Hugging Face's logo

seoulsky-field
/
KoBART_base_v2-trial

Text2Text Generation
Transformers
TensorBoard
Safetensors
bart
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
KoBART_base_v2-trial / README.md
Kyungmin Jeon
End of training
3ee4c7b
preview code
|
raw
history blame
2.7 kB
metadata
license: mit
base_model: gogamza/kobart-base-v2
tags:
  - generated_from_trainer
model-index:
  - name: KoBART_base_v2-trial
    results: []

KoBART_base_v2-trial

This model is a fine-tuned version of gogamza/kobart-base-v2 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1815

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: 0.0005
  • train_batch_size: 64
  • eval_batch_size: 64
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 20
  • num_epochs: 3
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
2.4147 0.11 50 0.5490
0.5457 0.22 100 0.4810
0.4642 0.32 150 0.3971
0.4364 0.43 200 0.3955
0.4111 0.54 250 0.3851
0.3888 0.65 300 0.3438
0.3586 0.76 350 0.3290
0.3304 0.87 400 0.3201
0.3337 0.97 450 0.2992
0.2677 1.08 500 0.3161
0.2576 1.19 550 0.2981
0.2467 1.3 600 0.2846
0.2369 1.41 650 0.2674
0.226 1.52 700 0.2529
0.2204 1.62 750 0.2446
0.204 1.73 800 0.2400
0.2071 1.84 850 0.2262
0.1911 1.95 900 0.2153
0.1591 2.06 950 0.2121
0.1338 2.16 1000 0.2090
0.1312 2.27 1050 0.1986
0.1336 2.38 1100 0.1947
0.1205 2.49 1150 0.1903
0.1162 2.6 1200 0.1867
0.1187 2.71 1250 0.1840
0.1171 2.81 1300 0.1821
0.1149 2.92 1350 0.1815

Framework versions

  • Transformers 4.36.0
  • Pytorch 2.0.1+cu117
  • Datasets 2.15.0
  • Tokenizers 0.15.0