mirrors/docling
1
0
Fork 0
mirror of https://github.com/DS4SD/docling.git synced 2025-07-24 19:14:23 +00:00
Code Issues Packages Projects Releases Wiki Activity

Move expensive imports closer to usage

Browse Source
This commit is contained in:
William Easton 2025-06-28 22:14:21 -05:00
parent bb99be6c24
commit 8a7412ce5b
No known key found for this signature in database
5 changed files with 17 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
docling/models/base_ocr_model.py
View File

@ -10,7 +10,6 @@ from docling_core.types.doc import BoundingBox, CoordOrigin
from docling_core.types.doc.page import TextCell from docling_core.types.doc.page import TextCell
from PIL import Image, ImageDraw from PIL import Image, ImageDraw
from rtree import index from rtree import index
from scipy.ndimage import binary_dilation, find_objects, label
from docling.datamodel.accelerator_options import AcceleratorOptions from docling.datamodel.accelerator_options import AcceleratorOptions
from docling.datamodel.base_models import Page from docling.datamodel.base_models import Page
@ -31,6 +30,8 @@ class BaseOcrModel(BasePageModel, BaseModelWithOptions):
options: OcrOptions, options: OcrOptions,
accelerator_options: AcceleratorOptions, accelerator_options: AcceleratorOptions,
): ):
from scipy.ndimage import binary_dilation, find_objects, label
self.enabled = enabled self.enabled = enabled
self.options = options self.options = options

3
docling/models/layout_model.py
View File

@ -7,7 +7,6 @@ from typing import Optional
import numpy as np import numpy as np
from docling_core.types.doc import DocItemLabel from docling_core.types.doc import DocItemLabel
from docling_ibm_models.layoutmodel.layout_predictor import LayoutPredictor
from PIL import Image from PIL import Image
from docling.datamodel.accelerator_options import AcceleratorOptions from docling.datamodel.accelerator_options import AcceleratorOptions
@ -51,6 +50,8 @@ class LayoutModel(BasePageModel):
def __init__( def __init__(
self, artifacts_path: Optional[Path], accelerator_options: AcceleratorOptions self, artifacts_path: Optional[Path], accelerator_options: AcceleratorOptions
): ):
from docling_ibm_models.layoutmodel.layout_predictor import LayoutPredictor
device = decide_device(accelerator_options.device) device = decide_device(accelerator_options.device)
if artifacts_path is None: if artifacts_path is None:

16
docling/models/plugins/defaults.py
View File

@ -1,13 +1,12 @@
from docling.models.easyocr_model import EasyOcrModel
from docling.models.ocr_mac_model import OcrMacModel
from docling.models.picture_description_api_model import PictureDescriptionApiModel
from docling.models.picture_description_vlm_model import PictureDescriptionVlmModel
from docling.models.rapid_ocr_model import RapidOcrModel
from docling.models.tesseract_ocr_cli_model import TesseractOcrCliModel
from docling.models.tesseract_ocr_model import TesseractOcrModel
def ocr_engines(): def ocr_engines():
from docling.models.easyocr_model import EasyOcrModel
from docling.models.ocr_mac_model import OcrMacModel
from docling.models.rapid_ocr_model import RapidOcrModel
from docling.models.tesseract_ocr_cli_model import TesseractOcrCliModel
from docling.models.tesseract_ocr_model import TesseractOcrModel
return { return {
"ocr_engines": [ "ocr_engines": [
EasyOcrModel, EasyOcrModel,
@ -20,6 +19,9 @@ def ocr_engines():
def picture_description(): def picture_description():
from docling.models.picture_description_api_model import PictureDescriptionApiModel
from docling.models.picture_description_vlm_model import PictureDescriptionVlmModel
return { return {
"picture_description": [ "picture_description": [
PictureDescriptionVlmModel, PictureDescriptionVlmModel,

4
docling/models/table_structure_model.py
View File

@ -10,7 +10,6 @@ from docling_core.types.doc.page import (
BoundingRectangle, BoundingRectangle,
TextCellUnit, TextCellUnit,
) )
from docling_ibm_models.tableformer.data_management.tf_predictor import TFPredictor
from PIL import ImageDraw from PIL import ImageDraw
from docling.datamodel.accelerator_options import AcceleratorDevice, AcceleratorOptions from docling.datamodel.accelerator_options import AcceleratorDevice, AcceleratorOptions
@ -70,6 +69,9 @@ class TableStructureModel(BasePageModel):
# Third Party # Third Party
import docling_ibm_models.tableformer.common as c import docling_ibm_models.tableformer.common as c
from docling_ibm_models.tableformer.data_management.tf_predictor import (
TFPredictor,
)
device = decide_device(accelerator_options.device) device = decide_device(accelerator_options.device)

3
docling/utils/accelerator_utils.py
View File

@ -1,8 +1,6 @@
import logging import logging
from typing import List, Optional from typing import List, Optional
import torch
from docling.datamodel.accelerator_options import AcceleratorDevice from docling.datamodel.accelerator_options import AcceleratorDevice
_log = logging.getLogger(__name__) _log = logging.getLogger(__name__)
@ -18,6 +16,7 @@ def decide_device(
1. AUTO: Check for the best available device on the system. 1. AUTO: Check for the best available device on the system.
2. User-defined: Check if the device actually exists, otherwise fall-back to CPU 2. User-defined: Check if the device actually exists, otherwise fall-back to CPU
""" """
import torch
device = "cpu" device = "cpu"
has_cuda = torch.backends.cuda.is_built() and torch.cuda.is_available() has_cuda = torch.backends.cuda.is_built() and torch.cuda.is_available()