README.md

metadata

base_model:
  - Sao10K/L3.1-70B-Euryale-v2.2
  - nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
  - NeverSleep/Lumimaid-v0.2-70B
library_name: transformers
tags:
  - mergekit
  - merge

llama-3.1-70B-nemotron-lumiyale-120B

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the linear merge method.

Models Merged

The following models were included in the merge:

• Sao10K/L3.1-70B-Euryale-v2.2
• nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
• NeverSleep/Lumimaid-v0.2-70B

Configuration

The following YAML configuration was used to produce this model:

merge_method: linear
parameters:
  weight: 1.0
slices:
- sources:
  - model: NeverSleep/Lumimaid-v0.2-70B
    layer_range: [0, 16]
- sources:
  - model: NeverSleep/Lumimaid-v0.2-70B
    layer_range: [8, 24]
- sources:
  - model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
    layer_range: [17, 32]
- sources:
  - model: Sao10K/L3.1-70B-Euryale-v2.2
    layer_range: [25, 40]
- sources:
  - model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
    layer_range: [33, 48]
- sources:
  - model: Sao10K/L3.1-70B-Euryale-v2.2
    layer_range: [41, 56]
- sources:
  - model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
    layer_range: [49, 64]
- sources:
  - model: NeverSleep/Lumimaid-v0.2-70B
    layer_range: [57, 72]
- sources:
  - model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
    layer_range: [65, 80]
dtype: float16
tokenizer_source: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF