NeRUBioS_xlm_RoBERTa_base_Training_Testing

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: 0.3585
• Negref Precision: 0.5638
• Negref Recall: 0.6035
• Negref F1: 0.5830
• Neg Precision: 0.9508
• Neg Recall: 0.9642
• Neg F1: 0.9575
• Nsco Precision: 0.8692
• Nsco Recall: 0.9047
• Nsco F1: 0.8866
• Unc Precision: 0.8005
• Unc Recall: 0.8846
• Unc F1: 0.8404
• Usco Precision: 0.6696
• Usco Recall: 0.7815
• Usco F1: 0.7212
• Precision: 0.8184
• Recall: 0.8628
• F1: 0.8400
• Accuracy: 0.9482

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: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 12

Training results

Training Loss	Epoch	Step	Validation Loss	Negref Precision	Negref Recall	Negref F1	Neg Precision	Neg Recall	Neg F1	Nsco Precision	Nsco Recall	Nsco F1	Unc Precision	Unc Recall	Unc F1	Usco Precision	Usco Recall	Usco F1	Precision	Recall	F1	Accuracy
0.2259	1.0	1729	0.2246	0.4076	0.4890	0.4446	0.9112	0.9515	0.9310	0.7928	0.8654	0.8275	0.7015	0.8256	0.7585	0.4629	0.6735	0.5487	0.7158	0.8122	0.7610	0.9287
0.16	2.0	3458	0.2028	0.5217	0.5301	0.5259	0.9283	0.9642	0.9459	0.8311	0.8896	0.8593	0.7812	0.8513	0.8147	0.5734	0.7532	0.6511	0.7817	0.8405	0.8100	0.9397
0.1235	3.0	5187	0.2148	0.5176	0.4963	0.5067	0.9520	0.9607	0.9563	0.8641	0.8850	0.8744	0.7684	0.8846	0.8224	0.6113	0.7481	0.6728	0.8038	0.8350	0.8191	0.9439
0.0949	4.0	6916	0.2261	0.5054	0.6211	0.5573	0.9327	0.9642	0.9482	0.8450	0.8828	0.8635	0.7976	0.8487	0.8224	0.6034	0.7352	0.6628	0.7818	0.8512	0.8150	0.9450
0.0633	5.0	8645	0.2354	0.5609	0.5947	0.5773	0.9417	0.9649	0.9532	0.8669	0.9062	0.8861	0.8062	0.8641	0.8342	0.6334	0.7506	0.6871	0.8118	0.8573	0.8340	0.9461
0.0495	6.0	10374	0.2829	0.5585	0.5962	0.5767	0.9445	0.9684	0.9563	0.8671	0.9077	0.8869	0.8116	0.8615	0.8358	0.6526	0.7532	0.6993	0.8151	0.8592	0.8366	0.9442
0.0365	7.0	12103	0.2699	0.5446	0.5830	0.5631	0.9552	0.9572	0.9562	0.8804	0.9024	0.8913	0.8080	0.8846	0.8446	0.6521	0.7661	0.7045	0.8182	0.8550	0.8362	0.9473
0.0265	8.0	13832	0.3082	0.5630	0.5580	0.5605	0.9466	0.9593	0.9529	0.8702	0.9024	0.8860	0.8038	0.8718	0.8364	0.6571	0.7635	0.7063	0.8194	0.8502	0.8345	0.9460
0.0216	9.0	15561	0.3286	0.5485	0.5977	0.5720	0.9388	0.9691	0.9537	0.8715	0.9077	0.8892	0.8085	0.8769	0.8413	0.6453	0.7763	0.7048	0.8105	0.8633	0.8361	0.9455
0.0133	10.0	17290	0.3503	0.5732	0.6094	0.5907	0.9481	0.9628	0.9554	0.8698	0.8994	0.8843	0.8137	0.8846	0.8477	0.6816	0.7815	0.7281	0.8223	0.8616	0.8415	0.9482
0.0088	11.0	19019	0.3476	0.5584	0.6182	0.5868	0.9450	0.9656	0.9552	0.8614	0.9070	0.8836	0.8080	0.8846	0.8446	0.6659	0.7789	0.7180	0.8126	0.8661	0.8385	0.9483
0.0093	12.0	20748	0.3585	0.5638	0.6035	0.5830	0.9508	0.9642	0.9575	0.8692	0.9047	0.8866	0.8005	0.8846	0.8404	0.6696	0.7815	0.7212	0.8184	0.8628	0.8400	0.9482

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2