README.md

---
license: apache-2.0
language:
- en
tags:
- zen
- zenlm
- hanzo-ai
- multilingual
- translation
- CJK
pipeline_tag: text-generation
library_name: transformers
base_model: zenlm/zen-next-80b-instruct
---

# Zen Multilingual

> **Parameters**: 32B | **Architecture**: Zen 3 Architecture | **Context**: 128K | **License**: Apache 2.0 | **Released**: 2024-12-01

Multilingual generation across 30+ languages: English, Chinese, Japanese, Korean, Arabic, Spanish, French, German, Portuguese, Russian, and more.

Strong at cross-lingual reasoning, code-switching, and multilingual instruction following.

Base weights: [zenlm/zen-next-80b-instruct](https://huggingface.co/zenlm/zen-next-80b-instruct)

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("zenlm/zen-next-80b-instruct", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained("zenlm/zen-next-80b-instruct")
messages = [{"role": "user", "content": "Your domain-specific prompt here"}]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(text, return_tensors="pt").to(model.device)
output = model.generate(**inputs, max_new_tokens=1024)
print(tokenizer.decode(output[0][inputs.input_ids.shape[-1]:], skip_special_tokens=True))
```

---
## The Zen LM Family

Joint research between **Hanzo AI** (Techstars '17), **Zoo Labs Foundation** (501c3), and **Lux Partners Limited**.

All weights Apache 2.0. Download, run locally, fine-tune, deploy commercially.

[HuggingFace](https://huggingface.co/zenlm) · [Chat](https://hanzo.chat) · [API](https://api.hanzo.ai) · [Docs](https://zenlm.org)