sn17-ch-job-sd

ComputeHorde + SN17 PoC job for running Stable Diffusion generation

Building job image

In order to schedule jobs on ComputeHorde, a docker image containing the code and model to execute must be first build and uploaded to a registry. The registry must be accessible by miners.

The ch-job-image directory contains a simple Dockerfile and a script (build-image.sh) to turn a "model config" into a ComputeHorde job image.

Prerequisites

• docker and git installed on your machine
• upload access to a docker registry

Preparation (one-time)

cd ch-job-image
git clone <image-generator-repo-address> image-generator

Building the image

./build-job-image.sh ./model-configs/sd15.yml <some.docker.registry/image-name>:latest
docker push <some.docker.registry/image-name>:latest

Executing a job

The ch-job-client directory contains a sample python project for submitting and validating jobs on ComputeHorde:

Prerequisites

• Python 3.13
• Bittensor wallet+hotkey for authentication - can be generated just for this purpose
• SS58 Address of a validator who has whitelisted the above wallet
• AWS bucket and credentials with read+write access to the bucket

Setup

(within the ch-job-client directory)

Initialize a venv and install requirements:

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Configure the client:

cp .env.example .env

Open the .env file with a text editor and fill in the configuration

Running

Run python submit_jobs.py.

This will:

• Scan each subdirectory in the batches, each containing a list of prompts
• Submit it as a ComputeHorde job using the configured image
• Pull in job results - generated images
• In the meantime, submit a "trusted" job with prompt samples to validate the results
• Cross-validate job results with the trusted job
• Report issues

Stable Diffusion License notice

This repository contains code that builds and packages a complete Docker image that is essentially a distribution of a Stable Diffusion model (SD 1.5, SD 2, SDXL).

The models are licensed under The CreativeML OpenRAIL M license available at https://huggingface.co/spaces/CompVis/stable-diffusion-license

Note that the license contains some use case restrictions. Therefore, you agree not to use the model or this distribution:

• In any way that violates any applicable national, federal, state, local or international law or regulation;
• For the purpose of exploiting, harming or attempting to exploit or harm minors in any way;
• To generate or disseminate verifiably false information and/or content with the purpose of harming others;
• To generate or disseminate personal identifiable information that can be used to harm an individual;
• To defame, disparage or otherwise harass others;
• For fully automated decision making that adversely impacts an individual’s legal rights or otherwise creates or modifies a binding, enforceable obligation;
• For any use intended to or which has the effect of discriminating against or harming individuals or groups based on online or offline social behavior or known or predicted personal or personality characteristics;
• To exploit any of the vulnerabilities of a specific group of persons based on their age, social, physical or mental characteristics, in order to materially distort the behavior of a person pertaining to that group in a manner that causes or is likely to cause that person or another person physical or psychological harm;
• For any use intended to or which has the effect of discriminating against individuals or groups based on legally protected characteristics or categories;
• To provide medical advice and medical results interpretation;
• To generate or disseminate information for the purpose to be used for administration of justice, law enforcement, immigration or asylum processes, such as predicting an individual will commit fraud/crime commitment (e.g. by text profiling, drawing causal relationships between assertions made in documents, indiscriminate and arbitrarily-targeted use).