Fix imports and add pyproject script

.gitignore

@@ -0,0 +1,2 @@
+outputs
+*.jsonl

pyproject.toml

@@ -5,6 +5,7 @@ description = "Add your description here"
 readme = "README.md"
 requires-python = ">=3.13"
 dependencies = [
+    "accelerate>=1.11.0",
     "boto3>=1.40.18",
     "matplotlib>=3.10.6",
     "numpy>=2.3.3",
@@ -20,18 +21,15 @@ dependencies = [
 requires = ["hatchling"]
 build-backend = "hatchling.build"
 
-[tool.hatch.build.targets.wheel]
-only-include = [
-    "evaluation.py",
-    "plot_precision_vs_cache_hit_ratio.py",
-]
-
 [dependency-groups]
 dev = [
     "black>=25.1.0",
     "isort>=6.0.1",
     "pylint>=3.3.7",
 ]
+llm-sim-eval = [
+    "llm-sim-eval>=0.2.0",
+]
 
 [tool.black]
 line-length = 120
@@ -43,3 +41,6 @@ line_length = 120
 
 [tool.pylint]
 max-line-length = 120
+
+[project.scripts]
+langcache-eval = "customer_analysis.cli.evaluation:run"

evaluation.py → src/customer_analysis/cli/evaluation.py

@@ -16,7 +16,7 @@
 except ImportError:
     HAS_LLM_SIM_EVAL = False
 
-from src.customer_analysis import (
+from customer_analysis import (
     FileHandler,
     generate_plots,
     load_data,
@@ -71,9 +71,7 @@ def run_chr_analysis(queries: pd.DataFrame, args):
 
     # Sweep cache hit ratios
     similarity_scores = queries["best_scores"].values
-    results_df = sweep_thresholds_on_results(
-        pd.DataFrame({"similarity_score": similarity_scores})
-    )
+    results_df = sweep_thresholds_on_results(pd.DataFrame({"similarity_score": similarity_scores}))
 
     # Save sweep results
     FileHandler.write_csv(results_df, args.output_dir, "chr_sweep.csv")
@@ -169,7 +167,7 @@ def main(args):
     print("\nDone!")
 
 
-if __name__ == "__main__":
+def run():
     parser = argparse.ArgumentParser(
         description="Semantic cache evaluation: CHR-only (default) or full evaluation with LLM-as-a-Judge (--full)"
     )
@@ -228,3 +226,7 @@ def main(args):
     args = parser.parse_args()
 
     main(args)
+
+
+if __name__ == "__main__":
+    run()

src/customer_analysis/data_processing.py

@@ -1,9 +1,9 @@
 import pandas as pd
 import torch
 
-from src.customer_analysis.embedding_interface import NeuralEmbedding
-from src.customer_analysis.file_handler import FileHandler
-from src.customer_analysis.query_engine import RedisVectorIndex
+from customer_analysis.embedding_interface import NeuralEmbedding
+from customer_analysis.file_handler import FileHandler
+from customer_analysis.query_engine import RedisVectorIndex
 
 RANDOM_SEED = 42
 

src/customer_analysis/file_handler.py

@@ -3,7 +3,7 @@
 import matplotlib.pyplot as plt
 import pandas as pd
 
-from src.customer_analysis.s3_util import s3_upload_dataframe_csv, s3_upload_matplotlib_png
+from customer_analysis.s3_util import s3_upload_dataframe_csv, s3_upload_matplotlib_png
 
 
 def make_output_path(output_dir: str, filename: str) -> str:

src/customer_analysis/plotting.py

@@ -1,7 +1,7 @@
 import matplotlib.pyplot as plt
 import pandas as pd
 
-from src.customer_analysis.file_handler import FileHandler
+from customer_analysis.file_handler import FileHandler
 
 
 def generate_plots(