Attempt to use commit hash into code

.dockerignore

@@ -5,7 +5,6 @@
 /tests
 .venv
 .flake8
-.git
 .github
 .mypy_cache
 .pytest_cache

Dockerfile

@@ -13,9 +13,9 @@ RUN pipx run poetry export --format requirements.txt --output requirements.txt
 
 FROM nvidia/cuda:12.2.2-cudnn8-runtime-ubuntu22.04
 
-# Install Python & ffmpeg
+# Install Python & ffmpeg & Git
 RUN apt-get update && \
-    apt-get install -y python3.11-dev python3-pip python-is-python3 ffmpeg && \
+    apt-get install -y python3.11-dev python3-pip python-is-python3 ffmpeg git && \
     rm -rf /var/lib/apt/lists/*
 
 # Ensure Python 3.11 is used (for some reason, 3.10 is also installed...)
@@ -27,7 +27,6 @@ RUN update-alternatives --install /usr/bin/python python /usr/bin/python3.11 1
 # - Storing the model: /model
 RUN mkdir /src /data /model
 
-
 WORKDIR /src
 
 COPY --from=req ./requirements.txt requirements.txt
@@ -36,8 +35,8 @@ RUN python -m pip install --no-cache-dir -r requirements.txt
 # copy the rest into the source dir
 COPY ./ /src
 
-# Keep commit hash for provenance of the versioning
-ARG GIT_COMMIT
-ENV GIT_COMMIT=$GIT_COMMIT
+# Get git commit info, then delete the .git folder
+RUN git rev-parse HEAD >> git_commit
+RUN rm -rf .git
 
 ENTRYPOINT ["./docker-entrypoint.sh"]

asr.py

@@ -1,7 +1,6 @@
 import logging
 import os
 import time
-import tomli
 
 from base_util import (
     get_asset_info,
@@ -27,14 +26,12 @@
 
 logger = logging.getLogger(__name__)
 
-
-def _get_project_meta():
-    with open("pyproject.toml", mode="rb") as pyproject:
-        return tomli.load(pyproject)["tool"]["poetry"]
-
-
-pkg_meta = _get_project_meta()
-version = os.environ["GIT_COMMIT"] if os.environ["GIT_COMMIT"] else str(pkg_meta["version"])
+# Get commit hash and use it as version in prov
+version = ""
+if os.path.exists("git_commit"):
+    with open("git_commit", "r") as f:
+        for line in f:
+            version = line.strip()
 
 
 def run(input_uri: str, output_uri: str, model=None) -> bool:

docker-compose.yml

@@ -13,8 +13,8 @@ services:
         max-size: 20m
     restart: no
     # Comment below to use the CPU instead of the GPU/CUDA
-    deploy:
-      resources:
-        reservations:
-          devices:
-          - capabilities: [gpu]
+    # deploy:
+    #   resources:
+    #     reservations:
+    #       devices:
+    #       - capabilities: [gpu]

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "whisper-asr-worker"
-version = "0.3.0"
+version = "1.0.0"
 description = "Whisper speech-to-text worker"
 authors = ["Dragos Alexandru Balan <[email protected]>"]
 license = "MIT"
@@ -14,7 +14,6 @@ validators = "^0.34.0"
 boto3 = "^1.35.10"
 fastapi = "^0.115.0"
 uvicorn = "^0.30.6"
-tomli = "^2.0.1"
 
 [tool.poetry.group.dev.dependencies]
 moto = "^5.0.13"