Edit model card

videomae-base-short-finetuned-ssv2-finetuned-rwf2000-epochs8-batch8

This model is a fine-tuned version of MCG-NJU/videomae-base-short-finetuned-ssv2 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7821
  • Accuracy: 0.6713

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: 3200

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.4247 0.06 200 0.4205 0.8063
0.4125 1.06 400 0.6749 0.72
0.3265 2.06 600 1.3838 0.5763
0.2204 3.06 800 0.6725 0.7275
0.2965 4.06 1000 0.4583 0.8263
0.1883 5.06 1200 0.3786 0.8488
0.1321 6.06 1400 1.6632 0.5962
0.369 7.06 1600 0.6018 0.8063
0.3764 8.06 1800 0.8546 0.74
0.2401 9.06 2000 0.5422 0.825
0.1943 10.06 2200 0.5868 0.8113
0.1352 11.06 2400 0.7111 0.8063
0.2276 12.06 2600 0.8847 0.7812
0.149 13.06 2800 0.8581 0.7837
0.0848 14.06 3000 0.8707 0.7788
0.046 15.06 3200 0.7914 0.7963

Framework versions

  • Transformers 4.25.1
  • Pytorch 1.13.1+cu117
  • Datasets 2.8.0
  • Tokenizers 0.13.2
Downloads last month
4
Inference API
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.