Perspicacious-LLaMa-70B

{{CAUTION: WIP}} This is a bit of an experiment, trying to merge some smart models together, which I will then combine with a RP focused merge.

merge

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

Merge Details

Merge Method

This model was merged using the SCE merge method using VAGOsolutions/Llama-3.1-SauerkrautLM-70b-Instruct as a base.

Models Merged

The following models were included in the merge:

• MaziyarPanahi/calme-2.3-llama3.1-70b
• deepseek-ai/DeepSeek-R1-Distill-Llama-70B
• Steelskull/L3.3-MS-Nevoria-70b
• Daemontatox/Llama3.3-70B-CogniLink
• Triangle104/Set-70b

Configuration

The following YAML configuration was used to produce this model:

models:
  # Pivot model
  - model: deepseek-ai/DeepSeek-R1-Distill-Llama-70B
  # Target models
  - model: Daemontatox/Llama3.3-70B-CogniLink
  - model: Triangle104/Set-70b
  - model: MaziyarPanahi/calme-2.3-llama3.1-70b
  - model: Steelskull/L3.3-MS-Nevoria-70b
merge_method: sce
base_model: VAGOsolutions/Llama-3.1-SauerkrautLM-70b-Instruct
parameters:
  select_topk: 1.0
dtype: bfloat16