Remove install.sh

.github/workflows/python-app.yml

@@ -26,9 +26,7 @@ jobs:
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        chmod +x install.sh
         pip install -e .
-        ./install.sh
         pip install flake8 pytest
     - name: Test with pytest
       run: |

README.md

@@ -21,21 +21,7 @@ A comprehensive toolkit for applying active learning techniques to natural langu
 
 ## 🔧 Installation
 
-1. Clone the repository:
-   ```bash
-   git clone https://github.com/Aktsvigun/atgen.git
-   cd atgen
-   ```
-
-2. Run the installation script:
-   ```bash
-   bash install.sh
-   ```
-
-   This will install:
-   - All required Python packages
-   - External metrics (submodlib, AlignScore)
-   - Required NLP resources
+`pip install atgen`
 
 ## 🚀 Usage
 
@@ -115,4 +101,4 @@ If you use this toolkit in your research, please cite:
   url = {https://github.com/Aktsvigun/atgen},
   year = {2025},
 }
-``` 
+```

pyproject.toml

@@ -4,15 +4,15 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "atgen"
-version = "0.0.0"
+version = "0.1.0"
 authors = [
   { name = "List of contributors: https://github.com/Aktsvigun/atgen/graphs/contributors", email = "artemshelmanov@gmail.com" },
 ]
 description = "Toolkit for Active Learning in Generative Tasks"
 readme = "README.md"
 
 keywords = ["NLP", "deep learning", "transformer", "pytorch", "peft", "inference", "active learning"]
-license = {file = "LICENSE.md"}
+license-files = ["LICENSE.md"]
 
 requires-python = ">=3.10"
 dynamic = ["dependencies"]

requirements.txt

@@ -1,3 +1,4 @@
+alignscore-SpeedOfMagic
 accelerate==1.5.2
 anthropic==0.49.0
 benepar==0.2.0

src/atgen/metrics/metrics.py

@@ -62,7 +62,7 @@
     # Going up 3 levels from metrics.py: src/atgen/metrics -> repository root
     os.path.join(
         Path(__file__).parents[3],
-        "external_metrics/AlignScore/model/AlignScore-base.ckpt",
+        "cache/AlignScore-base.ckpt",
     ),
 )
 

src/atgen/run_scripts/run_active_learning.py

@@ -32,6 +32,7 @@ def run_active_learning(config, workdir: Union[str, Path]):
         maybe_get_few_shot_examples,
         get_output_column_name_for_phase,
     )
+    from atgen.utils.installers import install_spacy, install_nltk
     from atgen.utils.load_model_tokenizer import load_model_tokenizer
     from atgen.utils.prepare_model_for_training import prepare_model_for_training
     from atgen.utils.training_utils import get_trainer
@@ -50,6 +51,10 @@ def run_active_learning(config, workdir: Union[str, Path]):
         check_performance_against_requirements,
     )
 
+    # TODO Figure out how to stop downloading it every time
+    install_spacy()
+    install_nltk()
+
     seed = config.seed
     cache_dir = config.cache_dir
     input_column_name = config.data.input_column_name

src/atgen/utils/installers.py

@@ -0,0 +1,10 @@
+import subprocess
+
+
+def install_spacy():
+    subprocess.run("python -m spacy download en_core_web_sm", shell=True)
+
+def install_nltk():
+    import nltk
+    nltk.download('punkt')
+    nltk.download('punkt_tab')