RLHF-V/RLHF-V issues and pull requests

#29 - MiniCPM-V Model Support with RLHF-V

Issue - State: open - Opened by DarioPTWR 2 months ago

#28 - Using a different model

Issue - State: closed - Opened by DanyalMaq 3 months ago - 1 comment

#27 - Any plans to release the MHumanEval benchmark?

Issue - State: closed - Opened by Molly-3000 5 months ago - 1 comment

#26 - Curves of "rewards_train/chosen"

Issue - State: closed - Opened by davidluciolu 5 months ago - 2 comments

#25 - waiting for the training code files

Issue - State: closed - Opened by nanamma 5 months ago - 2 comments

#24 - 您好，数据集好像都是英文的，是否有中文的数据集？

Issue - State: closed - Opened by Shubao-Zhao 6 months ago - 1 comment

#23 - KeyError: <InterpolationMode.BICUBIC: 'bicubic'>

Issue - State: closed - Opened by nicochen1118 6 months ago - 2 comments

#22 - 程序会一直停在waiting resource finalization

Issue - State: closed - Opened by shipengai 6 months ago - 1 comment

#21 - review_file_name=hall_obj_halbench_answer_-1.json

Issue - State: closed - Opened by shipengai 6 months ago

#20 - 请问一下，使用了什么llm?

Issue - State: closed - Opened by dreaming12580 6 months ago - 2 comments

#19 - 请问在cogvlm2模型上如何使用RLHF-V数据集进行DPO训练

Issue - State: closed - Opened by kaka-Cao 6 months ago - 1 comment

#18 - γ参数设置问题，原文建议是5，某一版提交后变成1.1了，这是因为有什么区别吗

Issue - State: closed - Opened by Stubborn-one 8 months ago - 1 comment

#17 - CUDA ordinal error within run_RLHFV.sh

Issue - State: closed - Opened by rezacopol 9 months ago - 1 comment

#16 - Request inference script for logps data

Issue - State: closed - Opened by tibetgao 9 months ago - 2 comments

#15 - Are parameters of vision tower freezed at SFT and DPO?

Issue - State: closed - Opened by davidluciolu 9 months ago - 2 comments

#14 - Does the data can be used on non qwen and none llava architecture model?

Issue - State: closed - Opened by OpenJarvisAI 9 months ago - 1 comment

#13 - Question about MHumanEval

Issue - State: closed - Opened by franciscoliu 10 months ago - 1 comment

#12 - vision tower

Issue - State: closed - Opened by Coach257 11 months ago - 1 comment

#11 - ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -9) local_rank: 0 (pid: 484375) of binary

Issue - State: closed - Opened by Cooperx521 12 months ago - 4 comments

#10 - About training code and some question about the DDPO

Issue - State: closed - Opened by Yxxxb almost 1 year ago - 2 comments

#9 - About model size

Issue - State: closed - Opened by yfzhang114 about 1 year ago - 1 comment

#8 - KeyError: 'RLHF-V-Dataset'

Issue - State: closed - Opened by DuoLong about 1 year ago - 10 comments

#7 - ImportError: cannot import name 'flash_attn_unpadded_qkvpacked_func' from 'flash_attn.flash_attn_interface'

Issue - State: closed - Opened by linhaojia13 about 1 year ago - 1 comment

#6 - Missing Data (tsv files)

Issue - State: closed - Opened by jacklishufan about 1 year ago - 3 comments

#5 - the dataset link returns `404`

Issue - State: closed - Opened by linhaojia13 about 1 year ago - 2 comments

#4 - corrections & hallucinations

Issue - State: closed - Opened by eyuansu62 about 1 year ago - 3 comments

#3 - TypeError: forward() got an unexpected keyword argument 'position_ids'

Issue - State: closed - Opened by findalexli about 1 year ago - 1 comment

#2 - eval_muffin_inference_logp.sh this script doesnt seem to be present

Issue - State: closed - Opened by findalexli about 1 year ago - 2 comments

#1 - 关于数据集的一点问题。

Issue - State: closed - Opened by shipengai about 1 year ago - 5 comments

Ecosyste.ms: Issues

GitHub / RLHF-V/RLHF-V issues and pull requests