LLM Code Gen Tester

I worked on developing a LLM code generator tester for Ollama. The tester was built using Python and was used to test the accuracy and performance of the LLM code generator.

I used IBM’s CodeNet dataset to train the LLM model and generate code snippets. The tester was used to evaluate the generated code snippets and provide feedback on the accuracy and performance of the model.

I also used Ollama to generate test cases for the problem statements and evaluate the performance of the LLM model on the test cases.