Timesformer_default_fold_10_10_epoch_noAug_batch8

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: 0.8574
• Accuracy: 0.7641

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: 9030

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.2337	0.1001	904	0.1879	0.9626
0.0508	1.1001	1808	0.0274	0.9938
0.0191	2.1001	2712	0.0239	0.9925
0.0169	3.1001	3616	0.0305	0.9913
0.0008	4.1001	4520	0.0126	0.9975
0.0001	5.1001	5424	0.0077	0.9963
0.0001	6.1001	6328	0.0228	0.9963
0.0001	7.1001	7232	0.0163	0.9975
0.0001	8.1001	8136	0.0157	0.9975
0.0001	9.0990	9030	0.0157	0.9975

Framework versions

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