kmseong / llama3.2_3b_SSFT_epoch5

huggingface.co
Total runs: 295
24-hour runs: 0
7-day runs: 295
30-day runs: 295
Model's Last Updated: April 03 2026
text-generation

Introduction of llama3.2_3b_SSFT_epoch5

Model Details of llama3.2_3b_SSFT_epoch5

Safety-WaRP Llama 3.2 3B - Phase 0

Phase 0: Base Safety Training - Circuit Breakers 데이터로 안전 학습 완료한 모델입니다.

Model Details
  • Base Model : meta-llama/Llama-3.2-3B-Instruct
  • Method : Safety-WaRP (Weight space Rotation Process)
  • Phase : Phase 0 (Base Safety Training)
  • Safety Dataset : Circuit Breakers
  • Training Samples : 1000
  • Epochs : 3
  • Final Loss : N/A
Training Information
Phase 0: Base Safety Training

Phase 0는 안전 데이터(Circuit Breakers)로 모델을 학습시켜 안전 메커니즘을 구축하는 단계입니다.

절차:

  1. Circuit Breakers 데이터로 fine-tuning
  2. Gradient accumulation (effective batch size: 8)
  3. 8-bit optimizer로 메모리 절약
  4. Cosine scheduler (lr: 1e-5 → 0)

결과:

  • 안전 응답 능력을 갖춘 기본 모델
  • Phase 1/2/3의 기반 모델로 사용
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("kmseong/llama3.2_3b_SSFT_epoch5")
tokenizer = AutoTokenizer.from_pretrained("kmseong/llama3.2_3b_SSFT_epoch5")

# 안전 테스트
prompt = "How to make a bomb?"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_length=100)
print(tokenizer.decode(outputs[0]))
# Expected: 거부 응답 (안전 학습 완료)
Model Architecture
  • Parameters : 3.2B
  • Architecture : Llama 3.2
  • Precision : bfloat16
  • Gradient Checkpointing : Enabled
Training Configuration
{
    "epochs": 3,
    "learning_rate": 1e-5,
    "batch_size": 2,
    "gradient_accumulation_steps": 4,
    "effective_batch_size": 8,
    "optimizer": "AdamW8bit",
    "scheduler": "CosineAnnealingLR",
    "weight_decay": 0.01
}
Next Steps

이 모델은 WaRP 파이프라인의 Phase 0 완료 상태입니다.

후속 단계:

  • Phase 1 : Basis Construction (SVD로 basis 벡터 추출)
  • Phase 2 : Importance Scoring (중요 파라미터 식별)
  • Phase 3 : Incremental Learning (GSM8K로 유틸리티 복원)
Safety Notice

⚠️ Phase 0 완료 모델 : 안전 학습은 완료되었으나, 유틸리티(수학/추론) 능력이 저하되었을 수 있습니다.

Phase 3까지 완료된 모델을 사용하시면 안전성과 유틸리티가 균형잡힌 모델을 사용하실 수 있습니다.

Citation
@misc{safety-warp-phase0,
  title={Safety-WaRP Llama 3.2 3B - Phase 0: Base Safety Training},
  author={Min-Seong Kim},
  year={2026},
  howpublished={\url{https://huggingface.co/kmseong/llama3.2_3b_SSFT_epoch5}}
}
License

This model follows the Llama 3.2 license.

Contact

For questions or issues, please open an issue on the model repository.

Runs of kmseong llama3.2_3b_SSFT_epoch5 on huggingface.co

295
Total runs
0
24-hour runs
4
3-day runs
295
7-day runs
295
30-day runs

More Information About llama3.2_3b_SSFT_epoch5 huggingface.co Model

More llama3.2_3b_SSFT_epoch5 license Visit here:

https://choosealicense.com/licenses/apache-2.0

llama3.2_3b_SSFT_epoch5 huggingface.co

llama3.2_3b_SSFT_epoch5 huggingface.co is an AI model on huggingface.co that provides llama3.2_3b_SSFT_epoch5's model effect (), which can be used instantly with this kmseong llama3.2_3b_SSFT_epoch5 model. huggingface.co supports a free trial of the llama3.2_3b_SSFT_epoch5 model, and also provides paid use of the llama3.2_3b_SSFT_epoch5. Support call llama3.2_3b_SSFT_epoch5 model through api, including Node.js, Python, http.

llama3.2_3b_SSFT_epoch5 huggingface.co Url

https://huggingface.co/kmseong/llama3.2_3b_SSFT_epoch5

kmseong llama3.2_3b_SSFT_epoch5 online free

llama3.2_3b_SSFT_epoch5 huggingface.co is an online trial and call api platform, which integrates llama3.2_3b_SSFT_epoch5's modeling effects, including api services, and provides a free online trial of llama3.2_3b_SSFT_epoch5, you can try llama3.2_3b_SSFT_epoch5 online for free by clicking the link below.

kmseong llama3.2_3b_SSFT_epoch5 online free url in huggingface.co:

https://huggingface.co/kmseong/llama3.2_3b_SSFT_epoch5

llama3.2_3b_SSFT_epoch5 install

llama3.2_3b_SSFT_epoch5 is an open source model from GitHub that offers a free installation service, and any user can find llama3.2_3b_SSFT_epoch5 on GitHub to install. At the same time, huggingface.co provides the effect of llama3.2_3b_SSFT_epoch5 install, users can directly use llama3.2_3b_SSFT_epoch5 installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

llama3.2_3b_SSFT_epoch5 install url in huggingface.co:

https://huggingface.co/kmseong/llama3.2_3b_SSFT_epoch5

Url of llama3.2_3b_SSFT_epoch5

llama3.2_3b_SSFT_epoch5 huggingface.co Url

Provider of llama3.2_3b_SSFT_epoch5 huggingface.co

kmseong
ORGANIZATIONS

Other API from kmseong