resnet152-FV-finetuned-memes

This model is a fine-tuned version of microsoft/resnet-152 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.6772
• Accuracy: 0.7558
• Precision: 0.7557
• Recall: 0.7558
• F1: 0.7546

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: 0.00012
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
1.5739	0.99	20	1.5427	0.4521	0.3131	0.4521	0.2880
1.4353	1.99	40	1.3786	0.4490	0.3850	0.4490	0.2791
1.3026	2.99	60	1.2734	0.4799	0.3073	0.4799	0.3393
1.1579	3.99	80	1.1378	0.5278	0.4300	0.5278	0.4143
1.0276	4.99	100	1.0231	0.5734	0.4497	0.5734	0.4865
0.8826	5.99	120	0.9228	0.6252	0.5983	0.6252	0.5637
0.766	6.99	140	0.8441	0.6662	0.6474	0.6662	0.6320
0.6732	7.99	160	0.8009	0.6901	0.6759	0.6901	0.6704
0.5653	8.99	180	0.7535	0.7218	0.7141	0.7218	0.7129
0.4957	9.99	200	0.7317	0.7257	0.7248	0.7257	0.7200
0.4534	10.99	220	0.6808	0.7434	0.7405	0.7434	0.7390
0.3792	11.99	240	0.6949	0.7450	0.7454	0.7450	0.7399
0.3489	12.99	260	0.6746	0.7496	0.7511	0.7496	0.7474
0.3113	13.99	280	0.6637	0.7573	0.7638	0.7573	0.7579
0.2947	14.99	300	0.6451	0.7589	0.7667	0.7589	0.7610
0.2776	15.99	320	0.6754	0.7543	0.7565	0.7543	0.7525
0.2611	16.99	340	0.6808	0.7550	0.7607	0.7550	0.7529
0.2428	17.99	360	0.7005	0.7457	0.7497	0.7457	0.7404
0.2346	18.99	380	0.6597	0.7573	0.7642	0.7573	0.7590
0.2367	19.99	400	0.6772	0.7558	0.7557	0.7558	0.7546

Framework versions

• Transformers 4.24.0.dev0
• Pytorch 1.11.0+cu102
• Datasets 2.6.1.dev0
• Tokenizers 0.13.1