feat: add ci pipeline (#1)

* ci: add pipeline

* ci: use conda-incubator/setup-miniconda@v3

* ci: misspelling

* ci: fix path

* ci: fix working-directory

* ci: fix compress downloads to fit ci server

* ci: use the shell bash -el

* ci: fix work dir

* ci: only load data and train toxsmi, skip rest

.github/workflows/ci.yml

@@ -0,0 +1,58 @@
+name: Make sure env is installable and models runnable 
+
+on:
+  push:
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+
+    defaults:
+      run:
+        shell: bash -el {0}
+
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Clone GT4SD
+      run: |
+        git clone https://github.com/GT4SD/gt4sd-core.git
+    
+    - uses: conda-incubator/setup-miniconda@v3
+      with:
+        activate-environment: gt4sd
+        environment-file: gt4sd-core/conda_cpu_linux.yml
+        python-version: 3.8
+        auto-activate-base: false
+        use-only-tar-bz2: true
+
+    - name: Install dependencies
+      run: |
+        conda activate gt4sd
+        pip install torch==1.12.1+cpu torchvision==0.13.1+cpu torchaudio==0.12.1 --extra-index-url https://download.pytorch.org/whl/cpu
+        pip install gt4sd
+        pip uninstall --yes toxsmi && pip install toxsmi
+
+    - name: Load datasets
+      run: |
+        conda activate gt4sd
+        mkdir data
+        python scripts/load_data.py \
+          --uniprot P05067 \
+          --affinity_type IC50 \
+          --affinity_cutoff 10000 \
+          --output_dir data/ \
+          --train_size 0.8 \
+          --binary_labels
+
+    - name: Train toxsmi model
+      run: |
+        conda activate gt4sd
+        python scripts/train_toxsmi.py \
+          --train data/train.csv \
+          --test data/valid.csv \
+          --smi data/mols.smi \
+          --language tokenizer \
+          --params config/toxsmi_conf.json \
+          --model models \
+          --name toxsmi_model