This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 1.4766
Accuracy: 0.5127

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: 3e-05
train_batch_size: 128
eval_batch_size: 128
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 1024
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 1000
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.9670	22	2.0708	0.1449
No log	1.9670	44	2.0653	0.1663
2.0961	2.9670	66	2.0564	0.1931
2.0961	3.9670	88	2.0423	0.2350
2.0586	4.9670	110	2.0171	0.2823
2.0586	5.9670	132	1.9638	0.3305
1.9128	6.9670	154	1.8130	0.3968
1.9128	7.9670	176	1.6647	0.4278
1.9128	8.9670	198	1.5676	0.4844
1.6466	9.9670	220	1.4766	0.5127

Framework versions

Transformers 4.51.0
Pytorch 2.5.1+cu124
Datasets 3.5.0
Tokenizers 0.21.0

NekoJar/trainer_output

trainer_output

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NekoJar/
trainer_output