Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

I've got different metrics, I wonder the hyperparameters for metrics posted in paper. #2

Open
PolarisRisingWar opened this issue Jun 28, 2022 · 0 comments

Comments

@PolarisRisingWar
Copy link

This is my result in test_metrics.json:
'macro_prec': 0.3009602427482605, 'macro_rec': 0.20516671240329742, 'macro_f1': 0.22117026150226593, 'micro_prec': 0.4229486087450312, 'micro_rec': 0.2341717429750442, 'micro_f1': 0.3014443629372929, 'jacc': 89.72397011976976, 'loss': 1.490009605884552

While the result in paper is:
image

And my hyperparameters are (some used the default hyperparameters, some used the ones posted on paper):

{
	"seed": 42,
	"do_train_dev": true,
	"do_test": true,
	"do_infer": false,
	"vocab_limit": null,
	"vocab_thresh": null,
	"weight_scheme": "tws",
	"tws_thresh": 10.0,
	"train_bs": 32,
	"dev_bs": 512,
	"test_bs": 512,
	"infer_bs": 512,
	"max_segments": 128,
	"max_segment_size": null,
	"num_mpath_samples": 8,
	"hidden_size": 200,
	"opt_lr": 1e-4,
	"opt_wt_decay": 1e-4,
	"sch_factor": 0.2,
	"sch_patience": 3,
	"num_epochs": 100,
	"pthresh": 0.65,
	"thetas": [1,2,3],
	"lambdas": [0.25,0.75],
	"dropout": 0.5
}

The results are different from the metrics posted on paper. So I wonder what are the hyperparamaters or other configurations used in your setting different from my configuration?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant