omegaconf error when loading RoBERTa

facebookresearch / fairseq

Facebook AI Research Sequence-to-Sequence Toolkit written in Python.

MIT License

30.56k stars 6.41k forks source link

🐛 Bug

Omegaconf encounters an error when loading RoBERTa.

To Reproduce

remove argument strict in function call to circumvent #3723

composed_cfg = compose("config", overrides=overrides) #, strict=False)

rum cmd

import torch roberta = torch.hub.load('pytorch/fairseq', 'roberta.large')

see error

Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/felix/.local/lib/python3.8/site-packages/torch/hub.py", line 364, in load model = _load_local(repo_or_dir, model, *args, **kwargs) File "/home/felix/.local/lib/python3.8/site-packages/torch/hub.py", line 393, in _load_local model = entry(*args, **kwargs) File "/home/felix/.cache/torch/hub/pytorch_fairseq_master/fairseq/models/roberta/model.py", line 277, in from_pretrained x = hub_utils.from_pretrained( File "/home/felix/.cache/torch/hub/pytorch_fairseq_master/fairseq/hub_utils.py", line 73, in from_pretrained models, args, task = checkpoint_utils.load_model_ensemble_and_task( File "/home/felix/.cache/torch/hub/pytorch_fairseq_master/fairseq/checkpoint_utils.py", line 421, in load_model_ensemble_and_task state = load_checkpoint_to_cpu(filename, arg_overrides) File "/home/felix/.cache/torch/hub/pytorch_fairseq_master/fairseq/checkpoint_utils.py", line 339, in load_checkpoint_to_cpu state = _upgrade_state_dict(state) File "/home/felix/.cache/torch/hub/pytorch_fairseq_master/fairseq/checkpoint_utils.py", line 643, in _upgrade_state_dict state["cfg"] = convert_namespace_to_omegaconf(state["args"]) File "/home/felix/.cache/torch/hub/pytorch_fairseq_master/fairseq/dataclass/utils.py", line 395, in convert_namespace_to_omegaconf composed_cfg[k] = None File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/dictconfig.py", line 310, in __setitem__ self._format_and_raise( File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/base.py", line 190, in _format_and_raise format_and_raise( File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/_utils.py", line 741, in format_and_raise _raise(ex, cause) File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/_utils.py", line 719, in _raise raise ex.with_traceback(sys.exc_info()[2]) # set end OC_CAUSE=1 for full backtrace File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/dictconfig.py", line 308, in __setitem__ self.__set_impl(key=key, value=value) File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/dictconfig.py", line 318, in __set_impl self._set_item_impl(key, value) File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/basecontainer.py", line 497, in _set_item_impl self._validate_set(key, value) File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/dictconfig.py", line 176, in _validate_set self._validate_non_optional(key, value) File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/dictconfig.py", line 246, in _validate_non_optional child = self._get_node(key) File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/dictconfig.py", line 465, in _get_node self._validate_get(key) File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/dictconfig.py", line 166, in _validate_get self._format_and_raise( File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/base.py", line 190, in _format_and_raise format_and_raise( File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/_utils.py", line 821, in format_and_raise _raise(ex, cause) File "/home/felix/.local/lib/python3.8/site-packages/omegaconf/_utils.py", line 719, in _raise raise ex.with_traceback(sys.exc_info()[2]) # set end OC_CAUSE=1 for full backtrace omegaconf.errors.ConfigKeyError: Key 'simul_type' not in 'FairseqConfig' full_key: simul_type object_type=FairseqConfig

Environment

fairseq Version: master

PyTorch Version: 1.9.0+cpu

OS: Ubuntu

How you installed fairseq: pip

Python version: 3.8.10