NeMo OpenFold LangChain for NVIDIA GenAI Contest
- nemo-toolkit 1.23.0
- langchain 0.2.5
-
Hardware Requirements: Ensure you have access to NVIDIA GPUs, ideally A100, to run the models efficiently. In my case I rent A100 GPU from Digital Ocean Paperspace. Please see the screenshot. Disk Size: 1TB
-
venv:
Ensure you have Python 3.10 or later installed.
cd ~ python3.11 -m venv venv source venv/bin/activate
-
First resolve YouTokenToMe Cython issue:
cd ~ git clone [email protected]:aidatatools/NeMoOpenFoldChain.git cd NeMoOpenFoldChain cd YouTokenToMe pip install -e .
-
NeMo Toolkit:
cd ~ cd NeMoOpenFoldChain pip install "nemo_toolkit[all]"
-
Install LangChain Dependencies:
cd ~ cd NeMoOpenFoldChain pip install langchain
-
Install OpenFold
#Install conda and then mamba cd ~ wget https://repo.anaconda.com/miniconda/Miniconda3-py311_24.4.0-0-Linux-x86_64.sh chmod +x Miniconda3-*-Linux-x86_64.sh ./Miniconda3-*-Linux-x86_64.sh
If you'd prefer that conda's base environment not be activated on startup, run the following command when conda is activated:
conda config --set auto_activate_base false
# Install Mamba via Conda cd ~ source ~/.bashrc conda install -c conda-forge mamba #~/miniconda3/bin/conda install mamba -n base -c conda-forge
https://openfold.readthedocs.io/en/latest/Installation.html
#In my environment, I have CUDA 12.1, so pay attention to the branch name cd ~ git clone -b pl_upgrades https://github.com/aqlaboratory/openfold.git cd openfold mamba env create -n openfold_env -f environment.yml #Activate the environment, e.g conda activate openfold_env scripts/install_third_party_dependencies.sh conda deactivate conda activate openfold_env export LD_LIBRARY_PATH=$CONDA_PREFIX/lib:$LD_LIBRARY_PATH ./scripts/download_openfold_params.sh ~/openfold/resources #Checking your build with unit tests: scripts/run_unit_tests.sh