issues
search
GanjinZero
/
RRHF
[NIPS2023] RRHF & Wombat
781
stars
49
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
关于ppl的方差
#54
skepsun
opened
5 months ago
1
请问基于Vicuna测试集的比较是如何进行比较的?
#53
IT-five
closed
6 months ago
0
如果我想将模型更改为baichuan2-7b-chat,需要做哪些方面的变动?
#52
IT-five
closed
5 months ago
3
算loss的时候求均值的时候是不是可以优化
#51
shyoulala
opened
6 months ago
6
数据构造问题
#50
lylcst
opened
6 months ago
1
Label Shifts
#49
yafuly
closed
7 months ago
0
bug 计算sft损失的时候
#48
shyoulala
opened
7 months ago
2
关于alpaca-7B和LLaMA-7B
#47
NEUBuffett
opened
8 months ago
3
有关IMDB数据集的问题
#46
stevie1023
opened
8 months ago
2
dummy_target的请教
#45
xunfengzhangyang
closed
8 months ago
5
dummy_target的请教
#44
xunfengzhangyang
closed
8 months ago
1
加载模型的问题
#43
LiangZhuuu
opened
10 months ago
11
损失函数
#42
xiayouhong
closed
10 months ago
3
训练过程OOM的问题
#41
Guochry
opened
10 months ago
1
Wombat与RRHF
#40
Guochry
opened
11 months ago
4
The generation config for evaluation
#39
stevie1023
closed
11 months ago
6
labels != -100的作用是什么
#38
LSX-Sneakerprogrammer
opened
11 months ago
3
RRHFTrainer.gather_logits_labels label in-place operation error
#37
asadfgglie
opened
11 months ago
8
The size of tensor a (8) must match the size of tensor b (2) at non-singleton dimension 1
#36
ZJXNEFU
closed
11 months ago
11
在单卡A100上训练出现torch.distributed.elastic.multiprocessing.api.SignalException: Process 2920830 got signal: 1
#35
Zhang-Each
closed
11 months ago
2
the evaluation script with average reward score (Dahoas/gptj-rm-static)
#34
stevie1023
closed
12 months ago
5
NameError: name 'save_fsdp_model' is not defined
#33
ZJXNEFU
closed
11 months ago
4
评估方法与位置有很大关系
#32
xiaoyuan1996
opened
1 year ago
2
期待LoRA或ptuning
#31
Noyce765103
opened
1 year ago
1
CUDA out of memory when trainer.model.state_dict()
#30
Akiraxty
closed
1 year ago
2
The calculation about rrhf loss in the code seems to be completely wrong
#29
yyhycx
closed
1 year ago
1
How to use it. Is there some code examples?
#28
Mr-IT007
opened
1 year ago
1
一些训练细节
#27
xiaoyuan1996
closed
1 year ago
2
PPL
#26
SuMeng123
closed
1 year ago
7
对于重复score答案样本的处理疑问
#25
yanhan19940405
opened
1 year ago
7
fix batch size bug
#24
echoht
opened
1 year ago
3
loss的代码关于batch size的处理有bug。
#23
echoht
opened
1 year ago
4
training with my own gpt2
#22
dyyzhmm
opened
1 year ago
1
wombat-7B的输出异常
#21
lx86110
closed
1 year ago
15
can RRHF train on v100 32G?
#20
akk-123
closed
1 year ago
24
PPO implementation
#19
yuzc19
opened
1 year ago
2
Wombat-7B,Wombat-7B-gpt4 and ChatGPT Results on Comparison based on Vicuna test set, evaluation by gpt-4.
#18
onlyfish79
opened
1 year ago
4
有关训练模型细节
#17
yanhan19940405
closed
1 year ago
12
Results on Comparison based on Vicuna test set
#16
LeeShiyang
opened
1 year ago
1
Why use HingeLoss instead of BPRLoss ?
#15
KID-22
opened
1 year ago
1
single_sentence_inference output is empty
#14
better629
closed
1 year ago
10
This loss seems to consume a lot of memory.
#13
piekey1994
opened
1 year ago
4
Error when try to inference
#12
oasis-0927
closed
1 year ago
5
We are trying to evaluate Wombat on Vicuna test set, but we do not have GPT4 API.
#11
GanjinZero
closed
1 year ago
0
docs: fix some typo
#10
zmsn-2077
closed
1 year ago
1
Update README.md
#9
eltociear
closed
1 year ago
0
RRHF only works on llama model.
#8
Taekyoon
closed
1 year ago
16
update & rearrange
#7
Yuanhy1997
closed
1 year ago
0
Add comparison workflow
#6
Yuanhy1997
closed
1 year ago
0
update case
#5
Chuanqi1992
closed
1 year ago
0
Next