smelborp_-_MixtralOrochi8x7B-gguf

RichardErkhov/smelborp_-_MixtralOrochi8x7B-gguf

smelborp_-_MixtralOrochi8x7B-gguf is a machine learning model from RichardErkhov. Use it for the machine learning task on the model card, and read the license before you ship it in a product.

Downloads · 30 days

136

9% of all-time downloads

All-time downloads

1.5K

Public

Repo size

541 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.gguf592 GB · 100%

At a glance

Access: Public
Created: Aug 29, 2024
Updated: Aug 30, 2024
SHA: 3a8b349e

Model card

From the Hugging Face model README

Quantization made by Richard Erkhov.

Github

Discord

Request more models

MixtralOrochi8x7B - GGUF

Model creator: https://huggingface.co/smelborp/
Original model: https://huggingface.co/smelborp/MixtralOrochi8x7B/

Name	Quant method	Size
MixtralOrochi8x7B.Q2_K.gguf	Q2_K	16.12GB
MixtralOrochi8x7B.IQ3_XS.gguf	IQ3_XS	18.02GB
MixtralOrochi8x7B.IQ3_S.gguf	IQ3_S	19.03GB
MixtralOrochi8x7B.Q3_K_S.gguf	Q3_K_S	19.03GB
MixtralOrochi8x7B.IQ3_M.gguf	IQ3_M	19.96GB
MixtralOrochi8x7B.Q3_K.gguf	Q3_K	21.0GB
MixtralOrochi8x7B.Q3_K_M.gguf	Q3_K_M	21.0GB
MixtralOrochi8x7B.Q3_K_L.gguf	Q3_K_L	22.51GB
MixtralOrochi8x7B.IQ4_XS.gguf	IQ4_XS	23.63GB
MixtralOrochi8x7B.Q4_0.gguf	Q4_0	24.63GB
MixtralOrochi8x7B.IQ4_NL.gguf	IQ4_NL	24.91GB
MixtralOrochi8x7B.Q4_K_S.gguf	Q4_K_S	24.91GB
MixtralOrochi8x7B.Q4_K.gguf	Q4_K	26.49GB
MixtralOrochi8x7B.Q4_K_M.gguf	Q4_K_M	26.49GB
MixtralOrochi8x7B.Q4_1.gguf	Q4_1	27.32GB
MixtralOrochi8x7B.Q5_0.gguf	Q5_0	30.02GB
MixtralOrochi8x7B.Q5_K_S.gguf	Q5_K_S	30.02GB
MixtralOrochi8x7B.Q5_K.gguf	Q5_K	30.95GB
MixtralOrochi8x7B.Q5_K_M.gguf	Q5_K_M	11.0GB
MixtralOrochi8x7B.Q5_1.gguf	Q5_1	32.71GB
MixtralOrochi8x7B.Q6_K.gguf	Q6_K	35.74GB
MixtralOrochi8x7B.Q8_0.gguf	Q8_0	46.22GB

Original model description:

license: cc-by-nc-4.0 language:

en tags:
mixtral
uncensored
high-intelligence

Orochi

Overview

Orochi is a cutting-edge language model based on the Mixtral architecture developed by Mistral. It represents a sophisticated merge of several prominent models, including Mixtral instruct, Noromaid, OpenBuddy, and several others, using mergekit with the DARE merge method. This model aims to provide highly intelligent responses unrestricted by content limitations. The name "Orochi" references the mythical Yamata-no-Orochi, symbolizing the model's multifaceted and powerful capabilities.

Goals

Uncensored Content: To provide unrestricted and comprehensive responses across various domains.
High Intelligence: Leverage the combined knowledge and capabilities of the merged models to deliver insightful and accurate information.
Innovation in Language Modeling: Push the boundaries of what's possible in natural language understanding and generation.

Model Details

Architecture: Mixtral, a Mixture of Experts model, underlies Orochi's design, enabling it to specialize and optimize its responses across different tasks and topics.
Merge Strategy: Utilizing mergekit and the DARE method, Orochi integrates aspects of various models to enhance its performance and capabilities.

Usage

Due to its uncensored nature, Orochi is best utilized in environments where intelligent, unrestricted dialogue is necessary. Users are encouraged to implement their own content moderation or alignment strategies appropriate for their use case.

Ethical Considerations

As an uncensored model, Orochi may generate content that is unsuitable for all audiences. Users are advised to consider the implications of using such a model and to implement suitable safeguards and ethical guidelines.

Acknowledgements

Orochi is a product of numerous contributions from the fields of machine learning and language modeling. Special thanks to the teams behind Mixtral, mergekit, and all the individual models integrated into Orochi.

Created: Aug 29, 2024
Updated: Aug 30, 2024