Donato Capitella

@UCP0QFok6EimQYTMj5qOLNow - 76.2K subscribers

I'm a Software Engineer and Ethical Hacker, but mostly a tech enthusiast who likes to discover how things work by breaking them apart.

Home Videos Live Playlists

Fast Finetuning of Gemma-3, Qwen-3 and GPT-OSS on Strix Halo using Unsloth and Multi-Node Setups

Fast Finetuning of Gemma-3, Qwen-3 and GPT-OSS on Strix Halo using Unsloth and Multi-Node Setups Donato Capitella

4.4K views - 3 days ago

ComfyUI Strix Halo Toolbox for Image and Video Generation (LTX2, Qwen Image, WAN 2.2, Hunyuan 1.5)

ComfyUI Strix Halo Toolbox for Image and Video Generation (LTX2, Qwen Image, WAN 2.2, Hunyuan 1.5) Donato Capitella

6K views - 3 weeks ago

Low-Latency Strix Halo Cluster with RDMA (RoCE/Intel E810) and vLLM, Framework Desktop Boards

Low-Latency Strix Halo Cluster with RDMA (RoCE/Intel E810) and vLLM, Framework Desktop Boards Donato Capitella

18.3K views - 4 weeks ago

Kimi-K2(1T)/GLM 4.7(355B) on a 4-Node Strix Halo Cluster - 512GB of Unified Memory

Kimi-K2(1T)/GLM 4.7(355B) on a 4-Node Strix Halo Cluster - 512GB of Unified Memory Donato Capitella

76.9K views - 4 weeks ago

ROCm+Linux Support on Strix Halo: It's finally stable in 2026!

ROCm+Linux Support on Strix Halo: It's finally stable in 2026! Donato Capitella

13.3K views - 4 weeks ago

Video and Image Generation on AMD R9700 AI PRO (Qwen Image, Wan 2.2, Hunyuan 1.5)

Video and Image Generation on AMD R9700 AI PRO (Qwen Image, Wan 2.2, Hunyuan 1.5) Donato Capitella

15.3K views - 1 month ago

Running vLLM on Strix Halo (AMD Ryzen AI MAX) + ROCm Performance Updates

Running vLLM on Strix Halo (AMD Ryzen AI MAX) + ROCm Performance Updates Donato Capitella

28.9K views - 1 month ago

vLLM on Dual AMD Radeon 9700 AI PRO: Tutorials, Benchmarks (vs RTX 5090/5000/4090/3090/A100)

vLLM on Dual AMD Radeon 9700 AI PRO: Tutorials, Benchmarks (vs RTX 5090/5000/4090/3090/A100) Donato Capitella

12.3K views - 1 month ago

Dual AMD Radeon 9700 AI PRO: Building a 64GB LLM/AI Server with Llama.cpp

Dual AMD Radeon 9700 AI PRO: Building a 64GB LLM/AI Server with Llama.cpp Donato Capitella

23.2K views - 3 months ago

Building a Two-Node AMD Strix Halo Cluster for LLMs with llama.cpp RPC (MiniMax-M2 & GLM 4.6)

Building a Two-Node AMD Strix Halo Cluster for LLMs with llama.cpp RPC (MiniMax-M2 & GLM 4.6) Donato Capitella

14.9K views - 4 months ago

Finetuning LLMs on Strix Halo – Full, LoRA, and QLoRA on Gemma-3, Qwen-3, and GPT-OSS-20B

Finetuning LLMs on Strix Halo – Full, LoRA, and QLoRA on Gemma-3, Qwen-3, and GPT-OSS-20B Donato Capitella

15.1K views - 4 months ago

VibeVoice (Speech Generation/Voice Cloning) on Framework Desktop with Strix Halo (AMD AI Ryzen MAX+)

VibeVoice (Speech Generation/Voice Cloning) on Framework Desktop with Strix Halo (AMD AI Ryzen MAX+) Donato Capitella

6.3K views - 5 months ago

Run Qwen Image and WAN 2.2 on Framework Desktop with Strix Halo (AMD AI Ryzen MAX+ 395) - Full Guide

Run Qwen Image and WAN 2.2 on Framework Desktop with Strix Halo (AMD AI Ryzen MAX+ 395) - Full Guide Donato Capitella

21.8K views - 6 months ago

GLM 4.5-Air-106B and Qwen3-235B on AMD "Strix Halo" AI Ryzen MAX+ 395 (HP Z2 G1a Mini Workstation)

GLM 4.5-Air-106B and Qwen3-235B on AMD "Strix Halo" AI Ryzen MAX+ 395 (HP Z2 G1a Mini Workstation) Donato Capitella

