feat: getting rid of broken & unused things; added "test_input.json"; updated README (#48)

* feat: getting rid of broken & unused things; added "test_input.json"

* docs: improve README

Author: TimPietrusky

.runpod/hub.json

@@ -1,11 +1,22 @@
 {
-  "title": "Automatic1111 Stable Diffusion",
-  "description": "This worker uses the Stable Diffusion model for AI image generation tasks. Built upon the Stable Diffusion WebUI, it provides an efficient interface for generating images with Stable Diffusion AI models.",
+  "title": "Automatic1111 Stable Diffusion web UI",
+  "description": "Built upon the Stable Diffusion WebUI, this worker provides an API to generate images with Stable Diffusion models.",
   "type": "serverless",
   "category": "image",
   "iconUrl": "https://dummyimage.com/100x100/ff5733/fff&text=SD",
   "config": {
     "runsOn": "GPU",
-    "containerDiskInGb": 20
+    "containerDiskInGb": 20,
+    "gpuTypeId": "NVIDIA GeForce RTX 4090",
+    "gpuCount": 1,
+    "allowedCudaVersions": [
+      "12.7",
+      "12.6",
+      "12.5",
+      "12.4",
+      "12.3",
+      "12.2",
+      "12.1"
+    ]
   }
-}
+}

Dockerfile

@@ -22,40 +22,29 @@ ENV DEBIAN_FRONTEND=noninteractive \
 
 SHELL ["/bin/bash", "-o", "pipefail", "-c"]
 
-RUN export COMMANDLINE_ARGS="--skip-torch-cuda-test --precision full --no-half"
-RUN export TORCH_COMMAND='pip install ---no-cache-dir torch==2.1.2+cu118 torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118'
-
 RUN apt-get update && \
     apt install -y \
     fonts-dejavu-core rsync git jq moreutils aria2 wget libgoogle-perftools-dev libtcmalloc-minimal4 procps libgl1 libglib2.0-0 && \
     apt-get autoremove -y && rm -rf /var/lib/apt/lists/* && apt-get clean -y
 
-RUN --mount=type=cache,target=/cache --mount=type=cache,target=/root/.cache/pip \
-    pip install --upgrade pip && \
-    ${TORCH_COMMAND} && \
-    pip install --no-cache-dir xformers==0.0.23.post1 --index-url https://download.pytorch.org/whl/cu118
-
 RUN --mount=type=cache,target=/root/.cache/pip \
     git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui.git && \
     cd stable-diffusion-webui && \
     git reset --hard ${A1111_RELEASE} && \
+    pip install xformers && \
+    pip install -r requirements_versions.txt && \
     python -c "from launch import prepare_environment; prepare_environment()" --skip-torch-cuda-test
 
 COPY --from=download /model.safetensors /model.safetensors
 
-# Copy requirements.txt and install dependencies
+# install dependencies
 COPY requirements.txt .
 RUN --mount=type=cache,target=/root/.cache/pip \
     pip install --no-cache-dir -r requirements.txt
 
-# Copy test input file
 COPY test_input.json .
 
 ADD src .
 
-COPY builder/cache.py /stable-diffusion-webui/cache.py
-RUN cd /stable-diffusion-webui && python cache.py --use-cpu=all --ckpt /model.safetensors
-
-# Set permissions and specify the command to run
 RUN chmod +x /start.sh
 CMD /start.sh

LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2023 runpod-workers
+Copyright (c) 2025 RunPod
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

README.md

@@ -1,33 +1,40 @@
 <div align="center">
 
-<h1>Automatic1111 | Worker</h1>
+<h1>Automatic1111 Stable Diffusion web UI</h1>
 
 [![RunPod](https://api.runpod.io/badge/runpod-workers/worker-a1111)](https://www.runpod.io/console/hub/runpod-workers/worker-a1111)
-[![CI | Test Worker](https://github.com/runpod-workers/worker-template/actions/workflows/CI-test_worker.yml/badge.svg)](https://github.com/runpod-workers/worker-template/actions/workflows/CI-test_worker.yml)
-&nbsp;
-[![Docker Image](https://github.com/runpod-workers/worker-template/actions/workflows/CD-docker_dev.yml/badge.svg)](https://github.com/runpod-workers/worker-template/actions/workflows/CD-docker_dev.yml)
 
-This worker is a RunPod worker that uses the Stable Diffusion model for AI tasks. The worker is built upon the Stable Diffusion WebUI, which is a user interface for Stable Diffusion AI models.
-</div>
-
-## Model
+Built upon the Stable Diffusion WebUI, this worker provides an API to generate images with Stable Diffusion models.
 
