gg-bert-base-uncased

This model is a fine-tuned version of FacebookAI/xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7791
• Accuracy: 0.752
• Precision: 0.7388
• Recall: 0.7570
• F1: 0.7396

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0002
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
2.2541	1.0	469	2.2063	0.1488	0.2698	0.1552	0.1036
1.8967	2.0	938	1.8773	0.5168	0.5264	0.5331	0.4788
1.5747	3.0	1407	1.5546	0.5984	0.6125	0.6118	0.5636
1.4206	4.0	1876	1.3029	0.6528	0.6732	0.6666	0.6224
1.2804	5.0	2345	1.1876	0.6928	0.6972	0.6989	0.6844
1.1587	6.0	2814	1.0644	0.7136	0.7105	0.7136	0.6858
1.1589	7.0	3283	0.9883	0.7216	0.7173	0.7261	0.7031
1.0745	8.0	3752	0.9485	0.728	0.7151	0.7318	0.7195
1.0348	9.0	4221	0.9278	0.7328	0.7306	0.7372	0.7166
1.0019	10.0	4690	0.9114	0.72	0.7316	0.7231	0.7006
1.0204	11.0	5159	0.8967	0.7152	0.7187	0.7215	0.6895
1.0651	12.0	5628	0.8574	0.7424	0.7446	0.7474	0.7327
0.9841	13.0	6097	0.8461	0.7328	0.7495	0.7370	0.7076
0.9794	14.0	6566	0.8510	0.7248	0.7157	0.7319	0.7022
1.0242	15.0	7035	0.8127	0.7264	0.7112	0.7300	0.6998
0.9614	16.0	7504	0.8146	0.7312	0.7210	0.7376	0.7149
0.9358	17.0	7973	0.8288	0.736	0.7487	0.7439	0.7275
0.9719	18.0	8442	0.7958	0.7488	0.7403	0.7530	0.7349
0.9159	19.0	8911	0.7973	0.7472	0.7388	0.7522	0.7357
0.9824	20.0	9380	0.7921	0.7504	0.7439	0.7562	0.7363
1.0215	21.0	9849	0.7831	0.7536	0.7415	0.7586	0.7392
0.9191	22.0	10318	0.7780	0.7504	0.7387	0.7554	0.7399
0.9087	23.0	10787	0.7843	0.7472	0.7352	0.7536	0.7345
0.9198	24.0	11256	0.7793	0.7504	0.7358	0.7554	0.7374
0.9162	25.0	11725	0.7791	0.752	0.7388	0.7570	0.7396

Framework versions

• Transformers 4.43.3
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1