Ppoyaa/L3-Inca-8B-v0.8

---

L3-Inca-8B-v0.8

L3-Inca-8B-v0.8 is a merge of the following models:

• Sao10K/L3-8B-Stheno-v3.2
• Nitral-AI/Hathor-L3-8B-v.02
• grimjim/Llama-3-Luminurse-v0.2-OAS-8B

using NurtureAI/Meta-Llama-3-8B-Instruct-32k as the base.

UPDATE: Changed the merging method from model_stock to ties and made Stheno have the most weight and density.

---

Quantized Models by mradermacher

• Static L3-Inca-8B-v0.8-GGUF

• Imatrix L3-Inca-8B-v0.8-i1-GGUF

---

Configuration

models:
  - model: Sao10K/L3-8B-Stheno-v3.2
    parameters:
      density: 0.85
      weight: 0.5
  - model: Nitral-AI/Hathor-L3-8B-v.02
    parameters:
      density: 0.75
      weight: 0.3
  - model: grimjim/Llama-3-Luminurse-v0.2-OAS-8B
    parameters:
      density: 0.75
      weight: 0.2
merge_method: ties
base_model: NurtureAI/Meta-Llama-3-8B-Instruct-32k
parameters:
  normalize: false
  int8_mask: true
dtype: bfloat16