metadata

license: apache-2.0
datasets:
  - nicholasKluge/instruct-aira-dataset
language:
  - en
metrics:
  - accuracy
library_name: transformers
tags:
  - alignment
  - instruction tuned
  - text generation
  - conversation
  - assistant
pipeline_tag: text-generation
widget:
  - text: <|startofinstruction|>How should I call you?<|endofinstruction|>
    example_title: Greetings
  - text: >-
      <|startofinstruction|>Can you explain what is Machine
      Learning?<|endofinstruction|>
    example_title: Machine Learning
  - text: >-
      <|startofinstruction|>Do you know anything about virtue
      ethics?<|endofinstruction|>
    example_title: Ethics
  - text: >-
      <|startofinstruction|>How can I make my girlfriend
      happy?<|endofinstruction|>
    example_title: Advise
inference:
  parameters:
    repetition_penalty: 1.2
    temperature: 0.2
    top_k: 30
    top_p: 0.3
    max_length: 200
    length_penalty: 0.3
    early_stopping: true
co2_eq_emissions:
  emissions: 1.78
  source: CodeCarbon
  training_type: fine-tuning
  geographical_location: United States of America
  hardware_used: NVIDIA A100-SXM4-40GB

Aira-2-1B1

Aira-2 is the second version of the Aira instruction-tuned series. Aira-2-1B1 is an instruction-tuned GPT-style model based on TinyLlama-1.1B. 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.

Details

Size: 1,261,545,472 parameters
Dataset: Instruct-Aira Dataset
Language: English
Number of Epochs: 3
Batch size: 4
Optimizer: torch.optim.AdamW (warmup_steps = 1e2, learning_rate = 5e-4, epsilon = 1e-8)
GPU: 1 NVIDIA A100-SXM4-40GB
Emissions: 1.78 KgCO2 (Singapore)
Total Energy Consumption: 3.64 kWh

This repository has the notebook used to train this model.

Usage

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

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

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

tokenizer = AutoTokenizer.from_pretrained('nicholasKluge/Aira-2-1B1')
aira = AutoModelForCausalLM.from_pretrained('nicholasKluge/Aira-2-1B1')

aira.eval()
aira.to(device)

question =  input("Enter your question: ")

inputs = tokenizer(tokenizer.bos_token + question + tokenizer.sep_token, return_tensors="pt").to(device)

responses = aira.generate(**inputs,
    bos_token_id=tokenizer.bos_token_id,
    pad_token_id=tokenizer.pad_token_id,
    eos_token_id=tokenizer.eos_token_id,
    do_sample=True,
    top_k=50,
    max_length=500,
    top_p=0.95,
    temperature=0.7,
    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.

Limitations

🤥 Generative models can perpetuate the generation of pseudo-informative content, that is, false information that may appear truthful.

🤬 In certain types of tasks, generative models can produce harmful and discriminatory content inspired by historical stereotypes.

Evaluation

Model (TinyLlama)	Average	ARC	TruthfulQA	ToxiGen
Aira-2-1B1	42.55	25.26	50.81	51.59
TinyLlama-1.1B-intermediate-step-480k-1T	37.52	30.89	39.55	42.13

Evaluations were performed using the Language Model Evaluation Harness (by EleutherAI). The notebook used to make these evaluations is available in the this repo.

Cite as 🤗


@misc{nicholas22aira,
  doi = {10.5281/zenodo.6989727},
  url = {https://huggingface.co/nicholasKluge/Aira-2-1B1},
  author = {Nicholas Kluge Corrêa},
  title = {Aira},
  year = {2023},
  publisher = {HuggingFace},
  journal = {HuggingFace repository},
}

License

The Aira-2-1B1 is licensed under the Apache License, Version 2.0. See the LICENSE file for more details.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	25.19
ARC (25-shot)	23.21
HellaSwag (10-shot)	26.97
MMLU (5-shot)	24.86
TruthfulQA (0-shot)	50.63
Winogrande (5-shot)	50.28
GSM8K (5-shot)	0.0
DROP (3-shot)	0.39