Downloads · 30 days
8
2% of all-time downloads
opus-mt-zh-de-tuned-Tatoeba-small
ykliu1892/opus-mt-zh-de-tuned-Tatoeba-small
opus-mt-zh-de-tuned-Tatoeba-small is a machine learning model from ykliu1892. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.
should probably proofread and complete it, then remove this comment. --
All-time downloads
412
Public
Repo size
930 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin310 MB · 99%
At a glance
- Library
- transformers
- License
- apache-2.0
- Model type
- marian
- Access
- Public
- Created
- Mar 2, 2022
- Updated
- Jan 2, 2022
- SHA
- a9a3017a
Model card
From the Hugging Face model README
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
opus-mt-zh-de-tuned-Tatoeba-small
This model is a fine-tuned version of Helsinki-NLP/opus-mt-zh-de on a refined dataset of Tatoeba German - Chinese corpus https://github.com/Helsinki-NLP/Tatoeba-Challenge/blob/master/data/README.md. It achieves the following results on the evaluation set:
- Loss: 2.2703
- Bleu: 16.504
- Gen Len: 16.6531
Model description
More information needed
Intended uses & limitations
Prefix used during fine-tuning: "将中文翻译成德语". This prefix is also recommended in prediction.
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: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 2.7229 | 0.24 | 16000 | 2.5605 | 14.1956 | 16.2206 |
| 2.5988 | 0.49 | 32000 | 2.4447 | 14.8619 | 16.2726 |
| 2.515 | 0.73 | 48000 | 2.3817 | 15.3212 | 16.2823 |
| 2.4683 | 0.97 | 64000 | 2.3367 | 15.9043 | 16.7138 |
| 2.3873 | 1.22 | 80000 | 2.3115 | 16.1037 | 16.6369 |
| 2.3792 | 1.46 | 96000 | 2.2919 | 16.2957 | 16.6304 |
| 2.3626 | 1.7 | 112000 | 2.2790 | 16.2995 | 16.6235 |
| 2.3353 | 1.95 | 128000 | 2.2703 | 16.504 | 16.6531 |
Framework versions
- Transformers 4.15.0
- Pytorch 1.10.0+cu111
- Datasets 1.17.0
- Tokenizers 0.10.3
- Library
- transformers
- Type
- marian
- License
- apache-2.0
- Created
- Mar 2, 2022
- Updated
- Jan 2, 2022