ml_grid.util.logger_setup

Classes

TeeWriter

Writes to both original stream and a file, preserving all stream behaviors.

Functions

setup_logger(→ logging.Logger)

Sets up a logger that writes to a file and the console.

Module Contents

class ml_grid.util.logger_setup.TeeWriter(original_stream, log_file_path)

Writes to both original stream and a file, preserving all stream behaviors.

original_stream

log_file_path

log_file

encoding

errors

mode

write(message: str) → int

Write to both streams, return bytes written.

flush() → None

Flush both streams.

close() → None

Close only the log file, not the original stream.

isatty() → bool

Check if original stream is a TTY.

fileno()

Return file descriptor of original stream if available.

ml_grid.util.logger_setup.setup_logger(experiment_dir: str, param_space_index: int, verbose: int = 1, redirect_stdout: bool = True) → logging.Logger

Sets up a logger that writes to a file and the console.

This function creates a timestamped main experiment folder and a sub-folder for the specific run (e.g., ‘run_0’). It configures a logger to save all messages to ‘run.log’ inside the sub-folder and also redirects stdout/stderr to this logger.

Parameters:

• experiment_dir (str) – The path to the parent directory for this group of experimental runs.

• param_space_index (int) – The index of the parameter space run.

• verbose (int) – The verbosity level. Higher numbers mean more logs to the console.

• redirect_stdout (bool) – If True, redirects stdout and stderr to the logger. Defaults to True.

Returns:

The configured logger instance.

Return type:

logging.Logger