cellpose_sam.yaml

# Cellpose-SAM v2's flows on jrc_mus-salivary-1:
#   cellmap_flow yaml example/cellpose_sam.yaml
# The cellpose type runs in the cellpose4 pixi environment without an `env`
# (Cellpose 4 cannot share the default environment's cellpose 3); give one
# to run it elsewhere. Its parameters, output modes and licence
# (non-commercial): "Cellpose" in docs/source/yaml_config.rst.
data_path: "/nrs/cellmap/data/jrc_mus-salivary-1/jrc_mus-salivary-1.zarr/recon-1/em/fibsem-uint8"

charge_group: "cellmap"

queue: "gpu_h100"

# Named as the Models tab's Cellpose panel names it, so its row shows it
# ticked; tick another output there (Probability, Masks) to run it beside.
models:
  # Flows and probability (flow_y, flow_x, cell): the layer colours each
  # voxel by its flow's direction, dimmed by the probability. Add
  # CellposeMasksPostprocessor on the Output tab to see masks made from them.
  cellpose_sam_v2:
    type: "cellpose"
    # Left out, the type's default model, cpsam_v2; "cpsam" for Cellpose-SAM,
    # or the path of finetuned Cellpose weights.
    # pretrained_model: "cpsam_v2"
    output: "flows"
    # nm; the server reads s3. Cellpose-SAM sees objects best about 30
    # voxels across, so pick the voxel size at which yours are roughly that.
    voxel_size: 64

# With output: "masks", make the per-chunk ids unique across chunks and show
# the layer as a segmentation:
# json_data:
#   postprocess:
#     - name: MortonSegmentationRelabeling