tsf-gs-rots-wtoken-DRPT0.3-r128-f150-6.6-h768-i3072-p32-b8-e50

This model is a fine-tuned version of facebook/timesformer-base-finetuned-k400 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.7417
• Accuracy: 0.6150

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: 8
• eval_batch_size: 8
• 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: 5400
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1273	0.0202	109	1.1116	0.3262
1.1488	1.0202	218	1.0997	0.3369
1.1365	2.0202	327	1.1016	0.3369
1.1232	3.0202	436	1.1009	0.3369
1.0927	4.0202	545	1.1017	0.3369
1.0807	5.0202	654	1.1055	0.3262
1.1257	6.0202	763	1.1005	0.3262
1.0961	7.0202	872	1.0999	0.3369
1.1192	8.0202	981	1.0997	0.3262
1.1039	9.0202	1090	1.0982	0.3369
1.1047	10.0202	1199	1.0988	0.3369
1.0662	11.0202	1308	1.3609	0.3743
1.197	12.0202	1417	1.1068	0.3369
1.1331	13.0202	1526	1.1085	0.3422
1.1174	14.0202	1635	1.1129	0.3262
1.0838	15.0202	1744	1.0893	0.4011
1.0943	16.0202	1853	1.0385	0.3904
1.0989	17.0202	1962	1.0832	0.5027
1.0326	18.0202	2071	0.9577	0.4813
1.0394	19.0202	2180	0.9385	0.6043
0.9952	20.0202	2289	0.8765	0.6096
0.9504	21.0202	2398	0.8307	0.6096
0.9256	22.0202	2507	0.8004	0.6471
0.8924	23.0202	2616	0.9152	0.5989
0.9158	24.0202	2725	0.7679	0.6952
0.8838	25.0202	2834	0.7533	0.6952
1.0359	26.0202	2943	0.7408	0.6845
0.8345	27.0202	3052	0.7069	0.7112
0.8803	28.0202	3161	0.7740	0.6684
0.7475	29.0202	3270	0.6999	0.7112
0.5596	30.0202	3379	0.8609	0.6364
0.8362	31.0202	3488	1.5082	0.4813
0.672	32.0202	3597	0.7459	0.7059
0.6874	33.0202	3706	0.9255	0.6845
0.6259	34.0202	3815	0.8475	0.6364
0.6356	35.0202	3924	0.8400	0.6791
0.6482	36.0202	4033	0.8579	0.6310
0.5495	37.0202	4142	1.5998	0.5241
0.6663	38.0202	4251	0.7969	0.7112
0.6363	39.0202	4360	1.1134	0.6845
0.6794	40.0202	4469	0.9227	0.6952
0.6632	41.0202	4578	1.1304	0.6631
0.7225	42.0202	4687	0.9182	0.7112
0.6032	43.0202	4796	1.2193	0.6578
0.5534	44.0202	4905	1.4561	0.6524
0.4216	45.0202	5014	1.3694	0.6364
0.6082	46.0202	5123	1.5731	0.5989
0.7025	47.0202	5232	1.8556	0.6257
0.4275	48.0202	5341	1.7699	0.6257
0.462	49.0109	5400	1.7417	0.6150

Framework versions

• Transformers 4.41.2
• Pytorch 1.13.0+cu117
• Datasets 2.20.0
• Tokenizers 0.19.1