---
library_name: transformers
license: apache-2.0
datasets:
- liswei/zhtw-news-and-articles-2B
base_model:
- liswei/Taiwan-ELM
- apple/OpenELM-270M
language:
- zh
metrics:
- perplexity
pipeline_tag: text-generation
---

<center>
    <img src="https://huggingface.co/liswei/Taiwan-ELM/resolve/main/Taiwan%20ELM%20Logo.jpeg" alt="Efficient LLM for Taiwan">
</center>

> Efficient LLM for Taiwan

# Taiwan ELM

Taiwan ELM is a family of Efficient LLMs for Taiwan base on [apple/OpenELM](https://huggingface.co/apple/OpenELM).
The project aims to provide an efficient model for researchers without access to large-scale computing resources.

The model is trained using a custom fork of [LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory) on 2B Traditional Chinese tokens and 500K instruction samples.
We will extend the model to train on larger data sets and different base models if there is sufficient demand.

## What is being released?

We release both pre-trained base models and instruction tuned variants with 270M and 1.1B parameters.
Along with the model, datasets used to train the base and instruction-tuned models are also released.

List of released models:
* Taiwan-ELM-270M
* Taiwan-ELM-1_1B
* Taiwan-ELM-270M-Instruct
* Taiwan-ELM-1_1B-Instruct

List of released datasets:
* [liswei/Taiwan-Text-Excellence-2B](https://huggingface.co/datasets/liswei/Taiwan-Text-Excellence-2B)
* [liswei/PromptPair-TW](https://huggingface.co/datasets/liswei/PromptPair-TW)

## Usage Examples

We adapt the LLaMA2 template:
```jinja2
<s>[INST] <<SYS>>
{{ system_prompt }}
<</SYS>>

{{ user_message }} [/INST]
```

The model could be load via `AutoModelForCausalLM` with `trust_remote_code=True`:
```python
taiwanelm_270m = AutoModelForCausalLM.from_pretrained("liswei/Taiwan-ELM-270M", trust_remote_code=True)
```

We also support additional generation methods and speculative generation, please find reference at [OpenELM#usage](https://huggingface.co/apple/OpenELM#usage).