README.md

metadata

license: mit
base_model: FacebookAI/xlm-roberta-large
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: fine_tuned_XLMROBERTA_cs_wikann
    results: []

fine_tuned_XLMROBERTA_cs_wikann

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

• Loss: 0.1699
• Precision: 0.9133
• Recall: 0.9319
• F1: 0.9225
• Accuracy: 0.9699

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: 8
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 7

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.7699	0.2	500	0.3588	0.5878	0.6990	0.6386	0.8894
0.3658	0.4	1000	0.2538	0.7427	0.8258	0.7821	0.9355
0.301	0.6	1500	0.2403	0.7649	0.8237	0.7932	0.9400
0.2796	0.8	2000	0.1828	0.7967	0.8509	0.8229	0.9456
0.258	1.0	2500	0.2223	0.7770	0.8322	0.8037	0.9400
0.2192	1.2	3000	0.1911	0.8156	0.8745	0.8440	0.9511
0.2161	1.4	3500	0.1878	0.8401	0.8858	0.8623	0.9551
0.2095	1.6	4000	0.1916	0.8306	0.8783	0.8538	0.9559
0.2137	1.8	4500	0.1657	0.8573	0.8874	0.8721	0.9585
0.1884	2.0	5000	0.2134	0.8486	0.8837	0.8658	0.9542
0.164	2.2	5500	0.2038	0.8619	0.9048	0.8828	0.9588
0.1564	2.4	6000	0.1707	0.8502	0.8874	0.8684	0.9582
0.1719	2.6	6500	0.1781	0.8645	0.8994	0.8816	0.9610
0.1565	2.8	7000	0.1908	0.8712	0.9021	0.8864	0.9614
0.1713	3.0	7500	0.1628	0.8672	0.8954	0.8811	0.9623
0.1359	3.2	8000	0.1890	0.8684	0.9072	0.8874	0.9624
0.1362	3.4	8500	0.1672	0.8653	0.9065	0.8854	0.9620
0.1301	3.6	9000	0.1866	0.8698	0.9069	0.8879	0.9631
0.1345	3.8	9500	0.1766	0.8759	0.9071	0.8913	0.9647
0.1363	4.0	10000	0.1817	0.8700	0.9137	0.8913	0.9626
0.1097	4.2	10500	0.1611	0.8861	0.9118	0.8987	0.9653
0.1045	4.4	11000	0.1743	0.8899	0.9123	0.9009	0.9659
0.1068	4.6	11500	0.1771	0.8870	0.9167	0.9016	0.9660
0.1168	4.8	12000	0.1704	0.8894	0.9174	0.9032	0.9660
0.1116	5.0	12500	0.1748	0.8926	0.9203	0.9062	0.9673
0.0979	5.2	13000	0.1726	0.8956	0.9255	0.9103	0.9672
0.0992	5.4	13500	0.1798	0.9058	0.9280	0.9168	0.9686
0.0929	5.6	14000	0.1740	0.9063	0.9304	0.9182	0.9693
0.098	5.8	14500	0.1690	0.8931	0.9262	0.9094	0.9683
0.0878	6.0	15000	0.1682	0.9065	0.9294	0.9178	0.9696
0.0925	6.2	15500	0.1691	0.9102	0.9308	0.9204	0.9694
0.0841	6.4	16000	0.1657	0.9138	0.9298	0.9217	0.9699
0.0748	6.6	16500	0.1696	0.9114	0.9313	0.9213	0.9695
0.0753	6.8	17000	0.1703	0.9118	0.9311	0.9214	0.9697
0.073	7.0	17500	0.1699	0.9133	0.9319	0.9225	0.9699

Framework versions

• Transformers 4.36.2
• Pytorch 2.1.2+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0