Introducing EverythingLM, a llama-2 based, general-purpose 13b model with 16k context thanks to LlongMa. The model is trained on the EverythingLM-V3 dataset, more info can be found on the dataset page.
The model is completely uncensored.
Despite being "uncensored", the base model might be resistant; you might have to prompt-engineer certain prompts.
General all around improvements thanks to the new dataset. Check out the dataset for more info.
Prompt format (Alpaca-chat):
USER: <prompt>
ASSISTANT:
Future plans:
Highest priority right now is V3.1 with more optimized training and iterative dataset improvements based on testing.
Note:
Through testing V2, I realized some alignment data had leaked in, causing the model to be less cooperative then intended. This model should do much better due to stricter filetering.
