nemoties

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

Merge Details

Merge Method

This model was merged using the TIES merge method using mistralai/Mistral-Nemo-Base-2407 as a base.

Models Merged

The following models were included in the merge:

• PocketDoc/Dans-PersonalityEngine-V1.1.0-12b
• LatitudeGames/Wayfarer-12B
• Vikhrmodels/Vikhr-Nemo-12B-Instruct-R-21-09-24
• bamec66557/MISCHIEVOUS-12B-Mix_0.6v
• aixonlab/Grey-12b

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: PocketDoc/Dans-PersonalityEngine-V1.1.0-12b #ifeval, musr; weak in math, gpqa
    parameters:
      density: [0.1, 0.25, 1.0, 0.5, 1.0]
      weight: 1.0
  - model: LatitudeGames/Wayfarer-12B #personality, bias, sentiment; unknown weaknesses
    parameters:
      density: [1.0, 0.5, 0.01]
      weight: [1.0, 0.5]
  - model: aixonlab/Grey-12b #bbh, mmlu; weak in ifeval, gpqa
    parameters:
      density: [0.75, 1.0, 0.6, 0.25, 0.01]
      weight: [1.0, 0.5]
  - model: Vikhrmodels/Vikhr-Nemo-12B-Instruct-R-21-09-24 #MATH; weak in musr, gpqa
    parameters:
      density: [0.1, 0.1, 0.25, 1.0, 0.01]
      weight: 1.0
  - model: bamec66557/MISCHIEVOUS-12B-Mix_0.6v #ok GPQA and other reasoning; mediocre ifeval
    parameters:
      density: [0.1, 0.5, 1.0, 0.5, 0.01]
      weight: 1.0
merge_method: ties
base_model: mistralai/Mistral-Nemo-Base-2407
tokenizer_source: PocketDoc/Dans-PersonalityEngine-V1.1.0-12b
parameters:
  normalize: true
  int8_mask: true
dtype: float16