---
title: "obliterate.exe (@chrisoo1) | Hashnode"
description: "hope!!!!!! Read the latest articles by obliterate.exe on Hashnode."
url: "https://hashnode.com/@chrisoo1"
---

@chrisoo1

# obliterate.exe

@chrisoo1 · Paris · Joined June 2026

hope!!!!!!

### [How i built a frontier level llm guardrail](https://chris001.hashnode.dev/how-i-built-a-frontier-level-llm-guardrail?utm_source=hashnode&utm_medium=feed)

This is built using the combination of microsoft presidio and qwen 3 guard , This production script offloads Presidio (CPU) and Qwen3Guard (GPU) to worker threads via asyncio.to\_thread and gathers the

4 min read · [chris001.hashnode.dev](https://chris001.hashnode.dev/how-i-built-a-frontier-level-llm-guardrail?utm_source=hashnode&utm_medium=feed)

[1d ago](https://hashnode.com/posts/how-i-built-a-frontier-level-llm-guardrail/6af31f9b57c24b226e1b586a)

### [4-bit NF4 quantization for llama guard 3](https://chris001.hashnode.dev/4-bit-nf4-quantization-for-llama-guard-3?utm_source=hashnode&utm_medium=feed)

this quantization will help for users with low vram and get desired result but note using 4 bit quantization has its drawback. import torch from transformers import AutoTokenizer, AutoModelForCausalLM

1 min read · [chris001.hashnode.dev](https://chris001.hashnode.dev/4-bit-nf4-quantization-for-llama-guard-3?utm_source=hashnode&utm_medium=feed)

[1d ago](https://hashnode.com/posts/4-bit-nf4-quantization-for-llama-guard-3/6af3153ea84302f8dc4ddd89)

### [Building the Interception Script: Middleware for Local LLMs](https://chris001.hashnode.dev/building-the-interception-script-middleware-for-local-llms?utm_source=hashnode&utm_medium=feed)

A guard model cannot defend a system if it is not placed directly in the path of network traffic. To utilize Llama Guard or Qwen3Guard effectively, you must construct an API gateway that intercepts al

7 min read · [chris001.hashnode.dev](https://chris001.hashnode.dev/building-the-interception-script-middleware-for-local-llms?utm_source=hashnode&utm_medium=feed)

[1d ago](https://hashnode.com/posts/building-the-interception-script-middleware-for-local-llms/6af30dd6df67fc3d1400e5ea)

### [Multilingual and Nuanced Filtering using Qwen3Guard](https://chris001.hashnode.dev/multilingual-and-nuanced-filtering-using-qwen3guard?utm_source=hashnode&utm_medium=feed)

While Llama Guard excels in binary classification, Alibaba's Qwen3Guard (released in late 2025) offers a more nuanced approach. Built on the Qwen3 architecture, the Qwen3Guard-Gen-8B model natively su

5 min read · [chris001.hashnode.dev](https://chris001.hashnode.dev/multilingual-and-nuanced-filtering-using-qwen3guard?utm_source=hashnode&utm_medium=feed)

[1d ago](https://hashnode.com/posts/multilingual-and-nuanced-filtering-using-qwen3guard/6af30cfb1cd13c05520bf451)

### [Implementing a Robust Input and Output Filter with Meta's Llama Guard 3](https://chris001.hashnode.dev/implementing-a-robust-input-and-output-filter-with-meta-s-llama-guard-3?utm_source=hashnode&utm_medium=feed)

When deploying a local Large Language Model (LLM), ensuring safety and alignment requires a dedicated moderation layer. Meta’s Llama Guard 3 (specifically the 8B parameter version) serves as a special

5 min read · [chris001.hashnode.dev](https://chris001.hashnode.dev/implementing-a-robust-input-and-output-filter-with-meta-s-llama-guard-3?utm_source=hashnode&utm_medium=feed)

[1d ago](https://hashnode.com/posts/implementing-a-robust-input-and-output-filter-with-meta-s-llama-guard-3/6af30bcbf8a5aa7c47630cea)