llama8_on_korean_conv-base-ltm
This model is a fine-tuned version of beomi/Llama-3-Open-Ko-8B-Instruct-preview on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1665
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: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 20
- total_train_batch_size: 20
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- training_steps: 2000
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
1.156 | 0.2563 | 100 | 1.1623 |
1.0573 | 0.5126 | 200 | 1.0621 |
0.9766 | 0.7688 | 300 | 1.0136 |
0.8888 | 1.0251 | 400 | 0.9799 |
0.7699 | 1.2814 | 500 | 0.9696 |
0.7575 | 1.5377 | 600 | 0.9534 |
0.7641 | 1.7940 | 700 | 0.9405 |
0.588 | 2.0502 | 800 | 0.9882 |
0.542 | 2.3065 | 900 | 0.9862 |
0.5494 | 2.5628 | 1000 | 0.9776 |
0.5314 | 2.8191 | 1100 | 0.9726 |
0.3521 | 3.0753 | 1200 | 1.0522 |
0.3379 | 3.3316 | 1300 | 1.0573 |
0.3875 | 3.5879 | 1400 | 1.0439 |
0.346 | 3.8442 | 1500 | 1.0578 |
0.2252 | 4.1005 | 1600 | 1.1292 |
0.2271 | 4.3567 | 1700 | 1.1488 |
0.2237 | 4.6130 | 1800 | 1.1526 |
0.2199 | 4.8693 | 1900 | 1.1563 |
0.1614 | 5.1256 | 2000 | 1.1665 |
Framework versions
- PEFT 0.11.1
- Transformers 4.41.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
- Downloads last month
- 0
Model tree for ghost613/llama8_on_korean_conv-base-ltm
Base model
beomi/Llama-3-Open-Ko-8B-Instruct-preview