afro-xlmr-base-sun-finetuned-augmentation-LUNAR

This model is a fine-tuned version of Davlan/afro-xlmr-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3556
• F1: 0.3987
• Roc Auc: 0.6273
• Accuracy: 0.5156

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 100
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.6264	1.0	57	0.4136	0.1420	0.5	0.5067
0.382	2.0	114	0.3981	0.1420	0.5	0.5067
0.4333	3.0	171	0.3526	0.2407	0.5539	0.5244
0.3472	4.0	228	0.3299	0.2767	0.5946	0.5511
0.325	5.0	285	0.3186	0.2669	0.6007	0.5156
0.3188	6.0	342	0.3278	0.2681	0.5975	0.5289
0.2909	7.0	399	0.3446	0.2675	0.5809	0.5422
0.2579	8.0	456	0.3238	0.2935	0.6150	0.5289
0.2779	9.0	513	0.3341	0.2891	0.6043	0.52
0.2547	10.0	570	0.3615	0.3142	0.5980	0.52
0.2266	11.0	627	0.3394	0.3499	0.6212	0.5289
0.2258	12.0	684	0.3587	0.3515	0.6061	0.5022
0.2159	13.0	741	0.3402	0.3677	0.6297	0.5333
0.2163	14.0	798	0.3485	0.3678	0.6198	0.4978
0.2007	15.0	855	0.3556	0.3987	0.6273	0.5156
0.1955	16.0	912	0.3552	0.3724	0.6195	0.5022
0.1806	17.0	969	0.3619	0.3744	0.6195	0.5111
0.189	18.0	1026	0.3559	0.3850	0.6227	0.4889
0.1837	19.0	1083	0.3561	0.3868	0.6241	0.4933

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.20.0