metadata
license: gemma
library_name: peft
tags:
- generated_from_trainer
base_model: google/gemma-2b
metrics:
- accuracy
- f1
- precision
- recall
model-index:
- name: gemma-finetuned-spam
results: []
gemma-finetuned-spam
This model is a fine-tuned version of google/gemma-2b on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0545
- Accuracy: 0.994
- F1: 0.994
- Precision: 0.994
- Recall: 0.994
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: 3.589634237431302e-05
- train_batch_size: 8
- eval_batch_size: 2
- seed: 3
- 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 | F1 | Precision | Recall |
|---|---|---|---|---|---|---|---|
| 0.0491 | 1.0 | 1125 | 0.0730 | 0.989 | 0.9890 | 0.9861 | 0.992 |
| 0.0253 | 2.0 | 2250 | 0.0516 | 0.99 | 0.9900 | 0.9920 | 0.988 |
| 0.006 | 3.0 | 3375 | 0.0546 | 0.993 | 0.9930 | 0.9920 | 0.994 |
| 0.0 | 4.0 | 4500 | 0.0545 | 0.994 | 0.994 | 0.994 | 0.994 |
| 0.0001 | 5.0 | 5625 | 0.0554 | 0.994 | 0.994 | 0.994 | 0.994 |
Framework versions
- PEFT 0.11.1
- Transformers 4.41.0
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1