Sigmoid or Softmax missing in evaluation

[shakib-root]

Shouldn't sigmoid(multilabel classification) or softmax(multiclass classification) be applied here to get the predictions? Otherwise, later the thresholds are being applied on the logits directly.

GoEmotions-pytorch/run_goemotions.py

Line 183 in 47f5afb

preds = logits.detach().cpu().numpy()

[monologg]

@shakib-root

It's totally true that there is a mistake.

I'm fixing the issue on this branch (d47dfed) and re-running the models. As training is over, I'll reupload the model & result!

Thank you for informing the issue

[blackcat84]

Does this affect the already fine-tuned models (like the ones on the model hub of Hugging Face) or is it just a matter of metrics computation while fine-tuning?

[monologg]

@blackcat84

This one doesn't affect the fine-tuned models, but there might be a possibility that the model from huggingface model hub isn't the best model cuz I've chose the model based on the wrong metrics

[blackcat84]

Right, makes perfectly sense. Looking forward to your new results! Thanks!

[monologg]

Fixed on b75f9d3 :)

Also huggingface s3 models are updated!