Dutch model does not seem to work out of the box

[tcbrouwer]

I tried to run the dutch model but I can not seem to get it to work. What am I doing wrong?

Here is a minimal example with the resulting error message.

# The minimal example of diaparser is:

# >>> from diaparser.parsers import Parser
# >>> parser = Parser.load('en_ewt-electra')
# >>> dataset = parser.predict([['She', 'enjoys', 'playing', 'tennis', '.']], prob=True)

# Let's try this with dutch

from diaparser.parsers import Parser
parser = Parser.load('nl_alpino_lassysmall.wietsedv')
dataset = parser.predict([['Zij', 'houdt', 'van', 'tennissen', '.']], prob=True)

# Note that using 'nl_alpino_lassysmall-wietsedv' as model name seems to just load the english model, which yields bad results for dutch.

Running the code yields the following error:

Some weights of BertModel were not initialized from the model checkpoint at wietsedv/bert-base-dutch-cased and are newly initialized: ['bert.pooler.dense.bias', 'bert.pooler.dense.weight']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.

---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
Cell In[15], line 10
      1 # The minimal example of diaparser is:
      2 
      3 # >>> from diaparser.parsers import Parser
   (...)
      6 
      7 # Let's try this with dutch
      9 from diaparser.parsers import Parser
---> 10 parser = Parser.load('nl_alpino_lassysmall.wietsedv')
     11 dataset = parser.predict([['Zij', 'houdt', 'van', 'tennissen', '.']], prob=True)
     13 # Note the using 'nl_alpino_lassysmall-wietsedv' as model name seems to just load the english model, which yields bad results for dutch.

File ~/Projects/UDParserEvaluation/venv/lib/python3.10/site-packages/diaparser/parsers/parser.py:263, in Parser.load(cls, name_or_path, lang, cache_dir, **kwargs)
    261 model = cls.MODEL(**args)
    262 model.load_pretrained(state['pretrained'])
--> 263 model.load_state_dict(state['state_dict'], False)
    264 model.to(args.device)
    265 transform = state['transform']

File ~/Projects/UDParserEvaluation/venv/lib/python3.10/site-packages/torch/nn/modules/module.py:2152, in Module.load_state_dict(self, state_dict, strict, assign)
   2147         error_msgs.insert(
   2148             0, 'Missing key(s) in state_dict: {}. '.format(
   2149                 ', '.join(f'"{k}"' for k in missing_keys)))
   2151 if len(error_msgs) > 0:
-> 2152     raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
   2153                        self.__class__.__name__, "\n\t".join(error_msgs)))
   2154 return _IncompatibleKeys(missing_keys, unexpected_keys)

RuntimeError: Error(s) in loading state_dict for BiaffineDependencyModel:
	size mismatch for feat_embed.bert.embeddings.word_embeddings.weight: copying a param with shape torch.Size([30000, 768]) from checkpoint, the shape in current model is torch.Size([30073, 768]).

Is the model not intended to be run this way?