metadata
library_name: peft
tags:
- generated_from_trainer
metrics:
- accuracy
base_model: meta-llama/Llama-2-7b-hf
model-index:
- name: llama2-7B-ReqORNot
results: []
llama2-7B-ReqORNot
This model is a fine-tuned version of meta-llama/Llama-2-7b-hf on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2597
- Accuracy: 0.8970
- Weighted precision: 0.8971
- Weighted recall: 0.8970
- Weighted f1: 0.8971
- Macro precision: 0.8969
- Macro recall: 0.8971
- Macro f1: 0.8970
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: 2e-05
- train_batch_size: 24
- eval_batch_size: 24
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Weighted precision | Weighted recall | Weighted f1 | Macro precision | Macro recall | Macro f1 |
|---|---|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 237 | 0.4807 | 0.7896 | 0.7895 | 0.7896 | 0.7895 | 0.7894 | 0.7891 | 0.7892 |
| No log | 2.0 | 474 | 0.3167 | 0.8605 | 0.8605 | 0.8605 | 0.8605 | 0.8604 | 0.8604 | 0.8604 |
| 0.5108 | 3.0 | 711 | 0.2709 | 0.8860 | 0.8869 | 0.8860 | 0.8860 | 0.8862 | 0.8866 | 0.8860 |
| 0.5108 | 4.0 | 948 | 0.2704 | 0.8880 | 0.8889 | 0.8880 | 0.8879 | 0.8894 | 0.8871 | 0.8876 |
| 0.1829 | 5.0 | 1185 | 0.2597 | 0.8970 | 0.8971 | 0.8970 | 0.8971 | 0.8969 | 0.8971 | 0.8970 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2