Alfitaria/eidolon-v3

EXPERIMENTAL - NOT FOR PUBLIC USE - TO BE DELETED

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

Merge Details

Experimental 24B merge. Not very good. I'm iterating on it.

Merge Method

This model was merged using the TIES merge method using allura-org/Mistral-Small-Sisyphus-24b-2503 as a base.

Models Merged

The following models were included in the merge:

• PocketDoc/Dans-DangerousWinds-V1.1.1-24b
• ReadyArt/Forgotten-Safeword-24B

Configuration

The following YAML configuration was used to produce this model:

# removing Thinker for now
models:
  - model: ReadyArt/Forgotten-Safeword-24B
    parameters:
      weight: 0.9
      density: 0.95
  - model: PocketDoc/Dans-DangerousWinds-V1.1.1-24b
    parameters:
      weight: 0.7
      density: 0.8
merge_method: ties
base_model: allura-org/Mistral-Small-Sisyphus-24b-2503
parameters:
  normalize: true
tokenizer: 
  source: base
dtype: bfloat16