chore: Update Dockerfile to improve build process and upgrade pip
Browse files- Dockerfile +25 -7
- README.md +2 -0
Dockerfile
CHANGED
|
@@ -1,21 +1,39 @@
|
|
| 1 |
# read the doc: https://huggingface.co/docs/hub/spaces-sdks-docker
|
| 2 |
# you will also find guides on how best to write your Dockerfile
|
| 3 |
|
|
|
|
| 4 |
FROM tiangolo/uvicorn-gunicorn:python3.10-slim
|
| 5 |
|
| 6 |
-
# Set
|
| 7 |
-
|
| 8 |
|
| 9 |
-
#
|
| 10 |
-
|
|
|
|
| 11 |
|
| 12 |
-
#
|
| 13 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
|
| 15 |
# Upgrade pip
|
| 16 |
RUN pip install --no-cache-dir --upgrade pip
|
| 17 |
|
| 18 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
RUN pip install --no-cache-dir --upgrade -r requirements.txt
|
| 20 |
|
| 21 |
# Expose the port the app runs on
|
|
|
|
| 1 |
# read the doc: https://huggingface.co/docs/hub/spaces-sdks-docker
|
| 2 |
# you will also find guides on how best to write your Dockerfile
|
| 3 |
|
| 4 |
+
# Use the official Python 3.10 slim image as the base image
|
| 5 |
FROM tiangolo/uvicorn-gunicorn:python3.10-slim
|
| 6 |
|
| 7 |
+
# Set up a new user named "user" with user ID 1000
|
| 8 |
+
RUN useradd -m -u 1000 user
|
| 9 |
|
| 10 |
+
# Create necessary directories and set permissions
|
| 11 |
+
RUN mkdir -p /home/user/.cache && chown -R user:user /home/user/.cache
|
| 12 |
+
RUN mkdir -p /home/user/.local && chown -R user:user /home/user/.local
|
| 13 |
|
| 14 |
+
# Switch to the "user" user
|
| 15 |
+
USER user
|
| 16 |
+
|
| 17 |
+
# Set home to the user's home directory
|
| 18 |
+
ENV HOME=/home/user \
|
| 19 |
+
PATH=/home/user/.local/bin:$PATH
|
| 20 |
+
|
| 21 |
+
# Set the working directory to the user's home directory
|
| 22 |
+
WORKDIR $HOME/app
|
| 23 |
|
| 24 |
# Upgrade pip
|
| 25 |
RUN pip install --no-cache-dir --upgrade pip
|
| 26 |
|
| 27 |
+
# Copy the current directory contents into the container at $HOME/app
|
| 28 |
+
COPY --chown=user . $HOME/app
|
| 29 |
+
|
| 30 |
+
# Set the working directory
|
| 31 |
+
WORKDIR /app
|
| 32 |
+
|
| 33 |
+
# List the contents of the working directory to debug
|
| 34 |
+
RUN ls -la
|
| 35 |
+
|
| 36 |
+
# Install the dependencies
|
| 37 |
RUN pip install --no-cache-dir --upgrade -r requirements.txt
|
| 38 |
|
| 39 |
# Expose the port the app runs on
|
README.md
CHANGED
|
@@ -8,3 +8,5 @@ pinned: false
|
|
| 8 |
---
|
| 9 |
|
| 10 |
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
|
|
|
|
|
|
|
|
| 8 |
---
|
| 9 |
|
| 10 |
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
| 11 |
+
|
| 12 |
+
SOTA open VLM is [InternVL-1.5](https://huggingface.co/spaces/opencompass/open_vlm_leaderboard), which is 22B, for practical deployment I choose moondream which is a model can answer real-world questions about images (378x378). It's tiny by today's models, with only 1.6B parameters. That enables it to run on a variety of devices, including mobile phones and edge devices.
|