Update Dockerfile

Dockerfile

@@ -1,8 +1,8 @@
-# 1. UPGRADE to Python 3.11 (Required by GECToR)
+# 1. Use Python 3.11 (Required by the new GECToR library)
 FROM python:3.11-slim
 
-# 2. Install System Tools
-# We switch back to openjdk-17 because it is the standard for Python 3.11 images
+# 2. Install System Tools & Java 21
+# We use openjdk-21 because this base image is Debian 13 (Trixie)
 RUN apt-get update && \
     apt-get install -y \
     git \
@@ -13,7 +13,7 @@ RUN apt-get update && \
     gcc \
     && apt-get clean
 
-# Set Java Home (Updated for Java 21)
+# Set Java Home
 ENV JAVA_HOME=/usr/lib/jvm/java-21-openjdk-amd64
 
 # 3. Set working directory
@@ -25,27 +25,26 @@ ENV TRANSFORMERS_CACHE=/app/cache
 ENV HF_HOME=/app/cache
 RUN mkdir -p /app/cache && chmod 777 /app/cache
 
-# 5. Install Dependencies
+# 5. Install Dependencies (From your cleaned requirements.txt)
 COPY requirements.txt .
 RUN pip install --no-cache-dir --upgrade pip && \
     pip install --no-cache-dir -r requirements.txt
 
 # 6. INSTALL GECToR LIBRARY
+# We clone the library and let pip resolve dependencies automatically
 RUN git clone https://github.com/gotutiyan/gector.git /app/gector_lib
-# This step failed before; now it will pass because we have Python 3.11
 RUN pip install --no-cache-dir /app/gector_lib
 
 # 7. SETUP DATA & MODEL
 RUN mkdir -p /app/data
-
-# Download Model Weights
+# Download the heavy model weights
 RUN wget -O /app/data/gector_model.th https://huggingface.co/gotutiyan/gector-roberta-base-5k/resolve/main/pytorch_model.bin
 
 # 8. Copy App Code
 COPY . .
 
-# 9. Move vocab file if needed (ensure it exists)
-RUN mv verb-form-vocab.txt /app/data/verb-form-vocab.txt || echo "File not found root, skipping move"
+# 9. Move vocab file (if uploaded to root)
+RUN mv verb-form-vocab.txt /app/data/verb-form-vocab.txt || echo "File not found in root, skipping move"
 
 # 10. Run App
 CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]