Inconsistent behaviour with --code flag for train cli

[uwaisiqbal]

How to reproduce the behaviour

I am trying to train a relation extraction model using spaCy and would like to be able to call the train method from Python. Following the documentation here (https://spacy.io/api/cli#train-function), I am trying to using the following code:

    train(config_path, output_path=model_output_path / "rel",
          overrides={"paths.train": f"{train_data_path}",
                     "paths.dev": f"{dev_data_path}",
                     "code": "custom_functions.py",
                     "components.ner.source": f"{best_ner_model_path}"
                     })

This results in the following error:

✘ Error parsing config overrides
code   not a section value that can be overwritten

However, I can use the CLI to successfully run training using the -c or --code flag to point to my custom_functions.py script.

There seems to be an inconsistency between the expected behaviour of the train method and the CLI command with the code override.

Info about spaCy

• spaCy version: 3.2.1
• Platform: Darwin-20.6.0-x86_64-i386-64bit
• Python version: 3.7.12
• Pipelines: en_core_web_sm (3.2.0)

[adrianeboyd]

Hi, --code isn't an override in this sense. Overrides are only for overriding a value in the config. You want to use the code_path option with the method train_cli (or manage this import in some other way in your script):

spaCy/spacy/cli/train.py

Lines 18 to 54 in 93e9bf6

	def train_cli(
	# fmt: off
	ctx: typer.Context, # This is only used to read additional arguments
	config_path: Path = Arg(..., help="Path to config file", exists=True, allow_dash=True),
	output_path: Optional[Path] = Opt(None, "--output", "--output-path", "-o", help="Output directory to store trained pipeline in"),
	code_path: Optional[Path] = Opt(None, "--code", "-c", help="Path to Python file with additional code (registered functions) to be imported"),
	verbose: bool = Opt(False, "--verbose", "-V", "-VV", help="Display more information for debugging purposes"),
	use_gpu: int = Opt(-1, "--gpu-id", "-g", help="GPU ID or -1 for CPU")
	# fmt: on
	):
	"""
	Train or update a spaCy pipeline. Requires data in spaCy's binary format. To
	convert data from other formats, use the `spacy convert` command. The
	config file includes all settings and hyperparameters used during training.
	To override settings in the config, e.g. settings that point to local
	paths or that you want to experiment with, you can override them as
	command line options. For instance, --training.batch_size 128 overrides
	the value of "batch_size" in the block "[training]". The --code argument
	lets you pass in a Python file that's imported before training. It can be
	used to register custom functions and architectures that can then be
	referenced in the config.
	DOCS: https://spacy.io/api/cli#train
	"""
	util.logger.setLevel(logging.DEBUG if verbose else logging.INFO)
	overrides = parse_config_overrides(ctx.args)
	import_code(code_path)
	train(config_path, output_path, use_gpu=use_gpu, overrides=overrides)
	def train(
	config_path: Union[str, Path],
	output_path: Optional[Union[str, Path]] = None,
	*,
	use_gpu: int = -1,
	overrides: Dict[str, Any] = util.SimpleFrozenDict(),
	):

[uwaisiqbal]

Ah ok thanks for pointing that out. Is there a way to pass in the same dictionary of overrides into the train_cli method? It looks like I'll just have to make a call to import_code before using the train method and it should work hopefully.

[adrianeboyd]

No, I think you'd have to write your own function to do exactly that, but it would be just as short as train_cli.

[uwaisiqbal]

Great thanks for your help! Much appreciated.