A RuntimeError occurred when use_bert_only==True

[DingQiang2018]

Hi authors,

I ran into a runtime error when I ran bash best_parser_training_script.sh with only --use-xlnet replaced with --use-bert-only. Here is my environment and the traceback of the error.

cython                    0.29.13 
nltk                      3.5 
numpy                     1.17.2   
pytorch-pretrained-bert   0.6.2   
sentencepiece             0.2.0     
torch                     2.4.1     
transformers              2.8.0    

Training...
This is  joint_xlnet_clean_large_3_layers_no_resdrop_lambda
Traceback (most recent call last):
  File "src_joint/main.py", line 788, in <module>
    main()
  File "src_joint/main.py", line 784, in main
    args.callback(args)
  File "src_joint/main.py", line 727, in <lambda>
    subparser.set_defaults(callback=lambda args: run_train(args, hparams))
  File "src_joint/main.py", line 491, in run_train
    _, loss = parser.parse_batch(subbatch_sentences, subbatch_trees)
  File "/data/dingqiang/LAL-Parser/src_joint/KM_parser.py", line 1827, in parse_batch
    = self.parse_from_annotations(fencepost_annotations_start[start:end,:], fencepost_annotations_end[start:end,:], sentences[i], i, gold=golds[i])
  File "/data/dingqiang/LAL-Parser/src_joint/KM_parser.py", line 1900, in parse_from_annotations
    label_scores_chart = self.label_scores_from_annotations(fencepost_annotations_start, fencepost_annotations_end)
  File "/data/dingqiang/LAL-Parser/src_joint/KM_parser.py", line 1885, in label_scores_from_annotations
    label_scores_chart = self.f_label(span_features)
  File "/home/dingqiang/miniconda3/envs/lal_parser/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "/home/dingqiang/miniconda3/envs/lal_parser/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/dingqiang/miniconda3/envs/lal_parser/lib/python3.8/site-packages/torch/nn/modules/container.py", line 219, in forward
    input = module(input)
  File "/home/dingqiang/miniconda3/envs/lal_parser/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "/home/dingqiang/miniconda3/envs/lal_parser/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/dingqiang/miniconda3/envs/lal_parser/lib/python3.8/site-packages/torch/nn/modules/linear.py", line 117, in forward
    return F.linear(input, self.weight, self.bias)
RuntimeError: mat1 and mat2 shapes cannot be multiplied (16x1024 and 14336x250)

Could you check and resolve this bug? I will be very grateful.