<img src="./assets/pdx.png" height="40">
# Hello, I'm PDX 👋
**PDX is a framework for prompt engineering and a dev-ops toolkit.**
At the core, it provides a mental-model on how to build and manage `agents`. An agent is a collection of prompts and/or prompt templates with information that is used to interact with the Language Models.
**Documentation**: [pdxlabs.io/docs](https://pdxlabs.io/docs)
**Website**: [pdxlabs.io](https://pdxlabs.io/)
**Quickstart**: [create-an-agent](https://pdxlabs.io/docs/getting-started/create-an-agent)
## Installation
```bash
pip install pdx
```
## Quickstart
To create your first agent, run the following command:
```bash
pdx create my_first_agent
```
Run and test out the agent by running:
```bash
pdx test my_first_agent --verbose
```
More information here: [PDX - Main Concepts](https://pdxlabs.io/docs/getting-started/main-concepts)
## Why use PDX?
- 🗃️ Low dependency footprint -> ease of production deployment and maintainance.
- 📂 Mental model to separate prompt templates from the application code. (Similar to Flask blueprint or FastAPI router).
- 📌 Version control the prompts along with their evaluation metrics.
- 📸 Logging and tracing of inputs, prompt render, and model response made easy.
- 🧯 Standardize Error handling and logging.
- 💾 Caching for lowering latency. (Coming soon)
- 📊 Observability out-of-the-box. (Coming soon)
- 📩 Log feedback of the user. (Coming soon)
- 🛎️ A/B testing of prompts. (Coming soon)
## [Demos](https://github.com/pdx-labs/demos)
Check our the demos in the [demos repository](https://github.com/pdx-labs/demos).
## Models (APIs) currently supported:
- OpenAI
- Anthropic
Raw data
{
"_id": null,
"home_page": "",
"name": "pdx",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": "",
"keywords": "prompt,LLM,prompt engineering,dev-ops,observability,apps",
"author": "Adithya Krishnan",
"author_email": "krishsandeep@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/44/e2/48901574f6e317824115bb5393e82e707cb50c9968eef002b479504d5f62/pdx-0.7.0.tar.gz",
"platform": null,
"description": "<img src=\"./assets/pdx.png\" height=\"40\">\n\n# Hello, I'm PDX \ud83d\udc4b\n\n**PDX is a framework for prompt engineering and a dev-ops toolkit.**\n\nAt the core, it provides a mental-model on how to build and manage `agents`. An agent is a collection of prompts and/or prompt templates with information that is used to interact with the Language Models.\n\n**Documentation**: [pdxlabs.io/docs](https://pdxlabs.io/docs)\n\n**Website**: [pdxlabs.io](https://pdxlabs.io/)\n\n**Quickstart**: [create-an-agent](https://pdxlabs.io/docs/getting-started/create-an-agent)\n\n## Installation\n\n```bash\npip install pdx\n```\n\n## Quickstart\n\nTo create your first agent, run the following command:\n\n```bash\npdx create my_first_agent\n```\n\nRun and test out the agent by running:\n\n```bash\npdx test my_first_agent --verbose\n```\n\nMore information here: [PDX - Main Concepts](https://pdxlabs.io/docs/getting-started/main-concepts)\n\n## Why use PDX?\n\n- \ud83d\uddc3\ufe0f Low dependency footprint -> ease of production deployment and maintainance.\n- \ud83d\udcc2 Mental model to separate prompt templates from the application code. (Similar to Flask blueprint or FastAPI router).\n- \ud83d\udccc Version control the prompts along with their evaluation metrics.\n- \ud83d\udcf8 Logging and tracing of inputs, prompt render, and model response made easy.\n- \ud83e\uddef Standardize Error handling and logging.\n- \ud83d\udcbe Caching for lowering latency. (Coming soon)\n- \ud83d\udcca Observability out-of-the-box. (Coming soon)\n- \ud83d\udce9 Log feedback of the user. (Coming soon)\n- \ud83d\udece\ufe0f A/B testing of prompts. (Coming soon)\n\n## [Demos](https://github.com/pdx-labs/demos)\n\nCheck our the demos in the [demos repository](https://github.com/pdx-labs/demos).\n\n## Models (APIs) currently supported:\n\n- OpenAI\n- Anthropic\n",
"bugtrack_url": null,
"license": "Apache-2.0",
"summary": "Prompt Engineering and Dev-Ops toolkits. A faster way to build and manage applications powered by Language Models.",
"version": "0.7.0",
"project_urls": null,
"split_keywords": [
"prompt",
"llm",
"prompt engineering",
"dev-ops",
"observability",
"apps"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "13ffb6fd406ad0227c7f749c6de638c9a54f9f31918145ee02d17f80ee00f9a9",
"md5": "1ef2574a2f921d93278cb4256081524b",
"sha256": "788704079aab8e50d9457755537f0f4a708fcf01b427d8473508231afb54ce72"
},
"downloads": -1,
"filename": "pdx-0.7.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "1ef2574a2f921d93278cb4256081524b",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 37414,
"upload_time": "2023-08-02T14:29:11",
"upload_time_iso_8601": "2023-08-02T14:29:11.407242Z",
"url": "https://files.pythonhosted.org/packages/13/ff/b6fd406ad0227c7f749c6de638c9a54f9f31918145ee02d17f80ee00f9a9/pdx-0.7.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "44e248901574f6e317824115bb5393e82e707cb50c9968eef002b479504d5f62",
"md5": "d2a9d0662736f71e3970a7d6a21567f5",
"sha256": "614f32ec21d73a7f42868876c53982249cf246ab177fa20ad565b9696aafdfb0"
},
"downloads": -1,
"filename": "pdx-0.7.0.tar.gz",
"has_sig": false,
"md5_digest": "d2a9d0662736f71e3970a7d6a21567f5",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 23918,
"upload_time": "2023-08-02T14:29:13",
"upload_time_iso_8601": "2023-08-02T14:29:13.121349Z",
"url": "https://files.pythonhosted.org/packages/44/e2/48901574f6e317824115bb5393e82e707cb50c9968eef002b479504d5f62/pdx-0.7.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-08-02 14:29:13",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "pdx"
}
JSON