bensapir's picture
update model card README.md
bcb037d
|
raw
history blame
3.9 kB
metadata
license: apache-2.0
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: pixel-base-finetune-sent
    results: []

pixel-base-finetune-sent

This model is a fine-tuned version of Team-PIXEL/pixel-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 4.4999
  • Accuracy: 0.3702
  • Qwk: 0.6030
  • Mae: 1.9328

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: 32
  • 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_steps: 200
  • training_steps: 50000

Training results

Training Loss Epoch Step Validation Loss Accuracy Qwk Mae
2.0132 0.58 1000 2.0384 0.2788 0.5991 2.0788
1.9246 1.17 2000 1.9957 0.2979 0.6192 2.0294
1.9128 1.75 3000 1.9440 0.3218 0.5979 2.0585
1.792 2.33 4000 1.9316 0.3271 0.6068 1.9224
1.7771 2.92 5000 1.8539 0.3557 0.6218 1.8866
1.6715 3.5 6000 1.8448 0.3665 0.6286 1.8692
1.5602 4.08 7000 1.8855 0.3692 0.5773 2.0124
1.5562 4.67 8000 1.8120 0.3837 0.6033 1.8889
1.4078 5.25 9000 1.8713 0.3884 0.6243 1.8740
1.4296 5.83 10000 1.8595 0.3865 0.6071 1.9174
1.269 6.42 11000 1.9294 0.3989 0.6260 1.8328
1.3173 7.0 12000 1.9485 0.3796 0.6218 1.9100
1.1227 7.58 13000 2.0674 0.3893 0.6181 1.8724
0.8926 8.17 14000 2.3098 0.3819 0.6120 1.8919
0.9693 8.75 15000 2.2659 0.3862 0.6195 1.8196
0.7348 9.33 16000 2.5195 0.3828 0.6031 1.9008
0.8165 9.92 17000 2.4748 0.3824 0.6125 1.8863
0.6071 10.5 18000 2.7954 0.3718 0.6138 1.8930
0.4994 11.09 19000 3.0574 0.3681 0.6113 1.9193
0.5026 11.67 20000 3.0452 0.3550 0.5989 1.9430
0.3619 12.25 21000 3.2941 0.3576 0.5921 1.9748
0.3928 12.84 22000 3.4092 0.3561 0.6070 1.9112
0.2999 13.42 23000 3.5744 0.3661 0.6001 1.9334
0.3142 14.0 24000 3.5801 0.3703 0.6258 1.8979
0.2513 14.59 25000 3.7402 0.3699 0.6177 1.9004
0.1896 15.17 26000 3.9153 0.3752 0.6147 1.8882
0.2204 15.75 27000 3.9247 0.3747 0.6165 1.9239
0.1522 16.34 28000 4.1680 0.3703 0.5994 1.9439
0.1844 16.92 29000 4.2140 0.3752 0.6022 1.9122
0.1331 17.5 30000 4.4904 0.3736 0.5998 1.8850
0.1141 18.09 31000 4.4999 0.3702 0.6030 1.9328

Framework versions

  • Transformers 4.17.0
  • Pytorch 2.5.1
  • Datasets 3.6.0
  • Tokenizers 0.21.1