
- 如何用多卡微调呢? HOT 5
- 微调和推理不能兼得 HOT 18
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from hcgf.