Aya Expanse 32B is an open-weight research release of a model with highly advanced multilingual capabilities. It focuses on pairing a highly performant pre-trained Command family of models with the result of a year’s dedicated research from Cohere Labs, including data arbitrage, multilingual preference training, safety tuning, and model merging. The result is a powerful multilingual large language model serving 23 languages.
This model card corresponds to the 32-billion version of the Aya Expanse model. We also released an 8-billion version which you can find here.
Install the transformers library and load Aya Expanse 32B as follows:
python
1from transformers import AutoTokenizer, AutoModelForCausalLM
23model_id ="CohereLabs/aya-expanse-32b"4tokenizer = AutoTokenizer.from_pretrained(model_id)5model = AutoModelForCausalLM.from_pretrained(model_id)67# Format message with the chat template8messages =[{"role":"user","content":"Anneme onu ne kadar sevdiğimi anlatan bir mektup yaz"}]9input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")10## <BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Anneme onu ne kadar sevdiğimi anlatan bir mektup yaz<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>1112gen_tokens = model.generate(13 input_ids,14 max_new_tokens=100,15 do_sample=True,16 temperature=0.3,17)1819gen_text = tokenizer.decode(gen_tokens[0])20print(gen_text)
Model Architecture: Aya Expanse 32B is an auto-regressive language model that uses an optimized transformer architecture. Post-training includes supervised finetuning, preference training, and model merging.
Languages covered: The model is particularly optimized for multilinguality and supports the following languages: Arabic, Chinese (simplified & traditional), Czech, Dutch, English, French, German, Greek, Hebrew, Hindi, Indonesian, Italian, Japanese, Korean, Persian, Polish, Portuguese, Romanian, Russian, Spanish, Turkish, Ukrainian, and Vietnamese
Context length: 128K
Evaluation
We evaluated Aya Expanse 32B against Gemma 2 27B, Llama 3.1 70B, Mixtral 8x22B, and Qwen 2.5 32B using the dolly_human_edited subset from the Aya Evaluation Suite dataset and m-ArenaHard, a dataset based on the Arena-Hard-Auto dataset and translated to the 23 languages we support in Aya Expanse. Win-rates were determined using gpt-4o-2024-08-06 as a judge. For a conservative benchmark, we report results from gpt-4o-2024-08-06, though gpt-4o-mini scores showed even stronger performance.
The m-ArenaHard dataset, used to evaluate Aya Expanse’s capabilities, is publicly available here.
Whatsapp Integration
You can also talk to Aya Expanse through the popular messaging service WhatsApp. Use this link to open a WhatsApp chatbox with Aya Expanse.
If you don’t have WhatsApp downloaded on your machine you might need to do that, or, if you have it on your phone, you can follow the on-screen instructions to link your phone and WhatsApp Web.
By the end, you should see a text window which you can use to chat with the model.
More details about our whatsapp integration are available here.
Model Card Contact
For errors or additional questions about details in this model card, contact labs@cohere.com
Terms of Use
We hope that the release of this model will make community-based research efforts more accessible, by releasing the weights of a highly performant multilingual model to researchers all over the world. This model is governed by a CC-BY-NC, requires also adhering to Cohere Lab's Acceptable Use Policy
Cite
You can cite Aya Expanse using:
@misc{dang2024ayaexpansecombiningresearch,
title={Aya Expanse: Combining Research Breakthroughs for a New Multilingual Frontier},
author={John Dang and Shivalika Singh and Daniel D'souza and Arash Ahmadian and Alejandro Salamanca and Madeline Smith and Aidan Peppin and Sungjin Hong and Manoj Govindassamy and Terrence Zhao and Sandra Kublik and Meor Amer and Viraat Aryabumi and Jon Ander Campos and Yi-Chern Tan and Tom Kocmi and Florian Strub and Nathan Grinsztajn and Yannis Flet-Berliac and Acyr Locatelli and Hangyu Lin and Dwarak Talupuru and Bharat Venkitesh and David Cairuz and Bowen Yang and Tim Chung and Wei-Yin Ko and Sylvie Shang Shi and Amir Shukayev and Sammie Bae and Aleksandra Piktus and Roman Castagné and Felipe Cruz-Salinas and Eddie Kim and Lucas Crawhall-Stein and Adrien Morisot and Sudip Roy and Phil Blunsom and Ivan Zhang and Aidan Gomez and Nick Frosst and Marzieh Fadaee and Beyza Ermis and Ahmet Üstün and Sara Hooker},
year={2024},
eprint={2412.04261},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2412.04261},
}
