ViViT_WLASL_200_epochs_p20

This model is a fine-tuned version of google/vivit-b-16x2-kinetics400 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 4.4366
  • Top 1 Accuracy: 0.2725
  • Top 5 Accuracy: 0.5651
  • Top 10 Accuracy: 0.6599
  • Accuracy: 0.2727
  • Precision: 0.2554
  • Recall: 0.2727
  • F1: 0.2419

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: 357200
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Top 1 Accuracy Top 5 Accuracy Top 10 Accuracy Accuracy Precision Recall F1
30.7847 0.005 1786 7.6710 0.0003 0.0013 0.0043 0.0003 0.0000 0.0003 0.0000
30.4187 1.0050 3572 7.6049 0.0020 0.0056 0.0115 0.0020 0.0003 0.0020 0.0004
29.7427 2.0050 5358 7.4861 0.0036 0.0161 0.0271 0.0036 0.0009 0.0036 0.0008
28.4534 3.0050 7145 7.2492 0.0115 0.0375 0.0605 0.0115 0.0049 0.0115 0.0044
26.8561 4.005 8931 6.9297 0.0317 0.0766 0.1208 0.0319 0.0115 0.0319 0.0125
24.997 5.0050 10717 6.5686 0.0460 0.1292 0.1946 0.0460 0.0212 0.0460 0.0210
22.8381 6.0050 12503 6.2049 0.0746 0.1859 0.2605 0.0748 0.0386 0.0748 0.0382
20.56 7.0050 14290 5.8050 0.1021 0.2554 0.3463 0.1021 0.0591 0.1021 0.0594
18.323 8.005 16076 5.4230 0.1318 0.3113 0.4101 0.1325 0.0720 0.1325 0.0773
15.752 9.0050 17862 5.0506 0.1622 0.3598 0.4847 0.1619 0.0954 0.1619 0.1013
13.1613 10.0050 19648 4.6589 0.1933 0.4321 0.5564 0.1941 0.1294 0.1941 0.1352
10.1883 11.0050 21435 4.3067 0.2263 0.4946 0.6139 0.2263 0.1633 0.2263 0.1684
8.1852 12.005 23221 4.0124 0.2474 0.5314 0.6527 0.2472 0.1886 0.2472 0.1927
5.88 13.0050 25007 3.7567 0.2651 0.5720 0.6885 0.2651 0.2188 0.2651 0.2186
4.5173 14.0050 26793 3.6064 0.2763 0.5927 0.7125 0.2763 0.2440 0.2763 0.2373
2.851 15.0050 28580 3.5264 0.2776 0.6083 0.7196 0.2778 0.2534 0.2778 0.2434
2.2493 16.005 30366 3.5001 0.2725 0.6085 0.7183 0.2725 0.2471 0.2725 0.2383
1.9096 17.0050 32152 3.4933 0.2809 0.6014 0.7127 0.2814 0.2487 0.2814 0.2432
1.7109 18.0050 33938 3.5175 0.2776 0.5919 0.7068 0.2778 0.2517 0.2778 0.2418
1.4479 19.0050 35725 3.5182 0.2835 0.5986 0.7153 0.2835 0.2580 0.2835 0.2478
1.2076 20.005 37511 3.5383 0.2842 0.5981 0.7071 0.2842 0.2600 0.2842 0.2493
1.4093 21.0050 39297 3.6272 0.2699 0.5899 0.6918 0.2702 0.2458 0.2702 0.2359
1.3122 22.0050 41083 3.6791 0.2725 0.5822 0.6951 0.2722 0.2420 0.2722 0.2364
1.4774 23.0050 42870 3.7163 0.2758 0.5812 0.6941 0.2760 0.2548 0.2760 0.2417
1.1338 24.005 44656 3.7577 0.2676 0.5817 0.6897 0.2674 0.2471 0.2674 0.2346
