SeanLee97's picture
End of training
a7e3559 verified
metadata
library_name: peft
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
base_model: NousResearch/Llama-2-7b-hf
model-index:
  - name: billm-llama-7b-conll03-ner
    results: []

billm-llama-7b-conll03-ner

This model is a fine-tuned version of NousResearch/Llama-2-7b-hf on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1783
  • Precision: 0.9150
  • Recall: 0.9330
  • F1: 0.9239
  • Accuracy: 0.9851

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: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
0.0477 1.0 1756 0.0998 0.9116 0.9283 0.9199 0.9842
0.0201 2.0 3512 0.0986 0.9152 0.9251 0.9201 0.9842
0.0089 3.0 5268 0.1195 0.9128 0.9278 0.9202 0.9843
0.0025 4.0 7024 0.1564 0.9129 0.9341 0.9234 0.9851
0.0013 5.0 8780 0.1669 0.9140 0.9316 0.9227 0.9850
0.0006 6.0 10536 0.1736 0.9155 0.9328 0.9241 0.9852
0.0003 7.0 12292 0.1755 0.9144 0.9325 0.9233 0.9851
0.0003 8.0 14048 0.1782 0.9145 0.9328 0.9236 0.9851
0.0003 9.0 15804 0.1782 0.9144 0.9326 0.9234 0.9851
0.0002 10.0 17560 0.1783 0.9150 0.9330 0.9239 0.9851

Framework versions

  • PEFT 0.9.0
  • Transformers 4.38.2
  • Pytorch 2.0.1
  • Datasets 2.16.0
  • Tokenizers 0.15.0