---
license: mit
library_name: peft
tags:
- trl
- dpo
- generated_from_trainer
base_model: HuggingFaceH4/zephyr-7b-beta
model-index:
- name: WeniGPT-2.6.1-Zephyr-7B-0.3-reduction-QA-1.0.1_DPO
  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-2.6.1-Zephyr-7B-0.3-reduction-QA-1.0.1_DPO

This model is a fine-tuned version of [HuggingFaceH4/zephyr-7b-beta](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6539
- Rewards/chosen: 0.9200
- Rewards/rejected: -1.4403
- Rewards/accuracies: 0.0562
- Rewards/margins: 2.3603
- Logps/rejected: -10.8805
- Logps/chosen: -5.1910
- Logits/rejected: -2.3586
- Logits/chosen: -2.3589

## 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: 0.0005
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- 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
- training_steps: 89
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
| 4.3155        | 0.56  | 50   | 0.7680          | -0.3150        | -1.9142          | 0.0500             | 1.5992          | -11.3544       | -6.4260      | -2.2698         | -2.2706       |


### Framework versions

- PEFT 0.8.2
- Transformers 4.38.2
- Pytorch 2.1.0+cu118
- Datasets 2.17.1
- Tokenizers 0.15.2