metadata
base_model: google/gemma-2-2b-it
datasets:
- GaetanMichelet/chat-60_ft_task-3_auto
library_name: peft
license: gemma
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
model-index:
- name: Gemma-2-2B_task-3_60-samples_config-1_full_auto
results: []
Gemma-2-2B_task-3_60-samples_config-1_full_auto
This model is a fine-tuned version of google/gemma-2-2b-it on the GaetanMichelet/chat-60_ft_task-3_auto dataset. It achieves the following results on the evaluation set:
- Loss: 0.9570
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: 0.0001
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 8
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.3456 | 0.8696 | 5 | 1.3519 |
| 1.3545 | 1.9130 | 11 | 1.2756 |
| 1.2474 | 2.9565 | 17 | 1.1765 |
| 1.1006 | 4.0 | 23 | 1.0925 |
| 0.995 | 4.8696 | 28 | 1.0253 |
| 0.9268 | 5.9130 | 34 | 0.9853 |
| 0.8574 | 6.9565 | 40 | 0.9666 |
| 0.8398 | 8.0 | 46 | 0.9581 |
| 0.8249 | 8.8696 | 51 | 0.9570 |
| 0.8095 | 9.9130 | 57 | 0.9601 |
| 0.7215 | 10.9565 | 63 | 0.9763 |
| 0.6764 | 12.0 | 69 | 0.9987 |
| 0.6692 | 12.8696 | 74 | 1.0207 |
| 0.5679 | 13.9130 | 80 | 1.0891 |
| 0.5212 | 14.9565 | 86 | 1.1447 |
| 0.5149 | 16.0 | 92 | 1.2058 |
Framework versions
- PEFT 0.12.0
- Transformers 4.44.0
- Pytorch 2.1.2+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1