wenliangdai / Multimodal-End2end-Sparse

The code repository for NAACL 2021 paper "Multimodal End-to-End Sparse Model for Emotion Recognition".
95 stars 16 forks source link

training error #8

Open HOUSHIJING opened 3 years ago

HOUSHIJING commented 3 years ago

Hello, I have some question about this code,after we tokenizer the text transcription into tensor, when we use the model 'mme2e', it occurred the error below ,thanks for you help. Traceback (most recent call last): File "/home5/HouSJ/code/Multimodal-End2end-Sparse-main/main.py", line 155, in trainer.train() File "/home5/HouSJ/code/Multimodal-End2end-Sparse-main/src/trainers/emotiontrainer.py", line 46, in train train_stats, train_thresholds = self.train_one_epoch() File "/home5/HouSJ/code/Multimodal-End2end-Sparse-main/src/trainers/emotiontrainer.py", line 169, in train_one_epoch logits = self.model(imgs, imgLens, specgrams, specgramLens, text) File "/home5/HouSJ/anaconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl return forward_call(*input, *kwargs) File "/home5/HouSJ/code/Multimodal-End2end-Sparse-main/src/models/e2e.py", line 90, in forward text_cls = self.T(text, get_cls=True) File "/home5/HouSJ/anaconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl return forward_call(input, kwargs) File "/home5/HouSJ/code/Multimodal-End2end-Sparse-main/src/models/e2e_t.py", line 16, in forward logits, hidden_states = self.albert(text, output_hidden_states=True) TypeError: AlbertModel object argument after ** must be a mapping, not Tensor Train: 0%| | 0/646 [00:00<?, ?it/s]