Search results

I can't find where the negative log-likelihood function is used in the code

Status: Open.

#11 In jacobswan1/ViTCAP;

· LT156 opened on May 20, 2023

How can I find the pt model named "Logit_Vilt_captioning_testing_batch-size_512_encoder_vit_base_patch16_384_lr_1e-4_iter_60_vitbfocal20_bert_tokenizer_tags_ENC-DEC_multiplier_0.1_expand_tag-classifier_emb.pt"

Status: Open.

#10 In jacobswan1/ViTCAP;

· Shunli-W opened on Mar 15, 2023

Thank you for your code and paper, I have gained a lot, there is a question how to implement distributed training, can you please share your PyTorch DDP mode, thank you very much.

Status: Open.

#9 In jacobswan1/ViTCAP;

· Markkk111 opened on Feb 22, 2023

Checkpoint model cannot be loaded

Status: Open.

#8 In jacobswan1/ViTCAP;

· Faiail opened on Jan 31, 2023

Freezing CTN and ViT during captioning

Status: Open.

#7 In jacobswan1/ViTCAP;

· thilinen opened on Dec 8, 2022

Use semantic segmentation map as channel as additional input

Status: Open.

#6 In jacobswan1/ViTCAP;

· thilinen opened on Nov 19, 2022

Thanks for your code. Could you elaborate the implementation details of ViLT-CAP you used as one of the baselines?

Status: Open.

#4 In jacobswan1/ViTCAP;

· meiling-fdu opened on Oct 25, 2022

What machine to use

Status: Open.

#2 In jacobswan1/ViTCAP;

· Gary-code opened on Aug 29, 2022