cellmap_flow.globals

Attributes

`logger`
`SERVER_CONFIG_PATH`
`SERVER_CONFIG_DEFAULTS`
`SERVER_CONFIG_KEYS`
`input_norms`
`postprocess`
`viewer`
`g`

Classes

`Flow`
`LogHandler`	Handler instances dispatch logging events to specific destinations.

Functions

`load_server_config_cache`(→ Optional[Dict[str, Any]])	Load server config from cache file. Returns None if not found.
`save_server_config_cache`(→ None)	Save server config to cache file.
`current_input_norm_config`(→ dict)	Return the dashboard's current input_norm as a JSON-serializable dict.
`get_blockwise_tasks_dir`()

Module Contents

cellmap_flow.globals.logger

cellmap_flow.globals.SERVER_CONFIG_PATH

cellmap_flow.globals.SERVER_CONFIG_DEFAULTS

cellmap_flow.globals.SERVER_CONFIG_KEYS = ['queue', 'charge_group', 'nb_cores_master', 'nb_cores_worker', 'nb_workers']

cellmap_flow.globals.load_server_config_cache() → Dict[str, Any] | None: Load server config from cache file. Returns None if not found.

cellmap_flow.globals.save_server_config_cache(config: Dict[str, Any]) → None: Save server config to cache file.

cellmap_flow.globals.input_norms = []

cellmap_flow.globals.postprocess = []

cellmap_flow.globals.viewer = None

class cellmap_flow.globals.Flow

jobs: List[Any]

models_config: List[Any]

servers: List[Any]

raw: Any | None

input_norms: List[Any]

postprocess: List[Any]

viewer: Any | None

dataset_path: str | None

model_catalog: dict

queue: str

charge_group: str

nb_cores_master: int

nb_cores_worker: int

nb_workers: int

tmp_dir: str | None

blockwise_tasks_dir: str | None

neuroglancer_thread: Any | None

pipeline_inputs: List[Any]

pipeline_outputs: List[Any]

pipeline_edges: List[Any]

pipeline_normalizers: List[Any]

pipeline_models: List[Any]

pipeline_postprocessors: List[Any]

shaders: dict

shader_controls: dict

log_buffer: collections.deque

log_clients: list

NEUROGLANCER_URL: str | None

INFERENCE_SERVER: Any | None

CUSTOM_CODE_FOLDER: str

bbx_generator_state: dict

property finetune_job_manager

minio_state: dict

annotation_volumes: dict

output_sessions: dict

to_dict()

save_server_config(): Save current server config attributes to cache.

get_output_dtype(model_output_dtype)

classmethod run(zarr_path, model_configs, queue='gpu_h100', charge_group='cellmap', input_normalizers=None, post_processors=None)

classmethod stop()

classmethod delete()

cellmap_flow.globals.g

class cellmap_flow.globals.LogHandler(level=NOTSET)

Handler instances dispatch logging events to specific destinations.

The base handler class. Acts as a placeholder which defines the Handler interface. Handlers can optionally use Formatter instances to format records as desired. By default, no formatter is specified; in this case, the ‘raw’ message as determined by record.message is logged.

emit(record)

Do whatever it takes to actually log the specified logging record.

This version is intended to be implemented by subclasses and so raises a NotImplementedError.

cellmap_flow.globals.current_input_norm_config() → dict

Return the dashboard’s current input_norm as a JSON-serializable dict.

Reads g.input_norm_config if populated; otherwise reconstructs the dict from the live g.input_norms instances via their .to_dict(). The fallback matters because some startup paths (yaml_cli) populate g.input_norms from the YAML at server boot but never touch input_norm_config – if the user submits training without first hitting /api/run, the manifest would otherwise be written empty.

cellmap_flow.globals.get_blockwise_tasks_dir()