distilbert-qa-checkpoint-v2

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3141

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
0.6564	1.0	658	0.3612
0.3397	2.0	1316	0.3361
0.2885	3.0	1974	0.3515
0.2407	4.0	2632	0.3672
0.2213	5.0	3290	0.3718
0.2197	6.0	3948	0.3967
0.1986	7.0	4606	0.4115
0.1932	8.0	5264	0.4152
0.19	9.0	5922	0.4208
0.1844	10.0	6580	0.4472
0.1824	11.0	7238	0.4466
0.1812	12.0	7896	0.2695
0.0078	13.0	8554	0.2824
0.0073	14.0	9212	0.2793
0.0048	15.0	9870	0.3107
0.0033	16.0	10528	0.3074
0.0022	17.0	11186	0.3073
0.0038	18.0	11844	0.3147
0.0013	19.0	12502	0.3160
0.0008	20.0	13160	0.3141

Framework versions

• Transformers 4.27.2
• Pytorch 1.13.1+cu117
• Datasets 2.11.0
• Tokenizers 0.13.3