worker: Move ai-worker package to go-livepeer

.github/workflows/validate-openapi-on-pr.yaml

@@ -37,13 +37,3 @@ jobs:
             echo "::error:: OpenAPI spec has changed. Please run 'python gen_openapi.py' in the 'runner' directory and commit the changes."
             exit 1
           fi
-
-      - name: Generate Go bindings
-        run: make
-
-      - name: Check for Go bindings changes
-        run: |
-          if ! git diff --exit-code; then
-            echo "::error::Go bindings have changed. Please run 'make' at the root of the repository and commit the changes."
-            exit 1
-          fi

README.md

@@ -1,38 +1,69 @@
-# ai-worker
+# ai-runner
 
 > [!WARNING]
 > The AI network is in it's **Beta** phase and although it is ready for production it is still under development. Please report any issues you encounter to the [Livepeer Discord](https://discord.gg/7nbPbTK).
 
-This repository hosts the AI worker and runner for processing inference requests on the Livepeer AI subnet.
+This repository hosts the AI runner for processing AI inference jobs on the Livepeer network.
 
 ## Overview
 
-The AI worker repository includes:
+The AI runner is a containerized Python application which processes inference requests on Livepeer AI's Pipelines and models. It loads models into GPU memory and exposes a REST API other programs like [the Livepeer node AI worker](../README.md) can use to request AI inference requests. The AI runner code sits in the [runner](https://github.com/livepeer/ai-runner/tree/main/runner) directory.
 
-- **Runner**: The [AI runner](https://github.com/livepeer/ai-worker/tree/main/runner), a containerized Python application, processes inference requests on Livepeer AI's Pipelines and models, providing a REST API for model interaction.
+## Build
 
-- **Worker**: The [AI worker](https://github.com/livepeer/ai-worker) allows the [ai-video](https://github.com/livepeer/go-livepeer/tree/ai-video) branch of [go-livepeer](https://github.com/livepeer/go-livepeer/tree/ai-video) to interact with the AI runner. It includes golang API bindings, a worker for routing inference requests, and a Docker manager for AI runner containers.
+To build the AI runner locally and run examples, follow these steps:
 
-### Runner
+1. Follow the instructions in this document to download model checkpoints and build the runner image.
+2. Generate Go bindings for the runner OpenAPI spec with `make codegen`.
+3. Run any examples in the `cmd/examples` directory, e.g., `go run cmd/examples/text-to-image/main.go <RUNS> <PROMPT>`.
 
-The AI runner's code is in the [runner](https://github.com/livepeer/ai-worker/tree/main/runner) directory. For more details, see the [AI runner README](./runner/README.md).
+## Architecture
 
-### Worker
+A high level sketch of how the runner is used:
 
-The AI worker's code is in the [worker](https://github.com/livepeer/ai-worker/tree/main/worker) directory. It includes:
+![Architecture](./docs/images/architecture.png)
 
-- **Golang API Bindings**: Generated from the AI runner's OpenAPI spec using `make codegen`.
-- **Worker**: Listens for inference requests from the Livepeer AI subnet and routes them to the AI runner.
-- **Docker Manager**: Manages AI runner containers.
+The AI runner, found in the [app](./runner/app) directory, consists of:
 
-## Build
+- **Routes**: FastAPI routes in [app/routes](./runner/app/routes) that handle requests and delegate them to the appropriate pipeline.
+- **Pipelines**: Modules in [app/pipelines](./runner/app/pipelines) that manage model loading, request processing, and response generation for specific AI tasks.
 
-The AI worker and runner are designed to work with the [ai-video](https://github.com/livepeer/go-livepeer/tree/ai-video) branch of [go-livepeer](https://github.com/livepeer/go-livepeer/tree/ai-video). You can run both independently for testing. To build the AI worker locally and run examples, follow these steps:
+It also includes utility scripts:
 
-1. Follow the [README](./runner/README.md) instructions in the [runner](./runner/README.md) directory to download model checkpoints and build the runner image.
-2. Generate Go bindings for the runner OpenAPI spec with `make codegen`.
-3. Run any examples in the `cmd/examples` directory, e.g., `go run cmd/examples/text-to-image/main.go <RUNS> <PROMPT>`.
+- **[bench.py](./runner/bench.py)**: Benchmarks the runner's performance.
+- **[gen_openapi.py](./runner/gen_openapi.py)**: Generates the OpenAPI specification for the runner's API endpoints.
+- **[dl_checkpoints.sh](./runner/dl_checkpoints.sh)**: Downloads model checkpoints from Hugging Face.
+- **[modal_app.py](./runner/modal_app.py)**: Deploys the runner on [Modal](https://modal.com/), a serverless GPU platform.
+
+## OpenAPI Specification
+
+Regenerate the OpenAPI specification for the AI runner's API endpoints with:
+
+```bash
+python gen_openapi.py
+```
+
+To correspondingly generate the Go client bindings in the go-livepeer repository,
+you should install `livepeer/go-livepeer` and run:
+```bash
+# in the go-livepeer repo
+make ai_worker_codegen
+```
+
+Alternatively, if you want to test the client from a development version of
+`ai-runner`, you can specify a commit hash or branch to generate from:
+```bash
+# for commit `aa7ab76`
+make ai_worker_codegen REF=aa7ab76
+# for branch `test`
+make ai_worker_codegen REF=refs/heads/vg/chore/test
+```
 
 ## Development documentation
 
-For more on developing and debugging the AI runner, see the [development documentation](./dev/README.md).
+For more on developing and debugging the AI runner, see the [development documentation](./docs/development-guide.md).
+
+## Credits
+
+Based off of [this repo](https://github.com/huggingface/api-inference-community/tree/main/docker_images/diffusers).
+