---
base_model: microsoft/Phi-3-mini-4k-instruct
library_name: peft
license: mit
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: phi3-mini-QLoRA
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# phi3-mini-QLoRA

This model is a fine-tuned version of [microsoft/Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5590

## 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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 32
- 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 |
|:-------------:|:------:|:----:|:---------------:|
| 0.9007        | 0.1810 | 100  | 0.6361          |
| 0.606         | 0.3619 | 200  | 0.5831          |
| 0.5789        | 0.5429 | 300  | 0.5755          |
| 0.5797        | 0.7238 | 400  | 0.5709          |
| 0.5653        | 0.9048 | 500  | 0.5687          |
| 0.573         | 1.0857 | 600  | 0.5663          |
| 0.5577        | 1.2667 | 700  | 0.5648          |
| 0.5586        | 1.4476 | 800  | 0.5636          |
| 0.572         | 1.6286 | 900  | 0.5623          |
| 0.5594        | 1.8095 | 1000 | 0.5616          |
| 0.5585        | 1.9905 | 1100 | 0.5614          |
| 0.5486        | 2.1715 | 1200 | 0.5609          |
| 0.5614        | 2.3524 | 1300 | 0.5599          |
| 0.5515        | 2.5334 | 1400 | 0.5597          |
| 0.5525        | 2.7143 | 1500 | 0.5593          |
| 0.5545        | 2.8953 | 1600 | 0.5590          |


### Framework versions

- PEFT 0.13.0
- Transformers 4.45.1
- Pytorch 2.2.2
- Datasets 3.0.1
- Tokenizers 0.20.0