---
base_model:
- KaraKaraWitch/BlenderCartel-llama33-70B-Pt1
- KaraKaraWitch/BlenderCartel-llama33-70B-Pt2
library_name: transformers
tags:
- mergekit
- mergekitty
- merge

---
# KaraKaraWitch/StackedBlenderCartel-llama33-120B

This is a merge of pre-trained language models created using [mergekitty](https://github.com/allura-org/mergekitty).

## But... __WHY?!__

... A stupid idea crept into my mind when I saw that I have 2 70B models of pretty good (:TM:) quality.

> [You could make a stacked model out of this.](https://knowyourmeme.com/memes/you-could-make-a-religion-out-of-this)

So... here we are.  

I know, contrary to what people have told [us](https://merge.moe/) [countless](https://huggingface.co/NobodyExistsOnTheInternet/K3-Q4-GGUF) [times](https://huggingface.co/Replete-AI/Kronos-703B) [not](https://huggingface.co/Kquant03/Phalanx-512x460M-MoE) [to](https://huggingface.co/alpindale/goliath-120b) [do](https://huggingface.co/alpindale/miquella-120b) [it](https://huggingface.co/NobodyExistsOnTheInternet/Llama-2-70b-x8-MoE-clown-truck), here we are once again with a stupid idea.

## Merge Details
### Merge Method

This model was merged using the `passthrough` merge method.

### Models Merged

The following models were included in the merge:
* [KaraKaraWitch/BlenderCartel-llama33-70B-Pt1](https://huggingface.co/KaraKaraWitch/BlenderCartel-llama33-70B-Pt1)
* [KaraKaraWitch/BlenderCartel-llama33-70B-Pt2](https://huggingface.co/KaraKaraWitch/BlenderCartel-llama33-70B-Pt2)

### Configuration

The following YAML configuration was used to produce this model:

```yaml

slices:
  - sources:
    - model: KaraKaraWitch/BlenderCartel-llama33-70B-Pt1
      layer_range: [0, 16]

  - sources:
    - model: KaraKaraWitch/BlenderCartel-llama33-70B-Pt2
      layer_range: [8, 24]

  - sources:
    - model: KaraKaraWitch/BlenderCartel-llama33-70B-Pt1
      layer_range: [17, 32]

  - sources:
    - model: KaraKaraWitch/BlenderCartel-llama33-70B-Pt2
      layer_range: [25, 40]

  - sources:
    - model: KaraKaraWitch/BlenderCartel-llama33-70B-Pt1
      layer_range: [33, 48]

  - sources:
    - model: KaraKaraWitch/BlenderCartel-llama33-70B-Pt2
      layer_range: [41, 56]

  - sources:
    - model: KaraKaraWitch/BlenderCartel-llama33-70B-Pt1
      layer_range: [49, 64]

  - sources:
    - model: KaraKaraWitch/BlenderCartel-llama33-70B-Pt2
      layer_range: [57, 72]

  - sources:
    - model: KaraKaraWitch/BlenderCartel-llama33-70B-Pt1
      layer_range: [65, 80]

merge_method: passthrough
dtype: float16


```