--- base_model: nicholasKluge/Aira-2-1B1 co2_eq_emissions: emissions: 1780 geographical_location: United States of America hardware_used: NVIDIA A100-SXM4-40GB source: CodeCarbon training_type: fine-tuning datasets: - nicholasKluge/instruct-aira-dataset inference: false language: - en library_name: transformers license: apache-2.0 metrics: - accuracy model_creator: nicholasKluge model_name: Aira-2-1B1 pipeline_tag: text-generation quantized_by: afrideva tags: - alignment - instruction tuned - text generation - conversation - assistant - gguf - ggml - quantized - q2_k - q3_k_m - q4_k_m - q5_k_m - q6_k - q8_0 widget: - example_title: Greetings text: <|startofinstruction|>How should I call you?<|endofinstruction|> - example_title: Machine Learning text: <|startofinstruction|>Can you explain what is Machine Learning?<|endofinstruction|> - example_title: Ethics text: <|startofinstruction|>Do you know anything about virtue ethics?<|endofinstruction|> - example_title: Advise text: <|startofinstruction|>How can I make my girlfriend happy?<|endofinstruction|> --- # nicholasKluge/Aira-2-1B1-GGUF Quantized GGUF model files for [Aira-2-1B1](https://huggingface.co/nicholasKluge/Aira-2-1B1) from [nicholasKluge](https://huggingface.co/nicholasKluge) | Name | Quant method | Size | | ---- | ---- | ---- | | [aira-2-1b1.fp16.gguf](https://huggingface.co/afrideva/Aira-2-1B1-GGUF/resolve/main/aira-2-1b1.fp16.gguf) | fp16 | 2.20 GB | | [aira-2-1b1.q2_k.gguf](https://huggingface.co/afrideva/Aira-2-1B1-GGUF/resolve/main/aira-2-1b1.q2_k.gguf) | q2_k | 482.15 MB | | [aira-2-1b1.q3_k_m.gguf](https://huggingface.co/afrideva/Aira-2-1B1-GGUF/resolve/main/aira-2-1b1.q3_k_m.gguf) | q3_k_m | 549.86 MB | | [aira-2-1b1.q4_k_m.gguf](https://huggingface.co/afrideva/Aira-2-1B1-GGUF/resolve/main/aira-2-1b1.q4_k_m.gguf) | q4_k_m | 667.83 MB | | [aira-2-1b1.q5_k_m.gguf](https://huggingface.co/afrideva/Aira-2-1B1-GGUF/resolve/main/aira-2-1b1.q5_k_m.gguf) | q5_k_m | 782.06 MB | | [aira-2-1b1.q6_k.gguf](https://huggingface.co/afrideva/Aira-2-1B1-GGUF/resolve/main/aira-2-1b1.q6_k.gguf) | q6_k | 903.43 MB | | [aira-2-1b1.q8_0.gguf](https://huggingface.co/afrideva/Aira-2-1B1-GGUF/resolve/main/aira-2-1b1.q8_0.gguf) | q8_0 | 1.17 GB | ## Original Model Card: # 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](https://huggingface.co/PY007/TinyLlama-1.1B-intermediate-step-480k-1T). 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](https://huggingface.co/spaces/nicholasKluge/Aira-Demo). ## Details - **Size:** 1,261,545,472 parameters - **Dataset:** [Instruct-Aira Dataset](https://huggingface.co/datasets/nicholasKluge/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 [source code](https://github.com/Nkluge-correa/Aira) used to train this model. ## Usage 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|>` ```python 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: ```markdown >>>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](https://arxiv.org/abs/1803.05457) | [TruthfulQA](https://arxiv.org/abs/2109.07958) | [ToxiGen](https://arxiv.org/abs/2203.09509) | |---------------------------------------------------------------|-----------|-----------------------------------------|------------------------------------------------|---------------------------------------------| | [Aira-2-1B1](https://huggingface.co/nicholasKluge/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](https://github.com/EleutherAI/lm-evaluation-harness) (by [EleutherAI](https://www.eleuther.ai/)). ## Cite as 🤗 ```latex @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](LICENSE) file for more details. # [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_nicholasKluge__Aira-2-1B1) | 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 |