9_mae_4

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

  • Loss: 0.4857
  • Accuracy: 0.8667

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: 1e-05
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • 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: 9700

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.6583 0.0201 195 0.7953 0.4667
0.6439 1.0201 390 0.7813 0.4667
0.5285 2.0201 585 0.5919 0.8
0.3275 3.0201 780 0.5570 0.6222
0.5265 4.0201 975 0.7656 0.7333
0.3472 5.0201 1170 1.0666 0.5556
0.3978 6.0201 1365 0.7593 0.7333
0.5104 7.0201 1560 0.6972 0.7111
0.5969 8.0201 1755 0.5279 0.7778
0.5841 9.0201 1950 0.6547 0.8
0.4973 10.0201 2145 0.5363 0.7556
0.3399 11.0201 2340 1.0321 0.6667
0.2955 12.0201 2535 1.2320 0.6
0.5854 13.0201 2730 1.3658 0.7333
0.153 14.0201 2925 1.0651 0.7333
0.3668 15.0201 3120 0.5162 0.7333
0.2628 16.0201 3315 0.7342 0.8
0.5817 17.0201 3510 0.4857 0.8667
0.7572 18.0201 3705 0.4588 0.8222
0.4044 19.0201 3900 1.1526 0.7111
0.5366 20.0201 4095 1.3653 0.6667
0.5051 21.0201 4290 0.6016 0.8222
0.3338 22.0201 4485 0.7427 0.8222
0.7609 23.0201 4680 0.6195 0.8667
0.2756 24.0201 4875 0.9184 0.8
0.2845 25.0201 5070 1.1702 0.7333
0.3813 26.0201 5265 0.7869 0.8444
0.1671 27.0201 5460 0.8510 0.8222
0.3501 28.0201 5655 1.3892 0.7111
0.5373 29.0201 5850 1.8380 0.6667
0.0084 30.0201 6045 0.9612 0.8222
0.6559 31.0201 6240 1.5071 0.7111
0.5831 32.0201 6435 1.5741 0.7556
0.0256 33.0201 6630 1.6111 0.7333
0.4461 34.0201 6825 0.7689 0.8222
0.0307 35.0201 7020 0.6986 0.8667
0.0007 36.0201 7215 1.4886 0.7778
0.0672 37.0201 7410 1.4161 0.7778
0.0106 38.0201 7605 1.1670 0.8
0.0009 39.0201 7800 1.0056 0.8222
0.5082 40.0201 7995 0.8138 0.8222
0.143 41.0201 8190 0.9801 0.8444
0.3104 42.0201 8385 1.0276 0.8
0.2251 43.0201 8580 1.2434 0.7556
0.2293 44.0201 8775 1.2502 0.7778
0.1605 45.0201 8970 1.0620 0.7778
0.1764 46.0201 9165 1.3246 0.7556
0.0096 47.0201 9360 1.3971 0.7778
0.0002 48.0201 9555 1.3673 0.7778
0.001 49.0149 9700 1.4292 0.7556

Framework versions

  • Transformers 4.46.2
  • Pytorch 2.0.1+cu117
  • Datasets 3.0.1
  • Tokenizers 0.20.0
Downloads last month
237
Safetensors
Model size
304M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.

Model tree for beingbatman/9_mae_4

Finetuned
(60)
this model