README.md

metadata

base_model:
  - kromcomp/L3.1-Rays-10B
  - kromcomp/L3.1-Bloat-10B
  - kromcomp/L3.1-Extend-10B
library_name: transformers
tags:
  - mergekit
  - merge

tiv

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 kromcomp/L3.1-Extend-10B as a base.

Models Merged

The following models were included in the merge:

• kromcomp/L3.1-Rays-10B
• kromcomp/L3.1-Bloat-10B

Configuration

The following YAML configuration was used to produce this model:

base_model: kromcomp/L3.1-Extend-10B
chat_template: llama3
dtype: float32
merge_method: della
parameters:
  int8_mask: 1.0
  normalize: 0.0
slices:
- sources:
  - layer_range: [0, 42]
    model: kromcomp/L3.1-Rays-10B
    parameters:
      density: 0.7
      epsilon: 0.05
      weight: [0.0, 0.2, 0.6]
  - layer_range: [0, 42]
    model: kromcomp/L3.1-Bloat-10B
    parameters:
      density: 0.7
      epsilon: 0.05
      weight: [0.9, 0.3, 0.3]
  - layer_range: [0, 42]
    model: kromcomp/L3.1-Extend-10B
    parameters:
      density: 0.65
      epsilon: 0.05
      weight: [0.1, 0.5, 0.1]
tokenizer: {}