Timesformer_BdSLW60_100_0.15splt_gradAcu

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: 1.1567
• Accuracy: 0.7163

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: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 22400

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.5832	0.0400	897	1.2956	0.7671
0.483	1.0401	1795	0.4015	0.9082
0.3155	2.0401	2693	0.1985	0.9588
0.2306	3.0401	3591	0.2235	0.9341
0.1954	4.0400	4488	0.2092	0.9529
0.1373	5.0401	5386	0.1641	0.9541
0.1823	6.0401	6284	0.1500	0.9647
0.1086	7.0401	7182	0.1330	0.9718
0.1537	8.0400	8079	0.1076	0.9706
0.1189	9.0401	8977	0.1491	0.96
0.1437	10.0401	9875	0.1375	0.9671
0.1009	11.0401	10773	0.1095	0.9682
0.0777	12.0400	11670	0.1060	0.9718

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1