princeton-nlp LESS issues

princeton-nlp / LESS

[ICML 2024] LESS: Selecting Influential Data for Targeted Instruction Tuning

MIT License

282 stars 21 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Fix the get_info/grad path in README.md

#23 kykim0 opened 2 weeks ago
0
At step1, single GPU works while multiple GPUs get stuck.

#22 timturing opened 2 weeks ago
0
Questions about the influence value

#21 simplelifetime opened 1 month ago
1
OOM Errors in Step 3.1

#20 ChangyuChen347 closed 3 weeks ago
3
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:6 and cuda:0!

#19 Haruka1307 opened 1 month ago
3
No optimizer.bin in Step 2

#18 mihara-bot opened 1 month ago
3
Error while deserializing header: InvalidHeaderDeserialization

#17 shangqing-liu closed 1 month ago
0
Error with fast_jl

#16 simplelifetime closed 3 weeks ago
1
a small mistake in the code

#15 ZZZZZccccc123 closed 3 weeks ago
3
some question about calulate socre？

#14 smashfan closed 3 weeks ago
1
Step 2: Building the gradient datastore: FileNotFoundError: [Errno 2] No such file or directory: '../out/llama2-7b-p0.05-lora-seed3/checkpoint-1688/optimizer.bin'

#13 minchaoyue closed 2 months ago
3
Question about accuracy

#12 fyf3 closed 3 weeks ago
1
tiny typo

#11 kgourgou closed 3 months ago
0
FileNotFoundError: optimizer.bin and KeyError: 'base_model.model.model.layers.0.self_attn.q_proj.lora_A.default.weight'

#10 yinochaos closed 3 months ago
2
Question about running stage 3 script

#9 2003pro closed 3 weeks ago
3
Fix missing return value

#8 leopoldwhite closed 3 months ago
0
Using Multiple GPUs

#7 SD325 closed 3 weeks ago
3
data selection results

#6 wang-debug closed 4 months ago
2
Are samples used for warmup training and gradient calculation the same?

#5 ZigeW closed 3 weeks ago
1
step 2 when run "/get_train_lora_grads.sh", load the optimizer.pt error is happend

#4 victorjiax opened 4 months ago
19
A concerned question about adding a new baseline.

#3 FanqingM closed 4 months ago
2
What if we have the optim states of specific tasks at the beginning?

#2 suhmily closed 4 months ago
2
Does it support baichuan or chatglm?

#1 amulil closed 4 months ago
2