Llama-3.2-2B-40pct-Compressed-3B-EN-V1

Llama-3.2-2B — 40% Compressed from Llama-3.2-3B (English)

Built with Llama. This repository is part of the Efficient and Robust AI System (E-AI) Project by Vincent-Daniel Yun. This model is a compressed edition of meta-llama/Llama-3.2-3B with 11 of 28 transformer layers removed (17 layers remain, ≈2.11B parameters), recovered training-free. It is a base (non–chat-tuned) model.

🔗 Project: https://www.worldwidedaniel.com/eai-project 📅 Release: 2026-06-28 · Version: V1

⚖️ License: governed by the Llama 3.2 Community License Agreement (https://huggingface.co/meta-llama/Llama-3.2-3B/blob/main/LICENSE.txt), and subject to the Llama 3.2 Acceptable Use Policy (https://www.llama.com/llama3_2/use-policy). By using this model you agree to those terms. "Llama" is a trademark of Meta Platforms, Inc. Built with Llama.

⚠️ Language: English-focused. Use as a discrimination / classification engine — open-ended long-form generation is degraded by compression (see PPL). For factual questions use retrieval (RAG).

About E-AI

Modern AI is powerful but heavy. The E-AI (Efficient-AI) project builds compact yet capable AI — making every model lightweight and fast — so AI can assist people in urgent, high-stakes moments.

Method

The pruning method and the training-free recovery method are proprietary, undisclosed methods created by Vincent-Daniel Yun and are not released. Only the resulting model is shared.

Results (measured, full test sets)

PPL on 2048-token context (↓ better); downstream & MMLU are 0-shot via lm-eval-harness (↑ better).

Task-suitability (vs base Llama-3.2-3B)

(Discrimination is largely preserved; generation/PPL degrades with compression. Yes/No safety-F1 benchmarks are omitted as the non-chat base is poorly calibrated for them.)

Efficiency (measured, fp16, single GPU)

Usage — Transformers

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
m = AutoModelForCausalLM.from_pretrained("daniel-eai/Llama-3.2-2B-40pct-Compressed-3B-EN-V1", trust_remote_code=True, dtype=torch.float16, device_map="cuda")
tok = AutoTokenizer.from_pretrained("daniel-eai/Llama-3.2-2B-40pct-Compressed-3B-EN-V1", trust_remote_code=True)
ids = tok("The capital of France is", return_tensors="pt").to("cuda")
print(tok.decode(m.generate(**ids, max_new_tokens=20)[0]))

trust_remote_code=True is required (custom decoder layer in modeling_llama_recovered.py).

License & Acknowledgements

Derivative of Llama 3.2, distributed under the Llama 3.2 Community License. Built with Llama. Thanks to Prof. Sai Praneeth Karimireddy (USC) and Prof. Sunwoo Lee (Inha University) for guidance, and to Meta for releasing Llama 3.2.