Edit model card

kd4_opus-mt-ko-en

This model is a fine-tuned version of Helsinki-NLP/opus-mt-ko-en on the kde4 dataset. It achieves the following results on the evaluation set:

  • Loss: 1.3924
  • Bleu: 32.1162

See translation_ko_en.ipynb

Model description

More information needed

Intended uses & limitations

More information needed

Usage

You can use this model directly with a pipeline for translation language modeling:

>>> from transformers import pipeline
>>> translator = pipeline('translation',model='chunwoolee0/kd4_opus-mt-ko-e')
>>> translator("์ ์‹ฌ ์‹์‚ฌ ํ›„์— ์‚ฐ์ฑ…๊ฐ€์ž.")

[{'translation_text': "Let's go for a walk after noon."}]

>>> translator("์ด ๊ฐ•์ขŒ๋Š” ํ—ˆ๊น…ํŽ˜์ด์Šค๊ฐ€ ๋งŒ๋“  ๊ฑฐ์•ผ.")
[{'translation_text': 'This is a course by Huggingspace.'}]

>>> translator("์˜ค๋Š˜์€ ๋Šฆ๊ฒŒ ์ผ์–ด๋‚ฌ๋‹ค.")
[{'translation_text': "I'm up late today."}]

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: 64
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 3

Training results

Step Training Loss 500 1.858500 1000 1.781400 1500 1.715200 2000 1.678100 2500 1.546600 3000 1.488700 3500 1.503500 4000 1.455100 4500 1.419100 5000 1.393400 5500 1.357100 6000 1.339400 TrainOutput(global_step=6474, training_loss=1.532715692246148, metrics={'train_runtime': 1035.7775, 'train_samples_per_second': 199.957, 'train_steps_per_second': 6.25, 'total_flos': 2551308264603648.0, 'train_loss': 1.532715692246148, 'epoch': 3.0})

Framework versions

  • Transformers 4.31.0
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.4
  • Tokenizers 0.13.3
Downloads last month
5
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for chunwoolee0/kd4_opus-mt-ko-en

Finetuned
(7)
this model

Dataset used to train chunwoolee0/kd4_opus-mt-ko-en

Evaluation results