This is a merge of pre-trained language models created using
mergekit.
This model was merged using the
TIES merge method using
liminerity/Phigments12 as a base.
1models:
2 - model: netcat420/MFANN3bv0.6
3 parameters:
4 density: [1, 0.7, 0.1] # density gradient
5 weight: 1.0
6 - model: netcat420/MFANN3bv0.9
7 parameters:
8 density: [1, 0.7, 0.1] # density gradient
9 weight: 1.0
10 - model: netcat420/MFANN3bv0.8.10
11 parameters:
12 density: [1, 0.7, 0.1] # density gradient
13 weight: 1.0
14merge_method: ties
15base_model: liminerity/Phigments12
16parameters:
17 normalize: true
18 int8_mask: true
19dtype: float16
20
this model was used as a base to train MFANNv0.10. and that model now has a context length of 16k!
