xlm-roberta-large-finetuned-augument-visquad2-5-4-2023-1

This model is a fine-tuned version of xlm-roberta-large on the None dataset. It achieves the following results on the evaluation set:

• Best F1: 76.9467
• Loss: 2.4868
• Exact: 38.7333
• F1: 56.7311
• Total: 3821
• Hasans Exact: 55.1451
• Hasans F1: 81.0666
• Hasans Total: 2653
• Noans Exact: 1.4555
• Noans F1: 1.4555
• Noans Total: 1168
• Best Exact: 60.7694
• Best Exact Thresh: 0.8788
• Best F1 Thresh: 0.9686

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: 5e-06
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.2
• lr_scheduler_warmup_steps: 1000
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Best F1	Validation Loss	Exact	F1	Total	Hasans Exact	Hasans F1	Hasans Total	Noans Exact	Noans F1	Noans Total	Best Exact	Best Exact Thresh	Best F1 Thresh
1.1009	1.0	4221	69.8405	1.2010	36.1685	54.2642	3821	52.0920	78.1543	2653	0.0	0.0	1168	55.4305	0.8416	0.9065
0.4716	2.0	8443	74.1358	1.0553	38.2884	56.0896	3821	54.9943	80.6326	2653	0.3425	0.3425	1168	58.8328	0.8002	0.9118
0.3487	3.0	12664	76.3875	1.1176	39.3876	56.5884	3821	56.6905	81.4641	2653	0.0856	0.0856	1168	61.3190	0.7923	0.9324
0.2747	4.0	16886	76.3938	1.2634	38.6548	56.3082	3821	55.6728	81.0982	2653	0.0	0.0	1168	60.4030	0.7414	0.9059
0.217	5.0	21107	76.5504	1.3581	39.3353	56.9569	3821	56.4644	81.8441	2653	0.4281	0.4281	1168	61.0050	0.8307	0.8701
0.1758	6.0	25329	77.2312	1.5473	39.4399	56.6673	3821	56.7282	81.5401	2653	0.1712	0.1712	1168	61.4761	0.8283	0.8996
0.1429	7.0	29550	77.2045	1.7840	38.8642	56.8934	3821	55.7105	81.6773	2653	0.5993	0.5993	1168	61.3190	0.7413	0.9449
0.1159	8.0	33772	76.6868	2.1179	38.7071	56.7210	3821	55.2959	81.2405	2653	1.0274	1.0274	1168	60.4292	0.5939	0.9917
0.0997	9.0	37993	77.1237	2.3697	38.6548	56.6046	3821	55.4467	81.2990	2653	0.5137	0.5137	1168	60.8479	0.9509	0.9636
0.0873	10.0	42210	76.9467	2.4868	38.7333	56.7311	3821	55.1451	81.0666	2653	1.4555	1.4555	1168	60.7694	0.8788	0.9686

Framework versions

• Transformers 4.27.3
• Pytorch 1.13.1+cu117
• Datasets 2.10.1
• Tokenizers 0.13.2