|
--- |
|
license: cc-by-nc-sa-4.0 |
|
language: |
|
- en |
|
- ko |
|
library_name: transformers |
|
--- |
|
|
|
|
|
|
|
Model Architecture |
|
|
|
OOM-13B_01 is an language model that uses an optimized transformer architecture based on Llama-2. |
|
|
|
|
|
## Model description |
|
|
|
Based on "beomi/llama-2-koen-13b" |
|
|
|
## Intended uses & limitations |
|
|
|
T.B.D. |
|
|
|
## Training and evaluation data |
|
|
|
T.B.D. |
|
|
|
## Training procedure |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 2e-04 |
|
- train_batch_size: 2 |
|
- eval_batch_size: 8 |
|
- seed: 24 |
|
- gradient_accumulation_steps: 1 |
|
- total_train_batch_size: |
|
- num_epochs: 2.0 |
|
|
|
### Training results |
|
|
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.37.2 |
|
- Pytorch 2.2.0+cu118 |
|
- Datasets 2.16.1 |
|
- Tokenizers 0.15.1 |
