Query and images

Sample out put that we need

[{StudentName:"Bonnie Stu" ,score :  10, subject : "English Assessment Rubric",  criteria:"A" },
{StudentName:"Cassie Katy" ,score :  6, subject : "English Assessment Rubric",  criteria:"A"},
{StudentName:"Cecelia Tang", score :  6, subject : "English Assessment Rubric",  criteria:"A"},
{StudentName:"Cynthia Wang", score :  12, subject : "English Assessment Rubric",  criteria:"A"},
{StudentName:"Fayth Lim", score :  5, subject : "English Assessment Rubric",  criteria:"A"},
{StudentName:"Hannah Ng" ,score :  7, subject : "English Assessment Rubric",  criteria:"A"},
{StudentName:"Karen Smith", score :  11 ,subject : "English Assessment Rubric",  criteria:"B"},
{StudentName:"Lisa Yeo", score :  11 ,subject : "English Assessment Rubric",  criteria:"B"},
{StudentName:"Oliver Wong" ,score :  8, subject : "English Assessment Rubric",  criteria:"B"},
{StudentName:"Patrick Tan" ,score :  7 ,subject : "English Assessment Rubric",  criteria:"A"},
{StudentName:"Subject Test" ,score :  10, subject : "English Assessment Rubric",  criteria:"A"},
{StudentName:"Test Adam", score :  6, subject : "English Assessment Rubric",  criteria:"B"}
{StudentName:"Test3 Test3" ,score :  12, subject : "English Assessment Rubric",  criteria:"B"},
{StudentName:"chan Mimi", score :  8, subject : "English Assessment Rubric",  criteria:"B"}]

1. Getting data from source and loaded into MindsDB through ETL process. With rating and percentage data
2. Create a model by training current dataset to train dataset to identify the subject which rating is 50% lower than max_rating.
3. predict the percentage based on a subject.

Conclusion, by doing these testing, it doesnt fulfill the requirements of regression or classification  methods, because the rating is not predicted based on some conditions like hour of learn, hour of study. A simple SQL query will solve the above problem.

1. Getting data, rating data from mysql source.

1. Getting data, rating data from mysql source.
2. write a python code, to train and embed data into embed model using vectorization method from openAI.
3. To query data based on test cases question, but the result is not expected.
4. Conclusion, the embedded model method is not suitable for statistical query. In fact, the embedded model is good peforming at text extraction, question answering