CoolFace
Modelpublic

appvoid/arco

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face
14likes61downloads
Model Card

<style> img{ user-select: none; transition: all 0.2s ease; border-radius: .5rem; } img:hover{ transform: rotate(2deg); filter: invert(100%); } @import url('https://fonts.googleapis.com/css2?family=Vollkorn:ital,wght@0,400..900;1,400..900&display=swap'); </style>

<div style="background-color: transparent; border-radius: .5rem; padding: 2rem; font-family: monospace; font-size: .85rem; text-align: justify;">

cubby

arco consistently outperforms every sota model below 600m parameters as well as some 1b base models on average on 5 core benchmarks and is competitive with the best 0.7b-1b llms. arco is a merge of multiple internal models fine-tuned on a diverse set of styles and finally merged with the several models (including palmer-004-turbo and danube3-chat), followed by a merge with base model to preserve knowledge.

prompt

there is no prompt intentionally set but this one worked really good for me:

The following is a conversation between a super smart AI assistant and an user.

user: <your question>

assistant:
benchmarks

zero-shot evaluations performed on current sota ~0.5b models.

ParametersModelMMLUARC-CHellaSwagPIQAWinograndeAverage
0.5bqwen244.1328.9249.0569.3156.9949.68
1.1btinyllama25.7730.2959.3573.2959.5949.66
0.5bdanube3-base24.8136.1860.4673.7861.0151.25
0.5bdanube3-chat25.5436.2660.7274.3261.4051.64
0.5bpalmer-004-turbo27.3635.5861.7973.6761.1751.91
0.5barco26.1737.2962.8874.3762.2752.60
supporters

<a href="https://ko-fi.com/appvoid" target="_blank"><img src="https://cdn.buymeacoffee.com/buttons/v2/default-yellow.png" alt="Buy Me A Coffee" style="height: 34px !important; margin-top: -4px;width: 128px !important; filter: contrast(2) grayscale(100%) brightness(100%);" ></a>

trivia

arco comes from spanish word "bow" which is always associated with arrows and hence, speed. </div>