search

SapienzaNLP / xl-amr

XL-AMR is a sequence-to-graph cross-lingual AMR parser that exploits transfer learning (EMNLP2020).
Other
16 stars 5 forks source link

TypeError when running prediction (problem with Predictor class) #5

Open meganndare opened 1 year ago

meganndare commented 1 year ago

Hi, I am trying to use the best checkpoint for ZH provided by you to predict some AMRs given chinese sentences. However, when I run the prediction script, I'm getting a type error (see below). Additionally the same exact error occurs when I tried to run project_test.sh. Any idea how to get around this? 

:~/code/xl-amr$ ./scripts/predict_multilingual.sh en models/xl-amr_lang_specific_zh_trans/
Predicting...Sun Apr 30 10:30:33 CEST 2023
/home/CE/mdare/anaconda3/envs/xlamr/lib/python3.6/site-packages/tqdm/__init__.py
Traceback (most recent call last):
  File "/home/CE/mdare/anaconda3/envs/xlamr/lib/python3.6/runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "/home/CE/mdare/anaconda3/envs/xlamr/lib/python3.6/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/commands/predict.py", line 52, in <module>
    from xlamr_stog.predictors.predictor import Predictor, JsonDict
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/predictors/__init__.py", line 1, in <module>
    from xlamr_stog.predictors.predictor import Predictor
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/predictors/predictor.py", line 7, in <module>
    from xlamr_stog.data import DatasetReader, Instance
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/data/__init__.py", line 1, in <module>
    from xlamr_stog.data.dataset_readers.dataset_reader import DatasetReader
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/data/dataset_readers/__init__.py", line 10, in <module>
    from xlamr_stog.data.dataset_readers.dataset_reader import DatasetReader
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/data/dataset_readers/dataset_reader.py", line 4, in <module>
    from xlamr_stog.data.instance import Instance
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/data/instance.py", line 3, in <module>
    from xlamr_stog.data.fields.field import DataArray, Field
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/data/fields/__init__.py", line 7, in <module>
    from xlamr_stog.data.fields.array_field import ArrayField
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/data/fields/array_field.py", line 10, in <module>
    class ArrayField(Field[numpy.ndarray]):
  File "/local/home/CE/mdare/code/xl-amr/xlamr_stog/data/fields/array_field.py", line 42, in ArrayField
    @overrides
  File "/home/CE/mdare/anaconda3/envs/xlamr/lib/python3.6/site-packages/overrides/overrides.py", line 83, in overrides
    return _overrides(method, check_signature, check_at_runtime)
  File "/home/CE/mdare/anaconda3/envs/xlamr/lib/python3.6/site-packages/overrides/overrides.py", line 170, in _overrides
    _validate_method(method, super_class, check_signature)
  File "/home/CE/mdare/anaconda3/envs/xlamr/lib/python3.6/site-packages/overrides/overrides.py", line 189, in _validate_method
    ensure_signature_is_compatible(super_method, method, is_static)
  File "/home/CE/mdare/anaconda3/envs/xlamr/lib/python3.6/site-packages/overrides/signature.py", line 102, in ensure_signature_is_compatible
    ensure_return_type_compatibility(super_type_hints, sub_type_hints, method_name)
  File "/home/CE/mdare/anaconda3/envs/xlamr/lib/python3.6/site-packages/overrides/signature.py", line 303, in ensure_return_type_compatibility
    f"{method_name}: return type `{sub_return}` is not a `{super_return}`."
TypeError: ArrayField.empty_field: return type `None` is not a `xlamr_stog.data.fields.field.Field`.