kromcomp/L3.1-Flatv1-8B

write

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

Merge Details

Merge Method

This model was merged using the Model Stock merge method using nvidia/Llama-3.1-8B-UltraLong-2M-Instruct as a base.

Models Merged

The following models were included in the merge:

• DISLab/Gen-8B-R2
• Undi95/Unholy-8B-DPO-OAS

Configuration

The following YAML configuration was used to produce this model:

base_model: nvidia/Llama-3.1-8B-UltraLong-2M-Instruct
chat_template: llama3
dtype: float32
merge_method: model_stock
modules:
  default:
    slices:
    - sources:
      - layer_range: [0, 32]
        model: Undi95/Unholy-8B-DPO-OAS
      - layer_range: [0, 32]
        model: DISLab/Gen-8B-R2
      - layer_range: [0, 32]
        model: nvidia/Llama-3.1-8B-UltraLong-2M-Instruct
parameters:
  int8_mask: 1.0
tokenizer:
  pad_to_multiple_of: 4