LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

Llama Guard 3 8B by meta-llama

 »  All LLMs  »  meta-llama  »  Llama Guard 3 8B   URL Share it on

  327878   
  Arxiv:2204.05862   Arxiv:2308.01263   Arxiv:2312.06674   Arxiv:2407.21783   Autotrain compatible Base model:finetune:meta-llama... Base model:meta-llama/llama-3....   Conversational   En   Endpoints compatible   Facebook   Llama   Llama-3   Meta   Pytorch   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/meta-llama/Llama-Guard-3-8B 

Llama Guard 3 8B Benchmarks

LLME Score: 0.32292
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Llama Guard 3 8B (meta-llama/Llama-Guard-3-8B)

Llama Guard 3 8B Parameters and Internals

Model Type 
content safety classification, LLM
Use Cases 
Areas:
Content moderation, Security systems, Search and code interpretation tools
Applications:
Enterprise content management systems, Search optimizations and secure AI deployments
Primary Use Cases:
Moderating harmful content in AI-generated inputs and outputs, Ensuring safety in search tool interactions and code interpretation applications
Limitations:
Performance limited by training data context and scope, Potential susceptibility to prompt injection attacks
Considerations:
Recommended for use alongside broader moderation systems for cases requiring up-to-date factual evaluations.
Additional Notes 
Dataset expansions include multilingual conversation data and challenges to border cases to reduce false positives further.
Supported Languages 
English (full), French (full), German (full), Hindi (full), Italian (full), Portuguese (full), Spanish (full), Thai (full)
Training Details 
Data Sources:
Llama Guard 1 and 2 generations, multilingual conversation data, Brave Search API query results
Methodology:
Fine-tuning on pre-trained Llama 3.1 model
Safety Evaluation 
Methodologies:
Aligned with MLCommons standardized hazards taxonomy, Internal tests against multilingual dataset, Comparison with prior versions and competitor models, Synthetic generation of safety data
Findings:
Improved safety classification while lowering false positive rates
Risk Categories:
Violent Crimes, Non-Violent Crimes, Sex-Related Crimes, Child Sexual Exploitation, Defamation, Specialized Advice, Privacy, Intellectual Property, Indiscriminate Weapons, Hate, Suicide & Self-Harm, Sexual Content, Elections, Code Interpreter Abuse
Ethical Considerations:
Focus on reducing false positives while ensuring comprehensive content moderation across categories
Responsible Ai Considerations 
Fairness:
Supports multilingual content moderation intended for a wide range of languages with consistent policy alignment.
Transparency:
Commercial conditions and usage thresholds are clearly outlined in the licensing terms.
Accountability:
Users are accountable for adhering to the Acceptable Use Policy and Meta's broader policy guidelines.
Mitigation Strategies:
Policies and thresholds are provided to manage usage levels, and community reporting mechanisms are in place for policy violations.
Input Output 
Input Format:
Prompt and response classification
Accepted Modalities:
text
Output Format:
Safety classification and content category indication
Performance Tips:
Use the model with specified transformer versions and follow community guidelines for optimal safety checks.
Release Notes 
Version:
Llama Guard 3
Date:
2024
Notes:
Improved safety evaluations, expanded multilingual capabilities, and refined moderation taxonomies.
LLM NameLlama Guard 3 8B
Repository 🤗https://huggingface.co/meta-llama/Llama-Guard-3-8B 
Base Model(s)  meta-llama/Meta-Llama-3.1-8B   meta-llama/Meta-Llama-3.1-8B
Model Size8b
Required VRAM16.1 GB
Updated2025-01-29
Maintainermeta-llama
Model Typellama
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 Languagesen
Model ArchitectureLlamaForCausalLM
Licensellama3.1
Context Length131072
Model Max Length131072
Transformers Version4.43.0.dev0
Tokenizer ClassPreTrainedTokenizerFast
Vocabulary Size128256
Torch Data Typebfloat16

Best Alternatives to Llama Guard 3 8B

Best Alternatives
Context / RAM
Downloads
Likes
...a 3 8B Instruct Gradient 1048K1024K / 16.1 GB3927680
MrRoboto ProLong 8B V4i1024K / 16.1 GB661
...o ProLongBASE Pt8 Unaligned 8B1024K / 16.1 GB240
MrRoboto BASE V2 Unholy 8B 64K1024K / 16.1 GB271
Mpasila Viking 8B1024K / 16.1 GB840
Thor V1.4 8B DARK FICTION1024K / 16.1 GB9412
41024K / 16.1 GB3220
Hel V2 8B DARK FICTION1024K / 16.1 GB220
161024K / 16.1 GB1690
...di95 LewdStorytellerMix 8B 64K1024K / 16.1 GB692
Note: green Score (e.g. "73.2") means that the model is better than meta-llama/Llama-Guard-3-8B.

Rank the Llama Guard 3 8B 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? 43470 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.
Release v20241227
  Support LLM Explorer