# GULL-API
GULL-API is a web application backend that can be used to run Large Language Models (LLMs). The interface between the front-end and the back-end is a JSON REST API.
## Features
- Exposes a `/api` route that returns a JSON file describing the parameters of the LLM.
- Provides a `/llm` route that accepts POST requests with JSON payloads to run the LLM with the specified parameters.
## Installation
### Using Docker
1. Build the Docker image:
```
docker build -t gull-api .
```
2. Run the Docker container:
```
docker run -p 8000:8000 gull-api
```
The API will be available at `http://localhost:8000`.
### Docker Test Mode
To build and run the Docker container in test mode, use the following commands:
```bash
docker build -t gull-api .
docker run -v $(pwd)/data:/app/data -v $(pwd)/example_cli.json:/app/cli.json -p 8000:8000 gull-api
```
In test mode, an included script echo_args.sh is used instead of a real LLM. This script echoes the arguments it receives, which can be helpful for local testing.
### Local Installation
1. Clone the repository:
```
git clone https://github.com/yourusername/gull-api.git
```
2. Change to the project directory:
```
cd gull-api
```
3. Install the dependencies:
```
pip install poetry
poetry install
```
4. Configure Environment Variables (Optional):
`GULL-API` can be configured using environment variables. To do this, create a file named `.env` in the root of the project directory, and set the environment variables there. For example:
```
DB_URI=sqlite:///mydatabase.db
CLI_JSON_PATH=/path/to/cli.json
```
`GULL-API` uses the `python-dotenv` package to load these environment variables when the application starts.
5. Run the application:
```
uvicorn gull_api.main:app --host 0.0.0.0 --port 8000
```
The API will be available at `http://localhost:8000`.
## Usage
### `/api` Route
Send a GET request to the `/api` route to retrieve a JSON file describing the parameters of the LLM:
```
GET http://localhost:8000/api
```
### `/llm` Route
Send a POST request to the `/llm` route with a JSON payload containing the LLM parameters:
```
POST http://localhost:8000/llm
Content-Type: application/json
{
"Prompt": "Once upon a time",
"Top P": 0.5
}
```
### Example Requests
```bash
curl -X POST "http://localhost:8000/llm" -H "accept: application/json" -H "Content-Type: application/json" -d "{\"Instruct mode\":false, \"Maximum length\":256, \"Prompt\":\"Hello, world\", \"Stop sequences\":\"Goodbye, world\", \"Temperature\":0.7, \"Top P\":0.95}"
curl -X GET "http://localhost:8000/api" -H "accept: application/json" | python -mjson.tool
```
### Example CLI JSON
An example CLI JSON file is provided in the repository as `example_cli.json`. This file provides an example of the expected structure for defining the command-line arguments for the LLM.
## License
See LICENSE
Raw data
{
"_id": null,
"home_page": "https://github.com/mdbecker/gull_api",
"name": "gull-api",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.10,<4.0",
"maintainer_email": "",
"keywords": "api,artificial-intelligence,automation,bot,deep-learning,fastapi,GPT,language-models,large-language-models,machine-learning,microservices,natural-language-processing,NLP,openai,REST,text,text-generation,web-api",
"author": "Michael Becker",
"author_email": "mdbecker@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/ce/80/2460f474eb0f1573a592ad3bb2e74db592dad97c0b0ecd16710084301b5c/gull_api-0.0.15.tar.gz",
"platform": null,
"description": "# GULL-API\n\n\n\n\n\n\nGULL-API is a web application backend that can be used to run Large Language Models (LLMs). The interface between the front-end and the back-end is a JSON REST API.\n\n## Features\n\n- Exposes a `/api` route that returns a JSON file describing the parameters of the LLM.\n- Provides a `/llm` route that accepts POST requests with JSON payloads to run the LLM with the specified parameters.\n\n## Installation\n\n### Using Docker\n\n1. Build the Docker image:\n\n ```\n docker build -t gull-api .\n ```\n\n2. Run the Docker container:\n\n ```\n docker run -p 8000:8000 gull-api\n ```\n\nThe API will be available at `http://localhost:8000`.\n\n### Docker Test Mode\n\nTo build and run the Docker container in test mode, use the following commands:\n\n```bash\ndocker build -t gull-api .