metadata

license: cc-by-nc-4.0
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: videomae-base-finetuned-ucf101-subset
    results: []

videomae-base-finetuned-ucf101-subset

This model is a fine-tuned version of MCG-NJU/videomae-base on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.3841
Accuracy: 0.8851

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: 4
eval_batch_size: 4
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
training_steps: 3750

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.369	0.02	75	2.2216	0.2973
1.8283	1.02	150	1.7584	0.4865
0.8729	2.02	225	1.0192	0.7027
0.4077	3.02	300	0.4849	0.8378
0.3742	4.02	375	0.1344	0.9730
0.094	5.02	450	0.2449	0.8919
0.1005	6.02	525	1.0794	0.7838
0.0053	7.02	600	0.2364	0.9459
0.0807	8.02	675	0.6659	0.8378
0.0031	9.02	750	0.4496	0.9189
0.0203	10.02	825	0.3399	0.9189
0.0093	11.02	900	0.3725	0.9459
0.0022	12.02	975	0.5498	0.9189
0.0017	13.02	1050	0.1698	0.9730
0.0014	14.02	1125	0.1923	0.9459
0.0014	15.02	1200	0.1571	0.9730
0.0474	16.02	1275	0.5193	0.8919
0.0011	17.02	1350	0.1408	0.9730
0.001	18.02	1425	0.3406	0.9459
0.0034	19.02	1500	0.2516	0.9459
0.0029	20.02	1575	0.2962	0.9189
0.0008	21.02	1650	0.4024	0.9189
0.0008	22.02	1725	0.4644	0.9189
0.1521	23.02	1800	0.4825	0.9189
0.001	24.02	1875	0.6340	0.9189
0.0245	25.02	1950	0.3779	0.9459
0.0007	26.02	2025	0.3376	0.9459
0.0011	27.02	2100	0.2833	0.9459
0.0008	28.02	2175	0.1593	0.9730
0.0008	29.02	2250	0.0856	0.9730
0.0005	30.02	2325	0.1049	0.9730
0.0005	31.02	2400	0.1132	0.9730
0.0005	32.02	2475	0.1164	0.9730
0.0005	33.02	2550	0.1243	0.9730
0.0005	34.02	2625	0.1306	0.9730
0.0005	35.02	2700	0.3919	0.9459
0.0004	36.02	2775	0.3630	0.9459
0.0004	37.02	2850	0.2762	0.9459
0.0005	38.02	2925	0.2368	0.9459
0.0004	39.02	3000	0.1935	0.9730
0.0004	40.02	3075	0.1931	0.9730
0.0004	41.02	3150	0.2139	0.9459
0.0004	42.02	3225	0.1900	0.9730
0.0006	43.02	3300	0.1751	0.9730
0.0004	44.02	3375	0.2978	0.9459
0.0004	45.02	3450	0.2777	0.9459
0.0004	46.02	3525	0.2706	0.9459
0.0004	47.02	3600	0.2638	0.9459
0.0004	48.02	3675	0.2123	0.9459
0.0004	49.02	3750	0.2106	0.9459

Framework versions

Transformers 4.24.0
Pytorch 1.8.0+cu111
Datasets 2.7.1
Tokenizers 0.13.2