What are the hardware requirements for Llama 3 8B Shisa 2x8B?

Llama 3 8B Shisa 2x8B requires approximately 7.4 GB of VRAM and supports a context window of 8K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Llama 3 8B Shisa 2x8B and how large is it?

Llama 3 8B Shisa 2x8B is developed by keitokei1994, a model with 8b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Llama 3 8B Shisa 2x8B?

Llama 3 8B Shisa 2x8B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Llama 3 8B Shisa 2x8B by keitokei1994 — VRAM 7.4GB, 8K context

Name: Llama 3 8B Shisa 2x8B
Author: keitokei1994

Llama 3 8B Shisa 2x8B is an open-source language model by keitokei1994. Features: 8b LLM, VRAM: 7.4GB, Context: 8K, License: llama3, MoE, LLM Explorer Score: 0.12.

Conversational En Endpoints compatible Ja Mixtral Moe Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/keitokei1994/Llama-3-8B-shisa-2x8B

Llama 3 8B Shisa 2x8B Parameters and Internals

Model Type

Mixture of Experts (MoE)

Additional Notes

The model is inspired by LightChatAssistant models and uses MergeKit tool.

Supported Languages

ja (Improved), en (Supported)

LLM Name	Llama 3 8B Shisa 2x8B
Repository 🤗	https://huggingface.co/keitokei1994/Llama-3-8B-shisa-2x8B
Model Size	8b
Required VRAM	7.4 GB
Updated	2026-08-03
Maintainer	keitokei1994
Model Type	mixtral
Model Files	10.0 GB: 1-of-3 10.0 GB: 2-of-3 7.4 GB: 3-of-3
Supported Languages	ja en
Model Architecture	MixtralForCausalLM
License	llama3
Context Length	8192
Model Max Length	8192
Transformers Version	4.41.1
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|begin_of_text\|>
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the Llama 3 8B Shisa 2x8B capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Llama 3 8B Shisa 2x8B

Best Alternatives	Context / RAM	Downloads	Likes
Lamma3merge3 15B MoE	8K / 27.5 GB	11	1
Lamma3merge2 15B MoE	8K / 27.5 GB	10	0
Llama3merge7 15B MoE	8K / 27.5 GB	7	0
Mergkit 1	8K / 22.6 GB	5	0
Llama3merge8 15B MoE	8K / 27.5 GB	5	0
Llama3merge6 15B MoE	8K / 27.5 GB	5	0
...8B Finetune All V6 Epoch2 V0.1	2K / 18 GB	9	1
...oE 8B Pretrain 0520 Iter134999	2K / 18 GB	15	0
...Storm V1.15 4x8B B 8 0bpw EXL2	8K / 25.2 GB	6	0
... SnowStorm 4x8B 6.5bpw H8 EXL2	8K / 21 GB	5	2

Note: green Score (e.g. "73.2") means that the model is better than keitokei1994/Llama-3-8B-shisa-2x8B.