CLI Argument Reference

This document provides a comprehensive reference of all CLI arguments used across CellMap Flow’s command-line interfaces.

Standardization Principles 

All CellMap Flow CLIs follow these naming conventions:

Hyphenated multi-word arguments: Use hyphens (--data-path), not underscores (--data_path)
Consistent terminology: --project for billing groups (not --charge_group)
Clear, descriptive names: Arguments should be self-documenting
Short and long forms: Commonly used arguments have single-letter shortcuts

Common Arguments 

These arguments are used consistently across multiple CLI interfaces.

–data-path, -d 

Type:: string
Required:: Yes (for most commands)
Default:: None
Description:: Path to the input dataset, typically a Zarr file or group

Examples:

--data-path /nrs/cellmap/data/dataset.zarr
-d /nrs/cellmap/data/dataset.zarr/s1

Used in:

cellmap_flow (all subcommands)
cellmap_flow_server (all subcommands)
cellmap_flow_multiple
cellmap_flow_yaml (in YAML as data_path)

–project, -P 

Type:: string
Required:: Yes (for job submission)
Default:: None
Description:: Project or chargeback group for billing purposes

Examples:

--project cellmap
-P my_lab_group

Used in:

cellmap_flow (all subcommands)
cellmap_flow_multiple
cellmap_flow_yaml (in YAML as charge_group or project)

Migration Note: Previously --charge_group. Updated to --project for consistency.

–queue, -q 

Type:: string
Required:: No
Default:: gpu_h100
Description:: Compute queue to submit jobs to

Examples:

--queue gpu_rtx
-q local

Used in:

cellmap_flow (all subcommands)
cellmap_flow_multiple
cellmap_flow_yaml (in YAML as queue)

–name, -n 

Type:: string
Required:: Depends on subcommand
Default:: Auto-generated from model/script path
Description:: Custom name for the model, used for job naming and output identification

Examples:

--name mitochondria_model
-n my_experiment_v2

Used in:

cellmap_flow cellmap-model
cellmap_flow_server cellmap-model
cellmap_flow_multiple (per-model in YAML)

Model Path Arguments 

–script-path, -s 

Type:: string
Required:: Yes (for script-based models)
Default:: None
Description:: Path to Python script containing model specification

Examples:

--script-path /groups/cellmap/scripts/my_model.py
-s ./model_definition.py

Used in:

cellmap_flow script
cellmap_flow_server script
cellmap_flow_multiple --script

–model-path, -m 

Type:: string
Required:: Yes (for bioimage.io models)
Default:: None
Description:: Path to bioimage.io model file or directory

Examples:

--model-path /models/mito_segmentation.bioimage.io
-m ./my_model.zip

Used in:

cellmap_flow bioimage
cellmap_flow_server bioimage
cellmap_flow_multiple --bioimage

–config-folder, -f 

Type:: string
Required:: Yes (for CellMap models)
Default:: None
Description:: Path to model configuration folder containing model definition files

Examples:

--config-folder /groups/cellmap/models/mito_16
-f ./model_configs/nuclei

Used in:

cellmap_flow cellmap-model
cellmap_flow_server cellmap-model
cellmap_flow_multiple --cellmap-model

DaCapo-Specific Arguments 

–run-name, -r 

Type:: string
Required:: Yes (for DaCapo models)
Default:: None
Description:: Name of the DaCapo training run to use

Examples:

--run-name mito_run_v3
-r nuclei_training_20241015

Used in:

cellmap_flow dacapo
cellmap_flow_server dacapo
cellmap_flow_multiple --dacapo

–iteration, -i 

Type:: integer
Required:: No
Default:: 0
Description:: Training iteration/checkpoint to use from the DaCapo run

Examples:

--iteration 60000
-i 40000

Used in:

cellmap_flow dacapo
cellmap_flow_server dacapo
cellmap_flow_multiple --dacapo

Fly Model-Specific Arguments 

–checkpoint, -c 

Type:: string
Required:: Yes (for Fly models)
Default:: None
Description:: Path to the Fly model checkpoint file

Examples:

--checkpoint /models/fly_checkpoint.pth
-c ./checkpoints/epoch_100.pth

Used in:

cellmap_flow_server fly
cellmap_flow_fly (in YAML as checkpoint)

–channels, -ch 

Type:: string (comma-separated)
Required:: Yes (for Fly models)
Default:: None
Description:: Comma-separated list of output channel names

Examples:

--channels mito,nucleus,golgi
-ch cell_membrane,er

Used in:

cellmap_flow_server fly
cellmap_flow_fly (in YAML as classes)

–input-voxel-size, -ivs 

Type:: string (comma-separated)
Required:: Yes (for Fly models)
Default:: None
Description:: Comma-separated input voxel size in nanometers (z,y,x)

Examples:

--input-voxel-size 8,8,8
-ivs 4,4,4

Used in:

cellmap_flow_server fly
cellmap_flow_fly (in YAML as input_voxel_size)

–output-voxel-size, -ovs 

Type:: string (comma-separated)
Required:: Yes (for Fly models)
Default:: None
Description:: Comma-separated output voxel size in nanometers (z,y,x)

Examples:

--output-voxel-size 8,8,8
-ovs 4,4,4

Used in:

cellmap_flow_server fly
cellmap_flow_fly (in YAML as output_voxel_size)

