prototype-0.4x105

prototype-0.4x105

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

Merge Details

Merge Method

This model was merged using the DARE TIES merge method using /workspace/prototype-0.4x102 as a base.

Models Merged

The following models were included in the merge:

• /workspace/cache/models--Daemontatox--Llama3.3-70B-CogniLink/snapshots/99ede7d64184a107a405eea01f0a3eb5dc9f669a
• /workspace/cache/models--nbeerbower--Llama-3.1-Nemotron-lorablated-70B/snapshots/713defaa340007a0163832318b7b70d1880770f1
• /workspace/cache/models--ValiantLabs--Llama3.1-70B-ShiningValiant2/snapshots/cfefbea0fa3e21a214e9be56557081566bc5b246

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: /workspace/cache/models--ValiantLabs--Llama3.1-70B-ShiningValiant2/snapshots/cfefbea0fa3e21a214e9be56557081566bc5b246
    parameters:
      weight: 0.25
      density: 0.5
  - model: /workspace/cache/models--Daemontatox--Llama3.3-70B-CogniLink/snapshots/99ede7d64184a107a405eea01f0a3eb5dc9f669a
    parameters:
      weight: 0.25
      density: 0.5
  - model: /workspace/cache/models--nbeerbower--Llama-3.1-Nemotron-lorablated-70B/snapshots/713defaa340007a0163832318b7b70d1880770f1
    parameters:
      weight: 0.25
      density: 0.5
  - model: /workspace/prototype-0.4x102
    parameters:
      weight: 0.25
      density: 0.5
merge_method: dare_ties
base_model: /workspace/prototype-0.4x102
parameters:
  normalize: false
  lambda: .95
dtype: bfloat16
int8_mask: true
chat_template: llama3
tokenizer:
 source: base