(Previously BeaverAI/Rocinante-12B-v2c-GGUF) | |
# UnslopNemo v1 (Experiment) | |
I unslopped roughly 80% of my RP dataset in an attempt to make the model more expressive. | |
## Feedback | |
- Feedback is very important for this one. If you have any thoughts, positive or negative, please share it in the discussion. | |
## Usage | |
- ChatML, Text Completion, or Mistral | |
- You may need to set `<|im_end|>` as a stop token | |
- DRY surprisingly brought up the slop, so I suggest playing around with your samplers |