TensorBoard S3 Docker Image

This Docker image provides a convenient way to run TensorBoard to visualize data stored in Amzaon S3/Compatible S3 Services. It comes pre-configured with support for S3 data storage and provides an easy way to specify the necessary environment variables for connecting to your S3 bucket.

Environment Variables

Before using this Docker image, make sure you have the following variables:

Environment Variable	Description	Required	Default Value	Value Example
S3_ENDPOINT	The endpoint URL for custom S3 endpoint	Yes	N/A	https://your.s3.provider.url (or you can use http://)
S3_BUCKET_NAME	The name of the Amazon S3 bucket	Yes	N/A	s3-bucket
LOGDIR	The directory path inside the S3 bucket	Yes	N/A	logdir_folder
AWS_ACCESS_KEY_ID	The AWS access key ID for S3 authentication	Yes	N/A	********
AWS_SECRET_ACCESS_KEY	The AWS secret access key for S3 authentication	Yes	N/A	*****************
PORT	The port number on which TensorBoard will listen	No	6006	6006
S3_VERIFY_SSL	Enable SSL verification for S3 endpoint	No	0	0
S3_USE_HTTPS	Use HTTPS for S3 endpoint	No	0	1 (If you're using https in the endpoint, this should be 1)

Usage

To use this Docker image, follow the steps below:

1. Pull the Docker image from Docker Hub:

docker pull kaenova/tensorboard-s3

2. Pull the Docker image from Docker Hub:

docker run -p 6006:6006 \
-e S3_USE_HTTPS=<your_s3_with_http_or_https> \
-e S3_BUCKET_NAME=<your_s3_bucket_name> \
-e LOGDIR=<your_logdir_path> \
-e AWS_ACCESS_KEY_ID=<your_aws_access_key_id> \
-e AWS_SECRET_ACCESS_KEY=<your_aws_secret_access_key> \
-e PORT=<optional_port_number> \
-e S3_ENDPOINT=<optional_s3_endpoint> \
-e S3_VERIFY_SSL=<optional_s3_verify_ssl> \
kaenova/tensorboard-s3

3. Access TensorBoard
   Once the container is running, you can access TensorBoard by navigating to http://localhost:6006 in your web browser, or to the custom port number you specified in the PORT environment variable.

Docker Compose Example

Here is an example of local deployment using docker compose S3 Tensorboard with MinIO for the S3 service.

version: "3.8"

services:
  minio:
    image: quay.io/minio/minio:latest
    command: ["server", "/data", "--console-address", ":9001"]
    ports:
      - "9000:9000"
      - "9001:9001"
    environment:
      - "MINIO_ROOT_USER=ROOTNAME"
      - "MINIO_ROOT_PASSWORD=CHANGEME123"

  tensorboard:
    image: kaenova/tensorboard-s3
    ports:
      - "6006:6006"
    environment:
      - "S3_BUCKET_NAME=s3-test"
      - "S3_ENDPOINT=http://minio:9000"
      - "LOGDIR=tensorboard"
      - "AWS_ACCESS_KEY_ID=IUbvAf6LS4zYsvZLaNLR"
      - "AWS_SECRET_ACCESS_KEY=quZScW8nZaATzVbwkKLk7DmlQvwqeHmHu3Qm5N3F"

---

That's it! You can now use TensorBoard to visualize your TensorFlow logs stored in S3 Service using this Docker image with the specified environment variables.

This markdown was created with the help of ChatGPT