LightGPT

The LightGPT-instruct-6B language model was developed by AWS Contributors using GPT-J 6B and fine-tuned on the OIG-small-chip2 instruction dataset, which contains around 200K training examples. The model is capable of generating text in response to prompts with specific instructions, following a standardized format. It recognizes the end of its response when the prompt ends with "### Response: