Comments (8)
话说照着你这样修改的话,原本的l2_loss就没有了吗? 最终的loss = loss + orthogonal_loss * lamda_1吗?
你自己加上就好了,又不冲突…… 只是我懒得写了
是直接用matched_modules进行计算吗?
l2_loss = 0. for name, param in matched_modules: l2_loss += torch.norm(param, p=2)
完全不对吧,
# l2-normalization for loranew_A/B
l2_loss = 0.
for name, param in self.model.named_parameters():
if "loranew_" in name:
l2_loss += torch.norm(param, p=2)
原本代码里面写的是新的loranew,那么简化代码以后目标是
# l2-normalization for loranew_A/B
l2_loss = 0.
for name, param in self.model.named_parameters():
if "lora_" in name:
l2_loss += torch.norm(param, p=2)
lora_ 就是原本的lora_new啊,l2正则肯定是对现在task的参数进行的啊
from o-lora.
可以的!👍
from o-lora.
哦对,有个问题我不懂就问了:)懒得再翻您改的PEFT代码了(不是
既然说是当前LoRA在之前LoRA的正交方向上更新的;那么当前的LoRA大概率是merge之前LoRA,以此为基础继续训练的吧?我没理解错吧
from o-lora.
哦对,有个问题我不懂就问了:)懒得再翻您改的PEFT代码了(不是 既然说是当前LoRA在之前LoRA的正交方向上更新的;那么当前的LoRA大概率是merge之前LoRA,以此为基础继续训练的吧?我没理解错吧
训练完会进行merge
#5 (comment)
from o-lora.
哦对,有个问题我不懂就问了:)懒得再翻您改的PEFT代码了(不是 既然说是当前LoRA在之前LoRA的正交方向上更新的;那么当前的LoRA大概率是merge之前LoRA,以此为基础继续训练的吧?我没理解错吧
训练完会进行merge #5 (comment)
我的疑惑在新的task的lora初始化上面,既然说是最后合并的,我姑且认为是随机初始化的~毕竟代码上loss要保证两个lora_a是正交的。
from o-lora.
话说照着你这样修改的话,原本的l2_loss就没有了吗?
最终的loss = loss + orthogonal_loss * lamda_1吗?
from o-lora.
话说照着你这样修改的话,原本的l2_loss就没有了吗? 最终的loss = loss + orthogonal_loss * lamda_1吗?
你自己加上就好了,又不冲突…… 只是我懒得写了
from o-lora.
话说照着你这样修改的话,原本的l2_loss就没有了吗? 最终的loss = loss + orthogonal_loss * lamda_1吗?
你自己加上就好了,又不冲突…… 只是我懒得写了
是直接用matched_modules进行计算吗?
l2_loss = 0.
for name, param in matched_modules:
l2_loss += torch.norm(param, p=2)
from o-lora.
Related Issues (20)
- ModuleNotFoundError: No module named 'datasets'
- 关于长序列任务上的讨论 HOT 2
- 关于数据集 HOT 6
- 关于数据集加载的报错 HOT 1
- Loss在yahoo数据集上骤降为0
- 代码里的UIE是什么的缩写? HOT 1
- Missing the parameter `r_sum` in class Linear8bitLt in lora.py
- 作者你好~ 请问一下为什么lora矩阵的形状中有0呢 HOT 5
- 代码运行时在loss那里报错
- Thanks! Great Projects! Could you please provide the example to finetune llama-3-instruct-8B with O-LoRA?
- 有order4,5,6的脚本吗? HOT 3
- 运行报错
- llama2的结果比论文中的llama1的结果低
- llama2 结果复现 HOT 4
- 关于standard benchmark
- 如何解决task-id问题
- 关于loss_mask
- 关于SeqLoRA和IncLoRA HOT 2
- 关于lora_b矩阵的更新问题 HOT 4
- 关于MTL
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 o-lora.