metadata
license: gemma
library_name: peft
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
base_model: google/gemma-7b
datasets:
- llama-duo/synth_classification_dataset_dedup
model-index:
- name: gemma7b-classification-gpt4o-100k
results: []
gemma7b-classification-gpt4o-100k
This model is a fine-tuned version of google/gemma-7b on the llama-duo/synth_classification_dataset_dedup dataset. It achieves the following results on the evaluation set:
- Loss: 4.4668
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: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 3
- gradient_accumulation_steps: 2
- total_train_batch_size: 24
- total_eval_batch_size: 12
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.0589 | 1.0 | 319 | 1.7605 |
| 1.0 | 2.0 | 638 | 1.7306 |
| 0.8741 | 3.0 | 957 | 1.8115 |
| 0.7735 | 4.0 | 1276 | 1.9239 |
| 0.6599 | 5.0 | 1595 | 2.0617 |
| 0.5764 | 6.0 | 1914 | 2.3235 |
| 0.4817 | 7.0 | 2233 | 2.6759 |
| 0.4016 | 8.0 | 2552 | 3.0133 |
| 0.327 | 9.0 | 2871 | 3.4537 |
| 0.2814 | 10.0 | 3190 | 3.8273 |
| 0.2539 | 11.0 | 3509 | 4.1355 |
| 0.233 | 12.0 | 3828 | 4.3549 |
| 0.2281 | 13.0 | 4147 | 4.4559 |
| 0.2274 | 14.0 | 4466 | 4.4672 |
| 0.2251 | 15.0 | 4785 | 4.4668 |
Framework versions
- PEFT 0.11.1
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1