d1mitriz
Final model pretrained on the greek-longformer after 6 epochs
7f6d761
|
raw
history blame
1.23 kB
metadata
language:
  - el
tags:
  - text
  - language-modeling
metrics:
  - accuracy
model-index:
  - name: greek-media-longformer-base-4096-uncased
    results: []

Greek Media Longformer

This model is a second-stage pretrained version of dimitriz/greek-longformer-base-4096 trained on the dimitriz/greek_media_texts dataset. It achieves the following results on the evaluation set:

  • Loss: 1.1424
  • Accuracy: 0.7574

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: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 64
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 6.0

Training results

Framework versions

  • Transformers 4.28.0.dev0
  • Pytorch 2.0.0+cu118
  • Datasets 2.11.0
  • Tokenizers 0.13.2