Upload 3 files

Dockerfile

@@ -1,30 +1,32 @@
 # Use the official vLLM Docker image as the base image
 FROM vllm/vllm-openai:latest
 
-# Use the root user to ensure write permissions (if needed)
+# Ensure we run as root (the default) so we can set permissions
 USER root
 
-# Set the working directory
-WORKDIR /app
-
-# Install system dependencies if needed (e.g., wget)
+# Install system dependencies
 RUN apt-get update && apt-get install -y wget && rm -rf /var/lib/apt/lists/*
 
-# Create and set permissions for cache directories (for Hugging Face, matplotlib, etc.)
+# Set working directory
+WORKDIR /app
+
+# Create and set permissions for cache directories
 RUN mkdir -p /app/.cache && chmod -R 777 /app/.cache \
-    && mkdir -p /app/.config/matplotlib && chmod -R 777 /app/.config/matplotlib
+    && mkdir -p /app/.config/matplotlib && chmod -R 777 /app/.config/matplotlib \
+    && mkdir -p /.EasyOCR && chmod -R 777 /.EasyOCR
 
-# Set environment variables for cache directories and user agent
+# Set environment variables for cache directories
 ENV HF_HOME=/app/.cache
 ENV XDG_CACHE_HOME=/app/.cache
 ENV MPLCONFIGDIR=/app/.config/matplotlib
 ENV USER_AGENT="my-gradio-app"
+ENV EASYOCR_CACHE_DIR=/app/.EasyOCR
 
-# Copy the requirements file and install additional Python dependencies (e.g., gradio)
+# Copy the requirements file and install dependencies
 COPY requirements.txt .
 RUN pip install --no-cache-dir -r requirements.txt
 
-# Copy your application code (including app.py and any other needed files)
+# Copy the rest of the application code
 COPY . .
 
 # Expose the port for Gradio (Spaces expects the app on port 7860)
@@ -33,8 +35,5 @@ EXPOSE 7860
 # Override the base image's entrypoint so our CMD is executed directly.
 ENTRYPOINT []
 
-# Create a writable log file (alternatively, you could redirect logs to /tmp)
-RUN touch /app/vllm.log && chmod 666 /app/vllm.log
-
 # Set the CMD to launch the vLLM server (for your new model) in the background and then start the Gradio app.
 CMD ["bash", "-c", "vllm.entrypoints.openai.api_server --model unsloth/llama-3-8b-Instruct-bnb-4bit --enable-auto-tool-choice --tool-call-parser llama3_json --chat-template examples/tool_chat_template_llama3.1_json.jinja --quantization bitsandbytes --load-format bitsandbytes --dtype half --max-model-len 8192 > /app/vllm.log 2>&1 & python3 app.py"]

app.py

@@ -2,7 +2,19 @@ from io import StringIO
 import sys
 
 import os
-from huggingface_hub import login
+# Set EasyOCR cache directory to a writable location
+os.environ["EASYOCR_CACHE_DIR"] = "/app/.EASYOCR"
+import easyocr
+# Monkey-patch the easyocr.Reader to force the model_storage directory parameter
+_original_init = easyocr.Reader.__init__
+def new_init(self, *args, **kwargs):
+    if args and "lang_list" in kwargs:
+        del kwargs["lang_list"]
+    kwargs.setdefault("model_storage_directory", "/app/.EasyOCR")
+    _original_init(self, *args, **kwargs)
+easyocr.Reader.__init__ = new_init
+
+#from huggingface_hub import login
 import gradio as gr
 import json
 import csv
@@ -56,11 +68,11 @@ logger = logging.getLogger(__name__)
 logging.disable(logging.WARNING)
 
 
-HF_TOKEN = os.getenv("HF_TOKEN")  # Read from environment variable
-if HF_TOKEN:
-    login(token=HF_TOKEN)  # Log in to Hugging Face Hub
-else:
-    print("Warning: HF_TOKEN not found in environment variables.")
+# HF_TOKEN = os.getenv("HF_TOKEN")  # Read from environment variable
+# if HF_TOKEN:
+#     login(token=HF_TOKEN)  # Log in to Hugging Face Hub
+# else:
+#     print("Warning: HF_TOKEN not found in environment variables.")
 
 # GROQ_API_KEY = os.getenv("GROQ_API_KEY")  # Read from environment variable
 # if not GROQ_API_KEY:
@@ -551,6 +563,7 @@ model = ChatOpenAI(
     base_url="http://localhost:8000/v1", # Use the VLLM instance URL
 )
 
+
 # model = ChatGroq(
 #     model_name="deepseek-r1-distill-llama-70b",
 #     temperature=TEMPERATURE,
@@ -991,11 +1004,13 @@ with gr.Blocks(theme="ocean") as AI_Tutor:
         type="messages",
         chatbot=gr.Chatbot(
             label="Chat Window",
-            height=500
+            height=500,
+            type="messages"
         ),
         textbox=gr.MultimodalTextbox(
+            interactive=True,
             file_count="multiple",
-            file_types=None,
+            file_types=[".pdf",".ppt",".pptx",".doc",".docx",".md","image"],
             sources=["upload"],
             label="Type your query here:",
             placeholder="Enter your question...",