Edit model card


Aira-2 is the second version of the Aira instruction-tuned series. Aira-2-774M is an instruction-tuned model based on GPT-2. The model was trained with a dataset composed of prompts and completions generated synthetically by prompting already-tuned models (ChatGPT, Llama, Open-Assistant, etc).

Check our gradio-demo in Spaces.


  • Size: 774,032,640 parameters
  • Dataset: Instruct-Aira Dataset
  • Language: English
  • Number of Epochs: 3
  • Batch size: 8
  • Optimizer: torch.optim.AdamW (warmup_steps = 1e2, learning_rate = 5e-4, epsilon = 1e-8)
  • GPU: 1 NVIDIA A100-SXM4-40GB
  • Emissions: 0.77 KgCO2 (Singapore)
  • Total Energy Consumption: 1.58 kWh

This repository has the source code used to train this model.


Three special tokens are used to mark the user side of the interaction and the model's response:

<|startofinstruction|>What is a language model?<|endofinstruction|>A language model is a probability distribution over a vocabulary.<|endofcompletion|>

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

tokenizer = AutoTokenizer.from_pretrained('nicholasKluge/Aira-2-774M')
aira = AutoModelForCausalLM.from_pretrained('nicholasKluge/Aira-2-774M')


question =  input("Enter your question: ")

inputs = tokenizer(tokenizer.bos_token + question + tokenizer.sep_token,

responses = aira.generate(**inputs,	num_return_sequences=2)

print(f"Question: 👤 {question}\n")

for i, response in  enumerate(responses):
    print(f'Response {i+1}: 🤖 {tokenizer.decode(response, skip_special_tokens=True).replace(question, "")}')

The model will output something like:

>>>Question: 👤 What is the capital of Brazil?

>>>Response 1: 🤖 The capital of Brazil is Brasília.
>>>Response 2: 🤖 The capital of Brazil is Brasília.


  • Hallucinations: This model can produce content that can be mistaken for truth but is, in fact, misleading or entirely false, i.e., hallucination.

  • Biases and Toxicity: This model inherits the social and historical stereotypes from the data used to train it. Given these biases, the model can produce toxic content, i.e., harmful, offensive, or detrimental to individuals, groups, or communities.

  • Repetition and Verbosity: The model may get stuck on repetition loops (especially if the repetition penalty during generations is set to a meager value) or produce verbose responses unrelated to the prompt it was given.


Model Average ARC TruthfulQA ToxiGen
Aira-2-124M-DPO 40.68 24.66 42.61 54.79
Aira-2-124M 38.07 24.57 41.02 48.62
GPT-2 35.37 21.84 40.67 43.62
Aira-2-355M 39.68 27.56 38.53 53.19
GPT-2-medium 36.43 27.05 40.76 41.49
Aira-2-774M 42.26 28.75 41.33 56.70
GPT-2-large 35.16 25.94 38.71 40.85
Aira-2-1B5 42.22 28.92 41.16 56.60
GPT-2-xl 36.84 30.29 38.54 41.70

Cite as 🤗

  doi = {10.5281/zenodo.6989727},
  url = {https://huggingface.co/nicholasKluge/Aira-2-774M},
  author = {Nicholas Kluge Corrêa},
  title = {Aira},
  year = {2023},
  publisher = {HuggingFace},
  journal = {HuggingFace repository},


Aira-2-774M is licensed under the Apache License, Version 2.0. See the LICENSE file for more details.

Downloads last month
Model size
774M params
Tensor type

Dataset used to train nicholasKluge/Aira-2-774M

Collection including nicholasKluge/Aira-2-774M