Welcome to the official repository for the AI Acceleration with AMD hands-on lab at HPE Tech Jam Vienna 2026! This repository is dedicated to sharing sample files, datasets, and materials relevant to the hands-on lab.
- fine_tuning_dataset.jsonl A JSONL file containing sample data for fine-tuning language models. Each line represents a training example with a prompt and corresponding completion.
-
vllm_bench_setup.sh Bash script to set up a Python virtual environment and install the vLLM library for benchmarking. Run this before using the benchmark script.
-
vllm_bench.sh Bash script to run a benchmark using the vLLM library against a chat completions API endpoint.
- streamed_response.py Example Python script demonstrating how to interact with a chat completions API using streamed responses. This approach allows the response to appear incrementally—simulating a real-time typing effect—rather than displaying the entire output at once.
- comfyui_sample.json A sample ComfyUI workflow (canvas file) for generating images using Flux model. Drag and drop this file into ComfyUI to load the workflow.
These materials were prepared for a hands-on lab run at the HPE Tech Jam Vienna 2026 event. The repository is published publicly to allow participants free access during the workshop.
This repository contains sample scripts and datasets intended for demonstration purposes only. Some example scripts in this repo use insecure defaults (e.g., disabling TLS verification); do not use those defaults in production. Always follow best practices for security, data privacy, and compliance when deploying AI solutions in real-world scenarios.
This project is licensed under the MIT License — see the included LICENSE file for full terms.