This is an experimental lima style model trained on a small subset of freedom-rp and erotica-analysis-16k. Due to the much smaller dataset size (about 1000 samples from each original dataset) it was much easier to edit and clean thoroughly. I also used a slightly lower learning rate of 0.00015.
The prompt format is chatml.
I have not tested the model yet, but I am hoping I can use this to help me create more training data for specific genres.
Please consider subscribing to my patreon or buying a giant candle dick on my etsy to show your support.
- Downloads last month
- 2
This model does not have enough activity to be deployed to Inference API (serverless) yet.
Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.