---
license: llama2
library_name: peft
tags:
- unsloth
- generated_from_trainer
base_model: meta-llama/Llama-2-13b-hf
model-index:
- name: llama_2_13b_MetaMathQA_40K_ortho
  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. -->

# llama_2_13b_MetaMathQA_40K_ortho

This model is a fine-tuned version of [meta-llama/Llama-2-13b-hf](https://huggingface.co/meta-llama/Llama-2-13b-hf) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4660

## 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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 0.02
- num_epochs: 1

### Training results

| Training Loss | Epoch  | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 0.7973        | 0.0211 | 13   | 0.6231          |
| 0.587         | 0.0421 | 26   | 0.5915          |
| 0.5618        | 0.0632 | 39   | 0.5752          |
| 0.5435        | 0.0842 | 52   | 0.5648          |
| 0.5329        | 0.1053 | 65   | 0.5548          |
| 0.5189        | 0.1264 | 78   | 0.5486          |
| 0.5265        | 0.1474 | 91   | 0.5423          |
| 0.507         | 0.1685 | 104  | 0.5367          |
| 0.5282        | 0.1896 | 117  | 0.5328          |
| 0.517         | 0.2106 | 130  | 0.5292          |
| 0.4969        | 0.2317 | 143  | 0.5249          |
| 0.481         | 0.2527 | 156  | 0.5207          |
| 0.4847        | 0.2738 | 169  | 0.5166          |
| 0.4876        | 0.2949 | 182  | 0.5137          |
| 0.4776        | 0.3159 | 195  | 0.5105          |
| 0.4857        | 0.3370 | 208  | 0.5071          |
| 0.4888        | 0.3580 | 221  | 0.5037          |
| 0.4833        | 0.3791 | 234  | 0.5016          |
| 0.4695        | 0.4002 | 247  | 0.4988          |
| 0.476         | 0.4212 | 260  | 0.4948          |
| 0.4873        | 0.4423 | 273  | 0.4931          |
| 0.4709        | 0.4633 | 286  | 0.4913          |
| 0.4765        | 0.4844 | 299  | 0.4892          |
| 0.4767        | 0.5055 | 312  | 0.4871          |
| 0.4599        | 0.5265 | 325  | 0.4856          |
| 0.4687        | 0.5476 | 338  | 0.4838          |
| 0.4525        | 0.5687 | 351  | 0.4826          |
| 0.4627        | 0.5897 | 364  | 0.4797          |
| 0.4535        | 0.6108 | 377  | 0.4781          |
| 0.4574        | 0.6318 | 390  | 0.4766          |
| 0.4597        | 0.6529 | 403  | 0.4758          |
| 0.4651        | 0.6740 | 416  | 0.4746          |
| 0.456         | 0.6950 | 429  | 0.4732          |
| 0.4557        | 0.7161 | 442  | 0.4720          |
| 0.4505        | 0.7371 | 455  | 0.4708          |
| 0.4446        | 0.7582 | 468  | 0.4699          |
| 0.4574        | 0.7793 | 481  | 0.4688          |
| 0.4459        | 0.8003 | 494  | 0.4678          |
| 0.4394        | 0.8214 | 507  | 0.4674          |
| 0.4587        | 0.8424 | 520  | 0.4672          |
| 0.4536        | 0.8635 | 533  | 0.4667          |
| 0.4488        | 0.8846 | 546  | 0.4666          |
| 0.4504        | 0.9056 | 559  | 0.4662          |
| 0.4428        | 0.9267 | 572  | 0.4661          |
| 0.4505        | 0.9478 | 585  | 0.4660          |
| 0.444         | 0.9688 | 598  | 0.4659          |
| 0.4496        | 0.9899 | 611  | 0.4660          |


### Framework versions

- PEFT 0.7.1
- Transformers 4.40.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1