Skip to content

Latest commit

 

History

History
49 lines (31 loc) · 1.51 KB

t5_README.md

File metadata and controls

49 lines (31 loc) · 1.51 KB

T5-BASE Model Integration

About T5-BASE

The T5 model is a Text-to-Text Transfer Transformer model that was developed by Google Research. It's a large-scale transformer-based language model that's designed to handle a wide range of NLP tasks, including translation, summarization, and question answering.

Requirements

  • Python (v3.10+ recommended)
  • Poetry (A Python packaging and dependency management tool)

Setup

  1. For the demo to work, you need to get HuggingFaceAPI Token:

    1. Visit HuggingFace.
    2. Sign up or log in.
    3. Navigate to Profile -> Settings -> Access Tokens.
    4. Copy an existing token or create a new one.

  2. Install Dependencies

    poetry install

  3. Running The Agent Script

    open terminal goto "t5-base/src", run below command to load environment variables and run the agent.

    export HUGGING_FACE_ACCESS_TOKEN="{Your HuggingFaceAPI Token}"
poetry run python agent.py

    Check the log for "adding t5-base agent to bureau" line and copy the {agent address}.

  4. Running The User Script

    open terminal and goto "t5-base/src",run below command to load environment variables and run the client.

    export T5_BASE_AGENT_ADDRESS="{ agent address from last step }"
poetry run python client.py

After running the command, a request is sent to the agent every 30 sec till its successful.

Modify INPUT_TEXT in t5_base_user.py to translate different sentence.