Quantization made by Richard Erkhov.

Aira-2-124M - GGUF

Model creator: https://huggingface.co/nicholasKluge/
Original model: https://huggingface.co/nicholasKluge/Aira-2-124M/

Name	Quant method	Size
Aira-2-124M.Q2_K.gguf	Q2_K	0.08GB
Aira-2-124M.IQ3_XS.gguf	IQ3_XS	0.08GB
Aira-2-124M.IQ3_S.gguf	IQ3_S	0.08GB
Aira-2-124M.Q3_K_S.gguf	Q3_K_S	0.08GB
Aira-2-124M.IQ3_M.gguf	IQ3_M	0.09GB
Aira-2-124M.Q3_K.gguf	Q3_K	0.09GB
Aira-2-124M.Q3_K_M.gguf	Q3_K_M	0.09GB
Aira-2-124M.Q3_K_L.gguf	Q3_K_L	0.1GB
Aira-2-124M.IQ4_XS.gguf	IQ4_XS	0.1GB
Aira-2-124M.Q4_0.gguf	Q4_0	0.1GB
Aira-2-124M.IQ4_NL.gguf	IQ4_NL	0.1GB
Aira-2-124M.Q4_K_S.gguf	Q4_K_S	0.1GB
Aira-2-124M.Q4_K.gguf	Q4_K	0.11GB
Aira-2-124M.Q4_K_M.gguf	Q4_K_M	0.11GB
Aira-2-124M.Q4_1.gguf	Q4_1	0.11GB
Aira-2-124M.Q5_0.gguf	Q5_0	0.11GB
Aira-2-124M.Q5_K_S.gguf	Q5_K_S	0.11GB
Aira-2-124M.Q5_K.gguf	Q5_K	0.12GB
Aira-2-124M.Q5_K_M.gguf	Q5_K_M	0.12GB
Aira-2-124M.Q5_1.gguf	Q5_1	0.12GB
Aira-2-124M.Q6_K.gguf	Q6_K	0.13GB
Aira-2-124M.Q8_0.gguf	Q8_0	0.17GB

Original model description:

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|>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_new_tokens: 200 length_penalty: 0.3 early_stopping: true co2_eq_emissions: emissions: 250 source: CodeCarbon training_type: fine-tuning geographical_location: United States of America hardware_used: NVIDIA A100-SXM4-40GB

Aira-2-124M

Aira-2 is the second version of the Aira instruction-tuned series. Aira-2-124M 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.

Details

Size: 124,441,344 parameters
Dataset: Instruct-Aira Dataset
Language: English
Number of Epochs: 5
Batch size: 32
Optimizer: torch.optim.AdamW (warmup_steps = 1e2, learning_rate = 5e-4, epsilon = 1e-8)
GPU: 1 NVIDIA A100-SXM4-40GB
Emissions: 0.25 KgCO2 (Singapore)
Total Energy Consumption: 0.52 kWh

This repository has the source code 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-2-124M')
aira = AutoModelForCausalLM.from_pretrained('nicholasKluge/Aira-2-124M')

aira.eval()
aira.to(device)

question =  input("Enter your question: ")

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

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.

Limitations

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.

Evaluation

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

Evaluations were performed using the Language Model Evaluation Harness (by EleutherAI).

Cite as 🤗

@misc{nicholas22aira,
  doi = {10.5281/zenodo.6989727},
  url = {https://github.com/Nkluge-correa/Aira},
  author = {Nicholas Kluge Corrêa},
  title = {Aira},
  year = {2023},
  publisher = {GitHub},
  journal = {GitHub repository},
}

@phdthesis{kluge2024dynamic,
  title={Dynamic Normativity},
  author={Kluge Corr{\^e}a, Nicholas},
  year={2024},
  school={Universit{\"a}ts-und Landesbibliothek Bonn}
}

License

Aira-2-124M is licensed under the Apache License, Version 2.0. See the LICENSE file for more details.