XMLRoberta_70KURL

This model is a fine-tuned version of FacebookAI/xlm-roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4150
• Accuracy: 0.9408
• F1: 0.9448

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2150
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.4651	200	0.4701	0.8955	0.8673
No log	0.9302	400	0.1893	0.9359	0.9368
No log	1.3953	600	0.2241	0.9128	0.9192
No log	1.8605	800	0.2777	0.8848	0.8984
0.382	2.3256	1000	0.1388	0.9504	0.9525
0.382	2.7907	1200	0.1028	0.9694	0.9701
0.382	3.2558	1400	0.1413	0.9557	0.9579
0.382	3.7209	1600	0.0929	0.9718	0.9722
0.1521	4.1860	1800	0.1008	0.9695	0.9702
0.1521	4.6512	2000	0.1939	0.9238	0.9306
0.1521	5.1163	2200	0.1550	0.9401	0.9443
0.1521	5.5814	2400	0.0813	0.9744	0.9750
0.1044	6.0465	2600	0.2088	0.9193	0.9267
0.1044	6.5116	2800	0.1343	0.9523	0.9548
0.1044	6.9767	3000	0.2172	0.9219	0.9289
0.1044	7.4419	3200	0.1097	0.9656	0.9668
0.1044	7.9070	3400	0.3044	0.9147	0.9230
0.0762	8.3721	3600	0.2122	0.9283	0.9341
0.0762	8.8372	3800	0.1430	0.9532	0.9556
0.0762	9.3023	4000	0.2332	0.9312	0.9368
0.0762	9.7674	4200	0.2167	0.9297	0.9353
0.0564	10.2326	4400	0.1904	0.9475	0.9506
0.0564	10.6977	4600	0.2916	0.9196	0.9270
0.0564	11.1628	4800	0.2317	0.9451	0.9484
0.0564	11.6279	5000	0.2430	0.9475	0.9503
0.042	12.0930	5200	0.4035	0.9248	0.9315
0.042	12.5581	5400	0.3060	0.9352	0.9398
0.042	13.0233	5600	0.2894	0.9359	0.9407
0.042	13.4884	5800	0.2804	0.9439	0.9474
0.042	13.9535	6000	0.2941	0.9456	0.9490
0.0316	14.4186	6200	0.3211	0.9424	0.9460
0.0316	14.8837	6400	0.3453	0.9371	0.9416
0.0316	15.3488	6600	0.2587	0.9548	0.9569
0.0316	15.8140	6800	0.3433	0.9432	0.9468
0.025	16.2791	7000	0.3454	0.9416	0.9455
0.025	16.7442	7200	0.2977	0.9450	0.9484
0.025	17.2093	7400	0.3622	0.9452	0.9486
0.025	17.6744	7600	0.3035	0.9550	0.9572
0.0196	18.1395	7800	0.3588	0.9464	0.9496
0.0196	18.6047	8000	0.3714	0.9467	0.9500
0.0196	19.0698	8200	0.4517	0.9341	0.9391
0.0196	19.5349	8400	0.4078	0.9411	0.9451
0.0148	20.0	8600	0.4150	0.9408	0.9448

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.2
• Tokenizers 0.19.1