README.md

metadata

license: apache-2.0
base_model: facebook/deit-base-patch16-224
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: hushem_1x_deit_base_adamax_00001_fold4
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: imagefolder
          type: imagefolder
          config: default
          split: test
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.7619047619047619

hushem_1x_deit_base_adamax_00001_fold4

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

• Loss: 0.5330
• Accuracy: 0.7619

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: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	6	1.3484	0.2857
1.3393	2.0	12	1.3138	0.4524
1.3393	3.0	18	1.2772	0.4048
1.1604	4.0	24	1.2261	0.4524
1.014	5.0	30	1.1685	0.4762
1.014	6.0	36	1.1130	0.5476
0.8569	7.0	42	1.0641	0.5476
0.8569	8.0	48	1.0213	0.5476
0.7145	9.0	54	0.9685	0.5714
0.5812	10.0	60	0.9109	0.6190
0.5812	11.0	66	0.8739	0.6905
0.4645	12.0	72	0.8376	0.6667
0.4645	13.0	78	0.8046	0.6667
0.3784	14.0	84	0.7821	0.6667
0.308	15.0	90	0.7516	0.6905
0.308	16.0	96	0.7309	0.7143
0.2446	17.0	102	0.7113	0.7381
0.2446	18.0	108	0.6911	0.7143
0.2032	19.0	114	0.6782	0.6905
0.1713	20.0	120	0.6649	0.7381
0.1713	21.0	126	0.6459	0.7381
0.1338	22.0	132	0.6300	0.7143
0.1338	23.0	138	0.6291	0.7619
0.113	24.0	144	0.6105	0.8095
0.0989	25.0	150	0.5999	0.7619
0.0989	26.0	156	0.5962	0.7857
0.0793	27.0	162	0.5828	0.7619
0.0793	28.0	168	0.5775	0.7857
0.0704	29.0	174	0.5718	0.7857
0.0586	30.0	180	0.5598	0.7857
0.0586	31.0	186	0.5576	0.7857
0.0498	32.0	192	0.5530	0.7857
0.0498	33.0	198	0.5470	0.7857
0.0487	34.0	204	0.5432	0.7857
0.0426	35.0	210	0.5430	0.7619
0.0426	36.0	216	0.5406	0.7619
0.0394	37.0	222	0.5370	0.7619
0.0394	38.0	228	0.5337	0.7619
0.039	39.0	234	0.5328	0.7619
0.0365	40.0	240	0.5330	0.7619
0.0365	41.0	246	0.5331	0.7619
0.0366	42.0	252	0.5330	0.7619
0.0366	43.0	258	0.5330	0.7619
0.0347	44.0	264	0.5330	0.7619
0.0374	45.0	270	0.5330	0.7619
0.0374	46.0	276	0.5330	0.7619
0.0363	47.0	282	0.5330	0.7619
0.0363	48.0	288	0.5330	0.7619
0.0346	49.0	294	0.5330	0.7619
0.0366	50.0	300	0.5330	0.7619

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.14.7
• Tokenizers 0.15.0