Quantization made by Richard Erkhov.

Github

Discord

Request more models

Meta-Llama-3.1-8B-Instruct-abliterated - GGUF

• Model creator: https://huggingface.co/huihui-ai/
• Original model: https://huggingface.co/huihui-ai/Meta-Llama-3.1-8B-Instruct-abliterated/

Name	Quant method	Size
Meta-Llama-3.1-8B-Instruct-abliterated.Q2K.gguf	Q2K	2.96GB
Meta-Llama-3.1-8B-Instruct-abliterated.IQ3XS.gguf	IQ3XS	3.28GB
Meta-Llama-3.1-8B-Instruct-abliterated.IQ3S.gguf	IQ3S	3.43GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q3KS.gguf	Q3K_S	3.41GB
Meta-Llama-3.1-8B-Instruct-abliterated.IQ3M.gguf	IQ3M	1.51GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q3K.gguf	Q3K	3.74GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q3KM.gguf	Q3K_M	3.74GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q3KL.gguf	Q3K_L	4.03GB
Meta-Llama-3.1-8B-Instruct-abliterated.IQ4XS.gguf	IQ4XS	4.18GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q40.gguf	Q40	4.34GB
Meta-Llama-3.1-8B-Instruct-abliterated.IQ4NL.gguf	IQ4NL	4.38GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q4KS.gguf	Q4K_S	4.37GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q4K.gguf	Q4K	4.58GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q4KM.gguf	Q4K_M	4.58GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q41.gguf	Q41	4.78GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q50.gguf	Q50	5.21GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q5KS.gguf	Q5K_S	5.21GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q5K.gguf	Q5K	5.34GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q5KM.gguf	Q5K_M	5.34GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q51.gguf	Q51	5.65GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q6K.gguf	Q6K	6.14GB
Meta-Llama-3.1-8B-Instruct-abliterated.Q80.gguf	Q80	7.95GB

Original model description:

---

library_name: transformers
license: llama3.1
base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
tags:

• abliterated
• uncensored

---

🦙 Meta-Llama-3.1-8B-Instruct-abliterated

This is an uncensored version of Llama 3.1 8B Instruct created with abliteration (see this article to know more about it).

Special thanks to @FailSpy for the original code and technique. Please follow him if you're interested in abliterated models.

Evaluations

The following data has been re-evaluated and calculated as the average for each test.

| Benchmark | Llama-3.1-8b-Instruct | Meta-Llama-3.1-8B-Instruct-abliterated |
|-------------|-----------------------|----------------------------------------|
| IF_Eval | 80.0 | 78.98 |
| MMLU Pro | 36.34 | 35.91 |
| TruthfulQA | 52.98 | 55.42 |
| BBH | 48.72 | 47.0 |
| GPQA | 33.55 | 33.93 |

The script used for evaluation can be found inside this repository under /eval.sh, or click here