Computes the text embedding for a code string. Args: code (str, list[str]): The code as a string or list

Get a simple evaluator for evolution testing

Get a simple program for evolution testing

Get config for integration tests with optillm

Get statistics for each island

Show what parameters would be used for each model

Get baseline performance for comparison

Get test cases with HTML content and expected results. These test cases include URLs that will be fetched by optillm's readurls plugin d

Get total size of all artifacts in bytes

Determines if two hexagons are disjoint given their parameters.

Determines if two hexagons are disjoint given their parameters.

Increment generation counter for an island

Checks if a point is inside a hexagon (given its parameters).

Checks if a point is inside a hexagon (given its parameters).

Check if optillm server is running on the given port

Check if the provided solution is valid. Args: problem: The original problem solution: The proposed

Check if the provided solution is valid. Args: problem: The original problem solution: The proposed

Check if the provided solution is valid. Args: problem: The original problem solution: The proposed

Check if the provided solution is valid. Args: problem: The original problem solution: The proposed

Check if the provided solution is valid. Args: problem: The original problem solution: The proposed

Check if the provided solution is valid. Args: problem: The original problem solution: The proposed

Check if the provided solution is valid. Args: problem: The original problem solution: The proposed

Validate an LU factorization A = P L U. Checks: - Presence of 'LU' with 'P','L','U' - Shapes match A (square)

Load baseline MLIR from file

Load the CustomGQAAttention class from the evolved program

Load results from iterative agent output.

Load results from OpenEvolve checkpoint.

Dynamically load the best_program.py module from the specified directory.

Load prompt template for a dataset.

Load API keys and configuration from a secrets file. Args: secrets_file: Path to the YAML secrets file Returns: Diction

Load traces from an HDF5 file Args: input_path: Path to the HDF5 file Returns: Tuple of (traces list, metadata dict)

Load traces from a JSON file Args: input_path: Path to the JSON file Returns: Tuple of (traces list, metadata dict)

Load traces from a JSONL file Args: input_path: Path to the JSONL file compress: Whether the file is compressed with gzip

Log in to Kattis. At least one of password or token needs to be provided. Returns a requests.Response with cookies needed to be able to subm

Run the complete benchmark suite

Entry point: parse CLI arguments, apply the hook, and run mlx_lm.generate. This function orchestrates the entire flow: 1. Parse command-

Main function to create OpenEvolve task files.

Main function.

Main function to generate OpenEvolve task files.

Run all tests

Run all tests

Main entry point for processing symbolic regression problems. Initializes datasets and processes problems in parallel using multiprocessing.

Main asynchronous entry point Returns: Exit code

Measure baseline performance using the original AlgoTune implementation. Args: task_instance: The AlgoTune task instance

Measure baseline performance using the original AlgoTune implementation. Args: task_instance: The AlgoTune task instance

Measure baseline performance using the original AlgoTune implementation. Args: task_instance: The AlgoTune task instance

Measure baseline performance using the original AlgoTune implementation. Args: task_instance: The AlgoTune task instance

Measure baseline performance using the original AlgoTune implementation. Args: task_instance: The AlgoTune task instance

Measure baseline performance using the original AlgoTune implementation. Args: task_instance: The AlgoTune task instance

Measure baseline performance using the original AlgoTune implementation. Args: task_instance: The AlgoTune task instance

Measure baseline performance using the original AlgoTune implementation. Args: task_instance: The AlgoTune task instance

Measure evolved solution performance. Args: program: The evolved program module problem: Problem to solve num_ru

Measure evolved solution performance. Args: program: The evolved program module problem: Problem to solve num_ru

Measure evolved solution performance. Args: program: The evolved program module problem: Problem to solve num_ru

Measure evolved solution performance. Args: program: The evolved program module problem: Problem to solve num_ru

Measure evolved solution performance. Args: program: The evolved program module problem: Problem to solve num_ru

Measure evolved solution performance. Args: program: The evolved program module problem: Problem to solve num_ru

Measure evolved solution performance. Args: program: The evolved program module problem: Problem to solve num_ru

Measure evolved solution performance. Args: program: The evolved program module problem: Problem to solve num_ru

Returns the width and height of the minimum circumscribing rectangle. Args: circles: A numpy array of shape (num_circles, 3), where each row

Normalizes a 2D vector.

Normalizes a 2D vector.

Determines whether two 1D intervals overlap, allowing for numerical tolerance.

Determines whether two 1D intervals overlap, allowing for numerical tolerance.

Parse command-line arguments.

Parse command-line arguments

Parse OpenEvolve results from the output directory. Args: task_dir: Path to the task directory task_name: Name of the ta

Compute pass@2 accuracy for a single ARC test case. Args: attempts: List of 2 numpy arrays representing model attempts. gt:

Generate comparison plot.

Determines if two polygons intersect using the Separating Axis Theorem.

Determines if two polygons intersect using the Separating Axis Theorem.

Print a comprehensive comparison summary.

Print summary comparison.

Complete pipeline to lower an MLIR file