---
license: apache-2.0
library_name: peft
tags:
- alignment-handbook
- trl
- dpo
- generated_from_trainer
base_model: mistralai/Mistral-7B-v0.1
datasets:
- HuggingFaceH4/ultrafeedback_binarized
model-index:
- name: zephyr-7b
  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-7b

This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-qlora](https://huggingface.co/alignment-handbook/zephyr-7b-sft-qlora) on the HuggingFaceH4/ultrafeedback_binarized dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6789
- Rewards/chosen: -0.5482
- Rewards/rejected: -0.8623
- Rewards/accuracies: 0.3591
- Rewards/margins: 0.3141
- Logps/rejected: -161.6313
- Logps/chosen: -123.7209
- Logits/rejected: 1.4916
- Logits/chosen: 1.3712
- Use Label: 17581.0469
- Pred Label: 2490.9524

## 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-06
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- total_eval_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1

### Training results

| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen | Use Label  | Pred Label |
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|:----------:|:----------:|
| 0.6685        | 0.1   | 100  | 0.6684          | -0.0306        | -0.0936          | 0.3353             | 0.0631          | -84.7626       | -71.9572     | -2.0796         | -2.1097       | 1856.0     | 0.0        |
| 0.676         | 0.21  | 200  | 0.6717          | -0.3729        | -0.4956          | 0.3214             | 0.1227          | -124.9563      | -106.1906    | -1.6889         | -1.7319       | 3898.4443  | 61.5556    |
| 0.6728        | 0.31  | 300  | 0.6784          | -0.4712        | -0.7059          | 0.3373             | 0.2347          | -145.9853      | -116.0199    | -0.6762         | -0.7414       | 5793.0317  | 270.9683   |
| 0.6715        | 0.42  | 400  | 0.6812          | -0.4462        | -0.7352          | 0.3552             | 0.2890          | -148.9146      | -113.5210    | 0.7648          | 0.6420        | 7595.3174  | 572.6826   |
| 0.6744        | 0.52  | 500  | 0.6722          | -0.5121        | -0.7576          | 0.3413             | 0.2455          | -151.1573      | -120.1133    | 0.7128          | 0.6149        | 9378.1592  | 893.8412   |
| 0.6784        | 0.63  | 600  | 0.6792          | -0.5107        | -0.8136          | 0.3512             | 0.3028          | -156.7531      | -119.9755    | 0.9939          | 0.8860        | 11169.8096 | 1206.1904  |
| 0.6783        | 0.73  | 700  | 0.6756          | -0.6634        | -0.9598          | 0.3671             | 0.2964          | -171.3761      | -135.2395    | 1.2995          | 1.1927        | 12921.4766 | 1558.5238  |
| 0.6776        | 0.84  | 800  | 0.6801          | -0.5500        | -0.8628          | 0.3532             | 0.3128          | -161.6791      | -123.9010    | 1.4789          | 1.3586        | 14683.5078 | 1900.4921  |
| 0.6751        | 0.94  | 900  | 0.6790          | -0.5476        | -0.8618          | 0.3571             | 0.3143          | -161.5806      | -123.6563    | 1.4905          | 1.3693        | 16436.9844 | 2251.0159  |


### Framework versions

- PEFT 0.7.1
- Transformers 4.38.2
- Pytorch 2.1.1+cu121
- Datasets 2.14.6
- Tokenizers 0.15.2