\ndocker run -v $(pwd)/data:/app/data -v $(pwd)/example_cli.json:/app/cli.json -p 8000:8000 gull-api\n```\n\nIn test mode, an included script echo_args.sh is used instead of a real LLM. This script echoes the arguments it receives, which can be helpful for local testing.\n\n\n### Local Installation\n\n1. Clone the repository:\n\n ```\n git clone https://github.com/yourusername/gull-api.git\n ```\n\n2. Change to the project directory:\n\n ```\n cd gull-api\n ```\n\n3. Install the dependencies:\n\n ```\n pip install poetry\n poetry install\n ```\n\n4. Configure Environment Variables (Optional):\n\n `GULL-API` can be configured using environment variables. To do this, create a file named `.env` in the root of the project directory, and set the environment variables there. For example:\n\n ```\n DB_URI=sqlite:///mydatabase.db\n CLI_JSON_PATH=/path/to/cli.json\n ```\n\n `GULL-API` uses the `python-dotenv` package to load these environment variables when the application starts.\n\n\n5. Run the application:\n\n ```\n uvicorn gull_api.main:app --host 0.0.0.0 --port 8000\n ```\n\nThe API will be available at `http://localhost:8000`.\n\n## Usage\n\n### `/api` Route\n\nSend a GET request to the `/api` route to retrieve a JSON file describing the parameters of the LLM:\n\n```\nGET http://localhost:8000/api\n```\n\n### `/llm` Route\n\nSend a POST request to the `/llm` route with a JSON payload containing the LLM parameters:\n\n```\nPOST http://localhost:8000/llm\nContent-Type: application/json\n\n{\n \"Prompt\": \"Once upon a time\",\n \"Top P\": 0.5\n}\n```\n\n### Example Requests\n\n```bash\ncurl -X POST \"http://localhost:8000/llm\" -H \"accept: application/json\" -H \"Content-Type: application/json\" -d \"{\\\"Instruct mode\\\":false, \\\"Maximum length\\\":256, \\\"Prompt\\\":\\\"Hello, world\\\", \\\"Stop sequences\\\":\\\"Goodbye, world\\\", \\\"Temperature\\\":0.7, \\\"Top P\\\":0.95}\"\ncurl -X GET \"http://localhost:8000/api\" -H \"accept: application/json\" | python -mjson.tool\n```\n\n### Example CLI JSON\n\nAn example CLI JSON file is provided in the repository as `example_cli.json`. This file provides an example of the expected structure for defining the command-line arguments for the LLM.\n\n## License\n\nSee LICENSE\n\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "A REST API for running Large Language Models",
"version": "0.0.15",
"project_urls": {
"Documentation": "https://github.com/mdbecker/gull_api/blob/main/README.md",
"Homepage": "https://github.com/mdbecker/gull_api",
"Repository": "https://github.com/mdbecker/gull_api"
},
"split_keywords": [
"api",
"artificial-intelligence",
"automation",
"bot",
"deep-learning",
"fastapi",
"gpt",
"language-models",
"large-language-models",
"machine-learning",
"microservices",
"natural-language-processing",
"nlp",
"openai",
"rest",
"text",
"text-generation",
"web-api"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "b4a13bbe2325265f94de23d07e2ffcd1614ec14b04edec87ad561d5ca90a431e",
"md5": "71b9f91095f1039b7ca2b948f56cb2fc",
"sha256": "cf1676278f2eee525736e40bf4a85aae2fc0b5509f2e80bf75c91600508db560"
},
"downloads": -1,
"filename": "gull_api-0.0.15-py3-none-any.whl",
"has_sig": false,
"md5_digest": "71b9f91095f1039b7ca2b948f56cb2fc",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.10,<4.0",
"size": 8333,
"upload_time": "2023-06-27T12:20:16",
"upload_time_iso_8601": "2023-06-27T12:20:16.021886Z",
"url": "https://files.pythonhosted.org/packages/b4/a1/3bbe2325265f94de23d07e2ffcd1614ec14b04edec87ad561d5ca90a431e/gull_api-0.0.15-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "ce802460f474eb0f1573a592ad3bb2e74db592dad97c0b0ecd16710084301b5c",
"md5": "4b9192074940b5f9a17abd0e40997c1c",
"sha256": "885a7e2a6e94843c1160edf1cd11097f61bec8039c25c56887ea20a0757d7d42"
},
"downloads": -1,
"filename": "gull_api-0.0.15.tar.gz",
"has_sig": false,
"md5_digest": "4b9192074940b5f9a17abd0e40997c1c",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.10,<4.0",
"size": 6883,
"upload_time": "2023-06-27T12:20:17",
"upload_time_iso_8601": "2023-06-27T12:20:17.318090Z",
"url": "https://files.pythonhosted.org/packages/ce/80/2460f474eb0f1573a592ad3bb2e74db592dad97c0b0ecd16710084301b5c/gull_api-0.0.15.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-06-27 12:20:17",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "mdbecker",
"github_project": "gull_api",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "gull-api"
}
JSON
