File size: 2,015 Bytes
d7c5a8c f335a58 d7c5a8c 28448ab eedf5d6 d7c5a8c eedf5d6 d7c5a8c eedf5d6 d7c5a8c eedf5d6 d7c5a8c dd7f5c0 eedf5d6 d7c5a8c eedf5d6 d7c5a8c eedf5d6 d7c5a8c eedf5d6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 |
---
base_model:
- nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
- mlabonne/Llama-3-70B-Instruct-abliterated-LORA
library_name: transformers
tags:
- mergekit
- merge
license: llama3.1
---
# Llama-3.1-Nemotron-lorablated-70B
An uncensored version of [nvidia/Llama-3.1-Nemotron-70B-Instruct-HF](https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct-HF) created by merging [mlabonne/Llama-3-70B-Instruct-abliterated-LORA](https://huggingface.co/mlabonne/Llama-3-70B-Instruct-abliterated-LORA) using [task arithmetic](https://arxiv.org/abs/2212.04089).
## Method
This model was created using [mergekit](https://github.com/cg123/mergekit).
From Ubuntu 24.04 (as root):
```
apt update
apt install pipx
git clone https://github.com/arcee-ai/mergekit.git
cd mergekit && pipx install -e .
mergekit-yaml config.yaml Llama-3.1-Nemotron-lorablated-70B --allow-crimes --lora-merge-cache=./cache
```
See [@mlabonne](https://huggingface.co/mlabonne)'s [Llama-3.1-70B-Instruct-lorablated](https://huggingface.co/mlabonne/Llama-3.1-70B-Instruct-lorablated) for more details on how the LoRA was extracted.
### Configuration
The following YAML configuration was used to produce this model:
```yaml
base_model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF+mlabonne/Llama-3-70B-Instruct-abliterated-LORA
dtype: bfloat16
merge_method: task_arithmetic
parameters:
normalize: false
slices:
- sources:
- layer_range: [0, 80]
model: nvidia/Llama-3.1-Nemotron-70B-Instruct-HF+mlabonne/Llama-3-70B-Instruct-abliterated-LORA
parameters:
weight: 1.0
```
### Acknowlegements
Thanks to [@mlabonne](https://huggingface.co/mlabonne), [@grimjim](https://huggingface.co/grimjim), and [@failspy](https://huggingface.co/failspy) for pioneering this technique for uncensoring models.
Compute provided by [Hetzner](https://www.hetzner.com/) and funded by [Schneewolf Labs](https://schneewolflabs.com/). |