Modelpublic

mergekit-community/MN-Sappho-i-12B

sourceHugging Faceupdated 2y agoView on Hugging Face

0likes14downloads

Model Card

merge

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

Merge Details

Merge Method

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

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

yaml

dtype: bfloat16
out_dtype: bfloat16
merge_method: della
base_model: mistralai/Mistral-Nemo-Instruct-2407
models:
  - model: mergekit-community/MN-Sappho-e-12B
    parameters:
      weight: 1.2
  - model: mergekit-community/MN-Sappho-g2-12B
    parameters:
      weight: 0.9
  - model: inflatebot/MN-12B-Mag-Mell-R1
    parameters:
      weight: 1
  - model: LatitudeGames/Wayfarer-12B
    parameters:
      weight: 1.1
  - model: nbeerbower/mistral-nemo-wissenschaft-12B+jtatman/mistral_nemo_12b_reasoning_psychology_lora
    parameters:
      weight: 1.1
  - model: mistralai/Mistral-Nemo-Base-2407
    parameters:
      weight: 1.2
parameters:
  lambda: 1.25
  density: 0.5
  epsilon: 0.25
tokenizer:
  source: mergekit-community/MN-Sappho-e-12B