Text Classification
PEFT
Safetensors
English
safety
guard
agent-safety
agent-trajectories
lora
alert-only
fast
Instructions to use LyzrAI/IntuitionGuard-3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use LyzrAI/IntuitionGuard-3B with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("mistralai/Ministral-3-3B-Instruct-2512-BF16") model = PeftModel.from_pretrained(base_model, "LyzrAI/IntuitionGuard-3B") - Notebooks
- Google Colab
- Kaggle
IntuitionGuard-3B: three w25 seed adapters (H54), model card, scorer, evaluation reports
ab3b173 verified |
Download evaluation/README.md from LyzrAI/IntuitionGuard-3B: direct link, hf CLI and curl.
- Browser
- Download file 395 Bytes
-
https://huggingface.co/LyzrAI/IntuitionGuard-3B/resolve/main/evaluation/README.md
- Command line
-
hf download hf://LyzrAI/IntuitionGuard-3B/evaluation/README.md
-
curl -L -o README.md https://huggingface.co/LyzrAI/IntuitionGuard-3B/resolve/main/evaluation/README.md
395 Bytes
Evaluation reports from the project harness (JSON). h54_3b_w25*.json: per-seed 3B metrics and paired tests against the released 4B guards and the calibrated cut; h54_ens_3b.json: the 3-seed ensemble; h49.json: the released AgentDoG1.5-Unified-4B in the same harness and every arm of ours against it. Set names: rjudge, assebench_strict, atbench, atbench_codex, atbench_claw(_native), tracesafe.