---
license: mit
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
base_model: HuggingFaceH4/zephyr-7b-beta
datasets:
- generator
model-index:
- name: WeniGPT-2.2.3-Zephyr-7B-LLM_Base_2.0.3_SFT
  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. -->

# WeniGPT-2.2.3-Zephyr-7B-LLM_Base_2.0.3_SFT

This model is a fine-tuned version of [HuggingFaceH4/zephyr-7b-beta](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) on the generator dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4088

## 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: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.03
- training_steps: 337
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.5761        | 0.83  | 50   | 0.4671          |
| 0.4488        | 1.66  | 100  | 0.4409          |
| 0.4272        | 2.49  | 150  | 0.4293          |
| 0.4145        | 3.32  | 200  | 0.4219          |
| 0.3999        | 4.15  | 250  | 0.4155          |
| 0.3872        | 4.98  | 300  | 0.4088          |


### Framework versions

- PEFT 0.7.1
- Transformers 4.37.0.dev0
- Pytorch 2.1.0+cu118
- Datasets 2.16.1
- Tokenizers 0.15.0

# Open Portuguese LLM Leaderboard Evaluation Results  

Detailed results can be found [here](https://huggingface.co/datasets/eduagarcia-temp/llm_pt_leaderboard_raw_results/tree/main/Weni/WeniGPT-2.2.3-Zephyr-7B-LLM_Base_2.0.3_SFT) and on the [🚀 Open Portuguese LLM Leaderboard](https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard)

|          Metric          |  Value  |
|--------------------------|---------|
|Average                   |**45.86**|
|ENEM Challenge (No Images)|    23.58|
|BLUEX (No Images)         |    28.79|
|OAB Exams                 |    26.33|
|Assin2 RTE                |    87.01|
|Assin2 STS                |    28.33|
|FaQuAD NLI                |    44.66|
|HateBR Binary             |    66.91|
|PT Hate Speech Binary     |    56.25|
|tweetSentBR               |    50.88|