THUDM/ImageReward

Issues

• Unable to pin memory: operation not supported

  #111 opened a month ago by lestersssss
  0

• 'gbk' codec can't decode byte 0xa2 in position 116

  #110 opened a month ago by aswordok
  0

• Running with no prompt

  #109 opened a month ago by benoriol
  0

• the vae module can not generate normal images

  #108 opened 2 months ago by deropty
  0

• 我是一名深度学习新手，我尝试自己做了一个训练集并运行您的训练脚本，遇到了一些问题

  #107 opened 3 months ago by Zhang-ZheWei
  0

• not find problem category label file corresponding to the image

  #104 opened 5 months ago by ctxya1207
  1

• Human Annotation Analysis

  #99 opened 6 months ago by 292725651
  5

• where i can find the train.json

  #103 opened 5 months ago by ctxya1207
  1

• range of perference scores

  #102 opened 5 months ago by ctxya1207
  1

• 数据集损坏问题

  #101 opened 5 months ago by psycho-ygq
  2

• self.noise_scheduler.reset_timesteps()没有reset_timesteps这个方法

  #105 opened 4 months ago by PINKRAYL
  1

• Training loss？

  #106 opened 4 months ago by wzczc
  0

• fine-tuning: NotImplementedError

  #100 opened 6 months ago by fenglincong
  1

• ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. deepseek-vl 1.0.0 requires timm>=0.9.16, but you have timm 0.6.13 which is incompatible.

  #98 opened 6 months ago by prantikseal
  2

• Can I use ReFL method to train another dataset for a new reward model?

  #96 opened 7 months ago by LOGIC-10
  1

• dependency_links are ignored; manual installation of CLIP is requried

  #80 opened a year ago by iuliaturc
  2

• The formation of the reward-loss function

  #83 opened 7 months ago by he-nantian
  2

• ReFL Training Performance

  #95 opened 7 months ago by LemonTwoL
  2

• Finetuned model weights

  #90 opened 7 months ago by rohitrango
  4

• Training step

  #93 opened 8 months ago by ucasyjz
  2

• (Webui) Image reward info text does not show like in demo video

  #91 opened 8 months ago by kappa54m
  2

• requires: timm==0.6.13, this causes comfyui to install timm back and forth due to different version requirements

  #86 opened 8 months ago by bluevisor
  1

• w/o prertain loss in the code

  #88 opened 8 months ago by dongbo811
  1

• any progress on updating imagereward model?

  #87 opened 9 months ago by wendellswa06
  1

• [dataset] Collecting rankings from different annotators

  #85 opened 9 months ago by dain5832
  1

• The fine-tuned SD released?

  #84 opened 10 months ago by XiaoBuL
  1

• 训练问题

  #82 opened a year ago by zshuo7615
  1

• 时间t的选择

  #81 opened a year ago by X-fxx
  1

• 一些问题

  #79 opened a year ago by zhou431496
  5

• Reproducing all numbers in Tab 3

  #78 opened a year ago by vishaal27
  1

• How to train stabilityai/stable-diffusion-xl-base-1.0 using ImageReward model

  #77 opened a year ago by emelpolaris
  5

• 推理阶段图像尺寸

  #76 opened a year ago by jianghonglin30
  1

• 缺陷分类对应的图片

  #75 opened a year ago by wzczc
  1

• 有没有可能把ImageReward用在生图过程中指导降噪？

  #74 opened a year ago by petercham
  1

• About the score range

  #73 opened a year ago by geknow
  2

• Training Novel Concepts

  #71 opened a year ago by adamgeddon1686
  1

• About the reward score

  #72 opened a year ago by KN1GHT9
  1

• Can we train for sdxl?

  #59 opened 2 years ago by sneccc
  3

• About the CUDA and torch version.

  #70 opened a year ago by Yxxxb
  1

• About loading CLIPScore and BLIPScore

  #69 opened a year ago by passing2961
  1

• How to understand the accuracy calculation method

  #68 opened a year ago by learn01one
  1

• cuda

  #67 opened a year ago by ckhui
  1

• Strange training dynamics for ImageReward model.

  #64 opened a year ago by bhattg
  3

• About Prompt Selection

  #66 opened a year ago by QMME
  1

• a nccl problem

  #65 opened a year ago by learn01one
  1

• LoRA models for ImageReward?

  #63 opened a year ago by CreativeBuilds
  1

• test custom model acc : 0.0%

  #62 opened a year ago by Shiyao-Huang
  2

• Yet another question on standard deviation

  #60 opened 2 years ago by tusharbhutt
  2

• Full refl_data.json for REFL

  #61 opened 2 years ago by fuqianya
  1

• ReFL implement details

  #57 opened 2 years ago by hkunzhe
  2