Update Dockerfile

Dockerfile

@@ -1,8 +1,7 @@
-# 1. Use Python 3.11 (Required by the new GECToR library)
+# 1. Use Python 3.11 (Required for GECToR)
 FROM python:3.11-slim
 
 # 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 \
@@ -25,26 +24,28 @@ ENV TRANSFORMERS_CACHE=/app/cache
 ENV HF_HOME=/app/cache
 RUN mkdir -p /app/cache && chmod 777 /app/cache
 
-# 5. Install Dependencies (From your cleaned requirements.txt)
+# 5. Install Dependencies
 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
 RUN pip install --no-cache-dir /app/gector_lib
 
-# 7. SETUP DATA & MODEL
+# 7. SETUP DATA (Force Fresh Download)
 RUN mkdir -p /app/data
-# 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
+
+# Force cache break using date (Ensures we actually download the files)
+RUN echo "Downloading Model Weights..." && \
+    wget -O /app/data/gector_model.th https://huggingface.co/gotutiyan/gector-roberta-base-5k/resolve/main/pytorch_model.bin
+
+# Download Vocab file directly (Fixes the 'No such file' error)
+RUN echo "Downloading Vocab File..." && \
+    wget -O /app/data/verb-form-vocab.txt https://github.com/grammarly/gector/raw/master/data/vocabulary/verb-form-vocab.txt
 
 # 8. Copy App Code
 COPY . .
 
-# 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
+# 9. Run App
 CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]