Llama Prompt Guard 2 86M
llama
86M-parameter (mDeBERTa-base) multilingual classifier that detects prompt injection and jailbreak attempts
Open WeightsLlama 4 Community License
Context Window
512
Max Output
512
Temperature
No
Open Weights
Yes
Knowledge Cutoff
N/A
Released
2025-04-29
Last Updated
2025-04-29
License
Llama 4 Community License
Modalities
Input:Text
→Output:Text
Available Providers (0)
Tidak ada provider yang terdaftar untuk model ini.