qywu / textgail Goto Github PK

Hello, I am a student in Japan.
Your research is very interesting.
I actually tried to run your prepared code in googlecolab, but
I got the following error
Is there any solution for this?

!python main.py --config /config/config.yaml task.name=CommonGEN

Traceback (most recent call last):
File "main.py", line 9, in
from torchfly.training import TrainerLoop
File "/content/TextGAIL/TorchFly/torchfly/training/init.py", line 2, in
from .flymodel import FlyModel
File "/content/TextGAIL/TorchFly/torchfly/training/flymodel.py", line 9, in
from torchfly.metrics import CategoricalAccuracy, Average, MovingAverage, Speed
File "/content/TextGAIL/TorchFly/torchfly/metrics/init.py", line 6, in
from .moving_average import MovingAverage
File "/content/TextGAIL/TorchFly/torchfly/metrics/moving_average.py", line 12, in
class MovingAverage(Metric):
File "/content/TextGAIL/TorchFly/torchfly/metrics/moving_average.py", line 43, in MovingAverage
def get_metric(self, reset: bool = False):
File "/usr/local/lib/python3.7/dist-packages/overrides/overrides.py", line 88, in overrides
return _overrides(method, check_signature, check_at_runtime)
File "/usr/local/lib/python3.7/dist-packages/overrides/overrides.py", line 114, in _overrides
_validate_method(method, super_class, check_signature)
File "/usr/local/lib/python3.7/dist-packages/overrides/overrides.py", line 135, in _validate_method
ensure_signature_is_compatible(super_method, method, is_static)
File "/usr/local/lib/python3.7/dist-packages/overrides/signature.py", line 93, in ensure_signature_is_compatible
ensure_return_type_compatibility(super_type_hints, sub_type_hints, method_name)
File "/usr/local/lib/python3.7/dist-packages/overrides/signature.py", line 288, in ensure_return_type_compatibility
f"{method_name}: return type {sub_return} is not a {super_return}."
TypeError: MovingAverage.get_metric: return type None is not a typing.Union[float, typing.Tuple[float, ...], typing.Dict[str, float], typing.Dict[str, typing.List[float]]].

when I ran python main.py --config config/config.yaml

Traceback (most recent call last):
  File "main.py", line 36, in <module>
    main()
  File "main.py", line 22, in main
    config = FlyConfig.load()
  File "/home/hx/github项目/TorchFly/torchfly/flyconfig/flyconfig.py", line 189, in load
    config.runtime.owd = os.getcwd()
AttributeError: 'NoneType' object has no attribute 'owd'

Hi,

I think your paper and opensource repo is pretty awesome! I am submitting this ticket because I cant seem to find how to reproduce the final plots. Inside of the "Quality Diversity.ipynb" file there are several names below:

mle_bleu = read_data(f"{task_name}/Quality-Diversity/mle_bleu.txt")
textgail_bleu = read_data(f"{task_name}/Quality-Diversity/textgail_bleu.txt")
mle_distinct = read_data(f"{task_name}/Quality-Diversity/mle_distinct.txt")
textgail_distinct = read_data(f"{task_name}/Quality-Diversity/textgail_distinct.txt")

How did you produce these files? Could you please share your eval scripts.

Hi authors,
Thanks for your great work titled "TextGAIL: Generative Adversarial Imitation Learning for Text Generation".
I have a question about your experiment metrics. The paper stated that "When using BLEU for unconditional generation tasks, the entire training corpus is used as references for BLEU".
I was wondering what is the justification of using training corpus instead of test corpus?
I looked up SeqGAN and other text+GAN papers but do not find reasons that can support this statement. Perhaps I omit the detail, could you point it out?

Thanks!