Skip to main content

AgentX: Seamlessly integrate intelligent agents into your projects. Empower your applications with advanced AI capabilities.

Project description

AgentX

AgentX is a versatile toolkit designed to seamlessly integrate intelligent agents into your projects. With AgentX, developers can effortlessly bridge the gap between different frameworks and platforms, empowering their applications with advanced AI capabilities. AgentX facilitates the connection and collaboration of diverse agents, enhancing the synergy between neural networks and decision-making processes. Whether you're working on machine learning, robotics, or data analysis projects, AgentX provides the essential tools to create dynamic and adaptive systems. Experience the power of intelligent fusion with AgentX and unlock new possibilities for your applications.

Features 🔮

  1. Seamless Integration: AgentX allows for the seamless integration of intelligent agents into your projects, regardless of the underlying frameworks or platforms.

  2. Versatility: With AgentX, you can connect and collaborate with diverse agents, enabling a wide range of applications in fields such as machine learning, robotics, and data analysis.

  3. Advanced AI Capabilities: Empower your applications with advanced AI capabilities by leveraging AgentX's tools and functionalities.

  4. Flexibility: AgentX offers flexibility in designing and implementing dynamic and adaptive systems, providing developers with the freedom to explore new possibilities.

  5. Framework Agnostic: AgentX is framework-agnostic, meaning it can work with various AI frameworks, allowing you to choose the best tools for your specific project needs.

  6. Efficiency: AgentX is designed for efficiency, ensuring optimal performance and resource utilization in your applications.

  7. Community Support: Join a vibrant community of developers using AgentX, where you can find support, share ideas, and collaborate on enhancing the toolkit.

  8. Documentation: Comprehensive documentation and examples are provided to help you get started quickly and make the most out of AgentX in your projects.

  9. Continuous Updates: Expect regular updates and improvements to AgentX, ensuring that you stay up-to-date with the latest advancements in AI integration and functionality.

  10. Open Source: AgentX is an open-source project, inviting contributions from the community to help enhance its features, usability, and effectiveness in real-world applications.

CLI Support

AgentX support usage Via CLI for more information just use

agentx serve --help

Torch Serving Example

Here's an Example of Serving and using Model with AgentX with torch Backend.

[!NOTE] You can just don't pass prompter or prompt template and just pass the tokenizer engine will use tokenizer chat template

from agentx import ServeEngine, PromptTemplates
import torch

engine = ServeEngine.from_torch_pretrained(
    huggingface_repo_id="meta-llama/Meta-Llama-3-8B-Instruct",
    sample_config=None,  # torch support Auto Set from Model Config.
    # sample_config=EngineGenerationConfig(
    #     max_new_tokens=8192,
    #     max_sequence_length=8192,
    #     top_k=20,
    #     top_p=0.95,
    #     temperature=0.2,
    # ),
    prompter=PromptTemplates.from_prompt_templates(
        "llama_3",
        eos_token=None,  # Auto Set is supported for some models
        bos_token=None  # Auto Set is supported for some models
    ),
    tokenizer_huggingface_repo_id=None,
    bnb_4bit_compute_dtype=torch.float16,
    device_map="auto",
    _attn_implementation="sdpa",
    bnb_4bit_quant_type="fp4"
)

response = engine.execute("You Are Using AgentX Execute Function")
for char in engine.process(
        "You Are Using AgentX Process/Stream Function"
):
    print(char, end="")

# Do you Need CHAT GUI?

engine.build_inference().launch(server_name="0.0.0.0", server_port=7860)

Ollama Serving Example

Here's an Example of Serving and using Model with AgentX with Ollama Backend. (Ollama and Llama CPP kinda have same usage examples.)

from agentx import ServeEngine, PromptTemplates, EngineGenerationConfig

engine = ServeEngine.from_ollama_model(
    ollama_model="LLAMA-3-OLLAMA",
    sample_config=EngineGenerationConfig(
        max_new_tokens=8192,
        max_sequence_length=8192,
        top_k=20,
        top_p=0.95,
        temperature=0.2,
    ),
    prompter=PromptTemplates.from_prompt_templates(
        "llama_3",
        eos_token=None,  # Auto Set is supported for some models
        bos_token=None  # Auto Set is supported for some models
    ),
)

response = engine.execute("You Are Using AgentX Execute Function")
for char in engine.process(
        "You Are Using AgentX Process/Stream Function"
):
    print(char, end="")

# Do you Need CHAT GUI?

engine.build_inference().launch(server_name="0.0.0.0", server_port=7860)

Contributing

If you would like to contribute to AgentX, please follow the guidelines outlined in the CONTRIBUTING.md file in the repository.

License

AgentX is licensed under the MIT. See the LICENSE.md file for more details.

Support

For any questions or issues, please get in touch with me at erfanzare810@gmail.com.

Thank you for using AgentX! We hope it will help you have a personal computer experience.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

agentx-0.0.27.tar.gz (26.7 kB view details)

Uploaded Source

Built Distribution

AgentX-0.0.27-py3-none-any.whl (36.8 kB view details)

Uploaded Python 3

File details

Details for the file agentx-0.0.27.tar.gz.

File metadata

  • Download URL: agentx-0.0.27.tar.gz
  • Upload date:
  • Size: 26.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.0.0 CPython/3.11.4

File hashes

Hashes for agentx-0.0.27.tar.gz
Algorithm Hash digest
SHA256 e2180e2ba0e0f821bff5ec26b1dc591ee44b5a7fc40de30dc519821578f1c67f
MD5 fd9afecc504ebd7b6924cc29a8d7da4b
BLAKE2b-256 564af6c2d3eb603d59f98df400a5c09dd1b3b5e65bd8e520f002e7875fe4a4e8

See more details on using hashes here.

File details

Details for the file AgentX-0.0.27-py3-none-any.whl.

File metadata

  • Download URL: AgentX-0.0.27-py3-none-any.whl
  • Upload date:
  • Size: 36.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.0.0 CPython/3.11.4

File hashes

Hashes for AgentX-0.0.27-py3-none-any.whl
Algorithm Hash digest
SHA256 92ff388e5610d3579e40314588498e4c42af6e5442b12a5ee0469d4f128c44e7
MD5 d23c8ab725da12a04bfcf295896b2cbd
BLAKE2b-256 0956d3e6f7c65f1f26b58398f7dcc8bab79d693af5efefb299cb15a812741ada

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page