Merge pull request #9 from digitalbuiltenvironment/development

Minor Fixes & Improvements

Dockerfile

@@ -1,31 +1,52 @@
-# Use the official Python Image
-FROM python:3.11.4
+ARG CUDA_IMAGE="12.1.1-devel-ubuntu22.04"
+# Build with Cuda 12.1.1 and Ubuntu 22.04
+FROM nvidia/cuda:${CUDA_IMAGE}
 
 # Set up a new user named "user" with user ID 1000
 RUN useradd -m -u 1000 user
+
 # Install the dependencies
-RUN apt-get update && apt-get -y install cmake protobuf-compiler libopenblas-dev liblapack-dev liblapacke-dev libeigen3-dev libboost-all-dev
+RUN apt-get update && apt-get upgrade -y \
+    && apt-get install -y git build-essential \
+    python3.11 gcc wget \
+    ocl-icd-opencl-dev opencl-headers clinfo \
+    cmake protobuf-compiler pkg-config \
+    libclblast-dev libopenblas-dev \
+    liblapack-dev liblapacke-dev libeigen3-dev libboost-all-dev \
+    && mkdir -p /etc/OpenCL/vendors && echo "libnvidia-opencl.so.1" > /etc/OpenCL/vendors/nvidia.icd
+
+# Install pip for python 3.11
+RUN wget https://bootstrap.pypa.io/get-pip.py && \
+    python3.11 get-pip.py && \
+    rm get-pip.py
 
 # Switch to the user 'user'
 USER user
 
+# Setting build related env vars
+ENV CUDA_DOCKER_ARCH=all
+ENV LLAMA_CUBLAS=1
+
 # Set home to the user's home directory and Poetry's environment variables
 ENV HOME=/home/user \
 	PATH=/home/user/.local/bin:$PATH \
+    PYTHONUNBUFFERED=1 \
     POETRY_NO_INTERACTION=1 \
     POETRY_VIRTUALENVS_IN_PROJECT=1 \
     POETRY_VIRTUALENVS_CREATE=1 \
     POETRY_CACHE_DIR=/tmp/poetry_cache \
-    CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS"
+    # Build llama-cpp-python with default cuda support
+    CMAKE_ARGS="-DLLAMA_CUBLAS=on"
+    # CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS"
 
 # Set the working directory to /app
 WORKDIR $HOME/app
 
 # Update pip and wheel
-RUN pip install --upgrade pip setuptools wheel
+RUN python3.11 -m pip install --upgrade pip setuptools wheel
 
 # Install poetry
-RUN pip install poetry
+RUN python3.11 -m pip install poetry
 
 # Copy the poetry files
 COPY --chown=user ./backend/pyproject.toml ./backend/poetry.lock $HOME/app/
@@ -34,8 +55,7 @@ COPY --chown=user ./backend/pyproject.toml ./backend/poetry.lock $HOME/app/
 COPY --chown=user ./backend $HOME/app
 
 # Install the dependencies
-RUN poetry lock --no-update
-RUN poetry install --without dev && \
+RUN poetry install --without dev,torch-cpu && \
     rm -rf /tmp/poetry_cache
 
 # Change to the package directory

backend/README.md

@@ -5,34 +5,56 @@ The backend is built using Python & [FastAPI](https://fastapi.tiangolo.com/) boo
 ## Requirements
 
 1. Python >= 3.11
-2. Poetry (To manage dependencies)
+2. Miniconda (To manage Python versions)
+   - [Windows](https://repo.anaconda.com/miniconda/Miniconda3-latest-Windows-x86_64.exe)
+   - [Linux](https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh)
+   - [MacOS](https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-x86_64.pkg)
+   - ```conda create -n SmartRetrieval python=3.11```
+3. Pipx (To manage Python packages)
+   - ```pip install pipx``` (If you already have pipx installed, you can skip this step)
+4. Cuda > 12.1 (if you have a Nvidia GPU)
+    - [Windows](https://developer.nvidia.com/cuda-downloads)
+    - [Linux](https://developer.nvidia.com/cuda-downloads)
+    - [MacOS](https://developer.nvidia.com/cuda-downloads)
+5. Poetry (To manage dependencies)
    - ```pipx install poetry```
 
 ## Getting Started
 
-First, setup the `pyproject.toml` file to install the correct version of PyTorch (CPU or GPU):
+First, ensure if you want to use the cuda version of pytorch, you have the correct version `cuda > 12.1` of cuda installed. You can check this by running `nvcc --version or nvidia-smi` in your terminal. If you do not have cuda installed, you can install it from [here](https://developer.nvidia.com/cuda-downloads).
 
