---
license: mit
base_model: microsoft/Phi-3-mini-4k-instruct
tags:
- generated_from_trainer
model-index:
- name: PHI30512HMAB4H
  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. -->

# PHI30512HMAB4H

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: 0.0863

## 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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine_with_restarts
- lr_scheduler_warmup_steps: 80
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 4.7212        | 0.09  | 10   | 1.7615          |
| 0.8243        | 0.18  | 20   | 0.3615          |
| 0.3349        | 0.27  | 30   | 0.2864          |
| 0.3202        | 0.36  | 40   | 0.2662          |
| 0.6542        | 0.45  | 50   | 0.2339          |
| 0.2334        | 0.54  | 60   | 0.2159          |
| 0.4715        | 0.63  | 70   | 0.7325          |
| 1.1442        | 0.73  | 80   | 0.9402          |
| 2.8468        | 0.82  | 90   | 4.9739          |
| 2.3962        | 0.91  | 100  | 0.9736          |
| 0.6771        | 1.0   | 110  | 0.4388          |
| 0.4592        | 1.09  | 120  | 0.3687          |
| 0.3319        | 1.18  | 130  | 0.2208          |
| 0.248         | 1.27  | 140  | 0.1911          |
| 0.2015        | 1.36  | 150  | 0.1909          |
| 0.2016        | 1.45  | 160  | 0.2083          |
| 0.2028        | 1.54  | 170  | 0.1676          |
| 0.1686        | 1.63  | 180  | 0.1661          |
| 0.1546        | 1.72  | 190  | 0.1469          |
| 0.1708        | 1.81  | 200  | 0.1622          |
| 0.1494        | 1.9   | 210  | 0.1380          |
| 0.1412        | 1.99  | 220  | 0.1445          |
| 0.1389        | 2.08  | 230  | 0.1363          |
| 0.1386        | 2.18  | 240  | 0.1266          |
| 0.1307        | 2.27  | 250  | 0.1289          |
| 0.1261        | 2.36  | 260  | 0.1213          |
| 0.123         | 2.45  | 270  | 0.1142          |
| 0.1098        | 2.54  | 280  | 0.1067          |
| 0.1063        | 2.63  | 290  | 0.1014          |
| 0.0987        | 2.72  | 300  | 0.0947          |
| 0.0955        | 2.81  | 310  | 0.0904          |
| 0.0946        | 2.9   | 320  | 0.0882          |
| 0.0854        | 2.99  | 330  | 0.0863          |


### Framework versions

- Transformers 4.36.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.14.0