metadata
license: other
library_name: peft
tags:
- llama-factory
- lora
- generated_from_trainer
base_model: /data1/model/llama2/meta-llama/Llama2-13b
model-index:
- name: news_commentary_es
results: []
news_commentary_es
This model is a fine-tuned version of /data1/model/llama2/meta-llama/Llama2-13b on the news_commentary_es dataset. It achieves the following results on the evaluation set:
- Loss: 0.5704
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 32
- total_eval_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 20
- num_epochs: 5.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.6296 | 0.25 | 200 | 0.6344 |
| 0.6272 | 0.5 | 400 | 0.6124 |
| 0.6167 | 0.75 | 600 | 0.6020 |
| 0.5882 | 1.0 | 800 | 0.5940 |
| 0.5522 | 1.25 | 1000 | 0.5889 |
| 0.5919 | 1.51 | 1200 | 0.5836 |
| 0.6038 | 1.76 | 1400 | 0.5801 |
| 0.5882 | 2.01 | 1600 | 0.5773 |
| 0.5302 | 2.26 | 1800 | 0.5754 |
| 0.5331 | 2.51 | 2000 | 0.5732 |
| 0.5669 | 2.76 | 2200 | 0.5714 |
| 0.5551 | 3.01 | 2400 | 0.5706 |
| 0.5499 | 3.26 | 2600 | 0.5716 |
| 0.545 | 3.51 | 2800 | 0.5704 |
| 0.5253 | 3.76 | 3000 | 0.5704 |
| 0.5418 | 4.02 | 3200 | 0.5702 |
| 0.5289 | 4.27 | 3400 | 0.5704 |
| 0.4983 | 4.52 | 3600 | 0.5706 |
| 0.5371 | 4.77 | 3800 | 0.5705 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.2.1
- Datasets 2.18.0
- Tokenizers 0.15.2