dandelin / ViLT

Code for the ICML 2021 (long talk) paper: "ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision"

dandelin/ViLT Issues

AttributeError: 'LightningDistributedDataParallel' object has no attribute '_sync_params'
Updated 4 months ago1
while read file idx 2740206 in conceptual_caption_train_0 -> image file is truncated
Updated 4 months ago1
Question about GCC
Updated 4 months ago
Checkpoint file for VQA might be wrong
Updated 4 months ago
what is the meaning of "split" in /vilt/utils/write_conceptual_caption.py
Updated 5 months ago
python run.py with data_root=content/datasets num_gpus=2 num_nodes=1 task_mlm_itm whole_word_masking=True step100k per_gpu_batchsize=64
Updated 6 months ago6
Question about pad_choice
Updated 6 months ago
Pre-training Time
Closed 6 months ago1
Question about ITM pretraining
Updated 6 months ago
Got better results than in the paper:
Updated 7 months ago
integrate with Lightning ecosystem CI
Updated 7 months ago
COCO split for pre-training
Updated 7 months ago
ITM Objectives task will not be enabled？
Updated 7 months ago2
self.mask_token at 553 line in vision_transformer.py #35 I find the self.mask_token has not been defined.
Updated 7 months ago
`self.mask_token` at 553 line in `vision_transformer.py`
Updated 7 months ago6
Flickr30k Finetune results does not match the provided checkpoint
Updated 7 months ago10
Reproduce Flickr30k Evaluation results - DataSet problem
Closed 7 months ago2
Question about GCC dataset download
Updated 7 months ago1
Do I have to download the pre-training weights of VILT or the pre-training weights of VIT
Updated 7 months ago3
Add ViLT to HuggingFace Transformers
Updated 7 months ago11
How to run programs on multiple machines, such as 4 machines with 8 gpus(4*8=32 in total)?
Closed 8 months ago
Download and process GCC and SBU
Updated 8 months ago10
ViltTransformerSS usage clarification
Updated 8 months ago9
Incorrect dataset link.
Closed 8 months ago1
About finetune on f30k.
Updated 8 months ago2
missing values in vqa_dict.json
Closed 8 months ago1
Visual Grounding
Updated 8 months ago1
Time Calculation
Updated 8 months ago6
small difference between paper and code about token type embedding
Closed 8 months ago2
Time cost when fine-tuning on COCO-irtr dataset
Closed 9 months ago2
inference image captioning
Updated 9 months ago2
Possible out-of-memory issue of dataloader
Closed 9 months ago4
Is it necessary to manually add the sampler in the dataloader?
Closed 9 months ago2
License
Closed 9 months ago2
itm and mlm training accuracy
Updated 9 months ago1
FLOPS calculation
Updated 9 months ago1
Visualizations for VQA dataset
Closed 9 months ago
Finetune on F30K IR/TR，why use itm
Closed 9 months ago
Masked Patch Prediction
Closed 10 months ago1
AttributeError: module 'vilt' has no attribute 'modules'
Closed 10 months ago4
got error when I turn on mpp (Masked Patch Prediction)
Closed a year ago8
Nearly 3 points below the provided VQAv2 fine-tuning result
Closed a year ago2
The time consumption for pretraining
Closed a year ago1
Regarding pretraining time
Closed a year ago2
Question about image transformation: short edge is still 384 for the fine-tuning task?
Closed a year ago2
Debug model
Updated a year ago1
I reproduced the code of pytorch version, but get different result
Closed a year ago4
A very large batchsize requires 64 GPUs
Updated a year ago12
Unable to reproduce the 100k results
Closed a year ago8
Finetuning VQA from checkpoint has unmatched keys
Closed a year ago1

ezoic increase your site revenue