xlm-roberta-base-finetuned-augmentation

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: 0.2338
• F1: 0.4959
• Roc Auc: 0.7179
• Accuracy: 0.6679

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
0.5295	1.0	70	0.4146	0.0	0.5	0.2599
0.3157	2.0	140	0.2945	0.2751	0.6036	0.5704
0.2656	3.0	210	0.2381	0.3269	0.6333	0.6318
0.2268	4.0	280	0.2267	0.4170	0.6785	0.6498
0.2006	5.0	350	0.2156	0.4383	0.6998	0.6498
0.199	6.0	420	0.2148	0.4382	0.6990	0.6606
0.1646	7.0	490	0.2182	0.4257	0.6864	0.6390
0.1452	8.0	560	0.2136	0.4425	0.7021	0.6498
0.1385	9.0	630	0.2163	0.4603	0.7170	0.6426
0.1258	10.0	700	0.2228	0.4528	0.7018	0.6534
0.116	11.0	770	0.2287	0.4383	0.6938	0.6462
0.1097	12.0	840	0.2300	0.4580	0.7088	0.6606
0.1149	13.0	910	0.2309	0.4770	0.7058	0.6679
0.08	14.0	980	0.2288	0.4770	0.7094	0.6715
0.0787	15.0	1050	0.2338	0.4959	0.7179	0.6679
0.0755	16.0	1120	0.2356	0.4936	0.7169	0.6679
0.0761	17.0	1190	0.2365	0.4841	0.7110	0.6715
0.0676	18.0	1260	0.2361	0.4894	0.7128	0.6715
0.0755	19.0	1330	0.2369	0.4959	0.7180	0.6751

Framework versions

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