Hugging Face's logo

waelChafei
/
RobertaTuned

Text Classification
Transformers
Safetensors
bert
Generated from Trainer
Inference Endpoints
Model card Files Community
RobertaTuned / README.md
waelChafei's picture
waelChafei
/kaggle/working/TTC4900Model
052159c verified
preview code
|
raw
history blame contribute delete
8.18 kB
metadata
license: mit
base_model: roberta-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
  - precision
  - recall
model-index:
  - name: TTC4900Model
    results: []

TTC4900Model

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

  • Loss: 0.5463
  • Accuracy: 0.8381
  • F1: 0.7260
  • Precision: 0.7596
  • Recall: 0.7047

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: 64
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 3
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy F1 Precision Recall
1.4742 0.04 50 1.3501 0.5138 0.1534 0.1802 0.2019
1.3854 0.08 100 1.2962 0.5492 0.1444 0.1464 0.1712
1.2044 0.12 150 1.1191 0.6135 0.1935 0.2802 0.2438
1.1934 0.16 200 1.1296 0.6328 0.2580 0.3115 0.2543
1.0589 0.2 250 0.9550 0.6909 0.3103 0.4154 0.3417
1.0216 0.24 300 0.9304 0.7036 0.3720 0.3727 0.3799
0.9372 0.28 350 0.8312 0.7185 0.4050 0.7342 0.3833
0.8818 0.33 400 0.8910 0.7197 0.4773 0.5010 0.4834
0.8957 0.37 450 0.7688 0.7512 0.4872 0.5636 0.4631
0.8311 0.41 500 0.7380 0.7638 0.5687 0.6187 0.5541
0.7595 0.45 550 0.7932 0.7435 0.5502 0.5931 0.5648
0.7677 0.49 600 0.7167 0.7746 0.5999 0.6017 0.6063
0.7386 0.53 650 0.6960 0.7776 0.5608 0.6716 0.5294
0.7731 0.57 700 0.6524 0.7973 0.6137 0.6762 0.5856
0.6949 0.61 750 0.6898 0.7880 0.6079 0.6804 0.5660
0.6982 0.65 800 0.6676 0.7882 0.6021 0.6450 0.5858
0.6805 0.69 850 0.6533 0.7989 0.6216 0.7456 0.6235
0.7633 0.73 900 0.7205 0.7796 0.5835 0.6257 0.6041
0.7712 0.77 950 0.7247 0.7838 0.5740 0.6818 0.5463
0.6768 0.81 1000 0.6328 0.8051 0.6448 0.7470 0.6220
0.671 0.85 1050 0.7261 0.7767 0.5529 0.6892 0.5497
0.6413 0.89 1100 0.6102 0.8100 0.6359 0.6886 0.6147
0.6398 0.93 1150 0.6881 0.7857 0.5860 0.8209 0.5796
0.6588 0.98 1200 0.6264 0.8056 0.6416 0.6564 0.6405
0.5952 1.02 1250 0.6763 0.8119 0.6407 0.6848 0.6231
0.5342 1.06 1300 0.7901 0.7930 0.5880 0.6963 0.5642
0.5187 1.1 1350 0.6499 0.8073 0.6686 0.7048 0.6669
0.5655 1.14 1400 0.6369 0.8061 0.6759 0.6753 0.6796
0.5522 1.18 1450 0.6168 0.8089 0.6496 0.6933 0.6619
0.5308 1.22 1500 0.6293 0.8173 0.6627 0.7965 0.6479
0.628 1.26 1550 0.6275 0.8086 0.6672 0.7533 0.6413
0.4993 1.3 1600 0.6286 0.8150 0.6753 0.7726 0.6521
0.5557 1.34 1650 0.6392 0.8145 0.6380 0.7942 0.6101
0.5315 1.38 1700 0.6072 0.8222 0.6863 0.7386 0.6572
0.5766 1.42 1750 0.6300 0.8120 0.6318 0.8268 0.6121
0.5225 1.46 1800 0.5962 0.8195 0.6903 0.7529 0.6648
0.5074 1.5 1850 0.6217 0.8196 0.6622 0.7711 0.6262
0.5613 1.54 1900 0.5924 0.8246 0.7053 0.7634 0.6756
0.5097 1.59 1950 0.5728 0.8233 0.6791 0.7823 0.6391
0.5001 1.63 2000 0.5828 0.8300 0.7151 0.7483 0.6918
0.5144 1.67 2050 0.5746 0.8256 0.6997 0.7606 0.6727
0.5462 1.71 2100 0.5792 0.8229 0.6932 0.7236 0.6943
0.5252 1.75 2150 0.5827 0.8266 0.6926 0.7896 0.6572
0.5369 1.79 2200 0.6034 0.8142 0.6867 0.7556 0.6558
0.5144 1.83 2250 0.5748 0.8280 0.7103 0.7445 0.6937
0.545 1.87 2300 0.5671 0.8243 0.6942 0.7573 0.6910
0.5151 1.91 2350 0.5685 0.8292 0.6961 0.7770 0.6678
0.5268 1.95 2400 0.5470 0.8318 0.7171 0.7650 0.6974
0.509 1.99 2450 0.5448 0.8336 0.7126 0.7736 0.6885
0.4062 2.03 2500 0.6064 0.8329 0.6949 0.7580 0.6716
0.452 2.07 2550 0.5852 0.8291 0.7058 0.7678 0.6852
0.488 2.11 2600 0.5741 0.8283 0.6993 0.7521 0.6897
0.4459 2.15 2650 0.5606 0.8319 0.7094 0.7706 0.6829
0.4588 2.2 2700 0.5834 0.8253 0.7106 0.7520 0.6914
0.4325 2.24 2750 0.5672 0.8299 0.7149 0.7590 0.6895
0.4182 2.28 2800 0.5661 0.8316 0.7190 0.7527 0.7071
0.4524 2.32 2850 0.5719 0.8329 0.7176 0.7715 0.6936
0.4078 2.36 2900 0.5574 0.8308 0.7149 0.7479 0.7035
0.3654 2.4 2950 0.5658 0.8353 0.7188 0.7521 0.7002
0.4095 2.44 3000 0.5608 0.8335 0.7213 0.7524 0.7019
0.379 2.48 3050 0.5666 0.8365 0.7211 0.7739 0.6949
0.3939 2.52 3100 0.5711 0.8296 0.7203 0.7621 0.6954
0.4039 2.56 3150 0.5748 0.8341 0.7213 0.7641 0.6942
0.4034 2.6 3200 0.5533 0.8348 0.7282 0.7593 0.7065
0.4412 2.64 3250 0.5490 0.8357 0.7250 0.7805 0.6944
0.386 2.68 3300 0.5675 0.8353 0.7296 0.7605 0.7093
0.4298 2.72 3350 0.5525 0.8344 0.7320 0.7583 0.7140
0.384 2.76 3400 0.5629 0.8355 0.7240 0.7734 0.7004
0.3909 2.8 3450 0.5586 0.8344 0.7269 0.7562 0.7132
0.3975 2.85 3500 0.5538 0.8356 0.7253 0.7679 0.7022
0.3906 2.89 3550 0.5566 0.8332 0.7246 0.7570 0.7091
0.3707 2.93 3600 0.5575 0.8359 0.7290 0.7619 0.7095
0.3995 2.97 3650 0.5529 0.8345 0.7296 0.7563 0.7131

Framework versions

  • Transformers 4.38.2
  • Pytorch 2.1.2
  • Datasets 2.1.0
  • Tokenizers 0.15.2