94.6K views - 7 months ago

LLM Chronicles #6.9: Design Patterns for Securing LLM Agents Against Prompt Injection (Paper Review)

LLM Chronicles #6.9: Design Patterns for Securing LLM Agents Against Prompt Injection (Paper Review) Donato Capitella

21.4K views - 8 months ago

Zero-Click M365 Copilot Exploit 'EchoLeak' - Deep Dive

Zero-Click M365 Copilot Exploit 'EchoLeak' - Deep Dive Donato Capitella

11.9K views - 8 months ago

Spikee #4 - Bypassing LLM Guardrails (Anti-spotlighting, Best of N attacks)

Spikee #4 - Bypassing LLM Guardrails (Anti-spotlighting, Best of N attacks) Donato Capitella

5.7K views - 9 months ago

Understanding Reasoning LLMs (o1/o3, DeepSeek-R1, Gemini Thinking, Grok 3, Claude 3.7)

Understanding Reasoning LLMs (o1/o3, DeepSeek-R1, Gemini Thinking, Grok 3, Claude 3.7) Donato Capitella

21.1K views - 11 months ago

Spikee #3 - Prompt Injection Testing of a GenAI Feature in an LLM Application

Spikee #3 - Prompt Injection Testing of a GenAI Feature in an LLM Application Donato Capitella

12K views - 1 year ago

Spikee #2 - Hands-on Prompt Injection Testing with Custom Dataset

Spikee #2 - Hands-on Prompt Injection Testing with Custom Dataset Donato Capitella

1.6K views - 1 year ago

Spikee #1 - LLM Benchmarking/Testing Tool for Prompt Injection

Spikee #1 - LLM Benchmarking/Testing Tool for Prompt Injection Donato Capitella

10.1K views - 1 year ago

Hacking and Securing LLM Applications - Should you let ChatGPT Control Your Browser [DeepSec 2024]

Hacking and Securing LLM Applications - Should you let ChatGPT Control Your Browser [DeepSec 2024] Donato Capitella

12.9K views - 1 year ago

Multi-Chain Prompt Injection and Jailbreaking of LLM Applications

Multi-Chain Prompt Injection and Jailbreaking of LLM Applications Donato Capitella

10.5K views - 1 year ago

LLM Chronicles #6.7: Faithfulness Checks for RAG pipelines (gpt4o-mini, Llama-index, RAGAS, Azure)

LLM Chronicles #6.7: Faithfulness Checks for RAG pipelines (gpt4o-mini, Llama-index, RAGAS, Azure) Donato Capitella

13.4K views - 1 year ago

LLM Chronicles #6.6: Hallucination Detection and Evaluation for RAG systems (RAGAS, Lynx)

LLM Chronicles #6.6: Hallucination Detection and Evaluation for RAG systems (RAGAS, Lynx) Donato Capitella

25.5K views - 1 year ago

LLM Chronicles #6.3a: OpenAI CLIP for Zero-Shot Image Classification and Similarity

LLM Chronicles #6.3a: OpenAI CLIP for Zero-Shot Image Classification and Similarity Donato Capitella

1.9K views - 1 year ago

LLM Chronicles #6.5: Build LLM Agents from Scratch: PAL, ReAct & Langchain in 1 Hour

LLM Chronicles #6.5: Build LLM Agents from Scratch: PAL, ReAct & Langchain in 1 Hour Donato Capitella

35.5K views - 1 year ago

Microsoft 365 Copilot Hack Breakdown [Black Hat 2024]

Microsoft 365 Copilot Hack Breakdown [Black Hat 2024] Donato Capitella

27.2K views - 1 year ago

LLM Chronicles #6.4: LLM Agents with ReAct (Reason + Act)

LLM Chronicles #6.4: LLM Agents with ReAct (Reason + Act) Donato Capitella

52K views - 1 year ago

LLM Chronicles #6.3: Multi-Modal LLMs for Image, Sound and Video

LLM Chronicles #6.3: Multi-Modal LLMs for Image, Sound and Video Donato Capitella

32K views - 1 year ago

The ADAM Optimizer, Momentum and RMSProp

The ADAM Optimizer, Momentum and RMSProp Donato Capitella

1K views - 1 year ago

Update Strategies: Full Batch / Incremental, Stochastic Gradient Descent with Mini-Batches

Update Strategies: Full Batch / Incremental, Stochastic Gradient Descent with Mini-Batches Donato Capitella

313 views - 1 year ago

The Training Loop of a Neural Network

The Training Loop of a Neural Network Donato Capitella

264 views - 1 year ago

LLM Chronicles #5.6: Limitations & Challenges of LLMs

LLM Chronicles #5.6: Limitations & Challenges of LLMs Donato Capitella

26.4K views - 1 year ago

Gradient Descent to Train a Neural Network

Gradient Descent to Train a Neural Network Donato Capitella

240 views - 1 year ago

Dataset Split (Train, Test, Validation)

Dataset Split (Train, Test, Validation) Donato Capitella

215 views - 1 year ago

Data Normalization

Data Normalization Donato Capitella

720 views - 1 year ago

What's the Loss Function?

