GPTNeo
GPTNeo is a model released by EleutherAI to try and provide an open source model with capabilities similar to OpenAI's GPT-3 model. One of the earliest such models, GPTNeo was trained on The Pile, Eleuther's corpus of web text.
A side-by-side comparison of release history, capabilities, licensing and model sizes.
| Attribute |
 GPTNeo
|
 Vicuna
|
|---|---|---|
| Developer | EleutherAI, Together | UC Berkeley, CMU, Stanford, MBZUAI, UCSD |
| Current generation | – | – |
| Initial release | 2021-03-21 | 2023-03-30 |
| Instruct tuned | No | Yes |
| Coding | No | No |
| Finetuning | Yes | Yes |
| Open source | Yes | Yes |
| License | Apache 2.0 | Noncommercial |
| Model sizes | 1.3B, 2.7B | 13B |
| Links |
GPTNeo is a model released by EleutherAI to try and provide an open source model with capabilities similar to OpenAI's GPT-3 model. One of the earliest such models, GPTNeo was trained on The Pile, Eleuther's corpus of web text.
Released alongside Koala, Vicuna is one of many descendants of the Meta LLaMA model trained on dialogue data collected from the ShareGPT website. According to the authors, Vicuna achieves more than 90% of ChatGPT's quality in user preference tests, while vastly outperforming Alpaca. As of May 2023, Vicuna seems to be the heir apparent of the instruct-finetuned LLaMA model family, though it is also restricted from commercial use.