What are the hardware requirements for Llama 3 8B 64K PoSE?

Llama 3 8B 64K PoSE requires approximately 16.1 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Llama 3 8B 64K PoSE and how large is it?

Llama 3 8B 64K PoSE is developed by winglian, 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 64K PoSE?

Llama 3 8B 64K PoSE is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Llama 3 8B 64K PoSE by winglian — VRAM 16.1GB, 128K context

Name: Llama 3 8B 64K PoSE
Author: winglian

Llama 3 8B 64K PoSE is an open-source language model by winglian. Features: 8b LLM, VRAM: 16.1GB, Context: 128K, LLM Explorer Score: 0.15.

Arxiv:2309.10400 Axolotl Deploy:azure En Endpoints compatible Facebook Llama Llama-3 Meta Pytorch Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/winglian/Llama-3-8b-64k-PoSE

Llama 3 8B 64K PoSE Parameters and Internals

Model Type

text generation, instruction tuned

Use Cases

Areas:

commercial applications, research

Applications:

assistant-like chat, natural language generation tasks

Primary Use Cases:

Instruction tuned applications for dialogue, Pretrained models adaptation for various tasks

Limitations:

Not intended for legal or medical advice

Considerations:

Developers should follow the Responsible Use Guide

Additional Notes

Model includes tools for responsible AI practices.

Supported Languages

English (Full proficiency)

Training Details

Data Sources:

RedPajama V1

Data Volume:

15 trillion tokens

Methodology:

SFT, RLHF

Context Length:

64000

Hardware Used:

H100-80GB GPUs, Meta's Research SuperCluster

Model Architecture:

Auto-regressive transformer

Safety Evaluation

Methodologies:

red teaming, adversarial evaluations

Risk Categories:

CBRNE threats, Cyber Security, Child Safety

Ethical Considerations:

Iterative testing and expert consultations conducted

Responsible Ai Considerations

Fairness:

Comprehensive steps were taken to ensure fairness.

Transparency:

Open community collaboration and resources provided for transparency.

Accountability:

Meta is responsible for the model's deployment.

Mitigation Strategies:

Use of safeguards like Meta Llama Guard 2 and Code Shield

Input Output

Input Format:

Text

Accepted Modalities:

text

Output Format:

Text and code

Performance Tips:

Use appropriate hardware and settings for extended context.

Release Notes

Version:

Version for transformers

Notes:

Optimized for usage with the transformer library.

LLM Name	Llama 3 8B 64K PoSE
Repository 🤗	https://huggingface.co/winglian/Llama-3-8b-64k-PoSE
Model Size	8b
Required VRAM	16.1 GB
Updated	2026-08-06
Maintainer	winglian
Model Type	llama
Model Files	5.0 GB: 1-of-4 5.0 GB: 2-of-4 4.9 GB: 3-of-4 1.2 GB: 4-of-4
Supported Languages	en
Model Architecture	LlamaForCausalLM
Context Length	131072
Model Max Length	131072
Transformers Version	4.40.0.dev0
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|end_of_text\|>
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the Llama 3 8B 64K PoSE 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

Quantized Models of the Llama 3 8B 64K PoSE

Model	Likes	Downloads	VRAM
Llama 3 8B 64K PoSE 8bit	1	12	8 GB

Best Alternatives to Llama 3 8B 64K PoSE

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 winglian/Llama-3-8b-64k-PoSE.