llava-rlhf LLaVA-RLHF issues

llava-rlhf / LLaVA-RLHF

Aligning LMMs with Factually Augmented RLHF

https://llava-rlhf.github.io/

GNU General Public License v3.0

323 stars 25 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Question about prompt (especially about reward model prompt)

#37 tunantu opened 4 weeks ago
0
Enquiry about the fintuned LLaVA-Fact-RM-13b-v1.5-336-lora-padding

#36 tunantu closed 1 month ago
0
Quetion about reward model's score

#35 DripNowhy closed 2 months ago
4
llava_ppo50k-aokvqa12k-vqa10k.json.数据怎么制作的呢？

#34 Spring24ch closed 2 months ago
1
Question about the optimization time

#33 JulioZhao97 closed 3 months ago
1
Question About the reward model

#32 tyxiong23 closed 4 months ago
2
Inquiry About Padding Strategies in LLaVA-RLHF Training

#31 zhyang2226 closed 4 months ago
1
Image Data for RM

#30 ChencongZJU closed 7 months ago
1
Question about padding side at RL model initialization.

#29 L4zyy closed 7 months ago
1
how to use the reward model isolatedly?

#28 jxgu1016 closed 4 months ago
1
reward base model missing

#27 Ritz111 closed 2 months ago
5
Model testing

#26 ernestoBocini closed 6 months ago
1
NotImplementedError in rl_trainer.py

#25 janak11111 closed 7 months ago
1
The accuracy of reward model seem to be low

#24 Wizardcoast closed 8 months ago
1
About 'hallucination' in preference dataset

#23 davidluciolu closed 8 months ago
1
复现RL训练时报错

#22 Mr-Loevan closed 4 months ago
12
how can I find the eval_image files while evaluating the llava bench?

#21 Amanda2024 closed 9 months ago
1
The performance of the released ckpt is much lower than the scores reported in the paper

#20 Weiyun1025 closed 10 months ago
11
evaluation images missing?

#19 findalexli closed 1 year ago
1
Training on RTX 4090

#18 luohaowen2003 closed 1 year ago
2
Question about insrtuction data

#17 zhang-jr closed 1 year ago
1
Question with regarding to training the reward model

#16 TianjinTeda closed 1 year ago
6
Question

#15 Fake10086 closed 1 year ago
6
Will the RM be released?

#14 findalexli closed 1 year ago
1
Detailed Results of models on MMHal-Bench

#13 vateye closed 1 year ago
1
RuntimeError: mat1 and mat2 must have the same dtype

#12 HarrySSH closed 1 year ago
4
RuntimeError: The size of tensor a (577) must match the size of tensor b (257) at non-singleton dimension 1

#11 HarrySSH closed 1 year ago
13
where is LLaVA-Fact-RM-13b-v1.5-336-lora-padding/checkpoint-200?

#10 HarrySSH closed 1 year ago
2
Merge the models

#9 ThierryDeruyttere closed 1 year ago
1
Cannot reproduce results

#8 Haoye17 closed 1 year ago
13
Can you use this with 4bit?

#7 ThierryDeruyttere closed 1 year ago
0
When will the training codes be released?

#6 feymanpriv closed 1 year ago
1
Any simple script to run this model using llava? (help)

#5 barshag closed 1 year ago
1
Images for the SFT dataset

#4 yuvalkirstain closed 1 year ago
2
error about call model

#3 LiqiangJing closed 1 year ago
7
how to use the model for testing

#2 LiqiangJing closed 1 year ago
1
Great work! Can I know if there is any implementation or script to call this model? Thanks.

#1 WilTay1 closed 1 year ago
2