jialicheng/unlearn_nlvr2_vilt_salun_10_42

42

This model is a fine-tuned version of dandelin/vilt-b32-finetuned-nlvr2 on the nlvr2 dataset. It achieves the following results on the evaluation set:

• Loss: 0.9471
• Accuracy: 0.7223
• Dt Accuracy: 0.7223
• Df Accuracy: 0.4254
• Unlearn Overall Accuracy: 0.4659
• Unlearn Time: None

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: 32
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 6

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Overall Accuracy	Unlearn Overall Accuracy	Time
0.3977	1.0	2700	0.5272	0.8690	0.3156	0.3156	None
0.3329	2.0	5400	0.5785	0.7129	0.3795	0.3795	None
0.2612	3.0	8100	0.6846	0.5871	0.4222	0.4222	None
0.2205	4.0	10800	0.8106	0.4994	0.4478	0.4478	None
0.1673	5.0	13500	0.9010	0.4432	0.4619	0.4619	None
0.1453	6.0	16200	0.9471	0.4254	0.4659	0.4659	None

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.2+cu118
• Datasets 2.18.0
• Tokenizers 0.15.2