Downloads · 30 days
14
25% of all-time downloads
ko-gangwon-nmt-v1
leadawon/ko-gangwon-nmt-v1
ko-gangwon-nmt-v1 is a text generation model from leadawon. Use it when you need the model to write or continue text. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
56
Public
Repo size
3 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin751 MB · 100%
At a glance
- Task
- Text Generation
- Library
- transformers
- Model type
- encoder-decoder
- Access
- Public
- Created
- Jan 26, 2023
- Updated
- Jan 26, 2023
- SHA
- 48ff1044
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. -->
jeolla-ko-nmt-v1
This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.199624
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: 24
- eval_batch_size: 24
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 96
- 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
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.525000 | 1.0 | 3017 | 0.424946 |
| 0.318700 | 2.0 | 6034 | 0.285191 |
| 0.244100 | 3.0 | 9051 | 0.237215 |
| 0.195900 | 4.0 | 12068 | 0.216691 |
| 0.160500 | 5.0 | 15085 | 0.203532 |
| 0.135400 | 6.0 | 18092 | 0.199624 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.13.1+cu116
- Tokenizers 0.13.2
- Task
- Text Generation
- Library
- transformers
- Type
- encoder-decoder
- Languages
- ko
- Created
- Jan 26, 2023
- Updated
- Jan 26, 2023