callmesan's picture
End of training
c0aa6c0 verified
metadata
library_name: transformers
license: mit
base_model: ai4bharat/indic-bert
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: indic-bert-abusive-comments-ta
    results: []

indic-bert-abusive-comments-ta

This model is a fine-tuned version of ai4bharat/indic-bert on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.1108
  • Accuracy: 0.6896
  • Precision: 0.2983
  • Recall: 0.2872
  • F1: 0.2740

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: 128
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 64
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Accuracy Precision Recall F1
1.1508 1.0 93 1.2143 0.6216 0.1373 0.1440 0.1253
1.1999 2.0 186 1.1389 0.6337 0.1632 0.1656 0.1545
1.0445 3.0 279 1.0623 0.6815 0.2057 0.2219 0.2109
0.882 4.0 372 1.0181 0.6996 0.3264 0.3092 0.2976
0.7892 5.0 465 1.0408 0.7036 0.3327 0.3097 0.3011
0.7381 6.0 558 1.1007 0.7103 0.3317 0.3060 0.3057
0.6175 7.0 651 1.2323 0.7070 0.3241 0.3014 0.3037
0.49 8.0 744 1.2686 0.6956 0.4590 0.3093 0.3164
0.4576 9.0 837 1.3347 0.6761 0.3285 0.3200 0.3199
0.3436 10.0 930 1.3501 0.6673 0.3363 0.3167 0.3204

Framework versions

  • Transformers 4.45.1
  • Pytorch 2.4.0
  • Datasets 3.0.1
  • Tokenizers 0.20.0