Trained on a transcript which was yoinked from this AO3 fanfic: https://archiveofourown.org/works/47722798/chapters/120296941 As the name suggests, this is trained on dialogue from this 'Hana' character, a chaotic gremlin.
Note that on the 4th reply, for some reason the model implodes and the replies become nonsensical. I'd get around this by limiting the memory to a maximum of 4 messages total.
- Downloads last month
- 6
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Model tree for Skrunbger/DialoGPT-medium-HANA
Base model
microsoft/DialoGPT-medium