README.md

---
license: apache-2.0
---

<!-- 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. -->

This model is a fine-tuned model for Chat based on [mosaicml/mpt-7b](https://huggingface.co/mosaicml/mpt-7b) with **max_seq_lenght=2048** on various open source dataset. For the details of the used dataset, please refer to [Intel/neural-chat-dataset-v1-1](https://huggingface.co/datasets/Intel/neural-chat-dataset-v1-1).

## Model date
Neural-chat-7b-v1.1 was trained between June and July 2023.

## Evaluation
We use the same evaluation metrics as [open_llm_leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) which uses [Eleuther AI Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness/tree/master), a unified framework to test generative language models on a large number of different evaluation tasks.

| Model | Average ⬆️| ARC (25-s) ⬆️ | HellaSwag (10-s) ⬆️ | MMLU (5-s) ⬆️| TruthfulQA (MC) (0-s) ⬆️ |
| --- | --- | --- | --- | --- | --- |
|[mosaicml/mpt-7b](https://huggingface.co/mosaicml/mpt-7b)| 47.4  | 47.61 | 77.56 | 31 | 33.43 |
| [mosaicml/mpt-7b-chat](https://huggingface.co/mosaicml/mpt-7b-chat) | **49.95** | 46.5 | 75.55 | 37.60 | 40.17 |
| **Ours** | **51.41** | 50.09 | 76.69 | 38.79 | 40.07 |

### Bias evaluation

Following the blog [evaluating-llm-bias](https://huggingface.co/blog/evaluating-llm-bias), we select 10000 samples randomly from [allenai/real-toxicity-prompts](https://huggingface.co/datasets/allenai/real-toxicity-prompts) to evaluate toxicity bias in Language Models

| Model | Toxicity Rito ↓|
| --- | --- |
|[mosaicml/mpt-7b](https://huggingface.co/mosaicml/mpt-7b)| 0.027 |
| **Ours** | 0.0264 |


## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 8
- 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: 3.0

## Inference with transformers

```shell
import transformers
model = transformers.AutoModelForCausalLM.from_pretrained(
  'Intel/neural-chat-7b-v1-1',
  trust_remote_code=True
)
```

## Inference with INT8
Follow the instructions [link](https://github.com/intel/intel-extension-for-transformers/tree/main/examples/huggingface/pytorch/text-generation/quantization) to install the necessary dependencies. Use the below command to quantize the model using Intel Neural Compressor [link](https://github.com/intel/neural-compressor) and accelerate the inference.

```shell
python run_generation.py \
    --model Intel/neural-chat-7b-v1-1 \
    --quantize \
    --sq \
    --alpha 0.95 \
    --ipex
```

### Examples

- code generation
![code-generation](examples/code.png)

- summarization
![summarization](examples/summarization.png)

- trip
![trip](examples/trip.png)

## Ethical Considerations and Limitations
neural-chat-7b-v1-1 can produce factually incorrect output, and should not be relied on to produce factually accurate information. neural-chat-7b-v1-1 was trained on various instruction/chat datasets based on [mosaicml/mpt-7b](https://huggingface.co/mosaicml/mpt-7b). 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-v1-1, 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)