What's the Loss Function? Donato Capitella

260 views - 1 year ago

Preparing Datasets for Training Neural Networks

Preparing Datasets for Training Neural Networks Donato Capitella

429 views - 1 year ago

[Google Gemini] Prompt Injection via Email for Social Engineering Attacks

[Google Gemini] Prompt Injection via Email for Social Engineering Attacks Donato Capitella

687 views - 1 year ago

Indirect Prompt Injection in Langchain/GPT4 Email Agent

Indirect Prompt Injection in Langchain/GPT4 Email Agent Donato Capitella

2.4K views - 1 year ago

Tensors and GPUs

Tensors and GPUs Donato Capitella

2K views - 1 year ago

Matrix Multiplications in Neural Networks

Matrix Multiplications in Neural Networks Donato Capitella

787 views - 1 year ago

Prompt Injection / JailBreaking a Banking LLM Agent (GPT-4, Langchain)

Prompt Injection / JailBreaking a Banking LLM Agent (GPT-4, Langchain) Donato Capitella

2.8K views - 1 year ago

Outputs of Neural Networks for Classification and Regression Tasks

Outputs of Neural Networks for Classification and Regression Tasks Donato Capitella

285 views - 1 year ago

Modelling inputs to a Neural Network

Modelling inputs to a Neural Network Donato Capitella

341 views - 1 year ago

Multi-Layer Perceptrons

Multi-Layer Perceptrons Donato Capitella

604 views - 1 year ago

Perceptrons and Artificial Neurons

Perceptrons and Artificial Neurons Donato Capitella

3.7K views - 1 year ago

[Webinar] Building LLM applications in a secure way (WithSecure™)

[Webinar] Building LLM applications in a secure way (WithSecure™) Donato Capitella

1.2K views - 1 year ago

LLM Chronicles #5.5: Running Gemma 2B and Llama-2 7B with Quantization

LLM Chronicles #5.5: Running Gemma 2B and Llama-2 7B with Quantization Donato Capitella

1.3K views - 1 year ago

LLM Chronicles #5.4: GPT, Instruction Fine-Tuning, RLHF

LLM Chronicles #5.4: GPT, Instruction Fine-Tuning, RLHF Donato Capitella

57.6K views - 1 year ago

Prompt Injection in LLM Browser Agents

Prompt Injection in LLM Browser Agents Donato Capitella

11.3K views - 1 year ago

LLM Chronicles #5.3: Fine-tuning DistilBERT for Sentiment Analysis (Lab)

LLM Chronicles #5.3: Fine-tuning DistilBERT for Sentiment Analysis (Lab) Donato Capitella

2.8K views - 1 year ago

LLM Chronicles: #5.2: Making LLMs from Transformers Part 1: BERT, Encoder-based

LLM Chronicles: #5.2: Making LLMs from Transformers Part 1: BERT, Encoder-based Donato Capitella

8.3K views - 2 years ago

LLM Chronicles #5.1: The Transformer Architecture

LLM Chronicles #5.1: The Transformer Architecture Donato Capitella

12.7K views - 2 years ago

LLM Chronicles #4.8: Adding Attention to the Language Translation RNN in PyTorch

LLM Chronicles #4.8: Adding Attention to the Language Translation RNN in PyTorch Donato Capitella

949 views - 2 years ago

LLM Chronicles #4.7: Attention Mechanism for Neural Networks

LLM Chronicles #4.7: Attention Mechanism for Neural Networks Donato Capitella

1.6K views - 2 years ago

LLM Chronicles #4.6: Building an Encoder/Decoder RNN in PyTorch to Translate from English to Italian

LLM Chronicles #4.6: Building an Encoder/Decoder RNN in PyTorch to Translate from English to Italian Donato Capitella

1.9K views - 2 years ago

LLM Chronicles #:6.2 RAG (Self-Query/Parent Document/HyDE) - Part 2

LLM Chronicles #:6.2 RAG (Self-Query/Parent Document/HyDE) - Part 2 Donato Capitella

2.6K views - 2 years ago

LLM Chronicles #4.5: Encoder / Decoder RNN for Language Translation

LLM Chronicles #4.5: Encoder / Decoder RNN for Language Translation Donato Capitella

2K views - 2 years ago