videomae_base_wlasl_2000_20ep_coR

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: 7.7331
• Accuracy: 0.0051

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 35720
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
30.647	0.05	1786	7.6329	0.0015
30.5394	1.0500	3572	7.6123	0.0010
30.4288	2.0500	5358	7.5918	0.0015
30.2564	3.0500	7145	7.5909	0.0026
30.0749	4.05	8931	7.6150	0.0028
29.8125	5.0500	10717	7.6600	0.0020
29.5133	6.0500	12503	7.6993	0.0023
29.1777	7.0500	14290	7.7246	0.0028
28.8751	8.05	16076	7.7638	0.0028
28.5476	9.0500	17862	7.7868	0.0031
28.2183	10.0500	19648	7.7697	0.0028
27.8581	11.0500	21435	7.7797	0.0031
27.5243	12.05	23221	7.7842	0.0049
27.1882	13.0500	25007	7.7781	0.0036
26.8459	14.0500	26793	7.7716	0.0043
26.507	15.0500	28580	7.7595	0.0041
26.2302	16.05	30366	7.7506	0.0051
25.9571	17.0500	32152	7.7419	0.0054
25.7257	18.0500	33938	7.7376	0.0051
25.5424	19.0499	35720	7.7331	0.0051

Framework versions

• Transformers 4.46.1
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.1