From 7f114098fd815adcbc55797aa2ff5703be969d48 Mon Sep 17 00:00:00 2001
From: Quentin McGaw <quentin.mcgaw@gmail.com>
Date: Mon, 25 Mar 2024 15:24:41 +0000
Subject: [PATCH] chore(docker): merge local and external dockerfiles -
 llama-cpp dependencies install does not occupy space in final image -
 llama-cpp dependencies install takes about 10 seconds, worth the trade off -
 Update docker.yml workflow to use the standard `Dockerfile`

---
 .github/workflows/docker.yml      |  2 -
 Dockerfile.external => Dockerfile |  8 ++++
 Dockerfile.local                  | 69 -------------------------------
 docker-compose.yaml               |  5 ++-
 4 files changed, 12 insertions(+), 72 deletions(-)
 rename Dockerfile.external => Dockerfile (90%)
 delete mode 100644 Dockerfile.local

diff --git a/.github/workflows/docker.yml b/.github/workflows/docker.yml
index 259cf5d..471106f 100644
--- a/.github/workflows/docker.yml
+++ b/.github/workflows/docker.yml
@@ -38,8 +38,6 @@ jobs:
       - name: Build and push Docker image
         uses: docker/build-push-action@v5
         with:
-          context: .
-          file: Dockerfile.external
           push: true
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
diff --git a/Dockerfile.external b/Dockerfile
similarity index 90%
rename from Dockerfile.external
rename to Dockerfile
index 8b01e9c..4db8e5f 100644
--- a/Dockerfile.external
+++ b/Dockerfile
@@ -11,8 +11,16 @@ ENV POETRY_VIRTUALENVS_IN_PROJECT=true
 
 FROM base as dependencies
 WORKDIR /home/worker/app
+# Dependencies to build llama-cpp
+RUN apt update && apt install -y \
+  libopenblas-dev\
+  ninja-build\
+  build-essential\
+  pkg-config\
+  wget
 COPY pyproject.toml poetry.lock ./
 
+# Extras possible are: llms-llama-cpp embeddings-huggingface
 ARG POETRY_EXTRAS="ui vector-stores-qdrant llms-ollama embeddings-ollama"
 RUN poetry install --no-root --extras "${POETRY_EXTRAS}"
 
diff --git a/Dockerfile.local b/Dockerfile.local
deleted file mode 100644
index d4ba8be..0000000
--- a/Dockerfile.local
+++ /dev/null
@@ -1,69 +0,0 @@
-### IMPORTANT, THIS IMAGE CAN ONLY BE RUN IN LINUX DOCKER
-### You will run into a segfault in mac
-FROM python:3.11.6-slim-bookworm as base
-
-# Install poetry
-RUN pip install pipx
-RUN pipx install poetry
-ENV PATH="/root/.local/bin:$PATH"
-ENV PATH=".venv/bin/:$PATH"
-# https://python-poetry.org/docs/configuration/#virtualenvsin-project
-ENV POETRY_VIRTUALENVS_IN_PROJECT=true
-
-FROM base as dependencies
-WORKDIR /home/worker/app
-# Dependencies to build llama-cpp
-RUN apt update && apt install -y \
-  libopenblas-dev\
-  ninja-build\
-  build-essential\
-  pkg-config\
-  wget
-COPY pyproject.toml poetry.lock ./
-ARG POETRY_EXTRAS="ui embeddings-huggingface llms-llama-cpp vector-stores-qdrant"
-RUN poetry install --no-root --extras "${POETRY_EXTRAS}"
-
-FROM base as app
-
-ENV \
-  PYTHONUNBUFFERED=1 \
-  PORT=8080 \
-  PYTHONPATH="$PYTHONPATH:/home/worker/app/private_gpt/" \
-  APP_ENV=prod \
-  PGPT_PROFILES=docker \
-  PGPT_MODE=mock \
-  PGPT_EMBEDDING_MODE=sagemaker \
-  PGPT_HF_REPO_ID=TheBloke/Mistral-7B-Instruct-v0.1-GGUF \
-  PGPT_HF_MODEL_FILE=mistral-7b-instruct-v0.1.Q4_K_M.gguf \
-  PGPT_EMBEDDING_HF_MODEL_NAME=BAAI/bge-small-en-v1.5 \
-  PGPT_SAGEMAKER_LLM_ENDPOINT_NAME= \
-  PGPT_SAGEMAKER_EMBEDDING_ENDPOINT_NAME= \
-  PGPT_OLLAMA_LLM_MODEL=mistral \
-  PGPT_OLLAMA_EMBEDDING_MODEL=nomic-embed-text \
-  PGPT_OLLAMA_API_BASE=http://ollama:11434 \
-  PGPT_OLLAMA_TFS_Z=1.0 \
-  PGPT_OLLAMA_TOP_K=40 \
-  PGPT_OLLAMA_TOP_P=0.9 \
-  PGPT_OLLAMA_REPEAT_LAST_N=64 \
-  PGPT_OLLAMA_REPEAT_PENALTY=1.2 \
-  PGPT_OLLAMA_REQUEST_TIMEOUT=600.0
-EXPOSE 8080
-
-# Prepare a non-root user
-ARG UID=100
-ARG GID=65534
-RUN adduser --group --gid ${GID} worker
-RUN adduser --system --ingroup worker --uid ${UID} --home /home/worker worker
-WORKDIR /home/worker/app
-
-RUN mkdir local_data && chown worker local_data
-RUN mkdir models && chown worker models
-COPY --chown=worker --from=dependencies /home/worker/app/.venv/ .venv
-COPY --chown=worker private_gpt/ private_gpt
-COPY --chown=worker fern/ fern
-COPY --chown=worker *.yaml ./
-COPY --chown=worker scripts/ scripts
-COPY --chown=worker Makefile .
-
-USER worker
-ENTRYPOINT python -m private_gpt
\ No newline at end of file
diff --git a/docker-compose.yaml b/docker-compose.yaml
index f12e1da..e208c86 100644
--- a/docker-compose.yaml
+++ b/docker-compose.yaml
@@ -1,7 +1,10 @@
 services:
   private-gpt:
     build:
-      dockerfile: Dockerfile.external
+      context: .
+      # context: https://github.com/zylon-ai/private-gpt.git#main
+      args:
+        - POETRY_EXTRAS=ui vector-stores-qdrant llms-ollama embeddings-ollama
     volumes:
       - ./local_data/:/home/worker/app/local_data
     ports: