container.def

Bootstrap: docker
From: python:3.12.4-slim-bookworm

%files
    ./praat_scripts/FilledPauses.praat /usr/local/share/praat/FilledPauses.praat
    ./praat_scripts/syllablenucleiv3.praat /usr/local/share/praat/syllablenucleiv3.praat

%post
    export HUGGINGFACE_HUB_CACHE="/usr/local/share/huggingface"
    apt update
    apt install -y ffmpeg wget libgtk-3-0 libpulse-dev libasound2 git
    apt autoremove -y
    pip install lingualabpy==0.0.6
    pip install https://github.com/lingualab/Text2Variable/archive/refs/tags/0.0.12.zip
    #pip install git+https://github.com/lingualab/Text2Variable.git@main
    python -m spacy download en_core_web_sm
    python -m spacy download en_core_web_lg
    python -m spacy download fr_core_news_sm
    python -m spacy download fr_core_news_lg
    python -m nltk.downloader -d /usr/local/share/nltk_data words
    python -c "from transformers import AutoTokenizer, AutoModelWithLMHead, pipeline;\
    AutoTokenizer.from_pretrained('mrm8488/t5-base-finetuned-emotion');\
    AutoModelWithLMHead.from_pretrained('mrm8488/t5-base-finetuned-emotion');\
    model_path = 'cardiffnlp/twitter-xlm-roberta-base-sentiment';\
    pipeline('sentiment-analysis', model=model_path, tokenizer=model_path)"
    wget https://www.fon.hum.uva.nl/praat/praat6427_linux-intel64.tar.gz
    tar zxf praat6427_linux-intel64.tar.gz -C /usr/bin
    rm praat6427_linux-intel64.tar.gz


%environment
    export HUGGINGFACE_HUB_CACHE="/usr/local/share/huggingface"
    export PRAAT_SCRIPTS_DIR="/usr/local/share/praat"