1.1446 25.0050 46442 3.7662 0.2745 0.5958 0.7074 0.2745 0.2533 0.2745 0.2430
1.0876 26.0050 48228 3.8689 0.2789 0.5781 0.6851 0.2786 0.2525 0.2786 0.2435
1.0755 27.0050 50015 3.8659 0.2773 0.5855 0.7010 0.2773 0.2626 0.2773 0.2465
1.1092 28.005 51801 3.9151 0.2786 0.5891 0.6966 0.2789 0.2540 0.2789 0.2448
0.9877 29.0050 53587 4.0145 0.2832 0.5853 0.6902 0.2832 0.2596 0.2832 0.2480
0.946 30.0050 55373 3.9974 0.2763 0.5884 0.6915 0.2771 0.2626 0.2771 0.2478
1.1572 31.0050 57160 4.0120 0.2868 0.5787 0.6849 0.2870 0.2664 0.2870 0.2545
1.0663 32.005 58946 4.1235 0.2763 0.5702 0.6734 0.2763 0.2552 0.2763 0.2446
1.085 33.0050 60732 4.1426 0.2715 0.5718 0.6782 0.2715 0.2528 0.2715 0.2403
1.3799 34.0050 62518 4.1017 0.2755 0.5743 0.6839 0.2755 0.2594 0.2755 0.2445
0.8419 35.0050 64305 4.1769 0.2794 0.5728 0.6836 0.2794 0.2620 0.2794 0.2487
1.1308 36.005 66091 4.1333 0.2796 0.5804 0.6890 0.2794 0.2569 0.2794 0.2468
0.9785 37.0050 67877 4.2391 0.2771 0.5592 0.6655 0.2771 0.2617 0.2771 0.2471
0.923 38.0050 69663 4.2274 0.2804 0.5769 0.6683 0.2804 0.2583 0.2804 0.2484
0.9857 39.0050 71450 4.2033 0.2814 0.5830 0.6862 0.2819 0.2610 0.2819 0.2498
0.7679 40.005 73236 4.1983 0.2845 0.5861 0.6834 0.2845 0.2640 0.2845 0.2518
0.8991 41.0050 75022 4.2099 0.2812 0.5873 0.6882 0.2812 0.2713 0.2812 0.2533
1.1176 42.0050 76808 4.3419 0.2768 0.5687 0.6734 0.2766 0.2596 0.2766 0.2455
1.2777 43.0050 78595 4.3104 0.2783 0.5774 0.6742 0.2781 0.2610 0.2781 0.2469
0.8072 44.005 80381 4.3708 0.2778 0.5638 0.6616 0.2781 0.2686 0.2781 0.2500
1.1258 45.0050 82167 4.3517 0.2799 0.5707 0.6731 0.2801 0.2688 0.2801 0.2510
0.9476 46.0050 83953 4.3712 0.2814 0.5684 0.6757 0.2812 0.2583 0.2812 0.2473
1.0757 47.0050 85740 4.4280 0.2679 0.5641 0.6668 0.2676 0.2519 0.2676 0.2372
0.7244 48.005 87526 4.4374 0.2676 0.5677 0.6685 0.2674 0.2529 0.2674 0.2388
0.9193 49.0050 89312 4.4014 0.2748 0.5636 0.6662 0.2748 0.2617 0.2748 0.2467
0.8372 50.0050 91098 4.4763 0.2735 0.5585 0.6545 0.2735 0.2555 0.2735 0.2429
0.7598 51.0050 92885 4.4366 0.2725 0.5651 0.6599 0.2727 0.2554 0.2727 0.2419

Framework versions

  • Transformers 4.46.1
  • Pytorch 2.5.1+cu124
  • Datasets 3.1.0
  • Tokenizers 0.20.1
Downloads last month
9
Safetensors
Model size
90.2M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported third-party Inference Providers, and the model is not deployed on the HF Inference API.

Model tree for Shawon16/ViViT_WLASL_200_epochs_p20

Finetuned
(60)
this model