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: alpaca_no_sys
results: []
alpaca_no_sys
This model is a fine-tuned version of /data1/model/llama2/meta-llama/Llama2-13b on the alpaca_no_sys dataset. It achieves the following results on the evaluation set:
- Loss: 1.0082
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: 0.0001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 16
- total_eval_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 20
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.0429 | 0.07 | 200 | 1.0298 |
| 1.0505 | 0.14 | 400 | 1.0238 |
| 1.044 | 0.22 | 600 | 1.0194 |
| 1.0169 | 0.29 | 800 | 1.0172 |
| 1.02 | 0.36 | 1000 | 1.0154 |
| 0.9492 | 0.43 | 1200 | 1.0134 |
| 1.0051 | 0.51 | 1400 | 1.0117 |
| 1.0469 | 0.58 | 1600 | 1.0106 |
| 0.9994 | 0.65 | 1800 | 1.0094 |
| 1.0141 | 0.72 | 2000 | 1.0082 |
| 1.0891 | 0.8 | 2200 | 1.0073 |
| 1.0141 | 0.87 | 2400 | 1.0063 |
| 1.0002 | 0.94 | 2600 | 1.0059 |
| 0.9686 | 1.01 | 2800 | 1.0086 |
| 0.9767 | 1.09 | 3000 | 1.0141 |
| 0.9494 | 1.16 | 3200 | 1.0160 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.2.1
- Datasets 2.18.0
- Tokenizers 0.15.2