vimalabs/VIMA

Issues

• LVLM Implementation

  #58 opened 3 months ago by GasolSun36
  0

• Implementation details of the loss scaling algorithm

  #53 opened 9 months ago by jiajingk
  1

• How to eval the baseline？

  #52 opened 9 months ago by clawnotfound
  1

• Asking for better explanation

  #50 opened a year ago by Zero-coder
  2

• Action dim not match

  #51 opened 10 months ago by Chris-Chow
  2

• How is the current inference frequency? Did someone try it on real world experiment?

  #57 opened 6 months ago by DeeDive
  0

• Franka in Isaac-sim

  #45 opened a year ago by peter05010402
  1

• Invalid GLX version: major 1, minor 2

  #54 opened 8 months ago by KangriX
  1

• Metrics related to evaluating the models.

  #56 opened 8 months ago by KangriX
  0

• [Test] The test result is not consistent with that reported in paper.

  #34 opened a year ago by aopolin-lv
  11

• Exception: Model "t5-base" on the Hub doesn't have a tokenizer

  #42 opened a year ago by 2000222
  4

• Request for Guidance on Implementing imitation_loss

  #39 opened a year ago by nisshimura
  3

• ImportError: cannot import name 'checkpoint' from 'transformers.models.t5.modeling_t5'

  #47 opened a year ago by Dylan-LDR
  15

• Question: how to specific state or restore the state of the environment?

  #40 opened a year ago by LongXinKou
  0

• Problems about using --device=cuda

  #30 opened a year ago by Deaddawn
  4

• The order of object token is always same

  #46 opened a year ago by shure-dev
  0

• RuntimeError: Failed to reset environment after 10 retries

  #41 opened a year ago by zhaoyangninglilith
  3

• The inference process of the VIMA strongly relies on the images in the prompts and objects in the environment

  #44 opened a year ago by shure-dev
  1

• Model "t5-base" on the Hub doesn't have a tokenizer

  #20 opened a year ago by lqiang2003cn
  7

• Some questions about the input observation

  #38 opened a year ago by wxz1996
  1

• Request for Training Code and Clarification on Parallelization & Batch Size

  #36 opened a year ago by nisshimura
  1

• Question of the object encoding and training details

  #35 opened a year ago by JisuHann
  1

• The detail of traning and evaluating process can be shared? How can I get a success rate?

  #15 opened a year ago by shure-dev
  2

• [Evaluation] How to get the result reported in the paper

  #33 opened a year ago by aopolin-lv
  0

• the physical meaning of the parameter 'actions'

  #32 opened a year ago by kaixin-bai
  1

• About Success Rate calculating details

  #31 opened a year ago by Deaddawn
  2

• How can we calculate the loss for vima?

  #28 opened a year ago by Mano176
  1

• How do you tokenize the pose actions?

  #27 opened a year ago by amitkparekh
  2

• Questions about prompt and Mask-RCNN

  #29 opened a year ago by oxFFFF-Q
  1

• Questions about VIMA Data Loading, RCNN Models, and Real-World Applications

  #26 opened a year ago by oxFFFF-Q
  2

• How to construct dataset for training?

  #19 opened a year ago by oydontbecrazy
  5

• Script/command for training?

  #25 opened a year ago by asmith26
  2

• The inputs are single objects but h

  #23 opened a year ago by zhufq00
  1

• As the input are images of single objects, so how does the model know the relative position and distance between objects

  #24 opened a year ago by zhufq00
  1

• An error is reported when running the demo, and other models also fail to load

  #21 opened a year ago by liuqinglong110
  1

• Clarification Required on ckpt_path Parameter for Running Demo

  #18 opened a year ago by oxFFFF-Q
  0

• how to solve this import error?

  #17 opened a year ago by torressliu
  1

• Questions about the training and evaluation pipelines

  #16 opened a year ago by Ji4chenLi
  3

• ModuleNotFoundError: No module named 'vima_bench'

  #14 opened 2 years ago by YUHANG-Ma
  1

• Can you release the Mask-RCNN model used for segmentation?

  #13 opened 2 years ago by amitkparekh
  1

• The choice of the actiion decoder

  #11 opened 2 years ago by SiyuanHuang95
  3

• Training Algorithm

  #12 opened 2 years ago by ciccio42
  5

• Training hyperparameters besides Table 6

  #9 opened 2 years ago by SiyuanHuang95
  9

• could you release the code of real robot experiment?

  #10 opened 2 years ago by Bailey-24
  1

• Any suggestion to reproduce your reseults?

  #8 opened 2 years ago by SiyuanHuang95
  10

• Question about Dataset creation

  #7 opened 2 years ago by zhanghuzhenyu
  1

• t5-large instead of t5-base

  #6 opened 2 years ago by yeknafar
  2

• Problem with rendering

  #5 opened 2 years ago by yeknafar
  1

• problem about "post_init"

  #4 opened 2 years ago by zhanghuzhenyu
  2

• How to render environment like in paper?

  #3 opened 2 years ago by ezhang7423
  1