README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - para_pat
metrics:
  - bleu
model-index:
  - name: opus-mt-ru-uk-finetuned-ru-to-uk
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: para_pat
          type: para_pat
          config: ru-uk
          split: train
          args: ru-uk
        metrics:
          - name: Bleu
            type: bleu
            value: 70.1044

opus-mt-ru-uk-finetuned-ru-to-uk

This model is a fine-tuned version of Helsinki-NLP/opus-mt-ru-uk on the para_pat dataset. It achieves the following results on the evaluation set:

• Loss: 0.5266
• Bleu: 70.1044
• Gen Len: 122.6206

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
0.9218	1.0	2150	0.6965	64.5544	123.2615
0.7195	2.0	4300	0.5750	68.3283	122.4044
0.6363	3.0	6450	0.5380	69.7152	122.6875
0.6086	4.0	8600	0.5266	70.1044	122.6206

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3