What are the hardware requirements for Meta Llama 3 8B Hf?

Meta Llama 3 8B Hf requires approximately 16.1 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 Meta Llama 3 8B Hf and how large is it?

Meta Llama 3 8B Hf is developed by Undi95, a model with 8b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Meta Llama 3 8B Hf perform on standard benchmarks?

Meta Llama 3 8B Hf has the following published scores: MMLU 66.49. Compare against reference models on this page or on the LLM Explorer leaderboards.

Meta Llama 3 8B Hf by Undi95 — VRAM 16.1GB, 8K context

Name: Meta Llama 3 8B Hf
Author: Undi95

Meta Llama 3 8B Hf is an open-source language model by Undi95. Features: 8b LLM, VRAM: 16.1GB, Context: 8K, License: other, LLM Explorer Score: 0.16, Arc: 59.2, HellaSwag: 82, MMLU: 66.5, GSM8K: 45.3.

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

Model Card on HF 🤗: https://huggingface.co/Undi95/Meta-Llama-3-8B-hf

Meta Llama 3 8B Hf Benchmarks

ARC: 59.22 vs 96.7 (so35)^-38.8%

HellaSwag: 82.02 vs 95.3 (gpt4)^-13.9%

MMLU: 66.49 vs 88.3 (so35)^-24.7%

GSM8K: 45.34 vs 96.4 (so35)^-53%

LLME Score: 0.15856

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

Meta Llama 3 8B Hf (Undi95/Meta-Llama-3-8B-hf)

🌟 Advertise your project 🚀

Meta Llama 3 8B Hf Parameters and Internals

Model Type

text generation

Use Cases

Areas:

Commercial, Research

Applications:

Instruction tuned models for assistant-like chat

Primary Use Cases:

Natural language generation, Multilingual dialogue interactions

Limitations:

Out-of-the-box use only in English, Potential inaccurate or biased responses

Considerations:

Developers should fine-tune based on specific needs.

Additional Notes

100% carbon emissions offset by Meta’s sustainability program.

Supported Languages

en (high)

Training Details

Data Sources:

publicly available online data

Data Volume:

15 trillion tokens

Methodology:

Supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF)

Context Length:

8000

Hardware Used:

H100-80GB GPU with a cumulative 7.7M GPU hours

Model Architecture:

Auto-regressive transformer architecture

Safety Evaluation

Methodologies:

Red teaming exercises, Adversarial evaluations

Risk Categories:

CBRNE, Cyber Security, Child Safety

Ethical Considerations:

Leverages best practices for safety and responsible deployment.

Responsible Ai Considerations

Fairness:

Inclusive and open approach, aiming to serve diverse user needs and perspectives.

Accountability:

Developers responsible for end-user safety evaluations.

Mitigation Strategies:

Tools like Meta Llama Guard 2 and Code Shield for layering safety measures.

Input Output

Input Format:

text

Accepted Modalities:

text

Output Format:

text and code

Performance Tips:

Fine-tune with language-specific data where appropriate.

Release Notes

Version:

Meta Llama 3 (8B, 70B)

Date:

April 18, 2024

Notes:

Initial release of pre-trained and instruction tuned variants.

LLM Name	Meta Llama 3 8B Hf
Repository 🤗	https://huggingface.co/Undi95/Meta-Llama-3-8B-hf
Model Size	8b
Required VRAM	16.1 GB
Updated	2026-05-22
Maintainer	Undi95
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
License	other
Context Length	8192
Model Max Length	8192
Transformers Version	4.40.0.dev0
Tokenizer Class	PreTrainedTokenizerFast
Vocabulary Size	128256
Torch Data Type	bfloat16

Quantized Models of the Meta Llama 3 8B Hf

Model	Likes	Downloads	VRAM
Meta Llama 3 8B AWQ	2	13	5 GB
Meta Llama 3 8B Hf AWQ	0	15	5 GB
Meta Llama 3 8B AWQ	0	15	5 GB
Meta Llama 3 8B AWQ	0	5	5 GB
Meta Llama 3 8B AWQ	0	5	5 GB

Best Alternatives to Meta Llama 3 8B Hf

Best Alternatives	Context / RAM	Downloads	Likes
...otron 8B UltraLong 4M Instruct	4192K / 32.1 GB	156	119
UltraLong Thinking	4192K / 16.1 GB	10	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	119	17
Cthulhu 8B V1.4	1048K / 16.1 GB	12	9
...otron 8B UltraLong 1M Instruct	1048K / 32.1 GB	2059	57
...a 3.1 8B UltraLong 1M Instruct	1048K / 32.1 GB	1387	29
...xis Bookwriter Llama3.1 8B Sft	1048K / 16.1 GB	17	4
Zero Llama 3.1 8B Beta6	1048K / 16.1 GB	4	1

Note: green Score (e.g. "73.2") means that the model is better than Undi95/Meta-Llama-3-8B-hf.

Rank the Meta Llama 3 8B Hf Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

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

What open-source LLMs or SLMs are you in search of? 53972 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Check out Ag3ntum — our secure, self-hosted AI agent for server management.

Release v20260328a

Support LLM Explorer

Meta Llama 3 8B Hf by Undi95