metadata
base_model: microsoft/Phi-3-mini-128k-instruct
library_name: peft
license: mit
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: phi-3-mini-LoRA
results: []
phi-3-mini-LoRA
This model is a fine-tuned version of microsoft/Phi-3-mini-128k-instruct on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3525
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: 2
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 2
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.3351 | 0.17 | 500 | 0.3755 |
| 0.3312 | 0.34 | 1000 | 0.3644 |
| 0.3079 | 0.51 | 1500 | 0.3597 |
| 0.3195 | 0.68 | 2000 | 0.3577 |
| 0.3218 | 0.85 | 2500 | 0.3557 |
| 0.3034 | 1.02 | 3000 | 0.3553 |
| 0.296 | 1.19 | 3500 | 0.3543 |
| 0.3175 | 1.36 | 4000 | 0.3539 |
| 0.3257 | 1.53 | 4500 | 0.3533 |
| 0.3263 | 1.7 | 5000 | 0.3526 |
| 0.3209 | 1.87 | 5500 | 0.3522 |
| 0.3221 | 2.04 | 6000 | 0.3528 |
| 0.2927 | 2.21 | 6500 | 0.3526 |
| 0.2922 | 2.38 | 7000 | 0.3527 |
| 0.2968 | 2.55 | 7500 | 0.3525 |
| 0.2968 | 2.72 | 8000 | 0.3526 |
| 0.3094 | 2.89 | 8500 | 0.3525 |
Framework versions
- PEFT 0.8.2
- Transformers 4.38.0
- Pytorch 2.2.1
- Datasets 2.17.0
- Tokenizers 0.15.2