metadata
license: other
library_name: peft
tags:
- generated_from_trainer
base_model: google/gemma-2b
model-index:
- name: gemma-2b-prompt-dict_fix
results: []
gemma-2b-prompt-dict_fix
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: 0.4486
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: 0.0004
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- training_steps: 2000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.9304 | 0.38 | 100 | 1.1381 |
| 0.8056 | 0.75 | 200 | 1.0223 |
| 0.8532 | 1.13 | 300 | 0.9144 |
| 0.9387 | 1.5 | 400 | 0.8460 |
| 0.8744 | 1.88 | 500 | 0.7882 |
| 0.8939 | 2.26 | 600 | 0.7415 |
| 0.9265 | 2.63 | 700 | 0.7008 |
| 0.4869 | 3.01 | 800 | 0.6631 |
| 0.4609 | 3.38 | 900 | 0.6485 |
| 0.4909 | 3.76 | 1000 | 0.6063 |
| 0.5568 | 4.14 | 1100 | 0.5645 |
| 0.4727 | 4.51 | 1200 | 0.5412 |
| 0.5277 | 4.89 | 1300 | 0.5235 |
| 0.7115 | 5.26 | 1400 | 0.5122 |
| 0.5917 | 5.64 | 1500 | 0.4972 |
| 0.4643 | 6.02 | 1600 | 0.4777 |
| 0.275 | 6.39 | 1700 | 0.4750 |
| 0.4726 | 6.77 | 1800 | 0.4611 |
| 0.3297 | 7.14 | 1900 | 0.4496 |
| 0.3426 | 7.52 | 2000 | 0.4486 |
Framework versions
- PEFT 0.9.0
- Transformers 4.39.0.dev0
- Pytorch 2.2.1+cu121
- Datasets 2.17.1
- Tokenizers 0.15.2