moecule-1.1b-m2

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 TinyLlama-1.1B-1T-OpenOrca as a base.

Models Merged

The following models were included in the merge:

• tiny-llama-1.1b-chat-medical
• TinyLlama-1.1B-intermediate-step-1195k-token-2.5T

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: tiny-llama-1.1b-chat-medical
    parameters:
      density: [1, 0.7, 0.1] # density gradient
      weight: 1.0
  - model: TinyLlama-1.1B-intermediate-step-1195k-token-2.5T
    parameters:
      density: 0.5
      weight: [0, 0.3, 0.7, 1] # weight gradient
merge_method: ties
base_model: TinyLlama-1.1B-1T-OpenOrca
parameters:
  normalize: true
  int8_mask: true
dtype: bfloat16