This repo contains a merge of pre-trained language models created using
mergekit.
Text generation length variable. Creativity is on the wild side. Attention could use improvement. Model safety questionable.
I'm uploading anyway, as this model has potential value as a merge contribution.
This model was merged using the SLERP merge method.
1models:
2 - model: Delta-Vector/Odin-9B
3 - model: anthracite-org/magnum-v3-9b-chatml
4merge_method: slerp
5base_model: Delta-Vector/Odin-9B
6parameters:
7 t:
8 - value: 0.4
9dtype: bfloat16
10