We provide some easy-to-use demos for you to directly experience the effect of Yival in the README on Github.

• google colab :
• notebook : headline_generation

The goal of this demo is to generate corresponding page headlines based on the names of startup companies. YiVal supports automatic generation of related prompts and test data according to this goal, and self-evaluation based on the configured evaluator. It provides different result selection methods such as AHP to confirm the final result.

For how to confirm the prompt used by the generator, we recommend using a step-by-step optimization pipeline mode, continuously adjusting based on the test case results. We provide an example :

• google colab :
• notebook : animal_story

Yival now supports image now , and in near future We will equip Yival with full modality capabilities (including sound, video, etc.) in near future.

In this demo, we only need to provide the initial prompt, and Yival will generate a variety of animal types and personalities, and write cute and concise animal stories based on different story templates. Finally, corresponding images are generated through Midjourney. What's surprising is that this series of actions is completely automated, you only need to provide a prompt.

• google colab:
• notebook : model_comparison

For NLP practitioners, a common issue is evaluating the capabilities of different models. Yival provides a variety of generalized evaluation methods, requiring only the provision of data and configuration files.In this demo, we conducted a horizontal comparison of the QA capabilities of the following four models:

• gpt-3.5-turbo
• llama-2-13b-chat
• llama-2-70b-chat
• vicuna-13b

To get started with a demo for basic interactive mode of YiVal, run the following command:

yival demo --basic_interactive

Once started, navigate to the following address in your web browser:

http://127.0.0.1:8073/interactive

For more details on this demo, check out the Basic Interactive Mode Demo.

yival demo --qa_expected_results

Once started, navigate to the following address in your web browser: http://127.0.0.1:8073/

Click to view the screenshot

For more details, check out the Question Answering with expected result evaluator