twitter_trainer

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

  • Loss: 0.7924
  • Accuracy: 86.8509
  • P: 102.7555
  • R: 100.3442
  • F1: 101.5355

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: 4
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 16
  • 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
  • num_epochs: 7

Training results

Training Loss Epoch Step Validation Loss Accuracy P R F1
3.6245 1.0 597 0.4451 84.1709 99.8149 103.1569 101.4584
1.8323 2.0 1194 0.3794 86.0972 102.3665 100.0625 101.2014
1.233 3.0 1791 0.3715 87.5209 100.9234 102.3408 101.6272
0.9132 4.0 2388 0.5171 87.1022 102.4483 100.4991 101.4643
0.6928 5.0 2985 0.6683 86.9347 102.6526 100.5006 101.5652
0.4037 6.0 3582 0.7477 87.3534 101.8838 101.3746 101.6286
0.3334 6.9891 4172 0.7924 86.8509 102.7555 100.3442 101.5355

Framework versions

  • Transformers 4.47.1
  • Pytorch 2.5.1+cu121
  • Datasets 3.2.0
  • Tokenizers 0.21.0
Downloads last month
11
Safetensors
Model size
108M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.

Model tree for Mikimi/twitter_trainer

Finetuned
(2179)
this model