麦地

Dear author, I appreciate your work in TransVG++ and I tried using CLIP-ViT to replace the original ResNet myself, but the results were not very good. So sincerely want to...

Using Groma to generate region descriptions

1

Thanks for your great work! I also want to learn more details about using groma.eval.run_groma to generate region description. As you mentioned in #20 ,For user prompt, you can follow...

invalid choice: 'snli_ve'

1

作者您好，我在用OFA的VE部分微调时，显示无法使用`snli_ve`，报错信息如下： > error: argument --task: invalid choice: 'snli_ve' (choose from 'hubert_pretraining', 'translation', 'translation_lev', 'online_backtranslation', 'speech_to_text', 'text_to_speech', 'cross_lingual_lm', 'frm_text_to_speech', 'translation_from_pretrained_bart', 'language_modeling', 'masked_lm', 'denoising', 'multilingual_denoising', 'simul_speech_to_text', 'simul_text_to_text', 'multilingual_masked_lm', 'audio_pretraining', 'audio_finetuning', 'sentence_ranking', 'translation_multi_simple_epoch',...