What are the hardware requirements for Ko Llama3 Luxia 8B?

Ko Llama3 Luxia 8B requires approximately 16.4 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 Ko Llama3 Luxia 8B and how large is it?

Ko Llama3 Luxia 8B is developed by saltlux, 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 Ko Llama3 Luxia 8B?

Ko Llama3 Luxia 8B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Ko Llama3 Luxia 8B by saltlux — VRAM 16.4GB, 8K context

Name: Ko Llama3 Luxia 8B
Author: saltlux

Ko Llama3 Luxia 8B is an open-source language model by saltlux. Features: 8b LLM, VRAM: 16.4GB, Context: 8K, License: llama3, LLM Explorer Score: 0.13.

Conversational Deploy:azure En Endpoints compatible Ko Llama Llama-3 Luxia Meta Pytorch Region:us Safetensors Saltlux Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/saltlux/Ko-Llama3-Luxia-8B

Ko Llama3 Luxia 8B Parameters and Internals

Model Type

text generation

Use Cases

Areas:

research, text generation

Primary Use Cases:

natural language generation

Additional Notes

Ko-Llama3-Luxia-8B is optimized for Korean language usage and is trained using an extended Korean vocabulary with Llama-3 Tokenizer.

Supported Languages

en (partial), ko (full)

Training Details

Data Sources:

news, law, patents, medicine, history, society, culture, conversation (written/spoken)

Data Volume:

100GB

Methodology:

Pretraining

Context Length:

8000

Training Time:

unknown

Hardware Used:

NVIDIA H100 80GB * 8 units

Model Architecture:

unknown

Input Output

Input Format:

text

Accepted Modalities:

text

Output Format:

text

LLM Name	Ko Llama3 Luxia 8B
Repository 🤗	https://huggingface.co/saltlux/Ko-Llama3-Luxia-8B
Model Size	8b
Required VRAM	16.4 GB
Updated	2026-07-30
Maintainer	saltlux
Model Type	llama
Model Files	4.9 GB: 1-of-4 4.9 GB: 2-of-4 5.0 GB: 3-of-4 1.6 GB: 4-of-4
Supported Languages	en ko
Model Architecture	LlamaForCausalLM
License	llama3
Context Length	8192
Model Max Length	8192
Transformers Version	4.37.2
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|end_of_text\|>
Vocabulary Size	145792
Torch Data Type	bfloat16

Rank the Ko Llama3 Luxia 8B 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 Ko Llama3 Luxia 8B

Best Alternatives	Context / RAM	Downloads	Likes
...otron 8B UltraLong 4M Instruct	4192K / 32.1 GB	1135	125
UltraLong Thinking	4192K / 16.1 GB	2	3
...a 3.1 8B UltraLong 4M Instruct	4192K / 32.1 GB	176	24
...a 3.1 8B UltraLong 2M Instruct	2096K / 32.1 GB	875	9
...otron 8B UltraLong 2M Instruct	2096K / 32.1 GB	124	18
Cthulhu 8B V1.4	1048K / 16.1 GB	10	10
...raLong 1M Instruct Abliterated	1048K / 32.1 GB	4	9
...a 3.1 8B UltraLong 1M Instruct	1048K / 32.1 GB	1387	29
...otron 8B UltraLong 1M Instruct	1048K / 32.1 GB	702	59
...xis Bookwriter Llama3.1 8B Sft	1048K / 16.1 GB	25	4

Note: green Score (e.g. "73.2") means that the model is better than saltlux/Ko-Llama3-Luxia-8B.