-Comment/Uncomment the following block depending on your system. Use CPU only if you do not have a supported Cuda Device.
+Ensure you have followed the steps in the requirements section above.
+
+Then activate the conda environment:
 
 ```bash
-# For CPU version: Windows and Linux and MacOS (arm64)
-torch = [
-    { url = "https://download.pytorch.org/whl/cpu/torch-2.1.1%2Bcpu-cp311-cp311-win_amd64.whl", markers = "sys_platform == 'win32'" },
-    { url = "https://download.pytorch.org/whl/cpu/torch-2.1.1%2Bcpu-cp311-cp311-linux_x86_64.whl", markers = "sys_platform == 'linux'" },
-    { url = "https://download.pytorch.org/whl/cpu/torch-2.1.1-cp311-none-macosx_11_0_arm64.whl", markers = "sys_platform == 'darwin'" },
-]
-## For GPU version: Windows and Linux and MacOS (arm64)
-# torch = [
-#     { url = "https://download.pytorch.org/whl/cu121/torch-2.1.1%2Bcu121-cp311-cp311-win_amd64.whl", markers = "sys_platform == 'win32'" },
-#     { url = "https://download.pytorch.org/whl/cu121/torch-2.1.1%2Bcu121-cp311-cp311-linux_x86_64.whl", markers = "sys_platform == 'linux'" },
-#     { url = "https://download.pytorch.org/whl/cu121/torch-2.1.1-cp311-none-macosx_11_0_arm64.whl", markers = "sys_platform == 'darwin'" },
-# ]
+conda activate SmartRetrieval
 ```
 
 Second, setup the environment:
 
 ```bash
-poetry install
+# Only run one of the following commands:
+-----------------------------------------------
+# Install dependencies and torch (cpu version)
+# Windows: Set env for llama-cpp-python with openblas support on cpu
+$env:CMAKE_ARGS = "-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS"
+# Linux: Set env for llama-cpp-python with openblas support on cpu
+CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS"
+# Then:
+poetry install --without torch-cuda
+-----------------------------------------------
+# Install dependencies and torch (cuda version)
+# Windows: Set env for llama-cpp-python with cuda support on gpu
+$env:CMAKE_ARGS = "-DLLAMA_CUBLAS=on"
+# Linux: Set env for llama-cpp-python with cuda support on gpu
+CMAKE_ARGS="-DLLAMA_CUBLAS=on"
+# Then:
+poetry install --without torch-cpu
+```
+
+```bash
+# Enter poetry shell
 poetry shell
 ```
 

backend/pyproject.toml

@@ -15,21 +15,24 @@ pypdf = "^3.17.0"
 python-dotenv = "^1.0.0"
 llama-cpp-python = "^0.2.18"
 transformers = "^4.35.2"
-# For CPU version: Windows and Linux and MacOS (arm64)
-torch = [
-    { url = "https://download.pytorch.org/whl/cpu/torch-2.1.1%2Bcpu-cp311-cp311-win_amd64.whl", markers = "sys_platform == 'win32'" },
-    { url = "https://download.pytorch.org/whl/cpu/torch-2.1.1%2Bcpu-cp311-cp311-linux_x86_64.whl", markers = "sys_platform == 'linux'" },
-]
-## For GPU version: Windows and Linux and MacOS (arm64)
-# torch = [
-#     { url = "https://download.pytorch.org/whl/cu121/torch-2.1.1%2Bcu121-cp311-cp311-win_amd64.whl", markers = "sys_platform == 'win32'" },
-#     { url = "https://download.pytorch.org/whl/cu121/torch-2.1.1%2Bcu121-cp311-cp311-linux_x86_64.whl", markers = "sys_platform == 'linux'" },
-# ]
 docx2txt = "^0.8"
 
 # Dev Dependencies here
 [tool.poetry.group.dev.dependencies]
 
+# For CPU torch version: Windows and Linux
+[tool.poetry.group.torch-cpu.dependencies]
+torch = [
+    { url = "https://download.pytorch.org/whl/cpu/torch-2.1.1%2Bcpu-cp311-cp311-win_amd64.whl", markers = "sys_platform == 'win32'" },
+    { url = "https://download.pytorch.org/whl/cpu/torch-2.1.1%2Bcpu-cp311-cp311-linux_x86_64.whl", markers = "sys_platform == 'linux'" },
+]
+
+## For Cuda torch version: Windows and Linux
+[tool.poetry.group.torch-cuda.dependencies]
+torch = [
+    { url = "https://download.pytorch.org/whl/cu121/torch-2.1.1%2Bcu121-cp311-cp311-win_amd64.whl", markers = "sys_platform == 'win32'" },
+    { url = "https://download.pytorch.org/whl/cu121/torch-2.1.1%2Bcu121-cp311-cp311-linux_x86_64.whl", markers = "sys_platform == 'linux'" },
+]
 
 [build-system]
 requires = ["poetry-core"]

frontend/app/components/ui/search/useSearch.tsx

@@ -32,7 +32,9 @@ const useSearch = (): UseSearchResult => {
                 setIsLoading(false);
                 return;
             }
-            const response = await fetch(`${search_api}?query=${query}`);
+            const response = await fetch(`${search_api}?query=${query}`, {
+                signal: AbortSignal.timeout(10000), // Abort the request if it takes longer than 10 seconds
+                });
             const data = await response.json();
             setSearchResults(data);
         } catch (error) {

frontend/app/status/page.tsx

@@ -12,7 +12,9 @@ const StatusPage = () => {
   const { data, error, isValidating, mutate } = useSWR(healthcheck_api, async (url) => {
     try {
       // Fetch the data
-      const response = await fetch(url);
+      const response = await fetch(url, {
+        signal: AbortSignal.timeout(5000), // Abort the request if it takes longer than 5 seconds
+        });
       if (!response.ok) {
         throw new Error(response.statusText || 'Unknown Error');
       }