README.md

metadata

license: apache-2.0
base_model: facebook/deit-base-distilled-patch16-224
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: deit-base-distilled-patch16-224-hasta-55-fold1
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: imagefolder
          type: imagefolder
          config: default
          split: train
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.6944444444444444

deit-base-distilled-patch16-224-hasta-55-fold1

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

• Loss: 0.9277
• Accuracy: 0.6944

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.5714	1	1.1037	0.4444
No log	1.7143	3	1.0649	0.3889
No log	2.8571	5	1.0727	0.4167
No log	4.0	7	1.0623	0.5
No log	4.5714	8	1.0530	0.4722
1.065	5.7143	10	1.0510	0.5278
1.065	6.8571	12	1.0745	0.6111
1.065	8.0	14	1.0511	0.5
1.065	8.5714	15	1.0158	0.5556
1.065	9.7143	17	0.9998	0.6389
1.065	10.8571	19	1.0472	0.5556
0.9239	12.0	21	0.9675	0.5833
0.9239	12.5714	22	0.9732	0.5278
0.9239	13.7143	24	0.9489	0.5
0.9239	14.8571	26	0.9277	0.6944
0.9239	16.0	28	0.9244	0.5833
0.9239	16.5714	29	0.9643	0.5833
0.7838	17.7143	31	0.9721	0.5278
0.7838	18.8571	33	0.9432	0.6111
0.7838	20.0	35	0.9337	0.6389
0.7838	20.5714	36	0.9369	0.5556
0.7838	21.7143	38	0.9481	0.6111
0.6124	22.8571	40	0.9643	0.6111
0.6124	24.0	42	0.9300	0.6111
0.6124	24.5714	43	0.9629	0.6389
0.6124	25.7143	45	0.9200	0.5556
0.6124	26.8571	47	1.0089	0.5
0.6124	28.0	49	0.9980	0.5833
0.5068	28.5714	50	0.9876	0.5556
0.5068	29.7143	52	1.0464	0.5
0.5068	30.8571	54	1.0312	0.5556
0.5068	32.0	56	1.0946	0.5278
0.5068	32.5714	57	1.1677	0.5
0.5068	33.7143	59	1.1616	0.4722
0.4375	34.8571	61	1.0658	0.5556
0.4375	36.0	63	1.0921	0.6111
0.4375	36.5714	64	1.0801	0.6389
0.4375	37.7143	66	1.0586	0.5556
0.4375	38.8571	68	1.2152	0.5
0.3932	40.0	70	1.1543	0.5
0.3932	40.5714	71	1.0655	0.5556
0.3932	41.7143	73	0.9952	0.5556
0.3932	42.8571	75	0.9986	0.5278
0.3932	44.0	77	1.0175	0.5556
0.3932	44.5714	78	1.0234	0.5556
0.3539	45.7143	80	1.0385	0.5556
0.3539	46.8571	82	1.0191	0.5278
0.3539	48.0	84	1.0151	0.5556
0.3539	48.5714	85	1.0203	0.5556
0.3539	49.7143	87	1.0341	0.5556
0.3539	50.8571	89	1.0720	0.5556
0.3257	52.0	91	1.0951	0.5556
0.3257	52.5714	92	1.0927	0.5556
0.3257	53.7143	94	1.0883	0.5556
0.3257	54.8571	96	1.0874	0.5556
0.3257	56.0	98	1.0883	0.5278
0.3257	56.5714	99	1.0868	0.5278
0.321	57.1429	100	1.0858	0.5278

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1