---
license: apache-2.0
---

## Finetuning on [habana](https://habana.ai/) HPU

This model is a fine-tuned model based on [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the open source dataset [Open-Orca/SlimOrca](https://huggingface.co/datasets/Open-Orca/SlimOrca). And align it with DPO algorithm. For more details about our DPO blog, you can refer: [NeuralChat: Simplifying Supervised Instruction Fine-Tuning and Reinforcement Aligning](https://medium.com/intel-analytics-software/neuralchat-simplifying-supervised-instruction-fine-tuning-and-reinforcement-aligning-for-chatbots-d034bca44f69).

## Model date
Neural-chat-7b-v3 was trained between September and October, 2023.

## Evaluation

We use the [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness/tree/master) to measure the metrics that are adopted by [open_llm_leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).

| Model | Average ⬆️| ARC (25-s) ⬆️ | HellaSwag (10-s) ⬆️ | MMLU (5-s) ⬆️| TruthfulQA (MC) (0-s) ⬆️ |
| --- | --- | --- | --- | --- | --- |
|[mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) | 62.4 | 59.58  | 83.31  | 64.16  | 42.15 | 
| **Ours** | **67.92** | 66.29 | 83.28 | 62.11  | 60.02 |


## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 1e-04
- train_batch_size: 1
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 16
- total_train_batch_size: 64
- total_eval_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.02
- num_epochs: 2.0

## Inference with transformers

```shell
import transformers
model = transformers.AutoModelForCausalLM.from_pretrained(
  'Intel/neural-chat-7b-v3'
)
```

## Ethical Considerations and Limitations
neural-chat-7b-v3 can produce factually incorrect output, and should not be relied on to produce factually accurate information. neural-chat-7b-v3 was trained on [Open-Orca/SlimOrca](https://huggingface.co/datasets/Open-Orca/SlimOrca) based on [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1). Because of the limitations of the pretrained model and the finetuning datasets, it is possible that this model could generate lewd, biased or otherwise offensive outputs.

Therefore, before deploying any applications of neural-chat-7b-v3, developers should perform safety testing.

## Disclaimer

The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please cosult an attorney before using this model for commercial purposes.

## Organizations developing the model

The NeuralChat team with members from Intel/SATG/AIA/AIPT. Core team members: Kaokao Lv, Liang Lv, Chang Wang, Wenxin Zhang, Xuhui Ren, and Haihao Shen.

## Useful links
* Intel Neural Compressor [link](https://github.com/intel/neural-compressor)
* Intel Extension for Transformers [link](https://github.com/intel/intel-extension-for-transformers)
* Intel Extension for PyTorch [link](https://github.com/intel/intel-extension-for-pytorch)