AILab-CVC/SEED-X

Issues

• Question about training data mixture. （训练数据混合问题）

  #30 opened a month ago by JLM-Z
  0

• The number of tokens are inconsistent in get_image_embeds

  #29 opened 2 months ago by CharlesGong12
  0

• seed part-1 dataset

  #27 opened 3 months ago by Eureka-Maggie
  0

• Force Polishing Generated Image

  #28 opened 3 months ago by CharlesGong12
  0

• How to generate multiple images at once?

  #26 opened 3 months ago by JasiRose
  0

• What is the batchsize when training de-tokenizer (the first stage)? help me please.

  #25 opened 3 months ago by huangjch526
  0

• Weird artifacts for editing model

  #21 opened 5 months ago by Ctrl-C-V4ever
  1

• About SEED-X Visual De-tokenization pretraining code

  #7 opened 5 months ago by qxhhou
  1

• Difference in detokenizer when the input “image_pil“ and “image_embeds“

  #14 opened 5 months ago by qxhhou
  1

• release the code of pre-training

  #24 opened 5 months ago by cxjtju
  0

• How LLM model from LLaMA to Baichuan?

  #23 opened 5 months ago by cxjtju
  0

• Data format of image generation

  #22 opened 5 months ago by Sammy20207109
  0

• Wired data operations.

  #15 opened 5 months ago by idealwei
  1

• Details about Data Preprocessing

  #13 opened 7 months ago by BolinLai
  2

• Details about Instruction FT

  #19 opened 6 months ago by Ctrl-C-V4ever
  3

• how do you handle multi image input？

  #10 opened 8 months ago by A-new-b
  3

• PEFT versions?

  #20 opened 6 months ago by naoto0804
  1

• Data structure

  #16 opened 7 months ago by floppycracken
  1

• multi-turn editing

  #12 opened 7 months ago by j-min
  1

• Release detokenizer training code

  #11 opened 7 months ago by hzphzp
  1

• Consider cite the paper.

  #9 opened 8 months ago by yuzhou914
  1

• model quantification

  #8 opened 8 months ago by Zhou2019
  1

• deepspeed bug?

  #17 opened 6 months ago by huangjch526
  1

• Questions about number of tokens in tokenizer~

  #18 opened 6 months ago by thaoshibe
  2

• Error from image editing inference script.

  #6 opened 8 months ago by j-min
  4

• About pretrain code

  #5 opened 8 months ago by HalvesChen
  3

• Does the input Resampler and output Resampler share same weights?

  #4 opened 8 months ago by lucasjinreal
  1

• Will you share the Gradio code?

  #3 opened 8 months ago by kadirnar
  1

• Questions about the model pre-training stage

  #1 opened 8 months ago by JunZhan2000
  1

• Questions about the visual tokenizer and de-tokenizer training stage

  #2 opened 8 months ago by friedrichor
  2