-The worker uses the Stable Diffusion model, which has been optimized for RunPod. This model is stored as a SafeTensors file, which is a format that facilitates efficient loading and execution of AI models. You may download the model file from the following link: here.
+</div>
 
-## Building the Worker
+## Features
 
-The worker is built using a Dockerfile. The Dockerfile specifies the base image, environment variables, system package dependencies, Python dependencies, and the steps to install and setup the Stable Diffusion WebUI. It also downloads the model and sets up the API server using supervisor.
+- Runs the [Automatic1111 Stable Diffusion WebUI](https://github.com/AUTOMATIC1111/stable-diffusion-webui) and exposes its txt2img API endpoint.
+- Comes pre-packaged with the [**Deliberate v6**](https://huggingface.co/XpucT/Deliberate) model (`Deliberate_v6.safetensors`).
 
-The Python dependencies are specified in requirements.txt. The primary dependency is runpod==0.9.4.
+## Usage
 
-## Running the Worker
+The `input` object within your job request is forwarded directly to the `/sdapi/v1/txt2img` endpoint of the internal Automatic1111 API.
 
-The worker can be run using the start.sh script. This script starts the init system and runs the serverless handler script.
+### Configuration
 
-## API
+The `input` object accepts any valid parameter for the Automatic1111 `/sdapi/v1/txt2img` endpoint. Refer to the [Automatic1111 API Documentation](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/API) for a full list of available parameters (like `seed`, `sampler_name`, `batch_size`, `styles`, `override_settings`, etc.).
 
-The worker provides an API for inference. The API is set up using supervisor, and the configuration is specified in webui_api.conf. The API runs on port 3000.
+### Example Request
 
-## Serverless Handler
+Here's an example payload to generate an image:
 
-The serverless handler (handler.py) is a Python script that handles inference requests. It defines a function handler(event) that takes an inference request, runs the inference using the Stable Diffusion model, and returns the output.
+```json
+{
+  "input": {
+    "prompt": "a photograph of an astronaut riding a horse",
+    "negative_prompt": "text, watermark, blurry, low quality",
+    "steps": 25,
+    "cfg_scale": 7,
+    "width": 512,
+    "height": 512,
+    "sampler_name": "DPM++ 2M Karras"
+  }
+}
+```

builder/cache.py

@@ -13,15 +13,14 @@ python /stable-diffusion-webui/webui.py \
   --skip-torch-cuda-test \
   --skip-install \
   --ckpt /model.safetensors \
-  --lowram \
   --opt-sdp-attention \
   --disable-safe-unpickle \
   --port 3000 \
   --api \
   --nowebui \
   --skip-version-check \
   --no-hashing \
-  --no-download-sd-model > /webui.log 2>&1 &
+  --no-download-sd-model &
 
 echo "Starting RunPod Handler"
-python -u /rp_handler.py
+python -u /handler.py

src/rp_handler.py

@@ -1,9 +1,11 @@
 {
-  "prompt": "a photo of an astronaut riding a horse on mars",
-  "negative_prompt": "blurry, bad quality",
-  "width": 512,
-  "height": 512,
-  "num_inference_steps": 20,
-  "guidance_scale": 7.5,
-  "seed": 42
-}
+  "input": {
+    "prompt": "a photo of an astronaut riding a horse on mars",
+    "negative_prompt": "blurry, bad quality",
+    "width": 512,
+    "height": 512,
+    "num_inference_steps": 20,
+    "guidance_scale": 7.5,
+    "seed": 42
+  }
+}