videomae-finetuned-nba-5-class-4-batch-8000-vid-multilabel-4

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

• Loss: 0.4886
• F1: 0.8863
• Roc Auc: 0.9209
• Accuracy: 0.7962

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: 1.5e-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: cosine
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 50000

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.771	0.04	2000	0.7660	0.6227	0.7324	0.2937
0.6225	1.04	4000	0.6446	0.7401	0.8162	0.4653
0.6093	2.04	6000	0.6517	0.7441	0.8223	0.5253
0.603	3.04	8000	0.5630	0.8151	0.8722	0.6558
0.6547	4.04	10000	0.5009	0.8425	0.8911	0.6695
0.6426	5.04	12000	0.5179	0.8422	0.9012	0.6642
0.4447	6.04	14000	0.5052	0.8537	0.8973	0.7147
0.6949	7.04	16000	0.5045	0.8548	0.8974	0.7337
0.509	8.04	18000	0.5262	0.8705	0.9089	0.7768
0.4341	9.04	20000	0.4731	0.8831	0.9163	0.7863
0.5037	10.04	22000	0.5040	0.8781	0.9154	0.7747
0.3592	11.04	24000	0.5091	0.8746	0.9101	0.7821
0.2829	12.04	26000	0.4709	0.8867	0.9230	0.7905
0.3599	13.04	28000	0.4722	0.8888	0.9230	0.7947
0.4152	14.04	30000	0.4744	0.8911	0.9247	0.8074
0.1167	15.04	32000	0.4817	0.8949	0.9241	0.8179
0.2721	16.04	34000	0.4627	0.9064	0.9337	0.8274
0.42	17.04	36000	0.4849	0.8992	0.9282	0.8295
0.5841	18.04	38000	0.4747	0.9026	0.9304	0.8379
0.509	19.04	40000	0.4779	0.8991	0.9273	0.8316
0.1599	20.04	42000	0.4945	0.9056	0.9319	0.8411
0.3645	21.04	44000	0.4875	0.8984	0.9277	0.8242
0.4402	22.04	46000	0.5156	0.9004	0.9283	0.8368
0.2769	23.04	48000	0.5063	0.9028	0.9296	0.8389
0.3009	24.04	50000	0.5122	0.8975	0.9257	0.8326

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1