chore(language): optimize Dockerfile

idp/package.json

@@ -1,7 +1,7 @@
 {
   "name": "voltaserve-idp",
   "version": "2.0.0",
-  "license": "MIT",
+  "license": "AGPL-3.0-only",
   "private": true,
   "scripts": {
     "start": "ts-node -r tsconfig-paths/register src/app.ts",

language/.dockerignore

@@ -1 +1,3 @@
-__pycache__
+__pycache__
+.pdm-python
+.venv

language/.gitignore

@@ -1,2 +1,3 @@
 /__pycache__
-/.idea
+/.venv
+/.pdm-python

language/Dockerfile

@@ -1,21 +1,21 @@
-FROM python:3.11-slim-bookworm
+FROM python:3.12-alpine
 
 WORKDIR /app
 
 COPY . .
 
-RUN apt-get update
+RUN apk update
+RUN apk add --no-cache build-base
 
-RUN apt-get install -y curl build-essential python3-poetry rust-all
+RUN pip3 install pipx
+ENV PATH="/root/.local/bin:$PATH"
 
-RUN poetry install --no-root
+RUN pipx install pdm --python $(which python)
 
-RUN pip3 install --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu
+RUN pdm install --prod --no-editable
+RUN .venv/bin/python3 -m ensurepip
+RUN .venv/bin/python3 -m spacy download xx_ent_wiki_sm
 
-ENV PIP_DEFAULT_TIMEOUT=900
-
-RUN poetry run spacy download xx_ent_wiki_sm
-
-ENTRYPOINT ["poetry", "run", "flask", "run", "--host=0.0.0.0", "--port=8084"]
+ENTRYPOINT ["pdm", "run", "flask", "run", "--host=0.0.0.0", "--port=8084"]
 
 EXPOSE 8084

language/README.md

@@ -2,47 +2,37 @@
 
 ## Getting Started
 
-Install [Poetry](https://python-poetry.org).
+Install [PDM](https://pdm-project.org).
+
+Install Python 3.12.
 
 Install dependencies:
 
 ```shell
-poetry install
+pdm install
 ```
 
-Spawn a shell within the project's virtual environment:
+Activate the virtual environment created by PDM:
 
 ```shell
-poetry shell
+source .venv/bin/activate
 ```
 
-Install spaCy model:
+Make sure PIP is available:
 
-```shell:
-poetry run spacy download xx_ent_wiki_sm
-poetry run spacy download zh_core_web_trf
-poetry run spacy download de_core_news_lg
-poetry run spacy download en_core_web_trf
-poetry run spacy download fr_core_news_lg
-poetry run spacy download it_core_news_lg
-poetry run spacy download ja_core_news_trf
-poetry run spacy download nl_core_news_lg
-poetry run spacy download pt_core_news_lg
-poetry run spacy download ru_core_news_lg
-poetry run spacy download es_core_news_lg
-poetry run spacy download sv_core_news_lg
+```shell
+python3 -m ensurepip
 ```
 
-On Apple Silicon or Intel Macs with supported AMD GPUs, do the following to install a hardware accelerated version of PyTorch:
+Install spaCy model:
 
-https://developer.apple.com/metal/pytorch
+```shell:
 
-```shell
-pip3 install --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu
+spacy download xx_ent_wiki_sm
 ```
 
 Run for development:
 
 ```shell
-poetry run flask run --host=0.0.0.0 --port=8084 --debug
+flask run --host=0.0.0.0 --port=8084 --debug
 ```

language/app.py

@@ -1,14 +1,6 @@
 from flask import Flask, request, jsonify
 import string
 import spacy
-import torch
-
-if torch.backends.mps.is_available():
-    mps_device = torch.device("mps")
-    x = torch.ones(1, device=mps_device)
-    print(f"🔥 MPS device is available: {x}")
-else:
-    print ("MPS device not found.")
 
 app = Flask(__name__)
 nlp = None
@@ -75,9 +67,12 @@ def ner_entities():
             result[key]["frequency"] += 1
         else:
             result[key] = {"text": entity["text"], "frequency": 1}
-    
+
     # Convert the dictionary back to a list of entities with the "frequency" field
-    result = [{"text": value["text"], "frequency": value["frequency"]} for value in result.values()]
+    result = [
+        {"text": value["text"], "frequency": value["frequency"]}
+        for value in result.values()
+    ]
 
     # Sort by descending order of frequency
     result.sort(key=lambda x: x["frequency"], reverse=True)