CF_CodeDebug.yaml

name: "CodeDebug_Flow"
description: "ToDO: add description"

max_rounds: 2 # ToDo: To increase to 4

input_interface:
  - "problem_description"
  - "input_description"
  - "output_description"
  - "io_examples_and_explanation"
  - "public_tests_individual_io"
output_interface:
  - "code"

subflows_config:
  CodeGenerator:
    _target_: .CF_Code.instantiate_from_default_config
    name: "CodeGenerator"
    model_name: "gpt-4"
    human_message_prompt_template:
      template: |2-
        {{testing_results_summary}}


        Consider the problem statement, the last proposed solution, and its issue. Provide a corrected version of the code that solves the original problem and resolves the issue, without any explanation, in the following format:
        ```python
        {{code_placeholder}}
        ```
      input_variables:
        - testing_results_summary
      partial_variables:
        code_placeholder: "{{python_code}}"
    input_interface_initialized:
      - "testing_results_summary"
  CodeTestingCritic:
    _target_: .CF_CodeTesting.instantiate_from_default_config

topology:
  # ~~~ Code Generator ~~~
  - goal: "Generate/refine a solution."

    ### Input Interface
    input_interface:
      _target_: flows.interfaces.KeyInterface
      additional_transformations:
        - _target_: flows.data_transformations.KeyMatchInput

    ### Flow Specification
    flow: CodeGenerator

    ### Output Interface
    output_interface:
      _target_: flows.interfaces.KeyInterface
      additional_transformations:
        - _target_: flows.data_transformations.RegexFirstOccurrenceExtractor
          regex: '(?<=```python)([\s\S]*?)(?=```)'
          regex_fallback: '(?<=```)([\s\S]*?)(?=```)'
          input_key: "api_output"
          output_key: "code"
          strip: True
          assert_unique: True
      keys_to_select:
        - "code"

    reset: false

  # ~~~ Code Testing Critic ~~~
  - goal: "Test the code on the public tests and provide a results summary."

    ### Input Interface
    input_interface:
      _target_: flows.interfaces.KeyInterface
      additional_transformations:
        - _target_: flows.data_transformations.KeyMatchInput

    ### Flow Specification
    flow: CodeTestingCritic

    ### Output Interface
    output_interface:
      _target_: flows.interfaces.KeyInterface
      additional_transformations:
        - _target_: martinjosifoski.CC_flows.src.data_transformations.CorrectnessFlag
          input_key: "public_tests_results"
          output_key: "all_tests_passed"
        - _target_: martinjosifoski.CC_flows.src.data_transformations.TestingResultsSummaryGeneration
          output_key: "testing_results_summary"

          single_test_error_message: True

          no_error_template: |2-
            ${.issue_title}
            All of the executed tests passed.

          compilation_error_template: |2-
            ${.issue_title}
            The execution resulted in a compilation error.
            ## Compilation error message:
            {{error_message}}
          timeout_error_template: |2-
            ${.issue_title}
            The execution timed out, the solution is not efficient enough.
          runtime_error_template: |2-
            ${.issue_title}
            The execution resulted in a runtime error on the following test.
            ## [Failed test] Input
            ```
            {{test_input}}
            ```
            ## [Failed test] Runtime error message
            {{error_message}}
          single_test_error_template: |2-
            ${.issue_title}
            The Python code does not solve the problem in the problem description due to logical errors. It fails the following test:
            ## [Failed test] Input
            ```
            {{test_input}}
            ```
            ## [Failed test] Expected output
            ```
            {{expected_output}}
            ```
            ## [Failed test] Generated output
            ```
            {{generated_output}}
            ```
          all_tests_header: |2-
            ${.issue_title}
            The Python code does not solve the problem in the problem description due to logical errors. It fails on the following tests.
          test_error_template: |2-
            ## [Failed test {{idx}}]
            ### [Failed test {{idx}}] Input
            ```
            {{test_input}}
            ```
            ### [Failed test {{idx}}] Expected output
            ```
            {{expected_output}}
            ```
            ### [Failed test {{idx}}] Generated output
            ```
            {{generated_output}}
            ```
          tests_separator: "\n\n"

          issue_title: "# Issue with the last proposed solution"

    reset: true