fine-tuned-DatasetQAS-IDK-MRC-with-xlm-roberta-large-without-ITTL-without-freeze-LR-1e-05

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

• Loss: 0.8673
• Exact Match: 74.0838
• F1: 81.0390

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: 1e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 32
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Exact Match	F1
6.2177	0.49	36	2.3043	45.2880	46.1924
3.4831	0.98	72	1.5333	51.3089	56.5227
1.6897	1.48	108	1.1604	60.2094	68.3733
1.6897	1.97	144	0.9852	65.3141	72.9935
1.1108	2.46	180	0.9487	65.4450	72.8064
0.8854	2.95	216	0.8634	68.0628	75.1967
0.7269	3.45	252	0.9271	69.7644	76.9429
0.7269	3.94	288	0.9044	69.3717	76.4864
0.648	4.44	324	0.8352	73.1675	79.8410
0.5446	4.92	360	0.8074	74.7382	81.2181
0.5446	5.42	396	0.8726	73.4293	80.5400
0.497	5.91	432	0.8598	73.6911	80.8239
0.4647	6.41	468	0.8673	74.0838	81.0390

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu117
• Datasets 2.2.0
• Tokenizers 0.13.2