README.md

---
license: apache-2.0
datasets:
- stingning/ultrachat
- TIGER-Lab/MathInstruct
- ise-uiuc/Magicoder-Evol-Instruct-110K
- OpenAssistant/oasst2
- teknium/openhermes
- bigcode/commitpackft
- Open-Orca/SlimOrca
- ise-uiuc/Magicoder-OSS-Instruct-75K
language:
- en
library_name: transformers
base_model:
- mllmTeam/PhoneLM-0.5B
---
PhoneLM-0.5B-Instruct is a 0.5 billion parameter decoder-only language model.

## Usage
```python
from transformers import AutoTokenizer, AutoModelForCausalLM

model_name = 'mllmTeam/PhoneLM-0.5B-Instruct'
question = "Hello, who are you?"
prompt = [{"role": "user", "content": question}]

model = AutoModelForCausalLM.from_pretrained(model_name, device_map='cuda', trust_remote_code=True)

tokenizer = AutoTokenizer.from_pretrained(model_name)
input_text = tokenizer.apply_chat_template(prompt, tokenize=False, add_generation_prompt=True)

inp = tokenizer(input_text, return_tensors="pt")
inp = {k: v.to('cuda') for k, v in inp.items()}
out = model.generate(**inp, 
                     max_length=256,
                     do_sample=True,
                     temperature=0.7,
                     top_p=0.7
                     )
text = tokenizer.decode(out[0], skip_special_tokens=True)
print(text)
```


## Model Details

* **Developed by**: mllmTeam
* **Model type**: `PhoneLM 0.5B` models are auto-regressive language models based on the transformer decoder architecture.
* **Language(s)**: English
* **Paper**: [PhoneLM Technical Report]()
* **Library**: [PhoneLM](https://github.com/UbiquitousLearning/PhoneLM)

### Model Architecture

The model is a decoder-only transformer architecture with the following modifications:

| Hidden Size | Layers | Heads | Sequence Length |
|-------------|--------|-------|-----------------|
| 1024        | 24     | 16    | 2048            |

* **Position Embeddings**: Rotary Position Embeddings ([Su et al., 2021](https://arxiv.org/abs/2104.09864)) applied to the first 25% of head embedding dimensions for improved throughput following [Black et al. (2022)](https://arxiv.org/pdf/2204.06745.pdf). PhoneLM quantized the sin and cos values in Rotary Position Embeddings to 8-bit integers.
* **Normalization**: LayerNorm ([Ba et al., 2016](https://arxiv.org/abs/1607.06450)) with learned bias terms as opposed to RMSNorm ([Zhang & Sennrich, 2019](https://arxiv.org/abs/1910.07467)).
* **Biases**: We remove all bias terms from the feed-forward networks and multi-head self-attention layers, except for the biases of the query, key, and value projections ([Bai et al., 2023](https://arxiv.org/abs/2309.16609)).
* **ReLU Activation Function**: ReLU([Glorot et al., 2011](https://proceedings.mlr.press/v15/glorot11a/glorot11a.pdf)) activation functions are adopted in feed-forward networks.
* **Tokenizer**: We use the SmolLM([Allal et al., 2024](https://huggingface.co/blog/smollm))'s tokenizer with a vocabulary size of 49,152. 

## License
* This repository is released under the [Apache-2.0](https://huggingface.co/mllmTeam/PhoneLM-0.5B-Instruct/blob/main/LICENSE) License.、

## Citation
```
@misc{yi2024phonelmanefficientcapablesmall,
      title={PhoneLM:an Efficient and Capable Small Language Model Family through Principled Pre-training}, 
      author={Rongjie Yi and Xiang Li and Weikai Xie and Zhenyan Lu and Chenghua Wang and Ao Zhou and Shangguang Wang and Xiwen Zhang and Mengwei Xu},
      year={2024},
      eprint={2411.05046},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2411.05046}, 
}
```