metadata
license: gemma
base_model: google/gemma-7b
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
- trl
- sft
- generated_from_trainer
datasets:
- GAIR/lima
model-index:
- name: gemma-lima
results: []
gemma-lima
This model is a fine-tuned version of google/gemma-7b on the GAIR/lima dataset. It achieves the following results on the evaluation set:
- Loss: 2.7259
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: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 16
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 10.4256 | 0.91 | 5 | 47.0001 |
| 6.0419 | 2.0 | 11 | 43.9691 |
| 5.2838 | 2.91 | 16 | 40.7857 |
| 4.8705 | 4.0 | 22 | 33.9282 |
| 4.196 | 4.91 | 27 | 17.5336 |
| 3.0724 | 6.0 | 33 | 2.7088 |
| 2.1966 | 6.91 | 38 | 2.7434 |
| 2.1116 | 8.0 | 44 | 2.7265 |
| 2.0641 | 8.91 | 49 | 2.7168 |
| 2.0467 | 9.09 | 50 | 2.7259 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.1.2
- Datasets 2.14.6
- Tokenizers 0.15.2