Bioimage-Specific Arguments 

–edge-length-to-process, -e 

Type:: integer
Required:: No
Default:: Auto-calculated
Description:: For 2D models, desired edge length of chunks to process; batch size (z) will be adjusted accordingly

Examples:

--edge-length-to-process 512
-e 1024

Used in:

cellmap_flow bioimage
cellmap_flow_server bioimage
cellmap_flow_multiple --bioimage

Server-Specific Arguments 

–port, -p 

Type:: integer
Required:: No
Default:: 0 (auto-assign)
Description:: Port number for the inference server to listen on

Examples:

--port 8080
-p 5000

Used in:

cellmap_flow_server (all model subcommands)

–debug 

Type:: flag (boolean)
Required:: No
Default:: False
Description:: Run server in debug mode with verbose logging

Examples:

--debug

Used in:

cellmap_flow_server (all model subcommands)

–certfile 

Type:: string
Required:: No (for HTTPS)
Default:: None
Description:: Path to SSL certificate file for HTTPS connections

Examples:

--certfile /path/to/cert.pem

Used in:

cellmap_flow_server (all model subcommands)

–keyfile 

Type:: string
Required:: No (for HTTPS)
Default:: None
Description:: Path to SSL private key file for HTTPS connections

Examples:

--keyfile /path/to/key.pem

Used in:

cellmap_flow_server (all model subcommands)

UI Server Arguments 

–neuroglancer-url 

Type:: string
Required:: Yes (for UI server)
Default:: None
Description:: Neuroglancer viewer URL to connect to

Examples:

--neuroglancer-url http://localhost:8080/neuroglancer

Used in:

cellmap_flow_server run-ui-server

–inference-host, -i 

Type:: string
Required:: Yes (for UI server)
Default:: None
Description:: Inference host address(es)

Examples:

--inference-host http://gpu-node-01:5000
-i http://localhost:8000,http://localhost:8001

Used in:

cellmap_flow_server run-ui-server

Optional/Advanced Arguments 

–scale, -r 

Type:: string
Required:: No
Default:: Auto-detected
Description:: Input scale level for multi-resolution datasets (e.g., s0, s1, s2)

Examples:

--scale s1
-r s2

Used in:

cellmap_flow_multiple (per-model)
cellmap_flow_yaml (in YAML as scale)

–log-level 

Type:: choice (DEBUG, INFO, WARNING, ERROR, CRITICAL)
Required:: No
Default:: INFO
Description:: Logging verbosity level

Examples:

--log-level DEBUG
--log-level WARNING

Used in:

cellmap_flow (main CLI group)
cellmap_flow_server (main CLI group)

Migration Guide 

Old vs New Argument Names 

If you have scripts using old argument names, use this table to update them:

Old Argument	New Argument	Notes
`--data_path`	`--data-path`	Standardized hyphenation
`--charge_group`	`--project`	Clearer terminology
`--script_path`	`--script-path`	Standardized hyphenation
`--model_path`	`--model-path`	Standardized hyphenation
`--config_folder`	`--config-folder`	Standardized hyphenation
`--edge_length_to_process`	`--edge-length-to-process`	Standardized hyphenation
`--input_voxel_size`	`--input-voxel-size`	Standardized hyphenation
`--output_voxel_size`	`--output-voxel-size`	Standardized hyphenation
`--neuroglancer_url`	`--neuroglancer-url`	Standardized hyphenation
`--inference_host`	`--inference-host`	Standardized hyphenation
`--folder_path`	`--folder-path`	Standardized hyphenation

Automated Migration Script 

For large codebases, you can use this sed script to update your YAML and shell scripts:

# Update argument names in shell scripts
sed -i 's/--data_path/--data-path/g' *.sh
sed -i 's/--charge_group/--project/g' *.sh
sed -i 's/--script_path/--script-path/g' *.sh
sed -i 's/--model_path/--model-path/g' *.sh
sed -i 's/--config_folder/--config-folder/g' *.sh

Examples by Use Case 

Single Model Inference 

# Script-based model
cellmap_flow script \
    --script-path /path/to/model.py \
    --data-path /data/volume.zarr/s1 \
    --project cellmap \
    --queue gpu_h100

# CellMap model
cellmap_flow cellmap-model \
    --config-folder /models/mito_config \
    --name mito_v2 \
    --data-path /data/volume.zarr/s1 \
    --project cellmap

Multiple Models 

cellmap_flow_multiple \
    --data-path /data/volume.zarr/s1 \
    --project cellmap \
    --queue gpu_h100 \
    --script -s /models/mito.py -n mito \
    --script -s /models/nucleus.py -n nucleus \
    --cellmap-model -f /models/golgi -n golgi

YAML Configuration 

data_path: "/data/volume.zarr/s1"
project: "cellmap"  # or use 'charge_group'
queue: "gpu_h100"

models:
  - type: script
    script_path: "/models/mito.py"
    name: "mito"
  - type: cellmap-model
    config_folder: "/models/nucleus_config"
    name: "nucleus"
    scale: "s2"

Manual Server Control 

# Start inference server
cellmap_flow_server script \
    --script-path /models/model.py \
    --data-path /data/volume.zarr \
    --port 8080 \
    --debug

# Start UI server
cellmap_flow_server run-ui-server \
    --neuroglancer-url http://localhost:8080/ng \
    --inference-host http://localhost:5000