---
license: other
base_model: HuggingFaceH4/zephyr-7b-gemma-sft-v0.1
tags:
- trl
- dpo
- generated_from_trainer
model-index:
- name: zephyr-gemma-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. -->

# zephyr-gemma-dpo

This model is a fine-tuned version of [HuggingFaceH4/zephyr-7b-gemma-sft-v0.1](https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-sft-v0.1) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4446
- Rewards/chosen: -1.7969
- Rewards/rejected: -3.6406
- Rewards/accuracies: 0.7447
- Rewards/margins: 1.8359
- Logps/rejected: -480.0
- Logps/chosen: -388.0
- Logits/rejected: 84.0
- Logits/chosen: 82.0

## 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: 5e-07
- train_batch_size: 2
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- total_eval_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- lr_scheduler_warmup_steps: 100
- num_epochs: 2

### Training results

| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
| 0.5324        | 0.95  | 100  | 0.4852          | -1.1641        | -2.4688          | 0.7340             | 1.3047          | -456.0         | -376.0       | 94.0            | 92.0          |
| 0.1822        | 1.9   | 200  | 0.4446          | -1.7969        | -3.6406          | 0.7447             | 1.8359          | -480.0         | -388.0       | 84.0            | 82.0          |


### Framework versions

- Transformers 4.38.0
- Pytorch 2.1.1+cu121
- Datasets 2.16.1
- Tokenizers 0.15.2