ssh-keygen -f hetzner
cat hetzner_box.pub | ssh -p23 uXXXXX-sub1@uXXXXX.your-storagebox.de install-ssh-key

vi ~/.config/rclone/rclone.conf

[storagebox]
type = sftp
host = uXXXXX.your-storagebox.de
user = uXXXXX
port = 23
pass = <obscured-password>

ls -la /dev/dri # expected to fail -> iGPU is disabled
vi /etc/modprobe.d/blacklist-hetzner.conf # comment out (disabled) i915 & i915_bdw
vi /etc/default/grub.d/hetzner.cfg # at GRUB_CMDLINE_LINUX_DEFAULT, 'nomodeset' to be removed
sudo grub-mkconfig -o /boot/grub/grub.cfg
sudo shutdown -r now
ls -la /dev/dri # shall give devices list
sudo lspci -v -s $(lspci | grep VGA | cut -d" " -f 1) # shall contain 'Kernel driver in use: i915'
sudo apt install intel-gpu-tools
sudo intel_gpu_top

LC_ALL=C lspci -v | grep -EA10 "3D|VGA" | grep 'prefetchable'

git clone https://github.com/flame/blis
cd blis
./configure --enable-cblas -t openmp,pthreads auto
make -j
sudo make install
cd ..
git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp
make GGML_BLIS=1 -j
LD_LIBRARY_PATH=/usr/local/lib ./llama-cli -m your_model.gguf -p "I believe the meaning of life is" -n 128

sudo su -
wget -qO - https://packages.lunarg.com/lunarg-signing-key-pub.asc | apt-key add -
# it's for Ubuntu 22.04! 24.04 shall be picked up from https://vulkan.lunarg.com/doc/view/latest/linux/getting_started_ubuntu.html
wget -qO /etc/apt/sources.list.d/lunarg-vulkan-jammy.list https://packages.lunarg.com/vulkan/lunarg-vulkan-jammy.list
apt update -y
apt-get install -y vulkan-sdk
# To verify the installation, use the command below:
vulkaninfo
vkvia # needs X11

git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp

sudo apt install cmake libvulkan-dev
cmake -B build -DGGML_VULKAN=1
cmake --build build --config Release
# Test the output binary (with "-ngl 33" to offload all layers to GPU)
./build/bin/llama-cli -m "PATH_TO_MODEL" -p "Hi you how are you" -n 50 -e -ngl 33 -t 4

# You should see in the output, ggml_vulkan detected your GPU. For example:
# ggml_vulkan: Using Intel(R) Graphics (ADL GT2) | uma: 1 | fp16: 1 | warp size: 32

git clone https://github.com/ggerganov/whisper.cpp.git
cd whisper
bash ./models/download-ggml-model.sh large-v3
cd models
python3.11 -m venv openvino_conv_env # 3.12 requires numpy woodoo: https://stackoverflow.com/questions/77364550/attributeerror-module-pkgutil-has-no-attribute-impimporter-did-you-mean/77364602#77364602
source openvino_conv_env/bin/activate
# python -m pip install --upgrade pip
# pip install setuptools
# pip install numpy==1.26.4 # for Python 3.12+
pip install -r requirements-openvino.txt
pip install openai-whisper
python convert-whisper-to-openvino.py --model large-v3
# This will produce ggml-base.en-encoder-openvino.xml/.bin IR model files. It's recommended to relocate these to the same folder as ggml models, as that is the default location that the OpenVINO extension will search at runtime.
ls ggml-base.en-encoder-openvino.*
# download https://github.com/openvinotoolkit/openvino/releases/tag/2023.0.0
sudo mkdir /opt/intel
# 24.04 instructions!
curl -L https://storage.openvinotoolkit.org/repositories/openvino/packages/2024.3/linux/l_openvino_toolkit_ubuntu24_2024.3.0.16041.1e3b88e4e3f_x86_64.tgz --output openvino_2024.3.0.tgz
tar -xf openvino_2024.3.0.tgz
sudo mv l_openvino_toolkit_ubuntu24_2024.3.0.16041.1e3b88e4e3f_x86_64 /opt/intel/openvino_2024.3.0
cd /opt/intel/openvino_2024.3.0
sudo -E ./install_dependencies/install_openvino_dependencies.sh
cd /opt/intel
sudo ln -s openvino_2024.3.0 openvino_2024
source /opt/intel/openvino_2024/setupvars.sh


# python -c "from openvino import Core; print(Core().available_devices)"
cd ..
cmake -B build -DWHISPER_OPENVINO=1
cmake --build build -j --config Release
 ./main -m models/ggml-base.en.bin -f samples/jfk.wav
 # The first time run on an OpenVINO device is slow, since the OpenVINO framework will compile the IR (Intermediate Representation) model to a device-specific 'blob'. This device-specific blob will get cached for the next run.

mkdir neo
cd neo
wget https://github.com/intel/intel-graphics-compiler/releases/download/igc-1.0.17193.4/intel-igc-core_1.0.17193.4_amd64.deb
wget https://github.com/intel/intel-graphics-compiler/releases/download/igc-1.0.17193.4/intel-igc-opencl_1.0.17193.4_amd64.deb
wget https://github.com/intel/compute-runtime/releases/download/24.26.30049.6/intel-level-zero-gpu-dbgsym_1.3.30049.6_amd64.ddeb
wget https://github.com/intel/compute-runtime/releases/download/24.26.30049.6/intel-level-zero-gpu_1.3.30049.6_amd64.deb
wget https://github.com/intel/compute-runtime/releases/download/24.26.30049.6/intel-opencl-icd-dbgsym_24.26.30049.6_amd64.ddeb
wget https://github.com/intel/compute-runtime/releases/download/24.26.30049.6/intel-opencl-icd_24.26.30049.6_amd64.deb
wget https://github.com/intel/compute-runtime/releases/download/24.26.30049.6/libigdgmm12_22.3.20_amd64.deb
sudo dpkg -i *.deb

./main -m ../../models/ggml-large-v3.bin -f ../../samples/jfk.wav -oved GPU # that works