This is a merge of pre-trained language models created using
mergekit.
This model was merged using the Passthrough merge method.
1
2merge_method: passthrough
3dtype: bfloat16
4tokenizer_source: union
5#base_model: huihui-ai/Hermes-3-Llama-3.2-3B-abliterated
6
7slices:
8 - sources:
9 - model: EpistemeAI/ReasoningCore-Llama-3.2-3B-r1-v1_2
10 layer_range: [0, 16]
11 parameters:
12 weight: 0.65 # Transcendent: HyperZone 1 - Efficiency Nexus (low layers with entropy scaling)
13 - sources:
14 - model: bunnycore/Llama-3.2-3b-RP-Toxic-Fuse
15 layer_range: [16, 27]
16 parameters:
17 weight: 0.6 # Hyperscaled for transcendent boost
18 - sources:
19 - model: merge # Reference the Step 1 output directory/model
20 layer_range: [26, 27]
21 # tensors: ["self_attn"]
22 parameters:
23 weight: 0.75 # Optimized via simulated genius annealing
24 # tensors: ["mlp"]
25 # parameters:
26 # weight: 0.95
27
28
