metadata
license: mit
library_name: peft
tags:
- generated_from_trainer
base_model: microsoft/Phi-3-mini-128k-instruct
model-index:
- name: working
results: []
working
This model is a fine-tuned version of microsoft/Phi-3-mini-128k-instruct on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4364
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: 6
- eval_batch_size: 6
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 24
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.8501 | 1.0 | 3 | 2.2605 |
| 2.1387 | 2.0 | 6 | 1.7344 |
| 1.5826 | 3.0 | 9 | 1.3666 |
| 1.2187 | 4.0 | 12 | 1.0485 |
| 0.8879 | 5.0 | 15 | 0.7558 |
| 0.6134 | 6.0 | 18 | 0.5396 |
| 0.4343 | 7.0 | 21 | 0.4304 |
| 0.3557 | 8.0 | 24 | 0.3943 |
| 0.3205 | 9.0 | 27 | 0.3689 |
| 0.2947 | 10.0 | 30 | 0.3580 |
| 0.2727 | 11.0 | 33 | 0.3371 |
| 0.2506 | 12.0 | 36 | 0.3361 |
| 0.2291 | 13.0 | 39 | 0.3342 |
| 0.2098 | 14.0 | 42 | 0.3332 |
| 0.1911 | 15.0 | 45 | 0.3446 |
| 0.1761 | 16.0 | 48 | 0.3334 |
| 0.159 | 17.0 | 51 | 0.3453 |
| 0.1399 | 18.0 | 54 | 0.3540 |
| 0.124 | 19.0 | 57 | 0.3631 |
| 0.1123 | 20.0 | 60 | 0.3636 |
| 0.0992 | 21.0 | 63 | 0.3778 |
| 0.0862 | 22.0 | 66 | 0.3862 |
| 0.0783 | 23.0 | 69 | 0.3966 |
| 0.0704 | 24.0 | 72 | 0.4072 |
| 0.0627 | 25.0 | 75 | 0.4178 |
| 0.0582 | 26.0 | 78 | 0.4200 |
| 0.0553 | 27.0 | 81 | 0.4283 |
| 0.0521 | 28.0 | 84 | 0.4338 |
| 0.0505 | 29.0 | 87 | 0.4366 |
| 0.0494 | 30.0 | 90 | 0.4364 |
Framework versions
- PEFT 0.10.0
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2