metadata
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
datasets:
- generator
base_model: NousResearch/Llama-2-7b-hf
model-index:
- name: output_llama2_instruct
results: []
output_llama2_instruct
This model is a fine-tuned version of NousResearch/Llama-2-7b-hf on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 1.6746
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: 0.0002
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_steps: 20
- training_steps: 500
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.9193 | 0.0 | 20 | 1.8124 |
| 1.8847 | 0.01 | 40 | 1.7835 |
| 1.813 | 0.01 | 60 | 1.7710 |
| 1.9026 | 0.01 | 80 | 1.7605 |
| 1.8352 | 0.01 | 100 | 1.7449 |
| 1.7327 | 0.02 | 120 | 1.7100 |
| 1.8625 | 0.02 | 140 | 1.7055 |
| 1.9379 | 0.02 | 160 | 1.7008 |
| 1.8597 | 0.02 | 180 | 1.6971 |
| 1.8703 | 0.03 | 200 | 1.6943 |
| 1.7749 | 0.03 | 220 | 1.6902 |
| 1.7645 | 0.03 | 240 | 1.6898 |
| 1.7894 | 0.04 | 260 | 1.6886 |
| 1.8492 | 0.04 | 280 | 1.6870 |
| 1.7331 | 0.04 | 300 | 1.6841 |
| 1.7278 | 0.04 | 320 | 1.6830 |
| 1.6963 | 0.05 | 340 | 1.6826 |
| 1.8226 | 0.05 | 360 | 1.6813 |
| 1.8246 | 0.05 | 380 | 1.6797 |
| 1.8577 | 0.05 | 400 | 1.6778 |
| 1.6691 | 0.06 | 420 | 1.6779 |
| 1.7001 | 0.06 | 440 | 1.6774 |
| 1.9008 | 0.06 | 460 | 1.6770 |
| 1.7856 | 0.07 | 480 | 1.6733 |
| 1.7674 | 0.07 | 500 | 1.6746 |
Framework versions
- PEFT 0.8.2
- Transformers 4.37.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1