microsoft LoRA issues - Githubissues

microsoft / LoRA

Code for loralib, an implementation of "LoRA: Low-Rank Adaptation of Large Language Models"

https://arxiv.org/abs/2106.09685

MIT License

10.43k stars 669 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

add local-rank as pytorch requested to avoid errors

#182 nncliff closed 1 month ago
0
How are the pretrained weigts set to be fixed during training?

#181 HaiminZhang opened 2 months ago
1
NLU experiments error

#180 jie040109 closed 2 months ago
1
Questions about replicating NLU experiments

#177 lky-violet opened 3 months ago
0
Questions about running the cola dataset script

#176 dengxingzhi opened 4 months ago
1
Modified.

#175 zhuopinz closed 5 months ago
0
add hhlayer file

#174 ppaimon opened 5 months ago
0
Where is the LoRA matrices saved?

#173 KelsenMa opened 5 months ago
0
Parameter count on GPT-2 medium

#172 Heimine opened 5 months ago
1
question for scale！

#171 Duperr opened 5 months ago
0
[Question about multi-gpu training]

#170 FindTheTruth opened 5 months ago
2
LORA on T5 model

#169 vivektreddy opened 5 months ago
0
lora-dim == lora-r ?

#168 ardand4708 opened 6 months ago
0
After joining Lora, the first few layers show a gradient of 0

#167 hluckye opened 6 months ago
0
Guidance Needed on Continuing Training with a New Dataset via LoRA

#166 TanTaiLi opened 6 months ago
0
Reproduce Lora results is close but not accurate

#165 harsh306 opened 6 months ago
3
Dynamic Lora Selection In Runtime❓

#164 teneous closed 6 months ago
1
clora added homotopy

#163 harsh306 closed 6 months ago
0
_conv_forward() error

#162 likaiucas opened 6 months ago
0
Cannot implement LoRA on a custom model containing transformer encoder from pytorch

#161 wsuSaiman opened 7 months ago
0
How to adjust LoRA into nn.ConvTranspose2d?

#160 vanmeruso opened 7 months ago
2
typo

#159 WrRan opened 7 months ago
0
[Embedding-bugfix]: reset_parameters

#158 WrRan opened 7 months ago
0
Fix Conv1d (Convnd) implementation

#157 sentient-codebot opened 7 months ago
2
Is it necessary to add `model = model.merge_and_unload()` when training a new LoRA adapter?

#156 4daJKong closed 8 months ago
0
Initialization for LoRA weights A and B initialized in the wrong way.

#155 arun477 opened 8 months ago
0
models are the same after loading lora parameters using peft library

#154 ra-MANUJ-an closed 8 months ago
0
how to improve the memory ability of lora fine tuning?

#153 K-Alex13 opened 8 months ago
1
Can not reproduce the result of Roberta-Base

#151 Ther-nullptr opened 9 months ago
2
Layers.py not being executed

#150 Aradhye2002 opened 9 months ago
1
Can't reproduce the results for GLUE and hyperparameter misalignment

#149 nbasyl opened 10 months ago
4
How to compute that GPT-2 M (FTTop2 ) trainable parameters number is 25.19M?

#148 floatingbigcat opened 10 months ago
0
[Minor] Possible typos in weight initialization

#146 awgu opened 10 months ago
0
Question about the test set of the GLUE benchmark

#145 James6Chou opened 11 months ago
1
Question about seed numbers.

#144 SEONHOK opened 11 months ago
0
fixed typo in readme

#143 cwingho opened 11 months ago
0
The description and the behavior don't match

#142 yongchanghao opened 11 months ago
2
LoRA adapter checkpoints not downloadable

#141 kibru9399 opened 11 months ago
9
Bump urllib3 from 1.26.5 to 1.26.18 in /examples/NLU/examples/research_projects/lxmert

#140 dependabot[bot] opened 11 months ago
0
Is the output the entire model？

#139 licy02 opened 11 months ago
3
Not able to reproduce the scores using provided checkpoint on NLG tasks

#138 ylli0218 opened 11 months ago
5
Bump gitpython from 3.0.2 to 3.1.37 in /examples/NLU/examples/research_projects/distillation

#137 dependabot[bot] opened 11 months ago
0
Support multi-lora fine tune in the same GPU

#136 merlintang opened 11 months ago
0
AttributeError: 'lora.Conv2d' object has no attribute 'weight'

#134 bjzhb666 closed 11 months ago
3
improve mark_only_lora_as_trainable

#133 Edenzzzz closed 10 months ago
0
Bump pillow from 8.3.2 to 10.0.1 in /examples/NLU/examples/research_projects/lxmert

#132 dependabot[bot] opened 12 months ago
0
Bump urllib3 from 1.26.5 to 1.26.17 in /examples/NLU/examples/research_projects/lxmert

#131 dependabot[bot] closed 11 months ago
1
Fix minor typos

#130 haipinglu opened 1 year ago
0
batch=1,why adapter latency so much vs. LoRA in paper???

#129 macqueen09 opened 1 year ago
3
T(w) problem

#128 fclearner opened 1 year ago
0