LLM EXPLORER 57,252 MODELS INDEXED

Mixllama3 8x8b Instruct V0.1 by sherazkhan

By sherazkhan · 2 downloads
HF Repo ↗ Copy URL

Mixllama3 8x8b Instruct V0.1 is an open-source language model by sherazkhan. Features: 47.5b LLM, VRAM: 95.3GB, Context: 8K, License: llama3, MoE, Instruction-Based, LLM Explorer Score: 0.12.

  Conversational   En   Endpoints compatible   Instruct   Mixtral   Moe   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/sherazkhan/Mixllama3-8x8b-Instruct-v0.1 

Mixllama3 8x8b Instruct V0.1 Parameters and Internals

Model Type 
text generation
Additional Notes 
The model is an experimental research model and might generate incorrect or harmful content. Outputs should not be considered as factual or representative of the creator's views. License and copyright are under Meta Platforms, Inc.
Supported Languages 
en (fluent)
Training Details 
Methodology:
The model is a Mixture of Experts (MoE) model combining 8 fine-tuned LLaMA 8B models, each focused on a specific set of tasks, to enhance performance and adaptability.
Model Architecture:
MoE based on LLaMA-3-8B
LLM NameMixllama3 8x8b Instruct V0.1
Repository 🤗https://huggingface.co/sherazkhan/Mixllama3-8x8b-Instruct-v0.1 
Model Size47.5b
Required VRAM95.3 GB
Updated2026-07-22
Maintainersherazkhan
Model Typemixtral
Instruction-BasedYes
Model Files  5.0 GB: 1-of-20   5.0 GB: 2-of-20   4.9 GB: 3-of-20   5.0 GB: 4-of-20   5.0 GB: 5-of-20   4.9 GB: 6-of-20   5.0 GB: 7-of-20   5.0 GB: 8-of-20   5.0 GB: 9-of-20   4.9 GB: 10-of-20   5.0 GB: 11-of-20   5.0 GB: 12-of-20   4.9 GB: 13-of-20   5.0 GB: 14-of-20   5.0 GB: 15-of-20   4.9 GB: 16-of-20   5.0 GB: 17-of-20   5.0 GB: 18-of-20   4.7 GB: 19-of-20   1.1 GB: 20-of-20
Supported Languagesen
Model ArchitectureMixtralForCausalLM
Licensellama3
Context Length8192
Model Max Length8192
Transformers Version4.40.0
Tokenizer ClassPreTrainedTokenizerFast
Vocabulary Size128256
Torch Data Typefloat16

Best Alternatives to Mixllama3 8x8b Instruct V0.1

Best Alternatives
Context / RAM
Downloads
Likes
Llama 3 8B Instruct MoE 48K / 95.2 GB70
Note: green Score (e.g. "73.2") means that the model is better than sherazkhan/Mixllama3-8x8b-Instruct-v0.1.