|
--- |
|
license: other |
|
library_name: peft |
|
tags: |
|
- generated_from_trainer |
|
base_model: beomi/Llama-3-Open-Ko-8B-Instruct-preview |
|
model-index: |
|
- name: llama8_on_korean_summary |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# llama8_on_korean_summary |
|
|
|
This model is a fine-tuned version of [beomi/Llama-3-Open-Ko-8B-Instruct-preview](https://huggingface.co/beomi/Llama-3-Open-Ko-8B-Instruct-preview) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.8536 |
|
|
|
## 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: 2 |
|
- eval_batch_size: 2 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 5 |
|
- total_train_batch_size: 10 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 50 |
|
- training_steps: 760 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 1.7767 | 0.26 | 20 | 1.5096 | |
|
| 1.2382 | 0.53 | 40 | 1.0459 | |
|
| 0.9915 | 0.79 | 60 | 0.9451 | |
|
| 0.9126 | 1.05 | 80 | 0.8893 | |
|
| 0.8501 | 1.32 | 100 | 0.8515 | |
|
| 0.8113 | 1.58 | 120 | 0.8192 | |
|
| 0.8019 | 1.84 | 140 | 0.7939 | |
|
| 0.7239 | 2.11 | 160 | 0.7795 | |
|
| 0.6621 | 2.37 | 180 | 0.7594 | |
|
| 0.6457 | 2.63 | 200 | 0.7433 | |
|
| 0.6417 | 2.89 | 220 | 0.7281 | |
|
| 0.5929 | 3.16 | 240 | 0.7305 | |
|
| 0.5245 | 3.42 | 260 | 0.7242 | |
|
| 0.5291 | 3.68 | 280 | 0.7154 | |
|
| 0.528 | 3.95 | 300 | 0.7109 | |
|
| 0.4696 | 4.21 | 320 | 0.7257 | |
|
| 0.4474 | 4.47 | 340 | 0.7251 | |
|
| 0.4572 | 4.74 | 360 | 0.7252 | |
|
| 0.4391 | 5.0 | 380 | 0.7202 | |
|
| 0.3794 | 5.26 | 400 | 0.7462 | |
|
| 0.3771 | 5.53 | 420 | 0.7568 | |
|
| 0.3754 | 5.79 | 440 | 0.7453 | |
|
| 0.3739 | 6.05 | 460 | 0.7597 | |
|
| 0.3179 | 6.32 | 480 | 0.7803 | |
|
| 0.3328 | 6.58 | 500 | 0.7699 | |
|
| 0.3259 | 6.84 | 520 | 0.7710 | |
|
| 0.3014 | 7.11 | 540 | 0.8083 | |
|
| 0.2759 | 7.37 | 560 | 0.8017 | |
|
| 0.2758 | 7.63 | 580 | 0.7954 | |
|
| 0.2798 | 7.89 | 600 | 0.8003 | |
|
| 0.2545 | 8.16 | 620 | 0.8325 | |
|
| 0.2451 | 8.42 | 640 | 0.8282 | |
|
| 0.2355 | 8.68 | 660 | 0.8318 | |
|
| 0.2382 | 8.95 | 680 | 0.8300 | |
|
| 0.2256 | 9.21 | 700 | 0.8544 | |
|
| 0.212 | 9.47 | 720 | 0.8532 | |
|
| 0.2108 | 9.74 | 740 | 0.8529 | |
|
| 0.2125 | 10.0 | 760 | 0.8536 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.8.2 |
|
- Transformers 4.38.2 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.17.0 |
|
- Tokenizers 0.15.0 |