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

Home/Models/mlabonne/Meta-Llama-3.1-8B-Instruct-abliterated

mlabonne/Meta-Llama-3.1-8B-Instruct-abliterated

kind: abliterated
quant: abliterated
family: llama
modality: text-generation
root model: meta-llama/Meta-Llama-3.1-8B-Instruct
provider: mlabonne
license: llama3.1
language: —
context: —
library: transformers
vRAM (GB): —
size on disk: —
downloads: 24,003
likes: 210
created: —
last modified: 2024-09-14T22:17:18+00:00 (2024)
huggingface: https://huggingface.co/mlabonne/Meta-Llama-3.1-8B-Instruct-abliterated
tags: transformerssafetensorsllamatext-generationabliterateduncensoredconversationalbase_model:meta-llama/Llama-3.1-8B-Instructbase_model:finetune:meta-llama/Llama-3.1-8B-Instructlicense:llama3.1model-indextext-generation-inferenceendpoints_compatibleregion:us

benchmark	category	score	context	conf	evidence
GPQA	reasoning	89.0	GPQA (0-shot)	medium	`\| GPQA (0-shot) \| 0.89 \|`
IFEval	instruction-following	73.29	IFEval (0-Shot)	medium	`\| IFEval (0-Shot) \| 73.29 \|`
MMLU-Pro	knowledge	27.81	MMLU-PRO (5-shot)	medium	`\| MMLU-PRO (5-shot) \| 27.81 \|`
BBH	reasoning	27.13	BBH (3-Shot)	medium	`\| BBH (3-Shot) \| 27.13 \|`
MATH	math	6.42	MATH Lvl 5 (4-Shot)	medium	`\| MATH Lvl 5 (4-Shot) \| 6.42 \|`
MUSR	reasoning	3.21	MuSR (0-shot)	medium	`\| MuSR (0-shot) \| 3.21 \|`

none scraped.