Uchile Human-Robot Interface

Repo for Speech recognition and text to speech

Requirements

In order to install the requirements you must install python 3.9 or greater.

Python 3.9 installation

(FG): Hay que tener ojo, esto solo sirve para el speech to text y el text to speech. Para instalar ollama, chromadb y todo eso, hay que revisar porque parece que hay que tener una versión más actualizada de python (más que python3.9) (pero no cacho porque solamente alcancé a usar el stt y el tts).

1. Update the packages list and install the prerequisites:

sudo apt update
sudo apt install software-properties-common

2. Add the deadsnakes PPA to your system’s sources list:

sudo add-apt-repository ppa:deadsnakes/ppa

3. Once the repository is enabled, you can install Python 3.9 by executing:

sudo apt install python3.9

4. Verify that the installation was successful by typing:

python3.9 --version

# Output:
Python 3.9.1+

Virtual Environment creation

To avoid conflicting installations of python packages, we recommend that you create a virtual environment (you can choose the name of your virtual environment):

cd ~/
python3.9 -m venv <name-of-environment>

Then to activate it, you must run the following command:

source ~/<name-of-environment>/bin/activate

CUDA clean install

Install CUDA and CUDNN

First you must update your OS

sudo apt update && sudo apt upgrade

Remove all CUDA, CUDNN and Nvidia Drivers:

sudo apt-get remove --purge -y '*nvidia*' '*cuda*' 'libcudnn*' 'libnccl*' '*cudnn*' '*nccl*'
sudo apt-get autoremove --purge -y
sudo apt-get clean

Check if there's any installed packages left:

dpkg -l | grep -E 'nvidia|cuda|cudnn|nccl'

If there is any packages left, you must run:

packages=('*nvidia*' '*cuda*' 'libcudnn*' 'libnccl*' '*cudnn*' '*nccl*')
for pack in "${packages[@]}"; do
    echo "Removing $pack..."
    sudo apt remove --purge -y "$pack"
done

Detecting and managing drivers on ubuntu

ubuntu-drivers devices

We will install the NVIDIA driver tagged recommended — Which indicates which drivers are recommended for each piece of hardware based on compatibility and performance.

sudo ubuntu-drivers autoinstall

Install Nvidia Drivers

My recommended version is 555, change “XYZ” in the following command to your recommended driver:

sudo apt install nvidia-driver-XYZ

Reboot the system for these changes to take effect.

reboot

Check Installation

After reboot verify that the following command works (in order to verify that the nvidia drivers are properly installed)

nvidia-smi

It's recommended that before installing CUDA you check PyTorch's website to install the newest supported version of PyTorch (and its corresponding CUDA version). In that website you can find the CUDA versions which are supported by torch. You should search the CUDA versions here.

In our case we'll be installing CUDA 12.4 for Ubuntu 20.04 with architecture x86_64. Before installing, you must check for any .deb or .pin files in home directory.

cd ~/
rm cuda*

Now we proceed with the installation:

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pin

sudo mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600

wget https://developer.download.nvidia.com/compute/cuda/12.4.0/local_installers/cuda-repo-ubuntu2004-12-4-local_12.4.0-550.54.14-1_amd64.deb

sudo dpkg -i cuda-repo-ubuntu2004-12-4-local_12.4.0-550.54.14-1_amd64.deb

sudo cp /var/cuda-repo-ubuntu2004-12-4-local/cuda-*-keyring.gpg /usr/share/keyrings/

sudo apt-get update

sudo apt-get -y install cuda-toolkit-12-4

Now we check if CUDA is installed correctly:

nvcc --version

If you are not getting the CUDA version as output, do the following:

Creating a symlink for easier reference

sudo ln -s /usr/local/cuda-12.4 /usr/local/cuda

Add the CUDA paths to your .bashrc file to ensure they are set up every time you open a terminal.

# if you're using bash terminal:
echo 'export PATH=/usr/local/cuda-12.4/bin:$PATH' >> ~/.bashrc

# if you're using zsh terminal:
echo 'export PATH=/usr/local/cuda-12.4/bin:$PATH' >> ~/.zshrc

Then close the terminal and open a new one. Then check the installation with the nvcc --version command. If that command outputs something similar to :

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Thu_Mar_28_02:18:24_PDT_2024
Cuda compilation tools, release 12.4, V12.4.131
Build cuda_12.4.r12.4/compiler.34097967_0

Then the CUDA installation was succesful.

Installing CUDNN

Now we'll install CUDNN. To check the compatible CUDNN version with your CUDA installation you must Check Here

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-keyring_1.1-1_all.deb
sudo dpkg -i cuda-keyring_1.1-1_all.deb
sudo apt-get update
sudo apt-get -y install cudnn9-cuda-12 # This command is for our case only, you must check the compatible CUDNN version for your CUDA installation

Installation of Pip packages

Activating the created virtual environment from earlier:

source ~/<name-of-environment>/bin/activate
pip3 install torch torchvision torchaudio TTS