step_val_25_gemma-2-2b_hs2_iter1_sftsd2
This model is a fine-tuned version of google/gemma-2-2b on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.0950
- Num Input Tokens Seen: 1470072
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: 8e-06
- train_batch_size: 8
- eval_batch_size: 16
- seed: 2
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_ratio: 0.05
- training_steps: 25
Training results
| Training Loss | Epoch | Step | Validation Loss | Input Tokens Seen |
|---|---|---|---|---|
| No log | 0 | 0 | 1.3956 | 0 |
| 1.188 | 0.0511 | 5 | 1.2111 | 295904 |
| 1.1647 | 0.1021 | 10 | 1.1523 | 591104 |
| 1.1386 | 0.1532 | 15 | 1.1210 | 883280 |
| 1.0904 | 0.2042 | 20 | 1.1020 | 1177976 |
| 1.0895 | 0.2553 | 25 | 1.0950 | 1470072 |
Framework versions
- Transformers 4.44.0
- Pytorch 2.4.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1